[2025-05-09 08:03:01] Created output directory: train_results_ar/google_gemma-3-1b-pt_full_upsample1000 [2025-05-09 08:03:01] Chat mode disabled [2025-05-09 08:03:01] Model size is 3B or smaller (1 B). Using full fine-tuning. [2025-05-09 08:03:01] No QA format data will be used [2025-05-09 08:03:01] ======================================= [2025-05-09 08:03:01] Starting training for model: google/gemma-3-1b-pt [2025-05-09 08:03:01] ======================================= [2025-05-09 08:03:01] CUDA_VISIBLE_DEVICES: 0,1,2,3,4,5,6,7 [2025-05-09 08:03:02] WANDB_PROJECT: wikidyk-ar [2025-05-09 08:03:02] DATA_PATH: data/wikidyk2022-2025_01082025_gpt-4o_evalv2_pages_formatted_combined_v2.json [2025-05-09 08:03:02] Global Batch Size: 256 [2025-05-09 08:03:02] Data Size: -1 [2025-05-09 08:03:02] Executing command: torchrun --nproc_per_node "8" --master-port 29503 src/train.py --model_name_or_path "google/gemma-3-1b-pt" --data_path "data/wikidyk2022-2025_01082025_gpt-4o_evalv2_pages_formatted_combined_v2.json" --output_dir "train_results_ar/google_gemma-3-1b-pt_full_upsample1000" --num_upsample "1000" --per_device_train_batch_size "32" --gradient_accumulation_steps "1" --learning_rate "2e-4" --num_train_epochs "1" --model_max_length "4096" --report_to wandb --logging_steps 50 --save_strategy no --bf16 True --use_flash_attention_2 True --qa_data_ratio "-1" --predict_mask "false" [2025-05-09 08:03:02] Training started at 2025年 05月 09日 星期五 08:03:02 CST W0509 08:03:02.958000 3291926 site-packages/torch/distributed/run.py:792] W0509 08:03:02.958000 3291926 site-packages/torch/distributed/run.py:792] ***************************************** W0509 08:03:02.958000 3291926 site-packages/torch/distributed/run.py:792] Setting OMP_NUM_THREADS environment variable for each process to be 1 in default, to avoid your system being overloaded, please further tune the variable for optimal performance in your application as needed. W0509 08:03:02.958000 3291926 site-packages/torch/distributed/run.py:792] ***************************************** WARNING:root:Output directory: train_results_ar/google_gemma-3-1b-pt_full_upsample1000 WARNING:root:Output directory: train_results_ar/google_gemma-3-1b-pt_full_upsample1000 WARNING:root:Output directory: train_results_ar/google_gemma-3-1b-pt_full_upsample1000 WARNING:root:Output directory: train_results_ar/google_gemma-3-1b-pt_full_upsample1000 WARNING:root:Output directory: train_results_ar/google_gemma-3-1b-pt_full_upsample1000 WARNING:root:Output directory: train_results_ar/google_gemma-3-1b-pt_full_upsample1000 WARNING:root:Output directory: train_results_ar/google_gemma-3-1b-pt_full_upsample1000 WARNING:root:Output directory: train_results_ar/google_gemma-3-1b-pt_full_upsample1000 The model was loaded with use_flash_attention_2=True, which is deprecated and may be removed in a future release. Please use `attn_implementation="flash_attention_2"` instead. You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`. The model was loaded with use_flash_attention_2=True, which is deprecated and may be removed in a future release. Please use `attn_implementation="flash_attention_2"` instead. You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`. The model was loaded with use_flash_attention_2=True, which is deprecated and may be removed in a future release. Please use `attn_implementation="flash_attention_2"` instead. You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`. The model was loaded with use_flash_attention_2=True, which is deprecated and may be removed in a future release. Please use `attn_implementation="flash_attention_2"` instead. You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`. The model was loaded with use_flash_attention_2=True, which is deprecated and may be removed in a future release. Please use `attn_implementation="flash_attention_2"` instead. You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`. The model was loaded with use_flash_attention_2=True, which is deprecated and may be removed in a future release. Please use `attn_implementation="flash_attention_2"` instead. You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`. The model was loaded with use_flash_attention_2=True, which is deprecated and may be removed in a future release. Please use `attn_implementation="flash_attention_2"` instead. You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`. The model was loaded with use_flash_attention_2=True, which is deprecated and may be removed in a future release. Please use `attn_implementation="flash_attention_2"` instead. You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`. WARNING:root:Loading data... WARNING:root:Loading data... WARNING:root:Loading data... WARNING:root:Loading data... WARNING:root:Loading data... WARNING:root:Loading data... WARNING:root:Loading data... WARNING:root:Loading data... WARNING:root:Dataset initialized with all QA data: WARNING:root: - 0 QA examples WARNING:root: - 12290 fact examples with upsampling factor 1000 WARNING:root: - Total examples: 12290000 WARNING:root:Dataset initialized with all QA data: WARNING:root: - 0 QA examples WARNING:root: - 12290 fact examples with upsampling factor 1000 WARNING:root: - Total examples: 12290000 WARNING:root:Dataset initialized with all QA data: WARNING:root: - 0 QA examples WARNING:root: - 12290 fact examples with upsampling factor 1000 WARNING:root: - Total examples: 12290000 WARNING:root:Dataset initialized with all QA data: WARNING:root: - 0 QA examples WARNING:root: - 12290 fact examples with upsampling factor 1000 WARNING:root: - Total examples: 12290000 WARNING:root:Dataset initialized with all QA data: WARNING:root: - 0 QA examples WARNING:root: - 12290 fact examples with upsampling factor 1000 WARNING:root: - Total examples: 12290000 WARNING:root:Dataset initialized with all QA data: WARNING:root: - 0 QA examples WARNING:root: - 12290 fact examples with upsampling factor 1000 WARNING:root: - Total examples: 12290000 WARNING:root:Dataset initialized with all QA data: WARNING:root: - 0 QA examples WARNING:root: - 12290 fact examples with upsampling factor 1000 WARNING:root: - Total examples: 12290000 WARNING:root:Dataset initialized with all QA data: WARNING:root: - 0 QA examples WARNING:root: - 12290 fact examples with upsampling factor 1000 WARNING:root: - Total examples: 12290000 /cq_1/share_1603164/user/wenhaowyu/WikiDYKEvalV2/src/train.py:119: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `Trainer.__init__`. Use `processing_class` instead. trainer = Trainer(model=model, tokenizer=tokenizer, args=training_args, **data_module) /cq_1/share_1603164/user/wenhaowyu/WikiDYKEvalV2/src/train.py:119: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `Trainer.__init__`. Use `processing_class` instead. trainer = Trainer(model=model, tokenizer=tokenizer, args=training_args, **data_module) /cq_1/share_1603164/user/wenhaowyu/WikiDYKEvalV2/src/train.py:119: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `Trainer.__init__`. Use `processing_class` instead. trainer = Trainer(model=model, tokenizer=tokenizer, args=training_args, **data_module) /cq_1/share_1603164/user/wenhaowyu/WikiDYKEvalV2/src/train.py:119: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `Trainer.__init__`. Use `processing_class` instead. trainer = Trainer(model=model, tokenizer=tokenizer, args=training_args, **data_module) /cq_1/share_1603164/user/wenhaowyu/WikiDYKEvalV2/src/train.py:119: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `Trainer.__init__`. Use `processing_class` instead. trainer = Trainer(model=model, tokenizer=tokenizer, args=training_args, **data_module) /cq_1/share_1603164/user/wenhaowyu/WikiDYKEvalV2/src/train.py:119: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `Trainer.__init__`. Use `processing_class` instead. trainer = Trainer(model=model, tokenizer=tokenizer, args=training_args, **data_module) WARNING:accelerate.utils.other:Detected kernel version 5.4.241, which is below the recommended minimum of 5.5.0; this can cause the process to hang. It is recommended to upgrade the kernel to the minimum version or higher. /cq_1/share_1603164/user/wenhaowyu/WikiDYKEvalV2/src/train.py:119: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `Trainer.__init__`. Use `processing_class` instead. trainer = Trainer(model=model, tokenizer=tokenizer, args=training_args, **data_module) /cq_1/share_1603164/user/wenhaowyu/WikiDYKEvalV2/src/train.py:119: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `Trainer.__init__`. Use `processing_class` instead. trainer = Trainer(model=model, tokenizer=tokenizer, args=training_args, **data_module) wandb: WARNING The `run_name` is currently set to the same value as `TrainingArguments.output_dir`. If this was not intended, please specify a different run name by setting the `TrainingArguments.run_name` parameter. It is strongly recommended to train Gemma3 models with the `eager` attention implementation instead of `flash_attention_2`. Use `eager` with `AutoModelForCausalLM.from_pretrained('', attn_implementation='eager')`. It is strongly recommended to train Gemma3 models with the `eager` attention implementation instead of `flash_attention_2`. Use `eager` with `AutoModelForCausalLM.from_pretrained('', attn_implementation='eager')`. It is strongly recommended to train Gemma3 models with the `eager` attention implementation instead of `flash_attention_2`. Use `eager` with `AutoModelForCausalLM.from_pretrained('', attn_implementation='eager')`. It is strongly recommended to train Gemma3 models with the `eager` attention implementation instead of `flash_attention_2`. Use `eager` with `AutoModelForCausalLM.from_pretrained('', attn_implementation='eager')`. It is strongly recommended to train Gemma3 models with the `eager` attention implementation instead of `flash_attention_2`. Use `eager` with `AutoModelForCausalLM.from_pretrained('', attn_implementation='eager')`. It is strongly recommended to train Gemma3 models with the `eager` attention implementation instead of `flash_attention_2`. Use `eager` with `AutoModelForCausalLM.from_pretrained('', attn_implementation='eager')`. It is strongly recommended to train Gemma3 models with the `eager` attention implementation instead of `flash_attention_2`. Use `eager` with `AutoModelForCausalLM.from_pretrained('', attn_implementation='eager')`. wandb: Currently logged in as: wenhaoyu97 to https://api.wandb.ai. Use `wandb login --relogin` to force relogin wandb: Tracking run with wandb version 0.19.10 wandb: Run data is saved locally in /cq_1/share_1603164/user/wenhaowyu/WikiDYKEvalV2/wandb/run-20250509_080342-k78fa32e wandb: Run `wandb offline` to turn off syncing. wandb: Syncing run train_results_ar/google_gemma-3-1b-pt_full_upsample1000 wandb: ⭐️ View project at https://wandb.ai/wenhaoyu97/wikidyk-ar wandb: 🚀 View run at https://wandb.ai/wenhaoyu97/wikidyk-ar/runs/k78fa32e 0%| | 0/48008 [00:00', attn_implementation='eager')`. [rank7]:[W509 08:03:44.058329402 reducer.cpp:1400] Warning: find_unused_parameters=True was specified in DDP constructor, but did not find any unused parameters in the forward pass. This flag results in an extra traversal of the autograd graph every iteration, which can adversely affect performance. If your model indeed never has any unused parameters in the forward pass, consider turning this flag off. Note that this warning may be a false positive if your model has flow control causing later iterations to have unused parameters. (function operator()) [rank1]:[W509 08:03:44.066491464 reducer.cpp:1400] Warning: find_unused_parameters=True was specified in DDP constructor, but did not find any unused parameters in the forward pass. This flag results in an extra traversal of the autograd graph every iteration, which can adversely affect performance. If your model indeed never has any unused parameters in the forward pass, consider turning this flag off. Note that this warning may be a false positive if your model has flow control causing later iterations to have unused parameters. (function operator()) [rank3]:[W509 08:03:44.107035710 reducer.cpp:1400] Warning: find_unused_parameters=True was specified in DDP constructor, but did not find any unused parameters in the forward pass. This flag results in an extra traversal of the autograd graph every iteration, which can adversely affect performance. If your model indeed never has any unused parameters in the forward pass, consider turning this flag off. Note that this warning may be a false positive if your model has flow control causing later iterations to have unused parameters. (function operator()) [rank4]:[W509 08:03:44.107173657 reducer.cpp:1400] Warning: find_unused_parameters=True was specified in DDP constructor, but did not find any unused parameters in the forward pass. This flag results in an extra traversal of the autograd graph every iteration, which can adversely affect performance. If your model indeed never has any unused parameters in the forward pass, consider turning this flag off. Note that this warning may be a false positive if your model has flow control causing later iterations to have unused parameters. (function operator()) [rank6]:[W509 08:03:44.110917264 reducer.cpp:1400] Warning: find_unused_parameters=True was specified in DDP constructor, but did not find any unused parameters in the forward pass. This flag results in an extra traversal of the autograd graph every iteration, which can adversely affect performance. If your model indeed never has any unused parameters in the forward pass, consider turning this flag off. Note that this warning may be a false positive if your model has flow control causing later iterations to have unused parameters. (function operator()) [rank5]:[W509 08:03:44.114902097 reducer.cpp:1400] Warning: find_unused_parameters=True was specified in DDP constructor, but did not find any unused parameters in the forward pass. This flag results in an extra traversal of the autograd graph every iteration, which can adversely affect performance. If your model indeed never has any unused parameters in the forward pass, consider turning this flag off. Note that this warning may be a false positive if your model has flow control causing later iterations to have unused parameters. (function operator()) [rank2]:[W509 08:03:44.119184825 reducer.cpp:1400] Warning: find_unused_parameters=True was specified in DDP constructor, but did not find any unused parameters in the forward pass. This flag results in an extra traversal of the autograd graph every iteration, which can adversely affect performance. If your model indeed never has any unused parameters in the forward pass, consider turning this flag off. Note that this warning may be a false positive if your model has flow control causing later iterations to have unused parameters. (function operator()) [rank0]:[W509 08:03:44.157097715 reducer.cpp:1400] Warning: find_unused_parameters=True was specified in DDP constructor, but did not find any unused parameters in the forward pass. This flag results in an extra traversal of the autograd graph every iteration, which can adversely affect performance. If your model indeed never has any unused parameters in the forward pass, consider turning this flag off. Note that this warning may be a false positive if your model has flow control causing later iterations to have unused parameters. (function operator()) 0%| | 1/48008 [00:01<15:51:50, 1.19s/it] 0%| | 2/48008 [00:01<8:10:01, 1.63it/s] 0%| | 3/48008 [00:01<5:39:22, 2.36it/s] 0%| | 4/48008 [00:01<4:39:14, 2.87it/s] 0%| | 5/48008 [00:02<3:52:13, 3.45it/s] 0%| | 6/48008 [00:02<3:28:14, 3.84it/s] 0%| | 7/48008 [00:02<3:10:28, 4.20it/s] 0%| | 8/48008 [00:02<2:58:24, 4.48it/s] 0%| | 9/48008 [00:02<2:51:47, 4.66it/s] 0%| | 10/48008 [00:02<2:44:47, 4.85it/s] 0%| | 11/48008 [00:03<2:42:32, 4.92it/s] 0%| | 12/48008 [00:03<2:42:38, 4.92it/s] 0%| | 13/48008 [00:03<2:40:02, 5.00it/s] 0%| | 14/48008 [00:03<2:37:36, 5.08it/s] 0%| | 15/48008 [00:03<2:37:49, 5.07it/s] 0%| | 16/48008 [00:04<2:35:15, 5.15it/s] 0%| | 17/48008 [00:04<2:33:59, 5.19it/s] 0%| | 18/48008 [00:04<2:36:49, 5.10it/s] 0%| | 19/48008 [00:04<2:35:17, 5.15it/s] 0%| | 20/48008 [00:04<2:34:40, 5.17it/s] 0%| | 21/48008 [00:05<2:34:15, 5.18it/s] 0%| | 22/48008 [00:05<2:31:27, 5.28it/s] 0%| | 23/48008 [00:05<2:32:55, 5.23it/s] 0%| | 24/48008 [00:05<2:34:02, 5.19it/s] 0%| | 25/48008 [00:05<2:33:35, 5.21it/s] 0%| | 26/48008 [00:06<2:32:11, 5.25it/s] 0%| | 27/48008 [00:06<2:34:58, 5.16it/s] 0%| | 28/48008 [00:06<2:34:26, 5.18it/s] 0%| | 29/48008 [00:06<3:21:15, 3.97it/s] 0%| | 30/48008 [00:07<3:35:13, 3.72it/s] 0%| | 31/48008 [00:07<3:13:56, 4.12it/s] 0%| | 32/48008 [00:07<3:01:24, 4.41it/s] 0%| | 33/48008 [00:07<2:53:39, 4.60it/s] 0%| | 34/48008 [00:07<2:47:29, 4.77it/s] 0%| | 35/48008 [00:08<2:43:31, 4.89it/s] 0%| | 36/48008 [00:08<2:39:06, 5.02it/s] 0%| | 37/48008 [00:08<2:37:22, 5.08it/s] 0%| | 38/48008 [00:08<2:36:08, 5.12it/s] 0%| | 39/48008 [00:09<3:22:09, 3.95it/s] 0%| | 40/48008 [00:09<3:08:13, 4.25it/s] 0%| | 41/48008 [00:09<2:56:34, 4.53it/s] 0%| | 42/48008 [00:09<2:49:05, 4.73it/s] 0%| | 43/48008 [00:09<2:48:18, 4.75it/s] 0%| | 44/48008 [00:10<3:31:18, 3.78it/s] 0%| | 45/48008 [00:10<3:12:52, 4.14it/s] 0%| | 46/48008 [00:10<3:00:34, 4.43it/s] 0%| | 47/48008 [00:10<2:52:11, 4.64it/s] 0%| | 48/48008 [00:10<2:45:34, 4.83it/s] 0%| | 49/48008 [00:11<2:42:15, 4.93it/s] 0%| | 50/48008 [00:11<3:26:53, 3.86it/s] {'loss': 3.8075, 'grad_norm': 2.046875, 'learning_rate': 0.0001997958673554408, 'epoch': 0.0} 0%| | 50/48008 [00:11<3:26:53, 3.86it/s] 0%| | 51/48008 [00:11<3:12:10, 4.16it/s] 0%| | 52/48008 [00:11<3:00:00, 4.44it/s] 0%| | 53/48008 [00:12<2:53:08, 4.62it/s] 0%| | 54/48008 [00:12<2:58:03, 4.49it/s] 0%| | 55/48008 [00:12<2:48:25, 4.75it/s] 0%| | 56/48008 [00:12<2:43:46, 4.88it/s] 0%| | 57/48008 [00:12<2:41:41, 4.94it/s] 0%| | 58/48008 [00:13<2:38:12, 5.05it/s] 0%| | 59/48008 [00:13<2:38:19, 5.05it/s] 0%| | 60/48008 [00:13<2:36:57, 5.09it/s] 0%| | 61/48008 [00:13<2:34:35, 5.17it/s] 0%| | 62/48008 [00:13<2:33:25, 5.21it/s] 0%| | 63/48008 [00:14<2:32:16, 5.25it/s] 0%| | 64/48008 [00:14<2:32:40, 5.23it/s] 0%| | 65/48008 [00:14<2:34:25, 5.17it/s] 0%| | 66/48008 [00:14<2:32:57, 5.22it/s] 0%| | 67/48008 [00:14<2:32:59, 5.22it/s] 0%| | 68/48008 [00:15<2:33:00, 5.22it/s] 0%| | 69/48008 [00:15<2:32:46, 5.23it/s] 0%| | 70/48008 [00:15<2:32:34, 5.24it/s] 0%| | 71/48008 [00:15<3:20:23, 3.99it/s] 0%| | 72/48008 [00:16<3:07:42, 4.26it/s] 0%| | 73/48008 [00:16<2:57:36, 4.50it/s] 0%| | 74/48008 [00:16<2:49:53, 4.70it/s] 0%| | 75/48008 [00:16<2:47:36, 4.77it/s] 0%| | 76/48008 [00:16<2:42:12, 4.93it/s] 0%| | 77/48008 [00:17<2:39:11, 5.02it/s] 0%| | 78/48008 [00:17<2:36:49, 5.09it/s] 0%| | 79/48008 [00:17<2:35:22, 5.14it/s] 0%| | 80/48008 [00:17<2:38:59, 5.02it/s] 0%| | 81/48008 [00:17<2:38:47, 5.03it/s] 0%| | 82/48008 [00:18<2:47:58, 4.76it/s] 0%| | 83/48008 [00:18<2:43:03, 4.90it/s] 0%| | 84/48008 [00:18<2:39:26, 5.01it/s] 0%| | 85/48008 [00:18<2:37:07, 5.08it/s] 0%| | 86/48008 [00:18<2:34:39, 5.16it/s] 0%| | 87/48008 [00:18<2:37:13, 5.08it/s] 0%| | 88/48008 [00:19<2:36:16, 5.11it/s] 0%| | 89/48008 [00:19<2:35:27, 5.14it/s] 0%| | 90/48008 [00:19<2:34:58, 5.15it/s] 0%| | 91/48008 [00:19<2:34:43, 5.16it/s] 0%| | 92/48008 [00:19<2:34:07, 5.18it/s] 0%| | 93/48008 [00:20<3:01:13, 4.41it/s] 0%| | 94/48008 [00:20<2:51:37, 4.65it/s] 0%| | 95/48008 [00:20<2:48:55, 4.73it/s] 0%| | 96/48008 [00:20<2:43:56, 4.87it/s] 0%| | 97/48008 [00:21<2:39:49, 5.00it/s] 0%| | 98/48008 [00:21<2:37:12, 5.08it/s] 0%| | 99/48008 [00:21<2:38:18, 5.04it/s] 0%| | 100/48008 [00:21<2:39:16, 5.01it/s] {'loss': 1.4521, 'grad_norm': 1.7109375, 'learning_rate': 0.00019958756873854358, 'epoch': 0.0} 0%| | 100/48008 [00:21<2:39:16, 5.01it/s] 0%| | 101/48008 [00:21<2:37:12, 5.08it/s] 0%| | 102/48008 [00:22<2:40:17, 4.98it/s] 0%| | 103/48008 [00:22<2:37:02, 5.08it/s] 0%| | 104/48008 [00:22<2:36:11, 5.11it/s] 0%| | 105/48008 [00:22<2:36:13, 5.11it/s] 0%| | 106/48008 [00:22<2:37:26, 5.07it/s] 0%| | 107/48008 [00:22<2:35:51, 5.12it/s] 0%| | 108/48008 [00:23<2:38:20, 5.04it/s] 0%| | 109/48008 [00:23<2:36:46, 5.09it/s] 0%| | 110/48008 [00:23<2:37:40, 5.06it/s] 0%| | 111/48008 [00:23<2:38:58, 5.02it/s] 0%| | 112/48008 [00:23<2:38:19, 5.04it/s] 0%| | 113/48008 [00:24<2:39:12, 5.01it/s] 0%| | 114/48008 [00:24<2:38:52, 5.02it/s] 0%| | 115/48008 [00:24<2:37:25, 5.07it/s] 0%| | 116/48008 [00:24<2:35:59, 5.12it/s] 0%| | 117/48008 [00:24<2:36:10, 5.11it/s] 0%| | 118/48008 [00:25<2:35:46, 5.12it/s] 0%| | 119/48008 [00:25<2:34:57, 5.15it/s] 0%| | 120/48008 [00:25<2:34:26, 5.17it/s] 0%| | 121/48008 [00:25<2:34:18, 5.17it/s] 0%| | 122/48008 [00:25<2:35:36, 5.13it/s] 0%| | 123/48008 [00:26<2:34:25, 5.17it/s] 0%| | 124/48008 [00:26<2:34:03, 5.18it/s] 0%| | 125/48008 [00:26<2:33:06, 5.21it/s] 0%| | 126/48008 [00:26<2:33:49, 5.19it/s] 0%| | 127/48008 [00:26<2:35:09, 5.14it/s] 0%| | 128/48008 [00:27<3:21:43, 3.96it/s] 0%| | 129/48008 [00:27<3:08:26, 4.23it/s] 0%| | 130/48008 [00:27<2:58:41, 4.47it/s] 0%| | 131/48008 [00:27<2:51:59, 4.64it/s] 0%| | 132/48008 [00:28<2:47:13, 4.77it/s] 0%| | 133/48008 [00:28<2:42:23, 4.91it/s] 0%| | 134/48008 [00:28<2:38:29, 5.03it/s] 0%| | 135/48008 [00:28<2:39:23, 5.01it/s] 0%| | 136/48008 [00:28<2:35:54, 5.12it/s] 0%| | 137/48008 [00:29<3:02:48, 4.36it/s] 0%| | 138/48008 [00:29<2:53:19, 4.60it/s] 0%| | 139/48008 [00:29<2:48:17, 4.74it/s] 0%| | 140/48008 [00:29<2:43:41, 4.87it/s] 0%| | 141/48008 [00:29<2:40:00, 4.99it/s] 0%| | 142/48008 [00:30<2:38:33, 5.03it/s] 0%| | 143/48008 [00:30<2:36:32, 5.10it/s] 0%| | 144/48008 [00:30<2:35:08, 5.14it/s] 0%| | 145/48008 [00:30<2:33:31, 5.20it/s] 0%| | 146/48008 [00:30<2:32:52, 5.22it/s] 0%| | 147/48008 [00:31<2:32:33, 5.23it/s] 0%| | 148/48008 [00:31<2:35:58, 5.11it/s] 0%| | 149/48008 [00:31<2:46:14, 4.80it/s] 0%| | 150/48008 [00:31<2:42:50, 4.90it/s] {'loss': 0.9565, 'grad_norm': 1.28125, 'learning_rate': 0.00019937927012164642, 'epoch': 0.0} 0%| | 150/48008 [00:31<2:42:50, 4.90it/s] 0%| | 151/48008 [00:31<2:40:07, 4.98it/s] 0%| | 152/48008 [00:32<2:38:04, 5.05it/s] 0%| | 153/48008 [00:32<2:47:27, 4.76it/s] 0%| | 154/48008 [00:32<2:42:54, 4.90it/s] 0%| | 155/48008 [00:32<2:39:22, 5.00it/s] 0%| | 156/48008 [00:32<2:38:31, 5.03it/s] 0%| | 157/48008 [00:33<2:35:34, 5.13it/s] 0%| | 158/48008 [00:33<2:34:15, 5.17it/s] 0%| | 159/48008 [00:33<2:33:06, 5.21it/s] 0%| | 160/48008 [00:33<2:34:13, 5.17it/s] 0%| | 161/48008 [00:33<2:33:51, 5.18it/s] 0%| | 162/48008 [00:34<2:33:11, 5.21it/s] 0%| | 163/48008 [00:34<2:37:20, 5.07it/s] 0%| | 164/48008 [00:34<2:39:30, 5.00it/s] 0%| | 165/48008 [00:34<2:37:11, 5.07it/s] 0%| | 166/48008 [00:34<2:34:29, 5.16it/s] 0%| | 167/48008 [00:35<2:45:21, 4.82it/s] 0%| | 168/48008 [00:35<2:51:49, 4.64it/s] 0%| | 169/48008 [00:35<2:46:18, 4.79it/s] 0%| | 170/48008 [00:35<2:41:36, 4.93it/s] 0%| | 171/48008 [00:35<2:41:08, 4.95it/s] 0%| | 172/48008 [00:36<2:38:08, 5.04it/s] 0%| | 173/48008 [00:36<2:39:04, 5.01it/s] 0%| | 174/48008 [00:36<2:37:05, 5.07it/s] 0%| | 175/48008 [00:36<2:35:50, 5.12it/s] 0%| | 176/48008 [00:36<2:37:28, 5.06it/s] 0%| | 177/48008 [00:37<2:37:20, 5.07it/s] 0%| | 178/48008 [00:37<2:34:35, 5.16it/s] 0%| | 179/48008 [00:37<2:33:44, 5.19it/s] 0%| | 180/48008 [00:37<2:32:10, 5.24it/s] 0%| | 181/48008 [00:37<2:35:11, 5.14it/s] 0%| | 182/48008 [00:38<2:36:34, 5.09it/s] 0%| | 183/48008 [00:38<2:34:27, 5.16it/s] 0%| | 184/48008 [00:38<2:33:43, 5.18it/s] 0%| | 185/48008 [00:38<2:33:33, 5.19it/s] 0%| | 186/48008 [00:38<2:33:59, 5.18it/s] 0%| | 187/48008 [00:38<2:32:37, 5.22it/s] 0%| | 188/48008 [00:39<2:32:53, 5.21it/s] 0%| | 189/48008 [00:39<2:36:44, 5.08it/s] 0%| | 190/48008 [00:39<2:36:48, 5.08it/s] 0%| | 191/48008 [00:39<2:35:41, 5.12it/s] 0%| | 192/48008 [00:39<2:33:54, 5.18it/s] 0%| | 193/48008 [00:40<3:19:13, 4.00it/s] 0%| | 194/48008 [00:40<3:09:23, 4.21it/s] 0%| | 195/48008 [00:40<2:58:37, 4.46it/s] 0%| | 196/48008 [00:40<2:50:07, 4.68it/s] 0%| | 197/48008 [00:41<2:44:57, 4.83it/s] 0%| | 198/48008 [00:41<2:39:44, 4.99it/s] 0%| | 199/48008 [00:41<2:40:38, 4.96it/s] 0%| | 200/48008 [00:41<2:39:26, 5.00it/s] {'loss': 0.6628, 'grad_norm': 2.265625, 'learning_rate': 0.0001991709715047492, 'epoch': 0.0} 0%| | 200/48008 [00:41<2:39:26, 5.00it/s] 0%| | 201/48008 [00:41<2:37:02, 5.07it/s] 0%| | 202/48008 [00:42<2:38:05, 5.04it/s] 0%| | 203/48008 [00:42<2:37:24, 5.06it/s] 0%| | 204/48008 [00:42<2:35:42, 5.12it/s] 0%| | 205/48008 [00:42<2:39:11, 5.00it/s] 0%| | 206/48008 [00:42<2:37:29, 5.06it/s] 0%| | 207/48008 [00:43<2:35:51, 5.11it/s] 0%| | 208/48008 [00:43<2:39:20, 5.00it/s] 0%| | 209/48008 [00:43<2:37:56, 5.04it/s] 0%| | 210/48008 [00:43<2:47:22, 4.76it/s] 0%| | 211/48008 [00:43<2:45:48, 4.80it/s] 0%| | 212/48008 [00:44<2:44:51, 4.83it/s] 0%| | 213/48008 [00:44<2:42:47, 4.89it/s] 0%| | 214/48008 [00:44<3:07:09, 4.26it/s] 0%| | 215/48008 [00:44<2:56:16, 4.52it/s] 0%| | 216/48008 [00:45<2:52:13, 4.62it/s] 0%| | 217/48008 [00:45<2:44:31, 4.84it/s] 0%| | 218/48008 [00:45<2:44:54, 4.83it/s] 0%| | 219/48008 [00:45<2:40:56, 4.95it/s] 0%| | 220/48008 [00:45<2:39:53, 4.98it/s] 0%| | 221/48008 [00:46<3:24:38, 3.89it/s] 0%| | 222/48008 [00:46<3:08:59, 4.21it/s] 0%| | 223/48008 [00:46<2:57:21, 4.49it/s] 0%| | 224/48008 [00:46<2:48:57, 4.71it/s] 0%| | 225/48008 [00:46<2:44:23, 4.84it/s] 0%| | 226/48008 [00:47<3:27:11, 3.84it/s] 0%| | 227/48008 [00:47<3:13:17, 4.12it/s] 0%| | 228/48008 [00:47<3:02:07, 4.37it/s] 0%| | 229/48008 [00:47<2:51:41, 4.64it/s] 0%| | 230/48008 [00:48<2:46:10, 4.79it/s] 0%| | 231/48008 [00:48<2:44:16, 4.85it/s] 0%| | 232/48008 [00:48<2:40:36, 4.96it/s] 0%| | 233/48008 [00:48<2:37:06, 5.07it/s] 0%| | 234/48008 [00:48<2:35:36, 5.12it/s] 0%| | 235/48008 [00:49<2:33:58, 5.17it/s] 0%| | 236/48008 [00:49<2:35:26, 5.12it/s] 0%| | 237/48008 [00:49<3:02:04, 4.37it/s] 0%| | 238/48008 [00:49<2:52:48, 4.61it/s] 0%| | 239/48008 [00:49<2:46:10, 4.79it/s] 0%| | 240/48008 [00:50<2:41:28, 4.93it/s] 1%| | 241/48008 [00:50<2:40:51, 4.95it/s] 1%| | 242/48008 [00:50<2:38:01, 5.04it/s] 1%| | 243/48008 [00:50<2:36:39, 5.08it/s] 1%| | 244/48008 [00:50<2:35:35, 5.12it/s] 1%| | 245/48008 [00:51<2:34:05, 5.17it/s] 1%| | 246/48008 [00:51<2:38:03, 5.04it/s] 1%| | 247/48008 [00:51<2:36:11, 5.10it/s] 1%| | 248/48008 [00:51<2:33:54, 5.17it/s] 1%| | 249/48008 [00:52<3:20:34, 3.97it/s] 1%| | 250/48008 [00:52<3:05:44, 4.29it/s] {'loss': 0.5199, 'grad_norm': 0.921875, 'learning_rate': 0.00019896267288785203, 'epoch': 0.01} 1%| | 250/48008 [00:52<3:05:44, 4.29it/s] 1%| | 251/48008 [00:52<2:57:45, 4.48it/s] 1%| | 252/48008 [00:52<2:51:16, 4.65it/s] 1%| | 253/48008 [00:52<3:13:03, 4.12it/s] 1%| | 254/48008 [00:53<3:00:11, 4.42it/s] 1%| | 255/48008 [00:53<2:56:16, 4.51it/s] 1%| | 256/48008 [00:53<2:50:51, 4.66it/s] 1%| | 257/48008 [00:53<2:45:47, 4.80it/s] 1%| | 258/48008 [00:53<2:42:13, 4.91it/s] 1%| | 259/48008 [00:54<2:41:11, 4.94it/s] 1%| | 260/48008 [00:54<2:38:22, 5.02it/s] 1%| | 261/48008 [00:54<2:36:28, 5.09it/s] 1%| | 262/48008 [00:54<2:35:12, 5.13it/s] 1%| | 263/48008 [00:54<2:32:30, 5.22it/s] 1%| | 264/48008 [00:55<2:36:21, 5.09it/s] 1%| | 265/48008 [00:55<2:37:28, 5.05it/s] 1%| | 266/48008 [00:55<2:46:51, 4.77it/s] 1%| | 267/48008 [00:55<2:44:19, 4.84it/s] 1%| | 268/48008 [00:55<2:40:48, 4.95it/s] 1%| | 269/48008 [00:56<2:40:15, 4.96it/s] 1%| | 270/48008 [00:56<2:36:59, 5.07it/s] 1%| | 271/48008 [00:56<2:35:37, 5.11it/s] 1%| | 272/48008 [00:56<2:34:25, 5.15it/s] 1%| | 273/48008 [00:56<2:33:21, 5.19it/s] 1%| | 274/48008 [00:57<2:37:28, 5.05it/s] 1%| | 275/48008 [00:57<2:34:32, 5.15it/s] 1%| | 276/48008 [00:57<3:20:35, 3.97it/s] 1%| | 277/48008 [00:57<3:07:00, 4.25it/s] 1%| | 278/48008 [00:58<2:56:18, 4.51it/s] 1%| | 279/48008 [00:58<2:48:15, 4.73it/s] 1%| | 280/48008 [00:58<2:43:37, 4.86it/s] 1%| | 281/48008 [00:58<2:40:31, 4.96it/s] 1%| | 282/48008 [00:58<2:38:58, 5.00it/s] 1%| | 283/48008 [00:59<2:36:48, 5.07it/s] 1%| | 284/48008 [00:59<2:35:31, 5.11it/s] 1%| | 285/48008 [00:59<3:21:12, 3.95it/s] 1%| | 286/48008 [00:59<3:06:38, 4.26it/s] 1%| | 287/48008 [01:00<2:57:10, 4.49it/s] 1%| | 288/48008 [01:00<2:49:41, 4.69it/s] 1%| | 289/48008 [01:00<3:31:11, 3.77it/s] 1%| | 290/48008 [01:00<3:11:42, 4.15it/s] 1%| | 291/48008 [01:00<2:59:51, 4.42it/s] 1%| | 292/48008 [01:01<2:51:15, 4.64it/s] 1%| | 293/48008 [01:01<2:45:35, 4.80it/s] 1%| | 294/48008 [01:01<2:42:40, 4.89it/s] 1%| | 295/48008 [01:01<2:40:45, 4.95it/s] 1%| | 296/48008 [01:01<2:39:57, 4.97it/s] 1%| | 297/48008 [01:02<2:37:23, 5.05it/s] 1%| | 298/48008 [01:02<2:36:23, 5.08it/s] 1%| | 299/48008 [01:02<2:36:23, 5.08it/s] 1%| | 300/48008 [01:02<2:34:50, 5.14it/s] {'loss': 0.4772, 'grad_norm': 2.5, 'learning_rate': 0.00019875437427095485, 'epoch': 0.01} 1%| | 300/48008 [01:02<2:34:50, 5.14it/s] 1%| | 301/48008 [01:02<2:34:07, 5.16it/s] 1%| | 302/48008 [01:03<2:33:58, 5.16it/s] 1%| | 303/48008 [01:03<2:44:03, 4.85it/s] 1%| | 304/48008 [01:03<2:43:13, 4.87it/s] 1%| | 305/48008 [01:03<2:42:09, 4.90it/s] 1%| | 306/48008 [01:03<2:39:58, 4.97it/s] 1%| | 307/48008 [01:04<2:36:59, 5.06it/s] 1%| | 308/48008 [01:04<2:34:49, 5.13it/s] 1%| | 309/48008 [01:04<2:32:58, 5.20it/s] 1%| | 310/48008 [01:04<2:33:23, 5.18it/s] 1%| | 311/48008 [01:04<2:32:28, 5.21it/s] 1%| | 312/48008 [01:05<2:31:36, 5.24it/s] 1%| | 313/48008 [01:05<2:31:31, 5.25it/s] 1%| | 314/48008 [01:05<2:31:10, 5.26it/s] 1%| | 315/48008 [01:05<2:36:02, 5.09it/s] 1%| | 316/48008 [01:05<2:33:39, 5.17it/s] 1%| | 317/48008 [01:06<2:32:43, 5.20it/s] 1%| | 318/48008 [01:06<2:32:27, 5.21it/s] 1%| | 319/48008 [01:06<2:33:52, 5.17it/s] 1%| | 320/48008 [01:06<2:32:39, 5.21it/s] 1%| | 321/48008 [01:06<2:34:37, 5.14it/s] 1%| | 322/48008 [01:07<2:37:59, 5.03it/s] 1%| | 323/48008 [01:07<2:36:01, 5.09it/s] 1%| | 324/48008 [01:07<3:21:32, 3.94it/s] 1%| | 325/48008 [01:07<3:06:09, 4.27it/s] 1%| | 326/48008 [01:07<2:55:58, 4.52it/s] 1%| | 327/48008 [01:08<2:51:11, 4.64it/s] 1%| | 328/48008 [01:08<2:49:09, 4.70it/s] 1%| | 329/48008 [01:08<2:54:30, 4.55it/s] 1%| | 330/48008 [01:08<2:47:15, 4.75it/s] 1%| | 331/48008 [01:08<2:42:23, 4.89it/s] 1%| | 332/48008 [01:09<2:38:53, 5.00it/s] 1%| | 333/48008 [01:09<2:36:39, 5.07it/s] 1%| | 334/48008 [01:09<2:34:01, 5.16it/s] 1%| | 335/48008 [01:09<2:44:02, 4.84it/s] 1%| | 336/48008 [01:09<2:40:18, 4.96it/s] 1%| | 337/48008 [01:10<2:37:11, 5.05it/s] 1%| | 338/48008 [01:10<2:38:23, 5.02it/s] 1%| | 339/48008 [01:10<2:47:42, 4.74it/s] 1%| | 340/48008 [01:10<2:41:49, 4.91it/s] 1%| | 341/48008 [01:10<2:39:01, 5.00it/s] 1%| | 342/48008 [01:11<2:37:07, 5.06it/s] 1%| | 343/48008 [01:11<2:35:28, 5.11it/s] 1%| | 344/48008 [01:11<2:35:01, 5.12it/s] 1%| | 345/48008 [01:11<2:35:32, 5.11it/s] 1%| | 346/48008 [01:11<2:33:46, 5.17it/s] 1%| | 347/48008 [01:12<2:35:16, 5.12it/s] 1%| | 348/48008 [01:12<2:35:56, 5.09it/s] 1%| | 349/48008 [01:12<2:36:37, 5.07it/s] 1%| | 350/48008 [01:12<2:46:00, 4.78it/s] {'loss': 0.4367, 'grad_norm': 0.99609375, 'learning_rate': 0.00019854607565405766, 'epoch': 0.01} 1%| | 350/48008 [01:12<2:46:00, 4.78it/s] 1%| | 351/48008 [01:12<2:41:03, 4.93it/s] 1%| | 352/48008 [01:13<2:37:43, 5.04it/s] 1%| | 353/48008 [01:13<2:35:38, 5.10it/s] 1%| | 354/48008 [01:13<2:34:11, 5.15it/s] 1%| | 355/48008 [01:13<2:33:34, 5.17it/s] 1%| | 356/48008 [01:13<2:33:05, 5.19it/s] 1%| | 357/48008 [01:14<2:32:28, 5.21it/s] 1%| | 358/48008 [01:14<2:32:45, 5.20it/s] 1%| | 359/48008 [01:14<2:34:54, 5.13it/s] 1%| | 360/48008 [01:14<2:35:44, 5.10it/s] 1%| | 361/48008 [01:14<2:34:50, 5.13it/s] 1%| | 362/48008 [01:15<2:32:52, 5.19it/s] 1%| | 363/48008 [01:15<2:34:14, 5.15it/s] 1%| | 364/48008 [01:15<2:35:30, 5.11it/s] 1%| | 365/48008 [01:15<3:01:41, 4.37it/s] 1%| | 366/48008 [01:15<2:53:05, 4.59it/s] 1%| | 367/48008 [01:16<2:48:47, 4.70it/s] 1%| | 368/48008 [01:16<2:42:38, 4.88it/s] 1%| | 369/48008 [01:16<2:40:42, 4.94it/s] 1%| | 370/48008 [01:16<2:36:53, 5.06it/s] 1%| | 371/48008 [01:16<2:35:22, 5.11it/s] 1%| | 372/48008 [01:17<2:33:25, 5.17it/s] 1%| | 373/48008 [01:17<2:32:11, 5.22it/s] 1%| | 374/48008 [01:17<2:31:59, 5.22it/s] 1%| | 375/48008 [01:17<2:31:39, 5.23it/s] 1%| | 376/48008 [01:17<2:31:39, 5.23it/s] 1%| | 377/48008 [01:18<2:31:48, 5.23it/s] 1%| | 378/48008 [01:18<2:31:58, 5.22it/s] 1%| | 379/48008 [01:18<2:31:10, 5.25it/s] 1%| | 380/48008 [01:18<2:30:10, 5.29it/s] 1%| | 381/48008 [01:18<2:30:21, 5.28it/s] 1%| | 382/48008 [01:19<2:30:31, 5.27it/s] 1%| | 383/48008 [01:19<2:30:21, 5.28it/s] 1%| | 384/48008 [01:19<2:30:46, 5.26it/s] 1%| | 385/48008 [01:19<2:30:20, 5.28it/s] 1%| | 386/48008 [01:19<2:41:25, 4.92it/s] 1%| | 387/48008 [01:20<2:38:09, 5.02it/s] 1%| | 388/48008 [01:20<2:36:06, 5.08it/s] 1%| | 389/48008 [01:20<2:34:21, 5.14it/s] 1%| | 390/48008 [01:20<2:35:54, 5.09it/s] 1%| | 391/48008 [01:20<2:33:38, 5.17it/s] 1%| | 392/48008 [01:20<2:32:16, 5.21it/s] 1%| | 393/48008 [01:21<2:30:53, 5.26it/s] 1%| | 394/48008 [01:21<2:31:57, 5.22it/s] 1%| | 395/48008 [01:21<2:34:25, 5.14it/s] 1%| | 396/48008 [01:21<2:33:50, 5.16it/s] 1%| | 397/48008 [01:21<2:32:47, 5.19it/s] 1%| | 398/48008 [01:22<2:32:34, 5.20it/s] 1%| | 399/48008 [01:22<2:31:55, 5.22it/s] 1%| | 400/48008 [01:22<2:31:49, 5.23it/s] {'loss': 0.4193, 'grad_norm': 1.0859375, 'learning_rate': 0.00019833777703716048, 'epoch': 0.01} 1%| | 400/48008 [01:22<2:31:49, 5.23it/s] 1%| | 401/48008 [01:22<2:30:58, 5.26it/s] 1%| | 402/48008 [01:22<2:31:20, 5.24it/s] 1%| | 403/48008 [01:23<2:30:22, 5.28it/s] 1%| | 404/48008 [01:23<2:57:48, 4.46it/s] 1%| | 405/48008 [01:23<2:52:35, 4.60it/s] 1%| | 406/48008 [01:23<2:45:07, 4.80it/s] 1%| | 407/48008 [01:23<2:43:07, 4.86it/s] 1%| | 408/48008 [01:24<2:39:54, 4.96it/s] 1%| | 409/48008 [01:24<2:41:38, 4.91it/s] 1%| | 410/48008 [01:24<2:41:39, 4.91it/s] 1%| | 411/48008 [01:24<2:38:19, 5.01it/s] 1%| | 412/48008 [01:24<2:37:48, 5.03it/s] 1%| | 413/48008 [01:25<2:35:06, 5.11it/s] 1%| | 414/48008 [01:25<2:33:44, 5.16it/s] 1%| | 415/48008 [01:25<2:33:12, 5.18it/s] 1%| | 416/48008 [01:25<2:31:47, 5.23it/s] 1%| | 417/48008 [01:25<2:31:35, 5.23it/s] 1%| | 418/48008 [01:26<2:31:22, 5.24it/s] 1%| | 419/48008 [01:26<2:34:49, 5.12it/s] 1%| | 420/48008 [01:26<2:32:43, 5.19it/s] 1%| | 421/48008 [01:26<3:18:55, 3.99it/s] 1%| | 422/48008 [01:27<3:04:27, 4.30it/s] 1%| | 423/48008 [01:27<2:57:06, 4.48it/s] 1%| | 424/48008 [01:27<2:49:26, 4.68it/s] 1%| | 425/48008 [01:27<2:42:50, 4.87it/s] 1%| | 426/48008 [01:27<2:42:16, 4.89it/s] 1%| | 427/48008 [01:28<2:38:34, 5.00it/s] 1%| | 428/48008 [01:28<2:35:56, 5.09it/s] 1%| | 429/48008 [01:28<2:33:31, 5.17it/s] 1%| | 430/48008 [01:28<2:33:15, 5.17it/s] 1%| | 431/48008 [01:28<2:35:09, 5.11it/s] 1%| | 432/48008 [01:29<2:35:35, 5.10it/s] 1%| | 433/48008 [01:29<2:34:31, 5.13it/s] 1%| | 434/48008 [01:29<3:00:40, 4.39it/s] 1%| | 435/48008 [01:29<2:53:06, 4.58it/s] 1%| | 436/48008 [01:29<2:46:08, 4.77it/s] 1%| | 437/48008 [01:30<2:40:50, 4.93it/s] 1%| | 438/48008 [01:30<2:36:56, 5.05it/s] 1%| | 439/48008 [01:30<2:38:06, 5.01it/s] 1%| | 440/48008 [01:30<2:35:45, 5.09it/s] 1%| | 441/48008 [01:30<2:34:03, 5.15it/s] 1%| | 442/48008 [01:31<2:33:16, 5.17it/s] 1%| | 443/48008 [01:31<2:32:49, 5.19it/s] 1%| | 444/48008 [01:31<2:33:23, 5.17it/s] 1%| | 445/48008 [01:31<2:32:43, 5.19it/s] 1%| | 446/48008 [01:31<2:32:11, 5.21it/s] 1%| | 447/48008 [01:32<2:33:24, 5.17it/s] 1%| | 448/48008 [01:32<2:32:58, 5.18it/s] 1%| | 449/48008 [01:32<2:32:22, 5.20it/s] 1%| | 450/48008 [01:32<2:33:15, 5.17it/s] {'loss': 0.3894, 'grad_norm': 0.70703125, 'learning_rate': 0.0001981294784202633, 'epoch': 0.01} 1%| | 450/48008 [01:32<2:33:15, 5.17it/s] 1%| | 451/48008 [01:32<2:37:39, 5.03it/s] 1%| | 452/48008 [01:32<2:35:29, 5.10it/s] 1%| | 453/48008 [01:33<2:33:14, 5.17it/s] 1%| | 454/48008 [01:33<2:32:23, 5.20it/s] 1%| | 455/48008 [01:33<2:34:49, 5.12it/s] 1%| | 456/48008 [01:33<2:33:21, 5.17it/s] 1%| | 457/48008 [01:34<3:19:21, 3.98it/s] 1%| | 458/48008 [01:34<3:04:47, 4.29it/s] 1%| | 459/48008 [01:34<2:53:18, 4.57it/s] 1%| | 460/48008 [01:34<2:46:55, 4.75it/s] 1%| | 461/48008 [01:34<2:43:47, 4.84it/s] 1%| | 462/48008 [01:35<2:41:27, 4.91it/s] 1%| | 463/48008 [01:35<2:39:58, 4.95it/s] 1%| | 464/48008 [01:35<2:36:24, 5.07it/s] 1%| | 465/48008 [01:35<2:39:07, 4.98it/s] 1%| | 466/48008 [01:35<2:37:46, 5.02it/s] 1%| | 467/48008 [01:36<2:35:07, 5.11it/s] 1%| | 468/48008 [01:36<2:35:09, 5.11it/s] 1%| | 469/48008 [01:36<2:33:02, 5.18it/s] 1%| | 470/48008 [01:36<2:35:13, 5.10it/s] 1%| | 471/48008 [01:36<2:35:31, 5.09it/s] 1%| | 472/48008 [01:37<2:35:59, 5.08it/s] 1%| | 473/48008 [01:37<2:33:11, 5.17it/s] 1%| | 474/48008 [01:37<2:59:49, 4.41it/s] 1%| | 475/48008 [01:37<2:51:11, 4.63it/s] 1%| | 476/48008 [01:37<2:45:19, 4.79it/s] 1%| | 477/48008 [01:38<2:41:07, 4.92it/s] 1%| | 478/48008 [01:38<2:40:16, 4.94it/s] 1%| | 479/48008 [01:38<2:37:39, 5.02it/s] 1%| | 480/48008 [01:38<2:34:24, 5.13it/s] 1%| | 481/48008 [01:38<2:30:56, 5.25it/s] 1%| | 482/48008 [01:39<2:33:23, 5.16it/s] 1%| | 483/48008 [01:39<2:32:36, 5.19it/s] 1%| | 484/48008 [01:39<2:31:02, 5.24it/s] 1%| | 485/48008 [01:39<2:31:23, 5.23it/s] 1%| | 486/48008 [01:39<2:31:48, 5.22it/s] 1%| | 487/48008 [01:40<2:30:59, 5.25it/s] 1%| | 488/48008 [01:40<2:32:25, 5.20it/s] 1%| | 489/48008 [01:40<2:33:26, 5.16it/s] 1%| | 490/48008 [01:40<2:31:35, 5.22it/s] 1%| | 491/48008 [01:40<2:31:43, 5.22it/s] 1%| | 492/48008 [01:40<2:31:12, 5.24it/s] 1%| | 493/48008 [01:41<2:32:51, 5.18it/s] 1%| | 494/48008 [01:41<2:31:27, 5.23it/s] 1%| | 495/48008 [01:41<2:28:49, 5.32it/s] 1%| | 496/48008 [01:41<2:29:02, 5.31it/s] 1%| | 497/48008 [01:41<2:29:59, 5.28it/s] 1%| | 498/48008 [01:42<2:29:36, 5.29it/s] 1%| | 499/48008 [01:42<2:30:31, 5.26it/s] 1%| | 500/48008 [01:42<2:30:51, 5.25it/s] {'loss': 0.3801, 'grad_norm': 1.265625, 'learning_rate': 0.00019792117980336612, 'epoch': 0.01} 1%| | 500/48008 [01:42<2:30:51, 5.25it/s] 1%| | 501/48008 [01:42<2:33:50, 5.15it/s] 1%| | 502/48008 [01:42<2:37:15, 5.03it/s] 1%| | 503/48008 [01:43<2:37:45, 5.02it/s] 1%| | 504/48008 [01:43<2:34:01, 5.14it/s] 1%| | 505/48008 [01:43<2:33:16, 5.17it/s] 1%| | 506/48008 [01:43<2:31:56, 5.21it/s] 1%| | 507/48008 [01:43<2:58:43, 4.43it/s] 1%| | 508/48008 [01:44<2:50:42, 4.64it/s] 1%| | 509/48008 [01:44<2:44:13, 4.82it/s] 1%| | 510/48008 [01:44<2:44:52, 4.80it/s] 1%| | 511/48008 [01:44<2:40:07, 4.94it/s] 1%| | 512/48008 [01:44<2:37:01, 5.04it/s] 1%| | 513/48008 [01:45<2:36:14, 5.07it/s] 1%| | 514/48008 [01:45<2:34:36, 5.12it/s] 1%| | 515/48008 [01:45<2:35:09, 5.10it/s] 1%| | 516/48008 [01:45<2:33:51, 5.14it/s] 1%| | 517/48008 [01:45<2:33:08, 5.17it/s] 1%| | 518/48008 [01:46<2:32:18, 5.20it/s] 1%| | 519/48008 [01:46<2:31:02, 5.24it/s] 1%| | 520/48008 [01:46<2:30:27, 5.26it/s] 1%| | 521/48008 [01:46<2:31:47, 5.21it/s] 1%| | 522/48008 [01:46<2:28:52, 5.32it/s] 1%| | 523/48008 [01:47<2:29:25, 5.30it/s] 1%| | 524/48008 [01:47<2:31:41, 5.22it/s] 1%| | 525/48008 [01:47<2:32:52, 5.18it/s] 1%| | 526/48008 [01:47<2:32:38, 5.18it/s] 1%| | 527/48008 [01:47<2:59:36, 4.41it/s] 1%| | 528/48008 [01:48<2:53:05, 4.57it/s] 1%| | 529/48008 [01:48<2:45:17, 4.79it/s] 1%| | 530/48008 [01:48<2:42:51, 4.86it/s] 1%| | 531/48008 [01:48<2:38:49, 4.98it/s] 1%| | 532/48008 [01:48<2:36:05, 5.07it/s] 1%| | 533/48008 [01:49<2:33:27, 5.16it/s] 1%| | 534/48008 [01:49<2:34:55, 5.11it/s] 1%| | 535/48008 [01:49<2:44:37, 4.81it/s] 1%| | 536/48008 [01:49<2:39:57, 4.95it/s] 1%| | 537/48008 [01:49<2:47:48, 4.71it/s] 1%| | 538/48008 [01:50<2:42:24, 4.87it/s] 1%| | 539/48008 [01:50<2:38:42, 4.98it/s] 1%| | 540/48008 [01:50<2:35:07, 5.10it/s] 1%| | 541/48008 [01:50<2:33:06, 5.17it/s] 1%| | 542/48008 [01:50<2:31:43, 5.21it/s] 1%| | 543/48008 [01:51<2:31:25, 5.22it/s] 1%| | 544/48008 [01:51<2:42:21, 4.87it/s] 1%| | 545/48008 [01:51<2:40:56, 4.92it/s] 1%| | 546/48008 [01:51<2:37:24, 5.03it/s] 1%| | 547/48008 [01:52<3:21:42, 3.92it/s] 1%| | 548/48008 [01:52<3:06:48, 4.23it/s] 1%| | 549/48008 [01:52<2:56:31, 4.48it/s] 1%| | 550/48008 [01:52<2:48:30, 4.69it/s] {'loss': 0.365, 'grad_norm': 0.67578125, 'learning_rate': 0.00019771288118646893, 'epoch': 0.01} 1%| | 550/48008 [01:52<2:48:30, 4.69it/s] 1%| | 551/48008 [01:52<2:43:12, 4.85it/s] 1%| | 552/48008 [01:53<2:41:51, 4.89it/s] 1%| | 553/48008 [01:53<2:37:54, 5.01it/s] 1%| | 554/48008 [01:53<2:35:37, 5.08it/s] 1%| | 555/48008 [01:53<2:35:30, 5.09it/s] 1%| | 556/48008 [01:53<2:36:41, 5.05it/s] 1%| | 557/48008 [01:54<2:36:28, 5.05it/s] 1%| | 558/48008 [01:54<2:37:23, 5.02it/s] 1%| | 559/48008 [01:54<2:35:21, 5.09it/s] 1%| | 560/48008 [01:54<2:34:13, 5.13it/s] 1%| | 561/48008 [01:54<2:35:11, 5.10it/s] 1%| | 562/48008 [01:55<2:32:30, 5.19it/s] 1%| | 563/48008 [01:55<2:33:48, 5.14it/s] 1%| | 564/48008 [01:55<2:34:38, 5.11it/s] 1%| | 565/48008 [01:55<2:34:44, 5.11it/s] 1%| | 566/48008 [01:55<2:44:48, 4.80it/s] 1%| | 567/48008 [01:56<2:41:14, 4.90it/s] 1%| | 568/48008 [01:56<2:42:47, 4.86it/s] 1%| | 569/48008 [01:56<2:39:25, 4.96it/s] 1%| | 570/48008 [01:56<2:37:16, 5.03it/s] 1%| | 571/48008 [01:56<2:38:19, 4.99it/s] 1%| | 572/48008 [01:57<2:36:18, 5.06it/s] 1%| | 573/48008 [01:57<2:33:38, 5.15it/s] 1%| | 574/48008 [01:57<2:32:45, 5.18it/s] 1%| | 575/48008 [01:57<2:33:29, 5.15it/s] 1%| | 576/48008 [01:57<2:32:47, 5.17it/s] 1%| | 577/48008 [01:57<2:30:57, 5.24it/s] 1%| | 578/48008 [01:58<2:30:40, 5.25it/s] 1%| | 579/48008 [01:58<2:30:42, 5.24it/s] 1%| | 580/48008 [01:58<2:29:09, 5.30it/s] 1%| | 581/48008 [01:58<2:27:53, 5.35it/s] 1%| | 582/48008 [01:58<2:31:12, 5.23it/s] 1%| | 583/48008 [01:59<2:30:00, 5.27it/s] 1%| | 584/48008 [01:59<2:41:16, 4.90it/s] 1%| | 585/48008 [01:59<2:37:17, 5.02it/s] 1%| | 586/48008 [01:59<2:34:39, 5.11it/s] 1%| | 587/48008 [01:59<2:33:20, 5.15it/s] 1%| | 588/48008 [02:00<2:35:20, 5.09it/s] 1%| | 589/48008 [02:00<2:36:38, 5.05it/s] 1%| | 590/48008 [02:00<2:34:36, 5.11it/s] 1%| | 591/48008 [02:00<2:36:12, 5.06it/s] 1%| | 592/48008 [02:00<2:34:45, 5.11it/s] 1%| | 593/48008 [02:01<2:32:09, 5.19it/s] 1%| | 594/48008 [02:01<2:36:13, 5.06it/s] 1%| | 595/48008 [02:01<2:33:37, 5.14it/s] 1%| | 596/48008 [02:02<5:16:46, 2.49it/s] 1%| | 597/48008 [02:02<4:28:15, 2.95it/s] 1%| | 598/48008 [02:02<3:52:42, 3.40it/s] 1%| | 599/48008 [02:02<3:28:10, 3.80it/s] 1%| | 600/48008 [02:03<3:14:12, 4.07it/s] {'loss': 0.364, 'grad_norm': 2.859375, 'learning_rate': 0.00019750458256957175, 'epoch': 0.01} 1%| | 600/48008 [02:03<3:14:12, 4.07it/s] 1%|▏ | 601/48008 [02:03<3:02:09, 4.34it/s] 1%|▏ | 602/48008 [02:03<2:53:21, 4.56it/s] 1%|▏ | 603/48008 [02:03<2:51:00, 4.62it/s] 1%|▏ | 604/48008 [02:03<2:45:08, 4.78it/s] 1%|▏ | 605/48008 [02:04<2:52:02, 4.59it/s] 1%|▏ | 606/48008 [02:04<2:45:52, 4.76it/s] 1%|▏ | 607/48008 [02:04<2:40:14, 4.93it/s] 1%|▏ | 608/48008 [02:04<2:40:16, 4.93it/s] 1%|▏ | 609/48008 [02:04<2:40:04, 4.94it/s] 1%|▏ | 610/48008 [02:05<2:36:34, 5.05it/s] 1%|▏ | 611/48008 [02:05<2:33:37, 5.14it/s] 1%|▏ | 612/48008 [02:05<2:32:49, 5.17it/s] 1%|▏ | 613/48008 [02:05<2:29:16, 5.29it/s] 1%|▏ | 614/48008 [02:05<2:29:51, 5.27it/s] 1%|▏ | 615/48008 [02:06<2:31:43, 5.21it/s] 1%|▏ | 616/48008 [02:06<2:31:47, 5.20it/s] 1%|▏ | 617/48008 [02:06<2:29:13, 5.29it/s] 1%|▏ | 618/48008 [02:06<2:30:55, 5.23it/s] 1%|▏ | 619/48008 [02:06<2:30:29, 5.25it/s] 1%|▏ | 620/48008 [02:07<2:31:41, 5.21it/s] 1%|▏ | 621/48008 [02:07<2:33:57, 5.13it/s] 1%|▏ | 622/48008 [02:07<2:31:39, 5.21it/s] 1%|▏ | 623/48008 [02:07<2:33:58, 5.13it/s] 1%|▏ | 624/48008 [02:07<2:32:54, 5.16it/s] 1%|▏ | 625/48008 [02:08<2:32:00, 5.20it/s] 1%|▏ | 626/48008 [02:08<2:42:49, 4.85it/s] 1%|▏ | 627/48008 [02:08<2:38:11, 4.99it/s] 1%|▏ | 628/48008 [02:08<2:36:29, 5.05it/s] 1%|▏ | 629/48008 [02:08<2:34:03, 5.13it/s] 1%|▏ | 630/48008 [02:09<2:34:55, 5.10it/s] 1%|▏ | 631/48008 [02:09<2:33:21, 5.15it/s] 1%|▏ | 632/48008 [02:09<2:31:53, 5.20it/s] 1%|▏ | 633/48008 [02:09<2:42:08, 4.87it/s] 1%|▏ | 634/48008 [02:09<2:37:21, 5.02it/s] 1%|▏ | 635/48008 [02:10<2:33:59, 5.13it/s] 1%|▏ | 636/48008 [02:10<2:43:29, 4.83it/s] 1%|▏ | 637/48008 [02:10<2:39:09, 4.96it/s] 1%|▏ | 638/48008 [02:10<2:36:29, 5.05it/s] 1%|▏ | 639/48008 [02:10<2:34:42, 5.10it/s] 1%|▏ | 640/48008 [02:10<2:32:50, 5.17it/s] 1%|▏ | 641/48008 [02:11<2:32:59, 5.16it/s] 1%|▏ | 642/48008 [02:11<2:34:58, 5.09it/s] 1%|▏ | 643/48008 [02:11<2:34:30, 5.11it/s] 1%|▏ | 644/48008 [02:11<2:32:52, 5.16it/s] 1%|▏ | 645/48008 [02:11<2:32:13, 5.19it/s] 1%|▏ | 646/48008 [02:12<2:31:48, 5.20it/s] 1%|▏ | 647/48008 [02:12<2:31:10, 5.22it/s] 1%|▏ | 648/48008 [02:12<2:29:44, 5.27it/s] 1%|▏ | 649/48008 [02:12<2:29:51, 5.27it/s] 1%|▏ | 650/48008 [02:12<2:29:45, 5.27it/s] {'loss': 0.3519, 'grad_norm': 0.625, 'learning_rate': 0.00019729628395267457, 'epoch': 0.01} 1%|▏ | 650/48008 [02:12<2:29:45, 5.27it/s] 1%|▏ | 651/48008 [02:13<2:31:51, 5.20it/s] 1%|▏ | 652/48008 [02:13<2:32:32, 5.17it/s] 1%|▏ | 653/48008 [02:13<2:31:46, 5.20it/s] 1%|▏ | 654/48008 [02:13<2:31:41, 5.20it/s] 1%|▏ | 655/48008 [02:13<2:32:25, 5.18it/s] 1%|▏ | 656/48008 [02:14<2:31:45, 5.20it/s] 1%|▏ | 657/48008 [02:14<2:33:50, 5.13it/s] 1%|▏ | 658/48008 [02:14<2:31:20, 5.21it/s] 1%|▏ | 659/48008 [02:14<2:29:35, 5.28it/s] 1%|▏ | 660/48008 [02:14<2:30:00, 5.26it/s] 1%|▏ | 661/48008 [02:15<2:31:30, 5.21it/s] 1%|▏ | 662/48008 [02:15<2:30:37, 5.24it/s] 1%|▏ | 663/48008 [02:15<2:30:41, 5.24it/s] 1%|▏ | 664/48008 [02:15<2:30:45, 5.23it/s] 1%|▏ | 665/48008 [02:15<2:30:58, 5.23it/s] 1%|▏ | 666/48008 [02:15<2:30:11, 5.25it/s] 1%|▏ | 667/48008 [02:16<2:30:03, 5.26it/s] 1%|▏ | 668/48008 [02:16<2:30:27, 5.24it/s] 1%|▏ | 669/48008 [02:16<2:30:27, 5.24it/s] 1%|▏ | 670/48008 [02:16<2:29:23, 5.28it/s] 1%|▏ | 671/48008 [02:16<2:29:54, 5.26it/s] 1%|▏ | 672/48008 [02:17<2:31:18, 5.21it/s] 1%|▏ | 673/48008 [02:17<2:30:43, 5.23it/s] 1%|▏ | 674/48008 [02:17<2:32:59, 5.16it/s] 1%|▏ | 675/48008 [02:17<2:33:16, 5.15it/s] 1%|▏ | 676/48008 [02:17<2:34:00, 5.12it/s] 1%|▏ | 677/48008 [02:18<2:34:04, 5.12it/s] 1%|▏ | 678/48008 [02:18<2:34:17, 5.11it/s] 1%|▏ | 679/48008 [02:18<2:33:17, 5.15it/s] 1%|▏ | 680/48008 [02:18<2:32:53, 5.16it/s] 1%|▏ | 681/48008 [02:18<2:33:43, 5.13it/s] 1%|▏ | 682/48008 [02:19<2:32:55, 5.16it/s] 1%|▏ | 683/48008 [02:19<2:32:42, 5.17it/s] 1%|▏ | 684/48008 [02:19<2:31:38, 5.20it/s] 1%|▏ | 685/48008 [02:19<2:42:22, 4.86it/s] 1%|▏ | 686/48008 [02:19<2:38:21, 4.98it/s] 1%|▏ | 687/48008 [02:20<2:37:19, 5.01it/s] 1%|▏ | 688/48008 [02:20<2:34:36, 5.10it/s] 1%|▏ | 689/48008 [02:20<2:32:35, 5.17it/s] 1%|▏ | 690/48008 [02:20<2:31:47, 5.20it/s] 1%|▏ | 691/48008 [02:20<2:30:31, 5.24it/s] 1%|▏ | 692/48008 [02:21<2:34:50, 5.09it/s] 1%|▏ | 693/48008 [02:21<3:19:38, 3.95it/s] 1%|▏ | 694/48008 [02:21<3:06:17, 4.23it/s] 1%|▏ | 695/48008 [02:21<2:56:01, 4.48it/s] 1%|▏ | 696/48008 [02:22<2:50:05, 4.64it/s] 1%|▏ | 697/48008 [02:22<2:43:56, 4.81it/s] 1%|▏ | 698/48008 [02:22<2:40:00, 4.93it/s] 1%|▏ | 699/48008 [02:22<2:38:20, 4.98it/s] 1%|▏ | 700/48008 [02:22<2:38:25, 4.98it/s] {'loss': 0.3536, 'grad_norm': 1.1015625, 'learning_rate': 0.00019708798533577736, 'epoch': 0.01} 1%|▏ | 700/48008 [02:22<2:38:25, 4.98it/s] 1%|▏ | 701/48008 [02:23<2:39:20, 4.95it/s] 1%|▏ | 702/48008 [02:23<2:36:31, 5.04it/s] 1%|▏ | 703/48008 [02:23<2:35:06, 5.08it/s] 1%|▏ | 704/48008 [02:23<2:33:42, 5.13it/s] 1%|▏ | 705/48008 [02:23<2:34:07, 5.12it/s] 1%|▏ | 706/48008 [02:23<2:37:15, 5.01it/s] 1%|▏ | 707/48008 [02:24<2:35:06, 5.08it/s] 1%|▏ | 708/48008 [02:24<2:34:25, 5.11it/s] 1%|▏ | 709/48008 [02:24<2:34:40, 5.10it/s] 1%|▏ | 710/48008 [02:24<3:19:24, 3.95it/s] 1%|▏ | 711/48008 [02:25<3:04:44, 4.27it/s] 1%|▏ | 712/48008 [02:25<2:55:20, 4.50it/s] 1%|▏ | 713/48008 [02:25<2:49:35, 4.65it/s] 1%|▏ | 714/48008 [02:25<2:43:36, 4.82it/s] 1%|▏ | 715/48008 [02:25<2:40:27, 4.91it/s] 1%|▏ | 716/48008 [02:26<2:37:19, 5.01it/s] 1%|▏ | 717/48008 [02:26<2:46:24, 4.74it/s] 1%|▏ | 718/48008 [02:26<2:40:51, 4.90it/s] 1%|▏ | 719/48008 [02:26<2:36:40, 5.03it/s] 1%|▏ | 720/48008 [02:26<2:36:35, 5.03it/s] 2%|▏ | 721/48008 [02:27<2:34:39, 5.10it/s] 2%|▏ | 722/48008 [02:27<2:35:19, 5.07it/s] 2%|▏ | 723/48008 [02:27<3:20:24, 3.93it/s] 2%|▏ | 724/48008 [02:27<3:05:25, 4.25it/s] 2%|▏ | 725/48008 [02:28<2:53:57, 4.53it/s] 2%|▏ | 726/48008 [02:28<2:57:51, 4.43it/s] 2%|▏ | 727/48008 [02:28<2:51:16, 4.60it/s] 2%|▏ | 728/48008 [02:28<2:45:04, 4.77it/s] 2%|▏ | 729/48008 [02:28<2:39:53, 4.93it/s] 2%|▏ | 730/48008 [02:29<2:36:53, 5.02it/s] 2%|▏ | 731/48008 [02:29<2:38:58, 4.96it/s] 2%|▏ | 732/48008 [02:29<2:36:13, 5.04it/s] 2%|▏ | 733/48008 [02:29<2:34:12, 5.11it/s] 2%|▏ | 734/48008 [02:29<2:34:56, 5.09it/s] 2%|▏ | 735/48008 [02:30<2:32:52, 5.15it/s] 2%|▏ | 736/48008 [02:30<2:36:20, 5.04it/s] 2%|▏ | 737/48008 [02:30<2:35:31, 5.07it/s] 2%|▏ | 738/48008 [02:30<2:34:16, 5.11it/s] 2%|▏ | 739/48008 [02:30<2:31:55, 5.19it/s] 2%|▏ | 740/48008 [02:31<2:32:01, 5.18it/s] 2%|▏ | 741/48008 [02:31<2:34:50, 5.09it/s] 2%|▏ | 742/48008 [02:31<2:33:18, 5.14it/s] 2%|▏ | 743/48008 [02:31<2:32:40, 5.16it/s] 2%|▏ | 744/48008 [02:31<2:31:40, 5.19it/s] 2%|▏ | 745/48008 [02:31<2:30:55, 5.22it/s] 2%|▏ | 746/48008 [02:32<2:34:45, 5.09it/s] 2%|▏ | 747/48008 [02:32<2:33:07, 5.14it/s] 2%|▏ | 748/48008 [02:32<2:31:30, 5.20it/s] 2%|▏ | 749/48008 [02:32<2:31:24, 5.20it/s] 2%|▏ | 750/48008 [02:32<2:30:55, 5.22it/s] {'loss': 0.3503, 'grad_norm': 1.3671875, 'learning_rate': 0.0001968796867188802, 'epoch': 0.02} 2%|▏ | 750/48008 [02:32<2:30:55, 5.22it/s] 2%|▏ | 751/48008 [02:33<2:30:11, 5.24it/s] 2%|▏ | 752/48008 [02:33<2:30:02, 5.25it/s] 2%|▏ | 753/48008 [02:33<2:31:11, 5.21it/s] 2%|▏ | 754/48008 [02:33<2:31:04, 5.21it/s] 2%|▏ | 755/48008 [02:33<2:41:35, 4.87it/s] 2%|▏ | 756/48008 [02:34<2:38:05, 4.98it/s] 2%|▏ | 757/48008 [02:34<2:37:41, 4.99it/s] 2%|▏ | 758/48008 [02:34<2:39:42, 4.93it/s] 2%|▏ | 759/48008 [02:34<2:35:53, 5.05it/s] 2%|▏ | 760/48008 [02:34<2:33:38, 5.13it/s] 2%|▏ | 761/48008 [02:35<2:31:39, 5.19it/s] 2%|▏ | 762/48008 [02:35<2:31:09, 5.21it/s] 2%|▏ | 763/48008 [02:35<2:32:35, 5.16it/s] 2%|▏ | 764/48008 [02:35<2:31:54, 5.18it/s] 2%|▏ | 765/48008 [02:35<2:33:04, 5.14it/s] 2%|▏ | 766/48008 [02:36<2:31:59, 5.18it/s] 2%|▏ | 767/48008 [02:36<2:30:27, 5.23it/s] 2%|▏ | 768/48008 [02:36<2:30:31, 5.23it/s] 2%|▏ | 769/48008 [02:36<2:30:51, 5.22it/s] 2%|▏ | 770/48008 [02:36<2:30:49, 5.22it/s] 2%|▏ | 771/48008 [02:37<2:32:27, 5.16it/s] 2%|▏ | 772/48008 [02:37<2:34:56, 5.08it/s] 2%|▏ | 773/48008 [02:37<2:34:13, 5.10it/s] 2%|▏ | 774/48008 [02:37<2:33:31, 5.13it/s] 2%|▏ | 775/48008 [02:37<2:43:26, 4.82it/s] 2%|▏ | 776/48008 [02:38<2:39:12, 4.94it/s] 2%|▏ | 777/48008 [02:38<2:35:55, 5.05it/s] 2%|▏ | 778/48008 [02:38<2:32:21, 5.17it/s] 2%|▏ | 779/48008 [02:38<2:32:04, 5.18it/s] 2%|▏ | 780/48008 [02:38<2:30:17, 5.24it/s] 2%|▏ | 781/48008 [02:39<2:29:25, 5.27it/s] 2%|▏ | 782/48008 [02:39<2:28:56, 5.28it/s] 2%|▏ | 783/48008 [02:39<2:29:21, 5.27it/s] 2%|▏ | 784/48008 [02:39<2:31:42, 5.19it/s] 2%|▏ | 785/48008 [02:39<2:31:29, 5.20it/s] 2%|▏ | 786/48008 [02:39<2:29:50, 5.25it/s] 2%|▏ | 787/48008 [02:40<2:29:26, 5.27it/s] 2%|▏ | 788/48008 [02:40<2:29:22, 5.27it/s] 2%|▏ | 789/48008 [02:40<2:29:47, 5.25it/s] 2%|▏ | 790/48008 [02:40<2:29:49, 5.25it/s] 2%|▏ | 791/48008 [02:40<2:29:00, 5.28it/s] 2%|▏ | 792/48008 [02:41<2:29:29, 5.26it/s] 2%|▏ | 793/48008 [02:41<2:29:44, 5.26it/s] 2%|▏ | 794/48008 [02:41<2:29:52, 5.25it/s] 2%|▏ | 795/48008 [02:41<2:29:07, 5.28it/s] 2%|▏ | 796/48008 [02:41<2:29:04, 5.28it/s] 2%|▏ | 797/48008 [02:42<2:40:03, 4.92it/s] 2%|▏ | 798/48008 [02:42<2:38:21, 4.97it/s] 2%|▏ | 799/48008 [02:42<2:37:09, 5.01it/s] 2%|▏ | 800/48008 [02:42<2:35:27, 5.06it/s] {'loss': 0.3406, 'grad_norm': 0.71875, 'learning_rate': 0.00019667138810198302, 'epoch': 0.02} 2%|▏ | 800/48008 [02:42<2:35:27, 5.06it/s] 2%|▏ | 801/48008 [02:42<2:35:47, 5.05it/s] 2%|▏ | 802/48008 [02:43<2:45:23, 4.76it/s] 2%|▏ | 803/48008 [02:43<2:41:04, 4.88it/s] 2%|▏ | 804/48008 [02:43<2:37:43, 4.99it/s] 2%|▏ | 805/48008 [02:43<2:46:32, 4.72it/s] 2%|▏ | 806/48008 [02:43<2:45:25, 4.76it/s] 2%|▏ | 807/48008 [02:44<2:51:57, 4.58it/s] 2%|▏ | 808/48008 [02:44<2:46:17, 4.73it/s] 2%|▏ | 809/48008 [02:44<2:43:23, 4.81it/s] 2%|▏ | 810/48008 [02:44<2:39:05, 4.94it/s] 2%|▏ | 811/48008 [02:44<2:39:25, 4.93it/s] 2%|▏ | 812/48008 [02:45<2:38:16, 4.97it/s] 2%|▏ | 813/48008 [02:45<2:38:25, 4.97it/s] 2%|▏ | 814/48008 [02:45<2:37:20, 5.00it/s] 2%|▏ | 815/48008 [02:45<2:39:56, 4.92it/s] 2%|▏ | 816/48008 [02:45<2:35:59, 5.04it/s] 2%|▏ | 817/48008 [02:46<2:33:55, 5.11it/s] 2%|▏ | 818/48008 [02:46<2:30:44, 5.22it/s] 2%|▏ | 819/48008 [02:46<2:31:49, 5.18it/s] 2%|▏ | 820/48008 [02:46<2:30:44, 5.22it/s] 2%|▏ | 821/48008 [02:46<2:29:20, 5.27it/s] 2%|▏ | 822/48008 [02:47<2:29:11, 5.27it/s] 2%|▏ | 823/48008 [02:47<2:28:05, 5.31it/s] 2%|▏ | 824/48008 [02:47<2:28:53, 5.28it/s] 2%|▏ | 825/48008 [02:47<2:32:02, 5.17it/s] 2%|▏ | 826/48008 [02:47<2:31:39, 5.18it/s] 2%|▏ | 827/48008 [02:48<2:30:38, 5.22it/s] 2%|▏ | 828/48008 [02:48<2:30:54, 5.21it/s] 2%|▏ | 829/48008 [02:48<2:30:47, 5.21it/s] 2%|▏ | 830/48008 [02:48<2:29:44, 5.25it/s] 2%|▏ | 831/48008 [02:48<2:40:43, 4.89it/s] 2%|▏ | 832/48008 [02:49<3:23:18, 3.87it/s] 2%|▏ | 833/48008 [02:49<3:07:11, 4.20it/s] 2%|▏ | 834/48008 [02:49<3:07:38, 4.19it/s] 2%|▏ | 835/48008 [02:49<3:00:36, 4.35it/s] 2%|▏ | 836/48008 [02:50<2:51:32, 4.58it/s] 2%|▏ | 837/48008 [02:50<2:44:53, 4.77it/s] 2%|▏ | 838/48008 [02:50<2:41:34, 4.87it/s] 2%|▏ | 839/48008 [02:50<3:24:12, 3.85it/s] 2%|▏ | 840/48008 [02:51<3:06:47, 4.21it/s] 2%|▏ | 841/48008 [02:51<2:53:55, 4.52it/s] 2%|▏ | 842/48008 [02:51<2:49:48, 4.63it/s] 2%|▏ | 843/48008 [02:51<2:43:29, 4.81it/s] 2%|▏ | 844/48008 [02:51<2:39:13, 4.94it/s] 2%|▏ | 845/48008 [02:51<2:35:07, 5.07it/s] 2%|▏ | 846/48008 [02:52<2:33:48, 5.11it/s] 2%|▏ | 847/48008 [02:52<2:32:43, 5.15it/s] 2%|▏ | 848/48008 [02:52<2:31:36, 5.18it/s] 2%|▏ | 849/48008 [02:52<2:32:05, 5.17it/s] 2%|▏ | 850/48008 [02:52<2:32:25, 5.16it/s] {'loss': 0.3353, 'grad_norm': 0.4375, 'learning_rate': 0.00019646308948508584, 'epoch': 0.02} 2%|▏ | 850/48008 [02:52<2:32:25, 5.16it/s] 2%|▏ | 851/48008 [02:53<2:32:01, 5.17it/s] 2%|▏ | 852/48008 [02:53<2:33:45, 5.11it/s] 2%|▏ | 853/48008 [02:53<2:32:31, 5.15it/s] 2%|▏ | 854/48008 [02:53<2:31:39, 5.18it/s] 2%|▏ | 855/48008 [02:53<2:42:20, 4.84it/s] 2%|▏ | 856/48008 [02:54<2:38:16, 4.97it/s] 2%|▏ | 857/48008 [02:54<2:35:43, 5.05it/s] 2%|▏ | 858/48008 [02:54<2:33:45, 5.11it/s] 2%|▏ | 859/48008 [02:54<2:31:55, 5.17it/s] 2%|▏ | 860/48008 [02:54<2:30:59, 5.20it/s] 2%|▏ | 861/48008 [02:55<2:30:44, 5.21it/s] 2%|▏ | 862/48008 [02:55<2:31:22, 5.19it/s] 2%|▏ | 863/48008 [02:55<2:33:16, 5.13it/s] 2%|▏ | 864/48008 [02:55<2:32:37, 5.15it/s] 2%|▏ | 865/48008 [02:55<2:32:31, 5.15it/s] 2%|▏ | 866/48008 [02:56<2:31:18, 5.19it/s] 2%|▏ | 867/48008 [02:56<2:30:39, 5.22it/s] 2%|▏ | 868/48008 [02:56<2:30:36, 5.22it/s] 2%|▏ | 869/48008 [02:56<2:31:24, 5.19it/s] 2%|▏ | 870/48008 [02:56<2:32:19, 5.16it/s] 2%|▏ | 871/48008 [02:57<2:32:44, 5.14it/s] 2%|▏ | 872/48008 [02:57<2:36:01, 5.04it/s] 2%|▏ | 873/48008 [02:57<2:35:55, 5.04it/s] 2%|▏ | 874/48008 [02:57<2:34:54, 5.07it/s] 2%|▏ | 875/48008 [02:57<2:35:01, 5.07it/s] 2%|▏ | 876/48008 [02:58<2:36:13, 5.03it/s] 2%|▏ | 877/48008 [02:58<2:34:41, 5.08it/s] 2%|▏ | 878/48008 [02:58<2:32:12, 5.16it/s] 2%|▏ | 879/48008 [02:58<2:31:54, 5.17it/s] 2%|▏ | 880/48008 [02:58<2:34:20, 5.09it/s] 2%|▏ | 881/48008 [02:59<2:33:23, 5.12it/s] 2%|▏ | 882/48008 [02:59<2:32:45, 5.14it/s] 2%|▏ | 883/48008 [02:59<2:31:58, 5.17it/s] 2%|▏ | 884/48008 [02:59<2:30:59, 5.20it/s] 2%|▏ | 885/48008 [02:59<2:29:31, 5.25it/s] 2%|▏ | 886/48008 [02:59<2:32:34, 5.15it/s] 2%|▏ | 887/48008 [03:00<2:31:03, 5.20it/s] 2%|▏ | 888/48008 [03:00<2:31:28, 5.18it/s] 2%|▏ | 889/48008 [03:00<2:30:58, 5.20it/s] 2%|▏ | 890/48008 [03:00<2:30:21, 5.22it/s] 2%|▏ | 891/48008 [03:00<2:29:39, 5.25it/s] 2%|▏ | 892/48008 [03:01<2:40:40, 4.89it/s] 2%|▏ | 893/48008 [03:01<2:39:46, 4.91it/s] 2%|▏ | 894/48008 [03:01<2:37:03, 5.00it/s] 2%|▏ | 895/48008 [03:01<2:35:17, 5.06it/s] 2%|▏ | 896/48008 [03:01<2:32:10, 5.16it/s] 2%|▏ | 897/48008 [03:02<2:32:07, 5.16it/s] 2%|▏ | 898/48008 [03:02<2:33:29, 5.12it/s] 2%|▏ | 899/48008 [03:02<2:43:07, 4.81it/s] 2%|▏ | 900/48008 [03:02<2:38:37, 4.95it/s] {'loss': 0.3359, 'grad_norm': 0.546875, 'learning_rate': 0.00019625479086818866, 'epoch': 0.02} 2%|▏ | 900/48008 [03:02<2:38:37, 4.95it/s] 2%|▏ | 901/48008 [03:02<2:35:42, 5.04it/s] 2%|▏ | 902/48008 [03:03<2:33:09, 5.13it/s] 2%|▏ | 903/48008 [03:03<2:30:59, 5.20it/s] 2%|▏ | 904/48008 [03:03<2:30:20, 5.22it/s] 2%|▏ | 905/48008 [03:03<2:29:59, 5.23it/s] 2%|▏ | 906/48008 [03:03<2:31:09, 5.19it/s] 2%|▏ | 907/48008 [03:04<2:29:22, 5.26it/s] 2%|▏ | 908/48008 [03:04<2:29:11, 5.26it/s] 2%|▏ | 909/48008 [03:04<2:29:47, 5.24it/s] 2%|▏ | 910/48008 [03:04<2:29:23, 5.25it/s] 2%|▏ | 911/48008 [03:04<2:29:31, 5.25it/s] 2%|▏ | 912/48008 [03:05<2:29:43, 5.24it/s] 2%|▏ | 913/48008 [03:05<2:29:02, 5.27it/s] 2%|▏ | 914/48008 [03:05<2:40:20, 4.90it/s] 2%|▏ | 915/48008 [03:05<2:37:49, 4.97it/s] 2%|▏ | 916/48008 [03:05<2:35:15, 5.06it/s] 2%|▏ | 917/48008 [03:06<2:36:20, 5.02it/s] 2%|▏ | 918/48008 [03:06<2:35:12, 5.06it/s] 2%|▏ | 919/48008 [03:06<2:34:01, 5.10it/s] 2%|▏ | 920/48008 [03:06<2:35:05, 5.06it/s] 2%|▏ | 921/48008 [03:06<2:33:59, 5.10it/s] 2%|▏ | 922/48008 [03:07<2:33:52, 5.10it/s] 2%|▏ | 923/48008 [03:07<2:35:31, 5.05it/s] 2%|▏ | 924/48008 [03:07<2:33:04, 5.13it/s] 2%|▏ | 925/48008 [03:07<2:32:46, 5.14it/s] 2%|▏ | 926/48008 [03:07<2:31:56, 5.16it/s] 2%|▏ | 927/48008 [03:07<2:30:09, 5.23it/s] 2%|▏ | 928/48008 [03:08<2:32:49, 5.13it/s] 2%|▏ | 929/48008 [03:08<2:32:00, 5.16it/s] 2%|▏ | 930/48008 [03:08<2:30:02, 5.23it/s] 2%|▏ | 931/48008 [03:08<2:29:21, 5.25it/s] 2%|▏ | 932/48008 [03:08<2:30:29, 5.21it/s] 2%|▏ | 933/48008 [03:09<2:34:39, 5.07it/s] 2%|▏ | 934/48008 [03:09<2:36:17, 5.02it/s] 2%|▏ | 935/48008 [03:09<2:33:15, 5.12it/s] 2%|▏ | 936/48008 [03:09<2:31:42, 5.17it/s] 2%|▏ | 937/48008 [03:09<2:42:30, 4.83it/s] 2%|▏ | 938/48008 [03:10<2:36:54, 5.00it/s] 2%|▏ | 939/48008 [03:10<2:35:05, 5.06it/s] 2%|▏ | 940/48008 [03:10<2:31:24, 5.18it/s] 2%|▏ | 941/48008 [03:10<2:33:30, 5.11it/s] 2%|▏ | 942/48008 [03:10<2:32:45, 5.14it/s] 2%|▏ | 943/48008 [03:11<2:42:35, 4.82it/s] 2%|▏ | 944/48008 [03:11<2:38:59, 4.93it/s] 2%|▏ | 945/48008 [03:11<3:22:05, 3.88it/s] 2%|▏ | 946/48008 [03:12<3:52:28, 3.37it/s] 2%|▏ | 947/48008 [03:12<3:26:25, 3.80it/s] 2%|▏ | 948/48008 [03:12<3:09:18, 4.14it/s] 2%|▏ | 949/48008 [03:12<2:56:50, 4.44it/s] 2%|▏ | 950/48008 [03:12<2:52:07, 4.56it/s] {'loss': 0.3229, 'grad_norm': 1.9140625, 'learning_rate': 0.00019604649225129145, 'epoch': 0.02} 2%|▏ | 950/48008 [03:12<2:52:07, 4.56it/s] 2%|▏ | 951/48008 [03:13<2:48:11, 4.66it/s] 2%|▏ | 952/48008 [03:13<2:42:57, 4.81it/s] 2%|▏ | 953/48008 [03:13<2:41:46, 4.85it/s] 2%|▏ | 954/48008 [03:13<2:37:19, 4.98it/s] 2%|▏ | 955/48008 [03:13<2:36:54, 5.00it/s] 2%|▏ | 956/48008 [03:14<2:36:21, 5.02it/s] 2%|▏ | 957/48008 [03:14<2:33:05, 5.12it/s] 2%|▏ | 958/48008 [03:14<2:32:22, 5.15it/s] 2%|▏ | 959/48008 [03:14<2:32:43, 5.13it/s] 2%|▏ | 960/48008 [03:14<2:31:21, 5.18it/s] 2%|▏ | 961/48008 [03:15<2:31:47, 5.17it/s] 2%|▏ | 962/48008 [03:15<2:34:23, 5.08it/s] 2%|▏ | 963/48008 [03:15<2:32:27, 5.14it/s] 2%|▏ | 964/48008 [03:15<2:31:53, 5.16it/s] 2%|▏ | 965/48008 [03:15<2:30:23, 5.21it/s] 2%|▏ | 966/48008 [03:15<2:30:32, 5.21it/s] 2%|▏ | 967/48008 [03:16<2:30:22, 5.21it/s] 2%|▏ | 968/48008 [03:16<2:30:02, 5.23it/s] 2%|▏ | 969/48008 [03:16<2:30:09, 5.22it/s] 2%|▏ | 970/48008 [03:16<2:30:04, 5.22it/s] 2%|▏ | 971/48008 [03:16<2:29:59, 5.23it/s] 2%|▏ | 972/48008 [03:17<2:31:27, 5.18it/s] 2%|▏ | 973/48008 [03:17<2:29:21, 5.25it/s] 2%|▏ | 974/48008 [03:17<2:29:25, 5.25it/s] 2%|▏ | 975/48008 [03:17<2:29:59, 5.23it/s] 2%|▏ | 976/48008 [03:17<2:33:00, 5.12it/s] 2%|▏ | 977/48008 [03:18<2:36:07, 5.02it/s] 2%|▏ | 978/48008 [03:18<2:33:35, 5.10it/s] 2%|▏ | 979/48008 [03:18<2:32:58, 5.12it/s] 2%|▏ | 980/48008 [03:18<2:31:04, 5.19it/s] 2%|▏ | 981/48008 [03:18<2:30:27, 5.21it/s] 2%|▏ | 982/48008 [03:19<2:34:16, 5.08it/s] 2%|▏ | 983/48008 [03:19<2:32:55, 5.12it/s] 2%|▏ | 984/48008 [03:19<2:33:36, 5.10it/s] 2%|▏ | 985/48008 [03:19<2:31:03, 5.19it/s] 2%|▏ | 986/48008 [03:19<2:57:41, 4.41it/s] 2%|▏ | 987/48008 [03:20<2:49:37, 4.62it/s] 2%|▏ | 988/48008 [03:20<2:43:06, 4.80it/s] 2%|▏ | 989/48008 [03:20<2:43:39, 4.79it/s] 2%|▏ | 990/48008 [03:20<2:38:51, 4.93it/s] 2%|▏ | 991/48008 [03:20<2:35:24, 5.04it/s] 2%|▏ | 992/48008 [03:21<2:33:58, 5.09it/s] 2%|▏ | 993/48008 [03:21<2:35:30, 5.04it/s] 2%|▏ | 994/48008 [03:21<2:44:39, 4.76it/s] 2%|▏ | 995/48008 [03:21<2:40:04, 4.89it/s] 2%|▏ | 996/48008 [03:21<2:38:59, 4.93it/s] 2%|▏ | 997/48008 [03:22<2:35:50, 5.03it/s] 2%|▏ | 998/48008 [03:22<2:35:58, 5.02it/s] 2%|▏ | 999/48008 [03:22<2:35:46, 5.03it/s] 2%|▏ | 1000/48008 [03:22<2:35:12, 5.05it/s] {'loss': 0.3237, 'grad_norm': 0.3984375, 'learning_rate': 0.0001958381936343943, 'epoch': 0.02} 2%|▏ | 1000/48008 [03:22<2:35:12, 5.05it/s] 2%|▏ | 1001/48008 [03:22<2:34:30, 5.07it/s] 2%|▏ | 1002/48008 [03:23<2:34:22, 5.07it/s] 2%|▏ | 1003/48008 [03:23<2:32:05, 5.15it/s] 2%|▏ | 1004/48008 [03:23<2:33:25, 5.11it/s] 2%|▏ | 1005/48008 [03:23<2:33:04, 5.12it/s] 2%|▏ | 1006/48008 [03:23<2:34:02, 5.09it/s] 2%|▏ | 1007/48008 [03:24<2:32:53, 5.12it/s] 2%|▏ | 1008/48008 [03:24<2:31:33, 5.17it/s] 2%|▏ | 1009/48008 [03:24<2:32:22, 5.14it/s] 2%|▏ | 1010/48008 [03:24<2:31:16, 5.18it/s] 2%|▏ | 1011/48008 [03:24<2:30:55, 5.19it/s] 2%|▏ | 1012/48008 [03:25<2:30:20, 5.21it/s] 2%|▏ | 1013/48008 [03:25<2:32:24, 5.14it/s] 2%|▏ | 1014/48008 [03:25<2:34:50, 5.06it/s] 2%|▏ | 1015/48008 [03:25<2:35:40, 5.03it/s] 2%|▏ | 1016/48008 [03:25<2:35:35, 5.03it/s] 2%|▏ | 1017/48008 [03:26<2:36:25, 5.01it/s] 2%|▏ | 1018/48008 [03:26<2:35:29, 5.04it/s] 2%|▏ | 1019/48008 [03:26<2:33:09, 5.11it/s] 2%|▏ | 1020/48008 [03:26<2:31:44, 5.16it/s] 2%|▏ | 1021/48008 [03:26<2:31:38, 5.16it/s] 2%|▏ | 1022/48008 [03:27<2:32:20, 5.14it/s] 2%|▏ | 1023/48008 [03:27<2:32:39, 5.13it/s] 2%|▏ | 1024/48008 [03:27<2:30:58, 5.19it/s] 2%|▏ | 1025/48008 [03:27<2:29:42, 5.23it/s] 2%|▏ | 1026/48008 [03:27<2:29:09, 5.25it/s] 2%|▏ | 1027/48008 [03:28<2:33:38, 5.10it/s] 2%|▏ | 1028/48008 [03:28<2:32:57, 5.12it/s] 2%|▏ | 1029/48008 [03:28<2:30:49, 5.19it/s] 2%|▏ | 1030/48008 [03:28<2:29:50, 5.23it/s] 2%|▏ | 1031/48008 [03:28<2:27:54, 5.29it/s] 2%|▏ | 1032/48008 [03:28<2:29:49, 5.23it/s] 2%|▏ | 1033/48008 [03:29<2:32:35, 5.13it/s] 2%|▏ | 1034/48008 [03:29<3:17:45, 3.96it/s] 2%|▏ | 1035/48008 [03:29<3:06:02, 4.21it/s] 2%|▏ | 1036/48008 [03:29<2:55:22, 4.46it/s] 2%|▏ | 1037/48008 [03:30<2:46:48, 4.69it/s] 2%|▏ | 1038/48008 [03:30<2:41:28, 4.85it/s] 2%|▏ | 1039/48008 [03:30<2:37:41, 4.96it/s] 2%|▏ | 1040/48008 [03:30<2:34:41, 5.06it/s] 2%|▏ | 1041/48008 [03:30<2:32:53, 5.12it/s] 2%|▏ | 1042/48008 [03:31<2:31:23, 5.17it/s] 2%|▏ | 1043/48008 [03:31<2:32:45, 5.12it/s] 2%|▏ | 1044/48008 [03:31<2:31:46, 5.16it/s] 2%|▏ | 1045/48008 [03:31<2:35:11, 5.04it/s] 2%|▏ | 1046/48008 [03:31<2:32:25, 5.13it/s] 2%|▏ | 1047/48008 [03:32<2:32:48, 5.12it/s] 2%|▏ | 1048/48008 [03:32<2:30:15, 5.21it/s] 2%|▏ | 1049/48008 [03:32<2:29:38, 5.23it/s] 2%|▏ | 1050/48008 [03:32<2:29:01, 5.25it/s] {'loss': 0.3187, 'grad_norm': 0.2314453125, 'learning_rate': 0.00019562989501749708, 'epoch': 0.02} 2%|▏ | 1050/48008 [03:32<2:29:01, 5.25it/s] 2%|▏ | 1051/48008 [03:32<2:29:24, 5.24it/s] 2%|▏ | 1052/48008 [03:33<2:28:54, 5.26it/s] 2%|▏ | 1053/48008 [03:33<2:28:52, 5.26it/s] 2%|▏ | 1054/48008 [03:34<6:27:38, 2.02it/s] 2%|▏ | 1055/48008 [03:34<5:16:31, 2.47it/s] 2%|▏ | 1056/48008 [03:34<4:25:29, 2.95it/s] 2%|▏ | 1057/48008 [03:34<3:50:23, 3.40it/s] 2%|▏ | 1058/48008 [03:35<3:27:42, 3.77it/s] 2%|▏ | 1059/48008 [03:35<3:10:07, 4.12it/s] 2%|▏ | 1060/48008 [03:35<2:57:40, 4.40it/s] 2%|▏ | 1061/48008 [03:35<2:50:13, 4.60it/s] 2%|▏ | 1062/48008 [03:35<2:46:52, 4.69it/s] 2%|▏ | 1063/48008 [03:36<2:41:18, 4.85it/s] 2%|▏ | 1064/48008 [03:36<2:36:50, 4.99it/s] 2%|▏ | 1065/48008 [03:36<2:34:03, 5.08it/s] 2%|▏ | 1066/48008 [03:36<2:31:48, 5.15it/s] 2%|▏ | 1067/48008 [03:36<2:41:29, 4.84it/s] 2%|▏ | 1068/48008 [03:37<2:38:08, 4.95it/s] 2%|▏ | 1069/48008 [03:37<2:36:25, 5.00it/s] 2%|▏ | 1070/48008 [03:37<2:36:06, 5.01it/s] 2%|▏ | 1071/48008 [03:37<2:36:43, 4.99it/s] 2%|▏ | 1072/48008 [03:37<2:45:26, 4.73it/s] 2%|▏ | 1073/48008 [03:38<2:44:45, 4.75it/s] 2%|▏ | 1074/48008 [03:38<2:39:12, 4.91it/s] 2%|▏ | 1075/48008 [03:38<2:36:30, 5.00it/s] 2%|▏ | 1076/48008 [03:38<2:34:38, 5.06it/s] 2%|▏ | 1077/48008 [03:38<2:32:58, 5.11it/s] 2%|▏ | 1078/48008 [03:39<2:31:45, 5.15it/s] 2%|▏ | 1079/48008 [03:39<2:30:24, 5.20it/s] 2%|▏ | 1080/48008 [03:39<2:41:09, 4.85it/s] 2%|▏ | 1081/48008 [03:39<2:39:25, 4.91it/s] 2%|▏ | 1082/48008 [03:39<2:38:32, 4.93it/s] 2%|▏ | 1083/48008 [03:40<3:21:33, 3.88it/s] 2%|▏ | 1084/48008 [03:40<3:16:48, 3.97it/s] 2%|▏ | 1085/48008 [03:40<3:01:11, 4.32it/s] 2%|▏ | 1086/48008 [03:40<3:02:30, 4.29it/s] 2%|▏ | 1087/48008 [03:41<2:52:54, 4.52it/s] 2%|▏ | 1088/48008 [03:41<2:46:23, 4.70it/s] 2%|▏ | 1089/48008 [03:41<2:43:49, 4.77it/s] 2%|▏ | 1090/48008 [03:41<2:40:12, 4.88it/s] 2%|▏ | 1091/48008 [03:42<3:03:58, 4.25it/s] 2%|▏ | 1092/48008 [03:42<2:56:11, 4.44it/s] 2%|▏ | 1093/48008 [03:42<2:47:29, 4.67it/s] 2%|▏ | 1094/48008 [03:42<2:43:33, 4.78it/s] 2%|▏ | 1095/48008 [03:42<2:39:03, 4.92it/s] 2%|▏ | 1096/48008 [03:43<2:35:51, 5.02it/s] 2%|▏ | 1097/48008 [03:43<2:32:58, 5.11it/s] 2%|▏ | 1098/48008 [03:43<2:32:05, 5.14it/s] 2%|▏ | 1099/48008 [03:43<2:33:01, 5.11it/s] 2%|▏ | 1100/48008 [03:43<2:36:11, 5.01it/s] {'loss': 0.316, 'grad_norm': 1.0703125, 'learning_rate': 0.00019542159640059992, 'epoch': 0.02} 2%|▏ | 1100/48008 [03:43<2:36:11, 5.01it/s] 2%|▏ | 1101/48008 [03:44<2:35:03, 5.04it/s] 2%|▏ | 1102/48008 [03:44<2:31:58, 5.14it/s] 2%|▏ | 1103/48008 [03:44<2:30:33, 5.19it/s] 2%|▏ | 1104/48008 [03:44<2:32:41, 5.12it/s] 2%|▏ | 1105/48008 [03:44<2:31:36, 5.16it/s] 2%|▏ | 1106/48008 [03:44<2:30:59, 5.18it/s] 2%|▏ | 1107/48008 [03:45<2:32:36, 5.12it/s] 2%|▏ | 1108/48008 [03:45<2:31:19, 5.17it/s] 2%|▏ | 1109/48008 [03:45<2:31:19, 5.17it/s] 2%|▏ | 1110/48008 [03:45<2:32:15, 5.13it/s] 2%|▏ | 1111/48008 [03:45<2:30:47, 5.18it/s] 2%|▏ | 1112/48008 [03:46<2:30:16, 5.20it/s] 2%|▏ | 1113/48008 [03:46<2:30:06, 5.21it/s] 2%|▏ | 1114/48008 [03:46<2:29:57, 5.21it/s] 2%|▏ | 1115/48008 [03:46<2:29:51, 5.22it/s] 2%|▏ | 1116/48008 [03:46<2:29:21, 5.23it/s] 2%|▏ | 1117/48008 [03:47<2:29:46, 5.22it/s] 2%|▏ | 1118/48008 [03:47<2:32:07, 5.14it/s] 2%|▏ | 1119/48008 [03:47<2:29:28, 5.23it/s] 2%|▏ | 1120/48008 [03:47<2:28:12, 5.27it/s] 2%|▏ | 1121/48008 [03:47<2:28:04, 5.28it/s] 2%|▏ | 1122/48008 [03:48<2:27:56, 5.28it/s] 2%|▏ | 1123/48008 [03:48<2:39:04, 4.91it/s] 2%|▏ | 1124/48008 [03:48<2:39:51, 4.89it/s] 2%|▏ | 1125/48008 [03:48<2:36:07, 5.01it/s] 2%|▏ | 1126/48008 [03:48<2:35:38, 5.02it/s] 2%|▏ | 1127/48008 [03:49<2:34:36, 5.05it/s] 2%|▏ | 1128/48008 [03:49<3:00:07, 4.34it/s] 2%|▏ | 1129/48008 [03:49<2:50:19, 4.59it/s] 2%|▏ | 1130/48008 [03:49<2:43:14, 4.79it/s] 2%|▏ | 1131/48008 [03:49<2:37:55, 4.95it/s] 2%|▏ | 1132/48008 [03:50<2:37:43, 4.95it/s] 2%|▏ | 1133/48008 [03:50<2:34:39, 5.05it/s] 2%|▏ | 1134/48008 [03:50<2:33:39, 5.08it/s] 2%|▏ | 1135/48008 [03:50<3:17:58, 3.95it/s] 2%|▏ | 1136/48008 [03:51<3:03:31, 4.26it/s] 2%|▏ | 1137/48008 [03:51<2:54:45, 4.47it/s] 2%|▏ | 1138/48008 [03:51<2:47:36, 4.66it/s] 2%|▏ | 1139/48008 [03:51<2:42:51, 4.80it/s] 2%|▏ | 1140/48008 [03:51<2:37:46, 4.95it/s] 2%|▏ | 1141/48008 [03:52<2:34:28, 5.06it/s] 2%|▏ | 1142/48008 [03:52<2:43:37, 4.77it/s] 2%|▏ | 1143/48008 [03:52<2:40:49, 4.86it/s] 2%|▏ | 1144/48008 [03:52<2:40:03, 4.88it/s] 2%|▏ | 1145/48008 [03:52<2:37:02, 4.97it/s] 2%|▏ | 1146/48008 [03:53<2:36:52, 4.98it/s] 2%|▏ | 1147/48008 [03:53<2:37:24, 4.96it/s] 2%|▏ | 1148/48008 [03:53<2:37:22, 4.96it/s] 2%|▏ | 1149/48008 [03:53<2:34:42, 5.05it/s] 2%|▏ | 1150/48008 [03:53<2:32:44, 5.11it/s] {'loss': 0.314, 'grad_norm': 0.43359375, 'learning_rate': 0.00019521329778370271, 'epoch': 0.02} 2%|▏ | 1150/48008 [03:53<2:32:44, 5.11it/s] 2%|▏ | 1151/48008 [03:54<2:34:48, 5.04it/s] 2%|▏ | 1152/48008 [03:54<2:44:07, 4.76it/s] 2%|▏ | 1153/48008 [03:54<2:38:45, 4.92it/s] 2%|▏ | 1154/48008 [03:54<2:36:26, 4.99it/s] 2%|▏ | 1155/48008 [03:54<2:33:00, 5.10it/s] 2%|▏ | 1156/48008 [03:55<2:35:02, 5.04it/s] 2%|▏ | 1157/48008 [03:55<2:31:48, 5.14it/s] 2%|▏ | 1158/48008 [03:55<2:30:00, 5.21it/s] 2%|▏ | 1159/48008 [03:55<2:28:50, 5.25it/s] 2%|▏ | 1160/48008 [03:55<2:27:49, 5.28it/s] 2%|▏ | 1161/48008 [03:56<2:30:02, 5.20it/s] 2%|▏ | 1162/48008 [03:56<2:29:21, 5.23it/s] 2%|▏ | 1163/48008 [03:56<2:27:57, 5.28it/s] 2%|▏ | 1164/48008 [03:56<2:25:47, 5.36it/s] 2%|▏ | 1165/48008 [03:56<2:25:27, 5.37it/s] 2%|▏ | 1166/48008 [03:56<2:30:49, 5.18it/s] 2%|▏ | 1167/48008 [03:57<3:15:43, 3.99it/s] 2%|▏ | 1168/48008 [03:57<3:02:05, 4.29it/s] 2%|▏ | 1169/48008 [03:57<2:54:11, 4.48it/s] 2%|▏ | 1170/48008 [03:57<2:47:04, 4.67it/s] 2%|▏ | 1171/48008 [03:58<2:41:34, 4.83it/s] 2%|▏ | 1172/48008 [03:58<2:36:13, 5.00it/s] 2%|▏ | 1173/48008 [03:58<2:34:12, 5.06it/s] 2%|▏ | 1174/48008 [03:58<2:33:58, 5.07it/s] 2%|▏ | 1175/48008 [03:58<2:34:12, 5.06it/s] 2%|▏ | 1176/48008 [03:59<2:32:28, 5.12it/s] 2%|▏ | 1177/48008 [03:59<2:31:17, 5.16it/s] 2%|▏ | 1178/48008 [03:59<2:30:29, 5.19it/s] 2%|▏ | 1179/48008 [03:59<2:32:49, 5.11it/s] 2%|▏ | 1180/48008 [03:59<2:31:42, 5.14it/s] 2%|▏ | 1181/48008 [04:00<2:41:24, 4.84it/s] 2%|▏ | 1182/48008 [04:00<2:37:32, 4.95it/s] 2%|▏ | 1183/48008 [04:00<2:36:23, 4.99it/s] 2%|▏ | 1184/48008 [04:00<2:34:29, 5.05it/s] 2%|▏ | 1185/48008 [04:01<2:59:49, 4.34it/s] 2%|▏ | 1186/48008 [04:01<2:48:59, 4.62it/s] 2%|▏ | 1187/48008 [04:01<2:42:51, 4.79it/s] 2%|▏ | 1188/48008 [04:01<2:38:39, 4.92it/s] 2%|▏ | 1189/48008 [04:01<2:35:52, 5.01it/s] 2%|▏ | 1190/48008 [04:01<2:31:58, 5.13it/s] 2%|▏ | 1191/48008 [04:02<2:31:54, 5.14it/s] 2%|▏ | 1192/48008 [04:02<2:30:09, 5.20it/s] 2%|▏ | 1193/48008 [04:02<2:27:59, 5.27it/s] 2%|▏ | 1194/48008 [04:02<2:28:39, 5.25it/s] 2%|▏ | 1195/48008 [04:02<2:30:26, 5.19it/s] 2%|▏ | 1196/48008 [04:03<2:29:21, 5.22it/s] 2%|▏ | 1197/48008 [04:03<2:28:21, 5.26it/s] 2%|▏ | 1198/48008 [04:03<2:30:20, 5.19it/s] 2%|▏ | 1199/48008 [04:03<2:30:09, 5.20it/s] 2%|▏ | 1200/48008 [04:04<3:15:36, 3.99it/s] {'loss': 0.3137, 'grad_norm': 0.205078125, 'learning_rate': 0.00019500499916680556, 'epoch': 0.02} 2%|▏ | 1200/48008 [04:04<3:15:36, 3.99it/s] 3%|▎ | 1201/48008 [04:04<3:01:00, 4.31it/s] 3%|▎ | 1202/48008 [04:04<3:02:15, 4.28it/s] 3%|▎ | 1203/48008 [04:04<2:56:14, 4.43it/s] 3%|▎ | 1204/48008 [04:04<2:49:06, 4.61it/s] 3%|▎ | 1205/48008 [04:05<2:42:53, 4.79it/s] 3%|▎ | 1206/48008 [04:05<2:37:24, 4.96it/s] 3%|▎ | 1207/48008 [04:05<3:01:49, 4.29it/s] 3%|▎ | 1208/48008 [04:05<2:52:25, 4.52it/s] 3%|▎ | 1209/48008 [04:05<2:46:01, 4.70it/s] 3%|▎ | 1210/48008 [04:06<2:51:48, 4.54it/s] 3%|▎ | 1211/48008 [04:06<2:44:56, 4.73it/s] 3%|▎ | 1212/48008 [04:06<2:40:40, 4.85it/s] 3%|▎ | 1213/48008 [04:06<2:38:03, 4.93it/s] 3%|▎ | 1214/48008 [04:06<2:35:08, 5.03it/s] 3%|▎ | 1215/48008 [04:07<2:37:42, 4.94it/s] 3%|▎ | 1216/48008 [04:07<2:37:40, 4.95it/s] 3%|▎ | 1217/48008 [04:07<2:35:02, 5.03it/s] 3%|▎ | 1218/48008 [04:07<2:33:24, 5.08it/s] 3%|▎ | 1219/48008 [04:07<2:30:57, 5.17it/s] 3%|▎ | 1220/48008 [04:08<2:30:11, 5.19it/s] 3%|▎ | 1221/48008 [04:08<2:27:11, 5.30it/s] 3%|▎ | 1222/48008 [04:08<2:27:56, 5.27it/s] 3%|▎ | 1223/48008 [04:08<2:29:57, 5.20it/s] 3%|▎ | 1224/48008 [04:08<2:29:32, 5.21it/s] 3%|▎ | 1225/48008 [04:09<2:28:02, 5.27it/s] 3%|▎ | 1226/48008 [04:09<2:27:38, 5.28it/s] 3%|▎ | 1227/48008 [04:09<2:27:41, 5.28it/s] 3%|▎ | 1228/48008 [04:09<2:55:06, 4.45it/s] 3%|▎ | 1229/48008 [04:09<2:46:46, 4.67it/s] 3%|▎ | 1230/48008 [04:10<2:41:06, 4.84it/s] 3%|▎ | 1231/48008 [04:10<2:38:54, 4.91it/s] 3%|▎ | 1232/48008 [04:10<2:35:55, 5.00it/s] 3%|▎ | 1233/48008 [04:10<2:33:37, 5.07it/s] 3%|▎ | 1234/48008 [04:10<2:34:55, 5.03it/s] 3%|▎ | 1235/48008 [04:11<2:32:56, 5.10it/s] 3%|▎ | 1236/48008 [04:11<2:30:43, 5.17it/s] 3%|▎ | 1237/48008 [04:11<2:30:43, 5.17it/s] 3%|▎ | 1238/48008 [04:12<5:11:31, 2.50it/s] 3%|▎ | 1239/48008 [04:12<4:24:06, 2.95it/s] 3%|▎ | 1240/48008 [04:12<3:49:30, 3.40it/s] 3%|▎ | 1241/48008 [04:12<3:25:42, 3.79it/s] 3%|▎ | 1242/48008 [04:13<3:07:45, 4.15it/s] 3%|▎ | 1243/48008 [04:13<2:56:01, 4.43it/s] 3%|▎ | 1244/48008 [04:13<2:47:40, 4.65it/s] 3%|▎ | 1245/48008 [04:13<2:41:59, 4.81it/s] 3%|▎ | 1246/48008 [04:13<2:36:59, 4.96it/s] 3%|▎ | 1247/48008 [04:14<2:45:57, 4.70it/s] 3%|▎ | 1248/48008 [04:14<2:41:19, 4.83it/s] 3%|▎ | 1249/48008 [04:14<2:38:13, 4.93it/s] 3%|▎ | 1250/48008 [04:14<2:36:51, 4.97it/s] {'loss': 0.3141, 'grad_norm': 0.6171875, 'learning_rate': 0.00019479670054990835, 'epoch': 0.03} 3%|▎ | 1250/48008 [04:14<2:36:51, 4.97it/s] 3%|▎ | 1251/48008 [04:14<2:33:38, 5.07it/s] 3%|▎ | 1252/48008 [04:15<2:31:39, 5.14it/s] 3%|▎ | 1253/48008 [04:15<2:33:02, 5.09it/s] 3%|▎ | 1254/48008 [04:15<2:31:34, 5.14it/s] 3%|▎ | 1255/48008 [04:15<2:32:13, 5.12it/s] 3%|▎ | 1256/48008 [04:15<2:31:06, 5.16it/s] 3%|▎ | 1257/48008 [04:16<2:32:12, 5.12it/s] 3%|▎ | 1258/48008 [04:16<2:30:44, 5.17it/s] 3%|▎ | 1259/48008 [04:16<2:29:14, 5.22it/s] 3%|▎ | 1260/48008 [04:16<2:28:18, 5.25it/s] 3%|▎ | 1261/48008 [04:16<2:26:49, 5.31it/s] 3%|▎ | 1262/48008 [04:17<2:27:18, 5.29it/s] 3%|▎ | 1263/48008 [04:17<2:30:35, 5.17it/s] 3%|▎ | 1264/48008 [04:17<2:29:38, 5.21it/s] 3%|▎ | 1265/48008 [04:17<2:34:24, 5.05it/s] 3%|▎ | 1266/48008 [04:17<2:33:28, 5.08it/s] 3%|▎ | 1267/48008 [04:17<2:30:49, 5.17it/s] 3%|▎ | 1268/48008 [04:18<3:15:34, 3.98it/s] 3%|▎ | 1269/48008 [04:18<3:00:56, 4.31it/s] 3%|▎ | 1270/48008 [04:18<2:52:24, 4.52it/s] 3%|▎ | 1271/48008 [04:18<2:44:18, 4.74it/s] 3%|▎ | 1272/48008 [04:19<2:39:54, 4.87it/s] 3%|▎ | 1273/48008 [04:19<2:36:54, 4.96it/s] 3%|▎ | 1274/48008 [04:19<2:34:04, 5.06it/s] 3%|▎ | 1275/48008 [04:19<2:31:48, 5.13it/s] 3%|▎ | 1276/48008 [04:19<2:31:19, 5.15it/s] 3%|▎ | 1277/48008 [04:20<2:29:03, 5.23it/s] 3%|▎ | 1278/48008 [04:20<2:28:49, 5.23it/s] 3%|▎ | 1279/48008 [04:20<2:28:48, 5.23it/s] 3%|▎ | 1280/48008 [04:20<2:29:54, 5.20it/s] 3%|▎ | 1281/48008 [04:20<2:29:33, 5.21it/s] 3%|▎ | 1282/48008 [04:21<2:29:20, 5.21it/s] 3%|▎ | 1283/48008 [04:21<2:28:11, 5.25it/s] 3%|▎ | 1284/48008 [04:21<2:28:17, 5.25it/s] 3%|▎ | 1285/48008 [04:21<2:28:37, 5.24it/s] 3%|▎ | 1286/48008 [04:21<2:28:48, 5.23it/s] 3%|▎ | 1287/48008 [04:22<2:27:33, 5.28it/s] 3%|▎ | 1288/48008 [04:22<2:28:03, 5.26it/s] 3%|▎ | 1289/48008 [04:22<2:30:11, 5.18it/s] 3%|▎ | 1290/48008 [04:22<2:32:38, 5.10it/s] 3%|▎ | 1291/48008 [04:22<2:30:58, 5.16it/s] 3%|▎ | 1292/48008 [04:22<2:29:45, 5.20it/s] 3%|▎ | 1293/48008 [04:23<2:32:47, 5.10it/s] 3%|▎ | 1294/48008 [04:23<2:31:10, 5.15it/s] 3%|▎ | 1295/48008 [04:23<2:30:09, 5.19it/s] 3%|▎ | 1296/48008 [04:23<2:29:05, 5.22it/s] 3%|▎ | 1297/48008 [04:23<2:31:05, 5.15it/s] 3%|▎ | 1298/48008 [04:24<2:31:33, 5.14it/s] 3%|▎ | 1299/48008 [04:24<2:33:34, 5.07it/s] 3%|▎ | 1300/48008 [04:24<2:32:29, 5.10it/s] {'loss': 0.3112, 'grad_norm': 0.2470703125, 'learning_rate': 0.00019458840193301117, 'epoch': 0.03} 3%|▎ | 1300/48008 [04:24<2:32:29, 5.10it/s] 3%|▎ | 1301/48008 [04:24<2:32:00, 5.12it/s] 3%|▎ | 1302/48008 [04:24<2:31:49, 5.13it/s] 3%|▎ | 1303/48008 [04:25<2:30:16, 5.18it/s] 3%|▎ | 1304/48008 [04:25<2:28:48, 5.23it/s] 3%|▎ | 1305/48008 [04:25<2:27:23, 5.28it/s] 3%|▎ | 1306/48008 [04:25<2:38:13, 4.92it/s] 3%|▎ | 1307/48008 [04:25<2:34:49, 5.03it/s] 3%|▎ | 1308/48008 [04:26<2:32:28, 5.10it/s] 3%|▎ | 1309/48008 [04:26<2:32:47, 5.09it/s] 3%|▎ | 1310/48008 [04:26<2:33:56, 5.06it/s] 3%|▎ | 1311/48008 [04:26<2:31:11, 5.15it/s] 3%|▎ | 1312/48008 [04:26<2:31:52, 5.12it/s] 3%|▎ | 1313/48008 [04:27<2:31:12, 5.15it/s] 3%|▎ | 1314/48008 [04:27<2:30:18, 5.18it/s] 3%|▎ | 1315/48008 [04:27<2:29:43, 5.20it/s] 3%|▎ | 1316/48008 [04:27<2:28:41, 5.23it/s] 3%|▎ | 1317/48008 [04:27<2:30:25, 5.17it/s] 3%|▎ | 1318/48008 [04:28<2:32:50, 5.09it/s] 3%|▎ | 1319/48008 [04:28<2:33:56, 5.05it/s] 3%|▎ | 1320/48008 [04:28<2:31:16, 5.14it/s] 3%|▎ | 1321/48008 [04:28<2:30:42, 5.16it/s] 3%|▎ | 1322/48008 [04:28<2:32:33, 5.10it/s] 3%|▎ | 1323/48008 [04:29<2:35:47, 4.99it/s] 3%|▎ | 1324/48008 [04:29<2:33:44, 5.06it/s] 3%|▎ | 1325/48008 [04:29<2:32:20, 5.11it/s] 3%|▎ | 1326/48008 [04:29<2:30:04, 5.18it/s] 3%|▎ | 1327/48008 [04:29<2:31:38, 5.13it/s] 3%|▎ | 1328/48008 [04:29<2:31:05, 5.15it/s] 3%|▎ | 1329/48008 [04:30<2:30:51, 5.16it/s] 3%|▎ | 1330/48008 [04:30<3:15:35, 3.98it/s] 3%|▎ | 1331/48008 [04:30<3:01:24, 4.29it/s] 3%|▎ | 1332/48008 [04:30<2:55:48, 4.42it/s] 3%|▎ | 1333/48008 [04:31<2:47:16, 4.65it/s] 3%|▎ | 1334/48008 [04:31<2:43:05, 4.77it/s] 3%|▎ | 1335/48008 [04:31<2:38:26, 4.91it/s] 3%|▎ | 1336/48008 [04:31<2:37:49, 4.93it/s] 3%|▎ | 1337/48008 [04:31<2:35:12, 5.01it/s] 3%|▎ | 1338/48008 [04:32<2:32:18, 5.11it/s] 3%|▎ | 1339/48008 [04:32<2:32:28, 5.10it/s] 3%|▎ | 1340/48008 [04:32<2:31:28, 5.13it/s] 3%|▎ | 1341/48008 [04:32<2:29:59, 5.19it/s] 3%|▎ | 1342/48008 [04:32<2:28:47, 5.23it/s] 3%|▎ | 1343/48008 [04:33<2:27:54, 5.26it/s] 3%|▎ | 1344/48008 [04:33<2:27:35, 5.27it/s] 3%|▎ | 1345/48008 [04:33<3:13:03, 4.03it/s] 3%|▎ | 1346/48008 [04:33<3:00:05, 4.32it/s] 3%|▎ | 1347/48008 [04:34<2:53:47, 4.47it/s] 3%|▎ | 1348/48008 [04:34<2:45:38, 4.70it/s] 3%|▎ | 1349/48008 [04:34<2:40:35, 4.84it/s] 3%|▎ | 1350/48008 [04:34<2:39:35, 4.87it/s] {'loss': 0.3103, 'grad_norm': 1.65625, 'learning_rate': 0.00019438010331611398, 'epoch': 0.03} 3%|▎ | 1350/48008 [04:34<2:39:35, 4.87it/s] 3%|▎ | 1351/48008 [04:34<2:39:22, 4.88it/s] 3%|▎ | 1352/48008 [04:35<2:35:05, 5.01it/s] 3%|▎ | 1353/48008 [04:35<2:35:36, 5.00it/s] 3%|▎ | 1354/48008 [04:35<2:32:11, 5.11it/s] 3%|▎ | 1355/48008 [04:35<2:41:52, 4.80it/s] 3%|▎ | 1356/48008 [04:35<2:37:31, 4.94it/s] 3%|▎ | 1357/48008 [04:36<2:35:45, 4.99it/s] 3%|▎ | 1358/48008 [04:36<2:34:58, 5.02it/s] 3%|▎ | 1359/48008 [04:36<2:33:31, 5.06it/s] 3%|▎ | 1360/48008 [04:36<2:32:30, 5.10it/s] 3%|▎ | 1361/48008 [04:36<2:30:07, 5.18it/s] 3%|▎ | 1362/48008 [04:37<2:32:26, 5.10it/s] 3%|▎ | 1363/48008 [04:37<2:31:42, 5.12it/s] 3%|▎ | 1364/48008 [04:37<2:30:40, 5.16it/s] 3%|▎ | 1365/48008 [04:37<2:29:26, 5.20it/s] 3%|▎ | 1366/48008 [04:37<2:29:07, 5.21it/s] 3%|▎ | 1367/48008 [04:37<2:28:38, 5.23it/s] 3%|▎ | 1368/48008 [04:38<2:29:57, 5.18it/s] 3%|▎ | 1369/48008 [04:38<2:32:05, 5.11it/s] 3%|▎ | 1370/48008 [04:38<2:30:05, 5.18it/s] 3%|▎ | 1371/48008 [04:38<2:30:40, 5.16it/s] 3%|▎ | 1372/48008 [04:38<2:28:37, 5.23it/s] 3%|▎ | 1373/48008 [04:39<2:28:48, 5.22it/s] 3%|▎ | 1374/48008 [04:39<2:30:58, 5.15it/s] 3%|▎ | 1375/48008 [04:39<2:29:34, 5.20it/s] 3%|▎ | 1376/48008 [04:39<2:33:01, 5.08it/s] 3%|▎ | 1377/48008 [04:39<2:32:40, 5.09it/s] 3%|▎ | 1378/48008 [04:40<2:31:36, 5.13it/s] 3%|▎ | 1379/48008 [04:40<2:30:31, 5.16it/s] 3%|▎ | 1380/48008 [04:40<2:29:26, 5.20it/s] 3%|▎ | 1381/48008 [04:40<2:28:23, 5.24it/s] 3%|▎ | 1382/48008 [04:40<2:27:09, 5.28it/s] 3%|▎ | 1383/48008 [04:41<2:29:09, 5.21it/s] 3%|▎ | 1384/48008 [04:41<2:29:45, 5.19it/s] 3%|▎ | 1385/48008 [04:41<2:40:04, 4.85it/s] 3%|▎ | 1386/48008 [04:41<2:40:32, 4.84it/s] 3%|▎ | 1387/48008 [04:41<2:36:47, 4.96it/s] 3%|▎ | 1388/48008 [04:42<2:33:31, 5.06it/s] 3%|▎ | 1389/48008 [04:42<2:31:58, 5.11it/s] 3%|▎ | 1390/48008 [04:42<2:41:44, 4.80it/s] 3%|▎ | 1391/48008 [04:42<2:38:08, 4.91it/s] 3%|▎ | 1392/48008 [04:42<2:36:59, 4.95it/s] 3%|▎ | 1393/48008 [04:43<2:32:59, 5.08it/s] 3%|▎ | 1394/48008 [04:43<2:31:33, 5.13it/s] 3%|▎ | 1395/48008 [04:43<2:29:15, 5.20it/s] 3%|▎ | 1396/48008 [04:43<2:31:32, 5.13it/s] 3%|▎ | 1397/48008 [04:43<2:33:55, 5.05it/s] 3%|▎ | 1398/48008 [04:44<2:35:13, 5.00it/s] 3%|▎ | 1399/48008 [04:44<2:32:34, 5.09it/s] 3%|▎ | 1400/48008 [04:44<2:33:59, 5.04it/s] {'loss': 0.311, 'grad_norm': 0.1875, 'learning_rate': 0.0001941718046992168, 'epoch': 0.03} 3%|▎ | 1400/48008 [04:44<2:33:59, 5.04it/s] 3%|▎ | 1401/48008 [04:44<2:33:17, 5.07it/s] 3%|▎ | 1402/48008 [04:44<2:31:35, 5.12it/s] 3%|▎ | 1403/48008 [04:45<2:30:26, 5.16it/s] 3%|▎ | 1404/48008 [04:45<2:32:26, 5.10it/s] 3%|▎ | 1405/48008 [04:45<2:35:07, 5.01it/s] 3%|▎ | 1406/48008 [04:45<2:43:39, 4.75it/s] 3%|▎ | 1407/48008 [04:45<3:05:54, 4.18it/s] 3%|▎ | 1408/48008 [04:46<2:54:13, 4.46it/s] 3%|▎ | 1409/48008 [04:46<2:46:42, 4.66it/s] 3%|▎ | 1410/48008 [04:46<2:39:26, 4.87it/s] 3%|▎ | 1411/48008 [04:46<2:36:03, 4.98it/s] 3%|▎ | 1412/48008 [04:46<2:33:39, 5.05it/s] 3%|▎ | 1413/48008 [04:47<3:17:05, 3.94it/s] 3%|▎ | 1414/48008 [04:47<3:03:43, 4.23it/s] 3%|▎ | 1415/48008 [04:47<2:53:21, 4.48it/s] 3%|▎ | 1416/48008 [04:47<2:45:10, 4.70it/s] 3%|▎ | 1417/48008 [04:48<2:39:00, 4.88it/s] 3%|▎ | 1418/48008 [04:48<2:35:51, 4.98it/s] 3%|▎ | 1419/48008 [04:48<2:33:31, 5.06it/s] 3%|▎ | 1420/48008 [04:48<2:31:06, 5.14it/s] 3%|▎ | 1421/48008 [04:48<2:30:31, 5.16it/s] 3%|▎ | 1422/48008 [04:49<3:14:52, 3.98it/s] 3%|▎ | 1423/48008 [04:49<3:00:45, 4.30it/s] 3%|▎ | 1424/48008 [04:49<2:50:21, 4.56it/s] 3%|▎ | 1425/48008 [04:49<2:54:09, 4.46it/s] 3%|▎ | 1426/48008 [04:50<2:47:39, 4.63it/s] 3%|▎ | 1427/48008 [04:50<3:08:39, 4.11it/s] 3%|▎ | 1428/48008 [04:50<2:58:51, 4.34it/s] 3%|▎ | 1429/48008 [04:50<2:49:28, 4.58it/s] 3%|▎ | 1430/48008 [04:50<2:45:37, 4.69it/s] 3%|▎ | 1431/48008 [04:51<2:42:23, 4.78it/s] 3%|▎ | 1432/48008 [04:51<2:38:01, 4.91it/s] 3%|▎ | 1433/48008 [04:51<3:01:50, 4.27it/s] 3%|▎ | 1434/48008 [04:51<3:02:29, 4.25it/s] 3%|▎ | 1435/48008 [04:52<2:52:43, 4.49it/s] 3%|▎ | 1436/48008 [04:52<2:48:04, 4.62it/s] 3%|▎ | 1437/48008 [04:52<3:27:38, 3.74it/s] 3%|▎ | 1438/48008 [04:52<3:11:17, 4.06it/s] 3%|▎ | 1439/48008 [04:53<2:57:53, 4.36it/s] 3%|▎ | 1440/48008 [04:53<2:50:09, 4.56it/s] 3%|▎ | 1441/48008 [04:53<2:44:05, 4.73it/s] 3%|▎ | 1442/48008 [04:53<2:38:13, 4.91it/s] 3%|▎ | 1443/48008 [04:53<2:35:20, 5.00it/s] 3%|▎ | 1444/48008 [04:53<2:33:57, 5.04it/s] 3%|▎ | 1445/48008 [04:54<2:32:00, 5.11it/s] 3%|▎ | 1446/48008 [04:54<2:33:05, 5.07it/s] 3%|▎ | 1447/48008 [04:54<2:31:11, 5.13it/s] 3%|▎ | 1448/48008 [04:54<2:31:12, 5.13it/s] 3%|▎ | 1449/48008 [04:54<2:33:22, 5.06it/s] 3%|▎ | 1450/48008 [04:55<2:30:46, 5.15it/s] {'loss': 0.3076, 'grad_norm': 0.205078125, 'learning_rate': 0.00019396350608231962, 'epoch': 0.03} 3%|▎ | 1450/48008 [04:55<2:30:46, 5.15it/s] 3%|▎ | 1451/48008 [04:55<2:28:40, 5.22it/s] 3%|▎ | 1452/48008 [04:55<2:31:03, 5.14it/s] 3%|▎ | 1453/48008 [04:55<2:29:46, 5.18it/s] 3%|▎ | 1454/48008 [04:55<2:30:41, 5.15it/s] 3%|▎ | 1455/48008 [04:56<2:30:43, 5.15it/s] 3%|▎ | 1456/48008 [04:56<2:26:53, 5.28it/s] 3%|▎ | 1457/48008 [04:56<2:26:23, 5.30it/s] 3%|▎ | 1458/48008 [04:56<2:26:53, 5.28it/s] 3%|▎ | 1459/48008 [04:56<2:27:39, 5.25it/s] 3%|▎ | 1460/48008 [04:57<2:26:43, 5.29it/s] 3%|▎ | 1461/48008 [04:57<2:25:49, 5.32it/s] 3%|▎ | 1462/48008 [04:57<2:26:02, 5.31it/s] 3%|▎ | 1463/48008 [04:57<2:26:23, 5.30it/s] 3%|▎ | 1464/48008 [04:57<2:28:29, 5.22it/s] 3%|▎ | 1465/48008 [04:58<3:13:48, 4.00it/s] 3%|▎ | 1466/48008 [04:58<3:00:18, 4.30it/s] 3%|▎ | 1467/48008 [04:58<2:50:05, 4.56it/s] 3%|▎ | 1468/48008 [04:58<2:45:57, 4.67it/s] 3%|▎ | 1469/48008 [04:58<2:40:54, 4.82it/s] 3%|▎ | 1470/48008 [04:59<2:37:13, 4.93it/s] 3%|▎ | 1471/48008 [04:59<2:33:18, 5.06it/s] 3%|▎ | 1472/48008 [04:59<2:32:06, 5.10it/s] 3%|▎ | 1473/48008 [04:59<2:32:44, 5.08it/s] 3%|▎ | 1474/48008 [04:59<2:30:52, 5.14it/s] 3%|▎ | 1475/48008 [05:00<2:29:00, 5.20it/s] 3%|▎ | 1476/48008 [05:00<2:27:08, 5.27it/s] 3%|▎ | 1477/48008 [05:00<2:30:04, 5.17it/s] 3%|▎ | 1478/48008 [05:00<2:32:43, 5.08it/s] 3%|▎ | 1479/48008 [05:00<2:31:30, 5.12it/s] 3%|▎ | 1480/48008 [05:01<2:32:54, 5.07it/s] 3%|▎ | 1481/48008 [05:01<2:32:52, 5.07it/s] 3%|▎ | 1482/48008 [05:01<2:31:42, 5.11it/s] 3%|▎ | 1483/48008 [05:01<2:29:58, 5.17it/s] 3%|▎ | 1484/48008 [05:01<2:29:44, 5.18it/s] 3%|▎ | 1485/48008 [05:02<2:31:14, 5.13it/s] 3%|▎ | 1486/48008 [05:02<2:30:18, 5.16it/s] 3%|▎ | 1487/48008 [05:02<2:32:36, 5.08it/s] 3%|▎ | 1488/48008 [05:02<2:33:18, 5.06it/s] 3%|▎ | 1489/48008 [05:02<2:34:23, 5.02it/s] 3%|▎ | 1490/48008 [05:03<2:32:11, 5.09it/s] 3%|▎ | 1491/48008 [05:03<2:30:26, 5.15it/s] 3%|▎ | 1492/48008 [05:03<2:31:13, 5.13it/s] 3%|▎ | 1493/48008 [05:03<2:30:20, 5.16it/s] 3%|▎ | 1494/48008 [05:03<2:29:44, 5.18it/s] 3%|▎ | 1495/48008 [05:04<2:28:20, 5.23it/s] 3%|▎ | 1496/48008 [05:04<2:55:13, 4.42it/s] 3%|▎ | 1497/48008 [05:04<2:47:25, 4.63it/s] 3%|▎ | 1498/48008 [05:04<2:40:56, 4.82it/s] 3%|▎ | 1499/48008 [05:04<2:37:28, 4.92it/s] 3%|▎ | 1500/48008 [05:05<2:33:12, 5.06it/s] {'loss': 0.3123, 'grad_norm': 0.33203125, 'learning_rate': 0.00019375520746542244, 'epoch': 0.03} 3%|▎ | 1500/48008 [05:05<2:33:12, 5.06it/s] 3%|▎ | 1501/48008 [05:05<2:32:16, 5.09it/s] 3%|▎ | 1502/48008 [05:05<3:16:19, 3.95it/s] 3%|▎ | 1503/48008 [05:05<3:06:26, 4.16it/s] 3%|▎ | 1504/48008 [05:06<2:53:30, 4.47it/s] 3%|▎ | 1505/48008 [05:06<2:44:55, 4.70it/s] 3%|▎ | 1506/48008 [05:06<2:42:28, 4.77it/s] 3%|▎ | 1507/48008 [05:06<2:37:46, 4.91it/s] 3%|▎ | 1508/48008 [05:06<2:34:29, 5.02it/s] 3%|▎ | 1509/48008 [05:07<2:32:34, 5.08it/s] 3%|▎ | 1510/48008 [05:07<2:31:07, 5.13it/s] 3%|▎ | 1511/48008 [05:07<2:30:00, 5.17it/s] 3%|▎ | 1512/48008 [05:07<2:30:19, 5.16it/s] 3%|▎ | 1513/48008 [05:07<2:29:43, 5.18it/s] 3%|▎ | 1514/48008 [05:08<2:32:52, 5.07it/s] 3%|▎ | 1515/48008 [05:08<2:30:41, 5.14it/s] 3%|▎ | 1516/48008 [05:08<2:29:41, 5.18it/s] 3%|▎ | 1517/48008 [05:08<2:31:51, 5.10it/s] 3%|▎ | 1518/48008 [05:08<2:29:59, 5.17it/s] 3%|▎ | 1519/48008 [05:08<2:28:02, 5.23it/s] 3%|▎ | 1520/48008 [05:09<2:28:15, 5.23it/s] 3%|▎ | 1521/48008 [05:09<2:26:32, 5.29it/s] 3%|▎ | 1522/48008 [05:09<2:28:18, 5.22it/s] 3%|▎ | 1523/48008 [05:09<2:27:06, 5.27it/s] 3%|▎ | 1524/48008 [05:09<2:26:55, 5.27it/s] 3%|▎ | 1525/48008 [05:10<3:12:34, 4.02it/s] 3%|▎ | 1526/48008 [05:10<3:00:14, 4.30it/s] 3%|▎ | 1527/48008 [05:10<2:50:33, 4.54it/s] 3%|▎ | 1528/48008 [05:10<2:42:39, 4.76it/s] 3%|▎ | 1529/48008 [05:11<2:38:14, 4.90it/s] 3%|▎ | 1530/48008 [05:11<2:35:16, 4.99it/s] 3%|▎ | 1531/48008 [05:11<2:32:48, 5.07it/s] 3%|▎ | 1532/48008 [05:11<2:31:15, 5.12it/s] 3%|▎ | 1533/48008 [05:11<2:32:19, 5.09it/s] 3%|▎ | 1534/48008 [05:12<2:33:25, 5.05it/s] 3%|▎ | 1535/48008 [05:12<2:31:51, 5.10it/s] 3%|▎ | 1536/48008 [05:12<2:30:24, 5.15it/s] 3%|▎ | 1537/48008 [05:12<2:29:42, 5.17it/s] 3%|▎ | 1538/48008 [05:12<2:30:32, 5.14it/s] 3%|▎ | 1539/48008 [05:12<2:29:40, 5.17it/s] 3%|▎ | 1540/48008 [05:13<2:33:12, 5.06it/s] 3%|▎ | 1541/48008 [05:13<2:31:24, 5.11it/s] 3%|▎ | 1542/48008 [05:13<2:31:32, 5.11it/s] 3%|▎ | 1543/48008 [05:13<2:30:24, 5.15it/s] 3%|▎ | 1544/48008 [05:13<2:30:29, 5.15it/s] 3%|▎ | 1545/48008 [05:14<2:30:27, 5.15it/s] 3%|▎ | 1546/48008 [05:14<2:28:38, 5.21it/s] 3%|▎ | 1547/48008 [05:14<2:29:18, 5.19it/s] 3%|▎ | 1548/48008 [05:14<2:28:01, 5.23it/s] 3%|▎ | 1549/48008 [05:14<2:27:47, 5.24it/s] 3%|▎ | 1550/48008 [05:15<2:28:02, 5.23it/s] {'loss': 0.3087, 'grad_norm': 0.94921875, 'learning_rate': 0.00019354690884852525, 'epoch': 0.03} 3%|▎ | 1550/48008 [05:15<2:28:02, 5.23it/s] 3%|▎ | 1551/48008 [05:15<2:28:31, 5.21it/s] 3%|▎ | 1552/48008 [05:15<2:28:29, 5.21it/s] 3%|▎ | 1553/48008 [05:15<2:29:52, 5.17it/s] 3%|▎ | 1554/48008 [05:15<2:28:57, 5.20it/s] 3%|▎ | 1555/48008 [05:16<2:29:01, 5.20it/s] 3%|▎ | 1556/48008 [05:16<2:28:15, 5.22it/s] 3%|▎ | 1557/48008 [05:16<2:27:55, 5.23it/s] 3%|▎ | 1558/48008 [05:16<2:28:43, 5.21it/s] 3%|▎ | 1559/48008 [05:16<2:28:36, 5.21it/s] 3%|▎ | 1560/48008 [05:17<2:28:10, 5.22it/s] 3%|▎ | 1561/48008 [05:17<2:27:31, 5.25it/s] 3%|▎ | 1562/48008 [05:17<2:27:10, 5.26it/s] 3%|▎ | 1563/48008 [05:17<2:28:39, 5.21it/s] 3%|▎ | 1564/48008 [05:17<2:27:27, 5.25it/s] 3%|▎ | 1565/48008 [05:17<2:26:22, 5.29it/s] 3%|▎ | 1566/48008 [05:18<2:28:15, 5.22it/s] 3%|▎ | 1567/48008 [05:18<2:30:24, 5.15it/s] 3%|▎ | 1568/48008 [05:18<2:28:43, 5.20it/s] 3%|▎ | 1569/48008 [05:18<2:29:55, 5.16it/s] 3%|▎ | 1570/48008 [05:18<2:32:00, 5.09it/s] 3%|▎ | 1571/48008 [05:19<2:33:22, 5.05it/s] 3%|▎ | 1572/48008 [05:19<3:16:55, 3.93it/s] 3%|▎ | 1573/48008 [05:19<3:02:41, 4.24it/s] 3%|▎ | 1574/48008 [05:19<2:52:22, 4.49it/s] 3%|▎ | 1575/48008 [05:20<2:46:14, 4.66it/s] 3%|▎ | 1576/48008 [05:20<2:39:28, 4.85it/s] 3%|▎ | 1577/48008 [05:20<2:36:01, 4.96it/s] 3%|▎ | 1578/48008 [05:20<2:33:44, 5.03it/s] 3%|▎ | 1579/48008 [05:20<2:30:11, 5.15it/s] 3%|▎ | 1580/48008 [05:21<2:30:24, 5.14it/s] 3%|▎ | 1581/48008 [05:21<2:29:09, 5.19it/s] 3%|▎ | 1582/48008 [05:21<2:28:31, 5.21it/s] 3%|▎ | 1583/48008 [05:21<2:28:42, 5.20it/s] 3%|▎ | 1584/48008 [05:21<2:28:31, 5.21it/s] 3%|▎ | 1585/48008 [05:22<2:26:47, 5.27it/s] 3%|▎ | 1586/48008 [05:22<2:26:22, 5.29it/s] 3%|▎ | 1587/48008 [05:22<2:30:59, 5.12it/s] 3%|▎ | 1588/48008 [05:22<2:29:52, 5.16it/s] 3%|▎ | 1589/48008 [05:23<3:14:22, 3.98it/s] 3%|▎ | 1590/48008 [05:23<3:04:25, 4.19it/s] 3%|▎ | 1591/48008 [05:23<2:55:52, 4.40it/s] 3%|▎ | 1592/48008 [05:23<2:47:12, 4.63it/s] 3%|▎ | 1593/48008 [05:23<2:41:05, 4.80it/s] 3%|▎ | 1594/48008 [05:23<2:37:11, 4.92it/s] 3%|▎ | 1595/48008 [05:24<2:33:43, 5.03it/s] 3%|▎ | 1596/48008 [05:24<2:34:09, 5.02it/s] 3%|▎ | 1597/48008 [05:24<2:36:16, 4.95it/s] 3%|▎ | 1598/48008 [05:24<2:33:33, 5.04it/s] 3%|▎ | 1599/48008 [05:24<2:31:33, 5.10it/s] 3%|▎ | 1600/48008 [05:25<2:30:42, 5.13it/s] {'loss': 0.3109, 'grad_norm': 1.7265625, 'learning_rate': 0.00019333861023162807, 'epoch': 0.03} 3%|▎ | 1600/48008 [05:25<2:30:42, 5.13it/s] 3%|▎ | 1601/48008 [05:25<2:30:14, 5.15it/s] 3%|▎ | 1602/48008 [05:25<2:30:07, 5.15it/s] 3%|▎ | 1603/48008 [05:25<2:32:04, 5.09it/s] 3%|▎ | 1604/48008 [05:25<2:33:07, 5.05it/s] 3%|▎ | 1605/48008 [05:26<2:31:11, 5.12it/s] 3%|▎ | 1606/48008 [05:26<2:29:47, 5.16it/s] 3%|▎ | 1607/48008 [05:26<3:14:35, 3.97it/s] 3%|▎ | 1608/48008 [05:26<2:59:55, 4.30it/s] 3%|▎ | 1609/48008 [05:27<2:50:27, 4.54it/s] 3%|▎ | 1610/48008 [05:27<2:43:54, 4.72it/s] 3%|▎ | 1611/48008 [05:27<2:38:10, 4.89it/s] 3%|▎ | 1612/48008 [05:27<2:35:19, 4.98it/s] 3%|▎ | 1613/48008 [05:27<2:32:27, 5.07it/s] 3%|▎ | 1614/48008 [05:28<2:33:57, 5.02it/s] 3%|▎ | 1615/48008 [05:28<2:32:00, 5.09it/s] 3%|▎ | 1616/48008 [05:28<2:28:20, 5.21it/s] 3%|▎ | 1617/48008 [05:28<2:29:31, 5.17it/s] 3%|▎ | 1618/48008 [05:28<2:31:21, 5.11it/s] 3%|▎ | 1619/48008 [05:29<2:31:43, 5.10it/s] 3%|▎ | 1620/48008 [05:29<2:30:39, 5.13it/s] 3%|▎ | 1621/48008 [05:29<2:29:13, 5.18it/s] 3%|▎ | 1622/48008 [05:29<2:29:42, 5.16it/s] 3%|▎ | 1623/48008 [05:29<2:39:24, 4.85it/s] 3%|▎ | 1624/48008 [05:30<2:35:38, 4.97it/s] 3%|▎ | 1625/48008 [05:30<2:31:48, 5.09it/s] 3%|▎ | 1626/48008 [05:30<2:33:12, 5.05it/s] 3%|▎ | 1627/48008 [05:30<2:31:47, 5.09it/s] 3%|▎ | 1628/48008 [05:30<2:30:21, 5.14it/s] 3%|▎ | 1629/48008 [05:31<3:14:51, 3.97it/s] 3%|▎ | 1630/48008 [05:31<3:00:27, 4.28it/s] 3%|▎ | 1631/48008 [05:31<2:50:41, 4.53it/s] 3%|▎ | 1632/48008 [05:31<2:43:58, 4.71it/s] 3%|▎ | 1633/48008 [05:31<2:40:45, 4.81it/s] 3%|▎ | 1634/48008 [05:32<3:03:16, 4.22it/s] 3%|▎ | 1635/48008 [05:32<2:51:45, 4.50it/s] 3%|▎ | 1636/48008 [05:32<2:45:33, 4.67it/s] 3%|▎ | 1637/48008 [05:32<2:40:19, 4.82it/s] 3%|▎ | 1638/48008 [05:33<2:35:51, 4.96it/s] 3%|▎ | 1639/48008 [05:33<3:18:48, 3.89it/s] 3%|▎ | 1640/48008 [05:33<3:03:38, 4.21it/s] 3%|▎ | 1641/48008 [05:33<2:54:01, 4.44it/s] 3%|▎ | 1642/48008 [05:33<2:48:49, 4.58it/s] 3%|▎ | 1643/48008 [05:34<2:40:29, 4.81it/s] 3%|▎ | 1644/48008 [05:34<3:21:47, 3.83it/s] 3%|▎ | 1645/48008 [05:34<3:06:34, 4.14it/s] 3%|▎ | 1646/48008 [05:34<2:54:28, 4.43it/s] 3%|▎ | 1647/48008 [05:35<2:46:06, 4.65it/s] 3%|▎ | 1648/48008 [05:35<2:41:51, 4.77it/s] 3%|▎ | 1649/48008 [05:35<2:37:00, 4.92it/s] 3%|▎ | 1650/48008 [05:35<2:38:38, 4.87it/s] {'loss': 0.3097, 'grad_norm': 0.1806640625, 'learning_rate': 0.0001931303116147309, 'epoch': 0.03} 3%|▎ | 1650/48008 [05:35<2:38:38, 4.87it/s] 3%|▎ | 1651/48008 [05:35<2:35:16, 4.98it/s] 3%|▎ | 1652/48008 [05:36<2:35:34, 4.97it/s] 3%|▎ | 1653/48008 [05:36<2:31:38, 5.09it/s] 3%|▎ | 1654/48008 [05:36<2:32:53, 5.05it/s] 3%|▎ | 1655/48008 [05:36<2:32:50, 5.05it/s] 3%|▎ | 1656/48008 [05:36<2:33:27, 5.03it/s] 3%|▎ | 1657/48008 [05:37<2:31:34, 5.10it/s] 3%|▎ | 1658/48008 [05:37<2:30:18, 5.14it/s] 3%|▎ | 1659/48008 [05:37<2:28:54, 5.19it/s] 3%|▎ | 1660/48008 [05:37<2:27:29, 5.24it/s] 3%|▎ | 1661/48008 [05:37<2:27:43, 5.23it/s] 3%|▎ | 1662/48008 [05:38<2:27:12, 5.25it/s] 3%|▎ | 1663/48008 [05:38<2:28:37, 5.20it/s] 3%|▎ | 1664/48008 [05:38<2:28:30, 5.20it/s] 3%|▎ | 1665/48008 [05:38<2:27:15, 5.25it/s] 3%|▎ | 1666/48008 [05:38<2:26:17, 5.28it/s] 3%|▎ | 1667/48008 [05:39<6:17:59, 2.04it/s] 3%|▎ | 1668/48008 [05:40<5:10:53, 2.48it/s] 3%|▎ | 1669/48008 [05:40<4:24:34, 2.92it/s] 3%|▎ | 1670/48008 [05:40<3:52:07, 3.33it/s] 3%|▎ | 1671/48008 [05:40<3:26:23, 3.74it/s] 3%|▎ | 1672/48008 [05:40<3:07:57, 4.11it/s] 3%|▎ | 1673/48008 [05:41<2:55:41, 4.40it/s] 3%|▎ | 1674/48008 [05:41<2:46:22, 4.64it/s] 3%|▎ | 1675/48008 [05:41<2:40:31, 4.81it/s] 3%|▎ | 1676/48008 [05:41<2:40:55, 4.80it/s] 3%|▎ | 1677/48008 [05:41<2:37:00, 4.92it/s] 3%|▎ | 1678/48008 [05:42<2:34:03, 5.01it/s] 3%|▎ | 1679/48008 [05:42<3:17:13, 3.92it/s] 3%|▎ | 1680/48008 [05:42<3:01:42, 4.25it/s] 4%|▎ | 1681/48008 [05:42<2:51:45, 4.50it/s] 4%|▎ | 1682/48008 [05:43<2:43:18, 4.73it/s] 4%|▎ | 1683/48008 [05:43<2:37:43, 4.90it/s] 4%|▎ | 1684/48008 [05:43<2:37:17, 4.91it/s] 4%|▎ | 1685/48008 [05:43<2:36:34, 4.93it/s] 4%|▎ | 1686/48008 [05:43<2:34:06, 5.01it/s] 4%|▎ | 1687/48008 [05:44<2:32:38, 5.06it/s] 4%|▎ | 1688/48008 [05:44<2:30:30, 5.13it/s] 4%|▎ | 1689/48008 [05:44<2:40:01, 4.82it/s] 4%|▎ | 1690/48008 [05:44<2:36:00, 4.95it/s] 4%|▎ | 1691/48008 [05:44<2:34:00, 5.01it/s] 4%|▎ | 1692/48008 [05:45<2:30:43, 5.12it/s] 4%|▎ | 1693/48008 [05:45<2:29:54, 5.15it/s] 4%|▎ | 1694/48008 [05:45<2:29:32, 5.16it/s] 4%|▎ | 1695/48008 [05:45<2:28:19, 5.20it/s] 4%|▎ | 1696/48008 [05:45<2:30:40, 5.12it/s] 4%|▎ | 1697/48008 [05:46<2:31:15, 5.10it/s] 4%|▎ | 1698/48008 [05:46<2:29:49, 5.15it/s] 4%|▎ | 1699/48008 [05:46<2:30:44, 5.12it/s] 4%|▎ | 1700/48008 [05:46<2:29:25, 5.17it/s] {'loss': 0.3075, 'grad_norm': 0.6015625, 'learning_rate': 0.0001929220129978337, 'epoch': 0.04} 4%|▎ | 1700/48008 [05:46<2:29:25, 5.17it/s] 4%|▎ | 1701/48008 [05:46<2:30:30, 5.13it/s] 4%|▎ | 1702/48008 [05:47<2:29:19, 5.17it/s] 4%|▎ | 1703/48008 [05:47<2:27:46, 5.22it/s] 4%|▎ | 1704/48008 [05:47<2:29:00, 5.18it/s] 4%|▎ | 1705/48008 [05:47<2:28:22, 5.20it/s] 4%|▎ | 1706/48008 [05:47<2:28:35, 5.19it/s] 4%|▎ | 1707/48008 [05:47<2:27:22, 5.24it/s] 4%|▎ | 1708/48008 [05:48<2:26:29, 5.27it/s] 4%|▎ | 1709/48008 [05:48<2:26:17, 5.27it/s] 4%|▎ | 1710/48008 [05:48<2:24:38, 5.33it/s] 4%|▎ | 1711/48008 [05:48<2:27:04, 5.25it/s] 4%|▎ | 1712/48008 [05:48<2:29:41, 5.15it/s] 4%|▎ | 1713/48008 [05:49<2:28:40, 5.19it/s] 4%|▎ | 1714/48008 [05:49<2:30:33, 5.12it/s] 4%|▎ | 1715/48008 [05:49<2:29:39, 5.16it/s] 4%|▎ | 1716/48008 [05:49<2:39:14, 4.84it/s] 4%|▎ | 1717/48008 [05:49<2:35:47, 4.95it/s] 4%|▎ | 1718/48008 [05:50<3:00:06, 4.28it/s] 4%|▎ | 1719/48008 [05:50<2:50:22, 4.53it/s] 4%|▎ | 1720/48008 [05:50<2:43:08, 4.73it/s] 4%|▎ | 1721/48008 [05:50<2:41:20, 4.78it/s] 4%|▎ | 1722/48008 [05:51<2:36:00, 4.94it/s] 4%|▎ | 1723/48008 [05:51<2:35:54, 4.95it/s] 4%|▎ | 1724/48008 [05:51<2:32:37, 5.05it/s] 4%|▎ | 1725/48008 [05:51<2:29:29, 5.16it/s] 4%|▎ | 1726/48008 [05:51<2:28:42, 5.19it/s] 4%|▎ | 1727/48008 [05:51<2:28:30, 5.19it/s] 4%|▎ | 1728/48008 [05:52<2:26:46, 5.26it/s] 4%|▎ | 1729/48008 [05:52<2:26:19, 5.27it/s] 4%|▎ | 1730/48008 [05:52<2:36:55, 4.92it/s] 4%|▎ | 1731/48008 [05:52<2:34:16, 5.00it/s] 4%|▎ | 1732/48008 [05:52<2:31:50, 5.08it/s] 4%|▎ | 1733/48008 [05:53<2:30:13, 5.13it/s] 4%|▎ | 1734/48008 [05:53<2:28:08, 5.21it/s] 4%|▎ | 1735/48008 [05:53<2:27:09, 5.24it/s] 4%|▎ | 1736/48008 [05:53<2:30:03, 5.14it/s] 4%|▎ | 1737/48008 [05:53<2:28:38, 5.19it/s] 4%|▎ | 1738/48008 [05:54<2:27:44, 5.22it/s] 4%|▎ | 1739/48008 [05:54<2:27:27, 5.23it/s] 4%|▎ | 1740/48008 [05:54<2:26:29, 5.26it/s] 4%|▎ | 1741/48008 [05:54<2:27:05, 5.24it/s] 4%|▎ | 1742/48008 [05:55<3:12:04, 4.01it/s] 4%|▎ | 1743/48008 [05:55<2:57:14, 4.35it/s] 4%|▎ | 1744/48008 [05:55<2:47:46, 4.60it/s] 4%|▎ | 1745/48008 [05:55<2:41:34, 4.77it/s] 4%|▎ | 1746/48008 [05:55<2:41:33, 4.77it/s] 4%|▎ | 1747/48008 [05:56<2:36:21, 4.93it/s] 4%|▎ | 1748/48008 [05:56<2:33:25, 5.03it/s] 4%|▎ | 1749/48008 [05:56<2:31:37, 5.08it/s] 4%|▎ | 1750/48008 [05:56<2:30:25, 5.13it/s] {'loss': 0.3093, 'grad_norm': 0.97265625, 'learning_rate': 0.00019271371438093652, 'epoch': 0.04} 4%|▎ | 1750/48008 [05:56<2:30:25, 5.13it/s] 4%|▎ | 1751/48008 [05:56<2:32:38, 5.05it/s] 4%|▎ | 1752/48008 [05:56<2:30:06, 5.14it/s] 4%|▎ | 1753/48008 [05:57<2:28:49, 5.18it/s] 4%|▎ | 1754/48008 [05:57<2:28:38, 5.19it/s] 4%|▎ | 1755/48008 [05:57<2:26:45, 5.25it/s] 4%|▎ | 1756/48008 [05:57<2:26:46, 5.25it/s] 4%|▎ | 1757/48008 [05:57<2:27:37, 5.22it/s] 4%|▎ | 1758/48008 [05:58<2:38:20, 4.87it/s] 4%|▎ | 1759/48008 [05:58<2:34:56, 4.98it/s] 4%|▎ | 1760/48008 [05:58<2:32:24, 5.06it/s] 4%|▎ | 1761/48008 [05:58<2:29:45, 5.15it/s] 4%|▎ | 1762/48008 [05:58<2:28:46, 5.18it/s] 4%|▎ | 1763/48008 [05:59<2:30:45, 5.11it/s] 4%|▎ | 1764/48008 [05:59<2:29:44, 5.15it/s] 4%|▎ | 1765/48008 [05:59<2:33:28, 5.02it/s] 4%|▎ | 1766/48008 [05:59<2:29:42, 5.15it/s] 4%|▎ | 1767/48008 [05:59<2:28:47, 5.18it/s] 4%|▎ | 1768/48008 [06:00<2:25:54, 5.28it/s] 4%|▎ | 1769/48008 [06:00<2:25:56, 5.28it/s] 4%|▎ | 1770/48008 [06:00<2:29:15, 5.16it/s] 4%|▎ | 1771/48008 [06:00<2:29:54, 5.14it/s] 4%|▎ | 1772/48008 [06:00<2:28:54, 5.18it/s] 4%|▎ | 1773/48008 [06:01<2:28:04, 5.20it/s] 4%|▎ | 1774/48008 [06:01<2:27:15, 5.23it/s] 4%|▎ | 1775/48008 [06:01<2:29:46, 5.14it/s] 4%|▎ | 1776/48008 [06:01<2:28:30, 5.19it/s] 4%|▎ | 1777/48008 [06:01<2:27:50, 5.21it/s] 4%|▎ | 1778/48008 [06:01<2:25:04, 5.31it/s] 4%|▎ | 1779/48008 [06:02<2:28:19, 5.19it/s] 4%|▎ | 1780/48008 [06:02<2:27:25, 5.23it/s] 4%|▎ | 1781/48008 [06:02<2:27:09, 5.24it/s] 4%|▎ | 1782/48008 [06:02<2:27:18, 5.23it/s] 4%|▎ | 1783/48008 [06:02<2:26:08, 5.27it/s] 4%|▎ | 1784/48008 [06:03<2:26:36, 5.25it/s] 4%|▎ | 1785/48008 [06:03<2:26:45, 5.25it/s] 4%|▎ | 1786/48008 [06:03<2:26:36, 5.25it/s] 4%|▎ | 1787/48008 [06:03<2:28:56, 5.17it/s] 4%|▎ | 1788/48008 [06:03<2:27:51, 5.21it/s] 4%|▎ | 1789/48008 [06:04<2:29:04, 5.17it/s] 4%|▎ | 1790/48008 [06:04<2:28:34, 5.18it/s] 4%|▎ | 1791/48008 [06:04<2:29:10, 5.16it/s] 4%|▎ | 1792/48008 [06:04<2:27:29, 5.22it/s] 4%|▎ | 1793/48008 [06:04<2:54:00, 4.43it/s] 4%|▎ | 1794/48008 [06:05<2:45:35, 4.65it/s] 4%|▎ | 1795/48008 [06:05<2:39:25, 4.83it/s] 4%|▎ | 1796/48008 [06:05<2:38:09, 4.87it/s] 4%|▎ | 1797/48008 [06:05<2:34:27, 4.99it/s] 4%|▎ | 1798/48008 [06:05<2:31:08, 5.10it/s] 4%|▎ | 1799/48008 [06:06<2:33:07, 5.03it/s] 4%|▎ | 1800/48008 [06:06<2:32:29, 5.05it/s] {'loss': 0.3108, 'grad_norm': 0.427734375, 'learning_rate': 0.00019250541576403934, 'epoch': 0.04} 4%|▎ | 1800/48008 [06:06<2:32:29, 5.05it/s] 4%|▍ | 1801/48008 [06:06<2:34:02, 5.00it/s] 4%|▍ | 1802/48008 [06:06<2:32:23, 5.05it/s] 4%|▍ | 1803/48008 [06:06<2:30:14, 5.13it/s] 4%|▍ | 1804/48008 [06:07<2:32:08, 5.06it/s] 4%|▍ | 1805/48008 [06:07<2:29:36, 5.15it/s] 4%|▍ | 1806/48008 [06:07<2:30:22, 5.12it/s] 4%|▍ | 1807/48008 [06:07<2:29:56, 5.14it/s] 4%|▍ | 1808/48008 [06:07<2:28:56, 5.17it/s] 4%|▍ | 1809/48008 [06:08<2:31:17, 5.09it/s] 4%|▍ | 1810/48008 [06:08<2:32:57, 5.03it/s] 4%|▍ | 1811/48008 [06:08<2:30:52, 5.10it/s] 4%|▍ | 1812/48008 [06:08<2:30:04, 5.13it/s] 4%|▍ | 1813/48008 [06:08<2:28:48, 5.17it/s] 4%|▍ | 1814/48008 [06:09<2:27:27, 5.22it/s] 4%|▍ | 1815/48008 [06:09<2:27:41, 5.21it/s] 4%|▍ | 1816/48008 [06:09<2:30:09, 5.13it/s] 4%|▍ | 1817/48008 [06:09<2:31:38, 5.08it/s] 4%|▍ | 1818/48008 [06:09<2:30:00, 5.13it/s] 4%|▍ | 1819/48008 [06:10<2:28:59, 5.17it/s] 4%|▍ | 1820/48008 [06:10<2:32:49, 5.04it/s] 4%|▍ | 1821/48008 [06:10<2:29:48, 5.14it/s] 4%|▍ | 1822/48008 [06:10<2:28:56, 5.17it/s] 4%|▍ | 1823/48008 [06:10<2:28:34, 5.18it/s] 4%|▍ | 1824/48008 [06:11<2:29:28, 5.15it/s] 4%|▍ | 1825/48008 [06:11<2:28:13, 5.19it/s] 4%|▍ | 1826/48008 [06:11<2:28:09, 5.20it/s] 4%|▍ | 1827/48008 [06:11<2:26:48, 5.24it/s] 4%|▍ | 1828/48008 [06:11<2:53:14, 4.44it/s] 4%|▍ | 1829/48008 [06:12<2:44:46, 4.67it/s] 4%|▍ | 1830/48008 [06:12<2:39:19, 4.83it/s] 4%|▍ | 1831/48008 [06:12<2:34:46, 4.97it/s] 4%|▍ | 1832/48008 [06:12<2:32:09, 5.06it/s] 4%|▍ | 1833/48008 [06:12<2:32:20, 5.05it/s] 4%|▍ | 1834/48008 [06:13<2:30:39, 5.11it/s] 4%|▍ | 1835/48008 [06:13<2:30:31, 5.11it/s] 4%|▍ | 1836/48008 [06:13<2:29:35, 5.14it/s] 4%|▍ | 1837/48008 [06:13<2:27:58, 5.20it/s] 4%|▍ | 1838/48008 [06:13<2:31:38, 5.07it/s] 4%|▍ | 1839/48008 [06:14<2:30:32, 5.11it/s] 4%|▍ | 1840/48008 [06:14<2:29:16, 5.15it/s] 4%|▍ | 1841/48008 [06:14<2:28:28, 5.18it/s] 4%|▍ | 1842/48008 [06:14<2:30:26, 5.11it/s] 4%|▍ | 1843/48008 [06:14<2:29:26, 5.15it/s] 4%|▍ | 1844/48008 [06:14<2:28:04, 5.20it/s] 4%|▍ | 1845/48008 [06:15<2:26:12, 5.26it/s] 4%|▍ | 1846/48008 [06:15<2:26:09, 5.26it/s] 4%|▍ | 1847/48008 [06:15<2:25:50, 5.28it/s] 4%|▍ | 1848/48008 [06:15<2:26:44, 5.24it/s] 4%|▍ | 1849/48008 [06:15<2:28:59, 5.16it/s] 4%|▍ | 1850/48008 [06:16<2:30:45, 5.10it/s] {'loss': 0.3075, 'grad_norm': 0.6640625, 'learning_rate': 0.00019229711714714216, 'epoch': 0.04} 4%|▍ | 1850/48008 [06:16<2:30:45, 5.10it/s] 4%|▍ | 1851/48008 [06:16<2:29:46, 5.14it/s] 4%|▍ | 1852/48008 [06:16<2:27:38, 5.21it/s] 4%|▍ | 1853/48008 [06:16<2:28:57, 5.16it/s] 4%|▍ | 1854/48008 [06:16<2:26:58, 5.23it/s] 4%|▍ | 1855/48008 [06:17<2:26:31, 5.25it/s] 4%|▍ | 1856/48008 [06:17<2:27:03, 5.23it/s] 4%|▍ | 1857/48008 [06:17<2:29:55, 5.13it/s] 4%|▍ | 1858/48008 [06:17<2:29:12, 5.16it/s] 4%|▍ | 1859/48008 [06:17<2:28:59, 5.16it/s] 4%|▍ | 1860/48008 [06:18<2:30:50, 5.10it/s] 4%|▍ | 1861/48008 [06:18<2:32:17, 5.05it/s] 4%|▍ | 1862/48008 [06:18<2:33:22, 5.01it/s] 4%|▍ | 1863/48008 [06:18<2:30:20, 5.12it/s] 4%|▍ | 1864/48008 [06:18<2:28:55, 5.16it/s] 4%|▍ | 1865/48008 [06:19<2:28:17, 5.19it/s] 4%|▍ | 1866/48008 [06:19<2:26:21, 5.25it/s] 4%|▍ | 1867/48008 [06:19<2:29:41, 5.14it/s] 4%|▍ | 1868/48008 [06:19<2:28:56, 5.16it/s] 4%|▍ | 1869/48008 [06:19<2:27:35, 5.21it/s] 4%|▍ | 1870/48008 [06:20<2:27:24, 5.22it/s] 4%|▍ | 1871/48008 [06:20<2:27:25, 5.22it/s] 4%|▍ | 1872/48008 [06:20<2:29:34, 5.14it/s] 4%|▍ | 1873/48008 [06:20<2:28:05, 5.19it/s] 4%|▍ | 1874/48008 [06:20<2:27:17, 5.22it/s] 4%|▍ | 1875/48008 [06:21<2:38:08, 4.86it/s] 4%|▍ | 1876/48008 [06:21<2:34:12, 4.99it/s] 4%|▍ | 1877/48008 [06:21<2:36:42, 4.91it/s] 4%|▍ | 1878/48008 [06:21<2:33:47, 5.00it/s] 4%|▍ | 1879/48008 [06:21<2:34:01, 4.99it/s] 4%|▍ | 1880/48008 [06:21<2:32:20, 5.05it/s] 4%|▍ | 1881/48008 [06:22<2:30:38, 5.10it/s] 4%|▍ | 1882/48008 [06:22<2:28:16, 5.18it/s] 4%|▍ | 1883/48008 [06:22<2:27:54, 5.20it/s] 4%|▍ | 1884/48008 [06:22<2:26:52, 5.23it/s] 4%|▍ | 1885/48008 [06:22<2:29:46, 5.13it/s] 4%|▍ | 1886/48008 [06:23<2:29:56, 5.13it/s] 4%|▍ | 1887/48008 [06:23<2:28:41, 5.17it/s] 4%|▍ | 1888/48008 [06:23<2:27:59, 5.19it/s] 4%|▍ | 1889/48008 [06:23<2:28:39, 5.17it/s] 4%|▍ | 1890/48008 [06:23<2:28:31, 5.18it/s] 4%|▍ | 1891/48008 [06:24<2:28:09, 5.19it/s] 4%|▍ | 1892/48008 [06:24<2:29:36, 5.14it/s] 4%|▍ | 1893/48008 [06:24<2:33:12, 5.02it/s] 4%|▍ | 1894/48008 [06:24<2:31:09, 5.08it/s] 4%|▍ | 1895/48008 [06:24<2:32:45, 5.03it/s] 4%|▍ | 1896/48008 [06:25<2:32:16, 5.05it/s] 4%|▍ | 1897/48008 [06:25<2:30:24, 5.11it/s] 4%|▍ | 1898/48008 [06:25<2:30:49, 5.10it/s] 4%|▍ | 1899/48008 [06:25<2:29:17, 5.15it/s] 4%|▍ | 1900/48008 [06:25<2:28:42, 5.17it/s] {'loss': 0.308, 'grad_norm': 0.43359375, 'learning_rate': 0.00019208881853024498, 'epoch': 0.04} 4%|▍ | 1900/48008 [06:25<2:28:42, 5.17it/s] 4%|▍ | 1901/48008 [06:26<2:27:55, 5.20it/s] 4%|▍ | 1902/48008 [06:26<2:29:08, 5.15it/s] 4%|▍ | 1903/48008 [06:26<2:28:44, 5.17it/s] 4%|▍ | 1904/48008 [06:26<2:27:30, 5.21it/s] 4%|▍ | 1905/48008 [06:27<3:12:01, 4.00it/s] 4%|▍ | 1906/48008 [06:27<2:59:55, 4.27it/s] 4%|▍ | 1907/48008 [06:27<2:49:56, 4.52it/s] 4%|▍ | 1908/48008 [06:27<2:44:08, 4.68it/s] 4%|▍ | 1909/48008 [06:27<2:38:27, 4.85it/s] 4%|▍ | 1910/48008 [06:27<2:35:04, 4.95it/s] 4%|▍ | 1911/48008 [06:28<2:35:35, 4.94it/s] 4%|▍ | 1912/48008 [06:28<2:32:48, 5.03it/s] 4%|▍ | 1913/48008 [06:28<2:35:23, 4.94it/s] 4%|▍ | 1914/48008 [06:28<2:31:28, 5.07it/s] 4%|▍ | 1915/48008 [06:28<2:30:16, 5.11it/s] 4%|▍ | 1916/48008 [06:29<2:31:03, 5.09it/s] 4%|▍ | 1917/48008 [06:29<2:29:46, 5.13it/s] 4%|▍ | 1918/48008 [06:29<2:29:13, 5.15it/s] 4%|▍ | 1919/48008 [06:29<2:28:44, 5.16it/s] 4%|▍ | 1920/48008 [06:29<2:27:58, 5.19it/s] 4%|▍ | 1921/48008 [06:30<2:27:22, 5.21it/s] 4%|▍ | 1922/48008 [06:30<2:28:32, 5.17it/s] 4%|▍ | 1923/48008 [06:30<2:28:07, 5.19it/s] 4%|▍ | 1924/48008 [06:30<2:27:39, 5.20it/s] 4%|▍ | 1925/48008 [06:30<2:27:05, 5.22it/s] 4%|▍ | 1926/48008 [06:31<2:29:25, 5.14it/s] 4%|▍ | 1927/48008 [06:31<2:28:24, 5.18it/s] 4%|▍ | 1928/48008 [06:31<2:27:10, 5.22it/s] 4%|▍ | 1929/48008 [06:31<2:29:22, 5.14it/s] 4%|▍ | 1930/48008 [06:31<2:29:38, 5.13it/s] 4%|▍ | 1931/48008 [06:32<2:28:23, 5.18it/s] 4%|▍ | 1932/48008 [06:32<2:28:42, 5.16it/s] 4%|▍ | 1933/48008 [06:32<2:28:00, 5.19it/s] 4%|▍ | 1934/48008 [06:32<2:26:29, 5.24it/s] 4%|▍ | 1935/48008 [06:32<2:29:03, 5.15it/s] 4%|▍ | 1936/48008 [06:33<2:28:43, 5.16it/s] 4%|▍ | 1937/48008 [06:33<2:28:37, 5.17it/s] 4%|▍ | 1938/48008 [06:33<2:31:47, 5.06it/s] 4%|▍ | 1939/48008 [06:33<2:30:30, 5.10it/s] 4%|▍ | 1940/48008 [06:33<2:29:13, 5.15it/s] 4%|▍ | 1941/48008 [06:34<2:28:06, 5.18it/s] 4%|▍ | 1942/48008 [06:34<2:26:18, 5.25it/s] 4%|▍ | 1943/48008 [06:34<2:26:29, 5.24it/s] 4%|▍ | 1944/48008 [06:34<2:27:59, 5.19it/s] 4%|▍ | 1945/48008 [06:34<2:28:45, 5.16it/s] 4%|▍ | 1946/48008 [06:34<2:28:07, 5.18it/s] 4%|▍ | 1947/48008 [06:35<2:30:04, 5.12it/s] 4%|▍ | 1948/48008 [06:35<2:29:16, 5.14it/s] 4%|▍ | 1949/48008 [06:35<2:27:15, 5.21it/s] 4%|▍ | 1950/48008 [06:35<2:26:48, 5.23it/s] {'loss': 0.3091, 'grad_norm': 0.369140625, 'learning_rate': 0.0001918805199133478, 'epoch': 0.04} 4%|▍ | 1950/48008 [06:35<2:26:48, 5.23it/s] 4%|▍ | 1951/48008 [06:35<2:27:51, 5.19it/s] 4%|▍ | 1952/48008 [06:36<2:27:34, 5.20it/s] 4%|▍ | 1953/48008 [06:36<2:26:54, 5.22it/s] 4%|▍ | 1954/48008 [06:36<2:25:59, 5.26it/s] 4%|▍ | 1955/48008 [06:36<2:25:01, 5.29it/s] 4%|▍ | 1956/48008 [06:36<2:23:59, 5.33it/s] 4%|▍ | 1957/48008 [06:37<2:24:17, 5.32it/s] 4%|▍ | 1958/48008 [06:37<2:24:43, 5.30it/s] 4%|▍ | 1959/48008 [06:37<2:27:45, 5.19it/s] 4%|▍ | 1960/48008 [06:37<2:27:32, 5.20it/s] 4%|▍ | 1961/48008 [06:37<2:26:54, 5.22it/s] 4%|▍ | 1962/48008 [06:38<2:26:52, 5.23it/s] 4%|▍ | 1963/48008 [06:38<2:27:33, 5.20it/s] 4%|▍ | 1964/48008 [06:38<2:27:03, 5.22it/s] 4%|▍ | 1965/48008 [06:38<2:26:35, 5.23it/s] 4%|▍ | 1966/48008 [06:38<2:24:43, 5.30it/s] 4%|▍ | 1967/48008 [06:38<2:24:25, 5.31it/s] 4%|▍ | 1968/48008 [06:39<2:51:06, 4.48it/s] 4%|▍ | 1969/48008 [06:39<2:46:10, 4.62it/s] 4%|▍ | 1970/48008 [06:39<2:41:10, 4.76it/s] 4%|▍ | 1971/48008 [06:39<2:38:14, 4.85it/s] 4%|▍ | 1972/48008 [06:40<2:34:48, 4.96it/s] 4%|▍ | 1973/48008 [06:40<2:32:31, 5.03it/s] 4%|▍ | 1974/48008 [06:40<2:29:01, 5.15it/s] 4%|▍ | 1975/48008 [06:40<2:28:12, 5.18it/s] 4%|▍ | 1976/48008 [06:40<2:27:20, 5.21it/s] 4%|▍ | 1977/48008 [06:41<2:26:55, 5.22it/s] 4%|▍ | 1978/48008 [06:41<2:26:05, 5.25it/s] 4%|▍ | 1979/48008 [06:41<2:24:32, 5.31it/s] 4%|▍ | 1980/48008 [06:41<2:25:28, 5.27it/s] 4%|▍ | 1981/48008 [06:41<2:36:33, 4.90it/s] 4%|▍ | 1982/48008 [06:42<2:44:22, 4.67it/s] 4%|▍ | 1983/48008 [06:42<2:38:55, 4.83it/s] 4%|▍ | 1984/48008 [06:42<2:35:00, 4.95it/s] 4%|▍ | 1985/48008 [06:42<2:31:33, 5.06it/s] 4%|▍ | 1986/48008 [06:42<2:29:39, 5.13it/s] 4%|▍ | 1987/48008 [06:42<2:29:55, 5.12it/s] 4%|▍ | 1988/48008 [06:43<2:29:10, 5.14it/s] 4%|▍ | 1989/48008 [06:43<2:27:08, 5.21it/s] 4%|▍ | 1990/48008 [06:43<2:29:17, 5.14it/s] 4%|▍ | 1991/48008 [06:43<2:29:27, 5.13it/s] 4%|▍ | 1992/48008 [06:43<2:27:44, 5.19it/s] 4%|▍ | 1993/48008 [06:44<2:28:53, 5.15it/s] 4%|▍ | 1994/48008 [06:44<2:27:04, 5.21it/s] 4%|▍ | 1995/48008 [06:44<2:30:10, 5.11it/s] 4%|▍ | 1996/48008 [06:44<2:29:09, 5.14it/s] 4%|▍ | 1997/48008 [06:44<2:29:44, 5.12it/s] 4%|▍ | 1998/48008 [06:45<2:39:24, 4.81it/s] 4%|▍ | 1999/48008 [06:45<2:35:25, 4.93it/s] 4%|▍ | 2000/48008 [06:45<2:33:07, 5.01it/s] {'loss': 0.3092, 'grad_norm': 0.330078125, 'learning_rate': 0.00019167222129645058, 'epoch': 0.04} 4%|▍ | 2000/48008 [06:45<2:33:07, 5.01it/s] 4%|▍ | 2001/48008 [06:45<2:33:40, 4.99it/s] 4%|▍ | 2002/48008 [06:45<2:31:26, 5.06it/s] 4%|▍ | 2003/48008 [06:46<2:32:35, 5.02it/s] 4%|▍ | 2004/48008 [06:46<2:33:17, 5.00it/s] 4%|▍ | 2005/48008 [06:46<2:29:59, 5.11it/s] 4%|▍ | 2006/48008 [06:46<2:28:57, 5.15it/s] 4%|▍ | 2007/48008 [06:46<2:27:29, 5.20it/s] 4%|▍ | 2008/48008 [06:47<2:26:15, 5.24it/s] 4%|▍ | 2009/48008 [06:47<2:29:17, 5.13it/s] 4%|▍ | 2010/48008 [06:47<2:28:24, 5.17it/s] 4%|▍ | 2011/48008 [06:47<2:29:30, 5.13it/s] 4%|▍ | 2012/48008 [06:47<2:29:31, 5.13it/s] 4%|▍ | 2013/48008 [06:48<2:27:59, 5.18it/s] 4%|▍ | 2014/48008 [06:48<2:30:05, 5.11it/s] 4%|▍ | 2015/48008 [06:48<2:26:46, 5.22it/s] 4%|▍ | 2016/48008 [06:48<2:29:16, 5.14it/s] 4%|▍ | 2017/48008 [06:48<2:27:39, 5.19it/s] 4%|▍ | 2018/48008 [06:49<2:28:35, 5.16it/s] 4%|▍ | 2019/48008 [06:49<2:28:46, 5.15it/s] 4%|▍ | 2020/48008 [06:49<2:29:22, 5.13it/s] 4%|▍ | 2021/48008 [06:49<2:30:59, 5.08it/s] 4%|▍ | 2022/48008 [06:49<2:30:54, 5.08it/s] 4%|▍ | 2023/48008 [06:50<2:33:37, 4.99it/s] 4%|▍ | 2024/48008 [06:50<2:31:05, 5.07it/s] 4%|▍ | 2025/48008 [06:50<2:29:40, 5.12it/s] 4%|▍ | 2026/48008 [06:50<2:27:24, 5.20it/s] 4%|▍ | 2027/48008 [06:50<2:25:36, 5.26it/s] 4%|▍ | 2028/48008 [06:50<2:24:59, 5.29it/s] 4%|▍ | 2029/48008 [06:51<2:25:05, 5.28it/s] 4%|▍ | 2030/48008 [06:51<2:27:12, 5.21it/s] 4%|▍ | 2031/48008 [06:51<2:25:47, 5.26it/s] 4%|▍ | 2032/48008 [06:51<2:26:13, 5.24it/s] 4%|▍ | 2033/48008 [06:51<2:25:16, 5.27it/s] 4%|▍ | 2034/48008 [06:52<2:27:03, 5.21it/s] 4%|▍ | 2035/48008 [06:52<2:26:49, 5.22it/s] 4%|▍ | 2036/48008 [06:52<2:26:06, 5.24it/s] 4%|▍ | 2037/48008 [06:52<2:25:20, 5.27it/s] 4%|▍ | 2038/48008 [06:52<2:27:19, 5.20it/s] 4%|▍ | 2039/48008 [06:53<2:29:13, 5.13it/s] 4%|▍ | 2040/48008 [06:53<2:28:16, 5.17it/s] 4%|▍ | 2041/48008 [06:53<2:27:24, 5.20it/s] 4%|▍ | 2042/48008 [06:53<2:27:11, 5.20it/s] 4%|▍ | 2043/48008 [06:53<2:26:18, 5.24it/s] 4%|▍ | 2044/48008 [06:54<2:30:19, 5.10it/s] 4%|▍ | 2045/48008 [06:54<2:27:49, 5.18it/s] 4%|▍ | 2046/48008 [06:54<2:25:11, 5.28it/s] 4%|▍ | 2047/48008 [06:54<2:28:00, 5.18it/s] 4%|▍ | 2048/48008 [06:54<2:26:53, 5.21it/s] 4%|▍ | 2049/48008 [06:55<2:25:51, 5.25it/s] 4%|▍ | 2050/48008 [06:55<2:29:39, 5.12it/s] {'loss': 0.3086, 'grad_norm': 0.259765625, 'learning_rate': 0.00019146392267955343, 'epoch': 0.04} 4%|▍ | 2050/48008 [06:55<2:29:39, 5.12it/s] 4%|▍ | 2051/48008 [06:55<2:28:05, 5.17it/s] 4%|▍ | 2052/48008 [06:55<2:27:07, 5.21it/s] 4%|▍ | 2053/48008 [06:55<2:30:39, 5.08it/s] 4%|▍ | 2054/48008 [06:56<2:39:50, 4.79it/s] 4%|▍ | 2055/48008 [06:56<2:35:50, 4.91it/s] 4%|▍ | 2056/48008 [06:56<2:32:55, 5.01it/s] 4%|▍ | 2057/48008 [06:56<2:31:23, 5.06it/s] 4%|▍ | 2058/48008 [06:56<2:28:40, 5.15it/s] 4%|▍ | 2059/48008 [06:56<2:26:38, 5.22it/s] 4%|▍ | 2060/48008 [06:57<2:25:48, 5.25it/s] 4%|▍ | 2061/48008 [06:57<2:25:37, 5.26it/s] 4%|▍ | 2062/48008 [06:57<2:25:16, 5.27it/s] 4%|▍ | 2063/48008 [06:57<2:25:06, 5.28it/s] 4%|▍ | 2064/48008 [06:57<2:25:51, 5.25it/s] 4%|▍ | 2065/48008 [06:58<2:26:16, 5.23it/s] 4%|▍ | 2066/48008 [06:58<2:36:35, 4.89it/s] 4%|▍ | 2067/48008 [06:58<2:33:43, 4.98it/s] 4%|▍ | 2068/48008 [06:58<2:31:13, 5.06it/s] 4%|▍ | 2069/48008 [06:58<2:30:11, 5.10it/s] 4%|▍ | 2070/48008 [06:59<2:27:42, 5.18it/s] 4%|▍ | 2071/48008 [06:59<2:27:29, 5.19it/s] 4%|▍ | 2072/48008 [06:59<2:29:49, 5.11it/s] 4%|▍ | 2073/48008 [06:59<2:39:01, 4.81it/s] 4%|▍ | 2074/48008 [06:59<2:34:58, 4.94it/s] 4%|▍ | 2075/48008 [07:00<2:43:29, 4.68it/s] 4%|▍ | 2076/48008 [07:00<2:37:24, 4.86it/s] 4%|▍ | 2077/48008 [07:00<2:33:17, 4.99it/s] 4%|▍ | 2078/48008 [07:00<2:30:38, 5.08it/s] 4%|▍ | 2079/48008 [07:00<2:39:59, 4.78it/s] 4%|▍ | 2080/48008 [07:01<2:36:02, 4.91it/s] 4%|▍ | 2081/48008 [07:01<2:33:57, 4.97it/s] 4%|▍ | 2082/48008 [07:01<2:32:05, 5.03it/s] 4%|▍ | 2083/48008 [07:01<2:29:16, 5.13it/s] 4%|▍ | 2084/48008 [07:01<2:28:01, 5.17it/s] 4%|▍ | 2085/48008 [07:02<2:26:17, 5.23it/s] 4%|▍ | 2086/48008 [07:02<3:11:02, 4.01it/s] 4%|▍ | 2087/48008 [07:02<2:57:03, 4.32it/s] 4%|▍ | 2088/48008 [07:02<2:52:21, 4.44it/s] 4%|▍ | 2089/48008 [07:03<2:44:04, 4.66it/s] 4%|▍ | 2090/48008 [07:03<2:38:07, 4.84it/s] 4%|▍ | 2091/48008 [07:03<2:33:32, 4.98it/s] 4%|▍ | 2092/48008 [07:03<2:31:14, 5.06it/s] 4%|▍ | 2093/48008 [07:03<2:31:58, 5.04it/s] 4%|▍ | 2094/48008 [07:04<2:31:51, 5.04it/s] 4%|▍ | 2095/48008 [07:04<2:30:16, 5.09it/s] 4%|▍ | 2096/48008 [07:04<2:27:48, 5.18it/s] 4%|▍ | 2097/48008 [07:04<2:26:04, 5.24it/s] 4%|▍ | 2098/48008 [07:04<2:26:16, 5.23it/s] 4%|▍ | 2099/48008 [07:05<2:26:34, 5.22it/s] 4%|▍ | 2100/48008 [07:05<2:27:36, 5.18it/s] {'loss': 0.309, 'grad_norm': 0.953125, 'learning_rate': 0.00019125562406265622, 'epoch': 0.04} 4%|▍ | 2100/48008 [07:05<2:27:36, 5.18it/s] 4%|▍ | 2101/48008 [07:05<2:26:25, 5.23it/s] 4%|▍ | 2102/48008 [07:05<2:25:32, 5.26it/s] 4%|▍ | 2103/48008 [07:05<2:28:23, 5.16it/s] 4%|▍ | 2104/48008 [07:05<2:27:06, 5.20it/s] 4%|▍ | 2105/48008 [07:06<2:25:44, 5.25it/s] 4%|▍ | 2106/48008 [07:06<2:25:28, 5.26it/s] 4%|▍ | 2107/48008 [07:06<2:28:00, 5.17it/s] 4%|▍ | 2108/48008 [07:06<2:29:58, 5.10it/s] 4%|▍ | 2109/48008 [07:06<2:27:34, 5.18it/s] 4%|▍ | 2110/48008 [07:07<2:29:31, 5.12it/s] 4%|▍ | 2111/48008 [07:07<2:26:04, 5.24it/s] 4%|▍ | 2112/48008 [07:07<2:27:49, 5.17it/s] 4%|▍ | 2113/48008 [07:07<2:53:55, 4.40it/s] 4%|▍ | 2114/48008 [07:08<2:47:03, 4.58it/s] 4%|▍ | 2115/48008 [07:08<2:40:32, 4.76it/s] 4%|▍ | 2116/48008 [07:08<2:40:34, 4.76it/s] 4%|▍ | 2117/48008 [07:08<2:35:07, 4.93it/s] 4%|▍ | 2118/48008 [07:09<3:17:00, 3.88it/s] 4%|▍ | 2119/48008 [07:09<3:03:38, 4.16it/s] 4%|▍ | 2120/48008 [07:09<2:55:04, 4.37it/s] 4%|▍ | 2121/48008 [07:09<2:48:54, 4.53it/s] 4%|▍ | 2122/48008 [07:09<2:43:21, 4.68it/s] 4%|▍ | 2123/48008 [07:09<2:36:50, 4.88it/s] 4%|▍ | 2124/48008 [07:10<2:32:51, 5.00it/s] 4%|▍ | 2125/48008 [07:10<2:33:59, 4.97it/s] 4%|▍ | 2126/48008 [07:10<2:30:49, 5.07it/s] 4%|▍ | 2127/48008 [07:10<2:29:15, 5.12it/s] 4%|▍ | 2128/48008 [07:10<2:27:07, 5.20it/s] 4%|▍ | 2129/48008 [07:11<2:26:57, 5.20it/s] 4%|▍ | 2130/48008 [07:11<2:27:04, 5.20it/s] 4%|▍ | 2131/48008 [07:11<2:27:23, 5.19it/s] 4%|▍ | 2132/48008 [07:11<2:29:35, 5.11it/s] 4%|▍ | 2133/48008 [07:11<2:27:56, 5.17it/s] 4%|▍ | 2134/48008 [07:12<2:27:31, 5.18it/s] 4%|▍ | 2135/48008 [07:12<2:28:24, 5.15it/s] 4%|▍ | 2136/48008 [07:12<2:27:59, 5.17it/s] 4%|▍ | 2137/48008 [07:12<3:11:34, 3.99it/s] 4%|▍ | 2138/48008 [07:13<2:58:08, 4.29it/s] 4%|▍ | 2139/48008 [07:13<2:48:18, 4.54it/s] 4%|▍ | 2140/48008 [07:13<2:52:07, 4.44it/s] 4%|▍ | 2141/48008 [07:13<2:48:16, 4.54it/s] 4%|▍ | 2142/48008 [07:13<2:40:31, 4.76it/s] 4%|▍ | 2143/48008 [07:14<2:46:32, 4.59it/s] 4%|▍ | 2144/48008 [07:14<2:39:15, 4.80it/s] 4%|▍ | 2145/48008 [07:14<2:36:44, 4.88it/s] 4%|▍ | 2146/48008 [07:14<2:35:10, 4.93it/s] 4%|▍ | 2147/48008 [07:14<2:32:09, 5.02it/s] 4%|▍ | 2148/48008 [07:15<2:29:45, 5.10it/s] 4%|▍ | 2149/48008 [07:15<2:30:11, 5.09it/s] 4%|▍ | 2150/48008 [07:15<2:28:52, 5.13it/s] {'loss': 0.3083, 'grad_norm': 0.19140625, 'learning_rate': 0.00019104732544575906, 'epoch': 0.04} 4%|▍ | 2150/48008 [07:15<2:28:52, 5.13it/s] 4%|▍ | 2151/48008 [07:15<2:28:08, 5.16it/s] 4%|▍ | 2152/48008 [07:15<2:26:19, 5.22it/s] 4%|▍ | 2153/48008 [07:16<2:25:33, 5.25it/s] 4%|▍ | 2154/48008 [07:16<2:25:33, 5.25it/s] 4%|▍ | 2155/48008 [07:16<2:25:39, 5.25it/s] 4%|▍ | 2156/48008 [07:16<2:25:46, 5.24it/s] 4%|▍ | 2157/48008 [07:16<2:26:30, 5.22it/s] 4%|▍ | 2158/48008 [07:17<2:26:19, 5.22it/s] 4%|▍ | 2159/48008 [07:17<2:25:04, 5.27it/s] 4%|▍ | 2160/48008 [07:17<2:25:43, 5.24it/s] 5%|▍ | 2161/48008 [07:17<2:25:43, 5.24it/s] 5%|▍ | 2162/48008 [07:17<2:28:14, 5.15it/s] 5%|▍ | 2163/48008 [07:17<2:27:57, 5.16it/s] 5%|▍ | 2164/48008 [07:18<2:26:31, 5.21it/s] 5%|▍ | 2165/48008 [07:18<2:27:57, 5.16it/s] 5%|▍ | 2166/48008 [07:18<2:30:04, 5.09it/s] 5%|▍ | 2167/48008 [07:18<2:31:48, 5.03it/s] 5%|▍ | 2168/48008 [07:18<2:29:02, 5.13it/s] 5%|▍ | 2169/48008 [07:19<2:31:04, 5.06it/s] 5%|▍ | 2170/48008 [07:19<2:29:16, 5.12it/s] 5%|▍ | 2171/48008 [07:19<2:38:46, 4.81it/s] 5%|▍ | 2172/48008 [07:19<2:45:30, 4.62it/s] 5%|▍ | 2173/48008 [07:20<2:42:14, 4.71it/s] 5%|▍ | 2174/48008 [07:20<2:37:03, 4.86it/s] 5%|▍ | 2175/48008 [07:20<2:33:51, 4.97it/s] 5%|▍ | 2176/48008 [07:20<2:32:55, 5.00it/s] 5%|▍ | 2177/48008 [07:20<2:33:15, 4.98it/s] 5%|▍ | 2178/48008 [07:20<2:30:19, 5.08it/s] 5%|▍ | 2179/48008 [07:21<2:28:15, 5.15it/s] 5%|▍ | 2180/48008 [07:21<2:38:31, 4.82it/s] 5%|▍ | 2181/48008 [07:21<2:34:50, 4.93it/s] 5%|▍ | 2182/48008 [07:21<2:31:59, 5.03it/s] 5%|▍ | 2183/48008 [07:21<2:30:12, 5.08it/s] 5%|▍ | 2184/48008 [07:22<2:28:19, 5.15it/s] 5%|▍ | 2185/48008 [07:22<2:27:32, 5.18it/s] 5%|▍ | 2186/48008 [07:22<2:26:23, 5.22it/s] 5%|▍ | 2187/48008 [07:22<2:27:06, 5.19it/s] 5%|▍ | 2188/48008 [07:22<2:26:49, 5.20it/s] 5%|▍ | 2189/48008 [07:23<2:26:48, 5.20it/s] 5%|▍ | 2190/48008 [07:23<2:26:12, 5.22it/s] 5%|▍ | 2191/48008 [07:23<2:24:08, 5.30it/s] 5%|▍ | 2192/48008 [07:23<2:24:44, 5.28it/s] 5%|▍ | 2193/48008 [07:23<2:35:41, 4.90it/s] 5%|▍ | 2194/48008 [07:24<2:33:52, 4.96it/s] 5%|▍ | 2195/48008 [07:24<2:31:28, 5.04it/s] 5%|▍ | 2196/48008 [07:24<2:28:56, 5.13it/s] 5%|▍ | 2197/48008 [07:24<2:29:13, 5.12it/s] 5%|▍ | 2198/48008 [07:24<2:27:30, 5.18it/s] 5%|▍ | 2199/48008 [07:25<2:27:13, 5.19it/s] 5%|▍ | 2200/48008 [07:25<2:26:45, 5.20it/s] {'loss': 0.3083, 'grad_norm': 0.98046875, 'learning_rate': 0.00019083902682886185, 'epoch': 0.05} 5%|▍ | 2200/48008 [07:25<2:26:45, 5.20it/s] 5%|▍ | 2201/48008 [07:25<2:27:49, 5.16it/s] 5%|▍ | 2202/48008 [07:25<2:27:18, 5.18it/s] 5%|▍ | 2203/48008 [07:25<2:29:34, 5.10it/s] 5%|▍ | 2204/48008 [07:26<2:32:41, 5.00it/s] 5%|▍ | 2205/48008 [07:26<2:34:38, 4.94it/s] 5%|▍ | 2206/48008 [07:26<2:32:02, 5.02it/s] 5%|▍ | 2207/48008 [07:26<3:14:40, 3.92it/s] 5%|▍ | 2208/48008 [07:27<2:59:28, 4.25it/s] 5%|▍ | 2209/48008 [07:27<3:00:08, 4.24it/s] 5%|▍ | 2210/48008 [07:27<2:49:47, 4.50it/s] 5%|▍ | 2211/48008 [07:27<2:41:32, 4.72it/s] 5%|▍ | 2212/48008 [07:27<2:36:32, 4.88it/s] 5%|▍ | 2213/48008 [07:28<2:33:23, 4.98it/s] 5%|▍ | 2214/48008 [07:28<2:35:07, 4.92it/s] 5%|▍ | 2215/48008 [07:28<2:31:41, 5.03it/s] 5%|▍ | 2216/48008 [07:28<2:32:15, 5.01it/s] 5%|▍ | 2217/48008 [07:28<2:30:01, 5.09it/s] 5%|▍ | 2218/48008 [07:29<2:31:27, 5.04it/s] 5%|▍ | 2219/48008 [07:29<2:29:33, 5.10it/s] 5%|▍ | 2220/48008 [07:29<2:28:24, 5.14it/s] 5%|▍ | 2221/48008 [07:29<2:27:43, 5.17it/s] 5%|▍ | 2222/48008 [07:29<2:28:05, 5.15it/s] 5%|▍ | 2223/48008 [07:30<2:53:29, 4.40it/s] 5%|▍ | 2224/48008 [07:30<2:49:18, 4.51it/s] 5%|▍ | 2225/48008 [07:30<2:40:48, 4.75it/s] 5%|▍ | 2226/48008 [07:30<2:37:46, 4.84it/s] 5%|▍ | 2227/48008 [07:30<2:33:37, 4.97it/s] 5%|▍ | 2228/48008 [07:31<2:30:43, 5.06it/s] 5%|▍ | 2229/48008 [07:31<2:26:45, 5.20it/s] 5%|▍ | 2230/48008 [07:31<2:26:34, 5.21it/s] 5%|▍ | 2231/48008 [07:31<2:37:02, 4.86it/s] 5%|▍ | 2232/48008 [07:31<2:32:58, 4.99it/s] 5%|▍ | 2233/48008 [07:32<2:30:18, 5.08it/s] 5%|▍ | 2234/48008 [07:32<2:29:46, 5.09it/s] 5%|▍ | 2235/48008 [07:32<2:27:36, 5.17it/s] 5%|▍ | 2236/48008 [07:32<2:30:09, 5.08it/s] 5%|▍ | 2237/48008 [07:32<2:29:04, 5.12it/s] 5%|▍ | 2238/48008 [07:33<2:28:15, 5.15it/s] 5%|▍ | 2239/48008 [07:33<2:38:06, 4.82it/s] 5%|▍ | 2240/48008 [07:33<2:33:43, 4.96it/s] 5%|▍ | 2241/48008 [07:33<2:32:14, 5.01it/s] 5%|▍ | 2242/48008 [07:33<2:29:02, 5.12it/s] 5%|▍ | 2243/48008 [07:34<2:27:41, 5.16it/s] 5%|▍ | 2244/48008 [07:34<2:29:29, 5.10it/s] 5%|▍ | 2245/48008 [07:34<2:29:30, 5.10it/s] 5%|▍ | 2246/48008 [07:34<2:28:42, 5.13it/s] 5%|▍ | 2247/48008 [07:34<2:27:24, 5.17it/s] 5%|▍ | 2248/48008 [07:35<2:26:58, 5.19it/s] 5%|▍ | 2249/48008 [07:35<2:26:23, 5.21it/s] 5%|▍ | 2250/48008 [07:35<2:25:41, 5.23it/s] {'loss': 0.3064, 'grad_norm': 0.248046875, 'learning_rate': 0.00019063072821196467, 'epoch': 0.05} 5%|▍ | 2250/48008 [07:35<2:25:41, 5.23it/s] 5%|▍ | 2251/48008 [07:35<2:25:51, 5.23it/s] 5%|▍ | 2252/48008 [07:35<2:25:53, 5.23it/s] 5%|▍ | 2253/48008 [07:35<2:25:39, 5.24it/s] 5%|▍ | 2254/48008 [07:36<2:26:00, 5.22it/s] 5%|▍ | 2255/48008 [07:36<2:28:23, 5.14it/s] 5%|▍ | 2256/48008 [07:36<2:28:39, 5.13it/s] 5%|▍ | 2257/48008 [07:36<2:28:23, 5.14it/s] 5%|▍ | 2258/48008 [07:36<2:27:48, 5.16it/s] 5%|▍ | 2259/48008 [07:37<2:27:55, 5.15it/s] 5%|▍ | 2260/48008 [07:37<2:27:52, 5.16it/s] 5%|▍ | 2261/48008 [07:37<2:27:15, 5.18it/s] 5%|▍ | 2262/48008 [07:37<2:29:01, 5.12it/s] 5%|▍ | 2263/48008 [07:37<2:27:18, 5.18it/s] 5%|▍ | 2264/48008 [07:38<2:31:26, 5.03it/s] 5%|▍ | 2265/48008 [07:38<2:27:49, 5.16it/s] 5%|▍ | 2266/48008 [07:38<2:27:16, 5.18it/s] 5%|▍ | 2267/48008 [07:38<2:26:41, 5.20it/s] 5%|▍ | 2268/48008 [07:38<2:26:07, 5.22it/s] 5%|▍ | 2269/48008 [07:39<2:27:43, 5.16it/s] 5%|▍ | 2270/48008 [07:39<2:27:18, 5.17it/s] 5%|▍ | 2271/48008 [07:39<2:27:02, 5.18it/s] 5%|▍ | 2272/48008 [07:39<2:26:27, 5.20it/s] 5%|▍ | 2273/48008 [07:39<2:24:50, 5.26it/s] 5%|▍ | 2274/48008 [07:40<2:51:04, 4.46it/s] 5%|▍ | 2275/48008 [07:40<2:43:45, 4.65it/s] 5%|▍ | 2276/48008 [07:40<2:37:56, 4.83it/s] 5%|▍ | 2277/48008 [07:40<2:33:20, 4.97it/s] 5%|▍ | 2278/48008 [07:41<3:15:46, 3.89it/s] 5%|▍ | 2279/48008 [07:41<2:59:38, 4.24it/s] 5%|▍ | 2280/48008 [07:41<2:50:20, 4.47it/s] 5%|▍ | 2281/48008 [07:41<2:43:02, 4.67it/s] 5%|▍ | 2282/48008 [07:41<2:37:54, 4.83it/s] 5%|▍ | 2283/48008 [07:42<2:44:49, 4.62it/s] 5%|▍ | 2284/48008 [07:42<2:39:24, 4.78it/s] 5%|▍ | 2285/48008 [07:42<2:35:55, 4.89it/s] 5%|▍ | 2286/48008 [07:42<2:32:11, 5.01it/s] 5%|▍ | 2287/48008 [07:42<2:28:37, 5.13it/s] 5%|▍ | 2288/48008 [07:43<2:27:41, 5.16it/s] 5%|▍ | 2289/48008 [07:43<2:26:41, 5.19it/s] 5%|▍ | 2290/48008 [07:43<3:10:55, 3.99it/s] 5%|▍ | 2291/48008 [07:43<2:59:30, 4.24it/s] 5%|▍ | 2292/48008 [07:43<2:48:02, 4.53it/s] 5%|▍ | 2293/48008 [07:44<2:43:00, 4.67it/s] 5%|▍ | 2294/48008 [07:44<2:38:08, 4.82it/s] 5%|▍ | 2295/48008 [07:44<2:35:25, 4.90it/s] 5%|▍ | 2296/48008 [07:44<2:43:09, 4.67it/s] 5%|▍ | 2297/48008 [07:45<2:38:40, 4.80it/s] 5%|▍ | 2298/48008 [07:45<2:35:41, 4.89it/s] 5%|▍ | 2299/48008 [07:45<2:32:26, 5.00it/s] 5%|▍ | 2300/48008 [07:45<2:32:37, 4.99it/s] {'loss': 0.3064, 'grad_norm': 0.1865234375, 'learning_rate': 0.00019042242959506751, 'epoch': 0.05} 5%|▍ | 2300/48008 [07:45<2:32:37, 4.99it/s] 5%|▍ | 2301/48008 [07:45<2:35:00, 4.91it/s] 5%|▍ | 2302/48008 [07:46<2:31:58, 5.01it/s] 5%|▍ | 2303/48008 [07:46<2:33:31, 4.96it/s] 5%|▍ | 2304/48008 [07:46<2:31:09, 5.04it/s] 5%|▍ | 2305/48008 [07:46<2:30:34, 5.06it/s] 5%|▍ | 2306/48008 [07:46<2:28:26, 5.13it/s] 5%|▍ | 2307/48008 [07:46<2:30:01, 5.08it/s] 5%|▍ | 2308/48008 [07:47<2:28:58, 5.11it/s] 5%|▍ | 2309/48008 [07:47<2:30:24, 5.06it/s] 5%|▍ | 2310/48008 [07:47<2:55:17, 4.34it/s] 5%|▍ | 2311/48008 [07:47<2:49:21, 4.50it/s] 5%|▍ | 2312/48008 [07:48<2:44:17, 4.64it/s] 5%|▍ | 2313/48008 [07:48<2:37:55, 4.82it/s] 5%|▍ | 2314/48008 [07:48<2:34:26, 4.93it/s] 5%|▍ | 2315/48008 [07:48<2:30:15, 5.07it/s] 5%|▍ | 2316/48008 [07:48<2:28:42, 5.12it/s] 5%|▍ | 2317/48008 [07:49<2:27:35, 5.16it/s] 5%|▍ | 2318/48008 [07:49<2:26:39, 5.19it/s] 5%|▍ | 2319/48008 [07:49<2:25:18, 5.24it/s] 5%|▍ | 2320/48008 [07:49<2:24:34, 5.27it/s] 5%|▍ | 2321/48008 [07:49<2:25:31, 5.23it/s] 5%|▍ | 2322/48008 [07:49<2:25:41, 5.23it/s] 5%|▍ | 2323/48008 [07:50<2:25:39, 5.23it/s] 5%|▍ | 2324/48008 [07:50<2:25:20, 5.24it/s] 5%|▍ | 2325/48008 [07:50<2:28:04, 5.14it/s] 5%|▍ | 2326/48008 [07:50<2:26:39, 5.19it/s] 5%|▍ | 2327/48008 [07:50<2:25:13, 5.24it/s] 5%|▍ | 2328/48008 [07:51<2:24:55, 5.25it/s] 5%|▍ | 2329/48008 [07:51<2:25:49, 5.22it/s] 5%|▍ | 2330/48008 [07:51<2:28:08, 5.14it/s] 5%|▍ | 2331/48008 [07:51<2:29:58, 5.08it/s] 5%|▍ | 2332/48008 [07:51<2:30:20, 5.06it/s] 5%|▍ | 2333/48008 [07:52<2:30:07, 5.07it/s] 5%|▍ | 2334/48008 [07:52<2:30:06, 5.07it/s] 5%|▍ | 2335/48008 [07:52<2:29:22, 5.10it/s] 5%|▍ | 2336/48008 [07:52<3:12:29, 3.95it/s] 5%|▍ | 2337/48008 [07:53<3:01:05, 4.20it/s] 5%|▍ | 2338/48008 [07:53<2:50:26, 4.47it/s] 5%|▍ | 2339/48008 [07:53<2:42:40, 4.68it/s] 5%|▍ | 2340/48008 [07:53<2:40:05, 4.75it/s] 5%|▍ | 2341/48008 [07:53<2:34:59, 4.91it/s] 5%|▍ | 2342/48008 [07:54<2:34:45, 4.92it/s] 5%|▍ | 2343/48008 [07:54<2:31:50, 5.01it/s] 5%|▍ | 2344/48008 [07:54<2:40:02, 4.76it/s] 5%|▍ | 2345/48008 [07:54<2:34:55, 4.91it/s] 5%|▍ | 2346/48008 [07:54<2:32:01, 5.01it/s] 5%|▍ | 2347/48008 [07:55<2:29:18, 5.10it/s] 5%|▍ | 2348/48008 [07:55<2:29:42, 5.08it/s] 5%|▍ | 2349/48008 [07:55<2:32:33, 4.99it/s] 5%|▍ | 2350/48008 [07:55<2:31:39, 5.02it/s] {'loss': 0.3086, 'grad_norm': 1.390625, 'learning_rate': 0.0001902141309781703, 'epoch': 0.05} 5%|▍ | 2350/48008 [07:55<2:31:39, 5.02it/s] 5%|▍ | 2351/48008 [07:55<2:29:38, 5.09it/s] 5%|▍ | 2352/48008 [07:56<2:38:11, 4.81it/s] 5%|▍ | 2353/48008 [07:56<2:35:52, 4.88it/s] 5%|▍ | 2354/48008 [07:57<5:06:24, 2.48it/s] 5%|▍ | 2355/48008 [07:57<4:21:44, 2.91it/s] 5%|▍ | 2356/48008 [07:57<3:46:16, 3.36it/s] 5%|▍ | 2357/48008 [07:57<3:24:38, 3.72it/s] 5%|▍ | 2358/48008 [07:57<3:06:40, 4.08it/s] 5%|▍ | 2359/48008 [07:58<2:53:07, 4.39it/s] 5%|▍ | 2360/48008 [07:58<2:45:44, 4.59it/s] 5%|▍ | 2361/48008 [07:58<2:39:42, 4.76it/s] 5%|▍ | 2362/48008 [07:58<2:39:45, 4.76it/s] 5%|▍ | 2363/48008 [07:58<2:45:41, 4.59it/s] 5%|▍ | 2364/48008 [07:59<2:38:36, 4.80it/s] 5%|▍ | 2365/48008 [07:59<2:37:03, 4.84it/s] 5%|▍ | 2366/48008 [07:59<2:33:27, 4.96it/s] 5%|▍ | 2367/48008 [07:59<2:30:06, 5.07it/s] 5%|▍ | 2368/48008 [08:00<2:54:49, 4.35it/s] 5%|▍ | 2369/48008 [08:00<2:50:35, 4.46it/s] 5%|▍ | 2370/48008 [08:00<2:45:41, 4.59it/s] 5%|▍ | 2371/48008 [08:00<2:38:34, 4.80it/s] 5%|▍ | 2372/48008 [08:00<2:44:29, 4.62it/s] 5%|▍ | 2373/48008 [08:01<2:40:10, 4.75it/s] 5%|▍ | 2374/48008 [08:01<2:38:32, 4.80it/s] 5%|▍ | 2375/48008 [08:01<2:35:06, 4.90it/s] 5%|▍ | 2376/48008 [08:01<2:42:26, 4.68it/s] 5%|▍ | 2377/48008 [08:01<2:37:25, 4.83it/s] 5%|▍ | 2378/48008 [08:02<2:33:15, 4.96it/s] 5%|▍ | 2379/48008 [08:02<2:31:22, 5.02it/s] 5%|▍ | 2380/48008 [08:02<2:29:20, 5.09it/s] 5%|▍ | 2381/48008 [08:02<3:12:42, 3.95it/s] 5%|▍ | 2382/48008 [08:03<2:58:12, 4.27it/s] 5%|▍ | 2383/48008 [08:03<2:49:16, 4.49it/s] 5%|▍ | 2384/48008 [08:03<2:43:22, 4.65it/s] 5%|▍ | 2385/48008 [08:03<2:42:26, 4.68it/s] 5%|▍ | 2386/48008 [08:03<2:39:00, 4.78it/s] 5%|▍ | 2387/48008 [08:04<2:34:58, 4.91it/s] 5%|▍ | 2388/48008 [08:04<2:34:27, 4.92it/s] 5%|▍ | 2389/48008 [08:04<2:31:21, 5.02it/s] 5%|▍ | 2390/48008 [08:04<2:29:48, 5.07it/s] 5%|▍ | 2391/48008 [08:04<2:30:03, 5.07it/s] 5%|▍ | 2392/48008 [08:05<2:27:27, 5.16it/s] 5%|▍ | 2393/48008 [08:05<2:52:39, 4.40it/s] 5%|▍ | 2394/48008 [08:05<2:44:21, 4.63it/s] 5%|▍ | 2395/48008 [08:05<2:39:09, 4.78it/s] 5%|▍ | 2396/48008 [08:05<2:35:21, 4.89it/s] 5%|▍ | 2397/48008 [08:06<2:35:28, 4.89it/s] 5%|▍ | 2398/48008 [08:06<2:31:06, 5.03it/s] 5%|▍ | 2399/48008 [08:06<2:31:53, 5.00it/s] 5%|▍ | 2400/48008 [08:06<2:28:54, 5.10it/s] {'loss': 0.3059, 'grad_norm': 0.232421875, 'learning_rate': 0.00019000583236127315, 'epoch': 0.05} 5%|▍ | 2400/48008 [08:06<2:28:54, 5.10it/s] 5%|▌ | 2401/48008 [08:06<2:27:56, 5.14it/s] 5%|▌ | 2402/48008 [08:07<2:27:15, 5.16it/s] 5%|▌ | 2403/48008 [08:07<2:37:10, 4.84it/s] 5%|▌ | 2404/48008 [08:07<2:33:02, 4.97it/s] 5%|▌ | 2405/48008 [08:07<2:30:42, 5.04it/s] 5%|▌ | 2406/48008 [08:07<2:29:02, 5.10it/s] 5%|▌ | 2407/48008 [08:08<2:27:53, 5.14it/s] 5%|▌ | 2408/48008 [08:08<2:26:12, 5.20it/s] 5%|▌ | 2409/48008 [08:08<2:25:36, 5.22it/s] 5%|▌ | 2410/48008 [08:08<2:25:41, 5.22it/s] 5%|▌ | 2411/48008 [08:08<2:27:52, 5.14it/s] 5%|▌ | 2412/48008 [08:09<3:11:28, 3.97it/s] 5%|▌ | 2413/48008 [08:09<2:57:00, 4.29it/s] 5%|▌ | 2414/48008 [08:09<2:48:36, 4.51it/s] 5%|▌ | 2415/48008 [08:09<2:40:13, 4.74it/s] 5%|▌ | 2416/48008 [08:09<2:35:50, 4.88it/s] 5%|▌ | 2417/48008 [08:10<2:32:28, 4.98it/s] 5%|▌ | 2418/48008 [08:10<2:30:27, 5.05it/s] 5%|▌ | 2419/48008 [08:10<2:28:23, 5.12it/s] 5%|▌ | 2420/48008 [08:10<2:29:49, 5.07it/s] 5%|▌ | 2421/48008 [08:10<2:27:21, 5.16it/s] 5%|▌ | 2422/48008 [08:11<2:26:27, 5.19it/s] 5%|▌ | 2423/48008 [08:11<2:30:20, 5.05it/s] 5%|▌ | 2424/48008 [08:11<2:28:08, 5.13it/s] 5%|▌ | 2425/48008 [08:11<2:27:10, 5.16it/s] 5%|▌ | 2426/48008 [08:11<2:30:25, 5.05it/s] 5%|▌ | 2427/48008 [08:12<2:28:47, 5.11it/s] 5%|▌ | 2428/48008 [08:12<2:26:41, 5.18it/s] 5%|▌ | 2429/48008 [08:12<2:24:55, 5.24it/s] 5%|▌ | 2430/48008 [08:12<2:24:05, 5.27it/s] 5%|▌ | 2431/48008 [08:12<2:24:18, 5.26it/s] 5%|▌ | 2432/48008 [08:13<2:25:51, 5.21it/s] 5%|▌ | 2433/48008 [08:13<2:25:59, 5.20it/s] 5%|▌ | 2434/48008 [08:13<3:10:29, 3.99it/s] 5%|▌ | 2435/48008 [08:13<2:56:47, 4.30it/s] 5%|▌ | 2436/48008 [08:14<2:46:49, 4.55it/s] 5%|▌ | 2437/48008 [08:14<2:42:24, 4.68it/s] 5%|▌ | 2438/48008 [08:14<2:36:09, 4.86it/s] 5%|▌ | 2439/48008 [08:14<2:32:39, 4.98it/s] 5%|▌ | 2440/48008 [08:14<2:32:05, 4.99it/s] 5%|▌ | 2441/48008 [08:14<2:29:13, 5.09it/s] 5%|▌ | 2442/48008 [08:15<2:28:07, 5.13it/s] 5%|▌ | 2443/48008 [08:15<2:27:12, 5.16it/s] 5%|▌ | 2444/48008 [08:15<2:26:42, 5.18it/s] 5%|▌ | 2445/48008 [08:15<2:30:09, 5.06it/s] 5%|▌ | 2446/48008 [08:16<3:12:58, 3.94it/s] 5%|▌ | 2447/48008 [08:16<2:57:50, 4.27it/s] 5%|▌ | 2448/48008 [08:16<2:47:57, 4.52it/s] 5%|▌ | 2449/48008 [08:16<2:41:18, 4.71it/s] 5%|▌ | 2450/48008 [08:16<2:39:05, 4.77it/s] {'loss': 0.3094, 'grad_norm': 0.337890625, 'learning_rate': 0.00018979753374437594, 'epoch': 0.05} 5%|▌ | 2450/48008 [08:16<2:39:05, 4.77it/s] 5%|▌ | 2451/48008 [08:17<2:36:21, 4.86it/s] 5%|▌ | 2452/48008 [08:17<2:34:33, 4.91it/s] 5%|▌ | 2453/48008 [08:17<2:32:25, 4.98it/s] 5%|▌ | 2454/48008 [08:17<2:29:01, 5.09it/s] 5%|▌ | 2455/48008 [08:17<2:27:47, 5.14it/s] 5%|▌ | 2456/48008 [08:18<2:26:05, 5.20it/s] 5%|▌ | 2457/48008 [08:18<2:25:14, 5.23it/s] 5%|▌ | 2458/48008 [08:18<2:27:55, 5.13it/s] 5%|▌ | 2459/48008 [08:18<2:26:43, 5.17it/s] 5%|▌ | 2460/48008 [08:18<2:26:28, 5.18it/s] 5%|▌ | 2461/48008 [08:19<2:24:37, 5.25it/s] 5%|▌ | 2462/48008 [08:19<2:25:42, 5.21it/s] 5%|▌ | 2463/48008 [08:19<2:25:40, 5.21it/s] 5%|▌ | 2464/48008 [08:19<2:27:36, 5.14it/s] 5%|▌ | 2465/48008 [08:19<2:26:34, 5.18it/s] 5%|▌ | 2466/48008 [08:20<2:26:51, 5.17it/s] 5%|▌ | 2467/48008 [08:20<2:27:44, 5.14it/s] 5%|▌ | 2468/48008 [08:20<2:26:34, 5.18it/s] 5%|▌ | 2469/48008 [08:20<2:25:13, 5.23it/s] 5%|▌ | 2470/48008 [08:20<2:24:36, 5.25it/s] 5%|▌ | 2471/48008 [08:20<2:27:02, 5.16it/s] 5%|▌ | 2472/48008 [08:21<2:30:22, 5.05it/s] 5%|▌ | 2473/48008 [08:21<2:31:30, 5.01it/s] 5%|▌ | 2474/48008 [08:21<2:28:44, 5.10it/s] 5%|▌ | 2475/48008 [08:21<2:27:15, 5.15it/s] 5%|▌ | 2476/48008 [08:21<2:25:46, 5.21it/s] 5%|▌ | 2477/48008 [08:22<2:25:16, 5.22it/s] 5%|▌ | 2478/48008 [08:22<2:24:51, 5.24it/s] 5%|▌ | 2479/48008 [08:22<2:35:27, 4.88it/s] 5%|▌ | 2480/48008 [08:22<2:36:32, 4.85it/s] 5%|▌ | 2481/48008 [08:22<2:32:39, 4.97it/s] 5%|▌ | 2482/48008 [08:23<2:28:07, 5.12it/s] 5%|▌ | 2483/48008 [08:23<2:26:30, 5.18it/s] 5%|▌ | 2484/48008 [08:23<2:27:10, 5.16it/s] 5%|▌ | 2485/48008 [08:23<2:26:29, 5.18it/s] 5%|▌ | 2486/48008 [08:23<2:26:13, 5.19it/s] 5%|▌ | 2487/48008 [08:24<2:27:38, 5.14it/s] 5%|▌ | 2488/48008 [08:24<2:25:37, 5.21it/s] 5%|▌ | 2489/48008 [08:24<2:24:12, 5.26it/s] 5%|▌ | 2490/48008 [08:24<2:25:31, 5.21it/s] 5%|▌ | 2491/48008 [08:25<3:09:21, 4.01it/s] 5%|▌ | 2492/48008 [08:25<2:56:30, 4.30it/s] 5%|▌ | 2493/48008 [08:25<2:47:29, 4.53it/s] 5%|▌ | 2494/48008 [08:25<2:39:49, 4.75it/s] 5%|▌ | 2495/48008 [08:25<2:34:37, 4.91it/s] 5%|▌ | 2496/48008 [08:26<2:32:20, 4.98it/s] 5%|▌ | 2497/48008 [08:26<2:29:54, 5.06it/s] 5%|▌ | 2498/48008 [08:26<2:29:37, 5.07it/s] 5%|▌ | 2499/48008 [08:26<2:29:46, 5.06it/s] 5%|▌ | 2500/48008 [08:26<2:28:04, 5.12it/s] {'loss': 0.3082, 'grad_norm': 0.326171875, 'learning_rate': 0.00018958923512747878, 'epoch': 0.05} 5%|▌ | 2500/48008 [08:26<2:28:04, 5.12it/s] 5%|▌ | 2501/48008 [08:26<2:31:08, 5.02it/s] 5%|▌ | 2502/48008 [08:27<2:28:29, 5.11it/s] 5%|▌ | 2503/48008 [08:27<2:28:16, 5.11it/s] 5%|▌ | 2504/48008 [08:27<3:11:46, 3.95it/s] 5%|▌ | 2505/48008 [08:27<2:59:30, 4.22it/s] 5%|▌ | 2506/48008 [08:28<2:49:15, 4.48it/s] 5%|▌ | 2507/48008 [08:28<2:43:02, 4.65it/s] 5%|▌ | 2508/48008 [08:28<2:39:06, 4.77it/s] 5%|▌ | 2509/48008 [08:28<2:34:28, 4.91it/s] 5%|▌ | 2510/48008 [08:28<2:30:49, 5.03it/s] 5%|▌ | 2511/48008 [08:29<2:28:43, 5.10it/s] 5%|▌ | 2512/48008 [08:29<2:26:39, 5.17it/s] 5%|▌ | 2513/48008 [08:29<2:25:50, 5.20it/s] 5%|▌ | 2514/48008 [08:29<2:27:47, 5.13it/s] 5%|▌ | 2515/48008 [08:29<2:37:34, 4.81it/s] 5%|▌ | 2516/48008 [08:30<2:44:13, 4.62it/s] 5%|▌ | 2517/48008 [08:30<2:40:44, 4.72it/s] 5%|▌ | 2518/48008 [08:30<2:34:50, 4.90it/s] 5%|▌ | 2519/48008 [08:30<3:15:49, 3.87it/s] 5%|▌ | 2520/48008 [08:31<2:59:55, 4.21it/s] 5%|▌ | 2521/48008 [08:31<2:48:26, 4.50it/s] 5%|▌ | 2522/48008 [08:31<2:40:39, 4.72it/s] 5%|▌ | 2523/48008 [08:31<2:35:58, 4.86it/s] 5%|▌ | 2524/48008 [08:31<2:33:47, 4.93it/s] 5%|▌ | 2525/48008 [08:32<2:29:43, 5.06it/s] 5%|▌ | 2526/48008 [08:32<2:29:55, 5.06it/s] 5%|▌ | 2527/48008 [08:32<2:27:09, 5.15it/s] 5%|▌ | 2528/48008 [08:32<2:28:08, 5.12it/s] 5%|▌ | 2529/48008 [08:32<2:27:54, 5.12it/s] 5%|▌ | 2530/48008 [08:33<2:26:18, 5.18it/s] 5%|▌ | 2531/48008 [08:33<2:29:33, 5.07it/s] 5%|▌ | 2532/48008 [08:33<2:31:05, 5.02it/s] 5%|▌ | 2533/48008 [08:33<3:12:51, 3.93it/s] 5%|▌ | 2534/48008 [08:34<2:57:21, 4.27it/s] 5%|▌ | 2535/48008 [08:34<2:48:21, 4.50it/s] 5%|▌ | 2536/48008 [08:34<2:45:15, 4.59it/s] 5%|▌ | 2537/48008 [08:34<2:36:52, 4.83it/s] 5%|▌ | 2538/48008 [08:34<2:33:32, 4.94it/s] 5%|▌ | 2539/48008 [08:34<2:31:31, 5.00it/s] 5%|▌ | 2540/48008 [08:35<2:30:39, 5.03it/s] 5%|▌ | 2541/48008 [08:35<2:27:57, 5.12it/s] 5%|▌ | 2542/48008 [08:35<2:26:39, 5.17it/s] 5%|▌ | 2543/48008 [08:35<2:25:50, 5.20it/s] 5%|▌ | 2544/48008 [08:35<2:26:26, 5.17it/s] 5%|▌ | 2545/48008 [08:36<2:25:42, 5.20it/s] 5%|▌ | 2546/48008 [08:36<2:25:13, 5.22it/s] 5%|▌ | 2547/48008 [08:36<2:26:33, 5.17it/s] 5%|▌ | 2548/48008 [08:36<2:26:14, 5.18it/s] 5%|▌ | 2549/48008 [08:36<2:36:36, 4.84it/s] 5%|▌ | 2550/48008 [08:37<2:31:17, 5.01it/s] {'loss': 0.3055, 'grad_norm': 0.236328125, 'learning_rate': 0.00018938093651058157, 'epoch': 0.05} 5%|▌ | 2550/48008 [08:37<2:31:17, 5.01it/s] 5%|▌ | 2551/48008 [08:37<2:40:15, 4.73it/s] 5%|▌ | 2552/48008 [08:37<2:35:13, 4.88it/s] 5%|▌ | 2553/48008 [08:37<2:32:59, 4.95it/s] 5%|▌ | 2554/48008 [08:37<2:29:52, 5.05it/s] 5%|▌ | 2555/48008 [08:38<2:27:19, 5.14it/s] 5%|▌ | 2556/48008 [08:38<2:27:14, 5.14it/s] 5%|▌ | 2557/48008 [08:38<2:26:05, 5.19it/s] 5%|▌ | 2558/48008 [08:38<2:25:04, 5.22it/s] 5%|▌ | 2559/48008 [08:38<2:24:47, 5.23it/s] 5%|▌ | 2560/48008 [08:39<2:27:04, 5.15it/s] 5%|▌ | 2561/48008 [08:39<2:26:24, 5.17it/s] 5%|▌ | 2562/48008 [08:39<2:28:40, 5.09it/s] 5%|▌ | 2563/48008 [08:39<2:30:06, 5.05it/s] 5%|▌ | 2564/48008 [08:39<2:28:20, 5.11it/s] 5%|▌ | 2565/48008 [08:40<2:26:31, 5.17it/s] 5%|▌ | 2566/48008 [08:40<2:27:32, 5.13it/s] 5%|▌ | 2567/48008 [08:40<2:27:09, 5.15it/s] 5%|▌ | 2568/48008 [08:40<2:36:58, 4.82it/s] 5%|▌ | 2569/48008 [08:40<2:44:57, 4.59it/s] 5%|▌ | 2570/48008 [08:41<2:38:58, 4.76it/s] 5%|▌ | 2571/48008 [08:41<2:38:36, 4.77it/s] 5%|▌ | 2572/48008 [08:41<2:34:19, 4.91it/s] 5%|▌ | 2573/48008 [08:41<2:30:59, 5.02it/s] 5%|▌ | 2574/48008 [08:41<2:28:47, 5.09it/s] 5%|▌ | 2575/48008 [08:42<2:27:32, 5.13it/s] 5%|▌ | 2576/48008 [08:42<2:37:37, 4.80it/s] 5%|▌ | 2577/48008 [08:42<3:18:04, 3.82it/s] 5%|▌ | 2578/48008 [08:42<3:03:41, 4.12it/s] 5%|▌ | 2579/48008 [08:43<2:54:13, 4.35it/s] 5%|▌ | 2580/48008 [08:43<2:45:11, 4.58it/s] 5%|▌ | 2581/48008 [08:43<2:49:31, 4.47it/s] 5%|▌ | 2582/48008 [08:43<2:44:43, 4.60it/s] 5%|▌ | 2583/48008 [08:43<2:39:48, 4.74it/s] 5%|▌ | 2584/48008 [08:44<2:35:00, 4.88it/s] 5%|▌ | 2585/48008 [08:44<2:35:51, 4.86it/s] 5%|▌ | 2586/48008 [08:44<2:58:30, 4.24it/s] 5%|▌ | 2587/48008 [08:44<2:47:35, 4.52it/s] 5%|▌ | 2588/48008 [08:45<2:40:16, 4.72it/s] 5%|▌ | 2589/48008 [08:45<2:35:13, 4.88it/s] 5%|▌ | 2590/48008 [08:45<2:36:06, 4.85it/s] 5%|▌ | 2591/48008 [08:45<2:32:32, 4.96it/s] 5%|▌ | 2592/48008 [08:45<2:30:19, 5.04it/s] 5%|▌ | 2593/48008 [08:46<2:28:42, 5.09it/s] 5%|▌ | 2594/48008 [08:46<2:30:06, 5.04it/s] 5%|▌ | 2595/48008 [08:46<2:27:04, 5.15it/s] 5%|▌ | 2596/48008 [08:46<2:24:56, 5.22it/s] 5%|▌ | 2597/48008 [08:46<2:24:05, 5.25it/s] 5%|▌ | 2598/48008 [08:47<2:34:55, 4.89it/s] 5%|▌ | 2599/48008 [08:47<2:34:24, 4.90it/s] 5%|▌ | 2600/48008 [08:47<2:31:14, 5.00it/s] {'loss': 0.3047, 'grad_norm': 0.69140625, 'learning_rate': 0.0001891726378936844, 'epoch': 0.05} 5%|▌ | 2600/48008 [08:47<2:31:14, 5.00it/s] 5%|▌ | 2601/48008 [08:47<2:28:31, 5.10it/s] 5%|▌ | 2602/48008 [08:47<2:26:38, 5.16it/s] 5%|▌ | 2603/48008 [08:47<2:24:02, 5.25it/s] 5%|▌ | 2604/48008 [08:48<2:24:10, 5.25it/s] 5%|▌ | 2605/48008 [08:48<2:22:42, 5.30it/s] 5%|▌ | 2606/48008 [08:48<2:22:03, 5.33it/s] 5%|▌ | 2607/48008 [08:48<2:26:31, 5.16it/s] 5%|▌ | 2608/48008 [08:48<2:25:18, 5.21it/s] 5%|▌ | 2609/48008 [08:49<2:24:57, 5.22it/s] 5%|▌ | 2610/48008 [08:49<2:23:39, 5.27it/s] 5%|▌ | 2611/48008 [08:49<2:34:26, 4.90it/s] 5%|▌ | 2612/48008 [08:49<2:34:08, 4.91it/s] 5%|▌ | 2613/48008 [08:49<2:32:29, 4.96it/s] 5%|▌ | 2614/48008 [08:50<2:27:56, 5.11it/s] 5%|▌ | 2615/48008 [08:50<2:27:07, 5.14it/s] 5%|▌ | 2616/48008 [08:50<2:25:42, 5.19it/s] 5%|▌ | 2617/48008 [08:50<2:26:49, 5.15it/s] 5%|▌ | 2618/48008 [08:50<2:25:55, 5.18it/s] 5%|▌ | 2619/48008 [08:51<2:25:10, 5.21it/s] 5%|▌ | 2620/48008 [08:51<2:26:32, 5.16it/s] 5%|▌ | 2621/48008 [08:51<2:25:46, 5.19it/s] 5%|▌ | 2622/48008 [08:51<2:26:04, 5.18it/s] 5%|▌ | 2623/48008 [08:51<2:25:19, 5.21it/s] 5%|▌ | 2624/48008 [08:52<2:24:48, 5.22it/s] 5%|▌ | 2625/48008 [08:52<2:22:17, 5.32it/s] 5%|▌ | 2626/48008 [08:52<2:27:06, 5.14it/s] 5%|▌ | 2627/48008 [08:52<2:25:51, 5.19it/s] 5%|▌ | 2628/48008 [08:52<2:25:25, 5.20it/s] 5%|▌ | 2629/48008 [08:52<2:24:42, 5.23it/s] 5%|▌ | 2630/48008 [08:53<2:23:46, 5.26it/s] 5%|▌ | 2631/48008 [08:53<2:23:16, 5.28it/s] 5%|▌ | 2632/48008 [08:53<2:25:50, 5.19it/s] 5%|▌ | 2633/48008 [08:53<2:24:04, 5.25it/s] 5%|▌ | 2634/48008 [08:53<2:25:11, 5.21it/s] 5%|▌ | 2635/48008 [08:54<2:26:24, 5.17it/s] 5%|▌ | 2636/48008 [08:54<3:09:47, 3.98it/s] 5%|▌ | 2637/48008 [08:54<2:56:09, 4.29it/s] 5%|▌ | 2638/48008 [08:54<2:49:18, 4.47it/s] 5%|▌ | 2639/48008 [08:55<2:41:52, 4.67it/s] 5%|▌ | 2640/48008 [08:55<2:38:52, 4.76it/s] 6%|▌ | 2641/48008 [08:55<2:36:09, 4.84it/s] 6%|▌ | 2642/48008 [08:55<2:32:02, 4.97it/s] 6%|▌ | 2643/48008 [08:55<2:29:13, 5.07it/s] 6%|▌ | 2644/48008 [08:56<2:27:40, 5.12it/s] 6%|▌ | 2645/48008 [08:56<2:52:20, 4.39it/s] 6%|▌ | 2646/48008 [08:56<2:44:18, 4.60it/s] 6%|▌ | 2647/48008 [08:56<2:37:19, 4.81it/s] 6%|▌ | 2648/48008 [08:56<2:32:06, 4.97it/s] 6%|▌ | 2649/48008 [08:57<2:30:06, 5.04it/s] 6%|▌ | 2650/48008 [08:57<2:29:41, 5.05it/s] {'loss': 0.3088, 'grad_norm': 0.5, 'learning_rate': 0.0001889643392767872, 'epoch': 0.06} 6%|▌ | 2650/48008 [08:57<2:29:41, 5.05it/s] 6%|▌ | 2651/48008 [08:57<2:28:02, 5.11it/s] 6%|▌ | 2652/48008 [08:57<2:27:11, 5.14it/s] 6%|▌ | 2653/48008 [08:57<2:24:49, 5.22it/s] 6%|▌ | 2654/48008 [08:58<2:23:48, 5.26it/s] 6%|▌ | 2655/48008 [08:58<2:23:14, 5.28it/s] 6%|▌ | 2656/48008 [08:58<2:23:02, 5.28it/s] 6%|▌ | 2657/48008 [08:58<2:25:45, 5.19it/s] 6%|▌ | 2658/48008 [08:58<2:25:00, 5.21it/s] 6%|▌ | 2659/48008 [08:59<2:26:47, 5.15it/s] 6%|▌ | 2660/48008 [08:59<2:26:21, 5.16it/s] 6%|▌ | 2661/48008 [08:59<2:25:47, 5.18it/s] 6%|▌ | 2662/48008 [08:59<2:26:13, 5.17it/s] 6%|▌ | 2663/48008 [08:59<2:24:59, 5.21it/s] 6%|▌ | 2664/48008 [09:00<2:25:30, 5.19it/s] 6%|▌ | 2665/48008 [09:00<2:25:31, 5.19it/s] 6%|▌ | 2666/48008 [09:00<2:27:31, 5.12it/s] 6%|▌ | 2667/48008 [09:00<2:28:15, 5.10it/s] 6%|▌ | 2668/48008 [09:00<2:26:33, 5.16it/s] 6%|▌ | 2669/48008 [09:01<2:27:04, 5.14it/s] 6%|▌ | 2670/48008 [09:01<2:28:40, 5.08it/s] 6%|▌ | 2671/48008 [09:01<2:26:34, 5.16it/s] 6%|▌ | 2672/48008 [09:01<2:25:55, 5.18it/s] 6%|▌ | 2673/48008 [09:01<2:23:33, 5.26it/s] 6%|▌ | 2674/48008 [09:01<2:25:05, 5.21it/s] 6%|▌ | 2675/48008 [09:02<2:24:56, 5.21it/s] 6%|▌ | 2676/48008 [09:02<2:26:53, 5.14it/s] 6%|▌ | 2677/48008 [09:02<2:52:43, 4.37it/s] 6%|▌ | 2678/48008 [09:02<2:43:46, 4.61it/s] 6%|▌ | 2679/48008 [09:03<2:40:13, 4.72it/s] 6%|▌ | 2680/48008 [09:03<2:34:44, 4.88it/s] 6%|▌ | 2681/48008 [09:03<2:33:48, 4.91it/s] 6%|▌ | 2682/48008 [09:03<2:28:26, 5.09it/s] 6%|▌ | 2683/48008 [09:03<2:37:42, 4.79it/s] 6%|▌ | 2684/48008 [09:04<3:17:52, 3.82it/s] 6%|▌ | 2685/48008 [09:04<3:00:48, 4.18it/s] 6%|▌ | 2686/48008 [09:04<2:49:40, 4.45it/s] 6%|▌ | 2687/48008 [09:04<2:42:28, 4.65it/s] 6%|▌ | 2688/48008 [09:05<2:37:32, 4.79it/s] 6%|▌ | 2689/48008 [09:05<2:34:51, 4.88it/s] 6%|▌ | 2690/48008 [09:05<2:34:15, 4.90it/s] 6%|▌ | 2691/48008 [09:05<2:41:24, 4.68it/s] 6%|▌ | 2692/48008 [09:05<2:35:05, 4.87it/s] 6%|▌ | 2693/48008 [09:06<2:31:45, 4.98it/s] 6%|▌ | 2694/48008 [09:06<2:29:17, 5.06it/s] 6%|▌ | 2695/48008 [09:06<2:37:57, 4.78it/s] 6%|▌ | 2696/48008 [09:06<2:32:40, 4.95it/s] 6%|▌ | 2697/48008 [09:06<2:34:20, 4.89it/s] 6%|▌ | 2698/48008 [09:07<2:57:01, 4.27it/s] 6%|▌ | 2699/48008 [09:07<2:47:56, 4.50it/s] 6%|▌ | 2700/48008 [09:07<2:51:33, 4.40it/s] {'loss': 0.3065, 'grad_norm': 0.158203125, 'learning_rate': 0.00018875604065989003, 'epoch': 0.06} 6%|▌ | 2700/48008 [09:07<2:51:33, 4.40it/s] 6%|▌ | 2701/48008 [09:07<2:43:50, 4.61it/s] 6%|▌ | 2702/48008 [09:07<2:36:48, 4.82it/s] 6%|▌ | 2703/48008 [09:08<2:32:42, 4.94it/s] 6%|▌ | 2704/48008 [09:08<2:30:57, 5.00it/s] 6%|▌ | 2705/48008 [09:08<2:28:31, 5.08it/s] 6%|▌ | 2706/48008 [09:08<2:26:34, 5.15it/s] 6%|▌ | 2707/48008 [09:08<2:24:44, 5.22it/s] 6%|▌ | 2708/48008 [09:09<2:24:34, 5.22it/s] 6%|▌ | 2709/48008 [09:09<2:25:19, 5.20it/s] 6%|▌ | 2710/48008 [09:09<2:25:08, 5.20it/s] 6%|▌ | 2711/48008 [09:09<2:27:40, 5.11it/s] 6%|▌ | 2712/48008 [09:09<2:25:49, 5.18it/s] 6%|▌ | 2713/48008 [09:10<2:25:04, 5.20it/s] 6%|▌ | 2714/48008 [09:10<2:29:01, 5.07it/s] 6%|▌ | 2715/48008 [09:10<2:38:17, 4.77it/s] 6%|▌ | 2716/48008 [09:10<2:33:31, 4.92it/s] 6%|▌ | 2717/48008 [09:10<2:30:16, 5.02it/s] 6%|▌ | 2718/48008 [09:11<2:29:36, 5.05it/s] 6%|▌ | 2719/48008 [09:11<2:27:47, 5.11it/s] 6%|▌ | 2720/48008 [09:11<2:26:21, 5.16it/s] 6%|▌ | 2721/48008 [09:11<2:28:46, 5.07it/s] 6%|▌ | 2722/48008 [09:11<2:26:46, 5.14it/s] 6%|▌ | 2723/48008 [09:12<2:25:44, 5.18it/s] 6%|▌ | 2724/48008 [09:12<2:26:31, 5.15it/s] 6%|▌ | 2725/48008 [09:12<2:25:57, 5.17it/s] 6%|▌ | 2726/48008 [09:12<2:27:19, 5.12it/s] 6%|▌ | 2727/48008 [09:12<2:26:01, 5.17it/s] 6%|▌ | 2728/48008 [09:13<2:26:55, 5.14it/s] 6%|▌ | 2729/48008 [09:13<2:26:23, 5.16it/s] 6%|▌ | 2730/48008 [09:13<2:26:36, 5.15it/s] 6%|▌ | 2731/48008 [09:13<2:25:46, 5.18it/s] 6%|▌ | 2732/48008 [09:13<2:24:32, 5.22it/s] 6%|▌ | 2733/48008 [09:13<2:24:11, 5.23it/s] 6%|▌ | 2734/48008 [09:14<2:34:32, 4.88it/s] 6%|▌ | 2735/48008 [09:14<2:32:59, 4.93it/s] 6%|▌ | 2736/48008 [09:14<2:34:07, 4.90it/s] 6%|▌ | 2737/48008 [09:14<2:32:31, 4.95it/s] 6%|▌ | 2738/48008 [09:15<2:55:51, 4.29it/s] 6%|▌ | 2739/48008 [09:15<2:45:37, 4.56it/s] 6%|▌ | 2740/48008 [09:15<2:41:29, 4.67it/s] 6%|▌ | 2741/48008 [09:15<2:38:49, 4.75it/s] 6%|▌ | 2742/48008 [09:15<2:33:33, 4.91it/s] 6%|▌ | 2743/48008 [09:16<2:33:14, 4.92it/s] 6%|▌ | 2744/48008 [09:16<2:31:49, 4.97it/s] 6%|▌ | 2745/48008 [09:16<2:55:22, 4.30it/s] 6%|▌ | 2746/48008 [09:16<2:48:46, 4.47it/s] 6%|▌ | 2747/48008 [09:16<2:41:12, 4.68it/s] 6%|▌ | 2748/48008 [09:17<3:20:02, 3.77it/s] 6%|▌ | 2749/48008 [09:17<3:03:34, 4.11it/s] 6%|▌ | 2750/48008 [09:17<2:51:50, 4.39it/s] {'loss': 0.3032, 'grad_norm': 0.625, 'learning_rate': 0.00018854774204299284, 'epoch': 0.06} 6%|▌ | 2750/48008 [09:17<2:51:50, 4.39it/s] 6%|▌ | 2751/48008 [09:17<2:43:24, 4.62it/s] 6%|▌ | 2752/48008 [09:18<2:37:28, 4.79it/s] 6%|▌ | 2753/48008 [09:18<2:33:22, 4.92it/s] 6%|▌ | 2754/48008 [09:18<2:29:32, 5.04it/s] 6%|▌ | 2755/48008 [09:18<2:27:47, 5.10it/s] 6%|▌ | 2756/48008 [09:19<2:52:57, 4.36it/s] 6%|▌ | 2757/48008 [09:19<2:43:44, 4.61it/s] 6%|▌ | 2758/48008 [09:19<2:37:39, 4.78it/s] 6%|▌ | 2759/48008 [09:19<2:32:19, 4.95it/s] 6%|▌ | 2760/48008 [09:19<2:32:32, 4.94it/s] 6%|▌ | 2761/48008 [09:19<2:30:09, 5.02it/s] 6%|▌ | 2762/48008 [09:20<2:27:51, 5.10it/s] 6%|▌ | 2763/48008 [09:20<2:29:31, 5.04it/s] 6%|▌ | 2764/48008 [09:20<2:27:25, 5.11it/s] 6%|▌ | 2765/48008 [09:20<2:25:08, 5.20it/s] 6%|▌ | 2766/48008 [09:20<2:24:49, 5.21it/s] 6%|▌ | 2767/48008 [09:21<2:26:15, 5.16it/s] 6%|▌ | 2768/48008 [09:21<2:24:05, 5.23it/s] 6%|▌ | 2769/48008 [09:21<2:26:19, 5.15it/s] 6%|▌ | 2770/48008 [09:21<2:25:09, 5.19it/s] 6%|▌ | 2771/48008 [09:21<2:23:45, 5.24it/s] 6%|▌ | 2772/48008 [09:22<2:22:44, 5.28it/s] 6%|▌ | 2773/48008 [09:22<2:22:50, 5.28it/s] 6%|▌ | 2774/48008 [09:22<2:22:42, 5.28it/s] 6%|▌ | 2775/48008 [09:22<2:23:42, 5.25it/s] 6%|▌ | 2776/48008 [09:22<2:27:29, 5.11it/s] 6%|▌ | 2777/48008 [09:23<2:27:46, 5.10it/s] 6%|▌ | 2778/48008 [09:23<2:27:40, 5.10it/s] 6%|▌ | 2779/48008 [09:23<2:24:58, 5.20it/s] 6%|▌ | 2780/48008 [09:23<2:25:29, 5.18it/s] 6%|▌ | 2781/48008 [09:23<2:27:30, 5.11it/s] 6%|▌ | 2782/48008 [09:24<2:26:08, 5.16it/s] 6%|▌ | 2783/48008 [09:24<5:21:21, 2.35it/s] 6%|▌ | 2784/48008 [09:25<4:32:12, 2.77it/s] 6%|▌ | 2785/48008 [09:25<3:53:20, 3.23it/s] 6%|▌ | 2786/48008 [09:25<3:25:24, 3.67it/s] 6%|▌ | 2787/48008 [09:25<3:06:51, 4.03it/s] 6%|▌ | 2788/48008 [09:26<3:37:58, 3.46it/s] 6%|▌ | 2789/48008 [09:27<6:59:05, 1.80it/s] 6%|▌ | 2790/48008 [09:27<5:39:48, 2.22it/s] 6%|▌ | 2791/48008 [09:27<4:40:51, 2.68it/s] 6%|▌ | 2792/48008 [09:27<4:00:15, 3.14it/s] 6%|▌ | 2793/48008 [09:28<3:31:16, 3.57it/s] 6%|▌ | 2794/48008 [09:28<3:10:37, 3.95it/s] 6%|▌ | 2795/48008 [09:28<2:59:05, 4.21it/s] 6%|▌ | 2796/48008 [09:28<2:48:13, 4.48it/s] 6%|▌ | 2797/48008 [09:28<2:40:15, 4.70it/s] 6%|▌ | 2798/48008 [09:29<2:35:15, 4.85it/s] 6%|▌ | 2799/48008 [09:29<2:32:45, 4.93it/s] 6%|▌ | 2800/48008 [09:29<2:29:42, 5.03it/s]{'loss': 0.3073, 'grad_norm': 0.263671875, 'learning_rate': 0.00018833944342609566, 'epoch': 0.06} 6%|▌ | 2800/48008 [09:29<2:29:42, 5.03it/s] 6%|▌ | 2801/48008 [09:29<2:32:40, 4.94it/s] 6%|▌ | 2802/48008 [09:29<2:30:50, 4.99it/s] 6%|▌ | 2803/48008 [09:30<2:29:37, 5.04it/s] 6%|▌ | 2804/48008 [09:30<2:27:11, 5.12it/s] 6%|▌ | 2805/48008 [09:30<2:26:10, 5.15it/s] 6%|▌ | 2806/48008 [09:30<2:25:47, 5.17it/s] 6%|▌ | 2807/48008 [09:30<2:25:24, 5.18it/s] 6%|▌ | 2808/48008 [09:31<2:28:48, 5.06it/s] 6%|▌ | 2809/48008 [09:31<2:27:14, 5.12it/s] 6%|▌ | 2810/48008 [09:31<2:26:09, 5.15it/s] 6%|▌ | 2811/48008 [09:31<2:25:02, 5.19it/s] 6%|▌ | 2812/48008 [09:31<2:22:44, 5.28it/s] 6%|▌ | 2813/48008 [09:31<2:22:00, 5.30it/s] 6%|▌ | 2814/48008 [09:32<2:22:26, 5.29it/s] 6%|▌ | 2815/48008 [09:32<2:48:58, 4.46it/s] 6%|▌ | 2816/48008 [09:32<2:40:53, 4.68it/s] 6%|▌ | 2817/48008 [09:32<2:35:54, 4.83it/s] 6%|▌ | 2818/48008 [09:33<2:32:31, 4.94it/s] 6%|▌ | 2819/48008 [09:33<2:28:35, 5.07it/s] 6%|▌ | 2820/48008 [09:33<2:31:33, 4.97it/s] 6%|▌ | 2821/48008 [09:33<2:28:54, 5.06it/s] 6%|▌ | 2822/48008 [09:33<2:27:33, 5.10it/s] 6%|▌ | 2823/48008 [09:34<2:26:28, 5.14it/s] 6%|▌ | 2824/48008 [09:34<2:24:31, 5.21it/s] 6%|▌ | 2825/48008 [09:34<2:23:25, 5.25it/s] 6%|▌ | 2826/48008 [09:34<2:24:18, 5.22it/s] 6%|▌ | 2827/48008 [09:34<2:24:22, 5.22it/s] 6%|▌ | 2828/48008 [09:34<2:24:02, 5.23it/s] 6%|▌ | 2829/48008 [09:35<2:26:21, 5.14it/s] 6%|▌ | 2830/48008 [09:35<2:25:16, 5.18it/s] 6%|▌ | 2831/48008 [09:35<2:24:52, 5.20it/s] 6%|▌ | 2832/48008 [09:35<2:35:09, 4.85it/s] 6%|▌ | 2833/48008 [09:35<2:29:26, 5.04it/s] 6%|▌ | 2834/48008 [09:36<2:30:28, 5.00it/s] 6%|▌ | 2835/48008 [09:36<2:27:17, 5.11it/s] 6%|▌ | 2836/48008 [09:36<2:25:44, 5.17it/s] 6%|▌ | 2837/48008 [09:36<2:28:15, 5.08it/s] 6%|▌ | 2838/48008 [09:36<2:27:58, 5.09it/s] 6%|▌ | 2839/48008 [09:37<2:28:36, 5.07it/s] 6%|▌ | 2840/48008 [09:37<2:31:44, 4.96it/s] 6%|▌ | 2841/48008 [09:37<2:33:05, 4.92it/s] 6%|▌ | 2842/48008 [09:37<2:29:16, 5.04it/s] 6%|▌ | 2843/48008 [09:37<2:28:06, 5.08it/s] 6%|▌ | 2844/48008 [09:38<2:27:05, 5.12it/s] 6%|▌ | 2845/48008 [09:38<2:28:13, 5.08it/s] 6%|▌ | 2846/48008 [09:38<2:27:21, 5.11it/s] 6%|▌ | 2847/48008 [09:38<2:28:01, 5.08it/s] 6%|▌ | 2848/48008 [09:38<2:25:29, 5.17it/s] 6%|▌ | 2849/48008 [09:39<2:24:11, 5.22it/s] 6%|▌ | 2850/48008 [09:39<2:23:52, 5.23it/s] {'loss': 0.3065, 'grad_norm': 0.80859375, 'learning_rate': 0.00018813114480919848, 'epoch': 0.06} 6%|▌ | 2850/48008 [09:39<2:23:52, 5.23it/s] 6%|▌ | 2851/48008 [09:39<2:22:39, 5.28it/s] 6%|▌ | 2852/48008 [09:39<2:27:24, 5.11it/s] 6%|▌ | 2853/48008 [09:39<2:26:35, 5.13it/s] 6%|▌ | 2854/48008 [09:40<2:25:28, 5.17it/s] 6%|▌ | 2855/48008 [09:40<2:24:57, 5.19it/s] 6%|▌ | 2856/48008 [09:40<2:24:37, 5.20it/s] 6%|▌ | 2857/48008 [09:40<2:23:15, 5.25it/s] 6%|▌ | 2858/48008 [09:40<2:23:03, 5.26it/s] 6%|▌ | 2859/48008 [09:41<2:22:09, 5.29it/s] 6%|▌ | 2860/48008 [09:41<2:32:50, 4.92it/s] 6%|▌ | 2861/48008 [09:41<2:31:22, 4.97it/s] 6%|▌ | 2862/48008 [09:41<2:28:27, 5.07it/s] 6%|▌ | 2863/48008 [09:41<2:27:16, 5.11it/s] 6%|▌ | 2864/48008 [09:42<2:28:07, 5.08it/s] 6%|▌ | 2865/48008 [09:42<2:25:41, 5.16it/s] 6%|▌ | 2866/48008 [09:42<2:23:47, 5.23it/s] 6%|▌ | 2867/48008 [09:42<2:23:41, 5.24it/s] 6%|▌ | 2868/48008 [09:42<2:23:31, 5.24it/s] 6%|▌ | 2869/48008 [09:42<2:22:02, 5.30it/s] 6%|▌ | 2870/48008 [09:43<2:22:55, 5.26it/s] 6%|▌ | 2871/48008 [09:43<2:21:36, 5.31it/s] 6%|▌ | 2872/48008 [09:43<2:32:25, 4.94it/s] 6%|▌ | 2873/48008 [09:43<2:29:07, 5.04it/s] 6%|▌ | 2874/48008 [09:43<2:29:01, 5.05it/s] 6%|▌ | 2875/48008 [09:44<2:31:08, 4.98it/s] 6%|▌ | 2876/48008 [09:44<2:28:24, 5.07it/s] 6%|▌ | 2877/48008 [09:44<2:26:42, 5.13it/s] 6%|▌ | 2878/48008 [09:44<2:26:50, 5.12it/s] 6%|▌ | 2879/48008 [09:44<2:25:57, 5.15it/s] 6%|▌ | 2880/48008 [09:45<2:25:00, 5.19it/s] 6%|▌ | 2881/48008 [09:45<2:24:17, 5.21it/s] 6%|▌ | 2882/48008 [09:45<2:23:38, 5.24it/s] 6%|▌ | 2883/48008 [09:45<2:25:17, 5.18it/s] 6%|▌ | 2884/48008 [09:45<2:28:41, 5.06it/s] 6%|▌ | 2885/48008 [09:46<2:26:32, 5.13it/s] 6%|▌ | 2886/48008 [09:46<2:29:39, 5.02it/s] 6%|▌ | 2887/48008 [09:46<2:27:27, 5.10it/s] 6%|▌ | 2888/48008 [09:46<2:28:35, 5.06it/s] 6%|▌ | 2889/48008 [09:46<2:24:46, 5.19it/s] 6%|▌ | 2890/48008 [09:47<2:25:59, 5.15it/s] 6%|▌ | 2891/48008 [09:47<2:29:18, 5.04it/s] 6%|▌ | 2892/48008 [09:47<2:28:02, 5.08it/s] 6%|▌ | 2893/48008 [09:47<2:26:14, 5.14it/s] 6%|▌ | 2894/48008 [09:47<2:25:11, 5.18it/s] 6%|▌ | 2895/48008 [09:48<2:23:47, 5.23it/s] 6%|▌ | 2896/48008 [09:48<2:22:29, 5.28it/s] 6%|▌ | 2897/48008 [09:48<2:22:57, 5.26it/s] 6%|▌ | 2898/48008 [09:48<2:23:06, 5.25it/s] 6%|▌ | 2899/48008 [09:48<2:26:08, 5.14it/s] 6%|▌ | 2900/48008 [09:48<2:24:43, 5.19it/s] {'loss': 0.3093, 'grad_norm': 0.177734375, 'learning_rate': 0.0001879228461923013, 'epoch': 0.06} 6%|▌ | 2900/48008 [09:49<2:24:43, 5.19it/s] 6%|▌ | 2901/48008 [09:49<2:23:54, 5.22it/s] 6%|▌ | 2902/48008 [09:49<2:24:53, 5.19it/s] 6%|▌ | 2903/48008 [09:49<2:24:38, 5.20it/s] 6%|▌ | 2904/48008 [09:49<2:24:36, 5.20it/s] 6%|▌ | 2905/48008 [09:49<2:23:13, 5.25it/s] 6%|▌ | 2906/48008 [09:50<2:24:46, 5.19it/s] 6%|▌ | 2907/48008 [09:50<2:26:33, 5.13it/s] 6%|▌ | 2908/48008 [09:50<2:25:25, 5.17it/s] 6%|▌ | 2909/48008 [09:50<2:25:14, 5.18it/s] 6%|▌ | 2910/48008 [09:50<2:24:37, 5.20it/s] 6%|▌ | 2911/48008 [09:51<2:24:51, 5.19it/s] 6%|▌ | 2912/48008 [09:51<2:23:43, 5.23it/s] 6%|▌ | 2913/48008 [09:51<2:23:18, 5.24it/s] 6%|▌ | 2914/48008 [09:51<2:23:46, 5.23it/s] 6%|▌ | 2915/48008 [09:51<2:22:48, 5.26it/s] 6%|▌ | 2916/48008 [09:52<2:24:14, 5.21it/s] 6%|▌ | 2917/48008 [09:52<2:25:01, 5.18it/s] 6%|▌ | 2918/48008 [09:52<2:23:51, 5.22it/s] 6%|▌ | 2919/48008 [09:52<3:07:29, 4.01it/s] 6%|▌ | 2920/48008 [09:53<2:52:21, 4.36it/s] 6%|▌ | 2921/48008 [09:53<2:54:19, 4.31it/s] 6%|▌ | 2922/48008 [09:53<2:47:26, 4.49it/s] 6%|▌ | 2923/48008 [09:53<2:40:04, 4.69it/s] 6%|▌ | 2924/48008 [09:53<2:37:42, 4.76it/s] 6%|▌ | 2925/48008 [09:54<2:34:53, 4.85it/s] 6%|▌ | 2926/48008 [09:54<2:41:34, 4.65it/s] 6%|▌ | 2927/48008 [09:54<2:35:45, 4.82it/s] 6%|▌ | 2928/48008 [09:54<2:32:05, 4.94it/s] 6%|▌ | 2929/48008 [09:54<2:29:29, 5.03it/s] 6%|▌ | 2930/48008 [09:55<2:27:33, 5.09it/s] 6%|▌ | 2931/48008 [09:55<2:25:26, 5.17it/s] 6%|▌ | 2932/48008 [09:55<2:24:38, 5.19it/s] 6%|▌ | 2933/48008 [09:55<2:23:23, 5.24it/s] 6%|▌ | 2934/48008 [09:55<2:22:55, 5.26it/s] 6%|▌ | 2935/48008 [09:55<2:24:16, 5.21it/s] 6%|▌ | 2936/48008 [09:56<2:22:27, 5.27it/s] 6%|▌ | 2937/48008 [09:56<2:22:03, 5.29it/s] 6%|▌ | 2938/48008 [09:56<2:22:05, 5.29it/s] 6%|▌ | 2939/48008 [09:56<2:21:38, 5.30it/s] 6%|▌ | 2940/48008 [09:56<2:22:42, 5.26it/s] 6%|▌ | 2941/48008 [09:57<2:22:35, 5.27it/s] 6%|▌ | 2942/48008 [09:57<2:22:58, 5.25it/s] 6%|▌ | 2943/48008 [09:57<2:22:34, 5.27it/s] 6%|▌ | 2944/48008 [09:57<2:21:10, 5.32it/s] 6%|▌ | 2945/48008 [09:57<2:21:43, 5.30it/s] 6%|▌ | 2946/48008 [09:58<2:23:15, 5.24it/s] 6%|▌ | 2947/48008 [09:58<2:22:35, 5.27it/s] 6%|▌ | 2948/48008 [09:58<2:25:53, 5.15it/s] 6%|▌ | 2949/48008 [09:58<2:24:05, 5.21it/s] 6%|▌ | 2950/48008 [09:58<2:23:45, 5.22it/s] {'loss': 0.3072, 'grad_norm': 0.1591796875, 'learning_rate': 0.00018771454757540409, 'epoch': 0.06} 6%|▌ | 2950/48008 [09:58<2:23:45, 5.22it/s] 6%|▌ | 2951/48008 [09:59<2:24:05, 5.21it/s] 6%|▌ | 2952/48008 [09:59<2:34:23, 4.86it/s] 6%|▌ | 2953/48008 [09:59<2:30:59, 4.97it/s] 6%|▌ | 2954/48008 [09:59<2:27:39, 5.09it/s] 6%|▌ | 2955/48008 [09:59<2:26:00, 5.14it/s] 6%|▌ | 2956/48008 [10:00<2:24:40, 5.19it/s] 6%|▌ | 2957/48008 [10:00<2:34:24, 4.86it/s] 6%|▌ | 2958/48008 [10:00<2:31:35, 4.95it/s] 6%|▌ | 2959/48008 [10:00<2:31:42, 4.95it/s] 6%|▌ | 2960/48008 [10:00<2:28:52, 5.04it/s] 6%|▌ | 2961/48008 [10:01<2:28:25, 5.06it/s] 6%|▌ | 2962/48008 [10:01<2:30:15, 5.00it/s] 6%|▌ | 2963/48008 [10:01<2:54:16, 4.31it/s] 6%|▌ | 2964/48008 [10:01<2:44:21, 4.57it/s] 6%|▌ | 2965/48008 [10:01<2:40:25, 4.68it/s] 6%|▌ | 2966/48008 [10:02<2:34:16, 4.87it/s] 6%|▌ | 2967/48008 [10:03<5:06:19, 2.45it/s] 6%|▌ | 2968/48008 [10:03<4:17:14, 2.92it/s] 6%|▌ | 2969/48008 [10:03<3:45:34, 3.33it/s] 6%|▌ | 2970/48008 [10:03<3:21:04, 3.73it/s] 6%|▌ | 2971/48008 [10:03<3:03:53, 4.08it/s] 6%|▌ | 2972/48008 [10:03<2:51:38, 4.37it/s] 6%|▌ | 2973/48008 [10:04<2:44:07, 4.57it/s] 6%|▌ | 2974/48008 [10:04<2:39:01, 4.72it/s] 6%|▌ | 2975/48008 [10:04<2:35:53, 4.81it/s] 6%|▌ | 2976/48008 [10:04<2:32:03, 4.94it/s] 6%|▌ | 2977/48008 [10:04<2:29:15, 5.03it/s] 6%|▌ | 2978/48008 [10:05<2:28:13, 5.06it/s] 6%|▌ | 2979/48008 [10:05<2:27:07, 5.10it/s] 6%|▌ | 2980/48008 [10:05<2:25:39, 5.15it/s] 6%|▌ | 2981/48008 [10:05<2:26:14, 5.13it/s] 6%|▌ | 2982/48008 [10:05<2:29:28, 5.02it/s] 6%|▌ | 2983/48008 [10:06<2:32:11, 4.93it/s] 6%|▌ | 2984/48008 [10:06<2:29:20, 5.02it/s] 6%|▌ | 2985/48008 [10:06<2:26:54, 5.11it/s] 6%|▌ | 2986/48008 [10:06<2:27:22, 5.09it/s] 6%|▌ | 2987/48008 [10:06<2:26:26, 5.12it/s] 6%|▌ | 2988/48008 [10:07<2:25:12, 5.17it/s] 6%|▌ | 2989/48008 [10:07<2:23:21, 5.23it/s] 6%|▌ | 2990/48008 [10:07<2:23:43, 5.22it/s] 6%|▌ | 2991/48008 [10:07<2:25:46, 5.15it/s] 6%|▌ | 2992/48008 [10:07<2:23:43, 5.22it/s] 6%|▌ | 2993/48008 [10:08<2:26:09, 5.13it/s] 6%|▌ | 2994/48008 [10:08<2:27:53, 5.07it/s] 6%|▌ | 2995/48008 [10:08<2:26:07, 5.13it/s] 6%|▌ | 2996/48008 [10:08<2:35:35, 4.82it/s] 6%|▌ | 2997/48008 [10:08<2:31:50, 4.94it/s] 6%|▌ | 2998/48008 [10:09<2:31:35, 4.95it/s] 6%|▌ | 2999/48008 [10:09<2:28:28, 5.05it/s] 6%|▌ | 3000/48008 [10:09<2:36:41, 4.79it/s] {'loss': 0.3042, 'grad_norm': 0.1943359375, 'learning_rate': 0.00018750624895850693, 'epoch': 0.06} 6%|▌ | 3000/48008 [10:09<2:36:41, 4.79it/s] 6%|▋ | 3001/48008 [10:09<2:32:31, 4.92it/s] 6%|▋ | 3002/48008 [10:09<2:29:26, 5.02it/s] 6%|▋ | 3003/48008 [10:10<2:31:46, 4.94it/s] 6%|▋ | 3004/48008 [10:10<2:28:09, 5.06it/s] 6%|▋ | 3005/48008 [10:10<2:26:36, 5.12it/s] 6%|▋ | 3006/48008 [10:10<2:24:39, 5.19it/s] 6%|▋ | 3007/48008 [10:10<2:23:59, 5.21it/s] 6%|▋ | 3008/48008 [10:11<2:25:15, 5.16it/s] 6%|▋ | 3009/48008 [10:11<2:25:39, 5.15it/s] 6%|▋ | 3010/48008 [10:11<2:25:09, 5.17it/s] 6%|▋ | 3011/48008 [10:11<2:23:12, 5.24it/s] 6%|▋ | 3012/48008 [10:11<2:22:32, 5.26it/s] 6%|▋ | 3013/48008 [10:11<2:21:21, 5.31it/s] 6%|▋ | 3014/48008 [10:12<2:31:56, 4.94it/s] 6%|▋ | 3015/48008 [10:12<2:29:31, 5.01it/s] 6%|▋ | 3016/48008 [10:12<2:38:06, 4.74it/s] 6%|▋ | 3017/48008 [10:12<2:59:19, 4.18it/s] 6%|▋ | 3018/48008 [10:13<2:47:39, 4.47it/s] 6%|▋ | 3019/48008 [10:13<2:40:38, 4.67it/s] 6%|▋ | 3020/48008 [10:13<2:32:36, 4.91it/s] 6%|▋ | 3021/48008 [10:13<2:29:39, 5.01it/s] 6%|▋ | 3022/48008 [10:13<2:27:23, 5.09it/s] 6%|▋ | 3023/48008 [10:14<2:24:35, 5.19it/s] 6%|▋ | 3024/48008 [10:14<3:07:42, 3.99it/s] 6%|▋ | 3025/48008 [10:14<2:53:02, 4.33it/s] 6%|▋ | 3026/48008 [10:14<2:43:14, 4.59it/s] 6%|▋ | 3027/48008 [10:15<2:36:20, 4.80it/s] 6%|▋ | 3028/48008 [10:15<2:31:37, 4.94it/s] 6%|▋ | 3029/48008 [10:15<2:29:04, 5.03it/s] 6%|▋ | 3030/48008 [10:15<2:27:12, 5.09it/s] 6%|▋ | 3031/48008 [10:15<2:24:48, 5.18it/s] 6%|▋ | 3032/48008 [10:15<2:23:43, 5.22it/s] 6%|▋ | 3033/48008 [10:16<2:23:17, 5.23it/s] 6%|▋ | 3034/48008 [10:16<2:22:32, 5.26it/s] 6%|▋ | 3035/48008 [10:16<2:21:31, 5.30it/s] 6%|▋ | 3036/48008 [10:16<2:21:57, 5.28it/s] 6%|▋ | 3037/48008 [10:16<2:22:30, 5.26it/s] 6%|▋ | 3038/48008 [10:17<2:26:32, 5.11it/s] 6%|▋ | 3039/48008 [10:17<2:24:56, 5.17it/s] 6%|▋ | 3040/48008 [10:17<2:24:10, 5.20it/s] 6%|▋ | 3041/48008 [10:17<2:34:30, 4.85it/s] 6%|▋ | 3042/48008 [10:17<2:30:35, 4.98it/s] 6%|▋ | 3043/48008 [10:18<2:30:00, 5.00it/s] 6%|▋ | 3044/48008 [10:18<2:28:11, 5.06it/s] 6%|▋ | 3045/48008 [10:18<2:30:29, 4.98it/s] 6%|▋ | 3046/48008 [10:18<2:38:35, 4.73it/s] 6%|▋ | 3047/48008 [10:18<2:34:09, 4.86it/s] 6%|▋ | 3048/48008 [10:19<2:29:58, 5.00it/s] 6%|▋ | 3049/48008 [10:19<2:29:30, 5.01it/s] 6%|▋ | 3050/48008 [10:19<2:27:08, 5.09it/s] {'loss': 0.3056, 'grad_norm': 0.203125, 'learning_rate': 0.00018729795034160975, 'epoch': 0.06} 6%|▋ | 3050/48008 [10:19<2:27:08, 5.09it/s] 6%|▋ | 3051/48008 [10:19<2:26:58, 5.10it/s] 6%|▋ | 3052/48008 [10:19<2:25:05, 5.16it/s] 6%|▋ | 3053/48008 [10:20<2:23:16, 5.23it/s] 6%|▋ | 3054/48008 [10:20<2:24:28, 5.19it/s] 6%|▋ | 3055/48008 [10:20<2:26:22, 5.12it/s] 6%|▋ | 3056/48008 [10:20<2:29:19, 5.02it/s] 6%|▋ | 3057/48008 [10:20<2:26:34, 5.11it/s] 6%|▋ | 3058/48008 [10:21<2:24:58, 5.17it/s] 6%|▋ | 3059/48008 [10:21<2:23:56, 5.20it/s] 6%|▋ | 3060/48008 [10:21<2:23:39, 5.21it/s] 6%|▋ | 3061/48008 [10:21<2:23:13, 5.23it/s] 6%|▋ | 3062/48008 [10:21<2:22:02, 5.27it/s] 6%|▋ | 3063/48008 [10:22<2:24:37, 5.18it/s] 6%|▋ | 3064/48008 [10:22<2:23:59, 5.20it/s] 6%|▋ | 3065/48008 [10:22<2:22:44, 5.25it/s] 6%|▋ | 3066/48008 [10:22<2:23:50, 5.21it/s] 6%|▋ | 3067/48008 [10:22<2:23:17, 5.23it/s] 6%|▋ | 3068/48008 [10:22<2:22:56, 5.24it/s] 6%|▋ | 3069/48008 [10:23<2:24:25, 5.19it/s] 6%|▋ | 3070/48008 [10:23<2:23:30, 5.22it/s] 6%|▋ | 3071/48008 [10:23<2:22:42, 5.25it/s] 6%|▋ | 3072/48008 [10:23<2:23:07, 5.23it/s] 6%|▋ | 3073/48008 [10:23<2:22:48, 5.24it/s] 6%|▋ | 3074/48008 [10:24<2:22:45, 5.25it/s] 6%|▋ | 3075/48008 [10:24<2:22:34, 5.25it/s] 6%|▋ | 3076/48008 [10:24<2:33:05, 4.89it/s] 6%|▋ | 3077/48008 [10:24<2:29:55, 4.99it/s] 6%|▋ | 3078/48008 [10:24<2:29:26, 5.01it/s] 6%|▋ | 3079/48008 [10:25<2:26:44, 5.10it/s] 6%|▋ | 3080/48008 [10:25<2:24:17, 5.19it/s] 6%|▋ | 3081/48008 [10:25<2:23:55, 5.20it/s] 6%|▋ | 3082/48008 [10:25<2:22:52, 5.24it/s] 6%|▋ | 3083/48008 [10:25<2:22:26, 5.26it/s] 6%|▋ | 3084/48008 [10:26<2:20:23, 5.33it/s] 6%|▋ | 3085/48008 [10:26<2:23:04, 5.23it/s] 6%|▋ | 3086/48008 [10:26<2:23:03, 5.23it/s] 6%|▋ | 3087/48008 [10:26<2:23:21, 5.22it/s] 6%|▋ | 3088/48008 [10:26<2:22:55, 5.24it/s] 6%|▋ | 3089/48008 [10:27<2:22:36, 5.25it/s] 6%|▋ | 3090/48008 [10:27<2:25:22, 5.15it/s] 6%|▋ | 3091/48008 [10:27<2:29:07, 5.02it/s] 6%|▋ | 3092/48008 [10:27<2:30:18, 4.98it/s] 6%|▋ | 3093/48008 [10:27<2:29:14, 5.02it/s] 6%|▋ | 3094/48008 [10:28<2:29:59, 4.99it/s] 6%|▋ | 3095/48008 [10:28<2:31:48, 4.93it/s] 6%|▋ | 3096/48008 [10:28<2:30:41, 4.97it/s] 6%|▋ | 3097/48008 [10:28<2:27:23, 5.08it/s] 6%|▋ | 3098/48008 [10:28<2:29:09, 5.02it/s] 6%|▋ | 3099/48008 [10:29<2:26:57, 5.09it/s] 6%|▋ | 3100/48008 [10:29<2:25:37, 5.14it/s] {'loss': 0.3079, 'grad_norm': 0.181640625, 'learning_rate': 0.00018708965172471256, 'epoch': 0.06} 6%|▋ | 3100/48008 [10:29<2:25:37, 5.14it/s] 6%|▋ | 3101/48008 [10:29<2:25:52, 5.13it/s] 6%|▋ | 3102/48008 [10:29<3:08:58, 3.96it/s] 6%|▋ | 3103/48008 [10:30<2:55:21, 4.27it/s] 6%|▋ | 3104/48008 [10:30<2:44:46, 4.54it/s] 6%|▋ | 3105/48008 [10:30<2:41:14, 4.64it/s] 6%|▋ | 3106/48008 [10:30<3:01:38, 4.12it/s] 6%|▋ | 3107/48008 [10:30<2:49:41, 4.41it/s] 6%|▋ | 3108/48008 [10:31<2:41:26, 4.64it/s] 6%|▋ | 3109/48008 [10:31<3:19:28, 3.75it/s] 6%|▋ | 3110/48008 [10:31<3:03:53, 4.07it/s] 6%|▋ | 3111/48008 [10:31<2:51:14, 4.37it/s] 6%|▋ | 3112/48008 [10:32<2:42:42, 4.60it/s] 6%|▋ | 3113/48008 [10:32<2:37:24, 4.75it/s] 6%|▋ | 3114/48008 [10:32<2:32:41, 4.90it/s] 6%|▋ | 3115/48008 [10:32<2:29:14, 5.01it/s] 6%|▋ | 3116/48008 [10:32<2:26:34, 5.10it/s] 6%|▋ | 3117/48008 [10:33<2:25:12, 5.15it/s] 6%|▋ | 3118/48008 [10:33<2:24:36, 5.17it/s] 6%|▋ | 3119/48008 [10:33<2:23:46, 5.20it/s] 6%|▋ | 3120/48008 [10:33<2:23:45, 5.20it/s] 7%|▋ | 3121/48008 [10:33<2:24:33, 5.18it/s] 7%|▋ | 3122/48008 [10:33<2:22:46, 5.24it/s] 7%|▋ | 3123/48008 [10:34<2:20:31, 5.32it/s] 7%|▋ | 3124/48008 [10:34<2:21:10, 5.30it/s] 7%|▋ | 3125/48008 [10:34<2:20:44, 5.31it/s] 7%|▋ | 3126/48008 [10:34<2:22:20, 5.26it/s] 7%|▋ | 3127/48008 [10:35<2:48:09, 4.45it/s] 7%|▋ | 3128/48008 [10:35<2:44:35, 4.54it/s] 7%|▋ | 3129/48008 [10:35<2:41:59, 4.62it/s] 7%|▋ | 3130/48008 [10:35<2:40:28, 4.66it/s] 7%|▋ | 3131/48008 [10:35<2:34:59, 4.83it/s] 7%|▋ | 3132/48008 [10:36<2:34:22, 4.85it/s] 7%|▋ | 3133/48008 [10:36<2:30:23, 4.97it/s] 7%|▋ | 3134/48008 [10:36<2:28:23, 5.04it/s] 7%|▋ | 3135/48008 [10:36<2:29:19, 5.01it/s] 7%|▋ | 3136/48008 [10:37<3:11:37, 3.90it/s] 7%|▋ | 3137/48008 [10:37<2:56:37, 4.23it/s] 7%|▋ | 3138/48008 [10:37<2:51:57, 4.35it/s] 7%|▋ | 3139/48008 [10:37<2:47:38, 4.46it/s] 7%|▋ | 3140/48008 [10:37<2:51:10, 4.37it/s] 7%|▋ | 3141/48008 [10:38<2:42:41, 4.60it/s] 7%|▋ | 3142/48008 [10:38<2:38:09, 4.73it/s] 7%|▋ | 3143/48008 [10:38<2:34:12, 4.85it/s] 7%|▋ | 3144/48008 [10:38<2:29:57, 4.99it/s] 7%|▋ | 3145/48008 [10:38<2:27:54, 5.06it/s] 7%|▋ | 3146/48008 [10:39<2:26:14, 5.11it/s] 7%|▋ | 3147/48008 [10:39<2:24:15, 5.18it/s] 7%|▋ | 3148/48008 [10:39<2:26:16, 5.11it/s] 7%|▋ | 3149/48008 [10:39<2:25:18, 5.15it/s] 7%|▋ | 3150/48008 [10:39<2:24:57, 5.16it/s] {'loss': 0.3045, 'grad_norm': 0.64453125, 'learning_rate': 0.00018688135310781538, 'epoch': 0.07} 7%|▋ | 3150/48008 [10:39<2:24:57, 5.16it/s] 7%|▋ | 3151/48008 [10:39<2:23:24, 5.21it/s] 7%|▋ | 3152/48008 [10:40<2:24:52, 5.16it/s] 7%|▋ | 3153/48008 [10:40<2:27:29, 5.07it/s] 7%|▋ | 3154/48008 [10:40<2:26:01, 5.12it/s] 7%|▋ | 3155/48008 [10:40<2:22:59, 5.23it/s] 7%|▋ | 3156/48008 [10:40<2:33:37, 4.87it/s] 7%|▋ | 3157/48008 [10:41<2:33:27, 4.87it/s] 7%|▋ | 3158/48008 [10:41<2:56:06, 4.24it/s] 7%|▋ | 3159/48008 [10:41<2:45:47, 4.51it/s] 7%|▋ | 3160/48008 [10:41<2:37:59, 4.73it/s] 7%|▋ | 3161/48008 [10:42<2:32:52, 4.89it/s] 7%|▋ | 3162/48008 [10:42<2:29:32, 5.00it/s] 7%|▋ | 3163/48008 [10:42<2:27:00, 5.08it/s] 7%|▋ | 3164/48008 [10:42<2:26:22, 5.11it/s] 7%|▋ | 3165/48008 [10:42<2:25:07, 5.15it/s] 7%|▋ | 3166/48008 [10:43<2:23:58, 5.19it/s] 7%|▋ | 3167/48008 [10:43<2:22:30, 5.24it/s] 7%|▋ | 3168/48008 [10:43<2:22:57, 5.23it/s] 7%|▋ | 3169/48008 [10:43<2:22:25, 5.25it/s] 7%|▋ | 3170/48008 [10:43<2:24:46, 5.16it/s] 7%|▋ | 3171/48008 [10:43<2:24:30, 5.17it/s] 7%|▋ | 3172/48008 [10:44<2:23:32, 5.21it/s] 7%|▋ | 3173/48008 [10:44<2:24:35, 5.17it/s] 7%|▋ | 3174/48008 [10:44<2:24:10, 5.18it/s] 7%|▋ | 3175/48008 [10:44<2:26:05, 5.11it/s] 7%|▋ | 3176/48008 [10:44<2:28:07, 5.04it/s] 7%|▋ | 3177/48008 [10:45<2:25:57, 5.12it/s] 7%|▋ | 3178/48008 [10:45<2:24:20, 5.18it/s] 7%|▋ | 3179/48008 [10:45<2:23:51, 5.19it/s] 7%|▋ | 3180/48008 [10:45<2:33:12, 4.88it/s] 7%|▋ | 3181/48008 [10:45<2:29:34, 5.00it/s] 7%|▋ | 3182/48008 [10:46<2:26:00, 5.12it/s] 7%|▋ | 3183/48008 [10:46<2:24:40, 5.16it/s] 7%|▋ | 3184/48008 [10:46<2:23:59, 5.19it/s] 7%|▋ | 3185/48008 [10:46<2:33:52, 4.86it/s] 7%|▋ | 3186/48008 [10:46<2:31:52, 4.92it/s] 7%|▋ | 3187/48008 [10:47<2:29:16, 5.00it/s] 7%|▋ | 3188/48008 [10:47<2:37:33, 4.74it/s] 7%|▋ | 3189/48008 [10:47<2:37:00, 4.76it/s] 7%|▋ | 3190/48008 [10:47<2:32:34, 4.90it/s] 7%|▋ | 3191/48008 [10:47<2:29:16, 5.00it/s] 7%|▋ | 3192/48008 [10:48<2:27:34, 5.06it/s] 7%|▋ | 3193/48008 [10:48<2:26:56, 5.08it/s] 7%|▋ | 3194/48008 [10:48<2:25:19, 5.14it/s] 7%|▋ | 3195/48008 [10:48<2:24:10, 5.18it/s] 7%|▋ | 3196/48008 [10:48<2:23:26, 5.21it/s] 7%|▋ | 3197/48008 [10:49<2:23:00, 5.22it/s] 7%|▋ | 3198/48008 [10:49<2:25:21, 5.14it/s] 7%|▋ | 3199/48008 [10:49<2:23:57, 5.19it/s] 7%|▋ | 3200/48008 [10:49<2:23:28, 5.20it/s] {'loss': 0.3058, 'grad_norm': 0.1533203125, 'learning_rate': 0.0001866730544909182, 'epoch': 0.07} 7%|▋ | 3200/48008 [10:49<2:23:28, 5.20it/s] 7%|▋ | 3201/48008 [10:49<2:23:28, 5.21it/s] 7%|▋ | 3202/48008 [10:50<2:22:39, 5.23it/s] 7%|▋ | 3203/48008 [10:50<2:32:50, 4.89it/s] 7%|▋ | 3204/48008 [10:50<2:29:53, 4.98it/s] 7%|▋ | 3205/48008 [10:50<2:27:03, 5.08it/s] 7%|▋ | 3206/48008 [10:50<2:25:20, 5.14it/s] 7%|▋ | 3207/48008 [10:51<2:24:47, 5.16it/s] 7%|▋ | 3208/48008 [10:51<2:25:49, 5.12it/s] 7%|▋ | 3209/48008 [10:51<2:26:14, 5.11it/s] 7%|▋ | 3210/48008 [10:51<2:26:06, 5.11it/s] 7%|▋ | 3211/48008 [10:51<2:24:51, 5.15it/s] 7%|▋ | 3212/48008 [10:52<2:23:23, 5.21it/s] 7%|▋ | 3213/48008 [10:52<2:23:15, 5.21it/s] 7%|▋ | 3214/48008 [10:52<2:25:56, 5.12it/s] 7%|▋ | 3215/48008 [10:52<2:24:55, 5.15it/s] 7%|▋ | 3216/48008 [10:52<2:25:43, 5.12it/s] 7%|▋ | 3217/48008 [10:53<2:24:25, 5.17it/s] 7%|▋ | 3218/48008 [10:53<2:23:29, 5.20it/s] 7%|▋ | 3219/48008 [10:53<2:22:59, 5.22it/s] 7%|▋ | 3220/48008 [10:53<2:25:19, 5.14it/s] 7%|▋ | 3221/48008 [10:53<2:27:11, 5.07it/s] 7%|▋ | 3222/48008 [10:53<2:25:21, 5.13it/s] 7%|▋ | 3223/48008 [10:54<2:25:40, 5.12it/s] 7%|▋ | 3224/48008 [10:54<3:08:38, 3.96it/s] 7%|▋ | 3225/48008 [10:54<2:57:13, 4.21it/s] 7%|▋ | 3226/48008 [10:54<2:48:09, 4.44it/s] 7%|▋ | 3227/48008 [10:55<2:39:57, 4.67it/s] 7%|▋ | 3228/48008 [10:55<2:33:30, 4.86it/s] 7%|▋ | 3229/48008 [10:55<2:32:57, 4.88it/s] 7%|▋ | 3230/48008 [10:55<3:13:28, 3.86it/s] 7%|▋ | 3231/48008 [10:56<2:57:58, 4.19it/s] 7%|▋ | 3232/48008 [10:56<2:46:40, 4.48it/s] 7%|▋ | 3233/48008 [10:56<2:38:11, 4.72it/s] 7%|▋ | 3234/48008 [10:56<2:32:41, 4.89it/s] 7%|▋ | 3235/48008 [10:56<2:28:57, 5.01it/s] 7%|▋ | 3236/48008 [10:57<2:27:00, 5.08it/s] 7%|▋ | 3237/48008 [10:57<2:25:33, 5.13it/s] 7%|▋ | 3238/48008 [10:57<2:25:16, 5.14it/s] 7%|▋ | 3239/48008 [10:57<2:24:04, 5.18it/s] 7%|▋ | 3240/48008 [10:57<2:23:43, 5.19it/s] 7%|▋ | 3241/48008 [10:58<2:23:22, 5.20it/s] 7%|▋ | 3242/48008 [10:58<2:23:10, 5.21it/s] 7%|▋ | 3243/48008 [10:58<2:23:15, 5.21it/s] 7%|▋ | 3244/48008 [10:58<2:23:17, 5.21it/s] 7%|▋ | 3245/48008 [10:58<2:22:39, 5.23it/s] 7%|▋ | 3246/48008 [10:58<2:21:23, 5.28it/s] 7%|▋ | 3247/48008 [10:59<2:21:43, 5.26it/s] 7%|▋ | 3248/48008 [10:59<2:22:01, 5.25it/s] 7%|▋ | 3249/48008 [10:59<2:21:21, 5.28it/s] 7%|▋ | 3250/48008 [10:59<2:21:52, 5.26it/s] {'loss': 0.3044, 'grad_norm': 0.1435546875, 'learning_rate': 0.00018646475587402102, 'epoch': 0.07} 7%|▋ | 3250/48008 [10:59<2:21:52, 5.26it/s] 7%|▋ | 3251/48008 [10:59<2:21:47, 5.26it/s] 7%|▋ | 3252/48008 [11:00<2:20:47, 5.30it/s] 7%|▋ | 3253/48008 [11:00<2:24:02, 5.18it/s] 7%|▋ | 3254/48008 [11:00<2:23:08, 5.21it/s] 7%|▋ | 3255/48008 [11:00<2:23:08, 5.21it/s] 7%|▋ | 3256/48008 [11:00<2:22:41, 5.23it/s] 7%|▋ | 3257/48008 [11:01<2:22:53, 5.22it/s] 7%|▋ | 3258/48008 [11:01<2:25:24, 5.13it/s] 7%|▋ | 3259/48008 [11:01<2:28:15, 5.03it/s] 7%|▋ | 3260/48008 [11:01<3:10:05, 3.92it/s] 7%|▋ | 3261/48008 [11:02<2:54:03, 4.28it/s] 7%|▋ | 3262/48008 [11:02<2:45:27, 4.51it/s] 7%|▋ | 3263/48008 [11:02<2:42:36, 4.59it/s] 7%|▋ | 3264/48008 [11:02<2:39:05, 4.69it/s] 7%|▋ | 3265/48008 [11:02<2:32:37, 4.89it/s] 7%|▋ | 3266/48008 [11:03<2:29:55, 4.97it/s] 7%|▋ | 3267/48008 [11:03<2:27:25, 5.06it/s] 7%|▋ | 3268/48008 [11:03<2:25:42, 5.12it/s] 7%|▋ | 3269/48008 [11:03<2:27:22, 5.06it/s] 7%|▋ | 3270/48008 [11:03<2:26:04, 5.10it/s] 7%|▋ | 3271/48008 [11:04<2:25:10, 5.14it/s] 7%|▋ | 3272/48008 [11:04<2:49:50, 4.39it/s] 7%|▋ | 3273/48008 [11:04<2:45:56, 4.49it/s] 7%|▋ | 3274/48008 [11:04<2:39:21, 4.68it/s] 7%|▋ | 3275/48008 [11:04<2:33:56, 4.84it/s] 7%|▋ | 3276/48008 [11:05<2:30:04, 4.97it/s] 7%|▋ | 3277/48008 [11:05<2:30:10, 4.96it/s] 7%|▋ | 3278/48008 [11:05<2:29:19, 4.99it/s] 7%|▋ | 3279/48008 [11:05<2:27:26, 5.06it/s] 7%|▋ | 3280/48008 [11:05<2:26:05, 5.10it/s] 7%|▋ | 3281/48008 [11:06<2:24:10, 5.17it/s] 7%|▋ | 3282/48008 [11:06<2:24:51, 5.15it/s] 7%|▋ | 3283/48008 [11:06<2:23:25, 5.20it/s] 7%|▋ | 3284/48008 [11:06<2:22:45, 5.22it/s] 7%|▋ | 3285/48008 [11:06<2:24:55, 5.14it/s] 7%|▋ | 3286/48008 [11:07<2:25:49, 5.11it/s] 7%|▋ | 3287/48008 [11:07<2:34:54, 4.81it/s] 7%|▋ | 3288/48008 [11:07<2:29:56, 4.97it/s] 7%|▋ | 3289/48008 [11:07<2:27:19, 5.06it/s] 7%|▋ | 3290/48008 [11:07<2:25:52, 5.11it/s] 7%|▋ | 3291/48008 [11:08<2:27:25, 5.06it/s] 7%|▋ | 3292/48008 [11:08<2:25:35, 5.12it/s] 7%|▋ | 3293/48008 [11:08<3:08:20, 3.96it/s] 7%|▋ | 3294/48008 [11:08<2:55:38, 4.24it/s] 7%|▋ | 3295/48008 [11:09<2:45:27, 4.50it/s] 7%|▋ | 3296/48008 [11:09<2:38:08, 4.71it/s] 7%|▋ | 3297/48008 [11:09<2:36:37, 4.76it/s] 7%|▋ | 3298/48008 [11:09<2:32:12, 4.90it/s] 7%|▋ | 3299/48008 [11:09<2:30:48, 4.94it/s] 7%|▋ | 3300/48008 [11:09<2:27:39, 5.05it/s] {'loss': 0.3033, 'grad_norm': 0.23046875, 'learning_rate': 0.0001862564572571238, 'epoch': 0.07} 7%|▋ | 3300/48008 [11:09<2:27:39, 5.05it/s] 7%|▋ | 3301/48008 [11:10<2:52:07, 4.33it/s] 7%|▋ | 3302/48008 [11:10<2:43:10, 4.57it/s] 7%|▋ | 3303/48008 [11:10<3:20:35, 3.71it/s] 7%|▋ | 3304/48008 [11:11<3:03:12, 4.07it/s] 7%|▋ | 3305/48008 [11:11<2:54:51, 4.26it/s] 7%|▋ | 3306/48008 [11:11<2:44:14, 4.54it/s] 7%|▋ | 3307/48008 [11:11<2:37:04, 4.74it/s] 7%|▋ | 3308/48008 [11:11<2:32:35, 4.88it/s] 7%|▋ | 3309/48008 [11:12<2:39:52, 4.66it/s] 7%|▋ | 3310/48008 [11:12<2:44:44, 4.52it/s] 7%|▋ | 3311/48008 [11:12<2:48:20, 4.43it/s] 7%|▋ | 3312/48008 [11:12<2:39:38, 4.67it/s] 7%|▋ | 3313/48008 [11:12<2:36:50, 4.75it/s] 7%|▋ | 3314/48008 [11:13<2:31:38, 4.91it/s] 7%|▋ | 3315/48008 [11:13<2:28:44, 5.01it/s] 7%|▋ | 3316/48008 [11:13<2:26:18, 5.09it/s] 7%|▋ | 3317/48008 [11:13<2:25:22, 5.12it/s] 7%|▋ | 3318/48008 [11:13<2:24:34, 5.15it/s] 7%|▋ | 3319/48008 [11:14<2:23:15, 5.20it/s] 7%|▋ | 3320/48008 [11:14<2:22:40, 5.22it/s] 7%|▋ | 3321/48008 [11:14<2:23:36, 5.19it/s] 7%|▋ | 3322/48008 [11:14<2:22:23, 5.23it/s] 7%|▋ | 3323/48008 [11:14<2:22:12, 5.24it/s] 7%|▋ | 3324/48008 [11:15<2:21:15, 5.27it/s] 7%|▋ | 3325/48008 [11:15<2:21:44, 5.25it/s] 7%|▋ | 3326/48008 [11:15<2:26:06, 5.10it/s] 7%|▋ | 3327/48008 [11:15<2:24:54, 5.14it/s] 7%|▋ | 3328/48008 [11:15<2:24:11, 5.16it/s] 7%|▋ | 3329/48008 [11:15<2:23:36, 5.19it/s] 7%|▋ | 3330/48008 [11:16<2:22:47, 5.21it/s] 7%|▋ | 3331/48008 [11:16<2:24:14, 5.16it/s] 7%|▋ | 3332/48008 [11:16<2:23:34, 5.19it/s] 7%|▋ | 3333/48008 [11:16<2:22:57, 5.21it/s] 7%|▋ | 3334/48008 [11:16<2:25:14, 5.13it/s] 7%|▋ | 3335/48008 [11:17<2:23:53, 5.17it/s] 7%|▋ | 3336/48008 [11:17<2:23:43, 5.18it/s] 7%|▋ | 3337/48008 [11:17<2:23:04, 5.20it/s] 7%|▋ | 3338/48008 [11:17<2:20:39, 5.29it/s] 7%|▋ | 3339/48008 [11:17<2:23:06, 5.20it/s] 7%|▋ | 3340/48008 [11:18<2:22:18, 5.23it/s] 7%|▋ | 3341/48008 [11:18<2:23:43, 5.18it/s] 7%|▋ | 3342/48008 [11:18<2:24:40, 5.15it/s] 7%|▋ | 3343/48008 [11:18<2:25:03, 5.13it/s] 7%|▋ | 3344/48008 [11:18<2:24:52, 5.14it/s] 7%|▋ | 3345/48008 [11:19<2:25:48, 5.10it/s] 7%|▋ | 3346/48008 [11:19<2:26:04, 5.10it/s] 7%|▋ | 3347/48008 [11:19<2:24:49, 5.14it/s] 7%|▋ | 3348/48008 [11:19<2:26:07, 5.09it/s] 7%|▋ | 3349/48008 [11:19<2:25:16, 5.12it/s] 7%|▋ | 3350/48008 [11:20<2:24:09, 5.16it/s] {'loss': 0.3058, 'grad_norm': 0.375, 'learning_rate': 0.00018604815864022665, 'epoch': 0.07} 7%|▋ | 3350/48008 [11:20<2:24:09, 5.16it/s] 7%|▋ | 3351/48008 [11:20<2:23:43, 5.18it/s] 7%|▋ | 3352/48008 [11:20<2:24:45, 5.14it/s] 7%|▋ | 3353/48008 [11:20<2:23:40, 5.18it/s] 7%|▋ | 3354/48008 [11:20<2:23:17, 5.19it/s] 7%|▋ | 3355/48008 [11:21<2:22:34, 5.22it/s] 7%|▋ | 3356/48008 [11:21<2:22:14, 5.23it/s] 7%|▋ | 3357/48008 [11:21<2:20:55, 5.28it/s] 7%|▋ | 3358/48008 [11:21<2:23:30, 5.19it/s] 7%|▋ | 3359/48008 [11:21<2:21:47, 5.25it/s] 7%|▋ | 3360/48008 [11:21<2:24:25, 5.15it/s] 7%|▋ | 3361/48008 [11:22<2:23:08, 5.20it/s] 7%|▋ | 3362/48008 [11:22<2:21:42, 5.25it/s] 7%|▋ | 3363/48008 [11:22<2:19:26, 5.34it/s] 7%|▋ | 3364/48008 [11:22<2:18:47, 5.36it/s] 7%|▋ | 3365/48008 [11:22<2:19:48, 5.32it/s] 7%|▋ | 3366/48008 [11:23<2:20:44, 5.29it/s] 7%|▋ | 3367/48008 [11:23<2:25:24, 5.12it/s] 7%|▋ | 3368/48008 [11:23<2:24:00, 5.17it/s] 7%|▋ | 3369/48008 [11:23<2:25:05, 5.13it/s] 7%|▋ | 3370/48008 [11:23<2:24:19, 5.15it/s] 7%|▋ | 3371/48008 [11:24<2:23:47, 5.17it/s] 7%|▋ | 3372/48008 [11:24<2:22:07, 5.23it/s] 7%|▋ | 3373/48008 [11:24<2:22:02, 5.24it/s] 7%|▋ | 3374/48008 [11:24<2:20:56, 5.28it/s] 7%|▋ | 3375/48008 [11:24<2:31:39, 4.91it/s] 7%|▋ | 3376/48008 [11:25<2:28:43, 5.00it/s] 7%|▋ | 3377/48008 [11:25<2:26:46, 5.07it/s] 7%|▋ | 3378/48008 [11:25<2:25:18, 5.12it/s] 7%|▋ | 3379/48008 [11:25<2:24:02, 5.16it/s] 7%|▋ | 3380/48008 [11:25<2:22:49, 5.21it/s] 7%|▋ | 3381/48008 [11:26<2:25:06, 5.13it/s] 7%|▋ | 3382/48008 [11:26<2:49:59, 4.38it/s] 7%|▋ | 3383/48008 [11:26<2:41:00, 4.62it/s] 7%|▋ | 3384/48008 [11:26<2:35:36, 4.78it/s] 7%|▋ | 3385/48008 [11:26<2:33:51, 4.83it/s] 7%|▋ | 3386/48008 [11:27<2:31:44, 4.90it/s] 7%|▋ | 3387/48008 [11:27<2:26:19, 5.08it/s] 7%|▋ | 3388/48008 [11:27<2:25:05, 5.13it/s] 7%|▋ | 3389/48008 [11:27<2:34:15, 4.82it/s] 7%|▋ | 3390/48008 [11:27<2:31:45, 4.90it/s] 7%|▋ | 3391/48008 [11:28<2:28:23, 5.01it/s] 7%|▋ | 3392/48008 [11:28<2:27:47, 5.03it/s] 7%|▋ | 3393/48008 [11:28<2:29:59, 4.96it/s] 7%|▋ | 3394/48008 [11:28<2:30:02, 4.96it/s] 7%|▋ | 3395/48008 [11:28<2:27:57, 5.03it/s] 7%|▋ | 3396/48008 [11:29<5:20:58, 2.32it/s] 7%|▋ | 3397/48008 [11:30<4:27:04, 2.78it/s] 7%|▋ | 3398/48008 [11:30<3:49:48, 3.24it/s] 7%|▋ | 3399/48008 [11:30<3:22:47, 3.67it/s] 7%|▋ | 3400/48008 [11:30<3:04:36, 4.03it/s] {'loss': 0.3059, 'grad_norm': 1.0078125, 'learning_rate': 0.00018583986002332944, 'epoch': 0.07} 7%|▋ | 3400/48008 [11:30<3:04:36, 4.03it/s] 7%|▋ | 3401/48008 [11:30<2:52:32, 4.31it/s] 7%|▋ | 3402/48008 [11:32<6:20:01, 1.96it/s] 7%|▋ | 3403/48008 [11:32<5:07:38, 2.42it/s] 7%|▋ | 3404/48008 [11:32<4:19:34, 2.86it/s] 7%|▋ | 3405/48008 [11:32<3:44:05, 3.32it/s] 7%|▋ | 3406/48008 [11:32<3:19:45, 3.72it/s] 7%|▋ | 3407/48008 [11:32<3:02:25, 4.07it/s] 7%|▋ | 3408/48008 [11:33<3:33:23, 3.48it/s] 7%|▋ | 3409/48008 [11:33<3:12:05, 3.87it/s] 7%|▋ | 3410/48008 [11:33<2:58:14, 4.17it/s] 7%|▋ | 3411/48008 [11:33<2:48:24, 4.41it/s] 7%|▋ | 3412/48008 [11:34<2:40:14, 4.64it/s] 7%|▋ | 3413/48008 [11:34<2:35:10, 4.79it/s] 7%|▋ | 3414/48008 [11:34<2:30:48, 4.93it/s] 7%|▋ | 3415/48008 [11:34<2:28:22, 5.01it/s] 7%|▋ | 3416/48008 [11:34<2:25:47, 5.10it/s] 7%|▋ | 3417/48008 [11:35<2:25:34, 5.11it/s] 7%|▋ | 3418/48008 [11:35<2:24:39, 5.14it/s] 7%|▋ | 3419/48008 [11:35<2:23:22, 5.18it/s] 7%|▋ | 3420/48008 [11:35<2:23:15, 5.19it/s] 7%|▋ | 3421/48008 [11:35<2:21:03, 5.27it/s] 7%|▋ | 3422/48008 [11:36<2:46:41, 4.46it/s] 7%|▋ | 3423/48008 [11:36<2:38:07, 4.70it/s] 7%|▋ | 3424/48008 [11:36<2:36:06, 4.76it/s] 7%|▋ | 3425/48008 [11:36<2:32:34, 4.87it/s] 7%|▋ | 3426/48008 [11:36<2:33:50, 4.83it/s] 7%|▋ | 3427/48008 [11:37<2:33:12, 4.85it/s] 7%|▋ | 3428/48008 [11:37<2:31:21, 4.91it/s] 7%|▋ | 3429/48008 [11:37<2:27:44, 5.03it/s] 7%|▋ | 3430/48008 [11:37<2:28:11, 5.01it/s] 7%|▋ | 3431/48008 [11:37<2:27:47, 5.03it/s] 7%|▋ | 3432/48008 [11:38<2:27:28, 5.04it/s] 7%|▋ | 3433/48008 [11:38<2:25:39, 5.10it/s] 7%|▋ | 3434/48008 [11:38<2:26:23, 5.07it/s] 7%|▋ | 3435/48008 [11:38<2:25:16, 5.11it/s] 7%|▋ | 3436/48008 [11:38<2:28:40, 5.00it/s] 7%|▋ | 3437/48008 [11:39<2:26:11, 5.08it/s] 7%|▋ | 3438/48008 [11:39<2:50:28, 4.36it/s] 7%|▋ | 3439/48008 [11:39<2:45:44, 4.48it/s] 7%|▋ | 3440/48008 [11:39<2:38:17, 4.69it/s] 7%|▋ | 3441/48008 [11:40<2:34:53, 4.80it/s] 7%|▋ | 3442/48008 [11:40<2:31:18, 4.91it/s] 7%|▋ | 3443/48008 [11:40<2:29:43, 4.96it/s] 7%|▋ | 3444/48008 [11:40<2:27:13, 5.04it/s] 7%|▋ | 3445/48008 [11:40<2:26:05, 5.08it/s] 7%|▋ | 3446/48008 [11:41<2:35:08, 4.79it/s] 7%|▋ | 3447/48008 [11:41<2:30:53, 4.92it/s] 7%|▋ | 3448/48008 [11:41<2:27:49, 5.02it/s] 7%|▋ | 3449/48008 [11:41<2:51:38, 4.33it/s] 7%|▋ | 3450/48008 [11:41<2:44:36, 4.51it/s] {'loss': 0.3028, 'grad_norm': 0.2314453125, 'learning_rate': 0.0001856315614064323, 'epoch': 0.07} 7%|▋ | 3450/48008 [11:41<2:44:36, 4.51it/s] 7%|▋ | 3451/48008 [11:42<2:37:31, 4.71it/s] 7%|▋ | 3452/48008 [11:42<2:32:27, 4.87it/s] 7%|▋ | 3453/48008 [11:42<2:33:33, 4.84it/s] 7%|▋ | 3454/48008 [11:42<2:31:16, 4.91it/s] 7%|▋ | 3455/48008 [11:42<2:31:14, 4.91it/s] 7%|▋ | 3456/48008 [11:43<2:30:03, 4.95it/s] 7%|▋ | 3457/48008 [11:43<2:53:17, 4.28it/s] 7%|▋ | 3458/48008 [11:43<2:43:09, 4.55it/s] 7%|▋ | 3459/48008 [11:43<2:37:38, 4.71it/s] 7%|▋ | 3460/48008 [11:43<2:34:08, 4.82it/s] 7%|▋ | 3461/48008 [11:44<2:30:32, 4.93it/s] 7%|▋ | 3462/48008 [11:44<2:27:51, 5.02it/s] 7%|▋ | 3463/48008 [11:44<2:25:35, 5.10it/s] 7%|▋ | 3464/48008 [11:44<2:24:12, 5.15it/s] 7%|▋ | 3465/48008 [11:44<2:26:09, 5.08it/s] 7%|▋ | 3466/48008 [11:45<2:25:36, 5.10it/s] 7%|▋ | 3467/48008 [11:45<2:25:41, 5.10it/s] 7%|▋ | 3468/48008 [11:45<2:24:39, 5.13it/s] 7%|▋ | 3469/48008 [11:45<2:22:54, 5.19it/s] 7%|▋ | 3470/48008 [11:45<2:22:37, 5.20it/s] 7%|▋ | 3471/48008 [11:46<2:22:55, 5.19it/s] 7%|▋ | 3472/48008 [11:46<2:24:44, 5.13it/s] 7%|▋ | 3473/48008 [11:46<2:23:06, 5.19it/s] 7%|▋ | 3474/48008 [11:46<2:25:16, 5.11it/s] 7%|▋ | 3475/48008 [11:46<2:23:38, 5.17it/s] 7%|▋ | 3476/48008 [11:47<2:22:40, 5.20it/s] 7%|▋ | 3477/48008 [11:47<2:22:55, 5.19it/s] 7%|▋ | 3478/48008 [11:47<2:22:47, 5.20it/s] 7%|▋ | 3479/48008 [11:47<2:21:16, 5.25it/s] 7%|▋ | 3480/48008 [11:47<2:20:17, 5.29it/s] 7%|▋ | 3481/48008 [11:48<2:20:42, 5.27it/s] 7%|▋ | 3482/48008 [11:48<2:23:29, 5.17it/s] 7%|▋ | 3483/48008 [11:48<2:22:19, 5.21it/s] 7%|▋ | 3484/48008 [11:48<2:32:33, 4.86it/s] 7%|▋ | 3485/48008 [11:48<2:29:35, 4.96it/s] 7%|▋ | 3486/48008 [11:49<2:26:40, 5.06it/s] 7%|▋ | 3487/48008 [11:49<2:25:28, 5.10it/s] 7%|▋ | 3488/48008 [11:49<2:24:25, 5.14it/s] 7%|▋ | 3489/48008 [11:49<2:23:31, 5.17it/s] 7%|▋ | 3490/48008 [11:49<2:24:03, 5.15it/s] 7%|▋ | 3491/48008 [11:49<2:23:12, 5.18it/s] 7%|▋ | 3492/48008 [11:50<2:25:10, 5.11it/s] 7%|▋ | 3493/48008 [11:50<2:23:42, 5.16it/s] 7%|▋ | 3494/48008 [11:50<2:22:45, 5.20it/s] 7%|▋ | 3495/48008 [11:50<2:24:47, 5.12it/s] 7%|▋ | 3496/48008 [11:50<2:23:19, 5.18it/s] 7%|▋ | 3497/48008 [11:51<2:24:26, 5.14it/s] 7%|▋ | 3498/48008 [11:51<2:26:00, 5.08it/s] 7%|▋ | 3499/48008 [11:51<2:25:38, 5.09it/s] 7%|▋ | 3500/48008 [11:51<2:34:46, 4.79it/s] {'loss': 0.3045, 'grad_norm': 0.380859375, 'learning_rate': 0.00018542326278953508, 'epoch': 0.07} 7%|▋ | 3500/48008 [11:51<2:34:46, 4.79it/s] 7%|▋ | 3501/48008 [11:51<2:33:22, 4.84it/s] 7%|▋ | 3502/48008 [11:52<2:55:21, 4.23it/s] 7%|▋ | 3503/48008 [11:52<2:43:14, 4.54it/s] 7%|▋ | 3504/48008 [11:52<2:36:02, 4.75it/s] 7%|▋ | 3505/48008 [11:52<2:31:45, 4.89it/s] 7%|▋ | 3506/48008 [11:53<2:28:39, 4.99it/s] 7%|▋ | 3507/48008 [11:53<2:26:32, 5.06it/s] 7%|▋ | 3508/48008 [11:53<2:27:43, 5.02it/s] 7%|▋ | 3509/48008 [11:53<2:28:06, 5.01it/s] 7%|▋ | 3510/48008 [11:53<2:36:11, 4.75it/s] 7%|▋ | 3511/48008 [11:54<2:30:31, 4.93it/s] 7%|▋ | 3512/48008 [11:54<2:27:06, 5.04it/s] 7%|▋ | 3513/48008 [11:54<2:24:59, 5.11it/s] 7%|▋ | 3514/48008 [11:54<2:24:10, 5.14it/s] 7%|▋ | 3515/48008 [11:54<2:27:32, 5.03it/s] 7%|▋ | 3516/48008 [11:55<2:25:52, 5.08it/s] 7%|▋ | 3517/48008 [11:55<2:23:52, 5.15it/s] 7%|▋ | 3518/48008 [11:55<2:23:16, 5.18it/s] 7%|▋ | 3519/48008 [11:55<2:22:50, 5.19it/s] 7%|▋ | 3520/48008 [11:55<2:24:23, 5.14it/s] 7%|▋ | 3521/48008 [11:56<2:26:04, 5.08it/s] 7%|▋ | 3522/48008 [11:56<2:24:06, 5.15it/s] 7%|▋ | 3523/48008 [11:56<2:48:41, 4.39it/s] 7%|▋ | 3524/48008 [11:56<2:40:37, 4.62it/s] 7%|▋ | 3525/48008 [11:56<2:32:24, 4.86it/s] 7%|▋ | 3526/48008 [11:57<2:39:42, 4.64it/s] 7%|▋ | 3527/48008 [11:57<2:34:00, 4.81it/s] 7%|▋ | 3528/48008 [11:57<2:30:32, 4.92it/s] 7%|▋ | 3529/48008 [11:57<2:27:33, 5.02it/s] 7%|▋ | 3530/48008 [11:57<2:24:59, 5.11it/s] 7%|▋ | 3531/48008 [11:58<2:23:56, 5.15it/s] 7%|▋ | 3532/48008 [11:58<2:22:49, 5.19it/s] 7%|▋ | 3533/48008 [11:58<2:22:16, 5.21it/s] 7%|▋ | 3534/48008 [11:58<2:24:22, 5.13it/s] 7%|▋ | 3535/48008 [11:58<2:27:21, 5.03it/s] 7%|▋ | 3536/48008 [11:59<2:25:04, 5.11it/s] 7%|▋ | 3537/48008 [11:59<2:24:07, 5.14it/s] 7%|▋ | 3538/48008 [11:59<2:24:33, 5.13it/s] 7%|▋ | 3539/48008 [11:59<2:23:41, 5.16it/s] 7%|▋ | 3540/48008 [11:59<2:33:21, 4.83it/s] 7%|▋ | 3541/48008 [12:00<2:28:31, 4.99it/s] 7%|▋ | 3542/48008 [12:00<2:28:52, 4.98it/s] 7%|▋ | 3543/48008 [12:00<2:26:49, 5.05it/s] 7%|▋ | 3544/48008 [12:00<2:25:20, 5.10it/s] 7%|▋ | 3545/48008 [12:00<2:23:11, 5.18it/s] 7%|▋ | 3546/48008 [12:00<2:22:29, 5.20it/s] 7%|▋ | 3547/48008 [12:01<2:24:36, 5.12it/s] 7%|▋ | 3548/48008 [12:01<2:23:15, 5.17it/s] 7%|▋ | 3549/48008 [12:01<2:22:46, 5.19it/s] 7%|▋ | 3550/48008 [12:01<3:05:26, 4.00it/s] {'loss': 0.3067, 'grad_norm': 0.482421875, 'learning_rate': 0.0001852149641726379, 'epoch': 0.07} 7%|▋ | 3550/48008 [12:01<3:05:26, 4.00it/s] 7%|▋ | 3551/48008 [12:02<2:52:27, 4.30it/s] 7%|▋ | 3552/48008 [12:02<2:43:00, 4.55it/s] 7%|▋ | 3553/48008 [12:02<2:46:32, 4.45it/s] 7%|▋ | 3554/48008 [12:02<2:38:38, 4.67it/s] 7%|▋ | 3555/48008 [12:02<2:33:31, 4.83it/s] 7%|▋ | 3556/48008 [12:03<2:28:51, 4.98it/s] 7%|▋ | 3557/48008 [12:03<2:25:43, 5.08it/s] 7%|▋ | 3558/48008 [12:03<2:23:05, 5.18it/s] 7%|▋ | 3559/48008 [12:03<2:21:34, 5.23it/s] 7%|▋ | 3560/48008 [12:03<2:24:10, 5.14it/s] 7%|▋ | 3561/48008 [12:04<2:22:22, 5.20it/s] 7%|▋ | 3562/48008 [12:04<2:21:07, 5.25it/s] 7%|▋ | 3563/48008 [12:04<2:21:15, 5.24it/s] 7%|▋ | 3564/48008 [12:04<2:20:33, 5.27it/s] 7%|▋ | 3565/48008 [12:04<2:19:12, 5.32it/s] 7%|▋ | 3566/48008 [12:05<2:18:48, 5.34it/s] 7%|▋ | 3567/48008 [12:05<2:22:01, 5.22it/s] 7%|▋ | 3568/48008 [12:05<2:21:49, 5.22it/s] 7%|▋ | 3569/48008 [12:05<2:24:15, 5.13it/s] 7%|▋ | 3570/48008 [12:05<2:24:21, 5.13it/s] 7%|▋ | 3571/48008 [12:06<2:22:53, 5.18it/s] 7%|▋ | 3572/48008 [12:06<2:22:14, 5.21it/s] 7%|▋ | 3573/48008 [12:06<2:21:55, 5.22it/s] 7%|▋ | 3574/48008 [12:06<2:21:47, 5.22it/s] 7%|▋ | 3575/48008 [12:06<2:24:06, 5.14it/s] 7%|▋ | 3576/48008 [12:06<2:21:52, 5.22it/s] 7%|▋ | 3577/48008 [12:07<2:20:35, 5.27it/s] 7%|▋ | 3578/48008 [12:07<2:21:35, 5.23it/s] 7%|▋ | 3579/48008 [12:07<2:19:01, 5.33it/s] 7%|▋ | 3580/48008 [12:07<2:21:13, 5.24it/s] 7%|▋ | 3581/48008 [12:07<2:24:56, 5.11it/s] 7%|▋ | 3582/48008 [12:08<2:24:59, 5.11it/s] 7%|▋ | 3583/48008 [12:08<2:22:32, 5.19it/s] 7%|▋ | 3584/48008 [12:08<2:22:24, 5.20it/s] 7%|▋ | 3585/48008 [12:08<2:24:13, 5.13it/s] 7%|▋ | 3586/48008 [12:08<2:24:54, 5.11it/s] 7%|▋ | 3587/48008 [12:09<2:49:20, 4.37it/s] 7%|▋ | 3588/48008 [12:09<2:44:01, 4.51it/s] 7%|▋ | 3589/48008 [12:09<2:36:44, 4.72it/s] 7%|▋ | 3590/48008 [12:09<2:31:39, 4.88it/s] 7%|▋ | 3591/48008 [12:09<2:27:37, 5.01it/s] 7%|▋ | 3592/48008 [12:10<2:24:50, 5.11it/s] 7%|▋ | 3593/48008 [12:10<2:22:55, 5.18it/s] 7%|▋ | 3594/48008 [12:10<2:22:03, 5.21it/s] 7%|▋ | 3595/48008 [12:10<2:32:11, 4.86it/s] 7%|▋ | 3596/48008 [12:11<2:39:12, 4.65it/s] 7%|▋ | 3597/48008 [12:11<2:33:18, 4.83it/s] 7%|▋ | 3598/48008 [12:11<2:39:50, 4.63it/s] 7%|▋ | 3599/48008 [12:11<2:37:59, 4.68it/s] 7%|▋ | 3600/48008 [12:11<2:32:35, 4.85it/s] {'loss': 0.3082, 'grad_norm': 0.4921875, 'learning_rate': 0.0001850066655557407, 'epoch': 0.07} 7%|▋ | 3600/48008 [12:11<2:32:35, 4.85it/s] 8%|▊ | 3601/48008 [12:12<2:29:20, 4.96it/s] 8%|▊ | 3602/48008 [12:12<2:26:54, 5.04it/s] 8%|▊ | 3603/48008 [12:12<2:25:10, 5.10it/s] 8%|▊ | 3604/48008 [12:12<2:23:09, 5.17it/s] 8%|▊ | 3605/48008 [12:12<2:22:18, 5.20it/s] 8%|▊ | 3606/48008 [12:12<2:21:16, 5.24it/s] 8%|▊ | 3607/48008 [12:13<2:21:20, 5.24it/s] 8%|▊ | 3608/48008 [12:13<2:21:03, 5.25it/s] 8%|▊ | 3609/48008 [12:13<2:23:30, 5.16it/s] 8%|▊ | 3610/48008 [12:13<2:27:20, 5.02it/s] 8%|▊ | 3611/48008 [12:13<2:25:27, 5.09it/s] 8%|▊ | 3612/48008 [12:14<2:23:07, 5.17it/s] 8%|▊ | 3613/48008 [12:14<2:22:05, 5.21it/s] 8%|▊ | 3614/48008 [12:14<2:21:31, 5.23it/s] 8%|▊ | 3615/48008 [12:14<2:20:57, 5.25it/s] 8%|▊ | 3616/48008 [12:14<2:20:56, 5.25it/s] 8%|▊ | 3617/48008 [12:15<3:04:18, 4.01it/s] 8%|▊ | 3618/48008 [12:15<3:01:31, 4.08it/s] 8%|▊ | 3619/48008 [12:15<2:51:42, 4.31it/s] 8%|▊ | 3620/48008 [12:15<2:43:29, 4.53it/s] 8%|▊ | 3621/48008 [12:16<2:39:54, 4.63it/s] 8%|▊ | 3622/48008 [12:16<2:34:03, 4.80it/s] 8%|▊ | 3623/48008 [12:16<2:34:10, 4.80it/s] 8%|▊ | 3624/48008 [12:16<2:30:16, 4.92it/s] 8%|▊ | 3625/48008 [12:16<2:27:31, 5.01it/s] 8%|▊ | 3626/48008 [12:17<2:25:47, 5.07it/s] 8%|▊ | 3627/48008 [12:17<2:23:30, 5.15it/s] 8%|▊ | 3628/48008 [12:17<2:21:44, 5.22it/s] 8%|▊ | 3629/48008 [12:17<2:21:32, 5.23it/s] 8%|▊ | 3630/48008 [12:17<2:21:05, 5.24it/s] 8%|▊ | 3631/48008 [12:18<2:22:18, 5.20it/s] 8%|▊ | 3632/48008 [12:18<3:05:01, 4.00it/s] 8%|▊ | 3633/48008 [12:18<2:51:28, 4.31it/s] 8%|▊ | 3634/48008 [12:18<2:45:08, 4.48it/s] 8%|▊ | 3635/48008 [12:19<2:36:38, 4.72it/s] 8%|▊ | 3636/48008 [12:19<2:34:42, 4.78it/s] 8%|▊ | 3637/48008 [12:19<2:30:10, 4.92it/s] 8%|▊ | 3638/48008 [12:19<2:27:13, 5.02it/s] 8%|▊ | 3639/48008 [12:19<2:25:22, 5.09it/s] 8%|▊ | 3640/48008 [12:19<2:25:22, 5.09it/s] 8%|▊ | 3641/48008 [12:20<2:26:51, 5.04it/s] 8%|▊ | 3642/48008 [12:20<2:26:32, 5.05it/s] 8%|▊ | 3643/48008 [12:20<2:28:52, 4.97it/s] 8%|▊ | 3644/48008 [12:20<2:26:13, 5.06it/s] 8%|▊ | 3645/48008 [12:20<2:24:55, 5.10it/s] 8%|▊ | 3646/48008 [12:21<2:23:34, 5.15it/s] 8%|▊ | 3647/48008 [12:21<2:22:26, 5.19it/s] 8%|▊ | 3648/48008 [12:21<2:21:55, 5.21it/s] 8%|▊ | 3649/48008 [12:21<2:25:53, 5.07it/s] 8%|▊ | 3650/48008 [12:22<3:07:38, 3.94it/s] {'loss': 0.3069, 'grad_norm': 0.2197265625, 'learning_rate': 0.00018479836693884353, 'epoch': 0.08} 8%|▊ | 3650/48008 [12:22<3:07:38, 3.94it/s] 8%|▊ | 3651/48008 [12:22<2:53:56, 4.25it/s] 8%|▊ | 3652/48008 [12:22<2:43:33, 4.52it/s] 8%|▊ | 3653/48008 [12:22<2:37:21, 4.70it/s] 8%|▊ | 3654/48008 [12:22<2:31:56, 4.87it/s] 8%|▊ | 3655/48008 [12:23<2:29:32, 4.94it/s] 8%|▊ | 3656/48008 [12:23<2:28:07, 4.99it/s] 8%|▊ | 3657/48008 [12:23<2:51:18, 4.32it/s] 8%|▊ | 3658/48008 [12:23<2:46:39, 4.44it/s] 8%|▊ | 3659/48008 [12:24<2:40:26, 4.61it/s] 8%|▊ | 3660/48008 [12:24<2:33:54, 4.80it/s] 8%|▊ | 3661/48008 [12:24<2:28:54, 4.96it/s] 8%|▊ | 3662/48008 [12:24<2:26:40, 5.04it/s] 8%|▊ | 3663/48008 [12:24<2:24:09, 5.13it/s] 8%|▊ | 3664/48008 [12:24<2:22:25, 5.19it/s] 8%|▊ | 3665/48008 [12:25<2:22:24, 5.19it/s] 8%|▊ | 3666/48008 [12:25<2:23:28, 5.15it/s] 8%|▊ | 3667/48008 [12:25<2:25:09, 5.09it/s] 8%|▊ | 3668/48008 [12:25<2:26:30, 5.04it/s] 8%|▊ | 3669/48008 [12:25<2:23:44, 5.14it/s] 8%|▊ | 3670/48008 [12:26<2:21:49, 5.21it/s] 8%|▊ | 3671/48008 [12:26<2:21:22, 5.23it/s] 8%|▊ | 3672/48008 [12:26<2:21:23, 5.23it/s] 8%|▊ | 3673/48008 [12:26<2:20:32, 5.26it/s] 8%|▊ | 3674/48008 [12:26<2:20:10, 5.27it/s] 8%|▊ | 3675/48008 [12:27<4:51:38, 2.53it/s] 8%|▊ | 3676/48008 [12:27<4:06:26, 3.00it/s] 8%|▊ | 3677/48008 [12:28<4:00:23, 3.07it/s] 8%|▊ | 3678/48008 [12:28<3:30:04, 3.52it/s] 8%|▊ | 3679/48008 [12:28<3:09:03, 3.91it/s] 8%|▊ | 3680/48008 [12:28<3:04:53, 4.00it/s] 8%|▊ | 3681/48008 [12:29<2:50:33, 4.33it/s] 8%|▊ | 3682/48008 [12:29<2:42:38, 4.54it/s] 8%|▊ | 3683/48008 [12:29<2:37:09, 4.70it/s] 8%|▊ | 3684/48008 [12:29<2:31:37, 4.87it/s] 8%|▊ | 3685/48008 [12:29<2:28:17, 4.98it/s] 8%|▊ | 3686/48008 [12:29<2:25:30, 5.08it/s] 8%|▊ | 3687/48008 [12:30<2:24:33, 5.11it/s] 8%|▊ | 3688/48008 [12:30<2:23:26, 5.15it/s] 8%|▊ | 3689/48008 [12:30<2:22:46, 5.17it/s] 8%|▊ | 3690/48008 [12:30<2:21:49, 5.21it/s] 8%|▊ | 3691/48008 [12:30<2:21:16, 5.23it/s] 8%|▊ | 3692/48008 [12:31<2:20:16, 5.27it/s] 8%|▊ | 3693/48008 [12:31<2:19:09, 5.31it/s] 8%|▊ | 3694/48008 [12:31<2:23:32, 5.15it/s] 8%|▊ | 3695/48008 [12:31<2:32:45, 4.83it/s] 8%|▊ | 3696/48008 [12:31<2:29:08, 4.95it/s] 8%|▊ | 3697/48008 [12:32<2:36:56, 4.71it/s] 8%|▊ | 3698/48008 [12:32<3:15:23, 3.78it/s] 8%|▊ | 3699/48008 [12:32<3:03:01, 4.03it/s] 8%|▊ | 3700/48008 [12:32<2:49:34, 4.35it/s] {'loss': 0.3055, 'grad_norm': 0.875, 'learning_rate': 0.00018459006832194635, 'epoch': 0.08} 8%|▊ | 3700/48008 [12:32<2:49:34, 4.35it/s] 8%|▊ | 3701/48008 [12:33<2:41:11, 4.58it/s] 8%|▊ | 3702/48008 [12:33<2:36:26, 4.72it/s] 8%|▊ | 3703/48008 [12:33<2:35:32, 4.75it/s] 8%|▊ | 3704/48008 [12:33<2:30:38, 4.90it/s] 8%|▊ | 3705/48008 [12:33<2:27:06, 5.02it/s] 8%|▊ | 3706/48008 [12:34<2:25:09, 5.09it/s] 8%|▊ | 3707/48008 [12:34<2:24:15, 5.12it/s] 8%|▊ | 3708/48008 [12:34<2:33:32, 4.81it/s] 8%|▊ | 3709/48008 [12:34<2:30:46, 4.90it/s] 8%|▊ | 3710/48008 [12:34<2:26:55, 5.03it/s] 8%|▊ | 3711/48008 [12:35<2:27:11, 5.02it/s] 8%|▊ | 3712/48008 [12:35<2:25:31, 5.07it/s] 8%|▊ | 3713/48008 [12:35<3:07:03, 3.95it/s] 8%|▊ | 3714/48008 [12:35<2:53:42, 4.25it/s] 8%|▊ | 3715/48008 [12:36<2:42:13, 4.55it/s] 8%|▊ | 3716/48008 [12:36<2:35:18, 4.75it/s] 8%|▊ | 3717/48008 [12:36<2:29:54, 4.92it/s] 8%|▊ | 3718/48008 [12:36<2:30:00, 4.92it/s] 8%|▊ | 3719/48008 [12:36<2:27:09, 5.02it/s] 8%|▊ | 3720/48008 [12:37<2:25:44, 5.06it/s] 8%|▊ | 3721/48008 [12:37<2:25:10, 5.08it/s] 8%|▊ | 3722/48008 [12:37<2:23:20, 5.15it/s] 8%|▊ | 3723/48008 [12:37<2:25:15, 5.08it/s] 8%|▊ | 3724/48008 [12:37<2:49:24, 4.36it/s] 8%|▊ | 3725/48008 [12:38<2:40:42, 4.59it/s] 8%|▊ | 3726/48008 [12:38<2:34:16, 4.78it/s] 8%|▊ | 3727/48008 [12:38<2:27:56, 4.99it/s] 8%|▊ | 3728/48008 [12:38<2:24:22, 5.11it/s] 8%|▊ | 3729/48008 [12:38<2:22:19, 5.19it/s] 8%|▊ | 3730/48008 [12:39<2:22:13, 5.19it/s] 8%|▊ | 3731/48008 [12:39<2:21:50, 5.20it/s] 8%|▊ | 3732/48008 [12:39<2:23:36, 5.14it/s] 8%|▊ | 3733/48008 [12:39<2:22:39, 5.17it/s] 8%|▊ | 3734/48008 [12:39<2:20:03, 5.27it/s] 8%|▊ | 3735/48008 [12:40<2:19:49, 5.28it/s] 8%|▊ | 3736/48008 [12:40<2:20:18, 5.26it/s] 8%|▊ | 3737/48008 [12:40<2:20:30, 5.25it/s] 8%|▊ | 3738/48008 [12:40<2:19:13, 5.30it/s] 8%|▊ | 3739/48008 [12:40<2:19:09, 5.30it/s] 8%|▊ | 3740/48008 [12:40<2:19:02, 5.31it/s] 8%|▊ | 3741/48008 [12:41<2:21:13, 5.22it/s] 8%|▊ | 3742/48008 [12:41<2:20:58, 5.23it/s] 8%|▊ | 3743/48008 [12:41<2:22:49, 5.17it/s] 8%|▊ | 3744/48008 [12:41<2:26:07, 5.05it/s] 8%|▊ | 3745/48008 [12:41<2:24:15, 5.11it/s] 8%|▊ | 3746/48008 [12:42<2:22:21, 5.18it/s] 8%|▊ | 3747/48008 [12:42<2:21:55, 5.20it/s] 8%|▊ | 3748/48008 [12:42<2:23:00, 5.16it/s] 8%|▊ | 3749/48008 [12:42<2:22:39, 5.17it/s] 8%|▊ | 3750/48008 [12:42<2:24:14, 5.11it/s] {'loss': 0.3057, 'grad_norm': 0.275390625, 'learning_rate': 0.00018438176970504916, 'epoch': 0.08} 8%|▊ | 3750/48008 [12:42<2:24:14, 5.11it/s] 8%|▊ | 3751/48008 [12:43<2:23:23, 5.14it/s] 8%|▊ | 3752/48008 [12:43<2:22:20, 5.18it/s] 8%|▊ | 3753/48008 [12:43<2:20:41, 5.24it/s] 8%|▊ | 3754/48008 [12:43<2:20:37, 5.25it/s] 8%|▊ | 3755/48008 [12:43<2:19:28, 5.29it/s] 8%|▊ | 3756/48008 [12:44<2:20:41, 5.24it/s] 8%|▊ | 3757/48008 [12:44<2:20:45, 5.24it/s] 8%|▊ | 3758/48008 [12:44<2:18:35, 5.32it/s] 8%|▊ | 3759/48008 [12:44<2:21:46, 5.20it/s] 8%|▊ | 3760/48008 [12:44<2:20:23, 5.25it/s] 8%|▊ | 3761/48008 [12:45<2:21:41, 5.20it/s] 8%|▊ | 3762/48008 [12:45<2:20:38, 5.24it/s] 8%|▊ | 3763/48008 [12:45<2:19:08, 5.30it/s] 8%|▊ | 3764/48008 [12:45<2:23:50, 5.13it/s] 8%|▊ | 3765/48008 [12:45<2:23:02, 5.16it/s] 8%|▊ | 3766/48008 [12:45<2:21:33, 5.21it/s] 8%|▊ | 3767/48008 [12:46<2:22:00, 5.19it/s] 8%|▊ | 3768/48008 [12:46<2:21:51, 5.20it/s] 8%|▊ | 3769/48008 [12:46<2:20:43, 5.24it/s] 8%|▊ | 3770/48008 [12:46<2:20:58, 5.23it/s] 8%|▊ | 3771/48008 [12:46<2:22:31, 5.17it/s] 8%|▊ | 3772/48008 [12:47<2:32:11, 4.84it/s] 8%|▊ | 3773/48008 [12:47<2:31:09, 4.88it/s] 8%|▊ | 3774/48008 [12:47<2:28:00, 4.98it/s] 8%|▊ | 3775/48008 [12:47<2:25:46, 5.06it/s] 8%|▊ | 3776/48008 [12:48<2:49:52, 4.34it/s] 8%|▊ | 3777/48008 [12:48<2:41:55, 4.55it/s] 8%|▊ | 3778/48008 [12:48<2:34:20, 4.78it/s] 8%|▊ | 3779/48008 [12:48<2:30:34, 4.90it/s] 8%|▊ | 3780/48008 [12:49<3:10:58, 3.86it/s] 8%|▊ | 3781/48008 [12:49<2:57:50, 4.14it/s] 8%|▊ | 3782/48008 [12:49<2:49:19, 4.35it/s] 8%|▊ | 3783/48008 [12:49<2:40:51, 4.58it/s] 8%|▊ | 3784/48008 [12:49<2:44:53, 4.47it/s] 8%|▊ | 3785/48008 [12:50<2:36:43, 4.70it/s] 8%|▊ | 3786/48008 [12:50<2:31:35, 4.86it/s] 8%|▊ | 3787/48008 [12:50<2:28:34, 4.96it/s] 8%|▊ | 3788/48008 [12:50<2:28:36, 4.96it/s] 8%|▊ | 3789/48008 [12:50<2:28:36, 4.96it/s] 8%|▊ | 3790/48008 [12:51<2:27:33, 4.99it/s] 8%|▊ | 3791/48008 [12:51<2:26:14, 5.04it/s] 8%|▊ | 3792/48008 [12:51<2:24:03, 5.12it/s] 8%|▊ | 3793/48008 [12:51<2:22:56, 5.16it/s] 8%|▊ | 3794/48008 [12:51<2:22:05, 5.19it/s] 8%|▊ | 3795/48008 [12:51<2:21:54, 5.19it/s] 8%|▊ | 3796/48008 [12:52<2:23:59, 5.12it/s] 8%|▊ | 3797/48008 [12:52<2:23:27, 5.14it/s] 8%|▊ | 3798/48008 [12:52<2:22:34, 5.17it/s] 8%|▊ | 3799/48008 [12:52<2:21:33, 5.21it/s] 8%|▊ | 3800/48008 [12:52<2:22:42, 5.16it/s] {'loss': 0.3043, 'grad_norm': 0.1435546875, 'learning_rate': 0.00018417347108815198, 'epoch': 0.08} 8%|▊ | 3800/48008 [12:52<2:22:42, 5.16it/s] 8%|▊ | 3801/48008 [12:53<2:22:10, 5.18it/s] 8%|▊ | 3802/48008 [12:53<2:21:19, 5.21it/s] 8%|▊ | 3803/48008 [12:53<2:23:21, 5.14it/s] 8%|▊ | 3804/48008 [12:53<2:20:59, 5.23it/s] 8%|▊ | 3805/48008 [12:53<2:20:32, 5.24it/s] 8%|▊ | 3806/48008 [12:54<2:20:59, 5.23it/s] 8%|▊ | 3807/48008 [12:54<2:23:12, 5.14it/s] 8%|▊ | 3808/48008 [12:54<2:26:24, 5.03it/s] 8%|▊ | 3809/48008 [12:54<2:27:05, 5.01it/s] 8%|▊ | 3810/48008 [12:54<2:25:01, 5.08it/s] 8%|▊ | 3811/48008 [12:55<2:24:17, 5.11it/s] 8%|▊ | 3812/48008 [12:55<2:25:53, 5.05it/s] 8%|▊ | 3813/48008 [12:55<3:07:30, 3.93it/s] 8%|▊ | 3814/48008 [12:55<2:55:48, 4.19it/s] 8%|▊ | 3815/48008 [12:56<2:43:56, 4.49it/s] 8%|▊ | 3816/48008 [12:56<2:37:17, 4.68it/s] 8%|▊ | 3817/48008 [12:56<2:33:06, 4.81it/s] 8%|▊ | 3818/48008 [12:56<2:33:29, 4.80it/s] 8%|▊ | 3819/48008 [12:56<2:29:09, 4.94it/s] 8%|▊ | 3820/48008 [12:57<2:25:06, 5.08it/s] 8%|▊ | 3821/48008 [12:57<2:23:29, 5.13it/s] 8%|▊ | 3822/48008 [12:57<2:22:18, 5.17it/s] 8%|▊ | 3823/48008 [12:57<2:21:59, 5.19it/s] 8%|▊ | 3824/48008 [12:57<2:21:25, 5.21it/s] 8%|▊ | 3825/48008 [12:57<2:20:52, 5.23it/s] 8%|▊ | 3826/48008 [12:58<2:23:49, 5.12it/s] 8%|▊ | 3827/48008 [12:58<2:25:16, 5.07it/s] 8%|▊ | 3828/48008 [12:58<2:24:17, 5.10it/s] 8%|▊ | 3829/48008 [12:58<2:24:06, 5.11it/s] 8%|▊ | 3830/48008 [12:59<3:05:51, 3.96it/s] 8%|▊ | 3831/48008 [12:59<2:52:47, 4.26it/s] 8%|▊ | 3832/48008 [12:59<2:45:30, 4.45it/s] 8%|▊ | 3833/48008 [12:59<2:37:17, 4.68it/s] 8%|▊ | 3834/48008 [12:59<2:32:10, 4.84it/s] 8%|▊ | 3835/48008 [13:00<2:28:22, 4.96it/s] 8%|▊ | 3836/48008 [13:00<2:25:25, 5.06it/s] 8%|▊ | 3837/48008 [13:00<2:49:27, 4.34it/s] 8%|▊ | 3838/48008 [13:00<2:40:44, 4.58it/s] 8%|▊ | 3839/48008 [13:01<2:34:51, 4.75it/s] 8%|▊ | 3840/48008 [13:01<2:32:44, 4.82it/s] 8%|▊ | 3841/48008 [13:01<2:29:33, 4.92it/s] 8%|▊ | 3842/48008 [13:01<2:27:42, 4.98it/s] 8%|▊ | 3843/48008 [13:01<2:28:05, 4.97it/s] 8%|▊ | 3844/48008 [13:01<2:26:15, 5.03it/s] 8%|▊ | 3845/48008 [13:02<2:27:02, 5.01it/s] 8%|▊ | 3846/48008 [13:02<2:27:56, 4.98it/s] 8%|▊ | 3847/48008 [13:02<2:24:47, 5.08it/s] 8%|▊ | 3848/48008 [13:02<2:25:41, 5.05it/s] 8%|▊ | 3849/48008 [13:02<2:24:05, 5.11it/s] 8%|▊ | 3850/48008 [13:03<2:24:38, 5.09it/s] {'loss': 0.3042, 'grad_norm': 0.625, 'learning_rate': 0.0001839651724712548, 'epoch': 0.08} 8%|▊ | 3850/48008 [13:03<2:24:38, 5.09it/s] 8%|▊ | 3851/48008 [13:03<2:23:27, 5.13it/s] 8%|▊ | 3852/48008 [13:03<2:22:28, 5.17it/s] 8%|▊ | 3853/48008 [13:03<2:22:14, 5.17it/s] 8%|▊ | 3854/48008 [13:03<2:20:21, 5.24it/s] 8%|▊ | 3855/48008 [13:04<2:23:07, 5.14it/s] 8%|▊ | 3856/48008 [13:04<2:22:12, 5.17it/s] 8%|▊ | 3857/48008 [13:04<2:20:56, 5.22it/s] 8%|▊ | 3858/48008 [13:04<2:21:19, 5.21it/s] 8%|▊ | 3859/48008 [13:04<2:23:41, 5.12it/s] 8%|▊ | 3860/48008 [13:05<2:32:42, 4.82it/s] 8%|▊ | 3861/48008 [13:05<2:31:30, 4.86it/s] 8%|▊ | 3862/48008 [13:05<2:29:46, 4.91it/s] 8%|▊ | 3863/48008 [13:05<2:29:36, 4.92it/s] 8%|▊ | 3864/48008 [13:05<2:26:29, 5.02it/s] 8%|▊ | 3865/48008 [13:06<2:25:11, 5.07it/s] 8%|▊ | 3866/48008 [13:06<2:23:00, 5.14it/s] 8%|▊ | 3867/48008 [13:06<2:22:18, 5.17it/s] 8%|▊ | 3868/48008 [13:06<2:22:51, 5.15it/s] 8%|▊ | 3869/48008 [13:06<2:24:49, 5.08it/s] 8%|▊ | 3870/48008 [13:07<2:23:03, 5.14it/s] 8%|▊ | 3871/48008 [13:07<2:22:31, 5.16it/s] 8%|▊ | 3872/48008 [13:07<2:22:15, 5.17it/s] 8%|▊ | 3873/48008 [13:07<2:46:58, 4.41it/s] 8%|▊ | 3874/48008 [13:07<2:39:01, 4.63it/s] 8%|▊ | 3875/48008 [13:08<2:32:58, 4.81it/s] 8%|▊ | 3876/48008 [13:08<2:27:45, 4.98it/s] 8%|▊ | 3877/48008 [13:08<2:25:30, 5.05it/s] 8%|▊ | 3878/48008 [13:08<2:23:53, 5.11it/s] 8%|▊ | 3879/48008 [13:08<2:25:47, 5.04it/s] 8%|▊ | 3880/48008 [13:09<2:24:21, 5.09it/s] 8%|▊ | 3881/48008 [13:09<2:21:58, 5.18it/s] 8%|▊ | 3882/48008 [13:09<2:21:28, 5.20it/s] 8%|▊ | 3883/48008 [13:09<2:22:38, 5.16it/s] 8%|▊ | 3884/48008 [13:09<2:21:05, 5.21it/s] 8%|▊ | 3885/48008 [13:10<2:20:38, 5.23it/s] 8%|▊ | 3886/48008 [13:10<2:20:10, 5.25it/s] 8%|▊ | 3887/48008 [13:10<2:19:24, 5.27it/s] 8%|▊ | 3888/48008 [13:10<2:30:07, 4.90it/s] 8%|▊ | 3889/48008 [13:10<2:26:14, 5.03it/s] 8%|▊ | 3890/48008 [13:11<2:27:15, 4.99it/s] 8%|▊ | 3891/48008 [13:11<2:28:08, 4.96it/s] 8%|▊ | 3892/48008 [13:11<2:27:17, 4.99it/s] 8%|▊ | 3893/48008 [13:11<2:25:40, 5.05it/s] 8%|▊ | 3894/48008 [13:11<2:26:19, 5.02it/s] 8%|▊ | 3895/48008 [13:12<2:24:07, 5.10it/s] 8%|▊ | 3896/48008 [13:12<2:23:13, 5.13it/s] 8%|▊ | 3897/48008 [13:12<2:21:54, 5.18it/s] 8%|▊ | 3898/48008 [13:12<2:21:26, 5.20it/s] 8%|▊ | 3899/48008 [13:12<2:23:02, 5.14it/s] 8%|▊ | 3900/48008 [13:13<2:22:09, 5.17it/s] {'loss': 0.3053, 'grad_norm': 0.326171875, 'learning_rate': 0.00018375687385435762, 'epoch': 0.08} 8%|▊ | 3900/48008 [13:13<2:22:09, 5.17it/s] 8%|▊ | 3901/48008 [13:13<2:21:35, 5.19it/s] 8%|▊ | 3902/48008 [13:13<2:21:47, 5.18it/s] 8%|▊ | 3903/48008 [13:13<2:23:20, 5.13it/s] 8%|▊ | 3904/48008 [13:13<2:22:32, 5.16it/s] 8%|▊ | 3905/48008 [13:14<2:21:59, 5.18it/s] 8%|▊ | 3906/48008 [13:14<2:21:18, 5.20it/s] 8%|▊ | 3907/48008 [13:14<2:20:18, 5.24it/s] 8%|▊ | 3908/48008 [13:14<2:19:22, 5.27it/s] 8%|▊ | 3909/48008 [13:14<2:19:02, 5.29it/s] 8%|▊ | 3910/48008 [13:14<2:19:26, 5.27it/s] 8%|▊ | 3911/48008 [13:15<2:19:00, 5.29it/s] 8%|▊ | 3912/48008 [13:15<2:29:55, 4.90it/s] 8%|▊ | 3913/48008 [13:15<2:27:16, 4.99it/s] 8%|▊ | 3914/48008 [13:15<2:26:37, 5.01it/s] 8%|▊ | 3915/48008 [13:15<2:24:24, 5.09it/s] 8%|▊ | 3916/48008 [13:16<2:26:08, 5.03it/s] 8%|▊ | 3917/48008 [13:16<2:23:55, 5.11it/s] 8%|▊ | 3918/48008 [13:16<2:24:02, 5.10it/s] 8%|▊ | 3919/48008 [13:16<2:21:42, 5.19it/s] 8%|▊ | 3920/48008 [13:16<2:21:04, 5.21it/s] 8%|▊ | 3921/48008 [13:17<2:23:07, 5.13it/s] 8%|▊ | 3922/48008 [13:17<2:24:25, 5.09it/s] 8%|▊ | 3923/48008 [13:17<2:24:34, 5.08it/s] 8%|▊ | 3924/48008 [13:17<2:25:09, 5.06it/s] 8%|▊ | 3925/48008 [13:17<2:25:55, 5.03it/s] 8%|▊ | 3926/48008 [13:18<2:24:54, 5.07it/s] 8%|▊ | 3927/48008 [13:18<2:23:30, 5.12it/s] 8%|▊ | 3928/48008 [13:18<2:22:25, 5.16it/s] 8%|▊ | 3929/48008 [13:18<2:20:47, 5.22it/s] 8%|▊ | 3930/48008 [13:18<2:20:27, 5.23it/s] 8%|▊ | 3931/48008 [13:19<2:45:55, 4.43it/s] 8%|▊ | 3932/48008 [13:19<2:38:56, 4.62it/s] 8%|▊ | 3933/48008 [13:19<2:33:27, 4.79it/s] 8%|▊ | 3934/48008 [13:19<2:32:07, 4.83it/s] 8%|▊ | 3935/48008 [13:19<2:28:42, 4.94it/s] 8%|▊ | 3936/48008 [13:20<3:09:19, 3.88it/s] 8%|▊ | 3937/48008 [13:20<2:58:46, 4.11it/s] 8%|▊ | 3938/48008 [13:20<2:49:09, 4.34it/s] 8%|▊ | 3939/48008 [13:20<2:43:08, 4.50it/s] 8%|▊ | 3940/48008 [13:21<2:35:19, 4.73it/s] 8%|▊ | 3941/48008 [13:21<2:32:06, 4.83it/s] 8%|▊ | 3942/48008 [13:21<2:28:06, 4.96it/s] 8%|▊ | 3943/48008 [13:21<2:25:42, 5.04it/s] 8%|▊ | 3944/48008 [13:21<2:23:47, 5.11it/s] 8%|▊ | 3945/48008 [13:22<2:22:44, 5.14it/s] 8%|▊ | 3946/48008 [13:22<2:21:54, 5.18it/s] 8%|▊ | 3947/48008 [13:22<2:47:10, 4.39it/s] 8%|▊ | 3948/48008 [13:22<2:40:54, 4.56it/s] 8%|▊ | 3949/48008 [13:22<2:36:00, 4.71it/s] 8%|▊ | 3950/48008 [13:23<2:30:09, 4.89it/s] {'loss': 0.3034, 'grad_norm': 0.2158203125, 'learning_rate': 0.00018354857523746043, 'epoch': 0.08} 8%|▊ | 3950/48008 [13:23<2:30:09, 4.89it/s] 8%|▊ | 3951/48008 [13:23<2:29:36, 4.91it/s] 8%|▊ | 3952/48008 [13:23<2:26:34, 5.01it/s] 8%|▊ | 3953/48008 [13:23<2:27:05, 4.99it/s] 8%|▊ | 3954/48008 [13:23<2:26:16, 5.02it/s] 8%|▊ | 3955/48008 [13:24<2:24:05, 5.10it/s] 8%|▊ | 3956/48008 [13:24<2:21:46, 5.18it/s] 8%|▊ | 3957/48008 [13:24<2:25:39, 5.04it/s] 8%|▊ | 3958/48008 [13:24<2:49:12, 4.34it/s] 8%|▊ | 3959/48008 [13:25<2:50:20, 4.31it/s] 8%|▊ | 3960/48008 [13:25<2:40:09, 4.58it/s] 8%|▊ | 3961/48008 [13:25<2:33:15, 4.79it/s] 8%|▊ | 3962/48008 [13:25<2:31:47, 4.84it/s] 8%|▊ | 3963/48008 [13:25<2:30:27, 4.88it/s] 8%|▊ | 3964/48008 [13:26<2:26:50, 5.00it/s] 8%|▊ | 3965/48008 [13:26<2:23:35, 5.11it/s] 8%|▊ | 3966/48008 [13:26<2:22:05, 5.17it/s] 8%|▊ | 3967/48008 [13:26<2:22:17, 5.16it/s] 8%|▊ | 3968/48008 [13:26<2:46:35, 4.41it/s] 8%|▊ | 3969/48008 [13:27<2:38:37, 4.63it/s] 8%|▊ | 3970/48008 [13:27<2:33:02, 4.80it/s] 8%|▊ | 3971/48008 [13:27<2:33:03, 4.80it/s] 8%|▊ | 3972/48008 [13:27<2:29:20, 4.91it/s] 8%|▊ | 3973/48008 [13:27<2:27:11, 4.99it/s] 8%|▊ | 3974/48008 [13:28<2:25:17, 5.05it/s] 8%|▊ | 3975/48008 [13:28<2:23:51, 5.10it/s] 8%|▊ | 3976/48008 [13:28<2:47:51, 4.37it/s] 8%|▊ | 3977/48008 [13:28<2:39:06, 4.61it/s] 8%|▊ | 3978/48008 [13:28<2:33:39, 4.78it/s] 8%|▊ | 3979/48008 [13:29<2:31:35, 4.84it/s] 8%|▊ | 3980/48008 [13:29<2:28:02, 4.96it/s] 8%|▊ | 3981/48008 [13:29<2:50:51, 4.29it/s] 8%|▊ | 3982/48008 [13:29<2:42:36, 4.51it/s] 8%|▊ | 3983/48008 [13:30<2:35:20, 4.72it/s] 8%|▊ | 3984/48008 [13:30<2:30:38, 4.87it/s] 8%|▊ | 3985/48008 [13:30<2:28:49, 4.93it/s] 8%|▊ | 3986/48008 [13:30<2:26:04, 5.02it/s] 8%|▊ | 3987/48008 [13:30<2:24:24, 5.08it/s] 8%|▊ | 3988/48008 [13:31<2:22:49, 5.14it/s] 8%|▊ | 3989/48008 [13:31<2:24:46, 5.07it/s] 8%|▊ | 3990/48008 [13:31<2:22:59, 5.13it/s] 8%|▊ | 3991/48008 [13:31<2:25:18, 5.05it/s] 8%|▊ | 3992/48008 [13:31<2:23:47, 5.10it/s] 8%|▊ | 3993/48008 [13:32<2:22:49, 5.14it/s] 8%|▊ | 3994/48008 [13:32<2:21:28, 5.19it/s] 8%|▊ | 3995/48008 [13:32<2:22:35, 5.14it/s] 8%|▊ | 3996/48008 [13:32<2:23:08, 5.12it/s] 8%|▊ | 3997/48008 [13:32<2:22:28, 5.15it/s] 8%|▊ | 3998/48008 [13:32<2:20:51, 5.21it/s] 8%|▊ | 3999/48008 [13:33<2:20:34, 5.22it/s] 8%|▊ | 4000/48008 [13:33<2:22:28, 5.15it/s] {'loss': 0.3055, 'grad_norm': 0.1455078125, 'learning_rate': 0.00018334027662056325, 'epoch': 0.08} 8%|▊ | 4000/48008 [13:33<2:22:28, 5.15it/s] 8%|▊ | 4001/48008 [13:33<2:21:43, 5.18it/s] 8%|▊ | 4002/48008 [13:33<2:20:39, 5.21it/s] 8%|▊ | 4003/48008 [13:33<2:21:28, 5.18it/s] 8%|▊ | 4004/48008 [13:34<2:20:16, 5.23it/s] 8%|▊ | 4005/48008 [13:34<2:30:18, 4.88it/s] 8%|▊ | 4006/48008 [13:34<2:26:43, 5.00it/s] 8%|▊ | 4007/48008 [13:34<2:23:56, 5.09it/s] 8%|▊ | 4008/48008 [13:34<2:21:36, 5.18it/s] 8%|▊ | 4009/48008 [13:35<2:19:51, 5.24it/s] 8%|▊ | 4010/48008 [13:35<2:20:05, 5.23it/s] 8%|▊ | 4011/48008 [13:35<2:29:33, 4.90it/s] 8%|▊ | 4012/48008 [13:35<2:31:43, 4.83it/s] 8%|▊ | 4013/48008 [13:36<2:53:17, 4.23it/s] 8%|▊ | 4014/48008 [13:36<2:43:23, 4.49it/s] 8%|▊ | 4015/48008 [13:36<3:18:48, 3.69it/s] 8%|▊ | 4016/48008 [13:36<3:26:25, 3.55it/s] 8%|▊ | 4017/48008 [13:37<3:05:39, 3.95it/s] 8%|▊ | 4018/48008 [13:37<2:51:37, 4.27it/s] 8%|▊ | 4019/48008 [13:37<2:46:14, 4.41it/s] 8%|▊ | 4020/48008 [13:37<2:39:06, 4.61it/s] 8%|▊ | 4021/48008 [13:37<2:33:55, 4.76it/s] 8%|▊ | 4022/48008 [13:38<2:30:21, 4.88it/s] 8%|▊ | 4023/48008 [13:38<2:24:38, 5.07it/s] 8%|▊ | 4024/48008 [13:38<2:23:33, 5.11it/s] 8%|▊ | 4025/48008 [13:38<2:22:27, 5.15it/s] 8%|▊ | 4026/48008 [13:38<2:21:36, 5.18it/s] 8%|▊ | 4027/48008 [13:39<2:20:44, 5.21it/s] 8%|▊ | 4028/48008 [13:39<2:20:07, 5.23it/s] 8%|▊ | 4029/48008 [13:39<2:21:00, 5.20it/s] 8%|▊ | 4030/48008 [13:39<2:21:11, 5.19it/s] 8%|▊ | 4031/48008 [13:39<2:21:05, 5.20it/s] 8%|▊ | 4032/48008 [13:40<2:31:06, 4.85it/s] 8%|▊ | 4033/48008 [13:40<2:29:07, 4.91it/s] 8%|▊ | 4034/48008 [13:40<2:26:02, 5.02it/s] 8%|▊ | 4035/48008 [13:40<2:25:48, 5.03it/s] 8%|▊ | 4036/48008 [13:40<2:24:44, 5.06it/s] 8%|▊ | 4037/48008 [13:41<2:22:33, 5.14it/s] 8%|▊ | 4038/48008 [13:41<2:23:20, 5.11it/s] 8%|▊ | 4039/48008 [13:41<2:22:27, 5.14it/s] 8%|▊ | 4040/48008 [13:41<2:22:40, 5.14it/s] 8%|▊ | 4041/48008 [13:41<2:26:35, 5.00it/s] 8%|▊ | 4042/48008 [13:42<2:24:36, 5.07it/s] 8%|▊ | 4043/48008 [13:42<2:23:14, 5.12it/s] 8%|▊ | 4044/48008 [13:42<2:23:26, 5.11it/s] 8%|▊ | 4045/48008 [13:42<2:21:34, 5.18it/s] 8%|▊ | 4046/48008 [13:42<2:21:41, 5.17it/s] 8%|▊ | 4047/48008 [13:42<2:21:08, 5.19it/s] 8%|▊ | 4048/48008 [13:43<2:20:48, 5.20it/s] 8%|▊ | 4049/48008 [13:43<2:19:19, 5.26it/s] 8%|▊ | 4050/48008 [13:43<2:23:42, 5.10it/s] {'loss': 0.303, 'grad_norm': 0.34765625, 'learning_rate': 0.00018313197800366607, 'epoch': 0.08} 8%|▊ | 4050/48008 [13:43<2:23:42, 5.10it/s] 8%|▊ | 4051/48008 [13:43<2:25:57, 5.02it/s] 8%|▊ | 4052/48008 [13:43<2:26:24, 5.00it/s] 8%|▊ | 4053/48008 [13:44<2:49:23, 4.32it/s] 8%|▊ | 4054/48008 [13:44<2:43:00, 4.49it/s] 8%|▊ | 4055/48008 [13:44<2:36:03, 4.69it/s] 8%|▊ | 4056/48008 [13:44<2:29:54, 4.89it/s] 8%|▊ | 4057/48008 [13:45<2:26:50, 4.99it/s] 8%|▊ | 4058/48008 [13:45<2:23:03, 5.12it/s] 8%|▊ | 4059/48008 [13:45<2:24:38, 5.06it/s] 8%|▊ | 4060/48008 [13:45<3:05:57, 3.94it/s] 8%|▊ | 4061/48008 [13:45<2:51:28, 4.27it/s] 8%|▊ | 4062/48008 [13:46<2:41:39, 4.53it/s] 8%|▊ | 4063/48008 [13:46<2:35:58, 4.70it/s] 8%|▊ | 4064/48008 [13:46<2:29:52, 4.89it/s] 8%|▊ | 4065/48008 [13:46<2:29:43, 4.89it/s] 8%|▊ | 4066/48008 [13:46<2:26:29, 5.00it/s] 8%|▊ | 4067/48008 [13:47<2:24:13, 5.08it/s] 8%|▊ | 4068/48008 [13:47<2:21:44, 5.17it/s] 8%|▊ | 4069/48008 [13:47<2:21:23, 5.18it/s] 8%|▊ | 4070/48008 [13:47<2:22:59, 5.12it/s] 8%|▊ | 4071/48008 [13:47<2:23:00, 5.12it/s] 8%|▊ | 4072/48008 [13:48<2:23:43, 5.09it/s] 8%|▊ | 4073/48008 [13:48<2:22:23, 5.14it/s] 8%|▊ | 4074/48008 [13:48<2:22:12, 5.15it/s] 8%|▊ | 4075/48008 [13:48<2:21:08, 5.19it/s] 8%|▊ | 4076/48008 [13:48<2:23:14, 5.11it/s] 8%|▊ | 4077/48008 [13:49<2:24:42, 5.06it/s] 8%|▊ | 4078/48008 [13:49<2:22:06, 5.15it/s] 8%|▊ | 4079/48008 [13:49<2:20:35, 5.21it/s] 8%|▊ | 4080/48008 [13:49<2:19:40, 5.24it/s] 9%|▊ | 4081/48008 [13:49<2:22:06, 5.15it/s] 9%|▊ | 4082/48008 [13:50<2:20:19, 5.22it/s] 9%|▊ | 4083/48008 [13:50<2:22:20, 5.14it/s] 9%|▊ | 4084/48008 [13:50<2:20:33, 5.21it/s] 9%|▊ | 4085/48008 [13:50<2:30:16, 4.87it/s] 9%|▊ | 4086/48008 [13:50<2:27:13, 4.97it/s] 9%|▊ | 4087/48008 [13:51<2:27:11, 4.97it/s] 9%|▊ | 4088/48008 [13:51<2:23:11, 5.11it/s] 9%|▊ | 4089/48008 [13:51<2:20:59, 5.19it/s] 9%|▊ | 4090/48008 [13:51<2:19:49, 5.24it/s] 9%|▊ | 4091/48008 [13:51<2:18:34, 5.28it/s] 9%|▊ | 4092/48008 [13:51<2:19:58, 5.23it/s] 9%|▊ | 4093/48008 [13:52<2:22:17, 5.14it/s] 9%|▊ | 4094/48008 [13:52<3:04:38, 3.96it/s] 9%|▊ | 4095/48008 [13:52<2:50:48, 4.28it/s] 9%|▊ | 4096/48008 [13:52<2:40:29, 4.56it/s] 9%|▊ | 4097/48008 [13:53<2:36:45, 4.67it/s] 9%|▊ | 4098/48008 [13:53<2:35:02, 4.72it/s] 9%|▊ | 4099/48008 [13:53<2:29:49, 4.88it/s] 9%|▊ | 4100/48008 [13:53<2:26:27, 5.00it/s] {'loss': 0.3043, 'grad_norm': 0.205078125, 'learning_rate': 0.00018292367938676889, 'epoch': 0.09} 9%|▊ | 4100/48008 [13:53<2:26:27, 5.00it/s] 9%|▊ | 4101/48008 [13:53<2:27:03, 4.98it/s] 9%|▊ | 4102/48008 [13:54<2:24:10, 5.08it/s] 9%|▊ | 4103/48008 [13:54<2:21:27, 5.17it/s] 9%|▊ | 4104/48008 [13:54<2:21:47, 5.16it/s] 9%|▊ | 4105/48008 [13:54<2:46:30, 4.39it/s] 9%|▊ | 4106/48008 [13:55<2:37:59, 4.63it/s] 9%|▊ | 4107/48008 [13:55<2:32:14, 4.81it/s] 9%|▊ | 4108/48008 [13:55<2:28:36, 4.92it/s] 9%|▊ | 4109/48008 [13:55<2:27:09, 4.97it/s] 9%|▊ | 4110/48008 [13:55<2:25:09, 5.04it/s] 9%|▊ | 4111/48008 [13:55<2:22:38, 5.13it/s] 9%|▊ | 4112/48008 [13:56<2:21:09, 5.18it/s] 9%|▊ | 4113/48008 [13:56<2:20:55, 5.19it/s] 9%|▊ | 4114/48008 [13:56<2:19:43, 5.24it/s] 9%|▊ | 4115/48008 [13:56<2:22:06, 5.15it/s] 9%|▊ | 4116/48008 [13:56<2:20:55, 5.19it/s] 9%|▊ | 4117/48008 [13:57<2:24:20, 5.07it/s] 9%|▊ | 4118/48008 [13:57<2:33:13, 4.77it/s] 9%|▊ | 4119/48008 [13:57<3:11:43, 3.82it/s] 9%|▊ | 4120/48008 [13:57<2:55:51, 4.16it/s] 9%|▊ | 4121/48008 [13:58<2:47:29, 4.37it/s] 9%|▊ | 4122/48008 [13:58<2:49:27, 4.32it/s] 9%|▊ | 4123/48008 [13:58<2:40:27, 4.56it/s] 9%|▊ | 4124/48008 [13:58<2:33:53, 4.75it/s] 9%|▊ | 4125/48008 [13:58<2:28:25, 4.93it/s] 9%|▊ | 4126/48008 [13:59<2:25:27, 5.03it/s] 9%|▊ | 4127/48008 [13:59<2:24:56, 5.05it/s] 9%|▊ | 4128/48008 [13:59<2:26:06, 5.01it/s] 9%|▊ | 4129/48008 [13:59<2:25:31, 5.03it/s] 9%|▊ | 4130/48008 [13:59<2:22:45, 5.12it/s] 9%|▊ | 4131/48008 [14:00<2:21:41, 5.16it/s] 9%|▊ | 4132/48008 [14:00<2:21:58, 5.15it/s] 9%|▊ | 4133/48008 [14:00<2:20:37, 5.20it/s] 9%|▊ | 4134/48008 [14:00<2:24:11, 5.07it/s] 9%|▊ | 4135/48008 [14:01<2:47:57, 4.35it/s] 9%|▊ | 4136/48008 [14:01<2:42:07, 4.51it/s] 9%|▊ | 4137/48008 [14:01<2:39:19, 4.59it/s] 9%|▊ | 4138/48008 [14:01<3:16:20, 3.72it/s] 9%|▊ | 4139/48008 [14:02<2:59:39, 4.07it/s] 9%|▊ | 4140/48008 [14:02<3:12:48, 3.79it/s] 9%|▊ | 4141/48008 [14:02<3:00:41, 4.05it/s] 9%|▊ | 4142/48008 [14:02<2:48:18, 4.34it/s] 9%|▊ | 4143/48008 [14:02<2:39:43, 4.58it/s] 9%|▊ | 4144/48008 [14:03<2:33:12, 4.77it/s] 9%|▊ | 4145/48008 [14:03<2:28:41, 4.92it/s] 9%|▊ | 4146/48008 [14:03<2:25:27, 5.03it/s] 9%|▊ | 4147/48008 [14:03<3:06:00, 3.93it/s] 9%|▊ | 4148/48008 [14:04<2:50:55, 4.28it/s] 9%|▊ | 4149/48008 [14:04<2:42:25, 4.50it/s] 9%|▊ | 4150/48008 [14:04<2:36:26, 4.67it/s] {'loss': 0.3017, 'grad_norm': 0.640625, 'learning_rate': 0.0001827153807698717, 'epoch': 0.09} 9%|▊ | 4150/48008 [14:04<2:36:26, 4.67it/s] 9%|▊ | 4151/48008 [14:04<2:33:09, 4.77it/s] 9%|▊ | 4152/48008 [14:04<2:32:44, 4.79it/s] 9%|▊ | 4153/48008 [14:05<2:28:50, 4.91it/s] 9%|▊ | 4154/48008 [14:05<2:27:04, 4.97it/s] 9%|▊ | 4155/48008 [14:05<2:28:09, 4.93it/s] 9%|▊ | 4156/48008 [14:05<2:25:39, 5.02it/s] 9%|▊ | 4157/48008 [14:05<2:23:59, 5.08it/s] 9%|▊ | 4158/48008 [14:05<2:21:42, 5.16it/s] 9%|▊ | 4159/48008 [14:06<2:21:38, 5.16it/s] 9%|▊ | 4160/48008 [14:06<2:20:20, 5.21it/s] 9%|▊ | 4161/48008 [14:06<2:19:52, 5.22it/s] 9%|▊ | 4162/48008 [14:06<2:20:05, 5.22it/s] 9%|▊ | 4163/48008 [14:06<2:20:01, 5.22it/s] 9%|▊ | 4164/48008 [14:07<2:45:40, 4.41it/s] 9%|▊ | 4165/48008 [14:07<2:36:51, 4.66it/s] 9%|▊ | 4166/48008 [14:07<3:14:25, 3.76it/s] 9%|▊ | 4167/48008 [14:08<2:59:28, 4.07it/s] 9%|▊ | 4168/48008 [14:08<2:47:19, 4.37it/s] 9%|▊ | 4169/48008 [14:08<2:38:28, 4.61it/s] 9%|▊ | 4170/48008 [14:08<2:35:00, 4.71it/s] 9%|▊ | 4171/48008 [14:08<2:29:13, 4.90it/s] 9%|▊ | 4172/48008 [14:08<2:28:21, 4.92it/s] 9%|▊ | 4173/48008 [14:09<2:25:31, 5.02it/s] 9%|▊ | 4174/48008 [14:09<2:24:23, 5.06it/s] 9%|▊ | 4175/48008 [14:09<2:25:24, 5.02it/s] 9%|▊ | 4176/48008 [14:09<2:23:56, 5.08it/s] 9%|▊ | 4177/48008 [14:09<2:21:15, 5.17it/s] 9%|▊ | 4178/48008 [14:10<2:19:53, 5.22it/s] 9%|▊ | 4179/48008 [14:10<2:20:40, 5.19it/s] 9%|▊ | 4180/48008 [14:10<2:20:16, 5.21it/s] 9%|▊ | 4181/48008 [14:10<2:19:46, 5.23it/s] 9%|▊ | 4182/48008 [14:10<2:20:47, 5.19it/s] 9%|▊ | 4183/48008 [14:11<2:22:06, 5.14it/s] 9%|▊ | 4184/48008 [14:11<2:24:03, 5.07it/s] 9%|▊ | 4185/48008 [14:11<2:23:38, 5.08it/s] 9%|▊ | 4186/48008 [14:11<2:21:15, 5.17it/s] 9%|▊ | 4187/48008 [14:11<2:20:50, 5.19it/s] 9%|▊ | 4188/48008 [14:12<2:19:46, 5.23it/s] 9%|▊ | 4189/48008 [14:12<2:18:39, 5.27it/s] 9%|▊ | 4190/48008 [14:12<2:20:59, 5.18it/s] 9%|▊ | 4191/48008 [14:12<2:21:44, 5.15it/s] 9%|▊ | 4192/48008 [14:12<2:20:53, 5.18it/s] 9%|▊ | 4193/48008 [14:13<2:20:13, 5.21it/s] 9%|▊ | 4194/48008 [14:13<2:21:51, 5.15it/s] 9%|▊ | 4195/48008 [14:13<2:31:08, 4.83it/s] 9%|▊ | 4196/48008 [14:13<2:28:28, 4.92it/s] 9%|▊ | 4197/48008 [14:13<2:25:20, 5.02it/s] 9%|▊ | 4198/48008 [14:14<2:23:37, 5.08it/s] 9%|▊ | 4199/48008 [14:14<2:22:12, 5.13it/s] 9%|▊ | 4200/48008 [14:14<2:31:20, 4.82it/s] 9%|▊ | 4200/48008 [14:14<2:31:20, 4.82it/s]{'loss': 0.3028, 'grad_norm': 0.7265625, 'learning_rate': 0.00018250708215297452, 'epoch': 0.09} 9%|▉ | 4201/48008 [14:14<2:28:13, 4.93it/s] 9%|▉ | 4202/48008 [14:15<3:08:19, 3.88it/s] 9%|▉ | 4203/48008 [14:15<2:52:21, 4.24it/s] 9%|▉ | 4204/48008 [14:15<2:44:16, 4.44it/s] 9%|▉ | 4205/48008 [14:15<2:36:29, 4.66it/s] 9%|▉ | 4206/48008 [14:15<2:30:55, 4.84it/s] 9%|▉ | 4207/48008 [14:16<2:27:17, 4.96it/s] 9%|▉ | 4208/48008 [14:16<2:24:53, 5.04it/s] 9%|▉ | 4209/48008 [14:16<2:25:18, 5.02it/s] 9%|▉ | 4210/48008 [14:16<2:23:06, 5.10it/s] 9%|▉ | 4211/48008 [14:16<2:23:30, 5.09it/s] 9%|▉ | 4212/48008 [14:16<2:21:44, 5.15it/s] 9%|▉ | 4213/48008 [14:17<2:22:30, 5.12it/s] 9%|▉ | 4214/48008 [14:17<2:22:56, 5.11it/s] 9%|▉ | 4215/48008 [14:17<2:24:11, 5.06it/s] 9%|▉ | 4216/48008 [14:17<2:21:38, 5.15it/s] 9%|▉ | 4217/48008 [14:17<2:22:07, 5.14it/s] 9%|▉ | 4218/48008 [14:18<2:20:37, 5.19it/s] 9%|▉ | 4219/48008 [14:18<2:19:53, 5.22it/s] 9%|▉ | 4220/48008 [14:18<2:23:03, 5.10it/s] 9%|▉ | 4221/48008 [14:18<2:47:24, 4.36it/s] 9%|▉ | 4222/48008 [14:19<2:41:10, 4.53it/s] 9%|▉ | 4223/48008 [14:19<2:33:59, 4.74it/s] 9%|▉ | 4224/48008 [14:19<2:29:04, 4.90it/s] 9%|▉ | 4225/48008 [14:19<2:27:04, 4.96it/s] 9%|▉ | 4226/48008 [14:19<2:24:47, 5.04it/s] 9%|▉ | 4227/48008 [14:20<2:24:44, 5.04it/s] 9%|▉ | 4228/48008 [14:20<2:23:06, 5.10it/s] 9%|▉ | 4229/48008 [14:20<2:21:28, 5.16it/s] 9%|▉ | 4230/48008 [14:20<2:19:38, 5.23it/s] 9%|▉ | 4231/48008 [14:20<2:44:35, 4.43it/s] 9%|▉ | 4232/48008 [14:21<2:36:38, 4.66it/s] 9%|▉ | 4233/48008 [14:21<2:41:34, 4.52it/s] 9%|▉ | 4234/48008 [14:21<2:35:48, 4.68it/s] 9%|▉ | 4235/48008 [14:21<2:30:45, 4.84it/s] 9%|▉ | 4236/48008 [14:21<2:27:20, 4.95it/s] 9%|▉ | 4237/48008 [14:22<2:34:56, 4.71it/s] 9%|▉ | 4238/48008 [14:22<2:30:55, 4.83it/s] 9%|▉ | 4239/48008 [14:22<2:28:48, 4.90it/s] 9%|▉ | 4240/48008 [14:22<2:27:29, 4.95it/s] 9%|▉ | 4241/48008 [14:22<2:27:23, 4.95it/s] 9%|▉ | 4242/48008 [14:23<2:24:34, 5.05it/s] 9%|▉ | 4243/48008 [14:23<2:22:35, 5.12it/s] 9%|▉ | 4244/48008 [14:23<2:21:49, 5.14it/s] 9%|▉ | 4245/48008 [14:23<2:46:16, 4.39it/s] 9%|▉ | 4246/48008 [14:23<2:37:23, 4.63it/s] 9%|▉ | 4247/48008 [14:24<2:34:29, 4.72it/s] 9%|▉ | 4248/48008 [14:24<2:29:36, 4.88it/s] 9%|▉ | 4249/48008 [14:24<2:25:59, 5.00it/s] 9%|▉ | 4250/48008 [14:24<2:24:01, 5.06it/s] {'loss': 0.3025, 'grad_norm': 0.24609375, 'learning_rate': 0.0001822987835360773, 'epoch': 0.09} 9%|▉ | 4250/48008 [14:24<2:24:01, 5.06it/s] 9%|▉ | 4251/48008 [14:24<2:23:13, 5.09it/s] 9%|▉ | 4252/48008 [14:25<2:23:37, 5.08it/s] 9%|▉ | 4253/48008 [14:25<2:22:09, 5.13it/s] 9%|▉ | 4254/48008 [14:25<2:23:29, 5.08it/s] 9%|▉ | 4255/48008 [14:25<2:22:11, 5.13it/s] 9%|▉ | 4256/48008 [14:25<2:21:49, 5.14it/s] 9%|▉ | 4257/48008 [14:26<2:20:16, 5.20it/s] 9%|▉ | 4258/48008 [14:26<3:02:42, 3.99it/s] 9%|▉ | 4259/48008 [14:26<2:48:12, 4.33it/s] 9%|▉ | 4260/48008 [14:26<2:39:13, 4.58it/s] 9%|▉ | 4261/48008 [14:27<2:33:01, 4.76it/s] 9%|▉ | 4262/48008 [14:27<2:32:31, 4.78it/s] 9%|▉ | 4263/48008 [14:27<2:27:54, 4.93it/s] 9%|▉ | 4264/48008 [14:27<2:29:44, 4.87it/s] 9%|▉ | 4265/48008 [14:27<2:26:23, 4.98it/s] 9%|▉ | 4266/48008 [14:28<2:24:13, 5.05it/s] 9%|▉ | 4267/48008 [14:28<2:21:49, 5.14it/s] 9%|▉ | 4268/48008 [14:28<3:03:19, 3.98it/s] 9%|▉ | 4269/48008 [14:28<2:53:08, 4.21it/s] 9%|▉ | 4270/48008 [14:29<2:46:50, 4.37it/s] 9%|▉ | 4271/48008 [14:29<2:40:48, 4.53it/s] 9%|▉ | 4272/48008 [14:29<2:35:22, 4.69it/s] 9%|▉ | 4273/48008 [14:29<2:30:37, 4.84it/s] 9%|▉ | 4274/48008 [14:29<2:26:58, 4.96it/s] 9%|▉ | 4275/48008 [14:29<2:24:47, 5.03it/s] 9%|▉ | 4276/48008 [14:30<2:22:44, 5.11it/s] 9%|▉ | 4277/48008 [14:30<2:20:25, 5.19it/s] 9%|▉ | 4278/48008 [14:30<2:29:49, 4.86it/s] 9%|▉ | 4279/48008 [14:30<2:26:05, 4.99it/s] 9%|▉ | 4280/48008 [14:30<2:24:06, 5.06it/s] 9%|▉ | 4281/48008 [14:31<2:32:27, 4.78it/s] 9%|▉ | 4282/48008 [14:31<2:28:37, 4.90it/s] 9%|▉ | 4283/48008 [14:31<2:25:54, 4.99it/s] 9%|▉ | 4284/48008 [14:31<2:27:54, 4.93it/s] 9%|▉ | 4285/48008 [14:31<2:22:41, 5.11it/s] 9%|▉ | 4286/48008 [14:32<2:20:52, 5.17it/s] 9%|▉ | 4287/48008 [14:32<2:19:16, 5.23it/s] 9%|▉ | 4288/48008 [14:32<2:21:24, 5.15it/s] 9%|▉ | 4289/48008 [14:32<2:20:53, 5.17it/s] 9%|▉ | 4290/48008 [14:32<2:19:43, 5.22it/s] 9%|▉ | 4291/48008 [14:33<2:20:01, 5.20it/s] 9%|▉ | 4292/48008 [14:33<2:20:49, 5.17it/s] 9%|▉ | 4293/48008 [14:33<2:19:06, 5.24it/s] 9%|▉ | 4294/48008 [14:33<2:18:25, 5.26it/s] 9%|▉ | 4295/48008 [14:33<2:18:30, 5.26it/s] 9%|▉ | 4296/48008 [14:34<2:18:39, 5.25it/s] 9%|▉ | 4297/48008 [14:34<2:17:41, 5.29it/s] 9%|▉ | 4298/48008 [14:34<2:19:03, 5.24it/s] 9%|▉ | 4299/48008 [14:34<2:18:13, 5.27it/s] 9%|▉ | 4300/48008 [14:34<2:18:41, 5.25it/s] {'loss': 0.3044, 'grad_norm': 0.59765625, 'learning_rate': 0.00018209048491918015, 'epoch': 0.09} 9%|▉ | 4300/48008 [14:34<2:18:41, 5.25it/s] 9%|▉ | 4301/48008 [14:35<2:18:56, 5.24it/s] 9%|▉ | 4302/48008 [14:35<2:19:33, 5.22it/s] 9%|▉ | 4303/48008 [14:35<2:20:24, 5.19it/s] 9%|▉ | 4304/48008 [14:35<2:19:57, 5.20it/s] 9%|▉ | 4305/48008 [14:35<2:18:54, 5.24it/s] 9%|▉ | 4306/48008 [14:36<2:18:19, 5.27it/s] 9%|▉ | 4307/48008 [14:36<2:18:32, 5.26it/s] 9%|▉ | 4308/48008 [14:36<2:18:06, 5.27it/s] 9%|▉ | 4309/48008 [14:36<2:19:25, 5.22it/s] 9%|▉ | 4310/48008 [14:36<3:01:45, 4.01it/s] 9%|▉ | 4311/48008 [14:37<2:50:08, 4.28it/s] 9%|▉ | 4312/48008 [14:37<2:40:44, 4.53it/s] 9%|▉ | 4313/48008 [14:37<2:35:37, 4.68it/s] 9%|▉ | 4314/48008 [14:37<2:29:49, 4.86it/s] 9%|▉ | 4315/48008 [14:37<2:26:38, 4.97it/s] 9%|▉ | 4316/48008 [14:38<2:49:28, 4.30it/s] 9%|▉ | 4317/48008 [14:38<2:40:56, 4.52it/s] 9%|▉ | 4318/48008 [14:38<2:34:40, 4.71it/s] 9%|▉ | 4319/48008 [14:38<2:30:54, 4.83it/s] 9%|▉ | 4320/48008 [14:39<2:27:39, 4.93it/s] 9%|▉ | 4321/48008 [14:39<2:25:14, 5.01it/s] 9%|▉ | 4322/48008 [14:39<2:26:50, 4.96it/s] 9%|▉ | 4323/48008 [14:39<2:27:43, 4.93it/s] 9%|▉ | 4324/48008 [14:39<2:28:31, 4.90it/s] 9%|▉ | 4325/48008 [14:40<2:25:12, 5.01it/s] 9%|▉ | 4326/48008 [14:40<2:23:28, 5.07it/s] 9%|▉ | 4327/48008 [14:40<2:22:45, 5.10it/s] 9%|▉ | 4328/48008 [14:40<2:24:21, 5.04it/s] 9%|▉ | 4329/48008 [14:40<2:22:55, 5.09it/s] 9%|▉ | 4330/48008 [14:41<2:46:14, 4.38it/s] 9%|▉ | 4331/48008 [14:41<2:38:09, 4.60it/s] 9%|▉ | 4332/48008 [14:41<2:31:21, 4.81it/s] 9%|▉ | 4333/48008 [14:41<2:27:03, 4.95it/s] 9%|▉ | 4334/48008 [14:41<2:25:03, 5.02it/s] 9%|▉ | 4335/48008 [14:42<2:24:44, 5.03it/s] 9%|▉ | 4336/48008 [14:42<2:24:31, 5.04it/s] 9%|▉ | 4337/48008 [14:42<2:32:23, 4.78it/s] 9%|▉ | 4338/48008 [14:42<2:28:21, 4.91it/s] 9%|▉ | 4339/48008 [14:42<2:26:47, 4.96it/s] 9%|▉ | 4340/48008 [14:43<2:24:51, 5.02it/s] 9%|▉ | 4341/48008 [14:43<2:22:57, 5.09it/s] 9%|▉ | 4342/48008 [14:43<2:21:50, 5.13it/s] 9%|▉ | 4343/48008 [14:43<2:20:10, 5.19it/s] 9%|▉ | 4344/48008 [14:43<2:21:12, 5.15it/s] 9%|▉ | 4345/48008 [14:44<3:03:18, 3.97it/s] 9%|▉ | 4346/48008 [14:44<3:32:26, 3.43it/s] 9%|▉ | 4347/48008 [14:44<3:10:23, 3.82it/s] 9%|▉ | 4348/48008 [14:44<2:54:36, 4.17it/s] 9%|▉ | 4349/48008 [14:45<2:44:13, 4.43it/s] 9%|▉ | 4350/48008 [14:45<2:36:35, 4.65it/s] {'loss': 0.3025, 'grad_norm': 0.369140625, 'learning_rate': 0.00018188218630228294, 'epoch': 0.09} 9%|▉ | 4350/48008 [14:45<2:36:35, 4.65it/s] 9%|▉ | 4351/48008 [14:45<2:32:35, 4.77it/s] 9%|▉ | 4352/48008 [14:45<2:30:22, 4.84it/s] 9%|▉ | 4353/48008 [14:45<2:28:20, 4.90it/s] 9%|▉ | 4354/48008 [14:46<2:25:10, 5.01it/s] 9%|▉ | 4355/48008 [14:46<2:26:33, 4.96it/s] 9%|▉ | 4356/48008 [14:46<2:23:18, 5.08it/s] 9%|▉ | 4357/48008 [14:46<2:21:54, 5.13it/s] 9%|▉ | 4358/48008 [14:46<2:21:01, 5.16it/s] 9%|▉ | 4359/48008 [14:47<2:21:01, 5.16it/s] 9%|▉ | 4360/48008 [14:47<2:19:54, 5.20it/s] 9%|▉ | 4361/48008 [14:47<2:19:49, 5.20it/s] 9%|▉ | 4362/48008 [14:47<2:22:09, 5.12it/s] 9%|▉ | 4363/48008 [14:47<2:20:33, 5.18it/s] 9%|▉ | 4364/48008 [14:48<2:19:44, 5.21it/s] 9%|▉ | 4365/48008 [14:48<2:20:10, 5.19it/s] 9%|▉ | 4366/48008 [14:48<2:19:57, 5.20it/s] 9%|▉ | 4367/48008 [14:48<2:22:14, 5.11it/s] 9%|▉ | 4368/48008 [14:49<3:03:47, 3.96it/s] 9%|▉ | 4369/48008 [14:49<2:49:26, 4.29it/s] 9%|▉ | 4370/48008 [14:49<2:39:48, 4.55it/s] 9%|▉ | 4371/48008 [14:49<2:33:14, 4.75it/s] 9%|▉ | 4372/48008 [14:49<2:28:31, 4.90it/s] 9%|▉ | 4373/48008 [14:49<2:25:54, 4.98it/s] 9%|▉ | 4374/48008 [14:50<2:23:58, 5.05it/s] 9%|▉ | 4375/48008 [14:50<2:22:18, 5.11it/s] 9%|▉ | 4376/48008 [14:50<2:22:27, 5.10it/s] 9%|▉ | 4377/48008 [14:50<2:23:55, 5.05it/s] 9%|▉ | 4378/48008 [14:51<3:04:56, 3.93it/s] 9%|▉ | 4379/48008 [14:51<2:53:44, 4.19it/s] 9%|▉ | 4380/48008 [14:51<2:42:03, 4.49it/s] 9%|▉ | 4381/48008 [14:51<2:33:50, 4.73it/s] 9%|▉ | 4382/48008 [14:51<2:29:09, 4.87it/s] 9%|▉ | 4383/48008 [14:52<2:26:02, 4.98it/s] 9%|▉ | 4384/48008 [14:52<2:23:11, 5.08it/s] 9%|▉ | 4385/48008 [14:52<2:21:21, 5.14it/s] 9%|▉ | 4386/48008 [14:52<2:20:01, 5.19it/s] 9%|▉ | 4387/48008 [14:52<2:18:53, 5.23it/s] 9%|▉ | 4388/48008 [14:53<2:19:06, 5.23it/s] 9%|▉ | 4389/48008 [14:53<2:16:47, 5.31it/s] 9%|▉ | 4390/48008 [14:53<2:18:36, 5.24it/s] 9%|▉ | 4391/48008 [14:53<2:20:24, 5.18it/s] 9%|▉ | 4392/48008 [14:53<2:20:16, 5.18it/s] 9%|▉ | 4393/48008 [14:54<2:19:52, 5.20it/s] 9%|▉ | 4394/48008 [14:54<2:19:04, 5.23it/s] 9%|▉ | 4395/48008 [14:54<2:17:59, 5.27it/s] 9%|▉ | 4396/48008 [14:54<2:18:13, 5.26it/s] 9%|▉ | 4397/48008 [14:54<2:18:22, 5.25it/s] 9%|▉ | 4398/48008 [14:54<2:21:16, 5.15it/s] 9%|▉ | 4399/48008 [14:55<2:21:54, 5.12it/s] 9%|▉ | 4400/48008 [14:55<2:21:57, 5.12it/s] {'loss': 0.3048, 'grad_norm': 0.1630859375, 'learning_rate': 0.0001816738876853858, 'epoch': 0.09} 9%|▉ | 4400/48008 [14:55<2:21:57, 5.12it/s] 9%|▉ | 4401/48008 [14:55<3:04:06, 3.95it/s] 9%|▉ | 4402/48008 [14:55<2:50:46, 4.26it/s] 9%|▉ | 4403/48008 [14:56<2:40:49, 4.52it/s] 9%|▉ | 4404/48008 [14:56<2:34:05, 4.72it/s] 9%|▉ | 4405/48008 [14:56<2:27:48, 4.92it/s] 9%|▉ | 4406/48008 [14:56<2:24:07, 5.04it/s] 9%|▉ | 4407/48008 [14:56<2:22:33, 5.10it/s] 9%|▉ | 4408/48008 [14:57<2:19:57, 5.19it/s] 9%|▉ | 4409/48008 [14:57<2:44:53, 4.41it/s] 9%|▉ | 4410/48008 [14:57<2:36:49, 4.63it/s] 9%|▉ | 4411/48008 [14:57<2:29:52, 4.85it/s] 9%|▉ | 4412/48008 [14:57<2:36:11, 4.65it/s] 9%|▉ | 4413/48008 [14:58<2:30:52, 4.82it/s] 9%|▉ | 4414/48008 [14:58<2:30:16, 4.84it/s] 9%|▉ | 4415/48008 [14:58<2:25:51, 4.98it/s] 9%|▉ | 4416/48008 [14:58<2:48:29, 4.31it/s] 9%|▉ | 4417/48008 [14:59<2:43:28, 4.44it/s] 9%|▉ | 4418/48008 [14:59<2:35:59, 4.66it/s] 9%|▉ | 4419/48008 [14:59<2:32:19, 4.77it/s] 9%|▉ | 4420/48008 [14:59<2:28:18, 4.90it/s] 9%|▉ | 4421/48008 [14:59<2:26:55, 4.94it/s] 9%|▉ | 4422/48008 [15:00<2:22:53, 5.08it/s] 9%|▉ | 4423/48008 [15:00<2:21:38, 5.13it/s] 9%|▉ | 4424/48008 [15:00<2:23:20, 5.07it/s] 9%|▉ | 4425/48008 [15:00<2:21:28, 5.13it/s] 9%|▉ | 4426/48008 [15:00<2:30:25, 4.83it/s] 9%|▉ | 4427/48008 [15:01<2:25:34, 4.99it/s] 9%|▉ | 4428/48008 [15:01<2:23:34, 5.06it/s] 9%|▉ | 4429/48008 [15:01<2:31:54, 4.78it/s] 9%|▉ | 4430/48008 [15:01<2:27:08, 4.94it/s] 9%|▉ | 4431/48008 [15:01<2:24:09, 5.04it/s] 9%|▉ | 4432/48008 [15:02<2:22:36, 5.09it/s] 9%|▉ | 4433/48008 [15:02<2:21:09, 5.14it/s] 9%|▉ | 4434/48008 [15:02<2:19:59, 5.19it/s] 9%|▉ | 4435/48008 [15:02<2:19:00, 5.22it/s] 9%|▉ | 4436/48008 [15:02<2:20:12, 5.18it/s] 9%|▉ | 4437/48008 [15:02<2:18:42, 5.24it/s] 9%|▉ | 4438/48008 [15:03<2:20:05, 5.18it/s] 9%|▉ | 4439/48008 [15:03<2:22:21, 5.10it/s] 9%|▉ | 4440/48008 [15:03<2:21:16, 5.14it/s] 9%|▉ | 4441/48008 [15:03<2:20:04, 5.18it/s] 9%|▉ | 4442/48008 [15:03<2:19:31, 5.20it/s] 9%|▉ | 4443/48008 [15:04<2:20:46, 5.16it/s] 9%|▉ | 4444/48008 [15:04<2:45:07, 4.40it/s] 9%|▉ | 4445/48008 [15:04<2:38:43, 4.57it/s] 9%|▉ | 4446/48008 [15:04<2:31:33, 4.79it/s] 9%|▉ | 4447/48008 [15:05<2:27:25, 4.92it/s] 9%|▉ | 4448/48008 [15:05<2:24:31, 5.02it/s] 9%|▉ | 4449/48008 [15:05<2:21:53, 5.12it/s] 9%|▉ | 4450/48008 [15:05<2:23:04, 5.07it/s] {'loss': 0.3058, 'grad_norm': 0.88671875, 'learning_rate': 0.00018146558906848858, 'epoch': 0.09} 9%|▉ | 4450/48008 [15:05<2:23:04, 5.07it/s] 9%|▉ | 4451/48008 [15:05<2:21:03, 5.15it/s] 9%|▉ | 4452/48008 [15:06<2:22:42, 5.09it/s] 9%|▉ | 4453/48008 [15:06<2:22:08, 5.11it/s] 9%|▉ | 4454/48008 [15:06<2:20:50, 5.15it/s] 9%|▉ | 4455/48008 [15:06<2:22:47, 5.08it/s] 9%|▉ | 4456/48008 [15:06<2:22:32, 5.09it/s] 9%|▉ | 4457/48008 [15:07<2:25:17, 5.00it/s] 9%|▉ | 4458/48008 [15:07<2:25:39, 4.98it/s] 9%|▉ | 4459/48008 [15:07<3:05:57, 3.90it/s] 9%|▉ | 4460/48008 [15:07<2:50:49, 4.25it/s] 9%|▉ | 4461/48008 [15:07<2:40:43, 4.52it/s] 9%|▉ | 4462/48008 [15:08<2:34:16, 4.70it/s] 9%|▉ | 4463/48008 [15:08<2:33:15, 4.74it/s] 9%|▉ | 4464/48008 [15:08<2:28:18, 4.89it/s] 9%|▉ | 4465/48008 [15:08<2:24:56, 5.01it/s] 9%|▉ | 4466/48008 [15:08<2:27:12, 4.93it/s] 9%|▉ | 4467/48008 [15:09<2:28:28, 4.89it/s] 9%|▉ | 4468/48008 [15:09<2:24:10, 5.03it/s] 9%|▉ | 4469/48008 [15:09<2:21:16, 5.14it/s] 9%|▉ | 4470/48008 [15:09<2:20:50, 5.15it/s] 9%|▉ | 4471/48008 [15:09<2:19:19, 5.21it/s] 9%|▉ | 4472/48008 [15:10<2:44:03, 4.42it/s] 9%|▉ | 4473/48008 [15:10<2:36:23, 4.64it/s] 9%|▉ | 4474/48008 [15:10<2:34:35, 4.69it/s] 9%|▉ | 4475/48008 [15:10<2:31:31, 4.79it/s] 9%|▉ | 4476/48008 [15:11<2:27:14, 4.93it/s] 9%|▉ | 4477/48008 [15:11<2:24:14, 5.03it/s] 9%|▉ | 4478/48008 [15:11<2:23:41, 5.05it/s] 9%|▉ | 4479/48008 [15:11<2:22:20, 5.10it/s] 9%|▉ | 4480/48008 [15:11<2:20:06, 5.18it/s] 9%|▉ | 4481/48008 [15:12<2:29:44, 4.84it/s] 9%|▉ | 4482/48008 [15:12<2:26:08, 4.96it/s] 9%|▉ | 4483/48008 [15:12<2:23:36, 5.05it/s] 9%|▉ | 4484/48008 [15:12<2:21:55, 5.11it/s] 9%|▉ | 4485/48008 [15:12<2:21:13, 5.14it/s] 9%|▉ | 4486/48008 [15:12<2:20:21, 5.17it/s] 9%|▉ | 4487/48008 [15:13<2:21:54, 5.11it/s] 9%|▉ | 4488/48008 [15:13<2:20:41, 5.16it/s] 9%|▉ | 4489/48008 [15:13<2:20:58, 5.15it/s] 9%|▉ | 4490/48008 [15:13<2:18:56, 5.22it/s] 9%|▉ | 4491/48008 [15:13<2:29:06, 4.86it/s] 9%|▉ | 4492/48008 [15:14<2:25:44, 4.98it/s] 9%|▉ | 4493/48008 [15:14<2:23:56, 5.04it/s] 9%|▉ | 4494/48008 [15:14<2:21:39, 5.12it/s] 9%|▉ | 4495/48008 [15:14<2:23:08, 5.07it/s] 9%|▉ | 4496/48008 [15:15<3:03:49, 3.95it/s] 9%|▉ | 4497/48008 [15:15<2:50:10, 4.26it/s] 9%|▉ | 4498/48008 [15:15<2:40:30, 4.52it/s] 9%|▉ | 4499/48008 [15:15<2:34:02, 4.71it/s] 9%|▉ | 4500/48008 [15:15<2:33:29, 4.72it/s] {'loss': 0.3011, 'grad_norm': 0.2119140625, 'learning_rate': 0.0001812572904515914, 'epoch': 0.09} 9%|▉ | 4500/48008 [15:15<2:33:29, 4.72it/s] 9%|▉ | 4501/48008 [15:16<2:28:38, 4.88it/s] 9%|▉ | 4502/48008 [15:16<2:25:27, 4.99it/s] 9%|▉ | 4503/48008 [15:16<2:23:05, 5.07it/s] 9%|▉ | 4504/48008 [15:16<2:21:08, 5.14it/s] 9%|▉ | 4505/48008 [15:16<2:20:50, 5.15it/s] 9%|▉ | 4506/48008 [15:17<2:19:48, 5.19it/s] 9%|▉ | 4507/48008 [15:17<2:20:41, 5.15it/s] 9%|▉ | 4508/48008 [15:17<2:18:50, 5.22it/s] 9%|▉ | 4509/48008 [15:17<2:21:24, 5.13it/s] 9%|▉ | 4510/48008 [15:17<2:20:56, 5.14it/s] 9%|▉ | 4511/48008 [15:18<2:20:05, 5.18it/s] 9%|▉ | 4512/48008 [15:19<5:11:58, 2.32it/s] 9%|▉ | 4513/48008 [15:19<4:19:50, 2.79it/s] 9%|▉ | 4514/48008 [15:19<3:44:35, 3.23it/s] 9%|▉ | 4515/48008 [15:19<3:20:15, 3.62it/s] 9%|▉ | 4516/48008 [15:19<3:01:27, 3.99it/s] 9%|▉ | 4517/48008 [15:19<2:50:54, 4.24it/s] 9%|▉ | 4518/48008 [15:20<5:13:34, 2.31it/s] 9%|▉ | 4519/48008 [15:21<4:20:19, 2.78it/s] 9%|▉ | 4520/48008 [15:21<3:45:13, 3.22it/s] 9%|▉ | 4521/48008 [15:21<3:20:03, 3.62it/s] 9%|▉ | 4522/48008 [15:21<3:01:51, 3.99it/s] 9%|▉ | 4523/48008 [15:21<2:49:43, 4.27it/s] 9%|▉ | 4524/48008 [15:23<6:17:40, 1.92it/s] 9%|▉ | 4525/48008 [15:23<5:48:15, 2.08it/s] 9%|▉ | 4526/48008 [15:23<4:47:53, 2.52it/s] 9%|▉ | 4527/48008 [15:23<4:02:50, 2.98it/s] 9%|▉ | 4528/48008 [15:24<3:56:26, 3.06it/s] 9%|▉ | 4529/48008 [15:24<3:27:16, 3.50it/s] 9%|▉ | 4530/48008 [15:24<3:06:37, 3.88it/s] 9%|▉ | 4531/48008 [15:24<2:52:14, 4.21it/s] 9%|▉ | 4532/48008 [15:24<2:42:55, 4.45it/s] 9%|▉ | 4533/48008 [15:25<2:39:10, 4.55it/s] 9%|▉ | 4534/48008 [15:25<2:31:51, 4.77it/s] 9%|▉ | 4535/48008 [15:25<2:26:36, 4.94it/s] 9%|▉ | 4536/48008 [15:25<2:23:34, 5.05it/s] 9%|▉ | 4537/48008 [15:25<2:21:26, 5.12it/s] 9%|▉ | 4538/48008 [15:26<3:02:53, 3.96it/s] 9%|▉ | 4539/48008 [15:26<2:52:11, 4.21it/s] 9%|▉ | 4540/48008 [15:26<2:42:35, 4.46it/s] 9%|▉ | 4541/48008 [15:26<2:35:04, 4.67it/s] 9%|▉ | 4542/48008 [15:27<2:31:48, 4.77it/s] 9%|▉ | 4543/48008 [15:27<2:26:45, 4.94it/s] 9%|▉ | 4544/48008 [15:27<2:24:14, 5.02it/s] 9%|▉ | 4545/48008 [15:27<2:21:57, 5.10it/s] 9%|▉ | 4546/48008 [15:27<2:23:18, 5.05it/s] 9%|▉ | 4547/48008 [15:27<2:23:20, 5.05it/s] 9%|▉ | 4548/48008 [15:28<2:21:58, 5.10it/s] 9%|▉ | 4549/48008 [15:28<2:22:05, 5.10it/s] 9%|▉ | 4550/48008 [15:28<2:20:29, 5.16it/s] {'loss': 0.304, 'grad_norm': 0.458984375, 'learning_rate': 0.00018104899183469424, 'epoch': 0.09} 9%|▉ | 4550/48008 [15:28<2:20:29, 5.16it/s] 9%|▉ | 4551/48008 [15:28<2:22:17, 5.09it/s] 9%|▉ | 4552/48008 [15:28<2:25:08, 4.99it/s] 9%|▉ | 4553/48008 [15:29<2:22:57, 5.07it/s] 9%|▉ | 4554/48008 [15:29<2:25:50, 4.97it/s] 9%|▉ | 4555/48008 [15:29<2:24:06, 5.03it/s] 9%|▉ | 4556/48008 [15:29<2:22:38, 5.08it/s] 9%|▉ | 4557/48008 [15:29<2:22:21, 5.09it/s] 9%|▉ | 4558/48008 [15:30<2:19:54, 5.18it/s] 9%|▉ | 4559/48008 [15:30<2:21:46, 5.11it/s] 9%|▉ | 4560/48008 [15:30<2:20:32, 5.15it/s] 10%|▉ | 4561/48008 [15:30<2:19:02, 5.21it/s] 10%|▉ | 4562/48008 [15:31<3:01:05, 4.00it/s] 10%|▉ | 4563/48008 [15:31<2:47:51, 4.31it/s] 10%|▉ | 4564/48008 [15:31<2:39:02, 4.55it/s] 10%|▉ | 4565/48008 [15:31<2:34:43, 4.68it/s] 10%|▉ | 4566/48008 [15:31<2:29:58, 4.83it/s] 10%|▉ | 4567/48008 [15:32<2:26:26, 4.94it/s] 10%|▉ | 4568/48008 [15:32<2:26:14, 4.95it/s] 10%|▉ | 4569/48008 [15:32<2:48:43, 4.29it/s] 10%|▉ | 4570/48008 [15:32<2:39:22, 4.54it/s] 10%|▉ | 4571/48008 [15:32<2:35:26, 4.66it/s] 10%|▉ | 4572/48008 [15:33<2:30:06, 4.82it/s] 10%|▉ | 4573/48008 [15:33<2:25:59, 4.96it/s] 10%|▉ | 4574/48008 [15:33<2:22:34, 5.08it/s] 10%|▉ | 4575/48008 [15:33<2:21:05, 5.13it/s] 10%|▉ | 4576/48008 [15:33<2:21:37, 5.11it/s] 10%|▉ | 4577/48008 [15:34<2:20:19, 5.16it/s] 10%|▉ | 4578/48008 [15:34<2:19:36, 5.18it/s] 10%|▉ | 4579/48008 [15:34<2:19:31, 5.19it/s] 10%|▉ | 4580/48008 [15:34<2:19:16, 5.20it/s] 10%|▉ | 4581/48008 [15:34<2:18:23, 5.23it/s] 10%|▉ | 4582/48008 [15:35<2:28:45, 4.87it/s] 10%|▉ | 4583/48008 [15:35<2:27:16, 4.91it/s] 10%|▉ | 4584/48008 [15:35<2:24:58, 4.99it/s] 10%|▉ | 4585/48008 [15:35<2:23:31, 5.04it/s] 10%|▉ | 4586/48008 [15:35<2:22:21, 5.08it/s] 10%|▉ | 4587/48008 [15:36<4:50:20, 2.49it/s] 10%|▉ | 4588/48008 [15:36<4:07:53, 2.92it/s] 10%|▉ | 4589/48008 [15:37<3:34:30, 3.37it/s] 10%|▉ | 4590/48008 [15:37<3:12:23, 3.76it/s] 10%|▉ | 4591/48008 [15:37<2:56:53, 4.09it/s] 10%|▉ | 4592/48008 [15:37<2:48:05, 4.30it/s] 10%|▉ | 4593/48008 [15:37<2:38:32, 4.56it/s] 10%|▉ | 4594/48008 [15:38<2:32:04, 4.76it/s] 10%|▉ | 4595/48008 [15:38<2:29:19, 4.85it/s] 10%|▉ | 4596/48008 [15:38<2:27:03, 4.92it/s] 10%|▉ | 4597/48008 [15:38<2:34:21, 4.69it/s] 10%|▉ | 4598/48008 [15:38<2:29:53, 4.83it/s] 10%|▉ | 4599/48008 [15:39<2:24:53, 4.99it/s] 10%|▉ | 4600/48008 [15:39<3:05:15, 3.91it/s] {'loss': 0.3027, 'grad_norm': 0.90234375, 'learning_rate': 0.00018084069321779703, 'epoch': 0.1} 10%|▉ | 4600/48008 [15:39<3:05:15, 3.91it/s] 10%|▉ | 4601/48008 [15:39<2:54:27, 4.15it/s] 10%|▉ | 4602/48008 [15:39<2:46:45, 4.34it/s] 10%|▉ | 4603/48008 [15:40<2:49:38, 4.26it/s] 10%|▉ | 4604/48008 [15:40<2:40:24, 4.51it/s] 10%|▉ | 4605/48008 [15:40<2:34:09, 4.69it/s] 10%|▉ | 4606/48008 [15:40<2:28:24, 4.87it/s] 10%|▉ | 4607/48008 [15:40<2:25:34, 4.97it/s] 10%|▉ | 4608/48008 [15:41<2:23:22, 5.05it/s] 10%|▉ | 4609/48008 [15:41<2:21:51, 5.10it/s] 10%|▉ | 4610/48008 [15:41<2:21:40, 5.11it/s] 10%|▉ | 4611/48008 [15:41<2:21:00, 5.13it/s] 10%|▉ | 4612/48008 [15:41<2:20:35, 5.14it/s] 10%|▉ | 4613/48008 [15:42<2:20:37, 5.14it/s] 10%|▉ | 4614/48008 [15:42<2:19:57, 5.17it/s] 10%|▉ | 4615/48008 [15:42<2:44:35, 4.39it/s] 10%|▉ | 4616/48008 [15:42<3:18:54, 3.64it/s] 10%|▉ | 4617/48008 [15:43<3:00:57, 4.00it/s] 10%|▉ | 4618/48008 [15:43<2:46:42, 4.34it/s] 10%|▉ | 4619/48008 [15:43<2:39:14, 4.54it/s] 10%|▉ | 4620/48008 [15:43<2:32:29, 4.74it/s] 10%|▉ | 4621/48008 [15:43<2:27:06, 4.92it/s] 10%|▉ | 4622/48008 [15:44<2:23:16, 5.05it/s] 10%|▉ | 4623/48008 [15:44<2:21:41, 5.10it/s] 10%|▉ | 4624/48008 [15:44<2:19:53, 5.17it/s] 10%|▉ | 4625/48008 [15:44<2:21:30, 5.11it/s] 10%|▉ | 4626/48008 [15:44<2:23:03, 5.05it/s] 10%|▉ | 4627/48008 [15:45<2:25:15, 4.98it/s] 10%|▉ | 4628/48008 [15:45<2:22:46, 5.06it/s] 10%|▉ | 4629/48008 [15:45<2:21:21, 5.11it/s] 10%|▉ | 4630/48008 [15:45<2:20:05, 5.16it/s] 10%|▉ | 4631/48008 [15:45<2:19:26, 5.18it/s] 10%|▉ | 4632/48008 [15:46<2:18:35, 5.22it/s] 10%|▉ | 4633/48008 [15:46<2:28:16, 4.88it/s] 10%|▉ | 4634/48008 [15:46<2:25:01, 4.98it/s] 10%|▉ | 4635/48008 [15:46<2:25:34, 4.97it/s] 10%|▉ | 4636/48008 [15:46<2:26:56, 4.92it/s] 10%|▉ | 4637/48008 [15:47<2:25:28, 4.97it/s] 10%|▉ | 4638/48008 [15:47<2:24:07, 5.02it/s] 10%|▉ | 4639/48008 [15:47<2:23:37, 5.03it/s] 10%|▉ | 4640/48008 [15:47<2:20:57, 5.13it/s] 10%|▉ | 4641/48008 [15:47<2:19:51, 5.17it/s] 10%|▉ | 4642/48008 [15:48<2:19:31, 5.18it/s] 10%|▉ | 4643/48008 [15:48<2:19:07, 5.19it/s] 10%|▉ | 4644/48008 [15:48<2:19:03, 5.20it/s] 10%|▉ | 4645/48008 [15:48<2:20:20, 5.15it/s] 10%|▉ | 4646/48008 [15:48<2:21:14, 5.12it/s] 10%|▉ | 4647/48008 [15:49<2:21:26, 5.11it/s] 10%|▉ | 4648/48008 [15:49<2:20:22, 5.15it/s] 10%|▉ | 4649/48008 [15:49<2:20:56, 5.13it/s] 10%|▉ | 4650/48008 [15:49<2:19:49, 5.17it/s] {'loss': 0.3028, 'grad_norm': 0.205078125, 'learning_rate': 0.00018063239460089988, 'epoch': 0.1} 10%|▉ | 4650/48008 [15:49<2:19:49, 5.17it/s] 10%|▉ | 4651/48008 [15:49<2:19:11, 5.19it/s] 10%|▉ | 4652/48008 [15:49<2:19:15, 5.19it/s] 10%|▉ | 4653/48008 [15:50<3:01:18, 3.99it/s] 10%|▉ | 4654/48008 [15:50<2:47:48, 4.31it/s] 10%|▉ | 4655/48008 [15:50<2:38:43, 4.55it/s] 10%|▉ | 4656/48008 [15:50<2:34:37, 4.67it/s] 10%|▉ | 4657/48008 [15:51<2:29:03, 4.85it/s] 10%|▉ | 4658/48008 [15:51<2:25:29, 4.97it/s] 10%|▉ | 4659/48008 [15:51<2:22:54, 5.06it/s] 10%|▉ | 4660/48008 [15:51<2:31:14, 4.78it/s] 10%|▉ | 4661/48008 [15:51<2:26:58, 4.92it/s] 10%|▉ | 4662/48008 [15:52<2:24:02, 5.02it/s] 10%|▉ | 4663/48008 [15:52<2:24:52, 4.99it/s] 10%|▉ | 4664/48008 [15:52<2:22:20, 5.08it/s] 10%|▉ | 4665/48008 [15:52<2:21:57, 5.09it/s] 10%|▉ | 4666/48008 [15:52<2:24:36, 5.00it/s] 10%|▉ | 4667/48008 [15:53<2:22:57, 5.05it/s] 10%|▉ | 4668/48008 [15:53<2:21:14, 5.11it/s] 10%|▉ | 4669/48008 [15:53<2:21:22, 5.11it/s] 10%|▉ | 4670/48008 [15:53<2:19:34, 5.17it/s] 10%|▉ | 4671/48008 [15:53<2:19:36, 5.17it/s] 10%|▉ | 4672/48008 [15:54<2:19:05, 5.19it/s] 10%|▉ | 4673/48008 [15:54<2:17:45, 5.24it/s] 10%|▉ | 4674/48008 [15:54<2:16:36, 5.29it/s] 10%|▉ | 4675/48008 [15:54<2:59:11, 4.03it/s] 10%|▉ | 4676/48008 [15:55<2:49:09, 4.27it/s] 10%|▉ | 4677/48008 [15:55<3:04:31, 3.91it/s] 10%|▉ | 4678/48008 [15:55<2:50:11, 4.24it/s] 10%|▉ | 4679/48008 [15:55<2:43:08, 4.43it/s] 10%|▉ | 4680/48008 [15:55<2:45:34, 4.36it/s] 10%|▉ | 4681/48008 [15:56<2:37:17, 4.59it/s] 10%|▉ | 4682/48008 [15:56<2:31:31, 4.77it/s] 10%|▉ | 4683/48008 [15:56<2:28:58, 4.85it/s] 10%|▉ | 4684/48008 [15:56<2:27:27, 4.90it/s] 10%|▉ | 4685/48008 [15:56<2:23:33, 5.03it/s] 10%|▉ | 4686/48008 [15:57<2:22:57, 5.05it/s] 10%|▉ | 4687/48008 [15:57<2:46:22, 4.34it/s] 10%|▉ | 4688/48008 [15:57<2:37:35, 4.58it/s] 10%|▉ | 4689/48008 [15:57<2:31:39, 4.76it/s] 10%|▉ | 4690/48008 [15:58<2:27:12, 4.90it/s] 10%|▉ | 4691/48008 [15:58<2:34:12, 4.68it/s] 10%|▉ | 4692/48008 [15:58<2:29:20, 4.83it/s] 10%|▉ | 4693/48008 [15:58<2:25:46, 4.95it/s] 10%|▉ | 4694/48008 [15:58<2:23:36, 5.03it/s] 10%|▉ | 4695/48008 [15:59<2:20:19, 5.14it/s] 10%|▉ | 4696/48008 [15:59<2:44:30, 4.39it/s] 10%|▉ | 4697/48008 [15:59<2:34:07, 4.68it/s] 10%|▉ | 4698/48008 [15:59<2:32:03, 4.75it/s] 10%|▉ | 4699/48008 [15:59<2:27:53, 4.88it/s] 10%|▉ | 4700/48008 [16:00<2:24:40, 4.99it/s] {'loss': 0.3021, 'grad_norm': 0.138671875, 'learning_rate': 0.00018042409598400267, 'epoch': 0.1} 10%|▉ | 4700/48008 [16:00<2:24:40, 4.99it/s] 10%|▉ | 4701/48008 [16:00<2:22:45, 5.06it/s] 10%|▉ | 4702/48008 [16:00<2:21:29, 5.10it/s] 10%|▉ | 4703/48008 [16:00<3:02:44, 3.95it/s] 10%|▉ | 4704/48008 [16:01<2:48:32, 4.28it/s] 10%|▉ | 4705/48008 [16:01<2:40:46, 4.49it/s] 10%|▉ | 4706/48008 [16:01<3:15:41, 3.69it/s] 10%|▉ | 4707/48008 [16:01<2:57:37, 4.06it/s] 10%|▉ | 4708/48008 [16:02<2:47:44, 4.30it/s] 10%|▉ | 4709/48008 [16:02<2:41:35, 4.47it/s] 10%|▉ | 4710/48008 [16:02<2:36:20, 4.62it/s] 10%|▉ | 4711/48008 [16:02<2:30:45, 4.79it/s] 10%|▉ | 4712/48008 [16:02<2:29:38, 4.82it/s] 10%|▉ | 4713/48008 [16:03<2:27:07, 4.90it/s] 10%|▉ | 4714/48008 [16:03<2:24:22, 5.00it/s] 10%|▉ | 4715/48008 [16:03<2:23:46, 5.02it/s] 10%|▉ | 4716/48008 [16:03<2:21:37, 5.09it/s] 10%|▉ | 4717/48008 [16:03<2:20:14, 5.14it/s] 10%|▉ | 4718/48008 [16:03<2:19:45, 5.16it/s] 10%|▉ | 4719/48008 [16:04<3:01:11, 3.98it/s] 10%|▉ | 4720/48008 [16:04<2:47:19, 4.31it/s] 10%|▉ | 4721/48008 [16:04<2:37:23, 4.58it/s] 10%|▉ | 4722/48008 [16:04<2:31:14, 4.77it/s] 10%|▉ | 4723/48008 [16:05<2:27:18, 4.90it/s] 10%|▉ | 4724/48008 [16:05<2:34:17, 4.68it/s] 10%|▉ | 4725/48008 [16:05<2:28:58, 4.84it/s] 10%|▉ | 4726/48008 [16:05<2:25:45, 4.95it/s] 10%|▉ | 4727/48008 [16:05<2:33:10, 4.71it/s] 10%|▉ | 4728/48008 [16:06<2:27:34, 4.89it/s] 10%|▉ | 4729/48008 [16:06<2:26:56, 4.91it/s] 10%|▉ | 4730/48008 [16:06<2:24:23, 5.00it/s] 10%|▉ | 4731/48008 [16:06<2:23:25, 5.03it/s] 10%|▉ | 4732/48008 [16:06<2:22:53, 5.05it/s] 10%|▉ | 4733/48008 [16:07<2:30:56, 4.78it/s] 10%|▉ | 4734/48008 [16:07<2:30:54, 4.78it/s] 10%|▉ | 4735/48008 [16:07<2:26:01, 4.94it/s] 10%|▉ | 4736/48008 [16:07<2:33:04, 4.71it/s] 10%|▉ | 4737/48008 [16:07<2:28:00, 4.87it/s] 10%|▉ | 4738/48008 [16:08<2:25:32, 4.96it/s] 10%|▉ | 4739/48008 [16:08<2:22:34, 5.06it/s] 10%|▉ | 4740/48008 [16:08<2:20:17, 5.14it/s] 10%|▉ | 4741/48008 [16:08<2:19:21, 5.17it/s] 10%|▉ | 4742/48008 [16:08<2:19:15, 5.18it/s] 10%|▉ | 4743/48008 [16:09<2:19:01, 5.19it/s] 10%|▉ | 4744/48008 [16:09<2:17:23, 5.25it/s] 10%|▉ | 4745/48008 [16:09<2:17:28, 5.24it/s] 10%|▉ | 4746/48008 [16:09<2:17:15, 5.25it/s] 10%|▉ | 4747/48008 [16:09<2:17:06, 5.26it/s] 10%|▉ | 4748/48008 [16:10<2:16:59, 5.26it/s] 10%|▉ | 4749/48008 [16:10<2:18:20, 5.21it/s] 10%|▉ | 4750/48008 [16:10<2:19:42, 5.16it/s] {'loss': 0.3026, 'grad_norm': 0.390625, 'learning_rate': 0.0001802157973671055, 'epoch': 0.1} 10%|▉ | 4750/48008 [16:10<2:19:42, 5.16it/s] 10%|▉ | 4751/48008 [16:10<2:21:49, 5.08it/s] 10%|▉ | 4752/48008 [16:10<2:22:07, 5.07it/s] 10%|▉ | 4753/48008 [16:11<2:20:45, 5.12it/s] 10%|▉ | 4754/48008 [16:11<2:19:40, 5.16it/s] 10%|▉ | 4755/48008 [16:11<2:20:16, 5.14it/s] 10%|▉ | 4756/48008 [16:11<2:20:47, 5.12it/s] 10%|▉ | 4757/48008 [16:11<2:22:12, 5.07it/s] 10%|▉ | 4758/48008 [16:12<2:21:00, 5.11it/s] 10%|▉ | 4759/48008 [16:12<2:19:10, 5.18it/s] 10%|▉ | 4760/48008 [16:12<2:18:35, 5.20it/s] 10%|▉ | 4761/48008 [16:12<2:18:09, 5.22it/s] 10%|▉ | 4762/48008 [16:12<2:16:19, 5.29it/s] 10%|▉ | 4763/48008 [16:12<2:16:25, 5.28it/s] 10%|▉ | 4764/48008 [16:13<2:18:19, 5.21it/s] 10%|▉ | 4765/48008 [16:13<2:18:24, 5.21it/s] 10%|▉ | 4766/48008 [16:13<2:17:21, 5.25it/s] 10%|▉ | 4767/48008 [16:13<2:17:14, 5.25it/s] 10%|▉ | 4768/48008 [16:13<2:16:32, 5.28it/s] 10%|▉ | 4769/48008 [16:14<2:17:12, 5.25it/s] 10%|▉ | 4770/48008 [16:14<2:17:25, 5.24it/s] 10%|▉ | 4771/48008 [16:14<2:20:12, 5.14it/s] 10%|▉ | 4772/48008 [16:14<2:20:05, 5.14it/s] 10%|▉ | 4773/48008 [16:14<2:19:59, 5.15it/s] 10%|▉ | 4774/48008 [16:15<2:21:16, 5.10it/s] 10%|▉ | 4775/48008 [16:15<2:20:09, 5.14it/s] 10%|▉ | 4776/48008 [16:15<2:19:13, 5.18it/s] 10%|▉ | 4777/48008 [16:15<2:18:44, 5.19it/s] 10%|▉ | 4778/48008 [16:15<2:17:47, 5.23it/s] 10%|▉ | 4779/48008 [16:16<2:17:18, 5.25it/s] 10%|▉ | 4780/48008 [16:16<2:16:52, 5.26it/s] 10%|▉ | 4781/48008 [16:16<2:14:50, 5.34it/s] 10%|▉ | 4782/48008 [16:16<2:15:09, 5.33it/s] 10%|▉ | 4783/48008 [16:16<2:15:17, 5.32it/s] 10%|▉ | 4784/48008 [16:17<2:17:25, 5.24it/s] 10%|▉ | 4785/48008 [16:17<2:16:56, 5.26it/s] 10%|▉ | 4786/48008 [16:17<2:19:28, 5.16it/s] 10%|▉ | 4787/48008 [16:17<2:17:40, 5.23it/s] 10%|▉ | 4788/48008 [16:17<2:17:34, 5.24it/s] 10%|▉ | 4789/48008 [16:17<2:16:45, 5.27it/s] 10%|▉ | 4790/48008 [16:18<2:17:27, 5.24it/s] 10%|▉ | 4791/48008 [16:18<2:17:07, 5.25it/s] 10%|▉ | 4792/48008 [16:18<2:17:06, 5.25it/s] 10%|▉ | 4793/48008 [16:18<2:15:48, 5.30it/s] 10%|▉ | 4794/48008 [16:18<2:18:35, 5.20it/s] 10%|▉ | 4795/48008 [16:19<2:27:49, 4.87it/s] 10%|▉ | 4796/48008 [16:19<2:49:37, 4.25it/s] 10%|▉ | 4797/48008 [16:19<2:42:41, 4.43it/s] 10%|▉ | 4798/48008 [16:19<2:36:57, 4.59it/s] 10%|▉ | 4799/48008 [16:20<2:31:07, 4.77it/s] 10%|▉ | 4800/48008 [16:20<2:27:17, 4.89it/s] {'loss': 0.3039, 'grad_norm': 0.70703125, 'learning_rate': 0.0001800074987502083, 'epoch': 0.1} 10%|▉ | 4800/48008 [16:20<2:27:17, 4.89it/s] 10%|█ | 4801/48008 [16:20<2:23:04, 5.03it/s] 10%|█ | 4802/48008 [16:20<2:23:21, 5.02it/s] 10%|█ | 4803/48008 [16:20<2:23:59, 5.00it/s] 10%|█ | 4804/48008 [16:21<2:21:28, 5.09it/s] 10%|█ | 4805/48008 [16:21<2:20:32, 5.12it/s] 10%|█ | 4806/48008 [16:21<2:22:03, 5.07it/s] 10%|█ | 4807/48008 [16:21<2:20:38, 5.12it/s] 10%|█ | 4808/48008 [16:21<2:19:29, 5.16it/s] 10%|█ | 4809/48008 [16:21<2:21:20, 5.09it/s] 10%|█ | 4810/48008 [16:22<2:19:41, 5.15it/s] 10%|█ | 4811/48008 [16:22<2:20:10, 5.14it/s] 10%|█ | 4812/48008 [16:22<2:19:00, 5.18it/s] 10%|█ | 4813/48008 [16:22<2:18:24, 5.20it/s] 10%|█ | 4814/48008 [16:22<2:19:10, 5.17it/s] 10%|█ | 4815/48008 [16:23<2:17:47, 5.22it/s] 10%|█ | 4816/48008 [16:23<2:17:20, 5.24it/s] 10%|█ | 4817/48008 [16:23<2:17:06, 5.25it/s] 10%|█ | 4818/48008 [16:23<2:16:31, 5.27it/s] 10%|█ | 4819/48008 [16:23<2:15:49, 5.30it/s] 10%|█ | 4820/48008 [16:24<2:16:32, 5.27it/s] 10%|█ | 4821/48008 [16:24<2:16:04, 5.29it/s] 10%|█ | 4822/48008 [16:24<2:18:37, 5.19it/s] 10%|█ | 4823/48008 [16:24<2:18:24, 5.20it/s] 10%|█ | 4824/48008 [16:24<2:20:27, 5.12it/s] 10%|█ | 4825/48008 [16:25<2:20:01, 5.14it/s] 10%|█ | 4826/48008 [16:25<2:18:54, 5.18it/s] 10%|█ | 4827/48008 [16:25<2:18:27, 5.20it/s] 10%|█ | 4828/48008 [16:25<2:18:17, 5.20it/s] 10%|█ | 4829/48008 [16:25<2:16:35, 5.27it/s] 10%|█ | 4830/48008 [16:26<2:17:46, 5.22it/s] 10%|█ | 4831/48008 [16:26<2:17:41, 5.23it/s] 10%|█ | 4832/48008 [16:26<2:17:36, 5.23it/s] 10%|█ | 4833/48008 [16:26<2:21:10, 5.10it/s] 10%|█ | 4834/48008 [16:26<2:19:46, 5.15it/s] 10%|█ | 4835/48008 [16:26<2:19:05, 5.17it/s] 10%|█ | 4836/48008 [16:27<2:20:53, 5.11it/s] 10%|█ | 4837/48008 [16:27<2:19:25, 5.16it/s] 10%|█ | 4838/48008 [16:27<2:19:03, 5.17it/s] 10%|█ | 4839/48008 [16:27<2:17:36, 5.23it/s] 10%|█ | 4840/48008 [16:27<2:17:23, 5.24it/s] 10%|█ | 4841/48008 [16:28<2:42:18, 4.43it/s] 10%|█ | 4842/48008 [16:28<2:34:54, 4.64it/s] 10%|█ | 4843/48008 [16:28<2:28:57, 4.83it/s] 10%|█ | 4844/48008 [16:28<2:25:09, 4.96it/s] 10%|█ | 4845/48008 [16:29<2:22:13, 5.06it/s] 10%|█ | 4846/48008 [16:29<2:20:05, 5.13it/s] 10%|█ | 4847/48008 [16:29<2:19:14, 5.17it/s] 10%|█ | 4848/48008 [16:29<2:20:02, 5.14it/s] 10%|█ | 4849/48008 [16:29<2:21:40, 5.08it/s] 10%|█ | 4850/48008 [16:29<2:19:14, 5.17it/s] {'loss': 0.3072, 'grad_norm': 0.3828125, 'learning_rate': 0.00017979920013331112, 'epoch': 0.1} 10%|█ | 4850/48008 [16:29<2:19:14, 5.17it/s] 10%|█ | 4851/48008 [16:30<2:18:37, 5.19it/s] 10%|█ | 4852/48008 [16:30<2:42:40, 4.42it/s] 10%|█ | 4853/48008 [16:30<2:36:03, 4.61it/s] 10%|█ | 4854/48008 [16:30<2:30:01, 4.79it/s] 10%|█ | 4855/48008 [16:31<2:26:02, 4.92it/s] 10%|█ | 4856/48008 [16:31<2:28:07, 4.86it/s] 10%|█ | 4857/48008 [16:31<2:27:20, 4.88it/s] 10%|█ | 4858/48008 [16:31<2:24:21, 4.98it/s] 10%|█ | 4859/48008 [16:31<2:22:17, 5.05it/s] 10%|█ | 4860/48008 [16:32<2:20:06, 5.13it/s] 10%|█ | 4861/48008 [16:32<2:23:24, 5.01it/s] 10%|█ | 4862/48008 [16:32<2:24:11, 4.99it/s] 10%|█ | 4863/48008 [16:32<2:22:25, 5.05it/s] 10%|█ | 4864/48008 [16:32<2:20:43, 5.11it/s] 10%|█ | 4865/48008 [16:33<2:18:58, 5.17it/s] 10%|█ | 4866/48008 [16:33<2:43:12, 4.41it/s] 10%|█ | 4867/48008 [16:33<2:37:13, 4.57it/s] 10%|█ | 4868/48008 [16:33<2:30:02, 4.79it/s] 10%|█ | 4869/48008 [16:33<2:26:21, 4.91it/s] 10%|█ | 4870/48008 [16:34<2:23:14, 5.02it/s] 10%|█ | 4871/48008 [16:34<2:46:28, 4.32it/s] 10%|█ | 4872/48008 [16:34<2:37:16, 4.57it/s] 10%|█ | 4873/48008 [16:34<2:30:41, 4.77it/s] 10%|█ | 4874/48008 [16:34<2:25:50, 4.93it/s] 10%|█ | 4875/48008 [16:35<2:22:39, 5.04it/s] 10%|█ | 4876/48008 [16:35<2:19:58, 5.14it/s] 10%|█ | 4877/48008 [16:35<2:19:39, 5.15it/s] 10%|█ | 4878/48008 [16:35<2:19:29, 5.15it/s] 10%|█ | 4879/48008 [16:35<2:28:42, 4.83it/s] 10%|█ | 4880/48008 [16:36<2:26:00, 4.92it/s] 10%|█ | 4881/48008 [16:36<2:25:35, 4.94it/s] 10%|█ | 4882/48008 [16:36<2:22:30, 5.04it/s] 10%|█ | 4883/48008 [16:36<2:19:25, 5.16it/s] 10%|█ | 4884/48008 [16:36<2:20:21, 5.12it/s] 10%|█ | 4885/48008 [16:37<2:20:07, 5.13it/s] 10%|█ | 4886/48008 [16:37<2:18:12, 5.20it/s] 10%|█ | 4887/48008 [16:37<2:18:41, 5.18it/s] 10%|█ | 4888/48008 [16:37<2:17:58, 5.21it/s] 10%|█ | 4889/48008 [16:37<2:16:50, 5.25it/s] 10%|█ | 4890/48008 [16:38<2:16:38, 5.26it/s] 10%|█ | 4891/48008 [16:38<2:16:53, 5.25it/s] 10%|█ | 4892/48008 [16:38<2:58:48, 4.02it/s] 10%|█ | 4893/48008 [16:38<2:46:29, 4.32it/s] 10%|█ | 4894/48008 [16:39<2:40:17, 4.48it/s] 10%|█ | 4895/48008 [16:39<2:33:31, 4.68it/s] 10%|█ | 4896/48008 [16:39<2:28:10, 4.85it/s] 10%|█ | 4897/48008 [16:39<2:23:58, 4.99it/s] 10%|█ | 4898/48008 [16:39<2:21:40, 5.07it/s] 10%|█ | 4899/48008 [16:39<2:19:55, 5.13it/s] 10%|█ | 4900/48008 [16:40<2:21:30, 5.08it/s] {'loss': 0.3014, 'grad_norm': 1.15625, 'learning_rate': 0.00017959090151641394, 'epoch': 0.1} 10%|█ | 4900/48008 [16:40<2:21:30, 5.08it/s] 10%|█ | 4901/48008 [16:40<2:20:19, 5.12it/s] 10%|█ | 4902/48008 [16:40<2:20:08, 5.13it/s] 10%|█ | 4903/48008 [16:40<2:19:13, 5.16it/s] 10%|█ | 4904/48008 [16:40<2:21:22, 5.08it/s] 10%|█ | 4905/48008 [16:41<2:19:50, 5.14it/s] 10%|█ | 4906/48008 [16:41<2:19:07, 5.16it/s] 10%|█ | 4907/48008 [16:41<2:17:21, 5.23it/s] 10%|█ | 4908/48008 [16:41<2:16:55, 5.25it/s] 10%|█ | 4909/48008 [16:41<2:15:26, 5.30it/s] 10%|█ | 4910/48008 [16:42<2:25:45, 4.93it/s] 10%|█ | 4911/48008 [16:42<2:27:02, 4.88it/s] 10%|█ | 4912/48008 [16:42<3:06:02, 3.86it/s] 10%|█ | 4913/48008 [16:42<2:53:34, 4.14it/s] 10%|█ | 4914/48008 [16:43<2:42:34, 4.42it/s] 10%|█ | 4915/48008 [16:43<2:36:22, 4.59it/s] 10%|█ | 4916/48008 [16:43<2:30:13, 4.78it/s] 10%|█ | 4917/48008 [16:43<2:26:27, 4.90it/s] 10%|█ | 4918/48008 [16:43<2:23:16, 5.01it/s] 10%|█ | 4919/48008 [16:44<2:22:06, 5.05it/s] 10%|█ | 4920/48008 [16:44<2:21:32, 5.07it/s] 10%|█ | 4921/48008 [16:44<2:20:05, 5.13it/s] 10%|█ | 4922/48008 [16:44<2:21:45, 5.07it/s] 10%|█ | 4923/48008 [16:44<2:22:42, 5.03it/s] 10%|█ | 4924/48008 [16:45<2:20:52, 5.10it/s] 10%|█ | 4925/48008 [16:45<2:18:57, 5.17it/s] 10%|█ | 4926/48008 [16:45<2:18:01, 5.20it/s] 10%|█ | 4927/48008 [16:45<2:19:20, 5.15it/s] 10%|█ | 4928/48008 [16:45<2:20:05, 5.13it/s] 10%|█ | 4929/48008 [16:46<2:21:48, 5.06it/s] 10%|█ | 4930/48008 [16:46<2:22:40, 5.03it/s] 10%|█ | 4931/48008 [16:46<2:21:47, 5.06it/s] 10%|█ | 4932/48008 [16:46<2:20:27, 5.11it/s] 10%|█ | 4933/48008 [16:46<2:21:58, 5.06it/s] 10%|█ | 4934/48008 [16:47<2:20:05, 5.12it/s] 10%|█ | 4935/48008 [16:47<2:21:25, 5.08it/s] 10%|█ | 4936/48008 [16:47<2:29:44, 4.79it/s] 10%|█ | 4937/48008 [16:47<2:27:18, 4.87it/s] 10%|█ | 4938/48008 [16:47<2:24:32, 4.97it/s] 10%|█ | 4939/48008 [16:48<2:24:55, 4.95it/s] 10%|█ | 4940/48008 [16:48<2:22:34, 5.03it/s] 10%|█ | 4941/48008 [16:48<2:21:03, 5.09it/s] 10%|█ | 4942/48008 [16:48<2:19:45, 5.14it/s] 10%|█ | 4943/48008 [16:48<2:18:38, 5.18it/s] 10%|█ | 4944/48008 [16:49<2:18:10, 5.19it/s] 10%|█ | 4945/48008 [16:49<2:19:07, 5.16it/s] 10%|█ | 4946/48008 [16:49<2:17:07, 5.23it/s] 10%|█ | 4947/48008 [16:49<2:17:10, 5.23it/s] 10%|█ | 4948/48008 [16:49<2:16:40, 5.25it/s] 10%|█ | 4949/48008 [16:49<2:16:57, 5.24it/s] 10%|█ | 4950/48008 [16:50<2:42:00, 4.43it/s] {'loss': 0.3033, 'grad_norm': 0.74609375, 'learning_rate': 0.00017938260289951675, 'epoch': 0.1} 10%|█ | 4950/48008 [16:50<2:42:00, 4.43it/s] 10%|█ | 4951/48008 [16:50<2:36:01, 4.60it/s] 10%|█ | 4952/48008 [16:50<2:40:23, 4.47it/s] 10%|█ | 4953/48008 [16:50<2:37:19, 4.56it/s] 10%|█ | 4954/48008 [16:51<2:33:56, 4.66it/s] 10%|█ | 4955/48008 [16:51<2:27:26, 4.87it/s] 10%|█ | 4956/48008 [16:51<2:23:04, 5.02it/s] 10%|█ | 4957/48008 [16:51<2:31:23, 4.74it/s] 10%|█ | 4958/48008 [16:51<2:37:00, 4.57it/s] 10%|█ | 4959/48008 [16:52<2:32:36, 4.70it/s] 10%|█ | 4960/48008 [16:52<2:28:08, 4.84it/s] 10%|█ | 4961/48008 [16:52<2:24:16, 4.97it/s] 10%|█ | 4962/48008 [16:52<2:21:53, 5.06it/s] 10%|█ | 4963/48008 [16:52<2:20:39, 5.10it/s] 10%|█ | 4964/48008 [16:53<2:21:49, 5.06it/s] 10%|█ | 4965/48008 [16:53<2:19:58, 5.13it/s] 10%|█ | 4966/48008 [16:53<2:21:20, 5.08it/s] 10%|█ | 4967/48008 [16:53<2:20:04, 5.12it/s] 10%|█ | 4968/48008 [16:53<2:18:39, 5.17it/s] 10%|█ | 4969/48008 [16:54<2:17:20, 5.22it/s] 10%|█ | 4970/48008 [16:54<2:15:55, 5.28it/s] 10%|█ | 4971/48008 [16:54<2:16:03, 5.27it/s] 10%|█ | 4972/48008 [16:54<2:18:31, 5.18it/s] 10%|█ | 4973/48008 [16:54<2:20:18, 5.11it/s] 10%|█ | 4974/48008 [16:55<2:21:45, 5.06it/s] 10%|█ | 4975/48008 [16:55<3:02:15, 3.94it/s] 10%|█ | 4976/48008 [16:55<2:58:20, 4.02it/s] 10%|█ | 4977/48008 [16:55<2:46:05, 4.32it/s] 10%|█ | 4978/48008 [16:56<2:37:22, 4.56it/s] 10%|█ | 4979/48008 [16:56<2:29:00, 4.81it/s] 10%|█ | 4980/48008 [16:56<2:24:49, 4.95it/s] 10%|█ | 4981/48008 [16:56<2:24:37, 4.96it/s] 10%|█ | 4982/48008 [16:56<2:22:09, 5.04it/s] 10%|█ | 4983/48008 [16:57<2:21:38, 5.06it/s] 10%|█ | 4984/48008 [16:57<2:19:14, 5.15it/s] 10%|█ | 4985/48008 [16:57<2:18:19, 5.18it/s] 10%|█ | 4986/48008 [16:57<2:20:13, 5.11it/s] 10%|█ | 4987/48008 [16:57<2:20:13, 5.11it/s] 10%|█ | 4988/48008 [16:57<2:21:15, 5.08it/s] 10%|█ | 4989/48008 [16:58<2:22:17, 5.04it/s] 10%|█ | 4990/48008 [16:58<2:20:44, 5.09it/s] 10%|█ | 4991/48008 [16:58<2:29:23, 4.80it/s] 10%|█ | 4992/48008 [16:58<2:25:05, 4.94it/s] 10%|█ | 4993/48008 [16:59<2:24:42, 4.95it/s] 10%|█ | 4994/48008 [16:59<2:32:06, 4.71it/s] 10%|█ | 4995/48008 [16:59<2:27:48, 4.85it/s] 10%|█ | 4996/48008 [16:59<2:25:42, 4.92it/s] 10%|█ | 4997/48008 [16:59<2:22:55, 5.02it/s] 10%|█ | 4998/48008 [17:00<2:30:53, 4.75it/s] 10%|█ | 4999/48008 [17:00<2:26:23, 4.90it/s] 10%|█ | 5000/48008 [17:00<2:25:00, 4.94it/s] {'loss': 0.3018, 'grad_norm': 0.302734375, 'learning_rate': 0.00017917430428261957, 'epoch': 0.1} 10%|█ | 5000/48008 [17:00<2:25:00, 4.94it/s] 10%|█ | 5001/48008 [17:00<2:24:16, 4.97it/s] 10%|█ | 5002/48008 [17:00<2:20:50, 5.09it/s] 10%|█ | 5003/48008 [17:01<2:21:38, 5.06it/s] 10%|█ | 5004/48008 [17:01<2:21:41, 5.06it/s] 10%|█ | 5005/48008 [17:01<2:19:32, 5.14it/s] 10%|█ | 5006/48008 [17:01<2:21:16, 5.07it/s] 10%|█ | 5007/48008 [17:01<2:20:14, 5.11it/s] 10%|█ | 5008/48008 [17:01<2:17:31, 5.21it/s] 10%|█ | 5009/48008 [17:02<2:17:53, 5.20it/s] 10%|█ | 5010/48008 [17:02<2:16:17, 5.26it/s] 10%|█ | 5011/48008 [17:02<2:15:31, 5.29it/s] 10%|█ | 5012/48008 [17:02<2:15:47, 5.28it/s] 10%|█ | 5013/48008 [17:02<2:15:59, 5.27it/s] 10%|█ | 5014/48008 [17:03<2:15:48, 5.28it/s] 10%|█ | 5015/48008 [17:03<2:15:11, 5.30it/s] 10%|█ | 5016/48008 [17:03<2:17:06, 5.23it/s] 10%|█ | 5017/48008 [17:03<2:17:14, 5.22it/s] 10%|█ | 5018/48008 [17:03<2:19:34, 5.13it/s] 10%|█ | 5019/48008 [17:04<2:17:53, 5.20it/s] 10%|█ | 5020/48008 [17:04<2:16:28, 5.25it/s] 10%|█ | 5021/48008 [17:04<2:15:40, 5.28it/s] 10%|█ | 5022/48008 [17:04<2:15:52, 5.27it/s] 10%|█ | 5023/48008 [17:04<2:15:23, 5.29it/s] 10%|█ | 5024/48008 [17:05<2:26:10, 4.90it/s] 10%|█ | 5025/48008 [17:05<2:23:34, 4.99it/s] 10%|█ | 5026/48008 [17:05<2:20:39, 5.09it/s] 10%|█ | 5027/48008 [17:05<2:19:00, 5.15it/s] 10%|█ | 5028/48008 [17:05<2:18:16, 5.18it/s] 10%|█ | 5029/48008 [17:06<2:17:54, 5.19it/s] 10%|█ | 5030/48008 [17:06<2:27:11, 4.87it/s] 10%|█ | 5031/48008 [17:06<2:23:32, 4.99it/s] 10%|█ | 5032/48008 [17:06<2:23:35, 4.99it/s] 10%|█ | 5033/48008 [17:06<2:20:19, 5.10it/s] 10%|█ | 5034/48008 [17:07<2:19:04, 5.15it/s] 10%|█ | 5035/48008 [17:07<2:18:09, 5.18it/s] 10%|█ | 5036/48008 [17:07<2:17:25, 5.21it/s] 10%|█ | 5037/48008 [17:07<2:16:27, 5.25it/s] 10%|█ | 5038/48008 [17:07<2:17:40, 5.20it/s] 10%|█ | 5039/48008 [17:07<2:17:17, 5.22it/s] 10%|█ | 5040/48008 [17:08<2:16:38, 5.24it/s] 11%|█ | 5041/48008 [17:08<2:16:47, 5.23it/s] 11%|█ | 5042/48008 [17:08<2:16:19, 5.25it/s] 11%|█ | 5043/48008 [17:08<2:15:34, 5.28it/s] 11%|█ | 5044/48008 [17:08<2:15:19, 5.29it/s] 11%|█ | 5045/48008 [17:09<2:15:49, 5.27it/s] 11%|█ | 5046/48008 [17:09<2:16:02, 5.26it/s] 11%|█ | 5047/48008 [17:09<2:15:57, 5.27it/s] 11%|█ | 5048/48008 [17:09<2:18:53, 5.15it/s] 11%|█ | 5049/48008 [17:09<2:18:38, 5.16it/s] 11%|█ | 5050/48008 [17:10<2:17:51, 5.19it/s] {'loss': 0.3059, 'grad_norm': 0.1396484375, 'learning_rate': 0.0001789660056657224, 'epoch': 0.11} 11%|█ | 5050/48008 [17:10<2:17:51, 5.19it/s] 11%|█ | 5051/48008 [17:10<2:18:38, 5.16it/s] 11%|█ | 5052/48008 [17:10<2:17:06, 5.22it/s] 11%|█ | 5053/48008 [17:10<2:16:41, 5.24it/s] 11%|█ | 5054/48008 [17:10<2:19:31, 5.13it/s] 11%|█ | 5055/48008 [17:11<2:18:12, 5.18it/s] 11%|█ | 5056/48008 [17:11<2:17:33, 5.20it/s] 11%|█ | 5057/48008 [17:11<2:18:20, 5.17it/s] 11%|█ | 5058/48008 [17:11<2:17:58, 5.19it/s] 11%|█ | 5059/48008 [17:11<2:20:15, 5.10it/s] 11%|█ | 5060/48008 [17:12<2:18:36, 5.16it/s] 11%|█ | 5061/48008 [17:12<2:17:18, 5.21it/s] 11%|█ | 5062/48008 [17:12<2:16:08, 5.26it/s] 11%|█ | 5063/48008 [17:12<2:15:49, 5.27it/s] 11%|█ | 5064/48008 [17:12<2:14:58, 5.30it/s] 11%|█ | 5065/48008 [17:12<2:16:19, 5.25it/s] 11%|█ | 5066/48008 [17:13<2:16:17, 5.25it/s] 11%|█ | 5067/48008 [17:13<2:16:34, 5.24it/s] 11%|█ | 5068/48008 [17:13<2:15:40, 5.28it/s] 11%|█ | 5069/48008 [17:13<2:16:15, 5.25it/s] 11%|█ | 5070/48008 [17:13<2:15:21, 5.29it/s] 11%|█ | 5071/48008 [17:14<2:15:45, 5.27it/s] 11%|█ | 5072/48008 [17:14<2:16:10, 5.26it/s] 11%|█ | 5073/48008 [17:14<2:16:32, 5.24it/s] 11%|█ | 5074/48008 [17:14<2:20:09, 5.11it/s] 11%|█ | 5075/48008 [17:14<2:29:01, 4.80it/s] 11%|█ | 5076/48008 [17:15<2:25:01, 4.93it/s] 11%|█ | 5077/48008 [17:15<2:21:07, 5.07it/s] 11%|█ | 5078/48008 [17:15<2:18:53, 5.15it/s] 11%|█ | 5079/48008 [17:15<2:19:02, 5.15it/s] 11%|█ | 5080/48008 [17:15<2:42:55, 4.39it/s] 11%|█ | 5081/48008 [17:16<2:37:24, 4.55it/s] 11%|█ | 5082/48008 [17:16<2:30:00, 4.77it/s] 11%|█ | 5083/48008 [17:16<2:28:31, 4.82it/s] 11%|█ | 5084/48008 [17:16<2:25:29, 4.92it/s] 11%|█ | 5085/48008 [17:16<2:22:15, 5.03it/s] 11%|█ | 5086/48008 [17:17<2:18:43, 5.16it/s] 11%|█ | 5087/48008 [17:17<2:18:05, 5.18it/s] 11%|█ | 5088/48008 [17:17<2:16:36, 5.24it/s] 11%|█ | 5089/48008 [17:17<2:41:12, 4.44it/s] 11%|█ | 5090/48008 [17:18<2:33:49, 4.65it/s] 11%|█ | 5091/48008 [17:18<2:30:49, 4.74it/s] 11%|█ | 5092/48008 [17:18<2:36:18, 4.58it/s] 11%|█ | 5093/48008 [17:18<2:29:55, 4.77it/s] 11%|█ | 5094/48008 [17:18<2:25:25, 4.92it/s] 11%|█ | 5095/48008 [17:19<2:22:41, 5.01it/s] 11%|█ | 5096/48008 [17:19<2:20:52, 5.08it/s] 11%|█ | 5097/48008 [17:19<2:21:53, 5.04it/s] 11%|█ | 5098/48008 [17:19<2:20:01, 5.11it/s] 11%|█ | 5099/48008 [17:19<2:28:45, 4.81it/s] 11%|█ | 5100/48008 [17:20<2:26:22, 4.89it/s] {'loss': 0.3067, 'grad_norm': 0.2734375, 'learning_rate': 0.0001787577070488252, 'epoch': 0.11} 11%|█ | 5100/48008 [17:20<2:26:22, 4.89it/s] 11%|█ | 5101/48008 [17:20<2:23:44, 4.97it/s] 11%|█ | 5102/48008 [17:20<2:21:27, 5.06it/s] 11%|█ | 5103/48008 [17:20<2:22:16, 5.03it/s] 11%|█ | 5104/48008 [17:20<2:30:15, 4.76it/s] 11%|█ | 5105/48008 [17:21<2:26:01, 4.90it/s] 11%|█ | 5106/48008 [17:21<2:22:55, 5.00it/s] 11%|█ | 5107/48008 [17:21<2:20:40, 5.08it/s] 11%|█ | 5108/48008 [17:21<2:21:51, 5.04it/s] 11%|█ | 5109/48008 [17:21<2:19:50, 5.11it/s] 11%|█ | 5110/48008 [17:22<2:19:55, 5.11it/s] 11%|█ | 5111/48008 [17:22<2:18:35, 5.16it/s] 11%|█ | 5112/48008 [17:22<2:17:44, 5.19it/s] 11%|█ | 5113/48008 [17:22<2:16:43, 5.23it/s] 11%|█ | 5114/48008 [17:22<2:16:33, 5.24it/s] 11%|█ | 5115/48008 [17:22<2:17:11, 5.21it/s] 11%|█ | 5116/48008 [17:23<2:58:46, 4.00it/s] 11%|█ | 5117/48008 [17:23<2:47:39, 4.26it/s] 11%|█ | 5118/48008 [17:23<2:38:08, 4.52it/s] 11%|█ | 5119/48008 [17:23<2:31:16, 4.73it/s] 11%|█ | 5120/48008 [17:24<2:25:51, 4.90it/s] 11%|█ | 5121/48008 [17:24<2:23:01, 5.00it/s] 11%|█ | 5122/48008 [17:24<2:21:09, 5.06it/s] 11%|█ | 5123/48008 [17:24<2:19:09, 5.14it/s] 11%|█ | 5124/48008 [17:24<2:18:10, 5.17it/s] 11%|█ | 5125/48008 [17:25<5:03:36, 2.35it/s] 11%|█ | 5126/48008 [17:26<4:13:13, 2.82it/s] 11%|█ | 5127/48008 [17:26<3:38:11, 3.28it/s] 11%|█ | 5128/48008 [17:26<3:13:25, 3.69it/s] 11%|█ | 5129/48008 [17:26<2:58:00, 4.01it/s] 11%|█ | 5130/48008 [17:26<2:44:29, 4.34it/s] 11%|█ | 5131/48008 [17:27<5:03:29, 2.35it/s] 11%|█ | 5132/48008 [17:27<4:13:40, 2.82it/s] 11%|█ | 5133/48008 [17:28<3:42:40, 3.21it/s] 11%|█ | 5134/48008 [17:28<3:15:49, 3.65it/s] 11%|█ | 5135/48008 [17:28<2:59:26, 3.98it/s] 11%|█ | 5136/48008 [17:28<2:45:36, 4.31it/s] 11%|█ | 5137/48008 [17:29<6:09:00, 1.94it/s] 11%|█ | 5138/48008 [17:30<4:58:37, 2.39it/s] 11%|█ | 5139/48008 [17:30<4:09:52, 2.86it/s] 11%|█ | 5140/48008 [17:30<3:35:32, 3.31it/s] 11%|█ | 5141/48008 [17:30<3:11:19, 3.73it/s] 11%|█ | 5142/48008 [17:30<2:53:39, 4.11it/s] 11%|█ | 5143/48008 [17:30<2:43:03, 4.38it/s] 11%|█ | 5144/48008 [17:31<2:35:33, 4.59it/s] 11%|█ | 5145/48008 [17:31<2:29:59, 4.76it/s] 11%|█ | 5146/48008 [17:31<3:07:48, 3.80it/s] 11%|█ | 5147/48008 [17:31<3:02:14, 3.92it/s] 11%|█ | 5148/48008 [17:32<2:48:04, 4.25it/s] 11%|█ | 5149/48008 [17:32<2:38:59, 4.49it/s] 11%|█ | 5150/48008 [17:32<2:32:18, 4.69it/s] {'loss': 0.3053, 'grad_norm': 0.384765625, 'learning_rate': 0.00017854940843192802, 'epoch': 0.11} 11%|█ | 5150/48008 [17:32<2:32:18, 4.69it/s] 11%|█ | 5151/48008 [17:32<2:27:31, 4.84it/s] 11%|█ | 5152/48008 [17:32<2:22:52, 5.00it/s] 11%|█ | 5153/48008 [17:33<2:21:54, 5.03it/s] 11%|█ | 5154/48008 [17:33<2:19:22, 5.12it/s] 11%|█ | 5155/48008 [17:33<2:21:09, 5.06it/s] 11%|█ | 5156/48008 [17:33<2:20:51, 5.07it/s] 11%|█ | 5157/48008 [17:33<2:19:39, 5.11it/s] 11%|█ | 5158/48008 [17:34<2:17:59, 5.18it/s] 11%|█ | 5159/48008 [17:34<2:17:20, 5.20it/s] 11%|█ | 5160/48008 [17:34<2:19:20, 5.12it/s] 11%|█ | 5161/48008 [17:34<2:19:15, 5.13it/s] 11%|█ | 5162/48008 [17:34<2:17:58, 5.18it/s] 11%|█ | 5163/48008 [17:35<2:20:02, 5.10it/s] 11%|█ | 5164/48008 [17:35<2:18:46, 5.15it/s] 11%|█ | 5165/48008 [17:35<2:19:18, 5.13it/s] 11%|█ | 5166/48008 [17:35<2:28:04, 4.82it/s] 11%|█ | 5167/48008 [17:35<2:23:00, 4.99it/s] 11%|█ | 5168/48008 [17:36<2:21:03, 5.06it/s] 11%|█ | 5169/48008 [17:36<2:19:25, 5.12it/s] 11%|█ | 5170/48008 [17:36<2:19:17, 5.13it/s] 11%|█ | 5171/48008 [17:36<2:16:32, 5.23it/s] 11%|█ | 5172/48008 [17:36<2:19:09, 5.13it/s] 11%|█ | 5173/48008 [17:37<2:28:14, 4.82it/s] 11%|█ | 5174/48008 [17:37<2:26:09, 4.88it/s] 11%|█ | 5175/48008 [17:37<2:23:27, 4.98it/s] 11%|█ | 5176/48008 [17:37<2:21:49, 5.03it/s] 11%|█ | 5177/48008 [17:37<2:22:13, 5.02it/s] 11%|█ | 5178/48008 [17:38<2:20:00, 5.10it/s] 11%|█ | 5179/48008 [17:38<2:21:27, 5.05it/s] 11%|█ | 5180/48008 [17:38<2:23:56, 4.96it/s] 11%|█ | 5181/48008 [17:38<2:21:54, 5.03it/s] 11%|█ | 5182/48008 [17:38<2:22:57, 4.99it/s] 11%|█ | 5183/48008 [17:39<2:23:32, 4.97it/s] 11%|█ | 5184/48008 [17:39<2:24:30, 4.94it/s] 11%|█ | 5185/48008 [17:39<2:26:19, 4.88it/s] 11%|█ | 5186/48008 [17:39<2:23:28, 4.97it/s] 11%|█ | 5187/48008 [17:39<2:31:25, 4.71it/s] 11%|█ | 5188/48008 [17:40<2:26:50, 4.86it/s] 11%|█ | 5189/48008 [17:40<2:23:30, 4.97it/s] 11%|█ | 5190/48008 [17:40<2:25:18, 4.91it/s] 11%|█ | 5191/48008 [17:40<2:22:29, 5.01it/s] 11%|█ | 5192/48008 [17:40<2:20:37, 5.07it/s] 11%|█ | 5193/48008 [17:41<2:19:30, 5.11it/s] 11%|█ | 5194/48008 [17:41<2:18:34, 5.15it/s] 11%|█ | 5195/48008 [17:41<2:20:03, 5.09it/s] 11%|█ | 5196/48008 [17:41<2:18:44, 5.14it/s] 11%|█ | 5197/48008 [17:41<2:21:58, 5.03it/s] 11%|█ | 5198/48008 [17:42<2:20:32, 5.08it/s] 11%|█ | 5199/48008 [17:42<2:20:08, 5.09it/s] 11%|█ | 5200/48008 [17:42<2:19:13, 5.12it/s] {'loss': 0.304, 'grad_norm': 0.443359375, 'learning_rate': 0.00017834110981503084, 'epoch': 0.11} 11%|█ | 5200/48008 [17:42<2:19:13, 5.12it/s] 11%|█ | 5201/48008 [17:42<2:18:59, 5.13it/s] 11%|█ | 5202/48008 [17:42<2:18:21, 5.16it/s] 11%|█ | 5203/48008 [17:43<2:17:19, 5.20it/s] 11%|█ | 5204/48008 [17:43<2:19:15, 5.12it/s] 11%|█ | 5205/48008 [17:43<2:17:08, 5.20it/s] 11%|█ | 5206/48008 [17:43<2:18:24, 5.15it/s] 11%|█ | 5207/48008 [17:43<2:19:16, 5.12it/s] 11%|█ | 5208/48008 [17:44<2:28:01, 4.82it/s] 11%|█ | 5209/48008 [17:44<2:27:06, 4.85it/s] 11%|█ | 5210/48008 [17:44<2:25:11, 4.91it/s] 11%|█ | 5211/48008 [17:44<2:23:03, 4.99it/s] 11%|█ | 5212/48008 [17:44<2:20:31, 5.08it/s] 11%|█ | 5213/48008 [17:45<2:21:58, 5.02it/s] 11%|█ | 5214/48008 [17:45<2:22:25, 5.01it/s] 11%|█ | 5215/48008 [17:45<2:20:03, 5.09it/s] 11%|█ | 5216/48008 [17:45<2:19:55, 5.10it/s] 11%|█ | 5217/48008 [17:45<2:18:54, 5.13it/s] 11%|█ | 5218/48008 [17:46<2:20:22, 5.08it/s] 11%|█ | 5219/48008 [17:46<2:21:29, 5.04it/s] 11%|█ | 5220/48008 [17:46<2:21:44, 5.03it/s] 11%|█ | 5221/48008 [17:46<2:21:12, 5.05it/s] 11%|█ | 5222/48008 [17:46<2:18:47, 5.14it/s] 11%|█ | 5223/48008 [17:46<2:17:54, 5.17it/s] 11%|█ | 5224/48008 [17:47<2:21:33, 5.04it/s] 11%|█ | 5225/48008 [17:47<2:19:56, 5.10it/s] 11%|█ | 5226/48008 [17:47<2:17:59, 5.17it/s] 11%|█ | 5227/48008 [17:47<2:16:56, 5.21it/s] 11%|█ | 5228/48008 [17:47<2:16:05, 5.24it/s] 11%|█ | 5229/48008 [17:48<2:16:13, 5.23it/s] 11%|█ | 5230/48008 [17:48<2:16:10, 5.24it/s] 11%|█ | 5231/48008 [17:48<2:15:56, 5.24it/s] 11%|█ | 5232/48008 [17:48<2:16:06, 5.24it/s] 11%|█ | 5233/48008 [17:48<2:15:26, 5.26it/s] 11%|█ | 5234/48008 [17:49<2:16:50, 5.21it/s] 11%|█ | 5235/48008 [17:49<2:58:14, 4.00it/s] 11%|█ | 5236/48008 [17:49<2:45:39, 4.30it/s] 11%|█ | 5237/48008 [17:49<2:36:06, 4.57it/s] 11%|█ | 5238/48008 [17:50<2:29:05, 4.78it/s] 11%|█ | 5239/48008 [17:50<2:27:26, 4.83it/s] 11%|█ | 5240/48008 [17:50<2:24:58, 4.92it/s] 11%|█ | 5241/48008 [17:50<2:21:52, 5.02it/s] 11%|█ | 5242/48008 [17:50<2:22:49, 4.99it/s] 11%|█ | 5243/48008 [17:51<2:20:30, 5.07it/s] 11%|█ | 5244/48008 [17:51<2:18:21, 5.15it/s] 11%|█ | 5245/48008 [17:51<2:17:35, 5.18it/s] 11%|█ | 5246/48008 [17:51<2:19:36, 5.11it/s] 11%|█ | 5247/48008 [17:51<2:18:42, 5.14it/s] 11%|█ | 5248/48008 [17:51<2:18:01, 5.16it/s] 11%|█ | 5249/48008 [17:52<2:18:16, 5.15it/s] 11%|█ | 5250/48008 [17:52<2:16:21, 5.23it/s] {'loss': 0.3037, 'grad_norm': 0.3515625, 'learning_rate': 0.00017813281119813366, 'epoch': 0.11} 11%|█ | 5250/48008 [17:52<2:16:21, 5.23it/s] 11%|█ | 5251/48008 [17:52<2:16:38, 5.22it/s] 11%|█ | 5252/48008 [17:52<2:17:23, 5.19it/s] 11%|█ | 5253/48008 [17:52<2:16:59, 5.20it/s] 11%|█ | 5254/48008 [17:53<2:18:13, 5.15it/s] 11%|█ | 5255/48008 [17:53<2:16:41, 5.21it/s] 11%|█ | 5256/48008 [17:53<2:17:47, 5.17it/s] 11%|█ | 5257/48008 [17:53<2:16:45, 5.21it/s] 11%|█ | 5258/48008 [17:53<2:16:17, 5.23it/s] 11%|█ | 5259/48008 [17:54<2:15:04, 5.27it/s] 11%|█ | 5260/48008 [17:54<2:14:37, 5.29it/s] 11%|█ | 5261/48008 [17:54<2:15:18, 5.27it/s] 11%|█ | 5262/48008 [17:54<2:15:08, 5.27it/s] 11%|█ | 5263/48008 [17:54<2:15:20, 5.26it/s] 11%|█ | 5264/48008 [17:55<2:16:00, 5.24it/s] 11%|█ | 5265/48008 [17:55<2:15:28, 5.26it/s] 11%|█ | 5266/48008 [17:55<2:15:33, 5.25it/s] 11%|█ | 5267/48008 [17:55<2:14:26, 5.30it/s] 11%|█ | 5268/48008 [17:55<2:15:12, 5.27it/s] 11%|█ | 5269/48008 [17:56<2:18:13, 5.15it/s] 11%|█ | 5270/48008 [17:56<2:17:07, 5.19it/s] 11%|█ | 5271/48008 [17:56<2:18:05, 5.16it/s] 11%|█ | 5272/48008 [17:56<2:17:52, 5.17it/s] 11%|█ | 5273/48008 [17:56<2:16:51, 5.20it/s] 11%|█ | 5274/48008 [17:56<2:16:52, 5.20it/s] 11%|█ | 5275/48008 [17:57<2:15:46, 5.25it/s] 11%|█ | 5276/48008 [17:57<2:15:48, 5.24it/s] 11%|█ | 5277/48008 [17:57<2:15:10, 5.27it/s] 11%|█ | 5278/48008 [17:57<2:15:19, 5.26it/s] 11%|█ | 5279/48008 [17:57<2:18:17, 5.15it/s] 11%|█ | 5280/48008 [17:58<2:19:14, 5.11it/s] 11%|█ | 5281/48008 [17:58<2:28:07, 4.81it/s] 11%|█ | 5282/48008 [17:58<2:24:32, 4.93it/s] 11%|█ | 5283/48008 [17:58<2:21:54, 5.02it/s] 11%|█ | 5284/48008 [17:58<2:24:22, 4.93it/s] 11%|█ | 5285/48008 [17:59<2:46:27, 4.28it/s] 11%|█ | 5286/48008 [17:59<2:40:08, 4.45it/s] 11%|█ | 5287/48008 [17:59<2:35:37, 4.58it/s] 11%|█ | 5288/48008 [17:59<2:28:51, 4.78it/s] 11%|█ | 5289/48008 [18:00<2:23:05, 4.98it/s] 11%|█ | 5290/48008 [18:00<2:20:32, 5.07it/s] 11%|█ | 5291/48008 [18:00<2:18:48, 5.13it/s] 11%|█ | 5292/48008 [18:00<2:17:45, 5.17it/s] 11%|█ | 5293/48008 [18:00<2:18:18, 5.15it/s] 11%|█ | 5294/48008 [18:01<2:19:53, 5.09it/s] 11%|█ | 5295/48008 [18:01<2:22:37, 4.99it/s] 11%|█ | 5296/48008 [18:01<2:22:11, 5.01it/s] 11%|█ | 5297/48008 [18:01<2:20:05, 5.08it/s] 11%|█ | 5298/48008 [18:01<2:18:02, 5.16it/s] 11%|█ | 5299/48008 [18:01<2:17:25, 5.18it/s] 11%|█ | 5300/48008 [18:02<2:20:59, 5.05it/s] {'loss': 0.3051, 'grad_norm': 0.7890625, 'learning_rate': 0.00017792451258123647, 'epoch': 0.11} 11%|█ | 5300/48008 [18:02<2:20:59, 5.05it/s] 11%|█ | 5301/48008 [18:02<2:19:10, 5.11it/s] 11%|█ | 5302/48008 [18:02<2:17:41, 5.17it/s] 11%|█ | 5303/48008 [18:02<2:16:14, 5.22it/s] 11%|█ | 5304/48008 [18:02<2:16:32, 5.21it/s] 11%|█ | 5305/48008 [18:03<2:15:17, 5.26it/s] 11%|█ | 5306/48008 [18:03<2:15:37, 5.25it/s] 11%|█ | 5307/48008 [18:03<2:16:00, 5.23it/s] 11%|█ | 5308/48008 [18:03<2:19:32, 5.10it/s] 11%|█ | 5309/48008 [18:03<2:17:57, 5.16it/s] 11%|█ | 5310/48008 [18:04<2:21:06, 5.04it/s] 11%|█ | 5311/48008 [18:04<2:19:24, 5.10it/s] 11%|█ | 5312/48008 [18:04<2:17:59, 5.16it/s] 11%|█ | 5313/48008 [18:04<2:20:53, 5.05it/s] 11%|█ | 5314/48008 [18:04<2:29:02, 4.77it/s] 11%|█ | 5315/48008 [18:05<2:25:15, 4.90it/s] 11%|█ | 5316/48008 [18:05<2:47:03, 4.26it/s] 11%|█ | 5317/48008 [18:05<2:39:51, 4.45it/s] 11%|█ | 5318/48008 [18:05<2:32:25, 4.67it/s] 11%|█ | 5319/48008 [18:06<4:53:56, 2.42it/s] 11%|█ | 5320/48008 [18:06<4:08:34, 2.86it/s] 11%|█ | 5321/48008 [18:07<3:37:09, 3.28it/s] 11%|█ | 5322/48008 [18:07<3:12:30, 3.70it/s] 11%|█ | 5323/48008 [18:07<2:54:56, 4.07it/s] 11%|█ | 5324/48008 [18:07<2:45:30, 4.30it/s] 11%|█ | 5325/48008 [18:07<2:37:06, 4.53it/s] 11%|█ | 5326/48008 [18:08<2:32:01, 4.68it/s] 11%|█ | 5327/48008 [18:08<2:26:02, 4.87it/s] 11%|█ | 5328/48008 [18:08<2:22:40, 4.99it/s] 11%|█ | 5329/48008 [18:08<2:21:26, 5.03it/s] 11%|█ | 5330/48008 [18:08<2:22:16, 5.00it/s] 11%|█ | 5331/48008 [18:09<2:19:56, 5.08it/s] 11%|█ | 5332/48008 [18:09<2:17:55, 5.16it/s] 11%|█ | 5333/48008 [18:09<2:20:09, 5.07it/s] 11%|█ | 5334/48008 [18:09<2:18:29, 5.14it/s] 11%|█ | 5335/48008 [18:09<2:18:48, 5.12it/s] 11%|█ | 5336/48008 [18:10<2:19:23, 5.10it/s] 11%|█ | 5337/48008 [18:10<2:19:28, 5.10it/s] 11%|█ | 5338/48008 [18:10<2:19:12, 5.11it/s] 11%|█ | 5339/48008 [18:10<2:20:20, 5.07it/s] 11%|█ | 5340/48008 [18:10<2:17:55, 5.16it/s] 11%|█ | 5341/48008 [18:10<2:17:01, 5.19it/s] 11%|█ | 5342/48008 [18:11<2:16:19, 5.22it/s] 11%|█ | 5343/48008 [18:11<2:16:01, 5.23it/s] 11%|█ | 5344/48008 [18:11<2:14:53, 5.27it/s] 11%|█ | 5345/48008 [18:11<2:13:54, 5.31it/s] 11%|█ | 5346/48008 [18:11<2:14:40, 5.28it/s] 11%|█ | 5347/48008 [18:12<2:14:19, 5.29it/s] 11%|█ | 5348/48008 [18:12<2:14:39, 5.28it/s] 11%|█ | 5349/48008 [18:12<2:13:59, 5.31it/s] 11%|█ | 5350/48008 [18:12<2:14:10, 5.30it/s] {'loss': 0.3022, 'grad_norm': 0.4453125, 'learning_rate': 0.0001777162139643393, 'epoch': 0.11} 11%|█ | 5350/48008 [18:12<2:14:10, 5.30it/s] 11%|█ | 5351/48008 [18:12<2:17:44, 5.16it/s] 11%|█ | 5352/48008 [18:13<2:18:15, 5.14it/s] 11%|█ | 5353/48008 [18:13<2:17:13, 5.18it/s] 11%|█ | 5354/48008 [18:13<2:16:40, 5.20it/s] 11%|█ | 5355/48008 [18:13<2:15:10, 5.26it/s] 11%|█ | 5356/48008 [18:13<2:40:08, 4.44it/s] 11%|█ | 5357/48008 [18:14<2:32:53, 4.65it/s] 11%|█ | 5358/48008 [18:14<2:26:50, 4.84it/s] 11%|█ | 5359/48008 [18:14<2:23:14, 4.96it/s] 11%|█ | 5360/48008 [18:14<2:20:34, 5.06it/s] 11%|█ | 5361/48008 [18:14<2:18:40, 5.13it/s] 11%|█ | 5362/48008 [18:15<2:17:44, 5.16it/s] 11%|█ | 5363/48008 [18:15<2:58:59, 3.97it/s] 11%|█ | 5364/48008 [18:15<2:46:55, 4.26it/s] 11%|█ | 5365/48008 [18:15<2:47:18, 4.25it/s] 11%|█ | 5366/48008 [18:16<2:36:52, 4.53it/s] 11%|█ | 5367/48008 [18:16<2:29:38, 4.75it/s] 11%|█ | 5368/48008 [18:16<2:25:45, 4.88it/s] 11%|█ | 5369/48008 [18:16<2:23:17, 4.96it/s] 11%|█ | 5370/48008 [18:16<2:22:31, 4.99it/s] 11%|█ | 5371/48008 [18:17<2:20:11, 5.07it/s] 11%|█ | 5372/48008 [18:17<2:17:49, 5.16it/s] 11%|█ | 5373/48008 [18:17<2:16:34, 5.20it/s] 11%|█ | 5374/48008 [18:17<2:15:28, 5.24it/s] 11%|█ | 5375/48008 [18:17<2:15:55, 5.23it/s] 11%|█ | 5376/48008 [18:18<2:17:08, 5.18it/s] 11%|█ | 5377/48008 [18:18<2:19:10, 5.11it/s] 11%|█ | 5378/48008 [18:18<2:18:03, 5.15it/s] 11%|█ | 5379/48008 [18:18<2:19:24, 5.10it/s] 11%|█ | 5380/48008 [18:18<2:20:23, 5.06it/s] 11%|█ | 5381/48008 [18:18<2:18:57, 5.11it/s] 11%|█ | 5382/48008 [18:19<2:19:47, 5.08it/s] 11%|█ | 5383/48008 [18:19<2:18:32, 5.13it/s] 11%|█ | 5384/48008 [18:19<2:16:21, 5.21it/s] 11%|█ | 5385/48008 [18:19<2:17:15, 5.18it/s] 11%|█ | 5386/48008 [18:19<2:19:39, 5.09it/s] 11%|█ | 5387/48008 [18:20<2:43:16, 4.35it/s] 11%|█ | 5388/48008 [18:20<2:35:34, 4.57it/s] 11%|█ | 5389/48008 [18:20<2:33:34, 4.63it/s] 11%|█ | 5390/48008 [18:20<2:27:47, 4.81it/s] 11%|█ | 5391/48008 [18:21<2:25:10, 4.89it/s] 11%|█ | 5392/48008 [18:21<2:21:54, 5.00it/s] 11%|█ | 5393/48008 [18:21<2:44:20, 4.32it/s] 11%|█ | 5394/48008 [18:21<2:35:46, 4.56it/s] 11%|█ | 5395/48008 [18:21<2:30:02, 4.73it/s] 11%|█ | 5396/48008 [18:22<2:25:50, 4.87it/s] 11%|█ | 5397/48008 [18:22<2:22:04, 5.00it/s] 11%|█ | 5398/48008 [18:22<2:19:28, 5.09it/s] 11%|█ | 5399/48008 [18:22<2:18:33, 5.13it/s] 11%|█ | 5400/48008 [18:22<2:17:14, 5.17it/s] {'loss': 0.3031, 'grad_norm': 0.5, 'learning_rate': 0.0001775079153474421, 'epoch': 0.11} 11%|█ | 5400/48008 [18:22<2:17:14, 5.17it/s] 11%|█▏ | 5401/48008 [18:23<2:18:08, 5.14it/s] 11%|█▏ | 5402/48008 [18:23<2:16:35, 5.20it/s] 11%|█▏ | 5403/48008 [18:23<2:16:07, 5.22it/s] 11%|█▏ | 5404/48008 [18:23<2:14:57, 5.26it/s] 11%|█▏ | 5405/48008 [18:23<2:15:05, 5.26it/s] 11%|█▏ | 5406/48008 [18:24<2:14:38, 5.27it/s] 11%|█▏ | 5407/48008 [18:24<2:17:05, 5.18it/s] 11%|█▏ | 5408/48008 [18:24<2:15:45, 5.23it/s] 11%|█▏ | 5409/48008 [18:24<2:25:30, 4.88it/s] 11%|█▏ | 5410/48008 [18:25<3:03:49, 3.86it/s] 11%|█▏ | 5411/48008 [18:25<2:48:57, 4.20it/s] 11%|█▏ | 5412/48008 [18:25<3:03:38, 3.87it/s] 11%|█▏ | 5413/48008 [18:25<2:59:18, 3.96it/s] 11%|█▏ | 5414/48008 [18:25<2:48:23, 4.22it/s] 11%|█▏ | 5415/48008 [18:26<2:40:58, 4.41it/s] 11%|█▏ | 5416/48008 [18:26<2:32:32, 4.65it/s] 11%|█▏ | 5417/48008 [18:26<2:27:20, 4.82it/s] 11%|█▏ | 5418/48008 [18:26<2:25:40, 4.87it/s] 11%|█▏ | 5419/48008 [18:26<2:21:58, 5.00it/s] 11%|█▏ | 5420/48008 [18:27<2:20:12, 5.06it/s] 11%|█▏ | 5421/48008 [18:27<2:18:12, 5.14it/s] 11%|█▏ | 5422/48008 [18:27<2:21:32, 5.01it/s] 11%|█▏ | 5423/48008 [18:27<2:22:13, 4.99it/s] 11%|█▏ | 5424/48008 [18:27<2:22:36, 4.98it/s] 11%|█▏ | 5425/48008 [18:28<2:20:17, 5.06it/s] 11%|█▏ | 5426/48008 [18:28<2:18:27, 5.13it/s] 11%|█▏ | 5427/48008 [18:28<2:17:11, 5.17it/s] 11%|█▏ | 5428/48008 [18:28<2:19:07, 5.10it/s] 11%|█▏ | 5429/48008 [18:28<2:19:03, 5.10it/s] 11%|█▏ | 5430/48008 [18:29<2:21:40, 5.01it/s] 11%|█▏ | 5431/48008 [18:29<2:19:19, 5.09it/s] 11%|█▏ | 5432/48008 [18:29<2:18:14, 5.13it/s] 11%|█▏ | 5433/48008 [18:29<2:18:20, 5.13it/s] 11%|█▏ | 5434/48008 [18:29<2:16:01, 5.22it/s] 11%|█▏ | 5435/48008 [18:30<2:25:27, 4.88it/s] 11%|█▏ | 5436/48008 [18:30<2:24:57, 4.89it/s] 11%|█▏ | 5437/48008 [18:30<2:24:29, 4.91it/s] 11%|█▏ | 5438/48008 [18:30<2:21:12, 5.02it/s] 11%|█▏ | 5439/48008 [18:30<2:19:07, 5.10it/s] 11%|█▏ | 5440/48008 [18:31<2:18:05, 5.14it/s] 11%|█▏ | 5441/48008 [18:31<2:18:37, 5.12it/s] 11%|█▏ | 5442/48008 [18:31<2:16:10, 5.21it/s] 11%|█▏ | 5443/48008 [18:31<2:15:54, 5.22it/s] 11%|█▏ | 5444/48008 [18:31<2:19:48, 5.07it/s] 11%|█▏ | 5445/48008 [18:32<2:22:10, 4.99it/s] 11%|█▏ | 5446/48008 [18:32<2:21:45, 5.00it/s] 11%|█▏ | 5447/48008 [18:32<2:22:18, 4.98it/s] 11%|█▏ | 5448/48008 [18:32<2:22:36, 4.97it/s] 11%|█▏ | 5449/48008 [18:32<2:21:36, 5.01it/s] 11%|█▏ | 5450/48008 [18:33<2:19:27, 5.09it/s] {'loss': 0.3023, 'grad_norm': 0.2373046875, 'learning_rate': 0.00017729961673054493, 'epoch': 0.11} 11%|█▏ | 5450/48008 [18:33<2:19:27, 5.09it/s] 11%|█▏ | 5451/48008 [18:33<2:18:24, 5.12it/s] 11%|█▏ | 5452/48008 [18:33<2:20:04, 5.06it/s] 11%|█▏ | 5453/48008 [18:33<2:18:36, 5.12it/s] 11%|█▏ | 5454/48008 [18:33<2:16:45, 5.19it/s] 11%|█▏ | 5455/48008 [18:34<2:17:07, 5.17it/s] 11%|█▏ | 5456/48008 [18:34<2:18:12, 5.13it/s] 11%|█▏ | 5457/48008 [18:34<2:17:41, 5.15it/s] 11%|█▏ | 5458/48008 [18:34<2:19:54, 5.07it/s] 11%|█▏ | 5459/48008 [18:34<2:18:22, 5.13it/s] 11%|█▏ | 5460/48008 [18:35<2:17:11, 5.17it/s] 11%|█▏ | 5461/48008 [18:35<2:15:50, 5.22it/s] 11%|█▏ | 5462/48008 [18:35<2:15:16, 5.24it/s] 11%|█▏ | 5463/48008 [18:35<2:15:30, 5.23it/s] 11%|█▏ | 5464/48008 [18:35<2:16:19, 5.20it/s] 11%|█▏ | 5465/48008 [18:35<2:16:04, 5.21it/s] 11%|█▏ | 5466/48008 [18:36<2:18:09, 5.13it/s] 11%|█▏ | 5467/48008 [18:36<2:17:32, 5.15it/s] 11%|█▏ | 5468/48008 [18:36<2:16:15, 5.20it/s] 11%|█▏ | 5469/48008 [18:36<2:15:13, 5.24it/s] 11%|█▏ | 5470/48008 [18:36<2:17:29, 5.16it/s] 11%|█▏ | 5471/48008 [18:37<2:15:43, 5.22it/s] 11%|█▏ | 5472/48008 [18:37<2:15:26, 5.23it/s] 11%|█▏ | 5473/48008 [18:37<2:17:52, 5.14it/s] 11%|█▏ | 5474/48008 [18:37<2:16:44, 5.18it/s] 11%|█▏ | 5475/48008 [18:37<2:18:45, 5.11it/s] 11%|█▏ | 5476/48008 [18:38<2:17:12, 5.17it/s] 11%|█▏ | 5477/48008 [18:38<2:16:39, 5.19it/s] 11%|█▏ | 5478/48008 [18:38<2:15:53, 5.22it/s] 11%|█▏ | 5479/48008 [18:38<2:40:04, 4.43it/s] 11%|█▏ | 5480/48008 [18:39<2:58:37, 3.97it/s] 11%|█▏ | 5481/48008 [18:39<3:10:21, 3.72it/s] 11%|█▏ | 5482/48008 [18:39<2:53:50, 4.08it/s] 11%|█▏ | 5483/48008 [18:39<2:40:48, 4.41it/s] 11%|█▏ | 5484/48008 [18:39<2:35:14, 4.57it/s] 11%|█▏ | 5485/48008 [18:40<2:29:37, 4.74it/s] 11%|█▏ | 5486/48008 [18:40<2:24:54, 4.89it/s] 11%|█▏ | 5487/48008 [18:40<2:22:39, 4.97it/s] 11%|█▏ | 5488/48008 [18:40<2:30:58, 4.69it/s] 11%|█▏ | 5489/48008 [18:40<2:23:57, 4.92it/s] 11%|█▏ | 5490/48008 [18:41<2:22:11, 4.98it/s] 11%|█▏ | 5491/48008 [18:41<2:18:45, 5.11it/s] 11%|█▏ | 5492/48008 [18:41<2:18:02, 5.13it/s] 11%|█▏ | 5493/48008 [18:41<2:19:24, 5.08it/s] 11%|█▏ | 5494/48008 [18:41<2:18:02, 5.13it/s] 11%|█▏ | 5495/48008 [18:42<2:17:25, 5.16it/s] 11%|█▏ | 5496/48008 [18:42<2:16:44, 5.18it/s] 11%|█▏ | 5497/48008 [18:42<2:16:38, 5.19it/s] 11%|█▏ | 5498/48008 [18:42<2:15:56, 5.21it/s] 11%|█▏ | 5499/48008 [18:42<2:15:14, 5.24it/s] 11%|█▏ | 5500/48008 [18:43<2:17:36, 5.15it/s] {'loss': 0.3037, 'grad_norm': 0.185546875, 'learning_rate': 0.00017709131811364774, 'epoch': 0.11} 11%|█▏ | 5500/48008 [18:43<2:17:36, 5.15it/s] 11%|█▏ | 5501/48008 [18:43<2:17:15, 5.16it/s] 11%|█▏ | 5502/48008 [18:43<2:16:30, 5.19it/s] 11%|█▏ | 5503/48008 [18:43<2:15:57, 5.21it/s] 11%|█▏ | 5504/48008 [18:43<2:16:13, 5.20it/s] 11%|█▏ | 5505/48008 [18:44<2:18:09, 5.13it/s] 11%|█▏ | 5506/48008 [18:44<2:17:30, 5.15it/s] 11%|█▏ | 5507/48008 [18:44<2:16:49, 5.18it/s] 11%|█▏ | 5508/48008 [18:44<2:16:35, 5.19it/s] 11%|█▏ | 5509/48008 [18:44<2:17:37, 5.15it/s] 11%|█▏ | 5510/48008 [18:45<2:17:04, 5.17it/s] 11%|█▏ | 5511/48008 [18:45<2:16:34, 5.19it/s] 11%|█▏ | 5512/48008 [18:45<2:16:50, 5.18it/s] 11%|█▏ | 5513/48008 [18:45<2:17:16, 5.16it/s] 11%|█▏ | 5514/48008 [18:45<2:17:58, 5.13it/s] 11%|█▏ | 5515/48008 [18:45<2:18:15, 5.12it/s] 11%|█▏ | 5516/48008 [18:46<2:27:01, 4.82it/s] 11%|█▏ | 5517/48008 [18:46<2:23:04, 4.95it/s] 11%|█▏ | 5518/48008 [18:46<2:19:34, 5.07it/s] 11%|█▏ | 5519/48008 [18:46<2:18:08, 5.13it/s] 11%|█▏ | 5520/48008 [18:46<2:17:20, 5.16it/s] 12%|█▏ | 5521/48008 [18:47<2:16:13, 5.20it/s] 12%|█▏ | 5522/48008 [18:47<2:16:07, 5.20it/s] 12%|█▏ | 5523/48008 [18:47<2:18:07, 5.13it/s] 12%|█▏ | 5524/48008 [18:47<2:19:25, 5.08it/s] 12%|█▏ | 5525/48008 [18:47<2:17:58, 5.13it/s] 12%|█▏ | 5526/48008 [18:48<2:16:19, 5.19it/s] 12%|█▏ | 5527/48008 [18:48<2:15:03, 5.24it/s] 12%|█▏ | 5528/48008 [18:48<2:14:01, 5.28it/s] 12%|█▏ | 5529/48008 [18:48<2:15:43, 5.22it/s] 12%|█▏ | 5530/48008 [18:48<2:14:58, 5.25it/s] 12%|█▏ | 5531/48008 [18:49<2:39:16, 4.44it/s] 12%|█▏ | 5532/48008 [18:49<2:34:24, 4.58it/s] 12%|█▏ | 5533/48008 [18:49<2:31:28, 4.67it/s] 12%|█▏ | 5534/48008 [18:49<2:26:27, 4.83it/s] 12%|█▏ | 5535/48008 [18:49<2:22:37, 4.96it/s] 12%|█▏ | 5536/48008 [18:50<2:20:14, 5.05it/s] 12%|█▏ | 5537/48008 [18:50<2:17:16, 5.16it/s] 12%|█▏ | 5538/48008 [18:50<2:15:38, 5.22it/s] 12%|█▏ | 5539/48008 [18:50<2:14:51, 5.25it/s] 12%|█▏ | 5540/48008 [18:50<2:14:24, 5.27it/s] 12%|█▏ | 5541/48008 [18:51<2:18:33, 5.11it/s] 12%|█▏ | 5542/48008 [18:51<2:16:34, 5.18it/s] 12%|█▏ | 5543/48008 [18:51<2:15:17, 5.23it/s] 12%|█▏ | 5544/48008 [18:51<2:16:21, 5.19it/s] 12%|█▏ | 5545/48008 [18:51<2:18:19, 5.12it/s] 12%|█▏ | 5546/48008 [18:52<2:18:30, 5.11it/s] 12%|█▏ | 5547/48008 [18:52<2:18:22, 5.11it/s] 12%|█▏ | 5548/48008 [18:52<2:17:23, 5.15it/s] 12%|█▏ | 5549/48008 [18:52<2:14:38, 5.26it/s] 12%|█▏ | 5550/48008 [18:52<2:17:37, 5.14it/s] {'loss': 0.3045, 'grad_norm': 0.2080078125, 'learning_rate': 0.00017688301949675053, 'epoch': 0.12} 12%|█▏ | 5550/48008 [18:52<2:17:37, 5.14it/s] 12%|█▏ | 5551/48008 [18:53<2:17:18, 5.15it/s] 12%|█▏ | 5552/48008 [18:53<2:15:08, 5.24it/s] 12%|█▏ | 5553/48008 [18:53<2:14:45, 5.25it/s] 12%|█▏ | 5554/48008 [18:53<2:14:12, 5.27it/s] 12%|█▏ | 5555/48008 [18:53<2:14:10, 5.27it/s] 12%|█▏ | 5556/48008 [18:54<2:55:23, 4.03it/s] 12%|█▏ | 5557/48008 [18:54<2:43:30, 4.33it/s] 12%|█▏ | 5558/48008 [18:54<2:34:45, 4.57it/s] 12%|█▏ | 5559/48008 [18:54<2:28:20, 4.77it/s] 12%|█▏ | 5560/48008 [18:54<2:23:20, 4.94it/s] 12%|█▏ | 5561/48008 [18:55<2:21:09, 5.01it/s] 12%|█▏ | 5562/48008 [18:55<2:18:22, 5.11it/s] 12%|█▏ | 5563/48008 [18:55<2:17:39, 5.14it/s] 12%|█▏ | 5564/48008 [18:55<2:18:15, 5.12it/s] 12%|█▏ | 5565/48008 [18:55<2:17:40, 5.14it/s] 12%|█▏ | 5566/48008 [18:56<2:41:32, 4.38it/s] 12%|█▏ | 5567/48008 [18:56<2:33:35, 4.61it/s] 12%|█▏ | 5568/48008 [18:56<2:30:13, 4.71it/s] 12%|█▏ | 5569/48008 [18:56<2:24:06, 4.91it/s] 12%|█▏ | 5570/48008 [18:56<2:21:01, 5.02it/s] 12%|█▏ | 5571/48008 [18:57<2:20:15, 5.04it/s] 12%|█▏ | 5572/48008 [18:57<2:17:44, 5.13it/s] 12%|█▏ | 5573/48008 [18:57<2:16:50, 5.17it/s] 12%|█▏ | 5574/48008 [18:57<2:16:32, 5.18it/s] 12%|█▏ | 5575/48008 [18:58<2:26:06, 4.84it/s] 12%|█▏ | 5576/48008 [18:58<2:22:24, 4.97it/s] 12%|█▏ | 5577/48008 [18:58<2:21:28, 5.00it/s] 12%|█▏ | 5578/48008 [18:58<2:29:12, 4.74it/s] 12%|█▏ | 5579/48008 [18:58<2:34:43, 4.57it/s] 12%|█▏ | 5580/48008 [18:59<2:29:36, 4.73it/s] 12%|█▏ | 5581/48008 [18:59<2:24:07, 4.91it/s] 12%|█▏ | 5582/48008 [18:59<2:20:24, 5.04it/s] 12%|█▏ | 5583/48008 [18:59<2:18:44, 5.10it/s] 12%|█▏ | 5584/48008 [18:59<2:17:13, 5.15it/s] 12%|█▏ | 5585/48008 [18:59<2:16:36, 5.18it/s] 12%|█▏ | 5586/48008 [19:00<2:17:45, 5.13it/s] 12%|█▏ | 5587/48008 [19:00<2:16:55, 5.16it/s] 12%|█▏ | 5588/48008 [19:00<2:15:12, 5.23it/s] 12%|█▏ | 5589/48008 [19:00<2:14:06, 5.27it/s] 12%|█▏ | 5590/48008 [19:00<2:13:48, 5.28it/s] 12%|█▏ | 5591/48008 [19:01<2:13:55, 5.28it/s] 12%|█▏ | 5592/48008 [19:01<2:13:33, 5.29it/s] 12%|█▏ | 5593/48008 [19:01<2:13:09, 5.31it/s] 12%|█▏ | 5594/48008 [19:01<2:13:58, 5.28it/s] 12%|█▏ | 5595/48008 [19:01<2:14:28, 5.26it/s] 12%|█▏ | 5596/48008 [19:02<2:16:27, 5.18it/s] 12%|█▏ | 5597/48008 [19:02<2:15:27, 5.22it/s] 12%|█▏ | 5598/48008 [19:02<2:15:21, 5.22it/s] 12%|█▏ | 5599/48008 [19:02<2:14:38, 5.25it/s] 12%|█▏ | 5600/48008 [19:02<2:14:15, 5.26it/s] {'loss': 0.306, 'grad_norm': 0.169921875, 'learning_rate': 0.00017667472087985338, 'epoch': 0.12} 12%|█▏ | 5600/48008 [19:02<2:14:15, 5.26it/s] 12%|█▏ | 5601/48008 [19:03<2:13:46, 5.28it/s] 12%|█▏ | 5602/48008 [19:03<2:18:01, 5.12it/s] 12%|█▏ | 5603/48008 [19:03<2:58:03, 3.97it/s] 12%|█▏ | 5604/48008 [19:03<2:45:55, 4.26it/s] 12%|█▏ | 5605/48008 [19:04<2:36:14, 4.52it/s] 12%|█▏ | 5606/48008 [19:04<2:29:56, 4.71it/s] 12%|█▏ | 5607/48008 [19:04<2:25:31, 4.86it/s] 12%|█▏ | 5608/48008 [19:04<2:22:16, 4.97it/s] 12%|█▏ | 5609/48008 [19:04<2:20:59, 5.01it/s] 12%|█▏ | 5610/48008 [19:04<2:21:13, 5.00it/s] 12%|█▏ | 5611/48008 [19:05<2:23:08, 4.94it/s] 12%|█▏ | 5612/48008 [19:05<2:22:53, 4.94it/s] 12%|█▏ | 5613/48008 [19:05<2:20:26, 5.03it/s] 12%|█▏ | 5614/48008 [19:05<2:18:51, 5.09it/s] 12%|█▏ | 5615/48008 [19:05<2:18:44, 5.09it/s] 12%|█▏ | 5616/48008 [19:06<2:17:39, 5.13it/s] 12%|█▏ | 5617/48008 [19:06<2:16:18, 5.18it/s] 12%|█▏ | 5618/48008 [19:06<2:16:07, 5.19it/s] 12%|█▏ | 5619/48008 [19:06<2:15:53, 5.20it/s] 12%|█▏ | 5620/48008 [19:06<2:14:23, 5.26it/s] 12%|█▏ | 5621/48008 [19:07<2:14:24, 5.26it/s] 12%|█▏ | 5622/48008 [19:07<2:13:33, 5.29it/s] 12%|█▏ | 5623/48008 [19:07<2:13:34, 5.29it/s] 12%|█▏ | 5624/48008 [19:07<2:14:12, 5.26it/s] 12%|█▏ | 5625/48008 [19:07<2:24:07, 4.90it/s] 12%|█▏ | 5626/48008 [19:08<2:21:19, 5.00it/s] 12%|█▏ | 5627/48008 [19:08<2:18:11, 5.11it/s] 12%|█▏ | 5628/48008 [19:08<2:18:22, 5.10it/s] 12%|█▏ | 5629/48008 [19:08<2:19:53, 5.05it/s] 12%|█▏ | 5630/48008 [19:08<2:17:39, 5.13it/s] 12%|█▏ | 5631/48008 [19:09<2:40:57, 4.39it/s] 12%|█▏ | 5632/48008 [19:09<2:33:14, 4.61it/s] 12%|█▏ | 5633/48008 [19:09<2:27:18, 4.79it/s] 12%|█▏ | 5634/48008 [19:09<2:23:23, 4.93it/s] 12%|█▏ | 5635/48008 [19:09<2:20:14, 5.04it/s] 12%|█▏ | 5636/48008 [19:10<2:18:18, 5.11it/s] 12%|█▏ | 5637/48008 [19:10<2:21:19, 5.00it/s] 12%|█▏ | 5638/48008 [19:10<2:21:34, 4.99it/s] 12%|█▏ | 5639/48008 [19:10<2:17:25, 5.14it/s] 12%|█▏ | 5640/48008 [19:10<2:18:09, 5.11it/s] 12%|█▏ | 5641/48008 [19:11<2:17:53, 5.12it/s] 12%|█▏ | 5642/48008 [19:11<2:17:15, 5.14it/s] 12%|█▏ | 5643/48008 [19:11<2:16:00, 5.19it/s] 12%|█▏ | 5644/48008 [19:11<2:15:52, 5.20it/s] 12%|█▏ | 5645/48008 [19:11<2:15:12, 5.22it/s] 12%|█▏ | 5646/48008 [19:12<2:15:41, 5.20it/s] 12%|█▏ | 5647/48008 [19:12<2:18:03, 5.11it/s] 12%|█▏ | 5648/48008 [19:12<2:17:12, 5.15it/s] 12%|█▏ | 5649/48008 [19:12<2:15:56, 5.19it/s] 12%|█▏ | 5650/48008 [19:12<2:14:34, 5.25it/s] {'loss': 0.3038, 'grad_norm': 0.65625, 'learning_rate': 0.00017646642226295617, 'epoch': 0.12} 12%|█▏ | 5650/48008 [19:12<2:14:34, 5.25it/s] 12%|█▏ | 5651/48008 [19:13<2:15:56, 5.19it/s] 12%|█▏ | 5652/48008 [19:13<2:17:58, 5.12it/s] 12%|█▏ | 5653/48008 [19:13<2:19:52, 5.05it/s] 12%|█▏ | 5654/48008 [19:13<2:19:07, 5.07it/s] 12%|█▏ | 5655/48008 [19:13<2:17:12, 5.14it/s] 12%|█▏ | 5656/48008 [19:14<2:16:21, 5.18it/s] 12%|█▏ | 5657/48008 [19:14<2:15:25, 5.21it/s] 12%|█▏ | 5658/48008 [19:14<2:14:56, 5.23it/s] 12%|█▏ | 5659/48008 [19:14<2:16:21, 5.18it/s] 12%|█▏ | 5660/48008 [19:14<2:15:41, 5.20it/s] 12%|█▏ | 5661/48008 [19:14<2:14:56, 5.23it/s] 12%|█▏ | 5662/48008 [19:15<2:15:55, 5.19it/s] 12%|█▏ | 5663/48008 [19:15<2:15:20, 5.21it/s] 12%|█▏ | 5664/48008 [19:15<2:14:57, 5.23it/s] 12%|█▏ | 5665/48008 [19:15<2:14:23, 5.25it/s] 12%|█▏ | 5666/48008 [19:15<2:13:51, 5.27it/s] 12%|█▏ | 5667/48008 [19:16<2:14:18, 5.25it/s] 12%|█▏ | 5668/48008 [19:16<2:14:19, 5.25it/s] 12%|█▏ | 5669/48008 [19:16<2:15:38, 5.20it/s] 12%|█▏ | 5670/48008 [19:16<2:17:48, 5.12it/s] 12%|█▏ | 5671/48008 [19:16<2:16:20, 5.18it/s] 12%|█▏ | 5672/48008 [19:17<2:14:46, 5.24it/s] 12%|█▏ | 5673/48008 [19:17<2:16:00, 5.19it/s] 12%|█▏ | 5674/48008 [19:17<2:15:35, 5.20it/s] 12%|█▏ | 5675/48008 [19:17<2:14:38, 5.24it/s] 12%|█▏ | 5676/48008 [19:17<2:14:08, 5.26it/s] 12%|█▏ | 5677/48008 [19:18<2:14:12, 5.26it/s] 12%|█▏ | 5678/48008 [19:18<2:16:52, 5.15it/s] 12%|█▏ | 5679/48008 [19:18<2:15:14, 5.22it/s] 12%|█▏ | 5680/48008 [19:18<2:17:04, 5.15it/s] 12%|█▏ | 5681/48008 [19:18<2:16:22, 5.17it/s] 12%|█▏ | 5682/48008 [19:19<2:15:09, 5.22it/s] 12%|█▏ | 5683/48008 [19:19<2:14:35, 5.24it/s] 12%|█▏ | 5684/48008 [19:19<2:13:09, 5.30it/s] 12%|█▏ | 5685/48008 [19:19<2:13:53, 5.27it/s] 12%|█▏ | 5686/48008 [19:19<2:15:20, 5.21it/s] 12%|█▏ | 5687/48008 [19:19<2:14:06, 5.26it/s] 12%|█▏ | 5688/48008 [19:20<2:15:16, 5.21it/s] 12%|█▏ | 5689/48008 [19:20<2:15:29, 5.21it/s] 12%|█▏ | 5690/48008 [19:20<2:15:33, 5.20it/s] 12%|█▏ | 5691/48008 [19:20<2:15:24, 5.21it/s] 12%|█▏ | 5692/48008 [19:20<2:15:09, 5.22it/s] 12%|█▏ | 5693/48008 [19:21<2:18:25, 5.09it/s] 12%|█▏ | 5694/48008 [19:21<2:17:30, 5.13it/s] 12%|█▏ | 5695/48008 [19:21<2:16:52, 5.15it/s] 12%|█▏ | 5696/48008 [19:21<2:16:00, 5.19it/s] 12%|█▏ | 5697/48008 [19:21<2:14:33, 5.24it/s] 12%|█▏ | 5698/48008 [19:22<2:14:24, 5.25it/s] 12%|█▏ | 5699/48008 [19:22<2:15:34, 5.20it/s] 12%|█▏ | 5700/48008 [19:22<2:25:25, 4.85it/s] {'loss': 0.3051, 'grad_norm': 0.1435546875, 'learning_rate': 0.00017625812364605901, 'epoch': 0.12} 12%|█▏ | 5700/48008 [19:22<2:25:25, 4.85it/s] 12%|█▏ | 5701/48008 [19:22<2:21:41, 4.98it/s] 12%|█▏ | 5702/48008 [19:22<2:19:40, 5.05it/s] 12%|█▏ | 5703/48008 [19:23<2:21:09, 4.99it/s] 12%|█▏ | 5704/48008 [19:23<2:18:12, 5.10it/s] 12%|█▏ | 5705/48008 [19:23<2:17:03, 5.14it/s] 12%|█▏ | 5706/48008 [19:23<2:15:19, 5.21it/s] 12%|█▏ | 5707/48008 [19:23<2:16:23, 5.17it/s] 12%|█▏ | 5708/48008 [19:24<2:16:00, 5.18it/s] 12%|█▏ | 5709/48008 [19:24<2:15:37, 5.20it/s] 12%|█▏ | 5710/48008 [19:24<2:15:19, 5.21it/s] 12%|█▏ | 5711/48008 [19:24<2:17:17, 5.13it/s] 12%|█▏ | 5712/48008 [19:24<2:16:17, 5.17it/s] 12%|█▏ | 5713/48008 [19:25<2:16:06, 5.18it/s] 12%|█▏ | 5714/48008 [19:25<2:16:38, 5.16it/s] 12%|█▏ | 5715/48008 [19:25<2:16:25, 5.17it/s] 12%|█▏ | 5716/48008 [19:25<2:15:31, 5.20it/s] 12%|█▏ | 5717/48008 [19:25<2:25:00, 4.86it/s] 12%|█▏ | 5718/48008 [19:26<2:21:44, 4.97it/s] 12%|█▏ | 5719/48008 [19:26<2:20:34, 5.01it/s] 12%|█▏ | 5720/48008 [19:26<2:19:09, 5.06it/s] 12%|█▏ | 5721/48008 [19:26<2:19:09, 5.06it/s] 12%|█▏ | 5722/48008 [19:26<2:17:15, 5.13it/s] 12%|█▏ | 5723/48008 [19:26<2:16:23, 5.17it/s] 12%|█▏ | 5724/48008 [19:27<2:16:26, 5.17it/s] 12%|█▏ | 5725/48008 [19:27<2:16:06, 5.18it/s] 12%|█▏ | 5726/48008 [19:27<2:16:43, 5.15it/s] 12%|█▏ | 5727/48008 [19:27<2:15:48, 5.19it/s] 12%|█▏ | 5728/48008 [19:27<2:14:52, 5.22it/s] 12%|█▏ | 5729/48008 [19:28<2:17:17, 5.13it/s] 12%|█▏ | 5730/48008 [19:28<2:19:07, 5.06it/s] 12%|█▏ | 5731/48008 [19:28<2:20:09, 5.03it/s] 12%|█▏ | 5732/48008 [19:28<2:16:31, 5.16it/s] 12%|█▏ | 5733/48008 [19:28<2:15:38, 5.19it/s] 12%|█▏ | 5734/48008 [19:29<2:15:55, 5.18it/s] 12%|█▏ | 5735/48008 [19:29<2:15:44, 5.19it/s] 12%|█▏ | 5736/48008 [19:29<2:19:45, 5.04it/s] 12%|█▏ | 5737/48008 [19:29<2:18:03, 5.10it/s] 12%|█▏ | 5738/48008 [19:29<2:19:04, 5.07it/s] 12%|█▏ | 5739/48008 [19:30<2:17:29, 5.12it/s] 12%|█▏ | 5740/48008 [19:30<2:21:29, 4.98it/s] 12%|█▏ | 5741/48008 [19:30<2:20:01, 5.03it/s] 12%|█▏ | 5742/48008 [19:30<2:17:50, 5.11it/s] 12%|█▏ | 5743/48008 [19:30<2:26:34, 4.81it/s] 12%|█▏ | 5744/48008 [19:31<2:22:37, 4.94it/s] 12%|█▏ | 5745/48008 [19:31<2:21:12, 4.99it/s] 12%|█▏ | 5746/48008 [19:31<2:20:02, 5.03it/s] 12%|█▏ | 5747/48008 [19:31<2:17:24, 5.13it/s] 12%|█▏ | 5748/48008 [19:31<2:16:46, 5.15it/s] 12%|█▏ | 5749/48008 [19:32<2:20:01, 5.03it/s] 12%|█▏ | 5750/48008 [19:32<2:19:21, 5.05it/s] {'loss': 0.3032, 'grad_norm': 0.56640625, 'learning_rate': 0.0001760498250291618, 'epoch': 0.12} 12%|█▏ | 5750/48008 [19:32<2:19:21, 5.05it/s] 12%|█▏ | 5751/48008 [19:32<2:17:59, 5.10it/s] 12%|█▏ | 5752/48008 [19:32<2:17:01, 5.14it/s] 12%|█▏ | 5753/48008 [19:32<2:20:08, 5.02it/s] 12%|█▏ | 5754/48008 [19:33<2:19:56, 5.03it/s] 12%|█▏ | 5755/48008 [19:33<2:17:39, 5.12it/s] 12%|█▏ | 5756/48008 [19:33<2:18:04, 5.10it/s] 12%|█▏ | 5757/48008 [19:33<2:17:00, 5.14it/s] 12%|█▏ | 5758/48008 [19:33<2:18:38, 5.08it/s] 12%|█▏ | 5759/48008 [19:34<2:21:06, 4.99it/s] 12%|█▏ | 5760/48008 [19:34<2:19:28, 5.05it/s] 12%|█▏ | 5761/48008 [19:34<2:17:43, 5.11it/s] 12%|█▏ | 5762/48008 [19:34<2:16:36, 5.15it/s] 12%|█▏ | 5763/48008 [19:34<2:15:07, 5.21it/s] 12%|█▏ | 5764/48008 [19:35<2:16:51, 5.14it/s] 12%|█▏ | 5765/48008 [19:35<2:15:44, 5.19it/s] 12%|█▏ | 5766/48008 [19:35<2:14:20, 5.24it/s] 12%|█▏ | 5767/48008 [19:35<2:14:28, 5.24it/s] 12%|█▏ | 5768/48008 [19:35<2:14:25, 5.24it/s] 12%|█▏ | 5769/48008 [19:35<2:13:09, 5.29it/s] 12%|█▏ | 5770/48008 [19:36<2:15:00, 5.21it/s] 12%|█▏ | 5771/48008 [19:36<2:13:20, 5.28it/s] 12%|█▏ | 5772/48008 [19:36<2:16:19, 5.16it/s] 12%|█▏ | 5773/48008 [19:36<2:16:35, 5.15it/s] 12%|█▏ | 5774/48008 [19:36<2:15:56, 5.18it/s] 12%|█▏ | 5775/48008 [19:37<2:16:00, 5.18it/s] 12%|█▏ | 5776/48008 [19:37<2:15:24, 5.20it/s] 12%|█▏ | 5777/48008 [19:37<2:15:19, 5.20it/s] 12%|█▏ | 5778/48008 [19:37<2:39:32, 4.41it/s] 12%|█▏ | 5779/48008 [19:38<2:31:54, 4.63it/s] 12%|█▏ | 5780/48008 [19:38<2:50:47, 4.12it/s] 12%|█▏ | 5781/48008 [19:38<2:39:57, 4.40it/s] 12%|█▏ | 5782/48008 [19:38<2:34:17, 4.56it/s] 12%|█▏ | 5783/48008 [19:38<2:28:54, 4.73it/s] 12%|█▏ | 5784/48008 [19:39<2:25:43, 4.83it/s] 12%|█▏ | 5785/48008 [19:39<2:23:40, 4.90it/s] 12%|█▏ | 5786/48008 [19:39<2:21:37, 4.97it/s] 12%|█▏ | 5787/48008 [19:39<2:19:21, 5.05it/s] 12%|█▏ | 5788/48008 [19:39<2:17:58, 5.10it/s] 12%|█▏ | 5789/48008 [19:40<2:17:01, 5.14it/s] 12%|█▏ | 5790/48008 [19:40<2:14:59, 5.21it/s] 12%|█▏ | 5791/48008 [19:40<2:14:49, 5.22it/s] 12%|█▏ | 5792/48008 [19:40<2:17:14, 5.13it/s] 12%|█▏ | 5793/48008 [19:40<2:18:28, 5.08it/s] 12%|█▏ | 5794/48008 [19:41<2:27:29, 4.77it/s] 12%|█▏ | 5795/48008 [19:41<2:22:40, 4.93it/s] 12%|█▏ | 5796/48008 [19:41<2:19:52, 5.03it/s] 12%|█▏ | 5797/48008 [19:41<2:59:31, 3.92it/s] 12%|█▏ | 5798/48008 [19:42<2:46:11, 4.23it/s] 12%|█▏ | 5799/48008 [19:42<2:37:51, 4.46it/s] 12%|█▏ | 5800/48008 [19:42<2:30:56, 4.66it/s] {'loss': 0.3028, 'grad_norm': 0.2138671875, 'learning_rate': 0.00017584152641226462, 'epoch': 0.12} 12%|█▏ | 5800/48008 [19:42<2:30:56, 4.66it/s] 12%|█▏ | 5801/48008 [19:42<2:25:29, 4.84it/s] 12%|█▏ | 5802/48008 [19:42<2:22:38, 4.93it/s] 12%|█▏ | 5803/48008 [19:43<2:19:39, 5.04it/s] 12%|█▏ | 5804/48008 [19:43<2:18:33, 5.08it/s] 12%|█▏ | 5805/48008 [19:43<2:16:46, 5.14it/s] 12%|█▏ | 5806/48008 [19:43<2:18:26, 5.08it/s] 12%|█▏ | 5807/48008 [19:43<2:17:34, 5.11it/s] 12%|█▏ | 5808/48008 [19:43<2:16:32, 5.15it/s] 12%|█▏ | 5809/48008 [19:44<2:15:57, 5.17it/s] 12%|█▏ | 5810/48008 [19:44<2:15:35, 5.19it/s] 12%|█▏ | 5811/48008 [19:44<2:15:23, 5.19it/s] 12%|█▏ | 5812/48008 [19:44<2:16:22, 5.16it/s] 12%|█▏ | 5813/48008 [19:44<2:15:49, 5.18it/s] 12%|█▏ | 5814/48008 [19:45<2:19:07, 5.05it/s] 12%|█▏ | 5815/48008 [19:45<2:17:30, 5.11it/s] 12%|█▏ | 5816/48008 [19:45<2:16:22, 5.16it/s] 12%|█▏ | 5817/48008 [19:45<2:39:57, 4.40it/s] 12%|█▏ | 5818/48008 [19:46<2:34:22, 4.55it/s] 12%|█▏ | 5819/48008 [19:46<2:27:28, 4.77it/s] 12%|█▏ | 5820/48008 [19:46<2:47:50, 4.19it/s] 12%|█▏ | 5821/48008 [19:46<2:38:54, 4.42it/s] 12%|█▏ | 5822/48008 [19:46<2:31:27, 4.64it/s] 12%|█▏ | 5823/48008 [19:47<2:25:16, 4.84it/s] 12%|█▏ | 5824/48008 [19:47<2:19:26, 5.04it/s] 12%|█▏ | 5825/48008 [19:47<2:16:48, 5.14it/s] 12%|█▏ | 5826/48008 [19:47<2:25:58, 4.82it/s] 12%|█▏ | 5827/48008 [19:47<2:23:55, 4.88it/s] 12%|█▏ | 5828/48008 [19:48<2:21:06, 4.98it/s] 12%|█▏ | 5829/48008 [19:48<2:18:55, 5.06it/s] 12%|█▏ | 5830/48008 [19:48<2:18:24, 5.08it/s] 12%|█▏ | 5831/48008 [19:48<2:19:35, 5.04it/s] 12%|█▏ | 5832/48008 [19:48<2:17:27, 5.11it/s] 12%|█▏ | 5833/48008 [19:49<2:15:56, 5.17it/s] 12%|█▏ | 5834/48008 [19:49<2:14:57, 5.21it/s] 12%|█▏ | 5835/48008 [19:49<2:14:57, 5.21it/s] 12%|█▏ | 5836/48008 [19:49<2:14:25, 5.23it/s] 12%|█▏ | 5837/48008 [19:49<2:38:32, 4.43it/s] 12%|█▏ | 5838/48008 [19:50<2:30:10, 4.68it/s] 12%|█▏ | 5839/48008 [19:50<2:24:05, 4.88it/s] 12%|█▏ | 5840/48008 [19:50<2:22:30, 4.93it/s] 12%|█▏ | 5841/48008 [19:50<3:00:59, 3.88it/s] 12%|█▏ | 5842/48008 [19:51<2:46:56, 4.21it/s] 12%|█▏ | 5843/48008 [19:51<2:35:58, 4.51it/s] 12%|█▏ | 5844/48008 [19:51<2:29:26, 4.70it/s] 12%|█▏ | 5845/48008 [19:51<2:26:15, 4.80it/s] 12%|█▏ | 5846/48008 [19:51<2:21:37, 4.96it/s] 12%|█▏ | 5847/48008 [19:52<2:22:04, 4.95it/s] 12%|█▏ | 5848/48008 [19:52<2:19:33, 5.03it/s] 12%|█▏ | 5849/48008 [19:52<2:17:59, 5.09it/s] 12%|█▏ | 5850/48008 [19:52<2:17:02, 5.13it/s] {'loss': 0.3037, 'grad_norm': 0.2265625, 'learning_rate': 0.00017563322779536744, 'epoch': 0.12} 12%|█▏ | 5850/48008 [19:52<2:17:02, 5.13it/s] 12%|█▏ | 5851/48008 [19:53<2:57:25, 3.96it/s] 12%|█▏ | 5852/48008 [19:53<2:46:26, 4.22it/s] 12%|█▏ | 5853/48008 [19:53<2:36:16, 4.50it/s] 12%|█▏ | 5854/48008 [19:53<2:29:32, 4.70it/s] 12%|█▏ | 5855/48008 [19:53<2:24:26, 4.86it/s] 12%|█▏ | 5856/48008 [19:54<2:31:06, 4.65it/s] 12%|█▏ | 5857/48008 [19:54<2:25:53, 4.82it/s] 12%|█▏ | 5858/48008 [19:54<2:24:31, 4.86it/s] 12%|█▏ | 5859/48008 [19:54<2:25:02, 4.84it/s] 12%|█▏ | 5860/48008 [19:54<2:23:57, 4.88it/s] 12%|█▏ | 5861/48008 [19:55<2:22:38, 4.92it/s] 12%|█▏ | 5862/48008 [19:55<2:21:36, 4.96it/s] 12%|█▏ | 5863/48008 [19:55<2:19:30, 5.04it/s] 12%|█▏ | 5864/48008 [19:55<2:27:30, 4.76it/s] 12%|█▏ | 5865/48008 [19:55<2:23:21, 4.90it/s] 12%|█▏ | 5866/48008 [19:56<2:20:18, 5.01it/s] 12%|█▏ | 5867/48008 [19:56<2:18:00, 5.09it/s] 12%|█▏ | 5868/48008 [19:56<2:17:10, 5.12it/s] 12%|█▏ | 5869/48008 [19:56<2:15:23, 5.19it/s] 12%|█▏ | 5870/48008 [19:56<2:15:22, 5.19it/s] 12%|█▏ | 5871/48008 [19:56<2:13:58, 5.24it/s] 12%|█▏ | 5872/48008 [19:57<2:15:21, 5.19it/s] 12%|█▏ | 5873/48008 [19:57<2:15:10, 5.20it/s] 12%|█▏ | 5874/48008 [19:57<2:16:48, 5.13it/s] 12%|█▏ | 5875/48008 [19:57<2:17:20, 5.11it/s] 12%|█▏ | 5876/48008 [19:57<2:15:43, 5.17it/s] 12%|█▏ | 5877/48008 [19:58<2:17:34, 5.10it/s] 12%|█▏ | 5878/48008 [19:58<2:20:01, 5.01it/s] 12%|█▏ | 5879/48008 [19:58<2:16:44, 5.13it/s] 12%|█▏ | 5880/48008 [19:58<2:18:20, 5.08it/s] 12%|█▏ | 5881/48008 [19:58<2:16:55, 5.13it/s] 12%|█▏ | 5882/48008 [19:59<2:14:43, 5.21it/s] 12%|█▏ | 5883/48008 [19:59<2:13:18, 5.27it/s] 12%|█▏ | 5884/48008 [19:59<2:14:17, 5.23it/s] 12%|█▏ | 5885/48008 [19:59<2:24:02, 4.87it/s] 12%|█▏ | 5886/48008 [19:59<2:21:29, 4.96it/s] 12%|█▏ | 5887/48008 [20:00<2:23:29, 4.89it/s] 12%|█▏ | 5888/48008 [20:00<2:20:28, 5.00it/s] 12%|█▏ | 5889/48008 [20:00<2:21:10, 4.97it/s] 12%|█▏ | 5890/48008 [20:00<2:18:53, 5.05it/s] 12%|█▏ | 5891/48008 [20:00<2:16:26, 5.14it/s] 12%|█▏ | 5892/48008 [20:01<2:15:55, 5.16it/s] 12%|█▏ | 5893/48008 [20:01<2:25:26, 4.83it/s] 12%|█▏ | 5894/48008 [20:01<2:20:55, 4.98it/s] 12%|█▏ | 5895/48008 [20:01<2:18:29, 5.07it/s] 12%|█▏ | 5896/48008 [20:01<2:16:33, 5.14it/s] 12%|█▏ | 5897/48008 [20:02<2:18:06, 5.08it/s] 12%|█▏ | 5898/48008 [20:02<2:16:42, 5.13it/s] 12%|█▏ | 5899/48008 [20:02<2:16:09, 5.15it/s] 12%|█▏ | 5900/48008 [20:02<2:15:38, 5.17it/s] {'loss': 0.3013, 'grad_norm': 0.63671875, 'learning_rate': 0.00017542492917847026, 'epoch': 0.12} 12%|█▏ | 5900/48008 [20:02<2:15:38, 5.17it/s] 12%|█▏ | 5901/48008 [20:02<2:17:39, 5.10it/s] 12%|█▏ | 5902/48008 [20:03<2:16:23, 5.15it/s] 12%|█▏ | 5903/48008 [20:03<2:15:31, 5.18it/s] 12%|█▏ | 5904/48008 [20:03<2:14:53, 5.20it/s] 12%|█▏ | 5905/48008 [20:03<2:16:51, 5.13it/s] 12%|█▏ | 5906/48008 [20:03<2:18:05, 5.08it/s] 12%|█▏ | 5907/48008 [20:04<2:26:26, 4.79it/s] 12%|█▏ | 5908/48008 [20:04<2:22:44, 4.92it/s] 12%|█▏ | 5909/48008 [20:04<2:20:13, 5.00it/s] 12%|█▏ | 5910/48008 [20:04<2:18:29, 5.07it/s] 12%|█▏ | 5911/48008 [20:04<2:16:52, 5.13it/s] 12%|█▏ | 5912/48008 [20:05<2:18:11, 5.08it/s] 12%|█▏ | 5913/48008 [20:05<2:17:40, 5.10it/s] 12%|█▏ | 5914/48008 [20:05<2:18:39, 5.06it/s] 12%|█▏ | 5915/48008 [20:05<2:16:38, 5.13it/s] 12%|█▏ | 5916/48008 [20:05<2:17:59, 5.08it/s] 12%|█▏ | 5917/48008 [20:06<2:17:19, 5.11it/s] 12%|█▏ | 5918/48008 [20:06<2:15:26, 5.18it/s] 12%|█▏ | 5919/48008 [20:06<2:14:22, 5.22it/s] 12%|█▏ | 5920/48008 [20:06<2:13:46, 5.24it/s] 12%|█▏ | 5921/48008 [20:06<2:13:10, 5.27it/s] 12%|█▏ | 5922/48008 [20:06<2:12:51, 5.28it/s] 12%|█▏ | 5923/48008 [20:07<2:12:28, 5.29it/s] 12%|█▏ | 5924/48008 [20:07<2:22:48, 4.91it/s] 12%|█▏ | 5925/48008 [20:07<2:19:23, 5.03it/s] 12%|█▏ | 5926/48008 [20:07<2:18:21, 5.07it/s] 12%|█▏ | 5927/48008 [20:07<2:16:35, 5.13it/s] 12%|█▏ | 5928/48008 [20:08<2:15:05, 5.19it/s] 12%|█▏ | 5929/48008 [20:08<2:17:01, 5.12it/s] 12%|█▏ | 5930/48008 [20:08<2:16:10, 5.15it/s] 12%|█▏ | 5931/48008 [20:08<2:14:24, 5.22it/s] 12%|█▏ | 5932/48008 [20:08<2:14:35, 5.21it/s] 12%|█▏ | 5933/48008 [20:09<2:18:10, 5.07it/s] 12%|█▏ | 5934/48008 [20:09<2:16:46, 5.13it/s] 12%|█▏ | 5935/48008 [20:09<2:15:52, 5.16it/s] 12%|█▏ | 5936/48008 [20:09<2:17:59, 5.08it/s] 12%|█▏ | 5937/48008 [20:09<2:18:15, 5.07it/s] 12%|█▏ | 5938/48008 [20:10<2:19:11, 5.04it/s] 12%|█▏ | 5939/48008 [20:10<2:19:25, 5.03it/s] 12%|█▏ | 5940/48008 [20:10<2:17:17, 5.11it/s] 12%|█▏ | 5941/48008 [20:10<2:17:37, 5.09it/s] 12%|█▏ | 5942/48008 [20:10<2:19:01, 5.04it/s] 12%|█▏ | 5943/48008 [20:11<2:17:37, 5.09it/s] 12%|█▏ | 5944/48008 [20:11<2:18:52, 5.05it/s] 12%|█▏ | 5945/48008 [20:11<2:17:01, 5.12it/s] 12%|█▏ | 5946/48008 [20:11<2:16:35, 5.13it/s] 12%|█▏ | 5947/48008 [20:11<2:14:48, 5.20it/s] 12%|█▏ | 5948/48008 [20:12<2:14:09, 5.22it/s] 12%|█▏ | 5949/48008 [20:12<2:16:41, 5.13it/s] 12%|█▏ | 5950/48008 [20:12<2:15:39, 5.17it/s] {'loss': 0.3028, 'grad_norm': 0.4453125, 'learning_rate': 0.00017521663056157307, 'epoch': 0.12} 12%|█▏ | 5950/48008 [20:12<2:15:39, 5.17it/s] 12%|█▏ | 5951/48008 [20:12<2:15:09, 5.19it/s] 12%|█▏ | 5952/48008 [20:12<2:24:09, 4.86it/s] 12%|█▏ | 5953/48008 [20:13<2:20:06, 5.00it/s] 12%|█▏ | 5954/48008 [20:13<2:17:11, 5.11it/s] 12%|█▏ | 5955/48008 [20:13<2:15:27, 5.17it/s] 12%|█▏ | 5956/48008 [20:13<2:16:34, 5.13it/s] 12%|█▏ | 5957/48008 [20:14<2:56:41, 3.97it/s] 12%|█▏ | 5958/48008 [20:14<2:46:22, 4.21it/s] 12%|█▏ | 5959/48008 [20:14<2:38:57, 4.41it/s] 12%|█▏ | 5960/48008 [20:14<2:31:15, 4.63it/s] 12%|█▏ | 5961/48008 [20:14<2:25:45, 4.81it/s] 12%|█▏ | 5962/48008 [20:14<2:22:13, 4.93it/s] 12%|█▏ | 5963/48008 [20:15<2:19:39, 5.02it/s] 12%|█▏ | 5964/48008 [20:15<2:27:15, 4.76it/s] 12%|█▏ | 5965/48008 [20:15<2:23:19, 4.89it/s] 12%|█▏ | 5966/48008 [20:16<3:01:27, 3.86it/s] 12%|█▏ | 5967/48008 [20:16<2:47:24, 4.19it/s] 12%|█▏ | 5968/48008 [20:16<2:36:35, 4.47it/s] 12%|█▏ | 5969/48008 [20:16<2:28:27, 4.72it/s] 12%|█▏ | 5970/48008 [20:16<2:26:41, 4.78it/s] 12%|█▏ | 5971/48008 [20:16<2:22:43, 4.91it/s] 12%|█▏ | 5972/48008 [20:17<2:19:57, 5.01it/s] 12%|█▏ | 5973/48008 [20:17<2:16:40, 5.13it/s] 12%|█▏ | 5974/48008 [20:17<2:15:08, 5.18it/s] 12%|█▏ | 5975/48008 [20:17<2:56:09, 3.98it/s] 12%|█▏ | 5976/48008 [20:18<2:43:40, 4.28it/s] 12%|█▏ | 5977/48008 [20:18<2:36:09, 4.49it/s] 12%|█▏ | 5978/48008 [20:18<2:29:30, 4.69it/s] 12%|█▏ | 5979/48008 [20:18<2:24:40, 4.84it/s] 12%|█▏ | 5980/48008 [20:18<2:21:26, 4.95it/s] 12%|█▏ | 5981/48008 [20:19<2:18:58, 5.04it/s] 12%|█▏ | 5982/48008 [20:19<2:16:46, 5.12it/s] 12%|█▏ | 5983/48008 [20:19<2:15:59, 5.15it/s] 12%|█▏ | 5984/48008 [20:19<2:17:34, 5.09it/s] 12%|█▏ | 5985/48008 [20:19<2:15:34, 5.17it/s] 12%|█▏ | 5986/48008 [20:20<2:16:05, 5.15it/s] 12%|█▏ | 5987/48008 [20:20<2:15:02, 5.19it/s] 12%|█▏ | 5988/48008 [20:20<2:14:38, 5.20it/s] 12%|█▏ | 5989/48008 [20:20<2:15:52, 5.15it/s] 12%|█▏ | 5990/48008 [20:20<2:14:48, 5.20it/s] 12%|█▏ | 5991/48008 [20:20<2:14:50, 5.19it/s] 12%|█▏ | 5992/48008 [20:21<2:14:09, 5.22it/s] 12%|█▏ | 5993/48008 [20:21<2:18:15, 5.06it/s] 12%|█▏ | 5994/48008 [20:21<2:16:56, 5.11it/s] 12%|█▏ | 5995/48008 [20:21<2:15:57, 5.15it/s] 12%|█▏ | 5996/48008 [20:21<2:14:21, 5.21it/s] 12%|█▏ | 5997/48008 [20:22<2:13:54, 5.23it/s] 12%|█▏ | 5998/48008 [20:22<2:12:33, 5.28it/s] 12%|█▏ | 5999/48008 [20:22<2:12:41, 5.28it/s] 12%|█▏ | 6000/48008 [20:22<2:11:57, 5.31it/s] {'loss': 0.3043, 'grad_norm': 1.0703125, 'learning_rate': 0.0001750083319446759, 'epoch': 0.12} 12%|█▏ | 6000/48008 [20:22<2:11:57, 5.31it/s] 12%|█▎ | 6001/48008 [20:22<2:12:03, 5.30it/s] 13%|█▎ | 6002/48008 [20:23<2:12:09, 5.30it/s] 13%|█▎ | 6003/48008 [20:23<2:12:30, 5.28it/s] 13%|█▎ | 6004/48008 [20:23<2:16:53, 5.11it/s] 13%|█▎ | 6005/48008 [20:23<2:16:51, 5.12it/s] 13%|█▎ | 6006/48008 [20:23<2:15:51, 5.15it/s] 13%|█▎ | 6007/48008 [20:24<2:17:33, 5.09it/s] 13%|█▎ | 6008/48008 [20:24<2:17:17, 5.10it/s] 13%|█▎ | 6009/48008 [20:24<2:26:12, 4.79it/s] 13%|█▎ | 6010/48008 [20:24<2:22:31, 4.91it/s] 13%|█▎ | 6011/48008 [20:24<2:29:27, 4.68it/s] 13%|█▎ | 6012/48008 [20:25<2:26:49, 4.77it/s] 13%|█▎ | 6013/48008 [20:25<2:22:55, 4.90it/s] 13%|█▎ | 6014/48008 [20:25<2:19:49, 5.01it/s] 13%|█▎ | 6015/48008 [20:25<2:17:26, 5.09it/s] 13%|█▎ | 6016/48008 [20:25<2:15:57, 5.15it/s] 13%|█▎ | 6017/48008 [20:26<2:14:20, 5.21it/s] 13%|█▎ | 6018/48008 [20:26<2:14:05, 5.22it/s] 13%|█▎ | 6019/48008 [20:26<2:14:09, 5.22it/s] 13%|█▎ | 6020/48008 [20:26<2:17:50, 5.08it/s] 13%|█▎ | 6021/48008 [20:27<2:57:31, 3.94it/s] 13%|█▎ | 6022/48008 [20:27<2:46:33, 4.20it/s] 13%|█▎ | 6023/48008 [20:27<2:46:05, 4.21it/s] 13%|█▎ | 6024/48008 [20:27<2:36:32, 4.47it/s] 13%|█▎ | 6025/48008 [20:28<3:10:39, 3.67it/s] 13%|█▎ | 6026/48008 [20:28<2:53:16, 4.04it/s] 13%|█▎ | 6027/48008 [20:28<2:40:33, 4.36it/s] 13%|█▎ | 6028/48008 [20:28<2:34:45, 4.52it/s] 13%|█▎ | 6029/48008 [20:28<2:31:01, 4.63it/s] 13%|█▎ | 6030/48008 [20:29<2:24:28, 4.84it/s] 13%|█▎ | 6031/48008 [20:29<2:20:47, 4.97it/s] 13%|█▎ | 6032/48008 [20:29<2:18:58, 5.03it/s] 13%|█▎ | 6033/48008 [20:29<2:17:09, 5.10it/s] 13%|█▎ | 6034/48008 [20:29<2:17:39, 5.08it/s] 13%|█▎ | 6035/48008 [20:30<2:16:50, 5.11it/s] 13%|█▎ | 6036/48008 [20:30<2:15:55, 5.15it/s] 13%|█▎ | 6037/48008 [20:30<2:17:27, 5.09it/s] 13%|█▎ | 6038/48008 [20:30<2:16:45, 5.11it/s] 13%|█▎ | 6039/48008 [20:30<2:39:48, 4.38it/s] 13%|█▎ | 6040/48008 [20:31<2:31:28, 4.62it/s] 13%|█▎ | 6041/48008 [20:31<2:27:03, 4.76it/s] 13%|█▎ | 6042/48008 [20:31<2:23:57, 4.86it/s] 13%|█▎ | 6043/48008 [20:31<3:01:45, 3.85it/s] 13%|█▎ | 6044/48008 [20:32<2:46:57, 4.19it/s] 13%|█▎ | 6045/48008 [20:32<3:17:39, 3.54it/s] 13%|█▎ | 6046/48008 [20:32<2:56:49, 3.96it/s] 13%|█▎ | 6047/48008 [20:32<2:45:28, 4.23it/s] 13%|█▎ | 6048/48008 [20:33<2:36:14, 4.48it/s] 13%|█▎ | 6049/48008 [20:33<2:28:37, 4.71it/s] 13%|█▎ | 6050/48008 [20:33<2:24:01, 4.86it/s] {'loss': 0.2987, 'grad_norm': 0.55859375, 'learning_rate': 0.00017480003332777874, 'epoch': 0.13} 13%|█▎ | 6050/48008 [20:33<2:24:01, 4.86it/s] 13%|█▎ | 6051/48008 [20:33<2:21:28, 4.94it/s] 13%|█▎ | 6052/48008 [20:33<2:21:14, 4.95it/s] 13%|█▎ | 6053/48008 [20:33<2:18:34, 5.05it/s] 13%|█▎ | 6054/48008 [20:34<2:18:14, 5.06it/s] 13%|█▎ | 6055/48008 [20:34<2:19:22, 5.02it/s] 13%|█▎ | 6056/48008 [20:34<2:17:58, 5.07it/s] 13%|█▎ | 6057/48008 [20:34<2:16:18, 5.13it/s] 13%|█▎ | 6058/48008 [20:34<2:17:55, 5.07it/s] 13%|█▎ | 6059/48008 [20:35<2:20:09, 4.99it/s] 13%|█▎ | 6060/48008 [20:35<2:17:37, 5.08it/s] 13%|█▎ | 6061/48008 [20:35<2:26:21, 4.78it/s] 13%|█▎ | 6062/48008 [20:35<2:21:17, 4.95it/s] 13%|█▎ | 6063/48008 [20:35<2:17:52, 5.07it/s] 13%|█▎ | 6064/48008 [20:36<2:16:39, 5.12it/s] 13%|█▎ | 6065/48008 [20:36<2:15:00, 5.18it/s] 13%|█▎ | 6066/48008 [20:36<2:18:35, 5.04it/s] 13%|█▎ | 6067/48008 [20:36<2:19:30, 5.01it/s] 13%|█▎ | 6068/48008 [20:36<2:27:09, 4.75it/s] 13%|█▎ | 6069/48008 [20:37<2:46:56, 4.19it/s] 13%|█▎ | 6070/48008 [20:37<2:36:31, 4.47it/s] 13%|█▎ | 6071/48008 [20:37<2:29:04, 4.69it/s] 13%|█▎ | 6072/48008 [20:37<2:24:09, 4.85it/s] 13%|█▎ | 6073/48008 [20:38<2:20:55, 4.96it/s] 13%|█▎ | 6074/48008 [20:38<2:18:24, 5.05it/s] 13%|█▎ | 6075/48008 [20:38<2:16:03, 5.14it/s] 13%|█▎ | 6076/48008 [20:38<2:15:51, 5.14it/s] 13%|█▎ | 6077/48008 [20:38<2:15:51, 5.14it/s] 13%|█▎ | 6078/48008 [20:39<2:14:44, 5.19it/s] 13%|█▎ | 6079/48008 [20:39<2:14:45, 5.19it/s] 13%|█▎ | 6080/48008 [20:39<2:14:13, 5.21it/s] 13%|█▎ | 6081/48008 [20:39<2:13:32, 5.23it/s] 13%|█▎ | 6082/48008 [20:39<2:13:08, 5.25it/s] 13%|█▎ | 6083/48008 [20:39<2:13:31, 5.23it/s] 13%|█▎ | 6084/48008 [20:40<2:13:37, 5.23it/s] 13%|█▎ | 6085/48008 [20:40<2:13:38, 5.23it/s] 13%|█▎ | 6086/48008 [20:40<2:17:52, 5.07it/s] 13%|█▎ | 6087/48008 [20:40<2:15:48, 5.14it/s] 13%|█▎ | 6088/48008 [20:40<2:24:47, 4.83it/s] 13%|█▎ | 6089/48008 [20:41<2:20:59, 4.96it/s] 13%|█▎ | 6090/48008 [20:41<2:17:40, 5.07it/s] 13%|█▎ | 6091/48008 [20:41<2:16:15, 5.13it/s] 13%|█▎ | 6092/48008 [20:41<2:15:21, 5.16it/s] 13%|█▎ | 6093/48008 [20:41<2:24:29, 4.83it/s] 13%|█▎ | 6094/48008 [20:42<2:21:11, 4.95it/s] 13%|█▎ | 6095/48008 [20:42<2:18:08, 5.06it/s] 13%|█▎ | 6096/48008 [20:42<2:19:00, 5.03it/s] 13%|█▎ | 6097/48008 [20:42<2:18:18, 5.05it/s] 13%|█▎ | 6098/48008 [20:42<2:16:43, 5.11it/s] 13%|█▎ | 6099/48008 [20:43<2:56:28, 3.96it/s] 13%|█▎ | 6100/48008 [20:43<2:42:43, 4.29it/s] {'loss': 0.3042, 'grad_norm': 0.15234375, 'learning_rate': 0.00017459173471088153, 'epoch': 0.13} 13%|█▎ | 6100/48008 [20:43<2:42:43, 4.29it/s] 13%|█▎ | 6101/48008 [20:43<2:32:40, 4.57it/s] 13%|█▎ | 6102/48008 [20:43<2:29:17, 4.68it/s] 13%|█▎ | 6103/48008 [20:44<2:26:47, 4.76it/s] 13%|█▎ | 6104/48008 [20:44<2:22:32, 4.90it/s] 13%|█▎ | 6105/48008 [20:44<2:29:31, 4.67it/s] 13%|█▎ | 6106/48008 [20:44<2:23:48, 4.86it/s] 13%|█▎ | 6107/48008 [20:44<2:20:55, 4.96it/s] 13%|█▎ | 6108/48008 [20:45<2:18:30, 5.04it/s] 13%|█▎ | 6109/48008 [20:45<2:16:31, 5.11it/s] 13%|█▎ | 6110/48008 [20:45<2:16:40, 5.11it/s] 13%|█▎ | 6111/48008 [20:45<2:16:14, 5.13it/s] 13%|█▎ | 6112/48008 [20:45<2:14:47, 5.18it/s] 13%|█▎ | 6113/48008 [20:46<2:14:26, 5.19it/s] 13%|█▎ | 6114/48008 [20:46<2:14:29, 5.19it/s] 13%|█▎ | 6115/48008 [20:46<2:13:52, 5.22it/s] 13%|█▎ | 6116/48008 [20:46<2:13:17, 5.24it/s] 13%|█▎ | 6117/48008 [20:46<2:13:09, 5.24it/s] 13%|█▎ | 6118/48008 [20:47<2:12:41, 5.26it/s] 13%|█▎ | 6119/48008 [20:47<2:13:38, 5.22it/s] 13%|█▎ | 6120/48008 [20:47<2:13:26, 5.23it/s] 13%|█▎ | 6121/48008 [20:47<2:12:30, 5.27it/s] 13%|█▎ | 6122/48008 [20:47<2:12:40, 5.26it/s] 13%|█▎ | 6123/48008 [20:47<2:14:20, 5.20it/s] 13%|█▎ | 6124/48008 [20:48<2:15:03, 5.17it/s] 13%|█▎ | 6125/48008 [20:48<2:14:10, 5.20it/s] 13%|█▎ | 6126/48008 [20:48<2:13:39, 5.22it/s] 13%|█▎ | 6127/48008 [20:48<2:13:20, 5.24it/s] 13%|█▎ | 6128/48008 [20:48<2:14:08, 5.20it/s] 13%|█▎ | 6129/48008 [20:49<2:38:03, 4.42it/s] 13%|█▎ | 6130/48008 [20:49<2:31:26, 4.61it/s] 13%|█▎ | 6131/48008 [20:49<2:26:04, 4.78it/s] 13%|█▎ | 6132/48008 [20:49<2:21:35, 4.93it/s] 13%|█▎ | 6133/48008 [20:50<2:19:14, 5.01it/s] 13%|█▎ | 6134/48008 [20:50<2:17:45, 5.07it/s] 13%|█▎ | 6135/48008 [20:50<2:26:14, 4.77it/s] 13%|█▎ | 6136/48008 [20:50<2:23:46, 4.85it/s] 13%|█▎ | 6137/48008 [20:50<2:19:59, 4.98it/s] 13%|█▎ | 6138/48008 [20:51<2:20:26, 4.97it/s] 13%|█▎ | 6139/48008 [20:51<2:18:13, 5.05it/s] 13%|█▎ | 6140/48008 [20:51<2:16:38, 5.11it/s] 13%|█▎ | 6141/48008 [20:51<2:14:57, 5.17it/s] 13%|█▎ | 6142/48008 [20:51<2:14:19, 5.19it/s] 13%|█▎ | 6143/48008 [20:52<2:38:15, 4.41it/s] 13%|█▎ | 6144/48008 [20:52<2:30:37, 4.63it/s] 13%|█▎ | 6145/48008 [20:52<2:27:01, 4.75it/s] 13%|█▎ | 6146/48008 [20:52<2:22:08, 4.91it/s] 13%|█▎ | 6147/48008 [20:52<2:19:06, 5.02it/s] 13%|█▎ | 6148/48008 [20:53<2:17:32, 5.07it/s] 13%|█▎ | 6149/48008 [20:53<2:15:02, 5.17it/s] 13%|█▎ | 6150/48008 [20:53<2:14:10, 5.20it/s] {'loss': 0.3025, 'grad_norm': 0.1396484375, 'learning_rate': 0.00017438343609398434, 'epoch': 0.13} 13%|█▎ | 6150/48008 [20:53<2:14:10, 5.20it/s] 13%|█▎ | 6151/48008 [20:53<2:14:14, 5.20it/s] 13%|█▎ | 6152/48008 [20:53<2:12:54, 5.25it/s] 13%|█▎ | 6153/48008 [20:54<2:53:40, 4.02it/s] 13%|█▎ | 6154/48008 [20:54<2:43:01, 4.28it/s] 13%|█▎ | 6155/48008 [20:54<2:33:38, 4.54it/s] 13%|█▎ | 6156/48008 [20:54<2:28:56, 4.68it/s] 13%|█▎ | 6157/48008 [20:54<2:24:00, 4.84it/s] 13%|█▎ | 6158/48008 [20:55<2:21:54, 4.91it/s] 13%|█▎ | 6159/48008 [20:55<2:20:10, 4.98it/s] 13%|█▎ | 6160/48008 [20:55<2:18:50, 5.02it/s] 13%|█▎ | 6161/48008 [20:55<2:17:17, 5.08it/s] 13%|█▎ | 6162/48008 [20:55<2:16:55, 5.09it/s] 13%|█▎ | 6163/48008 [20:56<2:15:18, 5.15it/s] 13%|█▎ | 6164/48008 [20:56<2:18:37, 5.03it/s] 13%|█▎ | 6165/48008 [20:56<2:19:05, 5.01it/s] 13%|█▎ | 6166/48008 [20:56<2:16:46, 5.10it/s] 13%|█▎ | 6167/48008 [20:56<2:15:15, 5.16it/s] 13%|█▎ | 6168/48008 [20:57<2:14:11, 5.20it/s] 13%|█▎ | 6169/48008 [20:57<2:14:13, 5.20it/s] 13%|█▎ | 6170/48008 [20:57<2:14:00, 5.20it/s] 13%|█▎ | 6171/48008 [20:57<2:13:12, 5.23it/s] 13%|█▎ | 6172/48008 [20:57<2:13:23, 5.23it/s] 13%|█▎ | 6173/48008 [20:58<2:12:31, 5.26it/s] 13%|█▎ | 6174/48008 [20:58<2:12:46, 5.25it/s] 13%|█▎ | 6175/48008 [20:58<2:12:01, 5.28it/s] 13%|█▎ | 6176/48008 [20:58<2:21:47, 4.92it/s] 13%|█▎ | 6177/48008 [20:58<2:18:05, 5.05it/s] 13%|█▎ | 6178/48008 [20:59<2:16:41, 5.10it/s] 13%|█▎ | 6179/48008 [20:59<2:17:17, 5.08it/s] 13%|█▎ | 6180/48008 [20:59<2:15:00, 5.16it/s] 13%|█▎ | 6181/48008 [20:59<2:14:38, 5.18it/s] 13%|█▎ | 6182/48008 [20:59<2:14:09, 5.20it/s] 13%|█▎ | 6183/48008 [21:00<2:15:15, 5.15it/s] 13%|█▎ | 6184/48008 [21:00<2:14:31, 5.18it/s] 13%|█▎ | 6185/48008 [21:00<2:13:15, 5.23it/s] 13%|█▎ | 6186/48008 [21:00<2:12:11, 5.27it/s] 13%|█▎ | 6187/48008 [21:00<2:14:51, 5.17it/s] 13%|█▎ | 6188/48008 [21:00<2:15:05, 5.16it/s] 13%|█▎ | 6189/48008 [21:01<2:14:27, 5.18it/s] 13%|█▎ | 6190/48008 [21:01<2:18:18, 5.04it/s] 13%|█▎ | 6191/48008 [21:01<2:15:30, 5.14it/s] 13%|█▎ | 6192/48008 [21:01<2:13:43, 5.21it/s] 13%|█▎ | 6193/48008 [21:01<2:13:42, 5.21it/s] 13%|█▎ | 6194/48008 [21:02<2:16:04, 5.12it/s] 13%|█▎ | 6195/48008 [21:02<2:14:01, 5.20it/s] 13%|█▎ | 6196/48008 [21:02<2:13:06, 5.24it/s] 13%|█▎ | 6197/48008 [21:02<2:13:04, 5.24it/s] 13%|█▎ | 6198/48008 [21:02<2:12:35, 5.26it/s] 13%|█▎ | 6199/48008 [21:03<2:11:27, 5.30it/s] 13%|█▎ | 6200/48008 [21:03<2:10:14, 5.35it/s] {'loss': 0.3029, 'grad_norm': 0.150390625, 'learning_rate': 0.00017417513747708716, 'epoch': 0.13} 13%|█▎ | 6200/48008 [21:03<2:10:14, 5.35it/s] 13%|█▎ | 6201/48008 [21:03<2:11:25, 5.30it/s] 13%|█▎ | 6202/48008 [21:03<2:12:06, 5.27it/s] 13%|█▎ | 6203/48008 [21:03<2:12:11, 5.27it/s] 13%|█▎ | 6204/48008 [21:04<2:12:56, 5.24it/s] 13%|█▎ | 6205/48008 [21:04<2:12:54, 5.24it/s] 13%|█▎ | 6206/48008 [21:04<2:12:12, 5.27it/s] 13%|█▎ | 6207/48008 [21:04<2:13:58, 5.20it/s] 13%|█▎ | 6208/48008 [21:04<2:13:43, 5.21it/s] 13%|█▎ | 6209/48008 [21:04<2:13:05, 5.23it/s] 13%|█▎ | 6210/48008 [21:05<2:12:15, 5.27it/s] 13%|█▎ | 6211/48008 [21:05<2:12:19, 5.26it/s] 13%|█▎ | 6212/48008 [21:05<2:12:11, 5.27it/s] 13%|█▎ | 6213/48008 [21:05<2:11:18, 5.30it/s] 13%|█▎ | 6214/48008 [21:05<2:11:55, 5.28it/s] 13%|█▎ | 6215/48008 [21:06<2:11:39, 5.29it/s] 13%|█▎ | 6216/48008 [21:06<2:14:15, 5.19it/s] 13%|█▎ | 6217/48008 [21:06<2:12:49, 5.24it/s] 13%|█▎ | 6218/48008 [21:06<2:13:12, 5.23it/s] 13%|█▎ | 6219/48008 [21:06<2:17:29, 5.07it/s] 13%|█▎ | 6220/48008 [21:07<2:15:33, 5.14it/s] 13%|█▎ | 6221/48008 [21:07<2:14:28, 5.18it/s] 13%|█▎ | 6222/48008 [21:07<2:12:50, 5.24it/s] 13%|█▎ | 6223/48008 [21:07<2:13:06, 5.23it/s] 13%|█▎ | 6224/48008 [21:07<2:13:13, 5.23it/s] 13%|█▎ | 6225/48008 [21:08<2:12:16, 5.26it/s] 13%|█▎ | 6226/48008 [21:08<2:12:14, 5.27it/s] 13%|█▎ | 6227/48008 [21:08<2:16:14, 5.11it/s] 13%|█▎ | 6228/48008 [21:08<2:14:55, 5.16it/s] 13%|█▎ | 6229/48008 [21:08<2:13:15, 5.23it/s] 13%|█▎ | 6230/48008 [21:09<2:13:06, 5.23it/s] 13%|█▎ | 6231/48008 [21:09<2:14:23, 5.18it/s] 13%|█▎ | 6232/48008 [21:09<2:17:17, 5.07it/s] 13%|█▎ | 6233/48008 [21:09<2:19:53, 4.98it/s] 13%|█▎ | 6234/48008 [21:09<2:18:15, 5.04it/s] 13%|█▎ | 6235/48008 [21:09<2:16:16, 5.11it/s] 13%|█▎ | 6236/48008 [21:10<2:17:42, 5.06it/s] 13%|█▎ | 6237/48008 [21:10<2:18:27, 5.03it/s] 13%|█▎ | 6238/48008 [21:10<2:17:19, 5.07it/s] 13%|█▎ | 6239/48008 [21:10<2:40:24, 4.34it/s] 13%|█▎ | 6240/48008 [21:11<2:35:02, 4.49it/s] 13%|█▎ | 6241/48008 [21:12<5:16:02, 2.20it/s] 13%|█▎ | 6242/48008 [21:12<4:21:38, 2.66it/s] 13%|█▎ | 6243/48008 [21:12<3:43:55, 3.11it/s] 13%|█▎ | 6244/48008 [21:12<3:15:32, 3.56it/s] 13%|█▎ | 6245/48008 [21:12<2:56:28, 3.94it/s] 13%|█▎ | 6246/48008 [21:13<2:43:35, 4.25it/s] 13%|█▎ | 6247/48008 [21:13<5:06:20, 2.27it/s] 13%|█▎ | 6248/48008 [21:14<4:23:35, 2.64it/s] 13%|█▎ | 6249/48008 [21:14<3:44:36, 3.10it/s] 13%|█▎ | 6250/48008 [21:14<3:17:50, 3.52it/s] {'loss': 0.3052, 'grad_norm': 0.1474609375, 'learning_rate': 0.00017396683886018998, 'epoch': 0.13} 13%|█▎ | 6250/48008 [21:14<3:17:50, 3.52it/s] 13%|█▎ | 6251/48008 [21:14<2:58:32, 3.90it/s] 13%|█▎ | 6252/48008 [21:15<2:54:45, 3.98it/s] 13%|█▎ | 6253/48008 [21:15<4:51:50, 2.38it/s] 13%|█▎ | 6254/48008 [21:16<4:05:00, 2.84it/s] 13%|█▎ | 6255/48008 [21:16<3:29:26, 3.32it/s] 13%|█▎ | 6256/48008 [21:16<3:06:18, 3.73it/s] 13%|█▎ | 6257/48008 [21:16<3:30:45, 3.30it/s] 13%|█▎ | 6258/48008 [21:16<3:07:30, 3.71it/s] 13%|█▎ | 6259/48008 [21:18<6:15:14, 1.85it/s] 13%|█▎ | 6260/48008 [21:18<5:02:38, 2.30it/s] 13%|█▎ | 6261/48008 [21:18<4:12:59, 2.75it/s] 13%|█▎ | 6262/48008 [21:18<3:38:22, 3.19it/s] 13%|█▎ | 6263/48008 [21:18<3:15:20, 3.56it/s] 13%|█▎ | 6264/48008 [21:19<2:56:26, 3.94it/s] 13%|█▎ | 6265/48008 [21:19<2:43:00, 4.27it/s] 13%|█▎ | 6266/48008 [21:19<2:33:33, 4.53it/s] 13%|█▎ | 6267/48008 [21:19<2:30:48, 4.61it/s] 13%|█▎ | 6268/48008 [21:19<2:24:50, 4.80it/s] 13%|█▎ | 6269/48008 [21:20<2:20:54, 4.94it/s] 13%|█▎ | 6270/48008 [21:20<2:18:39, 5.02it/s] 13%|█▎ | 6271/48008 [21:20<2:26:41, 4.74it/s] 13%|█▎ | 6272/48008 [21:20<3:03:22, 3.79it/s] 13%|█▎ | 6273/48008 [21:21<3:12:18, 3.62it/s] 13%|█▎ | 6274/48008 [21:21<2:56:46, 3.93it/s] 13%|█▎ | 6275/48008 [21:21<2:42:49, 4.27it/s] 13%|█▎ | 6276/48008 [21:21<2:35:16, 4.48it/s] 13%|█▎ | 6277/48008 [21:21<2:28:05, 4.70it/s] 13%|█▎ | 6278/48008 [21:22<2:23:11, 4.86it/s] 13%|█▎ | 6279/48008 [21:22<2:19:52, 4.97it/s] 13%|█▎ | 6280/48008 [21:22<2:19:55, 4.97it/s] 13%|█▎ | 6281/48008 [21:22<2:17:52, 5.04it/s] 13%|█▎ | 6282/48008 [21:22<2:17:21, 5.06it/s] 13%|█▎ | 6283/48008 [21:23<2:17:28, 5.06it/s] 13%|█▎ | 6284/48008 [21:24<4:37:46, 2.50it/s] 13%|█▎ | 6285/48008 [21:24<3:56:41, 2.94it/s] 13%|█▎ | 6286/48008 [21:24<3:23:55, 3.41it/s] 13%|█▎ | 6287/48008 [21:24<3:04:05, 3.78it/s] 13%|█▎ | 6288/48008 [21:24<2:48:15, 4.13it/s] 13%|█▎ | 6289/48008 [21:25<2:41:03, 4.32it/s] 13%|█▎ | 6290/48008 [21:25<2:31:20, 4.59it/s] 13%|█▎ | 6291/48008 [21:25<2:27:06, 4.73it/s] 13%|█▎ | 6292/48008 [21:25<2:25:11, 4.79it/s] 13%|█▎ | 6293/48008 [21:25<2:20:30, 4.95it/s] 13%|█▎ | 6294/48008 [21:25<2:16:52, 5.08it/s] 13%|█▎ | 6295/48008 [21:26<2:14:23, 5.17it/s] 13%|█▎ | 6296/48008 [21:26<2:13:28, 5.21it/s] 13%|█▎ | 6297/48008 [21:26<2:13:00, 5.23it/s] 13%|█▎ | 6298/48008 [21:26<2:13:09, 5.22it/s] 13%|█▎ | 6299/48008 [21:26<2:12:31, 5.25it/s] 13%|█▎ | 6300/48008 [21:27<2:11:41, 5.28it/s] {'loss': 0.3016, 'grad_norm': 0.146484375, 'learning_rate': 0.0001737585402432928, 'epoch': 0.13} 13%|█▎ | 6300/48008 [21:27<2:11:41, 5.28it/s] 13%|█▎ | 6301/48008 [21:27<2:14:14, 5.18it/s] 13%|█▎ | 6302/48008 [21:27<2:17:10, 5.07it/s] 13%|█▎ | 6303/48008 [21:27<2:16:16, 5.10it/s] 13%|█▎ | 6304/48008 [21:27<2:15:41, 5.12it/s] 13%|█▎ | 6305/48008 [21:28<2:16:05, 5.11it/s] 13%|█▎ | 6306/48008 [21:28<2:15:12, 5.14it/s] 13%|█▎ | 6307/48008 [21:28<2:15:14, 5.14it/s] 13%|█▎ | 6308/48008 [21:28<2:12:43, 5.24it/s] 13%|█▎ | 6309/48008 [21:28<2:14:18, 5.17it/s] 13%|█▎ | 6310/48008 [21:29<2:13:35, 5.20it/s] 13%|█▎ | 6311/48008 [21:29<2:14:23, 5.17it/s] 13%|█▎ | 6312/48008 [21:29<2:13:32, 5.20it/s] 13%|█▎ | 6313/48008 [21:29<2:11:08, 5.30it/s] 13%|█▎ | 6314/48008 [21:29<2:11:33, 5.28it/s] 13%|█▎ | 6315/48008 [21:29<2:11:53, 5.27it/s] 13%|█▎ | 6316/48008 [21:30<2:13:42, 5.20it/s] 13%|█▎ | 6317/48008 [21:30<2:13:21, 5.21it/s] 13%|█▎ | 6318/48008 [21:30<2:15:21, 5.13it/s] 13%|█▎ | 6319/48008 [21:30<2:16:57, 5.07it/s] 13%|█▎ | 6320/48008 [21:30<2:15:41, 5.12it/s] 13%|█▎ | 6321/48008 [21:31<2:14:30, 5.17it/s] 13%|█▎ | 6322/48008 [21:31<2:15:03, 5.14it/s] 13%|█▎ | 6323/48008 [21:31<2:14:27, 5.17it/s] 13%|█▎ | 6324/48008 [21:31<2:13:31, 5.20it/s] 13%|█▎ | 6325/48008 [21:31<2:22:39, 4.87it/s] 13%|█▎ | 6326/48008 [21:32<2:19:50, 4.97it/s] 13%|█▎ | 6327/48008 [21:32<2:17:39, 5.05it/s] 13%|█▎ | 6328/48008 [21:32<2:16:07, 5.10it/s] 13%|█▎ | 6329/48008 [21:32<2:24:54, 4.79it/s] 13%|█▎ | 6330/48008 [21:32<2:20:14, 4.95it/s] 13%|█▎ | 6331/48008 [21:33<2:20:18, 4.95it/s] 13%|█▎ | 6332/48008 [21:33<2:17:15, 5.06it/s] 13%|█▎ | 6333/48008 [21:33<2:15:49, 5.11it/s] 13%|█▎ | 6334/48008 [21:33<2:17:09, 5.06it/s] 13%|█▎ | 6335/48008 [21:33<2:14:27, 5.17it/s] 13%|█▎ | 6336/48008 [21:34<2:13:45, 5.19it/s] 13%|█▎ | 6337/48008 [21:34<2:12:44, 5.23it/s] 13%|█▎ | 6338/48008 [21:34<2:12:41, 5.23it/s] 13%|█▎ | 6339/48008 [21:34<2:12:31, 5.24it/s] 13%|█▎ | 6340/48008 [21:34<2:12:36, 5.24it/s] 13%|█▎ | 6341/48008 [21:35<2:14:43, 5.15it/s] 13%|█▎ | 6342/48008 [21:35<2:14:16, 5.17it/s] 13%|█▎ | 6343/48008 [21:35<2:16:05, 5.10it/s] 13%|█▎ | 6344/48008 [21:35<2:14:13, 5.17it/s] 13%|█▎ | 6345/48008 [21:35<2:15:59, 5.11it/s] 13%|█▎ | 6346/48008 [21:36<2:14:23, 5.17it/s] 13%|█▎ | 6347/48008 [21:36<2:13:48, 5.19it/s] 13%|█▎ | 6348/48008 [21:36<2:13:40, 5.19it/s] 13%|█▎ | 6349/48008 [21:36<2:13:35, 5.20it/s] 13%|█▎ | 6350/48008 [21:36<2:15:36, 5.12it/s] {'loss': 0.3042, 'grad_norm': 0.279296875, 'learning_rate': 0.0001735502416263956, 'epoch': 0.13} 13%|█▎ | 6350/48008 [21:36<2:15:36, 5.12it/s] 13%|█▎ | 6351/48008 [21:37<2:14:17, 5.17it/s] 13%|█▎ | 6352/48008 [21:37<2:13:25, 5.20it/s] 13%|█▎ | 6353/48008 [21:37<2:11:49, 5.27it/s] 13%|█▎ | 6354/48008 [21:37<2:12:16, 5.25it/s] 13%|█▎ | 6355/48008 [21:37<2:11:56, 5.26it/s] 13%|█▎ | 6356/48008 [21:37<2:11:41, 5.27it/s] 13%|█▎ | 6357/48008 [21:38<2:11:52, 5.26it/s] 13%|█▎ | 6358/48008 [21:38<2:12:25, 5.24it/s] 13%|█▎ | 6359/48008 [21:38<2:12:04, 5.26it/s] 13%|█▎ | 6360/48008 [21:38<2:14:06, 5.18it/s] 13%|█▎ | 6361/48008 [21:38<2:13:52, 5.19it/s] 13%|█▎ | 6362/48008 [21:39<2:13:21, 5.20it/s] 13%|█▎ | 6363/48008 [21:39<2:12:49, 5.23it/s] 13%|█▎ | 6364/48008 [21:39<2:13:18, 5.21it/s] 13%|█▎ | 6365/48008 [21:39<2:22:33, 4.87it/s] 13%|█▎ | 6366/48008 [21:39<2:19:17, 4.98it/s] 13%|█▎ | 6367/48008 [21:40<2:17:14, 5.06it/s] 13%|█▎ | 6368/48008 [21:40<2:17:52, 5.03it/s] 13%|█▎ | 6369/48008 [21:40<2:16:28, 5.08it/s] 13%|█▎ | 6370/48008 [21:40<2:15:23, 5.13it/s] 13%|█▎ | 6371/48008 [21:40<2:15:44, 5.11it/s] 13%|█▎ | 6372/48008 [21:41<2:14:43, 5.15it/s] 13%|█▎ | 6373/48008 [21:41<2:16:03, 5.10it/s] 13%|█▎ | 6374/48008 [21:41<2:15:58, 5.10it/s] 13%|█▎ | 6375/48008 [21:41<2:14:42, 5.15it/s] 13%|█▎ | 6376/48008 [21:42<2:38:50, 4.37it/s] 13%|█▎ | 6377/48008 [21:42<2:30:35, 4.61it/s] 13%|█▎ | 6378/48008 [21:42<2:27:29, 4.70it/s] 13%|█▎ | 6379/48008 [21:42<2:22:51, 4.86it/s] 13%|█▎ | 6380/48008 [21:42<2:19:50, 4.96it/s] 13%|█▎ | 6381/48008 [21:42<2:17:51, 5.03it/s] 13%|█▎ | 6382/48008 [21:43<2:16:30, 5.08it/s] 13%|█▎ | 6383/48008 [21:43<2:15:27, 5.12it/s] 13%|█▎ | 6384/48008 [21:43<2:14:07, 5.17it/s] 13%|█▎ | 6385/48008 [21:43<2:13:34, 5.19it/s] 13%|█▎ | 6386/48008 [21:43<2:12:58, 5.22it/s] 13%|█▎ | 6387/48008 [21:44<2:13:05, 5.21it/s] 13%|█▎ | 6388/48008 [21:44<2:12:38, 5.23it/s] 13%|█▎ | 6389/48008 [21:44<2:11:36, 5.27it/s] 13%|█▎ | 6390/48008 [21:44<2:12:50, 5.22it/s] 13%|█▎ | 6391/48008 [21:44<2:12:39, 5.23it/s] 13%|█▎ | 6392/48008 [21:45<2:13:11, 5.21it/s] 13%|█▎ | 6393/48008 [21:45<2:13:58, 5.18it/s] 13%|█▎ | 6394/48008 [21:45<2:15:46, 5.11it/s] 13%|█▎ | 6395/48008 [21:45<2:13:47, 5.18it/s] 13%|█▎ | 6396/48008 [21:45<2:14:08, 5.17it/s] 13%|█▎ | 6397/48008 [21:46<2:12:56, 5.22it/s] 13%|█▎ | 6398/48008 [21:46<2:12:36, 5.23it/s] 13%|█▎ | 6399/48008 [21:46<2:12:49, 5.22it/s] 13%|█▎ | 6400/48008 [21:46<2:12:42, 5.23it/s] {'loss': 0.305, 'grad_norm': 1.21875, 'learning_rate': 0.00017334194300949843, 'epoch': 0.13} 13%|█▎ | 6400/48008 [21:46<2:12:42, 5.23it/s] 13%|█▎ | 6401/48008 [21:46<2:12:57, 5.22it/s] 13%|█▎ | 6402/48008 [21:47<2:37:30, 4.40it/s] 13%|█▎ | 6403/48008 [21:47<2:31:29, 4.58it/s] 13%|█▎ | 6404/48008 [21:47<2:25:40, 4.76it/s] 13%|█▎ | 6405/48008 [21:47<2:24:08, 4.81it/s] 13%|█▎ | 6406/48008 [21:47<2:19:55, 4.96it/s] 13%|█▎ | 6407/48008 [21:48<2:17:37, 5.04it/s] 13%|█▎ | 6408/48008 [21:48<2:18:12, 5.02it/s] 13%|█▎ | 6409/48008 [21:48<2:16:57, 5.06it/s] 13%|█▎ | 6410/48008 [21:48<2:18:40, 5.00it/s] 13%|█▎ | 6411/48008 [21:48<2:17:39, 5.04it/s] 13%|█▎ | 6412/48008 [21:49<2:17:05, 5.06it/s] 13%|█▎ | 6413/48008 [21:49<2:18:20, 5.01it/s] 13%|█▎ | 6414/48008 [21:49<2:40:26, 4.32it/s] 13%|█▎ | 6415/48008 [21:49<2:31:58, 4.56it/s] 13%|█▎ | 6416/48008 [21:49<2:25:56, 4.75it/s] 13%|█▎ | 6417/48008 [21:50<2:19:39, 4.96it/s] 13%|█▎ | 6418/48008 [21:50<2:17:14, 5.05it/s] 13%|█▎ | 6419/48008 [21:50<2:17:38, 5.04it/s] 13%|█▎ | 6420/48008 [21:50<2:18:51, 4.99it/s] 13%|█▎ | 6421/48008 [21:50<2:17:03, 5.06it/s] 13%|█▎ | 6422/48008 [21:51<2:15:17, 5.12it/s] 13%|█▎ | 6423/48008 [21:51<2:14:05, 5.17it/s] 13%|█▎ | 6424/48008 [21:51<2:17:14, 5.05it/s] 13%|█▎ | 6425/48008 [21:51<2:15:54, 5.10it/s] 13%|█▎ | 6426/48008 [21:52<2:55:31, 3.95it/s] 13%|█▎ | 6427/48008 [21:52<2:46:14, 4.17it/s] 13%|█▎ | 6428/48008 [21:52<2:35:29, 4.46it/s] 13%|█▎ | 6429/48008 [21:52<2:27:37, 4.69it/s] 13%|█▎ | 6430/48008 [21:52<2:22:17, 4.87it/s] 13%|█▎ | 6431/48008 [21:53<2:23:14, 4.84it/s] 13%|█▎ | 6432/48008 [21:53<2:22:30, 4.86it/s] 13%|█▎ | 6433/48008 [21:53<2:19:04, 4.98it/s] 13%|█▎ | 6434/48008 [21:53<2:17:49, 5.03it/s] 13%|█▎ | 6435/48008 [21:53<2:20:09, 4.94it/s] 13%|█▎ | 6436/48008 [21:54<2:18:43, 4.99it/s] 13%|█▎ | 6437/48008 [21:54<2:19:05, 4.98it/s] 13%|█▎ | 6438/48008 [21:54<2:17:00, 5.06it/s] 13%|█▎ | 6439/48008 [21:54<2:15:52, 5.10it/s] 13%|█▎ | 6440/48008 [21:54<2:17:12, 5.05it/s] 13%|█▎ | 6441/48008 [21:55<2:17:15, 5.05it/s] 13%|█▎ | 6442/48008 [21:55<2:15:44, 5.10it/s] 13%|█▎ | 6443/48008 [21:55<2:15:16, 5.12it/s] 13%|█▎ | 6444/48008 [21:55<2:13:45, 5.18it/s] 13%|█▎ | 6445/48008 [21:55<2:12:46, 5.22it/s] 13%|█▎ | 6446/48008 [21:56<2:22:14, 4.87it/s] 13%|█▎ | 6447/48008 [21:56<2:19:09, 4.98it/s] 13%|█▎ | 6448/48008 [21:56<2:16:49, 5.06it/s] 13%|█▎ | 6449/48008 [21:56<2:19:24, 4.97it/s] 13%|█▎ | 6450/48008 [21:56<2:15:58, 5.09it/s] {'loss': 0.3043, 'grad_norm': 0.1669921875, 'learning_rate': 0.00017313364439260125, 'epoch': 0.13} 13%|█▎ | 6450/48008 [21:56<2:15:58, 5.09it/s] 13%|█▎ | 6451/48008 [21:57<2:16:21, 5.08it/s] 13%|█▎ | 6452/48008 [21:57<2:15:05, 5.13it/s] 13%|█▎ | 6453/48008 [21:57<2:14:37, 5.14it/s] 13%|█▎ | 6454/48008 [21:57<2:13:55, 5.17it/s] 13%|█▎ | 6455/48008 [21:57<2:54:09, 3.98it/s] 13%|█▎ | 6456/48008 [21:58<2:41:50, 4.28it/s] 13%|█▎ | 6457/48008 [21:58<2:33:02, 4.52it/s] 13%|█▎ | 6458/48008 [21:58<2:26:43, 4.72it/s] 13%|█▎ | 6459/48008 [21:58<2:24:41, 4.79it/s] 13%|█▎ | 6460/48008 [21:58<2:21:54, 4.88it/s] 13%|█▎ | 6461/48008 [21:59<2:18:46, 4.99it/s] 13%|█▎ | 6462/48008 [21:59<2:17:02, 5.05it/s] 13%|█▎ | 6463/48008 [21:59<2:16:19, 5.08it/s] 13%|█▎ | 6464/48008 [21:59<2:17:07, 5.05it/s] 13%|█▎ | 6465/48008 [21:59<2:15:43, 5.10it/s] 13%|█▎ | 6466/48008 [22:00<2:17:06, 5.05it/s] 13%|█▎ | 6467/48008 [22:00<2:15:06, 5.12it/s] 13%|█▎ | 6468/48008 [22:00<2:14:26, 5.15it/s] 13%|█▎ | 6469/48008 [22:00<2:13:52, 5.17it/s] 13%|█▎ | 6470/48008 [22:00<2:13:03, 5.20it/s] 13%|█▎ | 6471/48008 [22:01<2:13:18, 5.19it/s] 13%|█▎ | 6472/48008 [22:01<2:13:38, 5.18it/s] 13%|█▎ | 6473/48008 [22:01<2:13:47, 5.17it/s] 13%|█▎ | 6474/48008 [22:01<2:22:43, 4.85it/s] 13%|█▎ | 6475/48008 [22:01<2:20:57, 4.91it/s] 13%|█▎ | 6476/48008 [22:02<2:16:48, 5.06it/s] 13%|█▎ | 6477/48008 [22:02<2:16:11, 5.08it/s] 13%|█▎ | 6478/48008 [22:02<2:14:52, 5.13it/s] 13%|█▎ | 6479/48008 [22:02<2:15:02, 5.13it/s] 13%|█▎ | 6480/48008 [22:02<2:15:51, 5.09it/s] 13%|█▎ | 6481/48008 [22:03<2:15:27, 5.11it/s] 14%|█▎ | 6482/48008 [22:03<2:15:58, 5.09it/s] 14%|█▎ | 6483/48008 [22:03<2:16:22, 5.08it/s] 14%|█▎ | 6484/48008 [22:03<2:16:00, 5.09it/s] 14%|█▎ | 6485/48008 [22:03<2:13:53, 5.17it/s] 14%|█▎ | 6486/48008 [22:04<2:13:42, 5.18it/s] 14%|█▎ | 6487/48008 [22:04<2:12:51, 5.21it/s] 14%|█▎ | 6488/48008 [22:04<2:12:15, 5.23it/s] 14%|█▎ | 6489/48008 [22:04<2:12:15, 5.23it/s] 14%|█▎ | 6490/48008 [22:04<2:35:56, 4.44it/s] 14%|█▎ | 6491/48008 [22:05<2:28:38, 4.66it/s] 14%|█▎ | 6492/48008 [22:05<2:24:33, 4.79it/s] 14%|█▎ | 6493/48008 [22:05<2:21:41, 4.88it/s] 14%|█▎ | 6494/48008 [22:05<2:20:54, 4.91it/s] 14%|█▎ | 6495/48008 [22:05<2:18:14, 5.01it/s] 14%|█▎ | 6496/48008 [22:06<2:16:04, 5.08it/s] 14%|█▎ | 6497/48008 [22:06<2:14:25, 5.15it/s] 14%|█▎ | 6498/48008 [22:06<2:13:29, 5.18it/s] 14%|█▎ | 6499/48008 [22:06<2:13:31, 5.18it/s] 14%|█▎ | 6500/48008 [22:06<2:16:39, 5.06it/s] {'loss': 0.3033, 'grad_norm': 0.8359375, 'learning_rate': 0.00017292534577570404, 'epoch': 0.14} 14%|█▎ | 6500/48008 [22:06<2:16:39, 5.06it/s] 14%|█▎ | 6501/48008 [22:07<2:16:23, 5.07it/s] 14%|█▎ | 6502/48008 [22:07<2:15:04, 5.12it/s] 14%|█▎ | 6503/48008 [22:07<2:13:51, 5.17it/s] 14%|█▎ | 6504/48008 [22:07<2:11:41, 5.25it/s] 14%|█▎ | 6505/48008 [22:07<2:11:45, 5.25it/s] 14%|█▎ | 6506/48008 [22:07<2:12:08, 5.23it/s] 14%|█▎ | 6507/48008 [22:08<2:13:09, 5.19it/s] 14%|█▎ | 6508/48008 [22:08<2:22:30, 4.85it/s] 14%|█▎ | 6509/48008 [22:08<2:21:56, 4.87it/s] 14%|█▎ | 6510/48008 [22:08<2:21:05, 4.90it/s] 14%|█▎ | 6511/48008 [22:09<2:19:07, 4.97it/s] 14%|█▎ | 6512/48008 [22:09<2:19:18, 4.96it/s] 14%|█▎ | 6513/48008 [22:09<2:16:54, 5.05it/s] 14%|█▎ | 6514/48008 [22:09<2:14:58, 5.12it/s] 14%|█▎ | 6515/48008 [22:09<2:23:32, 4.82it/s] 14%|█▎ | 6516/48008 [22:10<2:19:45, 4.95it/s] 14%|█▎ | 6517/48008 [22:10<2:18:25, 5.00it/s] 14%|█▎ | 6518/48008 [22:10<2:56:43, 3.91it/s] 14%|█▎ | 6519/48008 [22:10<2:45:33, 4.18it/s] 14%|█▎ | 6520/48008 [22:11<2:36:15, 4.43it/s] 14%|█▎ | 6521/48008 [22:11<2:28:38, 4.65it/s] 14%|█▎ | 6522/48008 [22:11<2:23:13, 4.83it/s] 14%|█▎ | 6523/48008 [22:11<2:19:14, 4.97it/s] 14%|█▎ | 6524/48008 [22:11<2:16:45, 5.06it/s] 14%|█▎ | 6525/48008 [22:11<2:19:23, 4.96it/s] 14%|█▎ | 6526/48008 [22:12<2:16:55, 5.05it/s] 14%|█▎ | 6527/48008 [22:12<2:14:38, 5.13it/s] 14%|█▎ | 6528/48008 [22:12<2:15:01, 5.12it/s] 14%|█▎ | 6529/48008 [22:12<2:14:04, 5.16it/s] 14%|█▎ | 6530/48008 [22:12<2:12:26, 5.22it/s] 14%|█▎ | 6531/48008 [22:13<2:11:12, 5.27it/s] 14%|█▎ | 6532/48008 [22:13<2:11:13, 5.27it/s] 14%|█▎ | 6533/48008 [22:13<2:11:48, 5.24it/s] 14%|█▎ | 6534/48008 [22:13<2:14:17, 5.15it/s] 14%|█▎ | 6535/48008 [22:13<2:13:52, 5.16it/s] 14%|█▎ | 6536/48008 [22:14<2:13:45, 5.17it/s] 14%|█▎ | 6537/48008 [22:14<2:13:55, 5.16it/s] 14%|█▎ | 6538/48008 [22:14<2:13:25, 5.18it/s] 14%|█▎ | 6539/48008 [22:14<2:12:08, 5.23it/s] 14%|█▎ | 6540/48008 [22:14<2:11:58, 5.24it/s] 14%|█▎ | 6541/48008 [22:15<2:13:12, 5.19it/s] 14%|█▎ | 6542/48008 [22:15<2:11:50, 5.24it/s] 14%|█▎ | 6543/48008 [22:15<2:11:23, 5.26it/s] 14%|█▎ | 6544/48008 [22:15<2:13:01, 5.19it/s] 14%|█▎ | 6545/48008 [22:15<2:12:21, 5.22it/s] 14%|█▎ | 6546/48008 [22:16<2:13:11, 5.19it/s] 14%|█▎ | 6547/48008 [22:16<2:14:35, 5.13it/s] 14%|█▎ | 6548/48008 [22:16<2:13:43, 5.17it/s] 14%|█▎ | 6549/48008 [22:16<2:12:54, 5.20it/s] 14%|█▎ | 6550/48008 [22:16<2:13:28, 5.18it/s] 14%|█▎ | 6550/48008 [22:16<2:13:28, 5.18it/s]{'loss': 0.303, 'grad_norm': 0.1650390625, 'learning_rate': 0.00017271704715880688, 'epoch': 0.14} 14%|█▎ | 6551/48008 [22:17<2:23:01, 4.83it/s] 14%|█▎ | 6552/48008 [22:17<2:18:57, 4.97it/s] 14%|█▎ | 6553/48008 [22:17<2:16:33, 5.06it/s] 14%|█▎ | 6554/48008 [22:17<2:14:36, 5.13it/s] 14%|█▎ | 6555/48008 [22:17<2:13:14, 5.18it/s] 14%|█▎ | 6556/48008 [22:18<2:36:48, 4.41it/s] 14%|█▎ | 6557/48008 [22:18<2:31:27, 4.56it/s] 14%|█▎ | 6558/48008 [22:18<2:25:30, 4.75it/s] 14%|█▎ | 6559/48008 [22:18<2:21:25, 4.88it/s] 14%|█▎ | 6560/48008 [22:18<2:19:10, 4.96it/s] 14%|█▎ | 6561/48008 [22:19<2:18:16, 5.00it/s] 14%|█▎ | 6562/48008 [22:19<2:16:06, 5.08it/s] 14%|█▎ | 6563/48008 [22:19<2:17:04, 5.04it/s] 14%|█▎ | 6564/48008 [22:19<2:18:06, 5.00it/s] 14%|█▎ | 6565/48008 [22:19<2:39:55, 4.32it/s] 14%|█▎ | 6566/48008 [22:20<2:30:46, 4.58it/s] 14%|█▎ | 6567/48008 [22:20<2:24:48, 4.77it/s] 14%|█▎ | 6568/48008 [22:20<2:19:54, 4.94it/s] 14%|█▎ | 6569/48008 [22:20<2:16:56, 5.04it/s] 14%|█▎ | 6570/48008 [22:20<2:15:38, 5.09it/s] 14%|█▎ | 6571/48008 [22:21<2:15:28, 5.10it/s] 14%|█▎ | 6572/48008 [22:21<2:14:21, 5.14it/s] 14%|█▎ | 6573/48008 [22:21<2:14:38, 5.13it/s] 14%|█▎ | 6574/48008 [22:21<2:16:20, 5.06it/s] 14%|█▎ | 6575/48008 [22:21<2:16:02, 5.08it/s] 14%|█▎ | 6576/48008 [22:22<2:24:30, 4.78it/s] 14%|█▎ | 6577/48008 [22:22<2:22:06, 4.86it/s] 14%|█▎ | 6578/48008 [22:22<2:20:56, 4.90it/s] 14%|█▎ | 6579/48008 [22:22<2:18:05, 5.00it/s] 14%|█▎ | 6580/48008 [22:22<2:15:54, 5.08it/s] 14%|█▎ | 6581/48008 [22:23<2:24:01, 4.79it/s] 14%|█▎ | 6582/48008 [22:23<2:19:54, 4.94it/s] 14%|█▎ | 6583/48008 [22:23<2:17:05, 5.04it/s] 14%|█▎ | 6584/48008 [22:23<2:15:28, 5.10it/s] 14%|█▎ | 6585/48008 [22:23<2:13:50, 5.16it/s] 14%|█▎ | 6586/48008 [22:24<2:13:09, 5.18it/s] 14%|█▎ | 6587/48008 [22:24<2:13:03, 5.19it/s] 14%|█▎ | 6588/48008 [22:24<2:12:27, 5.21it/s] 14%|█▎ | 6589/48008 [22:24<2:22:11, 4.85it/s] 14%|█▎ | 6590/48008 [22:24<2:18:52, 4.97it/s] 14%|█▎ | 6591/48008 [22:25<2:17:02, 5.04it/s] 14%|█▎ | 6592/48008 [22:25<2:17:03, 5.04it/s] 14%|█▎ | 6593/48008 [22:25<2:15:27, 5.10it/s] 14%|█▎ | 6594/48008 [22:25<2:13:06, 5.19it/s] 14%|█▎ | 6595/48008 [22:25<2:12:03, 5.23it/s] 14%|█▎ | 6596/48008 [22:26<2:12:49, 5.20it/s] 14%|█▎ | 6597/48008 [22:26<2:11:50, 5.24it/s] 14%|█▎ | 6598/48008 [22:26<2:12:26, 5.21it/s] 14%|█▎ | 6599/48008 [22:26<2:12:09, 5.22it/s] 14%|█▎ | 6600/48008 [22:26<2:09:53, 5.31it/s] {'loss': 0.3036, 'grad_norm': 0.27734375, 'learning_rate': 0.00017250874854190967, 'epoch': 0.14} 14%|█▎ | 6600/48008 [22:26<2:09:53, 5.31it/s] 14%|█▎ | 6601/48008 [22:26<2:10:43, 5.28it/s] 14%|█▍ | 6602/48008 [22:27<2:09:46, 5.32it/s] 14%|█▍ | 6603/48008 [22:27<2:10:21, 5.29it/s] 14%|█▍ | 6604/48008 [22:27<2:51:06, 4.03it/s] 14%|█▍ | 6605/48008 [22:27<2:39:35, 4.32it/s] 14%|█▍ | 6606/48008 [22:28<2:30:04, 4.60it/s] 14%|█▍ | 6607/48008 [22:28<2:26:45, 4.70it/s] 14%|█▍ | 6608/48008 [22:28<2:21:56, 4.86it/s] 14%|█▍ | 6609/48008 [22:28<2:18:41, 4.97it/s] 14%|█▍ | 6610/48008 [22:28<2:19:56, 4.93it/s] 14%|█▍ | 6611/48008 [22:29<2:27:21, 4.68it/s] 14%|█▍ | 6612/48008 [22:29<2:22:29, 4.84it/s] 14%|█▍ | 6613/48008 [22:29<2:21:33, 4.87it/s] 14%|█▍ | 6614/48008 [22:29<2:17:46, 5.01it/s] 14%|█▍ | 6615/48008 [22:29<2:15:40, 5.08it/s] 14%|█▍ | 6616/48008 [22:30<2:14:39, 5.12it/s] 14%|█▍ | 6617/48008 [22:30<2:17:54, 5.00it/s] 14%|█▍ | 6618/48008 [22:30<2:18:48, 4.97it/s] 14%|█▍ | 6619/48008 [22:30<2:16:29, 5.05it/s] 14%|█▍ | 6620/48008 [22:30<2:15:53, 5.08it/s] 14%|█▍ | 6621/48008 [22:31<2:14:02, 5.15it/s] 14%|█▍ | 6622/48008 [22:31<2:13:54, 5.15it/s] 14%|█▍ | 6623/48008 [22:31<2:13:20, 5.17it/s] 14%|█▍ | 6624/48008 [22:31<2:22:21, 4.84it/s] 14%|█▍ | 6625/48008 [22:31<2:18:53, 4.97it/s] 14%|█▍ | 6626/48008 [22:32<2:16:05, 5.07it/s] 14%|█▍ | 6627/48008 [22:32<2:13:51, 5.15it/s] 14%|█▍ | 6628/48008 [22:32<2:12:58, 5.19it/s] 14%|█▍ | 6629/48008 [22:32<2:12:14, 5.21it/s] 14%|█▍ | 6630/48008 [22:32<2:12:25, 5.21it/s] 14%|█▍ | 6631/48008 [22:33<2:11:25, 5.25it/s] 14%|█▍ | 6632/48008 [22:33<2:10:50, 5.27it/s] 14%|█▍ | 6633/48008 [22:33<2:13:08, 5.18it/s] 14%|█▍ | 6634/48008 [22:33<2:12:50, 5.19it/s] 14%|█▍ | 6635/48008 [22:33<2:16:04, 5.07it/s] 14%|█▍ | 6636/48008 [22:34<2:55:10, 3.94it/s] 14%|█▍ | 6637/48008 [22:34<2:42:19, 4.25it/s] 14%|█▍ | 6638/48008 [22:34<2:32:59, 4.51it/s] 14%|█▍ | 6639/48008 [22:34<2:27:42, 4.67it/s] 14%|█▍ | 6640/48008 [22:34<2:22:46, 4.83it/s] 14%|█▍ | 6641/48008 [22:35<2:20:36, 4.90it/s] 14%|█▍ | 6642/48008 [22:35<2:19:14, 4.95it/s] 14%|█▍ | 6643/48008 [22:35<2:17:01, 5.03it/s] 14%|█▍ | 6644/48008 [22:35<2:15:18, 5.10it/s] 14%|█▍ | 6645/48008 [22:35<2:14:13, 5.14it/s] 14%|█▍ | 6646/48008 [22:36<2:23:15, 4.81it/s] 14%|█▍ | 6647/48008 [22:36<2:18:46, 4.97it/s] 14%|█▍ | 6648/48008 [22:36<2:16:23, 5.05it/s] 14%|█▍ | 6649/48008 [22:36<2:17:23, 5.02it/s] 14%|█▍ | 6650/48008 [22:36<2:17:37, 5.01it/s] {'loss': 0.304, 'grad_norm': 1.1015625, 'learning_rate': 0.00017230044992501252, 'epoch': 0.14} 14%|█▍ | 6650/48008 [22:36<2:17:37, 5.01it/s] 14%|█▍ | 6651/48008 [22:37<2:16:48, 5.04it/s] 14%|█▍ | 6652/48008 [22:37<2:39:04, 4.33it/s] 14%|█▍ | 6653/48008 [22:37<2:30:52, 4.57it/s] 14%|█▍ | 6654/48008 [22:37<2:25:58, 4.72it/s] 14%|█▍ | 6655/48008 [22:38<2:22:33, 4.83it/s] 14%|█▍ | 6656/48008 [22:38<2:18:12, 4.99it/s] 14%|█▍ | 6657/48008 [22:38<2:15:48, 5.07it/s] 14%|█▍ | 6658/48008 [22:38<2:16:37, 5.04it/s] 14%|█▍ | 6659/48008 [22:38<2:16:58, 5.03it/s] 14%|█▍ | 6660/48008 [22:39<2:15:38, 5.08it/s] 14%|█▍ | 6661/48008 [22:39<2:13:29, 5.16it/s] 14%|█▍ | 6662/48008 [22:39<2:12:38, 5.20it/s] 14%|█▍ | 6663/48008 [22:39<2:36:23, 4.41it/s] 14%|█▍ | 6664/48008 [22:39<2:29:32, 4.61it/s] 14%|█▍ | 6665/48008 [22:40<2:24:01, 4.78it/s] 14%|█▍ | 6666/48008 [22:40<2:22:26, 4.84it/s] 14%|█▍ | 6667/48008 [22:40<2:19:04, 4.95it/s] 14%|█▍ | 6668/48008 [22:40<2:40:30, 4.29it/s] 14%|█▍ | 6669/48008 [22:40<2:32:38, 4.51it/s] 14%|█▍ | 6670/48008 [22:41<2:26:37, 4.70it/s] 14%|█▍ | 6671/48008 [22:41<2:22:19, 4.84it/s] 14%|█▍ | 6672/48008 [22:41<2:19:17, 4.95it/s] 14%|█▍ | 6673/48008 [22:41<2:17:06, 5.02it/s] 14%|█▍ | 6674/48008 [22:41<2:18:06, 4.99it/s] 14%|█▍ | 6675/48008 [22:42<2:15:57, 5.07it/s] 14%|█▍ | 6676/48008 [22:42<2:13:40, 5.15it/s] 14%|█▍ | 6677/48008 [22:42<2:12:46, 5.19it/s] 14%|█▍ | 6678/48008 [22:42<2:12:36, 5.19it/s] 14%|█▍ | 6679/48008 [22:42<2:13:56, 5.14it/s] 14%|█▍ | 6680/48008 [22:43<2:13:27, 5.16it/s] 14%|█▍ | 6681/48008 [22:43<2:12:41, 5.19it/s] 14%|█▍ | 6682/48008 [22:43<2:16:02, 5.06it/s] 14%|█▍ | 6683/48008 [22:43<2:16:05, 5.06it/s] 14%|█▍ | 6684/48008 [22:43<2:15:19, 5.09it/s] 14%|█▍ | 6685/48008 [22:44<2:13:02, 5.18it/s] 14%|█▍ | 6686/48008 [22:44<2:15:17, 5.09it/s] 14%|█▍ | 6687/48008 [22:44<2:14:26, 5.12it/s] 14%|█▍ | 6688/48008 [22:44<2:12:22, 5.20it/s] 14%|█▍ | 6689/48008 [22:44<2:12:11, 5.21it/s] 14%|█▍ | 6690/48008 [22:45<2:21:39, 4.86it/s] 14%|█▍ | 6691/48008 [22:45<2:19:09, 4.95it/s] 14%|█▍ | 6692/48008 [22:45<2:17:44, 5.00it/s] 14%|█▍ | 6693/48008 [22:45<2:15:48, 5.07it/s] 14%|█▍ | 6694/48008 [22:45<2:14:22, 5.12it/s] 14%|█▍ | 6695/48008 [22:46<2:14:23, 5.12it/s] 14%|█▍ | 6696/48008 [22:46<2:12:15, 5.21it/s] 14%|█▍ | 6697/48008 [22:46<2:11:45, 5.23it/s] 14%|█▍ | 6698/48008 [22:46<2:10:39, 5.27it/s] 14%|█▍ | 6699/48008 [22:46<2:51:04, 4.02it/s] 14%|█▍ | 6700/48008 [22:47<2:38:43, 4.34it/s] {'loss': 0.3043, 'grad_norm': 0.39453125, 'learning_rate': 0.0001720921513081153, 'epoch': 0.14} 14%|█▍ | 6700/48008 [22:47<2:38:43, 4.34it/s] 14%|█▍ | 6701/48008 [22:47<2:30:55, 4.56it/s] 14%|█▍ | 6702/48008 [22:47<2:25:15, 4.74it/s] 14%|█▍ | 6703/48008 [22:47<2:21:13, 4.87it/s] 14%|█▍ | 6704/48008 [22:47<2:20:52, 4.89it/s] 14%|█▍ | 6705/48008 [22:48<2:18:09, 4.98it/s] 14%|█▍ | 6706/48008 [22:48<2:15:04, 5.10it/s] 14%|█▍ | 6707/48008 [22:48<2:13:06, 5.17it/s] 14%|█▍ | 6708/48008 [22:48<2:12:39, 5.19it/s] 14%|█▍ | 6709/48008 [22:48<2:11:18, 5.24it/s] 14%|█▍ | 6710/48008 [22:49<2:51:46, 4.01it/s] 14%|█▍ | 6711/48008 [22:49<2:41:12, 4.27it/s] 14%|█▍ | 6712/48008 [22:49<2:32:09, 4.52it/s] 14%|█▍ | 6713/48008 [22:49<2:27:07, 4.68it/s] 14%|█▍ | 6714/48008 [22:50<2:21:52, 4.85it/s] 14%|█▍ | 6715/48008 [22:50<2:18:14, 4.98it/s] 14%|█▍ | 6716/48008 [22:50<2:16:14, 5.05it/s] 14%|█▍ | 6717/48008 [22:50<2:14:37, 5.11it/s] 14%|█▍ | 6718/48008 [22:50<2:13:29, 5.16it/s] 14%|█▍ | 6719/48008 [22:51<2:12:14, 5.20it/s] 14%|█▍ | 6720/48008 [22:51<2:11:46, 5.22it/s] 14%|█▍ | 6721/48008 [22:51<2:12:41, 5.19it/s] 14%|█▍ | 6722/48008 [22:51<2:11:37, 5.23it/s] 14%|█▍ | 6723/48008 [22:51<2:10:51, 5.26it/s] 14%|█▍ | 6724/48008 [22:51<2:09:42, 5.30it/s] 14%|█▍ | 6725/48008 [22:52<2:12:21, 5.20it/s] 14%|█▍ | 6726/48008 [22:52<2:11:32, 5.23it/s] 14%|█▍ | 6727/48008 [22:52<2:10:37, 5.27it/s] 14%|█▍ | 6728/48008 [22:52<2:10:54, 5.26it/s] 14%|█▍ | 6729/48008 [22:52<2:10:31, 5.27it/s] 14%|█▍ | 6730/48008 [22:53<2:10:34, 5.27it/s] 14%|█▍ | 6731/48008 [22:53<2:12:28, 5.19it/s] 14%|█▍ | 6732/48008 [22:53<2:11:11, 5.24it/s] 14%|█▍ | 6733/48008 [22:53<2:11:19, 5.24it/s] 14%|█▍ | 6734/48008 [22:53<2:13:22, 5.16it/s] 14%|█▍ | 6735/48008 [22:54<2:12:45, 5.18it/s] 14%|█▍ | 6736/48008 [22:54<2:12:01, 5.21it/s] 14%|█▍ | 6737/48008 [22:54<2:12:08, 5.21it/s] 14%|█▍ | 6738/48008 [22:54<2:11:41, 5.22it/s] 14%|█▍ | 6739/48008 [22:54<2:12:13, 5.20it/s] 14%|█▍ | 6740/48008 [22:55<2:11:44, 5.22it/s] 14%|█▍ | 6741/48008 [22:55<2:11:56, 5.21it/s] 14%|█▍ | 6742/48008 [22:55<2:13:52, 5.14it/s] 14%|█▍ | 6743/48008 [22:55<2:13:23, 5.16it/s] 14%|█▍ | 6744/48008 [22:55<2:15:12, 5.09it/s] 14%|█▍ | 6745/48008 [22:56<2:17:34, 5.00it/s] 14%|█▍ | 6746/48008 [22:56<2:15:21, 5.08it/s] 14%|█▍ | 6747/48008 [22:56<2:17:57, 4.98it/s] 14%|█▍ | 6748/48008 [22:56<2:16:10, 5.05it/s] 14%|█▍ | 6749/48008 [22:56<2:16:28, 5.04it/s] 14%|█▍ | 6750/48008 [22:57<2:14:04, 5.13it/s] {'loss': 0.3044, 'grad_norm': 0.359375, 'learning_rate': 0.00017188385269121815, 'epoch': 0.14} 14%|█▍ | 6750/48008 [22:57<2:14:04, 5.13it/s] 14%|█▍ | 6751/48008 [22:57<2:12:34, 5.19it/s] 14%|█▍ | 6752/48008 [22:57<2:11:19, 5.24it/s] 14%|█▍ | 6753/48008 [22:57<2:11:27, 5.23it/s] 14%|█▍ | 6754/48008 [22:57<2:10:32, 5.27it/s] 14%|█▍ | 6755/48008 [22:57<2:09:49, 5.30it/s] 14%|█▍ | 6756/48008 [22:58<2:10:46, 5.26it/s] 14%|█▍ | 6757/48008 [22:58<2:13:37, 5.15it/s] 14%|█▍ | 6758/48008 [22:58<2:16:27, 5.04it/s] 14%|█▍ | 6759/48008 [22:58<2:17:22, 5.00it/s] 14%|█▍ | 6760/48008 [22:58<2:17:37, 5.00it/s] 14%|█▍ | 6761/48008 [22:59<2:17:30, 5.00it/s] 14%|█▍ | 6762/48008 [22:59<2:15:15, 5.08it/s] 14%|█▍ | 6763/48008 [22:59<2:13:56, 5.13it/s] 14%|█▍ | 6764/48008 [22:59<2:13:19, 5.16it/s] 14%|█▍ | 6765/48008 [22:59<2:12:36, 5.18it/s] 14%|█▍ | 6766/48008 [23:00<2:11:30, 5.23it/s] 14%|█▍ | 6767/48008 [23:00<2:11:05, 5.24it/s] 14%|█▍ | 6768/48008 [23:00<2:13:34, 5.15it/s] 14%|█▍ | 6769/48008 [23:00<2:13:51, 5.13it/s] 14%|█▍ | 6770/48008 [23:00<2:12:58, 5.17it/s] 14%|█▍ | 6771/48008 [23:01<2:12:16, 5.20it/s] 14%|█▍ | 6772/48008 [23:01<2:13:29, 5.15it/s] 14%|█▍ | 6773/48008 [23:01<2:13:27, 5.15it/s] 14%|█▍ | 6774/48008 [23:01<2:12:28, 5.19it/s] 14%|█▍ | 6775/48008 [23:01<2:12:01, 5.20it/s] 14%|█▍ | 6776/48008 [23:02<2:11:57, 5.21it/s] 14%|█▍ | 6777/48008 [23:02<2:11:39, 5.22it/s] 14%|█▍ | 6778/48008 [23:02<2:11:30, 5.23it/s] 14%|█▍ | 6779/48008 [23:02<2:11:22, 5.23it/s] 14%|█▍ | 6780/48008 [23:02<2:13:36, 5.14it/s] 14%|█▍ | 6781/48008 [23:03<2:16:32, 5.03it/s] 14%|█▍ | 6782/48008 [23:03<2:13:51, 5.13it/s] 14%|█▍ | 6783/48008 [23:03<2:13:08, 5.16it/s] 14%|█▍ | 6784/48008 [23:03<2:12:16, 5.19it/s] 14%|█▍ | 6785/48008 [23:03<2:10:44, 5.26it/s] 14%|█▍ | 6786/48008 [23:03<2:10:45, 5.25it/s] 14%|█▍ | 6787/48008 [23:04<2:09:10, 5.32it/s] 14%|█▍ | 6788/48008 [23:04<2:09:28, 5.31it/s] 14%|█▍ | 6789/48008 [23:04<2:09:48, 5.29it/s] 14%|█▍ | 6790/48008 [23:04<2:09:57, 5.29it/s] 14%|█▍ | 6791/48008 [23:04<2:10:11, 5.28it/s] 14%|█▍ | 6792/48008 [23:05<2:12:44, 5.17it/s] 14%|█▍ | 6793/48008 [23:05<2:16:00, 5.05it/s] 14%|█▍ | 6794/48008 [23:05<2:15:30, 5.07it/s] 14%|█▍ | 6795/48008 [23:05<2:13:18, 5.15it/s] 14%|█▍ | 6796/48008 [23:05<2:12:43, 5.18it/s] 14%|█▍ | 6797/48008 [23:06<2:12:17, 5.19it/s] 14%|█▍ | 6798/48008 [23:06<2:14:02, 5.12it/s] 14%|█▍ | 6799/48008 [23:06<2:13:00, 5.16it/s] 14%|█▍ | 6800/48008 [23:06<2:12:14, 5.19it/s] {'loss': 0.3053, 'grad_norm': 0.1552734375, 'learning_rate': 0.00017167555407432097, 'epoch': 0.14} 14%|█▍ | 6800/48008 [23:06<2:12:14, 5.19it/s] 14%|█▍ | 6801/48008 [23:06<2:12:09, 5.20it/s] 14%|█▍ | 6802/48008 [23:07<2:52:00, 3.99it/s] 14%|█▍ | 6803/48008 [23:07<2:39:58, 4.29it/s] 14%|█▍ | 6804/48008 [23:07<2:30:18, 4.57it/s] 14%|█▍ | 6805/48008 [23:07<2:25:59, 4.70it/s] 14%|█▍ | 6806/48008 [23:08<2:21:14, 4.86it/s] 14%|█▍ | 6807/48008 [23:08<2:18:11, 4.97it/s] 14%|█▍ | 6808/48008 [23:08<2:15:45, 5.06it/s] 14%|█▍ | 6809/48008 [23:08<2:15:09, 5.08it/s] 14%|█▍ | 6810/48008 [23:08<2:16:33, 5.03it/s] 14%|█▍ | 6811/48008 [23:09<2:24:36, 4.75it/s] 14%|█▍ | 6812/48008 [23:09<2:19:31, 4.92it/s] 14%|█▍ | 6813/48008 [23:09<2:16:56, 5.01it/s] 14%|█▍ | 6814/48008 [23:09<2:14:25, 5.11it/s] 14%|█▍ | 6815/48008 [23:09<2:13:30, 5.14it/s] 14%|█▍ | 6816/48008 [23:09<2:12:35, 5.18it/s] 14%|█▍ | 6817/48008 [23:10<2:11:41, 5.21it/s] 14%|█▍ | 6818/48008 [23:10<2:11:31, 5.22it/s] 14%|█▍ | 6819/48008 [23:10<2:11:25, 5.22it/s] 14%|█▍ | 6820/48008 [23:10<2:10:51, 5.25it/s] 14%|█▍ | 6821/48008 [23:10<2:10:56, 5.24it/s] 14%|█▍ | 6822/48008 [23:11<2:13:16, 5.15it/s] 14%|█▍ | 6823/48008 [23:11<2:11:20, 5.23it/s] 14%|█▍ | 6824/48008 [23:11<2:11:28, 5.22it/s] 14%|█▍ | 6825/48008 [23:11<2:10:52, 5.24it/s] 14%|█▍ | 6826/48008 [23:11<2:13:26, 5.14it/s] 14%|█▍ | 6827/48008 [23:12<2:12:20, 5.19it/s] 14%|█▍ | 6828/48008 [23:12<2:14:09, 5.12it/s] 14%|█▍ | 6829/48008 [23:12<2:15:12, 5.08it/s] 14%|█▍ | 6830/48008 [23:12<2:13:15, 5.15it/s] 14%|█▍ | 6831/48008 [23:12<2:12:34, 5.18it/s] 14%|█▍ | 6832/48008 [23:13<2:21:08, 4.86it/s] 14%|█▍ | 6833/48008 [23:13<2:18:01, 4.97it/s] 14%|█▍ | 6834/48008 [23:13<2:15:56, 5.05it/s] 14%|█▍ | 6835/48008 [23:13<2:14:41, 5.09it/s] 14%|█▍ | 6836/48008 [23:13<2:13:26, 5.14it/s] 14%|█▍ | 6837/48008 [23:14<2:14:16, 5.11it/s] 14%|█▍ | 6838/48008 [23:14<2:13:57, 5.12it/s] 14%|█▍ | 6839/48008 [23:14<2:12:04, 5.20it/s] 14%|█▍ | 6840/48008 [23:14<2:10:50, 5.24it/s] 14%|█▍ | 6841/48008 [23:14<2:11:16, 5.23it/s] 14%|█▍ | 6842/48008 [23:15<2:12:19, 5.19it/s] 14%|█▍ | 6843/48008 [23:15<2:11:44, 5.21it/s] 14%|█▍ | 6844/48008 [23:15<2:11:36, 5.21it/s] 14%|█▍ | 6845/48008 [23:15<2:11:02, 5.24it/s] 14%|█▍ | 6846/48008 [23:15<2:11:32, 5.22it/s] 14%|█▍ | 6847/48008 [23:15<2:11:33, 5.21it/s] 14%|█▍ | 6848/48008 [23:16<2:13:47, 5.13it/s] 14%|█▍ | 6849/48008 [23:16<2:12:40, 5.17it/s] 14%|█▍ | 6850/48008 [23:16<2:11:28, 5.22it/s] {'loss': 0.3041, 'grad_norm': 0.51953125, 'learning_rate': 0.00017146725545742376, 'epoch': 0.14} 14%|█▍ | 6850/48008 [23:16<2:11:28, 5.22it/s] 14%|█▍ | 6851/48008 [23:16<2:12:11, 5.19it/s] 14%|█▍ | 6852/48008 [23:16<2:11:49, 5.20it/s] 14%|█▍ | 6853/48008 [23:17<2:14:06, 5.11it/s] 14%|█▍ | 6854/48008 [23:18<5:03:59, 2.26it/s] 14%|█▍ | 6855/48008 [23:18<4:11:33, 2.73it/s] 14%|█▍ | 6856/48008 [23:18<3:34:56, 3.19it/s] 14%|█▍ | 6857/48008 [23:18<3:11:03, 3.59it/s] 14%|█▍ | 6858/48008 [23:18<2:52:51, 3.97it/s] 14%|█▍ | 6859/48008 [23:19<2:40:39, 4.27it/s] 14%|█▍ | 6860/48008 [23:20<5:01:45, 2.27it/s] 14%|█▍ | 6861/48008 [23:20<4:10:26, 2.74it/s] 14%|█▍ | 6862/48008 [23:20<3:34:07, 3.20it/s] 14%|█▍ | 6863/48008 [23:20<3:09:15, 3.62it/s] 14%|█▍ | 6864/48008 [23:20<2:53:09, 3.96it/s] 14%|█▍ | 6865/48008 [23:20<2:40:19, 4.28it/s] 14%|█▍ | 6866/48008 [23:21<4:41:39, 2.43it/s] 14%|█▍ | 6867/48008 [23:22<3:56:30, 2.90it/s] 14%|█▍ | 6868/48008 [23:22<3:25:22, 3.34it/s] 14%|█▍ | 6869/48008 [23:22<3:04:01, 3.73it/s] 14%|█▍ | 6870/48008 [23:22<2:48:21, 4.07it/s] 14%|█▍ | 6871/48008 [23:22<2:36:03, 4.39it/s] 14%|█▍ | 6872/48008 [23:23<5:52:04, 1.95it/s] 14%|█▍ | 6873/48008 [23:24<4:45:39, 2.40it/s] 14%|█▍ | 6874/48008 [23:24<3:59:11, 2.87it/s] 14%|█▍ | 6875/48008 [23:24<3:27:01, 3.31it/s] 14%|█▍ | 6876/48008 [23:24<3:04:05, 3.72it/s] 14%|█▍ | 6877/48008 [23:24<2:47:37, 4.09it/s] 14%|█▍ | 6878/48008 [23:25<2:36:55, 4.37it/s] 14%|█▍ | 6879/48008 [23:25<2:28:40, 4.61it/s] 14%|█▍ | 6880/48008 [23:25<2:23:18, 4.78it/s] 14%|█▍ | 6881/48008 [23:25<2:19:41, 4.91it/s] 14%|█▍ | 6882/48008 [23:25<2:16:46, 5.01it/s] 14%|█▍ | 6883/48008 [23:26<2:14:44, 5.09it/s] 14%|█▍ | 6884/48008 [23:26<2:12:52, 5.16it/s] 14%|█▍ | 6885/48008 [23:26<2:12:00, 5.19it/s] 14%|█▍ | 6886/48008 [23:26<2:10:52, 5.24it/s] 14%|█▍ | 6887/48008 [23:26<2:12:50, 5.16it/s] 14%|█▍ | 6888/48008 [23:27<2:11:12, 5.22it/s] 14%|█▍ | 6889/48008 [23:27<2:09:49, 5.28it/s] 14%|█▍ | 6890/48008 [23:27<2:10:41, 5.24it/s] 14%|█▍ | 6891/48008 [23:27<2:10:45, 5.24it/s] 14%|█▍ | 6892/48008 [23:27<2:12:57, 5.15it/s] 14%|█▍ | 6893/48008 [23:27<2:13:03, 5.15it/s] 14%|█▍ | 6894/48008 [23:28<2:16:27, 5.02it/s] 14%|█▍ | 6895/48008 [23:28<2:14:50, 5.08it/s] 14%|█▍ | 6896/48008 [23:28<2:14:07, 5.11it/s] 14%|█▍ | 6897/48008 [23:29<4:37:09, 2.47it/s] 14%|█▍ | 6898/48008 [23:29<3:55:10, 2.91it/s] 14%|█▍ | 6899/48008 [23:29<3:23:39, 3.36it/s] 14%|█▍ | 6900/48008 [23:30<3:02:13, 3.76it/s] 14%|█▍ | 6900/48008 [23:30<3:02:13, 3.76it/s]{'loss': 0.3052, 'grad_norm': 0.2578125, 'learning_rate': 0.0001712589568405266, 'epoch': 0.14} 14%|█▍ | 6901/48008 [23:30<2:46:52, 4.11it/s] 14%|█▍ | 6902/48008 [23:30<2:38:07, 4.33it/s] 14%|█▍ | 6903/48008 [23:30<2:27:47, 4.64it/s] 14%|█▍ | 6904/48008 [23:30<2:24:52, 4.73it/s] 14%|█▍ | 6905/48008 [23:31<2:21:58, 4.82it/s] 14%|█▍ | 6906/48008 [23:31<2:18:07, 4.96it/s] 14%|█▍ | 6907/48008 [23:31<2:16:12, 5.03it/s] 14%|█▍ | 6908/48008 [23:31<2:16:49, 5.01it/s] 14%|█▍ | 6909/48008 [23:31<2:14:09, 5.11it/s] 14%|█▍ | 6910/48008 [23:31<2:15:01, 5.07it/s] 14%|█▍ | 6911/48008 [23:32<2:11:28, 5.21it/s] 14%|█▍ | 6912/48008 [23:32<2:14:36, 5.09it/s] 14%|█▍ | 6913/48008 [23:32<2:14:10, 5.10it/s] 14%|█▍ | 6914/48008 [23:32<2:15:24, 5.06it/s] 14%|█▍ | 6915/48008 [23:32<2:15:09, 5.07it/s] 14%|█▍ | 6916/48008 [23:33<2:14:06, 5.11it/s] 14%|█▍ | 6917/48008 [23:33<2:12:14, 5.18it/s] 14%|█▍ | 6918/48008 [23:33<2:11:20, 5.21it/s] 14%|█▍ | 6919/48008 [23:33<2:11:42, 5.20it/s] 14%|█▍ | 6920/48008 [23:33<2:21:01, 4.86it/s] 14%|█▍ | 6921/48008 [23:34<2:18:09, 4.96it/s] 14%|█▍ | 6922/48008 [23:34<2:18:08, 4.96it/s] 14%|█▍ | 6923/48008 [23:34<2:17:11, 4.99it/s] 14%|█▍ | 6924/48008 [23:34<2:15:15, 5.06it/s] 14%|█▍ | 6925/48008 [23:34<2:13:01, 5.15it/s] 14%|█▍ | 6926/48008 [23:35<2:12:29, 5.17it/s] 14%|█▍ | 6927/48008 [23:35<2:12:59, 5.15it/s] 14%|█▍ | 6928/48008 [23:35<2:11:35, 5.20it/s] 14%|█▍ | 6929/48008 [23:35<2:12:52, 5.15it/s] 14%|█▍ | 6930/48008 [23:35<2:12:25, 5.17it/s] 14%|█▍ | 6931/48008 [23:36<2:13:40, 5.12it/s] 14%|█▍ | 6932/48008 [23:36<2:13:04, 5.14it/s] 14%|█▍ | 6933/48008 [23:36<2:22:06, 4.82it/s] 14%|█▍ | 6934/48008 [23:36<2:19:30, 4.91it/s] 14%|█▍ | 6935/48008 [23:36<2:16:58, 5.00it/s] 14%|█▍ | 6936/48008 [23:37<2:24:33, 4.74it/s] 14%|█▍ | 6937/48008 [23:37<2:20:39, 4.87it/s] 14%|█▍ | 6938/48008 [23:37<2:19:39, 4.90it/s] 14%|█▍ | 6939/48008 [23:37<2:16:58, 5.00it/s] 14%|█▍ | 6940/48008 [23:37<2:15:11, 5.06it/s] 14%|█▍ | 6941/48008 [23:38<2:15:00, 5.07it/s] 14%|█▍ | 6942/48008 [23:38<2:24:06, 4.75it/s] 14%|█▍ | 6943/48008 [23:38<2:20:41, 4.86it/s] 14%|█▍ | 6944/48008 [23:38<2:21:14, 4.85it/s] 14%|█▍ | 6945/48008 [23:38<2:17:14, 4.99it/s] 14%|█▍ | 6946/48008 [23:39<2:15:19, 5.06it/s] 14%|█▍ | 6947/48008 [23:39<2:13:03, 5.14it/s] 14%|█▍ | 6948/48008 [23:39<2:12:44, 5.16it/s] 14%|█▍ | 6949/48008 [23:39<2:15:03, 5.07it/s] 14%|█▍ | 6950/48008 [23:39<2:17:23, 4.98it/s] {'loss': 0.3012, 'grad_norm': 0.30859375, 'learning_rate': 0.0001710506582236294, 'epoch': 0.14} 14%|█▍ | 6950/48008 [23:39<2:17:23, 4.98it/s] 14%|█▍ | 6951/48008 [23:40<2:16:25, 5.02it/s] 14%|█▍ | 6952/48008 [23:40<2:15:31, 5.05it/s] 14%|█▍ | 6953/48008 [23:40<2:13:44, 5.12it/s] 14%|█▍ | 6954/48008 [23:40<2:14:48, 5.08it/s] 14%|█▍ | 6955/48008 [23:40<2:15:02, 5.07it/s] 14%|█▍ | 6956/48008 [23:41<2:16:17, 5.02it/s] 14%|█▍ | 6957/48008 [23:41<2:15:31, 5.05it/s] 14%|█▍ | 6958/48008 [23:41<2:14:56, 5.07it/s] 14%|█▍ | 6959/48008 [23:41<2:13:28, 5.13it/s] 14%|█▍ | 6960/48008 [23:41<2:13:17, 5.13it/s] 14%|█▍ | 6961/48008 [23:42<2:11:47, 5.19it/s] 15%|█▍ | 6962/48008 [23:42<2:11:26, 5.20it/s] 15%|█▍ | 6963/48008 [23:42<2:12:32, 5.16it/s] 15%|█▍ | 6964/48008 [23:42<2:15:47, 5.04it/s] 15%|█▍ | 6965/48008 [23:42<2:13:38, 5.12it/s] 15%|█▍ | 6966/48008 [23:43<2:12:45, 5.15it/s] 15%|█▍ | 6967/48008 [23:43<2:21:49, 4.82it/s] 15%|█▍ | 6968/48008 [23:43<2:20:25, 4.87it/s] 15%|█▍ | 6969/48008 [23:43<2:17:44, 4.97it/s] 15%|█▍ | 6970/48008 [23:43<2:17:48, 4.96it/s] 15%|█▍ | 6971/48008 [23:44<2:15:45, 5.04it/s] 15%|█▍ | 6972/48008 [23:44<2:14:00, 5.10it/s] 15%|█▍ | 6973/48008 [23:44<2:11:59, 5.18it/s] 15%|█▍ | 6974/48008 [23:44<2:15:00, 5.07it/s] 15%|█▍ | 6975/48008 [23:44<2:13:25, 5.13it/s] 15%|█▍ | 6976/48008 [23:45<2:15:06, 5.06it/s] 15%|█▍ | 6977/48008 [23:45<2:14:08, 5.10it/s] 15%|█▍ | 6978/48008 [23:45<2:16:38, 5.00it/s] 15%|█▍ | 6979/48008 [23:45<2:14:49, 5.07it/s] 15%|█▍ | 6980/48008 [23:45<2:12:28, 5.16it/s] 15%|█▍ | 6981/48008 [23:46<2:11:39, 5.19it/s] 15%|█▍ | 6982/48008 [23:46<2:11:18, 5.21it/s] 15%|█▍ | 6983/48008 [23:46<2:11:35, 5.20it/s] 15%|█▍ | 6984/48008 [23:46<2:12:59, 5.14it/s] 15%|█▍ | 6985/48008 [23:46<2:16:02, 5.03it/s] 15%|█▍ | 6986/48008 [23:47<2:54:18, 3.92it/s] 15%|█▍ | 6987/48008 [23:47<2:41:16, 4.24it/s] 15%|█▍ | 6988/48008 [23:47<2:34:38, 4.42it/s] 15%|█▍ | 6989/48008 [23:47<2:27:20, 4.64it/s] 15%|█▍ | 6990/48008 [23:47<2:23:11, 4.77it/s] 15%|█▍ | 6991/48008 [23:48<2:21:42, 4.82it/s] 15%|█▍ | 6992/48008 [23:48<2:19:54, 4.89it/s] 15%|█▍ | 6993/48008 [23:48<2:26:19, 4.67it/s] 15%|█▍ | 6994/48008 [23:48<2:21:07, 4.84it/s] 15%|█▍ | 6995/48008 [23:49<2:17:33, 4.97it/s] 15%|█▍ | 6996/48008 [23:49<2:14:27, 5.08it/s] 15%|█▍ | 6997/48008 [23:49<2:11:07, 5.21it/s] 15%|█▍ | 6998/48008 [23:49<2:11:29, 5.20it/s] 15%|█▍ | 6999/48008 [23:49<2:11:04, 5.21it/s] 15%|█▍ | 7000/48008 [23:49<2:10:51, 5.22it/s] {'loss': 0.3024, 'grad_norm': 0.55859375, 'learning_rate': 0.00017084235960673224, 'epoch': 0.15} 15%|█▍ | 7000/48008 [23:49<2:10:51, 5.22it/s] 15%|█▍ | 7001/48008 [23:50<2:09:53, 5.26it/s] 15%|█▍ | 7002/48008 [23:50<2:09:28, 5.28it/s] 15%|█▍ | 7003/48008 [23:50<2:09:57, 5.26it/s] 15%|█▍ | 7004/48008 [23:50<2:09:50, 5.26it/s] 15%|█▍ | 7005/48008 [23:50<2:10:34, 5.23it/s] 15%|█▍ | 7006/48008 [23:51<2:13:09, 5.13it/s] 15%|█▍ | 7007/48008 [23:51<2:11:33, 5.19it/s] 15%|█▍ | 7008/48008 [23:51<2:15:12, 5.05it/s] 15%|█▍ | 7009/48008 [23:51<2:17:29, 4.97it/s] 15%|█▍ | 7010/48008 [23:51<2:17:35, 4.97it/s] 15%|█▍ | 7011/48008 [23:52<2:14:22, 5.09it/s] 15%|█▍ | 7012/48008 [23:52<2:13:34, 5.12it/s] 15%|█▍ | 7013/48008 [23:52<2:13:09, 5.13it/s] 15%|█▍ | 7014/48008 [23:52<2:12:38, 5.15it/s] 15%|█▍ | 7015/48008 [23:52<2:12:16, 5.17it/s] 15%|█▍ | 7016/48008 [23:53<2:21:11, 4.84it/s] 15%|█▍ | 7017/48008 [23:53<2:21:52, 4.82it/s] 15%|█▍ | 7018/48008 [23:53<2:18:34, 4.93it/s] 15%|█▍ | 7019/48008 [23:53<2:18:21, 4.94it/s] 15%|█▍ | 7020/48008 [23:53<2:15:48, 5.03it/s] 15%|█▍ | 7021/48008 [23:54<2:18:28, 4.93it/s] 15%|█▍ | 7022/48008 [23:54<2:16:58, 4.99it/s] 15%|█▍ | 7023/48008 [23:54<2:13:26, 5.12it/s] 15%|█▍ | 7024/48008 [23:54<2:11:43, 5.19it/s] 15%|█▍ | 7025/48008 [23:54<2:11:30, 5.19it/s] 15%|█▍ | 7026/48008 [23:55<2:10:54, 5.22it/s] 15%|█▍ | 7027/48008 [23:55<2:10:54, 5.22it/s] 15%|█▍ | 7028/48008 [23:55<2:10:34, 5.23it/s] 15%|█▍ | 7029/48008 [23:55<2:11:04, 5.21it/s] 15%|█▍ | 7030/48008 [23:55<2:11:19, 5.20it/s] 15%|█▍ | 7031/48008 [23:56<2:11:25, 5.20it/s] 15%|█▍ | 7032/48008 [23:56<2:11:27, 5.20it/s] 15%|█▍ | 7033/48008 [23:56<2:34:31, 4.42it/s] 15%|█▍ | 7034/48008 [23:56<2:27:08, 4.64it/s] 15%|█▍ | 7035/48008 [23:56<2:22:12, 4.80it/s] 15%|█▍ | 7036/48008 [23:57<2:17:44, 4.96it/s] 15%|█▍ | 7037/48008 [23:57<2:15:30, 5.04it/s] 15%|█▍ | 7038/48008 [23:57<2:16:12, 5.01it/s] 15%|█▍ | 7039/48008 [23:57<2:14:40, 5.07it/s] 15%|█▍ | 7040/48008 [23:57<2:13:41, 5.11it/s] 15%|█▍ | 7041/48008 [23:58<2:11:38, 5.19it/s] 15%|█▍ | 7042/48008 [23:58<2:11:23, 5.20it/s] 15%|█▍ | 7043/48008 [23:58<2:09:39, 5.27it/s] 15%|█▍ | 7044/48008 [23:58<2:13:18, 5.12it/s] 15%|█▍ | 7045/48008 [23:58<2:11:57, 5.17it/s] 15%|█▍ | 7046/48008 [23:59<2:11:26, 5.19it/s] 15%|█▍ | 7047/48008 [23:59<2:13:30, 5.11it/s] 15%|█▍ | 7048/48008 [23:59<2:11:44, 5.18it/s] 15%|█▍ | 7049/48008 [23:59<2:11:37, 5.19it/s] 15%|█▍ | 7050/48008 [23:59<2:10:17, 5.24it/s] {'loss': 0.305, 'grad_norm': 0.36328125, 'learning_rate': 0.00017063406098983503, 'epoch': 0.15} 15%|█▍ | 7050/48008 [23:59<2:10:17, 5.24it/s] 15%|█▍ | 7051/48008 [24:00<2:34:18, 4.42it/s] 15%|█▍ | 7052/48008 [24:00<2:26:10, 4.67it/s] 15%|█▍ | 7053/48008 [24:00<2:22:34, 4.79it/s] 15%|█▍ | 7054/48008 [24:00<2:21:28, 4.82it/s] 15%|█▍ | 7055/48008 [24:00<2:18:36, 4.92it/s] 15%|█▍ | 7056/48008 [24:01<2:17:27, 4.97it/s] 15%|█▍ | 7057/48008 [24:01<2:16:43, 4.99it/s] 15%|█▍ | 7058/48008 [24:01<2:15:42, 5.03it/s] 15%|█▍ | 7059/48008 [24:01<2:16:43, 4.99it/s] 15%|█▍ | 7060/48008 [24:01<2:16:01, 5.02it/s] 15%|█▍ | 7061/48008 [24:02<2:14:09, 5.09it/s] 15%|█▍ | 7062/48008 [24:02<2:22:31, 4.79it/s] 15%|█▍ | 7063/48008 [24:02<2:20:17, 4.86it/s] 15%|█▍ | 7064/48008 [24:02<2:16:29, 5.00it/s] 15%|█▍ | 7065/48008 [24:02<2:16:41, 4.99it/s] 15%|█▍ | 7066/48008 [24:03<2:15:05, 5.05it/s] 15%|█▍ | 7067/48008 [24:03<2:15:48, 5.02it/s] 15%|█▍ | 7068/48008 [24:03<2:14:32, 5.07it/s] 15%|█▍ | 7069/48008 [24:03<2:13:45, 5.10it/s] 15%|█▍ | 7070/48008 [24:03<2:36:52, 4.35it/s] 15%|█▍ | 7071/48008 [24:04<2:30:39, 4.53it/s] 15%|█▍ | 7072/48008 [24:04<2:27:25, 4.63it/s] 15%|█▍ | 7073/48008 [24:04<2:22:03, 4.80it/s] 15%|█▍ | 7074/48008 [24:04<2:41:49, 4.22it/s] 15%|█▍ | 7075/48008 [24:05<2:32:41, 4.47it/s] 15%|█▍ | 7076/48008 [24:05<2:27:48, 4.62it/s] 15%|█▍ | 7077/48008 [24:05<2:22:03, 4.80it/s] 15%|█▍ | 7078/48008 [24:05<2:18:00, 4.94it/s] 15%|█▍ | 7079/48008 [24:05<2:16:00, 5.02it/s] 15%|█▍ | 7080/48008 [24:06<2:14:20, 5.08it/s] 15%|█▍ | 7081/48008 [24:06<2:13:08, 5.12it/s] 15%|█▍ | 7082/48008 [24:06<2:13:14, 5.12it/s] 15%|█▍ | 7083/48008 [24:06<2:12:07, 5.16it/s] 15%|█▍ | 7084/48008 [24:06<2:12:16, 5.16it/s] 15%|█▍ | 7085/48008 [24:06<2:14:23, 5.08it/s] 15%|█▍ | 7086/48008 [24:07<2:15:26, 5.04it/s] 15%|█▍ | 7087/48008 [24:07<2:37:20, 4.33it/s] 15%|█▍ | 7088/48008 [24:07<2:31:53, 4.49it/s] 15%|█▍ | 7089/48008 [24:07<2:24:28, 4.72it/s] 15%|█▍ | 7090/48008 [24:08<2:22:45, 4.78it/s] 15%|█▍ | 7091/48008 [24:08<2:18:09, 4.94it/s] 15%|█▍ | 7092/48008 [24:08<2:15:11, 5.04it/s] 15%|█▍ | 7093/48008 [24:08<2:13:30, 5.11it/s] 15%|█▍ | 7094/48008 [24:08<2:12:03, 5.16it/s] 15%|█▍ | 7095/48008 [24:09<2:14:58, 5.05it/s] 15%|█▍ | 7096/48008 [24:09<2:14:43, 5.06it/s] 15%|█▍ | 7097/48008 [24:09<2:13:45, 5.10it/s] 15%|█▍ | 7098/48008 [24:09<2:12:32, 5.14it/s] 15%|█▍ | 7099/48008 [24:09<2:12:10, 5.16it/s] 15%|█▍ | 7100/48008 [24:10<2:11:35, 5.18it/s] {'loss': 0.3034, 'grad_norm': 0.1494140625, 'learning_rate': 0.00017042576237293785, 'epoch': 0.15} 15%|█▍ | 7100/48008 [24:10<2:11:35, 5.18it/s] 15%|█▍ | 7101/48008 [24:10<2:35:19, 4.39it/s] 15%|█▍ | 7102/48008 [24:10<2:29:11, 4.57it/s] 15%|█▍ | 7103/48008 [24:10<2:23:43, 4.74it/s] 15%|█▍ | 7104/48008 [24:10<2:20:54, 4.84it/s] 15%|█▍ | 7105/48008 [24:11<2:17:02, 4.97it/s] 15%|█▍ | 7106/48008 [24:11<2:14:57, 5.05it/s] 15%|█▍ | 7107/48008 [24:11<2:13:29, 5.11it/s] 15%|█▍ | 7108/48008 [24:11<2:11:32, 5.18it/s] 15%|█▍ | 7109/48008 [24:11<2:11:43, 5.17it/s] 15%|█▍ | 7110/48008 [24:12<2:11:33, 5.18it/s] 15%|█▍ | 7111/48008 [24:12<2:11:06, 5.20it/s] 15%|█▍ | 7112/48008 [24:12<2:10:53, 5.21it/s] 15%|█▍ | 7113/48008 [24:12<2:10:29, 5.22it/s] 15%|█▍ | 7114/48008 [24:12<2:11:39, 5.18it/s] 15%|█▍ | 7115/48008 [24:13<2:13:34, 5.10it/s] 15%|█▍ | 7116/48008 [24:13<2:13:52, 5.09it/s] 15%|█▍ | 7117/48008 [24:13<2:14:15, 5.08it/s] 15%|█▍ | 7118/48008 [24:13<2:13:05, 5.12it/s] 15%|█▍ | 7119/48008 [24:13<2:14:35, 5.06it/s] 15%|█▍ | 7120/48008 [24:13<2:12:18, 5.15it/s] 15%|█▍ | 7121/48008 [24:14<2:12:19, 5.15it/s] 15%|█▍ | 7122/48008 [24:14<2:21:15, 4.82it/s] 15%|█▍ | 7123/48008 [24:14<2:27:11, 4.63it/s] 15%|█▍ | 7124/48008 [24:14<2:25:35, 4.68it/s] 15%|█▍ | 7125/48008 [24:15<2:23:36, 4.74it/s] 15%|█▍ | 7126/48008 [24:15<2:21:57, 4.80it/s] 15%|█▍ | 7127/48008 [24:15<2:27:31, 4.62it/s] 15%|█▍ | 7128/48008 [24:15<2:22:34, 4.78it/s] 15%|█▍ | 7129/48008 [24:15<2:22:20, 4.79it/s] 15%|█▍ | 7130/48008 [24:16<2:28:18, 4.59it/s] 15%|█▍ | 7131/48008 [24:16<2:21:38, 4.81it/s] 15%|█▍ | 7132/48008 [24:16<2:20:58, 4.83it/s] 15%|█▍ | 7133/48008 [24:16<2:17:29, 4.95it/s] 15%|█▍ | 7134/48008 [24:16<2:14:33, 5.06it/s] 15%|█▍ | 7135/48008 [24:17<2:17:11, 4.97it/s] 15%|█▍ | 7136/48008 [24:17<2:15:08, 5.04it/s] 15%|█▍ | 7137/48008 [24:17<2:13:59, 5.08it/s] 15%|█▍ | 7138/48008 [24:17<2:13:48, 5.09it/s] 15%|█▍ | 7139/48008 [24:17<2:12:32, 5.14it/s] 15%|█▍ | 7140/48008 [24:18<2:10:48, 5.21it/s] 15%|█▍ | 7141/48008 [24:18<2:13:07, 5.12it/s] 15%|█▍ | 7142/48008 [24:18<2:21:47, 4.80it/s] 15%|█▍ | 7143/48008 [24:18<2:57:55, 3.83it/s] 15%|█▍ | 7144/48008 [24:19<2:42:01, 4.20it/s] 15%|█▍ | 7145/48008 [24:19<2:33:39, 4.43it/s] 15%|█▍ | 7146/48008 [24:19<2:25:27, 4.68it/s] 15%|█▍ | 7147/48008 [24:19<2:21:44, 4.80it/s] 15%|█▍ | 7148/48008 [24:19<2:19:49, 4.87it/s] 15%|█▍ | 7149/48008 [24:20<2:17:09, 4.96it/s] 15%|█▍ | 7150/48008 [24:20<2:15:10, 5.04it/s] {'loss': 0.301, 'grad_norm': 0.173828125, 'learning_rate': 0.00017021746375604066, 'epoch': 0.15} 15%|█▍ | 7150/48008 [24:20<2:15:10, 5.04it/s] 15%|█▍ | 7151/48008 [24:20<2:13:39, 5.09it/s] 15%|█▍ | 7152/48008 [24:20<2:15:05, 5.04it/s] 15%|█▍ | 7153/48008 [24:20<2:14:04, 5.08it/s] 15%|█▍ | 7154/48008 [24:21<2:11:55, 5.16it/s] 15%|█▍ | 7155/48008 [24:21<2:13:30, 5.10it/s] 15%|█▍ | 7156/48008 [24:21<2:14:29, 5.06it/s] 15%|█▍ | 7157/48008 [24:21<2:14:14, 5.07it/s] 15%|█▍ | 7158/48008 [24:21<2:13:08, 5.11it/s] 15%|█▍ | 7159/48008 [24:22<2:11:04, 5.19it/s] 15%|█▍ | 7160/48008 [24:22<2:13:19, 5.11it/s] 15%|█▍ | 7161/48008 [24:22<2:11:09, 5.19it/s] 15%|█▍ | 7162/48008 [24:22<2:10:09, 5.23it/s] 15%|█▍ | 7163/48008 [24:22<2:12:36, 5.13it/s] 15%|█▍ | 7164/48008 [24:22<2:11:48, 5.16it/s] 15%|█▍ | 7165/48008 [24:23<2:15:21, 5.03it/s] 15%|█▍ | 7166/48008 [24:23<2:12:44, 5.13it/s] 15%|█▍ | 7167/48008 [24:23<2:11:12, 5.19it/s] 15%|█▍ | 7168/48008 [24:23<2:09:32, 5.25it/s] 15%|█▍ | 7169/48008 [24:23<2:09:02, 5.27it/s] 15%|█▍ | 7170/48008 [24:24<2:09:35, 5.25it/s] 15%|█▍ | 7171/48008 [24:24<2:49:53, 4.01it/s] 15%|█▍ | 7172/48008 [24:24<2:38:15, 4.30it/s] 15%|█▍ | 7173/48008 [24:24<2:29:52, 4.54it/s] 15%|█▍ | 7174/48008 [24:25<2:24:13, 4.72it/s] 15%|█▍ | 7175/48008 [24:25<2:22:39, 4.77it/s] 15%|█▍ | 7176/48008 [24:25<2:18:39, 4.91it/s] 15%|█▍ | 7177/48008 [24:25<2:16:04, 5.00it/s] 15%|█▍ | 7178/48008 [24:25<2:17:13, 4.96it/s] 15%|█▍ | 7179/48008 [24:26<2:14:17, 5.07it/s] 15%|█▍ | 7180/48008 [24:26<2:12:51, 5.12it/s] 15%|█▍ | 7181/48008 [24:26<2:11:21, 5.18it/s] 15%|█▍ | 7182/48008 [24:26<2:34:47, 4.40it/s] 15%|█▍ | 7183/48008 [24:26<2:26:35, 4.64it/s] 15%|█▍ | 7184/48008 [24:27<2:23:13, 4.75it/s] 15%|█▍ | 7185/48008 [24:27<2:20:25, 4.85it/s] 15%|█▍ | 7186/48008 [24:27<2:17:09, 4.96it/s] 15%|█▍ | 7187/48008 [24:27<2:16:10, 5.00it/s] 15%|█▍ | 7188/48008 [24:27<2:14:34, 5.06it/s] 15%|█▍ | 7189/48008 [24:28<2:15:17, 5.03it/s] 15%|█▍ | 7190/48008 [24:28<2:12:54, 5.12it/s] 15%|█▍ | 7191/48008 [24:28<2:11:02, 5.19it/s] 15%|█▍ | 7192/48008 [24:28<2:10:49, 5.20it/s] 15%|█▍ | 7193/48008 [24:28<2:11:51, 5.16it/s] 15%|█▍ | 7194/48008 [24:29<2:10:55, 5.20it/s] 15%|█▍ | 7195/48008 [24:29<2:12:53, 5.12it/s] 15%|█▍ | 7196/48008 [24:29<2:13:09, 5.11it/s] 15%|█▍ | 7197/48008 [24:29<2:11:45, 5.16it/s] 15%|█▍ | 7198/48008 [24:29<2:11:07, 5.19it/s] 15%|█▍ | 7199/48008 [24:30<2:12:47, 5.12it/s] 15%|█▍ | 7200/48008 [24:30<2:13:53, 5.08it/s] 15%|█▍ | 7200/48008 [24:30<2:13:53, 5.08it/s]{'loss': 0.3028, 'grad_norm': 0.50390625, 'learning_rate': 0.00017000916513914348, 'epoch': 0.15} 15%|█▍ | 7201/48008 [24:30<2:10:49, 5.20it/s] 15%|█▌ | 7202/48008 [24:30<2:10:43, 5.20it/s] 15%|█▌ | 7203/48008 [24:30<2:09:50, 5.24it/s] 15%|█▌ | 7204/48008 [24:30<2:08:35, 5.29it/s] 15%|█▌ | 7205/48008 [24:31<2:12:00, 5.15it/s] 15%|█▌ | 7206/48008 [24:31<2:13:53, 5.08it/s] 15%|█▌ | 7207/48008 [24:31<2:12:56, 5.12it/s] 15%|█▌ | 7208/48008 [24:31<2:14:17, 5.06it/s] 15%|█▌ | 7209/48008 [24:32<2:22:18, 4.78it/s] 15%|█▌ | 7210/48008 [24:32<2:18:00, 4.93it/s] 15%|█▌ | 7211/48008 [24:32<2:17:36, 4.94it/s] 15%|█▌ | 7212/48008 [24:32<2:15:29, 5.02it/s] 15%|█▌ | 7213/48008 [24:32<2:16:25, 4.98it/s] 15%|█▌ | 7214/48008 [24:33<2:15:54, 5.00it/s] 15%|█▌ | 7215/48008 [24:33<2:14:05, 5.07it/s] 15%|█▌ | 7216/48008 [24:33<2:13:11, 5.10it/s] 15%|█▌ | 7217/48008 [24:33<2:12:20, 5.14it/s] 15%|█▌ | 7218/48008 [24:33<2:10:05, 5.23it/s] 15%|█▌ | 7219/48008 [24:33<2:12:42, 5.12it/s] 15%|█▌ | 7220/48008 [24:34<2:12:08, 5.14it/s] 15%|█▌ | 7221/48008 [24:34<2:10:51, 5.19it/s] 15%|█▌ | 7222/48008 [24:34<2:09:28, 5.25it/s] 15%|█▌ | 7223/48008 [24:34<2:10:27, 5.21it/s] 15%|█▌ | 7224/48008 [24:34<2:10:53, 5.19it/s] 15%|█▌ | 7225/48008 [24:35<2:10:16, 5.22it/s] 15%|█▌ | 7226/48008 [24:35<2:10:36, 5.20it/s] 15%|█▌ | 7227/48008 [24:35<2:12:36, 5.13it/s] 15%|█▌ | 7228/48008 [24:35<2:12:47, 5.12it/s] 15%|█▌ | 7229/48008 [24:35<2:14:05, 5.07it/s] 15%|█▌ | 7230/48008 [24:36<2:14:58, 5.04it/s] 15%|█▌ | 7231/48008 [24:36<2:13:22, 5.10it/s] 15%|█▌ | 7232/48008 [24:36<2:12:42, 5.12it/s] 15%|█▌ | 7233/48008 [24:36<2:13:11, 5.10it/s] 15%|█▌ | 7234/48008 [24:36<2:13:19, 5.10it/s] 15%|█▌ | 7235/48008 [24:37<2:13:55, 5.07it/s] 15%|█▌ | 7236/48008 [24:37<2:11:49, 5.15it/s] 15%|█▌ | 7237/48008 [24:37<2:50:46, 3.98it/s] 15%|█▌ | 7238/48008 [24:37<2:38:30, 4.29it/s] 15%|█▌ | 7239/48008 [24:38<2:29:41, 4.54it/s] 15%|█▌ | 7240/48008 [24:38<2:22:32, 4.77it/s] 15%|█▌ | 7241/48008 [24:38<2:19:32, 4.87it/s] 15%|█▌ | 7242/48008 [24:38<2:16:46, 4.97it/s] 15%|█▌ | 7243/48008 [24:38<2:13:53, 5.07it/s] 15%|█▌ | 7244/48008 [24:39<2:12:42, 5.12it/s] 15%|█▌ | 7245/48008 [24:39<2:12:02, 5.15it/s] 15%|█▌ | 7246/48008 [24:39<2:20:46, 4.83it/s] 15%|█▌ | 7247/48008 [24:39<2:19:49, 4.86it/s] 15%|█▌ | 7248/48008 [24:39<2:17:01, 4.96it/s] 15%|█▌ | 7249/48008 [24:40<2:14:34, 5.05it/s] 15%|█▌ | 7250/48008 [24:40<2:12:50, 5.11it/s] {'loss': 0.3036, 'grad_norm': 0.1845703125, 'learning_rate': 0.0001698008665222463, 'epoch': 0.15} 15%|█▌ | 7250/48008 [24:40<2:12:50, 5.11it/s] 15%|█▌ | 7251/48008 [24:40<2:12:50, 5.11it/s] 15%|█▌ | 7252/48008 [24:40<2:11:06, 5.18it/s] 15%|█▌ | 7253/48008 [24:40<2:10:35, 5.20it/s] 15%|█▌ | 7254/48008 [24:40<2:11:58, 5.15it/s] 15%|█▌ | 7255/48008 [24:41<2:11:17, 5.17it/s] 15%|█▌ | 7256/48008 [24:41<2:11:07, 5.18it/s] 15%|█▌ | 7257/48008 [24:41<2:12:14, 5.14it/s] 15%|█▌ | 7258/48008 [24:41<2:22:25, 4.77it/s] 15%|█▌ | 7259/48008 [24:42<2:19:30, 4.87it/s] 15%|█▌ | 7260/48008 [24:42<2:16:03, 4.99it/s] 15%|█▌ | 7261/48008 [24:42<2:14:20, 5.06it/s] 15%|█▌ | 7262/48008 [24:42<2:13:19, 5.09it/s] 15%|█▌ | 7263/48008 [24:42<2:12:27, 5.13it/s] 15%|█▌ | 7264/48008 [24:42<2:15:33, 5.01it/s] 15%|█▌ | 7265/48008 [24:43<2:13:29, 5.09it/s] 15%|█▌ | 7266/48008 [24:43<2:11:49, 5.15it/s] 15%|█▌ | 7267/48008 [24:43<2:12:27, 5.13it/s] 15%|█▌ | 7268/48008 [24:43<2:11:18, 5.17it/s] 15%|█▌ | 7269/48008 [24:43<2:11:14, 5.17it/s] 15%|█▌ | 7270/48008 [24:44<2:10:33, 5.20it/s] 15%|█▌ | 7271/48008 [24:44<2:12:22, 5.13it/s] 15%|█▌ | 7272/48008 [24:44<2:13:52, 5.07it/s] 15%|█▌ | 7273/48008 [24:44<2:12:22, 5.13it/s] 15%|█▌ | 7274/48008 [24:44<2:10:46, 5.19it/s] 15%|█▌ | 7275/48008 [24:45<2:10:17, 5.21it/s] 15%|█▌ | 7276/48008 [24:45<2:11:56, 5.15it/s] 15%|█▌ | 7277/48008 [24:45<2:11:19, 5.17it/s] 15%|█▌ | 7278/48008 [24:45<2:11:37, 5.16it/s] 15%|█▌ | 7279/48008 [24:45<2:13:21, 5.09it/s] 15%|█▌ | 7280/48008 [24:46<2:14:39, 5.04it/s] 15%|█▌ | 7281/48008 [24:46<2:12:44, 5.11it/s] 15%|█▌ | 7282/48008 [24:46<2:11:22, 5.17it/s] 15%|█▌ | 7283/48008 [24:46<2:10:33, 5.20it/s] 15%|█▌ | 7284/48008 [24:46<2:10:48, 5.19it/s] 15%|█▌ | 7285/48008 [24:47<2:11:41, 5.15it/s] 15%|█▌ | 7286/48008 [24:47<2:10:04, 5.22it/s] 15%|█▌ | 7287/48008 [24:47<2:19:18, 4.87it/s] 15%|█▌ | 7288/48008 [24:47<2:15:49, 5.00it/s] 15%|█▌ | 7289/48008 [24:47<2:13:00, 5.10it/s] 15%|█▌ | 7290/48008 [24:48<2:12:06, 5.14it/s] 15%|█▌ | 7291/48008 [24:48<2:10:23, 5.20it/s] 15%|█▌ | 7292/48008 [24:48<2:09:24, 5.24it/s] 15%|█▌ | 7293/48008 [24:48<2:09:25, 5.24it/s] 15%|█▌ | 7294/48008 [24:48<2:09:39, 5.23it/s] 15%|█▌ | 7295/48008 [24:48<2:09:21, 5.25it/s] 15%|█▌ | 7296/48008 [24:49<2:09:06, 5.26it/s] 15%|█▌ | 7297/48008 [24:49<2:11:17, 5.17it/s] 15%|█▌ | 7298/48008 [24:49<2:11:34, 5.16it/s] 15%|█▌ | 7299/48008 [24:49<2:34:22, 4.39it/s] 15%|█▌ | 7300/48008 [24:50<3:06:40, 3.63it/s] {'loss': 0.3026, 'grad_norm': 0.1396484375, 'learning_rate': 0.00016959256790534912, 'epoch': 0.15} 15%|█▌ | 7300/48008 [24:50<3:06:40, 3.63it/s] 15%|█▌ | 7301/48008 [24:50<2:49:49, 4.00it/s] 15%|█▌ | 7302/48008 [24:50<2:37:22, 4.31it/s] 15%|█▌ | 7303/48008 [24:50<2:29:01, 4.55it/s] 15%|█▌ | 7304/48008 [24:51<2:23:49, 4.72it/s] 15%|█▌ | 7305/48008 [24:51<2:22:12, 4.77it/s] 15%|█▌ | 7306/48008 [24:51<2:18:24, 4.90it/s] 15%|█▌ | 7307/48008 [24:51<2:16:54, 4.95it/s] 15%|█▌ | 7308/48008 [24:51<2:14:48, 5.03it/s] 15%|█▌ | 7309/48008 [24:52<2:22:38, 4.76it/s] 15%|█▌ | 7310/48008 [24:52<2:17:56, 4.92it/s] 15%|█▌ | 7311/48008 [24:52<2:15:22, 5.01it/s] 15%|█▌ | 7312/48008 [24:52<2:16:12, 4.98it/s] 15%|█▌ | 7313/48008 [24:52<2:15:21, 5.01it/s] 15%|█▌ | 7314/48008 [24:53<2:13:51, 5.07it/s] 15%|█▌ | 7315/48008 [24:53<2:11:19, 5.16it/s] 15%|█▌ | 7316/48008 [24:53<2:08:36, 5.27it/s] 15%|█▌ | 7317/48008 [24:53<2:09:00, 5.26it/s] 15%|█▌ | 7318/48008 [24:53<2:11:22, 5.16it/s] 15%|█▌ | 7319/48008 [24:53<2:12:41, 5.11it/s] 15%|█▌ | 7320/48008 [24:54<2:10:54, 5.18it/s] 15%|█▌ | 7321/48008 [24:54<2:10:57, 5.18it/s] 15%|█▌ | 7322/48008 [24:54<2:10:10, 5.21it/s] 15%|█▌ | 7323/48008 [24:54<2:09:52, 5.22it/s] 15%|█▌ | 7324/48008 [24:54<2:11:02, 5.17it/s] 15%|█▌ | 7325/48008 [24:55<2:09:45, 5.23it/s] 15%|█▌ | 7326/48008 [24:55<2:09:09, 5.25it/s] 15%|█▌ | 7327/48008 [24:55<2:08:38, 5.27it/s] 15%|█▌ | 7328/48008 [24:55<2:09:22, 5.24it/s] 15%|█▌ | 7329/48008 [24:55<2:08:30, 5.28it/s] 15%|█▌ | 7330/48008 [24:56<2:08:36, 5.27it/s] 15%|█▌ | 7331/48008 [24:56<2:09:02, 5.25it/s] 15%|█▌ | 7332/48008 [24:56<2:11:10, 5.17it/s] 15%|█▌ | 7333/48008 [24:56<2:12:54, 5.10it/s] 15%|█▌ | 7334/48008 [24:56<2:11:58, 5.14it/s] 15%|█▌ | 7335/48008 [24:57<2:11:11, 5.17it/s] 15%|█▌ | 7336/48008 [24:57<2:10:28, 5.20it/s] 15%|█▌ | 7337/48008 [24:57<2:10:26, 5.20it/s] 15%|█▌ | 7338/48008 [24:57<2:10:14, 5.20it/s] 15%|█▌ | 7339/48008 [24:57<2:12:06, 5.13it/s] 15%|█▌ | 7340/48008 [24:58<2:10:33, 5.19it/s] 15%|█▌ | 7341/48008 [24:58<2:09:32, 5.23it/s] 15%|█▌ | 7342/48008 [24:58<2:09:35, 5.23it/s] 15%|█▌ | 7343/48008 [24:58<2:09:19, 5.24it/s] 15%|█▌ | 7344/48008 [24:58<2:11:29, 5.15it/s] 15%|█▌ | 7345/48008 [24:58<2:13:13, 5.09it/s] 15%|█▌ | 7346/48008 [24:59<2:35:16, 4.36it/s] 15%|█▌ | 7347/48008 [24:59<2:27:09, 4.61it/s] 15%|█▌ | 7348/48008 [24:59<2:21:46, 4.78it/s] 15%|█▌ | 7349/48008 [24:59<2:16:57, 4.95it/s] 15%|█▌ | 7350/48008 [25:00<2:16:11, 4.98it/s] {'loss': 0.306, 'grad_norm': 0.146484375, 'learning_rate': 0.00016938426928845193, 'epoch': 0.15} 15%|█▌ | 7350/48008 [25:00<2:16:11, 4.98it/s] 15%|█▌ | 7351/48008 [25:00<2:14:03, 5.05it/s] 15%|█▌ | 7352/48008 [25:00<2:36:21, 4.33it/s] 15%|█▌ | 7353/48008 [25:00<2:27:59, 4.58it/s] 15%|█▌ | 7354/48008 [25:00<2:31:53, 4.46it/s] 15%|█▌ | 7355/48008 [25:01<2:24:56, 4.67it/s] 15%|█▌ | 7356/48008 [25:01<2:21:25, 4.79it/s] 15%|█▌ | 7357/48008 [25:01<2:57:25, 3.82it/s] 15%|█▌ | 7358/48008 [25:01<2:45:10, 4.10it/s] 15%|█▌ | 7359/48008 [25:02<2:34:13, 4.39it/s] 15%|█▌ | 7360/48008 [25:02<2:26:51, 4.61it/s] 15%|█▌ | 7361/48008 [25:02<2:21:27, 4.79it/s] 15%|█▌ | 7362/48008 [25:02<2:27:15, 4.60it/s] 15%|█▌ | 7363/48008 [25:02<2:21:25, 4.79it/s] 15%|█▌ | 7364/48008 [25:03<2:17:48, 4.92it/s] 15%|█▌ | 7365/48008 [25:03<2:24:44, 4.68it/s] 15%|█▌ | 7366/48008 [25:03<2:23:57, 4.71it/s] 15%|█▌ | 7367/48008 [25:03<2:20:27, 4.82it/s] 15%|█▌ | 7368/48008 [25:03<2:16:37, 4.96it/s] 15%|█▌ | 7369/48008 [25:04<2:14:39, 5.03it/s] 15%|█▌ | 7370/48008 [25:04<2:14:24, 5.04it/s] 15%|█▌ | 7371/48008 [25:04<2:16:43, 4.95it/s] 15%|█▌ | 7372/48008 [25:04<2:14:12, 5.05it/s] 15%|█▌ | 7373/48008 [25:04<2:15:01, 5.02it/s] 15%|█▌ | 7374/48008 [25:05<2:15:28, 5.00it/s] 15%|█▌ | 7375/48008 [25:05<2:13:55, 5.06it/s] 15%|█▌ | 7376/48008 [25:05<2:14:57, 5.02it/s] 15%|█▌ | 7377/48008 [25:05<2:14:37, 5.03it/s] 15%|█▌ | 7378/48008 [25:05<2:12:48, 5.10it/s] 15%|█▌ | 7379/48008 [25:06<2:11:51, 5.14it/s] 15%|█▌ | 7380/48008 [25:06<2:12:17, 5.12it/s] 15%|█▌ | 7381/48008 [25:06<2:12:27, 5.11it/s] 15%|█▌ | 7382/48008 [25:06<2:13:42, 5.06it/s] 15%|█▌ | 7383/48008 [25:06<2:12:12, 5.12it/s] 15%|█▌ | 7384/48008 [25:07<2:11:24, 5.15it/s] 15%|█▌ | 7385/48008 [25:07<2:13:07, 5.09it/s] 15%|█▌ | 7386/48008 [25:07<2:13:47, 5.06it/s] 15%|█▌ | 7387/48008 [25:07<2:13:34, 5.07it/s] 15%|█▌ | 7388/48008 [25:07<2:11:04, 5.17it/s] 15%|█▌ | 7389/48008 [25:08<2:10:09, 5.20it/s] 15%|█▌ | 7390/48008 [25:08<2:10:59, 5.17it/s] 15%|█▌ | 7391/48008 [25:08<2:10:03, 5.20it/s] 15%|█▌ | 7392/48008 [25:08<2:09:49, 5.21it/s] 15%|█▌ | 7393/48008 [25:08<2:11:58, 5.13it/s] 15%|█▌ | 7394/48008 [25:09<2:10:23, 5.19it/s] 15%|█▌ | 7395/48008 [25:09<2:09:42, 5.22it/s] 15%|█▌ | 7396/48008 [25:09<2:09:09, 5.24it/s] 15%|█▌ | 7397/48008 [25:09<2:07:52, 5.29it/s] 15%|█▌ | 7398/48008 [25:09<2:12:18, 5.12it/s] 15%|█▌ | 7399/48008 [25:10<2:12:21, 5.11it/s] 15%|█▌ | 7400/48008 [25:10<2:11:05, 5.16it/s] {'loss': 0.3034, 'grad_norm': 0.203125, 'learning_rate': 0.00016917597067155475, 'epoch': 0.15} 15%|█▌ | 7400/48008 [25:10<2:11:05, 5.16it/s] 15%|█▌ | 7401/48008 [25:10<2:11:04, 5.16it/s] 15%|█▌ | 7402/48008 [25:10<2:09:42, 5.22it/s] 15%|█▌ | 7403/48008 [25:10<2:08:55, 5.25it/s] 15%|█▌ | 7404/48008 [25:10<2:10:04, 5.20it/s] 15%|█▌ | 7405/48008 [25:11<2:19:38, 4.85it/s] 15%|█▌ | 7406/48008 [25:11<2:16:29, 4.96it/s] 15%|█▌ | 7407/48008 [25:11<2:14:04, 5.05it/s] 15%|█▌ | 7408/48008 [25:11<2:16:17, 4.96it/s] 15%|█▌ | 7409/48008 [25:11<2:14:25, 5.03it/s] 15%|█▌ | 7410/48008 [25:12<2:12:35, 5.10it/s] 15%|█▌ | 7411/48008 [25:12<2:15:13, 5.00it/s] 15%|█▌ | 7412/48008 [25:12<2:12:57, 5.09it/s] 15%|█▌ | 7413/48008 [25:12<2:11:43, 5.14it/s] 15%|█▌ | 7414/48008 [25:12<2:10:21, 5.19it/s] 15%|█▌ | 7415/48008 [25:13<2:33:28, 4.41it/s] 15%|█▌ | 7416/48008 [25:13<2:26:20, 4.62it/s] 15%|█▌ | 7417/48008 [25:13<2:20:51, 4.80it/s] 15%|█▌ | 7418/48008 [25:13<2:17:22, 4.92it/s] 15%|█▌ | 7419/48008 [25:14<2:19:00, 4.87it/s] 15%|█▌ | 7420/48008 [25:14<2:15:52, 4.98it/s] 15%|█▌ | 7421/48008 [25:14<2:13:57, 5.05it/s] 15%|█▌ | 7422/48008 [25:14<2:13:21, 5.07it/s] 15%|█▌ | 7423/48008 [25:14<2:12:15, 5.11it/s] 15%|█▌ | 7424/48008 [25:15<2:10:58, 5.16it/s] 15%|█▌ | 7425/48008 [25:15<2:09:22, 5.23it/s] 15%|█▌ | 7426/48008 [25:15<2:08:57, 5.24it/s] 15%|█▌ | 7427/48008 [25:15<2:08:18, 5.27it/s] 15%|█▌ | 7428/48008 [25:15<2:10:42, 5.17it/s] 15%|█▌ | 7429/48008 [25:15<2:09:40, 5.22it/s] 15%|█▌ | 7430/48008 [25:16<2:09:33, 5.22it/s] 15%|█▌ | 7431/48008 [25:16<2:09:37, 5.22it/s] 15%|█▌ | 7432/48008 [25:16<2:19:16, 4.86it/s] 15%|█▌ | 7433/48008 [25:16<2:15:19, 5.00it/s] 15%|█▌ | 7434/48008 [25:16<2:12:59, 5.08it/s] 15%|█▌ | 7435/48008 [25:17<2:14:00, 5.05it/s] 15%|█▌ | 7436/48008 [25:17<2:14:43, 5.02it/s] 15%|█▌ | 7437/48008 [25:17<2:12:44, 5.09it/s] 15%|█▌ | 7438/48008 [25:17<2:11:56, 5.12it/s] 15%|█▌ | 7439/48008 [25:17<2:11:33, 5.14it/s] 15%|█▌ | 7440/48008 [25:18<2:10:42, 5.17it/s] 15%|█▌ | 7441/48008 [25:18<2:12:42, 5.09it/s] 16%|█▌ | 7442/48008 [25:18<2:51:14, 3.95it/s] 16%|█▌ | 7443/48008 [25:18<2:37:19, 4.30it/s] 16%|█▌ | 7444/48008 [25:19<2:31:28, 4.46it/s] 16%|█▌ | 7445/48008 [25:19<2:27:13, 4.59it/s] 16%|█▌ | 7446/48008 [25:19<2:21:38, 4.77it/s] 16%|█▌ | 7447/48008 [25:19<2:18:16, 4.89it/s] 16%|█▌ | 7448/48008 [25:19<2:15:27, 4.99it/s] 16%|█▌ | 7449/48008 [25:20<2:12:55, 5.09it/s] 16%|█▌ | 7450/48008 [25:20<2:14:12, 5.04it/s] {'loss': 0.3016, 'grad_norm': 0.2734375, 'learning_rate': 0.00016896767205465757, 'epoch': 0.16} 16%|█▌ | 7450/48008 [25:20<2:14:12, 5.04it/s] 16%|█▌ | 7451/48008 [25:20<2:12:45, 5.09it/s] 16%|█▌ | 7452/48008 [25:20<2:10:22, 5.18it/s] 16%|█▌ | 7453/48008 [25:20<2:14:00, 5.04it/s] 16%|█▌ | 7454/48008 [25:21<2:14:07, 5.04it/s] 16%|█▌ | 7455/48008 [25:21<2:11:21, 5.15it/s] 16%|█▌ | 7456/48008 [25:21<2:10:11, 5.19it/s] 16%|█▌ | 7457/48008 [25:21<2:09:31, 5.22it/s] 16%|█▌ | 7458/48008 [25:21<2:10:49, 5.17it/s] 16%|█▌ | 7459/48008 [25:22<2:19:59, 4.83it/s] 16%|█▌ | 7460/48008 [25:22<2:16:43, 4.94it/s] 16%|█▌ | 7461/48008 [25:22<2:12:59, 5.08it/s] 16%|█▌ | 7462/48008 [25:22<2:12:42, 5.09it/s] 16%|█▌ | 7463/48008 [25:22<2:13:36, 5.06it/s] 16%|█▌ | 7464/48008 [25:23<2:21:18, 4.78it/s] 16%|█▌ | 7465/48008 [25:23<2:40:47, 4.20it/s] 16%|█▌ | 7466/48008 [25:23<2:30:16, 4.50it/s] 16%|█▌ | 7467/48008 [25:23<2:33:28, 4.40it/s] 16%|█▌ | 7468/48008 [25:23<2:25:13, 4.65it/s] 16%|█▌ | 7469/48008 [25:24<2:20:03, 4.82it/s] 16%|█▌ | 7470/48008 [25:24<2:19:15, 4.85it/s] 16%|█▌ | 7471/48008 [25:24<2:16:56, 4.93it/s] 16%|█▌ | 7472/48008 [25:24<2:18:19, 4.88it/s] 16%|█▌ | 7473/48008 [25:24<2:15:47, 4.97it/s] 16%|█▌ | 7474/48008 [25:25<2:13:30, 5.06it/s] 16%|█▌ | 7475/48008 [25:25<2:12:15, 5.11it/s] 16%|█▌ | 7476/48008 [25:25<2:11:26, 5.14it/s] 16%|█▌ | 7477/48008 [25:25<2:10:54, 5.16it/s] 16%|█▌ | 7478/48008 [25:25<2:10:24, 5.18it/s] 16%|█▌ | 7479/48008 [25:26<2:09:08, 5.23it/s] 16%|█▌ | 7480/48008 [25:26<2:08:47, 5.24it/s] 16%|█▌ | 7481/48008 [25:26<2:08:11, 5.27it/s] 16%|█▌ | 7482/48008 [25:26<2:11:08, 5.15it/s] 16%|█▌ | 7483/48008 [25:26<2:10:36, 5.17it/s] 16%|█▌ | 7484/48008 [25:27<2:12:36, 5.09it/s] 16%|█▌ | 7485/48008 [25:27<2:10:32, 5.17it/s] 16%|█▌ | 7486/48008 [25:27<2:09:26, 5.22it/s] 16%|█▌ | 7487/48008 [25:27<2:09:27, 5.22it/s] 16%|█▌ | 7488/48008 [25:27<2:11:22, 5.14it/s] 16%|█▌ | 7489/48008 [25:28<2:10:58, 5.16it/s] 16%|█▌ | 7490/48008 [25:28<2:10:25, 5.18it/s] 16%|█▌ | 7491/48008 [25:28<2:19:18, 4.85it/s] 16%|█▌ | 7492/48008 [25:28<2:25:22, 4.65it/s] 16%|█▌ | 7493/48008 [25:28<2:19:35, 4.84it/s] 16%|█▌ | 7494/48008 [25:29<2:16:18, 4.95it/s] 16%|█▌ | 7495/48008 [25:29<2:13:56, 5.04it/s] 16%|█▌ | 7496/48008 [25:29<2:15:06, 5.00it/s] 16%|█▌ | 7497/48008 [25:29<2:13:01, 5.08it/s] 16%|█▌ | 7498/48008 [25:29<2:11:44, 5.12it/s] 16%|█▌ | 7499/48008 [25:30<2:10:46, 5.16it/s] 16%|█▌ | 7500/48008 [25:30<2:09:25, 5.22it/s] {'loss': 0.3044, 'grad_norm': 0.263671875, 'learning_rate': 0.00016875937343776038, 'epoch': 0.16} 16%|█▌ | 7500/48008 [25:30<2:09:25, 5.22it/s] 16%|█▌ | 7501/48008 [25:30<2:09:49, 5.20it/s] 16%|█▌ | 7502/48008 [25:30<2:08:08, 5.27it/s] 16%|█▌ | 7503/48008 [25:30<2:08:41, 5.25it/s] 16%|█▌ | 7504/48008 [25:31<2:07:56, 5.28it/s] 16%|█▌ | 7505/48008 [25:31<2:08:19, 5.26it/s] 16%|█▌ | 7506/48008 [25:31<2:10:58, 5.15it/s] 16%|█▌ | 7507/48008 [25:31<2:19:55, 4.82it/s] 16%|█▌ | 7508/48008 [25:31<2:15:26, 4.98it/s] 16%|█▌ | 7509/48008 [25:32<2:13:22, 5.06it/s] 16%|█▌ | 7510/48008 [25:32<2:11:18, 5.14it/s] 16%|█▌ | 7511/48008 [25:32<2:10:26, 5.17it/s] 16%|█▌ | 7512/48008 [25:32<2:10:58, 5.15it/s] 16%|█▌ | 7513/48008 [25:32<2:12:17, 5.10it/s] 16%|█▌ | 7514/48008 [25:33<2:20:32, 4.80it/s] 16%|█▌ | 7515/48008 [25:33<2:16:23, 4.95it/s] 16%|█▌ | 7516/48008 [25:33<2:15:19, 4.99it/s] 16%|█▌ | 7517/48008 [25:33<2:14:45, 5.01it/s] 16%|█▌ | 7518/48008 [25:33<2:15:17, 4.99it/s] 16%|█▌ | 7519/48008 [25:34<2:16:41, 4.94it/s] 16%|█▌ | 7520/48008 [25:34<2:16:31, 4.94it/s] 16%|█▌ | 7521/48008 [25:34<2:14:18, 5.02it/s] 16%|█▌ | 7522/48008 [25:34<2:16:33, 4.94it/s] 16%|█▌ | 7523/48008 [25:34<2:17:39, 4.90it/s] 16%|█▌ | 7524/48008 [25:35<2:24:53, 4.66it/s] 16%|█▌ | 7525/48008 [25:35<2:20:27, 4.80it/s] 16%|█▌ | 7526/48008 [25:35<2:19:19, 4.84it/s] 16%|█▌ | 7527/48008 [25:35<2:15:36, 4.98it/s] 16%|█▌ | 7528/48008 [25:35<2:13:29, 5.05it/s] 16%|█▌ | 7529/48008 [25:36<2:12:18, 5.10it/s] 16%|█▌ | 7530/48008 [25:36<2:11:02, 5.15it/s] 16%|█▌ | 7531/48008 [25:36<2:10:04, 5.19it/s] 16%|█▌ | 7532/48008 [25:36<2:13:34, 5.05it/s] 16%|█▌ | 7533/48008 [25:36<2:21:42, 4.76it/s] 16%|█▌ | 7534/48008 [25:37<2:17:58, 4.89it/s] 16%|█▌ | 7535/48008 [25:37<2:14:49, 5.00it/s] 16%|█▌ | 7536/48008 [25:37<2:12:52, 5.08it/s] 16%|█▌ | 7537/48008 [25:37<2:12:34, 5.09it/s] 16%|█▌ | 7538/48008 [25:37<2:20:25, 4.80it/s] 16%|█▌ | 7539/48008 [25:38<2:16:06, 4.96it/s] 16%|█▌ | 7540/48008 [25:38<2:13:38, 5.05it/s] 16%|█▌ | 7541/48008 [25:38<2:11:22, 5.13it/s] 16%|█▌ | 7542/48008 [25:38<2:11:07, 5.14it/s] 16%|█▌ | 7543/48008 [25:38<2:10:00, 5.19it/s] 16%|█▌ | 7544/48008 [25:38<2:09:01, 5.23it/s] 16%|█▌ | 7545/48008 [25:39<2:08:40, 5.24it/s] 16%|█▌ | 7546/48008 [25:39<2:08:19, 5.26it/s] 16%|█▌ | 7547/48008 [25:39<2:10:56, 5.15it/s] 16%|█▌ | 7548/48008 [25:39<2:12:27, 5.09it/s] 16%|█▌ | 7549/48008 [25:39<2:11:22, 5.13it/s] 16%|█▌ | 7550/48008 [25:40<2:12:41, 5.08it/s] {'loss': 0.3016, 'grad_norm': 0.2373046875, 'learning_rate': 0.0001685510748208632, 'epoch': 0.16} 16%|█▌ | 7550/48008 [25:40<2:12:41, 5.08it/s] 16%|█▌ | 7551/48008 [25:40<2:11:47, 5.12it/s] 16%|█▌ | 7552/48008 [25:40<2:11:36, 5.12it/s] 16%|█▌ | 7553/48008 [25:40<2:10:44, 5.16it/s] 16%|█▌ | 7554/48008 [25:40<2:10:58, 5.15it/s] 16%|█▌ | 7555/48008 [25:41<2:11:18, 5.13it/s] 16%|█▌ | 7556/48008 [25:41<2:11:57, 5.11it/s] 16%|█▌ | 7557/48008 [25:41<2:12:19, 5.09it/s] 16%|█▌ | 7558/48008 [25:41<2:11:23, 5.13it/s] 16%|█▌ | 7559/48008 [25:41<2:13:19, 5.06it/s] 16%|█▌ | 7560/48008 [25:42<2:10:24, 5.17it/s] 16%|█▌ | 7561/48008 [25:42<2:11:23, 5.13it/s] 16%|█▌ | 7562/48008 [25:42<2:10:57, 5.15it/s] 16%|█▌ | 7563/48008 [25:42<2:10:21, 5.17it/s] 16%|█▌ | 7564/48008 [25:42<2:12:07, 5.10it/s] 16%|█▌ | 7565/48008 [25:43<2:11:31, 5.13it/s] 16%|█▌ | 7566/48008 [25:43<2:11:16, 5.13it/s] 16%|█▌ | 7567/48008 [25:43<2:09:35, 5.20it/s] 16%|█▌ | 7568/48008 [25:43<2:09:40, 5.20it/s] 16%|█▌ | 7569/48008 [25:43<2:09:22, 5.21it/s] 16%|█▌ | 7570/48008 [25:44<2:11:27, 5.13it/s] 16%|█▌ | 7571/48008 [25:44<2:10:23, 5.17it/s] 16%|█▌ | 7572/48008 [25:44<2:08:39, 5.24it/s] 16%|█▌ | 7573/48008 [25:44<2:18:17, 4.87it/s] 16%|█▌ | 7574/48008 [25:44<2:15:22, 4.98it/s] 16%|█▌ | 7575/48008 [25:45<2:13:13, 5.06it/s] 16%|█▌ | 7576/48008 [25:45<2:11:08, 5.14it/s] 16%|█▌ | 7577/48008 [25:45<2:10:56, 5.15it/s] 16%|█▌ | 7578/48008 [25:45<2:10:06, 5.18it/s] 16%|█▌ | 7579/48008 [25:45<2:09:03, 5.22it/s] 16%|█▌ | 7580/48008 [25:46<2:08:54, 5.23it/s] 16%|█▌ | 7581/48008 [25:46<2:08:12, 5.26it/s] 16%|█▌ | 7582/48008 [25:46<2:08:35, 5.24it/s] 16%|█▌ | 7583/48008 [25:46<2:09:02, 5.22it/s] 16%|█▌ | 7584/48008 [25:46<2:08:47, 5.23it/s] 16%|█▌ | 7585/48008 [25:46<2:10:52, 5.15it/s] 16%|█▌ | 7586/48008 [25:47<2:11:37, 5.12it/s] 16%|█▌ | 7587/48008 [25:47<2:11:36, 5.12it/s] 16%|█▌ | 7588/48008 [25:47<2:10:47, 5.15it/s] 16%|█▌ | 7589/48008 [25:47<2:09:50, 5.19it/s] 16%|█▌ | 7590/48008 [25:47<2:09:14, 5.21it/s] 16%|█▌ | 7591/48008 [25:48<2:11:28, 5.12it/s] 16%|█▌ | 7592/48008 [25:48<2:12:56, 5.07it/s] 16%|█▌ | 7593/48008 [25:48<2:11:30, 5.12it/s] 16%|█▌ | 7594/48008 [25:48<2:10:54, 5.15it/s] 16%|█▌ | 7595/48008 [25:48<2:09:19, 5.21it/s] 16%|█▌ | 7596/48008 [25:49<2:08:37, 5.24it/s] 16%|█▌ | 7597/48008 [25:49<2:08:13, 5.25it/s] 16%|█▌ | 7598/48008 [25:49<2:07:00, 5.30it/s] 16%|█▌ | 7599/48008 [25:49<2:08:55, 5.22it/s] 16%|█▌ | 7600/48008 [25:49<2:09:06, 5.22it/s] {'loss': 0.3035, 'grad_norm': 0.33203125, 'learning_rate': 0.00016834277620396602, 'epoch': 0.16} 16%|█▌ | 7600/48008 [25:49<2:09:06, 5.22it/s] 16%|█▌ | 7601/48008 [25:50<2:09:36, 5.20it/s] 16%|█▌ | 7602/48008 [25:50<2:11:28, 5.12it/s] 16%|█▌ | 7603/48008 [25:50<2:10:42, 5.15it/s] 16%|█▌ | 7604/48008 [25:50<2:13:56, 5.03it/s] 16%|█▌ | 7605/48008 [25:50<2:11:46, 5.11it/s] 16%|█▌ | 7606/48008 [25:51<2:13:12, 5.05it/s] 16%|█▌ | 7607/48008 [25:51<2:11:40, 5.11it/s] 16%|█▌ | 7608/48008 [25:51<2:09:53, 5.18it/s] 16%|█▌ | 7609/48008 [25:51<2:09:48, 5.19it/s] 16%|█▌ | 7610/48008 [25:51<2:11:44, 5.11it/s] 16%|█▌ | 7611/48008 [25:52<2:11:44, 5.11it/s] 16%|█▌ | 7612/48008 [25:52<2:10:22, 5.16it/s] 16%|█▌ | 7613/48008 [25:52<2:09:36, 5.19it/s] 16%|█▌ | 7614/48008 [25:52<2:08:31, 5.24it/s] 16%|█▌ | 7615/48008 [25:52<2:17:58, 4.88it/s] 16%|█▌ | 7616/48008 [25:53<2:15:09, 4.98it/s] 16%|█▌ | 7617/48008 [25:53<2:12:54, 5.06it/s] 16%|█▌ | 7618/48008 [25:53<2:10:53, 5.14it/s] 16%|█▌ | 7619/48008 [25:53<2:10:48, 5.15it/s] 16%|█▌ | 7620/48008 [25:53<2:09:29, 5.20it/s] 16%|█▌ | 7621/48008 [25:53<2:13:06, 5.06it/s] 16%|█▌ | 7622/48008 [25:54<2:13:53, 5.03it/s] 16%|█▌ | 7623/48008 [25:54<2:12:22, 5.08it/s] 16%|█▌ | 7624/48008 [25:54<2:11:32, 5.12it/s] 16%|█▌ | 7625/48008 [25:54<2:11:58, 5.10it/s] 16%|█▌ | 7626/48008 [25:54<2:09:57, 5.18it/s] 16%|█▌ | 7627/48008 [25:55<2:19:05, 4.84it/s] 16%|█▌ | 7628/48008 [25:55<2:15:43, 4.96it/s] 16%|█▌ | 7629/48008 [25:55<2:13:30, 5.04it/s] 16%|█▌ | 7630/48008 [25:55<2:12:03, 5.10it/s] 16%|█▌ | 7631/48008 [25:55<2:09:39, 5.19it/s] 16%|█▌ | 7632/48008 [25:56<2:11:09, 5.13it/s] 16%|█▌ | 7633/48008 [25:56<2:33:42, 4.38it/s] 16%|█▌ | 7634/48008 [25:56<2:26:50, 4.58it/s] 16%|█▌ | 7635/48008 [25:56<2:21:03, 4.77it/s] 16%|█▌ | 7636/48008 [25:57<2:17:21, 4.90it/s] 16%|█▌ | 7637/48008 [25:57<2:14:43, 4.99it/s] 16%|█▌ | 7638/48008 [25:57<2:10:30, 5.16it/s] 16%|█▌ | 7639/48008 [25:57<2:09:22, 5.20it/s] 16%|█▌ | 7640/48008 [25:57<2:08:48, 5.22it/s] 16%|█▌ | 7641/48008 [25:58<2:48:00, 4.00it/s] 16%|█▌ | 7642/48008 [25:58<2:37:36, 4.27it/s] 16%|█▌ | 7643/48008 [25:58<2:28:54, 4.52it/s] 16%|█▌ | 7644/48008 [25:58<2:32:05, 4.42it/s] 16%|█▌ | 7645/48008 [25:58<2:25:09, 4.63it/s] 16%|█▌ | 7646/48008 [25:59<2:22:34, 4.72it/s] 16%|█▌ | 7647/48008 [25:59<2:20:28, 4.79it/s] 16%|█▌ | 7648/48008 [25:59<2:15:58, 4.95it/s] 16%|█▌ | 7649/48008 [25:59<2:13:07, 5.05it/s] 16%|█▌ | 7650/48008 [25:59<2:12:41, 5.07it/s] {'loss': 0.3025, 'grad_norm': 1.0546875, 'learning_rate': 0.00016813447758706884, 'epoch': 0.16} 16%|█▌ | 7650/48008 [25:59<2:12:41, 5.07it/s] 16%|█▌ | 7651/48008 [26:00<2:11:54, 5.10it/s] 16%|█▌ | 7652/48008 [26:00<2:10:58, 5.14it/s] 16%|█▌ | 7653/48008 [26:00<2:10:26, 5.16it/s] 16%|█▌ | 7654/48008 [26:00<2:11:23, 5.12it/s] 16%|█▌ | 7655/48008 [26:00<2:10:22, 5.16it/s] 16%|█▌ | 7656/48008 [26:01<2:49:09, 3.98it/s] 16%|█▌ | 7657/48008 [26:01<2:40:31, 4.19it/s] 16%|█▌ | 7658/48008 [26:01<2:32:54, 4.40it/s] 16%|█▌ | 7659/48008 [26:01<2:25:14, 4.63it/s] 16%|█▌ | 7660/48008 [26:02<2:19:50, 4.81it/s] 16%|█▌ | 7661/48008 [26:02<2:16:27, 4.93it/s] 16%|█▌ | 7662/48008 [26:02<2:14:08, 5.01it/s] 16%|█▌ | 7663/48008 [26:02<2:12:03, 5.09it/s] 16%|█▌ | 7664/48008 [26:02<2:11:05, 5.13it/s] 16%|█▌ | 7665/48008 [26:03<2:11:58, 5.09it/s] 16%|█▌ | 7666/48008 [26:03<2:10:50, 5.14it/s] 16%|█▌ | 7667/48008 [26:03<2:09:10, 5.20it/s] 16%|█▌ | 7668/48008 [26:03<2:08:57, 5.21it/s] 16%|█▌ | 7669/48008 [26:03<2:08:29, 5.23it/s] 16%|█▌ | 7670/48008 [26:04<2:08:34, 5.23it/s] 16%|█▌ | 7671/48008 [26:04<2:08:45, 5.22it/s] 16%|█▌ | 7672/48008 [26:04<2:08:42, 5.22it/s] 16%|█▌ | 7673/48008 [26:04<2:07:29, 5.27it/s] 16%|█▌ | 7674/48008 [26:04<2:08:23, 5.24it/s] 16%|█▌ | 7675/48008 [26:04<2:08:29, 5.23it/s] 16%|█▌ | 7676/48008 [26:05<2:07:29, 5.27it/s] 16%|█▌ | 7677/48008 [26:05<2:11:05, 5.13it/s] 16%|█▌ | 7678/48008 [26:05<2:12:54, 5.06it/s] 16%|█▌ | 7679/48008 [26:05<2:13:02, 5.05it/s] 16%|█▌ | 7680/48008 [26:05<2:13:46, 5.02it/s] 16%|█▌ | 7681/48008 [26:06<2:11:22, 5.12it/s] 16%|█▌ | 7682/48008 [26:06<2:09:30, 5.19it/s] 16%|█▌ | 7683/48008 [26:06<2:08:58, 5.21it/s] 16%|█▌ | 7684/48008 [26:06<2:08:10, 5.24it/s] 16%|█▌ | 7685/48008 [26:06<2:10:40, 5.14it/s] 16%|█▌ | 7686/48008 [26:07<2:09:46, 5.18it/s] 16%|█▌ | 7687/48008 [26:07<2:08:46, 5.22it/s] 16%|█▌ | 7688/48008 [26:07<2:08:24, 5.23it/s] 16%|█▌ | 7689/48008 [26:07<2:08:59, 5.21it/s] 16%|█▌ | 7690/48008 [26:07<2:09:01, 5.21it/s] 16%|█▌ | 7691/48008 [26:08<2:08:41, 5.22it/s] 16%|█▌ | 7692/48008 [26:08<2:10:54, 5.13it/s] 16%|█▌ | 7693/48008 [26:08<2:10:54, 5.13it/s] 16%|█▌ | 7694/48008 [26:08<2:10:03, 5.17it/s] 16%|█▌ | 7695/48008 [26:08<2:11:34, 5.11it/s] 16%|█▌ | 7696/48008 [26:09<2:10:58, 5.13it/s] 16%|█▌ | 7697/48008 [26:09<2:10:26, 5.15it/s] 16%|█▌ | 7698/48008 [26:09<2:09:51, 5.17it/s] 16%|█▌ | 7699/48008 [26:09<2:09:34, 5.18it/s] 16%|█▌ | 7700/48008 [26:09<2:09:12, 5.20it/s] {'loss': 0.303, 'grad_norm': 0.333984375, 'learning_rate': 0.00016792617897017165, 'epoch': 0.16} 16%|█▌ | 7700/48008 [26:09<2:09:12, 5.20it/s] 16%|█▌ | 7701/48008 [26:09<2:09:00, 5.21it/s] 16%|█▌ | 7702/48008 [26:10<2:08:39, 5.22it/s] 16%|█▌ | 7703/48008 [26:10<2:08:53, 5.21it/s] 16%|█▌ | 7704/48008 [26:10<2:07:58, 5.25it/s] 16%|█▌ | 7705/48008 [26:10<2:07:53, 5.25it/s] 16%|█▌ | 7706/48008 [26:10<2:07:10, 5.28it/s] 16%|█▌ | 7707/48008 [26:11<2:07:16, 5.28it/s] 16%|█▌ | 7708/48008 [26:11<2:07:32, 5.27it/s] 16%|█▌ | 7709/48008 [26:11<2:07:40, 5.26it/s] 16%|█▌ | 7710/48008 [26:11<2:07:17, 5.28it/s] 16%|█▌ | 7711/48008 [26:11<2:07:02, 5.29it/s] 16%|█▌ | 7712/48008 [26:12<2:06:35, 5.31it/s] 16%|█▌ | 7713/48008 [26:12<2:10:07, 5.16it/s] 16%|█▌ | 7714/48008 [26:12<2:09:23, 5.19it/s] 16%|█▌ | 7715/48008 [26:12<2:31:44, 4.43it/s] 16%|█▌ | 7716/48008 [26:12<2:24:28, 4.65it/s] 16%|█▌ | 7717/48008 [26:13<2:28:53, 4.51it/s] 16%|█▌ | 7718/48008 [26:13<2:24:14, 4.66it/s] 16%|█▌ | 7719/48008 [26:13<2:19:41, 4.81it/s] 16%|█▌ | 7720/48008 [26:13<2:15:18, 4.96it/s] 16%|█▌ | 7721/48008 [26:13<2:12:07, 5.08it/s] 16%|█▌ | 7722/48008 [26:14<2:12:05, 5.08it/s] 16%|█▌ | 7723/48008 [26:14<2:10:05, 5.16it/s] 16%|█▌ | 7724/48008 [26:14<2:09:36, 5.18it/s] 16%|█▌ | 7725/48008 [26:14<2:10:48, 5.13it/s] 16%|█▌ | 7726/48008 [26:14<2:09:16, 5.19it/s] 16%|█▌ | 7727/48008 [26:15<2:07:56, 5.25it/s] 16%|█▌ | 7728/48008 [26:15<2:07:02, 5.28it/s] 16%|█▌ | 7729/48008 [26:15<2:07:16, 5.27it/s] 16%|█▌ | 7730/48008 [26:15<2:07:38, 5.26it/s] 16%|█▌ | 7731/48008 [26:15<2:06:53, 5.29it/s] 16%|█▌ | 7732/48008 [26:16<2:46:32, 4.03it/s] 16%|█▌ | 7733/48008 [26:16<2:44:03, 4.09it/s] 16%|█▌ | 7734/48008 [26:16<2:32:43, 4.39it/s] 16%|█▌ | 7735/48008 [26:16<2:26:48, 4.57it/s] 16%|█▌ | 7736/48008 [26:17<2:21:19, 4.75it/s] 16%|█▌ | 7737/48008 [26:17<2:16:47, 4.91it/s] 16%|█▌ | 7738/48008 [26:17<2:17:01, 4.90it/s] 16%|█▌ | 7739/48008 [26:17<2:13:32, 5.03it/s] 16%|█▌ | 7740/48008 [26:17<2:35:17, 4.32it/s] 16%|█▌ | 7741/48008 [26:18<2:27:58, 4.54it/s] 16%|█▌ | 7742/48008 [26:18<2:20:46, 4.77it/s] 16%|█▌ | 7743/48008 [26:18<2:17:35, 4.88it/s] 16%|█▌ | 7744/48008 [26:18<2:14:55, 4.97it/s] 16%|█▌ | 7745/48008 [26:18<2:11:49, 5.09it/s] 16%|█▌ | 7746/48008 [26:19<2:10:53, 5.13it/s] 16%|█▌ | 7747/48008 [26:19<2:08:50, 5.21it/s] 16%|█▌ | 7748/48008 [26:19<2:07:45, 5.25it/s] 16%|█▌ | 7749/48008 [26:19<2:07:37, 5.26it/s] 16%|█▌ | 7750/48008 [26:19<2:07:23, 5.27it/s] {'loss': 0.3054, 'grad_norm': 0.15625, 'learning_rate': 0.00016771788035327447, 'epoch': 0.16} 16%|█▌ | 7750/48008 [26:19<2:07:23, 5.27it/s] 16%|█▌ | 7751/48008 [26:20<2:07:30, 5.26it/s] 16%|█▌ | 7752/48008 [26:20<2:06:16, 5.31it/s] 16%|█▌ | 7753/48008 [26:20<2:07:12, 5.27it/s] 16%|█▌ | 7754/48008 [26:20<2:06:22, 5.31it/s] 16%|█▌ | 7755/48008 [26:20<2:09:14, 5.19it/s] 16%|█▌ | 7756/48008 [26:20<2:09:36, 5.18it/s] 16%|█▌ | 7757/48008 [26:21<2:08:16, 5.23it/s] 16%|█▌ | 7758/48008 [26:21<2:12:10, 5.08it/s] 16%|█▌ | 7759/48008 [26:21<2:10:49, 5.13it/s] 16%|█▌ | 7760/48008 [26:21<2:13:37, 5.02it/s] 16%|█▌ | 7761/48008 [26:21<2:12:27, 5.06it/s] 16%|█▌ | 7762/48008 [26:22<2:13:08, 5.04it/s] 16%|█▌ | 7763/48008 [26:22<2:13:58, 5.01it/s] 16%|█▌ | 7764/48008 [26:22<2:14:35, 4.98it/s] 16%|█▌ | 7765/48008 [26:22<2:12:11, 5.07it/s] 16%|█▌ | 7766/48008 [26:23<2:50:17, 3.94it/s] 16%|█▌ | 7767/48008 [26:23<2:38:13, 4.24it/s] 16%|█▌ | 7768/48008 [26:23<2:52:15, 3.89it/s] 16%|█▌ | 7769/48008 [26:23<2:38:53, 4.22it/s] 16%|█▌ | 7770/48008 [26:24<2:31:53, 4.42it/s] 16%|█▌ | 7771/48008 [26:24<2:24:37, 4.64it/s] 16%|█▌ | 7772/48008 [26:24<2:19:11, 4.82it/s] 16%|█▌ | 7773/48008 [26:24<2:15:59, 4.93it/s] 16%|█▌ | 7774/48008 [26:24<2:14:33, 4.98it/s] 16%|█▌ | 7775/48008 [26:25<2:11:36, 5.09it/s] 16%|█▌ | 7776/48008 [26:25<2:10:28, 5.14it/s] 16%|█▌ | 7777/48008 [26:25<2:19:13, 4.82it/s] 16%|█▌ | 7778/48008 [26:25<2:38:44, 4.22it/s] 16%|█▌ | 7779/48008 [26:25<2:29:02, 4.50it/s] 16%|█▌ | 7780/48008 [26:26<2:46:04, 4.04it/s] 16%|█▌ | 7781/48008 [26:26<2:34:50, 4.33it/s] 16%|█▌ | 7782/48008 [26:26<2:26:46, 4.57it/s] 16%|█▌ | 7783/48008 [26:26<2:22:20, 4.71it/s] 16%|█▌ | 7784/48008 [26:27<2:41:17, 4.16it/s] 16%|█▌ | 7785/48008 [26:27<2:30:06, 4.47it/s] 16%|█▌ | 7786/48008 [26:27<2:22:59, 4.69it/s] 16%|█▌ | 7787/48008 [26:27<2:19:47, 4.80it/s] 16%|█▌ | 7788/48008 [26:27<2:17:07, 4.89it/s] 16%|█▌ | 7789/48008 [26:28<2:14:39, 4.98it/s] 16%|█▌ | 7790/48008 [26:28<2:14:02, 5.00it/s] 16%|█▌ | 7791/48008 [26:28<2:11:08, 5.11it/s] 16%|█▌ | 7792/48008 [26:28<2:10:25, 5.14it/s] 16%|█▌ | 7793/48008 [26:28<2:10:12, 5.15it/s] 16%|█▌ | 7794/48008 [26:29<2:10:10, 5.15it/s] 16%|█▌ | 7795/48008 [26:29<2:10:26, 5.14it/s] 16%|█▌ | 7796/48008 [26:29<2:11:28, 5.10it/s] 16%|█▌ | 7797/48008 [26:29<2:10:43, 5.13it/s] 16%|█▌ | 7798/48008 [26:29<2:11:45, 5.09it/s] 16%|█▌ | 7799/48008 [26:30<2:19:57, 4.79it/s] 16%|█▌ | 7800/48008 [26:30<2:16:54, 4.89it/s] {'loss': 0.3029, 'grad_norm': 0.353515625, 'learning_rate': 0.00016750958173637726, 'epoch': 0.16} 16%|█▌ | 7800/48008 [26:30<2:16:54, 4.89it/s] 16%|█▌ | 7801/48008 [26:30<2:13:29, 5.02it/s] 16%|█▋ | 7802/48008 [26:30<2:11:49, 5.08it/s] 16%|█▋ | 7803/48008 [26:30<2:11:36, 5.09it/s] 16%|█▋ | 7804/48008 [26:31<2:13:12, 5.03it/s] 16%|█▋ | 7805/48008 [26:31<2:14:27, 4.98it/s] 16%|█▋ | 7806/48008 [26:31<2:11:53, 5.08it/s] 16%|█▋ | 7807/48008 [26:31<2:10:27, 5.14it/s] 16%|█▋ | 7808/48008 [26:31<2:08:50, 5.20it/s] 16%|█▋ | 7809/48008 [26:32<2:07:36, 5.25it/s] 16%|█▋ | 7810/48008 [26:32<2:08:09, 5.23it/s] 16%|█▋ | 7811/48008 [26:32<2:07:40, 5.25it/s] 16%|█▋ | 7812/48008 [26:32<2:08:03, 5.23it/s] 16%|█▋ | 7813/48008 [26:32<2:31:17, 4.43it/s] 16%|█▋ | 7814/48008 [26:33<2:47:21, 4.00it/s] 16%|█▋ | 7815/48008 [26:33<2:34:31, 4.33it/s] 16%|█▋ | 7816/48008 [26:33<2:26:39, 4.57it/s] 16%|█▋ | 7817/48008 [26:33<2:21:10, 4.75it/s] 16%|█▋ | 7818/48008 [26:34<2:56:30, 3.79it/s] 16%|█▋ | 7819/48008 [26:34<2:41:59, 4.13it/s] 16%|█▋ | 7820/48008 [26:34<2:32:28, 4.39it/s] 16%|█▋ | 7821/48008 [26:34<2:25:10, 4.61it/s] 16%|█▋ | 7822/48008 [26:34<2:22:30, 4.70it/s] 16%|█▋ | 7823/48008 [26:35<2:18:06, 4.85it/s] 16%|█▋ | 7824/48008 [26:35<2:14:59, 4.96it/s] 16%|█▋ | 7825/48008 [26:35<2:12:18, 5.06it/s] 16%|█▋ | 7826/48008 [26:35<2:08:54, 5.20it/s] 16%|█▋ | 7827/48008 [26:35<2:18:15, 4.84it/s] 16%|█▋ | 7828/48008 [26:36<2:17:49, 4.86it/s] 16%|█▋ | 7829/48008 [26:36<2:14:16, 4.99it/s] 16%|█▋ | 7830/48008 [26:36<2:12:00, 5.07it/s] 16%|█▋ | 7831/48008 [26:36<2:11:42, 5.08it/s] 16%|█▋ | 7832/48008 [26:36<2:10:08, 5.15it/s] 16%|█▋ | 7833/48008 [26:37<2:10:22, 5.14it/s] 16%|█▋ | 7834/48008 [26:37<2:09:12, 5.18it/s] 16%|█▋ | 7835/48008 [26:37<2:08:49, 5.20it/s] 16%|█▋ | 7836/48008 [26:37<2:09:28, 5.17it/s] 16%|█▋ | 7837/48008 [26:37<2:08:41, 5.20it/s] 16%|█▋ | 7838/48008 [26:38<2:09:01, 5.19it/s] 16%|█▋ | 7839/48008 [26:38<2:06:50, 5.28it/s] 16%|█▋ | 7840/48008 [26:38<2:09:08, 5.18it/s] 16%|█▋ | 7841/48008 [26:38<2:18:01, 4.85it/s] 16%|█▋ | 7842/48008 [26:38<2:15:09, 4.95it/s] 16%|█▋ | 7843/48008 [26:39<2:13:05, 5.03it/s] 16%|█▋ | 7844/48008 [26:39<2:20:30, 4.76it/s] 16%|█▋ | 7845/48008 [26:39<2:16:57, 4.89it/s] 16%|█▋ | 7846/48008 [26:39<2:16:32, 4.90it/s] 16%|█▋ | 7847/48008 [26:39<2:17:13, 4.88it/s] 16%|█▋ | 7848/48008 [26:40<2:15:04, 4.96it/s] 16%|█▋ | 7849/48008 [26:40<2:13:19, 5.02it/s] 16%|█▋ | 7850/48008 [26:40<2:12:32, 5.05it/s] {'loss': 0.3025, 'grad_norm': 0.14453125, 'learning_rate': 0.0001673012831194801, 'epoch': 0.16} 16%|█▋ | 7850/48008 [26:40<2:12:32, 5.05it/s] 16%|█▋ | 7851/48008 [26:40<2:10:31, 5.13it/s] 16%|█▋ | 7852/48008 [26:40<2:11:01, 5.11it/s] 16%|█▋ | 7853/48008 [26:41<2:12:10, 5.06it/s] 16%|█▋ | 7854/48008 [26:41<2:10:48, 5.12it/s] 16%|█▋ | 7855/48008 [26:41<2:48:58, 3.96it/s] 16%|█▋ | 7856/48008 [26:41<2:45:42, 4.04it/s] 16%|█▋ | 7857/48008 [26:42<2:33:26, 4.36it/s] 16%|█▋ | 7858/48008 [26:42<2:27:42, 4.53it/s] 16%|█▋ | 7859/48008 [26:42<2:30:57, 4.43it/s] 16%|█▋ | 7860/48008 [26:42<2:22:40, 4.69it/s] 16%|█▋ | 7861/48008 [26:42<2:19:15, 4.80it/s] 16%|█▋ | 7862/48008 [26:43<2:15:55, 4.92it/s] 16%|█▋ | 7863/48008 [26:43<2:13:29, 5.01it/s] 16%|█▋ | 7864/48008 [26:43<2:21:06, 4.74it/s] 16%|█▋ | 7865/48008 [26:43<2:17:14, 4.87it/s] 16%|█▋ | 7866/48008 [26:43<2:12:09, 5.06it/s] 16%|█▋ | 7867/48008 [26:44<2:10:02, 5.14it/s] 16%|█▋ | 7868/48008 [26:44<2:08:30, 5.21it/s] 16%|█▋ | 7869/48008 [26:44<2:08:09, 5.22it/s] 16%|█▋ | 7870/48008 [26:44<2:08:46, 5.19it/s] 16%|█▋ | 7871/48008 [26:44<2:08:15, 5.22it/s] 16%|█▋ | 7872/48008 [26:45<2:17:26, 4.87it/s] 16%|█▋ | 7873/48008 [26:45<2:14:35, 4.97it/s] 16%|█▋ | 7874/48008 [26:45<2:10:06, 5.14it/s] 16%|█▋ | 7875/48008 [26:45<2:10:33, 5.12it/s] 16%|█▋ | 7876/48008 [26:45<2:09:19, 5.17it/s] 16%|█▋ | 7877/48008 [26:46<2:11:03, 5.10it/s] 16%|█▋ | 7878/48008 [26:46<2:10:34, 5.12it/s] 16%|█▋ | 7879/48008 [26:46<2:12:04, 5.06it/s] 16%|█▋ | 7880/48008 [26:46<2:10:59, 5.11it/s] 16%|█▋ | 7881/48008 [26:46<2:09:35, 5.16it/s] 16%|█▋ | 7882/48008 [26:46<2:09:11, 5.18it/s] 16%|█▋ | 7883/48008 [26:47<2:08:46, 5.19it/s] 16%|█▋ | 7884/48008 [26:47<2:08:28, 5.21it/s] 16%|█▋ | 7885/48008 [26:47<2:07:26, 5.25it/s] 16%|█▋ | 7886/48008 [26:47<2:07:06, 5.26it/s] 16%|█▋ | 7887/48008 [26:48<2:30:17, 4.45it/s] 16%|█▋ | 7888/48008 [26:48<2:32:40, 4.38it/s] 16%|█▋ | 7889/48008 [26:48<2:26:25, 4.57it/s] 16%|█▋ | 7890/48008 [26:48<2:20:29, 4.76it/s] 16%|█▋ | 7891/48008 [26:48<2:16:42, 4.89it/s] 16%|█▋ | 7892/48008 [26:49<2:17:49, 4.85it/s] 16%|█▋ | 7893/48008 [26:49<2:14:42, 4.96it/s] 16%|█▋ | 7894/48008 [26:49<2:12:36, 5.04it/s] 16%|█▋ | 7895/48008 [26:49<2:10:49, 5.11it/s] 16%|█▋ | 7896/48008 [26:49<2:09:27, 5.16it/s] 16%|█▋ | 7897/48008 [26:50<2:08:34, 5.20it/s] 16%|█▋ | 7898/48008 [26:50<2:07:41, 5.24it/s] 16%|█▋ | 7899/48008 [26:50<2:08:02, 5.22it/s] 16%|█▋ | 7900/48008 [26:50<2:10:01, 5.14it/s] {'loss': 0.3025, 'grad_norm': 0.1640625, 'learning_rate': 0.0001670929845025829, 'epoch': 0.16} 16%|█▋ | 7900/48008 [26:50<2:10:01, 5.14it/s] 16%|█▋ | 7901/48008 [26:50<2:18:58, 4.81it/s] 16%|█▋ | 7902/48008 [26:51<2:16:17, 4.90it/s] 16%|█▋ | 7903/48008 [26:51<2:13:53, 4.99it/s] 16%|█▋ | 7904/48008 [26:51<2:14:50, 4.96it/s] 16%|█▋ | 7905/48008 [26:51<2:12:38, 5.04it/s] 16%|█▋ | 7906/48008 [26:51<2:11:54, 5.07it/s] 16%|█▋ | 7907/48008 [26:51<2:09:56, 5.14it/s] 16%|█▋ | 7908/48008 [26:52<2:08:59, 5.18it/s] 16%|█▋ | 7909/48008 [26:52<2:08:56, 5.18it/s] 16%|█▋ | 7910/48008 [26:52<2:07:58, 5.22it/s] 16%|█▋ | 7911/48008 [26:52<2:08:45, 5.19it/s] 16%|█▋ | 7912/48008 [26:52<2:09:26, 5.16it/s] 16%|█▋ | 7913/48008 [26:53<2:10:51, 5.11it/s] 16%|█▋ | 7914/48008 [26:53<2:09:52, 5.15it/s] 16%|█▋ | 7915/48008 [26:53<2:08:53, 5.18it/s] 16%|█▋ | 7916/48008 [26:53<2:10:41, 5.11it/s] 16%|█▋ | 7917/48008 [26:53<2:11:46, 5.07it/s] 16%|█▋ | 7918/48008 [26:54<2:10:35, 5.12it/s] 16%|█▋ | 7919/48008 [26:54<2:11:21, 5.09it/s] 16%|█▋ | 7920/48008 [26:54<2:11:00, 5.10it/s] 16%|█▋ | 7921/48008 [26:54<2:09:36, 5.16it/s] 17%|█▋ | 7922/48008 [26:54<2:09:09, 5.17it/s] 17%|█▋ | 7923/48008 [26:55<2:10:35, 5.12it/s] 17%|█▋ | 7924/48008 [26:55<2:09:53, 5.14it/s] 17%|█▋ | 7925/48008 [26:55<2:18:38, 4.82it/s] 17%|█▋ | 7926/48008 [26:55<2:15:46, 4.92it/s] 17%|█▋ | 7927/48008 [26:55<2:13:10, 5.02it/s] 17%|█▋ | 7928/48008 [26:56<2:11:35, 5.08it/s] 17%|█▋ | 7929/48008 [26:56<2:12:24, 5.04it/s] 17%|█▋ | 7930/48008 [26:56<2:11:00, 5.10it/s] 17%|█▋ | 7931/48008 [26:56<2:10:05, 5.13it/s] 17%|█▋ | 7932/48008 [26:56<2:10:39, 5.11it/s] 17%|█▋ | 7933/48008 [26:57<2:08:52, 5.18it/s] 17%|█▋ | 7934/48008 [26:57<2:09:13, 5.17it/s] 17%|█▋ | 7935/48008 [26:57<2:07:42, 5.23it/s] 17%|█▋ | 7936/48008 [26:57<2:10:06, 5.13it/s] 17%|█▋ | 7937/48008 [26:57<2:09:42, 5.15it/s] 17%|█▋ | 7938/48008 [26:58<2:09:12, 5.17it/s] 17%|█▋ | 7939/48008 [26:58<2:31:44, 4.40it/s] 17%|█▋ | 7940/48008 [26:58<2:26:38, 4.55it/s] 17%|█▋ | 7941/48008 [26:58<2:21:31, 4.72it/s] 17%|█▋ | 7942/48008 [26:58<2:19:30, 4.79it/s] 17%|█▋ | 7943/48008 [26:59<2:15:48, 4.92it/s] 17%|█▋ | 7944/48008 [26:59<2:14:10, 4.98it/s] 17%|█▋ | 7945/48008 [26:59<2:11:25, 5.08it/s] 17%|█▋ | 7946/48008 [26:59<2:12:48, 5.03it/s] 17%|█▋ | 7947/48008 [26:59<2:12:28, 5.04it/s] 17%|█▋ | 7948/48008 [27:00<2:11:06, 5.09it/s] 17%|█▋ | 7949/48008 [27:00<2:09:13, 5.17it/s] 17%|█▋ | 7950/48008 [27:00<2:09:58, 5.14it/s] {'loss': 0.3026, 'grad_norm': 0.2236328125, 'learning_rate': 0.00016688468588568574, 'epoch': 0.17} 17%|█▋ | 7950/48008 [27:00<2:09:58, 5.14it/s] 17%|█▋ | 7951/48008 [27:00<2:09:07, 5.17it/s] 17%|█▋ | 7952/48008 [27:00<2:18:15, 4.83it/s] 17%|█▋ | 7953/48008 [27:01<2:14:56, 4.95it/s] 17%|█▋ | 7954/48008 [27:01<2:15:08, 4.94it/s] 17%|█▋ | 7955/48008 [27:01<2:12:51, 5.02it/s] 17%|█▋ | 7956/48008 [27:01<2:50:18, 3.92it/s] 17%|█▋ | 7957/48008 [27:02<2:39:34, 4.18it/s] 17%|█▋ | 7958/48008 [27:02<2:29:28, 4.47it/s] 17%|█▋ | 7959/48008 [27:02<2:23:26, 4.65it/s] 17%|█▋ | 7960/48008 [27:02<2:18:58, 4.80it/s] 17%|█▋ | 7961/48008 [27:02<2:14:39, 4.96it/s] 17%|█▋ | 7962/48008 [27:03<2:14:04, 4.98it/s] 17%|█▋ | 7963/48008 [27:03<2:12:16, 5.05it/s] 17%|█▋ | 7964/48008 [27:03<2:13:05, 5.01it/s] 17%|█▋ | 7965/48008 [27:03<2:10:44, 5.10it/s] 17%|█▋ | 7966/48008 [27:03<2:09:48, 5.14it/s] 17%|█▋ | 7967/48008 [27:04<2:08:23, 5.20it/s] 17%|█▋ | 7968/48008 [27:04<2:10:40, 5.11it/s] 17%|█▋ | 7969/48008 [27:04<2:11:25, 5.08it/s] 17%|█▋ | 7970/48008 [27:05<4:50:02, 2.30it/s] 17%|█▋ | 7971/48008 [27:05<4:01:28, 2.76it/s] 17%|█▋ | 7972/48008 [27:05<3:26:41, 3.23it/s] 17%|█▋ | 7973/48008 [27:05<3:02:37, 3.65it/s] 17%|█▋ | 7974/48008 [27:06<2:46:03, 4.02it/s] 17%|█▋ | 7975/48008 [27:06<2:33:36, 4.34it/s] 17%|█▋ | 7976/48008 [27:07<4:41:41, 2.37it/s] 17%|█▋ | 7977/48008 [27:07<3:56:52, 2.82it/s] 17%|█▋ | 7978/48008 [27:07<3:22:41, 3.29it/s] 17%|█▋ | 7979/48008 [27:07<3:00:06, 3.70it/s] 17%|█▋ | 7980/48008 [27:07<2:43:20, 4.08it/s] 17%|█▋ | 7981/48008 [27:08<2:30:39, 4.43it/s] 17%|█▋ | 7982/48008 [27:09<4:31:22, 2.46it/s] 17%|█▋ | 7983/48008 [27:09<3:47:43, 2.93it/s] 17%|█▋ | 7984/48008 [27:09<3:40:43, 3.02it/s] 17%|█▋ | 7985/48008 [27:09<3:13:42, 3.44it/s] 17%|█▋ | 7986/48008 [27:09<2:54:44, 3.82it/s] 17%|█▋ | 7987/48008 [27:10<2:39:55, 4.17it/s] 17%|█▋ | 7988/48008 [27:11<5:51:34, 1.90it/s] 17%|█▋ | 7989/48008 [27:11<4:44:29, 2.34it/s] 17%|█▋ | 7990/48008 [27:11<3:56:23, 2.82it/s] 17%|█▋ | 7991/48008 [27:11<3:24:33, 3.26it/s] 17%|█▋ | 7992/48008 [27:12<3:00:45, 3.69it/s] 17%|█▋ | 7993/48008 [27:12<2:43:39, 4.08it/s] 17%|█▋ | 7994/48008 [27:13<4:46:47, 2.33it/s] 17%|█▋ | 7995/48008 [27:13<4:01:00, 2.77it/s] 17%|█▋ | 7996/48008 [27:13<3:25:57, 3.24it/s] 17%|█▋ | 7997/48008 [27:13<3:01:57, 3.66it/s] 17%|█▋ | 7998/48008 [27:13<2:45:43, 4.02it/s] 17%|█▋ | 7999/48008 [27:14<2:35:13, 4.30it/s] 17%|█▋ | 8000/48008 [27:14<2:36:16, 4.27it/s] {'loss': 0.3038, 'grad_norm': 0.2060546875, 'learning_rate': 0.00016667638726878853, 'epoch': 0.17} 17%|█▋ | 8000/48008 [27:14<2:36:16, 4.27it/s] 17%|█▋ | 8001/48008 [27:14<2:29:07, 4.47it/s] 17%|█▋ | 8002/48008 [27:14<2:23:06, 4.66it/s] 17%|█▋ | 8003/48008 [27:14<2:17:48, 4.84it/s] 17%|█▋ | 8004/48008 [27:15<2:15:24, 4.92it/s] 17%|█▋ | 8005/48008 [27:15<2:12:25, 5.03it/s] 17%|█▋ | 8006/48008 [27:15<2:10:41, 5.10it/s] 17%|█▋ | 8007/48008 [27:15<2:10:58, 5.09it/s] 17%|█▋ | 8008/48008 [27:15<2:12:20, 5.04it/s] 17%|█▋ | 8009/48008 [27:16<2:11:15, 5.08it/s] 17%|█▋ | 8010/48008 [27:16<2:10:28, 5.11it/s] 17%|█▋ | 8011/48008 [27:16<2:09:11, 5.16it/s] 17%|█▋ | 8012/48008 [27:16<2:08:15, 5.20it/s] 17%|█▋ | 8013/48008 [27:17<4:22:15, 2.54it/s] 17%|█▋ | 8014/48008 [27:17<3:44:00, 2.98it/s] 17%|█▋ | 8015/48008 [27:17<3:14:28, 3.43it/s] 17%|█▋ | 8016/48008 [27:18<2:55:25, 3.80it/s] 17%|█▋ | 8017/48008 [27:18<3:19:57, 3.33it/s] 17%|█▋ | 8018/48008 [27:18<2:57:47, 3.75it/s] 17%|█▋ | 8019/48008 [27:18<2:42:16, 4.11it/s] 17%|█▋ | 8020/48008 [27:19<2:31:38, 4.39it/s] 17%|█▋ | 8021/48008 [27:19<2:24:08, 4.62it/s] 17%|█▋ | 8022/48008 [27:19<2:18:52, 4.80it/s] 17%|█▋ | 8023/48008 [27:19<2:17:56, 4.83it/s] 17%|█▋ | 8024/48008 [27:19<2:14:41, 4.95it/s] 17%|█▋ | 8025/48008 [27:19<2:15:03, 4.93it/s] 17%|█▋ | 8026/48008 [27:20<2:12:33, 5.03it/s] 17%|█▋ | 8027/48008 [27:20<2:11:17, 5.08it/s] 17%|█▋ | 8028/48008 [27:20<2:10:10, 5.12it/s] 17%|█▋ | 8029/48008 [27:20<2:09:16, 5.15it/s] 17%|█▋ | 8030/48008 [27:20<2:09:02, 5.16it/s] 17%|█▋ | 8031/48008 [27:21<2:08:36, 5.18it/s] 17%|█▋ | 8032/48008 [27:21<2:08:30, 5.18it/s] 17%|█▋ | 8033/48008 [27:21<2:09:52, 5.13it/s] 17%|█▋ | 8034/48008 [27:21<2:09:47, 5.13it/s] 17%|█▋ | 8035/48008 [27:21<2:09:31, 5.14it/s] 17%|█▋ | 8036/48008 [27:22<2:12:17, 5.04it/s] 17%|█▋ | 8037/48008 [27:22<2:20:18, 4.75it/s] 17%|█▋ | 8038/48008 [27:22<2:16:29, 4.88it/s] 17%|█▋ | 8039/48008 [27:22<2:52:57, 3.85it/s] 17%|█▋ | 8040/48008 [27:23<2:40:26, 4.15it/s] 17%|█▋ | 8041/48008 [27:23<2:29:34, 4.45it/s] 17%|█▋ | 8042/48008 [27:23<2:24:37, 4.61it/s] 17%|█▋ | 8043/48008 [27:23<2:18:39, 4.80it/s] 17%|█▋ | 8044/48008 [27:23<2:16:29, 4.88it/s] 17%|█▋ | 8045/48008 [27:24<2:52:22, 3.86it/s] 17%|█▋ | 8046/48008 [27:24<2:39:56, 4.16it/s] 17%|█▋ | 8047/48008 [27:24<2:31:42, 4.39it/s] 17%|█▋ | 8048/48008 [27:24<2:26:23, 4.55it/s] 17%|█▋ | 8049/48008 [27:25<2:21:16, 4.71it/s] 17%|█▋ | 8050/48008 [27:25<2:56:04, 3.78it/s] {'loss': 0.2992, 'grad_norm': 0.431640625, 'learning_rate': 0.00016646808865189138, 'epoch': 0.17} 17%|█▋ | 8050/48008 [27:25<2:56:04, 3.78it/s] 17%|█▋ | 8051/48008 [27:25<2:42:55, 4.09it/s] 17%|█▋ | 8052/48008 [27:25<2:32:21, 4.37it/s] 17%|█▋ | 8053/48008 [27:26<3:03:44, 3.62it/s] 17%|█▋ | 8054/48008 [27:26<2:47:56, 3.96it/s] 17%|█▋ | 8055/48008 [27:26<2:35:29, 4.28it/s] 17%|█▋ | 8056/48008 [27:26<2:27:03, 4.53it/s] 17%|█▋ | 8057/48008 [27:27<2:23:12, 4.65it/s] 17%|█▋ | 8058/48008 [27:27<2:17:40, 4.84it/s] 17%|█▋ | 8059/48008 [27:27<2:16:27, 4.88it/s] 17%|█▋ | 8060/48008 [27:27<2:14:33, 4.95it/s] 17%|█▋ | 8061/48008 [27:27<2:12:04, 5.04it/s] 17%|█▋ | 8062/48008 [27:27<2:09:58, 5.12it/s] 17%|█▋ | 8063/48008 [27:28<2:10:34, 5.10it/s] 17%|█▋ | 8064/48008 [27:28<2:09:43, 5.13it/s] 17%|█▋ | 8065/48008 [27:28<2:08:29, 5.18it/s] 17%|█▋ | 8066/48008 [27:28<2:12:18, 5.03it/s] 17%|█▋ | 8067/48008 [27:28<2:11:24, 5.07it/s] 17%|█▋ | 8068/48008 [27:29<2:09:21, 5.15it/s] 17%|█▋ | 8069/48008 [27:29<2:08:45, 5.17it/s] 17%|█▋ | 8070/48008 [27:29<2:09:17, 5.15it/s] 17%|█▋ | 8071/48008 [27:29<2:09:03, 5.16it/s] 17%|█▋ | 8072/48008 [27:29<2:08:17, 5.19it/s] 17%|█▋ | 8073/48008 [27:30<2:17:47, 4.83it/s] 17%|█▋ | 8074/48008 [27:30<2:16:24, 4.88it/s] 17%|█▋ | 8075/48008 [27:30<2:12:46, 5.01it/s] 17%|█▋ | 8076/48008 [27:30<2:11:19, 5.07it/s] 17%|█▋ | 8077/48008 [27:30<2:09:24, 5.14it/s] 17%|█▋ | 8078/48008 [27:31<2:09:11, 5.15it/s] 17%|█▋ | 8079/48008 [27:31<2:07:31, 5.22it/s] 17%|█▋ | 8080/48008 [27:31<2:08:39, 5.17it/s] 17%|█▋ | 8081/48008 [27:31<2:11:50, 5.05it/s] 17%|█▋ | 8082/48008 [27:31<2:10:08, 5.11it/s] 17%|█▋ | 8083/48008 [27:32<2:12:02, 5.04it/s] 17%|█▋ | 8084/48008 [27:32<2:10:48, 5.09it/s] 17%|█▋ | 8085/48008 [27:32<2:10:18, 5.11it/s] 17%|█▋ | 8086/48008 [27:32<2:09:06, 5.15it/s] 17%|█▋ | 8087/48008 [27:32<2:09:06, 5.15it/s] 17%|█▋ | 8088/48008 [27:33<2:08:16, 5.19it/s] 17%|█▋ | 8089/48008 [27:33<2:07:45, 5.21it/s] 17%|█▋ | 8090/48008 [27:33<2:07:34, 5.22it/s] 17%|█▋ | 8091/48008 [27:33<2:07:33, 5.22it/s] 17%|█▋ | 8092/48008 [27:33<2:08:03, 5.20it/s] 17%|█▋ | 8093/48008 [27:34<2:07:47, 5.21it/s] 17%|█▋ | 8094/48008 [27:34<2:08:24, 5.18it/s] 17%|█▋ | 8095/48008 [27:34<2:07:30, 5.22it/s] 17%|█▋ | 8096/48008 [27:34<2:09:42, 5.13it/s] 17%|█▋ | 8097/48008 [27:34<2:08:36, 5.17it/s] 17%|█▋ | 8098/48008 [27:35<2:17:24, 4.84it/s] 17%|█▋ | 8099/48008 [27:35<2:18:09, 4.81it/s] 17%|█▋ | 8100/48008 [27:35<2:54:01, 3.82it/s] {'loss': 0.3012, 'grad_norm': 0.1689453125, 'learning_rate': 0.00016625979003499417, 'epoch': 0.17} 17%|█▋ | 8100/48008 [27:35<2:54:01, 3.82it/s] 17%|█▋ | 8101/48008 [27:35<2:49:30, 3.92it/s] 17%|█▋ | 8102/48008 [27:36<2:36:15, 4.26it/s] 17%|█▋ | 8103/48008 [27:36<2:31:18, 4.40it/s] 17%|█▋ | 8104/48008 [27:36<2:23:38, 4.63it/s] 17%|█▋ | 8105/48008 [27:36<2:18:47, 4.79it/s] 17%|█▋ | 8106/48008 [27:36<2:17:30, 4.84it/s] 17%|█▋ | 8107/48008 [27:37<2:17:47, 4.83it/s] 17%|█▋ | 8108/48008 [27:37<2:16:35, 4.87it/s] 17%|█▋ | 8109/48008 [27:37<2:13:51, 4.97it/s] 17%|█▋ | 8110/48008 [27:37<2:13:16, 4.99it/s] 17%|█▋ | 8111/48008 [27:37<2:10:58, 5.08it/s] 17%|█▋ | 8112/48008 [27:38<2:09:57, 5.12it/s] 17%|█▋ | 8113/48008 [27:38<2:18:17, 4.81it/s] 17%|█▋ | 8114/48008 [27:38<2:16:29, 4.87it/s] 17%|█▋ | 8115/48008 [27:38<2:12:44, 5.01it/s] 17%|█▋ | 8116/48008 [27:38<2:10:52, 5.08it/s] 17%|█▋ | 8117/48008 [27:39<2:09:44, 5.12it/s] 17%|█▋ | 8118/48008 [27:39<2:08:35, 5.17it/s] 17%|█▋ | 8119/48008 [27:39<2:07:06, 5.23it/s] 17%|█▋ | 8120/48008 [27:39<2:06:47, 5.24it/s] 17%|█▋ | 8121/48008 [27:39<2:06:19, 5.26it/s] 17%|█▋ | 8122/48008 [27:40<2:06:20, 5.26it/s] 17%|█▋ | 8123/48008 [27:40<2:06:12, 5.27it/s] 17%|█▋ | 8124/48008 [27:40<2:06:11, 5.27it/s] 17%|█▋ | 8125/48008 [27:40<2:06:46, 5.24it/s] 17%|█▋ | 8126/48008 [27:40<2:06:22, 5.26it/s] 17%|█▋ | 8127/48008 [27:40<2:08:39, 5.17it/s] 17%|█▋ | 8128/48008 [27:41<2:10:00, 5.11it/s] 17%|█▋ | 8129/48008 [27:41<2:09:26, 5.13it/s] 17%|█▋ | 8130/48008 [27:41<2:08:40, 5.17it/s] 17%|█▋ | 8131/48008 [27:41<2:07:29, 5.21it/s] 17%|█▋ | 8132/48008 [27:41<2:07:26, 5.21it/s] 17%|█▋ | 8133/48008 [27:42<2:07:26, 5.22it/s] 17%|█▋ | 8134/48008 [27:42<2:09:45, 5.12it/s] 17%|█▋ | 8135/48008 [27:42<2:08:38, 5.17it/s] 17%|█▋ | 8136/48008 [27:42<2:08:12, 5.18it/s] 17%|█▋ | 8137/48008 [27:42<2:17:04, 4.85it/s] 17%|█▋ | 8138/48008 [27:43<2:15:22, 4.91it/s] 17%|█▋ | 8139/48008 [27:43<2:12:59, 5.00it/s] 17%|█▋ | 8140/48008 [27:43<2:11:16, 5.06it/s] 17%|█▋ | 8141/48008 [27:43<2:12:33, 5.01it/s] 17%|█▋ | 8142/48008 [27:43<2:10:57, 5.07it/s] 17%|█▋ | 8143/48008 [27:44<2:18:49, 4.79it/s] 17%|█▋ | 8144/48008 [27:44<2:15:06, 4.92it/s] 17%|█▋ | 8145/48008 [27:44<2:12:20, 5.02it/s] 17%|█▋ | 8146/48008 [27:44<2:12:36, 5.01it/s] 17%|█▋ | 8147/48008 [27:44<2:13:28, 4.98it/s] 17%|█▋ | 8148/48008 [27:45<2:11:13, 5.06it/s] 17%|█▋ | 8149/48008 [27:45<2:09:16, 5.14it/s] 17%|█▋ | 8150/48008 [27:45<2:08:58, 5.15it/s] {'loss': 0.305, 'grad_norm': 0.515625, 'learning_rate': 0.00016605149141809698, 'epoch': 0.17} 17%|█▋ | 8150/48008 [27:45<2:08:58, 5.15it/s] 17%|█▋ | 8151/48008 [27:45<2:09:37, 5.12it/s] 17%|█▋ | 8152/48008 [27:45<2:08:28, 5.17it/s] 17%|█▋ | 8153/48008 [27:46<2:07:37, 5.20it/s] 17%|█▋ | 8154/48008 [27:46<2:07:32, 5.21it/s] 17%|█▋ | 8155/48008 [27:46<2:07:26, 5.21it/s] 17%|█▋ | 8156/48008 [27:46<2:05:52, 5.28it/s] 17%|█▋ | 8157/48008 [27:46<2:05:54, 5.27it/s] 17%|█▋ | 8158/48008 [27:47<2:05:05, 5.31it/s] 17%|█▋ | 8159/48008 [27:47<2:04:35, 5.33it/s] 17%|█▋ | 8160/48008 [27:47<2:04:21, 5.34it/s] 17%|█▋ | 8161/48008 [27:47<2:05:08, 5.31it/s] 17%|█▋ | 8162/48008 [27:47<2:04:35, 5.33it/s] 17%|█▋ | 8163/48008 [27:47<2:06:54, 5.23it/s] 17%|█▋ | 8164/48008 [27:48<2:07:52, 5.19it/s] 17%|█▋ | 8165/48008 [27:48<2:11:23, 5.05it/s] 17%|█▋ | 8166/48008 [27:48<2:09:38, 5.12it/s] 17%|█▋ | 8167/48008 [27:48<2:08:22, 5.17it/s] 17%|█▋ | 8168/48008 [27:48<2:08:11, 5.18it/s] 17%|█▋ | 8169/48008 [27:49<2:07:09, 5.22it/s] 17%|█▋ | 8170/48008 [27:49<2:05:44, 5.28it/s] 17%|█▋ | 8171/48008 [27:49<2:05:16, 5.30it/s] 17%|█▋ | 8172/48008 [27:49<2:08:14, 5.18it/s] 17%|█▋ | 8173/48008 [27:49<2:10:11, 5.10it/s] 17%|█▋ | 8174/48008 [27:50<2:08:51, 5.15it/s] 17%|█▋ | 8175/48008 [27:50<2:07:28, 5.21it/s] 17%|█▋ | 8176/48008 [27:50<2:07:27, 5.21it/s] 17%|█▋ | 8177/48008 [27:50<2:08:50, 5.15it/s] 17%|█▋ | 8178/48008 [27:50<2:08:18, 5.17it/s] 17%|█▋ | 8179/48008 [27:51<2:46:53, 3.98it/s] 17%|█▋ | 8180/48008 [27:51<2:34:44, 4.29it/s] 17%|█▋ | 8181/48008 [27:51<2:26:41, 4.53it/s] 17%|█▋ | 8182/48008 [27:51<2:22:04, 4.67it/s] 17%|█▋ | 8183/48008 [27:52<2:19:29, 4.76it/s] 17%|█▋ | 8184/48008 [27:52<2:15:19, 4.90it/s] 17%|█▋ | 8185/48008 [27:52<2:14:27, 4.94it/s] 17%|█▋ | 8186/48008 [27:52<2:14:36, 4.93it/s] 17%|█▋ | 8187/48008 [27:52<2:11:59, 5.03it/s] 17%|█▋ | 8188/48008 [27:53<2:11:59, 5.03it/s] 17%|█▋ | 8189/48008 [27:53<2:13:16, 4.98it/s] 17%|█▋ | 8190/48008 [27:53<2:50:18, 3.90it/s] 17%|█▋ | 8191/48008 [27:53<2:37:01, 4.23it/s] 17%|█▋ | 8192/48008 [27:54<2:28:09, 4.48it/s] 17%|█▋ | 8193/48008 [27:54<2:25:07, 4.57it/s] 17%|█▋ | 8194/48008 [27:54<2:19:14, 4.77it/s] 17%|█▋ | 8195/48008 [27:54<2:14:39, 4.93it/s] 17%|█▋ | 8196/48008 [27:54<2:13:29, 4.97it/s] 17%|█▋ | 8197/48008 [27:54<2:13:34, 4.97it/s] 17%|█▋ | 8198/48008 [27:55<2:12:46, 5.00it/s] 17%|█▋ | 8199/48008 [27:55<2:11:57, 5.03it/s] 17%|█▋ | 8200/48008 [27:55<2:14:05, 4.95it/s] {'loss': 0.305, 'grad_norm': 0.25, 'learning_rate': 0.0001658431928011998, 'epoch': 0.17} 17%|█▋ | 8200/48008 [27:55<2:14:05, 4.95it/s] 17%|█▋ | 8201/48008 [27:55<2:16:16, 4.87it/s] 17%|█▋ | 8202/48008 [27:56<2:22:59, 4.64it/s] 17%|█▋ | 8203/48008 [27:56<2:17:54, 4.81it/s] 17%|█▋ | 8204/48008 [27:56<2:14:20, 4.94it/s] 17%|█▋ | 8205/48008 [27:56<2:14:30, 4.93it/s] 17%|█▋ | 8206/48008 [27:56<2:11:43, 5.04it/s] 17%|█▋ | 8207/48008 [27:57<2:10:41, 5.08it/s] 17%|█▋ | 8208/48008 [27:57<2:09:18, 5.13it/s] 17%|█▋ | 8209/48008 [27:57<2:08:14, 5.17it/s] 17%|█▋ | 8210/48008 [27:57<2:07:53, 5.19it/s] 17%|█▋ | 8211/48008 [27:57<2:07:08, 5.22it/s] 17%|█▋ | 8212/48008 [27:57<2:06:05, 5.26it/s] 17%|█▋ | 8213/48008 [27:58<2:29:05, 4.45it/s] 17%|█▋ | 8214/48008 [27:58<2:21:31, 4.69it/s] 17%|█▋ | 8215/48008 [27:58<2:16:01, 4.88it/s] 17%|█▋ | 8216/48008 [27:58<2:16:59, 4.84it/s] 17%|█▋ | 8217/48008 [27:59<2:17:38, 4.82it/s] 17%|█▋ | 8218/48008 [27:59<2:16:23, 4.86it/s] 17%|█▋ | 8219/48008 [27:59<2:15:52, 4.88it/s] 17%|█▋ | 8220/48008 [27:59<2:12:32, 5.00it/s] 17%|█▋ | 8221/48008 [27:59<2:20:05, 4.73it/s] 17%|█▋ | 8222/48008 [28:00<2:16:23, 4.86it/s] 17%|█▋ | 8223/48008 [28:00<2:13:34, 4.96it/s] 17%|█▋ | 8224/48008 [28:00<2:11:14, 5.05it/s] 17%|█▋ | 8225/48008 [28:00<2:09:40, 5.11it/s] 17%|█▋ | 8226/48008 [28:00<2:12:22, 5.01it/s] 17%|█▋ | 8227/48008 [28:01<2:14:48, 4.92it/s] 17%|█▋ | 8228/48008 [28:01<2:35:14, 4.27it/s] 17%|█▋ | 8229/48008 [28:01<2:26:27, 4.53it/s] 17%|█▋ | 8230/48008 [28:01<2:20:15, 4.73it/s] 17%|█▋ | 8231/48008 [28:01<2:16:55, 4.84it/s] 17%|█▋ | 8232/48008 [28:02<2:13:32, 4.96it/s] 17%|█▋ | 8233/48008 [28:02<2:10:40, 5.07it/s] 17%|█▋ | 8234/48008 [28:02<2:10:46, 5.07it/s] 17%|█▋ | 8235/48008 [28:02<2:08:56, 5.14it/s] 17%|█▋ | 8236/48008 [28:03<2:47:00, 3.97it/s] 17%|█▋ | 8237/48008 [28:03<2:35:24, 4.27it/s] 17%|█▋ | 8238/48008 [28:03<2:26:32, 4.52it/s] 17%|█▋ | 8239/48008 [28:03<2:19:24, 4.75it/s] 17%|█▋ | 8240/48008 [28:03<2:15:45, 4.88it/s] 17%|█▋ | 8241/48008 [28:04<2:15:03, 4.91it/s] 17%|█▋ | 8242/48008 [28:04<2:14:44, 4.92it/s] 17%|█▋ | 8243/48008 [28:04<2:11:14, 5.05it/s] 17%|█▋ | 8244/48008 [28:04<2:09:56, 5.10it/s] 17%|█▋ | 8245/48008 [28:04<2:09:52, 5.10it/s] 17%|█▋ | 8246/48008 [28:05<2:10:05, 5.09it/s] 17%|█▋ | 8247/48008 [28:05<2:09:05, 5.13it/s] 17%|█▋ | 8248/48008 [28:05<2:10:54, 5.06it/s] 17%|█▋ | 8249/48008 [28:05<2:10:58, 5.06it/s] 17%|█▋ | 8250/48008 [28:05<2:09:47, 5.11it/s] {'loss': 0.3014, 'grad_norm': 0.44140625, 'learning_rate': 0.00016563489418430262, 'epoch': 0.17} 17%|█▋ | 8250/48008 [28:05<2:09:47, 5.11it/s] 17%|█▋ | 8251/48008 [28:06<2:08:15, 5.17it/s] 17%|█▋ | 8252/48008 [28:06<2:07:46, 5.19it/s] 17%|█▋ | 8253/48008 [28:06<2:08:37, 5.15it/s] 17%|█▋ | 8254/48008 [28:06<2:08:28, 5.16it/s] 17%|█▋ | 8255/48008 [28:06<2:07:08, 5.21it/s] 17%|█▋ | 8256/48008 [28:06<2:06:11, 5.25it/s] 17%|█▋ | 8257/48008 [28:07<2:06:16, 5.25it/s] 17%|█▋ | 8258/48008 [28:07<2:06:00, 5.26it/s] 17%|█▋ | 8259/48008 [28:07<2:05:52, 5.26it/s] 17%|█▋ | 8260/48008 [28:07<2:05:00, 5.30it/s] 17%|█▋ | 8261/48008 [28:07<2:05:43, 5.27it/s] 17%|█▋ | 8262/48008 [28:08<2:08:02, 5.17it/s] 17%|█▋ | 8263/48008 [28:08<2:07:47, 5.18it/s] 17%|█▋ | 8264/48008 [28:08<2:07:20, 5.20it/s] 17%|█▋ | 8265/48008 [28:08<2:06:49, 5.22it/s] 17%|█▋ | 8266/48008 [28:08<2:06:26, 5.24it/s] 17%|█▋ | 8267/48008 [28:09<2:08:37, 5.15it/s] 17%|█▋ | 8268/48008 [28:09<2:10:05, 5.09it/s] 17%|█▋ | 8269/48008 [28:09<2:08:02, 5.17it/s] 17%|█▋ | 8270/48008 [28:09<2:06:13, 5.25it/s] 17%|█▋ | 8271/48008 [28:09<2:05:28, 5.28it/s] 17%|█▋ | 8272/48008 [28:10<2:04:59, 5.30it/s] 17%|█▋ | 8273/48008 [28:10<2:06:16, 5.24it/s] 17%|█▋ | 8274/48008 [28:10<2:05:53, 5.26it/s] 17%|█▋ | 8275/48008 [28:10<2:06:15, 5.24it/s] 17%|█▋ | 8276/48008 [28:10<2:05:30, 5.28it/s] 17%|█▋ | 8277/48008 [28:10<2:04:49, 5.30it/s] 17%|█▋ | 8278/48008 [28:11<2:05:29, 5.28it/s] 17%|█▋ | 8279/48008 [28:11<2:06:40, 5.23it/s] 17%|█▋ | 8280/48008 [28:11<2:07:27, 5.19it/s] 17%|█▋ | 8281/48008 [28:11<2:06:33, 5.23it/s] 17%|█▋ | 8282/48008 [28:11<2:05:33, 5.27it/s] 17%|█▋ | 8283/48008 [28:12<2:08:39, 5.15it/s] 17%|█▋ | 8284/48008 [28:12<2:07:23, 5.20it/s] 17%|█▋ | 8285/48008 [28:12<2:07:01, 5.21it/s] 17%|█▋ | 8286/48008 [28:12<2:06:38, 5.23it/s] 17%|█▋ | 8287/48008 [28:12<2:07:41, 5.18it/s] 17%|█▋ | 8288/48008 [28:13<2:05:31, 5.27it/s] 17%|█▋ | 8289/48008 [28:13<2:05:23, 5.28it/s] 17%|█▋ | 8290/48008 [28:13<2:05:50, 5.26it/s] 17%|█▋ | 8291/48008 [28:13<2:08:04, 5.17it/s] 17%|█▋ | 8292/48008 [28:13<2:16:36, 4.85it/s] 17%|█▋ | 8293/48008 [28:14<2:13:37, 4.95it/s] 17%|█▋ | 8294/48008 [28:14<2:11:30, 5.03it/s] 17%|█▋ | 8295/48008 [28:14<2:10:05, 5.09it/s] 17%|█▋ | 8296/48008 [28:14<2:09:02, 5.13it/s] 17%|█▋ | 8297/48008 [28:14<2:08:13, 5.16it/s] 17%|█▋ | 8298/48008 [28:15<2:08:40, 5.14it/s] 17%|█▋ | 8299/48008 [28:15<2:08:48, 5.14it/s] 17%|█▋ | 8300/48008 [28:15<2:07:08, 5.21it/s] {'loss': 0.3054, 'grad_norm': 0.1572265625, 'learning_rate': 0.00016542659556740546, 'epoch': 0.17} 17%|█▋ | 8300/48008 [28:15<2:07:08, 5.21it/s] 17%|█▋ | 8301/48008 [28:15<2:06:52, 5.22it/s] 17%|█▋ | 8302/48008 [28:15<2:06:24, 5.24it/s] 17%|█▋ | 8303/48008 [28:16<2:08:47, 5.14it/s] 17%|█▋ | 8304/48008 [28:16<2:07:31, 5.19it/s] 17%|█▋ | 8305/48008 [28:16<2:45:50, 3.99it/s] 17%|█▋ | 8306/48008 [28:16<2:33:51, 4.30it/s] 17%|█▋ | 8307/48008 [28:16<2:29:22, 4.43it/s] 17%|█▋ | 8308/48008 [28:17<2:22:00, 4.66it/s] 17%|█▋ | 8309/48008 [28:17<2:40:05, 4.13it/s] 17%|█▋ | 8310/48008 [28:17<2:30:30, 4.40it/s] 17%|█▋ | 8311/48008 [28:17<2:23:03, 4.62it/s] 17%|█▋ | 8312/48008 [28:18<2:18:13, 4.79it/s] 17%|█▋ | 8313/48008 [28:18<2:16:10, 4.86it/s] 17%|█▋ | 8314/48008 [28:18<2:12:59, 4.97it/s] 17%|█▋ | 8315/48008 [28:18<2:11:02, 5.05it/s] 17%|█▋ | 8316/48008 [28:18<2:10:31, 5.07it/s] 17%|█▋ | 8317/48008 [28:19<2:08:50, 5.13it/s] 17%|█▋ | 8318/48008 [28:19<2:07:49, 5.17it/s] 17%|█▋ | 8319/48008 [28:19<2:07:51, 5.17it/s] 17%|█▋ | 8320/48008 [28:19<2:07:51, 5.17it/s] 17%|█▋ | 8321/48008 [28:19<2:07:37, 5.18it/s] 17%|█▋ | 8322/48008 [28:19<2:10:46, 5.06it/s] 17%|█▋ | 8323/48008 [28:20<2:08:50, 5.13it/s] 17%|█▋ | 8324/48008 [28:20<2:08:12, 5.16it/s] 17%|█▋ | 8325/48008 [28:20<2:07:28, 5.19it/s] 17%|█▋ | 8326/48008 [28:20<2:29:50, 4.41it/s] 17%|█▋ | 8327/48008 [28:21<2:22:07, 4.65it/s] 17%|█▋ | 8328/48008 [28:21<2:17:00, 4.83it/s] 17%|█▋ | 8329/48008 [28:21<2:13:51, 4.94it/s] 17%|█▋ | 8330/48008 [28:21<2:11:46, 5.02it/s] 17%|█▋ | 8331/48008 [28:21<2:09:44, 5.10it/s] 17%|█▋ | 8332/48008 [28:22<2:08:31, 5.15it/s] 17%|█▋ | 8333/48008 [28:22<2:07:33, 5.18it/s] 17%|█▋ | 8334/48008 [28:22<2:10:01, 5.09it/s] 17%|█▋ | 8335/48008 [28:22<2:47:49, 3.94it/s] 17%|█▋ | 8336/48008 [28:22<2:35:04, 4.26it/s] 17%|█▋ | 8337/48008 [28:23<2:27:14, 4.49it/s] 17%|█▋ | 8338/48008 [28:23<2:20:59, 4.69it/s] 17%|█▋ | 8339/48008 [28:23<2:14:56, 4.90it/s] 17%|█▋ | 8340/48008 [28:23<2:13:33, 4.95it/s] 17%|█▋ | 8341/48008 [28:23<2:10:40, 5.06it/s] 17%|█▋ | 8342/48008 [28:24<2:08:27, 5.15it/s] 17%|█▋ | 8343/48008 [28:24<2:10:51, 5.05it/s] 17%|█▋ | 8344/48008 [28:24<2:13:11, 4.96it/s] 17%|█▋ | 8345/48008 [28:24<2:12:28, 4.99it/s] 17%|█▋ | 8346/48008 [28:24<2:14:11, 4.93it/s] 17%|█▋ | 8347/48008 [28:25<2:11:42, 5.02it/s] 17%|█▋ | 8348/48008 [28:25<2:10:10, 5.08it/s] 17%|█▋ | 8349/48008 [28:25<2:08:48, 5.13it/s] 17%|█▋ | 8350/48008 [28:25<2:07:10, 5.20it/s] {'loss': 0.3026, 'grad_norm': 0.29296875, 'learning_rate': 0.00016521829695050825, 'epoch': 0.17} 17%|█▋ | 8350/48008 [28:25<2:07:10, 5.20it/s] 17%|█▋ | 8351/48008 [28:25<2:06:18, 5.23it/s] 17%|█▋ | 8352/48008 [28:26<2:05:02, 5.29it/s] 17%|█▋ | 8353/48008 [28:26<2:04:01, 5.33it/s] 17%|█▋ | 8354/48008 [28:26<2:03:39, 5.34it/s] 17%|█▋ | 8355/48008 [28:26<2:04:31, 5.31it/s] 17%|█▋ | 8356/48008 [28:26<2:05:10, 5.28it/s] 17%|█▋ | 8357/48008 [28:27<2:07:58, 5.16it/s] 17%|█▋ | 8358/48008 [28:27<2:07:57, 5.16it/s] 17%|█▋ | 8359/48008 [28:27<2:06:33, 5.22it/s] 17%|█▋ | 8360/48008 [28:27<2:06:39, 5.22it/s] 17%|█▋ | 8361/48008 [28:27<2:08:36, 5.14it/s] 17%|█▋ | 8362/48008 [28:28<2:09:53, 5.09it/s] 17%|█▋ | 8363/48008 [28:28<2:08:07, 5.16it/s] 17%|█▋ | 8364/48008 [28:28<2:07:31, 5.18it/s] 17%|█▋ | 8365/48008 [28:28<2:16:11, 4.85it/s] 17%|█▋ | 8366/48008 [28:28<2:13:03, 4.97it/s] 17%|█▋ | 8367/48008 [28:29<2:14:53, 4.90it/s] 17%|█▋ | 8368/48008 [28:29<2:12:31, 4.99it/s] 17%|█▋ | 8369/48008 [28:29<2:11:35, 5.02it/s] 17%|█▋ | 8370/48008 [28:29<2:10:55, 5.05it/s] 17%|█▋ | 8371/48008 [28:29<2:09:29, 5.10it/s] 17%|█▋ | 8372/48008 [28:30<2:10:39, 5.06it/s] 17%|█▋ | 8373/48008 [28:30<2:08:59, 5.12it/s] 17%|█▋ | 8374/48008 [28:30<2:08:14, 5.15it/s] 17%|█▋ | 8375/48008 [28:30<2:09:34, 5.10it/s] 17%|█▋ | 8376/48008 [28:30<2:09:49, 5.09it/s] 17%|█▋ | 8377/48008 [28:30<2:11:23, 5.03it/s] 17%|█▋ | 8378/48008 [28:31<2:09:33, 5.10it/s] 17%|█▋ | 8379/48008 [28:31<2:07:47, 5.17it/s] 17%|█▋ | 8380/48008 [28:31<2:07:14, 5.19it/s] 17%|█▋ | 8381/48008 [28:31<2:06:28, 5.22it/s] 17%|█▋ | 8382/48008 [28:31<2:05:40, 5.26it/s] 17%|█▋ | 8383/48008 [28:32<2:05:37, 5.26it/s] 17%|█▋ | 8384/48008 [28:32<2:06:59, 5.20it/s] 17%|█▋ | 8385/48008 [28:32<2:15:50, 4.86it/s] 17%|█▋ | 8386/48008 [28:32<2:13:24, 4.95it/s] 17%|█▋ | 8387/48008 [28:32<2:11:51, 5.01it/s] 17%|█▋ | 8388/48008 [28:33<2:10:18, 5.07it/s] 17%|█▋ | 8389/48008 [28:33<2:32:03, 4.34it/s] 17%|█▋ | 8390/48008 [28:33<2:24:37, 4.57it/s] 17%|█▋ | 8391/48008 [28:33<2:19:31, 4.73it/s] 17%|█▋ | 8392/48008 [28:34<2:24:55, 4.56it/s] 17%|█▋ | 8393/48008 [28:34<2:19:11, 4.74it/s] 17%|█▋ | 8394/48008 [28:34<2:14:07, 4.92it/s] 17%|█▋ | 8395/48008 [28:34<2:10:36, 5.05it/s] 17%|█▋ | 8396/48008 [28:34<2:08:05, 5.15it/s] 17%|█▋ | 8397/48008 [28:35<2:07:19, 5.18it/s] 17%|█▋ | 8398/48008 [28:35<2:08:10, 5.15it/s] 17%|█▋ | 8399/48008 [28:35<2:07:41, 5.17it/s] 17%|█▋ | 8400/48008 [28:35<2:06:53, 5.20it/s] {'loss': 0.3037, 'grad_norm': 0.2109375, 'learning_rate': 0.00016500999833361107, 'epoch': 0.17} 17%|█▋ | 8400/48008 [28:35<2:06:53, 5.20it/s] 17%|█▋ | 8401/48008 [28:35<2:06:27, 5.22it/s] 18%|█▊ | 8402/48008 [28:35<2:08:26, 5.14it/s] 18%|█▊ | 8403/48008 [28:36<2:07:05, 5.19it/s] 18%|█▊ | 8404/48008 [28:36<2:16:04, 4.85it/s] 18%|█▊ | 8405/48008 [28:36<2:15:30, 4.87it/s] 18%|█▊ | 8406/48008 [28:36<2:13:04, 4.96it/s] 18%|█▊ | 8407/48008 [28:36<2:09:52, 5.08it/s] 18%|█▊ | 8408/48008 [28:37<2:08:35, 5.13it/s] 18%|█▊ | 8409/48008 [28:37<2:07:09, 5.19it/s] 18%|█▊ | 8410/48008 [28:37<2:07:10, 5.19it/s] 18%|█▊ | 8411/48008 [28:37<2:06:04, 5.23it/s] 18%|█▊ | 8412/48008 [28:37<2:05:18, 5.27it/s] 18%|█▊ | 8413/48008 [28:38<2:04:33, 5.30it/s] 18%|█▊ | 8414/48008 [28:38<2:04:26, 5.30it/s] 18%|█▊ | 8415/48008 [28:38<2:04:34, 5.30it/s] 18%|█▊ | 8416/48008 [28:38<2:04:08, 5.32it/s] 18%|█▊ | 8417/48008 [28:38<2:04:56, 5.28it/s] 18%|█▊ | 8418/48008 [28:39<2:05:31, 5.26it/s] 18%|█▊ | 8419/48008 [28:39<2:05:43, 5.25it/s] 18%|█▊ | 8420/48008 [28:39<2:15:06, 4.88it/s] 18%|█▊ | 8421/48008 [28:39<2:12:16, 4.99it/s] 18%|█▊ | 8422/48008 [28:39<2:19:43, 4.72it/s] 18%|█▊ | 8423/48008 [28:40<2:38:29, 4.16it/s] 18%|█▊ | 8424/48008 [28:40<2:27:45, 4.46it/s] 18%|█▊ | 8425/48008 [28:40<2:21:34, 4.66it/s] 18%|█▊ | 8426/48008 [28:40<2:17:01, 4.81it/s] 18%|█▊ | 8427/48008 [28:41<2:22:41, 4.62it/s] 18%|█▊ | 8428/48008 [28:41<2:16:38, 4.83it/s] 18%|█▊ | 8429/48008 [28:41<2:13:17, 4.95it/s] 18%|█▊ | 8430/48008 [28:41<2:12:37, 4.97it/s] 18%|█▊ | 8431/48008 [28:41<2:12:17, 4.99it/s] 18%|█▊ | 8432/48008 [28:41<2:10:20, 5.06it/s] 18%|█▊ | 8433/48008 [28:42<2:08:31, 5.13it/s] 18%|█▊ | 8434/48008 [28:42<2:07:35, 5.17it/s] 18%|█▊ | 8435/48008 [28:42<2:09:17, 5.10it/s] 18%|█▊ | 8436/48008 [28:42<2:08:10, 5.15it/s] 18%|█▊ | 8437/48008 [28:42<2:09:47, 5.08it/s] 18%|█▊ | 8438/48008 [28:43<2:08:11, 5.14it/s] 18%|█▊ | 8439/48008 [28:43<2:07:52, 5.16it/s] 18%|█▊ | 8440/48008 [28:43<2:07:40, 5.17it/s] 18%|█▊ | 8441/48008 [28:43<2:05:15, 5.26it/s] 18%|█▊ | 8442/48008 [28:43<2:05:15, 5.26it/s] 18%|█▊ | 8443/48008 [28:44<2:05:56, 5.24it/s] 18%|█▊ | 8444/48008 [28:44<2:05:58, 5.23it/s] 18%|█▊ | 8445/48008 [28:44<2:05:21, 5.26it/s] 18%|█▊ | 8446/48008 [28:44<2:04:52, 5.28it/s] 18%|█▊ | 8447/48008 [28:44<2:06:30, 5.21it/s] 18%|█▊ | 8448/48008 [28:45<2:06:43, 5.20it/s] 18%|█▊ | 8449/48008 [28:45<2:07:41, 5.16it/s] 18%|█▊ | 8450/48008 [28:45<2:07:08, 5.19it/s] {'loss': 0.3049, 'grad_norm': 0.359375, 'learning_rate': 0.0001648016997167139, 'epoch': 0.18} 18%|█▊ | 8450/48008 [28:45<2:07:08, 5.19it/s] 18%|█▊ | 8451/48008 [28:45<2:06:55, 5.19it/s] 18%|█▊ | 8452/48008 [28:45<2:07:31, 5.17it/s] 18%|█▊ | 8453/48008 [28:46<2:05:56, 5.23it/s] 18%|█▊ | 8454/48008 [28:46<2:05:48, 5.24it/s] 18%|█▊ | 8455/48008 [28:46<2:07:42, 5.16it/s] 18%|█▊ | 8456/48008 [28:46<2:08:07, 5.14it/s] 18%|█▊ | 8457/48008 [28:46<2:07:49, 5.16it/s] 18%|█▊ | 8458/48008 [28:47<2:08:03, 5.15it/s] 18%|█▊ | 8459/48008 [28:47<2:07:38, 5.16it/s] 18%|█▊ | 8460/48008 [28:47<2:06:15, 5.22it/s] 18%|█▊ | 8461/48008 [28:47<2:07:18, 5.18it/s] 18%|█▊ | 8462/48008 [28:47<2:07:51, 5.16it/s] 18%|█▊ | 8463/48008 [28:47<2:07:19, 5.18it/s] 18%|█▊ | 8464/48008 [28:48<2:16:13, 4.84it/s] 18%|█▊ | 8465/48008 [28:48<2:13:14, 4.95it/s] 18%|█▊ | 8466/48008 [28:48<2:12:08, 4.99it/s] 18%|█▊ | 8467/48008 [28:48<2:12:44, 4.96it/s] 18%|█▊ | 8468/48008 [28:48<2:13:41, 4.93it/s] 18%|█▊ | 8469/48008 [28:49<2:12:59, 4.95it/s] 18%|█▊ | 8470/48008 [28:49<2:11:02, 5.03it/s] 18%|█▊ | 8471/48008 [28:49<2:09:22, 5.09it/s] 18%|█▊ | 8472/48008 [28:49<2:07:56, 5.15it/s] 18%|█▊ | 8473/48008 [28:50<2:29:57, 4.39it/s] 18%|█▊ | 8474/48008 [28:50<2:21:32, 4.66it/s] 18%|█▊ | 8475/48008 [28:50<2:17:15, 4.80it/s] 18%|█▊ | 8476/48008 [28:50<2:14:13, 4.91it/s] 18%|█▊ | 8477/48008 [28:50<2:14:22, 4.90it/s] 18%|█▊ | 8478/48008 [28:51<2:34:34, 4.26it/s] 18%|█▊ | 8479/48008 [28:51<2:26:17, 4.50it/s] 18%|█▊ | 8480/48008 [28:51<2:20:36, 4.69it/s] 18%|█▊ | 8481/48008 [28:51<2:16:15, 4.83it/s] 18%|█▊ | 8482/48008 [28:51<2:13:21, 4.94it/s] 18%|█▊ | 8483/48008 [28:52<2:10:05, 5.06it/s] 18%|█▊ | 8484/48008 [28:52<2:09:49, 5.07it/s] 18%|█▊ | 8485/48008 [28:52<2:08:34, 5.12it/s] 18%|█▊ | 8486/48008 [28:52<2:07:36, 5.16it/s] 18%|█▊ | 8487/48008 [28:52<2:07:34, 5.16it/s] 18%|█▊ | 8488/48008 [28:53<2:07:22, 5.17it/s] 18%|█▊ | 8489/48008 [28:53<2:05:59, 5.23it/s] 18%|█▊ | 8490/48008 [28:53<2:06:37, 5.20it/s] 18%|█▊ | 8491/48008 [28:53<2:07:14, 5.18it/s] 18%|█▊ | 8492/48008 [28:53<2:06:31, 5.21it/s] 18%|█▊ | 8493/48008 [28:54<2:07:43, 5.16it/s] 18%|█▊ | 8494/48008 [28:54<2:16:05, 4.84it/s] 18%|█▊ | 8495/48008 [28:54<2:12:25, 4.97it/s] 18%|█▊ | 8496/48008 [28:54<2:09:57, 5.07it/s] 18%|█▊ | 8497/48008 [28:54<2:11:04, 5.02it/s] 18%|█▊ | 8498/48008 [28:55<2:11:31, 5.01it/s] 18%|█▊ | 8499/48008 [28:55<2:08:57, 5.11it/s] 18%|█▊ | 8500/48008 [28:55<2:08:05, 5.14it/s] {'loss': 0.3039, 'grad_norm': 0.4140625, 'learning_rate': 0.0001645934010998167, 'epoch': 0.18} 18%|█▊ | 8500/48008 [28:55<2:08:05, 5.14it/s] 18%|█▊ | 8501/48008 [28:55<2:06:39, 5.20it/s] 18%|█▊ | 8502/48008 [28:55<2:07:30, 5.16it/s] 18%|█▊ | 8503/48008 [28:56<2:07:49, 5.15it/s] 18%|█▊ | 8504/48008 [28:56<2:06:10, 5.22it/s] 18%|█▊ | 8505/48008 [28:56<2:06:08, 5.22it/s] 18%|█▊ | 8506/48008 [28:56<2:07:04, 5.18it/s] 18%|█▊ | 8507/48008 [28:56<2:07:55, 5.15it/s] 18%|█▊ | 8508/48008 [28:56<2:09:59, 5.06it/s] 18%|█▊ | 8509/48008 [28:57<2:09:17, 5.09it/s] 18%|█▊ | 8510/48008 [28:57<2:08:49, 5.11it/s] 18%|█▊ | 8511/48008 [28:57<2:10:02, 5.06it/s] 18%|█▊ | 8512/48008 [28:57<2:07:55, 5.15it/s] 18%|█▊ | 8513/48008 [28:57<2:08:00, 5.14it/s] 18%|█▊ | 8514/48008 [28:58<2:08:10, 5.14it/s] 18%|█▊ | 8515/48008 [28:58<2:07:18, 5.17it/s] 18%|█▊ | 8516/48008 [28:58<2:07:10, 5.18it/s] 18%|█▊ | 8517/48008 [28:58<2:06:45, 5.19it/s] 18%|█▊ | 8518/48008 [28:59<2:29:26, 4.40it/s] 18%|█▊ | 8519/48008 [28:59<2:21:43, 4.64it/s] 18%|█▊ | 8520/48008 [28:59<2:16:33, 4.82it/s] 18%|█▊ | 8521/48008 [28:59<2:12:46, 4.96it/s] 18%|█▊ | 8522/48008 [28:59<2:10:15, 5.05it/s] 18%|█▊ | 8523/48008 [28:59<2:09:19, 5.09it/s] 18%|█▊ | 8524/48008 [29:00<2:07:52, 5.15it/s] 18%|█▊ | 8525/48008 [29:00<2:09:49, 5.07it/s] 18%|█▊ | 8526/48008 [29:00<2:10:51, 5.03it/s] 18%|█▊ | 8527/48008 [29:00<2:08:47, 5.11it/s] 18%|█▊ | 8528/48008 [29:00<2:08:10, 5.13it/s] 18%|█▊ | 8529/48008 [29:01<2:06:37, 5.20it/s] 18%|█▊ | 8530/48008 [29:01<2:09:36, 5.08it/s] 18%|█▊ | 8531/48008 [29:01<2:08:14, 5.13it/s] 18%|█▊ | 8532/48008 [29:01<2:07:29, 5.16it/s] 18%|█▊ | 8533/48008 [29:01<2:07:11, 5.17it/s] 18%|█▊ | 8534/48008 [29:02<2:08:18, 5.13it/s] 18%|█▊ | 8535/48008 [29:02<2:07:20, 5.17it/s] 18%|█▊ | 8536/48008 [29:02<2:06:31, 5.20it/s] 18%|█▊ | 8537/48008 [29:02<2:06:00, 5.22it/s] 18%|█▊ | 8538/48008 [29:02<2:08:46, 5.11it/s] 18%|█▊ | 8539/48008 [29:03<2:07:55, 5.14it/s] 18%|█▊ | 8540/48008 [29:03<2:29:52, 4.39it/s] 18%|█▊ | 8541/48008 [29:03<2:22:11, 4.63it/s] 18%|█▊ | 8542/48008 [29:03<2:19:44, 4.71it/s] 18%|█▊ | 8543/48008 [29:03<2:15:13, 4.86it/s] 18%|█▊ | 8544/48008 [29:04<2:11:51, 4.99it/s] 18%|█▊ | 8545/48008 [29:04<2:12:18, 4.97it/s] 18%|█▊ | 8546/48008 [29:04<2:10:25, 5.04it/s] 18%|█▊ | 8547/48008 [29:04<2:08:24, 5.12it/s] 18%|█▊ | 8548/48008 [29:04<2:09:39, 5.07it/s] 18%|█▊ | 8549/48008 [29:05<2:09:56, 5.06it/s] 18%|█▊ | 8550/48008 [29:05<2:08:35, 5.11it/s] {'loss': 0.3037, 'grad_norm': 0.34375, 'learning_rate': 0.00016438510248291952, 'epoch': 0.18} 18%|█▊ | 8550/48008 [29:05<2:08:35, 5.11it/s] 18%|█▊ | 8551/48008 [29:05<2:07:04, 5.17it/s] 18%|█▊ | 8552/48008 [29:05<2:06:20, 5.21it/s] 18%|█▊ | 8553/48008 [29:05<2:05:47, 5.23it/s] 18%|█▊ | 8554/48008 [29:06<2:07:25, 5.16it/s] 18%|█▊ | 8555/48008 [29:06<2:07:17, 5.17it/s] 18%|█▊ | 8556/48008 [29:06<2:07:42, 5.15it/s] 18%|█▊ | 8557/48008 [29:06<2:06:39, 5.19it/s] 18%|█▊ | 8558/48008 [29:06<2:05:31, 5.24it/s] 18%|█▊ | 8559/48008 [29:07<2:06:45, 5.19it/s] 18%|█▊ | 8560/48008 [29:07<2:05:33, 5.24it/s] 18%|█▊ | 8561/48008 [29:07<2:08:46, 5.11it/s] 18%|█▊ | 8562/48008 [29:07<2:08:49, 5.10it/s] 18%|█▊ | 8563/48008 [29:07<2:09:16, 5.09it/s] 18%|█▊ | 8564/48008 [29:08<2:08:28, 5.12it/s] 18%|█▊ | 8565/48008 [29:08<2:08:04, 5.13it/s] 18%|█▊ | 8566/48008 [29:08<2:06:58, 5.18it/s] 18%|█▊ | 8567/48008 [29:08<2:15:26, 4.85it/s] 18%|█▊ | 8568/48008 [29:08<2:12:47, 4.95it/s] 18%|█▊ | 8569/48008 [29:09<2:11:10, 5.01it/s] 18%|█▊ | 8570/48008 [29:09<2:11:57, 4.98it/s] 18%|█▊ | 8571/48008 [29:09<2:09:10, 5.09it/s] 18%|█▊ | 8572/48008 [29:09<2:08:44, 5.11it/s] 18%|█▊ | 8573/48008 [29:09<2:07:24, 5.16it/s] 18%|█▊ | 8574/48008 [29:10<2:08:53, 5.10it/s] 18%|█▊ | 8575/48008 [29:10<2:10:27, 5.04it/s] 18%|█▊ | 8576/48008 [29:10<2:08:38, 5.11it/s] 18%|█▊ | 8577/48008 [29:10<2:09:06, 5.09it/s] 18%|█▊ | 8578/48008 [29:10<2:07:37, 5.15it/s] 18%|█▊ | 8579/48008 [29:11<2:06:48, 5.18it/s] 18%|█▊ | 8580/48008 [29:11<2:05:23, 5.24it/s] 18%|█▊ | 8581/48008 [29:11<2:06:20, 5.20it/s] 18%|█▊ | 8582/48008 [29:11<2:04:44, 5.27it/s] 18%|█▊ | 8583/48008 [29:12<4:45:38, 2.30it/s] 18%|█▊ | 8584/48008 [29:12<3:59:50, 2.74it/s] 18%|█▊ | 8585/48008 [29:12<3:25:45, 3.19it/s] 18%|█▊ | 8586/48008 [29:13<3:04:51, 3.55it/s] 18%|█▊ | 8587/48008 [29:13<2:46:47, 3.94it/s] 18%|█▊ | 8588/48008 [29:13<2:34:11, 4.26it/s] 18%|█▊ | 8589/48008 [29:14<4:46:28, 2.29it/s] 18%|█▊ | 8590/48008 [29:14<4:07:22, 2.66it/s] 18%|█▊ | 8591/48008 [29:14<3:30:31, 3.12it/s] 18%|█▊ | 8592/48008 [29:15<3:06:06, 3.53it/s] 18%|█▊ | 8593/48008 [29:15<2:49:01, 3.89it/s] 18%|█▊ | 8594/48008 [29:15<2:35:52, 4.21it/s] 18%|█▊ | 8595/48008 [29:16<4:28:23, 2.45it/s] 18%|█▊ | 8596/48008 [29:16<3:54:46, 2.80it/s] 18%|█▊ | 8597/48008 [29:16<3:21:54, 3.25it/s] 18%|█▊ | 8598/48008 [29:16<3:00:05, 3.65it/s] 18%|█▊ | 8599/48008 [29:17<2:45:19, 3.97it/s] 18%|█▊ | 8600/48008 [29:17<2:35:21, 4.23it/s] {'loss': 0.3031, 'grad_norm': 0.353515625, 'learning_rate': 0.00016417680386602234, 'epoch': 0.18} 18%|█▊ | 8600/48008 [29:17<2:35:21, 4.23it/s] 18%|█▊ | 8601/48008 [29:18<5:42:17, 1.92it/s] 18%|█▊ | 8602/48008 [29:18<4:46:43, 2.29it/s] 18%|█▊ | 8603/48008 [29:18<3:58:44, 2.75it/s] 18%|█▊ | 8604/48008 [29:19<3:25:02, 3.20it/s] 18%|█▊ | 8605/48008 [29:19<3:04:22, 3.56it/s] 18%|█▊ | 8606/48008 [29:19<2:46:45, 3.94it/s] 18%|█▊ | 8607/48008 [29:20<4:43:06, 2.32it/s] 18%|█▊ | 8608/48008 [29:20<3:55:27, 2.79it/s] 18%|█▊ | 8609/48008 [29:20<3:26:07, 3.19it/s] 18%|█▊ | 8610/48008 [29:20<3:04:01, 3.57it/s] 18%|█▊ | 8611/48008 [29:21<2:55:38, 3.74it/s] 18%|█▊ | 8612/48008 [29:21<2:40:20, 4.09it/s] 18%|█▊ | 8613/48008 [29:21<2:30:52, 4.35it/s] 18%|█▊ | 8614/48008 [29:21<2:23:58, 4.56it/s] 18%|█▊ | 8615/48008 [29:21<2:20:48, 4.66it/s] 18%|█▊ | 8616/48008 [29:22<2:18:16, 4.75it/s] 18%|█▊ | 8617/48008 [29:22<2:13:15, 4.93it/s] 18%|█▊ | 8618/48008 [29:22<2:11:01, 5.01it/s] 18%|█▊ | 8619/48008 [29:22<2:08:34, 5.11it/s] 18%|█▊ | 8620/48008 [29:22<2:08:50, 5.10it/s] 18%|█▊ | 8621/48008 [29:23<2:07:31, 5.15it/s] 18%|█▊ | 8622/48008 [29:23<2:08:56, 5.09it/s] 18%|█▊ | 8623/48008 [29:23<2:06:56, 5.17it/s] 18%|█▊ | 8624/48008 [29:23<2:07:16, 5.16it/s] 18%|█▊ | 8625/48008 [29:23<2:06:16, 5.20it/s] 18%|█▊ | 8626/48008 [29:24<4:23:49, 2.49it/s] 18%|█▊ | 8627/48008 [29:25<3:44:53, 2.92it/s] 18%|█▊ | 8628/48008 [29:25<3:14:25, 3.38it/s] 18%|█▊ | 8629/48008 [29:25<2:54:15, 3.77it/s] 18%|█▊ | 8630/48008 [29:25<2:40:51, 4.08it/s] 18%|█▊ | 8631/48008 [29:25<2:33:54, 4.26it/s] 18%|█▊ | 8632/48008 [29:25<2:26:19, 4.48it/s] 18%|█▊ | 8633/48008 [29:26<2:19:15, 4.71it/s] 18%|█▊ | 8634/48008 [29:26<2:16:32, 4.81it/s] 18%|█▊ | 8635/48008 [29:26<2:13:19, 4.92it/s] 18%|█▊ | 8636/48008 [29:26<2:10:28, 5.03it/s] 18%|█▊ | 8637/48008 [29:26<2:09:02, 5.08it/s] 18%|█▊ | 8638/48008 [29:27<2:06:54, 5.17it/s] 18%|█▊ | 8639/48008 [29:27<2:09:01, 5.09it/s] 18%|█▊ | 8640/48008 [29:27<2:07:15, 5.16it/s] 18%|█▊ | 8641/48008 [29:27<2:06:21, 5.19it/s] 18%|█▊ | 8642/48008 [29:27<2:05:53, 5.21it/s] 18%|█▊ | 8643/48008 [29:28<2:05:42, 5.22it/s] 18%|█▊ | 8644/48008 [29:28<2:07:55, 5.13it/s] 18%|█▊ | 8645/48008 [29:28<2:09:39, 5.06it/s] 18%|█▊ | 8646/48008 [29:28<2:08:27, 5.11it/s] 18%|█▊ | 8647/48008 [29:28<2:07:07, 5.16it/s] 18%|█▊ | 8648/48008 [29:29<2:07:59, 5.13it/s] 18%|█▊ | 8649/48008 [29:29<2:07:10, 5.16it/s] 18%|█▊ | 8650/48008 [29:29<2:44:41, 3.98it/s] {'loss': 0.3001, 'grad_norm': 0.1513671875, 'learning_rate': 0.00016396850524912516, 'epoch': 0.18} 18%|█▊ | 8650/48008 [29:29<2:44:41, 3.98it/s] 18%|█▊ | 8651/48008 [29:29<2:55:39, 3.73it/s] 18%|█▊ | 8652/48008 [29:30<2:39:39, 4.11it/s] 18%|█▊ | 8653/48008 [29:30<2:29:35, 4.38it/s] 18%|█▊ | 8654/48008 [29:30<2:21:13, 4.64it/s] 18%|█▊ | 8655/48008 [29:30<2:16:07, 4.82it/s] 18%|█▊ | 8656/48008 [29:30<2:12:56, 4.93it/s] 18%|█▊ | 8657/48008 [29:31<2:10:15, 5.03it/s] 18%|█▊ | 8658/48008 [29:31<2:08:42, 5.10it/s] 18%|█▊ | 8659/48008 [29:31<2:07:34, 5.14it/s] 18%|█▊ | 8660/48008 [29:31<2:07:45, 5.13it/s] 18%|█▊ | 8661/48008 [29:31<2:07:06, 5.16it/s] 18%|█▊ | 8662/48008 [29:32<2:06:24, 5.19it/s] 18%|█▊ | 8663/48008 [29:32<2:06:00, 5.20it/s] 18%|█▊ | 8664/48008 [29:32<2:05:16, 5.23it/s] 18%|█▊ | 8665/48008 [29:32<2:07:36, 5.14it/s] 18%|█▊ | 8666/48008 [29:32<2:07:50, 5.13it/s] 18%|█▊ | 8667/48008 [29:33<2:06:23, 5.19it/s] 18%|█▊ | 8668/48008 [29:33<2:05:22, 5.23it/s] 18%|█▊ | 8669/48008 [29:33<2:05:06, 5.24it/s] 18%|█▊ | 8670/48008 [29:33<2:05:05, 5.24it/s] 18%|█▊ | 8671/48008 [29:33<2:14:28, 4.88it/s] 18%|█▊ | 8672/48008 [29:34<2:34:23, 4.25it/s] 18%|█▊ | 8673/48008 [29:34<2:25:43, 4.50it/s] 18%|█▊ | 8674/48008 [29:34<2:18:13, 4.74it/s] 18%|█▊ | 8675/48008 [29:34<2:14:04, 4.89it/s] 18%|█▊ | 8676/48008 [29:34<2:11:47, 4.97it/s] 18%|█▊ | 8677/48008 [29:35<2:10:27, 5.02it/s] 18%|█▊ | 8678/48008 [29:35<2:08:57, 5.08it/s] 18%|█▊ | 8679/48008 [29:35<2:07:12, 5.15it/s] 18%|█▊ | 8680/48008 [29:35<2:06:49, 5.17it/s] 18%|█▊ | 8681/48008 [29:35<2:06:33, 5.18it/s] 18%|█▊ | 8682/48008 [29:36<2:08:17, 5.11it/s] 18%|█▊ | 8683/48008 [29:36<2:07:18, 5.15it/s] 18%|█▊ | 8684/48008 [29:36<2:06:03, 5.20it/s] 18%|█▊ | 8685/48008 [29:36<2:06:22, 5.19it/s] 18%|█▊ | 8686/48008 [29:36<2:06:07, 5.20it/s] 18%|█▊ | 8687/48008 [29:37<2:44:20, 3.99it/s] 18%|█▊ | 8688/48008 [29:37<2:34:17, 4.25it/s] 18%|█▊ | 8689/48008 [29:37<2:25:42, 4.50it/s] 18%|█▊ | 8690/48008 [29:37<2:19:44, 4.69it/s] 18%|█▊ | 8691/48008 [29:37<2:17:30, 4.77it/s] 18%|█▊ | 8692/48008 [29:38<2:13:38, 4.90it/s] 18%|█▊ | 8693/48008 [29:38<2:10:49, 5.01it/s] 18%|█▊ | 8694/48008 [29:38<2:08:11, 5.11it/s] 18%|█▊ | 8695/48008 [29:38<2:30:11, 4.36it/s] 18%|█▊ | 8696/48008 [29:39<3:01:16, 3.61it/s] 18%|█▊ | 8697/48008 [29:39<2:45:17, 3.96it/s] 18%|█▊ | 8698/48008 [29:39<2:34:57, 4.23it/s] 18%|█▊ | 8699/48008 [29:39<2:26:12, 4.48it/s] 18%|█▊ | 8700/48008 [29:40<2:20:18, 4.67it/s] {'loss': 0.3024, 'grad_norm': 0.52734375, 'learning_rate': 0.00016376020663222797, 'epoch': 0.18} 18%|█▊ | 8700/48008 [29:40<2:20:18, 4.67it/s] 18%|█▊ | 8701/48008 [29:40<2:16:16, 4.81it/s] 18%|█▊ | 8702/48008 [29:40<2:14:10, 4.88it/s] 18%|█▊ | 8703/48008 [29:40<2:11:30, 4.98it/s] 18%|█▊ | 8704/48008 [29:40<2:12:05, 4.96it/s] 18%|█▊ | 8705/48008 [29:41<2:10:36, 5.02it/s] 18%|█▊ | 8706/48008 [29:41<2:08:19, 5.10it/s] 18%|█▊ | 8707/48008 [29:41<2:07:02, 5.16it/s] 18%|█▊ | 8708/48008 [29:41<2:06:37, 5.17it/s] 18%|█▊ | 8709/48008 [29:41<2:07:01, 5.16it/s] 18%|█▊ | 8710/48008 [29:41<2:08:38, 5.09it/s] 18%|█▊ | 8711/48008 [29:42<2:07:34, 5.13it/s] 18%|█▊ | 8712/48008 [29:42<2:06:20, 5.18it/s] 18%|█▊ | 8713/48008 [29:42<2:06:26, 5.18it/s] 18%|█▊ | 8714/48008 [29:42<2:05:56, 5.20it/s] 18%|█▊ | 8715/48008 [29:42<2:05:35, 5.21it/s] 18%|█▊ | 8716/48008 [29:43<2:07:43, 5.13it/s] 18%|█▊ | 8717/48008 [29:43<2:06:22, 5.18it/s] 18%|█▊ | 8718/48008 [29:43<2:05:57, 5.20it/s] 18%|█▊ | 8719/48008 [29:43<2:06:23, 5.18it/s] 18%|█▊ | 8720/48008 [29:43<2:08:39, 5.09it/s] 18%|█▊ | 8721/48008 [29:44<2:08:15, 5.10it/s] 18%|█▊ | 8722/48008 [29:44<2:10:54, 5.00it/s] 18%|█▊ | 8723/48008 [29:44<2:12:02, 4.96it/s] 18%|█▊ | 8724/48008 [29:44<2:10:16, 5.03it/s] 18%|█▊ | 8725/48008 [29:44<2:08:42, 5.09it/s] 18%|█▊ | 8726/48008 [29:45<2:07:51, 5.12it/s] 18%|█▊ | 8727/48008 [29:45<2:05:43, 5.21it/s] 18%|█▊ | 8728/48008 [29:45<2:14:39, 4.86it/s] 18%|█▊ | 8729/48008 [29:45<2:12:01, 4.96it/s] 18%|█▊ | 8730/48008 [29:45<2:10:34, 5.01it/s] 18%|█▊ | 8731/48008 [29:46<2:09:21, 5.06it/s] 18%|█▊ | 8732/48008 [29:46<2:09:33, 5.05it/s] 18%|█▊ | 8733/48008 [29:46<2:08:32, 5.09it/s] 18%|█▊ | 8734/48008 [29:46<2:07:47, 5.12it/s] 18%|█▊ | 8735/48008 [29:46<2:06:45, 5.16it/s] 18%|█▊ | 8736/48008 [29:47<2:06:37, 5.17it/s] 18%|█▊ | 8737/48008 [29:47<2:05:41, 5.21it/s] 18%|█▊ | 8738/48008 [29:47<2:05:35, 5.21it/s] 18%|█▊ | 8739/48008 [29:47<2:05:32, 5.21it/s] 18%|█▊ | 8740/48008 [29:47<2:05:41, 5.21it/s] 18%|█▊ | 8741/48008 [29:48<2:14:45, 4.86it/s] 18%|█▊ | 8742/48008 [29:48<2:09:58, 5.04it/s] 18%|█▊ | 8743/48008 [29:48<2:07:53, 5.12it/s] 18%|█▊ | 8744/48008 [29:48<2:16:31, 4.79it/s] 18%|█▊ | 8745/48008 [29:48<2:13:28, 4.90it/s] 18%|█▊ | 8746/48008 [29:49<2:13:00, 4.92it/s] 18%|█▊ | 8747/48008 [29:49<2:14:28, 4.87it/s] 18%|█▊ | 8748/48008 [29:49<2:11:33, 4.97it/s] 18%|█▊ | 8749/48008 [29:49<2:10:42, 5.01it/s] 18%|█▊ | 8750/48008 [29:49<2:09:20, 5.06it/s] {'loss': 0.3028, 'grad_norm': 0.470703125, 'learning_rate': 0.0001635519080153308, 'epoch': 0.18} 18%|█▊ | 8750/48008 [29:49<2:09:20, 5.06it/s] 18%|█▊ | 8751/48008 [29:50<2:10:28, 5.01it/s] 18%|█▊ | 8752/48008 [29:50<2:08:47, 5.08it/s] 18%|█▊ | 8753/48008 [29:50<2:08:41, 5.08it/s] 18%|█▊ | 8754/48008 [29:50<2:11:27, 4.98it/s] 18%|█▊ | 8755/48008 [29:50<2:08:40, 5.08it/s] 18%|█▊ | 8756/48008 [29:51<2:07:53, 5.12it/s] 18%|█▊ | 8757/48008 [29:51<2:09:33, 5.05it/s] 18%|█▊ | 8758/48008 [29:51<2:10:28, 5.01it/s] 18%|█▊ | 8759/48008 [29:51<2:07:52, 5.12it/s] 18%|█▊ | 8760/48008 [29:51<2:07:32, 5.13it/s] 18%|█▊ | 8761/48008 [29:52<2:05:51, 5.20it/s] 18%|█▊ | 8762/48008 [29:52<2:05:54, 5.20it/s] 18%|█▊ | 8763/48008 [29:52<2:05:14, 5.22it/s] 18%|█▊ | 8764/48008 [29:52<2:05:52, 5.20it/s] 18%|█▊ | 8765/48008 [29:52<2:08:07, 5.11it/s] 18%|█▊ | 8766/48008 [29:52<2:07:40, 5.12it/s] 18%|█▊ | 8767/48008 [29:53<2:09:18, 5.06it/s] 18%|█▊ | 8768/48008 [29:53<2:07:28, 5.13it/s] 18%|█▊ | 8769/48008 [29:53<2:06:52, 5.15it/s] 18%|█▊ | 8770/48008 [29:53<2:06:37, 5.16it/s] 18%|█▊ | 8771/48008 [29:53<2:05:06, 5.23it/s] 18%|█▊ | 8772/48008 [29:54<2:06:34, 5.17it/s] 18%|█▊ | 8773/48008 [29:54<2:44:15, 3.98it/s] 18%|█▊ | 8774/48008 [29:54<2:33:29, 4.26it/s] 18%|█▊ | 8775/48008 [29:54<2:24:57, 4.51it/s] 18%|█▊ | 8776/48008 [29:55<2:19:35, 4.68it/s] 18%|█▊ | 8777/48008 [29:55<2:53:18, 3.77it/s] 18%|█▊ | 8778/48008 [29:55<2:39:11, 4.11it/s] 18%|█▊ | 8779/48008 [29:55<2:28:17, 4.41it/s] 18%|█▊ | 8780/48008 [29:56<2:30:22, 4.35it/s] 18%|█▊ | 8781/48008 [29:56<2:25:08, 4.50it/s] 18%|█▊ | 8782/48008 [29:56<2:19:18, 4.69it/s] 18%|█▊ | 8783/48008 [29:56<2:17:24, 4.76it/s] 18%|█▊ | 8784/48008 [29:56<2:15:53, 4.81it/s] 18%|█▊ | 8785/48008 [29:57<2:12:59, 4.92it/s] 18%|█▊ | 8786/48008 [29:57<2:09:56, 5.03it/s] 18%|█▊ | 8787/48008 [29:57<2:08:20, 5.09it/s] 18%|█▊ | 8788/48008 [29:57<2:07:49, 5.11it/s] 18%|█▊ | 8789/48008 [29:57<2:08:13, 5.10it/s] 18%|█▊ | 8790/48008 [29:58<2:16:22, 4.79it/s] 18%|█▊ | 8791/48008 [29:58<2:15:32, 4.82it/s] 18%|█▊ | 8792/48008 [29:58<2:11:49, 4.96it/s] 18%|█▊ | 8793/48008 [29:58<2:47:53, 3.89it/s] 18%|█▊ | 8794/48008 [29:59<2:34:35, 4.23it/s] 18%|█▊ | 8795/48008 [29:59<2:25:32, 4.49it/s] 18%|█▊ | 8796/48008 [29:59<2:21:23, 4.62it/s] 18%|█▊ | 8797/48008 [29:59<2:16:17, 4.79it/s] 18%|█▊ | 8798/48008 [29:59<2:13:18, 4.90it/s] 18%|█▊ | 8799/48008 [30:00<2:10:44, 5.00it/s] 18%|█▊ | 8800/48008 [30:00<2:11:04, 4.99it/s] {'loss': 0.3032, 'grad_norm': 0.498046875, 'learning_rate': 0.0001633436093984336, 'epoch': 0.18} 18%|█▊ | 8800/48008 [30:00<2:11:04, 4.99it/s] 18%|█▊ | 8801/48008 [30:00<2:10:35, 5.00it/s] 18%|█▊ | 8802/48008 [30:00<2:10:35, 5.00it/s] 18%|█▊ | 8803/48008 [30:00<2:09:15, 5.05it/s] 18%|█▊ | 8804/48008 [30:01<2:08:04, 5.10it/s] 18%|█▊ | 8805/48008 [30:01<2:08:43, 5.08it/s] 18%|█▊ | 8806/48008 [30:01<2:07:06, 5.14it/s] 18%|█▊ | 8807/48008 [30:01<2:06:00, 5.19it/s] 18%|█▊ | 8808/48008 [30:01<2:07:24, 5.13it/s] 18%|█▊ | 8809/48008 [30:02<2:07:18, 5.13it/s] 18%|█▊ | 8810/48008 [30:02<2:16:08, 4.80it/s] 18%|█▊ | 8811/48008 [30:02<2:51:14, 3.81it/s] 18%|█▊ | 8812/48008 [30:02<2:37:53, 4.14it/s] 18%|█▊ | 8813/48008 [30:03<2:28:20, 4.40it/s] 18%|█▊ | 8814/48008 [30:03<2:20:41, 4.64it/s] 18%|█▊ | 8815/48008 [30:03<2:15:37, 4.82it/s] 18%|█▊ | 8816/48008 [30:03<2:12:22, 4.93it/s] 18%|█▊ | 8817/48008 [30:03<2:12:29, 4.93it/s] 18%|█▊ | 8818/48008 [30:04<2:12:41, 4.92it/s] 18%|█▊ | 8819/48008 [30:04<2:09:50, 5.03it/s] 18%|█▊ | 8820/48008 [30:04<2:08:50, 5.07it/s] 18%|█▊ | 8821/48008 [30:04<2:10:17, 5.01it/s] 18%|█▊ | 8822/48008 [30:04<2:08:51, 5.07it/s] 18%|█▊ | 8823/48008 [30:04<2:09:45, 5.03it/s] 18%|█▊ | 8824/48008 [30:05<2:08:33, 5.08it/s] 18%|█▊ | 8825/48008 [30:05<2:07:28, 5.12it/s] 18%|█▊ | 8826/48008 [30:05<2:06:43, 5.15it/s] 18%|█▊ | 8827/48008 [30:05<2:09:40, 5.04it/s] 18%|█▊ | 8828/48008 [30:05<2:08:21, 5.09it/s] 18%|█▊ | 8829/48008 [30:06<2:16:22, 4.79it/s] 18%|█▊ | 8830/48008 [30:06<2:13:09, 4.90it/s] 18%|█▊ | 8831/48008 [30:06<2:10:37, 5.00it/s] 18%|█▊ | 8832/48008 [30:06<2:10:02, 5.02it/s] 18%|█▊ | 8833/48008 [30:06<2:07:55, 5.10it/s] 18%|█▊ | 8834/48008 [30:07<2:29:39, 4.36it/s] 18%|█▊ | 8835/48008 [30:07<2:21:48, 4.60it/s] 18%|█▊ | 8836/48008 [30:07<2:17:17, 4.76it/s] 18%|█▊ | 8837/48008 [30:07<2:13:02, 4.91it/s] 18%|█▊ | 8838/48008 [30:08<2:10:18, 5.01it/s] 18%|█▊ | 8839/48008 [30:08<2:08:42, 5.07it/s] 18%|█▊ | 8840/48008 [30:08<2:29:57, 4.35it/s] 18%|█▊ | 8841/48008 [30:08<2:21:27, 4.61it/s] 18%|█▊ | 8842/48008 [30:08<2:15:31, 4.82it/s] 18%|█▊ | 8843/48008 [30:09<2:11:56, 4.95it/s] 18%|█▊ | 8844/48008 [30:09<2:10:24, 5.01it/s] 18%|█▊ | 8845/48008 [30:09<2:10:13, 5.01it/s] 18%|█▊ | 8846/48008 [30:09<2:08:20, 5.09it/s] 18%|█▊ | 8847/48008 [30:09<2:07:30, 5.12it/s] 18%|█▊ | 8848/48008 [30:10<2:07:08, 5.13it/s] 18%|█▊ | 8849/48008 [30:10<2:07:18, 5.13it/s] 18%|█▊ | 8850/48008 [30:10<2:05:50, 5.19it/s] {'loss': 0.3, 'grad_norm': 0.1904296875, 'learning_rate': 0.0001631353107815364, 'epoch': 0.18} 18%|█▊ | 8850/48008 [30:10<2:05:50, 5.19it/s] 18%|█▊ | 8851/48008 [30:10<2:14:50, 4.84it/s] 18%|█▊ | 8852/48008 [30:10<2:11:54, 4.95it/s] 18%|█▊ | 8853/48008 [30:11<2:09:03, 5.06it/s] 18%|█▊ | 8854/48008 [30:11<2:07:44, 5.11it/s] 18%|█▊ | 8855/48008 [30:11<2:05:58, 5.18it/s] 18%|█▊ | 8856/48008 [30:11<2:05:06, 5.22it/s] 18%|█▊ | 8857/48008 [30:11<2:05:33, 5.20it/s] 18%|█▊ | 8858/48008 [30:12<2:14:22, 4.86it/s] 18%|█▊ | 8859/48008 [30:12<2:11:22, 4.97it/s] 18%|█▊ | 8860/48008 [30:12<2:11:22, 4.97it/s] 18%|█▊ | 8861/48008 [30:12<2:08:35, 5.07it/s] 18%|█▊ | 8862/48008 [30:12<2:08:25, 5.08it/s] 18%|█▊ | 8863/48008 [30:13<2:07:19, 5.12it/s] 18%|█▊ | 8864/48008 [30:13<2:08:56, 5.06it/s] 18%|█▊ | 8865/48008 [30:13<2:07:41, 5.11it/s] 18%|█▊ | 8866/48008 [30:13<2:05:27, 5.20it/s] 18%|█▊ | 8867/48008 [30:13<2:07:48, 5.10it/s] 18%|█▊ | 8868/48008 [30:14<2:06:41, 5.15it/s] 18%|█▊ | 8869/48008 [30:14<2:05:31, 5.20it/s] 18%|█▊ | 8870/48008 [30:14<2:06:29, 5.16it/s] 18%|█▊ | 8871/48008 [30:14<2:05:37, 5.19it/s] 18%|█▊ | 8872/48008 [30:14<2:05:07, 5.21it/s] 18%|█▊ | 8873/48008 [30:14<2:06:15, 5.17it/s] 18%|█▊ | 8874/48008 [30:15<2:05:48, 5.18it/s] 18%|█▊ | 8875/48008 [30:15<2:06:34, 5.15it/s] 18%|█▊ | 8876/48008 [30:15<2:06:23, 5.16it/s] 18%|█▊ | 8877/48008 [30:15<2:05:38, 5.19it/s] 18%|█▊ | 8878/48008 [30:15<2:07:10, 5.13it/s] 18%|█▊ | 8879/48008 [30:16<2:05:33, 5.19it/s] 18%|█▊ | 8880/48008 [30:16<2:04:05, 5.26it/s] 18%|█▊ | 8881/48008 [30:16<2:05:34, 5.19it/s] 19%|█▊ | 8882/48008 [30:16<2:07:05, 5.13it/s] 19%|█▊ | 8883/48008 [30:16<2:06:42, 5.15it/s] 19%|█▊ | 8884/48008 [30:17<2:44:12, 3.97it/s] 19%|█▊ | 8885/48008 [30:17<2:32:33, 4.27it/s] 19%|█▊ | 8886/48008 [30:17<2:24:01, 4.53it/s] 19%|█▊ | 8887/48008 [30:18<2:56:17, 3.70it/s] 19%|█▊ | 8888/48008 [30:18<2:42:48, 4.00it/s] 19%|█▊ | 8889/48008 [30:18<2:31:21, 4.31it/s] 19%|█▊ | 8890/48008 [30:18<2:23:40, 4.54it/s] 19%|█▊ | 8891/48008 [30:18<2:17:54, 4.73it/s] 19%|█▊ | 8892/48008 [30:19<2:14:42, 4.84it/s] 19%|█▊ | 8893/48008 [30:19<2:11:27, 4.96it/s] 19%|█▊ | 8894/48008 [30:19<2:09:18, 5.04it/s] 19%|█▊ | 8895/48008 [30:19<2:08:07, 5.09it/s] 19%|█▊ | 8896/48008 [30:19<2:07:03, 5.13it/s] 19%|█▊ | 8897/48008 [30:19<2:05:40, 5.19it/s] 19%|█▊ | 8898/48008 [30:20<2:07:26, 5.12it/s] 19%|█▊ | 8899/48008 [30:20<2:06:29, 5.15it/s] 19%|█▊ | 8900/48008 [30:20<2:05:58, 5.17it/s] {'loss': 0.3018, 'grad_norm': 0.1357421875, 'learning_rate': 0.00016292701216463924, 'epoch': 0.19} 19%|█▊ | 8900/48008 [30:20<2:05:58, 5.17it/s] 19%|█▊ | 8901/48008 [30:20<2:05:26, 5.20it/s] 19%|█▊ | 8902/48008 [30:20<2:04:57, 5.22it/s] 19%|█▊ | 8903/48008 [30:21<2:04:59, 5.21it/s] 19%|█▊ | 8904/48008 [30:21<2:04:57, 5.22it/s] 19%|█▊ | 8905/48008 [30:21<2:05:33, 5.19it/s] 19%|█▊ | 8906/48008 [30:21<2:04:42, 5.23it/s] 19%|█▊ | 8907/48008 [30:21<2:04:33, 5.23it/s] 19%|█▊ | 8908/48008 [30:22<2:04:48, 5.22it/s] 19%|█▊ | 8909/48008 [30:22<2:05:53, 5.18it/s] 19%|█▊ | 8910/48008 [30:22<2:07:32, 5.11it/s] 19%|█▊ | 8911/48008 [30:22<2:06:38, 5.15it/s] 19%|█▊ | 8912/48008 [30:22<2:08:11, 5.08it/s] 19%|█▊ | 8913/48008 [30:23<2:30:14, 4.34it/s] 19%|█▊ | 8914/48008 [30:23<2:22:24, 4.58it/s] 19%|█▊ | 8915/48008 [30:23<2:17:03, 4.75it/s] 19%|█▊ | 8916/48008 [30:23<2:13:56, 4.86it/s] 19%|█▊ | 8917/48008 [30:23<2:13:52, 4.87it/s] 19%|█▊ | 8918/48008 [30:24<2:33:33, 4.24it/s] 19%|█▊ | 8919/48008 [30:24<2:25:07, 4.49it/s] 19%|█▊ | 8920/48008 [30:24<2:19:54, 4.66it/s] 19%|█▊ | 8921/48008 [30:24<2:15:41, 4.80it/s] 19%|█▊ | 8922/48008 [30:25<2:12:03, 4.93it/s] 19%|█▊ | 8923/48008 [30:25<2:10:08, 5.01it/s] 19%|█▊ | 8924/48008 [30:25<2:09:09, 5.04it/s] 19%|█▊ | 8925/48008 [30:25<2:09:16, 5.04it/s] 19%|█▊ | 8926/48008 [30:25<2:07:51, 5.09it/s] 19%|█▊ | 8927/48008 [30:26<2:08:40, 5.06it/s] 19%|█▊ | 8928/48008 [30:26<2:09:25, 5.03it/s] 19%|█▊ | 8929/48008 [30:26<2:07:49, 5.10it/s] 19%|█▊ | 8930/48008 [30:26<2:07:03, 5.13it/s] 19%|█▊ | 8931/48008 [30:26<2:08:18, 5.08it/s] 19%|█▊ | 8932/48008 [30:27<2:09:25, 5.03it/s] 19%|█▊ | 8933/48008 [30:27<2:06:25, 5.15it/s] 19%|█▊ | 8934/48008 [30:27<2:04:40, 5.22it/s] 19%|█▊ | 8935/48008 [30:27<2:27:21, 4.42it/s] 19%|█▊ | 8936/48008 [30:27<2:21:13, 4.61it/s] 19%|█▊ | 8937/48008 [30:28<2:17:12, 4.75it/s] 19%|█▊ | 8938/48008 [30:28<2:14:23, 4.85it/s] 19%|█▊ | 8939/48008 [30:28<2:10:22, 4.99it/s] 19%|█▊ | 8940/48008 [30:28<2:07:40, 5.10it/s] 19%|█▊ | 8941/48008 [30:28<2:06:45, 5.14it/s] 19%|█▊ | 8942/48008 [30:29<2:07:20, 5.11it/s] 19%|█▊ | 8943/48008 [30:29<2:06:28, 5.15it/s] 19%|█▊ | 8944/48008 [30:29<2:05:31, 5.19it/s] 19%|█▊ | 8945/48008 [30:29<2:05:41, 5.18it/s] 19%|█▊ | 8946/48008 [30:29<2:05:14, 5.20it/s] 19%|█▊ | 8947/48008 [30:30<2:04:46, 5.22it/s] 19%|█▊ | 8948/48008 [30:30<2:04:51, 5.21it/s] 19%|█▊ | 8949/48008 [30:30<2:08:02, 5.08it/s] 19%|█▊ | 8950/48008 [30:30<2:07:11, 5.12it/s] {'loss': 0.3015, 'grad_norm': 0.271484375, 'learning_rate': 0.00016271871354774203, 'epoch': 0.19} 19%|█▊ | 8950/48008 [30:30<2:07:11, 5.12it/s] 19%|█▊ | 8951/48008 [30:30<2:06:54, 5.13it/s] 19%|█▊ | 8952/48008 [30:30<2:08:13, 5.08it/s] 19%|█▊ | 8953/48008 [30:31<2:06:39, 5.14it/s] 19%|█▊ | 8954/48008 [30:31<2:05:57, 5.17it/s] 19%|█▊ | 8955/48008 [30:31<2:05:20, 5.19it/s] 19%|█▊ | 8956/48008 [30:31<2:04:23, 5.23it/s] 19%|█▊ | 8957/48008 [30:31<2:04:01, 5.25it/s] 19%|█▊ | 8958/48008 [30:32<2:02:51, 5.30it/s] 19%|█▊ | 8959/48008 [30:32<2:06:27, 5.15it/s] 19%|█▊ | 8960/48008 [30:32<2:06:41, 5.14it/s] 19%|█▊ | 8961/48008 [30:32<2:05:56, 5.17it/s] 19%|█▊ | 8962/48008 [30:32<2:05:15, 5.20it/s] 19%|█▊ | 8963/48008 [30:33<2:02:48, 5.30it/s] 19%|█▊ | 8964/48008 [30:33<2:01:53, 5.34it/s] 19%|█▊ | 8965/48008 [30:33<2:01:53, 5.34it/s] 19%|█▊ | 8966/48008 [30:33<2:02:26, 5.31it/s] 19%|█▊ | 8967/48008 [30:33<2:01:36, 5.35it/s] 19%|█▊ | 8968/48008 [30:34<2:02:33, 5.31it/s] 19%|█▊ | 8969/48008 [30:34<2:05:04, 5.20it/s] 19%|█▊ | 8970/48008 [30:34<2:04:56, 5.21it/s] 19%|█▊ | 8971/48008 [30:34<2:07:02, 5.12it/s] 19%|█▊ | 8972/48008 [30:34<2:08:23, 5.07it/s] 19%|█▊ | 8973/48008 [30:35<2:08:53, 5.05it/s] 19%|█▊ | 8974/48008 [30:35<2:06:55, 5.13it/s] 19%|█▊ | 8975/48008 [30:35<2:06:12, 5.15it/s] 19%|█▊ | 8976/48008 [30:35<2:05:43, 5.17it/s] 19%|█▊ | 8977/48008 [30:35<2:07:14, 5.11it/s] 19%|█▊ | 8978/48008 [30:35<2:06:24, 5.15it/s] 19%|█▊ | 8979/48008 [30:36<2:05:05, 5.20it/s] 19%|█▊ | 8980/48008 [30:36<2:04:33, 5.22it/s] 19%|█▊ | 8981/48008 [30:36<2:05:52, 5.17it/s] 19%|█▊ | 8982/48008 [30:36<2:05:01, 5.20it/s] 19%|█▊ | 8983/48008 [30:36<2:06:15, 5.15it/s] 19%|█▊ | 8984/48008 [30:37<2:07:56, 5.08it/s] 19%|█▊ | 8985/48008 [30:37<2:08:50, 5.05it/s] 19%|█▊ | 8986/48008 [30:37<2:07:18, 5.11it/s] 19%|█▊ | 8987/48008 [30:37<2:05:50, 5.17it/s] 19%|█▊ | 8988/48008 [30:37<2:05:13, 5.19it/s] 19%|█▊ | 8989/48008 [30:38<2:05:51, 5.17it/s] 19%|█▊ | 8990/48008 [30:38<2:05:41, 5.17it/s] 19%|█▊ | 8991/48008 [30:38<2:04:59, 5.20it/s] 19%|█▊ | 8992/48008 [30:38<2:08:08, 5.07it/s] 19%|█▊ | 8993/48008 [30:38<2:07:49, 5.09it/s] 19%|█▊ | 8994/48008 [30:39<2:05:45, 5.17it/s] 19%|█▊ | 8995/48008 [30:39<2:05:04, 5.20it/s] 19%|█▊ | 8996/48008 [30:39<2:04:40, 5.22it/s] 19%|█▊ | 8997/48008 [30:39<2:07:49, 5.09it/s] 19%|█▊ | 8998/48008 [30:39<2:06:38, 5.13it/s] 19%|█▊ | 8999/48008 [30:40<2:05:43, 5.17it/s] 19%|█▊ | 9000/48008 [30:40<2:05:16, 5.19it/s] {'loss': 0.3031, 'grad_norm': 0.59375, 'learning_rate': 0.00016251041493084488, 'epoch': 0.19} 19%|█▊ | 9000/48008 [30:40<2:05:16, 5.19it/s] 19%|█▊ | 9001/48008 [30:40<2:05:32, 5.18it/s] 19%|█▉ | 9002/48008 [30:40<2:03:56, 5.25it/s] 19%|█▉ | 9003/48008 [30:40<2:26:39, 4.43it/s] 19%|█▉ | 9004/48008 [30:41<2:21:56, 4.58it/s] 19%|█▉ | 9005/48008 [30:41<2:15:30, 4.80it/s] 19%|█▉ | 9006/48008 [30:41<2:12:24, 4.91it/s] 19%|█▉ | 9007/48008 [30:41<2:10:00, 5.00it/s] 19%|█▉ | 9008/48008 [30:41<2:17:07, 4.74it/s] 19%|█▉ | 9009/48008 [30:42<2:13:28, 4.87it/s] 19%|█▉ | 9010/48008 [30:42<2:10:22, 4.99it/s] 19%|█▉ | 9011/48008 [30:42<2:07:58, 5.08it/s] 19%|█▉ | 9012/48008 [30:42<2:08:02, 5.08it/s] 19%|█▉ | 9013/48008 [30:42<2:06:40, 5.13it/s] 19%|█▉ | 9014/48008 [30:43<2:06:16, 5.15it/s] 19%|█▉ | 9015/48008 [30:43<2:05:09, 5.19it/s] 19%|█▉ | 9016/48008 [30:43<2:04:49, 5.21it/s] 19%|█▉ | 9017/48008 [30:43<2:05:47, 5.17it/s] 19%|█▉ | 9018/48008 [30:43<2:07:11, 5.11it/s] 19%|█▉ | 9019/48008 [30:44<2:06:09, 5.15it/s] 19%|█▉ | 9020/48008 [30:44<2:05:21, 5.18it/s] 19%|█▉ | 9021/48008 [30:44<2:04:44, 5.21it/s] 19%|█▉ | 9022/48008 [30:44<2:05:37, 5.17it/s] 19%|█▉ | 9023/48008 [30:44<2:07:45, 5.09it/s] 19%|█▉ | 9024/48008 [30:45<2:06:13, 5.15it/s] 19%|█▉ | 9025/48008 [30:45<2:04:46, 5.21it/s] 19%|█▉ | 9026/48008 [30:45<2:03:34, 5.26it/s] 19%|█▉ | 9027/48008 [30:45<2:05:58, 5.16it/s] 19%|█▉ | 9028/48008 [30:45<2:07:24, 5.10it/s] 19%|█▉ | 9029/48008 [30:46<2:06:29, 5.14it/s] 19%|█▉ | 9030/48008 [30:46<2:04:59, 5.20it/s] 19%|█▉ | 9031/48008 [30:46<2:03:47, 5.25it/s] 19%|█▉ | 9032/48008 [30:46<2:04:55, 5.20it/s] 19%|█▉ | 9033/48008 [30:46<2:05:17, 5.18it/s] 19%|█▉ | 9034/48008 [30:46<2:04:32, 5.22it/s] 19%|█▉ | 9035/48008 [30:47<2:04:40, 5.21it/s] 19%|█▉ | 9036/48008 [30:47<2:06:40, 5.13it/s] 19%|█▉ | 9037/48008 [30:47<2:09:15, 5.03it/s] 19%|█▉ | 9038/48008 [30:47<2:07:28, 5.10it/s] 19%|█▉ | 9039/48008 [30:47<2:06:28, 5.14it/s] 19%|█▉ | 9040/48008 [30:48<2:05:59, 5.15it/s] 19%|█▉ | 9041/48008 [30:48<2:07:22, 5.10it/s] 19%|█▉ | 9042/48008 [30:48<2:07:47, 5.08it/s] 19%|█▉ | 9043/48008 [30:48<2:06:32, 5.13it/s] 19%|█▉ | 9044/48008 [30:48<2:14:52, 4.81it/s] 19%|█▉ | 9045/48008 [30:49<2:11:26, 4.94it/s] 19%|█▉ | 9046/48008 [30:49<2:09:04, 5.03it/s] 19%|█▉ | 9047/48008 [30:49<2:06:36, 5.13it/s] 19%|█▉ | 9048/48008 [30:49<2:06:50, 5.12it/s] 19%|█▉ | 9049/48008 [30:49<2:04:51, 5.20it/s] 19%|█▉ | 9050/48008 [30:50<2:04:16, 5.22it/s] {'loss': 0.3019, 'grad_norm': 0.1357421875, 'learning_rate': 0.0001623021163139477, 'epoch': 0.19} 19%|█▉ | 9050/48008 [30:50<2:04:16, 5.22it/s] 19%|█▉ | 9051/48008 [30:50<2:04:12, 5.23it/s] 19%|█▉ | 9052/48008 [30:50<2:04:15, 5.23it/s] 19%|█▉ | 9053/48008 [30:50<2:03:21, 5.26it/s] 19%|█▉ | 9054/48008 [30:50<2:02:20, 5.31it/s] 19%|█▉ | 9055/48008 [30:51<2:03:09, 5.27it/s] 19%|█▉ | 9056/48008 [30:51<2:03:35, 5.25it/s] 19%|█▉ | 9057/48008 [30:51<2:06:16, 5.14it/s] 19%|█▉ | 9058/48008 [30:51<2:05:25, 5.18it/s] 19%|█▉ | 9059/48008 [30:51<2:04:18, 5.22it/s] 19%|█▉ | 9060/48008 [30:52<2:42:19, 4.00it/s] 19%|█▉ | 9061/48008 [30:52<2:30:46, 4.31it/s] 19%|█▉ | 9062/48008 [30:52<2:22:54, 4.54it/s] 19%|█▉ | 9063/48008 [30:52<2:17:41, 4.71it/s] 19%|█▉ | 9064/48008 [30:52<2:12:53, 4.88it/s] 19%|█▉ | 9065/48008 [30:53<2:10:15, 4.98it/s] 19%|█▉ | 9066/48008 [30:53<2:17:44, 4.71it/s] 19%|█▉ | 9067/48008 [30:53<2:12:16, 4.91it/s] 19%|█▉ | 9068/48008 [30:53<2:08:47, 5.04it/s] 19%|█▉ | 9069/48008 [30:53<2:06:52, 5.12it/s] 19%|█▉ | 9070/48008 [30:54<2:05:25, 5.17it/s] 19%|█▉ | 9071/48008 [30:54<2:07:53, 5.07it/s] 19%|█▉ | 9072/48008 [30:54<2:07:52, 5.07it/s] 19%|█▉ | 9073/48008 [30:54<2:06:29, 5.13it/s] 19%|█▉ | 9074/48008 [30:54<2:06:55, 5.11it/s] 19%|█▉ | 9075/48008 [30:55<2:06:06, 5.15it/s] 19%|█▉ | 9076/48008 [30:55<2:05:20, 5.18it/s] 19%|█▉ | 9077/48008 [30:55<2:05:09, 5.18it/s] 19%|█▉ | 9078/48008 [30:55<2:06:03, 5.15it/s] 19%|█▉ | 9079/48008 [30:55<2:07:27, 5.09it/s] 19%|█▉ | 9080/48008 [30:56<2:05:48, 5.16it/s] 19%|█▉ | 9081/48008 [30:56<2:08:44, 5.04it/s] 19%|█▉ | 9082/48008 [30:56<2:09:15, 5.02it/s] 19%|█▉ | 9083/48008 [30:56<2:08:25, 5.05it/s] 19%|█▉ | 9084/48008 [30:56<2:09:01, 5.03it/s] 19%|█▉ | 9085/48008 [30:57<2:07:28, 5.09it/s] 19%|█▉ | 9086/48008 [30:57<2:06:24, 5.13it/s] 19%|█▉ | 9087/48008 [30:57<2:05:02, 5.19it/s] 19%|█▉ | 9088/48008 [30:57<2:03:37, 5.25it/s] 19%|█▉ | 9089/48008 [30:57<2:05:04, 5.19it/s] 19%|█▉ | 9090/48008 [30:58<2:04:37, 5.20it/s] 19%|█▉ | 9091/48008 [30:58<2:06:45, 5.12it/s] 19%|█▉ | 9092/48008 [30:58<2:06:13, 5.14it/s] 19%|█▉ | 9093/48008 [30:58<2:14:48, 4.81it/s] 19%|█▉ | 9094/48008 [30:58<2:11:50, 4.92it/s] 19%|█▉ | 9095/48008 [30:59<2:09:15, 5.02it/s] 19%|█▉ | 9096/48008 [30:59<2:08:05, 5.06it/s] 19%|█▉ | 9097/48008 [30:59<2:15:58, 4.77it/s] 19%|█▉ | 9098/48008 [30:59<2:12:18, 4.90it/s] 19%|█▉ | 9099/48008 [30:59<2:10:04, 4.99it/s] 19%|█▉ | 9100/48008 [31:00<2:10:10, 4.98it/s] {'loss': 0.3046, 'grad_norm': 0.314453125, 'learning_rate': 0.00016209381769705049, 'epoch': 0.19} 19%|█▉ | 9100/48008 [31:00<2:10:10, 4.98it/s] 19%|█▉ | 9101/48008 [31:00<2:10:03, 4.99it/s] 19%|█▉ | 9102/48008 [31:00<2:07:40, 5.08it/s] 19%|█▉ | 9103/48008 [31:00<2:08:55, 5.03it/s] 19%|█▉ | 9104/48008 [31:00<2:07:24, 5.09it/s] 19%|█▉ | 9105/48008 [31:01<2:06:06, 5.14it/s] 19%|█▉ | 9106/48008 [31:01<2:03:34, 5.25it/s] 19%|█▉ | 9107/48008 [31:01<2:12:31, 4.89it/s] 19%|█▉ | 9108/48008 [31:01<2:11:10, 4.94it/s] 19%|█▉ | 9109/48008 [31:01<2:08:44, 5.04it/s] 19%|█▉ | 9110/48008 [31:02<2:06:59, 5.11it/s] 19%|█▉ | 9111/48008 [31:02<2:08:27, 5.05it/s] 19%|█▉ | 9112/48008 [31:02<2:44:55, 3.93it/s] 19%|█▉ | 9113/48008 [31:02<2:32:08, 4.26it/s] 19%|█▉ | 9114/48008 [31:03<2:26:01, 4.44it/s] 19%|█▉ | 9115/48008 [31:03<2:20:41, 4.61it/s] 19%|█▉ | 9116/48008 [31:03<2:19:08, 4.66it/s] 19%|█▉ | 9117/48008 [31:03<2:13:22, 4.86it/s] 19%|█▉ | 9118/48008 [31:03<2:10:28, 4.97it/s] 19%|█▉ | 9119/48008 [31:04<2:09:27, 5.01it/s] 19%|█▉ | 9120/48008 [31:04<2:07:42, 5.08it/s] 19%|█▉ | 9121/48008 [31:04<2:07:24, 5.09it/s] 19%|█▉ | 9122/48008 [31:04<2:09:25, 5.01it/s] 19%|█▉ | 9123/48008 [31:04<2:06:44, 5.11it/s] 19%|█▉ | 9124/48008 [31:04<2:05:58, 5.14it/s] 19%|█▉ | 9125/48008 [31:05<2:04:48, 5.19it/s] 19%|█▉ | 9126/48008 [31:05<2:04:37, 5.20it/s] 19%|█▉ | 9127/48008 [31:05<2:05:36, 5.16it/s] 19%|█▉ | 9128/48008 [31:05<2:04:48, 5.19it/s] 19%|█▉ | 9129/48008 [31:05<2:03:26, 5.25it/s] 19%|█▉ | 9130/48008 [31:06<2:03:11, 5.26it/s] 19%|█▉ | 9131/48008 [31:06<2:04:30, 5.20it/s] 19%|█▉ | 9132/48008 [31:06<2:26:42, 4.42it/s] 19%|█▉ | 9133/48008 [31:06<2:19:32, 4.64it/s] 19%|█▉ | 9134/48008 [31:07<2:14:46, 4.81it/s] 19%|█▉ | 9135/48008 [31:07<2:10:24, 4.97it/s] 19%|█▉ | 9136/48008 [31:07<2:07:54, 5.07it/s] 19%|█▉ | 9137/48008 [31:07<2:06:45, 5.11it/s] 19%|█▉ | 9138/48008 [31:07<2:05:59, 5.14it/s] 19%|█▉ | 9139/48008 [31:07<2:03:59, 5.23it/s] 19%|█▉ | 9140/48008 [31:08<2:03:31, 5.24it/s] 19%|█▉ | 9141/48008 [31:08<2:03:37, 5.24it/s] 19%|█▉ | 9142/48008 [31:08<2:03:26, 5.25it/s] 19%|█▉ | 9143/48008 [31:08<2:06:41, 5.11it/s] 19%|█▉ | 9144/48008 [31:08<2:05:28, 5.16it/s] 19%|█▉ | 9145/48008 [31:09<2:43:03, 3.97it/s] 19%|█▉ | 9146/48008 [31:09<2:30:51, 4.29it/s] 19%|█▉ | 9147/48008 [31:09<2:22:17, 4.55it/s] 19%|█▉ | 9148/48008 [31:09<2:19:02, 4.66it/s] 19%|█▉ | 9149/48008 [31:10<2:14:35, 4.81it/s] 19%|█▉ | 9150/48008 [31:10<2:10:59, 4.94it/s] {'loss': 0.3025, 'grad_norm': 0.40234375, 'learning_rate': 0.00016188551908015333, 'epoch': 0.19} 19%|█▉ | 9150/48008 [31:10<2:10:59, 4.94it/s] 19%|█▉ | 9151/48008 [31:10<2:46:56, 3.88it/s] 19%|█▉ | 9152/48008 [31:10<2:33:44, 4.21it/s] 19%|█▉ | 9153/48008 [31:11<3:02:34, 3.55it/s] 19%|█▉ | 9154/48008 [31:11<2:46:37, 3.89it/s] 19%|█▉ | 9155/48008 [31:11<2:36:06, 4.15it/s] 19%|█▉ | 9156/48008 [31:11<2:25:31, 4.45it/s] 19%|█▉ | 9157/48008 [31:12<2:21:32, 4.57it/s] 19%|█▉ | 9158/48008 [31:12<2:15:48, 4.77it/s] 19%|█▉ | 9159/48008 [31:12<2:12:06, 4.90it/s] 19%|█▉ | 9160/48008 [31:12<2:08:37, 5.03it/s] 19%|█▉ | 9161/48008 [31:12<2:44:57, 3.92it/s] 19%|█▉ | 9162/48008 [31:13<2:31:45, 4.27it/s] 19%|█▉ | 9163/48008 [31:13<2:25:36, 4.45it/s] 19%|█▉ | 9164/48008 [31:13<2:19:58, 4.63it/s] 19%|█▉ | 9165/48008 [31:13<2:52:55, 3.74it/s] 19%|█▉ | 9166/48008 [31:14<2:39:45, 4.05it/s] 19%|█▉ | 9167/48008 [31:14<2:28:43, 4.35it/s] 19%|█▉ | 9168/48008 [31:14<2:20:19, 4.61it/s] 19%|█▉ | 9169/48008 [31:14<2:15:11, 4.79it/s] 19%|█▉ | 9170/48008 [31:14<2:11:20, 4.93it/s] 19%|█▉ | 9171/48008 [31:15<2:09:08, 5.01it/s] 19%|█▉ | 9172/48008 [31:15<2:07:48, 5.06it/s] 19%|█▉ | 9173/48008 [31:15<2:15:43, 4.77it/s] 19%|█▉ | 9174/48008 [31:15<2:11:34, 4.92it/s] 19%|█▉ | 9175/48008 [31:15<2:09:19, 5.00it/s] 19%|█▉ | 9176/48008 [31:16<2:07:13, 5.09it/s] 19%|█▉ | 9177/48008 [31:16<2:07:00, 5.10it/s] 19%|█▉ | 9178/48008 [31:16<2:07:50, 5.06it/s] 19%|█▉ | 9179/48008 [31:16<2:05:45, 5.15it/s] 19%|█▉ | 9180/48008 [31:16<2:04:49, 5.18it/s] 19%|█▉ | 9181/48008 [31:17<2:03:27, 5.24it/s] 19%|█▉ | 9182/48008 [31:17<2:03:11, 5.25it/s] 19%|█▉ | 9183/48008 [31:17<2:03:02, 5.26it/s] 19%|█▉ | 9184/48008 [31:17<2:03:27, 5.24it/s] 19%|█▉ | 9185/48008 [31:17<2:03:47, 5.23it/s] 19%|█▉ | 9186/48008 [31:18<2:05:59, 5.14it/s] 19%|█▉ | 9187/48008 [31:18<2:04:57, 5.18it/s] 19%|█▉ | 9188/48008 [31:18<2:04:40, 5.19it/s] 19%|█▉ | 9189/48008 [31:18<2:04:17, 5.21it/s] 19%|█▉ | 9190/48008 [31:18<2:07:47, 5.06it/s] 19%|█▉ | 9191/48008 [31:18<2:08:49, 5.02it/s] 19%|█▉ | 9192/48008 [31:19<2:06:43, 5.10it/s] 19%|█▉ | 9193/48008 [31:19<2:05:42, 5.15it/s] 19%|█▉ | 9194/48008 [31:19<2:04:46, 5.18it/s] 19%|█▉ | 9195/48008 [31:19<2:03:23, 5.24it/s] 19%|█▉ | 9196/48008 [31:19<2:12:25, 4.88it/s] 19%|█▉ | 9197/48008 [31:20<2:08:21, 5.04it/s] 19%|█▉ | 9198/48008 [31:20<2:10:35, 4.95it/s] 19%|█▉ | 9199/48008 [31:20<2:08:02, 5.05it/s] 19%|█▉ | 9200/48008 [31:20<2:06:21, 5.12it/s] {'loss': 0.3022, 'grad_norm': 0.30859375, 'learning_rate': 0.00016167722046325612, 'epoch': 0.19} 19%|█▉ | 9200/48008 [31:20<2:06:21, 5.12it/s] 19%|█▉ | 9201/48008 [31:20<2:14:26, 4.81it/s] 19%|█▉ | 9202/48008 [31:21<2:12:13, 4.89it/s] 19%|█▉ | 9203/48008 [31:21<2:18:32, 4.67it/s] 19%|█▉ | 9204/48008 [31:21<2:13:36, 4.84it/s] 19%|█▉ | 9205/48008 [31:21<2:12:37, 4.88it/s] 19%|█▉ | 9206/48008 [31:22<2:13:24, 4.85it/s] 19%|█▉ | 9207/48008 [31:22<2:11:15, 4.93it/s] 19%|█▉ | 9208/48008 [31:22<2:07:49, 5.06it/s] 19%|█▉ | 9209/48008 [31:22<2:06:06, 5.13it/s] 19%|█▉ | 9210/48008 [31:22<2:06:31, 5.11it/s] 19%|█▉ | 9211/48008 [31:22<2:05:45, 5.14it/s] 19%|█▉ | 9212/48008 [31:23<2:04:35, 5.19it/s] 19%|█▉ | 9213/48008 [31:23<2:05:15, 5.16it/s] 19%|█▉ | 9214/48008 [31:23<2:08:28, 5.03it/s] 19%|█▉ | 9215/48008 [31:23<2:29:22, 4.33it/s] 19%|█▉ | 9216/48008 [31:24<2:21:50, 4.56it/s] 19%|█▉ | 9217/48008 [31:24<2:25:13, 4.45it/s] 19%|█▉ | 9218/48008 [31:24<2:18:34, 4.67it/s] 19%|█▉ | 9219/48008 [31:24<2:14:16, 4.81it/s] 19%|█▉ | 9220/48008 [31:24<2:10:54, 4.94it/s] 19%|█▉ | 9221/48008 [31:25<2:10:44, 4.94it/s] 19%|█▉ | 9222/48008 [31:25<2:08:20, 5.04it/s] 19%|█▉ | 9223/48008 [31:25<2:06:41, 5.10it/s] 19%|█▉ | 9224/48008 [31:25<2:06:17, 5.12it/s] 19%|█▉ | 9225/48008 [31:25<2:05:36, 5.15it/s] 19%|█▉ | 9226/48008 [31:26<2:04:42, 5.18it/s] 19%|█▉ | 9227/48008 [31:26<2:03:55, 5.22it/s] 19%|█▉ | 9228/48008 [31:26<2:03:03, 5.25it/s] 19%|█▉ | 9229/48008 [31:26<2:03:17, 5.24it/s] 19%|█▉ | 9230/48008 [31:26<2:02:49, 5.26it/s] 19%|█▉ | 9231/48008 [31:27<2:25:29, 4.44it/s] 19%|█▉ | 9232/48008 [31:27<2:19:53, 4.62it/s] 19%|█▉ | 9233/48008 [31:27<2:15:48, 4.76it/s] 19%|█▉ | 9234/48008 [31:27<2:11:48, 4.90it/s] 19%|█▉ | 9235/48008 [31:27<2:09:08, 5.00it/s] 19%|█▉ | 9236/48008 [31:28<2:06:33, 5.11it/s] 19%|█▉ | 9237/48008 [31:28<2:05:54, 5.13it/s] 19%|█▉ | 9238/48008 [31:28<2:04:33, 5.19it/s] 19%|█▉ | 9239/48008 [31:28<2:05:27, 5.15it/s] 19%|█▉ | 9240/48008 [31:28<2:04:48, 5.18it/s] 19%|█▉ | 9241/48008 [31:29<2:05:14, 5.16it/s] 19%|█▉ | 9242/48008 [31:29<2:05:44, 5.14it/s] 19%|█▉ | 9243/48008 [31:29<2:04:08, 5.20it/s] 19%|█▉ | 9244/48008 [31:29<2:03:08, 5.25it/s] 19%|█▉ | 9245/48008 [31:29<2:03:15, 5.24it/s] 19%|█▉ | 9246/48008 [31:30<2:25:49, 4.43it/s] 19%|█▉ | 9247/48008 [31:30<2:19:07, 4.64it/s] 19%|█▉ | 9248/48008 [31:30<2:13:55, 4.82it/s] 19%|█▉ | 9249/48008 [31:30<2:10:46, 4.94it/s] 19%|█▉ | 9250/48008 [31:30<2:08:42, 5.02it/s] {'loss': 0.3011, 'grad_norm': 0.275390625, 'learning_rate': 0.00016146892184635897, 'epoch': 0.19} 19%|█▉ | 9250/48008 [31:30<2:08:42, 5.02it/s] 19%|█▉ | 9251/48008 [31:31<2:08:50, 5.01it/s] 19%|█▉ | 9252/48008 [31:31<2:07:51, 5.05it/s] 19%|█▉ | 9253/48008 [31:31<2:05:37, 5.14it/s] 19%|█▉ | 9254/48008 [31:31<2:04:24, 5.19it/s] 19%|█▉ | 9255/48008 [31:31<2:03:09, 5.24it/s] 19%|█▉ | 9256/48008 [31:32<2:05:34, 5.14it/s] 19%|█▉ | 9257/48008 [31:32<2:07:28, 5.07it/s] 19%|█▉ | 9258/48008 [31:32<2:05:31, 5.15it/s] 19%|█▉ | 9259/48008 [31:32<2:03:00, 5.25it/s] 19%|█▉ | 9260/48008 [31:32<2:03:08, 5.24it/s] 19%|█▉ | 9261/48008 [31:32<2:04:26, 5.19it/s] 19%|█▉ | 9262/48008 [31:33<2:03:53, 5.21it/s] 19%|█▉ | 9263/48008 [31:33<2:05:00, 5.17it/s] 19%|█▉ | 9264/48008 [31:33<2:05:00, 5.17it/s] 19%|█▉ | 9265/48008 [31:33<2:04:10, 5.20it/s] 19%|█▉ | 9266/48008 [31:33<2:12:52, 4.86it/s] 19%|█▉ | 9267/48008 [31:34<2:09:45, 4.98it/s] 19%|█▉ | 9268/48008 [31:34<2:08:31, 5.02it/s] 19%|█▉ | 9269/48008 [31:34<2:09:00, 5.00it/s] 19%|█▉ | 9270/48008 [31:34<2:45:09, 3.91it/s] 19%|█▉ | 9271/48008 [31:35<2:33:22, 4.21it/s] 19%|█▉ | 9272/48008 [31:35<2:23:47, 4.49it/s] 19%|█▉ | 9273/48008 [31:35<2:39:55, 4.04it/s] 19%|█▉ | 9274/48008 [31:35<2:31:06, 4.27it/s] 19%|█▉ | 9275/48008 [31:36<2:21:49, 4.55it/s] 19%|█▉ | 9276/48008 [31:36<2:16:15, 4.74it/s] 19%|█▉ | 9277/48008 [31:36<2:21:05, 4.57it/s] 19%|█▉ | 9278/48008 [31:36<2:15:42, 4.76it/s] 19%|█▉ | 9279/48008 [31:36<2:11:03, 4.93it/s] 19%|█▉ | 9280/48008 [31:37<2:08:10, 5.04it/s] 19%|█▉ | 9281/48008 [31:37<2:06:58, 5.08it/s] 19%|█▉ | 9282/48008 [31:37<2:07:50, 5.05it/s] 19%|█▉ | 9283/48008 [31:37<2:06:08, 5.12it/s] 19%|█▉ | 9284/48008 [31:37<2:04:26, 5.19it/s] 19%|█▉ | 9285/48008 [31:38<2:05:31, 5.14it/s] 19%|█▉ | 9286/48008 [31:38<2:05:15, 5.15it/s] 19%|█▉ | 9287/48008 [31:38<2:04:07, 5.20it/s] 19%|█▉ | 9288/48008 [31:38<2:12:46, 4.86it/s] 19%|█▉ | 9289/48008 [31:38<2:09:46, 4.97it/s] 19%|█▉ | 9290/48008 [31:39<2:16:55, 4.71it/s] 19%|█▉ | 9291/48008 [31:39<2:12:38, 4.86it/s] 19%|█▉ | 9292/48008 [31:39<2:09:39, 4.98it/s] 19%|█▉ | 9293/48008 [31:39<2:07:27, 5.06it/s] 19%|█▉ | 9294/48008 [31:39<2:06:24, 5.10it/s] 19%|█▉ | 9295/48008 [31:40<2:05:29, 5.14it/s] 19%|█▉ | 9296/48008 [31:40<2:04:46, 5.17it/s] 19%|█▉ | 9297/48008 [31:40<2:03:54, 5.21it/s] 19%|█▉ | 9298/48008 [31:40<2:02:54, 5.25it/s] 19%|█▉ | 9299/48008 [31:40<2:02:49, 5.25it/s] 19%|█▉ | 9300/48008 [31:40<2:02:57, 5.25it/s] {'loss': 0.3034, 'grad_norm': 0.275390625, 'learning_rate': 0.00016126062322946176, 'epoch': 0.19} 19%|█▉ | 9300/48008 [31:40<2:02:57, 5.25it/s] 19%|█▉ | 9301/48008 [31:41<2:02:53, 5.25it/s] 19%|█▉ | 9302/48008 [31:41<2:05:22, 5.15it/s] 19%|█▉ | 9303/48008 [31:41<2:04:28, 5.18it/s] 19%|█▉ | 9304/48008 [31:41<2:03:54, 5.21it/s] 19%|█▉ | 9305/48008 [31:41<2:04:58, 5.16it/s] 19%|█▉ | 9306/48008 [31:42<2:04:11, 5.19it/s] 19%|█▉ | 9307/48008 [31:42<2:06:10, 5.11it/s] 19%|█▉ | 9308/48008 [31:42<2:07:07, 5.07it/s] 19%|█▉ | 9309/48008 [31:42<2:05:40, 5.13it/s] 19%|█▉ | 9310/48008 [31:42<2:04:28, 5.18it/s] 19%|█▉ | 9311/48008 [31:43<2:03:14, 5.23it/s] 19%|█▉ | 9312/48008 [31:43<2:03:00, 5.24it/s] 19%|█▉ | 9313/48008 [31:43<2:01:16, 5.32it/s] 19%|█▉ | 9314/48008 [31:43<2:03:44, 5.21it/s] 19%|█▉ | 9315/48008 [31:43<2:03:38, 5.22it/s] 19%|█▉ | 9316/48008 [31:44<2:05:51, 5.12it/s] 19%|█▉ | 9317/48008 [31:44<2:03:49, 5.21it/s] 19%|█▉ | 9318/48008 [31:44<2:03:20, 5.23it/s] 19%|█▉ | 9319/48008 [31:44<2:05:45, 5.13it/s] 19%|█▉ | 9320/48008 [31:44<2:13:34, 4.83it/s] 19%|█▉ | 9321/48008 [31:45<2:10:14, 4.95it/s] 19%|█▉ | 9322/48008 [31:45<2:16:53, 4.71it/s] 19%|█▉ | 9323/48008 [31:45<2:12:48, 4.85it/s] 19%|█▉ | 9324/48008 [31:45<2:09:40, 4.97it/s] 19%|█▉ | 9325/48008 [31:45<2:07:34, 5.05it/s] 19%|█▉ | 9326/48008 [31:46<2:06:00, 5.12it/s] 19%|█▉ | 9327/48008 [31:46<2:04:56, 5.16it/s] 19%|█▉ | 9328/48008 [31:46<2:03:22, 5.22it/s] 19%|█▉ | 9329/48008 [31:46<2:01:52, 5.29it/s] 19%|█▉ | 9330/48008 [31:46<2:04:21, 5.18it/s] 19%|█▉ | 9331/48008 [31:46<2:03:32, 5.22it/s] 19%|█▉ | 9332/48008 [31:47<2:03:32, 5.22it/s] 19%|█▉ | 9333/48008 [31:47<2:06:45, 5.09it/s] 19%|█▉ | 9334/48008 [31:47<2:05:23, 5.14it/s] 19%|█▉ | 9335/48008 [31:47<2:07:05, 5.07it/s] 19%|█▉ | 9336/48008 [31:47<2:06:03, 5.11it/s] 19%|█▉ | 9337/48008 [31:48<2:05:04, 5.15it/s] 19%|█▉ | 9338/48008 [31:48<2:26:53, 4.39it/s] 19%|█▉ | 9339/48008 [31:48<2:19:33, 4.62it/s] 19%|█▉ | 9340/48008 [31:48<2:12:41, 4.86it/s] 19%|█▉ | 9341/48008 [31:49<2:10:11, 4.95it/s] 19%|█▉ | 9342/48008 [31:49<2:07:48, 5.04it/s] 19%|█▉ | 9343/48008 [31:49<2:06:02, 5.11it/s] 19%|█▉ | 9344/48008 [31:49<2:05:09, 5.15it/s] 19%|█▉ | 9345/48008 [31:49<2:05:22, 5.14it/s] 19%|█▉ | 9346/48008 [31:50<4:23:59, 2.44it/s] 19%|█▉ | 9347/48008 [31:50<3:41:34, 2.91it/s] 19%|█▉ | 9348/48008 [31:51<3:20:53, 3.21it/s] 19%|█▉ | 9349/48008 [31:51<2:56:51, 3.64it/s] 19%|█▉ | 9350/48008 [31:51<2:40:08, 4.02it/s] {'loss': 0.302, 'grad_norm': 0.2353515625, 'learning_rate': 0.0001610523246125646, 'epoch': 0.19} 19%|█▉ | 9350/48008 [31:51<2:40:08, 4.02it/s] 19%|█▉ | 9351/48008 [31:51<2:31:28, 4.25it/s] 19%|█▉ | 9352/48008 [31:51<2:23:20, 4.49it/s] 19%|█▉ | 9353/48008 [31:52<2:19:33, 4.62it/s] 19%|█▉ | 9354/48008 [31:52<2:17:53, 4.67it/s] 19%|█▉ | 9355/48008 [31:52<2:36:13, 4.12it/s] 19%|█▉ | 9356/48008 [31:52<2:27:55, 4.35it/s] 19%|█▉ | 9357/48008 [31:53<2:20:07, 4.60it/s] 19%|█▉ | 9358/48008 [31:53<2:16:13, 4.73it/s] 19%|█▉ | 9359/48008 [31:53<2:12:17, 4.87it/s] 19%|█▉ | 9360/48008 [31:53<2:10:59, 4.92it/s] 19%|█▉ | 9361/48008 [31:53<2:08:36, 5.01it/s] 20%|█▉ | 9362/48008 [31:53<2:05:35, 5.13it/s] 20%|█▉ | 9363/48008 [31:54<2:14:18, 4.80it/s] 20%|█▉ | 9364/48008 [31:54<2:10:57, 4.92it/s] 20%|█▉ | 9365/48008 [31:54<2:08:46, 5.00it/s] 20%|█▉ | 9366/48008 [31:54<2:06:46, 5.08it/s] 20%|█▉ | 9367/48008 [31:54<2:05:18, 5.14it/s] 20%|█▉ | 9368/48008 [31:55<2:42:27, 3.96it/s] 20%|█▉ | 9369/48008 [31:55<2:30:40, 4.27it/s] 20%|█▉ | 9370/48008 [31:55<2:21:31, 4.55it/s] 20%|█▉ | 9371/48008 [31:55<2:16:09, 4.73it/s] 20%|█▉ | 9372/48008 [31:56<2:11:56, 4.88it/s] 20%|█▉ | 9373/48008 [31:56<2:46:55, 3.86it/s] 20%|█▉ | 9374/48008 [31:56<2:33:58, 4.18it/s] 20%|█▉ | 9375/48008 [31:56<2:24:14, 4.46it/s] 20%|█▉ | 9376/48008 [31:57<2:17:23, 4.69it/s] 20%|█▉ | 9377/48008 [31:57<2:35:48, 4.13it/s] 20%|█▉ | 9378/48008 [31:57<2:25:20, 4.43it/s] 20%|█▉ | 9379/48008 [31:57<2:18:34, 4.65it/s] 20%|█▉ | 9380/48008 [31:57<2:13:57, 4.81it/s] 20%|█▉ | 9381/48008 [31:58<2:11:10, 4.91it/s] 20%|█▉ | 9382/48008 [31:58<2:17:49, 4.67it/s] 20%|█▉ | 9383/48008 [31:58<2:13:59, 4.80it/s] 20%|█▉ | 9384/48008 [31:58<2:13:00, 4.84it/s] 20%|█▉ | 9385/48008 [31:59<2:13:51, 4.81it/s] 20%|█▉ | 9386/48008 [31:59<2:13:56, 4.81it/s] 20%|█▉ | 9387/48008 [31:59<2:10:20, 4.94it/s] 20%|█▉ | 9388/48008 [31:59<2:16:59, 4.70it/s] 20%|█▉ | 9389/48008 [31:59<2:12:35, 4.85it/s] 20%|█▉ | 9390/48008 [32:00<2:18:24, 4.65it/s] 20%|█▉ | 9391/48008 [32:00<2:13:08, 4.83it/s] 20%|█▉ | 9392/48008 [32:00<2:11:23, 4.90it/s] 20%|█▉ | 9393/48008 [32:00<2:10:50, 4.92it/s] 20%|█▉ | 9394/48008 [32:00<2:09:44, 4.96it/s] 20%|█▉ | 9395/48008 [32:01<2:08:58, 4.99it/s] 20%|█▉ | 9396/48008 [32:01<2:07:27, 5.05it/s] 20%|█▉ | 9397/48008 [32:01<2:09:54, 4.95it/s] 20%|█▉ | 9398/48008 [32:01<2:07:58, 5.03it/s] 20%|█▉ | 9399/48008 [32:01<2:07:24, 5.05it/s] 20%|█▉ | 9400/48008 [32:02<2:05:44, 5.12it/s] {'loss': 0.3004, 'grad_norm': 0.150390625, 'learning_rate': 0.0001608440259956674, 'epoch': 0.2} 20%|█▉ | 9400/48008 [32:02<2:05:44, 5.12it/s] 20%|█▉ | 9401/48008 [32:02<2:07:14, 5.06it/s] 20%|█▉ | 9402/48008 [32:02<2:05:10, 5.14it/s] 20%|█▉ | 9403/48008 [32:02<2:04:10, 5.18it/s] 20%|█▉ | 9404/48008 [32:02<2:05:47, 5.11it/s] 20%|█▉ | 9405/48008 [32:03<2:03:24, 5.21it/s] 20%|█▉ | 9406/48008 [32:03<2:03:19, 5.22it/s] 20%|█▉ | 9407/48008 [32:03<2:25:04, 4.43it/s] 20%|█▉ | 9408/48008 [32:03<2:20:12, 4.59it/s] 20%|█▉ | 9409/48008 [32:03<2:23:45, 4.47it/s] 20%|█▉ | 9410/48008 [32:04<2:26:24, 4.39it/s] 20%|█▉ | 9411/48008 [32:04<2:18:23, 4.65it/s] 20%|█▉ | 9412/48008 [32:04<2:13:28, 4.82it/s] 20%|█▉ | 9413/48008 [32:04<2:09:53, 4.95it/s] 20%|█▉ | 9414/48008 [32:04<2:10:32, 4.93it/s] 20%|█▉ | 9415/48008 [32:05<2:08:07, 5.02it/s] 20%|█▉ | 9416/48008 [32:05<2:07:48, 5.03it/s] 20%|█▉ | 9417/48008 [32:05<2:06:04, 5.10it/s] 20%|█▉ | 9418/48008 [32:05<2:04:50, 5.15it/s] 20%|█▉ | 9419/48008 [32:05<2:04:33, 5.16it/s] 20%|█▉ | 9420/48008 [32:06<2:04:29, 5.17it/s] 20%|█▉ | 9421/48008 [32:06<2:04:02, 5.18it/s] 20%|█▉ | 9422/48008 [32:06<2:02:37, 5.24it/s] 20%|█▉ | 9423/48008 [32:06<2:04:06, 5.18it/s] 20%|█▉ | 9424/48008 [32:06<2:03:41, 5.20it/s] 20%|█▉ | 9425/48008 [32:07<2:03:06, 5.22it/s] 20%|█▉ | 9426/48008 [32:07<2:02:40, 5.24it/s] 20%|█▉ | 9427/48008 [32:07<2:01:56, 5.27it/s] 20%|█▉ | 9428/48008 [32:07<2:02:38, 5.24it/s] 20%|█▉ | 9429/48008 [32:07<2:03:39, 5.20it/s] 20%|█▉ | 9430/48008 [32:08<2:05:33, 5.12it/s] 20%|█▉ | 9431/48008 [32:08<2:04:23, 5.17it/s] 20%|█▉ | 9432/48008 [32:08<2:05:48, 5.11it/s] 20%|█▉ | 9433/48008 [32:08<2:05:16, 5.13it/s] 20%|█▉ | 9434/48008 [32:08<2:04:37, 5.16it/s] 20%|█▉ | 9435/48008 [32:09<2:12:37, 4.85it/s] 20%|█▉ | 9436/48008 [32:09<2:08:48, 4.99it/s] 20%|█▉ | 9437/48008 [32:09<2:07:14, 5.05it/s] 20%|█▉ | 9438/48008 [32:09<2:05:07, 5.14it/s] 20%|█▉ | 9439/48008 [32:09<2:04:06, 5.18it/s] 20%|█▉ | 9440/48008 [32:09<2:03:24, 5.21it/s] 20%|█▉ | 9441/48008 [32:10<2:05:53, 5.11it/s] 20%|█▉ | 9442/48008 [32:10<2:04:15, 5.17it/s] 20%|█▉ | 9443/48008 [32:10<2:03:50, 5.19it/s] 20%|█▉ | 9444/48008 [32:10<2:03:26, 5.21it/s] 20%|█▉ | 9445/48008 [32:10<2:02:42, 5.24it/s] 20%|█▉ | 9446/48008 [32:11<2:03:37, 5.20it/s] 20%|█▉ | 9447/48008 [32:11<2:02:45, 5.24it/s] 20%|█▉ | 9448/48008 [32:11<2:02:18, 5.25it/s] 20%|█▉ | 9449/48008 [32:11<2:03:12, 5.22it/s] 20%|█▉ | 9450/48008 [32:11<2:00:58, 5.31it/s] {'loss': 0.3042, 'grad_norm': 0.5625, 'learning_rate': 0.0001606357273787702, 'epoch': 0.2} 20%|█▉ | 9450/48008 [32:11<2:00:58, 5.31it/s] 20%|█▉ | 9451/48008 [32:12<2:01:29, 5.29it/s] 20%|█▉ | 9452/48008 [32:12<2:00:58, 5.31it/s] 20%|█▉ | 9453/48008 [32:12<2:01:30, 5.29it/s] 20%|█▉ | 9454/48008 [32:12<2:01:30, 5.29it/s] 20%|█▉ | 9455/48008 [32:12<2:00:55, 5.31it/s] 20%|█▉ | 9456/48008 [32:13<2:00:21, 5.34it/s] 20%|█▉ | 9457/48008 [32:13<2:03:15, 5.21it/s] 20%|█▉ | 9458/48008 [32:13<2:40:37, 4.00it/s] 20%|█▉ | 9459/48008 [32:13<2:29:57, 4.28it/s] 20%|█▉ | 9460/48008 [32:13<2:21:25, 4.54it/s] 20%|█▉ | 9461/48008 [32:14<2:14:53, 4.76it/s] 20%|█▉ | 9462/48008 [32:14<2:10:22, 4.93it/s] 20%|█▉ | 9463/48008 [32:14<2:07:15, 5.05it/s] 20%|█▉ | 9464/48008 [32:14<2:07:50, 5.03it/s] 20%|█▉ | 9465/48008 [32:14<2:05:59, 5.10it/s] 20%|█▉ | 9466/48008 [32:15<2:07:33, 5.04it/s] 20%|█▉ | 9467/48008 [32:15<2:05:06, 5.13it/s] 20%|█▉ | 9468/48008 [32:15<2:03:40, 5.19it/s] 20%|█▉ | 9469/48008 [32:15<2:05:11, 5.13it/s] 20%|█▉ | 9470/48008 [32:15<2:04:18, 5.17it/s] 20%|█▉ | 9471/48008 [32:16<2:06:54, 5.06it/s] 20%|█▉ | 9472/48008 [32:16<2:07:51, 5.02it/s] 20%|█▉ | 9473/48008 [32:16<2:06:29, 5.08it/s] 20%|█▉ | 9474/48008 [32:16<2:05:26, 5.12it/s] 20%|█▉ | 9475/48008 [32:16<2:06:02, 5.10it/s] 20%|█▉ | 9476/48008 [32:17<2:06:12, 5.09it/s] 20%|█▉ | 9477/48008 [32:17<2:04:48, 5.15it/s] 20%|█▉ | 9478/48008 [32:17<2:07:01, 5.06it/s] 20%|█▉ | 9479/48008 [32:17<2:06:35, 5.07it/s] 20%|█▉ | 9480/48008 [32:17<2:04:29, 5.16it/s] 20%|█▉ | 9481/48008 [32:18<2:06:01, 5.10it/s] 20%|█▉ | 9482/48008 [32:18<2:05:01, 5.14it/s] 20%|█▉ | 9483/48008 [32:18<2:07:54, 5.02it/s] 20%|█▉ | 9484/48008 [32:18<2:05:24, 5.12it/s] 20%|█▉ | 9485/48008 [32:18<2:04:33, 5.15it/s] 20%|█▉ | 9486/48008 [32:19<2:03:44, 5.19it/s] 20%|█▉ | 9487/48008 [32:19<2:03:41, 5.19it/s] 20%|█▉ | 9488/48008 [32:19<2:02:22, 5.25it/s] 20%|█▉ | 9489/48008 [32:19<2:01:33, 5.28it/s] 20%|█▉ | 9490/48008 [32:19<2:01:52, 5.27it/s] 20%|█▉ | 9491/48008 [32:19<2:02:13, 5.25it/s] 20%|█▉ | 9492/48008 [32:20<2:02:22, 5.25it/s] 20%|█▉ | 9493/48008 [32:20<2:01:40, 5.28it/s] 20%|█▉ | 9494/48008 [32:20<2:03:21, 5.20it/s] 20%|█▉ | 9495/48008 [32:20<2:02:04, 5.26it/s] 20%|█▉ | 9496/48008 [32:20<2:02:54, 5.22it/s] 20%|█▉ | 9497/48008 [32:21<2:02:58, 5.22it/s] 20%|█▉ | 9498/48008 [32:21<2:03:17, 5.21it/s] 20%|█▉ | 9499/48008 [32:21<2:03:27, 5.20it/s] 20%|█▉ | 9500/48008 [32:21<2:05:20, 5.12it/s] {'loss': 0.305, 'grad_norm': 0.1669921875, 'learning_rate': 0.00016042742876187303, 'epoch': 0.2} 20%|█▉ | 9500/48008 [32:21<2:05:20, 5.12it/s] 20%|█▉ | 9501/48008 [32:21<2:07:36, 5.03it/s] 20%|█▉ | 9502/48008 [32:22<2:05:38, 5.11it/s] 20%|█▉ | 9503/48008 [32:22<2:07:01, 5.05it/s] 20%|█▉ | 9504/48008 [32:22<2:05:50, 5.10it/s] 20%|█▉ | 9505/48008 [32:22<2:06:58, 5.05it/s] 20%|█▉ | 9506/48008 [32:22<2:05:40, 5.11it/s] 20%|█▉ | 9507/48008 [32:23<2:07:12, 5.04it/s] 20%|█▉ | 9508/48008 [32:23<2:09:16, 4.96it/s] 20%|█▉ | 9509/48008 [32:23<2:07:30, 5.03it/s] 20%|█▉ | 9510/48008 [32:23<2:09:08, 4.97it/s] 20%|█▉ | 9511/48008 [32:23<2:06:17, 5.08it/s] 20%|█▉ | 9512/48008 [32:24<2:07:21, 5.04it/s] 20%|█▉ | 9513/48008 [32:24<2:06:01, 5.09it/s] 20%|█▉ | 9514/48008 [32:24<2:04:49, 5.14it/s] 20%|█▉ | 9515/48008 [32:24<2:04:32, 5.15it/s] 20%|█▉ | 9516/48008 [32:24<2:04:55, 5.14it/s] 20%|█▉ | 9517/48008 [32:25<2:07:54, 5.02it/s] 20%|█▉ | 9518/48008 [32:25<2:06:00, 5.09it/s] 20%|█▉ | 9519/48008 [32:25<2:05:10, 5.13it/s] 20%|█▉ | 9520/48008 [32:25<2:03:19, 5.20it/s] 20%|█▉ | 9521/48008 [32:25<2:04:19, 5.16it/s] 20%|█▉ | 9522/48008 [32:26<2:13:29, 4.81it/s] 20%|█▉ | 9523/48008 [32:26<2:08:41, 4.98it/s] 20%|█▉ | 9524/48008 [32:26<2:07:00, 5.05it/s] 20%|█▉ | 9525/48008 [32:26<2:05:31, 5.11it/s] 20%|█▉ | 9526/48008 [32:26<2:05:42, 5.10it/s] 20%|█▉ | 9527/48008 [32:27<2:04:36, 5.15it/s] 20%|█▉ | 9528/48008 [32:27<2:03:07, 5.21it/s] 20%|█▉ | 9529/48008 [32:27<2:01:24, 5.28it/s] 20%|█▉ | 9530/48008 [32:27<2:03:13, 5.20it/s] 20%|█▉ | 9531/48008 [32:27<2:03:13, 5.20it/s] 20%|█▉ | 9532/48008 [32:28<2:05:45, 5.10it/s] 20%|█▉ | 9533/48008 [32:28<2:08:36, 4.99it/s] 20%|█▉ | 9534/48008 [32:28<2:08:50, 4.98it/s] 20%|█▉ | 9535/48008 [32:28<2:06:41, 5.06it/s] 20%|█▉ | 9536/48008 [32:28<2:07:22, 5.03it/s] 20%|█▉ | 9537/48008 [32:29<2:06:02, 5.09it/s] 20%|█▉ | 9538/48008 [32:29<2:04:49, 5.14it/s] 20%|█▉ | 9539/48008 [32:29<2:04:02, 5.17it/s] 20%|█▉ | 9540/48008 [32:29<2:03:30, 5.19it/s] 20%|█▉ | 9541/48008 [32:29<2:11:58, 4.86it/s] 20%|█▉ | 9542/48008 [32:30<2:08:00, 5.01it/s] 20%|█▉ | 9543/48008 [32:30<2:08:52, 4.97it/s] 20%|█▉ | 9544/48008 [32:30<2:09:46, 4.94it/s] 20%|█▉ | 9545/48008 [32:30<2:07:41, 5.02it/s] 20%|█▉ | 9546/48008 [32:30<2:08:02, 5.01it/s] 20%|█▉ | 9547/48008 [32:30<2:05:49, 5.09it/s] 20%|█▉ | 9548/48008 [32:31<2:05:19, 5.12it/s] 20%|█▉ | 9549/48008 [32:31<2:03:16, 5.20it/s] 20%|█▉ | 9550/48008 [32:31<2:03:03, 5.21it/s] {'loss': 0.3015, 'grad_norm': 0.1474609375, 'learning_rate': 0.00016021913014497584, 'epoch': 0.2} 20%|█▉ | 9550/48008 [32:31<2:03:03, 5.21it/s] 20%|█▉ | 9551/48008 [32:31<2:04:04, 5.17it/s] 20%|█▉ | 9552/48008 [32:31<2:02:53, 5.22it/s] 20%|█▉ | 9553/48008 [32:32<2:03:10, 5.20it/s] 20%|█▉ | 9554/48008 [32:32<2:02:49, 5.22it/s] 20%|█▉ | 9555/48008 [32:32<2:11:47, 4.86it/s] 20%|█▉ | 9556/48008 [32:32<2:09:01, 4.97it/s] 20%|█▉ | 9557/48008 [32:32<2:07:01, 5.04it/s] 20%|█▉ | 9558/48008 [32:33<2:04:26, 5.15it/s] 20%|█▉ | 9559/48008 [32:33<2:03:51, 5.17it/s] 20%|█▉ | 9560/48008 [32:33<2:04:26, 5.15it/s] 20%|█▉ | 9561/48008 [32:33<2:03:25, 5.19it/s] 20%|█▉ | 9562/48008 [32:33<2:02:53, 5.21it/s] 20%|█▉ | 9563/48008 [32:34<2:03:05, 5.21it/s] 20%|█▉ | 9564/48008 [32:34<2:24:58, 4.42it/s] 20%|█▉ | 9565/48008 [32:34<2:17:03, 4.67it/s] 20%|█▉ | 9566/48008 [32:34<2:12:32, 4.83it/s] 20%|█▉ | 9567/48008 [32:34<2:09:30, 4.95it/s] 20%|█▉ | 9568/48008 [32:35<2:06:28, 5.07it/s] 20%|█▉ | 9569/48008 [32:35<2:03:50, 5.17it/s] 20%|█▉ | 9570/48008 [32:35<2:03:34, 5.18it/s] 20%|█▉ | 9571/48008 [32:35<2:03:18, 5.19it/s] 20%|█▉ | 9572/48008 [32:35<2:11:40, 4.86it/s] 20%|█▉ | 9573/48008 [32:36<2:07:28, 5.03it/s] 20%|█▉ | 9574/48008 [32:36<2:05:47, 5.09it/s] 20%|█▉ | 9575/48008 [32:36<2:05:04, 5.12it/s] 20%|█▉ | 9576/48008 [32:36<2:05:13, 5.11it/s] 20%|█▉ | 9577/48008 [32:36<2:03:59, 5.17it/s] 20%|█▉ | 9578/48008 [32:37<2:03:17, 5.20it/s] 20%|█▉ | 9579/48008 [32:37<2:02:33, 5.23it/s] 20%|█▉ | 9580/48008 [32:37<2:02:43, 5.22it/s] 20%|█▉ | 9581/48008 [32:37<2:02:35, 5.22it/s] 20%|█▉ | 9582/48008 [32:37<2:02:42, 5.22it/s] 20%|█▉ | 9583/48008 [32:38<2:02:18, 5.24it/s] 20%|█▉ | 9584/48008 [32:38<2:02:10, 5.24it/s] 20%|█▉ | 9585/48008 [32:38<2:03:29, 5.19it/s] 20%|█▉ | 9586/48008 [32:38<2:02:14, 5.24it/s] 20%|█▉ | 9587/48008 [32:38<2:02:15, 5.24it/s] 20%|█▉ | 9588/48008 [32:39<2:04:56, 5.13it/s] 20%|█▉ | 9589/48008 [32:39<2:03:23, 5.19it/s] 20%|█▉ | 9590/48008 [32:39<2:02:47, 5.21it/s] 20%|█▉ | 9591/48008 [32:39<2:03:26, 5.19it/s] 20%|█▉ | 9592/48008 [32:39<2:03:20, 5.19it/s] 20%|█▉ | 9593/48008 [32:39<2:02:57, 5.21it/s] 20%|█▉ | 9594/48008 [32:40<2:02:45, 5.22it/s] 20%|█▉ | 9595/48008 [32:40<2:02:35, 5.22it/s] 20%|█▉ | 9596/48008 [32:40<2:04:26, 5.14it/s] 20%|█▉ | 9597/48008 [32:40<2:05:57, 5.08it/s] 20%|█▉ | 9598/48008 [32:40<2:07:58, 5.00it/s] 20%|█▉ | 9599/48008 [32:41<2:05:53, 5.08it/s] 20%|█▉ | 9600/48008 [32:41<2:06:05, 5.08it/s] {'loss': 0.3029, 'grad_norm': 0.1435546875, 'learning_rate': 0.00016001083152807866, 'epoch': 0.2} 20%|█▉ | 9600/48008 [32:41<2:06:05, 5.08it/s] 20%|█▉ | 9601/48008 [32:41<2:06:36, 5.06it/s] 20%|██ | 9602/48008 [32:41<2:06:00, 5.08it/s] 20%|██ | 9603/48008 [32:41<2:08:33, 4.98it/s] 20%|██ | 9604/48008 [32:42<2:08:30, 4.98it/s] 20%|██ | 9605/48008 [32:42<2:08:23, 4.99it/s] 20%|██ | 9606/48008 [32:42<2:06:46, 5.05it/s] 20%|██ | 9607/48008 [32:42<2:05:29, 5.10it/s] 20%|██ | 9608/48008 [32:42<2:06:30, 5.06it/s] 20%|██ | 9609/48008 [32:43<2:04:42, 5.13it/s] 20%|██ | 9610/48008 [32:43<2:41:31, 3.96it/s] 20%|██ | 9611/48008 [32:43<2:28:52, 4.30it/s] 20%|██ | 9612/48008 [32:43<2:20:54, 4.54it/s] 20%|██ | 9613/48008 [32:44<2:23:37, 4.46it/s] 20%|██ | 9614/48008 [32:44<2:15:05, 4.74it/s] 20%|██ | 9615/48008 [32:44<2:11:54, 4.85it/s] 20%|██ | 9616/48008 [32:44<2:10:31, 4.90it/s] 20%|██ | 9617/48008 [32:44<2:07:42, 5.01it/s] 20%|██ | 9618/48008 [32:45<2:04:59, 5.12it/s] 20%|██ | 9619/48008 [32:45<2:04:38, 5.13it/s] 20%|██ | 9620/48008 [32:45<2:04:51, 5.12it/s] 20%|██ | 9621/48008 [32:45<2:12:46, 4.82it/s] 20%|██ | 9622/48008 [32:45<2:08:38, 4.97it/s] 20%|██ | 9623/48008 [32:46<2:06:27, 5.06it/s] 20%|██ | 9624/48008 [32:46<2:04:48, 5.13it/s] 20%|██ | 9625/48008 [32:46<2:04:04, 5.16it/s] 20%|██ | 9626/48008 [32:46<2:12:26, 4.83it/s] 20%|██ | 9627/48008 [32:46<2:11:25, 4.87it/s] 20%|██ | 9628/48008 [32:47<2:08:23, 4.98it/s] 20%|██ | 9629/48008 [32:47<2:08:02, 5.00it/s] 20%|██ | 9630/48008 [32:47<2:08:30, 4.98it/s] 20%|██ | 9631/48008 [32:47<2:06:16, 5.07it/s] 20%|██ | 9632/48008 [32:47<2:04:36, 5.13it/s] 20%|██ | 9633/48008 [32:48<2:05:47, 5.08it/s] 20%|██ | 9634/48008 [32:48<2:04:20, 5.14it/s] 20%|██ | 9635/48008 [32:48<2:04:35, 5.13it/s] 20%|██ | 9636/48008 [32:48<2:05:11, 5.11it/s] 20%|██ | 9637/48008 [32:48<2:03:54, 5.16it/s] 20%|██ | 9638/48008 [32:49<2:04:20, 5.14it/s] 20%|██ | 9639/48008 [32:49<2:03:36, 5.17it/s] 20%|██ | 9640/48008 [32:49<2:12:21, 4.83it/s] 20%|██ | 9641/48008 [32:49<2:09:13, 4.95it/s] 20%|██ | 9642/48008 [32:49<2:09:08, 4.95it/s] 20%|██ | 9643/48008 [32:50<2:06:48, 5.04it/s] 20%|██ | 9644/48008 [32:50<2:06:47, 5.04it/s] 20%|██ | 9645/48008 [32:50<2:05:22, 5.10it/s] 20%|██ | 9646/48008 [32:50<2:04:09, 5.15it/s] 20%|██ | 9647/48008 [32:50<2:12:13, 4.84it/s] 20%|██ | 9648/48008 [32:51<2:09:28, 4.94it/s] 20%|██ | 9649/48008 [32:51<2:07:26, 5.02it/s] 20%|██ | 9650/48008 [32:51<2:06:44, 5.04it/s] {'loss': 0.3001, 'grad_norm': 0.1279296875, 'learning_rate': 0.00015980253291118148, 'epoch': 0.2} 20%|██ | 9650/48008 [32:51<2:06:44, 5.04it/s] 20%|██ | 9651/48008 [32:51<2:06:26, 5.06it/s] 20%|██ | 9652/48008 [32:51<2:06:37, 5.05it/s] 20%|██ | 9653/48008 [32:52<2:07:47, 5.00it/s] 20%|██ | 9654/48008 [32:52<2:05:35, 5.09it/s] 20%|██ | 9655/48008 [32:52<2:26:29, 4.36it/s] 20%|██ | 9656/48008 [32:52<2:18:58, 4.60it/s] 20%|██ | 9657/48008 [32:52<2:15:10, 4.73it/s] 20%|██ | 9658/48008 [32:53<2:11:18, 4.87it/s] 20%|██ | 9659/48008 [32:53<2:11:56, 4.84it/s] 20%|██ | 9660/48008 [32:53<2:08:50, 4.96it/s] 20%|██ | 9661/48008 [32:53<2:06:55, 5.04it/s] 20%|██ | 9662/48008 [32:53<2:05:55, 5.08it/s] 20%|██ | 9663/48008 [32:54<2:06:02, 5.07it/s] 20%|██ | 9664/48008 [32:54<2:04:12, 5.15it/s] 20%|██ | 9665/48008 [32:54<2:04:04, 5.15it/s] 20%|██ | 9666/48008 [32:54<2:25:27, 4.39it/s] 20%|██ | 9667/48008 [32:54<2:18:22, 4.62it/s] 20%|██ | 9668/48008 [32:55<2:11:44, 4.85it/s] 20%|██ | 9669/48008 [32:55<2:07:47, 5.00it/s] 20%|██ | 9670/48008 [32:55<2:05:43, 5.08it/s] 20%|██ | 9671/48008 [32:55<2:04:38, 5.13it/s] 20%|██ | 9672/48008 [32:55<2:03:18, 5.18it/s] 20%|██ | 9673/48008 [32:56<2:05:10, 5.10it/s] 20%|██ | 9674/48008 [32:56<2:07:41, 5.00it/s] 20%|██ | 9675/48008 [32:56<2:08:06, 4.99it/s] 20%|██ | 9676/48008 [32:56<2:05:19, 5.10it/s] 20%|██ | 9677/48008 [32:56<2:04:31, 5.13it/s] 20%|██ | 9678/48008 [32:57<2:03:48, 5.16it/s] 20%|██ | 9679/48008 [32:57<2:02:58, 5.19it/s] 20%|██ | 9680/48008 [32:57<2:02:59, 5.19it/s] 20%|██ | 9681/48008 [32:57<2:02:07, 5.23it/s] 20%|██ | 9682/48008 [32:57<2:01:02, 5.28it/s] 20%|██ | 9683/48008 [32:58<2:00:49, 5.29it/s] 20%|██ | 9684/48008 [32:58<2:00:55, 5.28it/s] 20%|██ | 9685/48008 [32:58<1:59:50, 5.33it/s] 20%|██ | 9686/48008 [32:58<2:03:13, 5.18it/s] 20%|██ | 9687/48008 [32:58<2:01:52, 5.24it/s] 20%|██ | 9688/48008 [32:58<2:01:39, 5.25it/s] 20%|██ | 9689/48008 [32:59<2:02:07, 5.23it/s] 20%|██ | 9690/48008 [32:59<2:02:47, 5.20it/s] 20%|██ | 9691/48008 [32:59<2:03:23, 5.18it/s] 20%|██ | 9692/48008 [32:59<2:02:56, 5.19it/s] 20%|██ | 9693/48008 [32:59<2:03:28, 5.17it/s] 20%|██ | 9694/48008 [33:00<2:03:05, 5.19it/s] 20%|██ | 9695/48008 [33:00<2:03:08, 5.19it/s] 20%|██ | 9696/48008 [33:00<2:11:58, 4.84it/s] 20%|██ | 9697/48008 [33:00<2:10:53, 4.88it/s] 20%|██ | 9698/48008 [33:00<2:10:15, 4.90it/s] 20%|██ | 9699/48008 [33:01<4:36:16, 2.31it/s] 20%|██ | 9700/48008 [33:02<3:50:47, 2.77it/s] {'loss': 0.3032, 'grad_norm': 0.1396484375, 'learning_rate': 0.0001595942342942843, 'epoch': 0.2} 20%|██ | 9700/48008 [33:02<3:50:47, 2.77it/s] 20%|██ | 9701/48008 [33:02<3:17:11, 3.24it/s] 20%|██ | 9702/48008 [33:02<2:54:44, 3.65it/s] 20%|██ | 9703/48008 [33:02<3:00:55, 3.53it/s] 20%|██ | 9704/48008 [33:03<2:43:21, 3.91it/s] 20%|██ | 9705/48008 [33:03<4:53:41, 2.17it/s] 20%|██ | 9706/48008 [33:04<4:03:57, 2.62it/s] 20%|██ | 9707/48008 [33:04<3:27:13, 3.08it/s] 20%|██ | 9708/48008 [33:04<3:02:32, 3.50it/s] 20%|██ | 9709/48008 [33:04<2:44:48, 3.87it/s] 20%|██ | 9710/48008 [33:04<2:32:08, 4.20it/s] 20%|██ | 9711/48008 [33:05<4:35:12, 2.32it/s] 20%|██ | 9712/48008 [33:06<3:48:44, 2.79it/s] 20%|██ | 9713/48008 [33:06<3:15:39, 3.26it/s] 20%|██ | 9714/48008 [33:06<2:52:19, 3.70it/s] 20%|██ | 9715/48008 [33:06<2:37:12, 4.06it/s] 20%|██ | 9716/48008 [33:06<2:29:04, 4.28it/s] 20%|██ | 9717/48008 [33:07<5:35:18, 1.90it/s] 20%|██ | 9718/48008 [33:08<4:34:12, 2.33it/s] 20%|██ | 9719/48008 [33:08<3:48:17, 2.80it/s] 20%|██ | 9720/48008 [33:08<3:19:15, 3.20it/s] 20%|██ | 9721/48008 [33:08<2:56:01, 3.62it/s] 20%|██ | 9722/48008 [33:09<2:49:06, 3.77it/s] 20%|██ | 9723/48008 [33:09<2:34:18, 4.14it/s] 20%|██ | 9724/48008 [33:09<2:26:35, 4.35it/s] 20%|██ | 9725/48008 [33:09<2:19:08, 4.59it/s] 20%|██ | 9726/48008 [33:09<2:13:33, 4.78it/s] 20%|██ | 9727/48008 [33:09<2:10:20, 4.90it/s] 20%|██ | 9728/48008 [33:10<2:06:32, 5.04it/s] 20%|██ | 9729/48008 [33:10<4:06:33, 2.59it/s] 20%|██ | 9730/48008 [33:11<3:29:01, 3.05it/s] 20%|██ | 9731/48008 [33:11<3:01:43, 3.51it/s] 20%|██ | 9732/48008 [33:11<2:43:43, 3.90it/s] 20%|██ | 9733/48008 [33:11<2:30:50, 4.23it/s] 20%|██ | 9734/48008 [33:11<2:21:14, 4.52it/s] 20%|██ | 9735/48008 [33:12<2:14:41, 4.74it/s] 20%|██ | 9736/48008 [33:12<2:10:41, 4.88it/s] 20%|██ | 9737/48008 [33:12<2:08:08, 4.98it/s] 20%|██ | 9738/48008 [33:12<2:08:35, 4.96it/s] 20%|██ | 9739/48008 [33:12<2:07:10, 5.02it/s] 20%|██ | 9740/48008 [33:13<2:05:42, 5.07it/s] 20%|██ | 9741/48008 [33:13<2:04:28, 5.12it/s] 20%|██ | 9742/48008 [33:13<2:03:27, 5.17it/s] 20%|██ | 9743/48008 [33:13<2:01:57, 5.23it/s] 20%|██ | 9744/48008 [33:13<2:01:36, 5.24it/s] 20%|██ | 9745/48008 [33:14<2:23:47, 4.43it/s] 20%|██ | 9746/48008 [33:14<2:16:25, 4.67it/s] 20%|██ | 9747/48008 [33:14<2:11:14, 4.86it/s] 20%|██ | 9748/48008 [33:14<2:08:08, 4.98it/s] 20%|██ | 9749/48008 [33:14<2:06:21, 5.05it/s] 20%|██ | 9750/48008 [33:15<2:06:06, 5.06it/s] 20%|██ | 9750/48008 [33:15<2:06:06, 5.06it/s]{'loss': 0.3021, 'grad_norm': 0.251953125, 'learning_rate': 0.0001593859356773871, 'epoch': 0.2} 20%|██ | 9751/48008 [33:15<2:04:44, 5.11it/s] 20%|██ | 9752/48008 [33:15<2:04:35, 5.12it/s] 20%|██ | 9753/48008 [33:15<2:03:05, 5.18it/s] 20%|██ | 9754/48008 [33:15<2:02:18, 5.21it/s] 20%|██ | 9755/48008 [33:16<2:05:44, 5.07it/s] 20%|██ | 9756/48008 [33:16<2:08:06, 4.98it/s] 20%|██ | 9757/48008 [33:16<2:08:15, 4.97it/s] 20%|██ | 9758/48008 [33:16<2:06:33, 5.04it/s] 20%|██ | 9759/48008 [33:16<2:05:00, 5.10it/s] 20%|██ | 9760/48008 [33:17<2:26:00, 4.37it/s] 20%|██ | 9761/48008 [33:17<2:18:06, 4.62it/s] 20%|██ | 9762/48008 [33:17<2:12:59, 4.79it/s] 20%|██ | 9763/48008 [33:17<2:09:55, 4.91it/s] 20%|██ | 9764/48008 [33:18<2:44:24, 3.88it/s] 20%|██ | 9765/48008 [33:18<2:40:56, 3.96it/s] 20%|██ | 9766/48008 [33:18<2:30:19, 4.24it/s] 20%|██ | 9767/48008 [33:18<2:22:42, 4.47it/s] 20%|██ | 9768/48008 [33:18<2:15:41, 4.70it/s] 20%|██ | 9769/48008 [33:19<2:20:04, 4.55it/s] 20%|██ | 9770/48008 [33:19<2:14:09, 4.75it/s] 20%|██ | 9771/48008 [33:19<2:10:15, 4.89it/s] 20%|██ | 9772/48008 [33:19<2:06:11, 5.05it/s] 20%|██ | 9773/48008 [33:19<2:05:01, 5.10it/s] 20%|██ | 9774/48008 [33:20<2:26:07, 4.36it/s] 20%|██ | 9775/48008 [33:20<2:17:53, 4.62it/s] 20%|██ | 9776/48008 [33:20<2:11:51, 4.83it/s] 20%|██ | 9777/48008 [33:20<2:08:53, 4.94it/s] 20%|██ | 9778/48008 [33:20<2:06:42, 5.03it/s] 20%|██ | 9779/48008 [33:21<2:03:58, 5.14it/s] 20%|██ | 9780/48008 [33:21<2:03:08, 5.17it/s] 20%|██ | 9781/48008 [33:21<2:03:13, 5.17it/s] 20%|██ | 9782/48008 [33:21<2:03:53, 5.14it/s] 20%|██ | 9783/48008 [33:21<2:02:42, 5.19it/s] 20%|██ | 9784/48008 [33:22<2:04:44, 5.11it/s] 20%|██ | 9785/48008 [33:22<2:02:53, 5.18it/s] 20%|██ | 9786/48008 [33:22<2:01:20, 5.25it/s] 20%|██ | 9787/48008 [33:22<2:01:38, 5.24it/s] 20%|██ | 9788/48008 [33:22<2:00:46, 5.27it/s] 20%|██ | 9789/48008 [33:23<2:00:46, 5.27it/s] 20%|██ | 9790/48008 [33:23<2:01:54, 5.23it/s] 20%|██ | 9791/48008 [33:23<2:01:25, 5.25it/s] 20%|██ | 9792/48008 [33:23<2:00:24, 5.29it/s] 20%|██ | 9793/48008 [33:23<2:00:16, 5.30it/s] 20%|██ | 9794/48008 [33:24<2:00:10, 5.30it/s] 20%|██ | 9795/48008 [33:24<2:03:05, 5.17it/s] 20%|██ | 9796/48008 [33:24<2:04:42, 5.11it/s] 20%|██ | 9797/48008 [33:24<2:02:51, 5.18it/s] 20%|██ | 9798/48008 [33:24<2:02:07, 5.21it/s] 20%|██ | 9799/48008 [33:25<2:02:57, 5.18it/s] 20%|██ | 9800/48008 [33:25<2:02:35, 5.19it/s] {'loss': 0.305, 'grad_norm': 0.13671875, 'learning_rate': 0.00015917763706048993, 'epoch': 0.2} 20%|██ | 9800/48008 [33:25<2:02:35, 5.19it/s] 20%|██ | 9801/48008 [33:25<2:03:13, 5.17it/s] 20%|██ | 9802/48008 [33:25<2:01:59, 5.22it/s] 20%|██ | 9803/48008 [33:25<2:03:13, 5.17it/s] 20%|██ | 9804/48008 [33:25<2:04:09, 5.13it/s] 20%|██ | 9805/48008 [33:26<2:02:08, 5.21it/s] 20%|██ | 9806/48008 [33:26<2:23:56, 4.42it/s] 20%|██ | 9807/48008 [33:26<2:16:50, 4.65it/s] 20%|██ | 9808/48008 [33:26<2:12:10, 4.82it/s] 20%|██ | 9809/48008 [33:27<2:10:25, 4.88it/s] 20%|██ | 9810/48008 [33:27<2:08:16, 4.96it/s] 20%|██ | 9811/48008 [33:27<2:08:08, 4.97it/s] 20%|██ | 9812/48008 [33:27<2:06:21, 5.04it/s] 20%|██ | 9813/48008 [33:27<2:04:49, 5.10it/s] 20%|██ | 9814/48008 [33:28<2:02:48, 5.18it/s] 20%|██ | 9815/48008 [33:28<2:01:39, 5.23it/s] 20%|██ | 9816/48008 [33:28<2:01:50, 5.22it/s] 20%|██ | 9817/48008 [33:28<2:01:32, 5.24it/s] 20%|██ | 9818/48008 [33:28<2:01:05, 5.26it/s] 20%|██ | 9819/48008 [33:28<2:00:18, 5.29it/s] 20%|██ | 9820/48008 [33:29<2:00:02, 5.30it/s] 20%|██ | 9821/48008 [33:29<1:58:53, 5.35it/s] 20%|██ | 9822/48008 [33:29<2:02:02, 5.21it/s] 20%|██ | 9823/48008 [33:29<2:03:58, 5.13it/s] 20%|██ | 9824/48008 [33:29<2:03:18, 5.16it/s] 20%|██ | 9825/48008 [33:30<2:02:53, 5.18it/s] 20%|██ | 9826/48008 [33:30<2:02:21, 5.20it/s] 20%|██ | 9827/48008 [33:30<2:02:12, 5.21it/s] 20%|██ | 9828/48008 [33:30<2:01:29, 5.24it/s] 20%|██ | 9829/48008 [33:30<2:00:31, 5.28it/s] 20%|██ | 9830/48008 [33:31<2:04:21, 5.12it/s] 20%|██ | 9831/48008 [33:31<2:02:28, 5.20it/s] 20%|██ | 9832/48008 [33:31<2:04:27, 5.11it/s] 20%|██ | 9833/48008 [33:31<2:02:35, 5.19it/s] 20%|██ | 9834/48008 [33:31<2:03:59, 5.13it/s] 20%|██ | 9835/48008 [33:32<2:03:16, 5.16it/s] 20%|██ | 9836/48008 [33:32<2:06:29, 5.03it/s] 20%|██ | 9837/48008 [33:32<2:06:11, 5.04it/s] 20%|██ | 9838/48008 [33:32<2:04:29, 5.11it/s] 20%|██ | 9839/48008 [33:32<2:03:26, 5.15it/s] 20%|██ | 9840/48008 [33:33<2:12:09, 4.81it/s] 20%|██ | 9841/48008 [33:33<2:12:32, 4.80it/s] 21%|██ | 9842/48008 [33:33<2:12:41, 4.79it/s] 21%|██ | 9843/48008 [33:33<2:09:30, 4.91it/s] 21%|██ | 9844/48008 [33:33<2:06:47, 5.02it/s] 21%|██ | 9845/48008 [33:34<2:05:23, 5.07it/s] 21%|██ | 9846/48008 [33:34<2:04:25, 5.11it/s] 21%|██ | 9847/48008 [33:34<2:02:59, 5.17it/s] 21%|██ | 9848/48008 [33:34<2:11:03, 4.85it/s] 21%|██ | 9849/48008 [33:34<2:09:01, 4.93it/s] 21%|██ | 9850/48008 [33:35<2:09:09, 4.92it/s] {'loss': 0.3039, 'grad_norm': 0.390625, 'learning_rate': 0.00015896933844359275, 'epoch': 0.21} 21%|██ | 9850/48008 [33:35<2:09:09, 4.92it/s] 21%|██ | 9851/48008 [33:35<2:09:06, 4.93it/s] 21%|██ | 9852/48008 [33:35<2:07:07, 5.00it/s] 21%|██ | 9853/48008 [33:35<2:05:19, 5.07it/s] 21%|██ | 9854/48008 [33:35<2:03:16, 5.16it/s] 21%|██ | 9855/48008 [33:36<2:02:55, 5.17it/s] 21%|██ | 9856/48008 [33:36<2:04:01, 5.13it/s] 21%|██ | 9857/48008 [33:36<2:11:56, 4.82it/s] 21%|██ | 9858/48008 [33:36<2:09:51, 4.90it/s] 21%|██ | 9859/48008 [33:36<2:07:25, 4.99it/s] 21%|██ | 9860/48008 [33:37<2:05:20, 5.07it/s] 21%|██ | 9861/48008 [33:37<2:06:13, 5.04it/s] 21%|██ | 9862/48008 [33:37<2:04:04, 5.12it/s] 21%|██ | 9863/48008 [33:37<2:07:06, 5.00it/s] 21%|██ | 9864/48008 [33:37<2:08:02, 4.97it/s] 21%|██ | 9865/48008 [33:38<2:14:46, 4.72it/s] 21%|██ | 9866/48008 [33:38<2:10:33, 4.87it/s] 21%|██ | 9867/48008 [33:38<2:07:25, 4.99it/s] 21%|██ | 9868/48008 [33:38<2:05:13, 5.08it/s] 21%|██ | 9869/48008 [33:38<2:05:34, 5.06it/s] 21%|██ | 9870/48008 [33:39<2:04:24, 5.11it/s] 21%|██ | 9871/48008 [33:39<2:12:15, 4.81it/s] 21%|██ | 9872/48008 [33:39<2:09:59, 4.89it/s] 21%|██ | 9873/48008 [33:39<2:07:25, 4.99it/s] 21%|██ | 9874/48008 [33:39<2:06:55, 5.01it/s] 21%|██ | 9875/48008 [33:40<2:04:11, 5.12it/s] 21%|██ | 9876/48008 [33:40<2:02:59, 5.17it/s] 21%|██ | 9877/48008 [33:40<2:04:33, 5.10it/s] 21%|██ | 9878/48008 [33:40<2:02:46, 5.18it/s] 21%|██ | 9879/48008 [33:40<2:02:41, 5.18it/s] 21%|██ | 9880/48008 [33:41<2:03:21, 5.15it/s] 21%|██ | 9881/48008 [33:41<2:03:20, 5.15it/s] 21%|██ | 9882/48008 [33:41<2:04:49, 5.09it/s] 21%|██ | 9883/48008 [33:41<2:40:36, 3.96it/s] 21%|██ | 9884/48008 [33:42<2:30:54, 4.21it/s] 21%|██ | 9885/48008 [33:42<2:43:44, 3.88it/s] 21%|██ | 9886/48008 [33:42<2:31:53, 4.18it/s] 21%|██ | 9887/48008 [33:42<2:22:04, 4.47it/s] 21%|██ | 9888/48008 [33:42<2:15:54, 4.67it/s] 21%|██ | 9889/48008 [33:43<2:10:25, 4.87it/s] 21%|██ | 9890/48008 [33:43<2:09:33, 4.90it/s] 21%|██ | 9891/48008 [33:43<2:08:12, 4.96it/s] 21%|██ | 9892/48008 [33:43<2:06:14, 5.03it/s] 21%|██ | 9893/48008 [33:43<2:05:05, 5.08it/s] 21%|██ | 9894/48008 [33:44<2:05:04, 5.08it/s] 21%|██ | 9895/48008 [33:44<2:12:32, 4.79it/s] 21%|██ | 9896/48008 [33:44<2:17:47, 4.61it/s] 21%|██ | 9897/48008 [33:44<2:11:52, 4.82it/s] 21%|██ | 9898/48008 [33:44<2:07:50, 4.97it/s] 21%|██ | 9899/48008 [33:45<2:06:15, 5.03it/s] 21%|██ | 9900/48008 [33:45<2:04:33, 5.10it/s] {'loss': 0.3009, 'grad_norm': 0.86328125, 'learning_rate': 0.00015876103982669556, 'epoch': 0.21} 21%|██ | 9900/48008 [33:45<2:04:33, 5.10it/s] 21%|██ | 9901/48008 [33:45<2:03:37, 5.14it/s] 21%|██ | 9902/48008 [33:45<2:06:40, 5.01it/s] 21%|██ | 9903/48008 [33:45<2:04:12, 5.11it/s] 21%|██ | 9904/48008 [33:46<2:04:22, 5.11it/s] 21%|██ | 9905/48008 [33:46<2:05:53, 5.04it/s] 21%|██ | 9906/48008 [33:46<2:05:23, 5.06it/s] 21%|██ | 9907/48008 [33:46<2:03:18, 5.15it/s] 21%|██ | 9908/48008 [33:46<2:04:11, 5.11it/s] 21%|██ | 9909/48008 [33:47<2:02:18, 5.19it/s] 21%|██ | 9910/48008 [33:47<2:02:03, 5.20it/s] 21%|██ | 9911/48008 [33:47<2:03:40, 5.13it/s] 21%|██ | 9912/48008 [33:47<2:02:34, 5.18it/s] 21%|██ | 9913/48008 [33:47<2:02:13, 5.19it/s] 21%|██ | 9914/48008 [33:48<2:01:59, 5.20it/s] 21%|██ | 9915/48008 [33:48<2:02:30, 5.18it/s] 21%|██ | 9916/48008 [33:48<2:02:39, 5.18it/s] 21%|██ | 9917/48008 [33:48<2:02:04, 5.20it/s] 21%|██ | 9918/48008 [33:48<2:02:59, 5.16it/s] 21%|██ | 9919/48008 [33:48<2:06:05, 5.03it/s] 21%|██ | 9920/48008 [33:49<2:06:46, 5.01it/s] 21%|██ | 9921/48008 [33:49<2:05:21, 5.06it/s] 21%|██ | 9922/48008 [33:49<2:05:02, 5.08it/s] 21%|██ | 9923/48008 [33:49<2:03:07, 5.16it/s] 21%|██ | 9924/48008 [33:49<2:04:33, 5.10it/s] 21%|██ | 9925/48008 [33:50<2:03:24, 5.14it/s] 21%|██ | 9926/48008 [33:50<2:02:59, 5.16it/s] 21%|██ | 9927/48008 [33:50<2:11:03, 4.84it/s] 21%|██ | 9928/48008 [33:50<2:07:47, 4.97it/s] 21%|██ | 9929/48008 [33:50<2:05:06, 5.07it/s] 21%|██ | 9930/48008 [33:51<2:06:03, 5.03it/s] 21%|██ | 9931/48008 [33:51<2:04:29, 5.10it/s] 21%|██ | 9932/48008 [33:51<2:05:47, 5.04it/s] 21%|██ | 9933/48008 [33:51<2:03:44, 5.13it/s] 21%|██ | 9934/48008 [33:52<2:24:45, 4.38it/s] 21%|██ | 9935/48008 [33:52<2:19:58, 4.53it/s] 21%|██ | 9936/48008 [33:52<2:16:47, 4.64it/s] 21%|██ | 9937/48008 [33:52<2:14:37, 4.71it/s] 21%|██ | 9938/48008 [33:52<2:13:40, 4.75it/s] 21%|██ | 9939/48008 [33:53<2:09:12, 4.91it/s] 21%|██ | 9940/48008 [33:53<2:07:36, 4.97it/s] 21%|██ | 9941/48008 [33:53<2:05:07, 5.07it/s] 21%|██ | 9942/48008 [33:53<2:03:18, 5.14it/s] 21%|██ | 9943/48008 [33:53<2:00:51, 5.25it/s] 21%|██ | 9944/48008 [33:53<1:59:09, 5.32it/s] 21%|██ | 9945/48008 [33:54<2:02:10, 5.19it/s] 21%|██ | 9946/48008 [33:54<2:03:55, 5.12it/s] 21%|██ | 9947/48008 [33:54<2:05:13, 5.07it/s] 21%|██ | 9948/48008 [33:54<2:06:26, 5.02it/s] 21%|██ | 9949/48008 [33:54<2:04:46, 5.08it/s] 21%|██ | 9950/48008 [33:55<2:07:43, 4.97it/s] {'loss': 0.3048, 'grad_norm': 0.177734375, 'learning_rate': 0.00015855274120979838, 'epoch': 0.21} 21%|██ | 9950/48008 [33:55<2:07:43, 4.97it/s] 21%|██ | 9951/48008 [33:55<2:05:53, 5.04it/s] 21%|██ | 9952/48008 [33:55<2:13:16, 4.76it/s] 21%|██ | 9953/48008 [33:55<2:18:26, 4.58it/s] 21%|██ | 9954/48008 [33:56<2:15:17, 4.69it/s] 21%|██ | 9955/48008 [33:56<2:14:13, 4.72it/s] 21%|██ | 9956/48008 [33:56<2:11:16, 4.83it/s] 21%|██ | 9957/48008 [33:56<2:08:36, 4.93it/s] 21%|██ | 9958/48008 [33:56<2:07:21, 4.98it/s] 21%|██ | 9959/48008 [33:57<2:07:22, 4.98it/s] 21%|██ | 9960/48008 [33:57<2:04:33, 5.09it/s] 21%|██ | 9961/48008 [33:57<2:03:05, 5.15it/s] 21%|██ | 9962/48008 [33:57<2:02:24, 5.18it/s] 21%|██ | 9963/48008 [33:57<2:11:07, 4.84it/s] 21%|██ | 9964/48008 [33:58<2:45:33, 3.83it/s] 21%|██ | 9965/48008 [33:58<2:32:24, 4.16it/s] 21%|██ | 9966/48008 [33:58<2:24:25, 4.39it/s] 21%|██ | 9967/48008 [33:58<2:19:20, 4.55it/s] 21%|██ | 9968/48008 [33:59<2:16:18, 4.65it/s] 21%|██ | 9969/48008 [33:59<2:13:42, 4.74it/s] 21%|██ | 9970/48008 [33:59<2:08:42, 4.93it/s] 21%|██ | 9971/48008 [33:59<2:06:53, 5.00it/s] 21%|██ | 9972/48008 [33:59<2:05:10, 5.06it/s] 21%|██ | 9973/48008 [34:00<2:04:07, 5.11it/s] 21%|██ | 9974/48008 [34:00<2:25:05, 4.37it/s] 21%|██ | 9975/48008 [34:00<2:18:31, 4.58it/s] 21%|██ | 9976/48008 [34:00<2:12:22, 4.79it/s] 21%|██ | 9977/48008 [34:00<2:08:59, 4.91it/s] 21%|██ | 9978/48008 [34:01<2:07:58, 4.95it/s] 21%|██ | 9979/48008 [34:01<2:27:57, 4.28it/s] 21%|██ | 9980/48008 [34:01<2:18:58, 4.56it/s] 21%|██ | 9981/48008 [34:01<2:35:35, 4.07it/s] 21%|██ | 9982/48008 [34:02<2:28:16, 4.27it/s] 21%|██ | 9983/48008 [34:02<2:20:16, 4.52it/s] 21%|██ | 9984/48008 [34:02<2:13:05, 4.76it/s] 21%|██ | 9985/48008 [34:02<2:10:39, 4.85it/s] 21%|██ | 9986/48008 [34:02<2:06:59, 4.99it/s] 21%|██ | 9987/48008 [34:03<2:05:30, 5.05it/s] 21%|██ | 9988/48008 [34:03<2:04:15, 5.10it/s] 21%|██ | 9989/48008 [34:03<2:02:18, 5.18it/s] 21%|██ | 9990/48008 [34:03<2:01:50, 5.20it/s] 21%|██ | 9991/48008 [34:03<2:01:21, 5.22it/s] 21%|██ | 9992/48008 [34:04<2:00:22, 5.26it/s] 21%|██ | 9993/48008 [34:04<2:02:46, 5.16it/s] 21%|██ | 9994/48008 [34:04<2:01:25, 5.22it/s] 21%|██ | 9995/48008 [34:04<2:04:48, 5.08it/s] 21%|██ | 9996/48008 [34:04<2:06:16, 5.02it/s] 21%|██ | 9997/48008 [34:04<2:04:33, 5.09it/s] 21%|██ | 9998/48008 [34:05<2:03:23, 5.13it/s] 21%|██ | 9999/48008 [34:05<2:03:31, 5.13it/s] 21%|██ | 10000/48008 [34:05<2:02:06, 5.19it/s] {'loss': 0.2998, 'grad_norm': 0.578125, 'learning_rate': 0.0001583444425929012, 'epoch': 0.21} 21%|██ | 10000/48008 [34:05<2:02:06, 5.19it/s] 21%|██ | 10001/48008 [34:05<2:23:36, 4.41it/s] 21%|██ | 10002/48008 [34:06<2:16:31, 4.64it/s] 21%|██ | 10003/48008 [34:06<2:14:47, 4.70it/s] 21%|██ | 10004/48008 [34:06<2:10:39, 4.85it/s] 21%|██ | 10005/48008 [34:06<2:08:36, 4.93it/s] 21%|██ | 10006/48008 [34:06<2:15:01, 4.69it/s] 21%|██ | 10007/48008 [34:07<2:10:33, 4.85it/s] 21%|██ | 10008/48008 [34:07<2:08:40, 4.92it/s] 21%|██ | 10009/48008 [34:07<2:07:44, 4.96it/s] 21%|██ | 10010/48008 [34:07<2:05:26, 5.05it/s] 21%|██ | 10011/48008 [34:07<2:03:48, 5.12it/s] 21%|██ | 10012/48008 [34:08<2:02:19, 5.18it/s] 21%|██ | 10013/48008 [34:08<2:10:43, 4.84it/s] 21%|██ | 10014/48008 [34:08<2:07:43, 4.96it/s] 21%|██ | 10015/48008 [34:08<2:06:56, 4.99it/s] 21%|██ | 10016/48008 [34:08<2:07:10, 4.98it/s] 21%|██ | 10017/48008 [34:09<2:04:38, 5.08it/s] 21%|██ | 10018/48008 [34:09<2:03:43, 5.12it/s] 21%|██ | 10019/48008 [34:09<2:01:48, 5.20it/s] 21%|██ | 10020/48008 [34:09<2:01:27, 5.21it/s] 21%|██ | 10021/48008 [34:09<2:02:15, 5.18it/s] 21%|██ | 10022/48008 [34:10<2:38:58, 3.98it/s] 21%|██ | 10023/48008 [34:10<2:27:19, 4.30it/s] 21%|██ | 10024/48008 [34:10<2:18:43, 4.56it/s] 21%|██ | 10025/48008 [34:10<2:12:47, 4.77it/s] 21%|██ | 10026/48008 [34:10<2:09:48, 4.88it/s] 21%|██ | 10027/48008 [34:11<2:10:32, 4.85it/s] 21%|██ | 10028/48008 [34:11<2:07:38, 4.96it/s] 21%|██ | 10029/48008 [34:11<2:05:14, 5.05it/s] 21%|██ | 10030/48008 [34:11<2:04:08, 5.10it/s] 21%|██ | 10031/48008 [34:11<2:02:34, 5.16it/s] 21%|██ | 10032/48008 [34:12<2:10:47, 4.84it/s] 21%|██ | 10033/48008 [34:12<2:07:26, 4.97it/s] 21%|██ | 10034/48008 [34:12<2:04:38, 5.08it/s] 21%|██ | 10035/48008 [34:12<2:12:18, 4.78it/s] 21%|██ | 10036/48008 [34:12<2:08:46, 4.91it/s] 21%|██ | 10037/48008 [34:13<2:07:08, 4.98it/s] 21%|██ | 10038/48008 [34:13<2:04:30, 5.08it/s] 21%|██ | 10039/48008 [34:13<2:03:21, 5.13it/s] 21%|██ | 10040/48008 [34:13<2:04:53, 5.07it/s] 21%|██ | 10041/48008 [34:13<2:03:59, 5.10it/s] 21%|██ | 10042/48008 [34:14<2:02:50, 5.15it/s] 21%|██ | 10043/48008 [34:14<2:01:49, 5.19it/s] 21%|██ | 10044/48008 [34:14<2:01:42, 5.20it/s] 21%|██ | 10045/48008 [34:14<2:00:25, 5.25it/s] 21%|██ | 10046/48008 [34:14<2:04:00, 5.10it/s] 21%|██ | 10047/48008 [34:15<2:03:17, 5.13it/s] 21%|██ | 10048/48008 [34:15<2:02:19, 5.17it/s] 21%|██ | 10049/48008 [34:15<2:04:23, 5.09it/s] 21%|██ | 10050/48008 [34:15<2:02:56, 5.15it/s] {'loss': 0.3013, 'grad_norm': 0.2353515625, 'learning_rate': 0.00015813614397600402, 'epoch': 0.21} 21%|██ | 10050/48008 [34:15<2:02:56, 5.15it/s] 21%|██ | 10051/48008 [34:15<2:03:45, 5.11it/s] 21%|██ | 10052/48008 [34:16<2:02:59, 5.14it/s] 21%|██ | 10053/48008 [34:16<2:01:20, 5.21it/s] 21%|██ | 10054/48008 [34:16<2:00:06, 5.27it/s] 21%|██ | 10055/48008 [34:16<2:00:20, 5.26it/s] 21%|██ | 10056/48008 [34:16<2:00:59, 5.23it/s] 21%|██ | 10057/48008 [34:17<2:09:22, 4.89it/s] 21%|██ | 10058/48008 [34:17<2:06:16, 5.01it/s] 21%|██ | 10059/48008 [34:17<2:04:17, 5.09it/s] 21%|██ | 10060/48008 [34:17<2:02:10, 5.18it/s] 21%|██ | 10061/48008 [34:17<2:02:04, 5.18it/s] 21%|██ | 10062/48008 [34:18<2:01:37, 5.20it/s] 21%|██ | 10063/48008 [34:18<2:02:44, 5.15it/s] 21%|██ | 10064/48008 [34:18<2:11:11, 4.82it/s] 21%|██ | 10065/48008 [34:18<2:45:12, 3.83it/s] 21%|██ | 10066/48008 [34:19<2:32:27, 4.15it/s] 21%|██ | 10067/48008 [34:19<2:22:52, 4.43it/s] 21%|██ | 10068/48008 [34:19<2:17:26, 4.60it/s] 21%|██ | 10069/48008 [34:19<2:11:31, 4.81it/s] 21%|██ | 10070/48008 [34:19<2:08:12, 4.93it/s] 21%|██ | 10071/48008 [34:19<2:03:57, 5.10it/s] 21%|██ | 10072/48008 [34:20<2:02:37, 5.16it/s] 21%|██ | 10073/48008 [34:20<2:01:42, 5.19it/s] 21%|██ | 10074/48008 [34:20<2:01:02, 5.22it/s] 21%|██ | 10075/48008 [34:20<2:01:46, 5.19it/s] 21%|██ | 10076/48008 [34:20<2:02:35, 5.16it/s] 21%|██ | 10077/48008 [34:21<2:10:56, 4.83it/s] 21%|██ | 10078/48008 [34:21<2:08:53, 4.90it/s] 21%|██ | 10079/48008 [34:21<2:06:41, 4.99it/s] 21%|██ | 10080/48008 [34:21<2:05:00, 5.06it/s] 21%|██ | 10081/48008 [34:21<2:04:43, 5.07it/s] 21%|██ | 10082/48008 [34:22<2:40:32, 3.94it/s] 21%|██ | 10083/48008 [34:22<2:50:29, 3.71it/s] 21%|██ | 10084/48008 [34:23<3:12:25, 3.28it/s] 21%|██ | 10085/48008 [34:23<2:50:23, 3.71it/s] 21%|██ | 10086/48008 [34:23<2:35:09, 4.07it/s] 21%|██ | 10087/48008 [34:23<2:24:36, 4.37it/s] 21%|██ | 10088/48008 [34:23<2:16:31, 4.63it/s] 21%|██ | 10089/48008 [34:23<2:11:45, 4.80it/s] 21%|██ | 10090/48008 [34:24<2:10:47, 4.83it/s] 21%|██ | 10091/48008 [34:24<2:09:57, 4.86it/s] 21%|██ | 10092/48008 [34:24<2:09:40, 4.87it/s] 21%|██ | 10093/48008 [34:24<2:08:54, 4.90it/s] 21%|██ | 10094/48008 [34:24<2:06:13, 5.01it/s] 21%|██ | 10095/48008 [34:25<2:04:50, 5.06it/s] 21%|██ | 10096/48008 [34:25<2:03:13, 5.13it/s] 21%|██ | 10097/48008 [34:25<2:01:34, 5.20it/s] 21%|██ | 10098/48008 [34:25<2:01:14, 5.21it/s] 21%|██ | 10099/48008 [34:25<2:00:54, 5.23it/s] 21%|██ | 10100/48008 [34:26<2:02:06, 5.17it/s] {'loss': 0.3028, 'grad_norm': 0.515625, 'learning_rate': 0.00015792784535910683, 'epoch': 0.21} 21%|██ | 10100/48008 [34:26<2:02:06, 5.17it/s] 21%|██ | 10101/48008 [34:26<2:01:29, 5.20it/s] 21%|██ | 10102/48008 [34:26<2:00:28, 5.24it/s] 21%|██ | 10103/48008 [34:26<2:00:34, 5.24it/s] 21%|██ | 10104/48008 [34:26<2:00:17, 5.25it/s] 21%|██ | 10105/48008 [34:27<2:00:15, 5.25it/s] 21%|██ | 10106/48008 [34:27<1:59:47, 5.27it/s] 21%|██ | 10107/48008 [34:27<2:08:43, 4.91it/s] 21%|██ | 10108/48008 [34:27<2:08:46, 4.90it/s] 21%|██ | 10109/48008 [34:27<2:06:24, 5.00it/s] 21%|██ | 10110/48008 [34:28<2:04:10, 5.09it/s] 21%|██ | 10111/48008 [34:28<2:05:02, 5.05it/s] 21%|██ | 10112/48008 [34:28<2:03:23, 5.12it/s] 21%|██ | 10113/48008 [34:28<2:03:49, 5.10it/s] 21%|██ | 10114/48008 [34:29<4:11:33, 2.51it/s] 21%|██ | 10115/48008 [34:29<3:32:02, 2.98it/s] 21%|██ | 10116/48008 [34:29<3:13:09, 3.27it/s] 21%|██ | 10117/48008 [34:30<2:53:40, 3.64it/s] 21%|██ | 10118/48008 [34:30<2:37:37, 4.01it/s] 21%|██ | 10119/48008 [34:30<2:28:23, 4.26it/s] 21%|██ | 10120/48008 [34:30<2:19:04, 4.54it/s] 21%|██ | 10121/48008 [34:30<2:13:47, 4.72it/s] 21%|██ | 10122/48008 [34:31<2:11:57, 4.79it/s] 21%|██ | 10123/48008 [34:31<2:10:24, 4.84it/s] 21%|██ | 10124/48008 [34:31<2:07:42, 4.94it/s] 21%|██ | 10125/48008 [34:31<2:07:04, 4.97it/s] 21%|██ | 10126/48008 [34:31<2:03:56, 5.09it/s] 21%|██ | 10127/48008 [34:32<2:02:38, 5.15it/s] 21%|██ | 10128/48008 [34:32<2:01:51, 5.18it/s] 21%|██ | 10129/48008 [34:32<2:02:12, 5.17it/s] 21%|██ | 10130/48008 [34:32<2:04:03, 5.09it/s] 21%|██ | 10131/48008 [34:33<2:25:00, 4.35it/s] 21%|██ | 10132/48008 [34:33<2:20:13, 4.50it/s] 21%|██ | 10133/48008 [34:33<2:15:49, 4.65it/s] 21%|██ | 10134/48008 [34:33<2:11:31, 4.80it/s] 21%|██ | 10135/48008 [34:33<2:08:02, 4.93it/s] 21%|██ | 10136/48008 [34:33<2:05:30, 5.03it/s] 21%|██ | 10137/48008 [34:34<2:02:57, 5.13it/s] 21%|██ | 10138/48008 [34:34<2:01:41, 5.19it/s] 21%|██ | 10139/48008 [34:34<2:00:30, 5.24it/s] 21%|██ | 10140/48008 [34:34<1:59:13, 5.29it/s] 21%|██ | 10141/48008 [34:34<2:01:41, 5.19it/s] 21%|██ | 10142/48008 [34:35<2:00:36, 5.23it/s] 21%|██ | 10143/48008 [34:35<2:02:35, 5.15it/s] 21%|██ | 10144/48008 [34:35<2:02:02, 5.17it/s] 21%|██ | 10145/48008 [34:35<2:01:37, 5.19it/s] 21%|██ | 10146/48008 [34:35<2:00:22, 5.24it/s] 21%|██ | 10147/48008 [34:36<2:02:09, 5.17it/s] 21%|██ | 10148/48008 [34:36<2:01:37, 5.19it/s] 21%|██ | 10149/48008 [34:36<2:03:43, 5.10it/s] 21%|██ | 10150/48008 [34:36<2:02:28, 5.15it/s] {'loss': 0.3028, 'grad_norm': 0.177734375, 'learning_rate': 0.00015771954674220962, 'epoch': 0.21} 21%|██ | 10150/48008 [34:36<2:02:28, 5.15it/s] 21%|██ | 10151/48008 [34:36<2:02:18, 5.16it/s] 21%|██ | 10152/48008 [34:37<2:05:21, 5.03it/s] 21%|██ | 10153/48008 [34:37<2:04:00, 5.09it/s] 21%|██ | 10154/48008 [34:37<2:03:14, 5.12it/s] 21%|██ | 10155/48008 [34:37<2:02:38, 5.14it/s] 21%|██ | 10156/48008 [34:37<2:01:48, 5.18it/s] 21%|██ | 10157/48008 [34:38<2:00:21, 5.24it/s] 21%|██ | 10158/48008 [34:38<2:00:50, 5.22it/s] 21%|██ | 10159/48008 [34:38<2:00:25, 5.24it/s] 21%|██ | 10160/48008 [34:38<1:59:37, 5.27it/s] 21%|██ | 10161/48008 [34:38<1:59:28, 5.28it/s] 21%|██ | 10162/48008 [34:38<1:59:11, 5.29it/s] 21%|██ | 10163/48008 [34:39<1:59:18, 5.29it/s] 21%|██ | 10164/48008 [34:39<1:59:24, 5.28it/s] 21%|██ | 10165/48008 [34:39<2:00:44, 5.22it/s] 21%|██ | 10166/48008 [34:39<2:04:01, 5.09it/s] 21%|██ | 10167/48008 [34:39<2:03:21, 5.11it/s] 21%|██ | 10168/48008 [34:40<2:02:25, 5.15it/s] 21%|██ | 10169/48008 [34:40<2:01:39, 5.18it/s] 21%|██ | 10170/48008 [34:40<2:01:15, 5.20it/s] 21%|██ | 10171/48008 [34:40<2:01:02, 5.21it/s] 21%|██ | 10172/48008 [34:40<2:00:35, 5.23it/s] 21%|██ | 10173/48008 [34:41<2:00:13, 5.25it/s] 21%|██ | 10174/48008 [34:41<1:59:12, 5.29it/s] 21%|██ | 10175/48008 [34:41<2:00:11, 5.25it/s] 21%|██ | 10176/48008 [34:41<2:02:40, 5.14it/s] 21%|██ | 10177/48008 [34:41<2:01:52, 5.17it/s] 21%|██ | 10178/48008 [34:42<2:01:41, 5.18it/s] 21%|██ | 10179/48008 [34:42<2:03:03, 5.12it/s] 21%|██ | 10180/48008 [34:42<2:01:53, 5.17it/s] 21%|██ | 10181/48008 [34:42<2:10:07, 4.84it/s] 21%|██ | 10182/48008 [34:42<2:06:57, 4.97it/s] 21%|██ | 10183/48008 [34:43<2:05:59, 5.00it/s] 21%|██ | 10184/48008 [34:43<2:04:04, 5.08it/s] 21%|██ | 10185/48008 [34:43<2:02:30, 5.15it/s] 21%|██ | 10186/48008 [34:43<2:01:45, 5.18it/s] 21%|██ | 10187/48008 [34:43<2:01:09, 5.20it/s] 21%|██ | 10188/48008 [34:44<2:01:23, 5.19it/s] 21%|██ | 10189/48008 [34:44<2:03:14, 5.11it/s] 21%|██ | 10190/48008 [34:44<2:03:20, 5.11it/s] 21%|██ | 10191/48008 [34:44<2:01:44, 5.18it/s] 21%|██ | 10192/48008 [34:44<2:02:12, 5.16it/s] 21%|██ | 10193/48008 [34:44<2:00:49, 5.22it/s] 21%|██ | 10194/48008 [34:45<1:59:42, 5.26it/s] 21%|██ | 10195/48008 [34:45<2:08:34, 4.90it/s] 21%|██ | 10196/48008 [34:45<2:06:37, 4.98it/s] 21%|██ | 10197/48008 [34:45<2:03:25, 5.11it/s] 21%|██ | 10198/48008 [34:45<2:01:56, 5.17it/s] 21%|██ | 10199/48008 [34:46<2:04:34, 5.06it/s] 21%|██ | 10200/48008 [34:46<2:03:32, 5.10it/s] {'loss': 0.3038, 'grad_norm': 0.26171875, 'learning_rate': 0.00015751124812531247, 'epoch': 0.21} 21%|██ | 10200/48008 [34:46<2:03:32, 5.10it/s] 21%|██ | 10201/48008 [34:46<2:04:08, 5.08it/s] 21%|██▏ | 10202/48008 [34:46<2:05:06, 5.04it/s] 21%|██▏ | 10203/48008 [34:46<2:03:40, 5.09it/s] 21%|██▏ | 10204/48008 [34:47<2:02:13, 5.15it/s] 21%|██▏ | 10205/48008 [34:47<2:01:23, 5.19it/s] 21%|██▏ | 10206/48008 [34:47<1:59:20, 5.28it/s] 21%|██▏ | 10207/48008 [34:47<1:59:58, 5.25it/s] 21%|██▏ | 10208/48008 [34:47<1:59:24, 5.28it/s] 21%|██▏ | 10209/48008 [34:48<1:59:28, 5.27it/s] 21%|██▏ | 10210/48008 [34:48<1:58:52, 5.30it/s] 21%|██▏ | 10211/48008 [34:48<2:07:45, 4.93it/s] 21%|██▏ | 10212/48008 [34:48<2:07:49, 4.93it/s] 21%|██▏ | 10213/48008 [34:48<2:04:53, 5.04it/s] 21%|██▏ | 10214/48008 [34:49<2:03:57, 5.08it/s] 21%|██▏ | 10215/48008 [34:49<2:02:45, 5.13it/s] 21%|██▏ | 10216/48008 [34:49<2:03:10, 5.11it/s] 21%|██▏ | 10217/48008 [34:49<2:02:29, 5.14it/s] 21%|██▏ | 10218/48008 [34:49<2:04:16, 5.07it/s] 21%|██▏ | 10219/48008 [34:50<2:03:27, 5.10it/s] 21%|██▏ | 10220/48008 [34:50<2:04:51, 5.04it/s] 21%|██▏ | 10221/48008 [34:50<2:12:14, 4.76it/s] 21%|██▏ | 10222/48008 [34:50<2:08:31, 4.90it/s] 21%|██▏ | 10223/48008 [34:50<2:09:19, 4.87it/s] 21%|██▏ | 10224/48008 [34:51<2:07:51, 4.93it/s] 21%|██▏ | 10225/48008 [34:51<2:04:37, 5.05it/s] 21%|██▏ | 10226/48008 [34:51<2:05:15, 5.03it/s] 21%|██▏ | 10227/48008 [34:51<2:04:10, 5.07it/s] 21%|██▏ | 10228/48008 [34:51<2:03:05, 5.12it/s] 21%|██▏ | 10229/48008 [34:52<2:02:24, 5.14it/s] 21%|██▏ | 10230/48008 [34:52<2:04:26, 5.06it/s] 21%|██▏ | 10231/48008 [34:52<2:04:33, 5.05it/s] 21%|██▏ | 10232/48008 [34:52<2:03:15, 5.11it/s] 21%|██▏ | 10233/48008 [34:52<2:01:30, 5.18it/s] 21%|██▏ | 10234/48008 [34:53<2:00:05, 5.24it/s] 21%|██▏ | 10235/48008 [34:53<2:08:52, 4.88it/s] 21%|██▏ | 10236/48008 [34:53<2:06:07, 4.99it/s] 21%|██▏ | 10237/48008 [34:53<2:05:02, 5.03it/s] 21%|██▏ | 10238/48008 [34:53<2:03:50, 5.08it/s] 21%|██▏ | 10239/48008 [34:54<2:06:11, 4.99it/s] 21%|██▏ | 10240/48008 [34:54<2:03:42, 5.09it/s] 21%|██▏ | 10241/48008 [34:54<2:02:16, 5.15it/s] 21%|██▏ | 10242/48008 [34:54<2:01:29, 5.18it/s] 21%|██▏ | 10243/48008 [34:54<2:01:13, 5.19it/s] 21%|██▏ | 10244/48008 [34:55<2:01:53, 5.16it/s] 21%|██▏ | 10245/48008 [34:55<2:01:29, 5.18it/s] 21%|██▏ | 10246/48008 [34:55<2:24:17, 4.36it/s] 21%|██▏ | 10247/48008 [34:55<2:17:33, 4.58it/s] 21%|██▏ | 10248/48008 [34:55<2:13:00, 4.73it/s] 21%|██▏ | 10249/48008 [34:56<2:08:21, 4.90it/s] 21%|██▏ | 10250/48008 [34:56<2:15:25, 4.65it/s] {'loss': 0.3035, 'grad_norm': 0.349609375, 'learning_rate': 0.00015730294950841526, 'epoch': 0.21} 21%|██▏ | 10250/48008 [34:56<2:15:25, 4.65it/s] 21%|██▏ | 10251/48008 [34:56<2:09:54, 4.84it/s] 21%|██▏ | 10252/48008 [34:56<2:15:51, 4.63it/s] 21%|██▏ | 10253/48008 [34:56<2:11:58, 4.77it/s] 21%|██▏ | 10254/48008 [34:57<2:07:45, 4.93it/s] 21%|██▏ | 10255/48008 [34:57<2:04:45, 5.04it/s] 21%|██▏ | 10256/48008 [34:57<2:04:03, 5.07it/s] 21%|██▏ | 10257/48008 [34:57<2:05:15, 5.02it/s] 21%|██▏ | 10258/48008 [34:57<2:03:32, 5.09it/s] 21%|██▏ | 10259/48008 [34:58<2:11:14, 4.79it/s] 21%|██▏ | 10260/48008 [34:58<2:07:13, 4.95it/s] 21%|██▏ | 10261/48008 [34:58<2:04:36, 5.05it/s] 21%|██▏ | 10262/48008 [34:58<2:02:34, 5.13it/s] 21%|██▏ | 10263/48008 [34:58<2:01:03, 5.20it/s] 21%|██▏ | 10264/48008 [34:59<2:22:40, 4.41it/s] 21%|██▏ | 10265/48008 [34:59<2:16:04, 4.62it/s] 21%|██▏ | 10266/48008 [34:59<2:48:09, 3.74it/s] 21%|██▏ | 10267/48008 [34:59<2:34:09, 4.08it/s] 21%|██▏ | 10268/48008 [35:00<3:00:50, 3.48it/s] 21%|██▏ | 10269/48008 [35:00<2:43:04, 3.86it/s] 21%|██▏ | 10270/48008 [35:00<2:30:41, 4.17it/s] 21%|██▏ | 10271/48008 [35:00<2:22:36, 4.41it/s] 21%|██▏ | 10272/48008 [35:01<2:15:30, 4.64it/s] 21%|██▏ | 10273/48008 [35:01<2:10:35, 4.82it/s] 21%|██▏ | 10274/48008 [35:01<2:06:20, 4.98it/s] 21%|██▏ | 10275/48008 [35:01<2:03:30, 5.09it/s] 21%|██▏ | 10276/48008 [35:01<2:02:39, 5.13it/s] 21%|██▏ | 10277/48008 [35:02<2:04:09, 5.07it/s] 21%|██▏ | 10278/48008 [35:02<2:24:33, 4.35it/s] 21%|██▏ | 10279/48008 [35:02<2:17:23, 4.58it/s] 21%|██▏ | 10280/48008 [35:02<2:12:01, 4.76it/s] 21%|██▏ | 10281/48008 [35:02<2:08:18, 4.90it/s] 21%|██▏ | 10282/48008 [35:03<2:05:45, 5.00it/s] 21%|██▏ | 10283/48008 [35:03<2:02:59, 5.11it/s] 21%|██▏ | 10284/48008 [35:03<2:02:55, 5.11it/s] 21%|██▏ | 10285/48008 [35:03<2:01:31, 5.17it/s] 21%|██▏ | 10286/48008 [35:03<2:00:53, 5.20it/s] 21%|██▏ | 10287/48008 [35:04<2:00:17, 5.23it/s] 21%|██▏ | 10288/48008 [35:04<2:02:34, 5.13it/s] 21%|██▏ | 10289/48008 [35:04<2:02:58, 5.11it/s] 21%|██▏ | 10290/48008 [35:04<2:05:33, 5.01it/s] 21%|██▏ | 10291/48008 [35:04<2:03:13, 5.10it/s] 21%|██▏ | 10292/48008 [35:05<2:03:24, 5.09it/s] 21%|██▏ | 10293/48008 [35:05<2:02:31, 5.13it/s] 21%|██▏ | 10294/48008 [35:05<2:01:23, 5.18it/s] 21%|██▏ | 10295/48008 [35:05<2:01:11, 5.19it/s] 21%|██▏ | 10296/48008 [35:05<2:04:19, 5.06it/s] 21%|██▏ | 10297/48008 [35:06<2:02:32, 5.13it/s] 21%|██▏ | 10298/48008 [35:06<2:38:42, 3.96it/s] 21%|██▏ | 10299/48008 [35:06<2:26:48, 4.28it/s] 21%|██▏ | 10300/48008 [35:06<2:19:42, 4.50it/s] {'loss': 0.3033, 'grad_norm': 0.6796875, 'learning_rate': 0.0001570946508915181, 'epoch': 0.21} 21%|██▏ | 10300/48008 [35:06<2:19:42, 4.50it/s] 21%|██▏ | 10301/48008 [35:07<2:16:18, 4.61it/s] 21%|██▏ | 10302/48008 [35:07<2:11:11, 4.79it/s] 21%|██▏ | 10303/48008 [35:07<2:07:19, 4.94it/s] 21%|██▏ | 10304/48008 [35:07<2:04:55, 5.03it/s] 21%|██▏ | 10305/48008 [35:07<2:03:42, 5.08it/s] 21%|██▏ | 10306/48008 [35:08<2:02:01, 5.15it/s] 21%|██▏ | 10307/48008 [35:08<2:00:35, 5.21it/s] 21%|██▏ | 10308/48008 [35:08<2:01:08, 5.19it/s] 21%|██▏ | 10309/48008 [35:08<2:04:07, 5.06it/s] 21%|██▏ | 10310/48008 [35:08<2:39:56, 3.93it/s] 21%|██▏ | 10311/48008 [35:09<2:28:22, 4.23it/s] 21%|██▏ | 10312/48008 [35:10<5:01:28, 2.08it/s] 21%|██▏ | 10313/48008 [35:10<4:07:01, 2.54it/s] 21%|██▏ | 10314/48008 [35:10<3:26:56, 3.04it/s] 21%|██▏ | 10315/48008 [35:10<3:01:49, 3.46it/s] 21%|██▏ | 10316/48008 [35:10<2:42:31, 3.87it/s] 21%|██▏ | 10317/48008 [35:11<2:29:59, 4.19it/s] 21%|██▏ | 10318/48008 [35:12<4:35:30, 2.28it/s] 21%|██▏ | 10319/48008 [35:12<3:49:44, 2.73it/s] 21%|██▏ | 10320/48008 [35:12<3:16:00, 3.20it/s] 21%|██▏ | 10321/48008 [35:12<2:52:27, 3.64it/s] 22%|██▏ | 10322/48008 [35:12<2:45:19, 3.80it/s] 22%|██▏ | 10323/48008 [35:13<2:40:23, 3.92it/s] 22%|██▏ | 10324/48008 [35:13<4:24:48, 2.37it/s] 22%|██▏ | 10325/48008 [35:14<3:41:04, 2.84it/s] 22%|██▏ | 10326/48008 [35:14<3:10:28, 3.30it/s] 22%|██▏ | 10327/48008 [35:14<2:52:34, 3.64it/s] 22%|██▏ | 10328/48008 [35:14<2:40:04, 3.92it/s] 22%|██▏ | 10329/48008 [35:14<2:28:00, 4.24it/s] 22%|██▏ | 10330/48008 [35:16<5:22:59, 1.94it/s] 22%|██▏ | 10331/48008 [35:16<4:22:17, 2.39it/s] 22%|██▏ | 10332/48008 [35:16<3:40:33, 2.85it/s] 22%|██▏ | 10333/48008 [35:16<3:10:24, 3.30it/s] 22%|██▏ | 10334/48008 [35:16<2:51:10, 3.67it/s] 22%|██▏ | 10335/48008 [35:17<2:35:16, 4.04it/s] 22%|██▏ | 10336/48008 [35:17<2:25:50, 4.30it/s] 22%|██▏ | 10337/48008 [35:17<2:16:15, 4.61it/s] 22%|██▏ | 10338/48008 [35:17<2:10:08, 4.82it/s] 22%|██▏ | 10339/48008 [35:17<2:28:57, 4.21it/s] 22%|██▏ | 10340/48008 [35:18<2:21:27, 4.44it/s] 22%|██▏ | 10341/48008 [35:18<2:15:45, 4.62it/s] 22%|██▏ | 10342/48008 [35:19<4:11:45, 2.49it/s] 22%|██▏ | 10343/48008 [35:19<3:34:28, 2.93it/s] 22%|██▏ | 10344/48008 [35:19<3:06:57, 3.36it/s] 22%|██▏ | 10345/48008 [35:19<2:46:32, 3.77it/s] 22%|██▏ | 10346/48008 [35:19<2:34:10, 4.07it/s] 22%|██▏ | 10347/48008 [35:20<2:24:15, 4.35it/s] 22%|██▏ | 10348/48008 [35:20<2:16:43, 4.59it/s] 22%|██▏ | 10349/48008 [35:20<2:13:48, 4.69it/s] 22%|██▏ | 10350/48008 [35:20<2:09:46, 4.84it/s] {'loss': 0.3027, 'grad_norm': 0.7890625, 'learning_rate': 0.0001568863522746209, 'epoch': 0.22} 22%|██▏ | 10350/48008 [35:20<2:09:46, 4.84it/s] 22%|██▏ | 10351/48008 [35:20<2:09:28, 4.85it/s] 22%|██▏ | 10352/48008 [35:21<2:06:51, 4.95it/s] 22%|██▏ | 10353/48008 [35:21<2:04:21, 5.05it/s] 22%|██▏ | 10354/48008 [35:21<2:02:54, 5.11it/s] 22%|██▏ | 10355/48008 [35:21<2:04:22, 5.05it/s] 22%|██▏ | 10356/48008 [35:21<2:01:04, 5.18it/s] 22%|██▏ | 10357/48008 [35:22<2:00:31, 5.21it/s] 22%|██▏ | 10358/48008 [35:22<2:03:07, 5.10it/s] 22%|██▏ | 10359/48008 [35:22<2:04:04, 5.06it/s] 22%|██▏ | 10360/48008 [35:22<2:02:52, 5.11it/s] 22%|██▏ | 10361/48008 [35:22<2:02:46, 5.11it/s] 22%|██▏ | 10362/48008 [35:23<2:02:11, 5.13it/s] 22%|██▏ | 10363/48008 [35:23<2:02:16, 5.13it/s] 22%|██▏ | 10364/48008 [35:23<2:02:39, 5.12it/s] 22%|██▏ | 10365/48008 [35:23<2:02:34, 5.12it/s] 22%|██▏ | 10366/48008 [35:23<2:01:35, 5.16it/s] 22%|██▏ | 10367/48008 [35:24<2:01:31, 5.16it/s] 22%|██▏ | 10368/48008 [35:24<2:00:59, 5.19it/s] 22%|██▏ | 10369/48008 [35:24<2:01:33, 5.16it/s] 22%|██▏ | 10370/48008 [35:24<2:00:50, 5.19it/s] 22%|██▏ | 10371/48008 [35:24<2:04:54, 5.02it/s] 22%|██▏ | 10372/48008 [35:25<2:11:46, 4.76it/s] 22%|██▏ | 10373/48008 [35:25<2:10:44, 4.80it/s] 22%|██▏ | 10374/48008 [35:25<2:09:15, 4.85it/s] 22%|██▏ | 10375/48008 [35:25<2:06:24, 4.96it/s] 22%|██▏ | 10376/48008 [35:25<2:05:39, 4.99it/s] 22%|██▏ | 10377/48008 [35:26<2:05:09, 5.01it/s] 22%|██▏ | 10378/48008 [35:26<2:01:50, 5.15it/s] 22%|██▏ | 10379/48008 [35:26<2:01:13, 5.17it/s] 22%|██▏ | 10380/48008 [35:26<2:00:44, 5.19it/s] 22%|██▏ | 10381/48008 [35:26<2:01:14, 5.17it/s] 22%|██▏ | 10382/48008 [35:26<2:00:26, 5.21it/s] 22%|██▏ | 10383/48008 [35:27<2:00:54, 5.19it/s] 22%|██▏ | 10384/48008 [35:27<2:00:29, 5.20it/s] 22%|██▏ | 10385/48008 [35:27<2:04:42, 5.03it/s] 22%|██▏ | 10386/48008 [35:27<2:07:23, 4.92it/s] 22%|██▏ | 10387/48008 [35:27<2:05:07, 5.01it/s] 22%|██▏ | 10388/48008 [35:28<2:03:57, 5.06it/s] 22%|██▏ | 10389/48008 [35:28<2:03:35, 5.07it/s] 22%|██▏ | 10390/48008 [35:28<2:02:06, 5.13it/s] 22%|██▏ | 10391/48008 [35:28<2:01:33, 5.16it/s] 22%|██▏ | 10392/48008 [35:28<2:01:07, 5.18it/s] 22%|██▏ | 10393/48008 [35:29<2:02:46, 5.11it/s] 22%|██▏ | 10394/48008 [35:29<2:02:09, 5.13it/s] 22%|██▏ | 10395/48008 [35:29<2:38:40, 3.95it/s] 22%|██▏ | 10396/48008 [35:29<2:27:03, 4.26it/s] 22%|██▏ | 10397/48008 [35:30<2:17:47, 4.55it/s] 22%|██▏ | 10398/48008 [35:30<2:12:39, 4.73it/s] 22%|██▏ | 10399/48008 [35:30<2:07:58, 4.90it/s] 22%|██▏ | 10400/48008 [35:30<2:05:22, 5.00it/s] {'loss': 0.3024, 'grad_norm': 0.1318359375, 'learning_rate': 0.0001566780536577237, 'epoch': 0.22} 22%|██▏ | 10400/48008 [35:30<2:05:22, 5.00it/s] 22%|██▏ | 10401/48008 [35:30<2:03:49, 5.06it/s] 22%|██▏ | 10402/48008 [35:31<2:24:11, 4.35it/s] 22%|██▏ | 10403/48008 [35:31<2:18:43, 4.52it/s] 22%|██▏ | 10404/48008 [35:31<2:49:54, 3.69it/s] 22%|██▏ | 10405/48008 [35:31<2:37:02, 3.99it/s] 22%|██▏ | 10406/48008 [35:32<2:34:42, 4.05it/s] 22%|██▏ | 10407/48008 [35:32<2:25:53, 4.30it/s] 22%|██▏ | 10408/48008 [35:32<2:20:40, 4.45it/s] 22%|██▏ | 10409/48008 [35:32<2:15:26, 4.63it/s] 22%|██▏ | 10410/48008 [35:33<2:10:52, 4.79it/s] 22%|██▏ | 10411/48008 [35:33<2:06:49, 4.94it/s] 22%|██▏ | 10412/48008 [35:33<2:04:22, 5.04it/s] 22%|██▏ | 10413/48008 [35:33<2:02:58, 5.09it/s] 22%|██▏ | 10414/48008 [35:33<2:02:48, 5.10it/s] 22%|██▏ | 10415/48008 [35:33<2:01:59, 5.14it/s] 22%|██▏ | 10416/48008 [35:34<2:00:04, 5.22it/s] 22%|██▏ | 10417/48008 [35:34<2:00:04, 5.22it/s] 22%|██▏ | 10418/48008 [35:34<2:00:53, 5.18it/s] 22%|██▏ | 10419/48008 [35:34<2:03:26, 5.07it/s] 22%|██▏ | 10420/48008 [35:34<2:03:41, 5.06it/s] 22%|██▏ | 10421/48008 [35:35<2:01:47, 5.14it/s] 22%|██▏ | 10422/48008 [35:35<2:02:59, 5.09it/s] 22%|██▏ | 10423/48008 [35:35<2:02:56, 5.09it/s] 22%|██▏ | 10424/48008 [35:35<2:01:49, 5.14it/s] 22%|██▏ | 10425/48008 [35:35<2:00:58, 5.18it/s] 22%|██▏ | 10426/48008 [35:36<2:00:15, 5.21it/s] 22%|██▏ | 10427/48008 [35:36<1:59:47, 5.23it/s] 22%|██▏ | 10428/48008 [35:36<2:00:41, 5.19it/s] 22%|██▏ | 10429/48008 [35:36<2:00:28, 5.20it/s] 22%|██▏ | 10430/48008 [35:36<1:58:37, 5.28it/s] 22%|██▏ | 10431/48008 [35:37<2:01:18, 5.16it/s] 22%|██▏ | 10432/48008 [35:37<2:00:51, 5.18it/s] 22%|██▏ | 10433/48008 [35:37<2:02:28, 5.11it/s] 22%|██▏ | 10434/48008 [35:37<2:00:27, 5.20it/s] 22%|██▏ | 10435/48008 [35:37<2:00:10, 5.21it/s] 22%|██▏ | 10436/48008 [35:38<2:02:33, 5.11it/s] 22%|██▏ | 10437/48008 [35:38<2:01:18, 5.16it/s] 22%|██▏ | 10438/48008 [35:38<2:02:19, 5.12it/s] 22%|██▏ | 10439/48008 [35:38<2:01:15, 5.16it/s] 22%|██▏ | 10440/48008 [35:38<2:00:57, 5.18it/s] 22%|██▏ | 10441/48008 [35:38<1:59:50, 5.22it/s] 22%|██▏ | 10442/48008 [35:39<1:59:34, 5.24it/s] 22%|██▏ | 10443/48008 [35:39<1:59:56, 5.22it/s] 22%|██▏ | 10444/48008 [35:39<2:00:00, 5.22it/s] 22%|██▏ | 10445/48008 [35:39<2:02:02, 5.13it/s] 22%|██▏ | 10446/48008 [35:39<2:00:23, 5.20it/s] 22%|██▏ | 10447/48008 [35:40<2:03:04, 5.09it/s] 22%|██▏ | 10448/48008 [35:40<2:04:33, 5.03it/s] 22%|██▏ | 10449/48008 [35:40<2:03:03, 5.09it/s] 22%|██▏ | 10450/48008 [35:40<2:02:09, 5.12it/s] {'loss': 0.3012, 'grad_norm': 0.1484375, 'learning_rate': 0.00015646975504082653, 'epoch': 0.22} 22%|██▏ | 10450/48008 [35:40<2:02:09, 5.12it/s] 22%|██▏ | 10451/48008 [35:40<2:04:01, 5.05it/s] 22%|██▏ | 10452/48008 [35:41<2:02:22, 5.11it/s] 22%|██▏ | 10453/48008 [35:41<2:03:38, 5.06it/s] 22%|██▏ | 10454/48008 [35:41<2:03:37, 5.06it/s] 22%|██▏ | 10455/48008 [35:41<2:04:44, 5.02it/s] 22%|██▏ | 10456/48008 [35:42<2:24:21, 4.34it/s] 22%|██▏ | 10457/48008 [35:42<2:16:41, 4.58it/s] 22%|██▏ | 10458/48008 [35:42<2:11:26, 4.76it/s] 22%|██▏ | 10459/48008 [35:42<2:07:06, 4.92it/s] 22%|██▏ | 10460/48008 [35:42<2:04:46, 5.02it/s] 22%|██▏ | 10461/48008 [35:42<2:03:30, 5.07it/s] 22%|██▏ | 10462/48008 [35:43<2:05:10, 5.00it/s] 22%|██▏ | 10463/48008 [35:43<2:04:11, 5.04it/s] 22%|██▏ | 10464/48008 [35:43<2:04:47, 5.01it/s] 22%|██▏ | 10465/48008 [35:43<2:03:57, 5.05it/s] 22%|██▏ | 10466/48008 [35:43<2:03:39, 5.06it/s] 22%|██▏ | 10467/48008 [35:44<2:01:34, 5.15it/s] 22%|██▏ | 10468/48008 [35:44<2:00:42, 5.18it/s] 22%|██▏ | 10469/48008 [35:44<2:00:13, 5.20it/s] 22%|██▏ | 10470/48008 [35:44<2:01:59, 5.13it/s] 22%|██▏ | 10471/48008 [35:44<2:02:42, 5.10it/s] 22%|██▏ | 10472/48008 [35:45<2:01:47, 5.14it/s] 22%|██▏ | 10473/48008 [35:45<2:00:34, 5.19it/s] 22%|██▏ | 10474/48008 [35:45<2:00:56, 5.17it/s] 22%|██▏ | 10475/48008 [35:45<2:36:58, 3.99it/s] 22%|██▏ | 10476/48008 [35:46<2:28:05, 4.22it/s] 22%|██▏ | 10477/48008 [35:46<2:19:34, 4.48it/s] 22%|██▏ | 10478/48008 [35:46<2:13:34, 4.68it/s] 22%|██▏ | 10479/48008 [35:46<2:11:32, 4.75it/s] 22%|██▏ | 10480/48008 [35:46<2:06:47, 4.93it/s] 22%|██▏ | 10481/48008 [35:47<2:05:47, 4.97it/s] 22%|██▏ | 10482/48008 [35:47<2:01:54, 5.13it/s] 22%|██▏ | 10483/48008 [35:47<2:01:16, 5.16it/s] 22%|██▏ | 10484/48008 [35:47<2:00:46, 5.18it/s] 22%|██▏ | 10485/48008 [35:47<2:00:20, 5.20it/s] 22%|██▏ | 10486/48008 [35:48<1:59:02, 5.25it/s] 22%|██▏ | 10487/48008 [35:48<2:00:50, 5.18it/s] 22%|██▏ | 10488/48008 [35:48<2:00:11, 5.20it/s] 22%|██▏ | 10489/48008 [35:48<2:01:32, 5.14it/s] 22%|██▏ | 10490/48008 [35:48<2:04:15, 5.03it/s] 22%|██▏ | 10491/48008 [35:49<2:02:49, 5.09it/s] 22%|██▏ | 10492/48008 [35:49<2:01:58, 5.13it/s] 22%|██▏ | 10493/48008 [35:49<2:01:12, 5.16it/s] 22%|██▏ | 10494/48008 [35:49<2:00:48, 5.18it/s] 22%|██▏ | 10495/48008 [35:49<2:03:55, 5.05it/s] 22%|██▏ | 10496/48008 [35:49<2:01:42, 5.14it/s] 22%|██▏ | 10497/48008 [35:50<2:01:11, 5.16it/s] 22%|██▏ | 10498/48008 [35:50<2:03:07, 5.08it/s] 22%|██▏ | 10499/48008 [35:50<2:01:14, 5.16it/s] 22%|██▏ | 10500/48008 [35:50<2:00:44, 5.18it/s] {'loss': 0.3019, 'grad_norm': 0.140625, 'learning_rate': 0.00015626145642392935, 'epoch': 0.22} 22%|██▏ | 10500/48008 [35:50<2:00:44, 5.18it/s] 22%|██▏ | 10501/48008 [35:50<2:00:51, 5.17it/s] 22%|██▏ | 10502/48008 [35:51<1:59:57, 5.21it/s] 22%|██▏ | 10503/48008 [35:51<2:00:13, 5.20it/s] 22%|██▏ | 10504/48008 [35:51<1:59:44, 5.22it/s] 22%|██▏ | 10505/48008 [35:51<2:00:02, 5.21it/s] 22%|██▏ | 10506/48008 [35:51<2:00:00, 5.21it/s] 22%|██▏ | 10507/48008 [35:52<2:00:24, 5.19it/s] 22%|██▏ | 10508/48008 [35:52<1:59:16, 5.24it/s] 22%|██▏ | 10509/48008 [35:52<1:59:12, 5.24it/s] 22%|██▏ | 10510/48008 [35:52<2:01:27, 5.15it/s] 22%|██▏ | 10511/48008 [35:52<2:00:49, 5.17it/s] 22%|██▏ | 10512/48008 [35:53<1:59:15, 5.24it/s] 22%|██▏ | 10513/48008 [35:53<2:07:59, 4.88it/s] 22%|██▏ | 10514/48008 [35:53<2:07:43, 4.89it/s] 22%|██▏ | 10515/48008 [35:53<2:04:49, 5.01it/s] 22%|██▏ | 10516/48008 [35:53<2:04:00, 5.04it/s] 22%|██▏ | 10517/48008 [35:54<2:03:04, 5.08it/s] 22%|██▏ | 10518/48008 [35:54<2:02:04, 5.12it/s] 22%|██▏ | 10519/48008 [35:54<2:01:05, 5.16it/s] 22%|██▏ | 10520/48008 [35:54<2:00:38, 5.18it/s] 22%|██▏ | 10521/48008 [35:54<2:00:21, 5.19it/s] 22%|██▏ | 10522/48008 [35:55<1:59:59, 5.21it/s] 22%|██▏ | 10523/48008 [35:55<2:08:33, 4.86it/s] 22%|██▏ | 10524/48008 [35:55<2:05:52, 4.96it/s] 22%|██▏ | 10525/48008 [35:55<2:04:12, 5.03it/s] 22%|██▏ | 10526/48008 [35:55<2:02:22, 5.10it/s] 22%|██▏ | 10527/48008 [35:56<2:01:05, 5.16it/s] 22%|██▏ | 10528/48008 [35:56<2:02:18, 5.11it/s] 22%|██▏ | 10529/48008 [35:56<2:01:15, 5.15it/s] 22%|██▏ | 10530/48008 [35:56<2:03:25, 5.06it/s] 22%|██▏ | 10531/48008 [35:56<2:01:37, 5.14it/s] 22%|██▏ | 10532/48008 [35:57<2:01:17, 5.15it/s] 22%|██▏ | 10533/48008 [35:57<2:04:00, 5.04it/s] 22%|██▏ | 10534/48008 [35:57<2:02:53, 5.08it/s] 22%|██▏ | 10535/48008 [35:57<2:05:23, 4.98it/s] 22%|██▏ | 10536/48008 [35:57<2:03:42, 5.05it/s] 22%|██▏ | 10537/48008 [35:58<2:04:34, 5.01it/s] 22%|██▏ | 10538/48008 [35:58<2:02:58, 5.08it/s] 22%|██▏ | 10539/48008 [35:58<2:03:01, 5.08it/s] 22%|██▏ | 10540/48008 [35:58<2:01:32, 5.14it/s] 22%|██▏ | 10541/48008 [35:58<2:22:30, 4.38it/s] 22%|██▏ | 10542/48008 [35:59<2:15:02, 4.62it/s] 22%|██▏ | 10543/48008 [35:59<2:09:17, 4.83it/s] 22%|██▏ | 10544/48008 [35:59<2:06:29, 4.94it/s] 22%|██▏ | 10545/48008 [35:59<2:05:23, 4.98it/s] 22%|██▏ | 10546/48008 [35:59<2:12:06, 4.73it/s] 22%|██▏ | 10547/48008 [36:00<2:17:15, 4.55it/s] 22%|██▏ | 10548/48008 [36:00<2:11:37, 4.74it/s] 22%|██▏ | 10549/48008 [36:00<2:16:31, 4.57it/s] 22%|██▏ | 10550/48008 [36:00<2:11:24, 4.75it/s] {'loss': 0.3031, 'grad_norm': 0.28515625, 'learning_rate': 0.0001560531578070322, 'epoch': 0.22} 22%|██▏ | 10550/48008 [36:00<2:11:24, 4.75it/s] 22%|██▏ | 10551/48008 [36:00<2:09:47, 4.81it/s] 22%|██▏ | 10552/48008 [36:01<2:07:59, 4.88it/s] 22%|██▏ | 10553/48008 [36:01<2:04:47, 5.00it/s] 22%|██▏ | 10554/48008 [36:01<2:02:53, 5.08it/s] 22%|██▏ | 10555/48008 [36:01<2:01:21, 5.14it/s] 22%|██▏ | 10556/48008 [36:01<2:01:05, 5.15it/s] 22%|██▏ | 10557/48008 [36:02<2:02:28, 5.10it/s] 22%|██▏ | 10558/48008 [36:02<2:03:13, 5.06it/s] 22%|██▏ | 10559/48008 [36:02<2:00:29, 5.18it/s] 22%|██▏ | 10560/48008 [36:02<2:00:24, 5.18it/s] 22%|██▏ | 10561/48008 [36:02<1:59:23, 5.23it/s] 22%|██▏ | 10562/48008 [36:03<1:59:38, 5.22it/s] 22%|██▏ | 10563/48008 [36:03<1:58:44, 5.26it/s] 22%|██▏ | 10564/48008 [36:03<1:59:38, 5.22it/s] 22%|██▏ | 10565/48008 [36:03<1:58:41, 5.26it/s] 22%|██▏ | 10566/48008 [36:03<1:58:33, 5.26it/s] 22%|██▏ | 10567/48008 [36:04<1:56:52, 5.34it/s] 22%|██▏ | 10568/48008 [36:04<1:57:23, 5.32it/s] 22%|██▏ | 10569/48008 [36:04<1:58:57, 5.25it/s] 22%|██▏ | 10570/48008 [36:04<1:58:53, 5.25it/s] 22%|██▏ | 10571/48008 [36:04<1:58:14, 5.28it/s] 22%|██▏ | 10572/48008 [36:04<1:57:40, 5.30it/s] 22%|██▏ | 10573/48008 [36:05<1:56:58, 5.33it/s] 22%|██▏ | 10574/48008 [36:05<1:57:35, 5.31it/s] 22%|██▏ | 10575/48008 [36:05<2:00:14, 5.19it/s] 22%|██▏ | 10576/48008 [36:05<1:59:47, 5.21it/s] 22%|██▏ | 10577/48008 [36:05<1:59:35, 5.22it/s] 22%|██▏ | 10578/48008 [36:06<2:02:02, 5.11it/s] 22%|██▏ | 10579/48008 [36:06<2:02:27, 5.09it/s] 22%|██▏ | 10580/48008 [36:06<2:00:30, 5.18it/s] 22%|██▏ | 10581/48008 [36:06<2:03:21, 5.06it/s] 22%|██▏ | 10582/48008 [36:06<2:02:00, 5.11it/s] 22%|██▏ | 10583/48008 [36:07<2:00:57, 5.16it/s] 22%|██▏ | 10584/48008 [36:07<1:59:10, 5.23it/s] 22%|██▏ | 10585/48008 [36:07<1:59:01, 5.24it/s] 22%|██▏ | 10586/48008 [36:07<2:00:52, 5.16it/s] 22%|██▏ | 10587/48008 [36:07<2:00:14, 5.19it/s] 22%|██▏ | 10588/48008 [36:08<2:00:33, 5.17it/s] 22%|██▏ | 10589/48008 [36:08<2:00:07, 5.19it/s] 22%|██▏ | 10590/48008 [36:08<1:59:51, 5.20it/s] 22%|██▏ | 10591/48008 [36:08<1:58:00, 5.28it/s] 22%|██▏ | 10592/48008 [36:08<1:57:20, 5.31it/s] 22%|██▏ | 10593/48008 [36:09<1:57:44, 5.30it/s] 22%|██▏ | 10594/48008 [36:09<1:57:26, 5.31it/s] 22%|██▏ | 10595/48008 [36:09<1:57:09, 5.32it/s] 22%|██▏ | 10596/48008 [36:09<2:01:18, 5.14it/s] 22%|██▏ | 10597/48008 [36:09<2:00:13, 5.19it/s] 22%|██▏ | 10598/48008 [36:09<1:59:42, 5.21it/s] 22%|██▏ | 10599/48008 [36:10<1:59:36, 5.21it/s] 22%|██▏ | 10600/48008 [36:10<2:00:18, 5.18it/s] {'loss': 0.3039, 'grad_norm': 0.28515625, 'learning_rate': 0.00015584485919013498, 'epoch': 0.22} 22%|██▏ | 10600/48008 [36:10<2:00:18, 5.18it/s] 22%|██▏ | 10601/48008 [36:10<1:59:10, 5.23it/s] 22%|██▏ | 10602/48008 [36:10<1:59:21, 5.22it/s] 22%|██▏ | 10603/48008 [36:10<1:59:15, 5.23it/s] 22%|██▏ | 10604/48008 [36:11<2:07:51, 4.88it/s] 22%|██▏ | 10605/48008 [36:11<2:05:10, 4.98it/s] 22%|██▏ | 10606/48008 [36:11<2:03:03, 5.07it/s] 22%|██▏ | 10607/48008 [36:11<2:01:52, 5.11it/s] 22%|██▏ | 10608/48008 [36:11<2:03:14, 5.06it/s] 22%|██▏ | 10609/48008 [36:12<2:01:55, 5.11it/s] 22%|██▏ | 10610/48008 [36:12<2:00:57, 5.15it/s] 22%|██▏ | 10611/48008 [36:12<1:59:02, 5.24it/s] 22%|██▏ | 10612/48008 [36:12<2:00:05, 5.19it/s] 22%|██▏ | 10613/48008 [36:12<1:59:50, 5.20it/s] 22%|██▏ | 10614/48008 [36:13<1:58:37, 5.25it/s] 22%|██▏ | 10615/48008 [36:13<1:58:52, 5.24it/s] 22%|██▏ | 10616/48008 [36:13<1:59:03, 5.23it/s] 22%|██▏ | 10617/48008 [36:13<1:59:12, 5.23it/s] 22%|██▏ | 10618/48008 [36:13<2:00:26, 5.17it/s] 22%|██▏ | 10619/48008 [36:14<2:02:04, 5.10it/s] 22%|██▏ | 10620/48008 [36:14<2:01:13, 5.14it/s] 22%|██▏ | 10621/48008 [36:14<2:00:44, 5.16it/s] 22%|██▏ | 10622/48008 [36:14<2:01:24, 5.13it/s] 22%|██▏ | 10623/48008 [36:14<2:00:55, 5.15it/s] 22%|██▏ | 10624/48008 [36:15<2:00:07, 5.19it/s] 22%|██▏ | 10625/48008 [36:15<2:00:21, 5.18it/s] 22%|██▏ | 10626/48008 [36:15<2:00:04, 5.19it/s] 22%|██▏ | 10627/48008 [36:15<1:59:37, 5.21it/s] 22%|██▏ | 10628/48008 [36:15<1:59:02, 5.23it/s] 22%|██▏ | 10629/48008 [36:15<1:58:11, 5.27it/s] 22%|██▏ | 10630/48008 [36:16<1:58:18, 5.27it/s] 22%|██▏ | 10631/48008 [36:16<1:58:09, 5.27it/s] 22%|██▏ | 10632/48008 [36:16<1:59:33, 5.21it/s] 22%|██▏ | 10633/48008 [36:16<1:59:14, 5.22it/s] 22%|██▏ | 10634/48008 [36:16<1:59:56, 5.19it/s] 22%|██▏ | 10635/48008 [36:17<2:00:26, 5.17it/s] 22%|██▏ | 10636/48008 [36:17<1:59:38, 5.21it/s] 22%|██▏ | 10637/48008 [36:17<2:21:00, 4.42it/s] 22%|██▏ | 10638/48008 [36:17<2:13:10, 4.68it/s] 22%|██▏ | 10639/48008 [36:17<2:09:10, 4.82it/s] 22%|██▏ | 10640/48008 [36:18<2:06:19, 4.93it/s] 22%|██▏ | 10641/48008 [36:18<2:04:19, 5.01it/s] 22%|██▏ | 10642/48008 [36:18<2:05:51, 4.95it/s] 22%|██▏ | 10643/48008 [36:18<2:03:52, 5.03it/s] 22%|██▏ | 10644/48008 [36:18<2:02:46, 5.07it/s] 22%|██▏ | 10645/48008 [36:19<2:01:27, 5.13it/s] 22%|██▏ | 10646/48008 [36:19<2:00:27, 5.17it/s] 22%|██▏ | 10647/48008 [36:19<1:59:19, 5.22it/s] 22%|██▏ | 10648/48008 [36:19<1:58:56, 5.23it/s] 22%|██▏ | 10649/48008 [36:19<1:59:03, 5.23it/s] 22%|██▏ | 10650/48008 [36:20<1:58:57, 5.23it/s] {'loss': 0.3036, 'grad_norm': 0.349609375, 'learning_rate': 0.0001556365605732378, 'epoch': 0.22} 22%|██▏ | 10650/48008 [36:20<1:58:57, 5.23it/s] 22%|██▏ | 10651/48008 [36:20<1:59:13, 5.22it/s] 22%|██▏ | 10652/48008 [36:20<1:58:46, 5.24it/s] 22%|██▏ | 10653/48008 [36:20<2:02:19, 5.09it/s] 22%|██▏ | 10654/48008 [36:20<2:00:30, 5.17it/s] 22%|██▏ | 10655/48008 [36:21<1:59:24, 5.21it/s] 22%|██▏ | 10656/48008 [36:21<1:59:36, 5.20it/s] 22%|██▏ | 10657/48008 [36:21<1:59:27, 5.21it/s] 22%|██▏ | 10658/48008 [36:21<2:02:08, 5.10it/s] 22%|██▏ | 10659/48008 [36:21<2:03:12, 5.05it/s] 22%|██▏ | 10660/48008 [36:22<2:03:19, 5.05it/s] 22%|██▏ | 10661/48008 [36:22<2:03:27, 5.04it/s] 22%|██▏ | 10662/48008 [36:22<2:01:12, 5.14it/s] 22%|██▏ | 10663/48008 [36:22<2:00:26, 5.17it/s] 22%|██▏ | 10664/48008 [36:22<2:00:26, 5.17it/s] 22%|██▏ | 10665/48008 [36:23<2:08:16, 4.85it/s] 22%|██▏ | 10666/48008 [36:23<2:06:18, 4.93it/s] 22%|██▏ | 10667/48008 [36:23<2:12:38, 4.69it/s] 22%|██▏ | 10668/48008 [36:23<2:08:26, 4.85it/s] 22%|██▏ | 10669/48008 [36:23<2:05:27, 4.96it/s] 22%|██▏ | 10670/48008 [36:24<2:03:29, 5.04it/s] 22%|██▏ | 10671/48008 [36:24<2:04:04, 5.02it/s] 22%|██▏ | 10672/48008 [36:24<2:04:46, 4.99it/s] 22%|██▏ | 10673/48008 [36:24<2:04:15, 5.01it/s] 22%|██▏ | 10674/48008 [36:24<2:02:48, 5.07it/s] 22%|██▏ | 10675/48008 [36:25<2:01:08, 5.14it/s] 22%|██▏ | 10676/48008 [36:25<2:03:39, 5.03it/s] 22%|██▏ | 10677/48008 [36:25<2:02:15, 5.09it/s] 22%|██▏ | 10678/48008 [36:25<2:04:46, 4.99it/s] 22%|██▏ | 10679/48008 [36:25<2:02:41, 5.07it/s] 22%|██▏ | 10680/48008 [36:26<2:02:14, 5.09it/s] 22%|██▏ | 10681/48008 [36:26<2:01:01, 5.14it/s] 22%|██▏ | 10682/48008 [36:26<2:02:20, 5.08it/s] 22%|██▏ | 10683/48008 [36:26<2:03:17, 5.05it/s] 22%|██▏ | 10684/48008 [36:26<2:02:45, 5.07it/s] 22%|██▏ | 10685/48008 [36:27<2:02:20, 5.08it/s] 22%|██▏ | 10686/48008 [36:27<2:22:43, 4.36it/s] 22%|██▏ | 10687/48008 [36:27<2:15:38, 4.59it/s] 22%|██▏ | 10688/48008 [36:27<2:19:08, 4.47it/s] 22%|██▏ | 10689/48008 [36:27<2:12:39, 4.69it/s] 22%|██▏ | 10690/48008 [36:28<2:07:33, 4.88it/s] 22%|██▏ | 10691/48008 [36:28<2:07:54, 4.86it/s] 22%|██▏ | 10692/48008 [36:28<2:05:02, 4.97it/s] 22%|██▏ | 10693/48008 [36:28<2:11:27, 4.73it/s] 22%|██▏ | 10694/48008 [36:28<2:07:14, 4.89it/s] 22%|██▏ | 10695/48008 [36:29<2:04:57, 4.98it/s] 22%|██▏ | 10696/48008 [36:29<2:03:14, 5.05it/s] 22%|██▏ | 10697/48008 [36:29<2:01:40, 5.11it/s] 22%|██▏ | 10698/48008 [36:29<2:01:11, 5.13it/s] 22%|██▏ | 10699/48008 [36:29<2:00:02, 5.18it/s] 22%|██▏ | 10700/48008 [36:30<1:59:22, 5.21it/s] {'loss': 0.302, 'grad_norm': 0.1943359375, 'learning_rate': 0.00015542826195634062, 'epoch': 0.22} 22%|██▏ | 10700/48008 [36:30<1:59:22, 5.21it/s] 22%|██▏ | 10701/48008 [36:30<1:58:01, 5.27it/s] 22%|██▏ | 10702/48008 [36:30<1:57:58, 5.27it/s] 22%|██▏ | 10703/48008 [36:30<1:57:17, 5.30it/s] 22%|██▏ | 10704/48008 [36:30<1:56:45, 5.32it/s] 22%|██▏ | 10705/48008 [36:31<1:59:26, 5.20it/s] 22%|██▏ | 10706/48008 [36:31<1:59:13, 5.21it/s] 22%|██▏ | 10707/48008 [36:31<1:58:51, 5.23it/s] 22%|██▏ | 10708/48008 [36:31<1:58:44, 5.24it/s] 22%|██▏ | 10709/48008 [36:31<1:59:47, 5.19it/s] 22%|██▏ | 10710/48008 [36:32<1:59:31, 5.20it/s] 22%|██▏ | 10711/48008 [36:32<1:58:49, 5.23it/s] 22%|██▏ | 10712/48008 [36:32<2:01:28, 5.12it/s] 22%|██▏ | 10713/48008 [36:32<2:02:41, 5.07it/s] 22%|██▏ | 10714/48008 [36:32<2:02:18, 5.08it/s] 22%|██▏ | 10715/48008 [36:33<2:02:40, 5.07it/s] 22%|██▏ | 10716/48008 [36:33<2:01:31, 5.11it/s] 22%|██▏ | 10717/48008 [36:33<2:00:15, 5.17it/s] 22%|██▏ | 10718/48008 [36:33<1:59:04, 5.22it/s] 22%|██▏ | 10719/48008 [36:33<2:00:57, 5.14it/s] 22%|██▏ | 10720/48008 [36:33<1:59:23, 5.21it/s] 22%|██▏ | 10721/48008 [36:34<2:00:02, 5.18it/s] 22%|██▏ | 10722/48008 [36:34<1:59:43, 5.19it/s] 22%|██▏ | 10723/48008 [36:34<1:59:03, 5.22it/s] 22%|██▏ | 10724/48008 [36:34<1:58:58, 5.22it/s] 22%|██▏ | 10725/48008 [36:34<1:58:04, 5.26it/s] 22%|██▏ | 10726/48008 [36:35<1:57:56, 5.27it/s] 22%|██▏ | 10727/48008 [36:35<1:59:41, 5.19it/s] 22%|██▏ | 10728/48008 [36:35<1:59:12, 5.21it/s] 22%|██▏ | 10729/48008 [36:35<1:58:24, 5.25it/s] 22%|██▏ | 10730/48008 [36:35<1:58:16, 5.25it/s] 22%|██▏ | 10731/48008 [36:36<1:58:44, 5.23it/s] 22%|██▏ | 10732/48008 [36:36<1:59:00, 5.22it/s] 22%|██▏ | 10733/48008 [36:36<2:01:02, 5.13it/s] 22%|██▏ | 10734/48008 [36:36<2:00:05, 5.17it/s] 22%|██▏ | 10735/48008 [36:36<2:20:54, 4.41it/s] 22%|██▏ | 10736/48008 [36:37<2:13:17, 4.66it/s] 22%|██▏ | 10737/48008 [36:37<2:09:03, 4.81it/s] 22%|██▏ | 10738/48008 [36:37<2:05:15, 4.96it/s] 22%|██▏ | 10739/48008 [36:37<2:04:21, 4.99it/s] 22%|██▏ | 10740/48008 [36:37<2:02:38, 5.06it/s] 22%|██▏ | 10741/48008 [36:38<2:01:43, 5.10it/s] 22%|██▏ | 10742/48008 [36:38<2:01:08, 5.13it/s] 22%|██▏ | 10743/48008 [36:38<2:01:39, 5.11it/s] 22%|██▏ | 10744/48008 [36:38<2:37:05, 3.95it/s] 22%|██▏ | 10745/48008 [36:39<2:25:25, 4.27it/s] 22%|██▏ | 10746/48008 [36:39<2:25:37, 4.26it/s] 22%|██▏ | 10747/48008 [36:39<2:17:24, 4.52it/s] 22%|██▏ | 10748/48008 [36:39<2:11:04, 4.74it/s] 22%|██▏ | 10749/48008 [36:39<2:07:01, 4.89it/s] 22%|██▏ | 10750/48008 [36:40<2:05:44, 4.94it/s] {'loss': 0.3044, 'grad_norm': 0.1396484375, 'learning_rate': 0.00015521996333944343, 'epoch': 0.22} 22%|██▏ | 10750/48008 [36:40<2:05:44, 4.94it/s] 22%|██▏ | 10751/48008 [36:40<2:06:26, 4.91it/s] 22%|██▏ | 10752/48008 [36:40<2:03:49, 5.01it/s] 22%|██▏ | 10753/48008 [36:40<2:05:51, 4.93it/s] 22%|██▏ | 10754/48008 [36:40<2:03:38, 5.02it/s] 22%|██▏ | 10755/48008 [36:41<2:02:10, 5.08it/s] 22%|██▏ | 10756/48008 [36:41<2:09:46, 4.78it/s] 22%|██▏ | 10757/48008 [36:41<2:15:00, 4.60it/s] 22%|██▏ | 10758/48008 [36:41<2:09:19, 4.80it/s] 22%|██▏ | 10759/48008 [36:41<2:06:03, 4.93it/s] 22%|██▏ | 10760/48008 [36:42<2:03:55, 5.01it/s] 22%|██▏ | 10761/48008 [36:42<2:03:21, 5.03it/s] 22%|██▏ | 10762/48008 [36:42<2:01:34, 5.11it/s] 22%|██▏ | 10763/48008 [36:42<2:00:23, 5.16it/s] 22%|██▏ | 10764/48008 [36:42<2:21:30, 4.39it/s] 22%|██▏ | 10765/48008 [36:43<2:23:22, 4.33it/s] 22%|██▏ | 10766/48008 [36:43<2:19:07, 4.46it/s] 22%|██▏ | 10767/48008 [36:43<2:12:39, 4.68it/s] 22%|██▏ | 10768/48008 [36:43<2:07:37, 4.86it/s] 22%|██▏ | 10769/48008 [36:44<2:04:51, 4.97it/s] 22%|██▏ | 10770/48008 [36:44<2:02:08, 5.08it/s] 22%|██▏ | 10771/48008 [36:44<2:03:02, 5.04it/s] 22%|██▏ | 10772/48008 [36:44<2:04:00, 5.00it/s] 22%|██▏ | 10773/48008 [36:44<2:01:41, 5.10it/s] 22%|██▏ | 10774/48008 [36:44<1:59:52, 5.18it/s] 22%|██▏ | 10775/48008 [36:45<2:00:22, 5.16it/s] 22%|██▏ | 10776/48008 [36:45<2:02:07, 5.08it/s] 22%|██▏ | 10777/48008 [36:45<2:01:19, 5.11it/s] 22%|██▏ | 10778/48008 [36:45<2:01:56, 5.09it/s] 22%|██▏ | 10779/48008 [36:45<1:59:55, 5.17it/s] 22%|██▏ | 10780/48008 [36:46<1:59:24, 5.20it/s] 22%|██▏ | 10781/48008 [36:46<1:58:32, 5.23it/s] 22%|██▏ | 10782/48008 [36:46<1:58:32, 5.23it/s] 22%|██▏ | 10783/48008 [36:46<1:57:35, 5.28it/s] 22%|██▏ | 10784/48008 [36:46<1:56:56, 5.31it/s] 22%|██▏ | 10785/48008 [36:47<1:55:54, 5.35it/s] 22%|██▏ | 10786/48008 [36:47<1:56:35, 5.32it/s] 22%|██▏ | 10787/48008 [36:47<1:59:20, 5.20it/s] 22%|██▏ | 10788/48008 [36:47<1:59:14, 5.20it/s] 22%|██▏ | 10789/48008 [36:47<1:57:53, 5.26it/s] 22%|██▏ | 10790/48008 [36:48<2:00:13, 5.16it/s] 22%|██▏ | 10791/48008 [36:48<2:02:58, 5.04it/s] 22%|██▏ | 10792/48008 [36:48<2:03:28, 5.02it/s] 22%|██▏ | 10793/48008 [36:48<2:02:06, 5.08it/s] 22%|██▏ | 10794/48008 [36:48<2:00:29, 5.15it/s] 22%|██▏ | 10795/48008 [36:49<2:00:00, 5.17it/s] 22%|██▏ | 10796/48008 [36:49<2:08:06, 4.84it/s] 22%|██▏ | 10797/48008 [36:49<2:08:21, 4.83it/s] 22%|██▏ | 10798/48008 [36:49<2:05:30, 4.94it/s] 22%|██▏ | 10799/48008 [36:49<2:02:51, 5.05it/s] 22%|██▏ | 10800/48008 [36:50<2:04:38, 4.98it/s] {'loss': 0.3029, 'grad_norm': 0.248046875, 'learning_rate': 0.00015501166472254625, 'epoch': 0.22} 22%|██▏ | 10800/48008 [36:50<2:04:38, 4.98it/s] 22%|██▏ | 10801/48008 [36:50<2:02:54, 5.05it/s] 23%|██▎ | 10802/48008 [36:50<2:01:41, 5.10it/s] 23%|██▎ | 10803/48008 [36:50<2:02:52, 5.05it/s] 23%|██▎ | 10804/48008 [36:50<2:00:33, 5.14it/s] 23%|██▎ | 10805/48008 [36:51<2:00:50, 5.13it/s] 23%|██▎ | 10806/48008 [36:51<1:59:51, 5.17it/s] 23%|██▎ | 10807/48008 [36:51<1:59:20, 5.20it/s] 23%|██▎ | 10808/48008 [36:51<1:59:16, 5.20it/s] 23%|██▎ | 10809/48008 [36:51<1:59:03, 5.21it/s] 23%|██▎ | 10810/48008 [36:51<1:58:21, 5.24it/s] 23%|██▎ | 10811/48008 [36:52<1:58:19, 5.24it/s] 23%|██▎ | 10812/48008 [36:52<1:58:33, 5.23it/s] 23%|██▎ | 10813/48008 [36:52<1:58:14, 5.24it/s] 23%|██▎ | 10814/48008 [36:52<1:58:05, 5.25it/s] 23%|██▎ | 10815/48008 [36:52<2:00:19, 5.15it/s] 23%|██▎ | 10816/48008 [36:53<2:00:02, 5.16it/s] 23%|██▎ | 10817/48008 [36:53<1:59:22, 5.19it/s] 23%|██▎ | 10818/48008 [36:53<1:59:01, 5.21it/s] 23%|██▎ | 10819/48008 [36:53<1:58:51, 5.21it/s] 23%|██▎ | 10820/48008 [36:53<1:58:08, 5.25it/s] 23%|██▎ | 10821/48008 [36:54<2:01:22, 5.11it/s] 23%|██▎ | 10822/48008 [36:54<2:00:18, 5.15it/s] 23%|██▎ | 10823/48008 [36:54<2:00:03, 5.16it/s] 23%|██▎ | 10824/48008 [36:54<1:59:24, 5.19it/s] 23%|██▎ | 10825/48008 [36:54<1:59:05, 5.20it/s] 23%|██▎ | 10826/48008 [36:55<1:59:07, 5.20it/s] 23%|██▎ | 10827/48008 [36:55<1:59:16, 5.20it/s] 23%|██▎ | 10828/48008 [36:55<1:58:23, 5.23it/s] 23%|██▎ | 10829/48008 [36:55<1:57:46, 5.26it/s] 23%|██▎ | 10830/48008 [36:55<1:57:22, 5.28it/s] 23%|██▎ | 10831/48008 [36:56<1:57:55, 5.25it/s] 23%|██▎ | 10832/48008 [36:56<2:00:00, 5.16it/s] 23%|██▎ | 10833/48008 [36:56<2:00:22, 5.15it/s] 23%|██▎ | 10834/48008 [36:56<2:02:12, 5.07it/s] 23%|██▎ | 10835/48008 [36:56<2:01:14, 5.11it/s] 23%|██▎ | 10836/48008 [36:57<2:02:38, 5.05it/s] 23%|██▎ | 10837/48008 [36:57<2:03:34, 5.01it/s] 23%|██▎ | 10838/48008 [36:57<2:05:09, 4.95it/s] 23%|██▎ | 10839/48008 [36:57<2:02:45, 5.05it/s] 23%|██▎ | 10840/48008 [36:57<2:01:03, 5.12it/s] 23%|██▎ | 10841/48008 [36:57<2:01:07, 5.11it/s] 23%|██▎ | 10842/48008 [36:58<2:00:08, 5.16it/s] 23%|██▎ | 10843/48008 [36:58<1:59:08, 5.20it/s] 23%|██▎ | 10844/48008 [36:58<1:58:14, 5.24it/s] 23%|██▎ | 10845/48008 [36:58<1:58:49, 5.21it/s] 23%|██▎ | 10846/48008 [36:58<1:58:29, 5.23it/s] 23%|██▎ | 10847/48008 [36:59<2:01:03, 5.12it/s] 23%|██▎ | 10848/48008 [36:59<2:08:45, 4.81it/s] 23%|██▎ | 10849/48008 [36:59<2:05:37, 4.93it/s] 23%|██▎ | 10850/48008 [36:59<2:04:13, 4.99it/s] {'loss': 0.3046, 'grad_norm': 0.1357421875, 'learning_rate': 0.00015480336610564907, 'epoch': 0.23} 23%|██▎ | 10850/48008 [36:59<2:04:13, 4.99it/s] 23%|██▎ | 10851/48008 [36:59<2:02:12, 5.07it/s] 23%|██▎ | 10852/48008 [37:00<2:02:04, 5.07it/s] 23%|██▎ | 10853/48008 [37:00<2:02:30, 5.05it/s] 23%|██▎ | 10854/48008 [37:00<2:00:13, 5.15it/s] 23%|██▎ | 10855/48008 [37:00<1:59:42, 5.17it/s] 23%|██▎ | 10856/48008 [37:00<1:58:28, 5.23it/s] 23%|██▎ | 10857/48008 [37:01<1:58:36, 5.22it/s] 23%|██▎ | 10858/48008 [37:01<1:58:11, 5.24it/s] 23%|██▎ | 10859/48008 [37:01<1:58:26, 5.23it/s] 23%|██▎ | 10860/48008 [37:01<1:59:52, 5.16it/s] 23%|██▎ | 10861/48008 [37:01<1:59:35, 5.18it/s] 23%|██▎ | 10862/48008 [37:02<1:58:57, 5.20it/s] 23%|██▎ | 10863/48008 [37:02<1:58:47, 5.21it/s] 23%|██▎ | 10864/48008 [37:02<2:01:02, 5.11it/s] 23%|██▎ | 10865/48008 [37:02<2:00:07, 5.15it/s] 23%|██▎ | 10866/48008 [37:02<1:58:53, 5.21it/s] 23%|██▎ | 10867/48008 [37:03<1:59:26, 5.18it/s] 23%|██▎ | 10868/48008 [37:03<1:58:49, 5.21it/s] 23%|██▎ | 10869/48008 [37:03<1:58:27, 5.23it/s] 23%|██▎ | 10870/48008 [37:03<1:59:17, 5.19it/s] 23%|██▎ | 10871/48008 [37:03<1:58:17, 5.23it/s] 23%|██▎ | 10872/48008 [37:03<1:58:01, 5.24it/s] 23%|██▎ | 10873/48008 [37:04<1:59:44, 5.17it/s] 23%|██▎ | 10874/48008 [37:04<1:58:46, 5.21it/s] 23%|██▎ | 10875/48008 [37:04<1:58:05, 5.24it/s] 23%|██▎ | 10876/48008 [37:04<1:57:29, 5.27it/s] 23%|██▎ | 10877/48008 [37:04<1:57:46, 5.25it/s] 23%|██▎ | 10878/48008 [37:05<1:58:59, 5.20it/s] 23%|██▎ | 10879/48008 [37:05<1:57:36, 5.26it/s] 23%|██▎ | 10880/48008 [37:05<1:56:39, 5.30it/s] 23%|██▎ | 10881/48008 [37:05<1:57:03, 5.29it/s] 23%|██▎ | 10882/48008 [37:05<1:59:24, 5.18it/s] 23%|██▎ | 10883/48008 [37:06<1:59:50, 5.16it/s] 23%|██▎ | 10884/48008 [37:06<1:59:28, 5.18it/s] 23%|██▎ | 10885/48008 [37:06<2:02:32, 5.05it/s] 23%|██▎ | 10886/48008 [37:06<2:00:15, 5.14it/s] 23%|██▎ | 10887/48008 [37:06<2:00:35, 5.13it/s] 23%|██▎ | 10888/48008 [37:07<1:59:59, 5.16it/s] 23%|██▎ | 10889/48008 [37:07<1:59:06, 5.19it/s] 23%|██▎ | 10890/48008 [37:07<1:59:19, 5.18it/s] 23%|██▎ | 10891/48008 [37:07<1:59:12, 5.19it/s] 23%|██▎ | 10892/48008 [37:07<1:57:39, 5.26it/s] 23%|██▎ | 10893/48008 [37:08<1:59:54, 5.16it/s] 23%|██▎ | 10894/48008 [37:08<2:00:16, 5.14it/s] 23%|██▎ | 10895/48008 [37:08<1:59:40, 5.17it/s] 23%|██▎ | 10896/48008 [37:08<1:58:54, 5.20it/s] 23%|██▎ | 10897/48008 [37:08<1:57:30, 5.26it/s] 23%|██▎ | 10898/48008 [37:08<1:57:07, 5.28it/s] 23%|██▎ | 10899/48008 [37:09<1:57:42, 5.25it/s] 23%|██▎ | 10900/48008 [37:09<1:56:59, 5.29it/s] {'loss': 0.3044, 'grad_norm': 0.1513671875, 'learning_rate': 0.00015459506748875188, 'epoch': 0.23} 23%|██▎ | 10900/48008 [37:09<1:56:59, 5.29it/s] 23%|██▎ | 10901/48008 [37:09<1:57:28, 5.26it/s] 23%|██▎ | 10902/48008 [37:09<1:57:30, 5.26it/s] 23%|██▎ | 10903/48008 [37:09<2:06:06, 4.90it/s] 23%|██▎ | 10904/48008 [37:10<2:05:22, 4.93it/s] 23%|██▎ | 10905/48008 [37:10<2:02:24, 5.05it/s] 23%|██▎ | 10906/48008 [37:10<2:00:27, 5.13it/s] 23%|██▎ | 10907/48008 [37:10<1:59:45, 5.16it/s] 23%|██▎ | 10908/48008 [37:10<1:58:30, 5.22it/s] 23%|██▎ | 10909/48008 [37:11<1:59:43, 5.16it/s] 23%|██▎ | 10910/48008 [37:11<1:59:21, 5.18it/s] 23%|██▎ | 10911/48008 [37:11<1:58:22, 5.22it/s] 23%|██▎ | 10912/48008 [37:11<1:58:03, 5.24it/s] 23%|██▎ | 10913/48008 [37:11<1:58:16, 5.23it/s] 23%|██▎ | 10914/48008 [37:12<1:57:29, 5.26it/s] 23%|██▎ | 10915/48008 [37:12<1:58:55, 5.20it/s] 23%|██▎ | 10916/48008 [37:12<1:59:57, 5.15it/s] 23%|██▎ | 10917/48008 [37:12<1:58:27, 5.22it/s] 23%|██▎ | 10918/48008 [37:12<1:57:05, 5.28it/s] 23%|██▎ | 10919/48008 [37:13<1:57:01, 5.28it/s] 23%|██▎ | 10920/48008 [37:13<1:59:46, 5.16it/s] 23%|██▎ | 10921/48008 [37:13<1:59:14, 5.18it/s] 23%|██▎ | 10922/48008 [37:13<2:02:29, 5.05it/s] 23%|██▎ | 10923/48008 [37:13<2:00:06, 5.15it/s] 23%|██▎ | 10924/48008 [37:14<2:07:56, 4.83it/s] 23%|██▎ | 10925/48008 [37:14<2:04:23, 4.97it/s] 23%|██▎ | 10926/48008 [37:14<2:04:40, 4.96it/s] 23%|██▎ | 10927/48008 [37:14<2:02:31, 5.04it/s] 23%|██▎ | 10928/48008 [37:14<2:00:32, 5.13it/s] 23%|██▎ | 10929/48008 [37:15<2:08:19, 4.82it/s] 23%|██▎ | 10930/48008 [37:15<2:04:36, 4.96it/s] 23%|██▎ | 10931/48008 [37:15<2:04:09, 4.98it/s] 23%|██▎ | 10932/48008 [37:15<2:11:06, 4.71it/s] 23%|██▎ | 10933/48008 [37:15<2:08:13, 4.82it/s] 23%|██▎ | 10934/48008 [37:16<2:05:17, 4.93it/s] 23%|██▎ | 10935/48008 [37:16<2:03:43, 4.99it/s] 23%|██▎ | 10936/48008 [37:16<2:10:44, 4.73it/s] 23%|██▎ | 10937/48008 [37:16<2:09:11, 4.78it/s] 23%|██▎ | 10938/48008 [37:16<2:05:58, 4.90it/s] 23%|██▎ | 10939/48008 [37:17<2:03:19, 5.01it/s] 23%|██▎ | 10940/48008 [37:17<2:03:03, 5.02it/s] 23%|██▎ | 10941/48008 [37:17<2:03:34, 5.00it/s] 23%|██▎ | 10942/48008 [37:17<2:03:48, 4.99it/s] 23%|██▎ | 10943/48008 [37:17<2:01:14, 5.10it/s] 23%|██▎ | 10944/48008 [37:18<2:01:05, 5.10it/s] 23%|██▎ | 10945/48008 [37:18<2:00:08, 5.14it/s] 23%|██▎ | 10946/48008 [37:18<1:58:35, 5.21it/s] 23%|██▎ | 10947/48008 [37:18<1:57:42, 5.25it/s] 23%|██▎ | 10948/48008 [37:18<1:57:37, 5.25it/s] 23%|██▎ | 10949/48008 [37:19<1:57:03, 5.28it/s] 23%|██▎ | 10950/48008 [37:19<1:56:38, 5.30it/s] {'loss': 0.3019, 'grad_norm': 0.220703125, 'learning_rate': 0.0001543867688718547, 'epoch': 0.23} 23%|██▎ | 10950/48008 [37:19<1:56:38, 5.30it/s] 23%|██▎ | 10951/48008 [37:19<1:56:22, 5.31it/s] 23%|██▎ | 10952/48008 [37:19<1:57:05, 5.27it/s] 23%|██▎ | 10953/48008 [37:19<1:57:24, 5.26it/s] 23%|██▎ | 10954/48008 [37:19<1:57:26, 5.26it/s] 23%|██▎ | 10955/48008 [37:20<1:57:42, 5.25it/s] 23%|██▎ | 10956/48008 [37:20<1:57:43, 5.25it/s] 23%|██▎ | 10957/48008 [37:20<2:34:06, 4.01it/s] 23%|██▎ | 10958/48008 [37:20<2:22:22, 4.34it/s] 23%|██▎ | 10959/48008 [37:21<2:14:58, 4.57it/s] 23%|██▎ | 10960/48008 [37:21<2:09:39, 4.76it/s] 23%|██▎ | 10961/48008 [37:21<2:08:07, 4.82it/s] 23%|██▎ | 10962/48008 [37:21<2:04:41, 4.95it/s] 23%|██▎ | 10963/48008 [37:21<2:01:49, 5.07it/s] 23%|██▎ | 10964/48008 [37:22<1:58:31, 5.21it/s] 23%|██▎ | 10965/48008 [37:22<1:58:05, 5.23it/s] 23%|██▎ | 10966/48008 [37:22<1:57:02, 5.27it/s] 23%|██▎ | 10967/48008 [37:22<1:57:00, 5.28it/s] 23%|██▎ | 10968/48008 [37:22<1:56:50, 5.28it/s] 23%|██▎ | 10969/48008 [37:23<1:56:47, 5.29it/s] 23%|██▎ | 10970/48008 [37:23<1:56:32, 5.30it/s] 23%|██▎ | 10971/48008 [37:23<1:57:10, 5.27it/s] 23%|██▎ | 10972/48008 [37:23<1:57:15, 5.26it/s] 23%|██▎ | 10973/48008 [37:23<1:56:43, 5.29it/s] 23%|██▎ | 10974/48008 [37:23<1:57:05, 5.27it/s] 23%|██▎ | 10975/48008 [37:24<2:05:48, 4.91it/s] 23%|██▎ | 10976/48008 [37:24<2:24:37, 4.27it/s] 23%|██▎ | 10977/48008 [37:24<2:16:44, 4.51it/s] 23%|██▎ | 10978/48008 [37:24<2:13:23, 4.63it/s] 23%|██▎ | 10979/48008 [37:25<2:09:06, 4.78it/s] 23%|██▎ | 10980/48008 [37:25<2:07:56, 4.82it/s] 23%|██▎ | 10981/48008 [37:25<2:04:37, 4.95it/s] 23%|██▎ | 10982/48008 [37:25<2:02:02, 5.06it/s] 23%|██▎ | 10983/48008 [37:25<2:02:05, 5.05it/s] 23%|██▎ | 10984/48008 [37:26<2:00:33, 5.12it/s] 23%|██▎ | 10985/48008 [37:26<2:00:13, 5.13it/s] 23%|██▎ | 10986/48008 [37:26<1:58:35, 5.20it/s] 23%|██▎ | 10987/48008 [37:26<1:58:08, 5.22it/s] 23%|██▎ | 10988/48008 [37:26<2:06:22, 4.88it/s] 23%|██▎ | 10989/48008 [37:27<2:02:31, 5.04it/s] 23%|██▎ | 10990/48008 [37:27<2:09:48, 4.75it/s] 23%|██▎ | 10991/48008 [37:27<2:05:30, 4.92it/s] 23%|██▎ | 10992/48008 [37:27<2:03:15, 5.01it/s] 23%|██▎ | 10993/48008 [37:27<2:01:40, 5.07it/s] 23%|██▎ | 10994/48008 [37:28<1:59:25, 5.17it/s] 23%|██▎ | 10995/48008 [37:28<2:00:58, 5.10it/s] 23%|██▎ | 10996/48008 [37:28<2:00:45, 5.11it/s] 23%|██▎ | 10997/48008 [37:28<1:59:37, 5.16it/s] 23%|██▎ | 10998/48008 [37:28<1:58:47, 5.19it/s] 23%|██▎ | 10999/48008 [37:29<1:57:37, 5.24it/s] 23%|██▎ | 11000/48008 [37:29<1:57:27, 5.25it/s] {'loss': 0.3055, 'grad_norm': 0.283203125, 'learning_rate': 0.00015417847025495752, 'epoch': 0.23} 23%|██▎ | 11000/48008 [37:29<1:57:27, 5.25it/s] 23%|██▎ | 11001/48008 [37:29<1:57:51, 5.23it/s] 23%|██▎ | 11002/48008 [37:29<1:58:37, 5.20it/s] 23%|██▎ | 11003/48008 [37:29<2:00:35, 5.11it/s] 23%|██▎ | 11004/48008 [37:29<1:59:47, 5.15it/s] 23%|██▎ | 11005/48008 [37:30<1:58:51, 5.19it/s] 23%|██▎ | 11006/48008 [37:30<1:58:42, 5.20it/s] 23%|██▎ | 11007/48008 [37:31<4:04:21, 2.52it/s] 23%|██▎ | 11008/48008 [37:31<3:26:06, 2.99it/s] 23%|██▎ | 11009/48008 [37:31<2:58:59, 3.45it/s] 23%|██▎ | 11010/48008 [37:31<2:39:31, 3.87it/s] 23%|██▎ | 11011/48008 [37:31<2:26:47, 4.20it/s] 23%|██▎ | 11012/48008 [37:32<2:17:05, 4.50it/s] 23%|██▎ | 11013/48008 [37:32<2:13:34, 4.62it/s] 23%|██▎ | 11014/48008 [37:32<2:30:16, 4.10it/s] 23%|██▎ | 11015/48008 [37:32<2:20:53, 4.38it/s] 23%|██▎ | 11016/48008 [37:33<2:15:57, 4.53it/s] 23%|██▎ | 11017/48008 [37:33<2:10:19, 4.73it/s] 23%|██▎ | 11018/48008 [37:33<2:06:44, 4.86it/s] 23%|██▎ | 11019/48008 [37:33<2:03:55, 4.97it/s] 23%|██▎ | 11020/48008 [37:33<2:05:59, 4.89it/s] 23%|██▎ | 11021/48008 [37:34<2:02:50, 5.02it/s] 23%|██▎ | 11022/48008 [37:34<2:00:50, 5.10it/s] 23%|██▎ | 11023/48008 [37:34<1:59:03, 5.18it/s] 23%|██▎ | 11024/48008 [37:34<1:57:49, 5.23it/s] 23%|██▎ | 11025/48008 [37:34<1:59:03, 5.18it/s] 23%|██▎ | 11026/48008 [37:34<1:58:18, 5.21it/s] 23%|██▎ | 11027/48008 [37:35<1:57:21, 5.25it/s] 23%|██▎ | 11028/48008 [37:35<1:57:00, 5.27it/s] 23%|██▎ | 11029/48008 [37:35<2:18:22, 4.45it/s] 23%|██▎ | 11030/48008 [37:35<2:11:09, 4.70it/s] 23%|██▎ | 11031/48008 [37:36<2:07:05, 4.85it/s] 23%|██▎ | 11032/48008 [37:36<2:06:27, 4.87it/s] 23%|██▎ | 11033/48008 [37:36<2:25:17, 4.24it/s] 23%|██▎ | 11034/48008 [37:36<2:17:13, 4.49it/s] 23%|██▎ | 11035/48008 [37:36<2:20:10, 4.40it/s] 23%|██▎ | 11036/48008 [37:37<2:12:43, 4.64it/s] 23%|██▎ | 11037/48008 [37:37<2:10:18, 4.73it/s] 23%|██▎ | 11038/48008 [37:37<2:09:20, 4.76it/s] 23%|██▎ | 11039/48008 [37:37<2:05:46, 4.90it/s] 23%|██▎ | 11040/48008 [37:37<2:02:37, 5.02it/s] 23%|██▎ | 11041/48008 [37:38<2:01:17, 5.08it/s] 23%|██▎ | 11042/48008 [37:38<2:01:40, 5.06it/s] 23%|██▎ | 11043/48008 [37:38<2:01:59, 5.05it/s] 23%|██▎ | 11044/48008 [37:38<2:01:12, 5.08it/s] 23%|██▎ | 11045/48008 [37:38<1:59:21, 5.16it/s] 23%|██▎ | 11046/48008 [37:39<2:00:03, 5.13it/s] 23%|██▎ | 11047/48008 [37:39<1:58:49, 5.18it/s] 23%|██▎ | 11048/48008 [37:39<1:58:11, 5.21it/s] 23%|██▎ | 11049/48008 [37:39<1:57:39, 5.24it/s] 23%|██▎ | 11050/48008 [37:39<1:57:34, 5.24it/s] {'loss': 0.3026, 'grad_norm': 0.18359375, 'learning_rate': 0.00015397017163806034, 'epoch': 0.23} 23%|██▎ | 11050/48008 [37:39<1:57:34, 5.24it/s] 23%|██▎ | 11051/48008 [37:40<1:58:13, 5.21it/s] 23%|██▎ | 11052/48008 [37:40<1:58:20, 5.20it/s] 23%|██▎ | 11053/48008 [37:40<1:57:46, 5.23it/s] 23%|██▎ | 11054/48008 [37:40<1:57:29, 5.24it/s] 23%|██▎ | 11055/48008 [37:40<2:06:03, 4.89it/s] 23%|██▎ | 11056/48008 [37:41<2:02:41, 5.02it/s] 23%|██▎ | 11057/48008 [37:41<2:22:40, 4.32it/s] 23%|██▎ | 11058/48008 [37:41<2:15:36, 4.54it/s] 23%|██▎ | 11059/48008 [37:41<2:10:10, 4.73it/s] 23%|██▎ | 11060/48008 [37:41<2:06:01, 4.89it/s] 23%|██▎ | 11061/48008 [37:42<2:02:44, 5.02it/s] 23%|██▎ | 11062/48008 [37:42<2:00:30, 5.11it/s] 23%|██▎ | 11063/48008 [37:42<2:08:02, 4.81it/s] 23%|██▎ | 11064/48008 [37:42<2:04:18, 4.95it/s] 23%|██▎ | 11065/48008 [37:42<2:01:30, 5.07it/s] 23%|██▎ | 11066/48008 [37:43<2:00:27, 5.11it/s] 23%|██▎ | 11067/48008 [37:43<1:59:32, 5.15it/s] 23%|██▎ | 11068/48008 [37:43<1:58:02, 5.22it/s] 23%|██▎ | 11069/48008 [37:43<1:58:07, 5.21it/s] 23%|██▎ | 11070/48008 [37:43<1:57:56, 5.22it/s] 23%|██▎ | 11071/48008 [37:44<1:56:30, 5.28it/s] 23%|██▎ | 11072/48008 [37:44<1:56:42, 5.27it/s] 23%|██▎ | 11073/48008 [37:44<1:56:40, 5.28it/s] 23%|██▎ | 11074/48008 [37:44<1:56:59, 5.26it/s] 23%|██▎ | 11075/48008 [37:44<1:57:04, 5.26it/s] 23%|██▎ | 11076/48008 [37:45<1:59:43, 5.14it/s] 23%|██▎ | 11077/48008 [37:45<1:58:42, 5.19it/s] 23%|██▎ | 11078/48008 [37:45<2:00:31, 5.11it/s] 23%|██▎ | 11079/48008 [37:45<2:35:37, 3.95it/s] 23%|██▎ | 11080/48008 [37:46<2:24:10, 4.27it/s] 23%|██▎ | 11081/48008 [37:46<2:24:42, 4.25it/s] 23%|██▎ | 11082/48008 [37:46<2:16:50, 4.50it/s] 23%|██▎ | 11083/48008 [37:46<2:10:24, 4.72it/s] 23%|██▎ | 11084/48008 [37:46<2:07:54, 4.81it/s] 23%|██▎ | 11085/48008 [37:47<2:03:50, 4.97it/s] 23%|██▎ | 11086/48008 [37:47<2:02:04, 5.04it/s] 23%|██▎ | 11087/48008 [37:47<2:36:35, 3.93it/s] 23%|██▎ | 11088/48008 [37:47<2:27:23, 4.17it/s] 23%|██▎ | 11089/48008 [37:47<2:18:13, 4.45it/s] 23%|██▎ | 11090/48008 [37:48<2:13:27, 4.61it/s] 23%|██▎ | 11091/48008 [37:48<2:10:01, 4.73it/s] 23%|██▎ | 11092/48008 [37:48<2:07:39, 4.82it/s] 23%|██▎ | 11093/48008 [37:48<2:12:53, 4.63it/s] 23%|██▎ | 11094/48008 [37:49<2:08:32, 4.79it/s] 23%|██▎ | 11095/48008 [37:49<2:08:46, 4.78it/s] 23%|██▎ | 11096/48008 [37:49<2:04:39, 4.93it/s] 23%|██▎ | 11097/48008 [37:49<2:03:46, 4.97it/s] 23%|██▎ | 11098/48008 [37:49<2:02:07, 5.04it/s] 23%|██▎ | 11099/48008 [37:49<2:01:37, 5.06it/s] 23%|██▎ | 11100/48008 [37:50<2:08:47, 4.78it/s] {'loss': 0.3042, 'grad_norm': 0.1611328125, 'learning_rate': 0.00015376187302116313, 'epoch': 0.23} 23%|██▎ | 11100/48008 [37:50<2:08:47, 4.78it/s] 23%|██▎ | 11101/48008 [37:50<2:05:18, 4.91it/s] 23%|██▎ | 11102/48008 [37:50<2:06:20, 4.87it/s] 23%|██▎ | 11103/48008 [37:50<2:03:10, 4.99it/s] 23%|██▎ | 11104/48008 [37:51<2:00:44, 5.09it/s] 23%|██▎ | 11105/48008 [37:51<1:58:59, 5.17it/s] 23%|██▎ | 11106/48008 [37:51<1:58:14, 5.20it/s] 23%|██▎ | 11107/48008 [37:51<1:57:45, 5.22it/s] 23%|██▎ | 11108/48008 [37:51<2:33:46, 4.00it/s] 23%|██▎ | 11109/48008 [37:52<2:58:52, 3.44it/s] 23%|██▎ | 11110/48008 [37:52<2:40:23, 3.83it/s] 23%|██▎ | 11111/48008 [37:52<2:27:18, 4.17it/s] 23%|██▎ | 11112/48008 [37:52<2:18:48, 4.43it/s] 23%|██▎ | 11113/48008 [37:53<2:11:17, 4.68it/s] 23%|██▎ | 11114/48008 [37:53<2:08:58, 4.77it/s] 23%|██▎ | 11115/48008 [37:53<2:08:44, 4.78it/s] 23%|██▎ | 11116/48008 [37:53<2:04:14, 4.95it/s] 23%|██▎ | 11117/48008 [37:53<2:02:26, 5.02it/s] 23%|██▎ | 11118/48008 [37:54<2:09:26, 4.75it/s] 23%|██▎ | 11119/48008 [37:54<2:05:44, 4.89it/s] 23%|██▎ | 11120/48008 [37:54<2:06:34, 4.86it/s] 23%|██▎ | 11121/48008 [37:54<2:04:11, 4.95it/s] 23%|██▎ | 11122/48008 [37:54<2:02:58, 5.00it/s] 23%|██▎ | 11123/48008 [37:55<2:00:50, 5.09it/s] 23%|██▎ | 11124/48008 [37:55<2:02:05, 5.04it/s] 23%|██▎ | 11125/48008 [37:55<1:59:52, 5.13it/s] 23%|██▎ | 11126/48008 [37:55<1:59:54, 5.13it/s] 23%|██▎ | 11127/48008 [37:55<1:59:01, 5.16it/s] 23%|██▎ | 11128/48008 [37:56<2:00:52, 5.09it/s] 23%|██▎ | 11129/48008 [37:56<1:59:47, 5.13it/s] 23%|██▎ | 11130/48008 [37:56<2:00:30, 5.10it/s] 23%|██▎ | 11131/48008 [37:56<2:00:51, 5.09it/s] 23%|██▎ | 11132/48008 [37:56<1:59:18, 5.15it/s] 23%|██▎ | 11133/48008 [37:57<1:57:50, 5.22it/s] 23%|██▎ | 11134/48008 [37:57<1:57:40, 5.22it/s] 23%|██▎ | 11135/48008 [37:57<1:57:49, 5.22it/s] 23%|██▎ | 11136/48008 [37:57<1:57:37, 5.22it/s] 23%|██▎ | 11137/48008 [37:57<1:57:12, 5.24it/s] 23%|██▎ | 11138/48008 [37:58<1:59:16, 5.15it/s] 23%|██▎ | 11139/48008 [37:58<1:59:28, 5.14it/s] 23%|██▎ | 11140/48008 [37:58<2:00:43, 5.09it/s] 23%|██▎ | 11141/48008 [37:58<2:01:43, 5.05it/s] 23%|██▎ | 11142/48008 [37:58<2:08:53, 4.77it/s] 23%|██▎ | 11143/48008 [37:59<2:05:27, 4.90it/s] 23%|██▎ | 11144/48008 [37:59<2:02:45, 5.00it/s] 23%|██▎ | 11145/48008 [37:59<2:01:51, 5.04it/s] 23%|██▎ | 11146/48008 [37:59<2:00:21, 5.10it/s] 23%|██▎ | 11147/48008 [37:59<2:20:28, 4.37it/s] 23%|██▎ | 11148/48008 [38:00<2:15:39, 4.53it/s] 23%|██▎ | 11149/48008 [38:00<2:12:18, 4.64it/s] 23%|██▎ | 11150/48008 [38:00<2:07:24, 4.82it/s] {'loss': 0.3032, 'grad_norm': 0.33203125, 'learning_rate': 0.00015355357440426597, 'epoch': 0.23} 23%|██▎ | 11150/48008 [38:00<2:07:24, 4.82it/s] 23%|██▎ | 11151/48008 [38:00<2:05:33, 4.89it/s] 23%|██▎ | 11152/48008 [38:00<2:05:13, 4.91it/s] 23%|██▎ | 11153/48008 [38:01<2:01:43, 5.05it/s] 23%|██▎ | 11154/48008 [38:01<1:59:30, 5.14it/s] 23%|██▎ | 11155/48008 [38:01<1:59:04, 5.16it/s] 23%|██▎ | 11156/48008 [38:01<1:59:53, 5.12it/s] 23%|██▎ | 11157/48008 [38:01<1:58:56, 5.16it/s] 23%|██▎ | 11158/48008 [38:02<1:57:58, 5.21it/s] 23%|██▎ | 11159/48008 [38:02<1:57:52, 5.21it/s] 23%|██▎ | 11160/48008 [38:02<2:00:17, 5.11it/s] 23%|██▎ | 11161/48008 [38:02<1:58:24, 5.19it/s] 23%|██▎ | 11162/48008 [38:02<1:57:15, 5.24it/s] 23%|██▎ | 11163/48008 [38:03<1:57:18, 5.23it/s] 23%|██▎ | 11164/48008 [38:03<1:56:47, 5.26it/s] 23%|██▎ | 11165/48008 [38:03<1:59:55, 5.12it/s] 23%|██▎ | 11166/48008 [38:03<2:20:31, 4.37it/s] 23%|██▎ | 11167/48008 [38:03<2:13:48, 4.59it/s] 23%|██▎ | 11168/48008 [38:04<2:09:18, 4.75it/s] 23%|██▎ | 11169/48008 [38:04<2:26:59, 4.18it/s] 23%|██▎ | 11170/48008 [38:04<2:18:19, 4.44it/s] 23%|██▎ | 11171/48008 [38:04<2:11:12, 4.68it/s] 23%|██▎ | 11172/48008 [38:04<2:07:01, 4.83it/s] 23%|██▎ | 11173/48008 [38:05<2:05:16, 4.90it/s] 23%|██▎ | 11174/48008 [38:05<2:02:38, 5.01it/s] 23%|██▎ | 11175/48008 [38:05<2:01:38, 5.05it/s] 23%|██▎ | 11176/48008 [38:05<2:00:23, 5.10it/s] 23%|██▎ | 11177/48008 [38:05<2:00:35, 5.09it/s] 23%|██▎ | 11178/48008 [38:06<1:59:55, 5.12it/s] 23%|██▎ | 11179/48008 [38:06<1:59:14, 5.15it/s] 23%|██▎ | 11180/48008 [38:06<2:00:50, 5.08it/s] 23%|██▎ | 11181/48008 [38:06<2:02:10, 5.02it/s] 23%|██▎ | 11182/48008 [38:06<1:59:40, 5.13it/s] 23%|██▎ | 11183/48008 [38:07<2:19:58, 4.38it/s] 23%|██▎ | 11184/48008 [38:07<2:13:58, 4.58it/s] 23%|██▎ | 11185/48008 [38:07<2:08:28, 4.78it/s] 23%|██▎ | 11186/48008 [38:07<2:05:10, 4.90it/s] 23%|██▎ | 11187/48008 [38:07<2:03:07, 4.98it/s] 23%|██▎ | 11188/48008 [38:08<2:00:58, 5.07it/s] 23%|██▎ | 11189/48008 [38:08<2:00:28, 5.09it/s] 23%|██▎ | 11190/48008 [38:08<1:59:10, 5.15it/s] 23%|██▎ | 11191/48008 [38:08<1:57:18, 5.23it/s] 23%|██▎ | 11192/48008 [38:08<1:59:17, 5.14it/s] 23%|██▎ | 11193/48008 [38:09<2:07:18, 4.82it/s] 23%|██▎ | 11194/48008 [38:09<2:03:53, 4.95it/s] 23%|██▎ | 11195/48008 [38:09<2:04:09, 4.94it/s] 23%|██▎ | 11196/48008 [38:09<2:01:14, 5.06it/s] 23%|██▎ | 11197/48008 [38:09<1:59:28, 5.14it/s] 23%|██▎ | 11198/48008 [38:10<1:58:48, 5.16it/s] 23%|██▎ | 11199/48008 [38:10<2:00:31, 5.09it/s] 23%|██▎ | 11200/48008 [38:10<1:59:22, 5.14it/s] {'loss': 0.3001, 'grad_norm': 0.359375, 'learning_rate': 0.00015334527578736876, 'epoch': 0.23} 23%|██▎ | 11200/48008 [38:10<1:59:22, 5.14it/s] 23%|██▎ | 11201/48008 [38:10<1:59:18, 5.14it/s] 23%|██▎ | 11202/48008 [38:10<1:59:11, 5.15it/s] 23%|██▎ | 11203/48008 [38:11<1:57:40, 5.21it/s] 23%|██▎ | 11204/48008 [38:11<1:57:03, 5.24it/s] 23%|██▎ | 11205/48008 [38:11<1:57:09, 5.24it/s] 23%|██▎ | 11206/48008 [38:11<2:05:31, 4.89it/s] 23%|██▎ | 11207/48008 [38:11<2:04:05, 4.94it/s] 23%|██▎ | 11208/48008 [38:12<2:04:31, 4.93it/s] 23%|██▎ | 11209/48008 [38:12<2:03:07, 4.98it/s] 23%|██▎ | 11210/48008 [38:12<2:03:25, 4.97it/s] 23%|██▎ | 11211/48008 [38:12<2:01:10, 5.06it/s] 23%|██▎ | 11212/48008 [38:12<1:59:51, 5.12it/s] 23%|██▎ | 11213/48008 [38:13<1:59:22, 5.14it/s] 23%|██▎ | 11214/48008 [38:13<1:58:42, 5.17it/s] 23%|██▎ | 11215/48008 [38:13<2:00:25, 5.09it/s] 23%|██▎ | 11216/48008 [38:13<2:07:45, 4.80it/s] 23%|██▎ | 11217/48008 [38:13<2:04:54, 4.91it/s] 23%|██▎ | 11218/48008 [38:14<2:02:51, 4.99it/s] 23%|██▎ | 11219/48008 [38:14<2:03:01, 4.98it/s] 23%|██▎ | 11220/48008 [38:14<2:01:16, 5.06it/s] 23%|██▎ | 11221/48008 [38:14<1:59:57, 5.11it/s] 23%|██▎ | 11222/48008 [38:14<2:01:30, 5.05it/s] 23%|██▎ | 11223/48008 [38:15<2:00:08, 5.10it/s] 23%|██▎ | 11224/48008 [38:15<1:58:07, 5.19it/s] 23%|██▎ | 11225/48008 [38:15<1:59:32, 5.13it/s] 23%|██▎ | 11226/48008 [38:15<1:58:12, 5.19it/s] 23%|██▎ | 11227/48008 [38:15<2:06:28, 4.85it/s] 23%|██▎ | 11228/48008 [38:16<2:05:06, 4.90it/s] 23%|██▎ | 11229/48008 [38:16<2:05:08, 4.90it/s] 23%|██▎ | 11230/48008 [38:16<2:02:28, 5.00it/s] 23%|██▎ | 11231/48008 [38:16<2:00:57, 5.07it/s] 23%|██▎ | 11232/48008 [38:16<2:01:53, 5.03it/s] 23%|██▎ | 11233/48008 [38:17<2:00:20, 5.09it/s] 23%|██▎ | 11234/48008 [38:17<1:59:31, 5.13it/s] 23%|██▎ | 11235/48008 [38:17<1:58:12, 5.18it/s] 23%|██▎ | 11236/48008 [38:17<2:01:17, 5.05it/s] 23%|██▎ | 11237/48008 [38:17<2:00:28, 5.09it/s] 23%|██▎ | 11238/48008 [38:18<1:59:13, 5.14it/s] 23%|██▎ | 11239/48008 [38:18<1:59:34, 5.13it/s] 23%|██▎ | 11240/48008 [38:18<1:58:53, 5.15it/s] 23%|██▎ | 11241/48008 [38:18<1:58:28, 5.17it/s] 23%|██▎ | 11242/48008 [38:18<2:01:24, 5.05it/s] 23%|██▎ | 11243/48008 [38:19<1:59:56, 5.11it/s] 23%|██▎ | 11244/48008 [38:19<1:58:56, 5.15it/s] 23%|██▎ | 11245/48008 [38:19<1:57:11, 5.23it/s] 23%|██▎ | 11246/48008 [38:19<2:00:25, 5.09it/s] 23%|██▎ | 11247/48008 [38:19<2:00:43, 5.08it/s] 23%|██▎ | 11248/48008 [38:20<1:59:40, 5.12it/s] 23%|██▎ | 11249/48008 [38:20<1:58:48, 5.16it/s] 23%|██▎ | 11250/48008 [38:20<1:57:52, 5.20it/s] {'loss': 0.2998, 'grad_norm': 0.185546875, 'learning_rate': 0.0001531369771704716, 'epoch': 0.23} 23%|██▎ | 11250/48008 [38:20<1:57:52, 5.20it/s] 23%|██▎ | 11251/48008 [38:20<1:57:01, 5.23it/s] 23%|██▎ | 11252/48008 [38:20<1:57:04, 5.23it/s] 23%|██▎ | 11253/48008 [38:20<1:56:54, 5.24it/s] 23%|██▎ | 11254/48008 [38:21<1:57:52, 5.20it/s] 23%|██▎ | 11255/48008 [38:21<1:56:42, 5.25it/s] 23%|██▎ | 11256/48008 [38:21<1:56:36, 5.25it/s] 23%|██▎ | 11257/48008 [38:21<1:56:20, 5.26it/s] 23%|██▎ | 11258/48008 [38:21<1:55:43, 5.29it/s] 23%|██▎ | 11259/48008 [38:22<1:55:59, 5.28it/s] 23%|██▎ | 11260/48008 [38:22<1:55:59, 5.28it/s] 23%|██▎ | 11261/48008 [38:22<1:55:58, 5.28it/s] 23%|██▎ | 11262/48008 [38:22<1:55:51, 5.29it/s] 23%|██▎ | 11263/48008 [38:22<2:04:52, 4.90it/s] 23%|██▎ | 11264/48008 [38:23<2:02:08, 5.01it/s] 23%|██▎ | 11265/48008 [38:23<2:00:13, 5.09it/s] 23%|██▎ | 11266/48008 [38:23<1:59:02, 5.14it/s] 23%|██▎ | 11267/48008 [38:23<1:58:19, 5.18it/s] 23%|██▎ | 11268/48008 [38:23<1:57:29, 5.21it/s] 23%|██▎ | 11269/48008 [38:24<1:59:17, 5.13it/s] 23%|██▎ | 11270/48008 [38:24<1:58:08, 5.18it/s] 23%|██▎ | 11271/48008 [38:24<1:57:49, 5.20it/s] 23%|██▎ | 11272/48008 [38:24<1:56:57, 5.23it/s] 23%|██▎ | 11273/48008 [38:24<1:56:42, 5.25it/s] 23%|██▎ | 11274/48008 [38:24<1:55:31, 5.30it/s] 23%|██▎ | 11275/48008 [38:25<1:55:31, 5.30it/s] 23%|██▎ | 11276/48008 [38:25<1:57:12, 5.22it/s] 23%|██▎ | 11277/48008 [38:25<1:56:52, 5.24it/s] 23%|██▎ | 11278/48008 [38:25<1:59:39, 5.12it/s] 23%|██▎ | 11279/48008 [38:25<2:02:18, 5.00it/s] 23%|██▎ | 11280/48008 [38:26<2:00:41, 5.07it/s] 23%|██▎ | 11281/48008 [38:26<1:59:27, 5.12it/s] 24%|██▎ | 11282/48008 [38:26<1:59:32, 5.12it/s] 24%|██▎ | 11283/48008 [38:26<1:58:44, 5.15it/s] 24%|██▎ | 11284/48008 [38:26<1:57:48, 5.20it/s] 24%|██▎ | 11285/48008 [38:27<1:57:07, 5.23it/s] 24%|██▎ | 11286/48008 [38:27<1:57:00, 5.23it/s] 24%|██▎ | 11287/48008 [38:27<2:00:15, 5.09it/s] 24%|██▎ | 11288/48008 [38:27<1:59:01, 5.14it/s] 24%|██▎ | 11289/48008 [38:27<1:59:14, 5.13it/s] 24%|██▎ | 11290/48008 [38:28<2:00:47, 5.07it/s] 24%|██▎ | 11291/48008 [38:28<1:59:08, 5.14it/s] 24%|██▎ | 11292/48008 [38:28<1:58:31, 5.16it/s] 24%|██▎ | 11293/48008 [38:28<2:00:05, 5.10it/s] 24%|██▎ | 11294/48008 [38:28<1:59:27, 5.12it/s] 24%|██▎ | 11295/48008 [38:29<1:58:18, 5.17it/s] 24%|██▎ | 11296/48008 [38:29<1:57:05, 5.23it/s] 24%|██▎ | 11297/48008 [38:29<1:56:55, 5.23it/s] 24%|██▎ | 11298/48008 [38:29<1:57:06, 5.22it/s] 24%|██▎ | 11299/48008 [38:29<1:56:02, 5.27it/s] 24%|██▎ | 11300/48008 [38:30<1:56:12, 5.26it/s] 24%|██▎ | 11300/48008 [38:30<1:56:12, 5.26it/s]{'loss': 0.305, 'grad_norm': 0.275390625, 'learning_rate': 0.00015292867855357442, 'epoch': 0.24} 24%|██▎ | 11301/48008 [38:30<1:56:24, 5.26it/s] 24%|██▎ | 11302/48008 [38:30<1:56:26, 5.25it/s] 24%|██▎ | 11303/48008 [38:30<1:56:25, 5.25it/s] 24%|██▎ | 11304/48008 [38:30<1:56:05, 5.27it/s] 24%|██▎ | 11305/48008 [38:30<1:57:41, 5.20it/s] 24%|██▎ | 11306/48008 [38:31<1:58:35, 5.16it/s] 24%|██▎ | 11307/48008 [38:31<1:57:49, 5.19it/s] 24%|██▎ | 11308/48008 [38:31<1:57:06, 5.22it/s] 24%|██▎ | 11309/48008 [38:31<1:58:11, 5.17it/s] 24%|██▎ | 11310/48008 [38:31<1:57:58, 5.18it/s] 24%|██▎ | 11311/48008 [38:32<1:57:38, 5.20it/s] 24%|██▎ | 11312/48008 [38:32<1:57:17, 5.21it/s] 24%|██▎ | 11313/48008 [38:32<1:57:14, 5.22it/s] 24%|██▎ | 11314/48008 [38:32<2:18:17, 4.42it/s] 24%|██▎ | 11315/48008 [38:33<2:11:40, 4.64it/s] 24%|██▎ | 11316/48008 [38:33<2:07:13, 4.81it/s] 24%|██▎ | 11317/48008 [38:33<2:03:27, 4.95it/s] 24%|██▎ | 11318/48008 [38:33<2:01:06, 5.05it/s] 24%|██▎ | 11319/48008 [38:33<2:00:02, 5.09it/s] 24%|██▎ | 11320/48008 [38:33<1:58:52, 5.14it/s] 24%|██▎ | 11321/48008 [38:34<1:58:01, 5.18it/s] 24%|██▎ | 11322/48008 [38:34<1:57:05, 5.22it/s] 24%|██▎ | 11323/48008 [38:34<1:57:09, 5.22it/s] 24%|██▎ | 11324/48008 [38:34<1:57:12, 5.22it/s] 24%|██▎ | 11325/48008 [38:34<1:55:57, 5.27it/s] 24%|██▎ | 11326/48008 [38:35<1:56:20, 5.25it/s] 24%|██▎ | 11327/48008 [38:35<1:56:32, 5.25it/s] 24%|██▎ | 11328/48008 [38:35<1:56:19, 5.26it/s] 24%|██▎ | 11329/48008 [38:35<1:58:01, 5.18it/s] 24%|██▎ | 11330/48008 [38:35<1:57:36, 5.20it/s] 24%|██▎ | 11331/48008 [38:36<1:56:09, 5.26it/s] 24%|██▎ | 11332/48008 [38:36<1:55:35, 5.29it/s] 24%|██▎ | 11333/48008 [38:36<1:56:48, 5.23it/s] 24%|██▎ | 11334/48008 [38:36<1:57:57, 5.18it/s] 24%|██▎ | 11335/48008 [38:36<1:57:18, 5.21it/s] 24%|██▎ | 11336/48008 [38:37<1:56:46, 5.23it/s] 24%|██▎ | 11337/48008 [38:37<1:56:40, 5.24it/s] 24%|██▎ | 11338/48008 [38:37<1:56:32, 5.24it/s] 24%|██▎ | 11339/48008 [38:37<1:56:23, 5.25it/s] 24%|██▎ | 11340/48008 [38:37<1:56:37, 5.24it/s] 24%|██▎ | 11341/48008 [38:37<1:55:10, 5.31it/s] 24%|██▎ | 11342/48008 [38:38<1:55:37, 5.29it/s] 24%|██▎ | 11343/48008 [38:38<1:55:21, 5.30it/s] 24%|██▎ | 11344/48008 [38:38<1:55:26, 5.29it/s] 24%|██▎ | 11345/48008 [38:38<1:57:46, 5.19it/s] 24%|██▎ | 11346/48008 [38:38<1:59:40, 5.11it/s] 24%|██▎ | 11347/48008 [38:39<2:06:59, 4.81it/s] 24%|██▎ | 11348/48008 [38:39<2:05:54, 4.85it/s] 24%|██▎ | 11349/48008 [38:39<2:11:35, 4.64it/s] 24%|██▎ | 11350/48008 [38:39<2:07:06, 4.81it/s] {'loss': 0.3024, 'grad_norm': 0.3984375, 'learning_rate': 0.00015272037993667721, 'epoch': 0.24} 24%|██▎ | 11350/48008 [38:39<2:07:06, 4.81it/s] 24%|██▎ | 11351/48008 [38:40<2:04:00, 4.93it/s] 24%|██▎ | 11352/48008 [38:40<2:02:34, 4.98it/s] 24%|██▎ | 11353/48008 [38:40<1:58:48, 5.14it/s] 24%|██▎ | 11354/48008 [38:40<1:58:09, 5.17it/s] 24%|██▎ | 11355/48008 [38:40<2:01:03, 5.05it/s] 24%|██▎ | 11356/48008 [38:40<1:58:44, 5.14it/s] 24%|██▎ | 11357/48008 [38:41<1:57:10, 5.21it/s] 24%|██▎ | 11358/48008 [38:41<1:56:54, 5.23it/s] 24%|██▎ | 11359/48008 [38:41<1:56:49, 5.23it/s] 24%|██▎ | 11360/48008 [38:41<1:56:57, 5.22it/s] 24%|██▎ | 11361/48008 [38:41<1:57:11, 5.21it/s] 24%|██▎ | 11362/48008 [38:42<1:59:29, 5.11it/s] 24%|██▎ | 11363/48008 [38:42<1:59:38, 5.10it/s] 24%|██▎ | 11364/48008 [38:42<1:57:50, 5.18it/s] 24%|██▎ | 11365/48008 [38:42<1:59:22, 5.12it/s] 24%|██▎ | 11366/48008 [38:42<2:07:00, 4.81it/s] 24%|██▎ | 11367/48008 [38:43<2:03:33, 4.94it/s] 24%|██▎ | 11368/48008 [38:43<2:02:10, 5.00it/s] 24%|██▎ | 11369/48008 [38:43<2:00:12, 5.08it/s] 24%|██▎ | 11370/48008 [38:43<1:58:18, 5.16it/s] 24%|██▎ | 11371/48008 [38:43<1:57:38, 5.19it/s] 24%|██▎ | 11372/48008 [38:44<1:56:29, 5.24it/s] 24%|██▎ | 11373/48008 [38:44<1:57:31, 5.20it/s] 24%|██▎ | 11374/48008 [38:44<1:56:42, 5.23it/s] 24%|██▎ | 11375/48008 [38:44<1:57:23, 5.20it/s] 24%|██▎ | 11376/48008 [38:44<1:57:03, 5.22it/s] 24%|██▎ | 11377/48008 [38:45<1:56:45, 5.23it/s] 24%|██▎ | 11378/48008 [38:45<2:05:10, 4.88it/s] 24%|██▎ | 11379/48008 [38:45<2:02:08, 5.00it/s] 24%|██▎ | 11380/48008 [38:45<2:02:36, 4.98it/s] 24%|██▎ | 11381/48008 [38:45<2:00:53, 5.05it/s] 24%|██▎ | 11382/48008 [38:46<1:58:27, 5.15it/s] 24%|██▎ | 11383/48008 [38:46<1:57:11, 5.21it/s] 24%|██▎ | 11384/48008 [38:46<1:59:04, 5.13it/s] 24%|██▎ | 11385/48008 [38:46<1:59:48, 5.09it/s] 24%|██▎ | 11386/48008 [38:46<1:57:37, 5.19it/s] 24%|██▎ | 11387/48008 [38:47<1:57:00, 5.22it/s] 24%|██▎ | 11388/48008 [38:47<1:57:53, 5.18it/s] 24%|██▎ | 11389/48008 [38:47<2:33:14, 3.98it/s] 24%|██▎ | 11390/48008 [38:47<2:30:44, 4.05it/s] 24%|██▎ | 11391/48008 [38:48<2:23:49, 4.24it/s] 24%|██▎ | 11392/48008 [38:48<2:15:27, 4.51it/s] 24%|██▎ | 11393/48008 [38:48<2:09:04, 4.73it/s] 24%|██▎ | 11394/48008 [38:48<2:07:30, 4.79it/s] 24%|██▎ | 11395/48008 [38:48<2:04:19, 4.91it/s] 24%|██▎ | 11396/48008 [38:48<2:02:01, 5.00it/s] 24%|██▎ | 11397/48008 [38:49<1:59:29, 5.11it/s] 24%|██▎ | 11398/48008 [38:49<1:58:50, 5.13it/s] 24%|██▎ | 11399/48008 [38:49<2:01:48, 5.01it/s] 24%|██▎ | 11400/48008 [38:49<2:35:54, 3.91it/s] {'loss': 0.3042, 'grad_norm': 0.63671875, 'learning_rate': 0.00015251208131978006, 'epoch': 0.24} 24%|██▎ | 11400/48008 [38:49<2:35:54, 3.91it/s] 24%|██▎ | 11401/48008 [38:50<2:24:07, 4.23it/s] 24%|██▍ | 11402/48008 [38:50<2:18:02, 4.42it/s] 24%|██▍ | 11403/48008 [38:50<2:10:42, 4.67it/s] 24%|██▍ | 11404/48008 [38:50<2:06:16, 4.83it/s] 24%|██▍ | 11405/48008 [38:50<2:02:16, 4.99it/s] 24%|██▍ | 11406/48008 [38:51<2:01:39, 5.01it/s] 24%|██▍ | 11407/48008 [38:51<2:01:30, 5.02it/s] 24%|██▍ | 11408/48008 [38:51<1:59:40, 5.10it/s] 24%|██▍ | 11409/48008 [38:51<2:00:38, 5.06it/s] 24%|██▍ | 11410/48008 [38:51<1:58:45, 5.14it/s] 24%|██▍ | 11411/48008 [38:52<1:57:37, 5.19it/s] 24%|██▍ | 11412/48008 [38:52<2:00:24, 5.07it/s] 24%|██▍ | 11413/48008 [38:52<1:59:03, 5.12it/s] 24%|██▍ | 11414/48008 [38:52<2:01:30, 5.02it/s] 24%|██▍ | 11415/48008 [38:53<2:21:17, 4.32it/s] 24%|██▍ | 11416/48008 [38:53<2:13:44, 4.56it/s] 24%|██▍ | 11417/48008 [38:53<2:10:20, 4.68it/s] 24%|██▍ | 11418/48008 [38:53<2:04:23, 4.90it/s] 24%|██▍ | 11419/48008 [38:53<2:01:56, 5.00it/s] 24%|██▍ | 11420/48008 [38:53<2:01:09, 5.03it/s] 24%|██▍ | 11421/48008 [38:54<1:59:30, 5.10it/s] 24%|██▍ | 11422/48008 [38:54<2:02:15, 4.99it/s] 24%|██▍ | 11423/48008 [38:54<2:03:52, 4.92it/s] 24%|██▍ | 11424/48008 [38:54<2:02:48, 4.96it/s] 24%|██▍ | 11425/48008 [38:54<2:00:59, 5.04it/s] 24%|██▍ | 11426/48008 [38:55<1:59:29, 5.10it/s] 24%|██▍ | 11427/48008 [38:55<1:58:52, 5.13it/s] 24%|██▍ | 11428/48008 [38:56<4:20:22, 2.34it/s] 24%|██▍ | 11429/48008 [38:56<3:39:33, 2.78it/s] 24%|██▍ | 11430/48008 [38:56<3:08:48, 3.23it/s] 24%|██▍ | 11431/48008 [38:56<2:47:08, 3.65it/s] 24%|██▍ | 11432/48008 [38:57<2:33:48, 3.96it/s] 24%|██▍ | 11433/48008 [38:57<2:23:08, 4.26it/s] 24%|██▍ | 11434/48008 [38:58<4:26:16, 2.29it/s] 24%|██▍ | 11435/48008 [38:58<3:40:35, 2.76it/s] 24%|██▍ | 11436/48008 [38:58<3:09:10, 3.22it/s] 24%|██▍ | 11437/48008 [38:58<2:47:16, 3.64it/s] 24%|██▍ | 11438/48008 [38:58<2:31:54, 4.01it/s] 24%|██▍ | 11439/48008 [38:59<2:22:07, 4.29it/s] 24%|██▍ | 11440/48008 [38:59<2:16:27, 4.47it/s] 24%|██▍ | 11441/48008 [38:59<2:10:21, 4.68it/s] 24%|██▍ | 11442/48008 [38:59<2:04:54, 4.88it/s] 24%|██▍ | 11443/48008 [39:00<2:38:00, 3.86it/s] 24%|██▍ | 11444/48008 [39:00<2:24:51, 4.21it/s] 24%|██▍ | 11445/48008 [39:00<2:16:27, 4.47it/s] 24%|██▍ | 11446/48008 [39:01<5:09:49, 1.97it/s] 24%|██▍ | 11447/48008 [39:01<4:13:10, 2.41it/s] 24%|██▍ | 11448/48008 [39:02<3:35:19, 2.83it/s] 24%|██▍ | 11449/48008 [39:02<3:05:33, 3.28it/s] 24%|██▍ | 11450/48008 [39:02<2:44:57, 3.69it/s] {'loss': 0.3014, 'grad_norm': 0.1376953125, 'learning_rate': 0.00015230378270288285, 'epoch': 0.24} 24%|██▍ | 11450/48008 [39:02<2:44:57, 3.69it/s] 24%|██▍ | 11451/48008 [39:02<2:31:38, 4.02it/s] 24%|██▍ | 11452/48008 [39:02<2:21:24, 4.31it/s] 24%|██▍ | 11453/48008 [39:03<2:14:06, 4.54it/s] 24%|██▍ | 11454/48008 [39:03<2:08:49, 4.73it/s] 24%|██▍ | 11455/48008 [39:03<2:04:50, 4.88it/s] 24%|██▍ | 11456/48008 [39:03<2:01:26, 5.02it/s] 24%|██▍ | 11457/48008 [39:03<2:00:24, 5.06it/s] 24%|██▍ | 11458/48008 [39:03<1:59:42, 5.09it/s] 24%|██▍ | 11459/48008 [39:04<1:58:13, 5.15it/s] 24%|██▍ | 11460/48008 [39:04<1:57:26, 5.19it/s] 24%|██▍ | 11461/48008 [39:04<1:57:33, 5.18it/s] 24%|██▍ | 11462/48008 [39:04<1:56:12, 5.24it/s] 24%|██▍ | 11463/48008 [39:04<1:56:24, 5.23it/s] 24%|██▍ | 11464/48008 [39:05<3:47:47, 2.67it/s] 24%|██▍ | 11465/48008 [39:05<3:16:09, 3.10it/s] 24%|██▍ | 11466/48008 [39:06<2:52:42, 3.53it/s] 24%|██▍ | 11467/48008 [39:06<2:36:45, 3.88it/s] 24%|██▍ | 11468/48008 [39:06<2:59:56, 3.38it/s] 24%|██▍ | 11469/48008 [39:06<2:42:33, 3.75it/s] 24%|██▍ | 11470/48008 [39:07<2:28:45, 4.09it/s] 24%|██▍ | 11471/48008 [39:07<2:19:23, 4.37it/s] 24%|██▍ | 11472/48008 [39:07<2:21:02, 4.32it/s] 24%|██▍ | 11473/48008 [39:07<2:14:47, 4.52it/s] 24%|██▍ | 11474/48008 [39:07<2:10:18, 4.67it/s] 24%|██▍ | 11475/48008 [39:08<2:05:50, 4.84it/s] 24%|██▍ | 11476/48008 [39:08<2:03:03, 4.95it/s] 24%|██▍ | 11477/48008 [39:08<2:36:55, 3.88it/s] 24%|██▍ | 11478/48008 [39:08<2:26:00, 4.17it/s] 24%|██▍ | 11479/48008 [39:09<2:16:54, 4.45it/s] 24%|██▍ | 11480/48008 [39:09<2:10:11, 4.68it/s] 24%|██▍ | 11481/48008 [39:09<2:27:33, 4.13it/s] 24%|██▍ | 11482/48008 [39:09<2:19:27, 4.37it/s] 24%|██▍ | 11483/48008 [39:09<2:11:13, 4.64it/s] 24%|██▍ | 11484/48008 [39:10<2:10:09, 4.68it/s] 24%|██▍ | 11485/48008 [39:10<2:07:51, 4.76it/s] 24%|██▍ | 11486/48008 [39:10<2:03:35, 4.92it/s] 24%|██▍ | 11487/48008 [39:10<2:00:31, 5.05it/s] 24%|██▍ | 11488/48008 [39:10<2:02:15, 4.98it/s] 24%|██▍ | 11489/48008 [39:11<2:00:26, 5.05it/s] 24%|██▍ | 11490/48008 [39:11<2:00:31, 5.05it/s] 24%|██▍ | 11491/48008 [39:11<1:58:20, 5.14it/s] 24%|██▍ | 11492/48008 [39:11<1:57:37, 5.17it/s] 24%|██▍ | 11493/48008 [39:11<1:57:05, 5.20it/s] 24%|██▍ | 11494/48008 [39:12<1:55:58, 5.25it/s] 24%|██▍ | 11495/48008 [39:12<1:57:59, 5.16it/s] 24%|██▍ | 11496/48008 [39:12<1:57:20, 5.19it/s] 24%|██▍ | 11497/48008 [39:12<1:56:01, 5.24it/s] 24%|██▍ | 11498/48008 [39:12<1:55:54, 5.25it/s] 24%|██▍ | 11499/48008 [39:13<1:56:39, 5.22it/s] 24%|██▍ | 11500/48008 [39:13<1:55:32, 5.27it/s] {'loss': 0.3014, 'grad_norm': 0.1953125, 'learning_rate': 0.0001520954840859857, 'epoch': 0.24} 24%|██▍ | 11500/48008 [39:13<1:55:32, 5.27it/s] 24%|██▍ | 11501/48008 [39:13<1:55:31, 5.27it/s] 24%|██▍ | 11502/48008 [39:13<1:57:59, 5.16it/s] 24%|██▍ | 11503/48008 [39:13<1:57:29, 5.18it/s] 24%|██▍ | 11504/48008 [39:14<1:58:06, 5.15it/s] 24%|██▍ | 11505/48008 [39:14<2:33:02, 3.98it/s] 24%|██▍ | 11506/48008 [39:14<2:21:53, 4.29it/s] 24%|██▍ | 11507/48008 [39:14<2:23:00, 4.25it/s] 24%|██▍ | 11508/48008 [39:15<2:12:55, 4.58it/s] 24%|██▍ | 11509/48008 [39:15<2:07:32, 4.77it/s] 24%|██▍ | 11510/48008 [39:15<2:04:24, 4.89it/s] 24%|██▍ | 11511/48008 [39:15<2:01:57, 4.99it/s] 24%|██▍ | 11512/48008 [39:15<1:58:54, 5.12it/s] 24%|██▍ | 11513/48008 [39:15<1:59:58, 5.07it/s] 24%|██▍ | 11514/48008 [39:16<1:58:20, 5.14it/s] 24%|██▍ | 11515/48008 [39:16<1:57:48, 5.16it/s] 24%|██▍ | 11516/48008 [39:16<1:57:12, 5.19it/s] 24%|██▍ | 11517/48008 [39:16<1:57:20, 5.18it/s] 24%|██▍ | 11518/48008 [39:16<1:57:09, 5.19it/s] 24%|██▍ | 11519/48008 [39:17<1:59:04, 5.11it/s] 24%|██▍ | 11520/48008 [39:17<1:58:56, 5.11it/s] 24%|██▍ | 11521/48008 [39:17<1:59:00, 5.11it/s] 24%|██▍ | 11522/48008 [39:17<1:59:39, 5.08it/s] 24%|██▍ | 11523/48008 [39:17<1:58:18, 5.14it/s] 24%|██▍ | 11524/48008 [39:18<1:57:41, 5.17it/s] 24%|██▍ | 11525/48008 [39:18<1:57:20, 5.18it/s] 24%|██▍ | 11526/48008 [39:18<1:55:55, 5.25it/s] 24%|██▍ | 11527/48008 [39:18<2:16:52, 4.44it/s] 24%|██▍ | 11528/48008 [39:18<2:10:45, 4.65it/s] 24%|██▍ | 11529/48008 [39:19<2:09:58, 4.68it/s] 24%|██▍ | 11530/48008 [39:19<2:05:54, 4.83it/s] 24%|██▍ | 11531/48008 [39:19<2:04:57, 4.86it/s] 24%|██▍ | 11532/48008 [39:19<2:02:16, 4.97it/s] 24%|██▍ | 11533/48008 [39:20<2:03:28, 4.92it/s] 24%|██▍ | 11534/48008 [39:20<2:00:38, 5.04it/s] 24%|██▍ | 11535/48008 [39:20<1:59:30, 5.09it/s] 24%|██▍ | 11536/48008 [39:20<1:58:32, 5.13it/s] 24%|██▍ | 11537/48008 [39:20<1:59:08, 5.10it/s] 24%|██▍ | 11538/48008 [39:20<1:57:23, 5.18it/s] 24%|██▍ | 11539/48008 [39:21<2:18:00, 4.40it/s] 24%|██▍ | 11540/48008 [39:21<2:10:05, 4.67it/s] 24%|██▍ | 11541/48008 [39:21<2:07:59, 4.75it/s] 24%|██▍ | 11542/48008 [39:21<2:04:17, 4.89it/s] 24%|██▍ | 11543/48008 [39:22<2:01:33, 5.00it/s] 24%|██▍ | 11544/48008 [39:22<2:00:42, 5.03it/s] 24%|██▍ | 11545/48008 [39:22<1:59:34, 5.08it/s] 24%|██▍ | 11546/48008 [39:22<1:58:05, 5.15it/s] 24%|██▍ | 11547/48008 [39:22<1:57:48, 5.16it/s] 24%|██▍ | 11548/48008 [39:22<1:58:33, 5.13it/s] 24%|██▍ | 11549/48008 [39:23<1:57:54, 5.15it/s] 24%|██▍ | 11550/48008 [39:23<1:58:06, 5.14it/s] {'loss': 0.3016, 'grad_norm': 0.1826171875, 'learning_rate': 0.00015188718546908848, 'epoch': 0.24} 24%|██▍ | 11550/48008 [39:23<1:58:06, 5.14it/s] 24%|██▍ | 11551/48008 [39:23<1:58:54, 5.11it/s] 24%|██▍ | 11552/48008 [39:23<2:01:33, 5.00it/s] 24%|██▍ | 11553/48008 [39:23<2:00:12, 5.05it/s] 24%|██▍ | 11554/48008 [39:24<1:58:59, 5.11it/s] 24%|██▍ | 11555/48008 [39:24<2:33:24, 3.96it/s] 24%|██▍ | 11556/48008 [39:24<2:23:09, 4.24it/s] 24%|██▍ | 11557/48008 [39:24<2:14:45, 4.51it/s] 24%|██▍ | 11558/48008 [39:25<2:08:58, 4.71it/s] 24%|██▍ | 11559/48008 [39:25<2:03:53, 4.90it/s] 24%|██▍ | 11560/48008 [39:25<2:00:27, 5.04it/s] 24%|██▍ | 11561/48008 [39:25<1:58:28, 5.13it/s] 24%|██▍ | 11562/48008 [39:25<1:57:54, 5.15it/s] 24%|██▍ | 11563/48008 [39:26<1:56:51, 5.20it/s] 24%|██▍ | 11564/48008 [39:26<1:58:50, 5.11it/s] 24%|██▍ | 11565/48008 [39:26<1:59:09, 5.10it/s] 24%|██▍ | 11566/48008 [39:26<1:58:02, 5.14it/s] 24%|██▍ | 11567/48008 [39:26<1:57:39, 5.16it/s] 24%|██▍ | 11568/48008 [39:27<1:56:52, 5.20it/s] 24%|██▍ | 11569/48008 [39:27<1:56:02, 5.23it/s] 24%|██▍ | 11570/48008 [39:27<1:55:21, 5.26it/s] 24%|██▍ | 11571/48008 [39:27<2:16:25, 4.45it/s] 24%|██▍ | 11572/48008 [39:27<2:09:49, 4.68it/s] 24%|██▍ | 11573/48008 [39:28<2:08:03, 4.74it/s] 24%|██▍ | 11574/48008 [39:28<2:04:40, 4.87it/s] 24%|██▍ | 11575/48008 [39:28<2:10:39, 4.65it/s] 24%|██▍ | 11576/48008 [39:28<2:06:13, 4.81it/s] 24%|██▍ | 11577/48008 [39:28<2:11:35, 4.61it/s] 24%|██▍ | 11578/48008 [39:29<2:06:10, 4.81it/s] 24%|██▍ | 11579/48008 [39:29<2:01:49, 4.98it/s] 24%|██▍ | 11580/48008 [39:29<2:08:15, 4.73it/s] 24%|██▍ | 11581/48008 [39:29<2:05:20, 4.84it/s] 24%|██▍ | 11582/48008 [39:29<2:03:50, 4.90it/s] 24%|██▍ | 11583/48008 [39:30<2:02:32, 4.95it/s] 24%|██▍ | 11584/48008 [39:30<2:00:48, 5.03it/s] 24%|██▍ | 11585/48008 [39:30<1:58:41, 5.11it/s] 24%|██▍ | 11586/48008 [39:30<1:59:15, 5.09it/s] 24%|██▍ | 11587/48008 [39:30<1:59:32, 5.08it/s] 24%|██▍ | 11588/48008 [39:31<1:59:32, 5.08it/s] 24%|██▍ | 11589/48008 [39:31<1:58:57, 5.10it/s] 24%|██▍ | 11590/48008 [39:31<1:56:58, 5.19it/s] 24%|██▍ | 11591/48008 [39:31<1:58:42, 5.11it/s] 24%|██▍ | 11592/48008 [39:31<2:00:40, 5.03it/s] 24%|██▍ | 11593/48008 [39:32<1:59:01, 5.10it/s] 24%|██▍ | 11594/48008 [39:32<1:56:58, 5.19it/s] 24%|██▍ | 11595/48008 [39:32<1:56:29, 5.21it/s] 24%|██▍ | 11596/48008 [39:32<1:56:18, 5.22it/s] 24%|██▍ | 11597/48008 [39:32<1:55:56, 5.23it/s] 24%|██▍ | 11598/48008 [39:33<1:55:17, 5.26it/s] 24%|██▍ | 11599/48008 [39:33<1:55:38, 5.25it/s] 24%|██▍ | 11600/48008 [39:33<1:55:53, 5.24it/s] {'loss': 0.302, 'grad_norm': 0.380859375, 'learning_rate': 0.00015167888685219133, 'epoch': 0.24} 24%|██▍ | 11600/48008 [39:33<1:55:53, 5.24it/s] 24%|██▍ | 11601/48008 [39:33<1:55:50, 5.24it/s] 24%|██▍ | 11602/48008 [39:33<2:16:59, 4.43it/s] 24%|██▍ | 11603/48008 [39:34<2:14:22, 4.52it/s] 24%|██▍ | 11604/48008 [39:34<2:08:22, 4.73it/s] 24%|██▍ | 11605/48008 [39:34<2:12:52, 4.57it/s] 24%|██▍ | 11606/48008 [39:34<2:07:34, 4.76it/s] 24%|██▍ | 11607/48008 [39:35<2:12:32, 4.58it/s] 24%|██▍ | 11608/48008 [39:35<2:06:14, 4.81it/s] 24%|██▍ | 11609/48008 [39:35<2:03:21, 4.92it/s] 24%|██▍ | 11610/48008 [39:35<2:00:50, 5.02it/s] 24%|██▍ | 11611/48008 [39:35<1:59:02, 5.10it/s] 24%|██▍ | 11612/48008 [39:35<1:57:57, 5.14it/s] 24%|██▍ | 11613/48008 [39:36<2:32:59, 3.96it/s] 24%|██▍ | 11614/48008 [39:36<2:22:58, 4.24it/s] 24%|██▍ | 11615/48008 [39:36<2:17:02, 4.43it/s] 24%|██▍ | 11616/48008 [39:36<2:10:57, 4.63it/s] 24%|██▍ | 11617/48008 [39:37<2:07:49, 4.75it/s] 24%|██▍ | 11618/48008 [39:37<2:07:22, 4.76it/s] 24%|██▍ | 11619/48008 [39:37<2:03:33, 4.91it/s] 24%|██▍ | 11620/48008 [39:37<2:01:17, 5.00it/s] 24%|██▍ | 11621/48008 [39:37<1:59:03, 5.09it/s] 24%|██▍ | 11622/48008 [39:38<2:06:51, 4.78it/s] 24%|██▍ | 11623/48008 [39:38<2:05:55, 4.82it/s] 24%|██▍ | 11624/48008 [39:38<2:02:01, 4.97it/s] 24%|██▍ | 11625/48008 [39:38<2:02:12, 4.96it/s] 24%|██▍ | 11626/48008 [39:38<2:08:40, 4.71it/s] 24%|██▍ | 11627/48008 [39:39<2:04:56, 4.85it/s] 24%|██▍ | 11628/48008 [39:39<2:01:58, 4.97it/s] 24%|██▍ | 11629/48008 [39:39<2:08:27, 4.72it/s] 24%|██▍ | 11630/48008 [39:39<2:05:32, 4.83it/s] 24%|██▍ | 11631/48008 [39:39<2:03:42, 4.90it/s] 24%|██▍ | 11632/48008 [39:40<2:00:26, 5.03it/s] 24%|██▍ | 11633/48008 [39:40<1:58:31, 5.12it/s] 24%|██▍ | 11634/48008 [39:40<2:00:19, 5.04it/s] 24%|██▍ | 11635/48008 [39:40<2:00:22, 5.04it/s] 24%|██▍ | 11636/48008 [39:40<1:59:10, 5.09it/s] 24%|██▍ | 11637/48008 [39:41<1:58:00, 5.14it/s] 24%|██▍ | 11638/48008 [39:41<1:57:05, 5.18it/s] 24%|██▍ | 11639/48008 [39:41<1:56:31, 5.20it/s] 24%|██▍ | 11640/48008 [39:41<2:17:14, 4.42it/s] 24%|██▍ | 11641/48008 [39:42<2:09:40, 4.67it/s] 24%|██▍ | 11642/48008 [39:42<2:04:37, 4.86it/s] 24%|██▍ | 11643/48008 [39:42<2:02:10, 4.96it/s] 24%|██▍ | 11644/48008 [39:42<2:02:33, 4.95it/s] 24%|██▍ | 11645/48008 [39:42<2:01:47, 4.98it/s] 24%|██▍ | 11646/48008 [39:43<2:02:08, 4.96it/s] 24%|██▍ | 11647/48008 [39:43<1:59:57, 5.05it/s] 24%|██▍ | 11648/48008 [39:43<1:58:48, 5.10it/s] 24%|██▍ | 11649/48008 [39:43<1:59:48, 5.06it/s] 24%|██▍ | 11650/48008 [39:43<1:59:14, 5.08it/s] {'loss': 0.2974, 'grad_norm': 0.55078125, 'learning_rate': 0.00015147058823529412, 'epoch': 0.24} 24%|██▍ | 11650/48008 [39:43<1:59:14, 5.08it/s] 24%|██▍ | 11651/48008 [39:43<1:57:56, 5.14it/s] 24%|██▍ | 11652/48008 [39:44<1:56:44, 5.19it/s] 24%|██▍ | 11653/48008 [39:44<1:56:39, 5.19it/s] 24%|██▍ | 11654/48008 [39:44<2:04:44, 4.86it/s] 24%|██▍ | 11655/48008 [39:44<2:03:08, 4.92it/s] 24%|██▍ | 11656/48008 [39:44<2:01:15, 5.00it/s] 24%|██▍ | 11657/48008 [39:45<1:59:44, 5.06it/s] 24%|██▍ | 11658/48008 [39:45<1:58:09, 5.13it/s] 24%|██▍ | 11659/48008 [39:45<1:58:25, 5.12it/s] 24%|██▍ | 11660/48008 [39:45<1:57:32, 5.15it/s] 24%|██▍ | 11661/48008 [39:45<1:57:02, 5.18it/s] 24%|██▍ | 11662/48008 [39:46<1:58:50, 5.10it/s] 24%|██▍ | 11663/48008 [39:46<1:59:52, 5.05it/s] 24%|██▍ | 11664/48008 [39:46<1:58:55, 5.09it/s] 24%|██▍ | 11665/48008 [39:46<2:33:33, 3.94it/s] 24%|██▍ | 11666/48008 [39:47<2:22:41, 4.25it/s] 24%|██▍ | 11667/48008 [39:47<2:13:41, 4.53it/s] 24%|██▍ | 11668/48008 [39:47<2:08:16, 4.72it/s] 24%|██▍ | 11669/48008 [39:47<2:05:39, 4.82it/s] 24%|██▍ | 11670/48008 [39:47<2:01:36, 4.98it/s] 24%|██▍ | 11671/48008 [39:48<2:00:18, 5.03it/s] 24%|██▍ | 11672/48008 [39:48<1:57:16, 5.16it/s] 24%|██▍ | 11673/48008 [39:48<1:58:38, 5.10it/s] 24%|██▍ | 11674/48008 [39:48<1:56:46, 5.19it/s] 24%|██▍ | 11675/48008 [39:48<1:55:41, 5.23it/s] 24%|██▍ | 11676/48008 [39:49<1:55:23, 5.25it/s] 24%|██▍ | 11677/48008 [39:49<1:56:17, 5.21it/s] 24%|██▍ | 11678/48008 [39:49<1:56:02, 5.22it/s] 24%|██▍ | 11679/48008 [39:49<1:58:17, 5.12it/s] 24%|██▍ | 11680/48008 [39:49<1:58:42, 5.10it/s] 24%|██▍ | 11681/48008 [39:49<1:57:38, 5.15it/s] 24%|██▍ | 11682/48008 [39:50<2:32:42, 3.96it/s] 24%|██▍ | 11683/48008 [39:50<2:20:35, 4.31it/s] 24%|██▍ | 11684/48008 [39:50<2:12:57, 4.55it/s] 24%|██▍ | 11685/48008 [39:50<2:08:27, 4.71it/s] 24%|██▍ | 11686/48008 [39:51<2:04:02, 4.88it/s] 24%|██▍ | 11687/48008 [39:51<2:04:30, 4.86it/s] 24%|██▍ | 11688/48008 [39:51<2:01:24, 4.99it/s] 24%|██▍ | 11689/48008 [39:51<2:01:43, 4.97it/s] 24%|██▍ | 11690/48008 [39:51<2:01:55, 4.96it/s] 24%|██▍ | 11691/48008 [39:52<1:58:48, 5.09it/s] 24%|██▍ | 11692/48008 [39:52<2:01:15, 4.99it/s] 24%|██▍ | 11693/48008 [39:52<1:58:22, 5.11it/s] 24%|██▍ | 11694/48008 [39:52<1:57:51, 5.14it/s] 24%|██▍ | 11695/48008 [39:52<1:57:15, 5.16it/s] 24%|██▍ | 11696/48008 [39:53<1:59:02, 5.08it/s] 24%|██▍ | 11697/48008 [39:53<2:00:08, 5.04it/s] 24%|██▍ | 11698/48008 [39:53<2:00:48, 5.01it/s] 24%|██▍ | 11699/48008 [39:53<1:58:43, 5.10it/s] 24%|██▍ | 11700/48008 [39:53<1:57:40, 5.14it/s] {'loss': 0.3038, 'grad_norm': 0.1337890625, 'learning_rate': 0.00015126228961839694, 'epoch': 0.24} 24%|██▍ | 11700/48008 [39:53<1:57:40, 5.14it/s] 24%|██▍ | 11701/48008 [39:54<1:56:50, 5.18it/s] 24%|██▍ | 11702/48008 [39:54<1:57:00, 5.17it/s] 24%|██▍ | 11703/48008 [39:54<1:56:18, 5.20it/s] 24%|██▍ | 11704/48008 [39:54<1:56:56, 5.17it/s] 24%|██▍ | 11705/48008 [39:54<1:57:31, 5.15it/s] 24%|██▍ | 11706/48008 [39:55<1:57:22, 5.15it/s] 24%|██▍ | 11707/48008 [39:55<1:59:02, 5.08it/s] 24%|██▍ | 11708/48008 [39:55<1:58:20, 5.11it/s] 24%|██▍ | 11709/48008 [39:55<1:56:25, 5.20it/s] 24%|██▍ | 11710/48008 [39:55<1:56:07, 5.21it/s] 24%|██▍ | 11711/48008 [39:56<1:55:58, 5.22it/s] 24%|██▍ | 11712/48008 [39:56<1:55:34, 5.23it/s] 24%|██▍ | 11713/48008 [39:56<1:56:32, 5.19it/s] 24%|██▍ | 11714/48008 [39:57<4:01:16, 2.51it/s] 24%|██▍ | 11715/48008 [39:57<3:23:36, 2.97it/s] 24%|██▍ | 11716/48008 [39:57<2:57:20, 3.41it/s] 24%|██▍ | 11717/48008 [39:57<2:40:50, 3.76it/s] 24%|██▍ | 11718/48008 [39:58<2:26:43, 4.12it/s] 24%|██▍ | 11719/48008 [39:58<2:20:26, 4.31it/s] 24%|██▍ | 11720/48008 [39:58<2:13:10, 4.54it/s] 24%|██▍ | 11721/48008 [39:58<2:08:12, 4.72it/s] 24%|██▍ | 11722/48008 [39:58<2:04:02, 4.88it/s] 24%|██▍ | 11723/48008 [39:59<2:22:06, 4.26it/s] 24%|██▍ | 11724/48008 [39:59<2:14:42, 4.49it/s] 24%|██▍ | 11725/48008 [39:59<2:09:04, 4.69it/s] 24%|██▍ | 11726/48008 [39:59<2:04:48, 4.85it/s] 24%|██▍ | 11727/48008 [39:59<2:01:42, 4.97it/s] 24%|██▍ | 11728/48008 [40:00<1:59:23, 5.06it/s] 24%|██▍ | 11729/48008 [40:00<2:01:31, 4.98it/s] 24%|██▍ | 11730/48008 [40:00<2:20:48, 4.29it/s] 24%|██▍ | 11731/48008 [40:00<2:21:34, 4.27it/s] 24%|██▍ | 11732/48008 [40:01<2:16:57, 4.41it/s] 24%|██▍ | 11733/48008 [40:01<2:09:59, 4.65it/s] 24%|██▍ | 11734/48008 [40:01<2:05:57, 4.80it/s] 24%|██▍ | 11735/48008 [40:01<2:02:07, 4.95it/s] 24%|██▍ | 11736/48008 [40:01<2:08:23, 4.71it/s] 24%|██▍ | 11737/48008 [40:02<2:04:09, 4.87it/s] 24%|██▍ | 11738/48008 [40:02<1:59:19, 5.07it/s] 24%|██▍ | 11739/48008 [40:02<1:58:02, 5.12it/s] 24%|██▍ | 11740/48008 [40:02<1:58:40, 5.09it/s] 24%|██▍ | 11741/48008 [40:02<2:00:05, 5.03it/s] 24%|██▍ | 11742/48008 [40:03<2:34:15, 3.92it/s] 24%|██▍ | 11743/48008 [40:03<2:22:01, 4.26it/s] 24%|██▍ | 11744/48008 [40:03<2:14:06, 4.51it/s] 24%|██▍ | 11745/48008 [40:03<2:08:23, 4.71it/s] 24%|██▍ | 11746/48008 [40:03<2:04:20, 4.86it/s] 24%|██▍ | 11747/48008 [40:04<2:01:50, 4.96it/s] 24%|██▍ | 11748/48008 [40:04<2:01:14, 4.98it/s] 24%|██▍ | 11749/48008 [40:04<1:59:23, 5.06it/s] 24%|██▍ | 11750/48008 [40:04<1:58:14, 5.11it/s] {'loss': 0.3007, 'grad_norm': 0.216796875, 'learning_rate': 0.00015105399100149975, 'epoch': 0.24} 24%|██▍ | 11750/48008 [40:04<1:58:14, 5.11it/s] 24%|██▍ | 11751/48008 [40:04<2:05:53, 4.80it/s] 24%|██▍ | 11752/48008 [40:05<2:02:46, 4.92it/s] 24%|██▍ | 11753/48008 [40:05<2:00:12, 5.03it/s] 24%|██▍ | 11754/48008 [40:05<1:58:04, 5.12it/s] 24%|██▍ | 11755/48008 [40:05<1:58:11, 5.11it/s] 24%|██▍ | 11756/48008 [40:05<1:55:36, 5.23it/s] 24%|██▍ | 11757/48008 [40:06<1:55:53, 5.21it/s] 24%|██▍ | 11758/48008 [40:06<1:55:09, 5.25it/s] 24%|██▍ | 11759/48008 [40:06<1:57:32, 5.14it/s] 24%|██▍ | 11760/48008 [40:06<2:00:28, 5.01it/s] 24%|██▍ | 11761/48008 [40:06<2:00:11, 5.03it/s] 25%|██▍ | 11762/48008 [40:07<2:00:39, 5.01it/s] 25%|██▍ | 11763/48008 [40:07<2:34:40, 3.91it/s] 25%|██▍ | 11764/48008 [40:07<2:22:34, 4.24it/s] 25%|██▍ | 11765/48008 [40:07<2:14:34, 4.49it/s] 25%|██▍ | 11766/48008 [40:08<2:08:27, 4.70it/s] 25%|██▍ | 11767/48008 [40:08<2:04:41, 4.84it/s] 25%|██▍ | 11768/48008 [40:08<2:01:29, 4.97it/s] 25%|██▍ | 11769/48008 [40:08<1:59:26, 5.06it/s] 25%|██▍ | 11770/48008 [40:08<2:00:14, 5.02it/s] 25%|██▍ | 11771/48008 [40:09<1:58:48, 5.08it/s] 25%|██▍ | 11772/48008 [40:09<1:57:45, 5.13it/s] 25%|██▍ | 11773/48008 [40:09<1:56:54, 5.17it/s] 25%|██▍ | 11774/48008 [40:09<1:56:18, 5.19it/s] 25%|██▍ | 11775/48008 [40:09<1:57:52, 5.12it/s] 25%|██▍ | 11776/48008 [40:10<1:56:20, 5.19it/s] 25%|██▍ | 11777/48008 [40:10<1:55:32, 5.23it/s] 25%|██▍ | 11778/48008 [40:10<1:55:45, 5.22it/s] 25%|██▍ | 11779/48008 [40:10<1:55:40, 5.22it/s] 25%|██▍ | 11780/48008 [40:10<1:55:25, 5.23it/s] 25%|██▍ | 11781/48008 [40:10<1:55:05, 5.25it/s] 25%|██▍ | 11782/48008 [40:11<1:56:09, 5.20it/s] 25%|██▍ | 11783/48008 [40:11<1:57:50, 5.12it/s] 25%|██▍ | 11784/48008 [40:11<1:57:21, 5.14it/s] 25%|██▍ | 11785/48008 [40:11<1:56:05, 5.20it/s] 25%|██▍ | 11786/48008 [40:11<1:55:33, 5.22it/s] 25%|██▍ | 11787/48008 [40:12<1:55:38, 5.22it/s] 25%|██▍ | 11788/48008 [40:12<2:04:02, 4.87it/s] 25%|██▍ | 11789/48008 [40:12<2:00:48, 5.00it/s] 25%|██▍ | 11790/48008 [40:12<1:58:29, 5.09it/s] 25%|██▍ | 11791/48008 [40:12<1:59:37, 5.05it/s] 25%|██▍ | 11792/48008 [40:13<1:57:37, 5.13it/s] 25%|██▍ | 11793/48008 [40:13<1:56:05, 5.20it/s] 25%|██▍ | 11794/48008 [40:13<1:56:53, 5.16it/s] 25%|██▍ | 11795/48008 [40:13<1:55:35, 5.22it/s] 25%|██▍ | 11796/48008 [40:13<1:54:48, 5.26it/s] 25%|██▍ | 11797/48008 [40:14<1:54:18, 5.28it/s] 25%|██▍ | 11798/48008 [40:14<1:54:13, 5.28it/s] 25%|██▍ | 11799/48008 [40:14<1:53:31, 5.32it/s] 25%|██▍ | 11800/48008 [40:14<1:54:45, 5.26it/s] {'loss': 0.3037, 'grad_norm': 0.384765625, 'learning_rate': 0.00015084569238460257, 'epoch': 0.25} 25%|██▍ | 11800/48008 [40:14<1:54:45, 5.26it/s] 25%|██▍ | 11801/48008 [40:14<1:55:09, 5.24it/s] 25%|██▍ | 11802/48008 [40:15<1:57:34, 5.13it/s] 25%|██▍ | 11803/48008 [40:15<1:57:50, 5.12it/s] 25%|██▍ | 11804/48008 [40:15<1:55:45, 5.21it/s] 25%|██▍ | 11805/48008 [40:15<1:56:50, 5.16it/s] 25%|██▍ | 11806/48008 [40:15<2:31:24, 3.99it/s] 25%|██▍ | 11807/48008 [40:16<2:21:23, 4.27it/s] 25%|██▍ | 11808/48008 [40:16<2:13:32, 4.52it/s] 25%|██▍ | 11809/48008 [40:16<2:07:25, 4.73it/s] 25%|██▍ | 11810/48008 [40:16<2:03:22, 4.89it/s] 25%|██▍ | 11811/48008 [40:16<2:01:02, 4.98it/s] 25%|██▍ | 11812/48008 [40:17<1:59:21, 5.05it/s] 25%|██▍ | 11813/48008 [40:17<2:01:19, 4.97it/s] 25%|██▍ | 11814/48008 [40:17<2:34:57, 3.89it/s] 25%|██▍ | 11815/48008 [40:17<2:24:51, 4.16it/s] 25%|██▍ | 11816/48008 [40:18<2:16:06, 4.43it/s] 25%|██▍ | 11817/48008 [40:18<2:09:51, 4.65it/s] 25%|██▍ | 11818/48008 [40:18<2:07:43, 4.72it/s] 25%|██▍ | 11819/48008 [40:18<2:06:07, 4.78it/s] 25%|██▍ | 11820/48008 [40:18<2:03:46, 4.87it/s] 25%|██▍ | 11821/48008 [40:19<2:01:05, 4.98it/s] 25%|██▍ | 11822/48008 [40:19<2:03:05, 4.90it/s] 25%|██▍ | 11823/48008 [40:19<2:00:26, 5.01it/s] 25%|██▍ | 11824/48008 [40:19<1:59:04, 5.06it/s] 25%|██▍ | 11825/48008 [40:19<1:57:45, 5.12it/s] 25%|██▍ | 11826/48008 [40:20<1:57:08, 5.15it/s] 25%|██▍ | 11827/48008 [40:20<1:57:49, 5.12it/s] 25%|██▍ | 11828/48008 [40:20<1:56:23, 5.18it/s] 25%|██▍ | 11829/48008 [40:20<2:17:00, 4.40it/s] 25%|██▍ | 11830/48008 [40:21<2:19:08, 4.33it/s] 25%|██▍ | 11831/48008 [40:21<2:12:44, 4.54it/s] 25%|██▍ | 11832/48008 [40:21<2:07:06, 4.74it/s] 25%|██▍ | 11833/48008 [40:21<2:02:53, 4.91it/s] 25%|██▍ | 11834/48008 [40:21<2:00:56, 4.98it/s] 25%|██▍ | 11835/48008 [40:21<1:59:21, 5.05it/s] 25%|██▍ | 11836/48008 [40:22<1:57:26, 5.13it/s] 25%|██▍ | 11837/48008 [40:22<1:56:34, 5.17it/s] 25%|██▍ | 11838/48008 [40:22<1:56:12, 5.19it/s] 25%|██▍ | 11839/48008 [40:22<1:55:11, 5.23it/s] 25%|██▍ | 11840/48008 [40:22<1:55:16, 5.23it/s] 25%|██▍ | 11841/48008 [40:23<1:58:42, 5.08it/s] 25%|██▍ | 11842/48008 [40:23<1:57:26, 5.13it/s] 25%|██▍ | 11843/48008 [40:23<1:55:26, 5.22it/s] 25%|██▍ | 11844/48008 [40:23<1:57:42, 5.12it/s] 25%|██▍ | 11845/48008 [40:23<1:59:06, 5.06it/s] 25%|██▍ | 11846/48008 [40:24<1:58:05, 5.10it/s] 25%|██▍ | 11847/48008 [40:24<2:32:30, 3.95it/s] 25%|██▍ | 11848/48008 [40:24<2:21:09, 4.27it/s] 25%|██▍ | 11849/48008 [40:24<2:16:32, 4.41it/s] 25%|██▍ | 11850/48008 [40:25<2:10:05, 4.63it/s] {'loss': 0.3003, 'grad_norm': 0.455078125, 'learning_rate': 0.0001506373937677054, 'epoch': 0.25} 25%|██▍ | 11850/48008 [40:25<2:10:05, 4.63it/s] 25%|██▍ | 11851/48008 [40:25<2:07:36, 4.72it/s] 25%|██▍ | 11852/48008 [40:25<2:04:07, 4.85it/s] 25%|██▍ | 11853/48008 [40:25<2:09:44, 4.64it/s] 25%|██▍ | 11854/48008 [40:25<2:05:15, 4.81it/s] 25%|██▍ | 11855/48008 [40:26<2:01:58, 4.94it/s] 25%|██▍ | 11856/48008 [40:26<2:01:00, 4.98it/s] 25%|██▍ | 11857/48008 [40:26<2:01:31, 4.96it/s] 25%|██▍ | 11858/48008 [40:26<2:02:52, 4.90it/s] 25%|██▍ | 11859/48008 [40:26<2:08:37, 4.68it/s] 25%|██▍ | 11860/48008 [40:27<2:05:20, 4.81it/s] 25%|██▍ | 11861/48008 [40:27<2:02:00, 4.94it/s] 25%|██▍ | 11862/48008 [40:27<2:02:04, 4.93it/s] 25%|██▍ | 11863/48008 [40:27<2:08:23, 4.69it/s] 25%|██▍ | 11864/48008 [40:27<2:04:08, 4.85it/s] 25%|██▍ | 11865/48008 [40:28<2:09:28, 4.65it/s] 25%|██▍ | 11866/48008 [40:28<2:07:23, 4.73it/s] 25%|██▍ | 11867/48008 [40:28<2:02:57, 4.90it/s] 25%|██▍ | 11868/48008 [40:28<2:01:34, 4.95it/s] 25%|██▍ | 11869/48008 [40:28<1:59:51, 5.03it/s] 25%|██▍ | 11870/48008 [40:29<1:59:26, 5.04it/s] 25%|██▍ | 11871/48008 [40:29<1:58:11, 5.10it/s] 25%|██▍ | 11872/48008 [40:29<1:58:47, 5.07it/s] 25%|██▍ | 11873/48008 [40:29<1:57:20, 5.13it/s] 25%|██▍ | 11874/48008 [40:29<1:59:47, 5.03it/s] 25%|██▍ | 11875/48008 [40:30<1:58:03, 5.10it/s] 25%|██▍ | 11876/48008 [40:30<1:56:43, 5.16it/s] 25%|██▍ | 11877/48008 [40:30<1:56:07, 5.19it/s] 25%|██▍ | 11878/48008 [40:30<1:55:17, 5.22it/s] 25%|██▍ | 11879/48008 [40:30<1:56:35, 5.16it/s] 25%|██▍ | 11880/48008 [40:31<1:57:10, 5.14it/s] 25%|██▍ | 11881/48008 [40:31<1:56:27, 5.17it/s] 25%|██▍ | 11882/48008 [40:31<1:55:41, 5.20it/s] 25%|██▍ | 11883/48008 [40:31<1:55:12, 5.23it/s] 25%|██▍ | 11884/48008 [40:31<1:55:13, 5.23it/s] 25%|██▍ | 11885/48008 [40:32<1:57:12, 5.14it/s] 25%|██▍ | 11886/48008 [40:32<1:57:49, 5.11it/s] 25%|██▍ | 11887/48008 [40:32<1:56:34, 5.16it/s] 25%|██▍ | 11888/48008 [40:32<1:56:01, 5.19it/s] 25%|██▍ | 11889/48008 [40:32<1:55:28, 5.21it/s] 25%|██▍ | 11890/48008 [40:33<2:30:23, 4.00it/s] 25%|██▍ | 11891/48008 [40:33<2:18:49, 4.34it/s] 25%|██▍ | 11892/48008 [40:33<2:12:03, 4.56it/s] 25%|██▍ | 11893/48008 [40:33<2:06:56, 4.74it/s] 25%|██▍ | 11894/48008 [40:33<2:03:15, 4.88it/s] 25%|██▍ | 11895/48008 [40:34<2:00:47, 4.98it/s] 25%|██▍ | 11896/48008 [40:34<2:00:15, 5.01it/s] 25%|██▍ | 11897/48008 [40:34<1:57:55, 5.10it/s] 25%|██▍ | 11898/48008 [40:34<1:56:53, 5.15it/s] 25%|██▍ | 11899/48008 [40:34<1:56:49, 5.15it/s] 25%|██▍ | 11900/48008 [40:35<1:56:47, 5.15it/s] {'loss': 0.3025, 'grad_norm': 0.2158203125, 'learning_rate': 0.0001504290951508082, 'epoch': 0.25} 25%|██▍ | 11900/48008 [40:35<1:56:47, 5.15it/s] 25%|██▍ | 11901/48008 [40:35<1:57:02, 5.14it/s] 25%|██▍ | 11902/48008 [40:35<1:57:02, 5.14it/s] 25%|██▍ | 11903/48008 [40:35<1:56:23, 5.17it/s] 25%|██▍ | 11904/48008 [40:35<1:55:37, 5.20it/s] 25%|██▍ | 11905/48008 [40:36<1:55:23, 5.21it/s] 25%|██▍ | 11906/48008 [40:36<1:54:50, 5.24it/s] 25%|██▍ | 11907/48008 [40:36<1:54:20, 5.26it/s] 25%|██▍ | 11908/48008 [40:36<1:57:49, 5.11it/s] 25%|██▍ | 11909/48008 [40:36<1:56:48, 5.15it/s] 25%|██▍ | 11910/48008 [40:37<1:55:49, 5.19it/s] 25%|██▍ | 11911/48008 [40:37<1:55:55, 5.19it/s] 25%|██▍ | 11912/48008 [40:37<1:57:43, 5.11it/s] 25%|██▍ | 11913/48008 [40:37<2:00:04, 5.01it/s] 25%|██▍ | 11914/48008 [40:37<2:06:47, 4.74it/s] 25%|██▍ | 11915/48008 [40:38<2:03:14, 4.88it/s] 25%|██▍ | 11916/48008 [40:38<1:59:17, 5.04it/s] 25%|██▍ | 11917/48008 [40:38<1:57:18, 5.13it/s] 25%|██▍ | 11918/48008 [40:38<1:56:35, 5.16it/s] 25%|██▍ | 11919/48008 [40:38<1:55:23, 5.21it/s] 25%|██▍ | 11920/48008 [40:39<1:55:03, 5.23it/s] 25%|██▍ | 11921/48008 [40:39<1:54:17, 5.26it/s] 25%|██▍ | 11922/48008 [40:39<1:56:41, 5.15it/s] 25%|██▍ | 11923/48008 [40:39<1:55:51, 5.19it/s] 25%|██▍ | 11924/48008 [40:39<1:57:57, 5.10it/s] 25%|██▍ | 11925/48008 [40:40<1:56:14, 5.17it/s] 25%|██▍ | 11926/48008 [40:40<1:56:38, 5.16it/s] 25%|██▍ | 11927/48008 [40:40<1:55:25, 5.21it/s] 25%|██▍ | 11928/48008 [40:40<1:56:58, 5.14it/s] 25%|██▍ | 11929/48008 [40:40<1:58:26, 5.08it/s] 25%|██▍ | 11930/48008 [40:40<1:58:19, 5.08it/s] 25%|██▍ | 11931/48008 [40:41<1:57:29, 5.12it/s] 25%|██▍ | 11932/48008 [40:41<1:56:03, 5.18it/s] 25%|██▍ | 11933/48008 [40:41<1:56:37, 5.16it/s] 25%|██▍ | 11934/48008 [40:41<2:04:49, 4.82it/s] 25%|██▍ | 11935/48008 [40:42<2:04:08, 4.84it/s] 25%|██▍ | 11936/48008 [40:42<2:03:34, 4.86it/s] 25%|██▍ | 11937/48008 [40:42<2:03:05, 4.88it/s] 25%|██▍ | 11938/48008 [40:42<2:00:56, 4.97it/s] 25%|██▍ | 11939/48008 [40:42<1:59:24, 5.03it/s] 25%|██▍ | 11940/48008 [40:43<2:01:30, 4.95it/s] 25%|██▍ | 11941/48008 [40:43<2:00:39, 4.98it/s] 25%|██▍ | 11942/48008 [40:43<1:58:03, 5.09it/s] 25%|██▍ | 11943/48008 [40:43<1:57:22, 5.12it/s] 25%|██▍ | 11944/48008 [40:43<1:55:54, 5.19it/s] 25%|██▍ | 11945/48008 [40:43<1:53:42, 5.29it/s] 25%|██▍ | 11946/48008 [40:44<1:53:19, 5.30it/s] 25%|██▍ | 11947/48008 [40:44<1:56:08, 5.18it/s] 25%|██▍ | 11948/48008 [40:44<1:55:41, 5.19it/s] 25%|██▍ | 11949/48008 [40:44<1:54:44, 5.24it/s] 25%|██▍ | 11950/48008 [40:44<1:54:52, 5.23it/s] {'loss': 0.3015, 'grad_norm': 0.27734375, 'learning_rate': 0.00015022079653391102, 'epoch': 0.25} 25%|██▍ | 11950/48008 [40:44<1:54:52, 5.23it/s] 25%|██▍ | 11951/48008 [40:45<1:54:32, 5.25it/s] 25%|██▍ | 11952/48008 [40:45<1:54:42, 5.24it/s] 25%|██▍ | 11953/48008 [40:45<1:54:27, 5.25it/s] 25%|██▍ | 11954/48008 [40:45<1:54:46, 5.24it/s] 25%|██▍ | 11955/48008 [40:45<1:55:22, 5.21it/s] 25%|██▍ | 11956/48008 [40:46<1:55:32, 5.20it/s] 25%|██▍ | 11957/48008 [40:46<1:53:40, 5.29it/s] 25%|██▍ | 11958/48008 [40:46<1:54:55, 5.23it/s] 25%|██▍ | 11959/48008 [40:46<2:02:45, 4.89it/s] 25%|██▍ | 11960/48008 [40:46<2:00:41, 4.98it/s] 25%|██▍ | 11961/48008 [40:47<2:00:51, 4.97it/s] 25%|██▍ | 11962/48008 [40:47<2:01:05, 4.96it/s] 25%|██▍ | 11963/48008 [40:47<1:58:29, 5.07it/s] 25%|██▍ | 11964/48008 [40:47<1:58:00, 5.09it/s] 25%|██▍ | 11965/48008 [40:47<1:56:10, 5.17it/s] 25%|██▍ | 11966/48008 [40:48<1:57:21, 5.12it/s] 25%|██▍ | 11967/48008 [40:48<1:56:24, 5.16it/s] 25%|██▍ | 11968/48008 [40:48<1:55:40, 5.19it/s] 25%|██▍ | 11969/48008 [40:48<1:56:09, 5.17it/s] 25%|██▍ | 11970/48008 [40:48<1:55:28, 5.20it/s] 25%|██▍ | 11971/48008 [40:49<2:03:25, 4.87it/s] 25%|██▍ | 11972/48008 [40:49<2:03:54, 4.85it/s] 25%|██▍ | 11973/48008 [40:49<2:01:13, 4.95it/s] 25%|██▍ | 11974/48008 [40:49<2:02:10, 4.92it/s] 25%|██▍ | 11975/48008 [40:49<2:02:01, 4.92it/s] 25%|██▍ | 11976/48008 [40:50<1:59:46, 5.01it/s] 25%|██▍ | 11977/48008 [40:50<1:58:20, 5.07it/s] 25%|██▍ | 11978/48008 [40:50<1:59:29, 5.03it/s] 25%|██▍ | 11979/48008 [40:50<1:58:07, 5.08it/s] 25%|██▍ | 11980/48008 [40:50<1:57:18, 5.12it/s] 25%|██▍ | 11981/48008 [40:51<1:56:32, 5.15it/s] 25%|██▍ | 11982/48008 [40:51<1:55:51, 5.18it/s] 25%|██▍ | 11983/48008 [40:51<1:55:57, 5.18it/s] 25%|██▍ | 11984/48008 [40:51<1:55:18, 5.21it/s] 25%|██▍ | 11985/48008 [40:51<1:55:32, 5.20it/s] 25%|██▍ | 11986/48008 [40:51<1:55:05, 5.22it/s] 25%|██▍ | 11987/48008 [40:52<1:54:46, 5.23it/s] 25%|██▍ | 11988/48008 [40:52<1:58:18, 5.07it/s] 25%|██▍ | 11989/48008 [40:52<1:56:35, 5.15it/s] 25%|██▍ | 11990/48008 [40:52<1:56:11, 5.17it/s] 25%|██▍ | 11991/48008 [40:52<1:56:47, 5.14it/s] 25%|██▍ | 11992/48008 [40:53<1:56:11, 5.17it/s] 25%|██▍ | 11993/48008 [40:53<1:56:08, 5.17it/s] 25%|██▍ | 11994/48008 [40:53<2:15:59, 4.41it/s] 25%|██▍ | 11995/48008 [40:53<2:09:51, 4.62it/s] 25%|██▍ | 11996/48008 [40:54<2:04:23, 4.82it/s] 25%|██▍ | 11997/48008 [40:54<2:01:06, 4.96it/s] 25%|██▍ | 11998/48008 [40:54<2:02:36, 4.90it/s] 25%|██▍ | 11999/48008 [40:54<2:00:03, 5.00it/s] 25%|██▍ | 12000/48008 [40:54<1:58:26, 5.07it/s] {'loss': 0.3011, 'grad_norm': 0.130859375, 'learning_rate': 0.00015001249791701384, 'epoch': 0.25} 25%|██▍ | 12000/48008 [40:54<1:58:26, 5.07it/s] 25%|██▍ | 12001/48008 [40:54<1:56:26, 5.15it/s] 25%|██▌ | 12002/48008 [40:55<1:55:32, 5.19it/s] 25%|██▌ | 12003/48008 [40:55<1:58:56, 5.05it/s] 25%|██▌ | 12004/48008 [40:55<1:58:25, 5.07it/s] 25%|██▌ | 12005/48008 [40:55<1:57:14, 5.12it/s] 25%|██▌ | 12006/48008 [40:55<1:57:03, 5.13it/s] 25%|██▌ | 12007/48008 [40:56<1:56:32, 5.15it/s] 25%|██▌ | 12008/48008 [40:56<1:56:11, 5.16it/s] 25%|██▌ | 12009/48008 [40:56<2:03:42, 4.85it/s] 25%|██▌ | 12010/48008 [40:56<2:01:05, 4.95it/s] 25%|██▌ | 12011/48008 [40:56<2:01:29, 4.94it/s] 25%|██▌ | 12012/48008 [40:57<2:20:03, 4.28it/s] 25%|██▌ | 12013/48008 [40:57<2:12:17, 4.54it/s] 25%|██▌ | 12014/48008 [40:57<2:07:32, 4.70it/s] 25%|██▌ | 12015/48008 [40:57<2:04:53, 4.80it/s] 25%|██▌ | 12016/48008 [40:58<2:01:04, 4.95it/s] 25%|██▌ | 12017/48008 [40:58<1:59:19, 5.03it/s] 25%|██▌ | 12018/48008 [40:58<1:57:19, 5.11it/s] 25%|██▌ | 12019/48008 [40:58<1:56:33, 5.15it/s] 25%|██▌ | 12020/48008 [40:58<1:58:29, 5.06it/s] 25%|██▌ | 12021/48008 [40:59<1:57:17, 5.11it/s] 25%|██▌ | 12022/48008 [40:59<1:57:09, 5.12it/s] 25%|██▌ | 12023/48008 [40:59<1:56:47, 5.14it/s] 25%|██▌ | 12024/48008 [40:59<1:57:06, 5.12it/s] 25%|██▌ | 12025/48008 [40:59<1:56:01, 5.17it/s] 25%|██▌ | 12026/48008 [40:59<1:57:05, 5.12it/s] 25%|██▌ | 12027/48008 [41:00<2:31:23, 3.96it/s] 25%|██▌ | 12028/48008 [41:00<2:20:42, 4.26it/s] 25%|██▌ | 12029/48008 [41:00<2:16:29, 4.39it/s] 25%|██▌ | 12030/48008 [41:00<2:10:37, 4.59it/s] 25%|██▌ | 12031/48008 [41:01<2:05:56, 4.76it/s] 25%|██▌ | 12032/48008 [41:01<2:02:18, 4.90it/s] 25%|██▌ | 12033/48008 [41:01<1:59:06, 5.03it/s] 25%|██▌ | 12034/48008 [41:01<1:58:02, 5.08it/s] 25%|██▌ | 12035/48008 [41:01<1:56:25, 5.15it/s] 25%|██▌ | 12036/48008 [41:02<1:55:30, 5.19it/s] 25%|██▌ | 12037/48008 [41:02<1:55:08, 5.21it/s] 25%|██▌ | 12038/48008 [41:02<1:54:52, 5.22it/s] 25%|██▌ | 12039/48008 [41:02<1:53:47, 5.27it/s] 25%|██▌ | 12040/48008 [41:02<1:53:59, 5.26it/s] 25%|██▌ | 12041/48008 [41:03<4:22:26, 2.28it/s] 25%|██▌ | 12042/48008 [41:04<3:41:31, 2.71it/s] 25%|██▌ | 12043/48008 [41:04<3:09:23, 3.16it/s] 25%|██▌ | 12044/48008 [41:04<2:55:34, 3.41it/s] 25%|██▌ | 12045/48008 [41:04<2:38:03, 3.79it/s] 25%|██▌ | 12046/48008 [41:04<2:24:58, 4.13it/s] 25%|██▌ | 12047/48008 [41:05<4:31:01, 2.21it/s] 25%|██▌ | 12048/48008 [41:06<3:43:27, 2.68it/s] 25%|██▌ | 12049/48008 [41:06<3:19:19, 3.01it/s] 25%|██▌ | 12050/48008 [41:06<2:53:47, 3.45it/s] {'loss': 0.301, 'grad_norm': 0.15625, 'learning_rate': 0.00014980419930011666, 'epoch': 0.25} 25%|██▌ | 12050/48008 [41:06<2:53:47, 3.45it/s] 25%|██▌ | 12051/48008 [41:06<2:38:11, 3.79it/s] 25%|██▌ | 12052/48008 [41:06<2:24:54, 4.14it/s] 25%|██▌ | 12053/48008 [41:07<2:15:57, 4.41it/s] 25%|██▌ | 12054/48008 [41:07<2:10:53, 4.58it/s] 25%|██▌ | 12055/48008 [41:07<2:04:50, 4.80it/s] 25%|██▌ | 12056/48008 [41:07<2:00:53, 4.96it/s] 25%|██▌ | 12057/48008 [41:07<1:58:23, 5.06it/s] 25%|██▌ | 12058/48008 [41:08<1:57:22, 5.10it/s] 25%|██▌ | 12059/48008 [41:09<4:54:04, 2.04it/s] 25%|██▌ | 12060/48008 [41:09<4:03:42, 2.46it/s] 25%|██▌ | 12061/48008 [41:09<3:25:04, 2.92it/s] 25%|██▌ | 12062/48008 [41:09<2:59:03, 3.35it/s] 25%|██▌ | 12063/48008 [41:09<2:40:23, 3.73it/s] 25%|██▌ | 12064/48008 [41:10<2:26:10, 4.10it/s] 25%|██▌ | 12065/48008 [41:10<2:15:44, 4.41it/s] 25%|██▌ | 12066/48008 [41:10<2:09:23, 4.63it/s] 25%|██▌ | 12067/48008 [41:10<2:05:03, 4.79it/s] 25%|██▌ | 12068/48008 [41:10<2:02:07, 4.90it/s] 25%|██▌ | 12069/48008 [41:11<2:00:57, 4.95it/s] 25%|██▌ | 12070/48008 [41:11<2:01:20, 4.94it/s] 25%|██▌ | 12071/48008 [41:11<2:07:39, 4.69it/s] 25%|██▌ | 12072/48008 [41:11<2:03:04, 4.87it/s] 25%|██▌ | 12073/48008 [41:11<2:00:51, 4.96it/s] 25%|██▌ | 12074/48008 [41:12<1:59:27, 5.01it/s] 25%|██▌ | 12075/48008 [41:12<1:57:24, 5.10it/s] 25%|██▌ | 12076/48008 [41:12<1:58:30, 5.05it/s] 25%|██▌ | 12077/48008 [41:13<3:49:27, 2.61it/s] 25%|██▌ | 12078/48008 [41:13<3:14:43, 3.08it/s] 25%|██▌ | 12079/48008 [41:13<2:50:19, 3.52it/s] 25%|██▌ | 12080/48008 [41:13<2:31:48, 3.94it/s] 25%|██▌ | 12081/48008 [41:14<2:20:53, 4.25it/s] 25%|██▌ | 12082/48008 [41:14<2:14:09, 4.46it/s] 25%|██▌ | 12083/48008 [41:14<2:28:41, 4.03it/s] 25%|██▌ | 12084/48008 [41:14<2:21:05, 4.24it/s] 25%|██▌ | 12085/48008 [41:14<2:11:54, 4.54it/s] 25%|██▌ | 12086/48008 [41:15<2:08:30, 4.66it/s] 25%|██▌ | 12087/48008 [41:15<2:04:27, 4.81it/s] 25%|██▌ | 12088/48008 [41:15<2:01:26, 4.93it/s] 25%|██▌ | 12089/48008 [41:15<2:00:00, 4.99it/s] 25%|██▌ | 12090/48008 [41:15<1:59:26, 5.01it/s] 25%|██▌ | 12091/48008 [41:16<1:57:45, 5.08it/s] 25%|██▌ | 12092/48008 [41:16<1:57:03, 5.11it/s] 25%|██▌ | 12093/48008 [41:16<1:56:09, 5.15it/s] 25%|██▌ | 12094/48008 [41:16<2:30:43, 3.97it/s] 25%|██▌ | 12095/48008 [41:17<2:19:57, 4.28it/s] 25%|██▌ | 12096/48008 [41:17<2:14:28, 4.45it/s] 25%|██▌ | 12097/48008 [41:17<2:07:41, 4.69it/s] 25%|██▌ | 12098/48008 [41:17<2:03:39, 4.84it/s] 25%|██▌ | 12099/48008 [41:17<2:03:14, 4.86it/s] 25%|██▌ | 12100/48008 [41:18<2:09:00, 4.64it/s] {'loss': 0.303, 'grad_norm': 0.15234375, 'learning_rate': 0.00014959590068321947, 'epoch': 0.25} 25%|██▌ | 12100/48008 [41:18<2:09:00, 4.64it/s] 25%|██▌ | 12101/48008 [41:18<2:04:41, 4.80it/s] 25%|██▌ | 12102/48008 [41:18<2:09:45, 4.61it/s] 25%|██▌ | 12103/48008 [41:18<2:08:26, 4.66it/s] 25%|██▌ | 12104/48008 [41:18<2:03:51, 4.83it/s] 25%|██▌ | 12105/48008 [41:19<2:04:13, 4.82it/s] 25%|██▌ | 12106/48008 [41:19<2:00:24, 4.97it/s] 25%|██▌ | 12107/48008 [41:19<1:59:23, 5.01it/s] 25%|██▌ | 12108/48008 [41:19<1:58:03, 5.07it/s] 25%|██▌ | 12109/48008 [41:19<1:56:49, 5.12it/s] 25%|██▌ | 12110/48008 [41:20<1:58:10, 5.06it/s] 25%|██▌ | 12111/48008 [41:20<2:05:25, 4.77it/s] 25%|██▌ | 12112/48008 [41:20<2:01:57, 4.91it/s] 25%|██▌ | 12113/48008 [41:20<1:59:53, 4.99it/s] 25%|██▌ | 12114/48008 [41:20<1:57:38, 5.09it/s] 25%|██▌ | 12115/48008 [41:21<2:00:29, 4.96it/s] 25%|██▌ | 12116/48008 [41:21<2:00:36, 4.96it/s] 25%|██▌ | 12117/48008 [41:21<2:00:43, 4.95it/s] 25%|██▌ | 12118/48008 [41:21<2:06:55, 4.71it/s] 25%|██▌ | 12119/48008 [41:22<2:06:12, 4.74it/s] 25%|██▌ | 12120/48008 [41:22<2:02:43, 4.87it/s] 25%|██▌ | 12121/48008 [41:22<2:00:18, 4.97it/s] 25%|██▌ | 12122/48008 [41:22<1:58:35, 5.04it/s] 25%|██▌ | 12123/48008 [41:22<1:57:28, 5.09it/s] 25%|██▌ | 12124/48008 [41:22<1:57:02, 5.11it/s] 25%|██▌ | 12125/48008 [41:23<2:16:49, 4.37it/s] 25%|██▌ | 12126/48008 [41:23<2:10:00, 4.60it/s] 25%|██▌ | 12127/48008 [41:23<2:05:25, 4.77it/s] 25%|██▌ | 12128/48008 [41:23<2:02:16, 4.89it/s] 25%|██▌ | 12129/48008 [41:24<1:59:19, 5.01it/s] 25%|██▌ | 12130/48008 [41:24<1:57:49, 5.07it/s] 25%|██▌ | 12131/48008 [41:24<1:56:23, 5.14it/s] 25%|██▌ | 12132/48008 [41:24<1:54:37, 5.22it/s] 25%|██▌ | 12133/48008 [41:24<1:54:35, 5.22it/s] 25%|██▌ | 12134/48008 [41:25<2:02:20, 4.89it/s] 25%|██▌ | 12135/48008 [41:25<1:59:12, 5.02it/s] 25%|██▌ | 12136/48008 [41:25<1:57:11, 5.10it/s] 25%|██▌ | 12137/48008 [41:25<1:56:30, 5.13it/s] 25%|██▌ | 12138/48008 [41:25<1:56:45, 5.12it/s] 25%|██▌ | 12139/48008 [41:26<1:56:04, 5.15it/s] 25%|██▌ | 12140/48008 [41:26<1:56:06, 5.15it/s] 25%|██▌ | 12141/48008 [41:26<1:56:56, 5.11it/s] 25%|██▌ | 12142/48008 [41:26<1:56:15, 5.14it/s] 25%|██▌ | 12143/48008 [41:26<1:56:51, 5.12it/s] 25%|██▌ | 12144/48008 [41:26<1:56:23, 5.14it/s] 25%|██▌ | 12145/48008 [41:27<1:57:29, 5.09it/s] 25%|██▌ | 12146/48008 [41:27<1:55:50, 5.16it/s] 25%|██▌ | 12147/48008 [41:27<1:55:38, 5.17it/s] 25%|██▌ | 12148/48008 [41:27<1:54:47, 5.21it/s] 25%|██▌ | 12149/48008 [41:27<1:53:35, 5.26it/s] 25%|██▌ | 12150/48008 [41:28<1:53:12, 5.28it/s] {'loss': 0.3025, 'grad_norm': 0.435546875, 'learning_rate': 0.0001493876020663223, 'epoch': 0.25} 25%|██▌ | 12150/48008 [41:28<1:53:12, 5.28it/s] 25%|██▌ | 12151/48008 [41:28<1:53:56, 5.24it/s] 25%|██▌ | 12152/48008 [41:28<1:54:17, 5.23it/s] 25%|██▌ | 12153/48008 [41:28<1:53:21, 5.27it/s] 25%|██▌ | 12154/48008 [41:28<1:53:51, 5.25it/s] 25%|██▌ | 12155/48008 [41:29<1:54:02, 5.24it/s] 25%|██▌ | 12156/48008 [41:29<1:55:56, 5.15it/s] 25%|██▌ | 12157/48008 [41:29<1:54:31, 5.22it/s] 25%|██▌ | 12158/48008 [41:29<1:57:55, 5.07it/s] 25%|██▌ | 12159/48008 [41:29<1:56:01, 5.15it/s] 25%|██▌ | 12160/48008 [41:30<1:55:16, 5.18it/s] 25%|██▌ | 12161/48008 [41:30<1:54:04, 5.24it/s] 25%|██▌ | 12162/48008 [41:30<1:53:37, 5.26it/s] 25%|██▌ | 12163/48008 [41:30<1:52:48, 5.30it/s] 25%|██▌ | 12164/48008 [41:30<1:54:35, 5.21it/s] 25%|██▌ | 12165/48008 [41:31<1:55:37, 5.17it/s] 25%|██▌ | 12166/48008 [41:31<1:55:02, 5.19it/s] 25%|██▌ | 12167/48008 [41:31<1:56:51, 5.11it/s] 25%|██▌ | 12168/48008 [41:31<1:56:09, 5.14it/s] 25%|██▌ | 12169/48008 [41:31<1:54:52, 5.20it/s] 25%|██▌ | 12170/48008 [41:31<1:56:09, 5.14it/s] 25%|██▌ | 12171/48008 [41:32<1:55:53, 5.15it/s] 25%|██▌ | 12172/48008 [41:32<1:54:42, 5.21it/s] 25%|██▌ | 12173/48008 [41:32<1:56:48, 5.11it/s] 25%|██▌ | 12174/48008 [41:32<1:56:04, 5.15it/s] 25%|██▌ | 12175/48008 [41:32<1:54:15, 5.23it/s] 25%|██▌ | 12176/48008 [41:33<1:54:35, 5.21it/s] 25%|██▌ | 12177/48008 [41:33<1:54:18, 5.22it/s] 25%|██▌ | 12178/48008 [41:33<1:54:09, 5.23it/s] 25%|██▌ | 12179/48008 [41:33<1:53:41, 5.25it/s] 25%|██▌ | 12180/48008 [41:33<1:53:31, 5.26it/s] 25%|██▌ | 12181/48008 [41:34<1:54:41, 5.21it/s] 25%|██▌ | 12182/48008 [41:34<1:54:17, 5.22it/s] 25%|██▌ | 12183/48008 [41:34<1:54:28, 5.22it/s] 25%|██▌ | 12184/48008 [41:34<1:55:12, 5.18it/s] 25%|██▌ | 12185/48008 [41:34<1:53:53, 5.24it/s] 25%|██▌ | 12186/48008 [41:35<1:54:46, 5.20it/s] 25%|██▌ | 12187/48008 [41:35<1:54:39, 5.21it/s] 25%|██▌ | 12188/48008 [41:35<2:14:57, 4.42it/s] 25%|██▌ | 12189/48008 [41:35<2:29:29, 3.99it/s] 25%|██▌ | 12190/48008 [41:36<2:20:45, 4.24it/s] 25%|██▌ | 12191/48008 [41:36<2:12:36, 4.50it/s] 25%|██▌ | 12192/48008 [41:36<2:09:12, 4.62it/s] 25%|██▌ | 12193/48008 [41:36<2:03:48, 4.82it/s] 25%|██▌ | 12194/48008 [41:36<2:00:45, 4.94it/s] 25%|██▌ | 12195/48008 [41:37<1:58:21, 5.04it/s] 25%|██▌ | 12196/48008 [41:37<1:56:57, 5.10it/s] 25%|██▌ | 12197/48008 [41:37<1:56:23, 5.13it/s] 25%|██▌ | 12198/48008 [41:37<1:55:36, 5.16it/s] 25%|██▌ | 12199/48008 [41:37<1:55:05, 5.19it/s] 25%|██▌ | 12200/48008 [41:38<2:15:45, 4.40it/s] {'loss': 0.3032, 'grad_norm': 0.82421875, 'learning_rate': 0.0001491793034494251, 'epoch': 0.25} 25%|██▌ | 12200/48008 [41:38<2:15:45, 4.40it/s] 25%|██▌ | 12201/48008 [41:38<2:08:36, 4.64it/s] 25%|██▌ | 12202/48008 [41:38<2:12:42, 4.50it/s] 25%|██▌ | 12203/48008 [41:38<2:07:02, 4.70it/s] 25%|██▌ | 12204/48008 [41:38<2:03:32, 4.83it/s] 25%|██▌ | 12205/48008 [41:39<2:02:15, 4.88it/s] 25%|██▌ | 12206/48008 [41:39<2:01:55, 4.89it/s] 25%|██▌ | 12207/48008 [41:39<1:58:46, 5.02it/s] 25%|██▌ | 12208/48008 [41:39<1:57:21, 5.08it/s] 25%|██▌ | 12209/48008 [41:39<1:57:15, 5.09it/s] 25%|██▌ | 12210/48008 [41:40<1:56:01, 5.14it/s] 25%|██▌ | 12211/48008 [41:40<1:55:51, 5.15it/s] 25%|██▌ | 12212/48008 [41:40<1:55:19, 5.17it/s] 25%|██▌ | 12213/48008 [41:40<1:53:47, 5.24it/s] 25%|██▌ | 12214/48008 [41:40<1:53:38, 5.25it/s] 25%|██▌ | 12215/48008 [41:41<1:55:01, 5.19it/s] 25%|██▌ | 12216/48008 [41:41<1:55:36, 5.16it/s] 25%|██▌ | 12217/48008 [41:41<1:54:22, 5.22it/s] 25%|██▌ | 12218/48008 [41:41<1:54:28, 5.21it/s] 25%|██▌ | 12219/48008 [41:41<1:54:19, 5.22it/s] 25%|██▌ | 12220/48008 [41:41<1:55:05, 5.18it/s] 25%|██▌ | 12221/48008 [41:42<1:56:42, 5.11it/s] 25%|██▌ | 12222/48008 [41:42<1:55:30, 5.16it/s] 25%|██▌ | 12223/48008 [41:42<1:56:03, 5.14it/s] 25%|██▌ | 12224/48008 [41:42<1:55:29, 5.16it/s] 25%|██▌ | 12225/48008 [41:42<1:57:16, 5.09it/s] 25%|██▌ | 12226/48008 [41:43<1:59:40, 4.98it/s] 25%|██▌ | 12227/48008 [41:43<1:58:52, 5.02it/s] 25%|██▌ | 12228/48008 [41:43<2:00:34, 4.95it/s] 25%|██▌ | 12229/48008 [41:43<1:58:03, 5.05it/s] 25%|██▌ | 12230/48008 [41:43<1:56:08, 5.13it/s] 25%|██▌ | 12231/48008 [41:44<1:55:54, 5.14it/s] 25%|██▌ | 12232/48008 [41:44<1:56:40, 5.11it/s] 25%|██▌ | 12233/48008 [41:44<1:56:21, 5.12it/s] 25%|██▌ | 12234/48008 [41:44<1:55:54, 5.14it/s] 25%|██▌ | 12235/48008 [41:44<2:03:24, 4.83it/s] 25%|██▌ | 12236/48008 [41:45<1:59:46, 4.98it/s] 25%|██▌ | 12237/48008 [41:45<1:57:55, 5.06it/s] 25%|██▌ | 12238/48008 [41:45<1:57:44, 5.06it/s] 25%|██▌ | 12239/48008 [41:45<1:59:53, 4.97it/s] 25%|██▌ | 12240/48008 [41:45<1:59:13, 5.00it/s] 25%|██▌ | 12241/48008 [41:46<1:57:52, 5.06it/s] 25%|██▌ | 12242/48008 [41:46<1:56:18, 5.13it/s] 26%|██▌ | 12243/48008 [41:46<1:58:04, 5.05it/s] 26%|██▌ | 12244/48008 [41:46<1:56:45, 5.11it/s] 26%|██▌ | 12245/48008 [41:46<1:55:04, 5.18it/s] 26%|██▌ | 12246/48008 [41:47<1:54:05, 5.22it/s] 26%|██▌ | 12247/48008 [41:47<1:55:15, 5.17it/s] 26%|██▌ | 12248/48008 [41:47<1:54:57, 5.18it/s] 26%|██▌ | 12249/48008 [41:47<1:54:51, 5.19it/s] 26%|██▌ | 12250/48008 [41:47<1:57:57, 5.05it/s] {'loss': 0.3017, 'grad_norm': 0.2333984375, 'learning_rate': 0.00014897100483252793, 'epoch': 0.26} 26%|██▌ | 12250/48008 [41:47<1:57:57, 5.05it/s] 26%|██▌ | 12251/48008 [41:48<1:58:05, 5.05it/s] 26%|██▌ | 12252/48008 [41:48<1:57:33, 5.07it/s] 26%|██▌ | 12253/48008 [41:48<1:58:13, 5.04it/s] 26%|██▌ | 12254/48008 [41:48<1:57:45, 5.06it/s] 26%|██▌ | 12255/48008 [41:48<1:56:47, 5.10it/s] 26%|██▌ | 12256/48008 [41:49<1:55:07, 5.18it/s] 26%|██▌ | 12257/48008 [41:49<1:53:49, 5.23it/s] 26%|██▌ | 12258/48008 [41:49<1:54:50, 5.19it/s] 26%|██▌ | 12259/48008 [41:49<1:54:16, 5.21it/s] 26%|██▌ | 12260/48008 [41:49<1:56:11, 5.13it/s] 26%|██▌ | 12261/48008 [41:50<1:54:25, 5.21it/s] 26%|██▌ | 12262/48008 [41:50<1:55:22, 5.16it/s] 26%|██▌ | 12263/48008 [41:50<1:55:10, 5.17it/s] 26%|██▌ | 12264/48008 [41:50<1:54:36, 5.20it/s] 26%|██▌ | 12265/48008 [41:50<1:54:20, 5.21it/s] 26%|██▌ | 12266/48008 [41:50<1:54:40, 5.19it/s] 26%|██▌ | 12267/48008 [41:51<1:54:34, 5.20it/s] 26%|██▌ | 12268/48008 [41:51<1:56:26, 5.12it/s] 26%|██▌ | 12269/48008 [41:51<1:55:44, 5.15it/s] 26%|██▌ | 12270/48008 [41:51<1:54:11, 5.22it/s] 26%|██▌ | 12271/48008 [41:51<1:54:00, 5.22it/s] 26%|██▌ | 12272/48008 [41:52<1:54:48, 5.19it/s] 26%|██▌ | 12273/48008 [41:52<1:55:09, 5.17it/s] 26%|██▌ | 12274/48008 [41:52<2:03:03, 4.84it/s] 26%|██▌ | 12275/48008 [41:52<2:02:33, 4.86it/s] 26%|██▌ | 12276/48008 [41:52<1:59:25, 4.99it/s] 26%|██▌ | 12277/48008 [41:53<1:55:58, 5.13it/s] 26%|██▌ | 12278/48008 [41:53<1:55:02, 5.18it/s] 26%|██▌ | 12279/48008 [41:53<1:55:59, 5.13it/s] 26%|██▌ | 12280/48008 [41:53<1:57:21, 5.07it/s] 26%|██▌ | 12281/48008 [41:53<1:57:28, 5.07it/s] 26%|██▌ | 12282/48008 [41:54<1:55:48, 5.14it/s] 26%|██▌ | 12283/48008 [41:54<1:55:23, 5.16it/s] 26%|██▌ | 12284/48008 [41:54<1:55:05, 5.17it/s] 26%|██▌ | 12285/48008 [41:54<1:54:24, 5.20it/s] 26%|██▌ | 12286/48008 [41:54<1:55:43, 5.14it/s] 26%|██▌ | 12287/48008 [41:55<1:56:24, 5.11it/s] 26%|██▌ | 12288/48008 [41:55<1:55:42, 5.15it/s] 26%|██▌ | 12289/48008 [41:55<1:57:11, 5.08it/s] 26%|██▌ | 12290/48008 [41:55<1:54:26, 5.20it/s] 26%|██▌ | 12291/48008 [41:55<2:02:29, 4.86it/s] 26%|██▌ | 12292/48008 [41:56<2:00:55, 4.92it/s] 26%|██▌ | 12293/48008 [41:56<2:01:16, 4.91it/s] 26%|██▌ | 12294/48008 [41:56<2:00:17, 4.95it/s] 26%|██▌ | 12295/48008 [41:56<1:58:10, 5.04it/s] 26%|██▌ | 12296/48008 [41:56<2:00:51, 4.92it/s] 26%|██▌ | 12297/48008 [41:57<2:00:16, 4.95it/s] 26%|██▌ | 12298/48008 [41:57<2:00:43, 4.93it/s] 26%|██▌ | 12299/48008 [41:57<2:19:51, 4.26it/s] 26%|██▌ | 12300/48008 [41:57<2:21:12, 4.21it/s] {'loss': 0.3004, 'grad_norm': 0.255859375, 'learning_rate': 0.00014876270621563074, 'epoch': 0.26} 26%|██▌ | 12300/48008 [41:57<2:21:12, 4.21it/s] 26%|██▌ | 12301/48008 [41:58<2:14:29, 4.42it/s] 26%|██▌ | 12302/48008 [41:58<2:09:54, 4.58it/s] 26%|██▌ | 12303/48008 [41:58<2:05:06, 4.76it/s] 26%|██▌ | 12304/48008 [41:58<2:22:12, 4.18it/s] 26%|██▌ | 12305/48008 [41:58<2:13:51, 4.45it/s] 26%|██▌ | 12306/48008 [41:59<2:07:54, 4.65it/s] 26%|██▌ | 12307/48008 [41:59<2:03:19, 4.82it/s] 26%|██▌ | 12308/48008 [41:59<2:00:29, 4.94it/s] 26%|██▌ | 12309/48008 [41:59<2:00:11, 4.95it/s] 26%|██▌ | 12310/48008 [41:59<1:58:13, 5.03it/s] 26%|██▌ | 12311/48008 [42:00<1:58:08, 5.04it/s] 26%|██▌ | 12312/48008 [42:00<2:00:32, 4.94it/s] 26%|██▌ | 12313/48008 [42:00<1:59:00, 5.00it/s] 26%|██▌ | 12314/48008 [42:00<2:07:01, 4.68it/s] 26%|██▌ | 12315/48008 [42:00<2:02:53, 4.84it/s] 26%|██▌ | 12316/48008 [42:01<2:00:16, 4.95it/s] 26%|██▌ | 12317/48008 [42:01<1:57:10, 5.08it/s] 26%|██▌ | 12318/48008 [42:01<1:56:21, 5.11it/s] 26%|██▌ | 12319/48008 [42:01<1:55:26, 5.15it/s] 26%|██▌ | 12320/48008 [42:01<1:54:25, 5.20it/s] 26%|██▌ | 12321/48008 [42:02<1:54:49, 5.18it/s] 26%|██▌ | 12322/48008 [42:02<1:54:54, 5.18it/s] 26%|██▌ | 12323/48008 [42:02<1:55:50, 5.13it/s] 26%|██▌ | 12324/48008 [42:02<2:03:24, 4.82it/s] 26%|██▌ | 12325/48008 [42:02<2:00:50, 4.92it/s] 26%|██▌ | 12326/48008 [42:03<1:58:29, 5.02it/s] 26%|██▌ | 12327/48008 [42:03<1:56:56, 5.09it/s] 26%|██▌ | 12328/48008 [42:03<1:54:18, 5.20it/s] 26%|██▌ | 12329/48008 [42:03<2:28:56, 3.99it/s] 26%|██▌ | 12330/48008 [42:04<2:19:18, 4.27it/s] 26%|██▌ | 12331/48008 [42:04<2:11:35, 4.52it/s] 26%|██▌ | 12332/48008 [42:04<2:06:28, 4.70it/s] 26%|██▌ | 12333/48008 [42:04<2:02:02, 4.87it/s] 26%|██▌ | 12334/48008 [42:04<2:00:19, 4.94it/s] 26%|██▌ | 12335/48008 [42:05<1:58:30, 5.02it/s] 26%|██▌ | 12336/48008 [42:05<1:57:06, 5.08it/s] 26%|██▌ | 12337/48008 [42:05<1:56:06, 5.12it/s] 26%|██▌ | 12338/48008 [42:05<1:55:32, 5.15it/s] 26%|██▌ | 12339/48008 [42:05<1:55:01, 5.17it/s] 26%|██▌ | 12340/48008 [42:05<1:54:08, 5.21it/s] 26%|██▌ | 12341/48008 [42:06<2:14:37, 4.42it/s] 26%|██▌ | 12342/48008 [42:06<2:11:31, 4.52it/s] 26%|██▌ | 12343/48008 [42:06<2:05:53, 4.72it/s] 26%|██▌ | 12344/48008 [42:06<2:10:36, 4.55it/s] 26%|██▌ | 12345/48008 [42:07<2:04:41, 4.77it/s] 26%|██▌ | 12346/48008 [42:07<2:01:39, 4.89it/s] 26%|██▌ | 12347/48008 [42:07<2:00:18, 4.94it/s] 26%|██▌ | 12348/48008 [42:07<2:00:13, 4.94it/s] 26%|██▌ | 12349/48008 [42:07<1:59:23, 4.98it/s] 26%|██▌ | 12350/48008 [42:08<1:59:49, 4.96it/s] {'loss': 0.2997, 'grad_norm': 0.2451171875, 'learning_rate': 0.00014855440759873356, 'epoch': 0.26} 26%|██▌ | 12350/48008 [42:08<1:59:49, 4.96it/s] 26%|██▌ | 12351/48008 [42:08<2:00:17, 4.94it/s] 26%|██▌ | 12352/48008 [42:08<1:58:35, 5.01it/s] 26%|██▌ | 12353/48008 [42:08<2:00:32, 4.93it/s] 26%|██▌ | 12354/48008 [42:08<1:57:54, 5.04it/s] 26%|██▌ | 12355/48008 [42:09<1:56:40, 5.09it/s] 26%|██▌ | 12356/48008 [42:09<1:55:16, 5.15it/s] 26%|██▌ | 12357/48008 [42:09<1:54:22, 5.20it/s] 26%|██▌ | 12358/48008 [42:09<1:54:01, 5.21it/s] 26%|██▌ | 12359/48008 [42:09<1:53:49, 5.22it/s] 26%|██▌ | 12360/48008 [42:10<1:53:00, 5.26it/s] 26%|██▌ | 12361/48008 [42:10<2:13:40, 4.44it/s] 26%|██▌ | 12362/48008 [42:10<2:08:01, 4.64it/s] 26%|██▌ | 12363/48008 [42:10<2:04:03, 4.79it/s] 26%|██▌ | 12364/48008 [42:10<2:02:23, 4.85it/s] 26%|██▌ | 12365/48008 [42:11<1:59:08, 4.99it/s] 26%|██▌ | 12366/48008 [42:11<1:57:29, 5.06it/s] 26%|██▌ | 12367/48008 [42:11<2:16:51, 4.34it/s] 26%|██▌ | 12368/48008 [42:11<2:09:26, 4.59it/s] 26%|██▌ | 12369/48008 [42:12<2:03:41, 4.80it/s] 26%|██▌ | 12370/48008 [42:12<2:21:12, 4.21it/s] 26%|██▌ | 12371/48008 [42:12<2:11:41, 4.51it/s] 26%|██▌ | 12372/48008 [42:12<2:05:29, 4.73it/s] 26%|██▌ | 12373/48008 [42:12<2:01:36, 4.88it/s] 26%|██▌ | 12374/48008 [42:13<1:59:37, 4.96it/s] 26%|██▌ | 12375/48008 [42:13<1:57:58, 5.03it/s] 26%|██▌ | 12376/48008 [42:13<1:57:54, 5.04it/s] 26%|██▌ | 12377/48008 [42:13<1:56:53, 5.08it/s] 26%|██▌ | 12378/48008 [42:13<2:03:53, 4.79it/s] 26%|██▌ | 12379/48008 [42:14<2:01:38, 4.88it/s] 26%|██▌ | 12380/48008 [42:14<2:02:41, 4.84it/s] 26%|██▌ | 12381/48008 [42:14<2:02:13, 4.86it/s] 26%|██▌ | 12382/48008 [42:14<1:59:31, 4.97it/s] 26%|██▌ | 12383/48008 [42:14<1:57:45, 5.04it/s] 26%|██▌ | 12384/48008 [42:15<1:57:46, 5.04it/s] 26%|██▌ | 12385/48008 [42:15<1:56:50, 5.08it/s] 26%|██▌ | 12386/48008 [42:15<1:56:19, 5.10it/s] 26%|██▌ | 12387/48008 [42:15<1:57:26, 5.06it/s] 26%|██▌ | 12388/48008 [42:15<1:57:02, 5.07it/s] 26%|██▌ | 12389/48008 [42:16<1:55:24, 5.14it/s] 26%|██▌ | 12390/48008 [42:16<1:54:49, 5.17it/s] 26%|██▌ | 12391/48008 [42:16<2:02:58, 4.83it/s] 26%|██▌ | 12392/48008 [42:16<2:02:13, 4.86it/s] 26%|██▌ | 12393/48008 [42:16<2:00:02, 4.94it/s] 26%|██▌ | 12394/48008 [42:17<1:57:52, 5.04it/s] 26%|██▌ | 12395/48008 [42:17<2:31:25, 3.92it/s] 26%|██▌ | 12396/48008 [42:17<2:22:31, 4.16it/s] 26%|██▌ | 12397/48008 [42:17<2:13:14, 4.45it/s] 26%|██▌ | 12398/48008 [42:18<2:07:52, 4.64it/s] 26%|██▌ | 12399/48008 [42:18<2:03:45, 4.80it/s] 26%|██▌ | 12400/48008 [42:18<2:00:46, 4.91it/s] {'loss': 0.2999, 'grad_norm': 0.373046875, 'learning_rate': 0.00014834610898183635, 'epoch': 0.26} 26%|██▌ | 12400/48008 [42:18<2:00:46, 4.91it/s] 26%|██▌ | 12401/48008 [42:18<2:00:10, 4.94it/s] 26%|██▌ | 12402/48008 [42:18<1:57:29, 5.05it/s] 26%|██▌ | 12403/48008 [42:19<1:55:33, 5.14it/s] 26%|██▌ | 12404/48008 [42:19<1:55:08, 5.15it/s] 26%|██▌ | 12405/48008 [42:19<1:55:50, 5.12it/s] 26%|██▌ | 12406/48008 [42:19<1:55:28, 5.14it/s] 26%|██▌ | 12407/48008 [42:19<1:56:57, 5.07it/s] 26%|██▌ | 12408/48008 [42:19<1:56:03, 5.11it/s] 26%|██▌ | 12409/48008 [42:20<2:03:39, 4.80it/s] 26%|██▌ | 12410/48008 [42:20<2:01:07, 4.90it/s] 26%|██▌ | 12411/48008 [42:20<2:01:53, 4.87it/s] 26%|██▌ | 12412/48008 [42:20<2:02:58, 4.82it/s] 26%|██▌ | 12413/48008 [42:21<2:00:01, 4.94it/s] 26%|██▌ | 12414/48008 [42:21<2:01:15, 4.89it/s] 26%|██▌ | 12415/48008 [42:21<1:59:06, 4.98it/s] 26%|██▌ | 12416/48008 [42:21<1:59:25, 4.97it/s] 26%|██▌ | 12417/48008 [42:21<1:59:25, 4.97it/s] 26%|██▌ | 12418/48008 [42:22<2:01:27, 4.88it/s] 26%|██▌ | 12419/48008 [42:22<1:59:59, 4.94it/s] 26%|██▌ | 12420/48008 [42:22<1:58:19, 5.01it/s] 26%|██▌ | 12421/48008 [42:22<1:55:46, 5.12it/s] 26%|██▌ | 12422/48008 [42:22<1:55:04, 5.15it/s] 26%|██▌ | 12423/48008 [42:23<2:02:56, 4.82it/s] 26%|██▌ | 12424/48008 [42:23<2:00:33, 4.92it/s] 26%|██▌ | 12425/48008 [42:23<2:00:27, 4.92it/s] 26%|██▌ | 12426/48008 [42:23<1:58:14, 5.02it/s] 26%|██▌ | 12427/48008 [42:23<2:00:49, 4.91it/s] 26%|██▌ | 12428/48008 [42:24<1:58:35, 5.00it/s] 26%|██▌ | 12429/48008 [42:24<1:56:01, 5.11it/s] 26%|██▌ | 12430/48008 [42:24<1:58:40, 5.00it/s] 26%|██▌ | 12431/48008 [42:24<1:57:13, 5.06it/s] 26%|██▌ | 12432/48008 [42:24<1:59:27, 4.96it/s] 26%|██▌ | 12433/48008 [42:25<1:56:15, 5.10it/s] 26%|██▌ | 12434/48008 [42:25<1:56:32, 5.09it/s] 26%|██▌ | 12435/48008 [42:25<1:55:13, 5.15it/s] 26%|██▌ | 12436/48008 [42:25<1:55:01, 5.15it/s] 26%|██▌ | 12437/48008 [42:25<1:54:51, 5.16it/s] 26%|██▌ | 12438/48008 [42:25<1:55:45, 5.12it/s] 26%|██▌ | 12439/48008 [42:26<1:55:25, 5.14it/s] 26%|██▌ | 12440/48008 [42:26<1:54:52, 5.16it/s] 26%|██▌ | 12441/48008 [42:26<1:54:50, 5.16it/s] 26%|██▌ | 12442/48008 [42:26<2:03:09, 4.81it/s] 26%|██▌ | 12443/48008 [42:27<2:00:29, 4.92it/s] 26%|██▌ | 12444/48008 [42:27<1:58:25, 5.01it/s] 26%|██▌ | 12445/48008 [42:27<2:00:05, 4.94it/s] 26%|██▌ | 12446/48008 [42:27<1:58:27, 5.00it/s] 26%|██▌ | 12447/48008 [42:27<1:57:21, 5.05it/s] 26%|██▌ | 12448/48008 [42:27<1:57:11, 5.06it/s] 26%|██▌ | 12449/48008 [42:28<2:04:49, 4.75it/s] 26%|██▌ | 12450/48008 [42:28<2:02:26, 4.84it/s] {'loss': 0.3017, 'grad_norm': 0.875, 'learning_rate': 0.0001481378103649392, 'epoch': 0.26} 26%|██▌ | 12450/48008 [42:28<2:02:26, 4.84it/s] 26%|██▌ | 12451/48008 [42:28<1:59:26, 4.96it/s] 26%|██▌ | 12452/48008 [42:29<2:32:37, 3.88it/s] 26%|██▌ | 12453/48008 [42:29<2:20:32, 4.22it/s] 26%|██▌ | 12454/48008 [42:29<2:13:07, 4.45it/s] 26%|██▌ | 12455/48008 [42:29<2:06:51, 4.67it/s] 26%|██▌ | 12456/48008 [42:29<2:02:41, 4.83it/s] 26%|██▌ | 12457/48008 [42:29<2:03:19, 4.80it/s] 26%|██▌ | 12458/48008 [42:30<2:02:50, 4.82it/s] 26%|██▌ | 12459/48008 [42:30<2:00:22, 4.92it/s] 26%|██▌ | 12460/48008 [42:30<1:59:06, 4.97it/s] 26%|██▌ | 12461/48008 [42:30<1:57:41, 5.03it/s] 26%|██▌ | 12462/48008 [42:30<1:56:09, 5.10it/s] 26%|██▌ | 12463/48008 [42:31<1:56:11, 5.10it/s] 26%|██▌ | 12464/48008 [42:31<1:55:41, 5.12it/s] 26%|██▌ | 12465/48008 [42:31<1:55:39, 5.12it/s] 26%|██▌ | 12466/48008 [42:31<1:55:08, 5.14it/s] 26%|██▌ | 12467/48008 [42:31<1:54:23, 5.18it/s] 26%|██▌ | 12468/48008 [42:32<1:56:06, 5.10it/s] 26%|██▌ | 12469/48008 [42:33<4:02:51, 2.44it/s] 26%|██▌ | 12470/48008 [42:33<3:24:49, 2.89it/s] 26%|██▌ | 12471/48008 [42:33<2:57:33, 3.34it/s] 26%|██▌ | 12472/48008 [42:33<2:39:37, 3.71it/s] 26%|██▌ | 12473/48008 [42:33<2:24:07, 4.11it/s] 26%|██▌ | 12474/48008 [42:33<2:14:35, 4.40it/s] 26%|██▌ | 12475/48008 [42:34<2:07:28, 4.65it/s] 26%|██▌ | 12476/48008 [42:34<2:03:20, 4.80it/s] 26%|██▌ | 12477/48008 [42:34<2:01:44, 4.86it/s] 26%|██▌ | 12478/48008 [42:34<1:59:56, 4.94it/s] 26%|██▌ | 12479/48008 [42:34<1:58:21, 5.00it/s] 26%|██▌ | 12480/48008 [42:35<1:57:38, 5.03it/s] 26%|██▌ | 12481/48008 [42:35<2:31:17, 3.91it/s] 26%|██▌ | 12482/48008 [42:35<2:19:56, 4.23it/s] 26%|██▌ | 12483/48008 [42:35<2:15:26, 4.37it/s] 26%|██▌ | 12484/48008 [42:36<2:09:20, 4.58it/s] 26%|██▌ | 12485/48008 [42:36<2:05:44, 4.71it/s] 26%|██▌ | 12486/48008 [42:36<2:02:07, 4.85it/s] 26%|██▌ | 12487/48008 [42:36<1:59:19, 4.96it/s] 26%|██▌ | 12488/48008 [42:36<1:57:37, 5.03it/s] 26%|██▌ | 12489/48008 [42:37<1:58:15, 5.01it/s] 26%|██▌ | 12490/48008 [42:37<1:56:39, 5.07it/s] 26%|██▌ | 12491/48008 [42:37<1:57:11, 5.05it/s] 26%|██▌ | 12492/48008 [42:37<1:57:49, 5.02it/s] 26%|██▌ | 12493/48008 [42:37<1:58:18, 5.00it/s] 26%|██▌ | 12494/48008 [42:38<1:56:29, 5.08it/s] 26%|██▌ | 12495/48008 [42:38<2:04:10, 4.77it/s] 26%|██▌ | 12496/48008 [42:38<2:01:17, 4.88it/s] 26%|██▌ | 12497/48008 [42:38<1:58:59, 4.97it/s] 26%|██▌ | 12498/48008 [42:38<1:57:48, 5.02it/s] 26%|██▌ | 12499/48008 [42:39<1:57:54, 5.02it/s] 26%|██▌ | 12500/48008 [42:39<1:55:29, 5.12it/s] {'loss': 0.3024, 'grad_norm': 0.318359375, 'learning_rate': 0.00014792951174804199, 'epoch': 0.26} 26%|██▌ | 12500/48008 [42:39<1:55:29, 5.12it/s] 26%|██▌ | 12501/48008 [42:39<2:03:17, 4.80it/s] 26%|██▌ | 12502/48008 [42:39<2:00:17, 4.92it/s] 26%|██▌ | 12503/48008 [42:39<1:59:59, 4.93it/s] 26%|██▌ | 12504/48008 [42:40<1:58:50, 4.98it/s] 26%|██▌ | 12505/48008 [42:40<1:57:02, 5.06it/s] 26%|██▌ | 12506/48008 [42:40<1:56:37, 5.07it/s] 26%|██▌ | 12507/48008 [42:40<2:03:41, 4.78it/s] 26%|██▌ | 12508/48008 [42:40<2:00:06, 4.93it/s] 26%|██▌ | 12509/48008 [42:41<1:57:18, 5.04it/s] 26%|██▌ | 12510/48008 [42:41<1:55:39, 5.12it/s] 26%|██▌ | 12511/48008 [42:41<1:56:53, 5.06it/s] 26%|██▌ | 12512/48008 [42:41<1:56:37, 5.07it/s] 26%|██▌ | 12513/48008 [42:41<1:55:34, 5.12it/s] 26%|██▌ | 12514/48008 [42:42<1:55:24, 5.13it/s] 26%|██▌ | 12515/48008 [42:42<1:53:14, 5.22it/s] 26%|██▌ | 12516/48008 [42:42<1:56:55, 5.06it/s] 26%|██▌ | 12517/48008 [42:42<2:16:17, 4.34it/s] 26%|██▌ | 12518/48008 [42:43<2:11:15, 4.51it/s] 26%|██▌ | 12519/48008 [42:43<2:06:42, 4.67it/s] 26%|██▌ | 12520/48008 [42:43<2:02:41, 4.82it/s] 26%|██▌ | 12521/48008 [42:43<2:00:14, 4.92it/s] 26%|██▌ | 12522/48008 [42:43<1:56:43, 5.07it/s] 26%|██▌ | 12523/48008 [42:43<1:54:52, 5.15it/s] 26%|██▌ | 12524/48008 [42:44<1:57:19, 5.04it/s] 26%|██▌ | 12525/48008 [42:44<1:58:08, 5.01it/s] 26%|██▌ | 12526/48008 [42:44<1:56:27, 5.08it/s] 26%|██▌ | 12527/48008 [42:44<2:30:09, 3.94it/s] 26%|██▌ | 12528/48008 [42:45<2:20:05, 4.22it/s] 26%|██▌ | 12529/48008 [42:45<2:12:21, 4.47it/s] 26%|██▌ | 12530/48008 [42:45<2:06:49, 4.66it/s] 26%|██▌ | 12531/48008 [42:45<2:02:31, 4.83it/s] 26%|██▌ | 12532/48008 [42:45<1:58:31, 4.99it/s] 26%|██▌ | 12533/48008 [42:46<1:56:13, 5.09it/s] 26%|██▌ | 12534/48008 [42:46<1:55:05, 5.14it/s] 26%|██▌ | 12535/48008 [42:46<1:54:46, 5.15it/s] 26%|██▌ | 12536/48008 [42:46<1:53:52, 5.19it/s] 26%|██▌ | 12537/48008 [42:46<1:53:50, 5.19it/s] 26%|██▌ | 12538/48008 [42:47<2:01:12, 4.88it/s] 26%|██▌ | 12539/48008 [42:47<1:58:30, 4.99it/s] 26%|██▌ | 12540/48008 [42:47<1:55:55, 5.10it/s] 26%|██▌ | 12541/48008 [42:47<1:54:57, 5.14it/s] 26%|██▌ | 12542/48008 [42:47<1:53:23, 5.21it/s] 26%|██▌ | 12543/48008 [42:48<1:52:59, 5.23it/s] 26%|██▌ | 12544/48008 [42:48<1:56:08, 5.09it/s] 26%|██▌ | 12545/48008 [42:48<1:55:28, 5.12it/s] 26%|██▌ | 12546/48008 [42:48<2:14:55, 4.38it/s] 26%|██▌ | 12547/48008 [42:48<2:10:31, 4.53it/s] 26%|██▌ | 12548/48008 [42:49<2:05:27, 4.71it/s] 26%|██▌ | 12549/48008 [42:49<2:03:28, 4.79it/s] 26%|██▌ | 12550/48008 [42:49<1:59:23, 4.95it/s] 26%|██▌ | 12550/48008 [42:49<1:59:23, 4.95it/s]{'loss': 0.302, 'grad_norm': 0.427734375, 'learning_rate': 0.00014772121313114483, 'epoch': 0.26} 26%|██▌ | 12551/48008 [42:49<1:57:30, 5.03it/s] 26%|██▌ | 12552/48008 [42:49<1:56:19, 5.08it/s] 26%|██▌ | 12553/48008 [42:50<1:54:29, 5.16it/s] 26%|██▌ | 12554/48008 [42:50<1:55:14, 5.13it/s] 26%|██▌ | 12555/48008 [42:50<1:56:02, 5.09it/s] 26%|██▌ | 12556/48008 [42:50<1:55:12, 5.13it/s] 26%|██▌ | 12557/48008 [42:50<1:56:29, 5.07it/s] 26%|██▌ | 12558/48008 [42:51<1:56:08, 5.09it/s] 26%|██▌ | 12559/48008 [42:51<1:55:08, 5.13it/s] 26%|██▌ | 12560/48008 [42:51<1:53:53, 5.19it/s] 26%|██▌ | 12561/48008 [42:51<1:53:30, 5.20it/s] 26%|██▌ | 12562/48008 [42:51<1:52:00, 5.27it/s] 26%|██▌ | 12563/48008 [42:52<1:51:43, 5.29it/s] 26%|██▌ | 12564/48008 [42:52<2:26:41, 4.03it/s] 26%|██▌ | 12565/48008 [42:52<2:15:49, 4.35it/s] 26%|██▌ | 12566/48008 [42:52<2:09:20, 4.57it/s] 26%|██▌ | 12567/48008 [42:52<2:04:23, 4.75it/s] 26%|██▌ | 12568/48008 [42:53<2:00:29, 4.90it/s] 26%|██▌ | 12569/48008 [42:53<1:59:09, 4.96it/s] 26%|██▌ | 12570/48008 [42:53<1:57:24, 5.03it/s] 26%|██▌ | 12571/48008 [42:53<1:55:04, 5.13it/s] 26%|██▌ | 12572/48008 [42:53<1:54:29, 5.16it/s] 26%|██▌ | 12573/48008 [42:54<1:53:59, 5.18it/s] 26%|██▌ | 12574/48008 [42:54<1:53:25, 5.21it/s] 26%|██▌ | 12575/48008 [42:54<1:53:36, 5.20it/s] 26%|██▌ | 12576/48008 [42:54<1:53:44, 5.19it/s] 26%|██▌ | 12577/48008 [42:54<1:54:17, 5.17it/s] 26%|██▌ | 12578/48008 [42:55<1:53:41, 5.19it/s] 26%|██▌ | 12579/48008 [42:55<1:53:32, 5.20it/s] 26%|██▌ | 12580/48008 [42:55<2:28:16, 3.98it/s] 26%|██▌ | 12581/48008 [42:55<2:17:45, 4.29it/s] 26%|██▌ | 12582/48008 [42:56<2:10:12, 4.53it/s] 26%|██▌ | 12583/48008 [42:56<2:05:41, 4.70it/s] 26%|██▌ | 12584/48008 [42:56<2:09:54, 4.54it/s] 26%|██▌ | 12585/48008 [42:56<2:05:10, 4.72it/s] 26%|██▌ | 12586/48008 [42:56<2:01:21, 4.86it/s] 26%|██▌ | 12587/48008 [42:57<2:33:21, 3.85it/s] 26%|██▌ | 12588/48008 [42:57<2:23:17, 4.12it/s] 26%|██▌ | 12589/48008 [42:57<2:14:26, 4.39it/s] 26%|██▌ | 12590/48008 [42:57<2:07:31, 4.63it/s] 26%|██▌ | 12591/48008 [42:58<2:03:41, 4.77it/s] 26%|██▌ | 12592/48008 [42:58<1:59:27, 4.94it/s] 26%|██▌ | 12593/48008 [42:58<1:59:21, 4.95it/s] 26%|██▌ | 12594/48008 [42:58<1:59:03, 4.96it/s] 26%|██▌ | 12595/48008 [42:59<2:32:18, 3.88it/s] 26%|██▌ | 12596/48008 [42:59<2:20:03, 4.21it/s] 26%|██▌ | 12597/48008 [42:59<2:11:57, 4.47it/s] 26%|██▌ | 12598/48008 [42:59<2:08:39, 4.59it/s] 26%|██▌ | 12599/48008 [42:59<2:12:24, 4.46it/s] 26%|██▌ | 12600/48008 [43:00<2:06:09, 4.68it/s] {'loss': 0.3017, 'grad_norm': 0.298828125, 'learning_rate': 0.00014751291451424762, 'epoch': 0.26} 26%|██▌ | 12600/48008 [43:00<2:06:09, 4.68it/s] 26%|██▌ | 12601/48008 [43:00<2:02:18, 4.82it/s] 26%|██▌ | 12602/48008 [43:00<2:00:54, 4.88it/s] 26%|██▋ | 12603/48008 [43:00<1:58:43, 4.97it/s] 26%|██▋ | 12604/48008 [43:00<1:57:57, 5.00it/s] 26%|██▋ | 12605/48008 [43:01<1:57:27, 5.02it/s] 26%|██▋ | 12606/48008 [43:01<1:57:59, 5.00it/s] 26%|██▋ | 12607/48008 [43:01<1:56:28, 5.07it/s] 26%|██▋ | 12608/48008 [43:01<1:56:08, 5.08it/s] 26%|██▋ | 12609/48008 [43:01<1:55:05, 5.13it/s] 26%|██▋ | 12610/48008 [43:02<2:02:31, 4.81it/s] 26%|██▋ | 12611/48008 [43:02<2:01:25, 4.86it/s] 26%|██▋ | 12612/48008 [43:02<2:01:09, 4.87it/s] 26%|██▋ | 12613/48008 [43:02<2:00:16, 4.91it/s] 26%|██▋ | 12614/48008 [43:02<1:56:48, 5.05it/s] 26%|██▋ | 12615/48008 [43:03<1:57:29, 5.02it/s] 26%|██▋ | 12616/48008 [43:03<1:55:08, 5.12it/s] 26%|██▋ | 12617/48008 [43:03<1:55:26, 5.11it/s] 26%|██▋ | 12618/48008 [43:03<1:56:44, 5.05it/s] 26%|██▋ | 12619/48008 [43:03<1:55:54, 5.09it/s] 26%|██▋ | 12620/48008 [43:03<1:54:58, 5.13it/s] 26%|██▋ | 12621/48008 [43:04<1:55:08, 5.12it/s] 26%|██▋ | 12622/48008 [43:04<1:54:18, 5.16it/s] 26%|██▋ | 12623/48008 [43:04<1:53:50, 5.18it/s] 26%|██▋ | 12624/48008 [43:04<1:53:29, 5.20it/s] 26%|██▋ | 12625/48008 [43:05<2:13:17, 4.42it/s] 26%|██▋ | 12626/48008 [43:05<2:07:25, 4.63it/s] 26%|██▋ | 12627/48008 [43:05<2:02:40, 4.81it/s] 26%|██▋ | 12628/48008 [43:05<2:07:52, 4.61it/s] 26%|██▋ | 12629/48008 [43:05<2:03:04, 4.79it/s] 26%|██▋ | 12630/48008 [43:06<1:59:38, 4.93it/s] 26%|██▋ | 12631/48008 [43:06<1:56:58, 5.04it/s] 26%|██▋ | 12632/48008 [43:06<1:55:34, 5.10it/s] 26%|██▋ | 12633/48008 [43:06<1:56:40, 5.05it/s] 26%|██▋ | 12634/48008 [43:06<1:55:31, 5.10it/s] 26%|██▋ | 12635/48008 [43:07<1:54:23, 5.15it/s] 26%|██▋ | 12636/48008 [43:07<1:53:04, 5.21it/s] 26%|██▋ | 12637/48008 [43:07<1:53:06, 5.21it/s] 26%|██▋ | 12638/48008 [43:07<1:54:35, 5.14it/s] 26%|██▋ | 12639/48008 [43:07<1:54:30, 5.15it/s] 26%|██▋ | 12640/48008 [43:07<1:53:13, 5.21it/s] 26%|██▋ | 12641/48008 [43:08<1:52:23, 5.24it/s] 26%|██▋ | 12642/48008 [43:08<1:52:54, 5.22it/s] 26%|██▋ | 12643/48008 [43:08<2:01:17, 4.86it/s] 26%|██▋ | 12644/48008 [43:08<1:58:11, 4.99it/s] 26%|██▋ | 12645/48008 [43:08<1:56:39, 5.05it/s] 26%|██▋ | 12646/48008 [43:09<1:55:02, 5.12it/s] 26%|██▋ | 12647/48008 [43:09<1:56:12, 5.07it/s] 26%|██▋ | 12648/48008 [43:09<2:03:20, 4.78it/s] 26%|██▋ | 12649/48008 [43:09<2:00:24, 4.89it/s] 26%|██▋ | 12650/48008 [43:09<1:58:11, 4.99it/s] {'loss': 0.3031, 'grad_norm': 0.40234375, 'learning_rate': 0.00014730461589735044, 'epoch': 0.26} 26%|██▋ | 12650/48008 [43:10<1:58:11, 4.99it/s] 26%|██▋ | 12651/48008 [43:10<1:56:24, 5.06it/s] 26%|██▋ | 12652/48008 [43:10<1:55:25, 5.11it/s] 26%|██▋ | 12653/48008 [43:10<1:55:48, 5.09it/s] 26%|██▋ | 12654/48008 [43:10<1:55:11, 5.11it/s] 26%|██▋ | 12655/48008 [43:10<1:54:33, 5.14it/s] 26%|██▋ | 12656/48008 [43:11<1:55:59, 5.08it/s] 26%|██▋ | 12657/48008 [43:11<1:55:13, 5.11it/s] 26%|██▋ | 12658/48008 [43:11<2:02:45, 4.80it/s] 26%|██▋ | 12659/48008 [43:11<1:59:46, 4.92it/s] 26%|██▋ | 12660/48008 [43:11<1:57:08, 5.03it/s] 26%|██▋ | 12661/48008 [43:12<1:55:41, 5.09it/s] 26%|██▋ | 12662/48008 [43:12<1:58:24, 4.98it/s] 26%|██▋ | 12663/48008 [43:12<1:57:47, 5.00it/s] 26%|██▋ | 12664/48008 [43:12<1:56:59, 5.03it/s] 26%|██▋ | 12665/48008 [43:12<1:56:57, 5.04it/s] 26%|██▋ | 12666/48008 [43:13<1:56:39, 5.05it/s] 26%|██▋ | 12667/48008 [43:13<1:55:21, 5.11it/s] 26%|██▋ | 12668/48008 [43:13<1:53:55, 5.17it/s] 26%|██▋ | 12669/48008 [43:13<1:56:45, 5.04it/s] 26%|██▋ | 12670/48008 [43:13<1:57:04, 5.03it/s] 26%|██▋ | 12671/48008 [43:14<2:03:47, 4.76it/s] 26%|██▋ | 12672/48008 [43:14<2:00:40, 4.88it/s] 26%|██▋ | 12673/48008 [43:14<1:57:45, 5.00it/s] 26%|██▋ | 12674/48008 [43:14<1:56:38, 5.05it/s] 26%|██▋ | 12675/48008 [43:14<1:57:14, 5.02it/s] 26%|██▋ | 12676/48008 [43:15<1:56:04, 5.07it/s] 26%|██▋ | 12677/48008 [43:15<2:03:14, 4.78it/s] 26%|██▋ | 12678/48008 [43:15<2:03:29, 4.77it/s] 26%|██▋ | 12679/48008 [43:15<1:59:53, 4.91it/s] 26%|██▋ | 12680/48008 [43:15<1:57:57, 4.99it/s] 26%|██▋ | 12681/48008 [43:16<1:56:15, 5.06it/s] 26%|██▋ | 12682/48008 [43:16<2:03:34, 4.76it/s] 26%|██▋ | 12683/48008 [43:16<2:00:02, 4.90it/s] 26%|██▋ | 12684/48008 [43:16<1:59:46, 4.92it/s] 26%|██▋ | 12685/48008 [43:16<1:57:25, 5.01it/s] 26%|██▋ | 12686/48008 [43:17<2:04:07, 4.74it/s] 26%|██▋ | 12687/48008 [43:17<2:00:54, 4.87it/s] 26%|██▋ | 12688/48008 [43:17<2:00:34, 4.88it/s] 26%|██▋ | 12689/48008 [43:17<2:00:26, 4.89it/s] 26%|██▋ | 12690/48008 [43:18<1:57:45, 5.00it/s] 26%|██▋ | 12691/48008 [43:18<1:56:34, 5.05it/s] 26%|██▋ | 12692/48008 [43:18<1:54:53, 5.12it/s] 26%|██▋ | 12693/48008 [43:18<1:54:29, 5.14it/s] 26%|██▋ | 12694/48008 [43:18<1:54:08, 5.16it/s] 26%|██▋ | 12695/48008 [43:18<1:54:43, 5.13it/s] 26%|██▋ | 12696/48008 [43:19<1:53:53, 5.17it/s] 26%|██▋ | 12697/48008 [43:19<1:53:37, 5.18it/s] 26%|██▋ | 12698/48008 [43:19<1:54:09, 5.16it/s] 26%|██▋ | 12699/48008 [43:19<1:57:04, 5.03it/s] 26%|██▋ | 12700/48008 [43:19<1:56:31, 5.05it/s] {'loss': 0.301, 'grad_norm': 0.302734375, 'learning_rate': 0.00014709631728045326, 'epoch': 0.26} 26%|██▋ | 12700/48008 [43:19<1:56:31, 5.05it/s] 26%|██▋ | 12701/48008 [43:20<1:55:32, 5.09it/s] 26%|██▋ | 12702/48008 [43:20<1:55:09, 5.11it/s] 26%|██▋ | 12703/48008 [43:20<1:53:28, 5.19it/s] 26%|██▋ | 12704/48008 [43:20<1:53:28, 5.19it/s] 26%|██▋ | 12705/48008 [43:20<1:53:24, 5.19it/s] 26%|██▋ | 12706/48008 [43:21<1:51:32, 5.27it/s] 26%|██▋ | 12707/48008 [43:21<1:50:49, 5.31it/s] 26%|██▋ | 12708/48008 [43:21<1:49:14, 5.39it/s] 26%|██▋ | 12709/48008 [43:21<1:50:19, 5.33it/s] 26%|██▋ | 12710/48008 [43:21<2:11:32, 4.47it/s] 26%|██▋ | 12711/48008 [43:22<2:06:02, 4.67it/s] 26%|██▋ | 12712/48008 [43:22<2:01:49, 4.83it/s] 26%|██▋ | 12713/48008 [43:22<2:33:28, 3.83it/s] 26%|██▋ | 12714/48008 [43:22<2:22:34, 4.13it/s] 26%|██▋ | 12715/48008 [43:23<2:12:51, 4.43it/s] 26%|██▋ | 12716/48008 [43:23<2:06:38, 4.64it/s] 26%|██▋ | 12717/48008 [43:23<2:01:38, 4.84it/s] 26%|██▋ | 12718/48008 [43:23<2:06:51, 4.64it/s] 26%|██▋ | 12719/48008 [43:23<2:02:39, 4.79it/s] 26%|██▋ | 12720/48008 [43:24<1:59:48, 4.91it/s] 26%|██▋ | 12721/48008 [43:24<1:58:16, 4.97it/s] 26%|██▋ | 12722/48008 [43:24<1:58:39, 4.96it/s] 27%|██▋ | 12723/48008 [43:24<1:55:59, 5.07it/s] 27%|██▋ | 12724/48008 [43:24<1:53:47, 5.17it/s] 27%|██▋ | 12725/48008 [43:25<1:53:26, 5.18it/s] 27%|██▋ | 12726/48008 [43:25<1:53:09, 5.20it/s] 27%|██▋ | 12727/48008 [43:25<1:54:14, 5.15it/s] 27%|██▋ | 12728/48008 [43:25<2:02:10, 4.81it/s] 27%|██▋ | 12729/48008 [43:25<1:59:23, 4.92it/s] 27%|██▋ | 12730/48008 [43:26<1:58:12, 4.97it/s] 27%|██▋ | 12731/48008 [43:26<1:57:46, 4.99it/s] 27%|██▋ | 12732/48008 [43:26<1:56:33, 5.04it/s] 27%|██▋ | 12733/48008 [43:26<1:57:20, 5.01it/s] 27%|██▋ | 12734/48008 [43:26<1:55:50, 5.07it/s] 27%|██▋ | 12735/48008 [43:27<1:54:46, 5.12it/s] 27%|██▋ | 12736/48008 [43:27<1:56:59, 5.03it/s] 27%|██▋ | 12737/48008 [43:27<1:56:53, 5.03it/s] 27%|██▋ | 12738/48008 [43:27<1:58:36, 4.96it/s] 27%|██▋ | 12739/48008 [43:27<1:57:48, 4.99it/s] 27%|██▋ | 12740/48008 [43:28<1:56:05, 5.06it/s] 27%|██▋ | 12741/48008 [43:28<1:54:20, 5.14it/s] 27%|██▋ | 12742/48008 [43:28<1:53:35, 5.17it/s] 27%|██▋ | 12743/48008 [43:28<1:52:47, 5.21it/s] 27%|██▋ | 12744/48008 [43:28<1:55:41, 5.08it/s] 27%|██▋ | 12745/48008 [43:29<1:54:23, 5.14it/s] 27%|██▋ | 12746/48008 [43:29<1:55:52, 5.07it/s] 27%|██▋ | 12747/48008 [43:29<1:54:39, 5.13it/s] 27%|██▋ | 12748/48008 [43:29<1:53:21, 5.18it/s] 27%|██▋ | 12749/48008 [43:29<1:54:19, 5.14it/s] 27%|██▋ | 12750/48008 [43:30<1:52:49, 5.21it/s] {'loss': 0.3006, 'grad_norm': 0.451171875, 'learning_rate': 0.00014688801866355607, 'epoch': 0.27} 27%|██▋ | 12750/48008 [43:30<1:52:49, 5.21it/s] 27%|██▋ | 12751/48008 [43:30<1:52:45, 5.21it/s] 27%|██▋ | 12752/48008 [43:30<2:01:03, 4.85it/s] 27%|██▋ | 12753/48008 [43:30<2:00:39, 4.87it/s] 27%|██▋ | 12754/48008 [43:30<2:00:03, 4.89it/s] 27%|██▋ | 12755/48008 [43:31<1:59:46, 4.91it/s] 27%|██▋ | 12756/48008 [43:31<1:57:04, 5.02it/s] 27%|██▋ | 12757/48008 [43:31<1:57:56, 4.98it/s] 27%|██▋ | 12758/48008 [43:31<1:56:21, 5.05it/s] 27%|██▋ | 12759/48008 [43:31<1:53:50, 5.16it/s] 27%|██▋ | 12760/48008 [43:32<1:55:41, 5.08it/s] 27%|██▋ | 12761/48008 [43:32<1:54:11, 5.14it/s] 27%|██▋ | 12762/48008 [43:32<1:53:16, 5.19it/s] 27%|██▋ | 12763/48008 [43:32<1:53:46, 5.16it/s] 27%|██▋ | 12764/48008 [43:32<1:53:07, 5.19it/s] 27%|██▋ | 12765/48008 [43:33<1:56:23, 5.05it/s] 27%|██▋ | 12766/48008 [43:33<1:55:15, 5.10it/s] 27%|██▋ | 12767/48008 [43:33<1:57:38, 4.99it/s] 27%|██▋ | 12768/48008 [43:33<1:57:45, 4.99it/s] 27%|██▋ | 12769/48008 [43:33<1:56:16, 5.05it/s] 27%|██▋ | 12770/48008 [43:33<1:54:27, 5.13it/s] 27%|██▋ | 12771/48008 [43:34<1:53:00, 5.20it/s] 27%|██▋ | 12772/48008 [43:34<1:53:38, 5.17it/s] 27%|██▋ | 12773/48008 [43:34<1:55:11, 5.10it/s] 27%|██▋ | 12774/48008 [43:34<2:28:38, 3.95it/s] 27%|██▋ | 12775/48008 [43:35<2:17:58, 4.26it/s] 27%|██▋ | 12776/48008 [43:35<2:18:28, 4.24it/s] 27%|██▋ | 12777/48008 [43:35<2:10:24, 4.50it/s] 27%|██▋ | 12778/48008 [43:35<2:04:55, 4.70it/s] 27%|██▋ | 12779/48008 [43:35<2:01:19, 4.84it/s] 27%|██▋ | 12780/48008 [43:36<1:59:34, 4.91it/s] 27%|██▋ | 12781/48008 [43:36<1:58:26, 4.96it/s] 27%|██▋ | 12782/48008 [43:36<1:56:04, 5.06it/s] 27%|██▋ | 12783/48008 [43:36<1:55:10, 5.10it/s] 27%|██▋ | 12784/48008 [43:36<1:54:01, 5.15it/s] 27%|██▋ | 12785/48008 [43:37<1:53:39, 5.17it/s] 27%|██▋ | 12786/48008 [43:37<1:53:23, 5.18it/s] 27%|██▋ | 12787/48008 [43:37<1:52:30, 5.22it/s] 27%|██▋ | 12788/48008 [43:37<1:54:37, 5.12it/s] 27%|██▋ | 12789/48008 [43:37<1:53:11, 5.19it/s] 27%|██▋ | 12790/48008 [43:38<2:13:01, 4.41it/s] 27%|██▋ | 12791/48008 [43:38<2:09:11, 4.54it/s] 27%|██▋ | 12792/48008 [43:38<2:03:40, 4.75it/s] 27%|██▋ | 12793/48008 [43:38<1:59:49, 4.90it/s] 27%|██▋ | 12794/48008 [43:38<1:58:48, 4.94it/s] 27%|██▋ | 12795/48008 [43:39<1:58:39, 4.95it/s] 27%|██▋ | 12796/48008 [43:39<1:58:22, 4.96it/s] 27%|██▋ | 12797/48008 [43:39<1:56:14, 5.05it/s] 27%|██▋ | 12798/48008 [43:39<1:57:18, 5.00it/s] 27%|██▋ | 12799/48008 [43:39<1:55:21, 5.09it/s] 27%|██▋ | 12800/48008 [43:40<1:54:26, 5.13it/s] {'loss': 0.3005, 'grad_norm': 0.18359375, 'learning_rate': 0.0001466797200466589, 'epoch': 0.27} 27%|██▋ | 12800/48008 [43:40<1:54:26, 5.13it/s] 27%|██▋ | 12801/48008 [43:40<1:54:25, 5.13it/s] 27%|██▋ | 12802/48008 [43:40<1:54:22, 5.13it/s] 27%|██▋ | 12803/48008 [43:40<1:53:19, 5.18it/s] 27%|██▋ | 12804/48008 [43:40<1:52:48, 5.20it/s] 27%|██▋ | 12805/48008 [43:41<1:52:13, 5.23it/s] 27%|██▋ | 12806/48008 [43:41<1:52:10, 5.23it/s] 27%|██▋ | 12807/48008 [43:41<1:52:13, 5.23it/s] 27%|██▋ | 12808/48008 [43:41<1:53:22, 5.17it/s] 27%|██▋ | 12809/48008 [43:41<1:52:53, 5.20it/s] 27%|██▋ | 12810/48008 [43:42<1:52:25, 5.22it/s] 27%|██▋ | 12811/48008 [43:42<1:53:26, 5.17it/s] 27%|██▋ | 12812/48008 [43:42<1:52:26, 5.22it/s] 27%|██▋ | 12813/48008 [43:42<1:51:43, 5.25it/s] 27%|██▋ | 12814/48008 [43:42<1:51:54, 5.24it/s] 27%|██▋ | 12815/48008 [43:43<1:55:11, 5.09it/s] 27%|██▋ | 12816/48008 [43:43<1:56:05, 5.05it/s] 27%|██▋ | 12817/48008 [43:43<1:54:28, 5.12it/s] 27%|██▋ | 12818/48008 [43:43<1:52:08, 5.23it/s] 27%|██▋ | 12819/48008 [43:43<1:51:43, 5.25it/s] 27%|██▋ | 12820/48008 [43:44<1:51:35, 5.26it/s] 27%|██▋ | 12821/48008 [43:44<1:51:50, 5.24it/s] 27%|██▋ | 12822/48008 [43:44<2:26:05, 4.01it/s] 27%|██▋ | 12823/48008 [43:44<2:15:30, 4.33it/s] 27%|██▋ | 12824/48008 [43:44<2:07:47, 4.59it/s] 27%|██▋ | 12825/48008 [43:45<2:06:13, 4.65it/s] 27%|██▋ | 12826/48008 [43:45<2:01:21, 4.83it/s] 27%|██▋ | 12827/48008 [43:45<1:59:55, 4.89it/s] 27%|██▋ | 12828/48008 [43:45<2:01:06, 4.84it/s] 27%|██▋ | 12829/48008 [43:45<1:58:57, 4.93it/s] 27%|██▋ | 12830/48008 [43:46<1:57:00, 5.01it/s] 27%|██▋ | 12831/48008 [43:46<1:55:47, 5.06it/s] 27%|██▋ | 12832/48008 [43:46<1:55:50, 5.06it/s] 27%|██▋ | 12833/48008 [43:46<1:54:28, 5.12it/s] 27%|██▋ | 12834/48008 [43:46<1:53:13, 5.18it/s] 27%|██▋ | 12835/48008 [43:47<2:01:13, 4.84it/s] 27%|██▋ | 12836/48008 [43:47<1:58:33, 4.94it/s] 27%|██▋ | 12837/48008 [43:47<1:58:33, 4.94it/s] 27%|██▋ | 12838/48008 [43:47<1:56:08, 5.05it/s] 27%|██▋ | 12839/48008 [43:47<1:54:30, 5.12it/s] 27%|██▋ | 12840/48008 [43:48<2:01:53, 4.81it/s] 27%|██▋ | 12841/48008 [43:48<1:58:36, 4.94it/s] 27%|██▋ | 12842/48008 [43:48<1:58:49, 4.93it/s] 27%|██▋ | 12843/48008 [43:48<1:56:58, 5.01it/s] 27%|██▋ | 12844/48008 [43:48<1:57:20, 4.99it/s] 27%|██▋ | 12845/48008 [43:49<1:55:05, 5.09it/s] 27%|██▋ | 12846/48008 [43:49<1:53:56, 5.14it/s] 27%|██▋ | 12847/48008 [43:49<1:52:41, 5.20it/s] 27%|██▋ | 12848/48008 [43:49<1:52:43, 5.20it/s] 27%|██▋ | 12849/48008 [43:49<1:54:27, 5.12it/s] 27%|██▋ | 12850/48008 [43:50<1:53:24, 5.17it/s] {'loss': 0.3028, 'grad_norm': 0.1357421875, 'learning_rate': 0.0001464714214297617, 'epoch': 0.27} 27%|██▋ | 12850/48008 [43:50<1:53:24, 5.17it/s] 27%|██▋ | 12851/48008 [43:50<1:56:05, 5.05it/s] 27%|██▋ | 12852/48008 [43:50<1:56:54, 5.01it/s] 27%|██▋ | 12853/48008 [43:50<1:55:16, 5.08it/s] 27%|██▋ | 12854/48008 [43:50<1:54:25, 5.12it/s] 27%|██▋ | 12855/48008 [43:51<1:53:35, 5.16it/s] 27%|██▋ | 12856/48008 [43:51<1:54:54, 5.10it/s] 27%|██▋ | 12857/48008 [43:51<1:53:49, 5.15it/s] 27%|██▋ | 12858/48008 [43:51<1:51:41, 5.25it/s] 27%|██▋ | 12859/48008 [43:51<2:11:56, 4.44it/s] 27%|██▋ | 12860/48008 [43:52<2:05:50, 4.66it/s] 27%|██▋ | 12861/48008 [43:52<2:01:33, 4.82it/s] 27%|██▋ | 12862/48008 [43:52<1:58:31, 4.94it/s] 27%|██▋ | 12863/48008 [43:52<1:56:01, 5.05it/s] 27%|██▋ | 12864/48008 [43:52<1:56:55, 5.01it/s] 27%|██▋ | 12865/48008 [43:53<1:57:21, 4.99it/s] 27%|██▋ | 12866/48008 [43:53<2:03:52, 4.73it/s] 27%|██▋ | 12867/48008 [43:53<2:00:31, 4.86it/s] 27%|██▋ | 12868/48008 [43:53<1:56:52, 5.01it/s] 27%|██▋ | 12869/48008 [43:53<1:55:22, 5.08it/s] 27%|██▋ | 12870/48008 [43:54<1:54:24, 5.12it/s] 27%|██▋ | 12871/48008 [43:54<1:53:15, 5.17it/s] 27%|██▋ | 12872/48008 [43:54<1:53:53, 5.14it/s] 27%|██▋ | 12873/48008 [43:54<1:52:37, 5.20it/s] 27%|██▋ | 12874/48008 [43:54<1:52:19, 5.21it/s] 27%|██▋ | 12875/48008 [43:55<1:51:07, 5.27it/s] 27%|██▋ | 12876/48008 [43:55<1:53:27, 5.16it/s] 27%|██▋ | 12877/48008 [43:55<1:55:21, 5.08it/s] 27%|██▋ | 12878/48008 [43:55<1:56:40, 5.02it/s] 27%|██▋ | 12879/48008 [43:55<1:56:53, 5.01it/s] 27%|██▋ | 12880/48008 [43:56<1:55:13, 5.08it/s] 27%|██▋ | 12881/48008 [43:56<1:56:08, 5.04it/s] 27%|██▋ | 12882/48008 [43:56<1:56:36, 5.02it/s] 27%|██▋ | 12883/48008 [43:56<1:54:18, 5.12it/s] 27%|██▋ | 12884/48008 [43:56<1:53:17, 5.17it/s] 27%|██▋ | 12885/48008 [43:57<2:12:53, 4.41it/s] 27%|██▋ | 12886/48008 [43:57<2:07:44, 4.58it/s] 27%|██▋ | 12887/48008 [43:57<2:02:06, 4.79it/s] 27%|██▋ | 12888/48008 [43:57<1:59:33, 4.90it/s] 27%|██▋ | 12889/48008 [43:57<1:57:01, 5.00it/s] 27%|██▋ | 12890/48008 [43:58<1:55:48, 5.05it/s] 27%|██▋ | 12891/48008 [43:58<1:55:14, 5.08it/s] 27%|██▋ | 12892/48008 [43:58<1:53:57, 5.14it/s] 27%|██▋ | 12893/48008 [43:58<1:53:18, 5.17it/s] 27%|██▋ | 12894/48008 [43:58<1:52:07, 5.22it/s] 27%|██▋ | 12895/48008 [43:59<2:12:51, 4.40it/s] 27%|██▋ | 12896/48008 [43:59<2:07:04, 4.61it/s] 27%|██▋ | 12897/48008 [43:59<2:04:54, 4.68it/s] 27%|██▋ | 12898/48008 [43:59<2:00:54, 4.84it/s] 27%|██▋ | 12899/48008 [43:59<1:57:12, 4.99it/s] 27%|██▋ | 12900/48008 [44:00<1:56:39, 5.02it/s] {'loss': 0.3045, 'grad_norm': 0.3671875, 'learning_rate': 0.00014626312281286455, 'epoch': 0.27} 27%|██▋ | 12900/48008 [44:00<1:56:39, 5.02it/s] 27%|██▋ | 12901/48008 [44:00<1:55:08, 5.08it/s] 27%|██▋ | 12902/48008 [44:00<1:54:39, 5.10it/s] 27%|██▋ | 12903/48008 [44:00<2:02:18, 4.78it/s] 27%|██▋ | 12904/48008 [44:00<1:58:21, 4.94it/s] 27%|██▋ | 12905/48008 [44:01<1:56:00, 5.04it/s] 27%|██▋ | 12906/48008 [44:01<1:53:33, 5.15it/s] 27%|██▋ | 12907/48008 [44:01<1:55:02, 5.09it/s] 27%|██▋ | 12908/48008 [44:01<1:53:51, 5.14it/s] 27%|██▋ | 12909/48008 [44:01<1:52:19, 5.21it/s] 27%|██▋ | 12910/48008 [44:02<1:51:36, 5.24it/s] 27%|██▋ | 12911/48008 [44:02<1:53:45, 5.14it/s] 27%|██▋ | 12912/48008 [44:02<2:01:29, 4.81it/s] 27%|██▋ | 12913/48008 [44:02<1:58:31, 4.93it/s] 27%|██▋ | 12914/48008 [44:02<1:57:00, 5.00it/s] 27%|██▋ | 12915/48008 [44:03<1:54:51, 5.09it/s] 27%|██▋ | 12916/48008 [44:03<1:54:43, 5.10it/s] 27%|██▋ | 12917/48008 [44:03<1:54:01, 5.13it/s] 27%|██▋ | 12918/48008 [44:03<1:53:27, 5.15it/s] 27%|██▋ | 12919/48008 [44:03<1:54:18, 5.12it/s] 27%|██▋ | 12920/48008 [44:04<1:53:26, 5.16it/s] 27%|██▋ | 12921/48008 [44:04<1:53:56, 5.13it/s] 27%|██▋ | 12922/48008 [44:04<1:52:43, 5.19it/s] 27%|██▋ | 12923/48008 [44:04<1:51:53, 5.23it/s] 27%|██▋ | 12924/48008 [44:04<1:51:58, 5.22it/s] 27%|██▋ | 12925/48008 [44:05<1:51:59, 5.22it/s] 27%|██▋ | 12926/48008 [44:05<1:52:15, 5.21it/s] 27%|██▋ | 12927/48008 [44:05<1:51:25, 5.25it/s] 27%|██▋ | 12928/48008 [44:05<1:52:28, 5.20it/s] 27%|██▋ | 12929/48008 [44:05<1:52:05, 5.22it/s] 27%|██▋ | 12930/48008 [44:06<1:52:03, 5.22it/s] 27%|██▋ | 12931/48008 [44:06<1:54:06, 5.12it/s] 27%|██▋ | 12932/48008 [44:06<2:13:23, 4.38it/s] 27%|██▋ | 12933/48008 [44:06<2:41:09, 3.63it/s] 27%|██▋ | 12934/48008 [44:07<2:26:38, 3.99it/s] 27%|██▋ | 12935/48008 [44:07<2:17:36, 4.25it/s] 27%|██▋ | 12936/48008 [44:07<2:17:59, 4.24it/s] 27%|██▋ | 12937/48008 [44:07<2:10:25, 4.48it/s] 27%|██▋ | 12938/48008 [44:07<2:06:02, 4.64it/s] 27%|██▋ | 12939/48008 [44:08<2:01:23, 4.81it/s] 27%|██▋ | 12940/48008 [44:08<1:58:40, 4.93it/s] 27%|██▋ | 12941/48008 [44:08<2:16:51, 4.27it/s] 27%|██▋ | 12942/48008 [44:08<2:09:03, 4.53it/s] 27%|██▋ | 12943/48008 [44:08<2:03:59, 4.71it/s] 27%|██▋ | 12944/48008 [44:09<2:02:16, 4.78it/s] 27%|██▋ | 12945/48008 [44:09<1:59:55, 4.87it/s] 27%|██▋ | 12946/48008 [44:09<1:57:21, 4.98it/s] 27%|██▋ | 12947/48008 [44:09<1:56:57, 5.00it/s] 27%|██▋ | 12948/48008 [44:09<1:56:13, 5.03it/s] 27%|██▋ | 12949/48008 [44:10<1:56:51, 5.00it/s] 27%|██▋ | 12950/48008 [44:10<1:55:41, 5.05it/s] {'loss': 0.3002, 'grad_norm': 0.5078125, 'learning_rate': 0.00014605482419596734, 'epoch': 0.27} 27%|██▋ | 12950/48008 [44:10<1:55:41, 5.05it/s] 27%|██▋ | 12951/48008 [44:10<1:54:16, 5.11it/s] 27%|██▋ | 12952/48008 [44:10<1:54:40, 5.10it/s] 27%|██▋ | 12953/48008 [44:10<1:57:10, 4.99it/s] 27%|██▋ | 12954/48008 [44:11<1:55:29, 5.06it/s] 27%|██▋ | 12955/48008 [44:11<1:56:20, 5.02it/s] 27%|██▋ | 12956/48008 [44:11<1:54:55, 5.08it/s] 27%|██▋ | 12957/48008 [44:11<1:53:21, 5.15it/s] 27%|██▋ | 12958/48008 [44:11<1:55:52, 5.04it/s] 27%|██▋ | 12959/48008 [44:12<1:54:34, 5.10it/s] 27%|██▋ | 12960/48008 [44:12<1:53:43, 5.14it/s] 27%|██▋ | 12961/48008 [44:12<2:01:12, 4.82it/s] 27%|██▋ | 12962/48008 [44:12<2:00:16, 4.86it/s] 27%|██▋ | 12963/48008 [44:12<1:58:49, 4.92it/s] 27%|██▋ | 12964/48008 [44:13<1:56:13, 5.03it/s] 27%|██▋ | 12965/48008 [44:13<1:55:07, 5.07it/s] 27%|██▋ | 12966/48008 [44:13<1:54:54, 5.08it/s] 27%|██▋ | 12967/48008 [44:13<1:55:11, 5.07it/s] 27%|██▋ | 12968/48008 [44:13<1:54:03, 5.12it/s] 27%|██▋ | 12969/48008 [44:14<1:53:07, 5.16it/s] 27%|██▋ | 12970/48008 [44:14<1:54:28, 5.10it/s] 27%|██▋ | 12971/48008 [44:14<1:52:49, 5.18it/s] 27%|██▋ | 12972/48008 [44:14<1:53:43, 5.13it/s] 27%|██▋ | 12973/48008 [44:14<1:54:21, 5.11it/s] 27%|██▋ | 12974/48008 [44:15<1:53:22, 5.15it/s] 27%|██▋ | 12975/48008 [44:15<2:01:09, 4.82it/s] 27%|██▋ | 12976/48008 [44:15<1:58:07, 4.94it/s] 27%|██▋ | 12977/48008 [44:15<1:58:22, 4.93it/s] 27%|██▋ | 12978/48008 [44:15<1:56:11, 5.03it/s] 27%|██▋ | 12979/48008 [44:16<1:56:08, 5.03it/s] 27%|██▋ | 12980/48008 [44:16<1:53:51, 5.13it/s] 27%|██▋ | 12981/48008 [44:16<1:53:03, 5.16it/s] 27%|██▋ | 12982/48008 [44:16<1:55:55, 5.04it/s] 27%|██▋ | 12983/48008 [44:16<1:54:42, 5.09it/s] 27%|██▋ | 12984/48008 [44:17<2:27:55, 3.95it/s] 27%|██▋ | 12985/48008 [44:17<2:18:42, 4.21it/s] 27%|██▋ | 12986/48008 [44:17<2:10:39, 4.47it/s] 27%|██▋ | 12987/48008 [44:18<2:39:14, 3.67it/s] 27%|██▋ | 12988/48008 [44:18<2:24:52, 4.03it/s] 27%|██▋ | 12989/48008 [44:18<2:14:49, 4.33it/s] 27%|██▋ | 12990/48008 [44:18<2:41:46, 3.61it/s] 27%|██▋ | 12991/48008 [44:19<2:29:03, 3.92it/s] 27%|██▋ | 12992/48008 [44:19<2:38:00, 3.69it/s] 27%|██▋ | 12993/48008 [44:19<2:23:17, 4.07it/s] 27%|██▋ | 12994/48008 [44:19<2:13:09, 4.38it/s] 27%|██▋ | 12995/48008 [44:19<2:05:30, 4.65it/s] 27%|██▋ | 12996/48008 [44:20<2:01:15, 4.81it/s] 27%|██▋ | 12997/48008 [44:20<1:58:48, 4.91it/s] 27%|██▋ | 12998/48008 [44:20<1:56:04, 5.03it/s] 27%|██▋ | 12999/48008 [44:20<2:15:05, 4.32it/s] 27%|██▋ | 13000/48008 [44:20<2:07:31, 4.58it/s] {'loss': 0.3008, 'grad_norm': 0.388671875, 'learning_rate': 0.00014584652557907016, 'epoch': 0.27} 27%|██▋ | 13000/48008 [44:20<2:07:31, 4.58it/s] 27%|██▋ | 13001/48008 [44:21<2:06:08, 4.63it/s] 27%|██▋ | 13002/48008 [44:21<2:00:52, 4.83it/s] 27%|██▋ | 13003/48008 [44:21<1:58:39, 4.92it/s] 27%|██▋ | 13004/48008 [44:21<1:56:05, 5.03it/s] 27%|██▋ | 13005/48008 [44:21<1:56:44, 5.00it/s] 27%|██▋ | 13006/48008 [44:22<1:54:57, 5.07it/s] 27%|██▋ | 13007/48008 [44:22<1:53:15, 5.15it/s] 27%|██▋ | 13008/48008 [44:22<1:52:49, 5.17it/s] 27%|██▋ | 13009/48008 [44:22<1:52:39, 5.18it/s] 27%|██▋ | 13010/48008 [44:22<1:52:34, 5.18it/s] 27%|██▋ | 13011/48008 [44:23<1:51:34, 5.23it/s] 27%|██▋ | 13012/48008 [44:23<1:51:36, 5.23it/s] 27%|██▋ | 13013/48008 [44:23<1:52:06, 5.20it/s] 27%|██▋ | 13014/48008 [44:23<1:52:15, 5.20it/s] 27%|██▋ | 13015/48008 [44:23<1:52:16, 5.19it/s] 27%|██▋ | 13016/48008 [44:24<1:54:14, 5.10it/s] 27%|██▋ | 13017/48008 [44:24<1:53:38, 5.13it/s] 27%|██▋ | 13018/48008 [44:24<1:52:34, 5.18it/s] 27%|██▋ | 13019/48008 [44:24<1:52:16, 5.19it/s] 27%|██▋ | 13020/48008 [44:24<1:51:46, 5.22it/s] 27%|██▋ | 13021/48008 [44:25<1:51:40, 5.22it/s] 27%|██▋ | 13022/48008 [44:25<1:53:40, 5.13it/s] 27%|██▋ | 13023/48008 [44:25<1:55:03, 5.07it/s] 27%|██▋ | 13024/48008 [44:25<1:54:20, 5.10it/s] 27%|██▋ | 13025/48008 [44:25<1:52:55, 5.16it/s] 27%|██▋ | 13026/48008 [44:26<1:52:19, 5.19it/s] 27%|██▋ | 13027/48008 [44:26<1:51:54, 5.21it/s] 27%|██▋ | 13028/48008 [44:26<1:51:17, 5.24it/s] 27%|██▋ | 13029/48008 [44:26<1:52:06, 5.20it/s] 27%|██▋ | 13030/48008 [44:26<1:52:55, 5.16it/s] 27%|██▋ | 13031/48008 [44:26<1:52:40, 5.17it/s] 27%|██▋ | 13032/48008 [44:27<1:51:25, 5.23it/s] 27%|██▋ | 13033/48008 [44:27<1:53:21, 5.14it/s] 27%|██▋ | 13034/48008 [44:27<1:54:36, 5.09it/s] 27%|██▋ | 13035/48008 [44:27<1:53:58, 5.11it/s] 27%|██▋ | 13036/48008 [44:27<1:52:51, 5.16it/s] 27%|██▋ | 13037/48008 [44:28<1:52:06, 5.20it/s] 27%|██▋ | 13038/48008 [44:28<1:51:58, 5.21it/s] 27%|██▋ | 13039/48008 [44:28<1:52:29, 5.18it/s] 27%|██▋ | 13040/48008 [44:28<1:52:35, 5.18it/s] 27%|██▋ | 13041/48008 [44:28<1:52:33, 5.18it/s] 27%|██▋ | 13042/48008 [44:29<1:54:25, 5.09it/s] 27%|██▋ | 13043/48008 [44:29<2:27:28, 3.95it/s] 27%|██▋ | 13044/48008 [44:29<2:18:33, 4.21it/s] 27%|██▋ | 13045/48008 [44:29<2:11:46, 4.42it/s] 27%|██▋ | 13046/48008 [44:30<2:07:44, 4.56it/s] 27%|██▋ | 13047/48008 [44:30<2:05:47, 4.63it/s] 27%|██▋ | 13048/48008 [44:30<2:04:33, 4.68it/s] 27%|██▋ | 13049/48008 [44:30<1:59:50, 4.86it/s] 27%|██▋ | 13050/48008 [44:30<1:57:12, 4.97it/s] {'loss': 0.3019, 'grad_norm': 0.61328125, 'learning_rate': 0.00014563822696217298, 'epoch': 0.27} 27%|██▋ | 13050/48008 [44:30<1:57:12, 4.97it/s] 27%|██▋ | 13051/48008 [44:31<1:55:12, 5.06it/s] 27%|██▋ | 13052/48008 [44:31<1:53:55, 5.11it/s] 27%|██▋ | 13053/48008 [44:31<1:53:18, 5.14it/s] 27%|██▋ | 13054/48008 [44:31<1:54:30, 5.09it/s] 27%|██▋ | 13055/48008 [44:31<1:54:55, 5.07it/s] 27%|██▋ | 13056/48008 [44:32<1:55:56, 5.02it/s] 27%|██▋ | 13057/48008 [44:32<1:54:54, 5.07it/s] 27%|██▋ | 13058/48008 [44:32<1:54:04, 5.11it/s] 27%|██▋ | 13059/48008 [44:32<1:52:54, 5.16it/s] 27%|██▋ | 13060/48008 [44:32<1:53:50, 5.12it/s] 27%|██▋ | 13061/48008 [44:33<1:53:51, 5.12it/s] 27%|██▋ | 13062/48008 [44:33<1:52:17, 5.19it/s] 27%|██▋ | 13063/48008 [44:33<1:54:00, 5.11it/s] 27%|██▋ | 13064/48008 [44:33<1:53:01, 5.15it/s] 27%|██▋ | 13065/48008 [44:33<1:52:41, 5.17it/s] 27%|██▋ | 13066/48008 [44:34<1:51:44, 5.21it/s] 27%|██▋ | 13067/48008 [44:34<1:51:15, 5.23it/s] 27%|██▋ | 13068/48008 [44:34<1:51:08, 5.24it/s] 27%|██▋ | 13069/48008 [44:34<1:50:56, 5.25it/s] 27%|██▋ | 13070/48008 [44:34<1:53:56, 5.11it/s] 27%|██▋ | 13071/48008 [44:34<1:56:29, 5.00it/s] 27%|██▋ | 13072/48008 [44:35<1:54:22, 5.09it/s] 27%|██▋ | 13073/48008 [44:35<1:53:18, 5.14it/s] 27%|██▋ | 13074/48008 [44:35<1:52:07, 5.19it/s] 27%|██▋ | 13075/48008 [44:35<1:51:49, 5.21it/s] 27%|██▋ | 13076/48008 [44:35<1:51:40, 5.21it/s] 27%|██▋ | 13077/48008 [44:36<1:53:11, 5.14it/s] 27%|██▋ | 13078/48008 [44:36<2:26:57, 3.96it/s] 27%|██▋ | 13079/48008 [44:36<2:14:54, 4.32it/s] 27%|██▋ | 13080/48008 [44:36<2:09:45, 4.49it/s] 27%|██▋ | 13081/48008 [44:37<2:04:10, 4.69it/s] 27%|██▋ | 13082/48008 [44:37<4:00:02, 2.43it/s] 27%|██▋ | 13083/48008 [44:38<3:21:25, 2.89it/s] 27%|██▋ | 13084/48008 [44:38<2:54:59, 3.33it/s] 27%|██▋ | 13085/48008 [44:38<2:35:53, 3.73it/s] 27%|██▋ | 13086/48008 [44:38<2:23:50, 4.05it/s] 27%|██▋ | 13087/48008 [44:38<2:14:06, 4.34it/s] 27%|██▋ | 13088/48008 [44:39<2:07:18, 4.57it/s] 27%|██▋ | 13089/48008 [44:39<2:02:36, 4.75it/s] 27%|██▋ | 13090/48008 [44:39<1:59:13, 4.88it/s] 27%|██▋ | 13091/48008 [44:39<1:56:44, 4.98it/s] 27%|██▋ | 13092/48008 [44:39<1:56:14, 5.01it/s] 27%|██▋ | 13093/48008 [44:40<1:54:45, 5.07it/s] 27%|██▋ | 13094/48008 [44:40<1:52:55, 5.15it/s] 27%|██▋ | 13095/48008 [44:40<1:51:44, 5.21it/s] 27%|██▋ | 13096/48008 [44:40<2:25:37, 4.00it/s] 27%|██▋ | 13097/48008 [44:41<2:14:17, 4.33it/s] 27%|██▋ | 13098/48008 [44:41<2:07:45, 4.55it/s] 27%|██▋ | 13099/48008 [44:41<2:02:45, 4.74it/s] 27%|██▋ | 13100/48008 [44:41<1:59:43, 4.86it/s] {'loss': 0.3036, 'grad_norm': 0.1787109375, 'learning_rate': 0.0001454299283452758, 'epoch': 0.27} 27%|██▋ | 13100/48008 [44:41<1:59:43, 4.86it/s] 27%|██▋ | 13101/48008 [44:41<2:05:23, 4.64it/s] 27%|██▋ | 13102/48008 [44:42<2:01:22, 4.79it/s] 27%|██▋ | 13103/48008 [44:42<1:58:03, 4.93it/s] 27%|██▋ | 13104/48008 [44:42<1:59:13, 4.88it/s] 27%|██▋ | 13105/48008 [44:42<1:57:33, 4.95it/s] 27%|██▋ | 13106/48008 [44:42<1:57:32, 4.95it/s] 27%|██▋ | 13107/48008 [44:43<1:58:42, 4.90it/s] 27%|██▋ | 13108/48008 [44:43<1:56:18, 5.00it/s] 27%|██▋ | 13109/48008 [44:43<1:54:53, 5.06it/s] 27%|██▋ | 13110/48008 [44:43<2:01:59, 4.77it/s] 27%|██▋ | 13111/48008 [44:43<1:59:33, 4.86it/s] 27%|██▋ | 13112/48008 [44:44<1:57:12, 4.96it/s] 27%|██▋ | 13113/48008 [44:44<1:56:30, 4.99it/s] 27%|██▋ | 13114/48008 [44:44<1:54:20, 5.09it/s] 27%|██▋ | 13115/48008 [44:44<1:53:18, 5.13it/s] 27%|██▋ | 13116/48008 [44:44<1:52:18, 5.18it/s] 27%|██▋ | 13117/48008 [44:45<1:52:21, 5.18it/s] 27%|██▋ | 13118/48008 [44:45<1:52:36, 5.16it/s] 27%|██▋ | 13119/48008 [44:45<1:53:43, 5.11it/s] 27%|██▋ | 13120/48008 [44:45<1:52:17, 5.18it/s] 27%|██▋ | 13121/48008 [44:45<2:00:00, 4.84it/s] 27%|██▋ | 13122/48008 [44:46<1:57:18, 4.96it/s] 27%|██▋ | 13123/48008 [44:46<1:56:29, 4.99it/s] 27%|██▋ | 13124/48008 [44:46<1:54:32, 5.08it/s] 27%|██▋ | 13125/48008 [44:46<1:54:49, 5.06it/s] 27%|██▋ | 13126/48008 [44:46<1:55:18, 5.04it/s] 27%|██▋ | 13127/48008 [44:47<1:54:15, 5.09it/s] 27%|██▋ | 13128/48008 [44:47<1:51:56, 5.19it/s] 27%|██▋ | 13129/48008 [44:47<1:53:58, 5.10it/s] 27%|██▋ | 13130/48008 [44:47<1:53:12, 5.13it/s] 27%|██▋ | 13131/48008 [44:47<1:53:25, 5.12it/s] 27%|██▋ | 13132/48008 [44:47<1:53:05, 5.14it/s] 27%|██▋ | 13133/48008 [44:48<1:52:33, 5.16it/s] 27%|██▋ | 13134/48008 [44:48<1:52:07, 5.18it/s] 27%|██▋ | 13135/48008 [44:48<1:51:42, 5.20it/s] 27%|██▋ | 13136/48008 [44:48<1:51:31, 5.21it/s] 27%|██▋ | 13137/48008 [44:48<1:53:18, 5.13it/s] 27%|██▋ | 13138/48008 [44:49<1:52:28, 5.17it/s] 27%|██▋ | 13139/48008 [44:49<1:52:48, 5.15it/s] 27%|██▋ | 13140/48008 [44:49<2:26:24, 3.97it/s] 27%|██▋ | 13141/48008 [44:49<2:16:04, 4.27it/s] 27%|██▋ | 13142/48008 [44:50<2:07:57, 4.54it/s] 27%|██▋ | 13143/48008 [44:50<2:03:42, 4.70it/s] 27%|██▋ | 13144/48008 [44:50<1:59:39, 4.86it/s] 27%|██▋ | 13145/48008 [44:50<1:56:56, 4.97it/s] 27%|██▋ | 13146/48008 [44:50<1:57:27, 4.95it/s] 27%|██▋ | 13147/48008 [44:51<1:55:09, 5.05it/s] 27%|██▋ | 13148/48008 [44:51<1:54:58, 5.05it/s] 27%|██▋ | 13149/48008 [44:51<1:56:14, 5.00it/s] 27%|██▋ | 13150/48008 [44:51<1:55:09, 5.04it/s] {'loss': 0.3, 'grad_norm': 0.1357421875, 'learning_rate': 0.0001452216297283786, 'epoch': 0.27} 27%|██▋ | 13150/48008 [44:51<1:55:09, 5.04it/s] 27%|██▋ | 13151/48008 [44:51<1:53:57, 5.10it/s] 27%|██▋ | 13152/48008 [44:52<1:53:53, 5.10it/s] 27%|██▋ | 13153/48008 [44:52<1:53:12, 5.13it/s] 27%|██▋ | 13154/48008 [44:52<1:53:49, 5.10it/s] 27%|██▋ | 13155/48008 [44:52<1:53:16, 5.13it/s] 27%|██▋ | 13156/48008 [44:52<1:53:16, 5.13it/s] 27%|██▋ | 13157/48008 [44:53<1:52:30, 5.16it/s] 27%|██▋ | 13158/48008 [44:53<1:54:23, 5.08it/s] 27%|██▋ | 13159/48008 [44:53<1:55:11, 5.04it/s] 27%|██▋ | 13160/48008 [44:53<1:54:33, 5.07it/s] 27%|██▋ | 13161/48008 [44:53<1:53:07, 5.13it/s] 27%|██▋ | 13162/48008 [44:54<1:53:43, 5.11it/s] 27%|██▋ | 13163/48008 [44:55<4:14:15, 2.28it/s] 27%|██▋ | 13164/48008 [44:55<3:31:29, 2.75it/s] 27%|██▋ | 13165/48008 [44:55<3:01:37, 3.20it/s] 27%|██▋ | 13166/48008 [44:55<2:39:59, 3.63it/s] 27%|██▋ | 13167/48008 [44:55<2:25:12, 4.00it/s] 27%|██▋ | 13168/48008 [44:55<2:14:45, 4.31it/s] 27%|██▋ | 13169/48008 [44:56<4:12:40, 2.30it/s] 27%|██▋ | 13170/48008 [44:57<3:30:38, 2.76it/s] 27%|██▋ | 13171/48008 [44:57<3:00:31, 3.22it/s] 27%|██▋ | 13172/48008 [44:57<2:40:26, 3.62it/s] 27%|██▋ | 13173/48008 [44:57<2:28:45, 3.90it/s] 27%|██▋ | 13174/48008 [44:57<2:37:21, 3.69it/s] 27%|██▋ | 13175/48008 [44:59<5:15:08, 1.84it/s] 27%|██▋ | 13176/48008 [44:59<4:33:53, 2.12it/s] 27%|██▋ | 13177/48008 [44:59<3:45:38, 2.57it/s] 27%|██▋ | 13178/48008 [44:59<3:10:58, 3.04it/s] 27%|██▋ | 13179/48008 [45:00<2:46:08, 3.49it/s] 27%|██▋ | 13180/48008 [45:00<2:29:48, 3.87it/s] 27%|██▋ | 13181/48008 [45:01<4:06:35, 2.35it/s] 27%|██▋ | 13182/48008 [45:01<3:25:52, 2.82it/s] 27%|██▋ | 13183/48008 [45:01<3:05:23, 3.13it/s] 27%|██▋ | 13184/48008 [45:01<2:44:40, 3.52it/s] 27%|██▋ | 13185/48008 [45:01<2:28:27, 3.91it/s] 27%|██▋ | 13186/48008 [45:02<2:19:07, 4.17it/s] 27%|██▋ | 13187/48008 [45:02<2:10:53, 4.43it/s] 27%|██▋ | 13188/48008 [45:02<2:04:52, 4.65it/s] 27%|██▋ | 13189/48008 [45:02<2:00:42, 4.81it/s] 27%|██▋ | 13190/48008 [45:02<1:58:31, 4.90it/s] 27%|██▋ | 13191/48008 [45:03<2:30:22, 3.86it/s] 27%|██▋ | 13192/48008 [45:03<2:20:18, 4.14it/s] 27%|██▋ | 13193/48008 [45:03<2:10:16, 4.45it/s] 27%|██▋ | 13194/48008 [45:03<2:12:41, 4.37it/s] 27%|██▋ | 13195/48008 [45:04<2:06:00, 4.60it/s] 27%|██▋ | 13196/48008 [45:04<2:00:54, 4.80it/s] 27%|██▋ | 13197/48008 [45:04<1:58:46, 4.88it/s] 27%|██▋ | 13198/48008 [45:04<1:56:36, 4.98it/s] 27%|██▋ | 13199/48008 [45:05<3:42:29, 2.61it/s] 27%|██▋ | 13200/48008 [45:05<3:08:17, 3.08it/s] {'loss': 0.3009, 'grad_norm': 0.2890625, 'learning_rate': 0.00014501333111148143, 'epoch': 0.27} 27%|██▋ | 13200/48008 [45:05<3:08:17, 3.08it/s] 27%|██▋ | 13201/48008 [45:05<2:45:10, 3.51it/s] 27%|██▋ | 13202/48008 [45:05<2:28:06, 3.92it/s] 28%|██▊ | 13203/48008 [45:06<2:19:07, 4.17it/s] 28%|██▊ | 13204/48008 [45:06<2:10:56, 4.43it/s] 28%|██▊ | 13205/48008 [45:06<2:07:36, 4.55it/s] 28%|██▊ | 13206/48008 [45:06<2:04:17, 4.67it/s] 28%|██▊ | 13207/48008 [45:06<2:02:01, 4.75it/s] 28%|██▊ | 13208/48008 [45:07<2:07:18, 4.56it/s] 28%|██▊ | 13209/48008 [45:07<2:02:28, 4.74it/s] 28%|██▊ | 13210/48008 [45:07<1:59:04, 4.87it/s] 28%|██▊ | 13211/48008 [45:07<1:55:57, 5.00it/s] 28%|██▊ | 13212/48008 [45:07<1:54:48, 5.05it/s] 28%|██▊ | 13213/48008 [45:08<2:01:33, 4.77it/s] 28%|██▊ | 13214/48008 [45:08<1:57:41, 4.93it/s] 28%|██▊ | 13215/48008 [45:08<1:55:20, 5.03it/s] 28%|██▊ | 13216/48008 [45:08<1:54:12, 5.08it/s] 28%|██▊ | 13217/48008 [45:08<1:53:10, 5.12it/s] 28%|██▊ | 13218/48008 [45:09<1:52:48, 5.14it/s] 28%|██▊ | 13219/48008 [45:09<1:54:31, 5.06it/s] 28%|██▊ | 13220/48008 [45:09<1:53:09, 5.12it/s] 28%|██▊ | 13221/48008 [45:09<1:52:38, 5.15it/s] 28%|██▊ | 13222/48008 [45:09<1:52:08, 5.17it/s] 28%|██▊ | 13223/48008 [45:10<1:55:00, 5.04it/s] 28%|██▊ | 13224/48008 [45:10<1:55:42, 5.01it/s] 28%|██▊ | 13225/48008 [45:10<1:55:34, 5.02it/s] 28%|██▊ | 13226/48008 [45:10<1:53:48, 5.09it/s] 28%|██▊ | 13227/48008 [45:10<1:52:46, 5.14it/s] 28%|██▊ | 13228/48008 [45:11<1:52:23, 5.16it/s] 28%|██▊ | 13229/48008 [45:11<1:53:37, 5.10it/s] 28%|██▊ | 13230/48008 [45:11<1:55:14, 5.03it/s] 28%|██▊ | 13231/48008 [45:11<1:53:51, 5.09it/s] 28%|██▊ | 13232/48008 [45:11<1:52:48, 5.14it/s] 28%|██▊ | 13233/48008 [45:12<1:51:57, 5.18it/s] 28%|██▊ | 13234/48008 [45:12<1:51:49, 5.18it/s] 28%|██▊ | 13235/48008 [45:12<1:50:31, 5.24it/s] 28%|██▊ | 13236/48008 [45:12<1:50:28, 5.25it/s] 28%|██▊ | 13237/48008 [45:12<1:51:53, 5.18it/s] 28%|██▊ | 13238/48008 [45:13<1:50:36, 5.24it/s] 28%|██▊ | 13239/48008 [45:13<1:51:44, 5.19it/s] 28%|██▊ | 13240/48008 [45:13<1:51:19, 5.21it/s] 28%|██▊ | 13241/48008 [45:13<1:50:05, 5.26it/s] 28%|██▊ | 13242/48008 [45:13<1:49:12, 5.31it/s] 28%|██▊ | 13243/48008 [45:14<1:49:10, 5.31it/s] 28%|██▊ | 13244/48008 [45:14<1:49:33, 5.29it/s] 28%|██▊ | 13245/48008 [45:14<1:49:37, 5.28it/s] 28%|██▊ | 13246/48008 [45:14<1:49:29, 5.29it/s] 28%|██▊ | 13247/48008 [45:14<1:51:38, 5.19it/s] 28%|██▊ | 13248/48008 [45:14<1:52:47, 5.14it/s] 28%|██▊ | 13249/48008 [45:15<1:51:28, 5.20it/s] 28%|██▊ | 13250/48008 [45:15<1:51:18, 5.20it/s] {'loss': 0.3029, 'grad_norm': 0.1689453125, 'learning_rate': 0.00014480503249458425, 'epoch': 0.28} 28%|██▊ | 13250/48008 [45:15<1:51:18, 5.20it/s] 28%|██▊ | 13251/48008 [45:15<1:51:18, 5.20it/s] 28%|██▊ | 13252/48008 [45:15<1:52:48, 5.14it/s] 28%|██▊ | 13253/48008 [45:15<1:52:15, 5.16it/s] 28%|██▊ | 13254/48008 [45:16<1:54:55, 5.04it/s] 28%|██▊ | 13255/48008 [45:16<1:54:15, 5.07it/s] 28%|██▊ | 13256/48008 [45:16<2:13:38, 4.33it/s] 28%|██▊ | 13257/48008 [45:16<2:06:12, 4.59it/s] 28%|██▊ | 13258/48008 [45:17<2:00:34, 4.80it/s] 28%|██▊ | 13259/48008 [45:17<2:17:32, 4.21it/s] 28%|██▊ | 13260/48008 [45:17<2:12:23, 4.37it/s] 28%|██▊ | 13261/48008 [45:17<2:13:40, 4.33it/s] 28%|██▊ | 13262/48008 [45:17<2:05:50, 4.60it/s] 28%|██▊ | 13263/48008 [45:18<2:01:59, 4.75it/s] 28%|██▊ | 13264/48008 [45:18<1:58:13, 4.90it/s] 28%|██▊ | 13265/48008 [45:18<1:57:48, 4.92it/s] 28%|██▊ | 13266/48008 [45:18<1:55:07, 5.03it/s] 28%|██▊ | 13267/48008 [45:18<1:53:55, 5.08it/s] 28%|██▊ | 13268/48008 [45:19<1:56:20, 4.98it/s] 28%|██▊ | 13269/48008 [45:19<1:56:08, 4.98it/s] 28%|██▊ | 13270/48008 [45:19<1:54:10, 5.07it/s] 28%|██▊ | 13271/48008 [45:19<1:52:21, 5.15it/s] 28%|██▊ | 13272/48008 [45:19<1:51:38, 5.19it/s] 28%|██▊ | 13273/48008 [45:20<1:51:38, 5.19it/s] 28%|██▊ | 13274/48008 [45:20<1:51:15, 5.20it/s] 28%|██▊ | 13275/48008 [45:20<1:51:20, 5.20it/s] 28%|██▊ | 13276/48008 [45:20<1:51:35, 5.19it/s] 28%|██▊ | 13277/48008 [45:20<1:51:32, 5.19it/s] 28%|██▊ | 13278/48008 [45:21<1:53:35, 5.10it/s] 28%|██▊ | 13279/48008 [45:21<1:52:21, 5.15it/s] 28%|██▊ | 13280/48008 [45:21<1:51:58, 5.17it/s] 28%|██▊ | 13281/48008 [45:21<1:54:47, 5.04it/s] 28%|██▊ | 13282/48008 [45:21<2:01:34, 4.76it/s] 28%|██▊ | 13283/48008 [45:22<2:00:38, 4.80it/s] 28%|██▊ | 13284/48008 [45:22<2:01:10, 4.78it/s] 28%|██▊ | 13285/48008 [45:22<1:56:57, 4.95it/s] 28%|██▊ | 13286/48008 [45:22<1:54:08, 5.07it/s] 28%|██▊ | 13287/48008 [45:22<1:52:52, 5.13it/s] 28%|██▊ | 13288/48008 [45:23<1:53:33, 5.10it/s] 28%|██▊ | 13289/48008 [45:23<1:55:13, 5.02it/s] 28%|██▊ | 13290/48008 [45:23<1:55:56, 4.99it/s] 28%|██▊ | 13291/48008 [45:23<1:54:18, 5.06it/s] 28%|██▊ | 13292/48008 [45:23<1:53:17, 5.11it/s] 28%|██▊ | 13293/48008 [45:24<1:53:44, 5.09it/s] 28%|██▊ | 13294/48008 [45:24<1:52:05, 5.16it/s] 28%|██▊ | 13295/48008 [45:24<1:59:57, 4.82it/s] 28%|██▊ | 13296/48008 [45:24<1:56:12, 4.98it/s] 28%|██▊ | 13297/48008 [45:24<2:03:02, 4.70it/s] 28%|██▊ | 13298/48008 [45:25<2:00:55, 4.78it/s] 28%|██▊ | 13299/48008 [45:25<1:58:07, 4.90it/s] 28%|██▊ | 13300/48008 [45:25<1:59:26, 4.84it/s] {'loss': 0.3025, 'grad_norm': 0.2890625, 'learning_rate': 0.00014459673387768706, 'epoch': 0.28} 28%|██▊ | 13300/48008 [45:25<1:59:26, 4.84it/s] 28%|██▊ | 13301/48008 [45:25<1:57:09, 4.94it/s] 28%|██▊ | 13302/48008 [45:25<1:56:50, 4.95it/s] 28%|██▊ | 13303/48008 [45:26<1:56:18, 4.97it/s] 28%|██▊ | 13304/48008 [45:26<1:56:31, 4.96it/s] 28%|██▊ | 13305/48008 [45:26<1:54:32, 5.05it/s] 28%|██▊ | 13306/48008 [45:26<1:55:25, 5.01it/s] 28%|██▊ | 13307/48008 [45:26<1:53:07, 5.11it/s] 28%|██▊ | 13308/48008 [45:27<1:52:00, 5.16it/s] 28%|██▊ | 13309/48008 [45:27<1:51:38, 5.18it/s] 28%|██▊ | 13310/48008 [45:27<1:51:07, 5.20it/s] 28%|██▊ | 13311/48008 [45:27<2:24:38, 4.00it/s] 28%|██▊ | 13312/48008 [45:28<2:14:19, 4.30it/s] 28%|██▊ | 13313/48008 [45:28<2:05:20, 4.61it/s] 28%|██▊ | 13314/48008 [45:28<2:00:15, 4.81it/s] 28%|██▊ | 13315/48008 [45:28<1:58:42, 4.87it/s] 28%|██▊ | 13316/48008 [45:28<1:56:09, 4.98it/s] 28%|██▊ | 13317/48008 [45:29<1:54:10, 5.06it/s] 28%|██▊ | 13318/48008 [45:29<1:53:09, 5.11it/s] 28%|██▊ | 13319/48008 [45:29<1:53:15, 5.10it/s] 28%|██▊ | 13320/48008 [45:29<1:52:27, 5.14it/s] 28%|██▊ | 13321/48008 [45:29<1:51:36, 5.18it/s] 28%|██▊ | 13322/48008 [45:29<1:51:02, 5.21it/s] 28%|██▊ | 13323/48008 [45:30<1:51:10, 5.20it/s] 28%|██▊ | 13324/48008 [45:30<1:52:56, 5.12it/s] 28%|██▊ | 13325/48008 [45:30<1:51:45, 5.17it/s] 28%|██▊ | 13326/48008 [45:30<1:52:01, 5.16it/s] 28%|██▊ | 13327/48008 [45:30<1:52:08, 5.15it/s] 28%|██▊ | 13328/48008 [45:31<1:51:15, 5.20it/s] 28%|██▊ | 13329/48008 [45:31<1:50:36, 5.23it/s] 28%|██▊ | 13330/48008 [45:31<1:50:48, 5.22it/s] 28%|██▊ | 13331/48008 [45:31<1:49:48, 5.26it/s] 28%|██▊ | 13332/48008 [45:31<1:49:27, 5.28it/s] 28%|██▊ | 13333/48008 [45:32<1:49:43, 5.27it/s] 28%|██▊ | 13334/48008 [45:32<1:50:47, 5.22it/s] 28%|██▊ | 13335/48008 [45:32<1:50:40, 5.22it/s] 28%|██▊ | 13336/48008 [45:32<1:51:30, 5.18it/s] 28%|██▊ | 13337/48008 [45:32<1:50:22, 5.24it/s] 28%|██▊ | 13338/48008 [45:33<2:24:17, 4.00it/s] 28%|██▊ | 13339/48008 [45:33<2:16:11, 4.24it/s] 28%|██▊ | 13340/48008 [45:33<2:08:05, 4.51it/s] 28%|██▊ | 13341/48008 [45:33<2:02:58, 4.70it/s] 28%|██▊ | 13342/48008 [45:34<2:00:11, 4.81it/s] 28%|██▊ | 13343/48008 [45:34<1:59:01, 4.85it/s] 28%|██▊ | 13344/48008 [45:34<1:56:46, 4.95it/s] 28%|██▊ | 13345/48008 [45:34<1:54:44, 5.03it/s] 28%|██▊ | 13346/48008 [45:34<1:53:27, 5.09it/s] 28%|██▊ | 13347/48008 [45:34<1:52:47, 5.12it/s] 28%|██▊ | 13348/48008 [45:35<1:51:26, 5.18it/s] 28%|██▊ | 13349/48008 [45:35<1:50:20, 5.24it/s] 28%|██▊ | 13350/48008 [45:35<1:49:24, 5.28it/s] {'loss': 0.3055, 'grad_norm': 0.1875, 'learning_rate': 0.00014438843526078985, 'epoch': 0.28} 28%|██▊ | 13350/48008 [45:35<1:49:24, 5.28it/s] 28%|██▊ | 13351/48008 [45:35<1:49:58, 5.25it/s] 28%|██▊ | 13352/48008 [45:35<1:50:19, 5.24it/s] 28%|██▊ | 13353/48008 [45:36<1:52:06, 5.15it/s] 28%|██▊ | 13354/48008 [45:36<1:51:15, 5.19it/s] 28%|██▊ | 13355/48008 [45:36<1:51:10, 5.19it/s] 28%|██▊ | 13356/48008 [45:36<1:49:48, 5.26it/s] 28%|██▊ | 13357/48008 [45:36<2:09:58, 4.44it/s] 28%|██▊ | 13358/48008 [45:37<2:03:27, 4.68it/s] 28%|██▊ | 13359/48008 [45:37<2:00:49, 4.78it/s] 28%|██▊ | 13360/48008 [45:37<1:55:59, 4.98it/s] 28%|██▊ | 13361/48008 [45:37<1:53:09, 5.10it/s] 28%|██▊ | 13362/48008 [45:37<1:54:21, 5.05it/s] 28%|██▊ | 13363/48008 [45:38<1:55:21, 5.01it/s] 28%|██▊ | 13364/48008 [45:38<1:53:25, 5.09it/s] 28%|██▊ | 13365/48008 [45:38<2:00:15, 4.80it/s] 28%|██▊ | 13366/48008 [45:38<1:58:30, 4.87it/s] 28%|██▊ | 13367/48008 [45:38<1:57:04, 4.93it/s] 28%|██▊ | 13368/48008 [45:39<1:55:45, 4.99it/s] 28%|██▊ | 13369/48008 [45:39<1:54:58, 5.02it/s] 28%|██▊ | 13370/48008 [45:39<1:53:29, 5.09it/s] 28%|██▊ | 13371/48008 [45:39<1:54:36, 5.04it/s] 28%|██▊ | 13372/48008 [45:39<1:54:57, 5.02it/s] 28%|██▊ | 13373/48008 [45:40<1:53:06, 5.10it/s] 28%|██▊ | 13374/48008 [45:40<1:53:30, 5.09it/s] 28%|██▊ | 13375/48008 [45:40<1:53:45, 5.07it/s] 28%|██▊ | 13376/48008 [45:40<1:52:40, 5.12it/s] 28%|██▊ | 13377/48008 [45:40<1:54:19, 5.05it/s] 28%|██▊ | 13378/48008 [45:41<2:26:48, 3.93it/s] 28%|██▊ | 13379/48008 [45:41<2:23:45, 4.01it/s] 28%|██▊ | 13380/48008 [45:41<2:13:50, 4.31it/s] 28%|██▊ | 13381/48008 [45:41<2:06:50, 4.55it/s] 28%|██▊ | 13382/48008 [45:42<2:01:26, 4.75it/s] 28%|██▊ | 13383/48008 [45:42<1:57:55, 4.89it/s] 28%|██▊ | 13384/48008 [45:42<2:15:25, 4.26it/s] 28%|██▊ | 13385/48008 [45:42<2:10:49, 4.41it/s] 28%|██▊ | 13386/48008 [45:43<2:03:57, 4.66it/s] 28%|██▊ | 13387/48008 [45:43<2:00:52, 4.77it/s] 28%|██▊ | 13388/48008 [45:43<1:57:38, 4.90it/s] 28%|██▊ | 13389/48008 [45:43<1:55:26, 5.00it/s] 28%|██▊ | 13390/48008 [45:43<1:56:10, 4.97it/s] 28%|██▊ | 13391/48008 [45:44<1:55:36, 4.99it/s] 28%|██▊ | 13392/48008 [45:44<1:54:56, 5.02it/s] 28%|██▊ | 13393/48008 [45:44<1:52:54, 5.11it/s] 28%|██▊ | 13394/48008 [45:44<1:51:56, 5.15it/s] 28%|██▊ | 13395/48008 [45:44<1:51:05, 5.19it/s] 28%|██▊ | 13396/48008 [45:45<1:59:18, 4.84it/s] 28%|██▊ | 13397/48008 [45:45<1:56:59, 4.93it/s] 28%|██▊ | 13398/48008 [45:45<1:55:05, 5.01it/s] 28%|██▊ | 13399/48008 [45:45<1:55:43, 4.98it/s] 28%|██▊ | 13400/48008 [45:45<1:53:30, 5.08it/s] {'loss': 0.3015, 'grad_norm': 0.400390625, 'learning_rate': 0.0001441801366438927, 'epoch': 0.28} 28%|██▊ | 13400/48008 [45:45<1:53:30, 5.08it/s] 28%|██▊ | 13401/48008 [45:45<1:53:51, 5.07it/s] 28%|██▊ | 13402/48008 [45:46<1:53:44, 5.07it/s] 28%|██▊ | 13403/48008 [45:46<2:00:27, 4.79it/s] 28%|██▊ | 13404/48008 [45:46<1:59:40, 4.82it/s] 28%|██▊ | 13405/48008 [45:46<1:56:23, 4.95it/s] 28%|██▊ | 13406/48008 [45:47<1:54:56, 5.02it/s] 28%|██▊ | 13407/48008 [45:47<1:53:21, 5.09it/s] 28%|██▊ | 13408/48008 [45:47<1:52:14, 5.14it/s] 28%|██▊ | 13409/48008 [45:47<1:52:33, 5.12it/s] 28%|██▊ | 13410/48008 [45:47<1:51:51, 5.16it/s] 28%|██▊ | 13411/48008 [45:47<1:53:15, 5.09it/s] 28%|██▊ | 13412/48008 [45:48<1:53:38, 5.07it/s] 28%|██▊ | 13413/48008 [45:48<1:53:42, 5.07it/s] 28%|██▊ | 13414/48008 [45:48<1:51:48, 5.16it/s] 28%|██▊ | 13415/48008 [45:48<1:51:07, 5.19it/s] 28%|██▊ | 13416/48008 [45:48<1:52:14, 5.14it/s] 28%|██▊ | 13417/48008 [45:49<1:50:43, 5.21it/s] 28%|██▊ | 13418/48008 [45:49<1:51:28, 5.17it/s] 28%|██▊ | 13419/48008 [45:49<1:51:46, 5.16it/s] 28%|██▊ | 13420/48008 [45:49<1:50:57, 5.20it/s] 28%|██▊ | 13421/48008 [45:49<1:53:04, 5.10it/s] 28%|██▊ | 13422/48008 [45:50<1:51:26, 5.17it/s] 28%|██▊ | 13423/48008 [45:50<1:52:22, 5.13it/s] 28%|██▊ | 13424/48008 [45:50<1:51:29, 5.17it/s] 28%|██▊ | 13425/48008 [45:50<1:51:19, 5.18it/s] 28%|██▊ | 13426/48008 [45:50<1:58:59, 4.84it/s] 28%|██▊ | 13427/48008 [45:51<1:56:23, 4.95it/s] 28%|██▊ | 13428/48008 [45:51<1:56:18, 4.96it/s] 28%|██▊ | 13429/48008 [45:51<1:55:41, 4.98it/s] 28%|██▊ | 13430/48008 [45:51<1:54:09, 5.05it/s] 28%|██▊ | 13431/48008 [45:51<1:52:09, 5.14it/s] 28%|██▊ | 13432/48008 [45:52<1:51:42, 5.16it/s] 28%|██▊ | 13433/48008 [45:52<1:53:25, 5.08it/s] 28%|██▊ | 13434/48008 [45:52<1:52:14, 5.13it/s] 28%|██▊ | 13435/48008 [45:52<2:00:03, 4.80it/s] 28%|██▊ | 13436/48008 [45:52<1:57:16, 4.91it/s] 28%|██▊ | 13437/48008 [45:53<1:56:01, 4.97it/s] 28%|██▊ | 13438/48008 [45:53<1:53:41, 5.07it/s] 28%|██▊ | 13439/48008 [45:53<1:52:44, 5.11it/s] 28%|██▊ | 13440/48008 [45:53<1:52:47, 5.11it/s] 28%|██▊ | 13441/48008 [45:53<1:52:02, 5.14it/s] 28%|██▊ | 13442/48008 [45:54<1:59:42, 4.81it/s] 28%|██▊ | 13443/48008 [45:54<1:57:26, 4.91it/s] 28%|██▊ | 13444/48008 [45:54<1:55:25, 4.99it/s] 28%|██▊ | 13445/48008 [45:54<1:55:43, 4.98it/s] 28%|██▊ | 13446/48008 [45:54<2:03:05, 4.68it/s] 28%|██▊ | 13447/48008 [45:55<1:58:44, 4.85it/s] 28%|██▊ | 13448/48008 [45:55<1:56:25, 4.95it/s] 28%|██▊ | 13449/48008 [45:55<1:54:20, 5.04it/s] 28%|██▊ | 13450/48008 [45:55<1:52:52, 5.10it/s] {'loss': 0.2994, 'grad_norm': 0.5, 'learning_rate': 0.0001439718380269955, 'epoch': 0.28} 28%|██▊ | 13450/48008 [45:55<1:52:52, 5.10it/s] 28%|██▊ | 13451/48008 [45:55<1:52:01, 5.14it/s] 28%|██▊ | 13452/48008 [45:56<1:53:42, 5.07it/s] 28%|██▊ | 13453/48008 [45:56<2:26:02, 3.94it/s] 28%|██▊ | 13454/48008 [45:56<2:17:21, 4.19it/s] 28%|██▊ | 13455/48008 [45:56<2:08:54, 4.47it/s] 28%|██▊ | 13456/48008 [45:57<2:02:25, 4.70it/s] 28%|██▊ | 13457/48008 [45:57<1:59:17, 4.83it/s] 28%|██▊ | 13458/48008 [45:57<1:55:33, 4.98it/s] 28%|██▊ | 13459/48008 [45:57<1:54:07, 5.05it/s] 28%|██▊ | 13460/48008 [45:57<1:53:01, 5.09it/s] 28%|██▊ | 13461/48008 [45:58<1:51:37, 5.16it/s] 28%|██▊ | 13462/48008 [45:58<1:51:19, 5.17it/s] 28%|██▊ | 13463/48008 [45:58<1:54:10, 5.04it/s] 28%|██▊ | 13464/48008 [45:58<1:52:58, 5.10it/s] 28%|██▊ | 13465/48008 [45:58<1:53:19, 5.08it/s] 28%|██▊ | 13466/48008 [45:58<1:51:15, 5.17it/s] 28%|██▊ | 13467/48008 [45:59<1:53:19, 5.08it/s] 28%|██▊ | 13468/48008 [45:59<1:52:03, 5.14it/s] 28%|██▊ | 13469/48008 [45:59<1:50:48, 5.20it/s] 28%|██▊ | 13470/48008 [45:59<1:58:22, 4.86it/s] 28%|██▊ | 13471/48008 [46:00<1:56:08, 4.96it/s] 28%|██▊ | 13472/48008 [46:00<1:53:37, 5.07it/s] 28%|██▊ | 13473/48008 [46:00<1:52:28, 5.12it/s] 28%|██▊ | 13474/48008 [46:00<1:51:41, 5.15it/s] 28%|██▊ | 13475/48008 [46:00<1:51:09, 5.18it/s] 28%|██▊ | 13476/48008 [46:00<1:50:07, 5.23it/s] 28%|██▊ | 13477/48008 [46:01<1:50:11, 5.22it/s] 28%|██▊ | 13478/48008 [46:01<1:50:18, 5.22it/s] 28%|██▊ | 13479/48008 [46:01<1:50:04, 5.23it/s] 28%|██▊ | 13480/48008 [46:01<2:10:45, 4.40it/s] 28%|██▊ | 13481/48008 [46:02<2:03:49, 4.65it/s] 28%|██▊ | 13482/48008 [46:02<2:00:02, 4.79it/s] 28%|██▊ | 13483/48008 [46:02<1:56:04, 4.96it/s] 28%|██▊ | 13484/48008 [46:02<1:54:33, 5.02it/s] 28%|██▊ | 13485/48008 [46:02<2:26:54, 3.92it/s] 28%|██▊ | 13486/48008 [46:03<2:15:38, 4.24it/s] 28%|██▊ | 13487/48008 [46:03<2:08:26, 4.48it/s] 28%|██▊ | 13488/48008 [46:03<2:02:56, 4.68it/s] 28%|██▊ | 13489/48008 [46:03<2:01:04, 4.75it/s] 28%|██▊ | 13490/48008 [46:03<2:00:01, 4.79it/s] 28%|██▊ | 13491/48008 [46:04<1:59:44, 4.80it/s] 28%|██▊ | 13492/48008 [46:04<1:56:41, 4.93it/s] 28%|██▊ | 13493/48008 [46:04<2:28:11, 3.88it/s] 28%|██▊ | 13494/48008 [46:04<2:16:12, 4.22it/s] 28%|██▊ | 13495/48008 [46:05<2:07:44, 4.50it/s] 28%|██▊ | 13496/48008 [46:05<2:02:41, 4.69it/s] 28%|██▊ | 13497/48008 [46:05<2:02:17, 4.70it/s] 28%|██▊ | 13498/48008 [46:05<1:58:29, 4.85it/s] 28%|██▊ | 13499/48008 [46:05<1:55:51, 4.96it/s] 28%|██▊ | 13500/48008 [46:06<2:01:49, 4.72it/s] {'loss': 0.301, 'grad_norm': 0.78515625, 'learning_rate': 0.00014376353941009833, 'epoch': 0.28} 28%|██▊ | 13500/48008 [46:06<2:01:49, 4.72it/s] 28%|██▊ | 13501/48008 [46:06<1:58:09, 4.87it/s] 28%|██▊ | 13502/48008 [46:06<2:03:33, 4.65it/s] 28%|██▊ | 13503/48008 [46:06<2:19:11, 4.13it/s] 28%|██▊ | 13504/48008 [46:07<2:30:21, 3.82it/s] 28%|██▊ | 13505/48008 [46:07<2:18:13, 4.16it/s] 28%|██▊ | 13506/48008 [46:07<2:10:23, 4.41it/s] 28%|██▊ | 13507/48008 [46:07<2:05:30, 4.58it/s] 28%|██▊ | 13508/48008 [46:07<2:00:38, 4.77it/s] 28%|██▊ | 13509/48008 [46:08<1:56:22, 4.94it/s] 28%|██▊ | 13510/48008 [46:08<1:54:38, 5.02it/s] 28%|██▊ | 13511/48008 [46:08<1:52:25, 5.11it/s] 28%|██▊ | 13512/48008 [46:08<1:53:40, 5.06it/s] 28%|██▊ | 13513/48008 [46:08<1:52:08, 5.13it/s] 28%|██▊ | 13514/48008 [46:09<1:51:20, 5.16it/s] 28%|██▊ | 13515/48008 [46:09<1:51:02, 5.18it/s] 28%|██▊ | 13516/48008 [46:09<1:52:31, 5.11it/s] 28%|██▊ | 13517/48008 [46:09<1:52:51, 5.09it/s] 28%|██▊ | 13518/48008 [46:09<1:52:10, 5.12it/s] 28%|██▊ | 13519/48008 [46:10<1:51:26, 5.16it/s] 28%|██▊ | 13520/48008 [46:10<1:52:47, 5.10it/s] 28%|██▊ | 13521/48008 [46:10<1:50:12, 5.22it/s] 28%|██▊ | 13522/48008 [46:10<1:49:41, 5.24it/s] 28%|██▊ | 13523/48008 [46:10<1:50:36, 5.20it/s] 28%|██▊ | 13524/48008 [46:11<1:50:25, 5.20it/s] 28%|██▊ | 13525/48008 [46:11<1:52:21, 5.12it/s] 28%|██▊ | 13526/48008 [46:11<1:50:38, 5.19it/s] 28%|██▊ | 13527/48008 [46:11<1:51:58, 5.13it/s] 28%|██▊ | 13528/48008 [46:11<1:59:01, 4.83it/s] 28%|██▊ | 13529/48008 [46:12<1:56:21, 4.94it/s] 28%|██▊ | 13530/48008 [46:12<1:54:28, 5.02it/s] 28%|██▊ | 13531/48008 [46:12<1:53:12, 5.08it/s] 28%|██▊ | 13532/48008 [46:12<1:54:27, 5.02it/s] 28%|██▊ | 13533/48008 [46:12<1:52:51, 5.09it/s] 28%|██▊ | 13534/48008 [46:13<1:51:52, 5.14it/s] 28%|██▊ | 13535/48008 [46:13<1:50:53, 5.18it/s] 28%|██▊ | 13536/48008 [46:13<1:50:22, 5.21it/s] 28%|██▊ | 13537/48008 [46:13<1:49:48, 5.23it/s] 28%|██▊ | 13538/48008 [46:13<1:50:56, 5.18it/s] 28%|██▊ | 13539/48008 [46:13<1:49:59, 5.22it/s] 28%|██▊ | 13540/48008 [46:14<1:51:41, 5.14it/s] 28%|██▊ | 13541/48008 [46:14<1:50:28, 5.20it/s] 28%|██▊ | 13542/48008 [46:14<1:49:28, 5.25it/s] 28%|██▊ | 13543/48008 [46:14<1:47:50, 5.33it/s] 28%|██▊ | 13544/48008 [46:14<1:48:04, 5.31it/s] 28%|██▊ | 13545/48008 [46:15<1:48:27, 5.30it/s] 28%|██▊ | 13546/48008 [46:15<1:48:46, 5.28it/s] 28%|██▊ | 13547/48008 [46:15<1:49:12, 5.26it/s] 28%|██▊ | 13548/48008 [46:15<1:48:24, 5.30it/s] 28%|██▊ | 13549/48008 [46:15<1:49:09, 5.26it/s] 28%|██▊ | 13550/48008 [46:16<1:48:24, 5.30it/s] {'loss': 0.304, 'grad_norm': 0.1337890625, 'learning_rate': 0.00014355524079320112, 'epoch': 0.28} 28%|██▊ | 13550/48008 [46:16<1:48:24, 5.30it/s] 28%|██▊ | 13551/48008 [46:16<1:48:58, 5.27it/s] 28%|██▊ | 13552/48008 [46:16<1:49:15, 5.26it/s] 28%|██▊ | 13553/48008 [46:16<1:48:32, 5.29it/s] 28%|██▊ | 13554/48008 [46:16<1:49:54, 5.23it/s] 28%|██▊ | 13555/48008 [46:17<1:48:54, 5.27it/s] 28%|██▊ | 13556/48008 [46:17<1:48:52, 5.27it/s] 28%|██▊ | 13557/48008 [46:17<1:50:55, 5.18it/s] 28%|██▊ | 13558/48008 [46:17<1:50:58, 5.17it/s] 28%|██▊ | 13559/48008 [46:17<1:50:08, 5.21it/s] 28%|██▊ | 13560/48008 [46:17<1:51:59, 5.13it/s] 28%|██▊ | 13561/48008 [46:18<1:51:17, 5.16it/s] 28%|██▊ | 13562/48008 [46:18<1:50:41, 5.19it/s] 28%|██▊ | 13563/48008 [46:18<1:52:33, 5.10it/s] 28%|██▊ | 13564/48008 [46:18<1:51:26, 5.15it/s] 28%|██▊ | 13565/48008 [46:18<1:50:49, 5.18it/s] 28%|██▊ | 13566/48008 [46:19<1:52:18, 5.11it/s] 28%|██▊ | 13567/48008 [46:19<1:59:21, 4.81it/s] 28%|██▊ | 13568/48008 [46:19<1:57:03, 4.90it/s] 28%|██▊ | 13569/48008 [46:19<1:55:01, 4.99it/s] 28%|██▊ | 13570/48008 [46:20<2:26:48, 3.91it/s] 28%|██▊ | 13571/48008 [46:20<2:15:44, 4.23it/s] 28%|██▊ | 13572/48008 [46:20<2:06:47, 4.53it/s] 28%|██▊ | 13573/48008 [46:20<2:00:43, 4.75it/s] 28%|██▊ | 13574/48008 [46:20<1:57:32, 4.88it/s] 28%|██▊ | 13575/48008 [46:21<1:56:22, 4.93it/s] 28%|██▊ | 13576/48008 [46:21<1:53:33, 5.05it/s] 28%|██▊ | 13577/48008 [46:21<2:12:21, 4.34it/s] 28%|██▊ | 13578/48008 [46:21<2:06:54, 4.52it/s] 28%|██▊ | 13579/48008 [46:21<2:01:44, 4.71it/s] 28%|██▊ | 13580/48008 [46:22<1:58:15, 4.85it/s] 28%|██▊ | 13581/48008 [46:22<1:55:51, 4.95it/s] 28%|██▊ | 13582/48008 [46:22<1:53:01, 5.08it/s] 28%|██▊ | 13583/48008 [46:22<1:51:22, 5.15it/s] 28%|██▊ | 13584/48008 [46:22<1:50:18, 5.20it/s] 28%|██▊ | 13585/48008 [46:23<1:49:51, 5.22it/s] 28%|██▊ | 13586/48008 [46:23<1:49:56, 5.22it/s] 28%|██▊ | 13587/48008 [46:23<1:49:53, 5.22it/s] 28%|██▊ | 13588/48008 [46:23<1:49:23, 5.24it/s] 28%|██▊ | 13589/48008 [46:23<1:49:26, 5.24it/s] 28%|██▊ | 13590/48008 [46:24<1:49:55, 5.22it/s] 28%|██▊ | 13591/48008 [46:24<1:50:06, 5.21it/s] 28%|██▊ | 13592/48008 [46:24<1:50:43, 5.18it/s] 28%|██▊ | 13593/48008 [46:24<1:50:28, 5.19it/s] 28%|██▊ | 13594/48008 [46:24<1:49:30, 5.24it/s] 28%|██▊ | 13595/48008 [46:25<1:49:37, 5.23it/s] 28%|██▊ | 13596/48008 [46:25<1:51:27, 5.15it/s] 28%|██▊ | 13597/48008 [46:25<1:51:21, 5.15it/s] 28%|██▊ | 13598/48008 [46:25<1:50:52, 5.17it/s] 28%|██▊ | 13599/48008 [46:25<1:50:24, 5.19it/s] 28%|██▊ | 13600/48008 [46:26<1:52:04, 5.12it/s] {'loss': 0.3016, 'grad_norm': 0.1513671875, 'learning_rate': 0.00014334694217630397, 'epoch': 0.28} 28%|██▊ | 13600/48008 [46:26<1:52:04, 5.12it/s] 28%|██▊ | 13601/48008 [46:26<1:51:10, 5.16it/s] 28%|██▊ | 13602/48008 [46:26<1:50:50, 5.17it/s] 28%|██▊ | 13603/48008 [46:26<1:50:30, 5.19it/s] 28%|██▊ | 13604/48008 [46:26<1:49:52, 5.22it/s] 28%|██▊ | 13605/48008 [46:26<1:50:43, 5.18it/s] 28%|██▊ | 13606/48008 [46:27<1:49:22, 5.24it/s] 28%|██▊ | 13607/48008 [46:27<1:48:15, 5.30it/s] 28%|██▊ | 13608/48008 [46:27<1:48:57, 5.26it/s] 28%|██▊ | 13609/48008 [46:27<1:49:01, 5.26it/s] 28%|██▊ | 13610/48008 [46:27<1:49:01, 5.26it/s] 28%|██▊ | 13611/48008 [46:28<1:49:19, 5.24it/s] 28%|██▊ | 13612/48008 [46:28<1:49:03, 5.26it/s] 28%|██▊ | 13613/48008 [46:28<1:50:19, 5.20it/s] 28%|██▊ | 13614/48008 [46:28<1:51:45, 5.13it/s] 28%|██▊ | 13615/48008 [46:28<1:53:24, 5.05it/s] 28%|██▊ | 13616/48008 [46:29<1:53:00, 5.07it/s] 28%|██▊ | 13617/48008 [46:29<1:52:18, 5.10it/s] 28%|██▊ | 13618/48008 [46:29<1:52:50, 5.08it/s] 28%|██▊ | 13619/48008 [46:29<1:53:25, 5.05it/s] 28%|██▊ | 13620/48008 [46:29<1:51:38, 5.13it/s] 28%|██▊ | 13621/48008 [46:30<1:50:13, 5.20it/s] 28%|██▊ | 13622/48008 [46:30<1:51:53, 5.12it/s] 28%|██▊ | 13623/48008 [46:30<1:51:18, 5.15it/s] 28%|██▊ | 13624/48008 [46:30<2:10:26, 4.39it/s] 28%|██▊ | 13625/48008 [46:30<2:05:05, 4.58it/s] 28%|██▊ | 13626/48008 [46:31<2:00:35, 4.75it/s] 28%|██▊ | 13627/48008 [46:31<1:56:40, 4.91it/s] 28%|██▊ | 13628/48008 [46:31<1:54:20, 5.01it/s] 28%|██▊ | 13629/48008 [46:31<1:54:35, 5.00it/s] 28%|██▊ | 13630/48008 [46:31<1:54:24, 5.01it/s] 28%|██▊ | 13631/48008 [46:32<1:54:46, 4.99it/s] 28%|██▊ | 13632/48008 [46:32<1:52:15, 5.10it/s] 28%|██▊ | 13633/48008 [46:32<1:50:51, 5.17it/s] 28%|██▊ | 13634/48008 [46:32<1:50:10, 5.20it/s] 28%|██▊ | 13635/48008 [46:32<1:49:59, 5.21it/s] 28%|██▊ | 13636/48008 [46:33<1:49:53, 5.21it/s] 28%|██▊ | 13637/48008 [46:33<2:09:33, 4.42it/s] 28%|██▊ | 13638/48008 [46:33<2:03:34, 4.64it/s] 28%|██▊ | 13639/48008 [46:33<1:59:48, 4.78it/s] 28%|██▊ | 13640/48008 [46:33<1:56:48, 4.90it/s] 28%|██▊ | 13641/48008 [46:34<1:54:45, 4.99it/s] 28%|██▊ | 13642/48008 [46:34<1:53:11, 5.06it/s] 28%|██▊ | 13643/48008 [46:34<1:51:57, 5.12it/s] 28%|██▊ | 13644/48008 [46:34<1:51:04, 5.16it/s] 28%|██▊ | 13645/48008 [46:34<1:58:08, 4.85it/s] 28%|██▊ | 13646/48008 [46:35<1:57:57, 4.86it/s] 28%|██▊ | 13647/48008 [46:35<1:55:37, 4.95it/s] 28%|██▊ | 13648/48008 [46:35<1:53:48, 5.03it/s] 28%|██▊ | 13649/48008 [46:35<1:52:35, 5.09it/s] 28%|██▊ | 13650/48008 [46:35<1:51:04, 5.16it/s] {'loss': 0.3042, 'grad_norm': 0.2236328125, 'learning_rate': 0.00014313864355940679, 'epoch': 0.28} 28%|██▊ | 13650/48008 [46:35<1:51:04, 5.16it/s] 28%|██▊ | 13651/48008 [46:36<1:50:24, 5.19it/s] 28%|██▊ | 13652/48008 [46:36<1:50:07, 5.20it/s] 28%|██▊ | 13653/48008 [46:36<1:49:59, 5.21it/s] 28%|██▊ | 13654/48008 [46:36<1:50:10, 5.20it/s] 28%|██▊ | 13655/48008 [46:36<1:50:05, 5.20it/s] 28%|██▊ | 13656/48008 [46:37<1:49:19, 5.24it/s] 28%|██▊ | 13657/48008 [46:37<1:57:56, 4.85it/s] 28%|██▊ | 13658/48008 [46:37<1:54:26, 5.00it/s] 28%|██▊ | 13659/48008 [46:37<1:53:16, 5.05it/s] 28%|██▊ | 13660/48008 [46:37<1:54:14, 5.01it/s] 28%|██▊ | 13661/48008 [46:38<1:53:59, 5.02it/s] 28%|██▊ | 13662/48008 [46:38<1:54:25, 5.00it/s] 28%|██▊ | 13663/48008 [46:38<2:00:48, 4.74it/s] 28%|██▊ | 13664/48008 [46:38<1:58:48, 4.82it/s] 28%|██▊ | 13665/48008 [46:38<1:56:46, 4.90it/s] 28%|██▊ | 13666/48008 [46:39<1:53:50, 5.03it/s] 28%|██▊ | 13667/48008 [46:39<1:52:31, 5.09it/s] 28%|██▊ | 13668/48008 [46:39<1:53:39, 5.04it/s] 28%|██▊ | 13669/48008 [46:39<1:51:07, 5.15it/s] 28%|██▊ | 13670/48008 [46:39<1:51:12, 5.15it/s] 28%|██▊ | 13671/48008 [46:40<1:58:24, 4.83it/s] 28%|██▊ | 13672/48008 [46:40<1:58:53, 4.81it/s] 28%|██▊ | 13673/48008 [46:40<1:55:22, 4.96it/s] 28%|██▊ | 13674/48008 [46:40<1:52:45, 5.08it/s] 28%|██▊ | 13675/48008 [46:40<1:51:50, 5.12it/s] 28%|██▊ | 13676/48008 [46:41<1:51:03, 5.15it/s] 28%|██▊ | 13677/48008 [46:41<1:50:27, 5.18it/s] 28%|██▊ | 13678/48008 [46:41<1:58:20, 4.83it/s] 28%|██▊ | 13679/48008 [46:41<1:55:08, 4.97it/s] 28%|██▊ | 13680/48008 [46:41<1:54:37, 4.99it/s] 28%|██▊ | 13681/48008 [46:42<1:52:01, 5.11it/s] 28%|██▊ | 13682/48008 [46:42<2:24:58, 3.95it/s] 29%|██▊ | 13683/48008 [46:42<2:13:59, 4.27it/s] 29%|██▊ | 13684/48008 [46:42<2:06:51, 4.51it/s] 29%|██▊ | 13685/48008 [46:43<2:01:27, 4.71it/s] 29%|██▊ | 13686/48008 [46:43<1:57:49, 4.85it/s] 29%|██▊ | 13687/48008 [46:43<1:56:31, 4.91it/s] 29%|██▊ | 13688/48008 [46:43<1:56:23, 4.91it/s] 29%|██▊ | 13689/48008 [46:43<1:54:45, 4.98it/s] 29%|██▊ | 13690/48008 [46:44<1:53:06, 5.06it/s] 29%|██▊ | 13691/48008 [46:44<1:51:43, 5.12it/s] 29%|██▊ | 13692/48008 [46:44<2:10:40, 4.38it/s] 29%|██▊ | 13693/48008 [46:44<2:03:59, 4.61it/s] 29%|██▊ | 13694/48008 [46:44<2:02:55, 4.65it/s] 29%|██▊ | 13695/48008 [46:45<1:59:30, 4.79it/s] 29%|██▊ | 13696/48008 [46:45<2:30:03, 3.81it/s] 29%|██▊ | 13697/48008 [46:45<2:17:47, 4.15it/s] 29%|██▊ | 13698/48008 [46:45<2:08:31, 4.45it/s] 29%|██▊ | 13699/48008 [46:46<2:04:36, 4.59it/s] 29%|██▊ | 13700/48008 [46:46<2:00:19, 4.75it/s] {'loss': 0.3011, 'grad_norm': 0.283203125, 'learning_rate': 0.00014293034494250958, 'epoch': 0.29} 29%|██▊ | 13700/48008 [46:46<2:00:19, 4.75it/s] 29%|██▊ | 13701/48008 [46:46<1:57:22, 4.87it/s] 29%|██▊ | 13702/48008 [46:46<2:28:15, 3.86it/s] 29%|██▊ | 13703/48008 [46:47<2:24:36, 3.95it/s] 29%|██▊ | 13704/48008 [46:47<2:14:31, 4.25it/s] 29%|██▊ | 13705/48008 [46:47<2:06:24, 4.52it/s] 29%|██▊ | 13706/48008 [46:47<2:01:45, 4.70it/s] 29%|██▊ | 13707/48008 [46:47<1:57:13, 4.88it/s] 29%|██▊ | 13708/48008 [46:48<1:54:58, 4.97it/s] 29%|██▊ | 13709/48008 [46:48<1:52:43, 5.07it/s] 29%|██▊ | 13710/48008 [46:48<1:53:47, 5.02it/s] 29%|██▊ | 13711/48008 [46:48<1:52:42, 5.07it/s] 29%|██▊ | 13712/48008 [46:48<1:53:41, 5.03it/s] 29%|██▊ | 13713/48008 [46:49<1:52:11, 5.09it/s] 29%|██▊ | 13714/48008 [46:49<1:50:36, 5.17it/s] 29%|██▊ | 13715/48008 [46:49<1:50:17, 5.18it/s] 29%|██▊ | 13716/48008 [46:49<1:49:56, 5.20it/s] 29%|██▊ | 13717/48008 [46:49<1:49:42, 5.21it/s] 29%|██▊ | 13718/48008 [46:49<1:49:36, 5.21it/s] 29%|██▊ | 13719/48008 [46:50<1:48:53, 5.25it/s] 29%|██▊ | 13720/48008 [46:50<1:52:39, 5.07it/s] 29%|██▊ | 13721/48008 [46:50<1:59:54, 4.77it/s] 29%|██▊ | 13722/48008 [46:50<1:58:04, 4.84it/s] 29%|██▊ | 13723/48008 [46:51<1:57:45, 4.85it/s] 29%|██▊ | 13724/48008 [46:51<1:54:56, 4.97it/s] 29%|██▊ | 13725/48008 [46:51<1:53:03, 5.05it/s] 29%|██▊ | 13726/48008 [46:51<1:52:04, 5.10it/s] 29%|██▊ | 13727/48008 [46:51<1:51:15, 5.14it/s] 29%|██▊ | 13728/48008 [46:52<1:51:59, 5.10it/s] 29%|██▊ | 13729/48008 [46:52<1:58:57, 4.80it/s] 29%|██▊ | 13730/48008 [46:52<1:58:13, 4.83it/s] 29%|██▊ | 13731/48008 [46:52<1:55:18, 4.95it/s] 29%|██▊ | 13732/48008 [46:52<1:55:26, 4.95it/s] 29%|██▊ | 13733/48008 [46:53<1:53:23, 5.04it/s] 29%|██▊ | 13734/48008 [46:53<1:51:39, 5.12it/s] 29%|██▊ | 13735/48008 [46:53<1:51:08, 5.14it/s] 29%|██▊ | 13736/48008 [46:53<1:49:53, 5.20it/s] 29%|██▊ | 13737/48008 [46:53<1:48:06, 5.28it/s] 29%|██▊ | 13738/48008 [46:53<1:48:16, 5.28it/s] 29%|██▊ | 13739/48008 [46:54<1:48:06, 5.28it/s] 29%|██▊ | 13740/48008 [46:54<1:48:18, 5.27it/s] 29%|██▊ | 13741/48008 [46:54<1:48:50, 5.25it/s] 29%|██▊ | 13742/48008 [46:54<2:08:37, 4.44it/s] 29%|██▊ | 13743/48008 [46:55<2:02:05, 4.68it/s] 29%|██▊ | 13744/48008 [46:55<1:57:10, 4.87it/s] 29%|██▊ | 13745/48008 [46:55<1:55:40, 4.94it/s] 29%|██▊ | 13746/48008 [46:55<1:55:38, 4.94it/s] 29%|██▊ | 13747/48008 [46:55<2:02:00, 4.68it/s] 29%|██▊ | 13748/48008 [46:56<1:57:49, 4.85it/s] 29%|██▊ | 13749/48008 [46:56<1:55:01, 4.96it/s] 29%|██▊ | 13750/48008 [46:56<1:52:25, 5.08it/s] {'loss': 0.302, 'grad_norm': 0.498046875, 'learning_rate': 0.00014272204632561242, 'epoch': 0.29} 29%|██▊ | 13750/48008 [46:56<1:52:25, 5.08it/s] 29%|██▊ | 13751/48008 [46:56<1:51:40, 5.11it/s] 29%|██▊ | 13752/48008 [46:56<1:51:50, 5.10it/s] 29%|██▊ | 13753/48008 [46:56<1:51:06, 5.14it/s] 29%|██▊ | 13754/48008 [46:57<1:50:42, 5.16it/s] 29%|██▊ | 13755/48008 [46:57<1:50:15, 5.18it/s] 29%|██▊ | 13756/48008 [46:57<1:50:02, 5.19it/s] 29%|██▊ | 13757/48008 [46:57<1:51:22, 5.13it/s] 29%|██▊ | 13758/48008 [46:58<2:24:30, 3.95it/s] 29%|██▊ | 13759/48008 [46:58<2:13:35, 4.27it/s] 29%|██▊ | 13760/48008 [46:58<2:06:22, 4.52it/s] 29%|██▊ | 13761/48008 [46:58<2:02:38, 4.65it/s] 29%|██▊ | 13762/48008 [46:58<1:57:40, 4.85it/s] 29%|██▊ | 13763/48008 [46:59<1:55:47, 4.93it/s] 29%|██▊ | 13764/48008 [46:59<1:54:23, 4.99it/s] 29%|██▊ | 13765/48008 [46:59<1:54:43, 4.97it/s] 29%|██▊ | 13766/48008 [47:00<3:54:12, 2.44it/s] 29%|██▊ | 13767/48008 [47:00<3:16:49, 2.90it/s] 29%|██▊ | 13768/48008 [47:00<2:50:39, 3.34it/s] 29%|██▊ | 13769/48008 [47:01<2:32:45, 3.74it/s] 29%|██▊ | 13770/48008 [47:01<2:19:21, 4.09it/s] 29%|██▊ | 13771/48008 [47:01<2:10:08, 4.38it/s] 29%|██▊ | 13772/48008 [47:01<2:04:51, 4.57it/s] 29%|██▊ | 13773/48008 [47:01<2:00:28, 4.74it/s] 29%|██▊ | 13774/48008 [47:01<1:58:16, 4.82it/s] 29%|██▊ | 13775/48008 [47:02<1:58:47, 4.80it/s] 29%|██▊ | 13776/48008 [47:03<4:13:04, 2.25it/s] 29%|██▊ | 13777/48008 [47:03<3:30:14, 2.71it/s] 29%|██▊ | 13778/48008 [47:03<3:02:07, 3.13it/s] 29%|██▊ | 13779/48008 [47:03<2:40:08, 3.56it/s] 29%|██▊ | 13780/48008 [47:03<2:24:15, 3.95it/s] 29%|██▊ | 13781/48008 [47:04<2:21:40, 4.03it/s] 29%|██▊ | 13782/48008 [47:05<4:19:15, 2.20it/s] 29%|██▊ | 13783/48008 [47:05<3:37:05, 2.63it/s] 29%|██▊ | 13784/48008 [47:05<3:04:00, 3.10it/s] 29%|██▊ | 13785/48008 [47:05<2:43:26, 3.49it/s] 29%|██▊ | 13786/48008 [47:05<2:28:41, 3.84it/s] 29%|██▊ | 13787/48008 [47:06<2:36:02, 3.65it/s] 29%|██▊ | 13788/48008 [47:07<5:10:27, 1.84it/s] 29%|██▊ | 13789/48008 [47:07<4:10:28, 2.28it/s] 29%|██▊ | 13790/48008 [47:07<3:28:36, 2.73it/s] 29%|██▊ | 13791/48008 [47:08<3:31:59, 2.69it/s] 29%|██▊ | 13792/48008 [47:08<3:01:59, 3.13it/s] 29%|██▊ | 13793/48008 [47:08<2:39:42, 3.57it/s] 29%|██▊ | 13794/48008 [47:09<4:11:24, 2.27it/s] 29%|██▊ | 13795/48008 [47:09<3:28:15, 2.74it/s] 29%|██▊ | 13796/48008 [47:09<2:58:46, 3.19it/s] 29%|██▊ | 13797/48008 [47:09<2:37:34, 3.62it/s] 29%|██▊ | 13798/48008 [47:10<2:22:38, 4.00it/s] 29%|██▊ | 13799/48008 [47:10<2:12:09, 4.31it/s] 29%|██▊ | 13800/48008 [47:10<2:05:02, 4.56it/s] {'loss': 0.3, 'grad_norm': 0.2001953125, 'learning_rate': 0.0001425137477087152, 'epoch': 0.29} 29%|██▊ | 13800/48008 [47:10<2:05:02, 4.56it/s] 29%|██▊ | 13801/48008 [47:10<2:00:45, 4.72it/s] 29%|██▊ | 13802/48008 [47:10<1:56:52, 4.88it/s] 29%|██▉ | 13803/48008 [47:11<1:55:59, 4.91it/s] 29%|██▉ | 13804/48008 [47:11<1:54:55, 4.96it/s] 29%|██▉ | 13805/48008 [47:11<1:51:36, 5.11it/s] 29%|██▉ | 13806/48008 [47:11<2:24:10, 3.95it/s] 29%|██▉ | 13807/48008 [47:12<2:13:55, 4.26it/s] 29%|██▉ | 13808/48008 [47:12<2:06:39, 4.50it/s] 29%|██▉ | 13809/48008 [47:12<2:01:27, 4.69it/s] 29%|██▉ | 13810/48008 [47:12<1:57:29, 4.85it/s] 29%|██▉ | 13811/48008 [47:12<1:53:47, 5.01it/s] 29%|██▉ | 13812/48008 [47:13<3:44:24, 2.54it/s] 29%|██▉ | 13813/48008 [47:13<3:11:21, 2.98it/s] 29%|██▉ | 13814/48008 [47:14<2:45:56, 3.43it/s] 29%|██▉ | 13815/48008 [47:14<2:27:06, 3.87it/s] 29%|██▉ | 13816/48008 [47:14<2:15:50, 4.20it/s] 29%|██▉ | 13817/48008 [47:14<2:07:26, 4.47it/s] 29%|██▉ | 13818/48008 [47:14<2:01:44, 4.68it/s] 29%|██▉ | 13819/48008 [47:14<1:57:20, 4.86it/s] 29%|██▉ | 13820/48008 [47:15<1:54:25, 4.98it/s] 29%|██▉ | 13821/48008 [47:15<1:51:41, 5.10it/s] 29%|██▉ | 13822/48008 [47:15<1:50:22, 5.16it/s] 29%|██▉ | 13823/48008 [47:15<1:49:48, 5.19it/s] 29%|██▉ | 13824/48008 [47:15<1:49:20, 5.21it/s] 29%|██▉ | 13825/48008 [47:16<1:49:07, 5.22it/s] 29%|██▉ | 13826/48008 [47:16<1:49:58, 5.18it/s] 29%|██▉ | 13827/48008 [47:16<1:48:39, 5.24it/s] 29%|██▉ | 13828/48008 [47:16<1:50:18, 5.16it/s] 29%|██▉ | 13829/48008 [47:16<1:50:18, 5.16it/s] 29%|██▉ | 13830/48008 [47:17<1:51:13, 5.12it/s] 29%|██▉ | 13831/48008 [47:17<1:50:17, 5.16it/s] 29%|██▉ | 13832/48008 [47:17<1:50:07, 5.17it/s] 29%|██▉ | 13833/48008 [47:17<2:09:23, 4.40it/s] 29%|██▉ | 13834/48008 [47:17<2:03:42, 4.60it/s] 29%|██▉ | 13835/48008 [47:18<2:02:37, 4.64it/s] 29%|██▉ | 13836/48008 [47:18<2:31:45, 3.75it/s] 29%|██▉ | 13837/48008 [47:18<2:18:22, 4.12it/s] 29%|██▉ | 13838/48008 [47:18<2:09:43, 4.39it/s] 29%|██▉ | 13839/48008 [47:19<2:03:26, 4.61it/s] 29%|██▉ | 13840/48008 [47:19<2:02:27, 4.65it/s] 29%|██▉ | 13841/48008 [47:19<2:01:40, 4.68it/s] 29%|██▉ | 13842/48008 [47:19<2:06:01, 4.52it/s] 29%|██▉ | 13843/48008 [47:20<2:01:32, 4.69it/s] 29%|██▉ | 13844/48008 [47:20<2:00:59, 4.71it/s] 29%|██▉ | 13845/48008 [47:20<1:58:13, 4.82it/s] 29%|██▉ | 13846/48008 [47:20<1:58:07, 4.82it/s] 29%|██▉ | 13847/48008 [47:20<1:57:06, 4.86it/s] 29%|██▉ | 13848/48008 [47:21<1:56:32, 4.89it/s] 29%|██▉ | 13849/48008 [47:21<1:55:05, 4.95it/s] 29%|██▉ | 13850/48008 [47:21<1:56:40, 4.88it/s] {'loss': 0.3, 'grad_norm': 0.2060546875, 'learning_rate': 0.00014230544909181805, 'epoch': 0.29} 29%|██▉ | 13850/48008 [47:21<1:56:40, 4.88it/s] 29%|██▉ | 13851/48008 [47:21<1:53:37, 5.01it/s] 29%|██▉ | 13852/48008 [47:22<2:25:46, 3.91it/s] 29%|██▉ | 13853/48008 [47:22<2:22:33, 3.99it/s] 29%|██▉ | 13854/48008 [47:22<2:12:45, 4.29it/s] 29%|██▉ | 13855/48008 [47:22<2:04:31, 4.57it/s] 29%|██▉ | 13856/48008 [47:22<1:58:56, 4.79it/s] 29%|██▉ | 13857/48008 [47:22<1:55:32, 4.93it/s] 29%|██▉ | 13858/48008 [47:23<1:52:53, 5.04it/s] 29%|██▉ | 13859/48008 [47:23<1:51:01, 5.13it/s] 29%|██▉ | 13860/48008 [47:23<2:23:50, 3.96it/s] 29%|██▉ | 13861/48008 [47:23<2:13:25, 4.27it/s] 29%|██▉ | 13862/48008 [47:24<2:05:35, 4.53it/s] 29%|██▉ | 13863/48008 [47:24<1:59:48, 4.75it/s] 29%|██▉ | 13864/48008 [47:24<1:56:31, 4.88it/s] 29%|██▉ | 13865/48008 [47:24<1:56:16, 4.89it/s] 29%|██▉ | 13866/48008 [47:24<1:53:28, 5.01it/s] 29%|██▉ | 13867/48008 [47:25<1:54:11, 4.98it/s] 29%|██▉ | 13868/48008 [47:25<1:52:40, 5.05it/s] 29%|██▉ | 13869/48008 [47:25<1:50:59, 5.13it/s] 29%|██▉ | 13870/48008 [47:25<1:51:20, 5.11it/s] 29%|██▉ | 13871/48008 [47:25<1:50:29, 5.15it/s] 29%|██▉ | 13872/48008 [47:26<1:49:29, 5.20it/s] 29%|██▉ | 13873/48008 [47:26<1:49:38, 5.19it/s] 29%|██▉ | 13874/48008 [47:26<1:49:34, 5.19it/s] 29%|██▉ | 13875/48008 [47:26<1:49:28, 5.20it/s] 29%|██▉ | 13876/48008 [47:26<1:49:20, 5.20it/s] 29%|██▉ | 13877/48008 [47:27<1:49:37, 5.19it/s] 29%|██▉ | 13878/48008 [47:27<1:49:00, 5.22it/s] 29%|██▉ | 13879/48008 [47:27<1:49:07, 5.21it/s] 29%|██▉ | 13880/48008 [47:27<1:51:59, 5.08it/s] 29%|██▉ | 13881/48008 [47:27<1:52:58, 5.03it/s] 29%|██▉ | 13882/48008 [47:28<1:53:48, 5.00it/s] 29%|██▉ | 13883/48008 [47:28<1:53:32, 5.01it/s] 29%|██▉ | 13884/48008 [47:28<1:53:08, 5.03it/s] 29%|██▉ | 13885/48008 [47:28<1:51:47, 5.09it/s] 29%|██▉ | 13886/48008 [47:28<1:50:44, 5.14it/s] 29%|██▉ | 13887/48008 [47:28<1:50:14, 5.16it/s] 29%|██▉ | 13888/48008 [47:29<1:49:20, 5.20it/s] 29%|██▉ | 13889/48008 [47:29<1:48:41, 5.23it/s] 29%|██▉ | 13890/48008 [47:29<1:48:54, 5.22it/s] 29%|██▉ | 13891/48008 [47:29<1:48:52, 5.22it/s] 29%|██▉ | 13892/48008 [47:30<2:21:53, 4.01it/s] 29%|██▉ | 13893/48008 [47:30<2:12:40, 4.29it/s] 29%|██▉ | 13894/48008 [47:30<2:06:00, 4.51it/s] 29%|██▉ | 13895/48008 [47:30<2:01:07, 4.69it/s] 29%|██▉ | 13896/48008 [47:30<1:57:13, 4.85it/s] 29%|██▉ | 13897/48008 [47:31<1:54:49, 4.95it/s] 29%|██▉ | 13898/48008 [47:31<1:53:04, 5.03it/s] 29%|██▉ | 13899/48008 [47:31<1:53:05, 5.03it/s] 29%|██▉ | 13900/48008 [47:31<1:54:11, 4.98it/s] {'loss': 0.301, 'grad_norm': 0.12890625, 'learning_rate': 0.00014209715047492085, 'epoch': 0.29} 29%|██▉ | 13900/48008 [47:31<1:54:11, 4.98it/s] 29%|██▉ | 13901/48008 [47:31<1:53:15, 5.02it/s] 29%|██▉ | 13902/48008 [47:32<1:49:51, 5.17it/s] 29%|██▉ | 13903/48008 [47:32<1:48:54, 5.22it/s] 29%|██▉ | 13904/48008 [47:32<1:48:43, 5.23it/s] 29%|██▉ | 13905/48008 [47:32<1:48:48, 5.22it/s] 29%|██▉ | 13906/48008 [47:32<1:50:54, 5.12it/s] 29%|██▉ | 13907/48008 [47:33<1:49:37, 5.18it/s] 29%|██▉ | 13908/48008 [47:33<1:50:07, 5.16it/s] 29%|██▉ | 13909/48008 [47:33<2:09:21, 4.39it/s] 29%|██▉ | 13910/48008 [47:33<2:04:22, 4.57it/s] 29%|██▉ | 13911/48008 [47:33<2:00:14, 4.73it/s] 29%|██▉ | 13912/48008 [47:34<1:58:30, 4.80it/s] 29%|██▉ | 13913/48008 [47:34<1:54:44, 4.95it/s] 29%|██▉ | 13914/48008 [47:34<1:52:42, 5.04it/s] 29%|██▉ | 13915/48008 [47:34<1:52:48, 5.04it/s] 29%|██▉ | 13916/48008 [47:34<1:51:50, 5.08it/s] 29%|██▉ | 13917/48008 [47:35<1:50:59, 5.12it/s] 29%|██▉ | 13918/48008 [47:35<1:53:16, 5.02it/s] 29%|██▉ | 13919/48008 [47:35<1:52:05, 5.07it/s] 29%|██▉ | 13920/48008 [47:35<1:50:51, 5.13it/s] 29%|██▉ | 13921/48008 [47:35<1:50:13, 5.15it/s] 29%|██▉ | 13922/48008 [47:36<1:50:49, 5.13it/s] 29%|██▉ | 13923/48008 [47:36<1:58:01, 4.81it/s] 29%|██▉ | 13924/48008 [47:36<1:54:24, 4.97it/s] 29%|██▉ | 13925/48008 [47:36<1:52:39, 5.04it/s] 29%|██▉ | 13926/48008 [47:36<1:59:16, 4.76it/s] 29%|██▉ | 13927/48008 [47:37<1:56:02, 4.90it/s] 29%|██▉ | 13928/48008 [47:37<1:53:57, 4.98it/s] 29%|██▉ | 13929/48008 [47:37<1:51:14, 5.11it/s] 29%|██▉ | 13930/48008 [47:37<1:51:10, 5.11it/s] 29%|██▉ | 13931/48008 [47:37<1:51:36, 5.09it/s] 29%|██▉ | 13932/48008 [47:38<1:49:46, 5.17it/s] 29%|██▉ | 13933/48008 [47:38<1:50:06, 5.16it/s] 29%|██▉ | 13934/48008 [47:38<1:57:35, 4.83it/s] 29%|██▉ | 13935/48008 [47:38<1:54:28, 4.96it/s] 29%|██▉ | 13936/48008 [47:38<1:51:42, 5.08it/s] 29%|██▉ | 13937/48008 [47:39<1:53:55, 4.98it/s] 29%|██▉ | 13938/48008 [47:39<1:51:51, 5.08it/s] 29%|██▉ | 13939/48008 [47:39<1:50:43, 5.13it/s] 29%|██▉ | 13940/48008 [47:39<1:50:07, 5.16it/s] 29%|██▉ | 13941/48008 [47:39<1:48:39, 5.23it/s] 29%|██▉ | 13942/48008 [47:40<1:50:43, 5.13it/s] 29%|██▉ | 13943/48008 [47:40<1:50:14, 5.15it/s] 29%|██▉ | 13944/48008 [47:40<1:50:37, 5.13it/s] 29%|██▉ | 13945/48008 [47:40<1:49:14, 5.20it/s] 29%|██▉ | 13946/48008 [47:40<1:49:09, 5.20it/s] 29%|██▉ | 13947/48008 [47:41<1:49:14, 5.20it/s] 29%|██▉ | 13948/48008 [47:41<1:49:45, 5.17it/s] 29%|██▉ | 13949/48008 [47:41<1:52:31, 5.04it/s] 29%|██▉ | 13950/48008 [47:41<1:51:08, 5.11it/s] {'loss': 0.3045, 'grad_norm': 0.4296875, 'learning_rate': 0.00014188885185802366, 'epoch': 0.29} 29%|██▉ | 13950/48008 [47:41<1:51:08, 5.11it/s] 29%|██▉ | 13951/48008 [47:41<1:49:51, 5.17it/s] 29%|██▉ | 13952/48008 [47:41<1:50:24, 5.14it/s] 29%|██▉ | 13953/48008 [47:42<1:49:50, 5.17it/s] 29%|██▉ | 13954/48008 [47:42<1:48:50, 5.21it/s] 29%|██▉ | 13955/48008 [47:42<1:49:52, 5.17it/s] 29%|██▉ | 13956/48008 [47:42<1:49:29, 5.18it/s] 29%|██▉ | 13957/48008 [47:42<1:49:11, 5.20it/s] 29%|██▉ | 13958/48008 [47:43<1:49:06, 5.20it/s] 29%|██▉ | 13959/48008 [47:43<1:49:07, 5.20it/s] 29%|██▉ | 13960/48008 [47:43<1:56:43, 4.86it/s] 29%|██▉ | 13961/48008 [47:43<1:56:32, 4.87it/s] 29%|██▉ | 13962/48008 [47:43<1:53:32, 5.00it/s] 29%|██▉ | 13963/48008 [47:44<1:51:59, 5.07it/s] 29%|██▉ | 13964/48008 [47:44<1:51:07, 5.11it/s] 29%|██▉ | 13965/48008 [47:44<1:52:16, 5.05it/s] 29%|██▉ | 13966/48008 [47:44<1:50:49, 5.12it/s] 29%|██▉ | 13967/48008 [47:44<1:51:56, 5.07it/s] 29%|██▉ | 13968/48008 [47:45<1:50:09, 5.15it/s] 29%|██▉ | 13969/48008 [47:45<1:50:39, 5.13it/s] 29%|██▉ | 13970/48008 [47:45<1:51:10, 5.10it/s] 29%|██▉ | 13971/48008 [47:45<1:52:49, 5.03it/s] 29%|██▉ | 13972/48008 [47:45<1:51:47, 5.07it/s] 29%|██▉ | 13973/48008 [47:46<1:53:01, 5.02it/s] 29%|██▉ | 13974/48008 [47:46<1:51:47, 5.07it/s] 29%|██▉ | 13975/48008 [47:46<1:50:07, 5.15it/s] 29%|██▉ | 13976/48008 [47:46<1:52:42, 5.03it/s] 29%|██▉ | 13977/48008 [47:46<1:51:46, 5.07it/s] 29%|██▉ | 13978/48008 [47:47<1:51:01, 5.11it/s] 29%|██▉ | 13979/48008 [47:47<1:49:29, 5.18it/s] 29%|██▉ | 13980/48008 [47:47<1:51:18, 5.10it/s] 29%|██▉ | 13981/48008 [47:47<1:49:49, 5.16it/s] 29%|██▉ | 13982/48008 [47:47<1:49:04, 5.20it/s] 29%|██▉ | 13983/48008 [47:48<1:50:01, 5.15it/s] 29%|██▉ | 13984/48008 [47:48<1:49:27, 5.18it/s] 29%|██▉ | 13985/48008 [47:48<1:56:54, 4.85it/s] 29%|██▉ | 13986/48008 [47:48<1:54:07, 4.97it/s] 29%|██▉ | 13987/48008 [47:48<1:52:07, 5.06it/s] 29%|██▉ | 13988/48008 [47:49<1:58:55, 4.77it/s] 29%|██▉ | 13989/48008 [47:49<1:58:17, 4.79it/s] 29%|██▉ | 13990/48008 [47:49<1:55:07, 4.93it/s] 29%|██▉ | 13991/48008 [47:49<1:54:59, 4.93it/s] 29%|██▉ | 13992/48008 [47:49<1:52:51, 5.02it/s] 29%|██▉ | 13993/48008 [47:50<1:50:48, 5.12it/s] 29%|██▉ | 13994/48008 [47:50<1:50:42, 5.12it/s] 29%|██▉ | 13995/48008 [47:50<1:50:10, 5.15it/s] 29%|██▉ | 13996/48008 [47:50<1:48:52, 5.21it/s] 29%|██▉ | 13997/48008 [47:50<1:48:13, 5.24it/s] 29%|██▉ | 13998/48008 [47:51<1:48:30, 5.22it/s] 29%|██▉ | 13999/48008 [47:51<1:48:16, 5.23it/s] 29%|██▉ | 14000/48008 [47:51<1:50:36, 5.12it/s] {'loss': 0.3009, 'grad_norm': 0.345703125, 'learning_rate': 0.00014168055324112648, 'epoch': 0.29} 29%|██▉ | 14000/48008 [47:51<1:50:36, 5.12it/s] 29%|██▉ | 14001/48008 [47:51<1:50:07, 5.15it/s] 29%|██▉ | 14002/48008 [47:51<1:51:25, 5.09it/s] 29%|██▉ | 14003/48008 [47:52<1:58:22, 4.79it/s] 29%|██▉ | 14004/48008 [47:52<1:54:37, 4.94it/s] 29%|██▉ | 14005/48008 [47:52<1:52:49, 5.02it/s] 29%|██▉ | 14006/48008 [47:52<2:11:29, 4.31it/s] 29%|██▉ | 14007/48008 [47:52<2:04:32, 4.55it/s] 29%|██▉ | 14008/48008 [47:53<2:32:54, 3.71it/s] 29%|██▉ | 14009/48008 [47:53<2:17:56, 4.11it/s] 29%|██▉ | 14010/48008 [47:53<2:10:12, 4.35it/s] 29%|██▉ | 14011/48008 [47:53<2:02:44, 4.62it/s] 29%|██▉ | 14012/48008 [47:54<1:58:06, 4.80it/s] 29%|██▉ | 14013/48008 [47:54<1:56:42, 4.85it/s] 29%|██▉ | 14014/48008 [47:54<1:55:27, 4.91it/s] 29%|██▉ | 14015/48008 [47:54<1:54:09, 4.96it/s] 29%|██▉ | 14016/48008 [47:54<1:53:16, 5.00it/s] 29%|██▉ | 14017/48008 [47:55<1:51:45, 5.07it/s] 29%|██▉ | 14018/48008 [47:55<1:53:43, 4.98it/s] 29%|██▉ | 14019/48008 [47:55<1:53:08, 5.01it/s] 29%|██▉ | 14020/48008 [47:55<1:52:13, 5.05it/s] 29%|██▉ | 14021/48008 [47:55<1:54:24, 4.95it/s] 29%|██▉ | 14022/48008 [47:56<1:52:11, 5.05it/s] 29%|██▉ | 14023/48008 [47:56<1:51:49, 5.06it/s] 29%|██▉ | 14024/48008 [47:56<1:51:51, 5.06it/s] 29%|██▉ | 14025/48008 [47:56<1:50:19, 5.13it/s] 29%|██▉ | 14026/48008 [47:56<1:49:35, 5.17it/s] 29%|██▉ | 14027/48008 [47:57<1:49:24, 5.18it/s] 29%|██▉ | 14028/48008 [47:57<1:48:56, 5.20it/s] 29%|██▉ | 14029/48008 [47:57<1:49:51, 5.15it/s] 29%|██▉ | 14030/48008 [47:57<1:49:08, 5.19it/s] 29%|██▉ | 14031/48008 [47:57<1:52:19, 5.04it/s] 29%|██▉ | 14032/48008 [47:58<1:51:13, 5.09it/s] 29%|██▉ | 14033/48008 [47:58<1:58:20, 4.79it/s] 29%|██▉ | 14034/48008 [47:58<1:54:37, 4.94it/s] 29%|██▉ | 14035/48008 [47:58<1:53:42, 4.98it/s] 29%|██▉ | 14036/48008 [47:58<2:12:55, 4.26it/s] 29%|██▉ | 14037/48008 [47:59<2:06:22, 4.48it/s] 29%|██▉ | 14038/48008 [47:59<2:21:59, 3.99it/s] 29%|██▉ | 14039/48008 [47:59<2:12:14, 4.28it/s] 29%|██▉ | 14040/48008 [47:59<2:04:19, 4.55it/s] 29%|██▉ | 14041/48008 [48:00<2:01:38, 4.65it/s] 29%|██▉ | 14042/48008 [48:00<2:17:28, 4.12it/s] 29%|██▉ | 14043/48008 [48:00<2:08:44, 4.40it/s] 29%|██▉ | 14044/48008 [48:00<2:06:00, 4.49it/s] 29%|██▉ | 14045/48008 [48:00<2:00:58, 4.68it/s] 29%|██▉ | 14046/48008 [48:01<1:57:08, 4.83it/s] 29%|██▉ | 14047/48008 [48:01<1:54:21, 4.95it/s] 29%|██▉ | 14048/48008 [48:01<1:52:24, 5.04it/s] 29%|██▉ | 14049/48008 [48:01<1:50:54, 5.10it/s] 29%|██▉ | 14050/48008 [48:01<1:49:38, 5.16it/s] {'loss': 0.3012, 'grad_norm': 0.1298828125, 'learning_rate': 0.0001414722546242293, 'epoch': 0.29} 29%|██▉ | 14050/48008 [48:01<1:49:38, 5.16it/s] 29%|██▉ | 14051/48008 [48:02<1:48:48, 5.20it/s] 29%|██▉ | 14052/48008 [48:02<1:48:35, 5.21it/s] 29%|██▉ | 14053/48008 [48:02<1:49:01, 5.19it/s] 29%|██▉ | 14054/48008 [48:02<1:51:57, 5.05it/s] 29%|██▉ | 14055/48008 [48:02<2:10:13, 4.35it/s] 29%|██▉ | 14056/48008 [48:03<2:04:01, 4.56it/s] 29%|██▉ | 14057/48008 [48:03<1:58:22, 4.78it/s] 29%|██▉ | 14058/48008 [48:03<1:55:22, 4.90it/s] 29%|██▉ | 14059/48008 [48:03<1:53:44, 4.97it/s] 29%|██▉ | 14060/48008 [48:03<1:51:44, 5.06it/s] 29%|██▉ | 14061/48008 [48:04<1:50:37, 5.11it/s] 29%|██▉ | 14062/48008 [48:04<1:49:35, 5.16it/s] 29%|██▉ | 14063/48008 [48:04<1:48:47, 5.20it/s] 29%|██▉ | 14064/48008 [48:04<1:48:42, 5.20it/s] 29%|██▉ | 14065/48008 [48:04<1:48:08, 5.23it/s] 29%|██▉ | 14066/48008 [48:05<1:49:12, 5.18it/s] 29%|██▉ | 14067/48008 [48:05<1:50:32, 5.12it/s] 29%|██▉ | 14068/48008 [48:05<1:49:09, 5.18it/s] 29%|██▉ | 14069/48008 [48:05<1:50:55, 5.10it/s] 29%|██▉ | 14070/48008 [48:05<1:53:43, 4.97it/s] 29%|██▉ | 14071/48008 [48:06<1:52:30, 5.03it/s] 29%|██▉ | 14072/48008 [48:06<1:51:28, 5.07it/s] 29%|██▉ | 14073/48008 [48:06<1:50:15, 5.13it/s] 29%|██▉ | 14074/48008 [48:06<1:48:42, 5.20it/s] 29%|██▉ | 14075/48008 [48:06<1:49:09, 5.18it/s] 29%|██▉ | 14076/48008 [48:07<1:50:19, 5.13it/s] 29%|██▉ | 14077/48008 [48:07<1:49:15, 5.18it/s] 29%|██▉ | 14078/48008 [48:07<2:21:50, 3.99it/s] 29%|██▉ | 14079/48008 [48:07<2:10:45, 4.32it/s] 29%|██▉ | 14080/48008 [48:08<2:07:10, 4.45it/s] 29%|██▉ | 14081/48008 [48:08<2:01:22, 4.66it/s] 29%|██▉ | 14082/48008 [48:08<2:05:22, 4.51it/s] 29%|██▉ | 14083/48008 [48:08<1:59:19, 4.74it/s] 29%|██▉ | 14084/48008 [48:08<1:56:10, 4.87it/s] 29%|██▉ | 14085/48008 [48:09<1:55:13, 4.91it/s] 29%|██▉ | 14086/48008 [48:09<1:55:01, 4.92it/s] 29%|██▉ | 14087/48008 [48:09<1:52:56, 5.01it/s] 29%|██▉ | 14088/48008 [48:09<1:53:16, 4.99it/s] 29%|██▉ | 14089/48008 [48:09<1:51:10, 5.08it/s] 29%|██▉ | 14090/48008 [48:10<1:50:10, 5.13it/s] 29%|██▉ | 14091/48008 [48:10<1:48:49, 5.19it/s] 29%|██▉ | 14092/48008 [48:10<1:50:22, 5.12it/s] 29%|██▉ | 14093/48008 [48:10<1:58:04, 4.79it/s] 29%|██▉ | 14094/48008 [48:10<1:55:54, 4.88it/s] 29%|██▉ | 14095/48008 [48:11<1:53:27, 4.98it/s] 29%|██▉ | 14096/48008 [48:11<1:51:26, 5.07it/s] 29%|██▉ | 14097/48008 [48:11<1:49:30, 5.16it/s] 29%|██▉ | 14098/48008 [48:11<1:49:57, 5.14it/s] 29%|██▉ | 14099/48008 [48:11<1:49:11, 5.18it/s] 29%|██▉ | 14100/48008 [48:11<1:48:36, 5.20it/s] {'loss': 0.3017, 'grad_norm': 0.421875, 'learning_rate': 0.00014126395600733211, 'epoch': 0.29} 29%|██▉ | 14100/48008 [48:11<1:48:36, 5.20it/s] 29%|██▉ | 14101/48008 [48:12<1:50:37, 5.11it/s] 29%|██▉ | 14102/48008 [48:12<1:49:55, 5.14it/s] 29%|██▉ | 14103/48008 [48:12<1:48:59, 5.18it/s] 29%|██▉ | 14104/48008 [48:12<1:48:22, 5.21it/s] 29%|██▉ | 14105/48008 [48:12<1:47:59, 5.23it/s] 29%|██▉ | 14106/48008 [48:13<1:48:45, 5.20it/s] 29%|██▉ | 14107/48008 [48:13<1:48:38, 5.20it/s] 29%|██▉ | 14108/48008 [48:13<1:50:38, 5.11it/s] 29%|██▉ | 14109/48008 [48:13<1:48:43, 5.20it/s] 29%|██▉ | 14110/48008 [48:13<1:47:57, 5.23it/s] 29%|██▉ | 14111/48008 [48:14<1:47:37, 5.25it/s] 29%|██▉ | 14112/48008 [48:14<1:48:42, 5.20it/s] 29%|██▉ | 14113/48008 [48:14<1:48:33, 5.20it/s] 29%|██▉ | 14114/48008 [48:14<2:07:44, 4.42it/s] 29%|██▉ | 14115/48008 [48:14<2:02:07, 4.63it/s] 29%|██▉ | 14116/48008 [48:15<2:00:14, 4.70it/s] 29%|██▉ | 14117/48008 [48:15<1:58:30, 4.77it/s] 29%|██▉ | 14118/48008 [48:15<1:55:13, 4.90it/s] 29%|██▉ | 14119/48008 [48:15<1:51:41, 5.06it/s] 29%|██▉ | 14120/48008 [48:15<1:51:18, 5.07it/s] 29%|██▉ | 14121/48008 [48:16<1:58:11, 4.78it/s] 29%|██▉ | 14122/48008 [48:16<1:54:44, 4.92it/s] 29%|██▉ | 14123/48008 [48:16<1:52:55, 5.00it/s] 29%|██▉ | 14124/48008 [48:16<1:50:57, 5.09it/s] 29%|██▉ | 14125/48008 [48:16<1:51:07, 5.08it/s] 29%|██▉ | 14126/48008 [48:17<1:52:30, 5.02it/s] 29%|██▉ | 14127/48008 [48:17<1:53:32, 4.97it/s] 29%|██▉ | 14128/48008 [48:17<1:52:51, 5.00it/s] 29%|██▉ | 14129/48008 [48:17<1:51:31, 5.06it/s] 29%|██▉ | 14130/48008 [48:17<1:49:16, 5.17it/s] 29%|██▉ | 14131/48008 [48:18<1:50:47, 5.10it/s] 29%|██▉ | 14132/48008 [48:18<1:51:46, 5.05it/s] 29%|██▉ | 14133/48008 [48:18<1:50:33, 5.11it/s] 29%|██▉ | 14134/48008 [48:18<1:49:52, 5.14it/s] 29%|██▉ | 14135/48008 [48:18<1:57:21, 4.81it/s] 29%|██▉ | 14136/48008 [48:19<1:54:47, 4.92it/s] 29%|██▉ | 14137/48008 [48:19<1:52:08, 5.03it/s] 29%|██▉ | 14138/48008 [48:19<1:52:47, 5.00it/s] 29%|██▉ | 14139/48008 [48:19<1:51:16, 5.07it/s] 29%|██▉ | 14140/48008 [48:19<1:51:39, 5.06it/s] 29%|██▉ | 14141/48008 [48:20<1:50:22, 5.11it/s] 29%|██▉ | 14142/48008 [48:20<2:09:01, 4.37it/s] 29%|██▉ | 14143/48008 [48:20<2:04:54, 4.52it/s] 29%|██▉ | 14144/48008 [48:20<1:59:05, 4.74it/s] 29%|██▉ | 14145/48008 [48:21<1:57:29, 4.80it/s] 29%|██▉ | 14146/48008 [48:21<1:55:23, 4.89it/s] 29%|██▉ | 14147/48008 [48:21<2:26:13, 3.86it/s] 29%|██▉ | 14148/48008 [48:21<2:16:32, 4.13it/s] 29%|██▉ | 14149/48008 [48:22<2:08:00, 4.41it/s] 29%|██▉ | 14150/48008 [48:22<2:04:05, 4.55it/s] {'loss': 0.3014, 'grad_norm': 0.1552734375, 'learning_rate': 0.00014105565739043493, 'epoch': 0.29} 29%|██▉ | 14150/48008 [48:22<2:04:05, 4.55it/s] 29%|██▉ | 14151/48008 [48:22<1:59:33, 4.72it/s] 29%|██▉ | 14152/48008 [48:22<1:54:17, 4.94it/s] 29%|██▉ | 14153/48008 [48:22<1:52:10, 5.03it/s] 29%|██▉ | 14154/48008 [48:22<1:52:05, 5.03it/s] 29%|██▉ | 14155/48008 [48:23<1:50:39, 5.10it/s] 29%|██▉ | 14156/48008 [48:23<1:49:55, 5.13it/s] 29%|██▉ | 14157/48008 [48:23<1:49:11, 5.17it/s] 29%|██▉ | 14158/48008 [48:23<1:48:57, 5.18it/s] 29%|██▉ | 14159/48008 [48:23<1:48:45, 5.19it/s] 29%|██▉ | 14160/48008 [48:24<1:48:14, 5.21it/s] 29%|██▉ | 14161/48008 [48:24<1:48:25, 5.20it/s] 29%|██▉ | 14162/48008 [48:24<1:47:56, 5.23it/s] 30%|██▉ | 14163/48008 [48:24<1:49:25, 5.15it/s] 30%|██▉ | 14164/48008 [48:24<1:48:17, 5.21it/s] 30%|██▉ | 14165/48008 [48:25<1:48:50, 5.18it/s] 30%|██▉ | 14166/48008 [48:25<1:48:21, 5.21it/s] 30%|██▉ | 14167/48008 [48:25<1:48:07, 5.22it/s] 30%|██▉ | 14168/48008 [48:25<2:20:48, 4.01it/s] 30%|██▉ | 14169/48008 [48:26<2:09:04, 4.37it/s] 30%|██▉ | 14170/48008 [48:26<2:02:50, 4.59it/s] 30%|██▉ | 14171/48008 [48:26<1:59:05, 4.74it/s] 30%|██▉ | 14172/48008 [48:26<1:55:14, 4.89it/s] 30%|██▉ | 14173/48008 [48:26<1:52:34, 5.01it/s] 30%|██▉ | 14174/48008 [48:26<1:51:16, 5.07it/s] 30%|██▉ | 14175/48008 [48:27<1:50:49, 5.09it/s] 30%|██▉ | 14176/48008 [48:27<1:48:57, 5.18it/s] 30%|██▉ | 14177/48008 [48:27<1:48:38, 5.19it/s] 30%|██▉ | 14178/48008 [48:27<1:47:51, 5.23it/s] 30%|██▉ | 14179/48008 [48:27<1:51:10, 5.07it/s] 30%|██▉ | 14180/48008 [48:28<1:49:17, 5.16it/s] 30%|██▉ | 14181/48008 [48:28<1:48:43, 5.19it/s] 30%|██▉ | 14182/48008 [48:28<1:49:38, 5.14it/s] 30%|██▉ | 14183/48008 [48:28<1:49:58, 5.13it/s] 30%|██▉ | 14184/48008 [48:28<1:49:08, 5.17it/s] 30%|██▉ | 14185/48008 [48:29<1:48:03, 5.22it/s] 30%|██▉ | 14186/48008 [48:29<1:47:11, 5.26it/s] 30%|██▉ | 14187/48008 [48:29<2:20:20, 4.02it/s] 30%|██▉ | 14188/48008 [48:29<2:11:22, 4.29it/s] 30%|██▉ | 14189/48008 [48:30<2:03:48, 4.55it/s] 30%|██▉ | 14190/48008 [48:30<2:00:10, 4.69it/s] 30%|██▉ | 14191/48008 [48:30<1:56:32, 4.84it/s] 30%|██▉ | 14192/48008 [48:30<1:53:51, 4.95it/s] 30%|██▉ | 14193/48008 [48:30<1:51:21, 5.06it/s] 30%|██▉ | 14194/48008 [48:31<1:50:04, 5.12it/s] 30%|██▉ | 14195/48008 [48:31<1:52:29, 5.01it/s] 30%|██▉ | 14196/48008 [48:31<1:51:06, 5.07it/s] 30%|██▉ | 14197/48008 [48:31<1:52:02, 5.03it/s] 30%|██▉ | 14198/48008 [48:32<2:23:41, 3.92it/s] 30%|██▉ | 14199/48008 [48:32<2:12:21, 4.26it/s] 30%|██▉ | 14200/48008 [48:32<2:04:29, 4.53it/s] {'loss': 0.3041, 'grad_norm': 0.259765625, 'learning_rate': 0.00014084735877353775, 'epoch': 0.3} 30%|██▉ | 14200/48008 [48:32<2:04:29, 4.53it/s] 30%|██▉ | 14201/48008 [48:32<2:00:02, 4.69it/s] 30%|██▉ | 14202/48008 [48:32<1:56:24, 4.84it/s] 30%|██▉ | 14203/48008 [48:32<1:53:07, 4.98it/s] 30%|██▉ | 14204/48008 [48:33<1:53:04, 4.98it/s] 30%|██▉ | 14205/48008 [48:33<1:51:32, 5.05it/s] 30%|██▉ | 14206/48008 [48:33<1:50:14, 5.11it/s] 30%|██▉ | 14207/48008 [48:33<1:49:34, 5.14it/s] 30%|██▉ | 14208/48008 [48:34<2:21:57, 3.97it/s] 30%|██▉ | 14209/48008 [48:34<2:12:05, 4.26it/s] 30%|██▉ | 14210/48008 [48:34<2:05:47, 4.48it/s] 30%|██▉ | 14211/48008 [48:34<2:00:48, 4.66it/s] 30%|██▉ | 14212/48008 [48:34<1:57:07, 4.81it/s] 30%|██▉ | 14213/48008 [48:35<1:54:37, 4.91it/s] 30%|██▉ | 14214/48008 [48:35<1:52:50, 4.99it/s] 30%|██▉ | 14215/48008 [48:35<1:51:08, 5.07it/s] 30%|██▉ | 14216/48008 [48:35<1:50:28, 5.10it/s] 30%|██▉ | 14217/48008 [48:35<1:48:53, 5.17it/s] 30%|██▉ | 14218/48008 [48:36<2:21:56, 3.97it/s] 30%|██▉ | 14219/48008 [48:36<2:11:36, 4.28it/s] 30%|██▉ | 14220/48008 [48:36<2:04:31, 4.52it/s] 30%|██▉ | 14221/48008 [48:36<1:59:34, 4.71it/s] 30%|██▉ | 14222/48008 [48:37<1:56:41, 4.83it/s] 30%|██▉ | 14223/48008 [48:37<1:56:01, 4.85it/s] 30%|██▉ | 14224/48008 [48:37<1:53:31, 4.96it/s] 30%|██▉ | 14225/48008 [48:37<1:51:17, 5.06it/s] 30%|██▉ | 14226/48008 [48:37<1:58:09, 4.76it/s] 30%|██▉ | 14227/48008 [48:38<1:53:54, 4.94it/s] 30%|██▉ | 14228/48008 [48:38<1:50:46, 5.08it/s] 30%|██▉ | 14229/48008 [48:38<1:49:52, 5.12it/s] 30%|██▉ | 14230/48008 [48:38<1:49:04, 5.16it/s] 30%|██▉ | 14231/48008 [48:38<1:49:50, 5.12it/s] 30%|██▉ | 14232/48008 [48:38<1:48:43, 5.18it/s] 30%|██▉ | 14233/48008 [48:39<1:50:26, 5.10it/s] 30%|██▉ | 14234/48008 [48:39<1:50:57, 5.07it/s] 30%|██▉ | 14235/48008 [48:39<1:49:44, 5.13it/s] 30%|██▉ | 14236/48008 [48:39<1:49:19, 5.15it/s] 30%|██▉ | 14237/48008 [48:39<1:50:53, 5.08it/s] 30%|██▉ | 14238/48008 [48:40<1:49:08, 5.16it/s] 30%|██▉ | 14239/48008 [48:40<1:50:34, 5.09it/s] 30%|██▉ | 14240/48008 [48:40<1:49:34, 5.14it/s] 30%|██▉ | 14241/48008 [48:40<1:51:09, 5.06it/s] 30%|██▉ | 14242/48008 [48:40<1:50:21, 5.10it/s] 30%|██▉ | 14243/48008 [48:41<1:49:06, 5.16it/s] 30%|██▉ | 14244/48008 [48:41<1:48:42, 5.18it/s] 30%|██▉ | 14245/48008 [48:41<1:48:55, 5.17it/s] 30%|██▉ | 14246/48008 [48:41<1:48:32, 5.18it/s] 30%|██▉ | 14247/48008 [48:42<2:21:09, 3.99it/s] 30%|██▉ | 14248/48008 [48:42<2:10:10, 4.32it/s] 30%|██▉ | 14249/48008 [48:42<2:04:10, 4.53it/s] 30%|██▉ | 14250/48008 [48:42<1:59:30, 4.71it/s] {'loss': 0.3014, 'grad_norm': 0.2890625, 'learning_rate': 0.00014063906015664057, 'epoch': 0.3} 30%|██▉ | 14250/48008 [48:42<1:59:30, 4.71it/s] 30%|██▉ | 14251/48008 [48:42<1:59:06, 4.72it/s] 30%|██▉ | 14252/48008 [48:43<1:55:45, 4.86it/s] 30%|██▉ | 14253/48008 [48:43<1:54:48, 4.90it/s] 30%|██▉ | 14254/48008 [48:43<1:52:24, 5.00it/s] 30%|██▉ | 14255/48008 [48:43<1:50:32, 5.09it/s] 30%|██▉ | 14256/48008 [48:43<1:50:24, 5.10it/s] 30%|██▉ | 14257/48008 [48:44<1:49:25, 5.14it/s] 30%|██▉ | 14258/48008 [48:44<1:48:17, 5.19it/s] 30%|██▉ | 14259/48008 [48:44<1:48:19, 5.19it/s] 30%|██▉ | 14260/48008 [48:44<1:47:36, 5.23it/s] 30%|██▉ | 14261/48008 [48:44<1:47:38, 5.23it/s] 30%|██▉ | 14262/48008 [48:44<1:47:32, 5.23it/s] 30%|██▉ | 14263/48008 [48:45<1:47:38, 5.22it/s] 30%|██▉ | 14264/48008 [48:45<1:47:02, 5.25it/s] 30%|██▉ | 14265/48008 [48:45<1:46:39, 5.27it/s] 30%|██▉ | 14266/48008 [48:45<1:47:00, 5.26it/s] 30%|██▉ | 14267/48008 [48:45<1:47:00, 5.25it/s] 30%|██▉ | 14268/48008 [48:46<1:47:58, 5.21it/s] 30%|██▉ | 14269/48008 [48:46<1:47:37, 5.22it/s] 30%|██▉ | 14270/48008 [48:46<1:47:35, 5.23it/s] 30%|██▉ | 14271/48008 [48:46<1:47:38, 5.22it/s] 30%|██▉ | 14272/48008 [48:46<1:47:22, 5.24it/s] 30%|██▉ | 14273/48008 [48:47<1:55:05, 4.88it/s] 30%|██▉ | 14274/48008 [48:47<1:53:44, 4.94it/s] 30%|██▉ | 14275/48008 [48:47<1:53:02, 4.97it/s] 30%|██▉ | 14276/48008 [48:47<2:24:32, 3.89it/s] 30%|██▉ | 14277/48008 [48:48<2:13:43, 4.20it/s] 30%|██▉ | 14278/48008 [48:48<2:04:52, 4.50it/s] 30%|██▉ | 14279/48008 [48:48<2:32:22, 3.69it/s] 30%|██▉ | 14280/48008 [48:48<2:19:15, 4.04it/s] 30%|██▉ | 14281/48008 [48:49<2:09:19, 4.35it/s] 30%|██▉ | 14282/48008 [48:49<2:04:32, 4.51it/s] 30%|██▉ | 14283/48008 [48:49<1:59:31, 4.70it/s] 30%|██▉ | 14284/48008 [48:49<1:57:52, 4.77it/s] 30%|██▉ | 14285/48008 [48:49<1:55:23, 4.87it/s] 30%|██▉ | 14286/48008 [48:50<1:53:03, 4.97it/s] 30%|██▉ | 14287/48008 [48:50<2:24:17, 3.89it/s] 30%|██▉ | 14288/48008 [48:50<2:13:26, 4.21it/s] 30%|██▉ | 14289/48008 [48:50<2:07:51, 4.40it/s] 30%|██▉ | 14290/48008 [48:51<2:01:54, 4.61it/s] 30%|██▉ | 14291/48008 [48:51<1:56:20, 4.83it/s] 30%|██▉ | 14292/48008 [48:51<1:54:54, 4.89it/s] 30%|██▉ | 14293/48008 [48:51<1:52:50, 4.98it/s] 30%|██▉ | 14294/48008 [48:51<2:24:04, 3.90it/s] 30%|██▉ | 14295/48008 [48:52<2:12:24, 4.24it/s] 30%|██▉ | 14296/48008 [48:52<2:05:50, 4.46it/s] 30%|██▉ | 14297/48008 [48:52<2:02:24, 4.59it/s] 30%|██▉ | 14298/48008 [48:52<1:58:55, 4.72it/s] 30%|██▉ | 14299/48008 [48:52<1:55:50, 4.85it/s] 30%|██▉ | 14300/48008 [48:53<1:55:16, 4.87it/s] {'loss': 0.301, 'grad_norm': 0.2392578125, 'learning_rate': 0.00014043076153974338, 'epoch': 0.3} 30%|██▉ | 14300/48008 [48:53<1:55:16, 4.87it/s] 30%|██▉ | 14301/48008 [48:53<1:53:16, 4.96it/s] 30%|██▉ | 14302/48008 [48:53<1:51:54, 5.02it/s] 30%|██▉ | 14303/48008 [48:53<1:50:51, 5.07it/s] 30%|██▉ | 14304/48008 [48:53<1:49:42, 5.12it/s] 30%|██▉ | 14305/48008 [48:54<1:48:16, 5.19it/s] 30%|██▉ | 14306/48008 [48:54<1:49:41, 5.12it/s] 30%|██▉ | 14307/48008 [48:54<1:51:54, 5.02it/s] 30%|██▉ | 14308/48008 [48:54<1:50:40, 5.07it/s] 30%|██▉ | 14309/48008 [48:54<1:49:31, 5.13it/s] 30%|██▉ | 14310/48008 [48:55<1:51:00, 5.06it/s] 30%|██▉ | 14311/48008 [48:55<1:49:57, 5.11it/s] 30%|██▉ | 14312/48008 [48:55<1:48:57, 5.15it/s] 30%|██▉ | 14313/48008 [48:55<1:48:43, 5.17it/s] 30%|██▉ | 14314/48008 [48:55<1:46:52, 5.25it/s] 30%|██▉ | 14315/48008 [48:56<1:47:00, 5.25it/s] 30%|██▉ | 14316/48008 [48:56<1:47:06, 5.24it/s] 30%|██▉ | 14317/48008 [48:56<1:46:52, 5.25it/s] 30%|██▉ | 14318/48008 [48:56<1:46:10, 5.29it/s] 30%|██▉ | 14319/48008 [48:56<1:47:46, 5.21it/s] 30%|██▉ | 14320/48008 [48:57<1:47:26, 5.23it/s] 30%|██▉ | 14321/48008 [48:57<1:48:22, 5.18it/s] 30%|██▉ | 14322/48008 [48:57<2:20:31, 4.00it/s] 30%|██▉ | 14323/48008 [48:57<2:11:40, 4.26it/s] 30%|██▉ | 14324/48008 [48:57<2:03:26, 4.55it/s] 30%|██▉ | 14325/48008 [48:58<1:58:17, 4.75it/s] 30%|██▉ | 14326/48008 [48:58<1:56:46, 4.81it/s] 30%|██▉ | 14327/48008 [48:58<1:54:29, 4.90it/s] 30%|██▉ | 14328/48008 [48:58<1:51:47, 5.02it/s] 30%|██▉ | 14329/48008 [48:58<1:50:35, 5.08it/s] 30%|██▉ | 14330/48008 [48:59<1:49:28, 5.13it/s] 30%|██▉ | 14331/48008 [48:59<1:51:37, 5.03it/s] 30%|██▉ | 14332/48008 [48:59<1:50:57, 5.06it/s] 30%|██▉ | 14333/48008 [48:59<1:49:42, 5.12it/s] 30%|██▉ | 14334/48008 [48:59<1:51:48, 5.02it/s] 30%|██▉ | 14335/48008 [49:00<2:09:59, 4.32it/s] 30%|██▉ | 14336/48008 [49:00<2:03:37, 4.54it/s] 30%|██▉ | 14337/48008 [49:00<2:18:05, 4.06it/s] 30%|██▉ | 14338/48008 [49:00<2:07:37, 4.40it/s] 30%|██▉ | 14339/48008 [49:01<2:00:39, 4.65it/s] 30%|██▉ | 14340/48008 [49:01<2:29:27, 3.75it/s] 30%|██▉ | 14341/48008 [49:01<2:16:15, 4.12it/s] 30%|██▉ | 14342/48008 [49:01<2:06:33, 4.43it/s] 30%|██▉ | 14343/48008 [49:02<2:00:32, 4.65it/s] 30%|██▉ | 14344/48008 [49:02<1:56:56, 4.80it/s] 30%|██▉ | 14345/48008 [49:02<1:56:13, 4.83it/s] 30%|██▉ | 14346/48008 [49:02<1:53:30, 4.94it/s] 30%|██▉ | 14347/48008 [49:02<1:52:34, 4.98it/s] 30%|██▉ | 14348/48008 [49:03<1:51:02, 5.05it/s] 30%|██▉ | 14349/48008 [49:03<1:52:53, 4.97it/s] 30%|██▉ | 14350/48008 [49:03<1:50:43, 5.07it/s] {'loss': 0.3027, 'grad_norm': 0.59765625, 'learning_rate': 0.0001402224629228462, 'epoch': 0.3} 30%|██▉ | 14350/48008 [49:03<1:50:43, 5.07it/s] 30%|██▉ | 14351/48008 [49:03<1:52:02, 5.01it/s] 30%|██▉ | 14352/48008 [49:03<1:50:08, 5.09it/s] 30%|██▉ | 14353/48008 [49:04<1:51:28, 5.03it/s] 30%|██▉ | 14354/48008 [49:04<1:50:09, 5.09it/s] 30%|██▉ | 14355/48008 [49:04<1:52:02, 5.01it/s] 30%|██▉ | 14356/48008 [49:04<1:50:48, 5.06it/s] 30%|██▉ | 14357/48008 [49:04<1:49:30, 5.12it/s] 30%|██▉ | 14358/48008 [49:05<1:47:57, 5.19it/s] 30%|██▉ | 14359/48008 [49:05<1:47:58, 5.19it/s] 30%|██▉ | 14360/48008 [49:05<1:48:31, 5.17it/s] 30%|██▉ | 14361/48008 [49:05<1:48:09, 5.18it/s] 30%|██▉ | 14362/48008 [49:05<1:47:43, 5.21it/s] 30%|██▉ | 14363/48008 [49:05<1:47:52, 5.20it/s] 30%|██▉ | 14364/48008 [49:06<1:50:03, 5.09it/s] 30%|██▉ | 14365/48008 [49:06<1:49:48, 5.11it/s] 30%|██▉ | 14366/48008 [49:06<1:50:00, 5.10it/s] 30%|██▉ | 14367/48008 [49:06<1:48:37, 5.16it/s] 30%|██▉ | 14368/48008 [49:06<1:50:02, 5.09it/s] 30%|██▉ | 14369/48008 [49:07<1:50:23, 5.08it/s] 30%|██▉ | 14370/48008 [49:07<1:48:37, 5.16it/s] 30%|██▉ | 14371/48008 [49:07<1:47:23, 5.22it/s] 30%|██▉ | 14372/48008 [49:07<1:46:34, 5.26it/s] 30%|██▉ | 14373/48008 [49:07<1:47:44, 5.20it/s] 30%|██▉ | 14374/48008 [49:08<1:47:19, 5.22it/s] 30%|██▉ | 14375/48008 [49:08<1:47:08, 5.23it/s] 30%|██▉ | 14376/48008 [49:08<1:49:28, 5.12it/s] 30%|██▉ | 14377/48008 [49:08<1:48:07, 5.18it/s] 30%|██▉ | 14378/48008 [49:08<1:47:37, 5.21it/s] 30%|██▉ | 14379/48008 [49:09<1:47:02, 5.24it/s] 30%|██▉ | 14380/48008 [49:09<1:47:58, 5.19it/s] 30%|██▉ | 14381/48008 [49:09<1:48:03, 5.19it/s] 30%|██▉ | 14382/48008 [49:09<1:48:41, 5.16it/s] 30%|██▉ | 14383/48008 [49:09<1:49:23, 5.12it/s] 30%|██▉ | 14384/48008 [49:10<1:48:43, 5.15it/s] 30%|██▉ | 14385/48008 [49:10<1:51:11, 5.04it/s] 30%|██▉ | 14386/48008 [49:10<1:50:15, 5.08it/s] 30%|██▉ | 14387/48008 [49:10<1:50:22, 5.08it/s] 30%|██▉ | 14388/48008 [49:10<1:50:36, 5.07it/s] 30%|██▉ | 14389/48008 [49:11<1:50:45, 5.06it/s] 30%|██▉ | 14390/48008 [49:11<1:50:26, 5.07it/s] 30%|██▉ | 14391/48008 [49:11<1:49:24, 5.12it/s] 30%|██▉ | 14392/48008 [49:11<1:49:31, 5.12it/s] 30%|██▉ | 14393/48008 [49:11<1:49:54, 5.10it/s] 30%|██▉ | 14394/48008 [49:12<1:48:58, 5.14it/s] 30%|██▉ | 14395/48008 [49:12<1:49:30, 5.12it/s] 30%|██▉ | 14396/48008 [49:12<1:48:23, 5.17it/s] 30%|██▉ | 14397/48008 [49:12<1:47:21, 5.22it/s] 30%|██▉ | 14398/48008 [49:12<1:47:03, 5.23it/s] 30%|██▉ | 14399/48008 [49:12<1:46:13, 5.27it/s] 30%|██▉ | 14400/48008 [49:13<1:46:26, 5.26it/s] {'loss': 0.3034, 'grad_norm': 0.2734375, 'learning_rate': 0.00014001416430594902, 'epoch': 0.3} 30%|██▉ | 14400/48008 [49:13<1:46:26, 5.26it/s] 30%|██▉ | 14401/48008 [49:13<1:46:51, 5.24it/s] 30%|██▉ | 14402/48008 [49:13<1:46:05, 5.28it/s] 30%|███ | 14403/48008 [49:13<1:47:49, 5.19it/s] 30%|███ | 14404/48008 [49:13<1:48:48, 5.15it/s] 30%|███ | 14405/48008 [49:14<1:49:04, 5.13it/s] 30%|███ | 14406/48008 [49:14<1:47:41, 5.20it/s] 30%|███ | 14407/48008 [49:14<1:48:24, 5.17it/s] 30%|███ | 14408/48008 [49:14<1:48:20, 5.17it/s] 30%|███ | 14409/48008 [49:14<1:47:59, 5.19it/s] 30%|███ | 14410/48008 [49:15<1:49:50, 5.10it/s] 30%|███ | 14411/48008 [49:15<1:48:58, 5.14it/s] 30%|███ | 14412/48008 [49:15<1:48:58, 5.14it/s] 30%|███ | 14413/48008 [49:15<1:47:42, 5.20it/s] 30%|███ | 14414/48008 [49:15<1:47:35, 5.20it/s] 30%|███ | 14415/48008 [49:16<1:47:20, 5.22it/s] 30%|███ | 14416/48008 [49:16<1:46:54, 5.24it/s] 30%|███ | 14417/48008 [49:16<1:46:14, 5.27it/s] 30%|███ | 14418/48008 [49:16<1:46:13, 5.27it/s] 30%|███ | 14419/48008 [49:16<1:46:10, 5.27it/s] 30%|███ | 14420/48008 [49:16<1:46:18, 5.27it/s] 30%|███ | 14421/48008 [49:17<2:19:06, 4.02it/s] 30%|███ | 14422/48008 [49:17<2:09:16, 4.33it/s] 30%|███ | 14423/48008 [49:17<2:01:46, 4.60it/s] 30%|███ | 14424/48008 [49:17<1:57:02, 4.78it/s] 30%|███ | 14425/48008 [49:18<1:54:58, 4.87it/s] 30%|███ | 14426/48008 [49:18<1:50:46, 5.05it/s] 30%|███ | 14427/48008 [49:18<1:57:36, 4.76it/s] 30%|███ | 14428/48008 [49:18<1:53:57, 4.91it/s] 30%|███ | 14429/48008 [49:18<1:54:12, 4.90it/s] 30%|███ | 14430/48008 [49:19<2:24:50, 3.86it/s] 30%|███ | 14431/48008 [49:19<2:15:45, 4.12it/s] 30%|███ | 14432/48008 [49:19<2:06:21, 4.43it/s] 30%|███ | 14433/48008 [49:19<2:01:52, 4.59it/s] 30%|███ | 14434/48008 [49:20<1:57:25, 4.77it/s] 30%|███ | 14435/48008 [49:20<1:53:47, 4.92it/s] 30%|███ | 14436/48008 [49:20<1:51:51, 5.00it/s] 30%|███ | 14437/48008 [49:20<1:49:36, 5.11it/s] 30%|███ | 14438/48008 [49:20<1:49:14, 5.12it/s] 30%|███ | 14439/48008 [49:21<1:49:15, 5.12it/s] 30%|███ | 14440/48008 [49:21<1:49:29, 5.11it/s] 30%|███ | 14441/48008 [49:21<1:49:39, 5.10it/s] 30%|███ | 14442/48008 [49:21<1:49:07, 5.13it/s] 30%|███ | 14443/48008 [49:21<2:07:58, 4.37it/s] 30%|███ | 14444/48008 [49:22<2:02:58, 4.55it/s] 30%|███ | 14445/48008 [49:22<1:57:04, 4.78it/s] 30%|███ | 14446/48008 [49:22<1:53:57, 4.91it/s] 30%|███ | 14447/48008 [49:22<1:51:17, 5.03it/s] 30%|███ | 14448/48008 [49:22<1:50:24, 5.07it/s] 30%|███ | 14449/48008 [49:23<1:49:58, 5.09it/s] 30%|███ | 14450/48008 [49:23<1:49:58, 5.09it/s] {'loss': 0.3038, 'grad_norm': 0.419921875, 'learning_rate': 0.00013980586568905184, 'epoch': 0.3} 30%|███ | 14450/48008 [49:23<1:49:58, 5.09it/s] 30%|███ | 14451/48008 [49:23<1:50:58, 5.04it/s] 30%|███ | 14452/48008 [49:23<1:49:45, 5.10it/s] 30%|███ | 14453/48008 [49:23<1:49:01, 5.13it/s] 30%|███ | 14454/48008 [49:24<1:48:06, 5.17it/s] 30%|███ | 14455/48008 [49:24<1:47:04, 5.22it/s] 30%|███ | 14456/48008 [49:24<1:46:14, 5.26it/s] 30%|███ | 14457/48008 [49:24<1:47:31, 5.20it/s] 30%|███ | 14458/48008 [49:24<1:47:15, 5.21it/s] 30%|███ | 14459/48008 [49:25<1:54:54, 4.87it/s] 30%|███ | 14460/48008 [49:25<2:11:50, 4.24it/s] 30%|███ | 14461/48008 [49:25<2:05:24, 4.46it/s] 30%|███ | 14462/48008 [49:25<1:59:07, 4.69it/s] 30%|███ | 14463/48008 [49:25<1:56:51, 4.78it/s] 30%|███ | 14464/48008 [49:26<2:01:44, 4.59it/s] 30%|███ | 14465/48008 [49:26<1:58:08, 4.73it/s] 30%|███ | 14466/48008 [49:26<1:54:54, 4.87it/s] 30%|███ | 14467/48008 [49:26<1:53:49, 4.91it/s] 30%|███ | 14468/48008 [49:27<1:54:46, 4.87it/s] 30%|███ | 14469/48008 [49:27<1:52:48, 4.96it/s] 30%|███ | 14470/48008 [49:27<1:52:37, 4.96it/s] 30%|███ | 14471/48008 [49:27<1:50:43, 5.05it/s] 30%|███ | 14472/48008 [49:27<1:49:19, 5.11it/s] 30%|███ | 14473/48008 [49:27<1:48:38, 5.14it/s] 30%|███ | 14474/48008 [49:28<1:50:16, 5.07it/s] 30%|███ | 14475/48008 [49:28<1:49:42, 5.09it/s] 30%|███ | 14476/48008 [49:28<1:49:06, 5.12it/s] 30%|███ | 14477/48008 [49:28<1:50:22, 5.06it/s] 30%|███ | 14478/48008 [49:28<1:49:22, 5.11it/s] 30%|███ | 14479/48008 [49:29<1:48:10, 5.17it/s] 30%|███ | 14480/48008 [49:29<1:46:47, 5.23it/s] 30%|███ | 14481/48008 [49:29<1:54:41, 4.87it/s] 30%|███ | 14482/48008 [49:29<2:00:38, 4.63it/s] 30%|███ | 14483/48008 [49:30<1:56:11, 4.81it/s] 30%|███ | 14484/48008 [49:30<1:53:33, 4.92it/s] 30%|███ | 14485/48008 [49:30<1:50:52, 5.04it/s] 30%|███ | 14486/48008 [49:30<1:48:56, 5.13it/s] 30%|███ | 14487/48008 [49:30<2:07:40, 4.38it/s] 30%|███ | 14488/48008 [49:31<2:03:34, 4.52it/s] 30%|███ | 14489/48008 [49:31<2:06:33, 4.41it/s] 30%|███ | 14490/48008 [49:31<2:20:10, 3.99it/s] 30%|███ | 14491/48008 [49:31<2:10:24, 4.28it/s] 30%|███ | 14492/48008 [49:32<2:03:12, 4.53it/s] 30%|███ | 14493/48008 [49:32<1:58:16, 4.72it/s] 30%|███ | 14494/48008 [49:32<1:54:49, 4.86it/s] 30%|███ | 14495/48008 [49:32<2:00:20, 4.64it/s] 30%|███ | 14496/48008 [49:32<1:56:14, 4.81it/s] 30%|███ | 14497/48008 [49:33<1:53:40, 4.91it/s] 30%|███ | 14498/48008 [49:33<1:52:45, 4.95it/s] 30%|███ | 14499/48008 [49:33<1:50:28, 5.06it/s] 30%|███ | 14500/48008 [49:33<1:49:23, 5.11it/s] {'loss': 0.3014, 'grad_norm': 0.1875, 'learning_rate': 0.00013959756707215465, 'epoch': 0.3} 30%|███ | 14500/48008 [49:33<1:49:23, 5.11it/s] 30%|███ | 14501/48008 [49:33<1:48:37, 5.14it/s] 30%|███ | 14502/48008 [49:33<1:50:05, 5.07it/s] 30%|███ | 14503/48008 [49:34<1:48:34, 5.14it/s] 30%|███ | 14504/48008 [49:34<1:48:05, 5.17it/s] 30%|███ | 14505/48008 [49:34<1:50:39, 5.05it/s] 30%|███ | 14506/48008 [49:34<1:52:24, 4.97it/s] 30%|███ | 14507/48008 [49:34<1:50:50, 5.04it/s] 30%|███ | 14508/48008 [49:35<1:49:45, 5.09it/s] 30%|███ | 14509/48008 [49:35<1:48:07, 5.16it/s] 30%|███ | 14510/48008 [49:35<1:46:12, 5.26it/s] 30%|███ | 14511/48008 [49:35<1:46:32, 5.24it/s] 30%|███ | 14512/48008 [49:35<1:45:53, 5.27it/s] 30%|███ | 14513/48008 [49:36<1:46:48, 5.23it/s] 30%|███ | 14514/48008 [49:36<2:19:20, 4.01it/s] 30%|███ | 14515/48008 [49:36<2:10:25, 4.28it/s] 30%|███ | 14516/48008 [49:36<2:03:16, 4.53it/s] 30%|███ | 14517/48008 [49:37<1:58:03, 4.73it/s] 30%|███ | 14518/48008 [49:37<2:02:21, 4.56it/s] 30%|███ | 14519/48008 [49:37<1:58:29, 4.71it/s] 30%|███ | 14520/48008 [49:37<1:57:19, 4.76it/s] 30%|███ | 14521/48008 [49:37<1:54:43, 4.86it/s] 30%|███ | 14522/48008 [49:38<1:51:48, 4.99it/s] 30%|███ | 14523/48008 [49:38<1:49:55, 5.08it/s] 30%|███ | 14524/48008 [49:38<1:48:29, 5.14it/s] 30%|███ | 14525/48008 [49:38<1:47:26, 5.19it/s] 30%|███ | 14526/48008 [49:38<1:47:33, 5.19it/s] 30%|███ | 14527/48008 [49:39<1:47:37, 5.18it/s] 30%|███ | 14528/48008 [49:39<1:47:02, 5.21it/s] 30%|███ | 14529/48008 [49:39<1:54:31, 4.87it/s] 30%|███ | 14530/48008 [49:39<1:52:07, 4.98it/s] 30%|███ | 14531/48008 [49:39<1:50:34, 5.05it/s] 30%|███ | 14532/48008 [49:40<2:22:01, 3.93it/s] 30%|███ | 14533/48008 [49:40<2:11:48, 4.23it/s] 30%|███ | 14534/48008 [49:40<2:05:03, 4.46it/s] 30%|███ | 14535/48008 [49:40<2:01:37, 4.59it/s] 30%|███ | 14536/48008 [49:41<1:56:23, 4.79it/s] 30%|███ | 14537/48008 [49:41<1:54:14, 4.88it/s] 30%|███ | 14538/48008 [49:41<1:53:59, 4.89it/s] 30%|███ | 14539/48008 [49:41<1:51:43, 4.99it/s] 30%|███ | 14540/48008 [49:41<1:51:32, 5.00it/s] 30%|███ | 14541/48008 [49:42<1:50:41, 5.04it/s] 30%|███ | 14542/48008 [49:42<1:49:27, 5.10it/s] 30%|███ | 14543/48008 [49:42<1:48:13, 5.15it/s] 30%|███ | 14544/48008 [49:42<1:46:53, 5.22it/s] 30%|███ | 14545/48008 [49:42<1:48:44, 5.13it/s] 30%|███ | 14546/48008 [49:42<1:49:12, 5.11it/s] 30%|███ | 14547/48008 [49:43<1:47:52, 5.17it/s] 30%|███ | 14548/48008 [49:43<1:47:14, 5.20it/s] 30%|███ | 14549/48008 [49:43<1:46:46, 5.22it/s] 30%|███ | 14550/48008 [49:43<1:46:13, 5.25it/s] {'loss': 0.302, 'grad_norm': 0.66015625, 'learning_rate': 0.00013938926845525747, 'epoch': 0.3} 30%|███ | 14550/48008 [49:43<1:46:13, 5.25it/s] 30%|███ | 14551/48008 [49:43<1:49:20, 5.10it/s] 30%|███ | 14552/48008 [49:44<1:49:08, 5.11it/s] 30%|███ | 14553/48008 [49:44<1:48:47, 5.12it/s] 30%|███ | 14554/48008 [49:44<1:49:53, 5.07it/s] 30%|███ | 14555/48008 [49:44<1:48:38, 5.13it/s] 30%|███ | 14556/48008 [49:44<1:47:42, 5.18it/s] 30%|███ | 14557/48008 [49:45<1:47:34, 5.18it/s] 30%|███ | 14558/48008 [49:45<1:47:27, 5.19it/s] 30%|███ | 14559/48008 [49:45<1:48:21, 5.14it/s] 30%|███ | 14560/48008 [49:45<1:47:38, 5.18it/s] 30%|███ | 14561/48008 [49:45<1:46:53, 5.21it/s] 30%|███ | 14562/48008 [49:46<1:48:52, 5.12it/s] 30%|███ | 14563/48008 [49:46<1:50:47, 5.03it/s] 30%|███ | 14564/48008 [49:46<2:08:46, 4.33it/s] 30%|███ | 14565/48008 [49:46<2:01:33, 4.59it/s] 30%|███ | 14566/48008 [49:46<1:58:39, 4.70it/s] 30%|███ | 14567/48008 [49:47<1:54:56, 4.85it/s] 30%|███ | 14568/48008 [49:47<2:11:43, 4.23it/s] 30%|███ | 14569/48008 [49:47<2:05:20, 4.45it/s] 30%|███ | 14570/48008 [49:47<1:59:12, 4.67it/s] 30%|███ | 14571/48008 [49:48<1:55:33, 4.82it/s] 30%|███ | 14572/48008 [49:48<2:25:49, 3.82it/s] 30%|███ | 14573/48008 [49:48<2:14:13, 4.15it/s] 30%|███ | 14574/48008 [49:48<2:05:58, 4.42it/s] 30%|███ | 14575/48008 [49:49<2:03:00, 4.53it/s] 30%|███ | 14576/48008 [49:49<1:59:48, 4.65it/s] 30%|███ | 14577/48008 [49:49<1:55:28, 4.83it/s] 30%|███ | 14578/48008 [49:49<1:54:44, 4.86it/s] 30%|███ | 14579/48008 [49:49<1:52:14, 4.96it/s] 30%|███ | 14580/48008 [49:50<1:49:48, 5.07it/s] 30%|███ | 14581/48008 [49:50<1:49:40, 5.08it/s] 30%|███ | 14582/48008 [49:50<1:48:45, 5.12it/s] 30%|███ | 14583/48008 [49:50<1:48:28, 5.14it/s] 30%|███ | 14584/48008 [49:50<1:48:00, 5.16it/s] 30%|███ | 14585/48008 [49:50<1:47:19, 5.19it/s] 30%|███ | 14586/48008 [49:51<1:46:49, 5.21it/s] 30%|███ | 14587/48008 [49:51<1:48:59, 5.11it/s] 30%|███ | 14588/48008 [49:51<1:47:59, 5.16it/s] 30%|███ | 14589/48008 [49:51<1:47:06, 5.20it/s] 30%|███ | 14590/48008 [49:51<1:47:31, 5.18it/s] 30%|███ | 14591/48008 [49:52<2:19:48, 3.98it/s] 30%|███ | 14592/48008 [49:52<2:12:07, 4.22it/s] 30%|███ | 14593/48008 [49:52<2:06:26, 4.40it/s] 30%|███ | 14594/48008 [49:52<2:01:50, 4.57it/s] 30%|███ | 14595/48008 [49:53<2:00:57, 4.60it/s] 30%|███ | 14596/48008 [49:53<1:55:57, 4.80it/s] 30%|███ | 14597/48008 [49:53<1:55:16, 4.83it/s] 30%|███ | 14598/48008 [49:53<1:52:03, 4.97it/s] 30%|███ | 14599/48008 [49:53<1:49:43, 5.07it/s] 30%|███ | 14600/48008 [49:54<1:48:38, 5.13it/s] {'loss': 0.3006, 'grad_norm': 0.2890625, 'learning_rate': 0.0001391809698383603, 'epoch': 0.3} 30%|███ | 14600/48008 [49:54<1:48:38, 5.13it/s] 30%|███ | 14601/48008 [49:54<2:07:34, 4.36it/s] 30%|███ | 14602/48008 [49:54<2:00:17, 4.63it/s] 30%|███ | 14603/48008 [49:54<1:55:48, 4.81it/s] 30%|███ | 14604/48008 [49:54<1:53:25, 4.91it/s] 30%|███ | 14605/48008 [49:55<1:53:44, 4.89it/s] 30%|███ | 14606/48008 [49:55<1:51:29, 4.99it/s] 30%|███ | 14607/48008 [49:55<1:49:56, 5.06it/s] 30%|███ | 14608/48008 [49:55<1:49:06, 5.10it/s] 30%|███ | 14609/48008 [49:55<1:48:55, 5.11it/s] 30%|███ | 14610/48008 [49:56<1:47:48, 5.16it/s] 30%|███ | 14611/48008 [49:56<1:48:35, 5.13it/s] 30%|███ | 14612/48008 [49:56<1:49:17, 5.09it/s] 30%|███ | 14613/48008 [49:56<1:48:35, 5.13it/s] 30%|███ | 14614/48008 [49:56<1:49:00, 5.11it/s] 30%|███ | 14615/48008 [49:57<2:20:55, 3.95it/s] 30%|███ | 14616/48008 [49:57<2:12:40, 4.19it/s] 30%|███ | 14617/48008 [49:57<2:03:51, 4.49it/s] 30%|███ | 14618/48008 [49:57<1:57:36, 4.73it/s] 30%|███ | 14619/48008 [49:58<1:54:45, 4.85it/s] 30%|███ | 14620/48008 [49:58<1:53:22, 4.91it/s] 30%|███ | 14621/48008 [49:58<1:50:51, 5.02it/s] 30%|███ | 14622/48008 [49:58<1:49:33, 5.08it/s] 30%|███ | 14623/48008 [49:58<1:51:39, 4.98it/s] 30%|███ | 14624/48008 [49:59<1:52:48, 4.93it/s] 30%|███ | 14625/48008 [49:59<1:51:47, 4.98it/s] 30%|███ | 14626/48008 [49:59<1:49:49, 5.07it/s] 30%|███ | 14627/48008 [49:59<1:48:48, 5.11it/s] 30%|███ | 14628/48008 [49:59<1:49:39, 5.07it/s] 30%|███ | 14629/48008 [50:00<1:49:04, 5.10it/s] 30%|███ | 14630/48008 [50:00<1:47:37, 5.17it/s] 30%|███ | 14631/48008 [50:00<1:49:08, 5.10it/s] 30%|███ | 14632/48008 [50:00<1:47:35, 5.17it/s] 30%|███ | 14633/48008 [50:00<1:50:23, 5.04it/s] 30%|███ | 14634/48008 [50:01<1:49:56, 5.06it/s] 30%|███ | 14635/48008 [50:01<1:50:02, 5.05it/s] 30%|███ | 14636/48008 [50:01<1:48:12, 5.14it/s] 30%|███ | 14637/48008 [50:01<1:47:32, 5.17it/s] 30%|███ | 14638/48008 [50:01<1:47:27, 5.18it/s] 30%|███ | 14639/48008 [50:02<1:49:33, 5.08it/s] 30%|███ | 14640/48008 [50:02<1:48:05, 5.14it/s] 30%|███ | 14641/48008 [50:02<1:47:43, 5.16it/s] 30%|███ | 14642/48008 [50:02<1:47:00, 5.20it/s] 31%|███ | 14643/48008 [50:02<1:48:40, 5.12it/s] 31%|███ | 14644/48008 [50:02<1:48:55, 5.10it/s] 31%|███ | 14645/48008 [50:03<1:49:04, 5.10it/s] 31%|███ | 14646/48008 [50:03<1:48:01, 5.15it/s] 31%|███ | 14647/48008 [50:03<2:06:57, 4.38it/s] 31%|███ | 14648/48008 [50:03<2:00:47, 4.60it/s] 31%|███ | 14649/48008 [50:04<2:15:20, 4.11it/s] 31%|███ | 14650/48008 [50:04<2:06:03, 4.41it/s] {'loss': 0.3021, 'grad_norm': 0.75390625, 'learning_rate': 0.00013897267122146308, 'epoch': 0.31} 31%|███ | 14650/48008 [50:04<2:06:03, 4.41it/s] 31%|███ | 14651/48008 [50:04<2:02:05, 4.55it/s] 31%|███ | 14652/48008 [50:04<1:57:18, 4.74it/s] 31%|███ | 14653/48008 [50:04<1:54:37, 4.85it/s] 31%|███ | 14654/48008 [50:05<1:54:11, 4.87it/s] 31%|███ | 14655/48008 [50:05<1:51:49, 4.97it/s] 31%|███ | 14656/48008 [50:05<1:49:44, 5.07it/s] 31%|███ | 14657/48008 [50:05<1:48:19, 5.13it/s] 31%|███ | 14658/48008 [50:05<1:47:37, 5.16it/s] 31%|███ | 14659/48008 [50:06<1:47:15, 5.18it/s] 31%|███ | 14660/48008 [50:06<1:46:54, 5.20it/s] 31%|███ | 14661/48008 [50:06<1:46:33, 5.22it/s] 31%|███ | 14662/48008 [50:06<1:46:59, 5.19it/s] 31%|███ | 14663/48008 [50:06<1:47:57, 5.15it/s] 31%|███ | 14664/48008 [50:07<1:47:27, 5.17it/s] 31%|███ | 14665/48008 [50:07<1:48:57, 5.10it/s] 31%|███ | 14666/48008 [50:07<1:48:02, 5.14it/s] 31%|███ | 14667/48008 [50:07<1:48:15, 5.13it/s] 31%|███ | 14668/48008 [50:07<1:55:32, 4.81it/s] 31%|███ | 14669/48008 [50:08<1:52:12, 4.95it/s] 31%|███ | 14670/48008 [50:08<1:50:30, 5.03it/s] 31%|███ | 14671/48008 [50:08<1:51:57, 4.96it/s] 31%|███ | 14672/48008 [50:08<1:50:07, 5.05it/s] 31%|███ | 14673/48008 [50:08<1:49:34, 5.07it/s] 31%|███ | 14674/48008 [50:09<1:48:35, 5.12it/s] 31%|███ | 14675/48008 [50:09<1:47:45, 5.16it/s] 31%|███ | 14676/48008 [50:09<1:47:20, 5.18it/s] 31%|███ | 14677/48008 [50:09<1:46:23, 5.22it/s] 31%|███ | 14678/48008 [50:09<1:45:44, 5.25it/s] 31%|███ | 14679/48008 [50:10<1:45:09, 5.28it/s] 31%|███ | 14680/48008 [50:10<1:45:02, 5.29it/s] 31%|███ | 14681/48008 [50:10<2:17:40, 4.03it/s] 31%|███ | 14682/48008 [50:10<2:11:19, 4.23it/s] 31%|███ | 14683/48008 [50:10<2:03:55, 4.48it/s] 31%|███ | 14684/48008 [50:11<2:30:53, 3.68it/s] 31%|███ | 14685/48008 [50:11<2:17:26, 4.04it/s] 31%|███ | 14686/48008 [50:11<2:09:09, 4.30it/s] 31%|███ | 14687/48008 [50:11<2:04:08, 4.47it/s] 31%|███ | 14688/48008 [50:12<1:58:05, 4.70it/s] 31%|███ | 14689/48008 [50:12<1:53:56, 4.87it/s] 31%|███ | 14690/48008 [50:12<1:51:11, 4.99it/s] 31%|███ | 14691/48008 [50:12<1:57:38, 4.72it/s] 31%|███ | 14692/48008 [50:12<2:01:53, 4.56it/s] 31%|███ | 14693/48008 [50:13<1:57:08, 4.74it/s] 31%|███ | 14694/48008 [50:13<1:54:45, 4.84it/s] 31%|███ | 14695/48008 [50:13<1:53:55, 4.87it/s] 31%|███ | 14696/48008 [50:13<1:53:24, 4.90it/s] 31%|███ | 14697/48008 [50:13<1:51:58, 4.96it/s] 31%|███ | 14698/48008 [50:14<1:53:27, 4.89it/s] 31%|███ | 14699/48008 [50:14<1:51:13, 4.99it/s] 31%|███ | 14700/48008 [50:14<1:52:28, 4.94it/s] {'loss': 0.3004, 'grad_norm': 0.1591796875, 'learning_rate': 0.00013876437260456592, 'epoch': 0.31} 31%|███ | 14700/48008 [50:14<1:52:28, 4.94it/s] 31%|███ | 14701/48008 [50:14<1:53:54, 4.87it/s] 31%|███ | 14702/48008 [50:15<1:53:19, 4.90it/s] 31%|███ | 14703/48008 [50:15<1:51:25, 4.98it/s] 31%|███ | 14704/48008 [50:15<1:49:23, 5.07it/s] 31%|███ | 14705/48008 [50:15<1:48:29, 5.12it/s] 31%|███ | 14706/48008 [50:15<1:48:01, 5.14it/s] 31%|███ | 14707/48008 [50:15<1:46:58, 5.19it/s] 31%|███ | 14708/48008 [50:16<1:49:46, 5.06it/s] 31%|███ | 14709/48008 [50:16<1:48:29, 5.12it/s] 31%|███ | 14710/48008 [50:16<1:47:16, 5.17it/s] 31%|███ | 14711/48008 [50:16<1:47:03, 5.18it/s] 31%|███ | 14712/48008 [50:16<1:46:55, 5.19it/s] 31%|███ | 14713/48008 [50:17<1:46:39, 5.20it/s] 31%|███ | 14714/48008 [50:17<1:47:04, 5.18it/s] 31%|███ | 14715/48008 [50:17<1:46:59, 5.19it/s] 31%|███ | 14716/48008 [50:17<1:50:00, 5.04it/s] 31%|███ | 14717/48008 [50:17<1:48:05, 5.13it/s] 31%|███ | 14718/48008 [50:18<1:47:33, 5.16it/s] 31%|███ | 14719/48008 [50:18<1:46:03, 5.23it/s] 31%|███ | 14720/48008 [50:18<1:45:40, 5.25it/s] 31%|███ | 14721/48008 [50:18<1:46:41, 5.20it/s] 31%|███ | 14722/48008 [50:18<1:46:16, 5.22it/s] 31%|███ | 14723/48008 [50:19<1:48:00, 5.14it/s] 31%|███ | 14724/48008 [50:19<1:47:35, 5.16it/s] 31%|███ | 14725/48008 [50:19<1:54:41, 4.84it/s] 31%|███ | 14726/48008 [50:19<1:54:00, 4.87it/s] 31%|███ | 14727/48008 [50:19<1:51:08, 4.99it/s] 31%|███ | 14728/48008 [50:20<1:50:38, 5.01it/s] 31%|███ | 14729/48008 [50:20<1:51:32, 4.97it/s] 31%|███ | 14730/48008 [50:20<1:51:04, 4.99it/s] 31%|███ | 14731/48008 [50:20<1:49:37, 5.06it/s] 31%|███ | 14732/48008 [50:20<1:50:29, 5.02it/s] 31%|███ | 14733/48008 [50:21<1:52:08, 4.95it/s] 31%|███ | 14734/48008 [50:21<1:51:14, 4.99it/s] 31%|███ | 14735/48008 [50:21<1:48:53, 5.09it/s] 31%|███ | 14736/48008 [50:21<2:07:03, 4.36it/s] 31%|███ | 14737/48008 [50:21<2:00:45, 4.59it/s] 31%|███ | 14738/48008 [50:22<3:52:46, 2.38it/s] 31%|███ | 14739/48008 [50:23<3:22:18, 2.74it/s] 31%|███ | 14740/48008 [50:23<2:52:22, 3.22it/s] 31%|███ | 14741/48008 [50:23<2:34:19, 3.59it/s] 31%|███ | 14742/48008 [50:23<2:20:18, 3.95it/s] 31%|███ | 14743/48008 [50:23<2:11:58, 4.20it/s] 31%|███ | 14744/48008 [50:24<2:03:47, 4.48it/s] 31%|███ | 14745/48008 [50:24<1:58:07, 4.69it/s] 31%|███ | 14746/48008 [50:24<1:56:39, 4.75it/s] 31%|███ | 14747/48008 [50:24<1:53:27, 4.89it/s] 31%|███ | 14748/48008 [50:24<1:51:19, 4.98it/s] 31%|███ | 14749/48008 [50:25<1:51:40, 4.96it/s] 31%|███ | 14750/48008 [50:25<1:49:24, 5.07it/s] {'loss': 0.3005, 'grad_norm': 0.4453125, 'learning_rate': 0.0001385560739876687, 'epoch': 0.31} 31%|███ | 14750/48008 [50:25<1:49:24, 5.07it/s] 31%|███ | 14751/48008 [50:25<1:47:48, 5.14it/s] 31%|███ | 14752/48008 [50:25<1:48:25, 5.11it/s] 31%|███ | 14753/48008 [50:25<1:47:50, 5.14it/s] 31%|███ | 14754/48008 [50:26<1:49:37, 5.06it/s] 31%|███ | 14755/48008 [50:26<1:47:46, 5.14it/s] 31%|███ | 14756/48008 [50:26<1:48:08, 5.12it/s] 31%|███ | 14757/48008 [50:26<1:46:25, 5.21it/s] 31%|███ | 14758/48008 [50:26<1:45:50, 5.24it/s] 31%|███ | 14759/48008 [50:26<1:46:03, 5.22it/s] 31%|███ | 14760/48008 [50:27<1:46:07, 5.22it/s] 31%|███ | 14761/48008 [50:27<1:45:54, 5.23it/s] 31%|███ | 14762/48008 [50:27<1:46:37, 5.20it/s] 31%|███ | 14763/48008 [50:27<1:48:09, 5.12it/s] 31%|███ | 14764/48008 [50:27<1:48:21, 5.11it/s] 31%|███ | 14765/48008 [50:28<1:47:48, 5.14it/s] 31%|███ | 14766/48008 [50:28<1:46:42, 5.19it/s] 31%|███ | 14767/48008 [50:28<1:45:50, 5.23it/s] 31%|███ | 14768/48008 [50:28<1:45:53, 5.23it/s] 31%|███ | 14769/48008 [50:29<2:05:04, 4.43it/s] 31%|███ | 14770/48008 [50:29<1:58:56, 4.66it/s] 31%|███ | 14771/48008 [50:29<1:54:54, 4.82it/s] 31%|███ | 14772/48008 [50:29<1:53:59, 4.86it/s] 31%|███ | 14773/48008 [50:29<1:51:13, 4.98it/s] 31%|███ | 14774/48008 [50:29<1:51:19, 4.98it/s] 31%|███ | 14775/48008 [50:30<1:49:47, 5.04it/s] 31%|███ | 14776/48008 [50:30<1:50:15, 5.02it/s] 31%|███ | 14777/48008 [50:30<1:50:46, 5.00it/s] 31%|███ | 14778/48008 [50:30<1:50:46, 5.00it/s] 31%|███ | 14779/48008 [50:31<1:56:46, 4.74it/s] 31%|███ | 14780/48008 [50:31<1:56:17, 4.76it/s] 31%|███ | 14781/48008 [50:31<1:54:50, 4.82it/s] 31%|███ | 14782/48008 [50:31<1:52:09, 4.94it/s] 31%|███ | 14783/48008 [50:31<1:49:32, 5.05it/s] 31%|███ | 14784/48008 [50:32<1:51:14, 4.98it/s] 31%|███ | 14785/48008 [50:32<1:48:40, 5.09it/s] 31%|███ | 14786/48008 [50:32<2:07:00, 4.36it/s] 31%|███ | 14787/48008 [50:32<2:01:52, 4.54it/s] 31%|███ | 14788/48008 [50:32<1:57:08, 4.73it/s] 31%|███ | 14789/48008 [50:33<1:54:31, 4.83it/s] 31%|███ | 14790/48008 [50:33<1:51:42, 4.96it/s] 31%|███ | 14791/48008 [50:33<1:49:02, 5.08it/s] 31%|███ | 14792/48008 [50:33<1:48:17, 5.11it/s] 31%|███ | 14793/48008 [50:33<1:47:38, 5.14it/s] 31%|███ | 14794/48008 [50:34<1:48:34, 5.10it/s] 31%|███ | 14795/48008 [50:34<1:48:50, 5.09it/s] 31%|███ | 14796/48008 [50:34<1:47:58, 5.13it/s] 31%|███ | 14797/48008 [50:34<1:47:27, 5.15it/s] 31%|███ | 14798/48008 [50:34<1:48:52, 5.08it/s] 31%|███ | 14799/48008 [50:35<1:55:30, 4.79it/s] 31%|███ | 14800/48008 [50:35<1:51:54, 4.95it/s] {'loss': 0.3014, 'grad_norm': 0.56640625, 'learning_rate': 0.00013834777537077156, 'epoch': 0.31} 31%|███ | 14800/48008 [50:35<1:51:54, 4.95it/s] 31%|███ | 14801/48008 [50:35<1:50:02, 5.03it/s] 31%|███ | 14802/48008 [50:35<1:50:44, 5.00it/s] 31%|███ | 14803/48008 [50:35<1:49:17, 5.06it/s] 31%|███ | 14804/48008 [50:36<1:48:05, 5.12it/s] 31%|███ | 14805/48008 [50:36<1:47:18, 5.16it/s] 31%|███ | 14806/48008 [50:36<1:47:53, 5.13it/s] 31%|███ | 14807/48008 [50:36<1:47:39, 5.14it/s] 31%|███ | 14808/48008 [50:36<1:47:27, 5.15it/s] 31%|███ | 14809/48008 [50:36<1:47:07, 5.17it/s] 31%|███ | 14810/48008 [50:37<1:46:24, 5.20it/s] 31%|███ | 14811/48008 [50:37<1:49:13, 5.07it/s] 31%|███ | 14812/48008 [50:37<1:47:34, 5.14it/s] 31%|███ | 14813/48008 [50:37<1:47:15, 5.16it/s] 31%|███ | 14814/48008 [50:37<1:46:30, 5.19it/s] 31%|███ | 14815/48008 [50:38<1:46:37, 5.19it/s] 31%|███ | 14816/48008 [50:38<1:45:42, 5.23it/s] 31%|███ | 14817/48008 [50:38<1:47:20, 5.15it/s] 31%|███ | 14818/48008 [50:38<1:47:00, 5.17it/s] 31%|███ | 14819/48008 [50:38<1:47:00, 5.17it/s] 31%|███ | 14820/48008 [50:39<1:46:20, 5.20it/s] 31%|███ | 14821/48008 [50:39<1:48:05, 5.12it/s] 31%|███ | 14822/48008 [50:39<2:06:32, 4.37it/s] 31%|███ | 14823/48008 [50:39<2:02:51, 4.50it/s] 31%|███ | 14824/48008 [50:40<1:58:12, 4.68it/s] 31%|███ | 14825/48008 [50:40<1:54:22, 4.84it/s] 31%|███ | 14826/48008 [50:40<2:24:06, 3.84it/s] 31%|███ | 14827/48008 [50:40<2:13:35, 4.14it/s] 31%|███ | 14828/48008 [50:41<2:07:15, 4.35it/s] 31%|███ | 14829/48008 [50:41<2:00:14, 4.60it/s] 31%|███ | 14830/48008 [50:41<1:57:56, 4.69it/s] 31%|███ | 14831/48008 [50:41<1:53:45, 4.86it/s] 31%|███ | 14832/48008 [50:41<1:53:15, 4.88it/s] 31%|███ | 14833/48008 [50:41<1:51:16, 4.97it/s] 31%|███ | 14834/48008 [50:42<1:49:28, 5.05it/s] 31%|███ | 14835/48008 [50:42<1:48:33, 5.09it/s] 31%|███ | 14836/48008 [50:42<1:47:46, 5.13it/s] 31%|███ | 14837/48008 [50:42<1:48:07, 5.11it/s] 31%|███ | 14838/48008 [50:42<1:46:35, 5.19it/s] 31%|███ | 14839/48008 [50:43<1:47:27, 5.14it/s] 31%|███ | 14840/48008 [50:43<1:47:10, 5.16it/s] 31%|███ | 14841/48008 [50:43<1:46:13, 5.20it/s] 31%|███ | 14842/48008 [50:43<1:46:55, 5.17it/s] 31%|███ | 14843/48008 [50:43<1:49:20, 5.06it/s] 31%|███ | 14844/48008 [50:44<1:47:40, 5.13it/s] 31%|███ | 14845/48008 [50:44<1:46:59, 5.17it/s] 31%|███ | 14846/48008 [50:44<1:47:35, 5.14it/s] 31%|███ | 14847/48008 [50:44<1:49:05, 5.07it/s] 31%|███ | 14848/48008 [50:44<1:47:44, 5.13it/s] 31%|███ | 14849/48008 [50:45<1:46:59, 5.17it/s] 31%|███ | 14850/48008 [50:45<1:46:24, 5.19it/s] {'loss': 0.3036, 'grad_norm': 0.1484375, 'learning_rate': 0.00013813947675387435, 'epoch': 0.31} 31%|███ | 14850/48008 [50:45<1:46:24, 5.19it/s] 31%|███ | 14851/48008 [50:45<1:53:52, 4.85it/s] 31%|███ | 14852/48008 [50:45<1:51:43, 4.95it/s] 31%|███ | 14853/48008 [50:46<2:22:30, 3.88it/s] 31%|███ | 14854/48008 [50:46<2:11:22, 4.21it/s] 31%|███ | 14855/48008 [50:46<2:03:01, 4.49it/s] 31%|███ | 14856/48008 [50:46<1:58:25, 4.67it/s] 31%|███ | 14857/48008 [50:46<1:56:29, 4.74it/s] 31%|███ | 14858/48008 [50:47<1:52:22, 4.92it/s] 31%|███ | 14859/48008 [50:47<1:52:15, 4.92it/s] 31%|███ | 14860/48008 [50:47<1:52:45, 4.90it/s] 31%|███ | 14861/48008 [50:47<2:22:39, 3.87it/s] 31%|███ | 14862/48008 [50:48<2:11:15, 4.21it/s] 31%|███ | 14863/48008 [50:48<2:03:21, 4.48it/s] 31%|███ | 14864/48008 [50:48<1:57:20, 4.71it/s] 31%|███ | 14865/48008 [50:48<1:55:39, 4.78it/s] 31%|███ | 14866/48008 [50:48<1:50:59, 4.98it/s] 31%|███ | 14867/48008 [50:48<1:49:41, 5.04it/s] 31%|███ | 14868/48008 [50:49<1:49:44, 5.03it/s] 31%|███ | 14869/48008 [50:49<1:47:57, 5.12it/s] 31%|███ | 14870/48008 [50:49<1:46:44, 5.17it/s] 31%|███ | 14871/48008 [50:49<1:48:41, 5.08it/s] 31%|███ | 14872/48008 [50:49<1:48:46, 5.08it/s] 31%|███ | 14873/48008 [50:50<1:47:54, 5.12it/s] 31%|███ | 14874/48008 [50:50<1:48:55, 5.07it/s] 31%|███ | 14875/48008 [50:50<1:49:42, 5.03it/s] 31%|███ | 14876/48008 [50:50<1:49:04, 5.06it/s] 31%|███ | 14877/48008 [50:50<1:49:50, 5.03it/s] 31%|███ | 14878/48008 [50:51<1:48:26, 5.09it/s] 31%|███ | 14879/48008 [50:51<1:47:36, 5.13it/s] 31%|███ | 14880/48008 [50:51<1:50:24, 5.00it/s] 31%|███ | 14881/48008 [50:51<1:51:08, 4.97it/s] 31%|███ | 14882/48008 [50:51<1:51:23, 4.96it/s] 31%|███ | 14883/48008 [50:52<1:50:46, 4.98it/s] 31%|███ | 14884/48008 [50:52<1:49:00, 5.06it/s] 31%|███ | 14885/48008 [50:52<1:48:55, 5.07it/s] 31%|███ | 14886/48008 [50:52<2:06:45, 4.35it/s] 31%|███ | 14887/48008 [50:53<2:00:37, 4.58it/s] 31%|███ | 14888/48008 [50:53<1:56:56, 4.72it/s] 31%|███ | 14889/48008 [50:53<1:53:07, 4.88it/s] 31%|███ | 14890/48008 [50:53<1:51:51, 4.93it/s] 31%|███ | 14891/48008 [50:53<1:52:38, 4.90it/s] 31%|███ | 14892/48008 [50:54<4:05:45, 2.25it/s] 31%|███ | 14893/48008 [50:55<3:23:31, 2.71it/s] 31%|███ | 14894/48008 [50:55<2:53:50, 3.17it/s] 31%|███ | 14895/48008 [50:55<2:52:19, 3.20it/s] 31%|███ | 14896/48008 [50:55<2:40:22, 3.44it/s] 31%|███ | 14897/48008 [50:55<2:23:31, 3.84it/s] 31%|███ | 14898/48008 [50:56<4:06:05, 2.24it/s] 31%|███ | 14899/48008 [50:57<3:23:05, 2.72it/s] 31%|███ | 14900/48008 [50:57<2:54:08, 3.17it/s] {'loss': 0.3012, 'grad_norm': 0.287109375, 'learning_rate': 0.0001379311781369772, 'epoch': 0.31} 31%|███ | 14900/48008 [50:57<2:54:08, 3.17it/s] 31%|███ | 14901/48008 [50:57<2:35:36, 3.55it/s] 31%|███ | 14902/48008 [50:57<2:23:35, 3.84it/s] 31%|███ | 14903/48008 [50:57<2:11:10, 4.21it/s] 31%|███ | 14904/48008 [50:59<4:57:25, 1.86it/s] 31%|███ | 14905/48008 [50:59<4:19:04, 2.13it/s] 31%|███ | 14906/48008 [50:59<3:32:48, 2.59it/s] 31%|███ | 14907/48008 [50:59<3:01:24, 3.04it/s] 31%|███ | 14908/48008 [50:59<2:39:00, 3.47it/s] 31%|███ | 14909/48008 [51:00<2:22:47, 3.86it/s] 31%|███ | 14910/48008 [51:00<3:55:27, 2.34it/s] 31%|███ | 14911/48008 [51:01<3:18:24, 2.78it/s] 31%|███ | 14912/48008 [51:01<2:50:51, 3.23it/s] 31%|███ | 14913/48008 [51:01<2:31:36, 3.64it/s] 31%|███ | 14914/48008 [51:01<2:17:42, 4.01it/s] 31%|███ | 14915/48008 [51:01<2:06:38, 4.36it/s] 31%|███ | 14916/48008 [51:02<1:59:31, 4.61it/s] 31%|███ | 14917/48008 [51:02<1:56:16, 4.74it/s] 31%|███ | 14918/48008 [51:02<1:52:46, 4.89it/s] 31%|███ | 14919/48008 [51:02<1:50:50, 4.98it/s] 31%|███ | 14920/48008 [51:02<1:57:04, 4.71it/s] 31%|███ | 14921/48008 [51:03<1:56:53, 4.72it/s] 31%|███ | 14922/48008 [51:03<1:53:19, 4.87it/s] 31%|███ | 14923/48008 [51:03<1:51:18, 4.95it/s] 31%|███ | 14924/48008 [51:03<1:49:25, 5.04it/s] 31%|███ | 14925/48008 [51:03<1:48:17, 5.09it/s] 31%|███ | 14926/48008 [51:04<1:46:38, 5.17it/s] 31%|███ | 14927/48008 [51:04<1:49:49, 5.02it/s] 31%|███ | 14928/48008 [51:05<3:33:33, 2.58it/s] 31%|███ | 14929/48008 [51:05<3:00:01, 3.06it/s] 31%|███ | 14930/48008 [51:05<2:37:28, 3.50it/s] 31%|███ | 14931/48008 [51:05<2:21:12, 3.90it/s] 31%|███ | 14932/48008 [51:05<2:11:29, 4.19it/s] 31%|███ | 14933/48008 [51:06<2:05:10, 4.40it/s] 31%|███ | 14934/48008 [51:06<2:31:31, 3.64it/s] 31%|███ | 14935/48008 [51:06<2:17:43, 4.00it/s] 31%|███ | 14936/48008 [51:06<2:07:36, 4.32it/s] 31%|███ | 14937/48008 [51:07<2:00:44, 4.57it/s] 31%|███ | 14938/48008 [51:07<1:56:10, 4.74it/s] 31%|███ | 14939/48008 [51:07<1:52:44, 4.89it/s] 31%|███ | 14940/48008 [51:07<1:49:29, 5.03it/s] 31%|███ | 14941/48008 [51:07<1:48:06, 5.10it/s] 31%|███ | 14942/48008 [51:07<1:47:16, 5.14it/s] 31%|███ | 14943/48008 [51:08<1:46:44, 5.16it/s] 31%|███ | 14944/48008 [51:08<1:45:39, 5.22it/s] 31%|███ | 14945/48008 [51:08<1:47:19, 5.13it/s] 31%|███ | 14946/48008 [51:08<1:45:32, 5.22it/s] 31%|███ | 14947/48008 [51:08<1:45:07, 5.24it/s] 31%|███ | 14948/48008 [51:09<1:44:33, 5.27it/s] 31%|███ | 14949/48008 [51:09<1:45:59, 5.20it/s] 31%|███ | 14950/48008 [51:09<1:46:10, 5.19it/s] {'loss': 0.3019, 'grad_norm': 0.38671875, 'learning_rate': 0.00013772287952007998, 'epoch': 0.31} 31%|███ | 14950/48008 [51:09<1:46:10, 5.19it/s] 31%|███ | 14951/48008 [51:09<1:46:19, 5.18it/s] 31%|███ | 14952/48008 [51:09<1:46:39, 5.17it/s] 31%|███ | 14953/48008 [51:10<1:45:36, 5.22it/s] 31%|███ | 14954/48008 [51:10<1:47:17, 5.13it/s] 31%|███ | 14955/48008 [51:10<1:46:48, 5.16it/s] 31%|███ | 14956/48008 [51:10<1:48:28, 5.08it/s] 31%|███ | 14957/48008 [51:10<1:47:30, 5.12it/s] 31%|███ | 14958/48008 [51:11<1:46:59, 5.15it/s] 31%|███ | 14959/48008 [51:11<1:48:31, 5.08it/s] 31%|███ | 14960/48008 [51:11<1:47:34, 5.12it/s] 31%|███ | 14961/48008 [51:11<1:45:19, 5.23it/s] 31%|███ | 14962/48008 [51:11<1:47:11, 5.14it/s] 31%|███ | 14963/48008 [51:12<1:48:46, 5.06it/s] 31%|███ | 14964/48008 [51:12<2:06:42, 4.35it/s] 31%|███ | 14965/48008 [51:12<2:00:01, 4.59it/s] 31%|███ | 14966/48008 [51:12<1:55:01, 4.79it/s] 31%|███ | 14967/48008 [51:12<1:52:21, 4.90it/s] 31%|███ | 14968/48008 [51:13<1:48:50, 5.06it/s] 31%|███ | 14969/48008 [51:13<1:47:59, 5.10it/s] 31%|███ | 14970/48008 [51:13<1:46:49, 5.15it/s] 31%|███ | 14971/48008 [51:13<1:47:43, 5.11it/s] 31%|███ | 14972/48008 [51:13<1:46:26, 5.17it/s] 31%|███ | 14973/48008 [51:14<1:45:24, 5.22it/s] 31%|███ | 14974/48008 [51:14<1:46:14, 5.18it/s] 31%|███ | 14975/48008 [51:14<1:46:42, 5.16it/s] 31%|███ | 14976/48008 [51:14<1:46:37, 5.16it/s] 31%|███ | 14977/48008 [51:14<1:47:28, 5.12it/s] 31%|███ | 14978/48008 [51:15<1:45:43, 5.21it/s] 31%|███ | 14979/48008 [51:15<1:47:28, 5.12it/s] 31%|███ | 14980/48008 [51:15<1:46:56, 5.15it/s] 31%|███ | 14981/48008 [51:15<1:47:31, 5.12it/s] 31%|███ | 14982/48008 [51:15<1:46:50, 5.15it/s] 31%|███ | 14983/48008 [51:16<1:49:29, 5.03it/s] 31%|███ | 14984/48008 [51:16<1:47:14, 5.13it/s] 31%|███ | 14985/48008 [51:16<1:46:10, 5.18it/s] 31%|███ | 14986/48008 [51:16<1:45:48, 5.20it/s] 31%|███ | 14987/48008 [51:16<1:45:35, 5.21it/s] 31%|███ | 14988/48008 [51:17<1:53:03, 4.87it/s] 31%|███ | 14989/48008 [51:17<1:49:48, 5.01it/s] 31%|███ | 14990/48008 [51:17<1:50:14, 4.99it/s] 31%|███ | 14991/48008 [51:17<1:48:24, 5.08it/s] 31%|███ | 14992/48008 [51:17<1:48:14, 5.08it/s] 31%|███ | 14993/48008 [51:17<1:47:47, 5.10it/s] 31%|███ | 14994/48008 [51:18<1:46:53, 5.15it/s] 31%|███ | 14995/48008 [51:18<1:45:20, 5.22it/s] 31%|███ | 14996/48008 [51:18<1:44:58, 5.24it/s] 31%|███ | 14997/48008 [51:18<1:44:58, 5.24it/s] 31%|███ | 14998/48008 [51:18<1:44:48, 5.25it/s] 31%|███ | 14999/48008 [51:19<1:44:38, 5.26it/s] 31%|███ | 15000/48008 [51:19<1:44:17, 5.28it/s] {'loss': 0.3021, 'grad_norm': 0.2041015625, 'learning_rate': 0.0001375145809031828, 'epoch': 0.31} 31%|███ | 15000/48008 [51:19<1:44:17, 5.28it/s] 31%|███ | 15001/48008 [51:19<1:46:40, 5.16it/s] 31%|███ | 15002/48008 [51:19<1:47:12, 5.13it/s] 31%|███▏ | 15003/48008 [51:19<1:46:38, 5.16it/s] 31%|███▏ | 15004/48008 [51:20<1:46:23, 5.17it/s] 31%|███▏ | 15005/48008 [51:20<1:48:12, 5.08it/s] 31%|███▏ | 15006/48008 [51:20<1:46:16, 5.18it/s] 31%|███▏ | 15007/48008 [51:20<1:45:45, 5.20it/s] 31%|███▏ | 15008/48008 [51:20<1:45:40, 5.20it/s] 31%|███▏ | 15009/48008 [51:21<1:47:24, 5.12it/s] 31%|███▏ | 15010/48008 [51:21<1:48:44, 5.06it/s] 31%|███▏ | 15011/48008 [51:21<1:47:47, 5.10it/s] 31%|███▏ | 15012/48008 [51:21<1:46:48, 5.15it/s] 31%|███▏ | 15013/48008 [51:21<1:48:20, 5.08it/s] 31%|███▏ | 15014/48008 [51:22<1:47:17, 5.13it/s] 31%|███▏ | 15015/48008 [51:22<1:47:27, 5.12it/s] 31%|███▏ | 15016/48008 [51:22<1:45:59, 5.19it/s] 31%|███▏ | 15017/48008 [51:22<1:45:55, 5.19it/s] 31%|███▏ | 15018/48008 [51:22<1:46:04, 5.18it/s] 31%|███▏ | 15019/48008 [51:23<1:45:57, 5.19it/s] 31%|███▏ | 15020/48008 [51:23<1:46:40, 5.15it/s] 31%|███▏ | 15021/48008 [51:23<1:46:11, 5.18it/s] 31%|███▏ | 15022/48008 [51:23<2:17:40, 3.99it/s] 31%|███▏ | 15023/48008 [51:23<2:07:37, 4.31it/s] 31%|███▏ | 15024/48008 [51:24<2:00:11, 4.57it/s] 31%|███▏ | 15025/48008 [51:24<1:55:03, 4.78it/s] 31%|███▏ | 15026/48008 [51:24<1:52:06, 4.90it/s] 31%|███▏ | 15027/48008 [51:24<1:51:22, 4.94it/s] 31%|███▏ | 15028/48008 [51:24<1:49:19, 5.03it/s] 31%|███▏ | 15029/48008 [51:25<1:49:51, 5.00it/s] 31%|███▏ | 15030/48008 [51:25<1:49:07, 5.04it/s] 31%|███▏ | 15031/48008 [51:25<1:47:48, 5.10it/s] 31%|███▏ | 15032/48008 [51:25<1:46:34, 5.16it/s] 31%|███▏ | 15033/48008 [51:25<1:46:09, 5.18it/s] 31%|███▏ | 15034/48008 [51:26<1:45:16, 5.22it/s] 31%|███▏ | 15035/48008 [51:26<1:45:17, 5.22it/s] 31%|███▏ | 15036/48008 [51:26<1:52:35, 4.88it/s] 31%|███▏ | 15037/48008 [51:26<1:50:05, 4.99it/s] 31%|███▏ | 15038/48008 [51:26<1:47:39, 5.10it/s] 31%|███▏ | 15039/48008 [51:27<1:45:58, 5.18it/s] 31%|███▏ | 15040/48008 [51:27<1:46:43, 5.15it/s] 31%|███▏ | 15041/48008 [51:27<1:44:40, 5.25it/s] 31%|███▏ | 15042/48008 [51:27<1:46:27, 5.16it/s] 31%|███▏ | 15043/48008 [51:27<1:47:14, 5.12it/s] 31%|███▏ | 15044/48008 [51:28<1:46:39, 5.15it/s] 31%|███▏ | 15045/48008 [51:28<2:05:11, 4.39it/s] 31%|███▏ | 15046/48008 [51:28<1:59:00, 4.62it/s] 31%|███▏ | 15047/48008 [51:28<1:56:41, 4.71it/s] 31%|███▏ | 15048/48008 [51:28<1:53:02, 4.86it/s] 31%|███▏ | 15049/48008 [51:29<1:50:08, 4.99it/s] 31%|███▏ | 15050/48008 [51:29<1:47:55, 5.09it/s] {'loss': 0.301, 'grad_norm': 0.48046875, 'learning_rate': 0.00013730628228628562, 'epoch': 0.31} 31%|███▏ | 15050/48008 [51:29<1:47:55, 5.09it/s] 31%|███▏ | 15051/48008 [51:29<1:47:08, 5.13it/s] 31%|███▏ | 15052/48008 [51:29<1:46:29, 5.16it/s] 31%|███▏ | 15053/48008 [51:29<1:46:48, 5.14it/s] 31%|███▏ | 15054/48008 [51:30<1:45:36, 5.20it/s] 31%|███▏ | 15055/48008 [51:30<1:45:21, 5.21it/s] 31%|███▏ | 15056/48008 [51:30<1:44:46, 5.24it/s] 31%|███▏ | 15057/48008 [51:30<1:44:48, 5.24it/s] 31%|███▏ | 15058/48008 [51:30<1:45:59, 5.18it/s] 31%|███▏ | 15059/48008 [51:31<1:45:45, 5.19it/s] 31%|███▏ | 15060/48008 [51:31<1:44:53, 5.24it/s] 31%|███▏ | 15061/48008 [51:31<1:44:05, 5.28it/s] 31%|███▏ | 15062/48008 [51:31<1:44:01, 5.28it/s] 31%|███▏ | 15063/48008 [51:31<1:46:07, 5.17it/s] 31%|███▏ | 15064/48008 [51:32<1:47:36, 5.10it/s] 31%|███▏ | 15065/48008 [51:32<1:47:19, 5.12it/s] 31%|███▏ | 15066/48008 [51:32<1:47:02, 5.13it/s] 31%|███▏ | 15067/48008 [51:32<1:46:00, 5.18it/s] 31%|███▏ | 15068/48008 [51:32<1:45:56, 5.18it/s] 31%|███▏ | 15069/48008 [51:32<1:46:45, 5.14it/s] 31%|███▏ | 15070/48008 [51:33<1:48:35, 5.06it/s] 31%|███▏ | 15071/48008 [51:33<1:49:39, 5.01it/s] 31%|███▏ | 15072/48008 [51:33<1:48:02, 5.08it/s] 31%|███▏ | 15073/48008 [51:33<1:46:36, 5.15it/s] 31%|███▏ | 15074/48008 [51:33<1:46:48, 5.14it/s] 31%|███▏ | 15075/48008 [51:34<1:45:20, 5.21it/s] 31%|███▏ | 15076/48008 [51:34<2:17:10, 4.00it/s] 31%|███▏ | 15077/48008 [51:34<2:07:37, 4.30it/s] 31%|███▏ | 15078/48008 [51:34<2:00:09, 4.57it/s] 31%|███▏ | 15079/48008 [51:35<1:55:09, 4.77it/s] 31%|███▏ | 15080/48008 [51:35<1:52:20, 4.89it/s] 31%|███▏ | 15081/48008 [51:35<2:22:14, 3.86it/s] 31%|███▏ | 15082/48008 [51:35<2:10:37, 4.20it/s] 31%|███▏ | 15083/48008 [51:36<2:03:40, 4.44it/s] 31%|███▏ | 15084/48008 [51:36<1:57:14, 4.68it/s] 31%|███▏ | 15085/48008 [51:36<2:12:35, 4.14it/s] 31%|███▏ | 15086/48008 [51:36<2:06:08, 4.35it/s] 31%|███▏ | 15087/48008 [51:36<1:59:06, 4.61it/s] 31%|███▏ | 15088/48008 [51:37<1:54:07, 4.81it/s] 31%|███▏ | 15089/48008 [51:37<1:50:28, 4.97it/s] 31%|███▏ | 15090/48008 [51:37<1:56:54, 4.69it/s] 31%|███▏ | 15091/48008 [51:37<1:53:04, 4.85it/s] 31%|███▏ | 15092/48008 [51:37<1:50:46, 4.95it/s] 31%|███▏ | 15093/48008 [51:38<1:48:23, 5.06it/s] 31%|███▏ | 15094/48008 [51:38<1:47:54, 5.08it/s] 31%|███▏ | 15095/48008 [51:38<1:47:26, 5.11it/s] 31%|███▏ | 15096/48008 [51:38<1:47:04, 5.12it/s] 31%|███▏ | 15097/48008 [51:39<2:18:34, 3.96it/s] 31%|███▏ | 15098/48008 [51:39<2:09:10, 4.25it/s] 31%|███▏ | 15099/48008 [51:39<2:33:59, 3.56it/s] 31%|███▏ | 15100/48008 [51:39<2:21:21, 3.88it/s] {'loss': 0.3014, 'grad_norm': 0.58984375, 'learning_rate': 0.00013709798366938843, 'epoch': 0.31} 31%|███▏ | 15100/48008 [51:39<2:21:21, 3.88it/s] 31%|███▏ | 15101/48008 [51:40<2:12:14, 4.15it/s] 31%|███▏ | 15102/48008 [51:40<2:04:13, 4.41it/s] 31%|███▏ | 15103/48008 [51:40<1:57:20, 4.67it/s] 31%|███▏ | 15104/48008 [51:40<1:53:41, 4.82it/s] 31%|███▏ | 15105/48008 [51:40<1:52:09, 4.89it/s] 31%|███▏ | 15106/48008 [51:41<1:49:18, 5.02it/s] 31%|███▏ | 15107/48008 [51:41<1:47:12, 5.11it/s] 31%|███▏ | 15108/48008 [51:41<1:46:54, 5.13it/s] 31%|███▏ | 15109/48008 [51:41<1:45:55, 5.18it/s] 31%|███▏ | 15110/48008 [51:41<1:45:33, 5.19it/s] 31%|███▏ | 15111/48008 [51:42<1:46:12, 5.16it/s] 31%|███▏ | 15112/48008 [51:42<1:45:29, 5.20it/s] 31%|███▏ | 15113/48008 [51:42<1:44:34, 5.24it/s] 31%|███▏ | 15114/48008 [51:42<2:16:32, 4.02it/s] 31%|███▏ | 15115/48008 [51:42<2:07:02, 4.32it/s] 31%|███▏ | 15116/48008 [51:43<2:19:24, 3.93it/s] 31%|███▏ | 15117/48008 [51:43<2:08:59, 4.25it/s] 31%|███▏ | 15118/48008 [51:43<2:01:54, 4.50it/s] 31%|███▏ | 15119/48008 [51:43<1:58:21, 4.63it/s] 31%|███▏ | 15120/48008 [51:44<1:54:21, 4.79it/s] 31%|███▏ | 15121/48008 [51:44<1:53:20, 4.84it/s] 31%|███▏ | 15122/48008 [51:44<1:50:45, 4.95it/s] 32%|███▏ | 15123/48008 [51:44<1:51:08, 4.93it/s] 32%|███▏ | 15124/48008 [51:44<1:56:25, 4.71it/s] 32%|███▏ | 15125/48008 [51:45<1:52:42, 4.86it/s] 32%|███▏ | 15126/48008 [51:45<1:50:39, 4.95it/s] 32%|███▏ | 15127/48008 [51:45<1:48:48, 5.04it/s] 32%|███▏ | 15128/48008 [51:45<1:47:14, 5.11it/s] 32%|███▏ | 15129/48008 [51:45<1:46:15, 5.16it/s] 32%|███▏ | 15130/48008 [51:46<2:04:45, 4.39it/s] 32%|███▏ | 15131/48008 [51:46<2:01:45, 4.50it/s] 32%|███▏ | 15132/48008 [51:46<1:56:29, 4.70it/s] 32%|███▏ | 15133/48008 [51:46<1:53:13, 4.84it/s] 32%|███▏ | 15134/48008 [51:46<1:50:50, 4.94it/s] 32%|███▏ | 15135/48008 [51:47<1:48:41, 5.04it/s] 32%|███▏ | 15136/48008 [51:47<1:48:12, 5.06it/s] 32%|███▏ | 15137/48008 [51:47<1:48:43, 5.04it/s] 32%|███▏ | 15138/48008 [51:47<1:46:55, 5.12it/s] 32%|███▏ | 15139/48008 [51:47<1:47:48, 5.08it/s] 32%|███▏ | 15140/48008 [51:48<1:48:31, 5.05it/s] 32%|███▏ | 15141/48008 [51:48<1:46:41, 5.13it/s] 32%|███▏ | 15142/48008 [51:48<1:46:20, 5.15it/s] 32%|███▏ | 15143/48008 [51:48<1:45:49, 5.18it/s] 32%|███▏ | 15144/48008 [51:48<1:45:00, 5.22it/s] 32%|███▏ | 15145/48008 [51:49<1:44:58, 5.22it/s] 32%|███▏ | 15146/48008 [51:49<1:45:00, 5.22it/s] 32%|███▏ | 15147/48008 [51:49<1:52:26, 4.87it/s] 32%|███▏ | 15148/48008 [51:49<2:22:13, 3.85it/s] 32%|███▏ | 15149/48008 [51:50<2:13:13, 4.11it/s] 32%|███▏ | 15150/48008 [51:50<2:04:32, 4.40it/s] {'loss': 0.3015, 'grad_norm': 0.287109375, 'learning_rate': 0.00013688968505249128, 'epoch': 0.32} 32%|███▏ | 15150/48008 [51:50<2:04:32, 4.40it/s] 32%|███▏ | 15151/48008 [51:50<2:01:53, 4.49it/s] 32%|███▏ | 15152/48008 [51:50<1:56:40, 4.69it/s] 32%|███▏ | 15153/48008 [51:50<1:52:49, 4.85it/s] 32%|███▏ | 15154/48008 [51:51<1:51:39, 4.90it/s] 32%|███▏ | 15155/48008 [51:51<1:49:41, 4.99it/s] 32%|███▏ | 15156/48008 [51:51<1:48:24, 5.05it/s] 32%|███▏ | 15157/48008 [51:51<1:47:29, 5.09it/s] 32%|███▏ | 15158/48008 [51:51<2:05:48, 4.35it/s] 32%|███▏ | 15159/48008 [51:52<2:01:19, 4.51it/s] 32%|███▏ | 15160/48008 [51:52<1:56:41, 4.69it/s] 32%|███▏ | 15161/48008 [51:52<1:54:49, 4.77it/s] 32%|███▏ | 15162/48008 [51:52<1:53:02, 4.84it/s] 32%|███▏ | 15163/48008 [51:52<1:50:40, 4.95it/s] 32%|███▏ | 15164/48008 [51:53<1:49:04, 5.02it/s] 32%|███▏ | 15165/48008 [51:53<1:47:44, 5.08it/s] 32%|███▏ | 15166/48008 [51:53<1:47:00, 5.12it/s] 32%|███▏ | 15167/48008 [51:53<1:46:42, 5.13it/s] 32%|███▏ | 15168/48008 [51:53<1:47:10, 5.11it/s] 32%|███▏ | 15169/48008 [51:54<1:53:57, 4.80it/s] 32%|███▏ | 15170/48008 [51:54<1:51:34, 4.90it/s] 32%|███▏ | 15171/48008 [51:54<1:50:28, 4.95it/s] 32%|███▏ | 15172/48008 [51:54<1:49:31, 5.00it/s] 32%|███▏ | 15173/48008 [51:54<1:46:56, 5.12it/s] 32%|███▏ | 15174/48008 [51:55<1:46:55, 5.12it/s] 32%|███▏ | 15175/48008 [51:55<1:47:25, 5.09it/s] 32%|███▏ | 15176/48008 [51:55<2:05:33, 4.36it/s] 32%|███▏ | 15177/48008 [51:55<2:00:31, 4.54it/s] 32%|███▏ | 15178/48008 [51:55<1:57:33, 4.65it/s] 32%|███▏ | 15179/48008 [51:56<2:01:21, 4.51it/s] 32%|███▏ | 15180/48008 [51:56<1:56:10, 4.71it/s] 32%|███▏ | 15181/48008 [51:56<1:52:45, 4.85it/s] 32%|███▏ | 15182/48008 [51:56<1:52:27, 4.86it/s] 32%|███▏ | 15183/48008 [51:57<1:50:15, 4.96it/s] 32%|███▏ | 15184/48008 [51:57<1:48:13, 5.05it/s] 32%|███▏ | 15185/48008 [51:57<1:47:24, 5.09it/s] 32%|███▏ | 15186/48008 [51:57<1:46:31, 5.13it/s] 32%|███▏ | 15187/48008 [51:57<2:04:53, 4.38it/s] 32%|███▏ | 15188/48008 [51:58<2:06:05, 4.34it/s] 32%|███▏ | 15189/48008 [51:58<2:00:38, 4.53it/s] 32%|███▏ | 15190/48008 [51:58<1:56:53, 4.68it/s] 32%|███▏ | 15191/48008 [51:58<1:53:11, 4.83it/s] 32%|███▏ | 15192/48008 [51:58<1:52:27, 4.86it/s] 32%|███▏ | 15193/48008 [51:59<1:50:09, 4.96it/s] 32%|███▏ | 15194/48008 [51:59<1:48:30, 5.04it/s] 32%|███▏ | 15195/48008 [51:59<1:49:48, 4.98it/s] 32%|███▏ | 15196/48008 [51:59<1:49:55, 4.97it/s] 32%|███▏ | 15197/48008 [51:59<1:50:10, 4.96it/s] 32%|███▏ | 15198/48008 [52:00<1:48:58, 5.02it/s] 32%|███▏ | 15199/48008 [52:00<1:48:17, 5.05it/s] 32%|███▏ | 15200/48008 [52:00<1:50:25, 4.95it/s] {'loss': 0.3004, 'grad_norm': 0.1904296875, 'learning_rate': 0.00013668138643559407, 'epoch': 0.32} 32%|███▏ | 15200/48008 [52:00<1:50:25, 4.95it/s] 32%|███▏ | 15201/48008 [52:00<2:20:29, 3.89it/s] 32%|███▏ | 15202/48008 [52:01<2:09:07, 4.23it/s] 32%|███▏ | 15203/48008 [52:01<2:02:24, 4.47it/s] 32%|███▏ | 15204/48008 [52:01<1:56:53, 4.68it/s] 32%|███▏ | 15205/48008 [52:01<1:55:19, 4.74it/s] 32%|███▏ | 15206/48008 [52:01<1:53:26, 4.82it/s] 32%|███▏ | 15207/48008 [52:02<1:50:55, 4.93it/s] 32%|███▏ | 15208/48008 [52:02<1:48:46, 5.03it/s] 32%|███▏ | 15209/48008 [52:02<1:49:28, 4.99it/s] 32%|███▏ | 15210/48008 [52:02<1:48:16, 5.05it/s] 32%|███▏ | 15211/48008 [52:02<1:46:52, 5.11it/s] 32%|███▏ | 15212/48008 [52:03<1:45:20, 5.19it/s] 32%|███▏ | 15213/48008 [52:03<1:44:58, 5.21it/s] 32%|███▏ | 15214/48008 [52:03<1:45:52, 5.16it/s] 32%|███▏ | 15215/48008 [52:03<1:47:14, 5.10it/s] 32%|███▏ | 15216/48008 [52:03<1:48:11, 5.05it/s] 32%|███▏ | 15217/48008 [52:03<1:46:11, 5.15it/s] 32%|███▏ | 15218/48008 [52:04<1:45:28, 5.18it/s] 32%|███▏ | 15219/48008 [52:04<2:17:06, 3.99it/s] 32%|███▏ | 15220/48008 [52:04<2:07:37, 4.28it/s] 32%|███▏ | 15221/48008 [52:05<2:32:42, 3.58it/s] 32%|███▏ | 15222/48008 [52:05<2:21:19, 3.87it/s] 32%|███▏ | 15223/48008 [52:05<2:09:25, 4.22it/s] 32%|███▏ | 15224/48008 [52:05<2:02:16, 4.47it/s] 32%|███▏ | 15225/48008 [52:05<1:56:18, 4.70it/s] 32%|███▏ | 15226/48008 [52:06<1:53:58, 4.79it/s] 32%|███▏ | 15227/48008 [52:06<1:50:27, 4.95it/s] 32%|███▏ | 15228/48008 [52:06<1:48:07, 5.05it/s] 32%|███▏ | 15229/48008 [52:06<1:47:00, 5.11it/s] 32%|███▏ | 15230/48008 [52:06<1:48:15, 5.05it/s] 32%|███▏ | 15231/48008 [52:07<1:46:56, 5.11it/s] 32%|███▏ | 15232/48008 [52:07<1:46:17, 5.14it/s] 32%|███▏ | 15233/48008 [52:07<1:46:26, 5.13it/s] 32%|███▏ | 15234/48008 [52:07<1:45:41, 5.17it/s] 32%|███▏ | 15235/48008 [52:07<1:45:03, 5.20it/s] 32%|███▏ | 15236/48008 [52:08<1:46:53, 5.11it/s] 32%|███▏ | 15237/48008 [52:08<1:47:54, 5.06it/s] 32%|███▏ | 15238/48008 [52:08<1:50:17, 4.95it/s] 32%|███▏ | 15239/48008 [52:08<1:48:47, 5.02it/s] 32%|███▏ | 15240/48008 [52:08<1:49:21, 4.99it/s] 32%|███▏ | 15241/48008 [52:09<1:47:57, 5.06it/s] 32%|███▏ | 15242/48008 [52:09<1:49:43, 4.98it/s] 32%|███▏ | 15243/48008 [52:09<1:48:04, 5.05it/s] 32%|███▏ | 15244/48008 [52:09<1:46:51, 5.11it/s] 32%|███▏ | 15245/48008 [52:09<1:47:21, 5.09it/s] 32%|███▏ | 15246/48008 [52:10<1:47:15, 5.09it/s] 32%|███▏ | 15247/48008 [52:10<1:45:24, 5.18it/s] 32%|███▏ | 15248/48008 [52:10<1:46:18, 5.14it/s] 32%|███▏ | 15249/48008 [52:10<1:45:14, 5.19it/s] 32%|███▏ | 15250/48008 [52:10<1:44:35, 5.22it/s] {'loss': 0.3003, 'grad_norm': 0.244140625, 'learning_rate': 0.0001364730878186969, 'epoch': 0.32} 32%|███▏ | 15250/48008 [52:10<1:44:35, 5.22it/s] 32%|███▏ | 15251/48008 [52:11<1:46:15, 5.14it/s] 32%|███▏ | 15252/48008 [52:11<1:44:51, 5.21it/s] 32%|███▏ | 15253/48008 [52:11<1:44:14, 5.24it/s] 32%|███▏ | 15254/48008 [52:11<1:43:46, 5.26it/s] 32%|███▏ | 15255/48008 [52:11<1:44:07, 5.24it/s] 32%|███▏ | 15256/48008 [52:11<1:43:18, 5.28it/s] 32%|███▏ | 15257/48008 [52:12<1:42:58, 5.30it/s] 32%|███▏ | 15258/48008 [52:12<1:43:16, 5.29it/s] 32%|███▏ | 15259/48008 [52:12<2:02:31, 4.45it/s] 32%|███▏ | 15260/48008 [52:12<1:57:06, 4.66it/s] 32%|███▏ | 15261/48008 [52:13<1:52:48, 4.84it/s] 32%|███▏ | 15262/48008 [52:13<1:50:51, 4.92it/s] 32%|███▏ | 15263/48008 [52:13<1:50:35, 4.94it/s] 32%|███▏ | 15264/48008 [52:13<1:48:28, 5.03it/s] 32%|███▏ | 15265/48008 [52:13<1:47:27, 5.08it/s] 32%|███▏ | 15266/48008 [52:13<1:47:14, 5.09it/s] 32%|███▏ | 15267/48008 [52:14<1:49:26, 4.99it/s] 32%|███▏ | 15268/48008 [52:14<1:47:35, 5.07it/s] 32%|███▏ | 15269/48008 [52:14<2:05:29, 4.35it/s] 32%|███▏ | 15270/48008 [52:14<1:58:59, 4.59it/s] 32%|███▏ | 15271/48008 [52:15<1:54:05, 4.78it/s] 32%|███▏ | 15272/48008 [52:15<1:54:19, 4.77it/s] 32%|███▏ | 15273/48008 [52:15<1:59:01, 4.58it/s] 32%|███▏ | 15274/48008 [52:15<1:54:28, 4.77it/s] 32%|███▏ | 15275/48008 [52:15<1:53:29, 4.81it/s] 32%|███▏ | 15276/48008 [52:16<1:52:15, 4.86it/s] 32%|███▏ | 15277/48008 [52:16<1:49:58, 4.96it/s] 32%|███▏ | 15278/48008 [52:16<1:48:16, 5.04it/s] 32%|███▏ | 15279/48008 [52:16<1:54:46, 4.75it/s] 32%|███▏ | 15280/48008 [52:16<1:59:14, 4.57it/s] 32%|███▏ | 15281/48008 [52:17<1:53:08, 4.82it/s] 32%|███▏ | 15282/48008 [52:17<1:49:19, 4.99it/s] 32%|███▏ | 15283/48008 [52:17<1:47:32, 5.07it/s] 32%|███▏ | 15284/48008 [52:17<1:48:16, 5.04it/s] 32%|███▏ | 15285/48008 [52:17<1:46:52, 5.10it/s] 32%|███▏ | 15286/48008 [52:18<1:47:14, 5.09it/s] 32%|███▏ | 15287/48008 [52:18<2:05:00, 4.36it/s] 32%|███▏ | 15288/48008 [52:18<1:58:49, 4.59it/s] 32%|███▏ | 15289/48008 [52:18<1:55:11, 4.73it/s] 32%|███▏ | 15290/48008 [52:18<1:51:00, 4.91it/s] 32%|███▏ | 15291/48008 [52:19<1:48:49, 5.01it/s] 32%|███▏ | 15292/48008 [52:19<1:47:26, 5.07it/s] 32%|███▏ | 15293/48008 [52:19<1:46:53, 5.10it/s] 32%|███▏ | 15294/48008 [52:19<1:45:28, 5.17it/s] 32%|███▏ | 15295/48008 [52:19<1:52:37, 4.84it/s] 32%|███▏ | 15296/48008 [52:20<1:49:16, 4.99it/s] 32%|███▏ | 15297/48008 [52:20<1:50:54, 4.92it/s] 32%|███▏ | 15298/48008 [52:20<1:50:55, 4.91it/s] 32%|███▏ | 15299/48008 [52:20<1:48:55, 5.00it/s] 32%|███▏ | 15300/48008 [52:20<1:49:33, 4.98it/s] {'loss': 0.2999, 'grad_norm': 0.30078125, 'learning_rate': 0.0001362647892017997, 'epoch': 0.32} 32%|███▏ | 15300/48008 [52:21<1:49:33, 4.98it/s] 32%|███▏ | 15301/48008 [52:21<1:48:17, 5.03it/s] 32%|███▏ | 15302/48008 [52:21<1:50:16, 4.94it/s] 32%|███▏ | 15303/48008 [52:21<1:48:40, 5.02it/s] 32%|███▏ | 15304/48008 [52:21<2:19:16, 3.91it/s] 32%|███▏ | 15305/48008 [52:22<2:10:09, 4.19it/s] 32%|███▏ | 15306/48008 [52:22<2:02:34, 4.45it/s] 32%|███▏ | 15307/48008 [52:22<1:56:18, 4.69it/s] 32%|███▏ | 15308/48008 [52:22<1:53:17, 4.81it/s] 32%|███▏ | 15309/48008 [52:22<1:50:29, 4.93it/s] 32%|███▏ | 15310/48008 [52:23<1:49:41, 4.97it/s] 32%|███▏ | 15311/48008 [52:23<1:48:16, 5.03it/s] 32%|███▏ | 15312/48008 [52:23<1:48:55, 5.00it/s] 32%|███▏ | 15313/48008 [52:23<2:06:35, 4.30it/s] 32%|███▏ | 15314/48008 [52:24<2:00:04, 4.54it/s] 32%|███▏ | 15315/48008 [52:24<1:54:37, 4.75it/s] 32%|███▏ | 15316/48008 [52:24<1:51:11, 4.90it/s] 32%|███▏ | 15317/48008 [52:24<1:49:52, 4.96it/s] 32%|███▏ | 15318/48008 [52:24<1:51:24, 4.89it/s] 32%|███▏ | 15319/48008 [52:25<1:49:00, 5.00it/s] 32%|███▏ | 15320/48008 [52:25<1:47:29, 5.07it/s] 32%|███▏ | 15321/48008 [52:25<1:46:00, 5.14it/s] 32%|███▏ | 15322/48008 [52:25<1:46:50, 5.10it/s] 32%|███▏ | 15323/48008 [52:25<1:53:40, 4.79it/s] 32%|███▏ | 15324/48008 [52:26<1:50:43, 4.92it/s] 32%|███▏ | 15325/48008 [52:26<1:47:58, 5.04it/s] 32%|███▏ | 15326/48008 [52:26<1:45:43, 5.15it/s] 32%|███▏ | 15327/48008 [52:26<1:45:30, 5.16it/s] 32%|███▏ | 15328/48008 [52:26<1:46:54, 5.10it/s] 32%|███▏ | 15329/48008 [52:26<1:46:11, 5.13it/s] 32%|███▏ | 15330/48008 [52:27<1:45:52, 5.14it/s] 32%|███▏ | 15331/48008 [52:27<1:44:48, 5.20it/s] 32%|███▏ | 15332/48008 [52:27<1:44:53, 5.19it/s] 32%|███▏ | 15333/48008 [52:27<2:16:47, 3.98it/s] 32%|███▏ | 15334/48008 [52:28<2:08:09, 4.25it/s] 32%|███▏ | 15335/48008 [52:28<2:00:38, 4.51it/s] 32%|███▏ | 15336/48008 [52:28<1:55:09, 4.73it/s] 32%|███▏ | 15337/48008 [52:28<1:50:42, 4.92it/s] 32%|███▏ | 15338/48008 [52:28<1:48:27, 5.02it/s] 32%|███▏ | 15339/48008 [52:29<1:47:05, 5.08it/s] 32%|███▏ | 15340/48008 [52:29<1:44:51, 5.19it/s] 32%|███▏ | 15341/48008 [52:29<1:44:41, 5.20it/s] 32%|███▏ | 15342/48008 [52:29<1:45:28, 5.16it/s] 32%|███▏ | 15343/48008 [52:29<1:44:52, 5.19it/s] 32%|███▏ | 15344/48008 [52:30<1:44:43, 5.20it/s] 32%|███▏ | 15345/48008 [52:30<1:44:51, 5.19it/s] 32%|███▏ | 15346/48008 [52:30<1:44:38, 5.20it/s] 32%|███▏ | 15347/48008 [52:30<1:45:29, 5.16it/s] 32%|███▏ | 15348/48008 [52:30<1:44:27, 5.21it/s] 32%|███▏ | 15349/48008 [52:30<1:45:08, 5.18it/s] 32%|███▏ | 15350/48008 [52:31<1:44:26, 5.21it/s] {'loss': 0.3011, 'grad_norm': 0.478515625, 'learning_rate': 0.00013605649058490252, 'epoch': 0.32} 32%|███▏ | 15350/48008 [52:31<1:44:26, 5.21it/s] 32%|███▏ | 15351/48008 [52:31<1:42:46, 5.30it/s] 32%|███▏ | 15352/48008 [52:31<1:45:14, 5.17it/s] 32%|███▏ | 15353/48008 [52:31<1:44:33, 5.21it/s] 32%|███▏ | 15354/48008 [52:31<1:44:15, 5.22it/s] 32%|███▏ | 15355/48008 [52:32<1:43:18, 5.27it/s] 32%|███▏ | 15356/48008 [52:32<1:43:16, 5.27it/s] 32%|███▏ | 15357/48008 [52:32<1:44:29, 5.21it/s] 32%|███▏ | 15358/48008 [52:32<1:43:42, 5.25it/s] 32%|███▏ | 15359/48008 [52:32<1:43:52, 5.24it/s] 32%|███▏ | 15360/48008 [52:33<1:43:55, 5.24it/s] 32%|███▏ | 15361/48008 [52:33<1:44:22, 5.21it/s] 32%|███▏ | 15362/48008 [52:33<2:02:55, 4.43it/s] 32%|███▏ | 15363/48008 [52:33<1:58:28, 4.59it/s] 32%|███▏ | 15364/48008 [52:33<1:54:06, 4.77it/s] 32%|███▏ | 15365/48008 [52:34<1:50:57, 4.90it/s] 32%|███▏ | 15366/48008 [52:34<1:48:53, 5.00it/s] 32%|███▏ | 15367/48008 [52:34<1:46:58, 5.09it/s] 32%|███▏ | 15368/48008 [52:34<1:45:42, 5.15it/s] 32%|███▏ | 15369/48008 [52:34<1:44:37, 5.20it/s] 32%|███▏ | 15370/48008 [52:35<1:44:37, 5.20it/s] 32%|███▏ | 15371/48008 [52:35<1:43:53, 5.24it/s] 32%|███▏ | 15372/48008 [52:35<1:43:51, 5.24it/s] 32%|███▏ | 15373/48008 [52:35<1:45:40, 5.15it/s] 32%|███▏ | 15374/48008 [52:35<1:45:24, 5.16it/s] 32%|███▏ | 15375/48008 [52:36<1:45:46, 5.14it/s] 32%|███▏ | 15376/48008 [52:36<1:45:36, 5.15it/s] 32%|███▏ | 15377/48008 [52:36<1:44:50, 5.19it/s] 32%|███▏ | 15378/48008 [52:36<1:44:38, 5.20it/s] 32%|███▏ | 15379/48008 [52:36<1:45:24, 5.16it/s] 32%|███▏ | 15380/48008 [52:37<1:44:59, 5.18it/s] 32%|███▏ | 15381/48008 [52:37<1:44:57, 5.18it/s] 32%|███▏ | 15382/48008 [52:37<1:44:14, 5.22it/s] 32%|███▏ | 15383/48008 [52:37<1:51:34, 4.87it/s] 32%|███▏ | 15384/48008 [52:37<1:48:46, 5.00it/s] 32%|███▏ | 15385/48008 [52:38<1:47:24, 5.06it/s] 32%|███▏ | 15386/48008 [52:38<1:47:51, 5.04it/s] 32%|███▏ | 15387/48008 [52:38<1:46:03, 5.13it/s] 32%|███▏ | 15388/48008 [52:38<1:45:09, 5.17it/s] 32%|███▏ | 15389/48008 [52:38<2:03:34, 4.40it/s] 32%|███▏ | 15390/48008 [52:39<1:57:10, 4.64it/s] 32%|███▏ | 15391/48008 [52:39<2:25:04, 3.75it/s] 32%|███▏ | 15392/48008 [52:39<2:12:53, 4.09it/s] 32%|███▏ | 15393/48008 [52:39<2:03:48, 4.39it/s] 32%|███▏ | 15394/48008 [52:40<1:57:47, 4.61it/s] 32%|███▏ | 15395/48008 [52:40<1:53:52, 4.77it/s] 32%|███▏ | 15396/48008 [52:40<1:50:49, 4.90it/s] 32%|███▏ | 15397/48008 [52:40<1:48:59, 4.99it/s] 32%|███▏ | 15398/48008 [52:40<1:47:00, 5.08it/s] 32%|███▏ | 15399/48008 [52:41<1:45:58, 5.13it/s] 32%|███▏ | 15400/48008 [52:41<1:47:11, 5.07it/s] {'loss': 0.3006, 'grad_norm': 0.224609375, 'learning_rate': 0.00013584819196800534, 'epoch': 0.32} 32%|███▏ | 15400/48008 [52:41<1:47:11, 5.07it/s] 32%|███▏ | 15401/48008 [52:41<1:46:44, 5.09it/s] 32%|███▏ | 15402/48008 [52:41<1:45:56, 5.13it/s] 32%|███▏ | 15403/48008 [52:41<1:45:44, 5.14it/s] 32%|███▏ | 15404/48008 [52:42<1:45:26, 5.15it/s] 32%|███▏ | 15405/48008 [52:42<1:46:35, 5.10it/s] 32%|███▏ | 15406/48008 [52:42<1:45:55, 5.13it/s] 32%|███▏ | 15407/48008 [52:42<1:47:02, 5.08it/s] 32%|███▏ | 15408/48008 [52:42<1:47:56, 5.03it/s] 32%|███▏ | 15409/48008 [52:43<1:48:51, 4.99it/s] 32%|███▏ | 15410/48008 [52:43<2:06:00, 4.31it/s] 32%|███▏ | 15411/48008 [52:43<1:58:58, 4.57it/s] 32%|███▏ | 15412/48008 [52:43<1:54:33, 4.74it/s] 32%|███▏ | 15413/48008 [52:43<1:50:32, 4.91it/s] 32%|███▏ | 15414/48008 [52:44<1:49:18, 4.97it/s] 32%|███▏ | 15415/48008 [52:44<1:47:26, 5.06it/s] 32%|███▏ | 15416/48008 [52:44<1:45:40, 5.14it/s] 32%|███▏ | 15417/48008 [52:44<2:16:45, 3.97it/s] 32%|███▏ | 15418/48008 [52:45<2:07:10, 4.27it/s] 32%|███▏ | 15419/48008 [52:45<1:59:33, 4.54it/s] 32%|███▏ | 15420/48008 [52:45<1:57:30, 4.62it/s] 32%|███▏ | 15421/48008 [52:45<1:53:04, 4.80it/s] 32%|███▏ | 15422/48008 [52:45<1:50:04, 4.93it/s] 32%|███▏ | 15423/48008 [52:45<1:47:57, 5.03it/s] 32%|███▏ | 15424/48008 [52:46<1:48:34, 5.00it/s] 32%|███▏ | 15425/48008 [52:46<2:19:11, 3.90it/s] 32%|███▏ | 15426/48008 [52:46<2:08:46, 4.22it/s] 32%|███▏ | 15427/48008 [52:47<2:08:48, 4.22it/s] 32%|███▏ | 15428/48008 [52:47<2:03:06, 4.41it/s] 32%|███▏ | 15429/48008 [52:47<1:57:19, 4.63it/s] 32%|███▏ | 15430/48008 [52:47<1:52:22, 4.83it/s] 32%|███▏ | 15431/48008 [52:47<1:49:59, 4.94it/s] 32%|███▏ | 15432/48008 [52:47<1:48:12, 5.02it/s] 32%|███▏ | 15433/48008 [52:48<1:48:53, 4.99it/s] 32%|███▏ | 15434/48008 [52:48<1:47:30, 5.05it/s] 32%|███▏ | 15435/48008 [52:48<1:48:12, 5.02it/s] 32%|███▏ | 15436/48008 [52:48<1:46:38, 5.09it/s] 32%|███▏ | 15437/48008 [52:48<1:47:34, 5.05it/s] 32%|███▏ | 15438/48008 [52:49<1:46:10, 5.11it/s] 32%|███▏ | 15439/48008 [52:49<1:44:42, 5.18it/s] 32%|███▏ | 15440/48008 [52:49<1:44:17, 5.20it/s] 32%|███▏ | 15441/48008 [52:49<1:46:08, 5.11it/s] 32%|███▏ | 15442/48008 [52:49<1:47:11, 5.06it/s] 32%|███▏ | 15443/48008 [52:50<2:05:01, 4.34it/s] 32%|███▏ | 15444/48008 [52:50<2:00:30, 4.50it/s] 32%|███▏ | 15445/48008 [52:50<1:57:18, 4.63it/s] 32%|███▏ | 15446/48008 [52:50<2:00:46, 4.49it/s] 32%|███▏ | 15447/48008 [52:51<1:55:45, 4.69it/s] 32%|███▏ | 15448/48008 [52:51<1:52:57, 4.80it/s] 32%|███▏ | 15449/48008 [52:51<1:52:00, 4.84it/s] 32%|███▏ | 15450/48008 [52:51<1:49:30, 4.96it/s] {'loss': 0.3003, 'grad_norm': 0.248046875, 'learning_rate': 0.00013563989335110816, 'epoch': 0.32} 32%|███▏ | 15450/48008 [52:51<1:49:30, 4.96it/s] 32%|███▏ | 15451/48008 [52:51<1:49:51, 4.94it/s] 32%|███▏ | 15452/48008 [52:52<1:47:49, 5.03it/s] 32%|███▏ | 15453/48008 [52:52<1:46:22, 5.10it/s] 32%|███▏ | 15454/48008 [52:52<1:44:31, 5.19it/s] 32%|███▏ | 15455/48008 [52:52<1:44:17, 5.20it/s] 32%|███▏ | 15456/48008 [52:52<1:45:58, 5.12it/s] 32%|███▏ | 15457/48008 [52:53<1:47:14, 5.06it/s] 32%|███▏ | 15458/48008 [52:53<1:47:02, 5.07it/s] 32%|███▏ | 15459/48008 [52:53<1:45:24, 5.15it/s] 32%|███▏ | 15460/48008 [52:53<2:16:30, 3.97it/s] 32%|███▏ | 15461/48008 [52:54<2:08:20, 4.23it/s] 32%|███▏ | 15462/48008 [52:54<2:00:24, 4.50it/s] 32%|███▏ | 15463/48008 [52:54<1:55:14, 4.71it/s] 32%|███▏ | 15464/48008 [52:54<1:52:35, 4.82it/s] 32%|███▏ | 15465/48008 [52:54<1:49:12, 4.97it/s] 32%|███▏ | 15466/48008 [52:55<2:18:54, 3.90it/s] 32%|███▏ | 15467/48008 [52:55<2:10:19, 4.16it/s] 32%|███▏ | 15468/48008 [52:55<2:02:57, 4.41it/s] 32%|███▏ | 15469/48008 [52:55<1:57:18, 4.62it/s] 32%|███▏ | 15470/48008 [52:55<1:52:52, 4.80it/s] 32%|███▏ | 15471/48008 [52:56<1:49:57, 4.93it/s] 32%|███▏ | 15472/48008 [52:56<2:06:43, 4.28it/s] 32%|███▏ | 15473/48008 [52:56<1:59:53, 4.52it/s] 32%|███▏ | 15474/48008 [52:56<2:02:42, 4.42it/s] 32%|███▏ | 15475/48008 [52:57<1:58:55, 4.56it/s] 32%|███▏ | 15476/48008 [52:57<1:53:35, 4.77it/s] 32%|███▏ | 15477/48008 [52:57<1:51:41, 4.85it/s] 32%|███▏ | 15478/48008 [52:57<1:50:22, 4.91it/s] 32%|███▏ | 15479/48008 [52:57<1:47:42, 5.03it/s] 32%|███▏ | 15480/48008 [52:58<1:46:41, 5.08it/s] 32%|███▏ | 15481/48008 [52:58<1:53:16, 4.79it/s] 32%|███▏ | 15482/48008 [52:58<1:50:41, 4.90it/s] 32%|███▏ | 15483/48008 [52:58<1:49:05, 4.97it/s] 32%|███▏ | 15484/48008 [52:58<1:47:20, 5.05it/s] 32%|███▏ | 15485/48008 [52:59<1:46:24, 5.09it/s] 32%|███▏ | 15486/48008 [52:59<1:45:21, 5.15it/s] 32%|███▏ | 15487/48008 [52:59<1:44:39, 5.18it/s] 32%|███▏ | 15488/48008 [52:59<1:46:48, 5.07it/s] 32%|███▏ | 15489/48008 [52:59<1:45:33, 5.13it/s] 32%|███▏ | 15490/48008 [53:00<1:45:23, 5.14it/s] 32%|███▏ | 15491/48008 [53:00<1:46:55, 5.07it/s] 32%|███▏ | 15492/48008 [53:00<1:45:49, 5.12it/s] 32%|███▏ | 15493/48008 [53:00<1:44:53, 5.17it/s] 32%|███▏ | 15494/48008 [53:00<1:44:20, 5.19it/s] 32%|███▏ | 15495/48008 [53:00<1:44:17, 5.20it/s] 32%|███▏ | 15496/48008 [53:01<1:44:08, 5.20it/s] 32%|███▏ | 15497/48008 [53:01<1:43:34, 5.23it/s] 32%|███▏ | 15498/48008 [53:01<1:51:12, 4.87it/s] 32%|███▏ | 15499/48008 [53:01<1:49:11, 4.96it/s] 32%|███▏ | 15500/48008 [53:01<1:47:03, 5.06it/s] {'loss': 0.3009, 'grad_norm': 0.3125, 'learning_rate': 0.00013543159473421097, 'epoch': 0.32} 32%|███▏ | 15500/48008 [53:01<1:47:03, 5.06it/s] 32%|███▏ | 15501/48008 [53:02<1:47:10, 5.06it/s] 32%|███▏ | 15502/48008 [53:02<1:46:07, 5.10it/s] 32%|███▏ | 15503/48008 [53:02<1:45:10, 5.15it/s] 32%|███▏ | 15504/48008 [53:02<1:44:39, 5.18it/s] 32%|███▏ | 15505/48008 [53:03<3:57:20, 2.28it/s] 32%|███▏ | 15506/48008 [53:03<3:17:29, 2.74it/s] 32%|███▏ | 15507/48008 [53:04<2:49:31, 3.20it/s] 32%|███▏ | 15508/48008 [53:04<2:31:32, 3.57it/s] 32%|███▏ | 15509/48008 [53:04<2:17:01, 3.95it/s] 32%|███▏ | 15510/48008 [53:04<2:08:18, 4.22it/s] 32%|███▏ | 15511/48008 [53:05<3:53:25, 2.32it/s] 32%|███▏ | 15512/48008 [53:05<3:46:19, 2.39it/s] 32%|███▏ | 15513/48008 [53:06<3:08:37, 2.87it/s] 32%|███▏ | 15514/48008 [53:06<2:43:09, 3.32it/s] 32%|███▏ | 15515/48008 [53:06<2:25:32, 3.72it/s] 32%|███▏ | 15516/48008 [53:06<2:12:18, 4.09it/s] 32%|███▏ | 15517/48008 [53:07<4:41:44, 1.92it/s] 32%|███▏ | 15518/48008 [53:08<3:47:41, 2.38it/s] 32%|███▏ | 15519/48008 [53:08<3:10:36, 2.84it/s] 32%|███▏ | 15520/48008 [53:08<2:44:26, 3.29it/s] 32%|███▏ | 15521/48008 [53:08<2:25:56, 3.71it/s] 32%|███▏ | 15522/48008 [53:08<2:13:26, 4.06it/s] 32%|███▏ | 15523/48008 [53:09<4:03:46, 2.22it/s] 32%|███▏ | 15524/48008 [53:09<3:21:12, 2.69it/s] 32%|███▏ | 15525/48008 [53:10<2:52:06, 3.15it/s] 32%|███▏ | 15526/48008 [53:10<2:31:42, 3.57it/s] 32%|███▏ | 15527/48008 [53:10<2:17:24, 3.94it/s] 32%|███▏ | 15528/48008 [53:10<2:06:45, 4.27it/s] 32%|███▏ | 15529/48008 [53:10<1:59:59, 4.51it/s] 32%|███▏ | 15530/48008 [53:11<1:54:25, 4.73it/s] 32%|███▏ | 15531/48008 [53:11<1:50:58, 4.88it/s] 32%|███▏ | 15532/48008 [53:11<1:48:00, 5.01it/s] 32%|███▏ | 15533/48008 [53:11<1:46:47, 5.07it/s] 32%|███▏ | 15534/48008 [53:11<1:46:00, 5.11it/s] 32%|███▏ | 15535/48008 [53:12<2:03:25, 4.38it/s] 32%|███▏ | 15536/48008 [53:12<1:59:20, 4.53it/s] 32%|███▏ | 15537/48008 [53:12<1:53:07, 4.78it/s] 32%|███▏ | 15538/48008 [53:12<1:51:27, 4.86it/s] 32%|███▏ | 15539/48008 [53:12<1:48:27, 4.99it/s] 32%|███▏ | 15540/48008 [53:13<1:47:17, 5.04it/s] 32%|███▏ | 15541/48008 [53:13<3:27:07, 2.61it/s] 32%|███▏ | 15542/48008 [53:14<2:55:57, 3.08it/s] 32%|███▏ | 15543/48008 [53:14<2:33:17, 3.53it/s] 32%|███▏ | 15544/48008 [53:14<2:20:08, 3.86it/s] 32%|███▏ | 15545/48008 [53:14<2:09:20, 4.18it/s] 32%|███▏ | 15546/48008 [53:14<2:01:25, 4.46it/s] 32%|███▏ | 15547/48008 [53:15<1:55:42, 4.68it/s] 32%|███▏ | 15548/48008 [53:15<1:51:46, 4.84it/s] 32%|███▏ | 15549/48008 [53:15<1:49:30, 4.94it/s] 32%|███▏ | 15550/48008 [53:15<1:47:36, 5.03it/s] {'loss': 0.3028, 'grad_norm': 0.44921875, 'learning_rate': 0.0001352232961173138, 'epoch': 0.32} 32%|███▏ | 15550/48008 [53:15<1:47:36, 5.03it/s] 32%|███▏ | 15551/48008 [53:15<1:45:01, 5.15it/s] 32%|███▏ | 15552/48008 [53:16<1:44:22, 5.18it/s] 32%|███▏ | 15553/48008 [53:16<1:43:42, 5.22it/s] 32%|███▏ | 15554/48008 [53:16<1:43:07, 5.24it/s] 32%|███▏ | 15555/48008 [53:16<1:44:13, 5.19it/s] 32%|███▏ | 15556/48008 [53:16<1:45:17, 5.14it/s] 32%|███▏ | 15557/48008 [53:17<1:44:18, 5.19it/s] 32%|███▏ | 15558/48008 [53:17<1:45:47, 5.11it/s] 32%|███▏ | 15559/48008 [53:17<1:52:27, 4.81it/s] 32%|███▏ | 15560/48008 [53:18<3:45:59, 2.39it/s] 32%|███▏ | 15561/48008 [53:18<3:09:22, 2.86it/s] 32%|███▏ | 15562/48008 [53:18<2:43:57, 3.30it/s] 32%|███▏ | 15563/48008 [53:18<2:26:23, 3.69it/s] 32%|███▏ | 15564/48008 [53:19<2:13:38, 4.05it/s] 32%|███▏ | 15565/48008 [53:19<2:11:39, 4.11it/s] 32%|███▏ | 15566/48008 [53:19<2:05:01, 4.32it/s] 32%|███▏ | 15567/48008 [53:19<2:01:06, 4.46it/s] 32%|███▏ | 15568/48008 [53:19<1:55:47, 4.67it/s] 32%|███▏ | 15569/48008 [53:20<2:23:28, 3.77it/s] 32%|███▏ | 15570/48008 [53:20<2:11:47, 4.10it/s] 32%|███▏ | 15571/48008 [53:20<2:02:34, 4.41it/s] 32%|███▏ | 15572/48008 [53:20<1:56:35, 4.64it/s] 32%|███▏ | 15573/48008 [53:21<1:52:33, 4.80it/s] 32%|███▏ | 15574/48008 [53:21<1:49:54, 4.92it/s] 32%|███▏ | 15575/48008 [53:21<1:47:36, 5.02it/s] 32%|███▏ | 15576/48008 [53:21<1:46:12, 5.09it/s] 32%|███▏ | 15577/48008 [53:21<1:47:07, 5.05it/s] 32%|███▏ | 15578/48008 [53:22<1:53:24, 4.77it/s] 32%|███▏ | 15579/48008 [53:22<1:50:20, 4.90it/s] 32%|███▏ | 15580/48008 [53:22<1:49:28, 4.94it/s] 32%|███▏ | 15581/48008 [53:22<1:47:23, 5.03it/s] 32%|███▏ | 15582/48008 [53:22<1:46:48, 5.06it/s] 32%|███▏ | 15583/48008 [53:23<1:47:29, 5.03it/s] 32%|███▏ | 15584/48008 [53:23<1:46:04, 5.09it/s] 32%|███▏ | 15585/48008 [53:23<1:45:44, 5.11it/s] 32%|███▏ | 15586/48008 [53:23<1:43:59, 5.20it/s] 32%|███▏ | 15587/48008 [53:23<1:43:41, 5.21it/s] 32%|███▏ | 15588/48008 [53:24<1:43:26, 5.22it/s] 32%|███▏ | 15589/48008 [53:24<1:46:26, 5.08it/s] 32%|███▏ | 15590/48008 [53:24<1:45:39, 5.11it/s] 32%|███▏ | 15591/48008 [53:24<1:43:50, 5.20it/s] 32%|███▏ | 15592/48008 [53:24<1:43:38, 5.21it/s] 32%|███▏ | 15593/48008 [53:25<1:43:25, 5.22it/s] 32%|███▏ | 15594/48008 [53:25<1:50:51, 4.87it/s] 32%|███▏ | 15595/48008 [53:25<1:48:05, 5.00it/s] 32%|███▏ | 15596/48008 [53:25<1:46:24, 5.08it/s] 32%|███▏ | 15597/48008 [53:25<1:46:13, 5.09it/s] 32%|███▏ | 15598/48008 [53:26<1:45:34, 5.12it/s] 32%|███▏ | 15599/48008 [53:26<1:44:38, 5.16it/s] 32%|███▏ | 15600/48008 [53:26<1:43:19, 5.23it/s] {'loss': 0.3014, 'grad_norm': 0.4140625, 'learning_rate': 0.0001350149975004166, 'epoch': 0.32} 32%|███▏ | 15600/48008 [53:26<1:43:19, 5.23it/s] 32%|███▏ | 15601/48008 [53:26<2:01:53, 4.43it/s] 32%|███▏ | 15602/48008 [53:26<1:55:25, 4.68it/s] 33%|███▎ | 15603/48008 [53:27<1:59:00, 4.54it/s] 33%|███▎ | 15604/48008 [53:27<1:53:57, 4.74it/s] 33%|███▎ | 15605/48008 [53:27<1:50:33, 4.89it/s] 33%|███▎ | 15606/48008 [53:27<1:48:31, 4.98it/s] 33%|███▎ | 15607/48008 [53:27<1:45:52, 5.10it/s] 33%|███▎ | 15608/48008 [53:28<1:42:54, 5.25it/s] 33%|███▎ | 15609/48008 [53:28<1:41:51, 5.30it/s] 33%|███▎ | 15610/48008 [53:28<1:42:02, 5.29it/s] 33%|███▎ | 15611/48008 [53:28<1:43:21, 5.22it/s] 33%|███▎ | 15612/48008 [53:28<1:45:09, 5.13it/s] 33%|███▎ | 15613/48008 [53:29<1:46:09, 5.09it/s] 33%|███▎ | 15614/48008 [53:29<1:45:13, 5.13it/s] 33%|███▎ | 15615/48008 [53:29<1:43:27, 5.22it/s] 33%|███▎ | 15616/48008 [53:29<1:43:20, 5.22it/s] 33%|███▎ | 15617/48008 [53:29<1:42:45, 5.25it/s] 33%|███▎ | 15618/48008 [53:30<1:44:25, 5.17it/s] 33%|███▎ | 15619/48008 [53:30<1:43:13, 5.23it/s] 33%|███▎ | 15620/48008 [53:30<1:42:49, 5.25it/s] 33%|███▎ | 15621/48008 [53:30<1:42:19, 5.28it/s] 33%|███▎ | 15622/48008 [53:30<1:42:13, 5.28it/s] 33%|███▎ | 15623/48008 [53:30<1:42:08, 5.28it/s] 33%|███▎ | 15624/48008 [53:31<1:42:20, 5.27it/s] 33%|███▎ | 15625/48008 [53:31<1:44:26, 5.17it/s] 33%|███▎ | 15626/48008 [53:31<1:43:47, 5.20it/s] 33%|███▎ | 15627/48008 [53:31<1:51:05, 4.86it/s] 33%|███▎ | 15628/48008 [53:31<1:49:25, 4.93it/s] 33%|███▎ | 15629/48008 [53:32<1:49:13, 4.94it/s] 33%|███▎ | 15630/48008 [53:32<1:47:38, 5.01it/s] 33%|███▎ | 15631/48008 [53:32<1:48:17, 4.98it/s] 33%|███▎ | 15632/48008 [53:32<1:46:58, 5.04it/s] 33%|███▎ | 15633/48008 [53:32<1:45:27, 5.12it/s] 33%|███▎ | 15634/48008 [53:33<1:44:51, 5.15it/s] 33%|███▎ | 15635/48008 [53:33<1:47:03, 5.04it/s] 33%|███▎ | 15636/48008 [53:33<1:53:09, 4.77it/s] 33%|███▎ | 15637/48008 [53:33<1:52:17, 4.80it/s] 33%|███▎ | 15638/48008 [53:33<1:49:44, 4.92it/s] 33%|███▎ | 15639/48008 [53:34<1:46:58, 5.04it/s] 33%|███▎ | 15640/48008 [53:34<1:44:55, 5.14it/s] 33%|███▎ | 15641/48008 [53:34<1:44:09, 5.18it/s] 33%|███▎ | 15642/48008 [53:34<1:43:12, 5.23it/s] 33%|███▎ | 15643/48008 [53:34<1:43:09, 5.23it/s] 33%|███▎ | 15644/48008 [53:35<1:42:10, 5.28it/s] 33%|███▎ | 15645/48008 [53:35<1:45:24, 5.12it/s] 33%|███▎ | 15646/48008 [53:35<1:44:50, 5.14it/s] 33%|███▎ | 15647/48008 [53:35<1:45:36, 5.11it/s] 33%|███▎ | 15648/48008 [53:35<1:46:56, 5.04it/s] 33%|███▎ | 15649/48008 [53:36<1:45:54, 5.09it/s] 33%|███▎ | 15650/48008 [53:36<1:44:34, 5.16it/s] {'loss': 0.3007, 'grad_norm': 0.83203125, 'learning_rate': 0.00013480669888351943, 'epoch': 0.33} 33%|███▎ | 15650/48008 [53:36<1:44:34, 5.16it/s] 33%|███▎ | 15651/48008 [53:36<1:44:30, 5.16it/s] 33%|███▎ | 15652/48008 [53:36<1:46:46, 5.05it/s] 33%|███▎ | 15653/48008 [53:36<1:53:10, 4.76it/s] 33%|███▎ | 15654/48008 [53:37<1:49:12, 4.94it/s] 33%|███▎ | 15655/48008 [53:37<1:46:37, 5.06it/s] 33%|███▎ | 15656/48008 [53:37<1:45:07, 5.13it/s] 33%|███▎ | 15657/48008 [53:37<1:44:05, 5.18it/s] 33%|███▎ | 15658/48008 [53:37<1:43:11, 5.22it/s] 33%|███▎ | 15659/48008 [53:38<1:42:28, 5.26it/s] 33%|███▎ | 15660/48008 [53:38<1:43:23, 5.21it/s] 33%|███▎ | 15661/48008 [53:38<1:43:15, 5.22it/s] 33%|███▎ | 15662/48008 [53:38<2:14:33, 4.01it/s] 33%|███▎ | 15663/48008 [53:39<2:05:23, 4.30it/s] 33%|███▎ | 15664/48008 [53:39<2:06:15, 4.27it/s] 33%|███▎ | 15665/48008 [53:39<1:59:06, 4.53it/s] 33%|███▎ | 15666/48008 [53:39<1:54:16, 4.72it/s] 33%|███▎ | 15667/48008 [53:39<1:52:54, 4.77it/s] 33%|███▎ | 15668/48008 [53:40<1:49:46, 4.91it/s] 33%|███▎ | 15669/48008 [53:40<1:46:17, 5.07it/s] 33%|███▎ | 15670/48008 [53:40<1:45:31, 5.11it/s] 33%|███▎ | 15671/48008 [53:40<1:44:10, 5.17it/s] 33%|███▎ | 15672/48008 [53:40<1:43:32, 5.21it/s] 33%|███▎ | 15673/48008 [53:40<1:43:02, 5.23it/s] 33%|███▎ | 15674/48008 [53:41<1:42:20, 5.27it/s] 33%|███▎ | 15675/48008 [53:41<1:43:28, 5.21it/s] 33%|███▎ | 15676/48008 [53:41<1:43:06, 5.23it/s] 33%|███▎ | 15677/48008 [53:41<1:42:42, 5.25it/s] 33%|███▎ | 15678/48008 [53:41<1:42:22, 5.26it/s] 33%|███▎ | 15679/48008 [53:42<1:42:27, 5.26it/s] 33%|███▎ | 15680/48008 [53:42<1:44:25, 5.16it/s] 33%|███▎ | 15681/48008 [53:42<1:43:01, 5.23it/s] 33%|███▎ | 15682/48008 [53:42<1:42:42, 5.25it/s] 33%|███▎ | 15683/48008 [53:42<1:42:43, 5.24it/s] 33%|███▎ | 15684/48008 [53:43<1:41:55, 5.29it/s] 33%|███▎ | 15685/48008 [53:43<1:41:27, 5.31it/s] 33%|███▎ | 15686/48008 [53:43<1:41:29, 5.31it/s] 33%|███▎ | 15687/48008 [53:43<1:49:09, 4.93it/s] 33%|███▎ | 15688/48008 [53:43<1:49:27, 4.92it/s] 33%|███▎ | 15689/48008 [53:44<1:49:44, 4.91it/s] 33%|███▎ | 15690/48008 [53:44<1:47:15, 5.02it/s] 33%|███▎ | 15691/48008 [53:44<1:45:58, 5.08it/s] 33%|███▎ | 15692/48008 [53:44<1:44:37, 5.15it/s] 33%|███▎ | 15693/48008 [53:44<1:45:07, 5.12it/s] 33%|███▎ | 15694/48008 [53:45<1:44:23, 5.16it/s] 33%|███▎ | 15695/48008 [53:45<1:43:19, 5.21it/s] 33%|███▎ | 15696/48008 [53:45<1:46:17, 5.07it/s] 33%|███▎ | 15697/48008 [53:45<1:46:12, 5.07it/s] 33%|███▎ | 15698/48008 [53:45<1:45:31, 5.10it/s] 33%|███▎ | 15699/48008 [53:46<2:15:58, 3.96it/s] 33%|███▎ | 15700/48008 [53:46<2:06:09, 4.27it/s] {'loss': 0.3029, 'grad_norm': 0.271484375, 'learning_rate': 0.00013459840026662222, 'epoch': 0.33} 33%|███▎ | 15700/48008 [53:46<2:06:09, 4.27it/s] 33%|███▎ | 15701/48008 [53:46<1:59:27, 4.51it/s] 33%|███▎ | 15702/48008 [53:46<1:54:13, 4.71it/s] 33%|███▎ | 15703/48008 [53:46<1:50:28, 4.87it/s] 33%|███▎ | 15704/48008 [53:47<1:47:59, 4.99it/s] 33%|███▎ | 15705/48008 [53:47<1:46:10, 5.07it/s] 33%|███▎ | 15706/48008 [53:47<1:44:11, 5.17it/s] 33%|███▎ | 15707/48008 [53:47<1:51:13, 4.84it/s] 33%|███▎ | 15708/48008 [53:47<1:48:29, 4.96it/s] 33%|███▎ | 15709/48008 [53:48<1:54:16, 4.71it/s] 33%|███▎ | 15710/48008 [53:48<1:52:10, 4.80it/s] 33%|███▎ | 15711/48008 [53:48<1:48:33, 4.96it/s] 33%|███▎ | 15712/48008 [53:48<1:48:38, 4.95it/s] 33%|███▎ | 15713/48008 [53:48<1:45:50, 5.09it/s] 33%|███▎ | 15714/48008 [53:49<1:45:58, 5.08it/s] 33%|███▎ | 15715/48008 [53:49<1:44:59, 5.13it/s] 33%|███▎ | 15716/48008 [53:49<1:44:02, 5.17it/s] 33%|███▎ | 15717/48008 [53:49<1:43:37, 5.19it/s] 33%|███▎ | 15718/48008 [53:49<1:42:36, 5.24it/s] 33%|███▎ | 15719/48008 [53:50<1:43:15, 5.21it/s] 33%|███▎ | 15720/48008 [53:50<1:45:00, 5.12it/s] 33%|███▎ | 15721/48008 [53:50<1:46:14, 5.06it/s] 33%|███▎ | 15722/48008 [53:50<1:46:01, 5.08it/s] 33%|███▎ | 15723/48008 [53:50<1:46:14, 5.06it/s] 33%|███▎ | 15724/48008 [53:51<1:44:23, 5.15it/s] 33%|███▎ | 15725/48008 [53:51<1:45:58, 5.08it/s] 33%|███▎ | 15726/48008 [53:51<1:44:28, 5.15it/s] 33%|███▎ | 15727/48008 [53:51<1:44:02, 5.17it/s] 33%|███▎ | 15728/48008 [53:51<1:46:13, 5.06it/s] 33%|███▎ | 15729/48008 [53:52<1:44:57, 5.13it/s] 33%|███▎ | 15730/48008 [53:52<1:47:26, 5.01it/s] 33%|███▎ | 15731/48008 [53:52<1:47:49, 4.99it/s] 33%|███▎ | 15732/48008 [53:52<1:47:21, 5.01it/s] 33%|███▎ | 15733/48008 [53:52<1:45:45, 5.09it/s] 33%|███▎ | 15734/48008 [53:53<1:44:39, 5.14it/s] 33%|███▎ | 15735/48008 [53:53<1:43:19, 5.21it/s] 33%|███▎ | 15736/48008 [53:53<1:45:17, 5.11it/s] 33%|███▎ | 15737/48008 [53:53<1:44:01, 5.17it/s] 33%|███▎ | 15738/48008 [53:53<1:45:00, 5.12it/s] 33%|███▎ | 15739/48008 [53:54<1:43:59, 5.17it/s] 33%|███▎ | 15740/48008 [53:54<1:43:54, 5.18it/s] 33%|███▎ | 15741/48008 [53:54<1:44:00, 5.17it/s] 33%|███▎ | 15742/48008 [53:54<1:44:06, 5.17it/s] 33%|███▎ | 15743/48008 [53:54<1:42:33, 5.24it/s] 33%|███▎ | 15744/48008 [53:55<1:43:42, 5.19it/s] 33%|███▎ | 15745/48008 [53:55<1:43:24, 5.20it/s] 33%|███▎ | 15746/48008 [53:55<1:42:14, 5.26it/s] 33%|███▎ | 15747/48008 [53:55<1:44:19, 5.15it/s] 33%|███▎ | 15748/48008 [53:55<1:51:19, 4.83it/s] 33%|███▎ | 15749/48008 [53:56<1:50:04, 4.88it/s] 33%|███▎ | 15750/48008 [53:56<1:49:41, 4.90it/s] {'loss': 0.3013, 'grad_norm': 0.24609375, 'learning_rate': 0.00013439010164972506, 'epoch': 0.33} 33%|███▎ | 15750/48008 [53:56<1:49:41, 4.90it/s] 33%|███▎ | 15751/48008 [53:56<1:55:15, 4.66it/s] 33%|███▎ | 15752/48008 [53:56<2:22:39, 3.77it/s] 33%|███▎ | 15753/48008 [53:57<2:18:19, 3.89it/s] 33%|███▎ | 15754/48008 [53:57<2:07:40, 4.21it/s] 33%|███▎ | 15755/48008 [53:57<2:18:49, 3.87it/s] 33%|███▎ | 15756/48008 [53:57<2:27:18, 3.65it/s] 33%|███▎ | 15757/48008 [53:58<2:15:07, 3.98it/s] 33%|███▎ | 15758/48008 [53:58<2:05:37, 4.28it/s] 33%|███▎ | 15759/48008 [53:58<2:17:04, 3.92it/s] 33%|███▎ | 15760/48008 [53:58<2:06:44, 4.24it/s] 33%|███▎ | 15761/48008 [53:58<2:01:10, 4.44it/s] 33%|███▎ | 15762/48008 [53:59<1:57:13, 4.58it/s] 33%|███▎ | 15763/48008 [53:59<1:52:28, 4.78it/s] 33%|███▎ | 15764/48008 [53:59<1:49:32, 4.91it/s] 33%|███▎ | 15765/48008 [53:59<1:47:04, 5.02it/s] 33%|███▎ | 15766/48008 [53:59<1:45:29, 5.09it/s] 33%|███▎ | 15767/48008 [54:00<1:44:25, 5.15it/s] 33%|███▎ | 15768/48008 [54:00<1:43:14, 5.20it/s] 33%|███▎ | 15769/48008 [54:00<1:43:43, 5.18it/s] 33%|███▎ | 15770/48008 [54:00<1:44:54, 5.12it/s] 33%|███▎ | 15771/48008 [54:00<1:44:28, 5.14it/s] 33%|███▎ | 15772/48008 [54:01<1:43:17, 5.20it/s] 33%|███▎ | 15773/48008 [54:01<1:42:30, 5.24it/s] 33%|███▎ | 15774/48008 [54:01<1:45:51, 5.07it/s] 33%|███▎ | 15775/48008 [54:01<1:52:14, 4.79it/s] 33%|███▎ | 15776/48008 [54:02<2:07:49, 4.20it/s] 33%|███▎ | 15777/48008 [54:02<2:00:19, 4.46it/s] 33%|███▎ | 15778/48008 [54:02<1:56:30, 4.61it/s] 33%|███▎ | 15779/48008 [54:02<1:51:54, 4.80it/s] 33%|███▎ | 15780/48008 [54:02<1:48:48, 4.94it/s] 33%|███▎ | 15781/48008 [54:02<1:45:31, 5.09it/s] 33%|███▎ | 15782/48008 [54:03<1:44:31, 5.14it/s] 33%|███▎ | 15783/48008 [54:03<1:44:06, 5.16it/s] 33%|███▎ | 15784/48008 [54:03<1:44:40, 5.13it/s] 33%|███▎ | 15785/48008 [54:03<2:15:29, 3.96it/s] 33%|███▎ | 15786/48008 [54:04<2:05:44, 4.27it/s] 33%|███▎ | 15787/48008 [54:04<1:57:49, 4.56it/s] 33%|███▎ | 15788/48008 [54:04<1:52:42, 4.76it/s] 33%|███▎ | 15789/48008 [54:04<1:50:41, 4.85it/s] 33%|███▎ | 15790/48008 [54:04<1:47:31, 4.99it/s] 33%|███▎ | 15791/48008 [54:05<1:45:33, 5.09it/s] 33%|███▎ | 15792/48008 [54:05<1:51:46, 4.80it/s] 33%|███▎ | 15793/48008 [54:05<1:48:49, 4.93it/s] 33%|███▎ | 15794/48008 [54:05<1:46:27, 5.04it/s] 33%|███▎ | 15795/48008 [54:05<1:44:01, 5.16it/s] 33%|███▎ | 15796/48008 [54:06<1:43:11, 5.20it/s] 33%|███▎ | 15797/48008 [54:06<1:45:42, 5.08it/s] 33%|███▎ | 15798/48008 [54:06<1:48:00, 4.97it/s] 33%|███▎ | 15799/48008 [54:06<1:47:32, 4.99it/s] 33%|███▎ | 15800/48008 [54:07<2:17:37, 3.90it/s] {'loss': 0.299, 'grad_norm': 0.609375, 'learning_rate': 0.00013418180303282785, 'epoch': 0.33} 33%|███▎ | 15800/48008 [54:07<2:17:37, 3.90it/s] 33%|███▎ | 15801/48008 [54:07<2:07:10, 4.22it/s] 33%|███▎ | 15802/48008 [54:07<1:59:05, 4.51it/s] 33%|███▎ | 15803/48008 [54:07<1:54:15, 4.70it/s] 33%|███▎ | 15804/48008 [54:07<1:51:58, 4.79it/s] 33%|███▎ | 15805/48008 [54:08<1:49:01, 4.92it/s] 33%|███▎ | 15806/48008 [54:08<1:48:50, 4.93it/s] 33%|███▎ | 15807/48008 [54:08<1:48:42, 4.94it/s] 33%|███▎ | 15808/48008 [54:08<1:46:05, 5.06it/s] 33%|███▎ | 15809/48008 [54:08<1:45:11, 5.10it/s] 33%|███▎ | 15810/48008 [54:09<2:15:47, 3.95it/s] 33%|███▎ | 15811/48008 [54:09<2:05:37, 4.27it/s] 33%|███▎ | 15812/48008 [54:09<2:00:19, 4.46it/s] 33%|███▎ | 15813/48008 [54:09<1:54:24, 4.69it/s] 33%|███▎ | 15814/48008 [54:10<1:58:20, 4.53it/s] 33%|███▎ | 15815/48008 [54:10<1:53:33, 4.73it/s] 33%|███▎ | 15816/48008 [54:10<1:50:17, 4.86it/s] 33%|███▎ | 15817/48008 [54:10<1:48:47, 4.93it/s] 33%|███▎ | 15818/48008 [54:10<1:45:40, 5.08it/s] 33%|███▎ | 15819/48008 [54:11<2:16:08, 3.94it/s] 33%|███▎ | 15820/48008 [54:11<2:07:40, 4.20it/s] 33%|███▎ | 15821/48008 [54:11<1:59:57, 4.47it/s] 33%|███▎ | 15822/48008 [54:11<1:55:37, 4.64it/s] 33%|███▎ | 15823/48008 [54:11<1:51:09, 4.83it/s] 33%|███▎ | 15824/48008 [54:12<1:48:40, 4.94it/s] 33%|███▎ | 15825/48008 [54:12<1:46:40, 5.03it/s] 33%|███▎ | 15826/48008 [54:12<1:45:09, 5.10it/s] 33%|███▎ | 15827/48008 [54:12<1:43:39, 5.17it/s] 33%|███▎ | 15828/48008 [54:12<1:43:07, 5.20it/s] 33%|███▎ | 15829/48008 [54:13<1:43:39, 5.17it/s] 33%|███▎ | 15830/48008 [54:13<1:43:03, 5.20it/s] 33%|███▎ | 15831/48008 [54:13<1:42:38, 5.22it/s] 33%|███▎ | 15832/48008 [54:13<1:43:40, 5.17it/s] 33%|███▎ | 15833/48008 [54:13<1:43:17, 5.19it/s] 33%|███▎ | 15834/48008 [54:14<1:42:59, 5.21it/s] 33%|███▎ | 15835/48008 [54:14<1:42:26, 5.23it/s] 33%|███▎ | 15836/48008 [54:14<1:43:58, 5.16it/s] 33%|███▎ | 15837/48008 [54:14<1:43:29, 5.18it/s] 33%|███▎ | 15838/48008 [54:14<1:44:00, 5.16it/s] 33%|███▎ | 15839/48008 [54:15<1:43:19, 5.19it/s] 33%|███▎ | 15840/48008 [54:15<1:43:00, 5.20it/s] 33%|███▎ | 15841/48008 [54:15<1:42:59, 5.21it/s] 33%|███▎ | 15842/48008 [54:15<1:42:35, 5.23it/s] 33%|███▎ | 15843/48008 [54:15<1:44:28, 5.13it/s] 33%|███▎ | 15844/48008 [54:15<1:44:12, 5.14it/s] 33%|███▎ | 15845/48008 [54:16<1:43:28, 5.18it/s] 33%|███▎ | 15846/48008 [54:16<1:43:19, 5.19it/s] 33%|███▎ | 15847/48008 [54:16<1:42:19, 5.24it/s] 33%|███▎ | 15848/48008 [54:16<1:41:47, 5.27it/s] 33%|███▎ | 15849/48008 [54:16<1:44:48, 5.11it/s] 33%|███▎ | 15850/48008 [54:17<1:44:59, 5.11it/s] {'loss': 0.3016, 'grad_norm': 0.578125, 'learning_rate': 0.0001339735044159307, 'epoch': 0.33} 33%|███▎ | 15850/48008 [54:17<1:44:59, 5.11it/s] 33%|███▎ | 15851/48008 [54:17<2:15:37, 3.95it/s] 33%|███▎ | 15852/48008 [54:17<2:05:38, 4.27it/s] 33%|███▎ | 15853/48008 [54:17<1:58:52, 4.51it/s] 33%|███▎ | 15854/48008 [54:18<2:01:07, 4.42it/s] 33%|███▎ | 15855/48008 [54:18<1:55:42, 4.63it/s] 33%|███▎ | 15856/48008 [54:18<1:51:34, 4.80it/s] 33%|███▎ | 15857/48008 [54:18<1:48:21, 4.95it/s] 33%|███▎ | 15858/48008 [54:18<1:45:53, 5.06it/s] 33%|███▎ | 15859/48008 [54:19<1:44:33, 5.12it/s] 33%|███▎ | 15860/48008 [54:19<1:44:12, 5.14it/s] 33%|███▎ | 15861/48008 [54:19<1:43:57, 5.15it/s] 33%|███▎ | 15862/48008 [54:19<1:43:28, 5.18it/s] 33%|███▎ | 15863/48008 [54:19<1:50:37, 4.84it/s] 33%|███▎ | 15864/48008 [54:20<1:49:13, 4.90it/s] 33%|███▎ | 15865/48008 [54:20<2:05:40, 4.26it/s] 33%|███▎ | 15866/48008 [54:20<1:58:30, 4.52it/s] 33%|███▎ | 15867/48008 [54:20<1:53:29, 4.72it/s] 33%|███▎ | 15868/48008 [54:21<1:51:40, 4.80it/s] 33%|███▎ | 15869/48008 [54:21<1:49:05, 4.91it/s] 33%|███▎ | 15870/48008 [54:21<1:47:17, 4.99it/s] 33%|███▎ | 15871/48008 [54:21<1:45:38, 5.07it/s] 33%|███▎ | 15872/48008 [54:21<1:44:34, 5.12it/s] 33%|███▎ | 15873/48008 [54:22<1:51:25, 4.81it/s] 33%|███▎ | 15874/48008 [54:22<1:48:04, 4.96it/s] 33%|███▎ | 15875/48008 [54:22<1:45:59, 5.05it/s] 33%|███▎ | 15876/48008 [54:22<1:52:13, 4.77it/s] 33%|███▎ | 15877/48008 [54:22<1:52:22, 4.77it/s] 33%|███▎ | 15878/48008 [54:23<1:49:43, 4.88it/s] 33%|███▎ | 15879/48008 [54:23<1:47:34, 4.98it/s] 33%|███▎ | 15880/48008 [54:23<1:53:44, 4.71it/s] 33%|███▎ | 15881/48008 [54:23<1:51:46, 4.79it/s] 33%|███▎ | 15882/48008 [54:23<1:56:21, 4.60it/s] 33%|███▎ | 15883/48008 [54:24<1:52:16, 4.77it/s] 33%|███▎ | 15884/48008 [54:24<1:48:57, 4.91it/s] 33%|███▎ | 15885/48008 [54:24<1:47:15, 4.99it/s] 33%|███▎ | 15886/48008 [54:24<1:46:11, 5.04it/s] 33%|███▎ | 15887/48008 [54:24<1:49:02, 4.91it/s] 33%|███▎ | 15888/48008 [54:25<1:46:35, 5.02it/s] 33%|███▎ | 15889/48008 [54:25<1:45:51, 5.06it/s] 33%|███▎ | 15890/48008 [54:25<1:45:29, 5.07it/s] 33%|███▎ | 15891/48008 [54:25<1:44:58, 5.10it/s] 33%|███▎ | 15892/48008 [54:25<1:44:02, 5.14it/s] 33%|███▎ | 15893/48008 [54:26<1:45:24, 5.08it/s] 33%|███▎ | 15894/48008 [54:26<1:45:06, 5.09it/s] 33%|███▎ | 15895/48008 [54:26<1:43:23, 5.18it/s] 33%|███▎ | 15896/48008 [54:26<1:42:31, 5.22it/s] 33%|███▎ | 15897/48008 [54:26<1:41:58, 5.25it/s] 33%|███▎ | 15898/48008 [54:26<1:42:14, 5.23it/s] 33%|███▎ | 15899/48008 [54:27<1:42:06, 5.24it/s] 33%|███▎ | 15900/48008 [54:27<1:42:09, 5.24it/s] {'loss': 0.303, 'grad_norm': 0.2890625, 'learning_rate': 0.0001337652057990335, 'epoch': 0.33} 33%|███▎ | 15900/48008 [54:27<1:42:09, 5.24it/s] 33%|███▎ | 15901/48008 [54:27<1:42:39, 5.21it/s] 33%|███▎ | 15902/48008 [54:27<1:42:31, 5.22it/s] 33%|███▎ | 15903/48008 [54:27<1:43:38, 5.16it/s] 33%|███▎ | 15904/48008 [54:28<1:42:03, 5.24it/s] 33%|███▎ | 15905/48008 [54:28<1:41:36, 5.27it/s] 33%|███▎ | 15906/48008 [54:28<1:44:31, 5.12it/s] 33%|███▎ | 15907/48008 [54:28<1:45:25, 5.08it/s] 33%|███▎ | 15908/48008 [54:28<1:46:21, 5.03it/s] 33%|███▎ | 15909/48008 [54:29<1:44:40, 5.11it/s] 33%|███▎ | 15910/48008 [54:29<1:47:00, 5.00it/s] 33%|███▎ | 15911/48008 [54:29<1:47:50, 4.96it/s] 33%|███▎ | 15912/48008 [54:29<1:47:46, 4.96it/s] 33%|███▎ | 15913/48008 [54:29<1:45:27, 5.07it/s] 33%|███▎ | 15914/48008 [54:30<1:43:58, 5.14it/s] 33%|███▎ | 15915/48008 [54:30<1:51:12, 4.81it/s] 33%|███▎ | 15916/48008 [54:30<1:48:30, 4.93it/s] 33%|███▎ | 15917/48008 [54:30<1:48:40, 4.92it/s] 33%|███▎ | 15918/48008 [54:30<1:45:30, 5.07it/s] 33%|███▎ | 15919/48008 [54:31<1:44:46, 5.10it/s] 33%|███▎ | 15920/48008 [54:31<1:43:48, 5.15it/s] 33%|███▎ | 15921/48008 [54:31<1:43:17, 5.18it/s] 33%|███▎ | 15922/48008 [54:31<1:42:47, 5.20it/s] 33%|███▎ | 15923/48008 [54:31<1:42:29, 5.22it/s] 33%|███▎ | 15924/48008 [54:32<1:44:25, 5.12it/s] 33%|███▎ | 15925/48008 [54:32<1:43:27, 5.17it/s] 33%|███▎ | 15926/48008 [54:32<1:46:29, 5.02it/s] 33%|███▎ | 15927/48008 [54:32<1:44:58, 5.09it/s] 33%|███▎ | 15928/48008 [54:32<2:02:28, 4.37it/s] 33%|███▎ | 15929/48008 [54:33<1:56:32, 4.59it/s] 33%|███▎ | 15930/48008 [54:33<1:55:12, 4.64it/s] 33%|███▎ | 15931/48008 [54:33<1:51:07, 4.81it/s] 33%|███▎ | 15932/48008 [54:33<1:47:50, 4.96it/s] 33%|███▎ | 15933/48008 [54:33<1:46:26, 5.02it/s] 33%|███▎ | 15934/48008 [54:34<1:45:06, 5.09it/s] 33%|███▎ | 15935/48008 [54:34<1:45:02, 5.09it/s] 33%|███▎ | 15936/48008 [54:34<1:46:13, 5.03it/s] 33%|███▎ | 15937/48008 [54:34<1:44:43, 5.10it/s] 33%|███▎ | 15938/48008 [54:34<1:42:53, 5.19it/s] 33%|███▎ | 15939/48008 [54:35<2:13:43, 4.00it/s] 33%|███▎ | 15940/48008 [54:35<2:04:26, 4.29it/s] 33%|███▎ | 15941/48008 [54:35<1:57:36, 4.54it/s] 33%|███▎ | 15942/48008 [54:35<1:52:43, 4.74it/s] 33%|███▎ | 15943/48008 [54:36<1:50:14, 4.85it/s] 33%|███▎ | 15944/48008 [54:36<1:49:58, 4.86it/s] 33%|███▎ | 15945/48008 [54:36<1:47:47, 4.96it/s] 33%|███▎ | 15946/48008 [54:36<1:47:56, 4.95it/s] 33%|███▎ | 15947/48008 [54:36<1:46:13, 5.03it/s] 33%|███▎ | 15948/48008 [54:37<1:45:21, 5.07it/s] 33%|███▎ | 15949/48008 [54:37<1:44:19, 5.12it/s] 33%|███▎ | 15950/48008 [54:37<1:43:18, 5.17it/s] {'loss': 0.2999, 'grad_norm': 0.169921875, 'learning_rate': 0.0001335569071821363, 'epoch': 0.33} 33%|███▎ | 15950/48008 [54:37<1:43:18, 5.17it/s] 33%|███▎ | 15951/48008 [54:37<1:43:16, 5.17it/s] 33%|███▎ | 15952/48008 [54:37<1:42:39, 5.20it/s] 33%|███▎ | 15953/48008 [54:38<1:45:25, 5.07it/s] 33%|███▎ | 15954/48008 [54:38<1:44:30, 5.11it/s] 33%|███▎ | 15955/48008 [54:38<1:43:27, 5.16it/s] 33%|███▎ | 15956/48008 [54:38<1:43:47, 5.15it/s] 33%|███▎ | 15957/48008 [54:38<1:42:55, 5.19it/s] 33%|███▎ | 15958/48008 [54:38<1:42:46, 5.20it/s] 33%|███▎ | 15959/48008 [54:39<1:44:08, 5.13it/s] 33%|███▎ | 15960/48008 [54:39<1:45:25, 5.07it/s] 33%|███▎ | 15961/48008 [54:39<1:44:26, 5.11it/s] 33%|███▎ | 15962/48008 [54:39<1:43:48, 5.15it/s] 33%|███▎ | 15963/48008 [54:39<1:44:23, 5.12it/s] 33%|███▎ | 15964/48008 [54:40<1:43:54, 5.14it/s] 33%|███▎ | 15965/48008 [54:40<1:41:58, 5.24it/s] 33%|███▎ | 15966/48008 [54:40<1:40:52, 5.29it/s] 33%|███▎ | 15967/48008 [54:40<1:40:27, 5.32it/s] 33%|███▎ | 15968/48008 [54:40<1:40:04, 5.34it/s] 33%|███▎ | 15969/48008 [54:41<1:40:49, 5.30it/s] 33%|███▎ | 15970/48008 [54:41<1:41:29, 5.26it/s] 33%|███▎ | 15971/48008 [54:41<1:42:35, 5.20it/s] 33%|███▎ | 15972/48008 [54:41<1:42:32, 5.21it/s] 33%|███▎ | 15973/48008 [54:41<1:43:07, 5.18it/s] 33%|███▎ | 15974/48008 [54:42<1:42:48, 5.19it/s] 33%|███▎ | 15975/48008 [54:42<1:42:27, 5.21it/s] 33%|███▎ | 15976/48008 [54:42<1:41:32, 5.26it/s] 33%|███▎ | 15977/48008 [54:42<1:48:59, 4.90it/s] 33%|███▎ | 15978/48008 [54:42<1:46:38, 5.01it/s] 33%|███▎ | 15979/48008 [54:43<1:46:58, 4.99it/s] 33%|███▎ | 15980/48008 [54:43<1:53:01, 4.72it/s] 33%|███▎ | 15981/48008 [54:43<1:49:28, 4.88it/s] 33%|███▎ | 15982/48008 [54:43<1:47:09, 4.98it/s] 33%|███▎ | 15983/48008 [54:43<1:45:34, 5.06it/s] 33%|███▎ | 15984/48008 [54:44<1:46:15, 5.02it/s] 33%|███▎ | 15985/48008 [54:44<1:44:07, 5.13it/s] 33%|███▎ | 15986/48008 [54:44<1:43:44, 5.14it/s] 33%|███▎ | 15987/48008 [54:44<1:45:00, 5.08it/s] 33%|███▎ | 15988/48008 [54:44<1:45:46, 5.05it/s] 33%|███▎ | 15989/48008 [54:45<1:44:28, 5.11it/s] 33%|███▎ | 15990/48008 [54:45<1:44:18, 5.12it/s] 33%|███▎ | 15991/48008 [54:45<1:43:05, 5.18it/s] 33%|███▎ | 15992/48008 [54:45<1:41:51, 5.24it/s] 33%|███▎ | 15993/48008 [54:45<1:42:44, 5.19it/s] 33%|███▎ | 15994/48008 [54:46<1:44:31, 5.10it/s] 33%|███▎ | 15995/48008 [54:46<1:51:20, 4.79it/s] 33%|███▎ | 15996/48008 [54:46<1:46:45, 5.00it/s] 33%|███▎ | 15997/48008 [54:46<1:45:17, 5.07it/s] 33%|███▎ | 15998/48008 [54:46<1:42:36, 5.20it/s] 33%|███▎ | 15999/48008 [54:47<1:41:56, 5.23it/s] 33%|███▎ | 16000/48008 [54:47<1:40:56, 5.28it/s] {'loss': 0.304, 'grad_norm': 0.1845703125, 'learning_rate': 0.00013334860856523915, 'epoch': 0.33} 33%|███▎ | 16000/48008 [54:47<1:40:56, 5.28it/s] 33%|███▎ | 16001/48008 [54:47<1:42:06, 5.22it/s] 33%|███▎ | 16002/48008 [54:47<1:42:41, 5.19it/s] 33%|███▎ | 16003/48008 [54:47<1:45:14, 5.07it/s] 33%|███▎ | 16004/48008 [54:47<1:45:19, 5.06it/s] 33%|███▎ | 16005/48008 [54:48<1:45:17, 5.07it/s] 33%|███▎ | 16006/48008 [54:48<1:43:39, 5.15it/s] 33%|███▎ | 16007/48008 [54:48<1:42:39, 5.20it/s] 33%|███▎ | 16008/48008 [54:48<1:41:29, 5.25it/s] 33%|███▎ | 16009/48008 [54:48<1:40:58, 5.28it/s] 33%|███▎ | 16010/48008 [54:49<1:40:49, 5.29it/s] 33%|███▎ | 16011/48008 [54:49<1:48:33, 4.91it/s] 33%|███▎ | 16012/48008 [54:49<1:48:24, 4.92it/s] 33%|███▎ | 16013/48008 [54:49<1:47:50, 4.94it/s] 33%|███▎ | 16014/48008 [54:49<1:45:48, 5.04it/s] 33%|███▎ | 16015/48008 [54:50<1:44:58, 5.08it/s] 33%|███▎ | 16016/48008 [54:50<1:43:36, 5.15it/s] 33%|███▎ | 16017/48008 [54:50<1:43:16, 5.16it/s] 33%|███▎ | 16018/48008 [54:50<1:43:03, 5.17it/s] 33%|███▎ | 16019/48008 [54:50<1:42:17, 5.21it/s] 33%|███▎ | 16020/48008 [54:51<1:41:42, 5.24it/s] 33%|███▎ | 16021/48008 [54:51<1:41:54, 5.23it/s] 33%|███▎ | 16022/48008 [54:51<1:41:48, 5.24it/s] 33%|███▎ | 16023/48008 [54:51<1:41:35, 5.25it/s] 33%|███▎ | 16024/48008 [54:51<1:41:09, 5.27it/s] 33%|███▎ | 16025/48008 [54:52<1:41:11, 5.27it/s] 33%|███▎ | 16026/48008 [54:52<1:42:30, 5.20it/s] 33%|███▎ | 16027/48008 [54:52<1:43:54, 5.13it/s] 33%|███▎ | 16028/48008 [54:52<1:43:25, 5.15it/s] 33%|███▎ | 16029/48008 [54:52<1:43:10, 5.17it/s] 33%|███▎ | 16030/48008 [54:53<1:42:49, 5.18it/s] 33%|███▎ | 16031/48008 [54:53<1:42:39, 5.19it/s] 33%|███▎ | 16032/48008 [54:53<1:43:17, 5.16it/s] 33%|███▎ | 16033/48008 [54:53<1:42:34, 5.20it/s] 33%|███▎ | 16034/48008 [54:53<1:42:23, 5.20it/s] 33%|███▎ | 16035/48008 [54:53<1:41:52, 5.23it/s] 33%|███▎ | 16036/48008 [54:54<1:41:40, 5.24it/s] 33%|███▎ | 16037/48008 [54:54<1:41:15, 5.26it/s] 33%|███▎ | 16038/48008 [54:54<1:40:43, 5.29it/s] 33%|███▎ | 16039/48008 [54:54<1:44:05, 5.12it/s] 33%|███▎ | 16040/48008 [54:54<1:46:45, 4.99it/s] 33%|███▎ | 16041/48008 [54:55<1:45:00, 5.07it/s] 33%|███▎ | 16042/48008 [54:55<1:44:01, 5.12it/s] 33%|███▎ | 16043/48008 [54:55<1:45:25, 5.05it/s] 33%|███▎ | 16044/48008 [54:55<1:42:47, 5.18it/s] 33%|███▎ | 16045/48008 [54:55<1:42:10, 5.21it/s] 33%|███▎ | 16046/48008 [54:56<1:40:59, 5.27it/s] 33%|███▎ | 16047/48008 [54:56<1:59:39, 4.45it/s] 33%|███▎ | 16048/48008 [54:56<1:56:37, 4.57it/s] 33%|███▎ | 16049/48008 [54:56<1:53:13, 4.70it/s] 33%|███▎ | 16050/48008 [54:57<1:49:43, 4.85it/s] {'loss': 0.3018, 'grad_norm': 0.25390625, 'learning_rate': 0.00013314030994834194, 'epoch': 0.33} 33%|███▎ | 16050/48008 [54:57<1:49:43, 4.85it/s] 33%|███▎ | 16051/48008 [54:57<1:46:45, 4.99it/s] 33%|███▎ | 16052/48008 [54:57<1:44:53, 5.08it/s] 33%|███▎ | 16053/48008 [54:57<1:43:19, 5.15it/s] 33%|███▎ | 16054/48008 [54:57<1:50:17, 4.83it/s] 33%|███▎ | 16055/48008 [54:58<2:06:14, 4.22it/s] 33%|███▎ | 16056/48008 [54:58<1:57:42, 4.52it/s] 33%|███▎ | 16057/48008 [54:58<1:53:15, 4.70it/s] 33%|███▎ | 16058/48008 [54:58<1:50:05, 4.84it/s] 33%|███▎ | 16059/48008 [54:58<1:47:30, 4.95it/s] 33%|███▎ | 16060/48008 [54:59<1:45:54, 5.03it/s] 33%|███▎ | 16061/48008 [54:59<1:46:36, 4.99it/s] 33%|███▎ | 16062/48008 [54:59<1:45:53, 5.03it/s] 33%|███▎ | 16063/48008 [54:59<1:45:12, 5.06it/s] 33%|███▎ | 16064/48008 [54:59<1:43:50, 5.13it/s] 33%|███▎ | 16065/48008 [55:00<1:43:41, 5.13it/s] 33%|███▎ | 16066/48008 [55:00<1:43:13, 5.16it/s] 33%|███▎ | 16067/48008 [55:00<1:43:15, 5.16it/s] 33%|███▎ | 16068/48008 [55:00<1:50:28, 4.82it/s] 33%|███▎ | 16069/48008 [55:00<1:50:10, 4.83it/s] 33%|███▎ | 16070/48008 [55:01<1:47:48, 4.94it/s] 33%|███▎ | 16071/48008 [55:01<1:45:58, 5.02it/s] 33%|███▎ | 16072/48008 [55:01<1:44:50, 5.08it/s] 33%|███▎ | 16073/48008 [55:01<1:43:42, 5.13it/s] 33%|███▎ | 16074/48008 [55:01<1:42:12, 5.21it/s] 33%|███▎ | 16075/48008 [55:02<1:42:17, 5.20it/s] 33%|███▎ | 16076/48008 [55:02<1:45:20, 5.05it/s] 33%|███▎ | 16077/48008 [55:02<1:44:51, 5.08it/s] 33%|███▎ | 16078/48008 [55:02<1:43:19, 5.15it/s] 33%|███▎ | 16079/48008 [55:02<1:44:28, 5.09it/s] 33%|███▎ | 16080/48008 [55:02<1:43:25, 5.15it/s] 33%|███▎ | 16081/48008 [55:03<1:44:17, 5.10it/s] 33%|███▎ | 16082/48008 [55:03<2:02:09, 4.36it/s] 34%|███▎ | 16083/48008 [55:03<1:55:27, 4.61it/s] 34%|███▎ | 16084/48008 [55:03<1:50:41, 4.81it/s] 34%|███▎ | 16085/48008 [55:04<1:48:24, 4.91it/s] 34%|███▎ | 16086/48008 [55:04<1:46:49, 4.98it/s] 34%|███▎ | 16087/48008 [55:04<1:45:40, 5.03it/s] 34%|███▎ | 16088/48008 [55:04<1:44:32, 5.09it/s] 34%|███▎ | 16089/48008 [55:04<1:45:39, 5.04it/s] 34%|███▎ | 16090/48008 [55:05<1:44:29, 5.09it/s] 34%|███▎ | 16091/48008 [55:05<1:45:42, 5.03it/s] 34%|███▎ | 16092/48008 [55:05<1:44:09, 5.11it/s] 34%|███▎ | 16093/48008 [55:05<1:43:13, 5.15it/s] 34%|███▎ | 16094/48008 [55:05<1:42:33, 5.19it/s] 34%|███▎ | 16095/48008 [55:06<1:41:43, 5.23it/s] 34%|███▎ | 16096/48008 [55:06<1:41:49, 5.22it/s] 34%|███▎ | 16097/48008 [55:06<1:43:38, 5.13it/s] 34%|███▎ | 16098/48008 [55:06<1:43:25, 5.14it/s] 34%|███▎ | 16099/48008 [55:06<1:50:25, 4.82it/s] 34%|███▎ | 16100/48008 [55:07<1:47:45, 4.93it/s] {'loss': 0.3027, 'grad_norm': 0.29296875, 'learning_rate': 0.00013293201133144478, 'epoch': 0.34} 34%|███▎ | 16100/48008 [55:07<1:47:45, 4.93it/s] 34%|███▎ | 16101/48008 [55:07<1:47:35, 4.94it/s] 34%|███▎ | 16102/48008 [55:07<1:47:05, 4.97it/s] 34%|███▎ | 16103/48008 [55:07<1:45:25, 5.04it/s] 34%|███▎ | 16104/48008 [55:07<1:46:21, 5.00it/s] 34%|███▎ | 16105/48008 [55:08<1:44:24, 5.09it/s] 34%|███▎ | 16106/48008 [55:08<1:43:22, 5.14it/s] 34%|███▎ | 16107/48008 [55:08<1:44:45, 5.08it/s] 34%|███▎ | 16108/48008 [55:08<1:44:36, 5.08it/s] 34%|███▎ | 16109/48008 [55:08<1:44:52, 5.07it/s] 34%|███▎ | 16110/48008 [55:08<1:42:51, 5.17it/s] 34%|███▎ | 16111/48008 [55:09<1:41:41, 5.23it/s] 34%|███▎ | 16112/48008 [55:09<1:42:06, 5.21it/s] 34%|███▎ | 16113/48008 [55:09<1:41:40, 5.23it/s] 34%|███▎ | 16114/48008 [55:09<1:42:31, 5.18it/s] 34%|███▎ | 16115/48008 [55:09<1:42:29, 5.19it/s] 34%|███▎ | 16116/48008 [55:10<1:41:14, 5.25it/s] 34%|███▎ | 16117/48008 [55:10<1:42:37, 5.18it/s] 34%|███▎ | 16118/48008 [55:10<1:44:01, 5.11it/s] 34%|███▎ | 16119/48008 [55:10<1:43:08, 5.15it/s] 34%|███▎ | 16120/48008 [55:10<1:41:52, 5.22it/s] 34%|███▎ | 16121/48008 [55:11<1:41:17, 5.25it/s] 34%|███▎ | 16122/48008 [55:11<1:41:31, 5.23it/s] 34%|███▎ | 16123/48008 [55:11<1:41:05, 5.26it/s] 34%|███▎ | 16124/48008 [55:11<1:43:10, 5.15it/s] 34%|███▎ | 16125/48008 [55:11<1:41:40, 5.23it/s] 34%|███▎ | 16126/48008 [55:12<1:40:53, 5.27it/s] 34%|███▎ | 16127/48008 [55:12<1:40:50, 5.27it/s] 34%|███▎ | 16128/48008 [55:12<1:41:22, 5.24it/s] 34%|███▎ | 16129/48008 [55:12<1:42:37, 5.18it/s] 34%|███▎ | 16130/48008 [55:12<1:41:48, 5.22it/s] 34%|███▎ | 16131/48008 [55:13<1:40:48, 5.27it/s] 34%|███▎ | 16132/48008 [55:13<1:42:11, 5.20it/s] 34%|███▎ | 16133/48008 [55:13<1:43:40, 5.12it/s] 34%|███▎ | 16134/48008 [55:13<1:42:47, 5.17it/s] 34%|███▎ | 16135/48008 [55:13<1:41:50, 5.22it/s] 34%|███▎ | 16136/48008 [55:13<1:41:56, 5.21it/s] 34%|███▎ | 16137/48008 [55:14<1:42:39, 5.17it/s] 34%|███▎ | 16138/48008 [55:14<1:42:06, 5.20it/s] 34%|███▎ | 16139/48008 [55:14<1:40:56, 5.26it/s] 34%|███▎ | 16140/48008 [55:14<1:40:54, 5.26it/s] 34%|███▎ | 16141/48008 [55:14<1:43:40, 5.12it/s] 34%|███▎ | 16142/48008 [55:15<1:42:56, 5.16it/s] 34%|███▎ | 16143/48008 [55:15<1:49:37, 4.84it/s] 34%|███▎ | 16144/48008 [55:15<1:48:00, 4.92it/s] 34%|███▎ | 16145/48008 [55:15<1:46:14, 5.00it/s] 34%|███▎ | 16146/48008 [55:15<1:44:12, 5.10it/s] 34%|███▎ | 16147/48008 [55:16<1:44:55, 5.06it/s] 34%|███▎ | 16148/48008 [55:16<1:43:37, 5.12it/s] 34%|███▎ | 16149/48008 [55:16<1:45:42, 5.02it/s] 34%|███▎ | 16150/48008 [55:16<1:43:27, 5.13it/s] {'loss': 0.303, 'grad_norm': 0.50390625, 'learning_rate': 0.00013272371271454757, 'epoch': 0.34} 34%|███▎ | 16150/48008 [55:16<1:43:27, 5.13it/s] 34%|███▎ | 16151/48008 [55:16<1:42:09, 5.20it/s] 34%|███▎ | 16152/48008 [55:17<1:41:47, 5.22it/s] 34%|███▎ | 16153/48008 [55:17<1:41:24, 5.24it/s] 34%|███▎ | 16154/48008 [55:17<1:43:07, 5.15it/s] 34%|███▎ | 16155/48008 [55:17<1:42:38, 5.17it/s] 34%|███▎ | 16156/48008 [55:17<1:43:13, 5.14it/s] 34%|███▎ | 16157/48008 [55:18<1:42:41, 5.17it/s] 34%|███▎ | 16158/48008 [55:18<1:42:52, 5.16it/s] 34%|███▎ | 16159/48008 [55:18<1:43:29, 5.13it/s] 34%|███▎ | 16160/48008 [55:18<1:41:52, 5.21it/s] 34%|███▎ | 16161/48008 [55:18<1:41:09, 5.25it/s] 34%|███▎ | 16162/48008 [55:19<1:41:18, 5.24it/s] 34%|███▎ | 16163/48008 [55:19<2:12:19, 4.01it/s] 34%|███▎ | 16164/48008 [55:19<2:03:13, 4.31it/s] 34%|███▎ | 16165/48008 [55:19<1:58:21, 4.48it/s] 34%|███▎ | 16166/48008 [55:20<1:53:18, 4.68it/s] 34%|███▎ | 16167/48008 [55:20<1:49:31, 4.85it/s] 34%|███▎ | 16168/48008 [55:20<1:47:19, 4.94it/s] 34%|███▎ | 16169/48008 [55:20<1:44:55, 5.06it/s] 34%|███▎ | 16170/48008 [55:20<1:44:46, 5.06it/s] 34%|███▎ | 16171/48008 [55:20<1:43:46, 5.11it/s] 34%|███▎ | 16172/48008 [55:21<1:42:34, 5.17it/s] 34%|███▎ | 16173/48008 [55:21<1:43:50, 5.11it/s] 34%|███▎ | 16174/48008 [55:21<1:42:16, 5.19it/s] 34%|███▎ | 16175/48008 [55:21<1:49:26, 4.85it/s] 34%|███▎ | 16176/48008 [55:21<1:47:25, 4.94it/s] 34%|███▎ | 16177/48008 [55:22<1:46:23, 4.99it/s] 34%|███▎ | 16178/48008 [55:22<1:44:44, 5.07it/s] 34%|███▎ | 16179/48008 [55:22<1:43:57, 5.10it/s] 34%|███▎ | 16180/48008 [55:22<1:42:56, 5.15it/s] 34%|███▎ | 16181/48008 [55:22<1:42:10, 5.19it/s] 34%|███▎ | 16182/48008 [55:23<1:40:57, 5.25it/s] 34%|███▎ | 16183/48008 [55:23<1:41:16, 5.24it/s] 34%|███▎ | 16184/48008 [55:23<1:40:16, 5.29it/s] 34%|███▎ | 16185/48008 [55:23<1:40:22, 5.28it/s] 34%|███▎ | 16186/48008 [55:23<1:41:47, 5.21it/s] 34%|███▎ | 16187/48008 [55:24<1:41:24, 5.23it/s] 34%|███▎ | 16188/48008 [55:24<1:59:49, 4.43it/s] 34%|███▎ | 16189/48008 [55:24<1:55:58, 4.57it/s] 34%|███▎ | 16190/48008 [55:24<1:59:13, 4.45it/s] 34%|███▎ | 16191/48008 [55:25<1:55:39, 4.58it/s] 34%|███▎ | 16192/48008 [55:25<1:51:42, 4.75it/s] 34%|███▎ | 16193/48008 [55:25<1:48:13, 4.90it/s] 34%|███▎ | 16194/48008 [55:25<1:44:28, 5.07it/s] 34%|███▎ | 16195/48008 [55:25<1:43:21, 5.13it/s] 34%|███▎ | 16196/48008 [55:25<1:42:34, 5.17it/s] 34%|███▎ | 16197/48008 [55:26<1:42:14, 5.19it/s] 34%|███▎ | 16198/48008 [55:26<1:40:26, 5.28it/s] 34%|███▎ | 16199/48008 [55:26<1:39:41, 5.32it/s] 34%|███▎ | 16200/48008 [55:26<1:40:19, 5.28it/s] {'loss': 0.3024, 'grad_norm': 0.23046875, 'learning_rate': 0.0001325154140976504, 'epoch': 0.34} 34%|███▎ | 16200/48008 [55:26<1:40:19, 5.28it/s] 34%|███▎ | 16201/48008 [55:26<1:40:23, 5.28it/s] 34%|███▎ | 16202/48008 [55:27<1:39:50, 5.31it/s] 34%|███▍ | 16203/48008 [55:27<1:40:15, 5.29it/s] 34%|███▍ | 16204/48008 [55:27<1:40:42, 5.26it/s] 34%|███▍ | 16205/48008 [55:27<1:40:43, 5.26it/s] 34%|███▍ | 16206/48008 [55:27<1:42:36, 5.17it/s] 34%|███▍ | 16207/48008 [55:28<1:44:51, 5.05it/s] 34%|███▍ | 16208/48008 [55:28<1:43:58, 5.10it/s] 34%|███▍ | 16209/48008 [55:28<1:45:11, 5.04it/s] 34%|███▍ | 16210/48008 [55:28<2:02:14, 4.34it/s] 34%|███▍ | 16211/48008 [55:28<1:55:36, 4.58it/s] 34%|███▍ | 16212/48008 [55:29<1:53:16, 4.68it/s] 34%|███▍ | 16213/48008 [55:29<1:50:25, 4.80it/s] 34%|███▍ | 16214/48008 [55:29<2:18:47, 3.82it/s] 34%|███▍ | 16215/48008 [55:29<2:07:19, 4.16it/s] 34%|███▍ | 16216/48008 [55:30<2:01:36, 4.36it/s] 34%|███▍ | 16217/48008 [55:30<1:55:25, 4.59it/s] 34%|███▍ | 16218/48008 [55:30<1:51:21, 4.76it/s] 34%|███▍ | 16219/48008 [55:30<1:47:52, 4.91it/s] 34%|███▍ | 16220/48008 [55:30<1:46:06, 4.99it/s] 34%|███▍ | 16221/48008 [55:31<1:44:16, 5.08it/s] 34%|███▍ | 16222/48008 [55:31<1:43:18, 5.13it/s] 34%|███▍ | 16223/48008 [55:31<1:42:21, 5.18it/s] 34%|███▍ | 16224/48008 [55:31<1:41:31, 5.22it/s] 34%|███▍ | 16225/48008 [55:31<1:41:01, 5.24it/s] 34%|███▍ | 16226/48008 [55:32<1:40:58, 5.25it/s] 34%|███▍ | 16227/48008 [55:32<1:41:20, 5.23it/s] 34%|███▍ | 16228/48008 [55:32<1:41:35, 5.21it/s] 34%|███▍ | 16229/48008 [55:32<1:40:55, 5.25it/s] 34%|███▍ | 16230/48008 [55:32<1:39:28, 5.32it/s] 34%|███▍ | 16231/48008 [55:32<1:42:03, 5.19it/s] 34%|███▍ | 16232/48008 [55:33<1:41:44, 5.20it/s] 34%|███▍ | 16233/48008 [55:33<1:41:45, 5.20it/s] 34%|███▍ | 16234/48008 [55:33<1:41:34, 5.21it/s] 34%|███▍ | 16235/48008 [55:33<1:41:12, 5.23it/s] 34%|███▍ | 16236/48008 [55:33<1:41:05, 5.24it/s] 34%|███▍ | 16237/48008 [55:34<1:42:03, 5.19it/s] 34%|███▍ | 16238/48008 [55:34<1:41:56, 5.19it/s] 34%|███▍ | 16239/48008 [55:34<1:41:34, 5.21it/s] 34%|███▍ | 16240/48008 [55:34<1:41:22, 5.22it/s] 34%|███▍ | 16241/48008 [55:34<1:40:56, 5.24it/s] 34%|███▍ | 16242/48008 [55:35<1:41:27, 5.22it/s] 34%|███▍ | 16243/48008 [55:35<1:41:06, 5.24it/s] 34%|███▍ | 16244/48008 [55:36<3:31:13, 2.51it/s] 34%|███▍ | 16245/48008 [55:36<2:57:56, 2.98it/s] 34%|███▍ | 16246/48008 [55:36<2:34:36, 3.42it/s] 34%|███▍ | 16247/48008 [55:36<2:20:18, 3.77it/s] 34%|███▍ | 16248/48008 [55:36<2:08:47, 4.11it/s] 34%|███▍ | 16249/48008 [55:37<2:01:40, 4.35it/s] 34%|███▍ | 16250/48008 [55:37<1:56:23, 4.55it/s] {'loss': 0.3007, 'grad_norm': 0.2001953125, 'learning_rate': 0.0001323071154807532, 'epoch': 0.34} 34%|███▍ | 16250/48008 [55:37<1:56:23, 4.55it/s] 34%|███▍ | 16251/48008 [55:37<1:52:20, 4.71it/s] 34%|███▍ | 16252/48008 [55:37<1:56:15, 4.55it/s] 34%|███▍ | 16253/48008 [55:37<1:51:04, 4.76it/s] 34%|███▍ | 16254/48008 [55:38<1:47:24, 4.93it/s] 34%|███▍ | 16255/48008 [55:38<1:45:06, 5.03it/s] 34%|███▍ | 16256/48008 [55:38<1:44:14, 5.08it/s] 34%|███▍ | 16257/48008 [55:38<1:43:35, 5.11it/s] 34%|███▍ | 16258/48008 [55:38<1:44:28, 5.07it/s] 34%|███▍ | 16259/48008 [55:39<1:42:55, 5.14it/s] 34%|███▍ | 16260/48008 [55:39<1:42:12, 5.18it/s] 34%|███▍ | 16261/48008 [55:39<1:43:44, 5.10it/s] 34%|███▍ | 16262/48008 [55:39<1:43:10, 5.13it/s] 34%|███▍ | 16263/48008 [55:39<1:42:48, 5.15it/s] 34%|███▍ | 16264/48008 [55:40<1:42:24, 5.17it/s] 34%|███▍ | 16265/48008 [55:40<1:43:18, 5.12it/s] 34%|███▍ | 16266/48008 [55:40<1:43:50, 5.09it/s] 34%|███▍ | 16267/48008 [55:40<1:42:32, 5.16it/s] 34%|███▍ | 16268/48008 [55:40<1:41:45, 5.20it/s] 34%|███▍ | 16269/48008 [55:41<1:43:40, 5.10it/s] 34%|███▍ | 16270/48008 [55:41<1:43:25, 5.11it/s] 34%|███▍ | 16271/48008 [55:41<1:43:34, 5.11it/s] 34%|███▍ | 16272/48008 [55:41<1:42:31, 5.16it/s] 34%|███▍ | 16273/48008 [55:41<1:42:33, 5.16it/s] 34%|███▍ | 16274/48008 [55:42<1:41:39, 5.20it/s] 34%|███▍ | 16275/48008 [55:42<1:41:59, 5.19it/s] 34%|███▍ | 16276/48008 [55:42<1:41:08, 5.23it/s] 34%|███▍ | 16277/48008 [55:42<1:40:46, 5.25it/s] 34%|███▍ | 16278/48008 [55:42<1:43:30, 5.11it/s] 34%|███▍ | 16279/48008 [55:42<1:43:07, 5.13it/s] 34%|███▍ | 16280/48008 [55:43<1:42:00, 5.18it/s] 34%|███▍ | 16281/48008 [55:43<1:42:17, 5.17it/s] 34%|███▍ | 16282/48008 [55:43<1:40:59, 5.24it/s] 34%|███▍ | 16283/48008 [55:43<1:40:47, 5.25it/s] 34%|███▍ | 16284/48008 [55:43<1:40:34, 5.26it/s] 34%|███▍ | 16285/48008 [55:44<1:40:58, 5.24it/s] 34%|███▍ | 16286/48008 [55:44<1:40:57, 5.24it/s] 34%|███▍ | 16287/48008 [55:44<1:43:03, 5.13it/s] 34%|███▍ | 16288/48008 [55:44<1:43:26, 5.11it/s] 34%|███▍ | 16289/48008 [55:45<2:13:20, 3.96it/s] 34%|███▍ | 16290/48008 [55:45<2:03:12, 4.29it/s] 34%|███▍ | 16291/48008 [55:45<1:56:12, 4.55it/s] 34%|███▍ | 16292/48008 [55:45<1:53:36, 4.65it/s] 34%|███▍ | 16293/48008 [55:46<2:20:38, 3.76it/s] 34%|███▍ | 16294/48008 [55:46<2:09:22, 4.09it/s] 34%|███▍ | 16295/48008 [55:46<2:32:12, 3.47it/s] 34%|███▍ | 16296/48008 [55:46<2:16:44, 3.87it/s] 34%|███▍ | 16297/48008 [55:47<2:08:41, 4.11it/s] 34%|███▍ | 16298/48008 [55:47<2:00:33, 4.38it/s] 34%|███▍ | 16299/48008 [55:47<1:55:50, 4.56it/s] 34%|███▍ | 16300/48008 [55:47<1:51:41, 4.73it/s] {'loss': 0.3044, 'grad_norm': 0.392578125, 'learning_rate': 0.00013209881686385602, 'epoch': 0.34} 34%|███▍ | 16300/48008 [55:47<1:51:41, 4.73it/s] 34%|███▍ | 16301/48008 [55:47<1:48:41, 4.86it/s] 34%|███▍ | 16302/48008 [55:48<1:46:08, 4.98it/s] 34%|███▍ | 16303/48008 [55:48<1:44:45, 5.04it/s] 34%|███▍ | 16304/48008 [55:48<1:45:49, 4.99it/s] 34%|███▍ | 16305/48008 [55:48<1:45:32, 5.01it/s] 34%|███▍ | 16306/48008 [55:48<1:43:14, 5.12it/s] 34%|███▍ | 16307/48008 [55:49<1:43:42, 5.09it/s] 34%|███▍ | 16308/48008 [55:49<1:43:56, 5.08it/s] 34%|███▍ | 16309/48008 [55:49<1:43:38, 5.10it/s] 34%|███▍ | 16310/48008 [55:49<1:43:35, 5.10it/s] 34%|███▍ | 16311/48008 [55:49<1:43:49, 5.09it/s] 34%|███▍ | 16312/48008 [55:49<1:42:35, 5.15it/s] 34%|███▍ | 16313/48008 [55:50<1:41:58, 5.18it/s] 34%|███▍ | 16314/48008 [55:50<1:43:25, 5.11it/s] 34%|███▍ | 16315/48008 [55:50<1:43:00, 5.13it/s] 34%|███▍ | 16316/48008 [55:50<1:42:12, 5.17it/s] 34%|███▍ | 16317/48008 [55:50<1:40:45, 5.24it/s] 34%|███▍ | 16318/48008 [55:51<2:11:43, 4.01it/s] 34%|███▍ | 16319/48008 [55:51<2:04:08, 4.25it/s] 34%|███▍ | 16320/48008 [55:51<1:58:17, 4.46it/s] 34%|███▍ | 16321/48008 [55:51<1:53:16, 4.66it/s] 34%|███▍ | 16322/48008 [55:52<1:51:19, 4.74it/s] 34%|███▍ | 16323/48008 [55:52<1:48:02, 4.89it/s] 34%|███▍ | 16324/48008 [55:52<1:45:33, 5.00it/s] 34%|███▍ | 16325/48008 [55:52<2:02:50, 4.30it/s] 34%|███▍ | 16326/48008 [55:53<1:57:16, 4.50it/s] 34%|███▍ | 16327/48008 [55:53<1:54:10, 4.62it/s] 34%|███▍ | 16328/48008 [55:53<1:57:33, 4.49it/s] 34%|███▍ | 16329/48008 [55:53<1:59:44, 4.41it/s] 34%|███▍ | 16330/48008 [55:53<1:55:26, 4.57it/s] 34%|███▍ | 16331/48008 [55:54<1:51:20, 4.74it/s] 34%|███▍ | 16332/48008 [55:54<1:50:22, 4.78it/s] 34%|███▍ | 16333/48008 [55:54<1:47:05, 4.93it/s] 34%|███▍ | 16334/48008 [55:54<1:45:11, 5.02it/s] 34%|███▍ | 16335/48008 [55:54<1:44:12, 5.07it/s] 34%|███▍ | 16336/48008 [55:55<1:42:06, 5.17it/s] 34%|███▍ | 16337/48008 [55:55<1:42:02, 5.17it/s] 34%|███▍ | 16338/48008 [55:55<1:40:58, 5.23it/s] 34%|███▍ | 16339/48008 [55:55<1:40:13, 5.27it/s] 34%|███▍ | 16340/48008 [55:55<1:47:56, 4.89it/s] 34%|███▍ | 16341/48008 [55:56<1:45:59, 4.98it/s] 34%|███▍ | 16342/48008 [55:56<1:45:53, 4.98it/s] 34%|███▍ | 16343/48008 [55:56<1:44:13, 5.06it/s] 34%|███▍ | 16344/48008 [55:56<1:44:27, 5.05it/s] 34%|███▍ | 16345/48008 [55:56<1:41:48, 5.18it/s] 34%|███▍ | 16346/48008 [55:56<1:42:10, 5.16it/s] 34%|███▍ | 16347/48008 [55:57<1:40:48, 5.23it/s] 34%|███▍ | 16348/48008 [55:57<1:41:53, 5.18it/s] 34%|███▍ | 16349/48008 [55:57<1:41:28, 5.20it/s] 34%|███▍ | 16350/48008 [55:57<1:42:11, 5.16it/s] {'loss': 0.3011, 'grad_norm': 0.1435546875, 'learning_rate': 0.00013189051824695884, 'epoch': 0.34} 34%|███▍ | 16350/48008 [55:57<1:42:11, 5.16it/s] 34%|███▍ | 16351/48008 [55:57<1:42:58, 5.12it/s] 34%|███▍ | 16352/48008 [55:58<2:00:20, 4.38it/s] 34%|███▍ | 16353/48008 [55:58<1:53:22, 4.65it/s] 34%|███▍ | 16354/48008 [55:58<1:50:03, 4.79it/s] 34%|███▍ | 16355/48008 [55:58<1:48:45, 4.85it/s] 34%|███▍ | 16356/48008 [55:59<1:46:26, 4.96it/s] 34%|███▍ | 16357/48008 [55:59<2:02:59, 4.29it/s] 34%|███▍ | 16358/48008 [55:59<1:57:16, 4.50it/s] 34%|███▍ | 16359/48008 [55:59<1:52:17, 4.70it/s] 34%|███▍ | 16360/48008 [55:59<1:48:49, 4.85it/s] 34%|███▍ | 16361/48008 [56:00<1:49:09, 4.83it/s] 34%|███▍ | 16362/48008 [56:00<1:46:54, 4.93it/s] 34%|███▍ | 16363/48008 [56:00<1:47:45, 4.89it/s] 34%|███▍ | 16364/48008 [56:00<1:45:58, 4.98it/s] 34%|███▍ | 16365/48008 [56:00<1:44:42, 5.04it/s] 34%|███▍ | 16366/48008 [56:01<1:42:53, 5.13it/s] 34%|███▍ | 16367/48008 [56:01<1:42:01, 5.17it/s] 34%|███▍ | 16368/48008 [56:01<1:42:02, 5.17it/s] 34%|███▍ | 16369/48008 [56:01<2:12:22, 3.98it/s] 34%|███▍ | 16370/48008 [56:02<2:03:04, 4.28it/s] 34%|███▍ | 16371/48008 [56:02<1:56:22, 4.53it/s] 34%|███▍ | 16372/48008 [56:02<1:51:11, 4.74it/s] 34%|███▍ | 16373/48008 [56:02<1:48:12, 4.87it/s] 34%|███▍ | 16374/48008 [56:02<1:46:04, 4.97it/s] 34%|███▍ | 16375/48008 [56:03<1:45:59, 4.97it/s] 34%|███▍ | 16376/48008 [56:03<2:15:20, 3.90it/s] 34%|███▍ | 16377/48008 [56:03<2:06:09, 4.18it/s] 34%|███▍ | 16378/48008 [56:03<1:58:07, 4.46it/s] 34%|███▍ | 16379/48008 [56:03<1:53:00, 4.66it/s] 34%|███▍ | 16380/48008 [56:04<1:49:12, 4.83it/s] 34%|███▍ | 16381/48008 [56:04<1:48:37, 4.85it/s] 34%|███▍ | 16382/48008 [56:04<1:45:50, 4.98it/s] 34%|███▍ | 16383/48008 [56:04<1:47:22, 4.91it/s] 34%|███▍ | 16384/48008 [56:04<1:45:50, 4.98it/s] 34%|███▍ | 16385/48008 [56:05<1:43:21, 5.10it/s] 34%|███▍ | 16386/48008 [56:05<1:44:52, 5.03it/s] 34%|███▍ | 16387/48008 [56:05<1:43:09, 5.11it/s] 34%|███▍ | 16388/48008 [56:05<1:42:32, 5.14it/s] 34%|███▍ | 16389/48008 [56:05<1:42:01, 5.17it/s] 34%|███▍ | 16390/48008 [56:06<1:42:08, 5.16it/s] 34%|███▍ | 16391/48008 [56:06<1:41:29, 5.19it/s] 34%|███▍ | 16392/48008 [56:06<1:43:12, 5.11it/s] 34%|███▍ | 16393/48008 [56:06<1:44:06, 5.06it/s] 34%|███▍ | 16394/48008 [56:06<1:42:41, 5.13it/s] 34%|███▍ | 16395/48008 [56:07<1:42:13, 5.15it/s] 34%|███▍ | 16396/48008 [56:07<1:41:34, 5.19it/s] 34%|███▍ | 16397/48008 [56:07<1:41:11, 5.21it/s] 34%|███▍ | 16398/48008 [56:07<1:40:41, 5.23it/s] 34%|███▍ | 16399/48008 [56:07<1:40:01, 5.27it/s] 34%|███▍ | 16400/48008 [56:08<1:39:51, 5.28it/s] {'loss': 0.3023, 'grad_norm': 0.1494140625, 'learning_rate': 0.00013168221963006166, 'epoch': 0.34} 34%|███▍ | 16400/48008 [56:08<1:39:51, 5.28it/s] 34%|███▍ | 16401/48008 [56:08<1:40:20, 5.25it/s] 34%|███▍ | 16402/48008 [56:08<1:47:48, 4.89it/s] 34%|███▍ | 16403/48008 [56:08<1:48:48, 4.84it/s] 34%|███▍ | 16404/48008 [56:08<1:47:00, 4.92it/s] 34%|███▍ | 16405/48008 [56:09<1:44:12, 5.05it/s] 34%|███▍ | 16406/48008 [56:09<1:45:02, 5.01it/s] 34%|███▍ | 16407/48008 [56:09<1:43:50, 5.07it/s] 34%|███▍ | 16408/48008 [56:09<1:43:09, 5.11it/s] 34%|███▍ | 16409/48008 [56:09<1:42:24, 5.14it/s] 34%|███▍ | 16410/48008 [56:10<1:41:11, 5.20it/s] 34%|███▍ | 16411/48008 [56:10<1:40:18, 5.25it/s] 34%|███▍ | 16412/48008 [56:10<1:40:26, 5.24it/s] 34%|███▍ | 16413/48008 [56:10<1:40:19, 5.25it/s] 34%|███▍ | 16414/48008 [56:10<1:39:49, 5.27it/s] 34%|███▍ | 16415/48008 [56:11<1:40:47, 5.22it/s] 34%|███▍ | 16416/48008 [56:11<1:40:08, 5.26it/s] 34%|███▍ | 16417/48008 [56:11<1:40:07, 5.26it/s] 34%|███▍ | 16418/48008 [56:11<1:41:59, 5.16it/s] 34%|███▍ | 16419/48008 [56:11<1:42:47, 5.12it/s] 34%|███▍ | 16420/48008 [56:11<1:42:01, 5.16it/s] 34%|███▍ | 16421/48008 [56:12<1:41:30, 5.19it/s] 34%|███▍ | 16422/48008 [56:12<1:41:33, 5.18it/s] 34%|███▍ | 16423/48008 [56:12<1:40:55, 5.22it/s] 34%|███▍ | 16424/48008 [56:12<1:43:12, 5.10it/s] 34%|███▍ | 16425/48008 [56:12<1:43:37, 5.08it/s] 34%|███▍ | 16426/48008 [56:13<1:45:01, 5.01it/s] 34%|███▍ | 16427/48008 [56:13<1:43:56, 5.06it/s] 34%|███▍ | 16428/48008 [56:13<1:42:27, 5.14it/s] 34%|███▍ | 16429/48008 [56:13<1:41:16, 5.20it/s] 34%|███▍ | 16430/48008 [56:13<1:40:59, 5.21it/s] 34%|███▍ | 16431/48008 [56:14<1:40:01, 5.26it/s] 34%|███▍ | 16432/48008 [56:14<1:40:50, 5.22it/s] 34%|███▍ | 16433/48008 [56:14<1:40:32, 5.23it/s] 34%|███▍ | 16434/48008 [56:14<1:48:08, 4.87it/s] 34%|███▍ | 16435/48008 [56:14<1:48:48, 4.84it/s] 34%|███▍ | 16436/48008 [56:15<1:48:08, 4.87it/s] 34%|███▍ | 16437/48008 [56:15<1:46:59, 4.92it/s] 34%|███▍ | 16438/48008 [56:15<1:45:00, 5.01it/s] 34%|███▍ | 16439/48008 [56:15<1:42:29, 5.13it/s] 34%|███▍ | 16440/48008 [56:15<1:41:39, 5.18it/s] 34%|███▍ | 16441/48008 [56:16<1:42:52, 5.11it/s] 34%|███▍ | 16442/48008 [56:16<1:42:02, 5.16it/s] 34%|███▍ | 16443/48008 [56:16<1:40:41, 5.22it/s] 34%|███▍ | 16444/48008 [56:16<1:41:39, 5.18it/s] 34%|███▍ | 16445/48008 [56:16<1:59:32, 4.40it/s] 34%|███▍ | 16446/48008 [56:17<1:53:39, 4.63it/s] 34%|███▍ | 16447/48008 [56:17<1:49:47, 4.79it/s] 34%|███▍ | 16448/48008 [56:17<1:47:54, 4.87it/s] 34%|███▍ | 16449/48008 [56:17<1:45:25, 4.99it/s] 34%|███▍ | 16450/48008 [56:17<1:45:40, 4.98it/s] {'loss': 0.301, 'grad_norm': 0.435546875, 'learning_rate': 0.00013147392101316448, 'epoch': 0.34} 34%|███▍ | 16450/48008 [56:17<1:45:40, 4.98it/s] 34%|███▍ | 16451/48008 [56:18<1:44:06, 5.05it/s] 34%|███▍ | 16452/48008 [56:18<1:43:42, 5.07it/s] 34%|███▍ | 16453/48008 [56:18<1:42:34, 5.13it/s] 34%|███▍ | 16454/48008 [56:18<1:43:21, 5.09it/s] 34%|███▍ | 16455/48008 [56:18<1:42:24, 5.14it/s] 34%|███▍ | 16456/48008 [56:19<1:44:12, 5.05it/s] 34%|███▍ | 16457/48008 [56:19<1:43:03, 5.10it/s] 34%|███▍ | 16458/48008 [56:19<1:42:22, 5.14it/s] 34%|███▍ | 16459/48008 [56:19<1:41:38, 5.17it/s] 34%|███▍ | 16460/48008 [56:19<1:40:47, 5.22it/s] 34%|███▍ | 16461/48008 [56:20<1:40:03, 5.25it/s] 34%|███▍ | 16462/48008 [56:20<1:39:34, 5.28it/s] 34%|███▍ | 16463/48008 [56:20<1:40:51, 5.21it/s] 34%|███▍ | 16464/48008 [56:20<1:40:33, 5.23it/s] 34%|███▍ | 16465/48008 [56:20<1:42:18, 5.14it/s] 34%|███▍ | 16466/48008 [56:21<1:41:20, 5.19it/s] 34%|███▍ | 16467/48008 [56:21<1:41:06, 5.20it/s] 34%|███▍ | 16468/48008 [56:21<1:39:56, 5.26it/s] 34%|███▍ | 16469/48008 [56:21<1:39:08, 5.30it/s] 34%|███▍ | 16470/48008 [56:21<1:41:29, 5.18it/s] 34%|███▍ | 16471/48008 [56:21<1:40:23, 5.24it/s] 34%|███▍ | 16472/48008 [56:22<1:40:31, 5.23it/s] 34%|███▍ | 16473/48008 [56:22<1:40:13, 5.24it/s] 34%|███▍ | 16474/48008 [56:22<1:41:56, 5.16it/s] 34%|███▍ | 16475/48008 [56:22<1:41:00, 5.20it/s] 34%|███▍ | 16476/48008 [56:22<1:41:34, 5.17it/s] 34%|███▍ | 16477/48008 [56:23<1:44:06, 5.05it/s] 34%|███▍ | 16478/48008 [56:23<1:50:00, 4.78it/s] 34%|███▍ | 16479/48008 [56:23<1:48:17, 4.85it/s] 34%|███▍ | 16480/48008 [56:23<1:46:37, 4.93it/s] 34%|███▍ | 16481/48008 [56:23<1:44:29, 5.03it/s] 34%|███▍ | 16482/48008 [56:24<1:44:58, 5.01it/s] 34%|███▍ | 16483/48008 [56:24<1:44:38, 5.02it/s] 34%|███▍ | 16484/48008 [56:24<1:43:27, 5.08it/s] 34%|███▍ | 16485/48008 [56:24<1:41:51, 5.16it/s] 34%|███▍ | 16486/48008 [56:24<1:40:52, 5.21it/s] 34%|███▍ | 16487/48008 [56:25<1:48:09, 4.86it/s] 34%|███▍ | 16488/48008 [56:25<1:48:47, 4.83it/s] 34%|███▍ | 16489/48008 [56:25<1:46:50, 4.92it/s] 34%|███▍ | 16490/48008 [56:25<1:47:48, 4.87it/s] 34%|███▍ | 16491/48008 [56:25<1:45:07, 5.00it/s] 34%|███▍ | 16492/48008 [56:26<1:42:52, 5.11it/s] 34%|███▍ | 16493/48008 [56:26<1:44:22, 5.03it/s] 34%|███▍ | 16494/48008 [56:26<1:43:28, 5.08it/s] 34%|███▍ | 16495/48008 [56:26<1:42:15, 5.14it/s] 34%|███▍ | 16496/48008 [56:26<1:41:06, 5.19it/s] 34%|███▍ | 16497/48008 [56:27<1:40:07, 5.24it/s] 34%|███▍ | 16498/48008 [56:27<1:40:47, 5.21it/s] 34%|███▍ | 16499/48008 [56:27<1:39:30, 5.28it/s] 34%|███▍ | 16500/48008 [56:27<1:40:01, 5.25it/s] {'loss': 0.3038, 'grad_norm': 0.15234375, 'learning_rate': 0.0001312656223962673, 'epoch': 0.34} 34%|███▍ | 16500/48008 [56:27<1:40:01, 5.25it/s] 34%|███▍ | 16501/48008 [56:27<1:41:20, 5.18it/s] 34%|███▍ | 16502/48008 [56:28<1:42:39, 5.11it/s] 34%|███▍ | 16503/48008 [56:28<1:43:29, 5.07it/s] 34%|███▍ | 16504/48008 [56:28<1:42:29, 5.12it/s] 34%|███▍ | 16505/48008 [56:28<1:41:57, 5.15it/s] 34%|███▍ | 16506/48008 [56:28<1:41:34, 5.17it/s] 34%|███▍ | 16507/48008 [56:29<1:41:23, 5.18it/s] 34%|███▍ | 16508/48008 [56:29<1:41:18, 5.18it/s] 34%|███▍ | 16509/48008 [56:29<1:48:39, 4.83it/s] 34%|███▍ | 16510/48008 [56:29<1:47:20, 4.89it/s] 34%|███▍ | 16511/48008 [56:29<1:46:49, 4.91it/s] 34%|███▍ | 16512/48008 [56:30<1:44:47, 5.01it/s] 34%|███▍ | 16513/48008 [56:30<1:45:20, 4.98it/s] 34%|███▍ | 16514/48008 [56:30<1:45:24, 4.98it/s] 34%|███▍ | 16515/48008 [56:30<1:43:44, 5.06it/s] 34%|███▍ | 16516/48008 [56:30<1:42:06, 5.14it/s] 34%|███▍ | 16517/48008 [56:31<1:42:25, 5.12it/s] 34%|███▍ | 16518/48008 [56:31<1:43:42, 5.06it/s] 34%|███▍ | 16519/48008 [56:31<1:42:42, 5.11it/s] 34%|███▍ | 16520/48008 [56:31<1:42:02, 5.14it/s] 34%|███▍ | 16521/48008 [56:31<1:41:11, 5.19it/s] 34%|███▍ | 16522/48008 [56:32<1:42:59, 5.10it/s] 34%|███▍ | 16523/48008 [56:32<1:42:02, 5.14it/s] 34%|███▍ | 16524/48008 [56:32<1:43:45, 5.06it/s] 34%|███▍ | 16525/48008 [56:32<1:42:27, 5.12it/s] 34%|███▍ | 16526/48008 [56:32<1:42:50, 5.10it/s] 34%|███▍ | 16527/48008 [56:33<1:41:58, 5.15it/s] 34%|███▍ | 16528/48008 [56:33<1:43:31, 5.07it/s] 34%|███▍ | 16529/48008 [56:33<1:42:25, 5.12it/s] 34%|███▍ | 16530/48008 [56:33<1:44:09, 5.04it/s] 34%|███▍ | 16531/48008 [56:33<1:42:45, 5.11it/s] 34%|███▍ | 16532/48008 [56:34<1:42:41, 5.11it/s] 34%|███▍ | 16533/48008 [56:34<1:41:18, 5.18it/s] 34%|███▍ | 16534/48008 [56:34<1:40:15, 5.23it/s] 34%|███▍ | 16535/48008 [56:34<1:40:19, 5.23it/s] 34%|███▍ | 16536/48008 [56:34<1:42:39, 5.11it/s] 34%|███▍ | 16537/48008 [56:34<1:41:45, 5.15it/s] 34%|███▍ | 16538/48008 [56:35<1:40:59, 5.19it/s] 34%|███▍ | 16539/48008 [56:35<1:40:42, 5.21it/s] 34%|███▍ | 16540/48008 [56:35<1:40:35, 5.21it/s] 34%|███▍ | 16541/48008 [56:35<1:58:26, 4.43it/s] 34%|███▍ | 16542/48008 [56:36<1:53:04, 4.64it/s] 34%|███▍ | 16543/48008 [56:36<1:51:12, 4.72it/s] 34%|███▍ | 16544/48008 [56:36<1:47:06, 4.90it/s] 34%|███▍ | 16545/48008 [56:36<1:48:06, 4.85it/s] 34%|███▍ | 16546/48008 [56:36<1:45:32, 4.97it/s] 34%|███▍ | 16547/48008 [56:37<1:45:40, 4.96it/s] 34%|███▍ | 16548/48008 [56:37<1:46:59, 4.90it/s] 34%|███▍ | 16549/48008 [56:37<1:45:04, 4.99it/s] 34%|███▍ | 16550/48008 [56:37<1:43:11, 5.08it/s] {'loss': 0.3016, 'grad_norm': 0.1982421875, 'learning_rate': 0.0001310573237793701, 'epoch': 0.34} 34%|███▍ | 16550/48008 [56:37<1:43:11, 5.08it/s] 34%|███▍ | 16551/48008 [56:37<1:42:25, 5.12it/s] 34%|███▍ | 16552/48008 [56:38<1:41:30, 5.16it/s] 34%|███▍ | 16553/48008 [56:38<1:40:55, 5.19it/s] 34%|███▍ | 16554/48008 [56:38<1:42:42, 5.10it/s] 34%|███▍ | 16555/48008 [56:38<1:41:28, 5.17it/s] 34%|███▍ | 16556/48008 [56:38<1:40:47, 5.20it/s] 34%|███▍ | 16557/48008 [56:38<1:41:00, 5.19it/s] 34%|███▍ | 16558/48008 [56:39<1:41:23, 5.17it/s] 34%|███▍ | 16559/48008 [56:39<1:40:04, 5.24it/s] 34%|███▍ | 16560/48008 [56:39<1:43:20, 5.07it/s] 34%|███▍ | 16561/48008 [56:39<1:42:30, 5.11it/s] 34%|███▍ | 16562/48008 [56:39<1:41:38, 5.16it/s] 35%|███▍ | 16563/48008 [56:40<1:40:20, 5.22it/s] 35%|███▍ | 16564/48008 [56:40<1:38:35, 5.32it/s] 35%|███▍ | 16565/48008 [56:40<1:46:30, 4.92it/s] 35%|███▍ | 16566/48008 [56:40<1:44:31, 5.01it/s] 35%|███▍ | 16567/48008 [56:40<1:43:43, 5.05it/s] 35%|███▍ | 16568/48008 [56:41<1:43:11, 5.08it/s] 35%|███▍ | 16569/48008 [56:41<2:12:51, 3.94it/s] 35%|███▍ | 16570/48008 [56:41<2:03:13, 4.25it/s] 35%|███▍ | 16571/48008 [56:41<1:55:40, 4.53it/s] 35%|███▍ | 16572/48008 [56:42<1:52:03, 4.68it/s] 35%|███▍ | 16573/48008 [56:42<1:50:50, 4.73it/s] 35%|███▍ | 16574/48008 [56:42<1:48:24, 4.83it/s] 35%|███▍ | 16575/48008 [56:42<1:45:44, 4.95it/s] 35%|███▍ | 16576/48008 [56:42<1:44:13, 5.03it/s] 35%|███▍ | 16577/48008 [56:43<1:43:30, 5.06it/s] 35%|███▍ | 16578/48008 [56:43<1:42:00, 5.14it/s] 35%|███▍ | 16579/48008 [56:43<1:43:30, 5.06it/s] 35%|███▍ | 16580/48008 [56:43<1:42:20, 5.12it/s] 35%|███▍ | 16581/48008 [56:43<1:41:22, 5.17it/s] 35%|███▍ | 16582/48008 [56:44<1:43:05, 5.08it/s] 35%|███▍ | 16583/48008 [56:44<1:45:17, 4.97it/s] 35%|███▍ | 16584/48008 [56:44<1:43:06, 5.08it/s] 35%|███▍ | 16585/48008 [56:44<1:43:08, 5.08it/s] 35%|███▍ | 16586/48008 [56:44<1:41:48, 5.14it/s] 35%|███▍ | 16587/48008 [56:45<1:40:20, 5.22it/s] 35%|███▍ | 16588/48008 [56:45<1:39:37, 5.26it/s] 35%|███▍ | 16589/48008 [56:45<1:40:04, 5.23it/s] 35%|███▍ | 16590/48008 [56:45<1:39:32, 5.26it/s] 35%|███▍ | 16591/48008 [56:45<1:47:05, 4.89it/s] 35%|███▍ | 16592/48008 [56:46<1:45:51, 4.95it/s] 35%|███▍ | 16593/48008 [56:46<1:44:10, 5.03it/s] 35%|███▍ | 16594/48008 [56:46<1:42:19, 5.12it/s] 35%|███▍ | 16595/48008 [56:46<1:40:49, 5.19it/s] 35%|███▍ | 16596/48008 [56:46<1:40:24, 5.21it/s] 35%|███▍ | 16597/48008 [56:46<1:39:15, 5.27it/s] 35%|███▍ | 16598/48008 [56:47<1:40:39, 5.20it/s] 35%|███▍ | 16599/48008 [56:47<1:41:54, 5.14it/s] 35%|███▍ | 16600/48008 [56:47<1:40:49, 5.19it/s] {'loss': 0.3024, 'grad_norm': 0.2109375, 'learning_rate': 0.00013084902516247293, 'epoch': 0.35} 35%|███▍ | 16600/48008 [56:47<1:40:49, 5.19it/s] 35%|███▍ | 16601/48008 [56:47<1:41:59, 5.13it/s] 35%|███▍ | 16602/48008 [56:47<1:41:18, 5.17it/s] 35%|███▍ | 16603/48008 [56:48<1:59:09, 4.39it/s] 35%|███▍ | 16604/48008 [56:48<2:00:25, 4.35it/s] 35%|███▍ | 16605/48008 [56:48<1:55:12, 4.54it/s] 35%|███▍ | 16606/48008 [56:48<1:50:56, 4.72it/s] 35%|███▍ | 16607/48008 [56:49<1:47:56, 4.85it/s] 35%|███▍ | 16608/48008 [56:49<1:45:22, 4.97it/s] 35%|███▍ | 16609/48008 [56:49<1:43:53, 5.04it/s] 35%|███▍ | 16610/48008 [56:49<1:42:44, 5.09it/s] 35%|███▍ | 16611/48008 [56:49<1:41:44, 5.14it/s] 35%|███▍ | 16612/48008 [56:50<1:41:06, 5.17it/s] 35%|███▍ | 16613/48008 [56:50<1:41:08, 5.17it/s] 35%|███▍ | 16614/48008 [56:50<1:41:52, 5.14it/s] 35%|███▍ | 16615/48008 [56:50<1:48:22, 4.83it/s] 35%|███▍ | 16616/48008 [56:50<1:45:42, 4.95it/s] 35%|███▍ | 16617/48008 [56:51<1:43:23, 5.06it/s] 35%|███▍ | 16618/48008 [56:51<1:43:17, 5.07it/s] 35%|███▍ | 16619/48008 [56:51<1:42:21, 5.11it/s] 35%|███▍ | 16620/48008 [56:51<1:41:35, 5.15it/s] 35%|███▍ | 16621/48008 [56:52<3:49:50, 2.28it/s] 35%|███▍ | 16622/48008 [56:52<3:11:26, 2.73it/s] 35%|███▍ | 16623/48008 [56:52<2:43:52, 3.19it/s] 35%|███▍ | 16624/48008 [56:53<2:24:26, 3.62it/s] 35%|███▍ | 16625/48008 [56:53<2:12:32, 3.95it/s] 35%|███▍ | 16626/48008 [56:53<2:03:16, 4.24it/s] 35%|███▍ | 16627/48008 [56:54<3:45:41, 2.32it/s] 35%|███▍ | 16628/48008 [56:54<3:08:02, 2.78it/s] 35%|███▍ | 16629/48008 [56:54<2:41:54, 3.23it/s] 35%|███▍ | 16630/48008 [56:55<2:25:07, 3.60it/s] 35%|███▍ | 16631/48008 [56:55<2:11:44, 3.97it/s] 35%|███▍ | 16632/48008 [56:55<2:05:03, 4.18it/s] 35%|███▍ | 16633/48008 [56:56<4:31:24, 1.93it/s] 35%|███▍ | 16634/48008 [56:56<3:41:41, 2.36it/s] 35%|███▍ | 16635/48008 [56:57<3:06:55, 2.80it/s] 35%|███▍ | 16636/48008 [56:57<2:40:10, 3.26it/s] 35%|███▍ | 16637/48008 [56:57<2:21:17, 3.70it/s] 35%|███▍ | 16638/48008 [56:57<2:09:36, 4.03it/s] 35%|███▍ | 16639/48008 [56:58<3:35:57, 2.42it/s] 35%|███▍ | 16640/48008 [56:58<3:02:51, 2.86it/s] 35%|███▍ | 16641/48008 [56:58<2:39:49, 3.27it/s] 35%|███▍ | 16642/48008 [56:58<2:21:31, 3.69it/s] 35%|███▍ | 16643/48008 [56:59<2:10:58, 3.99it/s] 35%|███▍ | 16644/48008 [56:59<2:01:36, 4.30it/s] 35%|███▍ | 16645/48008 [56:59<1:56:53, 4.47it/s] 35%|███▍ | 16646/48008 [56:59<1:53:37, 4.60it/s] 35%|███▍ | 16647/48008 [57:00<2:20:09, 3.73it/s] 35%|███▍ | 16648/48008 [57:00<2:08:40, 4.06it/s] 35%|███▍ | 16649/48008 [57:00<1:59:58, 4.36it/s] 35%|███▍ | 16650/48008 [57:00<1:54:25, 4.57it/s] {'loss': 0.3025, 'grad_norm': 0.16015625, 'learning_rate': 0.00013064072654557575, 'epoch': 0.35} 35%|███▍ | 16650/48008 [57:00<1:54:25, 4.57it/s] 35%|███▍ | 16651/48008 [57:00<1:50:59, 4.71it/s] 35%|███▍ | 16652/48008 [57:01<1:47:59, 4.84it/s] 35%|███▍ | 16653/48008 [57:01<1:45:45, 4.94it/s] 35%|███▍ | 16654/48008 [57:01<1:47:26, 4.86it/s] 35%|███▍ | 16655/48008 [57:01<1:44:52, 4.98it/s] 35%|███▍ | 16656/48008 [57:01<1:45:05, 4.97it/s] 35%|███▍ | 16657/48008 [57:02<1:43:07, 5.07it/s] 35%|███▍ | 16658/48008 [57:02<1:42:56, 5.08it/s] 35%|███▍ | 16659/48008 [57:02<1:43:13, 5.06it/s] 35%|███▍ | 16660/48008 [57:02<1:44:55, 4.98it/s] 35%|███▍ | 16661/48008 [57:02<1:45:12, 4.97it/s] 35%|███▍ | 16662/48008 [57:03<1:43:04, 5.07it/s] 35%|███▍ | 16663/48008 [57:03<3:18:02, 2.64it/s] 35%|███▍ | 16664/48008 [57:04<2:48:28, 3.10it/s] 35%|███▍ | 16665/48008 [57:04<2:29:29, 3.49it/s] 35%|███▍ | 16666/48008 [57:04<2:14:39, 3.88it/s] 35%|███▍ | 16667/48008 [57:04<2:04:15, 4.20it/s] 35%|███▍ | 16668/48008 [57:04<1:56:48, 4.47it/s] 35%|███▍ | 16669/48008 [57:05<1:51:00, 4.71it/s] 35%|███▍ | 16670/48008 [57:05<1:47:44, 4.85it/s] 35%|███▍ | 16671/48008 [57:05<1:45:55, 4.93it/s] 35%|███▍ | 16672/48008 [57:05<1:43:46, 5.03it/s] 35%|███▍ | 16673/48008 [57:05<1:42:25, 5.10it/s] 35%|███▍ | 16674/48008 [57:06<1:41:06, 5.16it/s] 35%|███▍ | 16675/48008 [57:06<1:40:49, 5.18it/s] 35%|███▍ | 16676/48008 [57:06<1:39:34, 5.24it/s] 35%|███▍ | 16677/48008 [57:06<1:41:08, 5.16it/s] 35%|███▍ | 16678/48008 [57:06<1:40:46, 5.18it/s] 35%|███▍ | 16679/48008 [57:07<1:41:23, 5.15it/s] 35%|███▍ | 16680/48008 [57:07<1:41:05, 5.16it/s] 35%|███▍ | 16681/48008 [57:07<1:42:20, 5.10it/s] 35%|███▍ | 16682/48008 [57:07<1:43:18, 5.05it/s] 35%|███▍ | 16683/48008 [57:07<1:41:42, 5.13it/s] 35%|███▍ | 16684/48008 [57:07<1:40:48, 5.18it/s] 35%|███▍ | 16685/48008 [57:08<1:39:51, 5.23it/s] 35%|███▍ | 16686/48008 [57:08<1:46:55, 4.88it/s] 35%|███▍ | 16687/48008 [57:08<1:47:35, 4.85it/s] 35%|███▍ | 16688/48008 [57:08<1:45:09, 4.96it/s] 35%|███▍ | 16689/48008 [57:08<1:42:57, 5.07it/s] 35%|███▍ | 16690/48008 [57:09<1:41:50, 5.12it/s] 35%|███▍ | 16691/48008 [57:09<1:40:56, 5.17it/s] 35%|███▍ | 16692/48008 [57:09<1:41:52, 5.12it/s] 35%|███▍ | 16693/48008 [57:09<1:40:59, 5.17it/s] 35%|███▍ | 16694/48008 [57:09<1:40:35, 5.19it/s] 35%|███▍ | 16695/48008 [57:10<1:40:03, 5.22it/s] 35%|███▍ | 16696/48008 [57:10<2:10:43, 3.99it/s] 35%|███▍ | 16697/48008 [57:10<2:00:53, 4.32it/s] 35%|███▍ | 16698/48008 [57:10<1:54:12, 4.57it/s] 35%|███▍ | 16699/48008 [57:11<1:51:47, 4.67it/s] 35%|███▍ | 16700/48008 [57:11<1:50:04, 4.74it/s] {'loss': 0.3026, 'grad_norm': 0.4921875, 'learning_rate': 0.00013043242792867856, 'epoch': 0.35} 35%|███▍ | 16700/48008 [57:11<1:50:04, 4.74it/s] 35%|███▍ | 16701/48008 [57:11<1:46:51, 4.88it/s] 35%|███▍ | 16702/48008 [57:11<1:45:25, 4.95it/s] 35%|███▍ | 16703/48008 [57:11<1:44:27, 5.00it/s] 35%|███▍ | 16704/48008 [57:12<1:44:13, 5.01it/s] 35%|███▍ | 16705/48008 [57:12<1:42:01, 5.11it/s] 35%|███▍ | 16706/48008 [57:12<1:42:34, 5.09it/s] 35%|███▍ | 16707/48008 [57:12<1:42:09, 5.11it/s] 35%|███▍ | 16708/48008 [57:12<1:41:12, 5.15it/s] 35%|███▍ | 16709/48008 [57:13<1:40:15, 5.20it/s] 35%|███▍ | 16710/48008 [57:13<1:39:53, 5.22it/s] 35%|███▍ | 16711/48008 [57:13<1:39:32, 5.24it/s] 35%|███▍ | 16712/48008 [57:13<1:39:42, 5.23it/s] 35%|███▍ | 16713/48008 [57:13<1:41:36, 5.13it/s] 35%|███▍ | 16714/48008 [57:14<1:40:36, 5.18it/s] 35%|███▍ | 16715/48008 [57:14<1:40:27, 5.19it/s] 35%|███▍ | 16716/48008 [57:14<1:40:09, 5.21it/s] 35%|███▍ | 16717/48008 [57:14<1:41:03, 5.16it/s] 35%|███▍ | 16718/48008 [57:14<1:40:55, 5.17it/s] 35%|███▍ | 16719/48008 [57:14<1:39:55, 5.22it/s] 35%|███▍ | 16720/48008 [57:15<1:39:53, 5.22it/s] 35%|███▍ | 16721/48008 [57:15<1:39:29, 5.24it/s] 35%|███▍ | 16722/48008 [57:15<1:39:44, 5.23it/s] 35%|███▍ | 16723/48008 [57:15<1:39:21, 5.25it/s] 35%|███▍ | 16724/48008 [57:15<1:41:03, 5.16it/s] 35%|███▍ | 16725/48008 [57:16<1:40:08, 5.21it/s] 35%|███▍ | 16726/48008 [57:16<1:39:57, 5.22it/s] 35%|███▍ | 16727/48008 [57:16<1:39:34, 5.24it/s] 35%|███▍ | 16728/48008 [57:16<1:38:51, 5.27it/s] 35%|███▍ | 16729/48008 [57:16<1:38:26, 5.30it/s] 35%|███▍ | 16730/48008 [57:17<1:41:27, 5.14it/s] 35%|███▍ | 16731/48008 [57:17<1:42:27, 5.09it/s] 35%|███▍ | 16732/48008 [57:17<1:42:22, 5.09it/s] 35%|███▍ | 16733/48008 [57:17<1:59:38, 4.36it/s] 35%|███▍ | 16734/48008 [57:17<1:54:29, 4.55it/s] 35%|███▍ | 16735/48008 [57:18<1:50:11, 4.73it/s] 35%|███▍ | 16736/48008 [57:18<1:47:18, 4.86it/s] 35%|███▍ | 16737/48008 [57:18<1:45:08, 4.96it/s] 35%|███▍ | 16738/48008 [57:18<1:44:11, 5.00it/s] 35%|███▍ | 16739/48008 [57:18<1:42:25, 5.09it/s] 35%|███▍ | 16740/48008 [57:19<1:41:34, 5.13it/s] 35%|███▍ | 16741/48008 [57:19<1:42:00, 5.11it/s] 35%|███▍ | 16742/48008 [57:19<1:41:17, 5.14it/s] 35%|███▍ | 16743/48008 [57:19<1:40:50, 5.17it/s] 35%|███▍ | 16744/48008 [57:19<1:40:29, 5.19it/s] 35%|███▍ | 16745/48008 [57:20<1:40:18, 5.19it/s] 35%|███▍ | 16746/48008 [57:20<1:39:43, 5.22it/s] 35%|███▍ | 16747/48008 [57:20<1:41:16, 5.14it/s] 35%|███▍ | 16748/48008 [57:20<1:40:50, 5.17it/s] 35%|███▍ | 16749/48008 [57:20<1:42:14, 5.10it/s] 35%|███▍ | 16750/48008 [57:21<1:41:47, 5.12it/s] {'loss': 0.302, 'grad_norm': 0.201171875, 'learning_rate': 0.00013022412931178138, 'epoch': 0.35} 35%|███▍ | 16750/48008 [57:21<1:41:47, 5.12it/s] 35%|███▍ | 16751/48008 [57:21<1:41:10, 5.15it/s] 35%|███▍ | 16752/48008 [57:21<1:39:44, 5.22it/s] 35%|███▍ | 16753/48008 [57:21<1:47:13, 4.86it/s] 35%|███▍ | 16754/48008 [57:21<1:45:09, 4.95it/s] 35%|███▍ | 16755/48008 [57:22<1:50:38, 4.71it/s] 35%|███▍ | 16756/48008 [57:22<1:49:19, 4.76it/s] 35%|███▍ | 16757/48008 [57:22<1:48:14, 4.81it/s] 35%|███▍ | 16758/48008 [57:22<1:45:50, 4.92it/s] 35%|███▍ | 16759/48008 [57:22<1:43:31, 5.03it/s] 35%|███▍ | 16760/48008 [57:23<1:42:38, 5.07it/s] 35%|███▍ | 16761/48008 [57:23<1:41:44, 5.12it/s] 35%|███▍ | 16762/48008 [57:23<1:40:58, 5.16it/s] 35%|███▍ | 16763/48008 [57:23<1:42:10, 5.10it/s] 35%|███▍ | 16764/48008 [57:23<1:41:25, 5.13it/s] 35%|███▍ | 16765/48008 [57:24<1:41:37, 5.12it/s] 35%|███▍ | 16766/48008 [57:24<1:39:54, 5.21it/s] 35%|███▍ | 16767/48008 [57:24<1:41:33, 5.13it/s] 35%|███▍ | 16768/48008 [57:24<1:41:49, 5.11it/s] 35%|███▍ | 16769/48008 [57:24<1:41:22, 5.14it/s] 35%|███▍ | 16770/48008 [57:25<1:40:28, 5.18it/s] 35%|███▍ | 16771/48008 [57:25<1:39:36, 5.23it/s] 35%|███▍ | 16772/48008 [57:25<1:41:20, 5.14it/s] 35%|███▍ | 16773/48008 [57:25<1:40:55, 5.16it/s] 35%|███▍ | 16774/48008 [57:25<1:42:45, 5.07it/s] 35%|███▍ | 16775/48008 [57:26<1:42:39, 5.07it/s] 35%|███▍ | 16776/48008 [57:26<1:41:14, 5.14it/s] 35%|███▍ | 16777/48008 [57:26<1:40:31, 5.18it/s] 35%|███▍ | 16778/48008 [57:26<1:47:35, 4.84it/s] 35%|███▍ | 16779/48008 [57:26<1:44:57, 4.96it/s] 35%|███▍ | 16780/48008 [57:27<1:44:03, 5.00it/s] 35%|███▍ | 16781/48008 [57:27<1:44:30, 4.98it/s] 35%|███▍ | 16782/48008 [57:27<1:42:33, 5.07it/s] 35%|███▍ | 16783/48008 [57:27<1:43:12, 5.04it/s] 35%|███▍ | 16784/48008 [57:27<1:49:21, 4.76it/s] 35%|███▍ | 16785/48008 [57:28<1:53:59, 4.57it/s] 35%|███▍ | 16786/48008 [57:28<1:49:34, 4.75it/s] 35%|███▍ | 16787/48008 [57:28<1:46:14, 4.90it/s] 35%|███▍ | 16788/48008 [57:28<1:44:06, 5.00it/s] 35%|███▍ | 16789/48008 [57:28<1:42:17, 5.09it/s] 35%|███▍ | 16790/48008 [57:29<1:40:19, 5.19it/s] 35%|███▍ | 16791/48008 [57:29<1:39:35, 5.22it/s] 35%|███▍ | 16792/48008 [57:29<1:40:38, 5.17it/s] 35%|███▍ | 16793/48008 [57:29<1:47:48, 4.83it/s] 35%|███▍ | 16794/48008 [57:29<1:47:17, 4.85it/s] 35%|███▍ | 16795/48008 [57:30<1:45:58, 4.91it/s] 35%|███▍ | 16796/48008 [57:30<1:43:23, 5.03it/s] 35%|███▍ | 16797/48008 [57:30<1:41:37, 5.12it/s] 35%|███▍ | 16798/48008 [57:30<1:41:59, 5.10it/s] 35%|███▍ | 16799/48008 [57:30<1:41:24, 5.13it/s] 35%|███▍ | 16800/48008 [57:31<1:41:59, 5.10it/s] {'loss': 0.3012, 'grad_norm': 0.16015625, 'learning_rate': 0.0001300158306948842, 'epoch': 0.35} 35%|███▍ | 16800/48008 [57:31<1:41:59, 5.10it/s] 35%|███▍ | 16801/48008 [57:31<1:40:44, 5.16it/s] 35%|███▍ | 16802/48008 [57:31<1:40:03, 5.20it/s] 35%|███▌ | 16803/48008 [57:31<1:40:59, 5.15it/s] 35%|███▌ | 16804/48008 [57:31<1:39:39, 5.22it/s] 35%|███▌ | 16805/48008 [57:31<1:39:46, 5.21it/s] 35%|███▌ | 16806/48008 [57:32<1:39:35, 5.22it/s] 35%|███▌ | 16807/48008 [57:32<1:38:45, 5.27it/s] 35%|███▌ | 16808/48008 [57:32<1:39:08, 5.24it/s] 35%|███▌ | 16809/48008 [57:32<1:40:27, 5.18it/s] 35%|███▌ | 16810/48008 [57:32<1:39:34, 5.22it/s] 35%|███▌ | 16811/48008 [57:33<1:40:43, 5.16it/s] 35%|███▌ | 16812/48008 [57:33<1:40:29, 5.17it/s] 35%|███▌ | 16813/48008 [57:33<1:40:12, 5.19it/s] 35%|███▌ | 16814/48008 [57:33<1:47:02, 4.86it/s] 35%|███▌ | 16815/48008 [57:33<1:44:36, 4.97it/s] 35%|███▌ | 16816/48008 [57:34<1:42:54, 5.05it/s] 35%|███▌ | 16817/48008 [57:34<1:41:39, 5.11it/s] 35%|███▌ | 16818/48008 [57:34<1:41:00, 5.15it/s] 35%|███▌ | 16819/48008 [57:34<1:40:21, 5.18it/s] 35%|███▌ | 16820/48008 [57:34<1:39:46, 5.21it/s] 35%|███▌ | 16821/48008 [57:35<1:39:00, 5.25it/s] 35%|███▌ | 16822/48008 [57:35<1:39:09, 5.24it/s] 35%|███▌ | 16823/48008 [57:35<1:38:44, 5.26it/s] 35%|███▌ | 16824/48008 [57:35<1:38:56, 5.25it/s] 35%|███▌ | 16825/48008 [57:35<1:39:18, 5.23it/s] 35%|███▌ | 16826/48008 [57:36<1:38:56, 5.25it/s] 35%|███▌ | 16827/48008 [57:36<1:39:52, 5.20it/s] 35%|███▌ | 16828/48008 [57:36<1:39:24, 5.23it/s] 35%|███▌ | 16829/48008 [57:36<1:39:20, 5.23it/s] 35%|███▌ | 16830/48008 [57:36<1:40:10, 5.19it/s] 35%|███▌ | 16831/48008 [57:37<1:57:53, 4.41it/s] 35%|███▌ | 16832/48008 [57:37<1:53:05, 4.59it/s] 35%|███▌ | 16833/48008 [57:37<1:50:38, 4.70it/s] 35%|███▌ | 16834/48008 [57:37<1:47:41, 4.82it/s] 35%|███▌ | 16835/48008 [57:37<1:45:03, 4.95it/s] 35%|███▌ | 16836/48008 [57:38<1:44:04, 4.99it/s] 35%|███▌ | 16837/48008 [57:38<1:42:36, 5.06it/s] 35%|███▌ | 16838/48008 [57:38<1:43:39, 5.01it/s] 35%|███▌ | 16839/48008 [57:38<1:41:27, 5.12it/s] 35%|███▌ | 16840/48008 [57:38<1:40:22, 5.18it/s] 35%|███▌ | 16841/48008 [57:39<1:40:08, 5.19it/s] 35%|███▌ | 16842/48008 [57:39<1:39:24, 5.23it/s] 35%|███▌ | 16843/48008 [57:39<1:39:27, 5.22it/s] 35%|███▌ | 16844/48008 [57:39<1:38:33, 5.27it/s] 35%|███▌ | 16845/48008 [57:39<1:37:54, 5.31it/s] 35%|███▌ | 16846/48008 [57:40<1:39:17, 5.23it/s] 35%|███▌ | 16847/48008 [57:40<1:40:59, 5.14it/s] 35%|███▌ | 16848/48008 [57:40<1:41:30, 5.12it/s] 35%|███▌ | 16849/48008 [57:40<1:40:51, 5.15it/s] 35%|███▌ | 16850/48008 [57:40<1:41:42, 5.11it/s] {'loss': 0.3025, 'grad_norm': 0.2216796875, 'learning_rate': 0.00012980753207798702, 'epoch': 0.35} 35%|███▌ | 16850/48008 [57:40<1:41:42, 5.11it/s] 35%|███▌ | 16851/48008 [57:40<1:42:33, 5.06it/s] 35%|███▌ | 16852/48008 [57:41<1:42:10, 5.08it/s] 35%|███▌ | 16853/48008 [57:41<1:48:44, 4.77it/s] 35%|███▌ | 16854/48008 [57:41<1:52:53, 4.60it/s] 35%|███▌ | 16855/48008 [57:41<1:55:48, 4.48it/s] 35%|███▌ | 16856/48008 [57:42<1:49:52, 4.73it/s] 35%|███▌ | 16857/48008 [57:42<1:46:06, 4.89it/s] 35%|███▌ | 16858/48008 [57:42<1:43:48, 5.00it/s] 35%|███▌ | 16859/48008 [57:42<2:12:40, 3.91it/s] 35%|███▌ | 16860/48008 [57:43<2:04:36, 4.17it/s] 35%|███▌ | 16861/48008 [57:43<1:56:29, 4.46it/s] 35%|███▌ | 16862/48008 [57:43<1:50:30, 4.70it/s] 35%|███▌ | 16863/48008 [57:43<1:46:26, 4.88it/s] 35%|███▌ | 16864/48008 [57:43<1:44:18, 4.98it/s] 35%|███▌ | 16865/48008 [57:44<1:42:54, 5.04it/s] 35%|███▌ | 16866/48008 [57:44<1:42:11, 5.08it/s] 35%|███▌ | 16867/48008 [57:44<1:41:00, 5.14it/s] 35%|███▌ | 16868/48008 [57:44<1:41:32, 5.11it/s] 35%|███▌ | 16869/48008 [57:44<1:40:04, 5.19it/s] 35%|███▌ | 16870/48008 [57:44<1:39:38, 5.21it/s] 35%|███▌ | 16871/48008 [57:45<1:39:50, 5.20it/s] 35%|███▌ | 16872/48008 [57:45<1:41:23, 5.12it/s] 35%|███▌ | 16873/48008 [57:45<1:47:52, 4.81it/s] 35%|███▌ | 16874/48008 [57:45<1:47:52, 4.81it/s] 35%|███▌ | 16875/48008 [57:45<1:44:46, 4.95it/s] 35%|███▌ | 16876/48008 [57:46<1:42:19, 5.07it/s] 35%|███▌ | 16877/48008 [57:46<1:41:44, 5.10it/s] 35%|███▌ | 16878/48008 [57:46<1:41:10, 5.13it/s] 35%|███▌ | 16879/48008 [57:46<1:40:29, 5.16it/s] 35%|███▌ | 16880/48008 [57:46<1:40:12, 5.18it/s] 35%|███▌ | 16881/48008 [57:47<1:39:56, 5.19it/s] 35%|███▌ | 16882/48008 [57:47<1:38:58, 5.24it/s] 35%|███▌ | 16883/48008 [57:47<1:38:54, 5.25it/s] 35%|███▌ | 16884/48008 [57:47<1:40:14, 5.17it/s] 35%|███▌ | 16885/48008 [57:47<1:40:04, 5.18it/s] 35%|███▌ | 16886/48008 [57:48<2:10:08, 3.99it/s] 35%|███▌ | 16887/48008 [57:48<2:00:54, 4.29it/s] 35%|███▌ | 16888/48008 [57:48<1:53:58, 4.55it/s] 35%|███▌ | 16889/48008 [57:48<1:56:42, 4.44it/s] 35%|███▌ | 16890/48008 [57:49<1:50:44, 4.68it/s] 35%|███▌ | 16891/48008 [57:49<1:47:15, 4.84it/s] 35%|███▌ | 16892/48008 [57:49<1:44:37, 4.96it/s] 35%|███▌ | 16893/48008 [57:49<2:13:25, 3.89it/s] 35%|███▌ | 16894/48008 [57:50<2:02:18, 4.24it/s] 35%|███▌ | 16895/48008 [57:50<1:55:11, 4.50it/s] 35%|███▌ | 16896/48008 [57:50<1:51:10, 4.66it/s] 35%|███▌ | 16897/48008 [57:50<1:48:50, 4.76it/s] 35%|███▌ | 16898/48008 [57:50<1:45:59, 4.89it/s] 35%|███▌ | 16899/48008 [57:51<1:46:41, 4.86it/s] 35%|███▌ | 16900/48008 [57:51<1:46:15, 4.88it/s] {'loss': 0.2995, 'grad_norm': 0.2490234375, 'learning_rate': 0.00012959923346108983, 'epoch': 0.35} 35%|███▌ | 16900/48008 [57:51<1:46:15, 4.88it/s] 35%|███▌ | 16901/48008 [57:51<1:44:14, 4.97it/s] 35%|███▌ | 16902/48008 [57:51<1:42:59, 5.03it/s] 35%|███▌ | 16903/48008 [57:51<1:41:52, 5.09it/s] 35%|███▌ | 16904/48008 [57:52<1:42:51, 5.04it/s] 35%|███▌ | 16905/48008 [57:52<1:41:48, 5.09it/s] 35%|███▌ | 16906/48008 [57:52<1:42:50, 5.04it/s] 35%|███▌ | 16907/48008 [57:52<1:41:35, 5.10it/s] 35%|███▌ | 16908/48008 [57:52<1:40:22, 5.16it/s] 35%|███▌ | 16909/48008 [57:52<1:39:56, 5.19it/s] 35%|███▌ | 16910/48008 [57:53<1:39:09, 5.23it/s] 35%|███▌ | 16911/48008 [57:53<1:40:49, 5.14it/s] 35%|███▌ | 16912/48008 [57:53<1:39:36, 5.20it/s] 35%|███▌ | 16913/48008 [57:53<1:40:24, 5.16it/s] 35%|███▌ | 16914/48008 [57:53<1:41:42, 5.10it/s] 35%|███▌ | 16915/48008 [57:54<1:47:57, 4.80it/s] 35%|███▌ | 16916/48008 [57:54<1:45:43, 4.90it/s] 35%|███▌ | 16917/48008 [57:54<1:45:49, 4.90it/s] 35%|███▌ | 16918/48008 [57:54<1:43:36, 5.00it/s] 35%|███▌ | 16919/48008 [57:54<1:44:12, 4.97it/s] 35%|███▌ | 16920/48008 [57:55<1:42:06, 5.07it/s] 35%|███▌ | 16921/48008 [57:55<1:42:40, 5.05it/s] 35%|███▌ | 16922/48008 [57:55<1:44:13, 4.97it/s] 35%|███▌ | 16923/48008 [57:55<1:41:50, 5.09it/s] 35%|███▌ | 16924/48008 [57:55<1:41:36, 5.10it/s] 35%|███▌ | 16925/48008 [57:56<1:48:05, 4.79it/s] 35%|███▌ | 16926/48008 [57:56<1:44:50, 4.94it/s] 35%|███▌ | 16927/48008 [57:56<1:45:14, 4.92it/s] 35%|███▌ | 16928/48008 [57:56<1:42:57, 5.03it/s] 35%|███▌ | 16929/48008 [57:56<1:41:14, 5.12it/s] 35%|███▌ | 16930/48008 [57:57<1:41:37, 5.10it/s] 35%|███▌ | 16931/48008 [57:57<1:40:40, 5.14it/s] 35%|███▌ | 16932/48008 [57:57<1:47:23, 4.82it/s] 35%|███▌ | 16933/48008 [57:57<1:46:27, 4.87it/s] 35%|███▌ | 16934/48008 [57:57<1:45:07, 4.93it/s] 35%|███▌ | 16935/48008 [57:58<1:43:26, 5.01it/s] 35%|███▌ | 16936/48008 [57:58<1:41:37, 5.10it/s] 35%|███▌ | 16937/48008 [57:58<1:41:20, 5.11it/s] 35%|███▌ | 16938/48008 [57:58<1:41:34, 5.10it/s] 35%|███▌ | 16939/48008 [57:58<1:41:23, 5.11it/s] 35%|███▌ | 16940/48008 [57:59<1:40:14, 5.17it/s] 35%|███▌ | 16941/48008 [57:59<1:39:25, 5.21it/s] 35%|███▌ | 16942/48008 [57:59<1:39:35, 5.20it/s] 35%|███▌ | 16943/48008 [57:59<1:40:29, 5.15it/s] 35%|███▌ | 16944/48008 [57:59<1:39:35, 5.20it/s] 35%|███▌ | 16945/48008 [58:00<1:39:12, 5.22it/s] 35%|███▌ | 16946/48008 [58:00<1:38:53, 5.24it/s] 35%|███▌ | 16947/48008 [58:00<1:38:49, 5.24it/s] 35%|███▌ | 16948/48008 [58:00<1:38:57, 5.23it/s] 35%|███▌ | 16949/48008 [58:00<1:40:59, 5.13it/s] 35%|███▌ | 16950/48008 [58:01<1:42:00, 5.07it/s] {'loss': 0.3001, 'grad_norm': 0.640625, 'learning_rate': 0.00012939093484419265, 'epoch': 0.35} 35%|███▌ | 16950/48008 [58:01<1:42:00, 5.07it/s] 35%|███▌ | 16951/48008 [58:01<1:41:41, 5.09it/s] 35%|███▌ | 16952/48008 [58:01<1:42:00, 5.07it/s] 35%|███▌ | 16953/48008 [58:01<1:40:35, 5.15it/s] 35%|███▌ | 16954/48008 [58:01<1:41:28, 5.10it/s] 35%|███▌ | 16955/48008 [58:02<1:40:17, 5.16it/s] 35%|███▌ | 16956/48008 [58:02<1:40:47, 5.14it/s] 35%|███▌ | 16957/48008 [58:02<1:39:29, 5.20it/s] 35%|███▌ | 16958/48008 [58:02<1:39:50, 5.18it/s] 35%|███▌ | 16959/48008 [58:02<1:42:18, 5.06it/s] 35%|███▌ | 16960/48008 [58:03<1:41:33, 5.10it/s] 35%|███▌ | 16961/48008 [58:03<1:40:29, 5.15it/s] 35%|███▌ | 16962/48008 [58:03<1:43:25, 5.00it/s] 35%|███▌ | 16963/48008 [58:03<1:42:00, 5.07it/s] 35%|███▌ | 16964/48008 [58:03<1:41:06, 5.12it/s] 35%|███▌ | 16965/48008 [58:04<1:40:54, 5.13it/s] 35%|███▌ | 16966/48008 [58:04<1:39:36, 5.19it/s] 35%|███▌ | 16967/48008 [58:04<1:39:09, 5.22it/s] 35%|███▌ | 16968/48008 [58:04<1:38:44, 5.24it/s] 35%|███▌ | 16969/48008 [58:04<1:39:51, 5.18it/s] 35%|███▌ | 16970/48008 [58:04<1:40:59, 5.12it/s] 35%|███▌ | 16971/48008 [58:05<1:39:11, 5.22it/s] 35%|███▌ | 16972/48008 [58:05<1:38:42, 5.24it/s] 35%|███▌ | 16973/48008 [58:05<1:38:51, 5.23it/s] 35%|███▌ | 16974/48008 [58:05<1:38:18, 5.26it/s] 35%|███▌ | 16975/48008 [58:05<1:38:53, 5.23it/s] 35%|███▌ | 16976/48008 [58:06<1:40:03, 5.17it/s] 35%|███▌ | 16977/48008 [58:06<1:39:19, 5.21it/s] 35%|███▌ | 16978/48008 [58:06<1:38:51, 5.23it/s] 35%|███▌ | 16979/48008 [58:06<1:38:30, 5.25it/s] 35%|███▌ | 16980/48008 [58:06<1:38:41, 5.24it/s] 35%|███▌ | 16981/48008 [58:07<1:39:17, 5.21it/s] 35%|███▌ | 16982/48008 [58:07<1:39:26, 5.20it/s] 35%|███▌ | 16983/48008 [58:07<1:38:50, 5.23it/s] 35%|███▌ | 16984/48008 [58:07<1:38:12, 5.26it/s] 35%|███▌ | 16985/48008 [58:07<1:38:50, 5.23it/s] 35%|███▌ | 16986/48008 [58:08<1:39:49, 5.18it/s] 35%|███▌ | 16987/48008 [58:08<1:39:31, 5.20it/s] 35%|███▌ | 16988/48008 [58:08<1:39:00, 5.22it/s] 35%|███▌ | 16989/48008 [58:08<1:38:50, 5.23it/s] 35%|███▌ | 16990/48008 [58:08<1:40:47, 5.13it/s] 35%|███▌ | 16991/48008 [58:08<1:39:29, 5.20it/s] 35%|███▌ | 16992/48008 [58:09<1:38:30, 5.25it/s] 35%|███▌ | 16993/48008 [58:09<1:37:08, 5.32it/s] 35%|███▌ | 16994/48008 [58:09<1:37:31, 5.30it/s] 35%|███▌ | 16995/48008 [58:09<1:37:49, 5.28it/s] 35%|███▌ | 16996/48008 [58:09<1:39:37, 5.19it/s] 35%|███▌ | 16997/48008 [58:10<1:39:04, 5.22it/s] 35%|███▌ | 16998/48008 [58:10<1:39:32, 5.19it/s] 35%|███▌ | 16999/48008 [58:10<1:39:07, 5.21it/s] 35%|███▌ | 17000/48008 [58:10<1:40:52, 5.12it/s] {'loss': 0.3028, 'grad_norm': 0.1591796875, 'learning_rate': 0.00012918263622729544, 'epoch': 0.35} 35%|███▌ | 17000/48008 [58:10<1:40:52, 5.12it/s] 35%|███▌ | 17001/48008 [58:10<1:42:09, 5.06it/s] 35%|███▌ | 17002/48008 [58:11<1:40:46, 5.13it/s] 35%|███▌ | 17003/48008 [58:11<1:41:28, 5.09it/s] 35%|███▌ | 17004/48008 [58:11<1:40:40, 5.13it/s] 35%|███▌ | 17005/48008 [58:11<1:40:58, 5.12it/s] 35%|███▌ | 17006/48008 [58:11<1:39:33, 5.19it/s] 35%|███▌ | 17007/48008 [58:12<1:39:38, 5.19it/s] 35%|███▌ | 17008/48008 [58:12<1:39:03, 5.22it/s] 35%|███▌ | 17009/48008 [58:12<1:38:48, 5.23it/s] 35%|███▌ | 17010/48008 [58:12<1:38:25, 5.25it/s] 35%|███▌ | 17011/48008 [58:12<1:38:39, 5.24it/s] 35%|███▌ | 17012/48008 [58:13<1:40:45, 5.13it/s] 35%|███▌ | 17013/48008 [58:13<1:39:26, 5.19it/s] 35%|███▌ | 17014/48008 [58:13<1:41:14, 5.10it/s] 35%|███▌ | 17015/48008 [58:13<1:40:16, 5.15it/s] 35%|███▌ | 17016/48008 [58:13<1:40:44, 5.13it/s] 35%|███▌ | 17017/48008 [58:14<1:40:17, 5.15it/s] 35%|███▌ | 17018/48008 [58:14<1:39:55, 5.17it/s] 35%|███▌ | 17019/48008 [58:14<1:47:00, 4.83it/s] 35%|███▌ | 17020/48008 [58:14<1:47:16, 4.81it/s] 35%|███▌ | 17021/48008 [58:14<1:47:18, 4.81it/s] 35%|███▌ | 17022/48008 [58:15<1:45:18, 4.90it/s] 35%|███▌ | 17023/48008 [58:15<2:01:08, 4.26it/s] 35%|███▌ | 17024/48008 [58:15<2:01:29, 4.25it/s] 35%|███▌ | 17025/48008 [58:15<1:53:38, 4.54it/s] 35%|███▌ | 17026/48008 [58:15<1:48:57, 4.74it/s] 35%|███▌ | 17027/48008 [58:16<1:46:52, 4.83it/s] 35%|███▌ | 17028/48008 [58:16<1:46:11, 4.86it/s] 35%|███▌ | 17029/48008 [58:16<1:46:51, 4.83it/s] 35%|███▌ | 17030/48008 [58:16<1:44:05, 4.96it/s] 35%|███▌ | 17031/48008 [58:17<2:12:59, 3.88it/s] 35%|███▌ | 17032/48008 [58:17<2:02:07, 4.23it/s] 35%|███▌ | 17033/48008 [58:17<1:54:57, 4.49it/s] 35%|███▌ | 17034/48008 [58:17<1:50:54, 4.65it/s] 35%|███▌ | 17035/48008 [58:17<1:46:27, 4.85it/s] 35%|███▌ | 17036/48008 [58:18<1:44:02, 4.96it/s] 35%|███▌ | 17037/48008 [58:18<1:42:00, 5.06it/s] 35%|███▌ | 17038/48008 [58:18<1:40:18, 5.15it/s] 35%|███▌ | 17039/48008 [58:18<1:39:46, 5.17it/s] 35%|███▌ | 17040/48008 [58:18<1:38:47, 5.22it/s] 35%|███▌ | 17041/48008 [58:19<1:38:27, 5.24it/s] 35%|███▌ | 17042/48008 [58:19<1:38:43, 5.23it/s] 36%|███▌ | 17043/48008 [58:19<1:38:04, 5.26it/s] 36%|███▌ | 17044/48008 [58:19<1:38:38, 5.23it/s] 36%|███▌ | 17045/48008 [58:19<1:38:23, 5.24it/s] 36%|███▌ | 17046/48008 [58:20<1:38:23, 5.24it/s] 36%|███▌ | 17047/48008 [58:20<1:37:41, 5.28it/s] 36%|███▌ | 17048/48008 [58:20<1:37:44, 5.28it/s] 36%|███▌ | 17049/48008 [58:20<1:37:52, 5.27it/s] 36%|███▌ | 17050/48008 [58:20<1:37:45, 5.28it/s] {'loss': 0.3009, 'grad_norm': 0.15234375, 'learning_rate': 0.00012897433761039829, 'epoch': 0.36} 36%|███▌ | 17050/48008 [58:20<1:37:45, 5.28it/s] 36%|███▌ | 17051/48008 [58:20<1:38:16, 5.25it/s] 36%|███▌ | 17052/48008 [58:21<1:39:16, 5.20it/s] 36%|███▌ | 17053/48008 [58:21<1:40:18, 5.14it/s] 36%|███▌ | 17054/48008 [58:21<1:40:09, 5.15it/s] 36%|███▌ | 17055/48008 [58:21<1:39:09, 5.20it/s] 36%|███▌ | 17056/48008 [58:21<1:38:13, 5.25it/s] 36%|███▌ | 17057/48008 [58:22<1:38:11, 5.25it/s] 36%|███▌ | 17058/48008 [58:22<1:38:33, 5.23it/s] 36%|███▌ | 17059/48008 [58:22<1:38:22, 5.24it/s] 36%|███▌ | 17060/48008 [58:22<1:38:27, 5.24it/s] 36%|███▌ | 17061/48008 [58:22<1:38:25, 5.24it/s] 36%|███▌ | 17062/48008 [58:23<1:40:32, 5.13it/s] 36%|███▌ | 17063/48008 [58:23<1:39:13, 5.20it/s] 36%|███▌ | 17064/48008 [58:23<1:38:08, 5.25it/s] 36%|███▌ | 17065/48008 [58:23<1:38:19, 5.24it/s] 36%|███▌ | 17066/48008 [58:23<1:41:26, 5.08it/s] 36%|███▌ | 17067/48008 [58:24<1:40:22, 5.14it/s] 36%|███▌ | 17068/48008 [58:24<1:40:17, 5.14it/s] 36%|███▌ | 17069/48008 [58:24<1:39:12, 5.20it/s] 36%|███▌ | 17070/48008 [58:24<1:39:09, 5.20it/s] 36%|███▌ | 17071/48008 [58:25<3:29:59, 2.46it/s] 36%|███▌ | 17072/48008 [58:25<2:57:58, 2.90it/s] 36%|███▌ | 17073/48008 [58:25<2:34:10, 3.34it/s] 36%|███▌ | 17074/48008 [58:26<2:17:34, 3.75it/s] 36%|███▌ | 17075/48008 [58:26<2:13:01, 3.88it/s] 36%|███▌ | 17076/48008 [58:26<2:05:31, 4.11it/s] 36%|███▌ | 17077/48008 [58:26<1:57:18, 4.39it/s] 36%|███▌ | 17078/48008 [58:26<1:51:09, 4.64it/s] 36%|███▌ | 17079/48008 [58:27<1:47:12, 4.81it/s] 36%|███▌ | 17080/48008 [58:27<1:44:02, 4.95it/s] 36%|███▌ | 17081/48008 [58:27<1:41:50, 5.06it/s] 36%|███▌ | 17082/48008 [58:27<1:40:32, 5.13it/s] 36%|███▌ | 17083/48008 [58:27<1:41:14, 5.09it/s] 36%|███▌ | 17084/48008 [58:28<1:41:24, 5.08it/s] 36%|███▌ | 17085/48008 [58:28<1:58:02, 4.37it/s] 36%|███▌ | 17086/48008 [58:28<1:52:55, 4.56it/s] 36%|███▌ | 17087/48008 [58:28<1:47:52, 4.78it/s] 36%|███▌ | 17088/48008 [58:28<1:44:57, 4.91it/s] 36%|███▌ | 17089/48008 [58:29<1:42:56, 5.01it/s] 36%|███▌ | 17090/48008 [58:29<1:43:35, 4.97it/s] 36%|███▌ | 17091/48008 [58:29<1:43:52, 4.96it/s] 36%|███▌ | 17092/48008 [58:29<1:42:08, 5.04it/s] 36%|███▌ | 17093/48008 [58:29<1:42:50, 5.01it/s] 36%|███▌ | 17094/48008 [58:30<1:41:53, 5.06it/s] 36%|███▌ | 17095/48008 [58:30<1:42:24, 5.03it/s] 36%|███▌ | 17096/48008 [58:30<1:41:10, 5.09it/s] 36%|███▌ | 17097/48008 [58:30<1:40:38, 5.12it/s] 36%|███▌ | 17098/48008 [58:30<1:41:36, 5.07it/s] 36%|███▌ | 17099/48008 [58:31<1:39:52, 5.16it/s] 36%|███▌ | 17100/48008 [58:31<2:09:53, 3.97it/s] {'loss': 0.3002, 'grad_norm': 0.232421875, 'learning_rate': 0.00012876603899350108, 'epoch': 0.36} 36%|███▌ | 17100/48008 [58:31<2:09:53, 3.97it/s] 36%|███▌ | 17101/48008 [58:31<1:59:52, 4.30it/s] 36%|███▌ | 17102/48008 [58:31<1:52:52, 4.56it/s] 36%|███▌ | 17103/48008 [58:32<1:50:28, 4.66it/s] 36%|███▌ | 17104/48008 [58:32<1:47:04, 4.81it/s] 36%|███▌ | 17105/48008 [58:32<1:44:08, 4.95it/s] 36%|███▌ | 17106/48008 [58:32<2:00:03, 4.29it/s] 36%|███▌ | 17107/48008 [58:32<1:55:00, 4.48it/s] 36%|███▌ | 17108/48008 [58:33<1:49:47, 4.69it/s] 36%|███▌ | 17109/48008 [58:33<1:45:53, 4.86it/s] 36%|███▌ | 17110/48008 [58:33<1:46:36, 4.83it/s] 36%|███▌ | 17111/48008 [58:33<1:44:02, 4.95it/s] 36%|███▌ | 17112/48008 [58:33<1:42:24, 5.03it/s] 36%|███▌ | 17113/48008 [58:34<1:40:47, 5.11it/s] 36%|███▌ | 17114/48008 [58:34<1:39:02, 5.20it/s] 36%|███▌ | 17115/48008 [58:34<1:38:32, 5.22it/s] 36%|███▌ | 17116/48008 [58:34<1:37:44, 5.27it/s] 36%|███▌ | 17117/48008 [58:34<1:38:45, 5.21it/s] 36%|███▌ | 17118/48008 [58:35<1:37:56, 5.26it/s] 36%|███▌ | 17119/48008 [58:35<1:38:10, 5.24it/s] 36%|███▌ | 17120/48008 [58:35<1:38:28, 5.23it/s] 36%|███▌ | 17121/48008 [58:35<1:37:58, 5.25it/s] 36%|███▌ | 17122/48008 [58:35<1:39:53, 5.15it/s] 36%|███▌ | 17123/48008 [58:36<1:39:26, 5.18it/s] 36%|███▌ | 17124/48008 [58:36<1:39:03, 5.20it/s] 36%|███▌ | 17125/48008 [58:36<1:38:34, 5.22it/s] 36%|███▌ | 17126/48008 [58:36<1:40:24, 5.13it/s] 36%|███▌ | 17127/48008 [58:36<1:39:42, 5.16it/s] 36%|███▌ | 17128/48008 [58:37<1:41:04, 5.09it/s] 36%|███▌ | 17129/48008 [58:37<1:38:56, 5.20it/s] 36%|███▌ | 17130/48008 [58:37<1:40:34, 5.12it/s] 36%|███▌ | 17131/48008 [58:37<1:38:35, 5.22it/s] 36%|███▌ | 17132/48008 [58:37<1:38:37, 5.22it/s] 36%|███▌ | 17133/48008 [58:37<1:40:35, 5.12it/s] 36%|███▌ | 17134/48008 [58:38<1:39:38, 5.16it/s] 36%|███▌ | 17135/48008 [58:38<1:38:30, 5.22it/s] 36%|███▌ | 17136/48008 [58:38<1:38:44, 5.21it/s] 36%|███▌ | 17137/48008 [58:38<1:38:22, 5.23it/s] 36%|███▌ | 17138/48008 [58:38<1:38:36, 5.22it/s] 36%|███▌ | 17139/48008 [58:39<1:38:33, 5.22it/s] 36%|███▌ | 17140/48008 [58:39<1:39:26, 5.17it/s] 36%|███▌ | 17141/48008 [58:39<1:38:25, 5.23it/s] 36%|███▌ | 17142/48008 [58:39<1:37:28, 5.28it/s] 36%|███▌ | 17143/48008 [58:39<1:37:14, 5.29it/s] 36%|███▌ | 17144/48008 [58:40<1:37:37, 5.27it/s] 36%|███▌ | 17145/48008 [58:40<1:37:01, 5.30it/s] 36%|███▌ | 17146/48008 [58:40<1:37:41, 5.27it/s] 36%|███▌ | 17147/48008 [58:40<1:37:54, 5.25it/s] 36%|███▌ | 17148/48008 [58:40<1:40:28, 5.12it/s] 36%|███▌ | 17149/48008 [58:41<1:40:49, 5.10it/s] 36%|███▌ | 17150/48008 [58:41<1:41:13, 5.08it/s] {'loss': 0.3031, 'grad_norm': 0.2470703125, 'learning_rate': 0.00012855774037660392, 'epoch': 0.36} 36%|███▌ | 17150/48008 [58:41<1:41:13, 5.08it/s] 36%|███▌ | 17151/48008 [58:41<1:40:44, 5.11it/s] 36%|███▌ | 17152/48008 [58:41<1:41:50, 5.05it/s] 36%|███▌ | 17153/48008 [58:41<1:47:43, 4.77it/s] 36%|███▌ | 17154/48008 [58:42<1:44:39, 4.91it/s] 36%|███▌ | 17155/48008 [58:42<1:42:39, 5.01it/s] 36%|███▌ | 17156/48008 [58:42<1:44:37, 4.91it/s] 36%|███▌ | 17157/48008 [58:42<1:42:29, 5.02it/s] 36%|███▌ | 17158/48008 [58:42<1:41:50, 5.05it/s] 36%|███▌ | 17159/48008 [58:43<1:39:47, 5.15it/s] 36%|███▌ | 17160/48008 [58:43<1:39:26, 5.17it/s] 36%|███▌ | 17161/48008 [58:43<1:38:58, 5.19it/s] 36%|███▌ | 17162/48008 [58:43<1:38:32, 5.22it/s] 36%|███▌ | 17163/48008 [58:43<1:39:08, 5.19it/s] 36%|███▌ | 17164/48008 [58:43<1:38:29, 5.22it/s] 36%|███▌ | 17165/48008 [58:44<1:40:00, 5.14it/s] 36%|███▌ | 17166/48008 [58:44<1:39:37, 5.16it/s] 36%|███▌ | 17167/48008 [58:44<1:39:12, 5.18it/s] 36%|███▌ | 17168/48008 [58:44<1:39:11, 5.18it/s] 36%|███▌ | 17169/48008 [58:44<1:38:33, 5.22it/s] 36%|███▌ | 17170/48008 [58:45<1:39:51, 5.15it/s] 36%|███▌ | 17171/48008 [58:45<1:42:17, 5.02it/s] 36%|███▌ | 17172/48008 [58:45<1:40:07, 5.13it/s] 36%|███▌ | 17173/48008 [58:45<1:46:34, 4.82it/s] 36%|███▌ | 17174/48008 [58:45<1:43:12, 4.98it/s] 36%|███▌ | 17175/48008 [58:46<1:41:18, 5.07it/s] 36%|███▌ | 17176/48008 [58:46<1:58:17, 4.34it/s] 36%|███▌ | 17177/48008 [58:46<1:53:23, 4.53it/s] 36%|███▌ | 17178/48008 [58:46<1:49:46, 4.68it/s] 36%|███▌ | 17179/48008 [58:47<1:46:16, 4.83it/s] 36%|███▌ | 17180/48008 [58:47<1:44:37, 4.91it/s] 36%|███▌ | 17181/48008 [58:47<1:42:56, 4.99it/s] 36%|███▌ | 17182/48008 [58:47<1:41:13, 5.08it/s] 36%|███▌ | 17183/48008 [58:47<1:40:45, 5.10it/s] 36%|███▌ | 17184/48008 [58:48<1:39:38, 5.16it/s] 36%|███▌ | 17185/48008 [58:48<1:39:04, 5.19it/s] 36%|███▌ | 17186/48008 [58:48<1:39:20, 5.17it/s] 36%|███▌ | 17187/48008 [58:48<1:40:33, 5.11it/s] 36%|███▌ | 17188/48008 [58:48<1:40:40, 5.10it/s] 36%|███▌ | 17189/48008 [58:49<1:40:01, 5.13it/s] 36%|███▌ | 17190/48008 [58:49<1:41:02, 5.08it/s] 36%|███▌ | 17191/48008 [58:49<1:40:00, 5.14it/s] 36%|███▌ | 17192/48008 [58:49<1:40:59, 5.09it/s] 36%|███▌ | 17193/48008 [58:49<1:39:43, 5.15it/s] 36%|███▌ | 17194/48008 [58:49<1:42:35, 5.01it/s] 36%|███▌ | 17195/48008 [58:50<1:48:16, 4.74it/s] 36%|███▌ | 17196/48008 [58:50<1:45:03, 4.89it/s] 36%|███▌ | 17197/48008 [58:50<1:42:28, 5.01it/s] 36%|███▌ | 17198/48008 [58:50<1:40:29, 5.11it/s] 36%|███▌ | 17199/48008 [58:50<1:39:32, 5.16it/s] 36%|███▌ | 17200/48008 [58:51<1:38:57, 5.19it/s] {'loss': 0.2995, 'grad_norm': 0.150390625, 'learning_rate': 0.0001283494417597067, 'epoch': 0.36} 36%|███▌ | 17200/48008 [58:51<1:38:57, 5.19it/s] 36%|███▌ | 17201/48008 [58:51<1:38:28, 5.21it/s] 36%|███▌ | 17202/48008 [58:51<1:37:51, 5.25it/s] 36%|███▌ | 17203/48008 [58:51<1:38:24, 5.22it/s] 36%|███▌ | 17204/48008 [58:51<1:37:29, 5.27it/s] 36%|███▌ | 17205/48008 [58:52<1:37:31, 5.26it/s] 36%|███▌ | 17206/48008 [58:52<1:37:48, 5.25it/s] 36%|███▌ | 17207/48008 [58:52<1:37:25, 5.27it/s] 36%|███▌ | 17208/48008 [58:52<1:36:39, 5.31it/s] 36%|███▌ | 17209/48008 [58:52<1:37:54, 5.24it/s] 36%|███▌ | 17210/48008 [58:53<1:40:42, 5.10it/s] 36%|███▌ | 17211/48008 [58:53<1:39:12, 5.17it/s] 36%|███▌ | 17212/48008 [58:53<1:38:24, 5.22it/s] 36%|███▌ | 17213/48008 [58:53<1:41:06, 5.08it/s] 36%|███▌ | 17214/48008 [58:53<1:40:27, 5.11it/s] 36%|███▌ | 17215/48008 [58:54<1:39:10, 5.17it/s] 36%|███▌ | 17216/48008 [58:54<1:38:33, 5.21it/s] 36%|███▌ | 17217/48008 [58:54<1:37:24, 5.27it/s] 36%|███▌ | 17218/48008 [58:54<1:39:40, 5.15it/s] 36%|███▌ | 17219/48008 [58:54<1:39:19, 5.17it/s] 36%|███▌ | 17220/48008 [58:55<1:38:20, 5.22it/s] 36%|███▌ | 17221/48008 [58:55<1:38:09, 5.23it/s] 36%|███▌ | 17222/48008 [58:55<1:37:48, 5.25it/s] 36%|███▌ | 17223/48008 [58:55<1:36:09, 5.34it/s] 36%|███▌ | 17224/48008 [58:55<1:38:19, 5.22it/s] 36%|███▌ | 17225/48008 [58:55<1:38:29, 5.21it/s] 36%|███▌ | 17226/48008 [58:56<1:56:18, 4.41it/s] 36%|███▌ | 17227/48008 [58:56<1:50:54, 4.63it/s] 36%|███▌ | 17228/48008 [58:56<1:54:07, 4.50it/s] 36%|███▌ | 17229/48008 [58:56<1:49:06, 4.70it/s] 36%|███▌ | 17230/48008 [58:57<1:45:40, 4.85it/s] 36%|███▌ | 17231/48008 [58:57<1:42:47, 4.99it/s] 36%|███▌ | 17232/48008 [58:57<1:41:32, 5.05it/s] 36%|███▌ | 17233/48008 [58:57<1:40:18, 5.11it/s] 36%|███▌ | 17234/48008 [58:58<3:42:09, 2.31it/s] 36%|███▌ | 17235/48008 [58:58<3:04:59, 2.77it/s] 36%|███▌ | 17236/48008 [58:59<2:39:05, 3.22it/s] 36%|███▌ | 17237/48008 [58:59<2:20:21, 3.65it/s] 36%|███▌ | 17238/48008 [58:59<2:08:16, 4.00it/s] 36%|███▌ | 17239/48008 [58:59<1:58:51, 4.31it/s] 36%|███▌ | 17240/48008 [59:00<3:38:24, 2.35it/s] 36%|███▌ | 17241/48008 [59:00<3:01:56, 2.82it/s] 36%|███▌ | 17242/48008 [59:00<2:36:41, 3.27it/s] 36%|███▌ | 17243/48008 [59:01<2:18:31, 3.70it/s] 36%|███▌ | 17244/48008 [59:01<2:06:12, 4.06it/s] 36%|███▌ | 17245/48008 [59:01<1:57:35, 4.36it/s] 36%|███▌ | 17246/48008 [59:02<4:22:53, 1.95it/s] 36%|███▌ | 17247/48008 [59:02<3:33:28, 2.40it/s] 36%|███▌ | 17248/48008 [59:02<2:58:24, 2.87it/s] 36%|███▌ | 17249/48008 [59:03<2:35:14, 3.30it/s] 36%|███▌ | 17250/48008 [59:03<2:17:14, 3.74it/s] {'loss': 0.3046, 'grad_norm': 0.1591796875, 'learning_rate': 0.00012814114314280953, 'epoch': 0.36} 36%|███▌ | 17250/48008 [59:03<2:17:14, 3.74it/s] 36%|███▌ | 17251/48008 [59:03<2:05:48, 4.07it/s] 36%|███▌ | 17252/48008 [59:04<3:32:04, 2.42it/s] 36%|███▌ | 17253/48008 [59:04<2:57:53, 2.88it/s] 36%|███▌ | 17254/48008 [59:04<2:33:24, 3.34it/s] 36%|███▌ | 17255/48008 [59:04<2:18:27, 3.70it/s] 36%|███▌ | 17256/48008 [59:05<2:13:46, 3.83it/s] 36%|███▌ | 17257/48008 [59:05<2:03:13, 4.16it/s] 36%|███▌ | 17258/48008 [59:05<1:55:54, 4.42it/s] 36%|███▌ | 17259/48008 [59:05<1:50:36, 4.63it/s] 36%|███▌ | 17260/48008 [59:05<1:46:36, 4.81it/s] 36%|███▌ | 17261/48008 [59:06<1:45:55, 4.84it/s] 36%|███▌ | 17262/48008 [59:06<2:01:13, 4.23it/s] 36%|███▌ | 17263/48008 [59:06<1:53:46, 4.50it/s] 36%|███▌ | 17264/48008 [59:06<1:48:02, 4.74it/s] 36%|███▌ | 17265/48008 [59:07<1:45:59, 4.83it/s] 36%|███▌ | 17266/48008 [59:07<1:43:10, 4.97it/s] 36%|███▌ | 17267/48008 [59:07<1:41:11, 5.06it/s] 36%|███▌ | 17268/48008 [59:07<1:40:18, 5.11it/s] 36%|███▌ | 17269/48008 [59:07<1:38:35, 5.20it/s] 36%|███▌ | 17270/48008 [59:07<1:38:31, 5.20it/s] 36%|███▌ | 17271/48008 [59:08<1:40:01, 5.12it/s] 36%|███▌ | 17272/48008 [59:08<1:39:20, 5.16it/s] 36%|███▌ | 17273/48008 [59:08<1:39:58, 5.12it/s] 36%|███▌ | 17274/48008 [59:08<1:41:05, 5.07it/s] 36%|███▌ | 17275/48008 [59:08<1:39:53, 5.13it/s] 36%|███▌ | 17276/48008 [59:09<3:15:24, 2.62it/s] 36%|███▌ | 17277/48008 [59:09<2:45:31, 3.09it/s] 36%|███▌ | 17278/48008 [59:10<2:25:19, 3.52it/s] 36%|███▌ | 17279/48008 [59:10<2:10:46, 3.92it/s] 36%|███▌ | 17280/48008 [59:10<2:30:50, 3.40it/s] 36%|███▌ | 17281/48008 [59:10<2:14:21, 3.81it/s] 36%|███▌ | 17282/48008 [59:11<2:03:06, 4.16it/s] 36%|███▌ | 17283/48008 [59:11<1:57:06, 4.37it/s] 36%|███▌ | 17284/48008 [59:11<1:51:06, 4.61it/s] 36%|███▌ | 17285/48008 [59:11<1:48:54, 4.70it/s] 36%|███▌ | 17286/48008 [59:11<1:45:13, 4.87it/s] 36%|███▌ | 17287/48008 [59:12<1:44:44, 4.89it/s] 36%|███▌ | 17288/48008 [59:12<1:41:59, 5.02it/s] 36%|███▌ | 17289/48008 [59:12<1:41:46, 5.03it/s] 36%|███▌ | 17290/48008 [59:12<1:40:21, 5.10it/s] 36%|███▌ | 17291/48008 [59:12<1:41:04, 5.06it/s] 36%|███▌ | 17292/48008 [59:13<1:42:11, 5.01it/s] 36%|███▌ | 17293/48008 [59:13<1:40:15, 5.11it/s] 36%|███▌ | 17294/48008 [59:13<1:40:46, 5.08it/s] 36%|███▌ | 17295/48008 [59:13<1:40:20, 5.10it/s] 36%|███▌ | 17296/48008 [59:13<1:39:22, 5.15it/s] 36%|███▌ | 17297/48008 [59:14<1:46:25, 4.81it/s] 36%|███▌ | 17298/48008 [59:14<1:45:46, 4.84it/s] 36%|███▌ | 17299/48008 [59:14<1:42:56, 4.97it/s] 36%|███▌ | 17300/48008 [59:14<1:41:10, 5.06it/s] {'loss': 0.2999, 'grad_norm': 0.15234375, 'learning_rate': 0.00012793284452591234, 'epoch': 0.36} 36%|███▌ | 17300/48008 [59:14<1:41:10, 5.06it/s] 36%|███▌ | 17301/48008 [59:14<1:41:43, 5.03it/s] 36%|███▌ | 17302/48008 [59:15<1:39:56, 5.12it/s] 36%|███▌ | 17303/48008 [59:15<1:39:49, 5.13it/s] 36%|███▌ | 17304/48008 [59:15<1:37:58, 5.22it/s] 36%|███▌ | 17305/48008 [59:15<1:37:54, 5.23it/s] 36%|███▌ | 17306/48008 [59:15<1:38:38, 5.19it/s] 36%|███▌ | 17307/48008 [59:16<1:38:34, 5.19it/s] 36%|███▌ | 17308/48008 [59:16<1:38:09, 5.21it/s] 36%|███▌ | 17309/48008 [59:16<1:37:57, 5.22it/s] 36%|███▌ | 17310/48008 [59:16<1:37:34, 5.24it/s] 36%|███▌ | 17311/48008 [59:16<1:39:21, 5.15it/s] 36%|███▌ | 17312/48008 [59:17<1:56:29, 4.39it/s] 36%|███▌ | 17313/48008 [59:17<1:53:34, 4.50it/s] 36%|███▌ | 17314/48008 [59:17<1:49:43, 4.66it/s] 36%|███▌ | 17315/48008 [59:17<1:45:34, 4.85it/s] 36%|███▌ | 17316/48008 [59:17<1:42:42, 4.98it/s] 36%|███▌ | 17317/48008 [59:18<1:40:53, 5.07it/s] 36%|███▌ | 17318/48008 [59:18<1:40:01, 5.11it/s] 36%|███▌ | 17319/48008 [59:18<1:39:17, 5.15it/s] 36%|███▌ | 17320/48008 [59:18<1:39:06, 5.16it/s] 36%|███▌ | 17321/48008 [59:18<1:40:33, 5.09it/s] 36%|███▌ | 17322/48008 [59:19<1:39:32, 5.14it/s] 36%|███▌ | 17323/48008 [59:19<1:37:53, 5.22it/s] 36%|███▌ | 17324/48008 [59:19<1:39:40, 5.13it/s] 36%|███▌ | 17325/48008 [59:19<1:39:48, 5.12it/s] 36%|███▌ | 17326/48008 [59:19<1:38:19, 5.20it/s] 36%|███▌ | 17327/48008 [59:19<1:38:02, 5.22it/s] 36%|███▌ | 17328/48008 [59:20<1:38:21, 5.20it/s] 36%|███▌ | 17329/48008 [59:20<1:38:45, 5.18it/s] 36%|███▌ | 17330/48008 [59:20<1:37:46, 5.23it/s] 36%|███▌ | 17331/48008 [59:20<1:38:49, 5.17it/s] 36%|███▌ | 17332/48008 [59:20<1:40:12, 5.10it/s] 36%|███▌ | 17333/48008 [59:21<1:39:33, 5.14it/s] 36%|███▌ | 17334/48008 [59:21<1:38:08, 5.21it/s] 36%|███▌ | 17335/48008 [59:21<1:38:10, 5.21it/s] 36%|███▌ | 17336/48008 [59:21<1:37:15, 5.26it/s] 36%|███▌ | 17337/48008 [59:21<1:37:29, 5.24it/s] 36%|███▌ | 17338/48008 [59:22<1:36:43, 5.28it/s] 36%|███▌ | 17339/48008 [59:22<1:39:47, 5.12it/s] 36%|███▌ | 17340/48008 [59:22<1:41:47, 5.02it/s] 36%|███▌ | 17341/48008 [59:22<1:40:21, 5.09it/s] 36%|███▌ | 17342/48008 [59:22<1:40:19, 5.09it/s] 36%|███▌ | 17343/48008 [59:23<2:09:29, 3.95it/s] 36%|███▌ | 17344/48008 [59:23<2:02:33, 4.17it/s] 36%|███▌ | 17345/48008 [59:23<1:54:11, 4.48it/s] 36%|███▌ | 17346/48008 [59:23<1:50:23, 4.63it/s] 36%|███▌ | 17347/48008 [59:24<1:46:15, 4.81it/s] 36%|███▌ | 17348/48008 [59:24<1:43:37, 4.93it/s] 36%|███▌ | 17349/48008 [59:24<1:43:42, 4.93it/s] 36%|███▌ | 17350/48008 [59:24<1:42:55, 4.96it/s] {'loss': 0.3022, 'grad_norm': 0.60546875, 'learning_rate': 0.00012772454590901516, 'epoch': 0.36} 36%|███▌ | 17350/48008 [59:24<1:42:55, 4.96it/s] 36%|███▌ | 17351/48008 [59:24<1:41:40, 5.03it/s] 36%|███▌ | 17352/48008 [59:25<1:42:09, 5.00it/s] 36%|███▌ | 17353/48008 [59:25<1:40:49, 5.07it/s] 36%|███▌ | 17354/48008 [59:25<1:39:52, 5.12it/s] 36%|███▌ | 17355/48008 [59:25<1:39:16, 5.15it/s] 36%|███▌ | 17356/48008 [59:25<1:39:23, 5.14it/s] 36%|███▌ | 17357/48008 [59:26<1:38:14, 5.20it/s] 36%|███▌ | 17358/48008 [59:26<1:41:01, 5.06it/s] 36%|███▌ | 17359/48008 [59:26<1:39:48, 5.12it/s] 36%|███▌ | 17360/48008 [59:26<1:39:18, 5.14it/s] 36%|███▌ | 17361/48008 [59:26<1:38:25, 5.19it/s] 36%|███▌ | 17362/48008 [59:26<1:39:44, 5.12it/s] 36%|███▌ | 17363/48008 [59:27<1:40:31, 5.08it/s] 36%|███▌ | 17364/48008 [59:27<1:39:18, 5.14it/s] 36%|███▌ | 17365/48008 [59:27<1:38:43, 5.17it/s] 36%|███▌ | 17366/48008 [59:27<1:38:25, 5.19it/s] 36%|███▌ | 17367/48008 [59:27<1:39:32, 5.13it/s] 36%|███▌ | 17368/48008 [59:28<1:39:49, 5.12it/s] 36%|███▌ | 17369/48008 [59:28<1:38:24, 5.19it/s] 36%|███▌ | 17370/48008 [59:28<1:37:30, 5.24it/s] 36%|███▌ | 17371/48008 [59:28<1:38:25, 5.19it/s] 36%|███▌ | 17372/48008 [59:28<1:38:09, 5.20it/s] 36%|███▌ | 17373/48008 [59:29<1:45:08, 4.86it/s] 36%|███▌ | 17374/48008 [59:29<1:42:32, 4.98it/s] 36%|███▌ | 17375/48008 [59:29<1:40:45, 5.07it/s] 36%|███▌ | 17376/48008 [59:29<1:39:14, 5.14it/s] 36%|███▌ | 17377/48008 [59:29<1:40:42, 5.07it/s] 36%|███▌ | 17378/48008 [59:30<1:57:11, 4.36it/s] 36%|███▌ | 17379/48008 [59:30<1:51:35, 4.57it/s] 36%|███▌ | 17380/48008 [59:30<1:46:58, 4.77it/s] 36%|███▌ | 17381/48008 [59:30<1:45:57, 4.82it/s] 36%|███▌ | 17382/48008 [59:31<1:43:38, 4.92it/s] 36%|███▌ | 17383/48008 [59:31<1:43:42, 4.92it/s] 36%|███▌ | 17384/48008 [59:31<1:43:48, 4.92it/s] 36%|███▌ | 17385/48008 [59:31<1:42:54, 4.96it/s] 36%|███▌ | 17386/48008 [59:31<1:43:47, 4.92it/s] 36%|███▌ | 17387/48008 [59:32<1:41:35, 5.02it/s] 36%|███▌ | 17388/48008 [59:32<1:40:11, 5.09it/s] 36%|███▌ | 17389/48008 [59:32<1:38:36, 5.18it/s] 36%|███▌ | 17390/48008 [59:32<1:40:00, 5.10it/s] 36%|███▌ | 17391/48008 [59:32<1:39:14, 5.14it/s] 36%|███▌ | 17392/48008 [59:32<1:39:36, 5.12it/s] 36%|███▌ | 17393/48008 [59:33<1:40:49, 5.06it/s] 36%|███▌ | 17394/48008 [59:33<1:39:28, 5.13it/s] 36%|███▌ | 17395/48008 [59:33<1:38:13, 5.19it/s] 36%|███▌ | 17396/48008 [59:33<1:37:37, 5.23it/s] 36%|███▌ | 17397/48008 [59:34<1:55:15, 4.43it/s] 36%|███▌ | 17398/48008 [59:34<1:49:17, 4.67it/s] 36%|███▌ | 17399/48008 [59:34<1:45:23, 4.84it/s] 36%|███▌ | 17400/48008 [59:34<1:49:53, 4.64it/s] {'loss': 0.3008, 'grad_norm': 0.453125, 'learning_rate': 0.000127516247292118, 'epoch': 0.36} 36%|███▌ | 17400/48008 [59:34<1:49:53, 4.64it/s] 36%|███▌ | 17401/48008 [59:34<1:47:29, 4.75it/s] 36%|███▌ | 17402/48008 [59:35<1:44:25, 4.88it/s] 36%|███▋ | 17403/48008 [59:35<1:41:55, 5.00it/s] 36%|███▋ | 17404/48008 [59:35<1:40:33, 5.07it/s] 36%|███▋ | 17405/48008 [59:35<2:09:27, 3.94it/s] 36%|███▋ | 17406/48008 [59:36<2:06:56, 4.02it/s] 36%|███▋ | 17407/48008 [59:36<1:56:34, 4.38it/s] 36%|███▋ | 17408/48008 [59:36<1:50:51, 4.60it/s] 36%|███▋ | 17409/48008 [59:36<1:46:33, 4.79it/s] 36%|███▋ | 17410/48008 [59:36<1:43:33, 4.92it/s] 36%|███▋ | 17411/48008 [59:36<1:41:35, 5.02it/s] 36%|███▋ | 17412/48008 [59:37<1:47:05, 4.76it/s] 36%|███▋ | 17413/48008 [59:37<1:43:50, 4.91it/s] 36%|███▋ | 17414/48008 [59:37<1:41:50, 5.01it/s] 36%|███▋ | 17415/48008 [59:37<1:40:15, 5.09it/s] 36%|███▋ | 17416/48008 [59:37<1:39:01, 5.15it/s] 36%|███▋ | 17417/48008 [59:38<1:39:55, 5.10it/s] 36%|███▋ | 17418/48008 [59:38<1:39:45, 5.11it/s] 36%|███▋ | 17419/48008 [59:38<1:38:19, 5.18it/s] 36%|███▋ | 17420/48008 [59:38<1:39:34, 5.12it/s] 36%|███▋ | 17421/48008 [59:38<1:38:46, 5.16it/s] 36%|███▋ | 17422/48008 [59:39<1:37:34, 5.22it/s] 36%|███▋ | 17423/48008 [59:39<1:39:19, 5.13it/s] 36%|███▋ | 17424/48008 [59:39<1:38:22, 5.18it/s] 36%|███▋ | 17425/48008 [59:39<1:37:04, 5.25it/s] 36%|███▋ | 17426/48008 [59:39<1:38:38, 5.17it/s] 36%|███▋ | 17427/48008 [59:40<1:37:39, 5.22it/s] 36%|███▋ | 17428/48008 [59:40<1:40:34, 5.07it/s] 36%|███▋ | 17429/48008 [59:40<1:42:18, 4.98it/s] 36%|███▋ | 17430/48008 [59:40<1:41:01, 5.04it/s] 36%|███▋ | 17431/48008 [59:40<1:39:25, 5.13it/s] 36%|███▋ | 17432/48008 [59:41<1:39:00, 5.15it/s] 36%|███▋ | 17433/48008 [59:41<1:39:11, 5.14it/s] 36%|███▋ | 17434/48008 [59:41<1:39:35, 5.12it/s] 36%|███▋ | 17435/48008 [59:41<1:40:32, 5.07it/s] 36%|███▋ | 17436/48008 [59:41<1:39:22, 5.13it/s] 36%|███▋ | 17437/48008 [59:42<1:40:33, 5.07it/s] 36%|███▋ | 17438/48008 [59:42<1:39:06, 5.14it/s] 36%|███▋ | 17439/48008 [59:42<1:38:07, 5.19it/s] 36%|███▋ | 17440/48008 [59:42<1:37:52, 5.21it/s] 36%|███▋ | 17441/48008 [59:42<1:37:43, 5.21it/s] 36%|███▋ | 17442/48008 [59:43<1:36:52, 5.26it/s] 36%|███▋ | 17443/48008 [59:43<1:37:39, 5.22it/s] 36%|███▋ | 17444/48008 [59:43<1:37:06, 5.25it/s] 36%|███▋ | 17445/48008 [59:43<1:39:03, 5.14it/s] 36%|███▋ | 17446/48008 [59:43<1:38:33, 5.17it/s] 36%|███▋ | 17447/48008 [59:44<1:38:04, 5.19it/s] 36%|███▋ | 17448/48008 [59:44<1:37:52, 5.20it/s] 36%|███▋ | 17449/48008 [59:44<1:38:15, 5.18it/s] 36%|███▋ | 17450/48008 [59:44<1:39:36, 5.11it/s] {'loss': 0.3016, 'grad_norm': 0.466796875, 'learning_rate': 0.0001273079486752208, 'epoch': 0.36} 36%|███▋ | 17450/48008 [59:44<1:39:36, 5.11it/s] 36%|███▋ | 17451/48008 [59:44<1:39:18, 5.13it/s] 36%|███▋ | 17452/48008 [59:44<1:39:48, 5.10it/s] 36%|███▋ | 17453/48008 [59:45<1:39:03, 5.14it/s] 36%|███▋ | 17454/48008 [59:45<1:55:53, 4.39it/s] 36%|███▋ | 17455/48008 [59:45<1:49:58, 4.63it/s] 36%|███▋ | 17456/48008 [59:45<1:45:58, 4.81it/s] 36%|███▋ | 17457/48008 [59:46<1:45:24, 4.83it/s] 36%|███▋ | 17458/48008 [59:46<1:42:21, 4.97it/s] 36%|███▋ | 17459/48008 [59:46<1:40:54, 5.05it/s] 36%|███▋ | 17460/48008 [59:46<1:41:38, 5.01it/s] 36%|███▋ | 17461/48008 [59:46<1:40:12, 5.08it/s] 36%|███▋ | 17462/48008 [59:47<1:38:43, 5.16it/s] 36%|███▋ | 17463/48008 [59:47<1:40:14, 5.08it/s] 36%|███▋ | 17464/48008 [59:47<1:39:11, 5.13it/s] 36%|███▋ | 17465/48008 [59:47<1:39:35, 5.11it/s] 36%|███▋ | 17466/48008 [59:47<1:39:26, 5.12it/s] 36%|███▋ | 17467/48008 [59:47<1:38:49, 5.15it/s] 36%|███▋ | 17468/48008 [59:48<1:38:14, 5.18it/s] 36%|███▋ | 17469/48008 [59:48<1:38:02, 5.19it/s] 36%|███▋ | 17470/48008 [59:48<1:37:47, 5.20it/s] 36%|███▋ | 17471/48008 [59:48<1:39:09, 5.13it/s] 36%|███▋ | 17472/48008 [59:48<1:40:17, 5.07it/s] 36%|███▋ | 17473/48008 [59:49<1:39:43, 5.10it/s] 36%|███▋ | 17474/48008 [59:49<1:40:48, 5.05it/s] 36%|███▋ | 17475/48008 [59:49<1:38:40, 5.16it/s] 36%|███▋ | 17476/48008 [59:49<1:38:24, 5.17it/s] 36%|███▋ | 17477/48008 [59:49<1:37:24, 5.22it/s] 36%|███▋ | 17478/48008 [59:50<1:37:32, 5.22it/s] 36%|███▋ | 17479/48008 [59:50<1:37:49, 5.20it/s] 36%|███▋ | 17480/48008 [59:50<1:38:26, 5.17it/s] 36%|███▋ | 17481/48008 [59:50<1:40:30, 5.06it/s] 36%|███▋ | 17482/48008 [59:50<1:39:38, 5.11it/s] 36%|███▋ | 17483/48008 [59:51<1:40:34, 5.06it/s] 36%|███▋ | 17484/48008 [59:51<1:39:05, 5.13it/s] 36%|███▋ | 17485/48008 [59:51<1:40:24, 5.07it/s] 36%|███▋ | 17486/48008 [59:51<1:42:30, 4.96it/s] 36%|███▋ | 17487/48008 [59:51<1:42:00, 4.99it/s] 36%|███▋ | 17488/48008 [59:52<1:39:52, 5.09it/s] 36%|███▋ | 17489/48008 [59:52<1:39:00, 5.14it/s] 36%|███▋ | 17490/48008 [59:52<1:37:56, 5.19it/s] 36%|███▋ | 17491/48008 [59:52<1:37:16, 5.23it/s] 36%|███▋ | 17492/48008 [59:52<1:37:10, 5.23it/s] 36%|███▋ | 17493/48008 [59:53<1:38:16, 5.18it/s] 36%|███▋ | 17494/48008 [59:53<1:38:17, 5.17it/s] 36%|███▋ | 17495/48008 [59:53<1:37:32, 5.21it/s] 36%|███▋ | 17496/48008 [59:53<1:39:24, 5.12it/s] 36%|███▋ | 17497/48008 [59:53<1:40:28, 5.06it/s] 36%|███▋ | 17498/48008 [59:54<1:40:00, 5.08it/s] 36%|███▋ | 17499/48008 [59:54<1:46:01, 4.80it/s] 36%|███▋ | 17500/48008 [59:54<1:44:23, 4.87it/s] {'loss': 0.296, 'grad_norm': 0.58203125, 'learning_rate': 0.00012709965005832361, 'epoch': 0.36} 36%|███▋ | 17500/48008 [59:54<1:44:23, 4.87it/s] 36%|███▋ | 17501/48008 [59:54<1:43:15, 4.92it/s] 36%|███▋ | 17502/48008 [59:54<1:42:16, 4.97it/s] 36%|███▋ | 17503/48008 [59:55<1:42:04, 4.98it/s] 36%|███▋ | 17504/48008 [59:55<1:40:20, 5.07it/s] 36%|███▋ | 17505/48008 [59:55<1:46:27, 4.78it/s] 36%|███▋ | 17506/48008 [59:55<1:43:09, 4.93it/s] 36%|███▋ | 17507/48008 [59:55<1:40:34, 5.05it/s] 36%|███▋ | 17508/48008 [59:56<1:42:05, 4.98it/s] 36%|███▋ | 17509/48008 [59:56<1:40:15, 5.07it/s] 36%|███▋ | 17510/48008 [59:56<1:39:13, 5.12it/s] 36%|███▋ | 17511/48008 [59:56<1:38:17, 5.17it/s] 36%|███▋ | 17512/48008 [59:56<1:38:51, 5.14it/s] 36%|███▋ | 17513/48008 [59:57<1:37:57, 5.19it/s] 36%|███▋ | 17514/48008 [59:57<1:37:47, 5.20it/s] 36%|███▋ | 17515/48008 [59:57<1:37:08, 5.23it/s] 36%|███▋ | 17516/48008 [59:57<1:44:43, 4.85it/s] 36%|███▋ | 17517/48008 [59:57<1:41:22, 5.01it/s] 36%|███▋ | 17518/48008 [59:58<1:41:19, 5.02it/s] 36%|███▋ | 17519/48008 [59:58<1:41:35, 5.00it/s] 36%|███▋ | 17520/48008 [59:58<1:40:22, 5.06it/s] 36%|███▋ | 17521/48008 [59:58<1:46:21, 4.78it/s] 36%|███▋ | 17522/48008 [59:58<1:45:21, 4.82it/s] 37%|███▋ | 17523/48008 [59:59<1:42:29, 4.96it/s] 37%|███▋ | 17524/48008 [59:59<1:41:07, 5.02it/s] 37%|███▋ | 17525/48008 [59:59<1:40:50, 5.04it/s] 37%|███▋ | 17526/48008 [59:59<1:39:44, 5.09it/s] 37%|███▋ | 17527/48008 [59:59<1:40:23, 5.06it/s] 37%|███▋ | 17528/48008 [1:00:00<1:40:17, 5.06it/s] 37%|███▋ | 17529/48008 [1:00:00<1:42:04, 4.98it/s] 37%|███▋ | 17530/48008 [1:00:00<1:40:30, 5.05it/s] 37%|███▋ | 17531/48008 [1:00:00<1:39:44, 5.09it/s] 37%|███▋ | 17532/48008 [1:00:00<1:40:34, 5.05it/s] 37%|███▋ | 17533/48008 [1:00:01<1:39:49, 5.09it/s] 37%|███▋ | 17534/48008 [1:00:01<1:40:33, 5.05it/s] 37%|███▋ | 17535/48008 [1:00:01<1:42:26, 4.96it/s] 37%|███▋ | 17536/48008 [1:00:01<1:41:30, 5.00it/s] 37%|███▋ | 17537/48008 [1:00:01<1:40:16, 5.06it/s] 37%|███▋ | 17538/48008 [1:00:02<1:46:26, 4.77it/s] 37%|███▋ | 17539/48008 [1:00:02<1:43:54, 4.89it/s] 37%|███▋ | 17540/48008 [1:00:02<1:42:06, 4.97it/s] 37%|███▋ | 17541/48008 [1:00:02<1:40:41, 5.04it/s] 37%|███▋ | 17542/48008 [1:00:02<1:39:14, 5.12it/s] 37%|███▋ | 17543/48008 [1:00:03<1:37:33, 5.20it/s] 37%|███▋ | 17544/48008 [1:00:03<1:37:15, 5.22it/s] 37%|███▋ | 17545/48008 [1:00:03<1:39:58, 5.08it/s] 37%|███▋ | 17546/48008 [1:00:03<1:39:02, 5.13it/s] 37%|███▋ | 17547/48008 [1:00:03<1:40:19, 5.06it/s] 37%|███▋ | 17548/48008 [1:00:04<1:39:52, 5.08it/s] 37%|███▋ | 17549/48008 [1:00:04<1:38:55, 5.13it/s] 37%|███▋ | 17550/48008 [1:00:04<1:38:19, 5.16it/s] {'loss': 0.2996, 'grad_norm': 0.443359375, 'learning_rate': 0.00012689135144142643, 'epoch': 0.37} 37%|███▋ | 17550/48008 [1:00:04<1:38:19, 5.16it/s] 37%|███▋ | 17551/48008 [1:00:04<1:39:59, 5.08it/s] 37%|███▋ | 17552/48008 [1:00:04<1:41:55, 4.98it/s] 37%|███▋ | 17553/48008 [1:00:04<1:40:44, 5.04it/s] 37%|███▋ | 17554/48008 [1:00:05<2:09:19, 3.92it/s] 37%|███▋ | 17555/48008 [1:00:05<2:02:46, 4.13it/s] 37%|███▋ | 17556/48008 [1:00:05<2:02:07, 4.16it/s] 37%|███▋ | 17557/48008 [1:00:06<1:55:12, 4.41it/s] 37%|███▋ | 17558/48008 [1:00:06<1:49:55, 4.62it/s] 37%|███▋ | 17559/48008 [1:00:06<1:48:06, 4.69it/s] 37%|███▋ | 17560/48008 [1:00:06<1:47:28, 4.72it/s] 37%|███▋ | 17561/48008 [1:00:06<1:46:53, 4.75it/s] 37%|███▋ | 17562/48008 [1:00:07<1:50:49, 4.58it/s] 37%|███▋ | 17563/48008 [1:00:07<1:47:16, 4.73it/s] 37%|███▋ | 17564/48008 [1:00:07<1:43:21, 4.91it/s] 37%|███▋ | 17565/48008 [1:00:07<1:43:20, 4.91it/s] 37%|███▋ | 17566/48008 [1:00:07<1:40:46, 5.04it/s] 37%|███▋ | 17567/48008 [1:00:08<1:39:24, 5.10it/s] 37%|███▋ | 17568/48008 [1:00:08<1:38:33, 5.15it/s] 37%|███▋ | 17569/48008 [1:00:08<1:55:20, 4.40it/s] 37%|███▋ | 17570/48008 [1:00:08<1:50:26, 4.59it/s] 37%|███▋ | 17571/48008 [1:00:08<1:47:05, 4.74it/s] 37%|███▋ | 17572/48008 [1:00:09<1:45:18, 4.82it/s] 37%|███▋ | 17573/48008 [1:00:09<1:44:27, 4.86it/s] 37%|███▋ | 17574/48008 [1:00:09<1:43:11, 4.92it/s] 37%|███▋ | 17575/48008 [1:00:09<1:43:52, 4.88it/s] 37%|███▋ | 17576/48008 [1:00:09<1:43:41, 4.89it/s] 37%|███▋ | 17577/48008 [1:00:10<1:43:23, 4.91it/s] 37%|███▋ | 17578/48008 [1:00:10<1:41:38, 4.99it/s] 37%|███▋ | 17579/48008 [1:00:10<1:41:18, 5.01it/s] 37%|███▋ | 17580/48008 [1:00:10<1:41:06, 5.02it/s] 37%|███▋ | 17581/48008 [1:00:10<1:41:26, 5.00it/s] 37%|███▋ | 17582/48008 [1:00:11<1:39:27, 5.10it/s] 37%|███▋ | 17583/48008 [1:00:11<1:39:47, 5.08it/s] 37%|███▋ | 17584/48008 [1:00:11<1:39:35, 5.09it/s] 37%|███▋ | 17585/48008 [1:00:11<1:37:50, 5.18it/s] 37%|███▋ | 17586/48008 [1:00:11<1:37:19, 5.21it/s] 37%|███▋ | 17587/48008 [1:00:12<1:37:09, 5.22it/s] 37%|███▋ | 17588/48008 [1:00:12<1:38:32, 5.15it/s] 37%|███▋ | 17589/48008 [1:00:12<1:39:10, 5.11it/s] 37%|███▋ | 17590/48008 [1:00:12<1:38:14, 5.16it/s] 37%|███▋ | 17591/48008 [1:00:12<1:40:25, 5.05it/s] 37%|███▋ | 17592/48008 [1:00:13<1:39:15, 5.11it/s] 37%|███▋ | 17593/48008 [1:00:13<1:40:25, 5.05it/s] 37%|███▋ | 17594/48008 [1:00:13<1:39:07, 5.11it/s] 37%|███▋ | 17595/48008 [1:00:13<1:38:47, 5.13it/s] 37%|███▋ | 17596/48008 [1:00:13<1:39:07, 5.11it/s] 37%|███▋ | 17597/48008 [1:00:14<1:39:06, 5.11it/s] 37%|███▋ | 17598/48008 [1:00:14<1:41:22, 5.00it/s] 37%|███▋ | 17599/48008 [1:00:14<1:41:19, 5.00it/s] 37%|███▋ | 17600/48008 [1:00:14<1:40:38, 5.04it/s] {'loss': 0.296, 'grad_norm': 0.2255859375, 'learning_rate': 0.00012668305282452925, 'epoch': 0.37} 37%|███▋ | 17600/48008 [1:00:14<1:40:38, 5.04it/s] 37%|███▋ | 17601/48008 [1:00:14<1:39:24, 5.10it/s] 37%|███▋ | 17602/48008 [1:00:15<1:37:49, 5.18it/s] 37%|███▋ | 17603/48008 [1:00:15<1:37:44, 5.18it/s] 37%|███▋ | 17604/48008 [1:00:15<1:37:24, 5.20it/s] 37%|███▋ | 17605/48008 [1:00:15<1:40:12, 5.06it/s] 37%|███▋ | 17606/48008 [1:00:15<1:39:00, 5.12it/s] 37%|███▋ | 17607/48008 [1:00:15<1:37:52, 5.18it/s] 37%|███▋ | 17608/48008 [1:00:16<1:36:48, 5.23it/s] 37%|███▋ | 17609/48008 [1:00:16<1:36:29, 5.25it/s] 37%|███▋ | 17610/48008 [1:00:16<1:38:35, 5.14it/s] 37%|███▋ | 17611/48008 [1:00:16<1:38:06, 5.16it/s] 37%|███▋ | 17612/48008 [1:00:16<1:37:19, 5.21it/s] 37%|███▋ | 17613/48008 [1:00:17<1:36:41, 5.24it/s] 37%|███▋ | 17614/48008 [1:00:17<1:37:33, 5.19it/s] 37%|███▋ | 17615/48008 [1:00:17<1:37:10, 5.21it/s] 37%|███▋ | 17616/48008 [1:00:17<1:38:49, 5.13it/s] 37%|███▋ | 17617/48008 [1:00:17<1:39:34, 5.09it/s] 37%|███▋ | 17618/48008 [1:00:18<1:38:55, 5.12it/s] 37%|███▋ | 17619/48008 [1:00:18<1:38:43, 5.13it/s] 37%|███▋ | 17620/48008 [1:00:18<1:37:48, 5.18it/s] 37%|███▋ | 17621/48008 [1:00:18<1:39:00, 5.12it/s] 37%|███▋ | 17622/48008 [1:00:18<1:38:46, 5.13it/s] 37%|███▋ | 17623/48008 [1:00:19<1:37:35, 5.19it/s] 37%|███▋ | 17624/48008 [1:00:19<1:39:09, 5.11it/s] 37%|███▋ | 17625/48008 [1:00:19<1:40:01, 5.06it/s] 37%|███▋ | 17626/48008 [1:00:19<1:40:41, 5.03it/s] 37%|███▋ | 17627/48008 [1:00:19<1:39:47, 5.07it/s] 37%|███▋ | 17628/48008 [1:00:20<1:38:29, 5.14it/s] 37%|███▋ | 17629/48008 [1:00:20<1:44:50, 4.83it/s] 37%|███▋ | 17630/48008 [1:00:20<1:42:13, 4.95it/s] 37%|███▋ | 17631/48008 [1:00:20<1:39:44, 5.08it/s] 37%|███▋ | 17632/48008 [1:00:20<1:38:05, 5.16it/s] 37%|███▋ | 17633/48008 [1:00:21<1:38:58, 5.11it/s] 37%|███▋ | 17634/48008 [1:00:21<1:39:19, 5.10it/s] 37%|███▋ | 17635/48008 [1:00:21<1:38:24, 5.14it/s] 37%|███▋ | 17636/48008 [1:00:21<1:38:49, 5.12it/s] 37%|███▋ | 17637/48008 [1:00:21<1:38:03, 5.16it/s] 37%|███▋ | 17638/48008 [1:00:22<1:36:22, 5.25it/s] 37%|███▋ | 17639/48008 [1:00:22<1:35:57, 5.27it/s] 37%|███▋ | 17640/48008 [1:00:22<1:35:50, 5.28it/s] 37%|███▋ | 17641/48008 [1:00:22<1:36:04, 5.27it/s] 37%|███▋ | 17642/48008 [1:00:22<1:36:21, 5.25it/s] 37%|███▋ | 17643/48008 [1:00:22<1:36:17, 5.26it/s] 37%|███▋ | 17644/48008 [1:00:23<1:38:02, 5.16it/s] 37%|███▋ | 17645/48008 [1:00:23<1:37:29, 5.19it/s] 37%|███▋ | 17646/48008 [1:00:23<1:36:19, 5.25it/s] 37%|███▋ | 17647/48008 [1:00:23<1:54:00, 4.44it/s] 37%|███▋ | 17648/48008 [1:00:24<1:48:56, 4.64it/s] 37%|███▋ | 17649/48008 [1:00:24<1:44:44, 4.83it/s] 37%|███▋ | 17650/48008 [1:00:24<1:42:06, 4.95it/s] {'loss': 0.3012, 'grad_norm': 0.1337890625, 'learning_rate': 0.00012647475420763207, 'epoch': 0.37} 37%|███▋ | 17650/48008 [1:00:24<1:42:06, 4.95it/s] 37%|███▋ | 17651/48008 [1:00:24<1:39:57, 5.06it/s] 37%|███▋ | 17652/48008 [1:00:24<1:41:42, 4.97it/s] 37%|███▋ | 17653/48008 [1:00:25<1:40:06, 5.05it/s] 37%|███▋ | 17654/48008 [1:00:25<1:40:36, 5.03it/s] 37%|███▋ | 17655/48008 [1:00:25<1:38:59, 5.11it/s] 37%|███▋ | 17656/48008 [1:00:25<1:39:04, 5.11it/s] 37%|███▋ | 17657/48008 [1:00:25<1:38:31, 5.13it/s] 37%|███▋ | 17658/48008 [1:00:25<1:36:52, 5.22it/s] 37%|███▋ | 17659/48008 [1:00:26<1:39:12, 5.10it/s] 37%|███▋ | 17660/48008 [1:00:26<1:38:27, 5.14it/s] 37%|███▋ | 17661/48008 [1:00:26<1:38:11, 5.15it/s] 37%|███▋ | 17662/48008 [1:00:26<1:40:24, 5.04it/s] 37%|███▋ | 17663/48008 [1:00:26<1:39:52, 5.06it/s] 37%|███▋ | 17664/48008 [1:00:27<1:38:02, 5.16it/s] 37%|███▋ | 17665/48008 [1:00:27<1:37:24, 5.19it/s] 37%|███▋ | 17666/48008 [1:00:27<1:38:51, 5.12it/s] 37%|███▋ | 17667/48008 [1:00:27<1:37:51, 5.17it/s] 37%|███▋ | 17668/48008 [1:00:27<1:39:19, 5.09it/s] 37%|███▋ | 17669/48008 [1:00:28<1:38:51, 5.11it/s] 37%|███▋ | 17670/48008 [1:00:28<1:38:20, 5.14it/s] 37%|███▋ | 17671/48008 [1:00:28<1:37:41, 5.18it/s] 37%|███▋ | 17672/48008 [1:00:28<1:35:40, 5.28it/s] 37%|███▋ | 17673/48008 [1:00:28<1:35:20, 5.30it/s] 37%|███▋ | 17674/48008 [1:00:29<1:35:04, 5.32it/s] 37%|███▋ | 17675/48008 [1:00:29<1:35:42, 5.28it/s] 37%|███▋ | 17676/48008 [1:00:29<1:36:10, 5.26it/s] 37%|███▋ | 17677/48008 [1:00:29<1:36:30, 5.24it/s] 37%|███▋ | 17678/48008 [1:00:29<1:38:21, 5.14it/s] 37%|███▋ | 17679/48008 [1:00:30<1:45:01, 4.81it/s] 37%|███▋ | 17680/48008 [1:00:30<1:44:32, 4.84it/s] 37%|███▋ | 17681/48008 [1:00:30<1:41:52, 4.96it/s] 37%|███▋ | 17682/48008 [1:00:30<1:47:20, 4.71it/s] 37%|███▋ | 17683/48008 [1:00:30<1:43:37, 4.88it/s] 37%|███▋ | 17684/48008 [1:00:31<1:40:31, 5.03it/s] 37%|███▋ | 17685/48008 [1:00:31<1:39:22, 5.09it/s] 37%|███▋ | 17686/48008 [1:00:31<1:38:58, 5.11it/s] 37%|███▋ | 17687/48008 [1:00:31<1:38:21, 5.14it/s] 37%|███▋ | 17688/48008 [1:00:31<1:37:40, 5.17it/s] 37%|███▋ | 17689/48008 [1:00:32<1:36:54, 5.21it/s] 37%|███▋ | 17690/48008 [1:00:32<1:36:37, 5.23it/s] 37%|███▋ | 17691/48008 [1:00:32<1:36:05, 5.26it/s] 37%|███▋ | 17692/48008 [1:00:32<1:38:17, 5.14it/s] 37%|███▋ | 17693/48008 [1:00:32<1:37:47, 5.17it/s] 37%|███▋ | 17694/48008 [1:00:33<1:38:32, 5.13it/s] 37%|███▋ | 17695/48008 [1:00:33<1:37:23, 5.19it/s] 37%|███▋ | 17696/48008 [1:00:33<1:37:02, 5.21it/s] 37%|███▋ | 17697/48008 [1:00:33<1:35:54, 5.27it/s] 37%|███▋ | 17698/48008 [1:00:33<1:35:58, 5.26it/s] 37%|███▋ | 17699/48008 [1:00:34<1:43:00, 4.90it/s] 37%|███▋ | 17700/48008 [1:00:34<1:42:51, 4.91it/s] {'loss': 0.3007, 'grad_norm': 0.380859375, 'learning_rate': 0.00012626645559073488, 'epoch': 0.37} 37%|███▋ | 17700/48008 [1:00:34<1:42:51, 4.91it/s] 37%|███▋ | 17701/48008 [1:00:34<1:42:11, 4.94it/s] 37%|███▋ | 17702/48008 [1:00:34<1:40:15, 5.04it/s] 37%|███▋ | 17703/48008 [1:00:34<1:39:02, 5.10it/s] 37%|███▋ | 17704/48008 [1:00:34<1:38:09, 5.15it/s] 37%|███▋ | 17705/48008 [1:00:35<1:38:27, 5.13it/s] 37%|███▋ | 17706/48008 [1:00:35<1:44:56, 4.81it/s] 37%|███▋ | 17707/48008 [1:00:35<1:41:56, 4.95it/s] 37%|███▋ | 17708/48008 [1:00:35<1:39:23, 5.08it/s] 37%|███▋ | 17709/48008 [1:00:35<1:38:25, 5.13it/s] 37%|███▋ | 17710/48008 [1:00:36<1:36:52, 5.21it/s] 37%|███▋ | 17711/48008 [1:00:36<1:36:05, 5.26it/s] 37%|███▋ | 17712/48008 [1:00:36<1:37:41, 5.17it/s] 37%|███▋ | 17713/48008 [1:00:36<1:38:30, 5.13it/s] 37%|███▋ | 17714/48008 [1:00:36<1:38:48, 5.11it/s] 37%|███▋ | 17715/48008 [1:00:37<1:38:55, 5.10it/s] 37%|███▋ | 17716/48008 [1:00:37<1:45:02, 4.81it/s] 37%|███▋ | 17717/48008 [1:00:37<1:42:48, 4.91it/s] 37%|███▋ | 17718/48008 [1:00:37<1:40:51, 5.01it/s] 37%|███▋ | 17719/48008 [1:00:37<1:39:15, 5.09it/s] 37%|███▋ | 17720/48008 [1:00:38<2:08:04, 3.94it/s] 37%|███▋ | 17721/48008 [1:00:38<2:01:10, 4.17it/s] 37%|███▋ | 17722/48008 [1:00:38<1:53:46, 4.44it/s] 37%|███▋ | 17723/48008 [1:00:38<1:47:52, 4.68it/s] 37%|███▋ | 17724/48008 [1:00:39<1:46:27, 4.74it/s] 37%|███▋ | 17725/48008 [1:00:39<1:43:05, 4.90it/s] 37%|███▋ | 17726/48008 [1:00:39<1:40:57, 5.00it/s] 37%|███▋ | 17727/48008 [1:00:39<1:39:12, 5.09it/s] 37%|███▋ | 17728/48008 [1:00:39<1:38:33, 5.12it/s] 37%|███▋ | 17729/48008 [1:00:40<1:37:14, 5.19it/s] 37%|███▋ | 17730/48008 [1:00:40<1:36:35, 5.22it/s] 37%|███▋ | 17731/48008 [1:00:40<1:37:29, 5.18it/s] 37%|███▋ | 17732/48008 [1:00:40<1:36:12, 5.25it/s] 37%|███▋ | 17733/48008 [1:00:40<1:36:19, 5.24it/s] 37%|███▋ | 17734/48008 [1:00:41<1:53:57, 4.43it/s] 37%|███▋ | 17735/48008 [1:00:41<1:51:20, 4.53it/s] 37%|███▋ | 17736/48008 [1:00:41<1:46:20, 4.74it/s] 37%|███▋ | 17737/48008 [1:00:41<1:43:13, 4.89it/s] 37%|███▋ | 17738/48008 [1:00:41<1:41:22, 4.98it/s] 37%|███▋ | 17739/48008 [1:00:42<1:41:36, 4.97it/s] 37%|███▋ | 17740/48008 [1:00:42<1:39:48, 5.05it/s] 37%|███▋ | 17741/48008 [1:00:42<1:40:32, 5.02it/s] 37%|███▋ | 17742/48008 [1:00:42<1:38:40, 5.11it/s] 37%|███▋ | 17743/48008 [1:00:42<1:37:42, 5.16it/s] 37%|███▋ | 17744/48008 [1:00:43<1:39:41, 5.06it/s] 37%|███▋ | 17745/48008 [1:00:43<1:38:27, 5.12it/s] 37%|███▋ | 17746/48008 [1:00:43<1:36:51, 5.21it/s] 37%|███▋ | 17747/48008 [1:00:43<1:35:47, 5.27it/s] 37%|███▋ | 17748/48008 [1:00:43<1:35:56, 5.26it/s] 37%|███▋ | 17749/48008 [1:00:44<1:53:32, 4.44it/s] 37%|███▋ | 17750/48008 [1:00:44<1:50:09, 4.58it/s] {'loss': 0.2999, 'grad_norm': 0.578125, 'learning_rate': 0.0001260581569738377, 'epoch': 0.37} 37%|███▋ | 17750/48008 [1:00:44<1:50:09, 4.58it/s] 37%|███▋ | 17751/48008 [1:00:44<1:44:53, 4.81it/s] 37%|███▋ | 17752/48008 [1:00:44<1:42:24, 4.92it/s] 37%|███▋ | 17753/48008 [1:00:44<1:40:46, 5.00it/s] 37%|███▋ | 17754/48008 [1:00:45<1:40:49, 5.00it/s] 37%|███▋ | 17755/48008 [1:00:45<1:40:21, 5.02it/s] 37%|███▋ | 17756/48008 [1:00:45<1:39:04, 5.09it/s] 37%|███▋ | 17757/48008 [1:00:45<1:40:04, 5.04it/s] 37%|███▋ | 17758/48008 [1:00:45<1:39:40, 5.06it/s] 37%|███▋ | 17759/48008 [1:00:46<1:39:22, 5.07it/s] 37%|███▋ | 17760/48008 [1:00:46<1:40:08, 5.03it/s] 37%|███▋ | 17761/48008 [1:00:46<1:38:58, 5.09it/s] 37%|███▋ | 17762/48008 [1:00:46<1:38:06, 5.14it/s] 37%|███▋ | 17763/48008 [1:00:46<1:44:26, 4.83it/s] 37%|███▋ | 17764/48008 [1:00:47<1:41:32, 4.96it/s] 37%|███▋ | 17765/48008 [1:00:47<1:40:01, 5.04it/s] 37%|███▋ | 17766/48008 [1:00:47<1:40:36, 5.01it/s] 37%|███▋ | 17767/48008 [1:00:47<1:46:18, 4.74it/s] 37%|███▋ | 17768/48008 [1:00:47<1:42:42, 4.91it/s] 37%|███▋ | 17769/48008 [1:00:48<1:40:28, 5.02it/s] 37%|███▋ | 17770/48008 [1:00:48<1:38:27, 5.12it/s] 37%|███▋ | 17771/48008 [1:00:48<1:37:46, 5.15it/s] 37%|███▋ | 17772/48008 [1:00:48<1:37:13, 5.18it/s] 37%|███▋ | 17773/48008 [1:00:48<1:37:29, 5.17it/s] 37%|███▋ | 17774/48008 [1:00:49<1:37:24, 5.17it/s] 37%|███▋ | 17775/48008 [1:00:49<1:37:09, 5.19it/s] 37%|███▋ | 17776/48008 [1:00:49<1:37:46, 5.15it/s] 37%|███▋ | 17777/48008 [1:00:49<1:35:43, 5.26it/s] 37%|███▋ | 17778/48008 [1:00:49<1:35:10, 5.29it/s] 37%|███▋ | 17779/48008 [1:00:50<1:42:21, 4.92it/s] 37%|███▋ | 17780/48008 [1:00:50<1:40:36, 5.01it/s] 37%|███▋ | 17781/48008 [1:00:50<1:39:29, 5.06it/s] 37%|███▋ | 17782/48008 [1:00:50<1:37:58, 5.14it/s] 37%|███▋ | 17783/48008 [1:00:50<1:37:53, 5.15it/s] 37%|███▋ | 17784/48008 [1:00:51<1:37:40, 5.16it/s] 37%|███▋ | 17785/48008 [1:00:51<1:38:51, 5.10it/s] 37%|███▋ | 17786/48008 [1:00:51<1:39:04, 5.08it/s] 37%|███▋ | 17787/48008 [1:00:51<1:38:09, 5.13it/s] 37%|███▋ | 17788/48008 [1:00:51<1:36:45, 5.21it/s] 37%|███▋ | 17789/48008 [1:00:52<1:39:26, 5.07it/s] 37%|███▋ | 17790/48008 [1:00:52<1:40:13, 5.03it/s] 37%|███▋ | 17791/48008 [1:00:52<1:38:27, 5.11it/s] 37%|███▋ | 17792/48008 [1:00:52<1:37:30, 5.16it/s] 37%|███▋ | 17793/48008 [1:00:52<1:37:05, 5.19it/s] 37%|███▋ | 17794/48008 [1:00:52<1:36:41, 5.21it/s] 37%|███▋ | 17795/48008 [1:00:53<1:36:31, 5.22it/s] 37%|███▋ | 17796/48008 [1:00:53<1:36:17, 5.23it/s] 37%|███▋ | 17797/48008 [1:00:53<1:35:54, 5.25it/s] 37%|███▋ | 17798/48008 [1:00:53<1:35:24, 5.28it/s] 37%|███▋ | 17799/48008 [1:00:54<2:05:12, 4.02it/s] 37%|███▋ | 17800/48008 [1:00:54<1:56:22, 4.33it/s] {'loss': 0.2993, 'grad_norm': 0.388671875, 'learning_rate': 0.00012584985835694052, 'epoch': 0.37} 37%|███▋ | 17800/48008 [1:00:54<1:56:22, 4.33it/s] 37%|███▋ | 17801/48008 [1:00:54<2:19:48, 3.60it/s] 37%|███▋ | 17802/48008 [1:00:54<2:05:50, 4.00it/s] 37%|███▋ | 17803/48008 [1:00:55<1:56:39, 4.32it/s] 37%|███▋ | 17804/48008 [1:00:55<1:51:47, 4.50it/s] 37%|███▋ | 17805/48008 [1:00:55<1:46:30, 4.73it/s] 37%|███▋ | 17806/48008 [1:00:55<1:43:15, 4.88it/s] 37%|███▋ | 17807/48008 [1:00:55<1:41:04, 4.98it/s] 37%|███▋ | 17808/48008 [1:00:56<2:09:07, 3.90it/s] 37%|███▋ | 17809/48008 [1:00:56<1:59:04, 4.23it/s] 37%|███▋ | 17810/48008 [1:00:56<1:52:57, 4.46it/s] 37%|███▋ | 17811/48008 [1:00:56<1:48:06, 4.66it/s] 37%|███▋ | 17812/48008 [1:00:57<1:46:01, 4.75it/s] 37%|███▋ | 17813/48008 [1:00:57<1:43:51, 4.85it/s] 37%|███▋ | 17814/48008 [1:00:57<1:40:31, 5.01it/s] 37%|███▋ | 17815/48008 [1:00:57<1:38:55, 5.09it/s] 37%|███▋ | 17816/48008 [1:00:57<1:40:20, 5.02it/s] 37%|███▋ | 17817/48008 [1:00:58<1:45:33, 4.77it/s] 37%|███▋ | 17818/48008 [1:00:58<1:43:29, 4.86it/s] 37%|███▋ | 17819/48008 [1:00:58<1:41:26, 4.96it/s] 37%|███▋ | 17820/48008 [1:00:58<1:47:01, 4.70it/s] 37%|███▋ | 17821/48008 [1:00:58<1:44:25, 4.82it/s] 37%|███▋ | 17822/48008 [1:00:59<1:41:45, 4.94it/s] 37%|███▋ | 17823/48008 [1:00:59<1:42:10, 4.92it/s] 37%|███▋ | 17824/48008 [1:00:59<1:41:12, 4.97it/s] 37%|███▋ | 17825/48008 [1:00:59<1:39:44, 5.04it/s] 37%|███▋ | 17826/48008 [1:00:59<1:38:20, 5.11it/s] 37%|███▋ | 17827/48008 [1:01:00<1:38:19, 5.12it/s] 37%|███▋ | 17828/48008 [1:01:00<1:37:41, 5.15it/s] 37%|███▋ | 17829/48008 [1:01:00<1:39:55, 5.03it/s] 37%|███▋ | 17830/48008 [1:01:00<1:38:35, 5.10it/s] 37%|███▋ | 17831/48008 [1:01:00<1:38:47, 5.09it/s] 37%|███▋ | 17832/48008 [1:01:00<1:38:02, 5.13it/s] 37%|███▋ | 17833/48008 [1:01:01<1:40:02, 5.03it/s] 37%|███▋ | 17834/48008 [1:01:01<1:38:21, 5.11it/s] 37%|███▋ | 17835/48008 [1:01:01<1:37:30, 5.16it/s] 37%|███▋ | 17836/48008 [1:01:01<1:36:40, 5.20it/s] 37%|███▋ | 17837/48008 [1:01:01<1:36:13, 5.23it/s] 37%|███▋ | 17838/48008 [1:01:02<1:35:28, 5.27it/s] 37%|███▋ | 17839/48008 [1:01:02<1:36:32, 5.21it/s] 37%|███▋ | 17840/48008 [1:01:02<1:35:59, 5.24it/s] 37%|███▋ | 17841/48008 [1:01:02<1:37:40, 5.15it/s] 37%|███▋ | 17842/48008 [1:01:02<1:37:11, 5.17it/s] 37%|███▋ | 17843/48008 [1:01:03<1:38:50, 5.09it/s] 37%|███▋ | 17844/48008 [1:01:03<1:55:23, 4.36it/s] 37%|███▋ | 17845/48008 [1:01:03<1:48:51, 4.62it/s] 37%|███▋ | 17846/48008 [1:01:03<1:44:28, 4.81it/s] 37%|███▋ | 17847/48008 [1:01:03<1:41:57, 4.93it/s] 37%|███▋ | 17848/48008 [1:01:04<1:39:47, 5.04it/s] 37%|███▋ | 17849/48008 [1:01:04<1:38:13, 5.12it/s] 37%|███▋ | 17850/48008 [1:01:04<1:39:26, 5.05it/s] {'loss': 0.2998, 'grad_norm': 0.43359375, 'learning_rate': 0.00012564155974004334, 'epoch': 0.37} 37%|███▋ | 17850/48008 [1:01:04<1:39:26, 5.05it/s] 37%|███▋ | 17851/48008 [1:01:04<1:37:34, 5.15it/s] 37%|███▋ | 17852/48008 [1:01:04<1:37:11, 5.17it/s] 37%|███▋ | 17853/48008 [1:01:05<1:36:58, 5.18it/s] 37%|███▋ | 17854/48008 [1:01:05<1:36:40, 5.20it/s] 37%|███▋ | 17855/48008 [1:01:05<1:38:25, 5.11it/s] 37%|███▋ | 17856/48008 [1:01:05<1:37:17, 5.17it/s] 37%|███▋ | 17857/48008 [1:01:05<1:37:43, 5.14it/s] 37%|███▋ | 17858/48008 [1:01:06<1:37:11, 5.17it/s] 37%|███▋ | 17859/48008 [1:01:06<1:40:19, 5.01it/s] 37%|███▋ | 17860/48008 [1:01:06<1:40:15, 5.01it/s] 37%|███▋ | 17861/48008 [1:01:06<1:39:06, 5.07it/s] 37%|███▋ | 17862/48008 [1:01:07<1:55:11, 4.36it/s] 37%|███▋ | 17863/48008 [1:01:07<1:51:11, 4.52it/s] 37%|███▋ | 17864/48008 [1:01:07<2:03:43, 4.06it/s] 37%|███▋ | 17865/48008 [1:01:07<1:55:11, 4.36it/s] 37%|███▋ | 17866/48008 [1:01:07<1:49:21, 4.59it/s] 37%|███▋ | 17867/48008 [1:01:08<1:45:07, 4.78it/s] 37%|███▋ | 17868/48008 [1:01:08<1:59:30, 4.20it/s] 37%|███▋ | 17869/48008 [1:01:08<1:54:22, 4.39it/s] 37%|███▋ | 17870/48008 [1:01:08<1:50:37, 4.54it/s] 37%|███▋ | 17871/48008 [1:01:09<1:52:47, 4.45it/s] 37%|███▋ | 17872/48008 [1:01:09<1:47:37, 4.67it/s] 37%|███▋ | 17873/48008 [1:01:09<1:44:00, 4.83it/s] 37%|███▋ | 17874/48008 [1:01:09<1:41:17, 4.96it/s] 37%|███▋ | 17875/48008 [1:01:09<1:39:45, 5.03it/s] 37%|███▋ | 17876/48008 [1:01:10<1:39:36, 5.04it/s] 37%|███▋ | 17877/48008 [1:01:10<1:39:17, 5.06it/s] 37%|███▋ | 17878/48008 [1:01:10<1:37:25, 5.15it/s] 37%|███▋ | 17879/48008 [1:01:10<1:36:25, 5.21it/s] 37%|███▋ | 17880/48008 [1:01:10<1:36:22, 5.21it/s] 37%|███▋ | 17881/48008 [1:01:10<1:35:56, 5.23it/s] 37%|███▋ | 17882/48008 [1:01:11<1:36:26, 5.21it/s] 37%|███▋ | 17883/48008 [1:01:11<1:36:09, 5.22it/s] 37%|███▋ | 17884/48008 [1:01:11<1:34:43, 5.30it/s] 37%|███▋ | 17885/48008 [1:01:11<1:35:22, 5.26it/s] 37%|███▋ | 17886/48008 [1:01:11<1:33:59, 5.34it/s] 37%|███▋ | 17887/48008 [1:01:12<1:34:15, 5.33it/s] 37%|███▋ | 17888/48008 [1:01:12<1:34:48, 5.30it/s] 37%|███▋ | 17889/48008 [1:01:12<1:35:23, 5.26it/s] 37%|███▋ | 17890/48008 [1:01:12<1:38:33, 5.09it/s] 37%|███▋ | 17891/48008 [1:01:12<1:37:09, 5.17it/s] 37%|███▋ | 17892/48008 [1:01:13<1:36:52, 5.18it/s] 37%|███▋ | 17893/48008 [1:01:13<1:35:49, 5.24it/s] 37%|███▋ | 17894/48008 [1:01:13<1:35:10, 5.27it/s] 37%|███▋ | 17895/48008 [1:01:13<1:35:48, 5.24it/s] 37%|███▋ | 17896/48008 [1:01:14<2:05:41, 3.99it/s] 37%|███▋ | 17897/48008 [1:01:14<1:56:08, 4.32it/s] 37%|███▋ | 17898/48008 [1:01:14<1:50:09, 4.56it/s] 37%|███▋ | 17899/48008 [1:01:14<1:45:49, 4.74it/s] 37%|███▋ | 17900/48008 [1:01:14<1:43:58, 4.83it/s] {'loss': 0.3014, 'grad_norm': 0.1748046875, 'learning_rate': 0.00012543326112314615, 'epoch': 0.37} 37%|███▋ | 17900/48008 [1:01:14<1:43:58, 4.83it/s] 37%|███▋ | 17901/48008 [1:01:15<1:59:07, 4.21it/s] 37%|███▋ | 17902/48008 [1:01:15<1:53:55, 4.40it/s] 37%|███▋ | 17903/48008 [1:01:15<1:50:00, 4.56it/s] 37%|███▋ | 17904/48008 [1:01:15<1:45:45, 4.74it/s] 37%|███▋ | 17905/48008 [1:01:15<1:45:23, 4.76it/s] 37%|███▋ | 17906/48008 [1:01:16<1:43:09, 4.86it/s] 37%|███▋ | 17907/48008 [1:01:16<1:39:41, 5.03it/s] 37%|███▋ | 17908/48008 [1:01:16<1:38:16, 5.10it/s] 37%|███▋ | 17909/48008 [1:01:16<1:37:43, 5.13it/s] 37%|███▋ | 17910/48008 [1:01:16<1:36:31, 5.20it/s] 37%|███▋ | 17911/48008 [1:01:17<1:34:51, 5.29it/s] 37%|███▋ | 17912/48008 [1:01:17<2:04:39, 4.02it/s] 37%|███▋ | 17913/48008 [1:01:17<1:55:55, 4.33it/s] 37%|███▋ | 17914/48008 [1:01:17<1:49:00, 4.60it/s] 37%|███▋ | 17915/48008 [1:01:17<1:44:55, 4.78it/s] 37%|███▋ | 17916/48008 [1:01:18<1:41:50, 4.92it/s] 37%|███▋ | 17917/48008 [1:01:18<1:40:04, 5.01it/s] 37%|███▋ | 17918/48008 [1:01:18<1:38:12, 5.11it/s] 37%|███▋ | 17919/48008 [1:01:18<1:39:25, 5.04it/s] 37%|███▋ | 17920/48008 [1:01:18<1:38:14, 5.10it/s] 37%|███▋ | 17921/48008 [1:01:19<1:37:36, 5.14it/s] 37%|███▋ | 17922/48008 [1:01:19<1:38:34, 5.09it/s] 37%|███▋ | 17923/48008 [1:01:19<1:37:45, 5.13it/s] 37%|███▋ | 17924/48008 [1:01:19<1:36:44, 5.18it/s] 37%|███▋ | 17925/48008 [1:01:19<1:36:32, 5.19it/s] 37%|███▋ | 17926/48008 [1:01:20<1:37:17, 5.15it/s] 37%|███▋ | 17927/48008 [1:01:20<1:36:20, 5.20it/s] 37%|███▋ | 17928/48008 [1:01:20<2:05:21, 4.00it/s] 37%|███▋ | 17929/48008 [1:01:20<1:57:34, 4.26it/s] 37%|███▋ | 17930/48008 [1:01:21<1:50:07, 4.55it/s] 37%|███▋ | 17931/48008 [1:01:21<1:45:35, 4.75it/s] 37%|███▋ | 17932/48008 [1:01:21<1:43:32, 4.84it/s] 37%|███▋ | 17933/48008 [1:01:21<1:42:44, 4.88it/s] 37%|███▋ | 17934/48008 [1:01:21<1:40:04, 5.01it/s] 37%|███▋ | 17935/48008 [1:01:22<1:38:00, 5.11it/s] 37%|███▋ | 17936/48008 [1:01:22<1:37:04, 5.16it/s] 37%|███▋ | 17937/48008 [1:01:22<1:35:48, 5.23it/s] 37%|███▋ | 17938/48008 [1:01:22<1:35:21, 5.26it/s] 37%|███▋ | 17939/48008 [1:01:22<1:34:43, 5.29it/s] 37%|███▋ | 17940/48008 [1:01:22<1:36:04, 5.22it/s] 37%|███▋ | 17941/48008 [1:01:23<1:36:28, 5.19it/s] 37%|███▋ | 17942/48008 [1:01:23<1:38:00, 5.11it/s] 37%|███▋ | 17943/48008 [1:01:23<1:36:34, 5.19it/s] 37%|███▋ | 17944/48008 [1:01:23<1:36:13, 5.21it/s] 37%|███▋ | 17945/48008 [1:01:23<1:35:13, 5.26it/s] 37%|███▋ | 17946/48008 [1:01:24<1:33:44, 5.35it/s] 37%|███▋ | 17947/48008 [1:01:24<1:34:01, 5.33it/s] 37%|███▋ | 17948/48008 [1:01:24<1:34:27, 5.30it/s] 37%|███▋ | 17949/48008 [1:01:24<1:34:43, 5.29it/s] 37%|███▋ | 17950/48008 [1:01:24<1:35:06, 5.27it/s] {'loss': 0.3039, 'grad_norm': 0.1650390625, 'learning_rate': 0.00012522496250624894, 'epoch': 0.37} 37%|███▋ | 17950/48008 [1:01:24<1:35:06, 5.27it/s] 37%|███▋ | 17951/48008 [1:01:25<1:36:02, 5.22it/s] 37%|███▋ | 17952/48008 [1:01:25<1:38:31, 5.08it/s] 37%|███▋ | 17953/48008 [1:01:25<1:38:02, 5.11it/s] 37%|███▋ | 17954/48008 [1:01:25<1:39:11, 5.05it/s] 37%|███▋ | 17955/48008 [1:01:25<1:40:03, 5.01it/s] 37%|███▋ | 17956/48008 [1:01:26<1:39:36, 5.03it/s] 37%|███▋ | 17957/48008 [1:01:26<1:38:00, 5.11it/s] 37%|███▋ | 17958/48008 [1:01:26<1:38:04, 5.11it/s] 37%|███▋ | 17959/48008 [1:01:26<1:44:21, 4.80it/s] 37%|███▋ | 17960/48008 [1:01:26<1:40:28, 4.98it/s] 37%|███▋ | 17961/48008 [1:01:27<1:40:13, 5.00it/s] 37%|███▋ | 17962/48008 [1:01:27<1:38:26, 5.09it/s] 37%|███▋ | 17963/48008 [1:01:27<1:37:18, 5.15it/s] 37%|███▋ | 17964/48008 [1:01:27<1:36:54, 5.17it/s] 37%|███▋ | 17965/48008 [1:01:27<1:36:55, 5.17it/s] 37%|███▋ | 17966/48008 [1:01:28<1:37:12, 5.15it/s] 37%|███▋ | 17967/48008 [1:01:28<1:37:11, 5.15it/s] 37%|███▋ | 17968/48008 [1:01:28<1:36:37, 5.18it/s] 37%|███▋ | 17969/48008 [1:01:28<1:35:14, 5.26it/s] 37%|███▋ | 17970/48008 [1:01:28<1:35:19, 5.25it/s] 37%|███▋ | 17971/48008 [1:01:28<1:35:14, 5.26it/s] 37%|███▋ | 17972/48008 [1:01:29<1:34:27, 5.30it/s] 37%|███▋ | 17973/48008 [1:01:29<1:36:48, 5.17it/s] 37%|███▋ | 17974/48008 [1:01:29<2:06:00, 3.97it/s] 37%|███▋ | 17975/48008 [1:01:29<1:56:12, 4.31it/s] 37%|███▋ | 17976/48008 [1:01:30<1:50:54, 4.51it/s] 37%|███▋ | 17977/48008 [1:01:30<1:46:24, 4.70it/s] 37%|███▋ | 17978/48008 [1:01:30<1:44:57, 4.77it/s] 37%|███▋ | 17979/48008 [1:01:30<1:41:25, 4.93it/s] 37%|███▋ | 17980/48008 [1:01:30<1:41:28, 4.93it/s] 37%|███▋ | 17981/48008 [1:01:31<1:39:38, 5.02it/s] 37%|███▋ | 17982/48008 [1:01:31<1:40:06, 5.00it/s] 37%|███▋ | 17983/48008 [1:01:31<1:37:41, 5.12it/s] 37%|███▋ | 17984/48008 [1:01:31<1:36:35, 5.18it/s] 37%|███▋ | 17985/48008 [1:01:31<1:35:35, 5.23it/s] 37%|███▋ | 17986/48008 [1:01:32<1:34:54, 5.27it/s] 37%|███▋ | 17987/48008 [1:01:32<1:34:17, 5.31it/s] 37%|███▋ | 17988/48008 [1:01:32<1:35:43, 5.23it/s] 37%|███▋ | 17989/48008 [1:01:32<1:35:21, 5.25it/s] 37%|███▋ | 17990/48008 [1:01:32<1:37:30, 5.13it/s] 37%|███▋ | 17991/48008 [1:01:33<1:38:03, 5.10it/s] 37%|███▋ | 17992/48008 [1:01:33<1:36:55, 5.16it/s] 37%|███▋ | 17993/48008 [1:01:33<1:36:37, 5.18it/s] 37%|███▋ | 17994/48008 [1:01:33<1:35:55, 5.21it/s] 37%|███▋ | 17995/48008 [1:01:33<1:38:32, 5.08it/s] 37%|███▋ | 17996/48008 [1:01:34<4:03:45, 2.05it/s] 37%|███▋ | 17997/48008 [1:01:35<3:19:10, 2.51it/s] 37%|███▋ | 17998/48008 [1:01:35<2:47:54, 2.98it/s] 37%|███▋ | 17999/48008 [1:01:35<2:27:46, 3.38it/s] 37%|███▋ | 18000/48008 [1:01:36<3:55:46, 2.12it/s] {'loss': 0.3009, 'grad_norm': 0.404296875, 'learning_rate': 0.0001250166638893518, 'epoch': 0.37} 37%|███▋ | 18000/48008 [1:01:36<3:55:46, 2.12it/s] 37%|███▋ | 18001/48008 [1:01:36<3:12:49, 2.59it/s] 37%|███▋ | 18002/48008 [1:01:36<2:43:51, 3.05it/s] 38%|███▊ | 18003/48008 [1:01:37<2:23:36, 3.48it/s] 38%|███▊ | 18004/48008 [1:01:37<2:10:28, 3.83it/s] 38%|███▊ | 18005/48008 [1:01:37<1:59:42, 4.18it/s] 38%|███▊ | 18006/48008 [1:01:37<1:51:30, 4.48it/s] 38%|███▊ | 18007/48008 [1:01:37<1:46:59, 4.67it/s] 38%|███▊ | 18008/48008 [1:01:37<1:43:41, 4.82it/s] 38%|███▊ | 18009/48008 [1:01:38<1:40:32, 4.97it/s] 38%|███▊ | 18010/48008 [1:01:38<1:41:14, 4.94it/s] 38%|███▊ | 18011/48008 [1:01:38<1:40:37, 4.97it/s] 38%|███▊ | 18012/48008 [1:01:38<1:46:03, 4.71it/s] 38%|███▊ | 18013/48008 [1:01:38<1:42:32, 4.88it/s] 38%|███▊ | 18014/48008 [1:01:39<1:39:44, 5.01it/s] 38%|███▊ | 18015/48008 [1:01:39<1:38:35, 5.07it/s] 38%|███▊ | 18016/48008 [1:01:39<1:37:07, 5.15it/s] 38%|███▊ | 18017/48008 [1:01:39<1:37:03, 5.15it/s] 38%|███▊ | 18018/48008 [1:01:39<1:36:14, 5.19it/s] 38%|███▊ | 18019/48008 [1:01:40<1:37:02, 5.15it/s] 38%|███▊ | 18020/48008 [1:01:40<1:36:27, 5.18it/s] 38%|███▊ | 18021/48008 [1:01:40<1:43:03, 4.85it/s] 38%|███▊ | 18022/48008 [1:01:40<1:42:40, 4.87it/s] 38%|███▊ | 18023/48008 [1:01:40<1:40:12, 4.99it/s] 38%|███▊ | 18024/48008 [1:01:41<1:40:10, 4.99it/s] 38%|███▊ | 18025/48008 [1:01:41<1:38:53, 5.05it/s] 38%|███▊ | 18026/48008 [1:01:41<1:38:17, 5.08it/s] 38%|███▊ | 18027/48008 [1:01:41<1:37:37, 5.12it/s] 38%|███▊ | 18028/48008 [1:01:41<1:36:23, 5.18it/s] 38%|███▊ | 18029/48008 [1:01:42<1:35:44, 5.22it/s] 38%|███▊ | 18030/48008 [1:01:42<1:34:57, 5.26it/s] 38%|███▊ | 18031/48008 [1:01:42<1:35:08, 5.25it/s] 38%|███▊ | 18032/48008 [1:01:42<1:34:32, 5.28it/s] 38%|███▊ | 18033/48008 [1:01:42<1:36:45, 5.16it/s] 38%|███▊ | 18034/48008 [1:01:43<1:37:10, 5.14it/s] 38%|███▊ | 18035/48008 [1:01:43<1:35:39, 5.22it/s] 38%|███▊ | 18036/48008 [1:01:43<1:38:32, 5.07it/s] 38%|███▊ | 18037/48008 [1:01:43<1:39:08, 5.04it/s] 38%|███▊ | 18038/48008 [1:01:43<1:37:34, 5.12it/s] 38%|███▊ | 18039/48008 [1:01:44<1:36:28, 5.18it/s] 38%|███▊ | 18040/48008 [1:01:44<1:38:50, 5.05it/s] 38%|███▊ | 18041/48008 [1:01:44<1:37:48, 5.11it/s] 38%|███▊ | 18042/48008 [1:01:44<1:36:21, 5.18it/s] 38%|███▊ | 18043/48008 [1:01:44<1:35:55, 5.21it/s] 38%|███▊ | 18044/48008 [1:01:45<1:36:27, 5.18it/s] 38%|███▊ | 18045/48008 [1:01:45<1:35:55, 5.21it/s] 38%|███▊ | 18046/48008 [1:01:45<1:35:55, 5.21it/s] 38%|███▊ | 18047/48008 [1:01:45<1:35:15, 5.24it/s] 38%|███▊ | 18048/48008 [1:01:45<1:35:15, 5.24it/s] 38%|███▊ | 18049/48008 [1:01:45<1:35:12, 5.24it/s] 38%|███▊ | 18050/48008 [1:01:46<1:52:25, 4.44it/s] {'loss': 0.3028, 'grad_norm': 0.6171875, 'learning_rate': 0.00012480836527245458, 'epoch': 0.38} 38%|███▊ | 18050/48008 [1:01:46<1:52:25, 4.44it/s] 38%|███▊ | 18051/48008 [1:01:46<1:47:35, 4.64it/s] 38%|███▊ | 18052/48008 [1:01:46<1:43:58, 4.80it/s] 38%|███▊ | 18053/48008 [1:01:46<1:39:58, 4.99it/s] 38%|███▊ | 18054/48008 [1:01:47<1:45:53, 4.71it/s] 38%|███▊ | 18055/48008 [1:01:47<1:44:10, 4.79it/s] 38%|███▊ | 18056/48008 [1:01:47<1:41:35, 4.91it/s] 38%|███▊ | 18057/48008 [1:01:47<1:39:22, 5.02it/s] 38%|███▊ | 18058/48008 [1:01:47<1:38:05, 5.09it/s] 38%|███▊ | 18059/48008 [1:01:48<1:37:22, 5.13it/s] 38%|███▊ | 18060/48008 [1:01:48<1:36:50, 5.15it/s] 38%|███▊ | 18061/48008 [1:01:48<1:38:01, 5.09it/s] 38%|███▊ | 18062/48008 [1:01:48<1:44:15, 4.79it/s] 38%|███▊ | 18063/48008 [1:01:48<1:40:46, 4.95it/s] 38%|███▊ | 18064/48008 [1:01:49<1:38:51, 5.05it/s] 38%|███▊ | 18065/48008 [1:01:49<1:39:39, 5.01it/s] 38%|███▊ | 18066/48008 [1:01:49<1:38:19, 5.08it/s] 38%|███▊ | 18067/48008 [1:01:49<1:37:13, 5.13it/s] 38%|███▊ | 18068/48008 [1:01:49<1:37:18, 5.13it/s] 38%|███▊ | 18069/48008 [1:01:50<1:36:51, 5.15it/s] 38%|███▊ | 18070/48008 [1:01:50<1:35:31, 5.22it/s] 38%|███▊ | 18071/48008 [1:01:50<1:34:51, 5.26it/s] 38%|███▊ | 18072/48008 [1:01:50<1:34:46, 5.26it/s] 38%|███▊ | 18073/48008 [1:01:50<1:34:26, 5.28it/s] 38%|███▊ | 18074/48008 [1:01:50<1:34:01, 5.31it/s] 38%|███▊ | 18075/48008 [1:01:51<1:35:34, 5.22it/s] 38%|███▊ | 18076/48008 [1:01:51<1:36:23, 5.18it/s] 38%|███▊ | 18077/48008 [1:01:51<1:36:46, 5.16it/s] 38%|███▊ | 18078/48008 [1:01:51<1:36:17, 5.18it/s] 38%|███▊ | 18079/48008 [1:01:51<1:35:52, 5.20it/s] 38%|███▊ | 18080/48008 [1:01:52<1:35:22, 5.23it/s] 38%|███▊ | 18081/48008 [1:01:52<1:35:27, 5.22it/s] 38%|███▊ | 18082/48008 [1:01:52<1:36:13, 5.18it/s] 38%|███▊ | 18083/48008 [1:01:52<1:35:20, 5.23it/s] 38%|███▊ | 18084/48008 [1:01:52<1:36:14, 5.18it/s] 38%|███▊ | 18085/48008 [1:01:53<1:34:07, 5.30it/s] 38%|███▊ | 18086/48008 [1:01:53<1:34:18, 5.29it/s] 38%|███▊ | 18087/48008 [1:01:53<1:35:12, 5.24it/s] 38%|███▊ | 18088/48008 [1:01:53<1:36:41, 5.16it/s] 38%|███▊ | 18089/48008 [1:01:53<1:37:15, 5.13it/s] 38%|███▊ | 18090/48008 [1:01:54<1:38:30, 5.06it/s] 38%|███▊ | 18091/48008 [1:01:54<1:37:31, 5.11it/s] 38%|███▊ | 18092/48008 [1:01:54<1:43:57, 4.80it/s] 38%|███▊ | 18093/48008 [1:01:54<1:48:45, 4.58it/s] 38%|███▊ | 18094/48008 [1:01:54<1:46:18, 4.69it/s] 38%|███▊ | 18095/48008 [1:01:55<1:43:29, 4.82it/s] 38%|███▊ | 18096/48008 [1:01:55<1:40:15, 4.97it/s] 38%|███▊ | 18097/48008 [1:01:55<1:38:36, 5.06it/s] 38%|███▊ | 18098/48008 [1:01:55<1:37:19, 5.12it/s] 38%|███▊ | 18099/48008 [1:01:55<1:39:27, 5.01it/s] 38%|███▊ | 18100/48008 [1:01:56<1:38:25, 5.06it/s] {'loss': 0.3027, 'grad_norm': 0.28515625, 'learning_rate': 0.00012460006665555742, 'epoch': 0.38} 38%|███▊ | 18100/48008 [1:01:56<1:38:25, 5.06it/s] 38%|███▊ | 18101/48008 [1:01:56<1:39:18, 5.02it/s] 38%|███▊ | 18102/48008 [1:01:56<1:40:38, 4.95it/s] 38%|███▊ | 18103/48008 [1:01:56<1:45:38, 4.72it/s] 38%|███▊ | 18104/48008 [1:01:56<1:42:25, 4.87it/s] 38%|███▊ | 18105/48008 [1:01:57<1:40:08, 4.98it/s] 38%|███▊ | 18106/48008 [1:01:57<1:37:25, 5.11it/s] 38%|███▊ | 18107/48008 [1:01:57<1:36:30, 5.16it/s] 38%|███▊ | 18108/48008 [1:01:57<1:35:30, 5.22it/s] 38%|███▊ | 18109/48008 [1:01:57<1:35:23, 5.22it/s] 38%|███▊ | 18110/48008 [1:01:58<1:35:30, 5.22it/s] 38%|███▊ | 18111/48008 [1:01:58<1:36:50, 5.15it/s] 38%|███▊ | 18112/48008 [1:01:58<1:36:48, 5.15it/s] 38%|███▊ | 18113/48008 [1:01:58<1:43:07, 4.83it/s] 38%|███▊ | 18114/48008 [1:01:58<1:39:47, 4.99it/s] 38%|███▊ | 18115/48008 [1:01:59<1:40:24, 4.96it/s] 38%|███▊ | 18116/48008 [1:01:59<1:40:02, 4.98it/s] 38%|███▊ | 18117/48008 [1:01:59<1:40:54, 4.94it/s] 38%|███▊ | 18118/48008 [1:01:59<1:39:00, 5.03it/s] 38%|███▊ | 18119/48008 [1:01:59<1:40:20, 4.96it/s] 38%|███▊ | 18120/48008 [1:02:00<1:38:27, 5.06it/s] 38%|███▊ | 18121/48008 [1:02:00<1:37:10, 5.13it/s] 38%|███▊ | 18122/48008 [1:02:00<1:36:32, 5.16it/s] 38%|███▊ | 18123/48008 [1:02:00<1:35:51, 5.20it/s] 38%|███▊ | 18124/48008 [1:02:00<1:35:41, 5.20it/s] 38%|███▊ | 18125/48008 [1:02:01<1:35:33, 5.21it/s] 38%|███▊ | 18126/48008 [1:02:01<1:34:19, 5.28it/s] 38%|███▊ | 18127/48008 [1:02:01<1:36:43, 5.15it/s] 38%|███▊ | 18128/48008 [1:02:01<1:36:49, 5.14it/s] 38%|███▊ | 18129/48008 [1:02:01<1:36:29, 5.16it/s] 38%|███▊ | 18130/48008 [1:02:02<1:37:38, 5.10it/s] 38%|███▊ | 18131/48008 [1:02:02<1:36:29, 5.16it/s] 38%|███▊ | 18132/48008 [1:02:02<2:05:29, 3.97it/s] 38%|███▊ | 18133/48008 [1:02:02<1:58:09, 4.21it/s] 38%|███▊ | 18134/48008 [1:02:03<2:09:08, 3.86it/s] 38%|███▊ | 18135/48008 [1:02:03<1:59:10, 4.18it/s] 38%|███▊ | 18136/48008 [1:02:03<1:51:11, 4.48it/s] 38%|███▊ | 18137/48008 [1:02:03<1:46:22, 4.68it/s] 38%|███▊ | 18138/48008 [1:02:03<1:49:56, 4.53it/s] 38%|███▊ | 18139/48008 [1:02:04<1:44:21, 4.77it/s] 38%|███▊ | 18140/48008 [1:02:04<1:40:43, 4.94it/s] 38%|███▊ | 18141/48008 [1:02:04<1:39:01, 5.03it/s] 38%|███▊ | 18142/48008 [1:02:04<1:38:11, 5.07it/s] 38%|███▊ | 18143/48008 [1:02:04<1:37:33, 5.10it/s] 38%|███▊ | 18144/48008 [1:02:05<1:39:38, 5.00it/s] 38%|███▊ | 18145/48008 [1:02:05<1:37:48, 5.09it/s] 38%|███▊ | 18146/48008 [1:02:05<1:54:01, 4.36it/s] 38%|███▊ | 18147/48008 [1:02:05<1:51:13, 4.47it/s] 38%|███▊ | 18148/48008 [1:02:05<1:45:52, 4.70it/s] 38%|███▊ | 18149/48008 [1:02:06<1:42:50, 4.84it/s] 38%|███▊ | 18150/48008 [1:02:06<1:40:18, 4.96it/s] {'loss': 0.3004, 'grad_norm': 0.41015625, 'learning_rate': 0.00012439176803866024, 'epoch': 0.38} 38%|███▊ | 18150/48008 [1:02:06<1:40:18, 4.96it/s] 38%|███▊ | 18151/48008 [1:02:06<1:41:38, 4.90it/s] 38%|███▊ | 18152/48008 [1:02:06<1:38:45, 5.04it/s] 38%|███▊ | 18153/48008 [1:02:06<1:39:15, 5.01it/s] 38%|███▊ | 18154/48008 [1:02:07<1:38:16, 5.06it/s] 38%|███▊ | 18155/48008 [1:02:07<1:36:47, 5.14it/s] 38%|███▊ | 18156/48008 [1:02:07<1:35:36, 5.20it/s] 38%|███▊ | 18157/48008 [1:02:07<1:37:13, 5.12it/s] 38%|███▊ | 18158/48008 [1:02:07<1:36:29, 5.16it/s] 38%|███▊ | 18159/48008 [1:02:08<2:04:51, 3.98it/s] 38%|███▊ | 18160/48008 [1:02:08<1:58:44, 4.19it/s] 38%|███▊ | 18161/48008 [1:02:08<1:51:08, 4.48it/s] 38%|███▊ | 18162/48008 [1:02:08<1:48:01, 4.60it/s] 38%|███▊ | 18163/48008 [1:02:09<1:44:14, 4.77it/s] 38%|███▊ | 18164/48008 [1:02:09<1:42:32, 4.85it/s] 38%|███▊ | 18165/48008 [1:02:09<1:40:23, 4.95it/s] 38%|███▊ | 18166/48008 [1:02:09<1:38:22, 5.06it/s] 38%|███▊ | 18167/48008 [1:02:09<1:37:05, 5.12it/s] 38%|███▊ | 18168/48008 [1:02:10<1:36:24, 5.16it/s] 38%|███▊ | 18169/48008 [1:02:10<1:36:33, 5.15it/s] 38%|███▊ | 18170/48008 [1:02:10<1:35:36, 5.20it/s] 38%|███▊ | 18171/48008 [1:02:10<1:36:46, 5.14it/s] 38%|███▊ | 18172/48008 [1:02:10<1:36:44, 5.14it/s] 38%|███▊ | 18173/48008 [1:02:11<1:38:05, 5.07it/s] 38%|███▊ | 18174/48008 [1:02:11<2:06:22, 3.93it/s] 38%|███▊ | 18175/48008 [1:02:11<2:26:07, 3.40it/s] 38%|███▊ | 18176/48008 [1:02:11<2:12:37, 3.75it/s] 38%|███▊ | 18177/48008 [1:02:12<2:30:10, 3.31it/s] 38%|███▊ | 18178/48008 [1:02:12<2:14:14, 3.70it/s] 38%|███▊ | 18179/48008 [1:02:12<2:02:27, 4.06it/s] 38%|███▊ | 18180/48008 [1:02:12<1:54:49, 4.33it/s] 38%|███▊ | 18181/48008 [1:02:13<1:50:44, 4.49it/s] 38%|███▊ | 18182/48008 [1:02:13<1:46:39, 4.66it/s] 38%|███▊ | 18183/48008 [1:02:13<1:43:03, 4.82it/s] 38%|███▊ | 18184/48008 [1:02:13<1:39:55, 4.97it/s] 38%|███▊ | 18185/48008 [1:02:13<1:37:19, 5.11it/s] 38%|███▊ | 18186/48008 [1:02:14<1:36:33, 5.15it/s] 38%|███▊ | 18187/48008 [1:02:14<1:36:36, 5.14it/s] 38%|███▊ | 18188/48008 [1:02:14<1:35:51, 5.18it/s] 38%|███▊ | 18189/48008 [1:02:14<1:35:40, 5.19it/s] 38%|███▊ | 18190/48008 [1:02:14<1:35:32, 5.20it/s] 38%|███▊ | 18191/48008 [1:02:15<1:35:17, 5.22it/s] 38%|███▊ | 18192/48008 [1:02:15<2:04:09, 4.00it/s] 38%|███▊ | 18193/48008 [1:02:15<1:57:13, 4.24it/s] 38%|███▊ | 18194/48008 [1:02:15<1:52:03, 4.43it/s] 38%|███▊ | 18195/48008 [1:02:16<1:47:10, 4.64it/s] 38%|███▊ | 18196/48008 [1:02:16<1:43:18, 4.81it/s] 38%|███▊ | 18197/48008 [1:02:16<1:39:57, 4.97it/s] 38%|███▊ | 18198/48008 [1:02:16<1:38:19, 5.05it/s] 38%|███▊ | 18199/48008 [1:02:16<1:38:00, 5.07it/s] 38%|███▊ | 18200/48008 [1:02:16<1:36:30, 5.15it/s] {'loss': 0.3011, 'grad_norm': 0.392578125, 'learning_rate': 0.00012418346942176303, 'epoch': 0.38} 38%|███▊ | 18200/48008 [1:02:16<1:36:30, 5.15it/s] 38%|███▊ | 18201/48008 [1:02:17<1:37:28, 5.10it/s] 38%|███▊ | 18202/48008 [1:02:17<1:36:25, 5.15it/s] 38%|███▊ | 18203/48008 [1:02:17<1:35:54, 5.18it/s] 38%|███▊ | 18204/48008 [1:02:17<1:34:25, 5.26it/s] 38%|███▊ | 18205/48008 [1:02:17<1:33:09, 5.33it/s] 38%|███▊ | 18206/48008 [1:02:18<1:40:32, 4.94it/s] 38%|███▊ | 18207/48008 [1:02:18<1:38:32, 5.04it/s] 38%|███▊ | 18208/48008 [1:02:18<1:37:27, 5.10it/s] 38%|███▊ | 18209/48008 [1:02:18<1:38:02, 5.07it/s] 38%|███▊ | 18210/48008 [1:02:18<1:37:01, 5.12it/s] 38%|███▊ | 18211/48008 [1:02:19<1:35:47, 5.18it/s] 38%|███▊ | 18212/48008 [1:02:19<1:37:09, 5.11it/s] 38%|███▊ | 18213/48008 [1:02:19<1:36:20, 5.15it/s] 38%|███▊ | 18214/48008 [1:02:19<1:36:04, 5.17it/s] 38%|███▊ | 18215/48008 [1:02:19<1:35:45, 5.19it/s] 38%|███▊ | 18216/48008 [1:02:20<1:42:10, 4.86it/s] 38%|███▊ | 18217/48008 [1:02:20<1:40:02, 4.96it/s] 38%|███▊ | 18218/48008 [1:02:20<1:40:30, 4.94it/s] 38%|███▊ | 18219/48008 [1:02:20<1:39:54, 4.97it/s] 38%|███▊ | 18220/48008 [1:02:20<1:40:20, 4.95it/s] 38%|███▊ | 18221/48008 [1:02:21<1:39:26, 4.99it/s] 38%|███▊ | 18222/48008 [1:02:21<1:37:23, 5.10it/s] 38%|███▊ | 18223/48008 [1:02:21<1:36:43, 5.13it/s] 38%|███▊ | 18224/48008 [1:02:21<1:37:54, 5.07it/s] 38%|███▊ | 18225/48008 [1:02:21<1:36:52, 5.12it/s] 38%|███▊ | 18226/48008 [1:02:22<1:38:19, 5.05it/s] 38%|███▊ | 18227/48008 [1:02:22<1:44:10, 4.76it/s] 38%|███▊ | 18228/48008 [1:02:22<1:41:30, 4.89it/s] 38%|███▊ | 18229/48008 [1:02:22<1:40:25, 4.94it/s] 38%|███▊ | 18230/48008 [1:02:22<1:38:10, 5.06it/s] 38%|███▊ | 18231/48008 [1:02:23<1:37:05, 5.11it/s] 38%|███▊ | 18232/48008 [1:02:23<2:05:06, 3.97it/s] 38%|███▊ | 18233/48008 [1:02:23<1:55:57, 4.28it/s] 38%|███▊ | 18234/48008 [1:02:23<1:51:34, 4.45it/s] 38%|███▊ | 18235/48008 [1:02:24<1:48:17, 4.58it/s] 38%|███▊ | 18236/48008 [1:02:24<2:01:00, 4.10it/s] 38%|███▊ | 18237/48008 [1:02:24<1:54:50, 4.32it/s] 38%|███▊ | 18238/48008 [1:02:24<1:48:46, 4.56it/s] 38%|███▊ | 18239/48008 [1:02:24<1:44:47, 4.73it/s] 38%|███▊ | 18240/48008 [1:02:25<1:41:56, 4.87it/s] 38%|███▊ | 18241/48008 [1:02:25<1:39:33, 4.98it/s] 38%|███▊ | 18242/48008 [1:02:25<1:38:28, 5.04it/s] 38%|███▊ | 18243/48008 [1:02:25<1:36:55, 5.12it/s] 38%|███▊ | 18244/48008 [1:02:25<1:35:40, 5.19it/s] 38%|███▊ | 18245/48008 [1:02:26<1:35:12, 5.21it/s] 38%|███▊ | 18246/48008 [1:02:26<1:34:54, 5.23it/s] 38%|███▊ | 18247/48008 [1:02:26<1:33:57, 5.28it/s] 38%|███▊ | 18248/48008 [1:02:26<1:33:59, 5.28it/s] 38%|███▊ | 18249/48008 [1:02:26<1:34:12, 5.26it/s] 38%|███▊ | 18250/48008 [1:02:27<1:35:36, 5.19it/s] {'loss': 0.2984, 'grad_norm': 0.5, 'learning_rate': 0.00012397517080486587, 'epoch': 0.38} 38%|███▊ | 18250/48008 [1:02:27<1:35:36, 5.19it/s] 38%|███▊ | 18251/48008 [1:02:27<1:36:06, 5.16it/s] 38%|███▊ | 18252/48008 [1:02:27<1:42:40, 4.83it/s] 38%|███▊ | 18253/48008 [1:02:27<1:41:55, 4.87it/s] 38%|███▊ | 18254/48008 [1:02:28<1:56:42, 4.25it/s] 38%|███▊ | 18255/48008 [1:02:28<1:50:13, 4.50it/s] 38%|███▊ | 18256/48008 [1:02:28<1:46:17, 4.66it/s] 38%|███▊ | 18257/48008 [1:02:28<1:42:12, 4.85it/s] 38%|███▊ | 18258/48008 [1:02:28<1:41:43, 4.87it/s] 38%|███▊ | 18259/48008 [1:02:28<1:39:02, 5.01it/s] 38%|███▊ | 18260/48008 [1:02:29<1:37:17, 5.10it/s] 38%|███▊ | 18261/48008 [1:02:29<1:38:09, 5.05it/s] 38%|███▊ | 18262/48008 [1:02:29<1:38:06, 5.05it/s] 38%|███▊ | 18263/48008 [1:02:29<1:43:44, 4.78it/s] 38%|███▊ | 18264/48008 [1:02:29<1:40:59, 4.91it/s] 38%|███▊ | 18265/48008 [1:02:30<1:38:34, 5.03it/s] 38%|███▊ | 18266/48008 [1:02:30<1:37:19, 5.09it/s] 38%|███▊ | 18267/48008 [1:02:30<1:38:16, 5.04it/s] 38%|███▊ | 18268/48008 [1:02:30<1:38:53, 5.01it/s] 38%|███▊ | 18269/48008 [1:02:30<1:35:57, 5.17it/s] 38%|███▊ | 18270/48008 [1:02:31<1:35:56, 5.17it/s] 38%|███▊ | 18271/48008 [1:02:31<1:37:11, 5.10it/s] 38%|███▊ | 18272/48008 [1:02:31<1:36:17, 5.15it/s] 38%|███▊ | 18273/48008 [1:02:31<1:36:00, 5.16it/s] 38%|███▊ | 18274/48008 [1:02:32<2:04:15, 3.99it/s] 38%|███▊ | 18275/48008 [1:02:32<1:54:55, 4.31it/s] 38%|███▊ | 18276/48008 [1:02:32<1:48:07, 4.58it/s] 38%|███▊ | 18277/48008 [1:02:32<1:44:23, 4.75it/s] 38%|███▊ | 18278/48008 [1:02:32<1:41:34, 4.88it/s] 38%|███▊ | 18279/48008 [1:02:33<1:38:54, 5.01it/s] 38%|███▊ | 18280/48008 [1:02:33<1:44:35, 4.74it/s] 38%|███▊ | 18281/48008 [1:02:33<1:42:19, 4.84it/s] 38%|███▊ | 18282/48008 [1:02:33<1:40:31, 4.93it/s] 38%|███▊ | 18283/48008 [1:02:33<1:38:12, 5.04it/s] 38%|███▊ | 18284/48008 [1:02:34<1:38:36, 5.02it/s] 38%|███▊ | 18285/48008 [1:02:34<1:36:48, 5.12it/s] 38%|███▊ | 18286/48008 [1:02:34<1:37:13, 5.10it/s] 38%|███▊ | 18287/48008 [1:02:34<1:36:13, 5.15it/s] 38%|███▊ | 18288/48008 [1:02:34<1:37:27, 5.08it/s] 38%|███▊ | 18289/48008 [1:02:35<1:38:11, 5.04it/s] 38%|███▊ | 18290/48008 [1:02:35<1:38:50, 5.01it/s] 38%|███▊ | 18291/48008 [1:02:35<1:37:31, 5.08it/s] 38%|███▊ | 18292/48008 [1:02:35<1:37:31, 5.08it/s] 38%|███▊ | 18293/48008 [1:02:35<1:38:18, 5.04it/s] 38%|███▊ | 18294/48008 [1:02:36<1:37:17, 5.09it/s] 38%|███▊ | 18295/48008 [1:02:36<1:35:36, 5.18it/s] 38%|███▊ | 18296/48008 [1:02:36<1:35:41, 5.17it/s] 38%|███▊ | 18297/48008 [1:02:36<1:36:25, 5.14it/s] 38%|███▊ | 18298/48008 [1:02:36<1:35:56, 5.16it/s] 38%|███▊ | 18299/48008 [1:02:37<1:42:14, 4.84it/s] 38%|███▊ | 18300/48008 [1:02:37<1:39:54, 4.96it/s] {'loss': 0.3018, 'grad_norm': 0.1494140625, 'learning_rate': 0.00012376687218796867, 'epoch': 0.38} 38%|███▊ | 18300/48008 [1:02:37<1:39:54, 4.96it/s] 38%|███▊ | 18301/48008 [1:02:37<1:38:02, 5.05it/s] 38%|███▊ | 18302/48008 [1:02:37<1:36:34, 5.13it/s] 38%|███▊ | 18303/48008 [1:02:37<1:36:42, 5.12it/s] 38%|███▊ | 18304/48008 [1:02:38<2:05:12, 3.95it/s] 38%|███▊ | 18305/48008 [1:02:38<1:55:39, 4.28it/s] 38%|███▊ | 18306/48008 [1:02:38<1:49:04, 4.54it/s] 38%|███▊ | 18307/48008 [1:02:38<2:13:42, 3.70it/s] 38%|███▊ | 18308/48008 [1:02:39<2:02:11, 4.05it/s] 38%|███▊ | 18309/48008 [1:02:39<2:01:07, 4.09it/s] 38%|███▊ | 18310/48008 [1:02:39<1:53:13, 4.37it/s] 38%|███▊ | 18311/48008 [1:02:39<1:48:17, 4.57it/s] 38%|███▊ | 18312/48008 [1:02:39<1:43:18, 4.79it/s] 38%|███▊ | 18313/48008 [1:02:40<1:41:31, 4.87it/s] 38%|███▊ | 18314/48008 [1:02:40<1:39:31, 4.97it/s] 38%|███▊ | 18315/48008 [1:02:40<1:38:17, 5.03it/s] 38%|███▊ | 18316/48008 [1:02:40<1:38:21, 5.03it/s] 38%|███▊ | 18317/48008 [1:02:40<1:44:02, 4.76it/s] 38%|███▊ | 18318/48008 [1:02:41<1:42:41, 4.82it/s] 38%|███▊ | 18319/48008 [1:02:41<1:40:22, 4.93it/s] 38%|███▊ | 18320/48008 [1:02:41<1:41:16, 4.89it/s] 38%|███▊ | 18321/48008 [1:02:41<1:39:45, 4.96it/s] 38%|███▊ | 18322/48008 [1:02:41<1:39:57, 4.95it/s] 38%|███▊ | 18323/48008 [1:02:42<1:38:33, 5.02it/s] 38%|███▊ | 18324/48008 [1:02:42<1:37:22, 5.08it/s] 38%|███▊ | 18325/48008 [1:02:42<1:36:36, 5.12it/s] 38%|███▊ | 18326/48008 [1:02:42<1:35:44, 5.17it/s] 38%|███▊ | 18327/48008 [1:02:42<1:37:12, 5.09it/s] 38%|███▊ | 18328/48008 [1:02:43<1:38:19, 5.03it/s] 38%|███▊ | 18329/48008 [1:02:43<1:36:56, 5.10it/s] 38%|███▊ | 18330/48008 [1:02:43<1:36:19, 5.14it/s] 38%|███▊ | 18331/48008 [1:02:43<1:36:35, 5.12it/s] 38%|███▊ | 18332/48008 [1:02:43<1:37:01, 5.10it/s] 38%|███▊ | 18333/48008 [1:02:44<1:37:37, 5.07it/s] 38%|███▊ | 18334/48008 [1:02:44<1:36:30, 5.12it/s] 38%|███▊ | 18335/48008 [1:02:44<1:35:53, 5.16it/s] 38%|███▊ | 18336/48008 [1:02:44<1:35:16, 5.19it/s] 38%|███▊ | 18337/48008 [1:02:44<1:35:03, 5.20it/s] 38%|███▊ | 18338/48008 [1:02:45<1:37:30, 5.07it/s] 38%|███▊ | 18339/48008 [1:02:45<1:37:23, 5.08it/s] 38%|███▊ | 18340/48008 [1:02:45<1:36:44, 5.11it/s] 38%|███▊ | 18341/48008 [1:02:45<1:35:39, 5.17it/s] 38%|███▊ | 18342/48008 [1:02:45<1:35:54, 5.16it/s] 38%|███▊ | 18343/48008 [1:02:46<1:37:08, 5.09it/s] 38%|███▊ | 18344/48008 [1:02:46<1:36:26, 5.13it/s] 38%|███▊ | 18345/48008 [1:02:46<1:35:52, 5.16it/s] 38%|███▊ | 18346/48008 [1:02:46<1:35:12, 5.19it/s] 38%|███▊ | 18347/48008 [1:02:46<1:34:39, 5.22it/s] 38%|███▊ | 18348/48008 [1:02:47<1:34:32, 5.23it/s] 38%|███▊ | 18349/48008 [1:02:47<1:36:26, 5.13it/s] 38%|███▊ | 18350/48008 [1:02:47<1:35:41, 5.17it/s] {'loss': 0.2994, 'grad_norm': 0.36328125, 'learning_rate': 0.0001235585735710715, 'epoch': 0.38} 38%|███▊ | 18350/48008 [1:02:47<1:35:41, 5.17it/s] 38%|███▊ | 18351/48008 [1:02:47<1:35:45, 5.16it/s] 38%|███▊ | 18352/48008 [1:02:47<1:38:08, 5.04it/s] 38%|███▊ | 18353/48008 [1:02:48<1:36:51, 5.10it/s] 38%|███▊ | 18354/48008 [1:02:48<1:37:21, 5.08it/s] 38%|███▊ | 18355/48008 [1:02:48<1:36:49, 5.10it/s] 38%|███▊ | 18356/48008 [1:02:49<3:40:10, 2.24it/s] 38%|███▊ | 18357/48008 [1:02:49<3:01:34, 2.72it/s] 38%|███▊ | 18358/48008 [1:02:49<2:35:29, 3.18it/s] 38%|███▊ | 18359/48008 [1:02:50<2:17:00, 3.61it/s] 38%|███▊ | 18360/48008 [1:02:50<2:05:18, 3.94it/s] 38%|███▊ | 18361/48008 [1:02:50<1:55:43, 4.27it/s] 38%|███▊ | 18362/48008 [1:02:51<3:20:27, 2.46it/s] 38%|███▊ | 18363/48008 [1:02:51<2:48:23, 2.93it/s] 38%|███▊ | 18364/48008 [1:02:51<2:25:07, 3.40it/s] 38%|███▊ | 18365/48008 [1:02:51<2:11:01, 3.77it/s] 38%|███▊ | 18366/48008 [1:02:51<1:59:14, 4.14it/s] 38%|███▊ | 18367/48008 [1:02:52<1:51:25, 4.43it/s] 38%|███▊ | 18368/48008 [1:02:52<3:17:19, 2.50it/s] 38%|███▊ | 18369/48008 [1:02:53<2:49:10, 2.92it/s] 38%|███▊ | 18370/48008 [1:02:53<2:55:48, 2.81it/s] 38%|███▊ | 18371/48008 [1:02:53<2:31:38, 3.26it/s] 38%|███▊ | 18372/48008 [1:02:53<2:14:17, 3.68it/s] 38%|███▊ | 18373/48008 [1:02:54<2:01:35, 4.06it/s] 38%|███▊ | 18374/48008 [1:02:54<1:53:26, 4.35it/s] 38%|███▊ | 18375/48008 [1:02:54<1:46:37, 4.63it/s] 38%|███▊ | 18376/48008 [1:02:54<1:42:51, 4.80it/s] 38%|███▊ | 18377/48008 [1:02:54<1:39:27, 4.97it/s] 38%|███▊ | 18378/48008 [1:02:55<1:38:00, 5.04it/s] 38%|███▊ | 18379/48008 [1:02:55<1:35:55, 5.15it/s] 38%|███▊ | 18380/48008 [1:02:55<1:36:58, 5.09it/s] 38%|███▊ | 18381/48008 [1:02:55<1:37:46, 5.05it/s] 38%|███▊ | 18382/48008 [1:02:55<1:37:01, 5.09it/s] 38%|███▊ | 18383/48008 [1:02:56<1:36:03, 5.14it/s] 38%|███▊ | 18384/48008 [1:02:56<1:36:17, 5.13it/s] 38%|███▊ | 18385/48008 [1:02:56<1:36:03, 5.14it/s] 38%|███▊ | 18386/48008 [1:02:56<1:36:05, 5.14it/s] 38%|███▊ | 18387/48008 [1:02:56<1:34:44, 5.21it/s] 38%|███▊ | 18388/48008 [1:02:57<1:35:22, 5.18it/s] 38%|███▊ | 18389/48008 [1:02:57<1:35:10, 5.19it/s] 38%|███▊ | 18390/48008 [1:02:57<1:34:44, 5.21it/s] 38%|███▊ | 18391/48008 [1:02:57<1:33:44, 5.27it/s] 38%|███▊ | 18392/48008 [1:02:57<1:35:23, 5.17it/s] 38%|███▊ | 18393/48008 [1:02:57<1:34:32, 5.22it/s] 38%|███▊ | 18394/48008 [1:02:58<1:35:43, 5.16it/s] 38%|███▊ | 18395/48008 [1:02:58<1:34:46, 5.21it/s] 38%|███▊ | 18396/48008 [1:02:58<1:35:40, 5.16it/s] 38%|███▊ | 18397/48008 [1:02:58<1:35:35, 5.16it/s] 38%|███▊ | 18398/48008 [1:02:59<3:08:35, 2.62it/s] 38%|███▊ | 18399/48008 [1:02:59<2:47:37, 2.94it/s] 38%|███▊ | 18400/48008 [1:02:59<2:25:25, 3.39it/s] {'loss': 0.3062, 'grad_norm': 0.431640625, 'learning_rate': 0.0001233502749541743, 'epoch': 0.38} 38%|███▊ | 18400/48008 [1:03:00<2:25:25, 3.39it/s] 38%|███▊ | 18401/48008 [1:03:00<2:12:13, 3.73it/s] 38%|███▊ | 18402/48008 [1:03:00<2:01:07, 4.07it/s] 38%|███▊ | 18403/48008 [1:03:00<1:52:32, 4.38it/s] 38%|███▊ | 18404/48008 [1:03:00<1:53:49, 4.33it/s] 38%|███▊ | 18405/48008 [1:03:00<1:47:24, 4.59it/s] 38%|███▊ | 18406/48008 [1:03:01<1:44:08, 4.74it/s] 38%|███▊ | 18407/48008 [1:03:01<1:40:51, 4.89it/s] 38%|███▊ | 18408/48008 [1:03:01<1:38:46, 4.99it/s] 38%|███▊ | 18409/48008 [1:03:01<1:39:10, 4.97it/s] 38%|███▊ | 18410/48008 [1:03:02<2:06:45, 3.89it/s] 38%|███▊ | 18411/48008 [1:03:02<1:56:48, 4.22it/s] 38%|███▊ | 18412/48008 [1:03:02<1:49:35, 4.50it/s] 38%|███▊ | 18413/48008 [1:03:02<1:45:09, 4.69it/s] 38%|███▊ | 18414/48008 [1:03:02<1:41:54, 4.84it/s] 38%|███▊ | 18415/48008 [1:03:03<1:40:49, 4.89it/s] 38%|███▊ | 18416/48008 [1:03:03<1:40:31, 4.91it/s] 38%|███▊ | 18417/48008 [1:03:03<1:56:04, 4.25it/s] 38%|███▊ | 18418/48008 [1:03:03<1:49:05, 4.52it/s] 38%|███▊ | 18419/48008 [1:03:04<1:47:43, 4.58it/s] 38%|███▊ | 18420/48008 [1:03:04<1:44:08, 4.74it/s] 38%|███▊ | 18421/48008 [1:03:04<1:41:00, 4.88it/s] 38%|███▊ | 18422/48008 [1:03:04<1:56:09, 4.24it/s] 38%|███▊ | 18423/48008 [1:03:04<1:48:58, 4.52it/s] 38%|███▊ | 18424/48008 [1:03:05<1:43:46, 4.75it/s] 38%|███▊ | 18425/48008 [1:03:05<1:40:46, 4.89it/s] 38%|███▊ | 18426/48008 [1:03:05<1:38:54, 4.98it/s] 38%|███▊ | 18427/48008 [1:03:05<1:37:53, 5.04it/s] 38%|███▊ | 18428/48008 [1:03:05<1:35:52, 5.14it/s] 38%|███▊ | 18429/48008 [1:03:06<1:35:34, 5.16it/s] 38%|███▊ | 18430/48008 [1:03:06<1:35:20, 5.17it/s] 38%|███▊ | 18431/48008 [1:03:06<1:35:09, 5.18it/s] 38%|███▊ | 18432/48008 [1:03:06<1:34:56, 5.19it/s] 38%|███▊ | 18433/48008 [1:03:06<1:37:10, 5.07it/s] 38%|███▊ | 18434/48008 [1:03:07<1:36:14, 5.12it/s] 38%|███▊ | 18435/48008 [1:03:07<1:35:45, 5.15it/s] 38%|███▊ | 18436/48008 [1:03:07<1:37:15, 5.07it/s] 38%|███▊ | 18437/48008 [1:03:07<1:36:15, 5.12it/s] 38%|███▊ | 18438/48008 [1:03:07<1:35:54, 5.14it/s] 38%|███▊ | 18439/48008 [1:03:07<1:34:59, 5.19it/s] 38%|███▊ | 18440/48008 [1:03:08<1:34:48, 5.20it/s] 38%|███▊ | 18441/48008 [1:03:08<1:34:08, 5.23it/s] 38%|███▊ | 18442/48008 [1:03:08<1:35:21, 5.17it/s] 38%|███▊ | 18443/48008 [1:03:08<1:35:20, 5.17it/s] 38%|███▊ | 18444/48008 [1:03:08<1:34:37, 5.21it/s] 38%|███▊ | 18445/48008 [1:03:09<1:35:57, 5.13it/s] 38%|███▊ | 18446/48008 [1:03:09<1:35:02, 5.18it/s] 38%|███▊ | 18447/48008 [1:03:09<1:34:23, 5.22it/s] 38%|███▊ | 18448/48008 [1:03:09<1:33:41, 5.26it/s] 38%|███▊ | 18449/48008 [1:03:09<1:33:04, 5.29it/s] 38%|███▊ | 18450/48008 [1:03:10<1:32:33, 5.32it/s] {'loss': 0.3043, 'grad_norm': 0.1953125, 'learning_rate': 0.00012314197633727714, 'epoch': 0.38} 38%|███▊ | 18450/48008 [1:03:10<1:32:33, 5.32it/s] 38%|███▊ | 18451/48008 [1:03:10<1:50:21, 4.46it/s] 38%|███▊ | 18452/48008 [1:03:10<1:45:37, 4.66it/s] 38%|███▊ | 18453/48008 [1:03:10<1:42:21, 4.81it/s] 38%|███▊ | 18454/48008 [1:03:10<1:42:44, 4.79it/s] 38%|███▊ | 18455/48008 [1:03:11<1:47:11, 4.60it/s] 38%|███▊ | 18456/48008 [1:03:11<1:43:29, 4.76it/s] 38%|███▊ | 18457/48008 [1:03:11<1:47:25, 4.58it/s] 38%|███▊ | 18458/48008 [1:03:11<1:43:07, 4.78it/s] 38%|███▊ | 18459/48008 [1:03:12<1:47:02, 4.60it/s] 38%|███▊ | 18460/48008 [1:03:12<1:42:43, 4.79it/s] 38%|███▊ | 18461/48008 [1:03:12<1:42:05, 4.82it/s] 38%|███▊ | 18462/48008 [1:03:12<1:39:36, 4.94it/s] 38%|███▊ | 18463/48008 [1:03:12<1:38:07, 5.02it/s] 38%|███▊ | 18464/48008 [1:03:13<1:43:23, 4.76it/s] 38%|███▊ | 18465/48008 [1:03:13<1:43:27, 4.76it/s] 38%|███▊ | 18466/48008 [1:03:13<1:40:08, 4.92it/s] 38%|███▊ | 18467/48008 [1:03:13<1:39:58, 4.93it/s] 38%|███▊ | 18468/48008 [1:03:13<1:38:45, 4.98it/s] 38%|███▊ | 18469/48008 [1:03:14<1:38:15, 5.01it/s] 38%|███▊ | 18470/48008 [1:03:14<1:37:48, 5.03it/s] 38%|███▊ | 18471/48008 [1:03:14<1:35:38, 5.15it/s] 38%|███▊ | 18472/48008 [1:03:14<1:36:52, 5.08it/s] 38%|███▊ | 18473/48008 [1:03:14<1:36:01, 5.13it/s] 38%|███▊ | 18474/48008 [1:03:15<1:35:24, 5.16it/s] 38%|███▊ | 18475/48008 [1:03:15<1:34:40, 5.20it/s] 38%|███▊ | 18476/48008 [1:03:15<1:35:12, 5.17it/s] 38%|███▊ | 18477/48008 [1:03:15<1:35:26, 5.16it/s] 38%|███▊ | 18478/48008 [1:03:15<1:35:20, 5.16it/s] 38%|███▊ | 18479/48008 [1:03:16<1:36:29, 5.10it/s] 38%|███▊ | 18480/48008 [1:03:16<1:35:58, 5.13it/s] 38%|███▊ | 18481/48008 [1:03:16<1:35:46, 5.14it/s] 38%|███▊ | 18482/48008 [1:03:16<1:35:23, 5.16it/s] 38%|███▊ | 18483/48008 [1:03:16<1:34:54, 5.18it/s] 39%|███▊ | 18484/48008 [1:03:16<1:35:26, 5.16it/s] 39%|███▊ | 18485/48008 [1:03:17<1:35:05, 5.17it/s] 39%|███▊ | 18486/48008 [1:03:17<1:35:32, 5.15it/s] 39%|███▊ | 18487/48008 [1:03:17<1:36:14, 5.11it/s] 39%|███▊ | 18488/48008 [1:03:17<1:36:41, 5.09it/s] 39%|███▊ | 18489/48008 [1:03:18<1:53:13, 4.35it/s] 39%|███▊ | 18490/48008 [1:03:18<1:48:16, 4.54it/s] 39%|███▊ | 18491/48008 [1:03:18<1:45:06, 4.68it/s] 39%|███▊ | 18492/48008 [1:03:18<1:41:55, 4.83it/s] 39%|███▊ | 18493/48008 [1:03:18<1:41:45, 4.83it/s] 39%|███▊ | 18494/48008 [1:03:19<1:41:29, 4.85it/s] 39%|███▊ | 18495/48008 [1:03:19<1:39:21, 4.95it/s] 39%|███▊ | 18496/48008 [1:03:19<1:37:40, 5.04it/s] 39%|███▊ | 18497/48008 [1:03:19<1:36:05, 5.12it/s] 39%|███▊ | 18498/48008 [1:03:19<1:42:08, 4.81it/s] 39%|███▊ | 18499/48008 [1:03:20<1:39:42, 4.93it/s] 39%|███▊ | 18500/48008 [1:03:20<1:38:03, 5.02it/s] {'loss': 0.2999, 'grad_norm': 0.36328125, 'learning_rate': 0.00012293367772037993, 'epoch': 0.39} 39%|███▊ | 18500/48008 [1:03:20<1:38:03, 5.02it/s] 39%|███▊ | 18501/48008 [1:03:20<1:37:23, 5.05it/s] 39%|███▊ | 18502/48008 [1:03:20<1:39:24, 4.95it/s] 39%|███▊ | 18503/48008 [1:03:20<1:37:41, 5.03it/s] 39%|███▊ | 18504/48008 [1:03:21<1:36:14, 5.11it/s] 39%|███▊ | 18505/48008 [1:03:21<1:35:23, 5.16it/s] 39%|███▊ | 18506/48008 [1:03:21<1:52:35, 4.37it/s] 39%|███▊ | 18507/48008 [1:03:21<1:47:44, 4.56it/s] 39%|███▊ | 18508/48008 [1:03:21<1:43:44, 4.74it/s] 39%|███▊ | 18509/48008 [1:03:22<1:42:13, 4.81it/s] 39%|███▊ | 18510/48008 [1:03:22<1:39:06, 4.96it/s] 39%|███▊ | 18511/48008 [1:03:22<1:37:43, 5.03it/s] 39%|███▊ | 18512/48008 [1:03:22<1:36:39, 5.09it/s] 39%|███▊ | 18513/48008 [1:03:23<1:53:05, 4.35it/s] 39%|███▊ | 18514/48008 [1:03:23<1:47:18, 4.58it/s] 39%|███▊ | 18515/48008 [1:03:23<1:45:36, 4.65it/s] 39%|███▊ | 18516/48008 [1:03:23<1:41:48, 4.83it/s] 39%|███▊ | 18517/48008 [1:03:23<1:46:20, 4.62it/s] 39%|███▊ | 18518/48008 [1:03:24<1:43:28, 4.75it/s] 39%|███▊ | 18519/48008 [1:03:24<1:40:38, 4.88it/s] 39%|███▊ | 18520/48008 [1:03:24<1:40:14, 4.90it/s] 39%|███▊ | 18521/48008 [1:03:24<1:55:56, 4.24it/s] 39%|███▊ | 18522/48008 [1:03:24<1:48:56, 4.51it/s] 39%|███▊ | 18523/48008 [1:03:25<1:51:12, 4.42it/s] 39%|███▊ | 18524/48008 [1:03:25<1:45:21, 4.66it/s] 39%|███▊ | 18525/48008 [1:03:25<1:41:37, 4.84it/s] 39%|███▊ | 18526/48008 [1:03:25<1:38:37, 4.98it/s] 39%|███▊ | 18527/48008 [1:03:25<1:38:12, 5.00it/s] 39%|███▊ | 18528/48008 [1:03:26<1:38:33, 4.99it/s] 39%|███▊ | 18529/48008 [1:03:26<1:37:21, 5.05it/s] 39%|███▊ | 18530/48008 [1:03:26<1:36:36, 5.09it/s] 39%|███▊ | 18531/48008 [1:03:26<1:35:11, 5.16it/s] 39%|███▊ | 18532/48008 [1:03:26<1:35:01, 5.17it/s] 39%|███▊ | 18533/48008 [1:03:27<1:36:30, 5.09it/s] 39%|███▊ | 18534/48008 [1:03:27<1:35:40, 5.13it/s] 39%|███▊ | 18535/48008 [1:03:27<1:34:17, 5.21it/s] 39%|███▊ | 18536/48008 [1:03:27<1:33:07, 5.27it/s] 39%|███▊ | 18537/48008 [1:03:27<1:33:06, 5.28it/s] 39%|███▊ | 18538/48008 [1:03:28<1:35:31, 5.14it/s] 39%|███▊ | 18539/48008 [1:03:28<1:35:26, 5.15it/s] 39%|███▊ | 18540/48008 [1:03:28<1:36:23, 5.10it/s] 39%|███▊ | 18541/48008 [1:03:28<1:38:59, 4.96it/s] 39%|███▊ | 18542/48008 [1:03:28<1:36:44, 5.08it/s] 39%|███▊ | 18543/48008 [1:03:29<1:37:42, 5.03it/s] 39%|███▊ | 18544/48008 [1:03:29<1:37:31, 5.04it/s] 39%|███▊ | 18545/48008 [1:03:29<1:38:43, 4.97it/s] 39%|███▊ | 18546/48008 [1:03:29<1:37:58, 5.01it/s] 39%|███▊ | 18547/48008 [1:03:29<1:37:05, 5.06it/s] 39%|███▊ | 18548/48008 [1:03:30<1:35:56, 5.12it/s] 39%|███▊ | 18549/48008 [1:03:30<1:34:53, 5.17it/s] 39%|███▊ | 18550/48008 [1:03:30<1:41:21, 4.84it/s] {'loss': 0.3016, 'grad_norm': 0.140625, 'learning_rate': 0.00012272537910348275, 'epoch': 0.39} 39%|███▊ | 18550/48008 [1:03:30<1:41:21, 4.84it/s] 39%|███▊ | 18551/48008 [1:03:30<1:39:08, 4.95it/s] 39%|███▊ | 18552/48008 [1:03:30<1:37:44, 5.02it/s] 39%|███▊ | 18553/48008 [1:03:31<1:36:41, 5.08it/s] 39%|███▊ | 18554/48008 [1:03:31<1:34:21, 5.20it/s] 39%|███▊ | 18555/48008 [1:03:31<1:35:52, 5.12it/s] 39%|███▊ | 18556/48008 [1:03:31<1:35:05, 5.16it/s] 39%|███▊ | 18557/48008 [1:03:31<1:37:23, 5.04it/s] 39%|███▊ | 18558/48008 [1:03:32<1:36:07, 5.11it/s] 39%|███▊ | 18559/48008 [1:03:32<1:35:27, 5.14it/s] 39%|███▊ | 18560/48008 [1:03:32<1:35:06, 5.16it/s] 39%|███▊ | 18561/48008 [1:03:32<1:35:28, 5.14it/s] 39%|███▊ | 18562/48008 [1:03:32<1:34:01, 5.22it/s] 39%|███▊ | 18563/48008 [1:03:32<1:34:31, 5.19it/s] 39%|███▊ | 18564/48008 [1:03:33<1:51:08, 4.42it/s] 39%|███▊ | 18565/48008 [1:03:33<1:47:58, 4.54it/s] 39%|███▊ | 18566/48008 [1:03:33<1:43:55, 4.72it/s] 39%|███▊ | 18567/48008 [1:03:33<1:47:46, 4.55it/s] 39%|███▊ | 18568/48008 [1:03:34<1:43:19, 4.75it/s] 39%|███▊ | 18569/48008 [1:03:34<1:41:10, 4.85it/s] 39%|███▊ | 18570/48008 [1:03:34<1:39:47, 4.92it/s] 39%|███▊ | 18571/48008 [1:03:34<1:37:58, 5.01it/s] 39%|███▊ | 18572/48008 [1:03:34<1:36:49, 5.07it/s] 39%|███▊ | 18573/48008 [1:03:35<1:38:46, 4.97it/s] 39%|███▊ | 18574/48008 [1:03:35<1:38:05, 5.00it/s] 39%|███▊ | 18575/48008 [1:03:35<1:37:27, 5.03it/s] 39%|███▊ | 18576/48008 [1:03:35<1:35:22, 5.14it/s] 39%|███▊ | 18577/48008 [1:03:35<1:34:37, 5.18it/s] 39%|███▊ | 18578/48008 [1:03:36<1:37:01, 5.06it/s] 39%|███▊ | 18579/48008 [1:03:36<1:36:10, 5.10it/s] 39%|███▊ | 18580/48008 [1:03:36<1:35:40, 5.13it/s] 39%|███▊ | 18581/48008 [1:03:36<1:35:07, 5.16it/s] 39%|███▊ | 18582/48008 [1:03:36<1:34:13, 5.21it/s] 39%|███▊ | 18583/48008 [1:03:37<1:35:37, 5.13it/s] 39%|███▊ | 18584/48008 [1:03:37<1:35:50, 5.12it/s] 39%|███▊ | 18585/48008 [1:03:37<1:35:09, 5.15it/s] 39%|███▊ | 18586/48008 [1:03:37<1:35:53, 5.11it/s] 39%|███▊ | 18587/48008 [1:03:37<1:35:29, 5.13it/s] 39%|███▊ | 18588/48008 [1:03:38<1:36:37, 5.07it/s] 39%|███▊ | 18589/48008 [1:03:38<1:37:29, 5.03it/s] 39%|███▊ | 18590/48008 [1:03:38<1:36:22, 5.09it/s] 39%|███▊ | 18591/48008 [1:03:38<1:35:39, 5.13it/s] 39%|███▊ | 18592/48008 [1:03:38<1:36:36, 5.07it/s] 39%|███▊ | 18593/48008 [1:03:39<1:35:03, 5.16it/s] 39%|███▊ | 18594/48008 [1:03:39<1:51:37, 4.39it/s] 39%|███▊ | 18595/48008 [1:03:39<1:46:14, 4.61it/s] 39%|███▊ | 18596/48008 [1:03:39<1:43:04, 4.76it/s] 39%|███▊ | 18597/48008 [1:03:39<1:41:50, 4.81it/s] 39%|███▊ | 18598/48008 [1:03:40<1:39:29, 4.93it/s] 39%|███▊ | 18599/48008 [1:03:40<1:44:11, 4.70it/s] 39%|███▊ | 18600/48008 [1:03:40<1:47:43, 4.55it/s] {'loss': 0.3, 'grad_norm': 0.1396484375, 'learning_rate': 0.00012251708048658557, 'epoch': 0.39} 39%|███▊ | 18600/48008 [1:03:40<1:47:43, 4.55it/s] 39%|███▊ | 18601/48008 [1:03:40<1:43:02, 4.76it/s] 39%|███▊ | 18602/48008 [1:03:40<1:40:21, 4.88it/s] 39%|███▊ | 18603/48008 [1:03:41<1:38:24, 4.98it/s] 39%|███▉ | 18604/48008 [1:03:41<1:38:36, 4.97it/s] 39%|███▉ | 18605/48008 [1:03:41<1:37:29, 5.03it/s] 39%|███▉ | 18606/48008 [1:03:41<1:36:43, 5.07it/s] 39%|███▉ | 18607/48008 [1:03:41<1:37:31, 5.02it/s] 39%|███▉ | 18608/48008 [1:03:42<1:38:05, 5.00it/s] 39%|███▉ | 18609/48008 [1:03:42<1:37:24, 5.03it/s] 39%|███▉ | 18610/48008 [1:03:42<1:36:24, 5.08it/s] 39%|███▉ | 18611/48008 [1:03:42<1:35:38, 5.12it/s] 39%|███▉ | 18612/48008 [1:03:42<1:35:07, 5.15it/s] 39%|███▉ | 18613/48008 [1:03:43<1:34:10, 5.20it/s] 39%|███▉ | 18614/48008 [1:03:43<1:34:43, 5.17it/s] 39%|███▉ | 18615/48008 [1:03:43<1:34:08, 5.20it/s] 39%|███▉ | 18616/48008 [1:03:43<1:33:41, 5.23it/s] 39%|███▉ | 18617/48008 [1:03:43<1:35:17, 5.14it/s] 39%|███▉ | 18618/48008 [1:03:44<1:33:52, 5.22it/s] 39%|███▉ | 18619/48008 [1:03:44<1:33:54, 5.22it/s] 39%|███▉ | 18620/48008 [1:03:44<1:33:52, 5.22it/s] 39%|███▉ | 18621/48008 [1:03:44<1:35:22, 5.13it/s] 39%|███▉ | 18622/48008 [1:03:44<1:34:33, 5.18it/s] 39%|███▉ | 18623/48008 [1:03:45<1:37:03, 5.05it/s] 39%|███▉ | 18624/48008 [1:03:45<1:35:55, 5.11it/s] 39%|███▉ | 18625/48008 [1:03:45<1:36:39, 5.07it/s] 39%|███▉ | 18626/48008 [1:03:45<1:35:41, 5.12it/s] 39%|███▉ | 18627/48008 [1:03:45<1:35:02, 5.15it/s] 39%|███▉ | 18628/48008 [1:03:46<1:37:12, 5.04it/s] 39%|███▉ | 18629/48008 [1:03:46<1:35:59, 5.10it/s] 39%|███▉ | 18630/48008 [1:03:46<1:35:13, 5.14it/s] 39%|███▉ | 18631/48008 [1:03:46<1:33:50, 5.22it/s] 39%|███▉ | 18632/48008 [1:03:46<1:40:59, 4.85it/s] 39%|███▉ | 18633/48008 [1:03:47<1:38:19, 4.98it/s] 39%|███▉ | 18634/48008 [1:03:47<1:36:06, 5.09it/s] 39%|███▉ | 18635/48008 [1:03:47<1:35:25, 5.13it/s] 39%|███▉ | 18636/48008 [1:03:47<1:36:38, 5.07it/s] 39%|███▉ | 18637/48008 [1:03:47<1:35:27, 5.13it/s] 39%|███▉ | 18638/48008 [1:03:47<1:35:00, 5.15it/s] 39%|███▉ | 18639/48008 [1:03:48<1:34:48, 5.16it/s] 39%|███▉ | 18640/48008 [1:03:48<1:34:09, 5.20it/s] 39%|███▉ | 18641/48008 [1:03:48<1:33:37, 5.23it/s] 39%|███▉ | 18642/48008 [1:03:48<1:32:50, 5.27it/s] 39%|███▉ | 18643/48008 [1:03:48<1:33:13, 5.25it/s] 39%|███▉ | 18644/48008 [1:03:49<1:32:59, 5.26it/s] 39%|███▉ | 18645/48008 [1:03:49<1:32:59, 5.26it/s] 39%|███▉ | 18646/48008 [1:03:49<2:01:51, 4.02it/s] 39%|███▉ | 18647/48008 [1:03:49<1:53:10, 4.32it/s] 39%|███▉ | 18648/48008 [1:03:50<1:47:10, 4.57it/s] 39%|███▉ | 18649/48008 [1:03:50<1:45:04, 4.66it/s] 39%|███▉ | 18650/48008 [1:03:50<1:43:01, 4.75it/s] {'loss': 0.3022, 'grad_norm': 0.373046875, 'learning_rate': 0.0001223087818696884, 'epoch': 0.39} 39%|███▉ | 18650/48008 [1:03:50<1:43:01, 4.75it/s] 39%|███▉ | 18651/48008 [1:03:50<1:42:55, 4.75it/s] 39%|███▉ | 18652/48008 [1:03:50<1:43:06, 4.75it/s] 39%|███▉ | 18653/48008 [1:03:51<1:40:10, 4.88it/s] 39%|███▉ | 18654/48008 [1:03:51<1:37:38, 5.01it/s] 39%|███▉ | 18655/48008 [1:03:51<1:53:11, 4.32it/s] 39%|███▉ | 18656/48008 [1:03:51<1:47:14, 4.56it/s] 39%|███▉ | 18657/48008 [1:03:51<1:42:56, 4.75it/s] 39%|███▉ | 18658/48008 [1:03:52<1:39:46, 4.90it/s] 39%|███▉ | 18659/48008 [1:03:52<1:39:02, 4.94it/s] 39%|███▉ | 18660/48008 [1:03:52<1:36:59, 5.04it/s] 39%|███▉ | 18661/48008 [1:03:52<1:37:54, 5.00it/s] 39%|███▉ | 18662/48008 [1:03:52<1:35:54, 5.10it/s] 39%|███▉ | 18663/48008 [1:03:53<1:34:44, 5.16it/s] 39%|███▉ | 18664/48008 [1:03:53<1:34:25, 5.18it/s] 39%|███▉ | 18665/48008 [1:03:53<1:34:43, 5.16it/s] 39%|███▉ | 18666/48008 [1:03:53<1:34:22, 5.18it/s] 39%|███▉ | 18667/48008 [1:03:53<1:35:51, 5.10it/s] 39%|███▉ | 18668/48008 [1:03:54<1:36:02, 5.09it/s] 39%|███▉ | 18669/48008 [1:03:54<1:34:32, 5.17it/s] 39%|███▉ | 18670/48008 [1:03:54<1:33:52, 5.21it/s] 39%|███▉ | 18671/48008 [1:03:54<1:33:10, 5.25it/s] 39%|███▉ | 18672/48008 [1:03:54<1:34:54, 5.15it/s] 39%|███▉ | 18673/48008 [1:03:55<1:34:26, 5.18it/s] 39%|███▉ | 18674/48008 [1:03:55<1:35:44, 5.11it/s] 39%|███▉ | 18675/48008 [1:03:55<1:34:35, 5.17it/s] 39%|███▉ | 18676/48008 [1:03:55<1:33:44, 5.21it/s] 39%|███▉ | 18677/48008 [1:03:55<1:35:39, 5.11it/s] 39%|███▉ | 18678/48008 [1:03:56<1:35:02, 5.14it/s] 39%|███▉ | 18679/48008 [1:03:56<1:33:54, 5.21it/s] 39%|███▉ | 18680/48008 [1:03:56<1:33:49, 5.21it/s] 39%|███▉ | 18681/48008 [1:03:56<1:34:40, 5.16it/s] 39%|███▉ | 18682/48008 [1:03:56<1:33:38, 5.22it/s] 39%|███▉ | 18683/48008 [1:03:56<1:33:40, 5.22it/s] 39%|███▉ | 18684/48008 [1:03:57<1:32:36, 5.28it/s] 39%|███▉ | 18685/48008 [1:03:57<1:31:22, 5.35it/s] 39%|███▉ | 18686/48008 [1:03:57<1:32:18, 5.29it/s] 39%|███▉ | 18687/48008 [1:03:57<1:33:13, 5.24it/s] 39%|███▉ | 18688/48008 [1:03:57<1:33:18, 5.24it/s] 39%|███▉ | 18689/48008 [1:03:58<1:32:50, 5.26it/s] 39%|███▉ | 18690/48008 [1:03:58<1:33:04, 5.25it/s] 39%|███▉ | 18691/48008 [1:03:58<2:01:57, 4.01it/s] 39%|███▉ | 18692/48008 [1:03:58<2:00:19, 4.06it/s] 39%|███▉ | 18693/48008 [1:03:59<1:51:55, 4.37it/s] 39%|███▉ | 18694/48008 [1:03:59<1:46:16, 4.60it/s] 39%|███▉ | 18695/48008 [1:03:59<1:42:25, 4.77it/s] 39%|███▉ | 18696/48008 [1:03:59<1:40:36, 4.86it/s] 39%|███▉ | 18697/48008 [1:03:59<1:41:05, 4.83it/s] 39%|███▉ | 18698/48008 [1:04:00<1:38:02, 4.98it/s] 39%|███▉ | 18699/48008 [1:04:00<1:37:22, 5.02it/s] 39%|███▉ | 18700/48008 [1:04:00<1:36:10, 5.08it/s] {'loss': 0.3025, 'grad_norm': 0.1396484375, 'learning_rate': 0.0001221004832527912, 'epoch': 0.39} 39%|███▉ | 18700/48008 [1:04:00<1:36:10, 5.08it/s] 39%|███▉ | 18701/48008 [1:04:00<1:37:07, 5.03it/s] 39%|███▉ | 18702/48008 [1:04:00<1:36:09, 5.08it/s] 39%|███▉ | 18703/48008 [1:04:01<1:42:50, 4.75it/s] 39%|███▉ | 18704/48008 [1:04:01<1:41:39, 4.80it/s] 39%|███▉ | 18705/48008 [1:04:01<1:42:03, 4.79it/s] 39%|███▉ | 18706/48008 [1:04:01<1:39:50, 4.89it/s] 39%|███▉ | 18707/48008 [1:04:01<1:37:36, 5.00it/s] 39%|███▉ | 18708/48008 [1:04:02<1:37:16, 5.02it/s] 39%|███▉ | 18709/48008 [1:04:02<1:35:45, 5.10it/s] 39%|███▉ | 18710/48008 [1:04:02<1:35:20, 5.12it/s] 39%|███▉ | 18711/48008 [1:04:02<1:42:09, 4.78it/s] 39%|███▉ | 18712/48008 [1:04:02<1:39:25, 4.91it/s] 39%|███▉ | 18713/48008 [1:04:03<1:37:52, 4.99it/s] 39%|███▉ | 18714/48008 [1:04:03<1:39:15, 4.92it/s] 39%|███▉ | 18715/48008 [1:04:03<1:36:34, 5.06it/s] 39%|███▉ | 18716/48008 [1:04:03<1:38:25, 4.96it/s] 39%|███▉ | 18717/48008 [1:04:03<1:36:48, 5.04it/s] 39%|███▉ | 18718/48008 [1:04:04<1:36:39, 5.05it/s] 39%|███▉ | 18719/48008 [1:04:04<1:36:35, 5.05it/s] 39%|███▉ | 18720/48008 [1:04:04<1:35:32, 5.11it/s] 39%|███▉ | 18721/48008 [1:04:04<1:34:41, 5.15it/s] 39%|███▉ | 18722/48008 [1:04:04<1:33:49, 5.20it/s] 39%|███▉ | 18723/48008 [1:04:05<1:33:54, 5.20it/s] 39%|███▉ | 18724/48008 [1:04:05<2:02:41, 3.98it/s] 39%|███▉ | 18725/48008 [1:04:05<1:55:55, 4.21it/s] 39%|███▉ | 18726/48008 [1:04:05<1:48:28, 4.50it/s] 39%|███▉ | 18727/48008 [1:04:06<1:43:25, 4.72it/s] 39%|███▉ | 18728/48008 [1:04:06<2:09:03, 3.78it/s] 39%|███▉ | 18729/48008 [1:04:06<1:58:08, 4.13it/s] 39%|███▉ | 18730/48008 [1:04:06<1:51:27, 4.38it/s] 39%|███▉ | 18731/48008 [1:04:07<1:47:59, 4.52it/s] 39%|███▉ | 18732/48008 [1:04:07<1:42:33, 4.76it/s] 39%|███▉ | 18733/48008 [1:04:07<1:39:00, 4.93it/s] 39%|███▉ | 18734/48008 [1:04:07<1:38:05, 4.97it/s] 39%|███▉ | 18735/48008 [1:04:07<1:37:12, 5.02it/s] 39%|███▉ | 18736/48008 [1:04:07<1:37:04, 5.03it/s] 39%|███▉ | 18737/48008 [1:04:08<1:36:04, 5.08it/s] 39%|███▉ | 18738/48008 [1:04:08<1:35:12, 5.12it/s] 39%|███▉ | 18739/48008 [1:04:08<1:34:24, 5.17it/s] 39%|███▉ | 18740/48008 [1:04:08<1:33:56, 5.19it/s] 39%|███▉ | 18741/48008 [1:04:08<1:33:52, 5.20it/s] 39%|███▉ | 18742/48008 [1:04:09<1:33:48, 5.20it/s] 39%|███▉ | 18743/48008 [1:04:09<1:36:16, 5.07it/s] 39%|███▉ | 18744/48008 [1:04:09<1:34:15, 5.17it/s] 39%|███▉ | 18745/48008 [1:04:09<1:34:13, 5.18it/s] 39%|███▉ | 18746/48008 [1:04:09<1:32:55, 5.25it/s] 39%|███▉ | 18747/48008 [1:04:10<1:32:46, 5.26it/s] 39%|███▉ | 18748/48008 [1:04:10<1:33:04, 5.24it/s] 39%|███▉ | 18749/48008 [1:04:10<1:34:34, 5.16it/s] 39%|███▉ | 18750/48008 [1:04:10<1:34:19, 5.17it/s] {'loss': 0.3011, 'grad_norm': 0.375, 'learning_rate': 0.00012189218463589402, 'epoch': 0.39} 39%|███▉ | 18750/48008 [1:04:10<1:34:19, 5.17it/s] 39%|███▉ | 18751/48008 [1:04:10<1:34:33, 5.16it/s] 39%|███▉ | 18752/48008 [1:04:11<1:34:32, 5.16it/s] 39%|███▉ | 18753/48008 [1:04:11<2:02:40, 3.97it/s] 39%|███▉ | 18754/48008 [1:04:11<1:53:38, 4.29it/s] 39%|███▉ | 18755/48008 [1:04:11<1:47:07, 4.55it/s] 39%|███▉ | 18756/48008 [1:04:12<1:44:47, 4.65it/s] 39%|███▉ | 18757/48008 [1:04:12<1:40:27, 4.85it/s] 39%|███▉ | 18758/48008 [1:04:12<1:38:46, 4.94it/s] 39%|███▉ | 18759/48008 [1:04:12<1:37:19, 5.01it/s] 39%|███▉ | 18760/48008 [1:04:12<1:35:42, 5.09it/s] 39%|███▉ | 18761/48008 [1:04:12<1:36:41, 5.04it/s] 39%|███▉ | 18762/48008 [1:04:13<1:33:59, 5.19it/s] 39%|███▉ | 18763/48008 [1:04:13<1:35:13, 5.12it/s] 39%|███▉ | 18764/48008 [1:04:13<1:35:30, 5.10it/s] 39%|███▉ | 18765/48008 [1:04:13<1:35:01, 5.13it/s] 39%|███▉ | 18766/48008 [1:04:13<1:34:55, 5.13it/s] 39%|███▉ | 18767/48008 [1:04:14<1:33:41, 5.20it/s] 39%|███▉ | 18768/48008 [1:04:14<1:33:43, 5.20it/s] 39%|███▉ | 18769/48008 [1:04:14<1:33:58, 5.19it/s] 39%|███▉ | 18770/48008 [1:04:14<1:34:20, 5.16it/s] 39%|███▉ | 18771/48008 [1:04:15<2:02:47, 3.97it/s] 39%|███▉ | 18772/48008 [1:04:15<1:56:05, 4.20it/s] 39%|███▉ | 18773/48008 [1:04:15<1:48:47, 4.48it/s] 39%|███▉ | 18774/48008 [1:04:15<1:43:59, 4.69it/s] 39%|███▉ | 18775/48008 [1:04:15<1:47:35, 4.53it/s] 39%|███▉ | 18776/48008 [1:04:16<1:43:37, 4.70it/s] 39%|███▉ | 18777/48008 [1:04:16<1:42:49, 4.74it/s] 39%|███▉ | 18778/48008 [1:04:16<1:40:25, 4.85it/s] 39%|███▉ | 18779/48008 [1:04:16<1:55:10, 4.23it/s] 39%|███▉ | 18780/48008 [1:04:17<1:47:40, 4.52it/s] 39%|███▉ | 18781/48008 [1:04:17<1:43:13, 4.72it/s] 39%|███▉ | 18782/48008 [1:04:17<1:39:53, 4.88it/s] 39%|███▉ | 18783/48008 [1:04:17<1:37:27, 5.00it/s] 39%|███▉ | 18784/48008 [1:04:17<1:37:43, 4.98it/s] 39%|███▉ | 18785/48008 [1:04:17<1:36:45, 5.03it/s] 39%|███▉ | 18786/48008 [1:04:18<1:36:05, 5.07it/s] 39%|███▉ | 18787/48008 [1:04:18<1:37:23, 5.00it/s] 39%|███▉ | 18788/48008 [1:04:18<1:43:03, 4.73it/s] 39%|███▉ | 18789/48008 [1:04:18<1:41:45, 4.79it/s] 39%|███▉ | 18790/48008 [1:04:19<1:38:54, 4.92it/s] 39%|███▉ | 18791/48008 [1:04:19<1:38:39, 4.94it/s] 39%|███▉ | 18792/48008 [1:04:19<1:37:10, 5.01it/s] 39%|███▉ | 18793/48008 [1:04:19<1:35:32, 5.10it/s] 39%|███▉ | 18794/48008 [1:04:19<1:34:49, 5.13it/s] 39%|███▉ | 18795/48008 [1:04:19<1:34:46, 5.14it/s] 39%|███▉ | 18796/48008 [1:04:20<1:33:17, 5.22it/s] 39%|███▉ | 18797/48008 [1:04:20<1:33:14, 5.22it/s] 39%|███▉ | 18798/48008 [1:04:20<1:33:52, 5.19it/s] 39%|███▉ | 18799/48008 [1:04:20<1:32:49, 5.24it/s] 39%|███▉ | 18800/48008 [1:04:20<1:35:33, 5.09it/s] {'loss': 0.3003, 'grad_norm': 0.240234375, 'learning_rate': 0.00012168388601899683, 'epoch': 0.39} 39%|███▉ | 18800/48008 [1:04:20<1:35:33, 5.09it/s] 39%|███▉ | 18801/48008 [1:04:21<1:36:20, 5.05it/s] 39%|███▉ | 18802/48008 [1:04:21<1:36:47, 5.03it/s] 39%|███▉ | 18803/48008 [1:04:21<1:35:31, 5.10it/s] 39%|███▉ | 18804/48008 [1:04:21<1:36:41, 5.03it/s] 39%|███▉ | 18805/48008 [1:04:21<1:37:04, 5.01it/s] 39%|███▉ | 18806/48008 [1:04:22<1:37:47, 4.98it/s] 39%|███▉ | 18807/48008 [1:04:22<1:37:04, 5.01it/s] 39%|███▉ | 18808/48008 [1:04:22<1:37:34, 4.99it/s] 39%|███▉ | 18809/48008 [1:04:22<1:37:04, 5.01it/s] 39%|███▉ | 18810/48008 [1:04:22<1:35:43, 5.08it/s] 39%|███▉ | 18811/48008 [1:04:23<1:37:10, 5.01it/s] 39%|███▉ | 18812/48008 [1:04:23<1:36:40, 5.03it/s] 39%|███▉ | 18813/48008 [1:04:23<1:35:27, 5.10it/s] 39%|███▉ | 18814/48008 [1:04:23<1:34:26, 5.15it/s] 39%|███▉ | 18815/48008 [1:04:23<1:34:29, 5.15it/s] 39%|███▉ | 18816/48008 [1:04:24<1:33:52, 5.18it/s] 39%|███▉ | 18817/48008 [1:04:24<1:33:57, 5.18it/s] 39%|███▉ | 18818/48008 [1:04:24<1:33:45, 5.19it/s] 39%|███▉ | 18819/48008 [1:04:24<1:33:01, 5.23it/s] 39%|███▉ | 18820/48008 [1:04:24<1:35:53, 5.07it/s] 39%|███▉ | 18821/48008 [1:04:25<1:35:11, 5.11it/s] 39%|███▉ | 18822/48008 [1:04:25<1:35:42, 5.08it/s] 39%|███▉ | 18823/48008 [1:04:25<1:35:02, 5.12it/s] 39%|███▉ | 18824/48008 [1:04:25<1:41:03, 4.81it/s] 39%|███▉ | 18825/48008 [1:04:25<1:37:52, 4.97it/s] 39%|███▉ | 18826/48008 [1:04:26<1:36:36, 5.03it/s] 39%|███▉ | 18827/48008 [1:04:26<1:34:35, 5.14it/s] 39%|███▉ | 18828/48008 [1:04:26<1:36:11, 5.06it/s] 39%|███▉ | 18829/48008 [1:04:26<1:35:28, 5.09it/s] 39%|███▉ | 18830/48008 [1:04:26<1:35:47, 5.08it/s] 39%|███▉ | 18831/48008 [1:04:27<1:35:03, 5.12it/s] 39%|███▉ | 18832/48008 [1:04:27<1:35:44, 5.08it/s] 39%|███▉ | 18833/48008 [1:04:27<1:51:53, 4.35it/s] 39%|███▉ | 18834/48008 [1:04:27<1:46:16, 4.58it/s] 39%|███▉ | 18835/48008 [1:04:27<1:42:26, 4.75it/s] 39%|███▉ | 18836/48008 [1:04:28<1:39:14, 4.90it/s] 39%|███▉ | 18837/48008 [1:04:28<1:37:03, 5.01it/s] 39%|███▉ | 18838/48008 [1:04:28<1:35:32, 5.09it/s] 39%|███▉ | 18839/48008 [1:04:28<1:34:45, 5.13it/s] 39%|███▉ | 18840/48008 [1:04:29<3:19:07, 2.44it/s] 39%|███▉ | 18841/48008 [1:04:29<2:48:59, 2.88it/s] 39%|███▉ | 18842/48008 [1:04:30<2:26:27, 3.32it/s] 39%|███▉ | 18843/48008 [1:04:30<2:11:37, 3.69it/s] 39%|███▉ | 18844/48008 [1:04:30<2:01:34, 4.00it/s] 39%|███▉ | 18845/48008 [1:04:30<1:52:11, 4.33it/s] 39%|███▉ | 18846/48008 [1:04:30<1:46:55, 4.55it/s] 39%|███▉ | 18847/48008 [1:04:30<1:42:30, 4.74it/s] 39%|███▉ | 18848/48008 [1:04:31<1:39:53, 4.87it/s] 39%|███▉ | 18849/48008 [1:04:31<1:38:12, 4.95it/s] 39%|███▉ | 18850/48008 [1:04:31<1:38:15, 4.95it/s] {'loss': 0.3031, 'grad_norm': 0.345703125, 'learning_rate': 0.00012147558740209966, 'epoch': 0.39} 39%|███▉ | 18850/48008 [1:04:31<1:38:15, 4.95it/s] 39%|███▉ | 18851/48008 [1:04:31<1:37:21, 4.99it/s] 39%|███▉ | 18852/48008 [1:04:31<1:39:06, 4.90it/s] 39%|███▉ | 18853/48008 [1:04:32<1:38:36, 4.93it/s] 39%|███▉ | 18854/48008 [1:04:32<1:37:47, 4.97it/s] 39%|███▉ | 18855/48008 [1:04:32<2:05:04, 3.88it/s] 39%|███▉ | 18856/48008 [1:04:32<1:55:29, 4.21it/s] 39%|███▉ | 18857/48008 [1:04:33<1:48:30, 4.48it/s] 39%|███▉ | 18858/48008 [1:04:33<1:43:55, 4.67it/s] 39%|███▉ | 18859/48008 [1:04:33<1:39:40, 4.87it/s] 39%|███▉ | 18860/48008 [1:04:33<1:37:56, 4.96it/s] 39%|███▉ | 18861/48008 [1:04:33<1:36:36, 5.03it/s] 39%|███▉ | 18862/48008 [1:04:34<1:36:32, 5.03it/s] 39%|███▉ | 18863/48008 [1:04:34<1:36:38, 5.03it/s] 39%|███▉ | 18864/48008 [1:04:34<1:35:37, 5.08it/s] 39%|███▉ | 18865/48008 [1:04:34<1:36:25, 5.04it/s] 39%|███▉ | 18866/48008 [1:04:34<1:35:14, 5.10it/s] 39%|███▉ | 18867/48008 [1:04:35<1:35:21, 5.09it/s] 39%|███▉ | 18868/48008 [1:04:35<1:34:04, 5.16it/s] 39%|███▉ | 18869/48008 [1:04:35<1:32:33, 5.25it/s] 39%|███▉ | 18870/48008 [1:04:35<1:32:16, 5.26it/s] 39%|███▉ | 18871/48008 [1:04:35<1:32:08, 5.27it/s] 39%|███▉ | 18872/48008 [1:04:36<1:33:24, 5.20it/s] 39%|███▉ | 18873/48008 [1:04:36<1:33:54, 5.17it/s] 39%|███▉ | 18874/48008 [1:04:36<1:34:12, 5.15it/s] 39%|███▉ | 18875/48008 [1:04:36<1:33:44, 5.18it/s] 39%|███▉ | 18876/48008 [1:04:36<1:33:13, 5.21it/s] 39%|███▉ | 18877/48008 [1:04:37<1:33:42, 5.18it/s] 39%|███▉ | 18878/48008 [1:04:37<1:33:26, 5.20it/s] 39%|███▉ | 18879/48008 [1:04:37<1:33:27, 5.19it/s] 39%|███▉ | 18880/48008 [1:04:37<1:33:33, 5.19it/s] 39%|███▉ | 18881/48008 [1:04:37<1:33:07, 5.21it/s] 39%|███▉ | 18882/48008 [1:04:37<1:33:00, 5.22it/s] 39%|███▉ | 18883/48008 [1:04:38<1:32:37, 5.24it/s] 39%|███▉ | 18884/48008 [1:04:38<1:33:34, 5.19it/s] 39%|███▉ | 18885/48008 [1:04:38<1:33:03, 5.22it/s] 39%|███▉ | 18886/48008 [1:04:38<1:39:44, 4.87it/s] 39%|███▉ | 18887/48008 [1:04:38<1:37:01, 5.00it/s] 39%|███▉ | 18888/48008 [1:04:39<1:38:48, 4.91it/s] 39%|███▉ | 18889/48008 [1:04:39<1:38:28, 4.93it/s] 39%|███▉ | 18890/48008 [1:04:39<1:39:10, 4.89it/s] 39%|███▉ | 18891/48008 [1:04:39<1:38:21, 4.93it/s] 39%|███▉ | 18892/48008 [1:04:39<1:36:02, 5.05it/s] 39%|███▉ | 18893/48008 [1:04:40<1:34:52, 5.11it/s] 39%|███▉ | 18894/48008 [1:04:40<1:34:08, 5.15it/s] 39%|███▉ | 18895/48008 [1:04:40<1:32:53, 5.22it/s] 39%|███▉ | 18896/48008 [1:04:40<1:32:32, 5.24it/s] 39%|███▉ | 18897/48008 [1:04:40<1:32:32, 5.24it/s] 39%|███▉ | 18898/48008 [1:04:41<1:32:53, 5.22it/s] 39%|███▉ | 18899/48008 [1:04:41<1:34:21, 5.14it/s] 39%|███▉ | 18900/48008 [1:04:41<1:35:27, 5.08it/s] {'loss': 0.3023, 'grad_norm': 0.1611328125, 'learning_rate': 0.00012126728878520249, 'epoch': 0.39} 39%|███▉ | 18900/48008 [1:04:41<1:35:27, 5.08it/s] 39%|███▉ | 18901/48008 [1:04:41<2:03:02, 3.94it/s] 39%|███▉ | 18902/48008 [1:04:42<1:53:49, 4.26it/s] 39%|███▉ | 18903/48008 [1:04:42<1:47:31, 4.51it/s] 39%|███▉ | 18904/48008 [1:04:42<1:42:00, 4.76it/s] 39%|███▉ | 18905/48008 [1:04:42<1:39:10, 4.89it/s] 39%|███▉ | 18906/48008 [1:04:42<1:38:54, 4.90it/s] 39%|███▉ | 18907/48008 [1:04:43<1:38:01, 4.95it/s] 39%|███▉ | 18908/48008 [1:04:43<1:36:12, 5.04it/s] 39%|███▉ | 18909/48008 [1:04:43<1:36:39, 5.02it/s] 39%|███▉ | 18910/48008 [1:04:43<1:35:22, 5.08it/s] 39%|███▉ | 18911/48008 [1:04:43<1:35:31, 5.08it/s] 39%|███▉ | 18912/48008 [1:04:44<1:34:22, 5.14it/s] 39%|███▉ | 18913/48008 [1:04:44<1:36:33, 5.02it/s] 39%|███▉ | 18914/48008 [1:04:44<1:37:23, 4.98it/s] 39%|███▉ | 18915/48008 [1:04:44<1:36:39, 5.02it/s] 39%|███▉ | 18916/48008 [1:04:44<1:35:33, 5.07it/s] 39%|███▉ | 18917/48008 [1:04:45<1:34:36, 5.12it/s] 39%|███▉ | 18918/48008 [1:04:45<1:33:01, 5.21it/s] 39%|███▉ | 18919/48008 [1:04:45<1:34:40, 5.12it/s] 39%|███▉ | 18920/48008 [1:04:45<1:33:53, 5.16it/s] 39%|███▉ | 18921/48008 [1:04:45<1:35:25, 5.08it/s] 39%|███▉ | 18922/48008 [1:04:46<1:35:13, 5.09it/s] 39%|███▉ | 18923/48008 [1:04:46<1:36:16, 5.03it/s] 39%|███▉ | 18924/48008 [1:04:46<1:35:12, 5.09it/s] 39%|███▉ | 18925/48008 [1:04:46<2:03:13, 3.93it/s] 39%|███▉ | 18926/48008 [1:04:46<1:53:49, 4.26it/s] 39%|███▉ | 18927/48008 [1:04:47<1:47:02, 4.53it/s] 39%|███▉ | 18928/48008 [1:04:47<1:41:54, 4.76it/s] 39%|███▉ | 18929/48008 [1:04:47<1:46:01, 4.57it/s] 39%|███▉ | 18930/48008 [1:04:47<1:43:35, 4.68it/s] 39%|███▉ | 18931/48008 [1:04:47<1:40:15, 4.83it/s] 39%|███▉ | 18932/48008 [1:04:48<1:37:53, 4.95it/s] 39%|███▉ | 18933/48008 [1:04:48<1:35:52, 5.05it/s] 39%|███▉ | 18934/48008 [1:04:48<1:34:49, 5.11it/s] 39%|███▉ | 18935/48008 [1:04:48<1:33:59, 5.16it/s] 39%|███▉ | 18936/48008 [1:04:48<1:33:12, 5.20it/s] 39%|███▉ | 18937/48008 [1:04:49<1:34:39, 5.12it/s] 39%|███▉ | 18938/48008 [1:04:49<1:34:18, 5.14it/s] 39%|███▉ | 18939/48008 [1:04:49<1:32:53, 5.22it/s] 39%|███▉ | 18940/48008 [1:04:49<1:33:00, 5.21it/s] 39%|███▉ | 18941/48008 [1:04:49<1:32:50, 5.22it/s] 39%|███▉ | 18942/48008 [1:04:50<1:32:47, 5.22it/s] 39%|███▉ | 18943/48008 [1:04:50<1:32:02, 5.26it/s] 39%|███▉ | 18944/48008 [1:04:50<1:32:15, 5.25it/s] 39%|███▉ | 18945/48008 [1:04:50<1:32:31, 5.24it/s] 39%|███▉ | 18946/48008 [1:04:50<1:39:18, 4.88it/s] 39%|███▉ | 18947/48008 [1:04:51<1:37:07, 4.99it/s] 39%|███▉ | 18948/48008 [1:04:51<1:37:38, 4.96it/s] 39%|███▉ | 18949/48008 [1:04:51<1:35:48, 5.06it/s] 39%|███▉ | 18950/48008 [1:04:51<1:35:34, 5.07it/s] {'loss': 0.3013, 'grad_norm': 0.314453125, 'learning_rate': 0.00012105899016830529, 'epoch': 0.39} 39%|███▉ | 18950/48008 [1:04:51<1:35:34, 5.07it/s] 39%|███▉ | 18951/48008 [1:04:51<1:35:41, 5.06it/s] 39%|███▉ | 18952/48008 [1:04:52<1:34:12, 5.14it/s] 39%|███▉ | 18953/48008 [1:04:52<1:33:37, 5.17it/s] 39%|███▉ | 18954/48008 [1:04:52<1:34:44, 5.11it/s] 39%|███▉ | 18955/48008 [1:04:52<1:34:53, 5.10it/s] 39%|███▉ | 18956/48008 [1:04:52<1:33:26, 5.18it/s] 39%|███▉ | 18957/48008 [1:04:53<1:36:08, 5.04it/s] 39%|███▉ | 18958/48008 [1:04:53<1:34:03, 5.15it/s] 39%|███▉ | 18959/48008 [1:04:53<1:33:07, 5.20it/s] 39%|███▉ | 18960/48008 [1:04:53<1:32:26, 5.24it/s] 39%|███▉ | 18961/48008 [1:04:53<1:39:21, 4.87it/s] 39%|███▉ | 18962/48008 [1:04:54<1:37:52, 4.95it/s] 39%|███▉ | 18963/48008 [1:04:54<1:52:55, 4.29it/s] 40%|███▉ | 18964/48008 [1:04:54<1:48:41, 4.45it/s] 40%|███▉ | 18965/48008 [1:04:54<1:42:55, 4.70it/s] 40%|███▉ | 18966/48008 [1:04:54<1:39:14, 4.88it/s] 40%|███▉ | 18967/48008 [1:04:55<1:39:45, 4.85it/s] 40%|███▉ | 18968/48008 [1:04:55<1:36:57, 4.99it/s] 40%|███▉ | 18969/48008 [1:04:56<3:29:43, 2.31it/s] 40%|███▉ | 18970/48008 [1:04:56<2:53:33, 2.79it/s] 40%|███▉ | 18971/48008 [1:04:56<2:28:52, 3.25it/s] 40%|███▉ | 18972/48008 [1:04:56<2:11:53, 3.67it/s] 40%|███▉ | 18973/48008 [1:04:57<1:59:23, 4.05it/s] 40%|███▉ | 18974/48008 [1:04:57<1:51:26, 4.34it/s] 40%|███▉ | 18975/48008 [1:04:58<3:16:13, 2.47it/s] 40%|███▉ | 18976/48008 [1:04:58<2:45:15, 2.93it/s] 40%|███▉ | 18977/48008 [1:04:58<2:23:09, 3.38it/s] 40%|███▉ | 18978/48008 [1:04:58<2:09:11, 3.75it/s] 40%|███▉ | 18979/48008 [1:04:58<1:57:51, 4.11it/s] 40%|███▉ | 18980/48008 [1:04:59<1:49:37, 4.41it/s] 40%|███▉ | 18981/48008 [1:04:59<3:12:31, 2.51it/s] 40%|███▉ | 18982/48008 [1:05:00<2:43:25, 2.96it/s] 40%|███▉ | 18983/48008 [1:05:00<2:22:33, 3.39it/s] 40%|███▉ | 18984/48008 [1:05:00<2:14:07, 3.61it/s] 40%|███▉ | 18985/48008 [1:05:00<2:01:54, 3.97it/s] 40%|███▉ | 18986/48008 [1:05:00<1:52:55, 4.28it/s] 40%|███▉ | 18987/48008 [1:05:01<1:46:36, 4.54it/s] 40%|███▉ | 18988/48008 [1:05:01<1:42:36, 4.71it/s] 40%|███▉ | 18989/48008 [1:05:01<1:41:27, 4.77it/s] 40%|███▉ | 18990/48008 [1:05:01<1:39:23, 4.87it/s] 40%|███▉ | 18991/48008 [1:05:01<1:36:37, 5.01it/s] 40%|███▉ | 18992/48008 [1:05:01<1:35:34, 5.06it/s] 40%|███▉ | 18993/48008 [1:05:02<1:35:21, 5.07it/s] 40%|███▉ | 18994/48008 [1:05:02<1:34:13, 5.13it/s] 40%|███▉ | 18995/48008 [1:05:02<1:34:03, 5.14it/s] 40%|███▉ | 18996/48008 [1:05:02<1:32:35, 5.22it/s] 40%|███▉ | 18997/48008 [1:05:02<1:32:42, 5.22it/s] 40%|███▉ | 18998/48008 [1:05:03<1:32:41, 5.22it/s] 40%|███▉ | 18999/48008 [1:05:03<1:33:46, 5.16it/s] 40%|███▉ | 19000/48008 [1:05:03<1:35:09, 5.08it/s] {'loss': 0.3021, 'grad_norm': 0.62890625, 'learning_rate': 0.00012085069155140811, 'epoch': 0.4} 40%|███▉ | 19000/48008 [1:05:03<1:35:09, 5.08it/s] 40%|███▉ | 19001/48008 [1:05:03<1:35:19, 5.07it/s] 40%|███▉ | 19002/48008 [1:05:03<1:34:00, 5.14it/s] 40%|███▉ | 19003/48008 [1:05:04<1:50:26, 4.38it/s] 40%|███▉ | 19004/48008 [1:05:04<1:46:50, 4.52it/s] 40%|███▉ | 19005/48008 [1:05:04<1:44:15, 4.64it/s] 40%|███▉ | 19006/48008 [1:05:04<1:42:30, 4.72it/s] 40%|███▉ | 19007/48008 [1:05:05<1:39:24, 4.86it/s] 40%|███▉ | 19008/48008 [1:05:05<1:43:59, 4.65it/s] 40%|███▉ | 19009/48008 [1:05:05<1:39:43, 4.85it/s] 40%|███▉ | 19010/48008 [1:05:05<1:37:34, 4.95it/s] 40%|███▉ | 19011/48008 [1:05:06<3:04:14, 2.62it/s] 40%|███▉ | 19012/48008 [1:05:06<2:36:40, 3.08it/s] 40%|███▉ | 19013/48008 [1:05:06<2:19:14, 3.47it/s] 40%|███▉ | 19014/48008 [1:05:07<2:06:13, 3.83it/s] 40%|███▉ | 19015/48008 [1:05:07<1:56:38, 4.14it/s] 40%|███▉ | 19016/48008 [1:05:07<1:49:15, 4.42it/s] 40%|███▉ | 19017/48008 [1:05:07<2:00:49, 4.00it/s] 40%|███▉ | 19018/48008 [1:05:07<1:51:17, 4.34it/s] 40%|███▉ | 19019/48008 [1:05:08<1:45:45, 4.57it/s] 40%|███▉ | 19020/48008 [1:05:08<1:41:45, 4.75it/s] 40%|███▉ | 19021/48008 [1:05:08<1:38:53, 4.89it/s] 40%|███▉ | 19022/48008 [1:05:08<1:36:16, 5.02it/s] 40%|███▉ | 19023/48008 [1:05:08<1:36:26, 5.01it/s] 40%|███▉ | 19024/48008 [1:05:09<1:41:42, 4.75it/s] 40%|███▉ | 19025/48008 [1:05:09<1:39:34, 4.85it/s] 40%|███▉ | 19026/48008 [1:05:09<1:37:28, 4.96it/s] 40%|███▉ | 19027/48008 [1:05:09<1:35:44, 5.04it/s] 40%|███▉ | 19028/48008 [1:05:09<1:34:12, 5.13it/s] 40%|███▉ | 19029/48008 [1:05:10<1:34:22, 5.12it/s] 40%|███▉ | 19030/48008 [1:05:10<1:40:19, 4.81it/s] 40%|███▉ | 19031/48008 [1:05:10<1:38:06, 4.92it/s] 40%|███▉ | 19032/48008 [1:05:10<1:36:04, 5.03it/s] 40%|███▉ | 19033/48008 [1:05:10<1:36:46, 4.99it/s] 40%|███▉ | 19034/48008 [1:05:11<1:42:15, 4.72it/s] 40%|███▉ | 19035/48008 [1:05:11<1:39:15, 4.87it/s] 40%|███▉ | 19036/48008 [1:05:11<1:37:53, 4.93it/s] 40%|███▉ | 19037/48008 [1:05:11<1:35:42, 5.05it/s] 40%|███▉ | 19038/48008 [1:05:11<1:34:29, 5.11it/s] 40%|███▉ | 19039/48008 [1:05:12<1:33:07, 5.18it/s] 40%|███▉ | 19040/48008 [1:05:12<1:31:37, 5.27it/s] 40%|███▉ | 19041/48008 [1:05:12<1:31:23, 5.28it/s] 40%|███▉ | 19042/48008 [1:05:12<1:31:19, 5.29it/s] 40%|███▉ | 19043/48008 [1:05:12<1:30:56, 5.31it/s] 40%|███▉ | 19044/48008 [1:05:13<1:31:23, 5.28it/s] 40%|███▉ | 19045/48008 [1:05:13<1:34:37, 5.10it/s] 40%|███▉ | 19046/48008 [1:05:13<1:34:39, 5.10it/s] 40%|███▉ | 19047/48008 [1:05:13<1:40:39, 4.80it/s] 40%|███▉ | 19048/48008 [1:05:13<1:39:48, 4.84it/s] 40%|███▉ | 19049/48008 [1:05:14<1:37:29, 4.95it/s] 40%|███▉ | 19050/48008 [1:05:14<1:37:38, 4.94it/s] {'loss': 0.303, 'grad_norm': 0.6953125, 'learning_rate': 0.00012064239293451091, 'epoch': 0.4} 40%|███▉ | 19050/48008 [1:05:14<1:37:38, 4.94it/s] 40%|███▉ | 19051/48008 [1:05:14<1:36:27, 5.00it/s] 40%|███▉ | 19052/48008 [1:05:14<1:35:19, 5.06it/s] 40%|███▉ | 19053/48008 [1:05:14<1:34:21, 5.11it/s] 40%|███▉ | 19054/48008 [1:05:15<1:33:45, 5.15it/s] 40%|███▉ | 19055/48008 [1:05:15<1:32:42, 5.20it/s] 40%|███▉ | 19056/48008 [1:05:15<1:32:34, 5.21it/s] 40%|███▉ | 19057/48008 [1:05:15<1:48:55, 4.43it/s] 40%|███▉ | 19058/48008 [1:05:15<1:43:59, 4.64it/s] 40%|███▉ | 19059/48008 [1:05:16<2:08:39, 3.75it/s] 40%|███▉ | 19060/48008 [1:05:16<1:57:33, 4.10it/s] 40%|███▉ | 19061/48008 [1:05:16<1:49:46, 4.39it/s] 40%|███▉ | 19062/48008 [1:05:16<1:43:43, 4.65it/s] 40%|███▉ | 19063/48008 [1:05:17<1:40:14, 4.81it/s] 40%|███▉ | 19064/48008 [1:05:17<1:39:22, 4.85it/s] 40%|███▉ | 19065/48008 [1:05:17<1:37:11, 4.96it/s] 40%|███▉ | 19066/48008 [1:05:17<1:36:51, 4.98it/s] 40%|███▉ | 19067/48008 [1:05:17<1:35:24, 5.06it/s] 40%|███▉ | 19068/48008 [1:05:18<1:34:17, 5.12it/s] 40%|███▉ | 19069/48008 [1:05:18<1:34:18, 5.11it/s] 40%|███▉ | 19070/48008 [1:05:18<1:33:50, 5.14it/s] 40%|███▉ | 19071/48008 [1:05:18<1:35:09, 5.07it/s] 40%|███▉ | 19072/48008 [1:05:18<1:33:54, 5.14it/s] 40%|███▉ | 19073/48008 [1:05:18<1:34:36, 5.10it/s] 40%|███▉ | 19074/48008 [1:05:19<1:33:18, 5.17it/s] 40%|███▉ | 19075/48008 [1:05:19<1:32:38, 5.21it/s] 40%|███▉ | 19076/48008 [1:05:19<1:31:42, 5.26it/s] 40%|███▉ | 19077/48008 [1:05:19<1:31:57, 5.24it/s] 40%|███▉ | 19078/48008 [1:05:19<1:34:36, 5.10it/s] 40%|███▉ | 19079/48008 [1:05:20<1:33:14, 5.17it/s] 40%|███▉ | 19080/48008 [1:05:20<1:34:32, 5.10it/s] 40%|███▉ | 19081/48008 [1:05:20<1:35:44, 5.04it/s] 40%|███▉ | 19082/48008 [1:05:20<1:34:47, 5.09it/s] 40%|███▉ | 19083/48008 [1:05:20<1:36:23, 5.00it/s] 40%|███▉ | 19084/48008 [1:05:21<1:34:53, 5.08it/s] 40%|███▉ | 19085/48008 [1:05:21<1:34:16, 5.11it/s] 40%|███▉ | 19086/48008 [1:05:21<1:40:20, 4.80it/s] 40%|███▉ | 19087/48008 [1:05:21<1:37:39, 4.94it/s] 40%|███▉ | 19088/48008 [1:05:21<1:36:43, 4.98it/s] 40%|███▉ | 19089/48008 [1:05:22<1:34:19, 5.11it/s] 40%|███▉ | 19090/48008 [1:05:22<1:35:33, 5.04it/s] 40%|███▉ | 19091/48008 [1:05:22<1:34:25, 5.10it/s] 40%|███▉ | 19092/48008 [1:05:22<1:33:38, 5.15it/s] 40%|███▉ | 19093/48008 [1:05:22<1:32:15, 5.22it/s] 40%|███▉ | 19094/48008 [1:05:23<1:34:39, 5.09it/s] 40%|███▉ | 19095/48008 [1:05:23<1:34:01, 5.12it/s] 40%|███▉ | 19096/48008 [1:05:23<1:50:23, 4.36it/s] 40%|███▉ | 19097/48008 [1:05:23<1:45:15, 4.58it/s] 40%|███▉ | 19098/48008 [1:05:24<1:41:28, 4.75it/s] 40%|███▉ | 19099/48008 [1:05:24<1:39:36, 4.84it/s] 40%|███▉ | 19100/48008 [1:05:24<1:37:20, 4.95it/s] {'loss': 0.3025, 'grad_norm': 0.2041015625, 'learning_rate': 0.00012043409431761374, 'epoch': 0.4} 40%|███▉ | 19100/48008 [1:05:24<1:37:20, 4.95it/s] 40%|███▉ | 19101/48008 [1:05:24<1:35:41, 5.03it/s] 40%|███▉ | 19102/48008 [1:05:24<1:34:11, 5.11it/s] 40%|███▉ | 19103/48008 [1:05:24<1:33:20, 5.16it/s] 40%|███▉ | 19104/48008 [1:05:25<1:33:03, 5.18it/s] 40%|███▉ | 19105/48008 [1:05:25<1:32:24, 5.21it/s] 40%|███▉ | 19106/48008 [1:05:25<1:32:44, 5.19it/s] 40%|███▉ | 19107/48008 [1:05:25<1:33:35, 5.15it/s] 40%|███▉ | 19108/48008 [1:05:25<1:32:59, 5.18it/s] 40%|███▉ | 19109/48008 [1:05:26<1:32:44, 5.19it/s] 40%|███▉ | 19110/48008 [1:05:26<1:32:20, 5.22it/s] 40%|███▉ | 19111/48008 [1:05:26<1:31:02, 5.29it/s] 40%|███▉ | 19112/48008 [1:05:27<3:53:48, 2.06it/s] 40%|███▉ | 19113/48008 [1:05:27<3:11:04, 2.52it/s] 40%|███▉ | 19114/48008 [1:05:28<2:41:33, 2.98it/s] 40%|███▉ | 19115/48008 [1:05:28<2:20:34, 3.43it/s] 40%|███▉ | 19116/48008 [1:05:28<2:05:03, 3.85it/s] 40%|███▉ | 19117/48008 [1:05:28<1:55:05, 4.18it/s] 40%|███▉ | 19118/48008 [1:05:28<1:47:58, 4.46it/s] 40%|███▉ | 19119/48008 [1:05:29<1:43:47, 4.64it/s] 40%|███▉ | 19120/48008 [1:05:29<1:40:14, 4.80it/s] 40%|███▉ | 19121/48008 [1:05:29<1:44:32, 4.61it/s] 40%|███▉ | 19122/48008 [1:05:29<1:43:27, 4.65it/s] 40%|███▉ | 19123/48008 [1:05:29<1:39:51, 4.82it/s] 40%|███▉ | 19124/48008 [1:05:30<1:36:29, 4.99it/s] 40%|███▉ | 19125/48008 [1:05:30<1:35:40, 5.03it/s] 40%|███▉ | 19126/48008 [1:05:30<1:36:15, 5.00it/s] 40%|███▉ | 19127/48008 [1:05:30<1:35:19, 5.05it/s] 40%|███▉ | 19128/48008 [1:05:30<1:33:45, 5.13it/s] 40%|███▉ | 19129/48008 [1:05:30<1:33:20, 5.16it/s] 40%|███▉ | 19130/48008 [1:05:31<1:32:57, 5.18it/s] 40%|███▉ | 19131/48008 [1:05:31<1:32:02, 5.23it/s] 40%|███▉ | 19132/48008 [1:05:31<1:32:18, 5.21it/s] 40%|███▉ | 19133/48008 [1:05:31<1:31:53, 5.24it/s] 40%|███▉ | 19134/48008 [1:05:31<1:31:58, 5.23it/s] 40%|███▉ | 19135/48008 [1:05:32<1:31:55, 5.23it/s] 40%|███▉ | 19136/48008 [1:05:32<1:31:52, 5.24it/s] 40%|███▉ | 19137/48008 [1:05:32<1:31:48, 5.24it/s] 40%|███▉ | 19138/48008 [1:05:32<1:31:37, 5.25it/s] 40%|███▉ | 19139/48008 [1:05:32<1:31:31, 5.26it/s] 40%|███▉ | 19140/48008 [1:05:33<1:31:35, 5.25it/s] 40%|███▉ | 19141/48008 [1:05:33<1:33:37, 5.14it/s] 40%|███▉ | 19142/48008 [1:05:33<1:35:12, 5.05it/s] 40%|███▉ | 19143/48008 [1:05:33<1:33:43, 5.13it/s] 40%|███▉ | 19144/48008 [1:05:33<1:33:52, 5.12it/s] 40%|███▉ | 19145/48008 [1:05:34<1:33:31, 5.14it/s] 40%|███▉ | 19146/48008 [1:05:34<1:34:05, 5.11it/s] 40%|███▉ | 19147/48008 [1:05:34<1:32:01, 5.23it/s] 40%|███▉ | 19148/48008 [1:05:34<1:33:03, 5.17it/s] 40%|███▉ | 19149/48008 [1:05:34<1:34:16, 5.10it/s] 40%|███▉ | 19150/48008 [1:05:35<1:33:22, 5.15it/s] {'loss': 0.3012, 'grad_norm': 0.1591796875, 'learning_rate': 0.00012022579570071655, 'epoch': 0.4} 40%|███▉ | 19150/48008 [1:05:35<1:33:22, 5.15it/s] 40%|███▉ | 19151/48008 [1:05:35<1:49:41, 4.38it/s] 40%|███▉ | 19152/48008 [1:05:35<1:44:18, 4.61it/s] 40%|███▉ | 19153/48008 [1:05:35<1:42:15, 4.70it/s] 40%|███▉ | 19154/48008 [1:05:35<1:38:44, 4.87it/s] 40%|███▉ | 19155/48008 [1:05:36<1:36:24, 4.99it/s] 40%|███▉ | 19156/48008 [1:05:36<1:34:38, 5.08it/s] 40%|███▉ | 19157/48008 [1:05:36<1:33:13, 5.16it/s] 40%|███▉ | 19158/48008 [1:05:36<1:34:37, 5.08it/s] 40%|███▉ | 19159/48008 [1:05:36<1:34:32, 5.09it/s] 40%|███▉ | 19160/48008 [1:05:37<1:33:07, 5.16it/s] 40%|███▉ | 19161/48008 [1:05:37<1:33:28, 5.14it/s] 40%|███▉ | 19162/48008 [1:05:37<1:33:10, 5.16it/s] 40%|███▉ | 19163/48008 [1:05:37<1:34:35, 5.08it/s] 40%|███▉ | 19164/48008 [1:05:37<1:33:36, 5.14it/s] 40%|███▉ | 19165/48008 [1:05:38<1:32:27, 5.20it/s] 40%|███▉ | 19166/48008 [1:05:38<1:32:13, 5.21it/s] 40%|███▉ | 19167/48008 [1:05:38<1:31:52, 5.23it/s] 40%|███▉ | 19168/48008 [1:05:38<1:31:11, 5.27it/s] 40%|███▉ | 19169/48008 [1:05:38<1:31:08, 5.27it/s] 40%|███▉ | 19170/48008 [1:05:38<1:30:34, 5.31it/s] 40%|███▉ | 19171/48008 [1:05:39<1:31:00, 5.28it/s] 40%|███▉ | 19172/48008 [1:05:39<1:30:24, 5.32it/s] 40%|███▉ | 19173/48008 [1:05:39<1:31:41, 5.24it/s] 40%|███▉ | 19174/48008 [1:05:39<1:30:54, 5.29it/s] 40%|███▉ | 19175/48008 [1:05:39<1:31:24, 5.26it/s] 40%|███▉ | 19176/48008 [1:05:40<1:32:59, 5.17it/s] 40%|███▉ | 19177/48008 [1:05:40<1:31:59, 5.22it/s] 40%|███▉ | 19178/48008 [1:05:40<1:31:22, 5.26it/s] 40%|███▉ | 19179/48008 [1:05:40<1:30:48, 5.29it/s] 40%|███▉ | 19180/48008 [1:05:40<1:29:41, 5.36it/s] 40%|███▉ | 19181/48008 [1:05:41<1:37:04, 4.95it/s] 40%|███▉ | 19182/48008 [1:05:41<1:35:17, 5.04it/s] 40%|███▉ | 19183/48008 [1:05:41<1:35:53, 5.01it/s] 40%|███▉ | 19184/48008 [1:05:41<1:34:45, 5.07it/s] 40%|███▉ | 19185/48008 [1:05:41<1:33:25, 5.14it/s] 40%|███▉ | 19186/48008 [1:05:42<1:33:25, 5.14it/s] 40%|███▉ | 19187/48008 [1:05:42<1:33:06, 5.16it/s] 40%|███▉ | 19188/48008 [1:05:42<1:32:33, 5.19it/s] 40%|███▉ | 19189/48008 [1:05:42<1:31:51, 5.23it/s] 40%|███▉ | 19190/48008 [1:05:42<1:31:54, 5.23it/s] 40%|███▉ | 19191/48008 [1:05:43<1:30:52, 5.29it/s] 40%|███▉ | 19192/48008 [1:05:43<1:31:57, 5.22it/s] 40%|███▉ | 19193/48008 [1:05:43<1:32:05, 5.22it/s] 40%|███▉ | 19194/48008 [1:05:43<1:32:14, 5.21it/s] 40%|███▉ | 19195/48008 [1:05:43<1:32:00, 5.22it/s] 40%|███▉ | 19196/48008 [1:05:44<1:59:59, 4.00it/s] 40%|███▉ | 19197/48008 [1:05:44<1:53:17, 4.24it/s] 40%|███▉ | 19198/48008 [1:05:44<1:48:26, 4.43it/s] 40%|███▉ | 19199/48008 [1:05:44<1:43:10, 4.65it/s] 40%|███▉ | 19200/48008 [1:05:44<1:40:45, 4.77it/s] {'loss': 0.3025, 'grad_norm': 0.1826171875, 'learning_rate': 0.00012001749708381938, 'epoch': 0.4} 40%|███▉ | 19200/48008 [1:05:44<1:40:45, 4.77it/s] 40%|███▉ | 19201/48008 [1:05:45<1:37:53, 4.90it/s] 40%|███▉ | 19202/48008 [1:05:45<1:37:53, 4.90it/s] 40%|███▉ | 19203/48008 [1:05:45<1:35:46, 5.01it/s] 40%|████ | 19204/48008 [1:05:45<1:34:32, 5.08it/s] 40%|████ | 19205/48008 [1:05:46<1:50:12, 4.36it/s] 40%|████ | 19206/48008 [1:05:46<1:46:24, 4.51it/s] 40%|████ | 19207/48008 [1:05:46<1:41:24, 4.73it/s] 40%|████ | 19208/48008 [1:05:46<1:39:20, 4.83it/s] 40%|████ | 19209/48008 [1:05:46<1:36:17, 4.99it/s] 40%|████ | 19210/48008 [1:05:47<1:34:42, 5.07it/s] 40%|████ | 19211/48008 [1:05:47<1:33:52, 5.11it/s] 40%|████ | 19212/48008 [1:05:47<1:32:44, 5.18it/s] 40%|████ | 19213/48008 [1:05:47<1:32:11, 5.21it/s] 40%|████ | 19214/48008 [1:05:47<2:00:02, 4.00it/s] 40%|████ | 19215/48008 [1:05:48<1:52:55, 4.25it/s] 40%|████ | 19216/48008 [1:05:48<1:45:42, 4.54it/s] 40%|████ | 19217/48008 [1:05:48<2:09:38, 3.70it/s] 40%|████ | 19218/48008 [1:05:48<1:58:05, 4.06it/s] 40%|████ | 19219/48008 [1:05:49<1:49:37, 4.38it/s] 40%|████ | 19220/48008 [1:05:49<1:45:05, 4.57it/s] 40%|████ | 19221/48008 [1:05:49<1:41:14, 4.74it/s] 40%|████ | 19222/48008 [1:05:49<1:38:08, 4.89it/s] 40%|████ | 19223/48008 [1:05:49<1:35:37, 5.02it/s] 40%|████ | 19224/48008 [1:05:50<1:33:41, 5.12it/s] 40%|████ | 19225/48008 [1:05:50<1:32:56, 5.16it/s] 40%|████ | 19226/48008 [1:05:50<2:00:33, 3.98it/s] 40%|████ | 19227/48008 [1:05:50<1:50:53, 4.33it/s] 40%|████ | 19228/48008 [1:05:51<1:46:06, 4.52it/s] 40%|████ | 19229/48008 [1:05:51<1:41:40, 4.72it/s] 40%|████ | 19230/48008 [1:05:51<1:38:27, 4.87it/s] 40%|████ | 19231/48008 [1:05:51<1:36:05, 4.99it/s] 40%|████ | 19232/48008 [1:05:51<1:34:14, 5.09it/s] 40%|████ | 19233/48008 [1:05:51<1:32:48, 5.17it/s] 40%|████ | 19234/48008 [1:05:52<1:31:18, 5.25it/s] 40%|████ | 19235/48008 [1:05:52<1:31:20, 5.25it/s] 40%|████ | 19236/48008 [1:05:52<1:31:31, 5.24it/s] 40%|████ | 19237/48008 [1:05:52<1:31:33, 5.24it/s] 40%|████ | 19238/48008 [1:05:52<1:31:21, 5.25it/s] 40%|████ | 19239/48008 [1:05:53<1:31:30, 5.24it/s] 40%|████ | 19240/48008 [1:05:53<1:32:24, 5.19it/s] 40%|████ | 19241/48008 [1:05:53<1:31:30, 5.24it/s] 40%|████ | 19242/48008 [1:05:53<1:31:17, 5.25it/s] 40%|████ | 19243/48008 [1:05:53<1:31:25, 5.24it/s] 40%|████ | 19244/48008 [1:05:54<1:32:54, 5.16it/s] 40%|████ | 19245/48008 [1:05:54<1:32:44, 5.17it/s] 40%|████ | 19246/48008 [1:05:54<1:32:30, 5.18it/s] 40%|████ | 19247/48008 [1:05:54<1:31:36, 5.23it/s] 40%|████ | 19248/48008 [1:05:54<1:31:00, 5.27it/s] 40%|████ | 19249/48008 [1:05:55<1:29:37, 5.35it/s] 40%|████ | 19250/48008 [1:05:55<1:30:12, 5.31it/s] {'loss': 0.3029, 'grad_norm': 0.640625, 'learning_rate': 0.00011980919846692218, 'epoch': 0.4} 40%|████ | 19250/48008 [1:05:55<1:30:12, 5.31it/s] 40%|████ | 19251/48008 [1:05:55<1:32:07, 5.20it/s] 40%|████ | 19252/48008 [1:05:55<1:31:56, 5.21it/s] 40%|████ | 19253/48008 [1:05:55<1:34:22, 5.08it/s] 40%|████ | 19254/48008 [1:05:56<1:36:04, 4.99it/s] 40%|████ | 19255/48008 [1:05:56<1:34:49, 5.05it/s] 40%|████ | 19256/48008 [1:05:56<2:01:44, 3.94it/s] 40%|████ | 19257/48008 [1:05:56<1:54:18, 4.19it/s] 40%|████ | 19258/48008 [1:05:57<1:47:58, 4.44it/s] 40%|████ | 19259/48008 [1:05:57<1:42:43, 4.66it/s] 40%|████ | 19260/48008 [1:05:57<1:39:35, 4.81it/s] 40%|████ | 19261/48008 [1:05:57<1:36:37, 4.96it/s] 40%|████ | 19262/48008 [1:05:57<1:34:57, 5.05it/s] 40%|████ | 19263/48008 [1:05:57<1:33:38, 5.12it/s] 40%|████ | 19264/48008 [1:05:58<1:31:48, 5.22it/s] 40%|████ | 19265/48008 [1:05:58<1:31:49, 5.22it/s] 40%|████ | 19266/48008 [1:05:58<1:33:36, 5.12it/s] 40%|████ | 19267/48008 [1:05:58<1:34:45, 5.06it/s] 40%|████ | 19268/48008 [1:05:58<1:34:07, 5.09it/s] 40%|████ | 19269/48008 [1:05:59<1:40:25, 4.77it/s] 40%|████ | 19270/48008 [1:05:59<1:37:29, 4.91it/s] 40%|████ | 19271/48008 [1:05:59<1:52:48, 4.25it/s] 40%|████ | 19272/48008 [1:05:59<1:45:50, 4.52it/s] 40%|████ | 19273/48008 [1:06:00<1:41:45, 4.71it/s] 40%|████ | 19274/48008 [1:06:00<1:38:11, 4.88it/s] 40%|████ | 19275/48008 [1:06:00<1:35:41, 5.00it/s] 40%|████ | 19276/48008 [1:06:00<1:34:28, 5.07it/s] 40%|████ | 19277/48008 [1:06:00<1:33:05, 5.14it/s] 40%|████ | 19278/48008 [1:06:00<1:31:57, 5.21it/s] 40%|████ | 19279/48008 [1:06:01<1:32:01, 5.20it/s] 40%|████ | 19280/48008 [1:06:01<1:31:58, 5.21it/s] 40%|████ | 19281/48008 [1:06:01<1:31:57, 5.21it/s] 40%|████ | 19282/48008 [1:06:01<1:32:33, 5.17it/s] 40%|████ | 19283/48008 [1:06:01<1:32:40, 5.17it/s] 40%|████ | 19284/48008 [1:06:02<1:31:56, 5.21it/s] 40%|████ | 19285/48008 [1:06:02<1:31:56, 5.21it/s] 40%|████ | 19286/48008 [1:06:02<1:31:11, 5.25it/s] 40%|████ | 19287/48008 [1:06:02<1:33:08, 5.14it/s] 40%|████ | 19288/48008 [1:06:02<1:32:42, 5.16it/s] 40%|████ | 19289/48008 [1:06:03<1:34:17, 5.08it/s] 40%|████ | 19290/48008 [1:06:03<1:34:08, 5.08it/s] 40%|████ | 19291/48008 [1:06:03<1:32:43, 5.16it/s] 40%|████ | 19292/48008 [1:06:03<1:32:39, 5.17it/s] 40%|████ | 19293/48008 [1:06:03<1:32:20, 5.18it/s] 40%|████ | 19294/48008 [1:06:04<1:31:48, 5.21it/s] 40%|████ | 19295/48008 [1:06:04<1:31:48, 5.21it/s] 40%|████ | 19296/48008 [1:06:04<1:33:46, 5.10it/s] 40%|████ | 19297/48008 [1:06:04<1:34:07, 5.08it/s] 40%|████ | 19298/48008 [1:06:04<1:33:32, 5.12it/s] 40%|████ | 19299/48008 [1:06:05<1:32:22, 5.18it/s] 40%|████ | 19300/48008 [1:06:05<1:33:52, 5.10it/s] {'loss': 0.3025, 'grad_norm': 0.150390625, 'learning_rate': 0.000119600899850025, 'epoch': 0.4} 40%|████ | 19300/48008 [1:06:05<1:33:52, 5.10it/s] 40%|████ | 19301/48008 [1:06:05<1:33:09, 5.14it/s] 40%|████ | 19302/48008 [1:06:05<1:33:40, 5.11it/s] 40%|████ | 19303/48008 [1:06:05<1:34:08, 5.08it/s] 40%|████ | 19304/48008 [1:06:06<2:01:23, 3.94it/s] 40%|████ | 19305/48008 [1:06:06<2:08:51, 3.71it/s] 40%|████ | 19306/48008 [1:06:06<1:59:38, 4.00it/s] 40%|████ | 19307/48008 [1:06:06<1:50:33, 4.33it/s] 40%|████ | 19308/48008 [1:06:07<1:44:31, 4.58it/s] 40%|████ | 19309/48008 [1:06:07<1:57:04, 4.09it/s] 40%|████ | 19310/48008 [1:06:07<1:50:52, 4.31it/s] 40%|████ | 19311/48008 [1:06:07<1:45:09, 4.55it/s] 40%|████ | 19312/48008 [1:06:08<1:40:30, 4.76it/s] 40%|████ | 19313/48008 [1:06:08<1:37:05, 4.93it/s] 40%|████ | 19314/48008 [1:06:08<1:36:55, 4.93it/s] 40%|████ | 19315/48008 [1:06:08<1:34:32, 5.06it/s] 40%|████ | 19316/48008 [1:06:08<1:50:00, 4.35it/s] 40%|████ | 19317/48008 [1:06:09<1:45:14, 4.54it/s] 40%|████ | 19318/48008 [1:06:09<1:40:28, 4.76it/s] 40%|████ | 19319/48008 [1:06:09<1:37:51, 4.89it/s] 40%|████ | 19320/48008 [1:06:09<1:35:54, 4.99it/s] 40%|████ | 19321/48008 [1:06:09<1:34:22, 5.07it/s] 40%|████ | 19322/48008 [1:06:10<1:32:59, 5.14it/s] 40%|████ | 19323/48008 [1:06:10<1:31:26, 5.23it/s] 40%|████ | 19324/48008 [1:06:10<1:34:11, 5.08it/s] 40%|████ | 19325/48008 [1:06:10<1:34:42, 5.05it/s] 40%|████ | 19326/48008 [1:06:10<1:34:00, 5.09it/s] 40%|████ | 19327/48008 [1:06:11<1:32:49, 5.15it/s] 40%|████ | 19328/48008 [1:06:11<1:32:10, 5.19it/s] 40%|████ | 19329/48008 [1:06:11<1:32:01, 5.19it/s] 40%|████ | 19330/48008 [1:06:11<1:32:52, 5.15it/s] 40%|████ | 19331/48008 [1:06:11<1:32:20, 5.18it/s] 40%|████ | 19332/48008 [1:06:11<1:33:05, 5.13it/s] 40%|████ | 19333/48008 [1:06:12<1:48:53, 4.39it/s] 40%|████ | 19334/48008 [1:06:12<1:45:25, 4.53it/s] 40%|████ | 19335/48008 [1:06:12<1:40:48, 4.74it/s] 40%|████ | 19336/48008 [1:06:12<1:39:48, 4.79it/s] 40%|████ | 19337/48008 [1:06:13<1:37:18, 4.91it/s] 40%|████ | 19338/48008 [1:06:13<1:35:19, 5.01it/s] 40%|████ | 19339/48008 [1:06:13<1:33:18, 5.12it/s] 40%|████ | 19340/48008 [1:06:13<1:32:59, 5.14it/s] 40%|████ | 19341/48008 [1:06:13<1:32:30, 5.16it/s] 40%|████ | 19342/48008 [1:06:14<1:32:30, 5.16it/s] 40%|████ | 19343/48008 [1:06:14<1:33:07, 5.13it/s] 40%|████ | 19344/48008 [1:06:14<1:32:17, 5.18it/s] 40%|████ | 19345/48008 [1:06:14<1:32:03, 5.19it/s] 40%|████ | 19346/48008 [1:06:14<1:32:06, 5.19it/s] 40%|████ | 19347/48008 [1:06:14<1:33:24, 5.11it/s] 40%|████ | 19348/48008 [1:06:15<1:49:24, 4.37it/s] 40%|████ | 19349/48008 [1:06:15<1:43:29, 4.62it/s] 40%|████ | 19350/48008 [1:06:15<1:39:54, 4.78it/s] {'loss': 0.2997, 'grad_norm': 0.1943359375, 'learning_rate': 0.00011939260123312782, 'epoch': 0.4} 40%|████ | 19350/48008 [1:06:15<1:39:54, 4.78it/s] 40%|████ | 19351/48008 [1:06:15<1:37:20, 4.91it/s] 40%|████ | 19352/48008 [1:06:16<1:35:42, 4.99it/s] 40%|████ | 19353/48008 [1:06:16<1:33:34, 5.10it/s] 40%|████ | 19354/48008 [1:06:16<1:32:46, 5.15it/s] 40%|████ | 19355/48008 [1:06:16<1:38:47, 4.83it/s] 40%|████ | 19356/48008 [1:06:16<1:38:02, 4.87it/s] 40%|████ | 19357/48008 [1:06:17<1:37:03, 4.92it/s] 40%|████ | 19358/48008 [1:06:17<1:35:06, 5.02it/s] 40%|████ | 19359/48008 [1:06:17<1:33:56, 5.08it/s] 40%|████ | 19360/48008 [1:06:17<1:49:46, 4.35it/s] 40%|████ | 19361/48008 [1:06:17<1:43:34, 4.61it/s] 40%|████ | 19362/48008 [1:06:18<1:39:12, 4.81it/s] 40%|████ | 19363/48008 [1:06:18<1:37:05, 4.92it/s] 40%|████ | 19364/48008 [1:06:18<1:34:56, 5.03it/s] 40%|████ | 19365/48008 [1:06:18<1:33:28, 5.11it/s] 40%|████ | 19366/48008 [1:06:18<1:32:52, 5.14it/s] 40%|████ | 19367/48008 [1:06:19<1:31:39, 5.21it/s] 40%|████ | 19368/48008 [1:06:19<1:33:07, 5.13it/s] 40%|████ | 19369/48008 [1:06:19<1:32:34, 5.16it/s] 40%|████ | 19370/48008 [1:06:19<1:31:17, 5.23it/s] 40%|████ | 19371/48008 [1:06:19<1:31:12, 5.23it/s] 40%|████ | 19372/48008 [1:06:20<1:31:00, 5.24it/s] 40%|████ | 19373/48008 [1:06:20<1:32:35, 5.15it/s] 40%|████ | 19374/48008 [1:06:20<1:31:54, 5.19it/s] 40%|████ | 19375/48008 [1:06:20<1:30:40, 5.26it/s] 40%|████ | 19376/48008 [1:06:20<1:30:13, 5.29it/s] 40%|████ | 19377/48008 [1:06:20<1:30:28, 5.27it/s] 40%|████ | 19378/48008 [1:06:21<1:29:45, 5.32it/s] 40%|████ | 19379/48008 [1:06:21<1:29:38, 5.32it/s] 40%|████ | 19380/48008 [1:06:21<1:36:37, 4.94it/s] 40%|████ | 19381/48008 [1:06:21<1:35:39, 4.99it/s] 40%|████ | 19382/48008 [1:06:22<2:02:19, 3.90it/s] 40%|████ | 19383/48008 [1:06:22<1:53:28, 4.20it/s] 40%|████ | 19384/48008 [1:06:22<1:47:09, 4.45it/s] 40%|████ | 19385/48008 [1:06:22<1:42:38, 4.65it/s] 40%|████ | 19386/48008 [1:06:22<1:39:10, 4.81it/s] 40%|████ | 19387/48008 [1:06:23<1:36:43, 4.93it/s] 40%|████ | 19388/48008 [1:06:23<1:36:25, 4.95it/s] 40%|████ | 19389/48008 [1:06:23<1:34:12, 5.06it/s] 40%|████ | 19390/48008 [1:06:23<1:34:32, 5.04it/s] 40%|████ | 19391/48008 [1:06:23<1:33:15, 5.11it/s] 40%|████ | 19392/48008 [1:06:24<1:32:22, 5.16it/s] 40%|████ | 19393/48008 [1:06:24<1:32:01, 5.18it/s] 40%|████ | 19394/48008 [1:06:24<1:32:00, 5.18it/s] 40%|████ | 19395/48008 [1:06:24<1:31:49, 5.19it/s] 40%|████ | 19396/48008 [1:06:24<1:31:19, 5.22it/s] 40%|████ | 19397/48008 [1:06:25<1:31:32, 5.21it/s] 40%|████ | 19398/48008 [1:06:25<1:31:08, 5.23it/s] 40%|████ | 19399/48008 [1:06:25<1:31:56, 5.19it/s] 40%|████ | 19400/48008 [1:06:25<1:34:17, 5.06it/s] {'loss': 0.3013, 'grad_norm': 0.1279296875, 'learning_rate': 0.00011918430261623063, 'epoch': 0.4} 40%|████ | 19400/48008 [1:06:25<1:34:17, 5.06it/s] 40%|████ | 19401/48008 [1:06:25<1:34:32, 5.04it/s] 40%|████ | 19402/48008 [1:06:26<1:35:07, 5.01it/s] 40%|████ | 19403/48008 [1:06:26<1:32:10, 5.17it/s] 40%|████ | 19404/48008 [1:06:26<1:31:57, 5.18it/s] 40%|████ | 19405/48008 [1:06:26<1:30:51, 5.25it/s] 40%|████ | 19406/48008 [1:06:26<1:31:19, 5.22it/s] 40%|████ | 19407/48008 [1:06:27<1:32:08, 5.17it/s] 40%|████ | 19408/48008 [1:06:27<1:32:17, 5.17it/s] 40%|████ | 19409/48008 [1:06:27<1:32:06, 5.17it/s] 40%|████ | 19410/48008 [1:06:27<1:32:27, 5.16it/s] 40%|████ | 19411/48008 [1:06:27<1:38:59, 4.81it/s] 40%|████ | 19412/48008 [1:06:28<1:36:47, 4.92it/s] 40%|████ | 19413/48008 [1:06:28<1:35:40, 4.98it/s] 40%|████ | 19414/48008 [1:06:28<1:33:51, 5.08it/s] 40%|████ | 19415/48008 [1:06:28<1:39:52, 4.77it/s] 40%|████ | 19416/48008 [1:06:28<1:37:18, 4.90it/s] 40%|████ | 19417/48008 [1:06:29<1:38:17, 4.85it/s] 40%|████ | 19418/48008 [1:06:29<1:37:32, 4.89it/s] 40%|████ | 19419/48008 [1:06:29<1:35:42, 4.98it/s] 40%|████ | 19420/48008 [1:06:29<1:35:41, 4.98it/s] 40%|████ | 19421/48008 [1:06:29<1:34:39, 5.03it/s] 40%|████ | 19422/48008 [1:06:30<1:33:39, 5.09it/s] 40%|████ | 19423/48008 [1:06:30<1:33:11, 5.11it/s] 40%|████ | 19424/48008 [1:06:30<1:32:14, 5.16it/s] 40%|████ | 19425/48008 [1:06:30<1:48:20, 4.40it/s] 40%|████ | 19426/48008 [1:06:30<1:43:11, 4.62it/s] 40%|████ | 19427/48008 [1:06:31<1:39:24, 4.79it/s] 40%|████ | 19428/48008 [1:06:31<1:37:05, 4.91it/s] 40%|████ | 19429/48008 [1:06:31<1:35:16, 5.00it/s] 40%|████ | 19430/48008 [1:06:31<1:50:29, 4.31it/s] 40%|████ | 19431/48008 [1:06:31<1:44:40, 4.55it/s] 40%|████ | 19432/48008 [1:06:32<1:42:19, 4.65it/s] 40%|████ | 19433/48008 [1:06:32<1:41:38, 4.69it/s] 40%|████ | 19434/48008 [1:06:32<1:38:58, 4.81it/s] 40%|████ | 19435/48008 [1:06:32<1:38:20, 4.84it/s] 40%|████ | 19436/48008 [1:06:33<2:04:12, 3.83it/s] 40%|████ | 19437/48008 [1:06:33<1:54:44, 4.15it/s] 40%|████ | 19438/48008 [1:06:33<1:47:45, 4.42it/s] 40%|████ | 19439/48008 [1:06:33<1:42:48, 4.63it/s] 40%|████ | 19440/48008 [1:06:33<1:38:59, 4.81it/s] 40%|████ | 19441/48008 [1:06:34<1:37:21, 4.89it/s] 40%|████ | 19442/48008 [1:06:34<1:34:32, 5.04it/s] 40%|████ | 19443/48008 [1:06:34<1:33:32, 5.09it/s] 41%|████ | 19444/48008 [1:06:34<1:32:19, 5.16it/s] 41%|████ | 19445/48008 [1:06:34<1:31:36, 5.20it/s] 41%|████ | 19446/48008 [1:06:35<1:31:37, 5.20it/s] 41%|████ | 19447/48008 [1:06:35<1:31:33, 5.20it/s] 41%|████ | 19448/48008 [1:06:35<1:32:21, 5.15it/s] 41%|████ | 19449/48008 [1:06:35<1:48:29, 4.39it/s] 41%|████ | 19450/48008 [1:06:35<1:42:51, 4.63it/s] 41%|████ | 19450/48008 [1:06:36<1:42:51, 4.63it/s]{'loss': 0.3013, 'grad_norm': 0.53515625, 'learning_rate': 0.00011897600399933344, 'epoch': 0.41} 41%|████ | 19451/48008 [1:06:36<1:39:16, 4.79it/s] 41%|████ | 19452/48008 [1:06:36<1:43:16, 4.61it/s] 41%|████ | 19453/48008 [1:06:37<3:23:37, 2.34it/s] 41%|████ | 19454/48008 [1:06:37<2:49:37, 2.81it/s] 41%|████ | 19455/48008 [1:06:37<2:42:25, 2.93it/s] 41%|████ | 19456/48008 [1:06:38<2:20:27, 3.39it/s] 41%|████ | 19457/48008 [1:06:38<2:05:24, 3.79it/s] 41%|████ | 19458/48008 [1:06:38<1:55:17, 4.13it/s] 41%|████ | 19459/48008 [1:06:38<1:47:38, 4.42it/s] 41%|████ | 19460/48008 [1:06:38<1:42:30, 4.64it/s] 41%|████ | 19461/48008 [1:06:38<1:40:34, 4.73it/s] 41%|████ | 19462/48008 [1:06:39<1:37:42, 4.87it/s] 41%|████ | 19463/48008 [1:06:39<1:35:10, 5.00it/s] 41%|████ | 19464/48008 [1:06:39<1:33:44, 5.08it/s] 41%|████ | 19465/48008 [1:06:39<1:33:45, 5.07it/s] 41%|████ | 19466/48008 [1:06:39<1:32:38, 5.13it/s] 41%|████ | 19467/48008 [1:06:40<1:32:25, 5.15it/s] 41%|████ | 19468/48008 [1:06:40<1:31:29, 5.20it/s] 41%|████ | 19469/48008 [1:06:40<1:31:54, 5.17it/s] 41%|████ | 19470/48008 [1:06:40<1:33:30, 5.09it/s] 41%|████ | 19471/48008 [1:06:40<1:31:58, 5.17it/s] 41%|████ | 19472/48008 [1:06:41<1:31:17, 5.21it/s] 41%|████ | 19473/48008 [1:06:41<1:30:18, 5.27it/s] 41%|████ | 19474/48008 [1:06:41<1:29:46, 5.30it/s] 41%|████ | 19475/48008 [1:06:41<1:31:12, 5.21it/s] 41%|████ | 19476/48008 [1:06:41<1:32:53, 5.12it/s] 41%|████ | 19477/48008 [1:06:42<1:31:34, 5.19it/s] 41%|████ | 19478/48008 [1:06:42<1:31:45, 5.18it/s] 41%|████ | 19479/48008 [1:06:42<1:31:32, 5.19it/s] 41%|████ | 19480/48008 [1:06:42<1:30:56, 5.23it/s] 41%|████ | 19481/48008 [1:06:42<1:47:46, 4.41it/s] 41%|████ | 19482/48008 [1:06:43<1:42:25, 4.64it/s] 41%|████ | 19483/48008 [1:06:43<1:38:14, 4.84it/s] 41%|████ | 19484/48008 [1:06:43<1:36:38, 4.92it/s] 41%|████ | 19485/48008 [1:06:43<1:34:51, 5.01it/s] 41%|████ | 19486/48008 [1:06:43<1:34:45, 5.02it/s] 41%|████ | 19487/48008 [1:06:44<1:35:09, 5.00it/s] 41%|████ | 19488/48008 [1:06:44<1:34:01, 5.06it/s] 41%|████ | 19489/48008 [1:06:44<1:32:57, 5.11it/s] 41%|████ | 19490/48008 [1:06:44<1:31:19, 5.20it/s] 41%|████ | 19491/48008 [1:06:44<1:31:33, 5.19it/s] 41%|████ | 19492/48008 [1:06:45<1:30:53, 5.23it/s] 41%|████ | 19493/48008 [1:06:45<1:31:21, 5.20it/s] 41%|████ | 19494/48008 [1:06:45<1:32:13, 5.15it/s] 41%|████ | 19495/48008 [1:06:45<1:32:21, 5.15it/s] 41%|████ | 19496/48008 [1:06:45<1:31:34, 5.19it/s] 41%|████ | 19497/48008 [1:06:46<1:31:56, 5.17it/s] 41%|████ | 19498/48008 [1:06:46<1:31:02, 5.22it/s] 41%|████ | 19499/48008 [1:06:46<1:31:35, 5.19it/s] 41%|████ | 19500/48008 [1:06:46<1:30:57, 5.22it/s] {'loss': 0.3039, 'grad_norm': 0.1494140625, 'learning_rate': 0.00011876770538243627, 'epoch': 0.41} 41%|████ | 19500/48008 [1:06:46<1:30:57, 5.22it/s] 41%|████ | 19501/48008 [1:06:46<1:31:16, 5.21it/s] 41%|████ | 19502/48008 [1:06:46<1:30:33, 5.25it/s] 41%|████ | 19503/48008 [1:06:47<1:31:36, 5.19it/s] 41%|████ | 19504/48008 [1:06:47<1:31:39, 5.18it/s] 41%|████ | 19505/48008 [1:06:47<1:30:28, 5.25it/s] 41%|████ | 19506/48008 [1:06:47<1:32:25, 5.14it/s] 41%|████ | 19507/48008 [1:06:47<1:32:03, 5.16it/s] 41%|████ | 19508/48008 [1:06:48<1:32:14, 5.15it/s] 41%|████ | 19509/48008 [1:06:48<1:31:55, 5.17it/s] 41%|████ | 19510/48008 [1:06:48<1:32:24, 5.14it/s] 41%|████ | 19511/48008 [1:06:48<1:31:01, 5.22it/s] 41%|████ | 19512/48008 [1:06:48<1:30:29, 5.25it/s] 41%|████ | 19513/48008 [1:06:49<1:30:35, 5.24it/s] 41%|████ | 19514/48008 [1:06:49<1:30:48, 5.23it/s] 41%|████ | 19515/48008 [1:06:49<1:37:33, 4.87it/s] 41%|████ | 19516/48008 [1:06:49<1:35:48, 4.96it/s] 41%|████ | 19517/48008 [1:06:49<1:34:03, 5.05it/s] 41%|████ | 19518/48008 [1:06:50<1:32:58, 5.11it/s] 41%|████ | 19519/48008 [1:06:50<1:31:49, 5.17it/s] 41%|████ | 19520/48008 [1:06:50<1:47:52, 4.40it/s] 41%|████ | 19521/48008 [1:06:50<1:43:43, 4.58it/s] 41%|████ | 19522/48008 [1:06:50<1:39:26, 4.77it/s] 41%|████ | 19523/48008 [1:06:51<1:36:51, 4.90it/s] 41%|████ | 19524/48008 [1:06:51<1:35:09, 4.99it/s] 41%|████ | 19525/48008 [1:06:51<1:33:07, 5.10it/s] 41%|████ | 19526/48008 [1:06:51<1:34:02, 5.05it/s] 41%|████ | 19527/48008 [1:06:51<1:32:59, 5.10it/s] 41%|████ | 19528/48008 [1:06:52<1:31:54, 5.16it/s] 41%|████ | 19529/48008 [1:06:52<1:32:23, 5.14it/s] 41%|████ | 19530/48008 [1:06:52<1:32:48, 5.11it/s] 41%|████ | 19531/48008 [1:06:52<1:31:36, 5.18it/s] 41%|████ | 19532/48008 [1:06:52<1:32:19, 5.14it/s] 41%|████ | 19533/48008 [1:06:53<1:48:58, 4.36it/s] 41%|████ | 19534/48008 [1:06:53<1:44:04, 4.56it/s] 41%|████ | 19535/48008 [1:06:53<1:42:52, 4.61it/s] 41%|████ | 19536/48008 [1:06:53<1:38:53, 4.80it/s] 41%|████ | 19537/48008 [1:06:54<1:37:53, 4.85it/s] 41%|████ | 19538/48008 [1:06:54<1:34:55, 5.00it/s] 41%|████ | 19539/48008 [1:06:54<1:34:19, 5.03it/s] 41%|████ | 19540/48008 [1:06:54<1:33:19, 5.08it/s] 41%|████ | 19541/48008 [1:06:54<1:34:51, 5.00it/s] 41%|████ | 19542/48008 [1:06:55<2:01:35, 3.90it/s] 41%|████ | 19543/48008 [1:06:55<1:53:16, 4.19it/s] 41%|████ | 19544/48008 [1:06:55<1:47:17, 4.42it/s] 41%|████ | 19545/48008 [1:06:55<1:43:25, 4.59it/s] 41%|████ | 19546/48008 [1:06:55<1:39:54, 4.75it/s] 41%|████ | 19547/48008 [1:06:56<1:38:45, 4.80it/s] 41%|████ | 19548/48008 [1:06:56<1:35:54, 4.95it/s] 41%|████ | 19549/48008 [1:06:56<2:02:16, 3.88it/s] 41%|████ | 19550/48008 [1:06:56<1:53:17, 4.19it/s] {'loss': 0.3011, 'grad_norm': 0.345703125, 'learning_rate': 0.00011855940676553907, 'epoch': 0.41} 41%|████ | 19550/48008 [1:06:56<1:53:17, 4.19it/s] 41%|████ | 19551/48008 [1:06:57<1:48:15, 4.38it/s] 41%|████ | 19552/48008 [1:06:57<1:42:33, 4.62it/s] 41%|████ | 19553/48008 [1:06:57<2:06:55, 3.74it/s] 41%|████ | 19554/48008 [1:06:57<1:56:10, 4.08it/s] 41%|████ | 19555/48008 [1:06:58<1:48:05, 4.39it/s] 41%|████ | 19556/48008 [1:06:58<1:44:09, 4.55it/s] 41%|████ | 19557/48008 [1:06:58<1:42:22, 4.63it/s] 41%|████ | 19558/48008 [1:06:58<1:39:04, 4.79it/s] 41%|████ | 19559/48008 [1:06:58<1:36:39, 4.91it/s] 41%|████ | 19560/48008 [1:06:59<1:35:14, 4.98it/s] 41%|████ | 19561/48008 [1:06:59<1:33:34, 5.07it/s] 41%|████ | 19562/48008 [1:06:59<1:32:44, 5.11it/s] 41%|████ | 19563/48008 [1:06:59<1:31:57, 5.16it/s] 41%|████ | 19564/48008 [1:06:59<1:31:48, 5.16it/s] 41%|████ | 19565/48008 [1:07:00<1:32:24, 5.13it/s] 41%|████ | 19566/48008 [1:07:00<1:31:50, 5.16it/s] 41%|████ | 19567/48008 [1:07:00<1:32:56, 5.10it/s] 41%|████ | 19568/48008 [1:07:00<1:32:04, 5.15it/s] 41%|████ | 19569/48008 [1:07:00<1:32:26, 5.13it/s] 41%|████ | 19570/48008 [1:07:01<1:32:05, 5.15it/s] 41%|████ | 19571/48008 [1:07:01<1:32:18, 5.13it/s] 41%|████ | 19572/48008 [1:07:01<1:32:08, 5.14it/s] 41%|████ | 19573/48008 [1:07:01<1:34:19, 5.02it/s] 41%|████ | 19574/48008 [1:07:01<1:33:03, 5.09it/s] 41%|████ | 19575/48008 [1:07:02<1:32:37, 5.12it/s] 41%|████ | 19576/48008 [1:07:02<1:32:21, 5.13it/s] 41%|████ | 19577/48008 [1:07:02<1:32:07, 5.14it/s] 41%|████ | 19578/48008 [1:07:02<1:31:48, 5.16it/s] 41%|████ | 19579/48008 [1:07:02<1:31:30, 5.18it/s] 41%|████ | 19580/48008 [1:07:02<1:33:20, 5.08it/s] 41%|████ | 19581/48008 [1:07:03<1:33:20, 5.08it/s] 41%|████ | 19582/48008 [1:07:03<1:31:58, 5.15it/s] 41%|████ | 19583/48008 [1:07:03<1:31:45, 5.16it/s] 41%|████ | 19584/48008 [1:07:03<1:32:14, 5.14it/s] 41%|████ | 19585/48008 [1:07:03<1:31:45, 5.16it/s] 41%|████ | 19586/48008 [1:07:04<1:31:14, 5.19it/s] 41%|████ | 19587/48008 [1:07:04<1:37:39, 4.85it/s] 41%|████ | 19588/48008 [1:07:04<1:34:37, 5.01it/s] 41%|████ | 19589/48008 [1:07:04<1:33:24, 5.07it/s] 41%|████ | 19590/48008 [1:07:04<1:33:26, 5.07it/s] 41%|████ | 19591/48008 [1:07:05<1:32:45, 5.11it/s] 41%|████ | 19592/48008 [1:07:05<1:34:51, 4.99it/s] 41%|████ | 19593/48008 [1:07:05<1:33:24, 5.07it/s] 41%|████ | 19594/48008 [1:07:05<2:00:21, 3.93it/s] 41%|████ | 19595/48008 [1:07:06<1:51:34, 4.24it/s] 41%|████ | 19596/48008 [1:07:06<1:52:42, 4.20it/s] 41%|████ | 19597/48008 [1:07:06<2:13:27, 3.55it/s] 41%|████ | 19598/48008 [1:07:06<2:01:50, 3.89it/s] 41%|████ | 19599/48008 [1:07:07<1:52:42, 4.20it/s] 41%|████ | 19600/48008 [1:07:07<1:47:44, 4.39it/s] {'loss': 0.3017, 'grad_norm': 0.177734375, 'learning_rate': 0.0001183511081486419, 'epoch': 0.41} 41%|████ | 19600/48008 [1:07:07<1:47:44, 4.39it/s] 41%|████ | 19601/48008 [1:07:07<1:42:32, 4.62it/s] 41%|████ | 19602/48008 [1:07:07<1:40:37, 4.70it/s] 41%|████ | 19603/48008 [1:07:07<1:38:28, 4.81it/s] 41%|████ | 19604/48008 [1:07:08<1:35:58, 4.93it/s] 41%|████ | 19605/48008 [1:07:08<1:33:43, 5.05it/s] 41%|████ | 19606/48008 [1:07:08<1:34:44, 5.00it/s] 41%|████ | 19607/48008 [1:07:08<1:33:31, 5.06it/s] 41%|████ | 19608/48008 [1:07:08<1:32:43, 5.10it/s] 41%|████ | 19609/48008 [1:07:09<1:38:38, 4.80it/s] 41%|████ | 19610/48008 [1:07:09<1:36:21, 4.91it/s] 41%|████ | 19611/48008 [1:07:09<1:36:13, 4.92it/s] 41%|████ | 19612/48008 [1:07:09<1:36:18, 4.91it/s] 41%|████ | 19613/48008 [1:07:09<1:34:12, 5.02it/s] 41%|████ | 19614/48008 [1:07:10<1:32:47, 5.10it/s] 41%|████ | 19615/48008 [1:07:10<1:59:58, 3.94it/s] 41%|████ | 19616/48008 [1:07:10<1:50:55, 4.27it/s] 41%|████ | 19617/48008 [1:07:10<1:43:50, 4.56it/s] 41%|████ | 19618/48008 [1:07:11<1:40:02, 4.73it/s] 41%|████ | 19619/48008 [1:07:11<1:37:30, 4.85it/s] 41%|████ | 19620/48008 [1:07:11<1:35:52, 4.94it/s] 41%|████ | 19621/48008 [1:07:11<1:33:59, 5.03it/s] 41%|████ | 19622/48008 [1:07:11<1:32:46, 5.10it/s] 41%|████ | 19623/48008 [1:07:12<1:31:36, 5.16it/s] 41%|████ | 19624/48008 [1:07:12<1:37:44, 4.84it/s] 41%|████ | 19625/48008 [1:07:12<1:35:19, 4.96it/s] 41%|████ | 19626/48008 [1:07:12<1:34:32, 5.00it/s] 41%|████ | 19627/48008 [1:07:12<1:33:20, 5.07it/s] 41%|████ | 19628/48008 [1:07:13<1:32:49, 5.10it/s] 41%|████ | 19629/48008 [1:07:13<1:32:16, 5.13it/s] 41%|████ | 19630/48008 [1:07:13<1:31:28, 5.17it/s] 41%|████ | 19631/48008 [1:07:13<1:37:41, 4.84it/s] 41%|████ | 19632/48008 [1:07:13<1:35:55, 4.93it/s] 41%|████ | 19633/48008 [1:07:14<1:33:57, 5.03it/s] 41%|████ | 19634/48008 [1:07:14<1:33:06, 5.08it/s] 41%|████ | 19635/48008 [1:07:14<1:31:15, 5.18it/s] 41%|████ | 19636/48008 [1:07:14<1:31:11, 5.18it/s] 41%|████ | 19637/48008 [1:07:14<1:31:12, 5.18it/s] 41%|████ | 19638/48008 [1:07:14<1:32:38, 5.10it/s] 41%|████ | 19639/48008 [1:07:15<1:33:37, 5.05it/s] 41%|████ | 19640/48008 [1:07:15<1:32:50, 5.09it/s] 41%|████ | 19641/48008 [1:07:15<1:31:44, 5.15it/s] 41%|████ | 19642/48008 [1:07:15<1:30:36, 5.22it/s] 41%|████ | 19643/48008 [1:07:15<1:30:28, 5.23it/s] 41%|████ | 19644/48008 [1:07:16<1:30:08, 5.24it/s] 41%|████ | 19645/48008 [1:07:16<1:31:15, 5.18it/s] 41%|████ | 19646/48008 [1:07:16<1:30:17, 5.24it/s] 41%|████ | 19647/48008 [1:07:16<1:30:26, 5.23it/s] 41%|████ | 19648/48008 [1:07:16<1:32:09, 5.13it/s] 41%|████ | 19649/48008 [1:07:17<1:32:12, 5.13it/s] 41%|████ | 19650/48008 [1:07:17<1:31:25, 5.17it/s] {'loss': 0.301, 'grad_norm': 0.2294921875, 'learning_rate': 0.00011814280953174472, 'epoch': 0.41} 41%|████ | 19650/48008 [1:07:17<1:31:25, 5.17it/s] 41%|████ | 19651/48008 [1:07:17<1:31:46, 5.15it/s] 41%|████ | 19652/48008 [1:07:17<1:31:19, 5.18it/s] 41%|████ | 19653/48008 [1:07:17<1:31:32, 5.16it/s] 41%|████ | 19654/48008 [1:07:18<1:31:02, 5.19it/s] 41%|████ | 19655/48008 [1:07:18<1:30:38, 5.21it/s] 41%|████ | 19656/48008 [1:07:18<1:33:15, 5.07it/s] 41%|████ | 19657/48008 [1:07:18<1:31:59, 5.14it/s] 41%|████ | 19658/48008 [1:07:18<1:33:06, 5.07it/s] 41%|████ | 19659/48008 [1:07:19<1:32:27, 5.11it/s] 41%|████ | 19660/48008 [1:07:19<1:31:36, 5.16it/s] 41%|████ | 19661/48008 [1:07:19<1:31:25, 5.17it/s] 41%|████ | 19662/48008 [1:07:19<1:31:19, 5.17it/s] 41%|████ | 19663/48008 [1:07:19<1:31:33, 5.16it/s] 41%|████ | 19664/48008 [1:07:20<1:47:35, 4.39it/s] 41%|████ | 19665/48008 [1:07:20<1:43:43, 4.55it/s] 41%|████ | 19666/48008 [1:07:20<1:39:06, 4.77it/s] 41%|████ | 19667/48008 [1:07:20<1:36:17, 4.91it/s] 41%|████ | 19668/48008 [1:07:21<2:02:06, 3.87it/s] 41%|████ | 19669/48008 [1:07:21<1:52:27, 4.20it/s] 41%|████ | 19670/48008 [1:07:21<1:45:41, 4.47it/s] 41%|████ | 19671/48008 [1:07:21<1:41:10, 4.67it/s] 41%|████ | 19672/48008 [1:07:21<1:37:20, 4.85it/s] 41%|████ | 19673/48008 [1:07:22<1:35:26, 4.95it/s] 41%|████ | 19674/48008 [1:07:22<1:33:59, 5.02it/s] 41%|████ | 19675/48008 [1:07:22<1:32:22, 5.11it/s] 41%|████ | 19676/48008 [1:07:22<1:33:17, 5.06it/s] 41%|████ | 19677/48008 [1:07:22<1:32:24, 5.11it/s] 41%|████ | 19678/48008 [1:07:23<1:31:54, 5.14it/s] 41%|████ | 19679/48008 [1:07:23<1:31:15, 5.17it/s] 41%|████ | 19680/48008 [1:07:23<1:32:28, 5.11it/s] 41%|████ | 19681/48008 [1:07:23<1:31:07, 5.18it/s] 41%|████ | 19682/48008 [1:07:23<1:32:26, 5.11it/s] 41%|████ | 19683/48008 [1:07:23<1:31:39, 5.15it/s] 41%|████ | 19684/48008 [1:07:24<1:31:36, 5.15it/s] 41%|████ | 19685/48008 [1:07:24<1:30:16, 5.23it/s] 41%|████ | 19686/48008 [1:07:24<1:29:39, 5.27it/s] 41%|████ | 19687/48008 [1:07:24<1:32:12, 5.12it/s] 41%|████ | 19688/48008 [1:07:24<1:31:54, 5.14it/s] 41%|████ | 19689/48008 [1:07:25<1:30:55, 5.19it/s] 41%|████ | 19690/48008 [1:07:25<1:30:41, 5.20it/s] 41%|████ | 19691/48008 [1:07:25<1:30:30, 5.21it/s] 41%|████ | 19692/48008 [1:07:25<1:30:10, 5.23it/s] 41%|████ | 19693/48008 [1:07:25<1:30:24, 5.22it/s] 41%|████ | 19694/48008 [1:07:26<1:57:46, 4.01it/s] 41%|████ | 19695/48008 [1:07:26<1:48:57, 4.33it/s] 41%|████ | 19696/48008 [1:07:26<1:42:56, 4.58it/s] 41%|████ | 19697/48008 [1:07:26<1:39:26, 4.75it/s] 41%|████ | 19698/48008 [1:07:27<1:53:11, 4.17it/s] 41%|████ | 19699/48008 [1:07:27<1:47:19, 4.40it/s] 41%|████ | 19700/48008 [1:07:27<1:41:39, 4.64it/s] {'loss': 0.303, 'grad_norm': 0.6015625, 'learning_rate': 0.00011793451091484752, 'epoch': 0.41} 41%|████ | 19700/48008 [1:07:27<1:41:39, 4.64it/s] 41%|████ | 19701/48008 [1:07:27<1:38:24, 4.79it/s] 41%|████ | 19702/48008 [1:07:27<1:37:16, 4.85it/s] 41%|████ | 19703/48008 [1:07:28<1:35:05, 4.96it/s] 41%|████ | 19704/48008 [1:07:28<1:34:13, 5.01it/s] 41%|████ | 19705/48008 [1:07:28<1:32:24, 5.10it/s] 41%|████ | 19706/48008 [1:07:28<1:31:55, 5.13it/s] 41%|████ | 19707/48008 [1:07:28<1:31:07, 5.18it/s] 41%|████ | 19708/48008 [1:07:29<1:30:35, 5.21it/s] 41%|████ | 19709/48008 [1:07:29<1:36:54, 4.87it/s] 41%|████ | 19710/48008 [1:07:29<1:36:36, 4.88it/s] 41%|████ | 19711/48008 [1:07:29<1:37:03, 4.86it/s] 41%|████ | 19712/48008 [1:07:29<1:35:02, 4.96it/s] 41%|████ | 19713/48008 [1:07:30<1:33:10, 5.06it/s] 41%|████ | 19714/48008 [1:07:30<1:32:30, 5.10it/s] 41%|████ | 19715/48008 [1:07:30<1:31:49, 5.13it/s] 41%|████ | 19716/48008 [1:07:30<1:31:07, 5.17it/s] 41%|████ | 19717/48008 [1:07:30<1:30:32, 5.21it/s] 41%|████ | 19718/48008 [1:07:31<1:29:41, 5.26it/s] 41%|████ | 19719/48008 [1:07:31<1:29:04, 5.29it/s] 41%|████ | 19720/48008 [1:07:31<1:30:07, 5.23it/s] 41%|████ | 19721/48008 [1:07:31<1:30:13, 5.23it/s] 41%|████ | 19722/48008 [1:07:31<1:29:18, 5.28it/s] 41%|████ | 19723/48008 [1:07:32<1:29:04, 5.29it/s] 41%|████ | 19724/48008 [1:07:32<1:31:16, 5.16it/s] 41%|████ | 19725/48008 [1:07:32<1:31:10, 5.17it/s] 41%|████ | 19726/48008 [1:07:32<1:32:40, 5.09it/s] 41%|████ | 19727/48008 [1:07:32<1:32:48, 5.08it/s] 41%|████ | 19728/48008 [1:07:33<1:32:01, 5.12it/s] 41%|████ | 19729/48008 [1:07:33<1:32:49, 5.08it/s] 41%|████ | 19730/48008 [1:07:33<1:38:47, 4.77it/s] 41%|████ | 19731/48008 [1:07:33<1:37:21, 4.84it/s] 41%|████ | 19732/48008 [1:07:33<1:35:24, 4.94it/s] 41%|████ | 19733/48008 [1:07:34<1:32:57, 5.07it/s] 41%|████ | 19734/48008 [1:07:34<1:32:14, 5.11it/s] 41%|████ | 19735/48008 [1:07:34<1:31:17, 5.16it/s] 41%|████ | 19736/48008 [1:07:34<1:33:35, 5.03it/s] 41%|████ | 19737/48008 [1:07:34<1:34:12, 5.00it/s] 41%|████ | 19738/48008 [1:07:35<1:32:42, 5.08it/s] 41%|████ | 19739/48008 [1:07:35<1:31:41, 5.14it/s] 41%|████ | 19740/48008 [1:07:35<1:31:24, 5.15it/s] 41%|████ | 19741/48008 [1:07:35<1:30:54, 5.18it/s] 41%|████ | 19742/48008 [1:07:35<1:30:11, 5.22it/s] 41%|████ | 19743/48008 [1:07:35<1:31:47, 5.13it/s] 41%|████ | 19744/48008 [1:07:36<1:30:49, 5.19it/s] 41%|████ | 19745/48008 [1:07:36<1:30:31, 5.20it/s] 41%|████ | 19746/48008 [1:07:36<1:30:01, 5.23it/s] 41%|████ | 19747/48008 [1:07:36<1:30:05, 5.23it/s] 41%|████ | 19748/48008 [1:07:36<1:29:29, 5.26it/s] 41%|████ | 19749/48008 [1:07:37<1:31:10, 5.17it/s] 41%|████ | 19750/48008 [1:07:37<1:30:34, 5.20it/s] {'loss': 0.3024, 'grad_norm': 0.1572265625, 'learning_rate': 0.00011772621229795036, 'epoch': 0.41} 41%|████ | 19750/48008 [1:07:37<1:30:34, 5.20it/s] 41%|████ | 19751/48008 [1:07:37<1:31:25, 5.15it/s] 41%|████ | 19752/48008 [1:07:37<1:32:29, 5.09it/s] 41%|████ | 19753/48008 [1:07:37<1:31:31, 5.15it/s] 41%|████ | 19754/48008 [1:07:38<1:30:48, 5.19it/s] 41%|████ | 19755/48008 [1:07:38<1:31:36, 5.14it/s] 41%|████ | 19756/48008 [1:07:38<1:30:38, 5.19it/s] 41%|████ | 19757/48008 [1:07:38<1:30:24, 5.21it/s] 41%|████ | 19758/48008 [1:07:38<1:30:18, 5.21it/s] 41%|████ | 19759/48008 [1:07:39<1:57:36, 4.00it/s] 41%|████ | 19760/48008 [1:07:39<1:49:17, 4.31it/s] 41%|████ | 19761/48008 [1:07:39<1:43:40, 4.54it/s] 41%|████ | 19762/48008 [1:07:39<1:39:55, 4.71it/s] 41%|████ | 19763/48008 [1:07:40<1:36:17, 4.89it/s] 41%|████ | 19764/48008 [1:07:40<1:34:25, 4.99it/s] 41%|████ | 19765/48008 [1:07:40<1:33:36, 5.03it/s] 41%|████ | 19766/48008 [1:07:40<1:32:46, 5.07it/s] 41%|████ | 19767/48008 [1:07:40<1:32:50, 5.07it/s] 41%|████ | 19768/48008 [1:07:41<1:38:17, 4.79it/s] 41%|████ | 19769/48008 [1:07:41<1:37:20, 4.83it/s] 41%|████ | 19770/48008 [1:07:41<1:35:03, 4.95it/s] 41%|████ | 19771/48008 [1:07:41<1:36:09, 4.89it/s] 41%|████ | 19772/48008 [1:07:41<1:34:26, 4.98it/s] 41%|████ | 19773/48008 [1:07:42<1:33:30, 5.03it/s] 41%|████ | 19774/48008 [1:07:42<1:31:52, 5.12it/s] 41%|████ | 19775/48008 [1:07:42<1:31:26, 5.15it/s] 41%|████ | 19776/48008 [1:07:42<1:30:23, 5.21it/s] 41%|████ | 19777/48008 [1:07:42<1:29:44, 5.24it/s] 41%|████ | 19778/48008 [1:07:42<1:31:21, 5.15it/s] 41%|████ | 19779/48008 [1:07:43<1:30:13, 5.21it/s] 41%|████ | 19780/48008 [1:07:43<1:31:48, 5.12it/s] 41%|████ | 19781/48008 [1:07:43<1:32:22, 5.09it/s] 41%|████ | 19782/48008 [1:07:43<1:31:41, 5.13it/s] 41%|████ | 19783/48008 [1:07:43<1:31:01, 5.17it/s] 41%|████ | 19784/48008 [1:07:44<1:30:34, 5.19it/s] 41%|████ | 19785/48008 [1:07:44<1:30:11, 5.22it/s] 41%|████ | 19786/48008 [1:07:44<1:30:15, 5.21it/s] 41%|████ | 19787/48008 [1:07:44<1:29:34, 5.25it/s] 41%|████ | 19788/48008 [1:07:44<1:31:31, 5.14it/s] 41%|████ | 19789/48008 [1:07:45<1:30:14, 5.21it/s] 41%|████ | 19790/48008 [1:07:45<1:30:48, 5.18it/s] 41%|████ | 19791/48008 [1:07:45<1:29:55, 5.23it/s] 41%|████ | 19792/48008 [1:07:45<1:30:40, 5.19it/s] 41%|████ | 19793/48008 [1:07:45<1:30:24, 5.20it/s] 41%|████ | 19794/48008 [1:07:46<1:30:13, 5.21it/s] 41%|████ | 19795/48008 [1:07:46<1:30:08, 5.22it/s] 41%|████ | 19796/48008 [1:07:46<1:30:04, 5.22it/s] 41%|████ | 19797/48008 [1:07:46<1:31:31, 5.14it/s] 41%|████ | 19798/48008 [1:07:46<1:31:08, 5.16it/s] 41%|████ | 19799/48008 [1:07:47<1:32:26, 5.09it/s] 41%|████ | 19800/48008 [1:07:47<1:31:56, 5.11it/s] 41%|████ | 19800/48008 [1:07:47<1:31:56, 5.11it/s]{'loss': 0.3021, 'grad_norm': 0.1396484375, 'learning_rate': 0.00011751791368105316, 'epoch': 0.41} 41%|████ | 19801/48008 [1:07:47<1:32:16, 5.09it/s] 41%|████ | 19802/48008 [1:07:47<1:31:44, 5.12it/s] 41%|████ | 19803/48008 [1:07:47<1:30:48, 5.18it/s] 41%|████▏ | 19804/48008 [1:07:47<1:30:06, 5.22it/s] 41%|████▏ | 19805/48008 [1:07:48<1:30:06, 5.22it/s] 41%|████▏ | 19806/48008 [1:07:48<1:30:51, 5.17it/s] 41%|████▏ | 19807/48008 [1:07:48<1:30:41, 5.18it/s] 41%|████▏ | 19808/48008 [1:07:48<1:30:24, 5.20it/s] 41%|████▏ | 19809/48008 [1:07:48<1:28:45, 5.30it/s] 41%|████▏ | 19810/48008 [1:07:49<1:29:36, 5.24it/s] 41%|████▏ | 19811/48008 [1:07:49<1:36:15, 4.88it/s] 41%|████▏ | 19812/48008 [1:07:49<1:34:01, 5.00it/s] 41%|████▏ | 19813/48008 [1:07:49<2:00:21, 3.90it/s] 41%|████▏ | 19814/48008 [1:07:50<1:52:28, 4.18it/s] 41%|████▏ | 19815/48008 [1:07:50<1:45:07, 4.47it/s] 41%|████▏ | 19816/48008 [1:07:50<1:39:44, 4.71it/s] 41%|████▏ | 19817/48008 [1:07:50<1:36:03, 4.89it/s] 41%|████▏ | 19818/48008 [1:07:50<1:35:39, 4.91it/s] 41%|████▏ | 19819/48008 [1:07:51<1:36:19, 4.88it/s] 41%|████▏ | 19820/48008 [1:07:51<1:33:56, 5.00it/s] 41%|████▏ | 19821/48008 [1:07:51<1:33:03, 5.05it/s] 41%|████▏ | 19822/48008 [1:07:51<1:31:54, 5.11it/s] 41%|████▏ | 19823/48008 [1:07:51<1:31:19, 5.14it/s] 41%|████▏ | 19824/48008 [1:07:52<1:32:27, 5.08it/s] 41%|████▏ | 19825/48008 [1:07:52<1:32:34, 5.07it/s] 41%|████▏ | 19826/48008 [1:07:52<1:33:22, 5.03it/s] 41%|████▏ | 19827/48008 [1:07:52<1:38:51, 4.75it/s] 41%|████▏ | 19828/48008 [1:07:52<1:35:51, 4.90it/s] 41%|████▏ | 19829/48008 [1:07:53<1:35:26, 4.92it/s] 41%|████▏ | 19830/48008 [1:07:53<1:33:35, 5.02it/s] 41%|████▏ | 19831/48008 [1:07:53<1:31:54, 5.11it/s] 41%|████▏ | 19832/48008 [1:07:53<1:33:47, 5.01it/s] 41%|████▏ | 19833/48008 [1:07:53<1:32:21, 5.08it/s] 41%|████▏ | 19834/48008 [1:07:54<1:31:40, 5.12it/s] 41%|████▏ | 19835/48008 [1:07:54<1:32:22, 5.08it/s] 41%|████▏ | 19836/48008 [1:07:54<1:31:33, 5.13it/s] 41%|████▏ | 19837/48008 [1:07:54<1:30:20, 5.20it/s] 41%|████▏ | 19838/48008 [1:07:54<1:37:13, 4.83it/s] 41%|████▏ | 19839/48008 [1:07:55<1:34:47, 4.95it/s] 41%|████▏ | 19840/48008 [1:07:55<1:33:48, 5.00it/s] 41%|████▏ | 19841/48008 [1:07:55<1:32:46, 5.06it/s] 41%|████▏ | 19842/48008 [1:07:55<1:31:58, 5.10it/s] 41%|████▏ | 19843/48008 [1:07:56<1:58:44, 3.95it/s] 41%|████▏ | 19844/48008 [1:07:56<1:49:50, 4.27it/s] 41%|████▏ | 19845/48008 [1:07:56<1:46:21, 4.41it/s] 41%|████▏ | 19846/48008 [1:07:56<1:42:30, 4.58it/s] 41%|████▏ | 19847/48008 [1:07:56<1:45:15, 4.46it/s] 41%|████▏ | 19848/48008 [1:07:57<1:40:02, 4.69it/s] 41%|████▏ | 19849/48008 [1:07:57<1:36:50, 4.85it/s] 41%|████▏ | 19850/48008 [1:07:57<1:34:42, 4.96it/s] {'loss': 0.2998, 'grad_norm': 0.1572265625, 'learning_rate': 0.00011730961506415599, 'epoch': 0.41} 41%|████▏ | 19850/48008 [1:07:57<1:34:42, 4.96it/s] 41%|████▏ | 19851/48008 [1:07:57<1:33:04, 5.04it/s] 41%|████▏ | 19852/48008 [1:07:57<1:31:52, 5.11it/s] 41%|████▏ | 19853/48008 [1:07:58<1:30:52, 5.16it/s] 41%|████▏ | 19854/48008 [1:07:58<1:30:19, 5.20it/s] 41%|████▏ | 19855/48008 [1:07:58<1:29:15, 5.26it/s] 41%|████▏ | 19856/48008 [1:07:58<1:29:17, 5.26it/s] 41%|████▏ | 19857/48008 [1:07:58<1:29:10, 5.26it/s] 41%|████▏ | 19858/48008 [1:07:58<1:30:15, 5.20it/s] 41%|████▏ | 19859/48008 [1:07:59<1:31:54, 5.10it/s] 41%|████▏ | 19860/48008 [1:07:59<1:31:23, 5.13it/s] 41%|████▏ | 19861/48008 [1:07:59<1:32:41, 5.06it/s] 41%|████▏ | 19862/48008 [1:07:59<1:31:48, 5.11it/s] 41%|████▏ | 19863/48008 [1:07:59<1:32:42, 5.06it/s] 41%|████▏ | 19864/48008 [1:08:00<1:32:38, 5.06it/s] 41%|████▏ | 19865/48008 [1:08:00<1:38:12, 4.78it/s] 41%|████▏ | 19866/48008 [1:08:00<1:35:59, 4.89it/s] 41%|████▏ | 19867/48008 [1:08:00<1:34:29, 4.96it/s] 41%|████▏ | 19868/48008 [1:08:00<1:33:19, 5.03it/s] 41%|████▏ | 19869/48008 [1:08:01<1:59:29, 3.92it/s] 41%|████▏ | 19870/48008 [1:08:01<1:51:24, 4.21it/s] 41%|████▏ | 19871/48008 [1:08:01<1:47:26, 4.36it/s] 41%|████▏ | 19872/48008 [1:08:01<1:42:02, 4.60it/s] 41%|████▏ | 19873/48008 [1:08:02<1:39:43, 4.70it/s] 41%|████▏ | 19874/48008 [1:08:02<1:37:04, 4.83it/s] 41%|████▏ | 19875/48008 [1:08:02<1:36:25, 4.86it/s] 41%|████▏ | 19876/48008 [1:08:02<1:35:09, 4.93it/s] 41%|████▏ | 19877/48008 [1:08:02<1:33:45, 5.00it/s] 41%|████▏ | 19878/48008 [1:08:03<1:32:33, 5.07it/s] 41%|████▏ | 19879/48008 [1:08:03<1:32:35, 5.06it/s] 41%|████▏ | 19880/48008 [1:08:03<1:31:43, 5.11it/s] 41%|████▏ | 19881/48008 [1:08:03<1:30:56, 5.16it/s] 41%|████▏ | 19882/48008 [1:08:03<1:30:26, 5.18it/s] 41%|████▏ | 19883/48008 [1:08:04<1:30:14, 5.19it/s] 41%|████▏ | 19884/48008 [1:08:04<1:29:49, 5.22it/s] 41%|████▏ | 19885/48008 [1:08:04<1:29:44, 5.22it/s] 41%|████▏ | 19886/48008 [1:08:04<1:36:05, 4.88it/s] 41%|████▏ | 19887/48008 [1:08:04<1:34:46, 4.95it/s] 41%|████▏ | 19888/48008 [1:08:05<1:33:20, 5.02it/s] 41%|████▏ | 19889/48008 [1:08:05<1:32:49, 5.05it/s] 41%|████▏ | 19890/48008 [1:08:05<1:32:21, 5.07it/s] 41%|████▏ | 19891/48008 [1:08:05<1:30:59, 5.15it/s] 41%|████▏ | 19892/48008 [1:08:05<1:31:07, 5.14it/s] 41%|████▏ | 19893/48008 [1:08:06<1:29:57, 5.21it/s] 41%|████▏ | 19894/48008 [1:08:06<1:28:44, 5.28it/s] 41%|████▏ | 19895/48008 [1:08:06<1:29:10, 5.25it/s] 41%|████▏ | 19896/48008 [1:08:06<1:28:42, 5.28it/s] 41%|████▏ | 19897/48008 [1:08:06<1:30:54, 5.15it/s] 41%|████▏ | 19898/48008 [1:08:07<1:31:05, 5.14it/s] 41%|████▏ | 19899/48008 [1:08:07<1:32:02, 5.09it/s] 41%|████▏ | 19900/48008 [1:08:07<1:30:35, 5.17it/s] {'loss': 0.3016, 'grad_norm': 0.1728515625, 'learning_rate': 0.0001171013164472588, 'epoch': 0.41} 41%|████▏ | 19900/48008 [1:08:07<1:30:35, 5.17it/s] 41%|████▏ | 19901/48008 [1:08:07<1:32:01, 5.09it/s] 41%|████▏ | 19902/48008 [1:08:07<1:32:23, 5.07it/s] 41%|████▏ | 19903/48008 [1:08:08<1:31:14, 5.13it/s] 41%|████▏ | 19904/48008 [1:08:08<1:31:28, 5.12it/s] 41%|████▏ | 19905/48008 [1:08:08<1:47:07, 4.37it/s] 41%|████▏ | 19906/48008 [1:08:08<1:41:21, 4.62it/s] 41%|████▏ | 19907/48008 [1:08:08<1:37:59, 4.78it/s] 41%|████▏ | 19908/48008 [1:08:09<1:35:25, 4.91it/s] 41%|████▏ | 19909/48008 [1:08:09<1:33:23, 5.01it/s] 41%|████▏ | 19910/48008 [1:08:09<1:32:08, 5.08it/s] 41%|████▏ | 19911/48008 [1:08:09<1:31:16, 5.13it/s] 41%|████▏ | 19912/48008 [1:08:09<1:30:17, 5.19it/s] 41%|████▏ | 19913/48008 [1:08:10<1:30:06, 5.20it/s] 41%|████▏ | 19914/48008 [1:08:10<1:29:22, 5.24it/s] 41%|████▏ | 19915/48008 [1:08:10<1:29:34, 5.23it/s] 41%|████▏ | 19916/48008 [1:08:11<3:46:02, 2.07it/s] 41%|████▏ | 19917/48008 [1:08:11<3:06:07, 2.52it/s] 41%|████▏ | 19918/48008 [1:08:11<2:36:41, 2.99it/s] 41%|████▏ | 19919/48008 [1:08:12<2:17:17, 3.41it/s] 41%|████▏ | 19920/48008 [1:08:12<2:02:52, 3.81it/s] 41%|████▏ | 19921/48008 [1:08:12<1:52:58, 4.14it/s] 41%|████▏ | 19922/48008 [1:08:12<1:45:06, 4.45it/s] 41%|████▏ | 19923/48008 [1:08:12<1:40:41, 4.65it/s] 42%|████▏ | 19924/48008 [1:08:13<1:37:44, 4.79it/s] 42%|████▏ | 19925/48008 [1:08:13<1:35:09, 4.92it/s] 42%|████▏ | 19926/48008 [1:08:13<1:32:40, 5.05it/s] 42%|████▏ | 19927/48008 [1:08:13<1:31:43, 5.10it/s] 42%|████▏ | 19928/48008 [1:08:13<1:30:48, 5.15it/s] 42%|████▏ | 19929/48008 [1:08:14<1:29:38, 5.22it/s] 42%|████▏ | 19930/48008 [1:08:14<1:36:11, 4.87it/s] 42%|████▏ | 19931/48008 [1:08:14<1:33:59, 4.98it/s] 42%|████▏ | 19932/48008 [1:08:14<1:35:32, 4.90it/s] 42%|████▏ | 19933/48008 [1:08:14<1:35:09, 4.92it/s] 42%|████▏ | 19934/48008 [1:08:15<1:33:40, 4.99it/s] 42%|████▏ | 19935/48008 [1:08:15<1:38:59, 4.73it/s] 42%|████▏ | 19936/48008 [1:08:15<1:37:55, 4.78it/s] 42%|████▏ | 19937/48008 [1:08:15<1:35:59, 4.87it/s] 42%|████▏ | 19938/48008 [1:08:15<1:34:48, 4.93it/s] 42%|████▏ | 19939/48008 [1:08:16<1:34:04, 4.97it/s] 42%|████▏ | 19940/48008 [1:08:16<1:32:53, 5.04it/s] 42%|████▏ | 19941/48008 [1:08:16<1:31:33, 5.11it/s] 42%|████▏ | 19942/48008 [1:08:16<1:31:30, 5.11it/s] 42%|████▏ | 19943/48008 [1:08:16<1:32:04, 5.08it/s] 42%|████▏ | 19944/48008 [1:08:17<1:31:27, 5.11it/s] 42%|████▏ | 19945/48008 [1:08:17<1:31:11, 5.13it/s] 42%|████▏ | 19946/48008 [1:08:17<1:29:59, 5.20it/s] 42%|████▏ | 19947/48008 [1:08:17<1:29:33, 5.22it/s] 42%|████▏ | 19948/48008 [1:08:17<1:28:56, 5.26it/s] 42%|████▏ | 19949/48008 [1:08:18<1:28:14, 5.30it/s] 42%|████▏ | 19950/48008 [1:08:18<1:28:00, 5.31it/s] {'loss': 0.3021, 'grad_norm': 0.65625, 'learning_rate': 0.00011689301783036161, 'epoch': 0.42} 42%|████▏ | 19950/48008 [1:08:18<1:28:00, 5.31it/s] 42%|████▏ | 19951/48008 [1:08:18<1:29:30, 5.22it/s] 42%|████▏ | 19952/48008 [1:08:18<1:29:47, 5.21it/s] 42%|████▏ | 19953/48008 [1:08:18<1:31:12, 5.13it/s] 42%|████▏ | 19954/48008 [1:08:19<1:32:25, 5.06it/s] 42%|████▏ | 19955/48008 [1:08:19<1:31:52, 5.09it/s] 42%|████▏ | 19956/48008 [1:08:19<1:32:45, 5.04it/s] 42%|████▏ | 19957/48008 [1:08:19<1:30:52, 5.14it/s] 42%|████▏ | 19958/48008 [1:08:19<1:30:37, 5.16it/s] 42%|████▏ | 19959/48008 [1:08:20<1:57:38, 3.97it/s] 42%|████▏ | 19960/48008 [1:08:20<1:49:15, 4.28it/s] 42%|████▏ | 19961/48008 [1:08:20<1:43:04, 4.54it/s] 42%|████▏ | 19962/48008 [1:08:20<1:38:30, 4.75it/s] 42%|████▏ | 19963/48008 [1:08:20<1:35:25, 4.90it/s] 42%|████▏ | 19964/48008 [1:08:21<1:35:27, 4.90it/s] 42%|████▏ | 19965/48008 [1:08:21<1:33:27, 5.00it/s] 42%|████▏ | 19966/48008 [1:08:21<1:32:12, 5.07it/s] 42%|████▏ | 19967/48008 [1:08:21<1:32:52, 5.03it/s] 42%|████▏ | 19968/48008 [1:08:21<1:31:53, 5.09it/s] 42%|████▏ | 19969/48008 [1:08:22<1:29:28, 5.22it/s] 42%|████▏ | 19970/48008 [1:08:22<1:30:59, 5.14it/s] 42%|████▏ | 19971/48008 [1:08:22<1:30:46, 5.15it/s] 42%|████▏ | 19972/48008 [1:08:22<1:30:21, 5.17it/s] 42%|████▏ | 19973/48008 [1:08:22<1:30:08, 5.18it/s] 42%|████▏ | 19974/48008 [1:08:23<1:30:42, 5.15it/s] 42%|████▏ | 19975/48008 [1:08:23<1:29:32, 5.22it/s] 42%|████▏ | 19976/48008 [1:08:23<1:29:22, 5.23it/s] 42%|████▏ | 19977/48008 [1:08:23<1:29:21, 5.23it/s] 42%|████▏ | 19978/48008 [1:08:23<1:31:38, 5.10it/s] 42%|████▏ | 19979/48008 [1:08:24<1:31:04, 5.13it/s] 42%|████▏ | 19980/48008 [1:08:24<1:32:04, 5.07it/s] 42%|████▏ | 19981/48008 [1:08:24<1:30:49, 5.14it/s] 42%|████▏ | 19982/48008 [1:08:24<1:32:07, 5.07it/s] 42%|████▏ | 19983/48008 [1:08:24<1:33:48, 4.98it/s] 42%|████▏ | 19984/48008 [1:08:25<1:32:35, 5.04it/s] 42%|████▏ | 19985/48008 [1:08:25<1:34:06, 4.96it/s] 42%|████▏ | 19986/48008 [1:08:25<1:32:44, 5.04it/s] 42%|████▏ | 19987/48008 [1:08:25<1:31:42, 5.09it/s] 42%|████▏ | 19988/48008 [1:08:25<1:31:58, 5.08it/s] 42%|████▏ | 19989/48008 [1:08:26<1:31:44, 5.09it/s] 42%|████▏ | 19990/48008 [1:08:26<1:30:02, 5.19it/s] 42%|████▏ | 19991/48008 [1:08:26<1:30:02, 5.19it/s] 42%|████▏ | 19992/48008 [1:08:26<1:29:03, 5.24it/s] 42%|████▏ | 19993/48008 [1:08:26<1:29:22, 5.22it/s] 42%|████▏ | 19994/48008 [1:08:27<1:56:53, 3.99it/s] 42%|████▏ | 19995/48008 [1:08:27<1:48:29, 4.30it/s] 42%|████▏ | 19996/48008 [1:08:27<1:44:21, 4.47it/s] 42%|████▏ | 19997/48008 [1:08:27<1:39:55, 4.67it/s] 42%|████▏ | 19998/48008 [1:08:27<1:36:30, 4.84it/s] 42%|████▏ | 19999/48008 [1:08:28<2:01:36, 3.84it/s] 42%|████▏ | 20000/48008 [1:08:28<1:58:26, 3.94it/s] {'loss': 0.2994, 'grad_norm': 0.66796875, 'learning_rate': 0.00011668471921346443, 'epoch': 0.42} 42%|████▏ | 20000/48008 [1:08:28<1:58:26, 3.94it/s] 42%|████▏ | 20001/48008 [1:08:28<1:49:37, 4.26it/s] 42%|████▏ | 20002/48008 [1:08:28<1:43:33, 4.51it/s] 42%|████▏ | 20003/48008 [1:08:29<1:39:32, 4.69it/s] 42%|████▏ | 20004/48008 [1:08:29<1:36:15, 4.85it/s] 42%|████▏ | 20005/48008 [1:08:29<1:34:48, 4.92it/s] 42%|████▏ | 20006/48008 [1:08:29<1:34:33, 4.94it/s] 42%|████▏ | 20007/48008 [1:08:29<1:32:50, 5.03it/s] 42%|████▏ | 20008/48008 [1:08:30<1:31:38, 5.09it/s] 42%|████▏ | 20009/48008 [1:08:30<1:31:10, 5.12it/s] 42%|████▏ | 20010/48008 [1:08:30<1:36:50, 4.82it/s] 42%|████▏ | 20011/48008 [1:08:30<1:34:13, 4.95it/s] 42%|████▏ | 20012/48008 [1:08:30<1:34:22, 4.94it/s] 42%|████▏ | 20013/48008 [1:08:31<1:33:35, 4.99it/s] 42%|████▏ | 20014/48008 [1:08:31<1:31:46, 5.08it/s] 42%|████▏ | 20015/48008 [1:08:31<1:32:01, 5.07it/s] 42%|████▏ | 20016/48008 [1:08:31<1:30:44, 5.14it/s] 42%|████▏ | 20017/48008 [1:08:31<1:30:24, 5.16it/s] 42%|████▏ | 20018/48008 [1:08:32<1:30:36, 5.15it/s] 42%|████▏ | 20019/48008 [1:08:32<1:30:24, 5.16it/s] 42%|████▏ | 20020/48008 [1:08:32<1:30:34, 5.15it/s] 42%|████▏ | 20021/48008 [1:08:32<1:29:58, 5.18it/s] 42%|████▏ | 20022/48008 [1:08:32<1:31:30, 5.10it/s] 42%|████▏ | 20023/48008 [1:08:33<1:30:34, 5.15it/s] 42%|████▏ | 20024/48008 [1:08:33<1:29:05, 5.24it/s] 42%|████▏ | 20025/48008 [1:08:33<1:28:57, 5.24it/s] 42%|████▏ | 20026/48008 [1:08:33<1:30:08, 5.17it/s] 42%|████▏ | 20027/48008 [1:08:33<1:31:46, 5.08it/s] 42%|████▏ | 20028/48008 [1:08:34<1:31:01, 5.12it/s] 42%|████▏ | 20029/48008 [1:08:34<1:30:56, 5.13it/s] 42%|████▏ | 20030/48008 [1:08:34<1:30:45, 5.14it/s] 42%|████▏ | 20031/48008 [1:08:34<1:46:24, 4.38it/s] 42%|████▏ | 20032/48008 [1:08:34<1:40:42, 4.63it/s] 42%|████▏ | 20033/48008 [1:08:35<1:37:10, 4.80it/s] 42%|████▏ | 20034/48008 [1:08:35<1:37:08, 4.80it/s] 42%|████▏ | 20035/48008 [1:08:35<1:35:04, 4.90it/s] 42%|████▏ | 20036/48008 [1:08:35<1:35:24, 4.89it/s] 42%|████▏ | 20037/48008 [1:08:35<1:33:35, 4.98it/s] 42%|████▏ | 20038/48008 [1:08:36<1:33:04, 5.01it/s] 42%|████▏ | 20039/48008 [1:08:36<1:31:08, 5.11it/s] 42%|████▏ | 20040/48008 [1:08:36<1:37:01, 4.80it/s] 42%|████▏ | 20041/48008 [1:08:36<1:35:55, 4.86it/s] 42%|████▏ | 20042/48008 [1:08:36<1:33:53, 4.96it/s] 42%|████▏ | 20043/48008 [1:08:37<1:32:13, 5.05it/s] 42%|████▏ | 20044/48008 [1:08:37<1:32:51, 5.02it/s] 42%|████▏ | 20045/48008 [1:08:37<1:31:40, 5.08it/s] 42%|████▏ | 20046/48008 [1:08:37<1:37:05, 4.80it/s] 42%|████▏ | 20047/48008 [1:08:37<1:34:14, 4.95it/s] 42%|████▏ | 20048/48008 [1:08:38<1:32:14, 5.05it/s] 42%|████▏ | 20049/48008 [1:08:38<1:32:58, 5.01it/s] 42%|████▏ | 20050/48008 [1:08:38<1:33:00, 5.01it/s] {'loss': 0.3017, 'grad_norm': 0.265625, 'learning_rate': 0.00011647642059656725, 'epoch': 0.42} 42%|████▏ | 20050/48008 [1:08:38<1:33:00, 5.01it/s] 42%|████▏ | 20051/48008 [1:08:38<1:31:56, 5.07it/s] 42%|████▏ | 20052/48008 [1:08:38<1:30:57, 5.12it/s] 42%|████▏ | 20053/48008 [1:08:39<1:30:07, 5.17it/s] 42%|████▏ | 20054/48008 [1:08:39<1:32:32, 5.03it/s] 42%|████▏ | 20055/48008 [1:08:39<1:32:35, 5.03it/s] 42%|████▏ | 20056/48008 [1:08:39<1:31:30, 5.09it/s] 42%|████▏ | 20057/48008 [1:08:39<1:32:35, 5.03it/s] 42%|████▏ | 20058/48008 [1:08:40<1:31:01, 5.12it/s] 42%|████▏ | 20059/48008 [1:08:40<1:30:50, 5.13it/s] 42%|████▏ | 20060/48008 [1:08:40<1:31:53, 5.07it/s] 42%|████▏ | 20061/48008 [1:08:40<1:31:09, 5.11it/s] 42%|████▏ | 20062/48008 [1:08:40<1:30:45, 5.13it/s] 42%|████▏ | 20063/48008 [1:08:41<1:29:42, 5.19it/s] 42%|████▏ | 20064/48008 [1:08:41<1:30:08, 5.17it/s] 42%|████▏ | 20065/48008 [1:08:41<1:29:58, 5.18it/s] 42%|████▏ | 20066/48008 [1:08:41<1:29:48, 5.19it/s] 42%|████▏ | 20067/48008 [1:08:41<1:29:32, 5.20it/s] 42%|████▏ | 20068/48008 [1:08:42<1:35:52, 4.86it/s] 42%|████▏ | 20069/48008 [1:08:42<1:33:29, 4.98it/s] 42%|████▏ | 20070/48008 [1:08:42<1:32:02, 5.06it/s] 42%|████▏ | 20071/48008 [1:08:42<1:32:54, 5.01it/s] 42%|████▏ | 20072/48008 [1:08:42<1:34:09, 4.95it/s] 42%|████▏ | 20073/48008 [1:08:43<1:32:26, 5.04it/s] 42%|████▏ | 20074/48008 [1:08:43<1:37:54, 4.76it/s] 42%|████▏ | 20075/48008 [1:08:43<1:35:13, 4.89it/s] 42%|████▏ | 20076/48008 [1:08:43<1:33:16, 4.99it/s] 42%|████▏ | 20077/48008 [1:08:43<1:38:34, 4.72it/s] 42%|████▏ | 20078/48008 [1:08:44<1:35:59, 4.85it/s] 42%|████▏ | 20079/48008 [1:08:44<1:33:54, 4.96it/s] 42%|████▏ | 20080/48008 [1:08:44<1:34:18, 4.94it/s] 42%|████▏ | 20081/48008 [1:08:44<1:31:40, 5.08it/s] 42%|████▏ | 20082/48008 [1:08:44<1:31:39, 5.08it/s] 42%|████▏ | 20083/48008 [1:08:45<1:30:35, 5.14it/s] 42%|████▏ | 20084/48008 [1:08:45<1:32:39, 5.02it/s] 42%|████▏ | 20085/48008 [1:08:46<3:20:26, 2.32it/s] 42%|████▏ | 20086/48008 [1:08:46<2:53:39, 2.68it/s] 42%|████▏ | 20087/48008 [1:08:46<2:26:41, 3.17it/s] 42%|████▏ | 20088/48008 [1:08:46<2:09:13, 3.60it/s] 42%|████▏ | 20089/48008 [1:08:47<1:56:44, 3.99it/s] 42%|████▏ | 20090/48008 [1:08:47<1:48:40, 4.28it/s] 42%|████▏ | 20091/48008 [1:08:48<3:20:48, 2.32it/s] 42%|████▏ | 20092/48008 [1:08:48<2:53:43, 2.68it/s] 42%|████▏ | 20093/48008 [1:08:48<2:28:04, 3.14it/s] 42%|████▏ | 20094/48008 [1:08:48<2:37:25, 2.96it/s] 42%|████▏ | 20095/48008 [1:08:49<2:16:06, 3.42it/s] 42%|████▏ | 20096/48008 [1:08:49<2:01:54, 3.82it/s] 42%|████▏ | 20097/48008 [1:08:50<3:18:33, 2.34it/s] 42%|████▏ | 20098/48008 [1:08:50<2:45:34, 2.81it/s] 42%|████▏ | 20099/48008 [1:08:50<2:49:31, 2.74it/s] 42%|████▏ | 20100/48008 [1:08:50<2:24:38, 3.22it/s] 42%|████▏ | 20100/48008 [1:08:50<2:24:38, 3.22it/s]{'loss': 0.2997, 'grad_norm': 0.54296875, 'learning_rate': 0.00011626812197967005, 'epoch': 0.42} 42%|████▏ | 20101/48008 [1:08:51<2:08:07, 3.63it/s] 42%|████▏ | 20102/48008 [1:08:51<1:57:50, 3.95it/s] 42%|████▏ | 20103/48008 [1:08:51<1:49:28, 4.25it/s] 42%|████▏ | 20104/48008 [1:08:51<1:43:13, 4.51it/s] 42%|████▏ | 20105/48008 [1:08:51<1:37:47, 4.76it/s] 42%|████▏ | 20106/48008 [1:08:52<1:33:52, 4.95it/s] 42%|████▏ | 20107/48008 [1:08:52<1:38:59, 4.70it/s] 42%|████▏ | 20108/48008 [1:08:52<1:35:37, 4.86it/s] 42%|████▏ | 20109/48008 [1:08:52<1:33:06, 4.99it/s] 42%|████▏ | 20110/48008 [1:08:52<1:31:18, 5.09it/s] 42%|████▏ | 20111/48008 [1:08:53<1:31:34, 5.08it/s] 42%|████▏ | 20112/48008 [1:08:53<1:30:15, 5.15it/s] 42%|████▏ | 20113/48008 [1:08:53<1:31:25, 5.09it/s] 42%|████▏ | 20114/48008 [1:08:53<1:30:33, 5.13it/s] 42%|████▏ | 20115/48008 [1:08:53<1:29:41, 5.18it/s] 42%|████▏ | 20116/48008 [1:08:53<1:28:49, 5.23it/s] 42%|████▏ | 20117/48008 [1:08:54<1:29:00, 5.22it/s] 42%|████▏ | 20118/48008 [1:08:54<1:31:20, 5.09it/s] 42%|████▏ | 20119/48008 [1:08:54<1:33:11, 4.99it/s] 42%|████▏ | 20120/48008 [1:08:54<1:32:40, 5.02it/s] 42%|████▏ | 20121/48008 [1:08:54<1:31:28, 5.08it/s] 42%|████▏ | 20122/48008 [1:08:55<1:30:40, 5.13it/s] 42%|████▏ | 20123/48008 [1:08:55<1:32:49, 5.01it/s] 42%|████▏ | 20124/48008 [1:08:55<1:31:25, 5.08it/s] 42%|████▏ | 20125/48008 [1:08:55<1:30:49, 5.12it/s] 42%|████▏ | 20126/48008 [1:08:55<1:30:01, 5.16it/s] 42%|████▏ | 20127/48008 [1:08:56<2:55:30, 2.65it/s] 42%|████▏ | 20128/48008 [1:08:56<2:31:01, 3.08it/s] 42%|████▏ | 20129/48008 [1:08:57<2:12:38, 3.50it/s] 42%|████▏ | 20130/48008 [1:08:57<1:58:49, 3.91it/s] 42%|████▏ | 20131/48008 [1:08:57<1:50:02, 4.22it/s] 42%|████▏ | 20132/48008 [1:08:57<1:43:19, 4.50it/s] 42%|████▏ | 20133/48008 [1:08:57<1:38:56, 4.70it/s] 42%|████▏ | 20134/48008 [1:08:58<1:37:59, 4.74it/s] 42%|████▏ | 20135/48008 [1:08:58<1:35:13, 4.88it/s] 42%|████▏ | 20136/48008 [1:08:58<1:33:30, 4.97it/s] 42%|████▏ | 20137/48008 [1:08:58<1:31:59, 5.05it/s] 42%|████▏ | 20138/48008 [1:08:58<1:32:04, 5.04it/s] 42%|████▏ | 20139/48008 [1:08:59<1:30:59, 5.11it/s] 42%|████▏ | 20140/48008 [1:08:59<1:29:22, 5.20it/s] 42%|████▏ | 20141/48008 [1:08:59<1:29:48, 5.17it/s] 42%|████▏ | 20142/48008 [1:08:59<1:29:39, 5.18it/s] 42%|████▏ | 20143/48008 [1:08:59<1:28:29, 5.25it/s] 42%|████▏ | 20144/48008 [1:09:00<1:28:02, 5.27it/s] 42%|████▏ | 20145/48008 [1:09:00<1:30:44, 5.12it/s] 42%|████▏ | 20146/48008 [1:09:00<1:30:18, 5.14it/s] 42%|████▏ | 20147/48008 [1:09:00<1:29:42, 5.18it/s] 42%|████▏ | 20148/48008 [1:09:00<1:35:44, 4.85it/s] 42%|████▏ | 20149/48008 [1:09:01<1:32:54, 5.00it/s] 42%|████▏ | 20150/48008 [1:09:01<1:31:57, 5.05it/s] {'loss': 0.3038, 'grad_norm': 0.13671875, 'learning_rate': 0.00011605982336277288, 'epoch': 0.42} 42%|████▏ | 20150/48008 [1:09:01<1:31:57, 5.05it/s] 42%|████▏ | 20151/48008 [1:09:01<1:31:22, 5.08it/s] 42%|████▏ | 20152/48008 [1:09:01<1:30:17, 5.14it/s] 42%|████▏ | 20153/48008 [1:09:01<1:29:58, 5.16it/s] 42%|████▏ | 20154/48008 [1:09:02<1:29:43, 5.17it/s] 42%|████▏ | 20155/48008 [1:09:02<1:29:21, 5.19it/s] 42%|████▏ | 20156/48008 [1:09:02<1:29:57, 5.16it/s] 42%|████▏ | 20157/48008 [1:09:02<1:30:54, 5.11it/s] 42%|████▏ | 20158/48008 [1:09:02<1:30:37, 5.12it/s] 42%|████▏ | 20159/48008 [1:09:02<1:30:59, 5.10it/s] 42%|████▏ | 20160/48008 [1:09:03<1:31:39, 5.06it/s] 42%|████▏ | 20161/48008 [1:09:03<1:30:35, 5.12it/s] 42%|████▏ | 20162/48008 [1:09:03<1:29:55, 5.16it/s] 42%|████▏ | 20163/48008 [1:09:03<1:31:58, 5.05it/s] 42%|████▏ | 20164/48008 [1:09:03<1:30:31, 5.13it/s] 42%|████▏ | 20165/48008 [1:09:04<1:29:55, 5.16it/s] 42%|████▏ | 20166/48008 [1:09:04<1:29:31, 5.18it/s] 42%|████▏ | 20167/48008 [1:09:04<1:28:23, 5.25it/s] 42%|████▏ | 20168/48008 [1:09:04<1:28:18, 5.25it/s] 42%|████▏ | 20169/48008 [1:09:04<1:28:15, 5.26it/s] 42%|████▏ | 20170/48008 [1:09:05<1:29:11, 5.20it/s] 42%|████▏ | 20171/48008 [1:09:05<1:30:31, 5.13it/s] 42%|████▏ | 20172/48008 [1:09:05<1:31:40, 5.06it/s] 42%|████▏ | 20173/48008 [1:09:05<1:31:27, 5.07it/s] 42%|████▏ | 20174/48008 [1:09:05<1:31:26, 5.07it/s] 42%|████▏ | 20175/48008 [1:09:06<1:30:44, 5.11it/s] 42%|████▏ | 20176/48008 [1:09:06<1:31:42, 5.06it/s] 42%|████▏ | 20177/48008 [1:09:06<1:46:47, 4.34it/s] 42%|████▏ | 20178/48008 [1:09:06<1:41:25, 4.57it/s] 42%|████▏ | 20179/48008 [1:09:07<1:39:08, 4.68it/s] 42%|████▏ | 20180/48008 [1:09:07<1:36:15, 4.82it/s] 42%|████▏ | 20181/48008 [1:09:07<1:35:30, 4.86it/s] 42%|████▏ | 20182/48008 [1:09:07<1:34:07, 4.93it/s] 42%|████▏ | 20183/48008 [1:09:07<1:32:20, 5.02it/s] 42%|████▏ | 20184/48008 [1:09:07<1:31:36, 5.06it/s] 42%|████▏ | 20185/48008 [1:09:08<1:31:13, 5.08it/s] 42%|████▏ | 20186/48008 [1:09:08<1:30:38, 5.12it/s] 42%|████▏ | 20187/48008 [1:09:08<1:29:49, 5.16it/s] 42%|████▏ | 20188/48008 [1:09:08<1:31:15, 5.08it/s] 42%|████▏ | 20189/48008 [1:09:09<1:36:54, 4.78it/s] 42%|████▏ | 20190/48008 [1:09:09<1:35:52, 4.84it/s] 42%|████▏ | 20191/48008 [1:09:09<1:33:58, 4.93it/s] 42%|████▏ | 20192/48008 [1:09:09<1:32:17, 5.02it/s] 42%|████▏ | 20193/48008 [1:09:09<1:31:17, 5.08it/s] 42%|████▏ | 20194/48008 [1:09:09<1:30:44, 5.11it/s] 42%|████▏ | 20195/48008 [1:09:10<1:36:23, 4.81it/s] 42%|████▏ | 20196/48008 [1:09:10<1:34:09, 4.92it/s] 42%|████▏ | 20197/48008 [1:09:10<1:32:31, 5.01it/s] 42%|████▏ | 20198/48008 [1:09:10<1:30:52, 5.10it/s] 42%|████▏ | 20199/48008 [1:09:10<1:31:33, 5.06it/s] 42%|████▏ | 20200/48008 [1:09:11<1:30:17, 5.13it/s] {'loss': 0.3011, 'grad_norm': 0.408203125, 'learning_rate': 0.00011585152474587568, 'epoch': 0.42} 42%|████▏ | 20200/48008 [1:09:11<1:30:17, 5.13it/s] 42%|████▏ | 20201/48008 [1:09:11<1:29:45, 5.16it/s] 42%|████▏ | 20202/48008 [1:09:11<1:29:21, 5.19it/s] 42%|████▏ | 20203/48008 [1:09:11<1:28:57, 5.21it/s] 42%|████▏ | 20204/48008 [1:09:11<1:28:30, 5.24it/s] 42%|████▏ | 20205/48008 [1:09:12<1:29:29, 5.18it/s] 42%|████▏ | 20206/48008 [1:09:12<1:30:44, 5.11it/s] 42%|████▏ | 20207/48008 [1:09:12<1:30:16, 5.13it/s] 42%|████▏ | 20208/48008 [1:09:12<1:30:29, 5.12it/s] 42%|████▏ | 20209/48008 [1:09:12<1:29:42, 5.16it/s] 42%|████▏ | 20210/48008 [1:09:13<1:29:10, 5.20it/s] 42%|████▏ | 20211/48008 [1:09:13<1:28:59, 5.21it/s] 42%|████▏ | 20212/48008 [1:09:13<1:28:38, 5.23it/s] 42%|████▏ | 20213/48008 [1:09:13<1:28:12, 5.25it/s] 42%|████▏ | 20214/48008 [1:09:13<1:28:13, 5.25it/s] 42%|████▏ | 20215/48008 [1:09:14<1:28:21, 5.24it/s] 42%|████▏ | 20216/48008 [1:09:14<1:27:53, 5.27it/s] 42%|████▏ | 20217/48008 [1:09:14<1:27:49, 5.27it/s] 42%|████▏ | 20218/48008 [1:09:14<1:27:50, 5.27it/s] 42%|████▏ | 20219/48008 [1:09:14<1:27:41, 5.28it/s] 42%|████▏ | 20220/48008 [1:09:15<1:29:33, 5.17it/s] 42%|████▏ | 20221/48008 [1:09:15<1:30:12, 5.13it/s] 42%|████▏ | 20222/48008 [1:09:15<1:29:53, 5.15it/s] 42%|████▏ | 20223/48008 [1:09:15<1:28:47, 5.22it/s] 42%|████▏ | 20224/48008 [1:09:15<1:29:38, 5.17it/s] 42%|████▏ | 20225/48008 [1:09:15<1:30:50, 5.10it/s] 42%|████▏ | 20226/48008 [1:09:16<1:57:04, 3.95it/s] 42%|████▏ | 20227/48008 [1:09:17<3:24:59, 2.26it/s] 42%|████▏ | 20228/48008 [1:09:17<2:49:59, 2.72it/s] 42%|████▏ | 20229/48008 [1:09:17<2:27:09, 3.15it/s] 42%|████▏ | 20230/48008 [1:09:17<2:11:08, 3.53it/s] 42%|████▏ | 20231/48008 [1:09:18<1:57:43, 3.93it/s] 42%|████▏ | 20232/48008 [1:09:18<1:49:36, 4.22it/s] 42%|████▏ | 20233/48008 [1:09:18<1:45:46, 4.38it/s] 42%|████▏ | 20234/48008 [1:09:18<1:39:52, 4.63it/s] 42%|████▏ | 20235/48008 [1:09:18<1:37:24, 4.75it/s] 42%|████▏ | 20236/48008 [1:09:19<1:34:52, 4.88it/s] 42%|████▏ | 20237/48008 [1:09:19<1:33:36, 4.94it/s] 42%|████▏ | 20238/48008 [1:09:19<1:33:35, 4.95it/s] 42%|████▏ | 20239/48008 [1:09:19<1:31:57, 5.03it/s] 42%|████▏ | 20240/48008 [1:09:19<1:30:55, 5.09it/s] 42%|████▏ | 20241/48008 [1:09:19<1:29:35, 5.17it/s] 42%|████▏ | 20242/48008 [1:09:20<1:30:03, 5.14it/s] 42%|████▏ | 20243/48008 [1:09:20<1:29:15, 5.18it/s] 42%|████▏ | 20244/48008 [1:09:20<1:32:06, 5.02it/s] 42%|████▏ | 20245/48008 [1:09:20<1:31:01, 5.08it/s] 42%|████▏ | 20246/48008 [1:09:20<1:29:44, 5.16it/s] 42%|████▏ | 20247/48008 [1:09:21<1:29:21, 5.18it/s] 42%|████▏ | 20248/48008 [1:09:21<1:28:22, 5.24it/s] 42%|████▏ | 20249/48008 [1:09:21<1:30:53, 5.09it/s] 42%|████▏ | 20250/48008 [1:09:21<1:30:50, 5.09it/s] {'loss': 0.3001, 'grad_norm': 0.205078125, 'learning_rate': 0.00011564322612897852, 'epoch': 0.42} 42%|████▏ | 20250/48008 [1:09:21<1:30:50, 5.09it/s] 42%|████▏ | 20251/48008 [1:09:22<1:57:18, 3.94it/s] 42%|████▏ | 20252/48008 [1:09:22<1:48:41, 4.26it/s] 42%|████▏ | 20253/48008 [1:09:22<1:43:14, 4.48it/s] 42%|████▏ | 20254/48008 [1:09:22<1:38:42, 4.69it/s] 42%|████▏ | 20255/48008 [1:09:22<1:36:15, 4.81it/s] 42%|████▏ | 20256/48008 [1:09:23<1:34:34, 4.89it/s] 42%|████▏ | 20257/48008 [1:09:23<1:31:52, 5.03it/s] 42%|████▏ | 20258/48008 [1:09:23<1:30:16, 5.12it/s] 42%|████▏ | 20259/48008 [1:09:23<1:30:10, 5.13it/s] 42%|████▏ | 20260/48008 [1:09:23<1:29:13, 5.18it/s] 42%|████▏ | 20261/48008 [1:09:24<1:28:57, 5.20it/s] 42%|████▏ | 20262/48008 [1:09:24<1:35:12, 4.86it/s] 42%|████▏ | 20263/48008 [1:09:24<1:32:03, 5.02it/s] 42%|████▏ | 20264/48008 [1:09:24<1:31:09, 5.07it/s] 42%|████▏ | 20265/48008 [1:09:24<1:29:43, 5.15it/s] 42%|████▏ | 20266/48008 [1:09:25<1:30:06, 5.13it/s] 42%|████▏ | 20267/48008 [1:09:25<1:30:14, 5.12it/s] 42%|████▏ | 20268/48008 [1:09:25<1:35:57, 4.82it/s] 42%|████▏ | 20269/48008 [1:09:25<1:35:12, 4.86it/s] 42%|████▏ | 20270/48008 [1:09:25<1:33:17, 4.96it/s] 42%|████▏ | 20271/48008 [1:09:26<1:31:53, 5.03it/s] 42%|████▏ | 20272/48008 [1:09:26<1:30:19, 5.12it/s] 42%|████▏ | 20273/48008 [1:09:26<1:29:45, 5.15it/s] 42%|████▏ | 20274/48008 [1:09:26<1:29:12, 5.18it/s] 42%|████▏ | 20275/48008 [1:09:26<1:29:01, 5.19it/s] 42%|████▏ | 20276/48008 [1:09:27<1:28:16, 5.24it/s] 42%|████▏ | 20277/48008 [1:09:27<1:28:31, 5.22it/s] 42%|████▏ | 20278/48008 [1:09:27<1:27:49, 5.26it/s] 42%|████▏ | 20279/48008 [1:09:27<1:28:02, 5.25it/s] 42%|████▏ | 20280/48008 [1:09:27<1:27:43, 5.27it/s] 42%|████▏ | 20281/48008 [1:09:27<1:28:07, 5.24it/s] 42%|████▏ | 20282/48008 [1:09:28<1:28:34, 5.22it/s] 42%|████▏ | 20283/48008 [1:09:28<1:28:30, 5.22it/s] 42%|████▏ | 20284/48008 [1:09:28<1:30:05, 5.13it/s] 42%|████▏ | 20285/48008 [1:09:28<1:29:51, 5.14it/s] 42%|████▏ | 20286/48008 [1:09:28<1:29:31, 5.16it/s] 42%|████▏ | 20287/48008 [1:09:29<1:28:48, 5.20it/s] 42%|████▏ | 20288/48008 [1:09:29<1:28:44, 5.21it/s] 42%|████▏ | 20289/48008 [1:09:29<1:29:16, 5.17it/s] 42%|████▏ | 20290/48008 [1:09:29<1:31:16, 5.06it/s] 42%|████▏ | 20291/48008 [1:09:29<1:30:28, 5.11it/s] 42%|████▏ | 20292/48008 [1:09:30<1:29:02, 5.19it/s] 42%|████▏ | 20293/48008 [1:09:30<1:30:09, 5.12it/s] 42%|████▏ | 20294/48008 [1:09:30<1:36:05, 4.81it/s] 42%|████▏ | 20295/48008 [1:09:30<1:33:33, 4.94it/s] 42%|████▏ | 20296/48008 [1:09:31<1:47:38, 4.29it/s] 42%|████▏ | 20297/48008 [1:09:31<1:42:38, 4.50it/s] 42%|████▏ | 20298/48008 [1:09:31<1:41:10, 4.56it/s] 42%|████▏ | 20299/48008 [1:09:31<1:37:43, 4.73it/s] 42%|████▏ | 20300/48008 [1:09:31<1:34:54, 4.87it/s] {'loss': 0.3014, 'grad_norm': 0.40625, 'learning_rate': 0.00011543492751208132, 'epoch': 0.42} 42%|████▏ | 20300/48008 [1:09:31<1:34:54, 4.87it/s] 42%|████▏ | 20301/48008 [1:09:32<2:00:02, 3.85it/s] 42%|████▏ | 20302/48008 [1:09:32<1:50:14, 4.19it/s] 42%|████▏ | 20303/48008 [1:09:32<1:43:43, 4.45it/s] 42%|████▏ | 20304/48008 [1:09:32<1:55:27, 4.00it/s] 42%|████▏ | 20305/48008 [1:09:33<1:48:12, 4.27it/s] 42%|████▏ | 20306/48008 [1:09:33<1:41:28, 4.55it/s] 42%|████▏ | 20307/48008 [1:09:33<1:38:15, 4.70it/s] 42%|████▏ | 20308/48008 [1:09:33<1:35:08, 4.85it/s] 42%|████▏ | 20309/48008 [1:09:33<1:34:20, 4.89it/s] 42%|████▏ | 20310/48008 [1:09:34<1:32:28, 4.99it/s] 42%|████▏ | 20311/48008 [1:09:34<1:33:11, 4.95it/s] 42%|████▏ | 20312/48008 [1:09:34<1:32:31, 4.99it/s] 42%|████▏ | 20313/48008 [1:09:34<1:32:50, 4.97it/s] 42%|████▏ | 20314/48008 [1:09:34<1:37:50, 4.72it/s] 42%|████▏ | 20315/48008 [1:09:35<1:34:37, 4.88it/s] 42%|████▏ | 20316/48008 [1:09:35<1:35:34, 4.83it/s] 42%|████▏ | 20317/48008 [1:09:35<1:33:09, 4.95it/s] 42%|████▏ | 20318/48008 [1:09:35<1:31:18, 5.05it/s] 42%|████▏ | 20319/48008 [1:09:35<1:30:16, 5.11it/s] 42%|████▏ | 20320/48008 [1:09:36<1:28:17, 5.23it/s] 42%|████▏ | 20321/48008 [1:09:36<1:28:30, 5.21it/s] 42%|████▏ | 20322/48008 [1:09:36<1:28:09, 5.23it/s] 42%|████▏ | 20323/48008 [1:09:36<1:27:26, 5.28it/s] 42%|████▏ | 20324/48008 [1:09:36<1:27:10, 5.29it/s] 42%|████▏ | 20325/48008 [1:09:37<1:28:14, 5.23it/s] 42%|████▏ | 20326/48008 [1:09:37<1:28:14, 5.23it/s] 42%|████▏ | 20327/48008 [1:09:37<1:27:41, 5.26it/s] 42%|████▏ | 20328/48008 [1:09:37<1:34:02, 4.91it/s] 42%|████▏ | 20329/48008 [1:09:37<1:34:39, 4.87it/s] 42%|████▏ | 20330/48008 [1:09:38<1:32:45, 4.97it/s] 42%|████▏ | 20331/48008 [1:09:38<1:31:25, 5.05it/s] 42%|████▏ | 20332/48008 [1:09:38<1:30:14, 5.11it/s] 42%|████▏ | 20333/48008 [1:09:38<1:30:40, 5.09it/s] 42%|████▏ | 20334/48008 [1:09:38<1:29:49, 5.13it/s] 42%|████▏ | 20335/48008 [1:09:38<1:28:59, 5.18it/s] 42%|████▏ | 20336/48008 [1:09:39<1:28:16, 5.22it/s] 42%|████▏ | 20337/48008 [1:09:39<1:27:56, 5.24it/s] 42%|████▏ | 20338/48008 [1:09:39<1:29:44, 5.14it/s] 42%|████▏ | 20339/48008 [1:09:39<1:28:52, 5.19it/s] 42%|████▏ | 20340/48008 [1:09:39<1:30:24, 5.10it/s] 42%|████▏ | 20341/48008 [1:09:40<1:29:03, 5.18it/s] 42%|████▏ | 20342/48008 [1:09:40<1:28:30, 5.21it/s] 42%|████▏ | 20343/48008 [1:09:40<1:28:31, 5.21it/s] 42%|████▏ | 20344/48008 [1:09:40<1:29:54, 5.13it/s] 42%|████▏ | 20345/48008 [1:09:40<1:29:57, 5.13it/s] 42%|████▏ | 20346/48008 [1:09:41<1:30:22, 5.10it/s] 42%|████▏ | 20347/48008 [1:09:41<1:29:50, 5.13it/s] 42%|████▏ | 20348/48008 [1:09:41<1:29:30, 5.15it/s] 42%|████▏ | 20349/48008 [1:09:41<1:29:51, 5.13it/s] 42%|████▏ | 20350/48008 [1:09:41<1:29:23, 5.16it/s] {'loss': 0.3028, 'grad_norm': 0.19140625, 'learning_rate': 0.00011522662889518414, 'epoch': 0.42} 42%|████▏ | 20350/48008 [1:09:41<1:29:23, 5.16it/s] 42%|████▏ | 20351/48008 [1:09:42<1:29:03, 5.18it/s] 42%|████▏ | 20352/48008 [1:09:42<1:28:16, 5.22it/s] 42%|████▏ | 20353/48008 [1:09:42<1:29:48, 5.13it/s] 42%|████▏ | 20354/48008 [1:09:42<1:29:02, 5.18it/s] 42%|████▏ | 20355/48008 [1:09:42<1:29:49, 5.13it/s] 42%|████▏ | 20356/48008 [1:09:43<1:29:00, 5.18it/s] 42%|████▏ | 20357/48008 [1:09:43<1:28:45, 5.19it/s] 42%|████▏ | 20358/48008 [1:09:43<1:28:05, 5.23it/s] 42%|████▏ | 20359/48008 [1:09:43<1:28:53, 5.18it/s] 42%|████▏ | 20360/48008 [1:09:43<1:28:31, 5.21it/s] 42%|████▏ | 20361/48008 [1:09:44<1:28:34, 5.20it/s] 42%|████▏ | 20362/48008 [1:09:44<1:30:16, 5.10it/s] 42%|████▏ | 20363/48008 [1:09:44<1:29:53, 5.13it/s] 42%|████▏ | 20364/48008 [1:09:44<1:29:21, 5.16it/s] 42%|████▏ | 20365/48008 [1:09:44<1:29:21, 5.16it/s] 42%|████▏ | 20366/48008 [1:09:44<1:29:02, 5.17it/s] 42%|████▏ | 20367/48008 [1:09:45<1:29:08, 5.17it/s] 42%|████▏ | 20368/48008 [1:09:45<1:30:18, 5.10it/s] 42%|████▏ | 20369/48008 [1:09:45<1:29:38, 5.14it/s] 42%|████▏ | 20370/48008 [1:09:45<1:27:59, 5.24it/s] 42%|████▏ | 20371/48008 [1:09:45<1:28:39, 5.20it/s] 42%|████▏ | 20372/48008 [1:09:46<1:27:39, 5.25it/s] 42%|████▏ | 20373/48008 [1:09:46<1:29:22, 5.15it/s] 42%|████▏ | 20374/48008 [1:09:46<1:30:01, 5.12it/s] 42%|████▏ | 20375/48008 [1:09:46<1:31:02, 5.06it/s] 42%|████▏ | 20376/48008 [1:09:46<1:31:00, 5.06it/s] 42%|████▏ | 20377/48008 [1:09:47<1:45:58, 4.35it/s] 42%|████▏ | 20378/48008 [1:09:47<1:42:20, 4.50it/s] 42%|████▏ | 20379/48008 [1:09:47<2:05:06, 3.68it/s] 42%|████▏ | 20380/48008 [1:09:48<1:54:01, 4.04it/s] 42%|████▏ | 20381/48008 [1:09:48<1:46:23, 4.33it/s] 42%|████▏ | 20382/48008 [1:09:48<1:41:27, 4.54it/s] 42%|████▏ | 20383/48008 [1:09:48<1:38:10, 4.69it/s] 42%|████▏ | 20384/48008 [1:09:48<1:36:13, 4.78it/s] 42%|████▏ | 20385/48008 [1:09:49<1:34:57, 4.85it/s] 42%|████▏ | 20386/48008 [1:09:49<1:32:20, 4.99it/s] 42%|████▏ | 20387/48008 [1:09:49<1:31:38, 5.02it/s] 42%|████▏ | 20388/48008 [1:09:49<1:30:44, 5.07it/s] 42%|████▏ | 20389/48008 [1:09:49<1:30:49, 5.07it/s] 42%|████▏ | 20390/48008 [1:09:49<1:29:59, 5.12it/s] 42%|████▏ | 20391/48008 [1:09:50<1:29:43, 5.13it/s] 42%|████▏ | 20392/48008 [1:09:50<1:29:17, 5.15it/s] 42%|████▏ | 20393/48008 [1:09:50<1:30:38, 5.08it/s] 42%|████▏ | 20394/48008 [1:09:50<1:30:55, 5.06it/s] 42%|████▏ | 20395/48008 [1:09:50<1:30:17, 5.10it/s] 42%|████▏ | 20396/48008 [1:09:51<1:31:17, 5.04it/s] 42%|████▏ | 20397/48008 [1:09:51<1:30:24, 5.09it/s] 42%|████▏ | 20398/48008 [1:09:51<1:30:52, 5.06it/s] 42%|████▏ | 20399/48008 [1:09:51<1:30:02, 5.11it/s] 42%|████▏ | 20400/48008 [1:09:51<1:29:34, 5.14it/s] {'loss': 0.3006, 'grad_norm': 0.1396484375, 'learning_rate': 0.00011501833027828697, 'epoch': 0.42} 42%|████▏ | 20400/48008 [1:09:51<1:29:34, 5.14it/s] 42%|████▏ | 20401/48008 [1:09:52<1:29:13, 5.16it/s] 42%|████▏ | 20402/48008 [1:09:52<1:28:32, 5.20it/s] 42%|████▏ | 20403/48008 [1:09:52<1:34:34, 4.86it/s] 43%|████▎ | 20404/48008 [1:09:52<1:33:26, 4.92it/s] 43%|████▎ | 20405/48008 [1:09:52<1:31:53, 5.01it/s] 43%|████▎ | 20406/48008 [1:09:53<1:29:57, 5.11it/s] 43%|████▎ | 20407/48008 [1:09:53<1:29:32, 5.14it/s] 43%|████▎ | 20408/48008 [1:09:53<1:29:50, 5.12it/s] 43%|████▎ | 20409/48008 [1:09:53<1:29:25, 5.14it/s] 43%|████▎ | 20410/48008 [1:09:53<1:28:38, 5.19it/s] 43%|████▎ | 20411/48008 [1:09:54<1:28:17, 5.21it/s] 43%|████▎ | 20412/48008 [1:09:54<1:28:15, 5.21it/s] 43%|████▎ | 20413/48008 [1:09:54<1:27:58, 5.23it/s] 43%|████▎ | 20414/48008 [1:09:54<1:27:58, 5.23it/s] 43%|████▎ | 20415/48008 [1:09:54<1:27:22, 5.26it/s] 43%|████▎ | 20416/48008 [1:09:55<1:27:15, 5.27it/s] 43%|████▎ | 20417/48008 [1:09:55<1:27:15, 5.27it/s] 43%|████▎ | 20418/48008 [1:09:55<1:30:00, 5.11it/s] 43%|████▎ | 20419/48008 [1:09:55<1:29:50, 5.12it/s] 43%|████▎ | 20420/48008 [1:09:55<1:28:38, 5.19it/s] 43%|████▎ | 20421/48008 [1:09:56<1:28:15, 5.21it/s] 43%|████▎ | 20422/48008 [1:09:56<1:28:35, 5.19it/s] 43%|████▎ | 20423/48008 [1:09:56<1:28:43, 5.18it/s] 43%|████▎ | 20424/48008 [1:09:56<1:30:20, 5.09it/s] 43%|████▎ | 20425/48008 [1:09:56<1:29:39, 5.13it/s] 43%|████▎ | 20426/48008 [1:09:56<1:28:37, 5.19it/s] 43%|████▎ | 20427/48008 [1:09:57<1:27:36, 5.25it/s] 43%|████▎ | 20428/48008 [1:09:57<1:27:58, 5.22it/s] 43%|████▎ | 20429/48008 [1:09:57<1:27:10, 5.27it/s] 43%|████▎ | 20430/48008 [1:09:57<1:27:03, 5.28it/s] 43%|████▎ | 20431/48008 [1:09:58<1:54:13, 4.02it/s] 43%|████▎ | 20432/48008 [1:09:58<1:45:45, 4.35it/s] 43%|████▎ | 20433/48008 [1:09:58<1:40:54, 4.55it/s] 43%|████▎ | 20434/48008 [1:09:58<1:37:57, 4.69it/s] 43%|████▎ | 20435/48008 [1:09:58<1:36:43, 4.75it/s] 43%|████▎ | 20436/48008 [1:09:59<1:34:15, 4.88it/s] 43%|████▎ | 20437/48008 [1:09:59<1:32:21, 4.98it/s] 43%|████▎ | 20438/48008 [1:09:59<1:31:12, 5.04it/s] 43%|████▎ | 20439/48008 [1:09:59<1:29:14, 5.15it/s] 43%|████▎ | 20440/48008 [1:09:59<1:28:32, 5.19it/s] 43%|████▎ | 20441/48008 [1:10:00<1:28:39, 5.18it/s] 43%|████▎ | 20442/48008 [1:10:00<1:28:03, 5.22it/s] 43%|████▎ | 20443/48008 [1:10:00<1:28:11, 5.21it/s] 43%|████▎ | 20444/48008 [1:10:00<1:28:03, 5.22it/s] 43%|████▎ | 20445/48008 [1:10:00<1:29:58, 5.11it/s] 43%|████▎ | 20446/48008 [1:10:01<1:29:35, 5.13it/s] 43%|████▎ | 20447/48008 [1:10:01<1:28:52, 5.17it/s] 43%|████▎ | 20448/48008 [1:10:01<1:28:50, 5.17it/s] 43%|████▎ | 20449/48008 [1:10:01<1:34:55, 4.84it/s] 43%|████▎ | 20450/48008 [1:10:02<1:59:43, 3.84it/s] 43%|████▎ | 20450/48008 [1:10:02<1:59:43, 3.84it/s]{'loss': 0.3029, 'grad_norm': 0.4296875, 'learning_rate': 0.00011481003166138977, 'epoch': 0.43} 43%|████▎ | 20451/48008 [1:10:02<1:50:55, 4.14it/s] 43%|████▎ | 20452/48008 [1:10:02<1:44:23, 4.40it/s] 43%|████▎ | 20453/48008 [1:10:02<1:38:47, 4.65it/s] 43%|████▎ | 20454/48008 [1:10:02<1:35:15, 4.82it/s] 43%|████▎ | 20455/48008 [1:10:02<1:33:23, 4.92it/s] 43%|████▎ | 20456/48008 [1:10:03<1:38:07, 4.68it/s] 43%|████▎ | 20457/48008 [1:10:03<1:50:54, 4.14it/s] 43%|████▎ | 20458/48008 [1:10:03<1:46:07, 4.33it/s] 43%|████▎ | 20459/48008 [1:10:03<1:40:27, 4.57it/s] 43%|████▎ | 20460/48008 [1:10:04<1:36:50, 4.74it/s] 43%|████▎ | 20461/48008 [1:10:04<1:35:05, 4.83it/s] 43%|████▎ | 20462/48008 [1:10:04<1:33:50, 4.89it/s] 43%|████▎ | 20463/48008 [1:10:04<1:32:08, 4.98it/s] 43%|████▎ | 20464/48008 [1:10:04<1:29:23, 5.14it/s] 43%|████▎ | 20465/48008 [1:10:05<1:29:10, 5.15it/s] 43%|████▎ | 20466/48008 [1:10:05<1:29:50, 5.11it/s] 43%|████▎ | 20467/48008 [1:10:05<1:32:04, 4.99it/s] 43%|████▎ | 20468/48008 [1:10:05<1:32:02, 4.99it/s] 43%|████▎ | 20469/48008 [1:10:05<1:29:59, 5.10it/s] 43%|████▎ | 20470/48008 [1:10:06<1:30:00, 5.10it/s] 43%|████▎ | 20471/48008 [1:10:06<1:28:54, 5.16it/s] 43%|████▎ | 20472/48008 [1:10:06<1:28:35, 5.18it/s] 43%|████▎ | 20473/48008 [1:10:06<1:30:01, 5.10it/s] 43%|████▎ | 20474/48008 [1:10:06<1:28:30, 5.18it/s] 43%|████▎ | 20475/48008 [1:10:07<1:30:45, 5.06it/s] 43%|████▎ | 20476/48008 [1:10:07<1:29:43, 5.11it/s] 43%|████▎ | 20477/48008 [1:10:07<1:28:55, 5.16it/s] 43%|████▎ | 20478/48008 [1:10:07<1:28:48, 5.17it/s] 43%|████▎ | 20479/48008 [1:10:07<1:30:05, 5.09it/s] 43%|████▎ | 20480/48008 [1:10:08<1:29:17, 5.14it/s] 43%|████▎ | 20481/48008 [1:10:08<1:28:41, 5.17it/s] 43%|████▎ | 20482/48008 [1:10:08<1:28:08, 5.20it/s] 43%|████▎ | 20483/48008 [1:10:08<1:27:51, 5.22it/s] 43%|████▎ | 20484/48008 [1:10:08<1:34:12, 4.87it/s] 43%|████▎ | 20485/48008 [1:10:09<1:33:13, 4.92it/s] 43%|████▎ | 20486/48008 [1:10:09<1:31:05, 5.04it/s] 43%|████▎ | 20487/48008 [1:10:09<1:36:29, 4.75it/s] 43%|████▎ | 20488/48008 [1:10:09<1:35:37, 4.80it/s] 43%|████▎ | 20489/48008 [1:10:09<1:33:21, 4.91it/s] 43%|████▎ | 20490/48008 [1:10:10<1:31:43, 5.00it/s] 43%|████▎ | 20491/48008 [1:10:10<1:31:54, 4.99it/s] 43%|████▎ | 20492/48008 [1:10:10<1:37:07, 4.72it/s] 43%|████▎ | 20493/48008 [1:10:10<1:35:58, 4.78it/s] 43%|████▎ | 20494/48008 [1:10:10<1:33:37, 4.90it/s] 43%|████▎ | 20495/48008 [1:10:11<1:34:37, 4.85it/s] 43%|████▎ | 20496/48008 [1:10:11<1:32:50, 4.94it/s] 43%|████▎ | 20497/48008 [1:10:11<1:32:24, 4.96it/s] 43%|████▎ | 20498/48008 [1:10:11<1:31:10, 5.03it/s] 43%|████▎ | 20499/48008 [1:10:11<1:46:01, 4.32it/s] 43%|████▎ | 20500/48008 [1:10:12<1:41:45, 4.51it/s] {'loss': 0.2998, 'grad_norm': 0.1943359375, 'learning_rate': 0.0001146017330444926, 'epoch': 0.43} 43%|████▎ | 20500/48008 [1:10:12<1:41:45, 4.51it/s] 43%|████▎ | 20501/48008 [1:10:12<1:37:15, 4.71it/s] 43%|████▎ | 20502/48008 [1:10:12<1:34:12, 4.87it/s] 43%|████▎ | 20503/48008 [1:10:12<1:32:20, 4.96it/s] 43%|████▎ | 20504/48008 [1:10:12<1:31:13, 5.03it/s] 43%|████▎ | 20505/48008 [1:10:13<1:30:11, 5.08it/s] 43%|████▎ | 20506/48008 [1:10:13<1:29:31, 5.12it/s] 43%|████▎ | 20507/48008 [1:10:13<1:29:19, 5.13it/s] 43%|████▎ | 20508/48008 [1:10:13<1:35:12, 4.81it/s] 43%|████▎ | 20509/48008 [1:10:13<1:32:28, 4.96it/s] 43%|████▎ | 20510/48008 [1:10:14<1:31:11, 5.03it/s] 43%|████▎ | 20511/48008 [1:10:14<1:31:26, 5.01it/s] 43%|████▎ | 20512/48008 [1:10:14<1:30:27, 5.07it/s] 43%|████▎ | 20513/48008 [1:10:14<1:30:59, 5.04it/s] 43%|████▎ | 20514/48008 [1:10:14<1:29:59, 5.09it/s] 43%|████▎ | 20515/48008 [1:10:15<1:29:25, 5.12it/s] 43%|████▎ | 20516/48008 [1:10:15<1:28:49, 5.16it/s] 43%|████▎ | 20517/48008 [1:10:15<1:27:57, 5.21it/s] 43%|████▎ | 20518/48008 [1:10:15<1:28:57, 5.15it/s] 43%|████▎ | 20519/48008 [1:10:15<1:28:34, 5.17it/s] 43%|████▎ | 20520/48008 [1:10:16<1:28:14, 5.19it/s] 43%|████▎ | 20521/48008 [1:10:16<1:27:39, 5.23it/s] 43%|████▎ | 20522/48008 [1:10:16<1:27:17, 5.25it/s] 43%|████▎ | 20523/48008 [1:10:16<1:28:07, 5.20it/s] 43%|████▎ | 20524/48008 [1:10:16<1:30:24, 5.07it/s] 43%|████▎ | 20525/48008 [1:10:17<1:28:50, 5.16it/s] 43%|████▎ | 20526/48008 [1:10:17<1:28:28, 5.18it/s] 43%|████▎ | 20527/48008 [1:10:17<1:29:25, 5.12it/s] 43%|████▎ | 20528/48008 [1:10:17<1:28:41, 5.16it/s] 43%|████▎ | 20529/48008 [1:10:18<3:41:33, 2.07it/s] 43%|████▎ | 20530/48008 [1:10:18<3:00:47, 2.53it/s] 43%|████▎ | 20531/48008 [1:10:19<2:32:38, 3.00it/s] 43%|████▎ | 20532/48008 [1:10:19<2:12:47, 3.45it/s] 43%|████▎ | 20533/48008 [1:10:19<1:59:23, 3.84it/s] 43%|████▎ | 20534/48008 [1:10:19<1:49:39, 4.18it/s] 43%|████▎ | 20535/48008 [1:10:19<1:49:23, 4.19it/s] 43%|████▎ | 20536/48008 [1:10:20<1:42:38, 4.46it/s] 43%|████▎ | 20537/48008 [1:10:20<1:39:54, 4.58it/s] 43%|████▎ | 20538/48008 [1:10:20<1:35:20, 4.80it/s] 43%|████▎ | 20539/48008 [1:10:20<1:33:57, 4.87it/s] 43%|████▎ | 20540/48008 [1:10:20<1:31:29, 5.00it/s] 43%|████▎ | 20541/48008 [1:10:21<1:30:25, 5.06it/s] 43%|████▎ | 20542/48008 [1:10:21<1:28:54, 5.15it/s] 43%|████▎ | 20543/48008 [1:10:21<1:29:57, 5.09it/s] 43%|████▎ | 20544/48008 [1:10:21<1:28:57, 5.15it/s] 43%|████▎ | 20545/48008 [1:10:21<1:28:33, 5.17it/s] 43%|████▎ | 20546/48008 [1:10:22<1:29:40, 5.10it/s] 43%|████▎ | 20547/48008 [1:10:22<1:28:45, 5.16it/s] 43%|████▎ | 20548/48008 [1:10:22<1:28:18, 5.18it/s] 43%|████▎ | 20549/48008 [1:10:22<1:28:10, 5.19it/s] 43%|████▎ | 20550/48008 [1:10:22<1:28:51, 5.15it/s] {'loss': 0.3048, 'grad_norm': 0.1484375, 'learning_rate': 0.0001143934344275954, 'epoch': 0.43} 43%|████▎ | 20550/48008 [1:10:22<1:28:51, 5.15it/s] 43%|████▎ | 20551/48008 [1:10:23<1:29:35, 5.11it/s] 43%|████▎ | 20552/48008 [1:10:23<1:28:47, 5.15it/s] 43%|████▎ | 20553/48008 [1:10:23<1:28:26, 5.17it/s] 43%|████▎ | 20554/48008 [1:10:23<1:27:52, 5.21it/s] 43%|████▎ | 20555/48008 [1:10:23<1:29:08, 5.13it/s] 43%|████▎ | 20556/48008 [1:10:24<1:28:16, 5.18it/s] 43%|████▎ | 20557/48008 [1:10:24<1:28:02, 5.20it/s] 43%|████▎ | 20558/48008 [1:10:24<1:28:06, 5.19it/s] 43%|████▎ | 20559/48008 [1:10:24<1:28:43, 5.16it/s] 43%|████▎ | 20560/48008 [1:10:24<1:28:41, 5.16it/s] 43%|████▎ | 20561/48008 [1:10:24<1:28:32, 5.17it/s] 43%|████▎ | 20562/48008 [1:10:25<1:29:06, 5.13it/s] 43%|████▎ | 20563/48008 [1:10:25<1:34:50, 4.82it/s] 43%|████▎ | 20564/48008 [1:10:25<1:33:28, 4.89it/s] 43%|████▎ | 20565/48008 [1:10:25<1:31:50, 4.98it/s] 43%|████▎ | 20566/48008 [1:10:26<1:29:11, 5.13it/s] 43%|████▎ | 20567/48008 [1:10:26<1:30:18, 5.06it/s] 43%|████▎ | 20568/48008 [1:10:26<1:29:21, 5.12it/s] 43%|████▎ | 20569/48008 [1:10:26<1:44:30, 4.38it/s] 43%|████▎ | 20570/48008 [1:10:26<1:39:16, 4.61it/s] 43%|████▎ | 20571/48008 [1:10:27<1:37:41, 4.68it/s] 43%|████▎ | 20572/48008 [1:10:27<1:36:06, 4.76it/s] 43%|████▎ | 20573/48008 [1:10:27<1:33:22, 4.90it/s] 43%|████▎ | 20574/48008 [1:10:27<1:31:46, 4.98it/s] 43%|████▎ | 20575/48008 [1:10:27<1:30:19, 5.06it/s] 43%|████▎ | 20576/48008 [1:10:28<1:29:08, 5.13it/s] 43%|████▎ | 20577/48008 [1:10:28<1:27:56, 5.20it/s] 43%|████▎ | 20578/48008 [1:10:28<1:27:16, 5.24it/s] 43%|████▎ | 20579/48008 [1:10:28<1:30:02, 5.08it/s] 43%|████▎ | 20580/48008 [1:10:28<1:31:06, 5.02it/s] 43%|████▎ | 20581/48008 [1:10:29<1:29:28, 5.11it/s] 43%|████▎ | 20582/48008 [1:10:29<1:28:44, 5.15it/s] 43%|████▎ | 20583/48008 [1:10:29<1:27:40, 5.21it/s] 43%|████▎ | 20584/48008 [1:10:29<1:26:36, 5.28it/s] 43%|████▎ | 20585/48008 [1:10:29<1:26:22, 5.29it/s] 43%|████▎ | 20586/48008 [1:10:29<1:26:44, 5.27it/s] 43%|████▎ | 20587/48008 [1:10:30<1:27:01, 5.25it/s] 43%|████▎ | 20588/48008 [1:10:30<1:27:12, 5.24it/s] 43%|████▎ | 20589/48008 [1:10:30<1:26:15, 5.30it/s] 43%|████▎ | 20590/48008 [1:10:30<1:28:02, 5.19it/s] 43%|████▎ | 20591/48008 [1:10:30<1:27:47, 5.21it/s] 43%|████▎ | 20592/48008 [1:10:31<1:27:52, 5.20it/s] 43%|████▎ | 20593/48008 [1:10:31<1:28:09, 5.18it/s] 43%|████▎ | 20594/48008 [1:10:31<1:34:17, 4.85it/s] 43%|████▎ | 20595/48008 [1:10:31<1:34:14, 4.85it/s] 43%|████▎ | 20596/48008 [1:10:31<1:32:58, 4.91it/s] 43%|████▎ | 20597/48008 [1:10:32<1:31:02, 5.02it/s] 43%|████▎ | 20598/48008 [1:10:32<1:29:45, 5.09it/s] 43%|████▎ | 20599/48008 [1:10:32<1:30:40, 5.04it/s] 43%|████▎ | 20600/48008 [1:10:32<1:29:02, 5.13it/s] {'loss': 0.3015, 'grad_norm': 0.765625, 'learning_rate': 0.00011418513581069822, 'epoch': 0.43} 43%|████▎ | 20600/48008 [1:10:32<1:29:02, 5.13it/s] 43%|████▎ | 20601/48008 [1:10:32<1:30:21, 5.05it/s] 43%|████▎ | 20602/48008 [1:10:33<1:29:04, 5.13it/s] 43%|████▎ | 20603/48008 [1:10:33<1:34:50, 4.82it/s] 43%|████▎ | 20604/48008 [1:10:33<1:32:51, 4.92it/s] 43%|████▎ | 20605/48008 [1:10:33<1:31:00, 5.02it/s] 43%|████▎ | 20606/48008 [1:10:33<1:30:00, 5.07it/s] 43%|████▎ | 20607/48008 [1:10:34<1:29:14, 5.12it/s] 43%|████▎ | 20608/48008 [1:10:34<1:28:27, 5.16it/s] 43%|████▎ | 20609/48008 [1:10:34<1:27:59, 5.19it/s] 43%|████▎ | 20610/48008 [1:10:34<1:28:45, 5.14it/s] 43%|████▎ | 20611/48008 [1:10:34<1:29:04, 5.13it/s] 43%|████▎ | 20612/48008 [1:10:35<1:28:12, 5.18it/s] 43%|████▎ | 20613/48008 [1:10:35<1:30:37, 5.04it/s] 43%|████▎ | 20614/48008 [1:10:35<1:28:57, 5.13it/s] 43%|████▎ | 20615/48008 [1:10:35<1:28:38, 5.15it/s] 43%|████▎ | 20616/48008 [1:10:35<1:27:57, 5.19it/s] 43%|████▎ | 20617/48008 [1:10:36<1:26:25, 5.28it/s] 43%|████▎ | 20618/48008 [1:10:36<1:26:54, 5.25it/s] 43%|████▎ | 20619/48008 [1:10:36<1:27:03, 5.24it/s] 43%|████▎ | 20620/48008 [1:10:36<1:27:16, 5.23it/s] 43%|████▎ | 20621/48008 [1:10:36<1:27:14, 5.23it/s] 43%|████▎ | 20622/48008 [1:10:37<1:27:25, 5.22it/s] 43%|████▎ | 20623/48008 [1:10:37<1:29:10, 5.12it/s] 43%|████▎ | 20624/48008 [1:10:37<1:29:56, 5.07it/s] 43%|████▎ | 20625/48008 [1:10:37<1:29:11, 5.12it/s] 43%|████▎ | 20626/48008 [1:10:37<1:28:32, 5.15it/s] 43%|████▎ | 20627/48008 [1:10:38<1:28:52, 5.13it/s] 43%|████▎ | 20628/48008 [1:10:38<1:28:43, 5.14it/s] 43%|████▎ | 20629/48008 [1:10:38<1:27:39, 5.21it/s] 43%|████▎ | 20630/48008 [1:10:38<1:27:28, 5.22it/s] 43%|████▎ | 20631/48008 [1:10:38<1:27:52, 5.19it/s] 43%|████▎ | 20632/48008 [1:10:38<1:28:23, 5.16it/s] 43%|████▎ | 20633/48008 [1:10:39<1:27:51, 5.19it/s] 43%|████▎ | 20634/48008 [1:10:39<1:27:16, 5.23it/s] 43%|████▎ | 20635/48008 [1:10:39<1:27:24, 5.22it/s] 43%|████▎ | 20636/48008 [1:10:39<1:26:46, 5.26it/s] 43%|████▎ | 20637/48008 [1:10:39<1:27:37, 5.21it/s] 43%|████▎ | 20638/48008 [1:10:40<1:27:55, 5.19it/s] 43%|████▎ | 20639/48008 [1:10:40<1:27:56, 5.19it/s] 43%|████▎ | 20640/48008 [1:10:40<1:26:54, 5.25it/s] 43%|████▎ | 20641/48008 [1:10:40<1:27:35, 5.21it/s] 43%|████▎ | 20642/48008 [1:10:40<1:30:16, 5.05it/s] 43%|████▎ | 20643/48008 [1:10:41<1:31:14, 5.00it/s] 43%|████▎ | 20644/48008 [1:10:41<1:30:19, 5.05it/s] 43%|████▎ | 20645/48008 [1:10:41<1:30:19, 5.05it/s] 43%|████▎ | 20646/48008 [1:10:41<1:28:53, 5.13it/s] 43%|████▎ | 20647/48008 [1:10:41<1:30:15, 5.05it/s] 43%|████▎ | 20648/48008 [1:10:42<1:28:32, 5.15it/s] 43%|████▎ | 20649/48008 [1:10:42<1:28:11, 5.17it/s] 43%|████▎ | 20650/48008 [1:10:42<1:27:22, 5.22it/s] 43%|████▎ | 20650/48008 [1:10:42<1:27:22, 5.22it/s]{'loss': 0.3034, 'grad_norm': 0.158203125, 'learning_rate': 0.00011397683719380103, 'epoch': 0.43} 43%|████▎ | 20651/48008 [1:10:42<1:27:09, 5.23it/s] 43%|████▎ | 20652/48008 [1:10:42<1:28:02, 5.18it/s] 43%|████▎ | 20653/48008 [1:10:43<1:43:39, 4.40it/s] 43%|████▎ | 20654/48008 [1:10:43<1:45:14, 4.33it/s] 43%|████▎ | 20655/48008 [1:10:43<1:42:18, 4.46it/s] 43%|████▎ | 20656/48008 [1:10:43<1:38:28, 4.63it/s] 43%|████▎ | 20657/48008 [1:10:43<1:36:23, 4.73it/s] 43%|████▎ | 20658/48008 [1:10:44<1:33:34, 4.87it/s] 43%|████▎ | 20659/48008 [1:10:44<1:31:41, 4.97it/s] 43%|████▎ | 20660/48008 [1:10:44<1:31:51, 4.96it/s] 43%|████▎ | 20661/48008 [1:10:44<1:30:37, 5.03it/s] 43%|████▎ | 20662/48008 [1:10:44<1:29:44, 5.08it/s] 43%|████▎ | 20663/48008 [1:10:45<1:29:19, 5.10it/s] 43%|████▎ | 20664/48008 [1:10:45<1:28:41, 5.14it/s] 43%|████▎ | 20665/48008 [1:10:45<1:27:34, 5.20it/s] 43%|████▎ | 20666/48008 [1:10:45<1:27:23, 5.21it/s] 43%|████▎ | 20667/48008 [1:10:45<1:33:39, 4.87it/s] 43%|████▎ | 20668/48008 [1:10:46<1:31:15, 4.99it/s] 43%|████▎ | 20669/48008 [1:10:46<1:31:02, 5.01it/s] 43%|████▎ | 20670/48008 [1:10:46<1:29:14, 5.11it/s] 43%|████▎ | 20671/48008 [1:10:46<1:34:54, 4.80it/s] 43%|████▎ | 20672/48008 [1:10:46<1:32:08, 4.94it/s] 43%|████▎ | 20673/48008 [1:10:47<1:30:03, 5.06it/s] 43%|████▎ | 20674/48008 [1:10:47<1:29:20, 5.10it/s] 43%|████▎ | 20675/48008 [1:10:47<1:28:37, 5.14it/s] 43%|████▎ | 20676/48008 [1:10:47<1:28:20, 5.16it/s] 43%|████▎ | 20677/48008 [1:10:47<1:27:49, 5.19it/s] 43%|████▎ | 20678/48008 [1:10:48<1:28:00, 5.18it/s] 43%|████▎ | 20679/48008 [1:10:48<1:27:48, 5.19it/s] 43%|████▎ | 20680/48008 [1:10:48<1:28:29, 5.15it/s] 43%|████▎ | 20681/48008 [1:10:48<1:27:20, 5.21it/s] 43%|████▎ | 20682/48008 [1:10:48<1:27:44, 5.19it/s] 43%|████▎ | 20683/48008 [1:10:49<1:27:45, 5.19it/s] 43%|████▎ | 20684/48008 [1:10:49<1:27:55, 5.18it/s] 43%|████▎ | 20685/48008 [1:10:49<1:28:15, 5.16it/s] 43%|████▎ | 20686/48008 [1:10:49<1:27:43, 5.19it/s] 43%|████▎ | 20687/48008 [1:10:49<1:27:55, 5.18it/s] 43%|████▎ | 20688/48008 [1:10:50<1:27:24, 5.21it/s] 43%|████▎ | 20689/48008 [1:10:50<1:27:05, 5.23it/s] 43%|████▎ | 20690/48008 [1:10:50<1:27:04, 5.23it/s] 43%|████▎ | 20691/48008 [1:10:50<1:27:03, 5.23it/s] 43%|████▎ | 20692/48008 [1:10:50<1:26:19, 5.27it/s] 43%|████▎ | 20693/48008 [1:10:51<1:32:45, 4.91it/s] 43%|████▎ | 20694/48008 [1:10:51<1:32:03, 4.95it/s] 43%|████▎ | 20695/48008 [1:10:51<1:30:01, 5.06it/s] 43%|████▎ | 20696/48008 [1:10:51<1:30:02, 5.06it/s] 43%|████▎ | 20697/48008 [1:10:51<1:30:44, 5.02it/s] 43%|████▎ | 20698/48008 [1:10:52<3:28:28, 2.18it/s] 43%|████▎ | 20699/48008 [1:10:53<2:51:53, 2.65it/s] 43%|████▎ | 20700/48008 [1:10:53<2:27:11, 3.09it/s] 43%|████▎ | 20700/48008 [1:10:53<2:27:11, 3.09it/s]{'loss': 0.3005, 'grad_norm': 0.40234375, 'learning_rate': 0.00011376853857690386, 'epoch': 0.43} 43%|████▎ | 20701/48008 [1:10:53<2:08:54, 3.53it/s] 43%|████▎ | 20702/48008 [1:10:53<1:56:54, 3.89it/s] 43%|████▎ | 20703/48008 [1:10:53<1:47:17, 4.24it/s] 43%|████▎ | 20704/48008 [1:10:54<3:14:32, 2.34it/s] 43%|████▎ | 20705/48008 [1:10:54<2:42:11, 2.81it/s] 43%|████▎ | 20706/48008 [1:10:55<2:21:08, 3.22it/s] 43%|████▎ | 20707/48008 [1:10:55<2:06:21, 3.60it/s] 43%|████▎ | 20708/48008 [1:10:55<1:54:37, 3.97it/s] 43%|████▎ | 20709/48008 [1:10:55<1:47:57, 4.21it/s] 43%|████▎ | 20710/48008 [1:10:56<3:09:00, 2.41it/s] 43%|████▎ | 20711/48008 [1:10:56<2:38:27, 2.87it/s] 43%|████▎ | 20712/48008 [1:10:56<2:16:52, 3.32it/s] 43%|████▎ | 20713/48008 [1:10:57<2:17:47, 3.30it/s] 43%|████▎ | 20714/48008 [1:10:57<2:04:22, 3.66it/s] 43%|████▎ | 20715/48008 [1:10:57<1:53:48, 4.00it/s] 43%|████▎ | 20716/48008 [1:10:57<2:01:10, 3.75it/s] 43%|████▎ | 20717/48008 [1:10:58<1:50:22, 4.12it/s] 43%|████▎ | 20718/48008 [1:10:58<1:42:57, 4.42it/s] 43%|████▎ | 20719/48008 [1:10:58<1:38:00, 4.64it/s] 43%|████▎ | 20720/48008 [1:10:58<1:35:19, 4.77it/s] 43%|████▎ | 20721/48008 [1:10:58<1:32:38, 4.91it/s] 43%|████▎ | 20722/48008 [1:10:59<1:33:21, 4.87it/s] 43%|████▎ | 20723/48008 [1:10:59<1:29:55, 5.06it/s] 43%|████▎ | 20724/48008 [1:10:59<1:28:45, 5.12it/s] 43%|████▎ | 20725/48008 [1:10:59<1:55:08, 3.95it/s] 43%|████▎ | 20726/48008 [1:11:00<2:02:18, 3.72it/s] 43%|████▎ | 20727/48008 [1:11:00<1:51:04, 4.09it/s] 43%|████▎ | 20728/48008 [1:11:00<1:43:49, 4.38it/s] 43%|████▎ | 20729/48008 [1:11:00<1:38:39, 4.61it/s] 43%|████▎ | 20730/48008 [1:11:00<1:34:26, 4.81it/s] 43%|████▎ | 20731/48008 [1:11:01<1:58:55, 3.82it/s] 43%|████▎ | 20732/48008 [1:11:01<1:51:16, 4.09it/s] 43%|████▎ | 20733/48008 [1:11:01<1:43:57, 4.37it/s] 43%|████▎ | 20734/48008 [1:11:01<1:39:27, 4.57it/s] 43%|████▎ | 20735/48008 [1:11:02<1:35:27, 4.76it/s] 43%|████▎ | 20736/48008 [1:11:02<1:33:36, 4.86it/s] 43%|████▎ | 20737/48008 [1:11:02<1:37:40, 4.65it/s] 43%|████▎ | 20738/48008 [1:11:02<1:36:59, 4.69it/s] 43%|████▎ | 20739/48008 [1:11:02<1:33:34, 4.86it/s] 43%|████▎ | 20740/48008 [1:11:03<2:55:44, 2.59it/s] 43%|████▎ | 20741/48008 [1:11:03<2:29:41, 3.04it/s] 43%|████▎ | 20742/48008 [1:11:04<2:13:40, 3.40it/s] 43%|████▎ | 20743/48008 [1:11:04<2:00:55, 3.76it/s] 43%|████▎ | 20744/48008 [1:11:04<1:52:00, 4.06it/s] 43%|████▎ | 20745/48008 [1:11:04<1:44:40, 4.34it/s] 43%|████▎ | 20746/48008 [1:11:04<1:41:55, 4.46it/s] 43%|████▎ | 20747/48008 [1:11:05<1:37:56, 4.64it/s] 43%|████▎ | 20748/48008 [1:11:05<1:34:40, 4.80it/s] 43%|████▎ | 20749/48008 [1:11:05<1:32:14, 4.92it/s] 43%|████▎ | 20750/48008 [1:11:05<1:57:09, 3.88it/s] {'loss': 0.2995, 'grad_norm': 0.5, 'learning_rate': 0.00011356023996000666, 'epoch': 0.43} 43%|████▎ | 20750/48008 [1:11:05<1:57:09, 3.88it/s] 43%|████▎ | 20751/48008 [1:11:06<1:48:15, 4.20it/s] 43%|████▎ | 20752/48008 [1:11:06<1:44:17, 4.36it/s] 43%|████▎ | 20753/48008 [1:11:06<1:39:04, 4.58it/s] 43%|████▎ | 20754/48008 [1:11:06<1:34:28, 4.81it/s] 43%|████▎ | 20755/48008 [1:11:06<1:31:39, 4.96it/s] 43%|████▎ | 20756/48008 [1:11:07<1:46:06, 4.28it/s] 43%|████▎ | 20757/48008 [1:11:07<2:06:55, 3.58it/s] 43%|████▎ | 20758/48008 [1:11:07<1:55:01, 3.95it/s] 43%|████▎ | 20759/48008 [1:11:07<1:47:40, 4.22it/s] 43%|████▎ | 20760/48008 [1:11:08<1:40:58, 4.50it/s] 43%|████▎ | 20761/48008 [1:11:08<1:36:37, 4.70it/s] 43%|████▎ | 20762/48008 [1:11:08<1:35:43, 4.74it/s] 43%|████▎ | 20763/48008 [1:11:08<1:32:38, 4.90it/s] 43%|████▎ | 20764/48008 [1:11:08<1:29:58, 5.05it/s] 43%|████▎ | 20765/48008 [1:11:09<1:31:33, 4.96it/s] 43%|████▎ | 20766/48008 [1:11:09<1:29:33, 5.07it/s] 43%|████▎ | 20767/48008 [1:11:09<1:28:37, 5.12it/s] 43%|████▎ | 20768/48008 [1:11:09<1:28:19, 5.14it/s] 43%|████▎ | 20769/48008 [1:11:09<1:27:23, 5.20it/s] 43%|████▎ | 20770/48008 [1:11:10<1:27:11, 5.21it/s] 43%|████▎ | 20771/48008 [1:11:10<1:26:53, 5.22it/s] 43%|████▎ | 20772/48008 [1:11:10<1:42:35, 4.43it/s] 43%|████▎ | 20773/48008 [1:11:10<1:38:42, 4.60it/s] 43%|████▎ | 20774/48008 [1:11:10<1:35:02, 4.78it/s] 43%|████▎ | 20775/48008 [1:11:11<1:33:34, 4.85it/s] 43%|████▎ | 20776/48008 [1:11:11<1:30:49, 5.00it/s] 43%|████▎ | 20777/48008 [1:11:11<1:29:33, 5.07it/s] 43%|████▎ | 20778/48008 [1:11:11<1:30:05, 5.04it/s] 43%|████▎ | 20779/48008 [1:11:11<1:30:41, 5.00it/s] 43%|████▎ | 20780/48008 [1:11:12<1:31:12, 4.98it/s] 43%|████▎ | 20781/48008 [1:11:12<1:29:08, 5.09it/s] 43%|████▎ | 20782/48008 [1:11:12<1:34:16, 4.81it/s] 43%|████▎ | 20783/48008 [1:11:12<1:31:34, 4.95it/s] 43%|████▎ | 20784/48008 [1:11:12<1:31:47, 4.94it/s] 43%|████▎ | 20785/48008 [1:11:13<1:30:08, 5.03it/s] 43%|████▎ | 20786/48008 [1:11:13<1:29:17, 5.08it/s] 43%|████▎ | 20787/48008 [1:11:13<1:28:41, 5.11it/s] 43%|████▎ | 20788/48008 [1:11:13<1:29:39, 5.06it/s] 43%|████▎ | 20789/48008 [1:11:13<1:29:25, 5.07it/s] 43%|████▎ | 20790/48008 [1:11:14<1:29:58, 5.04it/s] 43%|████▎ | 20791/48008 [1:11:14<1:28:17, 5.14it/s] 43%|████▎ | 20792/48008 [1:11:14<1:28:45, 5.11it/s] 43%|████▎ | 20793/48008 [1:11:14<1:34:22, 4.81it/s] 43%|████▎ | 20794/48008 [1:11:14<1:31:29, 4.96it/s] 43%|████▎ | 20795/48008 [1:11:15<1:29:59, 5.04it/s] 43%|████▎ | 20796/48008 [1:11:15<1:30:49, 4.99it/s] 43%|████▎ | 20797/48008 [1:11:15<1:29:24, 5.07it/s] 43%|████▎ | 20798/48008 [1:11:15<1:34:47, 4.78it/s] 43%|████▎ | 20799/48008 [1:11:15<1:34:55, 4.78it/s] 43%|████▎ | 20800/48008 [1:11:16<1:33:09, 4.87it/s] {'loss': 0.299, 'grad_norm': 0.2890625, 'learning_rate': 0.00011335194134310949, 'epoch': 0.43} 43%|████▎ | 20800/48008 [1:11:16<1:33:09, 4.87it/s] 43%|████▎ | 20801/48008 [1:11:16<1:31:37, 4.95it/s] 43%|████▎ | 20802/48008 [1:11:16<1:29:14, 5.08it/s] 43%|████▎ | 20803/48008 [1:11:16<1:29:01, 5.09it/s] 43%|████▎ | 20804/48008 [1:11:16<1:28:15, 5.14it/s] 43%|████▎ | 20805/48008 [1:11:17<1:34:06, 4.82it/s] 43%|████▎ | 20806/48008 [1:11:17<1:31:46, 4.94it/s] 43%|████▎ | 20807/48008 [1:11:17<1:31:37, 4.95it/s] 43%|████▎ | 20808/48008 [1:11:17<1:30:15, 5.02it/s] 43%|████▎ | 20809/48008 [1:11:17<1:30:08, 5.03it/s] 43%|████▎ | 20810/48008 [1:11:18<1:29:04, 5.09it/s] 43%|████▎ | 20811/48008 [1:11:18<1:29:58, 5.04it/s] 43%|████▎ | 20812/48008 [1:11:18<1:29:31, 5.06it/s] 43%|████▎ | 20813/48008 [1:11:18<1:28:36, 5.11it/s] 43%|████▎ | 20814/48008 [1:11:18<1:27:35, 5.17it/s] 43%|████▎ | 20815/48008 [1:11:19<1:27:14, 5.19it/s] 43%|████▎ | 20816/48008 [1:11:19<1:53:37, 3.99it/s] 43%|████▎ | 20817/48008 [1:11:19<1:45:20, 4.30it/s] 43%|████▎ | 20818/48008 [1:11:19<1:39:28, 4.56it/s] 43%|████▎ | 20819/48008 [1:11:20<1:38:21, 4.61it/s] 43%|████▎ | 20820/48008 [1:11:20<1:35:48, 4.73it/s] 43%|████▎ | 20821/48008 [1:11:20<1:34:10, 4.81it/s] 43%|████▎ | 20822/48008 [1:11:20<1:32:06, 4.92it/s] 43%|████▎ | 20823/48008 [1:11:20<1:29:44, 5.05it/s] 43%|████▎ | 20824/48008 [1:11:21<1:28:53, 5.10it/s] 43%|████▎ | 20825/48008 [1:11:21<1:28:16, 5.13it/s] 43%|████▎ | 20826/48008 [1:11:21<1:27:41, 5.17it/s] 43%|████▎ | 20827/48008 [1:11:21<1:27:27, 5.18it/s] 43%|████▎ | 20828/48008 [1:11:21<1:28:45, 5.10it/s] 43%|████▎ | 20829/48008 [1:11:22<1:28:14, 5.13it/s] 43%|████▎ | 20830/48008 [1:11:22<1:30:15, 5.02it/s] 43%|████▎ | 20831/48008 [1:11:22<1:30:53, 4.98it/s] 43%|████▎ | 20832/48008 [1:11:22<1:29:31, 5.06it/s] 43%|████▎ | 20833/48008 [1:11:22<1:29:12, 5.08it/s] 43%|████▎ | 20834/48008 [1:11:22<1:28:23, 5.12it/s] 43%|████▎ | 20835/48008 [1:11:23<1:28:02, 5.14it/s] 43%|████▎ | 20836/48008 [1:11:23<1:29:11, 5.08it/s] 43%|████▎ | 20837/48008 [1:11:23<1:28:24, 5.12it/s] 43%|████▎ | 20838/48008 [1:11:23<1:28:49, 5.10it/s] 43%|████▎ | 20839/48008 [1:11:23<1:29:18, 5.07it/s] 43%|████▎ | 20840/48008 [1:11:24<1:27:53, 5.15it/s] 43%|████▎ | 20841/48008 [1:11:24<1:26:46, 5.22it/s] 43%|████▎ | 20842/48008 [1:11:24<1:26:21, 5.24it/s] 43%|████▎ | 20843/48008 [1:11:24<1:28:00, 5.14it/s] 43%|████▎ | 20844/48008 [1:11:24<1:27:03, 5.20it/s] 43%|████▎ | 20845/48008 [1:11:25<1:33:23, 4.85it/s] 43%|████▎ | 20846/48008 [1:11:25<1:31:26, 4.95it/s] 43%|████▎ | 20847/48008 [1:11:25<1:29:38, 5.05it/s] 43%|████▎ | 20848/48008 [1:11:25<1:28:31, 5.11it/s] 43%|████▎ | 20849/48008 [1:11:25<1:28:05, 5.14it/s] 43%|████▎ | 20850/48008 [1:11:26<1:28:58, 5.09it/s] {'loss': 0.3016, 'grad_norm': 0.5390625, 'learning_rate': 0.0001131436427262123, 'epoch': 0.43} 43%|████▎ | 20850/48008 [1:11:26<1:28:58, 5.09it/s] 43%|████▎ | 20851/48008 [1:11:26<1:30:03, 5.03it/s] 43%|████▎ | 20852/48008 [1:11:26<1:29:16, 5.07it/s] 43%|████▎ | 20853/48008 [1:11:26<1:28:36, 5.11it/s] 43%|████▎ | 20854/48008 [1:11:26<1:28:06, 5.14it/s] 43%|████▎ | 20855/48008 [1:11:27<1:29:05, 5.08it/s] 43%|████▎ | 20856/48008 [1:11:27<1:30:58, 4.97it/s] 43%|████▎ | 20857/48008 [1:11:27<1:28:59, 5.08it/s] 43%|████▎ | 20858/48008 [1:11:27<1:28:38, 5.10it/s] 43%|████▎ | 20859/48008 [1:11:27<1:28:30, 5.11it/s] 43%|████▎ | 20860/48008 [1:11:28<1:34:14, 4.80it/s] 43%|████▎ | 20861/48008 [1:11:28<1:31:57, 4.92it/s] 43%|████▎ | 20862/48008 [1:11:28<1:29:59, 5.03it/s] 43%|████▎ | 20863/48008 [1:11:28<1:28:27, 5.11it/s] 43%|████▎ | 20864/48008 [1:11:28<1:27:48, 5.15it/s] 43%|████▎ | 20865/48008 [1:11:29<1:27:23, 5.18it/s] 43%|████▎ | 20866/48008 [1:11:29<1:26:54, 5.21it/s] 43%|████▎ | 20867/48008 [1:11:29<1:26:37, 5.22it/s] 43%|████▎ | 20868/48008 [1:11:29<1:27:17, 5.18it/s] 43%|████▎ | 20869/48008 [1:11:29<1:33:17, 4.85it/s] 43%|████▎ | 20870/48008 [1:11:30<1:31:22, 4.95it/s] 43%|████▎ | 20871/48008 [1:11:30<1:30:07, 5.02it/s] 43%|████▎ | 20872/48008 [1:11:30<1:28:59, 5.08it/s] 43%|████▎ | 20873/48008 [1:11:30<1:28:25, 5.11it/s] 43%|████▎ | 20874/48008 [1:11:30<1:28:28, 5.11it/s] 43%|████▎ | 20875/48008 [1:11:31<1:27:37, 5.16it/s] 43%|████▎ | 20876/48008 [1:11:31<1:27:50, 5.15it/s] 43%|████▎ | 20877/48008 [1:11:31<1:26:44, 5.21it/s] 43%|████▎ | 20878/48008 [1:11:31<1:26:36, 5.22it/s] 43%|████▎ | 20879/48008 [1:11:31<1:26:40, 5.22it/s] 43%|████▎ | 20880/48008 [1:11:32<1:29:02, 5.08it/s] 43%|████▎ | 20881/48008 [1:11:32<1:28:37, 5.10it/s] 43%|████▎ | 20882/48008 [1:11:32<1:28:36, 5.10it/s] 43%|████▎ | 20883/48008 [1:11:32<1:28:08, 5.13it/s] 44%|████▎ | 20884/48008 [1:11:32<1:27:14, 5.18it/s] 44%|████▎ | 20885/48008 [1:11:33<1:28:07, 5.13it/s] 44%|████▎ | 20886/48008 [1:11:33<1:43:00, 4.39it/s] 44%|████▎ | 20887/48008 [1:11:33<1:38:04, 4.61it/s] 44%|████▎ | 20888/48008 [1:11:33<1:35:02, 4.76it/s] 44%|████▎ | 20889/48008 [1:11:33<1:31:20, 4.95it/s] 44%|████▎ | 20890/48008 [1:11:34<1:29:04, 5.07it/s] 44%|████▎ | 20891/48008 [1:11:34<1:28:22, 5.11it/s] 44%|████▎ | 20892/48008 [1:11:34<1:28:27, 5.11it/s] 44%|████▎ | 20893/48008 [1:11:34<1:34:04, 4.80it/s] 44%|████▎ | 20894/48008 [1:11:34<1:31:24, 4.94it/s] 44%|████▎ | 20895/48008 [1:11:35<1:36:06, 4.70it/s] 44%|████▎ | 20896/48008 [1:11:35<1:59:34, 3.78it/s] 44%|████▎ | 20897/48008 [1:11:35<1:48:57, 4.15it/s] 44%|████▎ | 20898/48008 [1:11:35<1:42:24, 4.41it/s] 44%|████▎ | 20899/48008 [1:11:36<1:37:31, 4.63it/s] 44%|████▎ | 20900/48008 [1:11:36<1:33:05, 4.85it/s] {'loss': 0.3015, 'grad_norm': 0.55078125, 'learning_rate': 0.00011293534410931513, 'epoch': 0.44} 44%|████▎ | 20900/48008 [1:11:36<1:33:05, 4.85it/s] 44%|████▎ | 20901/48008 [1:11:36<1:30:40, 4.98it/s] 44%|████▎ | 20902/48008 [1:11:36<1:30:57, 4.97it/s] 44%|████▎ | 20903/48008 [1:11:36<1:35:40, 4.72it/s] 44%|████▎ | 20904/48008 [1:11:37<1:33:05, 4.85it/s] 44%|████▎ | 20905/48008 [1:11:37<1:31:00, 4.96it/s] 44%|████▎ | 20906/48008 [1:11:37<1:32:13, 4.90it/s] 44%|████▎ | 20907/48008 [1:11:37<1:30:06, 5.01it/s] 44%|████▎ | 20908/48008 [1:11:37<1:28:47, 5.09it/s] 44%|████▎ | 20909/48008 [1:11:38<1:28:04, 5.13it/s] 44%|████▎ | 20910/48008 [1:11:38<1:27:35, 5.16it/s] 44%|████▎ | 20911/48008 [1:11:38<1:26:58, 5.19it/s] 44%|████▎ | 20912/48008 [1:11:38<1:27:31, 5.16it/s] 44%|████▎ | 20913/48008 [1:11:38<1:26:34, 5.22it/s] 44%|████▎ | 20914/48008 [1:11:39<1:26:53, 5.20it/s] 44%|████▎ | 20915/48008 [1:11:39<1:26:54, 5.20it/s] 44%|████▎ | 20916/48008 [1:11:39<1:42:24, 4.41it/s] 44%|████▎ | 20917/48008 [1:11:39<1:39:01, 4.56it/s] 44%|████▎ | 20918/48008 [1:11:40<1:51:10, 4.06it/s] 44%|████▎ | 20919/48008 [1:11:40<1:43:57, 4.34it/s] 44%|████▎ | 20920/48008 [1:11:40<1:39:00, 4.56it/s] 44%|████▎ | 20921/48008 [1:11:40<1:36:15, 4.69it/s] 44%|████▎ | 20922/48008 [1:11:40<1:34:01, 4.80it/s] 44%|████▎ | 20923/48008 [1:11:40<1:31:57, 4.91it/s] 44%|████▎ | 20924/48008 [1:11:41<1:29:57, 5.02it/s] 44%|████▎ | 20925/48008 [1:11:41<1:55:19, 3.91it/s] 44%|████▎ | 20926/48008 [1:11:41<1:46:13, 4.25it/s] 44%|████▎ | 20927/48008 [1:11:41<1:40:10, 4.51it/s] 44%|████▎ | 20928/48008 [1:11:42<1:37:46, 4.62it/s] 44%|████▎ | 20929/48008 [1:11:42<1:33:27, 4.83it/s] 44%|████▎ | 20930/48008 [1:11:42<1:32:51, 4.86it/s] 44%|████▎ | 20931/48008 [1:11:42<1:33:22, 4.83it/s] 44%|████▎ | 20932/48008 [1:11:42<1:32:46, 4.86it/s] 44%|████▎ | 20933/48008 [1:11:43<1:29:56, 5.02it/s] 44%|████▎ | 20934/48008 [1:11:43<1:44:44, 4.31it/s] 44%|████▎ | 20935/48008 [1:11:43<1:39:46, 4.52it/s] 44%|████▎ | 20936/48008 [1:11:43<1:37:38, 4.62it/s] 44%|████▎ | 20937/48008 [1:11:44<1:34:37, 4.77it/s] 44%|████▎ | 20938/48008 [1:11:44<1:31:54, 4.91it/s] 44%|████▎ | 20939/48008 [1:11:44<1:30:09, 5.00it/s] 44%|████▎ | 20940/48008 [1:11:44<1:28:57, 5.07it/s] 44%|████▎ | 20941/48008 [1:11:44<1:27:43, 5.14it/s] 44%|████▎ | 20942/48008 [1:11:44<1:26:48, 5.20it/s] 44%|████▎ | 20943/48008 [1:11:45<1:27:47, 5.14it/s] 44%|████▎ | 20944/48008 [1:11:45<1:42:43, 4.39it/s] 44%|████▎ | 20945/48008 [1:11:45<1:38:10, 4.59it/s] 44%|████▎ | 20946/48008 [1:11:45<1:34:26, 4.78it/s] 44%|████▎ | 20947/48008 [1:11:46<1:31:58, 4.90it/s] 44%|████▎ | 20948/48008 [1:11:46<1:36:50, 4.66it/s] 44%|████▎ | 20949/48008 [1:11:46<1:35:23, 4.73it/s] 44%|████▎ | 20950/48008 [1:11:46<1:35:15, 4.73it/s] {'loss': 0.3, 'grad_norm': 0.3828125, 'learning_rate': 0.00011272704549241793, 'epoch': 0.44} 44%|████▎ | 20950/48008 [1:11:46<1:35:15, 4.73it/s] 44%|████▎ | 20951/48008 [1:11:46<1:32:52, 4.86it/s] 44%|████▎ | 20952/48008 [1:11:47<1:30:47, 4.97it/s] 44%|████▎ | 20953/48008 [1:11:47<1:30:42, 4.97it/s] 44%|████▎ | 20954/48008 [1:11:47<1:31:06, 4.95it/s] 44%|████▎ | 20955/48008 [1:11:47<1:29:28, 5.04it/s] 44%|████▎ | 20956/48008 [1:11:47<1:28:13, 5.11it/s] 44%|████▎ | 20957/48008 [1:11:48<1:27:41, 5.14it/s] 44%|████▎ | 20958/48008 [1:11:48<1:26:49, 5.19it/s] 44%|████▎ | 20959/48008 [1:11:48<1:26:37, 5.20it/s] 44%|████▎ | 20960/48008 [1:11:48<1:28:05, 5.12it/s] 44%|████▎ | 20961/48008 [1:11:48<1:27:31, 5.15it/s] 44%|████▎ | 20962/48008 [1:11:49<1:27:23, 5.16it/s] 44%|████▎ | 20963/48008 [1:11:49<1:28:01, 5.12it/s] 44%|████▎ | 20964/48008 [1:11:49<1:27:47, 5.13it/s] 44%|████▎ | 20965/48008 [1:11:49<1:28:43, 5.08it/s] 44%|████▎ | 20966/48008 [1:11:49<1:28:35, 5.09it/s] 44%|████▎ | 20967/48008 [1:11:50<1:28:07, 5.11it/s] 44%|████▎ | 20968/48008 [1:11:50<1:29:09, 5.05it/s] 44%|████▎ | 20969/48008 [1:11:50<1:28:03, 5.12it/s] 44%|████▎ | 20970/48008 [1:11:50<1:26:56, 5.18it/s] 44%|████▎ | 20971/48008 [1:11:50<1:27:18, 5.16it/s] 44%|████▎ | 20972/48008 [1:11:50<1:26:39, 5.20it/s] 44%|████▎ | 20973/48008 [1:11:51<1:26:30, 5.21it/s] 44%|████▎ | 20974/48008 [1:11:51<1:26:34, 5.20it/s] 44%|████▎ | 20975/48008 [1:11:51<1:25:51, 5.25it/s] 44%|████▎ | 20976/48008 [1:11:51<1:25:02, 5.30it/s] 44%|████▎ | 20977/48008 [1:11:51<1:28:02, 5.12it/s] 44%|████▎ | 20978/48008 [1:11:52<1:28:07, 5.11it/s] 44%|████▎ | 20979/48008 [1:11:52<1:33:51, 4.80it/s] 44%|████▎ | 20980/48008 [1:11:52<1:31:31, 4.92it/s] 44%|████▎ | 20981/48008 [1:11:52<1:31:09, 4.94it/s] 44%|████▎ | 20982/48008 [1:11:52<1:29:56, 5.01it/s] 44%|████▎ | 20983/48008 [1:11:53<1:29:06, 5.05it/s] 44%|████▎ | 20984/48008 [1:11:53<1:27:57, 5.12it/s] 44%|████▎ | 20985/48008 [1:11:53<1:28:24, 5.09it/s] 44%|████▎ | 20986/48008 [1:11:53<1:27:11, 5.16it/s] 44%|████▎ | 20987/48008 [1:11:53<1:28:22, 5.10it/s] 44%|████▎ | 20988/48008 [1:11:54<1:33:50, 4.80it/s] 44%|████▎ | 20989/48008 [1:11:54<1:32:21, 4.88it/s] 44%|████▎ | 20990/48008 [1:11:54<1:31:28, 4.92it/s] 44%|████▎ | 20991/48008 [1:11:54<1:56:28, 3.87it/s] 44%|████▎ | 20992/48008 [1:11:55<1:53:47, 3.96it/s] 44%|████▎ | 20993/48008 [1:11:55<1:44:52, 4.29it/s] 44%|████▎ | 20994/48008 [1:11:56<3:11:20, 2.35it/s] 44%|████▎ | 20995/48008 [1:11:56<2:40:10, 2.81it/s] 44%|████▎ | 20996/48008 [1:11:56<2:17:51, 3.27it/s] 44%|████▎ | 20997/48008 [1:11:56<2:02:32, 3.67it/s] 44%|████▎ | 20998/48008 [1:11:57<1:53:44, 3.96it/s] 44%|████▎ | 20999/48008 [1:11:57<1:45:04, 4.28it/s] 44%|████▎ | 21000/48008 [1:11:57<1:38:50, 4.55it/s] {'loss': 0.3003, 'grad_norm': 0.201171875, 'learning_rate': 0.00011251874687552075, 'epoch': 0.44} 44%|████▎ | 21000/48008 [1:11:57<1:38:50, 4.55it/s] 44%|████▎ | 21001/48008 [1:11:57<1:36:21, 4.67it/s] 44%|████▎ | 21002/48008 [1:11:57<1:32:31, 4.86it/s] 44%|████▎ | 21003/48008 [1:11:58<1:31:45, 4.90it/s] 44%|████▍ | 21004/48008 [1:11:58<1:29:54, 5.01it/s] 44%|████▍ | 21005/48008 [1:11:58<1:28:48, 5.07it/s] 44%|████▍ | 21006/48008 [1:11:58<1:28:03, 5.11it/s] 44%|████▍ | 21007/48008 [1:11:58<1:29:12, 5.04it/s] 44%|████▍ | 21008/48008 [1:11:59<1:54:50, 3.92it/s] 44%|████▍ | 21009/48008 [1:11:59<1:47:58, 4.17it/s] 44%|████▍ | 21010/48008 [1:11:59<1:41:53, 4.42it/s] 44%|████▍ | 21011/48008 [1:11:59<1:38:01, 4.59it/s] 44%|████▍ | 21012/48008 [1:12:00<1:40:34, 4.47it/s] 44%|████▍ | 21013/48008 [1:12:00<1:42:25, 4.39it/s] 44%|████▍ | 21014/48008 [1:12:00<1:37:52, 4.60it/s] 44%|████▍ | 21015/48008 [1:12:00<1:40:49, 4.46it/s] 44%|████▍ | 21016/48008 [1:12:00<1:36:36, 4.66it/s] 44%|████▍ | 21017/48008 [1:12:01<1:32:58, 4.84it/s] 44%|████▍ | 21018/48008 [1:12:01<1:30:32, 4.97it/s] 44%|████▍ | 21019/48008 [1:12:01<1:35:41, 4.70it/s] 44%|████▍ | 21020/48008 [1:12:01<1:32:36, 4.86it/s] 44%|████▍ | 21021/48008 [1:12:01<1:30:40, 4.96it/s] 44%|████▍ | 21022/48008 [1:12:02<1:28:54, 5.06it/s] 44%|████▍ | 21023/48008 [1:12:02<1:28:01, 5.11it/s] 44%|████▍ | 21024/48008 [1:12:02<1:27:35, 5.13it/s] 44%|████▍ | 21025/48008 [1:12:02<1:34:07, 4.78it/s] 44%|████▍ | 21026/48008 [1:12:02<1:32:12, 4.88it/s] 44%|████▍ | 21027/48008 [1:12:03<1:36:43, 4.65it/s] 44%|████▍ | 21028/48008 [1:12:03<1:33:03, 4.83it/s] 44%|████▍ | 21029/48008 [1:12:03<1:31:26, 4.92it/s] 44%|████▍ | 21030/48008 [1:12:03<1:30:57, 4.94it/s] 44%|████▍ | 21031/48008 [1:12:03<1:29:37, 5.02it/s] 44%|████▍ | 21032/48008 [1:12:04<1:28:54, 5.06it/s] 44%|████▍ | 21033/48008 [1:12:04<1:28:17, 5.09it/s] 44%|████▍ | 21034/48008 [1:12:04<1:27:11, 5.16it/s] 44%|████▍ | 21035/48008 [1:12:04<1:26:36, 5.19it/s] 44%|████▍ | 21036/48008 [1:12:04<1:26:05, 5.22it/s] 44%|████▍ | 21037/48008 [1:12:05<1:26:12, 5.21it/s] 44%|████▍ | 21038/48008 [1:12:05<1:26:07, 5.22it/s] 44%|████▍ | 21039/48008 [1:12:05<1:28:43, 5.07it/s] 44%|████▍ | 21040/48008 [1:12:05<1:27:39, 5.13it/s] 44%|████▍ | 21041/48008 [1:12:05<1:43:08, 4.36it/s] 44%|████▍ | 21042/48008 [1:12:06<1:37:36, 4.60it/s] 44%|████▍ | 21043/48008 [1:12:06<1:33:44, 4.79it/s] 44%|████▍ | 21044/48008 [1:12:06<1:32:55, 4.84it/s] 44%|████▍ | 21045/48008 [1:12:06<1:30:49, 4.95it/s] 44%|████▍ | 21046/48008 [1:12:06<1:29:08, 5.04it/s] 44%|████▍ | 21047/48008 [1:12:07<1:31:15, 4.92it/s] 44%|████▍ | 21048/48008 [1:12:07<1:35:41, 4.70it/s] 44%|████▍ | 21049/48008 [1:12:07<1:39:38, 4.51it/s] 44%|████▍ | 21050/48008 [1:12:07<1:37:47, 4.59it/s] {'loss': 0.3002, 'grad_norm': 0.78125, 'learning_rate': 0.00011231044825862355, 'epoch': 0.44} 44%|████▍ | 21050/48008 [1:12:07<1:37:47, 4.59it/s] 44%|████▍ | 21051/48008 [1:12:08<1:34:26, 4.76it/s] 44%|████▍ | 21052/48008 [1:12:08<1:31:41, 4.90it/s] 44%|████▍ | 21053/48008 [1:12:08<1:31:20, 4.92it/s] 44%|████▍ | 21054/48008 [1:12:08<1:30:27, 4.97it/s] 44%|████▍ | 21055/48008 [1:12:08<1:29:38, 5.01it/s] 44%|████▍ | 21056/48008 [1:12:08<1:28:36, 5.07it/s] 44%|████▍ | 21057/48008 [1:12:09<1:27:24, 5.14it/s] 44%|████▍ | 21058/48008 [1:12:09<1:26:23, 5.20it/s] 44%|████▍ | 21059/48008 [1:12:09<1:25:38, 5.24it/s] 44%|████▍ | 21060/48008 [1:12:09<1:27:11, 5.15it/s] 44%|████▍ | 21061/48008 [1:12:09<1:28:17, 5.09it/s] 44%|████▍ | 21062/48008 [1:12:10<1:26:49, 5.17it/s] 44%|████▍ | 21063/48008 [1:12:10<1:25:55, 5.23it/s] 44%|████▍ | 21064/48008 [1:12:10<1:27:32, 5.13it/s] 44%|████▍ | 21065/48008 [1:12:10<1:53:24, 3.96it/s] 44%|████▍ | 21066/48008 [1:12:11<1:44:11, 4.31it/s] 44%|████▍ | 21067/48008 [1:12:11<1:38:51, 4.54it/s] 44%|████▍ | 21068/48008 [1:12:11<1:35:36, 4.70it/s] 44%|████▍ | 21069/48008 [1:12:11<1:30:54, 4.94it/s] 44%|████▍ | 21070/48008 [1:12:11<1:29:28, 5.02it/s] 44%|████▍ | 21071/48008 [1:12:12<1:31:08, 4.93it/s] 44%|████▍ | 21072/48008 [1:12:12<1:29:38, 5.01it/s] 44%|████▍ | 21073/48008 [1:12:12<1:27:52, 5.11it/s] 44%|████▍ | 21074/48008 [1:12:12<1:28:45, 5.06it/s] 44%|████▍ | 21075/48008 [1:12:12<1:27:47, 5.11it/s] 44%|████▍ | 21076/48008 [1:12:13<1:26:57, 5.16it/s] 44%|████▍ | 21077/48008 [1:12:13<1:26:19, 5.20it/s] 44%|████▍ | 21078/48008 [1:12:13<1:25:17, 5.26it/s] 44%|████▍ | 21079/48008 [1:12:13<1:25:43, 5.24it/s] 44%|████▍ | 21080/48008 [1:12:13<1:26:28, 5.19it/s] 44%|████▍ | 21081/48008 [1:12:13<1:26:06, 5.21it/s] 44%|████▍ | 21082/48008 [1:12:14<1:28:22, 5.08it/s] 44%|████▍ | 21083/48008 [1:12:14<1:27:35, 5.12it/s] 44%|████▍ | 21084/48008 [1:12:14<1:33:03, 4.82it/s] 44%|████▍ | 21085/48008 [1:12:14<1:30:22, 4.97it/s] 44%|████▍ | 21086/48008 [1:12:14<1:28:48, 5.05it/s] 44%|████▍ | 21087/48008 [1:12:15<1:30:30, 4.96it/s] 44%|████▍ | 21088/48008 [1:12:15<1:28:03, 5.10it/s] 44%|████▍ | 21089/48008 [1:12:15<1:27:37, 5.12it/s] 44%|████▍ | 21090/48008 [1:12:15<1:27:01, 5.16it/s] 44%|████▍ | 21091/48008 [1:12:15<1:26:48, 5.17it/s] 44%|████▍ | 21092/48008 [1:12:16<1:26:36, 5.18it/s] 44%|████▍ | 21093/48008 [1:12:16<1:26:52, 5.16it/s] 44%|████▍ | 21094/48008 [1:12:16<1:26:09, 5.21it/s] 44%|████▍ | 21095/48008 [1:12:16<1:25:23, 5.25it/s] 44%|████▍ | 21096/48008 [1:12:16<1:25:35, 5.24it/s] 44%|████▍ | 21097/48008 [1:12:17<1:27:38, 5.12it/s] 44%|████▍ | 21098/48008 [1:12:17<1:26:23, 5.19it/s] 44%|████▍ | 21099/48008 [1:12:17<1:26:01, 5.21it/s] 44%|████▍ | 21100/48008 [1:12:17<1:26:46, 5.17it/s] {'loss': 0.3034, 'grad_norm': 0.29296875, 'learning_rate': 0.00011210214964172638, 'epoch': 0.44} 44%|████▍ | 21100/48008 [1:12:17<1:26:46, 5.17it/s] 44%|████▍ | 21101/48008 [1:12:17<1:26:08, 5.21it/s] 44%|████▍ | 21102/48008 [1:12:18<1:25:37, 5.24it/s] 44%|████▍ | 21103/48008 [1:12:18<1:26:40, 5.17it/s] 44%|████▍ | 21104/48008 [1:12:18<1:27:38, 5.12it/s] 44%|████▍ | 21105/48008 [1:12:18<1:27:46, 5.11it/s] 44%|████▍ | 21106/48008 [1:12:18<1:26:54, 5.16it/s] 44%|████▍ | 21107/48008 [1:12:19<1:28:48, 5.05it/s] 44%|████▍ | 21108/48008 [1:12:19<1:33:57, 4.77it/s] 44%|████▍ | 21109/48008 [1:12:19<1:31:18, 4.91it/s] 44%|████▍ | 21110/48008 [1:12:19<1:29:08, 5.03it/s] 44%|████▍ | 21111/48008 [1:12:19<1:28:18, 5.08it/s] 44%|████▍ | 21112/48008 [1:12:20<1:27:36, 5.12it/s] 44%|████▍ | 21113/48008 [1:12:20<1:33:36, 4.79it/s] 44%|████▍ | 21114/48008 [1:12:20<1:30:36, 4.95it/s] 44%|████▍ | 21115/48008 [1:12:20<1:29:09, 5.03it/s] 44%|████▍ | 21116/48008 [1:12:20<1:27:53, 5.10it/s] 44%|████▍ | 21117/48008 [1:12:21<1:27:41, 5.11it/s] 44%|████▍ | 21118/48008 [1:12:21<1:28:04, 5.09it/s] 44%|████▍ | 21119/48008 [1:12:21<1:27:57, 5.09it/s] 44%|████▍ | 21120/48008 [1:12:21<1:27:20, 5.13it/s] 44%|████▍ | 21121/48008 [1:12:21<1:26:33, 5.18it/s] 44%|████▍ | 21122/48008 [1:12:22<1:26:01, 5.21it/s] 44%|████▍ | 21123/48008 [1:12:22<1:27:20, 5.13it/s] 44%|████▍ | 21124/48008 [1:12:22<1:26:34, 5.18it/s] 44%|████▍ | 21125/48008 [1:12:22<1:26:08, 5.20it/s] 44%|████▍ | 21126/48008 [1:12:22<1:24:32, 5.30it/s] 44%|████▍ | 21127/48008 [1:12:22<1:24:51, 5.28it/s] 44%|████▍ | 21128/48008 [1:12:23<1:23:57, 5.34it/s] 44%|████▍ | 21129/48008 [1:12:23<1:23:46, 5.35it/s] 44%|████▍ | 21130/48008 [1:12:23<1:24:08, 5.32it/s] 44%|████▍ | 21131/48008 [1:12:23<1:26:14, 5.19it/s] 44%|████▍ | 21132/48008 [1:12:23<1:25:57, 5.21it/s] 44%|████▍ | 21133/48008 [1:12:24<1:26:18, 5.19it/s] 44%|████▍ | 21134/48008 [1:12:24<1:26:02, 5.21it/s] 44%|████▍ | 21135/48008 [1:12:24<1:26:05, 5.20it/s] 44%|████▍ | 21136/48008 [1:12:24<1:25:36, 5.23it/s] 44%|████▍ | 21137/48008 [1:12:24<1:27:28, 5.12it/s] 44%|████▍ | 21138/48008 [1:12:25<1:26:18, 5.19it/s] 44%|████▍ | 21139/48008 [1:12:25<1:25:29, 5.24it/s] 44%|████▍ | 21140/48008 [1:12:25<1:25:26, 5.24it/s] 44%|████▍ | 21141/48008 [1:12:25<1:26:05, 5.20it/s] 44%|████▍ | 21142/48008 [1:12:25<1:25:58, 5.21it/s] 44%|████▍ | 21143/48008 [1:12:26<1:26:00, 5.21it/s] 44%|████▍ | 21144/48008 [1:12:26<1:26:52, 5.15it/s] 44%|████▍ | 21145/48008 [1:12:26<1:26:40, 5.17it/s] 44%|████▍ | 21146/48008 [1:12:26<1:25:34, 5.23it/s] 44%|████▍ | 21147/48008 [1:12:26<1:27:07, 5.14it/s] 44%|████▍ | 21148/48008 [1:12:27<1:26:04, 5.20it/s] 44%|████▍ | 21149/48008 [1:12:27<1:25:43, 5.22it/s] 44%|████▍ | 21150/48008 [1:12:27<1:25:59, 5.21it/s] {'loss': 0.3026, 'grad_norm': 0.1298828125, 'learning_rate': 0.00011189385102482921, 'epoch': 0.44} 44%|████▍ | 21150/48008 [1:12:27<1:25:59, 5.21it/s] 44%|████▍ | 21151/48008 [1:12:27<1:25:42, 5.22it/s] 44%|████▍ | 21152/48008 [1:12:27<1:51:55, 4.00it/s] 44%|████▍ | 21153/48008 [1:12:28<1:44:14, 4.29it/s] 44%|████▍ | 21154/48008 [1:12:28<1:38:14, 4.56it/s] 44%|████▍ | 21155/48008 [1:12:28<1:33:48, 4.77it/s] 44%|████▍ | 21156/48008 [1:12:28<1:30:50, 4.93it/s] 44%|████▍ | 21157/48008 [1:12:28<1:29:06, 5.02it/s] 44%|████▍ | 21158/48008 [1:12:29<1:27:46, 5.10it/s] 44%|████▍ | 21159/48008 [1:12:29<1:26:35, 5.17it/s] 44%|████▍ | 21160/48008 [1:12:29<1:26:02, 5.20it/s] 44%|████▍ | 21161/48008 [1:12:29<1:25:32, 5.23it/s] 44%|████▍ | 21162/48008 [1:12:29<1:26:51, 5.15it/s] 44%|████▍ | 21163/48008 [1:12:30<1:27:32, 5.11it/s] 44%|████▍ | 21164/48008 [1:12:30<1:28:38, 5.05it/s] 44%|████▍ | 21165/48008 [1:12:30<1:27:28, 5.11it/s] 44%|████▍ | 21166/48008 [1:12:30<1:26:51, 5.15it/s] 44%|████▍ | 21167/48008 [1:12:30<1:27:24, 5.12it/s] 44%|████▍ | 21168/48008 [1:12:31<1:26:21, 5.18it/s] 44%|████▍ | 21169/48008 [1:12:31<1:26:19, 5.18it/s] 44%|████▍ | 21170/48008 [1:12:31<1:27:47, 5.10it/s] 44%|████▍ | 21171/48008 [1:12:31<1:26:15, 5.19it/s] 44%|████▍ | 21172/48008 [1:12:31<1:26:05, 5.20it/s] 44%|████▍ | 21173/48008 [1:12:32<1:25:48, 5.21it/s] 44%|████▍ | 21174/48008 [1:12:32<1:31:59, 4.86it/s] 44%|████▍ | 21175/48008 [1:12:32<1:29:43, 4.98it/s] 44%|████▍ | 21176/48008 [1:12:32<1:29:26, 5.00it/s] 44%|████▍ | 21177/48008 [1:12:32<1:27:30, 5.11it/s] 44%|████▍ | 21178/48008 [1:12:33<1:33:15, 4.80it/s] 44%|████▍ | 21179/48008 [1:12:33<1:32:29, 4.83it/s] 44%|████▍ | 21180/48008 [1:12:33<1:31:30, 4.89it/s] 44%|████▍ | 21181/48008 [1:12:33<1:29:31, 4.99it/s] 44%|████▍ | 21182/48008 [1:12:33<1:30:57, 4.92it/s] 44%|████▍ | 21183/48008 [1:12:34<1:29:14, 5.01it/s] 44%|████▍ | 21184/48008 [1:12:34<1:27:55, 5.08it/s] 44%|████▍ | 21185/48008 [1:12:34<1:27:08, 5.13it/s] 44%|████▍ | 21186/48008 [1:12:34<1:27:35, 5.10it/s] 44%|████▍ | 21187/48008 [1:12:34<1:26:54, 5.14it/s] 44%|████▍ | 21188/48008 [1:12:35<1:26:20, 5.18it/s] 44%|████▍ | 21189/48008 [1:12:35<1:26:02, 5.20it/s] 44%|████▍ | 21190/48008 [1:12:35<1:26:09, 5.19it/s] 44%|████▍ | 21191/48008 [1:12:35<1:25:49, 5.21it/s] 44%|████▍ | 21192/48008 [1:12:35<1:25:47, 5.21it/s] 44%|████▍ | 21193/48008 [1:12:35<1:25:19, 5.24it/s] 44%|████▍ | 21194/48008 [1:12:36<1:25:02, 5.25it/s] 44%|████▍ | 21195/48008 [1:12:36<1:31:33, 4.88it/s] 44%|████▍ | 21196/48008 [1:12:36<1:29:51, 4.97it/s] 44%|████▍ | 21197/48008 [1:12:36<1:27:59, 5.08it/s] 44%|████▍ | 21198/48008 [1:12:36<1:26:37, 5.16it/s] 44%|████▍ | 21199/48008 [1:12:37<1:27:53, 5.08it/s] 44%|████▍ | 21200/48008 [1:12:37<1:27:14, 5.12it/s] {'loss': 0.3017, 'grad_norm': 0.150390625, 'learning_rate': 0.00011168555240793202, 'epoch': 0.44} 44%|████▍ | 21200/48008 [1:12:37<1:27:14, 5.12it/s] 44%|████▍ | 21201/48008 [1:12:37<1:26:46, 5.15it/s] 44%|████▍ | 21202/48008 [1:12:37<1:27:13, 5.12it/s] 44%|████▍ | 21203/48008 [1:12:37<1:27:37, 5.10it/s] 44%|████▍ | 21204/48008 [1:12:38<1:27:46, 5.09it/s] 44%|████▍ | 21205/48008 [1:12:38<1:27:09, 5.13it/s] 44%|████▍ | 21206/48008 [1:12:38<1:26:52, 5.14it/s] 44%|████▍ | 21207/48008 [1:12:38<1:28:15, 5.06it/s] 44%|████▍ | 21208/48008 [1:12:38<1:27:31, 5.10it/s] 44%|████▍ | 21209/48008 [1:12:39<1:26:46, 5.15it/s] 44%|████▍ | 21210/48008 [1:12:39<1:27:44, 5.09it/s] 44%|████▍ | 21211/48008 [1:12:39<1:26:52, 5.14it/s] 44%|████▍ | 21212/48008 [1:12:39<1:26:10, 5.18it/s] 44%|████▍ | 21213/48008 [1:12:39<1:26:50, 5.14it/s] 44%|████▍ | 21214/48008 [1:12:40<1:52:33, 3.97it/s] 44%|████▍ | 21215/48008 [1:12:40<1:45:47, 4.22it/s] 44%|████▍ | 21216/48008 [1:12:40<1:39:32, 4.49it/s] 44%|████▍ | 21217/48008 [1:12:40<1:35:40, 4.67it/s] 44%|████▍ | 21218/48008 [1:12:41<1:32:33, 4.82it/s] 44%|████▍ | 21219/48008 [1:12:41<1:31:01, 4.90it/s] 44%|████▍ | 21220/48008 [1:12:41<1:30:12, 4.95it/s] 44%|████▍ | 21221/48008 [1:12:41<1:28:07, 5.07it/s] 44%|████▍ | 21222/48008 [1:12:41<1:28:51, 5.02it/s] 44%|████▍ | 21223/48008 [1:12:42<1:27:48, 5.08it/s] 44%|████▍ | 21224/48008 [1:12:42<1:28:06, 5.07it/s] 44%|████▍ | 21225/48008 [1:12:42<1:33:24, 4.78it/s] 44%|████▍ | 21226/48008 [1:12:42<1:30:49, 4.91it/s] 44%|████▍ | 21227/48008 [1:12:42<1:32:15, 4.84it/s] 44%|████▍ | 21228/48008 [1:12:43<1:30:29, 4.93it/s] 44%|████▍ | 21229/48008 [1:12:43<1:29:03, 5.01it/s] 44%|████▍ | 21230/48008 [1:12:43<1:28:17, 5.05it/s] 44%|████▍ | 21231/48008 [1:12:43<1:28:44, 5.03it/s] 44%|████▍ | 21232/48008 [1:12:43<1:27:42, 5.09it/s] 44%|████▍ | 21233/48008 [1:12:44<1:27:47, 5.08it/s] 44%|████▍ | 21234/48008 [1:12:44<1:27:05, 5.12it/s] 44%|████▍ | 21235/48008 [1:12:44<1:26:17, 5.17it/s] 44%|████▍ | 21236/48008 [1:12:44<1:25:11, 5.24it/s] 44%|████▍ | 21237/48008 [1:12:44<1:26:40, 5.15it/s] 44%|████▍ | 21238/48008 [1:12:45<1:27:31, 5.10it/s] 44%|████▍ | 21239/48008 [1:12:45<1:27:06, 5.12it/s] 44%|████▍ | 21240/48008 [1:12:45<1:52:44, 3.96it/s] 44%|████▍ | 21241/48008 [1:12:45<1:44:14, 4.28it/s] 44%|████▍ | 21242/48008 [1:12:45<1:38:29, 4.53it/s] 44%|████▍ | 21243/48008 [1:12:46<1:33:54, 4.75it/s] 44%|████▍ | 21244/48008 [1:12:46<1:33:11, 4.79it/s] 44%|████▍ | 21245/48008 [1:12:46<1:29:55, 4.96it/s] 44%|████▍ | 21246/48008 [1:12:46<1:28:44, 5.03it/s] 44%|████▍ | 21247/48008 [1:12:46<1:28:16, 5.05it/s] 44%|████▍ | 21248/48008 [1:12:47<1:27:11, 5.12it/s] 44%|████▍ | 21249/48008 [1:12:47<1:26:05, 5.18it/s] 44%|████▍ | 21250/48008 [1:12:47<1:26:12, 5.17it/s] {'loss': 0.3021, 'grad_norm': 0.197265625, 'learning_rate': 0.00011147725379103484, 'epoch': 0.44} 44%|████▍ | 21250/48008 [1:12:47<1:26:12, 5.17it/s] 44%|████▍ | 21251/48008 [1:12:47<1:26:02, 5.18it/s] 44%|████▍ | 21252/48008 [1:12:47<1:25:48, 5.20it/s] 44%|████▍ | 21253/48008 [1:12:48<1:27:26, 5.10it/s] 44%|████▍ | 21254/48008 [1:12:48<1:27:46, 5.08it/s] 44%|████▍ | 21255/48008 [1:12:48<1:26:02, 5.18it/s] 44%|████▍ | 21256/48008 [1:12:48<1:27:15, 5.11it/s] 44%|████▍ | 21257/48008 [1:12:48<1:26:48, 5.14it/s] 44%|████▍ | 21258/48008 [1:12:49<1:26:11, 5.17it/s] 44%|████▍ | 21259/48008 [1:12:49<1:26:06, 5.18it/s] 44%|████▍ | 21260/48008 [1:12:49<1:25:32, 5.21it/s] 44%|████▍ | 21261/48008 [1:12:49<1:25:16, 5.23it/s] 44%|████▍ | 21262/48008 [1:12:49<1:40:31, 4.43it/s] 44%|████▍ | 21263/48008 [1:12:50<1:35:42, 4.66it/s] 44%|████▍ | 21264/48008 [1:12:50<1:33:59, 4.74it/s] 44%|████▍ | 21265/48008 [1:12:50<1:30:57, 4.90it/s] 44%|████▍ | 21266/48008 [1:12:50<1:29:15, 4.99it/s] 44%|████▍ | 21267/48008 [1:12:50<1:28:02, 5.06it/s] 44%|████▍ | 21268/48008 [1:12:51<1:27:24, 5.10it/s] 44%|████▍ | 21269/48008 [1:12:51<1:52:48, 3.95it/s] 44%|████▍ | 21270/48008 [1:12:51<1:43:54, 4.29it/s] 44%|████▍ | 21271/48008 [1:12:51<1:38:33, 4.52it/s] 44%|████▍ | 21272/48008 [1:12:52<2:00:47, 3.69it/s] 44%|████▍ | 21273/48008 [1:12:52<1:50:07, 4.05it/s] 44%|████▍ | 21274/48008 [1:12:52<1:43:54, 4.29it/s] 44%|████▍ | 21275/48008 [1:12:52<1:38:11, 4.54it/s] 44%|████▍ | 21276/48008 [1:12:53<1:34:35, 4.71it/s] 44%|████▍ | 21277/48008 [1:12:53<1:32:51, 4.80it/s] 44%|████▍ | 21278/48008 [1:12:53<1:30:29, 4.92it/s] 44%|████▍ | 21279/48008 [1:12:53<1:29:07, 5.00it/s] 44%|████▍ | 21280/48008 [1:12:53<1:28:06, 5.06it/s] 44%|████▍ | 21281/48008 [1:12:53<1:26:28, 5.15it/s] 44%|████▍ | 21282/48008 [1:12:54<1:26:44, 5.14it/s] 44%|████▍ | 21283/48008 [1:12:54<1:27:47, 5.07it/s] 44%|████▍ | 21284/48008 [1:12:54<1:33:34, 4.76it/s] 44%|████▍ | 21285/48008 [1:12:54<1:30:46, 4.91it/s] 44%|████▍ | 21286/48008 [1:12:54<1:29:09, 5.00it/s] 44%|████▍ | 21287/48008 [1:12:55<1:28:02, 5.06it/s] 44%|████▍ | 21288/48008 [1:12:55<1:27:15, 5.10it/s] 44%|████▍ | 21289/48008 [1:12:55<1:26:20, 5.16it/s] 44%|████▍ | 21290/48008 [1:12:55<1:26:25, 5.15it/s] 44%|████▍ | 21291/48008 [1:12:55<1:25:37, 5.20it/s] 44%|████▍ | 21292/48008 [1:12:56<1:25:07, 5.23it/s] 44%|████▍ | 21293/48008 [1:12:56<1:25:15, 5.22it/s] 44%|████▍ | 21294/48008 [1:12:56<1:25:04, 5.23it/s] 44%|████▍ | 21295/48008 [1:12:56<1:25:32, 5.21it/s] 44%|████▍ | 21296/48008 [1:12:56<1:24:53, 5.24it/s] 44%|████▍ | 21297/48008 [1:12:57<1:25:35, 5.20it/s] 44%|████▍ | 21298/48008 [1:12:57<1:25:09, 5.23it/s] 44%|████▍ | 21299/48008 [1:12:57<1:40:30, 4.43it/s] 44%|████▍ | 21300/48008 [1:12:57<1:35:40, 4.65it/s] {'loss': 0.2992, 'grad_norm': 0.33203125, 'learning_rate': 0.00011126895517413764, 'epoch': 0.44} 44%|████▍ | 21300/48008 [1:12:57<1:35:40, 4.65it/s] 44%|████▍ | 21301/48008 [1:12:58<1:58:32, 3.76it/s] 44%|████▍ | 21302/48008 [1:12:58<1:48:20, 4.11it/s] 44%|████▍ | 21303/48008 [1:12:58<1:41:33, 4.38it/s] 44%|████▍ | 21304/48008 [1:12:58<1:42:44, 4.33it/s] 44%|████▍ | 21305/48008 [1:12:58<1:38:07, 4.54it/s] 44%|████▍ | 21306/48008 [1:12:59<1:40:20, 4.44it/s] 44%|████▍ | 21307/48008 [1:12:59<1:35:25, 4.66it/s] 44%|████▍ | 21308/48008 [1:12:59<1:32:09, 4.83it/s] 44%|████▍ | 21309/48008 [1:12:59<1:30:26, 4.92it/s] 44%|████▍ | 21310/48008 [1:12:59<1:29:40, 4.96it/s] 44%|████▍ | 21311/48008 [1:13:00<1:28:45, 5.01it/s] 44%|████▍ | 21312/48008 [1:13:00<1:29:30, 4.97it/s] 44%|████▍ | 21313/48008 [1:13:00<1:28:54, 5.00it/s] 44%|████▍ | 21314/48008 [1:13:00<1:27:51, 5.06it/s] 44%|████▍ | 21315/48008 [1:13:00<1:26:50, 5.12it/s] 44%|████▍ | 21316/48008 [1:13:01<1:26:23, 5.15it/s] 44%|████▍ | 21317/48008 [1:13:01<1:26:53, 5.12it/s] 44%|████▍ | 21318/48008 [1:13:01<1:27:21, 5.09it/s] 44%|████▍ | 21319/48008 [1:13:01<1:27:01, 5.11it/s] 44%|████▍ | 21320/48008 [1:13:01<1:25:35, 5.20it/s] 44%|████▍ | 21321/48008 [1:13:02<1:26:50, 5.12it/s] 44%|████▍ | 21322/48008 [1:13:02<1:27:51, 5.06it/s] 44%|████▍ | 21323/48008 [1:13:02<1:26:06, 5.16it/s] 44%|████▍ | 21324/48008 [1:13:02<1:25:51, 5.18it/s] 44%|████▍ | 21325/48008 [1:13:02<1:25:30, 5.20it/s] 44%|████▍ | 21326/48008 [1:13:03<1:31:37, 4.85it/s] 44%|████▍ | 21327/48008 [1:13:03<1:29:58, 4.94it/s] 44%|████▍ | 21328/48008 [1:13:03<1:28:30, 5.02it/s] 44%|████▍ | 21329/48008 [1:13:03<1:27:55, 5.06it/s] 44%|████▍ | 21330/48008 [1:13:03<1:26:56, 5.11it/s] 44%|████▍ | 21331/48008 [1:13:04<1:26:21, 5.15it/s] 44%|████▍ | 21332/48008 [1:13:04<1:25:36, 5.19it/s] 44%|████▍ | 21333/48008 [1:13:04<1:27:00, 5.11it/s] 44%|████▍ | 21334/48008 [1:13:04<1:25:58, 5.17it/s] 44%|████▍ | 21335/48008 [1:13:04<1:32:00, 4.83it/s] 44%|████▍ | 21336/48008 [1:13:05<1:29:24, 4.97it/s] 44%|████▍ | 21337/48008 [1:13:05<1:27:59, 5.05it/s] 44%|████▍ | 21338/48008 [1:13:05<1:25:56, 5.17it/s] 44%|████▍ | 21339/48008 [1:13:05<1:26:15, 5.15it/s] 44%|████▍ | 21340/48008 [1:13:05<1:25:44, 5.18it/s] 44%|████▍ | 21341/48008 [1:13:06<1:25:50, 5.18it/s] 44%|████▍ | 21342/48008 [1:13:06<1:25:32, 5.20it/s] 44%|████▍ | 21343/48008 [1:13:06<1:25:11, 5.22it/s] 44%|████▍ | 21344/48008 [1:13:06<1:25:17, 5.21it/s] 44%|████▍ | 21345/48008 [1:13:06<1:31:17, 4.87it/s] 44%|████▍ | 21346/48008 [1:13:07<1:29:14, 4.98it/s] 44%|████▍ | 21347/48008 [1:13:07<1:27:59, 5.05it/s] 44%|████▍ | 21348/48008 [1:13:07<1:26:21, 5.15it/s] 44%|████▍ | 21349/48008 [1:13:07<1:26:19, 5.15it/s] 44%|████▍ | 21350/48008 [1:13:07<1:25:50, 5.18it/s] 44%|████▍ | 21350/48008 [1:13:07<1:25:50, 5.18it/s]{'loss': 0.2998, 'grad_norm': 0.140625, 'learning_rate': 0.00011106065655724047, 'epoch': 0.44} 44%|████▍ | 21351/48008 [1:13:08<1:26:12, 5.15it/s] 44%|████▍ | 21352/48008 [1:13:08<1:27:54, 5.05it/s] 44%|████▍ | 21353/48008 [1:13:08<1:28:15, 5.03it/s] 44%|████▍ | 21354/48008 [1:13:08<1:27:33, 5.07it/s] 44%|████▍ | 21355/48008 [1:13:08<1:26:37, 5.13it/s] 44%|████▍ | 21356/48008 [1:13:09<1:26:42, 5.12it/s] 44%|████▍ | 21357/48008 [1:13:09<1:25:45, 5.18it/s] 44%|████▍ | 21358/48008 [1:13:09<1:25:34, 5.19it/s] 44%|████▍ | 21359/48008 [1:13:09<1:25:32, 5.19it/s] 44%|████▍ | 21360/48008 [1:13:09<1:25:26, 5.20it/s] 44%|████▍ | 21361/48008 [1:13:09<1:25:35, 5.19it/s] 44%|████▍ | 21362/48008 [1:13:10<1:26:28, 5.14it/s] 44%|████▍ | 21363/48008 [1:13:10<1:27:44, 5.06it/s] 45%|████▍ | 21364/48008 [1:13:10<1:26:47, 5.12it/s] 45%|████▍ | 21365/48008 [1:13:10<1:26:40, 5.12it/s] 45%|████▍ | 21366/48008 [1:13:10<1:27:04, 5.10it/s] 45%|████▍ | 21367/48008 [1:13:11<1:26:17, 5.15it/s] 45%|████▍ | 21368/48008 [1:13:11<1:26:35, 5.13it/s] 45%|████▍ | 21369/48008 [1:13:11<1:26:34, 5.13it/s] 45%|████▍ | 21370/48008 [1:13:11<1:42:02, 4.35it/s] 45%|████▍ | 21371/48008 [1:13:12<1:37:37, 4.55it/s] 45%|████▍ | 21372/48008 [1:13:12<1:34:26, 4.70it/s] 45%|████▍ | 21373/48008 [1:13:12<1:31:28, 4.85it/s] 45%|████▍ | 21374/48008 [1:13:12<1:30:27, 4.91it/s] 45%|████▍ | 21375/48008 [1:13:12<1:29:50, 4.94it/s] 45%|████▍ | 21376/48008 [1:13:13<1:30:58, 4.88it/s] 45%|████▍ | 21377/48008 [1:13:13<1:29:33, 4.96it/s] 45%|████▍ | 21378/48008 [1:13:13<1:27:55, 5.05it/s] 45%|████▍ | 21379/48008 [1:13:13<1:26:56, 5.10it/s] 45%|████▍ | 21380/48008 [1:13:13<1:26:50, 5.11it/s] 45%|████▍ | 21381/48008 [1:13:14<1:26:17, 5.14it/s] 45%|████▍ | 21382/48008 [1:13:14<1:25:50, 5.17it/s] 45%|████▍ | 21383/48008 [1:13:14<1:25:54, 5.17it/s] 45%|████▍ | 21384/48008 [1:13:14<1:26:57, 5.10it/s] 45%|████▍ | 21385/48008 [1:13:14<1:26:24, 5.13it/s] 45%|████▍ | 21386/48008 [1:13:14<1:25:51, 5.17it/s] 45%|████▍ | 21387/48008 [1:13:15<1:27:03, 5.10it/s] 45%|████▍ | 21388/48008 [1:13:15<1:28:25, 5.02it/s] 45%|████▍ | 21389/48008 [1:13:15<1:27:26, 5.07it/s] 45%|████▍ | 21390/48008 [1:13:15<1:26:32, 5.13it/s] 45%|████▍ | 21391/48008 [1:13:15<1:27:41, 5.06it/s] 45%|████▍ | 21392/48008 [1:13:16<1:25:47, 5.17it/s] 45%|████▍ | 21393/48008 [1:13:16<1:26:31, 5.13it/s] 45%|████▍ | 21394/48008 [1:13:16<1:27:24, 5.07it/s] 45%|████▍ | 21395/48008 [1:13:16<1:25:55, 5.16it/s] 45%|████▍ | 21396/48008 [1:13:16<1:26:08, 5.15it/s] 45%|████▍ | 21397/48008 [1:13:17<1:26:19, 5.14it/s] 45%|████▍ | 21398/48008 [1:13:17<1:24:50, 5.23it/s] 45%|████▍ | 21399/48008 [1:13:17<1:24:42, 5.24it/s] 45%|████▍ | 21400/48008 [1:13:17<1:24:56, 5.22it/s] {'loss': 0.3013, 'grad_norm': 0.259765625, 'learning_rate': 0.00011085235794034327, 'epoch': 0.45} 45%|████▍ | 21400/48008 [1:13:17<1:24:56, 5.22it/s] 45%|████▍ | 21401/48008 [1:13:17<1:27:42, 5.06it/s] 45%|████▍ | 21402/48008 [1:13:18<1:27:14, 5.08it/s] 45%|████▍ | 21403/48008 [1:13:18<1:26:06, 5.15it/s] 45%|████▍ | 21404/48008 [1:13:18<1:25:23, 5.19it/s] 45%|████▍ | 21405/48008 [1:13:18<1:26:36, 5.12it/s] 45%|████▍ | 21406/48008 [1:13:18<1:25:54, 5.16it/s] 45%|████▍ | 21407/48008 [1:13:19<1:25:17, 5.20it/s] 45%|████▍ | 21408/48008 [1:13:19<1:26:28, 5.13it/s] 45%|████▍ | 21409/48008 [1:13:19<1:25:45, 5.17it/s] 45%|████▍ | 21410/48008 [1:13:19<1:25:37, 5.18it/s] 45%|████▍ | 21411/48008 [1:13:19<1:24:19, 5.26it/s] 45%|████▍ | 21412/48008 [1:13:20<1:23:39, 5.30it/s] 45%|████▍ | 21413/48008 [1:13:20<1:24:13, 5.26it/s] 45%|████▍ | 21414/48008 [1:13:20<1:24:24, 5.25it/s] 45%|████▍ | 21415/48008 [1:13:20<1:25:46, 5.17it/s] 45%|████▍ | 21416/48008 [1:13:20<1:25:35, 5.18it/s] 45%|████▍ | 21417/48008 [1:13:20<1:24:58, 5.22it/s] 45%|████▍ | 21418/48008 [1:13:21<1:24:43, 5.23it/s] 45%|████▍ | 21419/48008 [1:13:21<1:23:40, 5.30it/s] 45%|████▍ | 21420/48008 [1:13:21<1:23:51, 5.28it/s] 45%|████▍ | 21421/48008 [1:13:21<1:23:53, 5.28it/s] 45%|████▍ | 21422/48008 [1:13:21<1:24:00, 5.27it/s] 45%|████▍ | 21423/48008 [1:13:22<1:25:18, 5.19it/s] 45%|████▍ | 21424/48008 [1:13:22<1:26:40, 5.11it/s] 45%|████▍ | 21425/48008 [1:13:22<1:27:40, 5.05it/s] 45%|████▍ | 21426/48008 [1:13:22<1:27:19, 5.07it/s] 45%|████▍ | 21427/48008 [1:13:22<1:27:28, 5.06it/s] 45%|████▍ | 21428/48008 [1:13:23<1:26:44, 5.11it/s] 45%|████▍ | 21429/48008 [1:13:23<1:26:19, 5.13it/s] 45%|████▍ | 21430/48008 [1:13:23<1:25:21, 5.19it/s] 45%|████▍ | 21431/48008 [1:13:23<1:25:09, 5.20it/s] 45%|████▍ | 21432/48008 [1:13:23<1:25:15, 5.20it/s] 45%|████▍ | 21433/48008 [1:13:24<1:24:56, 5.21it/s] 45%|████▍ | 21434/48008 [1:13:24<1:24:56, 5.21it/s] 45%|████▍ | 21435/48008 [1:13:24<1:24:31, 5.24it/s] 45%|████▍ | 21436/48008 [1:13:24<1:24:38, 5.23it/s] 45%|████▍ | 21437/48008 [1:13:24<1:24:23, 5.25it/s] 45%|████▍ | 21438/48008 [1:13:25<1:24:03, 5.27it/s] 45%|████▍ | 21439/48008 [1:13:25<1:24:18, 5.25it/s] 45%|████▍ | 21440/48008 [1:13:25<1:25:02, 5.21it/s] 45%|████▍ | 21441/48008 [1:13:25<1:23:57, 5.27it/s] 45%|████▍ | 21442/48008 [1:13:25<1:24:16, 5.25it/s] 45%|████▍ | 21443/48008 [1:13:25<1:24:08, 5.26it/s] 45%|████▍ | 21444/48008 [1:13:26<1:24:06, 5.26it/s] 45%|████▍ | 21445/48008 [1:13:26<1:24:15, 5.25it/s] 45%|████▍ | 21446/48008 [1:13:26<1:24:13, 5.26it/s] 45%|████▍ | 21447/48008 [1:13:26<1:25:45, 5.16it/s] 45%|████▍ | 21448/48008 [1:13:27<1:51:20, 3.98it/s] 45%|████▍ | 21449/48008 [1:13:27<1:42:43, 4.31it/s] 45%|████▍ | 21450/48008 [1:13:27<1:37:08, 4.56it/s] {'loss': 0.3022, 'grad_norm': 0.1767578125, 'learning_rate': 0.0001106440593234461, 'epoch': 0.45} 45%|████▍ | 21450/48008 [1:13:27<1:37:08, 4.56it/s] 45%|████▍ | 21451/48008 [1:13:27<1:34:12, 4.70it/s] 45%|████▍ | 21452/48008 [1:13:27<1:31:35, 4.83it/s] 45%|████▍ | 21453/48008 [1:13:28<1:29:18, 4.96it/s] 45%|████▍ | 21454/48008 [1:13:28<1:27:41, 5.05it/s] 45%|████▍ | 21455/48008 [1:13:28<1:26:30, 5.12it/s] 45%|████▍ | 21456/48008 [1:13:28<1:25:15, 5.19it/s] 45%|████▍ | 21457/48008 [1:13:28<1:26:24, 5.12it/s] 45%|████▍ | 21458/48008 [1:13:29<1:25:47, 5.16it/s] 45%|████▍ | 21459/48008 [1:13:29<1:25:42, 5.16it/s] 45%|████▍ | 21460/48008 [1:13:29<1:27:32, 5.05it/s] 45%|████▍ | 21461/48008 [1:13:29<1:26:35, 5.11it/s] 45%|████▍ | 21462/48008 [1:13:29<1:26:01, 5.14it/s] 45%|████▍ | 21463/48008 [1:13:30<1:25:45, 5.16it/s] 45%|████▍ | 21464/48008 [1:13:30<1:31:40, 4.83it/s] 45%|████▍ | 21465/48008 [1:13:30<1:29:18, 4.95it/s] 45%|████▍ | 21466/48008 [1:13:30<1:42:58, 4.30it/s] 45%|████▍ | 21467/48008 [1:13:30<1:38:49, 4.48it/s] 45%|████▍ | 21468/48008 [1:13:31<1:35:50, 4.62it/s] 45%|████▍ | 21469/48008 [1:13:31<1:32:33, 4.78it/s] 45%|████▍ | 21470/48008 [1:13:31<1:29:45, 4.93it/s] 45%|████▍ | 21471/48008 [1:13:31<1:28:59, 4.97it/s] 45%|████▍ | 21472/48008 [1:13:31<1:28:27, 5.00it/s] 45%|████▍ | 21473/48008 [1:13:32<1:26:45, 5.10it/s] 45%|████▍ | 21474/48008 [1:13:32<1:25:39, 5.16it/s] 45%|████▍ | 21475/48008 [1:13:32<1:26:17, 5.13it/s] 45%|████▍ | 21476/48008 [1:13:32<1:27:49, 5.04it/s] 45%|████▍ | 21477/48008 [1:13:32<1:26:19, 5.12it/s] 45%|████▍ | 21478/48008 [1:13:33<1:25:41, 5.16it/s] 45%|████▍ | 21479/48008 [1:13:33<1:26:11, 5.13it/s] 45%|████▍ | 21480/48008 [1:13:33<1:27:38, 5.04it/s] 45%|████▍ | 21481/48008 [1:13:33<1:27:00, 5.08it/s] 45%|████▍ | 21482/48008 [1:13:33<1:27:01, 5.08it/s] 45%|████▍ | 21483/48008 [1:13:34<1:25:43, 5.16it/s] 45%|████▍ | 21484/48008 [1:13:34<1:25:08, 5.19it/s] 45%|████▍ | 21485/48008 [1:13:34<1:26:20, 5.12it/s] 45%|████▍ | 21486/48008 [1:13:34<1:25:59, 5.14it/s] 45%|████▍ | 21487/48008 [1:13:34<1:25:42, 5.16it/s] 45%|████▍ | 21488/48008 [1:13:35<1:25:21, 5.18it/s] 45%|████▍ | 21489/48008 [1:13:35<1:31:09, 4.85it/s] 45%|████▍ | 21490/48008 [1:13:35<1:35:25, 4.63it/s] 45%|████▍ | 21491/48008 [1:13:35<1:31:57, 4.81it/s] 45%|████▍ | 21492/48008 [1:13:35<1:29:31, 4.94it/s] 45%|████▍ | 21493/48008 [1:13:36<1:28:27, 5.00it/s] 45%|████▍ | 21494/48008 [1:13:36<1:33:27, 4.73it/s] 45%|████▍ | 21495/48008 [1:13:36<1:30:46, 4.87it/s] 45%|████▍ | 21496/48008 [1:13:36<1:29:36, 4.93it/s] 45%|████▍ | 21497/48008 [1:13:36<1:27:59, 5.02it/s] 45%|████▍ | 21498/48008 [1:13:37<1:26:54, 5.08it/s] 45%|████▍ | 21499/48008 [1:13:37<1:26:08, 5.13it/s] 45%|████▍ | 21500/48008 [1:13:37<1:25:42, 5.15it/s] {'loss': 0.3006, 'grad_norm': 0.1845703125, 'learning_rate': 0.00011043576070654891, 'epoch': 0.45} 45%|████▍ | 21500/48008 [1:13:37<1:25:42, 5.15it/s] 45%|████▍ | 21501/48008 [1:13:37<1:27:43, 5.04it/s] 45%|████▍ | 21502/48008 [1:13:37<1:26:46, 5.09it/s] 45%|████▍ | 21503/48008 [1:13:38<1:26:13, 5.12it/s] 45%|████▍ | 21504/48008 [1:13:38<1:25:43, 5.15it/s] 45%|████▍ | 21505/48008 [1:13:38<1:25:49, 5.15it/s] 45%|████▍ | 21506/48008 [1:13:38<1:25:03, 5.19it/s] 45%|████▍ | 21507/48008 [1:13:38<1:24:25, 5.23it/s] 45%|████▍ | 21508/48008 [1:13:39<1:25:25, 5.17it/s] 45%|████▍ | 21509/48008 [1:13:39<1:50:49, 3.99it/s] 45%|████▍ | 21510/48008 [1:13:39<1:44:21, 4.23it/s] 45%|████▍ | 21511/48008 [1:13:39<1:39:53, 4.42it/s] 45%|████▍ | 21512/48008 [1:13:40<1:35:44, 4.61it/s] 45%|████▍ | 21513/48008 [1:13:40<1:32:18, 4.78it/s] 45%|████▍ | 21514/48008 [1:13:40<1:30:43, 4.87it/s] 45%|████▍ | 21515/48008 [1:13:40<1:28:37, 4.98it/s] 45%|████▍ | 21516/48008 [1:13:40<1:26:53, 5.08it/s] 45%|████▍ | 21517/48008 [1:13:40<1:26:59, 5.07it/s] 45%|████▍ | 21518/48008 [1:13:41<1:26:58, 5.08it/s] 45%|████▍ | 21519/48008 [1:13:41<1:25:35, 5.16it/s] 45%|████▍ | 21520/48008 [1:13:41<1:24:31, 5.22it/s] 45%|████▍ | 21521/48008 [1:13:41<1:27:12, 5.06it/s] 45%|████▍ | 21522/48008 [1:13:41<1:26:38, 5.09it/s] 45%|████▍ | 21523/48008 [1:13:42<1:27:44, 5.03it/s] 45%|████▍ | 21524/48008 [1:13:42<1:27:51, 5.02it/s] 45%|████▍ | 21525/48008 [1:13:42<1:29:13, 4.95it/s] 45%|████▍ | 21526/48008 [1:13:42<1:27:56, 5.02it/s] 45%|████▍ | 21527/48008 [1:13:42<1:26:11, 5.12it/s] 45%|████▍ | 21528/48008 [1:13:43<1:25:47, 5.14it/s] 45%|████▍ | 21529/48008 [1:13:43<1:27:39, 5.03it/s] 45%|████▍ | 21530/48008 [1:13:43<1:26:36, 5.09it/s] 45%|████▍ | 21531/48008 [1:13:43<1:25:29, 5.16it/s] 45%|████▍ | 21532/48008 [1:13:43<1:26:01, 5.13it/s] 45%|████▍ | 21533/48008 [1:13:44<1:26:53, 5.08it/s] 45%|████▍ | 21534/48008 [1:13:44<1:26:01, 5.13it/s] 45%|████▍ | 21535/48008 [1:13:44<1:25:34, 5.16it/s] 45%|████▍ | 21536/48008 [1:13:44<1:25:24, 5.17it/s] 45%|████▍ | 21537/48008 [1:13:44<1:24:57, 5.19it/s] 45%|████▍ | 21538/48008 [1:13:45<1:24:49, 5.20it/s] 45%|████▍ | 21539/48008 [1:13:45<1:24:48, 5.20it/s] 45%|████▍ | 21540/48008 [1:13:45<1:24:40, 5.21it/s] 45%|████▍ | 21541/48008 [1:13:45<1:24:41, 5.21it/s] 45%|████▍ | 21542/48008 [1:13:45<1:24:26, 5.22it/s] 45%|████▍ | 21543/48008 [1:13:46<1:24:28, 5.22it/s] 45%|████▍ | 21544/48008 [1:13:46<1:25:58, 5.13it/s] 45%|████▍ | 21545/48008 [1:13:46<1:31:32, 4.82it/s] 45%|████▍ | 21546/48008 [1:13:46<1:29:15, 4.94it/s] 45%|████▍ | 21547/48008 [1:13:46<1:28:12, 5.00it/s] 45%|████▍ | 21548/48008 [1:13:47<1:33:09, 4.73it/s] 45%|████▍ | 21549/48008 [1:13:47<1:29:46, 4.91it/s] 45%|████▍ | 21550/48008 [1:13:47<1:34:23, 4.67it/s] {'loss': 0.301, 'grad_norm': 0.1396484375, 'learning_rate': 0.00011022746208965174, 'epoch': 0.45} 45%|████▍ | 21550/48008 [1:13:47<1:34:23, 4.67it/s] 45%|████▍ | 21551/48008 [1:13:47<1:33:20, 4.72it/s] 45%|████▍ | 21552/48008 [1:13:47<1:30:38, 4.86it/s] 45%|████▍ | 21553/48008 [1:13:48<1:27:36, 5.03it/s] 45%|████▍ | 21554/48008 [1:13:48<1:29:20, 4.93it/s] 45%|████▍ | 21555/48008 [1:13:48<1:28:52, 4.96it/s] 45%|████▍ | 21556/48008 [1:13:48<1:28:03, 5.01it/s] 45%|████▍ | 21557/48008 [1:13:48<1:27:05, 5.06it/s] 45%|████▍ | 21558/48008 [1:13:49<1:27:07, 5.06it/s] 45%|████▍ | 21559/48008 [1:13:49<1:28:04, 5.00it/s] 45%|████▍ | 21560/48008 [1:13:49<1:29:01, 4.95it/s] 45%|████▍ | 21561/48008 [1:13:49<1:29:11, 4.94it/s] 45%|████▍ | 21562/48008 [1:13:49<1:27:55, 5.01it/s] 45%|████▍ | 21563/48008 [1:13:50<1:26:30, 5.09it/s] 45%|████▍ | 21564/48008 [1:13:50<1:32:25, 4.77it/s] 45%|████▍ | 21565/48008 [1:13:50<1:31:42, 4.81it/s] 45%|████▍ | 21566/48008 [1:13:50<1:30:26, 4.87it/s] 45%|████▍ | 21567/48008 [1:13:50<1:30:03, 4.89it/s] 45%|████▍ | 21568/48008 [1:13:51<1:28:13, 5.00it/s] 45%|████▍ | 21569/48008 [1:13:51<1:28:55, 4.96it/s] 45%|████▍ | 21570/48008 [1:13:51<1:28:57, 4.95it/s] 45%|████▍ | 21571/48008 [1:13:51<1:27:37, 5.03it/s] 45%|████▍ | 21572/48008 [1:13:51<1:26:31, 5.09it/s] 45%|████▍ | 21573/48008 [1:13:52<1:27:05, 5.06it/s] 45%|████▍ | 21574/48008 [1:13:52<1:25:56, 5.13it/s] 45%|████▍ | 21575/48008 [1:13:52<1:28:06, 5.00it/s] 45%|████▍ | 21576/48008 [1:13:52<1:27:08, 5.05it/s] 45%|████▍ | 21577/48008 [1:13:52<1:26:40, 5.08it/s] 45%|████▍ | 21578/48008 [1:13:53<1:26:06, 5.12it/s] 45%|████▍ | 21579/48008 [1:13:53<1:41:34, 4.34it/s] 45%|████▍ | 21580/48008 [1:13:53<1:37:08, 4.53it/s] 45%|████▍ | 21581/48008 [1:13:53<1:34:45, 4.65it/s] 45%|████▍ | 21582/48008 [1:13:54<1:33:12, 4.72it/s] 45%|████▍ | 21583/48008 [1:13:54<1:33:17, 4.72it/s] 45%|████▍ | 21584/48008 [1:13:54<1:30:48, 4.85it/s] 45%|████▍ | 21585/48008 [1:13:54<1:29:30, 4.92it/s] 45%|████▍ | 21586/48008 [1:13:54<1:28:44, 4.96it/s] 45%|████▍ | 21587/48008 [1:13:55<1:29:07, 4.94it/s] 45%|████▍ | 21588/48008 [1:13:55<1:27:09, 5.05it/s] 45%|████▍ | 21589/48008 [1:13:55<1:26:29, 5.09it/s] 45%|████▍ | 21590/48008 [1:13:55<1:25:51, 5.13it/s] 45%|████▍ | 21591/48008 [1:13:55<1:25:32, 5.15it/s] 45%|████▍ | 21592/48008 [1:13:55<1:25:29, 5.15it/s] 45%|████▍ | 21593/48008 [1:13:56<1:25:26, 5.15it/s] 45%|████▍ | 21594/48008 [1:13:56<1:25:29, 5.15it/s] 45%|████▍ | 21595/48008 [1:13:56<1:25:42, 5.14it/s] 45%|████▍ | 21596/48008 [1:13:56<1:28:11, 4.99it/s] 45%|████▍ | 21597/48008 [1:13:56<1:26:46, 5.07it/s] 45%|████▍ | 21598/48008 [1:13:57<1:26:07, 5.11it/s] 45%|████▍ | 21599/48008 [1:13:57<1:25:46, 5.13it/s] 45%|████▍ | 21600/48008 [1:13:57<1:24:48, 5.19it/s] {'loss': 0.3015, 'grad_norm': 0.1494140625, 'learning_rate': 0.00011001916347275454, 'epoch': 0.45} 45%|████▍ | 21600/48008 [1:13:57<1:24:48, 5.19it/s] 45%|████▍ | 21601/48008 [1:13:57<1:24:53, 5.18it/s] 45%|████▍ | 21602/48008 [1:13:57<1:25:49, 5.13it/s] 45%|████▍ | 21603/48008 [1:13:58<1:25:58, 5.12it/s] 45%|████▌ | 21604/48008 [1:13:58<1:27:23, 5.04it/s] 45%|████▌ | 21605/48008 [1:13:58<1:26:51, 5.07it/s] 45%|████▌ | 21606/48008 [1:13:58<1:26:15, 5.10it/s] 45%|████▌ | 21607/48008 [1:13:58<1:25:43, 5.13it/s] 45%|████▌ | 21608/48008 [1:13:59<1:25:15, 5.16it/s] 45%|████▌ | 21609/48008 [1:13:59<1:25:02, 5.17it/s] 45%|████▌ | 21610/48008 [1:13:59<1:24:52, 5.18it/s] 45%|████▌ | 21611/48008 [1:13:59<1:24:50, 5.19it/s] 45%|████▌ | 21612/48008 [1:13:59<1:25:30, 5.14it/s] 45%|████▌ | 21613/48008 [1:14:00<1:25:22, 5.15it/s] 45%|████▌ | 21614/48008 [1:14:00<1:24:53, 5.18it/s] 45%|████▌ | 21615/48008 [1:14:00<1:25:26, 5.15it/s] 45%|████▌ | 21616/48008 [1:14:00<1:25:21, 5.15it/s] 45%|████▌ | 21617/48008 [1:14:00<1:26:52, 5.06it/s] 45%|████▌ | 21618/48008 [1:14:01<1:26:31, 5.08it/s] 45%|████▌ | 21619/48008 [1:14:01<1:28:21, 4.98it/s] 45%|████▌ | 21620/48008 [1:14:01<1:27:39, 5.02it/s] 45%|████▌ | 21621/48008 [1:14:01<1:29:25, 4.92it/s] 45%|████▌ | 21622/48008 [1:14:01<1:28:22, 4.98it/s] 45%|████▌ | 21623/48008 [1:14:02<1:27:25, 5.03it/s] 45%|████▌ | 21624/48008 [1:14:02<1:52:23, 3.91it/s] 45%|████▌ | 21625/48008 [1:14:02<1:46:31, 4.13it/s] 45%|████▌ | 21626/48008 [1:14:02<1:40:18, 4.38it/s] 45%|████▌ | 21627/48008 [1:14:03<1:35:32, 4.60it/s] 45%|████▌ | 21628/48008 [1:14:03<1:32:14, 4.77it/s] 45%|████▌ | 21629/48008 [1:14:03<1:31:57, 4.78it/s] 45%|████▌ | 21630/48008 [1:14:03<1:30:13, 4.87it/s] 45%|████▌ | 21631/48008 [1:14:03<1:30:06, 4.88it/s] 45%|████▌ | 21632/48008 [1:14:04<1:27:59, 5.00it/s] 45%|████▌ | 21633/48008 [1:14:04<1:26:35, 5.08it/s] 45%|████▌ | 21634/48008 [1:14:04<1:26:27, 5.08it/s] 45%|████▌ | 21635/48008 [1:14:04<1:27:20, 5.03it/s] 45%|████▌ | 21636/48008 [1:14:04<1:25:50, 5.12it/s] 45%|████▌ | 21637/48008 [1:14:05<1:25:22, 5.15it/s] 45%|████▌ | 21638/48008 [1:14:05<1:27:10, 5.04it/s] 45%|████▌ | 21639/48008 [1:14:05<1:26:45, 5.07it/s] 45%|████▌ | 21640/48008 [1:14:05<1:25:38, 5.13it/s] 45%|████▌ | 21641/48008 [1:14:05<1:26:03, 5.11it/s] 45%|████▌ | 21642/48008 [1:14:06<1:26:33, 5.08it/s] 45%|████▌ | 21643/48008 [1:14:06<1:25:03, 5.17it/s] 45%|████▌ | 21644/48008 [1:14:06<1:25:23, 5.15it/s] 45%|████▌ | 21645/48008 [1:14:06<1:27:38, 5.01it/s] 45%|████▌ | 21646/48008 [1:14:06<1:26:45, 5.06it/s] 45%|████▌ | 21647/48008 [1:14:06<1:27:15, 5.04it/s] 45%|████▌ | 21648/48008 [1:14:07<1:26:45, 5.06it/s] 45%|████▌ | 21649/48008 [1:14:07<1:28:19, 4.97it/s] 45%|████▌ | 21650/48008 [1:14:07<1:29:06, 4.93it/s] 45%|████▌ | 21650/48008 [1:14:07<1:29:06, 4.93it/s]{'loss': 0.3013, 'grad_norm': 0.443359375, 'learning_rate': 0.00010981086485585736, 'epoch': 0.45} 45%|████▌ | 21651/48008 [1:14:08<3:47:48, 1.93it/s] 45%|████▌ | 21652/48008 [1:14:09<3:04:43, 2.38it/s] 45%|████▌ | 21653/48008 [1:14:09<2:34:58, 2.83it/s] 45%|████▌ | 21654/48008 [1:14:09<2:14:13, 3.27it/s] 45%|████▌ | 21655/48008 [1:14:09<1:59:42, 3.67it/s] 45%|████▌ | 21656/48008 [1:14:09<1:48:53, 4.03it/s] 45%|████▌ | 21657/48008 [1:14:10<2:07:16, 3.45it/s] 45%|████▌ | 21658/48008 [1:14:10<1:56:09, 3.78it/s] 45%|████▌ | 21659/48008 [1:14:10<1:47:15, 4.09it/s] 45%|████▌ | 21660/48008 [1:14:10<1:40:34, 4.37it/s] 45%|████▌ | 21661/48008 [1:14:11<1:42:20, 4.29it/s] 45%|████▌ | 21662/48008 [1:14:11<1:36:24, 4.55it/s] 45%|████▌ | 21663/48008 [1:14:11<1:47:57, 4.07it/s] 45%|████▌ | 21664/48008 [1:14:11<1:41:00, 4.35it/s] 45%|████▌ | 21665/48008 [1:14:11<1:37:34, 4.50it/s] 45%|████▌ | 21666/48008 [1:14:12<1:33:42, 4.69it/s] 45%|████▌ | 21667/48008 [1:14:12<1:33:31, 4.69it/s] 45%|████▌ | 21668/48008 [1:14:12<1:30:59, 4.82it/s] 45%|████▌ | 21669/48008 [1:14:12<1:31:39, 4.79it/s] 45%|████▌ | 21670/48008 [1:14:12<1:30:59, 4.82it/s] 45%|████▌ | 21671/48008 [1:14:13<1:28:59, 4.93it/s] 45%|████▌ | 21672/48008 [1:14:13<1:27:34, 5.01it/s] 45%|████▌ | 21673/48008 [1:14:13<1:27:02, 5.04it/s] 45%|████▌ | 21674/48008 [1:14:13<1:32:36, 4.74it/s] 45%|████▌ | 21675/48008 [1:14:14<1:36:16, 4.56it/s] 45%|████▌ | 21676/48008 [1:14:14<1:34:02, 4.67it/s] 45%|████▌ | 21677/48008 [1:14:14<1:31:54, 4.77it/s] 45%|████▌ | 21678/48008 [1:14:14<1:35:34, 4.59it/s] 45%|████▌ | 21679/48008 [1:14:14<1:32:15, 4.76it/s] 45%|████▌ | 21680/48008 [1:14:15<1:29:43, 4.89it/s] 45%|████▌ | 21681/48008 [1:14:15<1:28:38, 4.95it/s] 45%|████▌ | 21682/48008 [1:14:15<1:42:34, 4.28it/s] 45%|████▌ | 21683/48008 [1:14:15<1:38:05, 4.47it/s] 45%|████▌ | 21684/48008 [1:14:15<1:36:02, 4.57it/s] 45%|████▌ | 21685/48008 [1:14:16<1:31:58, 4.77it/s] 45%|████▌ | 21686/48008 [1:14:16<1:31:51, 4.78it/s] 45%|████▌ | 21687/48008 [1:14:16<1:36:22, 4.55it/s] 45%|████▌ | 21688/48008 [1:14:16<1:32:42, 4.73it/s] 45%|████▌ | 21689/48008 [1:14:16<1:30:21, 4.85it/s] 45%|████▌ | 21690/48008 [1:14:17<1:31:30, 4.79it/s] 45%|████▌ | 21691/48008 [1:14:17<1:29:15, 4.91it/s] 45%|████▌ | 21692/48008 [1:14:17<1:29:00, 4.93it/s] 45%|████▌ | 21693/48008 [1:14:17<1:29:54, 4.88it/s] 45%|████▌ | 21694/48008 [1:14:17<1:28:11, 4.97it/s] 45%|████▌ | 21695/48008 [1:14:18<1:27:32, 5.01it/s] 45%|████▌ | 21696/48008 [1:14:18<1:26:41, 5.06it/s] 45%|████▌ | 21697/48008 [1:14:18<1:25:12, 5.15it/s] 45%|████▌ | 21698/48008 [1:14:18<1:24:50, 5.17it/s] 45%|████▌ | 21699/48008 [1:14:18<1:24:33, 5.19it/s] 45%|████▌ | 21700/48008 [1:14:19<1:25:44, 5.11it/s] {'loss': 0.3012, 'grad_norm': 0.2041015625, 'learning_rate': 0.00010960256623896016, 'epoch': 0.45} 45%|████▌ | 21700/48008 [1:14:19<1:25:44, 5.11it/s] 45%|████▌ | 21701/48008 [1:14:19<1:25:22, 5.14it/s] 45%|████▌ | 21702/48008 [1:14:19<1:26:33, 5.06it/s] 45%|████▌ | 21703/48008 [1:14:19<1:25:23, 5.13it/s] 45%|████▌ | 21704/48008 [1:14:19<1:26:19, 5.08it/s] 45%|████▌ | 21705/48008 [1:14:20<1:25:50, 5.11it/s] 45%|████▌ | 21706/48008 [1:14:20<1:25:57, 5.10it/s] 45%|████▌ | 21707/48008 [1:14:20<1:25:28, 5.13it/s] 45%|████▌ | 21708/48008 [1:14:20<1:25:07, 5.15it/s] 45%|████▌ | 21709/48008 [1:14:20<1:26:42, 5.06it/s] 45%|████▌ | 21710/48008 [1:14:21<1:25:47, 5.11it/s] 45%|████▌ | 21711/48008 [1:14:21<1:25:44, 5.11it/s] 45%|████▌ | 21712/48008 [1:14:21<1:25:47, 5.11it/s] 45%|████▌ | 21713/48008 [1:14:21<1:27:38, 5.00it/s] 45%|████▌ | 21714/48008 [1:14:21<1:29:37, 4.89it/s] 45%|████▌ | 21715/48008 [1:14:22<1:28:24, 4.96it/s] 45%|████▌ | 21716/48008 [1:14:22<1:32:50, 4.72it/s] 45%|████▌ | 21717/48008 [1:14:22<1:29:24, 4.90it/s] 45%|████▌ | 21718/48008 [1:14:22<1:28:26, 4.95it/s] 45%|████▌ | 21719/48008 [1:14:22<1:26:48, 5.05it/s] 45%|████▌ | 21720/48008 [1:14:23<1:26:05, 5.09it/s] 45%|████▌ | 21721/48008 [1:14:23<1:25:36, 5.12it/s] 45%|████▌ | 21722/48008 [1:14:23<1:24:55, 5.16it/s] 45%|████▌ | 21723/48008 [1:14:23<1:24:59, 5.15it/s] 45%|████▌ | 21724/48008 [1:14:23<1:25:35, 5.12it/s] 45%|████▌ | 21725/48008 [1:14:24<1:25:33, 5.12it/s] 45%|████▌ | 21726/48008 [1:14:24<1:24:50, 5.16it/s] 45%|████▌ | 21727/48008 [1:14:24<1:26:23, 5.07it/s] 45%|████▌ | 21728/48008 [1:14:24<1:25:34, 5.12it/s] 45%|████▌ | 21729/48008 [1:14:24<1:27:21, 5.01it/s] 45%|████▌ | 21730/48008 [1:14:25<1:26:21, 5.07it/s] 45%|████▌ | 21731/48008 [1:14:25<1:25:31, 5.12it/s] 45%|████▌ | 21732/48008 [1:14:25<1:26:20, 5.07it/s] 45%|████▌ | 21733/48008 [1:14:25<1:25:15, 5.14it/s] 45%|████▌ | 21734/48008 [1:14:25<1:25:19, 5.13it/s] 45%|████▌ | 21735/48008 [1:14:26<1:25:02, 5.15it/s] 45%|████▌ | 21736/48008 [1:14:26<1:26:31, 5.06it/s] 45%|████▌ | 21737/48008 [1:14:26<1:27:18, 5.02it/s] 45%|████▌ | 21738/48008 [1:14:26<1:25:56, 5.09it/s] 45%|████▌ | 21739/48008 [1:14:26<1:25:46, 5.10it/s] 45%|████▌ | 21740/48008 [1:14:27<1:26:55, 5.04it/s] 45%|████▌ | 21741/48008 [1:14:27<1:26:18, 5.07it/s] 45%|████▌ | 21742/48008 [1:14:27<1:41:16, 4.32it/s] 45%|████▌ | 21743/48008 [1:14:27<1:35:26, 4.59it/s] 45%|████▌ | 21744/48008 [1:14:27<1:32:41, 4.72it/s] 45%|████▌ | 21745/48008 [1:14:28<1:31:56, 4.76it/s] 45%|████▌ | 21746/48008 [1:14:28<1:29:41, 4.88it/s] 45%|████▌ | 21747/48008 [1:14:28<1:28:00, 4.97it/s] 45%|████▌ | 21748/48008 [1:14:28<1:26:00, 5.09it/s] 45%|████▌ | 21749/48008 [1:14:28<1:26:50, 5.04it/s] 45%|████▌ | 21750/48008 [1:14:29<1:26:22, 5.07it/s] {'loss': 0.3015, 'grad_norm': 0.4453125, 'learning_rate': 0.000109394267622063, 'epoch': 0.45} 45%|████▌ | 21750/48008 [1:14:29<1:26:22, 5.07it/s] 45%|████▌ | 21751/48008 [1:14:29<1:25:20, 5.13it/s] 45%|████▌ | 21752/48008 [1:14:29<1:24:49, 5.16it/s] 45%|████▌ | 21753/48008 [1:14:29<1:26:08, 5.08it/s] 45%|████▌ | 21754/48008 [1:14:29<1:25:35, 5.11it/s] 45%|████▌ | 21755/48008 [1:14:30<1:24:25, 5.18it/s] 45%|████▌ | 21756/48008 [1:14:30<1:24:11, 5.20it/s] 45%|████▌ | 21757/48008 [1:14:30<1:24:11, 5.20it/s] 45%|████▌ | 21758/48008 [1:14:30<1:24:04, 5.20it/s] 45%|████▌ | 21759/48008 [1:14:30<1:23:55, 5.21it/s] 45%|████▌ | 21760/48008 [1:14:31<1:39:05, 4.41it/s] 45%|████▌ | 21761/48008 [1:14:31<1:34:35, 4.62it/s] 45%|████▌ | 21762/48008 [1:14:31<1:32:11, 4.75it/s] 45%|████▌ | 21763/48008 [1:14:31<1:29:48, 4.87it/s] 45%|████▌ | 21764/48008 [1:14:31<1:28:10, 4.96it/s] 45%|████▌ | 21765/48008 [1:14:32<1:26:53, 5.03it/s] 45%|████▌ | 21766/48008 [1:14:32<1:25:58, 5.09it/s] 45%|████▌ | 21767/48008 [1:14:32<1:24:40, 5.17it/s] 45%|████▌ | 21768/48008 [1:14:32<1:30:46, 4.82it/s] 45%|████▌ | 21769/48008 [1:14:32<1:28:21, 4.95it/s] 45%|████▌ | 21770/48008 [1:14:33<1:27:02, 5.02it/s] 45%|████▌ | 21771/48008 [1:14:33<1:26:01, 5.08it/s] 45%|████▌ | 21772/48008 [1:14:33<1:25:13, 5.13it/s] 45%|████▌ | 21773/48008 [1:14:33<1:24:40, 5.16it/s] 45%|████▌ | 21774/48008 [1:14:33<1:25:44, 5.10it/s] 45%|████▌ | 21775/48008 [1:14:34<1:25:02, 5.14it/s] 45%|████▌ | 21776/48008 [1:14:34<1:24:13, 5.19it/s] 45%|████▌ | 21777/48008 [1:14:34<1:25:10, 5.13it/s] 45%|████▌ | 21778/48008 [1:14:34<1:24:37, 5.17it/s] 45%|████▌ | 21779/48008 [1:14:34<1:23:55, 5.21it/s] 45%|████▌ | 21780/48008 [1:14:35<1:24:11, 5.19it/s] 45%|████▌ | 21781/48008 [1:14:35<1:25:07, 5.14it/s] 45%|████▌ | 21782/48008 [1:14:35<1:24:27, 5.18it/s] 45%|████▌ | 21783/48008 [1:14:35<1:25:51, 5.09it/s] 45%|████▌ | 21784/48008 [1:14:35<1:25:19, 5.12it/s] 45%|████▌ | 21785/48008 [1:14:36<1:30:44, 4.82it/s] 45%|████▌ | 21786/48008 [1:14:36<1:30:02, 4.85it/s] 45%|████▌ | 21787/48008 [1:14:36<1:29:22, 4.89it/s] 45%|████▌ | 21788/48008 [1:14:36<1:27:26, 5.00it/s] 45%|████▌ | 21789/48008 [1:14:36<1:32:24, 4.73it/s] 45%|████▌ | 21790/48008 [1:14:37<1:31:45, 4.76it/s] 45%|████▌ | 21791/48008 [1:14:37<1:30:10, 4.85it/s] 45%|████▌ | 21792/48008 [1:14:37<1:28:23, 4.94it/s] 45%|████▌ | 21793/48008 [1:14:37<1:27:11, 5.01it/s] 45%|████▌ | 21794/48008 [1:14:37<1:32:14, 4.74it/s] 45%|████▌ | 21795/48008 [1:14:38<1:30:11, 4.84it/s] 45%|████▌ | 21796/48008 [1:14:38<1:29:57, 4.86it/s] 45%|████▌ | 21797/48008 [1:14:38<1:27:20, 5.00it/s] 45%|████▌ | 21798/48008 [1:14:38<1:26:01, 5.08it/s] 45%|████▌ | 21799/48008 [1:14:38<1:25:48, 5.09it/s] 45%|████▌ | 21800/48008 [1:14:39<1:27:04, 5.02it/s] {'loss': 0.2995, 'grad_norm': 0.177734375, 'learning_rate': 0.0001091859690051658, 'epoch': 0.45} 45%|████▌ | 21800/48008 [1:14:39<1:27:04, 5.02it/s] 45%|████▌ | 21801/48008 [1:14:39<1:25:31, 5.11it/s] 45%|████▌ | 21802/48008 [1:14:39<1:30:56, 4.80it/s] 45%|████▌ | 21803/48008 [1:14:39<1:28:06, 4.96it/s] 45%|████▌ | 21804/48008 [1:14:39<1:28:47, 4.92it/s] 45%|████▌ | 21805/48008 [1:14:40<1:27:36, 4.98it/s] 45%|████▌ | 21806/48008 [1:14:40<1:28:02, 4.96it/s] 45%|████▌ | 21807/48008 [1:14:40<1:26:27, 5.05it/s] 45%|████▌ | 21808/48008 [1:14:40<1:27:20, 5.00it/s] 45%|████▌ | 21809/48008 [1:14:40<1:26:56, 5.02it/s] 45%|████▌ | 21810/48008 [1:14:41<1:27:41, 4.98it/s] 45%|████▌ | 21811/48008 [1:14:41<1:26:41, 5.04it/s] 45%|████▌ | 21812/48008 [1:14:41<1:28:45, 4.92it/s] 45%|████▌ | 21813/48008 [1:14:41<1:27:57, 4.96it/s] 45%|████▌ | 21814/48008 [1:14:41<1:26:25, 5.05it/s] 45%|████▌ | 21815/48008 [1:14:42<1:25:27, 5.11it/s] 45%|████▌ | 21816/48008 [1:14:42<1:25:09, 5.13it/s] 45%|████▌ | 21817/48008 [1:14:42<1:27:05, 5.01it/s] 45%|████▌ | 21818/48008 [1:14:42<1:26:15, 5.06it/s] 45%|████▌ | 21819/48008 [1:14:42<1:27:00, 5.02it/s] 45%|████▌ | 21820/48008 [1:14:43<3:09:53, 2.30it/s] 45%|████▌ | 21821/48008 [1:14:44<2:39:50, 2.73it/s] 45%|████▌ | 21822/48008 [1:14:44<2:19:01, 3.14it/s] 45%|████▌ | 21823/48008 [1:14:44<2:08:36, 3.39it/s] 45%|████▌ | 21824/48008 [1:14:44<1:56:21, 3.75it/s] 45%|████▌ | 21825/48008 [1:14:44<1:48:16, 4.03it/s] 45%|████▌ | 21826/48008 [1:14:45<3:13:01, 2.26it/s] 45%|████▌ | 21827/48008 [1:14:46<2:39:52, 2.73it/s] 45%|████▌ | 21828/48008 [1:14:46<2:18:16, 3.16it/s] 45%|████▌ | 21829/48008 [1:14:46<2:08:24, 3.40it/s] 45%|████▌ | 21830/48008 [1:14:46<1:55:29, 3.78it/s] 45%|████▌ | 21831/48008 [1:14:46<1:45:58, 4.12it/s] 45%|████▌ | 21832/48008 [1:14:47<1:40:25, 4.34it/s] 45%|████▌ | 21833/48008 [1:14:47<1:36:58, 4.50it/s] 45%|████▌ | 21834/48008 [1:14:47<1:34:08, 4.63it/s] 45%|████▌ | 21835/48008 [1:14:47<1:31:30, 4.77it/s] 45%|████▌ | 21836/48008 [1:14:48<1:54:32, 3.81it/s] 45%|████▌ | 21837/48008 [1:14:48<1:51:16, 3.92it/s] 45%|████▌ | 21838/48008 [1:14:48<1:43:03, 4.23it/s] 45%|████▌ | 21839/48008 [1:14:48<1:38:10, 4.44it/s] 45%|████▌ | 21840/48008 [1:14:48<1:33:46, 4.65it/s] 45%|████▌ | 21841/48008 [1:14:49<1:32:08, 4.73it/s] 45%|████▌ | 21842/48008 [1:14:49<1:31:15, 4.78it/s] 45%|████▌ | 21843/48008 [1:14:49<1:30:48, 4.80it/s] 46%|████▌ | 21844/48008 [1:14:49<1:28:38, 4.92it/s] 46%|████▌ | 21845/48008 [1:14:49<1:27:50, 4.96it/s] 46%|████▌ | 21846/48008 [1:14:50<1:26:22, 5.05it/s] 46%|████▌ | 21847/48008 [1:14:50<1:25:47, 5.08it/s] 46%|████▌ | 21848/48008 [1:14:50<1:25:12, 5.12it/s] 46%|████▌ | 21849/48008 [1:14:50<1:25:28, 5.10it/s] 46%|████▌ | 21850/48008 [1:14:50<1:25:14, 5.11it/s] {'loss': 0.2982, 'grad_norm': 0.146484375, 'learning_rate': 0.00010897767038826863, 'epoch': 0.46} 46%|████▌ | 21850/48008 [1:14:50<1:25:14, 5.11it/s] 46%|████▌ | 21851/48008 [1:14:51<1:25:25, 5.10it/s] 46%|████▌ | 21852/48008 [1:14:51<1:30:39, 4.81it/s] 46%|████▌ | 21853/48008 [1:14:51<1:27:50, 4.96it/s] 46%|████▌ | 21854/48008 [1:14:51<1:26:28, 5.04it/s] 46%|████▌ | 21855/48008 [1:14:51<1:26:02, 5.07it/s] 46%|████▌ | 21856/48008 [1:14:52<2:46:45, 2.61it/s] 46%|████▌ | 21857/48008 [1:14:52<2:28:10, 2.94it/s] 46%|████▌ | 21858/48008 [1:14:53<2:09:31, 3.36it/s] 46%|████▌ | 21859/48008 [1:14:53<2:21:44, 3.07it/s] 46%|████▌ | 21860/48008 [1:14:53<2:04:45, 3.49it/s] 46%|████▌ | 21861/48008 [1:14:53<1:52:08, 3.89it/s] 46%|████▌ | 21862/48008 [1:14:54<1:43:47, 4.20it/s] 46%|████▌ | 21863/48008 [1:14:54<1:44:00, 4.19it/s] 46%|████▌ | 21864/48008 [1:14:54<1:37:51, 4.45it/s] 46%|████▌ | 21865/48008 [1:14:54<1:34:06, 4.63it/s] 46%|████▌ | 21866/48008 [1:14:54<1:32:35, 4.71it/s] 46%|████▌ | 21867/48008 [1:14:55<1:31:32, 4.76it/s] 46%|████▌ | 21868/48008 [1:14:55<1:54:33, 3.80it/s] 46%|████▌ | 21869/48008 [1:14:55<1:45:15, 4.14it/s] 46%|████▌ | 21870/48008 [1:14:55<1:38:17, 4.43it/s] 46%|████▌ | 21871/48008 [1:14:56<1:36:02, 4.54it/s] 46%|████▌ | 21872/48008 [1:14:56<1:32:18, 4.72it/s] 46%|████▌ | 21873/48008 [1:14:56<1:29:52, 4.85it/s] 46%|████▌ | 21874/48008 [1:14:56<1:28:01, 4.95it/s] 46%|████▌ | 21875/48008 [1:14:56<1:27:08, 5.00it/s] 46%|████▌ | 21876/48008 [1:14:57<1:25:56, 5.07it/s] 46%|████▌ | 21877/48008 [1:14:57<1:25:15, 5.11it/s] 46%|████▌ | 21878/48008 [1:14:57<1:27:27, 4.98it/s] 46%|████▌ | 21879/48008 [1:14:57<1:27:59, 4.95it/s] 46%|████▌ | 21880/48008 [1:14:57<1:26:22, 5.04it/s] 46%|████▌ | 21881/48008 [1:14:58<1:25:52, 5.07it/s] 46%|████▌ | 21882/48008 [1:14:58<1:25:46, 5.08it/s] 46%|████▌ | 21883/48008 [1:14:58<1:26:25, 5.04it/s] 46%|████▌ | 21884/48008 [1:14:58<1:25:53, 5.07it/s] 46%|████▌ | 21885/48008 [1:14:58<1:24:37, 5.14it/s] 46%|████▌ | 21886/48008 [1:14:59<1:24:42, 5.14it/s] 46%|████▌ | 21887/48008 [1:14:59<1:39:35, 4.37it/s] 46%|████▌ | 21888/48008 [1:14:59<1:35:28, 4.56it/s] 46%|████▌ | 21889/48008 [1:14:59<1:31:46, 4.74it/s] 46%|████▌ | 21890/48008 [1:14:59<1:29:58, 4.84it/s] 46%|████▌ | 21891/48008 [1:15:00<1:31:03, 4.78it/s] 46%|████▌ | 21892/48008 [1:15:00<1:28:54, 4.90it/s] 46%|████▌ | 21893/48008 [1:15:00<1:27:15, 4.99it/s] 46%|████▌ | 21894/48008 [1:15:00<1:26:19, 5.04it/s] 46%|████▌ | 21895/48008 [1:15:00<1:24:58, 5.12it/s] 46%|████▌ | 21896/48008 [1:15:01<1:25:39, 5.08it/s] 46%|████▌ | 21897/48008 [1:15:01<1:25:10, 5.11it/s] 46%|████▌ | 21898/48008 [1:15:01<1:25:05, 5.11it/s] 46%|████▌ | 21899/48008 [1:15:01<1:25:08, 5.11it/s] 46%|████▌ | 21900/48008 [1:15:01<1:26:09, 5.05it/s] 46%|████▌ | 21900/48008 [1:15:01<1:26:09, 5.05it/s]{'loss': 0.2994, 'grad_norm': 0.1669921875, 'learning_rate': 0.00010876937177137145, 'epoch': 0.46} 46%|████▌ | 21901/48008 [1:15:02<1:25:29, 5.09it/s] 46%|████▌ | 21902/48008 [1:15:02<1:24:23, 5.16it/s] 46%|████▌ | 21903/48008 [1:15:02<1:24:27, 5.15it/s] 46%|████▌ | 21904/48008 [1:15:02<1:24:08, 5.17it/s] 46%|████▌ | 21905/48008 [1:15:02<1:24:34, 5.14it/s] 46%|████▌ | 21906/48008 [1:15:03<1:24:27, 5.15it/s] 46%|████▌ | 21907/48008 [1:15:03<1:25:13, 5.10it/s] 46%|████▌ | 21908/48008 [1:15:03<1:24:55, 5.12it/s] 46%|████▌ | 21909/48008 [1:15:03<1:26:20, 5.04it/s] 46%|████▌ | 21910/48008 [1:15:03<1:25:56, 5.06it/s] 46%|████▌ | 21911/48008 [1:15:04<1:26:49, 5.01it/s] 46%|████▌ | 21912/48008 [1:15:04<1:26:32, 5.03it/s] 46%|████▌ | 21913/48008 [1:15:04<1:25:55, 5.06it/s] 46%|████▌ | 21914/48008 [1:15:04<1:25:24, 5.09it/s] 46%|████▌ | 21915/48008 [1:15:04<1:25:56, 5.06it/s] 46%|████▌ | 21916/48008 [1:15:05<1:27:39, 4.96it/s] 46%|████▌ | 21917/48008 [1:15:05<1:51:44, 3.89it/s] 46%|████▌ | 21918/48008 [1:15:05<1:44:08, 4.18it/s] 46%|████▌ | 21919/48008 [1:15:05<1:38:02, 4.44it/s] 46%|████▌ | 21920/48008 [1:15:05<1:34:04, 4.62it/s] 46%|████▌ | 21921/48008 [1:15:06<1:30:40, 4.79it/s] 46%|████▌ | 21922/48008 [1:15:06<1:28:48, 4.90it/s] 46%|████▌ | 21923/48008 [1:15:06<1:27:30, 4.97it/s] 46%|████▌ | 21924/48008 [1:15:06<1:26:57, 5.00it/s] 46%|████▌ | 21925/48008 [1:15:06<1:25:30, 5.08it/s] 46%|████▌ | 21926/48008 [1:15:07<1:24:56, 5.12it/s] 46%|████▌ | 21927/48008 [1:15:07<1:26:59, 5.00it/s] 46%|████▌ | 21928/48008 [1:15:07<1:26:09, 5.05it/s] 46%|████▌ | 21929/48008 [1:15:07<1:24:59, 5.11it/s] 46%|████▌ | 21930/48008 [1:15:07<1:24:28, 5.14it/s] 46%|████▌ | 21931/48008 [1:15:08<1:23:48, 5.19it/s] 46%|████▌ | 21932/48008 [1:15:08<1:24:06, 5.17it/s] 46%|████▌ | 21933/48008 [1:15:08<1:23:55, 5.18it/s] 46%|████▌ | 21934/48008 [1:15:08<1:23:20, 5.21it/s] 46%|████▌ | 21935/48008 [1:15:08<1:23:16, 5.22it/s] 46%|████▌ | 21936/48008 [1:15:09<1:25:20, 5.09it/s] 46%|████▌ | 21937/48008 [1:15:09<1:50:34, 3.93it/s] 46%|████▌ | 21938/48008 [1:15:09<1:42:15, 4.25it/s] 46%|████▌ | 21939/48008 [1:15:09<1:36:50, 4.49it/s] 46%|████▌ | 21940/48008 [1:15:10<1:32:29, 4.70it/s] 46%|████▌ | 21941/48008 [1:15:10<1:35:56, 4.53it/s] 46%|████▌ | 21942/48008 [1:15:10<1:38:48, 4.40it/s] 46%|████▌ | 21943/48008 [1:15:10<1:34:55, 4.58it/s] 46%|████▌ | 21944/48008 [1:15:10<1:32:30, 4.70it/s] 46%|████▌ | 21945/48008 [1:15:11<1:55:27, 3.76it/s] 46%|████▌ | 21946/48008 [1:15:11<1:46:02, 4.10it/s] 46%|████▌ | 21947/48008 [1:15:11<1:39:19, 4.37it/s] 46%|████▌ | 21948/48008 [1:15:11<1:34:10, 4.61it/s] 46%|████▌ | 21949/48008 [1:15:12<1:31:35, 4.74it/s] 46%|████▌ | 21950/48008 [1:15:12<1:29:54, 4.83it/s] {'loss': 0.3016, 'grad_norm': 0.287109375, 'learning_rate': 0.00010856107315447425, 'epoch': 0.46} 46%|████▌ | 21950/48008 [1:15:12<1:29:54, 4.83it/s] 46%|████▌ | 21951/48008 [1:15:12<1:28:00, 4.93it/s] 46%|████▌ | 21952/48008 [1:15:12<1:27:27, 4.97it/s] 46%|████▌ | 21953/48008 [1:15:12<1:27:07, 4.98it/s] 46%|████▌ | 21954/48008 [1:15:13<1:26:28, 5.02it/s] 46%|████▌ | 21955/48008 [1:15:13<1:27:05, 4.99it/s] 46%|████▌ | 21956/48008 [1:15:13<1:25:39, 5.07it/s] 46%|████▌ | 21957/48008 [1:15:13<1:24:59, 5.11it/s] 46%|████▌ | 21958/48008 [1:15:13<1:25:54, 5.05it/s] 46%|████▌ | 21959/48008 [1:15:14<1:26:43, 5.01it/s] 46%|████▌ | 21960/48008 [1:15:14<1:25:26, 5.08it/s] 46%|████▌ | 21961/48008 [1:15:14<1:25:03, 5.10it/s] 46%|████▌ | 21962/48008 [1:15:14<1:25:04, 5.10it/s] 46%|████▌ | 21963/48008 [1:15:14<1:25:43, 5.06it/s] 46%|████▌ | 21964/48008 [1:15:15<1:50:09, 3.94it/s] 46%|████▌ | 21965/48008 [1:15:15<1:42:27, 4.24it/s] 46%|████▌ | 21966/48008 [1:15:15<1:36:18, 4.51it/s] 46%|████▌ | 21967/48008 [1:15:15<1:33:42, 4.63it/s] 46%|████▌ | 21968/48008 [1:15:16<1:30:27, 4.80it/s] 46%|████▌ | 21969/48008 [1:15:16<1:29:50, 4.83it/s] 46%|████▌ | 21970/48008 [1:15:16<1:28:00, 4.93it/s] 46%|████▌ | 21971/48008 [1:15:16<1:28:07, 4.92it/s] 46%|████▌ | 21972/48008 [1:15:16<1:26:31, 5.01it/s] 46%|████▌ | 21973/48008 [1:15:17<1:26:02, 5.04it/s] 46%|████▌ | 21974/48008 [1:15:17<1:25:01, 5.10it/s] 46%|████▌ | 21975/48008 [1:15:17<1:24:06, 5.16it/s] 46%|████▌ | 21976/48008 [1:15:17<1:23:36, 5.19it/s] 46%|████▌ | 21977/48008 [1:15:17<1:23:30, 5.20it/s] 46%|████▌ | 21978/48008 [1:15:17<1:23:35, 5.19it/s] 46%|████▌ | 21979/48008 [1:15:18<1:48:42, 3.99it/s] 46%|████▌ | 21980/48008 [1:15:18<1:40:47, 4.30it/s] 46%|████▌ | 21981/48008 [1:15:18<1:35:07, 4.56it/s] 46%|████▌ | 21982/48008 [1:15:18<1:31:19, 4.75it/s] 46%|████▌ | 21983/48008 [1:15:19<1:28:46, 4.89it/s] 46%|████▌ | 21984/48008 [1:15:19<1:27:08, 4.98it/s] 46%|████▌ | 21985/48008 [1:15:19<1:26:32, 5.01it/s] 46%|████▌ | 21986/48008 [1:15:19<1:24:48, 5.11it/s] 46%|████▌ | 21987/48008 [1:15:19<1:24:09, 5.15it/s] 46%|████▌ | 21988/48008 [1:15:20<1:23:48, 5.17it/s] 46%|████▌ | 21989/48008 [1:15:20<1:24:02, 5.16it/s] 46%|████▌ | 21990/48008 [1:15:20<1:23:07, 5.22it/s] 46%|████▌ | 21991/48008 [1:15:20<1:22:58, 5.23it/s] 46%|████▌ | 21992/48008 [1:15:20<1:24:12, 5.15it/s] 46%|████▌ | 21993/48008 [1:15:21<1:38:47, 4.39it/s] 46%|████▌ | 21994/48008 [1:15:21<1:35:04, 4.56it/s] 46%|████▌ | 21995/48008 [1:15:21<1:46:31, 4.07it/s] 46%|████▌ | 21996/48008 [1:15:21<1:38:50, 4.39it/s] 46%|████▌ | 21997/48008 [1:15:22<1:33:39, 4.63it/s] 46%|████▌ | 21998/48008 [1:15:22<1:31:01, 4.76it/s] 46%|████▌ | 21999/48008 [1:15:22<1:28:25, 4.90it/s] 46%|████▌ | 22000/48008 [1:15:22<1:27:28, 4.96it/s] {'loss': 0.2995, 'grad_norm': 0.41796875, 'learning_rate': 0.00010835277453757708, 'epoch': 0.46} 46%|████▌ | 22000/48008 [1:15:22<1:27:28, 4.96it/s] 46%|████▌ | 22001/48008 [1:15:22<1:28:40, 4.89it/s] 46%|████▌ | 22002/48008 [1:15:23<1:28:02, 4.92it/s] 46%|████▌ | 22003/48008 [1:15:23<1:51:50, 3.88it/s] 46%|████▌ | 22004/48008 [1:15:23<1:42:57, 4.21it/s] 46%|████▌ | 22005/48008 [1:15:23<1:37:49, 4.43it/s] 46%|████▌ | 22006/48008 [1:15:23<1:33:57, 4.61it/s] 46%|████▌ | 22007/48008 [1:15:24<1:30:07, 4.81it/s] 46%|████▌ | 22008/48008 [1:15:25<2:55:45, 2.47it/s] 46%|████▌ | 22009/48008 [1:15:25<2:28:35, 2.92it/s] 46%|████▌ | 22010/48008 [1:15:25<2:08:58, 3.36it/s] 46%|████▌ | 22011/48008 [1:15:25<1:55:06, 3.76it/s] 46%|████▌ | 22012/48008 [1:15:25<1:46:43, 4.06it/s] 46%|████▌ | 22013/48008 [1:15:26<1:40:55, 4.29it/s] 46%|████▌ | 22014/48008 [1:15:26<1:35:03, 4.56it/s] 46%|████▌ | 22015/48008 [1:15:26<1:33:14, 4.65it/s] 46%|████▌ | 22016/48008 [1:15:26<1:30:10, 4.80it/s] 46%|████▌ | 22017/48008 [1:15:26<1:28:06, 4.92it/s] 46%|████▌ | 22018/48008 [1:15:26<1:26:11, 5.03it/s] 46%|████▌ | 22019/48008 [1:15:27<1:25:16, 5.08it/s] 46%|████▌ | 22020/48008 [1:15:27<1:23:45, 5.17it/s] 46%|████▌ | 22021/48008 [1:15:27<1:23:31, 5.19it/s] 46%|████▌ | 22022/48008 [1:15:27<1:23:22, 5.19it/s] 46%|████▌ | 22023/48008 [1:15:27<1:25:50, 5.05it/s] 46%|████▌ | 22024/48008 [1:15:28<1:24:26, 5.13it/s] 46%|████▌ | 22025/48008 [1:15:28<1:25:21, 5.07it/s] 46%|████▌ | 22026/48008 [1:15:28<1:23:59, 5.16it/s] 46%|████▌ | 22027/48008 [1:15:28<1:23:13, 5.20it/s] 46%|████▌ | 22028/48008 [1:15:28<1:24:33, 5.12it/s] 46%|████▌ | 22029/48008 [1:15:29<1:24:05, 5.15it/s] 46%|████▌ | 22030/48008 [1:15:29<1:25:01, 5.09it/s] 46%|████▌ | 22031/48008 [1:15:29<1:49:44, 3.95it/s] 46%|████▌ | 22032/48008 [1:15:29<1:42:12, 4.24it/s] 46%|████▌ | 22033/48008 [1:15:30<1:36:10, 4.50it/s] 46%|████▌ | 22034/48008 [1:15:30<1:32:09, 4.70it/s] 46%|████▌ | 22035/48008 [1:15:30<1:35:14, 4.54it/s] 46%|████▌ | 22036/48008 [1:15:30<1:33:47, 4.62it/s] 46%|████▌ | 22037/48008 [1:15:30<1:36:16, 4.50it/s] 46%|████▌ | 22038/48008 [1:15:31<1:31:40, 4.72it/s] 46%|████▌ | 22039/48008 [1:15:31<1:29:12, 4.85it/s] 46%|████▌ | 22040/48008 [1:15:31<1:27:28, 4.95it/s] 46%|████▌ | 22041/48008 [1:15:31<1:25:26, 5.07it/s] 46%|████▌ | 22042/48008 [1:15:31<1:23:58, 5.15it/s] 46%|████▌ | 22043/48008 [1:15:32<1:23:07, 5.21it/s] 46%|████▌ | 22044/48008 [1:15:32<1:23:45, 5.17it/s] 46%|████▌ | 22045/48008 [1:15:32<1:25:03, 5.09it/s] 46%|████▌ | 22046/48008 [1:15:32<1:23:40, 5.17it/s] 46%|████▌ | 22047/48008 [1:15:32<1:22:38, 5.24it/s] 46%|████▌ | 22048/48008 [1:15:33<1:22:33, 5.24it/s] 46%|████▌ | 22049/48008 [1:15:33<1:21:32, 5.31it/s] 46%|████▌ | 22050/48008 [1:15:33<1:22:45, 5.23it/s] {'loss': 0.3007, 'grad_norm': 0.34765625, 'learning_rate': 0.00010814447592067989, 'epoch': 0.46} 46%|████▌ | 22050/48008 [1:15:33<1:22:45, 5.23it/s] 46%|████▌ | 22051/48008 [1:15:33<1:22:49, 5.22it/s] 46%|████▌ | 22052/48008 [1:15:33<1:22:23, 5.25it/s] 46%|████▌ | 22053/48008 [1:15:34<1:22:21, 5.25it/s] 46%|████▌ | 22054/48008 [1:15:34<1:23:33, 5.18it/s] 46%|████▌ | 22055/48008 [1:15:34<1:23:08, 5.20it/s] 46%|████▌ | 22056/48008 [1:15:34<1:23:05, 5.21it/s] 46%|████▌ | 22057/48008 [1:15:34<1:22:46, 5.23it/s] 46%|████▌ | 22058/48008 [1:15:34<1:23:00, 5.21it/s] 46%|████▌ | 22059/48008 [1:15:35<1:21:32, 5.30it/s] 46%|████▌ | 22060/48008 [1:15:35<1:21:17, 5.32it/s] 46%|████▌ | 22061/48008 [1:15:35<1:21:18, 5.32it/s] 46%|████▌ | 22062/48008 [1:15:35<1:21:34, 5.30it/s] 46%|████▌ | 22063/48008 [1:15:35<1:21:52, 5.28it/s] 46%|████▌ | 22064/48008 [1:15:36<1:22:03, 5.27it/s] 46%|████▌ | 22065/48008 [1:15:36<1:22:32, 5.24it/s] 46%|████▌ | 22066/48008 [1:15:36<1:22:37, 5.23it/s] 46%|████▌ | 22067/48008 [1:15:36<1:22:18, 5.25it/s] 46%|████▌ | 22068/48008 [1:15:36<1:21:12, 5.32it/s] 46%|████▌ | 22069/48008 [1:15:37<1:36:28, 4.48it/s] 46%|████▌ | 22070/48008 [1:15:37<1:32:19, 4.68it/s] 46%|████▌ | 22071/48008 [1:15:37<1:29:48, 4.81it/s] 46%|████▌ | 22072/48008 [1:15:37<1:26:59, 4.97it/s] 46%|████▌ | 22073/48008 [1:15:37<1:26:24, 5.00it/s] 46%|████▌ | 22074/48008 [1:15:38<1:25:17, 5.07it/s] 46%|████▌ | 22075/48008 [1:15:38<1:24:48, 5.10it/s] 46%|████▌ | 22076/48008 [1:15:38<1:25:12, 5.07it/s] 46%|████▌ | 22077/48008 [1:15:38<1:25:49, 5.04it/s] 46%|████▌ | 22078/48008 [1:15:38<1:25:45, 5.04it/s] 46%|████▌ | 22079/48008 [1:15:39<1:27:37, 4.93it/s] 46%|████▌ | 22080/48008 [1:15:39<1:25:21, 5.06it/s] 46%|████▌ | 22081/48008 [1:15:39<1:24:28, 5.12it/s] 46%|████▌ | 22082/48008 [1:15:39<1:23:45, 5.16it/s] 46%|████▌ | 22083/48008 [1:15:39<1:23:26, 5.18it/s] 46%|████▌ | 22084/48008 [1:15:40<1:23:09, 5.20it/s] 46%|████▌ | 22085/48008 [1:15:40<1:22:58, 5.21it/s] 46%|████▌ | 22086/48008 [1:15:40<1:22:46, 5.22it/s] 46%|████▌ | 22087/48008 [1:15:40<1:23:27, 5.18it/s] 46%|████▌ | 22088/48008 [1:15:40<1:24:37, 5.11it/s] 46%|████▌ | 22089/48008 [1:15:41<1:23:17, 5.19it/s] 46%|████▌ | 22090/48008 [1:15:41<1:22:09, 5.26it/s] 46%|████▌ | 22091/48008 [1:15:41<1:24:30, 5.11it/s] 46%|████▌ | 22092/48008 [1:15:41<1:24:52, 5.09it/s] 46%|████▌ | 22093/48008 [1:15:41<1:24:03, 5.14it/s] 46%|████▌ | 22094/48008 [1:15:42<1:23:46, 5.16it/s] 46%|████▌ | 22095/48008 [1:15:42<1:23:13, 5.19it/s] 46%|████▌ | 22096/48008 [1:15:42<1:24:05, 5.14it/s] 46%|████▌ | 22097/48008 [1:15:42<1:23:39, 5.16it/s] 46%|████▌ | 22098/48008 [1:15:42<1:23:22, 5.18it/s] 46%|████▌ | 22099/48008 [1:15:42<1:23:36, 5.16it/s] 46%|████▌ | 22100/48008 [1:15:43<1:24:31, 5.11it/s] {'loss': 0.3028, 'grad_norm': 0.19921875, 'learning_rate': 0.00010793617730378272, 'epoch': 0.46} 46%|████▌ | 22100/48008 [1:15:43<1:24:31, 5.11it/s] 46%|████▌ | 22101/48008 [1:15:43<1:23:32, 5.17it/s] 46%|████▌ | 22102/48008 [1:15:43<1:22:54, 5.21it/s] 46%|████▌ | 22103/48008 [1:15:43<1:25:17, 5.06it/s] 46%|████▌ | 22104/48008 [1:15:43<1:24:36, 5.10it/s] 46%|████▌ | 22105/48008 [1:15:44<1:23:40, 5.16it/s] 46%|████▌ | 22106/48008 [1:15:44<1:23:15, 5.19it/s] 46%|████▌ | 22107/48008 [1:15:44<1:22:56, 5.20it/s] 46%|████▌ | 22108/48008 [1:15:44<1:22:32, 5.23it/s] 46%|████▌ | 22109/48008 [1:15:44<1:22:37, 5.22it/s] 46%|████▌ | 22110/48008 [1:15:45<1:22:45, 5.22it/s] 46%|████▌ | 22111/48008 [1:15:45<1:22:33, 5.23it/s] 46%|████▌ | 22112/48008 [1:15:45<1:47:36, 4.01it/s] 46%|████▌ | 22113/48008 [1:15:45<1:40:08, 4.31it/s] 46%|████▌ | 22114/48008 [1:15:46<1:34:31, 4.57it/s] 46%|████▌ | 22115/48008 [1:15:46<1:30:56, 4.75it/s] 46%|████▌ | 22116/48008 [1:15:46<1:28:29, 4.88it/s] 46%|████▌ | 22117/48008 [1:15:46<1:26:42, 4.98it/s] 46%|████▌ | 22118/48008 [1:15:46<1:24:52, 5.08it/s] 46%|████▌ | 22119/48008 [1:15:47<1:24:38, 5.10it/s] 46%|████▌ | 22120/48008 [1:15:47<1:25:35, 5.04it/s] 46%|████▌ | 22121/48008 [1:15:47<1:39:38, 4.33it/s] 46%|████▌ | 22122/48008 [1:15:47<1:34:33, 4.56it/s] 46%|████▌ | 22123/48008 [1:15:47<1:31:08, 4.73it/s] 46%|████▌ | 22124/48008 [1:15:48<1:34:57, 4.54it/s] 46%|████▌ | 22125/48008 [1:15:48<1:31:21, 4.72it/s] 46%|████▌ | 22126/48008 [1:15:48<1:29:12, 4.84it/s] 46%|████▌ | 22127/48008 [1:15:48<1:26:32, 4.98it/s] 46%|████▌ | 22128/48008 [1:15:48<1:25:17, 5.06it/s] 46%|████▌ | 22129/48008 [1:15:49<1:24:19, 5.12it/s] 46%|████▌ | 22130/48008 [1:15:49<1:23:20, 5.17it/s] 46%|████▌ | 22131/48008 [1:15:49<1:23:08, 5.19it/s] 46%|████▌ | 22132/48008 [1:15:49<1:22:43, 5.21it/s] 46%|████▌ | 22133/48008 [1:15:49<1:21:45, 5.27it/s] 46%|████▌ | 22134/48008 [1:15:50<1:22:38, 5.22it/s] 46%|████▌ | 22135/48008 [1:15:50<1:22:28, 5.23it/s] 46%|████▌ | 22136/48008 [1:15:50<1:23:48, 5.15it/s] 46%|████▌ | 22137/48008 [1:15:50<1:24:51, 5.08it/s] 46%|████▌ | 22138/48008 [1:15:50<1:24:40, 5.09it/s] 46%|████▌ | 22139/48008 [1:15:51<1:24:07, 5.12it/s] 46%|████▌ | 22140/48008 [1:15:51<1:23:20, 5.17it/s] 46%|████▌ | 22141/48008 [1:15:51<1:24:26, 5.11it/s] 46%|████▌ | 22142/48008 [1:15:51<1:23:54, 5.14it/s] 46%|████▌ | 22143/48008 [1:15:51<1:24:49, 5.08it/s] 46%|████▌ | 22144/48008 [1:15:52<1:24:01, 5.13it/s] 46%|████▌ | 22145/48008 [1:15:52<1:23:05, 5.19it/s] 46%|████▌ | 22146/48008 [1:15:52<1:37:50, 4.41it/s] 46%|████▌ | 22147/48008 [1:15:52<1:34:14, 4.57it/s] 46%|████▌ | 22148/48008 [1:15:52<1:31:39, 4.70it/s] 46%|████▌ | 22149/48008 [1:15:53<1:28:52, 4.85it/s] 46%|████▌ | 22150/48008 [1:15:53<1:27:26, 4.93it/s] {'loss': 0.3011, 'grad_norm': 0.36328125, 'learning_rate': 0.00010772787868688552, 'epoch': 0.46} 46%|████▌ | 22150/48008 [1:15:53<1:27:26, 4.93it/s] 46%|████▌ | 22151/48008 [1:15:53<1:27:39, 4.92it/s] 46%|████▌ | 22152/48008 [1:15:53<1:25:44, 5.03it/s] 46%|████▌ | 22153/48008 [1:15:53<1:24:28, 5.10it/s] 46%|████▌ | 22154/48008 [1:15:54<1:25:23, 5.05it/s] 46%|████▌ | 22155/48008 [1:15:54<1:25:35, 5.03it/s] 46%|████▌ | 22156/48008 [1:15:54<1:24:36, 5.09it/s] 46%|████▌ | 22157/48008 [1:15:54<1:23:46, 5.14it/s] 46%|████▌ | 22158/48008 [1:15:54<1:24:17, 5.11it/s] 46%|████▌ | 22159/48008 [1:15:55<1:23:29, 5.16it/s] 46%|████▌ | 22160/48008 [1:15:55<1:38:01, 4.39it/s] 46%|████▌ | 22161/48008 [1:15:55<1:34:44, 4.55it/s] 46%|████▌ | 22162/48008 [1:15:55<1:31:38, 4.70it/s] 46%|████▌ | 22163/48008 [1:15:55<1:29:56, 4.79it/s] 46%|████▌ | 22164/48008 [1:15:56<1:29:15, 4.83it/s] 46%|████▌ | 22165/48008 [1:15:56<1:28:57, 4.84it/s] 46%|████▌ | 22166/48008 [1:15:56<1:27:01, 4.95it/s] 46%|████▌ | 22167/48008 [1:15:56<1:25:31, 5.04it/s] 46%|████▌ | 22168/48008 [1:15:56<1:24:42, 5.08it/s] 46%|████▌ | 22169/48008 [1:15:57<1:23:52, 5.13it/s] 46%|████▌ | 22170/48008 [1:15:57<1:23:22, 5.16it/s] 46%|████▌ | 22171/48008 [1:15:57<1:23:12, 5.17it/s] 46%|████▌ | 22172/48008 [1:15:57<1:24:05, 5.12it/s] 46%|████▌ | 22173/48008 [1:15:58<1:48:50, 3.96it/s] 46%|████▌ | 22174/48008 [1:15:58<1:42:35, 4.20it/s] 46%|████▌ | 22175/48008 [1:15:58<1:35:46, 4.50it/s] 46%|████▌ | 22176/48008 [1:15:58<1:31:19, 4.71it/s] 46%|████▌ | 22177/48008 [1:15:58<1:30:02, 4.78it/s] 46%|████▌ | 22178/48008 [1:15:59<1:27:43, 4.91it/s] 46%|████▌ | 22179/48008 [1:15:59<1:26:44, 4.96it/s] 46%|████▌ | 22180/48008 [1:15:59<1:26:44, 4.96it/s] 46%|████▌ | 22181/48008 [1:15:59<1:24:56, 5.07it/s] 46%|████▌ | 22182/48008 [1:15:59<1:24:02, 5.12it/s] 46%|████▌ | 22183/48008 [1:16:00<1:38:27, 4.37it/s] 46%|████▌ | 22184/48008 [1:16:00<1:32:57, 4.63it/s] 46%|████▌ | 22185/48008 [1:16:00<1:30:48, 4.74it/s] 46%|████▌ | 22186/48008 [1:16:00<1:28:10, 4.88it/s] 46%|████▌ | 22187/48008 [1:16:00<1:27:10, 4.94it/s] 46%|████▌ | 22188/48008 [1:16:01<1:25:31, 5.03it/s] 46%|████▌ | 22189/48008 [1:16:01<1:24:11, 5.11it/s] 46%|████▌ | 22190/48008 [1:16:01<1:25:48, 5.01it/s] 46%|████▌ | 22191/48008 [1:16:01<1:30:39, 4.75it/s] 46%|████▌ | 22192/48008 [1:16:01<1:29:29, 4.81it/s] 46%|████▌ | 22193/48008 [1:16:02<1:28:36, 4.86it/s] 46%|████▌ | 22194/48008 [1:16:02<1:26:42, 4.96it/s] 46%|████▌ | 22195/48008 [1:16:02<1:25:25, 5.04it/s] 46%|████▌ | 22196/48008 [1:16:02<1:30:24, 4.76it/s] 46%|████▌ | 22197/48008 [1:16:02<1:28:39, 4.85it/s] 46%|████▌ | 22198/48008 [1:16:03<1:26:44, 4.96it/s] 46%|████▌ | 22199/48008 [1:16:03<1:25:57, 5.00it/s] 46%|████▌ | 22200/48008 [1:16:03<1:24:16, 5.10it/s] {'loss': 0.2993, 'grad_norm': 0.5078125, 'learning_rate': 0.00010751958006998835, 'epoch': 0.46} 46%|████▌ | 22200/48008 [1:16:03<1:24:16, 5.10it/s] 46%|████▌ | 22201/48008 [1:16:03<1:23:52, 5.13it/s] 46%|████▌ | 22202/48008 [1:16:03<1:23:17, 5.16it/s] 46%|████▌ | 22203/48008 [1:16:04<1:29:02, 4.83it/s] 46%|████▋ | 22204/48008 [1:16:04<1:26:54, 4.95it/s] 46%|████▋ | 22205/48008 [1:16:04<1:25:15, 5.04it/s] 46%|████▋ | 22206/48008 [1:16:04<1:24:35, 5.08it/s] 46%|████▋ | 22207/48008 [1:16:04<1:25:22, 5.04it/s] 46%|████▋ | 22208/48008 [1:16:05<1:24:15, 5.10it/s] 46%|████▋ | 22209/48008 [1:16:05<1:25:13, 5.05it/s] 46%|████▋ | 22210/48008 [1:16:05<1:26:49, 4.95it/s] 46%|████▋ | 22211/48008 [1:16:05<1:25:19, 5.04it/s] 46%|████▋ | 22212/48008 [1:16:05<1:24:24, 5.09it/s] 46%|████▋ | 22213/48008 [1:16:06<1:24:18, 5.10it/s] 46%|████▋ | 22214/48008 [1:16:06<1:24:07, 5.11it/s] 46%|████▋ | 22215/48008 [1:16:06<1:23:36, 5.14it/s] 46%|████▋ | 22216/48008 [1:16:06<1:24:02, 5.12it/s] 46%|████▋ | 22217/48008 [1:16:06<1:24:24, 5.09it/s] 46%|████▋ | 22218/48008 [1:16:07<1:23:23, 5.15it/s] 46%|████▋ | 22219/48008 [1:16:07<1:48:04, 3.98it/s] 46%|████▋ | 22220/48008 [1:16:07<1:41:15, 4.24it/s] 46%|████▋ | 22221/48008 [1:16:07<1:35:43, 4.49it/s] 46%|████▋ | 22222/48008 [1:16:08<1:31:37, 4.69it/s] 46%|████▋ | 22223/48008 [1:16:08<1:29:06, 4.82it/s] 46%|████▋ | 22224/48008 [1:16:08<1:28:12, 4.87it/s] 46%|████▋ | 22225/48008 [1:16:08<1:26:25, 4.97it/s] 46%|████▋ | 22226/48008 [1:16:08<1:26:41, 4.96it/s] 46%|████▋ | 22227/48008 [1:16:09<1:26:09, 4.99it/s] 46%|████▋ | 22228/48008 [1:16:09<1:24:24, 5.09it/s] 46%|████▋ | 22229/48008 [1:16:09<1:23:06, 5.17it/s] 46%|████▋ | 22230/48008 [1:16:09<1:28:41, 4.84it/s] 46%|████▋ | 22231/48008 [1:16:09<1:26:42, 4.96it/s] 46%|████▋ | 22232/48008 [1:16:10<1:26:08, 4.99it/s] 46%|████▋ | 22233/48008 [1:16:10<1:25:49, 5.01it/s] 46%|████▋ | 22234/48008 [1:16:10<1:23:56, 5.12it/s] 46%|████▋ | 22235/48008 [1:16:10<1:22:52, 5.18it/s] 46%|████▋ | 22236/48008 [1:16:10<1:22:39, 5.20it/s] 46%|████▋ | 22237/48008 [1:16:11<1:23:58, 5.11it/s] 46%|████▋ | 22238/48008 [1:16:11<1:22:13, 5.22it/s] 46%|████▋ | 22239/48008 [1:16:11<1:22:44, 5.19it/s] 46%|████▋ | 22240/48008 [1:16:11<1:22:33, 5.20it/s] 46%|████▋ | 22241/48008 [1:16:11<1:21:50, 5.25it/s] 46%|████▋ | 22242/48008 [1:16:11<1:21:39, 5.26it/s] 46%|████▋ | 22243/48008 [1:16:12<1:21:57, 5.24it/s] 46%|████▋ | 22244/48008 [1:16:12<1:21:57, 5.24it/s] 46%|████▋ | 22245/48008 [1:16:12<1:22:22, 5.21it/s] 46%|████▋ | 22246/48008 [1:16:12<1:21:49, 5.25it/s] 46%|████▋ | 22247/48008 [1:16:12<1:22:36, 5.20it/s] 46%|████▋ | 22248/48008 [1:16:13<1:24:22, 5.09it/s] 46%|████▋ | 22249/48008 [1:16:13<1:23:18, 5.15it/s] 46%|████▋ | 22250/48008 [1:16:13<1:22:54, 5.18it/s] {'loss': 0.3007, 'grad_norm': 0.369140625, 'learning_rate': 0.00010731128145309116, 'epoch': 0.46} 46%|████▋ | 22250/48008 [1:16:13<1:22:54, 5.18it/s] 46%|████▋ | 22251/48008 [1:16:13<1:24:27, 5.08it/s] 46%|████▋ | 22252/48008 [1:16:13<1:22:33, 5.20it/s] 46%|████▋ | 22253/48008 [1:16:14<1:22:03, 5.23it/s] 46%|████▋ | 22254/48008 [1:16:14<1:22:05, 5.23it/s] 46%|████▋ | 22255/48008 [1:16:14<1:28:06, 4.87it/s] 46%|████▋ | 22256/48008 [1:16:14<1:27:48, 4.89it/s] 46%|████▋ | 22257/48008 [1:16:14<1:26:45, 4.95it/s] 46%|████▋ | 22258/48008 [1:16:15<1:25:18, 5.03it/s] 46%|████▋ | 22259/48008 [1:16:15<1:23:48, 5.12it/s] 46%|████▋ | 22260/48008 [1:16:15<1:23:02, 5.17it/s] 46%|████▋ | 22261/48008 [1:16:15<1:22:43, 5.19it/s] 46%|████▋ | 22262/48008 [1:16:15<1:24:06, 5.10it/s] 46%|████▋ | 22263/48008 [1:16:16<1:29:27, 4.80it/s] 46%|████▋ | 22264/48008 [1:16:17<3:35:19, 1.99it/s] 46%|████▋ | 22265/48008 [1:16:17<2:55:49, 2.44it/s] 46%|████▋ | 22266/48008 [1:16:17<2:28:24, 2.89it/s] 46%|████▋ | 22267/48008 [1:16:17<2:08:00, 3.35it/s] 46%|████▋ | 22268/48008 [1:16:18<1:54:49, 3.74it/s] 46%|████▋ | 22269/48008 [1:16:18<1:46:37, 4.02it/s] 46%|████▋ | 22270/48008 [1:16:18<1:39:16, 4.32it/s] 46%|████▋ | 22271/48008 [1:16:18<1:35:43, 4.48it/s] 46%|████▋ | 22272/48008 [1:16:18<1:31:16, 4.70it/s] 46%|████▋ | 22273/48008 [1:16:19<1:28:03, 4.87it/s] 46%|████▋ | 22274/48008 [1:16:19<1:26:37, 4.95it/s] 46%|████▋ | 22275/48008 [1:16:19<1:25:04, 5.04it/s] 46%|████▋ | 22276/48008 [1:16:19<1:49:08, 3.93it/s] 46%|████▋ | 22277/48008 [1:16:20<1:40:52, 4.25it/s] 46%|████▋ | 22278/48008 [1:16:20<1:35:06, 4.51it/s] 46%|████▋ | 22279/48008 [1:16:20<1:31:10, 4.70it/s] 46%|████▋ | 22280/48008 [1:16:20<1:29:39, 4.78it/s] 46%|████▋ | 22281/48008 [1:16:20<1:27:04, 4.92it/s] 46%|████▋ | 22282/48008 [1:16:21<1:31:26, 4.69it/s] 46%|████▋ | 22283/48008 [1:16:21<1:30:14, 4.75it/s] 46%|████▋ | 22284/48008 [1:16:21<1:28:43, 4.83it/s] 46%|████▋ | 22285/48008 [1:16:21<1:26:34, 4.95it/s] 46%|████▋ | 22286/48008 [1:16:21<1:25:23, 5.02it/s] 46%|████▋ | 22287/48008 [1:16:22<1:24:53, 5.05it/s] 46%|████▋ | 22288/48008 [1:16:22<1:25:02, 5.04it/s] 46%|████▋ | 22289/48008 [1:16:22<1:25:39, 5.00it/s] 46%|████▋ | 22290/48008 [1:16:22<1:23:57, 5.10it/s] 46%|████▋ | 22291/48008 [1:16:22<1:29:19, 4.80it/s] 46%|████▋ | 22292/48008 [1:16:23<1:26:33, 4.95it/s] 46%|████▋ | 22293/48008 [1:16:23<1:25:42, 5.00it/s] 46%|████▋ | 22294/48008 [1:16:23<1:24:28, 5.07it/s] 46%|████▋ | 22295/48008 [1:16:23<1:22:29, 5.19it/s] 46%|████▋ | 22296/48008 [1:16:23<1:21:46, 5.24it/s] 46%|████▋ | 22297/48008 [1:16:24<1:46:54, 4.01it/s] 46%|████▋ | 22298/48008 [1:16:24<1:39:32, 4.30it/s] 46%|████▋ | 22299/48008 [1:16:24<1:34:21, 4.54it/s] 46%|████▋ | 22300/48008 [1:16:24<1:29:55, 4.76it/s] {'loss': 0.3004, 'grad_norm': 0.421875, 'learning_rate': 0.00010710298283619397, 'epoch': 0.46} 46%|████▋ | 22300/48008 [1:16:24<1:29:55, 4.76it/s] 46%|████▋ | 22301/48008 [1:16:24<1:29:15, 4.80it/s] 46%|████▋ | 22302/48008 [1:16:25<1:27:05, 4.92it/s] 46%|████▋ | 22303/48008 [1:16:25<1:26:55, 4.93it/s] 46%|████▋ | 22304/48008 [1:16:25<1:25:35, 5.00it/s] 46%|████▋ | 22305/48008 [1:16:25<1:24:11, 5.09it/s] 46%|████▋ | 22306/48008 [1:16:25<1:23:25, 5.13it/s] 46%|████▋ | 22307/48008 [1:16:26<1:23:58, 5.10it/s] 46%|████▋ | 22308/48008 [1:16:26<1:23:40, 5.12it/s] 46%|████▋ | 22309/48008 [1:16:26<1:23:59, 5.10it/s] 46%|████▋ | 22310/48008 [1:16:26<1:23:16, 5.14it/s] 46%|████▋ | 22311/48008 [1:16:26<1:23:01, 5.16it/s] 46%|████▋ | 22312/48008 [1:16:27<1:24:09, 5.09it/s] 46%|████▋ | 22313/48008 [1:16:27<1:48:46, 3.94it/s] 46%|████▋ | 22314/48008 [1:16:27<1:42:10, 4.19it/s] 46%|████▋ | 22315/48008 [1:16:27<1:37:32, 4.39it/s] 46%|████▋ | 22316/48008 [1:16:28<1:33:41, 4.57it/s] 46%|████▋ | 22317/48008 [1:16:28<1:30:27, 4.73it/s] 46%|████▋ | 22318/48008 [1:16:28<1:28:27, 4.84it/s] 46%|████▋ | 22319/48008 [1:16:28<1:26:34, 4.95it/s] 46%|████▋ | 22320/48008 [1:16:28<1:31:04, 4.70it/s] 46%|████▋ | 22321/48008 [1:16:29<1:34:29, 4.53it/s] 46%|████▋ | 22322/48008 [1:16:29<1:31:38, 4.67it/s] 46%|████▋ | 22323/48008 [1:16:29<1:29:19, 4.79it/s] 47%|████▋ | 22324/48008 [1:16:29<1:27:07, 4.91it/s] 47%|████▋ | 22325/48008 [1:16:29<1:25:44, 4.99it/s] 47%|████▋ | 22326/48008 [1:16:30<1:39:32, 4.30it/s] 47%|████▋ | 22327/48008 [1:16:30<1:35:41, 4.47it/s] 47%|████▋ | 22328/48008 [1:16:30<1:33:06, 4.60it/s] 47%|████▋ | 22329/48008 [1:16:30<1:29:41, 4.77it/s] 47%|████▋ | 22330/48008 [1:16:30<1:27:15, 4.90it/s] 47%|████▋ | 22331/48008 [1:16:31<1:26:53, 4.92it/s] 47%|████▋ | 22332/48008 [1:16:31<1:25:32, 5.00it/s] 47%|████▋ | 22333/48008 [1:16:31<1:23:56, 5.10it/s] 47%|████▋ | 22334/48008 [1:16:31<1:23:11, 5.14it/s] 47%|████▋ | 22335/48008 [1:16:31<1:23:26, 5.13it/s] 47%|████▋ | 22336/48008 [1:16:32<1:23:20, 5.13it/s] 47%|████▋ | 22337/48008 [1:16:32<1:23:51, 5.10it/s] 47%|████▋ | 22338/48008 [1:16:32<1:23:14, 5.14it/s] 47%|████▋ | 22339/48008 [1:16:32<1:23:10, 5.14it/s] 47%|████▋ | 22340/48008 [1:16:32<1:23:24, 5.13it/s] 47%|████▋ | 22341/48008 [1:16:33<1:23:21, 5.13it/s] 47%|████▋ | 22342/48008 [1:16:33<1:22:46, 5.17it/s] 47%|████▋ | 22343/48008 [1:16:33<1:24:22, 5.07it/s] 47%|████▋ | 22344/48008 [1:16:33<1:24:35, 5.06it/s] 47%|████▋ | 22345/48008 [1:16:33<1:23:35, 5.12it/s] 47%|████▋ | 22346/48008 [1:16:34<1:25:31, 5.00it/s] 47%|████▋ | 22347/48008 [1:16:34<1:25:24, 5.01it/s] 47%|████▋ | 22348/48008 [1:16:34<1:25:09, 5.02it/s] 47%|████▋ | 22349/48008 [1:16:34<1:25:27, 5.00it/s] 47%|████▋ | 22350/48008 [1:16:34<1:24:14, 5.08it/s] {'loss': 0.299, 'grad_norm': 0.1552734375, 'learning_rate': 0.00010689468421929678, 'epoch': 0.47} 47%|████▋ | 22350/48008 [1:16:34<1:24:14, 5.08it/s] 47%|████▋ | 22351/48008 [1:16:35<1:25:20, 5.01it/s] 47%|████▋ | 22352/48008 [1:16:35<1:24:31, 5.06it/s] 47%|████▋ | 22353/48008 [1:16:35<1:24:11, 5.08it/s] 47%|████▋ | 22354/48008 [1:16:35<1:23:31, 5.12it/s] 47%|████▋ | 22355/48008 [1:16:35<1:23:35, 5.11it/s] 47%|████▋ | 22356/48008 [1:16:36<1:22:55, 5.16it/s] 47%|████▋ | 22357/48008 [1:16:36<1:23:14, 5.14it/s] 47%|████▋ | 22358/48008 [1:16:36<1:23:27, 5.12it/s] 47%|████▋ | 22359/48008 [1:16:36<1:23:08, 5.14it/s] 47%|████▋ | 22360/48008 [1:16:36<1:22:43, 5.17it/s] 47%|████▋ | 22361/48008 [1:16:37<1:22:23, 5.19it/s] 47%|████▋ | 22362/48008 [1:16:37<1:21:39, 5.23it/s] 47%|████▋ | 22363/48008 [1:16:37<1:21:43, 5.23it/s] 47%|████▋ | 22364/48008 [1:16:37<1:21:24, 5.25it/s] 47%|████▋ | 22365/48008 [1:16:37<1:21:16, 5.26it/s] 47%|████▋ | 22366/48008 [1:16:38<1:23:22, 5.13it/s] 47%|████▋ | 22367/48008 [1:16:38<1:22:42, 5.17it/s] 47%|████▋ | 22368/48008 [1:16:38<1:23:18, 5.13it/s] 47%|████▋ | 22369/48008 [1:16:38<1:22:37, 5.17it/s] 47%|████▋ | 22370/48008 [1:16:38<1:22:13, 5.20it/s] 47%|████▋ | 22371/48008 [1:16:38<1:22:53, 5.15it/s] 47%|████▋ | 22372/48008 [1:16:39<1:22:42, 5.17it/s] 47%|████▋ | 22373/48008 [1:16:39<1:22:38, 5.17it/s] 47%|████▋ | 22374/48008 [1:16:39<1:21:29, 5.24it/s] 47%|████▋ | 22375/48008 [1:16:39<1:21:08, 5.27it/s] 47%|████▋ | 22376/48008 [1:16:39<1:22:10, 5.20it/s] 47%|████▋ | 22377/48008 [1:16:40<1:23:07, 5.14it/s] 47%|████▋ | 22378/48008 [1:16:40<1:22:42, 5.16it/s] 47%|████▋ | 22379/48008 [1:16:40<1:21:51, 5.22it/s] 47%|████▋ | 22380/48008 [1:16:40<1:21:59, 5.21it/s] 47%|████▋ | 22381/48008 [1:16:40<1:21:34, 5.24it/s] 47%|████▋ | 22382/48008 [1:16:41<1:36:21, 4.43it/s] 47%|████▋ | 22383/48008 [1:16:41<1:32:46, 4.60it/s] 47%|████▋ | 22384/48008 [1:16:41<1:31:09, 4.68it/s] 47%|████▋ | 22385/48008 [1:16:41<1:28:16, 4.84it/s] 47%|████▋ | 22386/48008 [1:16:42<1:28:19, 4.83it/s] 47%|████▋ | 22387/48008 [1:16:42<1:27:37, 4.87it/s] 47%|████▋ | 22388/48008 [1:16:42<1:25:58, 4.97it/s] 47%|████▋ | 22389/48008 [1:16:42<1:25:57, 4.97it/s] 47%|████▋ | 22390/48008 [1:16:42<1:24:32, 5.05it/s] 47%|████▋ | 22391/48008 [1:16:42<1:23:33, 5.11it/s] 47%|████▋ | 22392/48008 [1:16:43<1:23:01, 5.14it/s] 47%|████▋ | 22393/48008 [1:16:43<1:21:51, 5.21it/s] 47%|████▋ | 22394/48008 [1:16:43<1:20:46, 5.28it/s] 47%|████▋ | 22395/48008 [1:16:43<1:21:56, 5.21it/s] 47%|████▋ | 22396/48008 [1:16:43<1:21:57, 5.21it/s] 47%|████▋ | 22397/48008 [1:16:44<1:21:59, 5.21it/s] 47%|████▋ | 22398/48008 [1:16:44<1:21:51, 5.21it/s] 47%|████▋ | 22399/48008 [1:16:44<1:22:50, 5.15it/s] 47%|████▋ | 22400/48008 [1:16:44<1:24:49, 5.03it/s] {'loss': 0.2997, 'grad_norm': 0.30859375, 'learning_rate': 0.00010668638560239961, 'epoch': 0.47} 47%|████▋ | 22400/48008 [1:16:44<1:24:49, 5.03it/s] 47%|████▋ | 22401/48008 [1:16:44<1:23:25, 5.12it/s] 47%|████▋ | 22402/48008 [1:16:45<1:22:21, 5.18it/s] 47%|████▋ | 22403/48008 [1:16:45<1:22:10, 5.19it/s] 47%|████▋ | 22404/48008 [1:16:45<1:21:16, 5.25it/s] 47%|████▋ | 22405/48008 [1:16:45<1:22:15, 5.19it/s] 47%|████▋ | 22406/48008 [1:16:45<1:22:02, 5.20it/s] 47%|████▋ | 22407/48008 [1:16:46<1:21:34, 5.23it/s] 47%|████▋ | 22408/48008 [1:16:46<1:21:26, 5.24it/s] 47%|████▋ | 22409/48008 [1:16:46<1:21:26, 5.24it/s] 47%|████▋ | 22410/48008 [1:16:46<1:23:00, 5.14it/s] 47%|████▋ | 22411/48008 [1:16:46<1:22:32, 5.17it/s] 47%|████▋ | 22412/48008 [1:16:47<1:21:57, 5.20it/s] 47%|████▋ | 22413/48008 [1:16:47<1:22:36, 5.16it/s] 47%|████▋ | 22414/48008 [1:16:47<1:22:50, 5.15it/s] 47%|████▋ | 22415/48008 [1:16:47<1:22:29, 5.17it/s] 47%|████▋ | 22416/48008 [1:16:47<1:47:17, 3.98it/s] 47%|████▋ | 22417/48008 [1:16:48<1:42:04, 4.18it/s] 47%|████▋ | 22418/48008 [1:16:48<1:36:13, 4.43it/s] 47%|████▋ | 22419/48008 [1:16:48<1:31:54, 4.64it/s] 47%|████▋ | 22420/48008 [1:16:48<1:28:36, 4.81it/s] 47%|████▋ | 22421/48008 [1:16:48<1:26:26, 4.93it/s] 47%|████▋ | 22422/48008 [1:16:49<1:26:35, 4.92it/s] 47%|████▋ | 22423/48008 [1:16:49<1:50:01, 3.88it/s] 47%|████▋ | 22424/48008 [1:16:49<1:42:27, 4.16it/s] 47%|████▋ | 22425/48008 [1:16:49<1:38:00, 4.35it/s] 47%|████▋ | 22426/48008 [1:16:50<1:33:14, 4.57it/s] 47%|████▋ | 22427/48008 [1:16:50<1:29:34, 4.76it/s] 47%|████▋ | 22428/48008 [1:16:50<1:27:23, 4.88it/s] 47%|████▋ | 22429/48008 [1:16:50<1:25:36, 4.98it/s] 47%|████▋ | 22430/48008 [1:16:50<1:26:48, 4.91it/s] 47%|████▋ | 22431/48008 [1:16:51<1:25:18, 5.00it/s] 47%|████▋ | 22432/48008 [1:16:51<1:23:59, 5.07it/s] 47%|████▋ | 22433/48008 [1:16:52<3:05:09, 2.30it/s] 47%|████▋ | 22434/48008 [1:16:52<2:34:47, 2.75it/s] 47%|████▋ | 22435/48008 [1:16:52<2:13:47, 3.19it/s] 47%|████▋ | 22436/48008 [1:16:52<1:58:46, 3.59it/s] 47%|████▋ | 22437/48008 [1:16:53<1:47:50, 3.95it/s] 47%|████▋ | 22438/48008 [1:16:53<1:46:00, 4.02it/s] 47%|████▋ | 22439/48008 [1:16:54<3:09:52, 2.24it/s] 47%|████▋ | 22440/48008 [1:16:54<2:37:17, 2.71it/s] 47%|████▋ | 22441/48008 [1:16:54<2:14:04, 3.18it/s] 47%|████▋ | 22442/48008 [1:16:54<1:59:48, 3.56it/s] 47%|████▋ | 22443/48008 [1:16:55<1:48:21, 3.93it/s] 47%|████▋ | 22444/48008 [1:16:55<1:40:39, 4.23it/s] 47%|████▋ | 22445/48008 [1:16:55<1:35:30, 4.46it/s] 47%|████▋ | 22446/48008 [1:16:55<1:31:24, 4.66it/s] 47%|████▋ | 22447/48008 [1:16:55<1:43:18, 4.12it/s] 47%|████▋ | 22448/48008 [1:16:56<1:36:39, 4.41it/s] 47%|████▋ | 22449/48008 [1:16:56<1:31:57, 4.63it/s] 47%|████▋ | 22450/48008 [1:16:56<1:31:17, 4.67it/s] {'loss': 0.301, 'grad_norm': 0.2412109375, 'learning_rate': 0.00010647808698550241, 'epoch': 0.47} 47%|████▋ | 22450/48008 [1:16:56<1:31:17, 4.67it/s] 47%|████▋ | 22451/48008 [1:16:56<1:27:45, 4.85it/s] 47%|████▋ | 22452/48008 [1:16:56<1:26:06, 4.95it/s] 47%|████▋ | 22453/48008 [1:16:57<1:25:52, 4.96it/s] 47%|████▋ | 22454/48008 [1:16:57<1:23:55, 5.07it/s] 47%|████▋ | 22455/48008 [1:16:57<1:24:46, 5.02it/s] 47%|████▋ | 22456/48008 [1:16:57<1:23:40, 5.09it/s] 47%|████▋ | 22457/48008 [1:16:57<1:22:42, 5.15it/s] 47%|████▋ | 22458/48008 [1:16:58<1:22:25, 5.17it/s] 47%|████▋ | 22459/48008 [1:16:58<1:23:08, 5.12it/s] 47%|████▋ | 22460/48008 [1:16:58<1:22:49, 5.14it/s] 47%|████▋ | 22461/48008 [1:16:58<1:22:13, 5.18it/s] 47%|████▋ | 22462/48008 [1:16:58<1:21:19, 5.24it/s] 47%|████▋ | 22463/48008 [1:16:58<1:20:46, 5.27it/s] 47%|████▋ | 22464/48008 [1:16:59<1:20:38, 5.28it/s] 47%|████▋ | 22465/48008 [1:16:59<1:20:43, 5.27it/s] 47%|████▋ | 22466/48008 [1:16:59<1:20:52, 5.26it/s] 47%|████▋ | 22467/48008 [1:16:59<1:20:56, 5.26it/s] 47%|████▋ | 22468/48008 [1:16:59<1:20:50, 5.27it/s] 47%|████▋ | 22469/48008 [1:17:00<2:39:20, 2.67it/s] 47%|████▋ | 22470/48008 [1:17:00<2:17:42, 3.09it/s] 47%|████▋ | 22471/48008 [1:17:01<2:00:56, 3.52it/s] 47%|████▋ | 22472/48008 [1:17:01<2:14:12, 3.17it/s] 47%|████▋ | 22473/48008 [1:17:01<1:58:51, 3.58it/s] 47%|████▋ | 22474/48008 [1:17:01<1:48:11, 3.93it/s] 47%|████▋ | 22475/48008 [1:17:02<1:40:10, 4.25it/s] 47%|████▋ | 22476/48008 [1:17:02<1:35:38, 4.45it/s] 47%|████▋ | 22477/48008 [1:17:02<1:31:43, 4.64it/s] 47%|████▋ | 22478/48008 [1:17:02<1:28:10, 4.83it/s] 47%|████▋ | 22479/48008 [1:17:02<1:25:21, 4.98it/s] 47%|████▋ | 22480/48008 [1:17:03<1:24:22, 5.04it/s] 47%|████▋ | 22481/48008 [1:17:03<1:23:04, 5.12it/s] 47%|████▋ | 22482/48008 [1:17:03<1:24:51, 5.01it/s] 47%|████▋ | 22483/48008 [1:17:03<1:24:19, 5.05it/s] 47%|████▋ | 22484/48008 [1:17:03<1:23:59, 5.07it/s] 47%|████▋ | 22485/48008 [1:17:04<1:22:29, 5.16it/s] 47%|████▋ | 22486/48008 [1:17:04<1:21:59, 5.19it/s] 47%|████▋ | 22487/48008 [1:17:04<1:22:05, 5.18it/s] 47%|████▋ | 22488/48008 [1:17:04<1:22:55, 5.13it/s] 47%|████▋ | 22489/48008 [1:17:04<1:23:03, 5.12it/s] 47%|████▋ | 22490/48008 [1:17:05<1:22:16, 5.17it/s] 47%|████▋ | 22491/48008 [1:17:05<1:22:51, 5.13it/s] 47%|████▋ | 22492/48008 [1:17:05<1:22:19, 5.17it/s] 47%|████▋ | 22493/48008 [1:17:05<1:22:08, 5.18it/s] 47%|████▋ | 22494/48008 [1:17:05<1:23:27, 5.10it/s] 47%|████▋ | 22495/48008 [1:17:06<1:24:31, 5.03it/s] 47%|████▋ | 22496/48008 [1:17:06<1:29:30, 4.75it/s] 47%|████▋ | 22497/48008 [1:17:06<1:28:38, 4.80it/s] 47%|████▋ | 22498/48008 [1:17:06<1:25:41, 4.96it/s] 47%|████▋ | 22499/48008 [1:17:06<1:30:21, 4.71it/s] 47%|████▋ | 22500/48008 [1:17:07<1:27:28, 4.86it/s] 47%|████▋ | 22500/48008 [1:17:07<1:27:28, 4.86it/s]{'loss': 0.301, 'grad_norm': 0.400390625, 'learning_rate': 0.00010626978836860524, 'epoch': 0.47} 47%|████▋ | 22501/48008 [1:17:07<1:25:39, 4.96it/s] 47%|████▋ | 22502/48008 [1:17:07<1:24:26, 5.03it/s] 47%|████▋ | 22503/48008 [1:17:07<1:24:53, 5.01it/s] 47%|████▋ | 22504/48008 [1:17:07<1:25:10, 4.99it/s] 47%|████▋ | 22505/48008 [1:17:08<1:29:43, 4.74it/s] 47%|████▋ | 22506/48008 [1:17:08<1:27:03, 4.88it/s] 47%|████▋ | 22507/48008 [1:17:08<1:26:17, 4.93it/s] 47%|████▋ | 22508/48008 [1:17:08<1:24:19, 5.04it/s] 47%|████▋ | 22509/48008 [1:17:08<1:38:09, 4.33it/s] 47%|████▋ | 22510/48008 [1:17:09<1:35:28, 4.45it/s] 47%|████▋ | 22511/48008 [1:17:09<1:30:38, 4.69it/s] 47%|████▋ | 22512/48008 [1:17:09<1:27:20, 4.87it/s] 47%|████▋ | 22513/48008 [1:17:09<1:26:59, 4.88it/s] 47%|████▋ | 22514/48008 [1:17:09<1:26:33, 4.91it/s] 47%|████▋ | 22515/48008 [1:17:10<1:25:38, 4.96it/s] 47%|████▋ | 22516/48008 [1:17:10<1:23:23, 5.09it/s] 47%|████▋ | 22517/48008 [1:17:10<1:22:43, 5.14it/s] 47%|████▋ | 22518/48008 [1:17:10<1:21:47, 5.19it/s] 47%|████▋ | 22519/48008 [1:17:10<1:22:50, 5.13it/s] 47%|████▋ | 22520/48008 [1:17:11<1:36:57, 4.38it/s] 47%|████▋ | 22521/48008 [1:17:11<1:31:33, 4.64it/s] 47%|████▋ | 22522/48008 [1:17:11<1:28:32, 4.80it/s] 47%|████▋ | 22523/48008 [1:17:11<1:27:02, 4.88it/s] 47%|████▋ | 22524/48008 [1:17:11<1:25:05, 4.99it/s] 47%|████▋ | 22525/48008 [1:17:12<1:25:04, 4.99it/s] 47%|████▋ | 22526/48008 [1:17:12<1:24:45, 5.01it/s] 47%|████▋ | 22527/48008 [1:17:12<1:23:48, 5.07it/s] 47%|████▋ | 22528/48008 [1:17:12<1:23:16, 5.10it/s] 47%|████▋ | 22529/48008 [1:17:12<1:22:02, 5.18it/s] 47%|████▋ | 22530/48008 [1:17:13<1:21:49, 5.19it/s] 47%|████▋ | 22531/48008 [1:17:13<1:21:42, 5.20it/s] 47%|████▋ | 22532/48008 [1:17:13<1:21:30, 5.21it/s] 47%|████▋ | 22533/48008 [1:17:13<1:21:34, 5.20it/s] 47%|████▋ | 22534/48008 [1:17:13<1:21:33, 5.21it/s] 47%|████▋ | 22535/48008 [1:17:14<1:21:27, 5.21it/s] 47%|████▋ | 22536/48008 [1:17:14<1:21:02, 5.24it/s] 47%|████▋ | 22537/48008 [1:17:14<1:20:54, 5.25it/s] 47%|████▋ | 22538/48008 [1:17:14<1:21:45, 5.19it/s] 47%|████▋ | 22539/48008 [1:17:14<1:22:09, 5.17it/s] 47%|████▋ | 22540/48008 [1:17:15<1:23:21, 5.09it/s] 47%|████▋ | 22541/48008 [1:17:15<1:22:35, 5.14it/s] 47%|████▋ | 22542/48008 [1:17:15<1:22:09, 5.17it/s] 47%|████▋ | 22543/48008 [1:17:15<1:21:12, 5.23it/s] 47%|████▋ | 22544/48008 [1:17:15<1:21:48, 5.19it/s] 47%|████▋ | 22545/48008 [1:17:16<1:27:24, 4.86it/s] 47%|████▋ | 22546/48008 [1:17:16<1:25:22, 4.97it/s] 47%|████▋ | 22547/48008 [1:17:16<1:24:12, 5.04it/s] 47%|████▋ | 22548/48008 [1:17:16<1:24:48, 5.00it/s] 47%|████▋ | 22549/48008 [1:17:16<1:23:39, 5.07it/s] 47%|████▋ | 22550/48008 [1:17:17<1:23:01, 5.11it/s] {'loss': 0.301, 'grad_norm': 0.1572265625, 'learning_rate': 0.00010606148975170805, 'epoch': 0.47} 47%|████▋ | 22550/48008 [1:17:17<1:23:01, 5.11it/s] 47%|████▋ | 22551/48008 [1:17:17<1:28:35, 4.79it/s] 47%|████▋ | 22552/48008 [1:17:17<1:31:59, 4.61it/s] 47%|████▋ | 22553/48008 [1:17:17<1:28:43, 4.78it/s] 47%|████▋ | 22554/48008 [1:17:17<1:26:01, 4.93it/s] 47%|████▋ | 22555/48008 [1:17:18<1:24:38, 5.01it/s] 47%|████▋ | 22556/48008 [1:17:18<1:24:19, 5.03it/s] 47%|████▋ | 22557/48008 [1:17:18<1:23:25, 5.08it/s] 47%|████▋ | 22558/48008 [1:17:18<1:24:26, 5.02it/s] 47%|████▋ | 22559/48008 [1:17:18<1:22:57, 5.11it/s] 47%|████▋ | 22560/48008 [1:17:19<1:37:07, 4.37it/s] 47%|████▋ | 22561/48008 [1:17:19<1:31:25, 4.64it/s] 47%|████▋ | 22562/48008 [1:17:19<1:28:00, 4.82it/s] 47%|████▋ | 22563/48008 [1:17:19<1:27:27, 4.85it/s] 47%|████▋ | 22564/48008 [1:17:19<1:24:58, 4.99it/s] 47%|████▋ | 22565/48008 [1:17:20<1:23:38, 5.07it/s] 47%|████▋ | 22566/48008 [1:17:20<1:23:18, 5.09it/s] 47%|████▋ | 22567/48008 [1:17:20<1:22:38, 5.13it/s] 47%|████▋ | 22568/48008 [1:17:20<1:21:42, 5.19it/s] 47%|████▋ | 22569/48008 [1:17:20<1:21:43, 5.19it/s] 47%|████▋ | 22570/48008 [1:17:21<1:23:06, 5.10it/s] 47%|████▋ | 22571/48008 [1:17:21<1:22:27, 5.14it/s] 47%|████▋ | 22572/48008 [1:17:21<1:22:07, 5.16it/s] 47%|████▋ | 22573/48008 [1:17:21<1:21:48, 5.18it/s] 47%|████▋ | 22574/48008 [1:17:21<1:27:36, 4.84it/s] 47%|████▋ | 22575/48008 [1:17:22<1:25:39, 4.95it/s] 47%|████▋ | 22576/48008 [1:17:22<1:24:03, 5.04it/s] 47%|████▋ | 22577/48008 [1:17:22<1:23:01, 5.11it/s] 47%|████▋ | 22578/48008 [1:17:22<1:22:24, 5.14it/s] 47%|████▋ | 22579/48008 [1:17:22<1:21:06, 5.23it/s] 47%|████▋ | 22580/48008 [1:17:23<1:20:41, 5.25it/s] 47%|████▋ | 22581/48008 [1:17:23<1:20:21, 5.27it/s] 47%|████▋ | 22582/48008 [1:17:23<1:20:14, 5.28it/s] 47%|████▋ | 22583/48008 [1:17:23<1:20:09, 5.29it/s] 47%|████▋ | 22584/48008 [1:17:23<1:21:46, 5.18it/s] 47%|████▋ | 22585/48008 [1:17:24<1:21:35, 5.19it/s] 47%|████▋ | 22586/48008 [1:17:24<1:20:56, 5.23it/s] 47%|████▋ | 22587/48008 [1:17:24<1:23:04, 5.10it/s] 47%|████▋ | 22588/48008 [1:17:24<1:22:19, 5.15it/s] 47%|████▋ | 22589/48008 [1:17:24<1:27:41, 4.83it/s] 47%|████▋ | 22590/48008 [1:17:25<1:25:43, 4.94it/s] 47%|████▋ | 22591/48008 [1:17:25<1:24:50, 4.99it/s] 47%|████▋ | 22592/48008 [1:17:25<1:23:53, 5.05it/s] 47%|████▋ | 22593/48008 [1:17:25<1:23:42, 5.06it/s] 47%|████▋ | 22594/48008 [1:17:25<1:23:04, 5.10it/s] 47%|████▋ | 22595/48008 [1:17:26<1:23:53, 5.05it/s] 47%|████▋ | 22596/48008 [1:17:26<1:22:18, 5.15it/s] 47%|████▋ | 22597/48008 [1:17:26<1:46:39, 3.97it/s] 47%|████▋ | 22598/48008 [1:17:26<1:40:28, 4.21it/s] 47%|████▋ | 22599/48008 [1:17:26<1:36:15, 4.40it/s] 47%|████▋ | 22600/48008 [1:17:27<1:33:09, 4.55it/s] {'loss': 0.3011, 'grad_norm': 0.158203125, 'learning_rate': 0.00010585319113481086, 'epoch': 0.47} 47%|████▋ | 22600/48008 [1:17:27<1:33:09, 4.55it/s] 47%|████▋ | 22601/48008 [1:17:27<1:29:38, 4.72it/s] 47%|████▋ | 22602/48008 [1:17:27<1:27:38, 4.83it/s] 47%|████▋ | 22603/48008 [1:17:27<1:25:44, 4.94it/s] 47%|████▋ | 22604/48008 [1:17:27<1:24:12, 5.03it/s] 47%|████▋ | 22605/48008 [1:17:28<1:22:36, 5.13it/s] 47%|████▋ | 22606/48008 [1:17:28<1:23:38, 5.06it/s] 47%|████▋ | 22607/48008 [1:17:28<1:22:57, 5.10it/s] 47%|████▋ | 22608/48008 [1:17:28<1:46:53, 3.96it/s] 47%|████▋ | 22609/48008 [1:17:29<1:40:06, 4.23it/s] 47%|████▋ | 22610/48008 [1:17:29<1:34:07, 4.50it/s] 47%|████▋ | 22611/48008 [1:17:29<1:31:34, 4.62it/s] 47%|████▋ | 22612/48008 [1:17:29<1:28:09, 4.80it/s] 47%|████▋ | 22613/48008 [1:17:29<1:25:38, 4.94it/s] 47%|████▋ | 22614/48008 [1:17:30<1:23:24, 5.07it/s] 47%|████▋ | 22615/48008 [1:17:30<1:22:21, 5.14it/s] 47%|████▋ | 22616/48008 [1:17:30<1:21:57, 5.16it/s] 47%|████▋ | 22617/48008 [1:17:30<1:22:35, 5.12it/s] 47%|████▋ | 22618/48008 [1:17:30<1:21:38, 5.18it/s] 47%|████▋ | 22619/48008 [1:17:31<1:21:17, 5.21it/s] 47%|████▋ | 22620/48008 [1:17:31<1:20:45, 5.24it/s] 47%|████▋ | 22621/48008 [1:17:31<1:21:37, 5.18it/s] 47%|████▋ | 22622/48008 [1:17:31<1:22:05, 5.15it/s] 47%|████▋ | 22623/48008 [1:17:31<1:23:59, 5.04it/s] 47%|████▋ | 22624/48008 [1:17:32<1:47:51, 3.92it/s] 47%|████▋ | 22625/48008 [1:17:32<1:40:21, 4.22it/s] 47%|████▋ | 22626/48008 [1:17:32<1:33:55, 4.50it/s] 47%|████▋ | 22627/48008 [1:17:32<1:30:52, 4.66it/s] 47%|████▋ | 22628/48008 [1:17:32<1:28:07, 4.80it/s] 47%|████▋ | 22629/48008 [1:17:33<1:26:07, 4.91it/s] 47%|████▋ | 22630/48008 [1:17:33<1:25:12, 4.96it/s] 47%|████▋ | 22631/48008 [1:17:33<1:29:33, 4.72it/s] 47%|████▋ | 22632/48008 [1:17:33<1:27:08, 4.85it/s] 47%|████▋ | 22633/48008 [1:17:33<1:24:55, 4.98it/s] 47%|████▋ | 22634/48008 [1:17:34<1:38:17, 4.30it/s] 47%|████▋ | 22635/48008 [1:17:34<1:32:46, 4.56it/s] 47%|████▋ | 22636/48008 [1:17:34<1:28:48, 4.76it/s] 47%|████▋ | 22637/48008 [1:17:34<1:28:44, 4.77it/s] 47%|████▋ | 22638/48008 [1:17:35<1:25:56, 4.92it/s] 47%|████▋ | 22639/48008 [1:17:35<1:24:08, 5.02it/s] 47%|████▋ | 22640/48008 [1:17:35<1:37:57, 4.32it/s] 47%|████▋ | 22641/48008 [1:17:35<1:33:36, 4.52it/s] 47%|████▋ | 22642/48008 [1:17:35<1:29:32, 4.72it/s] 47%|████▋ | 22643/48008 [1:17:36<1:27:46, 4.82it/s] 47%|████▋ | 22644/48008 [1:17:36<1:40:18, 4.21it/s] 47%|████▋ | 22645/48008 [1:17:36<1:34:00, 4.50it/s] 47%|████▋ | 22646/48008 [1:17:36<1:29:45, 4.71it/s] 47%|████▋ | 22647/48008 [1:17:37<1:28:01, 4.80it/s] 47%|████▋ | 22648/48008 [1:17:37<1:25:25, 4.95it/s] 47%|████▋ | 22649/48008 [1:17:37<1:24:14, 5.02it/s] 47%|████▋ | 22650/48008 [1:17:37<1:22:37, 5.11it/s] {'loss': 0.2997, 'grad_norm': 0.33203125, 'learning_rate': 0.0001056448925179137, 'epoch': 0.47} 47%|████▋ | 22650/48008 [1:17:37<1:22:37, 5.11it/s] 47%|████▋ | 22651/48008 [1:17:37<1:27:51, 4.81it/s] 47%|████▋ | 22652/48008 [1:17:38<1:25:14, 4.96it/s] 47%|████▋ | 22653/48008 [1:17:38<1:23:28, 5.06it/s] 47%|████▋ | 22654/48008 [1:17:38<1:22:11, 5.14it/s] 47%|████▋ | 22655/48008 [1:17:38<1:20:59, 5.22it/s] 47%|████▋ | 22656/48008 [1:17:38<1:20:47, 5.23it/s] 47%|████▋ | 22657/48008 [1:17:38<1:21:10, 5.21it/s] 47%|████▋ | 22658/48008 [1:17:39<1:20:47, 5.23it/s] 47%|████▋ | 22659/48008 [1:17:39<1:35:27, 4.43it/s] 47%|████▋ | 22660/48008 [1:17:39<1:31:55, 4.60it/s] 47%|████▋ | 22661/48008 [1:17:39<1:28:32, 4.77it/s] 47%|████▋ | 22662/48008 [1:17:40<1:25:32, 4.94it/s] 47%|████▋ | 22663/48008 [1:17:40<1:23:49, 5.04it/s] 47%|████▋ | 22664/48008 [1:17:40<1:22:46, 5.10it/s] 47%|████▋ | 22665/48008 [1:17:40<1:21:42, 5.17it/s] 47%|████▋ | 22666/48008 [1:17:40<1:21:10, 5.20it/s] 47%|████▋ | 22667/48008 [1:17:40<1:21:12, 5.20it/s] 47%|████▋ | 22668/48008 [1:17:41<2:49:28, 2.49it/s] 47%|████▋ | 22669/48008 [1:17:42<2:23:17, 2.95it/s] 47%|████▋ | 22670/48008 [1:17:42<2:03:19, 3.42it/s] 47%|████▋ | 22671/48008 [1:17:42<1:50:34, 3.82it/s] 47%|████▋ | 22672/48008 [1:17:42<1:42:25, 4.12it/s] 47%|████▋ | 22673/48008 [1:17:42<1:41:51, 4.15it/s] 47%|████▋ | 22674/48008 [1:17:43<1:37:42, 4.32it/s] 47%|████▋ | 22675/48008 [1:17:43<1:32:04, 4.59it/s] 47%|████▋ | 22676/48008 [1:17:43<1:29:43, 4.71it/s] 47%|████▋ | 22677/48008 [1:17:43<1:26:58, 4.85it/s] 47%|████▋ | 22678/48008 [1:17:43<1:26:37, 4.87it/s] 47%|████▋ | 22679/48008 [1:17:44<1:24:14, 5.01it/s] 47%|████▋ | 22680/48008 [1:17:44<1:22:55, 5.09it/s] 47%|████▋ | 22681/48008 [1:17:44<1:22:21, 5.12it/s] 47%|████▋ | 22682/48008 [1:17:44<1:21:57, 5.15it/s] 47%|████▋ | 22683/48008 [1:17:44<1:21:11, 5.20it/s] 47%|████▋ | 22684/48008 [1:17:45<1:20:22, 5.25it/s] 47%|████▋ | 22685/48008 [1:17:45<1:20:52, 5.22it/s] 47%|████▋ | 22686/48008 [1:17:45<1:20:14, 5.26it/s] 47%|████▋ | 22687/48008 [1:17:45<1:21:03, 5.21it/s] 47%|████▋ | 22688/48008 [1:17:45<1:21:00, 5.21it/s] 47%|████▋ | 22689/48008 [1:17:45<1:20:28, 5.24it/s] 47%|████▋ | 22690/48008 [1:17:46<1:20:09, 5.26it/s] 47%|████▋ | 22691/48008 [1:17:46<1:20:44, 5.23it/s] 47%|████▋ | 22692/48008 [1:17:46<1:21:03, 5.21it/s] 47%|████▋ | 22693/48008 [1:17:46<1:22:23, 5.12it/s] 47%|████▋ | 22694/48008 [1:17:46<1:22:02, 5.14it/s] 47%|████▋ | 22695/48008 [1:17:47<1:21:27, 5.18it/s] 47%|████▋ | 22696/48008 [1:17:47<1:20:59, 5.21it/s] 47%|████▋ | 22697/48008 [1:17:47<1:45:33, 4.00it/s] 47%|████▋ | 22698/48008 [1:17:47<1:38:13, 4.29it/s] 47%|████▋ | 22699/48008 [1:17:48<1:33:04, 4.53it/s] 47%|████▋ | 22700/48008 [1:17:48<1:31:52, 4.59it/s] 47%|████▋ | 22700/48008 [1:17:48<1:31:52, 4.59it/s]{'loss': 0.3032, 'grad_norm': 0.318359375, 'learning_rate': 0.0001054365939010165, 'epoch': 0.47} 47%|████▋ | 22701/48008 [1:17:48<1:28:24, 4.77it/s] 47%|████▋ | 22702/48008 [1:17:48<1:26:04, 4.90it/s] 47%|████▋ | 22703/48008 [1:17:48<1:24:00, 5.02it/s] 47%|████▋ | 22704/48008 [1:17:49<1:23:06, 5.07it/s] 47%|████▋ | 22705/48008 [1:17:49<1:21:45, 5.16it/s] 47%|████▋ | 22706/48008 [1:17:49<1:21:16, 5.19it/s] 47%|████▋ | 22707/48008 [1:17:49<1:21:11, 5.19it/s] 47%|████▋ | 22708/48008 [1:17:49<1:21:45, 5.16it/s] 47%|████▋ | 22709/48008 [1:17:50<1:21:16, 5.19it/s] 47%|████▋ | 22710/48008 [1:17:50<1:20:59, 5.21it/s] 47%|████▋ | 22711/48008 [1:17:50<1:21:42, 5.16it/s] 47%|████▋ | 22712/48008 [1:17:50<1:21:40, 5.16it/s] 47%|████▋ | 22713/48008 [1:17:50<1:21:07, 5.20it/s] 47%|████▋ | 22714/48008 [1:17:50<1:21:21, 5.18it/s] 47%|████▋ | 22715/48008 [1:17:51<1:21:04, 5.20it/s] 47%|████▋ | 22716/48008 [1:17:51<1:22:30, 5.11it/s] 47%|████▋ | 22717/48008 [1:17:51<1:21:46, 5.15it/s] 47%|████▋ | 22718/48008 [1:17:51<1:21:00, 5.20it/s] 47%|████▋ | 22719/48008 [1:17:51<1:20:25, 5.24it/s] 47%|████▋ | 22720/48008 [1:17:52<1:20:40, 5.22it/s] 47%|████▋ | 22721/48008 [1:17:52<1:20:55, 5.21it/s] 47%|████▋ | 22722/48008 [1:17:52<1:21:53, 5.15it/s] 47%|████▋ | 22723/48008 [1:17:52<1:23:54, 5.02it/s] 47%|████▋ | 22724/48008 [1:17:52<1:23:11, 5.07it/s] 47%|████▋ | 22725/48008 [1:17:53<1:24:12, 5.00it/s] 47%|████▋ | 22726/48008 [1:17:53<1:23:28, 5.05it/s] 47%|████▋ | 22727/48008 [1:17:53<1:22:51, 5.09it/s] 47%|████▋ | 22728/48008 [1:17:53<1:23:09, 5.07it/s] 47%|████▋ | 22729/48008 [1:17:53<1:22:29, 5.11it/s] 47%|████▋ | 22730/48008 [1:17:54<1:22:14, 5.12it/s] 47%|████▋ | 22731/48008 [1:17:54<1:21:53, 5.14it/s] 47%|████▋ | 22732/48008 [1:17:54<1:21:00, 5.20it/s] 47%|████▋ | 22733/48008 [1:17:54<1:20:49, 5.21it/s] 47%|████▋ | 22734/48008 [1:17:54<1:20:45, 5.22it/s] 47%|████▋ | 22735/48008 [1:17:55<1:26:45, 4.85it/s] 47%|████▋ | 22736/48008 [1:17:55<1:24:57, 4.96it/s] 47%|████▋ | 22737/48008 [1:17:55<1:23:56, 5.02it/s] 47%|████▋ | 22738/48008 [1:17:55<1:22:54, 5.08it/s] 47%|████▋ | 22739/48008 [1:17:55<1:23:55, 5.02it/s] 47%|████▋ | 22740/48008 [1:17:56<1:23:14, 5.06it/s] 47%|████▋ | 22741/48008 [1:17:56<1:22:52, 5.08it/s] 47%|████▋ | 22742/48008 [1:17:56<1:22:32, 5.10it/s] 47%|████▋ | 22743/48008 [1:17:56<1:22:00, 5.13it/s] 47%|████▋ | 22744/48008 [1:17:56<1:23:56, 5.02it/s] 47%|████▋ | 22745/48008 [1:17:57<1:22:44, 5.09it/s] 47%|████▋ | 22746/48008 [1:17:57<1:23:54, 5.02it/s] 47%|████▋ | 22747/48008 [1:17:57<1:22:32, 5.10it/s] 47%|████▋ | 22748/48008 [1:17:57<1:27:55, 4.79it/s] 47%|████▋ | 22749/48008 [1:17:57<1:25:15, 4.94it/s] 47%|████▋ | 22750/48008 [1:17:58<1:25:11, 4.94it/s] {'loss': 0.3012, 'grad_norm': 0.19140625, 'learning_rate': 0.00010522829528411933, 'epoch': 0.47} 47%|████▋ | 22750/48008 [1:17:58<1:25:11, 4.94it/s] 47%|████▋ | 22751/48008 [1:17:58<1:23:54, 5.02it/s] 47%|████▋ | 22752/48008 [1:17:58<1:22:35, 5.10it/s] 47%|████▋ | 22753/48008 [1:17:58<1:21:03, 5.19it/s] 47%|████▋ | 22754/48008 [1:17:58<1:21:15, 5.18it/s] 47%|████▋ | 22755/48008 [1:17:59<1:21:16, 5.18it/s] 47%|████▋ | 22756/48008 [1:17:59<1:21:19, 5.18it/s] 47%|████▋ | 22757/48008 [1:17:59<1:22:58, 5.07it/s] 47%|████▋ | 22758/48008 [1:17:59<1:28:12, 4.77it/s] 47%|████▋ | 22759/48008 [1:17:59<1:26:07, 4.89it/s] 47%|████▋ | 22760/48008 [1:18:00<1:24:18, 4.99it/s] 47%|████▋ | 22761/48008 [1:18:00<1:23:20, 5.05it/s] 47%|████▋ | 22762/48008 [1:18:00<1:22:22, 5.11it/s] 47%|████▋ | 22763/48008 [1:18:00<1:23:24, 5.04it/s] 47%|████▋ | 22764/48008 [1:18:00<1:21:58, 5.13it/s] 47%|████▋ | 22765/48008 [1:18:01<1:21:02, 5.19it/s] 47%|████▋ | 22766/48008 [1:18:01<1:20:59, 5.19it/s] 47%|████▋ | 22767/48008 [1:18:01<1:20:28, 5.23it/s] 47%|████▋ | 22768/48008 [1:18:01<1:20:12, 5.24it/s] 47%|████▋ | 22769/48008 [1:18:01<1:19:54, 5.26it/s] 47%|████▋ | 22770/48008 [1:18:01<1:21:14, 5.18it/s] 47%|████▋ | 22771/48008 [1:18:02<1:20:42, 5.21it/s] 47%|████▋ | 22772/48008 [1:18:02<1:19:42, 5.28it/s] 47%|████▋ | 22773/48008 [1:18:02<1:21:23, 5.17it/s] 47%|████▋ | 22774/48008 [1:18:02<1:21:03, 5.19it/s] 47%|████▋ | 22775/48008 [1:18:02<1:21:06, 5.18it/s] 47%|████▋ | 22776/48008 [1:18:03<1:20:59, 5.19it/s] 47%|████▋ | 22777/48008 [1:18:03<1:22:09, 5.12it/s] 47%|████▋ | 22778/48008 [1:18:03<1:21:37, 5.15it/s] 47%|████▋ | 22779/48008 [1:18:03<1:21:52, 5.14it/s] 47%|████▋ | 22780/48008 [1:18:03<1:21:56, 5.13it/s] 47%|████▋ | 22781/48008 [1:18:04<1:21:50, 5.14it/s] 47%|████▋ | 22782/48008 [1:18:04<1:21:45, 5.14it/s] 47%|████▋ | 22783/48008 [1:18:04<1:20:58, 5.19it/s] 47%|████▋ | 22784/48008 [1:18:04<1:20:47, 5.20it/s] 47%|████▋ | 22785/48008 [1:18:04<1:20:19, 5.23it/s] 47%|████▋ | 22786/48008 [1:18:05<1:20:16, 5.24it/s] 47%|████▋ | 22787/48008 [1:18:05<1:22:36, 5.09it/s] 47%|████▋ | 22788/48008 [1:18:05<1:22:34, 5.09it/s] 47%|████▋ | 22789/48008 [1:18:05<1:22:48, 5.08it/s] 47%|████▋ | 22790/48008 [1:18:05<1:22:10, 5.12it/s] 47%|████▋ | 22791/48008 [1:18:06<1:21:32, 5.15it/s] 47%|████▋ | 22792/48008 [1:18:06<1:23:34, 5.03it/s] 47%|████▋ | 22793/48008 [1:18:06<1:22:05, 5.12it/s] 47%|████▋ | 22794/48008 [1:18:06<1:21:04, 5.18it/s] 47%|████▋ | 22795/48008 [1:18:07<1:45:16, 3.99it/s] 47%|████▋ | 22796/48008 [1:18:07<1:37:40, 4.30it/s] 47%|████▋ | 22797/48008 [1:18:07<1:32:20, 4.55it/s] 47%|████▋ | 22798/48008 [1:18:07<1:28:41, 4.74it/s] 47%|████▋ | 22799/48008 [1:18:07<1:26:35, 4.85it/s] 47%|████▋ | 22800/48008 [1:18:07<1:24:52, 4.95it/s] {'loss': 0.3025, 'grad_norm': 0.859375, 'learning_rate': 0.00010501999666722213, 'epoch': 0.47} 47%|████▋ | 22800/48008 [1:18:07<1:24:52, 4.95it/s] 47%|████▋ | 22801/48008 [1:18:08<1:23:28, 5.03it/s] 47%|████▋ | 22802/48008 [1:18:08<1:22:16, 5.11it/s] 47%|████▋ | 22803/48008 [1:18:08<1:22:35, 5.09it/s] 48%|████▊ | 22804/48008 [1:18:08<1:22:08, 5.11it/s] 48%|████▊ | 22805/48008 [1:18:08<1:21:55, 5.13it/s] 48%|████▊ | 22806/48008 [1:18:09<1:21:29, 5.15it/s] 48%|████▊ | 22807/48008 [1:18:09<1:20:58, 5.19it/s] 48%|████▊ | 22808/48008 [1:18:09<1:20:35, 5.21it/s] 48%|████▊ | 22809/48008 [1:18:09<1:35:04, 4.42it/s] 48%|████▊ | 22810/48008 [1:18:10<1:30:29, 4.64it/s] 48%|████▊ | 22811/48008 [1:18:10<1:27:15, 4.81it/s] 48%|████▊ | 22812/48008 [1:18:10<1:25:52, 4.89it/s] 48%|████▊ | 22813/48008 [1:18:10<1:23:55, 5.00it/s] 48%|████▊ | 22814/48008 [1:18:10<1:22:50, 5.07it/s] 48%|████▊ | 22815/48008 [1:18:10<1:22:09, 5.11it/s] 48%|████▊ | 22816/48008 [1:18:11<1:23:00, 5.06it/s] 48%|████▊ | 22817/48008 [1:18:11<1:23:04, 5.05it/s] 48%|████▊ | 22818/48008 [1:18:11<1:46:54, 3.93it/s] 48%|████▊ | 22819/48008 [1:18:11<1:38:56, 4.24it/s] 48%|████▊ | 22820/48008 [1:18:12<1:31:59, 4.56it/s] 48%|████▊ | 22821/48008 [1:18:12<1:27:48, 4.78it/s] 48%|████▊ | 22822/48008 [1:18:12<1:25:23, 4.92it/s] 48%|████▊ | 22823/48008 [1:18:12<1:23:40, 5.02it/s] 48%|████▊ | 22824/48008 [1:18:12<1:22:37, 5.08it/s] 48%|████▊ | 22825/48008 [1:18:13<1:23:14, 5.04it/s] 48%|████▊ | 22826/48008 [1:18:13<1:21:54, 5.12it/s] 48%|████▊ | 22827/48008 [1:18:13<1:21:22, 5.16it/s] 48%|████▊ | 22828/48008 [1:18:13<1:21:07, 5.17it/s] 48%|████▊ | 22829/48008 [1:18:13<1:35:16, 4.40it/s] 48%|████▊ | 22830/48008 [1:18:14<1:30:35, 4.63it/s] 48%|████▊ | 22831/48008 [1:18:14<1:27:35, 4.79it/s] 48%|████▊ | 22832/48008 [1:18:14<1:26:46, 4.84it/s] 48%|████▊ | 22833/48008 [1:18:14<1:24:52, 4.94it/s] 48%|████▊ | 22834/48008 [1:18:15<1:48:06, 3.88it/s] 48%|████▊ | 22835/48008 [1:18:15<1:39:12, 4.23it/s] 48%|████▊ | 22836/48008 [1:18:15<1:35:04, 4.41it/s] 48%|████▊ | 22837/48008 [1:18:15<1:30:21, 4.64it/s] 48%|████▊ | 22838/48008 [1:18:15<1:27:20, 4.80it/s] 48%|████▊ | 22839/48008 [1:18:16<1:26:04, 4.87it/s] 48%|████▊ | 22840/48008 [1:18:16<1:23:39, 5.01it/s] 48%|████▊ | 22841/48008 [1:18:16<1:24:11, 4.98it/s] 48%|████▊ | 22842/48008 [1:18:16<1:22:57, 5.06it/s] 48%|████▊ | 22843/48008 [1:18:16<1:22:13, 5.10it/s] 48%|████▊ | 22844/48008 [1:18:17<1:21:30, 5.15it/s] 48%|████▊ | 22845/48008 [1:18:17<1:21:32, 5.14it/s] 48%|████▊ | 22846/48008 [1:18:17<1:21:22, 5.15it/s] 48%|████▊ | 22847/48008 [1:18:17<1:20:20, 5.22it/s] 48%|████▊ | 22848/48008 [1:18:17<1:20:04, 5.24it/s] 48%|████▊ | 22849/48008 [1:18:18<1:21:15, 5.16it/s] 48%|████▊ | 22850/48008 [1:18:18<1:21:02, 5.17it/s] {'loss': 0.3007, 'grad_norm': 0.1640625, 'learning_rate': 0.00010481169805032496, 'epoch': 0.48} 48%|████▊ | 22850/48008 [1:18:18<1:21:02, 5.17it/s] 48%|████▊ | 22851/48008 [1:18:18<1:20:54, 5.18it/s] 48%|████▊ | 22852/48008 [1:18:18<1:20:44, 5.19it/s] 48%|████▊ | 22853/48008 [1:18:18<1:20:08, 5.23it/s] 48%|████▊ | 22854/48008 [1:18:18<1:20:10, 5.23it/s] 48%|████▊ | 22855/48008 [1:18:19<1:20:07, 5.23it/s] 48%|████▊ | 22856/48008 [1:18:19<1:34:34, 4.43it/s] 48%|████▊ | 22857/48008 [1:18:19<1:30:20, 4.64it/s] 48%|████▊ | 22858/48008 [1:18:19<1:29:44, 4.67it/s] 48%|████▊ | 22859/48008 [1:18:20<1:27:32, 4.79it/s] 48%|████▊ | 22860/48008 [1:18:20<1:26:26, 4.85it/s] 48%|████▊ | 22861/48008 [1:18:20<1:39:04, 4.23it/s] 48%|████▊ | 22862/48008 [1:18:20<1:34:42, 4.43it/s] 48%|████▊ | 22863/48008 [1:18:20<1:32:39, 4.52it/s] 48%|████▊ | 22864/48008 [1:18:21<1:29:27, 4.68it/s] 48%|████▊ | 22865/48008 [1:18:21<1:26:31, 4.84it/s] 48%|████▊ | 22866/48008 [1:18:21<1:24:32, 4.96it/s] 48%|████▊ | 22867/48008 [1:18:21<1:23:27, 5.02it/s] 48%|████▊ | 22868/48008 [1:18:21<1:22:44, 5.06it/s] 48%|████▊ | 22869/48008 [1:18:22<1:23:19, 5.03it/s] 48%|████▊ | 22870/48008 [1:18:22<1:28:12, 4.75it/s] 48%|████▊ | 22871/48008 [1:18:22<1:25:19, 4.91it/s] 48%|████▊ | 22872/48008 [1:18:22<1:25:14, 4.91it/s] 48%|████▊ | 22873/48008 [1:18:22<1:23:21, 5.03it/s] 48%|████▊ | 22874/48008 [1:18:23<1:22:57, 5.05it/s] 48%|████▊ | 22875/48008 [1:18:23<1:36:32, 4.34it/s] 48%|████▊ | 22876/48008 [1:18:23<1:31:30, 4.58it/s] 48%|████▊ | 22877/48008 [1:18:23<1:28:15, 4.75it/s] 48%|████▊ | 22878/48008 [1:18:24<1:25:52, 4.88it/s] 48%|████▊ | 22879/48008 [1:18:24<1:24:14, 4.97it/s] 48%|████▊ | 22880/48008 [1:18:24<1:24:15, 4.97it/s] 48%|████▊ | 22881/48008 [1:18:24<1:47:18, 3.90it/s] 48%|████▊ | 22882/48008 [1:18:25<1:38:43, 4.24it/s] 48%|████▊ | 22883/48008 [1:18:25<1:33:08, 4.50it/s] 48%|████▊ | 22884/48008 [1:18:25<1:29:48, 4.66it/s] 48%|████▊ | 22885/48008 [1:18:25<1:26:51, 4.82it/s] 48%|████▊ | 22886/48008 [1:18:25<1:25:02, 4.92it/s] 48%|████▊ | 22887/48008 [1:18:26<1:29:23, 4.68it/s] 48%|████▊ | 22888/48008 [1:18:26<1:26:28, 4.84it/s] 48%|████▊ | 22889/48008 [1:18:26<1:24:08, 4.98it/s] 48%|████▊ | 22890/48008 [1:18:26<1:25:17, 4.91it/s] 48%|████▊ | 22891/48008 [1:18:26<1:23:57, 4.99it/s] 48%|████▊ | 22892/48008 [1:18:27<1:22:29, 5.07it/s] 48%|████▊ | 22893/48008 [1:18:27<1:21:47, 5.12it/s] 48%|████▊ | 22894/48008 [1:18:27<1:21:18, 5.15it/s] 48%|████▊ | 22895/48008 [1:18:27<1:20:59, 5.17it/s] 48%|████▊ | 22896/48008 [1:18:27<1:20:50, 5.18it/s] 48%|████▊ | 22897/48008 [1:18:27<1:20:29, 5.20it/s] 48%|████▊ | 22898/48008 [1:18:28<1:21:05, 5.16it/s] 48%|████▊ | 22899/48008 [1:18:28<1:20:35, 5.19it/s] 48%|████▊ | 22900/48008 [1:18:28<1:20:31, 5.20it/s] {'loss': 0.2981, 'grad_norm': 0.259765625, 'learning_rate': 0.00010460339943342777, 'epoch': 0.48} 48%|████▊ | 22900/48008 [1:18:28<1:20:31, 5.20it/s] 48%|████▊ | 22901/48008 [1:18:28<1:26:06, 4.86it/s] 48%|████▊ | 22902/48008 [1:18:28<1:23:33, 5.01it/s] 48%|████▊ | 22903/48008 [1:18:29<1:22:22, 5.08it/s] 48%|████▊ | 22904/48008 [1:18:29<1:21:43, 5.12it/s] 48%|████▊ | 22905/48008 [1:18:29<1:20:53, 5.17it/s] 48%|████▊ | 22906/48008 [1:18:29<1:22:08, 5.09it/s] 48%|████▊ | 22907/48008 [1:18:29<1:21:08, 5.16it/s] 48%|████▊ | 22908/48008 [1:18:30<1:20:57, 5.17it/s] 48%|████▊ | 22909/48008 [1:18:30<1:20:20, 5.21it/s] 48%|████▊ | 22910/48008 [1:18:30<1:21:55, 5.11it/s] 48%|████▊ | 22911/48008 [1:18:30<1:21:11, 5.15it/s] 48%|████▊ | 22912/48008 [1:18:30<1:20:23, 5.20it/s] 48%|████▊ | 22913/48008 [1:18:31<1:19:52, 5.24it/s] 48%|████▊ | 22914/48008 [1:18:31<1:19:47, 5.24it/s] 48%|████▊ | 22915/48008 [1:18:31<1:19:46, 5.24it/s] 48%|████▊ | 22916/48008 [1:18:31<1:19:25, 5.27it/s] 48%|████▊ | 22917/48008 [1:18:31<1:19:14, 5.28it/s] 48%|████▊ | 22918/48008 [1:18:32<1:19:32, 5.26it/s] 48%|████▊ | 22919/48008 [1:18:32<1:21:00, 5.16it/s] 48%|████▊ | 22920/48008 [1:18:32<1:20:43, 5.18it/s] 48%|████▊ | 22921/48008 [1:18:32<1:21:45, 5.11it/s] 48%|████▊ | 22922/48008 [1:18:32<1:21:13, 5.15it/s] 48%|████▊ | 22923/48008 [1:18:32<1:20:17, 5.21it/s] 48%|████▊ | 22924/48008 [1:18:33<1:21:22, 5.14it/s] 48%|████▊ | 22925/48008 [1:18:33<1:20:24, 5.20it/s] 48%|████▊ | 22926/48008 [1:18:33<1:21:13, 5.15it/s] 48%|████▊ | 22927/48008 [1:18:33<1:20:55, 5.17it/s] 48%|████▊ | 22928/48008 [1:18:33<1:21:50, 5.11it/s] 48%|████▊ | 22929/48008 [1:18:34<1:21:12, 5.15it/s] 48%|████▊ | 22930/48008 [1:18:34<1:35:19, 4.38it/s] 48%|████▊ | 22931/48008 [1:18:34<1:32:52, 4.50it/s] 48%|████▊ | 22932/48008 [1:18:34<1:29:04, 4.69it/s] 48%|████▊ | 22933/48008 [1:18:35<1:25:57, 4.86it/s] 48%|████▊ | 22934/48008 [1:18:35<1:22:43, 5.05it/s] 48%|████▊ | 22935/48008 [1:18:35<1:22:37, 5.06it/s] 48%|████▊ | 22936/48008 [1:18:35<1:36:18, 4.34it/s] 48%|████▊ | 22937/48008 [1:18:36<1:55:40, 3.61it/s] 48%|████▊ | 22938/48008 [1:18:36<1:44:34, 4.00it/s] 48%|████▊ | 22939/48008 [1:18:36<1:36:21, 4.34it/s] 48%|████▊ | 22940/48008 [1:18:36<1:32:14, 4.53it/s] 48%|████▊ | 22941/48008 [1:18:36<1:28:30, 4.72it/s] 48%|████▊ | 22942/48008 [1:18:37<1:26:19, 4.84it/s] 48%|████▊ | 22943/48008 [1:18:37<1:25:54, 4.86it/s] 48%|████▊ | 22944/48008 [1:18:37<1:24:25, 4.95it/s] 48%|████▊ | 22945/48008 [1:18:37<1:22:48, 5.04it/s] 48%|████▊ | 22946/48008 [1:18:37<1:21:37, 5.12it/s] 48%|████▊ | 22947/48008 [1:18:38<1:22:28, 5.06it/s] 48%|████▊ | 22948/48008 [1:18:38<1:21:38, 5.12it/s] 48%|████▊ | 22949/48008 [1:18:38<1:19:50, 5.23it/s] 48%|████▊ | 22950/48008 [1:18:38<1:19:33, 5.25it/s] {'loss': 0.3024, 'grad_norm': 0.29296875, 'learning_rate': 0.00010439510081653059, 'epoch': 0.48} 48%|████▊ | 22950/48008 [1:18:38<1:19:33, 5.25it/s] 48%|████▊ | 22951/48008 [1:18:38<1:20:47, 5.17it/s] 48%|████▊ | 22952/48008 [1:18:39<1:20:05, 5.21it/s] 48%|████▊ | 22953/48008 [1:18:39<1:25:33, 4.88it/s] 48%|████▊ | 22954/48008 [1:18:39<1:29:39, 4.66it/s] 48%|████▊ | 22955/48008 [1:18:39<1:27:13, 4.79it/s] 48%|████▊ | 22956/48008 [1:18:39<1:27:08, 4.79it/s] 48%|████▊ | 22957/48008 [1:18:40<1:24:36, 4.93it/s] 48%|████▊ | 22958/48008 [1:18:40<1:23:23, 5.01it/s] 48%|████▊ | 22959/48008 [1:18:40<1:27:47, 4.76it/s] 48%|████▊ | 22960/48008 [1:18:40<1:25:10, 4.90it/s] 48%|████▊ | 22961/48008 [1:18:40<1:22:15, 5.07it/s] 48%|████▊ | 22962/48008 [1:18:41<1:21:23, 5.13it/s] 48%|████▊ | 22963/48008 [1:18:41<1:20:33, 5.18it/s] 48%|████▊ | 22964/48008 [1:18:41<1:21:30, 5.12it/s] 48%|████▊ | 22965/48008 [1:18:41<1:20:52, 5.16it/s] 48%|████▊ | 22966/48008 [1:18:41<1:20:25, 5.19it/s] 48%|████▊ | 22967/48008 [1:18:42<1:20:21, 5.19it/s] 48%|████▊ | 22968/48008 [1:18:42<1:20:15, 5.20it/s] 48%|████▊ | 22969/48008 [1:18:42<1:19:23, 5.26it/s] 48%|████▊ | 22970/48008 [1:18:42<1:19:25, 5.25it/s] 48%|████▊ | 22971/48008 [1:18:42<1:20:11, 5.20it/s] 48%|████▊ | 22972/48008 [1:18:42<1:19:58, 5.22it/s] 48%|████▊ | 22973/48008 [1:18:43<1:18:29, 5.32it/s] 48%|████▊ | 22974/48008 [1:18:43<1:18:54, 5.29it/s] 48%|████▊ | 22975/48008 [1:18:43<1:19:22, 5.26it/s] 48%|████▊ | 22976/48008 [1:18:43<1:18:40, 5.30it/s] 48%|████▊ | 22977/48008 [1:18:43<1:19:07, 5.27it/s] 48%|████▊ | 22978/48008 [1:18:44<1:20:16, 5.20it/s] 48%|████▊ | 22979/48008 [1:18:44<1:19:50, 5.22it/s] 48%|████▊ | 22980/48008 [1:18:44<1:20:25, 5.19it/s] 48%|████▊ | 22981/48008 [1:18:44<1:21:18, 5.13it/s] 48%|████▊ | 22982/48008 [1:18:44<1:20:42, 5.17it/s] 48%|████▊ | 22983/48008 [1:18:45<1:19:42, 5.23it/s] 48%|████▊ | 22984/48008 [1:18:45<1:19:39, 5.24it/s] 48%|████▊ | 22985/48008 [1:18:45<1:20:12, 5.20it/s] 48%|████▊ | 22986/48008 [1:18:45<1:20:26, 5.18it/s] 48%|████▊ | 22987/48008 [1:18:45<1:21:33, 5.11it/s] 48%|████▊ | 22988/48008 [1:18:46<1:21:38, 5.11it/s] 48%|████▊ | 22989/48008 [1:18:46<1:20:51, 5.16it/s] 48%|████▊ | 22990/48008 [1:18:46<1:21:06, 5.14it/s] 48%|████▊ | 22991/48008 [1:18:46<1:23:01, 5.02it/s] 48%|████▊ | 22992/48008 [1:18:46<1:21:57, 5.09it/s] 48%|████▊ | 22993/48008 [1:18:47<1:21:09, 5.14it/s] 48%|████▊ | 22994/48008 [1:18:47<1:45:01, 3.97it/s] 48%|████▊ | 22995/48008 [1:18:47<1:43:16, 4.04it/s] 48%|████▊ | 22996/48008 [1:18:47<1:36:57, 4.30it/s] 48%|████▊ | 22997/48008 [1:18:48<1:31:29, 4.56it/s] 48%|████▊ | 22998/48008 [1:18:48<1:27:56, 4.74it/s] 48%|████▊ | 22999/48008 [1:18:48<1:25:53, 4.85it/s] 48%|████▊ | 23000/48008 [1:18:48<1:26:20, 4.83it/s] {'loss': 0.302, 'grad_norm': 0.1328125, 'learning_rate': 0.00010418680219963339, 'epoch': 0.48} 48%|████▊ | 23000/48008 [1:18:48<1:26:20, 4.83it/s] 48%|████▊ | 23001/48008 [1:18:48<1:24:41, 4.92it/s] 48%|████▊ | 23002/48008 [1:18:49<1:22:41, 5.04it/s] 48%|████▊ | 23003/48008 [1:18:49<1:21:41, 5.10it/s] 48%|████▊ | 23004/48008 [1:18:49<1:26:46, 4.80it/s] 48%|████▊ | 23005/48008 [1:18:49<1:24:51, 4.91it/s] 48%|████▊ | 23006/48008 [1:18:49<1:23:19, 5.00it/s] 48%|████▊ | 23007/48008 [1:18:50<1:21:54, 5.09it/s] 48%|████▊ | 23008/48008 [1:18:50<1:21:51, 5.09it/s] 48%|████▊ | 23009/48008 [1:18:50<1:20:54, 5.15it/s] 48%|████▊ | 23010/48008 [1:18:50<1:20:29, 5.18it/s] 48%|████▊ | 23011/48008 [1:18:50<1:20:14, 5.19it/s] 48%|████▊ | 23012/48008 [1:18:50<1:20:55, 5.15it/s] 48%|████▊ | 23013/48008 [1:18:51<1:19:44, 5.22it/s] 48%|████▊ | 23014/48008 [1:18:51<1:20:08, 5.20it/s] 48%|████▊ | 23015/48008 [1:18:51<1:20:08, 5.20it/s] 48%|████▊ | 23016/48008 [1:18:51<1:20:14, 5.19it/s] 48%|████▊ | 23017/48008 [1:18:51<1:25:56, 4.85it/s] 48%|████▊ | 23018/48008 [1:18:52<1:38:29, 4.23it/s] 48%|████▊ | 23019/48008 [1:18:52<1:34:56, 4.39it/s] 48%|████▊ | 23020/48008 [1:18:52<1:30:47, 4.59it/s] 48%|████▊ | 23021/48008 [1:18:52<1:27:30, 4.76it/s] 48%|████▊ | 23022/48008 [1:18:53<1:25:01, 4.90it/s] 48%|████▊ | 23023/48008 [1:18:53<1:25:45, 4.86it/s] 48%|████▊ | 23024/48008 [1:18:53<1:29:52, 4.63it/s] 48%|████▊ | 23025/48008 [1:18:53<1:26:55, 4.79it/s] 48%|████▊ | 23026/48008 [1:18:53<1:25:30, 4.87it/s] 48%|████▊ | 23027/48008 [1:18:54<1:24:39, 4.92it/s] 48%|████▊ | 23028/48008 [1:18:54<1:23:09, 5.01it/s] 48%|████▊ | 23029/48008 [1:18:54<1:21:24, 5.11it/s] 48%|████▊ | 23030/48008 [1:18:54<1:20:56, 5.14it/s] 48%|████▊ | 23031/48008 [1:18:54<1:20:20, 5.18it/s] 48%|████▊ | 23032/48008 [1:18:55<1:21:07, 5.13it/s] 48%|████▊ | 23033/48008 [1:18:55<1:20:26, 5.17it/s] 48%|████▊ | 23034/48008 [1:18:55<1:20:20, 5.18it/s] 48%|████▊ | 23035/48008 [1:18:55<1:34:36, 4.40it/s] 48%|████▊ | 23036/48008 [1:18:55<1:29:33, 4.65it/s] 48%|████▊ | 23037/48008 [1:18:56<1:32:22, 4.51it/s] 48%|████▊ | 23038/48008 [1:18:56<1:29:55, 4.63it/s] 48%|████▊ | 23039/48008 [1:18:56<1:25:58, 4.84it/s] 48%|████▊ | 23040/48008 [1:18:56<1:23:52, 4.96it/s] 48%|████▊ | 23041/48008 [1:18:56<1:22:58, 5.02it/s] 48%|████▊ | 23042/48008 [1:18:57<1:21:47, 5.09it/s] 48%|████▊ | 23043/48008 [1:18:57<1:20:43, 5.15it/s] 48%|████▊ | 23044/48008 [1:18:57<1:20:24, 5.17it/s] 48%|████▊ | 23045/48008 [1:18:57<1:20:18, 5.18it/s] 48%|████▊ | 23046/48008 [1:18:57<1:21:40, 5.09it/s] 48%|████▊ | 23047/48008 [1:18:58<1:20:23, 5.17it/s] 48%|████▊ | 23048/48008 [1:18:58<1:20:51, 5.14it/s] 48%|████▊ | 23049/48008 [1:18:58<1:21:20, 5.11it/s] 48%|████▊ | 23050/48008 [1:18:58<1:20:42, 5.15it/s] 48%|████▊ | 23050/48008 [1:18:58<1:20:42, 5.15it/s]{'loss': 0.2991, 'grad_norm': 0.265625, 'learning_rate': 0.00010397850358273622, 'epoch': 0.48} 48%|████▊ | 23051/48008 [1:18:58<1:20:06, 5.19it/s] 48%|████▊ | 23052/48008 [1:18:59<1:20:11, 5.19it/s] 48%|████▊ | 23053/48008 [1:18:59<1:44:21, 3.99it/s] 48%|████▊ | 23054/48008 [1:18:59<1:36:40, 4.30it/s] 48%|████▊ | 23055/48008 [1:18:59<1:31:17, 4.56it/s] 48%|████▊ | 23056/48008 [1:19:00<1:29:27, 4.65it/s] 48%|████▊ | 23057/48008 [1:19:00<1:28:14, 4.71it/s] 48%|████▊ | 23058/48008 [1:19:00<1:25:47, 4.85it/s] 48%|████▊ | 23059/48008 [1:19:00<1:23:30, 4.98it/s] 48%|████▊ | 23060/48008 [1:19:00<1:23:09, 5.00it/s] 48%|████▊ | 23061/48008 [1:19:01<1:27:51, 4.73it/s] 48%|████▊ | 23062/48008 [1:19:01<1:25:24, 4.87it/s] 48%|████▊ | 23063/48008 [1:19:01<1:23:02, 5.01it/s] 48%|████▊ | 23064/48008 [1:19:01<1:23:43, 4.97it/s] 48%|████▊ | 23065/48008 [1:19:01<1:24:02, 4.95it/s] 48%|████▊ | 23066/48008 [1:19:02<1:23:20, 4.99it/s] 48%|████▊ | 23067/48008 [1:19:02<1:23:23, 4.98it/s] 48%|████▊ | 23068/48008 [1:19:02<1:22:55, 5.01it/s] 48%|████▊ | 23069/48008 [1:19:02<1:22:08, 5.06it/s] 48%|████▊ | 23070/48008 [1:19:02<1:21:54, 5.07it/s] 48%|████▊ | 23071/48008 [1:19:03<1:20:48, 5.14it/s] 48%|████▊ | 23072/48008 [1:19:03<1:21:48, 5.08it/s] 48%|████▊ | 23073/48008 [1:19:03<1:20:52, 5.14it/s] 48%|████▊ | 23074/48008 [1:19:03<1:20:42, 5.15it/s] 48%|████▊ | 23075/48008 [1:19:03<1:20:43, 5.15it/s] 48%|████▊ | 23076/48008 [1:19:04<1:21:19, 5.11it/s] 48%|████▊ | 23077/48008 [1:19:04<1:20:40, 5.15it/s] 48%|████▊ | 23078/48008 [1:19:04<1:20:05, 5.19it/s] 48%|████▊ | 23079/48008 [1:19:04<1:21:11, 5.12it/s] 48%|████▊ | 23080/48008 [1:19:04<1:21:32, 5.10it/s] 48%|████▊ | 23081/48008 [1:19:04<1:20:37, 5.15it/s] 48%|████▊ | 23082/48008 [1:19:05<1:19:21, 5.23it/s] 48%|████▊ | 23083/48008 [1:19:05<1:20:29, 5.16it/s] 48%|████▊ | 23084/48008 [1:19:05<1:21:41, 5.09it/s] 48%|████▊ | 23085/48008 [1:19:05<1:20:58, 5.13it/s] 48%|████▊ | 23086/48008 [1:19:05<1:21:19, 5.11it/s] 48%|████▊ | 23087/48008 [1:19:06<1:21:41, 5.08it/s] 48%|████▊ | 23088/48008 [1:19:06<1:20:44, 5.14it/s] 48%|████▊ | 23089/48008 [1:19:06<1:21:52, 5.07it/s] 48%|████▊ | 23090/48008 [1:19:06<1:20:41, 5.15it/s] 48%|████▊ | 23091/48008 [1:19:06<1:20:17, 5.17it/s] 48%|████▊ | 23092/48008 [1:19:07<1:20:03, 5.19it/s] 48%|████▊ | 23093/48008 [1:19:07<1:20:07, 5.18it/s] 48%|████▊ | 23094/48008 [1:19:07<1:19:35, 5.22it/s] 48%|████▊ | 23095/48008 [1:19:07<1:21:09, 5.12it/s] 48%|████▊ | 23096/48008 [1:19:07<1:22:06, 5.06it/s] 48%|████▊ | 23097/48008 [1:19:08<1:35:39, 4.34it/s] 48%|████▊ | 23098/48008 [1:19:08<1:30:53, 4.57it/s] 48%|████▊ | 23099/48008 [1:19:08<1:28:54, 4.67it/s] 48%|████▊ | 23100/48008 [1:19:08<1:28:21, 4.70it/s] 48%|████▊ | 23100/48008 [1:19:08<1:28:21, 4.70it/s]{'loss': 0.3006, 'grad_norm': 0.85546875, 'learning_rate': 0.00010377020496583902, 'epoch': 0.48} 48%|████▊ | 23101/48008 [1:19:09<1:26:51, 4.78it/s] 48%|████▊ | 23102/48008 [1:19:09<1:24:46, 4.90it/s] 48%|████▊ | 23103/48008 [1:19:09<1:22:34, 5.03it/s] 48%|████▊ | 23104/48008 [1:19:09<1:23:07, 4.99it/s] 48%|████▊ | 23105/48008 [1:19:09<1:22:05, 5.06it/s] 48%|████▊ | 23106/48008 [1:19:09<1:22:06, 5.05it/s] 48%|████▊ | 23107/48008 [1:19:10<1:21:40, 5.08it/s] 48%|████▊ | 23108/48008 [1:19:10<1:21:51, 5.07it/s] 48%|████▊ | 23109/48008 [1:19:10<1:21:03, 5.12it/s] 48%|████▊ | 23110/48008 [1:19:10<1:21:08, 5.11it/s] 48%|████▊ | 23111/48008 [1:19:10<1:20:33, 5.15it/s] 48%|████▊ | 23112/48008 [1:19:11<1:34:18, 4.40it/s] 48%|████▊ | 23113/48008 [1:19:11<1:31:21, 4.54it/s] 48%|████▊ | 23114/48008 [1:19:11<1:28:40, 4.68it/s] 48%|████▊ | 23115/48008 [1:19:11<1:26:38, 4.79it/s] 48%|████▊ | 23116/48008 [1:19:12<1:25:15, 4.87it/s] 48%|████▊ | 23117/48008 [1:19:12<1:25:10, 4.87it/s] 48%|████▊ | 23118/48008 [1:19:12<1:22:56, 5.00it/s] 48%|████▊ | 23119/48008 [1:19:12<1:21:54, 5.06it/s] 48%|████▊ | 23120/48008 [1:19:12<1:20:47, 5.13it/s] 48%|████▊ | 23121/48008 [1:19:13<1:21:58, 5.06it/s] 48%|████▊ | 23122/48008 [1:19:13<1:21:00, 5.12it/s] 48%|████▊ | 23123/48008 [1:19:13<1:21:59, 5.06it/s] 48%|████▊ | 23124/48008 [1:19:13<1:21:10, 5.11it/s] 48%|████▊ | 23125/48008 [1:19:13<1:20:15, 5.17it/s] 48%|████▊ | 23126/48008 [1:19:14<1:21:17, 5.10it/s] 48%|████▊ | 23127/48008 [1:19:14<1:20:28, 5.15it/s] 48%|████▊ | 23128/48008 [1:19:14<1:21:31, 5.09it/s] 48%|████▊ | 23129/48008 [1:19:14<1:20:04, 5.18it/s] 48%|████▊ | 23130/48008 [1:19:14<1:19:34, 5.21it/s] 48%|████▊ | 23131/48008 [1:19:14<1:19:33, 5.21it/s] 48%|████▊ | 23132/48008 [1:19:15<1:18:29, 5.28it/s] 48%|████▊ | 23133/48008 [1:19:15<1:18:42, 5.27it/s] 48%|████▊ | 23134/48008 [1:19:15<1:18:48, 5.26it/s] 48%|████▊ | 23135/48008 [1:19:15<1:18:03, 5.31it/s] 48%|████▊ | 23136/48008 [1:19:15<1:20:09, 5.17it/s] 48%|████▊ | 23137/48008 [1:19:16<1:20:06, 5.17it/s] 48%|████▊ | 23138/48008 [1:19:16<1:19:55, 5.19it/s] 48%|████▊ | 23139/48008 [1:19:16<1:19:32, 5.21it/s] 48%|████▊ | 23140/48008 [1:19:16<1:19:18, 5.23it/s] 48%|████▊ | 23141/48008 [1:19:16<1:19:29, 5.21it/s] 48%|████▊ | 23142/48008 [1:19:17<1:25:04, 4.87it/s] 48%|████▊ | 23143/48008 [1:19:17<1:23:26, 4.97it/s] 48%|████▊ | 23144/48008 [1:19:17<1:21:40, 5.07it/s] 48%|████▊ | 23145/48008 [1:19:17<1:21:14, 5.10it/s] 48%|████▊ | 23146/48008 [1:19:17<1:23:00, 4.99it/s] 48%|████▊ | 23147/48008 [1:19:18<1:21:37, 5.08it/s] 48%|████▊ | 23148/48008 [1:19:18<1:21:00, 5.11it/s] 48%|████▊ | 23149/48008 [1:19:18<1:20:12, 5.17it/s] 48%|████▊ | 23150/48008 [1:19:18<1:19:49, 5.19it/s] {'loss': 0.2998, 'grad_norm': 0.2177734375, 'learning_rate': 0.00010356190634894185, 'epoch': 0.48} 48%|████▊ | 23150/48008 [1:19:18<1:19:49, 5.19it/s] 48%|████▊ | 23151/48008 [1:19:18<1:20:40, 5.14it/s] 48%|████▊ | 23152/48008 [1:19:19<1:21:43, 5.07it/s] 48%|████▊ | 23153/48008 [1:19:19<1:20:59, 5.11it/s] 48%|████▊ | 23154/48008 [1:19:19<1:19:37, 5.20it/s] 48%|████▊ | 23155/48008 [1:19:19<1:18:54, 5.25it/s] 48%|████▊ | 23156/48008 [1:19:19<1:19:46, 5.19it/s] 48%|████▊ | 23157/48008 [1:19:20<1:20:40, 5.13it/s] 48%|████▊ | 23158/48008 [1:19:20<1:19:56, 5.18it/s] 48%|████▊ | 23159/48008 [1:19:20<1:19:42, 5.20it/s] 48%|████▊ | 23160/48008 [1:19:20<1:21:02, 5.11it/s] 48%|████▊ | 23161/48008 [1:19:20<1:20:32, 5.14it/s] 48%|████▊ | 23162/48008 [1:19:21<1:20:05, 5.17it/s] 48%|████▊ | 23163/48008 [1:19:21<1:19:09, 5.23it/s] 48%|████▊ | 23164/48008 [1:19:21<1:19:58, 5.18it/s] 48%|████▊ | 23165/48008 [1:19:21<1:20:34, 5.14it/s] 48%|████▊ | 23166/48008 [1:19:21<1:19:51, 5.18it/s] 48%|████▊ | 23167/48008 [1:19:21<1:19:28, 5.21it/s] 48%|████▊ | 23168/48008 [1:19:22<1:19:20, 5.22it/s] 48%|████▊ | 23169/48008 [1:19:22<1:19:25, 5.21it/s] 48%|████▊ | 23170/48008 [1:19:22<1:20:37, 5.13it/s] 48%|████▊ | 23171/48008 [1:19:22<1:21:31, 5.08it/s] 48%|████▊ | 23172/48008 [1:19:22<1:21:29, 5.08it/s] 48%|████▊ | 23173/48008 [1:19:23<1:22:04, 5.04it/s] 48%|████▊ | 23174/48008 [1:19:23<1:22:24, 5.02it/s] 48%|████▊ | 23175/48008 [1:19:23<1:20:36, 5.13it/s] 48%|████▊ | 23176/48008 [1:19:23<1:19:45, 5.19it/s] 48%|████▊ | 23177/48008 [1:19:23<1:19:23, 5.21it/s] 48%|████▊ | 23178/48008 [1:19:24<1:19:23, 5.21it/s] 48%|████▊ | 23179/48008 [1:19:24<1:24:48, 4.88it/s] 48%|████▊ | 23180/48008 [1:19:24<1:22:43, 5.00it/s] 48%|████▊ | 23181/48008 [1:19:24<1:35:48, 4.32it/s] 48%|████▊ | 23182/48008 [1:19:25<1:30:37, 4.57it/s] 48%|████▊ | 23183/48008 [1:19:25<1:26:35, 4.78it/s] 48%|████▊ | 23184/48008 [1:19:25<1:24:28, 4.90it/s] 48%|████▊ | 23185/48008 [1:19:25<1:24:23, 4.90it/s] 48%|████▊ | 23186/48008 [1:19:25<1:22:53, 4.99it/s] 48%|████▊ | 23187/48008 [1:19:25<1:22:15, 5.03it/s] 48%|████▊ | 23188/48008 [1:19:26<1:21:04, 5.10it/s] 48%|████▊ | 23189/48008 [1:19:26<1:20:09, 5.16it/s] 48%|████▊ | 23190/48008 [1:19:26<1:19:46, 5.18it/s] 48%|████▊ | 23191/48008 [1:19:26<1:19:12, 5.22it/s] 48%|████▊ | 23192/48008 [1:19:26<1:19:17, 5.22it/s] 48%|████▊ | 23193/48008 [1:19:27<1:20:47, 5.12it/s] 48%|████▊ | 23194/48008 [1:19:27<1:20:03, 5.17it/s] 48%|████▊ | 23195/48008 [1:19:27<1:19:30, 5.20it/s] 48%|████▊ | 23196/48008 [1:19:27<1:21:40, 5.06it/s] 48%|████▊ | 23197/48008 [1:19:27<1:20:06, 5.16it/s] 48%|████▊ | 23198/48008 [1:19:28<1:19:38, 5.19it/s] 48%|████▊ | 23199/48008 [1:19:28<1:20:00, 5.17it/s] 48%|████▊ | 23200/48008 [1:19:28<1:19:12, 5.22it/s] {'loss': 0.3035, 'grad_norm': 0.1630859375, 'learning_rate': 0.00010335360773204466, 'epoch': 0.48} 48%|████▊ | 23200/48008 [1:19:28<1:19:12, 5.22it/s] 48%|████▊ | 23201/48008 [1:19:28<1:21:18, 5.09it/s] 48%|████▊ | 23202/48008 [1:19:28<1:20:28, 5.14it/s] 48%|████▊ | 23203/48008 [1:19:29<1:25:51, 4.81it/s] 48%|████▊ | 23204/48008 [1:19:29<1:23:25, 4.96it/s] 48%|████▊ | 23205/48008 [1:19:29<1:21:29, 5.07it/s] 48%|████▊ | 23206/48008 [1:19:29<1:20:29, 5.14it/s] 48%|████▊ | 23207/48008 [1:19:29<1:19:10, 5.22it/s] 48%|████▊ | 23208/48008 [1:19:30<1:19:21, 5.21it/s] 48%|████▊ | 23209/48008 [1:19:30<1:19:05, 5.23it/s] 48%|████▊ | 23210/48008 [1:19:30<1:19:02, 5.23it/s] 48%|████▊ | 23211/48008 [1:19:30<1:18:57, 5.23it/s] 48%|████▊ | 23212/48008 [1:19:30<1:20:01, 5.16it/s] 48%|████▊ | 23213/48008 [1:19:31<1:20:06, 5.16it/s] 48%|████▊ | 23214/48008 [1:19:31<1:19:33, 5.19it/s] 48%|████▊ | 23215/48008 [1:19:31<1:18:48, 5.24it/s] 48%|████▊ | 23216/48008 [1:19:31<1:18:42, 5.25it/s] 48%|████▊ | 23217/48008 [1:19:31<1:18:18, 5.28it/s] 48%|████▊ | 23218/48008 [1:19:32<1:42:38, 4.03it/s] 48%|████▊ | 23219/48008 [1:19:32<1:37:01, 4.26it/s] 48%|████▊ | 23220/48008 [1:19:32<1:37:21, 4.24it/s] 48%|████▊ | 23221/48008 [1:19:32<1:32:55, 4.45it/s] 48%|████▊ | 23222/48008 [1:19:33<1:30:08, 4.58it/s] 48%|████▊ | 23223/48008 [1:19:33<1:28:26, 4.67it/s] 48%|████▊ | 23224/48008 [1:19:33<1:25:11, 4.85it/s] 48%|████▊ | 23225/48008 [1:19:33<1:37:39, 4.23it/s] 48%|████▊ | 23226/48008 [1:19:33<1:32:03, 4.49it/s] 48%|████▊ | 23227/48008 [1:19:34<1:28:09, 4.68it/s] 48%|████▊ | 23228/48008 [1:19:34<1:39:35, 4.15it/s] 48%|████▊ | 23229/48008 [1:19:34<1:38:58, 4.17it/s] 48%|████▊ | 23230/48008 [1:19:34<1:33:03, 4.44it/s] 48%|████▊ | 23231/48008 [1:19:35<1:30:01, 4.59it/s] 48%|████▊ | 23232/48008 [1:19:35<1:26:11, 4.79it/s] 48%|████▊ | 23233/48008 [1:19:35<1:24:03, 4.91it/s] 48%|████▊ | 23234/48008 [1:19:35<1:22:18, 5.02it/s] 48%|████▊ | 23235/48008 [1:19:35<1:20:26, 5.13it/s] 48%|████▊ | 23236/48008 [1:19:35<1:19:37, 5.19it/s] 48%|████▊ | 23237/48008 [1:19:36<1:21:05, 5.09it/s] 48%|████▊ | 23238/48008 [1:19:36<1:20:25, 5.13it/s] 48%|████▊ | 23239/48008 [1:19:36<1:21:32, 5.06it/s] 48%|████▊ | 23240/48008 [1:19:36<1:21:35, 5.06it/s] 48%|████▊ | 23241/48008 [1:19:36<1:20:49, 5.11it/s] 48%|████▊ | 23242/48008 [1:19:37<1:20:11, 5.15it/s] 48%|████▊ | 23243/48008 [1:19:37<1:20:19, 5.14it/s] 48%|████▊ | 23244/48008 [1:19:37<1:19:40, 5.18it/s] 48%|████▊ | 23245/48008 [1:19:37<1:19:29, 5.19it/s] 48%|████▊ | 23246/48008 [1:19:37<1:20:03, 5.15it/s] 48%|████▊ | 23247/48008 [1:19:38<1:21:10, 5.08it/s] 48%|████▊ | 23248/48008 [1:19:38<1:21:36, 5.06it/s] 48%|████▊ | 23249/48008 [1:19:38<1:21:03, 5.09it/s] 48%|████▊ | 23250/48008 [1:19:38<1:19:43, 5.18it/s] {'loss': 0.2998, 'grad_norm': 0.361328125, 'learning_rate': 0.00010314530911514748, 'epoch': 0.48} 48%|████▊ | 23250/48008 [1:19:38<1:19:43, 5.18it/s] 48%|████▊ | 23251/48008 [1:19:38<1:21:05, 5.09it/s] 48%|████▊ | 23252/48008 [1:19:39<1:20:20, 5.14it/s] 48%|████▊ | 23253/48008 [1:19:39<1:21:22, 5.07it/s] 48%|████▊ | 23254/48008 [1:19:39<1:20:35, 5.12it/s] 48%|████▊ | 23255/48008 [1:19:39<1:19:28, 5.19it/s] 48%|████▊ | 23256/48008 [1:19:39<1:18:24, 5.26it/s] 48%|████▊ | 23257/48008 [1:19:40<1:18:36, 5.25it/s] 48%|████▊ | 23258/48008 [1:19:40<1:18:29, 5.26it/s] 48%|████▊ | 23259/48008 [1:19:40<1:18:27, 5.26it/s] 48%|████▊ | 23260/48008 [1:19:40<1:21:07, 5.08it/s] 48%|████▊ | 23261/48008 [1:19:40<1:20:53, 5.10it/s] 48%|████▊ | 23262/48008 [1:19:41<1:26:05, 4.79it/s] 48%|████▊ | 23263/48008 [1:19:41<1:23:47, 4.92it/s] 48%|████▊ | 23264/48008 [1:19:41<1:22:02, 5.03it/s] 48%|████▊ | 23265/48008 [1:19:41<1:20:40, 5.11it/s] 48%|████▊ | 23266/48008 [1:19:41<1:19:37, 5.18it/s] 48%|████▊ | 23267/48008 [1:19:42<1:18:47, 5.23it/s] 48%|████▊ | 23268/48008 [1:19:42<1:18:45, 5.23it/s] 48%|████▊ | 23269/48008 [1:19:42<1:20:14, 5.14it/s] 48%|████▊ | 23270/48008 [1:19:42<1:19:26, 5.19it/s] 48%|████▊ | 23271/48008 [1:19:42<1:20:07, 5.15it/s] 48%|████▊ | 23272/48008 [1:19:43<1:21:12, 5.08it/s] 48%|████▊ | 23273/48008 [1:19:43<1:21:36, 5.05it/s] 48%|████▊ | 23274/48008 [1:19:43<1:20:50, 5.10it/s] 48%|████▊ | 23275/48008 [1:19:43<1:20:23, 5.13it/s] 48%|████▊ | 23276/48008 [1:19:43<1:19:57, 5.15it/s] 48%|████▊ | 23277/48008 [1:19:43<1:22:01, 5.02it/s] 48%|████▊ | 23278/48008 [1:19:44<1:22:31, 4.99it/s] 48%|████▊ | 23279/48008 [1:19:44<1:21:27, 5.06it/s] 48%|████▊ | 23280/48008 [1:19:44<1:20:42, 5.11it/s] 48%|████▊ | 23281/48008 [1:19:44<1:34:20, 4.37it/s] 48%|████▊ | 23282/48008 [1:19:45<1:53:48, 3.62it/s] 48%|████▊ | 23283/48008 [1:19:45<1:45:05, 3.92it/s] 49%|████▊ | 23284/48008 [1:19:45<1:36:59, 4.25it/s] 49%|████▊ | 23285/48008 [1:19:45<1:33:10, 4.42it/s] 49%|████▊ | 23286/48008 [1:19:46<1:34:47, 4.35it/s] 49%|████▊ | 23287/48008 [1:19:46<1:32:30, 4.45it/s] 49%|████▊ | 23288/48008 [1:19:46<1:29:15, 4.62it/s] 49%|████▊ | 23289/48008 [1:19:46<1:26:20, 4.77it/s] 49%|████▊ | 23290/48008 [1:19:46<1:29:38, 4.60it/s] 49%|████▊ | 23291/48008 [1:19:47<1:26:19, 4.77it/s] 49%|████▊ | 23292/48008 [1:19:47<1:24:09, 4.89it/s] 49%|████▊ | 23293/48008 [1:19:47<1:23:02, 4.96it/s] 49%|████▊ | 23294/48008 [1:19:47<1:21:53, 5.03it/s] 49%|████▊ | 23295/48008 [1:19:47<1:21:08, 5.08it/s] 49%|████▊ | 23296/48008 [1:19:48<1:20:30, 5.12it/s] 49%|████▊ | 23297/48008 [1:19:48<1:19:37, 5.17it/s] 49%|████▊ | 23298/48008 [1:19:48<1:19:58, 5.15it/s] 49%|████▊ | 23299/48008 [1:19:48<1:19:30, 5.18it/s] 49%|████▊ | 23300/48008 [1:19:48<1:19:23, 5.19it/s] {'loss': 0.3004, 'grad_norm': 0.5078125, 'learning_rate': 0.00010293701049825028, 'epoch': 0.49} 49%|████▊ | 23300/48008 [1:19:48<1:19:23, 5.19it/s] 49%|████▊ | 23301/48008 [1:19:49<1:18:57, 5.22it/s] 49%|████▊ | 23302/48008 [1:19:49<1:18:53, 5.22it/s] 49%|████▊ | 23303/48008 [1:19:49<1:20:07, 5.14it/s] 49%|████▊ | 23304/48008 [1:19:49<1:18:32, 5.24it/s] 49%|████▊ | 23305/48008 [1:19:49<1:18:44, 5.23it/s] 49%|████▊ | 23306/48008 [1:19:50<1:18:44, 5.23it/s] 49%|████▊ | 23307/48008 [1:19:50<1:18:32, 5.24it/s] 49%|████▊ | 23308/48008 [1:19:50<1:24:20, 4.88it/s] 49%|████▊ | 23309/48008 [1:19:50<1:21:54, 5.03it/s] 49%|████▊ | 23310/48008 [1:19:50<1:22:13, 5.01it/s] 49%|████▊ | 23311/48008 [1:19:51<1:21:36, 5.04it/s] 49%|████▊ | 23312/48008 [1:19:51<1:20:32, 5.11it/s] 49%|████▊ | 23313/48008 [1:19:51<1:19:53, 5.15it/s] 49%|████▊ | 23314/48008 [1:19:51<1:20:41, 5.10it/s] 49%|████▊ | 23315/48008 [1:19:51<1:21:07, 5.07it/s] 49%|████▊ | 23316/48008 [1:19:51<1:19:53, 5.15it/s] 49%|████▊ | 23317/48008 [1:19:52<1:18:53, 5.22it/s] 49%|████▊ | 23318/48008 [1:19:52<1:17:56, 5.28it/s] 49%|████▊ | 23319/48008 [1:19:52<1:18:06, 5.27it/s] 49%|████▊ | 23320/48008 [1:19:52<1:18:56, 5.21it/s] 49%|████▊ | 23321/48008 [1:19:52<1:18:42, 5.23it/s] 49%|████▊ | 23322/48008 [1:19:53<1:19:09, 5.20it/s] 49%|████▊ | 23323/48008 [1:19:53<1:20:17, 5.12it/s] 49%|████▊ | 23324/48008 [1:19:53<1:18:24, 5.25it/s] 49%|████▊ | 23325/48008 [1:19:53<1:18:28, 5.24it/s] 49%|████▊ | 23326/48008 [1:19:53<1:18:07, 5.27it/s] 49%|████▊ | 23327/48008 [1:19:54<1:23:55, 4.90it/s] 49%|████▊ | 23328/48008 [1:19:54<1:22:49, 4.97it/s] 49%|████▊ | 23329/48008 [1:19:54<1:21:49, 5.03it/s] 49%|████▊ | 23330/48008 [1:19:54<1:22:11, 5.00it/s] 49%|████▊ | 23331/48008 [1:19:54<1:21:01, 5.08it/s] 49%|████▊ | 23332/48008 [1:19:55<1:44:20, 3.94it/s] 49%|████▊ | 23333/48008 [1:19:55<1:36:30, 4.26it/s] 49%|████▊ | 23334/48008 [1:19:55<1:31:14, 4.51it/s] 49%|████▊ | 23335/48008 [1:19:55<1:28:06, 4.67it/s] 49%|████▊ | 23336/48008 [1:19:56<1:26:51, 4.73it/s] 49%|████▊ | 23337/48008 [1:19:56<1:25:06, 4.83it/s] 49%|████▊ | 23338/48008 [1:19:56<1:23:48, 4.91it/s] 49%|████▊ | 23339/48008 [1:19:56<1:23:17, 4.94it/s] 49%|████▊ | 23340/48008 [1:19:56<1:24:03, 4.89it/s] 49%|████▊ | 23341/48008 [1:19:57<1:23:08, 4.94it/s] 49%|████▊ | 23342/48008 [1:19:57<1:21:57, 5.02it/s] 49%|████▊ | 23343/48008 [1:19:57<1:20:09, 5.13it/s] 49%|████▊ | 23344/48008 [1:19:57<1:19:24, 5.18it/s] 49%|████▊ | 23345/48008 [1:19:57<1:19:05, 5.20it/s] 49%|████▊ | 23346/48008 [1:19:58<1:19:43, 5.16it/s] 49%|████▊ | 23347/48008 [1:19:58<1:18:53, 5.21it/s] 49%|████▊ | 23348/48008 [1:19:58<1:20:15, 5.12it/s] 49%|████▊ | 23349/48008 [1:19:58<1:19:36, 5.16it/s] 49%|████▊ | 23350/48008 [1:19:58<1:19:23, 5.18it/s] {'loss': 0.3008, 'grad_norm': 0.16015625, 'learning_rate': 0.00010272871188135311, 'epoch': 0.49} 49%|████▊ | 23350/48008 [1:19:58<1:19:23, 5.18it/s] 49%|████▊ | 23351/48008 [1:19:58<1:18:35, 5.23it/s] 49%|████▊ | 23352/48008 [1:19:59<1:19:41, 5.16it/s] 49%|████▊ | 23353/48008 [1:19:59<1:19:28, 5.17it/s] 49%|████▊ | 23354/48008 [1:19:59<1:19:09, 5.19it/s] 49%|████▊ | 23355/48008 [1:19:59<1:18:26, 5.24it/s] 49%|████▊ | 23356/48008 [1:19:59<1:19:46, 5.15it/s] 49%|████▊ | 23357/48008 [1:20:00<1:19:17, 5.18it/s] 49%|████▊ | 23358/48008 [1:20:00<1:20:27, 5.11it/s] 49%|████▊ | 23359/48008 [1:20:00<1:19:59, 5.14it/s] 49%|████▊ | 23360/48008 [1:20:00<1:18:52, 5.21it/s] 49%|████▊ | 23361/48008 [1:20:00<1:19:48, 5.15it/s] 49%|████▊ | 23362/48008 [1:20:01<1:18:52, 5.21it/s] 49%|████▊ | 23363/48008 [1:20:01<1:18:35, 5.23it/s] 49%|████▊ | 23364/48008 [1:20:01<1:20:53, 5.08it/s] 49%|████▊ | 23365/48008 [1:20:01<1:20:13, 5.12it/s] 49%|████▊ | 23366/48008 [1:20:01<1:19:18, 5.18it/s] 49%|████▊ | 23367/48008 [1:20:02<1:19:25, 5.17it/s] 49%|████▊ | 23368/48008 [1:20:02<1:19:44, 5.15it/s] 49%|████▊ | 23369/48008 [1:20:02<1:20:27, 5.10it/s] 49%|████▊ | 23370/48008 [1:20:02<1:22:15, 4.99it/s] 49%|████▊ | 23371/48008 [1:20:02<1:22:07, 5.00it/s] 49%|████▊ | 23372/48008 [1:20:03<1:21:05, 5.06it/s] 49%|████▊ | 23373/48008 [1:20:03<1:20:40, 5.09it/s] 49%|████▊ | 23374/48008 [1:20:03<1:19:51, 5.14it/s] 49%|████▊ | 23375/48008 [1:20:03<1:18:15, 5.25it/s] 49%|████▊ | 23376/48008 [1:20:04<1:42:47, 3.99it/s] 49%|████▊ | 23377/48008 [1:20:04<1:36:08, 4.27it/s] 49%|████▊ | 23378/48008 [1:20:04<1:32:18, 4.45it/s] 49%|████▊ | 23379/48008 [1:20:04<1:28:19, 4.65it/s] 49%|████▊ | 23380/48008 [1:20:05<3:27:03, 1.98it/s] 49%|████▊ | 23381/48008 [1:20:06<2:48:17, 2.44it/s] 49%|████▊ | 23382/48008 [1:20:06<2:21:09, 2.91it/s] 49%|████▊ | 23383/48008 [1:20:06<2:03:29, 3.32it/s] 49%|████▊ | 23384/48008 [1:20:06<1:49:29, 3.75it/s] 49%|████▊ | 23385/48008 [1:20:06<1:39:33, 4.12it/s] 49%|████▊ | 23386/48008 [1:20:06<1:33:39, 4.38it/s] 49%|████▊ | 23387/48008 [1:20:07<1:29:08, 4.60it/s] 49%|████▊ | 23388/48008 [1:20:07<1:26:33, 4.74it/s] 49%|████▊ | 23389/48008 [1:20:07<1:29:53, 4.56it/s] 49%|████▊ | 23390/48008 [1:20:07<1:26:45, 4.73it/s] 49%|████▊ | 23391/48008 [1:20:07<1:24:57, 4.83it/s] 49%|████▊ | 23392/48008 [1:20:08<1:24:00, 4.88it/s] 49%|████▊ | 23393/48008 [1:20:08<1:22:09, 4.99it/s] 49%|████▊ | 23394/48008 [1:20:08<1:35:13, 4.31it/s] 49%|████▊ | 23395/48008 [1:20:08<1:31:30, 4.48it/s] 49%|████▊ | 23396/48008 [1:20:09<1:33:08, 4.40it/s] 49%|████▊ | 23397/48008 [1:20:09<1:34:20, 4.35it/s] 49%|████▊ | 23398/48008 [1:20:09<1:32:02, 4.46it/s] 49%|████▊ | 23399/48008 [1:20:09<1:28:21, 4.64it/s] 49%|████▊ | 23400/48008 [1:20:09<1:25:08, 4.82it/s] {'loss': 0.3015, 'grad_norm': 0.4921875, 'learning_rate': 0.00010252041326445594, 'epoch': 0.49} 49%|████▊ | 23400/48008 [1:20:09<1:25:08, 4.82it/s] 49%|████▊ | 23401/48008 [1:20:10<1:24:08, 4.87it/s] 49%|████▊ | 23402/48008 [1:20:10<1:21:52, 5.01it/s] 49%|████▊ | 23403/48008 [1:20:10<1:19:40, 5.15it/s] 49%|████▉ | 23404/48008 [1:20:10<1:19:24, 5.16it/s] 49%|████▉ | 23405/48008 [1:20:10<1:20:30, 5.09it/s] 49%|████▉ | 23406/48008 [1:20:11<1:25:32, 4.79it/s] 49%|████▉ | 23407/48008 [1:20:11<1:22:43, 4.96it/s] 49%|████▉ | 23408/48008 [1:20:11<1:21:00, 5.06it/s] 49%|████▉ | 23409/48008 [1:20:11<1:19:11, 5.18it/s] 49%|████▉ | 23410/48008 [1:20:11<1:19:40, 5.15it/s] 49%|████▉ | 23411/48008 [1:20:12<1:19:55, 5.13it/s] 49%|████▉ | 23412/48008 [1:20:12<1:20:25, 5.10it/s] 49%|████▉ | 23413/48008 [1:20:12<1:33:52, 4.37it/s] 49%|████▉ | 23414/48008 [1:20:12<1:28:38, 4.62it/s] 49%|████▉ | 23415/48008 [1:20:12<1:26:54, 4.72it/s] 49%|████▉ | 23416/48008 [1:20:13<1:25:11, 4.81it/s] 49%|████▉ | 23417/48008 [1:20:13<1:22:50, 4.95it/s] 49%|████▉ | 23418/48008 [1:20:13<1:21:11, 5.05it/s] 49%|████▉ | 23419/48008 [1:20:13<1:44:15, 3.93it/s] 49%|████▉ | 23420/48008 [1:20:14<1:36:31, 4.25it/s] 49%|████▉ | 23421/48008 [1:20:14<1:32:40, 4.42it/s] 49%|████▉ | 23422/48008 [1:20:14<1:28:25, 4.63it/s] 49%|████▉ | 23423/48008 [1:20:14<1:25:37, 4.78it/s] 49%|████▉ | 23424/48008 [1:20:14<1:23:26, 4.91it/s] 49%|████▉ | 23425/48008 [1:20:15<1:23:03, 4.93it/s] 49%|████▉ | 23426/48008 [1:20:15<1:21:22, 5.03it/s] 49%|████▉ | 23427/48008 [1:20:15<1:19:51, 5.13it/s] 49%|████▉ | 23428/48008 [1:20:15<1:19:14, 5.17it/s] 49%|████▉ | 23429/48008 [1:20:15<1:18:24, 5.22it/s] 49%|████▉ | 23430/48008 [1:20:16<1:19:43, 5.14it/s] 49%|████▉ | 23431/48008 [1:20:16<1:19:17, 5.17it/s] 49%|████▉ | 23432/48008 [1:20:16<1:18:08, 5.24it/s] 49%|████▉ | 23433/48008 [1:20:16<1:17:06, 5.31it/s] 49%|████▉ | 23434/48008 [1:20:16<1:16:13, 5.37it/s] 49%|████▉ | 23435/48008 [1:20:17<1:16:53, 5.33it/s] 49%|████▉ | 23436/48008 [1:20:17<1:18:49, 5.20it/s] 49%|████▉ | 23437/48008 [1:20:17<1:18:50, 5.19it/s] 49%|████▉ | 23438/48008 [1:20:17<1:20:56, 5.06it/s] 49%|████▉ | 23439/48008 [1:20:17<1:20:01, 5.12it/s] 49%|████▉ | 23440/48008 [1:20:18<1:19:06, 5.18it/s] 49%|████▉ | 23441/48008 [1:20:18<1:18:26, 5.22it/s] 49%|████▉ | 23442/48008 [1:20:18<1:17:37, 5.27it/s] 49%|████▉ | 23443/48008 [1:20:18<1:19:03, 5.18it/s] 49%|████▉ | 23444/48008 [1:20:18<1:21:01, 5.05it/s] 49%|████▉ | 23445/48008 [1:20:18<1:19:42, 5.14it/s] 49%|████▉ | 23446/48008 [1:20:19<1:20:45, 5.07it/s] 49%|████▉ | 23447/48008 [1:20:19<1:25:46, 4.77it/s] 49%|████▉ | 23448/48008 [1:20:19<1:23:08, 4.92it/s] 49%|████▉ | 23449/48008 [1:20:19<1:21:33, 5.02it/s] 49%|████▉ | 23450/48008 [1:20:19<1:20:39, 5.07it/s] {'loss': 0.3009, 'grad_norm': 0.2158203125, 'learning_rate': 0.00010231211464755875, 'epoch': 0.49} 49%|████▉ | 23450/48008 [1:20:19<1:20:39, 5.07it/s] 49%|████▉ | 23451/48008 [1:20:20<1:20:36, 5.08it/s] 49%|████▉ | 23452/48008 [1:20:20<1:20:19, 5.09it/s] 49%|████▉ | 23453/48008 [1:20:20<1:19:16, 5.16it/s] 49%|████▉ | 23454/48008 [1:20:20<1:42:57, 3.97it/s] 49%|████▉ | 23455/48008 [1:20:21<1:37:08, 4.21it/s] 49%|████▉ | 23456/48008 [1:20:21<1:33:31, 4.37it/s] 49%|████▉ | 23457/48008 [1:20:21<1:29:01, 4.60it/s] 49%|████▉ | 23458/48008 [1:20:21<1:25:28, 4.79it/s] 49%|████▉ | 23459/48008 [1:20:21<1:23:02, 4.93it/s] 49%|████▉ | 23460/48008 [1:20:22<1:21:39, 5.01it/s] 49%|████▉ | 23461/48008 [1:20:22<1:21:28, 5.02it/s] 49%|████▉ | 23462/48008 [1:20:22<1:34:30, 4.33it/s] 49%|████▉ | 23463/48008 [1:20:22<1:30:18, 4.53it/s] 49%|████▉ | 23464/48008 [1:20:23<1:26:05, 4.75it/s] 49%|████▉ | 23465/48008 [1:20:23<1:23:45, 4.88it/s] 49%|████▉ | 23466/48008 [1:20:23<1:21:48, 5.00it/s] 49%|████▉ | 23467/48008 [1:20:23<1:20:48, 5.06it/s] 49%|████▉ | 23468/48008 [1:20:23<1:20:38, 5.07it/s] 49%|████▉ | 23469/48008 [1:20:23<1:19:59, 5.11it/s] 49%|████▉ | 23470/48008 [1:20:24<1:19:15, 5.16it/s] 49%|████▉ | 23471/48008 [1:20:24<1:24:33, 4.84it/s] 49%|████▉ | 23472/48008 [1:20:24<1:23:16, 4.91it/s] 49%|████▉ | 23473/48008 [1:20:24<1:21:50, 5.00it/s] 49%|████▉ | 23474/48008 [1:20:24<1:22:18, 4.97it/s] 49%|████▉ | 23475/48008 [1:20:25<1:22:13, 4.97it/s] 49%|████▉ | 23476/48008 [1:20:25<1:21:03, 5.04it/s] 49%|████▉ | 23477/48008 [1:20:25<1:20:25, 5.08it/s] 49%|████▉ | 23478/48008 [1:20:25<1:19:20, 5.15it/s] 49%|████▉ | 23479/48008 [1:20:25<1:18:43, 5.19it/s] 49%|████▉ | 23480/48008 [1:20:26<1:20:11, 5.10it/s] 49%|████▉ | 23481/48008 [1:20:26<1:19:34, 5.14it/s] 49%|████▉ | 23482/48008 [1:20:26<1:18:47, 5.19it/s] 49%|████▉ | 23483/48008 [1:20:26<1:19:25, 5.15it/s] 49%|████▉ | 23484/48008 [1:20:26<1:19:32, 5.14it/s] 49%|████▉ | 23485/48008 [1:20:27<1:20:53, 5.05it/s] 49%|████▉ | 23486/48008 [1:20:27<1:20:29, 5.08it/s] 49%|████▉ | 23487/48008 [1:20:27<1:33:56, 4.35it/s] 49%|████▉ | 23488/48008 [1:20:27<1:29:05, 4.59it/s] 49%|████▉ | 23489/48008 [1:20:28<1:25:34, 4.77it/s] 49%|████▉ | 23490/48008 [1:20:28<1:25:45, 4.77it/s] 49%|████▉ | 23491/48008 [1:20:28<1:22:51, 4.93it/s] 49%|████▉ | 23492/48008 [1:20:28<1:21:19, 5.02it/s] 49%|████▉ | 23493/48008 [1:20:28<1:19:57, 5.11it/s] 49%|████▉ | 23494/48008 [1:20:28<1:19:32, 5.14it/s] 49%|████▉ | 23495/48008 [1:20:29<1:20:19, 5.09it/s] 49%|████▉ | 23496/48008 [1:20:29<1:19:10, 5.16it/s] 49%|████▉ | 23497/48008 [1:20:29<1:18:27, 5.21it/s] 49%|████▉ | 23498/48008 [1:20:29<1:17:46, 5.25it/s] 49%|████▉ | 23499/48008 [1:20:29<1:17:09, 5.29it/s] 49%|████▉ | 23500/48008 [1:20:30<1:18:56, 5.17it/s] {'loss': 0.2986, 'grad_norm': 0.236328125, 'learning_rate': 0.00010210381603066158, 'epoch': 0.49} 49%|████▉ | 23500/48008 [1:20:30<1:18:56, 5.17it/s] 49%|████▉ | 23501/48008 [1:20:30<1:20:08, 5.10it/s] 49%|████▉ | 23502/48008 [1:20:30<1:18:39, 5.19it/s] 49%|████▉ | 23503/48008 [1:20:30<1:17:23, 5.28it/s] 49%|████▉ | 23504/48008 [1:20:30<1:17:30, 5.27it/s] 49%|████▉ | 23505/48008 [1:20:31<1:17:48, 5.25it/s] 49%|████▉ | 23506/48008 [1:20:31<1:19:43, 5.12it/s] 49%|████▉ | 23507/48008 [1:20:31<1:19:41, 5.12it/s] 49%|████▉ | 23508/48008 [1:20:31<1:19:04, 5.16it/s] 49%|████▉ | 23509/48008 [1:20:31<1:18:56, 5.17it/s] 49%|████▉ | 23510/48008 [1:20:32<1:18:55, 5.17it/s] 49%|████▉ | 23511/48008 [1:20:32<1:18:45, 5.18it/s] 49%|████▉ | 23512/48008 [1:20:32<1:18:34, 5.20it/s] 49%|████▉ | 23513/48008 [1:20:32<1:42:11, 3.99it/s] 49%|████▉ | 23514/48008 [1:20:33<1:34:47, 4.31it/s] 49%|████▉ | 23515/48008 [1:20:33<1:31:05, 4.48it/s] 49%|████▉ | 23516/48008 [1:20:33<1:28:29, 4.61it/s] 49%|████▉ | 23517/48008 [1:20:33<1:26:06, 4.74it/s] 49%|████▉ | 23518/48008 [1:20:33<1:23:34, 4.88it/s] 49%|████▉ | 23519/48008 [1:20:34<1:21:44, 4.99it/s] 49%|████▉ | 23520/48008 [1:20:34<1:21:24, 5.01it/s] 49%|████▉ | 23521/48008 [1:20:34<1:20:36, 5.06it/s] 49%|████▉ | 23522/48008 [1:20:34<1:21:12, 5.02it/s] 49%|████▉ | 23523/48008 [1:20:34<1:19:30, 5.13it/s] 49%|████▉ | 23524/48008 [1:20:35<1:24:46, 4.81it/s] 49%|████▉ | 23525/48008 [1:20:35<1:23:27, 4.89it/s] 49%|████▉ | 23526/48008 [1:20:35<1:21:58, 4.98it/s] 49%|████▉ | 23527/48008 [1:20:35<1:20:48, 5.05it/s] 49%|████▉ | 23528/48008 [1:20:35<1:19:37, 5.12it/s] 49%|████▉ | 23529/48008 [1:20:35<1:19:19, 5.14it/s] 49%|████▉ | 23530/48008 [1:20:36<1:18:21, 5.21it/s] 49%|████▉ | 23531/48008 [1:20:36<1:19:50, 5.11it/s] 49%|████▉ | 23532/48008 [1:20:36<1:21:32, 5.00it/s] 49%|████▉ | 23533/48008 [1:20:36<1:26:23, 4.72it/s] 49%|████▉ | 23534/48008 [1:20:37<1:24:40, 4.82it/s] 49%|████▉ | 23535/48008 [1:20:37<1:22:03, 4.97it/s] 49%|████▉ | 23536/48008 [1:20:37<1:20:40, 5.06it/s] 49%|████▉ | 23537/48008 [1:20:37<1:19:41, 5.12it/s] 49%|████▉ | 23538/48008 [1:20:37<1:20:35, 5.06it/s] 49%|████▉ | 23539/48008 [1:20:37<1:19:49, 5.11it/s] 49%|████▉ | 23540/48008 [1:20:38<1:19:10, 5.15it/s] 49%|████▉ | 23541/48008 [1:20:38<1:42:37, 3.97it/s] 49%|████▉ | 23542/48008 [1:20:38<1:34:47, 4.30it/s] 49%|████▉ | 23543/48008 [1:20:38<1:29:43, 4.54it/s] 49%|████▉ | 23544/48008 [1:20:39<1:27:02, 4.68it/s] 49%|████▉ | 23545/48008 [1:20:39<1:24:25, 4.83it/s] 49%|████▉ | 23546/48008 [1:20:39<1:22:06, 4.97it/s] 49%|████▉ | 23547/48008 [1:20:39<1:20:10, 5.08it/s] 49%|████▉ | 23548/48008 [1:20:39<1:19:15, 5.14it/s] 49%|████▉ | 23549/48008 [1:20:40<2:54:25, 2.34it/s] 49%|████▉ | 23550/48008 [1:20:41<2:27:05, 2.77it/s] {'loss': 0.2989, 'grad_norm': 0.423828125, 'learning_rate': 0.00010189551741376438, 'epoch': 0.49} 49%|████▉ | 23550/48008 [1:20:41<2:27:05, 2.77it/s] 49%|████▉ | 23551/48008 [1:20:41<2:06:32, 3.22it/s] 49%|████▉ | 23552/48008 [1:20:41<2:05:54, 3.24it/s] 49%|████▉ | 23553/48008 [1:20:41<1:50:58, 3.67it/s] 49%|████▉ | 23554/48008 [1:20:42<2:04:51, 3.26it/s] 49%|████▉ | 23555/48008 [1:20:43<3:16:56, 2.07it/s] 49%|████▉ | 23556/48008 [1:20:43<2:42:01, 2.52it/s] 49%|████▉ | 23557/48008 [1:20:43<2:16:34, 2.98it/s] 49%|████▉ | 23558/48008 [1:20:43<2:01:24, 3.36it/s] 49%|████▉ | 23559/48008 [1:20:43<1:48:14, 3.76it/s] 49%|████▉ | 23560/48008 [1:20:44<1:39:12, 4.11it/s] 49%|████▉ | 23561/48008 [1:20:44<2:47:42, 2.43it/s] 49%|████▉ | 23562/48008 [1:20:45<2:44:36, 2.48it/s] 49%|████▉ | 23563/48008 [1:20:45<2:18:40, 2.94it/s] 49%|████▉ | 23564/48008 [1:20:45<1:59:14, 3.42it/s] 49%|████▉ | 23565/48008 [1:20:45<1:46:44, 3.82it/s] 49%|████▉ | 23566/48008 [1:20:45<1:37:19, 4.19it/s] 49%|████▉ | 23567/48008 [1:20:46<1:31:30, 4.45it/s] 49%|████▉ | 23568/48008 [1:20:46<1:27:34, 4.65it/s] 49%|████▉ | 23569/48008 [1:20:46<1:23:57, 4.85it/s] 49%|████▉ | 23570/48008 [1:20:46<1:21:39, 4.99it/s] 49%|████▉ | 23571/48008 [1:20:46<1:21:58, 4.97it/s] 49%|████▉ | 23572/48008 [1:20:47<1:21:01, 5.03it/s] 49%|████▉ | 23573/48008 [1:20:47<1:44:01, 3.91it/s] 49%|████▉ | 23574/48008 [1:20:47<1:36:19, 4.23it/s] 49%|████▉ | 23575/48008 [1:20:47<1:32:57, 4.38it/s] 49%|████▉ | 23576/48008 [1:20:48<1:27:56, 4.63it/s] 49%|████▉ | 23577/48008 [1:20:48<1:24:48, 4.80it/s] 49%|████▉ | 23578/48008 [1:20:48<1:22:33, 4.93it/s] 49%|████▉ | 23579/48008 [1:20:48<1:20:50, 5.04it/s] 49%|████▉ | 23580/48008 [1:20:48<1:20:08, 5.08it/s] 49%|████▉ | 23581/48008 [1:20:49<1:21:38, 4.99it/s] 49%|████▉ | 23582/48008 [1:20:49<1:20:32, 5.05it/s] 49%|████▉ | 23583/48008 [1:20:49<1:19:12, 5.14it/s] 49%|████▉ | 23584/48008 [1:20:49<1:19:10, 5.14it/s] 49%|████▉ | 23585/48008 [1:20:49<1:18:24, 5.19it/s] 49%|████▉ | 23586/48008 [1:20:50<1:18:20, 5.20it/s] 49%|████▉ | 23587/48008 [1:20:50<1:18:16, 5.20it/s] 49%|████▉ | 23588/48008 [1:20:50<1:41:47, 4.00it/s] 49%|████▉ | 23589/48008 [1:20:50<1:34:34, 4.30it/s] 49%|████▉ | 23590/48008 [1:20:50<1:29:22, 4.55it/s] 49%|████▉ | 23591/48008 [1:20:51<2:41:54, 2.51it/s] 49%|████▉ | 23592/48008 [1:20:51<2:16:07, 2.99it/s] 49%|████▉ | 23593/48008 [1:20:52<1:58:30, 3.43it/s] 49%|████▉ | 23594/48008 [1:20:52<1:47:05, 3.80it/s] 49%|████▉ | 23595/48008 [1:20:52<1:39:14, 4.10it/s] 49%|████▉ | 23596/48008 [1:20:52<1:32:44, 4.39it/s] 49%|████▉ | 23597/48008 [1:20:52<1:28:38, 4.59it/s] 49%|████▉ | 23598/48008 [1:20:53<1:24:40, 4.80it/s] 49%|████▉ | 23599/48008 [1:20:53<1:23:36, 4.87it/s] 49%|████▉ | 23600/48008 [1:20:53<1:21:53, 4.97it/s] {'loss': 0.2994, 'grad_norm': 0.46875, 'learning_rate': 0.0001016872187968672, 'epoch': 0.49} 49%|████▉ | 23600/48008 [1:20:53<1:21:53, 4.97it/s] 49%|████▉ | 23601/48008 [1:20:53<1:21:01, 5.02it/s] 49%|████▉ | 23602/48008 [1:20:53<1:20:08, 5.08it/s] 49%|████▉ | 23603/48008 [1:20:54<2:42:53, 2.50it/s] 49%|████▉ | 23604/48008 [1:20:54<2:17:04, 2.97it/s] 49%|████▉ | 23605/48008 [1:20:55<1:59:03, 3.42it/s] 49%|████▉ | 23606/48008 [1:20:55<1:47:35, 3.78it/s] 49%|████▉ | 23607/48008 [1:20:55<1:38:27, 4.13it/s] 49%|████▉ | 23608/48008 [1:20:55<1:32:04, 4.42it/s] 49%|████▉ | 23609/48008 [1:20:55<1:28:14, 4.61it/s] 49%|████▉ | 23610/48008 [1:20:56<1:24:49, 4.79it/s] 49%|████▉ | 23611/48008 [1:20:56<1:22:16, 4.94it/s] 49%|████▉ | 23612/48008 [1:20:56<1:20:22, 5.06it/s] 49%|████▉ | 23613/48008 [1:20:56<1:25:08, 4.78it/s] 49%|████▉ | 23614/48008 [1:20:56<1:22:50, 4.91it/s] 49%|████▉ | 23615/48008 [1:20:57<1:22:14, 4.94it/s] 49%|████▉ | 23616/48008 [1:20:57<1:34:58, 4.28it/s] 49%|████▉ | 23617/48008 [1:20:57<1:29:56, 4.52it/s] 49%|████▉ | 23618/48008 [1:20:57<1:26:23, 4.71it/s] 49%|████▉ | 23619/48008 [1:20:58<1:25:18, 4.76it/s] 49%|████▉ | 23620/48008 [1:20:58<1:23:19, 4.88it/s] 49%|████▉ | 23621/48008 [1:20:58<1:21:35, 4.98it/s] 49%|████▉ | 23622/48008 [1:20:58<1:20:32, 5.05it/s] 49%|████▉ | 23623/48008 [1:20:58<1:19:53, 5.09it/s] 49%|████▉ | 23624/48008 [1:20:58<1:20:31, 5.05it/s] 49%|████▉ | 23625/48008 [1:20:59<1:22:01, 4.95it/s] 49%|████▉ | 23626/48008 [1:20:59<1:22:08, 4.95it/s] 49%|████▉ | 23627/48008 [1:20:59<1:20:44, 5.03it/s] 49%|████▉ | 23628/48008 [1:20:59<1:19:10, 5.13it/s] 49%|████▉ | 23629/48008 [1:20:59<1:18:13, 5.19it/s] 49%|████▉ | 23630/48008 [1:21:00<1:18:33, 5.17it/s] 49%|████▉ | 23631/48008 [1:21:00<1:18:58, 5.14it/s] 49%|████▉ | 23632/48008 [1:21:00<1:19:23, 5.12it/s] 49%|████▉ | 23633/48008 [1:21:00<1:20:08, 5.07it/s] 49%|████▉ | 23634/48008 [1:21:00<1:21:12, 5.00it/s] 49%|████▉ | 23635/48008 [1:21:01<1:20:19, 5.06it/s] 49%|████▉ | 23636/48008 [1:21:01<1:21:54, 4.96it/s] 49%|████▉ | 23637/48008 [1:21:01<1:19:54, 5.08it/s] 49%|████▉ | 23638/48008 [1:21:01<1:19:30, 5.11it/s] 49%|████▉ | 23639/48008 [1:21:01<1:18:55, 5.15it/s] 49%|████▉ | 23640/48008 [1:21:02<1:19:58, 5.08it/s] 49%|████▉ | 23641/48008 [1:21:02<1:19:21, 5.12it/s] 49%|████▉ | 23642/48008 [1:21:02<1:20:06, 5.07it/s] 49%|████▉ | 23643/48008 [1:21:02<1:20:11, 5.06it/s] 49%|████▉ | 23644/48008 [1:21:02<1:19:23, 5.11it/s] 49%|████▉ | 23645/48008 [1:21:03<1:19:16, 5.12it/s] 49%|████▉ | 23646/48008 [1:21:03<1:24:24, 4.81it/s] 49%|████▉ | 23647/48008 [1:21:03<1:23:10, 4.88it/s] 49%|████▉ | 23648/48008 [1:21:03<1:21:22, 4.99it/s] 49%|████▉ | 23649/48008 [1:21:03<1:19:32, 5.10it/s] 49%|████▉ | 23650/48008 [1:21:04<1:19:00, 5.14it/s] {'loss': 0.2998, 'grad_norm': 0.32421875, 'learning_rate': 0.00010147892017997, 'epoch': 0.49} 49%|████▉ | 23650/48008 [1:21:04<1:19:00, 5.14it/s] 49%|████▉ | 23651/48008 [1:21:04<1:18:44, 5.16it/s] 49%|████▉ | 23652/48008 [1:21:04<1:18:35, 5.17it/s] 49%|████▉ | 23653/48008 [1:21:04<1:17:32, 5.24it/s] 49%|████▉ | 23654/48008 [1:21:04<1:17:36, 5.23it/s] 49%|████▉ | 23655/48008 [1:21:05<1:17:44, 5.22it/s] 49%|████▉ | 23656/48008 [1:21:05<1:17:32, 5.23it/s] 49%|████▉ | 23657/48008 [1:21:05<1:17:07, 5.26it/s] 49%|████▉ | 23658/48008 [1:21:05<1:17:35, 5.23it/s] 49%|████▉ | 23659/48008 [1:21:05<1:18:49, 5.15it/s] 49%|████▉ | 23660/48008 [1:21:06<1:18:28, 5.17it/s] 49%|████▉ | 23661/48008 [1:21:06<1:18:20, 5.18it/s] 49%|████▉ | 23662/48008 [1:21:06<1:18:10, 5.19it/s] 49%|████▉ | 23663/48008 [1:21:06<1:32:00, 4.41it/s] 49%|████▉ | 23664/48008 [1:21:06<1:27:27, 4.64it/s] 49%|████▉ | 23665/48008 [1:21:07<1:25:24, 4.75it/s] 49%|████▉ | 23666/48008 [1:21:07<1:22:20, 4.93it/s] 49%|████▉ | 23667/48008 [1:21:07<1:21:25, 4.98it/s] 49%|████▉ | 23668/48008 [1:21:07<1:20:16, 5.05it/s] 49%|████▉ | 23669/48008 [1:21:07<1:19:07, 5.13it/s] 49%|████▉ | 23670/48008 [1:21:08<1:32:31, 4.38it/s] 49%|████▉ | 23671/48008 [1:21:08<1:27:57, 4.61it/s] 49%|████▉ | 23672/48008 [1:21:08<1:26:14, 4.70it/s] 49%|████▉ | 23673/48008 [1:21:08<1:24:43, 4.79it/s] 49%|████▉ | 23674/48008 [1:21:08<1:24:11, 4.82it/s] 49%|████▉ | 23675/48008 [1:21:09<1:21:28, 4.98it/s] 49%|████▉ | 23676/48008 [1:21:09<1:19:42, 5.09it/s] 49%|████▉ | 23677/48008 [1:21:09<1:18:50, 5.14it/s] 49%|████▉ | 23678/48008 [1:21:09<1:18:17, 5.18it/s] 49%|████▉ | 23679/48008 [1:21:09<1:18:13, 5.18it/s] 49%|████▉ | 23680/48008 [1:21:10<1:18:01, 5.20it/s] 49%|████▉ | 23681/48008 [1:21:10<1:17:43, 5.22it/s] 49%|████▉ | 23682/48008 [1:21:10<1:16:59, 5.27it/s] 49%|████▉ | 23683/48008 [1:21:10<1:18:26, 5.17it/s] 49%|████▉ | 23684/48008 [1:21:10<1:18:25, 5.17it/s] 49%|████▉ | 23685/48008 [1:21:11<1:18:12, 5.18it/s] 49%|████▉ | 23686/48008 [1:21:11<1:17:51, 5.21it/s] 49%|████▉ | 23687/48008 [1:21:11<1:17:44, 5.21it/s] 49%|████▉ | 23688/48008 [1:21:11<1:18:24, 5.17it/s] 49%|████▉ | 23689/48008 [1:21:11<1:17:14, 5.25it/s] 49%|████▉ | 23690/48008 [1:21:12<1:17:17, 5.24it/s] 49%|████▉ | 23691/48008 [1:21:12<1:16:41, 5.29it/s] 49%|████▉ | 23692/48008 [1:21:12<1:17:22, 5.24it/s] 49%|████▉ | 23693/48008 [1:21:12<1:17:06, 5.26it/s] 49%|████▉ | 23694/48008 [1:21:12<1:22:57, 4.89it/s] 49%|████▉ | 23695/48008 [1:21:13<1:20:38, 5.02it/s] 49%|████▉ | 23696/48008 [1:21:13<1:19:13, 5.11it/s] 49%|████▉ | 23697/48008 [1:21:13<1:18:23, 5.17it/s] 49%|████▉ | 23698/48008 [1:21:13<1:20:13, 5.05it/s] 49%|████▉ | 23699/48008 [1:21:13<1:18:48, 5.14it/s] 49%|████▉ | 23700/48008 [1:21:13<1:18:54, 5.13it/s] {'loss': 0.3034, 'grad_norm': 0.255859375, 'learning_rate': 0.00010127062156307283, 'epoch': 0.49} 49%|████▉ | 23700/48008 [1:21:13<1:18:54, 5.13it/s] 49%|████▉ | 23701/48008 [1:21:14<1:18:38, 5.15it/s] 49%|████▉ | 23702/48008 [1:21:14<1:19:19, 5.11it/s] 49%|████▉ | 23703/48008 [1:21:14<1:20:17, 5.04it/s] 49%|████▉ | 23704/48008 [1:21:14<1:43:07, 3.93it/s] 49%|████▉ | 23705/48008 [1:21:15<1:35:22, 4.25it/s] 49%|████▉ | 23706/48008 [1:21:15<1:32:07, 4.40it/s] 49%|████▉ | 23707/48008 [1:21:15<1:27:21, 4.64it/s] 49%|████▉ | 23708/48008 [1:21:15<1:24:19, 4.80it/s] 49%|████▉ | 23709/48008 [1:21:15<1:22:12, 4.93it/s] 49%|████▉ | 23710/48008 [1:21:16<1:26:17, 4.69it/s] 49%|████▉ | 23711/48008 [1:21:16<1:23:29, 4.85it/s] 49%|████▉ | 23712/48008 [1:21:16<1:21:29, 4.97it/s] 49%|████▉ | 23713/48008 [1:21:16<1:20:01, 5.06it/s] 49%|████▉ | 23714/48008 [1:21:16<1:19:28, 5.09it/s] 49%|████▉ | 23715/48008 [1:21:17<1:18:09, 5.18it/s] 49%|████▉ | 23716/48008 [1:21:17<1:17:39, 5.21it/s] 49%|████▉ | 23717/48008 [1:21:17<1:19:14, 5.11it/s] 49%|████▉ | 23718/48008 [1:21:17<1:18:26, 5.16it/s] 49%|████▉ | 23719/48008 [1:21:18<1:41:32, 3.99it/s] 49%|████▉ | 23720/48008 [1:21:18<1:48:09, 3.74it/s] 49%|████▉ | 23721/48008 [1:21:18<1:38:52, 4.09it/s] 49%|████▉ | 23722/48008 [1:21:18<1:31:55, 4.40it/s] 49%|████▉ | 23723/48008 [1:21:18<1:28:09, 4.59it/s] 49%|████▉ | 23724/48008 [1:21:19<1:24:56, 4.76it/s] 49%|████▉ | 23725/48008 [1:21:19<1:22:42, 4.89it/s] 49%|████▉ | 23726/48008 [1:21:19<1:22:38, 4.90it/s] 49%|████▉ | 23727/48008 [1:21:19<1:20:35, 5.02it/s] 49%|████▉ | 23728/48008 [1:21:19<1:20:13, 5.04it/s] 49%|████▉ | 23729/48008 [1:21:20<1:20:03, 5.05it/s] 49%|████▉ | 23730/48008 [1:21:20<1:19:15, 5.10it/s] 49%|████▉ | 23731/48008 [1:21:20<1:20:12, 5.04it/s] 49%|████▉ | 23732/48008 [1:21:20<1:19:21, 5.10it/s] 49%|████▉ | 23733/48008 [1:21:20<1:18:23, 5.16it/s] 49%|████▉ | 23734/48008 [1:21:21<1:17:24, 5.23it/s] 49%|████▉ | 23735/48008 [1:21:21<1:17:30, 5.22it/s] 49%|████▉ | 23736/48008 [1:21:21<1:23:08, 4.87it/s] 49%|████▉ | 23737/48008 [1:21:21<1:21:23, 4.97it/s] 49%|████▉ | 23738/48008 [1:21:21<1:21:26, 4.97it/s] 49%|████▉ | 23739/48008 [1:21:22<1:22:26, 4.91it/s] 49%|████▉ | 23740/48008 [1:21:22<1:21:34, 4.96it/s] 49%|████▉ | 23741/48008 [1:21:22<1:25:57, 4.70it/s] 49%|████▉ | 23742/48008 [1:21:22<1:22:57, 4.87it/s] 49%|████▉ | 23743/48008 [1:21:22<1:21:23, 4.97it/s] 49%|████▉ | 23744/48008 [1:21:23<1:25:37, 4.72it/s] 49%|████▉ | 23745/48008 [1:21:23<1:23:51, 4.82it/s] 49%|████▉ | 23746/48008 [1:21:23<1:21:26, 4.97it/s] 49%|████▉ | 23747/48008 [1:21:23<1:20:25, 5.03it/s] 49%|████▉ | 23748/48008 [1:21:23<1:20:55, 5.00it/s] 49%|████▉ | 23749/48008 [1:21:24<1:19:52, 5.06it/s] 49%|████▉ | 23750/48008 [1:21:24<1:19:02, 5.12it/s] {'loss': 0.298, 'grad_norm': 0.33203125, 'learning_rate': 0.00010106232294617564, 'epoch': 0.49} 49%|████▉ | 23750/48008 [1:21:24<1:19:02, 5.12it/s] 49%|████▉ | 23751/48008 [1:21:24<1:18:37, 5.14it/s] 49%|████▉ | 23752/48008 [1:21:24<1:17:11, 5.24it/s] 49%|████▉ | 23753/48008 [1:21:24<1:18:43, 5.13it/s] 49%|████▉ | 23754/48008 [1:21:25<1:19:51, 5.06it/s] 49%|████▉ | 23755/48008 [1:21:25<1:19:54, 5.06it/s] 49%|████▉ | 23756/48008 [1:21:25<1:24:34, 4.78it/s] 49%|████▉ | 23757/48008 [1:21:25<1:22:58, 4.87it/s] 49%|████▉ | 23758/48008 [1:21:25<1:21:05, 4.98it/s] 49%|████▉ | 23759/48008 [1:21:26<1:20:07, 5.04it/s] 49%|████▉ | 23760/48008 [1:21:26<1:19:24, 5.09it/s] 49%|████▉ | 23761/48008 [1:21:26<1:18:28, 5.15it/s] 49%|████▉ | 23762/48008 [1:21:26<1:18:03, 5.18it/s] 49%|████▉ | 23763/48008 [1:21:26<1:17:55, 5.19it/s] 50%|████▉ | 23764/48008 [1:21:27<1:17:51, 5.19it/s] 50%|████▉ | 23765/48008 [1:21:27<1:31:37, 4.41it/s] 50%|████▉ | 23766/48008 [1:21:27<1:26:28, 4.67it/s] 50%|████▉ | 23767/48008 [1:21:27<1:47:20, 3.76it/s] 50%|████▉ | 23768/48008 [1:21:28<1:38:17, 4.11it/s] 50%|████▉ | 23769/48008 [1:21:28<1:31:55, 4.39it/s] 50%|████▉ | 23770/48008 [1:21:28<1:27:35, 4.61it/s] 50%|████▉ | 23771/48008 [1:21:28<1:25:07, 4.75it/s] 50%|████▉ | 23772/48008 [1:21:28<1:24:50, 4.76it/s] 50%|████▉ | 23773/48008 [1:21:29<1:21:56, 4.93it/s] 50%|████▉ | 23774/48008 [1:21:29<1:21:00, 4.99it/s] 50%|████▉ | 23775/48008 [1:21:29<1:19:55, 5.05it/s] 50%|████▉ | 23776/48008 [1:21:29<1:18:22, 5.15it/s] 50%|████▉ | 23777/48008 [1:21:29<1:16:33, 5.27it/s] 50%|████▉ | 23778/48008 [1:21:30<1:16:50, 5.26it/s] 50%|████▉ | 23779/48008 [1:21:30<1:16:49, 5.26it/s] 50%|████▉ | 23780/48008 [1:21:30<1:17:32, 5.21it/s] 50%|████▉ | 23781/48008 [1:21:30<1:17:20, 5.22it/s] 50%|████▉ | 23782/48008 [1:21:30<1:16:47, 5.26it/s] 50%|████▉ | 23783/48008 [1:21:31<1:18:51, 5.12it/s] 50%|████▉ | 23784/48008 [1:21:31<1:18:07, 5.17it/s] 50%|████▉ | 23785/48008 [1:21:31<1:17:39, 5.20it/s] 50%|████▉ | 23786/48008 [1:21:31<1:16:40, 5.27it/s] 50%|████▉ | 23787/48008 [1:21:31<1:16:22, 5.29it/s] 50%|████▉ | 23788/48008 [1:21:31<1:16:45, 5.26it/s] 50%|████▉ | 23789/48008 [1:21:32<1:17:41, 5.20it/s] 50%|████▉ | 23790/48008 [1:21:32<1:18:01, 5.17it/s] 50%|████▉ | 23791/48008 [1:21:32<1:17:48, 5.19it/s] 50%|████▉ | 23792/48008 [1:21:32<1:17:05, 5.24it/s] 50%|████▉ | 23793/48008 [1:21:32<1:22:48, 4.87it/s] 50%|████▉ | 23794/48008 [1:21:33<1:20:39, 5.00it/s] 50%|████▉ | 23795/48008 [1:21:33<1:19:55, 5.05it/s] 50%|████▉ | 23796/48008 [1:21:33<1:20:45, 5.00it/s] 50%|████▉ | 23797/48008 [1:21:33<1:20:16, 5.03it/s] 50%|████▉ | 23798/48008 [1:21:33<1:19:19, 5.09it/s] 50%|████▉ | 23799/48008 [1:21:34<1:18:44, 5.12it/s] 50%|████▉ | 23800/48008 [1:21:34<1:17:42, 5.19it/s] 50%|████▉ | 23800/48008 [1:21:34<1:17:42, 5.19it/s]{'loss': 0.303, 'grad_norm': 0.1962890625, 'learning_rate': 0.00010085402432927847, 'epoch': 0.5} 50%|████▉ | 23801/48008 [1:21:34<1:19:12, 5.09it/s] 50%|████▉ | 23802/48008 [1:21:34<1:18:26, 5.14it/s] 50%|████▉ | 23803/48008 [1:21:34<1:18:01, 5.17it/s] 50%|████▉ | 23804/48008 [1:21:35<1:17:39, 5.19it/s] 50%|████▉ | 23805/48008 [1:21:35<1:17:15, 5.22it/s] 50%|████▉ | 23806/48008 [1:21:35<1:16:49, 5.25it/s] 50%|████▉ | 23807/48008 [1:21:35<1:16:29, 5.27it/s] 50%|████▉ | 23808/48008 [1:21:35<1:18:43, 5.12it/s] 50%|████▉ | 23809/48008 [1:21:36<1:18:21, 5.15it/s] 50%|████▉ | 23810/48008 [1:21:36<1:19:14, 5.09it/s] 50%|████▉ | 23811/48008 [1:21:36<1:18:26, 5.14it/s] 50%|████▉ | 23812/48008 [1:21:36<1:17:59, 5.17it/s] 50%|████▉ | 23813/48008 [1:21:36<1:17:25, 5.21it/s] 50%|████▉ | 23814/48008 [1:21:37<1:16:38, 5.26it/s] 50%|████▉ | 23815/48008 [1:21:37<1:16:40, 5.26it/s] 50%|████▉ | 23816/48008 [1:21:37<1:19:17, 5.09it/s] 50%|████▉ | 23817/48008 [1:21:37<1:19:07, 5.10it/s] 50%|████▉ | 23818/48008 [1:21:37<1:18:22, 5.14it/s] 50%|████▉ | 23819/48008 [1:21:38<1:17:56, 5.17it/s] 50%|████▉ | 23820/48008 [1:21:38<1:19:52, 5.05it/s] 50%|████▉ | 23821/48008 [1:21:38<1:18:45, 5.12it/s] 50%|████▉ | 23822/48008 [1:21:38<1:18:13, 5.15it/s] 50%|████▉ | 23823/48008 [1:21:38<1:17:58, 5.17it/s] 50%|████▉ | 23824/48008 [1:21:38<1:17:17, 5.21it/s] 50%|████▉ | 23825/48008 [1:21:39<1:17:08, 5.23it/s] 50%|████▉ | 23826/48008 [1:21:39<1:17:52, 5.18it/s] 50%|████▉ | 23827/48008 [1:21:39<1:17:01, 5.23it/s] 50%|████▉ | 23828/48008 [1:21:39<1:17:09, 5.22it/s] 50%|████▉ | 23829/48008 [1:21:39<1:17:05, 5.23it/s] 50%|████▉ | 23830/48008 [1:21:40<1:17:15, 5.22it/s] 50%|████▉ | 23831/48008 [1:21:40<1:17:16, 5.21it/s] 50%|████▉ | 23832/48008 [1:21:40<1:17:11, 5.22it/s] 50%|████▉ | 23833/48008 [1:21:40<1:17:20, 5.21it/s] 50%|████▉ | 23834/48008 [1:21:40<1:18:37, 5.12it/s] 50%|████▉ | 23835/48008 [1:21:41<1:18:13, 5.15it/s] 50%|████▉ | 23836/48008 [1:21:41<1:17:57, 5.17it/s] 50%|████▉ | 23837/48008 [1:21:41<1:17:54, 5.17it/s] 50%|████▉ | 23838/48008 [1:21:41<1:18:18, 5.14it/s] 50%|████▉ | 23839/48008 [1:21:41<1:18:36, 5.12it/s] 50%|████▉ | 23840/48008 [1:21:42<1:41:38, 3.96it/s] 50%|████▉ | 23841/48008 [1:21:42<1:34:14, 4.27it/s] 50%|████▉ | 23842/48008 [1:21:42<1:28:43, 4.54it/s] 50%|████▉ | 23843/48008 [1:21:42<1:24:51, 4.75it/s] 50%|████▉ | 23844/48008 [1:21:43<1:22:33, 4.88it/s] 50%|████▉ | 23845/48008 [1:21:43<1:22:29, 4.88it/s] 50%|████▉ | 23846/48008 [1:21:43<1:20:52, 4.98it/s] 50%|████▉ | 23847/48008 [1:21:43<1:19:56, 5.04it/s] 50%|████▉ | 23848/48008 [1:21:43<1:19:12, 5.08it/s] 50%|████▉ | 23849/48008 [1:21:44<1:19:18, 5.08it/s] 50%|████▉ | 23850/48008 [1:21:44<1:19:24, 5.07it/s] {'loss': 0.3014, 'grad_norm': 0.2080078125, 'learning_rate': 0.00010064572571238127, 'epoch': 0.5} 50%|████▉ | 23850/48008 [1:21:44<1:19:24, 5.07it/s] 50%|████▉ | 23851/48008 [1:21:44<1:19:17, 5.08it/s] 50%|████▉ | 23852/48008 [1:21:44<1:18:32, 5.13it/s] 50%|████▉ | 23853/48008 [1:21:44<1:32:05, 4.37it/s] 50%|████▉ | 23854/48008 [1:21:45<1:27:22, 4.61it/s] 50%|████▉ | 23855/48008 [1:21:45<1:23:59, 4.79it/s] 50%|████▉ | 23856/48008 [1:21:45<1:23:20, 4.83it/s] 50%|████▉ | 23857/48008 [1:21:45<1:21:06, 4.96it/s] 50%|████▉ | 23858/48008 [1:21:45<1:19:24, 5.07it/s] 50%|████▉ | 23859/48008 [1:21:46<1:19:56, 5.03it/s] 50%|████▉ | 23860/48008 [1:21:46<1:18:52, 5.10it/s] 50%|████▉ | 23861/48008 [1:21:46<1:18:06, 5.15it/s] 50%|████▉ | 23862/48008 [1:21:46<1:17:15, 5.21it/s] 50%|████▉ | 23863/48008 [1:21:46<1:17:14, 5.21it/s] 50%|████▉ | 23864/48008 [1:21:47<1:16:58, 5.23it/s] 50%|████▉ | 23865/48008 [1:21:47<1:17:04, 5.22it/s] 50%|████▉ | 23866/48008 [1:21:47<1:17:09, 5.22it/s] 50%|████▉ | 23867/48008 [1:21:47<1:16:56, 5.23it/s] 50%|████▉ | 23868/48008 [1:21:47<1:17:01, 5.22it/s] 50%|████▉ | 23869/48008 [1:21:47<1:16:48, 5.24it/s] 50%|████▉ | 23870/48008 [1:21:48<1:16:12, 5.28it/s] 50%|████▉ | 23871/48008 [1:21:48<1:17:41, 5.18it/s] 50%|████▉ | 23872/48008 [1:21:48<1:19:34, 5.05it/s] 50%|████▉ | 23873/48008 [1:21:48<1:18:42, 5.11it/s] 50%|████▉ | 23874/48008 [1:21:48<1:17:41, 5.18it/s] 50%|████▉ | 23875/48008 [1:21:49<1:17:10, 5.21it/s] 50%|████▉ | 23876/48008 [1:21:49<1:17:51, 5.17it/s] 50%|████▉ | 23877/48008 [1:21:49<1:16:46, 5.24it/s] 50%|████▉ | 23878/48008 [1:21:49<1:16:36, 5.25it/s] 50%|████▉ | 23879/48008 [1:21:49<1:16:39, 5.25it/s] 50%|████▉ | 23880/48008 [1:21:50<1:16:40, 5.24it/s] 50%|████▉ | 23881/48008 [1:21:50<1:30:42, 4.43it/s] 50%|████▉ | 23882/48008 [1:21:50<1:27:06, 4.62it/s] 50%|████▉ | 23883/48008 [1:21:50<1:23:59, 4.79it/s] 50%|████▉ | 23884/48008 [1:21:50<1:24:11, 4.78it/s] 50%|████▉ | 23885/48008 [1:21:51<1:23:21, 4.82it/s] 50%|████▉ | 23886/48008 [1:21:51<1:20:48, 4.98it/s] 50%|████▉ | 23887/48008 [1:21:51<1:19:12, 5.08it/s] 50%|████▉ | 23888/48008 [1:21:51<1:20:07, 5.02it/s] 50%|████▉ | 23889/48008 [1:21:51<1:19:08, 5.08it/s] 50%|████▉ | 23890/48008 [1:21:52<1:18:20, 5.13it/s] 50%|████▉ | 23891/48008 [1:21:52<1:17:35, 5.18it/s] 50%|████▉ | 23892/48008 [1:21:52<1:16:57, 5.22it/s] 50%|████▉ | 23893/48008 [1:21:52<1:17:40, 5.17it/s] 50%|████▉ | 23894/48008 [1:21:52<1:17:23, 5.19it/s] 50%|████▉ | 23895/48008 [1:21:53<1:16:41, 5.24it/s] 50%|████▉ | 23896/48008 [1:21:53<1:18:57, 5.09it/s] 50%|████▉ | 23897/48008 [1:21:53<1:18:18, 5.13it/s] 50%|████▉ | 23898/48008 [1:21:53<1:18:24, 5.12it/s] 50%|████▉ | 23899/48008 [1:21:53<1:17:24, 5.19it/s] 50%|████▉ | 23900/48008 [1:21:54<1:17:21, 5.19it/s] {'loss': 0.3031, 'grad_norm': 0.1318359375, 'learning_rate': 0.00010043742709548409, 'epoch': 0.5} 50%|████▉ | 23900/48008 [1:21:54<1:17:21, 5.19it/s] 50%|████▉ | 23901/48008 [1:21:54<1:18:25, 5.12it/s] 50%|████▉ | 23902/48008 [1:21:54<1:18:38, 5.11it/s] 50%|████▉ | 23903/48008 [1:21:54<1:20:25, 5.00it/s] 50%|████▉ | 23904/48008 [1:21:54<1:19:26, 5.06it/s] 50%|████▉ | 23905/48008 [1:21:55<1:17:44, 5.17it/s] 50%|████▉ | 23906/48008 [1:21:55<1:17:19, 5.20it/s] 50%|████▉ | 23907/48008 [1:21:55<1:18:30, 5.12it/s] 50%|████▉ | 23908/48008 [1:21:55<1:17:56, 5.15it/s] 50%|████▉ | 23909/48008 [1:21:55<1:18:21, 5.13it/s] 50%|████▉ | 23910/48008 [1:21:56<1:18:31, 5.12it/s] 50%|████▉ | 23911/48008 [1:21:56<1:20:07, 5.01it/s] 50%|████▉ | 23912/48008 [1:21:56<1:18:55, 5.09it/s] 50%|████▉ | 23913/48008 [1:21:56<1:18:08, 5.14it/s] 50%|████▉ | 23914/48008 [1:21:56<1:18:58, 5.08it/s] 50%|████▉ | 23915/48008 [1:21:57<1:18:34, 5.11it/s] 50%|████▉ | 23916/48008 [1:21:57<1:18:39, 5.10it/s] 50%|████▉ | 23917/48008 [1:21:57<1:17:54, 5.15it/s] 50%|████▉ | 23918/48008 [1:21:57<1:18:01, 5.15it/s] 50%|████▉ | 23919/48008 [1:21:57<1:17:28, 5.18it/s] 50%|████▉ | 23920/48008 [1:21:57<1:17:14, 5.20it/s] 50%|████▉ | 23921/48008 [1:21:58<1:17:10, 5.20it/s] 50%|████▉ | 23922/48008 [1:21:58<1:16:52, 5.22it/s] 50%|████▉ | 23923/48008 [1:21:58<1:16:54, 5.22it/s] 50%|████▉ | 23924/48008 [1:21:58<1:18:46, 5.10it/s] 50%|████▉ | 23925/48008 [1:21:58<1:18:14, 5.13it/s] 50%|████▉ | 23926/48008 [1:21:59<1:17:49, 5.16it/s] 50%|████▉ | 23927/48008 [1:21:59<1:17:35, 5.17it/s] 50%|████▉ | 23928/48008 [1:21:59<1:17:15, 5.19it/s] 50%|████▉ | 23929/48008 [1:21:59<1:17:56, 5.15it/s] 50%|████▉ | 23930/48008 [1:21:59<1:18:11, 5.13it/s] 50%|████▉ | 23931/48008 [1:22:00<1:18:18, 5.12it/s] 50%|████▉ | 23932/48008 [1:22:00<1:16:40, 5.23it/s] 50%|████▉ | 23933/48008 [1:22:00<1:16:06, 5.27it/s] 50%|████▉ | 23934/48008 [1:22:00<1:21:47, 4.91it/s] 50%|████▉ | 23935/48008 [1:22:00<1:19:56, 5.02it/s] 50%|████▉ | 23936/48008 [1:22:01<1:20:14, 5.00it/s] 50%|████▉ | 23937/48008 [1:22:01<1:19:13, 5.06it/s] 50%|████▉ | 23938/48008 [1:22:01<1:17:56, 5.15it/s] 50%|████▉ | 23939/48008 [1:22:01<1:17:38, 5.17it/s] 50%|████▉ | 23940/48008 [1:22:01<1:18:14, 5.13it/s] 50%|████▉ | 23941/48008 [1:22:02<1:17:32, 5.17it/s] 50%|████▉ | 23942/48008 [1:22:02<1:17:13, 5.19it/s] 50%|████▉ | 23943/48008 [1:22:02<1:30:50, 4.42it/s] 50%|████▉ | 23944/48008 [1:22:02<1:26:35, 4.63it/s] 50%|████▉ | 23945/48008 [1:22:02<1:25:29, 4.69it/s] 50%|████▉ | 23946/48008 [1:22:03<1:23:34, 4.80it/s] 50%|████▉ | 23947/48008 [1:22:03<1:21:53, 4.90it/s] 50%|████▉ | 23948/48008 [1:22:03<1:20:59, 4.95it/s] 50%|████▉ | 23949/48008 [1:22:03<1:20:58, 4.95it/s] 50%|████▉ | 23950/48008 [1:22:03<1:19:39, 5.03it/s] {'loss': 0.3009, 'grad_norm': 0.1474609375, 'learning_rate': 0.00010022912847858689, 'epoch': 0.5} 50%|████▉ | 23950/48008 [1:22:03<1:19:39, 5.03it/s] 50%|████▉ | 23951/48008 [1:22:04<1:18:38, 5.10it/s] 50%|████▉ | 23952/48008 [1:22:04<1:17:55, 5.15it/s] 50%|████▉ | 23953/48008 [1:22:04<1:17:44, 5.16it/s] 50%|████▉ | 23954/48008 [1:22:04<1:17:27, 5.18it/s] 50%|████▉ | 23955/48008 [1:22:04<1:17:19, 5.18it/s] 50%|████▉ | 23956/48008 [1:22:05<1:17:13, 5.19it/s] 50%|████▉ | 23957/48008 [1:22:05<1:16:34, 5.23it/s] 50%|████▉ | 23958/48008 [1:22:05<1:19:36, 5.03it/s] 50%|████▉ | 23959/48008 [1:22:05<1:18:15, 5.12it/s] 50%|████▉ | 23960/48008 [1:22:05<1:17:46, 5.15it/s] 50%|████▉ | 23961/48008 [1:22:06<1:41:05, 3.96it/s] 50%|████▉ | 23962/48008 [1:22:06<1:33:42, 4.28it/s] 50%|████▉ | 23963/48008 [1:22:06<1:27:52, 4.56it/s] 50%|████▉ | 23964/48008 [1:22:06<1:24:34, 4.74it/s] 50%|████▉ | 23965/48008 [1:22:07<1:22:16, 4.87it/s] 50%|████▉ | 23966/48008 [1:22:07<1:21:56, 4.89it/s] 50%|████▉ | 23967/48008 [1:22:07<1:20:30, 4.98it/s] 50%|████▉ | 23968/48008 [1:22:07<1:24:47, 4.73it/s] 50%|████▉ | 23969/48008 [1:22:07<1:22:31, 4.85it/s] 50%|████▉ | 23970/48008 [1:22:08<1:22:43, 4.84it/s] 50%|████▉ | 23971/48008 [1:22:08<1:22:03, 4.88it/s] 50%|████▉ | 23972/48008 [1:22:08<1:20:24, 4.98it/s] 50%|████▉ | 23973/48008 [1:22:08<1:19:00, 5.07it/s] 50%|████▉ | 23974/48008 [1:22:08<1:18:01, 5.13it/s] 50%|████▉ | 23975/48008 [1:22:09<1:23:04, 4.82it/s] 50%|████▉ | 23976/48008 [1:22:09<1:21:20, 4.92it/s] 50%|████▉ | 23977/48008 [1:22:09<1:33:30, 4.28it/s] 50%|████▉ | 23978/48008 [1:22:09<1:28:21, 4.53it/s] 50%|████▉ | 23979/48008 [1:22:09<1:24:30, 4.74it/s] 50%|████▉ | 23980/48008 [1:22:10<1:22:08, 4.88it/s] 50%|████▉ | 23981/48008 [1:22:10<1:21:57, 4.89it/s] 50%|████▉ | 23982/48008 [1:22:10<1:19:57, 5.01it/s] 50%|████▉ | 23983/48008 [1:22:10<1:19:54, 5.01it/s] 50%|████▉ | 23984/48008 [1:22:10<1:19:17, 5.05it/s] 50%|████▉ | 23985/48008 [1:22:11<1:18:38, 5.09it/s] 50%|████▉ | 23986/48008 [1:22:11<1:19:24, 5.04it/s] 50%|████▉ | 23987/48008 [1:22:11<1:18:09, 5.12it/s] 50%|████▉ | 23988/48008 [1:22:11<1:31:41, 4.37it/s] 50%|████▉ | 23989/48008 [1:22:12<1:32:34, 4.32it/s] 50%|████▉ | 23990/48008 [1:22:12<1:28:17, 4.53it/s] 50%|████▉ | 23991/48008 [1:22:12<1:24:39, 4.73it/s] 50%|████▉ | 23992/48008 [1:22:12<1:21:39, 4.90it/s] 50%|████▉ | 23993/48008 [1:22:13<3:16:29, 2.04it/s] 50%|████▉ | 23994/48008 [1:22:13<2:40:17, 2.50it/s] 50%|████▉ | 23995/48008 [1:22:14<2:14:49, 2.97it/s] 50%|████▉ | 23996/48008 [1:22:14<1:57:12, 3.41it/s] 50%|████▉ | 23997/48008 [1:22:14<1:45:41, 3.79it/s] 50%|████▉ | 23998/48008 [1:22:14<1:36:45, 4.14it/s] 50%|████▉ | 23999/48008 [1:22:14<1:30:34, 4.42it/s] 50%|████▉ | 24000/48008 [1:22:15<1:27:41, 4.56it/s] {'loss': 0.3015, 'grad_norm': 0.3125, 'learning_rate': 0.00010002082986168972, 'epoch': 0.5} 50%|████▉ | 24000/48008 [1:22:15<1:27:41, 4.56it/s] 50%|████▉ | 24001/48008 [1:22:15<1:23:20, 4.80it/s] 50%|████▉ | 24002/48008 [1:22:15<1:21:11, 4.93it/s] 50%|████▉ | 24003/48008 [1:22:15<1:19:51, 5.01it/s] 50%|█████ | 24004/48008 [1:22:15<1:20:18, 4.98it/s] 50%|█████ | 24005/48008 [1:22:16<1:19:13, 5.05it/s] 50%|█████ | 24006/48008 [1:22:16<1:19:37, 5.02it/s] 50%|█████ | 24007/48008 [1:22:16<1:24:25, 4.74it/s] 50%|█████ | 24008/48008 [1:22:16<1:23:17, 4.80it/s] 50%|█████ | 24009/48008 [1:22:17<1:35:05, 4.21it/s] 50%|█████ | 24010/48008 [1:22:17<1:29:08, 4.49it/s] 50%|█████ | 24011/48008 [1:22:17<1:30:50, 4.40it/s] 50%|█████ | 24012/48008 [1:22:17<1:26:17, 4.63it/s] 50%|█████ | 24013/48008 [1:22:17<1:25:14, 4.69it/s] 50%|█████ | 24014/48008 [1:22:18<1:22:36, 4.84it/s] 50%|█████ | 24015/48008 [1:22:18<1:20:29, 4.97it/s] 50%|█████ | 24016/48008 [1:22:18<1:19:13, 5.05it/s] 50%|█████ | 24017/48008 [1:22:18<1:19:35, 5.02it/s] 50%|█████ | 24018/48008 [1:22:18<1:18:26, 5.10it/s] 50%|█████ | 24019/48008 [1:22:19<1:18:03, 5.12it/s] 50%|█████ | 24020/48008 [1:22:19<1:19:46, 5.01it/s] 50%|█████ | 24021/48008 [1:22:19<1:18:46, 5.08it/s] 50%|█████ | 24022/48008 [1:22:19<1:23:39, 4.78it/s] 50%|█████ | 24023/48008 [1:22:19<1:20:51, 4.94it/s] 50%|█████ | 24024/48008 [1:22:20<1:19:37, 5.02it/s] 50%|█████ | 24025/48008 [1:22:20<1:20:12, 4.98it/s] 50%|█████ | 24026/48008 [1:22:20<1:19:01, 5.06it/s] 50%|█████ | 24027/48008 [1:22:20<1:17:52, 5.13it/s] 50%|█████ | 24028/48008 [1:22:20<1:16:50, 5.20it/s] 50%|█████ | 24029/48008 [1:22:20<1:16:18, 5.24it/s] 50%|█████ | 24030/48008 [1:22:21<1:17:51, 5.13it/s] 50%|█████ | 24031/48008 [1:22:21<1:16:58, 5.19it/s] 50%|█████ | 24032/48008 [1:22:21<1:30:18, 4.42it/s] 50%|█████ | 24033/48008 [1:22:21<1:26:10, 4.64it/s] 50%|█████ | 24034/48008 [1:22:22<1:23:16, 4.80it/s] 50%|█████ | 24035/48008 [1:22:22<1:34:52, 4.21it/s] 50%|█████ | 24036/48008 [1:22:22<1:29:19, 4.47it/s] 50%|█████ | 24037/48008 [1:22:22<1:24:14, 4.74it/s] 50%|█████ | 24038/48008 [1:22:22<1:23:16, 4.80it/s] 50%|█████ | 24039/48008 [1:22:23<1:21:50, 4.88it/s] 50%|█████ | 24040/48008 [1:22:23<1:19:49, 5.00it/s] 50%|█████ | 24041/48008 [1:22:23<1:18:03, 5.12it/s] 50%|█████ | 24042/48008 [1:22:23<1:17:18, 5.17it/s] 50%|█████ | 24043/48008 [1:22:23<1:16:51, 5.20it/s] 50%|█████ | 24044/48008 [1:22:24<1:15:55, 5.26it/s] 50%|█████ | 24045/48008 [1:22:24<1:15:48, 5.27it/s] 50%|█████ | 24046/48008 [1:22:24<1:17:27, 5.16it/s] 50%|█████ | 24047/48008 [1:22:24<1:16:50, 5.20it/s] 50%|█████ | 24048/48008 [1:22:24<1:15:27, 5.29it/s] 50%|█████ | 24049/48008 [1:22:25<1:15:04, 5.32it/s] 50%|█████ | 24050/48008 [1:22:25<1:14:52, 5.33it/s] {'loss': 0.3011, 'grad_norm': 0.228515625, 'learning_rate': 9.981253124479254e-05, 'epoch': 0.5} 50%|█████ | 24050/48008 [1:22:25<1:14:52, 5.33it/s] 50%|█████ | 24051/48008 [1:22:25<1:15:11, 5.31it/s] 50%|█████ | 24052/48008 [1:22:25<1:15:02, 5.32it/s] 50%|█████ | 24053/48008 [1:22:25<1:15:19, 5.30it/s] 50%|█████ | 24054/48008 [1:22:25<1:16:50, 5.20it/s] 50%|█████ | 24055/48008 [1:22:26<1:16:12, 5.24it/s] 50%|█████ | 24056/48008 [1:22:26<1:16:08, 5.24it/s] 50%|█████ | 24057/48008 [1:22:26<1:16:54, 5.19it/s] 50%|█████ | 24058/48008 [1:22:26<1:19:15, 5.04it/s] 50%|█████ | 24059/48008 [1:22:26<1:17:52, 5.13it/s] 50%|█████ | 24060/48008 [1:22:27<1:18:53, 5.06it/s] 50%|█████ | 24061/48008 [1:22:27<1:41:32, 3.93it/s] 50%|█████ | 24062/48008 [1:22:27<1:33:39, 4.26it/s] 50%|█████ | 24063/48008 [1:22:27<1:28:11, 4.52it/s] 50%|█████ | 24064/48008 [1:22:28<1:25:00, 4.69it/s] 50%|█████ | 24065/48008 [1:22:28<1:22:47, 4.82it/s] 50%|█████ | 24066/48008 [1:22:28<1:20:33, 4.95it/s] 50%|█████ | 24067/48008 [1:22:28<1:19:36, 5.01it/s] 50%|█████ | 24068/48008 [1:22:28<1:18:26, 5.09it/s] 50%|█████ | 24069/48008 [1:22:29<1:19:05, 5.04it/s] 50%|█████ | 24070/48008 [1:22:29<1:18:59, 5.05it/s] 50%|█████ | 24071/48008 [1:22:29<1:17:55, 5.12it/s] 50%|█████ | 24072/48008 [1:22:29<1:17:25, 5.15it/s] 50%|█████ | 24073/48008 [1:22:29<1:18:30, 5.08it/s] 50%|█████ | 24074/48008 [1:22:30<1:17:40, 5.14it/s] 50%|█████ | 24075/48008 [1:22:30<1:16:34, 5.21it/s] 50%|█████ | 24076/48008 [1:22:30<1:16:29, 5.21it/s] 50%|█████ | 24077/48008 [1:22:30<1:16:21, 5.22it/s] 50%|█████ | 24078/48008 [1:22:30<1:16:15, 5.23it/s] 50%|█████ | 24079/48008 [1:22:31<1:15:44, 5.27it/s] 50%|█████ | 24080/48008 [1:22:31<1:16:28, 5.21it/s] 50%|█████ | 24081/48008 [1:22:31<1:15:39, 5.27it/s] 50%|█████ | 24082/48008 [1:22:31<1:15:43, 5.27it/s] 50%|█████ | 24083/48008 [1:22:31<1:21:34, 4.89it/s] 50%|█████ | 24084/48008 [1:22:32<1:21:22, 4.90it/s] 50%|█████ | 24085/48008 [1:22:32<1:21:26, 4.90it/s] 50%|█████ | 24086/48008 [1:22:32<1:19:48, 5.00it/s] 50%|█████ | 24087/48008 [1:22:32<1:18:23, 5.09it/s] 50%|█████ | 24088/48008 [1:22:32<1:17:34, 5.14it/s] 50%|█████ | 24089/48008 [1:22:33<1:30:45, 4.39it/s] 50%|█████ | 24090/48008 [1:22:33<1:26:12, 4.62it/s] 50%|█████ | 24091/48008 [1:22:33<1:23:55, 4.75it/s] 50%|█████ | 24092/48008 [1:22:33<1:21:02, 4.92it/s] 50%|█████ | 24093/48008 [1:22:33<1:19:29, 5.01it/s] 50%|█████ | 24094/48008 [1:22:34<1:23:54, 4.75it/s] 50%|█████ | 24095/48008 [1:22:34<1:21:15, 4.90it/s] 50%|█████ | 24096/48008 [1:22:34<1:20:25, 4.96it/s] 50%|█████ | 24097/48008 [1:22:34<1:18:51, 5.05it/s] 50%|█████ | 24098/48008 [1:22:34<1:16:40, 5.20it/s] 50%|█████ | 24099/48008 [1:22:35<1:21:55, 4.86it/s] 50%|█████ | 24100/48008 [1:22:35<1:20:27, 4.95it/s] {'loss': 0.3026, 'grad_norm': 0.369140625, 'learning_rate': 9.960423262789536e-05, 'epoch': 0.5} 50%|█████ | 24100/48008 [1:22:35<1:20:27, 4.95it/s] 50%|█████ | 24101/48008 [1:22:35<1:18:42, 5.06it/s] 50%|█████ | 24102/48008 [1:22:35<1:41:14, 3.94it/s] 50%|█████ | 24103/48008 [1:22:36<1:34:14, 4.23it/s] 50%|█████ | 24104/48008 [1:22:36<1:34:04, 4.24it/s] 50%|█████ | 24105/48008 [1:22:36<1:28:31, 4.50it/s] 50%|█████ | 24106/48008 [1:22:36<1:26:17, 4.62it/s] 50%|█████ | 24107/48008 [1:22:36<1:22:36, 4.82it/s] 50%|█████ | 24108/48008 [1:22:37<1:26:12, 4.62it/s] 50%|█████ | 24109/48008 [1:22:37<1:23:10, 4.79it/s] 50%|█████ | 24110/48008 [1:22:37<1:21:00, 4.92it/s] 50%|█████ | 24111/48008 [1:22:37<1:20:51, 4.93it/s] 50%|█████ | 24112/48008 [1:22:38<1:33:07, 4.28it/s] 50%|█████ | 24113/48008 [1:22:38<1:28:31, 4.50it/s] 50%|█████ | 24114/48008 [1:22:38<1:24:26, 4.72it/s] 50%|█████ | 24115/48008 [1:22:38<1:21:49, 4.87it/s] 50%|█████ | 24116/48008 [1:22:38<1:19:09, 5.03it/s] 50%|█████ | 24117/48008 [1:22:38<1:19:51, 4.99it/s] 50%|█████ | 24118/48008 [1:22:39<1:19:08, 5.03it/s] 50%|█████ | 24119/48008 [1:22:39<1:18:14, 5.09it/s] 50%|█████ | 24120/48008 [1:22:39<1:19:09, 5.03it/s] 50%|█████ | 24121/48008 [1:22:39<1:19:09, 5.03it/s] 50%|█████ | 24122/48008 [1:22:39<1:17:26, 5.14it/s] 50%|█████ | 24123/48008 [1:22:40<1:17:29, 5.14it/s] 50%|█████ | 24124/48008 [1:22:40<1:17:06, 5.16it/s] 50%|█████ | 24125/48008 [1:22:40<1:30:32, 4.40it/s] 50%|█████ | 24126/48008 [1:22:40<1:25:34, 4.65it/s] 50%|█████ | 24127/48008 [1:22:41<1:22:15, 4.84it/s] 50%|█████ | 24128/48008 [1:22:41<1:22:37, 4.82it/s] 50%|█████ | 24129/48008 [1:22:41<1:20:32, 4.94it/s] 50%|█████ | 24130/48008 [1:22:41<1:18:54, 5.04it/s] 50%|█████ | 24131/48008 [1:22:41<1:17:26, 5.14it/s] 50%|█████ | 24132/48008 [1:22:41<1:16:46, 5.18it/s] 50%|█████ | 24133/48008 [1:22:42<1:16:39, 5.19it/s] 50%|█████ | 24134/48008 [1:22:42<1:17:44, 5.12it/s] 50%|█████ | 24135/48008 [1:22:42<1:17:14, 5.15it/s] 50%|█████ | 24136/48008 [1:22:42<1:40:06, 3.97it/s] 50%|█████ | 24137/48008 [1:22:43<1:35:21, 4.17it/s] 50%|█████ | 24138/48008 [1:22:43<1:30:44, 4.38it/s] 50%|█████ | 24139/48008 [1:22:43<1:25:51, 4.63it/s] 50%|█████ | 24140/48008 [1:22:43<1:24:12, 4.72it/s] 50%|█████ | 24141/48008 [1:22:43<1:22:06, 4.84it/s] 50%|█████ | 24142/48008 [1:22:44<1:22:02, 4.85it/s] 50%|█████ | 24143/48008 [1:22:44<1:21:34, 4.88it/s] 50%|█████ | 24144/48008 [1:22:44<1:21:17, 4.89it/s] 50%|█████ | 24145/48008 [1:22:44<1:21:48, 4.86it/s] 50%|█████ | 24146/48008 [1:22:44<1:20:11, 4.96it/s] 50%|█████ | 24147/48008 [1:22:45<1:20:30, 4.94it/s] 50%|█████ | 24148/48008 [1:22:45<1:18:39, 5.06it/s] 50%|█████ | 24149/48008 [1:22:45<1:17:28, 5.13it/s] 50%|█████ | 24150/48008 [1:22:45<1:19:05, 5.03it/s] {'loss': 0.298, 'grad_norm': 0.1904296875, 'learning_rate': 9.939593401099818e-05, 'epoch': 0.5} 50%|█████ | 24150/48008 [1:22:45<1:19:05, 5.03it/s] 50%|█████ | 24151/48008 [1:22:45<1:18:16, 5.08it/s] 50%|█████ | 24152/48008 [1:22:46<1:22:57, 4.79it/s] 50%|█████ | 24153/48008 [1:22:46<1:22:21, 4.83it/s] 50%|█████ | 24154/48008 [1:22:46<1:20:13, 4.96it/s] 50%|█████ | 24155/48008 [1:22:46<1:19:58, 4.97it/s] 50%|█████ | 24156/48008 [1:22:46<1:18:59, 5.03it/s] 50%|█████ | 24157/48008 [1:22:47<1:19:29, 5.00it/s] 50%|█████ | 24158/48008 [1:22:47<1:18:04, 5.09it/s] 50%|█████ | 24159/48008 [1:22:47<1:22:51, 4.80it/s] 50%|█████ | 24160/48008 [1:22:47<1:21:06, 4.90it/s] 50%|█████ | 24161/48008 [1:22:47<1:19:33, 5.00it/s] 50%|█████ | 24162/48008 [1:22:48<2:52:54, 2.30it/s] 50%|█████ | 24163/48008 [1:22:49<2:23:45, 2.76it/s] 50%|█████ | 24164/48008 [1:22:49<2:04:48, 3.18it/s] 50%|█████ | 24165/48008 [1:22:49<1:50:06, 3.61it/s] 50%|█████ | 24166/48008 [1:22:49<1:53:41, 3.50it/s] 50%|█████ | 24167/48008 [1:22:50<1:42:08, 3.89it/s] 50%|█████ | 24168/48008 [1:22:50<2:56:44, 2.25it/s] 50%|█████ | 24169/48008 [1:22:51<2:26:30, 2.71it/s] 50%|█████ | 24170/48008 [1:22:51<2:05:46, 3.16it/s] 50%|█████ | 24171/48008 [1:22:51<1:52:31, 3.53it/s] 50%|█████ | 24172/48008 [1:22:51<1:43:14, 3.85it/s] 50%|█████ | 24173/48008 [1:22:51<1:35:21, 4.17it/s] 50%|█████ | 24174/48008 [1:22:52<2:43:07, 2.44it/s] 50%|█████ | 24175/48008 [1:22:52<2:17:28, 2.89it/s] 50%|█████ | 24176/48008 [1:22:53<2:04:29, 3.19it/s] 50%|█████ | 24177/48008 [1:22:53<1:49:58, 3.61it/s] 50%|█████ | 24178/48008 [1:22:53<1:40:25, 3.96it/s] 50%|█████ | 24179/48008 [1:22:53<1:32:35, 4.29it/s] 50%|█████ | 24180/48008 [1:22:53<1:27:22, 4.54it/s] 50%|█████ | 24181/48008 [1:22:54<1:24:48, 4.68it/s] 50%|█████ | 24182/48008 [1:22:54<1:22:45, 4.80it/s] 50%|█████ | 24183/48008 [1:22:54<1:21:33, 4.87it/s] 50%|█████ | 24184/48008 [1:22:54<1:19:57, 4.97it/s] 50%|█████ | 24185/48008 [1:22:54<1:18:09, 5.08it/s] 50%|█████ | 24186/48008 [1:22:55<1:19:40, 4.98it/s] 50%|█████ | 24187/48008 [1:22:55<1:19:02, 5.02it/s] 50%|█████ | 24188/48008 [1:22:55<1:18:37, 5.05it/s] 50%|█████ | 24189/48008 [1:22:55<1:17:07, 5.15it/s] 50%|█████ | 24190/48008 [1:22:55<1:16:26, 5.19it/s] 50%|█████ | 24191/48008 [1:22:56<1:18:16, 5.07it/s] 50%|█████ | 24192/48008 [1:22:56<1:16:49, 5.17it/s] 50%|█████ | 24193/48008 [1:22:56<1:30:20, 4.39it/s] 50%|█████ | 24194/48008 [1:22:56<1:25:54, 4.62it/s] 50%|█████ | 24195/48008 [1:22:56<1:22:25, 4.82it/s] 50%|█████ | 24196/48008 [1:22:57<1:20:01, 4.96it/s] 50%|█████ | 24197/48008 [1:22:57<1:32:35, 4.29it/s] 50%|█████ | 24198/48008 [1:22:57<1:30:20, 4.39it/s] 50%|█████ | 24199/48008 [1:22:57<1:31:47, 4.32it/s] 50%|█████ | 24200/48008 [1:22:58<1:27:11, 4.55it/s] {'loss': 0.2995, 'grad_norm': 0.302734375, 'learning_rate': 9.918763539410099e-05, 'epoch': 0.5} 50%|█████ | 24200/48008 [1:22:58<1:27:11, 4.55it/s] 50%|█████ | 24201/48008 [1:22:58<1:24:19, 4.71it/s] 50%|█████ | 24202/48008 [1:22:58<1:21:38, 4.86it/s] 50%|█████ | 24203/48008 [1:22:58<1:43:01, 3.85it/s] 50%|█████ | 24204/48008 [1:22:59<2:49:53, 2.34it/s] 50%|█████ | 24205/48008 [1:22:59<2:21:47, 2.80it/s] 50%|█████ | 24206/48008 [1:23:00<2:02:52, 3.23it/s] 50%|█████ | 24207/48008 [1:23:00<1:49:26, 3.62it/s] 50%|█████ | 24208/48008 [1:23:00<1:40:03, 3.96it/s] 50%|█████ | 24209/48008 [1:23:00<1:38:19, 4.03it/s] 50%|█████ | 24210/48008 [1:23:00<1:31:19, 4.34it/s] 50%|█████ | 24211/48008 [1:23:01<1:27:03, 4.56it/s] 50%|█████ | 24212/48008 [1:23:01<1:23:56, 4.72it/s] 50%|█████ | 24213/48008 [1:23:01<1:44:41, 3.79it/s] 50%|█████ | 24214/48008 [1:23:01<1:35:46, 4.14it/s] 50%|█████ | 24215/48008 [1:23:02<1:30:32, 4.38it/s] 50%|█████ | 24216/48008 [1:23:02<1:49:19, 3.63it/s] 50%|█████ | 24217/48008 [1:23:02<1:40:16, 3.95it/s] 50%|█████ | 24218/48008 [1:23:02<1:38:44, 4.02it/s] 50%|█████ | 24219/48008 [1:23:03<1:31:58, 4.31it/s] 50%|█████ | 24220/48008 [1:23:03<1:28:56, 4.46it/s] 50%|█████ | 24221/48008 [1:23:03<1:25:52, 4.62it/s] 50%|█████ | 24222/48008 [1:23:03<1:22:14, 4.82it/s] 50%|█████ | 24223/48008 [1:23:03<1:20:29, 4.93it/s] 50%|█████ | 24224/48008 [1:23:04<1:21:25, 4.87it/s] 50%|█████ | 24225/48008 [1:23:04<1:20:18, 4.94it/s] 50%|█████ | 24226/48008 [1:23:04<1:19:14, 5.00it/s] 50%|█████ | 24227/48008 [1:23:04<1:41:25, 3.91it/s] 50%|█████ | 24228/48008 [1:23:05<1:33:30, 4.24it/s] 50%|█████ | 24229/48008 [1:23:05<1:27:57, 4.51it/s] 50%|█████ | 24230/48008 [1:23:05<1:24:31, 4.69it/s] 50%|█████ | 24231/48008 [1:23:05<1:22:02, 4.83it/s] 50%|█████ | 24232/48008 [1:23:05<1:20:35, 4.92it/s] 50%|█████ | 24233/48008 [1:23:05<1:18:01, 5.08it/s] 50%|█████ | 24234/48008 [1:23:06<1:17:17, 5.13it/s] 50%|█████ | 24235/48008 [1:23:06<1:17:27, 5.12it/s] 50%|█████ | 24236/48008 [1:23:06<1:17:49, 5.09it/s] 50%|█████ | 24237/48008 [1:23:06<1:17:13, 5.13it/s] 50%|█████ | 24238/48008 [1:23:06<1:15:56, 5.22it/s] 50%|█████ | 24239/48008 [1:23:07<1:15:37, 5.24it/s] 50%|█████ | 24240/48008 [1:23:07<1:15:37, 5.24it/s] 50%|█████ | 24241/48008 [1:23:07<1:15:39, 5.24it/s] 50%|█████ | 24242/48008 [1:23:07<1:15:55, 5.22it/s] 50%|█████ | 24243/48008 [1:23:07<1:16:48, 5.16it/s] 50%|█████ | 24244/48008 [1:23:08<1:16:30, 5.18it/s] 51%|█████ | 24245/48008 [1:23:08<1:17:25, 5.12it/s] 51%|█████ | 24246/48008 [1:23:08<1:19:00, 5.01it/s] 51%|█████ | 24247/48008 [1:23:08<1:17:50, 5.09it/s] 51%|█████ | 24248/48008 [1:23:08<1:17:14, 5.13it/s] 51%|█████ | 24249/48008 [1:23:09<1:16:41, 5.16it/s] 51%|█████ | 24250/48008 [1:23:09<1:17:20, 5.12it/s] {'loss': 0.3007, 'grad_norm': 0.1396484375, 'learning_rate': 9.89793367772038e-05, 'epoch': 0.51} 51%|█████ | 24250/48008 [1:23:09<1:17:20, 5.12it/s] 51%|█████ | 24251/48008 [1:23:09<1:17:06, 5.14it/s] 51%|█████ | 24252/48008 [1:23:09<1:16:45, 5.16it/s] 51%|█████ | 24253/48008 [1:23:09<1:16:20, 5.19it/s] 51%|█████ | 24254/48008 [1:23:10<1:15:58, 5.21it/s] 51%|█████ | 24255/48008 [1:23:10<1:16:04, 5.20it/s] 51%|█████ | 24256/48008 [1:23:10<1:15:25, 5.25it/s] 51%|█████ | 24257/48008 [1:23:10<1:15:16, 5.26it/s] 51%|█████ | 24258/48008 [1:23:10<1:16:39, 5.16it/s] 51%|█████ | 24259/48008 [1:23:10<1:16:16, 5.19it/s] 51%|█████ | 24260/48008 [1:23:11<1:15:40, 5.23it/s] 51%|█████ | 24261/48008 [1:23:11<1:16:07, 5.20it/s] 51%|█████ | 24262/48008 [1:23:11<1:16:01, 5.21it/s] 51%|█████ | 24263/48008 [1:23:11<1:15:45, 5.22it/s] 51%|█████ | 24264/48008 [1:23:11<1:16:54, 5.15it/s] 51%|█████ | 24265/48008 [1:23:12<1:16:36, 5.17it/s] 51%|█████ | 24266/48008 [1:23:12<1:17:41, 5.09it/s] 51%|█████ | 24267/48008 [1:23:12<1:30:53, 4.35it/s] 51%|█████ | 24268/48008 [1:23:12<1:26:13, 4.59it/s] 51%|█████ | 24269/48008 [1:23:13<1:46:18, 3.72it/s] 51%|█████ | 24270/48008 [1:23:13<1:37:18, 4.07it/s] 51%|█████ | 24271/48008 [1:23:13<1:30:30, 4.37it/s] 51%|█████ | 24272/48008 [1:23:13<1:25:46, 4.61it/s] 51%|█████ | 24273/48008 [1:23:13<1:22:41, 4.78it/s] 51%|█████ | 24274/48008 [1:23:14<1:20:05, 4.94it/s] 51%|█████ | 24275/48008 [1:23:14<1:18:32, 5.04it/s] 51%|█████ | 24276/48008 [1:23:14<1:18:56, 5.01it/s] 51%|█████ | 24277/48008 [1:23:14<1:16:48, 5.15it/s] 51%|█████ | 24278/48008 [1:23:14<1:17:46, 5.09it/s] 51%|█████ | 24279/48008 [1:23:15<1:17:22, 5.11it/s] 51%|█████ | 24280/48008 [1:23:15<1:16:51, 5.14it/s] 51%|█████ | 24281/48008 [1:23:15<1:16:35, 5.16it/s] 51%|█████ | 24282/48008 [1:23:15<1:15:55, 5.21it/s] 51%|█████ | 24283/48008 [1:23:15<1:15:07, 5.26it/s] 51%|█████ | 24284/48008 [1:23:16<1:14:48, 5.28it/s] 51%|█████ | 24285/48008 [1:23:16<1:15:00, 5.27it/s] 51%|█████ | 24286/48008 [1:23:16<1:14:34, 5.30it/s] 51%|█████ | 24287/48008 [1:23:16<1:14:49, 5.28it/s] 51%|█████ | 24288/48008 [1:23:16<1:14:28, 5.31it/s] 51%|█████ | 24289/48008 [1:23:17<1:15:05, 5.26it/s] 51%|█████ | 24290/48008 [1:23:17<1:15:27, 5.24it/s] 51%|█████ | 24291/48008 [1:23:17<1:15:58, 5.20it/s] 51%|█████ | 24292/48008 [1:23:17<1:16:03, 5.20it/s] 51%|█████ | 24293/48008 [1:23:17<1:18:21, 5.04it/s] 51%|█████ | 24294/48008 [1:23:18<1:19:02, 5.00it/s] 51%|█████ | 24295/48008 [1:23:18<1:17:49, 5.08it/s] 51%|█████ | 24296/48008 [1:23:18<1:17:22, 5.11it/s] 51%|█████ | 24297/48008 [1:23:18<1:16:42, 5.15it/s] 51%|█████ | 24298/48008 [1:23:18<1:15:48, 5.21it/s] 51%|█████ | 24299/48008 [1:23:18<1:15:17, 5.25it/s] 51%|█████ | 24300/48008 [1:23:19<1:16:37, 5.16it/s] {'loss': 0.3028, 'grad_norm': 0.54296875, 'learning_rate': 9.877103816030661e-05, 'epoch': 0.51} 51%|█████ | 24300/48008 [1:23:19<1:16:37, 5.16it/s] 51%|█████ | 24301/48008 [1:23:19<1:16:23, 5.17it/s] 51%|█████ | 24302/48008 [1:23:19<1:15:27, 5.24it/s] 51%|█████ | 24303/48008 [1:23:19<1:16:51, 5.14it/s] 51%|█████ | 24304/48008 [1:23:19<1:16:27, 5.17it/s] 51%|█████ | 24305/48008 [1:23:20<1:15:59, 5.20it/s] 51%|█████ | 24306/48008 [1:23:20<1:15:47, 5.21it/s] 51%|█████ | 24307/48008 [1:23:20<1:15:31, 5.23it/s] 51%|█████ | 24308/48008 [1:23:20<1:16:33, 5.16it/s] 51%|█████ | 24309/48008 [1:23:20<1:16:26, 5.17it/s] 51%|█████ | 24310/48008 [1:23:21<1:17:05, 5.12it/s] 51%|█████ | 24311/48008 [1:23:21<1:17:14, 5.11it/s] 51%|█████ | 24312/48008 [1:23:21<1:16:57, 5.13it/s] 51%|█████ | 24313/48008 [1:23:21<1:15:52, 5.21it/s] 51%|█████ | 24314/48008 [1:23:21<1:15:38, 5.22it/s] 51%|█████ | 24315/48008 [1:23:22<1:20:49, 4.89it/s] 51%|█████ | 24316/48008 [1:23:22<1:20:06, 4.93it/s] 51%|█████ | 24317/48008 [1:23:22<1:18:17, 5.04it/s] 51%|█████ | 24318/48008 [1:23:22<1:17:40, 5.08it/s] 51%|█████ | 24319/48008 [1:23:22<1:18:43, 5.02it/s] 51%|█████ | 24320/48008 [1:23:23<1:17:15, 5.11it/s] 51%|█████ | 24321/48008 [1:23:23<1:18:48, 5.01it/s] 51%|█████ | 24322/48008 [1:23:23<1:16:58, 5.13it/s] 51%|█████ | 24323/48008 [1:23:23<1:18:08, 5.05it/s] 51%|█████ | 24324/48008 [1:23:23<1:17:24, 5.10it/s] 51%|█████ | 24325/48008 [1:23:24<1:17:27, 5.10it/s] 51%|█████ | 24326/48008 [1:23:24<1:18:34, 5.02it/s] 51%|█████ | 24327/48008 [1:23:24<1:17:38, 5.08it/s] 51%|█████ | 24328/48008 [1:23:24<1:17:19, 5.10it/s] 51%|█████ | 24329/48008 [1:23:24<1:16:40, 5.15it/s] 51%|█████ | 24330/48008 [1:23:25<1:39:36, 3.96it/s] 51%|█████ | 24331/48008 [1:23:25<1:32:34, 4.26it/s] 51%|█████ | 24332/48008 [1:23:25<1:32:57, 4.25it/s] 51%|█████ | 24333/48008 [1:23:26<1:50:51, 3.56it/s] 51%|█████ | 24334/48008 [1:23:26<1:40:15, 3.94it/s] 51%|█████ | 24335/48008 [1:23:26<1:32:09, 4.28it/s] 51%|█████ | 24336/48008 [1:23:26<1:29:29, 4.41it/s] 51%|█████ | 24337/48008 [1:23:26<1:25:04, 4.64it/s] 51%|█████ | 24338/48008 [1:23:27<1:23:05, 4.75it/s] 51%|█████ | 24339/48008 [1:23:27<1:20:34, 4.90it/s] 51%|█████ | 24340/48008 [1:23:27<1:19:07, 4.99it/s] 51%|█████ | 24341/48008 [1:23:27<1:17:30, 5.09it/s] 51%|█████ | 24342/48008 [1:23:27<1:30:25, 4.36it/s] 51%|█████ | 24343/48008 [1:23:28<1:25:12, 4.63it/s] 51%|█████ | 24344/48008 [1:23:28<1:22:02, 4.81it/s] 51%|█████ | 24345/48008 [1:23:28<1:20:08, 4.92it/s] 51%|█████ | 24346/48008 [1:23:28<1:19:17, 4.97it/s] 51%|█████ | 24347/48008 [1:23:28<1:18:33, 5.02it/s] 51%|█████ | 24348/48008 [1:23:29<1:17:36, 5.08it/s] 51%|█████ | 24349/48008 [1:23:29<1:16:53, 5.13it/s] 51%|█████ | 24350/48008 [1:23:29<1:17:07, 5.11it/s] {'loss': 0.301, 'grad_norm': 0.25390625, 'learning_rate': 9.856273954340943e-05, 'epoch': 0.51} 51%|█████ | 24350/48008 [1:23:29<1:17:07, 5.11it/s] 51%|█████ | 24351/48008 [1:23:29<1:16:44, 5.14it/s] 51%|█████ | 24352/48008 [1:23:29<1:21:41, 4.83it/s] 51%|█████ | 24353/48008 [1:23:30<1:19:16, 4.97it/s] 51%|█████ | 24354/48008 [1:23:30<1:18:02, 5.05it/s] 51%|█████ | 24355/48008 [1:23:30<1:16:53, 5.13it/s] 51%|█████ | 24356/48008 [1:23:30<1:16:13, 5.17it/s] 51%|█████ | 24357/48008 [1:23:30<1:15:49, 5.20it/s] 51%|█████ | 24358/48008 [1:23:31<1:16:15, 5.17it/s] 51%|█████ | 24359/48008 [1:23:31<1:15:47, 5.20it/s] 51%|█████ | 24360/48008 [1:23:31<1:15:26, 5.22it/s] 51%|█████ | 24361/48008 [1:23:31<1:29:00, 4.43it/s] 51%|█████ | 24362/48008 [1:23:31<1:25:01, 4.64it/s] 51%|█████ | 24363/48008 [1:23:32<1:27:26, 4.51it/s] 51%|█████ | 24364/48008 [1:23:32<1:46:59, 3.68it/s] 51%|█████ | 24365/48008 [1:23:33<2:58:57, 2.20it/s] 51%|█████ | 24366/48008 [1:23:33<2:28:12, 2.66it/s] 51%|█████ | 24367/48008 [1:23:33<2:08:37, 3.06it/s] 51%|█████ | 24368/48008 [1:23:34<1:53:44, 3.46it/s] 51%|█████ | 24369/48008 [1:23:34<1:42:43, 3.84it/s] 51%|█████ | 24370/48008 [1:23:34<1:33:58, 4.19it/s] 51%|█████ | 24371/48008 [1:23:34<1:28:48, 4.44it/s] 51%|█████ | 24372/48008 [1:23:34<1:24:39, 4.65it/s] 51%|█████ | 24373/48008 [1:23:34<1:21:57, 4.81it/s] 51%|█████ | 24374/48008 [1:23:35<1:19:59, 4.92it/s] 51%|█████ | 24375/48008 [1:23:35<1:20:39, 4.88it/s] 51%|█████ | 24376/48008 [1:23:35<1:19:53, 4.93it/s] 51%|█████ | 24377/48008 [1:23:35<1:18:22, 5.03it/s] 51%|█████ | 24378/48008 [1:23:35<1:18:24, 5.02it/s] 51%|█████ | 24379/48008 [1:23:36<1:18:32, 5.01it/s] 51%|█████ | 24380/48008 [1:23:36<1:40:36, 3.91it/s] 51%|█████ | 24381/48008 [1:23:36<1:34:21, 4.17it/s] 51%|█████ | 24382/48008 [1:23:36<1:28:36, 4.44it/s] 51%|█████ | 24383/48008 [1:23:37<1:24:35, 4.65it/s] 51%|█████ | 24384/48008 [1:23:37<1:21:46, 4.81it/s] 51%|█████ | 24385/48008 [1:23:37<1:21:13, 4.85it/s] 51%|█████ | 24386/48008 [1:23:37<1:20:50, 4.87it/s] 51%|█████ | 24387/48008 [1:23:37<1:19:12, 4.97it/s] 51%|█████ | 24388/48008 [1:23:38<1:17:45, 5.06it/s] 51%|█████ | 24389/48008 [1:23:38<1:16:49, 5.12it/s] 51%|█████ | 24390/48008 [1:23:38<1:21:51, 4.81it/s] 51%|█████ | 24391/48008 [1:23:38<1:19:41, 4.94it/s] 51%|█████ | 24392/48008 [1:23:38<1:20:02, 4.92it/s] 51%|█████ | 24393/48008 [1:23:39<1:18:52, 4.99it/s] 51%|█████ | 24394/48008 [1:23:39<1:17:46, 5.06it/s] 51%|█████ | 24395/48008 [1:23:39<1:17:05, 5.11it/s] 51%|█████ | 24396/48008 [1:23:39<1:17:22, 5.09it/s] 51%|█████ | 24397/48008 [1:23:39<1:16:15, 5.16it/s] 51%|█████ | 24398/48008 [1:23:40<1:16:04, 5.17it/s] 51%|█████ | 24399/48008 [1:23:40<1:15:47, 5.19it/s] 51%|█████ | 24400/48008 [1:23:40<1:29:12, 4.41it/s] {'loss': 0.3002, 'grad_norm': 0.66015625, 'learning_rate': 9.835444092651225e-05, 'epoch': 0.51} 51%|█████ | 24400/48008 [1:23:40<1:29:12, 4.41it/s] 51%|█████ | 24401/48008 [1:23:40<1:24:35, 4.65it/s] 51%|█████ | 24402/48008 [1:23:40<1:21:59, 4.80it/s] 51%|█████ | 24403/48008 [1:23:41<1:19:32, 4.95it/s] 51%|█████ | 24404/48008 [1:23:41<1:18:04, 5.04it/s] 51%|█████ | 24405/48008 [1:23:41<1:16:57, 5.11it/s] 51%|█████ | 24406/48008 [1:23:41<1:16:26, 5.15it/s] 51%|█████ | 24407/48008 [1:23:41<1:16:43, 5.13it/s] 51%|█████ | 24408/48008 [1:23:42<1:16:15, 5.16it/s] 51%|█████ | 24409/48008 [1:23:42<1:17:25, 5.08it/s] 51%|█████ | 24410/48008 [1:23:42<1:17:47, 5.06it/s] 51%|█████ | 24411/48008 [1:23:42<1:40:01, 3.93it/s] 51%|█████ | 24412/48008 [1:23:43<1:32:14, 4.26it/s] 51%|█████ | 24413/48008 [1:23:43<1:27:03, 4.52it/s] 51%|█████ | 24414/48008 [1:23:43<1:23:22, 4.72it/s] 51%|█████ | 24415/48008 [1:23:43<1:26:22, 4.55it/s] 51%|█████ | 24416/48008 [1:23:43<1:25:04, 4.62it/s] 51%|█████ | 24417/48008 [1:23:44<1:22:08, 4.79it/s] 51%|█████ | 24418/48008 [1:23:44<1:21:11, 4.84it/s] 51%|█████ | 24419/48008 [1:23:44<1:19:08, 4.97it/s] 51%|█████ | 24420/48008 [1:23:44<1:17:30, 5.07it/s] 51%|█████ | 24421/48008 [1:23:44<1:17:14, 5.09it/s] 51%|█████ | 24422/48008 [1:23:45<1:17:27, 5.07it/s] 51%|█████ | 24423/48008 [1:23:45<1:18:11, 5.03it/s] 51%|█████ | 24424/48008 [1:23:45<1:19:35, 4.94it/s] 51%|█████ | 24425/48008 [1:23:45<1:18:49, 4.99it/s] 51%|█████ | 24426/48008 [1:23:45<1:17:53, 5.05it/s] 51%|█████ | 24427/48008 [1:23:46<1:17:06, 5.10it/s] 51%|█████ | 24428/48008 [1:23:46<1:16:38, 5.13it/s] 51%|█████ | 24429/48008 [1:23:46<1:16:22, 5.15it/s] 51%|█████ | 24430/48008 [1:23:46<1:16:43, 5.12it/s] 51%|█████ | 24431/48008 [1:23:46<1:16:53, 5.11it/s] 51%|█████ | 24432/48008 [1:23:47<1:16:16, 5.15it/s] 51%|█████ | 24433/48008 [1:23:47<1:15:49, 5.18it/s] 51%|█████ | 24434/48008 [1:23:47<1:14:58, 5.24it/s] 51%|█████ | 24435/48008 [1:23:47<1:16:27, 5.14it/s] 51%|█████ | 24436/48008 [1:23:47<1:15:35, 5.20it/s] 51%|█████ | 24437/48008 [1:23:48<1:28:58, 4.42it/s] 51%|█████ | 24438/48008 [1:23:48<1:24:48, 4.63it/s] 51%|█████ | 24439/48008 [1:23:48<1:21:45, 4.80it/s] 51%|█████ | 24440/48008 [1:23:48<1:20:54, 4.86it/s] 51%|█████ | 24441/48008 [1:23:48<1:19:09, 4.96it/s] 51%|█████ | 24442/48008 [1:23:49<1:41:08, 3.88it/s] 51%|█████ | 24443/48008 [1:23:49<1:33:04, 4.22it/s] 51%|█████ | 24444/48008 [1:23:49<1:27:50, 4.47it/s] 51%|█████ | 24445/48008 [1:23:49<1:29:47, 4.37it/s] 51%|█████ | 24446/48008 [1:23:50<1:25:27, 4.60it/s] 51%|█████ | 24447/48008 [1:23:50<1:22:18, 4.77it/s] 51%|█████ | 24448/48008 [1:23:50<1:21:29, 4.82it/s] 51%|█████ | 24449/48008 [1:23:50<1:20:20, 4.89it/s] 51%|█████ | 24450/48008 [1:23:50<1:18:06, 5.03it/s] {'loss': 0.2995, 'grad_norm': 0.19921875, 'learning_rate': 9.814614230961507e-05, 'epoch': 0.51} 51%|█████ | 24450/48008 [1:23:50<1:18:06, 5.03it/s] 51%|█████ | 24451/48008 [1:23:51<1:17:32, 5.06it/s] 51%|█████ | 24452/48008 [1:23:51<1:16:28, 5.13it/s] 51%|█████ | 24453/48008 [1:23:51<1:15:55, 5.17it/s] 51%|█████ | 24454/48008 [1:23:51<1:15:33, 5.20it/s] 51%|█████ | 24455/48008 [1:23:51<1:14:53, 5.24it/s] 51%|█████ | 24456/48008 [1:23:52<1:16:31, 5.13it/s] 51%|█████ | 24457/48008 [1:23:52<1:14:58, 5.24it/s] 51%|█████ | 24458/48008 [1:23:52<1:16:13, 5.15it/s] 51%|█████ | 24459/48008 [1:23:52<1:38:49, 3.97it/s] 51%|█████ | 24460/48008 [1:23:52<1:31:49, 4.27it/s] 51%|█████ | 24461/48008 [1:23:53<1:28:05, 4.45it/s] 51%|█████ | 24462/48008 [1:23:53<1:23:56, 4.67it/s] 51%|█████ | 24463/48008 [1:23:53<1:26:41, 4.53it/s] 51%|█████ | 24464/48008 [1:23:53<1:22:32, 4.75it/s] 51%|█████ | 24465/48008 [1:23:53<1:20:18, 4.89it/s] 51%|█████ | 24466/48008 [1:23:54<1:20:07, 4.90it/s] 51%|█████ | 24467/48008 [1:23:54<1:17:57, 5.03it/s] 51%|█████ | 24468/48008 [1:23:54<1:16:52, 5.10it/s] 51%|█████ | 24469/48008 [1:23:54<1:17:15, 5.08it/s] 51%|█████ | 24470/48008 [1:23:54<1:17:07, 5.09it/s] 51%|█████ | 24471/48008 [1:23:55<1:16:44, 5.11it/s] 51%|█████ | 24472/48008 [1:23:55<1:16:03, 5.16it/s] 51%|█████ | 24473/48008 [1:23:55<1:15:51, 5.17it/s] 51%|█████ | 24474/48008 [1:23:55<1:15:21, 5.20it/s] 51%|█████ | 24475/48008 [1:23:55<1:15:03, 5.22it/s] 51%|█████ | 24476/48008 [1:23:56<1:15:33, 5.19it/s] 51%|█████ | 24477/48008 [1:23:56<1:15:59, 5.16it/s] 51%|█████ | 24478/48008 [1:23:56<1:14:56, 5.23it/s] 51%|█████ | 24479/48008 [1:23:56<1:14:51, 5.24it/s] 51%|█████ | 24480/48008 [1:23:56<1:28:40, 4.42it/s] 51%|█████ | 24481/48008 [1:23:57<1:24:07, 4.66it/s] 51%|█████ | 24482/48008 [1:23:57<1:22:09, 4.77it/s] 51%|█████ | 24483/48008 [1:23:57<1:19:58, 4.90it/s] 51%|█████ | 24484/48008 [1:23:57<1:19:49, 4.91it/s] 51%|█████ | 24485/48008 [1:23:57<1:18:07, 5.02it/s] 51%|█████ | 24486/48008 [1:23:58<1:16:54, 5.10it/s] 51%|█████ | 24487/48008 [1:23:58<1:21:45, 4.79it/s] 51%|█████ | 24488/48008 [1:23:58<1:42:30, 3.82it/s] 51%|█████ | 24489/48008 [1:23:58<1:33:38, 4.19it/s] 51%|█████ | 24490/48008 [1:23:59<1:27:25, 4.48it/s] 51%|█████ | 24491/48008 [1:23:59<1:23:42, 4.68it/s] 51%|█████ | 24492/48008 [1:23:59<1:20:52, 4.85it/s] 51%|█████ | 24493/48008 [1:23:59<1:18:02, 5.02it/s] 51%|█████ | 24494/48008 [1:23:59<1:18:33, 4.99it/s] 51%|█████ | 24495/48008 [1:24:00<1:17:34, 5.05it/s] 51%|█████ | 24496/48008 [1:24:00<1:39:40, 3.93it/s] 51%|█████ | 24497/48008 [1:24:00<1:32:22, 4.24it/s] 51%|█████ | 24498/48008 [1:24:01<1:50:17, 3.55it/s] 51%|█████ | 24499/48008 [1:24:01<1:41:15, 3.87it/s] 51%|█████ | 24500/48008 [1:24:01<1:33:39, 4.18it/s] {'loss': 0.3002, 'grad_norm': 0.50390625, 'learning_rate': 9.793784369271788e-05, 'epoch': 0.51} 51%|█████ | 24500/48008 [1:24:01<1:33:39, 4.18it/s] 51%|█████ | 24501/48008 [1:24:01<1:27:38, 4.47it/s] 51%|█████ | 24502/48008 [1:24:01<1:24:19, 4.65it/s] 51%|█████ | 24503/48008 [1:24:02<1:20:59, 4.84it/s] 51%|█████ | 24504/48008 [1:24:02<1:21:19, 4.82it/s] 51%|█████ | 24505/48008 [1:24:02<1:18:40, 4.98it/s] 51%|█████ | 24506/48008 [1:24:02<1:17:58, 5.02it/s] 51%|█████ | 24507/48008 [1:24:02<1:17:07, 5.08it/s] 51%|█████ | 24508/48008 [1:24:03<1:16:28, 5.12it/s] 51%|█████ | 24509/48008 [1:24:03<1:16:13, 5.14it/s] 51%|█████ | 24510/48008 [1:24:03<1:15:49, 5.16it/s] 51%|█████ | 24511/48008 [1:24:03<1:15:19, 5.20it/s] 51%|█████ | 24512/48008 [1:24:03<1:16:55, 5.09it/s] 51%|█████ | 24513/48008 [1:24:03<1:17:34, 5.05it/s] 51%|█████ | 24514/48008 [1:24:04<1:16:49, 5.10it/s] 51%|█████ | 24515/48008 [1:24:04<1:16:16, 5.13it/s] 51%|█████ | 24516/48008 [1:24:04<1:17:45, 5.04it/s] 51%|█████ | 24517/48008 [1:24:04<1:16:24, 5.12it/s] 51%|█████ | 24518/48008 [1:24:04<1:16:16, 5.13it/s] 51%|█████ | 24519/48008 [1:24:05<1:15:56, 5.16it/s] 51%|█████ | 24520/48008 [1:24:05<1:15:24, 5.19it/s] 51%|█████ | 24521/48008 [1:24:05<1:16:19, 5.13it/s] 51%|█████ | 24522/48008 [1:24:05<1:16:00, 5.15it/s] 51%|█████ | 24523/48008 [1:24:05<1:15:32, 5.18it/s] 51%|█████ | 24524/48008 [1:24:06<1:15:26, 5.19it/s] 51%|█████ | 24525/48008 [1:24:06<1:15:34, 5.18it/s] 51%|█████ | 24526/48008 [1:24:06<1:18:13, 5.00it/s] 51%|█████ | 24527/48008 [1:24:06<1:18:06, 5.01it/s] 51%|█████ | 24528/48008 [1:24:06<1:17:16, 5.06it/s] 51%|█████ | 24529/48008 [1:24:07<1:17:30, 5.05it/s] 51%|█████ | 24530/48008 [1:24:07<1:18:30, 4.98it/s] 51%|█████ | 24531/48008 [1:24:07<1:22:48, 4.73it/s] 51%|█████ | 24532/48008 [1:24:07<1:20:27, 4.86it/s] 51%|█████ | 24533/48008 [1:24:07<1:18:23, 4.99it/s] 51%|█████ | 24534/48008 [1:24:08<1:17:25, 5.05it/s] 51%|█████ | 24535/48008 [1:24:08<1:16:36, 5.11it/s] 51%|█████ | 24536/48008 [1:24:08<1:15:59, 5.15it/s] 51%|█████ | 24537/48008 [1:24:08<1:15:56, 5.15it/s] 51%|█████ | 24538/48008 [1:24:08<1:16:25, 5.12it/s] 51%|█████ | 24539/48008 [1:24:09<1:15:57, 5.15it/s] 51%|█████ | 24540/48008 [1:24:09<1:15:20, 5.19it/s] 51%|█████ | 24541/48008 [1:24:09<1:15:38, 5.17it/s] 51%|█████ | 24542/48008 [1:24:09<1:28:56, 4.40it/s] 51%|█████ | 24543/48008 [1:24:09<1:24:45, 4.61it/s] 51%|█████ | 24544/48008 [1:24:10<1:22:13, 4.76it/s] 51%|█████ | 24545/48008 [1:24:10<1:20:12, 4.88it/s] 51%|█████ | 24546/48008 [1:24:10<1:18:31, 4.98it/s] 51%|█████ | 24547/48008 [1:24:10<1:18:00, 5.01it/s] 51%|█████ | 24548/48008 [1:24:10<1:18:21, 4.99it/s] 51%|█████ | 24549/48008 [1:24:11<1:18:24, 4.99it/s] 51%|█████ | 24550/48008 [1:24:11<1:17:33, 5.04it/s] {'loss': 0.3027, 'grad_norm': 0.353515625, 'learning_rate': 9.77295450758207e-05, 'epoch': 0.51} 51%|█████ | 24550/48008 [1:24:11<1:17:33, 5.04it/s] 51%|█████ | 24551/48008 [1:24:11<1:16:44, 5.09it/s] 51%|█████ | 24552/48008 [1:24:11<1:17:37, 5.04it/s] 51%|█████ | 24553/48008 [1:24:11<1:16:43, 5.09it/s] 51%|█████ | 24554/48008 [1:24:12<1:17:00, 5.08it/s] 51%|█████ | 24555/48008 [1:24:12<1:16:07, 5.13it/s] 51%|█████ | 24556/48008 [1:24:12<1:15:40, 5.17it/s] 51%|█████ | 24557/48008 [1:24:12<1:14:21, 5.26it/s] 51%|█████ | 24558/48008 [1:24:12<1:16:01, 5.14it/s] 51%|█████ | 24559/48008 [1:24:13<1:16:32, 5.11it/s] 51%|█████ | 24560/48008 [1:24:13<1:16:07, 5.13it/s] 51%|█████ | 24561/48008 [1:24:13<1:16:22, 5.12it/s] 51%|█████ | 24562/48008 [1:24:13<1:17:22, 5.05it/s] 51%|█████ | 24563/48008 [1:24:13<1:15:53, 5.15it/s] 51%|█████ | 24564/48008 [1:24:14<1:15:24, 5.18it/s] 51%|█████ | 24565/48008 [1:24:14<1:14:32, 5.24it/s] 51%|█████ | 24566/48008 [1:24:14<1:14:53, 5.22it/s] 51%|█████ | 24567/48008 [1:24:14<1:15:12, 5.19it/s] 51%|█████ | 24568/48008 [1:24:14<1:13:41, 5.30it/s] 51%|█████ | 24569/48008 [1:24:15<1:13:55, 5.28it/s] 51%|█████ | 24570/48008 [1:24:15<1:14:19, 5.26it/s] 51%|█████ | 24571/48008 [1:24:15<1:14:09, 5.27it/s] 51%|█████ | 24572/48008 [1:24:15<1:13:30, 5.31it/s] 51%|█████ | 24573/48008 [1:24:15<1:13:20, 5.33it/s] 51%|█████ | 24574/48008 [1:24:15<1:13:55, 5.28it/s] 51%|█████ | 24575/48008 [1:24:16<1:14:06, 5.27it/s] 51%|█████ | 24576/48008 [1:24:16<1:13:38, 5.30it/s] 51%|█████ | 24577/48008 [1:24:16<1:13:33, 5.31it/s] 51%|█████ | 24578/48008 [1:24:16<1:13:43, 5.30it/s] 51%|█████ | 24579/48008 [1:24:16<1:14:05, 5.27it/s] 51%|█████ | 24580/48008 [1:24:17<1:14:11, 5.26it/s] 51%|█████ | 24581/48008 [1:24:17<1:13:07, 5.34it/s] 51%|█████ | 24582/48008 [1:24:17<1:13:29, 5.31it/s] 51%|█████ | 24583/48008 [1:24:17<1:14:02, 5.27it/s] 51%|█████ | 24584/48008 [1:24:18<1:37:14, 4.01it/s] 51%|█████ | 24585/48008 [1:24:18<1:31:16, 4.28it/s] 51%|█████ | 24586/48008 [1:24:18<1:27:57, 4.44it/s] 51%|█████ | 24587/48008 [1:24:18<1:24:19, 4.63it/s] 51%|█████ | 24588/48008 [1:24:18<1:35:01, 4.11it/s] 51%|█████ | 24589/48008 [1:24:19<1:28:57, 4.39it/s] 51%|█████ | 24590/48008 [1:24:19<1:25:29, 4.57it/s] 51%|█████ | 24591/48008 [1:24:19<1:22:55, 4.71it/s] 51%|█████ | 24592/48008 [1:24:19<1:20:58, 4.82it/s] 51%|█████ | 24593/48008 [1:24:19<1:19:14, 4.92it/s] 51%|█████ | 24594/48008 [1:24:20<1:18:02, 5.00it/s] 51%|█████ | 24595/48008 [1:24:20<1:17:01, 5.07it/s] 51%|█████ | 24596/48008 [1:24:20<1:15:52, 5.14it/s] 51%|█████ | 24597/48008 [1:24:20<1:15:02, 5.20it/s] 51%|█████ | 24598/48008 [1:24:20<1:14:10, 5.26it/s] 51%|█████ | 24599/48008 [1:24:21<1:13:53, 5.28it/s] 51%|█████ | 24600/48008 [1:24:21<1:15:16, 5.18it/s] {'loss': 0.3035, 'grad_norm': 0.51171875, 'learning_rate': 9.752124645892352e-05, 'epoch': 0.51} 51%|█████ | 24600/48008 [1:24:21<1:15:16, 5.18it/s] 51%|█████ | 24601/48008 [1:24:21<1:15:06, 5.19it/s] 51%|█████ | 24602/48008 [1:24:21<1:15:40, 5.16it/s] 51%|█████ | 24603/48008 [1:24:21<1:16:43, 5.08it/s] 51%|█████ | 24604/48008 [1:24:22<1:15:45, 5.15it/s] 51%|█████▏ | 24605/48008 [1:24:22<1:15:20, 5.18it/s] 51%|█████▏ | 24606/48008 [1:24:22<1:15:52, 5.14it/s] 51%|█████▏ | 24607/48008 [1:24:22<1:14:41, 5.22it/s] 51%|█████▏ | 24608/48008 [1:24:22<1:14:29, 5.24it/s] 51%|█████▏ | 24609/48008 [1:24:23<1:14:14, 5.25it/s] 51%|█████▏ | 24610/48008 [1:24:23<1:37:15, 4.01it/s] 51%|█████▏ | 24611/48008 [1:24:23<1:30:29, 4.31it/s] 51%|█████▏ | 24612/48008 [1:24:23<1:26:21, 4.52it/s] 51%|█████▏ | 24613/48008 [1:24:23<1:22:51, 4.71it/s] 51%|█████▏ | 24614/48008 [1:24:24<1:19:51, 4.88it/s] 51%|█████▏ | 24615/48008 [1:24:24<1:18:16, 4.98it/s] 51%|█████▏ | 24616/48008 [1:24:24<1:17:35, 5.02it/s] 51%|█████▏ | 24617/48008 [1:24:24<1:16:43, 5.08it/s] 51%|█████▏ | 24618/48008 [1:24:24<1:21:24, 4.79it/s] 51%|█████▏ | 24619/48008 [1:24:25<1:19:19, 4.91it/s] 51%|█████▏ | 24620/48008 [1:24:25<1:18:13, 4.98it/s] 51%|█████▏ | 24621/48008 [1:24:25<1:17:48, 5.01it/s] 51%|█████▏ | 24622/48008 [1:24:25<1:18:09, 4.99it/s] 51%|█████▏ | 24623/48008 [1:24:25<1:16:58, 5.06it/s] 51%|█████▏ | 24624/48008 [1:24:26<1:16:13, 5.11it/s] 51%|█████▏ | 24625/48008 [1:24:26<1:15:33, 5.16it/s] 51%|█████▏ | 24626/48008 [1:24:26<1:37:53, 3.98it/s] 51%|█████▏ | 24627/48008 [1:24:26<1:36:31, 4.04it/s] 51%|█████▏ | 24628/48008 [1:24:27<1:30:04, 4.33it/s] 51%|█████▏ | 24629/48008 [1:24:27<1:25:05, 4.58it/s] 51%|█████▏ | 24630/48008 [1:24:27<1:23:12, 4.68it/s] 51%|█████▏ | 24631/48008 [1:24:27<1:20:14, 4.86it/s] 51%|█████▏ | 24632/48008 [1:24:27<1:18:27, 4.97it/s] 51%|█████▏ | 24633/48008 [1:24:28<1:22:25, 4.73it/s] 51%|█████▏ | 24634/48008 [1:24:28<1:20:01, 4.87it/s] 51%|█████▏ | 24635/48008 [1:24:28<1:41:08, 3.85it/s] 51%|█████▏ | 24636/48008 [1:24:28<1:32:46, 4.20it/s] 51%|█████▏ | 24637/48008 [1:24:29<1:27:23, 4.46it/s] 51%|█████▏ | 24638/48008 [1:24:29<1:23:20, 4.67it/s] 51%|█████▏ | 24639/48008 [1:24:29<1:22:01, 4.75it/s] 51%|█████▏ | 24640/48008 [1:24:29<1:19:02, 4.93it/s] 51%|█████▏ | 24641/48008 [1:24:29<1:17:25, 5.03it/s] 51%|█████▏ | 24642/48008 [1:24:30<1:16:35, 5.08it/s] 51%|█████▏ | 24643/48008 [1:24:30<1:15:56, 5.13it/s] 51%|█████▏ | 24644/48008 [1:24:30<1:15:33, 5.15it/s] 51%|█████▏ | 24645/48008 [1:24:30<1:20:29, 4.84it/s] 51%|█████▏ | 24646/48008 [1:24:30<1:18:33, 4.96it/s] 51%|█████▏ | 24647/48008 [1:24:31<1:22:49, 4.70it/s] 51%|█████▏ | 24648/48008 [1:24:31<1:33:47, 4.15it/s] 51%|█████▏ | 24649/48008 [1:24:31<1:30:01, 4.32it/s] 51%|█████▏ | 24650/48008 [1:24:31<1:25:14, 4.57it/s] {'loss': 0.2975, 'grad_norm': 0.2119140625, 'learning_rate': 9.731294784202634e-05, 'epoch': 0.51} 51%|█████▏ | 24650/48008 [1:24:31<1:25:14, 4.57it/s] 51%|█████▏ | 24651/48008 [1:24:32<1:21:42, 4.76it/s] 51%|█████▏ | 24652/48008 [1:24:32<1:20:18, 4.85it/s] 51%|█████▏ | 24653/48008 [1:24:32<1:17:56, 4.99it/s] 51%|█████▏ | 24654/48008 [1:24:32<1:18:00, 4.99it/s] 51%|█████▏ | 24655/48008 [1:24:32<1:16:46, 5.07it/s] 51%|█████▏ | 24656/48008 [1:24:32<1:15:37, 5.15it/s] 51%|█████▏ | 24657/48008 [1:24:33<1:14:30, 5.22it/s] 51%|█████▏ | 24658/48008 [1:24:33<1:14:11, 5.25it/s] 51%|█████▏ | 24659/48008 [1:24:33<1:14:00, 5.26it/s] 51%|█████▏ | 24660/48008 [1:24:33<1:13:42, 5.28it/s] 51%|█████▏ | 24661/48008 [1:24:33<1:13:55, 5.26it/s] 51%|█████▏ | 24662/48008 [1:24:34<1:19:26, 4.90it/s] 51%|█████▏ | 24663/48008 [1:24:34<1:19:04, 4.92it/s] 51%|█████▏ | 24664/48008 [1:24:34<1:19:21, 4.90it/s] 51%|█████▏ | 24665/48008 [1:24:34<1:17:45, 5.00it/s] 51%|█████▏ | 24666/48008 [1:24:34<1:16:17, 5.10it/s] 51%|█████▏ | 24667/48008 [1:24:35<1:15:48, 5.13it/s] 51%|█████▏ | 24668/48008 [1:24:35<1:20:51, 4.81it/s] 51%|█████▏ | 24669/48008 [1:24:35<1:18:16, 4.97it/s] 51%|█████▏ | 24670/48008 [1:24:35<1:17:45, 5.00it/s] 51%|█████▏ | 24671/48008 [1:24:35<1:16:28, 5.09it/s] 51%|█████▏ | 24672/48008 [1:24:36<1:15:56, 5.12it/s] 51%|█████▏ | 24673/48008 [1:24:36<1:16:54, 5.06it/s] 51%|█████▏ | 24674/48008 [1:24:36<1:18:08, 4.98it/s] 51%|█████▏ | 24675/48008 [1:24:36<1:17:03, 5.05it/s] 51%|█████▏ | 24676/48008 [1:24:36<1:15:42, 5.14it/s] 51%|█████▏ | 24677/48008 [1:24:37<1:16:09, 5.11it/s] 51%|█████▏ | 24678/48008 [1:24:37<1:15:05, 5.18it/s] 51%|█████▏ | 24679/48008 [1:24:37<1:14:38, 5.21it/s] 51%|█████▏ | 24680/48008 [1:24:37<1:14:34, 5.21it/s] 51%|█████▏ | 24681/48008 [1:24:37<1:14:30, 5.22it/s] 51%|█████▏ | 24682/48008 [1:24:38<1:14:40, 5.21it/s] 51%|█████▏ | 24683/48008 [1:24:38<1:14:11, 5.24it/s] 51%|█████▏ | 24684/48008 [1:24:38<1:14:15, 5.24it/s] 51%|█████▏ | 24685/48008 [1:24:38<1:14:26, 5.22it/s] 51%|█████▏ | 24686/48008 [1:24:38<1:14:19, 5.23it/s] 51%|█████▏ | 24687/48008 [1:24:39<1:13:42, 5.27it/s] 51%|█████▏ | 24688/48008 [1:24:39<1:13:32, 5.28it/s] 51%|█████▏ | 24689/48008 [1:24:39<1:13:36, 5.28it/s] 51%|█████▏ | 24690/48008 [1:24:39<1:15:15, 5.16it/s] 51%|█████▏ | 24691/48008 [1:24:39<1:14:44, 5.20it/s] 51%|█████▏ | 24692/48008 [1:24:40<1:27:53, 4.42it/s] 51%|█████▏ | 24693/48008 [1:24:40<1:23:35, 4.65it/s] 51%|█████▏ | 24694/48008 [1:24:40<1:20:44, 4.81it/s] 51%|█████▏ | 24695/48008 [1:24:40<1:18:35, 4.94it/s] 51%|█████▏ | 24696/48008 [1:24:40<1:22:39, 4.70it/s] 51%|█████▏ | 24697/48008 [1:24:41<1:20:18, 4.84it/s] 51%|█████▏ | 24698/48008 [1:24:41<1:18:09, 4.97it/s] 51%|█████▏ | 24699/48008 [1:24:41<1:16:25, 5.08it/s] 51%|█████▏ | 24700/48008 [1:24:41<1:15:48, 5.12it/s] {'loss': 0.3028, 'grad_norm': 0.2890625, 'learning_rate': 9.710464922512915e-05, 'epoch': 0.51} 51%|█████▏ | 24700/48008 [1:24:41<1:15:48, 5.12it/s] 51%|█████▏ | 24701/48008 [1:24:41<1:14:18, 5.23it/s] 51%|█████▏ | 24702/48008 [1:24:42<1:14:16, 5.23it/s] 51%|█████▏ | 24703/48008 [1:24:42<1:13:32, 5.28it/s] 51%|█████▏ | 24704/48008 [1:24:42<1:13:56, 5.25it/s] 51%|█████▏ | 24705/48008 [1:24:42<1:13:27, 5.29it/s] 51%|█████▏ | 24706/48008 [1:24:42<1:36:20, 4.03it/s] 51%|█████▏ | 24707/48008 [1:24:43<1:52:21, 3.46it/s] 51%|█████▏ | 24708/48008 [1:24:43<1:41:01, 3.84it/s] 51%|█████▏ | 24709/48008 [1:24:43<1:34:55, 4.09it/s] 51%|█████▏ | 24710/48008 [1:24:43<1:28:47, 4.37it/s] 51%|█████▏ | 24711/48008 [1:24:44<1:23:52, 4.63it/s] 51%|█████▏ | 24712/48008 [1:24:44<1:20:54, 4.80it/s] 51%|█████▏ | 24713/48008 [1:24:44<1:18:46, 4.93it/s] 51%|█████▏ | 24714/48008 [1:24:44<1:17:01, 5.04it/s] 51%|█████▏ | 24715/48008 [1:24:44<1:21:34, 4.76it/s] 51%|█████▏ | 24716/48008 [1:24:45<1:19:00, 4.91it/s] 51%|█████▏ | 24717/48008 [1:24:45<1:18:46, 4.93it/s] 51%|█████▏ | 24718/48008 [1:24:45<1:16:46, 5.06it/s] 51%|█████▏ | 24719/48008 [1:24:45<1:17:25, 5.01it/s] 51%|█████▏ | 24720/48008 [1:24:45<1:15:39, 5.13it/s] 51%|█████▏ | 24721/48008 [1:24:46<1:16:35, 5.07it/s] 51%|█████▏ | 24722/48008 [1:24:46<1:15:51, 5.12it/s] 51%|█████▏ | 24723/48008 [1:24:46<1:20:36, 4.81it/s] 51%|█████▏ | 24724/48008 [1:24:46<1:18:01, 4.97it/s] 52%|█████▏ | 24725/48008 [1:24:46<1:18:19, 4.95it/s] 52%|█████▏ | 24726/48008 [1:24:47<1:17:09, 5.03it/s] 52%|█████▏ | 24727/48008 [1:24:47<1:17:36, 5.00it/s] 52%|█████▏ | 24728/48008 [1:24:47<1:16:45, 5.05it/s] 52%|█████▏ | 24729/48008 [1:24:47<1:16:26, 5.08it/s] 52%|█████▏ | 24730/48008 [1:24:47<1:16:19, 5.08it/s] 52%|█████▏ | 24731/48008 [1:24:48<1:16:13, 5.09it/s] 52%|█████▏ | 24732/48008 [1:24:48<1:17:39, 4.99it/s] 52%|█████▏ | 24733/48008 [1:24:48<1:16:34, 5.07it/s] 52%|█████▏ | 24734/48008 [1:24:48<1:15:33, 5.13it/s] 52%|█████▏ | 24735/48008 [1:24:48<1:15:25, 5.14it/s] 52%|█████▏ | 24736/48008 [1:24:49<1:20:18, 4.83it/s] 52%|█████▏ | 24737/48008 [1:24:49<1:18:09, 4.96it/s] 52%|█████▏ | 24738/48008 [1:24:49<1:16:44, 5.05it/s] 52%|█████▏ | 24739/48008 [1:24:49<1:15:28, 5.14it/s] 52%|█████▏ | 24740/48008 [1:24:49<1:28:23, 4.39it/s] 52%|█████▏ | 24741/48008 [1:24:50<1:25:11, 4.55it/s] 52%|█████▏ | 24742/48008 [1:24:50<1:22:34, 4.70it/s] 52%|█████▏ | 24743/48008 [1:24:50<1:19:41, 4.87it/s] 52%|█████▏ | 24744/48008 [1:24:50<1:19:13, 4.89it/s] 52%|█████▏ | 24745/48008 [1:24:50<1:16:50, 5.05it/s] 52%|█████▏ | 24746/48008 [1:24:51<1:16:01, 5.10it/s] 52%|█████▏ | 24747/48008 [1:24:51<1:20:44, 4.80it/s] 52%|█████▏ | 24748/48008 [1:24:51<1:18:51, 4.92it/s] 52%|█████▏ | 24749/48008 [1:24:51<1:22:58, 4.67it/s] 52%|█████▏ | 24750/48008 [1:24:52<1:20:44, 4.80it/s] {'loss': 0.2999, 'grad_norm': 0.392578125, 'learning_rate': 9.689635060823197e-05, 'epoch': 0.52} 52%|█████▏ | 24750/48008 [1:24:52<1:20:44, 4.80it/s] 52%|█████▏ | 24751/48008 [1:24:52<1:41:38, 3.81it/s] 52%|█████▏ | 24752/48008 [1:24:52<1:33:16, 4.16it/s] 52%|█████▏ | 24753/48008 [1:24:52<1:27:33, 4.43it/s] 52%|█████▏ | 24754/48008 [1:24:52<1:23:41, 4.63it/s] 52%|█████▏ | 24755/48008 [1:24:53<1:20:16, 4.83it/s] 52%|█████▏ | 24756/48008 [1:24:53<1:19:56, 4.85it/s] 52%|█████▏ | 24757/48008 [1:24:53<1:18:58, 4.91it/s] 52%|█████▏ | 24758/48008 [1:24:53<1:19:01, 4.90it/s] 52%|█████▏ | 24759/48008 [1:24:53<1:18:14, 4.95it/s] 52%|█████▏ | 24760/48008 [1:24:54<1:17:08, 5.02it/s] 52%|█████▏ | 24761/48008 [1:24:54<1:21:20, 4.76it/s] 52%|█████▏ | 24762/48008 [1:24:54<1:18:35, 4.93it/s] 52%|█████▏ | 24763/48008 [1:24:54<1:17:03, 5.03it/s] 52%|█████▏ | 24764/48008 [1:24:54<1:16:05, 5.09it/s] 52%|█████▏ | 24765/48008 [1:24:55<1:20:51, 4.79it/s] 52%|█████▏ | 24766/48008 [1:24:55<1:17:59, 4.97it/s] 52%|█████▏ | 24767/48008 [1:24:55<1:16:20, 5.07it/s] 52%|█████▏ | 24768/48008 [1:24:55<1:15:26, 5.13it/s] 52%|█████▏ | 24769/48008 [1:24:55<1:16:17, 5.08it/s] 52%|█████▏ | 24770/48008 [1:24:56<1:15:41, 5.12it/s] 52%|█████▏ | 24771/48008 [1:24:56<1:14:25, 5.20it/s] 52%|█████▏ | 24772/48008 [1:24:56<1:15:55, 5.10it/s] 52%|█████▏ | 24773/48008 [1:24:56<1:14:49, 5.17it/s] 52%|█████▏ | 24774/48008 [1:24:56<1:14:20, 5.21it/s] 52%|█████▏ | 24775/48008 [1:24:57<1:14:05, 5.23it/s] 52%|█████▏ | 24776/48008 [1:24:57<1:14:14, 5.22it/s] 52%|█████▏ | 24777/48008 [1:24:57<1:14:04, 5.23it/s] 52%|█████▏ | 24778/48008 [1:24:57<1:14:56, 5.17it/s] 52%|█████▏ | 24779/48008 [1:24:57<1:14:47, 5.18it/s] 52%|█████▏ | 24780/48008 [1:24:58<1:14:40, 5.18it/s] 52%|█████▏ | 24781/48008 [1:24:58<1:14:28, 5.20it/s] 52%|█████▏ | 24782/48008 [1:24:58<1:14:18, 5.21it/s] 52%|█████▏ | 24783/48008 [1:24:58<1:14:44, 5.18it/s] 52%|█████▏ | 24784/48008 [1:24:58<1:14:53, 5.17it/s] 52%|█████▏ | 24785/48008 [1:24:59<1:13:55, 5.24it/s] 52%|█████▏ | 24786/48008 [1:24:59<1:13:35, 5.26it/s] 52%|█████▏ | 24787/48008 [1:24:59<1:15:57, 5.09it/s] 52%|█████▏ | 24788/48008 [1:24:59<1:14:45, 5.18it/s] 52%|█████▏ | 24789/48008 [1:24:59<1:14:46, 5.18it/s] 52%|█████▏ | 24790/48008 [1:25:00<1:14:30, 5.19it/s] 52%|█████▏ | 24791/48008 [1:25:00<1:13:58, 5.23it/s] 52%|█████▏ | 24792/48008 [1:25:00<1:13:44, 5.25it/s] 52%|█████▏ | 24793/48008 [1:25:00<1:13:43, 5.25it/s] 52%|█████▏ | 24794/48008 [1:25:00<1:14:15, 5.21it/s] 52%|█████▏ | 24795/48008 [1:25:00<1:13:40, 5.25it/s] 52%|█████▏ | 24796/48008 [1:25:01<1:36:19, 4.02it/s] 52%|█████▏ | 24797/48008 [1:25:01<1:30:50, 4.26it/s] 52%|█████▏ | 24798/48008 [1:25:01<1:26:31, 4.47it/s] 52%|█████▏ | 24799/48008 [1:25:01<1:22:23, 4.69it/s] 52%|█████▏ | 24800/48008 [1:25:02<1:19:35, 4.86it/s] {'loss': 0.3017, 'grad_norm': 0.609375, 'learning_rate': 9.668805199133479e-05, 'epoch': 0.52} 52%|█████▏ | 24800/48008 [1:25:02<1:19:35, 4.86it/s] 52%|█████▏ | 24801/48008 [1:25:02<1:18:48, 4.91it/s] 52%|█████▏ | 24802/48008 [1:25:02<1:18:45, 4.91it/s] 52%|█████▏ | 24803/48008 [1:25:02<1:17:15, 5.01it/s] 52%|█████▏ | 24804/48008 [1:25:02<1:16:06, 5.08it/s] 52%|█████▏ | 24805/48008 [1:25:03<1:15:25, 5.13it/s] 52%|█████▏ | 24806/48008 [1:25:03<1:15:02, 5.15it/s] 52%|█████▏ | 24807/48008 [1:25:03<1:14:14, 5.21it/s] 52%|█████▏ | 24808/48008 [1:25:03<1:19:32, 4.86it/s] 52%|█████▏ | 24809/48008 [1:25:03<1:17:41, 4.98it/s] 52%|█████▏ | 24810/48008 [1:25:04<1:17:45, 4.97it/s] 52%|█████▏ | 24811/48008 [1:25:04<1:17:41, 4.98it/s] 52%|█████▏ | 24812/48008 [1:25:04<1:15:58, 5.09it/s] 52%|█████▏ | 24813/48008 [1:25:04<1:15:23, 5.13it/s] 52%|█████▏ | 24814/48008 [1:25:04<1:14:54, 5.16it/s] 52%|█████▏ | 24815/48008 [1:25:05<1:15:58, 5.09it/s] 52%|█████▏ | 24816/48008 [1:25:05<1:14:50, 5.16it/s] 52%|█████▏ | 24817/48008 [1:25:05<1:14:47, 5.17it/s] 52%|█████▏ | 24818/48008 [1:25:05<1:14:29, 5.19it/s] 52%|█████▏ | 24819/48008 [1:25:05<1:14:43, 5.17it/s] 52%|█████▏ | 24820/48008 [1:25:06<1:15:48, 5.10it/s] 52%|█████▏ | 24821/48008 [1:25:06<1:16:25, 5.06it/s] 52%|█████▏ | 24822/48008 [1:25:06<1:17:11, 5.01it/s] 52%|█████▏ | 24823/48008 [1:25:06<1:15:42, 5.10it/s] 52%|█████▏ | 24824/48008 [1:25:06<1:15:24, 5.12it/s] 52%|█████▏ | 24825/48008 [1:25:07<1:15:24, 5.12it/s] 52%|█████▏ | 24826/48008 [1:25:07<1:15:34, 5.11it/s] 52%|█████▏ | 24827/48008 [1:25:07<1:28:32, 4.36it/s] 52%|█████▏ | 24828/48008 [1:25:07<1:25:02, 4.54it/s] 52%|█████▏ | 24829/48008 [1:25:07<1:22:54, 4.66it/s] 52%|█████▏ | 24830/48008 [1:25:08<1:21:34, 4.74it/s] 52%|█████▏ | 24831/48008 [1:25:08<1:19:13, 4.88it/s] 52%|█████▏ | 24832/48008 [1:25:08<1:17:29, 4.99it/s] 52%|█████▏ | 24833/48008 [1:25:08<1:16:15, 5.07it/s] 52%|█████▏ | 24834/48008 [1:25:08<1:16:52, 5.02it/s] 52%|█████▏ | 24835/48008 [1:25:09<1:16:05, 5.08it/s] 52%|█████▏ | 24836/48008 [1:25:09<1:16:37, 5.04it/s] 52%|█████▏ | 24837/48008 [1:25:09<1:16:15, 5.06it/s] 52%|█████▏ | 24838/48008 [1:25:09<1:28:57, 4.34it/s] 52%|█████▏ | 24839/48008 [1:25:09<1:24:25, 4.57it/s] 52%|█████▏ | 24840/48008 [1:25:10<1:21:07, 4.76it/s] 52%|█████▏ | 24841/48008 [1:25:10<1:19:01, 4.89it/s] 52%|█████▏ | 24842/48008 [1:25:10<1:17:21, 4.99it/s] 52%|█████▏ | 24843/48008 [1:25:10<1:16:22, 5.06it/s] 52%|█████▏ | 24844/48008 [1:25:10<1:15:55, 5.08it/s] 52%|█████▏ | 24845/48008 [1:25:11<1:15:01, 5.15it/s] 52%|█████▏ | 24846/48008 [1:25:11<1:14:53, 5.15it/s] 52%|█████▏ | 24847/48008 [1:25:11<1:15:44, 5.10it/s] 52%|█████▏ | 24848/48008 [1:25:11<1:15:35, 5.11it/s] 52%|█████▏ | 24849/48008 [1:25:11<1:14:32, 5.18it/s] 52%|█████▏ | 24850/48008 [1:25:12<1:15:54, 5.08it/s] {'loss': 0.2979, 'grad_norm': 0.3671875, 'learning_rate': 9.64797533744376e-05, 'epoch': 0.52} 52%|█████▏ | 24850/48008 [1:25:12<1:15:54, 5.08it/s] 52%|█████▏ | 24851/48008 [1:25:12<1:16:09, 5.07it/s] 52%|█████▏ | 24852/48008 [1:25:12<1:16:31, 5.04it/s] 52%|█████▏ | 24853/48008 [1:25:12<1:17:24, 4.99it/s] 52%|█████▏ | 24854/48008 [1:25:12<1:16:00, 5.08it/s] 52%|█████▏ | 24855/48008 [1:25:13<1:16:59, 5.01it/s] 52%|█████▏ | 24856/48008 [1:25:13<1:15:36, 5.10it/s] 52%|█████▏ | 24857/48008 [1:25:13<1:15:49, 5.09it/s] 52%|█████▏ | 24858/48008 [1:25:13<1:37:43, 3.95it/s] 52%|█████▏ | 24859/48008 [1:25:14<1:53:06, 3.41it/s] 52%|█████▏ | 24860/48008 [1:25:14<1:40:58, 3.82it/s] 52%|█████▏ | 24861/48008 [1:25:14<1:32:43, 4.16it/s] 52%|█████▏ | 24862/48008 [1:25:14<1:26:31, 4.46it/s] 52%|█████▏ | 24863/48008 [1:25:15<1:22:18, 4.69it/s] 52%|█████▏ | 24864/48008 [1:25:15<1:20:59, 4.76it/s] 52%|█████▏ | 24865/48008 [1:25:15<1:19:45, 4.84it/s] 52%|█████▏ | 24866/48008 [1:25:15<1:17:53, 4.95it/s] 52%|█████▏ | 24867/48008 [1:25:15<1:18:32, 4.91it/s] 52%|█████▏ | 24868/48008 [1:25:16<1:16:28, 5.04it/s] 52%|█████▏ | 24869/48008 [1:25:16<1:16:30, 5.04it/s] 52%|█████▏ | 24870/48008 [1:25:16<1:21:00, 4.76it/s] 52%|█████▏ | 24871/48008 [1:25:16<1:18:32, 4.91it/s] 52%|█████▏ | 24872/48008 [1:25:16<1:16:48, 5.02it/s] 52%|█████▏ | 24873/48008 [1:25:17<1:17:02, 5.00it/s] 52%|█████▏ | 24874/48008 [1:25:17<1:15:24, 5.11it/s] 52%|█████▏ | 24875/48008 [1:25:17<1:14:27, 5.18it/s] 52%|█████▏ | 24876/48008 [1:25:17<1:15:35, 5.10it/s] 52%|█████▏ | 24877/48008 [1:25:17<1:15:28, 5.11it/s] 52%|█████▏ | 24878/48008 [1:25:17<1:14:24, 5.18it/s] 52%|█████▏ | 24879/48008 [1:25:18<1:14:50, 5.15it/s] 52%|█████▏ | 24880/48008 [1:25:18<1:14:15, 5.19it/s] 52%|█████▏ | 24881/48008 [1:25:18<1:13:56, 5.21it/s] 52%|█████▏ | 24882/48008 [1:25:18<1:13:57, 5.21it/s] 52%|█████▏ | 24883/48008 [1:25:18<1:14:39, 5.16it/s] 52%|█████▏ | 24884/48008 [1:25:19<1:15:15, 5.12it/s] 52%|█████▏ | 24885/48008 [1:25:19<1:15:22, 5.11it/s] 52%|█████▏ | 24886/48008 [1:25:19<1:14:27, 5.18it/s] 52%|█████▏ | 24887/48008 [1:25:19<1:16:30, 5.04it/s] 52%|█████▏ | 24888/48008 [1:25:19<1:16:25, 5.04it/s] 52%|█████▏ | 24889/48008 [1:25:20<1:16:17, 5.05it/s] 52%|█████▏ | 24890/48008 [1:25:20<1:15:19, 5.12it/s] 52%|█████▏ | 24891/48008 [1:25:20<1:14:54, 5.14it/s] 52%|█████▏ | 24892/48008 [1:25:20<1:15:16, 5.12it/s] 52%|█████▏ | 24893/48008 [1:25:20<1:15:58, 5.07it/s] 52%|█████▏ | 24894/48008 [1:25:21<1:15:16, 5.12it/s] 52%|█████▏ | 24895/48008 [1:25:21<1:14:06, 5.20it/s] 52%|█████▏ | 24896/48008 [1:25:21<1:19:21, 4.85it/s] 52%|█████▏ | 24897/48008 [1:25:21<1:18:15, 4.92it/s] 52%|█████▏ | 24898/48008 [1:25:21<1:17:25, 4.98it/s] 52%|█████▏ | 24899/48008 [1:25:22<1:38:40, 3.90it/s] 52%|█████▏ | 24900/48008 [1:25:22<1:30:43, 4.24it/s] {'loss': 0.2996, 'grad_norm': 0.51953125, 'learning_rate': 9.627145475754041e-05, 'epoch': 0.52} 52%|█████▏ | 24900/48008 [1:25:22<1:30:43, 4.24it/s] 52%|█████▏ | 24901/48008 [1:25:22<1:26:16, 4.46it/s] 52%|█████▏ | 24902/48008 [1:25:22<1:22:11, 4.69it/s] 52%|█████▏ | 24903/48008 [1:25:23<1:20:55, 4.76it/s] 52%|█████▏ | 24904/48008 [1:25:23<1:32:05, 4.18it/s] 52%|█████▏ | 24905/48008 [1:25:23<1:26:00, 4.48it/s] 52%|█████▏ | 24906/48008 [1:25:23<1:27:33, 4.40it/s] 52%|█████▏ | 24907/48008 [1:25:24<1:23:17, 4.62it/s] 52%|█████▏ | 24908/48008 [1:25:24<1:22:43, 4.65it/s] 52%|█████▏ | 24909/48008 [1:25:24<1:21:19, 4.73it/s] 52%|█████▏ | 24910/48008 [1:25:24<1:18:31, 4.90it/s] 52%|█████▏ | 24911/48008 [1:25:24<1:17:12, 4.99it/s] 52%|█████▏ | 24912/48008 [1:25:24<1:15:53, 5.07it/s] 52%|█████▏ | 24913/48008 [1:25:25<1:16:42, 5.02it/s] 52%|█████▏ | 24914/48008 [1:25:25<1:29:17, 4.31it/s] 52%|█████▏ | 24915/48008 [1:25:25<1:24:18, 4.57it/s] 52%|█████▏ | 24916/48008 [1:25:25<1:22:03, 4.69it/s] 52%|█████▏ | 24917/48008 [1:25:26<1:19:14, 4.86it/s] 52%|█████▏ | 24918/48008 [1:25:26<1:30:47, 4.24it/s] 52%|█████▏ | 24919/48008 [1:25:26<1:25:18, 4.51it/s] 52%|█████▏ | 24920/48008 [1:25:26<1:21:18, 4.73it/s] 52%|█████▏ | 24921/48008 [1:25:26<1:19:17, 4.85it/s] 52%|█████▏ | 24922/48008 [1:25:27<1:22:56, 4.64it/s] 52%|█████▏ | 24923/48008 [1:25:27<1:20:12, 4.80it/s] 52%|█████▏ | 24924/48008 [1:25:27<1:18:37, 4.89it/s] 52%|█████▏ | 24925/48008 [1:25:27<1:16:53, 5.00it/s] 52%|█████▏ | 24926/48008 [1:25:28<1:21:13, 4.74it/s] 52%|█████▏ | 24927/48008 [1:25:28<1:20:22, 4.79it/s] 52%|█████▏ | 24928/48008 [1:25:28<1:18:44, 4.89it/s] 52%|█████▏ | 24929/48008 [1:25:28<1:17:26, 4.97it/s] 52%|█████▏ | 24930/48008 [1:25:28<1:29:33, 4.29it/s] 52%|█████▏ | 24931/48008 [1:25:29<1:25:31, 4.50it/s] 52%|█████▏ | 24932/48008 [1:25:29<1:21:44, 4.70it/s] 52%|█████▏ | 24933/48008 [1:25:29<1:24:39, 4.54it/s] 52%|█████▏ | 24934/48008 [1:25:29<1:21:06, 4.74it/s] 52%|█████▏ | 24935/48008 [1:25:29<1:20:12, 4.79it/s] 52%|█████▏ | 24936/48008 [1:25:30<1:31:30, 4.20it/s] 52%|█████▏ | 24937/48008 [1:25:30<1:25:55, 4.48it/s] 52%|█████▏ | 24938/48008 [1:25:30<1:21:39, 4.71it/s] 52%|█████▏ | 24939/48008 [1:25:30<1:18:07, 4.92it/s] 52%|█████▏ | 24940/48008 [1:25:30<1:18:01, 4.93it/s] 52%|█████▏ | 24941/48008 [1:25:31<1:17:06, 4.99it/s] 52%|█████▏ | 24942/48008 [1:25:31<1:15:49, 5.07it/s] 52%|█████▏ | 24943/48008 [1:25:31<1:15:56, 5.06it/s] 52%|█████▏ | 24944/48008 [1:25:31<1:15:27, 5.09it/s] 52%|█████▏ | 24945/48008 [1:25:31<1:14:25, 5.16it/s] 52%|█████▏ | 24946/48008 [1:25:32<1:13:59, 5.19it/s] 52%|█████▏ | 24947/48008 [1:25:32<1:13:45, 5.21it/s] 52%|█████▏ | 24948/48008 [1:25:32<1:12:51, 5.28it/s] 52%|█████▏ | 24949/48008 [1:25:32<1:13:11, 5.25it/s] 52%|█████▏ | 24950/48008 [1:25:32<1:13:18, 5.24it/s] {'loss': 0.2987, 'grad_norm': 0.142578125, 'learning_rate': 9.606315614064323e-05, 'epoch': 0.52} 52%|█████▏ | 24950/48008 [1:25:32<1:13:18, 5.24it/s] 52%|█████▏ | 24951/48008 [1:25:33<1:13:18, 5.24it/s] 52%|█████▏ | 24952/48008 [1:25:33<1:12:57, 5.27it/s] 52%|█████▏ | 24953/48008 [1:25:33<1:14:37, 5.15it/s] 52%|█████▏ | 24954/48008 [1:25:33<1:13:38, 5.22it/s] 52%|█████▏ | 24955/48008 [1:25:33<1:13:48, 5.21it/s] 52%|█████▏ | 24956/48008 [1:25:34<1:15:53, 5.06it/s] 52%|█████▏ | 24957/48008 [1:25:34<1:14:44, 5.14it/s] 52%|█████▏ | 24958/48008 [1:25:34<1:16:23, 5.03it/s] 52%|█████▏ | 24959/48008 [1:25:34<1:16:10, 5.04it/s] 52%|█████▏ | 24960/48008 [1:25:34<1:14:49, 5.13it/s] 52%|█████▏ | 24961/48008 [1:25:35<1:14:55, 5.13it/s] 52%|█████▏ | 24962/48008 [1:25:35<1:15:10, 5.11it/s] 52%|█████▏ | 24963/48008 [1:25:35<1:16:06, 5.05it/s] 52%|█████▏ | 24964/48008 [1:25:35<1:14:44, 5.14it/s] 52%|█████▏ | 24965/48008 [1:25:35<1:16:42, 5.01it/s] 52%|█████▏ | 24966/48008 [1:25:36<1:29:13, 4.30it/s] 52%|█████▏ | 24967/48008 [1:25:36<1:24:46, 4.53it/s] 52%|█████▏ | 24968/48008 [1:25:36<1:20:55, 4.75it/s] 52%|█████▏ | 24969/48008 [1:25:36<1:24:15, 4.56it/s] 52%|█████▏ | 24970/48008 [1:25:36<1:22:28, 4.66it/s] 52%|█████▏ | 24971/48008 [1:25:37<1:19:34, 4.82it/s] 52%|█████▏ | 24972/48008 [1:25:37<1:17:56, 4.93it/s] 52%|█████▏ | 24973/48008 [1:25:37<1:16:32, 5.02it/s] 52%|█████▏ | 24974/48008 [1:25:37<1:16:26, 5.02it/s] 52%|█████▏ | 24975/48008 [1:25:37<1:16:20, 5.03it/s] 52%|█████▏ | 24976/48008 [1:25:38<1:16:03, 5.05it/s] 52%|█████▏ | 24977/48008 [1:25:38<1:15:11, 5.11it/s] 52%|█████▏ | 24978/48008 [1:25:38<1:14:40, 5.14it/s] 52%|█████▏ | 24979/48008 [1:25:38<1:19:41, 4.82it/s] 52%|█████▏ | 24980/48008 [1:25:38<1:17:58, 4.92it/s] 52%|█████▏ | 24981/48008 [1:25:39<1:16:19, 5.03it/s] 52%|█████▏ | 24982/48008 [1:25:39<1:16:31, 5.01it/s] 52%|█████▏ | 24983/48008 [1:25:39<1:20:48, 4.75it/s] 52%|█████▏ | 24984/48008 [1:25:39<1:18:34, 4.88it/s] 52%|█████▏ | 24985/48008 [1:25:39<1:16:22, 5.02it/s] 52%|█████▏ | 24986/48008 [1:25:40<1:20:46, 4.75it/s] 52%|█████▏ | 24987/48008 [1:25:40<1:18:18, 4.90it/s] 52%|█████▏ | 24988/48008 [1:25:40<1:16:31, 5.01it/s] 52%|█████▏ | 24989/48008 [1:25:40<1:20:49, 4.75it/s] 52%|█████▏ | 24990/48008 [1:25:41<1:19:07, 4.85it/s] 52%|█████▏ | 24991/48008 [1:25:41<1:17:01, 4.98it/s] 52%|█████▏ | 24992/48008 [1:25:41<1:15:46, 5.06it/s] 52%|█████▏ | 24993/48008 [1:25:41<1:15:29, 5.08it/s] 52%|█████▏ | 24994/48008 [1:25:41<1:15:50, 5.06it/s] 52%|█████▏ | 24995/48008 [1:25:41<1:14:54, 5.12it/s] 52%|█████▏ | 24996/48008 [1:25:42<1:15:38, 5.07it/s] 52%|█████▏ | 24997/48008 [1:25:42<1:14:31, 5.15it/s] 52%|█████▏ | 24998/48008 [1:25:42<1:13:46, 5.20it/s] 52%|█████▏ | 24999/48008 [1:25:42<1:14:53, 5.12it/s] 52%|█████▏ | 25000/48008 [1:25:42<1:15:54, 5.05it/s] {'loss': 0.3011, 'grad_norm': 0.1796875, 'learning_rate': 9.585485752374604e-05, 'epoch': 0.52} 52%|█████▏ | 25000/48008 [1:25:42<1:15:54, 5.05it/s] 52%|█████▏ | 25001/48008 [1:25:43<1:20:24, 4.77it/s] 52%|█████▏ | 25002/48008 [1:25:43<1:18:20, 4.89it/s] 52%|█████▏ | 25003/48008 [1:25:43<1:16:07, 5.04it/s] 52%|█████▏ | 25004/48008 [1:25:43<1:15:53, 5.05it/s] 52%|█████▏ | 25005/48008 [1:25:43<1:16:32, 5.01it/s] 52%|█████▏ | 25006/48008 [1:25:44<1:15:21, 5.09it/s] 52%|█████▏ | 25007/48008 [1:25:44<1:14:13, 5.17it/s] 52%|█████▏ | 25008/48008 [1:25:44<1:14:32, 5.14it/s] 52%|█████▏ | 25009/48008 [1:25:44<1:14:15, 5.16it/s] 52%|█████▏ | 25010/48008 [1:25:44<1:14:32, 5.14it/s] 52%|█████▏ | 25011/48008 [1:25:45<1:14:37, 5.14it/s] 52%|█████▏ | 25012/48008 [1:25:45<1:13:23, 5.22it/s] 52%|█████▏ | 25013/48008 [1:25:45<1:14:36, 5.14it/s] 52%|█████▏ | 25014/48008 [1:25:45<1:15:32, 5.07it/s] 52%|█████▏ | 25015/48008 [1:25:45<1:14:02, 5.18it/s] 52%|█████▏ | 25016/48008 [1:25:46<1:13:19, 5.23it/s] 52%|█████▏ | 25017/48008 [1:25:46<1:13:18, 5.23it/s] 52%|█████▏ | 25018/48008 [1:25:46<1:13:16, 5.23it/s] 52%|█████▏ | 25019/48008 [1:25:46<1:13:14, 5.23it/s] 52%|█████▏ | 25020/48008 [1:25:46<1:12:37, 5.28it/s] 52%|█████▏ | 25021/48008 [1:25:47<1:12:57, 5.25it/s] 52%|█████▏ | 25022/48008 [1:25:47<1:12:36, 5.28it/s] 52%|█████▏ | 25023/48008 [1:25:47<1:12:48, 5.26it/s] 52%|█████▏ | 25024/48008 [1:25:47<1:12:44, 5.27it/s] 52%|█████▏ | 25025/48008 [1:25:47<1:26:06, 4.45it/s] 52%|█████▏ | 25026/48008 [1:25:48<1:21:38, 4.69it/s] 52%|█████▏ | 25027/48008 [1:25:48<1:18:42, 4.87it/s] 52%|█████▏ | 25028/48008 [1:25:48<1:18:00, 4.91it/s] 52%|█████▏ | 25029/48008 [1:25:48<1:17:05, 4.97it/s] 52%|█████▏ | 25030/48008 [1:25:48<1:16:28, 5.01it/s] 52%|█████▏ | 25031/48008 [1:25:49<1:15:31, 5.07it/s] 52%|█████▏ | 25032/48008 [1:25:49<1:14:55, 5.11it/s] 52%|█████▏ | 25033/48008 [1:25:49<1:13:57, 5.18it/s] 52%|█████▏ | 25034/48008 [1:25:49<1:12:48, 5.26it/s] 52%|█████▏ | 25035/48008 [1:25:49<1:12:27, 5.28it/s] 52%|█████▏ | 25036/48008 [1:25:50<1:13:13, 5.23it/s] 52%|█████▏ | 25037/48008 [1:25:50<1:12:53, 5.25it/s] 52%|█████▏ | 25038/48008 [1:25:50<1:14:08, 5.16it/s] 52%|█████▏ | 25039/48008 [1:25:50<1:13:43, 5.19it/s] 52%|█████▏ | 25040/48008 [1:25:50<1:36:10, 3.98it/s] 52%|█████▏ | 25041/48008 [1:25:51<1:29:17, 4.29it/s] 52%|█████▏ | 25042/48008 [1:25:51<1:24:19, 4.54it/s] 52%|█████▏ | 25043/48008 [1:25:51<1:21:39, 4.69it/s] 52%|█████▏ | 25044/48008 [1:25:51<1:19:16, 4.83it/s] 52%|█████▏ | 25045/48008 [1:25:51<1:17:06, 4.96it/s] 52%|█████▏ | 25046/48008 [1:25:52<1:15:39, 5.06it/s] 52%|█████▏ | 25047/48008 [1:25:52<1:14:35, 5.13it/s] 52%|█████▏ | 25048/48008 [1:25:52<1:15:51, 5.04it/s] 52%|█████▏ | 25049/48008 [1:25:52<1:16:19, 5.01it/s] 52%|█████▏ | 25050/48008 [1:25:52<1:15:45, 5.05it/s] {'loss': 0.3022, 'grad_norm': 0.31640625, 'learning_rate': 9.564655890684886e-05, 'epoch': 0.52} 52%|█████▏ | 25050/48008 [1:25:52<1:15:45, 5.05it/s] 52%|█████▏ | 25051/48008 [1:25:53<1:16:14, 5.02it/s] 52%|█████▏ | 25052/48008 [1:25:53<1:17:13, 4.95it/s] 52%|█████▏ | 25053/48008 [1:25:53<1:15:24, 5.07it/s] 52%|█████▏ | 25054/48008 [1:25:53<1:17:06, 4.96it/s] 52%|█████▏ | 25055/48008 [1:25:53<1:15:44, 5.05it/s] 52%|█████▏ | 25056/48008 [1:25:54<1:15:13, 5.09it/s] 52%|█████▏ | 25057/48008 [1:25:54<1:14:32, 5.13it/s] 52%|█████▏ | 25058/48008 [1:25:54<1:15:37, 5.06it/s] 52%|█████▏ | 25059/48008 [1:25:54<1:14:26, 5.14it/s] 52%|█████▏ | 25060/48008 [1:25:54<1:14:11, 5.16it/s] 52%|█████▏ | 25061/48008 [1:25:55<1:14:33, 5.13it/s] 52%|█████▏ | 25062/48008 [1:25:55<1:14:07, 5.16it/s] 52%|█████▏ | 25063/48008 [1:25:55<1:13:50, 5.18it/s] 52%|█████▏ | 25064/48008 [1:25:55<1:15:05, 5.09it/s] 52%|█████▏ | 25065/48008 [1:25:55<1:13:57, 5.17it/s] 52%|█████▏ | 25066/48008 [1:25:56<1:15:11, 5.09it/s] 52%|█████▏ | 25067/48008 [1:25:56<1:14:43, 5.12it/s] 52%|█████▏ | 25068/48008 [1:25:56<1:14:30, 5.13it/s] 52%|█████▏ | 25069/48008 [1:25:56<1:14:10, 5.15it/s] 52%|█████▏ | 25070/48008 [1:25:56<1:13:39, 5.19it/s] 52%|█████▏ | 25071/48008 [1:25:57<1:15:33, 5.06it/s] 52%|█████▏ | 25072/48008 [1:25:57<1:14:53, 5.10it/s] 52%|█████▏ | 25073/48008 [1:25:57<1:14:07, 5.16it/s] 52%|█████▏ | 25074/48008 [1:25:57<1:13:25, 5.21it/s] 52%|█████▏ | 25075/48008 [1:25:57<1:13:09, 5.22it/s] 52%|█████▏ | 25076/48008 [1:25:57<1:12:41, 5.26it/s] 52%|█████▏ | 25077/48008 [1:25:58<1:12:46, 5.25it/s] 52%|█████▏ | 25078/48008 [1:25:58<1:15:10, 5.08it/s] 52%|█████▏ | 25079/48008 [1:25:58<1:15:39, 5.05it/s] 52%|█████▏ | 25080/48008 [1:25:58<1:14:42, 5.12it/s] 52%|█████▏ | 25081/48008 [1:25:59<1:36:28, 3.96it/s] 52%|█████▏ | 25082/48008 [1:25:59<1:29:25, 4.27it/s] 52%|█████▏ | 25083/48008 [1:25:59<1:26:05, 4.44it/s] 52%|█████▏ | 25084/48008 [1:25:59<1:22:23, 4.64it/s] 52%|█████▏ | 25085/48008 [1:25:59<1:19:45, 4.79it/s] 52%|█████▏ | 25086/48008 [1:26:00<1:18:14, 4.88it/s] 52%|█████▏ | 25087/48008 [1:26:00<1:16:59, 4.96it/s] 52%|█████▏ | 25088/48008 [1:26:00<1:16:33, 4.99it/s] 52%|█████▏ | 25089/48008 [1:26:00<1:15:22, 5.07it/s] 52%|█████▏ | 25090/48008 [1:26:01<1:37:04, 3.93it/s] 52%|█████▏ | 25091/48008 [1:26:01<1:29:36, 4.26it/s] 52%|█████▏ | 25092/48008 [1:26:01<1:25:20, 4.48it/s] 52%|█████▏ | 25093/48008 [1:26:01<1:21:07, 4.71it/s] 52%|█████▏ | 25094/48008 [1:26:01<1:19:53, 4.78it/s] 52%|█████▏ | 25095/48008 [1:26:02<1:19:05, 4.83it/s] 52%|█████▏ | 25096/48008 [1:26:02<1:17:24, 4.93it/s] 52%|█████▏ | 25097/48008 [1:26:02<1:21:23, 4.69it/s] 52%|█████▏ | 25098/48008 [1:26:02<1:18:20, 4.87it/s] 52%|█████▏ | 25099/48008 [1:26:02<1:17:18, 4.94it/s] 52%|█████▏ | 25100/48008 [1:26:03<1:17:08, 4.95it/s] {'loss': 0.299, 'grad_norm': 0.2109375, 'learning_rate': 9.543826028995168e-05, 'epoch': 0.52} 52%|█████▏ | 25100/48008 [1:26:03<1:17:08, 4.95it/s] 52%|█████▏ | 25101/48008 [1:26:03<1:15:29, 5.06it/s] 52%|█████▏ | 25102/48008 [1:26:03<1:14:49, 5.10it/s] 52%|█████▏ | 25103/48008 [1:26:03<1:14:39, 5.11it/s] 52%|█████▏ | 25104/48008 [1:26:04<1:36:25, 3.96it/s] 52%|█████▏ | 25105/48008 [1:26:04<1:29:12, 4.28it/s] 52%|█████▏ | 25106/48008 [1:26:04<1:24:13, 4.53it/s] 52%|█████▏ | 25107/48008 [1:26:04<1:20:23, 4.75it/s] 52%|█████▏ | 25108/48008 [1:26:04<1:20:15, 4.76it/s] 52%|█████▏ | 25109/48008 [1:26:06<3:17:53, 1.93it/s] 52%|█████▏ | 25110/48008 [1:26:06<2:40:38, 2.38it/s] 52%|█████▏ | 25111/48008 [1:26:06<2:13:44, 2.85it/s] 52%|█████▏ | 25112/48008 [1:26:06<1:55:38, 3.30it/s] 52%|█████▏ | 25113/48008 [1:26:06<1:43:14, 3.70it/s] 52%|█████▏ | 25114/48008 [1:26:07<1:34:11, 4.05it/s] 52%|█████▏ | 25115/48008 [1:26:07<1:28:12, 4.33it/s] 52%|█████▏ | 25116/48008 [1:26:07<1:23:49, 4.55it/s] 52%|█████▏ | 25117/48008 [1:26:07<1:34:05, 4.05it/s] 52%|█████▏ | 25118/48008 [1:26:07<1:27:48, 4.34it/s] 52%|█████▏ | 25119/48008 [1:26:08<1:23:24, 4.57it/s] 52%|█████▏ | 25120/48008 [1:26:08<1:21:19, 4.69it/s] 52%|█████▏ | 25121/48008 [1:26:08<1:19:58, 4.77it/s] 52%|█████▏ | 25122/48008 [1:26:08<1:19:03, 4.82it/s] 52%|█████▏ | 25123/48008 [1:26:08<1:17:56, 4.89it/s] 52%|█████▏ | 25124/48008 [1:26:09<1:16:00, 5.02it/s] 52%|█████▏ | 25125/48008 [1:26:09<1:14:25, 5.12it/s] 52%|█████▏ | 25126/48008 [1:26:09<1:16:14, 5.00it/s] 52%|█████▏ | 25127/48008 [1:26:09<1:15:06, 5.08it/s] 52%|█████▏ | 25128/48008 [1:26:09<1:14:34, 5.11it/s] 52%|█████▏ | 25129/48008 [1:26:10<1:14:09, 5.14it/s] 52%|█████▏ | 25130/48008 [1:26:10<1:13:48, 5.17it/s] 52%|█████▏ | 25131/48008 [1:26:10<1:13:44, 5.17it/s] 52%|█████▏ | 25132/48008 [1:26:10<1:13:30, 5.19it/s] 52%|█████▏ | 25133/48008 [1:26:10<1:14:37, 5.11it/s] 52%|█████▏ | 25134/48008 [1:26:11<1:16:40, 4.97it/s] 52%|█████▏ | 25135/48008 [1:26:11<1:15:43, 5.03it/s] 52%|█████▏ | 25136/48008 [1:26:11<1:14:39, 5.11it/s] 52%|█████▏ | 25137/48008 [1:26:11<1:15:40, 5.04it/s] 52%|█████▏ | 25138/48008 [1:26:11<1:15:07, 5.07it/s] 52%|█████▏ | 25139/48008 [1:26:12<1:19:47, 4.78it/s] 52%|█████▏ | 25140/48008 [1:26:12<1:18:24, 4.86it/s] 52%|█████▏ | 25141/48008 [1:26:12<1:16:46, 4.96it/s] 52%|█████▏ | 25142/48008 [1:26:12<1:15:31, 5.05it/s] 52%|█████▏ | 25143/48008 [1:26:12<1:14:53, 5.09it/s] 52%|█████▏ | 25144/48008 [1:26:13<1:14:53, 5.09it/s] 52%|█████▏ | 25145/48008 [1:26:13<1:14:09, 5.14it/s] 52%|█████▏ | 25146/48008 [1:26:13<1:12:57, 5.22it/s] 52%|█████▏ | 25147/48008 [1:26:13<1:18:10, 4.87it/s] 52%|█████▏ | 25148/48008 [1:26:13<1:17:04, 4.94it/s] 52%|█████▏ | 25149/48008 [1:26:14<1:16:12, 5.00it/s] 52%|█████▏ | 25150/48008 [1:26:14<1:14:55, 5.08it/s] {'loss': 0.3003, 'grad_norm': 0.2412109375, 'learning_rate': 9.52299616730545e-05, 'epoch': 0.52} 52%|█████▏ | 25150/48008 [1:26:14<1:14:55, 5.08it/s] 52%|█████▏ | 25151/48008 [1:26:14<1:13:32, 5.18it/s] 52%|█████▏ | 25152/48008 [1:26:14<1:13:31, 5.18it/s] 52%|█████▏ | 25153/48008 [1:26:14<1:12:31, 5.25it/s] 52%|█████▏ | 25154/48008 [1:26:15<1:12:21, 5.26it/s] 52%|█████▏ | 25155/48008 [1:26:15<1:13:58, 5.15it/s] 52%|█████▏ | 25156/48008 [1:26:15<1:13:09, 5.21it/s] 52%|█████▏ | 25157/48008 [1:26:15<1:13:03, 5.21it/s] 52%|█████▏ | 25158/48008 [1:26:15<1:14:19, 5.12it/s] 52%|█████▏ | 25159/48008 [1:26:15<1:12:40, 5.24it/s] 52%|█████▏ | 25160/48008 [1:26:16<1:12:47, 5.23it/s] 52%|█████▏ | 25161/48008 [1:26:16<1:13:55, 5.15it/s] 52%|█████▏ | 25162/48008 [1:26:16<1:13:16, 5.20it/s] 52%|█████▏ | 25163/48008 [1:26:16<1:13:37, 5.17it/s] 52%|█████▏ | 25164/48008 [1:26:16<1:13:02, 5.21it/s] 52%|█████▏ | 25165/48008 [1:26:17<1:12:58, 5.22it/s] 52%|█████▏ | 25166/48008 [1:26:17<1:12:46, 5.23it/s] 52%|█████▏ | 25167/48008 [1:26:17<1:34:50, 4.01it/s] 52%|█████▏ | 25168/48008 [1:26:17<1:27:46, 4.34it/s] 52%|█████▏ | 25169/48008 [1:26:18<1:23:20, 4.57it/s] 52%|█████▏ | 25170/48008 [1:26:18<1:20:07, 4.75it/s] 52%|█████▏ | 25171/48008 [1:26:18<1:18:05, 4.87it/s] 52%|█████▏ | 25172/48008 [1:26:18<1:16:08, 5.00it/s] 52%|█████▏ | 25173/48008 [1:26:18<1:14:38, 5.10it/s] 52%|█████▏ | 25174/48008 [1:26:19<1:13:40, 5.17it/s] 52%|█████▏ | 25175/48008 [1:26:19<1:13:25, 5.18it/s] 52%|█████▏ | 25176/48008 [1:26:19<1:14:07, 5.13it/s] 52%|█████▏ | 25177/48008 [1:26:19<1:18:47, 4.83it/s] 52%|█████▏ | 25178/48008 [1:26:19<1:17:04, 4.94it/s] 52%|█████▏ | 25179/48008 [1:26:20<1:15:37, 5.03it/s] 52%|█████▏ | 25180/48008 [1:26:20<1:15:17, 5.05it/s] 52%|█████▏ | 25181/48008 [1:26:20<1:14:20, 5.12it/s] 52%|█████▏ | 25182/48008 [1:26:20<1:14:27, 5.11it/s] 52%|█████▏ | 25183/48008 [1:26:20<1:14:23, 5.11it/s] 52%|█████▏ | 25184/48008 [1:26:21<1:13:50, 5.15it/s] 52%|█████▏ | 25185/48008 [1:26:21<1:13:30, 5.17it/s] 52%|█████▏ | 25186/48008 [1:26:21<1:13:31, 5.17it/s] 52%|█████▏ | 25187/48008 [1:26:21<1:13:25, 5.18it/s] 52%|█████▏ | 25188/48008 [1:26:21<1:13:30, 5.17it/s] 52%|█████▏ | 25189/48008 [1:26:21<1:13:05, 5.20it/s] 52%|█████▏ | 25190/48008 [1:26:22<1:12:41, 5.23it/s] 52%|█████▏ | 25191/48008 [1:26:22<1:12:19, 5.26it/s] 52%|█████▏ | 25192/48008 [1:26:22<1:13:37, 5.16it/s] 52%|█████▏ | 25193/48008 [1:26:22<1:13:07, 5.20it/s] 52%|█████▏ | 25194/48008 [1:26:22<1:12:45, 5.23it/s] 52%|█████▏ | 25195/48008 [1:26:23<1:13:04, 5.20it/s] 52%|█████▏ | 25196/48008 [1:26:23<1:13:20, 5.18it/s] 52%|█████▏ | 25197/48008 [1:26:23<1:14:21, 5.11it/s] 52%|█████▏ | 25198/48008 [1:26:23<1:27:02, 4.37it/s] 52%|█████▏ | 25199/48008 [1:26:24<1:23:56, 4.53it/s] 52%|█████▏ | 25200/48008 [1:26:24<1:20:23, 4.73it/s] 52%|█████▏ | 25200/48008 [1:26:24<1:20:23, 4.73it/s]{'loss': 0.3024, 'grad_norm': 0.3046875, 'learning_rate': 9.502166305615731e-05, 'epoch': 0.52} 52%|█████▏ | 25201/48008 [1:26:24<1:17:23, 4.91it/s] 52%|█████▏ | 25202/48008 [1:26:24<1:16:28, 4.97it/s] 52%|█████▏ | 25203/48008 [1:26:24<1:16:06, 4.99it/s] 52%|█████▏ | 25204/48008 [1:26:24<1:13:58, 5.14it/s] 53%|█████▎ | 25205/48008 [1:26:25<1:13:20, 5.18it/s] 53%|█████▎ | 25206/48008 [1:26:25<1:13:04, 5.20it/s] 53%|█████▎ | 25207/48008 [1:26:25<1:14:20, 5.11it/s] 53%|█████▎ | 25208/48008 [1:26:25<1:13:27, 5.17it/s] 53%|█████▎ | 25209/48008 [1:26:25<1:12:29, 5.24it/s] 53%|█████▎ | 25210/48008 [1:26:26<1:12:17, 5.26it/s] 53%|█████▎ | 25211/48008 [1:26:26<1:12:22, 5.25it/s] 53%|█████▎ | 25212/48008 [1:26:26<1:12:20, 5.25it/s] 53%|█████▎ | 25213/48008 [1:26:26<1:17:48, 4.88it/s] 53%|█████▎ | 25214/48008 [1:26:26<1:16:21, 4.98it/s] 53%|█████▎ | 25215/48008 [1:26:27<1:15:29, 5.03it/s] 53%|█████▎ | 25216/48008 [1:26:27<1:20:05, 4.74it/s] 53%|█████▎ | 25217/48008 [1:26:27<1:17:47, 4.88it/s] 53%|█████▎ | 25218/48008 [1:26:27<1:16:08, 4.99it/s] 53%|█████▎ | 25219/48008 [1:26:27<1:14:56, 5.07it/s] 53%|█████▎ | 25220/48008 [1:26:28<1:14:06, 5.12it/s] 53%|█████▎ | 25221/48008 [1:26:28<1:13:44, 5.15it/s] 53%|█████▎ | 25222/48008 [1:26:28<1:13:14, 5.19it/s] 53%|█████▎ | 25223/48008 [1:26:28<1:13:36, 5.16it/s] 53%|█████▎ | 25224/48008 [1:26:28<1:14:42, 5.08it/s] 53%|█████▎ | 25225/48008 [1:26:29<1:14:03, 5.13it/s] 53%|█████▎ | 25226/48008 [1:26:29<1:13:19, 5.18it/s] 53%|█████▎ | 25227/48008 [1:26:29<1:14:20, 5.11it/s] 53%|█████▎ | 25228/48008 [1:26:29<1:15:19, 5.04it/s] 53%|█████▎ | 25229/48008 [1:26:29<1:14:19, 5.11it/s] 53%|█████▎ | 25230/48008 [1:26:30<1:13:49, 5.14it/s] 53%|█████▎ | 25231/48008 [1:26:30<1:13:33, 5.16it/s] 53%|█████▎ | 25232/48008 [1:26:30<1:13:22, 5.17it/s] 53%|█████▎ | 25233/48008 [1:26:30<1:13:50, 5.14it/s] 53%|█████▎ | 25234/48008 [1:26:30<1:13:19, 5.18it/s] 53%|█████▎ | 25235/48008 [1:26:31<1:14:21, 5.10it/s] 53%|█████▎ | 25236/48008 [1:26:31<1:14:04, 5.12it/s] 53%|█████▎ | 25237/48008 [1:26:31<1:14:26, 5.10it/s] 53%|█████▎ | 25238/48008 [1:26:31<1:14:55, 5.07it/s] 53%|█████▎ | 25239/48008 [1:26:31<1:14:11, 5.11it/s] 53%|█████▎ | 25240/48008 [1:26:32<1:26:48, 4.37it/s] 53%|█████▎ | 25241/48008 [1:26:32<1:22:34, 4.59it/s] 53%|█████▎ | 25242/48008 [1:26:32<1:18:52, 4.81it/s] 53%|█████▎ | 25243/48008 [1:26:32<1:22:09, 4.62it/s] 53%|█████▎ | 25244/48008 [1:26:32<1:24:22, 4.50it/s] 53%|█████▎ | 25245/48008 [1:26:33<1:20:31, 4.71it/s] 53%|█████▎ | 25246/48008 [1:26:33<1:18:51, 4.81it/s] 53%|█████▎ | 25247/48008 [1:26:33<1:17:02, 4.92it/s] 53%|█████▎ | 25248/48008 [1:26:33<1:16:03, 4.99it/s] 53%|█████▎ | 25249/48008 [1:26:33<1:15:53, 5.00it/s] 53%|█████▎ | 25250/48008 [1:26:34<1:16:07, 4.98it/s] {'loss': 0.2989, 'grad_norm': 0.423828125, 'learning_rate': 9.481336443926013e-05, 'epoch': 0.53} 53%|█████▎ | 25250/48008 [1:26:34<1:16:07, 4.98it/s] 53%|█████▎ | 25251/48008 [1:26:34<1:14:50, 5.07it/s] 53%|█████▎ | 25252/48008 [1:26:34<1:15:35, 5.02it/s] 53%|█████▎ | 25253/48008 [1:26:34<1:14:08, 5.11it/s] 53%|█████▎ | 25254/48008 [1:26:34<1:13:14, 5.18it/s] 53%|█████▎ | 25255/48008 [1:26:35<1:12:40, 5.22it/s] 53%|█████▎ | 25256/48008 [1:26:35<1:12:33, 5.23it/s] 53%|█████▎ | 25257/48008 [1:26:35<1:12:56, 5.20it/s] 53%|█████▎ | 25258/48008 [1:26:35<1:35:16, 3.98it/s] 53%|█████▎ | 25259/48008 [1:26:36<1:28:51, 4.27it/s] 53%|█████▎ | 25260/48008 [1:26:36<1:23:58, 4.51it/s] 53%|█████▎ | 25261/48008 [1:26:36<1:42:34, 3.70it/s] 53%|█████▎ | 25262/48008 [1:26:36<1:33:14, 4.07it/s] 53%|█████▎ | 25263/48008 [1:26:37<1:27:46, 4.32it/s] 53%|█████▎ | 25264/48008 [1:26:37<1:23:03, 4.56it/s] 53%|█████▎ | 25265/48008 [1:26:37<1:19:36, 4.76it/s] 53%|█████▎ | 25266/48008 [1:26:37<1:16:56, 4.93it/s] 53%|█████▎ | 25267/48008 [1:26:37<1:15:26, 5.02it/s] 53%|█████▎ | 25268/48008 [1:26:38<1:16:26, 4.96it/s] 53%|█████▎ | 25269/48008 [1:26:38<1:15:56, 4.99it/s] 53%|█████▎ | 25270/48008 [1:26:38<1:16:01, 4.99it/s] 53%|█████▎ | 25271/48008 [1:26:39<2:31:50, 2.50it/s] 53%|█████▎ | 25272/48008 [1:26:39<2:08:03, 2.96it/s] 53%|█████▎ | 25273/48008 [1:26:39<1:52:05, 3.38it/s] 53%|█████▎ | 25274/48008 [1:26:39<1:40:05, 3.79it/s] 53%|█████▎ | 25275/48008 [1:26:40<1:32:25, 4.10it/s] 53%|█████▎ | 25276/48008 [1:26:40<1:26:22, 4.39it/s] 53%|█████▎ | 25277/48008 [1:26:40<1:21:56, 4.62it/s] 53%|█████▎ | 25278/48008 [1:26:41<2:48:31, 2.25it/s] 53%|█████▎ | 25279/48008 [1:26:41<2:19:16, 2.72it/s] 53%|█████▎ | 25280/48008 [1:26:41<1:59:20, 3.17it/s] 53%|█████▎ | 25281/48008 [1:26:41<1:44:37, 3.62it/s] 53%|█████▎ | 25282/48008 [1:26:42<1:35:02, 3.99it/s] 53%|█████▎ | 25283/48008 [1:26:42<1:29:30, 4.23it/s] 53%|█████▎ | 25284/48008 [1:26:43<2:43:05, 2.32it/s] 53%|█████▎ | 25285/48008 [1:26:43<2:16:31, 2.77it/s] 53%|█████▎ | 25286/48008 [1:26:43<1:57:17, 3.23it/s] 53%|█████▎ | 25287/48008 [1:26:43<1:45:10, 3.60it/s] 53%|█████▎ | 25288/48008 [1:26:44<1:35:24, 3.97it/s] 53%|█████▎ | 25289/48008 [1:26:44<1:28:20, 4.29it/s] 53%|█████▎ | 25290/48008 [1:26:45<2:33:53, 2.46it/s] 53%|█████▎ | 25291/48008 [1:26:45<2:09:40, 2.92it/s] 53%|█████▎ | 25292/48008 [1:26:45<1:52:15, 3.37it/s] 53%|█████▎ | 25293/48008 [1:26:45<1:40:22, 3.77it/s] 53%|█████▎ | 25294/48008 [1:26:45<1:32:00, 4.11it/s] 53%|█████▎ | 25295/48008 [1:26:45<1:26:09, 4.39it/s] 53%|█████▎ | 25296/48008 [1:26:46<1:21:10, 4.66it/s] 53%|█████▎ | 25297/48008 [1:26:46<1:19:36, 4.76it/s] 53%|█████▎ | 25298/48008 [1:26:46<1:17:36, 4.88it/s] 53%|█████▎ | 25299/48008 [1:26:46<1:16:01, 4.98it/s] 53%|█████▎ | 25300/48008 [1:26:46<1:14:44, 5.06it/s] {'loss': 0.3018, 'grad_norm': 0.1845703125, 'learning_rate': 9.460506582236295e-05, 'epoch': 0.53} 53%|█████▎ | 25300/48008 [1:26:46<1:14:44, 5.06it/s] 53%|█████▎ | 25301/48008 [1:26:47<1:14:09, 5.10it/s] 53%|█████▎ | 25302/48008 [1:26:47<1:13:21, 5.16it/s] 53%|█████▎ | 25303/48008 [1:26:47<1:18:23, 4.83it/s] 53%|█████▎ | 25304/48008 [1:26:47<1:16:14, 4.96it/s] 53%|█████▎ | 25305/48008 [1:26:47<1:15:06, 5.04it/s] 53%|█████▎ | 25306/48008 [1:26:48<1:14:10, 5.10it/s] 53%|█████▎ | 25307/48008 [1:26:48<1:13:51, 5.12it/s] 53%|█████▎ | 25308/48008 [1:26:48<1:13:52, 5.12it/s] 53%|█████▎ | 25309/48008 [1:26:48<1:18:40, 4.81it/s] 53%|█████▎ | 25310/48008 [1:26:49<1:38:51, 3.83it/s] 53%|█████▎ | 25311/48008 [1:26:49<1:29:43, 4.22it/s] 53%|█████▎ | 25312/48008 [1:26:49<1:24:36, 4.47it/s] 53%|█████▎ | 25313/48008 [1:26:49<1:22:00, 4.61it/s] 53%|█████▎ | 25314/48008 [1:26:49<1:19:00, 4.79it/s] 53%|█████▎ | 25315/48008 [1:26:50<1:16:48, 4.92it/s] 53%|█████▎ | 25316/48008 [1:26:50<1:15:13, 5.03it/s] 53%|█████▎ | 25317/48008 [1:26:50<1:14:15, 5.09it/s] 53%|█████▎ | 25318/48008 [1:26:50<1:12:20, 5.23it/s] 53%|█████▎ | 25319/48008 [1:26:50<1:12:26, 5.22it/s] 53%|█████▎ | 25320/48008 [1:26:51<2:23:44, 2.63it/s] 53%|█████▎ | 25321/48008 [1:26:51<2:02:55, 3.08it/s] 53%|█████▎ | 25322/48008 [1:26:52<1:47:44, 3.51it/s] 53%|█████▎ | 25323/48008 [1:26:52<1:37:52, 3.86it/s] 53%|█████▎ | 25324/48008 [1:26:52<1:30:08, 4.19it/s] 53%|█████▎ | 25325/48008 [1:26:52<1:25:20, 4.43it/s] 53%|█████▎ | 25326/48008 [1:26:52<1:20:55, 4.67it/s] 53%|█████▎ | 25327/48008 [1:26:53<1:17:53, 4.85it/s] 53%|█████▎ | 25328/48008 [1:26:53<1:17:31, 4.88it/s] 53%|█████▎ | 25329/48008 [1:26:53<1:16:01, 4.97it/s] 53%|█████▎ | 25330/48008 [1:26:53<1:14:01, 5.11it/s] 53%|█████▎ | 25331/48008 [1:26:53<1:13:30, 5.14it/s] 53%|█████▎ | 25332/48008 [1:26:53<1:12:38, 5.20it/s] 53%|█████▎ | 25333/48008 [1:26:54<1:34:37, 3.99it/s] 53%|█████▎ | 25334/48008 [1:26:54<1:29:25, 4.23it/s] 53%|█████▎ | 25335/48008 [1:26:54<1:25:40, 4.41it/s] 53%|█████▎ | 25336/48008 [1:26:54<1:21:10, 4.66it/s] 53%|█████▎ | 25337/48008 [1:26:55<1:23:53, 4.50it/s] 53%|█████▎ | 25338/48008 [1:26:55<1:21:50, 4.62it/s] 53%|█████▎ | 25339/48008 [1:26:55<1:18:53, 4.79it/s] 53%|█████▎ | 25340/48008 [1:26:55<1:16:40, 4.93it/s] 53%|█████▎ | 25341/48008 [1:26:55<1:16:50, 4.92it/s] 53%|█████▎ | 25342/48008 [1:26:56<1:15:06, 5.03it/s] 53%|█████▎ | 25343/48008 [1:26:56<1:27:14, 4.33it/s] 53%|█████▎ | 25344/48008 [1:26:56<1:23:00, 4.55it/s] 53%|█████▎ | 25345/48008 [1:26:56<1:20:03, 4.72it/s] 53%|█████▎ | 25346/48008 [1:26:57<1:17:34, 4.87it/s] 53%|█████▎ | 25347/48008 [1:26:57<1:15:59, 4.97it/s] 53%|█████▎ | 25348/48008 [1:26:57<1:14:51, 5.05it/s] 53%|█████▎ | 25349/48008 [1:26:57<1:13:54, 5.11it/s] 53%|█████▎ | 25350/48008 [1:26:57<1:13:10, 5.16it/s] {'loss': 0.3005, 'grad_norm': 0.1455078125, 'learning_rate': 9.439676720546576e-05, 'epoch': 0.53} 53%|█████▎ | 25350/48008 [1:26:57<1:13:10, 5.16it/s] 53%|█████▎ | 25351/48008 [1:26:58<1:14:11, 5.09it/s] 53%|█████▎ | 25352/48008 [1:26:58<1:14:25, 5.07it/s] 53%|█████▎ | 25353/48008 [1:26:58<1:12:42, 5.19it/s] 53%|█████▎ | 25354/48008 [1:26:58<1:12:22, 5.22it/s] 53%|█████▎ | 25355/48008 [1:26:58<1:25:13, 4.43it/s] 53%|█████▎ | 25356/48008 [1:26:59<1:21:13, 4.65it/s] 53%|█████▎ | 25357/48008 [1:26:59<1:18:19, 4.82it/s] 53%|█████▎ | 25358/48008 [1:26:59<1:15:47, 4.98it/s] 53%|█████▎ | 25359/48008 [1:26:59<1:14:56, 5.04it/s] 53%|█████▎ | 25360/48008 [1:26:59<1:14:51, 5.04it/s] 53%|█████▎ | 25361/48008 [1:27:00<1:13:48, 5.11it/s] 53%|█████▎ | 25362/48008 [1:27:00<1:35:20, 3.96it/s] 53%|█████▎ | 25363/48008 [1:27:00<1:27:51, 4.30it/s] 53%|█████▎ | 25364/48008 [1:27:00<1:23:15, 4.53it/s] 53%|█████▎ | 25365/48008 [1:27:00<1:19:55, 4.72it/s] 53%|█████▎ | 25366/48008 [1:27:01<1:17:21, 4.88it/s] 53%|█████▎ | 25367/48008 [1:27:01<1:16:15, 4.95it/s] 53%|█████▎ | 25368/48008 [1:27:01<1:15:01, 5.03it/s] 53%|█████▎ | 25369/48008 [1:27:01<1:15:40, 4.99it/s] 53%|█████▎ | 25370/48008 [1:27:01<1:14:29, 5.06it/s] 53%|█████▎ | 25371/48008 [1:27:02<1:13:57, 5.10it/s] 53%|█████▎ | 25372/48008 [1:27:02<1:12:58, 5.17it/s] 53%|█████▎ | 25373/48008 [1:27:02<1:14:06, 5.09it/s] 53%|█████▎ | 25374/48008 [1:27:02<1:12:44, 5.19it/s] 53%|█████▎ | 25375/48008 [1:27:02<1:12:30, 5.20it/s] 53%|█████▎ | 25376/48008 [1:27:03<1:14:07, 5.09it/s] 53%|█████▎ | 25377/48008 [1:27:03<1:13:04, 5.16it/s] 53%|█████▎ | 25378/48008 [1:27:03<1:25:35, 4.41it/s] 53%|█████▎ | 25379/48008 [1:27:03<1:21:59, 4.60it/s] 53%|█████▎ | 25380/48008 [1:27:04<1:18:54, 4.78it/s] 53%|█████▎ | 25381/48008 [1:27:04<1:16:26, 4.93it/s] 53%|█████▎ | 25382/48008 [1:27:04<1:15:34, 4.99it/s] 53%|█████▎ | 25383/48008 [1:27:04<1:15:52, 4.97it/s] 53%|█████▎ | 25384/48008 [1:27:04<1:14:42, 5.05it/s] 53%|█████▎ | 25385/48008 [1:27:05<1:36:02, 3.93it/s] 53%|█████▎ | 25386/48008 [1:27:05<1:28:51, 4.24it/s] 53%|█████▎ | 25387/48008 [1:27:05<1:36:50, 3.89it/s] 53%|█████▎ | 25388/48008 [1:27:05<1:29:09, 4.23it/s] 53%|█████▎ | 25389/48008 [1:27:06<1:23:26, 4.52it/s] 53%|█████▎ | 25390/48008 [1:27:06<1:19:56, 4.72it/s] 53%|█████▎ | 25391/48008 [1:27:06<1:17:26, 4.87it/s] 53%|█████▎ | 25392/48008 [1:27:06<1:15:21, 5.00it/s] 53%|█████▎ | 25393/48008 [1:27:06<1:14:25, 5.06it/s] 53%|█████▎ | 25394/48008 [1:27:07<1:16:30, 4.93it/s] 53%|█████▎ | 25395/48008 [1:27:07<1:15:37, 4.98it/s] 53%|█████▎ | 25396/48008 [1:27:07<1:14:25, 5.06it/s] 53%|█████▎ | 25397/48008 [1:27:07<1:14:38, 5.05it/s] 53%|█████▎ | 25398/48008 [1:27:07<1:15:13, 5.01it/s] 53%|█████▎ | 25399/48008 [1:27:08<1:16:11, 4.95it/s] 53%|█████▎ | 25400/48008 [1:27:08<1:15:03, 5.02it/s] {'loss': 0.3001, 'grad_norm': 0.2138671875, 'learning_rate': 9.418846858856858e-05, 'epoch': 0.53} 53%|█████▎ | 25400/48008 [1:27:08<1:15:03, 5.02it/s] 53%|█████▎ | 25401/48008 [1:27:08<1:13:57, 5.09it/s] 53%|█████▎ | 25402/48008 [1:27:08<1:14:42, 5.04it/s] 53%|█████▎ | 25403/48008 [1:27:08<1:14:36, 5.05it/s] 53%|█████▎ | 25404/48008 [1:27:08<1:13:42, 5.11it/s] 53%|█████▎ | 25405/48008 [1:27:09<1:14:22, 5.07it/s] 53%|█████▎ | 25406/48008 [1:27:09<1:13:56, 5.09it/s] 53%|█████▎ | 25407/48008 [1:27:09<1:14:12, 5.08it/s] 53%|█████▎ | 25408/48008 [1:27:09<1:35:38, 3.94it/s] 53%|█████▎ | 25409/48008 [1:27:10<1:27:54, 4.28it/s] 53%|█████▎ | 25410/48008 [1:27:10<1:22:22, 4.57it/s] 53%|█████▎ | 25411/48008 [1:27:10<1:19:34, 4.73it/s] 53%|█████▎ | 25412/48008 [1:27:10<1:17:15, 4.87it/s] 53%|█████▎ | 25413/48008 [1:27:11<1:28:41, 4.25it/s] 53%|█████▎ | 25414/48008 [1:27:11<1:28:50, 4.24it/s] 53%|█████▎ | 25415/48008 [1:27:11<1:23:46, 4.49it/s] 53%|█████▎ | 25416/48008 [1:27:11<1:19:43, 4.72it/s] 53%|█████▎ | 25417/48008 [1:27:11<1:17:09, 4.88it/s] 53%|█████▎ | 25418/48008 [1:27:12<1:15:40, 4.98it/s] 53%|█████▎ | 25419/48008 [1:27:12<1:14:37, 5.05it/s] 53%|█████▎ | 25420/48008 [1:27:12<1:26:35, 4.35it/s] 53%|█████▎ | 25421/48008 [1:27:12<1:21:10, 4.64it/s] 53%|█████▎ | 25422/48008 [1:27:12<1:18:19, 4.81it/s] 53%|█████▎ | 25423/48008 [1:27:13<1:17:16, 4.87it/s] 53%|█████▎ | 25424/48008 [1:27:13<1:15:30, 4.98it/s] 53%|█████▎ | 25425/48008 [1:27:13<1:14:17, 5.07it/s] 53%|█████▎ | 25426/48008 [1:27:13<1:13:18, 5.13it/s] 53%|█████▎ | 25427/48008 [1:27:13<1:14:21, 5.06it/s] 53%|█████▎ | 25428/48008 [1:27:14<1:13:05, 5.15it/s] 53%|█████▎ | 25429/48008 [1:27:14<1:12:37, 5.18it/s] 53%|█████▎ | 25430/48008 [1:27:14<1:13:47, 5.10it/s] 53%|█████▎ | 25431/48008 [1:27:14<1:12:56, 5.16it/s] 53%|█████▎ | 25432/48008 [1:27:14<1:14:18, 5.06it/s] 53%|█████▎ | 25433/48008 [1:27:15<1:14:47, 5.03it/s] 53%|█████▎ | 25434/48008 [1:27:15<1:13:56, 5.09it/s] 53%|█████▎ | 25435/48008 [1:27:15<1:14:44, 5.03it/s] 53%|█████▎ | 25436/48008 [1:27:15<1:13:37, 5.11it/s] 53%|█████▎ | 25437/48008 [1:27:15<1:12:47, 5.17it/s] 53%|█████▎ | 25438/48008 [1:27:16<1:13:43, 5.10it/s] 53%|█████▎ | 25439/48008 [1:27:16<1:15:08, 5.01it/s] 53%|█████▎ | 25440/48008 [1:27:16<1:14:10, 5.07it/s] 53%|█████▎ | 25441/48008 [1:27:16<1:13:45, 5.10it/s] 53%|█████▎ | 25442/48008 [1:27:16<1:13:08, 5.14it/s] 53%|█████▎ | 25443/48008 [1:27:16<1:12:42, 5.17it/s] 53%|█████▎ | 25444/48008 [1:27:17<1:12:25, 5.19it/s] 53%|█████▎ | 25445/48008 [1:27:17<1:12:06, 5.21it/s] 53%|█████▎ | 25446/48008 [1:27:17<1:11:50, 5.23it/s] 53%|█████▎ | 25447/48008 [1:27:17<1:11:24, 5.27it/s] 53%|█████▎ | 25448/48008 [1:27:17<1:11:49, 5.23it/s] 53%|█████▎ | 25449/48008 [1:27:18<1:12:25, 5.19it/s] 53%|█████▎ | 25450/48008 [1:27:18<1:17:19, 4.86it/s] {'loss': 0.2988, 'grad_norm': 0.34375, 'learning_rate': 9.39801699716714e-05, 'epoch': 0.53} 53%|█████▎ | 25450/48008 [1:27:18<1:17:19, 4.86it/s] 53%|█████▎ | 25451/48008 [1:27:18<1:37:34, 3.85it/s] 53%|█████▎ | 25452/48008 [1:27:18<1:29:21, 4.21it/s] 53%|█████▎ | 25453/48008 [1:27:19<1:24:22, 4.46it/s] 53%|█████▎ | 25454/48008 [1:27:19<1:21:46, 4.60it/s] 53%|█████▎ | 25455/48008 [1:27:19<1:20:25, 4.67it/s] 53%|█████▎ | 25456/48008 [1:27:19<1:18:07, 4.81it/s] 53%|█████▎ | 25457/48008 [1:27:19<1:17:03, 4.88it/s] 53%|█████▎ | 25458/48008 [1:27:20<1:15:48, 4.96it/s] 53%|█████▎ | 25459/48008 [1:27:20<1:14:43, 5.03it/s] 53%|█████▎ | 25460/48008 [1:27:20<1:14:03, 5.07it/s] 53%|█████▎ | 25461/48008 [1:27:20<1:18:35, 4.78it/s] 53%|█████▎ | 25462/48008 [1:27:20<1:16:38, 4.90it/s] 53%|█████▎ | 25463/48008 [1:27:21<1:15:12, 5.00it/s] 53%|█████▎ | 25464/48008 [1:27:21<1:14:16, 5.06it/s] 53%|█████▎ | 25465/48008 [1:27:21<1:14:58, 5.01it/s] 53%|█████▎ | 25466/48008 [1:27:21<1:14:32, 5.04it/s] 53%|█████▎ | 25467/48008 [1:27:21<1:12:55, 5.15it/s] 53%|█████▎ | 25468/48008 [1:27:22<1:13:20, 5.12it/s] 53%|█████▎ | 25469/48008 [1:27:22<1:13:02, 5.14it/s] 53%|█████▎ | 25470/48008 [1:27:22<1:13:09, 5.13it/s] 53%|█████▎ | 25471/48008 [1:27:22<1:15:09, 5.00it/s] 53%|█████▎ | 25472/48008 [1:27:22<1:14:18, 5.05it/s] 53%|█████▎ | 25473/48008 [1:27:23<1:14:17, 5.06it/s] 53%|█████▎ | 25474/48008 [1:27:23<1:14:56, 5.01it/s] 53%|█████▎ | 25475/48008 [1:27:23<1:14:59, 5.01it/s] 53%|█████▎ | 25476/48008 [1:27:23<1:15:28, 4.98it/s] 53%|█████▎ | 25477/48008 [1:27:23<1:14:12, 5.06it/s] 53%|█████▎ | 25478/48008 [1:27:24<1:35:21, 3.94it/s] 53%|█████▎ | 25479/48008 [1:27:24<1:28:45, 4.23it/s] 53%|█████▎ | 25480/48008 [1:27:24<1:24:14, 4.46it/s] 53%|█████▎ | 25481/48008 [1:27:24<1:20:14, 4.68it/s] 53%|█████▎ | 25482/48008 [1:27:25<1:19:38, 4.71it/s] 53%|█████▎ | 25483/48008 [1:27:25<1:16:45, 4.89it/s] 53%|█████▎ | 25484/48008 [1:27:25<1:14:34, 5.03it/s] 53%|█████▎ | 25485/48008 [1:27:25<1:15:03, 5.00it/s] 53%|█████▎ | 25486/48008 [1:27:25<1:16:23, 4.91it/s] 53%|█████▎ | 25487/48008 [1:27:26<1:14:55, 5.01it/s] 53%|█████▎ | 25488/48008 [1:27:26<1:14:45, 5.02it/s] 53%|█████▎ | 25489/48008 [1:27:26<1:19:02, 4.75it/s] 53%|█████▎ | 25490/48008 [1:27:26<1:16:49, 4.89it/s] 53%|█████▎ | 25491/48008 [1:27:26<1:15:25, 4.98it/s] 53%|█████▎ | 25492/48008 [1:27:27<1:14:03, 5.07it/s] 53%|█████▎ | 25493/48008 [1:27:27<1:15:21, 4.98it/s] 53%|█████▎ | 25494/48008 [1:27:27<1:14:02, 5.07it/s] 53%|█████▎ | 25495/48008 [1:27:27<1:13:04, 5.13it/s] 53%|█████▎ | 25496/48008 [1:27:27<1:12:22, 5.18it/s] 53%|█████▎ | 25497/48008 [1:27:28<1:12:25, 5.18it/s] 53%|█████▎ | 25498/48008 [1:27:28<1:17:19, 4.85it/s] 53%|█████▎ | 25499/48008 [1:27:28<1:16:03, 4.93it/s] 53%|█████▎ | 25500/48008 [1:27:28<1:14:35, 5.03it/s] {'loss': 0.2997, 'grad_norm': 0.1513671875, 'learning_rate': 9.37718713547742e-05, 'epoch': 0.53} 53%|█████▎ | 25500/48008 [1:27:28<1:14:35, 5.03it/s] 53%|█████▎ | 25501/48008 [1:27:28<1:15:54, 4.94it/s] 53%|█████▎ | 25502/48008 [1:27:29<1:14:02, 5.07it/s] 53%|█████▎ | 25503/48008 [1:27:29<1:13:47, 5.08it/s] 53%|█████▎ | 25504/48008 [1:27:29<1:12:57, 5.14it/s] 53%|█████▎ | 25505/48008 [1:27:29<1:12:19, 5.19it/s] 53%|█████▎ | 25506/48008 [1:27:29<1:11:56, 5.21it/s] 53%|█████▎ | 25507/48008 [1:27:29<1:11:06, 5.27it/s] 53%|█████▎ | 25508/48008 [1:27:30<1:13:26, 5.11it/s] 53%|█████▎ | 25509/48008 [1:27:30<1:12:53, 5.14it/s] 53%|█████▎ | 25510/48008 [1:27:30<1:13:06, 5.13it/s] 53%|█████▎ | 25511/48008 [1:27:30<1:12:40, 5.16it/s] 53%|█████▎ | 25512/48008 [1:27:30<1:11:36, 5.24it/s] 53%|█████▎ | 25513/48008 [1:27:31<1:11:23, 5.25it/s] 53%|█████▎ | 25514/48008 [1:27:31<1:11:16, 5.26it/s] 53%|█████▎ | 25515/48008 [1:27:31<1:11:08, 5.27it/s] 53%|█████▎ | 25516/48008 [1:27:31<1:10:52, 5.29it/s] 53%|█████▎ | 25517/48008 [1:27:31<1:10:46, 5.30it/s] 53%|█████▎ | 25518/48008 [1:27:32<1:11:50, 5.22it/s] 53%|█████▎ | 25519/48008 [1:27:32<1:13:55, 5.07it/s] 53%|█████▎ | 25520/48008 [1:27:32<1:14:32, 5.03it/s] 53%|█████▎ | 25521/48008 [1:27:32<1:14:52, 5.01it/s] 53%|█████▎ | 25522/48008 [1:27:32<1:13:51, 5.07it/s] 53%|█████▎ | 25523/48008 [1:27:33<1:12:06, 5.20it/s] 53%|█████▎ | 25524/48008 [1:27:33<1:13:12, 5.12it/s] 53%|█████▎ | 25525/48008 [1:27:33<1:12:12, 5.19it/s] 53%|█████▎ | 25526/48008 [1:27:33<1:12:44, 5.15it/s] 53%|█████▎ | 25527/48008 [1:27:33<1:12:06, 5.20it/s] 53%|█████▎ | 25528/48008 [1:27:34<1:12:11, 5.19it/s] 53%|█████▎ | 25529/48008 [1:27:34<1:12:20, 5.18it/s] 53%|█████▎ | 25530/48008 [1:27:34<1:12:09, 5.19it/s] 53%|█████▎ | 25531/48008 [1:27:34<1:11:47, 5.22it/s] 53%|█████▎ | 25532/48008 [1:27:34<1:13:39, 5.09it/s] 53%|█████▎ | 25533/48008 [1:27:35<1:12:53, 5.14it/s] 53%|█████▎ | 25534/48008 [1:27:35<1:12:19, 5.18it/s] 53%|█████▎ | 25535/48008 [1:27:35<1:12:04, 5.20it/s] 53%|█████▎ | 25536/48008 [1:27:35<1:11:23, 5.25it/s] 53%|█████▎ | 25537/48008 [1:27:35<1:10:58, 5.28it/s] 53%|█████▎ | 25538/48008 [1:27:35<1:10:59, 5.28it/s] 53%|█████▎ | 25539/48008 [1:27:36<1:16:23, 4.90it/s] 53%|█████▎ | 25540/48008 [1:27:36<1:14:45, 5.01it/s] 53%|█████▎ | 25541/48008 [1:27:36<1:15:08, 4.98it/s] 53%|█████▎ | 25542/48008 [1:27:36<1:14:13, 5.04it/s] 53%|█████▎ | 25543/48008 [1:27:37<1:14:04, 5.05it/s] 53%|█████▎ | 25544/48008 [1:27:37<1:13:46, 5.07it/s] 53%|█████▎ | 25545/48008 [1:27:37<1:13:07, 5.12it/s] 53%|█████▎ | 25546/48008 [1:27:37<1:13:07, 5.12it/s] 53%|█████▎ | 25547/48008 [1:27:37<1:12:48, 5.14it/s] 53%|█████▎ | 25548/48008 [1:27:37<1:12:19, 5.18it/s] 53%|█████▎ | 25549/48008 [1:27:38<1:12:12, 5.18it/s] 53%|█████▎ | 25550/48008 [1:27:38<1:13:08, 5.12it/s] 53%|█████▎ | 25550/48008 [1:27:38<1:13:08, 5.12it/s]{'loss': 0.3026, 'grad_norm': 0.134765625, 'learning_rate': 9.356357273787702e-05, 'epoch': 0.53} 53%|█████▎ | 25551/48008 [1:27:38<1:12:43, 5.15it/s] 53%|█████▎ | 25552/48008 [1:27:38<1:12:57, 5.13it/s] 53%|█████▎ | 25553/48008 [1:27:38<1:12:31, 5.16it/s] 53%|█████▎ | 25554/48008 [1:27:39<1:12:11, 5.18it/s] 53%|█████▎ | 25555/48008 [1:27:39<1:12:13, 5.18it/s] 53%|█████▎ | 25556/48008 [1:27:39<1:12:47, 5.14it/s] 53%|█████▎ | 25557/48008 [1:27:39<1:12:16, 5.18it/s] 53%|█████▎ | 25558/48008 [1:27:39<1:12:01, 5.19it/s] 53%|█████▎ | 25559/48008 [1:27:40<1:11:26, 5.24it/s] 53%|█████▎ | 25560/48008 [1:27:40<1:10:45, 5.29it/s] 53%|█████▎ | 25561/48008 [1:27:40<1:10:56, 5.27it/s] 53%|█████▎ | 25562/48008 [1:27:40<1:11:09, 5.26it/s] 53%|█████▎ | 25563/48008 [1:27:40<1:11:11, 5.25it/s] 53%|█████▎ | 25564/48008 [1:27:41<1:11:02, 5.27it/s] 53%|█████▎ | 25565/48008 [1:27:41<1:11:08, 5.26it/s] 53%|█████▎ | 25566/48008 [1:27:41<1:10:50, 5.28it/s] 53%|█████▎ | 25567/48008 [1:27:41<1:10:45, 5.29it/s] 53%|█████▎ | 25568/48008 [1:27:41<1:10:47, 5.28it/s] 53%|█████▎ | 25569/48008 [1:27:42<1:16:11, 4.91it/s] 53%|█████▎ | 25570/48008 [1:27:42<1:14:38, 5.01it/s] 53%|█████▎ | 25571/48008 [1:27:42<1:14:56, 4.99it/s] 53%|█████▎ | 25572/48008 [1:27:42<1:15:20, 4.96it/s] 53%|█████▎ | 25573/48008 [1:27:42<1:14:15, 5.04it/s] 53%|█████▎ | 25574/48008 [1:27:43<1:14:35, 5.01it/s] 53%|█████▎ | 25575/48008 [1:27:43<1:15:37, 4.94it/s] 53%|█████▎ | 25576/48008 [1:27:43<1:14:24, 5.02it/s] 53%|█████▎ | 25577/48008 [1:27:43<1:13:30, 5.09it/s] 53%|█████▎ | 25578/48008 [1:27:43<1:12:38, 5.15it/s] 53%|█████▎ | 25579/48008 [1:27:43<1:12:19, 5.17it/s] 53%|█████▎ | 25580/48008 [1:27:44<1:12:47, 5.13it/s] 53%|█████▎ | 25581/48008 [1:27:44<1:11:58, 5.19it/s] 53%|█████▎ | 25582/48008 [1:27:44<1:11:38, 5.22it/s] 53%|█████▎ | 25583/48008 [1:27:44<1:12:43, 5.14it/s] 53%|█████▎ | 25584/48008 [1:27:44<1:12:15, 5.17it/s] 53%|█████▎ | 25585/48008 [1:27:45<1:12:01, 5.19it/s] 53%|█████▎ | 25586/48008 [1:27:45<1:13:49, 5.06it/s] 53%|█████▎ | 25587/48008 [1:27:45<1:14:09, 5.04it/s] 53%|█████▎ | 25588/48008 [1:27:45<1:13:37, 5.08it/s] 53%|█████▎ | 25589/48008 [1:27:45<1:13:12, 5.10it/s] 53%|█████▎ | 25590/48008 [1:27:46<1:12:34, 5.15it/s] 53%|█████▎ | 25591/48008 [1:27:46<1:12:15, 5.17it/s] 53%|█████▎ | 25592/48008 [1:27:46<1:12:29, 5.15it/s] 53%|█████▎ | 25593/48008 [1:27:46<1:11:38, 5.21it/s] 53%|█████▎ | 25594/48008 [1:27:46<1:11:18, 5.24it/s] 53%|█████▎ | 25595/48008 [1:27:47<1:11:18, 5.24it/s] 53%|█████▎ | 25596/48008 [1:27:47<1:11:26, 5.23it/s] 53%|█████▎ | 25597/48008 [1:27:47<1:11:25, 5.23it/s] 53%|█████▎ | 25598/48008 [1:27:47<1:12:57, 5.12it/s] 53%|█████▎ | 25599/48008 [1:27:47<1:12:58, 5.12it/s] 53%|█████▎ | 25600/48008 [1:27:48<1:12:24, 5.16it/s] {'loss': 0.2997, 'grad_norm': 0.2431640625, 'learning_rate': 9.335527412097984e-05, 'epoch': 0.53} 53%|█████▎ | 25600/48008 [1:27:48<1:12:24, 5.16it/s] 53%|█████▎ | 25601/48008 [1:27:48<1:11:57, 5.19it/s] 53%|█████▎ | 25602/48008 [1:27:48<1:11:26, 5.23it/s] 53%|█████▎ | 25603/48008 [1:27:48<1:10:59, 5.26it/s] 53%|█████▎ | 25604/48008 [1:27:48<1:11:04, 5.25it/s] 53%|█████▎ | 25605/48008 [1:27:49<1:13:08, 5.11it/s] 53%|█████▎ | 25606/48008 [1:27:49<1:13:45, 5.06it/s] 53%|█████▎ | 25607/48008 [1:27:49<1:12:17, 5.16it/s] 53%|█████▎ | 25608/48008 [1:27:49<1:11:34, 5.22it/s] 53%|█████▎ | 25609/48008 [1:27:49<1:11:24, 5.23it/s] 53%|█████▎ | 25610/48008 [1:27:49<1:10:41, 5.28it/s] 53%|█████▎ | 25611/48008 [1:27:50<1:10:53, 5.27it/s] 53%|█████▎ | 25612/48008 [1:27:50<1:16:02, 4.91it/s] 53%|█████▎ | 25613/48008 [1:27:50<1:15:53, 4.92it/s] 53%|█████▎ | 25614/48008 [1:27:50<1:14:19, 5.02it/s] 53%|█████▎ | 25615/48008 [1:27:51<1:18:29, 4.76it/s] 53%|█████▎ | 25616/48008 [1:27:51<1:16:23, 4.89it/s] 53%|█████▎ | 25617/48008 [1:27:51<1:14:17, 5.02it/s] 53%|█████▎ | 25618/48008 [1:27:51<1:12:54, 5.12it/s] 53%|█████▎ | 25619/48008 [1:27:51<1:12:01, 5.18it/s] 53%|█████▎ | 25620/48008 [1:27:51<1:11:44, 5.20it/s] 53%|█████▎ | 25621/48008 [1:27:52<1:11:52, 5.19it/s] 53%|█████▎ | 25622/48008 [1:27:52<1:24:43, 4.40it/s] 53%|█████▎ | 25623/48008 [1:27:52<1:21:45, 4.56it/s] 53%|█████▎ | 25624/48008 [1:27:52<1:18:26, 4.76it/s] 53%|█████▎ | 25625/48008 [1:27:53<1:16:32, 4.87it/s] 53%|█████▎ | 25626/48008 [1:27:53<1:14:33, 5.00it/s] 53%|█████▎ | 25627/48008 [1:27:53<1:13:20, 5.09it/s] 53%|█████▎ | 25628/48008 [1:27:53<1:15:01, 4.97it/s] 53%|█████▎ | 25629/48008 [1:27:53<1:14:31, 5.00it/s] 53%|█████▎ | 25630/48008 [1:27:54<1:18:47, 4.73it/s] 53%|█████▎ | 25631/48008 [1:27:54<1:16:39, 4.87it/s] 53%|█████▎ | 25632/48008 [1:27:54<1:16:13, 4.89it/s] 53%|█████▎ | 25633/48008 [1:27:54<1:14:33, 5.00it/s] 53%|█████▎ | 25634/48008 [1:27:54<1:15:12, 4.96it/s] 53%|█████▎ | 25635/48008 [1:27:55<1:14:16, 5.02it/s] 53%|█████▎ | 25636/48008 [1:27:55<1:35:14, 3.92it/s] 53%|█████▎ | 25637/48008 [1:27:55<1:29:09, 4.18it/s] 53%|█████▎ | 25638/48008 [1:27:55<1:25:53, 4.34it/s] 53%|█████▎ | 25639/48008 [1:27:56<1:21:27, 4.58it/s] 53%|█████▎ | 25640/48008 [1:27:56<1:18:24, 4.75it/s] 53%|█████▎ | 25641/48008 [1:27:56<1:16:48, 4.85it/s] 53%|█████▎ | 25642/48008 [1:27:56<1:16:30, 4.87it/s] 53%|█████▎ | 25643/48008 [1:27:56<1:14:49, 4.98it/s] 53%|█████▎ | 25644/48008 [1:27:57<1:14:52, 4.98it/s] 53%|█████▎ | 25645/48008 [1:27:57<1:14:23, 5.01it/s] 53%|█████▎ | 25646/48008 [1:27:57<1:15:00, 4.97it/s] 53%|█████▎ | 25647/48008 [1:27:57<1:15:57, 4.91it/s] 53%|█████▎ | 25648/48008 [1:27:57<1:15:10, 4.96it/s] 53%|█████▎ | 25649/48008 [1:27:58<1:14:19, 5.01it/s] 53%|█████▎ | 25650/48008 [1:27:58<1:13:31, 5.07it/s] 53%|█████▎ | 25650/48008 [1:27:58<1:13:31, 5.07it/s]{'loss': 0.3017, 'grad_norm': 0.703125, 'learning_rate': 9.314697550408266e-05, 'epoch': 0.53} 53%|█████▎ | 25651/48008 [1:27:58<1:14:02, 5.03it/s] 53%|█████▎ | 25652/48008 [1:27:58<1:13:40, 5.06it/s] 53%|█████▎ | 25653/48008 [1:27:58<1:12:55, 5.11it/s] 53%|█████▎ | 25654/48008 [1:27:59<1:12:52, 5.11it/s] 53%|█████▎ | 25655/48008 [1:27:59<1:12:17, 5.15it/s] 53%|█████▎ | 25656/48008 [1:27:59<1:12:07, 5.17it/s] 53%|█████▎ | 25657/48008 [1:27:59<1:13:10, 5.09it/s] 53%|█████▎ | 25658/48008 [1:27:59<1:12:17, 5.15it/s] 53%|█████▎ | 25659/48008 [1:27:59<1:11:12, 5.23it/s] 53%|█████▎ | 25660/48008 [1:28:00<1:10:03, 5.32it/s] 53%|█████▎ | 25661/48008 [1:28:00<1:11:44, 5.19it/s] 53%|█████▎ | 25662/48008 [1:28:00<1:11:18, 5.22it/s] 53%|█████▎ | 25663/48008 [1:28:00<1:11:16, 5.22it/s] 53%|█████▎ | 25664/48008 [1:28:00<1:16:30, 4.87it/s] 53%|█████▎ | 25665/48008 [1:28:01<1:14:40, 4.99it/s] 53%|█████▎ | 25666/48008 [1:28:01<1:14:49, 4.98it/s] 53%|█████▎ | 25667/48008 [1:28:01<1:14:22, 5.01it/s] 53%|█████▎ | 25668/48008 [1:28:01<1:13:09, 5.09it/s] 53%|█████▎ | 25669/48008 [1:28:01<1:12:40, 5.12it/s] 53%|█████▎ | 25670/48008 [1:28:02<1:11:49, 5.18it/s] 53%|█████▎ | 25671/48008 [1:28:02<1:11:09, 5.23it/s] 53%|█████▎ | 25672/48008 [1:28:02<1:16:18, 4.88it/s] 53%|█████▎ | 25673/48008 [1:28:02<1:20:02, 4.65it/s] 53%|█████▎ | 25674/48008 [1:28:02<1:17:25, 4.81it/s] 53%|█████▎ | 25675/48008 [1:28:03<1:16:37, 4.86it/s] 53%|█████▎ | 25676/48008 [1:28:03<1:15:15, 4.95it/s] 53%|█████▎ | 25677/48008 [1:28:03<1:16:00, 4.90it/s] 53%|█████▎ | 25678/48008 [1:28:03<1:14:35, 4.99it/s] 53%|█████▎ | 25679/48008 [1:28:03<1:14:16, 5.01it/s] 53%|█████▎ | 25680/48008 [1:28:04<1:13:15, 5.08it/s] 53%|█████▎ | 25681/48008 [1:28:04<1:12:29, 5.13it/s] 53%|█████▎ | 25682/48008 [1:28:04<1:33:59, 3.96it/s] 53%|█████▎ | 25683/48008 [1:28:04<1:27:54, 4.23it/s] 53%|█████▎ | 25684/48008 [1:28:05<1:23:33, 4.45it/s] 54%|█████▎ | 25685/48008 [1:28:05<1:21:13, 4.58it/s] 54%|█████▎ | 25686/48008 [1:28:05<1:18:27, 4.74it/s] 54%|█████▎ | 25687/48008 [1:28:05<1:16:29, 4.86it/s] 54%|█████▎ | 25688/48008 [1:28:05<1:14:42, 4.98it/s] 54%|█████▎ | 25689/48008 [1:28:06<1:13:54, 5.03it/s] 54%|█████▎ | 25690/48008 [1:28:06<1:13:16, 5.08it/s] 54%|█████▎ | 25691/48008 [1:28:06<1:12:23, 5.14it/s] 54%|█████▎ | 25692/48008 [1:28:06<1:12:18, 5.14it/s] 54%|█████▎ | 25693/48008 [1:28:06<1:13:28, 5.06it/s] 54%|█████▎ | 25694/48008 [1:28:07<1:12:30, 5.13it/s] 54%|█████▎ | 25695/48008 [1:28:07<1:12:04, 5.16it/s] 54%|█████▎ | 25696/48008 [1:28:07<1:13:10, 5.08it/s] 54%|█████▎ | 25697/48008 [1:28:07<1:12:45, 5.11it/s] 54%|█████▎ | 25698/48008 [1:28:07<1:12:03, 5.16it/s] 54%|█████▎ | 25699/48008 [1:28:08<1:11:44, 5.18it/s] 54%|█████▎ | 25700/48008 [1:28:08<1:13:43, 5.04it/s] {'loss': 0.2989, 'grad_norm': 0.330078125, 'learning_rate': 9.293867688718547e-05, 'epoch': 0.54} 54%|█████▎ | 25700/48008 [1:28:08<1:13:43, 5.04it/s] 54%|█████▎ | 25701/48008 [1:28:08<1:12:14, 5.15it/s] 54%|█████▎ | 25702/48008 [1:28:08<1:11:57, 5.17it/s] 54%|█████▎ | 25703/48008 [1:28:08<1:12:13, 5.15it/s] 54%|█████▎ | 25704/48008 [1:28:09<1:13:16, 5.07it/s] 54%|█████▎ | 25705/48008 [1:28:09<1:14:00, 5.02it/s] 54%|█████▎ | 25706/48008 [1:28:09<1:12:35, 5.12it/s] 54%|█████▎ | 25707/48008 [1:28:09<1:12:01, 5.16it/s] 54%|█████▎ | 25708/48008 [1:28:09<1:12:59, 5.09it/s] 54%|█████▎ | 25709/48008 [1:28:10<1:13:34, 5.05it/s] 54%|█████▎ | 25710/48008 [1:28:10<1:17:59, 4.77it/s] 54%|█████▎ | 25711/48008 [1:28:10<1:15:46, 4.90it/s] 54%|█████▎ | 25712/48008 [1:28:10<1:14:45, 4.97it/s] 54%|█████▎ | 25713/48008 [1:28:10<1:14:43, 4.97it/s] 54%|█████▎ | 25714/48008 [1:28:11<1:13:35, 5.05it/s] 54%|█████▎ | 25715/48008 [1:28:11<1:12:54, 5.10it/s] 54%|█████▎ | 25716/48008 [1:28:11<1:11:51, 5.17it/s] 54%|█████▎ | 25717/48008 [1:28:11<1:12:21, 5.13it/s] 54%|█████▎ | 25718/48008 [1:28:11<1:11:48, 5.17it/s] 54%|█████▎ | 25719/48008 [1:28:11<1:11:06, 5.22it/s] 54%|█████▎ | 25720/48008 [1:28:12<1:13:03, 5.08it/s] 54%|█████▎ | 25721/48008 [1:28:12<1:12:13, 5.14it/s] 54%|█████▎ | 25722/48008 [1:28:13<2:58:45, 2.08it/s] 54%|█████▎ | 25723/48008 [1:28:13<2:48:27, 2.20it/s] 54%|█████▎ | 25724/48008 [1:28:14<2:19:39, 2.66it/s] 54%|█████▎ | 25725/48008 [1:28:14<1:59:47, 3.10it/s] 54%|█████▎ | 25726/48008 [1:28:14<1:44:40, 3.55it/s] 54%|█████▎ | 25727/48008 [1:28:14<1:34:01, 3.95it/s] 54%|█████▎ | 25728/48008 [1:28:14<1:27:09, 4.26it/s] 54%|█████▎ | 25729/48008 [1:28:15<1:22:28, 4.50it/s] 54%|█████▎ | 25730/48008 [1:28:15<1:19:04, 4.70it/s] 54%|█████▎ | 25731/48008 [1:28:15<1:17:53, 4.77it/s] 54%|█████▎ | 25732/48008 [1:28:15<1:15:42, 4.90it/s] 54%|█████▎ | 25733/48008 [1:28:15<1:15:10, 4.94it/s] 54%|█████▎ | 25734/48008 [1:28:16<1:14:33, 4.98it/s] 54%|█████▎ | 25735/48008 [1:28:16<1:13:03, 5.08it/s] 54%|█████▎ | 25736/48008 [1:28:16<1:12:28, 5.12it/s] 54%|█████▎ | 25737/48008 [1:28:16<1:12:15, 5.14it/s] 54%|█████▎ | 25738/48008 [1:28:16<1:13:16, 5.07it/s] 54%|█████▎ | 25739/48008 [1:28:17<1:12:40, 5.11it/s] 54%|█████▎ | 25740/48008 [1:28:17<1:12:18, 5.13it/s] 54%|█████▎ | 25741/48008 [1:28:17<1:14:14, 5.00it/s] 54%|█████▎ | 25742/48008 [1:28:17<1:14:05, 5.01it/s] 54%|█████▎ | 25743/48008 [1:28:17<1:12:58, 5.08it/s] 54%|█████▎ | 25744/48008 [1:28:18<1:12:51, 5.09it/s] 54%|█████▎ | 25745/48008 [1:28:18<1:11:55, 5.16it/s] 54%|█████▎ | 25746/48008 [1:28:18<1:33:40, 3.96it/s] 54%|█████▎ | 25747/48008 [1:28:18<1:26:39, 4.28it/s] 54%|█████▎ | 25748/48008 [1:28:18<1:22:28, 4.50it/s] 54%|█████▎ | 25749/48008 [1:28:19<1:18:50, 4.71it/s] 54%|█████▎ | 25750/48008 [1:28:19<1:15:44, 4.90it/s] {'loss': 0.3014, 'grad_norm': 0.232421875, 'learning_rate': 9.273037827028829e-05, 'epoch': 0.54} 54%|█████▎ | 25750/48008 [1:28:19<1:15:44, 4.90it/s] 54%|█████▎ | 25751/48008 [1:28:19<1:15:09, 4.94it/s] 54%|█████▎ | 25752/48008 [1:28:19<1:14:01, 5.01it/s] 54%|█████▎ | 25753/48008 [1:28:19<1:13:26, 5.05it/s] 54%|█████▎ | 25754/48008 [1:28:20<1:12:53, 5.09it/s] 54%|█████▎ | 25755/48008 [1:28:20<1:12:09, 5.14it/s] 54%|█████▎ | 25756/48008 [1:28:20<1:11:42, 5.17it/s] 54%|█████▎ | 25757/48008 [1:28:20<1:11:33, 5.18it/s] 54%|█████▎ | 25758/48008 [1:28:20<1:12:04, 5.14it/s] 54%|█████▎ | 25759/48008 [1:28:21<1:12:15, 5.13it/s] 54%|█████▎ | 25760/48008 [1:28:21<1:11:35, 5.18it/s] 54%|█████▎ | 25761/48008 [1:28:21<1:12:15, 5.13it/s] 54%|█████▎ | 25762/48008 [1:28:21<1:11:51, 5.16it/s] 54%|█████▎ | 25763/48008 [1:28:21<1:11:24, 5.19it/s] 54%|█████▎ | 25764/48008 [1:28:22<1:11:01, 5.22it/s] 54%|█████▎ | 25765/48008 [1:28:22<1:16:14, 4.86it/s] 54%|█████▎ | 25766/48008 [1:28:22<1:14:01, 5.01it/s] 54%|█████▎ | 25767/48008 [1:28:22<1:13:38, 5.03it/s] 54%|█████▎ | 25768/48008 [1:28:22<1:13:28, 5.05it/s] 54%|█████▎ | 25769/48008 [1:28:23<1:12:55, 5.08it/s] 54%|█████▎ | 25770/48008 [1:28:23<1:11:38, 5.17it/s] 54%|█████▎ | 25771/48008 [1:28:23<1:11:21, 5.19it/s] 54%|█████▎ | 25772/48008 [1:28:23<1:23:56, 4.41it/s] 54%|█████▎ | 25773/48008 [1:28:23<1:19:48, 4.64it/s] 54%|█████▎ | 25774/48008 [1:28:24<1:19:10, 4.68it/s] 54%|█████▎ | 25775/48008 [1:28:24<1:17:00, 4.81it/s] 54%|█████▎ | 25776/48008 [1:28:24<1:14:36, 4.97it/s] 54%|█████▎ | 25777/48008 [1:28:24<1:13:38, 5.03it/s] 54%|█████▎ | 25778/48008 [1:28:24<1:14:37, 4.97it/s] 54%|█████▎ | 25779/48008 [1:28:25<1:13:14, 5.06it/s] 54%|█████▎ | 25780/48008 [1:28:25<1:12:18, 5.12it/s] 54%|█████▎ | 25781/48008 [1:28:25<1:17:02, 4.81it/s] 54%|█████▎ | 25782/48008 [1:28:25<1:14:30, 4.97it/s] 54%|█████▎ | 25783/48008 [1:28:25<1:13:25, 5.05it/s] 54%|█████▎ | 25784/48008 [1:28:26<1:14:09, 4.99it/s] 54%|█████▎ | 25785/48008 [1:28:26<1:13:35, 5.03it/s] 54%|█████▎ | 25786/48008 [1:28:26<1:12:31, 5.11it/s] 54%|█████▎ | 25787/48008 [1:28:26<1:11:54, 5.15it/s] 54%|█████▎ | 25788/48008 [1:28:26<1:11:24, 5.19it/s] 54%|█████▎ | 25789/48008 [1:28:27<1:12:34, 5.10it/s] 54%|█████▎ | 25790/48008 [1:28:27<1:12:34, 5.10it/s] 54%|█████▎ | 25791/48008 [1:28:27<1:12:42, 5.09it/s] 54%|█████▎ | 25792/48008 [1:28:27<1:12:12, 5.13it/s] 54%|█████▎ | 25793/48008 [1:28:27<1:11:42, 5.16it/s] 54%|█████▎ | 25794/48008 [1:28:28<1:11:22, 5.19it/s] 54%|█████▎ | 25795/48008 [1:28:28<1:10:37, 5.24it/s] 54%|█████▎ | 25796/48008 [1:28:28<1:15:53, 4.88it/s] 54%|█████▎ | 25797/48008 [1:28:28<1:15:26, 4.91it/s] 54%|█████▎ | 25798/48008 [1:28:28<1:15:18, 4.92it/s] 54%|█████▎ | 25799/48008 [1:28:29<1:15:56, 4.87it/s] 54%|█████▎ | 25800/48008 [1:28:29<1:27:10, 4.25it/s] {'loss': 0.2996, 'grad_norm': 0.37890625, 'learning_rate': 9.252207965339111e-05, 'epoch': 0.54} 54%|█████▎ | 25800/48008 [1:28:29<1:27:10, 4.25it/s] 54%|█████▎ | 25801/48008 [1:28:29<1:27:36, 4.23it/s] 54%|█████▎ | 25802/48008 [1:28:29<1:23:04, 4.46it/s] 54%|█████▎ | 25803/48008 [1:28:30<1:19:33, 4.65it/s] 54%|█████▎ | 25804/48008 [1:28:30<1:16:39, 4.83it/s] 54%|█████▍ | 25805/48008 [1:28:30<1:15:58, 4.87it/s] 54%|█████▍ | 25806/48008 [1:28:30<1:14:02, 5.00it/s] 54%|█████▍ | 25807/48008 [1:28:30<1:12:36, 5.10it/s] 54%|█████▍ | 25808/48008 [1:28:31<1:17:02, 4.80it/s] 54%|█████▍ | 25809/48008 [1:28:31<1:15:30, 4.90it/s] 54%|█████▍ | 25810/48008 [1:28:31<1:14:10, 4.99it/s] 54%|█████▍ | 25811/48008 [1:28:31<1:13:08, 5.06it/s] 54%|█████▍ | 25812/48008 [1:28:31<1:12:57, 5.07it/s] 54%|█████▍ | 25813/48008 [1:28:31<1:11:25, 5.18it/s] 54%|█████▍ | 25814/48008 [1:28:32<1:13:21, 5.04it/s] 54%|█████▍ | 25815/48008 [1:28:32<1:12:39, 5.09it/s] 54%|█████▍ | 25816/48008 [1:28:32<1:33:36, 3.95it/s] 54%|█████▍ | 25817/48008 [1:28:32<1:26:22, 4.28it/s] 54%|█████▍ | 25818/48008 [1:28:33<1:21:40, 4.53it/s] 54%|█████▍ | 25819/48008 [1:28:33<1:18:28, 4.71it/s] 54%|█████▍ | 25820/48008 [1:28:33<1:21:13, 4.55it/s] 54%|█████▍ | 25821/48008 [1:28:33<1:16:58, 4.80it/s] 54%|█████▍ | 25822/48008 [1:28:33<1:14:26, 4.97it/s] 54%|█████▍ | 25823/48008 [1:28:34<1:13:23, 5.04it/s] 54%|█████▍ | 25824/48008 [1:28:34<1:13:30, 5.03it/s] 54%|█████▍ | 25825/48008 [1:28:34<1:12:33, 5.10it/s] 54%|█████▍ | 25826/48008 [1:28:34<1:11:35, 5.16it/s] 54%|█████▍ | 25827/48008 [1:28:34<1:10:56, 5.21it/s] 54%|█████▍ | 25828/48008 [1:28:35<1:10:48, 5.22it/s] 54%|█████▍ | 25829/48008 [1:28:35<1:10:41, 5.23it/s] 54%|█████▍ | 25830/48008 [1:28:35<1:09:51, 5.29it/s] 54%|█████▍ | 25831/48008 [1:28:35<1:11:09, 5.19it/s] 54%|█████▍ | 25832/48008 [1:28:35<1:11:24, 5.18it/s] 54%|█████▍ | 25833/48008 [1:28:36<1:12:24, 5.10it/s] 54%|█████▍ | 25834/48008 [1:28:36<1:11:04, 5.20it/s] 54%|█████▍ | 25835/48008 [1:28:36<1:15:53, 4.87it/s] 54%|█████▍ | 25836/48008 [1:28:36<1:14:22, 4.97it/s] 54%|█████▍ | 25837/48008 [1:28:36<1:12:40, 5.08it/s] 54%|█████▍ | 25838/48008 [1:28:37<1:13:17, 5.04it/s] 54%|█████▍ | 25839/48008 [1:28:37<1:13:41, 5.01it/s] 54%|█████▍ | 25840/48008 [1:28:37<1:13:27, 5.03it/s] 54%|█████▍ | 25841/48008 [1:28:37<1:17:41, 4.76it/s] 54%|█████▍ | 25842/48008 [1:28:37<1:14:51, 4.94it/s] 54%|█████▍ | 25843/48008 [1:28:38<1:13:02, 5.06it/s] 54%|█████▍ | 25844/48008 [1:28:38<1:12:58, 5.06it/s] 54%|█████▍ | 25845/48008 [1:28:38<1:12:03, 5.13it/s] 54%|█████▍ | 25846/48008 [1:28:38<1:11:46, 5.15it/s] 54%|█████▍ | 25847/48008 [1:28:38<1:13:07, 5.05it/s] 54%|█████▍ | 25848/48008 [1:28:39<1:12:30, 5.09it/s] 54%|█████▍ | 25849/48008 [1:28:39<1:11:43, 5.15it/s] 54%|█████▍ | 25850/48008 [1:28:39<1:11:32, 5.16it/s] {'loss': 0.3019, 'grad_norm': 0.234375, 'learning_rate': 9.231378103649391e-05, 'epoch': 0.54} 54%|█████▍ | 25850/48008 [1:28:39<1:11:32, 5.16it/s] 54%|█████▍ | 25851/48008 [1:28:39<1:12:16, 5.11it/s] 54%|█████▍ | 25852/48008 [1:28:39<1:13:03, 5.05it/s] 54%|█████▍ | 25853/48008 [1:28:40<1:12:07, 5.12it/s] 54%|█████▍ | 25854/48008 [1:28:40<1:11:47, 5.14it/s] 54%|█████▍ | 25855/48008 [1:28:40<1:10:57, 5.20it/s] 54%|█████▍ | 25856/48008 [1:28:40<1:10:51, 5.21it/s] 54%|█████▍ | 25857/48008 [1:28:40<1:11:58, 5.13it/s] 54%|█████▍ | 25858/48008 [1:28:40<1:11:18, 5.18it/s] 54%|█████▍ | 25859/48008 [1:28:41<1:11:31, 5.16it/s] 54%|█████▍ | 25860/48008 [1:28:41<1:11:14, 5.18it/s] 54%|█████▍ | 25861/48008 [1:28:41<1:16:05, 4.85it/s] 54%|█████▍ | 25862/48008 [1:28:41<1:14:15, 4.97it/s] 54%|█████▍ | 25863/48008 [1:28:41<1:12:57, 5.06it/s] 54%|█████▍ | 25864/48008 [1:28:42<1:13:17, 5.04it/s] 54%|█████▍ | 25865/48008 [1:28:42<1:33:59, 3.93it/s] 54%|█████▍ | 25866/48008 [1:28:42<1:26:11, 4.28it/s] 54%|█████▍ | 25867/48008 [1:28:42<1:21:58, 4.50it/s] 54%|█████▍ | 25868/48008 [1:28:43<1:19:53, 4.62it/s] 54%|█████▍ | 25869/48008 [1:28:43<1:15:54, 4.86it/s] 54%|█████▍ | 25870/48008 [1:28:43<1:14:07, 4.98it/s] 54%|█████▍ | 25871/48008 [1:28:43<1:12:53, 5.06it/s] 54%|█████▍ | 25872/48008 [1:28:43<1:12:00, 5.12it/s] 54%|█████▍ | 25873/48008 [1:28:44<1:11:36, 5.15it/s] 54%|█████▍ | 25874/48008 [1:28:44<1:11:05, 5.19it/s] 54%|█████▍ | 25875/48008 [1:28:44<1:10:34, 5.23it/s] 54%|█████▍ | 25876/48008 [1:28:44<1:10:02, 5.27it/s] 54%|█████▍ | 25877/48008 [1:28:44<1:10:02, 5.27it/s] 54%|█████▍ | 25878/48008 [1:28:45<1:12:21, 5.10it/s] 54%|█████▍ | 25879/48008 [1:28:45<1:13:09, 5.04it/s] 54%|█████▍ | 25880/48008 [1:28:45<1:12:00, 5.12it/s] 54%|█████▍ | 25881/48008 [1:28:45<1:11:35, 5.15it/s] 54%|█████▍ | 25882/48008 [1:28:45<1:11:13, 5.18it/s] 54%|█████▍ | 25883/48008 [1:28:46<1:11:06, 5.19it/s] 54%|█████▍ | 25884/48008 [1:28:46<1:11:31, 5.16it/s] 54%|█████▍ | 25885/48008 [1:28:46<1:11:18, 5.17it/s] 54%|█████▍ | 25886/48008 [1:28:46<1:12:07, 5.11it/s] 54%|█████▍ | 25887/48008 [1:28:46<1:11:21, 5.17it/s] 54%|█████▍ | 25888/48008 [1:28:47<1:10:58, 5.19it/s] 54%|█████▍ | 25889/48008 [1:28:47<1:12:11, 5.11it/s] 54%|█████▍ | 25890/48008 [1:28:47<1:11:08, 5.18it/s] 54%|█████▍ | 25891/48008 [1:28:48<2:37:37, 2.34it/s] 54%|█████▍ | 25892/48008 [1:28:48<2:11:21, 2.81it/s] 54%|█████▍ | 25893/48008 [1:28:48<1:51:54, 3.29it/s] 54%|█████▍ | 25894/48008 [1:28:48<1:40:45, 3.66it/s] 54%|█████▍ | 25895/48008 [1:28:49<1:32:17, 3.99it/s] 54%|█████▍ | 25896/48008 [1:28:49<1:26:04, 4.28it/s] 54%|█████▍ | 25897/48008 [1:28:50<2:37:04, 2.35it/s] 54%|█████▍ | 25898/48008 [1:28:50<2:10:29, 2.82it/s] 54%|█████▍ | 25899/48008 [1:28:50<1:54:29, 3.22it/s] 54%|█████▍ | 25900/48008 [1:28:50<1:42:08, 3.61it/s] {'loss': 0.3017, 'grad_norm': 0.27734375, 'learning_rate': 9.210548241959673e-05, 'epoch': 0.54} 54%|█████▍ | 25900/48008 [1:28:50<1:42:08, 3.61it/s] 54%|█████▍ | 25901/48008 [1:28:50<1:32:28, 3.98it/s] 54%|█████▍ | 25902/48008 [1:28:51<1:25:56, 4.29it/s] 54%|█████▍ | 25903/48008 [1:28:51<2:28:51, 2.47it/s] 54%|█████▍ | 25904/48008 [1:28:52<2:04:13, 2.97it/s] 54%|█████▍ | 25905/48008 [1:28:52<1:48:11, 3.41it/s] 54%|█████▍ | 25906/48008 [1:28:52<1:38:25, 3.74it/s] 54%|█████▍ | 25907/48008 [1:28:52<1:31:59, 4.00it/s] 54%|█████▍ | 25908/48008 [1:28:52<1:26:57, 4.24it/s] 54%|█████▍ | 25909/48008 [1:28:53<1:21:36, 4.51it/s] 54%|█████▍ | 25910/48008 [1:28:53<1:18:59, 4.66it/s] 54%|█████▍ | 25911/48008 [1:28:53<1:17:13, 4.77it/s] 54%|█████▍ | 25912/48008 [1:28:53<1:14:56, 4.91it/s] 54%|█████▍ | 25913/48008 [1:28:53<1:12:47, 5.06it/s] 54%|█████▍ | 25914/48008 [1:28:54<1:11:22, 5.16it/s] 54%|█████▍ | 25915/48008 [1:28:54<1:11:01, 5.18it/s] 54%|█████▍ | 25916/48008 [1:28:54<1:11:54, 5.12it/s] 54%|█████▍ | 25917/48008 [1:28:54<1:11:33, 5.15it/s] 54%|█████▍ | 25918/48008 [1:28:54<1:10:29, 5.22it/s] 54%|█████▍ | 25919/48008 [1:28:55<1:10:28, 5.22it/s] 54%|█████▍ | 25920/48008 [1:28:55<1:23:05, 4.43it/s] 54%|█████▍ | 25921/48008 [1:28:55<1:20:30, 4.57it/s] 54%|█████▍ | 25922/48008 [1:28:55<1:18:38, 4.68it/s] 54%|█████▍ | 25923/48008 [1:28:55<1:16:50, 4.79it/s] 54%|█████▍ | 25924/48008 [1:28:56<1:14:34, 4.94it/s] 54%|█████▍ | 25925/48008 [1:28:56<1:34:44, 3.88it/s] 54%|█████▍ | 25926/48008 [1:28:56<1:28:03, 4.18it/s] 54%|█████▍ | 25927/48008 [1:28:56<1:22:36, 4.45it/s] 54%|█████▍ | 25928/48008 [1:28:57<1:40:25, 3.66it/s] 54%|█████▍ | 25929/48008 [1:28:57<1:32:13, 3.99it/s] 54%|█████▍ | 25930/48008 [1:28:57<1:25:23, 4.31it/s] 54%|█████▍ | 25931/48008 [1:28:57<1:20:20, 4.58it/s] 54%|█████▍ | 25932/48008 [1:28:58<1:17:26, 4.75it/s] 54%|█████▍ | 25933/48008 [1:28:58<2:23:36, 2.56it/s] 54%|█████▍ | 25934/48008 [1:28:59<2:03:50, 2.97it/s] 54%|█████▍ | 25935/48008 [1:28:59<1:47:12, 3.43it/s] 54%|█████▍ | 25936/48008 [1:28:59<1:35:54, 3.84it/s] 54%|█████▍ | 25937/48008 [1:28:59<1:29:32, 4.11it/s] 54%|█████▍ | 25938/48008 [1:28:59<1:23:41, 4.40it/s] 54%|█████▍ | 25939/48008 [1:29:00<1:20:11, 4.59it/s] 54%|█████▍ | 25940/48008 [1:29:00<1:18:41, 4.67it/s] 54%|█████▍ | 25941/48008 [1:29:00<1:16:07, 4.83it/s] 54%|█████▍ | 25942/48008 [1:29:00<1:14:25, 4.94it/s] 54%|█████▍ | 25943/48008 [1:29:00<1:12:33, 5.07it/s] 54%|█████▍ | 25944/48008 [1:29:01<1:12:01, 5.11it/s] 54%|█████▍ | 25945/48008 [1:29:01<1:11:03, 5.17it/s] 54%|█████▍ | 25946/48008 [1:29:01<1:10:50, 5.19it/s] 54%|█████▍ | 25947/48008 [1:29:01<1:10:41, 5.20it/s] 54%|█████▍ | 25948/48008 [1:29:01<1:10:02, 5.25it/s] 54%|█████▍ | 25949/48008 [1:29:02<1:11:03, 5.17it/s] 54%|█████▍ | 25950/48008 [1:29:02<1:11:23, 5.15it/s] {'loss': 0.2997, 'grad_norm': 0.28515625, 'learning_rate': 9.189718380269955e-05, 'epoch': 0.54} 54%|█████▍ | 25950/48008 [1:29:02<1:11:23, 5.15it/s] 54%|█████▍ | 25951/48008 [1:29:02<1:11:36, 5.13it/s] 54%|█████▍ | 25952/48008 [1:29:02<1:10:43, 5.20it/s] 54%|█████▍ | 25953/48008 [1:29:02<1:10:20, 5.23it/s] 54%|█████▍ | 25954/48008 [1:29:02<1:12:07, 5.10it/s] 54%|█████▍ | 25955/48008 [1:29:03<1:11:25, 5.15it/s] 54%|█████▍ | 25956/48008 [1:29:03<1:12:04, 5.10it/s] 54%|█████▍ | 25957/48008 [1:29:03<1:12:44, 5.05it/s] 54%|█████▍ | 25958/48008 [1:29:03<1:11:43, 5.12it/s] 54%|█████▍ | 25959/48008 [1:29:03<1:10:50, 5.19it/s] 54%|█████▍ | 25960/48008 [1:29:04<1:10:32, 5.21it/s] 54%|█████▍ | 25961/48008 [1:29:04<1:09:56, 5.25it/s] 54%|█████▍ | 25962/48008 [1:29:04<1:09:23, 5.30it/s] 54%|█████▍ | 25963/48008 [1:29:04<1:09:36, 5.28it/s] 54%|█████▍ | 25964/48008 [1:29:04<1:09:49, 5.26it/s] 54%|█████▍ | 25965/48008 [1:29:05<1:11:00, 5.17it/s] 54%|█████▍ | 25966/48008 [1:29:05<1:10:37, 5.20it/s] 54%|█████▍ | 25967/48008 [1:29:05<1:10:01, 5.25it/s] 54%|█████▍ | 25968/48008 [1:29:05<1:09:57, 5.25it/s] 54%|█████▍ | 25969/48008 [1:29:05<1:11:24, 5.14it/s] 54%|█████▍ | 25970/48008 [1:29:06<1:11:55, 5.11it/s] 54%|█████▍ | 25971/48008 [1:29:06<1:10:57, 5.18it/s] 54%|█████▍ | 25972/48008 [1:29:06<1:11:57, 5.10it/s] 54%|█████▍ | 25973/48008 [1:29:06<1:11:03, 5.17it/s] 54%|█████▍ | 25974/48008 [1:29:06<1:11:04, 5.17it/s] 54%|█████▍ | 25975/48008 [1:29:07<1:32:17, 3.98it/s] 54%|█████▍ | 25976/48008 [1:29:07<1:25:44, 4.28it/s] 54%|█████▍ | 25977/48008 [1:29:07<1:23:11, 4.41it/s] 54%|█████▍ | 25978/48008 [1:29:07<1:20:02, 4.59it/s] 54%|█████▍ | 25979/48008 [1:29:08<1:16:41, 4.79it/s] 54%|█████▍ | 25980/48008 [1:29:08<1:16:54, 4.77it/s] 54%|█████▍ | 25981/48008 [1:29:08<1:15:32, 4.86it/s] 54%|█████▍ | 25982/48008 [1:29:08<1:13:43, 4.98it/s] 54%|█████▍ | 25983/48008 [1:29:08<1:12:39, 5.05it/s] 54%|█████▍ | 25984/48008 [1:29:09<1:12:52, 5.04it/s] 54%|█████▍ | 25985/48008 [1:29:09<1:12:18, 5.08it/s] 54%|█████▍ | 25986/48008 [1:29:09<1:12:24, 5.07it/s] 54%|█████▍ | 25987/48008 [1:29:09<1:12:23, 5.07it/s] 54%|█████▍ | 25988/48008 [1:29:09<1:11:50, 5.11it/s] 54%|█████▍ | 25989/48008 [1:29:09<1:11:10, 5.16it/s] 54%|█████▍ | 25990/48008 [1:29:10<1:12:13, 5.08it/s] 54%|█████▍ | 25991/48008 [1:29:10<1:11:23, 5.14it/s] 54%|█████▍ | 25992/48008 [1:29:10<1:09:49, 5.26it/s] 54%|█████▍ | 25993/48008 [1:29:10<1:09:58, 5.24it/s] 54%|█████▍ | 25994/48008 [1:29:10<1:11:12, 5.15it/s] 54%|█████▍ | 25995/48008 [1:29:11<1:10:56, 5.17it/s] 54%|█████▍ | 25996/48008 [1:29:11<1:23:28, 4.39it/s] 54%|█████▍ | 25997/48008 [1:29:11<1:20:31, 4.56it/s] 54%|█████▍ | 25998/48008 [1:29:11<1:18:10, 4.69it/s] 54%|█████▍ | 25999/48008 [1:29:12<1:15:37, 4.85it/s] 54%|█████▍ | 26000/48008 [1:29:12<1:13:51, 4.97it/s] {'loss': 0.3004, 'grad_norm': 0.357421875, 'learning_rate': 9.168888518580238e-05, 'epoch': 0.54} 54%|█████▍ | 26000/48008 [1:29:12<1:13:51, 4.97it/s] 54%|█████▍ | 26001/48008 [1:29:12<1:12:53, 5.03it/s] 54%|█████▍ | 26002/48008 [1:29:12<1:11:54, 5.10it/s] 54%|█████▍ | 26003/48008 [1:29:12<1:11:26, 5.13it/s] 54%|█████▍ | 26004/48008 [1:29:12<1:10:48, 5.18it/s] 54%|█████▍ | 26005/48008 [1:29:13<1:10:33, 5.20it/s] 54%|█████▍ | 26006/48008 [1:29:13<1:10:20, 5.21it/s] 54%|█████▍ | 26007/48008 [1:29:13<1:10:33, 5.20it/s] 54%|█████▍ | 26008/48008 [1:29:13<1:09:36, 5.27it/s] 54%|█████▍ | 26009/48008 [1:29:13<1:09:38, 5.26it/s] 54%|█████▍ | 26010/48008 [1:29:14<1:09:51, 5.25it/s] 54%|█████▍ | 26011/48008 [1:29:14<1:11:12, 5.15it/s] 54%|█████▍ | 26012/48008 [1:29:14<1:12:01, 5.09it/s] 54%|█████▍ | 26013/48008 [1:29:14<1:11:24, 5.13it/s] 54%|█████▍ | 26014/48008 [1:29:14<1:10:34, 5.19it/s] 54%|█████▍ | 26015/48008 [1:29:15<1:11:35, 5.12it/s] 54%|█████▍ | 26016/48008 [1:29:15<1:11:24, 5.13it/s] 54%|█████▍ | 26017/48008 [1:29:15<1:11:49, 5.10it/s] 54%|█████▍ | 26018/48008 [1:29:15<1:11:07, 5.15it/s] 54%|█████▍ | 26019/48008 [1:29:15<1:12:22, 5.06it/s] 54%|█████▍ | 26020/48008 [1:29:16<1:11:28, 5.13it/s] 54%|█████▍ | 26021/48008 [1:29:16<1:10:52, 5.17it/s] 54%|█████▍ | 26022/48008 [1:29:16<1:10:16, 5.21it/s] 54%|█████▍ | 26023/48008 [1:29:16<1:10:37, 5.19it/s] 54%|█████▍ | 26024/48008 [1:29:16<1:11:12, 5.15it/s] 54%|█████▍ | 26025/48008 [1:29:17<1:10:58, 5.16it/s] 54%|█████▍ | 26026/48008 [1:29:17<1:10:15, 5.21it/s] 54%|█████▍ | 26027/48008 [1:29:17<1:10:22, 5.21it/s] 54%|█████▍ | 26028/48008 [1:29:17<1:10:02, 5.23it/s] 54%|█████▍ | 26029/48008 [1:29:17<1:22:53, 4.42it/s] 54%|█████▍ | 26030/48008 [1:29:18<1:19:05, 4.63it/s] 54%|█████▍ | 26031/48008 [1:29:18<1:16:03, 4.82it/s] 54%|█████▍ | 26032/48008 [1:29:18<1:15:08, 4.87it/s] 54%|█████▍ | 26033/48008 [1:29:18<1:13:42, 4.97it/s] 54%|█████▍ | 26034/48008 [1:29:18<1:12:49, 5.03it/s] 54%|█████▍ | 26035/48008 [1:29:19<1:12:03, 5.08it/s] 54%|█████▍ | 26036/48008 [1:29:19<1:11:59, 5.09it/s] 54%|█████▍ | 26037/48008 [1:29:19<1:12:07, 5.08it/s] 54%|█████▍ | 26038/48008 [1:29:19<1:12:20, 5.06it/s] 54%|█████▍ | 26039/48008 [1:29:19<1:11:29, 5.12it/s] 54%|█████▍ | 26040/48008 [1:29:20<1:11:28, 5.12it/s] 54%|█████▍ | 26041/48008 [1:29:20<1:10:34, 5.19it/s] 54%|█████▍ | 26042/48008 [1:29:20<1:23:16, 4.40it/s] 54%|█████▍ | 26043/48008 [1:29:20<1:18:51, 4.64it/s] 54%|█████▍ | 26044/48008 [1:29:21<1:28:52, 4.12it/s] 54%|█████▍ | 26045/48008 [1:29:21<1:22:48, 4.42it/s] 54%|█████▍ | 26046/48008 [1:29:21<1:20:22, 4.55it/s] 54%|█████▍ | 26047/48008 [1:29:21<1:22:37, 4.43it/s] 54%|█████▍ | 26048/48008 [1:29:21<1:18:39, 4.65it/s] 54%|█████▍ | 26049/48008 [1:29:22<1:17:27, 4.73it/s] 54%|█████▍ | 26050/48008 [1:29:22<1:15:08, 4.87it/s] 54%|█████▍ | 26050/48008 [1:29:22<1:15:08, 4.87it/s]{'loss': 0.3011, 'grad_norm': 0.189453125, 'learning_rate': 9.14805865689052e-05, 'epoch': 0.54} 54%|█████▍ | 26051/48008 [1:29:22<1:15:11, 4.87it/s] 54%|█████▍ | 26052/48008 [1:29:22<1:13:36, 4.97it/s] 54%|█████▍ | 26053/48008 [1:29:22<1:12:30, 5.05it/s] 54%|█████▍ | 26054/48008 [1:29:23<1:16:37, 4.78it/s] 54%|█████▍ | 26055/48008 [1:29:23<1:15:17, 4.86it/s] 54%|█████▍ | 26056/48008 [1:29:23<1:18:39, 4.65it/s] 54%|█████▍ | 26057/48008 [1:29:23<1:15:34, 4.84it/s] 54%|█████▍ | 26058/48008 [1:29:23<1:14:39, 4.90it/s] 54%|█████▍ | 26059/48008 [1:29:24<1:13:18, 4.99it/s] 54%|█████▍ | 26060/48008 [1:29:24<1:11:43, 5.10it/s] 54%|█████▍ | 26061/48008 [1:29:24<1:12:39, 5.03it/s] 54%|█████▍ | 26062/48008 [1:29:24<1:11:56, 5.08it/s] 54%|█████▍ | 26063/48008 [1:29:24<1:11:01, 5.15it/s] 54%|█████▍ | 26064/48008 [1:29:25<1:10:29, 5.19it/s] 54%|█████▍ | 26065/48008 [1:29:25<1:10:18, 5.20it/s] 54%|█████▍ | 26066/48008 [1:29:25<1:10:14, 5.21it/s] 54%|█████▍ | 26067/48008 [1:29:25<1:10:42, 5.17it/s] 54%|█████▍ | 26068/48008 [1:29:25<1:10:41, 5.17it/s] 54%|█████▍ | 26069/48008 [1:29:26<1:10:59, 5.15it/s] 54%|█████▍ | 26070/48008 [1:29:26<1:32:06, 3.97it/s] 54%|█████▍ | 26071/48008 [1:29:26<1:27:28, 4.18it/s] 54%|█████▍ | 26072/48008 [1:29:26<1:22:05, 4.45it/s] 54%|█████▍ | 26073/48008 [1:29:27<1:18:23, 4.66it/s] 54%|█████▍ | 26074/48008 [1:29:27<1:15:23, 4.85it/s] 54%|█████▍ | 26075/48008 [1:29:27<1:14:44, 4.89it/s] 54%|█████▍ | 26076/48008 [1:29:27<1:13:58, 4.94it/s] 54%|█████▍ | 26077/48008 [1:29:27<1:13:12, 4.99it/s] 54%|█████▍ | 26078/48008 [1:29:27<1:12:56, 5.01it/s] 54%|█████▍ | 26079/48008 [1:29:28<1:11:24, 5.12it/s] 54%|█████▍ | 26080/48008 [1:29:28<1:12:25, 5.05it/s] 54%|█████▍ | 26081/48008 [1:29:28<1:11:34, 5.11it/s] 54%|█████▍ | 26082/48008 [1:29:28<1:11:05, 5.14it/s] 54%|█████▍ | 26083/48008 [1:29:28<1:12:48, 5.02it/s] 54%|█████▍ | 26084/48008 [1:29:29<1:17:03, 4.74it/s] 54%|█████▍ | 26085/48008 [1:29:29<1:15:58, 4.81it/s] 54%|█████▍ | 26086/48008 [1:29:29<1:14:42, 4.89it/s] 54%|█████▍ | 26087/48008 [1:29:29<1:14:04, 4.93it/s] 54%|█████▍ | 26088/48008 [1:29:29<1:12:33, 5.03it/s] 54%|█████▍ | 26089/48008 [1:29:30<1:11:06, 5.14it/s] 54%|█████▍ | 26090/48008 [1:29:30<1:10:07, 5.21it/s] 54%|█████▍ | 26091/48008 [1:29:30<1:22:37, 4.42it/s] 54%|█████▍ | 26092/48008 [1:29:30<1:18:31, 4.65it/s] 54%|█████▍ | 26093/48008 [1:29:31<1:17:05, 4.74it/s] 54%|█████▍ | 26094/48008 [1:29:31<1:14:34, 4.90it/s] 54%|█████▍ | 26095/48008 [1:29:31<1:14:04, 4.93it/s] 54%|█████▍ | 26096/48008 [1:29:31<1:13:25, 4.97it/s] 54%|█████▍ | 26097/48008 [1:29:31<1:12:31, 5.04it/s] 54%|█████▍ | 26098/48008 [1:29:32<1:11:37, 5.10it/s] 54%|█████▍ | 26099/48008 [1:29:32<1:10:57, 5.15it/s] 54%|█████▍ | 26100/48008 [1:29:32<1:10:16, 5.20it/s] {'loss': 0.3004, 'grad_norm': 0.21484375, 'learning_rate': 9.127228795200801e-05, 'epoch': 0.54} 54%|█████▍ | 26100/48008 [1:29:32<1:10:16, 5.20it/s] 54%|█████▍ | 26101/48008 [1:29:32<1:09:53, 5.22it/s] 54%|█████▍ | 26102/48008 [1:29:32<1:09:12, 5.28it/s] 54%|█████▍ | 26103/48008 [1:29:32<1:10:48, 5.16it/s] 54%|█████▍ | 26104/48008 [1:29:33<1:09:57, 5.22it/s] 54%|█████▍ | 26105/48008 [1:29:33<1:09:53, 5.22it/s] 54%|█████▍ | 26106/48008 [1:29:33<1:09:53, 5.22it/s] 54%|█████▍ | 26107/48008 [1:29:33<1:09:17, 5.27it/s] 54%|█████▍ | 26108/48008 [1:29:33<1:09:00, 5.29it/s] 54%|█████▍ | 26109/48008 [1:29:34<1:09:59, 5.21it/s] 54%|█████▍ | 26110/48008 [1:29:34<1:11:33, 5.10it/s] 54%|█████▍ | 26111/48008 [1:29:34<1:10:41, 5.16it/s] 54%|█████▍ | 26112/48008 [1:29:34<1:09:55, 5.22it/s] 54%|█████▍ | 26113/48008 [1:29:34<1:09:20, 5.26it/s] 54%|█████▍ | 26114/48008 [1:29:35<1:09:30, 5.25it/s] 54%|█████▍ | 26115/48008 [1:29:35<1:09:25, 5.26it/s] 54%|█████▍ | 26116/48008 [1:29:35<1:30:48, 4.02it/s] 54%|█████▍ | 26117/48008 [1:29:35<1:25:02, 4.29it/s] 54%|█████▍ | 26118/48008 [1:29:36<1:22:29, 4.42it/s] 54%|█████▍ | 26119/48008 [1:29:36<1:18:49, 4.63it/s] 54%|█████▍ | 26120/48008 [1:29:36<1:37:28, 3.74it/s] 54%|█████▍ | 26121/48008 [1:29:36<1:29:06, 4.09it/s] 54%|█████▍ | 26122/48008 [1:29:37<1:22:39, 4.41it/s] 54%|█████▍ | 26123/48008 [1:29:37<1:18:17, 4.66it/s] 54%|█████▍ | 26124/48008 [1:29:37<1:15:46, 4.81it/s] 54%|█████▍ | 26125/48008 [1:29:37<1:15:07, 4.85it/s] 54%|█████▍ | 26126/48008 [1:29:37<1:14:23, 4.90it/s] 54%|█████▍ | 26127/48008 [1:29:37<1:11:49, 5.08it/s] 54%|█████▍ | 26128/48008 [1:29:38<1:11:03, 5.13it/s] 54%|█████▍ | 26129/48008 [1:29:38<1:15:51, 4.81it/s] 54%|█████▍ | 26130/48008 [1:29:38<1:14:08, 4.92it/s] 54%|█████▍ | 26131/48008 [1:29:38<1:12:49, 5.01it/s] 54%|█████▍ | 26132/48008 [1:29:38<1:12:57, 5.00it/s] 54%|█████▍ | 26133/48008 [1:29:39<1:11:55, 5.07it/s] 54%|█████▍ | 26134/48008 [1:29:39<1:12:19, 5.04it/s] 54%|█████▍ | 26135/48008 [1:29:39<1:11:47, 5.08it/s] 54%|█████▍ | 26136/48008 [1:29:39<1:11:34, 5.09it/s] 54%|█████▍ | 26137/48008 [1:29:39<1:11:36, 5.09it/s] 54%|█████▍ | 26138/48008 [1:29:40<1:12:16, 5.04it/s] 54%|█████▍ | 26139/48008 [1:29:40<1:11:32, 5.09it/s] 54%|█████▍ | 26140/48008 [1:29:40<1:10:41, 5.16it/s] 54%|█████▍ | 26141/48008 [1:29:40<1:12:29, 5.03it/s] 54%|█████▍ | 26142/48008 [1:29:40<1:12:12, 5.05it/s] 54%|█████▍ | 26143/48008 [1:29:41<1:12:18, 5.04it/s] 54%|█████▍ | 26144/48008 [1:29:41<1:11:14, 5.12it/s] 54%|█████▍ | 26145/48008 [1:29:41<1:11:34, 5.09it/s] 54%|█████▍ | 26146/48008 [1:29:41<1:10:59, 5.13it/s] 54%|█████▍ | 26147/48008 [1:29:41<1:15:35, 4.82it/s] 54%|█████▍ | 26148/48008 [1:29:42<1:13:54, 4.93it/s] 54%|█████▍ | 26149/48008 [1:29:42<1:12:16, 5.04it/s] 54%|█████▍ | 26150/48008 [1:29:42<1:12:08, 5.05it/s] {'loss': 0.3009, 'grad_norm': 0.333984375, 'learning_rate': 9.106398933511082e-05, 'epoch': 0.54} 54%|█████▍ | 26150/48008 [1:29:42<1:12:08, 5.05it/s] 54%|█████▍ | 26151/48008 [1:29:42<1:11:40, 5.08it/s] 54%|█████▍ | 26152/48008 [1:29:42<1:10:40, 5.15it/s] 54%|█████▍ | 26153/48008 [1:29:43<1:10:08, 5.19it/s] 54%|█████▍ | 26154/48008 [1:29:43<1:14:54, 4.86it/s] 54%|█████▍ | 26155/48008 [1:29:43<1:12:58, 4.99it/s] 54%|█████▍ | 26156/48008 [1:29:43<1:12:35, 5.02it/s] 54%|█████▍ | 26157/48008 [1:29:43<1:11:44, 5.08it/s] 54%|█████▍ | 26158/48008 [1:29:44<1:10:39, 5.15it/s] 54%|█████▍ | 26159/48008 [1:29:44<1:10:18, 5.18it/s] 54%|█████▍ | 26160/48008 [1:29:44<1:09:33, 5.23it/s] 54%|█████▍ | 26161/48008 [1:29:44<1:09:19, 5.25it/s] 54%|█████▍ | 26162/48008 [1:29:44<1:09:28, 5.24it/s] 54%|█████▍ | 26163/48008 [1:29:45<1:09:21, 5.25it/s] 54%|█████▍ | 26164/48008 [1:29:45<1:09:56, 5.21it/s] 55%|█████▍ | 26165/48008 [1:29:45<1:09:56, 5.20it/s] 55%|█████▍ | 26166/48008 [1:29:45<1:09:43, 5.22it/s] 55%|█████▍ | 26167/48008 [1:29:45<1:09:22, 5.25it/s] 55%|█████▍ | 26168/48008 [1:29:46<1:09:27, 5.24it/s] 55%|█████▍ | 26169/48008 [1:29:46<1:11:20, 5.10it/s] 55%|█████▍ | 26170/48008 [1:29:46<1:10:39, 5.15it/s] 55%|█████▍ | 26171/48008 [1:29:46<1:10:30, 5.16it/s] 55%|█████▍ | 26172/48008 [1:29:46<1:10:17, 5.18it/s] 55%|█████▍ | 26173/48008 [1:29:47<1:10:14, 5.18it/s] 55%|█████▍ | 26174/48008 [1:29:47<1:08:54, 5.28it/s] 55%|█████▍ | 26175/48008 [1:29:47<1:10:19, 5.17it/s] 55%|█████▍ | 26176/48008 [1:29:47<1:10:06, 5.19it/s] 55%|█████▍ | 26177/48008 [1:29:47<1:10:32, 5.16it/s] 55%|█████▍ | 26178/48008 [1:29:47<1:10:43, 5.14it/s] 55%|█████▍ | 26179/48008 [1:29:48<1:12:24, 5.02it/s] 55%|█████▍ | 26180/48008 [1:29:48<1:11:47, 5.07it/s] 55%|█████▍ | 26181/48008 [1:29:48<1:11:33, 5.08it/s] 55%|█████▍ | 26182/48008 [1:29:48<1:10:56, 5.13it/s] 55%|█████▍ | 26183/48008 [1:29:48<1:11:50, 5.06it/s] 55%|█████▍ | 26184/48008 [1:29:49<1:10:52, 5.13it/s] 55%|█████▍ | 26185/48008 [1:29:49<1:10:29, 5.16it/s] 55%|█████▍ | 26186/48008 [1:29:49<1:11:28, 5.09it/s] 55%|█████▍ | 26187/48008 [1:29:49<1:10:59, 5.12it/s] 55%|█████▍ | 26188/48008 [1:29:49<1:10:55, 5.13it/s] 55%|█████▍ | 26189/48008 [1:29:50<1:10:32, 5.15it/s] 55%|█████▍ | 26190/48008 [1:29:50<1:09:46, 5.21it/s] 55%|█████▍ | 26191/48008 [1:29:50<1:09:47, 5.21it/s] 55%|█████▍ | 26192/48008 [1:29:50<1:11:34, 5.08it/s] 55%|█████▍ | 26193/48008 [1:29:50<1:10:52, 5.13it/s] 55%|█████▍ | 26194/48008 [1:29:51<1:11:45, 5.07it/s] 55%|█████▍ | 26195/48008 [1:29:51<1:11:33, 5.08it/s] 55%|█████▍ | 26196/48008 [1:29:51<1:11:52, 5.06it/s] 55%|█████▍ | 26197/48008 [1:29:51<1:10:43, 5.14it/s] 55%|█████▍ | 26198/48008 [1:29:51<1:10:15, 5.17it/s] 55%|█████▍ | 26199/48008 [1:29:52<1:10:37, 5.15it/s] 55%|█████▍ | 26200/48008 [1:29:52<1:10:10, 5.18it/s] {'loss': 0.302, 'grad_norm': 0.154296875, 'learning_rate': 9.085569071821363e-05, 'epoch': 0.55} 55%|█████▍ | 26200/48008 [1:29:52<1:10:10, 5.18it/s] 55%|█████▍ | 26201/48008 [1:29:52<1:11:14, 5.10it/s] 55%|█████▍ | 26202/48008 [1:29:52<1:10:42, 5.14it/s] 55%|█████▍ | 26203/48008 [1:29:52<1:10:55, 5.12it/s] 55%|█████▍ | 26204/48008 [1:29:53<1:10:22, 5.16it/s] 55%|█████▍ | 26205/48008 [1:29:53<1:10:07, 5.18it/s] 55%|█████▍ | 26206/48008 [1:29:53<1:09:47, 5.21it/s] 55%|█████▍ | 26207/48008 [1:29:53<1:09:49, 5.20it/s] 55%|█████▍ | 26208/48008 [1:29:53<1:09:48, 5.20it/s] 55%|█████▍ | 26209/48008 [1:29:54<1:09:42, 5.21it/s] 55%|█████▍ | 26210/48008 [1:29:54<1:09:45, 5.21it/s] 55%|█████▍ | 26211/48008 [1:29:54<1:10:51, 5.13it/s] 55%|█████▍ | 26212/48008 [1:29:54<1:09:53, 5.20it/s] 55%|█████▍ | 26213/48008 [1:29:54<1:09:50, 5.20it/s] 55%|█████▍ | 26214/48008 [1:29:55<1:22:16, 4.41it/s] 55%|█████▍ | 26215/48008 [1:29:55<1:18:44, 4.61it/s] 55%|█████▍ | 26216/48008 [1:29:55<1:15:37, 4.80it/s] 55%|█████▍ | 26217/48008 [1:29:55<1:13:35, 4.93it/s] 55%|█████▍ | 26218/48008 [1:29:55<1:12:32, 5.01it/s] 55%|█████▍ | 26219/48008 [1:29:56<1:13:13, 4.96it/s] 55%|█████▍ | 26220/48008 [1:29:56<1:12:06, 5.04it/s] 55%|█████▍ | 26221/48008 [1:29:56<1:11:21, 5.09it/s] 55%|█████▍ | 26222/48008 [1:29:56<1:10:39, 5.14it/s] 55%|█████▍ | 26223/48008 [1:29:56<1:09:41, 5.21it/s] 55%|█████▍ | 26224/48008 [1:29:56<1:09:08, 5.25it/s] 55%|█████▍ | 26225/48008 [1:29:57<1:09:02, 5.26it/s] 55%|█████▍ | 26226/48008 [1:29:57<1:09:14, 5.24it/s] 55%|█████▍ | 26227/48008 [1:29:57<1:09:20, 5.24it/s] 55%|█████▍ | 26228/48008 [1:29:57<1:09:02, 5.26it/s] 55%|█████▍ | 26229/48008 [1:29:57<1:09:55, 5.19it/s] 55%|█████▍ | 26230/48008 [1:29:58<1:10:07, 5.18it/s] 55%|█████▍ | 26231/48008 [1:29:58<1:10:36, 5.14it/s] 55%|█████▍ | 26232/48008 [1:29:58<1:09:58, 5.19it/s] 55%|█████▍ | 26233/48008 [1:29:58<1:10:23, 5.16it/s] 55%|█████▍ | 26234/48008 [1:29:58<1:09:50, 5.20it/s] 55%|█████▍ | 26235/48008 [1:29:59<1:09:34, 5.22it/s] 55%|█████▍ | 26236/48008 [1:29:59<1:09:17, 5.24it/s] 55%|█████▍ | 26237/48008 [1:29:59<1:08:51, 5.27it/s] 55%|█████▍ | 26238/48008 [1:29:59<1:09:06, 5.25it/s] 55%|█████▍ | 26239/48008 [1:29:59<1:09:13, 5.24it/s] 55%|█████▍ | 26240/48008 [1:30:00<1:11:20, 5.09it/s] 55%|█████▍ | 26241/48008 [1:30:00<1:10:33, 5.14it/s] 55%|█████▍ | 26242/48008 [1:30:00<1:11:26, 5.08it/s] 55%|█████▍ | 26243/48008 [1:30:00<1:10:43, 5.13it/s] 55%|█████▍ | 26244/48008 [1:30:00<1:11:28, 5.07it/s] 55%|█████▍ | 26245/48008 [1:30:01<1:15:41, 4.79it/s] 55%|█████▍ | 26246/48008 [1:30:01<1:13:35, 4.93it/s] 55%|█████▍ | 26247/48008 [1:30:01<1:12:58, 4.97it/s] 55%|█████▍ | 26248/48008 [1:30:01<1:12:53, 4.98it/s] 55%|█████▍ | 26249/48008 [1:30:01<1:12:41, 4.99it/s] 55%|█████▍ | 26250/48008 [1:30:02<1:16:39, 4.73it/s] 55%|█████▍ | 26250/48008 [1:30:02<1:16:39, 4.73it/s]{'loss': 0.3019, 'grad_norm': 0.2265625, 'learning_rate': 9.064739210131645e-05, 'epoch': 0.55} 55%|█████▍ | 26251/48008 [1:30:02<1:14:26, 4.87it/s] 55%|█████▍ | 26252/48008 [1:30:02<1:17:46, 4.66it/s] 55%|█████▍ | 26253/48008 [1:30:02<1:15:12, 4.82it/s] 55%|█████▍ | 26254/48008 [1:30:02<1:13:19, 4.94it/s] 55%|█████▍ | 26255/48008 [1:30:03<1:13:20, 4.94it/s] 55%|█████▍ | 26256/48008 [1:30:03<1:12:03, 5.03it/s] 55%|█████▍ | 26257/48008 [1:30:03<1:10:38, 5.13it/s] 55%|█████▍ | 26258/48008 [1:30:03<1:09:50, 5.19it/s] 55%|█████▍ | 26259/48008 [1:30:03<1:09:23, 5.22it/s] 55%|█████▍ | 26260/48008 [1:30:04<1:30:32, 4.00it/s] 55%|█████▍ | 26261/48008 [1:30:04<1:24:04, 4.31it/s] 55%|█████▍ | 26262/48008 [1:30:04<1:20:20, 4.51it/s] 55%|█████▍ | 26263/48008 [1:30:04<1:17:08, 4.70it/s] 55%|█████▍ | 26264/48008 [1:30:05<1:14:40, 4.85it/s] 55%|█████▍ | 26265/48008 [1:30:05<1:12:44, 4.98it/s] 55%|█████▍ | 26266/48008 [1:30:05<1:11:42, 5.05it/s] 55%|█████▍ | 26267/48008 [1:30:05<1:15:58, 4.77it/s] 55%|█████▍ | 26268/48008 [1:30:05<1:13:43, 4.92it/s] 55%|█████▍ | 26269/48008 [1:30:06<1:13:39, 4.92it/s] 55%|█████▍ | 26270/48008 [1:30:06<1:11:59, 5.03it/s] 55%|█████▍ | 26271/48008 [1:30:06<1:12:25, 5.00it/s] 55%|█████▍ | 26272/48008 [1:30:06<1:11:01, 5.10it/s] 55%|█████▍ | 26273/48008 [1:30:06<1:10:19, 5.15it/s] 55%|█████▍ | 26274/48008 [1:30:07<1:09:57, 5.18it/s] 55%|█████▍ | 26275/48008 [1:30:07<1:09:43, 5.19it/s] 55%|█████▍ | 26276/48008 [1:30:07<1:09:13, 5.23it/s] 55%|█████▍ | 26277/48008 [1:30:07<1:10:25, 5.14it/s] 55%|█████▍ | 26278/48008 [1:30:07<1:10:08, 5.16it/s] 55%|█████▍ | 26279/48008 [1:30:07<1:10:19, 5.15it/s] 55%|█████▍ | 26280/48008 [1:30:08<1:10:38, 5.13it/s] 55%|█████▍ | 26281/48008 [1:30:08<1:09:41, 5.20it/s] 55%|█████▍ | 26282/48008 [1:30:08<1:09:18, 5.22it/s] 55%|█████▍ | 26283/48008 [1:30:08<1:08:49, 5.26it/s] 55%|█████▍ | 26284/48008 [1:30:08<1:09:05, 5.24it/s] 55%|█████▍ | 26285/48008 [1:30:09<1:09:46, 5.19it/s] 55%|█████▍ | 26286/48008 [1:30:09<1:10:51, 5.11it/s] 55%|█████▍ | 26287/48008 [1:30:09<1:22:48, 4.37it/s] 55%|█████▍ | 26288/48008 [1:30:09<1:18:45, 4.60it/s] 55%|█████▍ | 26289/48008 [1:30:10<1:17:11, 4.69it/s] 55%|█████▍ | 26290/48008 [1:30:10<1:14:43, 4.84it/s] 55%|█████▍ | 26291/48008 [1:30:10<1:13:03, 4.95it/s] 55%|█████▍ | 26292/48008 [1:30:10<1:11:55, 5.03it/s] 55%|█████▍ | 26293/48008 [1:30:10<1:10:48, 5.11it/s] 55%|█████▍ | 26294/48008 [1:30:11<1:15:20, 4.80it/s] 55%|█████▍ | 26295/48008 [1:30:11<1:13:29, 4.92it/s] 55%|█████▍ | 26296/48008 [1:30:11<1:12:21, 5.00it/s] 55%|█████▍ | 26297/48008 [1:30:11<1:12:24, 5.00it/s] 55%|█████▍ | 26298/48008 [1:30:11<1:11:55, 5.03it/s] 55%|█████▍ | 26299/48008 [1:30:12<1:10:45, 5.11it/s] 55%|█████▍ | 26300/48008 [1:30:12<1:09:45, 5.19it/s] {'loss': 0.3023, 'grad_norm': 0.16015625, 'learning_rate': 9.043909348441927e-05, 'epoch': 0.55} 55%|█████▍ | 26300/48008 [1:30:12<1:09:45, 5.19it/s] 55%|█████▍ | 26301/48008 [1:30:12<1:09:28, 5.21it/s] 55%|█████▍ | 26302/48008 [1:30:12<1:09:55, 5.17it/s] 55%|█████▍ | 26303/48008 [1:30:12<1:10:21, 5.14it/s] 55%|█████▍ | 26304/48008 [1:30:12<1:09:58, 5.17it/s] 55%|█████▍ | 26305/48008 [1:30:13<1:10:22, 5.14it/s] 55%|█████▍ | 26306/48008 [1:30:13<1:09:46, 5.18it/s] 55%|█████▍ | 26307/48008 [1:30:13<1:10:20, 5.14it/s] 55%|█████▍ | 26308/48008 [1:30:13<1:12:19, 5.00it/s] 55%|█████▍ | 26309/48008 [1:30:13<1:11:04, 5.09it/s] 55%|█████▍ | 26310/48008 [1:30:14<1:10:04, 5.16it/s] 55%|█████▍ | 26311/48008 [1:30:14<1:09:38, 5.19it/s] 55%|█████▍ | 26312/48008 [1:30:14<1:14:31, 4.85it/s] 55%|█████▍ | 26313/48008 [1:30:14<1:17:49, 4.65it/s] 55%|█████▍ | 26314/48008 [1:30:15<1:16:24, 4.73it/s] 55%|█████▍ | 26315/48008 [1:30:15<1:14:32, 4.85it/s] 55%|█████▍ | 26316/48008 [1:30:15<1:14:04, 4.88it/s] 55%|█████▍ | 26317/48008 [1:30:15<1:13:44, 4.90it/s] 55%|█████▍ | 26318/48008 [1:30:15<1:13:15, 4.94it/s] 55%|█████▍ | 26319/48008 [1:30:15<1:12:02, 5.02it/s] 55%|█████▍ | 26320/48008 [1:30:16<1:11:21, 5.07it/s] 55%|█████▍ | 26321/48008 [1:30:16<1:10:59, 5.09it/s] 55%|█████▍ | 26322/48008 [1:30:16<1:11:16, 5.07it/s] 55%|█████▍ | 26323/48008 [1:30:16<1:10:33, 5.12it/s] 55%|█████▍ | 26324/48008 [1:30:16<1:10:52, 5.10it/s] 55%|█████▍ | 26325/48008 [1:30:17<1:10:31, 5.12it/s] 55%|█████▍ | 26326/48008 [1:30:17<1:10:06, 5.15it/s] 55%|█████▍ | 26327/48008 [1:30:17<1:09:43, 5.18it/s] 55%|█████▍ | 26328/48008 [1:30:17<1:09:57, 5.16it/s] 55%|█████▍ | 26329/48008 [1:30:17<1:09:14, 5.22it/s] 55%|█████▍ | 26330/48008 [1:30:18<1:08:50, 5.25it/s] 55%|█████▍ | 26331/48008 [1:30:18<1:09:29, 5.20it/s] 55%|█████▍ | 26332/48008 [1:30:18<1:09:02, 5.23it/s] 55%|█████▍ | 26333/48008 [1:30:18<1:09:11, 5.22it/s] 55%|█████▍ | 26334/48008 [1:30:18<1:09:12, 5.22it/s] 55%|█████▍ | 26335/48008 [1:30:19<1:30:40, 3.98it/s] 55%|█████▍ | 26336/48008 [1:30:19<1:24:35, 4.27it/s] 55%|█████▍ | 26337/48008 [1:30:19<1:19:58, 4.52it/s] 55%|█████▍ | 26338/48008 [1:30:19<1:16:56, 4.69it/s] 55%|█████▍ | 26339/48008 [1:30:20<1:14:00, 4.88it/s] 55%|█████▍ | 26340/48008 [1:30:20<1:12:39, 4.97it/s] 55%|█████▍ | 26341/48008 [1:30:20<1:16:31, 4.72it/s] 55%|█████▍ | 26342/48008 [1:30:20<1:14:24, 4.85it/s] 55%|█████▍ | 26343/48008 [1:30:20<1:13:09, 4.94it/s] 55%|█████▍ | 26344/48008 [1:30:21<1:11:43, 5.03it/s] 55%|█████▍ | 26345/48008 [1:30:21<1:11:04, 5.08it/s] 55%|█████▍ | 26346/48008 [1:30:21<1:10:35, 5.11it/s] 55%|█████▍ | 26347/48008 [1:30:21<1:09:44, 5.18it/s] 55%|█████▍ | 26348/48008 [1:30:21<1:09:05, 5.22it/s] 55%|█████▍ | 26349/48008 [1:30:21<1:09:10, 5.22it/s] 55%|█████▍ | 26350/48008 [1:30:22<1:09:01, 5.23it/s] {'loss': 0.3016, 'grad_norm': 0.142578125, 'learning_rate': 9.023079486752209e-05, 'epoch': 0.55} 55%|█████▍ | 26350/48008 [1:30:22<1:09:01, 5.23it/s] 55%|█████▍ | 26351/48008 [1:30:22<1:09:20, 5.21it/s] 55%|█████▍ | 26352/48008 [1:30:22<1:09:13, 5.21it/s] 55%|█████▍ | 26353/48008 [1:30:22<1:10:03, 5.15it/s] 55%|█████▍ | 26354/48008 [1:30:22<1:10:10, 5.14it/s] 55%|█████▍ | 26355/48008 [1:30:23<1:10:07, 5.15it/s] 55%|█████▍ | 26356/48008 [1:30:23<1:11:02, 5.08it/s] 55%|█████▍ | 26357/48008 [1:30:23<1:10:26, 5.12it/s] 55%|█████▍ | 26358/48008 [1:30:23<1:10:30, 5.12it/s] 55%|█████▍ | 26359/48008 [1:30:23<1:09:55, 5.16it/s] 55%|█████▍ | 26360/48008 [1:30:24<1:10:48, 5.10it/s] 55%|█████▍ | 26361/48008 [1:30:24<1:10:17, 5.13it/s] 55%|█████▍ | 26362/48008 [1:30:24<1:09:55, 5.16it/s] 55%|█████▍ | 26363/48008 [1:30:24<1:09:45, 5.17it/s] 55%|█████▍ | 26364/48008 [1:30:24<1:09:03, 5.22it/s] 55%|█████▍ | 26365/48008 [1:30:25<1:09:11, 5.21it/s] 55%|█████▍ | 26366/48008 [1:30:25<1:08:33, 5.26it/s] 55%|█████▍ | 26367/48008 [1:30:25<1:10:18, 5.13it/s] 55%|█████▍ | 26368/48008 [1:30:25<1:12:02, 5.01it/s] 55%|█████▍ | 26369/48008 [1:30:25<1:10:46, 5.10it/s] 55%|█████▍ | 26370/48008 [1:30:26<1:22:40, 4.36it/s] 55%|█████▍ | 26371/48008 [1:30:26<1:18:34, 4.59it/s] 55%|█████▍ | 26372/48008 [1:30:26<1:15:56, 4.75it/s] 55%|█████▍ | 26373/48008 [1:30:26<1:13:40, 4.89it/s] 55%|█████▍ | 26374/48008 [1:30:26<1:12:13, 4.99it/s] 55%|█████▍ | 26375/48008 [1:30:27<1:10:48, 5.09it/s] 55%|█████▍ | 26376/48008 [1:30:27<1:10:01, 5.15it/s] 55%|█████▍ | 26377/48008 [1:30:27<1:09:28, 5.19it/s] 55%|█████▍ | 26378/48008 [1:30:27<1:21:36, 4.42it/s] 55%|█████▍ | 26379/48008 [1:30:28<1:18:12, 4.61it/s] 55%|█████▍ | 26380/48008 [1:30:28<1:15:48, 4.76it/s] 55%|█████▍ | 26381/48008 [1:30:28<1:13:42, 4.89it/s] 55%|█████▍ | 26382/48008 [1:30:28<1:17:13, 4.67it/s] 55%|█████▍ | 26383/48008 [1:30:28<1:14:38, 4.83it/s] 55%|█████▍ | 26384/48008 [1:30:29<1:12:51, 4.95it/s] 55%|█████▍ | 26385/48008 [1:30:29<1:11:39, 5.03it/s] 55%|█████▍ | 26386/48008 [1:30:29<1:10:38, 5.10it/s] 55%|█████▍ | 26387/48008 [1:30:30<2:25:07, 2.48it/s] 55%|█████▍ | 26388/48008 [1:30:30<2:01:58, 2.95it/s] 55%|█████▍ | 26389/48008 [1:30:30<1:51:10, 3.24it/s] 55%|█████▍ | 26390/48008 [1:30:30<1:39:47, 3.61it/s] 55%|█████▍ | 26391/48008 [1:30:31<1:35:22, 3.78it/s] 55%|█████▍ | 26392/48008 [1:30:31<1:27:56, 4.10it/s] 55%|█████▍ | 26393/48008 [1:30:31<1:22:16, 4.38it/s] 55%|█████▍ | 26394/48008 [1:30:31<1:17:13, 4.66it/s] 55%|█████▍ | 26395/48008 [1:30:31<1:16:01, 4.74it/s] 55%|█████▍ | 26396/48008 [1:30:32<1:15:05, 4.80it/s] 55%|█████▍ | 26397/48008 [1:30:32<1:13:54, 4.87it/s] 55%|█████▍ | 26398/48008 [1:30:32<1:12:37, 4.96it/s] 55%|█████▍ | 26399/48008 [1:30:32<1:12:39, 4.96it/s] 55%|█████▍ | 26400/48008 [1:30:32<1:11:27, 5.04it/s] 55%|█████▍ | 26400/48008 [1:30:32<1:11:27, 5.04it/s]{'loss': 0.2981, 'grad_norm': 0.283203125, 'learning_rate': 9.00224962506249e-05, 'epoch': 0.55} 55%|█████▍ | 26401/48008 [1:30:33<1:09:30, 5.18it/s] 55%|█████▍ | 26402/48008 [1:30:33<1:10:49, 5.08it/s] 55%|█████▍ | 26403/48008 [1:30:33<1:10:17, 5.12it/s] 55%|█████▍ | 26404/48008 [1:30:33<1:11:51, 5.01it/s] 55%|█████▌ | 26405/48008 [1:30:33<1:10:52, 5.08it/s] 55%|█████▌ | 26406/48008 [1:30:34<1:10:38, 5.10it/s] 55%|█████▌ | 26407/48008 [1:30:34<1:10:14, 5.13it/s] 55%|█████▌ | 26408/48008 [1:30:34<1:09:44, 5.16it/s] 55%|█████▌ | 26409/48008 [1:30:34<1:08:57, 5.22it/s] 55%|█████▌ | 26410/48008 [1:30:34<1:08:25, 5.26it/s] 55%|█████▌ | 26411/48008 [1:30:35<1:10:23, 5.11it/s] 55%|█████▌ | 26412/48008 [1:30:35<1:10:11, 5.13it/s] 55%|█████▌ | 26413/48008 [1:30:35<1:09:34, 5.17it/s] 55%|█████▌ | 26414/48008 [1:30:35<1:09:01, 5.21it/s] 55%|█████▌ | 26415/48008 [1:30:35<1:14:04, 4.86it/s] 55%|█████▌ | 26416/48008 [1:30:36<1:17:30, 4.64it/s] 55%|█████▌ | 26417/48008 [1:30:36<1:14:53, 4.80it/s] 55%|█████▌ | 26418/48008 [1:30:36<1:12:04, 4.99it/s] 55%|█████▌ | 26419/48008 [1:30:36<1:32:18, 3.90it/s] 55%|█████▌ | 26420/48008 [1:30:37<1:26:23, 4.16it/s] 55%|█████▌ | 26421/48008 [1:30:37<1:20:38, 4.46it/s] 55%|█████▌ | 26422/48008 [1:30:37<1:17:10, 4.66it/s] 55%|█████▌ | 26423/48008 [1:30:37<1:14:33, 4.82it/s] 55%|█████▌ | 26424/48008 [1:30:37<1:13:55, 4.87it/s] 55%|█████▌ | 26425/48008 [1:30:38<1:13:45, 4.88it/s] 55%|█████▌ | 26426/48008 [1:30:38<1:13:31, 4.89it/s] 55%|█████▌ | 26427/48008 [1:30:38<1:13:33, 4.89it/s] 55%|█████▌ | 26428/48008 [1:30:38<1:11:33, 5.03it/s] 55%|█████▌ | 26429/48008 [1:30:38<1:10:39, 5.09it/s] 55%|█████▌ | 26430/48008 [1:30:39<1:09:36, 5.17it/s] 55%|█████▌ | 26431/48008 [1:30:39<1:09:48, 5.15it/s] 55%|█████▌ | 26432/48008 [1:30:39<1:09:27, 5.18it/s] 55%|█████▌ | 26433/48008 [1:30:39<1:09:49, 5.15it/s] 55%|█████▌ | 26434/48008 [1:30:39<1:09:24, 5.18it/s] 55%|█████▌ | 26435/48008 [1:30:39<1:09:51, 5.15it/s] 55%|█████▌ | 26436/48008 [1:30:40<1:10:10, 5.12it/s] 55%|█████▌ | 26437/48008 [1:30:40<1:09:22, 5.18it/s] 55%|█████▌ | 26438/48008 [1:30:40<1:09:58, 5.14it/s] 55%|█████▌ | 26439/48008 [1:30:40<1:10:55, 5.07it/s] 55%|█████▌ | 26440/48008 [1:30:40<1:10:19, 5.11it/s] 55%|█████▌ | 26441/48008 [1:30:41<1:09:15, 5.19it/s] 55%|█████▌ | 26442/48008 [1:30:41<1:09:30, 5.17it/s] 55%|█████▌ | 26443/48008 [1:30:41<1:08:46, 5.23it/s] 55%|█████▌ | 26444/48008 [1:30:41<1:08:43, 5.23it/s] 55%|█████▌ | 26445/48008 [1:30:41<1:09:28, 5.17it/s] 55%|█████▌ | 26446/48008 [1:30:42<1:09:15, 5.19it/s] 55%|█████▌ | 26447/48008 [1:30:42<1:13:54, 4.86it/s] 55%|█████▌ | 26448/48008 [1:30:42<1:12:15, 4.97it/s] 55%|█████▌ | 26449/48008 [1:30:42<1:11:04, 5.06it/s] 55%|█████▌ | 26450/48008 [1:30:42<1:09:51, 5.14it/s] {'loss': 0.2985, 'grad_norm': 0.158203125, 'learning_rate': 8.981419763372772e-05, 'epoch': 0.55} 55%|█████▌ | 26450/48008 [1:30:42<1:09:51, 5.14it/s] 55%|█████▌ | 26451/48008 [1:30:43<1:09:32, 5.17it/s] 55%|█████▌ | 26452/48008 [1:30:43<1:08:45, 5.23it/s] 55%|█████▌ | 26453/48008 [1:30:43<1:08:33, 5.24it/s] 55%|█████▌ | 26454/48008 [1:30:43<1:08:30, 5.24it/s] 55%|█████▌ | 26455/48008 [1:30:43<1:08:05, 5.28it/s] 55%|█████▌ | 26456/48008 [1:30:44<1:08:12, 5.27it/s] 55%|█████▌ | 26457/48008 [1:30:44<1:08:42, 5.23it/s] 55%|█████▌ | 26458/48008 [1:30:44<1:09:10, 5.19it/s] 55%|█████▌ | 26459/48008 [1:30:44<1:09:12, 5.19it/s] 55%|█████▌ | 26460/48008 [1:30:44<1:09:20, 5.18it/s] 55%|█████▌ | 26461/48008 [1:30:45<1:11:12, 5.04it/s] 55%|█████▌ | 26462/48008 [1:30:45<1:15:16, 4.77it/s] 55%|█████▌ | 26463/48008 [1:30:45<1:12:13, 4.97it/s] 55%|█████▌ | 26464/48008 [1:30:45<1:11:00, 5.06it/s] 55%|█████▌ | 26465/48008 [1:30:45<1:10:20, 5.10it/s] 55%|█████▌ | 26466/48008 [1:30:47<2:55:26, 2.05it/s] 55%|█████▌ | 26467/48008 [1:30:47<2:23:42, 2.50it/s] 55%|█████▌ | 26468/48008 [1:30:47<2:01:24, 2.96it/s] 55%|█████▌ | 26469/48008 [1:30:47<1:46:20, 3.38it/s] 55%|█████▌ | 26470/48008 [1:30:47<1:34:53, 3.78it/s] 55%|█████▌ | 26471/48008 [1:30:47<1:26:49, 4.13it/s] 55%|█████▌ | 26472/48008 [1:30:48<1:22:56, 4.33it/s] 55%|█████▌ | 26473/48008 [1:30:48<1:18:29, 4.57it/s] 55%|█████▌ | 26474/48008 [1:30:48<1:17:20, 4.64it/s] 55%|█████▌ | 26475/48008 [1:30:48<1:14:12, 4.84it/s] 55%|█████▌ | 26476/48008 [1:30:48<1:12:16, 4.97it/s] 55%|█████▌ | 26477/48008 [1:30:49<1:12:22, 4.96it/s] 55%|█████▌ | 26478/48008 [1:30:49<1:11:56, 4.99it/s] 55%|█████▌ | 26479/48008 [1:30:49<1:11:17, 5.03it/s] 55%|█████▌ | 26480/48008 [1:30:49<1:11:43, 5.00it/s] 55%|█████▌ | 26481/48008 [1:30:49<1:10:16, 5.11it/s] 55%|█████▌ | 26482/48008 [1:30:50<1:10:23, 5.10it/s] 55%|█████▌ | 26483/48008 [1:30:50<1:09:37, 5.15it/s] 55%|█████▌ | 26484/48008 [1:30:50<1:09:24, 5.17it/s] 55%|█████▌ | 26485/48008 [1:30:50<1:09:47, 5.14it/s] 55%|█████▌ | 26486/48008 [1:30:50<1:09:26, 5.17it/s] 55%|█████▌ | 26487/48008 [1:30:51<1:30:08, 3.98it/s] 55%|█████▌ | 26488/48008 [1:30:51<1:23:34, 4.29it/s] 55%|█████▌ | 26489/48008 [1:30:51<1:19:01, 4.54it/s] 55%|█████▌ | 26490/48008 [1:30:51<1:15:39, 4.74it/s] 55%|█████▌ | 26491/48008 [1:30:52<1:14:04, 4.84it/s] 55%|█████▌ | 26492/48008 [1:30:52<1:11:45, 5.00it/s] 55%|█████▌ | 26493/48008 [1:30:52<1:12:43, 4.93it/s] 55%|█████▌ | 26494/48008 [1:30:52<1:11:21, 5.03it/s] 55%|█████▌ | 26495/48008 [1:30:52<1:11:40, 5.00it/s] 55%|█████▌ | 26496/48008 [1:30:53<1:10:27, 5.09it/s] 55%|█████▌ | 26497/48008 [1:30:53<1:09:43, 5.14it/s] 55%|█████▌ | 26498/48008 [1:30:53<1:10:27, 5.09it/s] 55%|█████▌ | 26499/48008 [1:30:53<1:09:45, 5.14it/s] 55%|█████▌ | 26500/48008 [1:30:53<1:09:38, 5.15it/s] {'loss': 0.301, 'grad_norm': 0.197265625, 'learning_rate': 8.960589901683052e-05, 'epoch': 0.55} 55%|█████▌ | 26500/48008 [1:30:53<1:09:38, 5.15it/s] 55%|█████▌ | 26501/48008 [1:30:53<1:09:10, 5.18it/s] 55%|█████▌ | 26502/48008 [1:30:54<1:10:10, 5.11it/s] 55%|█████▌ | 26503/48008 [1:30:54<1:22:02, 4.37it/s] 55%|█████▌ | 26504/48008 [1:30:54<1:17:36, 4.62it/s] 55%|█████▌ | 26505/48008 [1:30:54<1:14:14, 4.83it/s] 55%|█████▌ | 26506/48008 [1:30:55<1:12:21, 4.95it/s] 55%|█████▌ | 26507/48008 [1:30:55<1:11:02, 5.04it/s] 55%|█████▌ | 26508/48008 [1:30:55<1:10:06, 5.11it/s] 55%|█████▌ | 26509/48008 [1:30:55<1:09:35, 5.15it/s] 55%|█████▌ | 26510/48008 [1:30:55<1:09:11, 5.18it/s] 55%|█████▌ | 26511/48008 [1:30:56<1:09:42, 5.14it/s] 55%|█████▌ | 26512/48008 [1:30:56<1:09:05, 5.18it/s] 55%|█████▌ | 26513/48008 [1:30:56<1:08:12, 5.25it/s] 55%|█████▌ | 26514/48008 [1:30:56<1:07:58, 5.27it/s] 55%|█████▌ | 26515/48008 [1:30:56<1:07:28, 5.31it/s] 55%|█████▌ | 26516/48008 [1:30:56<1:07:40, 5.29it/s] 55%|█████▌ | 26517/48008 [1:30:57<1:07:36, 5.30it/s] 55%|█████▌ | 26518/48008 [1:30:57<1:08:03, 5.26it/s] 55%|█████▌ | 26519/48008 [1:30:57<1:08:14, 5.25it/s] 55%|█████▌ | 26520/48008 [1:30:57<1:08:14, 5.25it/s] 55%|█████▌ | 26521/48008 [1:30:58<1:29:14, 4.01it/s] 55%|█████▌ | 26522/48008 [1:30:58<1:24:06, 4.26it/s] 55%|█████▌ | 26523/48008 [1:30:58<1:19:08, 4.52it/s] 55%|█████▌ | 26524/48008 [1:30:58<1:15:50, 4.72it/s] 55%|█████▌ | 26525/48008 [1:30:58<1:13:12, 4.89it/s] 55%|█████▌ | 26526/48008 [1:30:59<1:11:47, 4.99it/s] 55%|█████▌ | 26527/48008 [1:30:59<1:10:37, 5.07it/s] 55%|█████▌ | 26528/48008 [1:30:59<1:09:48, 5.13it/s] 55%|█████▌ | 26529/48008 [1:30:59<1:09:10, 5.17it/s] 55%|█████▌ | 26530/48008 [1:30:59<1:10:22, 5.09it/s] 55%|█████▌ | 26531/48008 [1:31:00<1:10:14, 5.10it/s] 55%|█████▌ | 26532/48008 [1:31:00<1:14:34, 4.80it/s] 55%|█████▌ | 26533/48008 [1:31:00<1:12:42, 4.92it/s] 55%|█████▌ | 26534/48008 [1:31:00<1:32:29, 3.87it/s] 55%|█████▌ | 26535/48008 [1:31:01<1:26:25, 4.14it/s] 55%|█████▌ | 26536/48008 [1:31:01<1:26:09, 4.15it/s] 55%|█████▌ | 26537/48008 [1:31:01<1:21:19, 4.40it/s] 55%|█████▌ | 26538/48008 [1:31:01<1:17:29, 4.62it/s] 55%|█████▌ | 26539/48008 [1:31:01<1:15:24, 4.75it/s] 55%|█████▌ | 26540/48008 [1:31:02<1:12:27, 4.94it/s] 55%|█████▌ | 26541/48008 [1:31:02<1:11:08, 5.03it/s] 55%|█████▌ | 26542/48008 [1:31:02<1:11:16, 5.02it/s] 55%|█████▌ | 26543/48008 [1:31:02<1:10:14, 5.09it/s] 55%|█████▌ | 26544/48008 [1:31:02<1:10:18, 5.09it/s] 55%|█████▌ | 26545/48008 [1:31:03<1:09:35, 5.14it/s] 55%|█████▌ | 26546/48008 [1:31:03<1:09:19, 5.16it/s] 55%|█████▌ | 26547/48008 [1:31:03<1:08:48, 5.20it/s] 55%|█████▌ | 26548/48008 [1:31:03<1:08:11, 5.25it/s] 55%|█████▌ | 26549/48008 [1:31:03<1:08:46, 5.20it/s] 55%|█████▌ | 26550/48008 [1:31:03<1:07:25, 5.30it/s] {'loss': 0.3029, 'grad_norm': 0.2412109375, 'learning_rate': 8.939760039993334e-05, 'epoch': 0.55} 55%|█████▌ | 26550/48008 [1:31:03<1:07:25, 5.30it/s] 55%|█████▌ | 26551/48008 [1:31:04<1:09:52, 5.12it/s] 55%|█████▌ | 26552/48008 [1:31:04<1:08:55, 5.19it/s] 55%|█████▌ | 26553/48008 [1:31:04<1:09:00, 5.18it/s] 55%|█████▌ | 26554/48008 [1:31:04<1:08:25, 5.23it/s] 55%|█████▌ | 26555/48008 [1:31:04<1:08:00, 5.26it/s] 55%|█████▌ | 26556/48008 [1:31:05<1:10:01, 5.11it/s] 55%|█████▌ | 26557/48008 [1:31:05<1:09:49, 5.12it/s] 55%|█████▌ | 26558/48008 [1:31:05<1:09:13, 5.16it/s] 55%|█████▌ | 26559/48008 [1:31:05<1:09:00, 5.18it/s] 55%|█████▌ | 26560/48008 [1:31:05<1:09:17, 5.16it/s] 55%|█████▌ | 26561/48008 [1:31:06<1:09:34, 5.14it/s] 55%|█████▌ | 26562/48008 [1:31:06<1:08:47, 5.20it/s] 55%|█████▌ | 26563/48008 [1:31:06<1:08:43, 5.20it/s] 55%|█████▌ | 26564/48008 [1:31:06<1:07:57, 5.26it/s] 55%|█████▌ | 26565/48008 [1:31:06<1:09:13, 5.16it/s] 55%|█████▌ | 26566/48008 [1:31:07<1:08:59, 5.18it/s] 55%|█████▌ | 26567/48008 [1:31:07<1:10:46, 5.05it/s] 55%|█████▌ | 26568/48008 [1:31:07<1:11:22, 5.01it/s] 55%|█████▌ | 26569/48008 [1:31:07<1:11:42, 4.98it/s] 55%|█████▌ | 26570/48008 [1:31:07<1:10:48, 5.05it/s] 55%|█████▌ | 26571/48008 [1:31:08<1:11:53, 4.97it/s] 55%|█████▌ | 26572/48008 [1:31:08<1:12:22, 4.94it/s] 55%|█████▌ | 26573/48008 [1:31:08<1:12:35, 4.92it/s] 55%|█████▌ | 26574/48008 [1:31:08<1:11:25, 5.00it/s] 55%|█████▌ | 26575/48008 [1:31:08<1:10:34, 5.06it/s] 55%|█████▌ | 26576/48008 [1:31:09<1:10:35, 5.06it/s] 55%|█████▌ | 26577/48008 [1:31:09<1:09:57, 5.11it/s] 55%|█████▌ | 26578/48008 [1:31:09<1:09:36, 5.13it/s] 55%|█████▌ | 26579/48008 [1:31:09<1:08:54, 5.18it/s] 55%|█████▌ | 26580/48008 [1:31:09<1:08:37, 5.20it/s] 55%|█████▌ | 26581/48008 [1:31:10<1:08:23, 5.22it/s] 55%|█████▌ | 26582/48008 [1:31:10<1:07:55, 5.26it/s] 55%|█████▌ | 26583/48008 [1:31:10<1:08:25, 5.22it/s] 55%|█████▌ | 26584/48008 [1:31:10<1:08:47, 5.19it/s] 55%|█████▌ | 26585/48008 [1:31:10<1:08:06, 5.24it/s] 55%|█████▌ | 26586/48008 [1:31:11<1:13:13, 4.88it/s] 55%|█████▌ | 26587/48008 [1:31:11<1:12:01, 4.96it/s] 55%|█████▌ | 26588/48008 [1:31:11<1:11:03, 5.02it/s] 55%|█████▌ | 26589/48008 [1:31:11<1:10:27, 5.07it/s] 55%|█████▌ | 26590/48008 [1:31:11<1:09:55, 5.10it/s] 55%|█████▌ | 26591/48008 [1:31:12<1:10:36, 5.06it/s] 55%|█████▌ | 26592/48008 [1:31:12<1:09:50, 5.11it/s] 55%|█████▌ | 26593/48008 [1:31:12<1:08:45, 5.19it/s] 55%|█████▌ | 26594/48008 [1:31:12<1:10:05, 5.09it/s] 55%|█████▌ | 26595/48008 [1:31:12<1:10:34, 5.06it/s] 55%|█████▌ | 26596/48008 [1:31:12<1:10:38, 5.05it/s] 55%|█████▌ | 26597/48008 [1:31:13<1:10:41, 5.05it/s] 55%|█████▌ | 26598/48008 [1:31:13<1:30:53, 3.93it/s] 55%|█████▌ | 26599/48008 [1:31:13<1:23:31, 4.27it/s] 55%|█████▌ | 26600/48008 [1:31:13<1:20:23, 4.44it/s] {'loss': 0.2997, 'grad_norm': 0.19140625, 'learning_rate': 8.918930178303616e-05, 'epoch': 0.55} 55%|█████▌ | 26600/48008 [1:31:13<1:20:23, 4.44it/s] 55%|█████▌ | 26601/48008 [1:31:14<1:17:01, 4.63it/s] 55%|█████▌ | 26602/48008 [1:31:14<1:14:49, 4.77it/s] 55%|█████▌ | 26603/48008 [1:31:14<1:14:07, 4.81it/s] 55%|█████▌ | 26604/48008 [1:31:14<1:12:17, 4.93it/s] 55%|█████▌ | 26605/48008 [1:31:14<1:11:11, 5.01it/s] 55%|█████▌ | 26606/48008 [1:31:15<1:10:07, 5.09it/s] 55%|█████▌ | 26607/48008 [1:31:15<1:09:22, 5.14it/s] 55%|█████▌ | 26608/48008 [1:31:15<1:10:13, 5.08it/s] 55%|█████▌ | 26609/48008 [1:31:15<1:09:10, 5.16it/s] 55%|█████▌ | 26610/48008 [1:31:15<1:10:37, 5.05it/s] 55%|█████▌ | 26611/48008 [1:31:16<1:15:05, 4.75it/s] 55%|█████▌ | 26612/48008 [1:31:16<1:12:58, 4.89it/s] 55%|█████▌ | 26613/48008 [1:31:16<1:16:35, 4.66it/s] 55%|█████▌ | 26614/48008 [1:31:16<1:14:04, 4.81it/s] 55%|█████▌ | 26615/48008 [1:31:16<1:12:29, 4.92it/s] 55%|█████▌ | 26616/48008 [1:31:17<1:32:07, 3.87it/s] 55%|█████▌ | 26617/48008 [1:31:17<1:27:13, 4.09it/s] 55%|█████▌ | 26618/48008 [1:31:17<1:21:25, 4.38it/s] 55%|█████▌ | 26619/48008 [1:31:17<1:17:48, 4.58it/s] 55%|█████▌ | 26620/48008 [1:31:18<1:14:22, 4.79it/s] 55%|█████▌ | 26621/48008 [1:31:18<1:12:48, 4.90it/s] 55%|█████▌ | 26622/48008 [1:31:18<1:16:18, 4.67it/s] 55%|█████▌ | 26623/48008 [1:31:18<1:13:40, 4.84it/s] 55%|█████▌ | 26624/48008 [1:31:18<1:13:31, 4.85it/s] 55%|█████▌ | 26625/48008 [1:31:19<1:11:33, 4.98it/s] 55%|█████▌ | 26626/48008 [1:31:19<1:10:25, 5.06it/s] 55%|█████▌ | 26627/48008 [1:31:19<1:09:30, 5.13it/s] 55%|█████▌ | 26628/48008 [1:31:19<1:09:02, 5.16it/s] 55%|█████▌ | 26629/48008 [1:31:19<1:09:56, 5.09it/s] 55%|█████▌ | 26630/48008 [1:31:20<1:09:59, 5.09it/s] 55%|█████▌ | 26631/48008 [1:31:20<1:09:16, 5.14it/s] 55%|█████▌ | 26632/48008 [1:31:20<1:09:08, 5.15it/s] 55%|█████▌ | 26633/48008 [1:31:20<1:09:18, 5.14it/s] 55%|█████▌ | 26634/48008 [1:31:20<1:08:53, 5.17it/s] 55%|█████▌ | 26635/48008 [1:31:21<1:09:49, 5.10it/s] 55%|█████▌ | 26636/48008 [1:31:21<1:09:04, 5.16it/s] 55%|█████▌ | 26637/48008 [1:31:21<1:10:35, 5.05it/s] 55%|█████▌ | 26638/48008 [1:31:21<1:10:22, 5.06it/s] 55%|█████▌ | 26639/48008 [1:31:21<1:09:30, 5.12it/s] 55%|█████▌ | 26640/48008 [1:31:22<1:09:25, 5.13it/s] 55%|█████▌ | 26641/48008 [1:31:22<1:10:21, 5.06it/s] 55%|█████▌ | 26642/48008 [1:31:22<1:09:52, 5.10it/s] 55%|█████▌ | 26643/48008 [1:31:22<1:09:17, 5.14it/s] 55%|█████▌ | 26644/48008 [1:31:22<1:09:17, 5.14it/s] 56%|█████▌ | 26645/48008 [1:31:23<1:09:40, 5.11it/s] 56%|█████▌ | 26646/48008 [1:31:23<1:08:50, 5.17it/s] 56%|█████▌ | 26647/48008 [1:31:23<1:08:51, 5.17it/s] 56%|█████▌ | 26648/48008 [1:31:23<1:08:44, 5.18it/s] 56%|█████▌ | 26649/48008 [1:31:23<1:07:46, 5.25it/s] 56%|█████▌ | 26650/48008 [1:31:24<1:09:03, 5.15it/s] {'loss': 0.3025, 'grad_norm': 0.15234375, 'learning_rate': 8.898100316613898e-05, 'epoch': 0.56} 56%|█████▌ | 26650/48008 [1:31:24<1:09:03, 5.15it/s] 56%|█████▌ | 26651/48008 [1:31:24<1:08:59, 5.16it/s] 56%|█████▌ | 26652/48008 [1:31:24<1:08:55, 5.16it/s] 56%|█████▌ | 26653/48008 [1:31:24<1:08:32, 5.19it/s] 56%|█████▌ | 26654/48008 [1:31:24<1:08:29, 5.20it/s] 56%|█████▌ | 26655/48008 [1:31:24<1:08:45, 5.18it/s] 56%|█████▌ | 26656/48008 [1:31:25<1:09:44, 5.10it/s] 56%|█████▌ | 26657/48008 [1:31:25<1:09:16, 5.14it/s] 56%|█████▌ | 26658/48008 [1:31:25<1:08:26, 5.20it/s] 56%|█████▌ | 26659/48008 [1:31:25<1:09:03, 5.15it/s] 56%|█████▌ | 26660/48008 [1:31:25<1:08:49, 5.17it/s] 56%|█████▌ | 26661/48008 [1:31:26<1:08:44, 5.18it/s] 56%|█████▌ | 26662/48008 [1:31:26<1:08:04, 5.23it/s] 56%|█████▌ | 26663/48008 [1:31:26<1:07:56, 5.24it/s] 56%|█████▌ | 26664/48008 [1:31:26<1:07:36, 5.26it/s] 56%|█████▌ | 26665/48008 [1:31:26<1:08:40, 5.18it/s] 56%|█████▌ | 26666/48008 [1:31:27<1:07:45, 5.25it/s] 56%|█████▌ | 26667/48008 [1:31:27<1:07:52, 5.24it/s] 56%|█████▌ | 26668/48008 [1:31:27<1:07:02, 5.31it/s] 56%|█████▌ | 26669/48008 [1:31:27<1:07:47, 5.25it/s] 56%|█████▌ | 26670/48008 [1:31:27<1:07:48, 5.25it/s] 56%|█████▌ | 26671/48008 [1:31:28<1:07:13, 5.29it/s] 56%|█████▌ | 26672/48008 [1:31:28<1:08:10, 5.22it/s] 56%|█████▌ | 26673/48008 [1:31:28<1:13:05, 4.87it/s] 56%|█████▌ | 26674/48008 [1:31:28<1:32:34, 3.84it/s] 56%|█████▌ | 26675/48008 [1:31:29<1:25:44, 4.15it/s] 56%|█████▌ | 26676/48008 [1:31:29<1:19:58, 4.45it/s] 56%|█████▌ | 26677/48008 [1:31:29<1:15:40, 4.70it/s] 56%|█████▌ | 26678/48008 [1:31:29<1:13:22, 4.85it/s] 56%|█████▌ | 26679/48008 [1:31:29<1:11:43, 4.96it/s] 56%|█████▌ | 26680/48008 [1:31:30<1:10:40, 5.03it/s] 56%|█████▌ | 26681/48008 [1:31:30<1:10:05, 5.07it/s] 56%|█████▌ | 26682/48008 [1:31:30<1:09:35, 5.11it/s] 56%|█████▌ | 26683/48008 [1:31:30<1:11:04, 5.00it/s] 56%|█████▌ | 26684/48008 [1:31:30<1:10:02, 5.07it/s] 56%|█████▌ | 26685/48008 [1:31:30<1:10:58, 5.01it/s] 56%|█████▌ | 26686/48008 [1:31:31<1:10:28, 5.04it/s] 56%|█████▌ | 26687/48008 [1:31:31<1:09:53, 5.08it/s] 56%|█████▌ | 26688/48008 [1:31:31<1:09:31, 5.11it/s] 56%|█████▌ | 26689/48008 [1:31:31<1:10:02, 5.07it/s] 56%|█████▌ | 26690/48008 [1:31:31<1:09:41, 5.10it/s] 56%|█████▌ | 26691/48008 [1:31:32<1:09:15, 5.13it/s] 56%|█████▌ | 26692/48008 [1:31:32<1:08:58, 5.15it/s] 56%|█████▌ | 26693/48008 [1:31:32<1:08:41, 5.17it/s] 56%|█████▌ | 26694/48008 [1:31:32<1:08:36, 5.18it/s] 56%|█████▌ | 26695/48008 [1:31:32<1:08:10, 5.21it/s] 56%|█████▌ | 26696/48008 [1:31:33<1:08:16, 5.20it/s] 56%|█████▌ | 26697/48008 [1:31:33<1:09:23, 5.12it/s] 56%|█████▌ | 26698/48008 [1:31:33<1:08:56, 5.15it/s] 56%|█████▌ | 26699/48008 [1:31:33<1:09:05, 5.14it/s] 56%|█████▌ | 26700/48008 [1:31:33<1:08:58, 5.15it/s] {'loss': 0.3026, 'grad_norm': 0.208984375, 'learning_rate': 8.877270454924179e-05, 'epoch': 0.56} 56%|█████▌ | 26700/48008 [1:31:33<1:08:58, 5.15it/s] 56%|█████▌ | 26701/48008 [1:31:34<1:08:50, 5.16it/s] 56%|█████▌ | 26702/48008 [1:31:34<1:09:46, 5.09it/s] 56%|█████▌ | 26703/48008 [1:31:34<1:09:22, 5.12it/s] 56%|█████▌ | 26704/48008 [1:31:34<1:08:33, 5.18it/s] 56%|█████▌ | 26705/48008 [1:31:34<1:08:50, 5.16it/s] 56%|█████▌ | 26706/48008 [1:31:35<1:08:44, 5.16it/s] 56%|█████▌ | 26707/48008 [1:31:35<1:08:26, 5.19it/s] 56%|█████▌ | 26708/48008 [1:31:35<1:09:01, 5.14it/s] 56%|█████▌ | 26709/48008 [1:31:35<1:08:17, 5.20it/s] 56%|█████▌ | 26710/48008 [1:31:35<1:08:15, 5.20it/s] 56%|█████▌ | 26711/48008 [1:31:36<1:13:07, 4.85it/s] 56%|█████▌ | 26712/48008 [1:31:36<1:11:28, 4.97it/s] 56%|█████▌ | 26713/48008 [1:31:36<1:10:34, 5.03it/s] 56%|█████▌ | 26714/48008 [1:31:36<1:11:59, 4.93it/s] 56%|█████▌ | 26715/48008 [1:31:36<1:10:41, 5.02it/s] 56%|█████▌ | 26716/48008 [1:31:37<1:10:03, 5.07it/s] 56%|█████▌ | 26717/48008 [1:31:37<1:09:26, 5.11it/s] 56%|█████▌ | 26718/48008 [1:31:37<1:09:07, 5.13it/s] 56%|█████▌ | 26719/48008 [1:31:37<1:08:28, 5.18it/s] 56%|█████▌ | 26720/48008 [1:31:37<1:10:17, 5.05it/s] 56%|█████▌ | 26721/48008 [1:31:38<1:10:47, 5.01it/s] 56%|█████▌ | 26722/48008 [1:31:38<1:10:08, 5.06it/s] 56%|█████▌ | 26723/48008 [1:31:38<1:09:30, 5.10it/s] 56%|█████▌ | 26724/48008 [1:31:38<1:08:43, 5.16it/s] 56%|█████▌ | 26725/48008 [1:31:38<1:07:58, 5.22it/s] 56%|█████▌ | 26726/48008 [1:31:39<1:13:02, 4.86it/s] 56%|█████▌ | 26727/48008 [1:31:39<1:11:39, 4.95it/s] 56%|█████▌ | 26728/48008 [1:31:39<1:09:58, 5.07it/s] 56%|█████▌ | 26729/48008 [1:31:39<1:09:10, 5.13it/s] 56%|█████▌ | 26730/48008 [1:31:39<1:08:43, 5.16it/s] 56%|█████▌ | 26731/48008 [1:31:39<1:08:34, 5.17it/s] 56%|█████▌ | 26732/48008 [1:31:40<1:08:24, 5.18it/s] 56%|█████▌ | 26733/48008 [1:31:40<1:07:56, 5.22it/s] 56%|█████▌ | 26734/48008 [1:31:40<1:07:45, 5.23it/s] 56%|█████▌ | 26735/48008 [1:31:40<1:08:16, 5.19it/s] 56%|█████▌ | 26736/48008 [1:31:40<1:09:02, 5.13it/s] 56%|█████▌ | 26737/48008 [1:31:41<1:29:34, 3.96it/s] 56%|█████▌ | 26738/48008 [1:31:41<1:24:29, 4.20it/s] 56%|█████▌ | 26739/48008 [1:31:41<1:20:19, 4.41it/s] 56%|█████▌ | 26740/48008 [1:31:41<1:21:37, 4.34it/s] 56%|█████▌ | 26741/48008 [1:31:42<1:17:26, 4.58it/s] 56%|█████▌ | 26742/48008 [1:31:42<1:14:26, 4.76it/s] 56%|█████▌ | 26743/48008 [1:31:42<1:12:28, 4.89it/s] 56%|█████▌ | 26744/48008 [1:31:42<1:12:14, 4.91it/s] 56%|█████▌ | 26745/48008 [1:31:42<1:11:40, 4.94it/s] 56%|█████▌ | 26746/48008 [1:31:43<1:10:28, 5.03it/s] 56%|█████▌ | 26747/48008 [1:31:43<1:09:51, 5.07it/s] 56%|█████▌ | 26748/48008 [1:31:43<1:09:51, 5.07it/s] 56%|█████▌ | 26749/48008 [1:31:43<1:09:16, 5.12it/s] 56%|█████▌ | 26750/48008 [1:31:43<1:10:16, 5.04it/s] {'loss': 0.2991, 'grad_norm': 0.396484375, 'learning_rate': 8.856440593234462e-05, 'epoch': 0.56} 56%|█████▌ | 26750/48008 [1:31:43<1:10:16, 5.04it/s] 56%|█████▌ | 26751/48008 [1:31:44<1:09:09, 5.12it/s] 56%|█████▌ | 26752/48008 [1:31:44<1:09:09, 5.12it/s] 56%|█████▌ | 26753/48008 [1:31:44<1:10:03, 5.06it/s] 56%|█████▌ | 26754/48008 [1:31:44<1:08:59, 5.13it/s] 56%|█████▌ | 26755/48008 [1:31:44<1:08:23, 5.18it/s] 56%|█████▌ | 26756/48008 [1:31:45<1:08:15, 5.19it/s] 56%|█████▌ | 26757/48008 [1:31:45<1:07:34, 5.24it/s] 56%|█████▌ | 26758/48008 [1:31:45<1:07:06, 5.28it/s] 56%|█████▌ | 26759/48008 [1:31:45<1:06:46, 5.30it/s] 56%|█████▌ | 26760/48008 [1:31:45<1:07:13, 5.27it/s] 56%|█████▌ | 26761/48008 [1:31:46<1:07:28, 5.25it/s] 56%|█████▌ | 26762/48008 [1:31:46<1:08:59, 5.13it/s] 56%|█████▌ | 26763/48008 [1:31:46<1:08:44, 5.15it/s] 56%|█████▌ | 26764/48008 [1:31:46<1:09:08, 5.12it/s] 56%|█████▌ | 26765/48008 [1:31:46<1:08:50, 5.14it/s] 56%|█████▌ | 26766/48008 [1:31:47<1:08:36, 5.16it/s] 56%|█████▌ | 26767/48008 [1:31:47<1:08:21, 5.18it/s] 56%|█████▌ | 26768/48008 [1:31:47<1:08:16, 5.18it/s] 56%|█████▌ | 26769/48008 [1:31:47<1:07:29, 5.24it/s] 56%|█████▌ | 26770/48008 [1:31:47<1:07:09, 5.27it/s] 56%|█████▌ | 26771/48008 [1:31:47<1:07:09, 5.27it/s] 56%|█████▌ | 26772/48008 [1:31:48<1:06:57, 5.29it/s] 56%|█████▌ | 26773/48008 [1:31:48<1:12:06, 4.91it/s] 56%|█████▌ | 26774/48008 [1:31:48<1:11:28, 4.95it/s] 56%|█████▌ | 26775/48008 [1:31:48<1:10:28, 5.02it/s] 56%|█████▌ | 26776/48008 [1:31:48<1:10:48, 5.00it/s] 56%|█████▌ | 26777/48008 [1:31:49<1:09:57, 5.06it/s] 56%|█████▌ | 26778/48008 [1:31:49<1:10:46, 5.00it/s] 56%|█████▌ | 26779/48008 [1:31:49<1:22:08, 4.31it/s] 56%|█████▌ | 26780/48008 [1:31:49<1:17:40, 4.55it/s] 56%|█████▌ | 26781/48008 [1:31:50<1:14:11, 4.77it/s] 56%|█████▌ | 26782/48008 [1:31:50<1:12:34, 4.87it/s] 56%|█████▌ | 26783/48008 [1:31:50<1:10:39, 5.01it/s] 56%|█████▌ | 26784/48008 [1:31:50<1:09:39, 5.08it/s] 56%|█████▌ | 26785/48008 [1:31:50<1:09:11, 5.11it/s] 56%|█████▌ | 26786/48008 [1:31:51<1:08:51, 5.14it/s] 56%|█████▌ | 26787/48008 [1:31:51<1:08:55, 5.13it/s] 56%|█████▌ | 26788/48008 [1:31:51<1:20:56, 4.37it/s] 56%|█████▌ | 26789/48008 [1:31:51<1:18:57, 4.48it/s] 56%|█████▌ | 26790/48008 [1:31:51<1:16:54, 4.60it/s] 56%|█████▌ | 26791/48008 [1:31:52<1:13:38, 4.80it/s] 56%|█████▌ | 26792/48008 [1:31:52<1:11:56, 4.92it/s] 56%|█████▌ | 26793/48008 [1:31:52<1:12:07, 4.90it/s] 56%|█████▌ | 26794/48008 [1:31:52<1:12:09, 4.90it/s] 56%|█████▌ | 26795/48008 [1:31:52<1:11:58, 4.91it/s] 56%|█████▌ | 26796/48008 [1:31:53<1:10:10, 5.04it/s] 56%|█████▌ | 26797/48008 [1:31:53<1:09:25, 5.09it/s] 56%|█████▌ | 26798/48008 [1:31:53<1:11:01, 4.98it/s] 56%|█████▌ | 26799/48008 [1:31:53<1:09:40, 5.07it/s] 56%|█████▌ | 26800/48008 [1:31:53<1:14:17, 4.76it/s] {'loss': 0.3015, 'grad_norm': 0.5625, 'learning_rate': 8.835610731544743e-05, 'epoch': 0.56} 56%|█████▌ | 26800/48008 [1:31:53<1:14:17, 4.76it/s] 56%|█████▌ | 26801/48008 [1:31:54<1:12:11, 4.90it/s] 56%|█████▌ | 26802/48008 [1:31:54<1:10:31, 5.01it/s] 56%|█████▌ | 26803/48008 [1:31:54<1:09:44, 5.07it/s] 56%|█████▌ | 26804/48008 [1:31:54<1:09:59, 5.05it/s] 56%|█████▌ | 26805/48008 [1:31:54<1:08:34, 5.15it/s] 56%|█████▌ | 26806/48008 [1:31:55<1:08:22, 5.17it/s] 56%|█████▌ | 26807/48008 [1:31:55<1:08:16, 5.18it/s] 56%|█████▌ | 26808/48008 [1:31:55<1:07:57, 5.20it/s] 56%|█████▌ | 26809/48008 [1:31:55<1:07:42, 5.22it/s] 56%|█████▌ | 26810/48008 [1:31:55<1:07:42, 5.22it/s] 56%|█████▌ | 26811/48008 [1:31:56<1:07:48, 5.21it/s] 56%|█████▌ | 26812/48008 [1:31:56<1:08:26, 5.16it/s] 56%|█████▌ | 26813/48008 [1:31:56<1:08:10, 5.18it/s] 56%|█████▌ | 26814/48008 [1:31:56<1:08:00, 5.19it/s] 56%|█████▌ | 26815/48008 [1:31:56<1:10:04, 5.04it/s] 56%|█████▌ | 26816/48008 [1:31:57<1:09:18, 5.10it/s] 56%|█████▌ | 26817/48008 [1:31:57<1:08:22, 5.17it/s] 56%|█████▌ | 26818/48008 [1:31:57<1:08:30, 5.16it/s] 56%|█████▌ | 26819/48008 [1:31:57<1:29:20, 3.95it/s] 56%|█████▌ | 26820/48008 [1:31:58<1:22:23, 4.29it/s] 56%|█████▌ | 26821/48008 [1:31:58<1:18:07, 4.52it/s] 56%|█████▌ | 26822/48008 [1:31:58<1:14:23, 4.75it/s] 56%|█████▌ | 26823/48008 [1:31:58<1:12:07, 4.90it/s] 56%|█████▌ | 26824/48008 [1:31:58<1:10:15, 5.03it/s] 56%|█████▌ | 26825/48008 [1:31:58<1:09:36, 5.07it/s] 56%|█████▌ | 26826/48008 [1:31:59<1:09:11, 5.10it/s] 56%|█████▌ | 26827/48008 [1:31:59<1:08:34, 5.15it/s] 56%|█████▌ | 26828/48008 [1:31:59<1:08:04, 5.19it/s] 56%|█████▌ | 26829/48008 [1:31:59<1:07:40, 5.22it/s] 56%|█████▌ | 26830/48008 [1:31:59<1:07:23, 5.24it/s] 56%|█████▌ | 26831/48008 [1:32:00<1:07:21, 5.24it/s] 56%|█████▌ | 26832/48008 [1:32:00<1:07:56, 5.19it/s] 56%|█████▌ | 26833/48008 [1:32:00<1:07:21, 5.24it/s] 56%|█████▌ | 26834/48008 [1:32:00<1:28:13, 4.00it/s] 56%|█████▌ | 26835/48008 [1:32:01<1:23:36, 4.22it/s] 56%|█████▌ | 26836/48008 [1:32:01<1:19:22, 4.45it/s] 56%|█████▌ | 26837/48008 [1:32:01<1:36:23, 3.66it/s] 56%|█████▌ | 26838/48008 [1:32:01<1:28:53, 3.97it/s] 56%|█████▌ | 26839/48008 [1:32:02<1:23:34, 4.22it/s] 56%|█████▌ | 26840/48008 [1:32:02<1:18:22, 4.50it/s] 56%|█████▌ | 26841/48008 [1:32:02<1:14:36, 4.73it/s] 56%|█████▌ | 26842/48008 [1:32:02<1:13:12, 4.82it/s] 56%|█████▌ | 26843/48008 [1:32:02<1:12:08, 4.89it/s] 56%|█████▌ | 26844/48008 [1:32:03<1:12:34, 4.86it/s] 56%|█████▌ | 26845/48008 [1:32:03<1:10:47, 4.98it/s] 56%|█████▌ | 26846/48008 [1:32:03<1:09:45, 5.06it/s] 56%|█████▌ | 26847/48008 [1:32:03<1:08:56, 5.12it/s] 56%|█████▌ | 26848/48008 [1:32:03<1:08:14, 5.17it/s] 56%|█████▌ | 26849/48008 [1:32:03<1:08:11, 5.17it/s] 56%|█████▌ | 26850/48008 [1:32:04<1:09:09, 5.10it/s] {'loss': 0.3011, 'grad_norm': 0.2265625, 'learning_rate': 8.814780869855025e-05, 'epoch': 0.56} 56%|█████▌ | 26850/48008 [1:32:04<1:09:09, 5.10it/s] 56%|█████▌ | 26851/48008 [1:32:04<1:13:31, 4.80it/s] 56%|█████▌ | 26852/48008 [1:32:04<1:16:38, 4.60it/s] 56%|█████▌ | 26853/48008 [1:32:04<1:13:33, 4.79it/s] 56%|█████▌ | 26854/48008 [1:32:05<1:13:37, 4.79it/s] 56%|█████▌ | 26855/48008 [1:32:05<1:11:22, 4.94it/s] 56%|█████▌ | 26856/48008 [1:32:05<1:10:11, 5.02it/s] 56%|█████▌ | 26857/48008 [1:32:05<1:09:07, 5.10it/s] 56%|█████▌ | 26858/48008 [1:32:05<1:08:22, 5.16it/s] 56%|█████▌ | 26859/48008 [1:32:06<1:09:10, 5.10it/s] 56%|█████▌ | 26860/48008 [1:32:06<1:08:29, 5.15it/s] 56%|█████▌ | 26861/48008 [1:32:06<1:08:02, 5.18it/s] 56%|█████▌ | 26862/48008 [1:32:06<1:08:59, 5.11it/s] 56%|█████▌ | 26863/48008 [1:32:06<1:09:36, 5.06it/s] 56%|█████▌ | 26864/48008 [1:32:06<1:09:01, 5.11it/s] 56%|█████▌ | 26865/48008 [1:32:07<1:07:55, 5.19it/s] 56%|█████▌ | 26866/48008 [1:32:07<1:09:02, 5.10it/s] 56%|█████▌ | 26867/48008 [1:32:07<1:07:59, 5.18it/s] 56%|█████▌ | 26868/48008 [1:32:07<1:07:45, 5.20it/s] 56%|█████▌ | 26869/48008 [1:32:07<1:07:33, 5.21it/s] 56%|█████▌ | 26870/48008 [1:32:08<1:08:43, 5.13it/s] 56%|█████▌ | 26871/48008 [1:32:08<1:08:12, 5.16it/s] 56%|█████▌ | 26872/48008 [1:32:08<1:07:56, 5.19it/s] 56%|█████▌ | 26873/48008 [1:32:08<1:07:39, 5.21it/s] 56%|█████▌ | 26874/48008 [1:32:08<1:08:39, 5.13it/s] 56%|█████▌ | 26875/48008 [1:32:09<1:08:09, 5.17it/s] 56%|█████▌ | 26876/48008 [1:32:09<1:09:00, 5.10it/s] 56%|█████▌ | 26877/48008 [1:32:09<1:10:24, 5.00it/s] 56%|█████▌ | 26878/48008 [1:32:09<1:10:24, 5.00it/s] 56%|█████▌ | 26879/48008 [1:32:09<1:10:40, 4.98it/s] 56%|█████▌ | 26880/48008 [1:32:10<1:10:10, 5.02it/s] 56%|█████▌ | 26881/48008 [1:32:10<1:10:33, 4.99it/s] 56%|█████▌ | 26882/48008 [1:32:10<1:21:44, 4.31it/s] 56%|█████▌ | 26883/48008 [1:32:10<1:16:54, 4.58it/s] 56%|█████▌ | 26884/48008 [1:32:11<1:14:20, 4.74it/s] 56%|█████▌ | 26885/48008 [1:32:11<1:13:04, 4.82it/s] 56%|█████▌ | 26886/48008 [1:32:11<1:12:26, 4.86it/s] 56%|█████▌ | 26887/48008 [1:32:11<1:11:50, 4.90it/s] 56%|█████▌ | 26888/48008 [1:32:11<1:11:45, 4.91it/s] 56%|█████▌ | 26889/48008 [1:32:12<1:10:31, 4.99it/s] 56%|█████▌ | 26890/48008 [1:32:12<1:09:20, 5.08it/s] 56%|█████▌ | 26891/48008 [1:32:12<1:07:59, 5.18it/s] 56%|█████▌ | 26892/48008 [1:32:12<1:07:05, 5.25it/s] 56%|█████▌ | 26893/48008 [1:32:12<1:07:23, 5.22it/s] 56%|█████▌ | 26894/48008 [1:32:12<1:07:04, 5.25it/s] 56%|█████▌ | 26895/48008 [1:32:13<1:07:08, 5.24it/s] 56%|█████▌ | 26896/48008 [1:32:13<1:07:16, 5.23it/s] 56%|█████▌ | 26897/48008 [1:32:13<1:07:08, 5.24it/s] 56%|█████▌ | 26898/48008 [1:32:13<1:07:18, 5.23it/s] 56%|█████▌ | 26899/48008 [1:32:13<1:07:18, 5.23it/s] 56%|█████▌ | 26900/48008 [1:32:14<1:08:15, 5.15it/s] 56%|█████▌ | 26900/48008 [1:32:14<1:08:15, 5.15it/s]{'loss': 0.3003, 'grad_norm': 0.1884765625, 'learning_rate': 8.793951008165306e-05, 'epoch': 0.56} 56%|█████▌ | 26901/48008 [1:32:14<1:08:01, 5.17it/s] 56%|█████▌ | 26902/48008 [1:32:14<1:08:04, 5.17it/s] 56%|█████▌ | 26903/48008 [1:32:14<1:07:07, 5.24it/s] 56%|█████▌ | 26904/48008 [1:32:14<1:07:43, 5.19it/s] 56%|█████▌ | 26905/48008 [1:32:15<1:07:17, 5.23it/s] 56%|█████▌ | 26906/48008 [1:32:15<1:07:43, 5.19it/s] 56%|█████▌ | 26907/48008 [1:32:15<1:07:33, 5.21it/s] 56%|█████▌ | 26908/48008 [1:32:15<1:07:23, 5.22it/s] 56%|█████▌ | 26909/48008 [1:32:15<1:08:00, 5.17it/s] 56%|█████▌ | 26910/48008 [1:32:16<1:08:20, 5.14it/s] 56%|█████▌ | 26911/48008 [1:32:16<1:08:38, 5.12it/s] 56%|█████▌ | 26912/48008 [1:32:16<1:08:09, 5.16it/s] 56%|█████▌ | 26913/48008 [1:32:16<1:28:20, 3.98it/s] 56%|█████▌ | 26914/48008 [1:32:17<1:22:34, 4.26it/s] 56%|█████▌ | 26915/48008 [1:32:17<1:17:17, 4.55it/s] 56%|█████▌ | 26916/48008 [1:32:17<1:13:52, 4.76it/s] 56%|█████▌ | 26917/48008 [1:32:17<1:11:46, 4.90it/s] 56%|█████▌ | 26918/48008 [1:32:17<1:10:34, 4.98it/s] 56%|█████▌ | 26919/48008 [1:32:17<1:10:02, 5.02it/s] 56%|█████▌ | 26920/48008 [1:32:18<1:10:28, 4.99it/s] 56%|█████▌ | 26921/48008 [1:32:18<1:09:06, 5.09it/s] 56%|█████▌ | 26922/48008 [1:32:18<1:08:10, 5.16it/s] 56%|█████▌ | 26923/48008 [1:32:18<1:08:29, 5.13it/s] 56%|█████▌ | 26924/48008 [1:32:18<1:08:47, 5.11it/s] 56%|█████▌ | 26925/48008 [1:32:19<1:08:33, 5.13it/s] 56%|█████▌ | 26926/48008 [1:32:19<1:09:20, 5.07it/s] 56%|█████▌ | 26927/48008 [1:32:19<1:08:36, 5.12it/s] 56%|█████▌ | 26928/48008 [1:32:19<1:08:29, 5.13it/s] 56%|█████▌ | 26929/48008 [1:32:19<1:13:14, 4.80it/s] 56%|█████▌ | 26930/48008 [1:32:20<1:12:40, 4.83it/s] 56%|█████▌ | 26931/48008 [1:32:20<1:11:22, 4.92it/s] 56%|█████▌ | 26932/48008 [1:32:20<1:09:53, 5.03it/s] 56%|█████▌ | 26933/48008 [1:32:20<1:14:26, 4.72it/s] 56%|█████▌ | 26934/48008 [1:32:20<1:13:19, 4.79it/s] 56%|█████▌ | 26935/48008 [1:32:21<1:10:27, 4.98it/s] 56%|█████▌ | 26936/48008 [1:32:21<1:09:29, 5.05it/s] 56%|█████▌ | 26937/48008 [1:32:21<1:08:19, 5.14it/s] 56%|█████▌ | 26938/48008 [1:32:21<1:08:38, 5.12it/s] 56%|█████▌ | 26939/48008 [1:32:21<1:08:40, 5.11it/s] 56%|█████▌ | 26940/48008 [1:32:22<1:28:39, 3.96it/s] 56%|█████▌ | 26941/48008 [1:32:22<1:20:56, 4.34it/s] 56%|█████▌ | 26942/48008 [1:32:22<1:17:11, 4.55it/s] 56%|█████▌ | 26943/48008 [1:32:22<1:14:11, 4.73it/s] 56%|█████▌ | 26944/48008 [1:32:23<1:16:56, 4.56it/s] 56%|█████▌ | 26945/48008 [1:32:23<1:15:02, 4.68it/s] 56%|█████▌ | 26946/48008 [1:32:23<1:12:32, 4.84it/s] 56%|█████▌ | 26947/48008 [1:32:23<1:10:37, 4.97it/s] 56%|█████▌ | 26948/48008 [1:32:23<1:09:33, 5.05it/s] 56%|█████▌ | 26949/48008 [1:32:24<1:08:52, 5.10it/s] 56%|█████▌ | 26950/48008 [1:32:24<1:08:44, 5.11it/s] {'loss': 0.3008, 'grad_norm': 0.1787109375, 'learning_rate': 8.773121146475588e-05, 'epoch': 0.56} 56%|█████▌ | 26950/48008 [1:32:24<1:08:44, 5.11it/s] 56%|█████▌ | 26951/48008 [1:32:24<1:07:55, 5.17it/s] 56%|█████▌ | 26952/48008 [1:32:24<1:07:01, 5.24it/s] 56%|█████▌ | 26953/48008 [1:32:24<1:08:52, 5.10it/s] 56%|█████▌ | 26954/48008 [1:32:25<1:08:18, 5.14it/s] 56%|█████▌ | 26955/48008 [1:32:25<1:28:12, 3.98it/s] 56%|█████▌ | 26956/48008 [1:32:25<1:24:03, 4.17it/s] 56%|█████▌ | 26957/48008 [1:32:25<1:20:08, 4.38it/s] 56%|█████▌ | 26958/48008 [1:32:26<1:16:03, 4.61it/s] 56%|█████▌ | 26959/48008 [1:32:26<1:13:14, 4.79it/s] 56%|█████▌ | 26960/48008 [1:32:26<1:11:33, 4.90it/s] 56%|█████▌ | 26961/48008 [1:32:26<1:10:58, 4.94it/s] 56%|█████▌ | 26962/48008 [1:32:26<1:09:34, 5.04it/s] 56%|█████▌ | 26963/48008 [1:32:27<1:09:11, 5.07it/s] 56%|█████▌ | 26964/48008 [1:32:27<1:07:50, 5.17it/s] 56%|█████▌ | 26965/48008 [1:32:27<1:07:01, 5.23it/s] 56%|█████▌ | 26966/48008 [1:32:27<1:06:35, 5.27it/s] 56%|█████▌ | 26967/48008 [1:32:27<1:07:35, 5.19it/s] 56%|█████▌ | 26968/48008 [1:32:27<1:06:46, 5.25it/s] 56%|█████▌ | 26969/48008 [1:32:28<1:07:11, 5.22it/s] 56%|█████▌ | 26970/48008 [1:32:28<1:08:37, 5.11it/s] 56%|█████▌ | 26971/48008 [1:32:28<1:08:15, 5.14it/s] 56%|█████▌ | 26972/48008 [1:32:28<1:07:58, 5.16it/s] 56%|█████▌ | 26973/48008 [1:32:28<1:07:53, 5.16it/s] 56%|█████▌ | 26974/48008 [1:32:29<1:08:01, 5.15it/s] 56%|█████▌ | 26975/48008 [1:32:29<1:07:43, 5.18it/s] 56%|█████▌ | 26976/48008 [1:32:29<1:12:34, 4.83it/s] 56%|█████▌ | 26977/48008 [1:32:29<1:10:50, 4.95it/s] 56%|█████▌ | 26978/48008 [1:32:29<1:09:46, 5.02it/s] 56%|█████▌ | 26979/48008 [1:32:30<1:08:56, 5.08it/s] 56%|█████▌ | 26980/48008 [1:32:30<1:08:21, 5.13it/s] 56%|█████▌ | 26981/48008 [1:32:30<1:07:47, 5.17it/s] 56%|█████▌ | 26982/48008 [1:32:30<1:06:40, 5.26it/s] 56%|█████▌ | 26983/48008 [1:32:30<1:07:38, 5.18it/s] 56%|█████▌ | 26984/48008 [1:32:31<1:08:05, 5.15it/s] 56%|█████▌ | 26985/48008 [1:32:31<1:09:05, 5.07it/s] 56%|█████▌ | 26986/48008 [1:32:31<1:13:13, 4.78it/s] 56%|█████▌ | 26987/48008 [1:32:31<1:11:32, 4.90it/s] 56%|█████▌ | 26988/48008 [1:32:31<1:10:25, 4.97it/s] 56%|█████▌ | 26989/48008 [1:32:32<1:09:05, 5.07it/s] 56%|█████▌ | 26990/48008 [1:32:32<1:08:50, 5.09it/s] 56%|█████▌ | 26991/48008 [1:32:32<1:08:04, 5.15it/s] 56%|█████▌ | 26992/48008 [1:32:32<1:07:26, 5.19it/s] 56%|█████▌ | 26993/48008 [1:32:32<1:07:23, 5.20it/s] 56%|█████▌ | 26994/48008 [1:32:33<1:11:59, 4.86it/s] 56%|█████▌ | 26995/48008 [1:32:33<1:15:23, 4.65it/s] 56%|█████▌ | 26996/48008 [1:32:33<1:12:42, 4.82it/s] 56%|█████▌ | 26997/48008 [1:32:33<1:10:16, 4.98it/s] 56%|█████▌ | 26998/48008 [1:32:34<1:21:20, 4.30it/s] 56%|█████▌ | 26999/48008 [1:32:34<1:16:25, 4.58it/s] 56%|█████▌ | 27000/48008 [1:32:34<1:13:28, 4.77it/s] 56%|█████▌ | 27000/48008 [1:32:34<1:13:28, 4.77it/s]{'loss': 0.3004, 'grad_norm': 0.279296875, 'learning_rate': 8.75229128478587e-05, 'epoch': 0.56} 56%|█████▌ | 27001/48008 [1:32:34<1:11:26, 4.90it/s] 56%|█████▌ | 27002/48008 [1:32:34<1:10:07, 4.99it/s] 56%|█████▌ | 27003/48008 [1:32:34<1:08:53, 5.08it/s] 56%|█████▌ | 27004/48008 [1:32:35<1:08:19, 5.12it/s] 56%|█████▋ | 27005/48008 [1:32:35<1:07:41, 5.17it/s] 56%|█████▋ | 27006/48008 [1:32:35<1:07:26, 5.19it/s] 56%|█████▋ | 27007/48008 [1:32:35<1:07:07, 5.21it/s] 56%|█████▋ | 27008/48008 [1:32:35<1:07:07, 5.21it/s] 56%|█████▋ | 27009/48008 [1:32:36<1:06:53, 5.23it/s] 56%|█████▋ | 27010/48008 [1:32:36<1:06:32, 5.26it/s] 56%|█████▋ | 27011/48008 [1:32:36<1:07:20, 5.20it/s] 56%|█████▋ | 27012/48008 [1:32:36<1:08:19, 5.12it/s] 56%|█████▋ | 27013/48008 [1:32:37<2:31:37, 2.31it/s] 56%|█████▋ | 27014/48008 [1:32:37<2:05:56, 2.78it/s] 56%|█████▋ | 27015/48008 [1:32:38<1:50:03, 3.18it/s] 56%|█████▋ | 27016/48008 [1:32:38<1:37:15, 3.60it/s] 56%|█████▋ | 27017/48008 [1:32:38<1:28:21, 3.96it/s] 56%|█████▋ | 27018/48008 [1:32:38<1:23:51, 4.17it/s] 56%|█████▋ | 27019/48008 [1:32:39<2:27:56, 2.36it/s] 56%|█████▋ | 27020/48008 [1:32:39<2:15:45, 2.58it/s] 56%|█████▋ | 27021/48008 [1:32:40<1:54:51, 3.05it/s] 56%|█████▋ | 27022/48008 [1:32:40<1:40:25, 3.48it/s] 56%|█████▋ | 27023/48008 [1:32:40<1:30:47, 3.85it/s] 56%|█████▋ | 27024/48008 [1:32:40<1:23:22, 4.19it/s] 56%|█████▋ | 27025/48008 [1:32:40<1:19:33, 4.40it/s] 56%|█████▋ | 27026/48008 [1:32:40<1:15:43, 4.62it/s] 56%|█████▋ | 27027/48008 [1:32:41<1:13:07, 4.78it/s] 56%|█████▋ | 27028/48008 [1:32:41<1:12:34, 4.82it/s] 56%|█████▋ | 27029/48008 [1:32:41<1:10:55, 4.93it/s] 56%|█████▋ | 27030/48008 [1:32:41<1:10:57, 4.93it/s] 56%|█████▋ | 27031/48008 [1:32:41<1:10:32, 4.96it/s] 56%|█████▋ | 27032/48008 [1:32:42<1:09:02, 5.06it/s] 56%|█████▋ | 27033/48008 [1:32:42<1:08:19, 5.12it/s] 56%|█████▋ | 27034/48008 [1:32:42<1:07:32, 5.18it/s] 56%|█████▋ | 27035/48008 [1:32:42<1:07:21, 5.19it/s] 56%|█████▋ | 27036/48008 [1:32:42<1:06:50, 5.23it/s] 56%|█████▋ | 27037/48008 [1:32:43<1:07:43, 5.16it/s] 56%|█████▋ | 27038/48008 [1:32:43<1:07:53, 5.15it/s] 56%|█████▋ | 27039/48008 [1:32:43<1:09:31, 5.03it/s] 56%|█████▋ | 27040/48008 [1:32:43<1:08:47, 5.08it/s] 56%|█████▋ | 27041/48008 [1:32:43<1:08:05, 5.13it/s] 56%|█████▋ | 27042/48008 [1:32:44<1:08:58, 5.07it/s] 56%|█████▋ | 27043/48008 [1:32:44<1:08:24, 5.11it/s] 56%|█████▋ | 27044/48008 [1:32:44<1:07:42, 5.16it/s] 56%|█████▋ | 27045/48008 [1:32:44<1:08:08, 5.13it/s] 56%|█████▋ | 27046/48008 [1:32:44<1:12:32, 4.82it/s] 56%|█████▋ | 27047/48008 [1:32:45<1:10:46, 4.94it/s] 56%|█████▋ | 27048/48008 [1:32:45<1:10:21, 4.97it/s] 56%|█████▋ | 27049/48008 [1:32:45<1:08:56, 5.07it/s] 56%|█████▋ | 27050/48008 [1:32:45<1:09:35, 5.02it/s] {'loss': 0.2992, 'grad_norm': 0.162109375, 'learning_rate': 8.731461423096151e-05, 'epoch': 0.56} 56%|█████▋ | 27050/48008 [1:32:45<1:09:35, 5.02it/s] 56%|█████▋ | 27051/48008 [1:32:45<1:08:20, 5.11it/s] 56%|█████▋ | 27052/48008 [1:32:46<1:08:33, 5.09it/s] 56%|█████▋ | 27053/48008 [1:32:46<1:09:23, 5.03it/s] 56%|█████▋ | 27054/48008 [1:32:46<1:09:49, 5.00it/s] 56%|█████▋ | 27055/48008 [1:32:47<2:16:11, 2.56it/s] 56%|█████▋ | 27056/48008 [1:32:47<1:55:11, 3.03it/s] 56%|█████▋ | 27057/48008 [1:32:47<1:42:35, 3.40it/s] 56%|█████▋ | 27058/48008 [1:32:47<1:31:36, 3.81it/s] 56%|█████▋ | 27059/48008 [1:32:48<1:24:00, 4.16it/s] 56%|█████▋ | 27060/48008 [1:32:48<1:18:13, 4.46it/s] 56%|█████▋ | 27061/48008 [1:32:48<1:15:55, 4.60it/s] 56%|█████▋ | 27062/48008 [1:32:48<1:25:06, 4.10it/s] 56%|█████▋ | 27063/48008 [1:32:49<1:20:19, 4.35it/s] 56%|█████▋ | 27064/48008 [1:32:49<1:16:03, 4.59it/s] 56%|█████▋ | 27065/48008 [1:32:49<1:12:31, 4.81it/s] 56%|█████▋ | 27066/48008 [1:32:49<1:10:52, 4.92it/s] 56%|█████▋ | 27067/48008 [1:32:49<1:09:00, 5.06it/s] 56%|█████▋ | 27068/48008 [1:32:49<1:08:30, 5.09it/s] 56%|█████▋ | 27069/48008 [1:32:50<1:07:30, 5.17it/s] 56%|█████▋ | 27070/48008 [1:32:50<1:11:49, 4.86it/s] 56%|█████▋ | 27071/48008 [1:32:50<1:10:20, 4.96it/s] 56%|█████▋ | 27072/48008 [1:32:50<1:09:24, 5.03it/s] 56%|█████▋ | 27073/48008 [1:32:50<1:09:10, 5.04it/s] 56%|█████▋ | 27074/48008 [1:32:51<1:08:58, 5.06it/s] 56%|█████▋ | 27075/48008 [1:32:51<1:07:38, 5.16it/s] 56%|█████▋ | 27076/48008 [1:32:51<1:07:11, 5.19it/s] 56%|█████▋ | 27077/48008 [1:32:51<1:06:58, 5.21it/s] 56%|█████▋ | 27078/48008 [1:32:51<1:08:04, 5.12it/s] 56%|█████▋ | 27079/48008 [1:32:52<1:07:49, 5.14it/s] 56%|█████▋ | 27080/48008 [1:32:52<1:19:35, 4.38it/s] 56%|█████▋ | 27081/48008 [1:32:52<1:15:14, 4.64it/s] 56%|█████▋ | 27082/48008 [1:32:52<1:13:19, 4.76it/s] 56%|█████▋ | 27083/48008 [1:32:53<1:13:07, 4.77it/s] 56%|█████▋ | 27084/48008 [1:32:53<1:10:46, 4.93it/s] 56%|█████▋ | 27085/48008 [1:32:53<1:11:31, 4.88it/s] 56%|█████▋ | 27086/48008 [1:32:53<1:30:22, 3.86it/s] 56%|█████▋ | 27087/48008 [1:32:53<1:23:11, 4.19it/s] 56%|█████▋ | 27088/48008 [1:32:54<2:30:20, 2.32it/s] 56%|█████▋ | 27089/48008 [1:32:55<2:04:53, 2.79it/s] 56%|█████▋ | 27090/48008 [1:32:55<1:48:38, 3.21it/s] 56%|█████▋ | 27091/48008 [1:32:55<1:37:50, 3.56it/s] 56%|█████▋ | 27092/48008 [1:32:55<1:28:18, 3.95it/s] 56%|█████▋ | 27093/48008 [1:32:55<1:22:55, 4.20it/s] 56%|█████▋ | 27094/48008 [1:32:56<1:18:32, 4.44it/s] 56%|█████▋ | 27095/48008 [1:32:56<1:14:52, 4.66it/s] 56%|█████▋ | 27096/48008 [1:32:56<1:12:16, 4.82it/s] 56%|█████▋ | 27097/48008 [1:32:56<1:11:06, 4.90it/s] 56%|█████▋ | 27098/48008 [1:32:56<1:09:13, 5.03it/s] 56%|█████▋ | 27099/48008 [1:32:57<1:08:30, 5.09it/s] 56%|█████▋ | 27100/48008 [1:32:57<1:09:01, 5.05it/s] {'loss': 0.2993, 'grad_norm': 0.169921875, 'learning_rate': 8.710631561406433e-05, 'epoch': 0.56} 56%|█████▋ | 27100/48008 [1:32:57<1:09:01, 5.05it/s] 56%|█████▋ | 27101/48008 [1:32:57<1:09:27, 5.02it/s] 56%|█████▋ | 27102/48008 [1:32:57<1:08:25, 5.09it/s] 56%|█████▋ | 27103/48008 [1:32:57<1:08:54, 5.06it/s] 56%|█████▋ | 27104/48008 [1:32:57<1:07:55, 5.13it/s] 56%|█████▋ | 27105/48008 [1:32:58<1:07:10, 5.19it/s] 56%|█████▋ | 27106/48008 [1:32:58<1:07:06, 5.19it/s] 56%|█████▋ | 27107/48008 [1:32:58<1:07:02, 5.20it/s] 56%|█████▋ | 27108/48008 [1:32:58<1:07:38, 5.15it/s] 56%|█████▋ | 27109/48008 [1:32:59<1:12:04, 4.83it/s] 56%|█████▋ | 27110/48008 [1:32:59<1:10:57, 4.91it/s] 56%|█████▋ | 27111/48008 [1:32:59<1:09:37, 5.00it/s] 56%|█████▋ | 27112/48008 [1:32:59<1:08:51, 5.06it/s] 56%|█████▋ | 27113/48008 [1:32:59<1:08:05, 5.11it/s] 56%|█████▋ | 27114/48008 [1:32:59<1:07:44, 5.14it/s] 56%|█████▋ | 27115/48008 [1:33:00<1:08:41, 5.07it/s] 56%|█████▋ | 27116/48008 [1:33:00<1:08:21, 5.09it/s] 56%|█████▋ | 27117/48008 [1:33:00<1:07:39, 5.15it/s] 56%|█████▋ | 27118/48008 [1:33:00<1:19:17, 4.39it/s] 56%|█████▋ | 27119/48008 [1:33:01<1:15:29, 4.61it/s] 56%|█████▋ | 27120/48008 [1:33:01<1:12:42, 4.79it/s] 56%|█████▋ | 27121/48008 [1:33:01<1:10:59, 4.90it/s] 56%|█████▋ | 27122/48008 [1:33:01<1:09:46, 4.99it/s] 56%|█████▋ | 27123/48008 [1:33:01<1:09:56, 4.98it/s] 56%|█████▋ | 27124/48008 [1:33:02<1:08:30, 5.08it/s] 57%|█████▋ | 27125/48008 [1:33:02<1:07:16, 5.17it/s] 57%|█████▋ | 27126/48008 [1:33:02<1:06:59, 5.20it/s] 57%|█████▋ | 27127/48008 [1:33:02<1:06:40, 5.22it/s] 57%|█████▋ | 27128/48008 [1:33:02<1:06:15, 5.25it/s] 57%|█████▋ | 27129/48008 [1:33:02<1:06:04, 5.27it/s] 57%|█████▋ | 27130/48008 [1:33:03<1:06:01, 5.27it/s] 57%|█████▋ | 27131/48008 [1:33:03<1:26:45, 4.01it/s] 57%|█████▋ | 27132/48008 [1:33:03<1:20:01, 4.35it/s] 57%|█████▋ | 27133/48008 [1:33:03<1:16:04, 4.57it/s] 57%|█████▋ | 27134/48008 [1:33:04<1:12:39, 4.79it/s] 57%|█████▋ | 27135/48008 [1:33:04<1:11:00, 4.90it/s] 57%|█████▋ | 27136/48008 [1:33:04<1:10:17, 4.95it/s] 57%|█████▋ | 27137/48008 [1:33:04<1:10:39, 4.92it/s] 57%|█████▋ | 27138/48008 [1:33:04<1:09:01, 5.04it/s] 57%|█████▋ | 27139/48008 [1:33:05<1:07:32, 5.15it/s] 57%|█████▋ | 27140/48008 [1:33:05<1:06:53, 5.20it/s] 57%|█████▋ | 27141/48008 [1:33:05<1:19:01, 4.40it/s] 57%|█████▋ | 27142/48008 [1:33:05<1:15:43, 4.59it/s] 57%|█████▋ | 27143/48008 [1:33:05<1:12:51, 4.77it/s] 57%|█████▋ | 27144/48008 [1:33:06<1:12:06, 4.82it/s] 57%|█████▋ | 27145/48008 [1:33:06<1:10:29, 4.93it/s] 57%|█████▋ | 27146/48008 [1:33:06<1:21:15, 4.28it/s] 57%|█████▋ | 27147/48008 [1:33:06<1:16:55, 4.52it/s] 57%|█████▋ | 27148/48008 [1:33:07<1:15:05, 4.63it/s] 57%|█████▋ | 27149/48008 [1:33:07<1:12:54, 4.77it/s] 57%|█████▋ | 27150/48008 [1:33:07<1:10:34, 4.93it/s] 57%|█████▋ | 27150/48008 [1:33:07<1:10:34, 4.93it/s]{'loss': 0.2991, 'grad_norm': 0.2490234375, 'learning_rate': 8.689801699716714e-05, 'epoch': 0.57} 57%|█████▋ | 27151/48008 [1:33:07<1:09:25, 5.01it/s] 57%|█████▋ | 27152/48008 [1:33:07<1:09:42, 4.99it/s] 57%|█████▋ | 27153/48008 [1:33:08<1:08:59, 5.04it/s] 57%|█████▋ | 27154/48008 [1:33:08<1:08:24, 5.08it/s] 57%|█████▋ | 27155/48008 [1:33:08<1:07:42, 5.13it/s] 57%|█████▋ | 27156/48008 [1:33:08<1:07:24, 5.16it/s] 57%|█████▋ | 27157/48008 [1:33:08<1:07:05, 5.18it/s] 57%|█████▋ | 27158/48008 [1:33:08<1:06:45, 5.21it/s] 57%|█████▋ | 27159/48008 [1:33:09<1:06:53, 5.19it/s] 57%|█████▋ | 27160/48008 [1:33:09<1:06:49, 5.20it/s] 57%|█████▋ | 27161/48008 [1:33:09<1:08:38, 5.06it/s] 57%|█████▋ | 27162/48008 [1:33:09<1:09:46, 4.98it/s] 57%|█████▋ | 27163/48008 [1:33:09<1:08:43, 5.06it/s] 57%|█████▋ | 27164/48008 [1:33:10<1:09:24, 5.01it/s] 57%|█████▋ | 27165/48008 [1:33:10<1:08:15, 5.09it/s] 57%|█████▋ | 27166/48008 [1:33:10<1:07:22, 5.16it/s] 57%|█████▋ | 27167/48008 [1:33:10<1:07:06, 5.18it/s] 57%|█████▋ | 27168/48008 [1:33:10<1:07:10, 5.17it/s] 57%|█████▋ | 27169/48008 [1:33:11<1:27:24, 3.97it/s] 57%|█████▋ | 27170/48008 [1:33:11<1:22:09, 4.23it/s] 57%|█████▋ | 27171/48008 [1:33:11<1:37:39, 3.56it/s] 57%|█████▋ | 27172/48008 [1:33:12<1:27:44, 3.96it/s] 57%|█████▋ | 27173/48008 [1:33:12<1:21:31, 4.26it/s] 57%|█████▋ | 27174/48008 [1:33:12<1:18:16, 4.44it/s] 57%|█████▋ | 27175/48008 [1:33:12<1:15:20, 4.61it/s] 57%|█████▋ | 27176/48008 [1:33:12<1:13:44, 4.71it/s] 57%|█████▋ | 27177/48008 [1:33:13<1:12:49, 4.77it/s] 57%|█████▋ | 27178/48008 [1:33:13<1:10:49, 4.90it/s] 57%|█████▋ | 27179/48008 [1:33:13<1:10:49, 4.90it/s] 57%|█████▋ | 27180/48008 [1:33:13<1:09:34, 4.99it/s] 57%|█████▋ | 27181/48008 [1:33:13<1:08:27, 5.07it/s] 57%|█████▋ | 27182/48008 [1:33:14<1:12:44, 4.77it/s] 57%|█████▋ | 27183/48008 [1:33:14<1:12:40, 4.78it/s] 57%|█████▋ | 27184/48008 [1:33:14<1:12:03, 4.82it/s] 57%|█████▋ | 27185/48008 [1:33:14<1:10:20, 4.93it/s] 57%|█████▋ | 27186/48008 [1:33:14<1:09:02, 5.03it/s] 57%|█████▋ | 27187/48008 [1:33:15<1:07:59, 5.10it/s] 57%|█████▋ | 27188/48008 [1:33:15<1:07:37, 5.13it/s] 57%|█████▋ | 27189/48008 [1:33:15<1:06:53, 5.19it/s] 57%|█████▋ | 27190/48008 [1:33:15<1:08:29, 5.07it/s] 57%|█████▋ | 27191/48008 [1:33:15<1:06:54, 5.19it/s] 57%|█████▋ | 27192/48008 [1:33:16<1:06:36, 5.21it/s] 57%|█████▋ | 27193/48008 [1:33:16<1:06:10, 5.24it/s] 57%|█████▋ | 27194/48008 [1:33:16<1:06:51, 5.19it/s] 57%|█████▋ | 27195/48008 [1:33:16<1:06:01, 5.25it/s] 57%|█████▋ | 27196/48008 [1:33:16<1:05:57, 5.26it/s] 57%|█████▋ | 27197/48008 [1:33:16<1:05:31, 5.29it/s] 57%|█████▋ | 27198/48008 [1:33:17<1:07:13, 5.16it/s] 57%|█████▋ | 27199/48008 [1:33:17<1:06:41, 5.20it/s] 57%|█████▋ | 27200/48008 [1:33:17<1:08:27, 5.07it/s] {'loss': 0.2995, 'grad_norm': 0.59375, 'learning_rate': 8.668971838026995e-05, 'epoch': 0.57} 57%|█████▋ | 27200/48008 [1:33:17<1:08:27, 5.07it/s] 57%|█████▋ | 27201/48008 [1:33:17<1:08:28, 5.06it/s] 57%|█████▋ | 27202/48008 [1:33:17<1:08:01, 5.10it/s] 57%|█████▋ | 27203/48008 [1:33:18<1:12:24, 4.79it/s] 57%|█████▋ | 27204/48008 [1:33:18<1:11:15, 4.87it/s] 57%|█████▋ | 27205/48008 [1:33:18<1:10:01, 4.95it/s] 57%|█████▋ | 27206/48008 [1:33:18<1:10:37, 4.91it/s] 57%|█████▋ | 27207/48008 [1:33:19<1:09:04, 5.02it/s] 57%|█████▋ | 27208/48008 [1:33:19<1:08:02, 5.10it/s] 57%|█████▋ | 27209/48008 [1:33:19<1:07:19, 5.15it/s] 57%|█████▋ | 27210/48008 [1:33:19<1:08:00, 5.10it/s] 57%|█████▋ | 27211/48008 [1:33:19<1:06:47, 5.19it/s] 57%|█████▋ | 27212/48008 [1:33:19<1:06:09, 5.24it/s] 57%|█████▋ | 27213/48008 [1:33:20<1:06:02, 5.25it/s] 57%|█████▋ | 27214/48008 [1:33:20<1:06:16, 5.23it/s] 57%|█████▋ | 27215/48008 [1:33:20<1:06:28, 5.21it/s] 57%|█████▋ | 27216/48008 [1:33:20<1:06:20, 5.22it/s] 57%|█████▋ | 27217/48008 [1:33:20<1:06:03, 5.25it/s] 57%|█████▋ | 27218/48008 [1:33:21<1:05:51, 5.26it/s] 57%|█████▋ | 27219/48008 [1:33:21<1:06:01, 5.25it/s] 57%|█████▋ | 27220/48008 [1:33:21<1:06:15, 5.23it/s] 57%|█████▋ | 27221/48008 [1:33:21<1:07:38, 5.12it/s] 57%|█████▋ | 27222/48008 [1:33:21<1:06:56, 5.17it/s] 57%|█████▋ | 27223/48008 [1:33:22<1:07:07, 5.16it/s] 57%|█████▋ | 27224/48008 [1:33:22<1:07:22, 5.14it/s] 57%|█████▋ | 27225/48008 [1:33:22<1:07:30, 5.13it/s] 57%|█████▋ | 27226/48008 [1:33:22<1:08:12, 5.08it/s] 57%|█████▋ | 27227/48008 [1:33:22<1:07:31, 5.13it/s] 57%|█████▋ | 27228/48008 [1:33:23<1:07:10, 5.16it/s] 57%|█████▋ | 27229/48008 [1:33:23<1:06:02, 5.24it/s] 57%|█████▋ | 27230/48008 [1:33:23<1:06:09, 5.23it/s] 57%|█████▋ | 27231/48008 [1:33:23<1:07:24, 5.14it/s] 57%|█████▋ | 27232/48008 [1:33:23<1:12:02, 4.81it/s] 57%|█████▋ | 27233/48008 [1:33:24<1:14:58, 4.62it/s] 57%|█████▋ | 27234/48008 [1:33:24<1:12:11, 4.80it/s] 57%|█████▋ | 27235/48008 [1:33:24<1:10:37, 4.90it/s] 57%|█████▋ | 27236/48008 [1:33:24<1:11:14, 4.86it/s] 57%|█████▋ | 27237/48008 [1:33:24<1:10:46, 4.89it/s] 57%|█████▋ | 27238/48008 [1:33:25<1:09:24, 4.99it/s] 57%|█████▋ | 27239/48008 [1:33:25<1:08:24, 5.06it/s] 57%|█████▋ | 27240/48008 [1:33:25<1:07:43, 5.11it/s] 57%|█████▋ | 27241/48008 [1:33:25<1:08:52, 5.02it/s] 57%|█████▋ | 27242/48008 [1:33:25<1:08:21, 5.06it/s] 57%|█████▋ | 27243/48008 [1:33:26<1:07:44, 5.11it/s] 57%|█████▋ | 27244/48008 [1:33:26<1:07:08, 5.15it/s] 57%|█████▋ | 27245/48008 [1:33:26<1:18:52, 4.39it/s] 57%|█████▋ | 27246/48008 [1:33:26<1:14:47, 4.63it/s] 57%|█████▋ | 27247/48008 [1:33:26<1:12:25, 4.78it/s] 57%|█████▋ | 27248/48008 [1:33:27<1:10:40, 4.90it/s] 57%|█████▋ | 27249/48008 [1:33:27<1:10:37, 4.90it/s] 57%|█████▋ | 27250/48008 [1:33:27<1:09:49, 4.95it/s] {'loss': 0.2996, 'grad_norm': 0.302734375, 'learning_rate': 8.648141976337277e-05, 'epoch': 0.57} 57%|█████▋ | 27250/48008 [1:33:27<1:09:49, 4.95it/s] 57%|█████▋ | 27251/48008 [1:33:27<1:09:48, 4.96it/s] 57%|█████▋ | 27252/48008 [1:33:27<1:08:24, 5.06it/s] 57%|█████▋ | 27253/48008 [1:33:28<1:12:28, 4.77it/s] 57%|█████▋ | 27254/48008 [1:33:28<1:10:27, 4.91it/s] 57%|█████▋ | 27255/48008 [1:33:28<1:08:24, 5.06it/s] 57%|█████▋ | 27256/48008 [1:33:28<1:07:26, 5.13it/s] 57%|█████▋ | 27257/48008 [1:33:28<1:07:04, 5.16it/s] 57%|█████▋ | 27258/48008 [1:33:29<1:07:22, 5.13it/s] 57%|█████▋ | 27259/48008 [1:33:29<1:08:58, 5.01it/s] 57%|█████▋ | 27260/48008 [1:33:29<1:07:24, 5.13it/s] 57%|█████▋ | 27261/48008 [1:33:29<1:07:48, 5.10it/s] 57%|█████▋ | 27262/48008 [1:33:29<1:07:23, 5.13it/s] 57%|█████▋ | 27263/48008 [1:33:30<1:11:45, 4.82it/s] 57%|█████▋ | 27264/48008 [1:33:30<1:09:51, 4.95it/s] 57%|█████▋ | 27265/48008 [1:33:30<1:08:29, 5.05it/s] 57%|█████▋ | 27266/48008 [1:33:30<1:08:45, 5.03it/s] 57%|█████▋ | 27267/48008 [1:33:30<1:12:37, 4.76it/s] 57%|█████▋ | 27268/48008 [1:33:31<1:10:05, 4.93it/s] 57%|█████▋ | 27269/48008 [1:33:31<1:08:54, 5.02it/s] 57%|█████▋ | 27270/48008 [1:33:31<1:08:39, 5.03it/s] 57%|█████▋ | 27271/48008 [1:33:31<1:07:30, 5.12it/s] 57%|█████▋ | 27272/48008 [1:33:31<1:07:11, 5.14it/s] 57%|█████▋ | 27273/48008 [1:33:32<1:06:36, 5.19it/s] 57%|█████▋ | 27274/48008 [1:33:32<1:06:19, 5.21it/s] 57%|█████▋ | 27275/48008 [1:33:32<1:05:48, 5.25it/s] 57%|█████▋ | 27276/48008 [1:33:32<1:06:24, 5.20it/s] 57%|█████▋ | 27277/48008 [1:33:32<1:05:11, 5.30it/s] 57%|█████▋ | 27278/48008 [1:33:33<1:05:21, 5.29it/s] 57%|█████▋ | 27279/48008 [1:33:33<1:05:40, 5.26it/s] 57%|█████▋ | 27280/48008 [1:33:33<1:05:42, 5.26it/s] 57%|█████▋ | 27281/48008 [1:33:33<1:05:52, 5.24it/s] 57%|█████▋ | 27282/48008 [1:33:33<1:05:48, 5.25it/s] 57%|█████▋ | 27283/48008 [1:33:34<1:07:11, 5.14it/s] 57%|█████▋ | 27284/48008 [1:33:34<1:07:38, 5.11it/s] 57%|█████▋ | 27285/48008 [1:33:34<1:06:54, 5.16it/s] 57%|█████▋ | 27286/48008 [1:33:34<1:06:35, 5.19it/s] 57%|█████▋ | 27287/48008 [1:33:34<1:18:21, 4.41it/s] 57%|█████▋ | 27288/48008 [1:33:35<1:14:19, 4.65it/s] 57%|█████▋ | 27289/48008 [1:33:35<1:13:03, 4.73it/s] 57%|█████▋ | 27290/48008 [1:33:35<1:10:34, 4.89it/s] 57%|█████▋ | 27291/48008 [1:33:35<1:09:54, 4.94it/s] 57%|█████▋ | 27292/48008 [1:33:35<1:08:47, 5.02it/s] 57%|█████▋ | 27293/48008 [1:33:36<1:12:35, 4.76it/s] 57%|█████▋ | 27294/48008 [1:33:36<1:22:28, 4.19it/s] 57%|█████▋ | 27295/48008 [1:33:36<1:17:20, 4.46it/s] 57%|█████▋ | 27296/48008 [1:33:36<1:13:57, 4.67it/s] 57%|█████▋ | 27297/48008 [1:33:36<1:11:20, 4.84it/s] 57%|█████▋ | 27298/48008 [1:33:37<1:10:00, 4.93it/s] 57%|█████▋ | 27299/48008 [1:33:37<1:10:51, 4.87it/s] 57%|█████▋ | 27300/48008 [1:33:37<1:10:07, 4.92it/s] {'loss': 0.3005, 'grad_norm': 0.2216796875, 'learning_rate': 8.627312114647559e-05, 'epoch': 0.57} 57%|█████▋ | 27300/48008 [1:33:37<1:10:07, 4.92it/s] 57%|█████▋ | 27301/48008 [1:33:37<1:08:37, 5.03it/s] 57%|█████▋ | 27302/48008 [1:33:37<1:07:40, 5.10it/s] 57%|█████▋ | 27303/48008 [1:33:38<1:08:34, 5.03it/s] 57%|█████▋ | 27304/48008 [1:33:38<1:07:48, 5.09it/s] 57%|█████▋ | 27305/48008 [1:33:38<1:06:59, 5.15it/s] 57%|█████▋ | 27306/48008 [1:33:38<1:06:31, 5.19it/s] 57%|█████▋ | 27307/48008 [1:33:38<1:05:58, 5.23it/s] 57%|█████▋ | 27308/48008 [1:33:39<1:05:54, 5.23it/s] 57%|█████▋ | 27309/48008 [1:33:39<1:05:57, 5.23it/s] 57%|█████▋ | 27310/48008 [1:33:39<1:05:50, 5.24it/s] 57%|█████▋ | 27311/48008 [1:33:39<1:05:28, 5.27it/s] 57%|█████▋ | 27312/48008 [1:33:39<1:17:30, 4.45it/s] 57%|█████▋ | 27313/48008 [1:33:40<1:18:54, 4.37it/s] 57%|█████▋ | 27314/48008 [1:33:40<1:15:08, 4.59it/s] 57%|█████▋ | 27315/48008 [1:33:40<1:12:21, 4.77it/s] 57%|█████▋ | 27316/48008 [1:33:40<1:09:48, 4.94it/s] 57%|█████▋ | 27317/48008 [1:33:40<1:09:10, 4.99it/s] 57%|█████▋ | 27318/48008 [1:33:41<1:08:18, 5.05it/s] 57%|█████▋ | 27319/48008 [1:33:41<1:08:23, 5.04it/s] 57%|█████▋ | 27320/48008 [1:33:41<1:08:11, 5.06it/s] 57%|█████▋ | 27321/48008 [1:33:41<1:06:57, 5.15it/s] 57%|█████▋ | 27322/48008 [1:33:41<1:07:47, 5.09it/s] 57%|█████▋ | 27323/48008 [1:33:42<1:07:14, 5.13it/s] 57%|█████▋ | 27324/48008 [1:33:42<1:06:40, 5.17it/s] 57%|█████▋ | 27325/48008 [1:33:42<1:06:23, 5.19it/s] 57%|█████▋ | 27326/48008 [1:33:42<1:07:02, 5.14it/s] 57%|█████▋ | 27327/48008 [1:33:42<1:07:52, 5.08it/s] 57%|█████▋ | 27328/48008 [1:33:43<1:07:19, 5.12it/s] 57%|█████▋ | 27329/48008 [1:33:43<1:06:25, 5.19it/s] 57%|█████▋ | 27330/48008 [1:33:43<1:06:10, 5.21it/s] 57%|█████▋ | 27331/48008 [1:33:43<1:26:16, 3.99it/s] 57%|█████▋ | 27332/48008 [1:33:44<1:20:13, 4.30it/s] 57%|█████▋ | 27333/48008 [1:33:44<1:15:51, 4.54it/s] 57%|█████▋ | 27334/48008 [1:33:44<1:12:10, 4.77it/s] 57%|█████▋ | 27335/48008 [1:33:44<1:10:55, 4.86it/s] 57%|█████▋ | 27336/48008 [1:33:44<1:08:48, 5.01it/s] 57%|█████▋ | 27337/48008 [1:33:45<1:09:17, 4.97it/s] 57%|█████▋ | 27338/48008 [1:33:45<1:08:28, 5.03it/s] 57%|█████▋ | 27339/48008 [1:33:45<1:09:37, 4.95it/s] 57%|█████▋ | 27340/48008 [1:33:45<1:07:45, 5.08it/s] 57%|█████▋ | 27341/48008 [1:33:45<1:07:00, 5.14it/s] 57%|█████▋ | 27342/48008 [1:33:46<1:06:40, 5.17it/s] 57%|█████▋ | 27343/48008 [1:33:46<1:06:00, 5.22it/s] 57%|█████▋ | 27344/48008 [1:33:46<1:17:46, 4.43it/s] 57%|█████▋ | 27345/48008 [1:33:46<1:13:58, 4.66it/s] 57%|█████▋ | 27346/48008 [1:33:46<1:11:25, 4.82it/s] 57%|█████▋ | 27347/48008 [1:33:47<1:14:33, 4.62it/s] 57%|█████▋ | 27348/48008 [1:33:47<1:32:01, 3.74it/s] 57%|█████▋ | 27349/48008 [1:33:47<1:23:52, 4.10it/s] 57%|█████▋ | 27350/48008 [1:33:47<1:18:27, 4.39it/s] {'loss': 0.2987, 'grad_norm': 0.35546875, 'learning_rate': 8.60648225295784e-05, 'epoch': 0.57} 57%|█████▋ | 27350/48008 [1:33:47<1:18:27, 4.39it/s] 57%|█████▋ | 27351/48008 [1:33:48<1:14:32, 4.62it/s] 57%|█████▋ | 27352/48008 [1:33:48<1:14:04, 4.65it/s] 57%|█████▋ | 27353/48008 [1:33:48<1:11:43, 4.80it/s] 57%|█████▋ | 27354/48008 [1:33:48<1:09:43, 4.94it/s] 57%|█████▋ | 27355/48008 [1:33:48<1:07:28, 5.10it/s] 57%|█████▋ | 27356/48008 [1:33:49<1:06:25, 5.18it/s] 57%|█████▋ | 27357/48008 [1:33:49<1:06:08, 5.20it/s] 57%|█████▋ | 27358/48008 [1:33:49<1:07:24, 5.11it/s] 57%|█████▋ | 27359/48008 [1:33:49<1:06:24, 5.18it/s] 57%|█████▋ | 27360/48008 [1:33:49<1:05:59, 5.22it/s] 57%|█████▋ | 27361/48008 [1:33:50<1:06:25, 5.18it/s] 57%|█████▋ | 27362/48008 [1:33:50<1:06:07, 5.20it/s] 57%|█████▋ | 27363/48008 [1:33:50<1:10:41, 4.87it/s] 57%|█████▋ | 27364/48008 [1:33:50<1:09:01, 4.99it/s] 57%|█████▋ | 27365/48008 [1:33:50<1:07:44, 5.08it/s] 57%|█████▋ | 27366/48008 [1:33:51<1:07:10, 5.12it/s] 57%|█████▋ | 27367/48008 [1:33:51<1:11:30, 4.81it/s] 57%|█████▋ | 27368/48008 [1:33:51<1:09:48, 4.93it/s] 57%|█████▋ | 27369/48008 [1:33:51<1:08:35, 5.01it/s] 57%|█████▋ | 27370/48008 [1:33:51<1:08:56, 4.99it/s] 57%|█████▋ | 27371/48008 [1:33:52<1:08:39, 5.01it/s] 57%|█████▋ | 27372/48008 [1:33:52<1:07:05, 5.13it/s] 57%|█████▋ | 27373/48008 [1:33:52<1:06:20, 5.18it/s] 57%|█████▋ | 27374/48008 [1:33:52<1:07:12, 5.12it/s] 57%|█████▋ | 27375/48008 [1:33:52<1:06:31, 5.17it/s] 57%|█████▋ | 27376/48008 [1:33:52<1:06:23, 5.18it/s] 57%|█████▋ | 27377/48008 [1:33:53<1:08:14, 5.04it/s] 57%|█████▋ | 27378/48008 [1:33:53<1:07:41, 5.08it/s] 57%|█████▋ | 27379/48008 [1:33:53<1:08:52, 4.99it/s] 57%|█████▋ | 27380/48008 [1:33:53<1:19:49, 4.31it/s] 57%|█████▋ | 27381/48008 [1:33:54<1:17:02, 4.46it/s] 57%|█████▋ | 27382/48008 [1:33:54<1:13:28, 4.68it/s] 57%|█████▋ | 27383/48008 [1:33:54<1:10:39, 4.87it/s] 57%|█████▋ | 27384/48008 [1:33:54<1:13:59, 4.65it/s] 57%|█████▋ | 27385/48008 [1:33:54<1:11:34, 4.80it/s] 57%|█████▋ | 27386/48008 [1:33:55<1:09:37, 4.94it/s] 57%|█████▋ | 27387/48008 [1:33:55<1:13:14, 4.69it/s] 57%|█████▋ | 27388/48008 [1:33:55<1:10:28, 4.88it/s] 57%|█████▋ | 27389/48008 [1:33:55<1:10:24, 4.88it/s] 57%|█████▋ | 27390/48008 [1:33:55<1:10:09, 4.90it/s] 57%|█████▋ | 27391/48008 [1:33:56<1:08:34, 5.01it/s] 57%|█████▋ | 27392/48008 [1:33:56<1:09:10, 4.97it/s] 57%|█████▋ | 27393/48008 [1:33:56<1:07:59, 5.05it/s] 57%|█████▋ | 27394/48008 [1:33:56<1:07:19, 5.10it/s] 57%|█████▋ | 27395/48008 [1:33:56<1:06:16, 5.18it/s] 57%|█████▋ | 27396/48008 [1:33:57<1:06:15, 5.18it/s] 57%|█████▋ | 27397/48008 [1:33:57<1:05:31, 5.24it/s] 57%|█████▋ | 27398/48008 [1:33:57<1:25:38, 4.01it/s] 57%|█████▋ | 27399/48008 [1:33:57<1:20:56, 4.24it/s] 57%|█████▋ | 27400/48008 [1:33:58<1:16:22, 4.50it/s] {'loss': 0.3015, 'grad_norm': 0.1982421875, 'learning_rate': 8.585652391268122e-05, 'epoch': 0.57} 57%|█████▋ | 27400/48008 [1:33:58<1:16:22, 4.50it/s] 57%|█████▋ | 27401/48008 [1:33:58<1:17:49, 4.41it/s] 57%|█████▋ | 27402/48008 [1:33:58<1:14:12, 4.63it/s] 57%|█████▋ | 27403/48008 [1:33:58<1:11:08, 4.83it/s] 57%|█████▋ | 27404/48008 [1:33:58<1:09:12, 4.96it/s] 57%|█████▋ | 27405/48008 [1:33:59<1:07:59, 5.05it/s] 57%|█████▋ | 27406/48008 [1:33:59<1:08:32, 5.01it/s] 57%|█████▋ | 27407/48008 [1:33:59<1:07:15, 5.10it/s] 57%|█████▋ | 27408/48008 [1:33:59<1:08:04, 5.04it/s] 57%|█████▋ | 27409/48008 [1:33:59<1:19:03, 4.34it/s] 57%|█████▋ | 27410/48008 [1:34:00<1:14:27, 4.61it/s] 57%|█████▋ | 27411/48008 [1:34:00<1:12:22, 4.74it/s] 57%|█████▋ | 27412/48008 [1:34:00<1:10:21, 4.88it/s] 57%|█████▋ | 27413/48008 [1:34:00<1:08:28, 5.01it/s] 57%|█████▋ | 27414/48008 [1:34:00<1:07:12, 5.11it/s] 57%|█████▋ | 27415/48008 [1:34:01<1:06:39, 5.15it/s] 57%|█████▋ | 27416/48008 [1:34:01<1:07:17, 5.10it/s] 57%|█████▋ | 27417/48008 [1:34:01<1:06:48, 5.14it/s] 57%|█████▋ | 27418/48008 [1:34:01<1:06:35, 5.15it/s] 57%|█████▋ | 27419/48008 [1:34:01<1:08:04, 5.04it/s] 57%|█████▋ | 27420/48008 [1:34:02<1:06:48, 5.14it/s] 57%|█████▋ | 27421/48008 [1:34:02<1:06:19, 5.17it/s] 57%|█████▋ | 27422/48008 [1:34:02<1:05:32, 5.23it/s] 57%|█████▋ | 27423/48008 [1:34:02<1:05:35, 5.23it/s] 57%|█████▋ | 27424/48008 [1:34:02<1:06:46, 5.14it/s] 57%|█████▋ | 27425/48008 [1:34:03<1:06:30, 5.16it/s] 57%|█████▋ | 27426/48008 [1:34:03<1:06:39, 5.15it/s] 57%|█████▋ | 27427/48008 [1:34:03<1:07:21, 5.09it/s] 57%|█████▋ | 27428/48008 [1:34:03<1:06:48, 5.13it/s] 57%|█████▋ | 27429/48008 [1:34:03<1:06:04, 5.19it/s] 57%|█████▋ | 27430/48008 [1:34:03<1:06:21, 5.17it/s] 57%|█████▋ | 27431/48008 [1:34:04<1:06:28, 5.16it/s] 57%|█████▋ | 27432/48008 [1:34:04<1:08:05, 5.04it/s] 57%|█████▋ | 27433/48008 [1:34:04<1:06:19, 5.17it/s] 57%|█████▋ | 27434/48008 [1:34:04<1:05:36, 5.23it/s] 57%|█████▋ | 27435/48008 [1:34:04<1:05:29, 5.24it/s] 57%|█████▋ | 27436/48008 [1:34:05<1:05:18, 5.25it/s] 57%|█████▋ | 27437/48008 [1:34:05<1:07:01, 5.12it/s] 57%|█████▋ | 27438/48008 [1:34:05<1:06:40, 5.14it/s] 57%|█████▋ | 27439/48008 [1:34:05<1:06:11, 5.18it/s] 57%|█████▋ | 27440/48008 [1:34:05<1:06:36, 5.15it/s] 57%|█████▋ | 27441/48008 [1:34:06<1:08:04, 5.04it/s] 57%|█████▋ | 27442/48008 [1:34:06<1:07:21, 5.09it/s] 57%|█████▋ | 27443/48008 [1:34:06<1:06:33, 5.15it/s] 57%|█████▋ | 27444/48008 [1:34:06<1:05:33, 5.23it/s] 57%|█████▋ | 27445/48008 [1:34:06<1:05:22, 5.24it/s] 57%|█████▋ | 27446/48008 [1:34:07<1:05:13, 5.25it/s] 57%|█████▋ | 27447/48008 [1:34:07<1:06:05, 5.18it/s] 57%|█████▋ | 27448/48008 [1:34:07<1:05:42, 5.21it/s] 57%|█████▋ | 27449/48008 [1:34:07<1:05:33, 5.23it/s] 57%|█████▋ | 27450/48008 [1:34:07<1:06:46, 5.13it/s] {'loss': 0.3016, 'grad_norm': 0.5703125, 'learning_rate': 8.564822529578404e-05, 'epoch': 0.57} 57%|█████▋ | 27450/48008 [1:34:07<1:06:46, 5.13it/s] 57%|█████▋ | 27451/48008 [1:34:08<1:05:55, 5.20it/s] 57%|█████▋ | 27452/48008 [1:34:08<1:06:53, 5.12it/s] 57%|█████▋ | 27453/48008 [1:34:08<1:06:37, 5.14it/s] 57%|█████▋ | 27454/48008 [1:34:08<1:26:20, 3.97it/s] 57%|█████▋ | 27455/48008 [1:34:09<1:19:52, 4.29it/s] 57%|█████▋ | 27456/48008 [1:34:09<1:15:37, 4.53it/s] 57%|█████▋ | 27457/48008 [1:34:09<1:12:17, 4.74it/s] 57%|█████▋ | 27458/48008 [1:34:09<1:09:47, 4.91it/s] 57%|█████▋ | 27459/48008 [1:34:09<1:09:28, 4.93it/s] 57%|█████▋ | 27460/48008 [1:34:09<1:09:11, 4.95it/s] 57%|█████▋ | 27461/48008 [1:34:10<1:08:04, 5.03it/s] 57%|█████▋ | 27462/48008 [1:34:10<1:07:00, 5.11it/s] 57%|█████▋ | 27463/48008 [1:34:10<1:07:40, 5.06it/s] 57%|█████▋ | 27464/48008 [1:34:10<1:07:36, 5.06it/s] 57%|█████▋ | 27465/48008 [1:34:11<1:11:30, 4.79it/s] 57%|█████▋ | 27466/48008 [1:34:11<1:09:09, 4.95it/s] 57%|█████▋ | 27467/48008 [1:34:11<1:07:58, 5.04it/s] 57%|█████▋ | 27468/48008 [1:34:11<1:07:09, 5.10it/s] 57%|█████▋ | 27469/48008 [1:34:11<1:07:45, 5.05it/s] 57%|█████▋ | 27470/48008 [1:34:11<1:06:58, 5.11it/s] 57%|█████▋ | 27471/48008 [1:34:12<1:06:08, 5.18it/s] 57%|█████▋ | 27472/48008 [1:34:12<1:05:44, 5.21it/s] 57%|█████▋ | 27473/48008 [1:34:12<1:05:40, 5.21it/s] 57%|█████▋ | 27474/48008 [1:34:12<1:06:33, 5.14it/s] 57%|█████▋ | 27475/48008 [1:34:12<1:05:29, 5.23it/s] 57%|█████▋ | 27476/48008 [1:34:13<1:04:55, 5.27it/s] 57%|█████▋ | 27477/48008 [1:34:13<1:06:06, 5.18it/s] 57%|█████▋ | 27478/48008 [1:34:13<1:05:22, 5.23it/s] 57%|█████▋ | 27479/48008 [1:34:13<1:04:52, 5.27it/s] 57%|█████▋ | 27480/48008 [1:34:13<1:04:53, 5.27it/s] 57%|█████▋ | 27481/48008 [1:34:14<1:03:57, 5.35it/s] 57%|█████▋ | 27482/48008 [1:34:14<1:04:25, 5.31it/s] 57%|█████▋ | 27483/48008 [1:34:14<1:05:50, 5.20it/s] 57%|█████▋ | 27484/48008 [1:34:14<1:05:46, 5.20it/s] 57%|█████▋ | 27485/48008 [1:34:14<1:05:45, 5.20it/s] 57%|█████▋ | 27486/48008 [1:34:15<1:04:36, 5.29it/s] 57%|█████▋ | 27487/48008 [1:34:15<1:04:27, 5.31it/s] 57%|█████▋ | 27488/48008 [1:34:15<1:05:10, 5.25it/s] 57%|█████▋ | 27489/48008 [1:34:15<1:07:02, 5.10it/s] 57%|█████▋ | 27490/48008 [1:34:15<1:06:12, 5.16it/s] 57%|█████▋ | 27491/48008 [1:34:15<1:07:09, 5.09it/s] 57%|█████▋ | 27492/48008 [1:34:16<1:06:35, 5.14it/s] 57%|█████▋ | 27493/48008 [1:34:16<1:05:58, 5.18it/s] 57%|█████▋ | 27494/48008 [1:34:16<1:05:40, 5.21it/s] 57%|█████▋ | 27495/48008 [1:34:16<1:17:20, 4.42it/s] 57%|█████▋ | 27496/48008 [1:34:17<1:14:55, 4.56it/s] 57%|█████▋ | 27497/48008 [1:34:17<1:12:05, 4.74it/s] 57%|█████▋ | 27498/48008 [1:34:17<1:09:15, 4.94it/s] 57%|█████▋ | 27499/48008 [1:34:17<1:09:14, 4.94it/s] 57%|█████▋ | 27500/48008 [1:34:17<1:08:10, 5.01it/s] {'loss': 0.3023, 'grad_norm': 0.2216796875, 'learning_rate': 8.543992667888686e-05, 'epoch': 0.57} 57%|█████▋ | 27500/48008 [1:34:17<1:08:10, 5.01it/s] 57%|█████▋ | 27501/48008 [1:34:18<1:07:25, 5.07it/s] 57%|█████▋ | 27502/48008 [1:34:18<1:06:45, 5.12it/s] 57%|█████▋ | 27503/48008 [1:34:18<1:07:25, 5.07it/s] 57%|█████▋ | 27504/48008 [1:34:18<1:07:32, 5.06it/s] 57%|█████▋ | 27505/48008 [1:34:18<1:08:01, 5.02it/s] 57%|█████▋ | 27506/48008 [1:34:19<1:06:43, 5.12it/s] 57%|█████▋ | 27507/48008 [1:34:19<1:07:00, 5.10it/s] 57%|█████▋ | 27508/48008 [1:34:19<1:06:30, 5.14it/s] 57%|█████▋ | 27509/48008 [1:34:19<1:06:33, 5.13it/s] 57%|█████▋ | 27510/48008 [1:34:19<1:06:04, 5.17it/s] 57%|█████▋ | 27511/48008 [1:34:19<1:06:54, 5.11it/s] 57%|█████▋ | 27512/48008 [1:34:20<1:06:46, 5.12it/s] 57%|█████▋ | 27513/48008 [1:34:20<1:06:18, 5.15it/s] 57%|█████▋ | 27514/48008 [1:34:20<1:05:58, 5.18it/s] 57%|█████▋ | 27515/48008 [1:34:20<1:05:08, 5.24it/s] 57%|█████▋ | 27516/48008 [1:34:20<1:05:35, 5.21it/s] 57%|█████▋ | 27517/48008 [1:34:21<1:06:01, 5.17it/s] 57%|█████▋ | 27518/48008 [1:34:21<1:05:46, 5.19it/s] 57%|█████▋ | 27519/48008 [1:34:21<1:05:45, 5.19it/s] 57%|█████▋ | 27520/48008 [1:34:21<1:05:20, 5.23it/s] 57%|█████▋ | 27521/48008 [1:34:21<1:05:24, 5.22it/s] 57%|█████▋ | 27522/48008 [1:34:22<1:05:11, 5.24it/s] 57%|█████▋ | 27523/48008 [1:34:22<1:06:18, 5.15it/s] 57%|█████▋ | 27524/48008 [1:34:22<1:05:55, 5.18it/s] 57%|█████▋ | 27525/48008 [1:34:22<1:05:34, 5.21it/s] 57%|█████▋ | 27526/48008 [1:34:22<1:07:09, 5.08it/s] 57%|█████▋ | 27527/48008 [1:34:23<1:11:20, 4.79it/s] 57%|█████▋ | 27528/48008 [1:34:23<1:09:28, 4.91it/s] 57%|█████▋ | 27529/48008 [1:34:23<1:08:14, 5.00it/s] 57%|█████▋ | 27530/48008 [1:34:23<1:07:24, 5.06it/s] 57%|█████▋ | 27531/48008 [1:34:23<1:06:42, 5.12it/s] 57%|█████▋ | 27532/48008 [1:34:24<1:06:09, 5.16it/s] 57%|█████▋ | 27533/48008 [1:34:24<1:17:35, 4.40it/s] 57%|█████▋ | 27534/48008 [1:34:24<1:14:28, 4.58it/s] 57%|█████▋ | 27535/48008 [1:34:24<1:11:10, 4.79it/s] 57%|█████▋ | 27536/48008 [1:34:24<1:09:13, 4.93it/s] 57%|█████▋ | 27537/48008 [1:34:25<1:08:26, 4.98it/s] 57%|█████▋ | 27538/48008 [1:34:25<1:07:22, 5.06it/s] 57%|█████▋ | 27539/48008 [1:34:25<1:06:28, 5.13it/s] 57%|█████▋ | 27540/48008 [1:34:25<1:06:03, 5.16it/s] 57%|█████▋ | 27541/48008 [1:34:25<1:05:41, 5.19it/s] 57%|█████▋ | 27542/48008 [1:34:26<1:06:24, 5.14it/s] 57%|█████▋ | 27543/48008 [1:34:26<1:06:07, 5.16it/s] 57%|█████▋ | 27544/48008 [1:34:26<1:06:55, 5.10it/s] 57%|█████▋ | 27545/48008 [1:34:26<1:06:26, 5.13it/s] 57%|█████▋ | 27546/48008 [1:34:26<1:06:14, 5.15it/s] 57%|█████▋ | 27547/48008 [1:34:27<1:05:35, 5.20it/s] 57%|█████▋ | 27548/48008 [1:34:27<1:06:47, 5.11it/s] 57%|█████▋ | 27549/48008 [1:34:27<1:06:18, 5.14it/s] 57%|█████▋ | 27550/48008 [1:34:27<1:10:40, 4.82it/s] {'loss': 0.3009, 'grad_norm': 0.33984375, 'learning_rate': 8.523162806198967e-05, 'epoch': 0.57} 57%|█████▋ | 27550/48008 [1:34:27<1:10:40, 4.82it/s] 57%|█████▋ | 27551/48008 [1:34:27<1:10:51, 4.81it/s] 57%|█████▋ | 27552/48008 [1:34:28<1:09:05, 4.93it/s] 57%|█████▋ | 27553/48008 [1:34:28<1:07:40, 5.04it/s] 57%|█████▋ | 27554/48008 [1:34:28<1:18:45, 4.33it/s] 57%|█████▋ | 27555/48008 [1:34:28<1:15:57, 4.49it/s] 57%|█████▋ | 27556/48008 [1:34:28<1:12:42, 4.69it/s] 57%|█████▋ | 27557/48008 [1:34:29<1:10:23, 4.84it/s] 57%|█████▋ | 27558/48008 [1:34:29<1:08:25, 4.98it/s] 57%|█████▋ | 27559/48008 [1:34:29<1:07:05, 5.08it/s] 57%|█████▋ | 27560/48008 [1:34:29<1:07:03, 5.08it/s] 57%|█████▋ | 27561/48008 [1:34:29<1:08:22, 4.98it/s] 57%|█████▋ | 27562/48008 [1:34:30<1:12:04, 4.73it/s] 57%|█████▋ | 27563/48008 [1:34:30<1:14:49, 4.55it/s] 57%|█████▋ | 27564/48008 [1:34:30<1:12:28, 4.70it/s] 57%|█████▋ | 27565/48008 [1:34:30<1:10:13, 4.85it/s] 57%|█████▋ | 27566/48008 [1:34:31<1:08:44, 4.96it/s] 57%|█████▋ | 27567/48008 [1:34:31<1:07:08, 5.07it/s] 57%|█████▋ | 27568/48008 [1:34:31<1:06:30, 5.12it/s] 57%|█████▋ | 27569/48008 [1:34:31<1:05:47, 5.18it/s] 57%|█████▋ | 27570/48008 [1:34:31<1:05:10, 5.23it/s] 57%|█████▋ | 27571/48008 [1:34:31<1:05:13, 5.22it/s] 57%|█████▋ | 27572/48008 [1:34:32<1:05:17, 5.22it/s] 57%|█████▋ | 27573/48008 [1:34:32<1:05:47, 5.18it/s] 57%|█████▋ | 27574/48008 [1:34:32<1:05:43, 5.18it/s] 57%|█████▋ | 27575/48008 [1:34:32<1:10:27, 4.83it/s] 57%|█████▋ | 27576/48008 [1:34:32<1:08:42, 4.96it/s] 57%|█████▋ | 27577/48008 [1:34:33<1:07:39, 5.03it/s] 57%|█████▋ | 27578/48008 [1:34:33<1:07:43, 5.03it/s] 57%|█████▋ | 27579/48008 [1:34:33<1:06:21, 5.13it/s] 57%|█████▋ | 27580/48008 [1:34:33<1:05:36, 5.19it/s] 57%|█████▋ | 27581/48008 [1:34:33<1:07:08, 5.07it/s] 57%|█████▋ | 27582/48008 [1:34:35<2:46:56, 2.04it/s] 57%|█████▋ | 27583/48008 [1:34:35<2:21:05, 2.41it/s] 57%|█████▋ | 27584/48008 [1:34:35<1:59:09, 2.86it/s] 57%|█████▋ | 27585/48008 [1:34:35<1:42:27, 3.32it/s] 57%|█████▋ | 27586/48008 [1:34:35<1:30:56, 3.74it/s] 57%|█████▋ | 27587/48008 [1:34:36<1:24:23, 4.03it/s] 57%|█████▋ | 27588/48008 [1:34:36<1:18:09, 4.35it/s] 57%|█████▋ | 27589/48008 [1:34:36<1:14:40, 4.56it/s] 57%|█████▋ | 27590/48008 [1:34:36<1:11:20, 4.77it/s] 57%|█████▋ | 27591/48008 [1:34:36<1:09:36, 4.89it/s] 57%|█████▋ | 27592/48008 [1:34:37<1:09:30, 4.90it/s] 57%|█████▋ | 27593/48008 [1:34:37<1:07:32, 5.04it/s] 57%|█████▋ | 27594/48008 [1:34:37<1:06:54, 5.09it/s] 57%|█████▋ | 27595/48008 [1:34:37<1:05:44, 5.18it/s] 57%|█████▋ | 27596/48008 [1:34:37<1:05:01, 5.23it/s] 57%|█████▋ | 27597/48008 [1:34:38<1:04:56, 5.24it/s] 57%|█████▋ | 27598/48008 [1:34:38<1:16:38, 4.44it/s] 57%|█████▋ | 27599/48008 [1:34:38<1:12:22, 4.70it/s] 57%|█████▋ | 27600/48008 [1:34:38<1:10:09, 4.85it/s] {'loss': 0.3022, 'grad_norm': 0.2578125, 'learning_rate': 8.502332944509249e-05, 'epoch': 0.57} 57%|█████▋ | 27600/48008 [1:34:38<1:10:09, 4.85it/s] 57%|█████▋ | 27601/48008 [1:34:38<1:08:13, 4.99it/s] 57%|█████▋ | 27602/48008 [1:34:39<1:07:35, 5.03it/s] 57%|█████▋ | 27603/48008 [1:34:39<1:06:53, 5.08it/s] 57%|█████▋ | 27604/48008 [1:34:39<1:05:57, 5.16it/s] 58%|█████▊ | 27605/48008 [1:34:39<1:06:04, 5.15it/s] 58%|█████▊ | 27606/48008 [1:34:39<1:05:36, 5.18it/s] 58%|█████▊ | 27607/48008 [1:34:40<1:05:13, 5.21it/s] 58%|█████▊ | 27608/48008 [1:34:40<1:04:51, 5.24it/s] 58%|█████▊ | 27609/48008 [1:34:40<1:05:04, 5.22it/s] 58%|█████▊ | 27610/48008 [1:34:40<1:05:10, 5.22it/s] 58%|█████▊ | 27611/48008 [1:34:40<1:06:25, 5.12it/s] 58%|█████▊ | 27612/48008 [1:34:41<1:05:31, 5.19it/s] 58%|█████▊ | 27613/48008 [1:34:41<1:04:42, 5.25it/s] 58%|█████▊ | 27614/48008 [1:34:41<1:09:24, 4.90it/s] 58%|█████▊ | 27615/48008 [1:34:41<1:08:43, 4.95it/s] 58%|█████▊ | 27616/48008 [1:34:41<1:08:52, 4.93it/s] 58%|█████▊ | 27617/48008 [1:34:42<1:07:18, 5.05it/s] 58%|█████▊ | 27618/48008 [1:34:42<1:06:44, 5.09it/s] 58%|█████▊ | 27619/48008 [1:34:42<1:07:23, 5.04it/s] 58%|█████▊ | 27620/48008 [1:34:42<1:06:05, 5.14it/s] 58%|█████▊ | 27621/48008 [1:34:42<1:05:49, 5.16it/s] 58%|█████▊ | 27622/48008 [1:34:42<1:05:04, 5.22it/s] 58%|█████▊ | 27623/48008 [1:34:43<1:09:53, 4.86it/s] 58%|█████▊ | 27624/48008 [1:34:43<1:07:58, 5.00it/s] 58%|█████▊ | 27625/48008 [1:34:43<1:07:16, 5.05it/s] 58%|█████▊ | 27626/48008 [1:34:44<2:26:41, 2.32it/s] 58%|█████▊ | 27627/48008 [1:34:44<2:02:09, 2.78it/s] 58%|█████▊ | 27628/48008 [1:34:44<1:44:34, 3.25it/s] 58%|█████▊ | 27629/48008 [1:34:45<1:32:35, 3.67it/s] 58%|█████▊ | 27630/48008 [1:34:45<1:25:34, 3.97it/s] 58%|█████▊ | 27631/48008 [1:34:45<1:19:18, 4.28it/s] 58%|█████▊ | 27632/48008 [1:34:46<2:20:00, 2.43it/s] 58%|█████▊ | 27633/48008 [1:34:46<1:57:12, 2.90it/s] 58%|█████▊ | 27634/48008 [1:34:46<1:41:31, 3.34it/s] 58%|█████▊ | 27635/48008 [1:34:46<1:31:34, 3.71it/s] 58%|█████▊ | 27636/48008 [1:34:47<1:25:27, 3.97it/s] 58%|█████▊ | 27637/48008 [1:34:47<1:19:30, 4.27it/s] 58%|█████▊ | 27638/48008 [1:34:47<1:15:09, 4.52it/s] 58%|█████▊ | 27639/48008 [1:34:47<1:11:37, 4.74it/s] 58%|█████▊ | 27640/48008 [1:34:47<1:09:30, 4.88it/s] 58%|█████▊ | 27641/48008 [1:34:48<1:09:15, 4.90it/s] 58%|█████▊ | 27642/48008 [1:34:48<1:08:09, 4.98it/s] 58%|█████▊ | 27643/48008 [1:34:48<1:06:55, 5.07it/s] 58%|█████▊ | 27644/48008 [1:34:48<1:11:13, 4.76it/s] 58%|█████▊ | 27645/48008 [1:34:48<1:09:42, 4.87it/s] 58%|█████▊ | 27646/48008 [1:34:49<1:07:57, 4.99it/s] 58%|█████▊ | 27647/48008 [1:34:49<1:06:31, 5.10it/s] 58%|█████▊ | 27648/48008 [1:34:49<1:05:59, 5.14it/s] 58%|█████▊ | 27649/48008 [1:34:49<1:05:01, 5.22it/s] 58%|█████▊ | 27650/48008 [1:34:49<1:04:22, 5.27it/s] {'loss': 0.3038, 'grad_norm': 0.1396484375, 'learning_rate': 8.481503082819531e-05, 'epoch': 0.58} 58%|█████▊ | 27650/48008 [1:34:49<1:04:22, 5.27it/s] 58%|█████▊ | 27651/48008 [1:34:50<1:05:26, 5.18it/s] 58%|█████▊ | 27652/48008 [1:34:50<1:05:21, 5.19it/s] 58%|█████▊ | 27653/48008 [1:34:50<1:05:02, 5.22it/s] 58%|█████▊ | 27654/48008 [1:34:50<1:04:30, 5.26it/s] 58%|█████▊ | 27655/48008 [1:34:50<1:16:07, 4.46it/s] 58%|█████▊ | 27656/48008 [1:34:51<1:13:06, 4.64it/s] 58%|█████▊ | 27657/48008 [1:34:51<1:22:12, 4.13it/s] 58%|█████▊ | 27658/48008 [1:34:51<1:16:43, 4.42it/s] 58%|█████▊ | 27659/48008 [1:34:51<1:13:13, 4.63it/s] 58%|█████▊ | 27660/48008 [1:34:52<1:10:45, 4.79it/s] 58%|█████▊ | 27661/48008 [1:34:52<1:09:23, 4.89it/s] 58%|█████▊ | 27662/48008 [1:34:52<1:07:50, 5.00it/s] 58%|█████▊ | 27663/48008 [1:34:52<1:06:42, 5.08it/s] 58%|█████▊ | 27664/48008 [1:34:52<1:05:50, 5.15it/s] 58%|█████▊ | 27665/48008 [1:34:52<1:05:18, 5.19it/s] 58%|█████▊ | 27666/48008 [1:34:53<1:05:11, 5.20it/s] 58%|█████▊ | 27667/48008 [1:34:53<1:04:45, 5.24it/s] 58%|█████▊ | 27668/48008 [1:34:54<2:08:26, 2.64it/s] 58%|█████▊ | 27669/48008 [1:34:54<1:48:53, 3.11it/s] 58%|█████▊ | 27670/48008 [1:34:54<1:36:05, 3.53it/s] 58%|█████▊ | 27671/48008 [1:34:54<1:26:44, 3.91it/s] 58%|█████▊ | 27672/48008 [1:34:54<1:20:11, 4.23it/s] 58%|█████▊ | 27673/48008 [1:34:55<1:16:13, 4.45it/s] 58%|█████▊ | 27674/48008 [1:34:55<1:12:41, 4.66it/s] 58%|█████▊ | 27675/48008 [1:34:55<1:09:57, 4.84it/s] 58%|█████▊ | 27676/48008 [1:34:55<1:09:26, 4.88it/s] 58%|█████▊ | 27677/48008 [1:34:56<1:27:55, 3.85it/s] 58%|█████▊ | 27678/48008 [1:34:56<1:22:13, 4.12it/s] 58%|█████▊ | 27679/48008 [1:34:56<1:17:50, 4.35it/s] 58%|█████▊ | 27680/48008 [1:34:56<1:14:14, 4.56it/s] 58%|█████▊ | 27681/48008 [1:34:56<1:11:25, 4.74it/s] 58%|█████▊ | 27682/48008 [1:34:57<1:09:18, 4.89it/s] 58%|█████▊ | 27683/48008 [1:34:57<1:07:51, 4.99it/s] 58%|█████▊ | 27684/48008 [1:34:57<1:07:02, 5.05it/s] 58%|█████▊ | 27685/48008 [1:34:57<1:06:11, 5.12it/s] 58%|█████▊ | 27686/48008 [1:34:57<1:05:38, 5.16it/s] 58%|█████▊ | 27687/48008 [1:34:58<1:05:10, 5.20it/s] 58%|█████▊ | 27688/48008 [1:34:58<1:04:37, 5.24it/s] 58%|█████▊ | 27689/48008 [1:34:58<1:05:20, 5.18it/s] 58%|█████▊ | 27690/48008 [1:34:58<1:05:01, 5.21it/s] 58%|█████▊ | 27691/48008 [1:34:58<1:04:54, 5.22it/s] 58%|█████▊ | 27692/48008 [1:34:59<1:06:06, 5.12it/s] 58%|█████▊ | 27693/48008 [1:34:59<1:05:45, 5.15it/s] 58%|█████▊ | 27694/48008 [1:34:59<1:05:35, 5.16it/s] 58%|█████▊ | 27695/48008 [1:34:59<1:06:02, 5.13it/s] 58%|█████▊ | 27696/48008 [1:34:59<1:05:36, 5.16it/s] 58%|█████▊ | 27697/48008 [1:34:59<1:05:09, 5.20it/s] 58%|█████▊ | 27698/48008 [1:35:00<1:05:36, 5.16it/s] 58%|█████▊ | 27699/48008 [1:35:00<1:09:57, 4.84it/s] 58%|█████▊ | 27700/48008 [1:35:00<1:07:56, 4.98it/s] 58%|█████▊ | 27700/48008 [1:35:00<1:07:56, 4.98it/s]{'loss': 0.3012, 'grad_norm': 0.328125, 'learning_rate': 8.460673221129813e-05, 'epoch': 0.58} 58%|█████▊ | 27701/48008 [1:35:00<1:07:14, 5.03it/s] 58%|█████▊ | 27702/48008 [1:35:00<1:07:05, 5.04it/s] 58%|█████▊ | 27703/48008 [1:35:01<1:05:51, 5.14it/s] 58%|█████▊ | 27704/48008 [1:35:01<1:05:32, 5.16it/s] 58%|█████▊ | 27705/48008 [1:35:01<1:06:08, 5.12it/s] 58%|█████▊ | 27706/48008 [1:35:01<1:05:33, 5.16it/s] 58%|█████▊ | 27707/48008 [1:35:01<1:05:05, 5.20it/s] 58%|█████▊ | 27708/48008 [1:35:02<1:04:15, 5.26it/s] 58%|█████▊ | 27709/48008 [1:35:02<1:03:59, 5.29it/s] 58%|█████▊ | 27710/48008 [1:35:02<1:05:14, 5.19it/s] 58%|█████▊ | 27711/48008 [1:35:02<1:06:06, 5.12it/s] 58%|█████▊ | 27712/48008 [1:35:02<1:06:14, 5.11it/s] 58%|█████▊ | 27713/48008 [1:35:03<1:05:51, 5.14it/s] 58%|█████▊ | 27714/48008 [1:35:03<1:05:17, 5.18it/s] 58%|█████▊ | 27715/48008 [1:35:03<1:04:47, 5.22it/s] 58%|█████▊ | 27716/48008 [1:35:03<1:05:02, 5.20it/s] 58%|█████▊ | 27717/48008 [1:35:03<1:04:51, 5.21it/s] 58%|█████▊ | 27718/48008 [1:35:04<1:05:48, 5.14it/s] 58%|█████▊ | 27719/48008 [1:35:04<1:09:54, 4.84it/s] 58%|█████▊ | 27720/48008 [1:35:04<1:13:12, 4.62it/s] 58%|█████▊ | 27721/48008 [1:35:04<1:10:37, 4.79it/s] 58%|█████▊ | 27722/48008 [1:35:04<1:07:52, 4.98it/s] 58%|█████▊ | 27723/48008 [1:35:05<1:07:06, 5.04it/s] 58%|█████▊ | 27724/48008 [1:35:05<1:07:39, 5.00it/s] 58%|█████▊ | 27725/48008 [1:35:05<1:07:08, 5.03it/s] 58%|█████▊ | 27726/48008 [1:35:05<1:06:29, 5.08it/s] 58%|█████▊ | 27727/48008 [1:35:05<1:05:35, 5.15it/s] 58%|█████▊ | 27728/48008 [1:35:06<1:05:06, 5.19it/s] 58%|█████▊ | 27729/48008 [1:35:06<1:06:57, 5.05it/s] 58%|█████▊ | 27730/48008 [1:35:06<1:06:24, 5.09it/s] 58%|█████▊ | 27731/48008 [1:35:06<1:05:36, 5.15it/s] 58%|█████▊ | 27732/48008 [1:35:06<1:05:51, 5.13it/s] 58%|█████▊ | 27733/48008 [1:35:07<1:06:01, 5.12it/s] 58%|█████▊ | 27734/48008 [1:35:07<1:05:23, 5.17it/s] 58%|█████▊ | 27735/48008 [1:35:07<1:04:10, 5.26it/s] 58%|█████▊ | 27736/48008 [1:35:07<1:04:21, 5.25it/s] 58%|█████▊ | 27737/48008 [1:35:07<1:04:15, 5.26it/s] 58%|█████▊ | 27738/48008 [1:35:08<1:04:33, 5.23it/s] 58%|█████▊ | 27739/48008 [1:35:08<1:04:34, 5.23it/s] 58%|█████▊ | 27740/48008 [1:35:08<1:04:10, 5.26it/s] 58%|█████▊ | 27741/48008 [1:35:08<1:03:40, 5.31it/s] 58%|█████▊ | 27742/48008 [1:35:08<1:03:49, 5.29it/s] 58%|█████▊ | 27743/48008 [1:35:08<1:04:36, 5.23it/s] 58%|█████▊ | 27744/48008 [1:35:09<1:05:31, 5.15it/s] 58%|█████▊ | 27745/48008 [1:35:09<1:10:00, 4.82it/s] 58%|█████▊ | 27746/48008 [1:35:09<1:09:11, 4.88it/s] 58%|█████▊ | 27747/48008 [1:35:09<1:08:47, 4.91it/s] 58%|█████▊ | 27748/48008 [1:35:09<1:07:14, 5.02it/s] 58%|█████▊ | 27749/48008 [1:35:10<1:06:22, 5.09it/s] 58%|█████▊ | 27750/48008 [1:35:10<1:05:46, 5.13it/s] {'loss': 0.3019, 'grad_norm': 0.142578125, 'learning_rate': 8.439843359440094e-05, 'epoch': 0.58} 58%|█████▊ | 27750/48008 [1:35:10<1:05:46, 5.13it/s] 58%|█████▊ | 27751/48008 [1:35:10<1:06:29, 5.08it/s] 58%|█████▊ | 27752/48008 [1:35:10<1:05:24, 5.16it/s] 58%|█████▊ | 27753/48008 [1:35:10<1:05:49, 5.13it/s] 58%|█████▊ | 27754/48008 [1:35:11<1:04:54, 5.20it/s] 58%|█████▊ | 27755/48008 [1:35:11<1:05:29, 5.15it/s] 58%|█████▊ | 27756/48008 [1:35:11<1:06:33, 5.07it/s] 58%|█████▊ | 27757/48008 [1:35:11<1:06:35, 5.07it/s] 58%|█████▊ | 27758/48008 [1:35:11<1:05:48, 5.13it/s] 58%|█████▊ | 27759/48008 [1:35:12<1:05:24, 5.16it/s] 58%|█████▊ | 27760/48008 [1:35:12<1:05:16, 5.17it/s] 58%|█████▊ | 27761/48008 [1:35:12<1:05:02, 5.19it/s] 58%|█████▊ | 27762/48008 [1:35:12<1:04:56, 5.20it/s] 58%|█████▊ | 27763/48008 [1:35:12<1:04:28, 5.23it/s] 58%|█████▊ | 27764/48008 [1:35:13<1:04:03, 5.27it/s] 58%|█████▊ | 27765/48008 [1:35:13<1:05:14, 5.17it/s] 58%|█████▊ | 27766/48008 [1:35:13<1:05:23, 5.16it/s] 58%|█████▊ | 27767/48008 [1:35:13<1:06:53, 5.04it/s] 58%|█████▊ | 27768/48008 [1:35:13<1:05:54, 5.12it/s] 58%|█████▊ | 27769/48008 [1:35:14<1:05:42, 5.13it/s] 58%|█████▊ | 27770/48008 [1:35:14<1:05:15, 5.17it/s] 58%|█████▊ | 27771/48008 [1:35:14<1:06:07, 5.10it/s] 58%|█████▊ | 27772/48008 [1:35:14<1:05:28, 5.15it/s] 58%|█████▊ | 27773/48008 [1:35:14<1:05:56, 5.11it/s] 58%|█████▊ | 27774/48008 [1:35:15<1:06:48, 5.05it/s] 58%|█████▊ | 27775/48008 [1:35:15<1:05:33, 5.14it/s] 58%|█████▊ | 27776/48008 [1:35:15<1:05:06, 5.18it/s] 58%|█████▊ | 27777/48008 [1:35:15<1:09:31, 4.85it/s] 58%|█████▊ | 27778/48008 [1:35:15<1:08:36, 4.91it/s] 58%|█████▊ | 27779/48008 [1:35:16<1:07:25, 5.00it/s] 58%|█████▊ | 27780/48008 [1:35:16<1:18:01, 4.32it/s] 58%|█████▊ | 27781/48008 [1:35:16<1:14:04, 4.55it/s] 58%|█████▊ | 27782/48008 [1:35:16<1:11:42, 4.70it/s] 58%|█████▊ | 27783/48008 [1:35:16<1:10:05, 4.81it/s] 58%|█████▊ | 27784/48008 [1:35:17<1:08:01, 4.96it/s] 58%|█████▊ | 27785/48008 [1:35:17<1:06:55, 5.04it/s] 58%|█████▊ | 27786/48008 [1:35:17<1:06:04, 5.10it/s] 58%|█████▊ | 27787/48008 [1:35:17<1:05:02, 5.18it/s] 58%|█████▊ | 27788/48008 [1:35:17<1:04:18, 5.24it/s] 58%|█████▊ | 27789/48008 [1:35:18<1:04:38, 5.21it/s] 58%|█████▊ | 27790/48008 [1:35:18<1:04:46, 5.20it/s] 58%|█████▊ | 27791/48008 [1:35:18<1:04:29, 5.23it/s] 58%|█████▊ | 27792/48008 [1:35:18<1:05:53, 5.11it/s] 58%|█████▊ | 27793/48008 [1:35:18<1:05:31, 5.14it/s] 58%|█████▊ | 27794/48008 [1:35:19<1:05:36, 5.13it/s] 58%|█████▊ | 27795/48008 [1:35:19<1:05:26, 5.15it/s] 58%|█████▊ | 27796/48008 [1:35:19<1:04:49, 5.20it/s] 58%|█████▊ | 27797/48008 [1:35:19<1:04:45, 5.20it/s] 58%|█████▊ | 27798/48008 [1:35:19<1:04:52, 5.19it/s] 58%|█████▊ | 27799/48008 [1:35:19<1:04:45, 5.20it/s] 58%|█████▊ | 27800/48008 [1:35:20<1:04:44, 5.20it/s] {'loss': 0.3037, 'grad_norm': 0.1728515625, 'learning_rate': 8.419013497750375e-05, 'epoch': 0.58} 58%|█████▊ | 27800/48008 [1:35:20<1:04:44, 5.20it/s] 58%|█████▊ | 27801/48008 [1:35:20<1:04:39, 5.21it/s] 58%|█████▊ | 27802/48008 [1:35:20<1:04:48, 5.20it/s] 58%|█████▊ | 27803/48008 [1:35:20<1:05:03, 5.18it/s] 58%|█████▊ | 27804/48008 [1:35:20<1:04:32, 5.22it/s] 58%|█████▊ | 27805/48008 [1:35:21<1:03:51, 5.27it/s] 58%|█████▊ | 27806/48008 [1:35:21<1:03:49, 5.28it/s] 58%|█████▊ | 27807/48008 [1:35:21<1:03:59, 5.26it/s] 58%|█████▊ | 27808/48008 [1:35:21<1:08:44, 4.90it/s] 58%|█████▊ | 27809/48008 [1:35:21<1:07:09, 5.01it/s] 58%|█████▊ | 27810/48008 [1:35:22<1:06:12, 5.08it/s] 58%|█████▊ | 27811/48008 [1:35:22<1:05:44, 5.12it/s] 58%|█████▊ | 27812/48008 [1:35:22<1:05:18, 5.15it/s] 58%|█████▊ | 27813/48008 [1:35:22<1:06:55, 5.03it/s] 58%|█████▊ | 27814/48008 [1:35:22<1:06:35, 5.05it/s] 58%|█████▊ | 27815/48008 [1:35:23<1:05:44, 5.12it/s] 58%|█████▊ | 27816/48008 [1:35:23<1:05:08, 5.17it/s] 58%|█████▊ | 27817/48008 [1:35:23<1:04:31, 5.21it/s] 58%|█████▊ | 27818/48008 [1:35:23<1:04:31, 5.21it/s] 58%|█████▊ | 27819/48008 [1:35:23<1:04:05, 5.25it/s] 58%|█████▊ | 27820/48008 [1:35:24<1:09:00, 4.88it/s] 58%|█████▊ | 27821/48008 [1:35:24<1:09:17, 4.86it/s] 58%|█████▊ | 27822/48008 [1:35:24<1:07:15, 5.00it/s] 58%|█████▊ | 27823/48008 [1:35:24<1:07:25, 4.99it/s] 58%|█████▊ | 27824/48008 [1:35:24<1:06:00, 5.10it/s] 58%|█████▊ | 27825/48008 [1:35:25<1:04:58, 5.18it/s] 58%|█████▊ | 27826/48008 [1:35:25<1:06:08, 5.09it/s] 58%|█████▊ | 27827/48008 [1:35:25<1:05:22, 5.15it/s] 58%|█████▊ | 27828/48008 [1:35:25<1:05:22, 5.14it/s] 58%|█████▊ | 27829/48008 [1:35:25<1:05:20, 5.15it/s] 58%|█████▊ | 27830/48008 [1:35:26<1:05:02, 5.17it/s] 58%|█████▊ | 27831/48008 [1:35:26<1:04:36, 5.20it/s] 58%|█████▊ | 27832/48008 [1:35:26<1:03:59, 5.26it/s] 58%|█████▊ | 27833/48008 [1:35:26<1:23:44, 4.02it/s] 58%|█████▊ | 27834/48008 [1:35:27<1:17:55, 4.31it/s] 58%|█████▊ | 27835/48008 [1:35:27<1:13:34, 4.57it/s] 58%|█████▊ | 27836/48008 [1:35:27<1:30:28, 3.72it/s] 58%|█████▊ | 27837/48008 [1:35:27<1:22:32, 4.07it/s] 58%|█████▊ | 27838/48008 [1:35:27<1:17:13, 4.35it/s] 58%|█████▊ | 27839/48008 [1:35:28<1:12:54, 4.61it/s] 58%|█████▊ | 27840/48008 [1:35:28<1:10:27, 4.77it/s] 58%|█████▊ | 27841/48008 [1:35:28<1:08:38, 4.90it/s] 58%|█████▊ | 27842/48008 [1:35:28<1:08:27, 4.91it/s] 58%|█████▊ | 27843/48008 [1:35:28<1:08:18, 4.92it/s] 58%|█████▊ | 27844/48008 [1:35:29<2:15:46, 2.48it/s] 58%|█████▊ | 27845/48008 [1:35:30<1:54:24, 2.94it/s] 58%|█████▊ | 27846/48008 [1:35:30<1:40:32, 3.34it/s] 58%|█████▊ | 27847/48008 [1:35:30<1:30:24, 3.72it/s] 58%|█████▊ | 27848/48008 [1:35:30<1:27:15, 3.85it/s] 58%|█████▊ | 27849/48008 [1:35:30<1:20:49, 4.16it/s] 58%|█████▊ | 27850/48008 [1:35:31<1:16:00, 4.42it/s] {'loss': 0.3009, 'grad_norm': 0.1904296875, 'learning_rate': 8.398183636060657e-05, 'epoch': 0.58} 58%|█████▊ | 27850/48008 [1:35:31<1:16:00, 4.42it/s] 58%|█████▊ | 27851/48008 [1:35:31<1:32:09, 3.65it/s] 58%|█████▊ | 27852/48008 [1:35:31<1:23:31, 4.02it/s] 58%|█████▊ | 27853/48008 [1:35:31<1:17:18, 4.35it/s] 58%|█████▊ | 27854/48008 [1:35:31<1:13:30, 4.57it/s] 58%|█████▊ | 27855/48008 [1:35:32<1:10:30, 4.76it/s] 58%|█████▊ | 27856/48008 [1:35:32<1:08:32, 4.90it/s] 58%|█████▊ | 27857/48008 [1:35:32<1:07:46, 4.96it/s] 58%|█████▊ | 27858/48008 [1:35:32<1:06:31, 5.05it/s] 58%|█████▊ | 27859/48008 [1:35:32<1:05:52, 5.10it/s] 58%|█████▊ | 27860/48008 [1:35:33<1:09:55, 4.80it/s] 58%|█████▊ | 27861/48008 [1:35:33<1:10:11, 4.78it/s] 58%|█████▊ | 27862/48008 [1:35:33<1:09:13, 4.85it/s] 58%|█████▊ | 27863/48008 [1:35:33<1:07:46, 4.95it/s] 58%|█████▊ | 27864/48008 [1:35:33<1:06:27, 5.05it/s] 58%|█████▊ | 27865/48008 [1:35:34<1:07:12, 5.00it/s] 58%|█████▊ | 27866/48008 [1:35:34<1:06:50, 5.02it/s] 58%|█████▊ | 27867/48008 [1:35:34<1:07:13, 4.99it/s] 58%|█████▊ | 27868/48008 [1:35:34<1:05:52, 5.10it/s] 58%|█████▊ | 27869/48008 [1:35:34<1:06:36, 5.04it/s] 58%|█████▊ | 27870/48008 [1:35:35<1:05:52, 5.09it/s] 58%|█████▊ | 27871/48008 [1:35:35<1:06:33, 5.04it/s] 58%|█████▊ | 27872/48008 [1:35:35<1:05:38, 5.11it/s] 58%|█████▊ | 27873/48008 [1:35:35<1:05:24, 5.13it/s] 58%|█████▊ | 27874/48008 [1:35:35<1:06:17, 5.06it/s] 58%|█████▊ | 27875/48008 [1:35:36<1:04:35, 5.19it/s] 58%|█████▊ | 27876/48008 [1:35:36<1:04:58, 5.16it/s] 58%|█████▊ | 27877/48008 [1:35:36<1:04:14, 5.22it/s] 58%|█████▊ | 27878/48008 [1:35:36<1:04:06, 5.23it/s] 58%|█████▊ | 27879/48008 [1:35:36<1:03:55, 5.25it/s] 58%|█████▊ | 27880/48008 [1:35:37<1:03:55, 5.25it/s] 58%|█████▊ | 27881/48008 [1:35:37<1:03:38, 5.27it/s] 58%|█████▊ | 27882/48008 [1:35:37<1:04:01, 5.24it/s] 58%|█████▊ | 27883/48008 [1:35:37<1:03:21, 5.29it/s] 58%|█████▊ | 27884/48008 [1:35:38<1:23:31, 4.02it/s] 58%|█████▊ | 27885/48008 [1:35:38<1:17:49, 4.31it/s] 58%|█████▊ | 27886/48008 [1:35:38<1:14:16, 4.52it/s] 58%|█████▊ | 27887/48008 [1:35:38<1:11:19, 4.70it/s] 58%|█████▊ | 27888/48008 [1:35:38<1:09:50, 4.80it/s] 58%|█████▊ | 27889/48008 [1:35:39<1:08:00, 4.93it/s] 58%|█████▊ | 27890/48008 [1:35:39<1:06:47, 5.02it/s] 58%|█████▊ | 27891/48008 [1:35:39<1:05:16, 5.14it/s] 58%|█████▊ | 27892/48008 [1:35:39<1:04:34, 5.19it/s] 58%|█████▊ | 27893/48008 [1:35:39<1:04:07, 5.23it/s] 58%|█████▊ | 27894/48008 [1:35:39<1:04:08, 5.23it/s] 58%|█████▊ | 27895/48008 [1:35:40<1:04:01, 5.24it/s] 58%|█████▊ | 27896/48008 [1:35:40<1:03:51, 5.25it/s] 58%|█████▊ | 27897/48008 [1:35:40<1:04:37, 5.19it/s] 58%|█████▊ | 27898/48008 [1:35:40<1:03:57, 5.24it/s] 58%|█████▊ | 27899/48008 [1:35:40<1:03:54, 5.24it/s] 58%|█████▊ | 27900/48008 [1:35:41<1:04:03, 5.23it/s] 58%|█████▊ | 27900/48008 [1:35:41<1:04:03, 5.23it/s]{'loss': 0.3024, 'grad_norm': 0.6171875, 'learning_rate': 8.377353774370938e-05, 'epoch': 0.58} 58%|█████▊ | 27901/48008 [1:35:41<1:04:20, 5.21it/s] 58%|█████▊ | 27902/48008 [1:35:41<1:03:55, 5.24it/s] 58%|█████▊ | 27903/48008 [1:35:41<1:03:39, 5.26it/s] 58%|█████▊ | 27904/48008 [1:35:41<1:03:48, 5.25it/s] 58%|█████▊ | 27905/48008 [1:35:42<1:05:23, 5.12it/s] 58%|█████▊ | 27906/48008 [1:35:42<1:04:38, 5.18it/s] 58%|█████▊ | 27907/48008 [1:35:42<1:05:28, 5.12it/s] 58%|█████▊ | 27908/48008 [1:35:42<1:04:42, 5.18it/s] 58%|█████▊ | 27909/48008 [1:35:42<1:04:06, 5.23it/s] 58%|█████▊ | 27910/48008 [1:35:43<1:04:39, 5.18it/s] 58%|█████▊ | 27911/48008 [1:35:43<1:04:33, 5.19it/s] 58%|█████▊ | 27912/48008 [1:35:43<1:05:44, 5.09it/s] 58%|█████▊ | 27913/48008 [1:35:43<1:05:17, 5.13it/s] 58%|█████▊ | 27914/48008 [1:35:43<1:04:56, 5.16it/s] 58%|█████▊ | 27915/48008 [1:35:43<1:04:51, 5.16it/s] 58%|█████▊ | 27916/48008 [1:35:44<1:04:28, 5.19it/s] 58%|█████▊ | 27917/48008 [1:35:44<1:04:28, 5.19it/s] 58%|█████▊ | 27918/48008 [1:35:44<1:04:06, 5.22it/s] 58%|█████▊ | 27919/48008 [1:35:44<1:03:42, 5.26it/s] 58%|█████▊ | 27920/48008 [1:35:45<1:23:33, 4.01it/s] 58%|█████▊ | 27921/48008 [1:35:45<1:17:32, 4.32it/s] 58%|█████▊ | 27922/48008 [1:35:45<1:14:01, 4.52it/s] 58%|█████▊ | 27923/48008 [1:35:45<1:10:56, 4.72it/s] 58%|█████▊ | 27924/48008 [1:35:45<1:08:34, 4.88it/s] 58%|█████▊ | 27925/48008 [1:35:46<1:07:11, 4.98it/s] 58%|█████▊ | 27926/48008 [1:35:46<1:06:52, 5.00it/s] 58%|█████▊ | 27927/48008 [1:35:46<1:06:42, 5.02it/s] 58%|█████▊ | 27928/48008 [1:35:46<1:05:38, 5.10it/s] 58%|█████▊ | 27929/48008 [1:35:46<1:05:06, 5.14it/s] 58%|█████▊ | 27930/48008 [1:35:47<1:04:36, 5.18it/s] 58%|█████▊ | 27931/48008 [1:35:47<1:05:33, 5.10it/s] 58%|█████▊ | 27932/48008 [1:35:47<1:04:37, 5.18it/s] 58%|█████▊ | 27933/48008 [1:35:47<1:04:05, 5.22it/s] 58%|█████▊ | 27934/48008 [1:35:47<1:05:41, 5.09it/s] 58%|█████▊ | 27935/48008 [1:35:48<1:16:50, 4.35it/s] 58%|█████▊ | 27936/48008 [1:35:48<1:32:31, 3.62it/s] 58%|█████▊ | 27937/48008 [1:35:48<1:23:43, 4.00it/s] 58%|█████▊ | 27938/48008 [1:35:48<1:17:12, 4.33it/s] 58%|█████▊ | 27939/48008 [1:35:49<1:14:35, 4.48it/s] 58%|█████▊ | 27940/48008 [1:35:49<1:11:16, 4.69it/s] 58%|█████▊ | 27941/48008 [1:35:49<1:09:12, 4.83it/s] 58%|█████▊ | 27942/48008 [1:35:49<1:06:39, 5.02it/s] 58%|█████▊ | 27943/48008 [1:35:49<1:05:47, 5.08it/s] 58%|█████▊ | 27944/48008 [1:35:50<1:04:52, 5.15it/s] 58%|█████▊ | 27945/48008 [1:35:50<1:04:21, 5.20it/s] 58%|█████▊ | 27946/48008 [1:35:50<1:04:25, 5.19it/s] 58%|█████▊ | 27947/48008 [1:35:50<1:04:05, 5.22it/s] 58%|█████▊ | 27948/48008 [1:35:50<1:03:52, 5.23it/s] 58%|█████▊ | 27949/48008 [1:35:51<1:03:46, 5.24it/s] 58%|█████▊ | 27950/48008 [1:35:51<1:04:33, 5.18it/s] {'loss': 0.3018, 'grad_norm': 0.220703125, 'learning_rate': 8.35652391268122e-05, 'epoch': 0.58} 58%|█████▊ | 27950/48008 [1:35:51<1:04:33, 5.18it/s] 58%|█████▊ | 27951/48008 [1:35:51<1:04:28, 5.18it/s] 58%|█████▊ | 27952/48008 [1:35:51<1:04:11, 5.21it/s] 58%|█████▊ | 27953/48008 [1:35:51<1:05:55, 5.07it/s] 58%|█████▊ | 27954/48008 [1:35:51<1:05:18, 5.12it/s] 58%|█████▊ | 27955/48008 [1:35:52<1:04:53, 5.15it/s] 58%|█████▊ | 27956/48008 [1:35:52<1:04:35, 5.17it/s] 58%|█████▊ | 27957/48008 [1:35:52<1:04:05, 5.21it/s] 58%|█████▊ | 27958/48008 [1:35:52<1:04:02, 5.22it/s] 58%|█████▊ | 27959/48008 [1:35:52<1:03:23, 5.27it/s] 58%|█████▊ | 27960/48008 [1:35:53<1:04:44, 5.16it/s] 58%|█████▊ | 27961/48008 [1:35:53<1:04:21, 5.19it/s] 58%|█████▊ | 27962/48008 [1:35:53<1:04:23, 5.19it/s] 58%|█████▊ | 27963/48008 [1:35:53<1:05:18, 5.11it/s] 58%|█████▊ | 27964/48008 [1:35:53<1:04:57, 5.14it/s] 58%|█████▊ | 27965/48008 [1:35:54<1:05:04, 5.13it/s] 58%|█████▊ | 27966/48008 [1:35:54<1:04:41, 5.16it/s] 58%|█████▊ | 27967/48008 [1:35:54<1:04:37, 5.17it/s] 58%|█████▊ | 27968/48008 [1:35:54<1:04:19, 5.19it/s] 58%|█████▊ | 27969/48008 [1:35:54<1:04:05, 5.21it/s] 58%|█████▊ | 27970/48008 [1:35:55<1:08:24, 4.88it/s] 58%|█████▊ | 27971/48008 [1:35:55<1:06:34, 5.02it/s] 58%|█████▊ | 27972/48008 [1:35:55<1:17:17, 4.32it/s] 58%|█████▊ | 27973/48008 [1:35:55<1:17:57, 4.28it/s] 58%|█████▊ | 27974/48008 [1:35:56<1:14:53, 4.46it/s] 58%|█████▊ | 27975/48008 [1:35:56<1:11:23, 4.68it/s] 58%|█████▊ | 27976/48008 [1:35:56<1:09:04, 4.83it/s] 58%|█████▊ | 27977/48008 [1:35:56<1:12:11, 4.62it/s] 58%|█████▊ | 27978/48008 [1:35:56<1:11:04, 4.70it/s] 58%|█████▊ | 27979/48008 [1:35:57<1:08:27, 4.88it/s] 58%|█████▊ | 27980/48008 [1:35:57<1:06:57, 4.99it/s] 58%|█████▊ | 27981/48008 [1:35:57<1:06:29, 5.02it/s] 58%|█████▊ | 27982/48008 [1:35:57<1:05:04, 5.13it/s] 58%|█████▊ | 27983/48008 [1:35:57<1:04:25, 5.18it/s] 58%|█████▊ | 27984/48008 [1:35:58<1:04:07, 5.20it/s] 58%|█████▊ | 27985/48008 [1:35:58<1:04:04, 5.21it/s] 58%|█████▊ | 27986/48008 [1:35:58<1:04:01, 5.21it/s] 58%|█████▊ | 27987/48008 [1:35:58<1:03:38, 5.24it/s] 58%|█████▊ | 27988/48008 [1:35:58<1:05:39, 5.08it/s] 58%|█████▊ | 27989/48008 [1:35:58<1:04:55, 5.14it/s] 58%|█████▊ | 27990/48008 [1:35:59<1:05:35, 5.09it/s] 58%|█████▊ | 27991/48008 [1:35:59<1:04:58, 5.13it/s] 58%|█████▊ | 27992/48008 [1:35:59<1:09:12, 4.82it/s] 58%|█████▊ | 27993/48008 [1:35:59<1:07:20, 4.95it/s] 58%|█████▊ | 27994/48008 [1:36:00<1:07:28, 4.94it/s] 58%|█████▊ | 27995/48008 [1:36:00<1:06:17, 5.03it/s] 58%|█████▊ | 27996/48008 [1:36:00<1:06:31, 5.01it/s] 58%|█████▊ | 27997/48008 [1:36:00<1:06:42, 5.00it/s] 58%|█████▊ | 27998/48008 [1:36:00<1:05:40, 5.08it/s] 58%|█████▊ | 27999/48008 [1:36:00<1:04:48, 5.15it/s] 58%|█████▊ | 28000/48008 [1:36:01<1:04:10, 5.20it/s] {'loss': 0.3011, 'grad_norm': 0.5, 'learning_rate': 8.335694050991502e-05, 'epoch': 0.58} 58%|█████▊ | 28000/48008 [1:36:01<1:04:10, 5.20it/s] 58%|█████▊ | 28001/48008 [1:36:01<1:03:39, 5.24it/s] 58%|█████▊ | 28002/48008 [1:36:01<1:04:03, 5.20it/s] 58%|█████▊ | 28003/48008 [1:36:01<1:08:36, 4.86it/s] 58%|█████▊ | 28004/48008 [1:36:01<1:06:46, 4.99it/s] 58%|█████▊ | 28005/48008 [1:36:02<1:10:25, 4.73it/s] 58%|█████▊ | 28006/48008 [1:36:02<1:08:28, 4.87it/s] 58%|█████▊ | 28007/48008 [1:36:02<1:06:57, 4.98it/s] 58%|█████▊ | 28008/48008 [1:36:02<1:05:31, 5.09it/s] 58%|█████▊ | 28009/48008 [1:36:02<1:05:02, 5.12it/s] 58%|█████▊ | 28010/48008 [1:36:03<1:24:13, 3.96it/s] 58%|█████▊ | 28011/48008 [1:36:03<1:18:02, 4.27it/s] 58%|█████▊ | 28012/48008 [1:36:03<1:14:21, 4.48it/s] 58%|█████▊ | 28013/48008 [1:36:03<1:12:12, 4.61it/s] 58%|█████▊ | 28014/48008 [1:36:04<1:09:38, 4.79it/s] 58%|█████▊ | 28015/48008 [1:36:04<1:09:45, 4.78it/s] 58%|█████▊ | 28016/48008 [1:36:04<1:07:46, 4.92it/s] 58%|█████▊ | 28017/48008 [1:36:04<1:11:03, 4.69it/s] 58%|█████▊ | 28018/48008 [1:36:04<1:08:22, 4.87it/s] 58%|█████▊ | 28019/48008 [1:36:05<1:06:49, 4.99it/s] 58%|█████▊ | 28020/48008 [1:36:05<1:07:11, 4.96it/s] 58%|█████▊ | 28021/48008 [1:36:05<1:06:43, 4.99it/s] 58%|█████▊ | 28022/48008 [1:36:05<1:05:34, 5.08it/s] 58%|█████▊ | 28023/48008 [1:36:05<1:05:36, 5.08it/s] 58%|█████▊ | 28024/48008 [1:36:06<1:06:51, 4.98it/s] 58%|█████▊ | 28025/48008 [1:36:06<1:05:27, 5.09it/s] 58%|█████▊ | 28026/48008 [1:36:06<1:05:27, 5.09it/s] 58%|█████▊ | 28027/48008 [1:36:06<1:04:29, 5.16it/s] 58%|█████▊ | 28028/48008 [1:36:06<1:04:15, 5.18it/s] 58%|█████▊ | 28029/48008 [1:36:07<1:03:09, 5.27it/s] 58%|█████▊ | 28030/48008 [1:36:07<1:02:45, 5.31it/s] 58%|█████▊ | 28031/48008 [1:36:07<1:03:18, 5.26it/s] 58%|█████▊ | 28032/48008 [1:36:07<1:03:16, 5.26it/s] 58%|█████▊ | 28033/48008 [1:36:07<1:03:11, 5.27it/s] 58%|█████▊ | 28034/48008 [1:36:08<1:03:20, 5.26it/s] 58%|█████▊ | 28035/48008 [1:36:08<1:03:26, 5.25it/s] 58%|█████▊ | 28036/48008 [1:36:08<1:03:53, 5.21it/s] 58%|█████▊ | 28037/48008 [1:36:08<1:04:11, 5.18it/s] 58%|█████▊ | 28038/48008 [1:36:08<1:04:30, 5.16it/s] 58%|█████▊ | 28039/48008 [1:36:09<1:03:53, 5.21it/s] 58%|█████▊ | 28040/48008 [1:36:09<1:04:07, 5.19it/s] 58%|█████▊ | 28041/48008 [1:36:09<1:03:37, 5.23it/s] 58%|█████▊ | 28042/48008 [1:36:09<1:15:02, 4.43it/s] 58%|█████▊ | 28043/48008 [1:36:09<1:11:17, 4.67it/s] 58%|█████▊ | 28044/48008 [1:36:10<1:10:13, 4.74it/s] 58%|█████▊ | 28045/48008 [1:36:10<1:09:32, 4.78it/s] 58%|█████▊ | 28046/48008 [1:36:10<1:07:24, 4.94it/s] 58%|█████▊ | 28047/48008 [1:36:10<1:06:41, 4.99it/s] 58%|█████▊ | 28048/48008 [1:36:10<1:10:33, 4.71it/s] 58%|█████▊ | 28049/48008 [1:36:11<1:09:51, 4.76it/s] 58%|█████▊ | 28050/48008 [1:36:11<1:07:44, 4.91it/s] {'loss': 0.2996, 'grad_norm': 0.2099609375, 'learning_rate': 8.314864189301783e-05, 'epoch': 0.58} 58%|█████▊ | 28050/48008 [1:36:11<1:07:44, 4.91it/s] 58%|█████▊ | 28051/48008 [1:36:11<1:11:19, 4.66it/s] 58%|█████▊ | 28052/48008 [1:36:11<1:09:02, 4.82it/s] 58%|█████▊ | 28053/48008 [1:36:11<1:07:01, 4.96it/s] 58%|█████▊ | 28054/48008 [1:36:12<1:05:59, 5.04it/s] 58%|█████▊ | 28055/48008 [1:36:12<1:09:44, 4.77it/s] 58%|█████▊ | 28056/48008 [1:36:12<1:09:39, 4.77it/s] 58%|█████▊ | 28057/48008 [1:36:12<1:06:48, 4.98it/s] 58%|█████▊ | 28058/48008 [1:36:12<1:05:29, 5.08it/s] 58%|█████▊ | 28059/48008 [1:36:13<1:04:59, 5.12it/s] 58%|█████▊ | 28060/48008 [1:36:13<1:03:57, 5.20it/s] 58%|█████▊ | 28061/48008 [1:36:13<1:03:58, 5.20it/s] 58%|█████▊ | 28062/48008 [1:36:13<1:04:07, 5.18it/s] 58%|█████▊ | 28063/48008 [1:36:13<1:03:27, 5.24it/s] 58%|█████▊ | 28064/48008 [1:36:14<1:04:12, 5.18it/s] 58%|█████▊ | 28065/48008 [1:36:14<1:04:05, 5.19it/s] 58%|█████▊ | 28066/48008 [1:36:14<1:03:22, 5.24it/s] 58%|█████▊ | 28067/48008 [1:36:14<1:03:03, 5.27it/s] 58%|█████▊ | 28068/48008 [1:36:14<1:03:16, 5.25it/s] 58%|█████▊ | 28069/48008 [1:36:15<1:03:19, 5.25it/s] 58%|█████▊ | 28070/48008 [1:36:15<1:03:09, 5.26it/s] 58%|█████▊ | 28071/48008 [1:36:15<1:02:55, 5.28it/s] 58%|█████▊ | 28072/48008 [1:36:15<1:04:23, 5.16it/s] 58%|█████▊ | 28073/48008 [1:36:15<1:03:53, 5.20it/s] 58%|█████▊ | 28074/48008 [1:36:15<1:03:18, 5.25it/s] 58%|█████▊ | 28075/48008 [1:36:16<1:03:21, 5.24it/s] 58%|█████▊ | 28076/48008 [1:36:16<1:03:13, 5.25it/s] 58%|█████▊ | 28077/48008 [1:36:16<1:02:52, 5.28it/s] 58%|█████▊ | 28078/48008 [1:36:16<1:03:25, 5.24it/s] 58%|█████▊ | 28079/48008 [1:36:16<1:03:07, 5.26it/s] 58%|█████▊ | 28080/48008 [1:36:17<1:02:49, 5.29it/s] 58%|█████▊ | 28081/48008 [1:36:17<1:03:06, 5.26it/s] 58%|█████▊ | 28082/48008 [1:36:17<1:03:20, 5.24it/s] 58%|█████▊ | 28083/48008 [1:36:17<1:03:00, 5.27it/s] 58%|█████▊ | 28084/48008 [1:36:17<1:02:59, 5.27it/s] 59%|█████▊ | 28085/48008 [1:36:18<1:03:45, 5.21it/s] 59%|█████▊ | 28086/48008 [1:36:18<1:03:27, 5.23it/s] 59%|█████▊ | 28087/48008 [1:36:18<1:03:52, 5.20it/s] 59%|█████▊ | 28088/48008 [1:36:18<1:04:05, 5.18it/s] 59%|█████▊ | 28089/48008 [1:36:18<1:04:18, 5.16it/s] 59%|█████▊ | 28090/48008 [1:36:19<1:04:20, 5.16it/s] 59%|█████▊ | 28091/48008 [1:36:19<1:04:09, 5.17it/s] 59%|█████▊ | 28092/48008 [1:36:19<1:03:50, 5.20it/s] 59%|█████▊ | 28093/48008 [1:36:19<1:04:05, 5.18it/s] 59%|█████▊ | 28094/48008 [1:36:19<1:03:53, 5.19it/s] 59%|█████▊ | 28095/48008 [1:36:20<1:15:17, 4.41it/s] 59%|█████▊ | 28096/48008 [1:36:20<1:12:10, 4.60it/s] 59%|█████▊ | 28097/48008 [1:36:20<1:10:17, 4.72it/s] 59%|█████▊ | 28098/48008 [1:36:20<1:08:23, 4.85it/s] 59%|█████▊ | 28099/48008 [1:36:20<1:08:09, 4.87it/s] 59%|█████▊ | 28100/48008 [1:36:21<1:06:48, 4.97it/s] {'loss': 0.3009, 'grad_norm': 0.1630859375, 'learning_rate': 8.294034327612065e-05, 'epoch': 0.59} 59%|█████▊ | 28100/48008 [1:36:21<1:06:48, 4.97it/s] 59%|█████▊ | 28101/48008 [1:36:21<1:05:52, 5.04it/s] 59%|█████▊ | 28102/48008 [1:36:21<1:10:03, 4.74it/s] 59%|█████▊ | 28103/48008 [1:36:21<1:08:54, 4.81it/s] 59%|█████▊ | 28104/48008 [1:36:21<1:07:35, 4.91it/s] 59%|█████▊ | 28105/48008 [1:36:22<1:06:40, 4.97it/s] 59%|█████▊ | 28106/48008 [1:36:22<1:05:27, 5.07it/s] 59%|█████▊ | 28107/48008 [1:36:22<1:05:52, 5.03it/s] 59%|█████▊ | 28108/48008 [1:36:22<1:06:17, 5.00it/s] 59%|█████▊ | 28109/48008 [1:36:23<1:16:56, 4.31it/s] 59%|█████▊ | 28110/48008 [1:36:23<1:14:00, 4.48it/s] 59%|█████▊ | 28111/48008 [1:36:23<1:10:45, 4.69it/s] 59%|█████▊ | 28112/48008 [1:36:23<1:08:41, 4.83it/s] 59%|█████▊ | 28113/48008 [1:36:23<1:06:58, 4.95it/s] 59%|█████▊ | 28114/48008 [1:36:23<1:05:51, 5.03it/s] 59%|█████▊ | 28115/48008 [1:36:24<1:05:19, 5.08it/s] 59%|█████▊ | 28116/48008 [1:36:24<1:04:43, 5.12it/s] 59%|█████▊ | 28117/48008 [1:36:24<1:04:23, 5.15it/s] 59%|█████▊ | 28118/48008 [1:36:24<1:15:33, 4.39it/s] 59%|█████▊ | 28119/48008 [1:36:25<1:11:46, 4.62it/s] 59%|█████▊ | 28120/48008 [1:36:25<1:08:06, 4.87it/s] 59%|█████▊ | 28121/48008 [1:36:25<1:06:33, 4.98it/s] 59%|█████▊ | 28122/48008 [1:36:25<1:05:57, 5.03it/s] 59%|█████▊ | 28123/48008 [1:36:25<1:16:45, 4.32it/s] 59%|█████▊ | 28124/48008 [1:36:26<1:12:44, 4.56it/s] 59%|█████▊ | 28125/48008 [1:36:26<1:10:33, 4.70it/s] 59%|█████▊ | 28126/48008 [1:36:26<1:10:19, 4.71it/s] 59%|█████▊ | 28127/48008 [1:36:26<1:09:34, 4.76it/s] 59%|█████▊ | 28128/48008 [1:36:26<1:07:41, 4.89it/s] 59%|█████▊ | 28129/48008 [1:36:27<1:06:24, 4.99it/s] 59%|█████▊ | 28130/48008 [1:36:27<1:05:03, 5.09it/s] 59%|█████▊ | 28131/48008 [1:36:27<1:04:52, 5.11it/s] 59%|█████▊ | 28132/48008 [1:36:27<1:04:22, 5.15it/s] 59%|█████▊ | 28133/48008 [1:36:27<1:04:17, 5.15it/s] 59%|█████▊ | 28134/48008 [1:36:28<1:03:57, 5.18it/s] 59%|█████▊ | 28135/48008 [1:36:28<1:04:46, 5.11it/s] 59%|█████▊ | 28136/48008 [1:36:28<1:04:56, 5.10it/s] 59%|█████▊ | 28137/48008 [1:36:28<1:04:07, 5.16it/s] 59%|█████▊ | 28138/48008 [1:36:29<1:23:14, 3.98it/s] 59%|█████▊ | 28139/48008 [1:36:29<1:18:44, 4.21it/s] 59%|█████▊ | 28140/48008 [1:36:29<1:14:07, 4.47it/s] 59%|█████▊ | 28141/48008 [1:36:29<1:10:49, 4.68it/s] 59%|█████▊ | 28142/48008 [1:36:29<1:07:55, 4.87it/s] 59%|█████▊ | 28143/48008 [1:36:30<1:06:25, 4.98it/s] 59%|█████▊ | 28144/48008 [1:36:30<1:05:20, 5.07it/s] 59%|█████▊ | 28145/48008 [1:36:30<1:05:47, 5.03it/s] 59%|█████▊ | 28146/48008 [1:36:30<1:06:20, 4.99it/s] 59%|█████▊ | 28147/48008 [1:36:30<1:05:17, 5.07it/s] 59%|█████▊ | 28148/48008 [1:36:30<1:04:27, 5.13it/s] 59%|█████▊ | 28149/48008 [1:36:31<1:04:06, 5.16it/s] 59%|█████▊ | 28150/48008 [1:36:31<1:03:53, 5.18it/s] {'loss': 0.3016, 'grad_norm': 0.244140625, 'learning_rate': 8.273204465922346e-05, 'epoch': 0.59} 59%|█████▊ | 28150/48008 [1:36:31<1:03:53, 5.18it/s] 59%|█████▊ | 28151/48008 [1:36:31<1:03:39, 5.20it/s] 59%|█████▊ | 28152/48008 [1:36:31<1:05:19, 5.07it/s] 59%|█████▊ | 28153/48008 [1:36:31<1:06:38, 4.97it/s] 59%|█████▊ | 28154/48008 [1:36:32<1:05:36, 5.04it/s] 59%|█████▊ | 28155/48008 [1:36:32<1:04:30, 5.13it/s] 59%|█████▊ | 28156/48008 [1:36:32<1:04:47, 5.11it/s] 59%|█████▊ | 28157/48008 [1:36:32<1:05:00, 5.09it/s] 59%|█████▊ | 28158/48008 [1:36:32<1:04:36, 5.12it/s] 59%|█████▊ | 28159/48008 [1:36:33<1:04:19, 5.14it/s] 59%|█████▊ | 28160/48008 [1:36:33<1:05:07, 5.08it/s] 59%|█████▊ | 28161/48008 [1:36:33<1:04:13, 5.15it/s] 59%|█████▊ | 28162/48008 [1:36:33<1:03:25, 5.22it/s] 59%|█████▊ | 28163/48008 [1:36:33<1:08:02, 4.86it/s] 59%|█████▊ | 28164/48008 [1:36:34<1:06:37, 4.96it/s] 59%|█████▊ | 28165/48008 [1:36:34<1:24:55, 3.89it/s] 59%|█████▊ | 28166/48008 [1:36:34<1:18:10, 4.23it/s] 59%|█████▊ | 28167/48008 [1:36:34<1:13:38, 4.49it/s] 59%|█████▊ | 28168/48008 [1:36:35<1:09:58, 4.73it/s] 59%|█████▊ | 28169/48008 [1:36:35<1:08:26, 4.83it/s] 59%|█████▊ | 28170/48008 [1:36:35<1:07:21, 4.91it/s] 59%|█████▊ | 28171/48008 [1:36:35<1:06:27, 4.98it/s] 59%|█████▊ | 28172/48008 [1:36:35<1:05:17, 5.06it/s] 59%|█████▊ | 28173/48008 [1:36:36<1:03:55, 5.17it/s] 59%|█████▊ | 28174/48008 [1:36:36<1:03:28, 5.21it/s] 59%|█████▊ | 28175/48008 [1:36:36<1:04:29, 5.13it/s] 59%|█████▊ | 28176/48008 [1:36:36<1:05:48, 5.02it/s] 59%|█████▊ | 28177/48008 [1:36:36<1:04:41, 5.11it/s] 59%|█████▊ | 28178/48008 [1:36:37<1:23:37, 3.95it/s] 59%|█████▊ | 28179/48008 [1:36:37<1:36:37, 3.42it/s] 59%|█████▊ | 28180/48008 [1:36:37<1:26:07, 3.84it/s] 59%|█████▊ | 28181/48008 [1:36:37<1:19:17, 4.17it/s] 59%|█████▊ | 28182/48008 [1:36:38<1:14:20, 4.45it/s] 59%|█████▊ | 28183/48008 [1:36:38<1:11:23, 4.63it/s] 59%|█████▊ | 28184/48008 [1:36:38<1:07:50, 4.87it/s] 59%|█████▊ | 28185/48008 [1:36:38<1:08:19, 4.84it/s] 59%|█████▊ | 28186/48008 [1:36:38<1:06:38, 4.96it/s] 59%|█████▊ | 28187/48008 [1:36:39<1:05:29, 5.04it/s] 59%|█████▊ | 28188/48008 [1:36:39<1:04:28, 5.12it/s] 59%|█████▊ | 28189/48008 [1:36:39<1:08:36, 4.81it/s] 59%|█████▊ | 28190/48008 [1:36:39<1:06:36, 4.96it/s] 59%|█████▊ | 28191/48008 [1:36:39<1:05:10, 5.07it/s] 59%|█████▊ | 28192/48008 [1:36:40<1:06:14, 4.99it/s] 59%|█████▊ | 28193/48008 [1:36:40<1:04:53, 5.09it/s] 59%|█████▊ | 28194/48008 [1:36:40<1:03:52, 5.17it/s] 59%|█████▊ | 28195/48008 [1:36:41<2:40:29, 2.06it/s] 59%|█████▊ | 28196/48008 [1:36:41<2:12:27, 2.49it/s] 59%|█████▊ | 28197/48008 [1:36:42<2:10:54, 2.52it/s] 59%|█████▊ | 28198/48008 [1:36:42<1:50:30, 2.99it/s] 59%|█████▊ | 28199/48008 [1:36:42<1:36:10, 3.43it/s] 59%|█████▊ | 28200/48008 [1:36:42<1:26:46, 3.80it/s] {'loss': 0.3018, 'grad_norm': 0.435546875, 'learning_rate': 8.252374604232627e-05, 'epoch': 0.59} 59%|█████▊ | 28200/48008 [1:36:42<1:26:46, 3.80it/s] 59%|█████▊ | 28201/48008 [1:36:43<1:19:48, 4.14it/s] 59%|█████▊ | 28202/48008 [1:36:43<1:14:12, 4.45it/s] 59%|█████▊ | 28203/48008 [1:36:43<1:10:45, 4.66it/s] 59%|█████▊ | 28204/48008 [1:36:43<1:09:31, 4.75it/s] 59%|█████▉ | 28205/48008 [1:36:43<1:08:10, 4.84it/s] 59%|█████▉ | 28206/48008 [1:36:44<1:25:58, 3.84it/s] 59%|█████▉ | 28207/48008 [1:36:44<1:18:44, 4.19it/s] 59%|█████▉ | 28208/48008 [1:36:44<1:13:32, 4.49it/s] 59%|█████▉ | 28209/48008 [1:36:44<1:11:43, 4.60it/s] 59%|█████▉ | 28210/48008 [1:36:44<1:09:13, 4.77it/s] 59%|█████▉ | 28211/48008 [1:36:45<1:11:57, 4.59it/s] 59%|█████▉ | 28212/48008 [1:36:45<1:09:46, 4.73it/s] 59%|█████▉ | 28213/48008 [1:36:45<1:08:44, 4.80it/s] 59%|█████▉ | 28214/48008 [1:36:45<1:07:07, 4.91it/s] 59%|█████▉ | 28215/48008 [1:36:46<1:05:57, 5.00it/s] 59%|█████▉ | 28216/48008 [1:36:46<1:24:08, 3.92it/s] 59%|█████▉ | 28217/48008 [1:36:46<1:29:21, 3.69it/s] 59%|█████▉ | 28218/48008 [1:36:46<1:21:19, 4.06it/s] 59%|█████▉ | 28219/48008 [1:36:47<1:16:31, 4.31it/s] 59%|█████▉ | 28220/48008 [1:36:47<1:12:40, 4.54it/s] 59%|█████▉ | 28221/48008 [1:36:47<1:10:08, 4.70it/s] 59%|█████▉ | 28222/48008 [1:36:47<1:07:15, 4.90it/s] 59%|█████▉ | 28223/48008 [1:36:47<1:05:26, 5.04it/s] 59%|█████▉ | 28224/48008 [1:36:48<1:04:08, 5.14it/s] 59%|█████▉ | 28225/48008 [1:36:48<1:03:48, 5.17it/s] 59%|█████▉ | 28226/48008 [1:36:48<1:03:34, 5.19it/s] 59%|█████▉ | 28227/48008 [1:36:48<1:04:44, 5.09it/s] 59%|█████▉ | 28228/48008 [1:36:48<1:04:20, 5.12it/s] 59%|█████▉ | 28229/48008 [1:36:48<1:03:21, 5.20it/s] 59%|█████▉ | 28230/48008 [1:36:49<1:05:21, 5.04it/s] 59%|█████▉ | 28231/48008 [1:36:49<1:04:33, 5.11it/s] 59%|█████▉ | 28232/48008 [1:36:49<1:04:01, 5.15it/s] 59%|█████▉ | 28233/48008 [1:36:49<1:03:31, 5.19it/s] 59%|█████▉ | 28234/48008 [1:36:49<1:03:06, 5.22it/s] 59%|█████▉ | 28235/48008 [1:36:50<1:03:10, 5.22it/s] 59%|█████▉ | 28236/48008 [1:36:50<1:02:57, 5.23it/s] 59%|█████▉ | 28237/48008 [1:36:50<1:03:14, 5.21it/s] 59%|█████▉ | 28238/48008 [1:36:50<1:03:18, 5.20it/s] 59%|█████▉ | 28239/48008 [1:36:50<1:03:14, 5.21it/s] 59%|█████▉ | 28240/48008 [1:36:51<1:03:05, 5.22it/s] 59%|█████▉ | 28241/48008 [1:36:51<1:22:19, 4.00it/s] 59%|█████▉ | 28242/48008 [1:36:51<1:16:45, 4.29it/s] 59%|█████▉ | 28243/48008 [1:36:51<1:12:38, 4.53it/s] 59%|█████▉ | 28244/48008 [1:36:52<1:09:10, 4.76it/s] 59%|█████▉ | 28245/48008 [1:36:52<1:07:43, 4.86it/s] 59%|█████▉ | 28246/48008 [1:36:52<1:06:10, 4.98it/s] 59%|█████▉ | 28247/48008 [1:36:52<1:05:00, 5.07it/s] 59%|█████▉ | 28248/48008 [1:36:52<1:04:01, 5.14it/s] 59%|█████▉ | 28249/48008 [1:36:53<1:03:42, 5.17it/s] 59%|█████▉ | 28250/48008 [1:36:53<1:03:10, 5.21it/s] {'loss': 0.3017, 'grad_norm': 0.1650390625, 'learning_rate': 8.23154474254291e-05, 'epoch': 0.59} 59%|█████▉ | 28250/48008 [1:36:53<1:03:10, 5.21it/s] 59%|█████▉ | 28251/48008 [1:36:53<1:02:59, 5.23it/s] 59%|█████▉ | 28252/48008 [1:36:53<1:03:43, 5.17it/s] 59%|█████▉ | 28253/48008 [1:36:53<1:05:24, 5.03it/s] 59%|█████▉ | 28254/48008 [1:36:54<1:05:40, 5.01it/s] 59%|█████▉ | 28255/48008 [1:36:54<1:06:08, 4.98it/s] 59%|█████▉ | 28256/48008 [1:36:54<1:05:19, 5.04it/s] 59%|█████▉ | 28257/48008 [1:36:54<1:04:13, 5.13it/s] 59%|█████▉ | 28258/48008 [1:36:54<1:05:05, 5.06it/s] 59%|█████▉ | 28259/48008 [1:36:54<1:04:03, 5.14it/s] 59%|█████▉ | 28260/48008 [1:36:55<1:03:33, 5.18it/s] 59%|█████▉ | 28261/48008 [1:36:55<1:03:38, 5.17it/s] 59%|█████▉ | 28262/48008 [1:36:55<1:02:59, 5.22it/s] 59%|█████▉ | 28263/48008 [1:36:55<1:02:50, 5.24it/s] 59%|█████▉ | 28264/48008 [1:36:55<1:03:11, 5.21it/s] 59%|█████▉ | 28265/48008 [1:36:56<1:04:08, 5.13it/s] 59%|█████▉ | 28266/48008 [1:36:56<1:04:16, 5.12it/s] 59%|█████▉ | 28267/48008 [1:36:56<1:04:30, 5.10it/s] 59%|█████▉ | 28268/48008 [1:36:56<1:04:50, 5.07it/s] 59%|█████▉ | 28269/48008 [1:36:56<1:05:10, 5.05it/s] 59%|█████▉ | 28270/48008 [1:36:57<1:04:08, 5.13it/s] 59%|█████▉ | 28271/48008 [1:36:57<1:03:42, 5.16it/s] 59%|█████▉ | 28272/48008 [1:36:57<1:03:29, 5.18it/s] 59%|█████▉ | 28273/48008 [1:36:57<1:03:18, 5.19it/s] 59%|█████▉ | 28274/48008 [1:36:57<1:03:47, 5.16it/s] 59%|█████▉ | 28275/48008 [1:36:58<1:22:42, 3.98it/s] 59%|█████▉ | 28276/48008 [1:36:58<1:16:31, 4.30it/s] 59%|█████▉ | 28277/48008 [1:36:58<1:11:41, 4.59it/s] 59%|█████▉ | 28278/48008 [1:36:58<1:10:05, 4.69it/s] 59%|█████▉ | 28279/48008 [1:36:59<1:08:17, 4.81it/s] 59%|█████▉ | 28280/48008 [1:36:59<1:06:24, 4.95it/s] 59%|█████▉ | 28281/48008 [1:36:59<1:05:18, 5.03it/s] 59%|█████▉ | 28282/48008 [1:36:59<1:04:09, 5.12it/s] 59%|█████▉ | 28283/48008 [1:36:59<1:02:48, 5.23it/s] 59%|█████▉ | 28284/48008 [1:36:59<1:02:54, 5.23it/s] 59%|█████▉ | 28285/48008 [1:37:00<1:22:03, 4.01it/s] 59%|█████▉ | 28286/48008 [1:37:00<1:16:46, 4.28it/s] 59%|█████▉ | 28287/48008 [1:37:00<1:12:40, 4.52it/s] 59%|█████▉ | 28288/48008 [1:37:00<1:09:47, 4.71it/s] 59%|█████▉ | 28289/48008 [1:37:01<1:07:14, 4.89it/s] 59%|█████▉ | 28290/48008 [1:37:01<1:06:00, 4.98it/s] 59%|█████▉ | 28291/48008 [1:37:01<1:04:47, 5.07it/s] 59%|█████▉ | 28292/48008 [1:37:01<1:04:09, 5.12it/s] 59%|█████▉ | 28293/48008 [1:37:01<1:08:20, 4.81it/s] 59%|█████▉ | 28294/48008 [1:37:02<1:07:25, 4.87it/s] 59%|█████▉ | 28295/48008 [1:37:02<1:05:59, 4.98it/s] 59%|█████▉ | 28296/48008 [1:37:02<1:05:29, 5.02it/s] 59%|█████▉ | 28297/48008 [1:37:02<1:05:41, 5.00it/s] 59%|█████▉ | 28298/48008 [1:37:03<1:16:07, 4.32it/s] 59%|█████▉ | 28299/48008 [1:37:03<1:12:02, 4.56it/s] 59%|█████▉ | 28300/48008 [1:37:03<1:09:42, 4.71it/s] 59%|█████▉ | 28300/48008 [1:37:03<1:09:42, 4.71it/s]{'loss': 0.2989, 'grad_norm': 0.3046875, 'learning_rate': 8.210714880853192e-05, 'epoch': 0.59} 59%|█████▉ | 28301/48008 [1:37:03<1:07:41, 4.85it/s] 59%|█████▉ | 28302/48008 [1:37:03<1:06:21, 4.95it/s] 59%|█████▉ | 28303/48008 [1:37:04<1:04:52, 5.06it/s] 59%|█████▉ | 28304/48008 [1:37:04<1:03:39, 5.16it/s] 59%|█████▉ | 28305/48008 [1:37:04<1:02:54, 5.22it/s] 59%|█████▉ | 28306/48008 [1:37:04<1:02:51, 5.22it/s] 59%|█████▉ | 28307/48008 [1:37:04<1:02:23, 5.26it/s] 59%|█████▉ | 28308/48008 [1:37:04<1:03:37, 5.16it/s] 59%|█████▉ | 28309/48008 [1:37:05<1:04:32, 5.09it/s] 59%|█████▉ | 28310/48008 [1:37:05<1:04:30, 5.09it/s] 59%|█████▉ | 28311/48008 [1:37:05<1:04:35, 5.08it/s] 59%|█████▉ | 28312/48008 [1:37:05<1:04:43, 5.07it/s] 59%|█████▉ | 28313/48008 [1:37:05<1:05:32, 5.01it/s] 59%|█████▉ | 28314/48008 [1:37:06<1:04:46, 5.07it/s] 59%|█████▉ | 28315/48008 [1:37:06<1:04:31, 5.09it/s] 59%|█████▉ | 28316/48008 [1:37:06<1:05:48, 4.99it/s] 59%|█████▉ | 28317/48008 [1:37:06<1:04:48, 5.06it/s] 59%|█████▉ | 28318/48008 [1:37:06<1:05:22, 5.02it/s] 59%|█████▉ | 28319/48008 [1:37:07<1:04:45, 5.07it/s] 59%|█████▉ | 28320/48008 [1:37:07<1:05:24, 5.02it/s] 59%|█████▉ | 28321/48008 [1:37:07<1:04:24, 5.09it/s] 59%|█████▉ | 28322/48008 [1:37:07<1:15:14, 4.36it/s] 59%|█████▉ | 28323/48008 [1:37:08<1:10:59, 4.62it/s] 59%|█████▉ | 28324/48008 [1:37:08<1:08:58, 4.76it/s] 59%|█████▉ | 28325/48008 [1:37:08<1:18:27, 4.18it/s] 59%|█████▉ | 28326/48008 [1:37:08<1:13:33, 4.46it/s] 59%|█████▉ | 28327/48008 [1:37:08<1:10:30, 4.65it/s] 59%|█████▉ | 28328/48008 [1:37:09<1:09:25, 4.72it/s] 59%|█████▉ | 28329/48008 [1:37:09<1:07:39, 4.85it/s] 59%|█████▉ | 28330/48008 [1:37:09<1:17:25, 4.24it/s] 59%|█████▉ | 28331/48008 [1:37:09<1:12:37, 4.52it/s] 59%|█████▉ | 28332/48008 [1:37:09<1:09:51, 4.69it/s] 59%|█████▉ | 28333/48008 [1:37:10<1:07:49, 4.84it/s] 59%|█████▉ | 28334/48008 [1:37:10<1:06:06, 4.96it/s] 59%|█████▉ | 28335/48008 [1:37:10<1:04:58, 5.05it/s] 59%|█████▉ | 28336/48008 [1:37:10<1:04:24, 5.09it/s] 59%|█████▉ | 28337/48008 [1:37:10<1:04:17, 5.10it/s] 59%|█████▉ | 28338/48008 [1:37:11<1:05:28, 5.01it/s] 59%|█████▉ | 28339/48008 [1:37:11<1:09:21, 4.73it/s] 59%|█████▉ | 28340/48008 [1:37:11<1:07:10, 4.88it/s] 59%|█████▉ | 28341/48008 [1:37:11<1:05:40, 4.99it/s] 59%|█████▉ | 28342/48008 [1:37:11<1:04:23, 5.09it/s] 59%|█████▉ | 28343/48008 [1:37:12<1:03:44, 5.14it/s] 59%|█████▉ | 28344/48008 [1:37:12<1:02:58, 5.20it/s] 59%|█████▉ | 28345/48008 [1:37:12<1:03:49, 5.13it/s] 59%|█████▉ | 28346/48008 [1:37:12<1:03:36, 5.15it/s] 59%|█████▉ | 28347/48008 [1:37:12<1:04:27, 5.08it/s] 59%|█████▉ | 28348/48008 [1:37:13<1:03:55, 5.13it/s] 59%|█████▉ | 28349/48008 [1:37:13<1:04:52, 5.05it/s] 59%|█████▉ | 28350/48008 [1:37:13<1:03:51, 5.13it/s] {'loss': 0.3009, 'grad_norm': 0.1591796875, 'learning_rate': 8.189885019163474e-05, 'epoch': 0.59} 59%|█████▉ | 28350/48008 [1:37:13<1:03:51, 5.13it/s] 59%|█████▉ | 28351/48008 [1:37:13<1:03:35, 5.15it/s] 59%|█████▉ | 28352/48008 [1:37:13<1:03:03, 5.20it/s] 59%|█████▉ | 28353/48008 [1:37:14<1:02:42, 5.22it/s] 59%|█████▉ | 28354/48008 [1:37:14<1:22:07, 3.99it/s] 59%|█████▉ | 28355/48008 [1:37:14<1:16:21, 4.29it/s] 59%|█████▉ | 28356/48008 [1:37:14<1:11:39, 4.57it/s] 59%|█████▉ | 28357/48008 [1:37:15<1:09:00, 4.75it/s] 59%|█████▉ | 28358/48008 [1:37:15<1:06:39, 4.91it/s] 59%|█████▉ | 28359/48008 [1:37:15<1:05:33, 5.00it/s] 59%|█████▉ | 28360/48008 [1:37:15<1:04:38, 5.07it/s] 59%|█████▉ | 28361/48008 [1:37:15<1:05:07, 5.03it/s] 59%|█████▉ | 28362/48008 [1:37:16<1:04:39, 5.06it/s] 59%|█████▉ | 28363/48008 [1:37:16<1:03:58, 5.12it/s] 59%|█████▉ | 28364/48008 [1:37:16<1:03:17, 5.17it/s] 59%|█████▉ | 28365/48008 [1:37:16<1:03:20, 5.17it/s] 59%|█████▉ | 28366/48008 [1:37:16<1:03:30, 5.15it/s] 59%|█████▉ | 28367/48008 [1:37:16<1:03:59, 5.11it/s] 59%|█████▉ | 28368/48008 [1:37:17<1:08:08, 4.80it/s] 59%|█████▉ | 28369/48008 [1:37:17<1:06:23, 4.93it/s] 59%|█████▉ | 28370/48008 [1:37:17<1:05:21, 5.01it/s] 59%|█████▉ | 28371/48008 [1:37:17<1:06:18, 4.94it/s] 59%|█████▉ | 28372/48008 [1:37:18<1:05:48, 4.97it/s] 59%|█████▉ | 28373/48008 [1:37:18<1:04:45, 5.05it/s] 59%|█████▉ | 28374/48008 [1:37:18<1:04:11, 5.10it/s] 59%|█████▉ | 28375/48008 [1:37:18<1:03:41, 5.14it/s] 59%|█████▉ | 28376/48008 [1:37:18<1:03:08, 5.18it/s] 59%|█████▉ | 28377/48008 [1:37:18<1:02:44, 5.21it/s] 59%|█████▉ | 28378/48008 [1:37:19<1:02:51, 5.20it/s] 59%|█████▉ | 28379/48008 [1:37:19<1:02:35, 5.23it/s] 59%|█████▉ | 28380/48008 [1:37:19<1:02:46, 5.21it/s] 59%|█████▉ | 28381/48008 [1:37:19<1:03:53, 5.12it/s] 59%|█████▉ | 28382/48008 [1:37:19<1:04:38, 5.06it/s] 59%|█████▉ | 28383/48008 [1:37:20<1:03:32, 5.15it/s] 59%|█████▉ | 28384/48008 [1:37:20<1:03:45, 5.13it/s] 59%|█████▉ | 28385/48008 [1:37:20<1:02:46, 5.21it/s] 59%|█████▉ | 28386/48008 [1:37:20<1:02:16, 5.25it/s] 59%|█████▉ | 28387/48008 [1:37:20<1:04:05, 5.10it/s] 59%|█████▉ | 28388/48008 [1:37:21<1:04:47, 5.05it/s] 59%|█████▉ | 28389/48008 [1:37:21<1:05:08, 5.02it/s] 59%|█████▉ | 28390/48008 [1:37:21<1:05:12, 5.01it/s] 59%|█████▉ | 28391/48008 [1:37:21<1:04:56, 5.03it/s] 59%|█████▉ | 28392/48008 [1:37:21<1:04:17, 5.08it/s] 59%|█████▉ | 28393/48008 [1:37:22<1:04:09, 5.10it/s] 59%|█████▉ | 28394/48008 [1:37:22<1:04:02, 5.10it/s] 59%|█████▉ | 28395/48008 [1:37:22<1:03:30, 5.15it/s] 59%|█████▉ | 28396/48008 [1:37:22<1:04:42, 5.05it/s] 59%|█████▉ | 28397/48008 [1:37:22<1:04:50, 5.04it/s] 59%|█████▉ | 28398/48008 [1:37:23<1:05:49, 4.97it/s] 59%|█████▉ | 28399/48008 [1:37:23<1:05:10, 5.01it/s] 59%|█████▉ | 28400/48008 [1:37:23<1:04:50, 5.04it/s] {'loss': 0.3005, 'grad_norm': 0.298828125, 'learning_rate': 8.169055157473756e-05, 'epoch': 0.59} 59%|█████▉ | 28400/48008 [1:37:23<1:04:50, 5.04it/s] 59%|█████▉ | 28401/48008 [1:37:23<1:03:59, 5.11it/s] 59%|█████▉ | 28402/48008 [1:37:23<1:05:41, 4.97it/s] 59%|█████▉ | 28403/48008 [1:37:24<1:05:01, 5.02it/s] 59%|█████▉ | 28404/48008 [1:37:24<1:04:37, 5.06it/s] 59%|█████▉ | 28405/48008 [1:37:24<1:03:43, 5.13it/s] 59%|█████▉ | 28406/48008 [1:37:24<1:04:35, 5.06it/s] 59%|█████▉ | 28407/48008 [1:37:24<1:04:01, 5.10it/s] 59%|█████▉ | 28408/48008 [1:37:25<1:03:44, 5.12it/s] 59%|█████▉ | 28409/48008 [1:37:25<1:03:25, 5.15it/s] 59%|█████▉ | 28410/48008 [1:37:25<1:22:13, 3.97it/s] 59%|█████▉ | 28411/48008 [1:37:25<1:27:53, 3.72it/s] 59%|█████▉ | 28412/48008 [1:37:26<1:20:16, 4.07it/s] 59%|█████▉ | 28413/48008 [1:37:26<1:16:10, 4.29it/s] 59%|█████▉ | 28414/48008 [1:37:26<1:12:08, 4.53it/s] 59%|█████▉ | 28415/48008 [1:37:26<1:09:00, 4.73it/s] 59%|█████▉ | 28416/48008 [1:37:26<1:07:17, 4.85it/s] 59%|█████▉ | 28417/48008 [1:37:27<1:06:17, 4.93it/s] 59%|█████▉ | 28418/48008 [1:37:27<1:05:15, 5.00it/s] 59%|█████▉ | 28419/48008 [1:37:27<1:05:27, 4.99it/s] 59%|█████▉ | 28420/48008 [1:37:27<1:16:03, 4.29it/s] 59%|█████▉ | 28421/48008 [1:37:28<1:11:50, 4.54it/s] 59%|█████▉ | 28422/48008 [1:37:28<1:09:10, 4.72it/s] 59%|█████▉ | 28423/48008 [1:37:28<1:07:02, 4.87it/s] 59%|█████▉ | 28424/48008 [1:37:28<1:06:03, 4.94it/s] 59%|█████▉ | 28425/48008 [1:37:28<1:05:02, 5.02it/s] 59%|█████▉ | 28426/48008 [1:37:28<1:05:19, 5.00it/s] 59%|█████▉ | 28427/48008 [1:37:29<1:04:18, 5.07it/s] 59%|█████▉ | 28428/48008 [1:37:29<1:03:17, 5.16it/s] 59%|█████▉ | 28429/48008 [1:37:29<1:02:53, 5.19it/s] 59%|█████▉ | 28430/48008 [1:37:29<1:02:44, 5.20it/s] 59%|█████▉ | 28431/48008 [1:37:29<1:02:53, 5.19it/s] 59%|█████▉ | 28432/48008 [1:37:30<1:02:51, 5.19it/s] 59%|█████▉ | 28433/48008 [1:37:30<1:03:56, 5.10it/s] 59%|█████▉ | 28434/48008 [1:37:30<1:03:45, 5.12it/s] 59%|█████▉ | 28435/48008 [1:37:30<1:03:37, 5.13it/s] 59%|█████▉ | 28436/48008 [1:37:30<1:03:15, 5.16it/s] 59%|█████▉ | 28437/48008 [1:37:31<1:02:52, 5.19it/s] 59%|█████▉ | 28438/48008 [1:37:31<1:03:27, 5.14it/s] 59%|█████▉ | 28439/48008 [1:37:31<1:03:05, 5.17it/s] 59%|█████▉ | 28440/48008 [1:37:31<1:03:37, 5.13it/s] 59%|█████▉ | 28441/48008 [1:37:31<1:03:59, 5.10it/s] 59%|█████▉ | 28442/48008 [1:37:32<1:03:12, 5.16it/s] 59%|█████▉ | 28443/48008 [1:37:32<1:02:31, 5.22it/s] 59%|█████▉ | 28444/48008 [1:37:32<1:02:38, 5.21it/s] 59%|█████▉ | 28445/48008 [1:37:32<1:21:29, 4.00it/s] 59%|█████▉ | 28446/48008 [1:37:33<1:16:53, 4.24it/s] 59%|█████▉ | 28447/48008 [1:37:33<1:12:22, 4.50it/s] 59%|█████▉ | 28448/48008 [1:37:33<1:09:17, 4.70it/s] 59%|█████▉ | 28449/48008 [1:37:33<1:06:48, 4.88it/s] 59%|█████▉ | 28450/48008 [1:37:33<1:05:19, 4.99it/s] {'loss': 0.3016, 'grad_norm': 0.1982421875, 'learning_rate': 8.148225295784036e-05, 'epoch': 0.59} 59%|█████▉ | 28450/48008 [1:37:33<1:05:19, 4.99it/s] 59%|█████▉ | 28451/48008 [1:37:33<1:04:01, 5.09it/s] 59%|█████▉ | 28452/48008 [1:37:34<1:03:25, 5.14it/s] 59%|█████▉ | 28453/48008 [1:37:34<1:02:52, 5.18it/s] 59%|█████▉ | 28454/48008 [1:37:34<1:04:44, 5.03it/s] 59%|█████▉ | 28455/48008 [1:37:34<1:03:53, 5.10it/s] 59%|█████▉ | 28456/48008 [1:37:34<1:03:38, 5.12it/s] 59%|█████▉ | 28457/48008 [1:37:35<1:02:36, 5.20it/s] 59%|█████▉ | 28458/48008 [1:37:35<1:02:25, 5.22it/s] 59%|█████▉ | 28459/48008 [1:37:35<1:04:16, 5.07it/s] 59%|█████▉ | 28460/48008 [1:37:35<1:03:51, 5.10it/s] 59%|█████▉ | 28461/48008 [1:37:35<1:03:21, 5.14it/s] 59%|█████▉ | 28462/48008 [1:37:36<1:03:02, 5.17it/s] 59%|█████▉ | 28463/48008 [1:37:36<1:03:29, 5.13it/s] 59%|█████▉ | 28464/48008 [1:37:36<1:02:44, 5.19it/s] 59%|█████▉ | 28465/48008 [1:37:36<1:03:46, 5.11it/s] 59%|█████▉ | 28466/48008 [1:37:36<1:05:01, 5.01it/s] 59%|█████▉ | 28467/48008 [1:37:37<1:04:04, 5.08it/s] 59%|█████▉ | 28468/48008 [1:37:37<1:03:39, 5.12it/s] 59%|█████▉ | 28469/48008 [1:37:37<1:04:36, 5.04it/s] 59%|█████▉ | 28470/48008 [1:37:37<1:03:24, 5.14it/s] 59%|█████▉ | 28471/48008 [1:37:37<1:02:54, 5.18it/s] 59%|█████▉ | 28472/48008 [1:37:38<1:04:30, 5.05it/s] 59%|█████▉ | 28473/48008 [1:37:38<1:03:32, 5.12it/s] 59%|█████▉ | 28474/48008 [1:37:38<1:02:53, 5.18it/s] 59%|█████▉ | 28475/48008 [1:37:38<1:02:54, 5.17it/s] 59%|█████▉ | 28476/48008 [1:37:38<1:07:16, 4.84it/s] 59%|█████▉ | 28477/48008 [1:37:39<1:05:21, 4.98it/s] 59%|█████▉ | 28478/48008 [1:37:39<1:05:20, 4.98it/s] 59%|█████▉ | 28479/48008 [1:37:39<1:04:09, 5.07it/s] 59%|█████▉ | 28480/48008 [1:37:39<1:03:08, 5.15it/s] 59%|█████▉ | 28481/48008 [1:37:39<1:14:09, 4.39it/s] 59%|█████▉ | 28482/48008 [1:37:40<1:10:10, 4.64it/s] 59%|█████▉ | 28483/48008 [1:37:40<1:09:57, 4.65it/s] 59%|█████▉ | 28484/48008 [1:37:40<1:07:17, 4.84it/s] 59%|█████▉ | 28485/48008 [1:37:40<1:10:13, 4.63it/s] 59%|█████▉ | 28486/48008 [1:37:40<1:08:20, 4.76it/s] 59%|█████▉ | 28487/48008 [1:37:41<1:06:27, 4.90it/s] 59%|█████▉ | 28488/48008 [1:37:41<1:05:08, 4.99it/s] 59%|█████▉ | 28489/48008 [1:37:41<1:04:26, 5.05it/s] 59%|█████▉ | 28490/48008 [1:37:41<1:03:51, 5.09it/s] 59%|█████▉ | 28491/48008 [1:37:41<1:03:00, 5.16it/s] 59%|█████▉ | 28492/48008 [1:37:42<1:02:33, 5.20it/s] 59%|█████▉ | 28493/48008 [1:37:42<1:13:34, 4.42it/s] 59%|█████▉ | 28494/48008 [1:37:42<1:12:02, 4.51it/s] 59%|█████▉ | 28495/48008 [1:37:42<1:08:58, 4.71it/s] 59%|█████▉ | 28496/48008 [1:37:43<1:06:43, 4.87it/s] 59%|█████▉ | 28497/48008 [1:37:43<1:07:14, 4.84it/s] 59%|█████▉ | 28498/48008 [1:37:43<1:05:23, 4.97it/s] 59%|█████▉ | 28499/48008 [1:37:43<1:04:38, 5.03it/s] 59%|█████▉ | 28500/48008 [1:37:43<1:03:22, 5.13it/s] 59%|█████▉ | 28500/48008 [1:37:43<1:03:22, 5.13it/s]{'loss': 0.3006, 'grad_norm': 0.5703125, 'learning_rate': 8.127395434094318e-05, 'epoch': 0.59} 59%|█████▉ | 28501/48008 [1:37:44<1:03:44, 5.10it/s] 59%|█████▉ | 28502/48008 [1:37:44<1:02:58, 5.16it/s] 59%|█████▉ | 28503/48008 [1:37:44<1:02:44, 5.18it/s] 59%|█████▉ | 28504/48008 [1:37:44<1:02:49, 5.17it/s] 59%|█████▉ | 28505/48008 [1:37:44<1:02:12, 5.23it/s] 59%|█████▉ | 28506/48008 [1:37:44<1:01:44, 5.26it/s] 59%|█████▉ | 28507/48008 [1:37:45<1:03:04, 5.15it/s] 59%|█████▉ | 28508/48008 [1:37:45<1:01:50, 5.25it/s] 59%|█████▉ | 28509/48008 [1:37:45<1:01:52, 5.25it/s] 59%|█████▉ | 28510/48008 [1:37:45<1:01:50, 5.25it/s] 59%|█████▉ | 28511/48008 [1:37:45<1:01:52, 5.25it/s] 59%|█████▉ | 28512/48008 [1:37:46<1:01:19, 5.30it/s] 59%|█████▉ | 28513/48008 [1:37:46<1:01:04, 5.32it/s] 59%|█████▉ | 28514/48008 [1:37:46<1:01:22, 5.29it/s] 59%|█████▉ | 28515/48008 [1:37:46<1:06:04, 4.92it/s] 59%|█████▉ | 28516/48008 [1:37:46<1:04:14, 5.06it/s] 59%|█████▉ | 28517/48008 [1:37:47<1:03:35, 5.11it/s] 59%|█████▉ | 28518/48008 [1:37:47<1:04:15, 5.06it/s] 59%|█████▉ | 28519/48008 [1:37:47<1:03:39, 5.10it/s] 59%|█████▉ | 28520/48008 [1:37:47<1:03:19, 5.13it/s] 59%|█████▉ | 28521/48008 [1:37:47<1:02:33, 5.19it/s] 59%|█████▉ | 28522/48008 [1:37:48<1:02:20, 5.21it/s] 59%|█████▉ | 28523/48008 [1:37:48<1:02:02, 5.23it/s] 59%|█████▉ | 28524/48008 [1:37:48<1:01:22, 5.29it/s] 59%|█████▉ | 28525/48008 [1:37:48<1:01:33, 5.27it/s] 59%|█████▉ | 28526/48008 [1:37:48<1:01:46, 5.26it/s] 59%|█████▉ | 28527/48008 [1:37:49<1:01:54, 5.25it/s] 59%|█████▉ | 28528/48008 [1:37:49<1:02:02, 5.23it/s] 59%|█████▉ | 28529/48008 [1:37:49<1:03:20, 5.13it/s] 59%|█████▉ | 28530/48008 [1:37:49<1:03:29, 5.11it/s] 59%|█████▉ | 28531/48008 [1:37:49<1:04:08, 5.06it/s] 59%|█████▉ | 28532/48008 [1:37:49<1:03:49, 5.09it/s] 59%|█████▉ | 28533/48008 [1:37:50<1:02:50, 5.16it/s] 59%|█████▉ | 28534/48008 [1:37:50<1:03:36, 5.10it/s] 59%|█████▉ | 28535/48008 [1:37:50<1:02:28, 5.19it/s] 59%|█████▉ | 28536/48008 [1:37:50<1:01:59, 5.24it/s] 59%|█████▉ | 28537/48008 [1:37:50<1:02:02, 5.23it/s] 59%|█████▉ | 28538/48008 [1:37:51<1:03:11, 5.14it/s] 59%|█████▉ | 28539/48008 [1:37:51<1:02:23, 5.20it/s] 59%|█████▉ | 28540/48008 [1:37:51<1:02:57, 5.15it/s] 59%|█████▉ | 28541/48008 [1:37:51<1:02:31, 5.19it/s] 59%|█████▉ | 28542/48008 [1:37:51<1:02:52, 5.16it/s] 59%|█████▉ | 28543/48008 [1:37:52<1:02:33, 5.19it/s] 59%|█████▉ | 28544/48008 [1:37:52<1:02:30, 5.19it/s] 59%|█████▉ | 28545/48008 [1:37:52<1:02:11, 5.22it/s] 59%|█████▉ | 28546/48008 [1:37:52<1:02:19, 5.20it/s] 59%|█████▉ | 28547/48008 [1:37:52<1:02:01, 5.23it/s] 59%|█████▉ | 28548/48008 [1:37:53<1:02:00, 5.23it/s] 59%|█████▉ | 28549/48008 [1:37:53<1:02:13, 5.21it/s] 59%|█████▉ | 28550/48008 [1:37:53<1:01:58, 5.23it/s] 59%|█████▉ | 28550/48008 [1:37:53<1:01:58, 5.23it/s]{'loss': 0.3039, 'grad_norm': 0.15625, 'learning_rate': 8.1065655724046e-05, 'epoch': 0.59} 59%|█████▉ | 28551/48008 [1:37:53<1:02:39, 5.18it/s] 59%|█████▉ | 28552/48008 [1:37:53<1:02:07, 5.22it/s] 59%|█████▉ | 28553/48008 [1:37:54<1:03:10, 5.13it/s] 59%|█████▉ | 28554/48008 [1:37:54<1:02:35, 5.18it/s] 59%|█████▉ | 28555/48008 [1:37:54<1:02:30, 5.19it/s] 59%|█████▉ | 28556/48008 [1:37:54<1:02:06, 5.22it/s] 59%|█████▉ | 28557/48008 [1:37:54<1:02:46, 5.16it/s] 59%|█████▉ | 28558/48008 [1:37:54<1:02:16, 5.21it/s] 59%|█████▉ | 28559/48008 [1:37:55<1:04:05, 5.06it/s] 59%|█████▉ | 28560/48008 [1:37:55<1:03:44, 5.08it/s] 59%|█████▉ | 28561/48008 [1:37:55<1:03:08, 5.13it/s] 59%|█████▉ | 28562/48008 [1:37:55<1:03:44, 5.09it/s] 59%|█████▉ | 28563/48008 [1:37:55<1:03:08, 5.13it/s] 59%|█████▉ | 28564/48008 [1:37:56<1:02:44, 5.17it/s] 60%|█████▉ | 28565/48008 [1:37:56<1:02:55, 5.15it/s] 60%|█████▉ | 28566/48008 [1:37:56<1:02:26, 5.19it/s] 60%|█████▉ | 28567/48008 [1:37:56<1:02:24, 5.19it/s] 60%|█████▉ | 28568/48008 [1:37:56<1:02:14, 5.21it/s] 60%|█████▉ | 28569/48008 [1:37:57<1:02:18, 5.20it/s] 60%|█████▉ | 28570/48008 [1:37:57<1:21:20, 3.98it/s] 60%|█████▉ | 28571/48008 [1:37:57<1:15:44, 4.28it/s] 60%|█████▉ | 28572/48008 [1:37:57<1:12:16, 4.48it/s] 60%|█████▉ | 28573/48008 [1:37:58<1:09:00, 4.69it/s] 60%|█████▉ | 28574/48008 [1:37:58<1:07:12, 4.82it/s] 60%|█████▉ | 28575/48008 [1:37:58<1:10:03, 4.62it/s] 60%|█████▉ | 28576/48008 [1:37:58<1:07:43, 4.78it/s] 60%|█████▉ | 28577/48008 [1:37:58<1:06:20, 4.88it/s] 60%|█████▉ | 28578/48008 [1:37:59<1:09:28, 4.66it/s] 60%|█████▉ | 28579/48008 [1:37:59<1:07:05, 4.83it/s] 60%|█████▉ | 28580/48008 [1:37:59<1:06:09, 4.89it/s] 60%|█████▉ | 28581/48008 [1:37:59<1:04:33, 5.01it/s] 60%|█████▉ | 28582/48008 [1:37:59<1:02:56, 5.14it/s] 60%|█████▉ | 28583/48008 [1:38:00<1:02:06, 5.21it/s] 60%|█████▉ | 28584/48008 [1:38:00<1:02:08, 5.21it/s] 60%|█████▉ | 28585/48008 [1:38:00<1:01:32, 5.26it/s] 60%|█████▉ | 28586/48008 [1:38:00<1:02:24, 5.19it/s] 60%|█████▉ | 28587/48008 [1:38:00<1:02:18, 5.19it/s] 60%|█████▉ | 28588/48008 [1:38:01<1:03:33, 5.09it/s] 60%|█████▉ | 28589/48008 [1:38:01<1:02:48, 5.15it/s] 60%|█████▉ | 28590/48008 [1:38:01<1:02:04, 5.21it/s] 60%|█████▉ | 28591/48008 [1:38:01<1:01:40, 5.25it/s] 60%|█████▉ | 28592/48008 [1:38:01<1:02:17, 5.19it/s] 60%|█████▉ | 28593/48008 [1:38:02<1:01:46, 5.24it/s] 60%|█████▉ | 28594/48008 [1:38:02<1:01:55, 5.23it/s] 60%|█████▉ | 28595/48008 [1:38:02<1:02:56, 5.14it/s] 60%|█████▉ | 28596/48008 [1:38:02<1:02:38, 5.17it/s] 60%|█████▉ | 28597/48008 [1:38:02<1:01:27, 5.26it/s] 60%|█████▉ | 28598/48008 [1:38:03<1:06:01, 4.90it/s] 60%|█████▉ | 28599/48008 [1:38:03<1:04:59, 4.98it/s] 60%|█████▉ | 28600/48008 [1:38:03<1:04:21, 5.03it/s] {'loss': 0.3015, 'grad_norm': 0.140625, 'learning_rate': 8.085735710714881e-05, 'epoch': 0.6} 60%|█████▉ | 28600/48008 [1:38:03<1:04:21, 5.03it/s] 60%|█████▉ | 28601/48008 [1:38:03<1:03:46, 5.07it/s] 60%|█████▉ | 28602/48008 [1:38:03<1:04:30, 5.01it/s] 60%|█████▉ | 28603/48008 [1:38:04<1:03:40, 5.08it/s] 60%|█████▉ | 28604/48008 [1:38:04<1:03:14, 5.11it/s] 60%|█████▉ | 28605/48008 [1:38:04<1:03:51, 5.06it/s] 60%|█████▉ | 28606/48008 [1:38:04<1:02:28, 5.18it/s] 60%|█████▉ | 28607/48008 [1:38:04<1:01:52, 5.23it/s] 60%|█████▉ | 28608/48008 [1:38:04<1:01:40, 5.24it/s] 60%|█████▉ | 28609/48008 [1:38:05<1:01:38, 5.25it/s] 60%|█████▉ | 28610/48008 [1:38:05<1:02:43, 5.15it/s] 60%|█████▉ | 28611/48008 [1:38:05<1:03:03, 5.13it/s] 60%|█████▉ | 28612/48008 [1:38:05<1:04:32, 5.01it/s] 60%|█████▉ | 28613/48008 [1:38:05<1:03:23, 5.10it/s] 60%|█████▉ | 28614/48008 [1:38:06<1:02:35, 5.16it/s] 60%|█████▉ | 28615/48008 [1:38:06<1:02:22, 5.18it/s] 60%|█████▉ | 28616/48008 [1:38:06<1:02:05, 5.21it/s] 60%|█████▉ | 28617/48008 [1:38:06<1:02:35, 5.16it/s] 60%|█████▉ | 28618/48008 [1:38:06<1:02:15, 5.19it/s] 60%|█████▉ | 28619/48008 [1:38:07<1:01:58, 5.21it/s] 60%|█████▉ | 28620/48008 [1:38:07<1:03:03, 5.12it/s] 60%|█████▉ | 28621/48008 [1:38:07<1:03:33, 5.08it/s] 60%|█████▉ | 28622/48008 [1:38:07<1:02:50, 5.14it/s] 60%|█████▉ | 28623/48008 [1:38:07<1:02:01, 5.21it/s] 60%|█████▉ | 28624/48008 [1:38:08<1:01:32, 5.25it/s] 60%|█████▉ | 28625/48008 [1:38:08<1:02:11, 5.19it/s] 60%|█████▉ | 28626/48008 [1:38:08<1:01:28, 5.25it/s] 60%|█████▉ | 28627/48008 [1:38:08<1:02:07, 5.20it/s] 60%|█████▉ | 28628/48008 [1:38:08<1:01:35, 5.24it/s] 60%|█████▉ | 28629/48008 [1:38:09<1:01:37, 5.24it/s] 60%|█████▉ | 28630/48008 [1:38:09<1:01:20, 5.27it/s] 60%|█████▉ | 28631/48008 [1:38:09<1:01:03, 5.29it/s] 60%|█████▉ | 28632/48008 [1:38:09<1:01:26, 5.26it/s] 60%|█████▉ | 28633/48008 [1:38:09<1:01:42, 5.23it/s] 60%|█████▉ | 28634/48008 [1:38:09<1:01:22, 5.26it/s] 60%|█████▉ | 28635/48008 [1:38:10<1:01:18, 5.27it/s] 60%|█████▉ | 28636/48008 [1:38:10<1:06:15, 4.87it/s] 60%|█████▉ | 28637/48008 [1:38:10<1:04:22, 5.02it/s] 60%|█████▉ | 28638/48008 [1:38:10<1:03:48, 5.06it/s] 60%|█████▉ | 28639/48008 [1:38:10<1:03:01, 5.12it/s] 60%|█████▉ | 28640/48008 [1:38:11<1:02:40, 5.15it/s] 60%|█████▉ | 28641/48008 [1:38:11<1:02:23, 5.17it/s] 60%|█████▉ | 28642/48008 [1:38:11<1:03:18, 5.10it/s] 60%|█████▉ | 28643/48008 [1:38:11<1:02:54, 5.13it/s] 60%|█████▉ | 28644/48008 [1:38:11<1:03:28, 5.08it/s] 60%|█████▉ | 28645/48008 [1:38:12<1:02:32, 5.16it/s] 60%|█████▉ | 28646/48008 [1:38:12<1:02:08, 5.19it/s] 60%|█████▉ | 28647/48008 [1:38:12<1:02:07, 5.19it/s] 60%|█████▉ | 28648/48008 [1:38:12<1:02:07, 5.19it/s] 60%|█████▉ | 28649/48008 [1:38:12<1:01:38, 5.23it/s] 60%|█████▉ | 28650/48008 [1:38:13<1:01:40, 5.23it/s] {'loss': 0.3012, 'grad_norm': 0.140625, 'learning_rate': 8.064905849025163e-05, 'epoch': 0.6} 60%|█████▉ | 28650/48008 [1:38:13<1:01:40, 5.23it/s] 60%|█████▉ | 28651/48008 [1:38:13<1:02:03, 5.20it/s] 60%|█████▉ | 28652/48008 [1:38:13<1:01:49, 5.22it/s] 60%|█████▉ | 28653/48008 [1:38:13<1:02:33, 5.16it/s] 60%|█████▉ | 28654/48008 [1:38:13<1:01:56, 5.21it/s] 60%|█████▉ | 28655/48008 [1:38:14<1:01:43, 5.23it/s] 60%|█████▉ | 28656/48008 [1:38:14<1:01:53, 5.21it/s] 60%|█████▉ | 28657/48008 [1:38:14<1:01:38, 5.23it/s] 60%|█████▉ | 28658/48008 [1:38:14<1:02:39, 5.15it/s] 60%|█████▉ | 28659/48008 [1:38:14<1:02:31, 5.16it/s] 60%|█████▉ | 28660/48008 [1:38:15<1:02:13, 5.18it/s] 60%|█████▉ | 28661/48008 [1:38:15<1:02:15, 5.18it/s] 60%|█████▉ | 28662/48008 [1:38:15<1:02:08, 5.19it/s] 60%|█████▉ | 28663/48008 [1:38:15<1:02:23, 5.17it/s] 60%|█████▉ | 28664/48008 [1:38:15<1:02:04, 5.19it/s] 60%|█████▉ | 28665/48008 [1:38:15<1:01:59, 5.20it/s] 60%|█████▉ | 28666/48008 [1:38:16<1:01:46, 5.22it/s] 60%|█████▉ | 28667/48008 [1:38:16<1:02:18, 5.17it/s] 60%|█████▉ | 28668/48008 [1:38:16<1:06:57, 4.81it/s] 60%|█████▉ | 28669/48008 [1:38:16<1:05:51, 4.89it/s] 60%|█████▉ | 28670/48008 [1:38:16<1:04:34, 4.99it/s] 60%|█████▉ | 28671/48008 [1:38:17<1:03:46, 5.05it/s] 60%|█████▉ | 28672/48008 [1:38:17<1:03:03, 5.11it/s] 60%|█████▉ | 28673/48008 [1:38:17<1:03:50, 5.05it/s] 60%|█████▉ | 28674/48008 [1:38:17<1:04:13, 5.02it/s] 60%|█████▉ | 28675/48008 [1:38:17<1:03:20, 5.09it/s] 60%|█████▉ | 28676/48008 [1:38:18<1:02:57, 5.12it/s] 60%|█████▉ | 28677/48008 [1:38:18<1:03:13, 5.10it/s] 60%|█████▉ | 28678/48008 [1:38:18<1:03:01, 5.11it/s] 60%|█████▉ | 28679/48008 [1:38:18<1:02:44, 5.13it/s] 60%|█████▉ | 28680/48008 [1:38:18<1:02:30, 5.15it/s] 60%|█████▉ | 28681/48008 [1:38:19<1:01:52, 5.21it/s] 60%|█████▉ | 28682/48008 [1:38:19<1:02:22, 5.16it/s] 60%|█████▉ | 28683/48008 [1:38:19<1:02:13, 5.18it/s] 60%|█████▉ | 28684/48008 [1:38:19<1:02:13, 5.18it/s] 60%|█████▉ | 28685/48008 [1:38:20<1:13:12, 4.40it/s] 60%|█████▉ | 28686/48008 [1:38:20<1:20:41, 3.99it/s] 60%|█████▉ | 28687/48008 [1:38:20<1:26:18, 3.73it/s] 60%|█████▉ | 28688/48008 [1:38:20<1:18:59, 4.08it/s] 60%|█████▉ | 28689/48008 [1:38:21<1:13:45, 4.37it/s] 60%|█████▉ | 28690/48008 [1:38:21<1:11:30, 4.50it/s] 60%|█████▉ | 28691/48008 [1:38:21<1:09:09, 4.66it/s] 60%|█████▉ | 28692/48008 [1:38:21<1:06:54, 4.81it/s] 60%|█████▉ | 28693/48008 [1:38:21<1:05:40, 4.90it/s] 60%|█████▉ | 28694/48008 [1:38:22<1:05:31, 4.91it/s] 60%|█████▉ | 28695/48008 [1:38:22<1:04:23, 5.00it/s] 60%|█████▉ | 28696/48008 [1:38:22<1:03:47, 5.05it/s] 60%|█████▉ | 28697/48008 [1:38:22<1:07:34, 4.76it/s] 60%|█████▉ | 28698/48008 [1:38:22<1:07:20, 4.78it/s] 60%|█████▉ | 28699/48008 [1:38:23<1:05:35, 4.91it/s] 60%|█████▉ | 28700/48008 [1:38:23<1:05:39, 4.90it/s] {'loss': 0.299, 'grad_norm': 0.310546875, 'learning_rate': 8.044075987335445e-05, 'epoch': 0.6} 60%|█████▉ | 28700/48008 [1:38:23<1:05:39, 4.90it/s] 60%|█████▉ | 28701/48008 [1:38:23<1:05:12, 4.93it/s] 60%|█████▉ | 28702/48008 [1:38:23<1:08:37, 4.69it/s] 60%|█████▉ | 28703/48008 [1:38:23<1:07:07, 4.79it/s] 60%|█████▉ | 28704/48008 [1:38:24<1:06:46, 4.82it/s] 60%|█████▉ | 28705/48008 [1:38:24<1:07:01, 4.80it/s] 60%|█████▉ | 28706/48008 [1:38:24<1:05:10, 4.94it/s] 60%|█████▉ | 28707/48008 [1:38:24<1:04:18, 5.00it/s] 60%|█████▉ | 28708/48008 [1:38:24<1:03:04, 5.10it/s] 60%|█████▉ | 28709/48008 [1:38:25<1:02:34, 5.14it/s] 60%|█████▉ | 28710/48008 [1:38:25<1:03:34, 5.06it/s] 60%|█████▉ | 28711/48008 [1:38:25<1:04:05, 5.02it/s] 60%|█████▉ | 28712/48008 [1:38:25<1:03:34, 5.06it/s] 60%|█████▉ | 28713/48008 [1:38:25<1:02:45, 5.12it/s] 60%|█████▉ | 28714/48008 [1:38:26<1:01:56, 5.19it/s] 60%|█████▉ | 28715/48008 [1:38:26<1:02:30, 5.14it/s] 60%|█████▉ | 28716/48008 [1:38:26<1:01:44, 5.21it/s] 60%|█████▉ | 28717/48008 [1:38:26<1:01:18, 5.24it/s] 60%|█████▉ | 28718/48008 [1:38:26<1:01:19, 5.24it/s] 60%|█████▉ | 28719/48008 [1:38:26<1:01:53, 5.19it/s] 60%|█████▉ | 28720/48008 [1:38:27<2:07:43, 2.52it/s] 60%|█████▉ | 28721/48008 [1:38:28<1:47:52, 2.98it/s] 60%|█████▉ | 28722/48008 [1:38:28<1:34:00, 3.42it/s] 60%|█████▉ | 28723/48008 [1:38:28<1:24:21, 3.81it/s] 60%|█████▉ | 28724/48008 [1:38:28<1:17:51, 4.13it/s] 60%|█████▉ | 28725/48008 [1:38:28<1:12:58, 4.40it/s] 60%|█████▉ | 28726/48008 [1:38:29<1:28:21, 3.64it/s] 60%|█████▉ | 28727/48008 [1:38:29<1:20:56, 3.97it/s] 60%|█████▉ | 28728/48008 [1:38:29<1:16:11, 4.22it/s] 60%|█████▉ | 28729/48008 [1:38:29<1:11:53, 4.47it/s] 60%|█████▉ | 28730/48008 [1:38:29<1:09:12, 4.64it/s] 60%|█████▉ | 28731/48008 [1:38:30<1:06:15, 4.85it/s] 60%|█████▉ | 28732/48008 [1:38:30<1:04:27, 4.98it/s] 60%|█████▉ | 28733/48008 [1:38:30<1:03:28, 5.06it/s] 60%|█████▉ | 28734/48008 [1:38:30<1:02:44, 5.12it/s] 60%|█████▉ | 28735/48008 [1:38:30<1:06:41, 4.82it/s] 60%|█████▉ | 28736/48008 [1:38:31<1:09:34, 4.62it/s] 60%|█████▉ | 28737/48008 [1:38:31<1:07:11, 4.78it/s] 60%|█████▉ | 28738/48008 [1:38:31<1:06:03, 4.86it/s] 60%|█████▉ | 28739/48008 [1:38:31<1:04:33, 4.97it/s] 60%|█████▉ | 28740/48008 [1:38:31<1:03:44, 5.04it/s] 60%|█████▉ | 28741/48008 [1:38:32<1:03:36, 5.05it/s] 60%|█████▉ | 28742/48008 [1:38:33<2:23:48, 2.23it/s] 60%|█████▉ | 28743/48008 [1:38:33<1:59:13, 2.69it/s] 60%|█████▉ | 28744/48008 [1:38:33<1:41:34, 3.16it/s] 60%|█████▉ | 28745/48008 [1:38:33<1:30:01, 3.57it/s] 60%|█████▉ | 28746/48008 [1:38:33<1:21:42, 3.93it/s] 60%|█████▉ | 28747/48008 [1:38:34<1:17:13, 4.16it/s] 60%|█████▉ | 28748/48008 [1:38:35<2:13:43, 2.40it/s] 60%|█████▉ | 28749/48008 [1:38:35<1:51:29, 2.88it/s] 60%|█████▉ | 28750/48008 [1:38:35<1:37:08, 3.30it/s] {'loss': 0.3018, 'grad_norm': 0.458984375, 'learning_rate': 8.023246125645726e-05, 'epoch': 0.6} 60%|█████▉ | 28750/48008 [1:38:35<1:37:08, 3.30it/s] 60%|█████▉ | 28751/48008 [1:38:35<1:26:43, 3.70it/s] 60%|█████▉ | 28752/48008 [1:38:35<1:18:53, 4.07it/s] 60%|█████▉ | 28753/48008 [1:38:35<1:13:19, 4.38it/s] 60%|█████▉ | 28754/48008 [1:38:36<1:10:09, 4.57it/s] 60%|█████▉ | 28755/48008 [1:38:36<1:08:31, 4.68it/s] 60%|█████▉ | 28756/48008 [1:38:36<1:06:22, 4.83it/s] 60%|█████▉ | 28757/48008 [1:38:36<1:05:20, 4.91it/s] 60%|█████▉ | 28758/48008 [1:38:36<1:04:23, 4.98it/s] 60%|█████▉ | 28759/48008 [1:38:37<1:02:47, 5.11it/s] 60%|█████▉ | 28760/48008 [1:38:37<1:02:24, 5.14it/s] 60%|█████▉ | 28761/48008 [1:38:37<1:02:04, 5.17it/s] 60%|█████▉ | 28762/48008 [1:38:37<1:00:52, 5.27it/s] 60%|█████▉ | 28763/48008 [1:38:38<1:19:45, 4.02it/s] 60%|█████▉ | 28764/48008 [1:38:38<1:14:20, 4.31it/s] 60%|█████▉ | 28765/48008 [1:38:38<1:10:29, 4.55it/s] 60%|█████▉ | 28766/48008 [1:38:38<1:08:08, 4.71it/s] 60%|█████▉ | 28767/48008 [1:38:38<1:05:39, 4.88it/s] 60%|█████▉ | 28768/48008 [1:38:39<1:04:20, 4.98it/s] 60%|█████▉ | 28769/48008 [1:38:39<1:03:44, 5.03it/s] 60%|█████▉ | 28770/48008 [1:38:39<1:02:25, 5.14it/s] 60%|█████▉ | 28771/48008 [1:38:39<1:02:13, 5.15it/s] 60%|█████▉ | 28772/48008 [1:38:39<1:02:11, 5.15it/s] 60%|█████▉ | 28773/48008 [1:38:40<1:02:01, 5.17it/s] 60%|█████▉ | 28774/48008 [1:38:40<1:01:57, 5.17it/s] 60%|█████▉ | 28775/48008 [1:38:40<1:02:20, 5.14it/s] 60%|█████▉ | 28776/48008 [1:38:40<1:03:36, 5.04it/s] 60%|█████▉ | 28777/48008 [1:38:40<1:07:26, 4.75it/s] 60%|█████▉ | 28778/48008 [1:38:41<1:06:45, 4.80it/s] 60%|█████▉ | 28779/48008 [1:38:41<1:23:57, 3.82it/s] 60%|█████▉ | 28780/48008 [1:38:41<1:16:58, 4.16it/s] 60%|█████▉ | 28781/48008 [1:38:41<1:13:04, 4.38it/s] 60%|█████▉ | 28782/48008 [1:38:42<1:09:29, 4.61it/s] 60%|█████▉ | 28783/48008 [1:38:42<1:07:07, 4.77it/s] 60%|█████▉ | 28784/48008 [1:38:43<2:06:12, 2.54it/s] 60%|█████▉ | 28785/48008 [1:38:43<1:47:46, 2.97it/s] 60%|█████▉ | 28786/48008 [1:38:43<1:33:43, 3.42it/s] 60%|█████▉ | 28787/48008 [1:38:43<1:35:02, 3.37it/s] 60%|█████▉ | 28788/48008 [1:38:43<1:25:08, 3.76it/s] 60%|█████▉ | 28789/48008 [1:38:44<1:17:28, 4.13it/s] 60%|█████▉ | 28790/48008 [1:38:44<1:13:47, 4.34it/s] 60%|█████▉ | 28791/48008 [1:38:44<1:11:20, 4.49it/s] 60%|█████▉ | 28792/48008 [1:38:44<1:08:25, 4.68it/s] 60%|█████▉ | 28793/48008 [1:38:44<1:06:20, 4.83it/s] 60%|█████▉ | 28794/48008 [1:38:45<1:06:18, 4.83it/s] 60%|█████▉ | 28795/48008 [1:38:45<1:04:46, 4.94it/s] 60%|█████▉ | 28796/48008 [1:38:45<1:03:38, 5.03it/s] 60%|█████▉ | 28797/48008 [1:38:45<1:02:34, 5.12it/s] 60%|█████▉ | 28798/48008 [1:38:45<1:01:53, 5.17it/s] 60%|█████▉ | 28799/48008 [1:38:46<1:02:29, 5.12it/s] 60%|█████▉ | 28800/48008 [1:38:46<1:02:10, 5.15it/s] {'loss': 0.2999, 'grad_norm': 0.1845703125, 'learning_rate': 8.002416263956007e-05, 'epoch': 0.6} 60%|█████▉ | 28800/48008 [1:38:46<1:02:10, 5.15it/s] 60%|█████▉ | 28801/48008 [1:38:46<1:02:13, 5.14it/s] 60%|█████▉ | 28802/48008 [1:38:46<1:06:35, 4.81it/s] 60%|█████▉ | 28803/48008 [1:38:46<1:04:56, 4.93it/s] 60%|█████▉ | 28804/48008 [1:38:47<1:03:45, 5.02it/s] 60%|██████ | 28805/48008 [1:38:47<1:02:54, 5.09it/s] 60%|██████ | 28806/48008 [1:38:47<1:02:45, 5.10it/s] 60%|██████ | 28807/48008 [1:38:47<1:02:02, 5.16it/s] 60%|██████ | 28808/48008 [1:38:47<1:01:19, 5.22it/s] 60%|██████ | 28809/48008 [1:38:48<1:05:45, 4.87it/s] 60%|██████ | 28810/48008 [1:38:48<1:04:28, 4.96it/s] 60%|██████ | 28811/48008 [1:38:48<1:03:09, 5.07it/s] 60%|██████ | 28812/48008 [1:38:48<1:02:21, 5.13it/s] 60%|██████ | 28813/48008 [1:38:48<1:03:13, 5.06it/s] 60%|██████ | 28814/48008 [1:38:49<1:01:57, 5.16it/s] 60%|██████ | 28815/48008 [1:38:49<1:03:14, 5.06it/s] 60%|██████ | 28816/48008 [1:38:49<1:14:03, 4.32it/s] 60%|██████ | 28817/48008 [1:38:49<1:09:37, 4.59it/s] 60%|██████ | 28818/48008 [1:38:49<1:07:06, 4.77it/s] 60%|██████ | 28819/48008 [1:38:50<1:04:49, 4.93it/s] 60%|██████ | 28820/48008 [1:38:50<1:05:18, 4.90it/s] 60%|██████ | 28821/48008 [1:38:50<1:04:12, 4.98it/s] 60%|██████ | 28822/48008 [1:38:50<1:03:19, 5.05it/s] 60%|██████ | 28823/48008 [1:38:51<1:21:29, 3.92it/s] 60%|██████ | 28824/48008 [1:38:51<1:15:04, 4.26it/s] 60%|██████ | 28825/48008 [1:38:51<1:11:13, 4.49it/s] 60%|██████ | 28826/48008 [1:38:51<1:08:23, 4.68it/s] 60%|██████ | 28827/48008 [1:38:51<1:05:56, 4.85it/s] 60%|██████ | 28828/48008 [1:38:52<1:06:25, 4.81it/s] 60%|██████ | 28829/48008 [1:38:52<1:06:03, 4.84it/s] 60%|██████ | 28830/48008 [1:38:52<1:04:29, 4.96it/s] 60%|██████ | 28831/48008 [1:38:52<1:03:50, 5.01it/s] 60%|██████ | 28832/48008 [1:38:52<1:03:14, 5.05it/s] 60%|██████ | 28833/48008 [1:38:53<1:02:38, 5.10it/s] 60%|██████ | 28834/48008 [1:38:53<1:01:55, 5.16it/s] 60%|██████ | 28835/48008 [1:38:53<1:01:51, 5.17it/s] 60%|██████ | 28836/48008 [1:38:53<1:02:44, 5.09it/s] 60%|██████ | 28837/48008 [1:38:53<1:01:42, 5.18it/s] 60%|██████ | 28838/48008 [1:38:54<1:01:39, 5.18it/s] 60%|██████ | 28839/48008 [1:38:54<1:01:37, 5.18it/s] 60%|██████ | 28840/48008 [1:38:54<1:01:19, 5.21it/s] 60%|██████ | 28841/48008 [1:38:54<1:00:45, 5.26it/s] 60%|██████ | 28842/48008 [1:38:54<1:05:27, 4.88it/s] 60%|██████ | 28843/48008 [1:38:55<1:04:06, 4.98it/s] 60%|██████ | 28844/48008 [1:38:55<1:02:16, 5.13it/s] 60%|██████ | 28845/48008 [1:38:55<1:01:17, 5.21it/s] 60%|██████ | 28846/48008 [1:38:55<1:00:53, 5.24it/s] 60%|██████ | 28847/48008 [1:38:55<1:00:48, 5.25it/s] 60%|██████ | 28848/48008 [1:38:55<1:00:56, 5.24it/s] 60%|██████ | 28849/48008 [1:38:56<1:00:55, 5.24it/s] 60%|██████ | 28850/48008 [1:38:56<1:02:02, 5.15it/s] {'loss': 0.3024, 'grad_norm': 0.37890625, 'learning_rate': 7.981586402266289e-05, 'epoch': 0.6} 60%|██████ | 28850/48008 [1:38:56<1:02:02, 5.15it/s] 60%|██████ | 28851/48008 [1:38:56<1:12:56, 4.38it/s] 60%|██████ | 28852/48008 [1:38:56<1:11:18, 4.48it/s] 60%|██████ | 28853/48008 [1:38:57<1:07:46, 4.71it/s] 60%|██████ | 28854/48008 [1:38:57<1:05:39, 4.86it/s] 60%|██████ | 28855/48008 [1:38:57<1:04:17, 4.97it/s] 60%|██████ | 28856/48008 [1:38:57<1:03:24, 5.03it/s] 60%|██████ | 28857/48008 [1:38:57<1:04:30, 4.95it/s] 60%|██████ | 28858/48008 [1:38:58<1:05:05, 4.90it/s] 60%|██████ | 28859/48008 [1:38:58<1:04:26, 4.95it/s] 60%|██████ | 28860/48008 [1:38:58<1:03:35, 5.02it/s] 60%|██████ | 28861/48008 [1:38:58<1:02:33, 5.10it/s] 60%|██████ | 28862/48008 [1:38:58<1:02:00, 5.15it/s] 60%|██████ | 28863/48008 [1:38:58<1:01:52, 5.16it/s] 60%|██████ | 28864/48008 [1:38:59<1:01:51, 5.16it/s] 60%|██████ | 28865/48008 [1:38:59<1:01:36, 5.18it/s] 60%|██████ | 28866/48008 [1:38:59<1:00:51, 5.24it/s] 60%|██████ | 28867/48008 [1:38:59<1:01:26, 5.19it/s] 60%|██████ | 28868/48008 [1:38:59<1:00:45, 5.25it/s] 60%|██████ | 28869/48008 [1:39:00<1:00:18, 5.29it/s] 60%|██████ | 28870/48008 [1:39:00<1:01:09, 5.22it/s] 60%|██████ | 28871/48008 [1:39:00<1:05:26, 4.87it/s] 60%|██████ | 28872/48008 [1:39:00<1:04:41, 4.93it/s] 60%|██████ | 28873/48008 [1:39:00<1:04:14, 4.96it/s] 60%|██████ | 28874/48008 [1:39:01<1:03:27, 5.02it/s] 60%|██████ | 28875/48008 [1:39:01<1:02:41, 5.09it/s] 60%|██████ | 28876/48008 [1:39:01<1:01:47, 5.16it/s] 60%|██████ | 28877/48008 [1:39:01<1:01:40, 5.17it/s] 60%|██████ | 28878/48008 [1:39:01<1:02:09, 5.13it/s] 60%|██████ | 28879/48008 [1:39:02<1:02:46, 5.08it/s] 60%|██████ | 28880/48008 [1:39:02<1:03:56, 4.99it/s] 60%|██████ | 28881/48008 [1:39:02<1:03:10, 5.05it/s] 60%|██████ | 28882/48008 [1:39:02<1:02:19, 5.11it/s] 60%|██████ | 28883/48008 [1:39:02<1:01:47, 5.16it/s] 60%|██████ | 28884/48008 [1:39:03<1:02:58, 5.06it/s] 60%|██████ | 28885/48008 [1:39:03<1:01:57, 5.14it/s] 60%|██████ | 28886/48008 [1:39:03<1:02:31, 5.10it/s] 60%|██████ | 28887/48008 [1:39:03<1:02:07, 5.13it/s] 60%|██████ | 28888/48008 [1:39:03<1:01:38, 5.17it/s] 60%|██████ | 28889/48008 [1:39:04<1:02:41, 5.08it/s] 60%|██████ | 28890/48008 [1:39:04<1:02:54, 5.06it/s] 60%|██████ | 28891/48008 [1:39:04<1:20:53, 3.94it/s] 60%|██████ | 28892/48008 [1:39:04<1:16:08, 4.18it/s] 60%|██████ | 28893/48008 [1:39:05<1:11:17, 4.47it/s] 60%|██████ | 28894/48008 [1:39:05<1:08:22, 4.66it/s] 60%|██████ | 28895/48008 [1:39:05<1:06:38, 4.78it/s] 60%|██████ | 28896/48008 [1:39:05<1:04:29, 4.94it/s] 60%|██████ | 28897/48008 [1:39:05<1:04:33, 4.93it/s] 60%|██████ | 28898/48008 [1:39:06<1:03:19, 5.03it/s] 60%|██████ | 28899/48008 [1:39:06<1:02:11, 5.12it/s] 60%|██████ | 28900/48008 [1:39:06<1:01:31, 5.18it/s] 60%|██████ | 28900/48008 [1:39:06<1:01:31, 5.18it/s]{'loss': 0.3017, 'grad_norm': 0.1884765625, 'learning_rate': 7.96075654057657e-05, 'epoch': 0.6} 60%|██████ | 28901/48008 [1:39:06<1:01:29, 5.18it/s] 60%|██████ | 28902/48008 [1:39:06<1:02:01, 5.13it/s] 60%|██████ | 28903/48008 [1:39:06<1:01:03, 5.22it/s] 60%|██████ | 28904/48008 [1:39:07<1:01:06, 5.21it/s] 60%|██████ | 28905/48008 [1:39:07<1:01:06, 5.21it/s] 60%|██████ | 28906/48008 [1:39:07<1:02:06, 5.13it/s] 60%|██████ | 28907/48008 [1:39:07<1:03:25, 5.02it/s] 60%|██████ | 28908/48008 [1:39:07<1:02:49, 5.07it/s] 60%|██████ | 28909/48008 [1:39:08<1:01:56, 5.14it/s] 60%|██████ | 28910/48008 [1:39:08<1:01:16, 5.19it/s] 60%|██████ | 28911/48008 [1:39:08<1:05:39, 4.85it/s] 60%|██████ | 28912/48008 [1:39:08<1:04:14, 4.95it/s] 60%|██████ | 28913/48008 [1:39:08<1:03:15, 5.03it/s] 60%|██████ | 28914/48008 [1:39:09<1:03:12, 5.03it/s] 60%|██████ | 28915/48008 [1:39:09<1:03:46, 4.99it/s] 60%|██████ | 28916/48008 [1:39:09<1:02:58, 5.05it/s] 60%|██████ | 28917/48008 [1:39:09<1:02:23, 5.10it/s] 60%|██████ | 28918/48008 [1:39:09<1:03:53, 4.98it/s] 60%|██████ | 28919/48008 [1:39:10<1:04:19, 4.95it/s] 60%|██████ | 28920/48008 [1:39:10<1:03:54, 4.98it/s] 60%|██████ | 28921/48008 [1:39:10<1:02:31, 5.09it/s] 60%|██████ | 28922/48008 [1:39:10<1:02:04, 5.12it/s] 60%|██████ | 28923/48008 [1:39:10<1:01:42, 5.15it/s] 60%|██████ | 28924/48008 [1:39:11<1:01:15, 5.19it/s] 60%|██████ | 28925/48008 [1:39:11<1:01:20, 5.19it/s] 60%|██████ | 28926/48008 [1:39:11<1:01:14, 5.19it/s] 60%|██████ | 28927/48008 [1:39:11<1:01:05, 5.21it/s] 60%|██████ | 28928/48008 [1:39:11<1:01:09, 5.20it/s] 60%|██████ | 28929/48008 [1:39:12<1:00:55, 5.22it/s] 60%|██████ | 28930/48008 [1:39:12<1:01:15, 5.19it/s] 60%|██████ | 28931/48008 [1:39:12<1:01:00, 5.21it/s] 60%|██████ | 28932/48008 [1:39:12<1:01:01, 5.21it/s] 60%|██████ | 28933/48008 [1:39:12<1:01:26, 5.17it/s] 60%|██████ | 28934/48008 [1:39:13<1:12:14, 4.40it/s] 60%|██████ | 28935/48008 [1:39:13<1:08:41, 4.63it/s] 60%|██████ | 28936/48008 [1:39:13<1:07:01, 4.74it/s] 60%|██████ | 28937/48008 [1:39:13<1:04:47, 4.91it/s] 60%|██████ | 28938/48008 [1:39:13<1:04:48, 4.90it/s] 60%|██████ | 28939/48008 [1:39:14<1:03:16, 5.02it/s] 60%|██████ | 28940/48008 [1:39:14<1:02:26, 5.09it/s] 60%|██████ | 28941/48008 [1:39:14<1:01:27, 5.17it/s] 60%|██████ | 28942/48008 [1:39:14<1:01:29, 5.17it/s] 60%|██████ | 28943/48008 [1:39:14<1:01:16, 5.19it/s] 60%|██████ | 28944/48008 [1:39:15<1:01:07, 5.20it/s] 60%|██████ | 28945/48008 [1:39:15<1:00:56, 5.21it/s] 60%|██████ | 28946/48008 [1:39:15<1:00:42, 5.23it/s] 60%|██████ | 28947/48008 [1:39:15<1:00:42, 5.23it/s] 60%|██████ | 28948/48008 [1:39:15<1:00:29, 5.25it/s] 60%|██████ | 28949/48008 [1:39:16<1:00:16, 5.27it/s] 60%|██████ | 28950/48008 [1:39:16<1:00:32, 5.25it/s] 60%|██████ | 28950/48008 [1:39:16<1:00:32, 5.25it/s]{'loss': 0.3027, 'grad_norm': 0.138671875, 'learning_rate': 7.939926678886852e-05, 'epoch': 0.6} 60%|██████ | 28951/48008 [1:39:16<1:00:30, 5.25it/s] 60%|██████ | 28952/48008 [1:39:16<1:01:08, 5.19it/s] 60%|██████ | 28953/48008 [1:39:16<1:00:50, 5.22it/s] 60%|██████ | 28954/48008 [1:39:17<1:00:57, 5.21it/s] 60%|██████ | 28955/48008 [1:39:17<1:01:26, 5.17it/s] 60%|██████ | 28956/48008 [1:39:17<1:00:57, 5.21it/s] 60%|██████ | 28957/48008 [1:39:17<1:02:01, 5.12it/s] 60%|██████ | 28958/48008 [1:39:17<1:03:00, 5.04it/s] 60%|██████ | 28959/48008 [1:39:17<1:02:16, 5.10it/s] 60%|██████ | 28960/48008 [1:39:18<1:01:38, 5.15it/s] 60%|██████ | 28961/48008 [1:39:18<1:12:12, 4.40it/s] 60%|██████ | 28962/48008 [1:39:18<1:08:58, 4.60it/s] 60%|██████ | 28963/48008 [1:39:18<1:11:06, 4.46it/s] 60%|██████ | 28964/48008 [1:39:19<1:07:52, 4.68it/s] 60%|██████ | 28965/48008 [1:39:19<1:05:39, 4.83it/s] 60%|██████ | 28966/48008 [1:39:19<1:04:15, 4.94it/s] 60%|██████ | 28967/48008 [1:39:19<1:03:43, 4.98it/s] 60%|██████ | 28968/48008 [1:39:19<1:02:25, 5.08it/s] 60%|██████ | 28969/48008 [1:39:20<1:02:01, 5.12it/s] 60%|██████ | 28970/48008 [1:39:20<1:01:58, 5.12it/s] 60%|██████ | 28971/48008 [1:39:20<1:12:29, 4.38it/s] 60%|██████ | 28972/48008 [1:39:20<1:08:49, 4.61it/s] 60%|██████ | 28973/48008 [1:39:20<1:06:28, 4.77it/s] 60%|██████ | 28974/48008 [1:39:21<1:04:55, 4.89it/s] 60%|██████ | 28975/48008 [1:39:21<1:03:39, 4.98it/s] 60%|██████ | 28976/48008 [1:39:21<1:05:06, 4.87it/s] 60%|██████ | 28977/48008 [1:39:21<1:03:51, 4.97it/s] 60%|██████ | 28978/48008 [1:39:21<1:07:16, 4.71it/s] 60%|██████ | 28979/48008 [1:39:22<1:06:34, 4.76it/s] 60%|██████ | 28980/48008 [1:39:22<1:04:45, 4.90it/s] 60%|██████ | 28981/48008 [1:39:22<1:03:25, 5.00it/s] 60%|██████ | 28982/48008 [1:39:22<1:03:16, 5.01it/s] 60%|██████ | 28983/48008 [1:39:22<1:02:46, 5.05it/s] 60%|██████ | 28984/48008 [1:39:23<1:02:32, 5.07it/s] 60%|██████ | 28985/48008 [1:39:23<1:02:07, 5.10it/s] 60%|██████ | 28986/48008 [1:39:23<1:01:33, 5.15it/s] 60%|██████ | 28987/48008 [1:39:23<1:13:11, 4.33it/s] 60%|██████ | 28988/48008 [1:39:24<1:13:46, 4.30it/s] 60%|██████ | 28989/48008 [1:39:24<1:28:27, 3.58it/s] 60%|██████ | 28990/48008 [1:39:24<1:19:55, 3.97it/s] 60%|██████ | 28991/48008 [1:39:24<1:13:53, 4.29it/s] 60%|██████ | 28992/48008 [1:39:25<1:10:33, 4.49it/s] 60%|██████ | 28993/48008 [1:39:25<1:18:24, 4.04it/s] 60%|██████ | 28994/48008 [1:39:25<1:13:08, 4.33it/s] 60%|██████ | 28995/48008 [1:39:25<1:09:17, 4.57it/s] 60%|██████ | 28996/48008 [1:39:25<1:10:59, 4.46it/s] 60%|██████ | 28997/48008 [1:39:26<1:07:44, 4.68it/s] 60%|██████ | 28998/48008 [1:39:26<1:05:46, 4.82it/s] 60%|██████ | 28999/48008 [1:39:26<1:04:21, 4.92it/s] 60%|██████ | 29000/48008 [1:39:26<1:03:48, 4.96it/s] {'loss': 0.3, 'grad_norm': 0.384765625, 'learning_rate': 7.919096817197135e-05, 'epoch': 0.6} 60%|██████ | 29000/48008 [1:39:26<1:03:48, 4.96it/s] 60%|██████ | 29001/48008 [1:39:26<1:02:52, 5.04it/s] 60%|██████ | 29002/48008 [1:39:27<1:02:44, 5.05it/s] 60%|██████ | 29003/48008 [1:39:27<1:02:23, 5.08it/s] 60%|██████ | 29004/48008 [1:39:27<1:01:24, 5.16it/s] 60%|██████ | 29005/48008 [1:39:27<1:02:15, 5.09it/s] 60%|██████ | 29006/48008 [1:39:28<1:12:40, 4.36it/s] 60%|██████ | 29007/48008 [1:39:28<1:09:02, 4.59it/s] 60%|██████ | 29008/48008 [1:39:28<1:07:41, 4.68it/s] 60%|██████ | 29009/48008 [1:39:28<1:10:03, 4.52it/s] 60%|██████ | 29010/48008 [1:39:28<1:07:13, 4.71it/s] 60%|██████ | 29011/48008 [1:39:29<1:05:04, 4.87it/s] 60%|██████ | 29012/48008 [1:39:29<1:03:34, 4.98it/s] 60%|██████ | 29013/48008 [1:39:29<1:02:36, 5.06it/s] 60%|██████ | 29014/48008 [1:39:29<1:06:31, 4.76it/s] 60%|██████ | 29015/48008 [1:39:29<1:04:44, 4.89it/s] 60%|██████ | 29016/48008 [1:39:30<1:03:14, 5.00it/s] 60%|██████ | 29017/48008 [1:39:30<1:02:57, 5.03it/s] 60%|██████ | 29018/48008 [1:39:30<1:02:19, 5.08it/s] 60%|██████ | 29019/48008 [1:39:30<1:01:23, 5.15it/s] 60%|██████ | 29020/48008 [1:39:30<1:01:04, 5.18it/s] 60%|██████ | 29021/48008 [1:39:31<1:00:50, 5.20it/s] 60%|██████ | 29022/48008 [1:39:31<1:00:28, 5.23it/s] 60%|██████ | 29023/48008 [1:39:31<1:00:26, 5.24it/s] 60%|██████ | 29024/48008 [1:39:31<1:01:35, 5.14it/s] 60%|██████ | 29025/48008 [1:39:31<1:01:47, 5.12it/s] 60%|██████ | 29026/48008 [1:39:32<1:05:50, 4.80it/s] 60%|██████ | 29027/48008 [1:39:32<1:05:01, 4.87it/s] 60%|██████ | 29028/48008 [1:39:32<1:04:42, 4.89it/s] 60%|██████ | 29029/48008 [1:39:32<1:03:22, 4.99it/s] 60%|██████ | 29030/48008 [1:39:32<1:02:29, 5.06it/s] 60%|██████ | 29031/48008 [1:39:32<1:02:01, 5.10it/s] 60%|██████ | 29032/48008 [1:39:33<1:01:22, 5.15it/s] 60%|██████ | 29033/48008 [1:39:33<1:01:01, 5.18it/s] 60%|██████ | 29034/48008 [1:39:33<1:01:00, 5.18it/s] 60%|██████ | 29035/48008 [1:39:33<1:05:10, 4.85it/s] 60%|██████ | 29036/48008 [1:39:33<1:03:26, 4.98it/s] 60%|██████ | 29037/48008 [1:39:34<1:02:43, 5.04it/s] 60%|██████ | 29038/48008 [1:39:34<1:01:31, 5.14it/s] 60%|██████ | 29039/48008 [1:39:34<1:02:21, 5.07it/s] 60%|██████ | 29040/48008 [1:39:34<1:02:46, 5.04it/s] 60%|██████ | 29041/48008 [1:39:35<1:06:27, 4.76it/s] 60%|██████ | 29042/48008 [1:39:35<1:04:39, 4.89it/s] 60%|██████ | 29043/48008 [1:39:35<1:03:14, 5.00it/s] 60%|██████ | 29044/48008 [1:39:35<1:02:09, 5.08it/s] 61%|██████ | 29045/48008 [1:39:35<1:03:25, 4.98it/s] 61%|██████ | 29046/48008 [1:39:35<1:02:29, 5.06it/s] 61%|██████ | 29047/48008 [1:39:36<1:01:26, 5.14it/s] 61%|██████ | 29048/48008 [1:39:36<1:00:45, 5.20it/s] 61%|██████ | 29049/48008 [1:39:36<59:54, 5.27it/s] 61%|██████ | 29050/48008 [1:39:36<59:23, 5.32it/s] {'loss': 0.3023, 'grad_norm': 0.63671875, 'learning_rate': 7.898266955507417e-05, 'epoch': 0.61} 61%|██████ | 29050/48008 [1:39:36<59:23, 5.32it/s] 61%|██████ | 29051/48008 [1:39:36<59:39, 5.30it/s] 61%|██████ | 29052/48008 [1:39:37<1:00:55, 5.19it/s] 61%|██████ | 29053/48008 [1:39:37<1:00:54, 5.19it/s] 61%|██████ | 29054/48008 [1:39:37<1:01:22, 5.15it/s] 61%|██████ | 29055/48008 [1:39:37<1:01:00, 5.18it/s] 61%|██████ | 29056/48008 [1:39:37<1:02:35, 5.05it/s] 61%|██████ | 29057/48008 [1:39:38<1:01:58, 5.10it/s] 61%|██████ | 29058/48008 [1:39:38<1:05:52, 4.79it/s] 61%|██████ | 29059/48008 [1:39:38<1:03:44, 4.95it/s] 61%|██████ | 29060/48008 [1:39:38<1:02:28, 5.05it/s] 61%|██████ | 29061/48008 [1:39:38<1:01:28, 5.14it/s] 61%|██████ | 29062/48008 [1:39:39<1:01:10, 5.16it/s] 61%|██████ | 29063/48008 [1:39:39<1:00:48, 5.19it/s] 61%|██████ | 29064/48008 [1:39:39<1:01:14, 5.16it/s] 61%|██████ | 29065/48008 [1:39:39<1:00:31, 5.22it/s] 61%|██████ | 29066/48008 [1:39:39<1:00:29, 5.22it/s] 61%|██████ | 29067/48008 [1:39:40<1:00:20, 5.23it/s] 61%|██████ | 29068/48008 [1:39:40<1:00:27, 5.22it/s] 61%|██████ | 29069/48008 [1:39:40<1:00:34, 5.21it/s] 61%|██████ | 29070/48008 [1:39:40<59:54, 5.27it/s] 61%|██████ | 29071/48008 [1:39:40<59:53, 5.27it/s] 61%|██████ | 29072/48008 [1:39:40<58:57, 5.35it/s] 61%|██████ | 29073/48008 [1:39:41<59:13, 5.33it/s] 61%|██████ | 29074/48008 [1:39:41<1:00:14, 5.24it/s] 61%|██████ | 29075/48008 [1:39:41<59:49, 5.27it/s] 61%|██████ | 29076/48008 [1:39:41<59:58, 5.26it/s] 61%|██████ | 29077/48008 [1:39:41<1:00:04, 5.25it/s] 61%|██████ | 29078/48008 [1:39:42<1:00:17, 5.23it/s] 61%|██████ | 29079/48008 [1:39:42<1:00:57, 5.18it/s] 61%|██████ | 29080/48008 [1:39:42<1:05:11, 4.84it/s] 61%|██████ | 29081/48008 [1:39:42<1:03:15, 4.99it/s] 61%|██████ | 29082/48008 [1:39:42<1:02:20, 5.06it/s] 61%|██████ | 29083/48008 [1:39:43<1:01:35, 5.12it/s] 61%|██████ | 29084/48008 [1:39:43<1:01:04, 5.16it/s] 61%|██████ | 29085/48008 [1:39:43<1:00:47, 5.19it/s] 61%|██████ | 29086/48008 [1:39:43<1:01:02, 5.17it/s] 61%|██████ | 29087/48008 [1:39:43<1:01:47, 5.10it/s] 61%|██████ | 29088/48008 [1:39:44<1:02:18, 5.06it/s] 61%|██████ | 29089/48008 [1:39:44<1:01:50, 5.10it/s] 61%|██████ | 29090/48008 [1:39:44<1:01:01, 5.17it/s] 61%|██████ | 29091/48008 [1:39:44<1:02:40, 5.03it/s] 61%|██████ | 29092/48008 [1:39:44<1:01:47, 5.10it/s] 61%|██████ | 29093/48008 [1:39:45<1:02:00, 5.08it/s] 61%|██████ | 29094/48008 [1:39:45<1:01:28, 5.13it/s] 61%|██████ | 29095/48008 [1:39:45<1:02:03, 5.08it/s] 61%|██████ | 29096/48008 [1:39:45<1:01:33, 5.12it/s] 61%|██████ | 29097/48008 [1:39:45<1:02:54, 5.01it/s] 61%|██████ | 29098/48008 [1:39:46<1:01:59, 5.08it/s] 61%|██████ | 29099/48008 [1:39:46<1:02:37, 5.03it/s] 61%|██████ | 29100/48008 [1:39:46<1:02:33, 5.04it/s] 61%|██████ | 29100/48008 [1:39:46<1:02:33, 5.04it/s]{'loss': 0.3006, 'grad_norm': 0.419921875, 'learning_rate': 7.877437093817697e-05, 'epoch': 0.61} 61%|██████ | 29101/48008 [1:39:46<1:01:51, 5.09it/s] 61%|██████ | 29102/48008 [1:39:46<1:02:24, 5.05it/s] 61%|██████ | 29103/48008 [1:39:47<1:01:37, 5.11it/s] 61%|██████ | 29104/48008 [1:39:47<1:01:06, 5.16it/s] 61%|██████ | 29105/48008 [1:39:47<1:02:05, 5.07it/s] 61%|██████ | 29106/48008 [1:39:47<1:01:05, 5.16it/s] 61%|██████ | 29107/48008 [1:39:47<1:01:00, 5.16it/s] 61%|██████ | 29108/48008 [1:39:48<1:00:30, 5.21it/s] 61%|██████ | 29109/48008 [1:39:48<1:00:35, 5.20it/s] 61%|██████ | 29110/48008 [1:39:48<1:01:07, 5.15it/s] 61%|██████ | 29111/48008 [1:39:48<1:00:42, 5.19it/s] 61%|██████ | 29112/48008 [1:39:48<59:52, 5.26it/s] 61%|██████ | 29113/48008 [1:39:48<1:00:49, 5.18it/s] 61%|██████ | 29114/48008 [1:39:49<1:00:43, 5.19it/s] 61%|██████ | 29115/48008 [1:39:49<1:00:34, 5.20it/s] 61%|██████ | 29116/48008 [1:39:49<1:00:49, 5.18it/s] 61%|██████ | 29117/48008 [1:39:49<1:00:40, 5.19it/s] 61%|██████ | 29118/48008 [1:39:49<1:01:35, 5.11it/s] 61%|██████ | 29119/48008 [1:39:50<1:01:15, 5.14it/s] 61%|██████ | 29120/48008 [1:39:50<1:00:22, 5.21it/s] 61%|██████ | 29121/48008 [1:39:50<1:00:39, 5.19it/s] 61%|██████ | 29122/48008 [1:39:50<1:00:38, 5.19it/s] 61%|██████ | 29123/48008 [1:39:50<1:00:11, 5.23it/s] 61%|██████ | 29124/48008 [1:39:51<1:00:14, 5.22it/s] 61%|██████ | 29125/48008 [1:39:51<1:00:36, 5.19it/s] 61%|██████ | 29126/48008 [1:39:51<1:01:29, 5.12it/s] 61%|██████ | 29127/48008 [1:39:51<1:00:50, 5.17it/s] 61%|██████ | 29128/48008 [1:39:52<1:11:29, 4.40it/s] 61%|██████ | 29129/48008 [1:39:52<1:07:50, 4.64it/s] 61%|██████ | 29130/48008 [1:39:52<1:06:00, 4.77it/s] 61%|██████ | 29131/48008 [1:39:52<1:03:16, 4.97it/s] 61%|██████ | 29132/48008 [1:39:52<1:01:52, 5.08it/s] 61%|██████ | 29133/48008 [1:39:52<1:01:13, 5.14it/s] 61%|██████ | 29134/48008 [1:39:53<1:00:29, 5.20it/s] 61%|██████ | 29135/48008 [1:39:53<1:00:53, 5.17it/s] 61%|██████ | 29136/48008 [1:39:53<1:00:58, 5.16it/s] 61%|██████ | 29137/48008 [1:39:53<1:00:39, 5.19it/s] 61%|██████ | 29138/48008 [1:39:53<1:01:25, 5.12it/s] 61%|██████ | 29139/48008 [1:39:54<1:00:40, 5.18it/s] 61%|██████ | 29140/48008 [1:39:54<1:11:28, 4.40it/s] 61%|██████ | 29141/48008 [1:39:54<1:08:03, 4.62it/s] 61%|██████ | 29142/48008 [1:39:54<1:05:47, 4.78it/s] 61%|██████ | 29143/48008 [1:39:54<1:03:52, 4.92it/s] 61%|██████ | 29144/48008 [1:39:55<1:02:25, 5.04it/s] 61%|██████ | 29145/48008 [1:39:55<1:01:41, 5.10it/s] 61%|██████ | 29146/48008 [1:39:55<1:01:00, 5.15it/s] 61%|██████ | 29147/48008 [1:39:55<1:01:34, 5.11it/s] 61%|██████ | 29148/48008 [1:39:55<1:01:11, 5.14it/s] 61%|██████ | 29149/48008 [1:39:56<1:02:44, 5.01it/s] 61%|██████ | 29150/48008 [1:39:56<1:01:32, 5.11it/s] {'loss': 0.3004, 'grad_norm': 0.197265625, 'learning_rate': 7.856607232127979e-05, 'epoch': 0.61} 61%|██████ | 29150/48008 [1:39:56<1:01:32, 5.11it/s] 61%|██████ | 29151/48008 [1:39:56<1:01:01, 5.15it/s] 61%|██████ | 29152/48008 [1:39:56<1:00:10, 5.22it/s] 61%|██████ | 29153/48008 [1:39:56<1:04:30, 4.87it/s] 61%|██████ | 29154/48008 [1:39:57<1:03:13, 4.97it/s] 61%|██████ | 29155/48008 [1:39:57<1:01:57, 5.07it/s] 61%|██████ | 29156/48008 [1:39:57<1:01:05, 5.14it/s] 61%|██████ | 29157/48008 [1:39:57<1:01:14, 5.13it/s] 61%|██████ | 29158/48008 [1:39:57<1:00:53, 5.16it/s] 61%|██████ | 29159/48008 [1:39:58<1:00:26, 5.20it/s] 61%|██████ | 29160/48008 [1:39:58<59:52, 5.25it/s] 61%|██████ | 29161/48008 [1:39:58<59:42, 5.26it/s] 61%|██████ | 29162/48008 [1:39:58<59:19, 5.29it/s] 61%|██████ | 29163/48008 [1:39:58<59:34, 5.27it/s] 61%|██████ | 29164/48008 [1:39:59<1:01:00, 5.15it/s] 61%|██████ | 29165/48008 [1:39:59<1:00:06, 5.22it/s] 61%|██████ | 29166/48008 [1:39:59<1:01:07, 5.14it/s] 61%|██████ | 29167/48008 [1:39:59<1:00:46, 5.17it/s] 61%|██████ | 29168/48008 [1:39:59<1:00:32, 5.19it/s] 61%|██████ | 29169/48008 [1:40:00<1:00:31, 5.19it/s] 61%|██████ | 29170/48008 [1:40:00<1:00:08, 5.22it/s] 61%|██████ | 29171/48008 [1:40:00<1:01:12, 5.13it/s] 61%|██████ | 29172/48008 [1:40:00<1:00:23, 5.20it/s] 61%|██████ | 29173/48008 [1:40:00<1:00:52, 5.16it/s] 61%|██████ | 29174/48008 [1:40:00<1:00:25, 5.19it/s] 61%|██████ | 29175/48008 [1:40:01<59:41, 5.26it/s] 61%|██████ | 29176/48008 [1:40:01<58:43, 5.34it/s] 61%|██████ | 29177/48008 [1:40:01<1:00:18, 5.20it/s] 61%|██████ | 29178/48008 [1:40:01<1:00:39, 5.17it/s] 61%|██████ | 29179/48008 [1:40:01<1:00:54, 5.15it/s] 61%|██████ | 29180/48008 [1:40:02<1:00:28, 5.19it/s] 61%|██████ | 29181/48008 [1:40:02<1:04:38, 4.85it/s] 61%|██████ | 29182/48008 [1:40:02<1:02:55, 4.99it/s] 61%|██████ | 29183/48008 [1:40:02<1:01:54, 5.07it/s] 61%|██████ | 29184/48008 [1:40:02<1:01:09, 5.13it/s] 61%|██████ | 29185/48008 [1:40:03<1:00:53, 5.15it/s] 61%|██████ | 29186/48008 [1:40:03<1:00:18, 5.20it/s] 61%|██████ | 29187/48008 [1:40:03<1:11:14, 4.40it/s] 61%|██████ | 29188/48008 [1:40:03<1:08:53, 4.55it/s] 61%|██████ | 29189/48008 [1:40:04<1:05:54, 4.76it/s] 61%|██████ | 29190/48008 [1:40:04<1:04:13, 4.88it/s] 61%|██████ | 29191/48008 [1:40:04<1:03:31, 4.94it/s] 61%|██████ | 29192/48008 [1:40:04<1:06:45, 4.70it/s] 61%|██████ | 29193/48008 [1:40:04<1:04:30, 4.86it/s] 61%|██████ | 29194/48008 [1:40:05<1:03:36, 4.93it/s] 61%|██████ | 29195/48008 [1:40:05<1:02:52, 4.99it/s] 61%|██████ | 29196/48008 [1:40:05<1:01:15, 5.12it/s] 61%|██████ | 29197/48008 [1:40:05<1:00:30, 5.18it/s] 61%|██████ | 29198/48008 [1:40:05<59:53, 5.23it/s] 61%|██████ | 29199/48008 [1:40:05<1:00:58, 5.14it/s] 61%|██████ | 29200/48008 [1:40:06<1:00:16, 5.20it/s] {'loss': 0.3018, 'grad_norm': 0.330078125, 'learning_rate': 7.835777370438261e-05, 'epoch': 0.61} 61%|██████ | 29200/48008 [1:40:06<1:00:16, 5.20it/s] 61%|██████ | 29201/48008 [1:40:06<1:00:28, 5.18it/s] 61%|██████ | 29202/48008 [1:40:06<1:00:26, 5.19it/s] 61%|██████ | 29203/48008 [1:40:06<1:00:16, 5.20it/s] 61%|██████ | 29204/48008 [1:40:06<59:58, 5.23it/s] 61%|██████ | 29205/48008 [1:40:07<59:35, 5.26it/s] 61%|██████ | 29206/48008 [1:40:07<59:42, 5.25it/s] 61%|██████ | 29207/48008 [1:40:07<59:06, 5.30it/s] 61%|██████ | 29208/48008 [1:40:07<58:56, 5.32it/s] 61%|██████ | 29209/48008 [1:40:07<59:08, 5.30it/s] 61%|██████ | 29210/48008 [1:40:08<59:03, 5.31it/s] 61%|██████ | 29211/48008 [1:40:08<1:00:24, 5.19it/s] 61%|██████ | 29212/48008 [1:40:08<1:00:12, 5.20it/s] 61%|██████ | 29213/48008 [1:40:08<59:50, 5.23it/s] 61%|██████ | 29214/48008 [1:40:08<59:58, 5.22it/s] 61%|██████ | 29215/48008 [1:40:09<59:50, 5.23it/s] 61%|██████ | 29216/48008 [1:40:09<59:06, 5.30it/s] 61%|██████ | 29217/48008 [1:40:09<1:00:07, 5.21it/s] 61%|██████ | 29218/48008 [1:40:09<1:01:02, 5.13it/s] 61%|██████ | 29219/48008 [1:40:09<1:02:22, 5.02it/s] 61%|██████ | 29220/48008 [1:40:10<1:06:16, 4.72it/s] 61%|██████ | 29221/48008 [1:40:10<1:04:32, 4.85it/s] 61%|██████ | 29222/48008 [1:40:10<1:02:56, 4.97it/s] 61%|██████ | 29223/48008 [1:40:10<1:02:05, 5.04it/s] 61%|██████ | 29224/48008 [1:40:10<1:01:15, 5.11it/s] 61%|██████ | 29225/48008 [1:40:11<1:00:49, 5.15it/s] 61%|██████ | 29226/48008 [1:40:11<1:00:30, 5.17it/s] 61%|██████ | 29227/48008 [1:40:11<1:00:03, 5.21it/s] 61%|██████ | 29228/48008 [1:40:11<1:00:03, 5.21it/s] 61%|██████ | 29229/48008 [1:40:11<59:54, 5.22it/s] 61%|██████ | 29230/48008 [1:40:11<59:53, 5.23it/s] 61%|██████ | 29231/48008 [1:40:12<1:00:14, 5.20it/s] 61%|██████ | 29232/48008 [1:40:12<1:00:20, 5.19it/s] 61%|██████ | 29233/48008 [1:40:12<1:00:05, 5.21it/s] 61%|██████ | 29234/48008 [1:40:12<1:18:19, 4.00it/s] 61%|██████ | 29235/48008 [1:40:13<1:13:38, 4.25it/s] 61%|██████ | 29236/48008 [1:40:13<1:09:36, 4.49it/s] 61%|██████ | 29237/48008 [1:40:13<1:06:23, 4.71it/s] 61%|██████ | 29238/48008 [1:40:13<1:04:56, 4.82it/s] 61%|██████ | 29239/48008 [1:40:13<1:04:13, 4.87it/s] 61%|██████ | 29240/48008 [1:40:14<1:02:29, 5.01it/s] 61%|██████ | 29241/48008 [1:40:14<1:02:15, 5.02it/s] 61%|██████ | 29242/48008 [1:40:14<1:02:04, 5.04it/s] 61%|██████ | 29243/48008 [1:40:14<1:01:49, 5.06it/s] 61%|██████ | 29244/48008 [1:40:14<1:01:44, 5.07it/s] 61%|██████ | 29245/48008 [1:40:15<1:01:10, 5.11it/s] 61%|██████ | 29246/48008 [1:40:15<1:00:56, 5.13it/s] 61%|██████ | 29247/48008 [1:40:15<1:00:28, 5.17it/s] 61%|██████ | 29248/48008 [1:40:15<1:00:43, 5.15it/s] 61%|██████ | 29249/48008 [1:40:15<1:00:30, 5.17it/s] 61%|██████ | 29250/48008 [1:40:16<1:04:43, 4.83it/s] {'loss': 0.2976, 'grad_norm': 0.96484375, 'learning_rate': 7.814947508748542e-05, 'epoch': 0.61} 61%|██████ | 29250/48008 [1:40:16<1:04:43, 4.83it/s] 61%|██████ | 29251/48008 [1:40:16<1:07:41, 4.62it/s] 61%|██████ | 29252/48008 [1:40:16<1:05:09, 4.80it/s] 61%|██████ | 29253/48008 [1:40:16<1:03:39, 4.91it/s] 61%|██████ | 29254/48008 [1:40:16<1:02:29, 5.00it/s] 61%|██████ | 29255/48008 [1:40:17<1:01:52, 5.05it/s] 61%|██████ | 29256/48008 [1:40:17<1:00:40, 5.15it/s] 61%|██████ | 29257/48008 [1:40:17<1:00:21, 5.18it/s] 61%|██████ | 29258/48008 [1:40:17<1:00:18, 5.18it/s] 61%|██████ | 29259/48008 [1:40:17<59:58, 5.21it/s] 61%|██████ | 29260/48008 [1:40:18<59:29, 5.25it/s] 61%|██████ | 29261/48008 [1:40:18<1:00:41, 5.15it/s] 61%|██████ | 29262/48008 [1:40:18<1:00:25, 5.17it/s] 61%|██████ | 29263/48008 [1:40:18<59:38, 5.24it/s] 61%|██████ | 29264/48008 [1:40:18<1:01:19, 5.09it/s] 61%|██████ | 29265/48008 [1:40:19<1:00:08, 5.19it/s] 61%|██████ | 29266/48008 [1:40:19<59:56, 5.21it/s] 61%|██████ | 29267/48008 [1:40:19<1:00:01, 5.20it/s] 61%|██████ | 29268/48008 [1:40:19<1:04:16, 4.86it/s] 61%|██████ | 29269/48008 [1:40:19<1:02:01, 5.03it/s] 61%|██████ | 29270/48008 [1:40:20<1:02:20, 5.01it/s] 61%|██████ | 29271/48008 [1:40:20<1:00:49, 5.13it/s] 61%|██████ | 29272/48008 [1:40:20<1:00:18, 5.18it/s] 61%|██████ | 29273/48008 [1:40:20<1:01:19, 5.09it/s] 61%|██████ | 29274/48008 [1:40:20<1:00:29, 5.16it/s] 61%|██████ | 29275/48008 [1:40:20<59:27, 5.25it/s] 61%|██████ | 29276/48008 [1:40:21<59:28, 5.25it/s] 61%|██████ | 29277/48008 [1:40:21<1:00:31, 5.16it/s] 61%|██████ | 29278/48008 [1:40:21<1:00:16, 5.18it/s] 61%|██████ | 29279/48008 [1:40:21<59:50, 5.22it/s] 61%|██████ | 29280/48008 [1:40:21<1:01:27, 5.08it/s] 61%|██████ | 29281/48008 [1:40:22<1:01:47, 5.05it/s] 61%|██████ | 29282/48008 [1:40:22<1:02:45, 4.97it/s] 61%|██████ | 29283/48008 [1:40:22<1:01:24, 5.08it/s] 61%|██████ | 29284/48008 [1:40:22<1:00:43, 5.14it/s] 61%|██████ | 29285/48008 [1:40:22<1:04:52, 4.81it/s] 61%|██████ | 29286/48008 [1:40:23<1:03:40, 4.90it/s] 61%|██████ | 29287/48008 [1:40:23<1:03:00, 4.95it/s] 61%|██████ | 29288/48008 [1:40:23<1:02:07, 5.02it/s] 61%|██████ | 29289/48008 [1:40:23<1:01:01, 5.11it/s] 61%|██████ | 29290/48008 [1:40:23<1:00:15, 5.18it/s] 61%|██████ | 29291/48008 [1:40:24<1:00:05, 5.19it/s] 61%|██████ | 29292/48008 [1:40:24<59:44, 5.22it/s] 61%|██████ | 29293/48008 [1:40:24<1:00:06, 5.19it/s] 61%|██████ | 29294/48008 [1:40:24<59:59, 5.20it/s] 61%|██████ | 29295/48008 [1:40:24<1:01:32, 5.07it/s] 61%|██████ | 29296/48008 [1:40:25<1:01:52, 5.04it/s] 61%|██████ | 29297/48008 [1:40:25<1:01:58, 5.03it/s] 61%|██████ | 29298/48008 [1:40:25<1:01:19, 5.09it/s] 61%|██████ | 29299/48008 [1:40:25<1:01:49, 5.04it/s] 61%|██████ | 29300/48008 [1:40:25<1:05:37, 4.75it/s] {'loss': 0.3016, 'grad_norm': 0.7265625, 'learning_rate': 7.794117647058824e-05, 'epoch': 0.61} 61%|██████ | 29300/48008 [1:40:25<1:05:37, 4.75it/s] 61%|██████ | 29301/48008 [1:40:26<1:05:06, 4.79it/s] 61%|██████ | 29302/48008 [1:40:26<1:03:30, 4.91it/s] 61%|██████ | 29303/48008 [1:40:26<1:20:28, 3.87it/s] 61%|██████ | 29304/48008 [1:40:26<1:13:38, 4.23it/s] 61%|██████ | 29305/48008 [1:40:27<1:09:52, 4.46it/s] 61%|██████ | 29306/48008 [1:40:27<1:06:51, 4.66it/s] 61%|██████ | 29307/48008 [1:40:27<1:04:18, 4.85it/s] 61%|██████ | 29308/48008 [1:40:27<1:03:00, 4.95it/s] 61%|██████ | 29309/48008 [1:40:27<1:02:36, 4.98it/s] 61%|██████ | 29310/48008 [1:40:28<1:01:44, 5.05it/s] 61%|██████ | 29311/48008 [1:40:28<1:01:13, 5.09it/s] 61%|██████ | 29312/48008 [1:40:28<1:00:14, 5.17it/s] 61%|██████ | 29313/48008 [1:40:28<59:59, 5.19it/s] 61%|██████ | 29314/48008 [1:40:28<1:01:08, 5.10it/s] 61%|██████ | 29315/48008 [1:40:29<1:01:10, 5.09it/s] 61%|██████ | 29316/48008 [1:40:29<1:00:50, 5.12it/s] 61%|██████ | 29317/48008 [1:40:30<2:31:14, 2.06it/s] 61%|██████ | 29318/48008 [1:40:30<2:03:53, 2.51it/s] 61%|██████ | 29319/48008 [1:40:30<1:44:26, 2.98it/s] 61%|██████ | 29320/48008 [1:40:30<1:31:42, 3.40it/s] 61%|██████ | 29321/48008 [1:40:31<1:40:26, 3.10it/s] 61%|██████ | 29322/48008 [1:40:31<1:29:16, 3.49it/s] 61%|██████ | 29323/48008 [1:40:31<1:20:16, 3.88it/s] 61%|██████ | 29324/48008 [1:40:31<1:14:07, 4.20it/s] 61%|██████ | 29325/48008 [1:40:32<1:20:31, 3.87it/s] 61%|██████ | 29326/48008 [1:40:32<1:13:33, 4.23it/s] 61%|██████ | 29327/48008 [1:40:32<1:10:31, 4.42it/s] 61%|██████ | 29328/48008 [1:40:32<1:07:42, 4.60it/s] 61%|██████ | 29329/48008 [1:40:33<1:05:24, 4.76it/s] 61%|██████ | 29330/48008 [1:40:33<1:04:02, 4.86it/s] 61%|██████ | 29331/48008 [1:40:33<1:02:29, 4.98it/s] 61%|██████ | 29332/48008 [1:40:33<1:01:03, 5.10it/s] 61%|██████ | 29333/48008 [1:40:33<1:01:11, 5.09it/s] 61%|██████ | 29334/48008 [1:40:33<1:00:40, 5.13it/s] 61%|██████ | 29335/48008 [1:40:34<1:00:24, 5.15it/s] 61%|██████ | 29336/48008 [1:40:34<1:00:06, 5.18it/s] 61%|██████ | 29337/48008 [1:40:34<59:59, 5.19it/s] 61%|██████ | 29338/48008 [1:40:34<1:00:08, 5.17it/s] 61%|██████ | 29339/48008 [1:40:35<1:10:45, 4.40it/s] 61%|██████ | 29340/48008 [1:40:35<1:07:29, 4.61it/s] 61%|██████ | 29341/48008 [1:40:35<1:05:14, 4.77it/s] 61%|██████ | 29342/48008 [1:40:35<1:03:34, 4.89it/s] 61%|██████ | 29343/48008 [1:40:35<1:02:28, 4.98it/s] 61%|██████ | 29344/48008 [1:40:36<1:02:07, 5.01it/s] 61%|██████ | 29345/48008 [1:40:36<1:01:27, 5.06it/s] 61%|██████ | 29346/48008 [1:40:36<1:00:39, 5.13it/s] 61%|██████ | 29347/48008 [1:40:36<59:54, 5.19it/s] 61%|██████ | 29348/48008 [1:40:36<59:58, 5.19it/s] 61%|██████ | 29349/48008 [1:40:36<59:42, 5.21it/s] 61%|██████ | 29350/48008 [1:40:37<59:30, 5.23it/s] {'loss': 0.2999, 'grad_norm': 0.94140625, 'learning_rate': 7.773287785369106e-05, 'epoch': 0.61} 61%|██████ | 29350/48008 [1:40:37<59:30, 5.23it/s] 61%|██████ | 29351/48008 [1:40:37<59:05, 5.26it/s] 61%|██████ | 29352/48008 [1:40:37<59:06, 5.26it/s] 61%|██████ | 29353/48008 [1:40:37<59:10, 5.25it/s] 61%|██████ | 29354/48008 [1:40:37<59:12, 5.25it/s] 61%|██████ | 29355/48008 [1:40:38<2:10:02, 2.39it/s] 61%|██████ | 29356/48008 [1:40:39<1:48:55, 2.85it/s] 61%|██████ | 29357/48008 [1:40:39<1:33:38, 3.32it/s] 61%|██████ | 29358/48008 [1:40:39<1:41:35, 3.06it/s] 61%|██████ | 29359/48008 [1:40:39<1:29:03, 3.49it/s] 61%|██████ | 29360/48008 [1:40:40<1:20:12, 3.88it/s] 61%|██████ | 29361/48008 [1:40:40<2:14:21, 2.31it/s] 61%|██████ | 29362/48008 [1:40:41<1:52:17, 2.77it/s] 61%|██████ | 29363/48008 [1:40:41<1:36:37, 3.22it/s] 61%|██████ | 29364/48008 [1:40:41<1:27:01, 3.57it/s] 61%|██████ | 29365/48008 [1:40:41<1:36:53, 3.21it/s] 61%|██████ | 29366/48008 [1:40:42<1:29:51, 3.46it/s] 61%|██████ | 29367/48008 [1:40:42<1:38:57, 3.14it/s] 61%|██████ | 29368/48008 [1:40:42<1:28:48, 3.50it/s] 61%|██████ | 29369/48008 [1:40:43<1:38:13, 3.16it/s] 61%|██████ | 29370/48008 [1:40:43<1:26:09, 3.61it/s] 61%|██████ | 29371/48008 [1:40:43<1:18:05, 3.98it/s] 61%|██████ | 29372/48008 [1:40:43<1:12:22, 4.29it/s] 61%|██████ | 29373/48008 [1:40:43<1:09:21, 4.48it/s] 61%|██████ | 29374/48008 [1:40:44<1:06:00, 4.70it/s] 61%|██████ | 29375/48008 [1:40:44<1:03:45, 4.87it/s] 61%|██████ | 29376/48008 [1:40:44<1:02:17, 4.99it/s] 61%|██████ | 29377/48008 [1:40:44<1:01:24, 5.06it/s] 61%|██████ | 29378/48008 [1:40:44<1:00:46, 5.11it/s] 61%|██████ | 29379/48008 [1:40:44<59:55, 5.18it/s] 61%|██████ | 29380/48008 [1:40:45<1:00:18, 5.15it/s] 61%|██████ | 29381/48008 [1:40:45<1:00:36, 5.12it/s] 61%|██████ | 29382/48008 [1:40:45<1:00:37, 5.12it/s] 61%|██████ | 29383/48008 [1:40:45<1:00:01, 5.17it/s] 61%|██████ | 29384/48008 [1:40:45<1:01:04, 5.08it/s] 61%|██████ | 29385/48008 [1:40:46<1:00:01, 5.17it/s] 61%|██████ | 29386/48008 [1:40:46<59:09, 5.25it/s] 61%|██████ | 29387/48008 [1:40:46<59:01, 5.26it/s] 61%|██████ | 29388/48008 [1:40:46<1:03:24, 4.89it/s] 61%|██████ | 29389/48008 [1:40:46<1:06:29, 4.67it/s] 61%|██████ | 29390/48008 [1:40:47<1:04:11, 4.83it/s] 61%|██████ | 29391/48008 [1:40:47<1:02:09, 4.99it/s] 61%|██████ | 29392/48008 [1:40:47<1:01:19, 5.06it/s] 61%|██████ | 29393/48008 [1:40:47<1:18:50, 3.93it/s] 61%|██████ | 29394/48008 [1:40:48<1:12:17, 4.29it/s] 61%|██████ | 29395/48008 [1:40:48<1:09:31, 4.46it/s] 61%|██████ | 29396/48008 [1:40:48<1:07:02, 4.63it/s] 61%|██████ | 29397/48008 [1:40:49<2:02:48, 2.53it/s] 61%|██████ | 29398/48008 [1:40:50<2:47:54, 1.85it/s] 61%|██████ | 29399/48008 [1:40:50<2:15:56, 2.28it/s] 61%|██████ | 29400/48008 [1:40:50<1:54:12, 2.72it/s] {'loss': 0.2993, 'grad_norm': 0.1884765625, 'learning_rate': 7.752457923679388e-05, 'epoch': 0.61} 61%|██████ | 29400/48008 [1:40:50<1:54:12, 2.72it/s] 61%|██████ | 29401/48008 [1:40:50<1:38:11, 3.16it/s] 61%|██████ | 29402/48008 [1:40:51<1:26:34, 3.58it/s] 61%|██████ | 29403/48008 [1:40:51<1:18:21, 3.96it/s] 61%|██████ | 29404/48008 [1:40:51<1:12:49, 4.26it/s] 61%|██████▏ | 29405/48008 [1:40:51<1:10:20, 4.41it/s] 61%|██████▏ | 29406/48008 [1:40:51<1:11:19, 4.35it/s] 61%|██████▏ | 29407/48008 [1:40:52<1:07:24, 4.60it/s] 61%|██████▏ | 29408/48008 [1:40:52<1:04:38, 4.80it/s] 61%|██████▏ | 29409/48008 [1:40:52<1:07:17, 4.61it/s] 61%|██████▏ | 29410/48008 [1:40:52<1:04:20, 4.82it/s] 61%|██████▏ | 29411/48008 [1:40:52<1:03:01, 4.92it/s] 61%|██████▏ | 29412/48008 [1:40:53<1:01:14, 5.06it/s] 61%|██████▏ | 29413/48008 [1:40:53<1:00:17, 5.14it/s] 61%|██████▏ | 29414/48008 [1:40:53<1:04:17, 4.82it/s] 61%|██████▏ | 29415/48008 [1:40:53<1:02:41, 4.94it/s] 61%|██████▏ | 29416/48008 [1:40:53<1:02:18, 4.97it/s] 61%|██████▏ | 29417/48008 [1:40:54<1:01:19, 5.05it/s] 61%|██████▏ | 29418/48008 [1:40:54<1:02:01, 5.00it/s] 61%|██████▏ | 29419/48008 [1:40:54<1:00:50, 5.09it/s] 61%|██████▏ | 29420/48008 [1:40:54<1:00:18, 5.14it/s] 61%|██████▏ | 29421/48008 [1:40:54<1:00:03, 5.16it/s] 61%|██████▏ | 29422/48008 [1:40:55<1:10:31, 4.39it/s] 61%|██████▏ | 29423/48008 [1:40:55<1:08:19, 4.53it/s] 61%|██████▏ | 29424/48008 [1:40:55<1:06:28, 4.66it/s] 61%|██████▏ | 29425/48008 [1:40:55<1:04:08, 4.83it/s] 61%|██████▏ | 29426/48008 [1:40:55<1:04:16, 4.82it/s] 61%|██████▏ | 29427/48008 [1:40:56<1:01:57, 5.00it/s] 61%|██████▏ | 29428/48008 [1:40:56<1:02:07, 4.98it/s] 61%|██████▏ | 29429/48008 [1:40:56<1:00:10, 5.15it/s] 61%|██████▏ | 29430/48008 [1:40:56<59:53, 5.17it/s] 61%|██████▏ | 29431/48008 [1:40:56<59:46, 5.18it/s] 61%|██████▏ | 29432/48008 [1:40:57<59:32, 5.20it/s] 61%|██████▏ | 29433/48008 [1:40:57<59:17, 5.22it/s] 61%|██████▏ | 29434/48008 [1:40:57<59:17, 5.22it/s] 61%|██████▏ | 29435/48008 [1:40:57<59:42, 5.18it/s] 61%|██████▏ | 29436/48008 [1:40:57<59:46, 5.18it/s] 61%|██████▏ | 29437/48008 [1:40:58<59:44, 5.18it/s] 61%|██████▏ | 29438/48008 [1:40:58<59:35, 5.19it/s] 61%|██████▏ | 29439/48008 [1:40:58<1:03:57, 4.84it/s] 61%|██████▏ | 29440/48008 [1:40:58<1:03:00, 4.91it/s] 61%|██████▏ | 29441/48008 [1:40:58<1:01:50, 5.00it/s] 61%|██████▏ | 29442/48008 [1:40:59<1:00:54, 5.08it/s] 61%|██████▏ | 29443/48008 [1:40:59<1:11:02, 4.36it/s] 61%|██████▏ | 29444/48008 [1:40:59<1:07:26, 4.59it/s] 61%|██████▏ | 29445/48008 [1:40:59<1:05:12, 4.74it/s] 61%|██████▏ | 29446/48008 [1:40:59<1:03:26, 4.88it/s] 61%|██████▏ | 29447/48008 [1:41:00<1:01:58, 4.99it/s] 61%|██████▏ | 29448/48008 [1:41:00<1:01:44, 5.01it/s] 61%|██████▏ | 29449/48008 [1:41:00<1:01:28, 5.03it/s] 61%|██████▏ | 29450/48008 [1:41:00<1:00:53, 5.08it/s] 61%|██████▏ | 29450/48008 [1:41:00<1:00:53, 5.08it/s]{'loss': 0.3018, 'grad_norm': 0.1484375, 'learning_rate': 7.731628061989668e-05, 'epoch': 0.61} 61%|██████▏ | 29451/48008 [1:41:00<1:00:06, 5.14it/s] 61%|██████▏ | 29452/48008 [1:41:01<59:32, 5.19it/s] 61%|██████▏ | 29453/48008 [1:41:01<59:25, 5.20it/s] 61%|██████▏ | 29454/48008 [1:41:01<59:04, 5.23it/s] 61%|██████▏ | 29455/48008 [1:41:01<59:08, 5.23it/s] 61%|██████▏ | 29456/48008 [1:41:01<58:34, 5.28it/s] 61%|██████▏ | 29457/48008 [1:41:02<58:36, 5.27it/s] 61%|██████▏ | 29458/48008 [1:41:02<59:10, 5.22it/s] 61%|██████▏ | 29459/48008 [1:41:02<59:53, 5.16it/s] 61%|██████▏ | 29460/48008 [1:41:02<59:12, 5.22it/s] 61%|██████▏ | 29461/48008 [1:41:02<59:04, 5.23it/s] 61%|██████▏ | 29462/48008 [1:41:02<58:53, 5.25it/s] 61%|██████▏ | 29463/48008 [1:41:03<58:38, 5.27it/s] 61%|██████▏ | 29464/48008 [1:41:03<58:35, 5.27it/s] 61%|██████▏ | 29465/48008 [1:41:03<59:12, 5.22it/s] 61%|██████▏ | 29466/48008 [1:41:03<59:11, 5.22it/s] 61%|██████▏ | 29467/48008 [1:41:03<59:41, 5.18it/s] 61%|██████▏ | 29468/48008 [1:41:04<1:00:03, 5.14it/s] 61%|██████▏ | 29469/48008 [1:41:04<1:00:47, 5.08it/s] 61%|██████▏ | 29470/48008 [1:41:04<1:00:43, 5.09it/s] 61%|██████▏ | 29471/48008 [1:41:04<59:52, 5.16it/s] 61%|██████▏ | 29472/48008 [1:41:04<59:45, 5.17it/s] 61%|██████▏ | 29473/48008 [1:41:05<1:00:11, 5.13it/s] 61%|██████▏ | 29474/48008 [1:41:05<59:19, 5.21it/s] 61%|██████▏ | 29475/48008 [1:41:05<1:00:03, 5.14it/s] 61%|██████▏ | 29476/48008 [1:41:05<1:00:48, 5.08it/s] 61%|██████▏ | 29477/48008 [1:41:05<59:50, 5.16it/s] 61%|██████▏ | 29478/48008 [1:41:06<59:13, 5.21it/s] 61%|██████▏ | 29479/48008 [1:41:06<59:17, 5.21it/s] 61%|██████▏ | 29480/48008 [1:41:06<59:04, 5.23it/s] 61%|██████▏ | 29481/48008 [1:41:06<59:09, 5.22it/s] 61%|██████▏ | 29482/48008 [1:41:06<59:40, 5.17it/s] 61%|██████▏ | 29483/48008 [1:41:07<59:29, 5.19it/s] 61%|██████▏ | 29484/48008 [1:41:07<58:54, 5.24it/s] 61%|██████▏ | 29485/48008 [1:41:07<1:00:11, 5.13it/s] 61%|██████▏ | 29486/48008 [1:41:07<59:37, 5.18it/s] 61%|██████▏ | 29487/48008 [1:41:07<59:26, 5.19it/s] 61%|██████▏ | 29488/48008 [1:41:07<59:24, 5.20it/s] 61%|██████▏ | 29489/48008 [1:41:08<59:25, 5.19it/s] 61%|██████▏ | 29490/48008 [1:41:08<59:51, 5.16it/s] 61%|██████▏ | 29491/48008 [1:41:08<1:00:13, 5.12it/s] 61%|██████▏ | 29492/48008 [1:41:08<59:42, 5.17it/s] 61%|██████▏ | 29493/48008 [1:41:08<59:15, 5.21it/s] 61%|██████▏ | 29494/48008 [1:41:09<59:08, 5.22it/s] 61%|██████▏ | 29495/48008 [1:41:09<59:05, 5.22it/s] 61%|██████▏ | 29496/48008 [1:41:09<59:58, 5.14it/s] 61%|██████▏ | 29497/48008 [1:41:09<59:23, 5.19it/s] 61%|██████▏ | 29498/48008 [1:41:09<1:00:28, 5.10it/s] 61%|██████▏ | 29499/48008 [1:41:10<1:00:04, 5.13it/s] 61%|██████▏ | 29500/48008 [1:41:10<59:20, 5.20it/s] 61%|██████▏ | 29500/48008 [1:41:10<59:20, 5.20it/s]{'loss': 0.3025, 'grad_norm': 0.203125, 'learning_rate': 7.71079820029995e-05, 'epoch': 0.61} 61%|██████▏ | 29501/48008 [1:41:10<59:19, 5.20it/s] 61%|██████▏ | 29502/48008 [1:41:10<59:06, 5.22it/s] 61%|██████▏ | 29503/48008 [1:41:10<58:38, 5.26it/s] 61%|██████▏ | 29504/48008 [1:41:11<1:03:00, 4.89it/s] 61%|██████▏ | 29505/48008 [1:41:11<1:02:36, 4.93it/s] 61%|██████▏ | 29506/48008 [1:41:11<1:01:18, 5.03it/s] 61%|██████▏ | 29507/48008 [1:41:11<1:00:02, 5.14it/s] 61%|██████▏ | 29508/48008 [1:41:11<58:39, 5.26it/s] 61%|██████▏ | 29509/48008 [1:41:12<58:41, 5.25it/s] 61%|██████▏ | 29510/48008 [1:41:12<58:36, 5.26it/s] 61%|██████▏ | 29511/48008 [1:41:12<59:37, 5.17it/s] 61%|██████▏ | 29512/48008 [1:41:12<1:00:23, 5.11it/s] 61%|██████▏ | 29513/48008 [1:41:12<1:01:13, 5.04it/s] 61%|██████▏ | 29514/48008 [1:41:13<1:04:48, 4.76it/s] 61%|██████▏ | 29515/48008 [1:41:13<1:02:46, 4.91it/s] 61%|██████▏ | 29516/48008 [1:41:13<1:02:30, 4.93it/s] 61%|██████▏ | 29517/48008 [1:41:13<1:01:23, 5.02it/s] 61%|██████▏ | 29518/48008 [1:41:13<1:01:27, 5.01it/s] 61%|██████▏ | 29519/48008 [1:41:14<1:00:48, 5.07it/s] 61%|██████▏ | 29520/48008 [1:41:14<59:56, 5.14it/s] 61%|██████▏ | 29521/48008 [1:41:14<1:00:35, 5.09it/s] 61%|██████▏ | 29522/48008 [1:41:14<59:47, 5.15it/s] 61%|██████▏ | 29523/48008 [1:41:14<59:23, 5.19it/s] 61%|██████▏ | 29524/48008 [1:41:15<58:57, 5.23it/s] 62%|██████▏ | 29525/48008 [1:41:15<58:45, 5.24it/s] 62%|██████▏ | 29526/48008 [1:41:15<59:59, 5.13it/s] 62%|██████▏ | 29527/48008 [1:41:15<59:37, 5.17it/s] 62%|██████▏ | 29528/48008 [1:41:15<59:45, 5.15it/s] 62%|██████▏ | 29529/48008 [1:41:15<58:55, 5.23it/s] 62%|██████▏ | 29530/48008 [1:41:16<58:46, 5.24it/s] 62%|██████▏ | 29531/48008 [1:41:16<58:38, 5.25it/s] 62%|██████▏ | 29532/48008 [1:41:16<59:20, 5.19it/s] 62%|██████▏ | 29533/48008 [1:41:16<59:48, 5.15it/s] 62%|██████▏ | 29534/48008 [1:41:16<59:37, 5.16it/s] 62%|██████▏ | 29535/48008 [1:41:17<58:46, 5.24it/s] 62%|██████▏ | 29536/48008 [1:41:17<58:20, 5.28it/s] 62%|██████▏ | 29537/48008 [1:41:17<58:05, 5.30it/s] 62%|██████▏ | 29538/48008 [1:41:17<58:58, 5.22it/s] 62%|██████▏ | 29539/48008 [1:41:17<1:03:09, 4.87it/s] 62%|██████▏ | 29540/48008 [1:41:18<1:03:41, 4.83it/s] 62%|██████▏ | 29541/48008 [1:41:18<1:13:10, 4.21it/s] 62%|██████▏ | 29542/48008 [1:41:18<1:08:48, 4.47it/s] 62%|██████▏ | 29543/48008 [1:41:18<1:05:56, 4.67it/s] 62%|██████▏ | 29544/48008 [1:41:19<1:03:24, 4.85it/s] 62%|██████▏ | 29545/48008 [1:41:19<1:01:55, 4.97it/s] 62%|██████▏ | 29546/48008 [1:41:19<1:00:33, 5.08it/s] 62%|██████▏ | 29547/48008 [1:41:19<1:10:45, 4.35it/s] 62%|██████▏ | 29548/48008 [1:41:19<1:07:10, 4.58it/s] 62%|██████▏ | 29549/48008 [1:41:20<1:04:51, 4.74it/s] 62%|██████▏ | 29550/48008 [1:41:20<1:03:05, 4.88it/s] {'loss': 0.3007, 'grad_norm': 0.283203125, 'learning_rate': 7.689968338610232e-05, 'epoch': 0.62} 62%|██████▏ | 29550/48008 [1:41:20<1:03:05, 4.88it/s] 62%|██████▏ | 29551/48008 [1:41:20<1:02:30, 4.92it/s] 62%|██████▏ | 29552/48008 [1:41:20<1:01:16, 5.02it/s] 62%|██████▏ | 29553/48008 [1:41:20<1:01:13, 5.02it/s] 62%|██████▏ | 29554/48008 [1:41:21<1:00:04, 5.12it/s] 62%|██████▏ | 29555/48008 [1:41:21<59:10, 5.20it/s] 62%|██████▏ | 29556/48008 [1:41:21<59:19, 5.18it/s] 62%|██████▏ | 29557/48008 [1:41:21<59:07, 5.20it/s] 62%|██████▏ | 29558/48008 [1:41:21<58:25, 5.26it/s] 62%|██████▏ | 29559/48008 [1:41:22<1:02:50, 4.89it/s] 62%|██████▏ | 29560/48008 [1:41:22<1:01:39, 4.99it/s] 62%|██████▏ | 29561/48008 [1:41:22<1:00:38, 5.07it/s] 62%|██████▏ | 29562/48008 [1:41:22<59:48, 5.14it/s] 62%|██████▏ | 29563/48008 [1:41:22<59:18, 5.18it/s] 62%|██████▏ | 29564/48008 [1:41:23<1:00:25, 5.09it/s] 62%|██████▏ | 29565/48008 [1:41:23<1:01:01, 5.04it/s] 62%|██████▏ | 29566/48008 [1:41:23<1:00:48, 5.05it/s] 62%|██████▏ | 29567/48008 [1:41:23<1:00:15, 5.10it/s] 62%|██████▏ | 29568/48008 [1:41:23<59:41, 5.15it/s] 62%|██████▏ | 29569/48008 [1:41:23<59:23, 5.17it/s] 62%|██████▏ | 29570/48008 [1:41:24<58:41, 5.24it/s] 62%|██████▏ | 29571/48008 [1:41:24<59:42, 5.15it/s] 62%|██████▏ | 29572/48008 [1:41:24<59:08, 5.20it/s] 62%|██████▏ | 29573/48008 [1:41:24<1:09:43, 4.41it/s] 62%|██████▏ | 29574/48008 [1:41:25<1:06:18, 4.63it/s] 62%|██████▏ | 29575/48008 [1:41:25<1:04:00, 4.80it/s] 62%|██████▏ | 29576/48008 [1:41:25<1:02:12, 4.94it/s] 62%|██████▏ | 29577/48008 [1:41:25<1:01:07, 5.03it/s] 62%|██████▏ | 29578/48008 [1:41:25<1:00:12, 5.10it/s] 62%|██████▏ | 29579/48008 [1:41:26<1:00:55, 5.04it/s] 62%|██████▏ | 29580/48008 [1:41:26<1:00:41, 5.06it/s] 62%|██████▏ | 29581/48008 [1:41:26<59:53, 5.13it/s] 62%|██████▏ | 29582/48008 [1:41:26<59:27, 5.17it/s] 62%|██████▏ | 29583/48008 [1:41:26<58:42, 5.23it/s] 62%|██████▏ | 29584/48008 [1:41:27<1:02:52, 4.88it/s] 62%|██████▏ | 29585/48008 [1:41:27<1:01:55, 4.96it/s] 62%|██████▏ | 29586/48008 [1:41:27<1:00:34, 5.07it/s] 62%|██████▏ | 29587/48008 [1:41:27<1:00:55, 5.04it/s] 62%|██████▏ | 29588/48008 [1:41:27<1:00:14, 5.10it/s] 62%|██████▏ | 29589/48008 [1:41:28<1:04:00, 4.80it/s] 62%|██████▏ | 29590/48008 [1:41:28<1:03:36, 4.83it/s] 62%|██████▏ | 29591/48008 [1:41:28<1:02:15, 4.93it/s] 62%|██████▏ | 29592/48008 [1:41:28<1:00:46, 5.05it/s] 62%|██████▏ | 29593/48008 [1:41:28<1:00:46, 5.05it/s] 62%|██████▏ | 29594/48008 [1:41:29<59:46, 5.13it/s] 62%|██████▏ | 29595/48008 [1:41:29<1:00:22, 5.08it/s] 62%|██████▏ | 29596/48008 [1:41:29<1:00:26, 5.08it/s] 62%|██████▏ | 29597/48008 [1:41:29<58:52, 5.21it/s] 62%|██████▏ | 29598/48008 [1:41:29<1:16:47, 4.00it/s] 62%|██████▏ | 29599/48008 [1:41:30<1:12:17, 4.24it/s] 62%|██████▏ | 29600/48008 [1:41:30<1:09:21, 4.42it/s] 62%|██████▏ | 29600/48008 [1:41:30<1:09:21, 4.42it/s]{'loss': 0.2998, 'grad_norm': 0.1708984375, 'learning_rate': 7.669138476920513e-05, 'epoch': 0.62} 62%|██████▏ | 29601/48008 [1:41:30<1:06:17, 4.63it/s] 62%|██████▏ | 29602/48008 [1:41:30<1:04:10, 4.78it/s] 62%|██████▏ | 29603/48008 [1:41:30<1:02:37, 4.90it/s] 62%|██████▏ | 29604/48008 [1:41:31<1:02:09, 4.93it/s] 62%|██████▏ | 29605/48008 [1:41:31<1:00:22, 5.08it/s] 62%|██████▏ | 29606/48008 [1:41:31<59:46, 5.13it/s] 62%|██████▏ | 29607/48008 [1:41:31<59:27, 5.16it/s] 62%|██████▏ | 29608/48008 [1:41:31<1:00:21, 5.08it/s] 62%|██████▏ | 29609/48008 [1:41:32<1:00:30, 5.07it/s] 62%|██████▏ | 29610/48008 [1:41:32<59:58, 5.11it/s] 62%|██████▏ | 29611/48008 [1:41:32<1:17:34, 3.95it/s] 62%|██████▏ | 29612/48008 [1:41:32<1:12:00, 4.26it/s] 62%|██████▏ | 29613/48008 [1:41:33<1:08:02, 4.51it/s] 62%|██████▏ | 29614/48008 [1:41:33<1:06:19, 4.62it/s] 62%|██████▏ | 29615/48008 [1:41:33<1:03:59, 4.79it/s] 62%|██████▏ | 29616/48008 [1:41:33<1:02:26, 4.91it/s] 62%|██████▏ | 29617/48008 [1:41:33<1:01:15, 5.00it/s] 62%|██████▏ | 29618/48008 [1:41:34<59:58, 5.11it/s] 62%|██████▏ | 29619/48008 [1:41:34<59:45, 5.13it/s] 62%|██████▏ | 29620/48008 [1:41:34<59:15, 5.17it/s] 62%|██████▏ | 29621/48008 [1:41:34<1:17:32, 3.95it/s] 62%|██████▏ | 29622/48008 [1:41:35<1:12:03, 4.25it/s] 62%|██████▏ | 29623/48008 [1:41:35<1:07:52, 4.51it/s] 62%|██████▏ | 29624/48008 [1:41:35<1:05:05, 4.71it/s] 62%|██████▏ | 29625/48008 [1:41:35<1:07:27, 4.54it/s] 62%|██████▏ | 29626/48008 [1:41:35<1:04:24, 4.76it/s] 62%|██████▏ | 29627/48008 [1:41:36<1:02:36, 4.89it/s] 62%|██████▏ | 29628/48008 [1:41:36<1:01:14, 5.00it/s] 62%|██████▏ | 29629/48008 [1:41:36<1:02:15, 4.92it/s] 62%|██████▏ | 29630/48008 [1:41:36<1:01:18, 5.00it/s] 62%|██████▏ | 29631/48008 [1:41:36<1:00:16, 5.08it/s] 62%|██████▏ | 29632/48008 [1:41:37<1:01:05, 5.01it/s] 62%|██████▏ | 29633/48008 [1:41:37<1:00:15, 5.08it/s] 62%|██████▏ | 29634/48008 [1:41:37<1:01:01, 5.02it/s] 62%|██████▏ | 29635/48008 [1:41:37<1:04:43, 4.73it/s] 62%|██████▏ | 29636/48008 [1:41:37<1:07:04, 4.57it/s] 62%|██████▏ | 29637/48008 [1:41:38<1:04:32, 4.74it/s] 62%|██████▏ | 29638/48008 [1:41:38<1:02:54, 4.87it/s] 62%|██████▏ | 29639/48008 [1:41:38<1:01:42, 4.96it/s] 62%|██████▏ | 29640/48008 [1:41:38<1:18:53, 3.88it/s] 62%|██████▏ | 29641/48008 [1:41:39<1:13:24, 4.17it/s] 62%|██████▏ | 29642/48008 [1:41:39<1:09:14, 4.42it/s] 62%|██████▏ | 29643/48008 [1:41:39<1:07:10, 4.56it/s] 62%|██████▏ | 29644/48008 [1:41:39<1:04:36, 4.74it/s] 62%|██████▏ | 29645/48008 [1:41:39<1:02:48, 4.87it/s] 62%|██████▏ | 29646/48008 [1:41:40<1:05:44, 4.65it/s] 62%|██████▏ | 29647/48008 [1:41:40<1:03:07, 4.85it/s] 62%|██████▏ | 29648/48008 [1:41:40<1:05:56, 4.64it/s] 62%|██████▏ | 29649/48008 [1:41:40<1:03:33, 4.81it/s] 62%|██████▏ | 29650/48008 [1:41:40<1:01:48, 4.95it/s] {'loss': 0.2988, 'grad_norm': 0.2353515625, 'learning_rate': 7.648308615230795e-05, 'epoch': 0.62} 62%|██████▏ | 29650/48008 [1:41:40<1:01:48, 4.95it/s] 62%|██████▏ | 29651/48008 [1:41:41<1:01:17, 4.99it/s] 62%|██████▏ | 29652/48008 [1:41:41<1:00:31, 5.05it/s] 62%|██████▏ | 29653/48008 [1:41:41<59:57, 5.10it/s] 62%|██████▏ | 29654/48008 [1:41:41<1:17:28, 3.95it/s] 62%|██████▏ | 29655/48008 [1:41:42<1:11:50, 4.26it/s] 62%|██████▏ | 29656/48008 [1:41:42<1:07:28, 4.53it/s] 62%|██████▏ | 29657/48008 [1:41:42<1:05:47, 4.65it/s] 62%|██████▏ | 29658/48008 [1:41:42<1:04:09, 4.77it/s] 62%|██████▏ | 29659/48008 [1:41:42<1:03:32, 4.81it/s] 62%|██████▏ | 29660/48008 [1:41:43<1:02:03, 4.93it/s] 62%|██████▏ | 29661/48008 [1:41:43<1:01:28, 4.97it/s] 62%|██████▏ | 29662/48008 [1:41:43<1:00:31, 5.05it/s] 62%|██████▏ | 29663/48008 [1:41:43<59:50, 5.11it/s] 62%|██████▏ | 29664/48008 [1:41:43<1:00:20, 5.07it/s] 62%|██████▏ | 29665/48008 [1:41:43<59:32, 5.13it/s] 62%|██████▏ | 29666/48008 [1:41:44<59:08, 5.17it/s] 62%|██████▏ | 29667/48008 [1:41:44<1:03:09, 4.84it/s] 62%|██████▏ | 29668/48008 [1:41:44<1:02:55, 4.86it/s] 62%|██████▏ | 29669/48008 [1:41:44<1:01:24, 4.98it/s] 62%|██████▏ | 29670/48008 [1:41:44<1:00:03, 5.09it/s] 62%|██████▏ | 29671/48008 [1:41:45<1:00:24, 5.06it/s] 62%|██████▏ | 29672/48008 [1:41:45<1:00:58, 5.01it/s] 62%|██████▏ | 29673/48008 [1:41:45<1:00:56, 5.01it/s] 62%|██████▏ | 29674/48008 [1:41:45<1:00:08, 5.08it/s] 62%|██████▏ | 29675/48008 [1:41:45<1:00:05, 5.08it/s] 62%|██████▏ | 29676/48008 [1:41:46<59:19, 5.15it/s] 62%|██████▏ | 29677/48008 [1:41:46<58:34, 5.22it/s] 62%|██████▏ | 29678/48008 [1:41:46<1:09:09, 4.42it/s] 62%|██████▏ | 29679/48008 [1:41:46<1:05:52, 4.64it/s] 62%|██████▏ | 29680/48008 [1:41:47<1:04:42, 4.72it/s] 62%|██████▏ | 29681/48008 [1:41:47<1:02:48, 4.86it/s] 62%|██████▏ | 29682/48008 [1:41:47<1:01:29, 4.97it/s] 62%|██████▏ | 29683/48008 [1:41:47<1:00:34, 5.04it/s] 62%|██████▏ | 29684/48008 [1:41:47<1:00:01, 5.09it/s] 62%|██████▏ | 29685/48008 [1:41:48<1:00:39, 5.03it/s] 62%|██████▏ | 29686/48008 [1:41:48<1:00:02, 5.09it/s] 62%|██████▏ | 29687/48008 [1:41:48<59:09, 5.16it/s] 62%|██████▏ | 29688/48008 [1:41:48<58:23, 5.23it/s] 62%|██████▏ | 29689/48008 [1:41:48<58:24, 5.23it/s] 62%|██████▏ | 29690/48008 [1:41:48<59:34, 5.12it/s] 62%|██████▏ | 29691/48008 [1:41:49<59:10, 5.16it/s] 62%|██████▏ | 29692/48008 [1:41:49<59:31, 5.13it/s] 62%|██████▏ | 29693/48008 [1:41:49<59:45, 5.11it/s] 62%|██████▏ | 29694/48008 [1:41:49<58:50, 5.19it/s] 62%|██████▏ | 29695/48008 [1:41:49<58:45, 5.19it/s] 62%|██████▏ | 29696/48008 [1:41:50<59:41, 5.11it/s] 62%|██████▏ | 29697/48008 [1:41:50<59:22, 5.14it/s] 62%|██████▏ | 29698/48008 [1:41:50<58:40, 5.20it/s] 62%|██████▏ | 29699/48008 [1:41:50<58:52, 5.18it/s] 62%|██████▏ | 29700/48008 [1:41:50<58:41, 5.20it/s] {'loss': 0.3014, 'grad_norm': 0.1865234375, 'learning_rate': 7.627478753541077e-05, 'epoch': 0.62} 62%|██████▏ | 29700/48008 [1:41:50<58:41, 5.20it/s] 62%|██████▏ | 29701/48008 [1:41:51<58:07, 5.25it/s] 62%|██████▏ | 29702/48008 [1:41:51<58:21, 5.23it/s] 62%|██████▏ | 29703/48008 [1:41:51<58:15, 5.24it/s] 62%|██████▏ | 29704/48008 [1:41:51<59:26, 5.13it/s] 62%|██████▏ | 29705/48008 [1:41:51<58:32, 5.21it/s] 62%|██████▏ | 29706/48008 [1:41:52<1:00:00, 5.08it/s] 62%|██████▏ | 29707/48008 [1:41:52<59:41, 5.11it/s] 62%|██████▏ | 29708/48008 [1:41:52<59:46, 5.10it/s] 62%|██████▏ | 29709/48008 [1:41:52<58:54, 5.18it/s] 62%|██████▏ | 29710/48008 [1:41:52<58:25, 5.22it/s] 62%|██████▏ | 29711/48008 [1:41:53<1:16:10, 4.00it/s] 62%|██████▏ | 29712/48008 [1:41:53<1:10:14, 4.34it/s] 62%|██████▏ | 29713/48008 [1:41:53<1:06:33, 4.58it/s] 62%|██████▏ | 29714/48008 [1:41:53<1:04:36, 4.72it/s] 62%|██████▏ | 29715/48008 [1:41:53<1:02:49, 4.85it/s] 62%|██████▏ | 29716/48008 [1:41:54<1:00:59, 5.00it/s] 62%|██████▏ | 29717/48008 [1:41:54<59:57, 5.08it/s] 62%|██████▏ | 29718/48008 [1:41:54<59:27, 5.13it/s] 62%|██████▏ | 29719/48008 [1:41:54<58:23, 5.22it/s] 62%|██████▏ | 29720/48008 [1:41:54<58:25, 5.22it/s] 62%|██████▏ | 29721/48008 [1:41:55<58:09, 5.24it/s] 62%|██████▏ | 29722/48008 [1:41:55<1:08:40, 4.44it/s] 62%|██████▏ | 29723/48008 [1:41:55<1:05:25, 4.66it/s] 62%|██████▏ | 29724/48008 [1:41:55<1:04:00, 4.76it/s] 62%|██████▏ | 29725/48008 [1:41:55<1:01:52, 4.92it/s] 62%|██████▏ | 29726/48008 [1:41:56<1:00:24, 5.04it/s] 62%|██████▏ | 29727/48008 [1:41:56<59:24, 5.13it/s] 62%|██████▏ | 29728/48008 [1:41:56<58:45, 5.19it/s] 62%|██████▏ | 29729/48008 [1:41:56<59:34, 5.11it/s] 62%|██████▏ | 29730/48008 [1:41:56<59:21, 5.13it/s] 62%|██████▏ | 29731/48008 [1:41:57<1:00:02, 5.07it/s] 62%|██████▏ | 29732/48008 [1:41:57<58:48, 5.18it/s] 62%|██████▏ | 29733/48008 [1:41:57<59:42, 5.10it/s] 62%|██████▏ | 29734/48008 [1:41:57<59:08, 5.15it/s] 62%|██████▏ | 29735/48008 [1:41:57<1:03:04, 4.83it/s] 62%|██████▏ | 29736/48008 [1:41:58<1:01:22, 4.96it/s] 62%|██████▏ | 29737/48008 [1:41:58<1:00:21, 5.04it/s] 62%|██████▏ | 29738/48008 [1:41:58<59:56, 5.08it/s] 62%|██████▏ | 29739/48008 [1:41:58<59:03, 5.16it/s] 62%|██████▏ | 29740/48008 [1:41:58<58:57, 5.16it/s] 62%|██████▏ | 29741/48008 [1:41:59<58:40, 5.19it/s] 62%|██████▏ | 29742/48008 [1:41:59<58:23, 5.21it/s] 62%|██████▏ | 29743/48008 [1:41:59<58:18, 5.22it/s] 62%|██████▏ | 29744/48008 [1:41:59<59:33, 5.11it/s] 62%|██████▏ | 29745/48008 [1:41:59<59:27, 5.12it/s] 62%|██████▏ | 29746/48008 [1:42:00<58:38, 5.19it/s] 62%|██████▏ | 29747/48008 [1:42:00<58:17, 5.22it/s] 62%|██████▏ | 29748/48008 [1:42:00<58:09, 5.23it/s] 62%|██████▏ | 29749/48008 [1:42:00<58:02, 5.24it/s] 62%|██████▏ | 29750/48008 [1:42:00<1:08:32, 4.44it/s] {'loss': 0.3005, 'grad_norm': 0.44140625, 'learning_rate': 7.606648891851358e-05, 'epoch': 0.62} 62%|██████▏ | 29750/48008 [1:42:00<1:08:32, 4.44it/s] 62%|██████▏ | 29751/48008 [1:42:01<1:05:17, 4.66it/s] 62%|██████▏ | 29752/48008 [1:42:01<1:03:09, 4.82it/s] 62%|██████▏ | 29753/48008 [1:42:01<1:01:42, 4.93it/s] 62%|██████▏ | 29754/48008 [1:42:01<1:00:38, 5.02it/s] 62%|██████▏ | 29755/48008 [1:42:01<59:59, 5.07it/s] 62%|██████▏ | 29756/48008 [1:42:02<59:17, 5.13it/s] 62%|██████▏ | 29757/48008 [1:42:02<1:00:06, 5.06it/s] 62%|██████▏ | 29758/48008 [1:42:02<1:00:04, 5.06it/s] 62%|██████▏ | 29759/48008 [1:42:02<59:36, 5.10it/s] 62%|██████▏ | 29760/48008 [1:42:02<59:36, 5.10it/s] 62%|██████▏ | 29761/48008 [1:42:03<58:29, 5.20it/s] 62%|██████▏ | 29762/48008 [1:42:03<58:16, 5.22it/s] 62%|██████▏ | 29763/48008 [1:42:03<59:22, 5.12it/s] 62%|██████▏ | 29764/48008 [1:42:03<58:37, 5.19it/s] 62%|██████▏ | 29765/48008 [1:42:03<58:15, 5.22it/s] 62%|██████▏ | 29766/48008 [1:42:04<58:34, 5.19it/s] 62%|██████▏ | 29767/48008 [1:42:04<58:46, 5.17it/s] 62%|██████▏ | 29768/48008 [1:42:04<58:11, 5.22it/s] 62%|██████▏ | 29769/48008 [1:42:04<58:16, 5.22it/s] 62%|██████▏ | 29770/48008 [1:42:04<58:03, 5.24it/s] 62%|██████▏ | 29771/48008 [1:42:04<58:20, 5.21it/s] 62%|██████▏ | 29772/48008 [1:42:05<59:10, 5.14it/s] 62%|██████▏ | 29773/48008 [1:42:05<59:26, 5.11it/s] 62%|██████▏ | 29774/48008 [1:42:05<58:55, 5.16it/s] 62%|██████▏ | 29775/48008 [1:42:05<58:41, 5.18it/s] 62%|██████▏ | 29776/48008 [1:42:05<58:29, 5.20it/s] 62%|██████▏ | 29777/48008 [1:42:06<57:51, 5.25it/s] 62%|██████▏ | 29778/48008 [1:42:06<58:02, 5.23it/s] 62%|██████▏ | 29779/48008 [1:42:06<58:24, 5.20it/s] 62%|██████▏ | 29780/48008 [1:42:06<58:17, 5.21it/s] 62%|██████▏ | 29781/48008 [1:42:06<58:46, 5.17it/s] 62%|██████▏ | 29782/48008 [1:42:07<58:36, 5.18it/s] 62%|██████▏ | 29783/48008 [1:42:07<58:17, 5.21it/s] 62%|██████▏ | 29784/48008 [1:42:07<58:41, 5.18it/s] 62%|██████▏ | 29785/48008 [1:42:07<1:00:10, 5.05it/s] 62%|██████▏ | 29786/48008 [1:42:07<59:10, 5.13it/s] 62%|██████▏ | 29787/48008 [1:42:08<1:16:43, 3.96it/s] 62%|██████▏ | 29788/48008 [1:42:08<1:11:13, 4.26it/s] 62%|██████▏ | 29789/48008 [1:42:08<1:07:36, 4.49it/s] 62%|██████▏ | 29790/48008 [1:42:08<1:04:41, 4.69it/s] 62%|██████▏ | 29791/48008 [1:42:09<1:02:36, 4.85it/s] 62%|██████▏ | 29792/48008 [1:42:09<1:01:39, 4.92it/s] 62%|██████▏ | 29793/48008 [1:42:09<1:00:37, 5.01it/s] 62%|██████▏ | 29794/48008 [1:42:09<59:32, 5.10it/s] 62%|██████▏ | 29795/48008 [1:42:10<1:16:55, 3.95it/s] 62%|██████▏ | 29796/48008 [1:42:10<1:11:27, 4.25it/s] 62%|██████▏ | 29797/48008 [1:42:10<1:07:30, 4.50it/s] 62%|██████▏ | 29798/48008 [1:42:10<1:04:38, 4.70it/s] 62%|██████▏ | 29799/48008 [1:42:10<1:03:35, 4.77it/s] 62%|██████▏ | 29800/48008 [1:42:11<1:06:10, 4.59it/s] 62%|██████▏ | 29800/48008 [1:42:11<1:06:10, 4.59it/s]{'loss': 0.301, 'grad_norm': 0.208984375, 'learning_rate': 7.58581903016164e-05, 'epoch': 0.62} 62%|██████▏ | 29801/48008 [1:42:11<1:03:12, 4.80it/s] 62%|██████▏ | 29802/48008 [1:42:11<1:01:50, 4.91it/s] 62%|██████▏ | 29803/48008 [1:42:11<1:01:36, 4.92it/s] 62%|██████▏ | 29804/48008 [1:42:11<1:01:19, 4.95it/s] 62%|██████▏ | 29805/48008 [1:42:12<1:00:19, 5.03it/s] 62%|██████▏ | 29806/48008 [1:42:12<1:01:23, 4.94it/s] 62%|██████▏ | 29807/48008 [1:42:12<1:02:05, 4.89it/s] 62%|██████▏ | 29808/48008 [1:42:12<1:00:51, 4.98it/s] 62%|██████▏ | 29809/48008 [1:42:13<1:17:51, 3.90it/s] 62%|██████▏ | 29810/48008 [1:42:13<1:11:25, 4.25it/s] 62%|██████▏ | 29811/48008 [1:42:13<1:07:21, 4.50it/s] 62%|██████▏ | 29812/48008 [1:42:13<1:05:41, 4.62it/s] 62%|██████▏ | 29813/48008 [1:42:13<1:03:14, 4.80it/s] 62%|██████▏ | 29814/48008 [1:42:13<1:01:36, 4.92it/s] 62%|██████▏ | 29815/48008 [1:42:14<1:00:25, 5.02it/s] 62%|██████▏ | 29816/48008 [1:42:14<1:00:21, 5.02it/s] 62%|██████▏ | 29817/48008 [1:42:14<1:01:20, 4.94it/s] 62%|██████▏ | 29818/48008 [1:42:14<1:00:59, 4.97it/s] 62%|██████▏ | 29819/48008 [1:42:14<1:01:55, 4.90it/s] 62%|██████▏ | 29820/48008 [1:42:15<1:01:04, 4.96it/s] 62%|██████▏ | 29821/48008 [1:42:15<1:00:19, 5.02it/s] 62%|██████▏ | 29822/48008 [1:42:15<59:34, 5.09it/s] 62%|██████▏ | 29823/48008 [1:42:15<58:55, 5.14it/s] 62%|██████▏ | 29824/48008 [1:42:15<59:09, 5.12it/s] 62%|██████▏ | 29825/48008 [1:42:16<58:27, 5.18it/s] 62%|██████▏ | 29826/48008 [1:42:16<58:43, 5.16it/s] 62%|██████▏ | 29827/48008 [1:42:16<58:19, 5.20it/s] 62%|██████▏ | 29828/48008 [1:42:16<58:06, 5.21it/s] 62%|██████▏ | 29829/48008 [1:42:16<59:08, 5.12it/s] 62%|██████▏ | 29830/48008 [1:42:17<58:35, 5.17it/s] 62%|██████▏ | 29831/48008 [1:42:17<57:55, 5.23it/s] 62%|██████▏ | 29832/48008 [1:42:17<57:37, 5.26it/s] 62%|██████▏ | 29833/48008 [1:42:17<1:01:54, 4.89it/s] 62%|██████▏ | 29834/48008 [1:42:17<1:00:40, 4.99it/s] 62%|██████▏ | 29835/48008 [1:42:18<59:29, 5.09it/s] 62%|██████▏ | 29836/48008 [1:42:18<59:05, 5.12it/s] 62%|██████▏ | 29837/48008 [1:42:18<58:39, 5.16it/s] 62%|██████▏ | 29838/48008 [1:42:18<1:02:31, 4.84it/s] 62%|██████▏ | 29839/48008 [1:42:18<1:01:02, 4.96it/s] 62%|██████▏ | 29840/48008 [1:42:19<59:59, 5.05it/s] 62%|██████▏ | 29841/48008 [1:42:19<59:14, 5.11it/s] 62%|██████▏ | 29842/48008 [1:42:19<1:00:01, 5.04it/s] 62%|██████▏ | 29843/48008 [1:42:19<1:03:31, 4.77it/s] 62%|██████▏ | 29844/48008 [1:42:19<1:01:39, 4.91it/s] 62%|██████▏ | 29845/48008 [1:42:20<1:00:31, 5.00it/s] 62%|██████▏ | 29846/48008 [1:42:20<59:14, 5.11it/s] 62%|██████▏ | 29847/48008 [1:42:20<59:13, 5.11it/s] 62%|██████▏ | 29848/48008 [1:42:20<58:45, 5.15it/s] 62%|██████▏ | 29849/48008 [1:42:20<59:14, 5.11it/s] 62%|██████▏ | 29850/48008 [1:42:21<58:50, 5.14it/s] {'loss': 0.2994, 'grad_norm': 0.181640625, 'learning_rate': 7.564989168471922e-05, 'epoch': 0.62} 62%|██████▏ | 29850/48008 [1:42:21<58:50, 5.14it/s] 62%|██████▏ | 29851/48008 [1:42:21<58:23, 5.18it/s] 62%|██████▏ | 29852/48008 [1:42:21<58:47, 5.15it/s] 62%|██████▏ | 29853/48008 [1:42:21<58:05, 5.21it/s] 62%|██████▏ | 29854/48008 [1:42:21<58:07, 5.21it/s] 62%|██████▏ | 29855/48008 [1:42:22<57:48, 5.23it/s] 62%|██████▏ | 29856/48008 [1:42:22<57:45, 5.24it/s] 62%|██████▏ | 29857/48008 [1:42:22<57:54, 5.22it/s] 62%|██████▏ | 29858/48008 [1:42:22<59:34, 5.08it/s] 62%|██████▏ | 29859/48008 [1:42:22<59:11, 5.11it/s] 62%|██████▏ | 29860/48008 [1:42:23<1:02:49, 4.81it/s] 62%|██████▏ | 29861/48008 [1:42:23<1:02:52, 4.81it/s] 62%|██████▏ | 29862/48008 [1:42:23<1:01:12, 4.94it/s] 62%|██████▏ | 29863/48008 [1:42:23<59:29, 5.08it/s] 62%|██████▏ | 29864/48008 [1:42:23<59:04, 5.12it/s] 62%|██████▏ | 29865/48008 [1:42:24<58:39, 5.15it/s] 62%|██████▏ | 29866/48008 [1:42:24<57:58, 5.22it/s] 62%|██████▏ | 29867/48008 [1:42:24<57:50, 5.23it/s] 62%|██████▏ | 29868/48008 [1:42:24<57:52, 5.22it/s] 62%|██████▏ | 29869/48008 [1:42:24<58:28, 5.17it/s] 62%|██████▏ | 29870/48008 [1:42:25<1:08:40, 4.40it/s] 62%|██████▏ | 29871/48008 [1:42:25<1:05:06, 4.64it/s] 62%|██████▏ | 29872/48008 [1:42:25<1:07:04, 4.51it/s] 62%|██████▏ | 29873/48008 [1:42:25<1:21:52, 3.69it/s] 62%|██████▏ | 29874/48008 [1:42:26<1:15:19, 4.01it/s] 62%|██████▏ | 29875/48008 [1:42:26<1:10:26, 4.29it/s] 62%|██████▏ | 29876/48008 [1:42:26<1:06:53, 4.52it/s] 62%|██████▏ | 29877/48008 [1:42:26<1:04:07, 4.71it/s] 62%|██████▏ | 29878/48008 [1:42:27<1:19:51, 3.78it/s] 62%|██████▏ | 29879/48008 [1:42:27<1:13:10, 4.13it/s] 62%|██████▏ | 29880/48008 [1:42:27<1:08:40, 4.40it/s] 62%|██████▏ | 29881/48008 [1:42:27<1:06:30, 4.54it/s] 62%|██████▏ | 29882/48008 [1:42:27<1:03:18, 4.77it/s] 62%|██████▏ | 29883/48008 [1:42:28<1:01:32, 4.91it/s] 62%|██████▏ | 29884/48008 [1:42:28<1:00:22, 5.00it/s] 62%|██████▏ | 29885/48008 [1:42:28<59:43, 5.06it/s] 62%|██████▏ | 29886/48008 [1:42:28<59:49, 5.05it/s] 62%|██████▏ | 29887/48008 [1:42:28<59:04, 5.11it/s] 62%|██████▏ | 29888/48008 [1:42:28<59:31, 5.07it/s] 62%|██████▏ | 29889/48008 [1:42:29<59:08, 5.11it/s] 62%|██████▏ | 29890/48008 [1:42:29<58:41, 5.15it/s] 62%|██████▏ | 29891/48008 [1:42:29<58:48, 5.13it/s] 62%|██████▏ | 29892/48008 [1:42:29<59:27, 5.08it/s] 62%|██████▏ | 29893/48008 [1:42:29<59:08, 5.10it/s] 62%|██████▏ | 29894/48008 [1:42:30<58:37, 5.15it/s] 62%|██████▏ | 29895/48008 [1:42:30<59:00, 5.12it/s] 62%|██████▏ | 29896/48008 [1:42:30<58:41, 5.14it/s] 62%|██████▏ | 29897/48008 [1:42:30<58:25, 5.17it/s] 62%|██████▏ | 29898/48008 [1:42:30<57:46, 5.22it/s] 62%|██████▏ | 29899/48008 [1:42:31<59:00, 5.11it/s] 62%|██████▏ | 29900/48008 [1:42:31<58:28, 5.16it/s] {'loss': 0.3003, 'grad_norm': 0.2158203125, 'learning_rate': 7.544159306782204e-05, 'epoch': 0.62} 62%|██████▏ | 29900/48008 [1:42:31<58:28, 5.16it/s] 62%|██████▏ | 29901/48008 [1:42:31<58:21, 5.17it/s] 62%|██████▏ | 29902/48008 [1:42:31<59:27, 5.07it/s] 62%|██████▏ | 29903/48008 [1:42:31<58:47, 5.13it/s] 62%|██████▏ | 29904/48008 [1:42:32<59:39, 5.06it/s] 62%|██████▏ | 29905/48008 [1:42:32<59:31, 5.07it/s] 62%|██████▏ | 29906/48008 [1:42:32<58:50, 5.13it/s] 62%|██████▏ | 29907/48008 [1:42:32<58:25, 5.16it/s] 62%|██████▏ | 29908/48008 [1:42:32<58:22, 5.17it/s] 62%|██████▏ | 29909/48008 [1:42:33<57:48, 5.22it/s] 62%|██████▏ | 29910/48008 [1:42:33<57:55, 5.21it/s] 62%|██████▏ | 29911/48008 [1:42:33<58:21, 5.17it/s] 62%|██████▏ | 29912/48008 [1:42:33<59:21, 5.08it/s] 62%|██████▏ | 29913/48008 [1:42:33<58:46, 5.13it/s] 62%|██████▏ | 29914/48008 [1:42:34<58:32, 5.15it/s] 62%|██████▏ | 29915/48008 [1:42:34<58:41, 5.14it/s] 62%|██████▏ | 29916/48008 [1:42:34<58:38, 5.14it/s] 62%|██████▏ | 29917/48008 [1:42:34<58:18, 5.17it/s] 62%|██████▏ | 29918/48008 [1:42:34<58:50, 5.12it/s] 62%|██████▏ | 29919/48008 [1:42:35<58:10, 5.18it/s] 62%|██████▏ | 29920/48008 [1:42:35<58:32, 5.15it/s] 62%|██████▏ | 29921/48008 [1:42:35<58:35, 5.14it/s] 62%|██████▏ | 29922/48008 [1:42:35<58:21, 5.17it/s] 62%|██████▏ | 29923/48008 [1:42:35<58:06, 5.19it/s] 62%|██████▏ | 29924/48008 [1:42:35<57:54, 5.20it/s] 62%|██████▏ | 29925/48008 [1:42:36<58:31, 5.15it/s] 62%|██████▏ | 29926/48008 [1:42:36<58:13, 5.18it/s] 62%|██████▏ | 29927/48008 [1:42:36<57:40, 5.23it/s] 62%|██████▏ | 29928/48008 [1:42:36<57:32, 5.24it/s] 62%|██████▏ | 29929/48008 [1:42:36<57:08, 5.27it/s] 62%|██████▏ | 29930/48008 [1:42:38<2:24:10, 2.09it/s] 62%|██████▏ | 29931/48008 [1:42:38<1:57:56, 2.55it/s] 62%|██████▏ | 29932/48008 [1:42:38<1:41:01, 2.98it/s] 62%|██████▏ | 29933/48008 [1:42:38<1:27:56, 3.43it/s] 62%|██████▏ | 29934/48008 [1:42:38<1:19:42, 3.78it/s] 62%|██████▏ | 29935/48008 [1:42:39<1:12:44, 4.14it/s] 62%|██████▏ | 29936/48008 [1:42:39<1:08:15, 4.41it/s] 62%|██████▏ | 29937/48008 [1:42:39<1:06:11, 4.55it/s] 62%|██████▏ | 29938/48008 [1:42:39<1:03:34, 4.74it/s] 62%|██████▏ | 29939/48008 [1:42:39<1:03:30, 4.74it/s] 62%|██████▏ | 29940/48008 [1:42:40<1:02:16, 4.84it/s] 62%|██████▏ | 29941/48008 [1:42:40<1:01:23, 4.91it/s] 62%|██████▏ | 29942/48008 [1:42:40<59:55, 5.03it/s] 62%|██████▏ | 29943/48008 [1:42:40<1:03:26, 4.75it/s] 62%|██████▏ | 29944/48008 [1:42:40<1:02:40, 4.80it/s] 62%|██████▏ | 29945/48008 [1:42:41<1:00:47, 4.95it/s] 62%|██████▏ | 29946/48008 [1:42:41<59:46, 5.04it/s] 62%|██████▏ | 29947/48008 [1:42:41<59:09, 5.09it/s] 62%|██████▏ | 29948/48008 [1:42:41<58:32, 5.14it/s] 62%|██████▏ | 29949/48008 [1:42:41<58:02, 5.19it/s] 62%|██████▏ | 29950/48008 [1:42:42<57:31, 5.23it/s] 62%|██████▏ | 29950/48008 [1:42:42<57:31, 5.23it/s]{'loss': 0.3012, 'grad_norm': 0.3671875, 'learning_rate': 7.523329445092485e-05, 'epoch': 0.62} 62%|██████▏ | 29951/48008 [1:42:42<57:32, 5.23it/s] 62%|██████▏ | 29952/48008 [1:42:42<1:15:09, 4.00it/s] 62%|██████▏ | 29953/48008 [1:42:42<1:10:58, 4.24it/s] 62%|██████▏ | 29954/48008 [1:42:43<1:11:13, 4.22it/s] 62%|██████▏ | 29955/48008 [1:42:43<1:07:10, 4.48it/s] 62%|██████▏ | 29956/48008 [1:42:43<1:04:23, 4.67it/s] 62%|██████▏ | 29957/48008 [1:42:43<1:02:13, 4.83it/s] 62%|██████▏ | 29958/48008 [1:42:43<1:02:04, 4.85it/s] 62%|██████▏ | 29959/48008 [1:42:43<1:00:46, 4.95it/s] 62%|██████▏ | 29960/48008 [1:42:44<1:00:32, 4.97it/s] 62%|██████▏ | 29961/48008 [1:42:44<59:32, 5.05it/s] 62%|██████▏ | 29962/48008 [1:42:44<1:00:00, 5.01it/s] 62%|██████▏ | 29963/48008 [1:42:44<1:09:42, 4.31it/s] 62%|██████▏ | 29964/48008 [1:42:45<1:16:27, 3.93it/s] 62%|██████▏ | 29965/48008 [1:42:45<1:10:48, 4.25it/s] 62%|██████▏ | 29966/48008 [1:42:45<1:06:50, 4.50it/s] 62%|██████▏ | 29967/48008 [1:42:45<1:03:55, 4.70it/s] 62%|██████▏ | 29968/48008 [1:42:45<1:01:50, 4.86it/s] 62%|██████▏ | 29969/48008 [1:42:46<1:01:02, 4.93it/s] 62%|██████▏ | 29970/48008 [1:42:46<1:00:04, 5.00it/s] 62%|██████▏ | 29971/48008 [1:42:46<59:43, 5.03it/s] 62%|██████▏ | 29972/48008 [1:42:46<58:43, 5.12it/s] 62%|██████▏ | 29973/48008 [1:42:46<59:15, 5.07it/s] 62%|██████▏ | 29974/48008 [1:42:47<58:39, 5.12it/s] 62%|██████▏ | 29975/48008 [1:42:47<57:45, 5.20it/s] 62%|██████▏ | 29976/48008 [1:42:47<58:48, 5.11it/s] 62%|██████▏ | 29977/48008 [1:42:47<58:20, 5.15it/s] 62%|██████▏ | 29978/48008 [1:42:47<58:03, 5.18it/s] 62%|██████▏ | 29979/48008 [1:42:48<58:29, 5.14it/s] 62%|██████▏ | 29980/48008 [1:42:48<58:10, 5.16it/s] 62%|██████▏ | 29981/48008 [1:42:48<59:10, 5.08it/s] 62%|██████▏ | 29982/48008 [1:42:48<58:23, 5.15it/s] 62%|██████▏ | 29983/48008 [1:42:48<59:18, 5.07it/s] 62%|██████▏ | 29984/48008 [1:42:49<58:39, 5.12it/s] 62%|██████▏ | 29985/48008 [1:42:49<58:48, 5.11it/s] 62%|██████▏ | 29986/48008 [1:42:49<59:06, 5.08it/s] 62%|██████▏ | 29987/48008 [1:42:49<58:34, 5.13it/s] 62%|██████▏ | 29988/48008 [1:42:49<58:45, 5.11it/s] 62%|██████▏ | 29989/48008 [1:42:50<58:26, 5.14it/s] 62%|██████▏ | 29990/48008 [1:42:50<58:13, 5.16it/s] 62%|██████▏ | 29991/48008 [1:42:50<57:29, 5.22it/s] 62%|██████▏ | 29992/48008 [1:42:50<58:28, 5.13it/s] 62%|██████▏ | 29993/48008 [1:42:50<1:02:19, 4.82it/s] 62%|██████▏ | 29994/48008 [1:42:51<1:00:52, 4.93it/s] 62%|██████▏ | 29995/48008 [1:42:51<59:47, 5.02it/s] 62%|██████▏ | 29996/48008 [1:42:51<1:00:08, 4.99it/s] 62%|██████▏ | 29997/48008 [1:42:51<59:19, 5.06it/s] 62%|██████▏ | 29998/48008 [1:42:51<58:44, 5.11it/s] 62%|██████▏ | 29999/48008 [1:42:52<58:30, 5.13it/s] 62%|██████▏ | 30000/48008 [1:42:52<59:18, 5.06it/s] {'loss': 0.3016, 'grad_norm': 0.177734375, 'learning_rate': 7.502499583402767e-05, 'epoch': 0.62} 62%|██████▏ | 30000/48008 [1:42:52<59:18, 5.06it/s] 62%|██████▏ | 30001/48008 [1:42:52<1:09:19, 4.33it/s] 62%|██████▏ | 30002/48008 [1:42:52<1:06:52, 4.49it/s] 62%|██████▏ | 30003/48008 [1:42:52<1:04:06, 4.68it/s] 62%|██████▏ | 30004/48008 [1:42:53<1:02:16, 4.82it/s] 62%|██████▎ | 30005/48008 [1:42:53<1:01:00, 4.92it/s] 63%|██████▎ | 30006/48008 [1:42:53<59:33, 5.04it/s] 63%|██████▎ | 30007/48008 [1:42:53<59:01, 5.08it/s] 63%|██████▎ | 30008/48008 [1:42:53<58:02, 5.17it/s] 63%|██████▎ | 30009/48008 [1:42:54<58:18, 5.14it/s] 63%|██████▎ | 30010/48008 [1:42:54<1:02:14, 4.82it/s] 63%|██████▎ | 30011/48008 [1:42:54<1:00:52, 4.93it/s] 63%|██████▎ | 30012/48008 [1:42:54<1:00:26, 4.96it/s] 63%|██████▎ | 30013/48008 [1:42:54<59:42, 5.02it/s] 63%|██████▎ | 30014/48008 [1:42:55<58:33, 5.12it/s] 63%|██████▎ | 30015/48008 [1:42:55<58:10, 5.15it/s] 63%|██████▎ | 30016/48008 [1:42:55<58:01, 5.17it/s] 63%|██████▎ | 30017/48008 [1:42:55<58:31, 5.12it/s] 63%|██████▎ | 30018/48008 [1:42:55<58:07, 5.16it/s] 63%|██████▎ | 30019/48008 [1:42:56<58:57, 5.09it/s] 63%|██████▎ | 30020/48008 [1:42:56<58:29, 5.13it/s] 63%|██████▎ | 30021/48008 [1:42:56<58:12, 5.15it/s] 63%|██████▎ | 30022/48008 [1:42:56<1:08:15, 4.39it/s] 63%|██████▎ | 30023/48008 [1:42:56<1:04:30, 4.65it/s] 63%|██████▎ | 30024/48008 [1:42:57<1:03:52, 4.69it/s] 63%|██████▎ | 30025/48008 [1:42:57<1:01:29, 4.87it/s] 63%|██████▎ | 30026/48008 [1:42:57<1:00:13, 4.98it/s] 63%|██████▎ | 30027/48008 [1:42:57<59:16, 5.06it/s] 63%|██████▎ | 30028/48008 [1:42:57<58:51, 5.09it/s] 63%|██████▎ | 30029/48008 [1:42:58<58:24, 5.13it/s] 63%|██████▎ | 30030/48008 [1:42:58<1:08:32, 4.37it/s] 63%|██████▎ | 30031/48008 [1:42:58<1:06:03, 4.54it/s] 63%|██████▎ | 30032/48008 [1:42:58<1:05:00, 4.61it/s] 63%|██████▎ | 30033/48008 [1:42:59<1:02:17, 4.81it/s] 63%|██████▎ | 30034/48008 [1:42:59<1:02:33, 4.79it/s] 63%|██████▎ | 30035/48008 [1:42:59<1:01:02, 4.91it/s] 63%|██████▎ | 30036/48008 [1:42:59<1:00:04, 4.99it/s] 63%|██████▎ | 30037/48008 [1:42:59<59:32, 5.03it/s] 63%|██████▎ | 30038/48008 [1:43:00<59:15, 5.05it/s] 63%|██████▎ | 30039/48008 [1:43:00<59:08, 5.06it/s] 63%|██████▎ | 30040/48008 [1:43:00<1:16:02, 3.94it/s] 63%|██████▎ | 30041/48008 [1:43:00<1:10:58, 4.22it/s] 63%|██████▎ | 30042/48008 [1:43:00<1:06:23, 4.51it/s] 63%|██████▎ | 30043/48008 [1:43:01<1:03:43, 4.70it/s] 63%|██████▎ | 30044/48008 [1:43:01<1:03:10, 4.74it/s] 63%|██████▎ | 30045/48008 [1:43:01<1:02:54, 4.76it/s] 63%|██████▎ | 30046/48008 [1:43:01<1:01:47, 4.85it/s] 63%|██████▎ | 30047/48008 [1:43:01<1:00:21, 4.96it/s] 63%|██████▎ | 30048/48008 [1:43:02<1:09:43, 4.29it/s] 63%|██████▎ | 30049/48008 [1:43:02<1:05:59, 4.54it/s] 63%|██████▎ | 30050/48008 [1:43:02<1:03:06, 4.74it/s] {'loss': 0.2981, 'grad_norm': 0.1650390625, 'learning_rate': 7.481669721713049e-05, 'epoch': 0.63} 63%|██████▎ | 30050/48008 [1:43:02<1:03:06, 4.74it/s] 63%|██████▎ | 30051/48008 [1:43:02<1:02:34, 4.78it/s] 63%|██████▎ | 30052/48008 [1:43:03<1:00:32, 4.94it/s] 63%|██████▎ | 30053/48008 [1:43:03<1:00:34, 4.94it/s] 63%|██████▎ | 30054/48008 [1:43:03<59:48, 5.00it/s] 63%|██████▎ | 30055/48008 [1:43:03<58:59, 5.07it/s] 63%|██████▎ | 30056/48008 [1:43:03<58:22, 5.12it/s] 63%|██████▎ | 30057/48008 [1:43:04<58:25, 5.12it/s] 63%|██████▎ | 30058/48008 [1:43:04<1:02:07, 4.82it/s] 63%|██████▎ | 30059/48008 [1:43:04<1:01:00, 4.90it/s] 63%|██████▎ | 30060/48008 [1:43:04<59:55, 4.99it/s] 63%|██████▎ | 30061/48008 [1:43:04<58:41, 5.10it/s] 63%|██████▎ | 30062/48008 [1:43:05<58:03, 5.15it/s] 63%|██████▎ | 30063/48008 [1:43:05<57:48, 5.17it/s] 63%|██████▎ | 30064/48008 [1:43:05<57:44, 5.18it/s] 63%|██████▎ | 30065/48008 [1:43:05<58:20, 5.13it/s] 63%|██████▎ | 30066/48008 [1:43:05<59:16, 5.05it/s] 63%|██████▎ | 30067/48008 [1:43:05<58:29, 5.11it/s] 63%|██████▎ | 30068/48008 [1:43:06<58:10, 5.14it/s] 63%|██████▎ | 30069/48008 [1:43:06<57:43, 5.18it/s] 63%|██████▎ | 30070/48008 [1:43:06<57:36, 5.19it/s] 63%|██████▎ | 30071/48008 [1:43:06<58:02, 5.15it/s] 63%|██████▎ | 30072/48008 [1:43:06<58:43, 5.09it/s] 63%|██████▎ | 30073/48008 [1:43:07<58:16, 5.13it/s] 63%|██████▎ | 30074/48008 [1:43:07<59:03, 5.06it/s] 63%|██████▎ | 30075/48008 [1:43:07<59:07, 5.05it/s] 63%|██████▎ | 30076/48008 [1:43:07<58:24, 5.12it/s] 63%|██████▎ | 30077/48008 [1:43:07<58:41, 5.09it/s] 63%|██████▎ | 30078/48008 [1:43:08<59:11, 5.05it/s] 63%|██████▎ | 30079/48008 [1:43:08<58:17, 5.13it/s] 63%|██████▎ | 30080/48008 [1:43:08<57:23, 5.21it/s] 63%|██████▎ | 30081/48008 [1:43:08<57:04, 5.23it/s] 63%|██████▎ | 30082/48008 [1:43:08<56:44, 5.26it/s] 63%|██████▎ | 30083/48008 [1:43:09<56:58, 5.24it/s] 63%|██████▎ | 30084/48008 [1:43:09<56:58, 5.24it/s] 63%|██████▎ | 30085/48008 [1:43:09<56:29, 5.29it/s] 63%|██████▎ | 30086/48008 [1:43:09<57:39, 5.18it/s] 63%|██████▎ | 30087/48008 [1:43:09<1:01:39, 4.84it/s] 63%|██████▎ | 30088/48008 [1:43:10<1:00:58, 4.90it/s] 63%|██████▎ | 30089/48008 [1:43:10<1:01:37, 4.85it/s] 63%|██████▎ | 30090/48008 [1:43:10<1:00:08, 4.97it/s] 63%|██████▎ | 30091/48008 [1:43:10<59:55, 4.98it/s] 63%|██████▎ | 30092/48008 [1:43:11<1:16:34, 3.90it/s] 63%|██████▎ | 30093/48008 [1:43:11<1:28:22, 3.38it/s] 63%|██████▎ | 30094/48008 [1:43:11<1:19:37, 3.75it/s] 63%|██████▎ | 30095/48008 [1:43:11<1:13:15, 4.08it/s] 63%|██████▎ | 30096/48008 [1:43:12<1:12:34, 4.11it/s] 63%|██████▎ | 30097/48008 [1:43:12<1:08:12, 4.38it/s] 63%|██████▎ | 30098/48008 [1:43:12<1:22:29, 3.62it/s] 63%|██████▎ | 30099/48008 [1:43:12<1:14:10, 4.02it/s] 63%|██████▎ | 30100/48008 [1:43:13<1:08:55, 4.33it/s] {'loss': 0.2984, 'grad_norm': 0.255859375, 'learning_rate': 7.460839860023329e-05, 'epoch': 0.63} 63%|██████▎ | 30100/48008 [1:43:13<1:08:55, 4.33it/s] 63%|██████▎ | 30101/48008 [1:43:13<1:05:29, 4.56it/s] 63%|██████▎ | 30102/48008 [1:43:13<1:02:40, 4.76it/s] 63%|██████▎ | 30103/48008 [1:43:13<1:00:46, 4.91it/s] 63%|██████▎ | 30104/48008 [1:43:13<1:01:19, 4.87it/s] 63%|██████▎ | 30105/48008 [1:43:14<59:44, 4.99it/s] 63%|██████▎ | 30106/48008 [1:43:14<59:03, 5.05it/s] 63%|██████▎ | 30107/48008 [1:43:14<58:53, 5.07it/s] 63%|██████▎ | 30108/48008 [1:43:14<58:15, 5.12it/s] 63%|██████▎ | 30109/48008 [1:43:14<57:14, 5.21it/s] 63%|██████▎ | 30110/48008 [1:43:14<57:13, 5.21it/s] 63%|██████▎ | 30111/48008 [1:43:15<57:36, 5.18it/s] 63%|██████▎ | 30112/48008 [1:43:15<57:46, 5.16it/s] 63%|██████▎ | 30113/48008 [1:43:15<57:32, 5.18it/s] 63%|██████▎ | 30114/48008 [1:43:15<57:24, 5.19it/s] 63%|██████▎ | 30115/48008 [1:43:15<57:12, 5.21it/s] 63%|██████▎ | 30116/48008 [1:43:16<57:08, 5.22it/s] 63%|██████▎ | 30117/48008 [1:43:16<1:14:37, 4.00it/s] 63%|██████▎ | 30118/48008 [1:43:16<1:10:02, 4.26it/s] 63%|██████▎ | 30119/48008 [1:43:16<1:05:38, 4.54it/s] 63%|██████▎ | 30120/48008 [1:43:17<1:03:29, 4.70it/s] 63%|██████▎ | 30121/48008 [1:43:17<1:02:03, 4.80it/s] 63%|██████▎ | 30122/48008 [1:43:17<1:04:39, 4.61it/s] 63%|██████▎ | 30123/48008 [1:43:17<1:03:05, 4.72it/s] 63%|██████▎ | 30124/48008 [1:43:17<1:01:39, 4.83it/s] 63%|██████▎ | 30125/48008 [1:43:18<1:10:35, 4.22it/s] 63%|██████▎ | 30126/48008 [1:43:18<1:10:33, 4.22it/s] 63%|██████▎ | 30127/48008 [1:43:18<1:06:05, 4.51it/s] 63%|██████▎ | 30128/48008 [1:43:18<1:03:28, 4.70it/s] 63%|██████▎ | 30129/48008 [1:43:19<1:00:59, 4.89it/s] 63%|██████▎ | 30130/48008 [1:43:19<59:38, 5.00it/s] 63%|██████▎ | 30131/48008 [1:43:19<58:41, 5.08it/s] 63%|██████▎ | 30132/48008 [1:43:19<59:00, 5.05it/s] 63%|██████▎ | 30133/48008 [1:43:19<1:00:01, 4.96it/s] 63%|██████▎ | 30134/48008 [1:43:20<59:47, 4.98it/s] 63%|██████▎ | 30135/48008 [1:43:20<58:59, 5.05it/s] 63%|██████▎ | 30136/48008 [1:43:20<58:12, 5.12it/s] 63%|██████▎ | 30137/48008 [1:43:20<57:30, 5.18it/s] 63%|██████▎ | 30138/48008 [1:43:20<58:59, 5.05it/s] 63%|██████▎ | 30139/48008 [1:43:21<57:45, 5.16it/s] 63%|██████▎ | 30140/48008 [1:43:21<57:08, 5.21it/s] 63%|██████▎ | 30141/48008 [1:43:21<56:40, 5.25it/s] 63%|██████▎ | 30142/48008 [1:43:21<1:00:53, 4.89it/s] 63%|██████▎ | 30143/48008 [1:43:21<1:00:38, 4.91it/s] 63%|██████▎ | 30144/48008 [1:43:22<59:36, 4.99it/s] 63%|██████▎ | 30145/48008 [1:43:22<58:51, 5.06it/s] 63%|██████▎ | 30146/48008 [1:43:22<58:18, 5.11it/s] 63%|██████▎ | 30147/48008 [1:43:22<58:00, 5.13it/s] 63%|██████▎ | 30148/48008 [1:43:22<59:18, 5.02it/s] 63%|██████▎ | 30149/48008 [1:43:22<59:13, 5.03it/s] 63%|██████▎ | 30150/48008 [1:43:23<57:52, 5.14it/s] {'loss': 0.3, 'grad_norm': 0.447265625, 'learning_rate': 7.440009998333611e-05, 'epoch': 0.63} 63%|██████▎ | 30150/48008 [1:43:23<57:52, 5.14it/s] 63%|██████▎ | 30151/48008 [1:43:23<58:24, 5.10it/s] 63%|██████▎ | 30152/48008 [1:43:23<57:09, 5.21it/s] 63%|██████▎ | 30153/48008 [1:43:23<57:12, 5.20it/s] 63%|██████▎ | 30154/48008 [1:43:23<57:32, 5.17it/s] 63%|██████▎ | 30155/48008 [1:43:24<57:16, 5.20it/s] 63%|██████▎ | 30156/48008 [1:43:24<57:44, 5.15it/s] 63%|██████▎ | 30157/48008 [1:43:24<57:43, 5.15it/s] 63%|██████▎ | 30158/48008 [1:43:24<57:18, 5.19it/s] 63%|██████▎ | 30159/48008 [1:43:24<57:38, 5.16it/s] 63%|██████▎ | 30160/48008 [1:43:25<57:23, 5.18it/s] 63%|██████▎ | 30161/48008 [1:43:25<57:10, 5.20it/s] 63%|██████▎ | 30162/48008 [1:43:25<56:59, 5.22it/s] 63%|██████▎ | 30163/48008 [1:43:25<58:37, 5.07it/s] 63%|██████▎ | 30164/48008 [1:43:25<57:55, 5.13it/s] 63%|██████▎ | 30165/48008 [1:43:26<57:47, 5.15it/s] 63%|██████▎ | 30166/48008 [1:43:26<1:01:44, 4.82it/s] 63%|██████▎ | 30167/48008 [1:43:26<1:01:10, 4.86it/s] 63%|██████▎ | 30168/48008 [1:43:26<59:52, 4.97it/s] 63%|██████▎ | 30169/48008 [1:43:27<1:09:14, 4.29it/s] 63%|██████▎ | 30170/48008 [1:43:27<1:05:58, 4.51it/s] 63%|██████▎ | 30171/48008 [1:43:27<1:03:21, 4.69it/s] 63%|██████▎ | 30172/48008 [1:43:27<1:01:27, 4.84it/s] 63%|██████▎ | 30173/48008 [1:43:27<59:44, 4.98it/s] 63%|██████▎ | 30174/48008 [1:43:27<58:31, 5.08it/s] 63%|██████▎ | 30175/48008 [1:43:28<59:53, 4.96it/s] 63%|██████▎ | 30176/48008 [1:43:28<1:00:11, 4.94it/s] 63%|██████▎ | 30177/48008 [1:43:28<59:47, 4.97it/s] 63%|██████▎ | 30178/48008 [1:43:28<58:45, 5.06it/s] 63%|██████▎ | 30179/48008 [1:43:29<1:02:19, 4.77it/s] 63%|██████▎ | 30180/48008 [1:43:29<1:01:22, 4.84it/s] 63%|██████▎ | 30181/48008 [1:43:29<1:01:04, 4.87it/s] 63%|██████▎ | 30182/48008 [1:43:29<59:36, 4.98it/s] 63%|██████▎ | 30183/48008 [1:43:29<59:49, 4.97it/s] 63%|██████▎ | 30184/48008 [1:43:30<59:00, 5.03it/s] 63%|██████▎ | 30185/48008 [1:43:30<57:49, 5.14it/s] 63%|██████▎ | 30186/48008 [1:43:30<57:18, 5.18it/s] 63%|██████▎ | 30187/48008 [1:43:30<57:02, 5.21it/s] 63%|██████▎ | 30188/48008 [1:43:30<56:22, 5.27it/s] 63%|██████▎ | 30189/48008 [1:43:30<56:09, 5.29it/s] 63%|██████▎ | 30190/48008 [1:43:31<56:14, 5.28it/s] 63%|██████▎ | 30191/48008 [1:43:31<56:29, 5.26it/s] 63%|██████▎ | 30192/48008 [1:43:31<56:13, 5.28it/s] 63%|██████▎ | 30193/48008 [1:43:31<55:55, 5.31it/s] 63%|██████▎ | 30194/48008 [1:43:31<55:33, 5.34it/s] 63%|██████▎ | 30195/48008 [1:43:32<56:24, 5.26it/s] 63%|██████▎ | 30196/48008 [1:43:32<56:23, 5.26it/s] 63%|██████▎ | 30197/48008 [1:43:32<57:30, 5.16it/s] 63%|██████▎ | 30198/48008 [1:43:32<57:10, 5.19it/s] 63%|██████▎ | 30199/48008 [1:43:32<56:50, 5.22it/s] 63%|██████▎ | 30200/48008 [1:43:33<57:08, 5.19it/s] {'loss': 0.3005, 'grad_norm': 0.310546875, 'learning_rate': 7.419180136643893e-05, 'epoch': 0.63} 63%|██████▎ | 30200/48008 [1:43:33<57:08, 5.19it/s] 63%|██████▎ | 30201/48008 [1:43:33<57:14, 5.19it/s] 63%|██████▎ | 30202/48008 [1:43:33<56:35, 5.24it/s] 63%|██████▎ | 30203/48008 [1:43:33<56:10, 5.28it/s] 63%|██████▎ | 30204/48008 [1:43:33<56:32, 5.25it/s] 63%|██████▎ | 30205/48008 [1:43:33<56:37, 5.24it/s] 63%|██████▎ | 30206/48008 [1:43:34<56:26, 5.26it/s] 63%|██████▎ | 30207/48008 [1:43:34<56:10, 5.28it/s] 63%|██████▎ | 30208/48008 [1:43:34<56:00, 5.30it/s] 63%|██████▎ | 30209/48008 [1:43:34<1:00:22, 4.91it/s] 63%|██████▎ | 30210/48008 [1:43:34<59:30, 4.98it/s] 63%|██████▎ | 30211/48008 [1:43:35<58:38, 5.06it/s] 63%|██████▎ | 30212/48008 [1:43:35<57:55, 5.12it/s] 63%|██████▎ | 30213/48008 [1:43:35<58:27, 5.07it/s] 63%|██████▎ | 30214/48008 [1:43:35<58:12, 5.10it/s] 63%|██████▎ | 30215/48008 [1:43:35<59:41, 4.97it/s] 63%|██████▎ | 30216/48008 [1:43:36<58:56, 5.03it/s] 63%|██████▎ | 30217/48008 [1:43:36<58:07, 5.10it/s] 63%|██████▎ | 30218/48008 [1:43:36<57:06, 5.19it/s] 63%|██████▎ | 30219/48008 [1:43:36<57:10, 5.19it/s] 63%|██████▎ | 30220/48008 [1:43:36<58:33, 5.06it/s] 63%|██████▎ | 30221/48008 [1:43:37<59:02, 5.02it/s] 63%|██████▎ | 30222/48008 [1:43:37<58:35, 5.06it/s] 63%|██████▎ | 30223/48008 [1:43:37<59:36, 4.97it/s] 63%|██████▎ | 30224/48008 [1:43:37<59:03, 5.02it/s] 63%|██████▎ | 30225/48008 [1:43:37<58:30, 5.07it/s] 63%|██████▎ | 30226/48008 [1:43:38<57:19, 5.17it/s] 63%|██████▎ | 30227/48008 [1:43:38<58:18, 5.08it/s] 63%|██████▎ | 30228/48008 [1:43:38<58:09, 5.10it/s] 63%|██████▎ | 30229/48008 [1:43:38<57:53, 5.12it/s] 63%|██████▎ | 30230/48008 [1:43:38<57:56, 5.11it/s] 63%|██████▎ | 30231/48008 [1:43:39<57:19, 5.17it/s] 63%|██████▎ | 30232/48008 [1:43:39<57:08, 5.19it/s] 63%|██████▎ | 30233/48008 [1:43:39<56:50, 5.21it/s] 63%|██████▎ | 30234/48008 [1:43:39<56:42, 5.22it/s] 63%|██████▎ | 30235/48008 [1:43:39<56:32, 5.24it/s] 63%|██████▎ | 30236/48008 [1:43:40<56:28, 5.24it/s] 63%|██████▎ | 30237/48008 [1:43:40<56:14, 5.27it/s] 63%|██████▎ | 30238/48008 [1:43:40<56:30, 5.24it/s] 63%|██████▎ | 30239/48008 [1:43:40<56:22, 5.25it/s] 63%|██████▎ | 30240/48008 [1:43:40<56:58, 5.20it/s] 63%|██████▎ | 30241/48008 [1:43:41<56:54, 5.20it/s] 63%|██████▎ | 30242/48008 [1:43:41<56:36, 5.23it/s] 63%|██████▎ | 30243/48008 [1:43:41<57:11, 5.18it/s] 63%|██████▎ | 30244/48008 [1:43:41<58:44, 5.04it/s] 63%|██████▎ | 30245/48008 [1:43:41<57:56, 5.11it/s] 63%|██████▎ | 30246/48008 [1:43:41<57:24, 5.16it/s] 63%|██████▎ | 30247/48008 [1:43:42<57:12, 5.17it/s] 63%|██████▎ | 30248/48008 [1:43:42<57:02, 5.19it/s] 63%|██████▎ | 30249/48008 [1:43:42<55:53, 5.30it/s] 63%|██████▎ | 30250/48008 [1:43:42<56:11, 5.27it/s] {'loss': 0.303, 'grad_norm': 0.2138671875, 'learning_rate': 7.398350274954174e-05, 'epoch': 0.63} 63%|██████▎ | 30250/48008 [1:43:42<56:11, 5.27it/s] 63%|██████▎ | 30251/48008 [1:43:42<57:52, 5.11it/s] 63%|██████▎ | 30252/48008 [1:43:43<56:56, 5.20it/s] 63%|██████▎ | 30253/48008 [1:43:43<56:37, 5.23it/s] 63%|██████▎ | 30254/48008 [1:43:43<56:14, 5.26it/s] 63%|██████▎ | 30255/48008 [1:43:43<57:25, 5.15it/s] 63%|██████▎ | 30256/48008 [1:43:43<56:46, 5.21it/s] 63%|██████▎ | 30257/48008 [1:43:44<56:22, 5.25it/s] 63%|██████▎ | 30258/48008 [1:43:44<56:30, 5.24it/s] 63%|██████▎ | 30259/48008 [1:43:44<56:34, 5.23it/s] 63%|██████▎ | 30260/48008 [1:43:44<56:29, 5.24it/s] 63%|██████▎ | 30261/48008 [1:43:44<56:46, 5.21it/s] 63%|██████▎ | 30262/48008 [1:43:45<56:20, 5.25it/s] 63%|██████▎ | 30263/48008 [1:43:45<57:55, 5.11it/s] 63%|██████▎ | 30264/48008 [1:43:45<57:41, 5.13it/s] 63%|██████▎ | 30265/48008 [1:43:45<57:14, 5.17it/s] 63%|██████▎ | 30266/48008 [1:43:45<57:35, 5.13it/s] 63%|██████▎ | 30267/48008 [1:43:46<56:59, 5.19it/s] 63%|██████▎ | 30268/48008 [1:43:46<56:54, 5.19it/s] 63%|██████▎ | 30269/48008 [1:43:46<56:26, 5.24it/s] 63%|██████▎ | 30270/48008 [1:43:46<56:23, 5.24it/s] 63%|██████▎ | 30271/48008 [1:43:46<1:13:43, 4.01it/s] 63%|██████▎ | 30272/48008 [1:43:47<1:08:09, 4.34it/s] 63%|██████▎ | 30273/48008 [1:43:47<1:05:06, 4.54it/s] 63%|██████▎ | 30274/48008 [1:43:47<1:02:29, 4.73it/s] 63%|██████▎ | 30275/48008 [1:43:47<1:04:50, 4.56it/s] 63%|██████▎ | 30276/48008 [1:43:47<1:03:51, 4.63it/s] 63%|██████▎ | 30277/48008 [1:43:48<1:02:53, 4.70it/s] 63%|██████▎ | 30278/48008 [1:43:48<1:01:01, 4.84it/s] 63%|██████▎ | 30279/48008 [1:43:48<1:00:03, 4.92it/s] 63%|██████▎ | 30280/48008 [1:43:48<59:02, 5.00it/s] 63%|██████▎ | 30281/48008 [1:43:48<57:55, 5.10it/s] 63%|██████▎ | 30282/48008 [1:43:49<57:14, 5.16it/s] 63%|██████▎ | 30283/48008 [1:43:49<57:06, 5.17it/s] 63%|██████▎ | 30284/48008 [1:43:49<56:54, 5.19it/s] 63%|██████▎ | 30285/48008 [1:43:49<57:14, 5.16it/s] 63%|██████▎ | 30286/48008 [1:43:49<57:57, 5.10it/s] 63%|██████▎ | 30287/48008 [1:43:50<57:53, 5.10it/s] 63%|██████▎ | 30288/48008 [1:43:50<58:05, 5.08it/s] 63%|██████▎ | 30289/48008 [1:43:50<57:36, 5.13it/s] 63%|██████▎ | 30290/48008 [1:43:50<56:46, 5.20it/s] 63%|██████▎ | 30291/48008 [1:43:50<57:16, 5.16it/s] 63%|██████▎ | 30292/48008 [1:43:51<57:37, 5.12it/s] 63%|██████▎ | 30293/48008 [1:43:51<57:55, 5.10it/s] 63%|██████▎ | 30294/48008 [1:43:51<57:23, 5.14it/s] 63%|██████▎ | 30295/48008 [1:43:51<57:12, 5.16it/s] 63%|██████▎ | 30296/48008 [1:43:51<57:08, 5.17it/s] 63%|██████▎ | 30297/48008 [1:43:52<56:47, 5.20it/s] 63%|██████▎ | 30298/48008 [1:43:52<56:39, 5.21it/s] 63%|██████▎ | 30299/48008 [1:43:52<56:38, 5.21it/s] 63%|██████▎ | 30300/48008 [1:43:52<57:06, 5.17it/s] 63%|██████▎ | 30300/48008 [1:43:52<57:06, 5.17it/s]{'loss': 0.2986, 'grad_norm': 0.314453125, 'learning_rate': 7.377520413264456e-05, 'epoch': 0.63} 63%|██████▎ | 30301/48008 [1:43:52<56:51, 5.19it/s] 63%|██████▎ | 30302/48008 [1:43:53<56:25, 5.23it/s] 63%|██████▎ | 30303/48008 [1:43:53<57:23, 5.14it/s] 63%|██████▎ | 30304/48008 [1:43:53<58:11, 5.07it/s] 63%|██████▎ | 30305/48008 [1:43:53<57:33, 5.13it/s] 63%|██████▎ | 30306/48008 [1:43:53<58:53, 5.01it/s] 63%|██████▎ | 30307/48008 [1:43:54<59:19, 4.97it/s] 63%|██████▎ | 30308/48008 [1:43:54<59:12, 4.98it/s] 63%|██████▎ | 30309/48008 [1:43:54<58:20, 5.06it/s] 63%|██████▎ | 30310/48008 [1:43:54<57:36, 5.12it/s] 63%|██████▎ | 30311/48008 [1:43:54<56:58, 5.18it/s] 63%|██████▎ | 30312/48008 [1:43:55<1:06:58, 4.40it/s] 63%|██████▎ | 30313/48008 [1:43:55<1:04:54, 4.54it/s] 63%|██████▎ | 30314/48008 [1:43:55<1:03:03, 4.68it/s] 63%|██████▎ | 30315/48008 [1:43:55<1:00:53, 4.84it/s] 63%|██████▎ | 30316/48008 [1:43:55<1:01:21, 4.81it/s] 63%|██████▎ | 30317/48008 [1:43:56<1:00:36, 4.86it/s] 63%|██████▎ | 30318/48008 [1:43:56<59:09, 4.98it/s] 63%|██████▎ | 30319/48008 [1:43:56<57:58, 5.09it/s] 63%|██████▎ | 30320/48008 [1:43:56<58:22, 5.05it/s] 63%|██████▎ | 30321/48008 [1:43:56<58:53, 5.01it/s] 63%|██████▎ | 30322/48008 [1:43:57<58:11, 5.07it/s] 63%|██████▎ | 30323/48008 [1:43:57<58:45, 5.02it/s] 63%|██████▎ | 30324/48008 [1:43:57<58:20, 5.05it/s] 63%|██████▎ | 30325/48008 [1:43:57<57:43, 5.10it/s] 63%|██████▎ | 30326/48008 [1:43:57<1:01:22, 4.80it/s] 63%|██████▎ | 30327/48008 [1:43:58<59:24, 4.96it/s] 63%|██████▎ | 30328/48008 [1:43:58<58:21, 5.05it/s] 63%|██████▎ | 30329/48008 [1:43:58<59:19, 4.97it/s] 63%|██████▎ | 30330/48008 [1:43:58<58:24, 5.05it/s] 63%|██████▎ | 30331/48008 [1:43:58<57:52, 5.09it/s] 63%|██████▎ | 30332/48008 [1:43:59<56:59, 5.17it/s] 63%|██████▎ | 30333/48008 [1:43:59<57:54, 5.09it/s] 63%|██████▎ | 30334/48008 [1:43:59<57:10, 5.15it/s] 63%|██████▎ | 30335/48008 [1:43:59<57:47, 5.10it/s] 63%|██████▎ | 30336/48008 [1:43:59<57:26, 5.13it/s] 63%|██████▎ | 30337/48008 [1:44:00<58:06, 5.07it/s] 63%|██████▎ | 30338/48008 [1:44:00<57:34, 5.11it/s] 63%|██████▎ | 30339/48008 [1:44:00<58:15, 5.05it/s] 63%|██████▎ | 30340/48008 [1:44:00<57:52, 5.09it/s] 63%|██████▎ | 30341/48008 [1:44:00<57:16, 5.14it/s] 63%|██████▎ | 30342/48008 [1:44:01<57:00, 5.17it/s] 63%|██████▎ | 30343/48008 [1:44:01<58:28, 5.03it/s] 63%|██████▎ | 30344/48008 [1:44:01<57:41, 5.10it/s] 63%|██████▎ | 30345/48008 [1:44:01<57:13, 5.14it/s] 63%|██████▎ | 30346/48008 [1:44:01<56:56, 5.17it/s] 63%|██████▎ | 30347/48008 [1:44:02<56:44, 5.19it/s] 63%|██████▎ | 30348/48008 [1:44:02<56:51, 5.18it/s] 63%|██████▎ | 30349/48008 [1:44:02<56:20, 5.22it/s] 63%|██████▎ | 30350/48008 [1:44:02<56:11, 5.24it/s] {'loss': 0.2998, 'grad_norm': 0.1328125, 'learning_rate': 7.356690551574738e-05, 'epoch': 0.63} 63%|██████▎ | 30350/48008 [1:44:02<56:11, 5.24it/s] 63%|██████▎ | 30351/48008 [1:44:02<57:21, 5.13it/s] 63%|██████▎ | 30352/48008 [1:44:02<56:40, 5.19it/s] 63%|██████▎ | 30353/48008 [1:44:03<56:09, 5.24it/s] 63%|██████▎ | 30354/48008 [1:44:03<56:18, 5.23it/s] 63%|██████▎ | 30355/48008 [1:44:03<1:00:34, 4.86it/s] 63%|██████▎ | 30356/48008 [1:44:03<59:15, 4.97it/s] 63%|██████▎ | 30357/48008 [1:44:03<58:28, 5.03it/s] 63%|██████▎ | 30358/48008 [1:44:04<58:22, 5.04it/s] 63%|██████▎ | 30359/48008 [1:44:04<58:58, 4.99it/s] 63%|██████▎ | 30360/48008 [1:44:04<58:22, 5.04it/s] 63%|██████▎ | 30361/48008 [1:44:04<57:35, 5.11it/s] 63%|██████▎ | 30362/48008 [1:44:04<57:03, 5.15it/s] 63%|██████▎ | 30363/48008 [1:44:05<58:02, 5.07it/s] 63%|██████▎ | 30364/48008 [1:44:05<58:17, 5.05it/s] 63%|██████▎ | 30365/48008 [1:44:05<57:32, 5.11it/s] 63%|██████▎ | 30366/48008 [1:44:05<57:34, 5.11it/s] 63%|██████▎ | 30367/48008 [1:44:05<58:13, 5.05it/s] 63%|██████▎ | 30368/48008 [1:44:06<57:24, 5.12it/s] 63%|██████▎ | 30369/48008 [1:44:06<57:08, 5.15it/s] 63%|██████▎ | 30370/48008 [1:44:06<56:34, 5.20it/s] 63%|██████▎ | 30371/48008 [1:44:06<1:00:32, 4.85it/s] 63%|██████▎ | 30372/48008 [1:44:06<58:58, 4.98it/s] 63%|██████▎ | 30373/48008 [1:44:07<58:30, 5.02it/s] 63%|██████▎ | 30374/48008 [1:44:07<57:48, 5.08it/s] 63%|██████▎ | 30375/48008 [1:44:07<57:10, 5.14it/s] 63%|██████▎ | 30376/48008 [1:44:07<56:56, 5.16it/s] 63%|██████▎ | 30377/48008 [1:44:07<56:10, 5.23it/s] 63%|██████▎ | 30378/48008 [1:44:08<56:07, 5.23it/s] 63%|██████▎ | 30379/48008 [1:44:08<56:18, 5.22it/s] 63%|██████▎ | 30380/48008 [1:44:08<56:21, 5.21it/s] 63%|██████▎ | 30381/48008 [1:44:08<56:20, 5.21it/s] 63%|██████▎ | 30382/48008 [1:44:08<56:22, 5.21it/s] 63%|██████▎ | 30383/48008 [1:44:09<56:39, 5.18it/s] 63%|██████▎ | 30384/48008 [1:44:09<56:41, 5.18it/s] 63%|██████▎ | 30385/48008 [1:44:09<56:58, 5.16it/s] 63%|██████▎ | 30386/48008 [1:44:09<57:10, 5.14it/s] 63%|██████▎ | 30387/48008 [1:44:09<57:32, 5.10it/s] 63%|██████▎ | 30388/48008 [1:44:10<58:46, 5.00it/s] 63%|██████▎ | 30389/48008 [1:44:10<57:52, 5.07it/s] 63%|██████▎ | 30390/48008 [1:44:10<57:07, 5.14it/s] 63%|██████▎ | 30391/48008 [1:44:10<56:47, 5.17it/s] 63%|██████▎ | 30392/48008 [1:44:10<56:38, 5.18it/s] 63%|██████▎ | 30393/48008 [1:44:11<1:06:33, 4.41it/s] 63%|██████▎ | 30394/48008 [1:44:11<1:03:11, 4.65it/s] 63%|██████▎ | 30395/48008 [1:44:11<1:01:05, 4.81it/s] 63%|██████▎ | 30396/48008 [1:44:11<58:49, 4.99it/s] 63%|██████▎ | 30397/48008 [1:44:12<1:15:09, 3.91it/s] 63%|██████▎ | 30398/48008 [1:44:12<1:09:15, 4.24it/s] 63%|██████▎ | 30399/48008 [1:44:12<1:05:14, 4.50it/s] 63%|██████▎ | 30400/48008 [1:44:12<1:02:34, 4.69it/s] {'loss': 0.2996, 'grad_norm': 0.169921875, 'learning_rate': 7.33586068988502e-05, 'epoch': 0.63} 63%|██████▎ | 30400/48008 [1:44:12<1:02:34, 4.69it/s] 63%|██████▎ | 30401/48008 [1:44:12<59:58, 4.89it/s] 63%|██████▎ | 30402/48008 [1:44:13<1:16:01, 3.86it/s] 63%|██████▎ | 30403/48008 [1:44:13<1:09:56, 4.20it/s] 63%|██████▎ | 30404/48008 [1:44:13<1:06:22, 4.42it/s] 63%|██████▎ | 30405/48008 [1:44:13<1:03:09, 4.64it/s] 63%|██████▎ | 30406/48008 [1:44:13<1:00:38, 4.84it/s] 63%|██████▎ | 30407/48008 [1:44:14<59:12, 4.95it/s] 63%|██████▎ | 30408/48008 [1:44:14<58:08, 5.05it/s] 63%|██████▎ | 30409/48008 [1:44:14<57:20, 5.12it/s] 63%|██████▎ | 30410/48008 [1:44:14<56:37, 5.18it/s] 63%|██████▎ | 30411/48008 [1:44:14<56:35, 5.18it/s] 63%|██████▎ | 30412/48008 [1:44:15<1:00:24, 4.86it/s] 63%|██████▎ | 30413/48008 [1:44:15<1:00:10, 4.87it/s] 63%|██████▎ | 30414/48008 [1:44:15<1:02:51, 4.66it/s] 63%|██████▎ | 30415/48008 [1:44:15<1:00:26, 4.85it/s] 63%|██████▎ | 30416/48008 [1:44:16<1:09:20, 4.23it/s] 63%|██████▎ | 30417/48008 [1:44:16<1:05:54, 4.45it/s] 63%|██████▎ | 30418/48008 [1:44:16<1:03:20, 4.63it/s] 63%|██████▎ | 30419/48008 [1:44:16<1:05:01, 4.51it/s] 63%|██████▎ | 30420/48008 [1:44:16<1:02:58, 4.66it/s] 63%|██████▎ | 30421/48008 [1:44:17<1:02:36, 4.68it/s] 63%|██████▎ | 30422/48008 [1:44:17<1:01:16, 4.78it/s] 63%|██████▎ | 30423/48008 [1:44:17<59:53, 4.89it/s] 63%|██████▎ | 30424/48008 [1:44:17<58:24, 5.02it/s] 63%|██████▎ | 30425/48008 [1:44:17<57:15, 5.12it/s] 63%|██████▎ | 30426/48008 [1:44:18<58:27, 5.01it/s] 63%|██████▎ | 30427/48008 [1:44:18<57:48, 5.07it/s] 63%|██████▎ | 30428/48008 [1:44:18<57:59, 5.05it/s] 63%|██████▎ | 30429/48008 [1:44:18<58:29, 5.01it/s] 63%|██████▎ | 30430/48008 [1:44:18<58:56, 4.97it/s] 63%|██████▎ | 30431/48008 [1:44:19<58:06, 5.04it/s] 63%|██████▎ | 30432/48008 [1:44:19<57:18, 5.11it/s] 63%|██████▎ | 30433/48008 [1:44:19<56:02, 5.23it/s] 63%|██████▎ | 30434/48008 [1:44:19<56:36, 5.17it/s] 63%|██████▎ | 30435/48008 [1:44:19<56:04, 5.22it/s] 63%|██████▎ | 30436/48008 [1:44:20<57:07, 5.13it/s] 63%|██████▎ | 30437/48008 [1:44:20<56:55, 5.14it/s] 63%|██████▎ | 30438/48008 [1:44:20<57:03, 5.13it/s] 63%|██████▎ | 30439/48008 [1:44:20<56:43, 5.16it/s] 63%|██████▎ | 30440/48008 [1:44:20<57:32, 5.09it/s] 63%|██████▎ | 30441/48008 [1:44:21<57:07, 5.13it/s] 63%|██████▎ | 30442/48008 [1:44:21<1:06:47, 4.38it/s] 63%|██████▎ | 30443/48008 [1:44:21<1:03:30, 4.61it/s] 63%|██████▎ | 30444/48008 [1:44:21<1:05:11, 4.49it/s] 63%|██████▎ | 30445/48008 [1:44:21<1:02:27, 4.69it/s] 63%|██████▎ | 30446/48008 [1:44:22<1:00:39, 4.83it/s] 63%|██████▎ | 30447/48008 [1:44:22<59:29, 4.92it/s] 63%|██████▎ | 30448/48008 [1:44:22<58:21, 5.01it/s] 63%|██████▎ | 30449/48008 [1:44:22<58:14, 5.02it/s] 63%|██████▎ | 30450/48008 [1:44:22<58:22, 5.01it/s] {'loss': 0.2991, 'grad_norm': 0.15234375, 'learning_rate': 7.3150308281953e-05, 'epoch': 0.63} 63%|██████▎ | 30450/48008 [1:44:22<58:22, 5.01it/s] 63%|██████▎ | 30451/48008 [1:44:23<58:36, 4.99it/s] 63%|██████▎ | 30452/48008 [1:44:23<1:14:48, 3.91it/s] 63%|██████▎ | 30453/48008 [1:44:23<1:13:08, 4.00it/s] 63%|██████▎ | 30454/48008 [1:44:24<1:17:57, 3.75it/s] 63%|██████▎ | 30455/48008 [1:44:24<1:11:45, 4.08it/s] 63%|██████▎ | 30456/48008 [1:44:24<1:06:45, 4.38it/s] 63%|██████▎ | 30457/48008 [1:44:24<1:07:37, 4.33it/s] 63%|██████▎ | 30458/48008 [1:44:25<1:21:07, 3.61it/s] 63%|██████▎ | 30459/48008 [1:44:25<1:13:26, 3.98it/s] 63%|██████▎ | 30460/48008 [1:44:25<1:08:16, 4.28it/s] 63%|██████▎ | 30461/48008 [1:44:25<1:04:12, 4.55it/s] 63%|██████▎ | 30462/48008 [1:44:25<1:01:13, 4.78it/s] 63%|██████▎ | 30463/48008 [1:44:25<59:17, 4.93it/s] 63%|██████▎ | 30464/48008 [1:44:26<58:07, 5.03it/s] 63%|██████▎ | 30465/48008 [1:44:26<57:23, 5.09it/s] 63%|██████▎ | 30466/48008 [1:44:26<56:55, 5.14it/s] 63%|██████▎ | 30467/48008 [1:44:26<56:13, 5.20it/s] 63%|██████▎ | 30468/48008 [1:44:26<56:06, 5.21it/s] 63%|██████▎ | 30469/48008 [1:44:27<55:42, 5.25it/s] 63%|██████▎ | 30470/48008 [1:44:27<55:20, 5.28it/s] 63%|██████▎ | 30471/48008 [1:44:28<1:56:52, 2.50it/s] 63%|██████▎ | 30472/48008 [1:44:28<1:38:55, 2.95it/s] 63%|██████▎ | 30473/48008 [1:44:28<1:26:41, 3.37it/s] 63%|██████▎ | 30474/48008 [1:44:28<1:17:37, 3.77it/s] 63%|██████▎ | 30475/48008 [1:44:28<1:11:00, 4.12it/s] 63%|██████▎ | 30476/48008 [1:44:29<1:06:58, 4.36it/s] 63%|██████▎ | 30477/48008 [1:44:30<2:11:59, 2.21it/s] 63%|██████▎ | 30478/48008 [1:44:30<1:49:49, 2.66it/s] 63%|██████▎ | 30479/48008 [1:44:30<1:37:30, 3.00it/s] 63%|██████▎ | 30480/48008 [1:44:30<1:25:01, 3.44it/s] 63%|██████▎ | 30481/48008 [1:44:30<1:16:02, 3.84it/s] 63%|██████▎ | 30482/48008 [1:44:31<1:11:03, 4.11it/s] 63%|██████▎ | 30483/48008 [1:44:31<2:01:17, 2.41it/s] 63%|██████▎ | 30484/48008 [1:44:32<1:42:42, 2.84it/s] 63%|██████▎ | 30485/48008 [1:44:32<1:28:47, 3.29it/s] 64%|██████▎ | 30486/48008 [1:44:32<1:18:48, 3.71it/s] 64%|██████▎ | 30487/48008 [1:44:32<1:12:03, 4.05it/s] 64%|██████▎ | 30488/48008 [1:44:32<1:06:48, 4.37it/s] 64%|██████▎ | 30489/48008 [1:44:33<1:03:28, 4.60it/s] 64%|██████▎ | 30490/48008 [1:44:33<1:01:22, 4.76it/s] 64%|██████▎ | 30491/48008 [1:44:33<59:47, 4.88it/s] 64%|██████▎ | 30492/48008 [1:44:33<58:19, 5.00it/s] 64%|██████▎ | 30493/48008 [1:44:33<58:01, 5.03it/s] 64%|██████▎ | 30494/48008 [1:44:34<57:08, 5.11it/s] 64%|██████▎ | 30495/48008 [1:44:34<57:02, 5.12it/s] 64%|██████▎ | 30496/48008 [1:44:34<1:00:48, 4.80it/s] 64%|██████▎ | 30497/48008 [1:44:34<59:53, 4.87it/s] 64%|██████▎ | 30498/48008 [1:44:34<58:29, 4.99it/s] 64%|██████▎ | 30499/48008 [1:44:35<58:06, 5.02it/s] 64%|██████▎ | 30500/48008 [1:44:35<58:04, 5.02it/s] {'loss': 0.2985, 'grad_norm': 0.1923828125, 'learning_rate': 7.294200966505583e-05, 'epoch': 0.64} 64%|██████▎ | 30500/48008 [1:44:35<58:04, 5.02it/s] 64%|██████▎ | 30501/48008 [1:44:35<57:22, 5.09it/s] 64%|██████▎ | 30502/48008 [1:44:35<56:38, 5.15it/s] 64%|██████▎ | 30503/48008 [1:44:35<56:20, 5.18it/s] 64%|██████▎ | 30504/48008 [1:44:36<56:21, 5.18it/s] 64%|██████▎ | 30505/48008 [1:44:36<1:00:15, 4.84it/s] 64%|██████▎ | 30506/48008 [1:44:36<58:28, 4.99it/s] 64%|██████▎ | 30507/48008 [1:44:36<57:06, 5.11it/s] 64%|██████▎ | 30508/48008 [1:44:36<57:48, 5.05it/s] 64%|██████▎ | 30509/48008 [1:44:37<56:55, 5.12it/s] 64%|██████▎ | 30510/48008 [1:44:37<1:06:32, 4.38it/s] 64%|██████▎ | 30511/48008 [1:44:37<1:03:19, 4.61it/s] 64%|██████▎ | 30512/48008 [1:44:37<1:01:29, 4.74it/s] 64%|██████▎ | 30513/48008 [1:44:37<1:00:44, 4.80it/s] 64%|██████▎ | 30514/48008 [1:44:38<2:04:51, 2.34it/s] 64%|██████▎ | 30515/48008 [1:44:39<1:43:46, 2.81it/s] 64%|██████▎ | 30516/48008 [1:44:39<1:30:18, 3.23it/s] 64%|██████▎ | 30517/48008 [1:44:39<1:19:55, 3.65it/s] 64%|██████▎ | 30518/48008 [1:44:39<1:12:27, 4.02it/s] 64%|██████▎ | 30519/48008 [1:44:40<2:03:35, 2.36it/s] 64%|██████▎ | 30520/48008 [1:44:40<1:44:22, 2.79it/s] 64%|██████▎ | 30521/48008 [1:44:40<1:30:41, 3.21it/s] 64%|██████▎ | 30522/48008 [1:44:41<1:20:07, 3.64it/s] 64%|██████▎ | 30523/48008 [1:44:41<1:13:16, 3.98it/s] 64%|██████▎ | 30524/48008 [1:44:41<1:08:20, 4.26it/s] 64%|██████▎ | 30525/48008 [1:44:41<1:04:21, 4.53it/s] 64%|██████▎ | 30526/48008 [1:44:41<1:02:46, 4.64it/s] 64%|██████▎ | 30527/48008 [1:44:42<1:00:20, 4.83it/s] 64%|██████▎ | 30528/48008 [1:44:42<1:00:15, 4.83it/s] 64%|██████▎ | 30529/48008 [1:44:42<59:50, 4.87it/s] 64%|██████▎ | 30530/48008 [1:44:42<59:49, 4.87it/s] 64%|██████▎ | 30531/48008 [1:44:42<59:39, 4.88it/s] 64%|██████▎ | 30532/48008 [1:44:43<58:24, 4.99it/s] 64%|██████▎ | 30533/48008 [1:44:43<57:25, 5.07it/s] 64%|██████▎ | 30534/48008 [1:44:43<56:46, 5.13it/s] 64%|██████▎ | 30535/48008 [1:44:43<57:32, 5.06it/s] 64%|██████▎ | 30536/48008 [1:44:43<56:42, 5.13it/s] 64%|██████▎ | 30537/48008 [1:44:44<57:06, 5.10it/s] 64%|██████▎ | 30538/48008 [1:44:44<1:07:00, 4.35it/s] 64%|██████▎ | 30539/48008 [1:44:44<1:03:35, 4.58it/s] 64%|██████▎ | 30540/48008 [1:44:44<1:01:00, 4.77it/s] 64%|██████▎ | 30541/48008 [1:44:44<59:28, 4.89it/s] 64%|██████▎ | 30542/48008 [1:44:45<58:06, 5.01it/s] 64%|██████▎ | 30543/48008 [1:44:45<57:36, 5.05it/s] 64%|██████▎ | 30544/48008 [1:44:45<56:55, 5.11it/s] 64%|██████▎ | 30545/48008 [1:44:45<56:28, 5.15it/s] 64%|██████▎ | 30546/48008 [1:44:45<56:11, 5.18it/s] 64%|██████▎ | 30547/48008 [1:44:46<55:55, 5.20it/s] 64%|██████▎ | 30548/48008 [1:44:46<55:26, 5.25it/s] 64%|██████▎ | 30549/48008 [1:44:46<55:29, 5.24it/s] 64%|██████▎ | 30550/48008 [1:44:46<55:13, 5.27it/s] {'loss': 0.3008, 'grad_norm': 0.2890625, 'learning_rate': 7.273371104815865e-05, 'epoch': 0.64} 64%|██████▎ | 30550/48008 [1:44:46<55:13, 5.27it/s] 64%|██████▎ | 30551/48008 [1:44:46<56:57, 5.11it/s] 64%|██████▎ | 30552/48008 [1:44:47<1:00:50, 4.78it/s] 64%|██████▎ | 30553/48008 [1:44:47<59:22, 4.90it/s] 64%|██████▎ | 30554/48008 [1:44:47<58:05, 5.01it/s] 64%|██████▎ | 30555/48008 [1:44:47<57:04, 5.10it/s] 64%|██████▎ | 30556/48008 [1:44:47<56:01, 5.19it/s] 64%|██████▎ | 30557/48008 [1:44:48<55:52, 5.20it/s] 64%|██████▎ | 30558/48008 [1:44:48<55:47, 5.21it/s] 64%|██████▎ | 30559/48008 [1:44:48<57:08, 5.09it/s] 64%|██████▎ | 30560/48008 [1:44:48<1:13:38, 3.95it/s] 64%|██████▎ | 30561/48008 [1:44:49<1:08:15, 4.26it/s] 64%|██████▎ | 30562/48008 [1:44:49<1:08:19, 4.26it/s] 64%|██████▎ | 30563/48008 [1:44:49<1:05:31, 4.44it/s] 64%|██████▎ | 30564/48008 [1:44:49<1:02:46, 4.63it/s] 64%|██████▎ | 30565/48008 [1:44:49<1:00:57, 4.77it/s] 64%|██████▎ | 30566/48008 [1:44:50<59:11, 4.91it/s] 64%|██████▎ | 30567/48008 [1:44:50<57:13, 5.08it/s] 64%|██████▎ | 30568/48008 [1:44:50<56:16, 5.17it/s] 64%|██████▎ | 30569/48008 [1:44:50<56:48, 5.12it/s] 64%|██████▎ | 30570/48008 [1:44:50<56:29, 5.14it/s] 64%|██████▎ | 30571/48008 [1:44:50<56:10, 5.17it/s] 64%|██████▎ | 30572/48008 [1:44:51<56:25, 5.15it/s] 64%|██████▎ | 30573/48008 [1:44:51<56:05, 5.18it/s] 64%|██████▎ | 30574/48008 [1:44:51<55:51, 5.20it/s] 64%|██████▎ | 30575/48008 [1:44:51<55:24, 5.24it/s] 64%|██████▎ | 30576/48008 [1:44:51<55:16, 5.26it/s] 64%|██████▎ | 30577/48008 [1:44:52<55:23, 5.24it/s] 64%|██████▎ | 30578/48008 [1:44:52<56:21, 5.15it/s] 64%|██████▎ | 30579/48008 [1:44:52<56:31, 5.14it/s] 64%|██████▎ | 30580/48008 [1:44:52<56:16, 5.16it/s] 64%|██████▎ | 30581/48008 [1:44:53<1:13:00, 3.98it/s] 64%|██████▎ | 30582/48008 [1:44:53<1:07:42, 4.29it/s] 64%|██████▎ | 30583/48008 [1:44:53<1:07:53, 4.28it/s] 64%|██████▎ | 30584/48008 [1:44:53<1:05:24, 4.44it/s] 64%|██████▎ | 30585/48008 [1:44:53<1:03:44, 4.56it/s] 64%|██████▎ | 30586/48008 [1:44:54<1:02:28, 4.65it/s] 64%|██████▎ | 30587/48008 [1:44:54<1:01:22, 4.73it/s] 64%|██████▎ | 30588/48008 [1:44:54<59:28, 4.88it/s] 64%|██████▎ | 30589/48008 [1:44:54<58:20, 4.98it/s] 64%|██████▎ | 30590/48008 [1:44:54<57:24, 5.06it/s] 64%|██████▎ | 30591/48008 [1:44:55<56:45, 5.11it/s] 64%|██████▎ | 30592/48008 [1:44:55<55:24, 5.24it/s] 64%|██████▎ | 30593/48008 [1:44:55<55:16, 5.25it/s] 64%|██████▎ | 30594/48008 [1:44:55<54:58, 5.28it/s] 64%|██████▎ | 30595/48008 [1:44:55<56:32, 5.13it/s] 64%|██████▎ | 30596/48008 [1:44:56<55:49, 5.20it/s] 64%|██████▎ | 30597/48008 [1:44:56<55:48, 5.20it/s] 64%|██████▎ | 30598/48008 [1:44:56<59:42, 4.86it/s] 64%|██████▎ | 30599/48008 [1:44:56<58:21, 4.97it/s] 64%|██████▎ | 30600/48008 [1:44:56<57:23, 5.05it/s] 64%|██████▎ | 30600/48008 [1:44:56<57:23, 5.05it/s]{'loss': 0.3001, 'grad_norm': 0.373046875, 'learning_rate': 7.252541243126147e-05, 'epoch': 0.64} 64%|██████▎ | 30601/48008 [1:44:57<56:27, 5.14it/s] 64%|██████▎ | 30602/48008 [1:44:57<57:42, 5.03it/s] 64%|██████▎ | 30603/48008 [1:44:57<57:43, 5.02it/s] 64%|██████▎ | 30604/48008 [1:44:57<56:27, 5.14it/s] 64%|██████▎ | 30605/48008 [1:44:57<55:41, 5.21it/s] 64%|██████▍ | 30606/48008 [1:44:58<57:18, 5.06it/s] 64%|██████▍ | 30607/48008 [1:44:58<56:39, 5.12it/s] 64%|██████▍ | 30608/48008 [1:44:58<55:49, 5.19it/s] 64%|██████▍ | 30609/48008 [1:44:58<56:22, 5.14it/s] 64%|██████▍ | 30610/48008 [1:44:58<56:06, 5.17it/s] 64%|██████▍ | 30611/48008 [1:44:59<55:51, 5.19it/s] 64%|██████▍ | 30612/48008 [1:44:59<55:48, 5.19it/s] 64%|██████▍ | 30613/48008 [1:44:59<54:44, 5.30it/s] 64%|██████▍ | 30614/48008 [1:44:59<54:24, 5.33it/s] 64%|██████▍ | 30615/48008 [1:44:59<58:43, 4.94it/s] 64%|██████▍ | 30616/48008 [1:45:00<58:10, 4.98it/s] 64%|██████▍ | 30617/48008 [1:45:00<57:12, 5.07it/s] 64%|██████▍ | 30618/48008 [1:45:00<56:12, 5.16it/s] 64%|██████▍ | 30619/48008 [1:45:00<1:00:01, 4.83it/s] 64%|██████▍ | 30620/48008 [1:45:00<58:38, 4.94it/s] 64%|██████▍ | 30621/48008 [1:45:01<58:38, 4.94it/s] 64%|██████▍ | 30622/48008 [1:45:01<57:35, 5.03it/s] 64%|██████▍ | 30623/48008 [1:45:01<56:49, 5.10it/s] 64%|██████▍ | 30624/48008 [1:45:01<56:54, 5.09it/s] 64%|██████▍ | 30625/48008 [1:45:01<57:28, 5.04it/s] 64%|██████▍ | 30626/48008 [1:45:01<57:20, 5.05it/s] 64%|██████▍ | 30627/48008 [1:45:02<56:14, 5.15it/s] 64%|██████▍ | 30628/48008 [1:45:02<56:00, 5.17it/s] 64%|██████▍ | 30629/48008 [1:45:02<55:40, 5.20it/s] 64%|██████▍ | 30630/48008 [1:45:02<55:42, 5.20it/s] 64%|██████▍ | 30631/48008 [1:45:02<56:44, 5.10it/s] 64%|██████▍ | 30632/48008 [1:45:03<55:53, 5.18it/s] 64%|██████▍ | 30633/48008 [1:45:03<1:12:40, 3.98it/s] 64%|██████▍ | 30634/48008 [1:45:03<1:08:58, 4.20it/s] 64%|██████▍ | 30635/48008 [1:45:03<1:06:22, 4.36it/s] 64%|██████▍ | 30636/48008 [1:45:04<1:04:09, 4.51it/s] 64%|██████▍ | 30637/48008 [1:45:04<1:01:53, 4.68it/s] 64%|██████▍ | 30638/48008 [1:45:04<1:00:03, 4.82it/s] 64%|██████▍ | 30639/48008 [1:45:04<1:00:08, 4.81it/s] 64%|██████▍ | 30640/48008 [1:45:04<58:07, 4.98it/s] 64%|██████▍ | 30641/48008 [1:45:05<57:10, 5.06it/s] 64%|██████▍ | 30642/48008 [1:45:05<57:13, 5.06it/s] 64%|██████▍ | 30643/48008 [1:45:05<56:26, 5.13it/s] 64%|██████▍ | 30644/48008 [1:45:05<55:48, 5.19it/s] 64%|██████▍ | 30645/48008 [1:45:05<55:38, 5.20it/s] 64%|██████▍ | 30646/48008 [1:45:06<55:17, 5.23it/s] 64%|██████▍ | 30647/48008 [1:45:06<56:20, 5.14it/s] 64%|██████▍ | 30648/48008 [1:45:06<55:35, 5.21it/s] 64%|██████▍ | 30649/48008 [1:45:06<55:30, 5.21it/s] 64%|██████▍ | 30650/48008 [1:45:06<56:28, 5.12it/s] {'loss': 0.3022, 'grad_norm': 0.18359375, 'learning_rate': 7.231711381436428e-05, 'epoch': 0.64} 64%|██████▍ | 30650/48008 [1:45:06<56:28, 5.12it/s] 64%|██████▍ | 30651/48008 [1:45:07<1:06:07, 4.37it/s] 64%|██████▍ | 30652/48008 [1:45:07<1:02:47, 4.61it/s] 64%|██████▍ | 30653/48008 [1:45:07<1:01:23, 4.71it/s] 64%|██████▍ | 30654/48008 [1:45:07<1:00:33, 4.78it/s] 64%|██████▍ | 30655/48008 [1:45:07<58:38, 4.93it/s] 64%|██████▍ | 30656/48008 [1:45:08<57:45, 5.01it/s] 64%|██████▍ | 30657/48008 [1:45:08<57:10, 5.06it/s] 64%|██████▍ | 30658/48008 [1:45:08<56:21, 5.13it/s] 64%|██████▍ | 30659/48008 [1:45:08<55:47, 5.18it/s] 64%|██████▍ | 30660/48008 [1:45:08<56:16, 5.14it/s] 64%|██████▍ | 30661/48008 [1:45:09<56:00, 5.16it/s] 64%|██████▍ | 30662/48008 [1:45:09<55:37, 5.20it/s] 64%|██████▍ | 30663/48008 [1:45:09<56:05, 5.15it/s] 64%|██████▍ | 30664/48008 [1:45:09<56:10, 5.15it/s] 64%|██████▍ | 30665/48008 [1:45:09<55:21, 5.22it/s] 64%|██████▍ | 30666/48008 [1:45:10<1:12:18, 4.00it/s] 64%|██████▍ | 30667/48008 [1:45:10<1:07:00, 4.31it/s] 64%|██████▍ | 30668/48008 [1:45:10<1:03:47, 4.53it/s] 64%|██████▍ | 30669/48008 [1:45:10<1:01:13, 4.72it/s] 64%|██████▍ | 30670/48008 [1:45:11<58:44, 4.92it/s] 64%|██████▍ | 30671/48008 [1:45:11<59:15, 4.88it/s] 64%|██████▍ | 30672/48008 [1:45:11<1:08:09, 4.24it/s] 64%|██████▍ | 30673/48008 [1:45:11<1:05:07, 4.44it/s] 64%|██████▍ | 30674/48008 [1:45:11<1:01:39, 4.69it/s] 64%|██████▍ | 30675/48008 [1:45:12<1:00:14, 4.80it/s] 64%|██████▍ | 30676/48008 [1:45:12<58:26, 4.94it/s] 64%|██████▍ | 30677/48008 [1:45:12<57:35, 5.02it/s] 64%|██████▍ | 30678/48008 [1:45:12<57:09, 5.05it/s] 64%|██████▍ | 30679/48008 [1:45:12<56:23, 5.12it/s] 64%|██████▍ | 30680/48008 [1:45:13<55:57, 5.16it/s] 64%|██████▍ | 30681/48008 [1:45:13<55:55, 5.16it/s] 64%|██████▍ | 30682/48008 [1:45:13<55:47, 5.18it/s] 64%|██████▍ | 30683/48008 [1:45:13<56:38, 5.10it/s] 64%|██████▍ | 30684/48008 [1:45:13<56:27, 5.11it/s] 64%|██████▍ | 30685/48008 [1:45:14<59:59, 4.81it/s] 64%|██████▍ | 30686/48008 [1:45:14<58:32, 4.93it/s] 64%|██████▍ | 30687/48008 [1:45:14<57:45, 5.00it/s] 64%|██████▍ | 30688/48008 [1:45:14<57:40, 5.01it/s] 64%|██████▍ | 30689/48008 [1:45:14<57:29, 5.02it/s] 64%|██████▍ | 30690/48008 [1:45:15<56:46, 5.08it/s] 64%|██████▍ | 30691/48008 [1:45:15<57:40, 5.00it/s] 64%|██████▍ | 30692/48008 [1:45:15<57:22, 5.03it/s] 64%|██████▍ | 30693/48008 [1:45:15<56:19, 5.12it/s] 64%|██████▍ | 30694/48008 [1:45:15<55:51, 5.17it/s] 64%|██████▍ | 30695/48008 [1:45:16<59:41, 4.83it/s] 64%|██████▍ | 30696/48008 [1:45:16<58:17, 4.95it/s] 64%|██████▍ | 30697/48008 [1:45:16<57:13, 5.04it/s] 64%|██████▍ | 30698/48008 [1:45:16<56:35, 5.10it/s] 64%|██████▍ | 30699/48008 [1:45:16<56:09, 5.14it/s] 64%|██████▍ | 30700/48008 [1:45:17<56:46, 5.08it/s] {'loss': 0.3007, 'grad_norm': 0.1728515625, 'learning_rate': 7.21088151974671e-05, 'epoch': 0.64} 64%|██████▍ | 30700/48008 [1:45:17<56:46, 5.08it/s] 64%|██████▍ | 30701/48008 [1:45:17<1:00:25, 4.77it/s] 64%|██████▍ | 30702/48008 [1:45:17<59:30, 4.85it/s] 64%|██████▍ | 30703/48008 [1:45:17<58:16, 4.95it/s] 64%|██████▍ | 30704/48008 [1:45:17<57:03, 5.05it/s] 64%|██████▍ | 30705/48008 [1:45:18<57:24, 5.02it/s] 64%|██████▍ | 30706/48008 [1:45:18<56:19, 5.12it/s] 64%|██████▍ | 30707/48008 [1:45:18<56:26, 5.11it/s] 64%|██████▍ | 30708/48008 [1:45:18<55:55, 5.16it/s] 64%|██████▍ | 30709/48008 [1:45:18<55:48, 5.17it/s] 64%|██████▍ | 30710/48008 [1:45:19<55:34, 5.19it/s] 64%|██████▍ | 30711/48008 [1:45:19<54:58, 5.24it/s] 64%|██████▍ | 30712/48008 [1:45:19<56:08, 5.13it/s] 64%|██████▍ | 30713/48008 [1:45:19<55:48, 5.16it/s] 64%|██████▍ | 30714/48008 [1:45:19<55:40, 5.18it/s] 64%|██████▍ | 30715/48008 [1:45:19<55:25, 5.20it/s] 64%|██████▍ | 30716/48008 [1:45:20<54:58, 5.24it/s] 64%|██████▍ | 30717/48008 [1:45:20<55:36, 5.18it/s] 64%|██████▍ | 30718/48008 [1:45:20<56:16, 5.12it/s] 64%|██████▍ | 30719/48008 [1:45:20<56:17, 5.12it/s] 64%|██████▍ | 30720/48008 [1:45:20<55:49, 5.16it/s] 64%|██████▍ | 30721/48008 [1:45:21<56:01, 5.14it/s] 64%|██████▍ | 30722/48008 [1:45:21<55:33, 5.19it/s] 64%|██████▍ | 30723/48008 [1:45:21<55:45, 5.17it/s] 64%|██████▍ | 30724/48008 [1:45:21<56:11, 5.13it/s] 64%|██████▍ | 30725/48008 [1:45:21<59:46, 4.82it/s] 64%|██████▍ | 30726/48008 [1:45:22<57:50, 4.98it/s] 64%|██████▍ | 30727/48008 [1:45:22<56:31, 5.10it/s] 64%|██████▍ | 30728/48008 [1:45:22<56:26, 5.10it/s] 64%|██████▍ | 30729/48008 [1:45:22<55:53, 5.15it/s] 64%|██████▍ | 30730/48008 [1:45:22<56:42, 5.08it/s] 64%|██████▍ | 30731/48008 [1:45:23<56:18, 5.11it/s] 64%|██████▍ | 30732/48008 [1:45:23<1:05:56, 4.37it/s] 64%|██████▍ | 30733/48008 [1:45:23<1:02:40, 4.59it/s] 64%|██████▍ | 30734/48008 [1:45:23<1:00:08, 4.79it/s] 64%|██████▍ | 30735/48008 [1:45:23<58:37, 4.91it/s] 64%|██████▍ | 30736/48008 [1:45:24<1:14:28, 3.87it/s] 64%|██████▍ | 30737/48008 [1:45:24<1:08:56, 4.18it/s] 64%|██████▍ | 30738/48008 [1:45:24<1:04:33, 4.46it/s] 64%|██████▍ | 30739/48008 [1:45:24<1:01:34, 4.67it/s] 64%|██████▍ | 30740/48008 [1:45:25<59:53, 4.81it/s] 64%|██████▍ | 30741/48008 [1:45:25<58:07, 4.95it/s] 64%|██████▍ | 30742/48008 [1:45:25<58:20, 4.93it/s] 64%|██████▍ | 30743/48008 [1:45:25<59:09, 4.86it/s] 64%|██████▍ | 30744/48008 [1:45:25<57:48, 4.98it/s] 64%|██████▍ | 30745/48008 [1:45:26<56:39, 5.08it/s] 64%|██████▍ | 30746/48008 [1:45:26<57:36, 4.99it/s] 64%|██████▍ | 30747/48008 [1:45:26<57:52, 4.97it/s] 64%|██████▍ | 30748/48008 [1:45:26<57:06, 5.04it/s] 64%|██████▍ | 30749/48008 [1:45:26<56:13, 5.12it/s] 64%|██████▍ | 30750/48008 [1:45:27<1:05:51, 4.37it/s] {'loss': 0.3008, 'grad_norm': 0.32421875, 'learning_rate': 7.19005165805699e-05, 'epoch': 0.64} 64%|██████▍ | 30750/48008 [1:45:27<1:05:51, 4.37it/s] 64%|██████▍ | 30751/48008 [1:45:27<1:03:19, 4.54it/s] 64%|██████▍ | 30752/48008 [1:45:27<1:04:52, 4.43it/s] 64%|██████▍ | 30753/48008 [1:45:27<1:05:57, 4.36it/s] 64%|██████▍ | 30754/48008 [1:45:28<1:12:44, 3.95it/s] 64%|██████▍ | 30755/48008 [1:45:28<1:07:15, 4.28it/s] 64%|██████▍ | 30756/48008 [1:45:28<1:04:15, 4.47it/s] 64%|██████▍ | 30757/48008 [1:45:28<1:01:29, 4.68it/s] 64%|██████▍ | 30758/48008 [1:45:28<59:25, 4.84it/s] 64%|██████▍ | 30759/48008 [1:45:29<57:35, 4.99it/s] 64%|██████▍ | 30760/48008 [1:45:29<56:54, 5.05it/s] 64%|██████▍ | 30761/48008 [1:45:29<55:37, 5.17it/s] 64%|██████▍ | 30762/48008 [1:45:29<55:22, 5.19it/s] 64%|██████▍ | 30763/48008 [1:45:29<54:58, 5.23it/s] 64%|██████▍ | 30764/48008 [1:45:30<55:02, 5.22it/s] 64%|██████▍ | 30765/48008 [1:45:30<54:33, 5.27it/s] 64%|██████▍ | 30766/48008 [1:45:30<54:12, 5.30it/s] 64%|██████▍ | 30767/48008 [1:45:30<54:24, 5.28it/s] 64%|██████▍ | 30768/48008 [1:45:30<55:25, 5.18it/s] 64%|██████▍ | 30769/48008 [1:45:31<55:17, 5.20it/s] 64%|██████▍ | 30770/48008 [1:45:31<55:12, 5.20it/s] 64%|██████▍ | 30771/48008 [1:45:31<54:47, 5.24it/s] 64%|██████▍ | 30772/48008 [1:45:31<55:19, 5.19it/s] 64%|██████▍ | 30773/48008 [1:45:31<56:23, 5.09it/s] 64%|██████▍ | 30774/48008 [1:45:32<55:58, 5.13it/s] 64%|██████▍ | 30775/48008 [1:45:32<55:13, 5.20it/s] 64%|██████▍ | 30776/48008 [1:45:32<1:05:32, 4.38it/s] 64%|██████▍ | 30777/48008 [1:45:32<1:02:33, 4.59it/s] 64%|██████▍ | 30778/48008 [1:45:32<1:00:44, 4.73it/s] 64%|██████▍ | 30779/48008 [1:45:33<58:54, 4.87it/s] 64%|██████▍ | 30780/48008 [1:45:33<57:31, 4.99it/s] 64%|██████▍ | 30781/48008 [1:45:33<57:51, 4.96it/s] 64%|██████▍ | 30782/48008 [1:45:33<56:51, 5.05it/s] 64%|██████▍ | 30783/48008 [1:45:33<56:37, 5.07it/s] 64%|██████▍ | 30784/48008 [1:45:34<57:48, 4.97it/s] 64%|██████▍ | 30785/48008 [1:45:34<1:06:41, 4.30it/s] 64%|██████▍ | 30786/48008 [1:45:34<1:19:52, 3.59it/s] 64%|██████▍ | 30787/48008 [1:45:34<1:12:14, 3.97it/s] 64%|██████▍ | 30788/48008 [1:45:35<1:07:06, 4.28it/s] 64%|██████▍ | 30789/48008 [1:45:35<1:03:32, 4.52it/s] 64%|██████▍ | 30790/48008 [1:45:35<1:02:33, 4.59it/s] 64%|██████▍ | 30791/48008 [1:45:35<1:00:17, 4.76it/s] 64%|██████▍ | 30792/48008 [1:45:36<1:08:28, 4.19it/s] 64%|██████▍ | 30793/48008 [1:45:36<1:04:41, 4.44it/s] 64%|██████▍ | 30794/48008 [1:45:36<1:02:10, 4.61it/s] 64%|██████▍ | 30795/48008 [1:45:36<59:51, 4.79it/s] 64%|██████▍ | 30796/48008 [1:45:36<58:27, 4.91it/s] 64%|██████▍ | 30797/48008 [1:45:37<57:14, 5.01it/s] 64%|██████▍ | 30798/48008 [1:45:37<56:47, 5.05it/s] 64%|██████▍ | 30799/48008 [1:45:37<56:03, 5.12it/s] 64%|██████▍ | 30800/48008 [1:45:37<55:18, 5.19it/s] {'loss': 0.2992, 'grad_norm': 0.361328125, 'learning_rate': 7.169221796367272e-05, 'epoch': 0.64} 64%|██████▍ | 30800/48008 [1:45:37<55:18, 5.19it/s] 64%|██████▍ | 30801/48008 [1:45:37<55:10, 5.20it/s] 64%|██████▍ | 30802/48008 [1:45:38<55:16, 5.19it/s] 64%|██████▍ | 30803/48008 [1:45:38<55:05, 5.20it/s] 64%|██████▍ | 30804/48008 [1:45:38<58:56, 4.86it/s] 64%|██████▍ | 30805/48008 [1:45:38<57:03, 5.03it/s] 64%|██████▍ | 30806/48008 [1:45:38<56:11, 5.10it/s] 64%|██████▍ | 30807/48008 [1:45:39<56:40, 5.06it/s] 64%|██████▍ | 30808/48008 [1:45:39<57:06, 5.02it/s] 64%|██████▍ | 30809/48008 [1:45:39<1:13:16, 3.91it/s] 64%|██████▍ | 30810/48008 [1:45:39<1:07:53, 4.22it/s] 64%|██████▍ | 30811/48008 [1:45:39<1:03:58, 4.48it/s] 64%|██████▍ | 30812/48008 [1:45:40<1:01:26, 4.67it/s] 64%|██████▍ | 30813/48008 [1:45:40<59:36, 4.81it/s] 64%|██████▍ | 30814/48008 [1:45:40<59:33, 4.81it/s] 64%|██████▍ | 30815/48008 [1:45:40<57:53, 4.95it/s] 64%|██████▍ | 30816/48008 [1:45:40<56:31, 5.07it/s] 64%|██████▍ | 30817/48008 [1:45:41<1:12:41, 3.94it/s] 64%|██████▍ | 30818/48008 [1:45:41<1:07:47, 4.23it/s] 64%|██████▍ | 30819/48008 [1:45:41<1:03:52, 4.48it/s] 64%|██████▍ | 30820/48008 [1:45:41<1:00:45, 4.72it/s] 64%|██████▍ | 30821/48008 [1:45:42<59:52, 4.78it/s] 64%|██████▍ | 30822/48008 [1:45:42<59:18, 4.83it/s] 64%|██████▍ | 30823/48008 [1:45:42<58:05, 4.93it/s] 64%|██████▍ | 30824/48008 [1:45:42<56:40, 5.05it/s] 64%|██████▍ | 30825/48008 [1:45:42<56:14, 5.09it/s] 64%|██████▍ | 30826/48008 [1:45:43<55:22, 5.17it/s] 64%|██████▍ | 30827/48008 [1:45:43<55:10, 5.19it/s] 64%|██████▍ | 30828/48008 [1:45:43<55:40, 5.14it/s] 64%|██████▍ | 30829/48008 [1:45:43<55:13, 5.18it/s] 64%|██████▍ | 30830/48008 [1:45:43<56:24, 5.07it/s] 64%|██████▍ | 30831/48008 [1:45:44<55:31, 5.16it/s] 64%|██████▍ | 30832/48008 [1:45:44<55:46, 5.13it/s] 64%|██████▍ | 30833/48008 [1:45:44<54:54, 5.21it/s] 64%|██████▍ | 30834/48008 [1:45:44<55:25, 5.16it/s] 64%|██████▍ | 30835/48008 [1:45:44<56:10, 5.09it/s] 64%|██████▍ | 30836/48008 [1:45:45<56:50, 5.04it/s] 64%|██████▍ | 30837/48008 [1:45:45<57:20, 4.99it/s] 64%|██████▍ | 30838/48008 [1:45:45<57:36, 4.97it/s] 64%|██████▍ | 30839/48008 [1:45:45<57:09, 5.01it/s] 64%|██████▍ | 30840/48008 [1:45:45<55:52, 5.12it/s] 64%|██████▍ | 30841/48008 [1:45:46<55:06, 5.19it/s] 64%|██████▍ | 30842/48008 [1:45:46<56:02, 5.10it/s] 64%|██████▍ | 30843/48008 [1:45:46<56:43, 5.04it/s] 64%|██████▍ | 30844/48008 [1:45:46<55:36, 5.14it/s] 64%|██████▍ | 30845/48008 [1:45:46<56:17, 5.08it/s] 64%|██████▍ | 30846/48008 [1:45:46<55:48, 5.13it/s] 64%|██████▍ | 30847/48008 [1:45:47<55:03, 5.19it/s] 64%|██████▍ | 30848/48008 [1:45:47<54:26, 5.25it/s] 64%|██████▍ | 30849/48008 [1:45:47<54:26, 5.25it/s] 64%|██████▍ | 30850/48008 [1:45:47<54:31, 5.24it/s] {'loss': 0.2997, 'grad_norm': 0.21875, 'learning_rate': 7.148391934677554e-05, 'epoch': 0.64} 64%|██████▍ | 30850/48008 [1:45:47<54:31, 5.24it/s] 64%|██████▍ | 30851/48008 [1:45:47<54:25, 5.25it/s] 64%|██████▍ | 30852/48008 [1:45:48<54:41, 5.23it/s] 64%|██████▍ | 30853/48008 [1:45:48<54:26, 5.25it/s] 64%|██████▍ | 30854/48008 [1:45:48<54:49, 5.22it/s] 64%|██████▍ | 30855/48008 [1:45:48<54:52, 5.21it/s] 64%|██████▍ | 30856/48008 [1:45:48<54:44, 5.22it/s] 64%|██████▍ | 30857/48008 [1:45:49<54:40, 5.23it/s] 64%|██████▍ | 30858/48008 [1:45:49<54:55, 5.20it/s] 64%|██████▍ | 30859/48008 [1:45:49<54:40, 5.23it/s] 64%|██████▍ | 30860/48008 [1:45:49<55:07, 5.18it/s] 64%|██████▍ | 30861/48008 [1:45:49<54:57, 5.20it/s] 64%|██████▍ | 30862/48008 [1:45:50<54:55, 5.20it/s] 64%|██████▍ | 30863/48008 [1:45:50<55:30, 5.15it/s] 64%|██████▍ | 30864/48008 [1:45:50<55:18, 5.17it/s] 64%|██████▍ | 30865/48008 [1:45:50<54:35, 5.23it/s] 64%|██████▍ | 30866/48008 [1:45:50<55:12, 5.17it/s] 64%|██████▍ | 30867/48008 [1:45:51<55:12, 5.17it/s] 64%|██████▍ | 30868/48008 [1:45:51<54:43, 5.22it/s] 64%|██████▍ | 30869/48008 [1:45:51<54:31, 5.24it/s] 64%|██████▍ | 30870/48008 [1:45:51<54:40, 5.22it/s] 64%|██████▍ | 30871/48008 [1:45:51<54:45, 5.22it/s] 64%|██████▍ | 30872/48008 [1:45:51<54:31, 5.24it/s] 64%|██████▍ | 30873/48008 [1:45:52<53:58, 5.29it/s] 64%|██████▍ | 30874/48008 [1:45:52<54:40, 5.22it/s] 64%|██████▍ | 30875/48008 [1:45:52<54:45, 5.22it/s] 64%|██████▍ | 30876/48008 [1:45:52<54:39, 5.22it/s] 64%|██████▍ | 30877/48008 [1:45:52<54:36, 5.23it/s] 64%|██████▍ | 30878/48008 [1:45:53<54:55, 5.20it/s] 64%|██████▍ | 30879/48008 [1:45:53<54:46, 5.21it/s] 64%|██████▍ | 30880/48008 [1:45:53<56:05, 5.09it/s] 64%|██████▍ | 30881/48008 [1:45:53<55:10, 5.17it/s] 64%|██████▍ | 30882/48008 [1:45:53<54:24, 5.25it/s] 64%|██████▍ | 30883/48008 [1:45:54<54:27, 5.24it/s] 64%|██████▍ | 30884/48008 [1:45:54<54:32, 5.23it/s] 64%|██████▍ | 30885/48008 [1:45:54<55:21, 5.15it/s] 64%|██████▍ | 30886/48008 [1:45:54<55:05, 5.18it/s] 64%|██████▍ | 30887/48008 [1:45:54<56:23, 5.06it/s] 64%|██████▍ | 30888/48008 [1:45:55<55:46, 5.12it/s] 64%|██████▍ | 30889/48008 [1:45:55<55:00, 5.19it/s] 64%|██████▍ | 30890/48008 [1:45:55<55:12, 5.17it/s] 64%|██████▍ | 30891/48008 [1:45:55<54:55, 5.19it/s] 64%|██████▍ | 30892/48008 [1:45:55<54:37, 5.22it/s] 64%|██████▍ | 30893/48008 [1:45:56<55:13, 5.16it/s] 64%|██████▍ | 30894/48008 [1:45:56<55:52, 5.11it/s] 64%|██████▍ | 30895/48008 [1:45:56<55:36, 5.13it/s] 64%|██████▍ | 30896/48008 [1:45:56<55:38, 5.13it/s] 64%|██████▍ | 30897/48008 [1:45:56<54:56, 5.19it/s] 64%|██████▍ | 30898/48008 [1:45:57<56:16, 5.07it/s] 64%|██████▍ | 30899/48008 [1:45:57<55:40, 5.12it/s] 64%|██████▍ | 30900/48008 [1:45:57<55:25, 5.14it/s] {'loss': 0.3007, 'grad_norm': 0.146484375, 'learning_rate': 7.127562072987836e-05, 'epoch': 0.64} 64%|██████▍ | 30900/48008 [1:45:57<55:25, 5.14it/s] 64%|██████▍ | 30901/48008 [1:45:57<55:52, 5.10it/s] 64%|██████▍ | 30902/48008 [1:45:57<55:14, 5.16it/s] 64%|██████▍ | 30903/48008 [1:45:57<55:01, 5.18it/s] 64%|██████▍ | 30904/48008 [1:45:58<54:51, 5.20it/s] 64%|██████▍ | 30905/48008 [1:45:58<54:45, 5.21it/s] 64%|██████▍ | 30906/48008 [1:45:58<55:39, 5.12it/s] 64%|██████▍ | 30907/48008 [1:45:58<55:21, 5.15it/s] 64%|██████▍ | 30908/48008 [1:45:58<54:46, 5.20it/s] 64%|██████▍ | 30909/48008 [1:45:59<55:50, 5.10it/s] 64%|██████▍ | 30910/48008 [1:45:59<55:12, 5.16it/s] 64%|██████▍ | 30911/48008 [1:45:59<55:56, 5.09it/s] 64%|██████▍ | 30912/48008 [1:45:59<56:08, 5.08it/s] 64%|██████▍ | 30913/48008 [1:45:59<56:27, 5.05it/s] 64%|██████▍ | 30914/48008 [1:46:00<56:57, 5.00it/s] 64%|██████▍ | 30915/48008 [1:46:00<55:54, 5.10it/s] 64%|██████▍ | 30916/48008 [1:46:00<55:20, 5.15it/s] 64%|██████▍ | 30917/48008 [1:46:00<55:01, 5.18it/s] 64%|██████▍ | 30918/48008 [1:46:00<54:50, 5.19it/s] 64%|██████▍ | 30919/48008 [1:46:01<54:37, 5.21it/s] 64%|██████▍ | 30920/48008 [1:46:01<54:28, 5.23it/s] 64%|██████▍ | 30921/48008 [1:46:01<55:37, 5.12it/s] 64%|██████▍ | 30922/48008 [1:46:01<55:39, 5.12it/s] 64%|██████▍ | 30923/48008 [1:46:01<55:06, 5.17it/s] 64%|██████▍ | 30924/48008 [1:46:02<54:48, 5.19it/s] 64%|██████▍ | 30925/48008 [1:46:02<54:20, 5.24it/s] 64%|██████▍ | 30926/48008 [1:46:02<54:16, 5.25it/s] 64%|██████▍ | 30927/48008 [1:46:02<54:47, 5.20it/s] 64%|██████▍ | 30928/48008 [1:46:02<55:03, 5.17it/s] 64%|██████▍ | 30929/48008 [1:46:03<55:48, 5.10it/s] 64%|██████▍ | 30930/48008 [1:46:03<55:12, 5.16it/s] 64%|██████▍ | 30931/48008 [1:46:03<55:17, 5.15it/s] 64%|██████▍ | 30932/48008 [1:46:03<54:52, 5.19it/s] 64%|██████▍ | 30933/48008 [1:46:03<54:02, 5.27it/s] 64%|██████▍ | 30934/48008 [1:46:03<54:34, 5.21it/s] 64%|██████▍ | 30935/48008 [1:46:04<53:55, 5.28it/s] 64%|██████▍ | 30936/48008 [1:46:04<54:07, 5.26it/s] 64%|██████▍ | 30937/48008 [1:46:04<55:11, 5.16it/s] 64%|██████▍ | 30938/48008 [1:46:04<54:58, 5.17it/s] 64%|██████▍ | 30939/48008 [1:46:04<55:00, 5.17it/s] 64%|██████▍ | 30940/48008 [1:46:05<56:31, 5.03it/s] 64%|██████▍ | 30941/48008 [1:46:05<57:29, 4.95it/s] 64%|██████▍ | 30942/48008 [1:46:05<56:40, 5.02it/s] 64%|██████▍ | 30943/48008 [1:46:05<56:05, 5.07it/s] 64%|██████▍ | 30944/48008 [1:46:05<55:48, 5.10it/s] 64%|██████▍ | 30945/48008 [1:46:06<55:25, 5.13it/s] 64%|██████▍ | 30946/48008 [1:46:06<54:50, 5.19it/s] 64%|██████▍ | 30947/48008 [1:46:06<55:18, 5.14it/s] 64%|██████▍ | 30948/48008 [1:46:06<55:44, 5.10it/s] 64%|██████▍ | 30949/48008 [1:46:06<55:25, 5.13it/s] 64%|██████▍ | 30950/48008 [1:46:07<54:51, 5.18it/s] {'loss': 0.3019, 'grad_norm': 0.431640625, 'learning_rate': 7.106732211298117e-05, 'epoch': 0.64} 64%|██████▍ | 30950/48008 [1:46:07<54:51, 5.18it/s] 64%|██████▍ | 30951/48008 [1:46:07<55:01, 5.17it/s] 64%|██████▍ | 30952/48008 [1:46:07<1:04:27, 4.41it/s] 64%|██████▍ | 30953/48008 [1:46:07<1:02:08, 4.57it/s] 64%|██████▍ | 30954/48008 [1:46:07<1:00:48, 4.67it/s] 64%|██████▍ | 30955/48008 [1:46:08<58:48, 4.83it/s] 64%|██████▍ | 30956/48008 [1:46:08<57:50, 4.91it/s] 64%|██████▍ | 30957/48008 [1:46:08<56:48, 5.00it/s] 64%|██████▍ | 30958/48008 [1:46:08<57:01, 4.98it/s] 64%|██████▍ | 30959/48008 [1:46:08<57:17, 4.96it/s] 64%|██████▍ | 30960/48008 [1:46:09<57:21, 4.95it/s] 64%|██████▍ | 30961/48008 [1:46:09<56:06, 5.06it/s] 64%|██████▍ | 30962/48008 [1:46:09<54:56, 5.17it/s] 64%|██████▍ | 30963/48008 [1:46:09<1:11:22, 3.98it/s] 64%|██████▍ | 30964/48008 [1:46:10<1:06:19, 4.28it/s] 64%|██████▍ | 30965/48008 [1:46:10<1:02:41, 4.53it/s] 65%|██████▍ | 30966/48008 [1:46:10<1:09:56, 4.06it/s] 65%|██████▍ | 30967/48008 [1:46:10<1:05:16, 4.35it/s] 65%|██████▍ | 30968/48008 [1:46:11<1:01:50, 4.59it/s] 65%|██████▍ | 30969/48008 [1:46:11<59:07, 4.80it/s] 65%|██████▍ | 30970/48008 [1:46:11<58:39, 4.84it/s] 65%|██████▍ | 30971/48008 [1:46:11<57:04, 4.97it/s] 65%|██████▍ | 30972/48008 [1:46:11<57:14, 4.96it/s] 65%|██████▍ | 30973/48008 [1:46:11<56:14, 5.05it/s] 65%|██████▍ | 30974/48008 [1:46:12<56:32, 5.02it/s] 65%|██████▍ | 30975/48008 [1:46:12<56:02, 5.07it/s] 65%|██████▍ | 30976/48008 [1:46:12<54:50, 5.18it/s] 65%|██████▍ | 30977/48008 [1:46:12<54:42, 5.19it/s] 65%|██████▍ | 30978/48008 [1:46:12<56:00, 5.07it/s] 65%|██████▍ | 30979/48008 [1:46:13<55:28, 5.12it/s] 65%|██████▍ | 30980/48008 [1:46:13<55:35, 5.11it/s] 65%|██████▍ | 30981/48008 [1:46:13<54:52, 5.17it/s] 65%|██████▍ | 30982/48008 [1:46:13<54:22, 5.22it/s] 65%|██████▍ | 30983/48008 [1:46:13<58:02, 4.89it/s] 65%|██████▍ | 30984/48008 [1:46:14<56:59, 4.98it/s] 65%|██████▍ | 30985/48008 [1:46:14<56:31, 5.02it/s] 65%|██████▍ | 30986/48008 [1:46:14<56:35, 5.01it/s] 65%|██████▍ | 30987/48008 [1:46:14<55:36, 5.10it/s] 65%|██████▍ | 30988/48008 [1:46:14<55:28, 5.11it/s] 65%|██████▍ | 30989/48008 [1:46:15<54:59, 5.16it/s] 65%|██████▍ | 30990/48008 [1:46:15<54:43, 5.18it/s] 65%|██████▍ | 30991/48008 [1:46:15<54:24, 5.21it/s] 65%|██████▍ | 30992/48008 [1:46:15<58:33, 4.84it/s] 65%|██████▍ | 30993/48008 [1:46:15<56:41, 5.00it/s] 65%|██████▍ | 30994/48008 [1:46:16<55:36, 5.10it/s] 65%|██████▍ | 30995/48008 [1:46:16<54:48, 5.17it/s] 65%|██████▍ | 30996/48008 [1:46:16<54:40, 5.19it/s] 65%|██████▍ | 30997/48008 [1:46:16<54:29, 5.20it/s] 65%|██████▍ | 30998/48008 [1:46:16<54:18, 5.22it/s] 65%|██████▍ | 30999/48008 [1:46:17<55:23, 5.12it/s] 65%|██████▍ | 31000/48008 [1:46:17<54:52, 5.17it/s] {'loss': 0.3001, 'grad_norm': 0.28515625, 'learning_rate': 7.085902349608399e-05, 'epoch': 0.65} 65%|██████▍ | 31000/48008 [1:46:17<54:52, 5.17it/s] 65%|██████▍ | 31001/48008 [1:46:17<54:48, 5.17it/s] 65%|██████▍ | 31002/48008 [1:46:17<54:52, 5.16it/s] 65%|██████▍ | 31003/48008 [1:46:17<54:42, 5.18it/s] 65%|██████▍ | 31004/48008 [1:46:18<54:22, 5.21it/s] 65%|██████▍ | 31005/48008 [1:46:18<54:31, 5.20it/s] 65%|██████▍ | 31006/48008 [1:46:18<54:32, 5.20it/s] 65%|██████▍ | 31007/48008 [1:46:18<54:45, 5.17it/s] 65%|██████▍ | 31008/48008 [1:46:18<54:14, 5.22it/s] 65%|██████▍ | 31009/48008 [1:46:18<54:06, 5.24it/s] 65%|██████▍ | 31010/48008 [1:46:19<54:13, 5.22it/s] 65%|██████▍ | 31011/48008 [1:46:19<58:07, 4.87it/s] 65%|██████▍ | 31012/48008 [1:46:19<56:46, 4.99it/s] 65%|██████▍ | 31013/48008 [1:46:19<55:50, 5.07it/s] 65%|██████▍ | 31014/48008 [1:46:20<56:18, 5.03it/s] 65%|██████▍ | 31015/48008 [1:46:20<56:33, 5.01it/s] 65%|██████▍ | 31016/48008 [1:46:20<56:01, 5.06it/s] 65%|██████▍ | 31017/48008 [1:46:20<55:15, 5.12it/s] 65%|██████▍ | 31018/48008 [1:46:20<56:32, 5.01it/s] 65%|██████▍ | 31019/48008 [1:46:21<1:12:29, 3.91it/s] 65%|██████▍ | 31020/48008 [1:46:21<1:07:06, 4.22it/s] 65%|██████▍ | 31021/48008 [1:46:21<1:03:38, 4.45it/s] 65%|██████▍ | 31022/48008 [1:46:21<1:00:21, 4.69it/s] 65%|██████▍ | 31023/48008 [1:46:21<58:28, 4.84it/s] 65%|██████▍ | 31024/48008 [1:46:22<57:03, 4.96it/s] 65%|██████▍ | 31025/48008 [1:46:22<55:38, 5.09it/s] 65%|██████▍ | 31026/48008 [1:46:22<54:58, 5.15it/s] 65%|██████▍ | 31027/48008 [1:46:22<54:37, 5.18it/s] 65%|██████▍ | 31028/48008 [1:46:22<54:21, 5.21it/s] 65%|██████▍ | 31029/48008 [1:46:23<53:47, 5.26it/s] 65%|██████▍ | 31030/48008 [1:46:23<53:57, 5.24it/s] 65%|██████▍ | 31031/48008 [1:46:23<54:01, 5.24it/s] 65%|██████▍ | 31032/48008 [1:46:23<1:10:38, 4.00it/s] 65%|██████▍ | 31033/48008 [1:46:24<1:05:38, 4.31it/s] 65%|██████▍ | 31034/48008 [1:46:24<1:03:09, 4.48it/s] 65%|██████▍ | 31035/48008 [1:46:24<1:00:27, 4.68it/s] 65%|██████▍ | 31036/48008 [1:46:24<58:03, 4.87it/s] 65%|██████▍ | 31037/48008 [1:46:24<58:36, 4.83it/s] 65%|██████▍ | 31038/48008 [1:46:25<57:40, 4.90it/s] 65%|██████▍ | 31039/48008 [1:46:25<56:04, 5.04it/s] 65%|██████▍ | 31040/48008 [1:46:25<55:02, 5.14it/s] 65%|██████▍ | 31041/48008 [1:46:25<54:53, 5.15it/s] 65%|██████▍ | 31042/48008 [1:46:25<58:35, 4.83it/s] 65%|██████▍ | 31043/48008 [1:46:26<57:05, 4.95it/s] 65%|██████▍ | 31044/48008 [1:46:26<56:03, 5.04it/s] 65%|██████▍ | 31045/48008 [1:46:26<56:20, 5.02it/s] 65%|██████▍ | 31046/48008 [1:46:26<55:12, 5.12it/s] 65%|██████▍ | 31047/48008 [1:46:26<54:49, 5.16it/s] 65%|██████▍ | 31048/48008 [1:46:26<55:28, 5.09it/s] 65%|██████▍ | 31049/48008 [1:46:27<55:02, 5.14it/s] 65%|██████▍ | 31050/48008 [1:46:27<55:22, 5.10it/s] {'loss': 0.3022, 'grad_norm': 0.51953125, 'learning_rate': 7.065072487918681e-05, 'epoch': 0.65} 65%|██████▍ | 31050/48008 [1:46:27<55:22, 5.10it/s] 65%|██████▍ | 31051/48008 [1:46:27<54:54, 5.15it/s] 65%|██████▍ | 31052/48008 [1:46:28<2:22:51, 1.98it/s] 65%|██████▍ | 31053/48008 [1:46:29<1:57:23, 2.41it/s] 65%|██████▍ | 31054/48008 [1:46:29<1:38:08, 2.88it/s] 65%|██████▍ | 31055/48008 [1:46:29<1:34:43, 2.98it/s] 65%|██████▍ | 31056/48008 [1:46:29<1:32:14, 3.06it/s] 65%|██████▍ | 31057/48008 [1:46:30<1:21:48, 3.45it/s] 65%|██████▍ | 31058/48008 [1:46:30<1:13:26, 3.85it/s] 65%|██████▍ | 31059/48008 [1:46:30<1:07:21, 4.19it/s] 65%|██████▍ | 31060/48008 [1:46:30<1:03:17, 4.46it/s] 65%|██████▍ | 31061/48008 [1:46:30<1:00:08, 4.70it/s] 65%|██████▍ | 31062/48008 [1:46:30<58:02, 4.87it/s] 65%|██████▍ | 31063/48008 [1:46:31<57:41, 4.90it/s] 65%|██████▍ | 31064/48008 [1:46:31<57:06, 4.95it/s] 65%|██████▍ | 31065/48008 [1:46:31<56:16, 5.02it/s] 65%|██████▍ | 31066/48008 [1:46:31<57:03, 4.95it/s] 65%|██████▍ | 31067/48008 [1:46:31<57:02, 4.95it/s] 65%|██████▍ | 31068/48008 [1:46:32<1:05:46, 4.29it/s] 65%|██████▍ | 31069/48008 [1:46:32<1:01:52, 4.56it/s] 65%|██████▍ | 31070/48008 [1:46:32<59:19, 4.76it/s] 65%|██████▍ | 31071/48008 [1:46:32<57:41, 4.89it/s] 65%|██████▍ | 31072/48008 [1:46:33<56:16, 5.02it/s] 65%|██████▍ | 31073/48008 [1:46:33<56:41, 4.98it/s] 65%|██████▍ | 31074/48008 [1:46:33<55:48, 5.06it/s] 65%|██████▍ | 31075/48008 [1:46:33<55:22, 5.10it/s] 65%|██████▍ | 31076/48008 [1:46:33<54:46, 5.15it/s] 65%|██████▍ | 31077/48008 [1:46:33<54:43, 5.16it/s] 65%|██████▍ | 31078/48008 [1:46:34<54:14, 5.20it/s] 65%|██████▍ | 31079/48008 [1:46:34<54:02, 5.22it/s] 65%|██████▍ | 31080/48008 [1:46:34<54:00, 5.22it/s] 65%|██████▍ | 31081/48008 [1:46:34<55:59, 5.04it/s] 65%|██████▍ | 31082/48008 [1:46:34<56:21, 5.01it/s] 65%|██████▍ | 31083/48008 [1:46:35<56:34, 4.99it/s] 65%|██████▍ | 31084/48008 [1:46:36<1:53:51, 2.48it/s] 65%|██████▍ | 31085/48008 [1:46:36<1:36:07, 2.93it/s] 65%|██████▍ | 31086/48008 [1:46:36<1:23:39, 3.37it/s] 65%|██████▍ | 31087/48008 [1:46:36<1:15:05, 3.76it/s] 65%|██████▍ | 31088/48008 [1:46:36<1:08:42, 4.10it/s] 65%|██████▍ | 31089/48008 [1:46:37<1:04:08, 4.40it/s] 65%|██████▍ | 31090/48008 [1:46:37<2:06:35, 2.23it/s] 65%|██████▍ | 31091/48008 [1:46:38<1:45:20, 2.68it/s] 65%|██████▍ | 31092/48008 [1:46:38<1:29:52, 3.14it/s] 65%|██████▍ | 31093/48008 [1:46:38<1:19:35, 3.54it/s] 65%|██████▍ | 31094/48008 [1:46:38<1:12:04, 3.91it/s] 65%|██████▍ | 31095/48008 [1:46:38<1:06:48, 4.22it/s] 65%|██████▍ | 31096/48008 [1:46:39<1:55:09, 2.45it/s] 65%|██████▍ | 31097/48008 [1:46:39<1:36:09, 2.93it/s] 65%|██████▍ | 31098/48008 [1:46:40<1:23:16, 3.38it/s] 65%|██████▍ | 31099/48008 [1:46:40<1:14:25, 3.79it/s] 65%|██████▍ | 31100/48008 [1:46:40<1:07:55, 4.15it/s] {'loss': 0.3025, 'grad_norm': 0.1962890625, 'learning_rate': 7.044242626228961e-05, 'epoch': 0.65} 65%|██████▍ | 31100/48008 [1:46:40<1:07:55, 4.15it/s] 65%|██████▍ | 31101/48008 [1:46:40<1:03:42, 4.42it/s] 65%|██████▍ | 31102/48008 [1:46:40<1:00:33, 4.65it/s] 65%|██████▍ | 31103/48008 [1:46:41<58:26, 4.82it/s] 65%|██████▍ | 31104/48008 [1:46:41<57:56, 4.86it/s] 65%|██████▍ | 31105/48008 [1:46:41<55:44, 5.05it/s] 65%|██████▍ | 31106/48008 [1:46:41<55:11, 5.10it/s] 65%|██████▍ | 31107/48008 [1:46:41<54:27, 5.17it/s] 65%|██████▍ | 31108/48008 [1:46:42<53:57, 5.22it/s] 65%|██████▍ | 31109/48008 [1:46:42<53:12, 5.29it/s] 65%|██████▍ | 31110/48008 [1:46:42<53:32, 5.26it/s] 65%|██████▍ | 31111/48008 [1:46:42<54:14, 5.19it/s] 65%|██████▍ | 31112/48008 [1:46:42<54:31, 5.17it/s] 65%|██████▍ | 31113/48008 [1:46:42<54:15, 5.19it/s] 65%|██████▍ | 31114/48008 [1:46:43<54:09, 5.20it/s] 65%|██████▍ | 31115/48008 [1:46:43<54:39, 5.15it/s] 65%|██████▍ | 31116/48008 [1:46:43<54:36, 5.16it/s] 65%|██████▍ | 31117/48008 [1:46:43<54:30, 5.16it/s] 65%|██████▍ | 31118/48008 [1:46:43<54:20, 5.18it/s] 65%|██████▍ | 31119/48008 [1:46:44<54:10, 5.20it/s] 65%|██████▍ | 31120/48008 [1:46:44<53:38, 5.25it/s] 65%|██████▍ | 31121/48008 [1:46:44<53:47, 5.23it/s] 65%|██████▍ | 31122/48008 [1:46:44<53:44, 5.24it/s] 65%|██████▍ | 31123/48008 [1:46:44<53:42, 5.24it/s] 65%|██████▍ | 31124/48008 [1:46:45<53:27, 5.26it/s] 65%|██████▍ | 31125/48008 [1:46:45<54:30, 5.16it/s] 65%|██████▍ | 31126/48008 [1:46:45<1:04:04, 4.39it/s] 65%|██████▍ | 31127/48008 [1:46:45<1:00:39, 4.64it/s] 65%|██████▍ | 31128/48008 [1:46:45<58:35, 4.80it/s] 65%|██████▍ | 31129/48008 [1:46:46<56:52, 4.95it/s] 65%|██████▍ | 31130/48008 [1:46:46<56:17, 5.00it/s] 65%|██████▍ | 31131/48008 [1:46:46<56:08, 5.01it/s] 65%|██████▍ | 31132/48008 [1:46:47<1:55:16, 2.44it/s] 65%|██████▍ | 31133/48008 [1:46:47<1:36:43, 2.91it/s] 65%|██████▍ | 31134/48008 [1:46:48<1:40:21, 2.80it/s] 65%|██████▍ | 31135/48008 [1:46:48<1:26:23, 3.26it/s] 65%|██████▍ | 31136/48008 [1:46:48<1:18:03, 3.60it/s] 65%|██████▍ | 31137/48008 [1:46:48<1:27:11, 3.22it/s] 65%|██████▍ | 31138/48008 [1:46:49<1:17:38, 3.62it/s] 65%|██████▍ | 31139/48008 [1:46:49<1:10:21, 4.00it/s] 65%|██████▍ | 31140/48008 [1:46:49<1:06:15, 4.24it/s] 65%|██████▍ | 31141/48008 [1:46:49<1:02:04, 4.53it/s] 65%|██████▍ | 31142/48008 [1:46:49<59:24, 4.73it/s] 65%|██████▍ | 31143/48008 [1:46:49<57:06, 4.92it/s] 65%|██████▍ | 31144/48008 [1:46:50<55:52, 5.03it/s] 65%|██████▍ | 31145/48008 [1:46:50<59:01, 4.76it/s] 65%|██████▍ | 31146/48008 [1:46:50<57:31, 4.89it/s] 65%|██████▍ | 31147/48008 [1:46:50<57:26, 4.89it/s] 65%|██████▍ | 31148/48008 [1:46:51<57:03, 4.92it/s] 65%|██████▍ | 31149/48008 [1:46:51<56:07, 5.01it/s] 65%|██████▍ | 31150/48008 [1:46:51<55:20, 5.08it/s] {'loss': 0.2991, 'grad_norm': 0.41015625, 'learning_rate': 7.023412764539243e-05, 'epoch': 0.65} 65%|██████▍ | 31150/48008 [1:46:51<55:20, 5.08it/s] 65%|██████▍ | 31151/48008 [1:46:51<56:23, 4.98it/s] 65%|██████▍ | 31152/48008 [1:46:51<56:27, 4.98it/s] 65%|██████▍ | 31153/48008 [1:46:51<55:34, 5.06it/s] 65%|██████▍ | 31154/48008 [1:46:52<55:31, 5.06it/s] 65%|██████▍ | 31155/48008 [1:46:52<55:29, 5.06it/s] 65%|██████▍ | 31156/48008 [1:46:52<54:34, 5.15it/s] 65%|██████▍ | 31157/48008 [1:46:52<58:12, 4.82it/s] 65%|██████▍ | 31158/48008 [1:46:53<57:30, 4.88it/s] 65%|██████▍ | 31159/48008 [1:46:53<57:09, 4.91it/s] 65%|██████▍ | 31160/48008 [1:46:53<56:13, 4.99it/s] 65%|██████▍ | 31161/48008 [1:46:53<54:32, 5.15it/s] 65%|██████▍ | 31162/48008 [1:46:53<55:40, 5.04it/s] 65%|██████▍ | 31163/48008 [1:46:53<54:41, 5.13it/s] 65%|██████▍ | 31164/48008 [1:46:54<54:10, 5.18it/s] 65%|██████▍ | 31165/48008 [1:46:54<53:54, 5.21it/s] 65%|██████▍ | 31166/48008 [1:46:54<54:57, 5.11it/s] 65%|██████▍ | 31167/48008 [1:46:54<55:11, 5.09it/s] 65%|██████▍ | 31168/48008 [1:46:54<54:50, 5.12it/s] 65%|██████▍ | 31169/48008 [1:46:55<54:15, 5.17it/s] 65%|██████▍ | 31170/48008 [1:46:55<57:54, 4.85it/s] 65%|██████▍ | 31171/48008 [1:46:55<56:19, 4.98it/s] 65%|██████▍ | 31172/48008 [1:46:55<55:08, 5.09it/s] 65%|██████▍ | 31173/48008 [1:46:55<54:52, 5.11it/s] 65%|██████▍ | 31174/48008 [1:46:56<58:26, 4.80it/s] 65%|██████▍ | 31175/48008 [1:46:56<57:10, 4.91it/s] 65%|██████▍ | 31176/48008 [1:46:56<55:55, 5.02it/s] 65%|██████▍ | 31177/48008 [1:46:56<55:20, 5.07it/s] 65%|██████▍ | 31178/48008 [1:46:56<54:45, 5.12it/s] 65%|██████▍ | 31179/48008 [1:46:57<55:01, 5.10it/s] 65%|██████▍ | 31180/48008 [1:46:57<58:22, 4.80it/s] 65%|██████▍ | 31181/48008 [1:46:57<56:39, 4.95it/s] 65%|██████▍ | 31182/48008 [1:46:57<55:43, 5.03it/s] 65%|██████▍ | 31183/48008 [1:46:57<55:08, 5.08it/s] 65%|██████▍ | 31184/48008 [1:46:58<54:41, 5.13it/s] 65%|██████▍ | 31185/48008 [1:46:58<54:47, 5.12it/s] 65%|██████▍ | 31186/48008 [1:46:58<54:15, 5.17it/s] 65%|██████▍ | 31187/48008 [1:46:58<53:36, 5.23it/s] 65%|██████▍ | 31188/48008 [1:46:58<53:27, 5.24it/s] 65%|██████▍ | 31189/48008 [1:46:59<53:07, 5.28it/s] 65%|██████▍ | 31190/48008 [1:46:59<53:25, 5.25it/s] 65%|██████▍ | 31191/48008 [1:46:59<53:24, 5.25it/s] 65%|██████▍ | 31192/48008 [1:46:59<54:06, 5.18it/s] 65%|██████▍ | 31193/48008 [1:46:59<53:34, 5.23it/s] 65%|██████▍ | 31194/48008 [1:47:00<54:09, 5.17it/s] 65%|██████▍ | 31195/48008 [1:47:00<1:10:27, 3.98it/s] 65%|██████▍ | 31196/48008 [1:47:00<1:06:22, 4.22it/s] 65%|██████▍ | 31197/48008 [1:47:00<1:03:34, 4.41it/s] 65%|██████▍ | 31198/48008 [1:47:01<1:00:42, 4.62it/s] 65%|██████▍ | 31199/48008 [1:47:01<59:03, 4.74it/s] 65%|██████▍ | 31200/48008 [1:47:01<57:29, 4.87it/s] {'loss': 0.2992, 'grad_norm': 0.138671875, 'learning_rate': 7.002582902849525e-05, 'epoch': 0.65} 65%|██████▍ | 31200/48008 [1:47:01<57:29, 4.87it/s] 65%|██████▍ | 31201/48008 [1:47:01<56:13, 4.98it/s] 65%|██████▍ | 31202/48008 [1:47:01<55:25, 5.05it/s] 65%|██████▍ | 31203/48008 [1:47:01<54:33, 5.13it/s] 65%|██████▍ | 31204/48008 [1:47:02<53:19, 5.25it/s] 65%|██████▍ | 31205/48008 [1:47:02<53:14, 5.26it/s] 65%|██████▌ | 31206/48008 [1:47:02<53:24, 5.24it/s] 65%|██████▌ | 31207/48008 [1:47:02<1:09:47, 4.01it/s] 65%|██████▌ | 31208/48008 [1:47:03<1:21:15, 3.45it/s] 65%|██████▌ | 31209/48008 [1:47:03<1:12:57, 3.84it/s] 65%|██████▌ | 31210/48008 [1:47:03<1:07:28, 4.15it/s] 65%|██████▌ | 31211/48008 [1:47:03<1:03:25, 4.41it/s] 65%|██████▌ | 31212/48008 [1:47:04<1:00:16, 4.64it/s] 65%|██████▌ | 31213/48008 [1:47:04<1:14:34, 3.75it/s] 65%|██████▌ | 31214/48008 [1:47:04<1:08:15, 4.10it/s] 65%|██████▌ | 31215/48008 [1:47:04<1:03:48, 4.39it/s] 65%|██████▌ | 31216/48008 [1:47:05<1:00:42, 4.61it/s] 65%|██████▌ | 31217/48008 [1:47:05<58:24, 4.79it/s] 65%|██████▌ | 31218/48008 [1:47:05<58:20, 4.80it/s] 65%|██████▌ | 31219/48008 [1:47:05<56:56, 4.91it/s] 65%|██████▌ | 31220/48008 [1:47:05<1:05:26, 4.28it/s] 65%|██████▌ | 31221/48008 [1:47:06<1:01:31, 4.55it/s] 65%|██████▌ | 31222/48008 [1:47:06<59:08, 4.73it/s] 65%|██████▌ | 31223/48008 [1:47:06<57:45, 4.84it/s] 65%|██████▌ | 31224/48008 [1:47:06<56:15, 4.97it/s] 65%|██████▌ | 31225/48008 [1:47:07<1:11:49, 3.89it/s] 65%|██████▌ | 31226/48008 [1:47:07<1:06:43, 4.19it/s] 65%|██████▌ | 31227/48008 [1:47:07<1:12:22, 3.86it/s] 65%|██████▌ | 31228/48008 [1:47:07<1:07:35, 4.14it/s] 65%|██████▌ | 31229/48008 [1:47:08<1:03:46, 4.39it/s] 65%|██████▌ | 31230/48008 [1:47:08<1:00:14, 4.64it/s] 65%|██████▌ | 31231/48008 [1:47:08<58:08, 4.81it/s] 65%|██████▌ | 31232/48008 [1:47:08<56:26, 4.95it/s] 65%|██████▌ | 31233/48008 [1:47:08<55:00, 5.08it/s] 65%|██████▌ | 31234/48008 [1:47:08<54:03, 5.17it/s] 65%|██████▌ | 31235/48008 [1:47:09<54:17, 5.15it/s] 65%|██████▌ | 31236/48008 [1:47:09<54:01, 5.17it/s] 65%|██████▌ | 31237/48008 [1:47:09<53:30, 5.22it/s] 65%|██████▌ | 31238/48008 [1:47:09<53:58, 5.18it/s] 65%|██████▌ | 31239/48008 [1:47:09<53:41, 5.20it/s] 65%|██████▌ | 31240/48008 [1:47:10<54:32, 5.12it/s] 65%|██████▌ | 31241/48008 [1:47:10<53:50, 5.19it/s] 65%|██████▌ | 31242/48008 [1:47:10<53:33, 5.22it/s] 65%|██████▌ | 31243/48008 [1:47:10<53:56, 5.18it/s] 65%|██████▌ | 31244/48008 [1:47:10<53:20, 5.24it/s] 65%|██████▌ | 31245/48008 [1:47:11<52:53, 5.28it/s] 65%|██████▌ | 31246/48008 [1:47:11<53:42, 5.20it/s] 65%|██████▌ | 31247/48008 [1:47:11<53:39, 5.21it/s] 65%|██████▌ | 31248/48008 [1:47:11<53:18, 5.24it/s] 65%|██████▌ | 31249/48008 [1:47:11<53:47, 5.19it/s] 65%|██████▌ | 31250/48008 [1:47:12<54:04, 5.17it/s] 65%|██████▌ | 31250/48008 [1:47:12<54:04, 5.17it/s]{'loss': 0.3013, 'grad_norm': 0.1474609375, 'learning_rate': 6.981753041159808e-05, 'epoch': 0.65} 65%|██████▌ | 31251/48008 [1:47:12<54:18, 5.14it/s] 65%|██████▌ | 31252/48008 [1:47:12<54:36, 5.11it/s] 65%|██████▌ | 31253/48008 [1:47:12<54:13, 5.15it/s] 65%|██████▌ | 31254/48008 [1:47:12<53:52, 5.18it/s] 65%|██████▌ | 31255/48008 [1:47:12<53:22, 5.23it/s] 65%|██████▌ | 31256/48008 [1:47:13<53:36, 5.21it/s] 65%|██████▌ | 31257/48008 [1:47:13<54:10, 5.15it/s] 65%|██████▌ | 31258/48008 [1:47:13<53:16, 5.24it/s] 65%|██████▌ | 31259/48008 [1:47:13<52:49, 5.28it/s] 65%|██████▌ | 31260/48008 [1:47:13<53:02, 5.26it/s] 65%|██████▌ | 31261/48008 [1:47:14<53:08, 5.25it/s] 65%|██████▌ | 31262/48008 [1:47:14<54:44, 5.10it/s] 65%|██████▌ | 31263/48008 [1:47:14<54:17, 5.14it/s] 65%|██████▌ | 31264/48008 [1:47:14<54:15, 5.14it/s] 65%|██████▌ | 31265/48008 [1:47:14<54:04, 5.16it/s] 65%|██████▌ | 31266/48008 [1:47:15<54:05, 5.16it/s] 65%|██████▌ | 31267/48008 [1:47:15<53:05, 5.26it/s] 65%|██████▌ | 31268/48008 [1:47:15<52:57, 5.27it/s] 65%|██████▌ | 31269/48008 [1:47:15<52:53, 5.27it/s] 65%|██████▌ | 31270/48008 [1:47:15<52:13, 5.34it/s] 65%|██████▌ | 31271/48008 [1:47:16<53:26, 5.22it/s] 65%|██████▌ | 31272/48008 [1:47:16<53:22, 5.23it/s] 65%|██████▌ | 31273/48008 [1:47:16<53:06, 5.25it/s] 65%|██████▌ | 31274/48008 [1:47:16<53:14, 5.24it/s] 65%|██████▌ | 31275/48008 [1:47:16<53:05, 5.25it/s] 65%|██████▌ | 31276/48008 [1:47:17<53:20, 5.23it/s] 65%|██████▌ | 31277/48008 [1:47:17<53:10, 5.24it/s] 65%|██████▌ | 31278/48008 [1:47:17<52:41, 5.29it/s] 65%|██████▌ | 31279/48008 [1:47:17<1:09:12, 4.03it/s] 65%|██████▌ | 31280/48008 [1:47:17<1:04:01, 4.35it/s] 65%|██████▌ | 31281/48008 [1:47:18<1:00:52, 4.58it/s] 65%|██████▌ | 31282/48008 [1:47:18<58:37, 4.76it/s] 65%|██████▌ | 31283/48008 [1:47:18<57:01, 4.89it/s] 65%|██████▌ | 31284/48008 [1:47:18<55:18, 5.04it/s] 65%|██████▌ | 31285/48008 [1:47:18<54:35, 5.11it/s] 65%|██████▌ | 31286/48008 [1:47:19<54:39, 5.10it/s] 65%|██████▌ | 31287/48008 [1:47:19<1:03:50, 4.36it/s] 65%|██████▌ | 31288/48008 [1:47:19<1:00:45, 4.59it/s] 65%|██████▌ | 31289/48008 [1:47:19<58:02, 4.80it/s] 65%|██████▌ | 31290/48008 [1:47:19<57:10, 4.87it/s] 65%|██████▌ | 31291/48008 [1:47:20<56:58, 4.89it/s] 65%|██████▌ | 31292/48008 [1:47:20<1:05:34, 4.25it/s] 65%|██████▌ | 31293/48008 [1:47:20<1:11:29, 3.90it/s] 65%|██████▌ | 31294/48008 [1:47:21<1:15:45, 3.68it/s] 65%|██████▌ | 31295/48008 [1:47:21<1:08:57, 4.04it/s] 65%|██████▌ | 31296/48008 [1:47:21<1:04:46, 4.30it/s] 65%|██████▌ | 31297/48008 [1:47:21<1:01:29, 4.53it/s] 65%|██████▌ | 31298/48008 [1:47:21<58:34, 4.75it/s] 65%|██████▌ | 31299/48008 [1:47:22<58:02, 4.80it/s] 65%|██████▌ | 31300/48008 [1:47:22<56:41, 4.91it/s] {'loss': 0.3017, 'grad_norm': 0.1337890625, 'learning_rate': 6.96092317947009e-05, 'epoch': 0.65} 65%|██████▌ | 31300/48008 [1:47:22<56:41, 4.91it/s] 65%|██████▌ | 31301/48008 [1:47:22<56:02, 4.97it/s] 65%|██████▌ | 31302/48008 [1:47:22<55:09, 5.05it/s] 65%|██████▌ | 31303/48008 [1:47:22<54:26, 5.11it/s] 65%|██████▌ | 31304/48008 [1:47:23<54:02, 5.15it/s] 65%|██████▌ | 31305/48008 [1:47:23<53:46, 5.18it/s] 65%|██████▌ | 31306/48008 [1:47:23<54:13, 5.13it/s] 65%|██████▌ | 31307/48008 [1:47:23<54:10, 5.14it/s] 65%|██████▌ | 31308/48008 [1:47:23<1:10:04, 3.97it/s] 65%|██████▌ | 31309/48008 [1:47:24<1:05:00, 4.28it/s] 65%|██████▌ | 31310/48008 [1:47:24<1:01:25, 4.53it/s] 65%|██████▌ | 31311/48008 [1:47:24<59:56, 4.64it/s] 65%|██████▌ | 31312/48008 [1:47:24<57:53, 4.81it/s] 65%|██████▌ | 31313/48008 [1:47:25<1:00:24, 4.61it/s] 65%|██████▌ | 31314/48008 [1:47:25<1:14:26, 3.74it/s] 65%|██████▌ | 31315/48008 [1:47:25<1:09:05, 4.03it/s] 65%|██████▌ | 31316/48008 [1:47:25<1:03:54, 4.35it/s] 65%|██████▌ | 31317/48008 [1:47:25<1:00:29, 4.60it/s] 65%|██████▌ | 31318/48008 [1:47:26<58:12, 4.78it/s] 65%|██████▌ | 31319/48008 [1:47:26<57:07, 4.87it/s] 65%|██████▌ | 31320/48008 [1:47:26<57:09, 4.87it/s] 65%|██████▌ | 31321/48008 [1:47:26<56:02, 4.96it/s] 65%|██████▌ | 31322/48008 [1:47:26<57:02, 4.88it/s] 65%|██████▌ | 31323/48008 [1:47:27<56:54, 4.89it/s] 65%|██████▌ | 31324/48008 [1:47:27<55:48, 4.98it/s] 65%|██████▌ | 31325/48008 [1:47:27<54:56, 5.06it/s] 65%|██████▌ | 31326/48008 [1:47:27<54:25, 5.11it/s] 65%|██████▌ | 31327/48008 [1:47:27<54:05, 5.14it/s] 65%|██████▌ | 31328/48008 [1:47:28<53:43, 5.17it/s] 65%|██████▌ | 31329/48008 [1:47:28<1:03:15, 4.39it/s] 65%|██████▌ | 31330/48008 [1:47:28<1:00:20, 4.61it/s] 65%|██████▌ | 31331/48008 [1:47:28<58:21, 4.76it/s] 65%|██████▌ | 31332/48008 [1:47:29<57:49, 4.81it/s] 65%|██████▌ | 31333/48008 [1:47:29<56:19, 4.93it/s] 65%|██████▌ | 31334/48008 [1:47:29<55:42, 4.99it/s] 65%|██████▌ | 31335/48008 [1:47:29<54:48, 5.07it/s] 65%|██████▌ | 31336/48008 [1:47:29<54:34, 5.09it/s] 65%|██████▌ | 31337/48008 [1:47:29<54:18, 5.12it/s] 65%|██████▌ | 31338/48008 [1:47:30<53:37, 5.18it/s] 65%|██████▌ | 31339/48008 [1:47:30<53:21, 5.21it/s] 65%|██████▌ | 31340/48008 [1:47:30<53:22, 5.20it/s] 65%|██████▌ | 31341/48008 [1:47:30<53:29, 5.19it/s] 65%|██████▌ | 31342/48008 [1:47:30<53:20, 5.21it/s] 65%|██████▌ | 31343/48008 [1:47:31<53:00, 5.24it/s] 65%|██████▌ | 31344/48008 [1:47:31<56:43, 4.90it/s] 65%|██████▌ | 31345/48008 [1:47:31<55:05, 5.04it/s] 65%|██████▌ | 31346/48008 [1:47:31<54:41, 5.08it/s] 65%|██████▌ | 31347/48008 [1:47:31<54:14, 5.12it/s] 65%|██████▌ | 31348/48008 [1:47:32<53:44, 5.17it/s] 65%|██████▌ | 31349/48008 [1:47:32<53:21, 5.20it/s] 65%|██████▌ | 31350/48008 [1:47:32<53:18, 5.21it/s] {'loss': 0.3008, 'grad_norm': 0.435546875, 'learning_rate': 6.94009331778037e-05, 'epoch': 0.65} 65%|██████▌ | 31350/48008 [1:47:32<53:18, 5.21it/s] 65%|██████▌ | 31351/48008 [1:47:32<54:15, 5.12it/s] 65%|██████▌ | 31352/48008 [1:47:32<53:51, 5.15it/s] 65%|██████▌ | 31353/48008 [1:47:33<54:43, 5.07it/s] 65%|██████▌ | 31354/48008 [1:47:33<54:07, 5.13it/s] 65%|██████▌ | 31355/48008 [1:47:33<54:00, 5.14it/s] 65%|██████▌ | 31356/48008 [1:47:33<53:19, 5.20it/s] 65%|██████▌ | 31357/48008 [1:47:33<53:15, 5.21it/s] 65%|██████▌ | 31358/48008 [1:47:34<53:11, 5.22it/s] 65%|██████▌ | 31359/48008 [1:47:34<53:26, 5.19it/s] 65%|██████▌ | 31360/48008 [1:47:34<54:08, 5.13it/s] 65%|██████▌ | 31361/48008 [1:47:34<55:01, 5.04it/s] 65%|██████▌ | 31362/48008 [1:47:34<54:21, 5.10it/s] 65%|██████▌ | 31363/48008 [1:47:35<53:59, 5.14it/s] 65%|██████▌ | 31364/48008 [1:47:35<53:52, 5.15it/s] 65%|██████▌ | 31365/48008 [1:47:35<53:20, 5.20it/s] 65%|██████▌ | 31366/48008 [1:47:35<53:11, 5.21it/s] 65%|██████▌ | 31367/48008 [1:47:35<53:10, 5.22it/s] 65%|██████▌ | 31368/48008 [1:47:36<53:09, 5.22it/s] 65%|██████▌ | 31369/48008 [1:47:36<54:03, 5.13it/s] 65%|██████▌ | 31370/48008 [1:47:36<54:37, 5.08it/s] 65%|██████▌ | 31371/48008 [1:47:36<53:44, 5.16it/s] 65%|██████▌ | 31372/48008 [1:47:36<53:16, 5.20it/s] 65%|██████▌ | 31373/48008 [1:47:36<53:11, 5.21it/s] 65%|██████▌ | 31374/48008 [1:47:37<53:26, 5.19it/s] 65%|██████▌ | 31375/48008 [1:47:37<53:22, 5.19it/s] 65%|██████▌ | 31376/48008 [1:47:37<55:10, 5.02it/s] 65%|██████▌ | 31377/48008 [1:47:37<54:25, 5.09it/s] 65%|██████▌ | 31378/48008 [1:47:38<1:03:24, 4.37it/s] 65%|██████▌ | 31379/48008 [1:47:38<1:00:22, 4.59it/s] 65%|██████▌ | 31380/48008 [1:47:38<59:08, 4.69it/s] 65%|██████▌ | 31381/48008 [1:47:38<57:19, 4.83it/s] 65%|██████▌ | 31382/48008 [1:47:39<1:12:24, 3.83it/s] 65%|██████▌ | 31383/48008 [1:47:39<1:06:13, 4.18it/s] 65%|██████▌ | 31384/48008 [1:47:39<1:02:26, 4.44it/s] 65%|██████▌ | 31385/48008 [1:47:39<1:00:05, 4.61it/s] 65%|██████▌ | 31386/48008 [1:47:39<57:41, 4.80it/s] 65%|██████▌ | 31387/48008 [1:47:39<56:22, 4.91it/s] 65%|██████▌ | 31388/48008 [1:47:40<55:01, 5.03it/s] 65%|██████▌ | 31389/48008 [1:47:40<54:54, 5.05it/s] 65%|██████▌ | 31390/48008 [1:47:40<55:20, 5.01it/s] 65%|██████▌ | 31391/48008 [1:47:40<54:48, 5.05it/s] 65%|██████▌ | 31392/48008 [1:47:40<54:38, 5.07it/s] 65%|██████▌ | 31393/48008 [1:47:41<54:03, 5.12it/s] 65%|██████▌ | 31394/48008 [1:47:41<53:50, 5.14it/s] 65%|██████▌ | 31395/48008 [1:47:41<57:30, 4.82it/s] 65%|██████▌ | 31396/48008 [1:47:41<55:31, 4.99it/s] 65%|██████▌ | 31397/48008 [1:47:41<54:36, 5.07it/s] 65%|██████▌ | 31398/48008 [1:47:42<54:59, 5.03it/s] 65%|██████▌ | 31399/48008 [1:47:42<53:56, 5.13it/s] 65%|██████▌ | 31400/48008 [1:47:42<54:28, 5.08it/s] {'loss': 0.2983, 'grad_norm': 0.208984375, 'learning_rate': 6.919263456090652e-05, 'epoch': 0.65} 65%|██████▌ | 31400/48008 [1:47:42<54:28, 5.08it/s] 65%|██████▌ | 31401/48008 [1:47:42<54:06, 5.12it/s] 65%|██████▌ | 31402/48008 [1:47:42<53:46, 5.15it/s] 65%|██████▌ | 31403/48008 [1:47:43<53:31, 5.17it/s] 65%|██████▌ | 31404/48008 [1:47:43<1:02:56, 4.40it/s] 65%|██████▌ | 31405/48008 [1:47:43<59:39, 4.64it/s] 65%|██████▌ | 31406/48008 [1:47:43<58:27, 4.73it/s] 65%|██████▌ | 31407/48008 [1:47:44<56:46, 4.87it/s] 65%|██████▌ | 31408/48008 [1:47:44<55:40, 4.97it/s] 65%|██████▌ | 31409/48008 [1:47:44<54:40, 5.06it/s] 65%|██████▌ | 31410/48008 [1:47:44<54:08, 5.11it/s] 65%|██████▌ | 31411/48008 [1:47:44<53:14, 5.20it/s] 65%|██████▌ | 31412/48008 [1:47:44<53:05, 5.21it/s] 65%|██████▌ | 31413/48008 [1:47:45<52:49, 5.24it/s] 65%|██████▌ | 31414/48008 [1:47:45<53:51, 5.14it/s] 65%|██████▌ | 31415/48008 [1:47:45<53:52, 5.13it/s] 65%|██████▌ | 31416/48008 [1:47:45<54:40, 5.06it/s] 65%|██████▌ | 31417/48008 [1:47:45<53:55, 5.13it/s] 65%|██████▌ | 31418/48008 [1:47:46<53:30, 5.17it/s] 65%|██████▌ | 31419/48008 [1:47:46<53:09, 5.20it/s] 65%|██████▌ | 31420/48008 [1:47:46<54:03, 5.11it/s] 65%|██████▌ | 31421/48008 [1:47:46<53:29, 5.17it/s] 65%|██████▌ | 31422/48008 [1:47:46<53:24, 5.18it/s] 65%|██████▌ | 31423/48008 [1:47:47<53:09, 5.20it/s] 65%|██████▌ | 31424/48008 [1:47:47<56:55, 4.86it/s] 65%|██████▌ | 31425/48008 [1:47:47<55:50, 4.95it/s] 65%|██████▌ | 31426/48008 [1:47:47<56:01, 4.93it/s] 65%|██████▌ | 31427/48008 [1:47:47<55:20, 4.99it/s] 65%|██████▌ | 31428/48008 [1:47:48<1:04:19, 4.30it/s] 65%|██████▌ | 31429/48008 [1:47:48<1:01:17, 4.51it/s] 65%|██████▌ | 31430/48008 [1:47:48<1:00:13, 4.59it/s] 65%|██████▌ | 31431/48008 [1:47:48<58:47, 4.70it/s] 65%|██████▌ | 31432/48008 [1:47:49<56:45, 4.87it/s] 65%|██████▌ | 31433/48008 [1:47:49<55:01, 5.02it/s] 65%|██████▌ | 31434/48008 [1:47:49<54:44, 5.05it/s] 65%|██████▌ | 31435/48008 [1:47:49<54:21, 5.08it/s] 65%|██████▌ | 31436/48008 [1:47:49<53:37, 5.15it/s] 65%|██████▌ | 31437/48008 [1:47:50<1:02:47, 4.40it/s] 65%|██████▌ | 31438/48008 [1:47:50<1:00:16, 4.58it/s] 65%|██████▌ | 31439/48008 [1:47:50<58:34, 4.71it/s] 65%|██████▌ | 31440/48008 [1:47:50<57:27, 4.81it/s] 65%|██████▌ | 31441/48008 [1:47:50<56:06, 4.92it/s] 65%|██████▌ | 31442/48008 [1:47:51<1:04:45, 4.26it/s] 65%|██████▌ | 31443/48008 [1:47:51<1:01:14, 4.51it/s] 65%|██████▌ | 31444/48008 [1:47:51<58:29, 4.72it/s] 65%|██████▌ | 31445/48008 [1:47:51<56:19, 4.90it/s] 66%|██████▌ | 31446/48008 [1:47:51<54:56, 5.02it/s] 66%|██████▌ | 31447/48008 [1:47:52<55:43, 4.95it/s] 66%|██████▌ | 31448/48008 [1:47:52<54:55, 5.03it/s] 66%|██████▌ | 31449/48008 [1:47:52<53:52, 5.12it/s] 66%|██████▌ | 31450/48008 [1:47:52<53:34, 5.15it/s] {'loss': 0.3009, 'grad_norm': 0.2138671875, 'learning_rate': 6.898433594400933e-05, 'epoch': 0.66} 66%|██████▌ | 31450/48008 [1:47:52<53:34, 5.15it/s] 66%|██████▌ | 31451/48008 [1:47:52<54:23, 5.07it/s] 66%|██████▌ | 31452/48008 [1:47:53<54:16, 5.08it/s] 66%|██████▌ | 31453/48008 [1:47:53<53:53, 5.12it/s] 66%|██████▌ | 31454/48008 [1:47:53<53:11, 5.19it/s] 66%|██████▌ | 31455/48008 [1:47:53<52:43, 5.23it/s] 66%|██████▌ | 31456/48008 [1:47:53<53:40, 5.14it/s] 66%|██████▌ | 31457/48008 [1:47:54<53:29, 5.16it/s] 66%|██████▌ | 31458/48008 [1:47:54<53:29, 5.16it/s] 66%|██████▌ | 31459/48008 [1:47:54<53:31, 5.15it/s] 66%|██████▌ | 31460/48008 [1:47:54<53:00, 5.20it/s] 66%|██████▌ | 31461/48008 [1:47:54<53:21, 5.17it/s] 66%|██████▌ | 31462/48008 [1:47:55<52:43, 5.23it/s] 66%|██████▌ | 31463/48008 [1:47:55<52:27, 5.26it/s] 66%|██████▌ | 31464/48008 [1:47:55<52:14, 5.28it/s] 66%|██████▌ | 31465/48008 [1:47:55<52:26, 5.26it/s] 66%|██████▌ | 31466/48008 [1:47:55<52:51, 5.22it/s] 66%|██████▌ | 31467/48008 [1:47:56<52:50, 5.22it/s] 66%|██████▌ | 31468/48008 [1:47:56<53:41, 5.13it/s] 66%|██████▌ | 31469/48008 [1:47:56<53:03, 5.19it/s] 66%|██████▌ | 31470/48008 [1:47:56<53:48, 5.12it/s] 66%|██████▌ | 31471/48008 [1:47:56<53:39, 5.14it/s] 66%|██████▌ | 31472/48008 [1:47:56<53:02, 5.20it/s] 66%|██████▌ | 31473/48008 [1:47:57<53:28, 5.15it/s] 66%|██████▌ | 31474/48008 [1:47:57<53:16, 5.17it/s] 66%|██████▌ | 31475/48008 [1:47:57<54:04, 5.10it/s] 66%|██████▌ | 31476/48008 [1:47:57<53:43, 5.13it/s] 66%|██████▌ | 31477/48008 [1:47:57<54:19, 5.07it/s] 66%|██████▌ | 31478/48008 [1:47:58<53:39, 5.13it/s] 66%|██████▌ | 31479/48008 [1:47:58<1:02:46, 4.39it/s] 66%|██████▌ | 31480/48008 [1:47:58<59:18, 4.64it/s] 66%|██████▌ | 31481/48008 [1:47:58<57:13, 4.81it/s] 66%|██████▌ | 31482/48008 [1:47:59<57:17, 4.81it/s] 66%|██████▌ | 31483/48008 [1:47:59<57:22, 4.80it/s] 66%|██████▌ | 31484/48008 [1:47:59<55:58, 4.92it/s] 66%|██████▌ | 31485/48008 [1:47:59<54:34, 5.05it/s] 66%|██████▌ | 31486/48008 [1:47:59<54:31, 5.05it/s] 66%|██████▌ | 31487/48008 [1:48:00<54:02, 5.09it/s] 66%|██████▌ | 31488/48008 [1:48:00<53:09, 5.18it/s] 66%|██████▌ | 31489/48008 [1:48:00<52:44, 5.22it/s] 66%|██████▌ | 31490/48008 [1:48:00<52:11, 5.27it/s] 66%|██████▌ | 31491/48008 [1:48:00<51:52, 5.31it/s] 66%|██████▌ | 31492/48008 [1:48:00<51:56, 5.30it/s] 66%|██████▌ | 31493/48008 [1:48:01<52:22, 5.25it/s] 66%|██████▌ | 31494/48008 [1:48:01<53:12, 5.17it/s] 66%|██████▌ | 31495/48008 [1:48:01<1:09:03, 3.99it/s] 66%|██████▌ | 31496/48008 [1:48:01<1:04:45, 4.25it/s] 66%|██████▌ | 31497/48008 [1:48:02<1:00:49, 4.52it/s] 66%|██████▌ | 31498/48008 [1:48:02<58:16, 4.72it/s] 66%|██████▌ | 31499/48008 [1:48:02<56:33, 4.87it/s] 66%|██████▌ | 31500/48008 [1:48:02<55:54, 4.92it/s] {'loss': 0.3007, 'grad_norm': 0.177734375, 'learning_rate': 6.877603732711215e-05, 'epoch': 0.66} 66%|██████▌ | 31500/48008 [1:48:02<55:54, 4.92it/s] 66%|██████▌ | 31501/48008 [1:48:02<55:04, 5.00it/s] 66%|██████▌ | 31502/48008 [1:48:03<54:43, 5.03it/s] 66%|██████▌ | 31503/48008 [1:48:03<54:18, 5.06it/s] 66%|██████▌ | 31504/48008 [1:48:03<53:30, 5.14it/s] 66%|██████▌ | 31505/48008 [1:48:03<54:10, 5.08it/s] 66%|██████▌ | 31506/48008 [1:48:03<53:50, 5.11it/s] 66%|██████▌ | 31507/48008 [1:48:04<53:06, 5.18it/s] 66%|██████▌ | 31508/48008 [1:48:04<52:54, 5.20it/s] 66%|██████▌ | 31509/48008 [1:48:04<52:30, 5.24it/s] 66%|██████▌ | 31510/48008 [1:48:04<52:13, 5.26it/s] 66%|██████▌ | 31511/48008 [1:48:04<52:08, 5.27it/s] 66%|██████▌ | 31512/48008 [1:48:05<52:30, 5.24it/s] 66%|██████▌ | 31513/48008 [1:48:05<53:09, 5.17it/s] 66%|██████▌ | 31514/48008 [1:48:05<52:21, 5.25it/s] 66%|██████▌ | 31515/48008 [1:48:05<52:33, 5.23it/s] 66%|██████▌ | 31516/48008 [1:48:05<52:43, 5.21it/s] 66%|██████▌ | 31517/48008 [1:48:05<52:45, 5.21it/s] 66%|██████▌ | 31518/48008 [1:48:06<52:19, 5.25it/s] 66%|██████▌ | 31519/48008 [1:48:06<53:23, 5.15it/s] 66%|██████▌ | 31520/48008 [1:48:06<57:06, 4.81it/s] 66%|██████▌ | 31521/48008 [1:48:06<55:48, 4.92it/s] 66%|██████▌ | 31522/48008 [1:48:07<58:35, 4.69it/s] 66%|██████▌ | 31523/48008 [1:48:07<57:52, 4.75it/s] 66%|██████▌ | 31524/48008 [1:48:07<56:17, 4.88it/s] 66%|██████▌ | 31525/48008 [1:48:07<55:42, 4.93it/s] 66%|██████▌ | 31526/48008 [1:48:07<55:01, 4.99it/s] 66%|██████▌ | 31527/48008 [1:48:07<54:17, 5.06it/s] 66%|██████▌ | 31528/48008 [1:48:08<53:51, 5.10it/s] 66%|██████▌ | 31529/48008 [1:48:08<52:55, 5.19it/s] 66%|██████▌ | 31530/48008 [1:48:08<52:51, 5.20it/s] 66%|██████▌ | 31531/48008 [1:48:08<52:58, 5.18it/s] 66%|██████▌ | 31532/48008 [1:48:08<52:19, 5.25it/s] 66%|██████▌ | 31533/48008 [1:48:09<53:23, 5.14it/s] 66%|██████▌ | 31534/48008 [1:48:09<53:14, 5.16it/s] 66%|██████▌ | 31535/48008 [1:48:09<52:53, 5.19it/s] 66%|██████▌ | 31536/48008 [1:48:09<52:56, 5.19it/s] 66%|██████▌ | 31537/48008 [1:48:09<52:59, 5.18it/s] 66%|██████▌ | 31538/48008 [1:48:10<52:57, 5.18it/s] 66%|██████▌ | 31539/48008 [1:48:10<52:23, 5.24it/s] 66%|██████▌ | 31540/48008 [1:48:10<52:17, 5.25it/s] 66%|██████▌ | 31541/48008 [1:48:10<52:29, 5.23it/s] 66%|██████▌ | 31542/48008 [1:48:10<52:19, 5.24it/s] 66%|██████▌ | 31543/48008 [1:48:11<52:41, 5.21it/s] 66%|██████▌ | 31544/48008 [1:48:11<53:30, 5.13it/s] 66%|██████▌ | 31545/48008 [1:48:11<53:23, 5.14it/s] 66%|██████▌ | 31546/48008 [1:48:11<53:11, 5.16it/s] 66%|██████▌ | 31547/48008 [1:48:11<53:53, 5.09it/s] 66%|██████▌ | 31548/48008 [1:48:12<53:19, 5.14it/s] 66%|██████▌ | 31549/48008 [1:48:12<54:40, 5.02it/s] 66%|██████▌ | 31550/48008 [1:48:12<53:56, 5.09it/s] {'loss': 0.3027, 'grad_norm': 0.208984375, 'learning_rate': 6.856773871021497e-05, 'epoch': 0.66} 66%|██████▌ | 31550/48008 [1:48:12<53:56, 5.09it/s] 66%|██████▌ | 31551/48008 [1:48:12<54:26, 5.04it/s] 66%|██████▌ | 31552/48008 [1:48:12<53:51, 5.09it/s] 66%|██████▌ | 31553/48008 [1:48:13<53:23, 5.14it/s] 66%|██████▌ | 31554/48008 [1:48:13<53:08, 5.16it/s] 66%|██████▌ | 31555/48008 [1:48:13<1:09:04, 3.97it/s] 66%|██████▌ | 31556/48008 [1:48:13<1:05:02, 4.22it/s] 66%|██████▌ | 31557/48008 [1:48:14<1:01:02, 4.49it/s] 66%|██████▌ | 31558/48008 [1:48:14<58:10, 4.71it/s] 66%|██████▌ | 31559/48008 [1:48:14<1:00:07, 4.56it/s] 66%|██████▌ | 31560/48008 [1:48:14<58:54, 4.65it/s] 66%|██████▌ | 31561/48008 [1:48:14<56:29, 4.85it/s] 66%|██████▌ | 31562/48008 [1:48:15<55:07, 4.97it/s] 66%|██████▌ | 31563/48008 [1:48:15<54:36, 5.02it/s] 66%|██████▌ | 31564/48008 [1:48:15<54:01, 5.07it/s] 66%|██████▌ | 31565/48008 [1:48:15<54:11, 5.06it/s] 66%|██████▌ | 31566/48008 [1:48:15<52:42, 5.20it/s] 66%|██████▌ | 31567/48008 [1:48:15<52:47, 5.19it/s] 66%|██████▌ | 31568/48008 [1:48:16<53:46, 5.10it/s] 66%|██████▌ | 31569/48008 [1:48:16<54:14, 5.05it/s] 66%|██████▌ | 31570/48008 [1:48:16<53:20, 5.14it/s] 66%|██████▌ | 31571/48008 [1:48:16<52:33, 5.21it/s] 66%|██████▌ | 31572/48008 [1:48:16<53:19, 5.14it/s] 66%|██████▌ | 31573/48008 [1:48:17<53:11, 5.15it/s] 66%|██████▌ | 31574/48008 [1:48:17<52:58, 5.17it/s] 66%|██████▌ | 31575/48008 [1:48:17<52:41, 5.20it/s] 66%|██████▌ | 31576/48008 [1:48:17<1:08:34, 3.99it/s] 66%|██████▌ | 31577/48008 [1:48:18<1:03:58, 4.28it/s] 66%|██████▌ | 31578/48008 [1:48:18<1:01:24, 4.46it/s] 66%|██████▌ | 31579/48008 [1:48:18<58:49, 4.65it/s] 66%|██████▌ | 31580/48008 [1:48:18<57:28, 4.76it/s] 66%|██████▌ | 31581/48008 [1:48:18<56:25, 4.85it/s] 66%|██████▌ | 31582/48008 [1:48:19<55:48, 4.91it/s] 66%|██████▌ | 31583/48008 [1:48:19<55:17, 4.95it/s] 66%|██████▌ | 31584/48008 [1:48:19<54:42, 5.00it/s] 66%|██████▌ | 31585/48008 [1:48:19<55:01, 4.97it/s] 66%|██████▌ | 31586/48008 [1:48:19<54:05, 5.06it/s] 66%|██████▌ | 31587/48008 [1:48:20<55:06, 4.97it/s] 66%|██████▌ | 31588/48008 [1:48:20<55:12, 4.96it/s] 66%|██████▌ | 31589/48008 [1:48:20<54:21, 5.03it/s] 66%|██████▌ | 31590/48008 [1:48:20<53:52, 5.08it/s] 66%|██████▌ | 31591/48008 [1:48:20<53:20, 5.13it/s] 66%|██████▌ | 31592/48008 [1:48:21<52:46, 5.18it/s] 66%|██████▌ | 31593/48008 [1:48:21<53:04, 5.15it/s] 66%|██████▌ | 31594/48008 [1:48:21<54:32, 5.02it/s] 66%|██████▌ | 31595/48008 [1:48:21<53:22, 5.12it/s] 66%|██████▌ | 31596/48008 [1:48:21<53:33, 5.11it/s] 66%|██████▌ | 31597/48008 [1:48:22<1:02:42, 4.36it/s] 66%|██████▌ | 31598/48008 [1:48:22<1:00:33, 4.52it/s] 66%|██████▌ | 31599/48008 [1:48:22<58:33, 4.67it/s] 66%|██████▌ | 31600/48008 [1:48:22<56:53, 4.81it/s] {'loss': 0.3011, 'grad_norm': 0.162109375, 'learning_rate': 6.835944009331779e-05, 'epoch': 0.66} 66%|██████▌ | 31600/48008 [1:48:22<56:53, 4.81it/s] 66%|██████▌ | 31601/48008 [1:48:22<59:22, 4.61it/s] 66%|██████▌ | 31602/48008 [1:48:23<56:48, 4.81it/s] 66%|██████▌ | 31603/48008 [1:48:23<57:00, 4.80it/s] 66%|██████▌ | 31604/48008 [1:48:23<59:24, 4.60it/s] 66%|██████▌ | 31605/48008 [1:48:23<57:02, 4.79it/s] 66%|██████▌ | 31606/48008 [1:48:23<55:03, 4.97it/s] 66%|██████▌ | 31607/48008 [1:48:24<54:11, 5.04it/s] 66%|██████▌ | 31608/48008 [1:48:24<54:01, 5.06it/s] 66%|██████▌ | 31609/48008 [1:48:24<53:50, 5.08it/s] 66%|██████▌ | 31610/48008 [1:48:24<54:25, 5.02it/s] 66%|██████▌ | 31611/48008 [1:48:24<53:21, 5.12it/s] 66%|██████▌ | 31612/48008 [1:48:25<1:50:48, 2.47it/s] 66%|██████▌ | 31613/48008 [1:48:26<1:33:08, 2.93it/s] 66%|██████▌ | 31614/48008 [1:48:26<1:20:35, 3.39it/s] 66%|██████▌ | 31615/48008 [1:48:26<1:11:38, 3.81it/s] 66%|██████▌ | 31616/48008 [1:48:26<1:06:15, 4.12it/s] 66%|██████▌ | 31617/48008 [1:48:26<1:02:00, 4.41it/s] 66%|██████▌ | 31618/48008 [1:48:27<59:11, 4.61it/s] 66%|██████▌ | 31619/48008 [1:48:27<57:11, 4.78it/s] 66%|██████▌ | 31620/48008 [1:48:27<57:17, 4.77it/s] 66%|██████▌ | 31621/48008 [1:48:27<56:19, 4.85it/s] 66%|██████▌ | 31622/48008 [1:48:27<54:35, 5.00it/s] 66%|██████▌ | 31623/48008 [1:48:27<53:51, 5.07it/s] 66%|██████▌ | 31624/48008 [1:48:28<53:26, 5.11it/s] 66%|██████▌ | 31625/48008 [1:48:28<52:57, 5.16it/s] 66%|██████▌ | 31626/48008 [1:48:28<52:32, 5.20it/s] 66%|██████▌ | 31627/48008 [1:48:28<53:57, 5.06it/s] 66%|██████▌ | 31628/48008 [1:48:28<54:08, 5.04it/s] 66%|██████▌ | 31629/48008 [1:48:29<57:29, 4.75it/s] 66%|██████▌ | 31630/48008 [1:48:29<55:56, 4.88it/s] 66%|██████▌ | 31631/48008 [1:48:29<54:29, 5.01it/s] 66%|██████▌ | 31632/48008 [1:48:29<53:40, 5.08it/s] 66%|██████▌ | 31633/48008 [1:48:30<1:02:34, 4.36it/s] 66%|██████▌ | 31634/48008 [1:48:30<1:00:16, 4.53it/s] 66%|██████▌ | 31635/48008 [1:48:30<1:01:41, 4.42it/s] 66%|██████▌ | 31636/48008 [1:48:30<58:57, 4.63it/s] 66%|██████▌ | 31637/48008 [1:48:30<56:14, 4.85it/s] 66%|██████▌ | 31638/48008 [1:48:31<55:13, 4.94it/s] 66%|██████▌ | 31639/48008 [1:48:31<54:14, 5.03it/s] 66%|██████▌ | 31640/48008 [1:48:31<53:18, 5.12it/s] 66%|██████▌ | 31641/48008 [1:48:31<52:48, 5.17it/s] 66%|██████▌ | 31642/48008 [1:48:31<52:39, 5.18it/s] 66%|██████▌ | 31643/48008 [1:48:32<53:01, 5.14it/s] 66%|██████▌ | 31644/48008 [1:48:32<52:24, 5.20it/s] 66%|██████▌ | 31645/48008 [1:48:32<52:38, 5.18it/s] 66%|██████▌ | 31646/48008 [1:48:32<53:32, 5.09it/s] 66%|██████▌ | 31647/48008 [1:48:32<53:04, 5.14it/s] 66%|██████▌ | 31648/48008 [1:48:33<52:53, 5.16it/s] 66%|██████▌ | 31649/48008 [1:48:33<53:06, 5.13it/s] 66%|██████▌ | 31650/48008 [1:48:33<52:59, 5.14it/s] {'loss': 0.3011, 'grad_norm': 1.0, 'learning_rate': 6.81511414764206e-05, 'epoch': 0.66} 66%|██████▌ | 31650/48008 [1:48:33<52:59, 5.14it/s] 66%|██████▌ | 31651/48008 [1:48:33<53:54, 5.06it/s] 66%|██████▌ | 31652/48008 [1:48:33<53:28, 5.10it/s] 66%|██████▌ | 31653/48008 [1:48:33<53:09, 5.13it/s] 66%|██████▌ | 31654/48008 [1:48:34<52:50, 5.16it/s] 66%|██████▌ | 31655/48008 [1:48:34<53:00, 5.14it/s] 66%|██████▌ | 31656/48008 [1:48:34<1:02:08, 4.39it/s] 66%|██████▌ | 31657/48008 [1:48:35<1:15:13, 3.62it/s] 66%|██████▌ | 31658/48008 [1:48:35<1:08:15, 3.99it/s] 66%|██████▌ | 31659/48008 [1:48:35<1:03:16, 4.31it/s] 66%|██████▌ | 31660/48008 [1:48:35<59:59, 4.54it/s] 66%|██████▌ | 31661/48008 [1:48:35<58:00, 4.70it/s] 66%|██████▌ | 31662/48008 [1:48:36<56:49, 4.79it/s] 66%|██████▌ | 31663/48008 [1:48:36<55:27, 4.91it/s] 66%|██████▌ | 31664/48008 [1:48:36<54:33, 4.99it/s] 66%|██████▌ | 31665/48008 [1:48:37<2:20:52, 1.93it/s] 66%|██████▌ | 31666/48008 [1:48:37<2:03:59, 2.20it/s] 66%|██████▌ | 31667/48008 [1:48:38<1:46:29, 2.56it/s] 66%|██████▌ | 31668/48008 [1:48:38<1:30:02, 3.02it/s] 66%|██████▌ | 31669/48008 [1:48:38<1:18:32, 3.47it/s] 66%|██████▌ | 31670/48008 [1:48:38<1:10:41, 3.85it/s] 66%|██████▌ | 31671/48008 [1:48:38<1:04:46, 4.20it/s] 66%|██████▌ | 31672/48008 [1:48:39<1:17:09, 3.53it/s] 66%|██████▌ | 31673/48008 [1:48:39<1:09:24, 3.92it/s] 66%|██████▌ | 31674/48008 [1:48:39<1:04:10, 4.24it/s] 66%|██████▌ | 31675/48008 [1:48:39<1:04:27, 4.22it/s] 66%|██████▌ | 31676/48008 [1:48:40<1:16:46, 3.55it/s] 66%|██████▌ | 31677/48008 [1:48:40<1:09:38, 3.91it/s] 66%|██████▌ | 31678/48008 [1:48:40<1:03:58, 4.25it/s] 66%|██████▌ | 31679/48008 [1:48:40<1:01:03, 4.46it/s] 66%|██████▌ | 31680/48008 [1:48:41<58:47, 4.63it/s] 66%|██████▌ | 31681/48008 [1:48:41<57:00, 4.77it/s] 66%|██████▌ | 31682/48008 [1:48:41<55:35, 4.90it/s] 66%|██████▌ | 31683/48008 [1:48:41<54:11, 5.02it/s] 66%|██████▌ | 31684/48008 [1:48:41<55:04, 4.94it/s] 66%|██████▌ | 31685/48008 [1:48:42<54:07, 5.03it/s] 66%|██████▌ | 31686/48008 [1:48:42<54:45, 4.97it/s] 66%|██████▌ | 31687/48008 [1:48:42<54:29, 4.99it/s] 66%|██████▌ | 31688/48008 [1:48:42<53:39, 5.07it/s] 66%|██████▌ | 31689/48008 [1:48:42<53:10, 5.12it/s] 66%|██████▌ | 31690/48008 [1:48:43<53:42, 5.06it/s] 66%|██████▌ | 31691/48008 [1:48:43<54:01, 5.03it/s] 66%|██████▌ | 31692/48008 [1:48:43<53:35, 5.07it/s] 66%|██████▌ | 31693/48008 [1:48:43<53:10, 5.11it/s] 66%|██████▌ | 31694/48008 [1:48:43<52:48, 5.15it/s] 66%|██████▌ | 31695/48008 [1:48:44<53:14, 5.11it/s] 66%|██████▌ | 31696/48008 [1:48:44<53:49, 5.05it/s] 66%|██████▌ | 31697/48008 [1:48:44<53:14, 5.11it/s] 66%|██████▌ | 31698/48008 [1:48:44<52:43, 5.16it/s] 66%|██████▌ | 31699/48008 [1:48:44<53:02, 5.12it/s] 66%|██████▌ | 31700/48008 [1:48:45<52:50, 5.14it/s] {'loss': 0.2965, 'grad_norm': 0.32421875, 'learning_rate': 6.794284285952341e-05, 'epoch': 0.66} 66%|██████▌ | 31700/48008 [1:48:45<52:50, 5.14it/s] 66%|██████▌ | 31701/48008 [1:48:45<52:38, 5.16it/s] 66%|██████▌ | 31702/48008 [1:48:45<52:32, 5.17it/s] 66%|██████▌ | 31703/48008 [1:48:45<53:38, 5.07it/s] 66%|██████▌ | 31704/48008 [1:48:45<53:17, 5.10it/s] 66%|██████▌ | 31705/48008 [1:48:46<53:01, 5.12it/s] 66%|██████▌ | 31706/48008 [1:48:46<52:39, 5.16it/s] 66%|██████▌ | 31707/48008 [1:48:46<52:22, 5.19it/s] 66%|██████▌ | 31708/48008 [1:48:46<52:03, 5.22it/s] 66%|██████▌ | 31709/48008 [1:48:46<51:47, 5.25it/s] 66%|██████▌ | 31710/48008 [1:48:47<51:49, 5.24it/s] 66%|██████▌ | 31711/48008 [1:48:47<55:40, 4.88it/s] 66%|██████▌ | 31712/48008 [1:48:47<54:26, 4.99it/s] 66%|██████▌ | 31713/48008 [1:48:47<53:25, 5.08it/s] 66%|██████▌ | 31714/48008 [1:48:47<54:00, 5.03it/s] 66%|██████▌ | 31715/48008 [1:48:48<54:31, 4.98it/s] 66%|██████▌ | 31716/48008 [1:48:48<54:16, 5.00it/s] 66%|██████▌ | 31717/48008 [1:48:48<53:56, 5.03it/s] 66%|██████▌ | 31718/48008 [1:48:48<52:59, 5.12it/s] 66%|██████▌ | 31719/48008 [1:48:48<53:04, 5.12it/s] 66%|██████▌ | 31720/48008 [1:48:49<56:26, 4.81it/s] 66%|██████▌ | 31721/48008 [1:48:49<54:45, 4.96it/s] 66%|██████▌ | 31722/48008 [1:48:49<54:03, 5.02it/s] 66%|██████▌ | 31723/48008 [1:48:49<53:37, 5.06it/s] 66%|██████▌ | 31724/48008 [1:48:49<53:30, 5.07it/s] 66%|██████▌ | 31725/48008 [1:48:50<53:05, 5.11it/s] 66%|██████▌ | 31726/48008 [1:48:50<52:31, 5.17it/s] 66%|██████▌ | 31727/48008 [1:48:50<52:20, 5.18it/s] 66%|██████▌ | 31728/48008 [1:48:50<52:01, 5.22it/s] 66%|██████▌ | 31729/48008 [1:48:50<52:06, 5.21it/s] 66%|██████▌ | 31730/48008 [1:48:50<51:45, 5.24it/s] 66%|██████▌ | 31731/48008 [1:48:51<51:30, 5.27it/s] 66%|██████▌ | 31732/48008 [1:48:51<51:17, 5.29it/s] 66%|██████▌ | 31733/48008 [1:48:51<52:05, 5.21it/s] 66%|██████▌ | 31734/48008 [1:48:51<51:37, 5.25it/s] 66%|██████▌ | 31735/48008 [1:48:51<51:16, 5.29it/s] 66%|██████▌ | 31736/48008 [1:48:52<1:00:41, 4.47it/s] 66%|██████▌ | 31737/48008 [1:48:52<57:38, 4.70it/s] 66%|██████▌ | 31738/48008 [1:48:52<55:45, 4.86it/s] 66%|██████▌ | 31739/48008 [1:48:52<55:26, 4.89it/s] 66%|██████▌ | 31740/48008 [1:48:52<54:44, 4.95it/s] 66%|██████▌ | 31741/48008 [1:48:53<53:31, 5.07it/s] 66%|██████▌ | 31742/48008 [1:48:53<52:49, 5.13it/s] 66%|██████▌ | 31743/48008 [1:48:53<1:01:53, 4.38it/s] 66%|██████▌ | 31744/48008 [1:48:53<58:35, 4.63it/s] 66%|██████▌ | 31745/48008 [1:48:54<56:12, 4.82it/s] 66%|██████▌ | 31746/48008 [1:48:54<54:31, 4.97it/s] 66%|██████▌ | 31747/48008 [1:48:54<54:19, 4.99it/s] 66%|██████▌ | 31748/48008 [1:48:54<53:30, 5.06it/s] 66%|██████▌ | 31749/48008 [1:48:54<52:54, 5.12it/s] 66%|██████▌ | 31750/48008 [1:48:55<53:09, 5.10it/s] {'loss': 0.3014, 'grad_norm': 0.1953125, 'learning_rate': 6.773454424262623e-05, 'epoch': 0.66} 66%|██████▌ | 31750/48008 [1:48:55<53:09, 5.10it/s] 66%|██████▌ | 31751/48008 [1:48:55<52:57, 5.12it/s] 66%|██████▌ | 31752/48008 [1:48:55<52:35, 5.15it/s] 66%|██████▌ | 31753/48008 [1:48:55<53:39, 5.05it/s] 66%|██████▌ | 31754/48008 [1:48:55<53:07, 5.10it/s] 66%|██████▌ | 31755/48008 [1:48:56<56:30, 4.79it/s] 66%|██████▌ | 31756/48008 [1:48:56<55:27, 4.88it/s] 66%|██████▌ | 31757/48008 [1:48:56<57:56, 4.67it/s] 66%|██████▌ | 31758/48008 [1:48:56<57:06, 4.74it/s] 66%|██████▌ | 31759/48008 [1:48:56<55:58, 4.84it/s] 66%|██████▌ | 31760/48008 [1:48:57<55:06, 4.91it/s] 66%|██████▌ | 31761/48008 [1:48:57<54:12, 5.00it/s] 66%|██████▌ | 31762/48008 [1:48:57<53:26, 5.07it/s] 66%|██████▌ | 31763/48008 [1:48:57<52:39, 5.14it/s] 66%|██████▌ | 31764/48008 [1:48:57<52:32, 5.15it/s] 66%|██████▌ | 31765/48008 [1:48:58<52:12, 5.19it/s] 66%|██████▌ | 31766/48008 [1:48:58<52:34, 5.15it/s] 66%|██████▌ | 31767/48008 [1:48:58<1:08:02, 3.98it/s] 66%|██████▌ | 31768/48008 [1:48:58<1:02:52, 4.30it/s] 66%|██████▌ | 31769/48008 [1:48:59<1:08:47, 3.93it/s] 66%|██████▌ | 31770/48008 [1:48:59<1:03:44, 4.25it/s] 66%|██████▌ | 31771/48008 [1:48:59<1:03:49, 4.24it/s] 66%|██████▌ | 31772/48008 [1:48:59<1:01:10, 4.42it/s] 66%|██████▌ | 31773/48008 [1:48:59<57:55, 4.67it/s] 66%|██████▌ | 31774/48008 [1:49:00<56:56, 4.75it/s] 66%|██████▌ | 31775/48008 [1:49:00<55:19, 4.89it/s] 66%|██████▌ | 31776/48008 [1:49:00<54:11, 4.99it/s] 66%|██████▌ | 31777/48008 [1:49:00<53:21, 5.07it/s] 66%|██████▌ | 31778/48008 [1:49:00<53:11, 5.09it/s] 66%|██████▌ | 31779/48008 [1:49:01<52:46, 5.13it/s] 66%|██████▌ | 31780/48008 [1:49:01<52:07, 5.19it/s] 66%|██████▌ | 31781/48008 [1:49:01<52:36, 5.14it/s] 66%|██████▌ | 31782/48008 [1:49:01<52:02, 5.20it/s] 66%|██████▌ | 31783/48008 [1:49:01<51:49, 5.22it/s] 66%|██████▌ | 31784/48008 [1:49:02<51:21, 5.27it/s] 66%|██████▌ | 31785/48008 [1:49:02<51:18, 5.27it/s] 66%|██████▌ | 31786/48008 [1:49:02<51:15, 5.28it/s] 66%|██████▌ | 31787/48008 [1:49:02<52:14, 5.17it/s] 66%|██████▌ | 31788/48008 [1:49:02<53:03, 5.09it/s] 66%|██████▌ | 31789/48008 [1:49:02<54:03, 5.00it/s] 66%|██████▌ | 31790/48008 [1:49:03<53:24, 5.06it/s] 66%|██████▌ | 31791/48008 [1:49:03<52:42, 5.13it/s] 66%|██████▌ | 31792/48008 [1:49:03<52:32, 5.14it/s] 66%|██████▌ | 31793/48008 [1:49:03<53:21, 5.07it/s] 66%|██████▌ | 31794/48008 [1:49:04<56:37, 4.77it/s] 66%|██████▌ | 31795/48008 [1:49:04<55:01, 4.91it/s] 66%|██████▌ | 31796/48008 [1:49:04<54:03, 5.00it/s] 66%|██████▌ | 31797/48008 [1:49:04<53:05, 5.09it/s] 66%|██████▌ | 31798/48008 [1:49:04<53:05, 5.09it/s] 66%|██████▌ | 31799/48008 [1:49:04<52:35, 5.14it/s] 66%|██████▌ | 31800/48008 [1:49:05<52:07, 5.18it/s] {'loss': 0.2987, 'grad_norm': 0.5, 'learning_rate': 6.752624562572904e-05, 'epoch': 0.66} 66%|██████▌ | 31800/48008 [1:49:05<52:07, 5.18it/s] 66%|██████▌ | 31801/48008 [1:49:05<51:52, 5.21it/s] 66%|██████▌ | 31802/48008 [1:49:05<51:51, 5.21it/s] 66%|██████▌ | 31803/48008 [1:49:05<51:56, 5.20it/s] 66%|██████▌ | 31804/48008 [1:49:05<51:48, 5.21it/s] 66%|██████▌ | 31805/48008 [1:49:06<52:13, 5.17it/s] 66%|██████▋ | 31806/48008 [1:49:06<52:17, 5.16it/s] 66%|██████▋ | 31807/48008 [1:49:06<51:41, 5.22it/s] 66%|██████▋ | 31808/48008 [1:49:06<51:25, 5.25it/s] 66%|██████▋ | 31809/48008 [1:49:06<51:34, 5.24it/s] 66%|██████▋ | 31810/48008 [1:49:07<52:35, 5.13it/s] 66%|██████▋ | 31811/48008 [1:49:07<52:18, 5.16it/s] 66%|██████▋ | 31812/48008 [1:49:07<52:27, 5.15it/s] 66%|██████▋ | 31813/48008 [1:49:07<53:33, 5.04it/s] 66%|██████▋ | 31814/48008 [1:49:07<52:59, 5.09it/s] 66%|██████▋ | 31815/48008 [1:49:08<53:26, 5.05it/s] 66%|██████▋ | 31816/48008 [1:49:08<52:19, 5.16it/s] 66%|██████▋ | 31817/48008 [1:49:08<52:33, 5.13it/s] 66%|██████▋ | 31818/48008 [1:49:08<51:40, 5.22it/s] 66%|██████▋ | 31819/48008 [1:49:08<51:28, 5.24it/s] 66%|██████▋ | 31820/48008 [1:49:09<51:28, 5.24it/s] 66%|██████▋ | 31821/48008 [1:49:09<51:54, 5.20it/s] 66%|██████▋ | 31822/48008 [1:49:09<51:57, 5.19it/s] 66%|██████▋ | 31823/48008 [1:49:09<51:38, 5.22it/s] 66%|██████▋ | 31824/48008 [1:49:09<51:20, 5.25it/s] 66%|██████▋ | 31825/48008 [1:49:09<51:53, 5.20it/s] 66%|██████▋ | 31826/48008 [1:49:10<51:32, 5.23it/s] 66%|██████▋ | 31827/48008 [1:49:10<52:45, 5.11it/s] 66%|██████▋ | 31828/48008 [1:49:10<1:08:08, 3.96it/s] 66%|██████▋ | 31829/48008 [1:49:10<1:02:53, 4.29it/s] 66%|██████▋ | 31830/48008 [1:49:11<59:59, 4.49it/s] 66%|██████▋ | 31831/48008 [1:49:11<57:57, 4.65it/s] 66%|██████▋ | 31832/48008 [1:49:11<55:59, 4.81it/s] 66%|██████▋ | 31833/48008 [1:49:11<54:02, 4.99it/s] 66%|██████▋ | 31834/48008 [1:49:11<56:56, 4.73it/s] 66%|██████▋ | 31835/48008 [1:49:12<54:50, 4.92it/s] 66%|██████▋ | 31836/48008 [1:49:12<53:57, 4.99it/s] 66%|██████▋ | 31837/48008 [1:49:12<54:09, 4.98it/s] 66%|██████▋ | 31838/48008 [1:49:12<53:28, 5.04it/s] 66%|██████▋ | 31839/48008 [1:49:12<52:58, 5.09it/s] 66%|██████▋ | 31840/48008 [1:49:13<52:12, 5.16it/s] 66%|██████▋ | 31841/48008 [1:49:13<52:01, 5.18it/s] 66%|██████▋ | 31842/48008 [1:49:13<51:42, 5.21it/s] 66%|██████▋ | 31843/48008 [1:49:13<51:34, 5.22it/s] 66%|██████▋ | 31844/48008 [1:49:13<51:11, 5.26it/s] 66%|██████▋ | 31845/48008 [1:49:14<51:46, 5.20it/s] 66%|██████▋ | 31846/48008 [1:49:14<55:26, 4.86it/s] 66%|██████▋ | 31847/48008 [1:49:14<54:01, 4.99it/s] 66%|██████▋ | 31848/48008 [1:49:14<53:06, 5.07it/s] 66%|██████▋ | 31849/48008 [1:49:14<53:14, 5.06it/s] 66%|██████▋ | 31850/48008 [1:49:15<52:42, 5.11it/s] {'loss': 0.3021, 'grad_norm': 0.294921875, 'learning_rate': 6.731794700883186e-05, 'epoch': 0.66} 66%|██████▋ | 31850/48008 [1:49:15<52:42, 5.11it/s] 66%|██████▋ | 31851/48008 [1:49:15<56:03, 4.80it/s] 66%|██████▋ | 31852/48008 [1:49:15<54:44, 4.92it/s] 66%|██████▋ | 31853/48008 [1:49:15<53:13, 5.06it/s] 66%|██████▋ | 31854/48008 [1:49:15<52:32, 5.12it/s] 66%|██████▋ | 31855/48008 [1:49:16<53:39, 5.02it/s] 66%|██████▋ | 31856/48008 [1:49:16<53:08, 5.07it/s] 66%|██████▋ | 31857/48008 [1:49:16<53:38, 5.02it/s] 66%|██████▋ | 31858/48008 [1:49:16<53:27, 5.03it/s] 66%|██████▋ | 31859/48008 [1:49:16<53:14, 5.05it/s] 66%|██████▋ | 31860/48008 [1:49:17<52:28, 5.13it/s] 66%|██████▋ | 31861/48008 [1:49:17<51:48, 5.19it/s] 66%|██████▋ | 31862/48008 [1:49:17<51:52, 5.19it/s] 66%|██████▋ | 31863/48008 [1:49:17<51:49, 5.19it/s] 66%|██████▋ | 31864/48008 [1:49:17<51:43, 5.20it/s] 66%|██████▋ | 31865/48008 [1:49:18<51:37, 5.21it/s] 66%|██████▋ | 31866/48008 [1:49:18<51:14, 5.25it/s] 66%|██████▋ | 31867/48008 [1:49:18<51:25, 5.23it/s] 66%|██████▋ | 31868/48008 [1:49:18<51:09, 5.26it/s] 66%|██████▋ | 31869/48008 [1:49:18<50:58, 5.28it/s] 66%|██████▋ | 31870/48008 [1:49:18<50:22, 5.34it/s] 66%|██████▋ | 31871/48008 [1:49:19<54:38, 4.92it/s] 66%|██████▋ | 31872/48008 [1:49:19<53:43, 5.01it/s] 66%|██████▋ | 31873/48008 [1:49:19<53:12, 5.05it/s] 66%|██████▋ | 31874/48008 [1:49:19<53:04, 5.07it/s] 66%|██████▋ | 31875/48008 [1:49:19<52:13, 5.15it/s] 66%|██████▋ | 31876/48008 [1:49:20<52:00, 5.17it/s] 66%|██████▋ | 31877/48008 [1:49:20<1:07:30, 3.98it/s] 66%|██████▋ | 31878/48008 [1:49:20<1:02:38, 4.29it/s] 66%|██████▋ | 31879/48008 [1:49:20<59:11, 4.54it/s] 66%|██████▋ | 31880/48008 [1:49:21<57:30, 4.67it/s] 66%|██████▋ | 31881/48008 [1:49:21<56:38, 4.75it/s] 66%|██████▋ | 31882/48008 [1:49:21<55:00, 4.89it/s] 66%|██████▋ | 31883/48008 [1:49:21<53:24, 5.03it/s] 66%|██████▋ | 31884/48008 [1:49:21<56:31, 4.75it/s] 66%|██████▋ | 31885/48008 [1:49:22<54:56, 4.89it/s] 66%|██████▋ | 31886/48008 [1:49:22<53:46, 5.00it/s] 66%|██████▋ | 31887/48008 [1:49:22<53:51, 4.99it/s] 66%|██████▋ | 31888/48008 [1:49:22<52:57, 5.07it/s] 66%|██████▋ | 31889/48008 [1:49:22<52:04, 5.16it/s] 66%|██████▋ | 31890/48008 [1:49:23<51:56, 5.17it/s] 66%|██████▋ | 31891/48008 [1:49:23<51:27, 5.22it/s] 66%|██████▋ | 31892/48008 [1:49:23<51:31, 5.21it/s] 66%|██████▋ | 31893/48008 [1:49:23<55:09, 4.87it/s] 66%|██████▋ | 31894/48008 [1:49:23<54:30, 4.93it/s] 66%|██████▋ | 31895/48008 [1:49:24<54:20, 4.94it/s] 66%|██████▋ | 31896/48008 [1:49:24<53:45, 5.00it/s] 66%|██████▋ | 31897/48008 [1:49:24<53:06, 5.06it/s] 66%|██████▋ | 31898/48008 [1:49:24<52:17, 5.13it/s] 66%|██████▋ | 31899/48008 [1:49:24<51:41, 5.19it/s] 66%|██████▋ | 31900/48008 [1:49:25<53:00, 5.07it/s] {'loss': 0.3019, 'grad_norm': 0.181640625, 'learning_rate': 6.710964839193468e-05, 'epoch': 0.66} 66%|██████▋ | 31900/48008 [1:49:25<53:00, 5.07it/s] 66%|██████▋ | 31901/48008 [1:49:25<52:48, 5.08it/s] 66%|██████▋ | 31902/48008 [1:49:25<52:07, 5.15it/s] 66%|██████▋ | 31903/48008 [1:49:25<52:14, 5.14it/s] 66%|██████▋ | 31904/48008 [1:49:25<51:46, 5.18it/s] 66%|██████▋ | 31905/48008 [1:49:26<52:16, 5.13it/s] 66%|██████▋ | 31906/48008 [1:49:26<52:08, 5.15it/s] 66%|██████▋ | 31907/48008 [1:49:26<52:03, 5.15it/s] 66%|██████▋ | 31908/48008 [1:49:26<52:00, 5.16it/s] 66%|██████▋ | 31909/48008 [1:49:26<51:35, 5.20it/s] 66%|██████▋ | 31910/48008 [1:49:27<55:15, 4.85it/s] 66%|██████▋ | 31911/48008 [1:49:27<54:39, 4.91it/s] 66%|██████▋ | 31912/48008 [1:49:27<53:21, 5.03it/s] 66%|██████▋ | 31913/48008 [1:49:27<1:02:02, 4.32it/s] 66%|██████▋ | 31914/48008 [1:49:27<59:35, 4.50it/s] 66%|██████▋ | 31915/48008 [1:49:28<57:08, 4.69it/s] 66%|██████▋ | 31916/48008 [1:49:28<54:38, 4.91it/s] 66%|██████▋ | 31917/48008 [1:49:28<53:11, 5.04it/s] 66%|██████▋ | 31918/48008 [1:49:28<53:08, 5.05it/s] 66%|██████▋ | 31919/48008 [1:49:28<52:38, 5.09it/s] 66%|██████▋ | 31920/48008 [1:49:29<51:52, 5.17it/s] 66%|██████▋ | 31921/48008 [1:49:29<51:43, 5.18it/s] 66%|██████▋ | 31922/48008 [1:49:29<51:21, 5.22it/s] 66%|██████▋ | 31923/48008 [1:49:29<51:20, 5.22it/s] 66%|██████▋ | 31924/48008 [1:49:29<51:16, 5.23it/s] 66%|██████▋ | 31925/48008 [1:49:30<51:16, 5.23it/s] 67%|██████▋ | 31926/48008 [1:49:30<51:08, 5.24it/s] 67%|██████▋ | 31927/48008 [1:49:30<1:00:26, 4.43it/s] 67%|██████▋ | 31928/48008 [1:49:30<57:13, 4.68it/s] 67%|██████▋ | 31929/48008 [1:49:30<55:29, 4.83it/s] 67%|██████▋ | 31930/48008 [1:49:31<54:11, 4.94it/s] 67%|██████▋ | 31931/48008 [1:49:31<54:01, 4.96it/s] 67%|██████▋ | 31932/48008 [1:49:31<53:03, 5.05it/s] 67%|██████▋ | 31933/48008 [1:49:31<53:28, 5.01it/s] 67%|██████▋ | 31934/48008 [1:49:31<52:42, 5.08it/s] 67%|██████▋ | 31935/48008 [1:49:32<52:06, 5.14it/s] 67%|██████▋ | 31936/48008 [1:49:32<55:32, 4.82it/s] 67%|██████▋ | 31937/48008 [1:49:32<54:17, 4.93it/s] 67%|██████▋ | 31938/48008 [1:49:32<53:16, 5.03it/s] 67%|██████▋ | 31939/48008 [1:49:32<52:40, 5.08it/s] 67%|██████▋ | 31940/48008 [1:49:33<52:47, 5.07it/s] 67%|██████▋ | 31941/48008 [1:49:33<52:19, 5.12it/s] 67%|██████▋ | 31942/48008 [1:49:33<51:49, 5.17it/s] 67%|██████▋ | 31943/48008 [1:49:33<51:55, 5.16it/s] 67%|██████▋ | 31944/48008 [1:49:33<51:58, 5.15it/s] 67%|██████▋ | 31945/48008 [1:49:34<51:39, 5.18it/s] 67%|██████▋ | 31946/48008 [1:49:34<51:18, 5.22it/s] 67%|██████▋ | 31947/48008 [1:49:34<51:33, 5.19it/s] 67%|██████▋ | 31948/48008 [1:49:34<50:59, 5.25it/s] 67%|██████▋ | 31949/48008 [1:49:34<51:28, 5.20it/s] 67%|██████▋ | 31950/48008 [1:49:35<1:06:52, 4.00it/s] {'loss': 0.3007, 'grad_norm': 0.90234375, 'learning_rate': 6.69013497750375e-05, 'epoch': 0.67} 67%|██████▋ | 31950/48008 [1:49:35<1:06:52, 4.00it/s] 67%|██████▋ | 31951/48008 [1:49:35<1:11:26, 3.75it/s] 67%|██████▋ | 31952/48008 [1:49:35<1:09:07, 3.87it/s] 67%|██████▋ | 31953/48008 [1:49:35<1:03:45, 4.20it/s] 67%|██████▋ | 31954/48008 [1:49:36<1:00:15, 4.44it/s] 67%|██████▋ | 31955/48008 [1:49:36<1:01:14, 4.37it/s] 67%|██████▋ | 31956/48008 [1:49:36<58:08, 4.60it/s] 67%|██████▋ | 31957/48008 [1:49:36<55:59, 4.78it/s] 67%|██████▋ | 31958/48008 [1:49:36<58:35, 4.56it/s] 67%|██████▋ | 31959/48008 [1:49:37<56:20, 4.75it/s] 67%|██████▋ | 31960/48008 [1:49:37<55:30, 4.82it/s] 67%|██████▋ | 31961/48008 [1:49:37<57:50, 4.62it/s] 67%|██████▋ | 31962/48008 [1:49:37<55:41, 4.80it/s] 67%|██████▋ | 31963/48008 [1:49:38<58:04, 4.60it/s] 67%|██████▋ | 31964/48008 [1:49:38<55:41, 4.80it/s] 67%|██████▋ | 31965/48008 [1:49:38<54:51, 4.87it/s] 67%|██████▋ | 31966/48008 [1:49:38<53:50, 4.97it/s] 67%|██████▋ | 31967/48008 [1:49:38<52:53, 5.06it/s] 67%|██████▋ | 31968/48008 [1:49:39<1:08:07, 3.92it/s] 67%|██████▋ | 31969/48008 [1:49:39<1:03:47, 4.19it/s] 67%|██████▋ | 31970/48008 [1:49:39<59:32, 4.49it/s] 67%|██████▋ | 31971/48008 [1:49:39<56:57, 4.69it/s] 67%|██████▋ | 31972/48008 [1:49:39<55:17, 4.83it/s] 67%|██████▋ | 31973/48008 [1:49:40<54:15, 4.93it/s] 67%|██████▋ | 31974/48008 [1:49:40<53:14, 5.02it/s] 67%|██████▋ | 31975/48008 [1:49:40<1:02:03, 4.31it/s] 67%|██████▋ | 31976/48008 [1:49:40<59:08, 4.52it/s] 67%|██████▋ | 31977/48008 [1:49:41<56:43, 4.71it/s] 67%|██████▋ | 31978/48008 [1:49:41<55:09, 4.84it/s] 67%|██████▋ | 31979/48008 [1:49:41<53:56, 4.95it/s] 67%|██████▋ | 31980/48008 [1:49:41<53:02, 5.04it/s] 67%|██████▋ | 31981/48008 [1:49:41<52:37, 5.08it/s] 67%|██████▋ | 31982/48008 [1:49:41<52:00, 5.14it/s] 67%|██████▋ | 31983/48008 [1:49:42<51:34, 5.18it/s] 67%|██████▋ | 31984/48008 [1:49:42<51:57, 5.14it/s] 67%|██████▋ | 31985/48008 [1:49:42<51:42, 5.16it/s] 67%|██████▋ | 31986/48008 [1:49:42<51:07, 5.22it/s] 67%|██████▋ | 31987/48008 [1:49:42<51:07, 5.22it/s] 67%|██████▋ | 31988/48008 [1:49:43<52:13, 5.11it/s] 67%|██████▋ | 31989/48008 [1:49:43<52:07, 5.12it/s] 67%|██████▋ | 31990/48008 [1:49:43<52:01, 5.13it/s] 67%|██████▋ | 31991/48008 [1:49:43<51:31, 5.18it/s] 67%|██████▋ | 31992/48008 [1:49:43<51:53, 5.14it/s] 67%|██████▋ | 31993/48008 [1:49:44<51:38, 5.17it/s] 67%|██████▋ | 31994/48008 [1:49:44<51:20, 5.20it/s] 67%|██████▋ | 31995/48008 [1:49:44<51:09, 5.22it/s] 67%|██████▋ | 31996/48008 [1:49:44<52:08, 5.12it/s] 67%|██████▋ | 31997/48008 [1:49:44<51:57, 5.14it/s] 67%|██████▋ | 31998/48008 [1:49:45<51:42, 5.16it/s] 67%|██████▋ | 31999/48008 [1:49:45<52:13, 5.11it/s] 67%|██████▋ | 32000/48008 [1:49:45<51:46, 5.15it/s] {'loss': 0.3012, 'grad_norm': 0.158203125, 'learning_rate': 6.669305115814031e-05, 'epoch': 0.67} 67%|██████▋ | 32000/48008 [1:49:45<51:46, 5.15it/s] 67%|██████▋ | 32001/48008 [1:49:45<51:22, 5.19it/s] 67%|██████▋ | 32002/48008 [1:49:45<51:17, 5.20it/s] 67%|██████▋ | 32003/48008 [1:49:46<51:04, 5.22it/s] 67%|██████▋ | 32004/48008 [1:49:46<51:21, 5.19it/s] 67%|██████▋ | 32005/48008 [1:49:46<51:13, 5.21it/s] 67%|██████▋ | 32006/48008 [1:49:46<51:11, 5.21it/s] 67%|██████▋ | 32007/48008 [1:49:46<52:00, 5.13it/s] 67%|██████▋ | 32008/48008 [1:49:47<51:51, 5.14it/s] 67%|██████▋ | 32009/48008 [1:49:47<51:22, 5.19it/s] 67%|██████▋ | 32010/48008 [1:49:47<51:05, 5.22it/s] 67%|██████▋ | 32011/48008 [1:49:47<51:20, 5.19it/s] 67%|██████▋ | 32012/48008 [1:49:47<51:15, 5.20it/s] 67%|██████▋ | 32013/48008 [1:49:47<51:32, 5.17it/s] 67%|██████▋ | 32014/48008 [1:49:48<51:27, 5.18it/s] 67%|██████▋ | 32015/48008 [1:49:48<51:33, 5.17it/s] 67%|██████▋ | 32016/48008 [1:49:48<51:20, 5.19it/s] 67%|██████▋ | 32017/48008 [1:49:48<51:01, 5.22it/s] 67%|██████▋ | 32018/48008 [1:49:48<50:59, 5.23it/s] 67%|██████▋ | 32019/48008 [1:49:49<51:22, 5.19it/s] 67%|██████▋ | 32020/48008 [1:49:49<52:06, 5.11it/s] 67%|██████▋ | 32021/48008 [1:49:49<51:47, 5.14it/s] 67%|██████▋ | 32022/48008 [1:49:49<51:25, 5.18it/s] 67%|██████▋ | 32023/48008 [1:49:49<51:25, 5.18it/s] 67%|██████▋ | 32024/48008 [1:49:50<51:21, 5.19it/s] 67%|██████▋ | 32025/48008 [1:49:50<51:14, 5.20it/s] 67%|██████▋ | 32026/48008 [1:49:50<51:07, 5.21it/s] 67%|██████▋ | 32027/48008 [1:49:50<50:56, 5.23it/s] 67%|██████▋ | 32028/48008 [1:49:50<51:45, 5.15it/s] 67%|██████▋ | 32029/48008 [1:49:51<50:55, 5.23it/s] 67%|██████▋ | 32030/48008 [1:49:51<50:47, 5.24it/s] 67%|██████▋ | 32031/48008 [1:49:51<54:20, 4.90it/s] 67%|██████▋ | 32032/48008 [1:49:51<53:41, 4.96it/s] 67%|██████▋ | 32033/48008 [1:49:51<52:36, 5.06it/s] 67%|██████▋ | 32034/48008 [1:49:52<52:03, 5.11it/s] 67%|██████▋ | 32035/48008 [1:49:52<51:26, 5.17it/s] 67%|██████▋ | 32036/48008 [1:49:52<51:42, 5.15it/s] 67%|██████▋ | 32037/48008 [1:49:52<51:19, 5.19it/s] 67%|██████▋ | 32038/48008 [1:49:52<50:54, 5.23it/s] 67%|██████▋ | 32039/48008 [1:49:53<50:53, 5.23it/s] 67%|██████▋ | 32040/48008 [1:49:53<52:17, 5.09it/s] 67%|██████▋ | 32041/48008 [1:49:53<51:32, 5.16it/s] 67%|██████▋ | 32042/48008 [1:49:53<54:50, 4.85it/s] 67%|██████▋ | 32043/48008 [1:49:53<53:38, 4.96it/s] 67%|██████▋ | 32044/48008 [1:49:54<52:32, 5.06it/s] 67%|██████▋ | 32045/48008 [1:49:54<52:33, 5.06it/s] 67%|██████▋ | 32046/48008 [1:49:54<51:59, 5.12it/s] 67%|██████▋ | 32047/48008 [1:49:54<52:07, 5.10it/s] 67%|██████▋ | 32048/48008 [1:49:54<51:36, 5.15it/s] 67%|██████▋ | 32049/48008 [1:49:54<51:19, 5.18it/s] 67%|██████▋ | 32050/48008 [1:49:55<50:53, 5.23it/s] {'loss': 0.3031, 'grad_norm': 0.482421875, 'learning_rate': 6.648475254124313e-05, 'epoch': 0.67} 67%|██████▋ | 32050/48008 [1:49:55<50:53, 5.23it/s] 67%|██████▋ | 32051/48008 [1:49:55<51:40, 5.15it/s] 67%|██████▋ | 32052/48008 [1:49:55<51:34, 5.16it/s] 67%|██████▋ | 32053/48008 [1:49:55<51:02, 5.21it/s] 67%|██████▋ | 32054/48008 [1:49:55<50:50, 5.23it/s] 67%|██████▋ | 32055/48008 [1:49:56<50:44, 5.24it/s] 67%|██████▋ | 32056/48008 [1:49:56<51:02, 5.21it/s] 67%|██████▋ | 32057/48008 [1:49:56<54:41, 4.86it/s] 67%|██████▋ | 32058/48008 [1:49:56<53:25, 4.98it/s] 67%|██████▋ | 32059/48008 [1:49:56<52:15, 5.09it/s] 67%|██████▋ | 32060/48008 [1:49:57<51:32, 5.16it/s] 67%|██████▋ | 32061/48008 [1:49:57<51:01, 5.21it/s] 67%|██████▋ | 32062/48008 [1:49:57<50:42, 5.24it/s] 67%|██████▋ | 32063/48008 [1:49:57<50:30, 5.26it/s] 67%|██████▋ | 32064/48008 [1:49:57<51:33, 5.15it/s] 67%|██████▋ | 32065/48008 [1:49:58<51:28, 5.16it/s] 67%|██████▋ | 32066/48008 [1:49:58<50:44, 5.24it/s] 67%|██████▋ | 32067/48008 [1:49:58<51:35, 5.15it/s] 67%|██████▋ | 32068/48008 [1:49:58<51:16, 5.18it/s] 67%|██████▋ | 32069/48008 [1:49:58<50:45, 5.23it/s] 67%|██████▋ | 32070/48008 [1:49:59<51:44, 5.13it/s] 67%|██████▋ | 32071/48008 [1:49:59<51:00, 5.21it/s] 67%|██████▋ | 32072/48008 [1:49:59<50:31, 5.26it/s] 67%|██████▋ | 32073/48008 [1:49:59<50:15, 5.28it/s] 67%|██████▋ | 32074/48008 [1:49:59<50:20, 5.28it/s] 67%|██████▋ | 32075/48008 [1:50:00<50:59, 5.21it/s] 67%|██████▋ | 32076/48008 [1:50:00<50:33, 5.25it/s] 67%|██████▋ | 32077/48008 [1:50:00<50:42, 5.24it/s] 67%|██████▋ | 32078/48008 [1:50:00<50:37, 5.25it/s] 67%|██████▋ | 32079/48008 [1:50:00<51:22, 5.17it/s] 67%|██████▋ | 32080/48008 [1:50:00<51:18, 5.17it/s] 67%|██████▋ | 32081/48008 [1:50:01<54:53, 4.84it/s] 67%|██████▋ | 32082/48008 [1:50:01<53:39, 4.95it/s] 67%|██████▋ | 32083/48008 [1:50:01<53:09, 4.99it/s] 67%|██████▋ | 32084/48008 [1:50:01<53:28, 4.96it/s] 67%|██████▋ | 32085/48008 [1:50:01<52:51, 5.02it/s] 67%|██████▋ | 32086/48008 [1:50:02<52:00, 5.10it/s] 67%|██████▋ | 32087/48008 [1:50:02<1:07:07, 3.95it/s] 67%|██████▋ | 32088/48008 [1:50:02<1:02:01, 4.28it/s] 67%|██████▋ | 32089/48008 [1:50:02<58:41, 4.52it/s] 67%|██████▋ | 32090/48008 [1:50:03<57:57, 4.58it/s] 67%|██████▋ | 32091/48008 [1:50:03<55:24, 4.79it/s] 67%|██████▋ | 32092/48008 [1:50:03<54:00, 4.91it/s] 67%|██████▋ | 32093/48008 [1:50:03<52:58, 5.01it/s] 67%|██████▋ | 32094/48008 [1:50:03<52:16, 5.07it/s] 67%|██████▋ | 32095/48008 [1:50:04<51:37, 5.14it/s] 67%|██████▋ | 32096/48008 [1:50:04<51:11, 5.18it/s] 67%|██████▋ | 32097/48008 [1:50:04<52:37, 5.04it/s] 67%|██████▋ | 32098/48008 [1:50:04<51:58, 5.10it/s] 67%|██████▋ | 32099/48008 [1:50:04<51:17, 5.17it/s] 67%|██████▋ | 32100/48008 [1:50:05<54:50, 4.83it/s] {'loss': 0.3012, 'grad_norm': 0.2353515625, 'learning_rate': 6.627645392434595e-05, 'epoch': 0.67} 67%|██████▋ | 32100/48008 [1:50:05<54:50, 4.83it/s] 67%|██████▋ | 32101/48008 [1:50:05<53:39, 4.94it/s] 67%|██████▋ | 32102/48008 [1:50:05<53:38, 4.94it/s] 67%|██████▋ | 32103/48008 [1:50:05<52:45, 5.02it/s] 67%|██████▋ | 32104/48008 [1:50:05<51:46, 5.12it/s] 67%|██████▋ | 32105/48008 [1:50:06<52:49, 5.02it/s] 67%|██████▋ | 32106/48008 [1:50:06<55:45, 4.75it/s] 67%|██████▋ | 32107/48008 [1:50:06<54:06, 4.90it/s] 67%|██████▋ | 32108/48008 [1:50:06<52:42, 5.03it/s] 67%|██████▋ | 32109/48008 [1:50:06<52:31, 5.05it/s] 67%|██████▋ | 32110/48008 [1:50:07<52:04, 5.09it/s] 67%|██████▋ | 32111/48008 [1:50:07<51:30, 5.14it/s] 67%|██████▋ | 32112/48008 [1:50:07<51:13, 5.17it/s] 67%|██████▋ | 32113/48008 [1:50:07<51:10, 5.18it/s] 67%|██████▋ | 32114/48008 [1:50:07<50:26, 5.25it/s] 67%|██████▋ | 32115/48008 [1:50:08<50:24, 5.25it/s] 67%|██████▋ | 32116/48008 [1:50:08<49:52, 5.31it/s] 67%|██████▋ | 32117/48008 [1:50:08<50:29, 5.25it/s] 67%|██████▋ | 32118/48008 [1:50:08<50:53, 5.20it/s] 67%|██████▋ | 32119/48008 [1:50:08<52:16, 5.07it/s] 67%|██████▋ | 32120/48008 [1:50:09<51:25, 5.15it/s] 67%|██████▋ | 32121/48008 [1:50:09<52:08, 5.08it/s] 67%|██████▋ | 32122/48008 [1:50:09<51:41, 5.12it/s] 67%|██████▋ | 32123/48008 [1:50:09<51:24, 5.15it/s] 67%|██████▋ | 32124/48008 [1:50:09<51:10, 5.17it/s] 67%|██████▋ | 32125/48008 [1:50:09<51:00, 5.19it/s] 67%|██████▋ | 32126/48008 [1:50:10<51:18, 5.16it/s] 67%|██████▋ | 32127/48008 [1:50:10<51:12, 5.17it/s] 67%|██████▋ | 32128/48008 [1:50:10<50:51, 5.20it/s] 67%|██████▋ | 32129/48008 [1:50:10<50:21, 5.25it/s] 67%|██████▋ | 32130/48008 [1:50:10<51:23, 5.15it/s] 67%|██████▋ | 32131/48008 [1:50:11<51:15, 5.16it/s] 67%|██████▋ | 32132/48008 [1:50:11<50:41, 5.22it/s] 67%|██████▋ | 32133/48008 [1:50:11<50:30, 5.24it/s] 67%|██████▋ | 32134/48008 [1:50:11<50:33, 5.23it/s] 67%|██████▋ | 32135/48008 [1:50:11<50:22, 5.25it/s] 67%|██████▋ | 32136/48008 [1:50:12<50:14, 5.27it/s] 67%|██████▋ | 32137/48008 [1:50:12<50:23, 5.25it/s] 67%|██████▋ | 32138/48008 [1:50:12<54:00, 4.90it/s] 67%|██████▋ | 32139/48008 [1:50:12<52:57, 4.99it/s] 67%|██████▋ | 32140/48008 [1:50:12<52:44, 5.01it/s] 67%|██████▋ | 32141/48008 [1:50:13<52:39, 5.02it/s] 67%|██████▋ | 32142/48008 [1:50:13<52:51, 5.00it/s] 67%|██████▋ | 32143/48008 [1:50:13<51:57, 5.09it/s] 67%|██████▋ | 32144/48008 [1:50:13<51:36, 5.12it/s] 67%|██████▋ | 32145/48008 [1:50:13<51:15, 5.16it/s] 67%|██████▋ | 32146/48008 [1:50:14<50:34, 5.23it/s] 67%|██████▋ | 32147/48008 [1:50:14<50:35, 5.22it/s] 67%|██████▋ | 32148/48008 [1:50:14<52:02, 5.08it/s] 67%|██████▋ | 32149/48008 [1:50:14<52:18, 5.05it/s] 67%|██████▋ | 32150/48008 [1:50:14<52:44, 5.01it/s] {'loss': 0.3023, 'grad_norm': 0.1533203125, 'learning_rate': 6.606815530744876e-05, 'epoch': 0.67} 67%|██████▋ | 32150/48008 [1:50:14<52:44, 5.01it/s] 67%|██████▋ | 32151/48008 [1:50:15<52:59, 4.99it/s] 67%|██████▋ | 32152/48008 [1:50:15<53:09, 4.97it/s] 67%|██████▋ | 32153/48008 [1:50:15<52:29, 5.03it/s] 67%|██████▋ | 32154/48008 [1:50:15<51:35, 5.12it/s] 67%|██████▋ | 32155/48008 [1:50:15<51:22, 5.14it/s] 67%|██████▋ | 32156/48008 [1:50:16<51:11, 5.16it/s] 67%|██████▋ | 32157/48008 [1:50:16<51:10, 5.16it/s] 67%|██████▋ | 32158/48008 [1:50:16<50:56, 5.19it/s] 67%|██████▋ | 32159/48008 [1:50:16<50:51, 5.19it/s] 67%|██████▋ | 32160/48008 [1:50:16<51:19, 5.15it/s] 67%|██████▋ | 32161/48008 [1:50:17<51:02, 5.17it/s] 67%|██████▋ | 32162/48008 [1:50:17<50:46, 5.20it/s] 67%|██████▋ | 32163/48008 [1:50:17<50:30, 5.23it/s] 67%|██████▋ | 32164/48008 [1:50:17<50:21, 5.24it/s] 67%|██████▋ | 32165/48008 [1:50:17<50:24, 5.24it/s] 67%|██████▋ | 32166/48008 [1:50:17<50:32, 5.22it/s] 67%|██████▋ | 32167/48008 [1:50:18<50:50, 5.19it/s] 67%|██████▋ | 32168/48008 [1:50:18<50:17, 5.25it/s] 67%|██████▋ | 32169/48008 [1:50:18<50:19, 5.24it/s] 67%|██████▋ | 32170/48008 [1:50:18<50:27, 5.23it/s] 67%|██████▋ | 32171/48008 [1:50:19<1:05:58, 4.00it/s] 67%|██████▋ | 32172/48008 [1:50:19<1:00:59, 4.33it/s] 67%|██████▋ | 32173/48008 [1:50:19<58:13, 4.53it/s] 67%|██████▋ | 32174/48008 [1:50:19<56:17, 4.69it/s] 67%|██████▋ | 32175/48008 [1:50:19<54:34, 4.83it/s] 67%|██████▋ | 32176/48008 [1:50:20<53:14, 4.96it/s] 67%|██████▋ | 32177/48008 [1:50:20<52:30, 5.03it/s] 67%|██████▋ | 32178/48008 [1:50:20<52:19, 5.04it/s] 67%|██████▋ | 32179/48008 [1:50:20<52:03, 5.07it/s] 67%|██████▋ | 32180/48008 [1:50:20<51:22, 5.14it/s] 67%|██████▋ | 32181/48008 [1:50:21<51:40, 5.11it/s] 67%|██████▋ | 32182/48008 [1:50:21<50:58, 5.17it/s] 67%|██████▋ | 32183/48008 [1:50:21<51:37, 5.11it/s] 67%|██████▋ | 32184/48008 [1:50:21<50:49, 5.19it/s] 67%|██████▋ | 32185/48008 [1:50:21<50:39, 5.21it/s] 67%|██████▋ | 32186/48008 [1:50:22<50:36, 5.21it/s] 67%|██████▋ | 32187/48008 [1:50:22<51:10, 5.15it/s] 67%|██████▋ | 32188/48008 [1:50:22<50:59, 5.17it/s] 67%|██████▋ | 32189/48008 [1:50:22<50:51, 5.18it/s] 67%|██████▋ | 32190/48008 [1:50:22<50:54, 5.18it/s] 67%|██████▋ | 32191/48008 [1:50:22<50:51, 5.18it/s] 67%|██████▋ | 32192/48008 [1:50:23<50:58, 5.17it/s] 67%|██████▋ | 32193/48008 [1:50:23<52:04, 5.06it/s] 67%|██████▋ | 32194/48008 [1:50:23<51:41, 5.10it/s] 67%|██████▋ | 32195/48008 [1:50:23<51:19, 5.14it/s] 67%|██████▋ | 32196/48008 [1:50:23<50:34, 5.21it/s] 67%|██████▋ | 32197/48008 [1:50:24<50:07, 5.26it/s] 67%|██████▋ | 32198/48008 [1:50:24<50:57, 5.17it/s] 67%|██████▋ | 32199/48008 [1:50:24<50:22, 5.23it/s] 67%|██████▋ | 32200/48008 [1:50:25<1:44:15, 2.53it/s] {'loss': 0.3005, 'grad_norm': 0.208984375, 'learning_rate': 6.585985669055158e-05, 'epoch': 0.67} 67%|██████▋ | 32200/48008 [1:50:25<1:44:15, 2.53it/s] 67%|██████▋ | 32201/48008 [1:50:25<1:28:00, 2.99it/s] 67%|██████▋ | 32202/48008 [1:50:25<1:17:13, 3.41it/s] 67%|██████▋ | 32203/48008 [1:50:25<1:10:07, 3.76it/s] 67%|██████▋ | 32204/48008 [1:50:26<1:04:59, 4.05it/s] 67%|██████▋ | 32205/48008 [1:50:26<1:00:36, 4.35it/s] 67%|██████▋ | 32206/48008 [1:50:27<1:59:28, 2.20it/s] 67%|██████▋ | 32207/48008 [1:50:27<1:38:26, 2.68it/s] 67%|██████▋ | 32208/48008 [1:50:27<1:24:04, 3.13it/s] 67%|██████▋ | 32209/48008 [1:50:27<1:14:02, 3.56it/s] 67%|██████▋ | 32210/48008 [1:50:28<1:06:43, 3.95it/s] 67%|██████▋ | 32211/48008 [1:50:28<1:01:51, 4.26it/s] 67%|██████▋ | 32212/48008 [1:50:29<1:47:34, 2.45it/s] 67%|██████▋ | 32213/48008 [1:50:29<1:30:35, 2.91it/s] 67%|██████▋ | 32214/48008 [1:50:29<1:19:05, 3.33it/s] 67%|██████▋ | 32215/48008 [1:50:29<1:10:12, 3.75it/s] 67%|██████▋ | 32216/48008 [1:50:29<1:04:13, 4.10it/s] 67%|██████▋ | 32217/48008 [1:50:30<1:00:17, 4.36it/s] 67%|██████▋ | 32218/48008 [1:50:30<57:33, 4.57it/s] 67%|██████▋ | 32219/48008 [1:50:30<55:52, 4.71it/s] 67%|██████▋ | 32220/48008 [1:50:30<55:01, 4.78it/s] 67%|██████▋ | 32221/48008 [1:50:30<53:35, 4.91it/s] 67%|██████▋ | 32222/48008 [1:50:31<52:38, 5.00it/s] 67%|██████▋ | 32223/48008 [1:50:31<51:51, 5.07it/s] 67%|██████▋ | 32224/48008 [1:50:31<52:01, 5.06it/s] 67%|██████▋ | 32225/48008 [1:50:32<1:49:30, 2.40it/s] 67%|██████▋ | 32226/48008 [1:50:32<1:30:54, 2.89it/s] 67%|██████▋ | 32227/48008 [1:50:32<1:18:39, 3.34it/s] 67%|██████▋ | 32228/48008 [1:50:32<1:10:10, 3.75it/s] 67%|██████▋ | 32229/48008 [1:50:33<1:04:12, 4.10it/s] 67%|██████▋ | 32230/48008 [1:50:33<59:49, 4.40it/s] 67%|██████▋ | 32231/48008 [1:50:33<56:34, 4.65it/s] 67%|██████▋ | 32232/48008 [1:50:33<54:17, 4.84it/s] 67%|██████▋ | 32233/48008 [1:50:34<1:02:21, 4.22it/s] 67%|██████▋ | 32234/48008 [1:50:34<58:37, 4.48it/s] 67%|██████▋ | 32235/48008 [1:50:34<55:34, 4.73it/s] 67%|██████▋ | 32236/48008 [1:50:34<53:41, 4.90it/s] 67%|██████▋ | 32237/48008 [1:50:34<54:12, 4.85it/s] 67%|██████▋ | 32238/48008 [1:50:34<53:27, 4.92it/s] 67%|██████▋ | 32239/48008 [1:50:35<53:21, 4.93it/s] 67%|██████▋ | 32240/48008 [1:50:35<52:20, 5.02it/s] 67%|██████▋ | 32241/48008 [1:50:35<51:41, 5.08it/s] 67%|██████▋ | 32242/48008 [1:50:35<51:28, 5.10it/s] 67%|██████▋ | 32243/48008 [1:50:35<51:05, 5.14it/s] 67%|██████▋ | 32244/48008 [1:50:36<50:47, 5.17it/s] 67%|██████▋ | 32245/48008 [1:50:36<50:44, 5.18it/s] 67%|██████▋ | 32246/48008 [1:50:36<50:10, 5.24it/s] 67%|██████▋ | 32247/48008 [1:50:36<50:16, 5.23it/s] 67%|██████▋ | 32248/48008 [1:50:37<1:40:04, 2.62it/s] 67%|██████▋ | 32249/48008 [1:50:37<1:25:16, 3.08it/s] 67%|██████▋ | 32250/48008 [1:50:37<1:14:34, 3.52it/s] {'loss': 0.3023, 'grad_norm': 0.47265625, 'learning_rate': 6.56515580736544e-05, 'epoch': 0.67} 67%|██████▋ | 32250/48008 [1:50:37<1:14:34, 3.52it/s] 67%|██████▋ | 32251/48008 [1:50:38<1:08:41, 3.82it/s] 67%|██████▋ | 32252/48008 [1:50:38<1:03:10, 4.16it/s] 67%|██████▋ | 32253/48008 [1:50:38<1:14:50, 3.51it/s] 67%|██████▋ | 32254/48008 [1:50:38<1:07:20, 3.90it/s] 67%|██████▋ | 32255/48008 [1:50:39<1:02:53, 4.17it/s] 67%|██████▋ | 32256/48008 [1:50:39<59:38, 4.40it/s] 67%|██████▋ | 32257/48008 [1:50:39<58:03, 4.52it/s] 67%|██████▋ | 32258/48008 [1:50:39<55:15, 4.75it/s] 67%|██████▋ | 32259/48008 [1:50:39<53:22, 4.92it/s] 67%|██████▋ | 32260/48008 [1:50:40<53:28, 4.91it/s] 67%|██████▋ | 32261/48008 [1:50:40<52:30, 5.00it/s] 67%|██████▋ | 32262/48008 [1:50:40<52:08, 5.03it/s] 67%|██████▋ | 32263/48008 [1:50:40<51:43, 5.07it/s] 67%|██████▋ | 32264/48008 [1:50:40<51:27, 5.10it/s] 67%|██████▋ | 32265/48008 [1:50:41<51:05, 5.13it/s] 67%|██████▋ | 32266/48008 [1:50:41<50:52, 5.16it/s] 67%|██████▋ | 32267/48008 [1:50:41<51:30, 5.09it/s] 67%|██████▋ | 32268/48008 [1:50:41<51:03, 5.14it/s] 67%|██████▋ | 32269/48008 [1:50:41<50:38, 5.18it/s] 67%|██████▋ | 32270/48008 [1:50:41<50:12, 5.22it/s] 67%|██████▋ | 32271/48008 [1:50:42<49:24, 5.31it/s] 67%|██████▋ | 32272/48008 [1:50:42<49:15, 5.32it/s] 67%|██████▋ | 32273/48008 [1:50:42<49:37, 5.28it/s] 67%|██████▋ | 32274/48008 [1:50:42<49:42, 5.27it/s] 67%|██████▋ | 32275/48008 [1:50:42<49:25, 5.31it/s] 67%|██████▋ | 32276/48008 [1:50:43<50:33, 5.19it/s] 67%|██████▋ | 32277/48008 [1:50:43<51:26, 5.10it/s] 67%|██████▋ | 32278/48008 [1:50:43<51:51, 5.06it/s] 67%|██████▋ | 32279/48008 [1:50:43<51:26, 5.10it/s] 67%|██████▋ | 32280/48008 [1:50:43<50:57, 5.14it/s] 67%|██████▋ | 32281/48008 [1:50:44<59:43, 4.39it/s] 67%|██████▋ | 32282/48008 [1:50:44<56:40, 4.62it/s] 67%|██████▋ | 32283/48008 [1:50:44<54:33, 4.80it/s] 67%|██████▋ | 32284/48008 [1:50:44<53:19, 4.91it/s] 67%|██████▋ | 32285/48008 [1:50:44<52:30, 4.99it/s] 67%|██████▋ | 32286/48008 [1:50:45<52:42, 4.97it/s] 67%|██████▋ | 32287/48008 [1:50:45<52:06, 5.03it/s] 67%|██████▋ | 32288/48008 [1:50:45<51:12, 5.12it/s] 67%|██████▋ | 32289/48008 [1:50:45<50:35, 5.18it/s] 67%|██████▋ | 32290/48008 [1:50:45<50:26, 5.19it/s] 67%|██████▋ | 32291/48008 [1:50:46<49:58, 5.24it/s] 67%|██████▋ | 32292/48008 [1:50:46<49:40, 5.27it/s] 67%|██████▋ | 32293/48008 [1:50:46<49:50, 5.25it/s] 67%|██████▋ | 32294/48008 [1:50:46<49:42, 5.27it/s] 67%|██████▋ | 32295/48008 [1:50:46<49:39, 5.27it/s] 67%|██████▋ | 32296/48008 [1:50:47<49:43, 5.27it/s] 67%|██████▋ | 32297/48008 [1:50:47<49:53, 5.25it/s] 67%|██████▋ | 32298/48008 [1:50:47<53:28, 4.90it/s] 67%|██████▋ | 32299/48008 [1:50:47<53:11, 4.92it/s] 67%|██████▋ | 32300/48008 [1:50:47<52:37, 4.97it/s] {'loss': 0.301, 'grad_norm': 0.66015625, 'learning_rate': 6.544325945675722e-05, 'epoch': 0.67} 67%|██████▋ | 32300/48008 [1:50:47<52:37, 4.97it/s] 67%|██████▋ | 32301/48008 [1:50:48<55:45, 4.70it/s] 67%|██████▋ | 32302/48008 [1:50:48<53:57, 4.85it/s] 67%|██████▋ | 32303/48008 [1:50:48<52:42, 4.97it/s] 67%|██████▋ | 32304/48008 [1:50:48<51:37, 5.07it/s] 67%|██████▋ | 32305/48008 [1:50:48<52:32, 4.98it/s] 67%|██████▋ | 32306/48008 [1:50:49<51:42, 5.06it/s] 67%|██████▋ | 32307/48008 [1:50:49<52:14, 5.01it/s] 67%|██████▋ | 32308/48008 [1:50:49<51:24, 5.09it/s] 67%|██████▋ | 32309/48008 [1:50:49<52:34, 4.98it/s] 67%|██████▋ | 32310/48008 [1:50:49<51:46, 5.05it/s] 67%|██████▋ | 32311/48008 [1:50:50<51:08, 5.12it/s] 67%|██████▋ | 32312/48008 [1:50:50<51:02, 5.13it/s] 67%|██████▋ | 32313/48008 [1:50:50<51:34, 5.07it/s] 67%|██████▋ | 32314/48008 [1:50:50<51:08, 5.11it/s] 67%|██████▋ | 32315/48008 [1:50:50<50:54, 5.14it/s] 67%|██████▋ | 32316/48008 [1:50:51<51:01, 5.13it/s] 67%|██████▋ | 32317/48008 [1:50:51<59:40, 4.38it/s] 67%|██████▋ | 32318/48008 [1:50:51<56:41, 4.61it/s] 67%|██████▋ | 32319/48008 [1:50:51<58:27, 4.47it/s] 67%|██████▋ | 32320/48008 [1:50:52<55:07, 4.74it/s] 67%|██████▋ | 32321/48008 [1:50:52<53:51, 4.85it/s] 67%|██████▋ | 32322/48008 [1:50:52<52:14, 5.00it/s] 67%|██████▋ | 32323/48008 [1:50:52<51:34, 5.07it/s] 67%|██████▋ | 32324/48008 [1:50:52<51:18, 5.09it/s] 67%|██████▋ | 32325/48008 [1:50:52<51:53, 5.04it/s] 67%|██████▋ | 32326/48008 [1:50:53<50:58, 5.13it/s] 67%|██████▋ | 32327/48008 [1:50:53<50:30, 5.17it/s] 67%|██████▋ | 32328/48008 [1:50:53<50:27, 5.18it/s] 67%|██████▋ | 32329/48008 [1:50:53<53:58, 4.84it/s] 67%|██████▋ | 32330/48008 [1:50:53<52:41, 4.96it/s] 67%|██████▋ | 32331/48008 [1:50:54<51:57, 5.03it/s] 67%|██████▋ | 32332/48008 [1:50:54<51:24, 5.08it/s] 67%|██████▋ | 32333/48008 [1:50:54<51:07, 5.11it/s] 67%|██████▋ | 32334/48008 [1:50:54<50:38, 5.16it/s] 67%|██████▋ | 32335/48008 [1:50:54<50:39, 5.16it/s] 67%|██████▋ | 32336/48008 [1:50:55<50:35, 5.16it/s] 67%|██████▋ | 32337/48008 [1:50:55<50:18, 5.19it/s] 67%|██████▋ | 32338/48008 [1:50:55<51:11, 5.10it/s] 67%|██████▋ | 32339/48008 [1:50:55<50:42, 5.15it/s] 67%|██████▋ | 32340/48008 [1:50:55<51:33, 5.07it/s] 67%|██████▋ | 32341/48008 [1:50:56<50:56, 5.13it/s] 67%|██████▋ | 32342/48008 [1:50:56<50:48, 5.14it/s] 67%|██████▋ | 32343/48008 [1:50:56<54:12, 4.82it/s] 67%|██████▋ | 32344/48008 [1:50:56<52:37, 4.96it/s] 67%|██████▋ | 32345/48008 [1:50:56<51:48, 5.04it/s] 67%|██████▋ | 32346/48008 [1:50:57<51:19, 5.09it/s] 67%|██████▋ | 32347/48008 [1:50:57<51:20, 5.08it/s] 67%|██████▋ | 32348/48008 [1:50:57<50:50, 5.13it/s] 67%|██████▋ | 32349/48008 [1:50:57<50:52, 5.13it/s] 67%|██████▋ | 32350/48008 [1:50:57<50:08, 5.20it/s] {'loss': 0.3007, 'grad_norm': 0.158203125, 'learning_rate': 6.523496083986002e-05, 'epoch': 0.67} 67%|██████▋ | 32350/48008 [1:50:57<50:08, 5.20it/s] 67%|██████▋ | 32351/48008 [1:50:58<50:36, 5.16it/s] 67%|██████▋ | 32352/48008 [1:50:58<50:47, 5.14it/s] 67%|██████▋ | 32353/48008 [1:50:58<59:31, 4.38it/s] 67%|██████▋ | 32354/48008 [1:50:58<56:55, 4.58it/s] 67%|██████▋ | 32355/48008 [1:50:58<54:54, 4.75it/s] 67%|██████▋ | 32356/48008 [1:50:59<53:15, 4.90it/s] 67%|██████▋ | 32357/48008 [1:50:59<52:46, 4.94it/s] 67%|██████▋ | 32358/48008 [1:50:59<1:00:52, 4.28it/s] 67%|██████▋ | 32359/48008 [1:50:59<57:34, 4.53it/s] 67%|██████▋ | 32360/48008 [1:51:00<1:10:32, 3.70it/s] 67%|██████▋ | 32361/48008 [1:51:00<1:04:34, 4.04it/s] 67%|██████▋ | 32362/48008 [1:51:00<1:00:13, 4.33it/s] 67%|██████▋ | 32363/48008 [1:51:00<56:56, 4.58it/s] 67%|██████▋ | 32364/48008 [1:51:00<54:20, 4.80it/s] 67%|██████▋ | 32365/48008 [1:51:01<53:49, 4.84it/s] 67%|██████▋ | 32366/48008 [1:51:01<52:32, 4.96it/s] 67%|██████▋ | 32367/48008 [1:51:01<51:59, 5.01it/s] 67%|██████▋ | 32368/48008 [1:51:01<51:12, 5.09it/s] 67%|██████▋ | 32369/48008 [1:51:01<50:46, 5.13it/s] 67%|██████▋ | 32370/48008 [1:51:02<54:04, 4.82it/s] 67%|██████▋ | 32371/48008 [1:51:02<53:41, 4.85it/s] 67%|██████▋ | 32372/48008 [1:51:02<52:31, 4.96it/s] 67%|██████▋ | 32373/48008 [1:51:02<52:05, 5.00it/s] 67%|██████▋ | 32374/48008 [1:51:02<52:15, 4.99it/s] 67%|██████▋ | 32375/48008 [1:51:03<52:22, 4.98it/s] 67%|██████▋ | 32376/48008 [1:51:03<51:30, 5.06it/s] 67%|██████▋ | 32377/48008 [1:51:03<50:56, 5.11it/s] 67%|██████▋ | 32378/48008 [1:51:03<50:40, 5.14it/s] 67%|██████▋ | 32379/48008 [1:51:03<50:46, 5.13it/s] 67%|██████▋ | 32380/48008 [1:51:04<51:00, 5.11it/s] 67%|██████▋ | 32381/48008 [1:51:04<50:37, 5.14it/s] 67%|██████▋ | 32382/48008 [1:51:04<49:55, 5.22it/s] 67%|██████▋ | 32383/48008 [1:51:04<49:38, 5.25it/s] 67%|██████▋ | 32384/48008 [1:51:04<50:26, 5.16it/s] 67%|██████▋ | 32385/48008 [1:51:05<50:18, 5.18it/s] 67%|██████▋ | 32386/48008 [1:51:05<1:05:20, 3.98it/s] 67%|██████▋ | 32387/48008 [1:51:05<1:00:38, 4.29it/s] 67%|██████▋ | 32388/48008 [1:51:05<57:21, 4.54it/s] 67%|██████▋ | 32389/48008 [1:51:06<56:06, 4.64it/s] 67%|██████▋ | 32390/48008 [1:51:06<54:37, 4.77it/s] 67%|██████▋ | 32391/48008 [1:51:06<52:53, 4.92it/s] 67%|██████▋ | 32392/48008 [1:51:06<51:53, 5.02it/s] 67%|██████▋ | 32393/48008 [1:51:06<54:54, 4.74it/s] 67%|██████▋ | 32394/48008 [1:51:07<53:21, 4.88it/s] 67%|██████▋ | 32395/48008 [1:51:07<52:19, 4.97it/s] 67%|██████▋ | 32396/48008 [1:51:07<51:32, 5.05it/s] 67%|██████▋ | 32397/48008 [1:51:07<51:10, 5.08it/s] 67%|██████▋ | 32398/48008 [1:51:07<50:44, 5.13it/s] 67%|██████▋ | 32399/48008 [1:51:08<50:17, 5.17it/s] 67%|██████▋ | 32400/48008 [1:51:08<49:57, 5.21it/s] {'loss': 0.299, 'grad_norm': 0.408203125, 'learning_rate': 6.502666222296284e-05, 'epoch': 0.67} 67%|██████▋ | 32400/48008 [1:51:08<49:57, 5.21it/s] 67%|██████▋ | 32401/48008 [1:51:08<50:52, 5.11it/s] 67%|██████▋ | 32402/48008 [1:51:08<51:20, 5.07it/s] 67%|██████▋ | 32403/48008 [1:51:08<51:00, 5.10it/s] 67%|██████▋ | 32404/48008 [1:51:09<51:01, 5.10it/s] 67%|██████▋ | 32405/48008 [1:51:09<50:37, 5.14it/s] 68%|██████▊ | 32406/48008 [1:51:09<51:43, 5.03it/s] 68%|██████▊ | 32407/48008 [1:51:09<51:54, 5.01it/s] 68%|██████▊ | 32408/48008 [1:51:09<50:54, 5.11it/s] 68%|██████▊ | 32409/48008 [1:51:10<50:24, 5.16it/s] 68%|██████▊ | 32410/48008 [1:51:10<50:20, 5.16it/s] 68%|██████▊ | 32411/48008 [1:51:10<1:05:32, 3.97it/s] 68%|██████▊ | 32412/48008 [1:51:10<1:02:12, 4.18it/s] 68%|██████▊ | 32413/48008 [1:51:11<1:13:30, 3.54it/s] 68%|██████▊ | 32414/48008 [1:51:11<1:07:13, 3.87it/s] 68%|██████▊ | 32415/48008 [1:51:11<1:02:07, 4.18it/s] 68%|██████▊ | 32416/48008 [1:51:11<58:17, 4.46it/s] 68%|██████▊ | 32417/48008 [1:51:11<55:50, 4.65it/s] 68%|██████▊ | 32418/48008 [1:51:12<53:54, 4.82it/s] 68%|██████▊ | 32419/48008 [1:51:12<52:40, 4.93it/s] 68%|██████▊ | 32420/48008 [1:51:12<52:02, 4.99it/s] 68%|██████▊ | 32421/48008 [1:51:12<51:18, 5.06it/s] 68%|██████▊ | 32422/48008 [1:51:12<50:24, 5.15it/s] 68%|██████▊ | 32423/48008 [1:51:13<51:38, 5.03it/s] 68%|██████▊ | 32424/48008 [1:51:13<1:00:06, 4.32it/s] 68%|██████▊ | 32425/48008 [1:51:13<56:49, 4.57it/s] 68%|██████▊ | 32426/48008 [1:51:13<55:26, 4.68it/s] 68%|██████▊ | 32427/48008 [1:51:14<53:56, 4.81it/s] 68%|██████▊ | 32428/48008 [1:51:14<53:01, 4.90it/s] 68%|██████▊ | 32429/48008 [1:51:14<51:55, 5.00it/s] 68%|██████▊ | 32430/48008 [1:51:14<52:21, 4.96it/s] 68%|██████▊ | 32431/48008 [1:51:14<52:24, 4.95it/s] 68%|██████▊ | 32432/48008 [1:51:15<51:36, 5.03it/s] 68%|██████▊ | 32433/48008 [1:51:15<50:43, 5.12it/s] 68%|██████▊ | 32434/48008 [1:51:15<51:15, 5.06it/s] 68%|██████▊ | 32435/48008 [1:51:15<51:05, 5.08it/s] 68%|██████▊ | 32436/48008 [1:51:15<1:05:50, 3.94it/s] 68%|██████▊ | 32437/48008 [1:51:16<1:00:37, 4.28it/s] 68%|██████▊ | 32438/48008 [1:51:16<57:09, 4.54it/s] 68%|██████▊ | 32439/48008 [1:51:16<54:34, 4.76it/s] 68%|██████▊ | 32440/48008 [1:51:16<54:26, 4.77it/s] 68%|██████▊ | 32441/48008 [1:51:16<52:48, 4.91it/s] 68%|██████▊ | 32442/48008 [1:51:17<51:50, 5.00it/s] 68%|██████▊ | 32443/48008 [1:51:17<51:17, 5.06it/s] 68%|██████▊ | 32444/48008 [1:51:17<50:41, 5.12it/s] 68%|██████▊ | 32445/48008 [1:51:17<51:05, 5.08it/s] 68%|██████▊ | 32446/48008 [1:51:17<50:20, 5.15it/s] 68%|██████▊ | 32447/48008 [1:51:18<51:08, 5.07it/s] 68%|██████▊ | 32448/48008 [1:51:18<50:37, 5.12it/s] 68%|██████▊ | 32449/48008 [1:51:18<50:14, 5.16it/s] 68%|██████▊ | 32450/48008 [1:51:18<58:56, 4.40it/s] {'loss': 0.2972, 'grad_norm': 0.578125, 'learning_rate': 6.481836360606565e-05, 'epoch': 0.68} 68%|██████▊ | 32450/48008 [1:51:18<58:56, 4.40it/s] 68%|██████▊ | 32451/48008 [1:51:18<56:14, 4.61it/s] 68%|██████▊ | 32452/48008 [1:51:19<1:03:10, 4.10it/s] 68%|██████▊ | 32453/48008 [1:51:19<59:00, 4.39it/s] 68%|██████▊ | 32454/48008 [1:51:19<56:14, 4.61it/s] 68%|██████▊ | 32455/48008 [1:51:19<54:07, 4.79it/s] 68%|██████▊ | 32456/48008 [1:51:20<53:09, 4.88it/s] 68%|██████▊ | 32457/48008 [1:51:20<52:13, 4.96it/s] 68%|██████▊ | 32458/48008 [1:51:20<50:46, 5.10it/s] 68%|██████▊ | 32459/48008 [1:51:20<51:22, 5.04it/s] 68%|██████▊ | 32460/48008 [1:51:20<50:24, 5.14it/s] 68%|██████▊ | 32461/48008 [1:51:21<50:39, 5.12it/s] 68%|██████▊ | 32462/48008 [1:51:21<51:06, 5.07it/s] 68%|██████▊ | 32463/48008 [1:51:21<1:05:49, 3.94it/s] 68%|██████▊ | 32464/48008 [1:51:21<1:00:48, 4.26it/s] 68%|██████▊ | 32465/48008 [1:51:21<57:30, 4.50it/s] 68%|██████▊ | 32466/48008 [1:51:22<56:35, 4.58it/s] 68%|██████▊ | 32467/48008 [1:51:22<58:08, 4.45it/s] 68%|██████▊ | 32468/48008 [1:51:22<55:18, 4.68it/s] 68%|██████▊ | 32469/48008 [1:51:22<53:33, 4.84it/s] 68%|██████▊ | 32470/48008 [1:51:23<52:41, 4.91it/s] 68%|██████▊ | 32471/48008 [1:51:23<51:22, 5.04it/s] 68%|██████▊ | 32472/48008 [1:51:23<51:12, 5.06it/s] 68%|██████▊ | 32473/48008 [1:51:23<51:11, 5.06it/s] 68%|██████▊ | 32474/48008 [1:51:23<51:10, 5.06it/s] 68%|██████▊ | 32475/48008 [1:51:23<50:41, 5.11it/s] 68%|██████▊ | 32476/48008 [1:51:24<50:12, 5.16it/s] 68%|██████▊ | 32477/48008 [1:51:24<50:00, 5.18it/s] 68%|██████▊ | 32478/48008 [1:51:24<51:14, 5.05it/s] 68%|██████▊ | 32479/48008 [1:51:24<50:32, 5.12it/s] 68%|██████▊ | 32480/48008 [1:51:24<50:14, 5.15it/s] 68%|██████▊ | 32481/48008 [1:51:25<58:53, 4.39it/s] 68%|██████▊ | 32482/48008 [1:51:25<1:11:12, 3.63it/s] 68%|██████▊ | 32483/48008 [1:51:25<1:05:26, 3.95it/s] 68%|██████▊ | 32484/48008 [1:51:26<1:01:54, 4.18it/s] 68%|██████▊ | 32485/48008 [1:51:26<57:43, 4.48it/s] 68%|██████▊ | 32486/48008 [1:51:26<55:02, 4.70it/s] 68%|██████▊ | 32487/48008 [1:51:26<53:23, 4.84it/s] 68%|██████▊ | 32488/48008 [1:51:26<51:54, 4.98it/s] 68%|██████▊ | 32489/48008 [1:51:27<51:47, 4.99it/s] 68%|██████▊ | 32490/48008 [1:51:27<50:50, 5.09it/s] 68%|██████▊ | 32491/48008 [1:51:27<50:32, 5.12it/s] 68%|██████▊ | 32492/48008 [1:51:27<51:01, 5.07it/s] 68%|██████▊ | 32493/48008 [1:51:27<50:27, 5.12it/s] 68%|██████▊ | 32494/48008 [1:51:27<50:03, 5.17it/s] 68%|██████▊ | 32495/48008 [1:51:28<49:22, 5.24it/s] 68%|██████▊ | 32496/48008 [1:51:28<49:26, 5.23it/s] 68%|██████▊ | 32497/48008 [1:51:28<49:24, 5.23it/s] 68%|██████▊ | 32498/48008 [1:51:28<1:04:30, 4.01it/s] 68%|██████▊ | 32499/48008 [1:51:29<1:00:02, 4.30it/s] 68%|██████▊ | 32500/48008 [1:51:29<56:32, 4.57it/s] {'loss': 0.2993, 'grad_norm': 0.1533203125, 'learning_rate': 6.461006498916847e-05, 'epoch': 0.68} 68%|██████▊ | 32500/48008 [1:51:29<56:32, 4.57it/s] 68%|██████▊ | 32501/48008 [1:51:29<54:08, 4.77it/s] 68%|██████▊ | 32502/48008 [1:51:29<52:24, 4.93it/s] 68%|██████▊ | 32503/48008 [1:51:29<51:24, 5.03it/s] 68%|██████▊ | 32504/48008 [1:51:30<50:52, 5.08it/s] 68%|██████▊ | 32505/48008 [1:51:30<50:46, 5.09it/s] 68%|██████▊ | 32506/48008 [1:51:30<50:16, 5.14it/s] 68%|██████▊ | 32507/48008 [1:51:30<49:56, 5.17it/s] 68%|██████▊ | 32508/48008 [1:51:30<49:56, 5.17it/s] 68%|██████▊ | 32509/48008 [1:51:31<50:41, 5.10it/s] 68%|██████▊ | 32510/48008 [1:51:31<50:19, 5.13it/s] 68%|██████▊ | 32511/48008 [1:51:31<50:57, 5.07it/s] 68%|██████▊ | 32512/48008 [1:51:31<50:30, 5.11it/s] 68%|██████▊ | 32513/48008 [1:51:31<53:37, 4.82it/s] 68%|██████▊ | 32514/48008 [1:51:32<52:32, 4.92it/s] 68%|██████▊ | 32515/48008 [1:51:32<52:50, 4.89it/s] 68%|██████▊ | 32516/48008 [1:51:32<51:18, 5.03it/s] 68%|██████▊ | 32517/48008 [1:51:32<50:31, 5.11it/s] 68%|██████▊ | 32518/48008 [1:51:32<51:35, 5.00it/s] 68%|██████▊ | 32519/48008 [1:51:33<54:27, 4.74it/s] 68%|██████▊ | 32520/48008 [1:51:33<52:43, 4.90it/s] 68%|██████▊ | 32521/48008 [1:51:33<52:29, 4.92it/s] 68%|██████▊ | 32522/48008 [1:51:33<51:21, 5.03it/s] 68%|██████▊ | 32523/48008 [1:51:33<51:06, 5.05it/s] 68%|██████▊ | 32524/48008 [1:51:34<51:22, 5.02it/s] 68%|██████▊ | 32525/48008 [1:51:34<51:05, 5.05it/s] 68%|██████▊ | 32526/48008 [1:51:34<54:04, 4.77it/s] 68%|██████▊ | 32527/48008 [1:51:34<52:26, 4.92it/s] 68%|██████▊ | 32528/48008 [1:51:34<52:15, 4.94it/s] 68%|██████▊ | 32529/48008 [1:51:35<51:52, 4.97it/s] 68%|██████▊ | 32530/48008 [1:51:35<52:04, 4.95it/s] 68%|██████▊ | 32531/48008 [1:51:35<50:53, 5.07it/s] 68%|██████▊ | 32532/48008 [1:51:35<52:08, 4.95it/s] 68%|██████▊ | 32533/48008 [1:51:35<51:17, 5.03it/s] 68%|██████▊ | 32534/48008 [1:51:36<51:51, 4.97it/s] 68%|██████▊ | 32535/48008 [1:51:36<51:01, 5.05it/s] 68%|██████▊ | 32536/48008 [1:51:36<50:46, 5.08it/s] 68%|██████▊ | 32537/48008 [1:51:36<51:44, 4.98it/s] 68%|██████▊ | 32538/48008 [1:51:36<51:19, 5.02it/s] 68%|██████▊ | 32539/48008 [1:51:37<51:01, 5.05it/s] 68%|██████▊ | 32540/48008 [1:51:37<50:24, 5.11it/s] 68%|██████▊ | 32541/48008 [1:51:37<50:32, 5.10it/s] 68%|██████▊ | 32542/48008 [1:51:37<51:03, 5.05it/s] 68%|██████▊ | 32543/48008 [1:51:37<51:26, 5.01it/s] 68%|██████▊ | 32544/48008 [1:51:38<1:05:48, 3.92it/s] 68%|██████▊ | 32545/48008 [1:51:38<1:00:52, 4.23it/s] 68%|██████▊ | 32546/48008 [1:51:38<58:11, 4.43it/s] 68%|██████▊ | 32547/48008 [1:51:38<55:31, 4.64it/s] 68%|██████▊ | 32548/48008 [1:51:39<53:32, 4.81it/s] 68%|██████▊ | 32549/48008 [1:51:39<52:39, 4.89it/s] 68%|██████▊ | 32550/48008 [1:51:39<51:38, 4.99it/s] {'loss': 0.2996, 'grad_norm': 0.1328125, 'learning_rate': 6.440176637227129e-05, 'epoch': 0.68} 68%|██████▊ | 32550/48008 [1:51:39<51:38, 4.99it/s] 68%|██████▊ | 32551/48008 [1:51:39<51:01, 5.05it/s] 68%|██████▊ | 32552/48008 [1:51:39<51:23, 5.01it/s] 68%|██████▊ | 32553/48008 [1:51:39<50:25, 5.11it/s] 68%|██████▊ | 32554/48008 [1:51:40<49:58, 5.15it/s] 68%|██████▊ | 32555/48008 [1:51:40<50:40, 5.08it/s] 68%|██████▊ | 32556/48008 [1:51:40<50:37, 5.09it/s] 68%|██████▊ | 32557/48008 [1:51:40<49:57, 5.15it/s] 68%|██████▊ | 32558/48008 [1:51:40<50:42, 5.08it/s] 68%|██████▊ | 32559/48008 [1:51:41<50:08, 5.14it/s] 68%|██████▊ | 32560/48008 [1:51:41<50:25, 5.11it/s] 68%|██████▊ | 32561/48008 [1:51:41<49:48, 5.17it/s] 68%|██████▊ | 32562/48008 [1:51:41<50:00, 5.15it/s] 68%|██████▊ | 32563/48008 [1:51:41<53:21, 4.82it/s] 68%|██████▊ | 32564/48008 [1:51:42<52:12, 4.93it/s] 68%|██████▊ | 32565/48008 [1:51:42<1:00:07, 4.28it/s] 68%|██████▊ | 32566/48008 [1:51:42<58:12, 4.42it/s] 68%|██████▊ | 32567/48008 [1:51:42<55:16, 4.66it/s] 68%|██████▊ | 32568/48008 [1:51:43<53:33, 4.80it/s] 68%|██████▊ | 32569/48008 [1:51:43<52:35, 4.89it/s] 68%|██████▊ | 32570/48008 [1:51:43<51:31, 4.99it/s] 68%|██████▊ | 32571/48008 [1:51:43<54:21, 4.73it/s] 68%|██████▊ | 32572/48008 [1:51:43<52:45, 4.88it/s] 68%|██████▊ | 32573/48008 [1:51:44<51:32, 4.99it/s] 68%|██████▊ | 32574/48008 [1:51:44<50:44, 5.07it/s] 68%|██████▊ | 32575/48008 [1:51:44<51:08, 5.03it/s] 68%|██████▊ | 32576/48008 [1:51:44<50:26, 5.10it/s] 68%|██████▊ | 32577/48008 [1:51:44<58:53, 4.37it/s] 68%|██████▊ | 32578/48008 [1:51:45<55:36, 4.63it/s] 68%|██████▊ | 32579/48008 [1:51:45<54:18, 4.74it/s] 68%|██████▊ | 32580/48008 [1:51:45<52:30, 4.90it/s] 68%|██████▊ | 32581/48008 [1:51:45<51:26, 5.00it/s] 68%|██████▊ | 32582/48008 [1:51:46<59:35, 4.31it/s] 68%|██████▊ | 32583/48008 [1:51:46<56:26, 4.55it/s] 68%|██████▊ | 32584/48008 [1:51:46<54:05, 4.75it/s] 68%|██████▊ | 32585/48008 [1:51:46<52:29, 4.90it/s] 68%|██████▊ | 32586/48008 [1:51:46<51:27, 4.99it/s] 68%|██████▊ | 32587/48008 [1:51:46<51:10, 5.02it/s] 68%|██████▊ | 32588/48008 [1:51:47<50:18, 5.11it/s] 68%|██████▊ | 32589/48008 [1:51:47<53:28, 4.81it/s] 68%|██████▊ | 32590/48008 [1:51:47<52:58, 4.85it/s] 68%|██████▊ | 32591/48008 [1:51:47<52:16, 4.92it/s] 68%|██████▊ | 32592/48008 [1:51:47<51:11, 5.02it/s] 68%|██████▊ | 32593/48008 [1:51:48<50:39, 5.07it/s] 68%|██████▊ | 32594/48008 [1:51:48<49:52, 5.15it/s] 68%|██████▊ | 32595/48008 [1:51:48<49:24, 5.20it/s] 68%|██████▊ | 32596/48008 [1:51:48<49:19, 5.21it/s] 68%|██████▊ | 32597/48008 [1:51:48<50:03, 5.13it/s] 68%|██████▊ | 32598/48008 [1:51:49<51:15, 5.01it/s] 68%|██████▊ | 32599/48008 [1:51:49<51:16, 5.01it/s] 68%|██████▊ | 32600/48008 [1:51:49<1:05:44, 3.91it/s] {'loss': 0.2966, 'grad_norm': 0.47265625, 'learning_rate': 6.419346775537411e-05, 'epoch': 0.68} 68%|██████▊ | 32600/48008 [1:51:49<1:05:44, 3.91it/s] 68%|██████▊ | 32601/48008 [1:51:49<1:00:42, 4.23it/s] 68%|██████▊ | 32602/48008 [1:51:50<57:10, 4.49it/s] 68%|██████▊ | 32603/48008 [1:51:50<55:17, 4.64it/s] 68%|██████▊ | 32604/48008 [1:51:50<56:56, 4.51it/s] 68%|██████▊ | 32605/48008 [1:51:50<54:31, 4.71it/s] 68%|██████▊ | 32606/48008 [1:51:50<53:17, 4.82it/s] 68%|██████▊ | 32607/48008 [1:51:51<51:52, 4.95it/s] 68%|██████▊ | 32608/48008 [1:51:51<50:55, 5.04it/s] 68%|██████▊ | 32609/48008 [1:51:51<49:51, 5.15it/s] 68%|██████▊ | 32610/48008 [1:51:51<49:39, 5.17it/s] 68%|██████▊ | 32611/48008 [1:51:51<50:46, 5.05it/s] 68%|██████▊ | 32612/48008 [1:51:52<50:18, 5.10it/s] 68%|██████▊ | 32613/48008 [1:51:52<49:29, 5.19it/s] 68%|██████▊ | 32614/48008 [1:51:52<49:23, 5.19it/s] 68%|██████▊ | 32615/48008 [1:51:52<48:50, 5.25it/s] 68%|██████▊ | 32616/48008 [1:51:52<48:54, 5.25it/s] 68%|██████▊ | 32617/48008 [1:51:53<48:41, 5.27it/s] 68%|██████▊ | 32618/48008 [1:51:53<48:22, 5.30it/s] 68%|██████▊ | 32619/48008 [1:51:53<49:05, 5.22it/s] 68%|██████▊ | 32620/48008 [1:51:53<49:03, 5.23it/s] 68%|██████▊ | 32621/48008 [1:51:53<48:41, 5.27it/s] 68%|██████▊ | 32622/48008 [1:51:54<57:40, 4.45it/s] 68%|██████▊ | 32623/48008 [1:51:54<54:29, 4.71it/s] 68%|██████▊ | 32624/48008 [1:51:54<1:07:50, 3.78it/s] 68%|██████▊ | 32625/48008 [1:51:54<1:02:23, 4.11it/s] 68%|██████▊ | 32626/48008 [1:51:55<58:08, 4.41it/s] 68%|██████▊ | 32627/48008 [1:51:55<55:28, 4.62it/s] 68%|██████▊ | 32628/48008 [1:51:55<53:53, 4.76it/s] 68%|██████▊ | 32629/48008 [1:51:55<52:19, 4.90it/s] 68%|██████▊ | 32630/48008 [1:51:55<54:49, 4.68it/s] 68%|██████▊ | 32631/48008 [1:51:56<53:01, 4.83it/s] 68%|██████▊ | 32632/48008 [1:51:56<55:13, 4.64it/s] 68%|██████▊ | 32633/48008 [1:51:56<54:15, 4.72it/s] 68%|██████▊ | 32634/48008 [1:51:56<54:11, 4.73it/s] 68%|██████▊ | 32635/48008 [1:51:56<56:09, 4.56it/s] 68%|██████▊ | 32636/48008 [1:51:57<54:22, 4.71it/s] 68%|██████▊ | 32637/48008 [1:51:57<52:43, 4.86it/s] 68%|██████▊ | 32638/48008 [1:51:57<51:58, 4.93it/s] 68%|██████▊ | 32639/48008 [1:51:57<50:53, 5.03it/s] 68%|██████▊ | 32640/48008 [1:51:57<50:10, 5.10it/s] 68%|██████▊ | 32641/48008 [1:51:58<51:08, 5.01it/s] 68%|██████▊ | 32642/48008 [1:51:58<49:57, 5.13it/s] 68%|██████▊ | 32643/48008 [1:51:58<49:37, 5.16it/s] 68%|██████▊ | 32644/48008 [1:51:58<49:25, 5.18it/s] 68%|██████▊ | 32645/48008 [1:51:58<49:07, 5.21it/s] 68%|██████▊ | 32646/48008 [1:51:59<49:33, 5.17it/s] 68%|██████▊ | 32647/48008 [1:51:59<49:18, 5.19it/s] 68%|██████▊ | 32648/48008 [1:51:59<52:51, 4.84it/s] 68%|██████▊ | 32649/48008 [1:51:59<52:27, 4.88it/s] 68%|██████▊ | 32650/48008 [1:51:59<52:13, 4.90it/s] 68%|██████▊ | 32650/48008 [1:51:59<52:13, 4.90it/s]{'loss': 0.2998, 'grad_norm': 0.396484375, 'learning_rate': 6.398516913847692e-05, 'epoch': 0.68} 68%|██████▊ | 32651/48008 [1:52:00<51:10, 5.00it/s] 68%|██████▊ | 32652/48008 [1:52:00<50:50, 5.03it/s] 68%|██████▊ | 32653/48008 [1:52:00<50:00, 5.12it/s] 68%|██████▊ | 32654/48008 [1:52:00<50:11, 5.10it/s] 68%|██████▊ | 32655/48008 [1:52:01<1:04:49, 3.95it/s] 68%|██████▊ | 32656/48008 [1:52:01<59:58, 4.27it/s] 68%|██████▊ | 32657/48008 [1:52:01<57:33, 4.44it/s] 68%|██████▊ | 32658/48008 [1:52:01<58:37, 4.36it/s] 68%|██████▊ | 32659/48008 [1:52:01<55:26, 4.61it/s] 68%|██████▊ | 32660/48008 [1:52:02<53:30, 4.78it/s] 68%|██████▊ | 32661/48008 [1:52:02<52:06, 4.91it/s] 68%|██████▊ | 32662/48008 [1:52:02<51:00, 5.01it/s] 68%|██████▊ | 32663/48008 [1:52:02<50:44, 5.04it/s] 68%|██████▊ | 32664/48008 [1:52:02<50:22, 5.08it/s] 68%|██████▊ | 32665/48008 [1:52:03<50:08, 5.10it/s] 68%|██████▊ | 32666/48008 [1:52:03<53:22, 4.79it/s] 68%|██████▊ | 32667/48008 [1:52:03<51:54, 4.93it/s] 68%|██████▊ | 32668/48008 [1:52:03<52:19, 4.89it/s] 68%|██████▊ | 32669/48008 [1:52:03<51:49, 4.93it/s] 68%|██████▊ | 32670/48008 [1:52:04<50:59, 5.01it/s] 68%|██████▊ | 32671/48008 [1:52:04<51:14, 4.99it/s] 68%|██████▊ | 32672/48008 [1:52:04<50:31, 5.06it/s] 68%|██████▊ | 32673/48008 [1:52:04<49:46, 5.13it/s] 68%|██████▊ | 32674/48008 [1:52:04<49:34, 5.16it/s] 68%|██████▊ | 32675/48008 [1:52:05<49:11, 5.19it/s] 68%|██████▊ | 32676/48008 [1:52:05<50:04, 5.10it/s] 68%|██████▊ | 32677/48008 [1:52:05<51:00, 5.01it/s] 68%|██████▊ | 32678/48008 [1:52:05<50:09, 5.09it/s] 68%|██████▊ | 32679/48008 [1:52:05<49:35, 5.15it/s] 68%|██████▊ | 32680/48008 [1:52:06<49:21, 5.18it/s] 68%|██████▊ | 32681/48008 [1:52:06<49:13, 5.19it/s] 68%|██████▊ | 32682/48008 [1:52:06<48:58, 5.21it/s] 68%|██████▊ | 32683/48008 [1:52:06<49:52, 5.12it/s] 68%|██████▊ | 32684/48008 [1:52:06<49:14, 5.19it/s] 68%|██████▊ | 32685/48008 [1:52:06<49:37, 5.15it/s] 68%|██████▊ | 32686/48008 [1:52:07<49:53, 5.12it/s] 68%|██████▊ | 32687/48008 [1:52:07<53:03, 4.81it/s] 68%|██████▊ | 32688/48008 [1:52:07<52:04, 4.90it/s] 68%|██████▊ | 32689/48008 [1:52:07<52:29, 4.86it/s] 68%|██████▊ | 32690/48008 [1:52:08<51:05, 5.00it/s] 68%|██████▊ | 32691/48008 [1:52:08<51:24, 4.97it/s] 68%|██████▊ | 32692/48008 [1:52:08<54:04, 4.72it/s] 68%|██████▊ | 32693/48008 [1:52:08<52:25, 4.87it/s] 68%|██████▊ | 32694/48008 [1:52:08<52:07, 4.90it/s] 68%|██████▊ | 32695/48008 [1:52:09<52:05, 4.90it/s] 68%|██████▊ | 32696/48008 [1:52:09<50:52, 5.02it/s] 68%|██████▊ | 32697/48008 [1:52:09<49:51, 5.12it/s] 68%|██████▊ | 32698/48008 [1:52:09<49:33, 5.15it/s] 68%|██████▊ | 32699/48008 [1:52:09<50:07, 5.09it/s] 68%|██████▊ | 32700/48008 [1:52:10<50:15, 5.08it/s] {'loss': 0.3013, 'grad_norm': 0.306640625, 'learning_rate': 6.377687052157973e-05, 'epoch': 0.68} 68%|██████▊ | 32700/48008 [1:52:10<50:15, 5.08it/s] 68%|██████▊ | 32701/48008 [1:52:10<49:21, 5.17it/s] 68%|██████▊ | 32702/48008 [1:52:10<49:13, 5.18it/s] 68%|██████▊ | 32703/48008 [1:52:10<49:52, 5.11it/s] 68%|██████▊ | 32704/48008 [1:52:10<50:06, 5.09it/s] 68%|██████▊ | 32705/48008 [1:52:10<49:33, 5.15it/s] 68%|██████▊ | 32706/48008 [1:52:11<50:14, 5.08it/s] 68%|██████▊ | 32707/48008 [1:52:11<49:28, 5.15it/s] 68%|██████▊ | 32708/48008 [1:52:11<49:17, 5.17it/s] 68%|██████▊ | 32709/48008 [1:52:11<49:15, 5.18it/s] 68%|██████▊ | 32710/48008 [1:52:11<49:06, 5.19it/s] 68%|██████▊ | 32711/48008 [1:52:12<48:56, 5.21it/s] 68%|██████▊ | 32712/48008 [1:52:12<48:45, 5.23it/s] 68%|██████▊ | 32713/48008 [1:52:12<48:32, 5.25it/s] 68%|██████▊ | 32714/48008 [1:52:12<48:35, 5.25it/s] 68%|██████▊ | 32715/48008 [1:52:12<48:40, 5.24it/s] 68%|██████▊ | 32716/48008 [1:52:13<48:39, 5.24it/s] 68%|██████▊ | 32717/48008 [1:52:13<52:08, 4.89it/s] 68%|██████▊ | 32718/48008 [1:52:13<50:51, 5.01it/s] 68%|██████▊ | 32719/48008 [1:52:13<53:52, 4.73it/s] 68%|██████▊ | 32720/48008 [1:52:13<52:19, 4.87it/s] 68%|██████▊ | 32721/48008 [1:52:14<51:12, 4.98it/s] 68%|██████▊ | 32722/48008 [1:52:14<50:24, 5.05it/s] 68%|██████▊ | 32723/48008 [1:52:14<49:54, 5.10it/s] 68%|██████▊ | 32724/48008 [1:52:14<49:06, 5.19it/s] 68%|██████▊ | 32725/48008 [1:52:14<48:51, 5.21it/s] 68%|██████▊ | 32726/48008 [1:52:15<48:48, 5.22it/s] 68%|██████▊ | 32727/48008 [1:52:15<48:33, 5.25it/s] 68%|██████▊ | 32728/48008 [1:52:15<48:35, 5.24it/s] 68%|██████▊ | 32729/48008 [1:52:15<48:18, 5.27it/s] 68%|██████▊ | 32730/48008 [1:52:15<49:11, 5.18it/s] 68%|██████▊ | 32731/48008 [1:52:16<49:07, 5.18it/s] 68%|██████▊ | 32732/48008 [1:52:16<48:42, 5.23it/s] 68%|██████▊ | 32733/48008 [1:52:16<48:55, 5.20it/s] 68%|██████▊ | 32734/48008 [1:52:16<48:25, 5.26it/s] 68%|██████▊ | 32735/48008 [1:52:16<49:00, 5.19it/s] 68%|██████▊ | 32736/48008 [1:52:17<52:22, 4.86it/s] 68%|██████▊ | 32737/48008 [1:52:17<51:06, 4.98it/s] 68%|██████▊ | 32738/48008 [1:52:17<50:31, 5.04it/s] 68%|██████▊ | 32739/48008 [1:52:17<50:29, 5.04it/s] 68%|██████▊ | 32740/48008 [1:52:17<49:53, 5.10it/s] 68%|██████▊ | 32741/48008 [1:52:18<49:31, 5.14it/s] 68%|██████▊ | 32742/48008 [1:52:18<49:12, 5.17it/s] 68%|██████▊ | 32743/48008 [1:52:18<49:51, 5.10it/s] 68%|██████▊ | 32744/48008 [1:52:18<49:07, 5.18it/s] 68%|██████▊ | 32745/48008 [1:52:18<52:39, 4.83it/s] 68%|██████▊ | 32746/48008 [1:52:19<52:01, 4.89it/s] 68%|██████▊ | 32747/48008 [1:52:19<51:50, 4.91it/s] 68%|██████▊ | 32748/48008 [1:52:19<50:53, 5.00it/s] 68%|██████▊ | 32749/48008 [1:52:19<50:02, 5.08it/s] 68%|██████▊ | 32750/48008 [1:52:19<50:05, 5.08it/s] {'loss': 0.301, 'grad_norm': 0.1884765625, 'learning_rate': 6.356857190468256e-05, 'epoch': 0.68} 68%|██████▊ | 32750/48008 [1:52:19<50:05, 5.08it/s] 68%|██████▊ | 32751/48008 [1:52:19<49:32, 5.13it/s] 68%|██████▊ | 32752/48008 [1:52:20<49:18, 5.16it/s] 68%|██████▊ | 32753/48008 [1:52:20<49:02, 5.19it/s] 68%|██████▊ | 32754/48008 [1:52:20<49:05, 5.18it/s] 68%|██████▊ | 32755/48008 [1:52:20<49:52, 5.10it/s] 68%|██████▊ | 32756/48008 [1:52:21<52:54, 4.80it/s] 68%|██████▊ | 32757/48008 [1:52:21<51:38, 4.92it/s] 68%|██████▊ | 32758/48008 [1:52:21<50:33, 5.03it/s] 68%|██████▊ | 32759/48008 [1:52:21<50:03, 5.08it/s] 68%|██████▊ | 32760/48008 [1:52:21<49:29, 5.13it/s] 68%|██████▊ | 32761/48008 [1:52:21<49:10, 5.17it/s] 68%|██████▊ | 32762/48008 [1:52:22<48:40, 5.22it/s] 68%|██████▊ | 32763/48008 [1:52:22<48:49, 5.20it/s] 68%|██████▊ | 32764/48008 [1:52:22<48:45, 5.21it/s] 68%|██████▊ | 32765/48008 [1:52:22<48:40, 5.22it/s] 68%|██████▊ | 32766/48008 [1:52:22<49:34, 5.12it/s] 68%|██████▊ | 32767/48008 [1:52:23<49:03, 5.18it/s] 68%|██████▊ | 32768/48008 [1:52:23<49:21, 5.15it/s] 68%|██████▊ | 32769/48008 [1:52:23<48:47, 5.21it/s] 68%|██████▊ | 32770/48008 [1:52:23<49:03, 5.18it/s] 68%|██████▊ | 32771/48008 [1:52:23<50:16, 5.05it/s] 68%|██████▊ | 32772/48008 [1:52:24<49:40, 5.11it/s] 68%|██████▊ | 32773/48008 [1:52:24<49:17, 5.15it/s] 68%|██████▊ | 32774/48008 [1:52:24<49:02, 5.18it/s] 68%|██████▊ | 32775/48008 [1:52:24<48:47, 5.20it/s] 68%|██████▊ | 32776/48008 [1:52:24<48:41, 5.21it/s] 68%|██████▊ | 32777/48008 [1:52:25<48:44, 5.21it/s] 68%|██████▊ | 32778/48008 [1:52:25<48:43, 5.21it/s] 68%|██████▊ | 32779/48008 [1:52:25<48:37, 5.22it/s] 68%|██████▊ | 32780/48008 [1:52:25<48:22, 5.25it/s] 68%|██████▊ | 32781/48008 [1:52:25<48:26, 5.24it/s] 68%|██████▊ | 32782/48008 [1:52:26<48:28, 5.24it/s] 68%|██████▊ | 32783/48008 [1:52:26<49:13, 5.16it/s] 68%|██████▊ | 32784/48008 [1:52:26<48:50, 5.19it/s] 68%|██████▊ | 32785/48008 [1:52:26<48:46, 5.20it/s] 68%|██████▊ | 32786/48008 [1:52:26<48:23, 5.24it/s] 68%|██████▊ | 32787/48008 [1:52:27<2:02:05, 2.08it/s] 68%|██████▊ | 32788/48008 [1:52:28<1:54:55, 2.21it/s] 68%|██████▊ | 32789/48008 [1:52:28<1:34:50, 2.67it/s] 68%|██████▊ | 32790/48008 [1:52:28<1:20:51, 3.14it/s] 68%|██████▊ | 32791/48008 [1:52:28<1:10:50, 3.58it/s] 68%|██████▊ | 32792/48008 [1:52:29<1:03:45, 3.98it/s] 68%|██████▊ | 32793/48008 [1:52:29<59:16, 4.28it/s] 68%|██████▊ | 32794/48008 [1:52:29<55:54, 4.54it/s] 68%|██████▊ | 32795/48008 [1:52:29<53:40, 4.72it/s] 68%|██████▊ | 32796/48008 [1:52:29<52:10, 4.86it/s] 68%|██████▊ | 32797/48008 [1:52:30<51:02, 4.97it/s] 68%|██████▊ | 32798/48008 [1:52:30<50:35, 5.01it/s] 68%|██████▊ | 32799/48008 [1:52:30<53:17, 4.76it/s] 68%|██████▊ | 32800/48008 [1:52:30<51:44, 4.90it/s] {'loss': 0.3029, 'grad_norm': 0.1591796875, 'learning_rate': 6.336027328778538e-05, 'epoch': 0.68} 68%|██████▊ | 32800/48008 [1:52:30<51:44, 4.90it/s] 68%|██████▊ | 32801/48008 [1:52:30<50:45, 4.99it/s] 68%|██████▊ | 32802/48008 [1:52:31<53:38, 4.72it/s] 68%|██████▊ | 32803/48008 [1:52:31<53:01, 4.78it/s] 68%|██████▊ | 32804/48008 [1:52:31<51:33, 4.92it/s] 68%|██████▊ | 32805/48008 [1:52:31<50:49, 4.99it/s] 68%|██████▊ | 32806/48008 [1:52:31<53:42, 4.72it/s] 68%|██████▊ | 32807/48008 [1:52:32<51:43, 4.90it/s] 68%|██████▊ | 32808/48008 [1:52:32<51:06, 4.96it/s] 68%|██████▊ | 32809/48008 [1:52:32<50:35, 5.01it/s] 68%|██████▊ | 32810/48008 [1:52:32<49:43, 5.09it/s] 68%|██████▊ | 32811/48008 [1:52:32<49:26, 5.12it/s] 68%|██████▊ | 32812/48008 [1:52:33<49:30, 5.12it/s] 68%|██████▊ | 32813/48008 [1:52:33<1:43:02, 2.46it/s] 68%|██████▊ | 32814/48008 [1:52:34<1:26:35, 2.92it/s] 68%|██████▊ | 32815/48008 [1:52:34<1:14:59, 3.38it/s] 68%|██████▊ | 32816/48008 [1:52:34<1:08:34, 3.69it/s] 68%|██████▊ | 32817/48008 [1:52:34<1:02:25, 4.06it/s] 68%|██████▊ | 32818/48008 [1:52:34<57:50, 4.38it/s] 68%|██████▊ | 32819/48008 [1:52:35<1:54:53, 2.20it/s] 68%|██████▊ | 32820/48008 [1:52:36<1:34:57, 2.67it/s] 68%|██████▊ | 32821/48008 [1:52:36<1:20:56, 3.13it/s] 68%|██████▊ | 32822/48008 [1:52:36<1:11:27, 3.54it/s] 68%|██████▊ | 32823/48008 [1:52:36<1:04:56, 3.90it/s] 68%|██████▊ | 32824/48008 [1:52:36<1:00:21, 4.19it/s] 68%|██████▊ | 32825/48008 [1:52:37<1:43:56, 2.43it/s] 68%|██████▊ | 32826/48008 [1:52:37<1:27:38, 2.89it/s] 68%|██████▊ | 32827/48008 [1:52:38<1:15:40, 3.34it/s] 68%|██████▊ | 32828/48008 [1:52:38<1:07:34, 3.74it/s] 68%|██████▊ | 32829/48008 [1:52:38<1:02:54, 4.02it/s] 68%|██████▊ | 32830/48008 [1:52:38<1:00:14, 4.20it/s] 68%|██████▊ | 32831/48008 [1:52:38<56:46, 4.45it/s] 68%|██████▊ | 32832/48008 [1:52:39<54:50, 4.61it/s] 68%|██████▊ | 32833/48008 [1:52:39<52:52, 4.78it/s] 68%|██████▊ | 32834/48008 [1:52:39<52:01, 4.86it/s] 68%|██████▊ | 32835/48008 [1:52:39<50:42, 4.99it/s] 68%|██████▊ | 32836/48008 [1:52:39<50:16, 5.03it/s] 68%|██████▊ | 32837/48008 [1:52:40<50:37, 4.99it/s] 68%|██████▊ | 32838/48008 [1:52:40<49:54, 5.07it/s] 68%|██████▊ | 32839/48008 [1:52:40<49:17, 5.13it/s] 68%|██████▊ | 32840/48008 [1:52:40<49:03, 5.15it/s] 68%|██████▊ | 32841/48008 [1:52:40<48:33, 5.21it/s] 68%|██████▊ | 32842/48008 [1:52:41<48:42, 5.19it/s] 68%|██████▊ | 32843/48008 [1:52:41<49:33, 5.10it/s] 68%|██████▊ | 32844/48008 [1:52:41<49:06, 5.15it/s] 68%|██████▊ | 32845/48008 [1:52:41<48:45, 5.18it/s] 68%|██████▊ | 32846/48008 [1:52:41<49:27, 5.11it/s] 68%|██████▊ | 32847/48008 [1:52:41<48:58, 5.16it/s] 68%|██████▊ | 32848/48008 [1:52:42<48:29, 5.21it/s] 68%|██████▊ | 32849/48008 [1:52:42<48:30, 5.21it/s] 68%|██████▊ | 32850/48008 [1:52:42<48:17, 5.23it/s] {'loss': 0.3002, 'grad_norm': 0.412109375, 'learning_rate': 6.31519746708882e-05, 'epoch': 0.68} 68%|██████▊ | 32850/48008 [1:52:42<48:17, 5.23it/s] 68%|██████▊ | 32851/48008 [1:52:42<47:58, 5.27it/s] 68%|██████▊ | 32852/48008 [1:52:42<49:03, 5.15it/s] 68%|██████▊ | 32853/48008 [1:52:43<48:41, 5.19it/s] 68%|██████▊ | 32854/48008 [1:52:43<49:04, 5.15it/s] 68%|██████▊ | 32855/48008 [1:52:43<48:51, 5.17it/s] 68%|██████▊ | 32856/48008 [1:52:43<49:00, 5.15it/s] 68%|██████▊ | 32857/48008 [1:52:43<49:09, 5.14it/s] 68%|██████▊ | 32858/48008 [1:52:44<48:54, 5.16it/s] 68%|██████▊ | 32859/48008 [1:52:44<48:36, 5.19it/s] 68%|██████▊ | 32860/48008 [1:52:44<49:21, 5.12it/s] 68%|██████▊ | 32861/48008 [1:52:45<1:37:45, 2.58it/s] 68%|██████▊ | 32862/48008 [1:52:45<1:23:16, 3.03it/s] 68%|██████▊ | 32863/48008 [1:52:45<1:12:47, 3.47it/s] 68%|██████▊ | 32864/48008 [1:52:45<1:05:20, 3.86it/s] 68%|██████▊ | 32865/48008 [1:52:46<1:08:55, 3.66it/s] 68%|██████▊ | 32866/48008 [1:52:46<1:02:36, 4.03it/s] 68%|██████▊ | 32867/48008 [1:52:46<59:45, 4.22it/s] 68%|██████▊ | 32868/48008 [1:52:46<55:42, 4.53it/s] 68%|██████▊ | 32869/48008 [1:52:47<56:59, 4.43it/s] 68%|██████▊ | 32870/48008 [1:52:47<54:39, 4.62it/s] 68%|██████▊ | 32871/48008 [1:52:47<52:46, 4.78it/s] 68%|██████▊ | 32872/48008 [1:52:47<51:07, 4.93it/s] 68%|██████▊ | 32873/48008 [1:52:47<51:38, 4.89it/s] 68%|██████▊ | 32874/48008 [1:52:48<50:25, 5.00it/s] 68%|██████▊ | 32875/48008 [1:52:48<49:16, 5.12it/s] 68%|██████▊ | 32876/48008 [1:52:48<49:30, 5.09it/s] 68%|██████▊ | 32877/48008 [1:52:48<49:13, 5.12it/s] 68%|██████▊ | 32878/48008 [1:52:48<50:25, 5.00it/s] 68%|██████▊ | 32879/48008 [1:52:48<49:40, 5.08it/s] 68%|██████▊ | 32880/48008 [1:52:49<48:44, 5.17it/s] 68%|██████▊ | 32881/48008 [1:52:49<50:10, 5.02it/s] 68%|██████▊ | 32882/48008 [1:52:49<49:41, 5.07it/s] 68%|██████▊ | 32883/48008 [1:52:49<50:11, 5.02it/s] 68%|██████▊ | 32884/48008 [1:52:49<49:48, 5.06it/s] 68%|██████▊ | 32885/48008 [1:52:50<49:59, 5.04it/s] 69%|██████▊ | 32886/48008 [1:52:50<49:10, 5.12it/s] 69%|██████▊ | 32887/48008 [1:52:50<48:33, 5.19it/s] 69%|██████▊ | 32888/48008 [1:52:50<48:05, 5.24it/s] 69%|██████▊ | 32889/48008 [1:52:50<47:51, 5.27it/s] 69%|██████▊ | 32890/48008 [1:52:51<48:24, 5.20it/s] 69%|██████▊ | 32891/48008 [1:52:51<48:19, 5.21it/s] 69%|██████▊ | 32892/48008 [1:52:51<49:07, 5.13it/s] 69%|██████▊ | 32893/48008 [1:52:51<52:18, 4.82it/s] 69%|██████▊ | 32894/48008 [1:52:51<51:58, 4.85it/s] 69%|██████▊ | 32895/48008 [1:52:52<51:37, 4.88it/s] 69%|██████▊ | 32896/48008 [1:52:52<54:06, 4.65it/s] 69%|██████▊ | 32897/48008 [1:52:52<53:10, 4.74it/s] 69%|██████▊ | 32898/48008 [1:52:52<51:46, 4.86it/s] 69%|██████▊ | 32899/48008 [1:52:52<51:30, 4.89it/s] 69%|██████▊ | 32900/48008 [1:52:53<1:05:19, 3.85it/s] {'loss': 0.2987, 'grad_norm': 0.1865234375, 'learning_rate': 6.294367605399101e-05, 'epoch': 0.69} 69%|██████▊ | 32900/48008 [1:52:53<1:05:19, 3.85it/s] 69%|██████▊ | 32901/48008 [1:52:53<1:00:42, 4.15it/s] 69%|██████▊ | 32902/48008 [1:52:53<57:26, 4.38it/s] 69%|██████▊ | 32903/48008 [1:52:53<54:39, 4.61it/s] 69%|██████▊ | 32904/48008 [1:52:54<52:35, 4.79it/s] 69%|██████▊ | 32905/48008 [1:52:54<51:19, 4.90it/s] 69%|██████▊ | 32906/48008 [1:52:54<50:02, 5.03it/s] 69%|██████▊ | 32907/48008 [1:52:54<49:09, 5.12it/s] 69%|██████▊ | 32908/48008 [1:52:54<49:35, 5.07it/s] 69%|██████▊ | 32909/48008 [1:52:55<48:48, 5.16it/s] 69%|██████▊ | 32910/48008 [1:52:55<49:18, 5.10it/s] 69%|██████▊ | 32911/48008 [1:52:55<49:10, 5.12it/s] 69%|██████▊ | 32912/48008 [1:52:55<48:32, 5.18it/s] 69%|██████▊ | 32913/48008 [1:52:55<48:21, 5.20it/s] 69%|██████▊ | 32914/48008 [1:52:56<48:06, 5.23it/s] 69%|██████▊ | 32915/48008 [1:52:56<48:48, 5.15it/s] 69%|██████▊ | 32916/48008 [1:52:56<49:12, 5.11it/s] 69%|██████▊ | 32917/48008 [1:52:56<48:36, 5.17it/s] 69%|██████▊ | 32918/48008 [1:52:56<48:27, 5.19it/s] 69%|██████▊ | 32919/48008 [1:52:57<48:17, 5.21it/s] 69%|██████▊ | 32920/48008 [1:52:57<48:41, 5.16it/s] 69%|██████▊ | 32921/48008 [1:52:57<48:08, 5.22it/s] 69%|██████▊ | 32922/48008 [1:52:57<48:36, 5.17it/s] 69%|██████▊ | 32923/48008 [1:52:57<51:54, 4.84it/s] 69%|██████▊ | 32924/48008 [1:52:58<50:54, 4.94it/s] 69%|██████▊ | 32925/48008 [1:52:58<50:08, 5.01it/s] 69%|██████▊ | 32926/48008 [1:52:58<50:00, 5.03it/s] 69%|██████▊ | 32927/48008 [1:52:58<48:53, 5.14it/s] 69%|██████▊ | 32928/48008 [1:52:58<48:33, 5.18it/s] 69%|██████▊ | 32929/48008 [1:52:59<48:19, 5.20it/s] 69%|██████▊ | 32930/48008 [1:52:59<51:48, 4.85it/s] 69%|██████▊ | 32931/48008 [1:52:59<51:00, 4.93it/s] 69%|██████▊ | 32932/48008 [1:52:59<50:12, 5.00it/s] 69%|██████▊ | 32933/48008 [1:52:59<49:36, 5.07it/s] 69%|██████▊ | 32934/48008 [1:53:00<49:53, 5.04it/s] 69%|██████▊ | 32935/48008 [1:53:00<48:57, 5.13it/s] 69%|██████▊ | 32936/48008 [1:53:00<48:26, 5.19it/s] 69%|██████▊ | 32937/48008 [1:53:00<49:10, 5.11it/s] 69%|██████▊ | 32938/48008 [1:53:00<49:09, 5.11it/s] 69%|██████▊ | 32939/48008 [1:53:00<48:41, 5.16it/s] 69%|██████▊ | 32940/48008 [1:53:01<49:39, 5.06it/s] 69%|██████▊ | 32941/48008 [1:53:01<49:08, 5.11it/s] 69%|██████▊ | 32942/48008 [1:53:01<48:32, 5.17it/s] 69%|██████▊ | 32943/48008 [1:53:01<57:04, 4.40it/s] 69%|██████▊ | 32944/48008 [1:53:02<54:46, 4.58it/s] 69%|██████▊ | 32945/48008 [1:53:02<52:17, 4.80it/s] 69%|██████▊ | 32946/48008 [1:53:02<50:59, 4.92it/s] 69%|██████▊ | 32947/48008 [1:53:02<50:56, 4.93it/s] 69%|██████▊ | 32948/48008 [1:53:02<51:28, 4.88it/s] 69%|██████▊ | 32949/48008 [1:53:03<50:56, 4.93it/s] 69%|██████▊ | 32950/48008 [1:53:03<49:42, 5.05it/s] {'loss': 0.3015, 'grad_norm': 0.14453125, 'learning_rate': 6.273537743709383e-05, 'epoch': 0.69} 69%|██████▊ | 32950/48008 [1:53:03<49:42, 5.05it/s] 69%|██████▊ | 32951/48008 [1:53:03<48:54, 5.13it/s] 69%|██████▊ | 32952/48008 [1:53:03<50:04, 5.01it/s] 69%|██████▊ | 32953/48008 [1:53:03<52:55, 4.74it/s] 69%|██████▊ | 32954/48008 [1:53:04<51:27, 4.88it/s] 69%|██████▊ | 32955/48008 [1:53:04<50:20, 4.98it/s] 69%|██████▊ | 32956/48008 [1:53:04<51:00, 4.92it/s] 69%|██████▊ | 32957/48008 [1:53:04<50:01, 5.01it/s] 69%|██████▊ | 32958/48008 [1:53:04<50:21, 4.98it/s] 69%|██████▊ | 32959/48008 [1:53:05<49:29, 5.07it/s] 69%|██████▊ | 32960/48008 [1:53:05<48:49, 5.14it/s] 69%|██████▊ | 32961/48008 [1:53:05<48:27, 5.17it/s] 69%|██████▊ | 32962/48008 [1:53:05<48:18, 5.19it/s] 69%|██████▊ | 32963/48008 [1:53:05<49:05, 5.11it/s] 69%|██████▊ | 32964/48008 [1:53:06<50:05, 5.00it/s] 69%|██████▊ | 32965/48008 [1:53:06<49:54, 5.02it/s] 69%|██████▊ | 32966/48008 [1:53:06<49:40, 5.05it/s] 69%|██████▊ | 32967/48008 [1:53:06<49:11, 5.10it/s] 69%|██████▊ | 32968/48008 [1:53:06<50:10, 5.00it/s] 69%|██████▊ | 32969/48008 [1:53:07<49:24, 5.07it/s] 69%|██████▊ | 32970/48008 [1:53:07<48:52, 5.13it/s] 69%|██████▊ | 32971/48008 [1:53:07<48:23, 5.18it/s] 69%|██████▊ | 32972/48008 [1:53:07<48:21, 5.18it/s] 69%|██████▊ | 32973/48008 [1:53:07<48:56, 5.12it/s] 69%|██████▊ | 32974/48008 [1:53:07<48:37, 5.15it/s] 69%|██████▊ | 32975/48008 [1:53:08<48:13, 5.20it/s] 69%|██████▊ | 32976/48008 [1:53:08<48:32, 5.16it/s] 69%|██████▊ | 32977/48008 [1:53:08<48:27, 5.17it/s] 69%|██████▊ | 32978/48008 [1:53:08<48:02, 5.21it/s] 69%|██████▊ | 32979/48008 [1:53:08<48:08, 5.20it/s] 69%|██████▊ | 32980/48008 [1:53:09<47:58, 5.22it/s] 69%|██████▊ | 32981/48008 [1:53:09<47:33, 5.27it/s] 69%|██████▊ | 32982/48008 [1:53:09<47:34, 5.26it/s] 69%|██████▊ | 32983/48008 [1:53:09<48:03, 5.21it/s] 69%|██████▊ | 32984/48008 [1:53:09<49:29, 5.06it/s] 69%|██████▊ | 32985/48008 [1:53:10<48:46, 5.13it/s] 69%|██████▊ | 32986/48008 [1:53:10<48:28, 5.16it/s] 69%|██████▊ | 32987/48008 [1:53:10<49:02, 5.10it/s] 69%|██████▊ | 32988/48008 [1:53:10<48:50, 5.12it/s] 69%|██████▊ | 32989/48008 [1:53:10<48:26, 5.17it/s] 69%|██████▊ | 32990/48008 [1:53:11<51:43, 4.84it/s] 69%|██████▊ | 32991/48008 [1:53:11<50:29, 4.96it/s] 69%|██████▊ | 32992/48008 [1:53:11<49:36, 5.04it/s] 69%|██████▊ | 32993/48008 [1:53:11<48:53, 5.12it/s] 69%|██████▊ | 32994/48008 [1:53:11<49:30, 5.05it/s] 69%|██████▊ | 32995/48008 [1:53:12<48:55, 5.11it/s] 69%|██████▊ | 32996/48008 [1:53:12<48:26, 5.17it/s] 69%|██████▊ | 32997/48008 [1:53:12<47:57, 5.22it/s] 69%|██████▊ | 32998/48008 [1:53:12<47:20, 5.29it/s] 69%|██████▊ | 32999/48008 [1:53:12<47:05, 5.31it/s] 69%|██████▊ | 33000/48008 [1:53:13<47:04, 5.31it/s] {'loss': 0.3016, 'grad_norm': 0.15625, 'learning_rate': 6.252707882019663e-05, 'epoch': 0.69} 69%|██████▊ | 33000/48008 [1:53:13<47:04, 5.31it/s] 69%|██████▊ | 33001/48008 [1:53:13<47:54, 5.22it/s] 69%|██████▊ | 33002/48008 [1:53:13<47:58, 5.21it/s] 69%|██████▊ | 33003/48008 [1:53:13<48:40, 5.14it/s] 69%|██████▊ | 33004/48008 [1:53:13<48:03, 5.20it/s] 69%|██████▊ | 33005/48008 [1:53:13<47:36, 5.25it/s] 69%|██████▉ | 33006/48008 [1:53:14<47:34, 5.25it/s] 69%|██████▉ | 33007/48008 [1:53:14<47:45, 5.23it/s] 69%|██████▉ | 33008/48008 [1:53:14<47:23, 5.27it/s] 69%|██████▉ | 33009/48008 [1:53:14<47:27, 5.27it/s] 69%|██████▉ | 33010/48008 [1:53:14<47:39, 5.24it/s] 69%|██████▉ | 33011/48008 [1:53:15<1:02:23, 4.01it/s] 69%|██████▉ | 33012/48008 [1:53:15<58:10, 4.30it/s] 69%|██████▉ | 33013/48008 [1:53:15<58:32, 4.27it/s] 69%|██████▉ | 33014/48008 [1:53:15<55:18, 4.52it/s] 69%|██████▉ | 33015/48008 [1:53:16<53:20, 4.68it/s] 69%|██████▉ | 33016/48008 [1:53:16<51:31, 4.85it/s] 69%|██████▉ | 33017/48008 [1:53:16<50:51, 4.91it/s] 69%|██████▉ | 33018/48008 [1:53:16<49:40, 5.03it/s] 69%|██████▉ | 33019/48008 [1:53:16<49:35, 5.04it/s] 69%|██████▉ | 33020/48008 [1:53:17<48:42, 5.13it/s] 69%|██████▉ | 33021/48008 [1:53:17<48:51, 5.11it/s] 69%|██████▉ | 33022/48008 [1:53:17<48:34, 5.14it/s] 69%|██████▉ | 33023/48008 [1:53:17<48:16, 5.17it/s] 69%|██████▉ | 33024/48008 [1:53:17<48:19, 5.17it/s] 69%|██████▉ | 33025/48008 [1:53:18<48:17, 5.17it/s] 69%|██████▉ | 33026/48008 [1:53:18<48:13, 5.18it/s] 69%|██████▉ | 33027/48008 [1:53:18<48:22, 5.16it/s] 69%|██████▉ | 33028/48008 [1:53:18<47:47, 5.22it/s] 69%|██████▉ | 33029/48008 [1:53:18<48:11, 5.18it/s] 69%|██████▉ | 33030/48008 [1:53:19<48:40, 5.13it/s] 69%|██████▉ | 33031/48008 [1:53:19<48:53, 5.10it/s] 69%|██████▉ | 33032/48008 [1:53:19<49:46, 5.01it/s] 69%|██████▉ | 33033/48008 [1:53:19<49:09, 5.08it/s] 69%|██████▉ | 33034/48008 [1:53:19<49:08, 5.08it/s] 69%|██████▉ | 33035/48008 [1:53:20<48:37, 5.13it/s] 69%|██████▉ | 33036/48008 [1:53:20<48:54, 5.10it/s] 69%|██████▉ | 33037/48008 [1:53:20<48:55, 5.10it/s] 69%|██████▉ | 33038/48008 [1:53:20<48:42, 5.12it/s] 69%|██████▉ | 33039/48008 [1:53:20<48:20, 5.16it/s] 69%|██████▉ | 33040/48008 [1:53:20<48:10, 5.18it/s] 69%|██████▉ | 33041/48008 [1:53:21<48:47, 5.11it/s] 69%|██████▉ | 33042/48008 [1:53:21<48:45, 5.12it/s] 69%|██████▉ | 33043/48008 [1:53:21<49:13, 5.07it/s] 69%|██████▉ | 33044/48008 [1:53:21<49:33, 5.03it/s] 69%|██████▉ | 33045/48008 [1:53:21<48:43, 5.12it/s] 69%|██████▉ | 33046/48008 [1:53:22<49:11, 5.07it/s] 69%|██████▉ | 33047/48008 [1:53:22<49:32, 5.03it/s] 69%|██████▉ | 33048/48008 [1:53:22<48:27, 5.15it/s] 69%|██████▉ | 33049/48008 [1:53:22<48:41, 5.12it/s] 69%|██████▉ | 33050/48008 [1:53:22<47:59, 5.19it/s] {'loss': 0.2995, 'grad_norm': 0.166015625, 'learning_rate': 6.231878020329945e-05, 'epoch': 0.69} 69%|██████▉ | 33050/48008 [1:53:22<47:59, 5.19it/s] 69%|██████▉ | 33051/48008 [1:53:23<47:58, 5.20it/s] 69%|██████▉ | 33052/48008 [1:53:23<48:41, 5.12it/s] 69%|██████▉ | 33053/48008 [1:53:23<48:21, 5.15it/s] 69%|██████▉ | 33054/48008 [1:53:23<48:10, 5.17it/s] 69%|██████▉ | 33055/48008 [1:53:23<48:37, 5.13it/s] 69%|██████▉ | 33056/48008 [1:53:24<48:10, 5.17it/s] 69%|██████▉ | 33057/48008 [1:53:24<48:02, 5.19it/s] 69%|██████▉ | 33058/48008 [1:53:24<47:51, 5.21it/s] 69%|██████▉ | 33059/48008 [1:53:24<47:44, 5.22it/s] 69%|██████▉ | 33060/48008 [1:53:24<47:30, 5.24it/s] 69%|██████▉ | 33061/48008 [1:53:25<47:18, 5.27it/s] 69%|██████▉ | 33062/48008 [1:53:25<47:24, 5.25it/s] 69%|██████▉ | 33063/48008 [1:53:25<47:42, 5.22it/s] 69%|██████▉ | 33064/48008 [1:53:26<1:37:24, 2.56it/s] 69%|██████▉ | 33065/48008 [1:53:26<1:22:57, 3.00it/s] 69%|██████▉ | 33066/48008 [1:53:26<1:12:22, 3.44it/s] 69%|██████▉ | 33067/48008 [1:53:26<1:05:02, 3.83it/s] 69%|██████▉ | 33068/48008 [1:53:27<59:36, 4.18it/s] 69%|██████▉ | 33069/48008 [1:53:27<56:01, 4.44it/s] 69%|██████▉ | 33070/48008 [1:53:27<54:30, 4.57it/s] 69%|██████▉ | 33071/48008 [1:53:27<52:10, 4.77it/s] 69%|██████▉ | 33072/48008 [1:53:27<50:47, 4.90it/s] 69%|██████▉ | 33073/48008 [1:53:28<58:33, 4.25it/s] 69%|██████▉ | 33074/48008 [1:53:28<56:08, 4.43it/s] 69%|██████▉ | 33075/48008 [1:53:28<53:31, 4.65it/s] 69%|██████▉ | 33076/48008 [1:53:28<51:42, 4.81it/s] 69%|██████▉ | 33077/48008 [1:53:28<51:19, 4.85it/s] 69%|██████▉ | 33078/48008 [1:53:29<50:48, 4.90it/s] 69%|██████▉ | 33079/48008 [1:53:29<50:21, 4.94it/s] 69%|██████▉ | 33080/48008 [1:53:29<49:27, 5.03it/s] 69%|██████▉ | 33081/48008 [1:53:29<48:51, 5.09it/s] 69%|██████▉ | 33082/48008 [1:53:29<51:59, 4.79it/s] 69%|██████▉ | 33083/48008 [1:53:30<50:39, 4.91it/s] 69%|██████▉ | 33084/48008 [1:53:30<49:31, 5.02it/s] 69%|██████▉ | 33085/48008 [1:53:30<50:00, 4.97it/s] 69%|██████▉ | 33086/48008 [1:53:30<49:10, 5.06it/s] 69%|██████▉ | 33087/48008 [1:53:30<48:19, 5.15it/s] 69%|██████▉ | 33088/48008 [1:53:31<51:25, 4.84it/s] 69%|██████▉ | 33089/48008 [1:53:31<50:47, 4.89it/s] 69%|██████▉ | 33090/48008 [1:53:31<50:50, 4.89it/s] 69%|██████▉ | 33091/48008 [1:53:31<49:42, 5.00it/s] 69%|██████▉ | 33092/48008 [1:53:31<49:17, 5.04it/s] 69%|██████▉ | 33093/48008 [1:53:32<48:37, 5.11it/s] 69%|██████▉ | 33094/48008 [1:53:32<49:16, 5.04it/s] 69%|██████▉ | 33095/48008 [1:53:32<48:19, 5.14it/s] 69%|██████▉ | 33096/48008 [1:53:32<48:56, 5.08it/s] 69%|██████▉ | 33097/48008 [1:53:32<48:42, 5.10it/s] 69%|██████▉ | 33098/48008 [1:53:33<1:02:55, 3.95it/s] 69%|██████▉ | 33099/48008 [1:53:33<58:20, 4.26it/s] 69%|██████▉ | 33100/48008 [1:53:33<55:05, 4.51it/s] {'loss': 0.3001, 'grad_norm': 0.294921875, 'learning_rate': 6.211048158640227e-05, 'epoch': 0.69} 69%|██████▉ | 33100/48008 [1:53:33<55:05, 4.51it/s] 69%|██████▉ | 33101/48008 [1:53:33<53:23, 4.65it/s] 69%|██████▉ | 33102/48008 [1:53:34<52:44, 4.71it/s] 69%|██████▉ | 33103/48008 [1:53:34<51:54, 4.79it/s] 69%|██████▉ | 33104/48008 [1:53:34<50:20, 4.93it/s] 69%|██████▉ | 33105/48008 [1:53:34<50:21, 4.93it/s] 69%|██████▉ | 33106/48008 [1:53:34<49:34, 5.01it/s] 69%|██████▉ | 33107/48008 [1:53:35<48:57, 5.07it/s] 69%|██████▉ | 33108/48008 [1:53:35<48:18, 5.14it/s] 69%|██████▉ | 33109/48008 [1:53:35<1:02:30, 3.97it/s] 69%|██████▉ | 33110/48008 [1:53:35<57:31, 4.32it/s] 69%|██████▉ | 33111/48008 [1:53:36<54:47, 4.53it/s] 69%|██████▉ | 33112/48008 [1:53:36<52:43, 4.71it/s] 69%|██████▉ | 33113/48008 [1:53:36<51:19, 4.84it/s] 69%|██████▉ | 33114/48008 [1:53:36<50:26, 4.92it/s] 69%|██████▉ | 33115/48008 [1:53:36<50:06, 4.95it/s] 69%|██████▉ | 33116/48008 [1:53:37<50:12, 4.94it/s] 69%|██████▉ | 33117/48008 [1:53:37<49:36, 5.00it/s] 69%|██████▉ | 33118/48008 [1:53:37<49:04, 5.06it/s] 69%|██████▉ | 33119/48008 [1:53:37<49:02, 5.06it/s] 69%|██████▉ | 33120/48008 [1:53:37<48:30, 5.12it/s] 69%|██████▉ | 33121/48008 [1:53:37<48:05, 5.16it/s] 69%|██████▉ | 33122/48008 [1:53:38<48:41, 5.10it/s] 69%|██████▉ | 33123/48008 [1:53:38<49:22, 5.02it/s] 69%|██████▉ | 33124/48008 [1:53:38<48:47, 5.08it/s] 69%|██████▉ | 33125/48008 [1:53:38<48:19, 5.13it/s] 69%|██████▉ | 33126/48008 [1:53:39<56:36, 4.38it/s] 69%|██████▉ | 33127/48008 [1:53:39<53:40, 4.62it/s] 69%|██████▉ | 33128/48008 [1:53:39<51:57, 4.77it/s] 69%|██████▉ | 33129/48008 [1:53:39<50:44, 4.89it/s] 69%|██████▉ | 33130/48008 [1:53:39<53:10, 4.66it/s] 69%|██████▉ | 33131/48008 [1:53:40<51:30, 4.81it/s] 69%|██████▉ | 33132/48008 [1:53:40<49:57, 4.96it/s] 69%|██████▉ | 33133/48008 [1:53:40<48:44, 5.09it/s] 69%|██████▉ | 33134/48008 [1:53:40<49:18, 5.03it/s] 69%|██████▉ | 33135/48008 [1:53:40<57:15, 4.33it/s] 69%|██████▉ | 33136/48008 [1:53:41<54:14, 4.57it/s] 69%|██████▉ | 33137/48008 [1:53:41<52:09, 4.75it/s] 69%|██████▉ | 33138/48008 [1:53:41<50:46, 4.88it/s] 69%|██████▉ | 33139/48008 [1:53:41<58:37, 4.23it/s] 69%|██████▉ | 33140/48008 [1:53:42<55:39, 4.45it/s] 69%|██████▉ | 33141/48008 [1:53:42<53:05, 4.67it/s] 69%|██████▉ | 33142/48008 [1:53:42<51:22, 4.82it/s] 69%|██████▉ | 33143/48008 [1:53:42<50:15, 4.93it/s] 69%|██████▉ | 33144/48008 [1:53:42<49:49, 4.97it/s] 69%|██████▉ | 33145/48008 [1:53:43<48:59, 5.06it/s] 69%|██████▉ | 33146/48008 [1:53:43<49:26, 5.01it/s] 69%|██████▉ | 33147/48008 [1:53:43<48:57, 5.06it/s] 69%|██████▉ | 33148/48008 [1:53:43<48:33, 5.10it/s] 69%|██████▉ | 33149/48008 [1:53:43<48:10, 5.14it/s] 69%|██████▉ | 33150/48008 [1:53:43<48:01, 5.16it/s] {'loss': 0.2976, 'grad_norm': 0.2060546875, 'learning_rate': 6.190218296950508e-05, 'epoch': 0.69} 69%|██████▉ | 33150/48008 [1:53:43<48:01, 5.16it/s] 69%|██████▉ | 33151/48008 [1:53:44<47:51, 5.17it/s] 69%|██████▉ | 33152/48008 [1:53:44<47:41, 5.19it/s] 69%|██████▉ | 33153/48008 [1:53:44<47:11, 5.25it/s] 69%|██████▉ | 33154/48008 [1:53:44<46:44, 5.30it/s] 69%|██████▉ | 33155/48008 [1:53:44<46:47, 5.29it/s] 69%|██████▉ | 33156/48008 [1:53:45<46:46, 5.29it/s] 69%|██████▉ | 33157/48008 [1:53:45<46:47, 5.29it/s] 69%|██████▉ | 33158/48008 [1:53:45<46:49, 5.28it/s] 69%|██████▉ | 33159/48008 [1:53:45<47:00, 5.26it/s] 69%|██████▉ | 33160/48008 [1:53:45<47:57, 5.16it/s] 69%|██████▉ | 33161/48008 [1:53:46<47:50, 5.17it/s] 69%|██████▉ | 33162/48008 [1:53:46<47:42, 5.19it/s] 69%|██████▉ | 33163/48008 [1:53:46<47:23, 5.22it/s] 69%|██████▉ | 33164/48008 [1:53:46<47:09, 5.25it/s] 69%|██████▉ | 33165/48008 [1:53:46<46:43, 5.30it/s] 69%|██████▉ | 33166/48008 [1:53:47<47:00, 5.26it/s] 69%|██████▉ | 33167/48008 [1:53:47<47:35, 5.20it/s] 69%|██████▉ | 33168/48008 [1:53:47<48:04, 5.14it/s] 69%|██████▉ | 33169/48008 [1:53:47<47:31, 5.20it/s] 69%|██████▉ | 33170/48008 [1:53:47<47:46, 5.18it/s] 69%|██████▉ | 33171/48008 [1:53:47<47:19, 5.22it/s] 69%|██████▉ | 33172/48008 [1:53:48<47:36, 5.19it/s] 69%|██████▉ | 33173/48008 [1:53:48<48:13, 5.13it/s] 69%|██████▉ | 33174/48008 [1:53:48<47:49, 5.17it/s] 69%|██████▉ | 33175/48008 [1:53:48<47:37, 5.19it/s] 69%|██████▉ | 33176/48008 [1:53:48<47:47, 5.17it/s] 69%|██████▉ | 33177/48008 [1:53:49<47:27, 5.21it/s] 69%|██████▉ | 33178/48008 [1:53:49<47:21, 5.22it/s] 69%|██████▉ | 33179/48008 [1:53:49<47:04, 5.25it/s] 69%|██████▉ | 33180/48008 [1:53:49<47:07, 5.24it/s] 69%|██████▉ | 33181/48008 [1:53:49<46:36, 5.30it/s] 69%|██████▉ | 33182/48008 [1:53:50<46:10, 5.35it/s] 69%|██████▉ | 33183/48008 [1:53:50<47:59, 5.15it/s] 69%|██████▉ | 33184/48008 [1:53:50<47:44, 5.18it/s] 69%|██████▉ | 33185/48008 [1:53:50<48:26, 5.10it/s] 69%|██████▉ | 33186/48008 [1:53:50<48:44, 5.07it/s] 69%|██████▉ | 33187/48008 [1:53:51<48:22, 5.11it/s] 69%|██████▉ | 33188/48008 [1:53:51<48:18, 5.11it/s] 69%|██████▉ | 33189/48008 [1:53:51<48:52, 5.05it/s] 69%|██████▉ | 33190/48008 [1:53:51<56:55, 4.34it/s] 69%|██████▉ | 33191/48008 [1:53:51<54:00, 4.57it/s] 69%|██████▉ | 33192/48008 [1:53:52<51:50, 4.76it/s] 69%|██████▉ | 33193/48008 [1:53:52<50:24, 4.90it/s] 69%|██████▉ | 33194/48008 [1:53:52<49:56, 4.94it/s] 69%|██████▉ | 33195/48008 [1:53:52<48:43, 5.07it/s] 69%|██████▉ | 33196/48008 [1:53:52<48:17, 5.11it/s] 69%|██████▉ | 33197/48008 [1:53:53<47:52, 5.16it/s] 69%|██████▉ | 33198/48008 [1:53:53<47:46, 5.17it/s] 69%|██████▉ | 33199/48008 [1:53:53<47:35, 5.19it/s] 69%|██████▉ | 33200/48008 [1:53:53<46:57, 5.26it/s] {'loss': 0.3025, 'grad_norm': 0.2734375, 'learning_rate': 6.16938843526079e-05, 'epoch': 0.69} 69%|██████▉ | 33200/48008 [1:53:53<46:57, 5.26it/s] 69%|██████▉ | 33201/48008 [1:53:53<47:35, 5.19it/s] 69%|██████▉ | 33202/48008 [1:53:54<47:27, 5.20it/s] 69%|██████▉ | 33203/48008 [1:53:54<47:25, 5.20it/s] 69%|██████▉ | 33204/48008 [1:53:54<47:56, 5.15it/s] 69%|██████▉ | 33205/48008 [1:53:54<47:39, 5.18it/s] 69%|██████▉ | 33206/48008 [1:53:54<47:32, 5.19it/s] 69%|██████▉ | 33207/48008 [1:53:55<48:15, 5.11it/s] 69%|██████▉ | 33208/48008 [1:53:55<47:51, 5.15it/s] 69%|██████▉ | 33209/48008 [1:53:55<49:02, 5.03it/s] 69%|██████▉ | 33210/48008 [1:53:55<51:59, 4.74it/s] 69%|██████▉ | 33211/48008 [1:53:55<50:26, 4.89it/s] 69%|██████▉ | 33212/48008 [1:53:56<49:19, 5.00it/s] 69%|██████▉ | 33213/48008 [1:53:56<48:40, 5.07it/s] 69%|██████▉ | 33214/48008 [1:53:56<48:16, 5.11it/s] 69%|██████▉ | 33215/48008 [1:53:56<47:43, 5.17it/s] 69%|██████▉ | 33216/48008 [1:53:56<47:23, 5.20it/s] 69%|██████▉ | 33217/48008 [1:53:57<47:20, 5.21it/s] 69%|██████▉ | 33218/48008 [1:53:57<47:06, 5.23it/s] 69%|██████▉ | 33219/48008 [1:53:57<47:08, 5.23it/s] 69%|██████▉ | 33220/48008 [1:53:57<48:10, 5.12it/s] 69%|██████▉ | 33221/48008 [1:53:57<49:17, 5.00it/s] 69%|██████▉ | 33222/48008 [1:53:58<48:20, 5.10it/s] 69%|██████▉ | 33223/48008 [1:53:58<56:26, 4.37it/s] 69%|██████▉ | 33224/48008 [1:53:58<53:58, 4.57it/s] 69%|██████▉ | 33225/48008 [1:53:58<55:21, 4.45it/s] 69%|██████▉ | 33226/48008 [1:53:58<52:36, 4.68it/s] 69%|██████▉ | 33227/48008 [1:53:59<51:15, 4.81it/s] 69%|██████▉ | 33228/48008 [1:53:59<50:44, 4.86it/s] 69%|██████▉ | 33229/48008 [1:53:59<49:28, 4.98it/s] 69%|██████▉ | 33230/48008 [1:53:59<49:38, 4.96it/s] 69%|██████▉ | 33231/48008 [1:53:59<49:13, 5.00it/s] 69%|██████▉ | 33232/48008 [1:54:00<48:56, 5.03it/s] 69%|██████▉ | 33233/48008 [1:54:00<48:18, 5.10it/s] 69%|██████▉ | 33234/48008 [1:54:00<48:02, 5.12it/s] 69%|██████▉ | 33235/48008 [1:54:00<48:47, 5.05it/s] 69%|██████▉ | 33236/48008 [1:54:00<48:12, 5.11it/s] 69%|██████▉ | 33237/48008 [1:54:01<47:35, 5.17it/s] 69%|██████▉ | 33238/48008 [1:54:01<47:29, 5.18it/s] 69%|██████▉ | 33239/48008 [1:54:01<47:20, 5.20it/s] 69%|██████▉ | 33240/48008 [1:54:01<47:01, 5.23it/s] 69%|██████▉ | 33241/48008 [1:54:01<47:24, 5.19it/s] 69%|██████▉ | 33242/48008 [1:54:02<47:17, 5.20it/s] 69%|██████▉ | 33243/48008 [1:54:02<47:33, 5.18it/s] 69%|██████▉ | 33244/48008 [1:54:02<47:01, 5.23it/s] 69%|██████▉ | 33245/48008 [1:54:02<47:46, 5.15it/s] 69%|██████▉ | 33246/48008 [1:54:02<47:49, 5.15it/s] 69%|██████▉ | 33247/48008 [1:54:03<47:31, 5.18it/s] 69%|██████▉ | 33248/48008 [1:54:03<47:17, 5.20it/s] 69%|██████▉ | 33249/48008 [1:54:03<47:05, 5.22it/s] 69%|██████▉ | 33250/48008 [1:54:03<46:56, 5.24it/s] 69%|██████▉ | 33250/48008 [1:54:03<46:56, 5.24it/s]{'loss': 0.2982, 'grad_norm': 0.154296875, 'learning_rate': 6.148558573571072e-05, 'epoch': 0.69} 69%|██████▉ | 33251/48008 [1:54:03<46:46, 5.26it/s] 69%|██████▉ | 33252/48008 [1:54:03<46:21, 5.31it/s] 69%|██████▉ | 33253/48008 [1:54:04<46:33, 5.28it/s] 69%|██████▉ | 33254/48008 [1:54:04<46:35, 5.28it/s] 69%|██████▉ | 33255/48008 [1:54:04<47:06, 5.22it/s] 69%|██████▉ | 33256/48008 [1:54:04<47:11, 5.21it/s] 69%|██████▉ | 33257/48008 [1:54:04<47:32, 5.17it/s] 69%|██████▉ | 33258/48008 [1:54:05<47:45, 5.15it/s] 69%|██████▉ | 33259/48008 [1:54:05<47:31, 5.17it/s] 69%|██████▉ | 33260/48008 [1:54:05<47:00, 5.23it/s] 69%|██████▉ | 33261/48008 [1:54:05<46:39, 5.27it/s] 69%|██████▉ | 33262/48008 [1:54:05<46:44, 5.26it/s] 69%|██████▉ | 33263/48008 [1:54:06<46:43, 5.26it/s] 69%|██████▉ | 33264/48008 [1:54:06<1:01:09, 4.02it/s] 69%|██████▉ | 33265/48008 [1:54:06<57:21, 4.28it/s] 69%|██████▉ | 33266/48008 [1:54:06<54:45, 4.49it/s] 69%|██████▉ | 33267/48008 [1:54:07<52:26, 4.68it/s] 69%|██████▉ | 33268/48008 [1:54:07<50:05, 4.90it/s] 69%|██████▉ | 33269/48008 [1:54:07<49:23, 4.97it/s] 69%|██████▉ | 33270/48008 [1:54:07<48:38, 5.05it/s] 69%|██████▉ | 33271/48008 [1:54:07<48:12, 5.09it/s] 69%|██████▉ | 33272/48008 [1:54:07<47:48, 5.14it/s] 69%|██████▉ | 33273/48008 [1:54:08<47:38, 5.16it/s] 69%|██████▉ | 33274/48008 [1:54:08<47:19, 5.19it/s] 69%|██████▉ | 33275/48008 [1:54:08<47:39, 5.15it/s] 69%|██████▉ | 33276/48008 [1:54:08<55:48, 4.40it/s] 69%|██████▉ | 33277/48008 [1:54:09<52:59, 4.63it/s] 69%|██████▉ | 33278/48008 [1:54:09<51:12, 4.79it/s] 69%|██████▉ | 33279/48008 [1:54:09<49:34, 4.95it/s] 69%|██████▉ | 33280/48008 [1:54:09<48:45, 5.04it/s] 69%|██████▉ | 33281/48008 [1:54:09<48:47, 5.03it/s] 69%|██████▉ | 33282/48008 [1:54:10<48:00, 5.11it/s] 69%|██████▉ | 33283/48008 [1:54:10<47:41, 5.15it/s] 69%|██████▉ | 33284/48008 [1:54:10<48:25, 5.07it/s] 69%|██████▉ | 33285/48008 [1:54:10<48:04, 5.10it/s] 69%|██████▉ | 33286/48008 [1:54:10<47:30, 5.16it/s] 69%|██████▉ | 33287/48008 [1:54:11<1:01:37, 3.98it/s] 69%|██████▉ | 33288/48008 [1:54:11<57:14, 4.29it/s] 69%|██████▉ | 33289/48008 [1:54:11<54:13, 4.52it/s] 69%|██████▉ | 33290/48008 [1:54:11<52:46, 4.65it/s] 69%|██████▉ | 33291/48008 [1:54:11<52:00, 4.72it/s] 69%|██████▉ | 33292/48008 [1:54:12<50:34, 4.85it/s] 69%|██████▉ | 33293/48008 [1:54:12<50:00, 4.90it/s] 69%|██████▉ | 33294/48008 [1:54:12<49:51, 4.92it/s] 69%|██████▉ | 33295/48008 [1:54:12<48:30, 5.06it/s] 69%|██████▉ | 33296/48008 [1:54:12<47:48, 5.13it/s] 69%|██████▉ | 33297/48008 [1:54:13<47:47, 5.13it/s] 69%|██████▉ | 33298/48008 [1:54:13<47:16, 5.19it/s] 69%|██████▉ | 33299/48008 [1:54:13<47:26, 5.17it/s] 69%|██████▉ | 33300/48008 [1:54:13<47:07, 5.20it/s] {'loss': 0.2994, 'grad_norm': 0.796875, 'learning_rate': 6.127728711881354e-05, 'epoch': 0.69} 69%|██████▉ | 33300/48008 [1:54:13<47:07, 5.20it/s] 69%|██████▉ | 33301/48008 [1:54:13<46:53, 5.23it/s] 69%|██████▉ | 33302/48008 [1:54:14<46:37, 5.26it/s] 69%|██████▉ | 33303/48008 [1:54:14<47:29, 5.16it/s] 69%|██████▉ | 33304/48008 [1:54:14<46:51, 5.23it/s] 69%|██████▉ | 33305/48008 [1:54:14<48:07, 5.09it/s] 69%|██████▉ | 33306/48008 [1:54:14<47:42, 5.14it/s] 69%|██████▉ | 33307/48008 [1:54:15<47:35, 5.15it/s] 69%|██████▉ | 33308/48008 [1:54:15<48:15, 5.08it/s] 69%|██████▉ | 33309/48008 [1:54:15<47:33, 5.15it/s] 69%|██████▉ | 33310/48008 [1:54:15<47:01, 5.21it/s] 69%|██████▉ | 33311/48008 [1:54:15<46:39, 5.25it/s] 69%|██████▉ | 33312/48008 [1:54:16<46:54, 5.22it/s] 69%|██████▉ | 33313/48008 [1:54:16<46:58, 5.21it/s] 69%|██████▉ | 33314/48008 [1:54:16<47:37, 5.14it/s] 69%|██████▉ | 33315/48008 [1:54:16<48:15, 5.07it/s] 69%|██████▉ | 33316/48008 [1:54:16<47:02, 5.21it/s] 69%|██████▉ | 33317/48008 [1:54:16<46:44, 5.24it/s] 69%|██████▉ | 33318/48008 [1:54:17<46:57, 5.21it/s] 69%|██████▉ | 33319/48008 [1:54:17<48:12, 5.08it/s] 69%|██████▉ | 33320/48008 [1:54:17<47:25, 5.16it/s] 69%|██████▉ | 33321/48008 [1:54:17<46:50, 5.23it/s] 69%|██████▉ | 33322/48008 [1:54:17<47:43, 5.13it/s] 69%|██████▉ | 33323/48008 [1:54:18<47:42, 5.13it/s] 69%|██████▉ | 33324/48008 [1:54:18<47:33, 5.15it/s] 69%|██████▉ | 33325/48008 [1:54:18<46:56, 5.21it/s] 69%|██████▉ | 33326/48008 [1:54:18<47:16, 5.18it/s] 69%|██████▉ | 33327/48008 [1:54:18<47:02, 5.20it/s] 69%|██████▉ | 33328/48008 [1:54:19<47:46, 5.12it/s] 69%|██████▉ | 33329/48008 [1:54:19<47:27, 5.15it/s] 69%|██████▉ | 33330/48008 [1:54:19<46:43, 5.24it/s] 69%|██████▉ | 33331/48008 [1:54:19<47:03, 5.20it/s] 69%|██████▉ | 33332/48008 [1:54:19<47:39, 5.13it/s] 69%|██████▉ | 33333/48008 [1:54:20<47:18, 5.17it/s] 69%|██████▉ | 33334/48008 [1:54:20<46:53, 5.22it/s] 69%|██████▉ | 33335/48008 [1:54:20<50:13, 4.87it/s] 69%|██████▉ | 33336/48008 [1:54:20<52:37, 4.65it/s] 69%|██████▉ | 33337/48008 [1:54:20<50:58, 4.80it/s] 69%|██████▉ | 33338/48008 [1:54:21<50:29, 4.84it/s] 69%|██████▉ | 33339/48008 [1:54:21<49:22, 4.95it/s] 69%|██████▉ | 33340/48008 [1:54:21<50:00, 4.89it/s] 69%|██████▉ | 33341/48008 [1:54:21<50:05, 4.88it/s] 69%|██████▉ | 33342/48008 [1:54:21<48:58, 4.99it/s] 69%|██████▉ | 33343/48008 [1:54:22<48:46, 5.01it/s] 69%|██████▉ | 33344/48008 [1:54:22<48:37, 5.03it/s] 69%|██████▉ | 33345/48008 [1:54:22<1:02:06, 3.93it/s] 69%|██████▉ | 33346/48008 [1:54:22<57:10, 4.27it/s] 69%|██████▉ | 33347/48008 [1:54:23<53:39, 4.55it/s] 69%|██████▉ | 33348/48008 [1:54:23<51:37, 4.73it/s] 69%|██████▉ | 33349/48008 [1:54:23<50:02, 4.88it/s] 69%|██████▉ | 33350/48008 [1:54:23<48:59, 4.99it/s] {'loss': 0.3002, 'grad_norm': 0.20703125, 'learning_rate': 6.106898850191634e-05, 'epoch': 0.69} 69%|██████▉ | 33350/48008 [1:54:23<48:59, 4.99it/s] 69%|██████▉ | 33351/48008 [1:54:23<48:48, 5.00it/s] 69%|██████▉ | 33352/48008 [1:54:24<48:16, 5.06it/s] 69%|██████▉ | 33353/48008 [1:54:24<47:45, 5.11it/s] 69%|██████▉ | 33354/48008 [1:54:24<46:58, 5.20it/s] 69%|██████▉ | 33355/48008 [1:54:24<55:27, 4.40it/s] 69%|██████▉ | 33356/48008 [1:54:24<53:19, 4.58it/s] 69%|██████▉ | 33357/48008 [1:54:25<51:17, 4.76it/s] 69%|██████▉ | 33358/48008 [1:54:25<51:18, 4.76it/s] 69%|██████▉ | 33359/48008 [1:54:25<50:03, 4.88it/s] 69%|██████▉ | 33360/48008 [1:54:25<49:17, 4.95it/s] 69%|██████▉ | 33361/48008 [1:54:25<49:18, 4.95it/s] 69%|██████▉ | 33362/48008 [1:54:26<48:00, 5.08it/s] 69%|██████▉ | 33363/48008 [1:54:26<48:04, 5.08it/s] 69%|██████▉ | 33364/48008 [1:54:26<48:26, 5.04it/s] 69%|██████▉ | 33365/48008 [1:54:26<47:43, 5.11it/s] 70%|██████▉ | 33366/48008 [1:54:26<48:23, 5.04it/s] 70%|██████▉ | 33367/48008 [1:54:27<48:39, 5.02it/s] 70%|██████▉ | 33368/48008 [1:54:27<47:54, 5.09it/s] 70%|██████▉ | 33369/48008 [1:54:27<46:54, 5.20it/s] 70%|██████▉ | 33370/48008 [1:54:27<46:43, 5.22it/s] 70%|██████▉ | 33371/48008 [1:54:27<46:30, 5.25it/s] 70%|██████▉ | 33372/48008 [1:54:28<46:59, 5.19it/s] 70%|██████▉ | 33373/48008 [1:54:28<50:16, 4.85it/s] 70%|██████▉ | 33374/48008 [1:54:28<49:42, 4.91it/s] 70%|██████▉ | 33375/48008 [1:54:28<50:03, 4.87it/s] 70%|██████▉ | 33376/48008 [1:54:28<48:41, 5.01it/s] 70%|██████▉ | 33377/48008 [1:54:29<47:44, 5.11it/s] 70%|██████▉ | 33378/48008 [1:54:29<47:23, 5.15it/s] 70%|██████▉ | 33379/48008 [1:54:29<47:36, 5.12it/s] 70%|██████▉ | 33380/48008 [1:54:29<47:15, 5.16it/s] 70%|██████▉ | 33381/48008 [1:54:29<47:46, 5.10it/s] 70%|██████▉ | 33382/48008 [1:54:30<47:03, 5.18it/s] 70%|██████▉ | 33383/48008 [1:54:30<47:17, 5.15it/s] 70%|██████▉ | 33384/48008 [1:54:30<47:07, 5.17it/s] 70%|██████▉ | 33385/48008 [1:54:30<46:47, 5.21it/s] 70%|██████▉ | 33386/48008 [1:54:30<47:12, 5.16it/s] 70%|██████▉ | 33387/48008 [1:54:31<46:47, 5.21it/s] 70%|██████▉ | 33388/48008 [1:54:31<55:06, 4.42it/s] 70%|██████▉ | 33389/48008 [1:54:31<52:12, 4.67it/s] 70%|██████▉ | 33390/48008 [1:54:31<50:48, 4.80it/s] 70%|██████▉ | 33391/48008 [1:54:31<49:26, 4.93it/s] 70%|██████▉ | 33392/48008 [1:54:32<48:25, 5.03it/s] 70%|██████▉ | 33393/48008 [1:54:32<47:42, 5.11it/s] 70%|██████▉ | 33394/48008 [1:54:32<46:56, 5.19it/s] 70%|██████▉ | 33395/48008 [1:54:32<46:46, 5.21it/s] 70%|██████▉ | 33396/48008 [1:54:32<47:00, 5.18it/s] 70%|██████▉ | 33397/48008 [1:54:33<47:36, 5.12it/s] 70%|██████▉ | 33398/48008 [1:54:33<47:43, 5.10it/s] 70%|██████▉ | 33399/48008 [1:54:33<47:12, 5.16it/s] 70%|██████▉ | 33400/48008 [1:54:34<1:57:44, 2.07it/s] {'loss': 0.2993, 'grad_norm': 0.2294921875, 'learning_rate': 6.0860689885019164e-05, 'epoch': 0.7} 70%|██████▉ | 33400/48008 [1:54:34<1:57:44, 2.07it/s] 70%|██████▉ | 33401/48008 [1:54:34<1:36:39, 2.52it/s] 70%|██████▉ | 33402/48008 [1:54:34<1:22:56, 2.93it/s] 70%|██████▉ | 33403/48008 [1:54:35<1:12:20, 3.36it/s] 70%|██████▉ | 33404/48008 [1:54:35<1:12:59, 3.33it/s] 70%|██████▉ | 33405/48008 [1:54:35<1:06:14, 3.67it/s] 70%|██████▉ | 33406/48008 [1:54:35<1:00:16, 4.04it/s] 70%|██████▉ | 33407/48008 [1:54:36<56:27, 4.31it/s] 70%|██████▉ | 33408/48008 [1:54:36<53:12, 4.57it/s] 70%|██████▉ | 33409/48008 [1:54:36<51:35, 4.72it/s] 70%|██████▉ | 33410/48008 [1:54:36<50:41, 4.80it/s] 70%|██████▉ | 33411/48008 [1:54:36<49:58, 4.87it/s] 70%|██████▉ | 33412/48008 [1:54:37<48:56, 4.97it/s] 70%|██████▉ | 33413/48008 [1:54:37<48:27, 5.02it/s] 70%|██████▉ | 33414/48008 [1:54:37<48:18, 5.04it/s] 70%|██████▉ | 33415/48008 [1:54:37<47:43, 5.10it/s] 70%|██████▉ | 33416/48008 [1:54:37<47:07, 5.16it/s] 70%|██████▉ | 33417/48008 [1:54:38<47:04, 5.17it/s] 70%|██████▉ | 33418/48008 [1:54:38<46:47, 5.20it/s] 70%|██████▉ | 33419/48008 [1:54:38<48:04, 5.06it/s] 70%|██████▉ | 33420/48008 [1:54:38<47:24, 5.13it/s] 70%|██████▉ | 33421/48008 [1:54:38<47:13, 5.15it/s] 70%|██████▉ | 33422/48008 [1:54:39<50:28, 4.82it/s] 70%|██████▉ | 33423/48008 [1:54:39<49:14, 4.94it/s] 70%|██████▉ | 33424/48008 [1:54:39<48:23, 5.02it/s] 70%|██████▉ | 33425/48008 [1:54:39<48:39, 4.99it/s] 70%|██████▉ | 33426/48008 [1:54:39<48:31, 5.01it/s] 70%|██████▉ | 33427/48008 [1:54:40<48:11, 5.04it/s] 70%|██████▉ | 33428/48008 [1:54:40<48:31, 5.01it/s] 70%|██████▉ | 33429/48008 [1:54:40<48:24, 5.02it/s] 70%|██████▉ | 33430/48008 [1:54:40<48:00, 5.06it/s] 70%|██████▉ | 33431/48008 [1:54:40<47:24, 5.12it/s] 70%|██████▉ | 33432/48008 [1:54:40<47:57, 5.07it/s] 70%|██████▉ | 33433/48008 [1:54:41<48:20, 5.02it/s] 70%|██████▉ | 33434/48008 [1:54:41<48:41, 4.99it/s] 70%|██████▉ | 33435/48008 [1:54:41<47:54, 5.07it/s] 70%|██████▉ | 33436/48008 [1:54:41<48:17, 5.03it/s] 70%|██████▉ | 33437/48008 [1:54:41<48:37, 5.00it/s] 70%|██████▉ | 33438/48008 [1:54:42<51:16, 4.74it/s] 70%|██████▉ | 33439/48008 [1:54:42<49:45, 4.88it/s] 70%|██████▉ | 33440/48008 [1:54:42<49:36, 4.90it/s] 70%|██████▉ | 33441/48008 [1:54:42<49:23, 4.92it/s] 70%|██████▉ | 33442/48008 [1:54:43<48:23, 5.02it/s] 70%|██████▉ | 33443/48008 [1:54:43<47:36, 5.10it/s] 70%|██████▉ | 33444/48008 [1:54:43<47:12, 5.14it/s] 70%|██████▉ | 33445/48008 [1:54:43<47:21, 5.12it/s] 70%|██████▉ | 33446/48008 [1:54:43<47:51, 5.07it/s] 70%|██████▉ | 33447/48008 [1:54:44<49:03, 4.95it/s] 70%|██████▉ | 33448/48008 [1:54:44<48:35, 4.99it/s] 70%|██████▉ | 33449/48008 [1:54:44<48:01, 5.05it/s] 70%|██████▉ | 33450/48008 [1:54:44<47:50, 5.07it/s] {'loss': 0.2984, 'grad_norm': 0.1953125, 'learning_rate': 6.0652391268121975e-05, 'epoch': 0.7} 70%|██████▉ | 33450/48008 [1:54:44<47:50, 5.07it/s] 70%|██████▉ | 33451/48008 [1:54:44<47:29, 5.11it/s] 70%|██████▉ | 33452/48008 [1:54:44<47:55, 5.06it/s] 70%|██████▉ | 33453/48008 [1:54:45<47:31, 5.10it/s] 70%|██████▉ | 33454/48008 [1:54:45<46:47, 5.18it/s] 70%|██████▉ | 33455/48008 [1:54:45<46:25, 5.22it/s] 70%|██████▉ | 33456/48008 [1:54:45<46:50, 5.18it/s] 70%|██████▉ | 33457/48008 [1:54:45<46:33, 5.21it/s] 70%|██████▉ | 33458/48008 [1:54:46<46:27, 5.22it/s] 70%|██████▉ | 33459/48008 [1:54:46<46:29, 5.22it/s] 70%|██████▉ | 33460/48008 [1:54:46<46:10, 5.25it/s] 70%|██████▉ | 33461/48008 [1:54:46<47:34, 5.10it/s] 70%|██████▉ | 33462/48008 [1:54:46<47:13, 5.13it/s] 70%|██████▉ | 33463/48008 [1:54:47<47:49, 5.07it/s] 70%|██████▉ | 33464/48008 [1:54:47<46:55, 5.17it/s] 70%|██████▉ | 33465/48008 [1:54:47<1:00:56, 3.98it/s] 70%|██████▉ | 33466/48008 [1:54:47<57:27, 4.22it/s] 70%|██████▉ | 33467/48008 [1:54:48<54:05, 4.48it/s] 70%|██████▉ | 33468/48008 [1:54:48<51:54, 4.67it/s] 70%|██████▉ | 33469/48008 [1:54:48<49:55, 4.85it/s] 70%|██████▉ | 33470/48008 [1:54:48<48:29, 5.00it/s] 70%|██████▉ | 33471/48008 [1:54:48<47:52, 5.06it/s] 70%|██████▉ | 33472/48008 [1:54:49<47:26, 5.11it/s] 70%|██████▉ | 33473/48008 [1:54:49<46:52, 5.17it/s] 70%|██████▉ | 33474/48008 [1:54:49<46:31, 5.21it/s] 70%|██████▉ | 33475/48008 [1:54:49<46:37, 5.19it/s] 70%|██████▉ | 33476/48008 [1:54:49<46:24, 5.22it/s] 70%|██████▉ | 33477/48008 [1:54:49<46:01, 5.26it/s] 70%|██████▉ | 33478/48008 [1:54:50<45:57, 5.27it/s] 70%|██████▉ | 33479/48008 [1:54:50<46:18, 5.23it/s] 70%|██████▉ | 33480/48008 [1:54:50<45:47, 5.29it/s] 70%|██████▉ | 33481/48008 [1:54:50<45:45, 5.29it/s] 70%|██████▉ | 33482/48008 [1:54:50<45:49, 5.28it/s] 70%|██████▉ | 33483/48008 [1:54:51<46:44, 5.18it/s] 70%|██████▉ | 33484/48008 [1:54:51<46:42, 5.18it/s] 70%|██████▉ | 33485/48008 [1:54:51<46:05, 5.25it/s] 70%|██████▉ | 33486/48008 [1:54:51<46:10, 5.24it/s] 70%|██████▉ | 33487/48008 [1:54:51<46:11, 5.24it/s] 70%|██████▉ | 33488/48008 [1:54:52<46:34, 5.20it/s] 70%|██████▉ | 33489/48008 [1:54:52<45:56, 5.27it/s] 70%|██████▉ | 33490/48008 [1:54:52<45:33, 5.31it/s] 70%|██████▉ | 33491/48008 [1:54:52<46:36, 5.19it/s] 70%|██████▉ | 33492/48008 [1:54:52<46:29, 5.20it/s] 70%|██████▉ | 33493/48008 [1:54:53<47:11, 5.13it/s] 70%|██████▉ | 33494/48008 [1:54:53<47:22, 5.11it/s] 70%|██████▉ | 33495/48008 [1:54:53<46:50, 5.16it/s] 70%|██████▉ | 33496/48008 [1:54:53<46:45, 5.17it/s] 70%|██████▉ | 33497/48008 [1:54:53<46:25, 5.21it/s] 70%|██████▉ | 33498/48008 [1:54:54<46:14, 5.23it/s] 70%|██████▉ | 33499/48008 [1:54:54<47:07, 5.13it/s] 70%|██████▉ | 33500/48008 [1:54:54<46:37, 5.19it/s] {'loss': 0.3031, 'grad_norm': 0.138671875, 'learning_rate': 6.0444092651224806e-05, 'epoch': 0.7} 70%|██████▉ | 33500/48008 [1:54:54<46:37, 5.19it/s] 70%|██████▉ | 33501/48008 [1:54:54<54:55, 4.40it/s] 70%|██████▉ | 33502/48008 [1:54:54<52:37, 4.59it/s] 70%|██████▉ | 33503/48008 [1:54:55<50:30, 4.79it/s] 70%|██████▉ | 33504/48008 [1:54:55<49:57, 4.84it/s] 70%|██████▉ | 33505/48008 [1:54:55<48:46, 4.96it/s] 70%|██████▉ | 33506/48008 [1:54:55<47:41, 5.07it/s] 70%|██████▉ | 33507/48008 [1:54:55<47:15, 5.11it/s] 70%|██████▉ | 33508/48008 [1:54:56<46:43, 5.17it/s] 70%|██████▉ | 33509/48008 [1:54:56<46:38, 5.18it/s] 70%|██████▉ | 33510/48008 [1:54:56<46:21, 5.21it/s] 70%|██████▉ | 33511/48008 [1:54:56<45:59, 5.25it/s] 70%|██████▉ | 33512/48008 [1:54:56<46:18, 5.22it/s] 70%|██████▉ | 33513/48008 [1:54:56<46:18, 5.22it/s] 70%|██████▉ | 33514/48008 [1:54:57<46:41, 5.17it/s] 70%|██████▉ | 33515/48008 [1:54:57<46:34, 5.19it/s] 70%|██████▉ | 33516/48008 [1:54:57<1:00:42, 3.98it/s] 70%|██████▉ | 33517/48008 [1:54:57<56:02, 4.31it/s] 70%|██████▉ | 33518/48008 [1:54:58<53:51, 4.48it/s] 70%|██████▉ | 33519/48008 [1:54:58<52:15, 4.62it/s] 70%|██████▉ | 33520/48008 [1:54:58<50:00, 4.83it/s] 70%|██████▉ | 33521/48008 [1:54:58<48:28, 4.98it/s] 70%|██████▉ | 33522/48008 [1:54:58<47:49, 5.05it/s] 70%|██████▉ | 33523/48008 [1:54:59<47:46, 5.05it/s] 70%|██████▉ | 33524/48008 [1:54:59<47:17, 5.10it/s] 70%|██████▉ | 33525/48008 [1:54:59<46:58, 5.14it/s] 70%|██████▉ | 33526/48008 [1:54:59<46:40, 5.17it/s] 70%|██████▉ | 33527/48008 [1:54:59<46:37, 5.18it/s] 70%|██████▉ | 33528/48008 [1:55:00<46:34, 5.18it/s] 70%|██████▉ | 33529/48008 [1:55:00<46:20, 5.21it/s] 70%|██████▉ | 33530/48008 [1:55:00<46:15, 5.22it/s] 70%|██████▉ | 33531/48008 [1:55:00<45:51, 5.26it/s] 70%|██████▉ | 33532/48008 [1:55:00<46:54, 5.14it/s] 70%|██████▉ | 33533/48008 [1:55:01<46:12, 5.22it/s] 70%|██████▉ | 33534/48008 [1:55:01<46:07, 5.23it/s] 70%|██████▉ | 33535/48008 [1:55:01<46:11, 5.22it/s] 70%|██████▉ | 33536/48008 [1:55:01<47:34, 5.07it/s] 70%|██████▉ | 33537/48008 [1:55:01<46:23, 5.20it/s] 70%|██████▉ | 33538/48008 [1:55:02<45:50, 5.26it/s] 70%|██████▉ | 33539/48008 [1:55:02<45:36, 5.29it/s] 70%|██████▉ | 33540/48008 [1:55:02<46:05, 5.23it/s] 70%|██████▉ | 33541/48008 [1:55:02<46:46, 5.16it/s] 70%|██████▉ | 33542/48008 [1:55:02<46:35, 5.18it/s] 70%|██████▉ | 33543/48008 [1:55:02<46:18, 5.21it/s] 70%|██████▉ | 33544/48008 [1:55:03<46:13, 5.21it/s] 70%|██████▉ | 33545/48008 [1:55:03<45:56, 5.25it/s] 70%|██████▉ | 33546/48008 [1:55:03<46:02, 5.23it/s] 70%|██████▉ | 33547/48008 [1:55:03<45:54, 5.25it/s] 70%|██████▉ | 33548/48008 [1:55:04<1:00:03, 4.01it/s] 70%|██████▉ | 33549/48008 [1:55:04<55:28, 4.34it/s] 70%|██████▉ | 33550/48008 [1:55:04<1:01:00, 3.95it/s] 70%|██████▉ | 33550/48008 [1:55:04<1:01:00, 3.95it/s]{'loss': 0.3022, 'grad_norm': 0.57421875, 'learning_rate': 6.0235794034327617e-05, 'epoch': 0.7} 70%|██████▉ | 33551/48008 [1:55:04<56:12, 4.29it/s] 70%|██████▉ | 33552/48008 [1:55:05<56:19, 4.28it/s] 70%|██████▉ | 33553/48008 [1:55:05<53:32, 4.50it/s] 70%|██████▉ | 33554/48008 [1:55:05<51:50, 4.65it/s] 70%|██████▉ | 33555/48008 [1:55:05<1:04:12, 3.75it/s] 70%|██████▉ | 33556/48008 [1:55:05<58:47, 4.10it/s] 70%|██████▉ | 33557/48008 [1:55:06<58:13, 4.14it/s] 70%|██████▉ | 33558/48008 [1:55:06<54:29, 4.42it/s] 70%|██████▉ | 33559/48008 [1:55:06<51:50, 4.64it/s] 70%|██████▉ | 33560/48008 [1:55:06<58:33, 4.11it/s] 70%|██████▉ | 33561/48008 [1:55:07<54:41, 4.40it/s] 70%|██████▉ | 33562/48008 [1:55:07<51:43, 4.66it/s] 70%|██████▉ | 33563/48008 [1:55:07<49:36, 4.85it/s] 70%|██████▉ | 33564/48008 [1:55:07<48:36, 4.95it/s] 70%|██████▉ | 33565/48008 [1:55:07<47:51, 5.03it/s] 70%|██████▉ | 33566/48008 [1:55:08<55:28, 4.34it/s] 70%|██████▉ | 33567/48008 [1:55:08<52:40, 4.57it/s] 70%|██████▉ | 33568/48008 [1:55:08<50:43, 4.74it/s] 70%|██████▉ | 33569/48008 [1:55:08<49:16, 4.88it/s] 70%|██████▉ | 33570/48008 [1:55:08<48:14, 4.99it/s] 70%|██████▉ | 33571/48008 [1:55:09<47:26, 5.07it/s] 70%|██████▉ | 33572/48008 [1:55:09<55:05, 4.37it/s] 70%|██████▉ | 33573/48008 [1:55:09<52:10, 4.61it/s] 70%|██████▉ | 33574/48008 [1:55:09<50:09, 4.80it/s] 70%|██████▉ | 33575/48008 [1:55:09<48:57, 4.91it/s] 70%|██████▉ | 33576/48008 [1:55:10<49:13, 4.89it/s] 70%|██████▉ | 33577/48008 [1:55:10<47:55, 5.02it/s] 70%|██████▉ | 33578/48008 [1:55:10<47:23, 5.07it/s] 70%|██████▉ | 33579/48008 [1:55:10<47:02, 5.11it/s] 70%|██████▉ | 33580/48008 [1:55:10<46:20, 5.19it/s] 70%|██████▉ | 33581/48008 [1:55:11<46:38, 5.15it/s] 70%|██████▉ | 33582/48008 [1:55:11<46:06, 5.21it/s] 70%|██████▉ | 33583/48008 [1:55:11<46:29, 5.17it/s] 70%|██████▉ | 33584/48008 [1:55:11<47:47, 5.03it/s] 70%|██████▉ | 33585/48008 [1:55:11<50:31, 4.76it/s] 70%|██████▉ | 33586/48008 [1:55:12<49:26, 4.86it/s] 70%|██████▉ | 33587/48008 [1:55:12<47:47, 5.03it/s] 70%|██████▉ | 33588/48008 [1:55:12<47:08, 5.10it/s] 70%|██████▉ | 33589/48008 [1:55:12<46:42, 5.14it/s] 70%|██████▉ | 33590/48008 [1:55:13<54:45, 4.39it/s] 70%|██████▉ | 33591/48008 [1:55:13<52:46, 4.55it/s] 70%|██████▉ | 33592/48008 [1:55:13<50:20, 4.77it/s] 70%|██████▉ | 33593/48008 [1:55:13<48:46, 4.93it/s] 70%|██████▉ | 33594/48008 [1:55:13<47:51, 5.02it/s] 70%|██████▉ | 33595/48008 [1:55:14<46:48, 5.13it/s] 70%|██████▉ | 33596/48008 [1:55:14<46:44, 5.14it/s] 70%|██████▉ | 33597/48008 [1:55:14<46:11, 5.20it/s] 70%|██████▉ | 33598/48008 [1:55:14<46:07, 5.21it/s] 70%|██████▉ | 33599/48008 [1:55:14<47:25, 5.06it/s] 70%|██████▉ | 33600/48008 [1:55:14<46:59, 5.11it/s] {'loss': 0.2994, 'grad_norm': 0.5, 'learning_rate': 6.0027495417430434e-05, 'epoch': 0.7} 70%|██████▉ | 33600/48008 [1:55:14<46:59, 5.11it/s] 70%|██████▉ | 33601/48008 [1:55:15<47:11, 5.09it/s] 70%|██████▉ | 33602/48008 [1:55:15<55:01, 4.36it/s] 70%|██████▉ | 33603/48008 [1:55:15<51:45, 4.64it/s] 70%|██████▉ | 33604/48008 [1:55:15<49:39, 4.83it/s] 70%|██████▉ | 33605/48008 [1:55:16<1:02:42, 3.83it/s] 70%|███████ | 33606/48008 [1:55:16<57:35, 4.17it/s] 70%|███████ | 33607/48008 [1:55:16<55:03, 4.36it/s] 70%|███████ | 33608/48008 [1:55:16<53:04, 4.52it/s] 70%|███████ | 33609/48008 [1:55:17<51:40, 4.64it/s] 70%|███████ | 33610/48008 [1:55:17<50:22, 4.76it/s] 70%|███████ | 33611/48008 [1:55:17<48:45, 4.92it/s] 70%|███████ | 33612/48008 [1:55:17<47:44, 5.02it/s] 70%|███████ | 33613/48008 [1:55:17<47:11, 5.08it/s] 70%|███████ | 33614/48008 [1:55:17<46:46, 5.13it/s] 70%|███████ | 33615/48008 [1:55:18<46:15, 5.19it/s] 70%|███████ | 33616/48008 [1:55:18<46:57, 5.11it/s] 70%|███████ | 33617/48008 [1:55:18<46:12, 5.19it/s] 70%|███████ | 33618/48008 [1:55:18<47:00, 5.10it/s] 70%|███████ | 33619/48008 [1:55:18<46:40, 5.14it/s] 70%|███████ | 33620/48008 [1:55:19<46:39, 5.14it/s] 70%|███████ | 33621/48008 [1:55:19<46:05, 5.20it/s] 70%|███████ | 33622/48008 [1:55:19<46:49, 5.12it/s] 70%|███████ | 33623/48008 [1:55:19<46:34, 5.15it/s] 70%|███████ | 33624/48008 [1:55:19<46:31, 5.15it/s] 70%|███████ | 33625/48008 [1:55:20<47:10, 5.08it/s] 70%|███████ | 33626/48008 [1:55:20<47:04, 5.09it/s] 70%|███████ | 33627/48008 [1:55:20<47:18, 5.07it/s] 70%|███████ | 33628/48008 [1:55:20<46:43, 5.13it/s] 70%|███████ | 33629/48008 [1:55:20<47:19, 5.06it/s] 70%|███████ | 33630/48008 [1:55:21<47:15, 5.07it/s] 70%|███████ | 33631/48008 [1:55:21<50:05, 4.78it/s] 70%|███████ | 33632/48008 [1:55:21<49:33, 4.83it/s] 70%|███████ | 33633/48008 [1:55:21<48:30, 4.94it/s] 70%|███████ | 33634/48008 [1:55:21<48:02, 4.99it/s] 70%|███████ | 33635/48008 [1:55:22<50:41, 4.73it/s] 70%|███████ | 33636/48008 [1:55:22<52:37, 4.55it/s] 70%|███████ | 33637/48008 [1:55:22<50:33, 4.74it/s] 70%|███████ | 33638/48008 [1:55:22<49:11, 4.87it/s] 70%|███████ | 33639/48008 [1:55:23<51:24, 4.66it/s] 70%|███████ | 33640/48008 [1:55:23<49:38, 4.82it/s] 70%|███████ | 33641/48008 [1:55:23<49:46, 4.81it/s] 70%|███████ | 33642/48008 [1:55:23<48:36, 4.93it/s] 70%|███████ | 33643/48008 [1:55:23<48:19, 4.95it/s] 70%|███████ | 33644/48008 [1:55:24<47:56, 4.99it/s] 70%|███████ | 33645/48008 [1:55:24<48:07, 4.97it/s] 70%|███████ | 33646/48008 [1:55:24<47:12, 5.07it/s] 70%|███████ | 33647/48008 [1:55:24<46:54, 5.10it/s] 70%|███████ | 33648/48008 [1:55:24<46:41, 5.13it/s] 70%|███████ | 33649/48008 [1:55:25<46:46, 5.12it/s] 70%|███████ | 33650/48008 [1:55:25<46:20, 5.16it/s] {'loss': 0.2986, 'grad_norm': 0.2060546875, 'learning_rate': 5.981919680053325e-05, 'epoch': 0.7} 70%|███████ | 33650/48008 [1:55:25<46:20, 5.16it/s] 70%|███████ | 33651/48008 [1:55:25<46:13, 5.18it/s] 70%|███████ | 33652/48008 [1:55:25<46:05, 5.19it/s] 70%|███████ | 33653/48008 [1:55:25<49:14, 4.86it/s] 70%|███████ | 33654/48008 [1:55:26<48:09, 4.97it/s] 70%|███████ | 33655/48008 [1:55:26<48:50, 4.90it/s] 70%|███████ | 33656/48008 [1:55:26<49:11, 4.86it/s] 70%|███████ | 33657/48008 [1:55:26<48:02, 4.98it/s] 70%|███████ | 33658/48008 [1:55:26<47:02, 5.08it/s] 70%|███████ | 33659/48008 [1:55:26<46:21, 5.16it/s] 70%|███████ | 33660/48008 [1:55:27<46:54, 5.10it/s] 70%|███████ | 33661/48008 [1:55:27<47:30, 5.03it/s] 70%|███████ | 33662/48008 [1:55:27<50:18, 4.75it/s] 70%|███████ | 33663/48008 [1:55:27<50:16, 4.76it/s] 70%|███████ | 33664/48008 [1:55:28<49:20, 4.85it/s] 70%|███████ | 33665/48008 [1:55:28<48:12, 4.96it/s] 70%|███████ | 33666/48008 [1:55:28<47:16, 5.06it/s] 70%|███████ | 33667/48008 [1:55:28<50:11, 4.76it/s] 70%|███████ | 33668/48008 [1:55:28<49:23, 4.84it/s] 70%|███████ | 33669/48008 [1:55:29<48:57, 4.88it/s] 70%|███████ | 33670/48008 [1:55:29<47:43, 5.01it/s] 70%|███████ | 33671/48008 [1:55:29<47:17, 5.05it/s] 70%|███████ | 33672/48008 [1:55:29<46:27, 5.14it/s] 70%|███████ | 33673/48008 [1:55:29<46:54, 5.09it/s] 70%|███████ | 33674/48008 [1:55:30<47:00, 5.08it/s] 70%|███████ | 33675/48008 [1:55:30<46:33, 5.13it/s] 70%|███████ | 33676/48008 [1:55:30<46:03, 5.19it/s] 70%|███████ | 33677/48008 [1:55:30<45:39, 5.23it/s] 70%|███████ | 33678/48008 [1:55:30<45:50, 5.21it/s] 70%|███████ | 33679/48008 [1:55:30<46:07, 5.18it/s] 70%|███████ | 33680/48008 [1:55:31<46:16, 5.16it/s] 70%|███████ | 33681/48008 [1:55:31<46:03, 5.18it/s] 70%|███████ | 33682/48008 [1:55:31<45:14, 5.28it/s] 70%|███████ | 33683/48008 [1:55:31<45:33, 5.24it/s] 70%|███████ | 33684/48008 [1:55:31<45:40, 5.23it/s] 70%|███████ | 33685/48008 [1:55:32<45:14, 5.28it/s] 70%|███████ | 33686/48008 [1:55:32<45:22, 5.26it/s] 70%|███████ | 33687/48008 [1:55:32<45:07, 5.29it/s] 70%|███████ | 33688/48008 [1:55:32<45:23, 5.26it/s] 70%|███████ | 33689/48008 [1:55:32<45:33, 5.24it/s] 70%|███████ | 33690/48008 [1:55:33<45:33, 5.24it/s] 70%|███████ | 33691/48008 [1:55:33<46:01, 5.18it/s] 70%|███████ | 33692/48008 [1:55:33<45:39, 5.23it/s] 70%|███████ | 33693/48008 [1:55:33<45:58, 5.19it/s] 70%|███████ | 33694/48008 [1:55:33<45:31, 5.24it/s] 70%|███████ | 33695/48008 [1:55:34<45:40, 5.22it/s] 70%|███████ | 33696/48008 [1:55:34<45:41, 5.22it/s] 70%|███████ | 33697/48008 [1:55:34<45:06, 5.29it/s] 70%|███████ | 33698/48008 [1:55:34<45:39, 5.22it/s] 70%|███████ | 33699/48008 [1:55:34<46:55, 5.08it/s] 70%|███████ | 33700/48008 [1:55:35<46:40, 5.11it/s] 70%|███████ | 33700/48008 [1:55:35<46:40, 5.11it/s]{'loss': 0.3007, 'grad_norm': 0.1455078125, 'learning_rate': 5.961089818363607e-05, 'epoch': 0.7} 70%|███████ | 33701/48008 [1:55:35<47:11, 5.05it/s] 70%|███████ | 33702/48008 [1:55:35<46:26, 5.13it/s] 70%|███████ | 33703/48008 [1:55:35<45:59, 5.18it/s] 70%|███████ | 33704/48008 [1:55:35<45:42, 5.22it/s] 70%|███████ | 33705/48008 [1:55:35<45:20, 5.26it/s] 70%|███████ | 33706/48008 [1:55:36<45:27, 5.24it/s] 70%|███████ | 33707/48008 [1:55:36<45:10, 5.28it/s] 70%|███████ | 33708/48008 [1:55:36<45:20, 5.26it/s] 70%|███████ | 33709/48008 [1:55:36<45:27, 5.24it/s] 70%|███████ | 33710/48008 [1:55:36<45:33, 5.23it/s] 70%|███████ | 33711/48008 [1:55:37<45:26, 5.24it/s] 70%|███████ | 33712/48008 [1:55:37<46:07, 5.17it/s] 70%|███████ | 33713/48008 [1:55:37<46:11, 5.16it/s] 70%|███████ | 33714/48008 [1:55:37<46:06, 5.17it/s] 70%|███████ | 33715/48008 [1:55:37<46:09, 5.16it/s] 70%|███████ | 33716/48008 [1:55:38<46:04, 5.17it/s] 70%|███████ | 33717/48008 [1:55:38<45:43, 5.21it/s] 70%|███████ | 33718/48008 [1:55:38<45:58, 5.18it/s] 70%|███████ | 33719/48008 [1:55:38<46:51, 5.08it/s] 70%|███████ | 33720/48008 [1:55:38<46:36, 5.11it/s] 70%|███████ | 33721/48008 [1:55:39<47:26, 5.02it/s] 70%|███████ | 33722/48008 [1:55:39<47:22, 5.03it/s] 70%|███████ | 33723/48008 [1:55:39<47:36, 5.00it/s] 70%|███████ | 33724/48008 [1:55:39<47:11, 5.04it/s] 70%|███████ | 33725/48008 [1:55:39<46:49, 5.08it/s] 70%|███████ | 33726/48008 [1:55:40<46:15, 5.15it/s] 70%|███████ | 33727/48008 [1:55:40<46:22, 5.13it/s] 70%|███████ | 33728/48008 [1:55:40<46:27, 5.12it/s] 70%|███████ | 33729/48008 [1:55:40<46:00, 5.17it/s] 70%|███████ | 33730/48008 [1:55:40<46:08, 5.16it/s] 70%|███████ | 33731/48008 [1:55:41<45:58, 5.18it/s] 70%|███████ | 33732/48008 [1:55:41<45:52, 5.19it/s] 70%|███████ | 33733/48008 [1:55:41<45:48, 5.19it/s] 70%|███████ | 33734/48008 [1:55:41<45:26, 5.23it/s] 70%|███████ | 33735/48008 [1:55:41<45:06, 5.27it/s] 70%|███████ | 33736/48008 [1:55:41<45:13, 5.26it/s] 70%|███████ | 33737/48008 [1:55:42<45:46, 5.20it/s] 70%|███████ | 33738/48008 [1:55:42<45:41, 5.21it/s] 70%|███████ | 33739/48008 [1:55:42<45:30, 5.23it/s] 70%|███████ | 33740/48008 [1:55:42<46:15, 5.14it/s] 70%|███████ | 33741/48008 [1:55:42<46:03, 5.16it/s] 70%|███████ | 33742/48008 [1:55:43<45:58, 5.17it/s] 70%|███████ | 33743/48008 [1:55:43<45:52, 5.18it/s] 70%|███████ | 33744/48008 [1:55:43<45:36, 5.21it/s] 70%|███████ | 33745/48008 [1:55:43<45:23, 5.24it/s] 70%|███████ | 33746/48008 [1:55:43<46:20, 5.13it/s] 70%|███████ | 33747/48008 [1:55:44<46:04, 5.16it/s] 70%|███████ | 33748/48008 [1:55:44<46:03, 5.16it/s] 70%|███████ | 33749/48008 [1:55:44<49:10, 4.83it/s] 70%|███████ | 33750/48008 [1:55:44<48:19, 4.92it/s] {'loss': 0.301, 'grad_norm': 0.2216796875, 'learning_rate': 5.940259956673888e-05, 'epoch': 0.7} 70%|███████ | 33750/48008 [1:55:44<48:19, 4.92it/s] 70%|███████ | 33751/48008 [1:55:44<47:36, 4.99it/s] 70%|███████ | 33752/48008 [1:55:45<46:42, 5.09it/s] 70%|███████ | 33753/48008 [1:55:45<46:03, 5.16it/s] 70%|███████ | 33754/48008 [1:55:45<46:20, 5.13it/s] 70%|███████ | 33755/48008 [1:55:45<47:25, 5.01it/s] 70%|███████ | 33756/48008 [1:55:45<47:15, 5.03it/s] 70%|███████ | 33757/48008 [1:55:46<47:37, 4.99it/s] 70%|███████ | 33758/48008 [1:55:46<47:00, 5.05it/s] 70%|███████ | 33759/48008 [1:55:46<46:27, 5.11it/s] 70%|███████ | 33760/48008 [1:55:46<45:54, 5.17it/s] 70%|███████ | 33761/48008 [1:55:46<49:07, 4.83it/s] 70%|███████ | 33762/48008 [1:55:47<48:28, 4.90it/s] 70%|███████ | 33763/48008 [1:55:47<55:55, 4.25it/s] 70%|███████ | 33764/48008 [1:55:47<52:52, 4.49it/s] 70%|███████ | 33765/48008 [1:55:47<51:31, 4.61it/s] 70%|███████ | 33766/48008 [1:55:48<49:23, 4.81it/s] 70%|███████ | 33767/48008 [1:55:48<49:02, 4.84it/s] 70%|███████ | 33768/48008 [1:55:48<47:55, 4.95it/s] 70%|███████ | 33769/48008 [1:55:48<47:59, 4.94it/s] 70%|███████ | 33770/48008 [1:55:48<46:51, 5.06it/s] 70%|███████ | 33771/48008 [1:55:48<47:10, 5.03it/s] 70%|███████ | 33772/48008 [1:55:49<46:29, 5.10it/s] 70%|███████ | 33773/48008 [1:55:49<45:44, 5.19it/s] 70%|███████ | 33774/48008 [1:55:49<46:54, 5.06it/s] 70%|███████ | 33775/48008 [1:55:49<46:40, 5.08it/s] 70%|███████ | 33776/48008 [1:55:49<47:13, 5.02it/s] 70%|███████ | 33777/48008 [1:55:50<54:51, 4.32it/s] 70%|███████ | 33778/48008 [1:55:50<52:28, 4.52it/s] 70%|███████ | 33779/48008 [1:55:50<50:38, 4.68it/s] 70%|███████ | 33780/48008 [1:55:50<49:48, 4.76it/s] 70%|███████ | 33781/48008 [1:55:51<48:34, 4.88it/s] 70%|███████ | 33782/48008 [1:55:51<47:24, 5.00it/s] 70%|███████ | 33783/48008 [1:55:51<46:28, 5.10it/s] 70%|███████ | 33784/48008 [1:55:51<46:52, 5.06it/s] 70%|███████ | 33785/48008 [1:55:51<46:16, 5.12it/s] 70%|███████ | 33786/48008 [1:55:52<46:43, 5.07it/s] 70%|███████ | 33787/48008 [1:55:52<46:42, 5.07it/s] 70%|███████ | 33788/48008 [1:55:52<46:13, 5.13it/s] 70%|███████ | 33789/48008 [1:55:52<46:42, 5.07it/s] 70%|███████ | 33790/48008 [1:55:52<46:13, 5.13it/s] 70%|███████ | 33791/48008 [1:55:53<45:50, 5.17it/s] 70%|███████ | 33792/48008 [1:55:53<45:41, 5.19it/s] 70%|███████ | 33793/48008 [1:55:53<45:30, 5.21it/s] 70%|███████ | 33794/48008 [1:55:53<45:43, 5.18it/s] 70%|███████ | 33795/48008 [1:55:53<45:34, 5.20it/s] 70%|███████ | 33796/48008 [1:55:53<45:29, 5.21it/s] 70%|███████ | 33797/48008 [1:55:54<45:09, 5.25it/s] 70%|███████ | 33798/48008 [1:55:54<45:18, 5.23it/s] 70%|███████ | 33799/48008 [1:55:54<45:37, 5.19it/s] 70%|███████ | 33800/48008 [1:55:54<46:10, 5.13it/s] 70%|███████ | 33800/48008 [1:55:54<46:10, 5.13it/s]{'loss': 0.3002, 'grad_norm': 0.140625, 'learning_rate': 5.91943009498417e-05, 'epoch': 0.7} 70%|███████ | 33801/48008 [1:55:54<46:02, 5.14it/s] 70%|███████ | 33802/48008 [1:55:55<45:53, 5.16it/s] 70%|███████ | 33803/48008 [1:55:55<45:34, 5.19it/s] 70%|███████ | 33804/48008 [1:55:55<45:29, 5.20it/s] 70%|███████ | 33805/48008 [1:55:55<45:15, 5.23it/s] 70%|███████ | 33806/48008 [1:55:55<46:06, 5.13it/s] 70%|███████ | 33807/48008 [1:55:56<45:28, 5.20it/s] 70%|███████ | 33808/48008 [1:55:56<45:04, 5.25it/s] 70%|███████ | 33809/48008 [1:55:56<46:01, 5.14it/s] 70%|███████ | 33810/48008 [1:55:56<45:29, 5.20it/s] 70%|███████ | 33811/48008 [1:55:56<45:49, 5.16it/s] 70%|███████ | 33812/48008 [1:55:57<45:23, 5.21it/s] 70%|███████ | 33813/48008 [1:55:57<45:27, 5.20it/s] 70%|███████ | 33814/48008 [1:55:57<46:06, 5.13it/s] 70%|███████ | 33815/48008 [1:55:57<45:20, 5.22it/s] 70%|███████ | 33816/48008 [1:55:57<46:29, 5.09it/s] 70%|███████ | 33817/48008 [1:55:58<47:30, 4.98it/s] 70%|███████ | 33818/48008 [1:55:58<48:11, 4.91it/s] 70%|███████ | 33819/48008 [1:55:58<47:13, 5.01it/s] 70%|███████ | 33820/48008 [1:55:58<46:44, 5.06it/s] 70%|███████ | 33821/48008 [1:55:58<45:53, 5.15it/s] 70%|███████ | 33822/48008 [1:55:59<46:10, 5.12it/s] 70%|███████ | 33823/48008 [1:55:59<45:41, 5.17it/s] 70%|███████ | 33824/48008 [1:55:59<45:00, 5.25it/s] 70%|███████ | 33825/48008 [1:55:59<44:55, 5.26it/s] 70%|███████ | 33826/48008 [1:55:59<48:19, 4.89it/s] 70%|███████ | 33827/48008 [1:56:00<1:01:10, 3.86it/s] 70%|███████ | 33828/48008 [1:56:00<56:53, 4.15it/s] 70%|███████ | 33829/48008 [1:56:00<54:09, 4.36it/s] 70%|███████ | 33830/48008 [1:56:00<51:19, 4.60it/s] 70%|███████ | 33831/48008 [1:56:01<52:37, 4.49it/s] 70%|███████ | 33832/48008 [1:56:01<51:14, 4.61it/s] 70%|███████ | 33833/48008 [1:56:01<49:16, 4.79it/s] 70%|███████ | 33834/48008 [1:56:01<48:23, 4.88it/s] 70%|███████ | 33835/48008 [1:56:01<47:02, 5.02it/s] 70%|███████ | 33836/48008 [1:56:02<54:32, 4.33it/s] 70%|███████ | 33837/48008 [1:56:02<50:57, 4.64it/s] 70%|███████ | 33838/48008 [1:56:02<49:19, 4.79it/s] 70%|███████ | 33839/48008 [1:56:02<51:19, 4.60it/s] 70%|███████ | 33840/48008 [1:56:02<49:04, 4.81it/s] 70%|███████ | 33841/48008 [1:56:03<47:42, 4.95it/s] 70%|███████ | 33842/48008 [1:56:03<46:47, 5.05it/s] 70%|███████ | 33843/48008 [1:56:03<47:32, 4.97it/s] 70%|███████ | 33844/48008 [1:56:03<47:33, 4.96it/s] 70%|███████ | 33845/48008 [1:56:03<47:20, 4.99it/s] 71%|███████ | 33846/48008 [1:56:04<46:21, 5.09it/s] 71%|███████ | 33847/48008 [1:56:04<45:14, 5.22it/s] 71%|███████ | 33848/48008 [1:56:04<44:44, 5.28it/s] 71%|███████ | 33849/48008 [1:56:04<45:42, 5.16it/s] 71%|███████ | 33850/48008 [1:56:04<46:41, 5.05it/s] {'loss': 0.2997, 'grad_norm': 0.3359375, 'learning_rate': 5.8986002332944514e-05, 'epoch': 0.71} 71%|███████ | 33850/48008 [1:56:04<46:41, 5.05it/s] 71%|███████ | 33851/48008 [1:56:05<54:23, 4.34it/s] 71%|███████ | 33852/48008 [1:56:05<51:42, 4.56it/s] 71%|███████ | 33853/48008 [1:56:05<49:21, 4.78it/s] 71%|███████ | 33854/48008 [1:56:05<47:46, 4.94it/s] 71%|███████ | 33855/48008 [1:56:05<47:02, 5.01it/s] 71%|███████ | 33856/48008 [1:56:06<46:25, 5.08it/s] 71%|███████ | 33857/48008 [1:56:06<45:59, 5.13it/s] 71%|███████ | 33858/48008 [1:56:06<45:41, 5.16it/s] 71%|███████ | 33859/48008 [1:56:06<46:00, 5.13it/s] 71%|███████ | 33860/48008 [1:56:06<46:33, 5.06it/s] 71%|███████ | 33861/48008 [1:56:07<47:28, 4.97it/s] 71%|███████ | 33862/48008 [1:56:07<46:26, 5.08it/s] 71%|███████ | 33863/48008 [1:56:07<45:58, 5.13it/s] 71%|███████ | 33864/48008 [1:56:07<53:48, 4.38it/s] 71%|███████ | 33865/48008 [1:56:07<50:51, 4.63it/s] 71%|███████ | 33866/48008 [1:56:08<52:22, 4.50it/s] 71%|███████ | 33867/48008 [1:56:08<50:01, 4.71it/s] 71%|███████ | 33868/48008 [1:56:08<48:22, 4.87it/s] 71%|███████ | 33869/48008 [1:56:08<48:08, 4.90it/s] 71%|███████ | 33870/48008 [1:56:08<47:07, 5.00it/s] 71%|███████ | 33871/48008 [1:56:09<46:28, 5.07it/s] 71%|███████ | 33872/48008 [1:56:09<45:43, 5.15it/s] 71%|███████ | 33873/48008 [1:56:09<46:06, 5.11it/s] 71%|███████ | 33874/48008 [1:56:09<46:09, 5.10it/s] 71%|███████ | 33875/48008 [1:56:09<46:00, 5.12it/s] 71%|███████ | 33876/48008 [1:56:10<47:10, 4.99it/s] 71%|███████ | 33877/48008 [1:56:10<46:40, 5.05it/s] 71%|███████ | 33878/48008 [1:56:10<46:27, 5.07it/s] 71%|███████ | 33879/48008 [1:56:10<46:24, 5.07it/s] 71%|███████ | 33880/48008 [1:56:10<45:59, 5.12it/s] 71%|███████ | 33881/48008 [1:56:11<46:13, 5.09it/s] 71%|███████ | 33882/48008 [1:56:11<59:45, 3.94it/s] 71%|███████ | 33883/48008 [1:56:11<56:13, 4.19it/s] 71%|███████ | 33884/48008 [1:56:11<54:02, 4.36it/s] 71%|███████ | 33885/48008 [1:56:12<51:35, 4.56it/s] 71%|███████ | 33886/48008 [1:56:12<49:14, 4.78it/s] 71%|███████ | 33887/48008 [1:56:12<48:29, 4.85it/s] 71%|███████ | 33888/48008 [1:56:12<48:13, 4.88it/s] 71%|███████ | 33889/48008 [1:56:12<47:50, 4.92it/s] 71%|███████ | 33890/48008 [1:56:13<47:19, 4.97it/s] 71%|███████ | 33891/48008 [1:56:13<46:23, 5.07it/s] 71%|███████ | 33892/48008 [1:56:13<46:01, 5.11it/s] 71%|███████ | 33893/48008 [1:56:13<45:41, 5.15it/s] 71%|███████ | 33894/48008 [1:56:13<45:46, 5.14it/s] 71%|███████ | 33895/48008 [1:56:14<46:15, 5.09it/s] 71%|███████ | 33896/48008 [1:56:14<46:40, 5.04it/s] 71%|███████ | 33897/48008 [1:56:14<47:03, 5.00it/s] 71%|███████ | 33898/48008 [1:56:14<46:23, 5.07it/s] 71%|███████ | 33899/48008 [1:56:14<54:03, 4.35it/s] 71%|███████ | 33900/48008 [1:56:15<51:14, 4.59it/s] {'loss': 0.2988, 'grad_norm': 0.3828125, 'learning_rate': 5.8777703716047325e-05, 'epoch': 0.71} 71%|███████ | 33900/48008 [1:56:15<51:14, 4.59it/s] 71%|███████ | 33901/48008 [1:56:15<50:19, 4.67it/s] 71%|███████ | 33902/48008 [1:56:15<49:32, 4.75it/s] 71%|███████ | 33903/48008 [1:56:15<48:06, 4.89it/s] 71%|███████ | 33904/48008 [1:56:15<46:45, 5.03it/s] 71%|███████ | 33905/48008 [1:56:16<59:54, 3.92it/s] 71%|███████ | 33906/48008 [1:56:16<55:38, 4.22it/s] 71%|███████ | 33907/48008 [1:56:16<52:30, 4.48it/s] 71%|███████ | 33908/48008 [1:56:16<50:16, 4.67it/s] 71%|███████ | 33909/48008 [1:56:17<1:02:29, 3.76it/s] 71%|███████ | 33910/48008 [1:56:17<57:05, 4.12it/s] 71%|███████ | 33911/48008 [1:56:17<53:43, 4.37it/s] 71%|███████ | 33912/48008 [1:56:17<50:48, 4.62it/s] 71%|███████ | 33913/48008 [1:56:18<48:38, 4.83it/s] 71%|███████ | 33914/48008 [1:56:18<47:34, 4.94it/s] 71%|███████ | 33915/48008 [1:56:18<47:03, 4.99it/s] 71%|███████ | 33916/48008 [1:56:18<46:18, 5.07it/s] 71%|███████ | 33917/48008 [1:56:18<46:41, 5.03it/s] 71%|███████ | 33918/48008 [1:56:19<46:08, 5.09it/s] 71%|███████ | 33919/48008 [1:56:19<45:35, 5.15it/s] 71%|███████ | 33920/48008 [1:56:19<46:11, 5.08it/s] 71%|███████ | 33921/48008 [1:56:19<46:18, 5.07it/s] 71%|███████ | 33922/48008 [1:56:19<46:07, 5.09it/s] 71%|███████ | 33923/48008 [1:56:20<45:43, 5.13it/s] 71%|███████ | 33924/48008 [1:56:20<45:26, 5.17it/s] 71%|███████ | 33925/48008 [1:56:20<45:30, 5.16it/s] 71%|███████ | 33926/48008 [1:56:20<45:24, 5.17it/s] 71%|███████ | 33927/48008 [1:56:20<45:32, 5.15it/s] 71%|███████ | 33928/48008 [1:56:20<46:03, 5.10it/s] 71%|███████ | 33929/48008 [1:56:21<1:34:53, 2.47it/s] 71%|███████ | 33930/48008 [1:56:22<1:19:43, 2.94it/s] 71%|███████ | 33931/48008 [1:56:22<1:09:14, 3.39it/s] 71%|███████ | 33932/48008 [1:56:22<1:01:54, 3.79it/s] 71%|███████ | 33933/48008 [1:56:22<1:04:48, 3.62it/s] 71%|███████ | 33934/48008 [1:56:22<58:30, 4.01it/s] 71%|███████ | 33935/48008 [1:56:23<1:49:54, 2.13it/s] 71%|███████ | 33936/48008 [1:56:24<1:31:15, 2.57it/s] 71%|███████ | 33937/48008 [1:56:24<1:17:14, 3.04it/s] 71%|███████ | 33938/48008 [1:56:24<1:08:14, 3.44it/s] 71%|███████ | 33939/48008 [1:56:24<1:01:17, 3.83it/s] 71%|███████ | 33940/48008 [1:56:24<55:45, 4.20it/s] 71%|███████ | 33941/48008 [1:56:25<52:41, 4.45it/s] 71%|███████ | 33942/48008 [1:56:25<50:10, 4.67it/s] 71%|███████ | 33943/48008 [1:56:25<48:19, 4.85it/s] 71%|███████ | 33944/48008 [1:56:25<47:41, 4.91it/s] 71%|███████ | 33945/48008 [1:56:25<47:37, 4.92it/s] 71%|███████ | 33946/48008 [1:56:26<46:47, 5.01it/s] 71%|███████ | 33947/48008 [1:56:26<1:29:59, 2.60it/s] 71%|███████ | 33948/48008 [1:56:27<1:16:03, 3.08it/s] 71%|███████ | 33949/48008 [1:56:27<1:07:33, 3.47it/s] 71%|███████ | 33950/48008 [1:56:27<1:00:32, 3.87it/s] {'loss': 0.3017, 'grad_norm': 0.1455078125, 'learning_rate': 5.856940509915014e-05, 'epoch': 0.71} 71%|███████ | 33950/48008 [1:56:27<1:00:32, 3.87it/s] 71%|███████ | 33951/48008 [1:56:27<55:55, 4.19it/s] 71%|███████ | 33952/48008 [1:56:27<53:58, 4.34it/s] 71%|███████ | 33953/48008 [1:56:28<51:02, 4.59it/s] 71%|███████ | 33954/48008 [1:56:28<49:03, 4.77it/s] 71%|███████ | 33955/48008 [1:56:28<47:35, 4.92it/s] 71%|███████ | 33956/48008 [1:56:28<46:31, 5.03it/s] 71%|███████ | 33957/48008 [1:56:28<46:27, 5.04it/s] 71%|███████ | 33958/48008 [1:56:28<45:48, 5.11it/s] 71%|███████ | 33959/48008 [1:56:29<45:30, 5.15it/s] 71%|███████ | 33960/48008 [1:56:29<45:17, 5.17it/s] 71%|███████ | 33961/48008 [1:56:29<44:53, 5.21it/s] 71%|███████ | 33962/48008 [1:56:29<44:54, 5.21it/s] 71%|███████ | 33963/48008 [1:56:29<44:24, 5.27it/s] 71%|███████ | 33964/48008 [1:56:30<44:29, 5.26it/s] 71%|███████ | 33965/48008 [1:56:30<44:34, 5.25it/s] 71%|███████ | 33966/48008 [1:56:30<44:42, 5.23it/s] 71%|███████ | 33967/48008 [1:56:30<45:33, 5.14it/s] 71%|███████ | 33968/48008 [1:56:30<45:00, 5.20it/s] 71%|███████ | 33969/48008 [1:56:31<44:45, 5.23it/s] 71%|███████ | 33970/48008 [1:56:31<44:22, 5.27it/s] 71%|███████ | 33971/48008 [1:56:31<44:39, 5.24it/s] 71%|███████ | 33972/48008 [1:56:31<44:32, 5.25it/s] 71%|███████ | 33973/48008 [1:56:31<44:57, 5.20it/s] 71%|███████ | 33974/48008 [1:56:32<44:06, 5.30it/s] 71%|███████ | 33975/48008 [1:56:32<44:47, 5.22it/s] 71%|███████ | 33976/48008 [1:56:32<45:08, 5.18it/s] 71%|███████ | 33977/48008 [1:56:32<44:57, 5.20it/s] 71%|███████ | 33978/48008 [1:56:32<44:35, 5.24it/s] 71%|███████ | 33979/48008 [1:56:32<44:18, 5.28it/s] 71%|███████ | 33980/48008 [1:56:33<44:35, 5.24it/s] 71%|███████ | 33981/48008 [1:56:33<45:14, 5.17it/s] 71%|███████ | 33982/48008 [1:56:33<44:36, 5.24it/s] 71%|███████ | 33983/48008 [1:56:34<1:28:53, 2.63it/s] 71%|███████ | 33984/48008 [1:56:34<1:15:00, 3.12it/s] 71%|███████ | 33985/48008 [1:56:34<1:05:44, 3.56it/s] 71%|███████ | 33986/48008 [1:56:34<59:31, 3.93it/s] 71%|███████ | 33987/48008 [1:56:35<55:23, 4.22it/s] 71%|███████ | 33988/48008 [1:56:35<52:00, 4.49it/s] 71%|███████ | 33989/48008 [1:56:35<49:49, 4.69it/s] 71%|███████ | 33990/48008 [1:56:35<48:07, 4.86it/s] 71%|███████ | 33991/48008 [1:56:35<48:02, 4.86it/s] 71%|███████ | 33992/48008 [1:56:36<46:44, 5.00it/s] 71%|███████ | 33993/48008 [1:56:36<46:39, 5.01it/s] 71%|███████ | 33994/48008 [1:56:36<59:43, 3.91it/s] 71%|███████ | 33995/48008 [1:56:36<55:17, 4.22it/s] 71%|███████ | 33996/48008 [1:56:37<52:27, 4.45it/s] 71%|███████ | 33997/48008 [1:56:37<50:12, 4.65it/s] 71%|███████ | 33998/48008 [1:56:37<48:39, 4.80it/s] 71%|███████ | 33999/48008 [1:56:37<50:43, 4.60it/s] 71%|███████ | 34000/48008 [1:56:37<48:57, 4.77it/s] {'loss': 0.3028, 'grad_norm': 0.25390625, 'learning_rate': 5.836110648225296e-05, 'epoch': 0.71} 71%|███████ | 34000/48008 [1:56:37<48:57, 4.77it/s] 71%|███████ | 34001/48008 [1:56:38<47:52, 4.88it/s] 71%|███████ | 34002/48008 [1:56:38<47:18, 4.94it/s] 71%|███████ | 34003/48008 [1:56:38<47:44, 4.89it/s] 71%|███████ | 34004/48008 [1:56:38<47:07, 4.95it/s] 71%|███████ | 34005/48008 [1:56:38<46:32, 5.01it/s] 71%|███████ | 34006/48008 [1:56:39<46:24, 5.03it/s] 71%|███████ | 34007/48008 [1:56:39<45:23, 5.14it/s] 71%|███████ | 34008/48008 [1:56:39<46:17, 5.04it/s] 71%|███████ | 34009/48008 [1:56:39<45:21, 5.14it/s] 71%|███████ | 34010/48008 [1:56:39<45:12, 5.16it/s] 71%|███████ | 34011/48008 [1:56:40<45:05, 5.17it/s] 71%|███████ | 34012/48008 [1:56:40<44:36, 5.23it/s] 71%|███████ | 34013/48008 [1:56:40<44:15, 5.27it/s] 71%|███████ | 34014/48008 [1:56:40<44:30, 5.24it/s] 71%|███████ | 34015/48008 [1:56:40<44:36, 5.23it/s] 71%|███████ | 34016/48008 [1:56:41<44:28, 5.24it/s] 71%|███████ | 34017/48008 [1:56:41<44:32, 5.24it/s] 71%|███████ | 34018/48008 [1:56:41<44:22, 5.25it/s] 71%|███████ | 34019/48008 [1:56:41<44:25, 5.25it/s] 71%|███████ | 34020/48008 [1:56:41<44:33, 5.23it/s] 71%|███████ | 34021/48008 [1:56:41<44:10, 5.28it/s] 71%|███████ | 34022/48008 [1:56:42<44:18, 5.26it/s] 71%|███████ | 34023/48008 [1:56:42<44:04, 5.29it/s] 71%|███████ | 34024/48008 [1:56:42<44:18, 5.26it/s] 71%|███████ | 34025/48008 [1:56:42<44:16, 5.26it/s] 71%|███████ | 34026/48008 [1:56:42<44:28, 5.24it/s] 71%|███████ | 34027/48008 [1:56:43<44:17, 5.26it/s] 71%|███████ | 34028/48008 [1:56:43<44:25, 5.24it/s] 71%|███████ | 34029/48008 [1:56:43<45:30, 5.12it/s] 71%|███████ | 34030/48008 [1:56:43<45:04, 5.17it/s] 71%|███████ | 34031/48008 [1:56:43<45:41, 5.10it/s] 71%|███████ | 34032/48008 [1:56:44<45:13, 5.15it/s] 71%|███████ | 34033/48008 [1:56:44<48:09, 4.84it/s] 71%|███████ | 34034/48008 [1:56:44<47:45, 4.88it/s] 71%|███████ | 34035/48008 [1:56:44<46:33, 5.00it/s] 71%|███████ | 34036/48008 [1:56:44<45:45, 5.09it/s] 71%|███████ | 34037/48008 [1:56:45<45:19, 5.14it/s] 71%|███████ | 34038/48008 [1:56:45<45:21, 5.13it/s] 71%|███████ | 34039/48008 [1:56:45<45:32, 5.11it/s] 71%|███████ | 34040/48008 [1:56:45<45:13, 5.15it/s] 71%|███████ | 34041/48008 [1:56:45<44:53, 5.19it/s] 71%|███████ | 34042/48008 [1:56:46<44:51, 5.19it/s] 71%|███████ | 34043/48008 [1:56:46<44:19, 5.25it/s] 71%|███████ | 34044/48008 [1:56:46<44:22, 5.24it/s] 71%|███████ | 34045/48008 [1:56:46<44:47, 5.19it/s] 71%|███████ | 34046/48008 [1:56:46<44:57, 5.18it/s] 71%|███████ | 34047/48008 [1:56:47<45:25, 5.12it/s] 71%|███████ | 34048/48008 [1:56:47<45:15, 5.14it/s] 71%|███████ | 34049/48008 [1:56:47<45:03, 5.16it/s] 71%|███████ | 34050/48008 [1:56:47<45:20, 5.13it/s] {'loss': 0.303, 'grad_norm': 0.1552734375, 'learning_rate': 5.815280786535578e-05, 'epoch': 0.71} 71%|███████ | 34050/48008 [1:56:47<45:20, 5.13it/s] 71%|███████ | 34051/48008 [1:56:47<44:55, 5.18it/s] 71%|███████ | 34052/48008 [1:56:47<45:41, 5.09it/s] 71%|███████ | 34053/48008 [1:56:48<46:13, 5.03it/s] 71%|███████ | 34054/48008 [1:56:48<46:09, 5.04it/s] 71%|███████ | 34055/48008 [1:56:48<45:33, 5.11it/s] 71%|███████ | 34056/48008 [1:56:48<45:59, 5.06it/s] 71%|███████ | 34057/48008 [1:56:48<45:26, 5.12it/s] 71%|███████ | 34058/48008 [1:56:49<45:09, 5.15it/s] 71%|███████ | 34059/48008 [1:56:49<45:23, 5.12it/s] 71%|███████ | 34060/48008 [1:56:49<45:03, 5.16it/s] 71%|███████ | 34061/48008 [1:56:49<46:07, 5.04it/s] 71%|███████ | 34062/48008 [1:56:49<45:43, 5.08it/s] 71%|███████ | 34063/48008 [1:56:50<45:31, 5.10it/s] 71%|███████ | 34064/48008 [1:56:50<58:49, 3.95it/s] 71%|███████ | 34065/48008 [1:56:50<54:13, 4.29it/s] 71%|███████ | 34066/48008 [1:56:50<54:22, 4.27it/s] 71%|███████ | 34067/48008 [1:56:51<51:23, 4.52it/s] 71%|███████ | 34068/48008 [1:56:51<49:18, 4.71it/s] 71%|███████ | 34069/48008 [1:56:51<47:52, 4.85it/s] 71%|███████ | 34070/48008 [1:56:51<46:22, 5.01it/s] 71%|███████ | 34071/48008 [1:56:51<45:51, 5.07it/s] 71%|███████ | 34072/48008 [1:56:52<45:14, 5.13it/s] 71%|███████ | 34073/48008 [1:56:52<45:06, 5.15it/s] 71%|███████ | 34074/48008 [1:56:52<45:48, 5.07it/s] 71%|███████ | 34075/48008 [1:56:52<45:22, 5.12it/s] 71%|███████ | 34076/48008 [1:56:52<45:49, 5.07it/s] 71%|███████ | 34077/48008 [1:56:53<45:47, 5.07it/s] 71%|███████ | 34078/48008 [1:56:53<45:18, 5.12it/s] 71%|███████ | 34079/48008 [1:56:53<45:01, 5.16it/s] 71%|███████ | 34080/48008 [1:56:53<58:30, 3.97it/s] 71%|███████ | 34081/48008 [1:56:54<54:05, 4.29it/s] 71%|███████ | 34082/48008 [1:56:54<51:03, 4.55it/s] 71%|███████ | 34083/48008 [1:56:54<49:25, 4.70it/s] 71%|███████ | 34084/48008 [1:56:54<48:33, 4.78it/s] 71%|███████ | 34085/48008 [1:56:54<47:09, 4.92it/s] 71%|███████ | 34086/48008 [1:56:55<46:02, 5.04it/s] 71%|███████ | 34087/48008 [1:56:55<44:54, 5.17it/s] 71%|███████ | 34088/48008 [1:56:55<46:01, 5.04it/s] 71%|███████ | 34089/48008 [1:56:55<45:19, 5.12it/s] 71%|███████ | 34090/48008 [1:56:55<46:16, 5.01it/s] 71%|███████ | 34091/48008 [1:56:55<45:47, 5.07it/s] 71%|███████ | 34092/48008 [1:56:56<45:34, 5.09it/s] 71%|███████ | 34093/48008 [1:56:56<45:54, 5.05it/s] 71%|███████ | 34094/48008 [1:56:56<45:09, 5.13it/s] 71%|███████ | 34095/48008 [1:56:56<45:46, 5.07it/s] 71%|███████ | 34096/48008 [1:56:56<46:08, 5.03it/s] 71%|███████ | 34097/48008 [1:56:57<45:34, 5.09it/s] 71%|███████ | 34098/48008 [1:56:57<46:11, 5.02it/s] 71%|███████ | 34099/48008 [1:56:57<45:42, 5.07it/s] 71%|███████ | 34100/48008 [1:56:57<45:59, 5.04it/s] 71%|███████ | 34100/48008 [1:56:57<45:59, 5.04it/s]{'loss': 0.2981, 'grad_norm': 0.40625, 'learning_rate': 5.794450924845859e-05, 'epoch': 0.71} 71%|███████ | 34101/48008 [1:56:57<46:27, 4.99it/s] 71%|███████ | 34102/48008 [1:56:58<45:50, 5.06it/s] 71%|███████ | 34103/48008 [1:56:58<44:57, 5.15it/s] 71%|███████ | 34104/48008 [1:56:58<45:04, 5.14it/s] 71%|███████ | 34105/48008 [1:56:58<44:46, 5.18it/s] 71%|███████ | 34106/48008 [1:56:58<44:42, 5.18it/s] 71%|███████ | 34107/48008 [1:56:59<45:17, 5.12it/s] 71%|███████ | 34108/48008 [1:56:59<44:52, 5.16it/s] 71%|███████ | 34109/48008 [1:56:59<44:38, 5.19it/s] 71%|███████ | 34110/48008 [1:56:59<45:12, 5.12it/s] 71%|███████ | 34111/48008 [1:56:59<44:55, 5.16it/s] 71%|███████ | 34112/48008 [1:57:00<45:29, 5.09it/s] 71%|███████ | 34113/48008 [1:57:00<45:48, 5.06it/s] 71%|███████ | 34114/48008 [1:57:00<46:07, 5.02it/s] 71%|███████ | 34115/48008 [1:57:00<45:09, 5.13it/s] 71%|███████ | 34116/48008 [1:57:00<45:41, 5.07it/s] 71%|███████ | 34117/48008 [1:57:01<45:11, 5.12it/s] 71%|███████ | 34118/48008 [1:57:01<44:44, 5.17it/s] 71%|███████ | 34119/48008 [1:57:01<44:35, 5.19it/s] 71%|███████ | 34120/48008 [1:57:01<44:57, 5.15it/s] 71%|███████ | 34121/48008 [1:57:01<44:42, 5.18it/s] 71%|███████ | 34122/48008 [1:57:02<44:26, 5.21it/s] 71%|███████ | 34123/48008 [1:57:02<44:50, 5.16it/s] 71%|███████ | 34124/48008 [1:57:02<45:08, 5.13it/s] 71%|███████ | 34125/48008 [1:57:02<45:17, 5.11it/s] 71%|███████ | 34126/48008 [1:57:02<44:53, 5.15it/s] 71%|███████ | 34127/48008 [1:57:03<45:32, 5.08it/s] 71%|███████ | 34128/48008 [1:57:03<44:49, 5.16it/s] 71%|███████ | 34129/48008 [1:57:03<44:19, 5.22it/s] 71%|███████ | 34130/48008 [1:57:03<44:24, 5.21it/s] 71%|███████ | 34131/48008 [1:57:03<44:24, 5.21it/s] 71%|███████ | 34132/48008 [1:57:03<44:00, 5.25it/s] 71%|███████ | 34133/48008 [1:57:04<43:38, 5.30it/s] 71%|███████ | 34134/48008 [1:57:04<44:13, 5.23it/s] 71%|███████ | 34135/48008 [1:57:04<45:30, 5.08it/s] 71%|███████ | 34136/48008 [1:57:04<45:33, 5.08it/s] 71%|███████ | 34137/48008 [1:57:04<45:11, 5.12it/s] 71%|███████ | 34138/48008 [1:57:05<45:45, 5.05it/s] 71%|███████ | 34139/48008 [1:57:05<46:04, 5.02it/s] 71%|███████ | 34140/48008 [1:57:05<45:04, 5.13it/s] 71%|███████ | 34141/48008 [1:57:05<44:48, 5.16it/s] 71%|███████ | 34142/48008 [1:57:05<44:07, 5.24it/s] 71%|███████ | 34143/48008 [1:57:06<44:06, 5.24it/s] 71%|███████ | 34144/48008 [1:57:06<44:04, 5.24it/s] 71%|███████ | 34145/48008 [1:57:06<44:07, 5.24it/s] 71%|███████ | 34146/48008 [1:57:06<44:09, 5.23it/s] 71%|███████ | 34147/48008 [1:57:06<43:54, 5.26it/s] 71%|███████ | 34148/48008 [1:57:07<43:48, 5.27it/s] 71%|███████ | 34149/48008 [1:57:07<43:51, 5.27it/s] 71%|███████ | 34150/48008 [1:57:07<43:43, 5.28it/s] {'loss': 0.3006, 'grad_norm': 0.3515625, 'learning_rate': 5.7736210631561405e-05, 'epoch': 0.71} 71%|███████ | 34150/48008 [1:57:07<43:43, 5.28it/s] 71%|███████ | 34151/48008 [1:57:07<43:54, 5.26it/s] 71%|███████ | 34152/48008 [1:57:07<44:36, 5.18it/s] 71%|███████ | 34153/48008 [1:57:08<44:20, 5.21it/s] 71%|███████ | 34154/48008 [1:57:08<43:59, 5.25it/s] 71%|███████ | 34155/48008 [1:57:08<43:45, 5.28it/s] 71%|███████ | 34156/48008 [1:57:08<45:18, 5.10it/s] 71%|███████ | 34157/48008 [1:57:08<44:59, 5.13it/s] 71%|███████ | 34158/48008 [1:57:08<44:44, 5.16it/s] 71%|███████ | 34159/48008 [1:57:09<45:26, 5.08it/s] 71%|███████ | 34160/48008 [1:57:09<53:03, 4.35it/s] 71%|███████ | 34161/48008 [1:57:09<51:19, 4.50it/s] 71%|███████ | 34162/48008 [1:57:09<50:01, 4.61it/s] 71%|███████ | 34163/48008 [1:57:10<49:08, 4.70it/s] 71%|███████ | 34164/48008 [1:57:10<47:48, 4.83it/s] 71%|███████ | 34165/48008 [1:57:10<46:44, 4.94it/s] 71%|███████ | 34166/48008 [1:57:10<46:13, 4.99it/s] 71%|███████ | 34167/48008 [1:57:10<45:21, 5.09it/s] 71%|███████ | 34168/48008 [1:57:11<53:03, 4.35it/s] 71%|███████ | 34169/48008 [1:57:11<50:18, 4.59it/s] 71%|███████ | 34170/48008 [1:57:11<1:01:58, 3.72it/s] 71%|███████ | 34171/48008 [1:57:11<57:06, 4.04it/s] 71%|███████ | 34172/48008 [1:57:12<53:47, 4.29it/s] 71%|███████ | 34173/48008 [1:57:12<50:56, 4.53it/s] 71%|███████ | 34174/48008 [1:57:12<49:21, 4.67it/s] 71%|███████ | 34175/48008 [1:57:12<48:26, 4.76it/s] 71%|███████ | 34176/48008 [1:57:12<47:09, 4.89it/s] 71%|███████ | 34177/48008 [1:57:13<45:55, 5.02it/s] 71%|███████ | 34178/48008 [1:57:13<45:05, 5.11it/s] 71%|███████ | 34179/48008 [1:57:13<44:45, 5.15it/s] 71%|███████ | 34180/48008 [1:57:14<1:31:30, 2.52it/s] 71%|███████ | 34181/48008 [1:57:14<1:17:16, 2.98it/s] 71%|███████ | 34182/48008 [1:57:14<1:07:27, 3.42it/s] 71%|███████ | 34183/48008 [1:57:14<59:58, 3.84it/s] 71%|███████ | 34184/48008 [1:57:15<55:15, 4.17it/s] 71%|███████ | 34185/48008 [1:57:15<52:04, 4.42it/s] 71%|███████ | 34186/48008 [1:57:15<49:38, 4.64it/s] 71%|███████ | 34187/48008 [1:57:15<48:18, 4.77it/s] 71%|███████ | 34188/48008 [1:57:15<46:50, 4.92it/s] 71%|███████ | 34189/48008 [1:57:16<46:10, 4.99it/s] 71%|███████ | 34190/48008 [1:57:16<45:27, 5.07it/s] 71%|███████ | 34191/48008 [1:57:16<44:55, 5.13it/s] 71%|███████ | 34192/48008 [1:57:16<43:59, 5.23it/s] 71%|███████ | 34193/48008 [1:57:16<44:25, 5.18it/s] 71%|███████ | 34194/48008 [1:57:17<44:22, 5.19it/s] 71%|███████ | 34195/48008 [1:57:17<45:09, 5.10it/s] 71%|███████ | 34196/48008 [1:57:17<44:52, 5.13it/s] 71%|███████ | 34197/48008 [1:57:17<47:46, 4.82it/s] 71%|███████ | 34198/48008 [1:57:17<46:26, 4.96it/s] 71%|███████ | 34199/48008 [1:57:18<46:37, 4.94it/s] 71%|███████ | 34200/48008 [1:57:18<45:11, 5.09it/s] {'loss': 0.3, 'grad_norm': 0.28515625, 'learning_rate': 5.752791201466422e-05, 'epoch': 0.71} 71%|███████ | 34200/48008 [1:57:18<45:11, 5.09it/s] 71%|███████ | 34201/48008 [1:57:18<44:50, 5.13it/s] 71%|███████ | 34202/48008 [1:57:18<45:00, 5.11it/s] 71%|███████ | 34203/48008 [1:57:18<44:17, 5.19it/s] 71%|███████ | 34204/48008 [1:57:19<44:04, 5.22it/s] 71%|███████ | 34205/48008 [1:57:19<43:50, 5.25it/s] 71%|███████▏ | 34206/48008 [1:57:19<44:19, 5.19it/s] 71%|███████▏ | 34207/48008 [1:57:19<44:14, 5.20it/s] 71%|███████▏ | 34208/48008 [1:57:19<44:00, 5.23it/s] 71%|███████▏ | 34209/48008 [1:57:19<43:46, 5.25it/s] 71%|███████▏ | 34210/48008 [1:57:20<43:56, 5.23it/s] 71%|███████▏ | 34211/48008 [1:57:20<43:39, 5.27it/s] 71%|███████▏ | 34212/48008 [1:57:20<43:47, 5.25it/s] 71%|███████▏ | 34213/48008 [1:57:20<43:56, 5.23it/s] 71%|███████▏ | 34214/48008 [1:57:20<47:16, 4.86it/s] 71%|███████▏ | 34215/48008 [1:57:21<46:15, 4.97it/s] 71%|███████▏ | 34216/48008 [1:57:21<45:56, 5.00it/s] 71%|███████▏ | 34217/48008 [1:57:21<45:04, 5.10it/s] 71%|███████▏ | 34218/48008 [1:57:21<44:37, 5.15it/s] 71%|███████▏ | 34219/48008 [1:57:21<44:33, 5.16it/s] 71%|███████▏ | 34220/48008 [1:57:22<44:29, 5.16it/s] 71%|███████▏ | 34221/48008 [1:57:22<44:04, 5.21it/s] 71%|███████▏ | 34222/48008 [1:57:22<44:03, 5.21it/s] 71%|███████▏ | 34223/48008 [1:57:22<43:54, 5.23it/s] 71%|███████▏ | 34224/48008 [1:57:22<43:43, 5.25it/s] 71%|███████▏ | 34225/48008 [1:57:23<51:45, 4.44it/s] 71%|███████▏ | 34226/48008 [1:57:23<48:54, 4.70it/s] 71%|███████▏ | 34227/48008 [1:57:23<48:05, 4.78it/s] 71%|███████▏ | 34228/48008 [1:57:23<46:55, 4.89it/s] 71%|███████▏ | 34229/48008 [1:57:24<59:33, 3.86it/s] 71%|███████▏ | 34230/48008 [1:57:24<55:11, 4.16it/s] 71%|███████▏ | 34231/48008 [1:57:24<51:55, 4.42it/s] 71%|███████▏ | 34232/48008 [1:57:24<50:31, 4.54it/s] 71%|███████▏ | 34233/48008 [1:57:24<48:23, 4.74it/s] 71%|███████▏ | 34234/48008 [1:57:25<47:01, 4.88it/s] 71%|███████▏ | 34235/48008 [1:57:25<45:56, 5.00it/s] 71%|███████▏ | 34236/48008 [1:57:25<53:16, 4.31it/s] 71%|███████▏ | 34237/48008 [1:57:25<50:43, 4.52it/s] 71%|███████▏ | 34238/48008 [1:57:26<48:58, 4.69it/s] 71%|███████▏ | 34239/48008 [1:57:26<47:41, 4.81it/s] 71%|███████▏ | 34240/48008 [1:57:26<46:36, 4.92it/s] 71%|███████▏ | 34241/48008 [1:57:26<46:59, 4.88it/s] 71%|███████▏ | 34242/48008 [1:57:26<45:49, 5.01it/s] 71%|███████▏ | 34243/48008 [1:57:27<53:13, 4.31it/s] 71%|███████▏ | 34244/48008 [1:57:27<50:43, 4.52it/s] 71%|███████▏ | 34245/48008 [1:57:27<49:31, 4.63it/s] 71%|███████▏ | 34246/48008 [1:57:27<48:45, 4.70it/s] 71%|███████▏ | 34247/48008 [1:57:27<47:25, 4.84it/s] 71%|███████▏ | 34248/48008 [1:57:28<46:48, 4.90it/s] 71%|███████▏ | 34249/48008 [1:57:28<45:44, 5.01it/s] 71%|███████▏ | 34250/48008 [1:57:28<45:51, 5.00it/s] {'loss': 0.3011, 'grad_norm': 0.166015625, 'learning_rate': 5.7319613397767046e-05, 'epoch': 0.71} 71%|███████▏ | 34250/48008 [1:57:28<45:51, 5.00it/s] 71%|███████▏ | 34251/48008 [1:57:28<45:13, 5.07it/s] 71%|███████▏ | 34252/48008 [1:57:29<58:21, 3.93it/s] 71%|███████▏ | 34253/48008 [1:57:29<54:27, 4.21it/s] 71%|███████▏ | 34254/48008 [1:57:29<51:10, 4.48it/s] 71%|███████▏ | 34255/48008 [1:57:29<49:26, 4.64it/s] 71%|███████▏ | 34256/48008 [1:57:29<48:30, 4.73it/s] 71%|███████▏ | 34257/48008 [1:57:30<47:27, 4.83it/s] 71%|███████▏ | 34258/48008 [1:57:30<46:40, 4.91it/s] 71%|███████▏ | 34259/48008 [1:57:30<46:08, 4.97it/s] 71%|███████▏ | 34260/48008 [1:57:30<53:23, 4.29it/s] 71%|███████▏ | 34261/48008 [1:57:30<50:34, 4.53it/s] 71%|███████▏ | 34262/48008 [1:57:31<56:27, 4.06it/s] 71%|███████▏ | 34263/48008 [1:57:31<53:04, 4.32it/s] 71%|███████▏ | 34264/48008 [1:57:31<50:08, 4.57it/s] 71%|███████▏ | 34265/48008 [1:57:31<48:55, 4.68it/s] 71%|███████▏ | 34266/48008 [1:57:32<47:17, 4.84it/s] 71%|███████▏ | 34267/48008 [1:57:32<46:16, 4.95it/s] 71%|███████▏ | 34268/48008 [1:57:32<45:35, 5.02it/s] 71%|███████▏ | 34269/48008 [1:57:32<45:54, 4.99it/s] 71%|███████▏ | 34270/48008 [1:57:32<45:31, 5.03it/s] 71%|███████▏ | 34271/48008 [1:57:33<44:51, 5.10it/s] 71%|███████▏ | 34272/48008 [1:57:33<44:18, 5.17it/s] 71%|███████▏ | 34273/48008 [1:57:33<57:33, 3.98it/s] 71%|███████▏ | 34274/48008 [1:57:33<53:27, 4.28it/s] 71%|███████▏ | 34275/48008 [1:57:33<50:32, 4.53it/s] 71%|███████▏ | 34276/48008 [1:57:34<48:23, 4.73it/s] 71%|███████▏ | 34277/48008 [1:57:34<46:39, 4.91it/s] 71%|███████▏ | 34278/48008 [1:57:34<45:27, 5.03it/s] 71%|███████▏ | 34279/48008 [1:57:34<44:37, 5.13it/s] 71%|███████▏ | 34280/48008 [1:57:34<44:53, 5.10it/s] 71%|███████▏ | 34281/48008 [1:57:35<45:53, 4.99it/s] 71%|███████▏ | 34282/48008 [1:57:35<48:16, 4.74it/s] 71%|███████▏ | 34283/48008 [1:57:35<47:32, 4.81it/s] 71%|███████▏ | 34284/48008 [1:57:35<46:16, 4.94it/s] 71%|███████▏ | 34285/48008 [1:57:35<46:49, 4.88it/s] 71%|███████▏ | 34286/48008 [1:57:36<46:08, 4.96it/s] 71%|███████▏ | 34287/48008 [1:57:36<45:25, 5.03it/s] 71%|███████▏ | 34288/48008 [1:57:36<46:10, 4.95it/s] 71%|███████▏ | 34289/48008 [1:57:36<46:33, 4.91it/s] 71%|███████▏ | 34290/48008 [1:57:36<45:19, 5.04it/s] 71%|███████▏ | 34291/48008 [1:57:37<44:47, 5.10it/s] 71%|███████▏ | 34292/48008 [1:57:37<45:11, 5.06it/s] 71%|███████▏ | 34293/48008 [1:57:37<45:07, 5.07it/s] 71%|███████▏ | 34294/48008 [1:57:37<44:56, 5.09it/s] 71%|███████▏ | 34295/48008 [1:57:37<45:33, 5.02it/s] 71%|███████▏ | 34296/48008 [1:57:38<45:40, 5.00it/s] 71%|███████▏ | 34297/48008 [1:57:38<44:47, 5.10it/s] 71%|███████▏ | 34298/48008 [1:57:38<43:59, 5.19it/s] 71%|███████▏ | 34299/48008 [1:57:38<43:55, 5.20it/s] 71%|███████▏ | 34300/48008 [1:57:38<44:39, 5.12it/s] {'loss': 0.301, 'grad_norm': 0.150390625, 'learning_rate': 5.7111314780869864e-05, 'epoch': 0.71} 71%|███████▏ | 34300/48008 [1:57:38<44:39, 5.12it/s] 71%|███████▏ | 34301/48008 [1:57:39<44:25, 5.14it/s] 71%|███████▏ | 34302/48008 [1:57:39<44:25, 5.14it/s] 71%|███████▏ | 34303/48008 [1:57:39<44:04, 5.18it/s] 71%|███████▏ | 34304/48008 [1:57:39<52:00, 4.39it/s] 71%|███████▏ | 34305/48008 [1:57:40<52:36, 4.34it/s] 71%|███████▏ | 34306/48008 [1:57:40<49:57, 4.57it/s] 71%|███████▏ | 34307/48008 [1:57:40<51:13, 4.46it/s] 71%|███████▏ | 34308/48008 [1:57:40<49:06, 4.65it/s] 71%|███████▏ | 34309/48008 [1:57:40<47:47, 4.78it/s] 71%|███████▏ | 34310/48008 [1:57:41<46:50, 4.87it/s] 71%|███████▏ | 34311/48008 [1:57:41<47:11, 4.84it/s] 71%|███████▏ | 34312/48008 [1:57:41<49:16, 4.63it/s] 71%|███████▏ | 34313/48008 [1:57:41<47:14, 4.83it/s] 71%|███████▏ | 34314/48008 [1:57:41<46:10, 4.94it/s] 71%|███████▏ | 34315/48008 [1:57:42<46:16, 4.93it/s] 71%|███████▏ | 34316/48008 [1:57:42<45:20, 5.03it/s] 71%|███████▏ | 34317/48008 [1:57:42<44:26, 5.13it/s] 71%|███████▏ | 34318/48008 [1:57:42<45:09, 5.05it/s] 71%|███████▏ | 34319/48008 [1:57:42<44:46, 5.09it/s] 71%|███████▏ | 34320/48008 [1:57:43<44:27, 5.13it/s] 71%|███████▏ | 34321/48008 [1:57:43<44:24, 5.14it/s] 71%|███████▏ | 34322/48008 [1:57:43<44:05, 5.17it/s] 71%|███████▏ | 34323/48008 [1:57:43<43:59, 5.19it/s] 71%|███████▏ | 34324/48008 [1:57:43<44:40, 5.11it/s] 71%|███████▏ | 34325/48008 [1:57:44<44:16, 5.15it/s] 72%|███████▏ | 34326/48008 [1:57:44<44:06, 5.17it/s] 72%|███████▏ | 34327/48008 [1:57:44<43:45, 5.21it/s] 72%|███████▏ | 34328/48008 [1:57:44<43:43, 5.21it/s] 72%|███████▏ | 34329/48008 [1:57:44<43:45, 5.21it/s] 72%|███████▏ | 34330/48008 [1:57:44<44:31, 5.12it/s] 72%|███████▏ | 34331/48008 [1:57:45<44:17, 5.15it/s] 72%|███████▏ | 34332/48008 [1:57:45<45:18, 5.03it/s] 72%|███████▏ | 34333/48008 [1:57:45<44:55, 5.07it/s] 72%|███████▏ | 34334/48008 [1:57:45<44:14, 5.15it/s] 72%|███████▏ | 34335/48008 [1:57:45<44:21, 5.14it/s] 72%|███████▏ | 34336/48008 [1:57:46<47:20, 4.81it/s] 72%|███████▏ | 34337/48008 [1:57:46<45:49, 4.97it/s] 72%|███████▏ | 34338/48008 [1:57:46<44:50, 5.08it/s] 72%|███████▏ | 34339/48008 [1:57:46<44:13, 5.15it/s] 72%|███████▏ | 34340/48008 [1:57:46<44:46, 5.09it/s] 72%|███████▏ | 34341/48008 [1:57:47<44:42, 5.09it/s] 72%|███████▏ | 34342/48008 [1:57:47<47:27, 4.80it/s] 72%|███████▏ | 34343/48008 [1:57:47<46:47, 4.87it/s] 72%|███████▏ | 34344/48008 [1:57:47<45:51, 4.97it/s] 72%|███████▏ | 34345/48008 [1:57:47<46:06, 4.94it/s] 72%|███████▏ | 34346/48008 [1:57:48<46:02, 4.95it/s] 72%|███████▏ | 34347/48008 [1:57:48<45:05, 5.05it/s] 72%|███████▏ | 34348/48008 [1:57:48<47:47, 4.76it/s] 72%|███████▏ | 34349/48008 [1:57:48<46:20, 4.91it/s] 72%|███████▏ | 34350/48008 [1:57:48<45:49, 4.97it/s] {'loss': 0.2997, 'grad_norm': 0.2490234375, 'learning_rate': 5.690301616397268e-05, 'epoch': 0.72} 72%|███████▏ | 34350/48008 [1:57:49<45:49, 4.97it/s] 72%|███████▏ | 34351/48008 [1:57:49<45:17, 5.03it/s] 72%|███████▏ | 34352/48008 [1:57:49<45:30, 5.00it/s] 72%|███████▏ | 34353/48008 [1:57:49<45:39, 4.98it/s] 72%|███████▏ | 34354/48008 [1:57:49<44:47, 5.08it/s] 72%|███████▏ | 34355/48008 [1:57:49<44:25, 5.12it/s] 72%|███████▏ | 34356/48008 [1:57:50<44:21, 5.13it/s] 72%|███████▏ | 34357/48008 [1:57:50<44:01, 5.17it/s] 72%|███████▏ | 34358/48008 [1:57:50<43:04, 5.28it/s] 72%|███████▏ | 34359/48008 [1:57:50<43:23, 5.24it/s] 72%|███████▏ | 34360/48008 [1:57:50<43:22, 5.24it/s] 72%|███████▏ | 34361/48008 [1:57:51<43:23, 5.24it/s] 72%|███████▏ | 34362/48008 [1:57:51<43:28, 5.23it/s] 72%|███████▏ | 34363/48008 [1:57:51<43:26, 5.23it/s] 72%|███████▏ | 34364/48008 [1:57:51<46:46, 4.86it/s] 72%|███████▏ | 34365/48008 [1:57:51<45:35, 4.99it/s] 72%|███████▏ | 34366/48008 [1:57:52<45:10, 5.03it/s] 72%|███████▏ | 34367/48008 [1:57:52<44:39, 5.09it/s] 72%|███████▏ | 34368/48008 [1:57:52<44:01, 5.16it/s] 72%|███████▏ | 34369/48008 [1:57:52<43:36, 5.21it/s] 72%|███████▏ | 34370/48008 [1:57:52<43:56, 5.17it/s] 72%|███████▏ | 34371/48008 [1:57:53<44:36, 5.10it/s] 72%|███████▏ | 34372/48008 [1:57:53<44:14, 5.14it/s] 72%|███████▏ | 34373/48008 [1:57:53<51:59, 4.37it/s] 72%|███████▏ | 34374/48008 [1:57:53<49:12, 4.62it/s] 72%|███████▏ | 34375/48008 [1:57:53<47:15, 4.81it/s] 72%|███████▏ | 34376/48008 [1:57:54<45:59, 4.94it/s] 72%|███████▏ | 34377/48008 [1:57:54<44:36, 5.09it/s] 72%|███████▏ | 34378/48008 [1:57:54<43:58, 5.17it/s] 72%|███████▏ | 34379/48008 [1:57:54<43:48, 5.18it/s] 72%|███████▏ | 34380/48008 [1:57:54<43:20, 5.24it/s] 72%|███████▏ | 34381/48008 [1:57:55<43:29, 5.22it/s] 72%|███████▏ | 34382/48008 [1:57:55<43:12, 5.26it/s] 72%|███████▏ | 34383/48008 [1:57:55<42:57, 5.29it/s] 72%|███████▏ | 34384/48008 [1:57:55<42:39, 5.32it/s] 72%|███████▏ | 34385/48008 [1:57:55<42:44, 5.31it/s] 72%|███████▏ | 34386/48008 [1:57:56<42:52, 5.30it/s] 72%|███████▏ | 34387/48008 [1:57:56<43:05, 5.27it/s] 72%|███████▏ | 34388/48008 [1:57:56<56:27, 4.02it/s] 72%|███████▏ | 34389/48008 [1:57:56<52:21, 4.34it/s] 72%|███████▏ | 34390/48008 [1:57:56<49:39, 4.57it/s] 72%|███████▏ | 34391/48008 [1:57:57<47:35, 4.77it/s] 72%|███████▏ | 34392/48008 [1:57:57<46:44, 4.85it/s] 72%|███████▏ | 34393/48008 [1:57:57<45:42, 4.97it/s] 72%|███████▏ | 34394/48008 [1:57:57<45:49, 4.95it/s] 72%|███████▏ | 34395/48008 [1:57:57<45:09, 5.02it/s] 72%|███████▏ | 34396/48008 [1:57:58<45:16, 5.01it/s] 72%|███████▏ | 34397/48008 [1:57:58<44:24, 5.11it/s] 72%|███████▏ | 34398/48008 [1:57:58<43:51, 5.17it/s] 72%|███████▏ | 34399/48008 [1:57:58<43:37, 5.20it/s] 72%|███████▏ | 34400/48008 [1:57:58<43:34, 5.21it/s] {'loss': 0.3008, 'grad_norm': 0.2353515625, 'learning_rate': 5.669471754707549e-05, 'epoch': 0.72} 72%|███████▏ | 34400/48008 [1:57:58<43:34, 5.21it/s] 72%|███████▏ | 34401/48008 [1:57:59<43:11, 5.25it/s] 72%|███████▏ | 34402/48008 [1:57:59<43:05, 5.26it/s] 72%|███████▏ | 34403/48008 [1:57:59<43:09, 5.25it/s] 72%|███████▏ | 34404/48008 [1:57:59<43:15, 5.24it/s] 72%|███████▏ | 34405/48008 [1:57:59<44:07, 5.14it/s] 72%|███████▏ | 34406/48008 [1:58:00<43:39, 5.19it/s] 72%|███████▏ | 34407/48008 [1:58:00<43:49, 5.17it/s] 72%|███████▏ | 34408/48008 [1:58:00<43:16, 5.24it/s] 72%|███████▏ | 34409/48008 [1:58:00<44:01, 5.15it/s] 72%|███████▏ | 34410/48008 [1:58:00<44:32, 5.09it/s] 72%|███████▏ | 34411/48008 [1:58:01<45:27, 4.99it/s] 72%|███████▏ | 34412/48008 [1:58:01<44:49, 5.05it/s] 72%|███████▏ | 34413/48008 [1:58:01<45:17, 5.00it/s] 72%|███████▏ | 34414/48008 [1:58:01<44:37, 5.08it/s] 72%|███████▏ | 34415/48008 [1:58:01<44:56, 5.04it/s] 72%|███████▏ | 34416/48008 [1:58:02<44:00, 5.15it/s] 72%|███████▏ | 34417/48008 [1:58:02<44:06, 5.14it/s] 72%|███████▏ | 34418/48008 [1:58:02<43:45, 5.18it/s] 72%|███████▏ | 34419/48008 [1:58:02<44:28, 5.09it/s] 72%|███████▏ | 34420/48008 [1:58:02<43:55, 5.16it/s] 72%|███████▏ | 34421/48008 [1:58:03<46:57, 4.82it/s] 72%|███████▏ | 34422/48008 [1:58:03<45:35, 4.97it/s] 72%|███████▏ | 34423/48008 [1:58:03<45:38, 4.96it/s] 72%|███████▏ | 34424/48008 [1:58:03<44:36, 5.08it/s] 72%|███████▏ | 34425/48008 [1:58:03<43:58, 5.15it/s] 72%|███████▏ | 34426/48008 [1:58:04<43:50, 5.16it/s] 72%|███████▏ | 34427/48008 [1:58:04<44:29, 5.09it/s] 72%|███████▏ | 34428/48008 [1:58:04<43:53, 5.16it/s] 72%|███████▏ | 34429/48008 [1:58:04<44:56, 5.04it/s] 72%|███████▏ | 34430/48008 [1:58:04<44:43, 5.06it/s] 72%|███████▏ | 34431/48008 [1:58:04<44:09, 5.12it/s] 72%|███████▏ | 34432/48008 [1:58:05<43:50, 5.16it/s] 72%|███████▏ | 34433/48008 [1:58:05<43:56, 5.15it/s] 72%|███████▏ | 34434/48008 [1:58:05<43:49, 5.16it/s] 72%|███████▏ | 34435/48008 [1:58:05<43:39, 5.18it/s] 72%|███████▏ | 34436/48008 [1:58:05<43:36, 5.19it/s] 72%|███████▏ | 34437/48008 [1:58:06<43:11, 5.24it/s] 72%|███████▏ | 34438/48008 [1:58:06<42:49, 5.28it/s] 72%|███████▏ | 34439/48008 [1:58:06<42:52, 5.27it/s] 72%|███████▏ | 34440/48008 [1:58:06<42:42, 5.29it/s] 72%|███████▏ | 34441/48008 [1:58:07<56:03, 4.03it/s] 72%|███████▏ | 34442/48008 [1:58:07<52:05, 4.34it/s] 72%|███████▏ | 34443/48008 [1:58:07<49:18, 4.58it/s] 72%|███████▏ | 34444/48008 [1:58:07<47:30, 4.76it/s] 72%|███████▏ | 34445/48008 [1:58:07<46:00, 4.91it/s] 72%|███████▏ | 34446/48008 [1:58:08<44:53, 5.03it/s] 72%|███████▏ | 34447/48008 [1:58:08<44:22, 5.09it/s] 72%|███████▏ | 34448/48008 [1:58:08<44:19, 5.10it/s] 72%|███████▏ | 34449/48008 [1:58:08<43:54, 5.15it/s] 72%|███████▏ | 34450/48008 [1:58:08<43:19, 5.22it/s] {'loss': 0.3002, 'grad_norm': 0.24609375, 'learning_rate': 5.648641893017831e-05, 'epoch': 0.72} 72%|███████▏ | 34450/48008 [1:58:08<43:19, 5.22it/s] 72%|███████▏ | 34451/48008 [1:58:08<44:02, 5.13it/s] 72%|███████▏ | 34452/48008 [1:58:09<43:22, 5.21it/s] 72%|███████▏ | 34453/48008 [1:58:09<42:53, 5.27it/s] 72%|███████▏ | 34454/48008 [1:58:09<43:51, 5.15it/s] 72%|███████▏ | 34455/48008 [1:58:09<43:20, 5.21it/s] 72%|███████▏ | 34456/48008 [1:58:09<44:29, 5.08it/s] 72%|███████▏ | 34457/48008 [1:58:10<45:00, 5.02it/s] 72%|███████▏ | 34458/48008 [1:58:10<44:26, 5.08it/s] 72%|███████▏ | 34459/48008 [1:58:10<44:32, 5.07it/s] 72%|███████▏ | 34460/48008 [1:58:10<44:06, 5.12it/s] 72%|███████▏ | 34461/48008 [1:58:10<43:55, 5.14it/s] 72%|███████▏ | 34462/48008 [1:58:11<44:30, 5.07it/s] 72%|███████▏ | 34463/48008 [1:58:11<44:32, 5.07it/s] 72%|███████▏ | 34464/48008 [1:58:11<44:34, 5.06it/s] 72%|███████▏ | 34465/48008 [1:58:11<44:05, 5.12it/s] 72%|███████▏ | 34466/48008 [1:58:11<47:04, 4.79it/s] 72%|███████▏ | 34467/48008 [1:58:12<45:58, 4.91it/s] 72%|███████▏ | 34468/48008 [1:58:12<44:58, 5.02it/s] 72%|███████▏ | 34469/48008 [1:58:12<44:04, 5.12it/s] 72%|███████▏ | 34470/48008 [1:58:12<43:48, 5.15it/s] 72%|███████▏ | 34471/48008 [1:58:12<43:37, 5.17it/s] 72%|███████▏ | 34472/48008 [1:58:13<43:33, 5.18it/s] 72%|███████▏ | 34473/48008 [1:58:13<51:20, 4.39it/s] 72%|███████▏ | 34474/48008 [1:58:13<48:57, 4.61it/s] 72%|███████▏ | 34475/48008 [1:58:13<46:53, 4.81it/s] 72%|███████▏ | 34476/48008 [1:58:13<45:38, 4.94it/s] 72%|███████▏ | 34477/48008 [1:58:14<44:43, 5.04it/s] 72%|███████▏ | 34478/48008 [1:58:14<43:46, 5.15it/s] 72%|███████▏ | 34479/48008 [1:58:14<56:49, 3.97it/s] 72%|███████▏ | 34480/48008 [1:58:14<53:25, 4.22it/s] 72%|███████▏ | 34481/48008 [1:58:15<50:16, 4.48it/s] 72%|███████▏ | 34482/48008 [1:58:15<48:54, 4.61it/s] 72%|███████▏ | 34483/48008 [1:58:15<47:10, 4.78it/s] 72%|███████▏ | 34484/48008 [1:58:15<45:34, 4.94it/s] 72%|███████▏ | 34485/48008 [1:58:15<44:50, 5.03it/s] 72%|███████▏ | 34486/48008 [1:58:16<44:19, 5.08it/s] 72%|███████▏ | 34487/48008 [1:58:16<44:29, 5.07it/s] 72%|███████▏ | 34488/48008 [1:58:16<44:48, 5.03it/s] 72%|███████▏ | 34489/48008 [1:58:16<44:03, 5.11it/s] 72%|███████▏ | 34490/48008 [1:58:16<43:47, 5.15it/s] 72%|███████▏ | 34491/48008 [1:58:17<56:48, 3.97it/s] 72%|███████▏ | 34492/48008 [1:58:17<52:40, 4.28it/s] 72%|███████▏ | 34493/48008 [1:58:17<49:48, 4.52it/s] 72%|███████▏ | 34494/48008 [1:58:17<47:39, 4.73it/s] 72%|███████▏ | 34495/48008 [1:58:18<46:44, 4.82it/s] 72%|███████▏ | 34496/48008 [1:58:18<45:23, 4.96it/s] 72%|███████▏ | 34497/48008 [1:58:18<45:26, 4.96it/s] 72%|███████▏ | 34498/48008 [1:58:18<44:29, 5.06it/s] 72%|███████▏ | 34499/48008 [1:58:18<44:05, 5.11it/s] 72%|███████▏ | 34500/48008 [1:58:19<43:48, 5.14it/s] 72%|███████▏ | 34500/48008 [1:58:19<43:48, 5.14it/s]{'loss': 0.2994, 'grad_norm': 0.14453125, 'learning_rate': 5.6278120313281126e-05, 'epoch': 0.72} 72%|███████▏ | 34501/48008 [1:58:19<43:11, 5.21it/s] 72%|███████▏ | 34502/48008 [1:58:19<43:14, 5.21it/s] 72%|███████▏ | 34503/48008 [1:58:19<43:17, 5.20it/s] 72%|███████▏ | 34504/48008 [1:58:19<44:29, 5.06it/s] 72%|███████▏ | 34505/48008 [1:58:19<44:03, 5.11it/s] 72%|███████▏ | 34506/48008 [1:58:20<43:46, 5.14it/s] 72%|███████▏ | 34507/48008 [1:58:20<44:17, 5.08it/s] 72%|███████▏ | 34508/48008 [1:58:20<47:01, 4.78it/s] 72%|███████▏ | 34509/48008 [1:58:20<46:42, 4.82it/s] 72%|███████▏ | 34510/48008 [1:58:21<45:34, 4.94it/s] 72%|███████▏ | 34511/48008 [1:58:21<44:42, 5.03it/s] 72%|███████▏ | 34512/48008 [1:58:21<44:12, 5.09it/s] 72%|███████▏ | 34513/48008 [1:58:21<44:10, 5.09it/s] 72%|███████▏ | 34514/48008 [1:58:21<44:58, 5.00it/s] 72%|███████▏ | 34515/48008 [1:58:21<44:20, 5.07it/s] 72%|███████▏ | 34516/48008 [1:58:22<44:02, 5.11it/s] 72%|███████▏ | 34517/48008 [1:58:22<44:15, 5.08it/s] 72%|███████▏ | 34518/48008 [1:58:22<44:37, 5.04it/s] 72%|███████▏ | 34519/48008 [1:58:22<43:22, 5.18it/s] 72%|███████▏ | 34520/48008 [1:58:22<43:11, 5.21it/s] 72%|███████▏ | 34521/48008 [1:58:23<43:03, 5.22it/s] 72%|███████▏ | 34522/48008 [1:58:24<1:48:06, 2.08it/s] 72%|███████▏ | 34523/48008 [1:58:24<1:28:16, 2.55it/s] 72%|███████▏ | 34524/48008 [1:58:24<1:14:37, 3.01it/s] 72%|███████▏ | 34525/48008 [1:58:24<1:05:22, 3.44it/s] 72%|███████▏ | 34526/48008 [1:58:25<59:06, 3.80it/s] 72%|███████▏ | 34527/48008 [1:58:25<54:39, 4.11it/s] 72%|███████▏ | 34528/48008 [1:58:25<51:12, 4.39it/s] 72%|███████▏ | 34529/48008 [1:58:25<48:45, 4.61it/s] 72%|███████▏ | 34530/48008 [1:58:25<46:51, 4.79it/s] 72%|███████▏ | 34531/48008 [1:58:26<45:40, 4.92it/s] 72%|███████▏ | 34532/48008 [1:58:26<44:43, 5.02it/s] 72%|███████▏ | 34533/48008 [1:58:26<57:22, 3.91it/s] 72%|███████▏ | 34534/48008 [1:58:26<52:51, 4.25it/s] 72%|███████▏ | 34535/48008 [1:58:26<49:56, 4.50it/s] 72%|███████▏ | 34536/48008 [1:58:27<47:44, 4.70it/s] 72%|███████▏ | 34537/48008 [1:58:27<49:21, 4.55it/s] 72%|███████▏ | 34538/48008 [1:58:27<47:25, 4.73it/s] 72%|███████▏ | 34539/48008 [1:58:27<45:53, 4.89it/s] 72%|███████▏ | 34540/48008 [1:58:27<44:44, 5.02it/s] 72%|███████▏ | 34541/48008 [1:58:28<43:59, 5.10it/s] 72%|███████▏ | 34542/48008 [1:58:29<1:30:09, 2.49it/s] 72%|███████▏ | 34543/48008 [1:58:29<1:15:45, 2.96it/s] 72%|███████▏ | 34544/48008 [1:58:29<1:05:51, 3.41it/s] 72%|███████▏ | 34545/48008 [1:58:29<1:00:18, 3.72it/s] 72%|███████▏ | 34546/48008 [1:58:29<54:33, 4.11it/s] 72%|███████▏ | 34547/48008 [1:58:30<50:41, 4.43it/s] 72%|███████▏ | 34548/48008 [1:58:30<1:40:46, 2.23it/s] 72%|███████▏ | 34549/48008 [1:58:31<1:23:15, 2.69it/s] 72%|███████▏ | 34550/48008 [1:58:31<1:11:02, 3.16it/s] {'loss': 0.3004, 'grad_norm': 0.1689453125, 'learning_rate': 5.606982169638394e-05, 'epoch': 0.72} 72%|███████▏ | 34550/48008 [1:58:31<1:11:02, 3.16it/s] 72%|███████▏ | 34551/48008 [1:58:31<1:03:27, 3.53it/s] 72%|███████▏ | 34552/48008 [1:58:31<57:36, 3.89it/s] 72%|███████▏ | 34553/48008 [1:58:31<53:56, 4.16it/s] 72%|███████▏ | 34554/48008 [1:58:32<51:17, 4.37it/s] 72%|███████▏ | 34555/48008 [1:58:32<48:41, 4.60it/s] 72%|███████▏ | 34556/48008 [1:58:32<46:16, 4.85it/s] 72%|███████▏ | 34557/48008 [1:58:32<44:57, 4.99it/s] 72%|███████▏ | 34558/48008 [1:58:32<44:01, 5.09it/s] 72%|███████▏ | 34559/48008 [1:58:33<43:47, 5.12it/s] 72%|███████▏ | 34560/48008 [1:58:33<1:26:48, 2.58it/s] 72%|███████▏ | 34561/48008 [1:58:34<1:13:31, 3.05it/s] 72%|███████▏ | 34562/48008 [1:58:34<1:04:19, 3.48it/s] 72%|███████▏ | 34563/48008 [1:58:34<57:30, 3.90it/s] 72%|███████▏ | 34564/48008 [1:58:34<52:54, 4.23it/s] 72%|███████▏ | 34565/48008 [1:58:34<50:34, 4.43it/s] 72%|███████▏ | 34566/48008 [1:58:35<48:23, 4.63it/s] 72%|███████▏ | 34567/48008 [1:58:35<46:48, 4.79it/s] 72%|███████▏ | 34568/48008 [1:58:35<46:22, 4.83it/s] 72%|███████▏ | 34569/48008 [1:58:35<45:11, 4.96it/s] 72%|███████▏ | 34570/48008 [1:58:35<44:27, 5.04it/s] 72%|███████▏ | 34571/48008 [1:58:36<44:29, 5.03it/s] 72%|███████▏ | 34572/48008 [1:58:36<43:37, 5.13it/s] 72%|███████▏ | 34573/48008 [1:58:36<43:25, 5.16it/s] 72%|███████▏ | 34574/48008 [1:58:36<43:23, 5.16it/s] 72%|███████▏ | 34575/48008 [1:58:36<43:10, 5.18it/s] 72%|███████▏ | 34576/48008 [1:58:36<42:40, 5.25it/s] 72%|███████▏ | 34577/48008 [1:58:37<43:58, 5.09it/s] 72%|███████▏ | 34578/48008 [1:58:37<43:50, 5.10it/s] 72%|███████▏ | 34579/48008 [1:58:37<44:29, 5.03it/s] 72%|███████▏ | 34580/48008 [1:58:37<43:47, 5.11it/s] 72%|███████▏ | 34581/48008 [1:58:37<43:26, 5.15it/s] 72%|███████▏ | 34582/48008 [1:58:38<44:04, 5.08it/s] 72%|███████▏ | 34583/48008 [1:58:38<51:31, 4.34it/s] 72%|███████▏ | 34584/48008 [1:58:38<49:04, 4.56it/s] 72%|███████▏ | 34585/48008 [1:58:38<47:00, 4.76it/s] 72%|███████▏ | 34586/48008 [1:58:39<46:02, 4.86it/s] 72%|███████▏ | 34587/48008 [1:58:39<45:15, 4.94it/s] 72%|███████▏ | 34588/48008 [1:58:39<47:43, 4.69it/s] 72%|███████▏ | 34589/48008 [1:58:39<46:28, 4.81it/s] 72%|███████▏ | 34590/48008 [1:58:39<45:26, 4.92it/s] 72%|███████▏ | 34591/48008 [1:58:40<44:50, 4.99it/s] 72%|███████▏ | 34592/48008 [1:58:40<44:12, 5.06it/s] 72%|███████▏ | 34593/48008 [1:58:40<44:16, 5.05it/s] 72%|███████▏ | 34594/48008 [1:58:40<47:01, 4.75it/s] 72%|███████▏ | 34595/48008 [1:58:41<58:53, 3.80it/s] 72%|███████▏ | 34596/48008 [1:58:41<1:35:58, 2.33it/s] 72%|███████▏ | 34597/48008 [1:58:42<1:20:06, 2.79it/s] 72%|███████▏ | 34598/48008 [1:58:42<1:08:53, 3.24it/s] 72%|███████▏ | 34599/48008 [1:58:42<1:01:10, 3.65it/s] 72%|███████▏ | 34600/48008 [1:58:42<55:49, 4.00it/s] 72%|███████▏ | 34600/48008 [1:58:42<55:49, 4.00it/s]{'loss': 0.301, 'grad_norm': 0.27734375, 'learning_rate': 5.5861523079486754e-05, 'epoch': 0.72} 72%|███████▏ | 34601/48008 [1:58:42<53:15, 4.20it/s] 72%|███████▏ | 34602/48008 [1:58:43<50:23, 4.43it/s] 72%|███████▏ | 34603/48008 [1:58:43<48:50, 4.57it/s] 72%|███████▏ | 34604/48008 [1:58:43<46:41, 4.79it/s] 72%|███████▏ | 34605/48008 [1:58:43<58:31, 3.82it/s] 72%|███████▏ | 34606/48008 [1:58:44<53:26, 4.18it/s] 72%|███████▏ | 34607/48008 [1:58:44<49:58, 4.47it/s] 72%|███████▏ | 34608/48008 [1:58:44<55:34, 4.02it/s] 72%|███████▏ | 34609/48008 [1:58:44<51:34, 4.33it/s] 72%|███████▏ | 34610/48008 [1:58:44<52:00, 4.29it/s] 72%|███████▏ | 34611/48008 [1:58:45<49:10, 4.54it/s] 72%|███████▏ | 34612/48008 [1:58:45<47:32, 4.70it/s] 72%|███████▏ | 34613/48008 [1:58:45<46:23, 4.81it/s] 72%|███████▏ | 34614/48008 [1:58:45<45:12, 4.94it/s] 72%|███████▏ | 34615/48008 [1:58:45<44:23, 5.03it/s] 72%|███████▏ | 34616/48008 [1:58:46<44:39, 5.00it/s] 72%|███████▏ | 34617/48008 [1:58:46<43:46, 5.10it/s] 72%|███████▏ | 34618/48008 [1:58:46<43:24, 5.14it/s] 72%|███████▏ | 34619/48008 [1:58:46<43:06, 5.18it/s] 72%|███████▏ | 34620/48008 [1:58:46<42:48, 5.21it/s] 72%|███████▏ | 34621/48008 [1:58:47<42:45, 5.22it/s] 72%|███████▏ | 34622/48008 [1:58:47<43:25, 5.14it/s] 72%|███████▏ | 34623/48008 [1:58:47<42:58, 5.19it/s] 72%|███████▏ | 34624/48008 [1:58:47<43:31, 5.12it/s] 72%|███████▏ | 34625/48008 [1:58:47<43:40, 5.11it/s] 72%|███████▏ | 34626/48008 [1:58:48<44:21, 5.03it/s] 72%|███████▏ | 34627/48008 [1:58:48<44:43, 4.99it/s] 72%|███████▏ | 34628/48008 [1:58:48<44:11, 5.05it/s] 72%|███████▏ | 34629/48008 [1:58:48<43:46, 5.09it/s] 72%|███████▏ | 34630/48008 [1:58:48<43:29, 5.13it/s] 72%|███████▏ | 34631/48008 [1:58:49<43:17, 5.15it/s] 72%|███████▏ | 34632/48008 [1:58:49<43:20, 5.14it/s] 72%|███████▏ | 34633/48008 [1:58:49<43:11, 5.16it/s] 72%|███████▏ | 34634/48008 [1:58:49<43:41, 5.10it/s] 72%|███████▏ | 34635/48008 [1:58:49<43:31, 5.12it/s] 72%|███████▏ | 34636/48008 [1:58:50<43:42, 5.10it/s] 72%|███████▏ | 34637/48008 [1:58:50<44:32, 5.00it/s] 72%|███████▏ | 34638/48008 [1:58:50<43:53, 5.08it/s] 72%|███████▏ | 34639/48008 [1:58:50<51:15, 4.35it/s] 72%|███████▏ | 34640/48008 [1:58:50<49:57, 4.46it/s] 72%|███████▏ | 34641/48008 [1:58:51<55:39, 4.00it/s] 72%|███████▏ | 34642/48008 [1:58:51<51:44, 4.31it/s] 72%|███████▏ | 34643/48008 [1:58:51<49:21, 4.51it/s] 72%|███████▏ | 34644/48008 [1:58:51<47:25, 4.70it/s] 72%|███████▏ | 34645/48008 [1:58:52<46:18, 4.81it/s] 72%|███████▏ | 34646/48008 [1:58:52<45:02, 4.94it/s] 72%|███████▏ | 34647/48008 [1:58:52<44:10, 5.04it/s] 72%|███████▏ | 34648/48008 [1:58:52<43:36, 5.11it/s] 72%|███████▏ | 34649/48008 [1:58:52<44:15, 5.03it/s] 72%|███████▏ | 34650/48008 [1:58:53<43:53, 5.07it/s] 72%|███████▏ | 34650/48008 [1:58:53<43:53, 5.07it/s]{'loss': 0.2989, 'grad_norm': 0.1650390625, 'learning_rate': 5.565322446258957e-05, 'epoch': 0.72} 72%|███████▏ | 34651/48008 [1:58:53<44:01, 5.06it/s] 72%|███████▏ | 34652/48008 [1:58:53<43:38, 5.10it/s] 72%|███████▏ | 34653/48008 [1:58:53<43:14, 5.15it/s] 72%|███████▏ | 34654/48008 [1:58:53<43:09, 5.16it/s] 72%|███████▏ | 34655/48008 [1:58:53<43:01, 5.17it/s] 72%|███████▏ | 34656/48008 [1:58:54<43:40, 5.10it/s] 72%|███████▏ | 34657/48008 [1:58:54<46:25, 4.79it/s] 72%|███████▏ | 34658/48008 [1:58:54<45:15, 4.92it/s] 72%|███████▏ | 34659/48008 [1:58:54<45:09, 4.93it/s] 72%|███████▏ | 34660/48008 [1:58:55<44:49, 4.96it/s] 72%|███████▏ | 34661/48008 [1:58:55<43:58, 5.06it/s] 72%|███████▏ | 34662/48008 [1:58:55<44:03, 5.05it/s] 72%|███████▏ | 34663/48008 [1:58:55<51:19, 4.33it/s] 72%|███████▏ | 34664/48008 [1:58:55<48:43, 4.56it/s] 72%|███████▏ | 34665/48008 [1:58:56<46:44, 4.76it/s] 72%|███████▏ | 34666/48008 [1:58:56<45:15, 4.91it/s] 72%|███████▏ | 34667/48008 [1:58:56<44:09, 5.03it/s] 72%|███████▏ | 34668/48008 [1:58:56<44:21, 5.01it/s] 72%|███████▏ | 34669/48008 [1:58:56<44:07, 5.04it/s] 72%|███████▏ | 34670/48008 [1:58:57<43:31, 5.11it/s] 72%|███████▏ | 34671/48008 [1:58:57<43:08, 5.15it/s] 72%|███████▏ | 34672/48008 [1:58:57<44:09, 5.03it/s] 72%|███████▏ | 34673/48008 [1:58:57<44:00, 5.05it/s] 72%|███████▏ | 34674/48008 [1:58:57<43:37, 5.09it/s] 72%|███████▏ | 34675/48008 [1:58:58<43:29, 5.11it/s] 72%|███████▏ | 34676/48008 [1:58:58<42:59, 5.17it/s] 72%|███████▏ | 34677/48008 [1:58:58<43:02, 5.16it/s] 72%|███████▏ | 34678/48008 [1:58:58<42:54, 5.18it/s] 72%|███████▏ | 34679/48008 [1:58:58<42:53, 5.18it/s] 72%|███████▏ | 34680/48008 [1:58:59<43:05, 5.15it/s] 72%|███████▏ | 34681/48008 [1:58:59<43:20, 5.12it/s] 72%|███████▏ | 34682/48008 [1:58:59<42:42, 5.20it/s] 72%|███████▏ | 34683/48008 [1:58:59<42:22, 5.24it/s] 72%|███████▏ | 34684/48008 [1:58:59<42:11, 5.26it/s] 72%|███████▏ | 34685/48008 [1:58:59<42:17, 5.25it/s] 72%|███████▏ | 34686/48008 [1:59:00<42:36, 5.21it/s] 72%|███████▏ | 34687/48008 [1:59:00<43:12, 5.14it/s] 72%|███████▏ | 34688/48008 [1:59:00<42:57, 5.17it/s] 72%|███████▏ | 34689/48008 [1:59:00<42:34, 5.21it/s] 72%|███████▏ | 34690/48008 [1:59:00<42:33, 5.21it/s] 72%|███████▏ | 34691/48008 [1:59:01<42:27, 5.23it/s] 72%|███████▏ | 34692/48008 [1:59:01<42:21, 5.24it/s] 72%|███████▏ | 34693/48008 [1:59:01<43:37, 5.09it/s] 72%|███████▏ | 34694/48008 [1:59:01<43:50, 5.06it/s] 72%|███████▏ | 34695/48008 [1:59:01<43:25, 5.11it/s] 72%|███████▏ | 34696/48008 [1:59:02<43:10, 5.14it/s] 72%|███████▏ | 34697/48008 [1:59:02<45:53, 4.83it/s] 72%|███████▏ | 34698/48008 [1:59:02<44:30, 4.98it/s] 72%|███████▏ | 34699/48008 [1:59:02<43:36, 5.09it/s] 72%|███████▏ | 34700/48008 [1:59:02<43:17, 5.12it/s] {'loss': 0.3003, 'grad_norm': 0.416015625, 'learning_rate': 5.544492584569239e-05, 'epoch': 0.72} 72%|███████▏ | 34700/48008 [1:59:02<43:17, 5.12it/s] 72%|███████▏ | 34701/48008 [1:59:03<42:41, 5.19it/s] 72%|███████▏ | 34702/48008 [1:59:03<42:02, 5.28it/s] 72%|███████▏ | 34703/48008 [1:59:03<43:23, 5.11it/s] 72%|███████▏ | 34704/48008 [1:59:03<46:06, 4.81it/s] 72%|███████▏ | 34705/48008 [1:59:03<45:51, 4.83it/s] 72%|███████▏ | 34706/48008 [1:59:04<44:24, 4.99it/s] 72%|███████▏ | 34707/48008 [1:59:04<44:12, 5.01it/s] 72%|███████▏ | 34708/48008 [1:59:04<43:39, 5.08it/s] 72%|███████▏ | 34709/48008 [1:59:04<43:17, 5.12it/s] 72%|███████▏ | 34710/48008 [1:59:04<43:15, 5.12it/s] 72%|███████▏ | 34711/48008 [1:59:05<42:57, 5.16it/s] 72%|███████▏ | 34712/48008 [1:59:05<43:10, 5.13it/s] 72%|███████▏ | 34713/48008 [1:59:05<43:02, 5.15it/s] 72%|███████▏ | 34714/48008 [1:59:05<42:47, 5.18it/s] 72%|███████▏ | 34715/48008 [1:59:05<42:40, 5.19it/s] 72%|███████▏ | 34716/48008 [1:59:06<43:04, 5.14it/s] 72%|███████▏ | 34717/48008 [1:59:06<43:57, 5.04it/s] 72%|███████▏ | 34718/48008 [1:59:06<43:36, 5.08it/s] 72%|███████▏ | 34719/48008 [1:59:06<43:21, 5.11it/s] 72%|███████▏ | 34720/48008 [1:59:06<43:06, 5.14it/s] 72%|███████▏ | 34721/48008 [1:59:07<43:01, 5.15it/s] 72%|███████▏ | 34722/48008 [1:59:07<43:48, 5.05it/s] 72%|███████▏ | 34723/48008 [1:59:07<43:18, 5.11it/s] 72%|███████▏ | 34724/48008 [1:59:07<42:58, 5.15it/s] 72%|███████▏ | 34725/48008 [1:59:07<43:35, 5.08it/s] 72%|███████▏ | 34726/48008 [1:59:07<43:30, 5.09it/s] 72%|███████▏ | 34727/48008 [1:59:08<42:54, 5.16it/s] 72%|███████▏ | 34728/48008 [1:59:08<43:26, 5.10it/s] 72%|███████▏ | 34729/48008 [1:59:08<43:05, 5.14it/s] 72%|███████▏ | 34730/48008 [1:59:08<42:31, 5.20it/s] 72%|███████▏ | 34731/48008 [1:59:08<42:51, 5.16it/s] 72%|███████▏ | 34732/48008 [1:59:09<43:26, 5.09it/s] 72%|███████▏ | 34733/48008 [1:59:09<43:01, 5.14it/s] 72%|███████▏ | 34734/48008 [1:59:09<43:34, 5.08it/s] 72%|███████▏ | 34735/48008 [1:59:09<43:06, 5.13it/s] 72%|███████▏ | 34736/48008 [1:59:09<42:49, 5.16it/s] 72%|███████▏ | 34737/48008 [1:59:10<42:28, 5.21it/s] 72%|███████▏ | 34738/48008 [1:59:10<42:14, 5.24it/s] 72%|███████▏ | 34739/48008 [1:59:10<42:19, 5.23it/s] 72%|███████▏ | 34740/48008 [1:59:10<45:13, 4.89it/s] 72%|███████▏ | 34741/48008 [1:59:10<44:22, 4.98it/s] 72%|███████▏ | 34742/48008 [1:59:11<43:53, 5.04it/s] 72%|███████▏ | 34743/48008 [1:59:11<44:13, 5.00it/s] 72%|███████▏ | 34744/48008 [1:59:11<43:57, 5.03it/s] 72%|███████▏ | 34745/48008 [1:59:11<43:25, 5.09it/s] 72%|███████▏ | 34746/48008 [1:59:11<43:25, 5.09it/s] 72%|███████▏ | 34747/48008 [1:59:12<43:05, 5.13it/s] 72%|███████▏ | 34748/48008 [1:59:12<43:04, 5.13it/s] 72%|███████▏ | 34749/48008 [1:59:12<43:38, 5.06it/s] 72%|███████▏ | 34750/48008 [1:59:12<43:04, 5.13it/s] 72%|███████▏ | 34750/48008 [1:59:12<43:04, 5.13it/s]{'loss': 0.2981, 'grad_norm': 0.1923828125, 'learning_rate': 5.52366272287952e-05, 'epoch': 0.72} 72%|███████▏ | 34751/48008 [1:59:12<43:17, 5.10it/s] 72%|███████▏ | 34752/48008 [1:59:13<44:09, 5.00it/s] 72%|███████▏ | 34753/48008 [1:59:13<44:23, 4.98it/s] 72%|███████▏ | 34754/48008 [1:59:13<43:38, 5.06it/s] 72%|███████▏ | 34755/48008 [1:59:13<46:08, 4.79it/s] 72%|███████▏ | 34756/48008 [1:59:13<44:38, 4.95it/s] 72%|███████▏ | 34757/48008 [1:59:14<44:35, 4.95it/s] 72%|███████▏ | 34758/48008 [1:59:14<51:26, 4.29it/s] 72%|███████▏ | 34759/48008 [1:59:14<49:10, 4.49it/s] 72%|███████▏ | 34760/48008 [1:59:14<46:53, 4.71it/s] 72%|███████▏ | 34761/48008 [1:59:15<45:25, 4.86it/s] 72%|███████▏ | 34762/48008 [1:59:15<44:33, 4.95it/s] 72%|███████▏ | 34763/48008 [1:59:15<43:41, 5.05it/s] 72%|███████▏ | 34764/48008 [1:59:15<44:28, 4.96it/s] 72%|███████▏ | 34765/48008 [1:59:15<43:52, 5.03it/s] 72%|███████▏ | 34766/48008 [1:59:15<43:59, 5.02it/s] 72%|███████▏ | 34767/48008 [1:59:16<51:07, 4.32it/s] 72%|███████▏ | 34768/48008 [1:59:16<48:32, 4.55it/s] 72%|███████▏ | 34769/48008 [1:59:16<46:35, 4.74it/s] 72%|███████▏ | 34770/48008 [1:59:16<44:47, 4.93it/s] 72%|███████▏ | 34771/48008 [1:59:17<43:43, 5.04it/s] 72%|███████▏ | 34772/48008 [1:59:17<43:00, 5.13it/s] 72%|███████▏ | 34773/48008 [1:59:17<55:42, 3.96it/s] 72%|███████▏ | 34774/48008 [1:59:17<51:39, 4.27it/s] 72%|███████▏ | 34775/48008 [1:59:18<49:01, 4.50it/s] 72%|███████▏ | 34776/48008 [1:59:18<47:48, 4.61it/s] 72%|███████▏ | 34777/48008 [1:59:18<46:06, 4.78it/s] 72%|███████▏ | 34778/48008 [1:59:18<44:39, 4.94it/s] 72%|███████▏ | 34779/48008 [1:59:18<43:48, 5.03it/s] 72%|███████▏ | 34780/48008 [1:59:18<43:10, 5.11it/s] 72%|███████▏ | 34781/48008 [1:59:19<43:43, 5.04it/s] 72%|███████▏ | 34782/48008 [1:59:19<43:39, 5.05it/s] 72%|███████▏ | 34783/48008 [1:59:19<43:13, 5.10it/s] 72%|███████▏ | 34784/48008 [1:59:19<42:34, 5.18it/s] 72%|███████▏ | 34785/48008 [1:59:19<42:44, 5.16it/s] 72%|███████▏ | 34786/48008 [1:59:20<43:20, 5.09it/s] 72%|███████▏ | 34787/48008 [1:59:20<44:12, 4.98it/s] 72%|███████▏ | 34788/48008 [1:59:20<43:23, 5.08it/s] 72%|███████▏ | 34789/48008 [1:59:20<43:03, 5.12it/s] 72%|███████▏ | 34790/48008 [1:59:20<42:37, 5.17it/s] 72%|███████▏ | 34791/48008 [1:59:21<42:30, 5.18it/s] 72%|███████▏ | 34792/48008 [1:59:21<42:16, 5.21it/s] 72%|███████▏ | 34793/48008 [1:59:21<42:06, 5.23it/s] 72%|███████▏ | 34794/48008 [1:59:21<42:02, 5.24it/s] 72%|███████▏ | 34795/48008 [1:59:21<42:06, 5.23it/s] 72%|███████▏ | 34796/48008 [1:59:22<41:51, 5.26it/s] 72%|███████▏ | 34797/48008 [1:59:22<41:57, 5.25it/s] 72%|███████▏ | 34798/48008 [1:59:22<41:30, 5.31it/s] 72%|███████▏ | 34799/48008 [1:59:22<42:54, 5.13it/s] 72%|███████▏ | 34800/48008 [1:59:22<42:21, 5.20it/s] {'loss': 0.3011, 'grad_norm': 0.1474609375, 'learning_rate': 5.502832861189802e-05, 'epoch': 0.72} 72%|███████▏ | 34800/48008 [1:59:22<42:21, 5.20it/s] 72%|███████▏ | 34801/48008 [1:59:23<43:32, 5.06it/s] 72%|███████▏ | 34802/48008 [1:59:23<43:01, 5.11it/s] 72%|███████▏ | 34803/48008 [1:59:23<42:37, 5.16it/s] 72%|███████▏ | 34804/48008 [1:59:23<42:28, 5.18it/s] 72%|███████▏ | 34805/48008 [1:59:23<45:24, 4.85it/s] 73%|███████▎ | 34806/48008 [1:59:24<47:24, 4.64it/s] 73%|███████▎ | 34807/48008 [1:59:24<45:46, 4.81it/s] 73%|███████▎ | 34808/48008 [1:59:24<44:17, 4.97it/s] 73%|███████▎ | 34809/48008 [1:59:24<43:16, 5.08it/s] 73%|███████▎ | 34810/48008 [1:59:24<42:59, 5.12it/s] 73%|███████▎ | 34811/48008 [1:59:25<42:44, 5.15it/s] 73%|███████▎ | 34812/48008 [1:59:25<42:50, 5.13it/s] 73%|███████▎ | 34813/48008 [1:59:25<42:37, 5.16it/s] 73%|███████▎ | 34814/48008 [1:59:25<43:03, 5.11it/s] 73%|███████▎ | 34815/48008 [1:59:25<43:42, 5.03it/s] 73%|███████▎ | 34816/48008 [1:59:26<44:24, 4.95it/s] 73%|███████▎ | 34817/48008 [1:59:26<43:39, 5.04it/s] 73%|███████▎ | 34818/48008 [1:59:26<44:23, 4.95it/s] 73%|███████▎ | 34819/48008 [1:59:26<44:15, 4.97it/s] 73%|███████▎ | 34820/48008 [1:59:26<43:40, 5.03it/s] 73%|███████▎ | 34821/48008 [1:59:27<43:26, 5.06it/s] 73%|███████▎ | 34822/48008 [1:59:27<43:06, 5.10it/s] 73%|███████▎ | 34823/48008 [1:59:27<42:50, 5.13it/s] 73%|███████▎ | 34824/48008 [1:59:27<42:16, 5.20it/s] 73%|███████▎ | 34825/48008 [1:59:27<42:00, 5.23it/s] 73%|███████▎ | 34826/48008 [1:59:27<42:23, 5.18it/s] 73%|███████▎ | 34827/48008 [1:59:28<42:59, 5.11it/s] 73%|███████▎ | 34828/48008 [1:59:28<43:19, 5.07it/s] 73%|███████▎ | 34829/48008 [1:59:28<42:58, 5.11it/s] 73%|███████▎ | 34830/48008 [1:59:28<43:02, 5.10it/s] 73%|███████▎ | 34831/48008 [1:59:28<42:52, 5.12it/s] 73%|███████▎ | 34832/48008 [1:59:29<50:11, 4.37it/s] 73%|███████▎ | 34833/48008 [1:59:29<48:28, 4.53it/s] 73%|███████▎ | 34834/48008 [1:59:29<46:54, 4.68it/s] 73%|███████▎ | 34835/48008 [1:59:29<46:39, 4.71it/s] 73%|███████▎ | 34836/48008 [1:59:30<45:10, 4.86it/s] 73%|███████▎ | 34837/48008 [1:59:30<44:56, 4.88it/s] 73%|███████▎ | 34838/48008 [1:59:30<44:14, 4.96it/s] 73%|███████▎ | 34839/48008 [1:59:30<43:38, 5.03it/s] 73%|███████▎ | 34840/48008 [1:59:30<43:18, 5.07it/s] 73%|███████▎ | 34841/48008 [1:59:31<45:55, 4.78it/s] 73%|███████▎ | 34842/48008 [1:59:31<44:36, 4.92it/s] 73%|███████▎ | 34843/48008 [1:59:31<43:55, 4.99it/s] 73%|███████▎ | 34844/48008 [1:59:31<43:22, 5.06it/s] 73%|███████▎ | 34845/48008 [1:59:31<42:46, 5.13it/s] 73%|███████▎ | 34846/48008 [1:59:32<42:24, 5.17it/s] 73%|███████▎ | 34847/48008 [1:59:32<42:33, 5.15it/s] 73%|███████▎ | 34848/48008 [1:59:32<41:44, 5.25it/s] 73%|███████▎ | 34849/48008 [1:59:32<42:54, 5.11it/s] 73%|███████▎ | 34850/48008 [1:59:32<42:32, 5.16it/s] {'loss': 0.2994, 'grad_norm': 0.462890625, 'learning_rate': 5.4820029995000834e-05, 'epoch': 0.73} 73%|███████▎ | 34850/48008 [1:59:32<42:32, 5.16it/s] 73%|███████▎ | 34851/48008 [1:59:33<42:10, 5.20it/s] 73%|███████▎ | 34852/48008 [1:59:33<42:50, 5.12it/s] 73%|███████▎ | 34853/48008 [1:59:33<42:27, 5.16it/s] 73%|███████▎ | 34854/48008 [1:59:33<42:28, 5.16it/s] 73%|███████▎ | 34855/48008 [1:59:33<42:21, 5.18it/s] 73%|███████▎ | 34856/48008 [1:59:33<43:10, 5.08it/s] 73%|███████▎ | 34857/48008 [1:59:34<45:50, 4.78it/s] 73%|███████▎ | 34858/48008 [1:59:34<45:03, 4.86it/s] 73%|███████▎ | 34859/48008 [1:59:34<44:11, 4.96it/s] 73%|███████▎ | 34860/48008 [1:59:34<43:25, 5.05it/s] 73%|███████▎ | 34861/48008 [1:59:35<42:53, 5.11it/s] 73%|███████▎ | 34862/48008 [1:59:35<42:31, 5.15it/s] 73%|███████▎ | 34863/48008 [1:59:35<42:11, 5.19it/s] 73%|███████▎ | 34864/48008 [1:59:35<42:13, 5.19it/s] 73%|███████▎ | 34865/48008 [1:59:35<42:14, 5.19it/s] 73%|███████▎ | 34866/48008 [1:59:35<42:34, 5.15it/s] 73%|███████▎ | 34867/48008 [1:59:36<41:58, 5.22it/s] 73%|███████▎ | 34868/48008 [1:59:36<42:38, 5.14it/s] 73%|███████▎ | 34869/48008 [1:59:36<43:43, 5.01it/s] 73%|███████▎ | 34870/48008 [1:59:36<44:34, 4.91it/s] 73%|███████▎ | 34871/48008 [1:59:36<43:42, 5.01it/s] 73%|███████▎ | 34872/48008 [1:59:37<43:15, 5.06it/s] 73%|███████▎ | 34873/48008 [1:59:37<42:39, 5.13it/s] 73%|███████▎ | 34874/48008 [1:59:37<42:56, 5.10it/s] 73%|███████▎ | 34875/48008 [1:59:37<42:32, 5.14it/s] 73%|███████▎ | 34876/48008 [1:59:37<42:06, 5.20it/s] 73%|███████▎ | 34877/48008 [1:59:38<42:04, 5.20it/s] 73%|███████▎ | 34878/48008 [1:59:38<42:11, 5.19it/s] 73%|███████▎ | 34879/48008 [1:59:38<41:36, 5.26it/s] 73%|███████▎ | 34880/48008 [1:59:38<41:38, 5.25it/s] 73%|███████▎ | 34881/48008 [1:59:38<41:29, 5.27it/s] 73%|███████▎ | 34882/48008 [1:59:39<54:21, 4.02it/s] 73%|███████▎ | 34883/48008 [1:59:39<50:39, 4.32it/s] 73%|███████▎ | 34884/48008 [1:59:39<47:39, 4.59it/s] 73%|███████▎ | 34885/48008 [1:59:39<45:49, 4.77it/s] 73%|███████▎ | 34886/48008 [1:59:40<44:42, 4.89it/s] 73%|███████▎ | 34887/48008 [1:59:40<43:44, 5.00it/s] 73%|███████▎ | 34888/48008 [1:59:40<42:59, 5.09it/s] 73%|███████▎ | 34889/48008 [1:59:40<43:55, 4.98it/s] 73%|███████▎ | 34890/48008 [1:59:40<43:11, 5.06it/s] 73%|███████▎ | 34891/48008 [1:59:40<42:49, 5.10it/s] 73%|███████▎ | 34892/48008 [1:59:41<42:40, 5.12it/s] 73%|███████▎ | 34893/48008 [1:59:41<43:11, 5.06it/s] 73%|███████▎ | 34894/48008 [1:59:41<42:49, 5.10it/s] 73%|███████▎ | 34895/48008 [1:59:41<42:26, 5.15it/s] 73%|███████▎ | 34896/48008 [1:59:41<42:24, 5.15it/s] 73%|███████▎ | 34897/48008 [1:59:42<42:17, 5.17it/s] 73%|███████▎ | 34898/48008 [1:59:42<42:23, 5.15it/s] 73%|███████▎ | 34899/48008 [1:59:42<42:07, 5.19it/s] 73%|███████▎ | 34900/48008 [1:59:42<45:00, 4.85it/s] {'loss': 0.2996, 'grad_norm': 0.5, 'learning_rate': 5.4611731378103645e-05, 'epoch': 0.73} 73%|███████▎ | 34900/48008 [1:59:42<45:00, 4.85it/s] 73%|███████▎ | 34901/48008 [1:59:42<44:14, 4.94it/s] 73%|███████▎ | 34902/48008 [1:59:43<43:09, 5.06it/s] 73%|███████▎ | 34903/48008 [1:59:43<42:45, 5.11it/s] 73%|███████▎ | 34904/48008 [1:59:43<45:34, 4.79it/s] 73%|███████▎ | 34905/48008 [1:59:43<44:33, 4.90it/s] 73%|███████▎ | 34906/48008 [1:59:43<44:06, 4.95it/s] 73%|███████▎ | 34907/48008 [1:59:44<42:57, 5.08it/s] 73%|███████▎ | 34908/48008 [1:59:44<42:25, 5.15it/s] 73%|███████▎ | 34909/48008 [1:59:44<42:27, 5.14it/s] 73%|███████▎ | 34910/48008 [1:59:44<42:52, 5.09it/s] 73%|███████▎ | 34911/48008 [1:59:44<42:33, 5.13it/s] 73%|███████▎ | 34912/48008 [1:59:45<42:06, 5.18it/s] 73%|███████▎ | 34913/48008 [1:59:45<42:02, 5.19it/s] 73%|███████▎ | 34914/48008 [1:59:45<41:54, 5.21it/s] 73%|███████▎ | 34915/48008 [1:59:45<49:22, 4.42it/s] 73%|███████▎ | 34916/48008 [1:59:46<47:04, 4.63it/s] 73%|███████▎ | 34917/48008 [1:59:46<45:09, 4.83it/s] 73%|███████▎ | 34918/48008 [1:59:46<44:56, 4.85it/s] 73%|███████▎ | 34919/48008 [1:59:46<51:20, 4.25it/s] 73%|███████▎ | 34920/48008 [1:59:46<48:44, 4.48it/s] 73%|███████▎ | 34921/48008 [1:59:47<46:53, 4.65it/s] 73%|███████▎ | 34922/48008 [1:59:47<45:15, 4.82it/s] 73%|███████▎ | 34923/48008 [1:59:47<44:12, 4.93it/s] 73%|███████▎ | 34924/48008 [1:59:47<43:08, 5.05it/s] 73%|███████▎ | 34925/48008 [1:59:47<42:18, 5.15it/s] 73%|███████▎ | 34926/48008 [1:59:48<41:46, 5.22it/s] 73%|███████▎ | 34927/48008 [1:59:48<41:45, 5.22it/s] 73%|███████▎ | 34928/48008 [1:59:48<44:37, 4.88it/s] 73%|███████▎ | 34929/48008 [1:59:48<43:03, 5.06it/s] 73%|███████▎ | 34930/48008 [1:59:48<42:27, 5.13it/s] 73%|███████▎ | 34931/48008 [1:59:49<42:07, 5.17it/s] 73%|███████▎ | 34932/48008 [1:59:49<41:56, 5.20it/s] 73%|███████▎ | 34933/48008 [1:59:49<42:13, 5.16it/s] 73%|███████▎ | 34934/48008 [1:59:49<41:42, 5.23it/s] 73%|███████▎ | 34935/48008 [1:59:49<41:29, 5.25it/s] 73%|███████▎ | 34936/48008 [1:59:49<41:53, 5.20it/s] 73%|███████▎ | 34937/48008 [1:59:50<41:30, 5.25it/s] 73%|███████▎ | 34938/48008 [1:59:50<41:22, 5.26it/s] 73%|███████▎ | 34939/48008 [1:59:50<41:08, 5.29it/s] 73%|███████▎ | 34940/48008 [1:59:50<41:59, 5.19it/s] 73%|███████▎ | 34941/48008 [1:59:50<41:34, 5.24it/s] 73%|███████▎ | 34942/48008 [1:59:51<42:00, 5.18it/s] 73%|███████▎ | 34943/48008 [1:59:51<42:45, 5.09it/s] 73%|███████▎ | 34944/48008 [1:59:51<42:14, 5.15it/s] 73%|███████▎ | 34945/48008 [1:59:51<42:48, 5.09it/s] 73%|███████▎ | 34946/48008 [1:59:51<42:28, 5.12it/s] 73%|███████▎ | 34947/48008 [1:59:52<42:11, 5.16it/s] 73%|███████▎ | 34948/48008 [1:59:52<42:20, 5.14it/s] 73%|███████▎ | 34949/48008 [1:59:52<42:49, 5.08it/s] 73%|███████▎ | 34950/48008 [1:59:52<42:44, 5.09it/s] {'loss': 0.3005, 'grad_norm': 0.2109375, 'learning_rate': 5.440343276120646e-05, 'epoch': 0.73} 73%|███████▎ | 34950/48008 [1:59:52<42:44, 5.09it/s] 73%|███████▎ | 34951/48008 [1:59:52<42:36, 5.11it/s] 73%|███████▎ | 34952/48008 [1:59:53<42:18, 5.14it/s] 73%|███████▎ | 34953/48008 [1:59:53<42:31, 5.12it/s] 73%|███████▎ | 34954/48008 [1:59:53<41:51, 5.20it/s] 73%|███████▎ | 34955/48008 [1:59:53<41:18, 5.27it/s] 73%|███████▎ | 34956/48008 [1:59:53<41:07, 5.29it/s] 73%|███████▎ | 34957/48008 [1:59:54<41:10, 5.28it/s] 73%|███████▎ | 34958/48008 [1:59:54<41:39, 5.22it/s] 73%|███████▎ | 34959/48008 [1:59:54<54:21, 4.00it/s] 73%|███████▎ | 34960/48008 [1:59:54<50:43, 4.29it/s] 73%|███████▎ | 34961/48008 [1:59:54<47:45, 4.55it/s] 73%|███████▎ | 34962/48008 [1:59:55<45:29, 4.78it/s] 73%|███████▎ | 34963/48008 [1:59:55<43:50, 4.96it/s] 73%|███████▎ | 34964/48008 [1:59:55<43:08, 5.04it/s] 73%|███████▎ | 34965/48008 [1:59:55<43:48, 4.96it/s] 73%|███████▎ | 34966/48008 [1:59:55<42:58, 5.06it/s] 73%|███████▎ | 34967/48008 [1:59:56<42:32, 5.11it/s] 73%|███████▎ | 34968/48008 [1:59:56<42:04, 5.16it/s] 73%|███████▎ | 34969/48008 [1:59:56<42:16, 5.14it/s] 73%|███████▎ | 34970/48008 [1:59:56<42:06, 5.16it/s] 73%|███████▎ | 34971/48008 [1:59:57<49:28, 4.39it/s] 73%|███████▎ | 34972/48008 [1:59:57<47:00, 4.62it/s] 73%|███████▎ | 34973/48008 [1:59:57<45:19, 4.79it/s] 73%|███████▎ | 34974/48008 [1:59:57<44:08, 4.92it/s] 73%|███████▎ | 34975/48008 [1:59:57<43:22, 5.01it/s] 73%|███████▎ | 34976/48008 [1:59:57<42:23, 5.12it/s] 73%|███████▎ | 34977/48008 [1:59:58<41:50, 5.19it/s] 73%|███████▎ | 34978/48008 [1:59:58<41:41, 5.21it/s] 73%|███████▎ | 34979/48008 [1:59:58<44:38, 4.86it/s] 73%|███████▎ | 34980/48008 [1:59:58<43:33, 4.99it/s] 73%|███████▎ | 34981/48008 [1:59:58<42:56, 5.06it/s] 73%|███████▎ | 34982/48008 [1:59:59<42:02, 5.16it/s] 73%|███████▎ | 34983/48008 [1:59:59<41:10, 5.27it/s] 73%|███████▎ | 34984/48008 [1:59:59<41:41, 5.21it/s] 73%|███████▎ | 34985/48008 [1:59:59<41:40, 5.21it/s] 73%|███████▎ | 34986/48008 [1:59:59<44:41, 4.86it/s] 73%|███████▎ | 34987/48008 [2:00:00<46:39, 4.65it/s] 73%|███████▎ | 34988/48008 [2:00:00<45:01, 4.82it/s] 73%|███████▎ | 34989/48008 [2:00:00<43:44, 4.96it/s] 73%|███████▎ | 34990/48008 [2:00:00<44:18, 4.90it/s] 73%|███████▎ | 34991/48008 [2:00:00<43:21, 5.00it/s] 73%|███████▎ | 34992/48008 [2:00:01<42:45, 5.07it/s] 73%|███████▎ | 34993/48008 [2:00:01<42:25, 5.11it/s] 73%|███████▎ | 34994/48008 [2:00:01<42:01, 5.16it/s] 73%|███████▎ | 34995/48008 [2:00:01<41:41, 5.20it/s] 73%|███████▎ | 34996/48008 [2:00:01<42:36, 5.09it/s] 73%|███████▎ | 34997/48008 [2:00:02<49:44, 4.36it/s] 73%|███████▎ | 34998/48008 [2:00:02<47:11, 4.59it/s] 73%|███████▎ | 34999/48008 [2:00:02<48:34, 4.46it/s] 73%|███████▎ | 35000/48008 [2:00:02<47:12, 4.59it/s] 73%|███████▎ | 35000/48008 [2:00:02<47:12, 4.59it/s]{'loss': 0.3014, 'grad_norm': 0.177734375, 'learning_rate': 5.4195134144309286e-05, 'epoch': 0.73} 73%|███████▎ | 35001/48008 [2:00:03<45:30, 4.76it/s] 73%|███████▎ | 35002/48008 [2:00:03<44:12, 4.90it/s] 73%|███████▎ | 35003/48008 [2:00:03<43:13, 5.01it/s] 73%|███████▎ | 35004/48008 [2:00:03<42:39, 5.08it/s] 73%|███████▎ | 35005/48008 [2:00:03<42:46, 5.07it/s] 73%|███████▎ | 35006/48008 [2:00:04<43:02, 5.03it/s] 73%|███████▎ | 35007/48008 [2:00:04<42:29, 5.10it/s] 73%|███████▎ | 35008/48008 [2:00:04<42:22, 5.11it/s] 73%|███████▎ | 35009/48008 [2:00:04<42:13, 5.13it/s] 73%|███████▎ | 35010/48008 [2:00:04<42:07, 5.14it/s] 73%|███████▎ | 35011/48008 [2:00:05<41:52, 5.17it/s] 73%|███████▎ | 35012/48008 [2:00:05<41:27, 5.22it/s] 73%|███████▎ | 35013/48008 [2:00:05<41:19, 5.24it/s] 73%|███████▎ | 35014/48008 [2:00:05<41:17, 5.25it/s] 73%|███████▎ | 35015/48008 [2:00:05<41:16, 5.25it/s] 73%|███████▎ | 35016/48008 [2:00:05<41:03, 5.27it/s] 73%|███████▎ | 35017/48008 [2:00:06<41:08, 5.26it/s] 73%|███████▎ | 35018/48008 [2:00:06<41:37, 5.20it/s] 73%|███████▎ | 35019/48008 [2:00:06<41:38, 5.20it/s] 73%|███████▎ | 35020/48008 [2:00:06<42:12, 5.13it/s] 73%|███████▎ | 35021/48008 [2:00:06<42:41, 5.07it/s] 73%|███████▎ | 35022/48008 [2:00:07<43:14, 5.01it/s] 73%|███████▎ | 35023/48008 [2:00:07<43:04, 5.02it/s] 73%|███████▎ | 35024/48008 [2:00:07<42:59, 5.03it/s] 73%|███████▎ | 35025/48008 [2:00:07<42:26, 5.10it/s] 73%|███████▎ | 35026/48008 [2:00:07<42:47, 5.06it/s] 73%|███████▎ | 35027/48008 [2:00:08<42:14, 5.12it/s] 73%|███████▎ | 35028/48008 [2:00:08<42:17, 5.11it/s] 73%|███████▎ | 35029/48008 [2:00:08<41:50, 5.17it/s] 73%|███████▎ | 35030/48008 [2:00:08<41:30, 5.21it/s] 73%|███████▎ | 35031/48008 [2:00:08<41:33, 5.20it/s] 73%|███████▎ | 35032/48008 [2:00:09<41:08, 5.26it/s] 73%|███████▎ | 35033/48008 [2:00:09<41:12, 5.25it/s] 73%|███████▎ | 35034/48008 [2:00:09<40:55, 5.28it/s] 73%|███████▎ | 35035/48008 [2:00:09<40:53, 5.29it/s] 73%|███████▎ | 35036/48008 [2:00:09<41:50, 5.17it/s] 73%|███████▎ | 35037/48008 [2:00:10<41:33, 5.20it/s] 73%|███████▎ | 35038/48008 [2:00:10<41:23, 5.22it/s] 73%|███████▎ | 35039/48008 [2:00:10<41:19, 5.23it/s] 73%|███████▎ | 35040/48008 [2:00:10<40:53, 5.29it/s] 73%|███████▎ | 35041/48008 [2:00:10<40:55, 5.28it/s] 73%|███████▎ | 35042/48008 [2:00:10<40:42, 5.31it/s] 73%|███████▎ | 35043/48008 [2:00:11<40:44, 5.30it/s] 73%|███████▎ | 35044/48008 [2:00:11<40:50, 5.29it/s] 73%|███████▎ | 35045/48008 [2:00:11<41:23, 5.22it/s] 73%|███████▎ | 35046/48008 [2:00:11<41:14, 5.24it/s] 73%|███████▎ | 35047/48008 [2:00:11<41:05, 5.26it/s] 73%|███████▎ | 35048/48008 [2:00:12<41:47, 5.17it/s] 73%|███████▎ | 35049/48008 [2:00:12<41:15, 5.23it/s] 73%|███████▎ | 35050/48008 [2:00:12<42:00, 5.14it/s] {'loss': 0.3015, 'grad_norm': 0.212890625, 'learning_rate': 5.3986835527412104e-05, 'epoch': 0.73} 73%|███████▎ | 35050/48008 [2:00:12<42:00, 5.14it/s] 73%|███████▎ | 35051/48008 [2:00:12<41:32, 5.20it/s] 73%|███████▎ | 35052/48008 [2:00:12<41:34, 5.19it/s] 73%|███████▎ | 35053/48008 [2:00:13<41:07, 5.25it/s] 73%|███████▎ | 35054/48008 [2:00:13<41:21, 5.22it/s] 73%|███████▎ | 35055/48008 [2:00:13<40:52, 5.28it/s] 73%|███████▎ | 35056/48008 [2:00:13<41:04, 5.26it/s] 73%|███████▎ | 35057/48008 [2:00:13<41:02, 5.26it/s] 73%|███████▎ | 35058/48008 [2:00:14<41:35, 5.19it/s] 73%|███████▎ | 35059/48008 [2:00:14<41:17, 5.23it/s] 73%|███████▎ | 35060/48008 [2:00:14<41:09, 5.24it/s] 73%|███████▎ | 35061/48008 [2:00:14<41:53, 5.15it/s] 73%|███████▎ | 35062/48008 [2:00:14<42:01, 5.14it/s] 73%|███████▎ | 35063/48008 [2:00:15<42:04, 5.13it/s] 73%|███████▎ | 35064/48008 [2:00:15<42:09, 5.12it/s] 73%|███████▎ | 35065/48008 [2:00:15<42:07, 5.12it/s] 73%|███████▎ | 35066/48008 [2:00:15<41:51, 5.15it/s] 73%|███████▎ | 35067/48008 [2:00:15<41:39, 5.18it/s] 73%|███████▎ | 35068/48008 [2:00:15<41:15, 5.23it/s] 73%|███████▎ | 35069/48008 [2:00:16<42:22, 5.09it/s] 73%|███████▎ | 35070/48008 [2:00:16<42:02, 5.13it/s] 73%|███████▎ | 35071/48008 [2:00:16<41:49, 5.15it/s] 73%|███████▎ | 35072/48008 [2:00:16<42:07, 5.12it/s] 73%|███████▎ | 35073/48008 [2:00:16<41:56, 5.14it/s] 73%|███████▎ | 35074/48008 [2:00:17<41:49, 5.15it/s] 73%|███████▎ | 35075/48008 [2:00:17<42:03, 5.13it/s] 73%|███████▎ | 35076/48008 [2:00:17<41:51, 5.15it/s] 73%|███████▎ | 35077/48008 [2:00:17<41:33, 5.19it/s] 73%|███████▎ | 35078/48008 [2:00:17<41:24, 5.20it/s] 73%|███████▎ | 35079/48008 [2:00:18<41:24, 5.20it/s] 73%|███████▎ | 35080/48008 [2:00:18<41:26, 5.20it/s] 73%|███████▎ | 35081/48008 [2:00:18<41:19, 5.21it/s] 73%|███████▎ | 35082/48008 [2:00:18<41:15, 5.22it/s] 73%|███████▎ | 35083/48008 [2:00:19<53:53, 4.00it/s] 73%|███████▎ | 35084/48008 [2:00:19<57:29, 3.75it/s] 73%|███████▎ | 35085/48008 [2:00:19<52:16, 4.12it/s] 73%|███████▎ | 35086/48008 [2:00:19<49:20, 4.37it/s] 73%|███████▎ | 35087/48008 [2:00:19<46:52, 4.59it/s] 73%|███████▎ | 35088/48008 [2:00:20<44:53, 4.80it/s] 73%|███████▎ | 35089/48008 [2:00:20<43:30, 4.95it/s] 73%|███████▎ | 35090/48008 [2:00:20<42:50, 5.03it/s] 73%|███████▎ | 35091/48008 [2:00:20<42:01, 5.12it/s] 73%|███████▎ | 35092/48008 [2:00:20<41:27, 5.19it/s] 73%|███████▎ | 35093/48008 [2:00:21<41:45, 5.15it/s] 73%|███████▎ | 35094/48008 [2:00:21<41:15, 5.22it/s] 73%|███████▎ | 35095/48008 [2:00:21<42:23, 5.08it/s] 73%|███████▎ | 35096/48008 [2:00:21<42:03, 5.12it/s] 73%|███████▎ | 35097/48008 [2:00:21<49:20, 4.36it/s] 73%|███████▎ | 35098/48008 [2:00:22<59:36, 3.61it/s] 73%|███████▎ | 35099/48008 [2:00:22<54:08, 3.97it/s] 73%|███████▎ | 35100/48008 [2:00:22<51:02, 4.21it/s] {'loss': 0.3009, 'grad_norm': 0.42578125, 'learning_rate': 5.377853691051492e-05, 'epoch': 0.73} 73%|███████▎ | 35100/48008 [2:00:22<51:02, 4.21it/s] 73%|███████▎ | 35101/48008 [2:00:23<1:00:46, 3.54it/s] 73%|███████▎ | 35102/48008 [2:00:23<54:30, 3.95it/s] 73%|███████▎ | 35103/48008 [2:00:23<50:09, 4.29it/s] 73%|███████▎ | 35104/48008 [2:00:23<47:48, 4.50it/s] 73%|███████▎ | 35105/48008 [2:00:23<45:35, 4.72it/s] 73%|███████▎ | 35106/48008 [2:00:24<44:07, 4.87it/s] 73%|███████▎ | 35107/48008 [2:00:24<43:13, 4.97it/s] 73%|███████▎ | 35108/48008 [2:00:24<42:43, 5.03it/s] 73%|███████▎ | 35109/48008 [2:00:24<42:42, 5.03it/s] 73%|███████▎ | 35110/48008 [2:00:24<42:20, 5.08it/s] 73%|███████▎ | 35111/48008 [2:00:25<42:16, 5.08it/s] 73%|███████▎ | 35112/48008 [2:00:25<41:50, 5.14it/s] 73%|███████▎ | 35113/48008 [2:00:25<44:33, 4.82it/s] 73%|███████▎ | 35114/48008 [2:00:25<43:34, 4.93it/s] 73%|███████▎ | 35115/48008 [2:00:25<42:51, 5.01it/s] 73%|███████▎ | 35116/48008 [2:00:26<54:54, 3.91it/s] 73%|███████▎ | 35117/48008 [2:00:26<51:01, 4.21it/s] 73%|███████▎ | 35118/48008 [2:00:26<48:50, 4.40it/s] 73%|███████▎ | 35119/48008 [2:00:26<46:27, 4.62it/s] 73%|███████▎ | 35120/48008 [2:00:27<44:48, 4.79it/s] 73%|███████▎ | 35121/48008 [2:00:27<43:45, 4.91it/s] 73%|███████▎ | 35122/48008 [2:00:27<42:53, 5.01it/s] 73%|███████▎ | 35123/48008 [2:00:27<42:22, 5.07it/s] 73%|███████▎ | 35124/48008 [2:00:28<54:32, 3.94it/s] 73%|███████▎ | 35125/48008 [2:00:28<57:53, 3.71it/s] 73%|███████▎ | 35126/48008 [2:00:28<53:40, 4.00it/s] 73%|███████▎ | 35127/48008 [2:00:29<1:34:06, 2.28it/s] 73%|███████▎ | 35128/48008 [2:00:29<1:18:30, 2.73it/s] 73%|███████▎ | 35129/48008 [2:00:29<1:07:37, 3.17it/s] 73%|███████▎ | 35130/48008 [2:00:29<59:47, 3.59it/s] 73%|███████▎ | 35131/48008 [2:00:30<54:05, 3.97it/s] 73%|███████▎ | 35132/48008 [2:00:30<50:13, 4.27it/s] 73%|███████▎ | 35133/48008 [2:00:30<47:20, 4.53it/s] 73%|███████▎ | 35134/48008 [2:00:30<46:02, 4.66it/s] 73%|███████▎ | 35135/48008 [2:00:31<1:47:07, 2.00it/s] 73%|███████▎ | 35136/48008 [2:00:32<1:30:17, 2.38it/s] 73%|███████▎ | 35137/48008 [2:00:32<1:15:34, 2.84it/s] 73%|███████▎ | 35138/48008 [2:00:32<1:05:04, 3.30it/s] 73%|███████▎ | 35139/48008 [2:00:32<58:07, 3.69it/s] 73%|███████▎ | 35140/48008 [2:00:32<54:06, 3.96it/s] 73%|███████▎ | 35141/48008 [2:00:33<49:59, 4.29it/s] 73%|███████▎ | 35142/48008 [2:00:33<48:18, 4.44it/s] 73%|███████▎ | 35143/48008 [2:00:33<47:21, 4.53it/s] 73%|███████▎ | 35144/48008 [2:00:33<52:53, 4.05it/s] 73%|███████▎ | 35145/48008 [2:00:34<49:19, 4.35it/s] 73%|███████▎ | 35146/48008 [2:00:34<46:29, 4.61it/s] 73%|███████▎ | 35147/48008 [2:00:34<44:59, 4.76it/s] 73%|███████▎ | 35148/48008 [2:00:34<43:43, 4.90it/s] 73%|███████▎ | 35149/48008 [2:00:34<42:38, 5.03it/s] 73%|███████▎ | 35150/48008 [2:00:35<45:07, 4.75it/s] {'loss': 0.2978, 'grad_norm': 0.341796875, 'learning_rate': 5.357023829361774e-05, 'epoch': 0.73} 73%|███████▎ | 35150/48008 [2:00:35<45:07, 4.75it/s] 73%|███████▎ | 35151/48008 [2:00:35<43:53, 4.88it/s] 73%|███████▎ | 35152/48008 [2:00:35<42:54, 4.99it/s] 73%|███████▎ | 35153/48008 [2:00:35<42:17, 5.07it/s] 73%|███████▎ | 35154/48008 [2:00:35<42:42, 5.02it/s] 73%|███████▎ | 35155/48008 [2:00:36<41:50, 5.12it/s] 73%|███████▎ | 35156/48008 [2:00:36<41:29, 5.16it/s] 73%|███████▎ | 35157/48008 [2:00:36<40:36, 5.27it/s] 73%|███████▎ | 35158/48008 [2:00:36<41:04, 5.21it/s] 73%|███████▎ | 35159/48008 [2:00:36<40:52, 5.24it/s] 73%|███████▎ | 35160/48008 [2:00:36<41:40, 5.14it/s] 73%|███████▎ | 35161/48008 [2:00:37<41:48, 5.12it/s] 73%|███████▎ | 35162/48008 [2:00:37<41:13, 5.19it/s] 73%|███████▎ | 35163/48008 [2:00:37<48:25, 4.42it/s] 73%|███████▎ | 35164/48008 [2:00:37<47:15, 4.53it/s] 73%|███████▎ | 35165/48008 [2:00:38<45:38, 4.69it/s] 73%|███████▎ | 35166/48008 [2:00:38<43:54, 4.87it/s] 73%|███████▎ | 35167/48008 [2:00:38<43:21, 4.94it/s] 73%|███████▎ | 35168/48008 [2:00:38<45:37, 4.69it/s] 73%|███████▎ | 35169/48008 [2:00:38<43:57, 4.87it/s] 73%|███████▎ | 35170/48008 [2:00:39<43:07, 4.96it/s] 73%|███████▎ | 35171/48008 [2:00:39<42:49, 5.00it/s] 73%|███████▎ | 35172/48008 [2:00:39<42:26, 5.04it/s] 73%|███████▎ | 35173/48008 [2:00:39<41:45, 5.12it/s] 73%|███████▎ | 35174/48008 [2:00:39<41:32, 5.15it/s] 73%|███████▎ | 35175/48008 [2:00:40<41:19, 5.18it/s] 73%|███████▎ | 35176/48008 [2:00:40<42:22, 5.05it/s] 73%|███████▎ | 35177/48008 [2:00:40<41:58, 5.09it/s] 73%|███████▎ | 35178/48008 [2:00:40<42:18, 5.05it/s] 73%|███████▎ | 35179/48008 [2:00:40<41:34, 5.14it/s] 73%|███████▎ | 35180/48008 [2:00:40<41:25, 5.16it/s] 73%|███████▎ | 35181/48008 [2:00:41<41:32, 5.15it/s] 73%|███████▎ | 35182/48008 [2:00:41<41:18, 5.18it/s] 73%|███████▎ | 35183/48008 [2:00:41<40:51, 5.23it/s] 73%|███████▎ | 35184/48008 [2:00:41<41:11, 5.19it/s] 73%|███████▎ | 35185/48008 [2:00:41<40:52, 5.23it/s] 73%|███████▎ | 35186/48008 [2:00:42<40:49, 5.23it/s] 73%|███████▎ | 35187/48008 [2:00:42<41:13, 5.18it/s] 73%|███████▎ | 35188/48008 [2:00:42<44:04, 4.85it/s] 73%|███████▎ | 35189/48008 [2:00:42<43:15, 4.94it/s] 73%|███████▎ | 35190/48008 [2:00:42<42:33, 5.02it/s] 73%|███████▎ | 35191/48008 [2:00:43<42:17, 5.05it/s] 73%|███████▎ | 35192/48008 [2:00:43<42:24, 5.04it/s] 73%|███████▎ | 35193/48008 [2:00:43<42:04, 5.08it/s] 73%|███████▎ | 35194/48008 [2:00:43<41:30, 5.15it/s] 73%|███████▎ | 35195/48008 [2:00:43<40:53, 5.22it/s] 73%|███████▎ | 35196/48008 [2:00:44<43:51, 4.87it/s] 73%|███████▎ | 35197/48008 [2:00:44<42:41, 5.00it/s] 73%|███████▎ | 35198/48008 [2:00:44<42:33, 5.02it/s] 73%|███████▎ | 35199/48008 [2:00:44<42:07, 5.07it/s] 73%|███████▎ | 35200/48008 [2:00:44<41:43, 5.12it/s] {'loss': 0.3, 'grad_norm': 0.5390625, 'learning_rate': 5.336193967672055e-05, 'epoch': 0.73} 73%|███████▎ | 35200/48008 [2:00:44<41:43, 5.12it/s] 73%|███████▎ | 35201/48008 [2:00:45<41:48, 5.10it/s] 73%|███████▎ | 35202/48008 [2:00:45<41:34, 5.13it/s] 73%|███████▎ | 35203/48008 [2:00:45<42:03, 5.07it/s] 73%|███████▎ | 35204/48008 [2:00:45<41:42, 5.12it/s] 73%|███████▎ | 35205/48008 [2:00:45<41:21, 5.16it/s] 73%|███████▎ | 35206/48008 [2:00:46<53:39, 3.98it/s] 73%|███████▎ | 35207/48008 [2:00:46<49:43, 4.29it/s] 73%|███████▎ | 35208/48008 [2:00:46<48:09, 4.43it/s] 73%|███████▎ | 35209/48008 [2:00:46<46:16, 4.61it/s] 73%|███████▎ | 35210/48008 [2:00:47<44:21, 4.81it/s] 73%|███████▎ | 35211/48008 [2:00:47<43:43, 4.88it/s] 73%|███████▎ | 35212/48008 [2:00:47<44:00, 4.85it/s] 73%|███████▎ | 35213/48008 [2:00:47<42:45, 4.99it/s] 73%|███████▎ | 35214/48008 [2:00:47<42:13, 5.05it/s] 73%|███████▎ | 35215/48008 [2:00:48<42:02, 5.07it/s] 73%|███████▎ | 35216/48008 [2:00:48<41:40, 5.12it/s] 73%|███████▎ | 35217/48008 [2:00:48<41:18, 5.16it/s] 73%|███████▎ | 35218/48008 [2:00:48<41:25, 5.15it/s] 73%|███████▎ | 35219/48008 [2:00:48<48:31, 4.39it/s] 73%|███████▎ | 35220/48008 [2:00:49<46:46, 4.56it/s] 73%|███████▎ | 35221/48008 [2:00:49<45:00, 4.74it/s] 73%|███████▎ | 35222/48008 [2:00:49<43:45, 4.87it/s] 73%|███████▎ | 35223/48008 [2:00:49<43:33, 4.89it/s] 73%|███████▎ | 35224/48008 [2:00:49<42:24, 5.02it/s] 73%|███████▎ | 35225/48008 [2:00:50<41:42, 5.11it/s] 73%|███████▎ | 35226/48008 [2:00:50<41:15, 5.16it/s] 73%|███████▎ | 35227/48008 [2:00:50<41:14, 5.16it/s] 73%|███████▎ | 35228/48008 [2:00:50<41:07, 5.18it/s] 73%|███████▎ | 35229/48008 [2:00:50<41:01, 5.19it/s] 73%|███████▎ | 35230/48008 [2:00:51<43:46, 4.87it/s] 73%|███████▎ | 35231/48008 [2:00:51<42:41, 4.99it/s] 73%|███████▎ | 35232/48008 [2:00:51<42:10, 5.05it/s] 73%|███████▎ | 35233/48008 [2:00:51<41:30, 5.13it/s] 73%|███████▎ | 35234/48008 [2:00:51<41:29, 5.13it/s] 73%|███████▎ | 35235/48008 [2:00:52<41:16, 5.16it/s] 73%|███████▎ | 35236/48008 [2:00:52<42:12, 5.04it/s] 73%|███████▎ | 35237/48008 [2:00:52<41:29, 5.13it/s] 73%|███████▎ | 35238/48008 [2:00:52<40:59, 5.19it/s] 73%|███████▎ | 35239/48008 [2:00:53<53:24, 3.98it/s] 73%|███████▎ | 35240/48008 [2:00:53<49:17, 4.32it/s] 73%|███████▎ | 35241/48008 [2:00:53<49:48, 4.27it/s] 73%|███████▎ | 35242/48008 [2:00:53<50:00, 4.25it/s] 73%|███████▎ | 35243/48008 [2:00:53<47:03, 4.52it/s] 73%|███████▎ | 35244/48008 [2:00:54<46:13, 4.60it/s] 73%|███████▎ | 35245/48008 [2:00:54<44:31, 4.78it/s] 73%|███████▎ | 35246/48008 [2:00:54<43:10, 4.93it/s] 73%|███████▎ | 35247/48008 [2:00:54<45:24, 4.68it/s] 73%|███████▎ | 35248/48008 [2:00:54<44:13, 4.81it/s] 73%|███████▎ | 35249/48008 [2:00:55<50:23, 4.22it/s] 73%|███████▎ | 35250/48008 [2:00:55<47:12, 4.50it/s] 73%|███████▎ | 35250/48008 [2:00:55<47:12, 4.50it/s]{'loss': 0.2988, 'grad_norm': 0.44140625, 'learning_rate': 5.3153641059823366e-05, 'epoch': 0.73} 73%|███████▎ | 35251/48008 [2:00:55<44:40, 4.76it/s] 73%|███████▎ | 35252/48008 [2:00:55<43:31, 4.88it/s] 73%|███████▎ | 35253/48008 [2:00:55<42:36, 4.99it/s] 73%|███████▎ | 35254/48008 [2:00:56<41:57, 5.07it/s] 73%|███████▎ | 35255/48008 [2:00:56<41:35, 5.11it/s] 73%|███████▎ | 35256/48008 [2:00:56<41:11, 5.16it/s] 73%|███████▎ | 35257/48008 [2:00:56<53:26, 3.98it/s] 73%|███████▎ | 35258/48008 [2:00:57<49:31, 4.29it/s] 73%|███████▎ | 35259/48008 [2:00:57<46:30, 4.57it/s] 73%|███████▎ | 35260/48008 [2:00:57<44:54, 4.73it/s] 73%|███████▎ | 35261/48008 [2:00:57<43:21, 4.90it/s] 73%|███████▎ | 35262/48008 [2:00:57<42:45, 4.97it/s] 73%|███████▎ | 35263/48008 [2:00:58<42:54, 4.95it/s] 73%|███████▎ | 35264/48008 [2:00:58<45:04, 4.71it/s] 73%|███████▎ | 35265/48008 [2:00:58<43:40, 4.86it/s] 73%|███████▎ | 35266/48008 [2:00:58<43:43, 4.86it/s] 73%|███████▎ | 35267/48008 [2:00:58<42:52, 4.95it/s] 73%|███████▎ | 35268/48008 [2:00:59<42:06, 5.04it/s] 73%|███████▎ | 35269/48008 [2:00:59<41:32, 5.11it/s] 73%|███████▎ | 35270/48008 [2:00:59<41:16, 5.14it/s] 73%|███████▎ | 35271/48008 [2:00:59<40:43, 5.21it/s] 73%|███████▎ | 35272/48008 [2:00:59<40:41, 5.22it/s] 73%|███████▎ | 35273/48008 [2:01:00<40:38, 5.22it/s] 73%|███████▎ | 35274/48008 [2:01:00<40:32, 5.23it/s] 73%|███████▎ | 35275/48008 [2:01:00<40:29, 5.24it/s] 73%|███████▎ | 35276/48008 [2:01:00<40:24, 5.25it/s] 73%|███████▎ | 35277/48008 [2:01:00<40:23, 5.25it/s] 73%|███████▎ | 35278/48008 [2:01:00<40:31, 5.24it/s] 73%|███████▎ | 35279/48008 [2:01:01<40:53, 5.19it/s] 73%|███████▎ | 35280/48008 [2:01:01<40:09, 5.28it/s] 73%|███████▎ | 35281/48008 [2:01:01<40:20, 5.26it/s] 73%|███████▎ | 35282/48008 [2:01:01<40:18, 5.26it/s] 73%|███████▎ | 35283/48008 [2:01:01<40:27, 5.24it/s] 73%|███████▎ | 35284/48008 [2:01:02<40:15, 5.27it/s] 73%|███████▎ | 35285/48008 [2:01:02<40:04, 5.29it/s] 74%|███████▎ | 35286/48008 [2:01:02<40:08, 5.28it/s] 74%|███████▎ | 35287/48008 [2:01:02<40:33, 5.23it/s] 74%|███████▎ | 35288/48008 [2:01:02<40:20, 5.25it/s] 74%|███████▎ | 35289/48008 [2:01:03<40:24, 5.25it/s] 74%|███████▎ | 35290/48008 [2:01:03<40:19, 5.26it/s] 74%|███████▎ | 35291/48008 [2:01:03<40:14, 5.27it/s] 74%|███████▎ | 35292/48008 [2:01:03<40:22, 5.25it/s] 74%|███████▎ | 35293/48008 [2:01:03<40:22, 5.25it/s] 74%|███████▎ | 35294/48008 [2:01:04<40:23, 5.25it/s] 74%|███████▎ | 35295/48008 [2:01:04<47:47, 4.43it/s] 74%|███████▎ | 35296/48008 [2:01:04<46:21, 4.57it/s] 74%|███████▎ | 35297/48008 [2:01:04<44:36, 4.75it/s] 74%|███████▎ | 35298/48008 [2:01:04<43:05, 4.92it/s] 74%|███████▎ | 35299/48008 [2:01:05<42:21, 5.00it/s] 74%|███████▎ | 35300/48008 [2:01:05<41:41, 5.08it/s] {'loss': 0.3007, 'grad_norm': 0.2431640625, 'learning_rate': 5.2945342442926184e-05, 'epoch': 0.74} 74%|███████▎ | 35300/48008 [2:01:05<41:41, 5.08it/s] 74%|███████▎ | 35301/48008 [2:01:05<44:15, 4.78it/s] 74%|███████▎ | 35302/48008 [2:01:05<43:01, 4.92it/s] 74%|███████▎ | 35303/48008 [2:01:05<42:58, 4.93it/s] 74%|███████▎ | 35304/48008 [2:01:06<42:27, 4.99it/s] 74%|███████▎ | 35305/48008 [2:01:06<41:30, 5.10it/s] 74%|███████▎ | 35306/48008 [2:01:06<40:55, 5.17it/s] 74%|███████▎ | 35307/48008 [2:01:06<48:06, 4.40it/s] 74%|███████▎ | 35308/48008 [2:01:06<45:58, 4.60it/s] 74%|███████▎ | 35309/48008 [2:01:07<44:07, 4.80it/s] 74%|███████▎ | 35310/48008 [2:01:07<42:57, 4.93it/s] 74%|███████▎ | 35311/48008 [2:01:07<42:54, 4.93it/s] 74%|███████▎ | 35312/48008 [2:01:07<42:19, 5.00it/s] 74%|███████▎ | 35313/48008 [2:01:07<41:33, 5.09it/s] 74%|███████▎ | 35314/48008 [2:01:08<41:23, 5.11it/s] 74%|███████▎ | 35315/48008 [2:01:08<40:51, 5.18it/s] 74%|███████▎ | 35316/48008 [2:01:08<41:03, 5.15it/s] 74%|███████▎ | 35317/48008 [2:01:08<41:35, 5.09it/s] 74%|███████▎ | 35318/48008 [2:01:08<41:18, 5.12it/s] 74%|███████▎ | 35319/48008 [2:01:09<48:19, 4.38it/s] 74%|███████▎ | 35320/48008 [2:01:09<46:18, 4.57it/s] 74%|███████▎ | 35321/48008 [2:01:09<44:21, 4.77it/s] 74%|███████▎ | 35322/48008 [2:01:09<43:15, 4.89it/s] 74%|███████▎ | 35323/48008 [2:01:10<43:27, 4.86it/s] 74%|███████▎ | 35324/48008 [2:01:10<42:23, 4.99it/s] 74%|███████▎ | 35325/48008 [2:01:10<42:29, 4.97it/s] 74%|███████▎ | 35326/48008 [2:01:10<41:55, 5.04it/s] 74%|███████▎ | 35327/48008 [2:01:10<41:08, 5.14it/s] 74%|███████▎ | 35328/48008 [2:01:10<41:06, 5.14it/s] 74%|███████▎ | 35329/48008 [2:01:11<40:53, 5.17it/s] 74%|███████▎ | 35330/48008 [2:01:11<41:29, 5.09it/s] 74%|███████▎ | 35331/48008 [2:01:11<53:40, 3.94it/s] 74%|███████▎ | 35332/48008 [2:01:11<50:43, 4.17it/s] 74%|███████▎ | 35333/48008 [2:01:12<47:28, 4.45it/s] 74%|███████▎ | 35334/48008 [2:01:12<45:01, 4.69it/s] 74%|███████▎ | 35335/48008 [2:01:12<43:55, 4.81it/s] 74%|███████▎ | 35336/48008 [2:01:12<42:46, 4.94it/s] 74%|███████▎ | 35337/48008 [2:01:12<42:21, 4.99it/s] 74%|███████▎ | 35338/48008 [2:01:13<41:56, 5.03it/s] 74%|███████▎ | 35339/48008 [2:01:13<44:26, 4.75it/s] 74%|███████▎ | 35340/48008 [2:01:13<42:48, 4.93it/s] 74%|███████▎ | 35341/48008 [2:01:13<42:52, 4.92it/s] 74%|███████▎ | 35342/48008 [2:01:14<49:25, 4.27it/s] 74%|███████▎ | 35343/48008 [2:01:14<46:42, 4.52it/s] 74%|███████▎ | 35344/48008 [2:01:14<44:58, 4.69it/s] 74%|███████▎ | 35345/48008 [2:01:14<44:40, 4.72it/s] 74%|███████▎ | 35346/48008 [2:01:14<43:14, 4.88it/s] 74%|███████▎ | 35347/48008 [2:01:15<42:19, 4.98it/s] 74%|███████▎ | 35348/48008 [2:01:15<42:08, 5.01it/s] 74%|███████▎ | 35349/48008 [2:01:15<41:46, 5.05it/s] 74%|███████▎ | 35350/48008 [2:01:15<42:07, 5.01it/s] 74%|███████▎ | 35350/48008 [2:01:15<42:07, 5.01it/s]{'loss': 0.2991, 'grad_norm': 0.45703125, 'learning_rate': 5.2737043826029e-05, 'epoch': 0.74} 74%|███████▎ | 35351/48008 [2:01:15<42:49, 4.93it/s] 74%|███████▎ | 35352/48008 [2:01:16<41:55, 5.03it/s] 74%|███████▎ | 35353/48008 [2:01:16<41:48, 5.04it/s] 74%|███████▎ | 35354/48008 [2:01:16<41:15, 5.11it/s] 74%|███████▎ | 35355/48008 [2:01:16<40:54, 5.16it/s] 74%|███████▎ | 35356/48008 [2:01:16<41:20, 5.10it/s] 74%|███████▎ | 35357/48008 [2:01:16<40:57, 5.15it/s] 74%|███████▎ | 35358/48008 [2:01:17<40:41, 5.18it/s] 74%|███████▎ | 35359/48008 [2:01:17<40:38, 5.19it/s] 74%|███████▎ | 35360/48008 [2:01:17<40:12, 5.24it/s] 74%|███████▎ | 35361/48008 [2:01:17<39:55, 5.28it/s] 74%|███████▎ | 35362/48008 [2:01:17<40:23, 5.22it/s] 74%|███████▎ | 35363/48008 [2:01:18<40:17, 5.23it/s] 74%|███████▎ | 35364/48008 [2:01:18<40:25, 5.21it/s] 74%|███████▎ | 35365/48008 [2:01:18<40:47, 5.17it/s] 74%|███████▎ | 35366/48008 [2:01:18<41:04, 5.13it/s] 74%|███████▎ | 35367/48008 [2:01:18<40:48, 5.16it/s] 74%|███████▎ | 35368/48008 [2:01:19<41:05, 5.13it/s] 74%|███████▎ | 35369/48008 [2:01:19<40:45, 5.17it/s] 74%|███████▎ | 35370/48008 [2:01:19<40:43, 5.17it/s] 74%|███████▎ | 35371/48008 [2:01:19<41:32, 5.07it/s] 74%|███████▎ | 35372/48008 [2:01:19<41:20, 5.09it/s] 74%|███████▎ | 35373/48008 [2:01:20<40:45, 5.17it/s] 74%|███████▎ | 35374/48008 [2:01:20<41:50, 5.03it/s] 74%|███████▎ | 35375/48008 [2:01:20<41:46, 5.04it/s] 74%|███████▎ | 35376/48008 [2:01:20<40:38, 5.18it/s] 74%|███████▎ | 35377/48008 [2:01:20<40:47, 5.16it/s] 74%|███████▎ | 35378/48008 [2:01:21<41:30, 5.07it/s] 74%|███████▎ | 35379/48008 [2:01:21<41:22, 5.09it/s] 74%|███████▎ | 35380/48008 [2:01:21<43:53, 4.79it/s] 74%|███████▎ | 35381/48008 [2:01:21<42:36, 4.94it/s] 74%|███████▎ | 35382/48008 [2:01:21<41:38, 5.05it/s] 74%|███████▎ | 35383/48008 [2:01:22<41:59, 5.01it/s] 74%|███████▎ | 35384/48008 [2:01:22<41:32, 5.07it/s] 74%|███████▎ | 35385/48008 [2:01:22<53:39, 3.92it/s] 74%|███████▎ | 35386/48008 [2:01:22<49:24, 4.26it/s] 74%|███████▎ | 35387/48008 [2:01:23<46:42, 4.50it/s] 74%|███████▎ | 35388/48008 [2:01:23<44:51, 4.69it/s] 74%|███████▎ | 35389/48008 [2:01:23<43:06, 4.88it/s] 74%|███████▎ | 35390/48008 [2:01:23<42:57, 4.90it/s] 74%|███████▎ | 35391/48008 [2:01:23<42:02, 5.00it/s] 74%|███████▎ | 35392/48008 [2:01:24<41:36, 5.05it/s] 74%|███████▎ | 35393/48008 [2:01:24<41:17, 5.09it/s] 74%|███████▎ | 35394/48008 [2:01:24<43:55, 4.79it/s] 74%|███████▎ | 35395/48008 [2:01:24<43:11, 4.87it/s] 74%|███████▎ | 35396/48008 [2:01:24<42:23, 4.96it/s] 74%|███████▎ | 35397/48008 [2:01:25<41:46, 5.03it/s] 74%|███████▎ | 35398/48008 [2:01:25<41:11, 5.10it/s] 74%|███████▎ | 35399/48008 [2:01:25<40:37, 5.17it/s] 74%|███████▎ | 35400/48008 [2:01:25<41:19, 5.09it/s] {'loss': 0.2995, 'grad_norm': 0.279296875, 'learning_rate': 5.252874520913181e-05, 'epoch': 0.74} 74%|███████▎ | 35400/48008 [2:01:25<41:19, 5.09it/s] 74%|███████▎ | 35401/48008 [2:01:25<41:01, 5.12it/s] 74%|███████▎ | 35402/48008 [2:01:25<40:40, 5.17it/s] 74%|███████▎ | 35403/48008 [2:01:26<41:04, 5.11it/s] 74%|███████▎ | 35404/48008 [2:01:26<40:58, 5.13it/s] 74%|███████▎ | 35405/48008 [2:01:26<40:51, 5.14it/s] 74%|███████▍ | 35406/48008 [2:01:26<52:56, 3.97it/s] 74%|███████▍ | 35407/48008 [2:01:27<52:02, 4.04it/s] 74%|███████▍ | 35408/48008 [2:01:27<48:39, 4.32it/s] 74%|███████▍ | 35409/48008 [2:01:27<46:04, 4.56it/s] 74%|███████▍ | 35410/48008 [2:01:27<47:10, 4.45it/s] 74%|███████▍ | 35411/48008 [2:01:28<45:13, 4.64it/s] 74%|███████▍ | 35412/48008 [2:01:28<43:24, 4.84it/s] 74%|███████▍ | 35413/48008 [2:01:28<42:00, 5.00it/s] 74%|███████▍ | 35414/48008 [2:01:28<42:04, 4.99it/s] 74%|███████▍ | 35415/48008 [2:01:28<42:07, 4.98it/s] 74%|███████▍ | 35416/48008 [2:01:29<44:32, 4.71it/s] 74%|███████▍ | 35417/48008 [2:01:29<42:50, 4.90it/s] 74%|███████▍ | 35418/48008 [2:01:29<42:05, 4.99it/s] 74%|███████▍ | 35419/48008 [2:01:29<41:33, 5.05it/s] 74%|███████▍ | 35420/48008 [2:01:29<41:01, 5.11it/s] 74%|███████▍ | 35421/48008 [2:01:29<40:51, 5.14it/s] 74%|███████▍ | 35422/48008 [2:01:30<52:56, 3.96it/s] 74%|███████▍ | 35423/48008 [2:01:30<49:43, 4.22it/s] 74%|███████▍ | 35424/48008 [2:01:30<46:53, 4.47it/s] 74%|███████▍ | 35425/48008 [2:01:30<44:54, 4.67it/s] 74%|███████▍ | 35426/48008 [2:01:31<44:40, 4.69it/s] 74%|███████▍ | 35427/48008 [2:01:31<44:10, 4.75it/s] 74%|███████▍ | 35428/48008 [2:01:31<42:59, 4.88it/s] 74%|███████▍ | 35429/48008 [2:01:31<42:26, 4.94it/s] 74%|███████▍ | 35430/48008 [2:01:31<41:23, 5.07it/s] 74%|███████▍ | 35431/48008 [2:01:32<40:59, 5.11it/s] 74%|███████▍ | 35432/48008 [2:01:32<40:46, 5.14it/s] 74%|███████▍ | 35433/48008 [2:01:32<40:39, 5.15it/s] 74%|███████▍ | 35434/48008 [2:01:32<40:57, 5.12it/s] 74%|███████▍ | 35435/48008 [2:01:32<40:49, 5.13it/s] 74%|███████▍ | 35436/48008 [2:01:33<40:37, 5.16it/s] 74%|███████▍ | 35437/48008 [2:01:33<40:21, 5.19it/s] 74%|███████▍ | 35438/48008 [2:01:33<40:45, 5.14it/s] 74%|███████▍ | 35439/48008 [2:01:33<40:26, 5.18it/s] 74%|███████▍ | 35440/48008 [2:01:33<40:10, 5.21it/s] 74%|███████▍ | 35441/48008 [2:01:34<40:36, 5.16it/s] 74%|███████▍ | 35442/48008 [2:01:34<40:07, 5.22it/s] 74%|███████▍ | 35443/48008 [2:01:34<40:50, 5.13it/s] 74%|███████▍ | 35444/48008 [2:01:34<40:31, 5.17it/s] 74%|███████▍ | 35445/48008 [2:01:34<40:21, 5.19it/s] 74%|███████▍ | 35446/48008 [2:01:35<40:11, 5.21it/s] 74%|███████▍ | 35447/48008 [2:01:35<41:03, 5.10it/s] 74%|███████▍ | 35448/48008 [2:01:35<40:45, 5.14it/s] 74%|███████▍ | 35449/48008 [2:01:35<40:32, 5.16it/s] 74%|███████▍ | 35450/48008 [2:01:35<40:11, 5.21it/s] {'loss': 0.299, 'grad_norm': 0.357421875, 'learning_rate': 5.232044659223463e-05, 'epoch': 0.74} 74%|███████▍ | 35450/48008 [2:01:35<40:11, 5.21it/s] 74%|███████▍ | 35451/48008 [2:01:36<41:02, 5.10it/s] 74%|███████▍ | 35452/48008 [2:01:36<41:27, 5.05it/s] 74%|███████▍ | 35453/48008 [2:01:36<41:46, 5.01it/s] 74%|███████▍ | 35454/48008 [2:01:36<42:02, 4.98it/s] 74%|███████▍ | 35455/48008 [2:01:36<41:23, 5.05it/s] 74%|███████▍ | 35456/48008 [2:01:37<41:03, 5.10it/s] 74%|███████▍ | 35457/48008 [2:01:37<40:49, 5.12it/s] 74%|███████▍ | 35458/48008 [2:01:37<40:40, 5.14it/s] 74%|███████▍ | 35459/48008 [2:01:37<40:12, 5.20it/s] 74%|███████▍ | 35460/48008 [2:01:37<41:21, 5.06it/s] 74%|███████▍ | 35461/48008 [2:01:37<41:13, 5.07it/s] 74%|███████▍ | 35462/48008 [2:01:38<40:26, 5.17it/s] 74%|███████▍ | 35463/48008 [2:01:38<39:47, 5.25it/s] 74%|███████▍ | 35464/48008 [2:01:38<40:01, 5.22it/s] 74%|███████▍ | 35465/48008 [2:01:38<39:51, 5.25it/s] 74%|███████▍ | 35466/48008 [2:01:38<39:41, 5.27it/s] 74%|███████▍ | 35467/48008 [2:01:39<39:54, 5.24it/s] 74%|███████▍ | 35468/48008 [2:01:39<40:12, 5.20it/s] 74%|███████▍ | 35469/48008 [2:01:39<40:49, 5.12it/s] 74%|███████▍ | 35470/48008 [2:01:39<40:34, 5.15it/s] 74%|███████▍ | 35471/48008 [2:01:39<40:12, 5.20it/s] 74%|███████▍ | 35472/48008 [2:01:40<43:07, 4.84it/s] 74%|███████▍ | 35473/48008 [2:01:40<42:09, 4.96it/s] 74%|███████▍ | 35474/48008 [2:01:40<41:36, 5.02it/s] 74%|███████▍ | 35475/48008 [2:01:40<41:06, 5.08it/s] 74%|███████▍ | 35476/48008 [2:01:41<48:17, 4.33it/s] 74%|███████▍ | 35477/48008 [2:01:41<45:40, 4.57it/s] 74%|███████▍ | 35478/48008 [2:01:41<44:54, 4.65it/s] 74%|███████▍ | 35479/48008 [2:01:41<43:24, 4.81it/s] 74%|███████▍ | 35480/48008 [2:01:41<42:06, 4.96it/s] 74%|███████▍ | 35481/48008 [2:01:41<41:33, 5.02it/s] 74%|███████▍ | 35482/48008 [2:01:42<40:51, 5.11it/s] 74%|███████▍ | 35483/48008 [2:01:42<40:28, 5.16it/s] 74%|███████▍ | 35484/48008 [2:01:42<40:16, 5.18it/s] 74%|███████▍ | 35485/48008 [2:01:42<40:30, 5.15it/s] 74%|███████▍ | 35486/48008 [2:01:42<40:40, 5.13it/s] 74%|███████▍ | 35487/48008 [2:01:43<40:20, 5.17it/s] 74%|███████▍ | 35488/48008 [2:01:43<41:13, 5.06it/s] 74%|███████▍ | 35489/48008 [2:01:43<40:55, 5.10it/s] 74%|███████▍ | 35490/48008 [2:01:43<40:46, 5.12it/s] 74%|███████▍ | 35491/48008 [2:01:43<40:36, 5.14it/s] 74%|███████▍ | 35492/48008 [2:01:44<41:42, 5.00it/s] 74%|███████▍ | 35493/48008 [2:01:44<41:04, 5.08it/s] 74%|███████▍ | 35494/48008 [2:01:44<40:36, 5.14it/s] 74%|███████▍ | 35495/48008 [2:01:44<39:59, 5.21it/s] 74%|███████▍ | 35496/48008 [2:01:44<40:00, 5.21it/s] 74%|███████▍ | 35497/48008 [2:01:45<39:50, 5.23it/s] 74%|███████▍ | 35498/48008 [2:01:45<39:55, 5.22it/s] 74%|███████▍ | 35499/48008 [2:01:45<40:06, 5.20it/s] 74%|███████▍ | 35500/48008 [2:01:45<40:22, 5.16it/s] {'loss': 0.2996, 'grad_norm': 0.142578125, 'learning_rate': 5.2112147975337446e-05, 'epoch': 0.74} 74%|███████▍ | 35500/48008 [2:01:45<40:22, 5.16it/s] 74%|███████▍ | 35501/48008 [2:01:45<40:01, 5.21it/s] 74%|███████▍ | 35502/48008 [2:01:46<39:50, 5.23it/s] 74%|███████▍ | 35503/48008 [2:01:46<39:11, 5.32it/s] 74%|███████▍ | 35504/48008 [2:01:46<39:26, 5.28it/s] 74%|███████▍ | 35505/48008 [2:01:46<39:46, 5.24it/s] 74%|███████▍ | 35506/48008 [2:01:46<40:33, 5.14it/s] 74%|███████▍ | 35507/48008 [2:01:47<40:11, 5.18it/s] 74%|███████▍ | 35508/48008 [2:01:47<39:48, 5.23it/s] 74%|███████▍ | 35509/48008 [2:01:47<39:51, 5.23it/s] 74%|███████▍ | 35510/48008 [2:01:47<39:47, 5.23it/s] 74%|███████▍ | 35511/48008 [2:01:47<39:32, 5.27it/s] 74%|███████▍ | 35512/48008 [2:01:47<40:13, 5.18it/s] 74%|███████▍ | 35513/48008 [2:01:48<39:54, 5.22it/s] 74%|███████▍ | 35514/48008 [2:01:48<39:41, 5.25it/s] 74%|███████▍ | 35515/48008 [2:01:48<40:02, 5.20it/s] 74%|███████▍ | 35516/48008 [2:01:48<39:59, 5.21it/s] 74%|███████▍ | 35517/48008 [2:01:48<39:57, 5.21it/s] 74%|███████▍ | 35518/48008 [2:01:49<39:49, 5.23it/s] 74%|███████▍ | 35519/48008 [2:01:49<39:45, 5.24it/s] 74%|███████▍ | 35520/48008 [2:01:49<39:41, 5.24it/s] 74%|███████▍ | 35521/48008 [2:01:49<47:01, 4.43it/s] 74%|███████▍ | 35522/48008 [2:01:49<44:59, 4.62it/s] 74%|███████▍ | 35523/48008 [2:01:50<44:10, 4.71it/s] 74%|███████▍ | 35524/48008 [2:01:50<43:34, 4.78it/s] 74%|███████▍ | 35525/48008 [2:01:50<42:17, 4.92it/s] 74%|███████▍ | 35526/48008 [2:01:50<42:09, 4.93it/s] 74%|███████▍ | 35527/48008 [2:01:50<41:47, 4.98it/s] 74%|███████▍ | 35528/48008 [2:01:51<41:12, 5.05it/s] 74%|███████▍ | 35529/48008 [2:01:51<40:44, 5.11it/s] 74%|███████▍ | 35530/48008 [2:01:51<40:28, 5.14it/s] 74%|███████▍ | 35531/48008 [2:01:51<40:11, 5.17it/s] 74%|███████▍ | 35532/48008 [2:01:51<40:12, 5.17it/s] 74%|███████▍ | 35533/48008 [2:01:52<41:05, 5.06it/s] 74%|███████▍ | 35534/48008 [2:01:52<40:46, 5.10it/s] 74%|███████▍ | 35535/48008 [2:01:52<40:20, 5.15it/s] 74%|███████▍ | 35536/48008 [2:01:52<39:56, 5.20it/s] 74%|███████▍ | 35537/48008 [2:01:52<40:58, 5.07it/s] 74%|███████▍ | 35538/48008 [2:01:53<52:47, 3.94it/s] 74%|███████▍ | 35539/48008 [2:01:53<48:55, 4.25it/s] 74%|███████▍ | 35540/48008 [2:01:53<45:55, 4.53it/s] 74%|███████▍ | 35541/48008 [2:01:53<44:08, 4.71it/s] 74%|███████▍ | 35542/48008 [2:01:54<42:34, 4.88it/s] 74%|███████▍ | 35543/48008 [2:01:54<41:47, 4.97it/s] 74%|███████▍ | 35544/48008 [2:01:54<41:51, 4.96it/s] 74%|███████▍ | 35545/48008 [2:01:54<40:30, 5.13it/s] 74%|███████▍ | 35546/48008 [2:01:54<41:26, 5.01it/s] 74%|███████▍ | 35547/48008 [2:01:55<43:55, 4.73it/s] 74%|███████▍ | 35548/48008 [2:01:55<43:47, 4.74it/s] 74%|███████▍ | 35549/48008 [2:01:55<42:27, 4.89it/s] 74%|███████▍ | 35550/48008 [2:01:55<53:47, 3.86it/s] {'loss': 0.3014, 'grad_norm': 0.69140625, 'learning_rate': 5.190384935844026e-05, 'epoch': 0.74} 74%|███████▍ | 35550/48008 [2:01:55<53:47, 3.86it/s] 74%|███████▍ | 35551/48008 [2:01:56<49:29, 4.20it/s] 74%|███████▍ | 35552/48008 [2:01:56<46:37, 4.45it/s] 74%|███████▍ | 35553/48008 [2:01:56<44:28, 4.67it/s] 74%|███████▍ | 35554/48008 [2:01:56<45:54, 4.52it/s] 74%|███████▍ | 35555/48008 [2:01:56<44:01, 4.71it/s] 74%|███████▍ | 35556/48008 [2:01:57<42:42, 4.86it/s] 74%|███████▍ | 35557/48008 [2:01:57<42:03, 4.93it/s] 74%|███████▍ | 35558/48008 [2:01:57<41:11, 5.04it/s] 74%|███████▍ | 35559/48008 [2:01:57<40:41, 5.10it/s] 74%|███████▍ | 35560/48008 [2:01:57<40:24, 5.14it/s] 74%|███████▍ | 35561/48008 [2:01:58<39:47, 5.21it/s] 74%|███████▍ | 35562/48008 [2:01:58<40:50, 5.08it/s] 74%|███████▍ | 35563/48008 [2:01:58<41:01, 5.06it/s] 74%|███████▍ | 35564/48008 [2:01:58<40:24, 5.13it/s] 74%|███████▍ | 35565/48008 [2:01:58<39:56, 5.19it/s] 74%|███████▍ | 35566/48008 [2:01:58<39:43, 5.22it/s] 74%|███████▍ | 35567/48008 [2:01:59<39:26, 5.26it/s] 74%|███████▍ | 35568/48008 [2:01:59<39:35, 5.24it/s] 74%|███████▍ | 35569/48008 [2:01:59<39:26, 5.26it/s] 74%|███████▍ | 35570/48008 [2:01:59<39:26, 5.26it/s] 74%|███████▍ | 35571/48008 [2:01:59<39:20, 5.27it/s] 74%|███████▍ | 35572/48008 [2:02:00<46:39, 4.44it/s] 74%|███████▍ | 35573/48008 [2:02:00<45:00, 4.60it/s] 74%|███████▍ | 35574/48008 [2:02:00<43:03, 4.81it/s] 74%|███████▍ | 35575/48008 [2:02:00<41:47, 4.96it/s] 74%|███████▍ | 35576/48008 [2:02:01<40:58, 5.06it/s] 74%|███████▍ | 35577/48008 [2:02:01<40:21, 5.13it/s] 74%|███████▍ | 35578/48008 [2:02:01<40:12, 5.15it/s] 74%|███████▍ | 35579/48008 [2:02:01<42:55, 4.83it/s] 74%|███████▍ | 35580/48008 [2:02:01<41:32, 4.99it/s] 74%|███████▍ | 35581/48008 [2:02:02<41:44, 4.96it/s] 74%|███████▍ | 35582/48008 [2:02:02<41:44, 4.96it/s] 74%|███████▍ | 35583/48008 [2:02:02<41:51, 4.95it/s] 74%|███████▍ | 35584/48008 [2:02:02<41:27, 5.00it/s] 74%|███████▍ | 35585/48008 [2:02:02<40:56, 5.06it/s] 74%|███████▍ | 35586/48008 [2:02:03<47:35, 4.35it/s] 74%|███████▍ | 35587/48008 [2:02:03<45:02, 4.60it/s] 74%|███████▍ | 35588/48008 [2:02:03<43:40, 4.74it/s] 74%|███████▍ | 35589/48008 [2:02:03<42:15, 4.90it/s] 74%|███████▍ | 35590/48008 [2:02:04<53:34, 3.86it/s] 74%|███████▍ | 35591/48008 [2:02:04<49:15, 4.20it/s] 74%|███████▍ | 35592/48008 [2:02:04<46:37, 4.44it/s] 74%|███████▍ | 35593/48008 [2:02:04<44:15, 4.68it/s] 74%|███████▍ | 35594/48008 [2:02:04<42:36, 4.86it/s] 74%|███████▍ | 35595/48008 [2:02:05<41:33, 4.98it/s] 74%|███████▍ | 35596/48008 [2:02:05<40:58, 5.05it/s] 74%|███████▍ | 35597/48008 [2:02:05<40:18, 5.13it/s] 74%|███████▍ | 35598/48008 [2:02:05<40:04, 5.16it/s] 74%|███████▍ | 35599/48008 [2:02:05<39:56, 5.18it/s] 74%|███████▍ | 35600/48008 [2:02:05<39:49, 5.19it/s] {'loss': 0.3018, 'grad_norm': 0.1513671875, 'learning_rate': 5.1695550741543074e-05, 'epoch': 0.74} 74%|███████▍ | 35600/48008 [2:02:05<39:49, 5.19it/s] 74%|███████▍ | 35601/48008 [2:02:06<39:30, 5.23it/s] 74%|███████▍ | 35602/48008 [2:02:06<42:32, 4.86it/s] 74%|███████▍ | 35603/48008 [2:02:06<41:25, 4.99it/s] 74%|███████▍ | 35604/48008 [2:02:06<41:33, 4.98it/s] 74%|███████▍ | 35605/48008 [2:02:06<41:25, 4.99it/s] 74%|███████▍ | 35606/48008 [2:02:07<41:29, 4.98it/s] 74%|███████▍ | 35607/48008 [2:02:07<40:52, 5.06it/s] 74%|███████▍ | 35608/48008 [2:02:07<40:31, 5.10it/s] 74%|███████▍ | 35609/48008 [2:02:07<40:17, 5.13it/s] 74%|███████▍ | 35610/48008 [2:02:07<40:02, 5.16it/s] 74%|███████▍ | 35611/48008 [2:02:08<39:59, 5.17it/s] 74%|███████▍ | 35612/48008 [2:02:08<39:49, 5.19it/s] 74%|███████▍ | 35613/48008 [2:02:08<39:35, 5.22it/s] 74%|███████▍ | 35614/48008 [2:02:08<39:28, 5.23it/s] 74%|███████▍ | 35615/48008 [2:02:08<39:55, 5.17it/s] 74%|███████▍ | 35616/48008 [2:02:09<39:47, 5.19it/s] 74%|███████▍ | 35617/48008 [2:02:09<46:47, 4.41it/s] 74%|███████▍ | 35618/48008 [2:02:09<44:32, 4.64it/s] 74%|███████▍ | 35619/48008 [2:02:09<43:03, 4.80it/s] 74%|███████▍ | 35620/48008 [2:02:10<42:32, 4.85it/s] 74%|███████▍ | 35621/48008 [2:02:10<42:24, 4.87it/s] 74%|███████▍ | 35622/48008 [2:02:10<41:21, 4.99it/s] 74%|███████▍ | 35623/48008 [2:02:10<40:44, 5.07it/s] 74%|███████▍ | 35624/48008 [2:02:10<40:36, 5.08it/s] 74%|███████▍ | 35625/48008 [2:02:10<39:58, 5.16it/s] 74%|███████▍ | 35626/48008 [2:02:11<39:47, 5.19it/s] 74%|███████▍ | 35627/48008 [2:02:11<39:08, 5.27it/s] 74%|███████▍ | 35628/48008 [2:02:11<39:05, 5.28it/s] 74%|███████▍ | 35629/48008 [2:02:11<39:05, 5.28it/s] 74%|███████▍ | 35630/48008 [2:02:11<39:33, 5.22it/s] 74%|███████▍ | 35631/48008 [2:02:12<39:40, 5.20it/s] 74%|███████▍ | 35632/48008 [2:02:12<39:35, 5.21it/s] 74%|███████▍ | 35633/48008 [2:02:12<39:51, 5.17it/s] 74%|███████▍ | 35634/48008 [2:02:12<42:40, 4.83it/s] 74%|███████▍ | 35635/48008 [2:02:12<41:41, 4.95it/s] 74%|███████▍ | 35636/48008 [2:02:13<41:26, 4.98it/s] 74%|███████▍ | 35637/48008 [2:02:13<40:54, 5.04it/s] 74%|███████▍ | 35638/48008 [2:02:13<40:30, 5.09it/s] 74%|███████▍ | 35639/48008 [2:02:13<40:12, 5.13it/s] 74%|███████▍ | 35640/48008 [2:02:13<40:01, 5.15it/s] 74%|███████▍ | 35641/48008 [2:02:14<40:06, 5.14it/s] 74%|███████▍ | 35642/48008 [2:02:14<39:54, 5.16it/s] 74%|███████▍ | 35643/48008 [2:02:14<39:52, 5.17it/s] 74%|███████▍ | 35644/48008 [2:02:14<39:45, 5.18it/s] 74%|███████▍ | 35645/48008 [2:02:14<39:46, 5.18it/s] 74%|███████▍ | 35646/48008 [2:02:15<46:45, 4.41it/s] 74%|███████▍ | 35647/48008 [2:02:15<44:42, 4.61it/s] 74%|███████▍ | 35648/48008 [2:02:15<45:57, 4.48it/s] 74%|███████▍ | 35649/48008 [2:02:15<44:01, 4.68it/s] 74%|███████▍ | 35650/48008 [2:02:15<42:35, 4.84it/s] {'loss': 0.299, 'grad_norm': 0.248046875, 'learning_rate': 5.148725212464589e-05, 'epoch': 0.74} 74%|███████▍ | 35650/48008 [2:02:15<42:35, 4.84it/s] 74%|███████▍ | 35651/48008 [2:02:16<41:45, 4.93it/s] 74%|███████▍ | 35652/48008 [2:02:16<42:10, 4.88it/s] 74%|███████▍ | 35653/48008 [2:02:16<41:17, 4.99it/s] 74%|███████▍ | 35654/48008 [2:02:16<41:20, 4.98it/s] 74%|███████▍ | 35655/48008 [2:02:16<40:38, 5.06it/s] 74%|███████▍ | 35656/48008 [2:02:17<41:19, 4.98it/s] 74%|███████▍ | 35657/48008 [2:02:17<40:47, 5.05it/s] 74%|███████▍ | 35658/48008 [2:02:18<1:23:09, 2.48it/s] 74%|███████▍ | 35659/48008 [2:02:18<1:09:55, 2.94it/s] 74%|███████▍ | 35660/48008 [2:02:18<1:00:37, 3.39it/s] 74%|███████▍ | 35661/48008 [2:02:18<54:39, 3.77it/s] 74%|███████▍ | 35662/48008 [2:02:19<49:46, 4.13it/s] 74%|███████▍ | 35663/48008 [2:02:19<46:36, 4.41it/s] 74%|███████▍ | 35664/48008 [2:02:20<1:22:53, 2.48it/s] 74%|███████▍ | 35665/48008 [2:02:20<1:09:43, 2.95it/s] 74%|███████▍ | 35666/48008 [2:02:20<1:00:58, 3.37it/s] 74%|███████▍ | 35667/48008 [2:02:20<57:19, 3.59it/s] 74%|███████▍ | 35668/48008 [2:02:20<51:54, 3.96it/s] 74%|███████▍ | 35669/48008 [2:02:21<48:12, 4.27it/s] 74%|███████▍ | 35670/48008 [2:02:22<1:34:29, 2.18it/s] 74%|███████▍ | 35671/48008 [2:02:22<1:17:41, 2.65it/s] 74%|███████▍ | 35672/48008 [2:02:22<1:05:44, 3.13it/s] 74%|███████▍ | 35673/48008 [2:02:22<57:41, 3.56it/s] 74%|███████▍ | 35674/48008 [2:02:22<52:17, 3.93it/s] 74%|███████▍ | 35675/48008 [2:02:22<48:46, 4.21it/s] 74%|███████▍ | 35676/48008 [2:02:23<1:23:29, 2.46it/s] 74%|███████▍ | 35677/48008 [2:02:23<1:10:05, 2.93it/s] 74%|███████▍ | 35678/48008 [2:02:24<1:00:42, 3.39it/s] 74%|███████▍ | 35679/48008 [2:02:24<55:04, 3.73it/s] 74%|███████▍ | 35680/48008 [2:02:24<51:00, 4.03it/s] 74%|███████▍ | 35681/48008 [2:02:24<47:09, 4.36it/s] 74%|███████▍ | 35682/48008 [2:02:24<45:10, 4.55it/s] 74%|███████▍ | 35683/48008 [2:02:25<43:02, 4.77it/s] 74%|███████▍ | 35684/48008 [2:02:25<43:05, 4.77it/s] 74%|███████▍ | 35685/48008 [2:02:25<41:51, 4.91it/s] 74%|███████▍ | 35686/48008 [2:02:25<41:23, 4.96it/s] 74%|███████▍ | 35687/48008 [2:02:25<40:31, 5.07it/s] 74%|███████▍ | 35688/48008 [2:02:26<40:07, 5.12it/s] 74%|███████▍ | 35689/48008 [2:02:26<39:54, 5.14it/s] 74%|███████▍ | 35690/48008 [2:02:26<39:30, 5.20it/s] 74%|███████▍ | 35691/48008 [2:02:26<40:06, 5.12it/s] 74%|███████▍ | 35692/48008 [2:02:26<39:33, 5.19it/s] 74%|███████▍ | 35693/48008 [2:02:27<39:54, 5.14it/s] 74%|███████▍ | 35694/48008 [2:02:27<40:19, 5.09it/s] 74%|███████▍ | 35695/48008 [2:02:27<39:46, 5.16it/s] 74%|███████▍ | 35696/48008 [2:02:27<46:46, 4.39it/s] 74%|███████▍ | 35697/48008 [2:02:27<45:16, 4.53it/s] 74%|███████▍ | 35698/48008 [2:02:28<44:03, 4.66it/s] 74%|███████▍ | 35699/48008 [2:02:28<42:25, 4.84it/s] 74%|███████▍ | 35700/48008 [2:02:28<42:06, 4.87it/s] {'loss': 0.3017, 'grad_norm': 0.259765625, 'learning_rate': 5.127895350774871e-05, 'epoch': 0.74} 74%|███████▍ | 35700/48008 [2:02:28<42:06, 4.87it/s] 74%|███████▍ | 35701/48008 [2:02:28<41:10, 4.98it/s] 74%|███████▍ | 35702/48008 [2:02:28<41:00, 5.00it/s] 74%|███████▍ | 35703/48008 [2:02:29<40:27, 5.07it/s] 74%|███████▍ | 35704/48008 [2:02:29<42:59, 4.77it/s] 74%|███████▍ | 35705/48008 [2:02:29<41:36, 4.93it/s] 74%|███████▍ | 35706/48008 [2:02:29<41:30, 4.94it/s] 74%|███████▍ | 35707/48008 [2:02:29<40:24, 5.07it/s] 74%|███████▍ | 35708/48008 [2:02:30<40:31, 5.06it/s] 74%|███████▍ | 35709/48008 [2:02:30<39:50, 5.14it/s] 74%|███████▍ | 35710/48008 [2:02:30<39:44, 5.16it/s] 74%|███████▍ | 35711/48008 [2:02:30<39:09, 5.23it/s] 74%|███████▍ | 35712/48008 [2:02:30<39:02, 5.25it/s] 74%|███████▍ | 35713/48008 [2:02:31<39:45, 5.15it/s] 74%|███████▍ | 35714/48008 [2:02:31<39:48, 5.15it/s] 74%|███████▍ | 35715/48008 [2:02:31<39:40, 5.16it/s] 74%|███████▍ | 35716/48008 [2:02:31<39:31, 5.18it/s] 74%|███████▍ | 35717/48008 [2:02:31<39:21, 5.20it/s] 74%|███████▍ | 35718/48008 [2:02:32<1:16:36, 2.67it/s] 74%|███████▍ | 35719/48008 [2:02:32<1:05:24, 3.13it/s] 74%|███████▍ | 35720/48008 [2:02:33<57:26, 3.57it/s] 74%|███████▍ | 35721/48008 [2:02:33<53:08, 3.85it/s] 74%|███████▍ | 35722/48008 [2:02:33<50:09, 4.08it/s] 74%|███████▍ | 35723/48008 [2:02:33<46:43, 4.38it/s] 74%|███████▍ | 35724/48008 [2:02:33<51:28, 3.98it/s] 74%|███████▍ | 35725/48008 [2:02:34<48:47, 4.20it/s] 74%|███████▍ | 35726/48008 [2:02:34<45:38, 4.48it/s] 74%|███████▍ | 35727/48008 [2:02:34<43:46, 4.68it/s] 74%|███████▍ | 35728/48008 [2:02:34<42:18, 4.84it/s] 74%|███████▍ | 35729/48008 [2:02:34<41:56, 4.88it/s] 74%|███████▍ | 35730/48008 [2:02:35<41:25, 4.94it/s] 74%|███████▍ | 35731/48008 [2:02:35<41:09, 4.97it/s] 74%|███████▍ | 35732/48008 [2:02:35<41:26, 4.94it/s] 74%|███████▍ | 35733/48008 [2:02:35<40:45, 5.02it/s] 74%|███████▍ | 35734/48008 [2:02:35<40:35, 5.04it/s] 74%|███████▍ | 35735/48008 [2:02:36<39:40, 5.16it/s] 74%|███████▍ | 35736/48008 [2:02:36<39:29, 5.18it/s] 74%|███████▍ | 35737/48008 [2:02:36<39:06, 5.23it/s] 74%|███████▍ | 35738/48008 [2:02:36<38:50, 5.26it/s] 74%|███████▍ | 35739/48008 [2:02:36<38:48, 5.27it/s] 74%|███████▍ | 35740/48008 [2:02:37<1:21:08, 2.52it/s] 74%|███████▍ | 35741/48008 [2:02:37<1:08:23, 2.99it/s] 74%|███████▍ | 35742/48008 [2:02:38<59:47, 3.42it/s] 74%|███████▍ | 35743/48008 [2:02:38<53:54, 3.79it/s] 74%|███████▍ | 35744/48008 [2:02:38<49:15, 4.15it/s] 74%|███████▍ | 35745/48008 [2:02:38<46:11, 4.42it/s] 74%|███████▍ | 35746/48008 [2:02:38<43:54, 4.65it/s] 74%|███████▍ | 35747/48008 [2:02:39<49:29, 4.13it/s] 74%|███████▍ | 35748/48008 [2:02:39<46:03, 4.44it/s] 74%|███████▍ | 35749/48008 [2:02:39<44:12, 4.62it/s] 74%|███████▍ | 35750/48008 [2:02:39<42:24, 4.82it/s] {'loss': 0.3029, 'grad_norm': 0.1884765625, 'learning_rate': 5.107065489085153e-05, 'epoch': 0.74} 74%|███████▍ | 35750/48008 [2:02:39<42:24, 4.82it/s] 74%|███████▍ | 35751/48008 [2:02:39<41:24, 4.93it/s] 74%|███████▍ | 35752/48008 [2:02:40<40:48, 5.01it/s] 74%|███████▍ | 35753/48008 [2:02:40<40:04, 5.10it/s] 74%|███████▍ | 35754/48008 [2:02:40<39:55, 5.11it/s] 74%|███████▍ | 35755/48008 [2:02:40<39:37, 5.15it/s] 74%|███████▍ | 35756/48008 [2:02:40<40:07, 5.09it/s] 74%|███████▍ | 35757/48008 [2:02:41<39:27, 5.17it/s] 74%|███████▍ | 35758/48008 [2:02:41<39:51, 5.12it/s] 74%|███████▍ | 35759/48008 [2:02:41<39:40, 5.15it/s] 74%|███████▍ | 35760/48008 [2:02:41<40:11, 5.08it/s] 74%|███████▍ | 35761/48008 [2:02:41<39:57, 5.11it/s] 74%|███████▍ | 35762/48008 [2:02:42<39:41, 5.14it/s] 74%|███████▍ | 35763/48008 [2:02:42<39:23, 5.18it/s] 74%|███████▍ | 35764/48008 [2:02:42<39:32, 5.16it/s] 74%|███████▍ | 35765/48008 [2:02:42<39:22, 5.18it/s] 75%|███████▍ | 35766/48008 [2:02:42<39:33, 5.16it/s] 75%|███████▍ | 35767/48008 [2:02:43<39:19, 5.19it/s] 75%|███████▍ | 35768/48008 [2:02:43<39:17, 5.19it/s] 75%|███████▍ | 35769/48008 [2:02:43<39:02, 5.22it/s] 75%|███████▍ | 35770/48008 [2:02:43<39:06, 5.22it/s] 75%|███████▍ | 35771/48008 [2:02:43<38:53, 5.24it/s] 75%|███████▍ | 35772/48008 [2:02:44<38:59, 5.23it/s] 75%|███████▍ | 35773/48008 [2:02:44<39:53, 5.11it/s] 75%|███████▍ | 35774/48008 [2:02:44<39:47, 5.12it/s] 75%|███████▍ | 35775/48008 [2:02:44<39:20, 5.18it/s] 75%|███████▍ | 35776/48008 [2:02:44<39:23, 5.17it/s] 75%|███████▍ | 35777/48008 [2:02:44<39:16, 5.19it/s] 75%|███████▍ | 35778/48008 [2:02:45<39:04, 5.22it/s] 75%|███████▍ | 35779/48008 [2:02:45<38:56, 5.23it/s] 75%|███████▍ | 35780/48008 [2:02:45<38:54, 5.24it/s] 75%|███████▍ | 35781/48008 [2:02:45<39:40, 5.14it/s] 75%|███████▍ | 35782/48008 [2:02:45<39:17, 5.19it/s] 75%|███████▍ | 35783/48008 [2:02:46<38:59, 5.22it/s] 75%|███████▍ | 35784/48008 [2:02:46<38:51, 5.24it/s] 75%|███████▍ | 35785/48008 [2:02:46<38:51, 5.24it/s] 75%|███████▍ | 35786/48008 [2:02:46<39:12, 5.19it/s] 75%|███████▍ | 35787/48008 [2:02:46<39:23, 5.17it/s] 75%|███████▍ | 35788/48008 [2:02:47<39:04, 5.21it/s] 75%|███████▍ | 35789/48008 [2:02:47<51:07, 3.98it/s] 75%|███████▍ | 35790/48008 [2:02:47<47:22, 4.30it/s] 75%|███████▍ | 35791/48008 [2:02:47<44:42, 4.55it/s] 75%|███████▍ | 35792/48008 [2:02:48<42:56, 4.74it/s] 75%|███████▍ | 35793/48008 [2:02:48<41:40, 4.88it/s] 75%|███████▍ | 35794/48008 [2:02:48<40:25, 5.04it/s] 75%|███████▍ | 35795/48008 [2:02:48<39:59, 5.09it/s] 75%|███████▍ | 35796/48008 [2:02:48<39:43, 5.12it/s] 75%|███████▍ | 35797/48008 [2:02:49<51:28, 3.95it/s] 75%|███████▍ | 35798/48008 [2:02:49<47:29, 4.29it/s] 75%|███████▍ | 35799/48008 [2:02:49<45:32, 4.47it/s] 75%|███████▍ | 35800/48008 [2:02:49<43:30, 4.68it/s] {'loss': 0.3012, 'grad_norm': 0.146484375, 'learning_rate': 5.086235627395435e-05, 'epoch': 0.75} 75%|███████▍ | 35800/48008 [2:02:49<43:30, 4.68it/s] 75%|███████▍ | 35801/48008 [2:02:50<45:01, 4.52it/s] 75%|███████▍ | 35802/48008 [2:02:50<43:16, 4.70it/s] 75%|███████▍ | 35803/48008 [2:02:50<42:00, 4.84it/s] 75%|███████▍ | 35804/48008 [2:02:50<41:24, 4.91it/s] 75%|███████▍ | 35805/48008 [2:02:50<40:34, 5.01it/s] 75%|███████▍ | 35806/48008 [2:02:50<40:05, 5.07it/s] 75%|███████▍ | 35807/48008 [2:02:51<40:36, 5.01it/s] 75%|███████▍ | 35808/48008 [2:02:51<40:35, 5.01it/s] 75%|███████▍ | 35809/48008 [2:02:51<40:13, 5.05it/s] 75%|███████▍ | 35810/48008 [2:02:51<39:56, 5.09it/s] 75%|███████▍ | 35811/48008 [2:02:51<39:41, 5.12it/s] 75%|███████▍ | 35812/48008 [2:02:52<39:35, 5.14it/s] 75%|███████▍ | 35813/48008 [2:02:52<39:15, 5.18it/s] 75%|███████▍ | 35814/48008 [2:02:52<41:56, 4.85it/s] 75%|███████▍ | 35815/48008 [2:02:52<40:53, 4.97it/s] 75%|███████▍ | 35816/48008 [2:02:52<40:53, 4.97it/s] 75%|███████▍ | 35817/48008 [2:02:53<40:11, 5.06it/s] 75%|███████▍ | 35818/48008 [2:02:53<39:21, 5.16it/s] 75%|███████▍ | 35819/48008 [2:02:53<39:16, 5.17it/s] 75%|███████▍ | 35820/48008 [2:02:53<39:07, 5.19it/s] 75%|███████▍ | 35821/48008 [2:02:53<39:13, 5.18it/s] 75%|███████▍ | 35822/48008 [2:02:54<39:11, 5.18it/s] 75%|███████▍ | 35823/48008 [2:02:54<39:15, 5.17it/s] 75%|███████▍ | 35824/48008 [2:02:54<39:53, 5.09it/s] 75%|███████▍ | 35825/48008 [2:02:54<39:38, 5.12it/s] 75%|███████▍ | 35826/48008 [2:02:54<39:04, 5.20it/s] 75%|███████▍ | 35827/48008 [2:02:55<38:52, 5.22it/s] 75%|███████▍ | 35828/48008 [2:02:55<38:43, 5.24it/s] 75%|███████▍ | 35829/48008 [2:02:55<38:50, 5.23it/s] 75%|███████▍ | 35830/48008 [2:02:55<40:01, 5.07it/s] 75%|███████▍ | 35831/48008 [2:02:55<39:32, 5.13it/s] 75%|███████▍ | 35832/48008 [2:02:56<39:26, 5.15it/s] 75%|███████▍ | 35833/48008 [2:02:56<51:07, 3.97it/s] 75%|███████▍ | 35834/48008 [2:02:56<48:18, 4.20it/s] 75%|███████▍ | 35835/48008 [2:02:56<46:12, 4.39it/s] 75%|███████▍ | 35836/48008 [2:02:57<43:58, 4.61it/s] 75%|███████▍ | 35837/48008 [2:02:57<42:45, 4.74it/s] 75%|███████▍ | 35838/48008 [2:02:57<41:18, 4.91it/s] 75%|███████▍ | 35839/48008 [2:02:57<41:17, 4.91it/s] 75%|███████▍ | 35840/48008 [2:02:57<40:36, 4.99it/s] 75%|███████▍ | 35841/48008 [2:02:58<40:16, 5.04it/s] 75%|███████▍ | 35842/48008 [2:02:58<39:54, 5.08it/s] 75%|███████▍ | 35843/48008 [2:02:58<46:40, 4.34it/s] 75%|███████▍ | 35844/48008 [2:02:58<44:26, 4.56it/s] 75%|███████▍ | 35845/48008 [2:02:58<42:55, 4.72it/s] 75%|███████▍ | 35846/48008 [2:02:59<53:41, 3.78it/s] 75%|███████▍ | 35847/48008 [2:02:59<49:34, 4.09it/s] 75%|███████▍ | 35848/48008 [2:02:59<45:45, 4.43it/s] 75%|███████▍ | 35849/48008 [2:02:59<44:16, 4.58it/s] 75%|███████▍ | 35850/48008 [2:03:00<42:35, 4.76it/s] {'loss': 0.2992, 'grad_norm': 0.1962890625, 'learning_rate': 5.065405765705716e-05, 'epoch': 0.75} 75%|███████▍ | 35850/48008 [2:03:00<42:35, 4.76it/s] 75%|███████▍ | 35851/48008 [2:03:00<41:19, 4.90it/s] 75%|███████▍ | 35852/48008 [2:03:00<40:36, 4.99it/s] 75%|███████▍ | 35853/48008 [2:03:00<40:02, 5.06it/s] 75%|███████▍ | 35854/48008 [2:03:00<40:45, 4.97it/s] 75%|███████▍ | 35855/48008 [2:03:01<42:57, 4.71it/s] 75%|███████▍ | 35856/48008 [2:03:01<41:48, 4.84it/s] 75%|███████▍ | 35857/48008 [2:03:01<41:11, 4.92it/s] 75%|███████▍ | 35858/48008 [2:03:01<40:23, 5.01it/s] 75%|███████▍ | 35859/48008 [2:03:01<39:50, 5.08it/s] 75%|███████▍ | 35860/48008 [2:03:02<39:30, 5.13it/s] 75%|███████▍ | 35861/48008 [2:03:02<39:12, 5.16it/s] 75%|███████▍ | 35862/48008 [2:03:02<38:49, 5.21it/s] 75%|███████▍ | 35863/48008 [2:03:02<45:47, 4.42it/s] 75%|███████▍ | 35864/48008 [2:03:02<44:45, 4.52it/s] 75%|███████▍ | 35865/48008 [2:03:03<43:20, 4.67it/s] 75%|███████▍ | 35866/48008 [2:03:03<42:22, 4.77it/s] 75%|███████▍ | 35867/48008 [2:03:03<41:11, 4.91it/s] 75%|███████▍ | 35868/48008 [2:03:03<40:16, 5.02it/s] 75%|███████▍ | 35869/48008 [2:03:03<40:08, 5.04it/s] 75%|███████▍ | 35870/48008 [2:03:04<39:43, 5.09it/s] 75%|███████▍ | 35871/48008 [2:03:04<39:25, 5.13it/s] 75%|███████▍ | 35872/48008 [2:03:04<39:26, 5.13it/s] 75%|███████▍ | 35873/48008 [2:03:04<39:10, 5.16it/s] 75%|███████▍ | 35874/48008 [2:03:04<39:07, 5.17it/s] 75%|███████▍ | 35875/48008 [2:03:05<40:05, 5.04it/s] 75%|███████▍ | 35876/48008 [2:03:05<39:31, 5.12it/s] 75%|███████▍ | 35877/48008 [2:03:05<39:20, 5.14it/s] 75%|███████▍ | 35878/48008 [2:03:05<38:59, 5.19it/s] 75%|███████▍ | 35879/48008 [2:03:05<38:24, 5.26it/s] 75%|███████▍ | 35880/48008 [2:03:06<38:48, 5.21it/s] 75%|███████▍ | 35881/48008 [2:03:06<38:49, 5.21it/s] 75%|███████▍ | 35882/48008 [2:03:06<38:29, 5.25it/s] 75%|███████▍ | 35883/48008 [2:03:06<38:55, 5.19it/s] 75%|███████▍ | 35884/48008 [2:03:06<41:44, 4.84it/s] 75%|███████▍ | 35885/48008 [2:03:07<40:50, 4.95it/s] 75%|███████▍ | 35886/48008 [2:03:07<40:06, 5.04it/s] 75%|███████▍ | 35887/48008 [2:03:07<39:35, 5.10it/s] 75%|███████▍ | 35888/48008 [2:03:07<38:59, 5.18it/s] 75%|███████▍ | 35889/48008 [2:03:07<39:37, 5.10it/s] 75%|███████▍ | 35890/48008 [2:03:08<39:25, 5.12it/s] 75%|███████▍ | 35891/48008 [2:03:08<39:14, 5.15it/s] 75%|███████▍ | 35892/48008 [2:03:08<38:39, 5.22it/s] 75%|███████▍ | 35893/48008 [2:03:08<50:29, 4.00it/s] 75%|███████▍ | 35894/48008 [2:03:08<46:47, 4.32it/s] 75%|███████▍ | 35895/48008 [2:03:09<44:17, 4.56it/s] 75%|███████▍ | 35896/48008 [2:03:09<42:32, 4.75it/s] 75%|███████▍ | 35897/48008 [2:03:09<41:12, 4.90it/s] 75%|███████▍ | 35898/48008 [2:03:09<40:26, 4.99it/s] 75%|███████▍ | 35899/48008 [2:03:09<39:34, 5.10it/s] 75%|███████▍ | 35900/48008 [2:03:10<40:18, 5.01it/s] 75%|███████▍ | 35900/48008 [2:03:10<40:18, 5.01it/s]{'loss': 0.3009, 'grad_norm': 0.466796875, 'learning_rate': 5.044575904015998e-05, 'epoch': 0.75} 75%|███████▍ | 35901/48008 [2:03:10<39:58, 5.05it/s] 75%|███████▍ | 35902/48008 [2:03:10<39:23, 5.12it/s] 75%|███████▍ | 35903/48008 [2:03:10<39:32, 5.10it/s] 75%|███████▍ | 35904/48008 [2:03:10<39:40, 5.08it/s] 75%|███████▍ | 35905/48008 [2:03:11<40:09, 5.02it/s] 75%|███████▍ | 35906/48008 [2:03:11<39:44, 5.07it/s] 75%|███████▍ | 35907/48008 [2:03:11<40:29, 4.98it/s] 75%|███████▍ | 35908/48008 [2:03:11<40:41, 4.96it/s] 75%|███████▍ | 35909/48008 [2:03:11<39:51, 5.06it/s] 75%|███████▍ | 35910/48008 [2:03:12<42:11, 4.78it/s] 75%|███████▍ | 35911/48008 [2:03:12<41:25, 4.87it/s] 75%|███████▍ | 35912/48008 [2:03:12<40:28, 4.98it/s] 75%|███████▍ | 35913/48008 [2:03:12<39:36, 5.09it/s] 75%|███████▍ | 35914/48008 [2:03:12<39:07, 5.15it/s] 75%|███████▍ | 35915/48008 [2:03:13<38:57, 5.17it/s] 75%|███████▍ | 35916/48008 [2:03:13<39:24, 5.11it/s] 75%|███████▍ | 35917/48008 [2:03:13<39:10, 5.14it/s] 75%|███████▍ | 35918/48008 [2:03:13<38:51, 5.19it/s] 75%|███████▍ | 35919/48008 [2:03:13<38:39, 5.21it/s] 75%|███████▍ | 35920/48008 [2:03:14<38:21, 5.25it/s] 75%|███████▍ | 35921/48008 [2:03:14<38:48, 5.19it/s] 75%|███████▍ | 35922/48008 [2:03:14<38:42, 5.20it/s] 75%|███████▍ | 35923/48008 [2:03:14<38:39, 5.21it/s] 75%|███████▍ | 35924/48008 [2:03:14<38:32, 5.23it/s] 75%|███████▍ | 35925/48008 [2:03:14<38:29, 5.23it/s] 75%|███████▍ | 35926/48008 [2:03:15<38:19, 5.25it/s] 75%|███████▍ | 35927/48008 [2:03:15<38:19, 5.25it/s] 75%|███████▍ | 35928/48008 [2:03:15<39:01, 5.16it/s] 75%|███████▍ | 35929/48008 [2:03:15<39:25, 5.11it/s] 75%|███████▍ | 35930/48008 [2:03:15<39:52, 5.05it/s] 75%|███████▍ | 35931/48008 [2:03:16<39:07, 5.14it/s] 75%|███████▍ | 35932/48008 [2:03:16<39:45, 5.06it/s] 75%|███████▍ | 35933/48008 [2:03:16<39:21, 5.11it/s] 75%|███████▍ | 35934/48008 [2:03:16<38:58, 5.16it/s] 75%|███████▍ | 35935/48008 [2:03:16<39:28, 5.10it/s] 75%|███████▍ | 35936/48008 [2:03:17<39:37, 5.08it/s] 75%|███████▍ | 35937/48008 [2:03:17<39:33, 5.09it/s] 75%|███████▍ | 35938/48008 [2:03:17<39:57, 5.03it/s] 75%|███████▍ | 35939/48008 [2:03:17<51:20, 3.92it/s] 75%|███████▍ | 35940/48008 [2:03:18<59:19, 3.39it/s] 75%|███████▍ | 35941/48008 [2:03:18<53:52, 3.73it/s] 75%|███████▍ | 35942/48008 [2:03:18<49:15, 4.08it/s] 75%|███████▍ | 35943/48008 [2:03:18<45:55, 4.38it/s] 75%|███████▍ | 35944/48008 [2:03:19<43:44, 4.60it/s] 75%|███████▍ | 35945/48008 [2:03:19<42:14, 4.76it/s] 75%|███████▍ | 35946/48008 [2:03:19<41:42, 4.82it/s] 75%|███████▍ | 35947/48008 [2:03:19<41:53, 4.80it/s] 75%|███████▍ | 35948/48008 [2:03:19<40:57, 4.91it/s] 75%|███████▍ | 35949/48008 [2:03:20<42:59, 4.67it/s] 75%|███████▍ | 35950/48008 [2:03:20<41:20, 4.86it/s] 75%|███████▍ | 35950/48008 [2:03:20<41:20, 4.86it/s]{'loss': 0.2989, 'grad_norm': 0.408203125, 'learning_rate': 5.0237460423262796e-05, 'epoch': 0.75} 75%|███████▍ | 35951/48008 [2:03:20<40:45, 4.93it/s] 75%|███████▍ | 35952/48008 [2:03:20<39:46, 5.05it/s] 75%|███████▍ | 35953/48008 [2:03:20<39:06, 5.14it/s] 75%|███████▍ | 35954/48008 [2:03:21<38:54, 5.16it/s] 75%|███████▍ | 35955/48008 [2:03:21<38:21, 5.24it/s] 75%|███████▍ | 35956/48008 [2:03:21<38:22, 5.24it/s] 75%|███████▍ | 35957/48008 [2:03:21<38:24, 5.23it/s] 75%|███████▍ | 35958/48008 [2:03:21<38:21, 5.24it/s] 75%|███████▍ | 35959/48008 [2:03:22<38:24, 5.23it/s] 75%|███████▍ | 35960/48008 [2:03:22<38:15, 5.25it/s] 75%|███████▍ | 35961/48008 [2:03:22<38:38, 5.20it/s] 75%|███████▍ | 35962/48008 [2:03:22<39:16, 5.11it/s] 75%|███████▍ | 35963/48008 [2:03:22<39:03, 5.14it/s] 75%|███████▍ | 35964/48008 [2:03:23<38:37, 5.20it/s] 75%|███████▍ | 35965/48008 [2:03:23<38:37, 5.20it/s] 75%|███████▍ | 35966/48008 [2:03:23<38:41, 5.19it/s] 75%|███████▍ | 35967/48008 [2:03:23<38:14, 5.25it/s] 75%|███████▍ | 35968/48008 [2:03:23<39:03, 5.14it/s] 75%|███████▍ | 35969/48008 [2:03:23<39:14, 5.11it/s] 75%|███████▍ | 35970/48008 [2:03:24<40:01, 5.01it/s] 75%|███████▍ | 35971/48008 [2:03:24<39:14, 5.11it/s] 75%|███████▍ | 35972/48008 [2:03:24<38:52, 5.16it/s] 75%|███████▍ | 35973/48008 [2:03:24<39:26, 5.09it/s] 75%|███████▍ | 35974/48008 [2:03:24<39:06, 5.13it/s] 75%|███████▍ | 35975/48008 [2:03:25<38:44, 5.18it/s] 75%|███████▍ | 35976/48008 [2:03:25<38:30, 5.21it/s] 75%|███████▍ | 35977/48008 [2:03:25<38:26, 5.22it/s] 75%|███████▍ | 35978/48008 [2:03:25<38:40, 5.18it/s] 75%|███████▍ | 35979/48008 [2:03:25<39:10, 5.12it/s] 75%|███████▍ | 35980/48008 [2:03:26<45:44, 4.38it/s] 75%|███████▍ | 35981/48008 [2:03:26<43:08, 4.65it/s] 75%|███████▍ | 35982/48008 [2:03:26<42:42, 4.69it/s] 75%|███████▍ | 35983/48008 [2:03:26<41:35, 4.82it/s] 75%|███████▍ | 35984/48008 [2:03:27<40:35, 4.94it/s] 75%|███████▍ | 35985/48008 [2:03:27<39:51, 5.03it/s] 75%|███████▍ | 35986/48008 [2:03:27<42:05, 4.76it/s] 75%|███████▍ | 35987/48008 [2:03:27<43:40, 4.59it/s] 75%|███████▍ | 35988/48008 [2:03:27<41:54, 4.78it/s] 75%|███████▍ | 35989/48008 [2:03:28<41:30, 4.83it/s] 75%|███████▍ | 35990/48008 [2:03:28<40:46, 4.91it/s] 75%|███████▍ | 35991/48008 [2:03:28<39:54, 5.02it/s] 75%|███████▍ | 35992/48008 [2:03:28<39:27, 5.07it/s] 75%|███████▍ | 35993/48008 [2:03:28<39:00, 5.13it/s] 75%|███████▍ | 35994/48008 [2:03:29<38:50, 5.15it/s] 75%|███████▍ | 35995/48008 [2:03:29<38:33, 5.19it/s] 75%|███████▍ | 35996/48008 [2:03:29<38:28, 5.20it/s] 75%|███████▍ | 35997/48008 [2:03:29<38:07, 5.25it/s] 75%|███████▍ | 35998/48008 [2:03:29<38:07, 5.25it/s] 75%|███████▍ | 35999/48008 [2:03:29<38:54, 5.14it/s] 75%|███████▍ | 36000/48008 [2:03:30<39:53, 5.02it/s] 75%|███████▍ | 36000/48008 [2:03:30<39:53, 5.02it/s]{'loss': 0.3012, 'grad_norm': 0.341796875, 'learning_rate': 5.0029161806365607e-05, 'epoch': 0.75} 75%|███████▍ | 36001/48008 [2:03:30<39:54, 5.01it/s] 75%|███████▍ | 36002/48008 [2:03:30<39:19, 5.09it/s] 75%|███████▍ | 36003/48008 [2:03:30<38:59, 5.13it/s] 75%|███████▍ | 36004/48008 [2:03:30<38:42, 5.17it/s] 75%|███████▍ | 36005/48008 [2:03:31<38:33, 5.19it/s] 75%|███████▌ | 36006/48008 [2:03:31<38:22, 5.21it/s] 75%|███████▌ | 36007/48008 [2:03:31<39:22, 5.08it/s] 75%|███████▌ | 36008/48008 [2:03:31<38:48, 5.15it/s] 75%|███████▌ | 36009/48008 [2:03:31<39:44, 5.03it/s] 75%|███████▌ | 36010/48008 [2:03:32<39:53, 5.01it/s] 75%|███████▌ | 36011/48008 [2:03:32<39:08, 5.11it/s] 75%|███████▌ | 36012/48008 [2:03:32<38:37, 5.18it/s] 75%|███████▌ | 36013/48008 [2:03:32<39:07, 5.11it/s] 75%|███████▌ | 36014/48008 [2:03:32<38:58, 5.13it/s] 75%|███████▌ | 36015/48008 [2:03:33<38:50, 5.15it/s] 75%|███████▌ | 36016/48008 [2:03:33<38:45, 5.16it/s] 75%|███████▌ | 36017/48008 [2:03:33<45:29, 4.39it/s] 75%|███████▌ | 36018/48008 [2:03:33<43:46, 4.57it/s] 75%|███████▌ | 36019/48008 [2:03:34<49:03, 4.07it/s] 75%|███████▌ | 36020/48008 [2:03:34<46:54, 4.26it/s] 75%|███████▌ | 36021/48008 [2:03:34<44:49, 4.46it/s] 75%|███████▌ | 36022/48008 [2:03:34<43:36, 4.58it/s] 75%|███████▌ | 36023/48008 [2:03:34<41:58, 4.76it/s] 75%|███████▌ | 36024/48008 [2:03:35<41:26, 4.82it/s] 75%|███████▌ | 36025/48008 [2:03:35<40:24, 4.94it/s] 75%|███████▌ | 36026/48008 [2:03:35<40:13, 4.96it/s] 75%|███████▌ | 36027/48008 [2:03:35<39:45, 5.02it/s] 75%|███████▌ | 36028/48008 [2:03:35<39:10, 5.10it/s] 75%|███████▌ | 36029/48008 [2:03:36<38:29, 5.19it/s] 75%|███████▌ | 36030/48008 [2:03:36<39:09, 5.10it/s] 75%|███████▌ | 36031/48008 [2:03:36<38:49, 5.14it/s] 75%|███████▌ | 36032/48008 [2:03:36<38:32, 5.18it/s] 75%|███████▌ | 36033/48008 [2:03:36<41:18, 4.83it/s] 75%|███████▌ | 36034/48008 [2:03:37<40:25, 4.94it/s] 75%|███████▌ | 36035/48008 [2:03:37<40:08, 4.97it/s] 75%|███████▌ | 36036/48008 [2:03:37<39:08, 5.10it/s] 75%|███████▌ | 36037/48008 [2:03:37<50:30, 3.95it/s] 75%|███████▌ | 36038/48008 [2:03:38<47:28, 4.20it/s] 75%|███████▌ | 36039/48008 [2:03:38<44:37, 4.47it/s] 75%|███████▌ | 36040/48008 [2:03:38<43:21, 4.60it/s] 75%|███████▌ | 36041/48008 [2:03:38<48:47, 4.09it/s] 75%|███████▌ | 36042/48008 [2:03:38<46:31, 4.29it/s] 75%|███████▌ | 36043/48008 [2:03:39<43:56, 4.54it/s] 75%|███████▌ | 36044/48008 [2:03:39<43:01, 4.63it/s] 75%|███████▌ | 36045/48008 [2:03:39<41:56, 4.75it/s] 75%|███████▌ | 36046/48008 [2:03:39<40:36, 4.91it/s] 75%|███████▌ | 36047/48008 [2:03:39<40:04, 4.98it/s] 75%|███████▌ | 36048/48008 [2:03:40<40:01, 4.98it/s] 75%|███████▌ | 36049/48008 [2:03:40<39:42, 5.02it/s] 75%|███████▌ | 36050/48008 [2:03:40<39:23, 5.06it/s] 75%|███████▌ | 36050/48008 [2:03:40<39:23, 5.06it/s]{'loss': 0.3004, 'grad_norm': 0.1552734375, 'learning_rate': 4.9820863189468424e-05, 'epoch': 0.75} 75%|███████▌ | 36051/48008 [2:03:40<38:58, 5.11it/s] 75%|███████▌ | 36052/48008 [2:03:40<39:31, 5.04it/s] 75%|███████▌ | 36053/48008 [2:03:41<39:13, 5.08it/s] 75%|███████▌ | 36054/48008 [2:03:41<38:51, 5.13it/s] 75%|███████▌ | 36055/48008 [2:03:41<41:24, 4.81it/s] 75%|███████▌ | 36056/48008 [2:03:41<40:22, 4.93it/s] 75%|███████▌ | 36057/48008 [2:03:42<46:36, 4.27it/s] 75%|███████▌ | 36058/48008 [2:03:42<44:06, 4.52it/s] 75%|███████▌ | 36059/48008 [2:03:42<42:16, 4.71it/s] 75%|███████▌ | 36060/48008 [2:03:42<41:28, 4.80it/s] 75%|███████▌ | 36061/48008 [2:03:42<41:02, 4.85it/s] 75%|███████▌ | 36062/48008 [2:03:43<39:53, 4.99it/s] 75%|███████▌ | 36063/48008 [2:03:43<39:35, 5.03it/s] 75%|███████▌ | 36064/48008 [2:03:43<39:32, 5.04it/s] 75%|███████▌ | 36065/48008 [2:03:43<38:55, 5.11it/s] 75%|███████▌ | 36066/48008 [2:03:43<45:32, 4.37it/s] 75%|███████▌ | 36067/48008 [2:03:44<43:00, 4.63it/s] 75%|███████▌ | 36068/48008 [2:03:44<42:22, 4.70it/s] 75%|███████▌ | 36069/48008 [2:03:44<42:15, 4.71it/s] 75%|███████▌ | 36070/48008 [2:03:44<41:10, 4.83it/s] 75%|███████▌ | 36071/48008 [2:03:44<41:02, 4.85it/s] 75%|███████▌ | 36072/48008 [2:03:45<40:14, 4.94it/s] 75%|███████▌ | 36073/48008 [2:03:45<40:10, 4.95it/s] 75%|███████▌ | 36074/48008 [2:03:45<40:13, 4.94it/s] 75%|███████▌ | 36075/48008 [2:03:45<39:31, 5.03it/s] 75%|███████▌ | 36076/48008 [2:03:45<38:56, 5.11it/s] 75%|███████▌ | 36077/48008 [2:03:46<38:58, 5.10it/s] 75%|███████▌ | 36078/48008 [2:03:46<38:56, 5.11it/s] 75%|███████▌ | 36079/48008 [2:03:46<38:30, 5.16it/s] 75%|███████▌ | 36080/48008 [2:03:46<38:24, 5.18it/s] 75%|███████▌ | 36081/48008 [2:03:46<38:18, 5.19it/s] 75%|███████▌ | 36082/48008 [2:03:47<38:09, 5.21it/s] 75%|███████▌ | 36083/48008 [2:03:47<38:43, 5.13it/s] 75%|███████▌ | 36084/48008 [2:03:47<38:14, 5.20it/s] 75%|███████▌ | 36085/48008 [2:03:47<38:25, 5.17it/s] 75%|███████▌ | 36086/48008 [2:03:47<37:58, 5.23it/s] 75%|███████▌ | 36087/48008 [2:03:48<38:04, 5.22it/s] 75%|███████▌ | 36088/48008 [2:03:48<38:13, 5.20it/s] 75%|███████▌ | 36089/48008 [2:03:48<38:07, 5.21it/s] 75%|███████▌ | 36090/48008 [2:03:48<38:25, 5.17it/s] 75%|███████▌ | 36091/48008 [2:03:48<38:52, 5.11it/s] 75%|███████▌ | 36092/48008 [2:03:49<41:20, 4.80it/s] 75%|███████▌ | 36093/48008 [2:03:49<40:16, 4.93it/s] 75%|███████▌ | 36094/48008 [2:03:49<39:56, 4.97it/s] 75%|███████▌ | 36095/48008 [2:03:49<40:07, 4.95it/s] 75%|███████▌ | 36096/48008 [2:03:49<39:37, 5.01it/s] 75%|███████▌ | 36097/48008 [2:03:50<38:50, 5.11it/s] 75%|███████▌ | 36098/48008 [2:03:50<39:09, 5.07it/s] 75%|███████▌ | 36099/48008 [2:03:50<39:26, 5.03it/s] 75%|███████▌ | 36100/48008 [2:03:50<39:27, 5.03it/s] {'loss': 0.299, 'grad_norm': 0.138671875, 'learning_rate': 4.961256457257124e-05, 'epoch': 0.75} 75%|███████▌ | 36100/48008 [2:03:50<39:27, 5.03it/s] 75%|███████▌ | 36101/48008 [2:03:50<39:05, 5.08it/s] 75%|███████▌ | 36102/48008 [2:03:50<38:05, 5.21it/s] 75%|███████▌ | 36103/48008 [2:03:51<38:03, 5.21it/s] 75%|███████▌ | 36104/48008 [2:03:51<37:46, 5.25it/s] 75%|███████▌ | 36105/48008 [2:03:51<37:39, 5.27it/s] 75%|███████▌ | 36106/48008 [2:03:51<38:22, 5.17it/s] 75%|███████▌ | 36107/48008 [2:03:51<38:14, 5.19it/s] 75%|███████▌ | 36108/48008 [2:03:52<37:49, 5.24it/s] 75%|███████▌ | 36109/48008 [2:03:52<37:42, 5.26it/s] 75%|███████▌ | 36110/48008 [2:03:52<37:34, 5.28it/s] 75%|███████▌ | 36111/48008 [2:03:52<37:26, 5.30it/s] 75%|███████▌ | 36112/48008 [2:03:52<37:34, 5.28it/s] 75%|███████▌ | 36113/48008 [2:03:53<38:51, 5.10it/s] 75%|███████▌ | 36114/48008 [2:03:53<38:57, 5.09it/s] 75%|███████▌ | 36115/48008 [2:03:53<38:23, 5.16it/s] 75%|███████▌ | 36116/48008 [2:03:53<38:06, 5.20it/s] 75%|███████▌ | 36117/48008 [2:03:53<38:46, 5.11it/s] 75%|███████▌ | 36118/48008 [2:03:54<38:16, 5.18it/s] 75%|███████▌ | 36119/48008 [2:03:54<38:07, 5.20it/s] 75%|███████▌ | 36120/48008 [2:03:54<38:13, 5.18it/s] 75%|███████▌ | 36121/48008 [2:03:54<38:05, 5.20it/s] 75%|███████▌ | 36122/48008 [2:03:54<38:16, 5.18it/s] 75%|███████▌ | 36123/48008 [2:03:55<38:01, 5.21it/s] 75%|███████▌ | 36124/48008 [2:03:55<37:59, 5.21it/s] 75%|███████▌ | 36125/48008 [2:03:55<44:48, 4.42it/s] 75%|███████▌ | 36126/48008 [2:03:55<42:43, 4.63it/s] 75%|███████▌ | 36127/48008 [2:03:55<41:57, 4.72it/s] 75%|███████▌ | 36128/48008 [2:03:56<40:55, 4.84it/s] 75%|███████▌ | 36129/48008 [2:03:56<40:36, 4.87it/s] 75%|███████▌ | 36130/48008 [2:03:56<39:26, 5.02it/s] 75%|███████▌ | 36131/48008 [2:03:56<39:10, 5.05it/s] 75%|███████▌ | 36132/48008 [2:03:56<38:52, 5.09it/s] 75%|███████▌ | 36133/48008 [2:03:57<39:29, 5.01it/s] 75%|███████▌ | 36134/48008 [2:03:57<39:51, 4.97it/s] 75%|███████▌ | 36135/48008 [2:03:57<39:50, 4.97it/s] 75%|███████▌ | 36136/48008 [2:03:57<39:18, 5.03it/s] 75%|███████▌ | 36137/48008 [2:03:57<38:57, 5.08it/s] 75%|███████▌ | 36138/48008 [2:03:58<38:26, 5.15it/s] 75%|███████▌ | 36139/48008 [2:03:58<38:09, 5.18it/s] 75%|███████▌ | 36140/48008 [2:03:58<38:20, 5.16it/s] 75%|███████▌ | 36141/48008 [2:03:58<40:52, 4.84it/s] 75%|███████▌ | 36142/48008 [2:03:58<40:38, 4.87it/s] 75%|███████▌ | 36143/48008 [2:03:59<39:36, 4.99it/s] 75%|███████▌ | 36144/48008 [2:03:59<39:00, 5.07it/s] 75%|███████▌ | 36145/48008 [2:03:59<39:17, 5.03it/s] 75%|███████▌ | 36146/48008 [2:03:59<39:01, 5.07it/s] 75%|███████▌ | 36147/48008 [2:03:59<39:21, 5.02it/s] 75%|███████▌ | 36148/48008 [2:04:00<39:00, 5.07it/s] 75%|███████▌ | 36149/48008 [2:04:00<38:40, 5.11it/s] 75%|███████▌ | 36150/48008 [2:04:00<38:45, 5.10it/s] 75%|███████▌ | 36150/48008 [2:04:00<38:45, 5.10it/s]{'loss': 0.3, 'grad_norm': 0.2197265625, 'learning_rate': 4.940426595567406e-05, 'epoch': 0.75} 75%|███████▌ | 36151/48008 [2:04:00<39:04, 5.06it/s] 75%|███████▌ | 36152/48008 [2:04:00<38:04, 5.19it/s] 75%|███████▌ | 36153/48008 [2:04:01<37:35, 5.26it/s] 75%|███████▌ | 36154/48008 [2:04:01<37:25, 5.28it/s] 75%|███████▌ | 36155/48008 [2:04:01<37:35, 5.26it/s] 75%|███████▌ | 36156/48008 [2:04:01<37:37, 5.25it/s] 75%|███████▌ | 36157/48008 [2:04:01<37:48, 5.22it/s] 75%|███████▌ | 36158/48008 [2:04:01<37:45, 5.23it/s] 75%|███████▌ | 36159/48008 [2:04:02<38:58, 5.07it/s] 75%|███████▌ | 36160/48008 [2:04:02<41:17, 4.78it/s] 75%|███████▌ | 36161/48008 [2:04:02<40:58, 4.82it/s] 75%|███████▌ | 36162/48008 [2:04:02<40:02, 4.93it/s] 75%|███████▌ | 36163/48008 [2:04:03<39:18, 5.02it/s] 75%|███████▌ | 36164/48008 [2:04:03<39:59, 4.94it/s] 75%|███████▌ | 36165/48008 [2:04:03<40:00, 4.93it/s] 75%|███████▌ | 36166/48008 [2:04:03<39:26, 5.00it/s] 75%|███████▌ | 36167/48008 [2:04:03<38:53, 5.07it/s] 75%|███████▌ | 36168/48008 [2:04:04<38:55, 5.07it/s] 75%|███████▌ | 36169/48008 [2:04:04<38:56, 5.07it/s] 75%|███████▌ | 36170/48008 [2:04:04<38:38, 5.11it/s] 75%|███████▌ | 36171/48008 [2:04:04<38:14, 5.16it/s] 75%|███████▌ | 36172/48008 [2:04:04<49:38, 3.97it/s] 75%|███████▌ | 36173/48008 [2:04:05<45:55, 4.30it/s] 75%|███████▌ | 36174/48008 [2:04:05<43:10, 4.57it/s] 75%|███████▌ | 36175/48008 [2:04:05<41:12, 4.79it/s] 75%|███████▌ | 36176/48008 [2:04:05<40:02, 4.92it/s] 75%|███████▌ | 36177/48008 [2:04:05<39:06, 5.04it/s] 75%|███████▌ | 36178/48008 [2:04:06<39:03, 5.05it/s] 75%|███████▌ | 36179/48008 [2:04:06<39:19, 5.01it/s] 75%|███████▌ | 36180/48008 [2:04:06<41:43, 4.73it/s] 75%|███████▌ | 36181/48008 [2:04:06<41:11, 4.78it/s] 75%|███████▌ | 36182/48008 [2:04:06<40:06, 4.91it/s] 75%|███████▌ | 36183/48008 [2:04:07<39:45, 4.96it/s] 75%|███████▌ | 36184/48008 [2:04:07<39:29, 4.99it/s] 75%|███████▌ | 36185/48008 [2:04:07<40:00, 4.93it/s] 75%|███████▌ | 36186/48008 [2:04:07<39:11, 5.03it/s] 75%|███████▌ | 36187/48008 [2:04:07<38:42, 5.09it/s] 75%|███████▌ | 36188/48008 [2:04:08<39:03, 5.04it/s] 75%|███████▌ | 36189/48008 [2:04:08<38:32, 5.11it/s] 75%|███████▌ | 36190/48008 [2:04:08<38:04, 5.17it/s] 75%|███████▌ | 36191/48008 [2:04:08<37:56, 5.19it/s] 75%|███████▌ | 36192/48008 [2:04:09<49:35, 3.97it/s] 75%|███████▌ | 36193/48008 [2:04:09<45:47, 4.30it/s] 75%|███████▌ | 36194/48008 [2:04:09<43:25, 4.53it/s] 75%|███████▌ | 36195/48008 [2:04:09<42:19, 4.65it/s] 75%|███████▌ | 36196/48008 [2:04:09<40:46, 4.83it/s] 75%|███████▌ | 36197/48008 [2:04:10<39:49, 4.94it/s] 75%|███████▌ | 36198/48008 [2:04:10<38:59, 5.05it/s] 75%|███████▌ | 36199/48008 [2:04:10<38:46, 5.08it/s] 75%|███████▌ | 36200/48008 [2:04:10<38:01, 5.18it/s] {'loss': 0.3001, 'grad_norm': 0.439453125, 'learning_rate': 4.919596733877687e-05, 'epoch': 0.75} 75%|███████▌ | 36200/48008 [2:04:10<38:01, 5.18it/s] 75%|███████▌ | 36201/48008 [2:04:10<37:47, 5.21it/s] 75%|███████▌ | 36202/48008 [2:04:10<37:44, 5.21it/s] 75%|███████▌ | 36203/48008 [2:04:11<38:22, 5.13it/s] 75%|███████▌ | 36204/48008 [2:04:11<38:11, 5.15it/s] 75%|███████▌ | 36205/48008 [2:04:11<38:03, 5.17it/s] 75%|███████▌ | 36206/48008 [2:04:11<38:00, 5.18it/s] 75%|███████▌ | 36207/48008 [2:04:12<44:49, 4.39it/s] 75%|███████▌ | 36208/48008 [2:04:12<43:44, 4.50it/s] 75%|███████▌ | 36209/48008 [2:04:12<41:52, 4.70it/s] 75%|███████▌ | 36210/48008 [2:04:12<40:31, 4.85it/s] 75%|███████▌ | 36211/48008 [2:04:13<51:11, 3.84it/s] 75%|███████▌ | 36212/48008 [2:04:13<47:50, 4.11it/s] 75%|███████▌ | 36213/48008 [2:04:13<44:26, 4.42it/s] 75%|███████▌ | 36214/48008 [2:04:13<42:07, 4.67it/s] 75%|███████▌ | 36215/48008 [2:04:13<40:32, 4.85it/s] 75%|███████▌ | 36216/48008 [2:04:14<39:23, 4.99it/s] 75%|███████▌ | 36217/48008 [2:04:14<38:51, 5.06it/s] 75%|███████▌ | 36218/48008 [2:04:14<38:49, 5.06it/s] 75%|███████▌ | 36219/48008 [2:04:14<38:19, 5.13it/s] 75%|███████▌ | 36220/48008 [2:04:14<37:59, 5.17it/s] 75%|███████▌ | 36221/48008 [2:04:14<37:53, 5.18it/s] 75%|███████▌ | 36222/48008 [2:04:15<38:03, 5.16it/s] 75%|███████▌ | 36223/48008 [2:04:15<37:55, 5.18it/s] 75%|███████▌ | 36224/48008 [2:04:15<37:42, 5.21it/s] 75%|███████▌ | 36225/48008 [2:04:15<40:20, 4.87it/s] 75%|███████▌ | 36226/48008 [2:04:15<39:48, 4.93it/s] 75%|███████▌ | 36227/48008 [2:04:16<39:48, 4.93it/s] 75%|███████▌ | 36228/48008 [2:04:16<39:32, 4.97it/s] 75%|███████▌ | 36229/48008 [2:04:16<38:43, 5.07it/s] 75%|███████▌ | 36230/48008 [2:04:16<38:15, 5.13it/s] 75%|███████▌ | 36231/48008 [2:04:16<38:15, 5.13it/s] 75%|███████▌ | 36232/48008 [2:04:17<40:39, 4.83it/s] 75%|███████▌ | 36233/48008 [2:04:17<40:45, 4.82it/s] 75%|███████▌ | 36234/48008 [2:04:17<40:29, 4.85it/s] 75%|███████▌ | 36235/48008 [2:04:17<39:33, 4.96it/s] 75%|███████▌ | 36236/48008 [2:04:17<39:00, 5.03it/s] 75%|███████▌ | 36237/48008 [2:04:18<39:14, 5.00it/s] 75%|███████▌ | 36238/48008 [2:04:18<45:32, 4.31it/s] 75%|███████▌ | 36239/48008 [2:04:18<43:02, 4.56it/s] 75%|███████▌ | 36240/48008 [2:04:19<52:47, 3.71it/s] 75%|███████▌ | 36241/48008 [2:04:19<48:41, 4.03it/s] 75%|███████▌ | 36242/48008 [2:04:19<45:01, 4.36it/s] 75%|███████▌ | 36243/48008 [2:04:19<45:29, 4.31it/s] 75%|███████▌ | 36244/48008 [2:04:19<43:47, 4.48it/s] 75%|███████▌ | 36245/48008 [2:04:20<41:46, 4.69it/s] 75%|███████▌ | 36246/48008 [2:04:20<40:37, 4.83it/s] 76%|███████▌ | 36247/48008 [2:04:20<39:49, 4.92it/s] 76%|███████▌ | 36248/48008 [2:04:20<41:54, 4.68it/s] 76%|███████▌ | 36249/48008 [2:04:20<40:50, 4.80it/s] 76%|███████▌ | 36250/48008 [2:04:21<39:29, 4.96it/s] 76%|███████▌ | 36250/48008 [2:04:21<39:29, 4.96it/s]{'loss': 0.2988, 'grad_norm': 0.1923828125, 'learning_rate': 4.898766872187969e-05, 'epoch': 0.76} 76%|███████▌ | 36251/48008 [2:04:22<1:43:07, 1.90it/s] 76%|███████▌ | 36252/48008 [2:04:22<1:23:23, 2.35it/s] 76%|███████▌ | 36253/48008 [2:04:22<1:10:12, 2.79it/s] 76%|███████▌ | 36254/48008 [2:04:22<1:00:18, 3.25it/s] 76%|███████▌ | 36255/48008 [2:04:23<53:32, 3.66it/s] 76%|███████▌ | 36256/48008 [2:04:23<51:40, 3.79it/s] 76%|███████▌ | 36257/48008 [2:04:23<48:16, 4.06it/s] 76%|███████▌ | 36258/48008 [2:04:23<45:41, 4.29it/s] 76%|███████▌ | 36259/48008 [2:04:23<43:06, 4.54it/s] 76%|███████▌ | 36260/48008 [2:04:24<42:06, 4.65it/s] 76%|███████▌ | 36261/48008 [2:04:24<40:49, 4.80it/s] 76%|███████▌ | 36262/48008 [2:04:24<39:40, 4.93it/s] 76%|███████▌ | 36263/48008 [2:04:24<38:47, 5.05it/s] 76%|███████▌ | 36264/48008 [2:04:24<39:09, 5.00it/s] 76%|███████▌ | 36265/48008 [2:04:25<38:35, 5.07it/s] 76%|███████▌ | 36266/48008 [2:04:25<38:04, 5.14it/s] 76%|███████▌ | 36267/48008 [2:04:25<37:46, 5.18it/s] 76%|███████▌ | 36268/48008 [2:04:25<37:22, 5.24it/s] 76%|███████▌ | 36269/48008 [2:04:25<37:05, 5.27it/s] 76%|███████▌ | 36270/48008 [2:04:26<36:59, 5.29it/s] 76%|███████▌ | 36271/48008 [2:04:27<1:20:38, 2.43it/s] 76%|███████▌ | 36272/48008 [2:04:27<1:07:31, 2.90it/s] 76%|███████▌ | 36273/48008 [2:04:27<58:34, 3.34it/s] 76%|███████▌ | 36274/48008 [2:04:27<53:04, 3.69it/s] 76%|███████▌ | 36275/48008 [2:04:27<48:22, 4.04it/s] 76%|███████▌ | 36276/48008 [2:04:27<45:04, 4.34it/s] 76%|███████▌ | 36277/48008 [2:04:28<1:20:12, 2.44it/s] 76%|███████▌ | 36278/48008 [2:04:29<1:07:16, 2.91it/s] 76%|███████▌ | 36279/48008 [2:04:29<59:02, 3.31it/s] 76%|███████▌ | 36280/48008 [2:04:29<52:10, 3.75it/s] 76%|███████▌ | 36281/48008 [2:04:29<47:28, 4.12it/s] 76%|███████▌ | 36282/48008 [2:04:29<44:32, 4.39it/s] 76%|███████▌ | 36283/48008 [2:04:30<1:30:34, 2.16it/s] 76%|███████▌ | 36284/48008 [2:04:30<1:14:31, 2.62it/s] 76%|███████▌ | 36285/48008 [2:04:31<1:03:08, 3.09it/s] 76%|███████▌ | 36286/48008 [2:04:31<55:32, 3.52it/s] 76%|███████▌ | 36287/48008 [2:04:31<50:03, 3.90it/s] 76%|███████▌ | 36288/48008 [2:04:31<46:16, 4.22it/s] 76%|███████▌ | 36289/48008 [2:04:32<1:19:32, 2.46it/s] 76%|███████▌ | 36290/48008 [2:04:32<1:18:24, 2.49it/s] 76%|███████▌ | 36291/48008 [2:04:33<1:06:30, 2.94it/s] 76%|███████▌ | 36292/48008 [2:04:33<57:39, 3.39it/s] 76%|███████▌ | 36293/48008 [2:04:33<51:27, 3.79it/s] 76%|███████▌ | 36294/48008 [2:04:33<47:12, 4.14it/s] 76%|███████▌ | 36295/48008 [2:04:33<44:57, 4.34it/s] 76%|███████▌ | 36296/48008 [2:04:34<42:22, 4.61it/s] 76%|███████▌ | 36297/48008 [2:04:34<40:41, 4.80it/s] 76%|███████▌ | 36298/48008 [2:04:34<40:09, 4.86it/s] 76%|███████▌ | 36299/48008 [2:04:34<39:34, 4.93it/s] 76%|███████▌ | 36300/48008 [2:04:34<38:55, 5.01it/s] 76%|███████▌ | 36300/48008 [2:04:34<38:55, 5.01it/s]{'loss': 0.3006, 'grad_norm': 0.1455078125, 'learning_rate': 4.8779370104982504e-05, 'epoch': 0.76} 76%|███████▌ | 36301/48008 [2:04:35<38:21, 5.09it/s] 76%|███████▌ | 36302/48008 [2:04:35<37:54, 5.15it/s] 76%|███████▌ | 36303/48008 [2:04:35<37:40, 5.18it/s] 76%|███████▌ | 36304/48008 [2:04:35<38:11, 5.11it/s] 76%|███████▌ | 36305/48008 [2:04:35<38:57, 5.01it/s] 76%|███████▌ | 36306/48008 [2:04:36<37:50, 5.15it/s] 76%|███████▌ | 36307/48008 [2:04:36<49:13, 3.96it/s] 76%|███████▌ | 36308/48008 [2:04:36<46:01, 4.24it/s] 76%|███████▌ | 36309/48008 [2:04:36<43:16, 4.51it/s] 76%|███████▌ | 36310/48008 [2:04:37<41:54, 4.65it/s] 76%|███████▌ | 36311/48008 [2:04:37<40:32, 4.81it/s] 76%|███████▌ | 36312/48008 [2:04:37<39:31, 4.93it/s] 76%|███████▌ | 36313/48008 [2:04:37<38:38, 5.04it/s] 76%|███████▌ | 36314/48008 [2:04:37<38:15, 5.09it/s] 76%|███████▌ | 36315/48008 [2:04:37<37:43, 5.17it/s] 76%|███████▌ | 36316/48008 [2:04:38<37:23, 5.21it/s] 76%|███████▌ | 36317/48008 [2:04:38<37:22, 5.21it/s] 76%|███████▌ | 36318/48008 [2:04:38<37:42, 5.17it/s] 76%|███████▌ | 36319/48008 [2:04:38<37:32, 5.19it/s] 76%|███████▌ | 36320/48008 [2:04:38<38:30, 5.06it/s] 76%|███████▌ | 36321/48008 [2:04:39<38:13, 5.10it/s] 76%|███████▌ | 36322/48008 [2:04:39<37:36, 5.18it/s] 76%|███████▌ | 36323/48008 [2:04:39<37:41, 5.17it/s] 76%|███████▌ | 36324/48008 [2:04:39<38:13, 5.09it/s] 76%|███████▌ | 36325/48008 [2:04:39<38:08, 5.11it/s] 76%|███████▌ | 36326/48008 [2:04:40<38:57, 5.00it/s] 76%|███████▌ | 36327/48008 [2:04:40<38:28, 5.06it/s] 76%|███████▌ | 36328/48008 [2:04:40<39:17, 4.96it/s] 76%|███████▌ | 36329/48008 [2:04:40<38:45, 5.02it/s] 76%|███████▌ | 36330/48008 [2:04:40<40:58, 4.75it/s] 76%|███████▌ | 36331/48008 [2:04:41<1:16:23, 2.55it/s] 76%|███████▌ | 36332/48008 [2:04:41<1:04:31, 3.02it/s] 76%|███████▌ | 36333/48008 [2:04:42<56:06, 3.47it/s] 76%|███████▌ | 36334/48008 [2:04:42<50:29, 3.85it/s] 76%|███████▌ | 36335/48008 [2:04:42<47:12, 4.12it/s] 76%|███████▌ | 36336/48008 [2:04:42<45:19, 4.29it/s] 76%|███████▌ | 36337/48008 [2:04:42<42:32, 4.57it/s] 76%|███████▌ | 36338/48008 [2:04:43<40:42, 4.78it/s] 76%|███████▌ | 36339/48008 [2:04:43<39:36, 4.91it/s] 76%|███████▌ | 36340/48008 [2:04:43<39:09, 4.97it/s] 76%|███████▌ | 36341/48008 [2:04:43<38:35, 5.04it/s] 76%|███████▌ | 36342/48008 [2:04:43<38:24, 5.06it/s] 76%|███████▌ | 36343/48008 [2:04:44<38:17, 5.08it/s] 76%|███████▌ | 36344/48008 [2:04:44<40:48, 4.76it/s] 76%|███████▌ | 36345/48008 [2:04:44<42:38, 4.56it/s] 76%|███████▌ | 36346/48008 [2:04:44<41:03, 4.73it/s] 76%|███████▌ | 36347/48008 [2:04:44<39:52, 4.87it/s] 76%|███████▌ | 36348/48008 [2:04:45<39:31, 4.92it/s] 76%|███████▌ | 36349/48008 [2:04:45<39:59, 4.86it/s] 76%|███████▌ | 36350/48008 [2:04:45<39:10, 4.96it/s] {'loss': 0.3027, 'grad_norm': 0.380859375, 'learning_rate': 4.857107148808532e-05, 'epoch': 0.76} 76%|███████▌ | 36350/48008 [2:04:45<39:10, 4.96it/s] 76%|███████▌ | 36351/48008 [2:04:45<38:29, 5.05it/s] 76%|███████▌ | 36352/48008 [2:04:45<38:10, 5.09it/s] 76%|███████▌ | 36353/48008 [2:04:46<37:55, 5.12it/s] 76%|███████▌ | 36354/48008 [2:04:46<37:57, 5.12it/s] 76%|███████▌ | 36355/48008 [2:04:46<38:20, 5.06it/s] 76%|███████▌ | 36356/48008 [2:04:46<37:42, 5.15it/s] 76%|███████▌ | 36357/48008 [2:04:46<37:39, 5.16it/s] 76%|███████▌ | 36358/48008 [2:04:47<37:31, 5.17it/s] 76%|███████▌ | 36359/48008 [2:04:47<37:33, 5.17it/s] 76%|███████▌ | 36360/48008 [2:04:47<38:25, 5.05it/s] 76%|███████▌ | 36361/48008 [2:04:47<38:25, 5.05it/s] 76%|███████▌ | 36362/48008 [2:04:47<38:00, 5.11it/s] 76%|███████▌ | 36363/48008 [2:04:48<38:05, 5.09it/s] 76%|███████▌ | 36364/48008 [2:04:48<37:49, 5.13it/s] 76%|███████▌ | 36365/48008 [2:04:48<37:00, 5.24it/s] 76%|███████▌ | 36366/48008 [2:04:48<37:30, 5.17it/s] 76%|███████▌ | 36367/48008 [2:04:48<37:45, 5.14it/s] 76%|███████▌ | 36368/48008 [2:04:49<37:36, 5.16it/s] 76%|███████▌ | 36369/48008 [2:04:49<37:51, 5.12it/s] 76%|███████▌ | 36370/48008 [2:04:49<38:13, 5.07it/s] 76%|███████▌ | 36371/48008 [2:04:49<38:23, 5.05it/s] 76%|███████▌ | 36372/48008 [2:04:49<38:07, 5.09it/s] 76%|███████▌ | 36373/48008 [2:04:50<37:49, 5.13it/s] 76%|███████▌ | 36374/48008 [2:04:50<38:02, 5.10it/s] 76%|███████▌ | 36375/48008 [2:04:50<37:42, 5.14it/s] 76%|███████▌ | 36376/48008 [2:04:50<37:30, 5.17it/s] 76%|███████▌ | 36377/48008 [2:04:50<37:34, 5.16it/s] 76%|███████▌ | 36378/48008 [2:04:51<48:41, 3.98it/s] 76%|███████▌ | 36379/48008 [2:04:51<44:57, 4.31it/s] 76%|███████▌ | 36380/48008 [2:04:51<42:52, 4.52it/s] 76%|███████▌ | 36381/48008 [2:04:51<52:29, 3.69it/s] 76%|███████▌ | 36382/48008 [2:04:52<47:34, 4.07it/s] 76%|███████▌ | 36383/48008 [2:04:52<44:19, 4.37it/s] 76%|███████▌ | 36384/48008 [2:04:52<41:57, 4.62it/s] 76%|███████▌ | 36385/48008 [2:04:52<40:28, 4.79it/s] 76%|███████▌ | 36386/48008 [2:04:52<39:24, 4.91it/s] 76%|███████▌ | 36387/48008 [2:04:53<38:53, 4.98it/s] 76%|███████▌ | 36388/48008 [2:04:53<41:02, 4.72it/s] 76%|███████▌ | 36389/48008 [2:04:53<39:56, 4.85it/s] 76%|███████▌ | 36390/48008 [2:04:53<39:05, 4.95it/s] 76%|███████▌ | 36391/48008 [2:04:53<39:06, 4.95it/s] 76%|███████▌ | 36392/48008 [2:04:54<38:20, 5.05it/s] 76%|███████▌ | 36393/48008 [2:04:54<37:38, 5.14it/s] 76%|███████▌ | 36394/48008 [2:04:54<37:26, 5.17it/s] 76%|███████▌ | 36395/48008 [2:04:54<37:26, 5.17it/s] 76%|███████▌ | 36396/48008 [2:04:54<37:03, 5.22it/s] 76%|███████▌ | 36397/48008 [2:04:55<39:42, 4.87it/s] 76%|███████▌ | 36398/48008 [2:04:55<41:26, 4.67it/s] 76%|███████▌ | 36399/48008 [2:04:55<40:03, 4.83it/s] 76%|███████▌ | 36400/48008 [2:04:55<39:19, 4.92it/s] {'loss': 0.3, 'grad_norm': 0.1865234375, 'learning_rate': 4.836277287118814e-05, 'epoch': 0.76} 76%|███████▌ | 36400/48008 [2:04:55<39:19, 4.92it/s] 76%|███████▌ | 36401/48008 [2:04:55<38:59, 4.96it/s] 76%|███████▌ | 36402/48008 [2:04:56<38:16, 5.05it/s] 76%|███████▌ | 36403/48008 [2:04:56<37:52, 5.11it/s] 76%|███████▌ | 36404/48008 [2:04:56<38:16, 5.05it/s] 76%|███████▌ | 36405/48008 [2:04:56<37:34, 5.15it/s] 76%|███████▌ | 36406/48008 [2:04:56<37:25, 5.17it/s] 76%|███████▌ | 36407/48008 [2:04:57<37:09, 5.20it/s] 76%|███████▌ | 36408/48008 [2:04:57<36:58, 5.23it/s] 76%|███████▌ | 36409/48008 [2:04:57<38:05, 5.08it/s] 76%|███████▌ | 36410/48008 [2:04:57<37:34, 5.15it/s] 76%|███████▌ | 36411/48008 [2:04:57<37:23, 5.17it/s] 76%|███████▌ | 36412/48008 [2:04:58<37:23, 5.17it/s] 76%|███████▌ | 36413/48008 [2:04:58<37:43, 5.12it/s] 76%|███████▌ | 36414/48008 [2:04:58<37:30, 5.15it/s] 76%|███████▌ | 36415/48008 [2:04:58<37:20, 5.17it/s] 76%|███████▌ | 36416/48008 [2:04:58<37:00, 5.22it/s] 76%|███████▌ | 36417/48008 [2:04:59<36:58, 5.22it/s] 76%|███████▌ | 36418/48008 [2:04:59<36:57, 5.23it/s] 76%|███████▌ | 36419/48008 [2:04:59<36:34, 5.28it/s] 76%|███████▌ | 36420/48008 [2:04:59<36:25, 5.30it/s] 76%|███████▌ | 36421/48008 [2:04:59<37:11, 5.19it/s] 76%|███████▌ | 36422/48008 [2:04:59<36:57, 5.23it/s] 76%|███████▌ | 36423/48008 [2:05:00<36:40, 5.26it/s] 76%|███████▌ | 36424/48008 [2:05:00<36:36, 5.27it/s] 76%|███████▌ | 36425/48008 [2:05:00<36:42, 5.26it/s] 76%|███████▌ | 36426/48008 [2:05:00<36:38, 5.27it/s] 76%|███████▌ | 36427/48008 [2:05:00<36:27, 5.29it/s] 76%|███████▌ | 36428/48008 [2:05:01<36:36, 5.27it/s] 76%|███████▌ | 36429/48008 [2:05:01<37:07, 5.20it/s] 76%|███████▌ | 36430/48008 [2:05:01<36:55, 5.23it/s] 76%|███████▌ | 36431/48008 [2:05:01<36:50, 5.24it/s] 76%|███████▌ | 36432/48008 [2:05:01<37:53, 5.09it/s] 76%|███████▌ | 36433/48008 [2:05:02<40:14, 4.79it/s] 76%|███████▌ | 36434/48008 [2:05:02<39:24, 4.90it/s] 76%|███████▌ | 36435/48008 [2:05:02<49:52, 3.87it/s] 76%|███████▌ | 36436/48008 [2:05:02<45:46, 4.21it/s] 76%|███████▌ | 36437/48008 [2:05:03<45:48, 4.21it/s] 76%|███████▌ | 36438/48008 [2:05:03<43:05, 4.47it/s] 76%|███████▌ | 36439/48008 [2:05:03<41:50, 4.61it/s] 76%|███████▌ | 36440/48008 [2:05:03<39:57, 4.82it/s] 76%|███████▌ | 36441/48008 [2:05:03<39:04, 4.93it/s] 76%|███████▌ | 36442/48008 [2:05:04<39:07, 4.93it/s] 76%|███████▌ | 36443/48008 [2:05:04<39:02, 4.94it/s] 76%|███████▌ | 36444/48008 [2:05:04<38:12, 5.05it/s] 76%|███████▌ | 36445/48008 [2:05:04<49:01, 3.93it/s] 76%|███████▌ | 36446/48008 [2:05:05<45:22, 4.25it/s] 76%|███████▌ | 36447/48008 [2:05:05<42:31, 4.53it/s] 76%|███████▌ | 36448/48008 [2:05:05<40:38, 4.74it/s] 76%|███████▌ | 36449/48008 [2:05:05<39:22, 4.89it/s] 76%|███████▌ | 36450/48008 [2:05:05<38:03, 5.06it/s] {'loss': 0.2993, 'grad_norm': 0.5625, 'learning_rate': 4.8154474254290956e-05, 'epoch': 0.76} 76%|███████▌ | 36450/48008 [2:05:05<38:03, 5.06it/s] 76%|███████▌ | 36451/48008 [2:05:06<37:51, 5.09it/s] 76%|███████▌ | 36452/48008 [2:05:06<37:30, 5.13it/s] 76%|███████▌ | 36453/48008 [2:05:06<37:11, 5.18it/s] 76%|███████▌ | 36454/48008 [2:05:06<36:55, 5.21it/s] 76%|███████▌ | 36455/48008 [2:05:06<36:56, 5.21it/s] 76%|███████▌ | 36456/48008 [2:05:06<36:45, 5.24it/s] 76%|███████▌ | 36457/48008 [2:05:07<36:50, 5.22it/s] 76%|███████▌ | 36458/48008 [2:05:07<37:25, 5.14it/s] 76%|███████▌ | 36459/48008 [2:05:07<37:10, 5.18it/s] 76%|███████▌ | 36460/48008 [2:05:07<36:51, 5.22it/s] 76%|███████▌ | 36461/48008 [2:05:07<36:34, 5.26it/s] 76%|███████▌ | 36462/48008 [2:05:08<36:43, 5.24it/s] 76%|███████▌ | 36463/48008 [2:05:08<36:48, 5.23it/s] 76%|███████▌ | 36464/48008 [2:05:08<36:39, 5.25it/s] 76%|███████▌ | 36465/48008 [2:05:08<36:21, 5.29it/s] 76%|███████▌ | 36466/48008 [2:05:08<36:31, 5.27it/s] 76%|███████▌ | 36467/48008 [2:05:09<37:20, 5.15it/s] 76%|███████▌ | 36468/48008 [2:05:09<37:54, 5.07it/s] 76%|███████▌ | 36469/48008 [2:05:09<37:12, 5.17it/s] 76%|███████▌ | 36470/48008 [2:05:09<37:05, 5.18it/s] 76%|███████▌ | 36471/48008 [2:05:09<37:33, 5.12it/s] 76%|███████▌ | 36472/48008 [2:05:10<37:11, 5.17it/s] 76%|███████▌ | 36473/48008 [2:05:10<37:52, 5.07it/s] 76%|███████▌ | 36474/48008 [2:05:10<37:37, 5.11it/s] 76%|███████▌ | 36475/48008 [2:05:10<37:24, 5.14it/s] 76%|███████▌ | 36476/48008 [2:05:10<37:37, 5.11it/s] 76%|███████▌ | 36477/48008 [2:05:11<38:06, 5.04it/s] 76%|███████▌ | 36478/48008 [2:05:11<37:26, 5.13it/s] 76%|███████▌ | 36479/48008 [2:05:11<40:06, 4.79it/s] 76%|███████▌ | 36480/48008 [2:05:11<39:06, 4.91it/s] 76%|███████▌ | 36481/48008 [2:05:11<38:22, 5.01it/s] 76%|███████▌ | 36482/48008 [2:05:12<40:55, 4.69it/s] 76%|███████▌ | 36483/48008 [2:05:12<40:50, 4.70it/s] 76%|███████▌ | 36484/48008 [2:05:12<39:55, 4.81it/s] 76%|███████▌ | 36485/48008 [2:05:12<38:50, 4.94it/s] 76%|███████▌ | 36486/48008 [2:05:12<38:19, 5.01it/s] 76%|███████▌ | 36487/48008 [2:05:13<38:15, 5.02it/s] 76%|███████▌ | 36488/48008 [2:05:13<38:16, 5.02it/s] 76%|███████▌ | 36489/48008 [2:05:13<38:13, 5.02it/s] 76%|███████▌ | 36490/48008 [2:05:13<37:39, 5.10it/s] 76%|███████▌ | 36491/48008 [2:05:13<37:46, 5.08it/s] 76%|███████▌ | 36492/48008 [2:05:14<37:28, 5.12it/s] 76%|███████▌ | 36493/48008 [2:05:14<37:25, 5.13it/s] 76%|███████▌ | 36494/48008 [2:05:14<38:00, 5.05it/s] 76%|███████▌ | 36495/48008 [2:05:14<37:37, 5.10it/s] 76%|███████▌ | 36496/48008 [2:05:14<37:42, 5.09it/s] 76%|███████▌ | 36497/48008 [2:05:15<38:11, 5.02it/s] 76%|███████▌ | 36498/48008 [2:05:15<37:40, 5.09it/s] 76%|███████▌ | 36499/48008 [2:05:15<38:33, 4.97it/s] 76%|███████▌ | 36500/48008 [2:05:15<37:44, 5.08it/s] 76%|███████▌ | 36500/48008 [2:05:15<37:44, 5.08it/s]{'loss': 0.3011, 'grad_norm': 0.56640625, 'learning_rate': 4.794617563739377e-05, 'epoch': 0.76} 76%|███████▌ | 36501/48008 [2:05:15<37:28, 5.12it/s] 76%|███████▌ | 36502/48008 [2:05:16<37:13, 5.15it/s] 76%|███████▌ | 36503/48008 [2:05:16<36:48, 5.21it/s] 76%|███████▌ | 36504/48008 [2:05:16<36:40, 5.23it/s] 76%|███████▌ | 36505/48008 [2:05:16<36:59, 5.18it/s] 76%|███████▌ | 36506/48008 [2:05:16<36:51, 5.20it/s] 76%|███████▌ | 36507/48008 [2:05:16<37:22, 5.13it/s] 76%|███████▌ | 36508/48008 [2:05:17<37:49, 5.07it/s] 76%|███████▌ | 36509/48008 [2:05:17<38:15, 5.01it/s] 76%|███████▌ | 36510/48008 [2:05:17<37:40, 5.09it/s] 76%|███████▌ | 36511/48008 [2:05:17<37:14, 5.14it/s] 76%|███████▌ | 36512/48008 [2:05:17<36:50, 5.20it/s] 76%|███████▌ | 36513/48008 [2:05:18<36:54, 5.19it/s] 76%|███████▌ | 36514/48008 [2:05:18<36:46, 5.21it/s] 76%|███████▌ | 36515/48008 [2:05:18<36:41, 5.22it/s] 76%|███████▌ | 36516/48008 [2:05:18<39:19, 4.87it/s] 76%|███████▌ | 36517/48008 [2:05:18<38:52, 4.93it/s] 76%|███████▌ | 36518/48008 [2:05:19<38:33, 4.97it/s] 76%|███████▌ | 36519/48008 [2:05:19<38:18, 5.00it/s] 76%|███████▌ | 36520/48008 [2:05:19<37:50, 5.06it/s] 76%|███████▌ | 36521/48008 [2:05:19<37:30, 5.10it/s] 76%|███████▌ | 36522/48008 [2:05:19<37:00, 5.17it/s] 76%|███████▌ | 36523/48008 [2:05:20<37:28, 5.11it/s] 76%|███████▌ | 36524/48008 [2:05:20<37:09, 5.15it/s] 76%|███████▌ | 36525/48008 [2:05:20<37:30, 5.10it/s] 76%|███████▌ | 36526/48008 [2:05:20<37:21, 5.12it/s] 76%|███████▌ | 36527/48008 [2:05:20<37:13, 5.14it/s] 76%|███████▌ | 36528/48008 [2:05:21<37:06, 5.16it/s] 76%|███████▌ | 36529/48008 [2:05:21<37:41, 5.08it/s] 76%|███████▌ | 36530/48008 [2:05:21<37:30, 5.10it/s] 76%|███████▌ | 36531/48008 [2:05:21<37:15, 5.13it/s] 76%|███████▌ | 36532/48008 [2:05:21<37:06, 5.16it/s] 76%|███████▌ | 36533/48008 [2:05:22<36:30, 5.24it/s] 76%|███████▌ | 36534/48008 [2:05:22<36:48, 5.20it/s] 76%|███████▌ | 36535/48008 [2:05:22<36:37, 5.22it/s] 76%|███████▌ | 36536/48008 [2:05:22<36:54, 5.18it/s] 76%|███████▌ | 36537/48008 [2:05:22<37:03, 5.16it/s] 76%|███████▌ | 36538/48008 [2:05:23<36:45, 5.20it/s] 76%|███████▌ | 36539/48008 [2:05:23<37:31, 5.09it/s] 76%|███████▌ | 36540/48008 [2:05:23<37:16, 5.13it/s] 76%|███████▌ | 36541/48008 [2:05:23<37:03, 5.16it/s] 76%|███████▌ | 36542/48008 [2:05:23<39:33, 4.83it/s] 76%|███████▌ | 36543/48008 [2:05:24<39:04, 4.89it/s] 76%|███████▌ | 36544/48008 [2:05:24<38:34, 4.95it/s] 76%|███████▌ | 36545/48008 [2:05:24<37:54, 5.04it/s] 76%|███████▌ | 36546/48008 [2:05:24<37:33, 5.09it/s] 76%|███████▌ | 36547/48008 [2:05:24<37:15, 5.13it/s] 76%|███████▌ | 36548/48008 [2:05:25<37:27, 5.10it/s] 76%|███████▌ | 36549/48008 [2:05:25<37:06, 5.15it/s] 76%|███████▌ | 36550/48008 [2:05:25<36:52, 5.18it/s] {'loss': 0.2991, 'grad_norm': 0.1494140625, 'learning_rate': 4.7737877020496584e-05, 'epoch': 0.76} 76%|███████▌ | 36550/48008 [2:05:25<36:52, 5.18it/s] 76%|███████▌ | 36551/48008 [2:05:25<36:43, 5.20it/s] 76%|███████▌ | 36552/48008 [2:05:25<36:40, 5.21it/s] 76%|███████▌ | 36553/48008 [2:05:26<37:16, 5.12it/s] 76%|███████▌ | 36554/48008 [2:05:26<37:08, 5.14it/s] 76%|███████▌ | 36555/48008 [2:05:26<37:13, 5.13it/s] 76%|███████▌ | 36556/48008 [2:05:26<37:05, 5.15it/s] 76%|███████▌ | 36557/48008 [2:05:26<37:11, 5.13it/s] 76%|███████▌ | 36558/48008 [2:05:26<37:17, 5.12it/s] 76%|███████▌ | 36559/48008 [2:05:27<37:13, 5.13it/s] 76%|███████▌ | 36560/48008 [2:05:27<37:07, 5.14it/s] 76%|███████▌ | 36561/48008 [2:05:27<36:58, 5.16it/s] 76%|███████▌ | 36562/48008 [2:05:27<39:23, 4.84it/s] 76%|███████▌ | 36563/48008 [2:05:27<38:36, 4.94it/s] 76%|███████▌ | 36564/48008 [2:05:28<40:42, 4.69it/s] 76%|███████▌ | 36565/48008 [2:05:28<42:03, 4.54it/s] 76%|███████▌ | 36566/48008 [2:05:28<40:07, 4.75it/s] 76%|███████▌ | 36567/48008 [2:05:28<38:46, 4.92it/s] 76%|███████▌ | 36568/48008 [2:05:29<37:53, 5.03it/s] 76%|███████▌ | 36569/48008 [2:05:29<38:20, 4.97it/s] 76%|███████▌ | 36570/48008 [2:05:29<37:40, 5.06it/s] 76%|███████▌ | 36571/48008 [2:05:29<37:04, 5.14it/s] 76%|███████▌ | 36572/48008 [2:05:29<36:52, 5.17it/s] 76%|███████▌ | 36573/48008 [2:05:29<36:47, 5.18it/s] 76%|███████▌ | 36574/48008 [2:05:30<36:43, 5.19it/s] 76%|███████▌ | 36575/48008 [2:05:30<36:46, 5.18it/s] 76%|███████▌ | 36576/48008 [2:05:30<36:41, 5.19it/s] 76%|███████▌ | 36577/48008 [2:05:30<36:42, 5.19it/s] 76%|███████▌ | 36578/48008 [2:05:30<36:14, 5.26it/s] 76%|███████▌ | 36579/48008 [2:05:31<36:18, 5.25it/s] 76%|███████▌ | 36580/48008 [2:05:31<47:32, 4.01it/s] 76%|███████▌ | 36581/48008 [2:05:31<43:52, 4.34it/s] 76%|███████▌ | 36582/48008 [2:05:31<42:25, 4.49it/s] 76%|███████▌ | 36583/48008 [2:05:32<40:22, 4.72it/s] 76%|███████▌ | 36584/48008 [2:05:32<39:09, 4.86it/s] 76%|███████▌ | 36585/48008 [2:05:32<38:30, 4.94it/s] 76%|███████▌ | 36586/48008 [2:05:32<37:39, 5.06it/s] 76%|███████▌ | 36587/48008 [2:05:32<37:08, 5.13it/s] 76%|███████▌ | 36588/48008 [2:05:33<39:36, 4.81it/s] 76%|███████▌ | 36589/48008 [2:05:33<39:20, 4.84it/s] 76%|███████▌ | 36590/48008 [2:05:33<38:19, 4.97it/s] 76%|███████▌ | 36591/48008 [2:05:33<37:38, 5.05it/s] 76%|███████▌ | 36592/48008 [2:05:33<37:51, 5.03it/s] 76%|███████▌ | 36593/48008 [2:05:34<37:22, 5.09it/s] 76%|███████▌ | 36594/48008 [2:05:34<36:50, 5.16it/s] 76%|███████▌ | 36595/48008 [2:05:34<36:43, 5.18it/s] 76%|███████▌ | 36596/48008 [2:05:34<36:46, 5.17it/s] 76%|███████▌ | 36597/48008 [2:05:34<36:37, 5.19it/s] 76%|███████▌ | 36598/48008 [2:05:35<36:08, 5.26it/s] 76%|███████▌ | 36599/48008 [2:05:35<36:07, 5.26it/s] 76%|███████▌ | 36600/48008 [2:05:35<36:05, 5.27it/s] {'loss': 0.3009, 'grad_norm': 0.197265625, 'learning_rate': 4.75295784035994e-05, 'epoch': 0.76} 76%|███████▌ | 36600/48008 [2:05:35<36:05, 5.27it/s] 76%|███████▌ | 36601/48008 [2:05:35<36:28, 5.21it/s] 76%|███████▌ | 36602/48008 [2:05:35<36:25, 5.22it/s] 76%|███████▌ | 36603/48008 [2:05:35<36:19, 5.23it/s] 76%|███████▌ | 36604/48008 [2:05:36<37:23, 5.08it/s] 76%|███████▌ | 36605/48008 [2:05:36<48:13, 3.94it/s] 76%|███████▌ | 36606/48008 [2:05:36<44:24, 4.28it/s] 76%|███████▋ | 36607/48008 [2:05:36<41:36, 4.57it/s] 76%|███████▋ | 36608/48008 [2:05:37<40:12, 4.73it/s] 76%|███████▋ | 36609/48008 [2:05:37<39:41, 4.79it/s] 76%|███████▋ | 36610/48008 [2:05:37<38:39, 4.91it/s] 76%|███████▋ | 36611/48008 [2:05:37<49:01, 3.88it/s] 76%|███████▋ | 36612/48008 [2:05:38<45:05, 4.21it/s] 76%|███████▋ | 36613/48008 [2:05:38<42:19, 4.49it/s] 76%|███████▋ | 36614/48008 [2:05:38<51:35, 3.68it/s] 76%|███████▋ | 36615/48008 [2:05:38<47:03, 4.04it/s] 76%|███████▋ | 36616/48008 [2:05:39<43:50, 4.33it/s] 76%|███████▋ | 36617/48008 [2:05:39<42:20, 4.48it/s] 76%|███████▋ | 36618/48008 [2:05:39<40:33, 4.68it/s] 76%|███████▋ | 36619/48008 [2:05:39<39:11, 4.84it/s] 76%|███████▋ | 36620/48008 [2:05:39<39:00, 4.86it/s] 76%|███████▋ | 36621/48008 [2:05:40<38:52, 4.88it/s] 76%|███████▋ | 36622/48008 [2:05:40<49:13, 3.85it/s] 76%|███████▋ | 36623/48008 [2:05:40<45:11, 4.20it/s] 76%|███████▋ | 36624/48008 [2:05:40<42:43, 4.44it/s] 76%|███████▋ | 36625/48008 [2:05:41<40:52, 4.64it/s] 76%|███████▋ | 36626/48008 [2:05:41<39:32, 4.80it/s] 76%|███████▋ | 36627/48008 [2:05:41<38:14, 4.96it/s] 76%|███████▋ | 36628/48008 [2:05:41<48:43, 3.89it/s] 76%|███████▋ | 36629/48008 [2:05:41<45:32, 4.16it/s] 76%|███████▋ | 36630/48008 [2:05:42<42:44, 4.44it/s] 76%|███████▋ | 36631/48008 [2:05:42<41:48, 4.53it/s] 76%|███████▋ | 36632/48008 [2:05:42<39:44, 4.77it/s] 76%|███████▋ | 36633/48008 [2:05:42<38:40, 4.90it/s] 76%|███████▋ | 36634/48008 [2:05:42<37:37, 5.04it/s] 76%|███████▋ | 36635/48008 [2:05:43<37:28, 5.06it/s] 76%|███████▋ | 36636/48008 [2:05:43<37:48, 5.01it/s] 76%|███████▋ | 36637/48008 [2:05:43<37:28, 5.06it/s] 76%|███████▋ | 36638/48008 [2:05:43<37:05, 5.11it/s] 76%|███████▋ | 36639/48008 [2:05:43<36:55, 5.13it/s] 76%|███████▋ | 36640/48008 [2:05:44<36:43, 5.16it/s] 76%|███████▋ | 36641/48008 [2:05:44<37:35, 5.04it/s] 76%|███████▋ | 36642/48008 [2:05:44<37:05, 5.11it/s] 76%|███████▋ | 36643/48008 [2:05:44<37:00, 5.12it/s] 76%|███████▋ | 36644/48008 [2:05:44<37:31, 5.05it/s] 76%|███████▋ | 36645/48008 [2:05:45<36:51, 5.14it/s] 76%|███████▋ | 36646/48008 [2:05:45<37:15, 5.08it/s] 76%|███████▋ | 36647/48008 [2:05:45<36:57, 5.12it/s] 76%|███████▋ | 36648/48008 [2:05:45<37:47, 5.01it/s] 76%|███████▋ | 36649/48008 [2:05:45<39:47, 4.76it/s] 76%|███████▋ | 36650/48008 [2:05:46<38:40, 4.90it/s] {'loss': 0.2985, 'grad_norm': 0.24609375, 'learning_rate': 4.732127978670222e-05, 'epoch': 0.76} 76%|███████▋ | 36650/48008 [2:05:46<38:40, 4.90it/s] 76%|███████▋ | 36651/48008 [2:05:46<37:35, 5.03it/s] 76%|███████▋ | 36652/48008 [2:05:46<37:16, 5.08it/s] 76%|███████▋ | 36653/48008 [2:05:46<37:11, 5.09it/s] 76%|███████▋ | 36654/48008 [2:05:46<36:52, 5.13it/s] 76%|███████▋ | 36655/48008 [2:05:47<36:33, 5.17it/s] 76%|███████▋ | 36656/48008 [2:05:47<36:09, 5.23it/s] 76%|███████▋ | 36657/48008 [2:05:47<35:53, 5.27it/s] 76%|███████▋ | 36658/48008 [2:05:47<36:45, 5.15it/s] 76%|███████▋ | 36659/48008 [2:05:47<36:34, 5.17it/s] 76%|███████▋ | 36660/48008 [2:05:48<36:51, 5.13it/s] 76%|███████▋ | 36661/48008 [2:05:48<36:40, 5.16it/s] 76%|███████▋ | 36662/48008 [2:05:48<37:06, 5.10it/s] 76%|███████▋ | 36663/48008 [2:05:48<36:25, 5.19it/s] 76%|███████▋ | 36664/48008 [2:05:48<36:13, 5.22it/s] 76%|███████▋ | 36665/48008 [2:05:49<36:09, 5.23it/s] 76%|███████▋ | 36666/48008 [2:05:49<37:11, 5.08it/s] 76%|███████▋ | 36667/48008 [2:05:49<37:05, 5.10it/s] 76%|███████▋ | 36668/48008 [2:05:49<37:28, 5.04it/s] 76%|███████▋ | 36669/48008 [2:05:49<37:08, 5.09it/s] 76%|███████▋ | 36670/48008 [2:05:50<37:25, 5.05it/s] 76%|███████▋ | 36671/48008 [2:05:50<36:48, 5.13it/s] 76%|███████▋ | 36672/48008 [2:05:50<37:01, 5.10it/s] 76%|███████▋ | 36673/48008 [2:05:50<36:50, 5.13it/s] 76%|███████▋ | 36674/48008 [2:05:50<36:43, 5.14it/s] 76%|███████▋ | 36675/48008 [2:05:50<37:19, 5.06it/s] 76%|███████▋ | 36676/48008 [2:05:51<36:43, 5.14it/s] 76%|███████▋ | 36677/48008 [2:05:51<37:10, 5.08it/s] 76%|███████▋ | 36678/48008 [2:05:51<36:53, 5.12it/s] 76%|███████▋ | 36679/48008 [2:05:51<36:20, 5.19it/s] 76%|███████▋ | 36680/48008 [2:05:51<36:09, 5.22it/s] 76%|███████▋ | 36681/48008 [2:05:52<36:56, 5.11it/s] 76%|███████▋ | 36682/48008 [2:05:52<36:38, 5.15it/s] 76%|███████▋ | 36683/48008 [2:05:52<37:03, 5.09it/s] 76%|███████▋ | 36684/48008 [2:05:52<37:20, 5.06it/s] 76%|███████▋ | 36685/48008 [2:05:52<37:03, 5.09it/s] 76%|███████▋ | 36686/48008 [2:05:53<37:42, 5.00it/s] 76%|███████▋ | 36687/48008 [2:05:53<37:02, 5.09it/s] 76%|███████▋ | 36688/48008 [2:05:53<36:44, 5.14it/s] 76%|███████▋ | 36689/48008 [2:05:53<42:57, 4.39it/s] 76%|███████▋ | 36690/48008 [2:05:54<40:52, 4.61it/s] 76%|███████▋ | 36691/48008 [2:05:54<39:10, 4.82it/s] 76%|███████▋ | 36692/48008 [2:05:54<38:06, 4.95it/s] 76%|███████▋ | 36693/48008 [2:05:54<37:30, 5.03it/s] 76%|███████▋ | 36694/48008 [2:05:54<48:05, 3.92it/s] 76%|███████▋ | 36695/48008 [2:05:55<44:29, 4.24it/s] 76%|███████▋ | 36696/48008 [2:05:55<41:44, 4.52it/s] 76%|███████▋ | 36697/48008 [2:05:55<40:03, 4.71it/s] 76%|███████▋ | 36698/48008 [2:05:55<38:36, 4.88it/s] 76%|███████▋ | 36699/48008 [2:05:55<38:52, 4.85it/s] 76%|███████▋ | 36700/48008 [2:05:56<38:00, 4.96it/s] {'loss': 0.3001, 'grad_norm': 0.267578125, 'learning_rate': 4.7112981169805036e-05, 'epoch': 0.76} 76%|███████▋ | 36700/48008 [2:05:56<38:00, 4.96it/s] 76%|███████▋ | 36701/48008 [2:05:56<37:18, 5.05it/s] 76%|███████▋ | 36702/48008 [2:05:56<36:49, 5.12it/s] 76%|███████▋ | 36703/48008 [2:05:56<35:57, 5.24it/s] 76%|███████▋ | 36704/48008 [2:05:56<36:33, 5.15it/s] 76%|███████▋ | 36705/48008 [2:05:57<35:56, 5.24it/s] 76%|███████▋ | 36706/48008 [2:05:57<35:47, 5.26it/s] 76%|███████▋ | 36707/48008 [2:05:57<35:50, 5.25it/s] 76%|███████▋ | 36708/48008 [2:05:57<35:51, 5.25it/s] 76%|███████▋ | 36709/48008 [2:05:57<36:26, 5.17it/s] 76%|███████▋ | 36710/48008 [2:05:58<36:05, 5.22it/s] 76%|███████▋ | 36711/48008 [2:05:58<36:00, 5.23it/s] 76%|███████▋ | 36712/48008 [2:05:58<36:10, 5.20it/s] 76%|███████▋ | 36713/48008 [2:05:58<36:27, 5.16it/s] 76%|███████▋ | 36714/48008 [2:05:58<36:13, 5.20it/s] 76%|███████▋ | 36715/48008 [2:05:59<36:29, 5.16it/s] 76%|███████▋ | 36716/48008 [2:05:59<36:27, 5.16it/s] 76%|███████▋ | 36717/48008 [2:05:59<36:12, 5.20it/s] 76%|███████▋ | 36718/48008 [2:05:59<38:42, 4.86it/s] 76%|███████▋ | 36719/48008 [2:05:59<37:52, 4.97it/s] 76%|███████▋ | 36720/48008 [2:06:00<37:38, 5.00it/s] 76%|███████▋ | 36721/48008 [2:06:00<37:06, 5.07it/s] 76%|███████▋ | 36722/48008 [2:06:00<47:45, 3.94it/s] 76%|███████▋ | 36723/48008 [2:06:00<44:01, 4.27it/s] 76%|███████▋ | 36724/48008 [2:06:00<41:30, 4.53it/s] 76%|███████▋ | 36725/48008 [2:06:01<39:43, 4.73it/s] 76%|███████▋ | 36726/48008 [2:06:01<38:25, 4.89it/s] 77%|███████▋ | 36727/48008 [2:06:01<37:32, 5.01it/s] 77%|███████▋ | 36728/48008 [2:06:01<37:22, 5.03it/s] 77%|███████▋ | 36729/48008 [2:06:01<36:57, 5.09it/s] 77%|███████▋ | 36730/48008 [2:06:02<36:35, 5.14it/s] 77%|███████▋ | 36731/48008 [2:06:02<36:20, 5.17it/s] 77%|███████▋ | 36732/48008 [2:06:02<36:31, 5.15it/s] 77%|███████▋ | 36733/48008 [2:06:02<37:12, 5.05it/s] 77%|███████▋ | 36734/48008 [2:06:02<36:46, 5.11it/s] 77%|███████▋ | 36735/48008 [2:06:03<37:39, 4.99it/s] 77%|███████▋ | 36736/48008 [2:06:03<37:23, 5.02it/s] 77%|███████▋ | 36737/48008 [2:06:03<36:58, 5.08it/s] 77%|███████▋ | 36738/48008 [2:06:03<37:36, 5.00it/s] 77%|███████▋ | 36739/48008 [2:06:03<36:45, 5.11it/s] 77%|███████▋ | 36740/48008 [2:06:04<37:36, 4.99it/s] 77%|███████▋ | 36741/48008 [2:06:04<37:08, 5.06it/s] 77%|███████▋ | 36742/48008 [2:06:04<37:50, 4.96it/s] 77%|███████▋ | 36743/48008 [2:06:04<37:15, 5.04it/s] 77%|███████▋ | 36744/48008 [2:06:04<36:49, 5.10it/s] 77%|███████▋ | 36745/48008 [2:06:05<37:09, 5.05it/s] 77%|███████▋ | 36746/48008 [2:06:05<47:44, 3.93it/s] 77%|███████▋ | 36747/48008 [2:06:05<44:30, 4.22it/s] 77%|███████▋ | 36748/48008 [2:06:05<41:38, 4.51it/s] 77%|███████▋ | 36749/48008 [2:06:06<39:58, 4.69it/s] 77%|███████▋ | 36750/48008 [2:06:06<38:57, 4.82it/s] {'loss': 0.3002, 'grad_norm': 0.208984375, 'learning_rate': 4.6904682552907854e-05, 'epoch': 0.77} 77%|███████▋ | 36750/48008 [2:06:06<38:57, 4.82it/s] 77%|███████▋ | 36751/48008 [2:06:06<38:22, 4.89it/s] 77%|███████▋ | 36752/48008 [2:06:06<37:20, 5.02it/s] 77%|███████▋ | 36753/48008 [2:06:06<36:51, 5.09it/s] 77%|███████▋ | 36754/48008 [2:06:07<36:19, 5.16it/s] 77%|███████▋ | 36755/48008 [2:06:07<36:14, 5.18it/s] 77%|███████▋ | 36756/48008 [2:06:07<35:50, 5.23it/s] 77%|███████▋ | 36757/48008 [2:06:07<36:04, 5.20it/s] 77%|███████▋ | 36758/48008 [2:06:07<35:55, 5.22it/s] 77%|███████▋ | 36759/48008 [2:06:07<35:47, 5.24it/s] 77%|███████▋ | 36760/48008 [2:06:08<35:44, 5.25it/s] 77%|███████▋ | 36761/48008 [2:06:08<36:24, 5.15it/s] 77%|███████▋ | 36762/48008 [2:06:08<36:49, 5.09it/s] 77%|███████▋ | 36763/48008 [2:06:08<36:25, 5.15it/s] 77%|███████▋ | 36764/48008 [2:06:08<35:49, 5.23it/s] 77%|███████▋ | 36765/48008 [2:06:09<42:18, 4.43it/s] 77%|███████▋ | 36766/48008 [2:06:09<40:08, 4.67it/s] 77%|███████▋ | 36767/48008 [2:06:09<38:37, 4.85it/s] 77%|███████▋ | 36768/48008 [2:06:09<37:45, 4.96it/s] 77%|███████▋ | 36769/48008 [2:06:09<37:09, 5.04it/s] 77%|███████▋ | 36770/48008 [2:06:10<37:14, 5.03it/s] 77%|███████▋ | 36771/48008 [2:06:10<36:59, 5.06it/s] 77%|███████▋ | 36772/48008 [2:06:10<36:36, 5.11it/s] 77%|███████▋ | 36773/48008 [2:06:10<35:59, 5.20it/s] 77%|███████▋ | 36774/48008 [2:06:10<36:15, 5.16it/s] 77%|███████▋ | 36775/48008 [2:06:11<38:42, 4.84it/s] 77%|███████▋ | 36776/48008 [2:06:11<37:56, 4.93it/s] 77%|███████▋ | 36777/48008 [2:06:11<37:53, 4.94it/s] 77%|███████▋ | 36778/48008 [2:06:11<37:15, 5.02it/s] 77%|███████▋ | 36779/48008 [2:06:11<36:47, 5.09it/s] 77%|███████▋ | 36780/48008 [2:06:12<36:42, 5.10it/s] 77%|███████▋ | 36781/48008 [2:06:12<36:04, 5.19it/s] 77%|███████▋ | 36782/48008 [2:06:12<35:59, 5.20it/s] 77%|███████▋ | 36783/48008 [2:06:12<35:52, 5.22it/s] 77%|███████▋ | 36784/48008 [2:06:12<36:22, 5.14it/s] 77%|███████▋ | 36785/48008 [2:06:13<35:56, 5.21it/s] 77%|███████▋ | 36786/48008 [2:06:13<35:36, 5.25it/s] 77%|███████▋ | 36787/48008 [2:06:13<35:38, 5.25it/s] 77%|███████▋ | 36788/48008 [2:06:13<35:37, 5.25it/s] 77%|███████▋ | 36789/48008 [2:06:13<35:30, 5.27it/s] 77%|███████▋ | 36790/48008 [2:06:14<36:37, 5.10it/s] 77%|███████▋ | 36791/48008 [2:06:14<36:25, 5.13it/s] 77%|███████▋ | 36792/48008 [2:06:14<37:14, 5.02it/s] 77%|███████▋ | 36793/48008 [2:06:14<36:48, 5.08it/s] 77%|███████▋ | 36794/48008 [2:06:14<36:25, 5.13it/s] 77%|███████▋ | 36795/48008 [2:06:15<36:04, 5.18it/s] 77%|███████▋ | 36796/48008 [2:06:15<35:59, 5.19it/s] 77%|███████▋ | 36797/48008 [2:06:15<35:40, 5.24it/s] 77%|███████▋ | 36798/48008 [2:06:15<35:43, 5.23it/s] 77%|███████▋ | 36799/48008 [2:06:15<36:23, 5.13it/s] 77%|███████▋ | 36800/48008 [2:06:16<36:12, 5.16it/s] {'loss': 0.3015, 'grad_norm': 0.2470703125, 'learning_rate': 4.669638393601067e-05, 'epoch': 0.77} 77%|███████▋ | 36800/48008 [2:06:16<36:12, 5.16it/s] 77%|███████▋ | 36801/48008 [2:06:16<36:05, 5.18it/s] 77%|███████▋ | 36802/48008 [2:06:16<35:56, 5.20it/s] 77%|███████▋ | 36803/48008 [2:06:16<36:03, 5.18it/s] 77%|███████▋ | 36804/48008 [2:06:16<35:56, 5.20it/s] 77%|███████▋ | 36805/48008 [2:06:16<35:39, 5.24it/s] 77%|███████▋ | 36806/48008 [2:06:17<35:27, 5.27it/s] 77%|███████▋ | 36807/48008 [2:06:17<35:19, 5.28it/s] 77%|███████▋ | 36808/48008 [2:06:17<36:02, 5.18it/s] 77%|███████▋ | 36809/48008 [2:06:17<36:38, 5.09it/s] 77%|███████▋ | 36810/48008 [2:06:17<36:57, 5.05it/s] 77%|███████▋ | 36811/48008 [2:06:18<36:59, 5.04it/s] 77%|███████▋ | 36812/48008 [2:06:18<37:09, 5.02it/s] 77%|███████▋ | 36813/48008 [2:06:18<37:19, 5.00it/s] 77%|███████▋ | 36814/48008 [2:06:18<43:12, 4.32it/s] 77%|███████▋ | 36815/48008 [2:06:19<40:55, 4.56it/s] 77%|███████▋ | 36816/48008 [2:06:19<39:41, 4.70it/s] 77%|███████▋ | 36817/48008 [2:06:19<38:28, 4.85it/s] 77%|███████▋ | 36818/48008 [2:06:19<37:30, 4.97it/s] 77%|███████▋ | 36819/48008 [2:06:19<36:39, 5.09it/s] 77%|███████▋ | 36820/48008 [2:06:20<36:06, 5.16it/s] 77%|███████▋ | 36821/48008 [2:06:20<35:53, 5.19it/s] 77%|███████▋ | 36822/48008 [2:06:20<35:58, 5.18it/s] 77%|███████▋ | 36823/48008 [2:06:20<35:49, 5.20it/s] 77%|███████▋ | 36824/48008 [2:06:20<35:49, 5.20it/s] 77%|███████▋ | 36825/48008 [2:06:20<36:23, 5.12it/s] 77%|███████▋ | 36826/48008 [2:06:21<36:19, 5.13it/s] 77%|███████▋ | 36827/48008 [2:06:21<36:12, 5.15it/s] 77%|███████▋ | 36828/48008 [2:06:21<35:59, 5.18it/s] 77%|███████▋ | 36829/48008 [2:06:21<35:52, 5.19it/s] 77%|███████▋ | 36830/48008 [2:06:21<35:38, 5.23it/s] 77%|███████▋ | 36831/48008 [2:06:22<35:35, 5.24it/s] 77%|███████▋ | 36832/48008 [2:06:22<36:35, 5.09it/s] 77%|███████▋ | 36833/48008 [2:06:22<42:42, 4.36it/s] 77%|███████▋ | 36834/48008 [2:06:22<40:28, 4.60it/s] 77%|███████▋ | 36835/48008 [2:06:23<39:14, 4.75it/s] 77%|███████▋ | 36836/48008 [2:06:23<38:51, 4.79it/s] 77%|███████▋ | 36837/48008 [2:06:23<38:42, 4.81it/s] 77%|███████▋ | 36838/48008 [2:06:23<38:07, 4.88it/s] 77%|███████▋ | 36839/48008 [2:06:23<38:04, 4.89it/s] 77%|███████▋ | 36840/48008 [2:06:24<43:50, 4.25it/s] 77%|███████▋ | 36841/48008 [2:06:24<41:56, 4.44it/s] 77%|███████▋ | 36842/48008 [2:06:24<39:52, 4.67it/s] 77%|███████▋ | 36843/48008 [2:06:24<38:43, 4.80it/s] 77%|███████▋ | 36844/48008 [2:06:24<37:48, 4.92it/s] 77%|███████▋ | 36845/48008 [2:06:25<37:07, 5.01it/s] 77%|███████▋ | 36846/48008 [2:06:25<37:05, 5.02it/s] 77%|███████▋ | 36847/48008 [2:06:25<39:14, 4.74it/s] 77%|███████▋ | 36848/48008 [2:06:25<38:06, 4.88it/s] 77%|███████▋ | 36849/48008 [2:06:25<37:11, 5.00it/s] 77%|███████▋ | 36850/48008 [2:06:26<36:43, 5.06it/s] {'loss': 0.3002, 'grad_norm': 0.15234375, 'learning_rate': 4.648808531911348e-05, 'epoch': 0.77} 77%|███████▋ | 36850/48008 [2:06:26<36:43, 5.06it/s] 77%|███████▋ | 36851/48008 [2:06:26<36:59, 5.03it/s] 77%|███████▋ | 36852/48008 [2:06:26<36:12, 5.14it/s] 77%|███████▋ | 36853/48008 [2:06:26<36:12, 5.14it/s] 77%|███████▋ | 36854/48008 [2:06:26<36:08, 5.14it/s] 77%|███████▋ | 36855/48008 [2:06:27<36:20, 5.11it/s] 77%|███████▋ | 36856/48008 [2:06:27<1:14:45, 2.49it/s] 77%|███████▋ | 36857/48008 [2:06:28<1:02:57, 2.95it/s] 77%|███████▋ | 36858/48008 [2:06:28<54:39, 3.40it/s] 77%|███████▋ | 36859/48008 [2:06:28<49:13, 3.78it/s] 77%|███████▋ | 36860/48008 [2:06:28<55:57, 3.32it/s] 77%|███████▋ | 36861/48008 [2:06:29<49:46, 3.73it/s] 77%|███████▋ | 36862/48008 [2:06:29<46:07, 4.03it/s] 77%|███████▋ | 36863/48008 [2:06:29<43:14, 4.29it/s] 77%|███████▋ | 36864/48008 [2:06:30<1:35:57, 1.94it/s] 77%|███████▋ | 36865/48008 [2:06:30<1:18:02, 2.38it/s] 77%|███████▋ | 36866/48008 [2:06:31<1:04:57, 2.86it/s] 77%|███████▋ | 36867/48008 [2:06:31<56:08, 3.31it/s] 77%|███████▋ | 36868/48008 [2:06:31<49:45, 3.73it/s] 77%|███████▋ | 36869/48008 [2:06:31<45:26, 4.08it/s] 77%|███████▋ | 36870/48008 [2:06:31<42:13, 4.40it/s] 77%|███████▋ | 36871/48008 [2:06:32<39:51, 4.66it/s] 77%|███████▋ | 36872/48008 [2:06:32<38:40, 4.80it/s] 77%|███████▋ | 36873/48008 [2:06:32<37:40, 4.93it/s] 77%|███████▋ | 36874/48008 [2:06:32<37:57, 4.89it/s] 77%|███████▋ | 36875/48008 [2:06:32<37:15, 4.98it/s] 77%|███████▋ | 36876/48008 [2:06:33<39:19, 4.72it/s] 77%|███████▋ | 36877/48008 [2:06:33<38:01, 4.88it/s] 77%|███████▋ | 36878/48008 [2:06:33<37:16, 4.98it/s] 77%|███████▋ | 36879/48008 [2:06:33<36:37, 5.07it/s] 77%|███████▋ | 36880/48008 [2:06:33<36:22, 5.10it/s] 77%|███████▋ | 36881/48008 [2:06:34<36:02, 5.15it/s] 77%|███████▋ | 36882/48008 [2:06:34<35:55, 5.16it/s] 77%|███████▋ | 36883/48008 [2:06:34<35:48, 5.18it/s] 77%|███████▋ | 36884/48008 [2:06:34<35:57, 5.16it/s] 77%|███████▋ | 36885/48008 [2:06:34<35:47, 5.18it/s] 77%|███████▋ | 36886/48008 [2:06:34<35:17, 5.25it/s] 77%|███████▋ | 36887/48008 [2:06:35<46:10, 4.01it/s] 77%|███████▋ | 36888/48008 [2:06:35<42:44, 4.34it/s] 77%|███████▋ | 36889/48008 [2:06:35<40:30, 4.57it/s] 77%|███████▋ | 36890/48008 [2:06:35<39:19, 4.71it/s] 77%|███████▋ | 36891/48008 [2:06:36<39:06, 4.74it/s] 77%|███████▋ | 36892/48008 [2:06:36<37:50, 4.90it/s] 77%|███████▋ | 36893/48008 [2:06:36<37:40, 4.92it/s] 77%|███████▋ | 36894/48008 [2:06:36<37:16, 4.97it/s] 77%|███████▋ | 36895/48008 [2:06:36<37:17, 4.97it/s] 77%|███████▋ | 36896/48008 [2:06:37<36:34, 5.06it/s] 77%|███████▋ | 36897/48008 [2:06:37<36:29, 5.08it/s] 77%|███████▋ | 36898/48008 [2:06:37<36:04, 5.13it/s] 77%|███████▋ | 36899/48008 [2:06:37<36:17, 5.10it/s] 77%|███████▋ | 36900/48008 [2:06:37<36:02, 5.14it/s] {'loss': 0.3007, 'grad_norm': 0.216796875, 'learning_rate': 4.62797867022163e-05, 'epoch': 0.77} 77%|███████▋ | 36900/48008 [2:06:37<36:02, 5.14it/s] 77%|███████▋ | 36901/48008 [2:06:38<35:40, 5.19it/s] 77%|███████▋ | 36902/48008 [2:06:38<35:27, 5.22it/s] 77%|███████▋ | 36903/48008 [2:06:38<35:52, 5.16it/s] 77%|███████▋ | 36904/48008 [2:06:38<35:37, 5.19it/s] 77%|███████▋ | 36905/48008 [2:06:38<35:09, 5.26it/s] 77%|███████▋ | 36906/48008 [2:06:39<35:01, 5.28it/s] 77%|███████▋ | 36907/48008 [2:06:39<35:17, 5.24it/s] 77%|███████▋ | 36908/48008 [2:06:39<36:14, 5.10it/s] 77%|███████▋ | 36909/48008 [2:06:39<47:02, 3.93it/s] 77%|███████▋ | 36910/48008 [2:06:40<43:50, 4.22it/s] 77%|███████▋ | 36911/48008 [2:06:40<41:18, 4.48it/s] 77%|███████▋ | 36912/48008 [2:06:40<39:21, 4.70it/s] 77%|███████▋ | 36913/48008 [2:06:40<38:03, 4.86it/s] 77%|███████▋ | 36914/48008 [2:06:40<37:15, 4.96it/s] 77%|███████▋ | 36915/48008 [2:06:40<37:26, 4.94it/s] 77%|███████▋ | 36916/48008 [2:06:41<36:40, 5.04it/s] 77%|███████▋ | 36917/48008 [2:06:41<36:21, 5.08it/s] 77%|███████▋ | 36918/48008 [2:06:41<36:26, 5.07it/s] 77%|███████▋ | 36919/48008 [2:06:41<37:01, 4.99it/s] 77%|███████▋ | 36920/48008 [2:06:41<37:37, 4.91it/s] 77%|███████▋ | 36921/48008 [2:06:42<37:40, 4.90it/s] 77%|███████▋ | 36922/48008 [2:06:42<38:03, 4.86it/s] 77%|███████▋ | 36923/48008 [2:06:42<37:01, 4.99it/s] 77%|███████▋ | 36924/48008 [2:06:42<47:24, 3.90it/s] 77%|███████▋ | 36925/48008 [2:06:43<43:47, 4.22it/s] 77%|███████▋ | 36926/48008 [2:06:43<41:18, 4.47it/s] 77%|███████▋ | 36927/48008 [2:06:43<39:31, 4.67it/s] 77%|███████▋ | 36928/48008 [2:06:43<38:18, 4.82it/s] 77%|███████▋ | 36929/48008 [2:06:43<37:26, 4.93it/s] 77%|███████▋ | 36930/48008 [2:06:44<37:30, 4.92it/s] 77%|███████▋ | 36931/48008 [2:06:44<47:37, 3.88it/s] 77%|███████▋ | 36932/48008 [2:06:44<44:16, 4.17it/s] 77%|███████▋ | 36933/48008 [2:06:44<41:36, 4.44it/s] 77%|███████▋ | 36934/48008 [2:06:45<40:41, 4.54it/s] 77%|███████▋ | 36935/48008 [2:06:45<39:06, 4.72it/s] 77%|███████▋ | 36936/48008 [2:06:45<37:57, 4.86it/s] 77%|███████▋ | 36937/48008 [2:06:45<37:06, 4.97it/s] 77%|███████▋ | 36938/48008 [2:06:45<39:09, 4.71it/s] 77%|███████▋ | 36939/48008 [2:06:46<44:25, 4.15it/s] 77%|███████▋ | 36940/48008 [2:06:46<44:07, 4.18it/s] 77%|███████▋ | 36941/48008 [2:06:46<41:22, 4.46it/s] 77%|███████▋ | 36942/48008 [2:06:46<39:29, 4.67it/s] 77%|███████▋ | 36943/48008 [2:06:47<38:27, 4.80it/s] 77%|███████▋ | 36944/48008 [2:06:47<37:55, 4.86it/s] 77%|███████▋ | 36945/48008 [2:06:47<37:41, 4.89it/s] 77%|███████▋ | 36946/48008 [2:06:47<37:05, 4.97it/s] 77%|███████▋ | 36947/48008 [2:06:47<36:14, 5.09it/s] 77%|███████▋ | 36948/48008 [2:06:48<35:52, 5.14it/s] 77%|███████▋ | 36949/48008 [2:06:48<35:36, 5.18it/s] 77%|███████▋ | 36950/48008 [2:06:48<36:02, 5.11it/s] 77%|███████▋ | 36950/48008 [2:06:48<36:02, 5.11it/s]{'loss': 0.2993, 'grad_norm': 0.3125, 'learning_rate': 4.6071488085319116e-05, 'epoch': 0.77} 77%|███████▋ | 36951/48008 [2:06:48<35:46, 5.15it/s] 77%|███████▋ | 36952/48008 [2:06:48<35:30, 5.19it/s] 77%|███████▋ | 36953/48008 [2:06:48<35:22, 5.21it/s] 77%|███████▋ | 36954/48008 [2:06:49<37:54, 4.86it/s] 77%|███████▋ | 36955/48008 [2:06:49<36:51, 5.00it/s] 77%|███████▋ | 36956/48008 [2:06:49<36:22, 5.06it/s] 77%|███████▋ | 36957/48008 [2:06:49<35:59, 5.12it/s] 77%|███████▋ | 36958/48008 [2:06:49<35:25, 5.20it/s] 77%|███████▋ | 36959/48008 [2:06:50<35:18, 5.22it/s] 77%|███████▋ | 36960/48008 [2:06:50<35:02, 5.25it/s] 77%|███████▋ | 36961/48008 [2:06:50<35:44, 5.15it/s] 77%|███████▋ | 36962/48008 [2:06:50<35:29, 5.19it/s] 77%|███████▋ | 36963/48008 [2:06:50<35:25, 5.20it/s] 77%|███████▋ | 36964/48008 [2:06:51<35:06, 5.24it/s] 77%|███████▋ | 36965/48008 [2:06:51<35:27, 5.19it/s] 77%|███████▋ | 36966/48008 [2:06:51<41:46, 4.41it/s] 77%|███████▋ | 36967/48008 [2:06:51<39:25, 4.67it/s] 77%|███████▋ | 36968/48008 [2:06:52<38:43, 4.75it/s] 77%|███████▋ | 36969/48008 [2:06:52<48:25, 3.80it/s] 77%|███████▋ | 36970/48008 [2:06:52<44:48, 4.11it/s] 77%|███████▋ | 36971/48008 [2:06:52<41:55, 4.39it/s] 77%|███████▋ | 36972/48008 [2:06:52<39:48, 4.62it/s] 77%|███████▋ | 36973/48008 [2:06:53<39:02, 4.71it/s] 77%|███████▋ | 36974/48008 [2:06:53<37:49, 4.86it/s] 77%|███████▋ | 36975/48008 [2:06:53<38:19, 4.80it/s] 77%|███████▋ | 36976/48008 [2:06:53<37:11, 4.94it/s] 77%|███████▋ | 36977/48008 [2:06:53<36:29, 5.04it/s] 77%|███████▋ | 36978/48008 [2:06:54<35:49, 5.13it/s] 77%|███████▋ | 36979/48008 [2:06:54<35:17, 5.21it/s] 77%|███████▋ | 36980/48008 [2:06:54<35:44, 5.14it/s] 77%|███████▋ | 36981/48008 [2:06:54<35:30, 5.18it/s] 77%|███████▋ | 36982/48008 [2:06:54<35:20, 5.20it/s] 77%|███████▋ | 36983/48008 [2:06:55<35:36, 5.16it/s] 77%|███████▋ | 36984/48008 [2:06:55<35:27, 5.18it/s] 77%|███████▋ | 36985/48008 [2:06:55<35:21, 5.20it/s] 77%|███████▋ | 36986/48008 [2:06:55<35:15, 5.21it/s] 77%|███████▋ | 36987/48008 [2:06:55<35:00, 5.25it/s] 77%|███████▋ | 36988/48008 [2:06:56<34:48, 5.28it/s] 77%|███████▋ | 36989/48008 [2:06:56<34:52, 5.27it/s] 77%|███████▋ | 36990/48008 [2:06:56<35:29, 5.17it/s] 77%|███████▋ | 36991/48008 [2:06:56<35:56, 5.11it/s] 77%|███████▋ | 36992/48008 [2:06:56<36:40, 5.01it/s] 77%|███████▋ | 36993/48008 [2:06:57<36:47, 4.99it/s] 77%|███████▋ | 36994/48008 [2:06:57<36:00, 5.10it/s] 77%|███████▋ | 36995/48008 [2:06:57<35:25, 5.18it/s] 77%|███████▋ | 36996/48008 [2:06:57<34:54, 5.26it/s] 77%|███████▋ | 36997/48008 [2:06:57<36:01, 5.09it/s] 77%|███████▋ | 36998/48008 [2:06:58<35:26, 5.18it/s] 77%|███████▋ | 36999/48008 [2:06:58<35:16, 5.20it/s] 77%|███████▋ | 37000/48008 [2:06:58<35:55, 5.11it/s] {'loss': 0.3013, 'grad_norm': 0.38671875, 'learning_rate': 4.586318946842193e-05, 'epoch': 0.77} 77%|███████▋ | 37000/48008 [2:06:58<35:55, 5.11it/s] 77%|███████▋ | 37001/48008 [2:06:58<35:23, 5.18it/s] 77%|███████▋ | 37002/48008 [2:06:58<35:48, 5.12it/s] 77%|███████▋ | 37003/48008 [2:06:58<35:30, 5.17it/s] 77%|███████▋ | 37004/48008 [2:06:59<35:21, 5.19it/s] 77%|███████▋ | 37005/48008 [2:06:59<35:12, 5.21it/s] 77%|███████▋ | 37006/48008 [2:06:59<35:43, 5.13it/s] 77%|███████▋ | 37007/48008 [2:06:59<46:14, 3.97it/s] 77%|███████▋ | 37008/48008 [2:07:00<42:45, 4.29it/s] 77%|███████▋ | 37009/48008 [2:07:00<40:26, 4.53it/s] 77%|███████▋ | 37010/48008 [2:07:00<38:26, 4.77it/s] 77%|███████▋ | 37011/48008 [2:07:00<37:05, 4.94it/s] 77%|███████▋ | 37012/48008 [2:07:00<36:12, 5.06it/s] 77%|███████▋ | 37013/48008 [2:07:01<46:31, 3.94it/s] 77%|███████▋ | 37014/48008 [2:07:01<43:03, 4.25it/s] 77%|███████▋ | 37015/48008 [2:07:01<41:13, 4.44it/s] 77%|███████▋ | 37016/48008 [2:07:01<39:23, 4.65it/s] 77%|███████▋ | 37017/48008 [2:07:02<38:03, 4.81it/s] 77%|███████▋ | 37018/48008 [2:07:02<37:09, 4.93it/s] 77%|███████▋ | 37019/48008 [2:07:02<36:31, 5.01it/s] 77%|███████▋ | 37020/48008 [2:07:02<36:01, 5.08it/s] 77%|███████▋ | 37021/48008 [2:07:02<36:42, 4.99it/s] 77%|███████▋ | 37022/48008 [2:07:03<35:56, 5.09it/s] 77%|███████▋ | 37023/48008 [2:07:03<38:09, 4.80it/s] 77%|███████▋ | 37024/48008 [2:07:03<37:12, 4.92it/s] 77%|███████▋ | 37025/48008 [2:07:03<36:18, 5.04it/s] 77%|███████▋ | 37026/48008 [2:07:03<35:48, 5.11it/s] 77%|███████▋ | 37027/48008 [2:07:04<41:47, 4.38it/s] 77%|███████▋ | 37028/48008 [2:07:04<39:38, 4.62it/s] 77%|███████▋ | 37029/48008 [2:07:04<38:29, 4.75it/s] 77%|███████▋ | 37030/48008 [2:07:04<37:11, 4.92it/s] 77%|███████▋ | 37031/48008 [2:07:04<36:11, 5.06it/s] 77%|███████▋ | 37032/48008 [2:07:05<35:40, 5.13it/s] 77%|███████▋ | 37033/48008 [2:07:05<35:27, 5.16it/s] 77%|███████▋ | 37034/48008 [2:07:05<35:25, 5.16it/s] 77%|███████▋ | 37035/48008 [2:07:05<35:11, 5.20it/s] 77%|███████▋ | 37036/48008 [2:07:05<35:03, 5.22it/s] 77%|███████▋ | 37037/48008 [2:07:06<35:03, 5.22it/s] 77%|███████▋ | 37038/48008 [2:07:06<35:14, 5.19it/s] 77%|███████▋ | 37039/48008 [2:07:06<35:05, 5.21it/s] 77%|███████▋ | 37040/48008 [2:07:06<34:56, 5.23it/s] 77%|███████▋ | 37041/48008 [2:07:06<34:53, 5.24it/s] 77%|███████▋ | 37042/48008 [2:07:06<34:55, 5.23it/s] 77%|███████▋ | 37043/48008 [2:07:07<34:55, 5.23it/s] 77%|███████▋ | 37044/48008 [2:07:07<35:56, 5.08it/s] 77%|███████▋ | 37045/48008 [2:07:07<35:42, 5.12it/s] 77%|███████▋ | 37046/48008 [2:07:07<35:25, 5.16it/s] 77%|███████▋ | 37047/48008 [2:07:07<35:10, 5.19it/s] 77%|███████▋ | 37048/48008 [2:07:08<35:04, 5.21it/s] 77%|███████▋ | 37049/48008 [2:07:08<34:55, 5.23it/s] 77%|███████▋ | 37050/48008 [2:07:08<34:57, 5.22it/s] {'loss': 0.3008, 'grad_norm': 0.2216796875, 'learning_rate': 4.5654890851524744e-05, 'epoch': 0.77} 77%|███████▋ | 37050/48008 [2:07:08<34:57, 5.22it/s] 77%|███████▋ | 37051/48008 [2:07:08<37:31, 4.87it/s] 77%|███████▋ | 37052/48008 [2:07:08<37:37, 4.85it/s] 77%|███████▋ | 37053/48008 [2:07:09<36:49, 4.96it/s] 77%|███████▋ | 37054/48008 [2:07:09<46:59, 3.88it/s] 77%|███████▋ | 37055/48008 [2:07:09<43:17, 4.22it/s] 77%|███████▋ | 37056/48008 [2:07:09<40:36, 4.50it/s] 77%|███████▋ | 37057/48008 [2:07:10<38:56, 4.69it/s] 77%|███████▋ | 37058/48008 [2:07:10<38:08, 4.79it/s] 77%|███████▋ | 37059/48008 [2:07:10<37:06, 4.92it/s] 77%|███████▋ | 37060/48008 [2:07:10<36:33, 4.99it/s] 77%|███████▋ | 37061/48008 [2:07:10<36:07, 5.05it/s] 77%|███████▋ | 37062/48008 [2:07:11<35:47, 5.10it/s] 77%|███████▋ | 37063/48008 [2:07:11<35:43, 5.11it/s] 77%|███████▋ | 37064/48008 [2:07:11<35:23, 5.15it/s] 77%|███████▋ | 37065/48008 [2:07:11<34:39, 5.26it/s] 77%|███████▋ | 37066/48008 [2:07:11<34:20, 5.31it/s] 77%|███████▋ | 37067/48008 [2:07:12<34:27, 5.29it/s] 77%|███████▋ | 37068/48008 [2:07:12<34:19, 5.31it/s] 77%|███████▋ | 37069/48008 [2:07:12<34:42, 5.25it/s] 77%|███████▋ | 37070/48008 [2:07:12<34:41, 5.26it/s] 77%|███████▋ | 37071/48008 [2:07:12<34:45, 5.24it/s] 77%|███████▋ | 37072/48008 [2:07:12<34:46, 5.24it/s] 77%|███████▋ | 37073/48008 [2:07:13<35:20, 5.16it/s] 77%|███████▋ | 37074/48008 [2:07:13<35:53, 5.08it/s] 77%|███████▋ | 37075/48008 [2:07:13<35:33, 5.13it/s] 77%|███████▋ | 37076/48008 [2:07:13<37:43, 4.83it/s] 77%|███████▋ | 37077/48008 [2:07:14<37:52, 4.81it/s] 77%|███████▋ | 37078/48008 [2:07:14<36:55, 4.93it/s] 77%|███████▋ | 37079/48008 [2:07:14<36:40, 4.97it/s] 77%|███████▋ | 37080/48008 [2:07:14<36:23, 5.00it/s] 77%|███████▋ | 37081/48008 [2:07:14<35:46, 5.09it/s] 77%|███████▋ | 37082/48008 [2:07:14<35:11, 5.17it/s] 77%|███████▋ | 37083/48008 [2:07:15<35:07, 5.18it/s] 77%|███████▋ | 37084/48008 [2:07:15<35:56, 5.06it/s] 77%|███████▋ | 37085/48008 [2:07:15<35:31, 5.12it/s] 77%|███████▋ | 37086/48008 [2:07:15<35:11, 5.17it/s] 77%|███████▋ | 37087/48008 [2:07:15<35:11, 5.17it/s] 77%|███████▋ | 37088/48008 [2:07:16<35:09, 5.18it/s] 77%|███████▋ | 37089/48008 [2:07:16<35:49, 5.08it/s] 77%|███████▋ | 37090/48008 [2:07:16<35:26, 5.13it/s] 77%|███████▋ | 37091/48008 [2:07:16<35:52, 5.07it/s] 77%|███████▋ | 37092/48008 [2:07:16<35:28, 5.13it/s] 77%|███████▋ | 37093/48008 [2:07:17<35:07, 5.18it/s] 77%|███████▋ | 37094/48008 [2:07:17<35:00, 5.20it/s] 77%|███████▋ | 37095/48008 [2:07:17<35:35, 5.11it/s] 77%|███████▋ | 37096/48008 [2:07:17<35:16, 5.16it/s] 77%|███████▋ | 37097/48008 [2:07:17<35:16, 5.15it/s] 77%|███████▋ | 37098/48008 [2:07:18<36:16, 5.01it/s] 77%|███████▋ | 37099/48008 [2:07:18<35:44, 5.09it/s] 77%|███████▋ | 37100/48008 [2:07:18<35:29, 5.12it/s] {'loss': 0.3005, 'grad_norm': 0.2275390625, 'learning_rate': 4.544659223462757e-05, 'epoch': 0.77} 77%|███████▋ | 37100/48008 [2:07:18<35:29, 5.12it/s] 77%|███████▋ | 37101/48008 [2:07:18<35:19, 5.15it/s] 77%|███████▋ | 37102/48008 [2:07:18<34:50, 5.22it/s] 77%|███████▋ | 37103/48008 [2:07:19<35:21, 5.14it/s] 77%|███████▋ | 37104/48008 [2:07:19<35:03, 5.18it/s] 77%|███████▋ | 37105/48008 [2:07:19<34:42, 5.23it/s] 77%|███████▋ | 37106/48008 [2:07:19<34:38, 5.25it/s] 77%|███████▋ | 37107/48008 [2:07:19<34:29, 5.27it/s] 77%|███████▋ | 37108/48008 [2:07:20<34:31, 5.26it/s] 77%|███████▋ | 37109/48008 [2:07:20<34:32, 5.26it/s] 77%|███████▋ | 37110/48008 [2:07:20<34:24, 5.28it/s] 77%|███████▋ | 37111/48008 [2:07:20<35:32, 5.11it/s] 77%|███████▋ | 37112/48008 [2:07:20<34:50, 5.21it/s] 77%|███████▋ | 37113/48008 [2:07:20<34:45, 5.22it/s] 77%|███████▋ | 37114/48008 [2:07:21<34:32, 5.26it/s] 77%|███████▋ | 37115/48008 [2:07:21<35:07, 5.17it/s] 77%|███████▋ | 37116/48008 [2:07:21<35:36, 5.10it/s] 77%|███████▋ | 37117/48008 [2:07:21<35:54, 5.05it/s] 77%|███████▋ | 37118/48008 [2:07:21<35:35, 5.10it/s] 77%|███████▋ | 37119/48008 [2:07:22<35:15, 5.15it/s] 77%|███████▋ | 37120/48008 [2:07:22<34:59, 5.19it/s] 77%|███████▋ | 37121/48008 [2:07:22<34:55, 5.20it/s] 77%|███████▋ | 37122/48008 [2:07:22<35:18, 5.14it/s] 77%|███████▋ | 37123/48008 [2:07:22<34:49, 5.21it/s] 77%|███████▋ | 37124/48008 [2:07:23<34:47, 5.21it/s] 77%|███████▋ | 37125/48008 [2:07:23<34:22, 5.28it/s] 77%|███████▋ | 37126/48008 [2:07:23<34:20, 5.28it/s] 77%|███████▋ | 37127/48008 [2:07:23<34:29, 5.26it/s] 77%|███████▋ | 37128/48008 [2:07:23<34:30, 5.25it/s] 77%|███████▋ | 37129/48008 [2:07:24<34:17, 5.29it/s] 77%|███████▋ | 37130/48008 [2:07:24<34:16, 5.29it/s] 77%|███████▋ | 37131/48008 [2:07:24<35:05, 5.17it/s] 77%|███████▋ | 37132/48008 [2:07:24<35:02, 5.17it/s] 77%|███████▋ | 37133/48008 [2:07:24<34:56, 5.19it/s] 77%|███████▋ | 37134/48008 [2:07:25<45:26, 3.99it/s] 77%|███████▋ | 37135/48008 [2:07:25<48:29, 3.74it/s] 77%|███████▋ | 37136/48008 [2:07:25<46:42, 3.88it/s] 77%|███████▋ | 37137/48008 [2:07:25<42:57, 4.22it/s] 77%|███████▋ | 37138/48008 [2:07:26<40:31, 4.47it/s] 77%|███████▋ | 37139/48008 [2:07:26<38:29, 4.71it/s] 77%|███████▋ | 37140/48008 [2:07:26<37:34, 4.82it/s] 77%|███████▋ | 37141/48008 [2:07:26<36:42, 4.93it/s] 77%|███████▋ | 37142/48008 [2:07:26<35:55, 5.04it/s] 77%|███████▋ | 37143/48008 [2:07:27<36:31, 4.96it/s] 77%|███████▋ | 37144/48008 [2:07:27<35:52, 5.05it/s] 77%|███████▋ | 37145/48008 [2:07:27<35:29, 5.10it/s] 77%|███████▋ | 37146/48008 [2:07:27<34:59, 5.17it/s] 77%|███████▋ | 37147/48008 [2:07:27<34:52, 5.19it/s] 77%|███████▋ | 37148/48008 [2:07:28<34:36, 5.23it/s] 77%|███████▋ | 37149/48008 [2:07:28<34:32, 5.24it/s] 77%|███████▋ | 37150/48008 [2:07:28<34:33, 5.24it/s] {'loss': 0.3011, 'grad_norm': 0.177734375, 'learning_rate': 4.523829361773038e-05, 'epoch': 0.77} 77%|███████▋ | 37150/48008 [2:07:28<34:33, 5.24it/s] 77%|███████▋ | 37151/48008 [2:07:28<34:44, 5.21it/s] 77%|███████▋ | 37152/48008 [2:07:28<34:40, 5.22it/s] 77%|███████▋ | 37153/48008 [2:07:29<35:01, 5.17it/s] 77%|███████▋ | 37154/48008 [2:07:29<35:31, 5.09it/s] 77%|███████▋ | 37155/48008 [2:07:29<35:15, 5.13it/s] 77%|███████▋ | 37156/48008 [2:07:29<35:41, 5.07it/s] 77%|███████▋ | 37157/48008 [2:07:29<35:21, 5.12it/s] 77%|███████▋ | 37158/48008 [2:07:30<35:12, 5.14it/s] 77%|███████▋ | 37159/48008 [2:07:30<34:46, 5.20it/s] 77%|███████▋ | 37160/48008 [2:07:30<34:26, 5.25it/s] 77%|███████▋ | 37161/48008 [2:07:30<40:43, 4.44it/s] 77%|███████▋ | 37162/48008 [2:07:30<38:40, 4.67it/s] 77%|███████▋ | 37163/48008 [2:07:31<37:28, 4.82it/s] 77%|███████▋ | 37164/48008 [2:07:31<37:14, 4.85it/s] 77%|███████▋ | 37165/48008 [2:07:31<37:24, 4.83it/s] 77%|███████▋ | 37166/48008 [2:07:31<36:51, 4.90it/s] 77%|███████▋ | 37167/48008 [2:07:31<36:07, 5.00it/s] 77%|███████▋ | 37168/48008 [2:07:32<35:32, 5.08it/s] 77%|███████▋ | 37169/48008 [2:07:32<35:22, 5.11it/s] 77%|███████▋ | 37170/48008 [2:07:32<34:58, 5.16it/s] 77%|███████▋ | 37171/48008 [2:07:32<34:35, 5.22it/s] 77%|███████▋ | 37172/48008 [2:07:32<34:38, 5.21it/s] 77%|███████▋ | 37173/48008 [2:07:33<35:00, 5.16it/s] 77%|███████▋ | 37174/48008 [2:07:33<34:44, 5.20it/s] 77%|███████▋ | 37175/48008 [2:07:33<34:27, 5.24it/s] 77%|███████▋ | 37176/48008 [2:07:33<34:36, 5.22it/s] 77%|███████▋ | 37177/48008 [2:07:33<34:23, 5.25it/s] 77%|███████▋ | 37178/48008 [2:07:34<36:53, 4.89it/s] 77%|███████▋ | 37179/48008 [2:07:34<36:14, 4.98it/s] 77%|███████▋ | 37180/48008 [2:07:34<35:37, 5.07it/s] 77%|███████▋ | 37181/48008 [2:07:34<35:17, 5.11it/s] 77%|███████▋ | 37182/48008 [2:07:34<36:06, 5.00it/s] 77%|███████▋ | 37183/48008 [2:07:34<35:31, 5.08it/s] 77%|███████▋ | 37184/48008 [2:07:35<35:48, 5.04it/s] 77%|███████▋ | 37185/48008 [2:07:35<35:56, 5.02it/s] 77%|███████▋ | 37186/48008 [2:07:35<36:26, 4.95it/s] 77%|███████▋ | 37187/48008 [2:07:35<35:38, 5.06it/s] 77%|███████▋ | 37188/48008 [2:07:35<35:13, 5.12it/s] 77%|███████▋ | 37189/48008 [2:07:36<37:29, 4.81it/s] 77%|███████▋ | 37190/48008 [2:07:36<36:38, 4.92it/s] 77%|███████▋ | 37191/48008 [2:07:36<35:53, 5.02it/s] 77%|███████▋ | 37192/48008 [2:07:36<38:05, 4.73it/s] 77%|███████▋ | 37193/48008 [2:07:37<37:19, 4.83it/s] 77%|███████▋ | 37194/48008 [2:07:37<36:27, 4.94it/s] 77%|███████▋ | 37195/48008 [2:07:37<35:48, 5.03it/s] 77%|███████▋ | 37196/48008 [2:07:37<36:03, 5.00it/s] 77%|███████▋ | 37197/48008 [2:07:37<35:28, 5.08it/s] 77%|███████▋ | 37198/48008 [2:07:37<34:57, 5.15it/s] 77%|███████▋ | 37199/48008 [2:07:38<35:31, 5.07it/s] 77%|███████▋ | 37200/48008 [2:07:38<35:07, 5.13it/s] {'loss': 0.2982, 'grad_norm': 0.4296875, 'learning_rate': 4.5029995000833196e-05, 'epoch': 0.77} 77%|███████▋ | 37200/48008 [2:07:38<35:07, 5.13it/s] 77%|███████▋ | 37201/48008 [2:07:38<34:56, 5.15it/s] 77%|███████▋ | 37202/48008 [2:07:38<34:48, 5.17it/s] 77%|███████▋ | 37203/48008 [2:07:38<35:20, 5.10it/s] 77%|███████▋ | 37204/48008 [2:07:39<34:48, 5.17it/s] 77%|███████▋ | 37205/48008 [2:07:39<34:18, 5.25it/s] 77%|███████▋ | 37206/48008 [2:07:39<34:27, 5.22it/s] 78%|███████▊ | 37207/48008 [2:07:39<34:19, 5.24it/s] 78%|███████▊ | 37208/48008 [2:07:39<34:33, 5.21it/s] 78%|███████▊ | 37209/48008 [2:07:40<35:43, 5.04it/s] 78%|███████▊ | 37210/48008 [2:07:40<35:04, 5.13it/s] 78%|███████▊ | 37211/48008 [2:07:40<34:53, 5.16it/s] 78%|███████▊ | 37212/48008 [2:07:40<34:50, 5.16it/s] 78%|███████▊ | 37213/48008 [2:07:40<34:38, 5.19it/s] 78%|███████▊ | 37214/48008 [2:07:41<34:55, 5.15it/s] 78%|███████▊ | 37215/48008 [2:07:41<34:35, 5.20it/s] 78%|███████▊ | 37216/48008 [2:07:41<35:02, 5.13it/s] 78%|███████▊ | 37217/48008 [2:07:41<35:30, 5.06it/s] 78%|███████▊ | 37218/48008 [2:07:41<35:26, 5.07it/s] 78%|███████▊ | 37219/48008 [2:07:42<35:32, 5.06it/s] 78%|███████▊ | 37220/48008 [2:07:42<35:08, 5.12it/s] 78%|███████▊ | 37221/48008 [2:07:42<35:00, 5.13it/s] 78%|███████▊ | 37222/48008 [2:07:42<37:22, 4.81it/s] 78%|███████▊ | 37223/48008 [2:07:43<47:01, 3.82it/s] 78%|███████▊ | 37224/48008 [2:07:43<43:13, 4.16it/s] 78%|███████▊ | 37225/48008 [2:07:43<40:33, 4.43it/s] 78%|███████▊ | 37226/48008 [2:07:43<38:42, 4.64it/s] 78%|███████▊ | 37227/48008 [2:07:43<38:05, 4.72it/s] 78%|███████▊ | 37228/48008 [2:07:44<36:58, 4.86it/s] 78%|███████▊ | 37229/48008 [2:07:44<36:16, 4.95it/s] 78%|███████▊ | 37230/48008 [2:07:44<35:41, 5.03it/s] 78%|███████▊ | 37231/48008 [2:07:44<35:25, 5.07it/s] 78%|███████▊ | 37232/48008 [2:07:44<35:31, 5.05it/s] 78%|███████▊ | 37233/48008 [2:07:45<35:14, 5.10it/s] 78%|███████▊ | 37234/48008 [2:07:45<35:06, 5.11it/s] 78%|███████▊ | 37235/48008 [2:07:45<34:43, 5.17it/s] 78%|███████▊ | 37236/48008 [2:07:45<34:38, 5.18it/s] 78%|███████▊ | 37237/48008 [2:07:45<34:41, 5.17it/s] 78%|███████▊ | 37238/48008 [2:07:46<35:37, 5.04it/s] 78%|███████▊ | 37239/48008 [2:07:46<35:18, 5.08it/s] 78%|███████▊ | 37240/48008 [2:07:46<35:35, 5.04it/s] 78%|███████▊ | 37241/48008 [2:07:46<35:11, 5.10it/s] 78%|███████▊ | 37242/48008 [2:07:46<34:52, 5.14it/s] 78%|███████▊ | 37243/48008 [2:07:46<35:28, 5.06it/s] 78%|███████▊ | 37244/48008 [2:07:47<35:28, 5.06it/s] 78%|███████▊ | 37245/48008 [2:07:47<35:07, 5.11it/s] 78%|███████▊ | 37246/48008 [2:07:47<34:48, 5.15it/s] 78%|███████▊ | 37247/48008 [2:07:47<35:38, 5.03it/s] 78%|███████▊ | 37248/48008 [2:07:47<35:11, 5.10it/s] 78%|███████▊ | 37249/48008 [2:07:48<35:17, 5.08it/s] 78%|███████▊ | 37250/48008 [2:07:48<41:06, 4.36it/s] {'loss': 0.2997, 'grad_norm': 0.68359375, 'learning_rate': 4.4821696383936014e-05, 'epoch': 0.78} 78%|███████▊ | 37250/48008 [2:07:48<41:06, 4.36it/s] 78%|███████▊ | 37251/48008 [2:07:48<39:02, 4.59it/s] 78%|███████▊ | 37252/48008 [2:07:48<37:23, 4.79it/s] 78%|███████▊ | 37253/48008 [2:07:49<36:25, 4.92it/s] 78%|███████▊ | 37254/48008 [2:07:49<35:25, 5.06it/s] 78%|███████▊ | 37255/48008 [2:07:49<35:10, 5.09it/s] 78%|███████▊ | 37256/48008 [2:07:49<37:26, 4.79it/s] 78%|███████▊ | 37257/48008 [2:07:49<36:24, 4.92it/s] 78%|███████▊ | 37258/48008 [2:07:50<36:09, 4.95it/s] 78%|███████▊ | 37259/48008 [2:07:50<35:33, 5.04it/s] 78%|███████▊ | 37260/48008 [2:07:50<34:51, 5.14it/s] 78%|███████▊ | 37261/48008 [2:07:50<37:13, 4.81it/s] 78%|███████▊ | 37262/48008 [2:07:50<36:25, 4.92it/s] 78%|███████▊ | 37263/48008 [2:07:51<35:45, 5.01it/s] 78%|███████▊ | 37264/48008 [2:07:51<35:19, 5.07it/s] 78%|███████▊ | 37265/48008 [2:07:51<35:58, 4.98it/s] 78%|███████▊ | 37266/48008 [2:07:51<35:21, 5.06it/s] 78%|███████▊ | 37267/48008 [2:07:51<35:20, 5.06it/s] 78%|███████▊ | 37268/48008 [2:07:52<35:13, 5.08it/s] 78%|███████▊ | 37269/48008 [2:07:52<34:45, 5.15it/s] 78%|███████▊ | 37270/48008 [2:07:52<40:52, 4.38it/s] 78%|███████▊ | 37271/48008 [2:07:52<38:51, 4.61it/s] 78%|███████▊ | 37272/48008 [2:07:52<38:34, 4.64it/s] 78%|███████▊ | 37273/48008 [2:07:53<37:20, 4.79it/s] 78%|███████▊ | 37274/48008 [2:07:53<36:24, 4.91it/s] 78%|███████▊ | 37275/48008 [2:07:53<36:15, 4.93it/s] 78%|███████▊ | 37276/48008 [2:07:53<36:11, 4.94it/s] 78%|███████▊ | 37277/48008 [2:07:53<35:35, 5.02it/s] 78%|███████▊ | 37278/48008 [2:07:54<34:59, 5.11it/s] 78%|███████▊ | 37279/48008 [2:07:54<35:21, 5.06it/s] 78%|███████▊ | 37280/48008 [2:07:54<35:05, 5.09it/s] 78%|███████▊ | 37281/48008 [2:07:54<34:35, 5.17it/s] 78%|███████▊ | 37282/48008 [2:07:54<35:05, 5.09it/s] 78%|███████▊ | 37283/48008 [2:07:55<35:45, 5.00it/s] 78%|███████▊ | 37284/48008 [2:07:55<35:11, 5.08it/s] 78%|███████▊ | 37285/48008 [2:07:55<34:37, 5.16it/s] 78%|███████▊ | 37286/48008 [2:07:55<34:28, 5.18it/s] 78%|███████▊ | 37287/48008 [2:07:55<34:33, 5.17it/s] 78%|███████▊ | 37288/48008 [2:07:56<44:54, 3.98it/s] 78%|███████▊ | 37289/48008 [2:07:56<42:17, 4.22it/s] 78%|███████▊ | 37290/48008 [2:07:56<42:20, 4.22it/s] 78%|███████▊ | 37291/48008 [2:07:56<40:32, 4.41it/s] 78%|███████▊ | 37292/48008 [2:07:57<38:37, 4.62it/s] 78%|███████▊ | 37293/48008 [2:07:57<39:47, 4.49it/s] 78%|███████▊ | 37294/48008 [2:07:57<38:44, 4.61it/s] 78%|███████▊ | 37295/48008 [2:07:57<37:25, 4.77it/s] 78%|███████▊ | 37296/48008 [2:07:57<36:56, 4.83it/s] 78%|███████▊ | 37297/48008 [2:07:58<36:58, 4.83it/s] 78%|███████▊ | 37298/48008 [2:07:58<42:20, 4.22it/s] 78%|███████▊ | 37299/48008 [2:07:58<40:44, 4.38it/s] 78%|███████▊ | 37300/48008 [2:07:58<39:10, 4.56it/s] {'loss': 0.2981, 'grad_norm': 0.15625, 'learning_rate': 4.461339776703883e-05, 'epoch': 0.78} 78%|███████▊ | 37300/48008 [2:07:58<39:10, 4.56it/s] 78%|███████▊ | 37301/48008 [2:07:59<38:19, 4.66it/s] 78%|███████▊ | 37302/48008 [2:07:59<37:02, 4.82it/s] 78%|███████▊ | 37303/48008 [2:07:59<36:11, 4.93it/s] 78%|███████▊ | 37304/48008 [2:07:59<35:29, 5.03it/s] 78%|███████▊ | 37305/48008 [2:07:59<35:21, 5.05it/s] 78%|███████▊ | 37306/48008 [2:08:00<41:22, 4.31it/s] 78%|███████▊ | 37307/48008 [2:08:00<39:11, 4.55it/s] 78%|███████▊ | 37308/48008 [2:08:00<37:16, 4.78it/s] 78%|███████▊ | 37309/48008 [2:08:00<37:17, 4.78it/s] 78%|███████▊ | 37310/48008 [2:08:00<36:57, 4.82it/s] 78%|███████▊ | 37311/48008 [2:08:01<35:53, 4.97it/s] 78%|███████▊ | 37312/48008 [2:08:01<35:11, 5.06it/s] 78%|███████▊ | 37313/48008 [2:08:01<35:09, 5.07it/s] 78%|███████▊ | 37314/48008 [2:08:01<45:15, 3.94it/s] 78%|███████▊ | 37315/48008 [2:08:02<41:51, 4.26it/s] 78%|███████▊ | 37316/48008 [2:08:02<39:24, 4.52it/s] 78%|███████▊ | 37317/48008 [2:08:02<38:22, 4.64it/s] 78%|███████▊ | 37318/48008 [2:08:02<36:57, 4.82it/s] 78%|███████▊ | 37319/48008 [2:08:02<37:01, 4.81it/s] 78%|███████▊ | 37320/48008 [2:08:03<35:57, 4.95it/s] 78%|███████▊ | 37321/48008 [2:08:03<35:28, 5.02it/s] 78%|███████▊ | 37322/48008 [2:08:03<35:37, 5.00it/s] 78%|███████▊ | 37323/48008 [2:08:03<34:56, 5.10it/s] 78%|███████▊ | 37324/48008 [2:08:03<37:08, 4.80it/s] 78%|███████▊ | 37325/48008 [2:08:04<36:12, 4.92it/s] 78%|███████▊ | 37326/48008 [2:08:04<35:28, 5.02it/s] 78%|███████▊ | 37327/48008 [2:08:04<34:45, 5.12it/s] 78%|███████▊ | 37328/48008 [2:08:04<37:01, 4.81it/s] 78%|███████▊ | 37329/48008 [2:08:04<36:21, 4.89it/s] 78%|███████▊ | 37330/48008 [2:08:05<36:12, 4.92it/s] 78%|███████▊ | 37331/48008 [2:08:05<38:04, 4.67it/s] 78%|███████▊ | 37332/48008 [2:08:05<36:53, 4.82it/s] 78%|███████▊ | 37333/48008 [2:08:05<36:02, 4.94it/s] 78%|███████▊ | 37334/48008 [2:08:05<35:26, 5.02it/s] 78%|███████▊ | 37335/48008 [2:08:06<35:42, 4.98it/s] 78%|███████▊ | 37336/48008 [2:08:06<35:12, 5.05it/s] 78%|███████▊ | 37337/48008 [2:08:06<34:37, 5.14it/s] 78%|███████▊ | 37338/48008 [2:08:06<34:16, 5.19it/s] 78%|███████▊ | 37339/48008 [2:08:06<34:03, 5.22it/s] 78%|███████▊ | 37340/48008 [2:08:07<34:17, 5.19it/s] 78%|███████▊ | 37341/48008 [2:08:07<36:36, 4.86it/s] 78%|███████▊ | 37342/48008 [2:08:07<35:50, 4.96it/s] 78%|███████▊ | 37343/48008 [2:08:07<35:26, 5.02it/s] 78%|███████▊ | 37344/48008 [2:08:07<35:47, 4.96it/s] 78%|███████▊ | 37345/48008 [2:08:08<35:08, 5.06it/s] 78%|███████▊ | 37346/48008 [2:08:08<34:32, 5.14it/s] 78%|███████▊ | 37347/48008 [2:08:08<34:20, 5.17it/s] 78%|███████▊ | 37348/48008 [2:08:08<34:03, 5.22it/s] 78%|███████▊ | 37349/48008 [2:08:08<34:00, 5.22it/s] 78%|███████▊ | 37350/48008 [2:08:08<33:56, 5.23it/s] {'loss': 0.2974, 'grad_norm': 0.1630859375, 'learning_rate': 4.440509915014164e-05, 'epoch': 0.78} 78%|███████▊ | 37350/48008 [2:08:08<33:56, 5.23it/s] 78%|███████▊ | 37351/48008 [2:08:09<33:43, 5.27it/s] 78%|███████▊ | 37352/48008 [2:08:09<33:53, 5.24it/s] 78%|███████▊ | 37353/48008 [2:08:09<33:48, 5.25it/s] 78%|███████▊ | 37354/48008 [2:08:09<33:47, 5.25it/s] 78%|███████▊ | 37355/48008 [2:08:09<33:22, 5.32it/s] 78%|███████▊ | 37356/48008 [2:08:10<33:18, 5.33it/s] 78%|███████▊ | 37357/48008 [2:08:10<33:28, 5.30it/s] 78%|███████▊ | 37358/48008 [2:08:10<33:19, 5.33it/s] 78%|███████▊ | 37359/48008 [2:08:10<33:43, 5.26it/s] 78%|███████▊ | 37360/48008 [2:08:10<34:02, 5.21it/s] 78%|███████▊ | 37361/48008 [2:08:11<34:02, 5.21it/s] 78%|███████▊ | 37362/48008 [2:08:11<34:35, 5.13it/s] 78%|███████▊ | 37363/48008 [2:08:11<34:56, 5.08it/s] 78%|███████▊ | 37364/48008 [2:08:11<34:33, 5.13it/s] 78%|███████▊ | 37365/48008 [2:08:11<34:23, 5.16it/s] 78%|███████▊ | 37366/48008 [2:08:12<34:17, 5.17it/s] 78%|███████▊ | 37367/48008 [2:08:12<34:21, 5.16it/s] 78%|███████▊ | 37368/48008 [2:08:12<34:14, 5.18it/s] 78%|███████▊ | 37369/48008 [2:08:12<34:20, 5.16it/s] 78%|███████▊ | 37370/48008 [2:08:12<34:58, 5.07it/s] 78%|███████▊ | 37371/48008 [2:08:13<34:32, 5.13it/s] 78%|███████▊ | 37372/48008 [2:08:13<34:09, 5.19it/s] 78%|███████▊ | 37373/48008 [2:08:13<33:48, 5.24it/s] 78%|███████▊ | 37374/48008 [2:08:13<34:06, 5.20it/s] 78%|███████▊ | 37375/48008 [2:08:13<40:13, 4.40it/s] 78%|███████▊ | 37376/48008 [2:08:14<38:20, 4.62it/s] 78%|███████▊ | 37377/48008 [2:08:14<36:44, 4.82it/s] 78%|███████▊ | 37378/48008 [2:08:14<36:26, 4.86it/s] 78%|███████▊ | 37379/48008 [2:08:14<35:34, 4.98it/s] 78%|███████▊ | 37380/48008 [2:08:14<35:45, 4.95it/s] 78%|███████▊ | 37381/48008 [2:08:15<35:10, 5.04it/s] 78%|███████▊ | 37382/48008 [2:08:15<40:54, 4.33it/s] 78%|███████▊ | 37383/48008 [2:08:15<38:58, 4.54it/s] 78%|███████▊ | 37384/48008 [2:08:15<37:20, 4.74it/s] 78%|███████▊ | 37385/48008 [2:08:15<36:02, 4.91it/s] 78%|███████▊ | 37386/48008 [2:08:16<35:30, 4.98it/s] 78%|███████▊ | 37387/48008 [2:08:17<1:14:44, 2.37it/s] 78%|███████▊ | 37388/48008 [2:08:17<1:03:02, 2.81it/s] 78%|███████▊ | 37389/48008 [2:08:17<54:15, 3.26it/s] 78%|███████▊ | 37390/48008 [2:08:17<48:04, 3.68it/s] 78%|███████▊ | 37391/48008 [2:08:17<43:41, 4.05it/s] 78%|███████▊ | 37392/48008 [2:08:18<40:50, 4.33it/s] 78%|███████▊ | 37393/48008 [2:08:18<1:13:27, 2.41it/s] 78%|███████▊ | 37394/48008 [2:08:19<1:01:27, 2.88it/s] 78%|███████▊ | 37395/48008 [2:08:19<53:31, 3.31it/s] 78%|███████▊ | 37396/48008 [2:08:19<47:25, 3.73it/s] 78%|███████▊ | 37397/48008 [2:08:19<43:41, 4.05it/s] 78%|███████▊ | 37398/48008 [2:08:19<40:38, 4.35it/s] 78%|███████▊ | 37399/48008 [2:08:20<38:31, 4.59it/s] 78%|███████▊ | 37400/48008 [2:08:20<36:56, 4.79it/s] {'loss': 0.3008, 'grad_norm': 0.177734375, 'learning_rate': 4.419680053324446e-05, 'epoch': 0.78} 78%|███████▊ | 37400/48008 [2:08:20<36:56, 4.79it/s] 78%|███████▊ | 37401/48008 [2:08:20<35:59, 4.91it/s] 78%|███████▊ | 37402/48008 [2:08:20<35:15, 5.01it/s] 78%|███████▊ | 37403/48008 [2:08:20<45:09, 3.91it/s] 78%|███████▊ | 37404/48008 [2:08:21<41:35, 4.25it/s] 78%|███████▊ | 37405/48008 [2:08:22<1:20:22, 2.20it/s] 78%|███████▊ | 37406/48008 [2:08:22<1:06:15, 2.67it/s] 78%|███████▊ | 37407/48008 [2:08:22<56:26, 3.13it/s] 78%|███████▊ | 37408/48008 [2:08:22<49:43, 3.55it/s] 78%|███████▊ | 37409/48008 [2:08:22<44:49, 3.94it/s] 78%|███████▊ | 37410/48008 [2:08:23<41:42, 4.23it/s] 78%|███████▊ | 37411/48008 [2:08:23<1:14:06, 2.38it/s] 78%|███████▊ | 37412/48008 [2:08:24<1:01:46, 2.86it/s] 78%|███████▊ | 37413/48008 [2:08:24<53:02, 3.33it/s] 78%|███████▊ | 37414/48008 [2:08:24<47:13, 3.74it/s] 78%|███████▊ | 37415/48008 [2:08:24<43:13, 4.08it/s] 78%|███████▊ | 37416/48008 [2:08:25<50:41, 3.48it/s] 78%|███████▊ | 37417/48008 [2:08:25<45:39, 3.87it/s] 78%|███████▊ | 37418/48008 [2:08:25<42:11, 4.18it/s] 78%|███████▊ | 37419/48008 [2:08:25<39:44, 4.44it/s] 78%|███████▊ | 37420/48008 [2:08:25<37:50, 4.66it/s] 78%|███████▊ | 37421/48008 [2:08:26<36:26, 4.84it/s] 78%|███████▊ | 37422/48008 [2:08:26<36:18, 4.86it/s] 78%|███████▊ | 37423/48008 [2:08:26<35:24, 4.98it/s] 78%|███████▊ | 37424/48008 [2:08:26<34:59, 5.04it/s] 78%|███████▊ | 37425/48008 [2:08:26<34:34, 5.10it/s] 78%|███████▊ | 37426/48008 [2:08:27<34:22, 5.13it/s] 78%|███████▊ | 37427/48008 [2:08:27<34:15, 5.15it/s] 78%|███████▊ | 37428/48008 [2:08:27<34:25, 5.12it/s] 78%|███████▊ | 37429/48008 [2:08:27<34:50, 5.06it/s] 78%|███████▊ | 37430/48008 [2:08:27<34:58, 5.04it/s] 78%|███████▊ | 37431/48008 [2:08:27<34:31, 5.10it/s] 78%|███████▊ | 37432/48008 [2:08:28<33:58, 5.19it/s] 78%|███████▊ | 37433/48008 [2:08:28<36:25, 4.84it/s] 78%|███████▊ | 37434/48008 [2:08:28<36:29, 4.83it/s] 78%|███████▊ | 37435/48008 [2:08:28<36:09, 4.87it/s] 78%|███████▊ | 37436/48008 [2:08:29<35:11, 5.01it/s] 78%|███████▊ | 37437/48008 [2:08:29<35:09, 5.01it/s] 78%|███████▊ | 37438/48008 [2:08:29<34:54, 5.05it/s] 78%|███████▊ | 37439/48008 [2:08:29<34:31, 5.10it/s] 78%|███████▊ | 37440/48008 [2:08:29<34:26, 5.11it/s] 78%|███████▊ | 37441/48008 [2:08:29<34:04, 5.17it/s] 78%|███████▊ | 37442/48008 [2:08:30<34:26, 5.11it/s] 78%|███████▊ | 37443/48008 [2:08:30<44:28, 3.96it/s] 78%|███████▊ | 37444/48008 [2:08:30<42:06, 4.18it/s] 78%|███████▊ | 37445/48008 [2:08:30<39:36, 4.44it/s] 78%|███████▊ | 37446/48008 [2:08:31<38:31, 4.57it/s] 78%|███████▊ | 37447/48008 [2:08:32<1:11:00, 2.48it/s] 78%|███████▊ | 37448/48008 [2:08:32<1:00:04, 2.93it/s] 78%|███████▊ | 37449/48008 [2:08:32<52:09, 3.37it/s] 78%|███████▊ | 37450/48008 [2:08:32<46:17, 3.80it/s] {'loss': 0.302, 'grad_norm': 0.3515625, 'learning_rate': 4.398850191634728e-05, 'epoch': 0.78} 78%|███████▊ | 37450/48008 [2:08:32<46:17, 3.80it/s] 78%|███████▊ | 37451/48008 [2:08:32<45:06, 3.90it/s] 78%|███████▊ | 37452/48008 [2:08:33<42:13, 4.17it/s] 78%|███████▊ | 37453/48008 [2:08:33<39:59, 4.40it/s] 78%|███████▊ | 37454/48008 [2:08:33<37:53, 4.64it/s] 78%|███████▊ | 37455/48008 [2:08:33<39:05, 4.50it/s] 78%|███████▊ | 37456/48008 [2:08:33<38:05, 4.62it/s] 78%|███████▊ | 37457/48008 [2:08:34<36:37, 4.80it/s] 78%|███████▊ | 37458/48008 [2:08:34<38:09, 4.61it/s] 78%|███████▊ | 37459/48008 [2:08:34<37:23, 4.70it/s] 78%|███████▊ | 37460/48008 [2:08:34<36:07, 4.87it/s] 78%|███████▊ | 37461/48008 [2:08:34<35:19, 4.98it/s] 78%|███████▊ | 37462/48008 [2:08:35<35:21, 4.97it/s] 78%|███████▊ | 37463/48008 [2:08:35<34:41, 5.07it/s] 78%|███████▊ | 37464/48008 [2:08:35<34:26, 5.10it/s] 78%|███████▊ | 37465/48008 [2:08:35<34:15, 5.13it/s] 78%|███████▊ | 37466/48008 [2:08:35<34:03, 5.16it/s] 78%|███████▊ | 37467/48008 [2:08:36<34:00, 5.17it/s] 78%|███████▊ | 37468/48008 [2:08:36<33:57, 5.17it/s] 78%|███████▊ | 37469/48008 [2:08:36<33:35, 5.23it/s] 78%|███████▊ | 37470/48008 [2:08:36<33:27, 5.25it/s] 78%|███████▊ | 37471/48008 [2:08:36<34:31, 5.09it/s] 78%|███████▊ | 37472/48008 [2:08:37<35:23, 4.96it/s] 78%|███████▊ | 37473/48008 [2:08:37<35:07, 5.00it/s] 78%|███████▊ | 37474/48008 [2:08:37<34:25, 5.10it/s] 78%|███████▊ | 37475/48008 [2:08:37<34:32, 5.08it/s] 78%|███████▊ | 37476/48008 [2:08:37<34:07, 5.14it/s] 78%|███████▊ | 37477/48008 [2:08:37<33:54, 5.18it/s] 78%|███████▊ | 37478/48008 [2:08:38<34:00, 5.16it/s] 78%|███████▊ | 37479/48008 [2:08:38<33:35, 5.23it/s] 78%|███████▊ | 37480/48008 [2:08:38<33:32, 5.23it/s] 78%|███████▊ | 37481/48008 [2:08:38<33:35, 5.22it/s] 78%|███████▊ | 37482/48008 [2:08:38<33:47, 5.19it/s] 78%|███████▊ | 37483/48008 [2:08:39<34:10, 5.13it/s] 78%|███████▊ | 37484/48008 [2:08:39<34:15, 5.12it/s] 78%|███████▊ | 37485/48008 [2:08:39<34:14, 5.12it/s] 78%|███████▊ | 37486/48008 [2:08:39<34:00, 5.16it/s] 78%|███████▊ | 37487/48008 [2:08:40<39:51, 4.40it/s] 78%|███████▊ | 37488/48008 [2:08:40<37:55, 4.62it/s] 78%|███████▊ | 37489/48008 [2:08:40<36:39, 4.78it/s] 78%|███████▊ | 37490/48008 [2:08:40<35:44, 4.90it/s] 78%|███████▊ | 37491/48008 [2:08:40<35:06, 4.99it/s] 78%|███████▊ | 37492/48008 [2:08:40<35:03, 5.00it/s] 78%|███████▊ | 37493/48008 [2:08:41<34:40, 5.05it/s] 78%|███████▊ | 37494/48008 [2:08:41<35:04, 5.00it/s] 78%|███████▊ | 37495/48008 [2:08:41<34:35, 5.07it/s] 78%|███████▊ | 37496/48008 [2:08:41<34:23, 5.09it/s] 78%|███████▊ | 37497/48008 [2:08:41<33:43, 5.20it/s] 78%|███████▊ | 37498/48008 [2:08:42<33:47, 5.18it/s] 78%|███████▊ | 37499/48008 [2:08:42<33:50, 5.18it/s] 78%|███████▊ | 37500/48008 [2:08:42<34:04, 5.14it/s] {'loss': 0.2987, 'grad_norm': 0.392578125, 'learning_rate': 4.3780203299450094e-05, 'epoch': 0.78} 78%|███████▊ | 37500/48008 [2:08:42<34:04, 5.14it/s] 78%|███████▊ | 37501/48008 [2:08:42<33:58, 5.15it/s] 78%|███████▊ | 37502/48008 [2:08:42<34:01, 5.15it/s] 78%|███████▊ | 37503/48008 [2:08:43<34:05, 5.14it/s] 78%|███████▊ | 37504/48008 [2:08:43<33:49, 5.18it/s] 78%|███████▊ | 37505/48008 [2:08:43<33:37, 5.21it/s] 78%|███████▊ | 37506/48008 [2:08:43<33:25, 5.24it/s] 78%|███████▊ | 37507/48008 [2:08:43<33:39, 5.20it/s] 78%|███████▊ | 37508/48008 [2:08:44<33:35, 5.21it/s] 78%|███████▊ | 37509/48008 [2:08:44<34:02, 5.14it/s] 78%|███████▊ | 37510/48008 [2:08:44<44:18, 3.95it/s] 78%|███████▊ | 37511/48008 [2:08:44<41:02, 4.26it/s] 78%|███████▊ | 37512/48008 [2:08:45<48:59, 3.57it/s] 78%|███████▊ | 37513/48008 [2:08:45<44:08, 3.96it/s] 78%|███████▊ | 37514/48008 [2:08:45<41:32, 4.21it/s] 78%|███████▊ | 37515/48008 [2:08:45<38:59, 4.48it/s] 78%|███████▊ | 37516/48008 [2:08:46<37:33, 4.65it/s] 78%|███████▊ | 37517/48008 [2:08:46<36:23, 4.80it/s] 78%|███████▊ | 37518/48008 [2:08:46<35:42, 4.90it/s] 78%|███████▊ | 37519/48008 [2:08:46<34:58, 5.00it/s] 78%|███████▊ | 37520/48008 [2:08:46<34:19, 5.09it/s] 78%|███████▊ | 37521/48008 [2:08:46<33:44, 5.18it/s] 78%|███████▊ | 37522/48008 [2:08:47<34:01, 5.14it/s] 78%|███████▊ | 37523/48008 [2:08:47<36:17, 4.82it/s] 78%|███████▊ | 37524/48008 [2:08:47<45:47, 3.82it/s] 78%|███████▊ | 37525/48008 [2:08:47<42:37, 4.10it/s] 78%|███████▊ | 37526/48008 [2:08:48<39:35, 4.41it/s] 78%|███████▊ | 37527/48008 [2:08:48<37:35, 4.65it/s] 78%|███████▊ | 37528/48008 [2:08:48<37:14, 4.69it/s] 78%|███████▊ | 37529/48008 [2:08:48<46:15, 3.78it/s] 78%|███████▊ | 37530/48008 [2:08:49<42:06, 4.15it/s] 78%|███████▊ | 37531/48008 [2:08:49<39:36, 4.41it/s] 78%|███████▊ | 37532/48008 [2:08:49<38:23, 4.55it/s] 78%|███████▊ | 37533/48008 [2:08:49<37:31, 4.65it/s] 78%|███████▊ | 37534/48008 [2:08:49<36:47, 4.74it/s] 78%|███████▊ | 37535/48008 [2:08:50<35:39, 4.90it/s] 78%|███████▊ | 37536/48008 [2:08:50<35:15, 4.95it/s] 78%|███████▊ | 37537/48008 [2:08:50<35:22, 4.93it/s] 78%|███████▊ | 37538/48008 [2:08:50<34:39, 5.04it/s] 78%|███████▊ | 37539/48008 [2:08:50<34:10, 5.10it/s] 78%|███████▊ | 37540/48008 [2:08:51<1:10:53, 2.46it/s] 78%|███████▊ | 37541/48008 [2:08:52<59:37, 2.93it/s] 78%|███████▊ | 37542/48008 [2:08:52<51:37, 3.38it/s] 78%|███████▊ | 37543/48008 [2:08:52<46:08, 3.78it/s] 78%|███████▊ | 37544/48008 [2:08:52<43:00, 4.06it/s] 78%|███████▊ | 37545/48008 [2:08:52<40:36, 4.29it/s] 78%|███████▊ | 37546/48008 [2:08:52<38:43, 4.50it/s] 78%|███████▊ | 37547/48008 [2:08:53<37:41, 4.63it/s] 78%|███████▊ | 37548/48008 [2:08:53<36:45, 4.74it/s] 78%|███████▊ | 37549/48008 [2:08:53<35:49, 4.87it/s] 78%|███████▊ | 37550/48008 [2:08:53<35:09, 4.96it/s] 78%|███████▊ | 37550/48008 [2:08:53<35:09, 4.96it/s]{'loss': 0.2975, 'grad_norm': 0.2412109375, 'learning_rate': 4.357190468255291e-05, 'epoch': 0.78} 78%|███████▊ | 37551/48008 [2:08:53<34:26, 5.06it/s] 78%|███████▊ | 37552/48008 [2:08:54<34:38, 5.03it/s] 78%|███████▊ | 37553/48008 [2:08:54<34:09, 5.10it/s] 78%|███████▊ | 37554/48008 [2:08:54<33:57, 5.13it/s] 78%|███████▊ | 37555/48008 [2:08:54<33:53, 5.14it/s] 78%|███████▊ | 37556/48008 [2:08:54<33:27, 5.21it/s] 78%|███████▊ | 37557/48008 [2:08:55<33:21, 5.22it/s] 78%|███████▊ | 37558/48008 [2:08:55<33:16, 5.23it/s] 78%|███████▊ | 37559/48008 [2:08:55<33:36, 5.18it/s] 78%|███████▊ | 37560/48008 [2:08:55<32:57, 5.28it/s] 78%|███████▊ | 37561/48008 [2:08:55<32:52, 5.30it/s] 78%|███████▊ | 37562/48008 [2:08:56<32:52, 5.30it/s] 78%|███████▊ | 37563/48008 [2:08:56<32:46, 5.31it/s] 78%|███████▊ | 37564/48008 [2:08:56<32:49, 5.30it/s] 78%|███████▊ | 37565/48008 [2:08:56<32:57, 5.28it/s] 78%|███████▊ | 37566/48008 [2:08:56<32:56, 5.28it/s] 78%|███████▊ | 37567/48008 [2:08:57<32:54, 5.29it/s] 78%|███████▊ | 37568/48008 [2:08:57<33:04, 5.26it/s] 78%|███████▊ | 37569/48008 [2:08:57<32:57, 5.28it/s] 78%|███████▊ | 37570/48008 [2:08:57<33:43, 5.16it/s] 78%|███████▊ | 37571/48008 [2:08:57<34:32, 5.03it/s] 78%|███████▊ | 37572/48008 [2:08:57<33:56, 5.13it/s] 78%|███████▊ | 37573/48008 [2:08:58<33:44, 5.15it/s] 78%|███████▊ | 37574/48008 [2:08:58<33:12, 5.24it/s] 78%|███████▊ | 37575/48008 [2:08:58<33:16, 5.23it/s] 78%|███████▊ | 37576/48008 [2:08:58<33:16, 5.23it/s] 78%|███████▊ | 37577/48008 [2:08:58<33:20, 5.21it/s] 78%|███████▊ | 37578/48008 [2:08:59<33:26, 5.20it/s] 78%|███████▊ | 37579/48008 [2:08:59<34:01, 5.11it/s] 78%|███████▊ | 37580/48008 [2:08:59<33:26, 5.20it/s] 78%|███████▊ | 37581/48008 [2:08:59<33:22, 5.21it/s] 78%|███████▊ | 37582/48008 [2:08:59<35:43, 4.86it/s] 78%|███████▊ | 37583/48008 [2:09:00<34:49, 4.99it/s] 78%|███████▊ | 37584/48008 [2:09:00<34:16, 5.07it/s] 78%|███████▊ | 37585/48008 [2:09:00<34:25, 5.05it/s] 78%|███████▊ | 37586/48008 [2:09:00<44:14, 3.93it/s] 78%|███████▊ | 37587/48008 [2:09:01<40:59, 4.24it/s] 78%|███████▊ | 37588/48008 [2:09:01<38:40, 4.49it/s] 78%|███████▊ | 37589/48008 [2:09:01<36:56, 4.70it/s] 78%|███████▊ | 37590/48008 [2:09:01<35:51, 4.84it/s] 78%|███████▊ | 37591/48008 [2:09:01<34:40, 5.01it/s] 78%|███████▊ | 37592/48008 [2:09:02<34:13, 5.07it/s] 78%|███████▊ | 37593/48008 [2:09:02<34:34, 5.02it/s] 78%|███████▊ | 37594/48008 [2:09:02<33:51, 5.13it/s] 78%|███████▊ | 37595/48008 [2:09:02<33:01, 5.26it/s] 78%|███████▊ | 37596/48008 [2:09:02<33:12, 5.23it/s] 78%|███████▊ | 37597/48008 [2:09:03<33:04, 5.25it/s] 78%|███████▊ | 37598/48008 [2:09:03<32:47, 5.29it/s] 78%|███████▊ | 37599/48008 [2:09:03<32:53, 5.28it/s] 78%|███████▊ | 37600/48008 [2:09:03<32:58, 5.26it/s] {'loss': 0.3028, 'grad_norm': 0.1533203125, 'learning_rate': 4.336360606565573e-05, 'epoch': 0.78} 78%|███████▊ | 37600/48008 [2:09:03<32:58, 5.26it/s] 78%|███████▊ | 37601/48008 [2:09:03<33:06, 5.24it/s] 78%|███████▊ | 37602/48008 [2:09:03<33:27, 5.18it/s] 78%|███████▊ | 37603/48008 [2:09:04<33:10, 5.23it/s] 78%|███████▊ | 37604/48008 [2:09:04<43:15, 4.01it/s] 78%|███████▊ | 37605/48008 [2:09:04<40:50, 4.24it/s] 78%|███████▊ | 37606/48008 [2:09:04<40:57, 4.23it/s] 78%|███████▊ | 37607/48008 [2:09:05<39:08, 4.43it/s] 78%|███████▊ | 37608/48008 [2:09:05<37:20, 4.64it/s] 78%|███████▊ | 37609/48008 [2:09:05<36:39, 4.73it/s] 78%|███████▊ | 37610/48008 [2:09:05<41:38, 4.16it/s] 78%|███████▊ | 37611/48008 [2:09:06<39:03, 4.44it/s] 78%|███████▊ | 37612/48008 [2:09:06<37:17, 4.65it/s] 78%|███████▊ | 37613/48008 [2:09:06<37:05, 4.67it/s] 78%|███████▊ | 37614/48008 [2:09:06<35:52, 4.83it/s] 78%|███████▊ | 37615/48008 [2:09:06<35:07, 4.93it/s] 78%|███████▊ | 37616/48008 [2:09:07<35:26, 4.89it/s] 78%|███████▊ | 37617/48008 [2:09:07<34:18, 5.05it/s] 78%|███████▊ | 37618/48008 [2:09:07<33:42, 5.14it/s] 78%|███████▊ | 37619/48008 [2:09:07<33:12, 5.21it/s] 78%|███████▊ | 37620/48008 [2:09:07<33:23, 5.18it/s] 78%|███████▊ | 37621/48008 [2:09:08<33:46, 5.13it/s] 78%|███████▊ | 37622/48008 [2:09:08<33:28, 5.17it/s] 78%|███████▊ | 37623/48008 [2:09:08<33:18, 5.20it/s] 78%|███████▊ | 37624/48008 [2:09:08<33:32, 5.16it/s] 78%|███████▊ | 37625/48008 [2:09:08<39:23, 4.39it/s] 78%|███████▊ | 37626/48008 [2:09:09<37:13, 4.65it/s] 78%|███████▊ | 37627/48008 [2:09:09<35:56, 4.81it/s] 78%|███████▊ | 37628/48008 [2:09:09<34:51, 4.96it/s] 78%|███████▊ | 37629/48008 [2:09:09<34:07, 5.07it/s] 78%|███████▊ | 37630/48008 [2:09:09<33:48, 5.12it/s] 78%|███████▊ | 37631/48008 [2:09:10<33:45, 5.12it/s] 78%|███████▊ | 37632/48008 [2:09:10<33:20, 5.19it/s] 78%|███████▊ | 37633/48008 [2:09:10<33:07, 5.22it/s] 78%|███████▊ | 37634/48008 [2:09:10<33:08, 5.22it/s] 78%|███████▊ | 37635/48008 [2:09:10<33:08, 5.22it/s] 78%|███████▊ | 37636/48008 [2:09:10<33:12, 5.21it/s] 78%|███████▊ | 37637/48008 [2:09:11<32:59, 5.24it/s] 78%|███████▊ | 37638/48008 [2:09:11<33:00, 5.24it/s] 78%|███████▊ | 37639/48008 [2:09:11<32:47, 5.27it/s] 78%|███████▊ | 37640/48008 [2:09:11<32:46, 5.27it/s] 78%|███████▊ | 37641/48008 [2:09:11<33:46, 5.12it/s] 78%|███████▊ | 37642/48008 [2:09:12<33:22, 5.18it/s] 78%|███████▊ | 37643/48008 [2:09:12<33:06, 5.22it/s] 78%|███████▊ | 37644/48008 [2:09:12<32:42, 5.28it/s] 78%|███████▊ | 37645/48008 [2:09:12<32:44, 5.28it/s] 78%|███████▊ | 37646/48008 [2:09:12<33:07, 5.21it/s] 78%|███████▊ | 37647/48008 [2:09:13<32:59, 5.23it/s] 78%|███████▊ | 37648/48008 [2:09:13<32:46, 5.27it/s] 78%|███████▊ | 37649/48008 [2:09:13<32:47, 5.26it/s] 78%|███████▊ | 37650/48008 [2:09:13<33:10, 5.20it/s] {'loss': 0.2998, 'grad_norm': 0.287109375, 'learning_rate': 4.315530744875854e-05, 'epoch': 0.78} 78%|███████▊ | 37650/48008 [2:09:13<33:10, 5.20it/s] 78%|███████▊ | 37651/48008 [2:09:13<33:42, 5.12it/s] 78%|███████▊ | 37652/48008 [2:09:14<33:16, 5.19it/s] 78%|███████▊ | 37653/48008 [2:09:14<33:02, 5.22it/s] 78%|███████▊ | 37654/48008 [2:09:14<33:04, 5.22it/s] 78%|███████▊ | 37655/48008 [2:09:14<33:06, 5.21it/s] 78%|███████▊ | 37656/48008 [2:09:14<32:52, 5.25it/s] 78%|███████▊ | 37657/48008 [2:09:15<33:31, 5.15it/s] 78%|███████▊ | 37658/48008 [2:09:15<33:12, 5.19it/s] 78%|███████▊ | 37659/48008 [2:09:15<33:03, 5.22it/s] 78%|███████▊ | 37660/48008 [2:09:15<32:52, 5.25it/s] 78%|███████▊ | 37661/48008 [2:09:15<33:07, 5.21it/s] 78%|███████▊ | 37662/48008 [2:09:15<33:03, 5.22it/s] 78%|███████▊ | 37663/48008 [2:09:16<43:09, 3.99it/s] 78%|███████▊ | 37664/48008 [2:09:16<39:47, 4.33it/s] 78%|███████▊ | 37665/48008 [2:09:16<38:04, 4.53it/s] 78%|███████▊ | 37666/48008 [2:09:16<36:33, 4.71it/s] 78%|███████▊ | 37667/48008 [2:09:17<35:48, 4.81it/s] 78%|███████▊ | 37668/48008 [2:09:17<34:51, 4.94it/s] 78%|███████▊ | 37669/48008 [2:09:17<33:51, 5.09it/s] 78%|███████▊ | 37670/48008 [2:09:17<34:03, 5.06it/s] 78%|███████▊ | 37671/48008 [2:09:17<33:59, 5.07it/s] 78%|███████▊ | 37672/48008 [2:09:18<34:08, 5.05it/s] 78%|███████▊ | 37673/48008 [2:09:18<34:54, 4.93it/s] 78%|███████▊ | 37674/48008 [2:09:18<34:48, 4.95it/s] 78%|███████▊ | 37675/48008 [2:09:18<33:59, 5.07it/s] 78%|███████▊ | 37676/48008 [2:09:18<33:38, 5.12it/s] 78%|███████▊ | 37677/48008 [2:09:19<33:21, 5.16it/s] 78%|███████▊ | 37678/48008 [2:09:19<35:45, 4.82it/s] 78%|███████▊ | 37679/48008 [2:09:19<35:28, 4.85it/s] 78%|███████▊ | 37680/48008 [2:09:19<34:49, 4.94it/s] 78%|███████▊ | 37681/48008 [2:09:19<36:38, 4.70it/s] 78%|███████▊ | 37682/48008 [2:09:20<35:26, 4.85it/s] 78%|███████▊ | 37683/48008 [2:09:20<34:35, 4.97it/s] 78%|███████▊ | 37684/48008 [2:09:20<34:33, 4.98it/s] 78%|███████▊ | 37685/48008 [2:09:20<33:55, 5.07it/s] 78%|███████▊ | 37686/48008 [2:09:20<33:48, 5.09it/s] 79%|███████▊ | 37687/48008 [2:09:21<33:26, 5.14it/s] 79%|███████▊ | 37688/48008 [2:09:21<33:11, 5.18it/s] 79%|███████▊ | 37689/48008 [2:09:21<32:59, 5.21it/s] 79%|███████▊ | 37690/48008 [2:09:21<33:04, 5.20it/s] 79%|███████▊ | 37691/48008 [2:09:21<33:37, 5.11it/s] 79%|███████▊ | 37692/48008 [2:09:22<33:29, 5.13it/s] 79%|███████▊ | 37693/48008 [2:09:22<33:02, 5.20it/s] 79%|███████▊ | 37694/48008 [2:09:22<32:45, 5.25it/s] 79%|███████▊ | 37695/48008 [2:09:22<32:44, 5.25it/s] 79%|███████▊ | 37696/48008 [2:09:22<33:19, 5.16it/s] 79%|███████▊ | 37697/48008 [2:09:23<33:03, 5.20it/s] 79%|███████▊ | 37698/48008 [2:09:23<32:40, 5.26it/s] 79%|███████▊ | 37699/48008 [2:09:23<32:31, 5.28it/s] 79%|███████▊ | 37700/48008 [2:09:23<32:31, 5.28it/s] 79%|███████▊ | 37700/48008 [2:09:23<32:31, 5.28it/s]{'loss': 0.3007, 'grad_norm': 0.1455078125, 'learning_rate': 4.2947008831861356e-05, 'epoch': 0.79} 79%|███████▊ | 37701/48008 [2:09:23<32:43, 5.25it/s] 79%|███████▊ | 37702/48008 [2:09:23<33:19, 5.15it/s] 79%|███████▊ | 37703/48008 [2:09:24<33:39, 5.10it/s] 79%|███████▊ | 37704/48008 [2:09:24<33:23, 5.14it/s] 79%|███████▊ | 37705/48008 [2:09:24<33:33, 5.12it/s] 79%|███████▊ | 37706/48008 [2:09:24<33:19, 5.15it/s] 79%|███████▊ | 37707/48008 [2:09:24<33:30, 5.12it/s] 79%|███████▊ | 37708/48008 [2:09:25<33:01, 5.20it/s] 79%|███████▊ | 37709/48008 [2:09:25<33:09, 5.18it/s] 79%|███████▊ | 37710/48008 [2:09:25<33:34, 5.11it/s] 79%|███████▊ | 37711/48008 [2:09:25<33:19, 5.15it/s] 79%|███████▊ | 37712/48008 [2:09:25<32:47, 5.23it/s] 79%|███████▊ | 37713/48008 [2:09:26<32:34, 5.27it/s] 79%|███████▊ | 37714/48008 [2:09:26<32:32, 5.27it/s] 79%|███████▊ | 37715/48008 [2:09:26<32:37, 5.26it/s] 79%|███████▊ | 37716/48008 [2:09:26<32:49, 5.23it/s] 79%|███████▊ | 37717/48008 [2:09:26<32:52, 5.22it/s] 79%|███████▊ | 37718/48008 [2:09:27<32:49, 5.22it/s] 79%|███████▊ | 37719/48008 [2:09:27<32:45, 5.24it/s] 79%|███████▊ | 37720/48008 [2:09:27<32:33, 5.27it/s] 79%|███████▊ | 37721/48008 [2:09:27<38:34, 4.44it/s] 79%|███████▊ | 37722/48008 [2:09:27<36:37, 4.68it/s] 79%|███████▊ | 37723/48008 [2:09:28<35:55, 4.77it/s] 79%|███████▊ | 37724/48008 [2:09:28<34:59, 4.90it/s] 79%|███████▊ | 37725/48008 [2:09:28<34:40, 4.94it/s] 79%|███████▊ | 37726/48008 [2:09:28<34:02, 5.04it/s] 79%|███████▊ | 37727/48008 [2:09:28<33:14, 5.16it/s] 79%|███████▊ | 37728/48008 [2:09:29<32:57, 5.20it/s] 79%|███████▊ | 37729/48008 [2:09:29<35:20, 4.85it/s] 79%|███████▊ | 37730/48008 [2:09:29<34:36, 4.95it/s] 79%|███████▊ | 37731/48008 [2:09:29<34:06, 5.02it/s] 79%|███████▊ | 37732/48008 [2:09:29<33:35, 5.10it/s] 79%|███████▊ | 37733/48008 [2:09:30<33:14, 5.15it/s] 79%|███████▊ | 37734/48008 [2:09:30<32:52, 5.21it/s] 79%|███████▊ | 37735/48008 [2:09:30<33:10, 5.16it/s] 79%|███████▊ | 37736/48008 [2:09:30<33:41, 5.08it/s] 79%|███████▊ | 37737/48008 [2:09:31<43:30, 3.93it/s] 79%|███████▊ | 37738/48008 [2:09:31<40:51, 4.19it/s] 79%|███████▊ | 37739/48008 [2:09:31<38:55, 4.40it/s] 79%|███████▊ | 37740/48008 [2:09:31<37:04, 4.62it/s] 79%|███████▊ | 37741/48008 [2:09:31<35:44, 4.79it/s] 79%|███████▊ | 37742/48008 [2:09:32<34:24, 4.97it/s] 79%|███████▊ | 37743/48008 [2:09:32<33:50, 5.06it/s] 79%|███████▊ | 37744/48008 [2:09:32<33:23, 5.12it/s] 79%|███████▊ | 37745/48008 [2:09:32<33:11, 5.15it/s] 79%|███████▊ | 37746/48008 [2:09:32<33:39, 5.08it/s] 79%|███████▊ | 37747/48008 [2:09:33<35:42, 4.79it/s] 79%|███████▊ | 37748/48008 [2:09:33<34:42, 4.93it/s] 79%|███████▊ | 37749/48008 [2:09:33<34:03, 5.02it/s] 79%|███████▊ | 37750/48008 [2:09:33<33:33, 5.09it/s] 79%|███████▊ | 37750/48008 [2:09:33<33:33, 5.09it/s]{'loss': 0.301, 'grad_norm': 0.2578125, 'learning_rate': 4.2738710214964174e-05, 'epoch': 0.79} 79%|███████▊ | 37751/48008 [2:09:33<32:48, 5.21it/s] 79%|███████▊ | 37752/48008 [2:09:34<38:45, 4.41it/s] 79%|███████▊ | 37753/48008 [2:09:34<36:56, 4.63it/s] 79%|███████▊ | 37754/48008 [2:09:34<35:45, 4.78it/s] 79%|███████▊ | 37755/48008 [2:09:34<34:53, 4.90it/s] 79%|███████▊ | 37756/48008 [2:09:34<34:31, 4.95it/s] 79%|███████▊ | 37757/48008 [2:09:35<33:56, 5.03it/s] 79%|███████▊ | 37758/48008 [2:09:35<33:34, 5.09it/s] 79%|███████▊ | 37759/48008 [2:09:35<39:10, 4.36it/s] 79%|███████▊ | 37760/48008 [2:09:35<36:54, 4.63it/s] 79%|███████▊ | 37761/48008 [2:09:35<36:13, 4.71it/s] 79%|███████▊ | 37762/48008 [2:09:36<35:12, 4.85it/s] 79%|███████▊ | 37763/48008 [2:09:36<34:18, 4.98it/s] 79%|███████▊ | 37764/48008 [2:09:36<34:32, 4.94it/s] 79%|███████▊ | 37765/48008 [2:09:36<34:01, 5.02it/s] 79%|███████▊ | 37766/48008 [2:09:36<33:33, 5.09it/s] 79%|███████▊ | 37767/48008 [2:09:37<33:12, 5.14it/s] 79%|███████▊ | 37768/48008 [2:09:37<32:58, 5.18it/s] 79%|███████▊ | 37769/48008 [2:09:37<42:49, 3.98it/s] 79%|███████▊ | 37770/48008 [2:09:37<39:29, 4.32it/s] 79%|███████▊ | 37771/48008 [2:09:38<37:22, 4.56it/s] 79%|███████▊ | 37772/48008 [2:09:38<36:27, 4.68it/s] 79%|███████▊ | 37773/48008 [2:09:38<35:16, 4.84it/s] 79%|███████▊ | 37774/48008 [2:09:38<34:39, 4.92it/s] 79%|███████▊ | 37775/48008 [2:09:38<34:39, 4.92it/s] 79%|███████▊ | 37776/48008 [2:09:39<33:55, 5.03it/s] 79%|███████▊ | 37777/48008 [2:09:39<33:55, 5.03it/s] 79%|███████▊ | 37778/48008 [2:09:39<33:53, 5.03it/s] 79%|███████▊ | 37779/48008 [2:09:39<33:32, 5.08it/s] 79%|███████▊ | 37780/48008 [2:09:39<33:16, 5.12it/s] 79%|███████▊ | 37781/48008 [2:09:40<33:21, 5.11it/s] 79%|███████▊ | 37782/48008 [2:09:40<33:09, 5.14it/s] 79%|███████▊ | 37783/48008 [2:09:40<33:04, 5.15it/s] 79%|███████▊ | 37784/48008 [2:09:40<32:55, 5.18it/s] 79%|███████▊ | 37785/48008 [2:09:40<32:55, 5.17it/s] 79%|███████▊ | 37786/48008 [2:09:40<32:48, 5.19it/s] 79%|███████▊ | 37787/48008 [2:09:41<33:20, 5.11it/s] 79%|███████▊ | 37788/48008 [2:09:41<33:35, 5.07it/s] 79%|███████▊ | 37789/48008 [2:09:41<33:58, 5.01it/s] 79%|███████▊ | 37790/48008 [2:09:41<33:28, 5.09it/s] 79%|███████▊ | 37791/48008 [2:09:41<33:28, 5.09it/s] 79%|███████▊ | 37792/48008 [2:09:42<39:10, 4.35it/s] 79%|███████▊ | 37793/48008 [2:09:42<37:47, 4.50it/s] 79%|███████▊ | 37794/48008 [2:09:42<36:05, 4.72it/s] 79%|███████▊ | 37795/48008 [2:09:42<35:30, 4.79it/s] 79%|███████▊ | 37796/48008 [2:09:43<34:29, 4.93it/s] 79%|███████▊ | 37797/48008 [2:09:43<33:49, 5.03it/s] 79%|███████▊ | 37798/48008 [2:09:43<33:20, 5.10it/s] 79%|███████▊ | 37799/48008 [2:09:43<35:21, 4.81it/s] 79%|███████▊ | 37800/48008 [2:09:43<34:27, 4.94it/s] {'loss': 0.2992, 'grad_norm': 0.21484375, 'learning_rate': 4.253041159806699e-05, 'epoch': 0.79} 79%|███████▊ | 37800/48008 [2:09:43<34:27, 4.94it/s] 79%|███████▊ | 37801/48008 [2:09:44<33:51, 5.02it/s] 79%|███████▊ | 37802/48008 [2:09:44<34:10, 4.98it/s] 79%|███████▊ | 37803/48008 [2:09:44<33:39, 5.05it/s] 79%|███████▊ | 37804/48008 [2:09:44<34:10, 4.98it/s] 79%|███████▊ | 37805/48008 [2:09:44<34:15, 4.96it/s] 79%|███████▊ | 37806/48008 [2:09:45<33:57, 5.01it/s] 79%|███████▉ | 37807/48008 [2:09:45<33:34, 5.06it/s] 79%|███████▉ | 37808/48008 [2:09:45<33:23, 5.09it/s] 79%|███████▉ | 37809/48008 [2:09:45<32:57, 5.16it/s] 79%|███████▉ | 37810/48008 [2:09:45<33:14, 5.11it/s] 79%|███████▉ | 37811/48008 [2:09:46<33:19, 5.10it/s] 79%|███████▉ | 37812/48008 [2:09:46<33:09, 5.12it/s] 79%|███████▉ | 37813/48008 [2:09:46<33:08, 5.13it/s] 79%|███████▉ | 37814/48008 [2:09:46<32:48, 5.18it/s] 79%|███████▉ | 37815/48008 [2:09:46<32:36, 5.21it/s] 79%|███████▉ | 37816/48008 [2:09:47<33:09, 5.12it/s] 79%|███████▉ | 37817/48008 [2:09:47<32:47, 5.18it/s] 79%|███████▉ | 37818/48008 [2:09:47<33:41, 5.04it/s] 79%|███████▉ | 37819/48008 [2:09:47<33:18, 5.10it/s] 79%|███████▉ | 37820/48008 [2:09:47<32:57, 5.15it/s] 79%|███████▉ | 37821/48008 [2:09:47<33:00, 5.14it/s] 79%|███████▉ | 37822/48008 [2:09:48<42:46, 3.97it/s] 79%|███████▉ | 37823/48008 [2:09:48<39:41, 4.28it/s] 79%|███████▉ | 37824/48008 [2:09:48<38:08, 4.45it/s] 79%|███████▉ | 37825/48008 [2:09:48<37:06, 4.57it/s] 79%|███████▉ | 37826/48008 [2:09:49<38:00, 4.47it/s] 79%|███████▉ | 37827/48008 [2:09:49<46:17, 3.67it/s] 79%|███████▉ | 37828/48008 [2:09:49<44:23, 3.82it/s] 79%|███████▉ | 37829/48008 [2:09:50<40:47, 4.16it/s] 79%|███████▉ | 37830/48008 [2:09:50<38:04, 4.46it/s] 79%|███████▉ | 37831/48008 [2:09:50<36:01, 4.71it/s] 79%|███████▉ | 37832/48008 [2:09:50<37:13, 4.56it/s] 79%|███████▉ | 37833/48008 [2:09:50<35:47, 4.74it/s] 79%|███████▉ | 37834/48008 [2:09:51<34:48, 4.87it/s] 79%|███████▉ | 37835/48008 [2:09:51<33:49, 5.01it/s] 79%|███████▉ | 37836/48008 [2:09:51<33:10, 5.11it/s] 79%|███████▉ | 37837/48008 [2:09:51<32:58, 5.14it/s] 79%|███████▉ | 37838/48008 [2:09:51<32:45, 5.17it/s] 79%|███████▉ | 37839/48008 [2:09:51<33:15, 5.10it/s] 79%|███████▉ | 37840/48008 [2:09:52<32:40, 5.19it/s] 79%|███████▉ | 37841/48008 [2:09:52<32:37, 5.19it/s] 79%|███████▉ | 37842/48008 [2:09:52<33:11, 5.11it/s] 79%|███████▉ | 37843/48008 [2:09:52<32:45, 5.17it/s] 79%|███████▉ | 37844/48008 [2:09:52<35:01, 4.84it/s] 79%|███████▉ | 37845/48008 [2:09:53<35:05, 4.83it/s] 79%|███████▉ | 37846/48008 [2:09:53<34:58, 4.84it/s] 79%|███████▉ | 37847/48008 [2:09:53<34:01, 4.98it/s] 79%|███████▉ | 37848/48008 [2:09:53<34:03, 4.97it/s] 79%|███████▉ | 37849/48008 [2:09:53<33:08, 5.11it/s] 79%|███████▉ | 37850/48008 [2:09:54<32:39, 5.18it/s] {'loss': 0.3002, 'grad_norm': 0.244140625, 'learning_rate': 4.232211298116981e-05, 'epoch': 0.79} 79%|███████▉ | 37850/48008 [2:09:54<32:39, 5.18it/s] 79%|███████▉ | 37851/48008 [2:09:54<32:39, 5.18it/s] 79%|███████▉ | 37852/48008 [2:09:54<33:35, 5.04it/s] 79%|███████▉ | 37853/48008 [2:09:54<33:12, 5.10it/s] 79%|███████▉ | 37854/48008 [2:09:54<32:52, 5.15it/s] 79%|███████▉ | 37855/48008 [2:09:55<32:43, 5.17it/s] 79%|███████▉ | 37856/48008 [2:09:55<32:29, 5.21it/s] 79%|███████▉ | 37857/48008 [2:09:55<33:03, 5.12it/s] 79%|███████▉ | 37858/48008 [2:09:55<33:11, 5.10it/s] 79%|███████▉ | 37859/48008 [2:09:55<33:17, 5.08it/s] 79%|███████▉ | 37860/48008 [2:09:56<33:30, 5.05it/s] 79%|███████▉ | 37861/48008 [2:09:56<32:54, 5.14it/s] 79%|███████▉ | 37862/48008 [2:09:56<32:43, 5.17it/s] 79%|███████▉ | 37863/48008 [2:09:56<32:35, 5.19it/s] 79%|███████▉ | 37864/48008 [2:09:56<32:20, 5.23it/s] 79%|███████▉ | 37865/48008 [2:09:57<32:17, 5.23it/s] 79%|███████▉ | 37866/48008 [2:09:57<32:53, 5.14it/s] 79%|███████▉ | 37867/48008 [2:09:57<32:48, 5.15it/s] 79%|███████▉ | 37868/48008 [2:09:57<32:31, 5.20it/s] 79%|███████▉ | 37869/48008 [2:09:57<32:29, 5.20it/s] 79%|███████▉ | 37870/48008 [2:09:58<42:16, 4.00it/s] 79%|███████▉ | 37871/48008 [2:09:58<39:02, 4.33it/s] 79%|███████▉ | 37872/48008 [2:09:58<37:33, 4.50it/s] 79%|███████▉ | 37873/48008 [2:09:58<35:51, 4.71it/s] 79%|███████▉ | 37874/48008 [2:09:59<37:04, 4.56it/s] 79%|███████▉ | 37875/48008 [2:09:59<35:22, 4.77it/s] 79%|███████▉ | 37876/48008 [2:09:59<44:16, 3.81it/s] 79%|███████▉ | 37877/48008 [2:09:59<40:30, 4.17it/s] 79%|███████▉ | 37878/48008 [2:10:00<40:19, 4.19it/s] 79%|███████▉ | 37879/48008 [2:10:00<37:38, 4.49it/s] 79%|███████▉ | 37880/48008 [2:10:00<35:50, 4.71it/s] 79%|███████▉ | 37881/48008 [2:10:00<34:36, 4.88it/s] 79%|███████▉ | 37882/48008 [2:10:00<34:04, 4.95it/s] 79%|███████▉ | 37883/48008 [2:10:00<33:26, 5.05it/s] 79%|███████▉ | 37884/48008 [2:10:01<43:00, 3.92it/s] 79%|███████▉ | 37885/48008 [2:10:01<39:32, 4.27it/s] 79%|███████▉ | 37886/48008 [2:10:01<37:26, 4.51it/s] 79%|███████▉ | 37887/48008 [2:10:01<35:54, 4.70it/s] 79%|███████▉ | 37888/48008 [2:10:02<34:24, 4.90it/s] 79%|███████▉ | 37889/48008 [2:10:02<33:47, 4.99it/s] 79%|███████▉ | 37890/48008 [2:10:02<33:25, 5.05it/s] 79%|███████▉ | 37891/48008 [2:10:02<33:06, 5.09it/s] 79%|███████▉ | 37892/48008 [2:10:02<35:12, 4.79it/s] 79%|███████▉ | 37893/48008 [2:10:03<34:23, 4.90it/s] 79%|███████▉ | 37894/48008 [2:10:03<33:37, 5.01it/s] 79%|███████▉ | 37895/48008 [2:10:03<33:44, 4.99it/s] 79%|███████▉ | 37896/48008 [2:10:03<33:31, 5.03it/s] 79%|███████▉ | 37897/48008 [2:10:03<33:31, 5.03it/s] 79%|███████▉ | 37898/48008 [2:10:04<32:58, 5.11it/s] 79%|███████▉ | 37899/48008 [2:10:04<32:34, 5.17it/s] 79%|███████▉ | 37900/48008 [2:10:04<32:29, 5.18it/s] {'loss': 0.3001, 'grad_norm': 0.22265625, 'learning_rate': 4.2113814364272626e-05, 'epoch': 0.79} 79%|███████▉ | 37900/48008 [2:10:04<32:29, 5.18it/s] 79%|███████▉ | 37901/48008 [2:10:04<32:43, 5.15it/s] 79%|███████▉ | 37902/48008 [2:10:04<32:43, 5.15it/s] 79%|███████▉ | 37903/48008 [2:10:05<32:24, 5.20it/s] 79%|███████▉ | 37904/48008 [2:10:05<32:35, 5.17it/s] 79%|███████▉ | 37905/48008 [2:10:05<32:18, 5.21it/s] 79%|███████▉ | 37906/48008 [2:10:05<32:51, 5.12it/s] 79%|███████▉ | 37907/48008 [2:10:05<32:51, 5.12it/s] 79%|███████▉ | 37908/48008 [2:10:06<32:44, 5.14it/s] 79%|███████▉ | 37909/48008 [2:10:06<33:31, 5.02it/s] 79%|███████▉ | 37910/48008 [2:10:06<32:56, 5.11it/s] 79%|███████▉ | 37911/48008 [2:10:06<32:44, 5.14it/s] 79%|███████▉ | 37912/48008 [2:10:06<32:51, 5.12it/s] 79%|███████▉ | 37913/48008 [2:10:07<32:33, 5.17it/s] 79%|███████▉ | 37914/48008 [2:10:07<32:56, 5.11it/s] 79%|███████▉ | 37915/48008 [2:10:07<33:02, 5.09it/s] 79%|███████▉ | 37916/48008 [2:10:07<32:39, 5.15it/s] 79%|███████▉ | 37917/48008 [2:10:07<32:35, 5.16it/s] 79%|███████▉ | 37918/48008 [2:10:07<32:47, 5.13it/s] 79%|███████▉ | 37919/48008 [2:10:08<32:31, 5.17it/s] 79%|███████▉ | 37920/48008 [2:10:08<34:47, 4.83it/s] 79%|███████▉ | 37921/48008 [2:10:08<33:57, 4.95it/s] 79%|███████▉ | 37922/48008 [2:10:08<33:30, 5.02it/s] 79%|███████▉ | 37923/48008 [2:10:08<33:12, 5.06it/s] 79%|███████▉ | 37924/48008 [2:10:09<38:38, 4.35it/s] 79%|███████▉ | 37925/48008 [2:10:09<36:55, 4.55it/s] 79%|███████▉ | 37926/48008 [2:10:09<35:30, 4.73it/s] 79%|███████▉ | 37927/48008 [2:10:09<34:23, 4.89it/s] 79%|███████▉ | 37928/48008 [2:10:10<35:03, 4.79it/s] 79%|███████▉ | 37929/48008 [2:10:10<34:12, 4.91it/s] 79%|███████▉ | 37930/48008 [2:10:10<33:40, 4.99it/s] 79%|███████▉ | 37931/48008 [2:10:10<33:25, 5.03it/s] 79%|███████▉ | 37932/48008 [2:10:10<33:55, 4.95it/s] 79%|███████▉ | 37933/48008 [2:10:11<33:28, 5.02it/s] 79%|███████▉ | 37934/48008 [2:10:11<33:35, 5.00it/s] 79%|███████▉ | 37935/48008 [2:10:11<33:18, 5.04it/s] 79%|███████▉ | 37936/48008 [2:10:11<33:52, 4.96it/s] 79%|███████▉ | 37937/48008 [2:10:11<33:04, 5.07it/s] 79%|███████▉ | 37938/48008 [2:10:12<32:48, 5.12it/s] 79%|███████▉ | 37939/48008 [2:10:12<32:25, 5.18it/s] 79%|███████▉ | 37940/48008 [2:10:12<32:20, 5.19it/s] 79%|███████▉ | 37941/48008 [2:10:12<32:21, 5.18it/s] 79%|███████▉ | 37942/48008 [2:10:12<32:09, 5.22it/s] 79%|███████▉ | 37943/48008 [2:10:13<32:04, 5.23it/s] 79%|███████▉ | 37944/48008 [2:10:13<32:36, 5.14it/s] 79%|███████▉ | 37945/48008 [2:10:13<32:56, 5.09it/s] 79%|███████▉ | 37946/48008 [2:10:13<32:27, 5.17it/s] 79%|███████▉ | 37947/48008 [2:10:13<34:39, 4.84it/s] 79%|███████▉ | 37948/48008 [2:10:14<33:34, 4.99it/s] 79%|███████▉ | 37949/48008 [2:10:14<33:55, 4.94it/s] 79%|███████▉ | 37950/48008 [2:10:14<33:23, 5.02it/s] {'loss': 0.301, 'grad_norm': 0.275390625, 'learning_rate': 4.190551574737544e-05, 'epoch': 0.79} 79%|███████▉ | 37950/48008 [2:10:14<33:23, 5.02it/s] 79%|███████▉ | 37951/48008 [2:10:14<33:31, 5.00it/s] 79%|███████▉ | 37952/48008 [2:10:14<33:32, 5.00it/s] 79%|███████▉ | 37953/48008 [2:10:15<32:58, 5.08it/s] 79%|███████▉ | 37954/48008 [2:10:15<38:27, 4.36it/s] 79%|███████▉ | 37955/48008 [2:10:15<37:00, 4.53it/s] 79%|███████▉ | 37956/48008 [2:10:15<35:16, 4.75it/s] 79%|███████▉ | 37957/48008 [2:10:15<35:05, 4.77it/s] 79%|███████▉ | 37958/48008 [2:10:16<33:49, 4.95it/s] 79%|███████▉ | 37959/48008 [2:10:16<43:04, 3.89it/s] 79%|███████▉ | 37960/48008 [2:10:16<39:33, 4.23it/s] 79%|███████▉ | 37961/48008 [2:10:16<37:53, 4.42it/s] 79%|███████▉ | 37962/48008 [2:10:17<36:12, 4.63it/s] 79%|███████▉ | 37963/48008 [2:10:17<37:11, 4.50it/s] 79%|███████▉ | 37964/48008 [2:10:17<35:37, 4.70it/s] 79%|███████▉ | 37965/48008 [2:10:17<34:23, 4.87it/s] 79%|███████▉ | 37966/48008 [2:10:17<34:03, 4.92it/s] 79%|███████▉ | 37967/48008 [2:10:18<39:07, 4.28it/s] 79%|███████▉ | 37968/48008 [2:10:18<36:50, 4.54it/s] 79%|███████▉ | 37969/48008 [2:10:18<35:18, 4.74it/s] 79%|███████▉ | 37970/48008 [2:10:18<34:27, 4.86it/s] 79%|███████▉ | 37971/48008 [2:10:18<33:43, 4.96it/s] 79%|███████▉ | 37972/48008 [2:10:19<39:00, 4.29it/s] 79%|███████▉ | 37973/48008 [2:10:19<37:47, 4.43it/s] 79%|███████▉ | 37974/48008 [2:10:19<36:51, 4.54it/s] 79%|███████▉ | 37975/48008 [2:10:19<36:20, 4.60it/s] 79%|███████▉ | 37976/48008 [2:10:20<34:48, 4.80it/s] 79%|███████▉ | 37977/48008 [2:10:20<33:56, 4.93it/s] 79%|███████▉ | 37978/48008 [2:10:20<35:33, 4.70it/s] 79%|███████▉ | 37979/48008 [2:10:20<34:55, 4.79it/s] 79%|███████▉ | 37980/48008 [2:10:21<1:23:12, 2.01it/s] 79%|███████▉ | 37981/48008 [2:10:22<1:08:06, 2.45it/s] 79%|███████▉ | 37982/48008 [2:10:22<56:59, 2.93it/s] 79%|███████▉ | 37983/48008 [2:10:22<49:13, 3.39it/s] 79%|███████▉ | 37984/48008 [2:10:22<44:05, 3.79it/s] 79%|███████▉ | 37985/48008 [2:10:22<40:59, 4.08it/s] 79%|███████▉ | 37986/48008 [2:10:23<38:36, 4.33it/s] 79%|███████▉ | 37987/48008 [2:10:23<36:22, 4.59it/s] 79%|███████▉ | 37988/48008 [2:10:23<35:21, 4.72it/s] 79%|███████▉ | 37989/48008 [2:10:23<34:21, 4.86it/s] 79%|███████▉ | 37990/48008 [2:10:23<33:31, 4.98it/s] 79%|███████▉ | 37991/48008 [2:10:23<33:37, 4.97it/s] 79%|███████▉ | 37992/48008 [2:10:24<32:59, 5.06it/s] 79%|███████▉ | 37993/48008 [2:10:24<32:37, 5.12it/s] 79%|███████▉ | 37994/48008 [2:10:24<33:01, 5.05it/s] 79%|███████▉ | 37995/48008 [2:10:24<32:28, 5.14it/s] 79%|███████▉ | 37996/48008 [2:10:24<32:49, 5.08it/s] 79%|███████▉ | 37997/48008 [2:10:25<32:14, 5.17it/s] 79%|███████▉ | 37998/48008 [2:10:25<32:09, 5.19it/s] 79%|███████▉ | 37999/48008 [2:10:25<32:14, 5.17it/s] 79%|███████▉ | 38000/48008 [2:10:26<1:06:30, 2.51it/s] {'loss': 0.3006, 'grad_norm': 0.435546875, 'learning_rate': 4.1697217130478254e-05, 'epoch': 0.79} 79%|███████▉ | 38000/48008 [2:10:26<1:06:30, 2.51it/s] 79%|███████▉ | 38001/48008 [2:10:26<56:43, 2.94it/s] 79%|███████▉ | 38002/48008 [2:10:26<49:16, 3.38it/s] 79%|███████▉ | 38003/48008 [2:10:26<44:01, 3.79it/s] 79%|███████▉ | 38004/48008 [2:10:27<40:43, 4.09it/s] 79%|███████▉ | 38005/48008 [2:10:27<37:45, 4.42it/s] 79%|███████▉ | 38006/48008 [2:10:28<1:07:28, 2.47it/s] 79%|███████▉ | 38007/48008 [2:10:28<1:06:36, 2.50it/s] 79%|███████▉ | 38008/48008 [2:10:28<1:01:53, 2.69it/s] 79%|███████▉ | 38009/48008 [2:10:29<52:52, 3.15it/s] 79%|███████▉ | 38010/48008 [2:10:29<46:36, 3.58it/s] 79%|███████▉ | 38011/48008 [2:10:29<42:13, 3.95it/s] 79%|███████▉ | 38012/48008 [2:10:29<40:02, 4.16it/s] 79%|███████▉ | 38013/48008 [2:10:29<37:24, 4.45it/s] 79%|███████▉ | 38014/48008 [2:10:30<36:31, 4.56it/s] 79%|███████▉ | 38015/48008 [2:10:30<34:51, 4.78it/s] 79%|███████▉ | 38016/48008 [2:10:30<36:12, 4.60it/s] 79%|███████▉ | 38017/48008 [2:10:30<35:25, 4.70it/s] 79%|███████▉ | 38018/48008 [2:10:31<1:13:18, 2.27it/s] 79%|███████▉ | 38019/48008 [2:10:31<1:00:53, 2.73it/s] 79%|███████▉ | 38020/48008 [2:10:32<52:45, 3.16it/s] 79%|███████▉ | 38021/48008 [2:10:32<46:13, 3.60it/s] 79%|███████▉ | 38022/48008 [2:10:32<41:37, 4.00it/s] 79%|███████▉ | 38023/48008 [2:10:32<38:44, 4.30it/s] 79%|███████▉ | 38024/48008 [2:10:33<1:08:30, 2.43it/s] 79%|███████▉ | 38025/48008 [2:10:33<57:34, 2.89it/s] 79%|███████▉ | 38026/48008 [2:10:33<50:28, 3.30it/s] 79%|███████▉ | 38027/48008 [2:10:34<45:32, 3.65it/s] 79%|███████▉ | 38028/48008 [2:10:34<41:09, 4.04it/s] 79%|███████▉ | 38029/48008 [2:10:34<38:19, 4.34it/s] 79%|███████▉ | 38030/48008 [2:10:34<36:23, 4.57it/s] 79%|███████▉ | 38031/48008 [2:10:34<34:44, 4.79it/s] 79%|███████▉ | 38032/48008 [2:10:35<34:34, 4.81it/s] 79%|███████▉ | 38033/48008 [2:10:35<33:52, 4.91it/s] 79%|███████▉ | 38034/48008 [2:10:35<33:13, 5.00it/s] 79%|███████▉ | 38035/48008 [2:10:35<33:49, 4.91it/s] 79%|███████▉ | 38036/48008 [2:10:35<33:09, 5.01it/s] 79%|███████▉ | 38037/48008 [2:10:35<32:45, 5.07it/s] 79%|███████▉ | 38038/48008 [2:10:36<31:59, 5.19it/s] 79%|███████▉ | 38039/48008 [2:10:36<31:43, 5.24it/s] 79%|███████▉ | 38040/48008 [2:10:36<31:29, 5.28it/s] 79%|███████▉ | 38041/48008 [2:10:36<31:37, 5.25it/s] 79%|███████▉ | 38042/48008 [2:10:36<31:41, 5.24it/s] 79%|███████▉ | 38043/48008 [2:10:37<32:02, 5.18it/s] 79%|███████▉ | 38044/48008 [2:10:37<31:51, 5.21it/s] 79%|███████▉ | 38045/48008 [2:10:37<32:00, 5.19it/s] 79%|███████▉ | 38046/48008 [2:10:37<32:05, 5.17it/s] 79%|███████▉ | 38047/48008 [2:10:37<31:46, 5.23it/s] 79%|███████▉ | 38048/48008 [2:10:38<34:08, 4.86it/s] 79%|███████▉ | 38049/48008 [2:10:38<33:27, 4.96it/s] 79%|███████▉ | 38050/48008 [2:10:38<33:22, 4.97it/s] {'loss': 0.302, 'grad_norm': 0.255859375, 'learning_rate': 4.148891851358107e-05, 'epoch': 0.79} 79%|███████▉ | 38050/48008 [2:10:38<33:22, 4.97it/s] 79%|███████▉ | 38051/48008 [2:10:38<32:54, 5.04it/s] 79%|███████▉ | 38052/48008 [2:10:38<32:19, 5.13it/s] 79%|███████▉ | 38053/48008 [2:10:39<41:49, 3.97it/s] 79%|███████▉ | 38054/48008 [2:10:39<39:06, 4.24it/s] 79%|███████▉ | 38055/48008 [2:10:39<36:54, 4.49it/s] 79%|███████▉ | 38056/48008 [2:10:39<35:51, 4.63it/s] 79%|███████▉ | 38057/48008 [2:10:40<35:09, 4.72it/s] 79%|███████▉ | 38058/48008 [2:10:40<34:07, 4.86it/s] 79%|███████▉ | 38059/48008 [2:10:40<33:13, 4.99it/s] 79%|███████▉ | 38060/48008 [2:10:41<1:04:27, 2.57it/s] 79%|███████▉ | 38061/48008 [2:10:41<54:47, 3.03it/s] 79%|███████▉ | 38062/48008 [2:10:41<47:46, 3.47it/s] 79%|███████▉ | 38063/48008 [2:10:41<42:59, 3.86it/s] 79%|███████▉ | 38064/48008 [2:10:42<40:08, 4.13it/s] 79%|███████▉ | 38065/48008 [2:10:42<37:32, 4.41it/s] 79%|███████▉ | 38066/48008 [2:10:42<35:49, 4.63it/s] 79%|███████▉ | 38067/48008 [2:10:42<34:19, 4.83it/s] 79%|███████▉ | 38068/48008 [2:10:42<33:28, 4.95it/s] 79%|███████▉ | 38069/48008 [2:10:43<33:07, 5.00it/s] 79%|███████▉ | 38070/48008 [2:10:43<32:34, 5.08it/s] 79%|███████▉ | 38071/48008 [2:10:43<32:06, 5.16it/s] 79%|███████▉ | 38072/48008 [2:10:43<31:56, 5.18it/s] 79%|███████▉ | 38073/48008 [2:10:43<31:51, 5.20it/s] 79%|███████▉ | 38074/48008 [2:10:43<31:35, 5.24it/s] 79%|███████▉ | 38075/48008 [2:10:44<32:11, 5.14it/s] 79%|███████▉ | 38076/48008 [2:10:44<32:01, 5.17it/s] 79%|███████▉ | 38077/48008 [2:10:44<32:49, 5.04it/s] 79%|███████▉ | 38078/48008 [2:10:44<32:56, 5.02it/s] 79%|███████▉ | 38079/48008 [2:10:44<32:33, 5.08it/s] 79%|███████▉ | 38080/48008 [2:10:45<32:32, 5.08it/s] 79%|███████▉ | 38081/48008 [2:10:45<32:19, 5.12it/s] 79%|███████▉ | 38082/48008 [2:10:45<32:12, 5.14it/s] 79%|███████▉ | 38083/48008 [2:10:45<31:53, 5.19it/s] 79%|███████▉ | 38084/48008 [2:10:45<32:23, 5.11it/s] 79%|███████▉ | 38085/48008 [2:10:46<32:43, 5.05it/s] 79%|███████▉ | 38086/48008 [2:10:46<32:24, 5.10it/s] 79%|███████▉ | 38087/48008 [2:10:46<32:11, 5.14it/s] 79%|███████▉ | 38088/48008 [2:10:46<32:02, 5.16it/s] 79%|███████▉ | 38089/48008 [2:10:46<32:30, 5.09it/s] 79%|███████▉ | 38090/48008 [2:10:47<32:15, 5.12it/s] 79%|███████▉ | 38091/48008 [2:10:47<32:26, 5.10it/s] 79%|███████▉ | 38092/48008 [2:10:47<32:19, 5.11it/s] 79%|███████▉ | 38093/48008 [2:10:47<32:43, 5.05it/s] 79%|███████▉ | 38094/48008 [2:10:47<32:18, 5.11it/s] 79%|███████▉ | 38095/48008 [2:10:48<32:55, 5.02it/s] 79%|███████▉ | 38096/48008 [2:10:48<32:33, 5.07it/s] 79%|███████▉ | 38097/48008 [2:10:48<31:56, 5.17it/s] 79%|███████▉ | 38098/48008 [2:10:48<32:04, 5.15it/s] 79%|███████▉ | 38099/48008 [2:10:48<31:51, 5.19it/s] 79%|███████▉ | 38100/48008 [2:10:49<32:18, 5.11it/s] {'loss': 0.3, 'grad_norm': 0.1474609375, 'learning_rate': 4.128061989668389e-05, 'epoch': 0.79} 79%|███████▉ | 38100/48008 [2:10:49<32:18, 5.11it/s] 79%|███████▉ | 38101/48008 [2:10:49<31:49, 5.19it/s] 79%|███████▉ | 38102/48008 [2:10:49<31:49, 5.19it/s] 79%|███████▉ | 38103/48008 [2:10:49<31:44, 5.20it/s] 79%|███████▉ | 38104/48008 [2:10:49<31:38, 5.22it/s] 79%|███████▉ | 38105/48008 [2:10:50<31:32, 5.23it/s] 79%|███████▉ | 38106/48008 [2:10:50<31:14, 5.28it/s] 79%|███████▉ | 38107/48008 [2:10:50<31:26, 5.25it/s] 79%|███████▉ | 38108/48008 [2:10:50<31:39, 5.21it/s] 79%|███████▉ | 38109/48008 [2:10:50<31:55, 5.17it/s] 79%|███████▉ | 38110/48008 [2:10:50<31:54, 5.17it/s] 79%|███████▉ | 38111/48008 [2:10:51<31:50, 5.18it/s] 79%|███████▉ | 38112/48008 [2:10:51<32:34, 5.06it/s] 79%|███████▉ | 38113/48008 [2:10:51<32:14, 5.11it/s] 79%|███████▉ | 38114/48008 [2:10:51<32:31, 5.07it/s] 79%|███████▉ | 38115/48008 [2:10:51<32:05, 5.14it/s] 79%|███████▉ | 38116/48008 [2:10:52<31:53, 5.17it/s] 79%|███████▉ | 38117/48008 [2:10:52<31:26, 5.24it/s] 79%|███████▉ | 38118/48008 [2:10:52<31:15, 5.27it/s] 79%|███████▉ | 38119/48008 [2:10:52<31:14, 5.28it/s] 79%|███████▉ | 38120/48008 [2:10:52<31:00, 5.31it/s] 79%|███████▉ | 38121/48008 [2:10:53<31:05, 5.30it/s] 79%|███████▉ | 38122/48008 [2:10:53<31:14, 5.27it/s] 79%|███████▉ | 38123/48008 [2:10:53<31:13, 5.28it/s] 79%|███████▉ | 38124/48008 [2:10:53<33:27, 4.92it/s] 79%|███████▉ | 38125/48008 [2:10:53<32:45, 5.03it/s] 79%|███████▉ | 38126/48008 [2:10:54<32:02, 5.14it/s] 79%|███████▉ | 38127/48008 [2:10:54<31:50, 5.17it/s] 79%|███████▉ | 38128/48008 [2:10:54<31:28, 5.23it/s] 79%|███████▉ | 38129/48008 [2:10:54<31:23, 5.24it/s] 79%|███████▉ | 38130/48008 [2:10:54<31:34, 5.21it/s] 79%|███████▉ | 38131/48008 [2:10:55<31:19, 5.25it/s] 79%|███████▉ | 38132/48008 [2:10:55<31:21, 5.25it/s] 79%|███████▉ | 38133/48008 [2:10:55<31:52, 5.16it/s] 79%|███████▉ | 38134/48008 [2:10:55<31:41, 5.19it/s] 79%|███████▉ | 38135/48008 [2:10:55<31:34, 5.21it/s] 79%|███████▉ | 38136/48008 [2:10:55<31:25, 5.24it/s] 79%|███████▉ | 38137/48008 [2:10:56<31:14, 5.27it/s] 79%|███████▉ | 38138/48008 [2:10:56<31:51, 5.16it/s] 79%|███████▉ | 38139/48008 [2:10:56<31:43, 5.18it/s] 79%|███████▉ | 38140/48008 [2:10:56<31:43, 5.19it/s] 79%|███████▉ | 38141/48008 [2:10:56<32:16, 5.09it/s] 79%|███████▉ | 38142/48008 [2:10:57<32:52, 5.00it/s] 79%|███████▉ | 38143/48008 [2:10:57<32:12, 5.10it/s] 79%|███████▉ | 38144/48008 [2:10:57<31:39, 5.19it/s] 79%|███████▉ | 38145/48008 [2:10:57<31:33, 5.21it/s] 79%|███████▉ | 38146/48008 [2:10:57<31:49, 5.16it/s] 79%|███████▉ | 38147/48008 [2:10:58<31:45, 5.18it/s] 79%|███████▉ | 38148/48008 [2:10:58<31:20, 5.24it/s] 79%|███████▉ | 38149/48008 [2:10:58<31:22, 5.24it/s] 79%|███████▉ | 38150/48008 [2:10:58<31:23, 5.24it/s] {'loss': 0.3005, 'grad_norm': 0.15625, 'learning_rate': 4.10723212797867e-05, 'epoch': 0.79} 79%|███████▉ | 38150/48008 [2:10:58<31:23, 5.24it/s] 79%|███████▉ | 38151/48008 [2:10:58<31:32, 5.21it/s] 79%|███████▉ | 38152/48008 [2:10:59<32:28, 5.06it/s] 79%|███████▉ | 38153/48008 [2:10:59<32:20, 5.08it/s] 79%|███████▉ | 38154/48008 [2:10:59<31:55, 5.14it/s] 79%|███████▉ | 38155/48008 [2:10:59<31:41, 5.18it/s] 79%|███████▉ | 38156/48008 [2:10:59<31:30, 5.21it/s] 79%|███████▉ | 38157/48008 [2:11:00<31:15, 5.25it/s] 79%|███████▉ | 38158/48008 [2:11:00<31:05, 5.28it/s] 79%|███████▉ | 38159/48008 [2:11:00<31:23, 5.23it/s] 79%|███████▉ | 38160/48008 [2:11:00<31:09, 5.27it/s] 79%|███████▉ | 38161/48008 [2:11:01<40:43, 4.03it/s] 79%|███████▉ | 38162/48008 [2:11:01<38:45, 4.23it/s] 79%|███████▉ | 38163/48008 [2:11:01<37:07, 4.42it/s] 79%|███████▉ | 38164/48008 [2:11:01<34:54, 4.70it/s] 79%|███████▉ | 38165/48008 [2:11:01<33:51, 4.85it/s] 79%|███████▉ | 38166/48008 [2:11:01<33:33, 4.89it/s] 80%|███████▉ | 38167/48008 [2:11:02<35:06, 4.67it/s] 80%|███████▉ | 38168/48008 [2:11:02<34:29, 4.75it/s] 80%|███████▉ | 38169/48008 [2:11:02<33:49, 4.85it/s] 80%|███████▉ | 38170/48008 [2:11:02<33:17, 4.93it/s] 80%|███████▉ | 38171/48008 [2:11:03<32:36, 5.03it/s] 80%|███████▉ | 38172/48008 [2:11:03<32:07, 5.10it/s] 80%|███████▉ | 38173/48008 [2:11:03<34:10, 4.80it/s] 80%|███████▉ | 38174/48008 [2:11:03<33:23, 4.91it/s] 80%|███████▉ | 38175/48008 [2:11:03<33:13, 4.93it/s] 80%|███████▉ | 38176/48008 [2:11:04<32:46, 5.00it/s] 80%|███████▉ | 38177/48008 [2:11:04<32:27, 5.05it/s] 80%|███████▉ | 38178/48008 [2:11:04<32:23, 5.06it/s] 80%|███████▉ | 38179/48008 [2:11:04<32:09, 5.10it/s] 80%|███████▉ | 38180/48008 [2:11:04<32:07, 5.10it/s] 80%|███████▉ | 38181/48008 [2:11:04<31:42, 5.16it/s] 80%|███████▉ | 38182/48008 [2:11:05<31:17, 5.23it/s] 80%|███████▉ | 38183/48008 [2:11:05<31:14, 5.24it/s] 80%|███████▉ | 38184/48008 [2:11:05<31:32, 5.19it/s] 80%|███████▉ | 38185/48008 [2:11:05<31:17, 5.23it/s] 80%|███████▉ | 38186/48008 [2:11:05<31:04, 5.27it/s] 80%|███████▉ | 38187/48008 [2:11:06<31:03, 5.27it/s] 80%|███████▉ | 38188/48008 [2:11:06<31:11, 5.25it/s] 80%|███████▉ | 38189/48008 [2:11:06<31:47, 5.15it/s] 80%|███████▉ | 38190/48008 [2:11:06<31:31, 5.19it/s] 80%|███████▉ | 38191/48008 [2:11:06<31:41, 5.16it/s] 80%|███████▉ | 38192/48008 [2:11:07<31:03, 5.27it/s] 80%|███████▉ | 38193/48008 [2:11:07<31:00, 5.28it/s] 80%|███████▉ | 38194/48008 [2:11:07<30:58, 5.28it/s] 80%|███████▉ | 38195/48008 [2:11:07<31:01, 5.27it/s] 80%|███████▉ | 38196/48008 [2:11:07<31:37, 5.17it/s] 80%|███████▉ | 38197/48008 [2:11:08<31:20, 5.22it/s] 80%|███████▉ | 38198/48008 [2:11:08<33:41, 4.85it/s] 80%|███████▉ | 38199/48008 [2:11:08<32:58, 4.96it/s] 80%|███████▉ | 38200/48008 [2:11:08<32:31, 5.03it/s] {'loss': 0.3006, 'grad_norm': 0.337890625, 'learning_rate': 4.086402266288952e-05, 'epoch': 0.8} 80%|███████▉ | 38200/48008 [2:11:08<32:31, 5.03it/s] 80%|███████▉ | 38201/48008 [2:11:08<32:10, 5.08it/s] 80%|███████▉ | 38202/48008 [2:11:09<32:28, 5.03it/s] 80%|███████▉ | 38203/48008 [2:11:09<31:59, 5.11it/s] 80%|███████▉ | 38204/48008 [2:11:09<31:36, 5.17it/s] 80%|███████▉ | 38205/48008 [2:11:09<32:05, 5.09it/s] 80%|███████▉ | 38206/48008 [2:11:09<31:35, 5.17it/s] 80%|███████▉ | 38207/48008 [2:11:10<37:07, 4.40it/s] 80%|███████▉ | 38208/48008 [2:11:10<35:57, 4.54it/s] 80%|███████▉ | 38209/48008 [2:11:10<34:57, 4.67it/s] 80%|███████▉ | 38210/48008 [2:11:10<33:53, 4.82it/s] 80%|███████▉ | 38211/48008 [2:11:10<33:04, 4.94it/s] 80%|███████▉ | 38212/48008 [2:11:11<32:19, 5.05it/s] 80%|███████▉ | 38213/48008 [2:11:11<32:01, 5.10it/s] 80%|███████▉ | 38214/48008 [2:11:11<31:48, 5.13it/s] 80%|███████▉ | 38215/48008 [2:11:11<31:51, 5.12it/s] 80%|███████▉ | 38216/48008 [2:11:11<31:50, 5.12it/s] 80%|███████▉ | 38217/48008 [2:11:12<31:49, 5.13it/s] 80%|███████▉ | 38218/48008 [2:11:12<31:40, 5.15it/s] 80%|███████▉ | 38219/48008 [2:11:12<31:35, 5.16it/s] 80%|███████▉ | 38220/48008 [2:11:12<31:16, 5.22it/s] 80%|███████▉ | 38221/48008 [2:11:12<31:35, 5.16it/s] 80%|███████▉ | 38222/48008 [2:11:13<31:30, 5.18it/s] 80%|███████▉ | 38223/48008 [2:11:13<31:26, 5.19it/s] 80%|███████▉ | 38224/48008 [2:11:13<31:27, 5.18it/s] 80%|███████▉ | 38225/48008 [2:11:13<31:23, 5.19it/s] 80%|███████▉ | 38226/48008 [2:11:13<31:50, 5.12it/s] 80%|███████▉ | 38227/48008 [2:11:14<31:38, 5.15it/s] 80%|███████▉ | 38228/48008 [2:11:14<31:30, 5.17it/s] 80%|███████▉ | 38229/48008 [2:11:14<32:19, 5.04it/s] 80%|███████▉ | 38230/48008 [2:11:14<37:43, 4.32it/s] 80%|███████▉ | 38231/48008 [2:11:15<45:22, 3.59it/s] 80%|███████▉ | 38232/48008 [2:11:15<41:07, 3.96it/s] 80%|███████▉ | 38233/48008 [2:11:15<38:11, 4.27it/s] 80%|███████▉ | 38234/48008 [2:11:15<36:12, 4.50it/s] 80%|███████▉ | 38235/48008 [2:11:15<34:42, 4.69it/s] 80%|███████▉ | 38236/48008 [2:11:16<33:41, 4.83it/s] 80%|███████▉ | 38237/48008 [2:11:16<32:54, 4.95it/s] 80%|███████▉ | 38238/48008 [2:11:16<32:48, 4.96it/s] 80%|███████▉ | 38239/48008 [2:11:16<32:48, 4.96it/s] 80%|███████▉ | 38240/48008 [2:11:16<32:02, 5.08it/s] 80%|███████▉ | 38241/48008 [2:11:17<31:42, 5.13it/s] 80%|███████▉ | 38242/48008 [2:11:17<32:23, 5.03it/s] 80%|███████▉ | 38243/48008 [2:11:17<34:16, 4.75it/s] 80%|███████▉ | 38244/48008 [2:11:17<35:34, 4.57it/s] 80%|███████▉ | 38245/48008 [2:11:17<34:06, 4.77it/s] 80%|███████▉ | 38246/48008 [2:11:18<32:57, 4.94it/s] 80%|███████▉ | 38247/48008 [2:11:18<32:53, 4.95it/s] 80%|███████▉ | 38248/48008 [2:11:18<32:11, 5.05it/s] 80%|███████▉ | 38249/48008 [2:11:18<31:42, 5.13it/s] 80%|███████▉ | 38250/48008 [2:11:18<31:27, 5.17it/s] {'loss': 0.2971, 'grad_norm': 0.296875, 'learning_rate': 4.065572404599234e-05, 'epoch': 0.8} 80%|███████▉ | 38250/48008 [2:11:18<31:27, 5.17it/s] 80%|███████▉ | 38251/48008 [2:11:19<31:19, 5.19it/s] 80%|███████▉ | 38252/48008 [2:11:19<31:02, 5.24it/s] 80%|███████▉ | 38253/48008 [2:11:19<31:36, 5.14it/s] 80%|███████▉ | 38254/48008 [2:11:20<1:06:57, 2.43it/s] 80%|███████▉ | 38255/48008 [2:11:20<55:56, 2.91it/s] 80%|███████▉ | 38256/48008 [2:11:20<48:18, 3.36it/s] 80%|███████▉ | 38257/48008 [2:11:20<43:36, 3.73it/s] 80%|███████▉ | 38258/48008 [2:11:21<39:33, 4.11it/s] 80%|███████▉ | 38259/48008 [2:11:21<37:14, 4.36it/s] 80%|███████▉ | 38260/48008 [2:11:21<37:40, 4.31it/s] 80%|███████▉ | 38261/48008 [2:11:21<35:26, 4.58it/s] 80%|███████▉ | 38262/48008 [2:11:21<34:21, 4.73it/s] 80%|███████▉ | 38263/48008 [2:11:22<33:43, 4.82it/s] 80%|███████▉ | 38264/48008 [2:11:22<32:55, 4.93it/s] 80%|███████▉ | 38265/48008 [2:11:22<32:34, 4.99it/s] 80%|███████▉ | 38266/48008 [2:11:22<32:04, 5.06it/s] 80%|███████▉ | 38267/48008 [2:11:22<32:17, 5.03it/s] 80%|███████▉ | 38268/48008 [2:11:23<31:44, 5.12it/s] 80%|███████▉ | 38269/48008 [2:11:23<31:34, 5.14it/s] 80%|███████▉ | 38270/48008 [2:11:23<31:37, 5.13it/s] 80%|███████▉ | 38271/48008 [2:11:23<40:55, 3.97it/s] 80%|███████▉ | 38272/48008 [2:11:24<37:55, 4.28it/s] 80%|███████▉ | 38273/48008 [2:11:24<36:19, 4.47it/s] 80%|███████▉ | 38274/48008 [2:11:24<35:13, 4.60it/s] 80%|███████▉ | 38275/48008 [2:11:24<33:49, 4.80it/s] 80%|███████▉ | 38276/48008 [2:11:24<33:10, 4.89it/s] 80%|███████▉ | 38277/48008 [2:11:25<32:57, 4.92it/s] 80%|███████▉ | 38278/48008 [2:11:25<32:01, 5.06it/s] 80%|███████▉ | 38279/48008 [2:11:25<31:38, 5.13it/s] 80%|███████▉ | 38280/48008 [2:11:25<31:24, 5.16it/s] 80%|███████▉ | 38281/48008 [2:11:25<31:17, 5.18it/s] 80%|███████▉ | 38282/48008 [2:11:26<31:28, 5.15it/s] 80%|███████▉ | 38283/48008 [2:11:26<31:56, 5.07it/s] 80%|███████▉ | 38284/48008 [2:11:26<32:16, 5.02it/s] 80%|███████▉ | 38285/48008 [2:11:26<31:53, 5.08it/s] 80%|███████▉ | 38286/48008 [2:11:26<32:17, 5.02it/s] 80%|███████▉ | 38287/48008 [2:11:27<31:52, 5.08it/s] 80%|███████▉ | 38288/48008 [2:11:27<31:37, 5.12it/s] 80%|███████▉ | 38289/48008 [2:11:27<31:15, 5.18it/s] 80%|███████▉ | 38290/48008 [2:11:27<31:58, 5.07it/s] 80%|███████▉ | 38291/48008 [2:11:27<31:34, 5.13it/s] 80%|███████▉ | 38292/48008 [2:11:27<31:11, 5.19it/s] 80%|███████▉ | 38293/48008 [2:11:28<30:52, 5.24it/s] 80%|███████▉ | 38294/48008 [2:11:28<36:26, 4.44it/s] 80%|███████▉ | 38295/48008 [2:11:28<34:33, 4.68it/s] 80%|███████▉ | 38296/48008 [2:11:28<33:31, 4.83it/s] 80%|███████▉ | 38297/48008 [2:11:29<32:13, 5.02it/s] 80%|███████▉ | 38298/48008 [2:11:29<37:22, 4.33it/s] 80%|███████▉ | 38299/48008 [2:11:29<35:58, 4.50it/s] 80%|███████▉ | 38300/48008 [2:11:29<34:26, 4.70it/s] {'loss': 0.2992, 'grad_norm': 0.5078125, 'learning_rate': 4.044742542909515e-05, 'epoch': 0.8} 80%|███████▉ | 38300/48008 [2:11:29<34:26, 4.70it/s] 80%|███████▉ | 38301/48008 [2:11:29<33:27, 4.83it/s] 80%|███████▉ | 38302/48008 [2:11:30<33:40, 4.80it/s] 80%|███████▉ | 38303/48008 [2:11:30<32:58, 4.91it/s] 80%|███████▉ | 38304/48008 [2:11:30<32:26, 4.98it/s] 80%|███████▉ | 38305/48008 [2:11:30<31:57, 5.06it/s] 80%|███████▉ | 38306/48008 [2:11:31<41:03, 3.94it/s] 80%|███████▉ | 38307/48008 [2:11:31<38:07, 4.24it/s] 80%|███████▉ | 38308/48008 [2:11:31<36:36, 4.42it/s] 80%|███████▉ | 38309/48008 [2:11:31<35:53, 4.50it/s] 80%|███████▉ | 38310/48008 [2:11:31<34:26, 4.69it/s] 80%|███████▉ | 38311/48008 [2:11:32<42:52, 3.77it/s] 80%|███████▉ | 38312/48008 [2:11:32<39:09, 4.13it/s] 80%|███████▉ | 38313/48008 [2:11:32<36:53, 4.38it/s] 80%|███████▉ | 38314/48008 [2:11:32<34:45, 4.65it/s] 80%|███████▉ | 38315/48008 [2:11:33<35:53, 4.50it/s] 80%|███████▉ | 38316/48008 [2:11:33<34:17, 4.71it/s] 80%|███████▉ | 38317/48008 [2:11:33<32:57, 4.90it/s] 80%|███████▉ | 38318/48008 [2:11:33<32:20, 4.99it/s] 80%|███████▉ | 38319/48008 [2:11:33<32:28, 4.97it/s] 80%|███████▉ | 38320/48008 [2:11:34<32:00, 5.04it/s] 80%|███████▉ | 38321/48008 [2:11:34<31:39, 5.10it/s] 80%|███████▉ | 38322/48008 [2:11:34<31:45, 5.08it/s] 80%|███████▉ | 38323/48008 [2:11:34<31:59, 5.05it/s] 80%|███████▉ | 38324/48008 [2:11:34<31:30, 5.12it/s] 80%|███████▉ | 38325/48008 [2:11:35<31:22, 5.14it/s] 80%|███████▉ | 38326/48008 [2:11:35<31:10, 5.18it/s] 80%|███████▉ | 38327/48008 [2:11:35<31:00, 5.20it/s] 80%|███████▉ | 38328/48008 [2:11:35<31:41, 5.09it/s] 80%|███████▉ | 38329/48008 [2:11:35<31:46, 5.08it/s] 80%|███████▉ | 38330/48008 [2:11:35<31:25, 5.13it/s] 80%|███████▉ | 38331/48008 [2:11:36<31:12, 5.17it/s] 80%|███████▉ | 38332/48008 [2:11:36<30:59, 5.20it/s] 80%|███████▉ | 38333/48008 [2:11:36<31:01, 5.20it/s] 80%|███████▉ | 38334/48008 [2:11:36<31:03, 5.19it/s] 80%|███████▉ | 38335/48008 [2:11:37<36:32, 4.41it/s] 80%|███████▉ | 38336/48008 [2:11:37<35:19, 4.56it/s] 80%|███████▉ | 38337/48008 [2:11:37<34:07, 4.72it/s] 80%|███████▉ | 38338/48008 [2:11:37<33:11, 4.86it/s] 80%|███████▉ | 38339/48008 [2:11:37<32:34, 4.95it/s] 80%|███████▉ | 38340/48008 [2:11:38<41:35, 3.87it/s] 80%|███████▉ | 38341/48008 [2:11:38<38:14, 4.21it/s] 80%|███████▉ | 38342/48008 [2:11:38<38:16, 4.21it/s] 80%|███████▉ | 38343/48008 [2:11:38<36:02, 4.47it/s] 80%|███████▉ | 38344/48008 [2:11:39<34:27, 4.67it/s] 80%|███████▉ | 38345/48008 [2:11:39<33:26, 4.82it/s] 80%|███████▉ | 38346/48008 [2:11:39<32:33, 4.95it/s] 80%|███████▉ | 38347/48008 [2:11:39<32:22, 4.97it/s] 80%|███████▉ | 38348/48008 [2:11:39<32:47, 4.91it/s] 80%|███████▉ | 38349/48008 [2:11:40<32:20, 4.98it/s] 80%|███████▉ | 38350/48008 [2:11:40<31:51, 5.05it/s] {'loss': 0.2986, 'grad_norm': 0.228515625, 'learning_rate': 4.023912681219797e-05, 'epoch': 0.8} 80%|███████▉ | 38350/48008 [2:11:40<31:51, 5.05it/s] 80%|███████▉ | 38351/48008 [2:11:40<31:38, 5.09it/s] 80%|███████▉ | 38352/48008 [2:11:40<31:22, 5.13it/s] 80%|███████▉ | 38353/48008 [2:11:40<31:10, 5.16it/s] 80%|███████▉ | 38354/48008 [2:11:40<30:51, 5.22it/s] 80%|███████▉ | 38355/48008 [2:11:41<30:40, 5.25it/s] 80%|███████▉ | 38356/48008 [2:11:41<30:43, 5.24it/s] 80%|███████▉ | 38357/48008 [2:11:41<30:11, 5.33it/s] 80%|███████▉ | 38358/48008 [2:11:41<39:46, 4.04it/s] 80%|███████▉ | 38359/48008 [2:11:42<46:29, 3.46it/s] 80%|███████▉ | 38360/48008 [2:11:42<41:50, 3.84it/s] 80%|███████▉ | 38361/48008 [2:11:42<38:27, 4.18it/s] 80%|███████▉ | 38362/48008 [2:11:42<36:19, 4.43it/s] 80%|███████▉ | 38363/48008 [2:11:43<34:39, 4.64it/s] 80%|███████▉ | 38364/48008 [2:11:43<33:40, 4.77it/s] 80%|███████▉ | 38365/48008 [2:11:43<32:54, 4.88it/s] 80%|███████▉ | 38366/48008 [2:11:43<32:06, 5.00it/s] 80%|███████▉ | 38367/48008 [2:11:43<31:47, 5.06it/s] 80%|███████▉ | 38368/48008 [2:11:44<31:24, 5.11it/s] 80%|███████▉ | 38369/48008 [2:11:44<40:36, 3.96it/s] 80%|███████▉ | 38370/48008 [2:11:44<43:13, 3.72it/s] 80%|███████▉ | 38371/48008 [2:11:44<39:37, 4.05it/s] 80%|███████▉ | 38372/48008 [2:11:45<37:12, 4.32it/s] 80%|███████▉ | 38373/48008 [2:11:45<35:23, 4.54it/s] 80%|███████▉ | 38374/48008 [2:11:45<33:51, 4.74it/s] 80%|███████▉ | 38375/48008 [2:11:45<32:59, 4.87it/s] 80%|███████▉ | 38376/48008 [2:11:45<32:12, 4.98it/s] 80%|███████▉ | 38377/48008 [2:11:46<32:18, 4.97it/s] 80%|███████▉ | 38378/48008 [2:11:46<31:36, 5.08it/s] 80%|███████▉ | 38379/48008 [2:11:46<31:20, 5.12it/s] 80%|███████▉ | 38380/48008 [2:11:46<31:07, 5.16it/s] 80%|███████▉ | 38381/48008 [2:11:46<30:53, 5.19it/s] 80%|███████▉ | 38382/48008 [2:11:47<31:19, 5.12it/s] 80%|███████▉ | 38383/48008 [2:11:47<31:01, 5.17it/s] 80%|███████▉ | 38384/48008 [2:11:47<30:42, 5.22it/s] 80%|███████▉ | 38385/48008 [2:11:47<30:29, 5.26it/s] 80%|███████▉ | 38386/48008 [2:11:47<30:36, 5.24it/s] 80%|███████▉ | 38387/48008 [2:11:48<40:00, 4.01it/s] 80%|███████▉ | 38388/48008 [2:11:48<37:00, 4.33it/s] 80%|███████▉ | 38389/48008 [2:11:48<35:01, 4.58it/s] 80%|███████▉ | 38390/48008 [2:11:48<34:32, 4.64it/s] 80%|███████▉ | 38391/48008 [2:11:48<33:34, 4.77it/s] 80%|███████▉ | 38392/48008 [2:11:49<33:02, 4.85it/s] 80%|███████▉ | 38393/48008 [2:11:49<32:03, 5.00it/s] 80%|███████▉ | 38394/48008 [2:11:49<31:34, 5.08it/s] 80%|███████▉ | 38395/48008 [2:11:49<31:17, 5.12it/s] 80%|███████▉ | 38396/48008 [2:11:49<30:59, 5.17it/s] 80%|███████▉ | 38397/48008 [2:11:50<31:23, 5.10it/s] 80%|███████▉ | 38398/48008 [2:11:50<31:00, 5.16it/s] 80%|███████▉ | 38399/48008 [2:11:50<30:52, 5.19it/s] 80%|███████▉ | 38400/48008 [2:11:50<30:38, 5.23it/s] {'loss': 0.3011, 'grad_norm': 0.4375, 'learning_rate': 4.0030828195300786e-05, 'epoch': 0.8} 80%|███████▉ | 38400/48008 [2:11:50<30:38, 5.23it/s] 80%|███████▉ | 38401/48008 [2:11:51<36:09, 4.43it/s] 80%|███████▉ | 38402/48008 [2:11:51<34:23, 4.65it/s] 80%|███████▉ | 38403/48008 [2:11:51<32:55, 4.86it/s] 80%|███████▉ | 38404/48008 [2:11:51<32:43, 4.89it/s] 80%|███████▉ | 38405/48008 [2:11:51<32:05, 4.99it/s] 80%|███████▉ | 38406/48008 [2:11:51<31:16, 5.12it/s] 80%|████████ | 38407/48008 [2:11:52<33:13, 4.82it/s] 80%|████████ | 38408/48008 [2:11:52<32:18, 4.95it/s] 80%|████████ | 38409/48008 [2:11:52<31:40, 5.05it/s] 80%|████████ | 38410/48008 [2:11:52<31:21, 5.10it/s] 80%|████████ | 38411/48008 [2:11:52<31:54, 5.01it/s] 80%|████████ | 38412/48008 [2:11:53<31:42, 5.04it/s] 80%|████████ | 38413/48008 [2:11:53<31:18, 5.11it/s] 80%|████████ | 38414/48008 [2:11:53<31:00, 5.16it/s] 80%|████████ | 38415/48008 [2:11:53<30:37, 5.22it/s] 80%|████████ | 38416/48008 [2:11:53<31:28, 5.08it/s] 80%|████████ | 38417/48008 [2:11:54<40:31, 3.94it/s] 80%|████████ | 38418/48008 [2:11:54<37:17, 4.29it/s] 80%|████████ | 38419/48008 [2:11:54<37:29, 4.26it/s] 80%|████████ | 38420/48008 [2:11:54<35:25, 4.51it/s] 80%|████████ | 38421/48008 [2:11:55<33:57, 4.70it/s] 80%|████████ | 38422/48008 [2:11:55<33:26, 4.78it/s] 80%|████████ | 38423/48008 [2:11:55<32:29, 4.92it/s] 80%|████████ | 38424/48008 [2:11:55<31:54, 5.01it/s] 80%|████████ | 38425/48008 [2:11:55<31:44, 5.03it/s] 80%|████████ | 38426/48008 [2:11:56<31:15, 5.11it/s] 80%|████████ | 38427/48008 [2:11:56<30:58, 5.16it/s] 80%|████████ | 38428/48008 [2:11:56<30:43, 5.20it/s] 80%|████████ | 38429/48008 [2:11:56<31:35, 5.05it/s] 80%|████████ | 38430/48008 [2:11:56<30:55, 5.16it/s] 80%|████████ | 38431/48008 [2:11:57<31:14, 5.11it/s] 80%|████████ | 38432/48008 [2:11:57<31:21, 5.09it/s] 80%|████████ | 38433/48008 [2:11:57<31:05, 5.13it/s] 80%|████████ | 38434/48008 [2:11:57<30:49, 5.18it/s] 80%|████████ | 38435/48008 [2:11:57<30:36, 5.21it/s] 80%|████████ | 38436/48008 [2:11:58<30:31, 5.23it/s] 80%|████████ | 38437/48008 [2:11:58<31:02, 5.14it/s] 80%|████████ | 38438/48008 [2:11:58<30:56, 5.15it/s] 80%|████████ | 38439/48008 [2:11:58<31:22, 5.08it/s] 80%|████████ | 38440/48008 [2:11:58<31:11, 5.11it/s] 80%|████████ | 38441/48008 [2:11:59<31:02, 5.14it/s] 80%|████████ | 38442/48008 [2:11:59<31:23, 5.08it/s] 80%|████████ | 38443/48008 [2:11:59<30:53, 5.16it/s] 80%|████████ | 38444/48008 [2:11:59<30:38, 5.20it/s] 80%|████████ | 38445/48008 [2:11:59<30:30, 5.23it/s] 80%|████████ | 38446/48008 [2:11:59<30:25, 5.24it/s] 80%|████████ | 38447/48008 [2:12:00<30:18, 5.26it/s] 80%|████████ | 38448/48008 [2:12:00<29:54, 5.33it/s] 80%|████████ | 38449/48008 [2:12:00<30:02, 5.30it/s] 80%|████████ | 38450/48008 [2:12:00<30:04, 5.30it/s] {'loss': 0.3031, 'grad_norm': 0.185546875, 'learning_rate': 3.9822529578403603e-05, 'epoch': 0.8} 80%|████████ | 38450/48008 [2:12:00<30:04, 5.30it/s] 80%|████████ | 38451/48008 [2:12:00<29:57, 5.32it/s] 80%|████████ | 38452/48008 [2:12:01<30:07, 5.29it/s] 80%|████████ | 38453/48008 [2:12:01<30:49, 5.17it/s] 80%|████████ | 38454/48008 [2:12:01<30:37, 5.20it/s] 80%|████████ | 38455/48008 [2:12:01<30:17, 5.26it/s] 80%|████████ | 38456/48008 [2:12:01<30:21, 5.24it/s] 80%|████████ | 38457/48008 [2:12:02<30:16, 5.26it/s] 80%|████████ | 38458/48008 [2:12:02<32:33, 4.89it/s] 80%|████████ | 38459/48008 [2:12:02<31:45, 5.01it/s] 80%|████████ | 38460/48008 [2:12:02<31:02, 5.13it/s] 80%|████████ | 38461/48008 [2:12:02<30:51, 5.16it/s] 80%|████████ | 38462/48008 [2:12:03<30:26, 5.23it/s] 80%|████████ | 38463/48008 [2:12:03<30:21, 5.24it/s] 80%|████████ | 38464/48008 [2:12:03<30:21, 5.24it/s] 80%|████████ | 38465/48008 [2:12:03<30:17, 5.25it/s] 80%|████████ | 38466/48008 [2:12:03<31:07, 5.11it/s] 80%|████████ | 38467/48008 [2:12:04<30:54, 5.14it/s] 80%|████████ | 38468/48008 [2:12:04<31:02, 5.12it/s] 80%|████████ | 38469/48008 [2:12:04<30:40, 5.18it/s] 80%|████████ | 38470/48008 [2:12:04<30:33, 5.20it/s] 80%|████████ | 38471/48008 [2:12:04<30:16, 5.25it/s] 80%|████████ | 38472/48008 [2:12:04<30:04, 5.28it/s] 80%|████████ | 38473/48008 [2:12:05<32:24, 4.90it/s] 80%|████████ | 38474/48008 [2:12:05<32:36, 4.87it/s] 80%|████████ | 38475/48008 [2:12:05<31:51, 4.99it/s] 80%|████████ | 38476/48008 [2:12:05<31:24, 5.06it/s] 80%|████████ | 38477/48008 [2:12:05<30:51, 5.15it/s] 80%|████████ | 38478/48008 [2:12:06<30:34, 5.20it/s] 80%|████████ | 38479/48008 [2:12:06<32:44, 4.85it/s] 80%|████████ | 38480/48008 [2:12:06<32:20, 4.91it/s] 80%|████████ | 38481/48008 [2:12:06<31:41, 5.01it/s] 80%|████████ | 38482/48008 [2:12:06<31:15, 5.08it/s] 80%|████████ | 38483/48008 [2:12:07<33:11, 4.78it/s] 80%|████████ | 38484/48008 [2:12:07<32:15, 4.92it/s] 80%|████████ | 38485/48008 [2:12:07<32:14, 4.92it/s] 80%|████████ | 38486/48008 [2:12:07<31:30, 5.04it/s] 80%|████████ | 38487/48008 [2:12:07<31:04, 5.11it/s] 80%|████████ | 38488/48008 [2:12:08<31:20, 5.06it/s] 80%|████████ | 38489/48008 [2:12:08<31:59, 4.96it/s] 80%|████████ | 38490/48008 [2:12:08<33:38, 4.72it/s] 80%|████████ | 38491/48008 [2:12:08<32:23, 4.90it/s] 80%|████████ | 38492/48008 [2:12:09<31:39, 5.01it/s] 80%|████████ | 38493/48008 [2:12:09<31:39, 5.01it/s] 80%|████████ | 38494/48008 [2:12:09<30:59, 5.12it/s] 80%|████████ | 38495/48008 [2:12:09<31:17, 5.07it/s] 80%|████████ | 38496/48008 [2:12:09<31:15, 5.07it/s] 80%|████████ | 38497/48008 [2:12:09<30:29, 5.20it/s] 80%|████████ | 38498/48008 [2:12:10<30:58, 5.12it/s] 80%|████████ | 38499/48008 [2:12:10<30:58, 5.12it/s] 80%|████████ | 38500/48008 [2:12:10<30:58, 5.11it/s] {'loss': 0.3007, 'grad_norm': 0.1865234375, 'learning_rate': 3.9614230961506414e-05, 'epoch': 0.8} 80%|████████ | 38500/48008 [2:12:10<30:58, 5.11it/s] 80%|████████ | 38501/48008 [2:12:10<30:43, 5.16it/s] 80%|████████ | 38502/48008 [2:12:10<30:35, 5.18it/s] 80%|████████ | 38503/48008 [2:12:11<30:58, 5.11it/s] 80%|████████ | 38504/48008 [2:12:11<30:43, 5.16it/s] 80%|████████ | 38505/48008 [2:12:11<30:29, 5.19it/s] 80%|████████ | 38506/48008 [2:12:11<30:36, 5.17it/s] 80%|████████ | 38507/48008 [2:12:11<30:16, 5.23it/s] 80%|████████ | 38508/48008 [2:12:12<30:12, 5.24it/s] 80%|████████ | 38509/48008 [2:12:12<30:07, 5.25it/s] 80%|████████ | 38510/48008 [2:12:12<31:02, 5.10it/s] 80%|████████ | 38511/48008 [2:12:12<31:23, 5.04it/s] 80%|████████ | 38512/48008 [2:12:12<31:39, 5.00it/s] 80%|████████ | 38513/48008 [2:12:13<31:20, 5.05it/s] 80%|████████ | 38514/48008 [2:12:13<31:28, 5.03it/s] 80%|████████ | 38515/48008 [2:12:13<31:19, 5.05it/s] 80%|████████ | 38516/48008 [2:12:13<31:48, 4.97it/s] 80%|████████ | 38517/48008 [2:12:13<31:25, 5.03it/s] 80%|████████ | 38518/48008 [2:12:14<31:06, 5.08it/s] 80%|████████ | 38519/48008 [2:12:14<30:38, 5.16it/s] 80%|████████ | 38520/48008 [2:12:14<30:18, 5.22it/s] 80%|████████ | 38521/48008 [2:12:14<30:16, 5.22it/s] 80%|████████ | 38522/48008 [2:12:14<30:07, 5.25it/s] 80%|████████ | 38523/48008 [2:12:15<30:45, 5.14it/s] 80%|████████ | 38524/48008 [2:12:15<30:55, 5.11it/s] 80%|████████ | 38525/48008 [2:12:15<30:39, 5.16it/s] 80%|████████ | 38526/48008 [2:12:15<30:29, 5.18it/s] 80%|████████ | 38527/48008 [2:12:15<30:25, 5.19it/s] 80%|████████ | 38528/48008 [2:12:16<29:59, 5.27it/s] 80%|████████ | 38529/48008 [2:12:16<29:57, 5.27it/s] 80%|████████ | 38530/48008 [2:12:16<32:09, 4.91it/s] 80%|████████ | 38531/48008 [2:12:16<32:02, 4.93it/s] 80%|████████ | 38532/48008 [2:12:16<32:01, 4.93it/s] 80%|████████ | 38533/48008 [2:12:17<31:59, 4.94it/s] 80%|████████ | 38534/48008 [2:12:17<31:24, 5.03it/s] 80%|████████ | 38535/48008 [2:12:17<30:58, 5.10it/s] 80%|████████ | 38536/48008 [2:12:17<30:46, 5.13it/s] 80%|████████ | 38537/48008 [2:12:17<30:35, 5.16it/s] 80%|████████ | 38538/48008 [2:12:17<30:16, 5.21it/s] 80%|████████ | 38539/48008 [2:12:18<30:09, 5.23it/s] 80%|████████ | 38540/48008 [2:12:18<30:12, 5.22it/s] 80%|████████ | 38541/48008 [2:12:18<30:20, 5.20it/s] 80%|████████ | 38542/48008 [2:12:18<30:12, 5.22it/s] 80%|████████ | 38543/48008 [2:12:18<30:09, 5.23it/s] 80%|████████ | 38544/48008 [2:12:19<30:30, 5.17it/s] 80%|████████ | 38545/48008 [2:12:19<30:35, 5.16it/s] 80%|████████ | 38546/48008 [2:12:19<30:53, 5.10it/s] 80%|████████ | 38547/48008 [2:12:19<30:40, 5.14it/s] 80%|████████ | 38548/48008 [2:12:19<30:32, 5.16it/s] 80%|████████ | 38549/48008 [2:12:20<32:30, 4.85it/s] 80%|████████ | 38550/48008 [2:12:20<32:15, 4.89it/s] {'loss': 0.3009, 'grad_norm': 0.3203125, 'learning_rate': 3.940593234460923e-05, 'epoch': 0.8} 80%|████████ | 38550/48008 [2:12:20<32:15, 4.89it/s] 80%|████████ | 38551/48008 [2:12:20<31:36, 4.99it/s] 80%|████████ | 38552/48008 [2:12:20<31:00, 5.08it/s] 80%|████████ | 38553/48008 [2:12:20<30:37, 5.15it/s] 80%|████████ | 38554/48008 [2:12:21<31:01, 5.08it/s] 80%|████████ | 38555/48008 [2:12:21<30:46, 5.12it/s] 80%|████████ | 38556/48008 [2:12:21<30:34, 5.15it/s] 80%|████████ | 38557/48008 [2:12:21<30:21, 5.19it/s] 80%|████████ | 38558/48008 [2:12:21<30:14, 5.21it/s] 80%|████████ | 38559/48008 [2:12:22<30:17, 5.20it/s] 80%|████████ | 38560/48008 [2:12:22<32:26, 4.85it/s] 80%|████████ | 38561/48008 [2:12:22<31:45, 4.96it/s] 80%|████████ | 38562/48008 [2:12:22<31:10, 5.05it/s] 80%|████████ | 38563/48008 [2:12:22<30:39, 5.13it/s] 80%|████████ | 38564/48008 [2:12:23<31:03, 5.07it/s] 80%|████████ | 38565/48008 [2:12:23<30:45, 5.12it/s] 80%|████████ | 38566/48008 [2:12:23<30:42, 5.13it/s] 80%|████████ | 38567/48008 [2:12:23<31:01, 5.07it/s] 80%|████████ | 38568/48008 [2:12:23<30:34, 5.15it/s] 80%|████████ | 38569/48008 [2:12:24<30:21, 5.18it/s] 80%|████████ | 38570/48008 [2:12:24<30:00, 5.24it/s] 80%|████████ | 38571/48008 [2:12:24<29:48, 5.28it/s] 80%|████████ | 38572/48008 [2:12:24<29:26, 5.34it/s] 80%|████████ | 38573/48008 [2:12:24<29:52, 5.26it/s] 80%|████████ | 38574/48008 [2:12:25<30:25, 5.17it/s] 80%|████████ | 38575/48008 [2:12:25<30:29, 5.16it/s] 80%|████████ | 38576/48008 [2:12:25<35:46, 4.39it/s] 80%|████████ | 38577/48008 [2:12:25<33:56, 4.63it/s] 80%|████████ | 38578/48008 [2:12:25<32:34, 4.83it/s] 80%|████████ | 38579/48008 [2:12:26<32:17, 4.87it/s] 80%|████████ | 38580/48008 [2:12:26<31:52, 4.93it/s] 80%|████████ | 38581/48008 [2:12:26<31:49, 4.94it/s] 80%|████████ | 38582/48008 [2:12:26<31:16, 5.02it/s] 80%|████████ | 38583/48008 [2:12:26<31:21, 5.01it/s] 80%|████████ | 38584/48008 [2:12:27<30:46, 5.10it/s] 80%|████████ | 38585/48008 [2:12:27<31:22, 5.01it/s] 80%|████████ | 38586/48008 [2:12:27<31:05, 5.05it/s] 80%|████████ | 38587/48008 [2:12:27<31:43, 4.95it/s] 80%|████████ | 38588/48008 [2:12:27<31:42, 4.95it/s] 80%|████████ | 38589/48008 [2:12:28<31:25, 5.00it/s] 80%|████████ | 38590/48008 [2:12:28<33:11, 4.73it/s] 80%|████████ | 38591/48008 [2:12:28<32:47, 4.79it/s] 80%|████████ | 38592/48008 [2:12:28<31:49, 4.93it/s] 80%|████████ | 38593/48008 [2:12:29<1:17:23, 2.03it/s] 80%|████████ | 38594/48008 [2:12:30<1:03:00, 2.49it/s] 80%|████████ | 38595/48008 [2:12:30<1:02:18, 2.52it/s] 80%|████████ | 38596/48008 [2:12:30<52:32, 2.99it/s] 80%|████████ | 38597/48008 [2:12:30<45:43, 3.43it/s] 80%|████████ | 38598/48008 [2:12:31<40:59, 3.83it/s] 80%|████████ | 38599/48008 [2:12:31<38:05, 4.12it/s] 80%|████████ | 38600/48008 [2:12:31<36:33, 4.29it/s] {'loss': 0.2986, 'grad_norm': 0.1484375, 'learning_rate': 3.919763372771205e-05, 'epoch': 0.8} 80%|████████ | 38600/48008 [2:12:31<36:33, 4.29it/s] 80%|████████ | 38601/48008 [2:12:31<34:35, 4.53it/s] 80%|████████ | 38602/48008 [2:12:31<32:54, 4.76it/s] 80%|████████ | 38603/48008 [2:12:32<32:00, 4.90it/s] 80%|████████ | 38604/48008 [2:12:32<31:16, 5.01it/s] 80%|████████ | 38605/48008 [2:12:32<31:30, 4.97it/s] 80%|████████ | 38606/48008 [2:12:32<31:00, 5.05it/s] 80%|████████ | 38607/48008 [2:12:32<30:43, 5.10it/s] 80%|████████ | 38608/48008 [2:12:32<30:14, 5.18it/s] 80%|████████ | 38609/48008 [2:12:33<31:08, 5.03it/s] 80%|████████ | 38610/48008 [2:12:33<30:35, 5.12it/s] 80%|████████ | 38611/48008 [2:12:33<30:57, 5.06it/s] 80%|████████ | 38612/48008 [2:12:33<30:26, 5.14it/s] 80%|████████ | 38613/48008 [2:12:33<30:19, 5.16it/s] 80%|████████ | 38614/48008 [2:12:34<30:08, 5.20it/s] 80%|████████ | 38615/48008 [2:12:34<30:00, 5.22it/s] 80%|████████ | 38616/48008 [2:12:34<30:12, 5.18it/s] 80%|████████ | 38617/48008 [2:12:34<29:54, 5.23it/s] 80%|████████ | 38618/48008 [2:12:34<30:21, 5.15it/s] 80%|████████ | 38619/48008 [2:12:35<30:13, 5.18it/s] 80%|████████ | 38620/48008 [2:12:35<29:47, 5.25it/s] 80%|████████ | 38621/48008 [2:12:35<30:18, 5.16it/s] 80%|████████ | 38622/48008 [2:12:35<30:36, 5.11it/s] 80%|████████ | 38623/48008 [2:12:35<32:30, 4.81it/s] 80%|████████ | 38624/48008 [2:12:36<31:37, 4.95it/s] 80%|████████ | 38625/48008 [2:12:36<31:04, 5.03it/s] 80%|████████ | 38626/48008 [2:12:36<30:40, 5.10it/s] 80%|████████ | 38627/48008 [2:12:36<30:29, 5.13it/s] 80%|████████ | 38628/48008 [2:12:37<39:28, 3.96it/s] 80%|████████ | 38629/48008 [2:12:37<36:34, 4.27it/s] 80%|████████ | 38630/48008 [2:12:37<35:02, 4.46it/s] 80%|████████ | 38631/48008 [2:12:37<33:24, 4.68it/s] 80%|████████ | 38632/48008 [2:12:37<32:17, 4.84it/s] 80%|████████ | 38633/48008 [2:12:38<31:46, 4.92it/s] 80%|████████ | 38634/48008 [2:12:38<31:12, 5.01it/s] 80%|████████ | 38635/48008 [2:12:38<30:27, 5.13it/s] 80%|████████ | 38636/48008 [2:12:38<30:13, 5.17it/s] 80%|████████ | 38637/48008 [2:12:38<30:04, 5.19it/s] 80%|████████ | 38638/48008 [2:12:39<32:05, 4.87it/s] 80%|████████ | 38639/48008 [2:12:39<31:24, 4.97it/s] 80%|████████ | 38640/48008 [2:12:39<31:32, 4.95it/s] 80%|████████ | 38641/48008 [2:12:39<30:52, 5.06it/s] 80%|████████ | 38642/48008 [2:12:39<30:54, 5.05it/s] 80%|████████ | 38643/48008 [2:12:40<30:37, 5.10it/s] 80%|████████ | 38644/48008 [2:12:40<31:15, 4.99it/s] 80%|████████ | 38645/48008 [2:12:40<30:39, 5.09it/s] 80%|████████ | 38646/48008 [2:12:40<30:53, 5.05it/s] 81%|████████ | 38647/48008 [2:12:40<30:58, 5.04it/s] 81%|████████ | 38648/48008 [2:12:40<30:22, 5.14it/s] 81%|████████ | 38649/48008 [2:12:41<30:18, 5.15it/s] 81%|████████ | 38650/48008 [2:12:41<30:14, 5.16it/s] {'loss': 0.3005, 'grad_norm': 0.2138671875, 'learning_rate': 3.8989335110814866e-05, 'epoch': 0.81} 81%|████████ | 38650/48008 [2:12:41<30:14, 5.16it/s] 81%|████████ | 38651/48008 [2:12:41<30:59, 5.03it/s] 81%|████████ | 38652/48008 [2:12:41<31:06, 5.01it/s] 81%|████████ | 38653/48008 [2:12:41<30:35, 5.10it/s] 81%|████████ | 38654/48008 [2:12:42<30:17, 5.15it/s] 81%|████████ | 38655/48008 [2:12:42<29:51, 5.22it/s] 81%|████████ | 38656/48008 [2:12:42<29:32, 5.28it/s] 81%|████████ | 38657/48008 [2:12:42<30:52, 5.05it/s] 81%|████████ | 38658/48008 [2:12:42<30:20, 5.14it/s] 81%|████████ | 38659/48008 [2:12:43<30:52, 5.05it/s] 81%|████████ | 38660/48008 [2:12:43<31:04, 5.01it/s] 81%|████████ | 38661/48008 [2:12:43<30:41, 5.08it/s] 81%|████████ | 38662/48008 [2:12:43<30:56, 5.03it/s] 81%|████████ | 38663/48008 [2:12:43<30:18, 5.14it/s] 81%|████████ | 38664/48008 [2:12:44<30:12, 5.16it/s] 81%|████████ | 38665/48008 [2:12:44<30:24, 5.12it/s] 81%|████████ | 38666/48008 [2:12:44<30:51, 5.05it/s] 81%|████████ | 38667/48008 [2:12:44<31:06, 5.00it/s] 81%|████████ | 38668/48008 [2:12:44<31:30, 4.94it/s] 81%|████████ | 38669/48008 [2:12:45<31:00, 5.02it/s] 81%|████████ | 38670/48008 [2:12:45<31:25, 4.95it/s] 81%|████████ | 38671/48008 [2:12:45<30:57, 5.03it/s] 81%|████████ | 38672/48008 [2:12:45<30:27, 5.11it/s] 81%|████████ | 38673/48008 [2:12:45<29:58, 5.19it/s] 81%|████████ | 38674/48008 [2:12:46<29:51, 5.21it/s] 81%|████████ | 38675/48008 [2:12:46<29:49, 5.22it/s] 81%|████████ | 38676/48008 [2:12:46<29:26, 5.28it/s] 81%|████████ | 38677/48008 [2:12:46<29:54, 5.20it/s] 81%|████████ | 38678/48008 [2:12:46<30:29, 5.10it/s] 81%|████████ | 38679/48008 [2:12:47<30:34, 5.08it/s] 81%|████████ | 38680/48008 [2:12:47<30:29, 5.10it/s] 81%|████████ | 38681/48008 [2:12:47<39:24, 3.94it/s] 81%|████████ | 38682/48008 [2:12:47<36:40, 4.24it/s] 81%|████████ | 38683/48008 [2:12:48<34:33, 4.50it/s] 81%|████████ | 38684/48008 [2:12:48<42:10, 3.68it/s] 81%|████████ | 38685/48008 [2:12:48<38:40, 4.02it/s] 81%|████████ | 38686/48008 [2:12:48<36:16, 4.28it/s] 81%|████████ | 38687/48008 [2:12:49<34:17, 4.53it/s] 81%|████████ | 38688/48008 [2:12:49<32:57, 4.71it/s] 81%|████████ | 38689/48008 [2:12:49<31:44, 4.89it/s] 81%|████████ | 38690/48008 [2:12:49<31:03, 5.00it/s] 81%|████████ | 38691/48008 [2:12:49<30:31, 5.09it/s] 81%|████████ | 38692/48008 [2:12:49<30:16, 5.13it/s] 81%|████████ | 38693/48008 [2:12:50<30:26, 5.10it/s] 81%|████████ | 38694/48008 [2:12:50<30:05, 5.16it/s] 81%|████████ | 38695/48008 [2:12:50<29:49, 5.20it/s] 81%|████████ | 38696/48008 [2:12:50<35:10, 4.41it/s] 81%|████████ | 38697/48008 [2:12:51<33:15, 4.67it/s] 81%|████████ | 38698/48008 [2:12:51<32:12, 4.82it/s] 81%|████████ | 38699/48008 [2:12:51<32:18, 4.80it/s] 81%|████████ | 38700/48008 [2:12:51<31:28, 4.93it/s] {'loss': 0.3002, 'grad_norm': 0.1552734375, 'learning_rate': 3.8781036493917683e-05, 'epoch': 0.81} 81%|████████ | 38700/48008 [2:12:51<31:28, 4.93it/s] 81%|████████ | 38701/48008 [2:12:51<31:30, 4.92it/s] 81%|████████ | 38702/48008 [2:12:52<31:26, 4.93it/s] 81%|████████ | 38703/48008 [2:12:52<31:13, 4.97it/s] 81%|████████ | 38704/48008 [2:12:52<30:55, 5.01it/s] 81%|████████ | 38705/48008 [2:12:52<31:02, 5.00it/s] 81%|████████ | 38706/48008 [2:12:52<31:05, 4.99it/s] 81%|████████ | 38707/48008 [2:12:53<32:47, 4.73it/s] 81%|████████ | 38708/48008 [2:12:53<31:45, 4.88it/s] 81%|████████ | 38709/48008 [2:12:53<30:46, 5.04it/s] 81%|████████ | 38710/48008 [2:12:53<30:36, 5.06it/s] 81%|████████ | 38711/48008 [2:12:53<30:11, 5.13it/s] 81%|████████ | 38712/48008 [2:12:54<30:03, 5.16it/s] 81%|████████ | 38713/48008 [2:12:54<29:55, 5.18it/s] 81%|████████ | 38714/48008 [2:12:54<29:46, 5.20it/s] 81%|████████ | 38715/48008 [2:12:54<30:00, 5.16it/s] 81%|████████ | 38716/48008 [2:12:54<29:56, 5.17it/s] 81%|████████ | 38717/48008 [2:12:54<30:00, 5.16it/s] 81%|████████ | 38718/48008 [2:12:55<29:54, 5.18it/s] 81%|████████ | 38719/48008 [2:12:55<29:59, 5.16it/s] 81%|████████ | 38720/48008 [2:12:55<30:02, 5.15it/s] 81%|████████ | 38721/48008 [2:12:55<29:34, 5.23it/s] 81%|████████ | 38722/48008 [2:12:55<29:37, 5.23it/s] 81%|████████ | 38723/48008 [2:12:56<30:09, 5.13it/s] 81%|████████ | 38724/48008 [2:12:56<29:54, 5.17it/s] 81%|████████ | 38725/48008 [2:12:56<29:38, 5.22it/s] 81%|████████ | 38726/48008 [2:12:56<29:41, 5.21it/s] 81%|████████ | 38727/48008 [2:12:57<38:37, 4.00it/s] 81%|████████ | 38728/48008 [2:12:57<35:43, 4.33it/s] 81%|████████ | 38729/48008 [2:12:57<33:55, 4.56it/s] 81%|████████ | 38730/48008 [2:12:57<33:06, 4.67it/s] 81%|████████ | 38731/48008 [2:12:57<31:52, 4.85it/s] 81%|████████ | 38732/48008 [2:12:58<31:45, 4.87it/s] 81%|████████ | 38733/48008 [2:12:58<31:42, 4.87it/s] 81%|████████ | 38734/48008 [2:12:58<30:58, 4.99it/s] 81%|████████ | 38735/48008 [2:12:58<31:04, 4.97it/s] 81%|████████ | 38736/48008 [2:12:58<30:38, 5.04it/s] 81%|████████ | 38737/48008 [2:12:59<31:00, 4.98it/s] 81%|████████ | 38738/48008 [2:12:59<30:33, 5.05it/s] 81%|████████ | 38739/48008 [2:12:59<30:16, 5.10it/s] 81%|████████ | 38740/48008 [2:12:59<39:05, 3.95it/s] 81%|████████ | 38741/48008 [2:13:00<36:45, 4.20it/s] 81%|████████ | 38742/48008 [2:13:00<43:33, 3.55it/s] 81%|████████ | 38743/48008 [2:13:00<39:19, 3.93it/s] 81%|████████ | 38744/48008 [2:13:00<36:45, 4.20it/s] 81%|████████ | 38745/48008 [2:13:00<34:31, 4.47it/s] 81%|████████ | 38746/48008 [2:13:01<35:08, 4.39it/s] 81%|████████ | 38747/48008 [2:13:01<33:07, 4.66it/s] 81%|████████ | 38748/48008 [2:13:01<31:47, 4.85it/s] 81%|████████ | 38749/48008 [2:13:01<30:48, 5.01it/s] 81%|████████ | 38750/48008 [2:13:01<30:24, 5.07it/s] {'loss': 0.2991, 'grad_norm': 0.216796875, 'learning_rate': 3.85727378770205e-05, 'epoch': 0.81} 81%|████████ | 38750/48008 [2:13:01<30:24, 5.07it/s] 81%|████████ | 38751/48008 [2:13:02<30:46, 5.01it/s] 81%|████████ | 38752/48008 [2:13:02<30:19, 5.09it/s] 81%|████████ | 38753/48008 [2:13:02<29:59, 5.14it/s] 81%|████████ | 38754/48008 [2:13:02<30:41, 5.03it/s] 81%|████████ | 38755/48008 [2:13:02<30:05, 5.13it/s] 81%|████████ | 38756/48008 [2:13:03<30:42, 5.02it/s] 81%|████████ | 38757/48008 [2:13:03<30:49, 5.00it/s] 81%|████████ | 38758/48008 [2:13:03<30:10, 5.11it/s] 81%|████████ | 38759/48008 [2:13:03<29:46, 5.18it/s] 81%|████████ | 38760/48008 [2:13:03<30:10, 5.11it/s] 81%|████████ | 38761/48008 [2:13:04<29:50, 5.17it/s] 81%|████████ | 38762/48008 [2:13:04<29:44, 5.18it/s] 81%|████████ | 38763/48008 [2:13:04<29:37, 5.20it/s] 81%|████████ | 38764/48008 [2:13:04<29:30, 5.22it/s] 81%|████████ | 38765/48008 [2:13:04<30:01, 5.13it/s] 81%|████████ | 38766/48008 [2:13:05<29:43, 5.18it/s] 81%|████████ | 38767/48008 [2:13:05<34:57, 4.41it/s] 81%|████████ | 38768/48008 [2:13:05<33:25, 4.61it/s] 81%|████████ | 38769/48008 [2:13:05<37:34, 4.10it/s] 81%|████████ | 38770/48008 [2:13:06<34:55, 4.41it/s] 81%|████████ | 38771/48008 [2:13:06<33:15, 4.63it/s] 81%|████████ | 38772/48008 [2:13:06<32:07, 4.79it/s] 81%|████████ | 38773/48008 [2:13:06<31:16, 4.92it/s] 81%|████████ | 38774/48008 [2:13:06<30:58, 4.97it/s] 81%|████████ | 38775/48008 [2:13:07<31:20, 4.91it/s] 81%|████████ | 38776/48008 [2:13:07<30:45, 5.00it/s] 81%|████████ | 38777/48008 [2:13:07<30:07, 5.11it/s] 81%|████████ | 38778/48008 [2:13:07<29:38, 5.19it/s] 81%|████████ | 38779/48008 [2:13:07<29:19, 5.24it/s] 81%|████████ | 38780/48008 [2:13:08<29:33, 5.20it/s] 81%|████████ | 38781/48008 [2:13:08<30:06, 5.11it/s] 81%|████████ | 38782/48008 [2:13:08<30:05, 5.11it/s] 81%|████████ | 38783/48008 [2:13:08<30:09, 5.10it/s] 81%|████████ | 38784/48008 [2:13:08<29:53, 5.14it/s] 81%|████████ | 38785/48008 [2:13:08<29:40, 5.18it/s] 81%|████████ | 38786/48008 [2:13:09<29:29, 5.21it/s] 81%|████████ | 38787/48008 [2:13:09<29:29, 5.21it/s] 81%|████████ | 38788/48008 [2:13:09<29:45, 5.16it/s] 81%|████████ | 38789/48008 [2:13:09<29:38, 5.18it/s] 81%|████████ | 38790/48008 [2:13:10<38:32, 3.99it/s] 81%|████████ | 38791/48008 [2:13:10<35:47, 4.29it/s] 81%|████████ | 38792/48008 [2:13:10<34:36, 4.44it/s] 81%|████████ | 38793/48008 [2:13:10<32:54, 4.67it/s] 81%|████████ | 38794/48008 [2:13:10<31:41, 4.85it/s] 81%|████████ | 38795/48008 [2:13:11<30:44, 5.00it/s] 81%|████████ | 38796/48008 [2:13:11<30:06, 5.10it/s] 81%|████████ | 38797/48008 [2:13:11<30:40, 5.01it/s] 81%|████████ | 38798/48008 [2:13:11<30:08, 5.09it/s] 81%|████████ | 38799/48008 [2:13:11<30:31, 5.03it/s] 81%|████████ | 38800/48008 [2:13:12<30:38, 5.01it/s] {'loss': 0.2999, 'grad_norm': 0.279296875, 'learning_rate': 3.836443926012331e-05, 'epoch': 0.81} 81%|████████ | 38800/48008 [2:13:12<30:38, 5.01it/s] 81%|████████ | 38801/48008 [2:13:12<30:15, 5.07it/s] 81%|████████ | 38802/48008 [2:13:12<29:59, 5.12it/s] 81%|████████ | 38803/48008 [2:13:12<29:50, 5.14it/s] 81%|████████ | 38804/48008 [2:13:12<29:35, 5.18it/s] 81%|████████ | 38805/48008 [2:13:13<29:30, 5.20it/s] 81%|████████ | 38806/48008 [2:13:13<29:32, 5.19it/s] 81%|████████ | 38807/48008 [2:13:13<29:26, 5.21it/s] 81%|████████ | 38808/48008 [2:13:13<29:06, 5.27it/s] 81%|████████ | 38809/48008 [2:13:13<29:25, 5.21it/s] 81%|████████ | 38810/48008 [2:13:14<29:24, 5.21it/s] 81%|████████ | 38811/48008 [2:13:14<29:23, 5.21it/s] 81%|████████ | 38812/48008 [2:13:14<29:40, 5.16it/s] 81%|████████ | 38813/48008 [2:13:14<29:21, 5.22it/s] 81%|████████ | 38814/48008 [2:13:14<29:08, 5.26it/s] 81%|████████ | 38815/48008 [2:13:14<29:13, 5.24it/s] 81%|████████ | 38816/48008 [2:13:15<29:21, 5.22it/s] 81%|████████ | 38817/48008 [2:13:15<29:36, 5.17it/s] 81%|████████ | 38818/48008 [2:13:15<29:42, 5.16it/s] 81%|████████ | 38819/48008 [2:13:15<29:28, 5.20it/s] 81%|████████ | 38820/48008 [2:13:15<29:21, 5.22it/s] 81%|████████ | 38821/48008 [2:13:16<29:09, 5.25it/s] 81%|████████ | 38822/48008 [2:13:16<29:42, 5.15it/s] 81%|████████ | 38823/48008 [2:13:16<29:38, 5.16it/s] 81%|████████ | 38824/48008 [2:13:16<29:31, 5.18it/s] 81%|████████ | 38825/48008 [2:13:16<29:21, 5.21it/s] 81%|████████ | 38826/48008 [2:13:17<29:11, 5.24it/s] 81%|████████ | 38827/48008 [2:13:17<28:45, 5.32it/s] 81%|████████ | 38828/48008 [2:13:17<30:57, 4.94it/s] 81%|████████ | 38829/48008 [2:13:17<31:16, 4.89it/s] 81%|████████ | 38830/48008 [2:13:17<30:23, 5.03it/s] 81%|████████ | 38831/48008 [2:13:18<29:42, 5.15it/s] 81%|████████ | 38832/48008 [2:13:18<29:34, 5.17it/s] 81%|████████ | 38833/48008 [2:13:18<29:19, 5.22it/s] 81%|████████ | 38834/48008 [2:13:18<29:16, 5.22it/s] 81%|████████ | 38835/48008 [2:13:18<29:30, 5.18it/s] 81%|████████ | 38836/48008 [2:13:19<29:09, 5.24it/s] 81%|████████ | 38837/48008 [2:13:19<29:25, 5.19it/s] 81%|████████ | 38838/48008 [2:13:19<29:10, 5.24it/s] 81%|████████ | 38839/48008 [2:13:19<29:28, 5.18it/s] 81%|████████ | 38840/48008 [2:13:19<29:26, 5.19it/s] 81%|████████ | 38841/48008 [2:13:19<29:33, 5.17it/s] 81%|████████ | 38842/48008 [2:13:20<29:41, 5.15it/s] 81%|████████ | 38843/48008 [2:13:20<29:04, 5.25it/s] 81%|████████ | 38844/48008 [2:13:20<28:55, 5.28it/s] 81%|████████ | 38845/48008 [2:13:20<28:59, 5.27it/s] 81%|████████ | 38846/48008 [2:13:20<29:05, 5.25it/s] 81%|████████ | 38847/48008 [2:13:21<29:06, 5.25it/s] 81%|████████ | 38848/48008 [2:13:21<29:25, 5.19it/s] 81%|████████ | 38849/48008 [2:13:21<29:07, 5.24it/s] 81%|████████ | 38850/48008 [2:13:21<29:03, 5.25it/s] {'loss': 0.303, 'grad_norm': 0.142578125, 'learning_rate': 3.815614064322613e-05, 'epoch': 0.81} 81%|████████ | 38850/48008 [2:13:21<29:03, 5.25it/s] 81%|████████ | 38851/48008 [2:13:21<29:03, 5.25it/s] 81%|████████ | 38852/48008 [2:13:22<28:56, 5.27it/s] 81%|████████ | 38853/48008 [2:13:22<28:57, 5.27it/s] 81%|████████ | 38854/48008 [2:13:22<29:02, 5.25it/s] 81%|████████ | 38855/48008 [2:13:22<29:09, 5.23it/s] 81%|████████ | 38856/48008 [2:13:22<28:57, 5.27it/s] 81%|████████ | 38857/48008 [2:13:23<28:53, 5.28it/s] 81%|████████ | 38858/48008 [2:13:23<31:01, 4.92it/s] 81%|████████ | 38859/48008 [2:13:23<30:09, 5.06it/s] 81%|████████ | 38860/48008 [2:13:23<30:25, 5.01it/s] 81%|████████ | 38861/48008 [2:13:23<30:39, 4.97it/s] 81%|████████ | 38862/48008 [2:13:24<30:03, 5.07it/s] 81%|████████ | 38863/48008 [2:13:24<38:39, 3.94it/s] 81%|████████ | 38864/48008 [2:13:24<35:31, 4.29it/s] 81%|████████ | 38865/48008 [2:13:24<34:10, 4.46it/s] 81%|████████ | 38866/48008 [2:13:25<32:52, 4.63it/s] 81%|████████ | 38867/48008 [2:13:25<31:44, 4.80it/s] 81%|████████ | 38868/48008 [2:13:25<30:41, 4.96it/s] 81%|████████ | 38869/48008 [2:13:25<30:09, 5.05it/s] 81%|████████ | 38870/48008 [2:13:25<29:41, 5.13it/s] 81%|████████ | 38871/48008 [2:13:25<30:04, 5.06it/s] 81%|████████ | 38872/48008 [2:13:26<29:42, 5.12it/s] 81%|████████ | 38873/48008 [2:13:26<29:32, 5.15it/s] 81%|████████ | 38874/48008 [2:13:26<29:56, 5.08it/s] 81%|████████ | 38875/48008 [2:13:26<29:36, 5.14it/s] 81%|████████ | 38876/48008 [2:13:26<29:30, 5.16it/s] 81%|████████ | 38877/48008 [2:13:27<29:25, 5.17it/s] 81%|████████ | 38878/48008 [2:13:27<29:50, 5.10it/s] 81%|████████ | 38879/48008 [2:13:27<29:30, 5.16it/s] 81%|████████ | 38880/48008 [2:13:27<29:19, 5.19it/s] 81%|████████ | 38881/48008 [2:13:27<29:01, 5.24it/s] 81%|████████ | 38882/48008 [2:13:28<29:08, 5.22it/s] 81%|████████ | 38883/48008 [2:13:28<28:53, 5.27it/s] 81%|████████ | 38884/48008 [2:13:28<37:48, 4.02it/s] 81%|████████ | 38885/48008 [2:13:28<34:53, 4.36it/s] 81%|████████ | 38886/48008 [2:13:29<33:20, 4.56it/s] 81%|████████ | 38887/48008 [2:13:29<40:53, 3.72it/s] 81%|████████ | 38888/48008 [2:13:29<37:18, 4.07it/s] 81%|████████ | 38889/48008 [2:13:29<34:54, 4.35it/s] 81%|████████ | 38890/48008 [2:13:30<33:40, 4.51it/s] 81%|████████ | 38891/48008 [2:13:30<32:01, 4.74it/s] 81%|████████ | 38892/48008 [2:13:30<31:27, 4.83it/s] 81%|████████ | 38893/48008 [2:13:30<30:49, 4.93it/s] 81%|████████ | 38894/48008 [2:13:30<30:09, 5.04it/s] 81%|████████ | 38895/48008 [2:13:30<29:44, 5.11it/s] 81%|████████ | 38896/48008 [2:13:31<29:20, 5.18it/s] 81%|████████ | 38897/48008 [2:13:31<29:48, 5.09it/s] 81%|████████ | 38898/48008 [2:13:31<29:26, 5.16it/s] 81%|████████ | 38899/48008 [2:13:31<29:26, 5.16it/s] 81%|████████ | 38900/48008 [2:13:31<29:24, 5.16it/s] {'loss': 0.3007, 'grad_norm': 0.2177734375, 'learning_rate': 3.7947842026328946e-05, 'epoch': 0.81} 81%|████████ | 38900/48008 [2:13:31<29:24, 5.16it/s] 81%|████████ | 38901/48008 [2:13:32<29:35, 5.13it/s] 81%|████████ | 38902/48008 [2:13:32<29:16, 5.18it/s] 81%|████████ | 38903/48008 [2:13:32<29:28, 5.15it/s] 81%|████████ | 38904/48008 [2:13:32<29:05, 5.22it/s] 81%|████████ | 38905/48008 [2:13:32<29:03, 5.22it/s] 81%|████████ | 38906/48008 [2:13:33<28:57, 5.24it/s] 81%|████████ | 38907/48008 [2:13:33<29:29, 5.14it/s] 81%|████████ | 38908/48008 [2:13:33<38:13, 3.97it/s] 81%|████████ | 38909/48008 [2:13:33<36:01, 4.21it/s] 81%|████████ | 38910/48008 [2:13:34<33:48, 4.48it/s] 81%|████████ | 38911/48008 [2:13:34<32:22, 4.68it/s] 81%|████████ | 38912/48008 [2:13:34<31:19, 4.84it/s] 81%|████████ | 38913/48008 [2:13:34<30:54, 4.90it/s] 81%|████████ | 38914/48008 [2:13:34<30:31, 4.96it/s] 81%|████████ | 38915/48008 [2:13:35<30:35, 4.95it/s] 81%|████████ | 38916/48008 [2:13:35<30:41, 4.94it/s] 81%|████████ | 38917/48008 [2:13:35<30:06, 5.03it/s] 81%|████████ | 38918/48008 [2:13:35<29:30, 5.13it/s] 81%|████████ | 38919/48008 [2:13:35<29:18, 5.17it/s] 81%|████████ | 38920/48008 [2:13:36<28:59, 5.22it/s] 81%|████████ | 38921/48008 [2:13:36<28:57, 5.23it/s] 81%|████████ | 38922/48008 [2:13:36<28:38, 5.29it/s] 81%|████████ | 38923/48008 [2:13:36<28:12, 5.37it/s] 81%|████████ | 38924/48008 [2:13:36<28:58, 5.23it/s] 81%|████████ | 38925/48008 [2:13:36<29:13, 5.18it/s] 81%|████████ | 38926/48008 [2:13:37<31:11, 4.85it/s] 81%|████████ | 38927/48008 [2:13:37<30:42, 4.93it/s] 81%|████████ | 38928/48008 [2:13:37<29:53, 5.06it/s] 81%|████████ | 38929/48008 [2:13:37<30:27, 4.97it/s] 81%|████████ | 38930/48008 [2:13:37<29:53, 5.06it/s] 81%|████████ | 38931/48008 [2:13:38<29:24, 5.14it/s] 81%|████████ | 38932/48008 [2:13:38<29:19, 5.16it/s] 81%|████████ | 38933/48008 [2:13:38<28:56, 5.23it/s] 81%|████████ | 38934/48008 [2:13:38<28:57, 5.22it/s] 81%|████████ | 38935/48008 [2:13:38<28:38, 5.28it/s] 81%|████████ | 38936/48008 [2:13:39<29:17, 5.16it/s] 81%|████████ | 38937/48008 [2:13:39<29:01, 5.21it/s] 81%|████████ | 38938/48008 [2:13:39<28:57, 5.22it/s] 81%|████████ | 38939/48008 [2:13:39<28:57, 5.22it/s] 81%|████████ | 38940/48008 [2:13:39<28:50, 5.24it/s] 81%|████████ | 38941/48008 [2:13:40<28:53, 5.23it/s] 81%|████████ | 38942/48008 [2:13:40<28:50, 5.24it/s] 81%|████████ | 38943/48008 [2:13:40<29:22, 5.14it/s] 81%|████████ | 38944/48008 [2:13:40<29:27, 5.13it/s] 81%|████████ | 38945/48008 [2:13:40<29:51, 5.06it/s] 81%|████████ | 38946/48008 [2:13:41<29:30, 5.12it/s] 81%|████████ | 38947/48008 [2:13:41<29:38, 5.09it/s] 81%|████████ | 38948/48008 [2:13:41<29:28, 5.12it/s] 81%|████████ | 38949/48008 [2:13:41<29:50, 5.06it/s] 81%|████████ | 38950/48008 [2:13:41<29:27, 5.13it/s] {'loss': 0.302, 'grad_norm': 0.455078125, 'learning_rate': 3.7739543409431764e-05, 'epoch': 0.81} 81%|████████ | 38950/48008 [2:13:41<29:27, 5.13it/s] 81%|████████ | 38951/48008 [2:13:42<31:22, 4.81it/s] 81%|████████ | 38952/48008 [2:13:42<30:37, 4.93it/s] 81%|████████ | 38953/48008 [2:13:42<29:59, 5.03it/s] 81%|████████ | 38954/48008 [2:13:42<29:40, 5.08it/s] 81%|████████ | 38955/48008 [2:13:42<29:21, 5.14it/s] 81%|████████ | 38956/48008 [2:13:43<30:02, 5.02it/s] 81%|████████ | 38957/48008 [2:13:43<29:38, 5.09it/s] 81%|████████ | 38958/48008 [2:13:43<29:04, 5.19it/s] 81%|████████ | 38959/48008 [2:13:43<28:48, 5.24it/s] 81%|████████ | 38960/48008 [2:13:43<28:40, 5.26it/s] 81%|████████ | 38961/48008 [2:13:43<28:34, 5.28it/s] 81%|████████ | 38962/48008 [2:13:44<28:57, 5.21it/s] 81%|████████ | 38963/48008 [2:13:44<28:54, 5.21it/s] 81%|████████ | 38964/48008 [2:13:44<29:05, 5.18it/s] 81%|████████ | 38965/48008 [2:13:44<28:49, 5.23it/s] 81%|████████ | 38966/48008 [2:13:44<28:51, 5.22it/s] 81%|████████ | 38967/48008 [2:13:45<28:48, 5.23it/s] 81%|████████ | 38968/48008 [2:13:45<28:41, 5.25it/s] 81%|████████ | 38969/48008 [2:13:45<28:42, 5.25it/s] 81%|████████ | 38970/48008 [2:13:45<28:59, 5.20it/s] 81%|████████ | 38971/48008 [2:13:45<28:44, 5.24it/s] 81%|████████ | 38972/48008 [2:13:46<28:48, 5.23it/s] 81%|████████ | 38973/48008 [2:13:46<29:05, 5.18it/s] 81%|████████ | 38974/48008 [2:13:46<28:57, 5.20it/s] 81%|████████ | 38975/48008 [2:13:46<29:22, 5.12it/s] 81%|████████ | 38976/48008 [2:13:46<29:07, 5.17it/s] 81%|████████ | 38977/48008 [2:13:47<29:19, 5.13it/s] 81%|████████ | 38978/48008 [2:13:47<29:04, 5.18it/s] 81%|████████ | 38979/48008 [2:13:47<29:03, 5.18it/s] 81%|████████ | 38980/48008 [2:13:47<28:59, 5.19it/s] 81%|████████ | 38981/48008 [2:13:47<29:06, 5.17it/s] 81%|████████ | 38982/48008 [2:13:48<29:00, 5.19it/s] 81%|████████ | 38983/48008 [2:13:48<28:56, 5.20it/s] 81%|████████ | 38984/48008 [2:13:48<29:19, 5.13it/s] 81%|████████ | 38985/48008 [2:13:48<28:48, 5.22it/s] 81%|████████ | 38986/48008 [2:13:48<29:20, 5.12it/s] 81%|████████ | 38987/48008 [2:13:48<28:48, 5.22it/s] 81%|████████ | 38988/48008 [2:13:49<29:07, 5.16it/s] 81%|████████ | 38989/48008 [2:13:49<34:11, 4.40it/s] 81%|████████ | 38990/48008 [2:13:49<33:08, 4.53it/s] 81%|████████ | 38991/48008 [2:13:49<32:11, 4.67it/s] 81%|████████ | 38992/48008 [2:13:50<31:40, 4.74it/s] 81%|████████ | 38993/48008 [2:13:50<30:29, 4.93it/s] 81%|████████ | 38994/48008 [2:13:50<29:42, 5.06it/s] 81%|████████ | 38995/48008 [2:13:50<29:21, 5.12it/s] 81%|████████ | 38996/48008 [2:13:50<29:03, 5.17it/s] 81%|████████ | 38997/48008 [2:13:51<29:31, 5.09it/s] 81%|████████ | 38998/48008 [2:13:51<29:16, 5.13it/s] 81%|████████ | 38999/48008 [2:13:51<28:44, 5.22it/s] 81%|████████ | 39000/48008 [2:13:51<30:50, 4.87it/s] {'loss': 0.3017, 'grad_norm': 0.1767578125, 'learning_rate': 3.753124479253458e-05, 'epoch': 0.81} 81%|████████ | 39000/48008 [2:13:51<30:50, 4.87it/s] 81%|████████ | 39001/48008 [2:13:51<30:11, 4.97it/s] 81%|████████ | 39002/48008 [2:13:52<29:46, 5.04it/s] 81%|████████ | 39003/48008 [2:13:52<29:26, 5.10it/s] 81%|████████ | 39004/48008 [2:13:52<29:47, 5.04it/s] 81%|████████ | 39005/48008 [2:13:52<29:27, 5.09it/s] 81%|████████ | 39006/48008 [2:13:52<28:59, 5.17it/s] 81%|████████▏ | 39007/48008 [2:13:53<28:44, 5.22it/s] 81%|████████▏ | 39008/48008 [2:13:53<28:52, 5.20it/s] 81%|████████▏ | 39009/48008 [2:13:53<28:45, 5.22it/s] 81%|████████▏ | 39010/48008 [2:13:53<28:31, 5.26it/s] 81%|████████▏ | 39011/48008 [2:13:53<28:30, 5.26it/s] 81%|████████▏ | 39012/48008 [2:13:53<29:00, 5.17it/s] 81%|████████▏ | 39013/48008 [2:13:54<29:23, 5.10it/s] 81%|████████▏ | 39014/48008 [2:13:54<29:07, 5.15it/s] 81%|████████▏ | 39015/48008 [2:13:54<29:29, 5.08it/s] 81%|████████▏ | 39016/48008 [2:13:54<29:30, 5.08it/s] 81%|████████▏ | 39017/48008 [2:13:54<29:12, 5.13it/s] 81%|████████▏ | 39018/48008 [2:13:55<29:01, 5.16it/s] 81%|████████▏ | 39019/48008 [2:13:55<29:11, 5.13it/s] 81%|████████▏ | 39020/48008 [2:13:55<29:17, 5.11it/s] 81%|████████▏ | 39021/48008 [2:13:55<28:59, 5.17it/s] 81%|████████▏ | 39022/48008 [2:13:55<28:48, 5.20it/s] 81%|████████▏ | 39023/48008 [2:13:56<28:40, 5.22it/s] 81%|████████▏ | 39024/48008 [2:13:56<33:47, 4.43it/s] 81%|████████▏ | 39025/48008 [2:13:56<32:43, 4.58it/s] 81%|████████▏ | 39026/48008 [2:13:56<31:21, 4.77it/s] 81%|████████▏ | 39027/48008 [2:13:56<30:19, 4.94it/s] 81%|████████▏ | 39028/48008 [2:13:57<30:22, 4.93it/s] 81%|████████▏ | 39029/48008 [2:13:57<29:40, 5.04it/s] 81%|████████▏ | 39030/48008 [2:13:57<29:17, 5.11it/s] 81%|████████▏ | 39031/48008 [2:13:57<28:45, 5.20it/s] 81%|████████▏ | 39032/48008 [2:13:57<29:00, 5.16it/s] 81%|████████▏ | 39033/48008 [2:13:58<28:51, 5.18it/s] 81%|████████▏ | 39034/48008 [2:13:58<28:40, 5.22it/s] 81%|████████▏ | 39035/48008 [2:13:58<28:49, 5.19it/s] 81%|████████▏ | 39036/48008 [2:13:58<29:33, 5.06it/s] 81%|████████▏ | 39037/48008 [2:13:58<29:43, 5.03it/s] 81%|████████▏ | 39038/48008 [2:13:59<29:24, 5.08it/s] 81%|████████▏ | 39039/48008 [2:13:59<28:52, 5.18it/s] 81%|████████▏ | 39040/48008 [2:13:59<28:44, 5.20it/s] 81%|████████▏ | 39041/48008 [2:13:59<28:44, 5.20it/s] 81%|████████▏ | 39042/48008 [2:13:59<29:12, 5.12it/s] 81%|████████▏ | 39043/48008 [2:14:00<29:29, 5.07it/s] 81%|████████▏ | 39044/48008 [2:14:00<29:48, 5.01it/s] 81%|████████▏ | 39045/48008 [2:14:00<29:20, 5.09it/s] 81%|████████▏ | 39046/48008 [2:14:00<29:02, 5.14it/s] 81%|████████▏ | 39047/48008 [2:14:00<29:39, 5.04it/s] 81%|████████▏ | 39048/48008 [2:14:01<29:22, 5.08it/s] 81%|████████▏ | 39049/48008 [2:14:01<29:18, 5.09it/s] 81%|████████▏ | 39050/48008 [2:14:01<28:56, 5.16it/s] {'loss': 0.302, 'grad_norm': 0.341796875, 'learning_rate': 3.73229461756374e-05, 'epoch': 0.81} 81%|████████▏ | 39050/48008 [2:14:01<28:56, 5.16it/s] 81%|████████▏ | 39051/48008 [2:14:01<28:34, 5.23it/s] 81%|████████▏ | 39052/48008 [2:14:01<28:32, 5.23it/s] 81%|████████▏ | 39053/48008 [2:14:02<29:05, 5.13it/s] 81%|████████▏ | 39054/48008 [2:14:02<28:42, 5.20it/s] 81%|████████▏ | 39055/48008 [2:14:02<28:26, 5.25it/s] 81%|████████▏ | 39056/48008 [2:14:02<28:22, 5.26it/s] 81%|████████▏ | 39057/48008 [2:14:02<28:24, 5.25it/s] 81%|████████▏ | 39058/48008 [2:14:02<28:13, 5.29it/s] 81%|████████▏ | 39059/48008 [2:14:03<28:51, 5.17it/s] 81%|████████▏ | 39060/48008 [2:14:03<28:34, 5.22it/s] 81%|████████▏ | 39061/48008 [2:14:03<28:34, 5.22it/s] 81%|████████▏ | 39062/48008 [2:14:03<28:28, 5.24it/s] 81%|████████▏ | 39063/48008 [2:14:03<28:45, 5.18it/s] 81%|████████▏ | 39064/48008 [2:14:04<29:08, 5.12it/s] 81%|████████▏ | 39065/48008 [2:14:04<28:55, 5.15it/s] 81%|████████▏ | 39066/48008 [2:14:04<28:54, 5.15it/s] 81%|████████▏ | 39067/48008 [2:14:04<28:45, 5.18it/s] 81%|████████▏ | 39068/48008 [2:14:04<28:40, 5.20it/s] 81%|████████▏ | 39069/48008 [2:14:05<28:37, 5.20it/s] 81%|████████▏ | 39070/48008 [2:14:05<28:35, 5.21it/s] 81%|████████▏ | 39071/48008 [2:14:05<28:56, 5.15it/s] 81%|████████▏ | 39072/48008 [2:14:05<28:47, 5.17it/s] 81%|████████▏ | 39073/48008 [2:14:05<29:15, 5.09it/s] 81%|████████▏ | 39074/48008 [2:14:06<28:58, 5.14it/s] 81%|████████▏ | 39075/48008 [2:14:06<28:45, 5.18it/s] 81%|████████▏ | 39076/48008 [2:14:06<30:52, 4.82it/s] 81%|████████▏ | 39077/48008 [2:14:06<30:37, 4.86it/s] 81%|████████▏ | 39078/48008 [2:14:06<29:53, 4.98it/s] 81%|████████▏ | 39079/48008 [2:14:07<29:35, 5.03it/s] 81%|████████▏ | 39080/48008 [2:14:07<29:07, 5.11it/s] 81%|████████▏ | 39081/48008 [2:14:07<30:56, 4.81it/s] 81%|████████▏ | 39082/48008 [2:14:07<32:14, 4.61it/s] 81%|████████▏ | 39083/48008 [2:14:07<30:50, 4.82it/s] 81%|████████▏ | 39084/48008 [2:14:08<30:08, 4.94it/s] 81%|████████▏ | 39085/48008 [2:14:08<29:44, 5.00it/s] 81%|████████▏ | 39086/48008 [2:14:08<29:41, 5.01it/s] 81%|████████▏ | 39087/48008 [2:14:09<1:00:17, 2.47it/s] 81%|████████▏ | 39088/48008 [2:14:09<50:40, 2.93it/s] 81%|████████▏ | 39089/48008 [2:14:09<43:50, 3.39it/s] 81%|████████▏ | 39090/48008 [2:14:09<38:55, 3.82it/s] 81%|████████▏ | 39091/48008 [2:14:10<36:17, 4.10it/s] 81%|████████▏ | 39092/48008 [2:14:10<33:57, 4.37it/s] 81%|████████▏ | 39093/48008 [2:14:10<37:28, 3.97it/s] 81%|████████▏ | 39094/48008 [2:14:10<34:35, 4.30it/s] 81%|████████▏ | 39095/48008 [2:14:11<32:39, 4.55it/s] 81%|████████▏ | 39096/48008 [2:14:11<31:24, 4.73it/s] 81%|████████▏ | 39097/48008 [2:14:11<30:36, 4.85it/s] 81%|████████▏ | 39098/48008 [2:14:11<30:21, 4.89it/s] 81%|████████▏ | 39099/48008 [2:14:11<29:46, 4.99it/s] 81%|████████▏ | 39100/48008 [2:14:12<29:38, 5.01it/s] {'loss': 0.2986, 'grad_norm': 0.1884765625, 'learning_rate': 3.711464755874021e-05, 'epoch': 0.81} 81%|████████▏ | 39100/48008 [2:14:12<29:38, 5.01it/s] 81%|████████▏ | 39101/48008 [2:14:12<29:07, 5.10it/s] 81%|████████▏ | 39102/48008 [2:14:12<28:48, 5.15it/s] 81%|████████▏ | 39103/48008 [2:14:12<37:21, 3.97it/s] 81%|████████▏ | 39104/48008 [2:14:13<34:35, 4.29it/s] 81%|████████▏ | 39105/48008 [2:14:13<33:11, 4.47it/s] 81%|████████▏ | 39106/48008 [2:14:13<31:51, 4.66it/s] 81%|████████▏ | 39107/48008 [2:14:13<30:47, 4.82it/s] 81%|████████▏ | 39108/48008 [2:14:13<30:16, 4.90it/s] 81%|████████▏ | 39109/48008 [2:14:13<30:14, 4.90it/s] 81%|████████▏ | 39110/48008 [2:14:14<30:08, 4.92it/s] 81%|████████▏ | 39111/48008 [2:14:14<30:22, 4.88it/s] 81%|████████▏ | 39112/48008 [2:14:14<29:27, 5.03it/s] 81%|████████▏ | 39113/48008 [2:14:14<28:58, 5.12it/s] 81%|████████▏ | 39114/48008 [2:14:15<37:28, 3.96it/s] 81%|████████▏ | 39115/48008 [2:14:15<34:37, 4.28it/s] 81%|████████▏ | 39116/48008 [2:14:16<1:03:41, 2.33it/s] 81%|████████▏ | 39117/48008 [2:14:16<53:33, 2.77it/s] 81%|████████▏ | 39118/48008 [2:14:16<46:02, 3.22it/s] 81%|████████▏ | 39119/48008 [2:14:16<41:19, 3.59it/s] 81%|████████▏ | 39120/48008 [2:14:17<37:41, 3.93it/s] 81%|████████▏ | 39121/48008 [2:14:17<35:06, 4.22it/s] 81%|████████▏ | 39122/48008 [2:14:18<1:02:21, 2.37it/s] 81%|████████▏ | 39123/48008 [2:14:18<52:08, 2.84it/s] 81%|████████▏ | 39124/48008 [2:14:18<45:00, 3.29it/s] 81%|████████▏ | 39125/48008 [2:14:18<40:24, 3.66it/s] 81%|████████▏ | 39126/48008 [2:14:18<36:39, 4.04it/s] 82%|████████▏ | 39127/48008 [2:14:19<34:01, 4.35it/s] 82%|████████▏ | 39128/48008 [2:14:19<1:00:52, 2.43it/s] 82%|████████▏ | 39129/48008 [2:14:20<51:03, 2.90it/s] 82%|████████▏ | 39130/48008 [2:14:20<44:12, 3.35it/s] 82%|████████▏ | 39131/48008 [2:14:20<39:36, 3.74it/s] 82%|████████▏ | 39132/48008 [2:14:20<36:16, 4.08it/s] 82%|████████▏ | 39133/48008 [2:14:20<34:40, 4.27it/s] 82%|████████▏ | 39134/48008 [2:14:21<1:08:38, 2.15it/s] 82%|████████▏ | 39135/48008 [2:14:22<56:46, 2.60it/s] 82%|████████▏ | 39136/48008 [2:14:22<48:12, 3.07it/s] 82%|████████▏ | 39137/48008 [2:14:22<42:46, 3.46it/s] 82%|████████▏ | 39138/48008 [2:14:22<38:25, 3.85it/s] 82%|████████▏ | 39139/48008 [2:14:22<35:36, 4.15it/s] 82%|████████▏ | 39140/48008 [2:14:23<1:02:14, 2.37it/s] 82%|████████▏ | 39141/48008 [2:14:23<51:52, 2.85it/s] 82%|████████▏ | 39142/48008 [2:14:24<44:55, 3.29it/s] 82%|████████▏ | 39143/48008 [2:14:24<40:30, 3.65it/s] 82%|████████▏ | 39144/48008 [2:14:24<37:05, 3.98it/s] 82%|████████▏ | 39145/48008 [2:14:24<34:12, 4.32it/s] 82%|████████▏ | 39146/48008 [2:14:24<32:42, 4.52it/s] 82%|████████▏ | 39147/48008 [2:14:25<31:25, 4.70it/s] 82%|████████▏ | 39148/48008 [2:14:25<30:42, 4.81it/s] 82%|████████▏ | 39149/48008 [2:14:25<29:56, 4.93it/s] 82%|████████▏ | 39150/48008 [2:14:25<29:28, 5.01it/s] 82%|████████▏ | 39150/48008 [2:14:25<29:28, 5.01it/s]{'loss': 0.297, 'grad_norm': 0.35546875, 'learning_rate': 3.6906348941843026e-05, 'epoch': 0.82} 82%|████████▏ | 39151/48008 [2:14:25<29:10, 5.06it/s] 82%|████████▏ | 39152/48008 [2:14:25<28:53, 5.11it/s] 82%|████████▏ | 39153/48008 [2:14:26<28:39, 5.15it/s] 82%|████████▏ | 39154/48008 [2:14:26<28:58, 5.09it/s] 82%|████████▏ | 39155/48008 [2:14:26<28:36, 5.16it/s] 82%|████████▏ | 39156/48008 [2:14:26<30:34, 4.83it/s] 82%|████████▏ | 39157/48008 [2:14:27<30:40, 4.81it/s] 82%|████████▏ | 39158/48008 [2:14:27<29:57, 4.92it/s] 82%|████████▏ | 39159/48008 [2:14:27<29:13, 5.05it/s] 82%|████████▏ | 39160/48008 [2:14:27<29:20, 5.03it/s] 82%|████████▏ | 39161/48008 [2:14:27<29:46, 4.95it/s] 82%|████████▏ | 39162/48008 [2:14:28<31:15, 4.72it/s] 82%|████████▏ | 39163/48008 [2:14:28<30:19, 4.86it/s] 82%|████████▏ | 39164/48008 [2:14:28<29:13, 5.04it/s] 82%|████████▏ | 39165/48008 [2:14:28<28:36, 5.15it/s] 82%|████████▏ | 39166/48008 [2:14:28<28:14, 5.22it/s] 82%|████████▏ | 39167/48008 [2:14:28<28:16, 5.21it/s] 82%|████████▏ | 39168/48008 [2:14:29<27:56, 5.27it/s] 82%|████████▏ | 39169/48008 [2:14:29<27:52, 5.29it/s] 82%|████████▏ | 39170/48008 [2:14:29<27:58, 5.26it/s] 82%|████████▏ | 39171/48008 [2:14:29<28:35, 5.15it/s] 82%|████████▏ | 39172/48008 [2:14:29<28:25, 5.18it/s] 82%|████████▏ | 39173/48008 [2:14:30<28:16, 5.21it/s] 82%|████████▏ | 39174/48008 [2:14:30<28:27, 5.17it/s] 82%|████████▏ | 39175/48008 [2:14:30<28:19, 5.20it/s] 82%|████████▏ | 39176/48008 [2:14:30<28:05, 5.24it/s] 82%|████████▏ | 39177/48008 [2:14:30<28:04, 5.24it/s] 82%|████████▏ | 39178/48008 [2:14:31<27:53, 5.28it/s] 82%|████████▏ | 39179/48008 [2:14:31<27:58, 5.26it/s] 82%|████████▏ | 39180/48008 [2:14:31<28:15, 5.21it/s] 82%|████████▏ | 39181/48008 [2:14:31<28:13, 5.21it/s] 82%|████████▏ | 39182/48008 [2:14:32<57:00, 2.58it/s] 82%|████████▏ | 39183/48008 [2:14:32<48:14, 3.05it/s] 82%|████████▏ | 39184/48008 [2:14:32<42:18, 3.48it/s] 82%|████████▏ | 39185/48008 [2:14:33<38:33, 3.81it/s] 82%|████████▏ | 39186/48008 [2:14:33<35:25, 4.15it/s] 82%|████████▏ | 39187/48008 [2:14:33<33:44, 4.36it/s] 82%|████████▏ | 39188/48008 [2:14:33<32:18, 4.55it/s] 82%|████████▏ | 39189/48008 [2:14:33<31:22, 4.68it/s] 82%|████████▏ | 39190/48008 [2:14:34<30:10, 4.87it/s] 82%|████████▏ | 39191/48008 [2:14:34<29:29, 4.98it/s] 82%|████████▏ | 39192/48008 [2:14:34<29:05, 5.05it/s] 82%|████████▏ | 39193/48008 [2:14:34<28:53, 5.08it/s] 82%|████████▏ | 39194/48008 [2:14:34<28:32, 5.15it/s] 82%|████████▏ | 39195/48008 [2:14:35<28:41, 5.12it/s] 82%|████████▏ | 39196/48008 [2:14:35<28:48, 5.10it/s] 82%|████████▏ | 39197/48008 [2:14:35<28:40, 5.12it/s] 82%|████████▏ | 39198/48008 [2:14:35<28:41, 5.12it/s] 82%|████████▏ | 39199/48008 [2:14:35<28:31, 5.15it/s] 82%|████████▏ | 39200/48008 [2:14:36<30:23, 4.83it/s] {'loss': 0.3005, 'grad_norm': 0.23046875, 'learning_rate': 3.6698050324945844e-05, 'epoch': 0.82} 82%|████████▏ | 39200/48008 [2:14:36<30:23, 4.83it/s] 82%|████████▏ | 39201/48008 [2:14:36<30:30, 4.81it/s] 82%|████████▏ | 39202/48008 [2:14:36<34:53, 4.21it/s] 82%|████████▏ | 39203/48008 [2:14:36<32:49, 4.47it/s] 82%|████████▏ | 39204/48008 [2:14:36<31:31, 4.65it/s] 82%|████████▏ | 39205/48008 [2:14:37<30:40, 4.78it/s] 82%|████████▏ | 39206/48008 [2:14:37<30:12, 4.86it/s] 82%|████████▏ | 39207/48008 [2:14:37<29:34, 4.96it/s] 82%|████████▏ | 39208/48008 [2:14:37<28:51, 5.08it/s] 82%|████████▏ | 39209/48008 [2:14:38<33:41, 4.35it/s] 82%|████████▏ | 39210/48008 [2:14:38<31:56, 4.59it/s] 82%|████████▏ | 39211/48008 [2:14:38<35:47, 4.10it/s] 82%|████████▏ | 39212/48008 [2:14:38<33:22, 4.39it/s] 82%|████████▏ | 39213/48008 [2:14:38<31:41, 4.62it/s] 82%|████████▏ | 39214/48008 [2:14:39<30:35, 4.79it/s] 82%|████████▏ | 39215/48008 [2:14:39<29:33, 4.96it/s] 82%|████████▏ | 39216/48008 [2:14:39<29:57, 4.89it/s] 82%|████████▏ | 39217/48008 [2:14:39<29:54, 4.90it/s] 82%|████████▏ | 39218/48008 [2:14:39<29:37, 4.95it/s] 82%|████████▏ | 39219/48008 [2:14:40<28:42, 5.10it/s] 82%|████████▏ | 39220/48008 [2:14:40<28:32, 5.13it/s] 82%|████████▏ | 39221/48008 [2:14:40<28:58, 5.06it/s] 82%|████████▏ | 39222/48008 [2:14:40<29:29, 4.96it/s] 82%|████████▏ | 39223/48008 [2:14:40<29:03, 5.04it/s] 82%|████████▏ | 39224/48008 [2:14:41<29:11, 5.02it/s] 82%|████████▏ | 39225/48008 [2:14:41<29:06, 5.03it/s] 82%|████████▏ | 39226/48008 [2:14:41<29:25, 4.97it/s] 82%|████████▏ | 39227/48008 [2:14:41<29:11, 5.01it/s] 82%|████████▏ | 39228/48008 [2:14:41<28:51, 5.07it/s] 82%|████████▏ | 39229/48008 [2:14:42<28:45, 5.09it/s] 82%|████████▏ | 39230/48008 [2:14:42<28:23, 5.15it/s] 82%|████████▏ | 39231/48008 [2:14:42<28:37, 5.11it/s] 82%|████████▏ | 39232/48008 [2:14:42<29:09, 5.02it/s] 82%|████████▏ | 39233/48008 [2:14:42<28:47, 5.08it/s] 82%|████████▏ | 39234/48008 [2:14:43<28:29, 5.13it/s] 82%|████████▏ | 39235/48008 [2:14:43<28:15, 5.17it/s] 82%|████████▏ | 39236/48008 [2:14:43<28:09, 5.19it/s] 82%|████████▏ | 39237/48008 [2:14:43<28:09, 5.19it/s] 82%|████████▏ | 39238/48008 [2:14:43<28:29, 5.13it/s] 82%|████████▏ | 39239/48008 [2:14:43<28:21, 5.15it/s] 82%|████████▏ | 39240/48008 [2:14:44<28:20, 5.16it/s] 82%|████████▏ | 39241/48008 [2:14:44<28:14, 5.17it/s] 82%|████████▏ | 39242/48008 [2:14:44<28:24, 5.14it/s] 82%|████████▏ | 39243/48008 [2:14:44<28:33, 5.11it/s] 82%|████████▏ | 39244/48008 [2:14:44<28:11, 5.18it/s] 82%|████████▏ | 39245/48008 [2:14:45<27:52, 5.24it/s] 82%|████████▏ | 39246/48008 [2:14:45<29:52, 4.89it/s] 82%|████████▏ | 39247/48008 [2:14:45<29:11, 5.00it/s] 82%|████████▏ | 39248/48008 [2:14:45<28:49, 5.06it/s] 82%|████████▏ | 39249/48008 [2:14:45<28:37, 5.10it/s] 82%|████████▏ | 39250/48008 [2:14:46<28:57, 5.04it/s] 82%|████████▏ | 39250/48008 [2:14:46<28:57, 5.04it/s]{'loss': 0.2987, 'grad_norm': 0.263671875, 'learning_rate': 3.648975170804866e-05, 'epoch': 0.82} 82%|████████▏ | 39251/48008 [2:14:46<28:42, 5.09it/s] 82%|████████▏ | 39252/48008 [2:14:46<28:29, 5.12it/s] 82%|████████▏ | 39253/48008 [2:14:46<28:19, 5.15it/s] 82%|████████▏ | 39254/48008 [2:14:46<28:02, 5.20it/s] 82%|████████▏ | 39255/48008 [2:14:47<28:18, 5.15it/s] 82%|████████▏ | 39256/48008 [2:14:47<28:11, 5.17it/s] 82%|████████▏ | 39257/48008 [2:14:47<28:02, 5.20it/s] 82%|████████▏ | 39258/48008 [2:14:47<27:48, 5.25it/s] 82%|████████▏ | 39259/48008 [2:14:47<29:56, 4.87it/s] 82%|████████▏ | 39260/48008 [2:14:48<29:12, 4.99it/s] 82%|████████▏ | 39261/48008 [2:14:48<28:42, 5.08it/s] 82%|████████▏ | 39262/48008 [2:14:48<28:39, 5.09it/s] 82%|████████▏ | 39263/48008 [2:14:48<28:24, 5.13it/s] 82%|████████▏ | 39264/48008 [2:14:48<29:05, 5.01it/s] 82%|████████▏ | 39265/48008 [2:14:49<28:54, 5.04it/s] 82%|████████▏ | 39266/48008 [2:14:49<28:34, 5.10it/s] 82%|████████▏ | 39267/48008 [2:14:49<28:02, 5.19it/s] 82%|████████▏ | 39268/48008 [2:14:49<27:44, 5.25it/s] 82%|████████▏ | 39269/48008 [2:14:49<32:48, 4.44it/s] 82%|████████▏ | 39270/48008 [2:14:50<31:29, 4.62it/s] 82%|████████▏ | 39271/48008 [2:14:50<30:17, 4.81it/s] 82%|████████▏ | 39272/48008 [2:14:50<29:24, 4.95it/s] 82%|████████▏ | 39273/48008 [2:14:50<29:44, 4.89it/s] 82%|████████▏ | 39274/48008 [2:14:50<29:09, 4.99it/s] 82%|████████▏ | 39275/48008 [2:14:51<29:01, 5.02it/s] 82%|████████▏ | 39276/48008 [2:14:51<29:04, 5.01it/s] 82%|████████▏ | 39277/48008 [2:14:51<28:30, 5.10it/s] 82%|████████▏ | 39278/48008 [2:14:51<28:10, 5.16it/s] 82%|████████▏ | 39279/48008 [2:14:51<28:02, 5.19it/s] 82%|████████▏ | 39280/48008 [2:14:52<27:56, 5.21it/s] 82%|████████▏ | 39281/48008 [2:14:52<27:47, 5.23it/s] 82%|████████▏ | 39282/48008 [2:14:52<28:33, 5.09it/s] 82%|████████▏ | 39283/48008 [2:14:52<28:48, 5.05it/s] 82%|████████▏ | 39284/48008 [2:14:52<28:27, 5.11it/s] 82%|████████▏ | 39285/48008 [2:14:53<29:12, 4.98it/s] 82%|████████▏ | 39286/48008 [2:14:53<29:31, 4.92it/s] 82%|████████▏ | 39287/48008 [2:14:53<28:55, 5.02it/s] 82%|████████▏ | 39288/48008 [2:14:53<28:25, 5.11it/s] 82%|████████▏ | 39289/48008 [2:14:53<30:11, 4.81it/s] 82%|████████▏ | 39290/48008 [2:14:54<29:25, 4.94it/s] 82%|████████▏ | 39291/48008 [2:14:54<28:55, 5.02it/s] 82%|████████▏ | 39292/48008 [2:14:54<28:41, 5.06it/s] 82%|████████▏ | 39293/48008 [2:14:54<28:23, 5.12it/s] 82%|████████▏ | 39294/48008 [2:14:54<28:09, 5.16it/s] 82%|████████▏ | 39295/48008 [2:14:55<27:56, 5.20it/s] 82%|████████▏ | 39296/48008 [2:14:55<32:51, 4.42it/s] 82%|████████▏ | 39297/48008 [2:14:55<31:18, 4.64it/s] 82%|████████▏ | 39298/48008 [2:14:55<30:13, 4.80it/s] 82%|████████▏ | 39299/48008 [2:14:55<29:46, 4.87it/s] 82%|████████▏ | 39300/48008 [2:14:56<29:07, 4.98it/s] {'loss': 0.3002, 'grad_norm': 0.248046875, 'learning_rate': 3.628145309115147e-05, 'epoch': 0.82} 82%|████████▏ | 39300/48008 [2:14:56<29:07, 4.98it/s] 82%|████████▏ | 39301/48008 [2:14:56<29:07, 4.98it/s] 82%|████████▏ | 39302/48008 [2:14:56<28:57, 5.01it/s] 82%|████████▏ | 39303/48008 [2:14:56<28:21, 5.12it/s] 82%|████████▏ | 39304/48008 [2:14:56<28:03, 5.17it/s] 82%|████████▏ | 39305/48008 [2:14:57<27:51, 5.21it/s] 82%|████████▏ | 39306/48008 [2:14:57<27:36, 5.25it/s] 82%|████████▏ | 39307/48008 [2:14:57<27:43, 5.23it/s] 82%|████████▏ | 39308/48008 [2:14:57<28:11, 5.14it/s] 82%|████████▏ | 39309/48008 [2:14:57<27:58, 5.18it/s] 82%|████████▏ | 39310/48008 [2:14:58<36:22, 3.98it/s] 82%|████████▏ | 39311/48008 [2:14:58<34:14, 4.23it/s] 82%|████████▏ | 39312/48008 [2:14:58<37:15, 3.89it/s] 82%|████████▏ | 39313/48008 [2:14:58<34:21, 4.22it/s] 82%|████████▏ | 39314/48008 [2:14:59<32:14, 4.49it/s] 82%|████████▏ | 39315/48008 [2:14:59<30:56, 4.68it/s] 82%|████████▏ | 39316/48008 [2:14:59<29:58, 4.83it/s] 82%|████████▏ | 39317/48008 [2:14:59<37:44, 3.84it/s] 82%|████████▏ | 39318/48008 [2:15:00<34:37, 4.18it/s] 82%|████████▏ | 39319/48008 [2:15:00<33:01, 4.39it/s] 82%|████████▏ | 39320/48008 [2:15:00<31:41, 4.57it/s] 82%|████████▏ | 39321/48008 [2:15:00<30:54, 4.69it/s] 82%|████████▏ | 39322/48008 [2:15:00<29:58, 4.83it/s] 82%|████████▏ | 39323/48008 [2:15:01<29:06, 4.97it/s] 82%|████████▏ | 39324/48008 [2:15:01<29:09, 4.96it/s] 82%|████████▏ | 39325/48008 [2:15:01<28:31, 5.07it/s] 82%|████████▏ | 39326/48008 [2:15:01<28:16, 5.12it/s] 82%|████████▏ | 39327/48008 [2:15:01<28:05, 5.15it/s] 82%|████████▏ | 39328/48008 [2:15:02<27:51, 5.19it/s] 82%|████████▏ | 39329/48008 [2:15:02<27:45, 5.21it/s] 82%|████████▏ | 39330/48008 [2:15:02<27:48, 5.20it/s] 82%|████████▏ | 39331/48008 [2:15:02<27:45, 5.21it/s] 82%|████████▏ | 39332/48008 [2:15:02<27:34, 5.24it/s] 82%|████████▏ | 39333/48008 [2:15:03<27:28, 5.26it/s] 82%|████████▏ | 39334/48008 [2:15:03<27:35, 5.24it/s] 82%|████████▏ | 39335/48008 [2:15:03<28:04, 5.15it/s] 82%|████████▏ | 39336/48008 [2:15:03<28:20, 5.10it/s] 82%|████████▏ | 39337/48008 [2:15:03<28:34, 5.06it/s] 82%|████████▏ | 39338/48008 [2:15:03<28:06, 5.14it/s] 82%|████████▏ | 39339/48008 [2:15:04<27:54, 5.18it/s] 82%|████████▏ | 39340/48008 [2:15:04<27:40, 5.22it/s] 82%|████████▏ | 39341/48008 [2:15:04<27:39, 5.22it/s] 82%|████████▏ | 39342/48008 [2:15:04<27:41, 5.22it/s] 82%|████████▏ | 39343/48008 [2:15:04<27:41, 5.22it/s] 82%|████████▏ | 39344/48008 [2:15:05<32:42, 4.42it/s] 82%|████████▏ | 39345/48008 [2:15:05<31:08, 4.64it/s] 82%|████████▏ | 39346/48008 [2:15:05<30:36, 4.72it/s] 82%|████████▏ | 39347/48008 [2:15:05<29:40, 4.87it/s] 82%|████████▏ | 39348/48008 [2:15:06<29:00, 4.98it/s] 82%|████████▏ | 39349/48008 [2:15:06<28:51, 5.00it/s] 82%|████████▏ | 39350/48008 [2:15:06<28:20, 5.09it/s] {'loss': 0.2999, 'grad_norm': 0.14453125, 'learning_rate': 3.6073154474254296e-05, 'epoch': 0.82} 82%|████████▏ | 39350/48008 [2:15:06<28:20, 5.09it/s] 82%|████████▏ | 39351/48008 [2:15:06<28:05, 5.14it/s] 82%|████████▏ | 39352/48008 [2:15:06<27:51, 5.18it/s] 82%|████████▏ | 39353/48008 [2:15:06<27:48, 5.19it/s] 82%|████████▏ | 39354/48008 [2:15:07<28:29, 5.06it/s] 82%|████████▏ | 39355/48008 [2:15:07<28:42, 5.02it/s] 82%|████████▏ | 39356/48008 [2:15:07<28:21, 5.08it/s] 82%|████████▏ | 39357/48008 [2:15:07<28:18, 5.09it/s] 82%|████████▏ | 39358/48008 [2:15:07<27:48, 5.18it/s] 82%|████████▏ | 39359/48008 [2:15:08<27:40, 5.21it/s] 82%|████████▏ | 39360/48008 [2:15:08<27:48, 5.18it/s] 82%|████████▏ | 39361/48008 [2:15:08<27:28, 5.25it/s] 82%|████████▏ | 39362/48008 [2:15:08<27:25, 5.25it/s] 82%|████████▏ | 39363/48008 [2:15:08<27:58, 5.15it/s] 82%|████████▏ | 39364/48008 [2:15:09<27:36, 5.22it/s] 82%|████████▏ | 39365/48008 [2:15:09<27:43, 5.19it/s] 82%|████████▏ | 39366/48008 [2:15:09<29:38, 4.86it/s] 82%|████████▏ | 39367/48008 [2:15:09<28:55, 4.98it/s] 82%|████████▏ | 39368/48008 [2:15:09<28:56, 4.97it/s] 82%|████████▏ | 39369/48008 [2:15:10<28:30, 5.05it/s] 82%|████████▏ | 39370/48008 [2:15:10<28:11, 5.11it/s] 82%|████████▏ | 39371/48008 [2:15:10<27:59, 5.14it/s] 82%|████████▏ | 39372/48008 [2:15:10<27:49, 5.17it/s] 82%|████████▏ | 39373/48008 [2:15:10<28:06, 5.12it/s] 82%|████████▏ | 39374/48008 [2:15:11<27:53, 5.16it/s] 82%|████████▏ | 39375/48008 [2:15:11<27:59, 5.14it/s] 82%|████████▏ | 39376/48008 [2:15:11<27:45, 5.18it/s] 82%|████████▏ | 39377/48008 [2:15:11<28:06, 5.12it/s] 82%|████████▏ | 39378/48008 [2:15:12<36:21, 3.96it/s] 82%|████████▏ | 39379/48008 [2:15:12<42:07, 3.41it/s] 82%|████████▏ | 39380/48008 [2:15:12<37:42, 3.81it/s] 82%|████████▏ | 39381/48008 [2:15:12<34:25, 4.18it/s] 82%|████████▏ | 39382/48008 [2:15:13<32:24, 4.44it/s] 82%|████████▏ | 39383/48008 [2:15:13<39:22, 3.65it/s] 82%|████████▏ | 39384/48008 [2:15:13<36:11, 3.97it/s] 82%|████████▏ | 39385/48008 [2:15:13<33:36, 4.28it/s] 82%|████████▏ | 39386/48008 [2:15:13<32:08, 4.47it/s] 82%|████████▏ | 39387/48008 [2:15:14<30:17, 4.74it/s] 82%|████████▏ | 39388/48008 [2:15:14<29:36, 4.85it/s] 82%|████████▏ | 39389/48008 [2:15:14<28:43, 5.00it/s] 82%|████████▏ | 39390/48008 [2:15:14<28:18, 5.07it/s] 82%|████████▏ | 39391/48008 [2:15:14<28:02, 5.12it/s] 82%|████████▏ | 39392/48008 [2:15:15<27:41, 5.18it/s] 82%|████████▏ | 39393/48008 [2:15:15<27:55, 5.14it/s] 82%|████████▏ | 39394/48008 [2:15:15<28:28, 5.04it/s] 82%|████████▏ | 39395/48008 [2:15:15<27:52, 5.15it/s] 82%|████████▏ | 39396/48008 [2:15:16<36:07, 3.97it/s] 82%|████████▏ | 39397/48008 [2:15:16<33:43, 4.25it/s] 82%|████████▏ | 39398/48008 [2:15:16<31:46, 4.52it/s] 82%|████████▏ | 39399/48008 [2:15:16<30:24, 4.72it/s] 82%|████████▏ | 39400/48008 [2:15:16<29:26, 4.87it/s] 82%|████████▏ | 39400/48008 [2:15:16<29:26, 4.87it/s]{'loss': 0.2996, 'grad_norm': 0.34765625, 'learning_rate': 3.586485585735711e-05, 'epoch': 0.82} 82%|████████▏ | 39401/48008 [2:15:17<28:35, 5.02it/s] 82%|████████▏ | 39402/48008 [2:15:17<27:57, 5.13it/s] 82%|████████▏ | 39403/48008 [2:15:17<28:03, 5.11it/s] 82%|████████▏ | 39404/48008 [2:15:17<27:39, 5.19it/s] 82%|████████▏ | 39405/48008 [2:15:17<32:32, 4.41it/s] 82%|████████▏ | 39406/48008 [2:15:18<30:28, 4.70it/s] 82%|████████▏ | 39407/48008 [2:15:18<29:25, 4.87it/s] 82%|████████▏ | 39408/48008 [2:15:18<28:48, 4.98it/s] 82%|████████▏ | 39409/48008 [2:15:18<28:20, 5.06it/s] 82%|████████▏ | 39410/48008 [2:15:18<28:03, 5.11it/s] 82%|████████▏ | 39411/48008 [2:15:19<28:19, 5.06it/s] 82%|████████▏ | 39412/48008 [2:15:19<28:32, 5.02it/s] 82%|████████▏ | 39413/48008 [2:15:19<28:57, 4.95it/s] 82%|████████▏ | 39414/48008 [2:15:19<28:56, 4.95it/s] 82%|████████▏ | 39415/48008 [2:15:20<36:45, 3.90it/s] 82%|████████▏ | 39416/48008 [2:15:20<33:50, 4.23it/s] 82%|████████▏ | 39417/48008 [2:15:20<31:55, 4.49it/s] 82%|████████▏ | 39418/48008 [2:15:20<30:57, 4.62it/s] 82%|████████▏ | 39419/48008 [2:15:20<29:51, 4.79it/s] 82%|████████▏ | 39420/48008 [2:15:21<28:55, 4.95it/s] 82%|████████▏ | 39421/48008 [2:15:21<28:20, 5.05it/s] 82%|████████▏ | 39422/48008 [2:15:21<27:57, 5.12it/s] 82%|████████▏ | 39423/48008 [2:15:21<27:42, 5.16it/s] 82%|████████▏ | 39424/48008 [2:15:21<27:31, 5.20it/s] 82%|████████▏ | 39425/48008 [2:15:21<27:32, 5.19it/s] 82%|████████▏ | 39426/48008 [2:15:22<32:28, 4.41it/s] 82%|████████▏ | 39427/48008 [2:15:22<30:52, 4.63it/s] 82%|████████▏ | 39428/48008 [2:15:22<30:14, 4.73it/s] 82%|████████▏ | 39429/48008 [2:15:22<29:55, 4.78it/s] 82%|████████▏ | 39430/48008 [2:15:23<29:06, 4.91it/s] 82%|████████▏ | 39431/48008 [2:15:23<29:20, 4.87it/s] 82%|████████▏ | 39432/48008 [2:15:23<28:41, 4.98it/s] 82%|████████▏ | 39433/48008 [2:15:23<28:50, 4.95it/s] 82%|████████▏ | 39434/48008 [2:15:23<28:49, 4.96it/s] 82%|████████▏ | 39435/48008 [2:15:24<28:25, 5.03it/s] 82%|████████▏ | 39436/48008 [2:15:24<28:02, 5.10it/s] 82%|████████▏ | 39437/48008 [2:15:24<27:48, 5.14it/s] 82%|████████▏ | 39438/48008 [2:15:24<27:34, 5.18it/s] 82%|████████▏ | 39439/48008 [2:15:24<27:29, 5.19it/s] 82%|████████▏ | 39440/48008 [2:15:25<27:52, 5.12it/s] 82%|████████▏ | 39441/48008 [2:15:25<27:35, 5.17it/s] 82%|████████▏ | 39442/48008 [2:15:25<27:55, 5.11it/s] 82%|████████▏ | 39443/48008 [2:15:25<27:43, 5.15it/s] 82%|████████▏ | 39444/48008 [2:15:25<27:49, 5.13it/s] 82%|████████▏ | 39445/48008 [2:15:26<29:35, 4.82it/s] 82%|████████▏ | 39446/48008 [2:15:26<28:51, 4.94it/s] 82%|████████▏ | 39447/48008 [2:15:26<28:16, 5.05it/s] 82%|████████▏ | 39448/48008 [2:15:26<28:28, 5.01it/s] 82%|████████▏ | 39449/48008 [2:15:26<28:15, 5.05it/s] 82%|████████▏ | 39450/48008 [2:15:27<27:48, 5.13it/s] 82%|████████▏ | 39450/48008 [2:15:27<27:48, 5.13it/s]{'loss': 0.2982, 'grad_norm': 0.1787109375, 'learning_rate': 3.5656557240459924e-05, 'epoch': 0.82} 82%|████████▏ | 39451/48008 [2:15:27<27:30, 5.19it/s] 82%|████████▏ | 39452/48008 [2:15:27<27:22, 5.21it/s] 82%|████████▏ | 39453/48008 [2:15:27<27:39, 5.15it/s] 82%|████████▏ | 39454/48008 [2:15:27<27:51, 5.12it/s] 82%|████████▏ | 39455/48008 [2:15:27<27:43, 5.14it/s] 82%|████████▏ | 39456/48008 [2:15:28<32:26, 4.39it/s] 82%|████████▏ | 39457/48008 [2:15:28<30:50, 4.62it/s] 82%|████████▏ | 39458/48008 [2:15:28<34:43, 4.10it/s] 82%|████████▏ | 39459/48008 [2:15:28<32:24, 4.40it/s] 82%|████████▏ | 39460/48008 [2:15:29<30:36, 4.66it/s] 82%|████████▏ | 39461/48008 [2:15:29<29:35, 4.81it/s] 82%|████████▏ | 39462/48008 [2:15:29<28:46, 4.95it/s] 82%|████████▏ | 39463/48008 [2:15:29<28:35, 4.98it/s] 82%|████████▏ | 39464/48008 [2:15:29<28:56, 4.92it/s] 82%|████████▏ | 39465/48008 [2:15:30<28:25, 5.01it/s] 82%|████████▏ | 39466/48008 [2:15:30<28:12, 5.05it/s] 82%|████████▏ | 39467/48008 [2:15:30<29:52, 4.76it/s] 82%|████████▏ | 39468/48008 [2:15:30<29:23, 4.84it/s] 82%|████████▏ | 39469/48008 [2:15:30<28:39, 4.97it/s] 82%|████████▏ | 39470/48008 [2:15:31<28:09, 5.05it/s] 82%|████████▏ | 39471/48008 [2:15:31<27:50, 5.11it/s] 82%|████████▏ | 39472/48008 [2:15:31<28:27, 5.00it/s] 82%|████████▏ | 39473/48008 [2:15:31<28:04, 5.07it/s] 82%|████████▏ | 39474/48008 [2:15:31<27:40, 5.14it/s] 82%|████████▏ | 39475/48008 [2:15:32<27:07, 5.24it/s] 82%|████████▏ | 39476/48008 [2:15:32<27:02, 5.26it/s] 82%|████████▏ | 39477/48008 [2:15:32<27:15, 5.22it/s] 82%|████████▏ | 39478/48008 [2:15:32<27:01, 5.26it/s] 82%|████████▏ | 39479/48008 [2:15:32<28:55, 4.91it/s] 82%|████████▏ | 39480/48008 [2:15:33<28:15, 5.03it/s] 82%|████████▏ | 39481/48008 [2:15:33<27:49, 5.11it/s] 82%|████████▏ | 39482/48008 [2:15:33<28:06, 5.06it/s] 82%|████████▏ | 39483/48008 [2:15:33<27:56, 5.08it/s] 82%|████████▏ | 39484/48008 [2:15:33<27:39, 5.14it/s] 82%|████████▏ | 39485/48008 [2:15:34<27:38, 5.14it/s] 82%|████████▏ | 39486/48008 [2:15:34<27:09, 5.23it/s] 82%|████████▏ | 39487/48008 [2:15:34<29:10, 4.87it/s] 82%|████████▏ | 39488/48008 [2:15:34<28:34, 4.97it/s] 82%|████████▏ | 39489/48008 [2:15:34<28:07, 5.05it/s] 82%|████████▏ | 39490/48008 [2:15:35<27:37, 5.14it/s] 82%|████████▏ | 39491/48008 [2:15:35<28:01, 5.06it/s] 82%|████████▏ | 39492/48008 [2:15:35<27:32, 5.15it/s] 82%|████████▏ | 39493/48008 [2:15:35<28:00, 5.07it/s] 82%|████████▏ | 39494/48008 [2:15:35<27:50, 5.10it/s] 82%|████████▏ | 39495/48008 [2:15:36<27:30, 5.16it/s] 82%|████████▏ | 39496/48008 [2:15:36<27:10, 5.22it/s] 82%|████████▏ | 39497/48008 [2:15:36<27:07, 5.23it/s] 82%|████████▏ | 39498/48008 [2:15:36<32:00, 4.43it/s] 82%|████████▏ | 39499/48008 [2:15:36<30:25, 4.66it/s] 82%|████████▏ | 39500/48008 [2:15:37<29:12, 4.86it/s] {'loss': 0.3009, 'grad_norm': 0.32421875, 'learning_rate': 3.544825862356274e-05, 'epoch': 0.82} 82%|████████▏ | 39500/48008 [2:15:37<29:12, 4.86it/s] 82%|████████▏ | 39501/48008 [2:15:37<28:26, 4.99it/s] 82%|████████▏ | 39502/48008 [2:15:37<29:55, 4.74it/s] 82%|████████▏ | 39503/48008 [2:15:37<29:12, 4.85it/s] 82%|████████▏ | 39504/48008 [2:15:37<28:22, 4.99it/s] 82%|████████▏ | 39505/48008 [2:15:38<27:57, 5.07it/s] 82%|████████▏ | 39506/48008 [2:15:38<28:06, 5.04it/s] 82%|████████▏ | 39507/48008 [2:15:38<27:55, 5.07it/s] 82%|████████▏ | 39508/48008 [2:15:38<28:11, 5.03it/s] 82%|████████▏ | 39509/48008 [2:15:38<27:23, 5.17it/s] 82%|████████▏ | 39510/48008 [2:15:39<27:16, 5.19it/s] 82%|████████▏ | 39511/48008 [2:15:39<26:54, 5.26it/s] 82%|████████▏ | 39512/48008 [2:15:39<26:55, 5.26it/s] 82%|████████▏ | 39513/48008 [2:15:39<26:49, 5.28it/s] 82%|████████▏ | 39514/48008 [2:15:39<27:11, 5.21it/s] 82%|████████▏ | 39515/48008 [2:15:40<27:54, 5.07it/s] 82%|████████▏ | 39516/48008 [2:15:40<27:40, 5.11it/s] 82%|████████▏ | 39517/48008 [2:15:40<27:46, 5.09it/s] 82%|████████▏ | 39518/48008 [2:15:40<27:22, 5.17it/s] 82%|████████▏ | 39519/48008 [2:15:40<29:17, 4.83it/s] 82%|████████▏ | 39520/48008 [2:15:41<28:33, 4.95it/s] 82%|████████▏ | 39521/48008 [2:15:41<28:03, 5.04it/s] 82%|████████▏ | 39522/48008 [2:15:41<27:31, 5.14it/s] 82%|████████▏ | 39523/48008 [2:15:41<27:19, 5.17it/s] 82%|████████▏ | 39524/48008 [2:15:41<27:39, 5.11it/s] 82%|████████▏ | 39525/48008 [2:15:41<27:14, 5.19it/s] 82%|████████▏ | 39526/48008 [2:15:42<27:00, 5.23it/s] 82%|████████▏ | 39527/48008 [2:15:42<27:28, 5.15it/s] 82%|████████▏ | 39528/48008 [2:15:42<27:20, 5.17it/s] 82%|████████▏ | 39529/48008 [2:15:42<27:40, 5.11it/s] 82%|████████▏ | 39530/48008 [2:15:42<27:25, 5.15it/s] 82%|████████▏ | 39531/48008 [2:15:43<27:16, 5.18it/s] 82%|████████▏ | 39532/48008 [2:15:43<27:08, 5.21it/s] 82%|████████▏ | 39533/48008 [2:15:43<26:55, 5.25it/s] 82%|████████▏ | 39534/48008 [2:15:43<26:43, 5.28it/s] 82%|████████▏ | 39535/48008 [2:15:43<26:49, 5.27it/s] 82%|████████▏ | 39536/48008 [2:15:44<27:01, 5.23it/s] 82%|████████▏ | 39537/48008 [2:15:44<26:52, 5.25it/s] 82%|████████▏ | 39538/48008 [2:15:44<26:55, 5.24it/s] 82%|████████▏ | 39539/48008 [2:15:44<27:07, 5.20it/s] 82%|████████▏ | 39540/48008 [2:15:44<27:25, 5.14it/s] 82%|████████▏ | 39541/48008 [2:15:45<29:14, 4.83it/s] 82%|████████▏ | 39542/48008 [2:15:45<28:36, 4.93it/s] 82%|████████▏ | 39543/48008 [2:15:45<28:24, 4.97it/s] 82%|████████▏ | 39544/48008 [2:15:45<27:54, 5.05it/s] 82%|████████▏ | 39545/48008 [2:15:45<28:07, 5.02it/s] 82%|████████▏ | 39546/48008 [2:15:46<27:38, 5.10it/s] 82%|████████▏ | 39547/48008 [2:15:46<27:31, 5.12it/s] 82%|████████▏ | 39548/48008 [2:15:46<27:20, 5.16it/s] 82%|████████▏ | 39549/48008 [2:15:46<27:41, 5.09it/s] 82%|████████▏ | 39550/48008 [2:15:46<27:22, 5.15it/s] {'loss': 0.3018, 'grad_norm': 0.1513671875, 'learning_rate': 3.523996000666556e-05, 'epoch': 0.82} 82%|████████▏ | 39550/48008 [2:15:46<27:22, 5.15it/s] 82%|████████▏ | 39551/48008 [2:15:47<27:25, 5.14it/s] 82%|████████▏ | 39552/48008 [2:15:47<27:05, 5.20it/s] 82%|████████▏ | 39553/48008 [2:15:47<27:17, 5.16it/s] 82%|████████▏ | 39554/48008 [2:15:47<27:12, 5.18it/s] 82%|████████▏ | 39555/48008 [2:15:48<35:28, 3.97it/s] 82%|████████▏ | 39556/48008 [2:15:48<33:28, 4.21it/s] 82%|████████▏ | 39557/48008 [2:15:48<32:04, 4.39it/s] 82%|████████▏ | 39558/48008 [2:15:48<30:43, 4.58it/s] 82%|████████▏ | 39559/48008 [2:15:48<29:54, 4.71it/s] 82%|████████▏ | 39560/48008 [2:15:49<28:55, 4.87it/s] 82%|████████▏ | 39561/48008 [2:15:49<28:05, 5.01it/s] 82%|████████▏ | 39562/48008 [2:15:49<28:05, 5.01it/s] 82%|████████▏ | 39563/48008 [2:15:49<27:45, 5.07it/s] 82%|████████▏ | 39564/48008 [2:15:49<27:45, 5.07it/s] 82%|████████▏ | 39565/48008 [2:15:49<27:27, 5.13it/s] 82%|████████▏ | 39566/48008 [2:15:50<27:34, 5.10it/s] 82%|████████▏ | 39567/48008 [2:15:50<27:21, 5.14it/s] 82%|████████▏ | 39568/48008 [2:15:50<27:01, 5.21it/s] 82%|████████▏ | 39569/48008 [2:15:50<26:59, 5.21it/s] 82%|████████▏ | 39570/48008 [2:15:50<26:55, 5.22it/s] 82%|████████▏ | 39571/48008 [2:15:51<27:10, 5.17it/s] 82%|████████▏ | 39572/48008 [2:15:51<27:04, 5.19it/s] 82%|████████▏ | 39573/48008 [2:15:51<27:16, 5.15it/s] 82%|████████▏ | 39574/48008 [2:15:51<27:04, 5.19it/s] 82%|████████▏ | 39575/48008 [2:15:51<27:10, 5.17it/s] 82%|████████▏ | 39576/48008 [2:15:52<27:32, 5.10it/s] 82%|████████▏ | 39577/48008 [2:15:52<27:33, 5.10it/s] 82%|████████▏ | 39578/48008 [2:15:52<27:21, 5.13it/s] 82%|████████▏ | 39579/48008 [2:15:52<27:21, 5.14it/s] 82%|████████▏ | 39580/48008 [2:15:52<29:11, 4.81it/s] 82%|████████▏ | 39581/48008 [2:15:53<28:25, 4.94it/s] 82%|████████▏ | 39582/48008 [2:15:53<29:50, 4.71it/s] 82%|████████▏ | 39583/48008 [2:15:53<28:55, 4.86it/s] 82%|████████▏ | 39584/48008 [2:15:53<30:12, 4.65it/s] 82%|████████▏ | 39585/48008 [2:15:53<29:52, 4.70it/s] 82%|████████▏ | 39586/48008 [2:15:54<28:56, 4.85it/s] 82%|████████▏ | 39587/48008 [2:15:54<28:21, 4.95it/s] 82%|████████▏ | 39588/48008 [2:15:54<32:46, 4.28it/s] 82%|████████▏ | 39589/48008 [2:15:54<31:30, 4.45it/s] 82%|████████▏ | 39590/48008 [2:15:55<30:09, 4.65it/s] 82%|████████▏ | 39591/48008 [2:15:55<29:08, 4.81it/s] 82%|████████▏ | 39592/48008 [2:15:55<28:26, 4.93it/s] 82%|████████▏ | 39593/48008 [2:15:55<27:56, 5.02it/s] 82%|████████▏ | 39594/48008 [2:15:55<27:39, 5.07it/s] 82%|████████▏ | 39595/48008 [2:15:56<27:27, 5.11it/s] 82%|████████▏ | 39596/48008 [2:15:56<27:12, 5.15it/s] 82%|████████▏ | 39597/48008 [2:15:56<27:11, 5.15it/s] 82%|████████▏ | 39598/48008 [2:15:56<27:05, 5.17it/s] 82%|████████▏ | 39599/48008 [2:15:56<26:59, 5.19it/s] 82%|████████▏ | 39600/48008 [2:15:57<35:05, 3.99it/s] {'loss': 0.2979, 'grad_norm': 0.55859375, 'learning_rate': 3.503166138976837e-05, 'epoch': 0.82} 82%|████████▏ | 39600/48008 [2:15:57<35:05, 3.99it/s] 82%|████████▏ | 39601/48008 [2:15:57<32:35, 4.30it/s] 82%|████████▏ | 39602/48008 [2:15:57<30:43, 4.56it/s] 82%|████████▏ | 39603/48008 [2:15:57<29:15, 4.79it/s] 82%|████████▏ | 39604/48008 [2:15:57<28:59, 4.83it/s] 82%|████████▏ | 39605/48008 [2:15:58<28:28, 4.92it/s] 82%|████████▏ | 39606/48008 [2:15:58<27:49, 5.03it/s] 83%|████████▎ | 39607/48008 [2:15:58<27:54, 5.02it/s] 83%|████████▎ | 39608/48008 [2:15:58<27:17, 5.13it/s] 83%|████████▎ | 39609/48008 [2:15:58<27:38, 5.06it/s] 83%|████████▎ | 39610/48008 [2:15:59<27:23, 5.11it/s] 83%|████████▎ | 39611/48008 [2:15:59<26:59, 5.18it/s] 83%|████████▎ | 39612/48008 [2:15:59<26:51, 5.21it/s] 83%|████████▎ | 39613/48008 [2:15:59<34:59, 4.00it/s] 83%|████████▎ | 39614/48008 [2:16:00<32:39, 4.28it/s] 83%|████████▎ | 39615/48008 [2:16:00<30:52, 4.53it/s] 83%|████████▎ | 39616/48008 [2:16:00<29:42, 4.71it/s] 83%|████████▎ | 39617/48008 [2:16:00<28:59, 4.82it/s] 83%|████████▎ | 39618/48008 [2:16:00<28:17, 4.94it/s] 83%|████████▎ | 39619/48008 [2:16:01<29:49, 4.69it/s] 83%|████████▎ | 39620/48008 [2:16:01<29:36, 4.72it/s] 83%|████████▎ | 39621/48008 [2:16:01<28:49, 4.85it/s] 83%|████████▎ | 39622/48008 [2:16:01<28:23, 4.92it/s] 83%|████████▎ | 39623/48008 [2:16:01<27:41, 5.05it/s] 83%|████████▎ | 39624/48008 [2:16:02<35:30, 3.94it/s] 83%|████████▎ | 39625/48008 [2:16:02<32:41, 4.27it/s] 83%|████████▎ | 39626/48008 [2:16:02<30:57, 4.51it/s] 83%|████████▎ | 39627/48008 [2:16:02<29:31, 4.73it/s] 83%|████████▎ | 39628/48008 [2:16:03<29:25, 4.75it/s] 83%|████████▎ | 39629/48008 [2:16:03<36:46, 3.80it/s] 83%|████████▎ | 39630/48008 [2:16:03<33:50, 4.13it/s] 83%|████████▎ | 39631/48008 [2:16:03<31:44, 4.40it/s] 83%|████████▎ | 39632/48008 [2:16:03<30:04, 4.64it/s] 83%|████████▎ | 39633/48008 [2:16:04<28:53, 4.83it/s] 83%|████████▎ | 39634/48008 [2:16:04<28:13, 4.94it/s] 83%|████████▎ | 39635/48008 [2:16:04<28:11, 4.95it/s] 83%|████████▎ | 39636/48008 [2:16:04<27:37, 5.05it/s] 83%|████████▎ | 39637/48008 [2:16:04<27:39, 5.05it/s] 83%|████████▎ | 39638/48008 [2:16:05<27:21, 5.10it/s] 83%|████████▎ | 39639/48008 [2:16:05<27:10, 5.13it/s] 83%|████████▎ | 39640/48008 [2:16:05<26:52, 5.19it/s] 83%|████████▎ | 39641/48008 [2:16:05<26:52, 5.19it/s] 83%|████████▎ | 39642/48008 [2:16:05<27:10, 5.13it/s] 83%|████████▎ | 39643/48008 [2:16:06<27:47, 5.02it/s] 83%|████████▎ | 39644/48008 [2:16:06<27:54, 4.99it/s] 83%|████████▎ | 39645/48008 [2:16:06<27:45, 5.02it/s] 83%|████████▎ | 39646/48008 [2:16:06<27:08, 5.13it/s] 83%|████████▎ | 39647/48008 [2:16:06<27:26, 5.08it/s] 83%|████████▎ | 39648/48008 [2:16:07<31:59, 4.36it/s] 83%|████████▎ | 39649/48008 [2:16:07<30:17, 4.60it/s] 83%|████████▎ | 39650/48008 [2:16:07<29:06, 4.79it/s] {'loss': 0.2987, 'grad_norm': 0.203125, 'learning_rate': 3.4823362772871186e-05, 'epoch': 0.83} 83%|████████▎ | 39650/48008 [2:16:07<29:06, 4.79it/s] 83%|████████▎ | 39651/48008 [2:16:07<36:27, 3.82it/s] 83%|████████▎ | 39652/48008 [2:16:08<33:21, 4.17it/s] 83%|████████▎ | 39653/48008 [2:16:08<31:19, 4.44it/s] 83%|████████▎ | 39654/48008 [2:16:08<29:59, 4.64it/s] 83%|████████▎ | 39655/48008 [2:16:08<29:31, 4.71it/s] 83%|████████▎ | 39656/48008 [2:16:08<28:49, 4.83it/s] 83%|████████▎ | 39657/48008 [2:16:09<28:19, 4.91it/s] 83%|████████▎ | 39658/48008 [2:16:09<27:39, 5.03it/s] 83%|████████▎ | 39659/48008 [2:16:09<29:13, 4.76it/s] 83%|████████▎ | 39660/48008 [2:16:09<36:32, 3.81it/s] 83%|████████▎ | 39661/48008 [2:16:10<33:13, 4.19it/s] 83%|████████▎ | 39662/48008 [2:16:10<33:10, 4.19it/s] 83%|████████▎ | 39663/48008 [2:16:10<31:15, 4.45it/s] 83%|████████▎ | 39664/48008 [2:16:10<30:09, 4.61it/s] 83%|████████▎ | 39665/48008 [2:16:10<29:49, 4.66it/s] 83%|████████▎ | 39666/48008 [2:16:11<28:39, 4.85it/s] 83%|████████▎ | 39667/48008 [2:16:11<27:53, 4.98it/s] 83%|████████▎ | 39668/48008 [2:16:11<27:13, 5.10it/s] 83%|████████▎ | 39669/48008 [2:16:11<27:21, 5.08it/s] 83%|████████▎ | 39670/48008 [2:16:11<26:57, 5.15it/s] 83%|████████▎ | 39671/48008 [2:16:12<26:56, 5.16it/s] 83%|████████▎ | 39672/48008 [2:16:12<26:51, 5.17it/s] 83%|████████▎ | 39673/48008 [2:16:12<26:40, 5.21it/s] 83%|████████▎ | 39674/48008 [2:16:12<26:19, 5.27it/s] 83%|████████▎ | 39675/48008 [2:16:12<26:50, 5.17it/s] 83%|████████▎ | 39676/48008 [2:16:13<26:51, 5.17it/s] 83%|████████▎ | 39677/48008 [2:16:13<26:46, 5.18it/s] 83%|████████▎ | 39678/48008 [2:16:13<26:17, 5.28it/s] 83%|████████▎ | 39679/48008 [2:16:13<26:29, 5.24it/s] 83%|████████▎ | 39680/48008 [2:16:13<26:24, 5.26it/s] 83%|████████▎ | 39681/48008 [2:16:14<26:17, 5.28it/s] 83%|████████▎ | 39682/48008 [2:16:14<26:19, 5.27it/s] 83%|████████▎ | 39683/48008 [2:16:14<26:17, 5.28it/s] 83%|████████▎ | 39684/48008 [2:16:14<26:30, 5.23it/s] 83%|████████▎ | 39685/48008 [2:16:14<26:24, 5.25it/s] 83%|████████▎ | 39686/48008 [2:16:14<26:16, 5.28it/s] 83%|████████▎ | 39687/48008 [2:16:15<26:03, 5.32it/s] 83%|████████▎ | 39688/48008 [2:16:15<26:15, 5.28it/s] 83%|████████▎ | 39689/48008 [2:16:15<26:29, 5.23it/s] 83%|████████▎ | 39690/48008 [2:16:15<26:30, 5.23it/s] 83%|████████▎ | 39691/48008 [2:16:15<26:30, 5.23it/s] 83%|████████▎ | 39692/48008 [2:16:16<28:21, 4.89it/s] 83%|████████▎ | 39693/48008 [2:16:16<27:50, 4.98it/s] 83%|████████▎ | 39694/48008 [2:16:16<27:43, 5.00it/s] 83%|████████▎ | 39695/48008 [2:16:16<27:23, 5.06it/s] 83%|████████▎ | 39696/48008 [2:16:16<26:40, 5.19it/s] 83%|████████▎ | 39697/48008 [2:16:17<27:04, 5.12it/s] 83%|████████▎ | 39698/48008 [2:16:17<27:18, 5.07it/s] 83%|████████▎ | 39699/48008 [2:16:17<27:30, 5.03it/s] 83%|████████▎ | 39700/48008 [2:16:17<27:10, 5.10it/s] {'loss': 0.3001, 'grad_norm': 0.1884765625, 'learning_rate': 3.461506415597401e-05, 'epoch': 0.83} 83%|████████▎ | 39700/48008 [2:16:17<27:10, 5.10it/s] 83%|████████▎ | 39701/48008 [2:16:17<26:52, 5.15it/s] 83%|████████▎ | 39702/48008 [2:16:18<26:58, 5.13it/s] 83%|████████▎ | 39703/48008 [2:16:18<26:52, 5.15it/s] 83%|████████▎ | 39704/48008 [2:16:18<26:47, 5.17it/s] 83%|████████▎ | 39705/48008 [2:16:18<26:52, 5.15it/s] 83%|████████▎ | 39706/48008 [2:16:18<26:42, 5.18it/s] 83%|████████▎ | 39707/48008 [2:16:19<26:33, 5.21it/s] 83%|████████▎ | 39708/48008 [2:16:19<26:28, 5.22it/s] 83%|████████▎ | 39709/48008 [2:16:20<1:07:41, 2.04it/s] 83%|████████▎ | 39710/48008 [2:16:20<55:43, 2.48it/s] 83%|████████▎ | 39711/48008 [2:16:20<46:51, 2.95it/s] 83%|████████▎ | 39712/48008 [2:16:21<40:42, 3.40it/s] 83%|████████▎ | 39713/48008 [2:16:21<36:18, 3.81it/s] 83%|████████▎ | 39714/48008 [2:16:21<33:19, 4.15it/s] 83%|████████▎ | 39715/48008 [2:16:21<31:08, 4.44it/s] 83%|████████▎ | 39716/48008 [2:16:21<29:35, 4.67it/s] 83%|████████▎ | 39717/48008 [2:16:21<28:48, 4.80it/s] 83%|████████▎ | 39718/48008 [2:16:22<28:07, 4.91it/s] 83%|████████▎ | 39719/48008 [2:16:22<27:34, 5.01it/s] 83%|████████▎ | 39720/48008 [2:16:22<27:13, 5.07it/s] 83%|████████▎ | 39721/48008 [2:16:22<26:58, 5.12it/s] 83%|████████▎ | 39722/48008 [2:16:22<26:46, 5.16it/s] 83%|████████▎ | 39723/48008 [2:16:23<26:32, 5.20it/s] 83%|████████▎ | 39724/48008 [2:16:23<26:57, 5.12it/s] 83%|████████▎ | 39725/48008 [2:16:23<28:38, 4.82it/s] 83%|████████▎ | 39726/48008 [2:16:23<27:52, 4.95it/s] 83%|████████▎ | 39727/48008 [2:16:23<27:55, 4.94it/s] 83%|████████▎ | 39728/48008 [2:16:24<32:09, 4.29it/s] 83%|████████▎ | 39729/48008 [2:16:25<59:03, 2.34it/s] 83%|████████▎ | 39730/48008 [2:16:25<49:18, 2.80it/s] 83%|████████▎ | 39731/48008 [2:16:25<42:24, 3.25it/s] 83%|████████▎ | 39732/48008 [2:16:25<37:28, 3.68it/s] 83%|████████▎ | 39733/48008 [2:16:25<34:09, 4.04it/s] 83%|████████▎ | 39734/48008 [2:16:26<31:48, 4.33it/s] 83%|████████▎ | 39735/48008 [2:16:26<56:01, 2.46it/s] 83%|████████▎ | 39736/48008 [2:16:27<47:02, 2.93it/s] 83%|████████▎ | 39737/48008 [2:16:27<41:08, 3.35it/s] 83%|████████▎ | 39738/48008 [2:16:27<37:12, 3.70it/s] 83%|████████▎ | 39739/48008 [2:16:27<34:10, 4.03it/s] 83%|████████▎ | 39740/48008 [2:16:27<31:42, 4.35it/s] 83%|████████▎ | 39741/48008 [2:16:28<1:00:29, 2.28it/s] 83%|████████▎ | 39742/48008 [2:16:28<50:21, 2.74it/s] 83%|████████▎ | 39743/48008 [2:16:29<43:09, 3.19it/s] 83%|████████▎ | 39744/48008 [2:16:29<38:06, 3.61it/s] 83%|████████▎ | 39745/48008 [2:16:29<34:38, 3.98it/s] 83%|████████▎ | 39746/48008 [2:16:29<32:01, 4.30it/s] 83%|████████▎ | 39747/48008 [2:16:30<1:03:02, 2.18it/s] 83%|████████▎ | 39748/48008 [2:16:30<51:46, 2.66it/s] 83%|████████▎ | 39749/48008 [2:16:31<44:05, 3.12it/s] 83%|████████▎ | 39750/48008 [2:16:31<39:11, 3.51it/s] {'loss': 0.2987, 'grad_norm': 0.376953125, 'learning_rate': 3.440676553907682e-05, 'epoch': 0.83} 83%|████████▎ | 39750/48008 [2:16:31<39:11, 3.51it/s] 83%|████████▎ | 39751/48008 [2:16:31<35:17, 3.90it/s] 83%|████████▎ | 39752/48008 [2:16:31<32:57, 4.18it/s] 83%|████████▎ | 39753/48008 [2:16:32<57:05, 2.41it/s] 83%|████████▎ | 39754/48008 [2:16:32<48:12, 2.85it/s] 83%|████████▎ | 39755/48008 [2:16:32<43:29, 3.16it/s] 83%|████████▎ | 39756/48008 [2:16:33<38:26, 3.58it/s] 83%|████████▎ | 39757/48008 [2:16:33<34:33, 3.98it/s] 83%|████████▎ | 39758/48008 [2:16:33<32:08, 4.28it/s] 83%|████████▎ | 39759/48008 [2:16:33<30:23, 4.52it/s] 83%|████████▎ | 39760/48008 [2:16:34<33:55, 4.05it/s] 83%|████████▎ | 39761/48008 [2:16:34<32:05, 4.28it/s] 83%|████████▎ | 39762/48008 [2:16:34<30:12, 4.55it/s] 83%|████████▎ | 39763/48008 [2:16:34<28:49, 4.77it/s] 83%|████████▎ | 39764/48008 [2:16:34<28:01, 4.90it/s] 83%|████████▎ | 39765/48008 [2:16:34<27:27, 5.00it/s] 83%|████████▎ | 39766/48008 [2:16:35<27:04, 5.07it/s] 83%|████████▎ | 39767/48008 [2:16:35<26:48, 5.12it/s] 83%|████████▎ | 39768/48008 [2:16:35<27:06, 5.06it/s] 83%|████████▎ | 39769/48008 [2:16:35<26:48, 5.12it/s] 83%|████████▎ | 39770/48008 [2:16:35<26:33, 5.17it/s] 83%|████████▎ | 39771/48008 [2:16:36<26:22, 5.20it/s] 83%|████████▎ | 39772/48008 [2:16:36<34:20, 4.00it/s] 83%|████████▎ | 39773/48008 [2:16:36<31:42, 4.33it/s] 83%|████████▎ | 39774/48008 [2:16:36<31:58, 4.29it/s] 83%|████████▎ | 39775/48008 [2:16:37<30:10, 4.55it/s] 83%|████████▎ | 39776/48008 [2:16:37<33:47, 4.06it/s] 83%|████████▎ | 39777/48008 [2:16:37<31:31, 4.35it/s] 83%|████████▎ | 39778/48008 [2:16:37<34:36, 3.96it/s] 83%|████████▎ | 39779/48008 [2:16:38<32:33, 4.21it/s] 83%|████████▎ | 39780/48008 [2:16:38<30:35, 4.48it/s] 83%|████████▎ | 39781/48008 [2:16:38<29:11, 4.70it/s] 83%|████████▎ | 39782/48008 [2:16:38<28:41, 4.78it/s] 83%|████████▎ | 39783/48008 [2:16:38<27:52, 4.92it/s] 83%|████████▎ | 39784/48008 [2:16:39<26:55, 5.09it/s] 83%|████████▎ | 39785/48008 [2:16:39<26:42, 5.13it/s] 83%|████████▎ | 39786/48008 [2:16:39<27:01, 5.07it/s] 83%|████████▎ | 39787/48008 [2:16:39<26:37, 5.15it/s] 83%|████████▎ | 39788/48008 [2:16:39<26:26, 5.18it/s] 83%|████████▎ | 39789/48008 [2:16:40<26:16, 5.21it/s] 83%|████████▎ | 39790/48008 [2:16:40<26:11, 5.23it/s] 83%|████████▎ | 39791/48008 [2:16:40<26:15, 5.22it/s] 83%|████████▎ | 39792/48008 [2:16:40<26:06, 5.24it/s] 83%|████████▎ | 39793/48008 [2:16:40<26:29, 5.17it/s] 83%|████████▎ | 39794/48008 [2:16:41<26:23, 5.19it/s] 83%|████████▎ | 39795/48008 [2:16:41<52:34, 2.60it/s] 83%|████████▎ | 39796/48008 [2:16:42<44:32, 3.07it/s] 83%|████████▎ | 39797/48008 [2:16:42<40:52, 3.35it/s] 83%|████████▎ | 39798/48008 [2:16:42<36:23, 3.76it/s] 83%|████████▎ | 39799/48008 [2:16:42<33:48, 4.05it/s] 83%|████████▎ | 39800/48008 [2:16:42<31:56, 4.28it/s] {'loss': 0.3, 'grad_norm': 0.216796875, 'learning_rate': 3.419846692217964e-05, 'epoch': 0.83} 83%|████████▎ | 39800/48008 [2:16:42<31:56, 4.28it/s] 83%|████████▎ | 39801/48008 [2:16:43<30:23, 4.50it/s] 83%|████████▎ | 39802/48008 [2:16:43<28:51, 4.74it/s] 83%|████████▎ | 39803/48008 [2:16:43<28:06, 4.87it/s] 83%|████████▎ | 39804/48008 [2:16:43<27:21, 5.00it/s] 83%|████████▎ | 39805/48008 [2:16:43<31:52, 4.29it/s] 83%|████████▎ | 39806/48008 [2:16:44<30:07, 4.54it/s] 83%|████████▎ | 39807/48008 [2:16:44<28:59, 4.72it/s] 83%|████████▎ | 39808/48008 [2:16:44<28:09, 4.85it/s] 83%|████████▎ | 39809/48008 [2:16:44<28:04, 4.87it/s] 83%|████████▎ | 39810/48008 [2:16:44<27:28, 4.97it/s] 83%|████████▎ | 39811/48008 [2:16:45<27:13, 5.02it/s] 83%|████████▎ | 39812/48008 [2:16:45<27:04, 5.05it/s] 83%|████████▎ | 39813/48008 [2:16:45<26:33, 5.14it/s] 83%|████████▎ | 39814/48008 [2:16:45<26:24, 5.17it/s] 83%|████████▎ | 39815/48008 [2:16:45<26:37, 5.13it/s] 83%|████████▎ | 39816/48008 [2:16:46<26:28, 5.16it/s] 83%|████████▎ | 39817/48008 [2:16:46<26:12, 5.21it/s] 83%|████████▎ | 39818/48008 [2:16:46<25:59, 5.25it/s] 83%|████████▎ | 39819/48008 [2:16:46<26:07, 5.22it/s] 83%|████████▎ | 39820/48008 [2:16:46<26:02, 5.24it/s] 83%|████████▎ | 39821/48008 [2:16:47<26:13, 5.20it/s] 83%|████████▎ | 39822/48008 [2:16:47<26:37, 5.13it/s] 83%|████████▎ | 39823/48008 [2:16:47<26:27, 5.16it/s] 83%|████████▎ | 39824/48008 [2:16:47<26:15, 5.19it/s] 83%|████████▎ | 39825/48008 [2:16:47<28:03, 4.86it/s] 83%|████████▎ | 39826/48008 [2:16:48<28:15, 4.82it/s] 83%|████████▎ | 39827/48008 [2:16:48<27:28, 4.96it/s] 83%|████████▎ | 39828/48008 [2:16:48<26:52, 5.07it/s] 83%|████████▎ | 39829/48008 [2:16:48<26:37, 5.12it/s] 83%|████████▎ | 39830/48008 [2:16:48<26:14, 5.20it/s] 83%|████████▎ | 39831/48008 [2:16:49<30:51, 4.42it/s] 83%|████████▎ | 39832/48008 [2:16:49<29:14, 4.66it/s] 83%|████████▎ | 39833/48008 [2:16:49<28:23, 4.80it/s] 83%|████████▎ | 39834/48008 [2:16:49<27:42, 4.92it/s] 83%|████████▎ | 39835/48008 [2:16:49<27:28, 4.96it/s] 83%|████████▎ | 39836/48008 [2:16:50<26:43, 5.10it/s] 83%|████████▎ | 39837/48008 [2:16:50<26:22, 5.16it/s] 83%|████████▎ | 39838/48008 [2:16:50<26:18, 5.18it/s] 83%|████████▎ | 39839/48008 [2:16:50<26:09, 5.20it/s] 83%|████████▎ | 39840/48008 [2:16:50<26:12, 5.19it/s] 83%|████████▎ | 39841/48008 [2:16:51<26:05, 5.22it/s] 83%|████████▎ | 39842/48008 [2:16:51<25:58, 5.24it/s] 83%|████████▎ | 39843/48008 [2:16:51<25:53, 5.25it/s] 83%|████████▎ | 39844/48008 [2:16:51<26:04, 5.22it/s] 83%|████████▎ | 39845/48008 [2:16:51<27:52, 4.88it/s] 83%|████████▎ | 39846/48008 [2:16:52<27:15, 4.99it/s] 83%|████████▎ | 39847/48008 [2:16:52<26:50, 5.07it/s] 83%|████████▎ | 39848/48008 [2:16:52<26:59, 5.04it/s] 83%|████████▎ | 39849/48008 [2:16:52<26:40, 5.10it/s] 83%|████████▎ | 39850/48008 [2:16:52<26:58, 5.04it/s] {'loss': 0.3005, 'grad_norm': 0.30078125, 'learning_rate': 3.3990168305282456e-05, 'epoch': 0.83} 83%|████████▎ | 39850/48008 [2:16:52<26:58, 5.04it/s] 83%|████████▎ | 39851/48008 [2:16:53<27:09, 5.01it/s] 83%|████████▎ | 39852/48008 [2:16:53<27:00, 5.03it/s] 83%|████████▎ | 39853/48008 [2:16:53<26:51, 5.06it/s] 83%|████████▎ | 39854/48008 [2:16:53<31:14, 4.35it/s] 83%|████████▎ | 39855/48008 [2:16:53<29:36, 4.59it/s] 83%|████████▎ | 39856/48008 [2:16:54<28:24, 4.78it/s] 83%|████████▎ | 39857/48008 [2:16:54<27:38, 4.91it/s] 83%|████████▎ | 39858/48008 [2:16:54<26:40, 5.09it/s] 83%|████████▎ | 39859/48008 [2:16:54<26:26, 5.14it/s] 83%|████████▎ | 39860/48008 [2:16:54<26:27, 5.13it/s] 83%|████████▎ | 39861/48008 [2:16:55<26:05, 5.20it/s] 83%|████████▎ | 39862/48008 [2:16:55<26:04, 5.21it/s] 83%|████████▎ | 39863/48008 [2:16:55<26:27, 5.13it/s] 83%|████████▎ | 39864/48008 [2:16:55<28:09, 4.82it/s] 83%|████████▎ | 39865/48008 [2:16:55<27:22, 4.96it/s] 83%|████████▎ | 39866/48008 [2:16:56<27:43, 4.89it/s] 83%|████████▎ | 39867/48008 [2:16:56<29:03, 4.67it/s] 83%|████████▎ | 39868/48008 [2:16:56<30:01, 4.52it/s] 83%|████████▎ | 39869/48008 [2:16:56<29:16, 4.63it/s] 83%|████████▎ | 39870/48008 [2:16:56<27:58, 4.85it/s] 83%|████████▎ | 39871/48008 [2:16:57<27:14, 4.98it/s] 83%|████████▎ | 39872/48008 [2:16:57<26:46, 5.06it/s] 83%|████████▎ | 39873/48008 [2:16:57<26:24, 5.13it/s] 83%|████████▎ | 39874/48008 [2:16:57<26:13, 5.17it/s] 83%|████████▎ | 39875/48008 [2:16:57<26:50, 5.05it/s] 83%|████████▎ | 39876/48008 [2:16:58<26:44, 5.07it/s] 83%|████████▎ | 39877/48008 [2:16:58<26:20, 5.14it/s] 83%|████████▎ | 39878/48008 [2:16:58<25:53, 5.23it/s] 83%|████████▎ | 39879/48008 [2:16:58<25:53, 5.23it/s] 83%|████████▎ | 39880/48008 [2:16:58<25:42, 5.27it/s] 83%|████████▎ | 39881/48008 [2:16:59<25:35, 5.29it/s] 83%|████████▎ | 39882/48008 [2:16:59<25:22, 5.34it/s] 83%|████████▎ | 39883/48008 [2:16:59<25:28, 5.32it/s] 83%|████████▎ | 39884/48008 [2:16:59<25:46, 5.25it/s] 83%|████████▎ | 39885/48008 [2:16:59<25:45, 5.26it/s] 83%|████████▎ | 39886/48008 [2:16:59<25:30, 5.31it/s] 83%|████████▎ | 39887/48008 [2:17:00<26:00, 5.20it/s] 83%|████████▎ | 39888/48008 [2:17:00<25:43, 5.26it/s] 83%|████████▎ | 39889/48008 [2:17:00<25:59, 5.21it/s] 83%|████████▎ | 39890/48008 [2:17:00<25:44, 5.26it/s] 83%|████████▎ | 39891/48008 [2:17:00<25:45, 5.25it/s] 83%|████████▎ | 39892/48008 [2:17:01<25:45, 5.25it/s] 83%|████████▎ | 39893/48008 [2:17:01<25:41, 5.26it/s] 83%|████████▎ | 39894/48008 [2:17:01<25:16, 5.35it/s] 83%|████████▎ | 39895/48008 [2:17:01<25:13, 5.36it/s] 83%|████████▎ | 39896/48008 [2:17:01<25:21, 5.33it/s] 83%|████████▎ | 39897/48008 [2:17:02<25:30, 5.30it/s] 83%|████████▎ | 39898/48008 [2:17:02<25:27, 5.31it/s] 83%|████████▎ | 39899/48008 [2:17:02<25:37, 5.27it/s] 83%|████████▎ | 39900/48008 [2:17:02<25:27, 5.31it/s] {'loss': 0.3045, 'grad_norm': 0.2236328125, 'learning_rate': 3.378186968838527e-05, 'epoch': 0.83} 83%|████████▎ | 39900/48008 [2:17:02<25:27, 5.31it/s] 83%|████████▎ | 39901/48008 [2:17:02<25:49, 5.23it/s] 83%|████████▎ | 39902/48008 [2:17:03<30:29, 4.43it/s] 83%|████████▎ | 39903/48008 [2:17:03<29:07, 4.64it/s] 83%|████████▎ | 39904/48008 [2:17:03<27:47, 4.86it/s] 83%|████████▎ | 39905/48008 [2:17:03<27:04, 4.99it/s] 83%|████████▎ | 39906/48008 [2:17:03<26:41, 5.06it/s] 83%|████████▎ | 39907/48008 [2:17:04<26:12, 5.15it/s] 83%|████████▎ | 39908/48008 [2:17:04<25:55, 5.21it/s] 83%|████████▎ | 39909/48008 [2:17:04<25:48, 5.23it/s] 83%|████████▎ | 39910/48008 [2:17:04<25:43, 5.25it/s] 83%|████████▎ | 39911/48008 [2:17:04<25:36, 5.27it/s] 83%|████████▎ | 39912/48008 [2:17:04<25:38, 5.26it/s] 83%|████████▎ | 39913/48008 [2:17:05<25:51, 5.22it/s] 83%|████████▎ | 39914/48008 [2:17:05<26:19, 5.13it/s] 83%|████████▎ | 39915/48008 [2:17:05<26:28, 5.09it/s] 83%|████████▎ | 39916/48008 [2:17:05<26:33, 5.08it/s] 83%|████████▎ | 39917/48008 [2:17:05<26:15, 5.14it/s] 83%|████████▎ | 39918/48008 [2:17:06<26:00, 5.19it/s] 83%|████████▎ | 39919/48008 [2:17:06<25:52, 5.21it/s] 83%|████████▎ | 39920/48008 [2:17:06<25:49, 5.22it/s] 83%|████████▎ | 39921/48008 [2:17:06<26:00, 5.18it/s] 83%|████████▎ | 39922/48008 [2:17:06<26:22, 5.11it/s] 83%|████████▎ | 39923/48008 [2:17:07<26:13, 5.14it/s] 83%|████████▎ | 39924/48008 [2:17:07<26:04, 5.17it/s] 83%|████████▎ | 39925/48008 [2:17:07<25:59, 5.18it/s] 83%|████████▎ | 39926/48008 [2:17:07<26:07, 5.16it/s] 83%|████████▎ | 39927/48008 [2:17:07<26:20, 5.11it/s] 83%|████████▎ | 39928/48008 [2:17:08<26:36, 5.06it/s] 83%|████████▎ | 39929/48008 [2:17:08<26:06, 5.16it/s] 83%|████████▎ | 39930/48008 [2:17:08<26:01, 5.17it/s] 83%|████████▎ | 39931/48008 [2:17:08<26:14, 5.13it/s] 83%|████████▎ | 39932/48008 [2:17:08<25:52, 5.20it/s] 83%|████████▎ | 39933/48008 [2:17:09<25:47, 5.22it/s] 83%|████████▎ | 39934/48008 [2:17:09<25:45, 5.23it/s] 83%|████████▎ | 39935/48008 [2:17:09<25:30, 5.27it/s] 83%|████████▎ | 39936/48008 [2:17:09<25:44, 5.23it/s] 83%|████████▎ | 39937/48008 [2:17:09<25:56, 5.18it/s] 83%|████████▎ | 39938/48008 [2:17:10<27:47, 4.84it/s] 83%|████████▎ | 39939/48008 [2:17:10<27:08, 4.96it/s] 83%|████████▎ | 39940/48008 [2:17:10<26:26, 5.09it/s] 83%|████████▎ | 39941/48008 [2:17:10<26:13, 5.13it/s] 83%|████████▎ | 39942/48008 [2:17:10<26:07, 5.14it/s] 83%|████████▎ | 39943/48008 [2:17:11<26:27, 5.08it/s] 83%|████████▎ | 39944/48008 [2:17:11<26:14, 5.12it/s] 83%|████████▎ | 39945/48008 [2:17:11<26:06, 5.15it/s] 83%|████████▎ | 39946/48008 [2:17:11<25:56, 5.18it/s] 83%|████████▎ | 39947/48008 [2:17:11<25:38, 5.24it/s] 83%|████████▎ | 39948/48008 [2:17:11<25:39, 5.24it/s] 83%|████████▎ | 39949/48008 [2:17:12<25:30, 5.27it/s] 83%|████████▎ | 39950/48008 [2:17:12<25:31, 5.26it/s] {'loss': 0.303, 'grad_norm': 0.3046875, 'learning_rate': 3.3573571071488084e-05, 'epoch': 0.83} 83%|████████▎ | 39950/48008 [2:17:12<25:31, 5.26it/s] 83%|████████▎ | 39951/48008 [2:17:12<25:22, 5.29it/s] 83%|████████▎ | 39952/48008 [2:17:12<25:29, 5.27it/s] 83%|████████▎ | 39953/48008 [2:17:12<25:30, 5.26it/s] 83%|████████▎ | 39954/48008 [2:17:13<25:28, 5.27it/s] 83%|████████▎ | 39955/48008 [2:17:13<25:28, 5.27it/s] 83%|████████▎ | 39956/48008 [2:17:13<25:22, 5.29it/s] 83%|████████▎ | 39957/48008 [2:17:13<25:19, 5.30it/s] 83%|████████▎ | 39958/48008 [2:17:13<26:07, 5.14it/s] 83%|████████▎ | 39959/48008 [2:17:14<25:47, 5.20it/s] 83%|████████▎ | 39960/48008 [2:17:14<33:36, 3.99it/s] 83%|████████▎ | 39961/48008 [2:17:14<30:52, 4.34it/s] 83%|████████▎ | 39962/48008 [2:17:14<29:20, 4.57it/s] 83%|████████▎ | 39963/48008 [2:17:15<28:08, 4.76it/s] 83%|████████▎ | 39964/48008 [2:17:15<27:51, 4.81it/s] 83%|████████▎ | 39965/48008 [2:17:15<27:43, 4.84it/s] 83%|████████▎ | 39966/48008 [2:17:15<27:18, 4.91it/s] 83%|████████▎ | 39967/48008 [2:17:16<34:36, 3.87it/s] 83%|████████▎ | 39968/48008 [2:17:16<32:20, 4.14it/s] 83%|████████▎ | 39969/48008 [2:17:16<30:08, 4.44it/s] 83%|████████▎ | 39970/48008 [2:17:16<28:29, 4.70it/s] 83%|████████▎ | 39971/48008 [2:17:16<27:46, 4.82it/s] 83%|████████▎ | 39972/48008 [2:17:16<26:59, 4.96it/s] 83%|████████▎ | 39973/48008 [2:17:17<27:03, 4.95it/s] 83%|████████▎ | 39974/48008 [2:17:17<26:25, 5.07it/s] 83%|████████▎ | 39975/48008 [2:17:17<25:46, 5.19it/s] 83%|████████▎ | 39976/48008 [2:17:17<25:30, 5.25it/s] 83%|████████▎ | 39977/48008 [2:17:17<25:30, 5.25it/s] 83%|████████▎ | 39978/48008 [2:17:18<25:22, 5.28it/s] 83%|████████▎ | 39979/48008 [2:17:18<25:49, 5.18it/s] 83%|████████▎ | 39980/48008 [2:17:18<26:08, 5.12it/s] 83%|████████▎ | 39981/48008 [2:17:18<25:59, 5.15it/s] 83%|████████▎ | 39982/48008 [2:17:18<25:52, 5.17it/s] 83%|████████▎ | 39983/48008 [2:17:19<26:02, 5.13it/s] 83%|████████▎ | 39984/48008 [2:17:19<25:47, 5.18it/s] 83%|████████▎ | 39985/48008 [2:17:19<25:40, 5.21it/s] 83%|████████▎ | 39986/48008 [2:17:19<25:51, 5.17it/s] 83%|████████▎ | 39987/48008 [2:17:19<25:50, 5.17it/s] 83%|████████▎ | 39988/48008 [2:17:20<25:45, 5.19it/s] 83%|████████▎ | 39989/48008 [2:17:20<25:36, 5.22it/s] 83%|████████▎ | 39990/48008 [2:17:20<25:37, 5.22it/s] 83%|████████▎ | 39991/48008 [2:17:20<25:43, 5.19it/s] 83%|████████▎ | 39992/48008 [2:17:20<25:39, 5.21it/s] 83%|████████▎ | 39993/48008 [2:17:21<25:29, 5.24it/s] 83%|████████▎ | 39994/48008 [2:17:21<25:27, 5.25it/s] 83%|████████▎ | 39995/48008 [2:17:21<25:38, 5.21it/s] 83%|████████▎ | 39996/48008 [2:17:21<25:37, 5.21it/s] 83%|████████▎ | 39997/48008 [2:17:21<25:29, 5.24it/s] 83%|████████▎ | 39998/48008 [2:17:21<25:43, 5.19it/s] 83%|████████▎ | 39999/48008 [2:17:22<25:51, 5.16it/s] 83%|████████▎ | 40000/48008 [2:17:22<25:31, 5.23it/s] {'loss': 0.3035, 'grad_norm': 0.357421875, 'learning_rate': 3.33652724545909e-05, 'epoch': 0.83} 83%|████████▎ | 40000/48008 [2:17:22<25:31, 5.23it/s] 83%|████████▎ | 40001/48008 [2:17:22<25:23, 5.26it/s] 83%|████████▎ | 40002/48008 [2:17:22<25:11, 5.30it/s] 83%|████████▎ | 40003/48008 [2:17:22<25:14, 5.29it/s] 83%|████████▎ | 40004/48008 [2:17:23<25:47, 5.17it/s] 83%|████████▎ | 40005/48008 [2:17:23<25:32, 5.22it/s] 83%|████████▎ | 40006/48008 [2:17:23<25:29, 5.23it/s] 83%|████████▎ | 40007/48008 [2:17:23<25:23, 5.25it/s] 83%|████████▎ | 40008/48008 [2:17:23<25:08, 5.30it/s] 83%|████████▎ | 40009/48008 [2:17:24<25:02, 5.32it/s] 83%|████████▎ | 40010/48008 [2:17:24<25:11, 5.29it/s] 83%|████████▎ | 40011/48008 [2:17:24<25:16, 5.27it/s] 83%|████████▎ | 40012/48008 [2:17:24<25:08, 5.30it/s] 83%|████████▎ | 40013/48008 [2:17:24<27:05, 4.92it/s] 83%|████████▎ | 40014/48008 [2:17:25<31:13, 4.27it/s] 83%|████████▎ | 40015/48008 [2:17:25<29:42, 4.48it/s] 83%|████████▎ | 40016/48008 [2:17:26<55:59, 2.38it/s] 83%|████████▎ | 40017/48008 [2:17:26<47:09, 2.82it/s] 83%|████████▎ | 40018/48008 [2:17:26<48:27, 2.75it/s] 83%|████████▎ | 40019/48008 [2:17:27<41:20, 3.22it/s] 83%|████████▎ | 40020/48008 [2:17:27<36:36, 3.64it/s] 83%|████████▎ | 40021/48008 [2:17:27<41:02, 3.24it/s] 83%|████████▎ | 40022/48008 [2:17:27<36:34, 3.64it/s] 83%|████████▎ | 40023/48008 [2:17:27<33:05, 4.02it/s] 83%|████████▎ | 40024/48008 [2:17:28<30:46, 4.32it/s] 83%|████████▎ | 40025/48008 [2:17:28<29:39, 4.49it/s] 83%|████████▎ | 40026/48008 [2:17:28<30:09, 4.41it/s] 83%|████████▎ | 40027/48008 [2:17:28<28:42, 4.63it/s] 83%|████████▎ | 40028/48008 [2:17:29<27:50, 4.78it/s] 83%|████████▎ | 40029/48008 [2:17:29<27:02, 4.92it/s] 83%|████████▎ | 40030/48008 [2:17:29<26:31, 5.01it/s] 83%|████████▎ | 40031/48008 [2:17:29<26:12, 5.07it/s] 83%|████████▎ | 40032/48008 [2:17:29<26:22, 5.04it/s] 83%|████████▎ | 40033/48008 [2:17:29<26:32, 5.01it/s] 83%|████████▎ | 40034/48008 [2:17:30<28:00, 4.74it/s] 83%|████████▎ | 40035/48008 [2:17:30<27:01, 4.92it/s] 83%|████████▎ | 40036/48008 [2:17:30<26:31, 5.01it/s] 83%|████████▎ | 40037/48008 [2:17:30<25:59, 5.11it/s] 83%|████████▎ | 40038/48008 [2:17:30<25:34, 5.19it/s] 83%|████████▎ | 40039/48008 [2:17:31<30:06, 4.41it/s] 83%|████████▎ | 40040/48008 [2:17:31<28:40, 4.63it/s] 83%|████████▎ | 40041/48008 [2:17:31<27:23, 4.85it/s] 83%|████████▎ | 40042/48008 [2:17:31<26:46, 4.96it/s] 83%|████████▎ | 40043/48008 [2:17:32<26:24, 5.03it/s] 83%|████████▎ | 40044/48008 [2:17:32<26:30, 5.01it/s] 83%|████████▎ | 40045/48008 [2:17:32<33:53, 3.92it/s] 83%|████████▎ | 40046/48008 [2:17:32<31:07, 4.26it/s] 83%|████████▎ | 40047/48008 [2:17:33<29:53, 4.44it/s] 83%|████████▎ | 40048/48008 [2:17:33<28:32, 4.65it/s] 83%|████████▎ | 40049/48008 [2:17:33<27:30, 4.82it/s] 83%|████████▎ | 40050/48008 [2:17:33<27:02, 4.90it/s] {'loss': 0.2993, 'grad_norm': 0.44921875, 'learning_rate': 3.315697383769372e-05, 'epoch': 0.83} 83%|████████▎ | 40050/48008 [2:17:33<27:02, 4.90it/s] 83%|████████▎ | 40051/48008 [2:17:33<26:36, 4.99it/s] 83%|████████▎ | 40052/48008 [2:17:33<26:27, 5.01it/s] 83%|████████▎ | 40053/48008 [2:17:34<25:55, 5.12it/s] 83%|████████▎ | 40054/48008 [2:17:34<25:31, 5.19it/s] 83%|████████▎ | 40055/48008 [2:17:34<25:38, 5.17it/s] 83%|████████▎ | 40056/48008 [2:17:34<25:34, 5.18it/s] 83%|████████▎ | 40057/48008 [2:17:34<25:46, 5.14it/s] 83%|████████▎ | 40058/48008 [2:17:35<25:47, 5.14it/s] 83%|████████▎ | 40059/48008 [2:17:35<25:25, 5.21it/s] 83%|████████▎ | 40060/48008 [2:17:35<29:56, 4.42it/s] 83%|████████▎ | 40061/48008 [2:17:35<28:19, 4.68it/s] 83%|████████▎ | 40062/48008 [2:17:35<27:21, 4.84it/s] 83%|████████▎ | 40063/48008 [2:17:36<27:08, 4.88it/s] 83%|████████▎ | 40064/48008 [2:17:36<27:04, 4.89it/s] 83%|████████▎ | 40065/48008 [2:17:36<26:30, 4.99it/s] 83%|████████▎ | 40066/48008 [2:17:36<26:03, 5.08it/s] 83%|████████▎ | 40067/48008 [2:17:36<26:15, 5.04it/s] 83%|████████▎ | 40068/48008 [2:17:37<26:24, 5.01it/s] 83%|████████▎ | 40069/48008 [2:17:37<25:54, 5.11it/s] 83%|████████▎ | 40070/48008 [2:17:37<25:31, 5.18it/s] 83%|████████▎ | 40071/48008 [2:17:37<26:09, 5.06it/s] 83%|████████▎ | 40072/48008 [2:17:37<25:43, 5.14it/s] 83%|████████▎ | 40073/48008 [2:17:38<25:49, 5.12it/s] 83%|████████▎ | 40074/48008 [2:17:38<25:37, 5.16it/s] 83%|████████▎ | 40075/48008 [2:17:38<26:06, 5.07it/s] 83%|████████▎ | 40076/48008 [2:17:38<25:52, 5.11it/s] 83%|████████▎ | 40077/48008 [2:17:38<25:43, 5.14it/s] 83%|████████▎ | 40078/48008 [2:17:39<27:23, 4.82it/s] 83%|████████▎ | 40079/48008 [2:17:39<26:37, 4.96it/s] 83%|████████▎ | 40080/48008 [2:17:39<28:01, 4.72it/s] 83%|████████▎ | 40081/48008 [2:17:39<34:52, 3.79it/s] 83%|████████▎ | 40082/48008 [2:17:40<32:22, 4.08it/s] 83%|████████▎ | 40083/48008 [2:17:40<30:08, 4.38it/s] 83%|████████▎ | 40084/48008 [2:17:40<28:40, 4.61it/s] 83%|████████▎ | 40085/48008 [2:17:40<27:37, 4.78it/s] 83%|████████▎ | 40086/48008 [2:17:40<27:04, 4.88it/s] 84%|████████▎ | 40087/48008 [2:17:41<28:24, 4.65it/s] 84%|████████▎ | 40088/48008 [2:17:41<27:28, 4.80it/s] 84%|████████▎ | 40089/48008 [2:17:41<27:31, 4.79it/s] 84%|████████▎ | 40090/48008 [2:17:41<26:42, 4.94it/s] 84%|████████▎ | 40091/48008 [2:17:41<26:13, 5.03it/s] 84%|████████▎ | 40092/48008 [2:17:42<25:49, 5.11it/s] 84%|████████▎ | 40093/48008 [2:17:42<25:31, 5.17it/s] 84%|████████▎ | 40094/48008 [2:17:42<25:30, 5.17it/s] 84%|████████▎ | 40095/48008 [2:17:42<25:31, 5.17it/s] 84%|████████▎ | 40096/48008 [2:17:42<25:41, 5.13it/s] 84%|████████▎ | 40097/48008 [2:17:43<25:43, 5.13it/s] 84%|████████▎ | 40098/48008 [2:17:43<25:38, 5.14it/s] 84%|████████▎ | 40099/48008 [2:17:43<25:29, 5.17it/s] 84%|████████▎ | 40100/48008 [2:17:43<25:34, 5.15it/s] {'loss': 0.2985, 'grad_norm': 0.373046875, 'learning_rate': 3.2948675220796536e-05, 'epoch': 0.84} 84%|████████▎ | 40100/48008 [2:17:43<25:34, 5.15it/s] 84%|████████▎ | 40101/48008 [2:17:44<33:21, 3.95it/s] 84%|████████▎ | 40102/48008 [2:17:44<30:53, 4.26it/s] 84%|████████▎ | 40103/48008 [2:17:44<29:07, 4.52it/s] 84%|████████▎ | 40104/48008 [2:17:44<28:27, 4.63it/s] 84%|████████▎ | 40105/48008 [2:17:44<27:28, 4.79it/s] 84%|████████▎ | 40106/48008 [2:17:45<26:44, 4.92it/s] 84%|████████▎ | 40107/48008 [2:17:45<28:08, 4.68it/s] 84%|████████▎ | 40108/48008 [2:17:45<29:07, 4.52it/s] 84%|████████▎ | 40109/48008 [2:17:45<27:58, 4.71it/s] 84%|████████▎ | 40110/48008 [2:17:45<27:00, 4.87it/s] 84%|████████▎ | 40111/48008 [2:17:46<26:24, 4.98it/s] 84%|████████▎ | 40112/48008 [2:17:46<26:01, 5.06it/s] 84%|████████▎ | 40113/48008 [2:17:46<25:43, 5.12it/s] 84%|████████▎ | 40114/48008 [2:17:46<25:15, 5.21it/s] 84%|████████▎ | 40115/48008 [2:17:46<25:07, 5.24it/s] 84%|████████▎ | 40116/48008 [2:17:47<25:53, 5.08it/s] 84%|████████▎ | 40117/48008 [2:17:47<25:36, 5.13it/s] 84%|████████▎ | 40118/48008 [2:17:47<25:31, 5.15it/s] 84%|████████▎ | 40119/48008 [2:17:47<25:12, 5.21it/s] 84%|████████▎ | 40120/48008 [2:17:47<25:18, 5.19it/s] 84%|████████▎ | 40121/48008 [2:17:48<25:02, 5.25it/s] 84%|████████▎ | 40122/48008 [2:17:48<25:24, 5.17it/s] 84%|████████▎ | 40123/48008 [2:17:48<27:10, 4.84it/s] 84%|████████▎ | 40124/48008 [2:17:48<26:33, 4.95it/s] 84%|████████▎ | 40125/48008 [2:17:48<26:33, 4.95it/s] 84%|████████▎ | 40126/48008 [2:17:49<26:22, 4.98it/s] 84%|████████▎ | 40127/48008 [2:17:49<26:24, 4.97it/s] 84%|████████▎ | 40128/48008 [2:17:49<26:44, 4.91it/s] 84%|████████▎ | 40129/48008 [2:17:49<26:06, 5.03it/s] 84%|████████▎ | 40130/48008 [2:17:49<26:15, 5.00it/s] 84%|████████▎ | 40131/48008 [2:17:50<27:42, 4.74it/s] 84%|████████▎ | 40132/48008 [2:17:50<27:06, 4.84it/s] 84%|████████▎ | 40133/48008 [2:17:50<26:39, 4.92it/s] 84%|████████▎ | 40134/48008 [2:17:50<26:33, 4.94it/s] 84%|████████▎ | 40135/48008 [2:17:50<25:57, 5.06it/s] 84%|████████▎ | 40136/48008 [2:17:51<25:44, 5.10it/s] 84%|████████▎ | 40137/48008 [2:17:51<25:27, 5.15it/s] 84%|████████▎ | 40138/48008 [2:17:51<25:07, 5.22it/s] 84%|████████▎ | 40139/48008 [2:17:51<25:20, 5.18it/s] 84%|████████▎ | 40140/48008 [2:17:51<27:03, 4.85it/s] 84%|████████▎ | 40141/48008 [2:17:52<26:27, 4.96it/s] 84%|████████▎ | 40142/48008 [2:17:52<25:52, 5.07it/s] 84%|████████▎ | 40143/48008 [2:17:52<25:48, 5.08it/s] 84%|████████▎ | 40144/48008 [2:17:52<26:02, 5.03it/s] 84%|████████▎ | 40145/48008 [2:17:52<25:44, 5.09it/s] 84%|████████▎ | 40146/48008 [2:17:53<25:44, 5.09it/s] 84%|████████▎ | 40147/48008 [2:17:53<25:32, 5.13it/s] 84%|████████▎ | 40148/48008 [2:17:53<25:15, 5.19it/s] 84%|████████▎ | 40149/48008 [2:17:53<25:10, 5.20it/s] 84%|████████▎ | 40150/48008 [2:17:53<25:08, 5.21it/s] {'loss': 0.3009, 'grad_norm': 0.140625, 'learning_rate': 3.274037660389935e-05, 'epoch': 0.84} 84%|████████▎ | 40150/48008 [2:17:53<25:08, 5.21it/s] 84%|████████▎ | 40151/48008 [2:17:53<24:57, 5.25it/s] 84%|████████▎ | 40152/48008 [2:17:54<24:57, 5.24it/s] 84%|████████▎ | 40153/48008 [2:17:54<25:31, 5.13it/s] 84%|████████▎ | 40154/48008 [2:17:54<26:14, 4.99it/s] 84%|████████▎ | 40155/48008 [2:17:54<25:59, 5.04it/s] 84%|████████▎ | 40156/48008 [2:17:54<25:43, 5.09it/s] 84%|████████▎ | 40157/48008 [2:17:55<26:13, 4.99it/s] 84%|████████▎ | 40158/48008 [2:17:55<25:55, 5.05it/s] 84%|████████▎ | 40159/48008 [2:17:55<25:48, 5.07it/s] 84%|████████▎ | 40160/48008 [2:17:55<25:23, 5.15it/s] 84%|████████▎ | 40161/48008 [2:17:55<25:50, 5.06it/s] 84%|████████▎ | 40162/48008 [2:17:56<26:00, 5.03it/s] 84%|████████▎ | 40163/48008 [2:17:56<25:41, 5.09it/s] 84%|████████▎ | 40164/48008 [2:17:56<25:40, 5.09it/s] 84%|████████▎ | 40165/48008 [2:17:56<25:14, 5.18it/s] 84%|████████▎ | 40166/48008 [2:17:56<25:25, 5.14it/s] 84%|████████▎ | 40167/48008 [2:17:57<27:04, 4.83it/s] 84%|████████▎ | 40168/48008 [2:17:57<26:21, 4.96it/s] 84%|████████▎ | 40169/48008 [2:17:57<27:42, 4.72it/s] 84%|████████▎ | 40170/48008 [2:17:57<27:07, 4.82it/s] 84%|████████▎ | 40171/48008 [2:17:57<26:21, 4.95it/s] 84%|████████▎ | 40172/48008 [2:17:58<25:52, 5.05it/s] 84%|████████▎ | 40173/48008 [2:17:58<25:56, 5.03it/s] 84%|████████▎ | 40174/48008 [2:17:58<25:50, 5.05it/s] 84%|████████▎ | 40175/48008 [2:17:58<26:18, 4.96it/s] 84%|████████▎ | 40176/48008 [2:17:58<26:01, 5.02it/s] 84%|████████▎ | 40177/48008 [2:17:59<26:08, 4.99it/s] 84%|████████▎ | 40178/48008 [2:17:59<25:39, 5.09it/s] 84%|████████▎ | 40179/48008 [2:17:59<25:17, 5.16it/s] 84%|████████▎ | 40180/48008 [2:17:59<25:03, 5.21it/s] 84%|████████▎ | 40181/48008 [2:17:59<25:01, 5.21it/s] 84%|████████▎ | 40182/48008 [2:18:00<25:14, 5.17it/s] 84%|████████▎ | 40183/48008 [2:18:00<25:09, 5.18it/s] 84%|████████▎ | 40184/48008 [2:18:00<25:00, 5.22it/s] 84%|████████▎ | 40185/48008 [2:18:00<24:57, 5.22it/s] 84%|████████▎ | 40186/48008 [2:18:00<24:53, 5.24it/s] 84%|████████▎ | 40187/48008 [2:18:01<25:04, 5.20it/s] 84%|████████▎ | 40188/48008 [2:18:01<24:56, 5.22it/s] 84%|████████▎ | 40189/48008 [2:18:01<25:01, 5.21it/s] 84%|████████▎ | 40190/48008 [2:18:01<24:59, 5.21it/s] 84%|████████▎ | 40191/48008 [2:18:01<24:50, 5.24it/s] 84%|████████▎ | 40192/48008 [2:18:02<24:50, 5.25it/s] 84%|████████▎ | 40193/48008 [2:18:02<24:49, 5.25it/s] 84%|████████▎ | 40194/48008 [2:18:02<25:08, 5.18it/s] 84%|████████▎ | 40195/48008 [2:18:02<29:33, 4.40it/s] 84%|████████▎ | 40196/48008 [2:18:02<28:09, 4.62it/s] 84%|████████▎ | 40197/48008 [2:18:03<28:56, 4.50it/s] 84%|████████▎ | 40198/48008 [2:18:03<27:35, 4.72it/s] 84%|████████▎ | 40199/48008 [2:18:03<26:46, 4.86it/s] 84%|████████▎ | 40200/48008 [2:18:03<26:28, 4.92it/s] {'loss': 0.3, 'grad_norm': 0.6171875, 'learning_rate': 3.253207798700217e-05, 'epoch': 0.84} 84%|████████▎ | 40200/48008 [2:18:03<26:28, 4.92it/s] 84%|████████▎ | 40201/48008 [2:18:03<25:53, 5.03it/s] 84%|████████▎ | 40202/48008 [2:18:04<25:34, 5.09it/s] 84%|████████▎ | 40203/48008 [2:18:04<26:13, 4.96it/s] 84%|████████▎ | 40204/48008 [2:18:04<25:52, 5.03it/s] 84%|████████▎ | 40205/48008 [2:18:04<25:20, 5.13it/s] 84%|████████▎ | 40206/48008 [2:18:04<25:12, 5.16it/s] 84%|████████▍ | 40207/48008 [2:18:05<25:11, 5.16it/s] 84%|████████▍ | 40208/48008 [2:18:05<24:53, 5.22it/s] 84%|████████▍ | 40209/48008 [2:18:05<25:08, 5.17it/s] 84%|████████▍ | 40210/48008 [2:18:05<29:33, 4.40it/s] 84%|████████▍ | 40211/48008 [2:18:05<28:16, 4.60it/s] 84%|████████▍ | 40212/48008 [2:18:06<27:16, 4.76it/s] 84%|████████▍ | 40213/48008 [2:18:06<26:32, 4.89it/s] 84%|████████▍ | 40214/48008 [2:18:06<26:14, 4.95it/s] 84%|████████▍ | 40215/48008 [2:18:06<25:53, 5.02it/s] 84%|████████▍ | 40216/48008 [2:18:06<25:29, 5.09it/s] 84%|████████▍ | 40217/48008 [2:18:07<25:27, 5.10it/s] 84%|████████▍ | 40218/48008 [2:18:07<25:57, 5.00it/s] 84%|████████▍ | 40219/48008 [2:18:07<25:59, 5.00it/s] 84%|████████▍ | 40220/48008 [2:18:07<25:54, 5.01it/s] 84%|████████▍ | 40221/48008 [2:18:07<25:54, 5.01it/s] 84%|████████▍ | 40222/48008 [2:18:08<25:31, 5.08it/s] 84%|████████▍ | 40223/48008 [2:18:08<25:19, 5.12it/s] 84%|████████▍ | 40224/48008 [2:18:08<25:26, 5.10it/s] 84%|████████▍ | 40225/48008 [2:18:08<25:08, 5.16it/s] 84%|████████▍ | 40226/48008 [2:18:08<25:00, 5.19it/s] 84%|████████▍ | 40227/48008 [2:18:09<24:54, 5.20it/s] 84%|████████▍ | 40228/48008 [2:18:09<24:53, 5.21it/s] 84%|████████▍ | 40229/48008 [2:18:09<24:46, 5.23it/s] 84%|████████▍ | 40230/48008 [2:18:09<24:41, 5.25it/s] 84%|████████▍ | 40231/48008 [2:18:09<24:57, 5.19it/s] 84%|████████▍ | 40232/48008 [2:18:10<24:42, 5.25it/s] 84%|████████▍ | 40233/48008 [2:18:10<25:28, 5.09it/s] 84%|████████▍ | 40234/48008 [2:18:10<25:11, 5.14it/s] 84%|████████▍ | 40235/48008 [2:18:10<24:49, 5.22it/s] 84%|████████▍ | 40236/48008 [2:18:10<25:03, 5.17it/s] 84%|████████▍ | 40237/48008 [2:18:11<25:27, 5.09it/s] 84%|████████▍ | 40238/48008 [2:18:11<25:08, 5.15it/s] 84%|████████▍ | 40239/48008 [2:18:11<24:52, 5.20it/s] 84%|████████▍ | 40240/48008 [2:18:11<24:41, 5.24it/s] 84%|████████▍ | 40241/48008 [2:18:11<24:47, 5.22it/s] 84%|████████▍ | 40242/48008 [2:18:12<32:21, 4.00it/s] 84%|████████▍ | 40243/48008 [2:18:12<30:01, 4.31it/s] 84%|████████▍ | 40244/48008 [2:18:12<28:10, 4.59it/s] 84%|████████▍ | 40245/48008 [2:18:12<28:54, 4.47it/s] 84%|████████▍ | 40246/48008 [2:18:12<27:40, 4.67it/s] 84%|████████▍ | 40247/48008 [2:18:13<27:04, 4.78it/s] 84%|████████▍ | 40248/48008 [2:18:13<26:19, 4.91it/s] 84%|████████▍ | 40249/48008 [2:18:13<25:40, 5.04it/s] 84%|████████▍ | 40250/48008 [2:18:13<25:16, 5.11it/s] {'loss': 0.2988, 'grad_norm': 0.2158203125, 'learning_rate': 3.232377937010498e-05, 'epoch': 0.84} 84%|████████▍ | 40250/48008 [2:18:13<25:16, 5.11it/s] 84%|████████▍ | 40251/48008 [2:18:13<24:57, 5.18it/s] 84%|████████▍ | 40252/48008 [2:18:14<24:46, 5.22it/s] 84%|████████▍ | 40253/48008 [2:18:14<26:30, 4.87it/s] 84%|████████▍ | 40254/48008 [2:18:14<30:23, 4.25it/s] 84%|████████▍ | 40255/48008 [2:18:14<28:25, 4.54it/s] 84%|████████▍ | 40256/48008 [2:18:15<27:18, 4.73it/s] 84%|████████▍ | 40257/48008 [2:18:15<26:25, 4.89it/s] 84%|████████▍ | 40258/48008 [2:18:15<26:37, 4.85it/s] 84%|████████▍ | 40259/48008 [2:18:15<26:02, 4.96it/s] 84%|████████▍ | 40260/48008 [2:18:15<25:57, 4.98it/s] 84%|████████▍ | 40261/48008 [2:18:16<25:41, 5.03it/s] 84%|████████▍ | 40262/48008 [2:18:16<25:14, 5.11it/s] 84%|████████▍ | 40263/48008 [2:18:16<24:58, 5.17it/s] 84%|████████▍ | 40264/48008 [2:18:16<24:49, 5.20it/s] 84%|████████▍ | 40265/48008 [2:18:16<25:01, 5.16it/s] 84%|████████▍ | 40266/48008 [2:18:16<24:38, 5.24it/s] 84%|████████▍ | 40267/48008 [2:18:17<24:41, 5.22it/s] 84%|████████▍ | 40268/48008 [2:18:17<24:56, 5.17it/s] 84%|████████▍ | 40269/48008 [2:18:17<24:53, 5.18it/s] 84%|████████▍ | 40270/48008 [2:18:17<24:38, 5.23it/s] 84%|████████▍ | 40271/48008 [2:18:17<24:37, 5.24it/s] 84%|████████▍ | 40272/48008 [2:18:18<24:32, 5.25it/s] 84%|████████▍ | 40273/48008 [2:18:18<24:31, 5.26it/s] 84%|████████▍ | 40274/48008 [2:18:18<24:37, 5.23it/s] 84%|████████▍ | 40275/48008 [2:18:18<24:26, 5.27it/s] 84%|████████▍ | 40276/48008 [2:18:18<24:43, 5.21it/s] 84%|████████▍ | 40277/48008 [2:18:19<32:17, 3.99it/s] 84%|████████▍ | 40278/48008 [2:18:19<29:56, 4.30it/s] 84%|████████▍ | 40279/48008 [2:18:19<28:11, 4.57it/s] 84%|████████▍ | 40280/48008 [2:18:19<27:07, 4.75it/s] 84%|████████▍ | 40281/48008 [2:18:20<26:16, 4.90it/s] 84%|████████▍ | 40282/48008 [2:18:20<27:32, 4.68it/s] 84%|████████▍ | 40283/48008 [2:18:20<26:40, 4.83it/s] 84%|████████▍ | 40284/48008 [2:18:20<26:00, 4.95it/s] 84%|████████▍ | 40285/48008 [2:18:20<25:30, 5.05it/s] 84%|████████▍ | 40286/48008 [2:18:21<25:20, 5.08it/s] 84%|████████▍ | 40287/48008 [2:18:21<25:17, 5.09it/s] 84%|████████▍ | 40288/48008 [2:18:21<25:05, 5.13it/s] 84%|████████▍ | 40289/48008 [2:18:21<24:57, 5.15it/s] 84%|████████▍ | 40290/48008 [2:18:21<24:58, 5.15it/s] 84%|████████▍ | 40291/48008 [2:18:21<25:02, 5.14it/s] 84%|████████▍ | 40292/48008 [2:18:22<29:22, 4.38it/s] 84%|████████▍ | 40293/48008 [2:18:22<27:48, 4.62it/s] 84%|████████▍ | 40294/48008 [2:18:22<34:22, 3.74it/s] 84%|████████▍ | 40295/48008 [2:18:23<31:56, 4.02it/s] 84%|████████▍ | 40296/48008 [2:18:23<30:14, 4.25it/s] 84%|████████▍ | 40297/48008 [2:18:23<28:55, 4.44it/s] 84%|████████▍ | 40298/48008 [2:18:23<27:40, 4.64it/s] 84%|████████▍ | 40299/48008 [2:18:23<27:17, 4.71it/s] 84%|████████▍ | 40300/48008 [2:18:24<26:29, 4.85it/s]{'loss': 0.2987, 'grad_norm': 0.267578125, 'learning_rate': 3.21154807532078e-05, 'epoch': 0.84} 84%|████████▍ | 40300/48008 [2:18:24<26:29, 4.85it/s] 84%|████████▍ | 40301/48008 [2:18:24<26:24, 4.86it/s] 84%|████████▍ | 40302/48008 [2:18:24<25:48, 4.98it/s] 84%|████████▍ | 40303/48008 [2:18:24<25:53, 4.96it/s] 84%|████████▍ | 40304/48008 [2:18:24<25:53, 4.96it/s] 84%|████████▍ | 40305/48008 [2:18:25<25:29, 5.04it/s] 84%|████████▍ | 40306/48008 [2:18:25<25:17, 5.08it/s] 84%|████████▍ | 40307/48008 [2:18:25<25:14, 5.09it/s] 84%|████████▍ | 40308/48008 [2:18:25<25:41, 5.00it/s] 84%|████████▍ | 40309/48008 [2:18:25<25:11, 5.09it/s] 84%|████████▍ | 40310/48008 [2:18:26<25:00, 5.13it/s] 84%|████████▍ | 40311/48008 [2:18:26<25:19, 5.07it/s] 84%|████████▍ | 40312/48008 [2:18:26<25:18, 5.07it/s] 84%|████████▍ | 40313/48008 [2:18:26<24:58, 5.14it/s] 84%|████████▍ | 40314/48008 [2:18:26<25:18, 5.07it/s] 84%|████████▍ | 40315/48008 [2:18:27<25:15, 5.08it/s] 84%|████████▍ | 40316/48008 [2:18:27<24:56, 5.14it/s] 84%|████████▍ | 40317/48008 [2:18:27<24:49, 5.16it/s] 84%|████████▍ | 40318/48008 [2:18:27<25:00, 5.13it/s] 84%|████████▍ | 40319/48008 [2:18:27<24:41, 5.19it/s] 84%|████████▍ | 40320/48008 [2:18:27<24:41, 5.19it/s] 84%|████████▍ | 40321/48008 [2:18:28<24:33, 5.22it/s] 84%|████████▍ | 40322/48008 [2:18:29<1:03:48, 2.01it/s] 84%|████████▍ | 40323/48008 [2:18:29<51:52, 2.47it/s] 84%|████████▍ | 40324/48008 [2:18:29<44:22, 2.89it/s] 84%|████████▍ | 40325/48008 [2:18:29<38:22, 3.34it/s] 84%|████████▍ | 40326/48008 [2:18:30<34:09, 3.75it/s] 84%|████████▍ | 40327/48008 [2:18:30<31:08, 4.11it/s] 84%|████████▍ | 40328/48008 [2:18:30<29:03, 4.40it/s] 84%|████████▍ | 40329/48008 [2:18:30<27:53, 4.59it/s] 84%|████████▍ | 40330/48008 [2:18:31<34:25, 3.72it/s] 84%|████████▍ | 40331/48008 [2:18:31<31:17, 4.09it/s] 84%|████████▍ | 40332/48008 [2:18:31<29:06, 4.40it/s] 84%|████████▍ | 40333/48008 [2:18:31<35:11, 3.63it/s] 84%|████████▍ | 40334/48008 [2:18:32<31:48, 4.02it/s] 84%|████████▍ | 40335/48008 [2:18:32<29:27, 4.34it/s] 84%|████████▍ | 40336/48008 [2:18:32<27:53, 4.59it/s] 84%|████████▍ | 40337/48008 [2:18:32<26:57, 4.74it/s] 84%|████████▍ | 40338/48008 [2:18:32<26:24, 4.84it/s] 84%|████████▍ | 40339/48008 [2:18:33<25:29, 5.01it/s] 84%|████████▍ | 40340/48008 [2:18:33<25:11, 5.07it/s] 84%|████████▍ | 40341/48008 [2:18:33<24:55, 5.13it/s] 84%|████████▍ | 40342/48008 [2:18:33<25:12, 5.07it/s] 84%|████████▍ | 40343/48008 [2:18:33<24:58, 5.12it/s] 84%|████████▍ | 40344/48008 [2:18:34<25:30, 5.01it/s] 84%|████████▍ | 40345/48008 [2:18:34<24:53, 5.13it/s] 84%|████████▍ | 40346/48008 [2:18:34<32:12, 3.96it/s] 84%|████████▍ | 40347/48008 [2:18:34<31:39, 4.03it/s] 84%|████████▍ | 40348/48008 [2:18:35<29:26, 4.34it/s] 84%|████████▍ | 40349/48008 [2:18:35<28:23, 4.50it/s] 84%|████████▍ | 40350/48008 [2:18:35<27:25, 4.65it/s] 84%|████████▍ | 40350/48008 [2:18:35<27:25, 4.65it/s]{'loss': 0.3001, 'grad_norm': 0.23828125, 'learning_rate': 3.1907182136310616e-05, 'epoch': 0.84} 84%|████████▍ | 40351/48008 [2:18:35<28:24, 4.49it/s] 84%|████████▍ | 40352/48008 [2:18:35<26:59, 4.73it/s] 84%|████████▍ | 40353/48008 [2:18:36<26:17, 4.85it/s] 84%|████████▍ | 40354/48008 [2:18:36<25:34, 4.99it/s] 84%|████████▍ | 40355/48008 [2:18:36<25:23, 5.02it/s] 84%|████████▍ | 40356/48008 [2:18:36<24:57, 5.11it/s] 84%|████████▍ | 40357/48008 [2:18:36<24:43, 5.16it/s] 84%|████████▍ | 40358/48008 [2:18:36<24:20, 5.24it/s] 84%|████████▍ | 40359/48008 [2:18:37<24:24, 5.22it/s] 84%|████████▍ | 40360/48008 [2:18:37<24:26, 5.21it/s] 84%|████████▍ | 40361/48008 [2:18:37<24:18, 5.24it/s] 84%|████████▍ | 40362/48008 [2:18:37<24:23, 5.22it/s] 84%|████████▍ | 40363/48008 [2:18:37<24:19, 5.24it/s] 84%|████████▍ | 40364/48008 [2:18:38<24:22, 5.23it/s] 84%|████████▍ | 40365/48008 [2:18:38<24:24, 5.22it/s] 84%|████████▍ | 40366/48008 [2:18:38<25:13, 5.05it/s] 84%|████████▍ | 40367/48008 [2:18:38<25:09, 5.06it/s] 84%|████████▍ | 40368/48008 [2:18:39<32:26, 3.93it/s] 84%|████████▍ | 40369/48008 [2:18:39<30:00, 4.24it/s] 84%|████████▍ | 40370/48008 [2:18:39<28:41, 4.44it/s] 84%|████████▍ | 40371/48008 [2:18:39<27:12, 4.68it/s] 84%|████████▍ | 40372/48008 [2:18:39<26:06, 4.87it/s] 84%|████████▍ | 40373/48008 [2:18:40<25:30, 4.99it/s] 84%|████████▍ | 40374/48008 [2:18:40<25:04, 5.07it/s] 84%|████████▍ | 40375/48008 [2:18:40<24:56, 5.10it/s] 84%|████████▍ | 40376/48008 [2:18:40<24:43, 5.15it/s] 84%|████████▍ | 40377/48008 [2:18:40<25:12, 5.04it/s] 84%|████████▍ | 40378/48008 [2:18:41<24:59, 5.09it/s] 84%|████████▍ | 40379/48008 [2:18:41<25:09, 5.06it/s] 84%|████████▍ | 40380/48008 [2:18:41<25:02, 5.08it/s] 84%|████████▍ | 40381/48008 [2:18:41<24:47, 5.13it/s] 84%|████████▍ | 40382/48008 [2:18:41<24:41, 5.15it/s] 84%|████████▍ | 40383/48008 [2:18:42<24:32, 5.18it/s] 84%|████████▍ | 40384/48008 [2:18:42<24:35, 5.17it/s] 84%|████████▍ | 40385/48008 [2:18:42<25:02, 5.07it/s] 84%|████████▍ | 40386/48008 [2:18:42<24:29, 5.19it/s] 84%|████████▍ | 40387/48008 [2:18:42<24:27, 5.19it/s] 84%|████████▍ | 40388/48008 [2:18:42<24:26, 5.19it/s] 84%|████████▍ | 40389/48008 [2:18:43<24:25, 5.20it/s] 84%|████████▍ | 40390/48008 [2:18:43<24:42, 5.14it/s] 84%|████████▍ | 40391/48008 [2:18:43<26:25, 4.80it/s] 84%|████████▍ | 40392/48008 [2:18:43<25:50, 4.91it/s] 84%|████████▍ | 40393/48008 [2:18:43<25:13, 5.03it/s] 84%|████████▍ | 40394/48008 [2:18:44<25:07, 5.05it/s] 84%|████████▍ | 40395/48008 [2:18:44<24:56, 5.09it/s] 84%|████████▍ | 40396/48008 [2:18:44<24:37, 5.15it/s] 84%|████████▍ | 40397/48008 [2:18:44<24:23, 5.20it/s] 84%|████████▍ | 40398/48008 [2:18:44<24:22, 5.20it/s] 84%|████████▍ | 40399/48008 [2:18:45<24:51, 5.10it/s] 84%|████████▍ | 40400/48008 [2:18:45<25:09, 5.04it/s] {'loss': 0.3013, 'grad_norm': 0.2373046875, 'learning_rate': 3.169888351941343e-05, 'epoch': 0.84} 84%|████████▍ | 40400/48008 [2:18:45<25:09, 5.04it/s] 84%|████████▍ | 40401/48008 [2:18:45<24:58, 5.08it/s] 84%|████████▍ | 40402/48008 [2:18:45<24:41, 5.13it/s] 84%|████████▍ | 40403/48008 [2:18:45<24:42, 5.13it/s] 84%|████████▍ | 40404/48008 [2:18:46<24:32, 5.16it/s] 84%|████████▍ | 40405/48008 [2:18:46<24:11, 5.24it/s] 84%|████████▍ | 40406/48008 [2:18:46<24:21, 5.20it/s] 84%|████████▍ | 40407/48008 [2:18:46<24:45, 5.12it/s] 84%|████████▍ | 40408/48008 [2:18:46<24:54, 5.08it/s] 84%|████████▍ | 40409/48008 [2:18:47<24:50, 5.10it/s] 84%|████████▍ | 40410/48008 [2:18:47<25:10, 5.03it/s] 84%|████████▍ | 40411/48008 [2:18:47<24:49, 5.10it/s] 84%|████████▍ | 40412/48008 [2:18:47<24:53, 5.09it/s] 84%|████████▍ | 40413/48008 [2:18:48<29:03, 4.36it/s] 84%|████████▍ | 40414/48008 [2:18:48<27:35, 4.59it/s] 84%|████████▍ | 40415/48008 [2:18:48<34:01, 3.72it/s] 84%|████████▍ | 40416/48008 [2:18:48<31:16, 4.05it/s] 84%|████████▍ | 40417/48008 [2:18:48<29:33, 4.28it/s] 84%|████████▍ | 40418/48008 [2:18:49<27:41, 4.57it/s] 84%|████████▍ | 40419/48008 [2:18:49<26:47, 4.72it/s] 84%|████████▍ | 40420/48008 [2:18:49<26:01, 4.86it/s] 84%|████████▍ | 40421/48008 [2:18:49<26:16, 4.81it/s] 84%|████████▍ | 40422/48008 [2:18:49<25:39, 4.93it/s] 84%|████████▍ | 40423/48008 [2:18:50<25:13, 5.01it/s] 84%|████████▍ | 40424/48008 [2:18:50<25:16, 5.00it/s] 84%|████████▍ | 40425/48008 [2:18:50<24:44, 5.11it/s] 84%|████████▍ | 40426/48008 [2:18:50<24:58, 5.06it/s] 84%|████████▍ | 40427/48008 [2:18:51<32:05, 3.94it/s] 84%|████████▍ | 40428/48008 [2:18:51<29:31, 4.28it/s] 84%|████████▍ | 40429/48008 [2:18:51<28:03, 4.50it/s] 84%|████████▍ | 40430/48008 [2:18:51<27:04, 4.66it/s] 84%|████████▍ | 40431/48008 [2:18:51<26:37, 4.74it/s] 84%|████████▍ | 40432/48008 [2:18:52<26:06, 4.84it/s] 84%|████████▍ | 40433/48008 [2:18:52<26:14, 4.81it/s] 84%|████████▍ | 40434/48008 [2:18:52<25:32, 4.94it/s] 84%|████████▍ | 40435/48008 [2:18:52<25:02, 5.04it/s] 84%|████████▍ | 40436/48008 [2:18:52<24:46, 5.09it/s] 84%|████████▍ | 40437/48008 [2:18:53<24:30, 5.15it/s] 84%|████████▍ | 40438/48008 [2:18:53<24:53, 5.07it/s] 84%|████████▍ | 40439/48008 [2:18:53<24:40, 5.11it/s] 84%|████████▍ | 40440/48008 [2:18:53<24:43, 5.10it/s] 84%|████████▍ | 40441/48008 [2:18:53<24:33, 5.13it/s] 84%|████████▍ | 40442/48008 [2:18:54<24:26, 5.16it/s] 84%|████████▍ | 40443/48008 [2:18:54<24:11, 5.21it/s] 84%|████████▍ | 40444/48008 [2:18:54<24:13, 5.20it/s] 84%|████████▍ | 40445/48008 [2:18:54<24:09, 5.22it/s] 84%|████████▍ | 40446/48008 [2:18:54<24:03, 5.24it/s] 84%|████████▍ | 40447/48008 [2:18:55<24:07, 5.22it/s] 84%|████████▍ | 40448/48008 [2:18:55<23:56, 5.26it/s] 84%|████████▍ | 40449/48008 [2:18:55<23:52, 5.28it/s] 84%|████████▍ | 40450/48008 [2:18:55<23:56, 5.26it/s] {'loss': 0.2997, 'grad_norm': 0.283203125, 'learning_rate': 3.149058490251625e-05, 'epoch': 0.84} 84%|████████▍ | 40450/48008 [2:18:55<23:56, 5.26it/s] 84%|████████▍ | 40451/48008 [2:18:55<24:17, 5.18it/s] 84%|████████▍ | 40452/48008 [2:18:55<24:03, 5.23it/s] 84%|████████▍ | 40453/48008 [2:18:56<24:01, 5.24it/s] 84%|████████▍ | 40454/48008 [2:18:56<31:28, 4.00it/s] 84%|████████▍ | 40455/48008 [2:18:56<29:17, 4.30it/s] 84%|████████▍ | 40456/48008 [2:18:56<28:22, 4.43it/s] 84%|████████▍ | 40457/48008 [2:18:57<27:32, 4.57it/s] 84%|████████▍ | 40458/48008 [2:18:57<26:19, 4.78it/s] 84%|████████▍ | 40459/48008 [2:18:57<26:05, 4.82it/s] 84%|████████▍ | 40460/48008 [2:18:57<25:24, 4.95it/s] 84%|████████▍ | 40461/48008 [2:18:57<24:56, 5.04it/s] 84%|████████▍ | 40462/48008 [2:18:58<24:31, 5.13it/s] 84%|████████▍ | 40463/48008 [2:18:58<24:21, 5.16it/s] 84%|████████▍ | 40464/48008 [2:18:58<24:10, 5.20it/s] 84%|████████▍ | 40465/48008 [2:18:58<23:59, 5.24it/s] 84%|████████▍ | 40466/48008 [2:18:58<23:58, 5.24it/s] 84%|████████▍ | 40467/48008 [2:18:59<24:24, 5.15it/s] 84%|████████▍ | 40468/48008 [2:18:59<24:06, 5.21it/s] 84%|████████▍ | 40469/48008 [2:18:59<24:07, 5.21it/s] 84%|████████▍ | 40470/48008 [2:18:59<24:07, 5.21it/s] 84%|████████▍ | 40471/48008 [2:18:59<24:02, 5.22it/s] 84%|████████▍ | 40472/48008 [2:19:00<31:24, 4.00it/s] 84%|████████▍ | 40473/48008 [2:19:00<29:16, 4.29it/s] 84%|████████▍ | 40474/48008 [2:19:00<28:09, 4.46it/s] 84%|████████▍ | 40475/48008 [2:19:00<26:51, 4.68it/s] 84%|████████▍ | 40476/48008 [2:19:00<26:00, 4.83it/s] 84%|████████▍ | 40477/48008 [2:19:01<25:21, 4.95it/s] 84%|████████▍ | 40478/48008 [2:19:01<24:55, 5.04it/s] 84%|████████▍ | 40479/48008 [2:19:01<24:37, 5.09it/s] 84%|████████▍ | 40480/48008 [2:19:01<24:20, 5.15it/s] 84%|████████▍ | 40481/48008 [2:19:01<24:30, 5.12it/s] 84%|████████▍ | 40482/48008 [2:19:02<24:29, 5.12it/s] 84%|████████▍ | 40483/48008 [2:19:02<26:06, 4.80it/s] 84%|████████▍ | 40484/48008 [2:19:02<25:36, 4.90it/s] 84%|████████▍ | 40485/48008 [2:19:02<25:34, 4.90it/s] 84%|████████▍ | 40486/48008 [2:19:02<25:23, 4.94it/s] 84%|████████▍ | 40487/48008 [2:19:03<25:01, 5.01it/s] 84%|████████▍ | 40488/48008 [2:19:03<24:51, 5.04it/s] 84%|████████▍ | 40489/48008 [2:19:03<24:22, 5.14it/s] 84%|████████▍ | 40490/48008 [2:19:03<24:28, 5.12it/s] 84%|████████▍ | 40491/48008 [2:19:03<24:20, 5.15it/s] 84%|████████▍ | 40492/48008 [2:19:04<24:18, 5.15it/s] 84%|████████▍ | 40493/48008 [2:19:04<24:37, 5.08it/s] 84%|████████▍ | 40494/48008 [2:19:04<24:32, 5.10it/s] 84%|████████▍ | 40495/48008 [2:19:04<24:24, 5.13it/s] 84%|████████▍ | 40496/48008 [2:19:04<24:59, 5.01it/s] 84%|████████▍ | 40497/48008 [2:19:05<29:02, 4.31it/s] 84%|████████▍ | 40498/48008 [2:19:05<27:27, 4.56it/s] 84%|████████▍ | 40499/48008 [2:19:05<26:09, 4.78it/s] 84%|████████▍ | 40500/48008 [2:19:05<25:35, 4.89it/s] {'loss': 0.3022, 'grad_norm': 0.57421875, 'learning_rate': 3.128228628561907e-05, 'epoch': 0.84} 84%|████████▍ | 40500/48008 [2:19:05<25:35, 4.89it/s] 84%|████████▍ | 40501/48008 [2:19:06<26:51, 4.66it/s] 84%|████████▍ | 40502/48008 [2:19:06<25:58, 4.82it/s] 84%|████████▍ | 40503/48008 [2:19:06<24:58, 5.01it/s] 84%|████████▍ | 40504/48008 [2:19:06<25:11, 4.96it/s] 84%|████████▍ | 40505/48008 [2:19:06<24:42, 5.06it/s] 84%|████████▍ | 40506/48008 [2:19:06<24:26, 5.12it/s] 84%|████████▍ | 40507/48008 [2:19:07<24:03, 5.20it/s] 84%|████████▍ | 40508/48008 [2:19:07<24:21, 5.13it/s] 84%|████████▍ | 40509/48008 [2:19:07<24:07, 5.18it/s] 84%|████████▍ | 40510/48008 [2:19:07<23:58, 5.21it/s] 84%|████████▍ | 40511/48008 [2:19:07<23:53, 5.23it/s] 84%|████████▍ | 40512/48008 [2:19:08<24:10, 5.17it/s] 84%|████████▍ | 40513/48008 [2:19:08<24:05, 5.18it/s] 84%|████████▍ | 40514/48008 [2:19:08<24:19, 5.13it/s] 84%|████████▍ | 40515/48008 [2:19:08<24:36, 5.07it/s] 84%|████████▍ | 40516/48008 [2:19:08<24:23, 5.12it/s] 84%|████████▍ | 40517/48008 [2:19:09<24:19, 5.13it/s] 84%|████████▍ | 40518/48008 [2:19:09<24:24, 5.11it/s] 84%|████████▍ | 40519/48008 [2:19:09<24:16, 5.14it/s] 84%|████████▍ | 40520/48008 [2:19:09<24:01, 5.19it/s] 84%|████████▍ | 40521/48008 [2:19:09<23:31, 5.30it/s] 84%|████████▍ | 40522/48008 [2:19:10<23:32, 5.30it/s] 84%|████████▍ | 40523/48008 [2:19:10<24:01, 5.19it/s] 84%|████████▍ | 40524/48008 [2:19:10<23:45, 5.25it/s] 84%|████████▍ | 40525/48008 [2:19:10<23:47, 5.24it/s] 84%|████████▍ | 40526/48008 [2:19:10<23:55, 5.21it/s] 84%|████████▍ | 40527/48008 [2:19:11<23:58, 5.20it/s] 84%|████████▍ | 40528/48008 [2:19:11<23:53, 5.22it/s] 84%|████████▍ | 40529/48008 [2:19:11<23:47, 5.24it/s] 84%|████████▍ | 40530/48008 [2:19:11<24:06, 5.17it/s] 84%|████████▍ | 40531/48008 [2:19:11<23:48, 5.24it/s] 84%|████████▍ | 40532/48008 [2:19:11<23:51, 5.22it/s] 84%|████████▍ | 40533/48008 [2:19:12<23:49, 5.23it/s] 84%|████████▍ | 40534/48008 [2:19:12<24:19, 5.12it/s] 84%|████████▍ | 40535/48008 [2:19:12<24:13, 5.14it/s] 84%|████████▍ | 40536/48008 [2:19:12<24:08, 5.16it/s] 84%|████████▍ | 40537/48008 [2:19:12<24:21, 5.11it/s] 84%|████████▍ | 40538/48008 [2:19:13<24:19, 5.12it/s] 84%|████████▍ | 40539/48008 [2:19:13<24:05, 5.17it/s] 84%|████████▍ | 40540/48008 [2:19:13<23:41, 5.25it/s] 84%|████████▍ | 40541/48008 [2:19:13<23:41, 5.25it/s] 84%|████████▍ | 40542/48008 [2:19:13<23:48, 5.23it/s] 84%|████████▍ | 40543/48008 [2:19:14<24:12, 5.14it/s] 84%|████████▍ | 40544/48008 [2:19:14<24:18, 5.12it/s] 84%|████████▍ | 40545/48008 [2:19:14<24:36, 5.06it/s] 84%|████████▍ | 40546/48008 [2:19:14<24:27, 5.09it/s] 84%|████████▍ | 40547/48008 [2:19:14<24:20, 5.11it/s] 84%|████████▍ | 40548/48008 [2:19:15<28:31, 4.36it/s] 84%|████████▍ | 40549/48008 [2:19:15<27:07, 4.58it/s] 84%|████████▍ | 40550/48008 [2:19:15<26:22, 4.71it/s] {'loss': 0.3019, 'grad_norm': 0.26171875, 'learning_rate': 3.1073987668721885e-05, 'epoch': 0.84} 84%|████████▍ | 40550/48008 [2:19:15<26:22, 4.71it/s] 84%|████████▍ | 40551/48008 [2:19:15<25:35, 4.86it/s] 84%|████████▍ | 40552/48008 [2:19:16<25:17, 4.91it/s] 84%|████████▍ | 40553/48008 [2:19:16<24:56, 4.98it/s] 84%|████████▍ | 40554/48008 [2:19:16<24:24, 5.09it/s] 84%|████████▍ | 40555/48008 [2:19:16<23:56, 5.19it/s] 84%|████████▍ | 40556/48008 [2:19:16<24:16, 5.12it/s] 84%|████████▍ | 40557/48008 [2:19:16<24:36, 5.05it/s] 84%|████████▍ | 40558/48008 [2:19:17<24:09, 5.14it/s] 84%|████████▍ | 40559/48008 [2:19:17<24:22, 5.09it/s] 84%|████████▍ | 40560/48008 [2:19:17<24:11, 5.13it/s] 84%|████████▍ | 40561/48008 [2:19:17<23:54, 5.19it/s] 84%|████████▍ | 40562/48008 [2:19:17<23:47, 5.21it/s] 84%|████████▍ | 40563/48008 [2:19:18<23:34, 5.26it/s] 84%|████████▍ | 40564/48008 [2:19:18<24:03, 5.16it/s] 84%|████████▍ | 40565/48008 [2:19:18<23:57, 5.18it/s] 84%|████████▍ | 40566/48008 [2:19:18<23:47, 5.21it/s] 85%|████████▍ | 40567/48008 [2:19:18<25:34, 4.85it/s] 85%|████████▍ | 40568/48008 [2:19:19<25:02, 4.95it/s] 85%|████████▍ | 40569/48008 [2:19:19<24:40, 5.03it/s] 85%|████████▍ | 40570/48008 [2:19:19<24:25, 5.08it/s] 85%|████████▍ | 40571/48008 [2:19:19<24:22, 5.08it/s] 85%|████████▍ | 40572/48008 [2:19:19<23:53, 5.19it/s] 85%|████████▍ | 40573/48008 [2:19:20<25:35, 4.84it/s] 85%|████████▍ | 40574/48008 [2:19:20<25:23, 4.88it/s] 85%|████████▍ | 40575/48008 [2:19:20<24:53, 4.98it/s] 85%|████████▍ | 40576/48008 [2:19:20<24:56, 4.97it/s] 85%|████████▍ | 40577/48008 [2:19:20<24:27, 5.06it/s] 85%|████████▍ | 40578/48008 [2:19:21<24:06, 5.14it/s] 85%|████████▍ | 40579/48008 [2:19:21<28:14, 4.38it/s] 85%|████████▍ | 40580/48008 [2:19:21<26:53, 4.60it/s] 85%|████████▍ | 40581/48008 [2:19:21<25:52, 4.78it/s] 85%|████████▍ | 40582/48008 [2:19:22<25:49, 4.79it/s] 85%|████████▍ | 40583/48008 [2:19:22<25:12, 4.91it/s] 85%|████████▍ | 40584/48008 [2:19:22<24:44, 5.00it/s] 85%|████████▍ | 40585/48008 [2:19:22<25:06, 4.93it/s] 85%|████████▍ | 40586/48008 [2:19:22<24:42, 5.01it/s] 85%|████████▍ | 40587/48008 [2:19:22<24:50, 4.98it/s] 85%|████████▍ | 40588/48008 [2:19:23<24:19, 5.08it/s] 85%|████████▍ | 40589/48008 [2:19:23<24:06, 5.13it/s] 85%|████████▍ | 40590/48008 [2:19:23<23:48, 5.19it/s] 85%|████████▍ | 40591/48008 [2:19:23<23:45, 5.20it/s] 85%|████████▍ | 40592/48008 [2:19:23<23:52, 5.18it/s] 85%|████████▍ | 40593/48008 [2:19:24<25:31, 4.84it/s] 85%|████████▍ | 40594/48008 [2:19:24<25:10, 4.91it/s] 85%|████████▍ | 40595/48008 [2:19:24<24:40, 5.01it/s] 85%|████████▍ | 40596/48008 [2:19:24<24:22, 5.07it/s] 85%|████████▍ | 40597/48008 [2:19:24<24:10, 5.11it/s] 85%|████████▍ | 40598/48008 [2:19:25<25:41, 4.81it/s] 85%|████████▍ | 40599/48008 [2:19:25<25:03, 4.93it/s] 85%|████████▍ | 40600/48008 [2:19:25<24:25, 5.05it/s] {'loss': 0.3006, 'grad_norm': 0.1650390625, 'learning_rate': 3.0865689051824696e-05, 'epoch': 0.85} 85%|████████▍ | 40600/48008 [2:19:25<24:25, 5.05it/s] 85%|████████▍ | 40601/48008 [2:19:25<23:52, 5.17it/s] 85%|████████▍ | 40602/48008 [2:19:25<24:14, 5.09it/s] 85%|████████▍ | 40603/48008 [2:19:26<24:26, 5.05it/s] 85%|████████▍ | 40604/48008 [2:19:26<24:12, 5.10it/s] 85%|████████▍ | 40605/48008 [2:19:26<24:23, 5.06it/s] 85%|████████▍ | 40606/48008 [2:19:26<24:35, 5.02it/s] 85%|████████▍ | 40607/48008 [2:19:26<24:21, 5.06it/s] 85%|████████▍ | 40608/48008 [2:19:27<24:13, 5.09it/s] 85%|████████▍ | 40609/48008 [2:19:27<24:24, 5.05it/s] 85%|████████▍ | 40610/48008 [2:19:27<31:23, 3.93it/s] 85%|████████▍ | 40611/48008 [2:19:27<28:52, 4.27it/s] 85%|████████▍ | 40612/48008 [2:19:28<27:11, 4.53it/s] 85%|████████▍ | 40613/48008 [2:19:28<26:31, 4.65it/s] 85%|████████▍ | 40614/48008 [2:19:28<25:33, 4.82it/s] 85%|████████▍ | 40615/48008 [2:19:28<24:34, 5.01it/s] 85%|████████▍ | 40616/48008 [2:19:28<24:03, 5.12it/s] 85%|████████▍ | 40617/48008 [2:19:29<23:54, 5.15it/s] 85%|████████▍ | 40618/48008 [2:19:29<24:00, 5.13it/s] 85%|████████▍ | 40619/48008 [2:19:29<23:39, 5.21it/s] 85%|████████▍ | 40620/48008 [2:19:29<23:37, 5.21it/s] 85%|████████▍ | 40621/48008 [2:19:29<23:41, 5.20it/s] 85%|████████▍ | 40622/48008 [2:19:30<23:33, 5.23it/s] 85%|████████▍ | 40623/48008 [2:19:30<23:49, 5.16it/s] 85%|████████▍ | 40624/48008 [2:19:30<23:53, 5.15it/s] 85%|████████▍ | 40625/48008 [2:19:30<30:57, 3.98it/s] 85%|████████▍ | 40626/48008 [2:19:30<28:43, 4.28it/s] 85%|████████▍ | 40627/48008 [2:19:31<27:12, 4.52it/s] 85%|████████▍ | 40628/48008 [2:19:31<26:05, 4.71it/s] 85%|████████▍ | 40629/48008 [2:19:32<50:36, 2.43it/s] 85%|████████▍ | 40630/48008 [2:19:32<43:04, 2.85it/s] 85%|████████▍ | 40631/48008 [2:19:32<38:48, 3.17it/s] 85%|████████▍ | 40632/48008 [2:19:32<34:35, 3.55it/s] 85%|████████▍ | 40633/48008 [2:19:33<31:04, 3.95it/s] 85%|████████▍ | 40634/48008 [2:19:33<28:41, 4.28it/s] 85%|████████▍ | 40635/48008 [2:19:33<27:31, 4.47it/s] 85%|████████▍ | 40636/48008 [2:19:33<26:02, 4.72it/s] 85%|████████▍ | 40637/48008 [2:19:33<25:11, 4.88it/s] 85%|████████▍ | 40638/48008 [2:19:34<24:37, 4.99it/s] 85%|████████▍ | 40639/48008 [2:19:34<24:30, 5.01it/s] 85%|████████▍ | 40640/48008 [2:19:34<28:24, 4.32it/s] 85%|████████▍ | 40641/48008 [2:19:34<26:46, 4.59it/s] 85%|████████▍ | 40642/48008 [2:19:34<25:41, 4.78it/s] 85%|████████▍ | 40643/48008 [2:19:35<25:04, 4.89it/s] 85%|████████▍ | 40644/48008 [2:19:35<25:19, 4.85it/s] 85%|████████▍ | 40645/48008 [2:19:35<24:47, 4.95it/s] 85%|████████▍ | 40646/48008 [2:19:35<24:23, 5.03it/s] 85%|████████▍ | 40647/48008 [2:19:35<25:46, 4.76it/s] 85%|████████▍ | 40648/48008 [2:19:36<25:44, 4.77it/s] 85%|████████▍ | 40649/48008 [2:19:36<25:00, 4.90it/s] 85%|████████▍ | 40650/48008 [2:19:36<24:36, 4.98it/s] {'loss': 0.3005, 'grad_norm': 0.154296875, 'learning_rate': 3.065739043492751e-05, 'epoch': 0.85} 85%|████████▍ | 40650/48008 [2:19:36<24:36, 4.98it/s] 85%|████████▍ | 40651/48008 [2:19:36<24:13, 5.06it/s] 85%|████████▍ | 40652/48008 [2:19:36<23:41, 5.17it/s] 85%|████████▍ | 40653/48008 [2:19:37<24:18, 5.04it/s] 85%|████████▍ | 40654/48008 [2:19:37<24:11, 5.07it/s] 85%|████████▍ | 40655/48008 [2:19:37<24:11, 5.07it/s] 85%|████████▍ | 40656/48008 [2:19:37<24:05, 5.09it/s] 85%|████████▍ | 40657/48008 [2:19:37<23:38, 5.18it/s] 85%|████████▍ | 40658/48008 [2:19:38<23:29, 5.21it/s] 85%|████████▍ | 40659/48008 [2:19:38<23:23, 5.23it/s] 85%|████████▍ | 40660/48008 [2:19:38<23:26, 5.23it/s] 85%|████████▍ | 40661/48008 [2:19:38<23:24, 5.23it/s] 85%|████████▍ | 40662/48008 [2:19:38<23:41, 5.17it/s] 85%|████████▍ | 40663/48008 [2:19:39<24:02, 5.09it/s] 85%|████████▍ | 40664/48008 [2:19:39<24:05, 5.08it/s] 85%|████████▍ | 40665/48008 [2:19:39<23:46, 5.15it/s] 85%|████████▍ | 40666/48008 [2:19:39<23:34, 5.19it/s] 85%|████████▍ | 40667/48008 [2:19:39<27:42, 4.42it/s] 85%|████████▍ | 40668/48008 [2:19:40<26:14, 4.66it/s] 85%|████████▍ | 40669/48008 [2:19:40<25:23, 4.82it/s] 85%|████████▍ | 40670/48008 [2:19:40<24:33, 4.98it/s] 85%|████████▍ | 40671/48008 [2:19:40<24:10, 5.06it/s] 85%|████████▍ | 40672/48008 [2:19:40<23:56, 5.11it/s] 85%|████████▍ | 40673/48008 [2:19:41<23:50, 5.13it/s] 85%|████████▍ | 40674/48008 [2:19:41<23:45, 5.15it/s] 85%|████████▍ | 40675/48008 [2:19:41<23:37, 5.17it/s] 85%|████████▍ | 40676/48008 [2:19:41<24:12, 5.05it/s] 85%|████████▍ | 40677/48008 [2:19:41<23:35, 5.18it/s] 85%|████████▍ | 40678/48008 [2:19:42<23:33, 5.19it/s] 85%|████████▍ | 40679/48008 [2:19:42<23:33, 5.19it/s] 85%|████████▍ | 40680/48008 [2:19:42<23:39, 5.16it/s] 85%|████████▍ | 40681/48008 [2:19:42<23:22, 5.23it/s] 85%|████████▍ | 40682/48008 [2:19:42<23:19, 5.24it/s] 85%|████████▍ | 40683/48008 [2:19:42<23:42, 5.15it/s] 85%|████████▍ | 40684/48008 [2:19:43<23:28, 5.20it/s] 85%|████████▍ | 40685/48008 [2:19:43<24:20, 5.01it/s] 85%|████████▍ | 40686/48008 [2:19:43<25:48, 4.73it/s] 85%|████████▍ | 40687/48008 [2:19:43<24:49, 4.91it/s] 85%|████████▍ | 40688/48008 [2:19:44<24:23, 5.00it/s] 85%|████████▍ | 40689/48008 [2:19:44<23:58, 5.09it/s] 85%|████████▍ | 40690/48008 [2:19:44<23:50, 5.11it/s] 85%|████████▍ | 40691/48008 [2:19:44<23:46, 5.13it/s] 85%|████████▍ | 40692/48008 [2:19:44<23:23, 5.21it/s] 85%|████████▍ | 40693/48008 [2:19:44<23:18, 5.23it/s] 85%|████████▍ | 40694/48008 [2:19:45<23:40, 5.15it/s] 85%|████████▍ | 40695/48008 [2:19:45<23:24, 5.21it/s] 85%|████████▍ | 40696/48008 [2:19:45<23:14, 5.24it/s] 85%|████████▍ | 40697/48008 [2:19:45<23:16, 5.23it/s] 85%|████████▍ | 40698/48008 [2:19:45<23:12, 5.25it/s] 85%|████████▍ | 40699/48008 [2:19:46<23:12, 5.25it/s] 85%|████████▍ | 40700/48008 [2:19:46<23:24, 5.20it/s] {'loss': 0.3018, 'grad_norm': 0.1640625, 'learning_rate': 3.0449091818030327e-05, 'epoch': 0.85} 85%|████████▍ | 40700/48008 [2:19:46<23:24, 5.20it/s] 85%|████████▍ | 40701/48008 [2:19:46<23:12, 5.25it/s] 85%|████████▍ | 40702/48008 [2:19:46<23:58, 5.08it/s] 85%|████████▍ | 40703/48008 [2:19:46<23:47, 5.12it/s] 85%|████████▍ | 40704/48008 [2:19:47<23:35, 5.16it/s] 85%|████████▍ | 40705/48008 [2:19:47<23:58, 5.08it/s] 85%|████████▍ | 40706/48008 [2:19:47<24:23, 4.99it/s] 85%|████████▍ | 40707/48008 [2:19:47<25:42, 4.73it/s] 85%|████████▍ | 40708/48008 [2:19:47<24:48, 4.90it/s] 85%|████████▍ | 40709/48008 [2:19:48<24:12, 5.02it/s] 85%|████████▍ | 40710/48008 [2:19:48<23:48, 5.11it/s] 85%|████████▍ | 40711/48008 [2:19:48<23:35, 5.15it/s] 85%|████████▍ | 40712/48008 [2:19:48<23:31, 5.17it/s] 85%|████████▍ | 40713/48008 [2:19:48<23:41, 5.13it/s] 85%|████████▍ | 40714/48008 [2:19:49<23:37, 5.14it/s] 85%|████████▍ | 40715/48008 [2:19:49<25:15, 4.81it/s] 85%|████████▍ | 40716/48008 [2:19:49<24:35, 4.94it/s] 85%|████████▍ | 40717/48008 [2:19:49<24:15, 5.01it/s] 85%|████████▍ | 40718/48008 [2:19:49<24:19, 5.00it/s] 85%|████████▍ | 40719/48008 [2:19:50<24:24, 4.98it/s] 85%|████████▍ | 40720/48008 [2:19:50<24:04, 5.05it/s] 85%|████████▍ | 40721/48008 [2:19:50<24:25, 4.97it/s] 85%|████████▍ | 40722/48008 [2:19:50<23:53, 5.08it/s] 85%|████████▍ | 40723/48008 [2:19:50<23:37, 5.14it/s] 85%|████████▍ | 40724/48008 [2:19:51<24:06, 5.03it/s] 85%|████████▍ | 40725/48008 [2:19:51<23:49, 5.09it/s] 85%|████████▍ | 40726/48008 [2:19:51<23:46, 5.10it/s] 85%|████████▍ | 40727/48008 [2:19:51<23:59, 5.06it/s] 85%|████████▍ | 40728/48008 [2:19:51<23:39, 5.13it/s] 85%|████████▍ | 40729/48008 [2:19:52<23:24, 5.18it/s] 85%|████████▍ | 40730/48008 [2:19:52<23:28, 5.17it/s] 85%|████████▍ | 40731/48008 [2:19:52<23:08, 5.24it/s] 85%|████████▍ | 40732/48008 [2:19:52<23:13, 5.22it/s] 85%|████████▍ | 40733/48008 [2:19:52<23:14, 5.22it/s] 85%|████████▍ | 40734/48008 [2:19:53<23:23, 5.18it/s] 85%|████████▍ | 40735/48008 [2:19:53<24:00, 5.05it/s] 85%|████████▍ | 40736/48008 [2:19:53<24:14, 5.00it/s] 85%|████████▍ | 40737/48008 [2:19:53<31:00, 3.91it/s] 85%|████████▍ | 40738/48008 [2:19:54<29:00, 4.18it/s] 85%|████████▍ | 40739/48008 [2:19:54<27:12, 4.45it/s] 85%|████████▍ | 40740/48008 [2:19:54<26:23, 4.59it/s] 85%|████████▍ | 40741/48008 [2:19:54<25:24, 4.77it/s] 85%|████████▍ | 40742/48008 [2:19:54<25:05, 4.82it/s] 85%|████████▍ | 40743/48008 [2:19:54<24:24, 4.96it/s] 85%|████████▍ | 40744/48008 [2:19:55<23:58, 5.05it/s] 85%|████████▍ | 40745/48008 [2:19:55<23:44, 5.10it/s] 85%|████████▍ | 40746/48008 [2:19:55<23:22, 5.18it/s] 85%|████████▍ | 40747/48008 [2:19:55<23:30, 5.15it/s] 85%|████████▍ | 40748/48008 [2:19:55<23:13, 5.21it/s] 85%|████████▍ | 40749/48008 [2:19:56<23:09, 5.23it/s] 85%|████████▍ | 40750/48008 [2:19:56<23:05, 5.24it/s] {'loss': 0.2996, 'grad_norm': 0.18359375, 'learning_rate': 3.0240793201133145e-05, 'epoch': 0.85} 85%|████████▍ | 40750/48008 [2:19:56<23:05, 5.24it/s] 85%|████████▍ | 40751/48008 [2:19:56<22:53, 5.28it/s] 85%|████████▍ | 40752/48008 [2:19:56<23:05, 5.24it/s] 85%|████████▍ | 40753/48008 [2:19:56<23:14, 5.20it/s] 85%|████████▍ | 40754/48008 [2:19:57<23:12, 5.21it/s] 85%|████████▍ | 40755/48008 [2:19:57<23:21, 5.18it/s] 85%|████████▍ | 40756/48008 [2:19:57<23:43, 5.09it/s] 85%|████████▍ | 40757/48008 [2:19:57<23:31, 5.14it/s] 85%|████████▍ | 40758/48008 [2:19:57<23:14, 5.20it/s] 85%|████████▍ | 40759/48008 [2:19:58<23:34, 5.13it/s] 85%|████████▍ | 40760/48008 [2:19:58<23:38, 5.11it/s] 85%|████████▍ | 40761/48008 [2:19:58<23:24, 5.16it/s] 85%|████████▍ | 40762/48008 [2:19:58<23:25, 5.15it/s] 85%|████████▍ | 40763/48008 [2:19:58<23:18, 5.18it/s] 85%|████████▍ | 40764/48008 [2:19:59<23:29, 5.14it/s] 85%|████████▍ | 40765/48008 [2:19:59<23:43, 5.09it/s] 85%|████████▍ | 40766/48008 [2:19:59<23:43, 5.09it/s] 85%|████████▍ | 40767/48008 [2:19:59<25:11, 4.79it/s] 85%|████████▍ | 40768/48008 [2:19:59<24:36, 4.90it/s] 85%|████████▍ | 40769/48008 [2:20:00<24:47, 4.87it/s] 85%|████████▍ | 40770/48008 [2:20:00<24:00, 5.02it/s] 85%|████████▍ | 40771/48008 [2:20:00<23:38, 5.10it/s] 85%|████████▍ | 40772/48008 [2:20:00<23:20, 5.17it/s] 85%|████████▍ | 40773/48008 [2:20:00<23:22, 5.16it/s] 85%|████████▍ | 40774/48008 [2:20:01<23:15, 5.19it/s] 85%|████████▍ | 40775/48008 [2:20:01<23:08, 5.21it/s] 85%|████████▍ | 40776/48008 [2:20:01<23:06, 5.22it/s] 85%|████████▍ | 40777/48008 [2:20:01<22:57, 5.25it/s] 85%|████████▍ | 40778/48008 [2:20:01<23:00, 5.24it/s] 85%|████████▍ | 40779/48008 [2:20:01<23:02, 5.23it/s] 85%|████████▍ | 40780/48008 [2:20:02<22:54, 5.26it/s] 85%|████████▍ | 40781/48008 [2:20:02<22:57, 5.24it/s] 85%|████████▍ | 40782/48008 [2:20:02<22:53, 5.26it/s] 85%|████████▍ | 40783/48008 [2:20:02<22:57, 5.24it/s] 85%|████████▍ | 40784/48008 [2:20:02<23:25, 5.14it/s] 85%|████████▍ | 40785/48008 [2:20:03<23:44, 5.07it/s] 85%|████████▍ | 40786/48008 [2:20:03<23:48, 5.06it/s] 85%|████████▍ | 40787/48008 [2:20:03<23:38, 5.09it/s] 85%|████████▍ | 40788/48008 [2:20:03<23:40, 5.08it/s] 85%|████████▍ | 40789/48008 [2:20:03<23:57, 5.02it/s] 85%|████████▍ | 40790/48008 [2:20:04<23:43, 5.07it/s] 85%|████████▍ | 40791/48008 [2:20:04<23:31, 5.11it/s] 85%|████████▍ | 40792/48008 [2:20:04<23:36, 5.09it/s] 85%|████████▍ | 40793/48008 [2:20:04<23:21, 5.15it/s] 85%|████████▍ | 40794/48008 [2:20:04<24:01, 5.01it/s] 85%|████████▍ | 40795/48008 [2:20:05<23:48, 5.05it/s] 85%|████████▍ | 40796/48008 [2:20:05<23:28, 5.12it/s] 85%|████████▍ | 40797/48008 [2:20:05<23:19, 5.15it/s] 85%|████████▍ | 40798/48008 [2:20:05<23:09, 5.19it/s] 85%|████████▍ | 40799/48008 [2:20:05<23:06, 5.20it/s] 85%|████████▍ | 40800/48008 [2:20:06<23:07, 5.19it/s] 85%|████████▍ | 40800/48008 [2:20:06<23:07, 5.19it/s]{'loss': 0.3018, 'grad_norm': 0.3203125, 'learning_rate': 3.003249458423596e-05, 'epoch': 0.85} 85%|████████▍ | 40801/48008 [2:20:06<23:05, 5.20it/s] 85%|████████▍ | 40802/48008 [2:20:06<23:30, 5.11it/s] 85%|████████▍ | 40803/48008 [2:20:06<23:15, 5.16it/s] 85%|████████▍ | 40804/48008 [2:20:06<23:11, 5.18it/s] 85%|████████▍ | 40805/48008 [2:20:07<23:05, 5.20it/s] 85%|████████▍ | 40806/48008 [2:20:07<23:26, 5.12it/s] 85%|████████▌ | 40807/48008 [2:20:07<23:03, 5.21it/s] 85%|████████▌ | 40808/48008 [2:20:07<23:11, 5.17it/s] 85%|████████▌ | 40809/48008 [2:20:07<23:38, 5.08it/s] 85%|████████▌ | 40810/48008 [2:20:07<23:09, 5.18it/s] 85%|████████▌ | 40811/48008 [2:20:08<23:10, 5.18it/s] 85%|████████▌ | 40812/48008 [2:20:08<23:07, 5.19it/s] 85%|████████▌ | 40813/48008 [2:20:08<23:11, 5.17it/s] 85%|████████▌ | 40814/48008 [2:20:08<23:12, 5.17it/s] 85%|████████▌ | 40815/48008 [2:20:08<23:06, 5.19it/s] 85%|████████▌ | 40816/48008 [2:20:09<23:16, 5.15it/s] 85%|████████▌ | 40817/48008 [2:20:09<23:50, 5.03it/s] 85%|████████▌ | 40818/48008 [2:20:09<23:34, 5.08it/s] 85%|████████▌ | 40819/48008 [2:20:09<25:07, 4.77it/s] 85%|████████▌ | 40820/48008 [2:20:09<24:28, 4.89it/s] 85%|████████▌ | 40821/48008 [2:20:10<24:03, 4.98it/s] 85%|████████▌ | 40822/48008 [2:20:10<23:53, 5.01it/s] 85%|████████▌ | 40823/48008 [2:20:10<23:58, 4.99it/s] 85%|████████▌ | 40824/48008 [2:20:10<23:39, 5.06it/s] 85%|████████▌ | 40825/48008 [2:20:10<23:23, 5.12it/s] 85%|████████▌ | 40826/48008 [2:20:11<23:43, 5.04it/s] 85%|████████▌ | 40827/48008 [2:20:11<23:33, 5.08it/s] 85%|████████▌ | 40828/48008 [2:20:11<25:09, 4.76it/s] 85%|████████▌ | 40829/48008 [2:20:11<24:58, 4.79it/s] 85%|████████▌ | 40830/48008 [2:20:12<24:32, 4.87it/s] 85%|████████▌ | 40831/48008 [2:20:12<24:03, 4.97it/s] 85%|████████▌ | 40832/48008 [2:20:12<23:45, 5.04it/s] 85%|████████▌ | 40833/48008 [2:20:12<23:25, 5.11it/s] 85%|████████▌ | 40834/48008 [2:20:12<23:06, 5.17it/s] 85%|████████▌ | 40835/48008 [2:20:12<23:04, 5.18it/s] 85%|████████▌ | 40836/48008 [2:20:13<23:12, 5.15it/s] 85%|████████▌ | 40837/48008 [2:20:13<23:08, 5.16it/s] 85%|████████▌ | 40838/48008 [2:20:13<22:49, 5.23it/s] 85%|████████▌ | 40839/48008 [2:20:13<22:52, 5.22it/s] 85%|████████▌ | 40840/48008 [2:20:13<22:55, 5.21it/s] 85%|████████▌ | 40841/48008 [2:20:14<23:16, 5.13it/s] 85%|████████▌ | 40842/48008 [2:20:14<23:06, 5.17it/s] 85%|████████▌ | 40843/48008 [2:20:14<22:58, 5.20it/s] 85%|████████▌ | 40844/48008 [2:20:14<23:16, 5.13it/s] 85%|████████▌ | 40845/48008 [2:20:15<49:15, 2.42it/s] 85%|████████▌ | 40846/48008 [2:20:15<41:58, 2.84it/s] 85%|████████▌ | 40847/48008 [2:20:16<36:06, 3.31it/s] 85%|████████▌ | 40848/48008 [2:20:16<33:42, 3.54it/s] 85%|████████▌ | 40849/48008 [2:20:16<30:26, 3.92it/s] 85%|████████▌ | 40850/48008 [2:20:16<28:07, 4.24it/s] {'loss': 0.2996, 'grad_norm': 0.224609375, 'learning_rate': 2.982419596733878e-05, 'epoch': 0.85} 85%|████████▌ | 40850/48008 [2:20:16<28:07, 4.24it/s] 85%|████████▌ | 40851/48008 [2:20:16<26:32, 4.49it/s] 85%|████████▌ | 40852/48008 [2:20:17<25:10, 4.74it/s] 85%|████████▌ | 40853/48008 [2:20:17<24:30, 4.87it/s] 85%|████████▌ | 40854/48008 [2:20:17<30:56, 3.85it/s] 85%|████████▌ | 40855/48008 [2:20:17<28:53, 4.13it/s] 85%|████████▌ | 40856/48008 [2:20:17<26:57, 4.42it/s] 85%|████████▌ | 40857/48008 [2:20:18<48:10, 2.47it/s] 85%|████████▌ | 40858/48008 [2:20:18<40:40, 2.93it/s] 85%|████████▌ | 40859/48008 [2:20:19<35:31, 3.35it/s] 85%|████████▌ | 40860/48008 [2:20:19<31:37, 3.77it/s] 85%|████████▌ | 40861/48008 [2:20:19<28:50, 4.13it/s] 85%|████████▌ | 40862/48008 [2:20:19<27:21, 4.35it/s] 85%|████████▌ | 40863/48008 [2:20:20<53:49, 2.21it/s] 85%|████████▌ | 40864/48008 [2:20:20<44:29, 2.68it/s] 85%|████████▌ | 40865/48008 [2:20:21<38:09, 3.12it/s] 85%|████████▌ | 40866/48008 [2:20:21<33:22, 3.57it/s] 85%|████████▌ | 40867/48008 [2:20:21<30:10, 3.94it/s] 85%|████████▌ | 40868/48008 [2:20:21<28:25, 4.19it/s] 85%|████████▌ | 40869/48008 [2:20:22<49:21, 2.41it/s] 85%|████████▌ | 40870/48008 [2:20:22<41:22, 2.87it/s] 85%|████████▌ | 40871/48008 [2:20:22<35:52, 3.32it/s] 85%|████████▌ | 40872/48008 [2:20:23<31:46, 3.74it/s] 85%|████████▌ | 40873/48008 [2:20:23<28:59, 4.10it/s] 85%|████████▌ | 40874/48008 [2:20:23<27:29, 4.32it/s] 85%|████████▌ | 40875/48008 [2:20:23<26:00, 4.57it/s] 85%|████████▌ | 40876/48008 [2:20:23<25:15, 4.71it/s] 85%|████████▌ | 40877/48008 [2:20:24<31:25, 3.78it/s] 85%|████████▌ | 40878/48008 [2:20:24<30:28, 3.90it/s] 85%|████████▌ | 40879/48008 [2:20:24<28:05, 4.23it/s] 85%|████████▌ | 40880/48008 [2:20:24<26:25, 4.50it/s] 85%|████████▌ | 40881/48008 [2:20:25<25:21, 4.68it/s] 85%|████████▌ | 40882/48008 [2:20:25<24:54, 4.77it/s] 85%|████████▌ | 40883/48008 [2:20:25<24:15, 4.90it/s] 85%|████████▌ | 40884/48008 [2:20:25<23:46, 4.99it/s] 85%|████████▌ | 40885/48008 [2:20:25<23:44, 5.00it/s] 85%|████████▌ | 40886/48008 [2:20:26<23:22, 5.08it/s] 85%|████████▌ | 40887/48008 [2:20:26<23:33, 5.04it/s] 85%|████████▌ | 40888/48008 [2:20:26<23:42, 5.00it/s] 85%|████████▌ | 40889/48008 [2:20:26<23:49, 4.98it/s] 85%|████████▌ | 40890/48008 [2:20:26<23:19, 5.09it/s] 85%|████████▌ | 40891/48008 [2:20:27<23:03, 5.15it/s] 85%|████████▌ | 40892/48008 [2:20:27<23:31, 5.04it/s] 85%|████████▌ | 40893/48008 [2:20:27<27:22, 4.33it/s] 85%|████████▌ | 40894/48008 [2:20:27<25:51, 4.58it/s] 85%|████████▌ | 40895/48008 [2:20:27<24:44, 4.79it/s] 85%|████████▌ | 40896/48008 [2:20:28<24:30, 4.84it/s] 85%|████████▌ | 40897/48008 [2:20:28<23:44, 4.99it/s] 85%|████████▌ | 40898/48008 [2:20:28<23:44, 4.99it/s] 85%|████████▌ | 40899/48008 [2:20:28<23:24, 5.06it/s] 85%|████████▌ | 40900/48008 [2:20:28<23:08, 5.12it/s] {'loss': 0.297, 'grad_norm': 0.2001953125, 'learning_rate': 2.9615897350441597e-05, 'epoch': 0.85} 85%|████████▌ | 40900/48008 [2:20:28<23:08, 5.12it/s] 85%|████████▌ | 40901/48008 [2:20:29<23:01, 5.15it/s] 85%|████████▌ | 40902/48008 [2:20:29<22:49, 5.19it/s] 85%|████████▌ | 40903/48008 [2:20:29<22:39, 5.23it/s] 85%|████████▌ | 40904/48008 [2:20:29<22:41, 5.22it/s] 85%|████████▌ | 40905/48008 [2:20:29<22:39, 5.23it/s] 85%|████████▌ | 40906/48008 [2:20:30<22:57, 5.15it/s] 85%|████████▌ | 40907/48008 [2:20:30<22:47, 5.19it/s] 85%|████████▌ | 40908/48008 [2:20:30<22:44, 5.20it/s] 85%|████████▌ | 40909/48008 [2:20:30<24:16, 4.87it/s] 85%|████████▌ | 40910/48008 [2:20:30<23:42, 4.99it/s] 85%|████████▌ | 40911/48008 [2:20:31<23:25, 5.05it/s] 85%|████████▌ | 40912/48008 [2:20:31<23:11, 5.10it/s] 85%|████████▌ | 40913/48008 [2:20:31<23:01, 5.13it/s] 85%|████████▌ | 40914/48008 [2:20:31<22:51, 5.17it/s] 85%|████████▌ | 40915/48008 [2:20:31<23:00, 5.14it/s] 85%|████████▌ | 40916/48008 [2:20:32<22:53, 5.16it/s] 85%|████████▌ | 40917/48008 [2:20:32<46:13, 2.56it/s] 85%|████████▌ | 40918/48008 [2:20:33<40:43, 2.90it/s] 85%|████████▌ | 40919/48008 [2:20:33<35:24, 3.34it/s] 85%|████████▌ | 40920/48008 [2:20:33<31:25, 3.76it/s] 85%|████████▌ | 40921/48008 [2:20:33<28:42, 4.11it/s] 85%|████████▌ | 40922/48008 [2:20:33<26:55, 4.39it/s] 85%|████████▌ | 40923/48008 [2:20:34<25:33, 4.62it/s] 85%|████████▌ | 40924/48008 [2:20:34<24:23, 4.84it/s] 85%|████████▌ | 40925/48008 [2:20:34<24:02, 4.91it/s] 85%|████████▌ | 40926/48008 [2:20:34<23:58, 4.92it/s] 85%|████████▌ | 40927/48008 [2:20:34<23:33, 5.01it/s] 85%|████████▌ | 40928/48008 [2:20:35<23:45, 4.97it/s] 85%|████████▌ | 40929/48008 [2:20:35<27:27, 4.30it/s] 85%|████████▌ | 40930/48008 [2:20:35<26:23, 4.47it/s] 85%|████████▌ | 40931/48008 [2:20:35<25:05, 4.70it/s] 85%|████████▌ | 40932/48008 [2:20:35<24:26, 4.82it/s] 85%|████████▌ | 40933/48008 [2:20:36<23:48, 4.95it/s] 85%|████████▌ | 40934/48008 [2:20:36<25:05, 4.70it/s] 85%|████████▌ | 40935/48008 [2:20:36<25:54, 4.55it/s] 85%|████████▌ | 40936/48008 [2:20:36<25:04, 4.70it/s] 85%|████████▌ | 40937/48008 [2:20:36<24:10, 4.88it/s] 85%|████████▌ | 40938/48008 [2:20:37<23:41, 4.98it/s] 85%|████████▌ | 40939/48008 [2:20:37<23:10, 5.09it/s] 85%|████████▌ | 40940/48008 [2:20:37<22:43, 5.18it/s] 85%|████████▌ | 40941/48008 [2:20:37<22:45, 5.18it/s] 85%|████████▌ | 40942/48008 [2:20:37<22:34, 5.22it/s] 85%|████████▌ | 40943/48008 [2:20:38<22:30, 5.23it/s] 85%|████████▌ | 40944/48008 [2:20:38<22:53, 5.14it/s] 85%|████████▌ | 40945/48008 [2:20:38<22:36, 5.21it/s] 85%|████████▌ | 40946/48008 [2:20:38<22:34, 5.21it/s] 85%|████████▌ | 40947/48008 [2:20:38<22:27, 5.24it/s] 85%|████████▌ | 40948/48008 [2:20:39<22:23, 5.26it/s] 85%|████████▌ | 40949/48008 [2:20:39<22:24, 5.25it/s] 85%|████████▌ | 40950/48008 [2:20:39<22:20, 5.27it/s] 85%|████████▌ | 40950/48008 [2:20:39<22:20, 5.27it/s]{'loss': 0.3005, 'grad_norm': 0.408203125, 'learning_rate': 2.940759873354441e-05, 'epoch': 0.85} 85%|████████▌ | 40951/48008 [2:20:39<22:20, 5.26it/s] 85%|████████▌ | 40952/48008 [2:20:39<22:09, 5.31it/s] 85%|████████▌ | 40953/48008 [2:20:40<22:17, 5.27it/s] 85%|████████▌ | 40954/48008 [2:20:40<22:57, 5.12it/s] 85%|████████▌ | 40955/48008 [2:20:40<22:40, 5.18it/s] 85%|████████▌ | 40956/48008 [2:20:40<22:45, 5.16it/s] 85%|████████▌ | 40957/48008 [2:20:40<22:36, 5.20it/s] 85%|████████▌ | 40958/48008 [2:20:40<22:34, 5.20it/s] 85%|████████▌ | 40959/48008 [2:20:41<22:55, 5.12it/s] 85%|████████▌ | 40960/48008 [2:20:41<22:38, 5.19it/s] 85%|████████▌ | 40961/48008 [2:20:41<22:31, 5.22it/s] 85%|████████▌ | 40962/48008 [2:20:41<22:29, 5.22it/s] 85%|████████▌ | 40963/48008 [2:20:41<24:06, 4.87it/s] 85%|████████▌ | 40964/48008 [2:20:42<23:38, 4.96it/s] 85%|████████▌ | 40965/48008 [2:20:42<23:19, 5.03it/s] 85%|████████▌ | 40966/48008 [2:20:42<23:17, 5.04it/s] 85%|████████▌ | 40967/48008 [2:20:42<23:00, 5.10it/s] 85%|████████▌ | 40968/48008 [2:20:42<23:14, 5.05it/s] 85%|████████▌ | 40969/48008 [2:20:43<22:57, 5.11it/s] 85%|████████▌ | 40970/48008 [2:20:43<22:36, 5.19it/s] 85%|████████▌ | 40971/48008 [2:20:43<22:31, 5.21it/s] 85%|████████▌ | 40972/48008 [2:20:43<22:26, 5.23it/s] 85%|████████▌ | 40973/48008 [2:20:43<22:36, 5.18it/s] 85%|████████▌ | 40974/48008 [2:20:44<22:22, 5.24it/s] 85%|████████▌ | 40975/48008 [2:20:44<22:17, 5.26it/s] 85%|████████▌ | 40976/48008 [2:20:44<22:52, 5.12it/s] 85%|████████▌ | 40977/48008 [2:20:44<23:07, 5.07it/s] 85%|████████▌ | 40978/48008 [2:20:44<22:42, 5.16it/s] 85%|████████▌ | 40979/48008 [2:20:45<22:46, 5.14it/s] 85%|████████▌ | 40980/48008 [2:20:45<22:33, 5.19it/s] 85%|████████▌ | 40981/48008 [2:20:45<22:30, 5.20it/s] 85%|████████▌ | 40982/48008 [2:20:45<22:25, 5.22it/s] 85%|████████▌ | 40983/48008 [2:20:45<22:20, 5.24it/s] 85%|████████▌ | 40984/48008 [2:20:46<22:12, 5.27it/s] 85%|████████▌ | 40985/48008 [2:20:46<22:16, 5.26it/s] 85%|████████▌ | 40986/48008 [2:20:46<22:12, 5.27it/s] 85%|████████▌ | 40987/48008 [2:20:46<22:27, 5.21it/s] 85%|████████▌ | 40988/48008 [2:20:46<22:26, 5.21it/s] 85%|████████▌ | 40989/48008 [2:20:46<22:16, 5.25it/s] 85%|████████▌ | 40990/48008 [2:20:47<26:21, 4.44it/s] 85%|████████▌ | 40991/48008 [2:20:47<25:32, 4.58it/s] 85%|████████▌ | 40992/48008 [2:20:47<24:34, 4.76it/s] 85%|████████▌ | 40993/48008 [2:20:47<24:37, 4.75it/s] 85%|████████▌ | 40994/48008 [2:20:48<23:58, 4.88it/s] 85%|████████▌ | 40995/48008 [2:20:48<30:16, 3.86it/s] 85%|████████▌ | 40996/48008 [2:20:48<27:56, 4.18it/s] 85%|████████▌ | 40997/48008 [2:20:48<26:09, 4.47it/s] 85%|████████▌ | 40998/48008 [2:20:49<25:01, 4.67it/s] 85%|████████▌ | 40999/48008 [2:20:49<24:13, 4.82it/s] 85%|████████▌ | 41000/48008 [2:20:49<23:36, 4.95it/s] {'loss': 0.3008, 'grad_norm': 0.29296875, 'learning_rate': 2.9199300116647228e-05, 'epoch': 0.85} 85%|████████▌ | 41000/48008 [2:20:49<23:36, 4.95it/s] 85%|████████▌ | 41001/48008 [2:20:49<23:15, 5.02it/s] 85%|████████▌ | 41002/48008 [2:20:49<23:02, 5.07it/s] 85%|████████▌ | 41003/48008 [2:20:50<23:10, 5.04it/s] 85%|████████▌ | 41004/48008 [2:20:50<23:04, 5.06it/s] 85%|████████▌ | 41005/48008 [2:20:50<22:58, 5.08it/s] 85%|████████▌ | 41006/48008 [2:20:50<23:12, 5.03it/s] 85%|████████▌ | 41007/48008 [2:20:50<22:51, 5.11it/s] 85%|████████▌ | 41008/48008 [2:20:50<23:09, 5.04it/s] 85%|████████▌ | 41009/48008 [2:20:51<24:30, 4.76it/s] 85%|████████▌ | 41010/48008 [2:20:51<23:49, 4.90it/s] 85%|████████▌ | 41011/48008 [2:20:51<23:15, 5.01it/s] 85%|████████▌ | 41012/48008 [2:20:51<23:23, 4.99it/s] 85%|████████▌ | 41013/48008 [2:20:52<23:02, 5.06it/s] 85%|████████▌ | 41014/48008 [2:20:52<29:35, 3.94it/s] 85%|████████▌ | 41015/48008 [2:20:52<27:18, 4.27it/s] 85%|████████▌ | 41016/48008 [2:20:52<25:48, 4.52it/s] 85%|████████▌ | 41017/48008 [2:20:52<24:42, 4.72it/s] 85%|████████▌ | 41018/48008 [2:20:53<24:20, 4.79it/s] 85%|████████▌ | 41019/48008 [2:20:53<27:43, 4.20it/s] 85%|████████▌ | 41020/48008 [2:20:53<26:01, 4.48it/s] 85%|████████▌ | 41021/48008 [2:20:53<24:54, 4.67it/s] 85%|████████▌ | 41022/48008 [2:20:54<24:03, 4.84it/s] 85%|████████▌ | 41023/48008 [2:20:54<23:42, 4.91it/s] 85%|████████▌ | 41024/48008 [2:20:54<23:10, 5.02it/s] 85%|████████▌ | 41025/48008 [2:20:54<29:44, 3.91it/s] 85%|████████▌ | 41026/48008 [2:20:55<28:13, 4.12it/s] 85%|████████▌ | 41027/48008 [2:20:55<26:15, 4.43it/s] 85%|████████▌ | 41028/48008 [2:20:55<25:16, 4.60it/s] 85%|████████▌ | 41029/48008 [2:20:55<24:15, 4.80it/s] 85%|████████▌ | 41030/48008 [2:20:55<23:52, 4.87it/s] 85%|████████▌ | 41031/48008 [2:20:55<23:19, 4.99it/s] 85%|████████▌ | 41032/48008 [2:20:56<23:12, 5.01it/s] 85%|████████▌ | 41033/48008 [2:20:56<22:48, 5.10it/s] 85%|████████▌ | 41034/48008 [2:20:56<22:37, 5.14it/s] 85%|████████▌ | 41035/48008 [2:20:56<23:12, 5.01it/s] 85%|████████▌ | 41036/48008 [2:20:56<23:01, 5.05it/s] 85%|████████▌ | 41037/48008 [2:20:57<26:45, 4.34it/s] 85%|████████▌ | 41038/48008 [2:20:57<25:19, 4.59it/s] 85%|████████▌ | 41039/48008 [2:20:57<24:40, 4.71it/s] 85%|████████▌ | 41040/48008 [2:20:57<23:50, 4.87it/s] 85%|████████▌ | 41041/48008 [2:20:58<30:08, 3.85it/s] 85%|████████▌ | 41042/48008 [2:20:58<27:47, 4.18it/s] 85%|████████▌ | 41043/48008 [2:20:58<26:06, 4.45it/s] 85%|████████▌ | 41044/48008 [2:20:58<25:19, 4.58it/s] 85%|████████▌ | 41045/48008 [2:20:59<24:21, 4.76it/s] 85%|████████▌ | 41046/48008 [2:20:59<23:42, 4.89it/s] 86%|████████▌ | 41047/48008 [2:20:59<23:22, 4.96it/s] 86%|████████▌ | 41048/48008 [2:20:59<23:30, 4.94it/s] 86%|████████▌ | 41049/48008 [2:20:59<23:03, 5.03it/s] 86%|████████▌ | 41050/48008 [2:21:00<29:32, 3.93it/s] {'loss': 0.298, 'grad_norm': 0.435546875, 'learning_rate': 2.8991001499750042e-05, 'epoch': 0.86} 86%|████████▌ | 41050/48008 [2:21:00<29:32, 3.93it/s] 86%|████████▌ | 41051/48008 [2:21:00<27:22, 4.23it/s] 86%|████████▌ | 41052/48008 [2:21:00<25:56, 4.47it/s] 86%|████████▌ | 41053/48008 [2:21:00<24:50, 4.67it/s] 86%|████████▌ | 41054/48008 [2:21:00<24:03, 4.82it/s] 86%|████████▌ | 41055/48008 [2:21:01<23:26, 4.94it/s] 86%|████████▌ | 41056/48008 [2:21:01<23:04, 5.02it/s] 86%|████████▌ | 41057/48008 [2:21:01<23:25, 4.94it/s] 86%|████████▌ | 41058/48008 [2:21:01<23:04, 5.02it/s] 86%|████████▌ | 41059/48008 [2:21:01<22:43, 5.10it/s] 86%|████████▌ | 41060/48008 [2:21:02<22:44, 5.09it/s] 86%|████████▌ | 41061/48008 [2:21:02<22:34, 5.13it/s] 86%|████████▌ | 41062/48008 [2:21:02<24:03, 4.81it/s] 86%|████████▌ | 41063/48008 [2:21:02<23:31, 4.92it/s] 86%|████████▌ | 41064/48008 [2:21:02<23:03, 5.02it/s] 86%|████████▌ | 41065/48008 [2:21:03<22:41, 5.10it/s] 86%|████████▌ | 41066/48008 [2:21:03<22:26, 5.15it/s] 86%|████████▌ | 41067/48008 [2:21:03<22:21, 5.17it/s] 86%|████████▌ | 41068/48008 [2:21:03<22:40, 5.10it/s] 86%|████████▌ | 41069/48008 [2:21:03<22:41, 5.09it/s] 86%|████████▌ | 41070/48008 [2:21:04<24:06, 4.80it/s] 86%|████████▌ | 41071/48008 [2:21:04<25:08, 4.60it/s] 86%|████████▌ | 41072/48008 [2:21:04<24:12, 4.78it/s] 86%|████████▌ | 41073/48008 [2:21:04<23:31, 4.91it/s] 86%|████████▌ | 41074/48008 [2:21:04<23:30, 4.92it/s] 86%|████████▌ | 41075/48008 [2:21:05<22:53, 5.05it/s] 86%|████████▌ | 41076/48008 [2:21:05<22:40, 5.09it/s] 86%|████████▌ | 41077/48008 [2:21:05<22:20, 5.17it/s] 86%|████████▌ | 41078/48008 [2:21:05<22:12, 5.20it/s] 86%|████████▌ | 41079/48008 [2:21:05<21:59, 5.25it/s] 86%|████████▌ | 41080/48008 [2:21:06<22:01, 5.24it/s] 86%|████████▌ | 41081/48008 [2:21:06<22:27, 5.14it/s] 86%|████████▌ | 41082/48008 [2:21:06<22:23, 5.16it/s] 86%|████████▌ | 41083/48008 [2:21:06<22:13, 5.19it/s] 86%|████████▌ | 41084/48008 [2:21:06<22:00, 5.24it/s] 86%|████████▌ | 41085/48008 [2:21:07<22:07, 5.22it/s] 86%|████████▌ | 41086/48008 [2:21:07<26:07, 4.41it/s] 86%|████████▌ | 41087/48008 [2:21:07<25:04, 4.60it/s] 86%|████████▌ | 41088/48008 [2:21:07<24:01, 4.80it/s] 86%|████████▌ | 41089/48008 [2:21:07<23:24, 4.93it/s] 86%|████████▌ | 41090/48008 [2:21:08<24:32, 4.70it/s] 86%|████████▌ | 41091/48008 [2:21:08<23:47, 4.85it/s] 86%|████████▌ | 41092/48008 [2:21:08<23:10, 4.97it/s] 86%|████████▌ | 41093/48008 [2:21:08<22:46, 5.06it/s] 86%|████████▌ | 41094/48008 [2:21:08<22:32, 5.11it/s] 86%|████████▌ | 41095/48008 [2:21:09<22:22, 5.15it/s] 86%|████████▌ | 41096/48008 [2:21:09<22:10, 5.19it/s] 86%|████████▌ | 41097/48008 [2:21:09<22:28, 5.12it/s] 86%|████████▌ | 41098/48008 [2:21:09<22:46, 5.06it/s] 86%|████████▌ | 41099/48008 [2:21:09<22:35, 5.10it/s] 86%|████████▌ | 41100/48008 [2:21:10<22:11, 5.19it/s] {'loss': 0.2991, 'grad_norm': 0.2177734375, 'learning_rate': 2.8782702882852856e-05, 'epoch': 0.86} 86%|████████▌ | 41100/48008 [2:21:10<22:11, 5.19it/s] 86%|████████▌ | 41101/48008 [2:21:10<22:30, 5.11it/s] 86%|████████▌ | 41102/48008 [2:21:10<22:23, 5.14it/s] 86%|████████▌ | 41103/48008 [2:21:10<22:13, 5.18it/s] 86%|████████▌ | 41104/48008 [2:21:10<21:58, 5.24it/s] 86%|████████▌ | 41105/48008 [2:21:11<22:00, 5.23it/s] 86%|████████▌ | 41106/48008 [2:21:11<23:32, 4.89it/s] 86%|████████▌ | 41107/48008 [2:21:11<22:57, 5.01it/s] 86%|████████▌ | 41108/48008 [2:21:11<22:30, 5.11it/s] 86%|████████▌ | 41109/48008 [2:21:11<22:10, 5.18it/s] 86%|████████▌ | 41110/48008 [2:21:12<21:49, 5.27it/s] 86%|████████▌ | 41111/48008 [2:21:12<21:33, 5.33it/s] 86%|████████▌ | 41112/48008 [2:21:12<21:31, 5.34it/s] 86%|████████▌ | 41113/48008 [2:21:12<21:43, 5.29it/s] 86%|████████▌ | 41114/48008 [2:21:12<21:56, 5.24it/s] 86%|████████▌ | 41115/48008 [2:21:12<21:58, 5.23it/s] 86%|████████▌ | 41116/48008 [2:21:13<22:26, 5.12it/s] 86%|████████▌ | 41117/48008 [2:21:13<22:15, 5.16it/s] 86%|████████▌ | 41118/48008 [2:21:13<26:03, 4.41it/s] 86%|████████▌ | 41119/48008 [2:21:13<24:38, 4.66it/s] 86%|████████▌ | 41120/48008 [2:21:14<25:27, 4.51it/s] 86%|████████▌ | 41121/48008 [2:21:14<24:18, 4.72it/s] 86%|████████▌ | 41122/48008 [2:21:14<23:36, 4.86it/s] 86%|████████▌ | 41123/48008 [2:21:14<23:00, 4.99it/s] 86%|████████▌ | 41124/48008 [2:21:14<22:37, 5.07it/s] 86%|████████▌ | 41125/48008 [2:21:15<22:23, 5.12it/s] 86%|████████▌ | 41126/48008 [2:21:15<22:20, 5.13it/s] 86%|████████▌ | 41127/48008 [2:21:15<22:06, 5.19it/s] 86%|████████▌ | 41128/48008 [2:21:15<22:12, 5.16it/s] 86%|████████▌ | 41129/48008 [2:21:15<22:00, 5.21it/s] 86%|████████▌ | 41130/48008 [2:21:16<28:45, 3.99it/s] 86%|████████▌ | 41131/48008 [2:21:16<33:25, 3.43it/s] 86%|████████▌ | 41132/48008 [2:21:16<29:59, 3.82it/s] 86%|████████▌ | 41133/48008 [2:21:16<27:33, 4.16it/s] 86%|████████▌ | 41134/48008 [2:21:17<25:50, 4.43it/s] 86%|████████▌ | 41135/48008 [2:21:17<24:37, 4.65it/s] 86%|████████▌ | 41136/48008 [2:21:17<23:50, 4.80it/s] 86%|████████▌ | 41137/48008 [2:21:17<23:05, 4.96it/s] 86%|████████▌ | 41138/48008 [2:21:17<22:41, 5.05it/s] 86%|████████▌ | 41139/48008 [2:21:18<22:22, 5.11it/s] 86%|████████▌ | 41140/48008 [2:21:18<22:15, 5.14it/s] 86%|████████▌ | 41141/48008 [2:21:18<22:15, 5.14it/s] 86%|████████▌ | 41142/48008 [2:21:18<22:06, 5.17it/s] 86%|████████▌ | 41143/48008 [2:21:18<21:59, 5.20it/s] 86%|████████▌ | 41144/48008 [2:21:19<22:34, 5.07it/s] 86%|████████▌ | 41145/48008 [2:21:19<22:20, 5.12it/s] 86%|████████▌ | 41146/48008 [2:21:19<22:11, 5.15it/s] 86%|████████▌ | 41147/48008 [2:21:19<22:12, 5.15it/s] 86%|████████▌ | 41148/48008 [2:21:19<22:04, 5.18it/s] 86%|████████▌ | 41149/48008 [2:21:20<23:32, 4.86it/s] 86%|████████▌ | 41150/48008 [2:21:20<26:57, 4.24it/s] 86%|████████▌ | 41150/48008 [2:21:20<26:57, 4.24it/s]{'loss': 0.301, 'grad_norm': 0.197265625, 'learning_rate': 2.8574404265955673e-05, 'epoch': 0.86} 86%|████████▌ | 41151/48008 [2:21:20<25:20, 4.51it/s] 86%|████████▌ | 41152/48008 [2:21:20<24:10, 4.73it/s] 86%|████████▌ | 41153/48008 [2:21:20<23:26, 4.87it/s] 86%|████████▌ | 41154/48008 [2:21:21<23:06, 4.94it/s] 86%|████████▌ | 41155/48008 [2:21:21<23:04, 4.95it/s] 86%|████████▌ | 41156/48008 [2:21:21<23:02, 4.95it/s] 86%|████████▌ | 41157/48008 [2:21:21<22:30, 5.07it/s] 86%|████████▌ | 41158/48008 [2:21:21<22:06, 5.16it/s] 86%|████████▌ | 41159/48008 [2:21:22<22:15, 5.13it/s] 86%|████████▌ | 41160/48008 [2:21:22<22:10, 5.15it/s] 86%|████████▌ | 41161/48008 [2:21:22<22:14, 5.13it/s] 86%|████████▌ | 41162/48008 [2:21:22<22:04, 5.17it/s] 86%|████████▌ | 41163/48008 [2:21:22<22:21, 5.10it/s] 86%|████████▌ | 41164/48008 [2:21:23<22:30, 5.07it/s] 86%|████████▌ | 41165/48008 [2:21:23<22:38, 5.04it/s] 86%|████████▌ | 41166/48008 [2:21:23<22:23, 5.09it/s] 86%|████████▌ | 41167/48008 [2:21:23<22:25, 5.08it/s] 86%|████████▌ | 41168/48008 [2:21:23<22:08, 5.15it/s] 86%|████████▌ | 41169/48008 [2:21:24<21:52, 5.21it/s] 86%|████████▌ | 41170/48008 [2:21:24<21:46, 5.23it/s] 86%|████████▌ | 41171/48008 [2:21:24<21:34, 5.28it/s] 86%|████████▌ | 41172/48008 [2:21:24<22:17, 5.11it/s] 86%|████████▌ | 41173/48008 [2:21:24<21:58, 5.18it/s] 86%|████████▌ | 41174/48008 [2:21:25<23:26, 4.86it/s] 86%|████████▌ | 41175/48008 [2:21:25<23:01, 4.95it/s] 86%|████████▌ | 41176/48008 [2:21:25<22:38, 5.03it/s] 86%|████████▌ | 41177/48008 [2:21:25<22:12, 5.13it/s] 86%|████████▌ | 41178/48008 [2:21:25<22:07, 5.15it/s] 86%|████████▌ | 41179/48008 [2:21:26<22:04, 5.16it/s] 86%|████████▌ | 41180/48008 [2:21:26<21:36, 5.27it/s] 86%|████████▌ | 41181/48008 [2:21:26<21:28, 5.30it/s] 86%|████████▌ | 41182/48008 [2:21:26<21:21, 5.33it/s] 86%|████████▌ | 41183/48008 [2:21:26<21:23, 5.32it/s] 86%|████████▌ | 41184/48008 [2:21:26<21:23, 5.31it/s] 86%|████████▌ | 41185/48008 [2:21:27<21:20, 5.33it/s] 86%|████████▌ | 41186/48008 [2:21:27<21:24, 5.31it/s] 86%|████████▌ | 41187/48008 [2:21:27<21:20, 5.32it/s] 86%|████████▌ | 41188/48008 [2:21:27<25:22, 4.48it/s] 86%|████████▌ | 41189/48008 [2:21:28<24:32, 4.63it/s] 86%|████████▌ | 41190/48008 [2:21:28<23:42, 4.79it/s] 86%|████████▌ | 41191/48008 [2:21:28<23:30, 4.83it/s] 86%|████████▌ | 41192/48008 [2:21:28<22:51, 4.97it/s] 86%|████████▌ | 41193/48008 [2:21:28<22:27, 5.06it/s] 86%|████████▌ | 41194/48008 [2:21:29<23:48, 4.77it/s] 86%|████████▌ | 41195/48008 [2:21:29<22:59, 4.94it/s] 86%|████████▌ | 41196/48008 [2:21:29<22:43, 5.00it/s] 86%|████████▌ | 41197/48008 [2:21:29<23:01, 4.93it/s] 86%|████████▌ | 41198/48008 [2:21:29<22:40, 5.00it/s] 86%|████████▌ | 41199/48008 [2:21:30<22:19, 5.08it/s] 86%|████████▌ | 41200/48008 [2:21:30<22:23, 5.07it/s] {'loss': 0.301, 'grad_norm': 0.150390625, 'learning_rate': 2.8366105649058494e-05, 'epoch': 0.86} 86%|████████▌ | 41200/48008 [2:21:30<22:23, 5.07it/s] 86%|████████▌ | 41201/48008 [2:21:30<22:13, 5.10it/s] 86%|████████▌ | 41202/48008 [2:21:30<22:16, 5.09it/s] 86%|████████▌ | 41203/48008 [2:21:30<22:27, 5.05it/s] 86%|████████▌ | 41204/48008 [2:21:31<22:14, 5.10it/s] 86%|████████▌ | 41205/48008 [2:21:31<22:23, 5.06it/s] 86%|████████▌ | 41206/48008 [2:21:31<22:10, 5.11it/s] 86%|████████▌ | 41207/48008 [2:21:31<21:57, 5.16it/s] 86%|████████▌ | 41208/48008 [2:21:31<22:12, 5.10it/s] 86%|████████▌ | 41209/48008 [2:21:31<21:52, 5.18it/s] 86%|████████▌ | 41210/48008 [2:21:32<21:51, 5.18it/s] 86%|████████▌ | 41211/48008 [2:21:32<21:45, 5.21it/s] 86%|████████▌ | 41212/48008 [2:21:32<22:18, 5.08it/s] 86%|████████▌ | 41213/48008 [2:21:32<22:42, 4.99it/s] 86%|████████▌ | 41214/48008 [2:21:32<22:36, 5.01it/s] 86%|████████▌ | 41215/48008 [2:21:33<22:19, 5.07it/s] 86%|████████▌ | 41216/48008 [2:21:33<23:39, 4.78it/s] 86%|████████▌ | 41217/48008 [2:21:33<23:16, 4.86it/s] 86%|████████▌ | 41218/48008 [2:21:33<29:26, 3.84it/s] 86%|████████▌ | 41219/48008 [2:21:34<27:06, 4.17it/s] 86%|████████▌ | 41220/48008 [2:21:34<25:36, 4.42it/s] 86%|████████▌ | 41221/48008 [2:21:34<24:58, 4.53it/s] 86%|████████▌ | 41222/48008 [2:21:34<24:06, 4.69it/s] 86%|████████▌ | 41223/48008 [2:21:34<23:27, 4.82it/s] 86%|████████▌ | 41224/48008 [2:21:35<22:50, 4.95it/s] 86%|████████▌ | 41225/48008 [2:21:35<22:42, 4.98it/s] 86%|████████▌ | 41226/48008 [2:21:35<26:16, 4.30it/s] 86%|████████▌ | 41227/48008 [2:21:35<24:47, 4.56it/s] 86%|████████▌ | 41228/48008 [2:21:36<24:02, 4.70it/s] 86%|████████▌ | 41229/48008 [2:21:36<23:11, 4.87it/s] 86%|████████▌ | 41230/48008 [2:21:36<23:03, 4.90it/s] 86%|████████▌ | 41231/48008 [2:21:36<22:36, 4.99it/s] 86%|████████▌ | 41232/48008 [2:21:36<23:52, 4.73it/s] 86%|████████▌ | 41233/48008 [2:21:37<23:11, 4.87it/s] 86%|████████▌ | 41234/48008 [2:21:37<23:02, 4.90it/s] 86%|████████▌ | 41235/48008 [2:21:37<22:45, 4.96it/s] 86%|████████▌ | 41236/48008 [2:21:37<22:21, 5.05it/s] 86%|████████▌ | 41237/48008 [2:21:37<22:12, 5.08it/s] 86%|████████▌ | 41238/48008 [2:21:38<22:01, 5.12it/s] 86%|████████▌ | 41239/48008 [2:21:38<21:52, 5.16it/s] 86%|████████▌ | 41240/48008 [2:21:38<21:42, 5.20it/s] 86%|████████▌ | 41241/48008 [2:21:38<21:42, 5.20it/s] 86%|████████▌ | 41242/48008 [2:21:38<21:54, 5.15it/s] 86%|████████▌ | 41243/48008 [2:21:38<21:45, 5.18it/s] 86%|████████▌ | 41244/48008 [2:21:39<22:04, 5.11it/s] 86%|████████▌ | 41245/48008 [2:21:39<21:52, 5.15it/s] 86%|████████▌ | 41246/48008 [2:21:39<21:34, 5.23it/s] 86%|████████▌ | 41247/48008 [2:21:39<21:42, 5.19it/s] 86%|████████▌ | 41248/48008 [2:21:39<22:13, 5.07it/s] 86%|████████▌ | 41249/48008 [2:21:40<22:09, 5.08it/s] 86%|████████▌ | 41250/48008 [2:21:40<25:48, 4.36it/s] {'loss': 0.2968, 'grad_norm': 0.419921875, 'learning_rate': 2.8157807032161308e-05, 'epoch': 0.86} 86%|████████▌ | 41250/48008 [2:21:40<25:48, 4.36it/s] 86%|████████▌ | 41251/48008 [2:21:40<24:32, 4.59it/s] 86%|████████▌ | 41252/48008 [2:21:40<23:49, 4.73it/s] 86%|████████▌ | 41253/48008 [2:21:41<23:03, 4.88it/s] 86%|████████▌ | 41254/48008 [2:21:41<29:10, 3.86it/s] 86%|████████▌ | 41255/48008 [2:21:41<33:27, 3.36it/s] 86%|████████▌ | 41256/48008 [2:21:42<29:56, 3.76it/s] 86%|████████▌ | 41257/48008 [2:21:42<27:13, 4.13it/s] 86%|████████▌ | 41258/48008 [2:21:42<25:20, 4.44it/s] 86%|████████▌ | 41259/48008 [2:21:42<24:06, 4.67it/s] 86%|████████▌ | 41260/48008 [2:21:42<23:24, 4.80it/s] 86%|████████▌ | 41261/48008 [2:21:42<22:46, 4.94it/s] 86%|████████▌ | 41262/48008 [2:21:43<22:31, 4.99it/s] 86%|████████▌ | 41263/48008 [2:21:43<26:11, 4.29it/s] 86%|████████▌ | 41264/48008 [2:21:43<25:21, 4.43it/s] 86%|████████▌ | 41265/48008 [2:21:43<24:20, 4.62it/s] 86%|████████▌ | 41266/48008 [2:21:44<23:22, 4.81it/s] 86%|████████▌ | 41267/48008 [2:21:44<23:25, 4.80it/s] 86%|████████▌ | 41268/48008 [2:21:44<22:57, 4.89it/s] 86%|████████▌ | 41269/48008 [2:21:44<22:24, 5.01it/s] 86%|████████▌ | 41270/48008 [2:21:44<22:09, 5.07it/s] 86%|████████▌ | 41271/48008 [2:21:45<45:23, 2.47it/s] 86%|████████▌ | 41272/48008 [2:21:45<38:09, 2.94it/s] 86%|████████▌ | 41273/48008 [2:21:46<33:08, 3.39it/s] 86%|████████▌ | 41274/48008 [2:21:46<29:28, 3.81it/s] 86%|████████▌ | 41275/48008 [2:21:46<27:26, 4.09it/s] 86%|████████▌ | 41276/48008 [2:21:46<25:34, 4.39it/s] 86%|████████▌ | 41277/48008 [2:21:46<24:20, 4.61it/s] 86%|████████▌ | 41278/48008 [2:21:47<23:28, 4.78it/s] 86%|████████▌ | 41279/48008 [2:21:47<22:37, 4.96it/s] 86%|████████▌ | 41280/48008 [2:21:47<22:21, 5.01it/s] 86%|████████▌ | 41281/48008 [2:21:47<22:14, 5.04it/s] 86%|████████▌ | 41282/48008 [2:21:47<22:03, 5.08it/s] 86%|████████▌ | 41283/48008 [2:21:48<21:47, 5.14it/s] 86%|████████▌ | 41284/48008 [2:21:48<21:41, 5.16it/s] 86%|████████▌ | 41285/48008 [2:21:48<21:36, 5.19it/s] 86%|████████▌ | 41286/48008 [2:21:48<21:28, 5.22it/s] 86%|████████▌ | 41287/48008 [2:21:48<21:27, 5.22it/s] 86%|████████▌ | 41288/48008 [2:21:48<21:17, 5.26it/s] 86%|████████▌ | 41289/48008 [2:21:49<21:41, 5.16it/s] 86%|████████▌ | 41290/48008 [2:21:49<21:37, 5.18it/s] 86%|████████▌ | 41291/48008 [2:21:49<21:52, 5.12it/s] 86%|████████▌ | 41292/48008 [2:21:49<21:40, 5.16it/s] 86%|████████▌ | 41293/48008 [2:21:49<21:36, 5.18it/s] 86%|████████▌ | 41294/48008 [2:21:50<22:01, 5.08it/s] 86%|████████▌ | 41295/48008 [2:21:50<22:10, 5.05it/s] 86%|████████▌ | 41296/48008 [2:21:50<21:55, 5.10it/s] 86%|████████▌ | 41297/48008 [2:21:50<21:49, 5.13it/s] 86%|████████▌ | 41298/48008 [2:21:50<21:26, 5.22it/s] 86%|████████▌ | 41299/48008 [2:21:51<21:39, 5.16it/s] 86%|████████▌ | 41300/48008 [2:21:51<28:06, 3.98it/s] {'loss': 0.3002, 'grad_norm': 0.19921875, 'learning_rate': 2.7949508415264126e-05, 'epoch': 0.86} 86%|████████▌ | 41300/48008 [2:21:51<28:06, 3.98it/s] 86%|████████▌ | 41301/48008 [2:21:51<26:03, 4.29it/s] 86%|████████▌ | 41302/48008 [2:21:51<24:40, 4.53it/s] 86%|████████▌ | 41303/48008 [2:21:52<24:03, 4.65it/s] 86%|████████▌ | 41304/48008 [2:21:52<23:10, 4.82it/s] 86%|████████▌ | 41305/48008 [2:21:52<22:39, 4.93it/s] 86%|████████▌ | 41306/48008 [2:21:52<21:53, 5.10it/s] 86%|████████▌ | 41307/48008 [2:21:52<21:56, 5.09it/s] 86%|████████▌ | 41308/48008 [2:21:53<22:07, 5.05it/s] 86%|████████▌ | 41309/48008 [2:21:53<21:54, 5.10it/s] 86%|████████▌ | 41310/48008 [2:21:53<21:47, 5.12it/s] 86%|████████▌ | 41311/48008 [2:21:53<22:12, 5.02it/s] 86%|████████▌ | 41312/48008 [2:21:53<22:06, 5.05it/s] 86%|████████▌ | 41313/48008 [2:21:54<22:17, 5.00it/s] 86%|████████▌ | 41314/48008 [2:21:54<21:57, 5.08it/s] 86%|████████▌ | 41315/48008 [2:21:54<21:48, 5.11it/s] 86%|████████▌ | 41316/48008 [2:21:54<22:13, 5.02it/s] 86%|████████▌ | 41317/48008 [2:21:54<21:52, 5.10it/s] 86%|████████▌ | 41318/48008 [2:21:55<21:36, 5.16it/s] 86%|████████▌ | 41319/48008 [2:21:55<21:24, 5.21it/s] 86%|████████▌ | 41320/48008 [2:21:55<21:12, 5.25it/s] 86%|████████▌ | 41321/48008 [2:21:55<21:20, 5.22it/s] 86%|████████▌ | 41322/48008 [2:21:55<21:14, 5.25it/s] 86%|████████▌ | 41323/48008 [2:21:55<21:05, 5.28it/s] 86%|████████▌ | 41324/48008 [2:21:56<21:22, 5.21it/s] 86%|████████▌ | 41325/48008 [2:21:56<21:12, 5.25it/s] 86%|████████▌ | 41326/48008 [2:21:56<21:14, 5.24it/s] 86%|████████▌ | 41327/48008 [2:21:56<25:04, 4.44it/s] 86%|████████▌ | 41328/48008 [2:21:57<23:56, 4.65it/s] 86%|████████▌ | 41329/48008 [2:21:57<23:26, 4.75it/s] 86%|████████▌ | 41330/48008 [2:21:57<23:08, 4.81it/s] 86%|████████▌ | 41331/48008 [2:21:57<22:36, 4.92it/s] 86%|████████▌ | 41332/48008 [2:21:57<22:03, 5.04it/s] 86%|████████▌ | 41333/48008 [2:21:58<21:47, 5.10it/s] 86%|████████▌ | 41334/48008 [2:21:58<21:29, 5.18it/s] 86%|████████▌ | 41335/48008 [2:21:58<21:18, 5.22it/s] 86%|████████▌ | 41336/48008 [2:21:58<21:39, 5.13it/s] 86%|████████▌ | 41337/48008 [2:21:58<21:47, 5.10it/s] 86%|████████▌ | 41338/48008 [2:21:58<21:40, 5.13it/s] 86%|████████▌ | 41339/48008 [2:21:59<21:40, 5.13it/s] 86%|████████▌ | 41340/48008 [2:21:59<21:23, 5.19it/s] 86%|████████▌ | 41341/48008 [2:21:59<21:20, 5.21it/s] 86%|████████▌ | 41342/48008 [2:21:59<21:34, 5.15it/s] 86%|████████▌ | 41343/48008 [2:21:59<21:32, 5.16it/s] 86%|████████▌ | 41344/48008 [2:22:00<21:22, 5.19it/s] 86%|████████▌ | 41345/48008 [2:22:00<21:21, 5.20it/s] 86%|████████▌ | 41346/48008 [2:22:00<21:21, 5.20it/s] 86%|████████▌ | 41347/48008 [2:22:00<21:23, 5.19it/s] 86%|████████▌ | 41348/48008 [2:22:00<21:34, 5.15it/s] 86%|████████▌ | 41349/48008 [2:22:01<21:54, 5.06it/s] 86%|████████▌ | 41350/48008 [2:22:01<21:44, 5.11it/s] 86%|████████▌ | 41350/48008 [2:22:01<21:44, 5.11it/s]{'loss': 0.2996, 'grad_norm': 0.14453125, 'learning_rate': 2.774120979836694e-05, 'epoch': 0.86} 86%|████████▌ | 41351/48008 [2:22:01<21:34, 5.14it/s] 86%|████████▌ | 41352/48008 [2:22:01<21:27, 5.17it/s] 86%|████████▌ | 41353/48008 [2:22:01<21:23, 5.19it/s] 86%|████████▌ | 41354/48008 [2:22:02<21:27, 5.17it/s] 86%|████████▌ | 41355/48008 [2:22:02<21:20, 5.20it/s] 86%|████████▌ | 41356/48008 [2:22:02<21:20, 5.19it/s] 86%|████████▌ | 41357/48008 [2:22:02<21:18, 5.20it/s] 86%|████████▌ | 41358/48008 [2:22:02<21:36, 5.13it/s] 86%|████████▌ | 41359/48008 [2:22:03<21:28, 5.16it/s] 86%|████████▌ | 41360/48008 [2:22:03<21:34, 5.14it/s] 86%|████████▌ | 41361/48008 [2:22:03<21:27, 5.16it/s] 86%|████████▌ | 41362/48008 [2:22:03<21:31, 5.14it/s] 86%|████████▌ | 41363/48008 [2:22:03<21:24, 5.17it/s] 86%|████████▌ | 41364/48008 [2:22:04<21:20, 5.19it/s] 86%|████████▌ | 41365/48008 [2:22:04<21:03, 5.26it/s] 86%|████████▌ | 41366/48008 [2:22:04<20:58, 5.28it/s] 86%|████████▌ | 41367/48008 [2:22:04<21:02, 5.26it/s] 86%|████████▌ | 41368/48008 [2:22:04<21:05, 5.25it/s] 86%|████████▌ | 41369/48008 [2:22:04<20:54, 5.29it/s] 86%|████████▌ | 41370/48008 [2:22:05<20:59, 5.27it/s] 86%|████████▌ | 41371/48008 [2:22:05<22:30, 4.91it/s] 86%|████████▌ | 41372/48008 [2:22:05<22:07, 5.00it/s] 86%|████████▌ | 41373/48008 [2:22:05<21:37, 5.12it/s] 86%|████████▌ | 41374/48008 [2:22:05<22:07, 5.00it/s] 86%|████████▌ | 41375/48008 [2:22:06<21:46, 5.08it/s] 86%|████████▌ | 41376/48008 [2:22:06<21:37, 5.11it/s] 86%|████████▌ | 41377/48008 [2:22:06<21:25, 5.16it/s] 86%|████████▌ | 41378/48008 [2:22:06<21:20, 5.18it/s] 86%|████████▌ | 41379/48008 [2:22:06<21:27, 5.15it/s] 86%|████████▌ | 41380/48008 [2:22:07<21:12, 5.21it/s] 86%|████████▌ | 41381/48008 [2:22:07<22:40, 4.87it/s] 86%|████████▌ | 41382/48008 [2:22:07<22:07, 4.99it/s] 86%|████████▌ | 41383/48008 [2:22:07<21:38, 5.10it/s] 86%|████████▌ | 41384/48008 [2:22:08<27:57, 3.95it/s] 86%|████████▌ | 41385/48008 [2:22:08<26:09, 4.22it/s] 86%|████████▌ | 41386/48008 [2:22:08<24:26, 4.52it/s] 86%|████████▌ | 41387/48008 [2:22:08<23:19, 4.73it/s] 86%|████████▌ | 41388/48008 [2:22:08<22:59, 4.80it/s] 86%|████████▌ | 41389/48008 [2:22:09<22:27, 4.91it/s] 86%|████████▌ | 41390/48008 [2:22:09<21:58, 5.02it/s] 86%|████████▌ | 41391/48008 [2:22:09<21:49, 5.05it/s] 86%|████████▌ | 41392/48008 [2:22:09<21:58, 5.02it/s] 86%|████████▌ | 41393/48008 [2:22:09<21:37, 5.10it/s] 86%|████████▌ | 41394/48008 [2:22:10<21:15, 5.19it/s] 86%|████████▌ | 41395/48008 [2:22:10<20:56, 5.26it/s] 86%|████████▌ | 41396/48008 [2:22:10<21:19, 5.17it/s] 86%|████████▌ | 41397/48008 [2:22:10<21:03, 5.23it/s] 86%|████████▌ | 41398/48008 [2:22:10<21:43, 5.07it/s] 86%|████████▌ | 41399/48008 [2:22:11<21:25, 5.14it/s] 86%|████████▌ | 41400/48008 [2:22:11<21:12, 5.19it/s] {'loss': 0.302, 'grad_norm': 0.150390625, 'learning_rate': 2.7532911181469757e-05, 'epoch': 0.86} 86%|████████▌ | 41400/48008 [2:22:11<21:12, 5.19it/s] 86%|████████▌ | 41401/48008 [2:22:11<21:06, 5.22it/s] 86%|████████▌ | 41402/48008 [2:22:11<21:10, 5.20it/s] 86%|████████▌ | 41403/48008 [2:22:11<21:15, 5.18it/s] 86%|████████▌ | 41404/48008 [2:22:11<21:15, 5.18it/s] 86%|████████▌ | 41405/48008 [2:22:12<21:16, 5.17it/s] 86%|████████▌ | 41406/48008 [2:22:12<21:18, 5.17it/s] 86%|████████▋ | 41407/48008 [2:22:12<21:15, 5.18it/s] 86%|████████▋ | 41408/48008 [2:22:12<21:10, 5.20it/s] 86%|████████▋ | 41409/48008 [2:22:12<21:22, 5.14it/s] 86%|████████▋ | 41410/48008 [2:22:13<21:15, 5.17it/s] 86%|████████▋ | 41411/48008 [2:22:13<21:01, 5.23it/s] 86%|████████▋ | 41412/48008 [2:22:13<21:00, 5.23it/s] 86%|████████▋ | 41413/48008 [2:22:13<21:03, 5.22it/s] 86%|████████▋ | 41414/48008 [2:22:13<21:03, 5.22it/s] 86%|████████▋ | 41415/48008 [2:22:14<21:25, 5.13it/s] 86%|████████▋ | 41416/48008 [2:22:14<21:12, 5.18it/s] 86%|████████▋ | 41417/48008 [2:22:14<27:37, 3.98it/s] 86%|████████▋ | 41418/48008 [2:22:14<25:39, 4.28it/s] 86%|████████▋ | 41419/48008 [2:22:15<24:18, 4.52it/s] 86%|████████▋ | 41420/48008 [2:22:15<27:05, 4.05it/s] 86%|████████▋ | 41421/48008 [2:22:15<25:39, 4.28it/s] 86%|████████▋ | 41422/48008 [2:22:15<24:15, 4.52it/s] 86%|████████▋ | 41423/48008 [2:22:15<23:12, 4.73it/s] 86%|████████▋ | 41424/48008 [2:22:16<22:30, 4.88it/s] 86%|████████▋ | 41425/48008 [2:22:16<22:03, 4.98it/s] 86%|████████▋ | 41426/48008 [2:22:16<21:31, 5.10it/s] 86%|████████▋ | 41427/48008 [2:22:16<21:32, 5.09it/s] 86%|████████▋ | 41428/48008 [2:22:16<21:33, 5.09it/s] 86%|████████▋ | 41429/48008 [2:22:17<21:47, 5.03it/s] 86%|████████▋ | 41430/48008 [2:22:17<21:49, 5.02it/s] 86%|████████▋ | 41431/48008 [2:22:17<21:29, 5.10it/s] 86%|████████▋ | 41432/48008 [2:22:17<21:13, 5.16it/s] 86%|████████▋ | 41433/48008 [2:22:17<21:10, 5.18it/s] 86%|████████▋ | 41434/48008 [2:22:18<21:13, 5.16it/s] 86%|████████▋ | 41435/48008 [2:22:18<21:01, 5.21it/s] 86%|████████▋ | 41436/48008 [2:22:18<20:47, 5.27it/s] 86%|████████▋ | 41437/48008 [2:22:18<20:58, 5.22it/s] 86%|████████▋ | 41438/48008 [2:22:19<27:23, 4.00it/s] 86%|████████▋ | 41439/48008 [2:22:19<25:18, 4.33it/s] 86%|████████▋ | 41440/48008 [2:22:19<23:38, 4.63it/s] 86%|████████▋ | 41441/48008 [2:22:19<29:15, 3.74it/s] 86%|████████▋ | 41442/48008 [2:22:19<27:09, 4.03it/s] 86%|████████▋ | 41443/48008 [2:22:20<25:10, 4.35it/s] 86%|████████▋ | 41444/48008 [2:22:21<55:55, 1.96it/s] 86%|████████▋ | 41445/48008 [2:22:21<46:57, 2.33it/s] 86%|████████▋ | 41446/48008 [2:22:21<38:58, 2.81it/s] 86%|████████▋ | 41447/48008 [2:22:22<39:56, 2.74it/s] 86%|████████▋ | 41448/48008 [2:22:22<34:15, 3.19it/s] 86%|████████▋ | 41449/48008 [2:22:22<30:17, 3.61it/s] 86%|████████▋ | 41450/48008 [2:22:22<27:42, 3.95it/s] {'loss': 0.2981, 'grad_norm': 0.1728515625, 'learning_rate': 2.732461256457257e-05, 'epoch': 0.86} 86%|████████▋ | 41450/48008 [2:22:22<27:42, 3.95it/s] 86%|████████▋ | 41451/48008 [2:22:22<26:05, 4.19it/s] 86%|████████▋ | 41452/48008 [2:22:23<24:16, 4.50it/s] 86%|████████▋ | 41453/48008 [2:22:23<23:47, 4.59it/s] 86%|████████▋ | 41454/48008 [2:22:23<22:44, 4.80it/s] 86%|████████▋ | 41455/48008 [2:22:23<22:23, 4.88it/s] 86%|████████▋ | 41456/48008 [2:22:23<21:46, 5.01it/s] 86%|████████▋ | 41457/48008 [2:22:24<21:30, 5.08it/s] 86%|████████▋ | 41458/48008 [2:22:24<43:37, 2.50it/s] 86%|████████▋ | 41459/48008 [2:22:25<37:01, 2.95it/s] 86%|████████▋ | 41460/48008 [2:22:25<32:08, 3.40it/s] 86%|████████▋ | 41461/48008 [2:22:25<28:36, 3.81it/s] 86%|████████▋ | 41462/48008 [2:22:25<26:12, 4.16it/s] 86%|████████▋ | 41463/48008 [2:22:25<24:34, 4.44it/s] 86%|████████▋ | 41464/48008 [2:22:26<23:40, 4.61it/s] 86%|████████▋ | 41465/48008 [2:22:26<22:50, 4.77it/s] 86%|████████▋ | 41466/48008 [2:22:26<22:37, 4.82it/s] 86%|████████▋ | 41467/48008 [2:22:26<22:37, 4.82it/s] 86%|████████▋ | 41468/48008 [2:22:26<22:06, 4.93it/s] 86%|████████▋ | 41469/48008 [2:22:27<21:41, 5.02it/s] 86%|████████▋ | 41470/48008 [2:22:27<42:35, 2.56it/s] 86%|████████▋ | 41471/48008 [2:22:28<36:04, 3.02it/s] 86%|████████▋ | 41472/48008 [2:22:28<31:26, 3.46it/s] 86%|████████▋ | 41473/48008 [2:22:28<28:14, 3.86it/s] 86%|████████▋ | 41474/48008 [2:22:28<26:18, 4.14it/s] 86%|████████▋ | 41475/48008 [2:22:28<24:38, 4.42it/s] 86%|████████▋ | 41476/48008 [2:22:29<49:07, 2.22it/s] 86%|████████▋ | 41477/48008 [2:22:30<40:37, 2.68it/s] 86%|████████▋ | 41478/48008 [2:22:30<34:41, 3.14it/s] 86%|████████▋ | 41479/48008 [2:22:30<31:58, 3.40it/s] 86%|████████▋ | 41480/48008 [2:22:30<28:33, 3.81it/s] 86%|████████▋ | 41481/48008 [2:22:30<26:00, 4.18it/s] 86%|████████▋ | 41482/48008 [2:22:31<45:02, 2.42it/s] 86%|████████▋ | 41483/48008 [2:22:31<37:46, 2.88it/s] 86%|████████▋ | 41484/48008 [2:22:32<32:42, 3.32it/s] 86%|████████▋ | 41485/48008 [2:22:32<28:59, 3.75it/s] 86%|████████▋ | 41486/48008 [2:22:32<28:00, 3.88it/s] 86%|████████▋ | 41487/48008 [2:22:32<25:53, 4.20it/s] 86%|████████▋ | 41488/48008 [2:22:32<24:33, 4.42it/s] 86%|████████▋ | 41489/48008 [2:22:33<23:26, 4.64it/s] 86%|████████▋ | 41490/48008 [2:22:33<22:35, 4.81it/s] 86%|████████▋ | 41491/48008 [2:22:33<22:04, 4.92it/s] 86%|████████▋ | 41492/48008 [2:22:33<21:48, 4.98it/s] 86%|████████▋ | 41493/48008 [2:22:33<25:15, 4.30it/s] 86%|████████▋ | 41494/48008 [2:22:34<23:48, 4.56it/s] 86%|████████▋ | 41495/48008 [2:22:34<22:49, 4.76it/s] 86%|████████▋ | 41496/48008 [2:22:34<22:31, 4.82it/s] 86%|████████▋ | 41497/48008 [2:22:34<21:57, 4.94it/s] 86%|████████▋ | 41498/48008 [2:22:34<21:32, 5.04it/s] 86%|████████▋ | 41499/48008 [2:22:35<21:13, 5.11it/s] 86%|████████▋ | 41500/48008 [2:22:35<21:40, 5.00it/s] {'loss': 0.3029, 'grad_norm': 0.18359375, 'learning_rate': 2.7116313947675388e-05, 'epoch': 0.86} 86%|████████▋ | 41500/48008 [2:22:35<21:40, 5.00it/s] 86%|████████▋ | 41501/48008 [2:22:35<22:07, 4.90it/s] 86%|████████▋ | 41502/48008 [2:22:35<21:40, 5.00it/s] 86%|████████▋ | 41503/48008 [2:22:35<21:42, 4.99it/s] 86%|████████▋ | 41504/48008 [2:22:36<21:37, 5.01it/s] 86%|████████▋ | 41505/48008 [2:22:36<21:14, 5.10it/s] 86%|████████▋ | 41506/48008 [2:22:36<21:14, 5.10it/s] 86%|████████▋ | 41507/48008 [2:22:36<21:05, 5.14it/s] 86%|████████▋ | 41508/48008 [2:22:36<21:08, 5.12it/s] 86%|████████▋ | 41509/48008 [2:22:37<21:01, 5.15it/s] 86%|████████▋ | 41510/48008 [2:22:37<20:50, 5.20it/s] 86%|████████▋ | 41511/48008 [2:22:37<20:44, 5.22it/s] 86%|████████▋ | 41512/48008 [2:22:37<20:44, 5.22it/s] 86%|████████▋ | 41513/48008 [2:22:37<20:46, 5.21it/s] 86%|████████▋ | 41514/48008 [2:22:38<20:41, 5.23it/s] 86%|████████▋ | 41515/48008 [2:22:38<20:32, 5.27it/s] 86%|████████▋ | 41516/48008 [2:22:38<20:34, 5.26it/s] 86%|████████▋ | 41517/48008 [2:22:38<20:59, 5.15it/s] 86%|████████▋ | 41518/48008 [2:22:38<20:55, 5.17it/s] 86%|████████▋ | 41519/48008 [2:22:39<21:14, 5.09it/s] 86%|████████▋ | 41520/48008 [2:22:39<21:02, 5.14it/s] 86%|████████▋ | 41521/48008 [2:22:39<20:53, 5.17it/s] 86%|████████▋ | 41522/48008 [2:22:39<22:23, 4.83it/s] 86%|████████▋ | 41523/48008 [2:22:39<22:05, 4.89it/s] 86%|████████▋ | 41524/48008 [2:22:40<21:39, 4.99it/s] 86%|████████▋ | 41525/48008 [2:22:40<21:21, 5.06it/s] 86%|████████▋ | 41526/48008 [2:22:40<22:39, 4.77it/s] 87%|████████▋ | 41527/48008 [2:22:40<22:03, 4.90it/s] 87%|████████▋ | 41528/48008 [2:22:40<22:15, 4.85it/s] 87%|████████▋ | 41529/48008 [2:22:41<21:44, 4.97it/s] 87%|████████▋ | 41530/48008 [2:22:41<41:55, 2.58it/s] 87%|████████▋ | 41531/48008 [2:22:42<35:33, 3.04it/s] 87%|████████▋ | 41532/48008 [2:22:42<31:16, 3.45it/s] 87%|████████▋ | 41533/48008 [2:22:42<28:01, 3.85it/s] 87%|████████▋ | 41534/48008 [2:22:42<25:51, 4.17it/s] 87%|████████▋ | 41535/48008 [2:22:42<24:24, 4.42it/s] 87%|████████▋ | 41536/48008 [2:22:43<23:15, 4.64it/s] 87%|████████▋ | 41537/48008 [2:22:43<22:27, 4.80it/s] 87%|████████▋ | 41538/48008 [2:22:43<21:53, 4.92it/s] 87%|████████▋ | 41539/48008 [2:22:43<21:49, 4.94it/s] 87%|████████▋ | 41540/48008 [2:22:43<21:30, 5.01it/s] 87%|████████▋ | 41541/48008 [2:22:43<21:12, 5.08it/s] 87%|████████▋ | 41542/48008 [2:22:44<21:06, 5.11it/s] 87%|████████▋ | 41543/48008 [2:22:44<20:53, 5.16it/s] 87%|████████▋ | 41544/48008 [2:22:44<20:52, 5.16it/s] 87%|████████▋ | 41545/48008 [2:22:44<20:49, 5.17it/s] 87%|████████▋ | 41546/48008 [2:22:44<20:44, 5.19it/s] 87%|████████▋ | 41547/48008 [2:22:45<20:39, 5.21it/s] 87%|████████▋ | 41548/48008 [2:22:45<20:42, 5.20it/s] 87%|████████▋ | 41549/48008 [2:22:45<20:41, 5.20it/s] 87%|████████▋ | 41550/48008 [2:22:45<20:29, 5.25it/s] {'loss': 0.3011, 'grad_norm': 0.421875, 'learning_rate': 2.6908015330778202e-05, 'epoch': 0.87} 87%|████████▋ | 41550/48008 [2:22:45<20:29, 5.25it/s] 87%|████████▋ | 41551/48008 [2:22:45<20:58, 5.13it/s] 87%|████████▋ | 41552/48008 [2:22:46<20:55, 5.14it/s] 87%|████████▋ | 41553/48008 [2:22:46<20:37, 5.22it/s] 87%|████████▋ | 41554/48008 [2:22:46<20:56, 5.13it/s] 87%|████████▋ | 41555/48008 [2:22:46<20:41, 5.20it/s] 87%|████████▋ | 41556/48008 [2:22:46<20:57, 5.13it/s] 87%|████████▋ | 41557/48008 [2:22:47<20:48, 5.17it/s] 87%|████████▋ | 41558/48008 [2:22:47<20:39, 5.20it/s] 87%|████████▋ | 41559/48008 [2:22:47<24:19, 4.42it/s] 87%|████████▋ | 41560/48008 [2:22:47<23:02, 4.66it/s] 87%|████████▋ | 41561/48008 [2:22:47<22:24, 4.80it/s] 87%|████████▋ | 41562/48008 [2:22:48<22:15, 4.83it/s] 87%|████████▋ | 41563/48008 [2:22:48<21:47, 4.93it/s] 87%|████████▋ | 41564/48008 [2:22:48<21:26, 5.01it/s] 87%|████████▋ | 41565/48008 [2:22:48<20:59, 5.11it/s] 87%|████████▋ | 41566/48008 [2:22:48<20:52, 5.14it/s] 87%|████████▋ | 41567/48008 [2:22:49<20:43, 5.18it/s] 87%|████████▋ | 41568/48008 [2:22:49<22:07, 4.85it/s] 87%|████████▋ | 41569/48008 [2:22:49<21:35, 4.97it/s] 87%|████████▋ | 41570/48008 [2:22:49<21:18, 5.04it/s] 87%|████████▋ | 41571/48008 [2:22:49<20:57, 5.12it/s] 87%|████████▋ | 41572/48008 [2:22:50<21:04, 5.09it/s] 87%|████████▋ | 41573/48008 [2:22:50<21:18, 5.03it/s] 87%|████████▋ | 41574/48008 [2:22:50<21:34, 4.97it/s] 87%|████████▋ | 41575/48008 [2:22:50<21:26, 5.00it/s] 87%|████████▋ | 41576/48008 [2:22:51<27:24, 3.91it/s] 87%|████████▋ | 41577/48008 [2:22:51<25:47, 4.16it/s] 87%|████████▋ | 41578/48008 [2:22:51<24:08, 4.44it/s] 87%|████████▋ | 41579/48008 [2:22:51<23:16, 4.60it/s] 87%|████████▋ | 41580/48008 [2:22:51<22:26, 4.77it/s] 87%|████████▋ | 41581/48008 [2:22:52<21:50, 4.91it/s] 87%|████████▋ | 41582/48008 [2:22:52<21:23, 5.01it/s] 87%|████████▋ | 41583/48008 [2:22:52<21:03, 5.08it/s] 87%|████████▋ | 41584/48008 [2:22:52<21:04, 5.08it/s] 87%|████████▋ | 41585/48008 [2:22:52<20:43, 5.17it/s] 87%|████████▋ | 41586/48008 [2:22:53<20:41, 5.17it/s] 87%|████████▋ | 41587/48008 [2:22:53<20:44, 5.16it/s] 87%|████████▋ | 41588/48008 [2:22:53<21:11, 5.05it/s] 87%|████████▋ | 41589/48008 [2:22:53<20:58, 5.10it/s] 87%|████████▋ | 41590/48008 [2:22:53<22:16, 4.80it/s] 87%|████████▋ | 41591/48008 [2:22:54<21:55, 4.88it/s] 87%|████████▋ | 41592/48008 [2:22:54<21:25, 4.99it/s] 87%|████████▋ | 41593/48008 [2:22:54<21:12, 5.04it/s] 87%|████████▋ | 41594/48008 [2:22:54<20:58, 5.10it/s] 87%|████████▋ | 41595/48008 [2:22:54<20:49, 5.13it/s] 87%|████████▋ | 41596/48008 [2:22:55<20:45, 5.15it/s] 87%|████████▋ | 41597/48008 [2:22:55<24:23, 4.38it/s] 87%|████████▋ | 41598/48008 [2:22:55<23:32, 4.54it/s] 87%|████████▋ | 41599/48008 [2:22:55<26:13, 4.07it/s] 87%|████████▋ | 41600/48008 [2:22:56<24:23, 4.38it/s] {'loss': 0.2973, 'grad_norm': 0.1982421875, 'learning_rate': 2.6699716713881023e-05, 'epoch': 0.87} 87%|████████▋ | 41600/48008 [2:22:56<24:23, 4.38it/s] 87%|████████▋ | 41601/48008 [2:22:56<23:44, 4.50it/s] 87%|████████▋ | 41602/48008 [2:22:56<22:47, 4.68it/s] 87%|████████▋ | 41603/48008 [2:22:56<22:14, 4.80it/s] 87%|████████▋ | 41604/48008 [2:22:56<21:43, 4.91it/s] 87%|████████▋ | 41605/48008 [2:22:57<21:42, 4.92it/s] 87%|████████▋ | 41606/48008 [2:22:57<21:19, 5.00it/s] 87%|████████▋ | 41607/48008 [2:22:57<21:19, 5.00it/s] 87%|████████▋ | 41608/48008 [2:22:57<21:38, 4.93it/s] 87%|████████▋ | 41609/48008 [2:22:57<21:16, 5.01it/s] 87%|████████▋ | 41610/48008 [2:22:58<21:14, 5.02it/s] 87%|████████▋ | 41611/48008 [2:22:58<20:52, 5.11it/s] 87%|████████▋ | 41612/48008 [2:22:58<21:06, 5.05it/s] 87%|████████▋ | 41613/48008 [2:22:58<20:53, 5.10it/s] 87%|████████▋ | 41614/48008 [2:22:58<20:36, 5.17it/s] 87%|████████▋ | 41615/48008 [2:22:58<20:26, 5.21it/s] 87%|████████▋ | 41616/48008 [2:22:59<20:30, 5.20it/s] 87%|████████▋ | 41617/48008 [2:22:59<20:35, 5.17it/s] 87%|████████▋ | 41618/48008 [2:22:59<20:37, 5.17it/s] 87%|████████▋ | 41619/48008 [2:22:59<20:40, 5.15it/s] 87%|████████▋ | 41620/48008 [2:22:59<20:54, 5.09it/s] 87%|████████▋ | 41621/48008 [2:23:00<22:12, 4.79it/s] 87%|████████▋ | 41622/48008 [2:23:00<21:37, 4.92it/s] 87%|████████▋ | 41623/48008 [2:23:00<21:10, 5.02it/s] 87%|████████▋ | 41624/48008 [2:23:00<21:03, 5.05it/s] 87%|████████▋ | 41625/48008 [2:23:01<24:27, 4.35it/s] 87%|████████▋ | 41626/48008 [2:23:01<23:14, 4.58it/s] 87%|████████▋ | 41627/48008 [2:23:01<22:19, 4.77it/s] 87%|████████▋ | 41628/48008 [2:23:01<25:28, 4.17it/s] 87%|████████▋ | 41629/48008 [2:23:01<24:03, 4.42it/s] 87%|████████▋ | 41630/48008 [2:23:02<22:50, 4.65it/s] 87%|████████▋ | 41631/48008 [2:23:02<22:28, 4.73it/s] 87%|████████▋ | 41632/48008 [2:23:02<25:29, 4.17it/s] 87%|████████▋ | 41633/48008 [2:23:02<23:54, 4.44it/s] 87%|████████▋ | 41634/48008 [2:23:03<22:49, 4.65it/s] 87%|████████▋ | 41635/48008 [2:23:03<21:57, 4.84it/s] 87%|████████▋ | 41636/48008 [2:23:03<21:28, 4.95it/s] 87%|████████▋ | 41637/48008 [2:23:03<21:04, 5.04it/s] 87%|████████▋ | 41638/48008 [2:23:03<20:42, 5.13it/s] 87%|████████▋ | 41639/48008 [2:23:03<20:37, 5.15it/s] 87%|████████▋ | 41640/48008 [2:23:04<20:28, 5.18it/s] 87%|████████▋ | 41641/48008 [2:23:04<20:32, 5.16it/s] 87%|████████▋ | 41642/48008 [2:23:04<20:22, 5.21it/s] 87%|████████▋ | 41643/48008 [2:23:04<20:19, 5.22it/s] 87%|████████▋ | 41644/48008 [2:23:04<20:14, 5.24it/s] 87%|████████▋ | 41645/48008 [2:23:05<26:25, 4.01it/s] 87%|████████▋ | 41646/48008 [2:23:05<24:31, 4.32it/s] 87%|████████▋ | 41647/48008 [2:23:05<23:20, 4.54it/s] 87%|████████▋ | 41648/48008 [2:23:05<22:29, 4.71it/s] 87%|████████▋ | 41649/48008 [2:23:06<23:16, 4.55it/s] 87%|████████▋ | 41650/48008 [2:23:06<22:20, 4.74it/s] {'loss': 0.2984, 'grad_norm': 0.228515625, 'learning_rate': 2.649141809698384e-05, 'epoch': 0.87} 87%|████████▋ | 41650/48008 [2:23:06<22:20, 4.74it/s] 87%|████████▋ | 41651/48008 [2:23:06<21:57, 4.82it/s] 87%|████████▋ | 41652/48008 [2:23:06<21:27, 4.93it/s] 87%|████████▋ | 41653/48008 [2:23:06<21:10, 5.00it/s] 87%|████████▋ | 41654/48008 [2:23:07<20:54, 5.07it/s] 87%|████████▋ | 41655/48008 [2:23:07<20:41, 5.12it/s] 87%|████████▋ | 41656/48008 [2:23:07<20:44, 5.10it/s] 87%|████████▋ | 41657/48008 [2:23:07<20:40, 5.12it/s] 87%|████████▋ | 41658/48008 [2:23:07<21:03, 5.03it/s] 87%|████████▋ | 41659/48008 [2:23:08<20:47, 5.09it/s] 87%|████████▋ | 41660/48008 [2:23:08<20:56, 5.05it/s] 87%|████████▋ | 41661/48008 [2:23:08<24:19, 4.35it/s] 87%|████████▋ | 41662/48008 [2:23:08<23:03, 4.59it/s] 87%|████████▋ | 41663/48008 [2:23:08<22:13, 4.76it/s] 87%|████████▋ | 41664/48008 [2:23:09<21:24, 4.94it/s] 87%|████████▋ | 41665/48008 [2:23:09<20:54, 5.06it/s] 87%|████████▋ | 41666/48008 [2:23:09<21:13, 4.98it/s] 87%|████████▋ | 41667/48008 [2:23:09<20:55, 5.05it/s] 87%|████████▋ | 41668/48008 [2:23:09<20:29, 5.16it/s] 87%|████████▋ | 41669/48008 [2:23:10<20:16, 5.21it/s] 87%|████████▋ | 41670/48008 [2:23:10<20:21, 5.19it/s] 87%|████████▋ | 41671/48008 [2:23:10<20:16, 5.21it/s] 87%|████████▋ | 41672/48008 [2:23:10<20:36, 5.12it/s] 87%|████████▋ | 41673/48008 [2:23:10<20:53, 5.05it/s] 87%|████████▋ | 41674/48008 [2:23:11<20:39, 5.11it/s] 87%|████████▋ | 41675/48008 [2:23:11<21:59, 4.80it/s] 87%|████████▋ | 41676/48008 [2:23:11<21:51, 4.83it/s] 87%|████████▋ | 41677/48008 [2:23:11<21:41, 4.86it/s] 87%|████████▋ | 41678/48008 [2:23:11<21:14, 4.97it/s] 87%|████████▋ | 41679/48008 [2:23:12<20:54, 5.04it/s] 87%|████████▋ | 41680/48008 [2:23:12<20:41, 5.10it/s] 87%|████████▋ | 41681/48008 [2:23:12<20:23, 5.17it/s] 87%|████████▋ | 41682/48008 [2:23:12<20:43, 5.09it/s] 87%|████████▋ | 41683/48008 [2:23:12<20:31, 5.14it/s] 87%|████████▋ | 41684/48008 [2:23:13<20:21, 5.18it/s] 87%|████████▋ | 41685/48008 [2:23:13<20:17, 5.19it/s] 87%|████████▋ | 41686/48008 [2:23:13<20:10, 5.22it/s] 87%|████████▋ | 41687/48008 [2:23:13<20:15, 5.20it/s] 87%|████████▋ | 41688/48008 [2:23:13<20:08, 5.23it/s] 87%|████████▋ | 41689/48008 [2:23:14<20:00, 5.26it/s] 87%|████████▋ | 41690/48008 [2:23:14<19:58, 5.27it/s] 87%|████████▋ | 41691/48008 [2:23:14<19:52, 5.30it/s] 87%|████████▋ | 41692/48008 [2:23:14<19:58, 5.27it/s] 87%|████████▋ | 41693/48008 [2:23:14<19:57, 5.27it/s] 87%|████████▋ | 41694/48008 [2:23:14<20:00, 5.26it/s] 87%|████████▋ | 41695/48008 [2:23:15<20:03, 5.24it/s] 87%|████████▋ | 41696/48008 [2:23:15<20:00, 5.26it/s] 87%|████████▋ | 41697/48008 [2:23:15<20:02, 5.25it/s] 87%|████████▋ | 41698/48008 [2:23:15<19:55, 5.28it/s] 87%|████████▋ | 41699/48008 [2:23:15<19:54, 5.28it/s] 87%|████████▋ | 41700/48008 [2:23:16<20:00, 5.25it/s] 87%|████████▋ | 41700/48008 [2:23:16<20:00, 5.25it/s]{'loss': 0.3006, 'grad_norm': 0.1796875, 'learning_rate': 2.6283119480086654e-05, 'epoch': 0.87} 87%|████████▋ | 41701/48008 [2:23:16<20:27, 5.14it/s] 87%|████████▋ | 41702/48008 [2:23:16<20:39, 5.09it/s] 87%|████████▋ | 41703/48008 [2:23:16<20:27, 5.13it/s] 87%|████████▋ | 41704/48008 [2:23:16<20:33, 5.11it/s] 87%|████████▋ | 41705/48008 [2:23:17<21:00, 5.00it/s] 87%|████████▋ | 41706/48008 [2:23:17<20:40, 5.08it/s] 87%|████████▋ | 41707/48008 [2:23:17<20:33, 5.11it/s] 87%|████████▋ | 41708/48008 [2:23:17<20:35, 5.10it/s] 87%|████████▋ | 41709/48008 [2:23:17<20:37, 5.09it/s] 87%|████████▋ | 41710/48008 [2:23:18<20:45, 5.06it/s] 87%|████████▋ | 41711/48008 [2:23:18<21:59, 4.77it/s] 87%|████████▋ | 41712/48008 [2:23:18<21:27, 4.89it/s] 87%|████████▋ | 41713/48008 [2:23:18<21:33, 4.87it/s] 87%|████████▋ | 41714/48008 [2:23:18<20:54, 5.02it/s] 87%|████████▋ | 41715/48008 [2:23:19<22:04, 4.75it/s] 87%|████████▋ | 41716/48008 [2:23:19<21:21, 4.91it/s] 87%|████████▋ | 41717/48008 [2:23:19<20:53, 5.02it/s] 87%|████████▋ | 41718/48008 [2:23:19<20:37, 5.08it/s] 87%|████████▋ | 41719/48008 [2:23:19<20:14, 5.18it/s] 87%|████████▋ | 41720/48008 [2:23:20<20:09, 5.20it/s] 87%|████████▋ | 41721/48008 [2:23:20<20:18, 5.16it/s] 87%|████████▋ | 41722/48008 [2:23:20<20:35, 5.09it/s] 87%|████████▋ | 41723/48008 [2:23:20<20:50, 5.02it/s] 87%|████████▋ | 41724/48008 [2:23:20<20:29, 5.11it/s] 87%|████████▋ | 41725/48008 [2:23:21<20:39, 5.07it/s] 87%|████████▋ | 41726/48008 [2:23:21<20:19, 5.15it/s] 87%|████████▋ | 41727/48008 [2:23:21<20:06, 5.21it/s] 87%|████████▋ | 41728/48008 [2:23:21<20:14, 5.17it/s] 87%|████████▋ | 41729/48008 [2:23:21<23:48, 4.40it/s] 87%|████████▋ | 41730/48008 [2:23:22<23:13, 4.51it/s] 87%|████████▋ | 41731/48008 [2:23:22<25:51, 4.05it/s] 87%|████████▋ | 41732/48008 [2:23:22<23:54, 4.37it/s] 87%|████████▋ | 41733/48008 [2:23:22<22:43, 4.60it/s] 87%|████████▋ | 41734/48008 [2:23:23<22:13, 4.71it/s] 87%|████████▋ | 41735/48008 [2:23:23<21:54, 4.77it/s] 87%|████████▋ | 41736/48008 [2:23:23<21:13, 4.92it/s] 87%|████████▋ | 41737/48008 [2:23:23<24:29, 4.27it/s] 87%|████████▋ | 41738/48008 [2:23:23<23:04, 4.53it/s] 87%|████████▋ | 41739/48008 [2:23:24<23:34, 4.43it/s] 87%|████████▋ | 41740/48008 [2:23:24<22:53, 4.56it/s] 87%|████████▋ | 41741/48008 [2:23:24<22:23, 4.67it/s] 87%|████████▋ | 41742/48008 [2:23:24<21:51, 4.78it/s] 87%|████████▋ | 41743/48008 [2:23:24<21:11, 4.93it/s] 87%|████████▋ | 41744/48008 [2:23:25<22:17, 4.69it/s] 87%|████████▋ | 41745/48008 [2:23:25<21:29, 4.86it/s] 87%|████████▋ | 41746/48008 [2:23:25<21:07, 4.94it/s] 87%|████████▋ | 41747/48008 [2:23:25<20:42, 5.04it/s] 87%|████████▋ | 41748/48008 [2:23:25<20:20, 5.13it/s] 87%|████████▋ | 41749/48008 [2:23:26<20:10, 5.17it/s] 87%|████████▋ | 41750/48008 [2:23:26<20:17, 5.14it/s] {'loss': 0.2973, 'grad_norm': 0.2041015625, 'learning_rate': 2.607482086318947e-05, 'epoch': 0.87} 87%|████████▋ | 41750/48008 [2:23:26<20:17, 5.14it/s] 87%|████████▋ | 41751/48008 [2:23:26<20:03, 5.20it/s] 87%|████████▋ | 41752/48008 [2:23:26<20:01, 5.21it/s] 87%|████████▋ | 41753/48008 [2:23:27<26:04, 4.00it/s] 87%|████████▋ | 41754/48008 [2:23:27<24:12, 4.31it/s] 87%|████████▋ | 41755/48008 [2:23:27<22:59, 4.53it/s] 87%|████████▋ | 41756/48008 [2:23:27<22:04, 4.72it/s] 87%|████████▋ | 41757/48008 [2:23:27<21:38, 4.81it/s] 87%|████████▋ | 41758/48008 [2:23:28<21:13, 4.91it/s] 87%|████████▋ | 41759/48008 [2:23:28<20:47, 5.01it/s] 87%|████████▋ | 41760/48008 [2:23:28<20:43, 5.03it/s] 87%|████████▋ | 41761/48008 [2:23:28<20:24, 5.10it/s] 87%|████████▋ | 41762/48008 [2:23:28<20:18, 5.13it/s] 87%|████████▋ | 41763/48008 [2:23:29<20:20, 5.12it/s] 87%|████████▋ | 41764/48008 [2:23:29<20:23, 5.10it/s] 87%|████████▋ | 41765/48008 [2:23:29<20:21, 5.11it/s] 87%|████████▋ | 41766/48008 [2:23:29<20:23, 5.10it/s] 87%|████████▋ | 41767/48008 [2:23:29<20:15, 5.13it/s] 87%|████████▋ | 41768/48008 [2:23:30<20:24, 5.10it/s] 87%|████████▋ | 41769/48008 [2:23:30<20:18, 5.12it/s] 87%|████████▋ | 41770/48008 [2:23:30<20:18, 5.12it/s] 87%|████████▋ | 41771/48008 [2:23:30<20:07, 5.17it/s] 87%|████████▋ | 41772/48008 [2:23:30<20:01, 5.19it/s] 87%|████████▋ | 41773/48008 [2:23:31<20:15, 5.13it/s] 87%|████████▋ | 41774/48008 [2:23:31<26:14, 3.96it/s] 87%|████████▋ | 41775/48008 [2:23:31<24:08, 4.30it/s] 87%|████████▋ | 41776/48008 [2:23:31<22:49, 4.55it/s] 87%|████████▋ | 41777/48008 [2:23:31<21:52, 4.75it/s] 87%|████████▋ | 41778/48008 [2:23:32<21:10, 4.90it/s] 87%|████████▋ | 41779/48008 [2:23:32<20:47, 4.99it/s] 87%|████████▋ | 41780/48008 [2:23:32<21:03, 4.93it/s] 87%|████████▋ | 41781/48008 [2:23:32<20:38, 5.03it/s] 87%|████████▋ | 41782/48008 [2:23:32<20:16, 5.12it/s] 87%|████████▋ | 41783/48008 [2:23:33<20:11, 5.14it/s] 87%|████████▋ | 41784/48008 [2:23:33<20:09, 5.14it/s] 87%|████████▋ | 41785/48008 [2:23:33<20:16, 5.12it/s] 87%|████████▋ | 41786/48008 [2:23:33<19:58, 5.19it/s] 87%|████████▋ | 41787/48008 [2:23:33<20:04, 5.16it/s] 87%|████████▋ | 41788/48008 [2:23:34<20:13, 5.13it/s] 87%|████████▋ | 41789/48008 [2:23:34<20:07, 5.15it/s] 87%|████████▋ | 41790/48008 [2:23:34<19:53, 5.21it/s] 87%|████████▋ | 41791/48008 [2:23:34<19:56, 5.20it/s] 87%|████████▋ | 41792/48008 [2:23:34<20:00, 5.18it/s] 87%|████████▋ | 41793/48008 [2:23:35<19:50, 5.22it/s] 87%|████████▋ | 41794/48008 [2:23:35<20:03, 5.16it/s] 87%|████████▋ | 41795/48008 [2:23:35<20:01, 5.17it/s] 87%|████████▋ | 41796/48008 [2:23:35<19:49, 5.22it/s] 87%|████████▋ | 41797/48008 [2:23:35<19:59, 5.18it/s] 87%|████████▋ | 41798/48008 [2:23:36<20:23, 5.08it/s] 87%|████████▋ | 41799/48008 [2:23:36<20:36, 5.02it/s] 87%|████████▋ | 41800/48008 [2:23:36<24:02, 4.30it/s] {'loss': 0.2994, 'grad_norm': 0.3125, 'learning_rate': 2.5866522246292286e-05, 'epoch': 0.87} 87%|████████▋ | 41800/48008 [2:23:36<24:02, 4.30it/s] 87%|████████▋ | 41801/48008 [2:23:36<22:50, 4.53it/s] 87%|████████▋ | 41802/48008 [2:23:36<22:11, 4.66it/s] 87%|████████▋ | 41803/48008 [2:23:37<21:32, 4.80it/s] 87%|████████▋ | 41804/48008 [2:23:37<20:47, 4.97it/s] 87%|████████▋ | 41805/48008 [2:23:37<20:29, 5.05it/s] 87%|████████▋ | 41806/48008 [2:23:37<20:34, 5.02it/s] 87%|████████▋ | 41807/48008 [2:23:37<20:19, 5.09it/s] 87%|████████▋ | 41808/48008 [2:23:38<20:10, 5.12it/s] 87%|████████▋ | 41809/48008 [2:23:38<19:57, 5.18it/s] 87%|████████▋ | 41810/48008 [2:23:38<19:49, 5.21it/s] 87%|████████▋ | 41811/48008 [2:23:38<19:48, 5.21it/s] 87%|████████▋ | 41812/48008 [2:23:38<19:43, 5.24it/s] 87%|████████▋ | 41813/48008 [2:23:39<19:58, 5.17it/s] 87%|████████▋ | 41814/48008 [2:23:39<20:13, 5.10it/s] 87%|████████▋ | 41815/48008 [2:23:39<20:04, 5.14it/s] 87%|████████▋ | 41816/48008 [2:23:39<20:03, 5.15it/s] 87%|████████▋ | 41817/48008 [2:23:39<19:52, 5.19it/s] 87%|████████▋ | 41818/48008 [2:23:40<20:02, 5.15it/s] 87%|████████▋ | 41819/48008 [2:23:40<19:57, 5.17it/s] 87%|████████▋ | 41820/48008 [2:23:40<25:57, 3.97it/s] 87%|████████▋ | 41821/48008 [2:23:40<24:27, 4.22it/s] 87%|████████▋ | 41822/48008 [2:23:40<22:52, 4.51it/s] 87%|████████▋ | 41823/48008 [2:23:41<21:47, 4.73it/s] 87%|████████▋ | 41824/48008 [2:23:41<21:09, 4.87it/s] 87%|████████▋ | 41825/48008 [2:23:41<21:03, 4.90it/s] 87%|████████▋ | 41826/48008 [2:23:41<20:49, 4.95it/s] 87%|████████▋ | 41827/48008 [2:23:41<20:28, 5.03it/s] 87%|████████▋ | 41828/48008 [2:23:42<20:18, 5.07it/s] 87%|████████▋ | 41829/48008 [2:23:42<20:32, 5.01it/s] 87%|████████▋ | 41830/48008 [2:23:42<20:39, 4.98it/s] 87%|████████▋ | 41831/48008 [2:23:42<20:20, 5.06it/s] 87%|████████▋ | 41832/48008 [2:23:42<20:09, 5.11it/s] 87%|████████▋ | 41833/48008 [2:23:43<19:57, 5.16it/s] 87%|████████▋ | 41834/48008 [2:23:43<19:50, 5.19it/s] 87%|████████▋ | 41835/48008 [2:23:43<20:01, 5.14it/s] 87%|████████▋ | 41836/48008 [2:23:43<19:52, 5.18it/s] 87%|████████▋ | 41837/48008 [2:23:43<19:51, 5.18it/s] 87%|████████▋ | 41838/48008 [2:23:44<19:52, 5.17it/s] 87%|████████▋ | 41839/48008 [2:23:44<20:17, 5.07it/s] 87%|████████▋ | 41840/48008 [2:23:44<20:07, 5.11it/s] 87%|████████▋ | 41841/48008 [2:23:44<19:58, 5.14it/s] 87%|████████▋ | 41842/48008 [2:23:45<23:29, 4.38it/s] 87%|████████▋ | 41843/48008 [2:23:45<22:24, 4.59it/s] 87%|████████▋ | 41844/48008 [2:23:45<21:29, 4.78it/s] 87%|████████▋ | 41845/48008 [2:23:45<21:22, 4.81it/s] 87%|████████▋ | 41846/48008 [2:23:45<20:47, 4.94it/s] 87%|████████▋ | 41847/48008 [2:23:45<20:11, 5.08it/s] 87%|████████▋ | 41848/48008 [2:23:46<26:00, 3.95it/s] 87%|████████▋ | 41849/48008 [2:23:46<23:52, 4.30it/s] 87%|████████▋ | 41850/48008 [2:23:46<22:57, 4.47it/s] {'loss': 0.2997, 'grad_norm': 0.55078125, 'learning_rate': 2.56582236293951e-05, 'epoch': 0.87} 87%|████████▋ | 41850/48008 [2:23:46<22:57, 4.47it/s] 87%|████████▋ | 41851/48008 [2:23:46<21:58, 4.67it/s] 87%|████████▋ | 41852/48008 [2:23:47<21:48, 4.70it/s] 87%|████████▋ | 41853/48008 [2:23:47<21:05, 4.86it/s] 87%|████████▋ | 41854/48008 [2:23:47<20:41, 4.96it/s] 87%|████████▋ | 41855/48008 [2:23:47<20:32, 4.99it/s] 87%|████████▋ | 41856/48008 [2:23:47<21:38, 4.74it/s] 87%|████████▋ | 41857/48008 [2:23:48<21:12, 4.83it/s] 87%|████████▋ | 41858/48008 [2:23:48<20:53, 4.90it/s] 87%|████████▋ | 41859/48008 [2:23:48<20:23, 5.03it/s] 87%|████████▋ | 41860/48008 [2:23:48<20:06, 5.10it/s] 87%|████████▋ | 41861/48008 [2:23:49<25:56, 3.95it/s] 87%|████████▋ | 41862/48008 [2:23:49<23:59, 4.27it/s] 87%|████████▋ | 41863/48008 [2:23:49<22:37, 4.53it/s] 87%|████████▋ | 41864/48008 [2:23:49<21:31, 4.76it/s] 87%|████████▋ | 41865/48008 [2:23:49<24:34, 4.17it/s] 87%|████████▋ | 41866/48008 [2:23:50<23:02, 4.44it/s] 87%|████████▋ | 41867/48008 [2:23:50<23:28, 4.36it/s] 87%|████████▋ | 41868/48008 [2:23:50<22:17, 4.59it/s] 87%|████████▋ | 41869/48008 [2:23:50<21:28, 4.76it/s] 87%|████████▋ | 41870/48008 [2:23:50<20:55, 4.89it/s] 87%|████████▋ | 41871/48008 [2:23:51<20:27, 5.00it/s] 87%|████████▋ | 41872/48008 [2:23:51<20:01, 5.11it/s] 87%|████████▋ | 41873/48008 [2:23:51<20:28, 5.00it/s] 87%|████████▋ | 41874/48008 [2:23:51<20:08, 5.07it/s] 87%|████████▋ | 41875/48008 [2:23:51<19:50, 5.15it/s] 87%|████████▋ | 41876/48008 [2:23:52<21:10, 4.83it/s] 87%|████████▋ | 41877/48008 [2:23:52<20:50, 4.90it/s] 87%|████████▋ | 41878/48008 [2:23:52<20:38, 4.95it/s] 87%|████████▋ | 41879/48008 [2:23:52<20:12, 5.06it/s] 87%|████████▋ | 41880/48008 [2:23:52<20:01, 5.10it/s] 87%|████████▋ | 41881/48008 [2:23:53<20:12, 5.05it/s] 87%|████████▋ | 41882/48008 [2:23:53<20:11, 5.06it/s] 87%|████████▋ | 41883/48008 [2:23:53<19:49, 5.15it/s] 87%|████████▋ | 41884/48008 [2:23:53<19:46, 5.16it/s] 87%|████████▋ | 41885/48008 [2:23:53<20:07, 5.07it/s] 87%|████████▋ | 41886/48008 [2:23:54<19:56, 5.12it/s] 87%|████████▋ | 41887/48008 [2:23:54<19:50, 5.14it/s] 87%|████████▋ | 41888/48008 [2:23:54<19:45, 5.16it/s] 87%|████████▋ | 41889/48008 [2:23:54<19:37, 5.20it/s] 87%|████████▋ | 41890/48008 [2:23:54<19:35, 5.21it/s] 87%|████████▋ | 41891/48008 [2:23:55<25:31, 4.00it/s] 87%|████████▋ | 41892/48008 [2:23:55<23:42, 4.30it/s] 87%|████████▋ | 41893/48008 [2:23:55<22:45, 4.48it/s] 87%|████████▋ | 41894/48008 [2:23:55<21:41, 4.70it/s] 87%|████████▋ | 41895/48008 [2:23:56<21:01, 4.85it/s] 87%|████████▋ | 41896/48008 [2:23:56<20:35, 4.95it/s] 87%|████████▋ | 41897/48008 [2:23:56<20:34, 4.95it/s] 87%|████████▋ | 41898/48008 [2:23:56<20:07, 5.06it/s] 87%|████████▋ | 41899/48008 [2:23:56<19:46, 5.15it/s] 87%|████████▋ | 41900/48008 [2:23:57<19:44, 5.16it/s] {'loss': 0.2997, 'grad_norm': 0.15234375, 'learning_rate': 2.5449925012497917e-05, 'epoch': 0.87} 87%|████████▋ | 41900/48008 [2:23:57<19:44, 5.16it/s] 87%|████████▋ | 41901/48008 [2:23:57<19:33, 5.20it/s] 87%|████████▋ | 41902/48008 [2:23:57<19:53, 5.12it/s] 87%|████████▋ | 41903/48008 [2:23:57<23:16, 4.37it/s] 87%|████████▋ | 41904/48008 [2:23:57<23:30, 4.33it/s] 87%|████████▋ | 41905/48008 [2:23:58<22:15, 4.57it/s] 87%|████████▋ | 41906/48008 [2:23:58<21:27, 4.74it/s] 87%|████████▋ | 41907/48008 [2:23:58<20:49, 4.88it/s] 87%|████████▋ | 41908/48008 [2:23:58<20:26, 4.97it/s] 87%|████████▋ | 41909/48008 [2:23:58<20:02, 5.07it/s] 87%|████████▋ | 41910/48008 [2:23:59<20:08, 5.05it/s] 87%|████████▋ | 41911/48008 [2:23:59<19:50, 5.12it/s] 87%|████████▋ | 41912/48008 [2:23:59<19:40, 5.16it/s] 87%|████████▋ | 41913/48008 [2:23:59<19:34, 5.19it/s] 87%|████████▋ | 41914/48008 [2:23:59<19:32, 5.20it/s] 87%|████████▋ | 41915/48008 [2:24:00<19:31, 5.20it/s] 87%|████████▋ | 41916/48008 [2:24:00<20:03, 5.06it/s] 87%|████████▋ | 41917/48008 [2:24:00<19:50, 5.11it/s] 87%|████████▋ | 41918/48008 [2:24:00<19:39, 5.16it/s] 87%|████████▋ | 41919/48008 [2:24:00<19:30, 5.20it/s] 87%|████████▋ | 41920/48008 [2:24:01<19:25, 5.22it/s] 87%|████████▋ | 41921/48008 [2:24:01<20:47, 4.88it/s] 87%|████████▋ | 41922/48008 [2:24:01<20:22, 4.98it/s] 87%|████████▋ | 41923/48008 [2:24:01<20:05, 5.05it/s] 87%|████████▋ | 41924/48008 [2:24:01<19:55, 5.09it/s] 87%|████████▋ | 41925/48008 [2:24:02<19:49, 5.11it/s] 87%|████████▋ | 41926/48008 [2:24:02<19:41, 5.15it/s] 87%|████████▋ | 41927/48008 [2:24:02<19:59, 5.07it/s] 87%|████████▋ | 41928/48008 [2:24:02<19:45, 5.13it/s] 87%|████████▋ | 41929/48008 [2:24:02<19:35, 5.17it/s] 87%|████████▋ | 41930/48008 [2:24:03<19:35, 5.17it/s] 87%|████████▋ | 41931/48008 [2:24:03<20:55, 4.84it/s] 87%|████████▋ | 41932/48008 [2:24:03<20:45, 4.88it/s] 87%|████████▋ | 41933/48008 [2:24:03<20:17, 4.99it/s] 87%|████████▋ | 41934/48008 [2:24:03<19:58, 5.07it/s] 87%|████████▋ | 41935/48008 [2:24:04<20:08, 5.03it/s] 87%|████████▋ | 41936/48008 [2:24:04<19:48, 5.11it/s] 87%|████████▋ | 41937/48008 [2:24:04<20:06, 5.03it/s] 87%|████████▋ | 41938/48008 [2:24:04<19:48, 5.11it/s] 87%|████████▋ | 41939/48008 [2:24:04<19:43, 5.13it/s] 87%|████████▋ | 41940/48008 [2:24:05<19:57, 5.07it/s] 87%|████████▋ | 41941/48008 [2:24:05<19:40, 5.14it/s] 87%|████████▋ | 41942/48008 [2:24:05<19:29, 5.19it/s] 87%|████████▋ | 41943/48008 [2:24:05<19:32, 5.17it/s] 87%|████████▋ | 41944/48008 [2:24:05<19:31, 5.17it/s] 87%|████████▋ | 41945/48008 [2:24:05<19:28, 5.19it/s] 87%|████████▋ | 41946/48008 [2:24:06<19:15, 5.25it/s] 87%|████████▋ | 41947/48008 [2:24:06<19:05, 5.29it/s] 87%|████████▋ | 41948/48008 [2:24:06<19:32, 5.17it/s] 87%|████████▋ | 41949/48008 [2:24:06<19:24, 5.20it/s] 87%|████████▋ | 41950/48008 [2:24:06<19:28, 5.18it/s] {'loss': 0.3015, 'grad_norm': 0.263671875, 'learning_rate': 2.5241626395600738e-05, 'epoch': 0.87} 87%|████████▋ | 41950/48008 [2:24:06<19:28, 5.18it/s] 87%|████████▋ | 41951/48008 [2:24:07<19:37, 5.15it/s] 87%|████████▋ | 41952/48008 [2:24:07<19:46, 5.10it/s] 87%|████████▋ | 41953/48008 [2:24:07<19:37, 5.14it/s] 87%|████████▋ | 41954/48008 [2:24:07<19:28, 5.18it/s] 87%|████████▋ | 41955/48008 [2:24:07<19:42, 5.12it/s] 87%|████████▋ | 41956/48008 [2:24:08<19:32, 5.16it/s] 87%|████████▋ | 41957/48008 [2:24:08<19:25, 5.19it/s] 87%|████████▋ | 41958/48008 [2:24:08<19:29, 5.17it/s] 87%|████████▋ | 41959/48008 [2:24:08<19:22, 5.20it/s] 87%|████████▋ | 41960/48008 [2:24:08<19:24, 5.19it/s] 87%|████████▋ | 41961/48008 [2:24:09<19:25, 5.19it/s] 87%|████████▋ | 41962/48008 [2:24:09<22:52, 4.41it/s] 87%|████████▋ | 41963/48008 [2:24:09<21:39, 4.65it/s] 87%|████████▋ | 41964/48008 [2:24:09<20:53, 4.82it/s] 87%|████████▋ | 41965/48008 [2:24:09<20:09, 4.99it/s] 87%|████████▋ | 41966/48008 [2:24:10<25:46, 3.91it/s] 87%|████████▋ | 41967/48008 [2:24:10<24:06, 4.18it/s] 87%|████████▋ | 41968/48008 [2:24:10<22:45, 4.42it/s] 87%|████████▋ | 41969/48008 [2:24:10<21:56, 4.59it/s] 87%|████████▋ | 41970/48008 [2:24:11<20:57, 4.80it/s] 87%|████████▋ | 41971/48008 [2:24:11<20:27, 4.92it/s] 87%|████████▋ | 41972/48008 [2:24:11<20:24, 4.93it/s] 87%|████████▋ | 41973/48008 [2:24:11<20:24, 4.93it/s] 87%|████████▋ | 41974/48008 [2:24:11<20:06, 5.00it/s] 87%|████████▋ | 41975/48008 [2:24:12<19:51, 5.07it/s] 87%|████████▋ | 41976/48008 [2:24:12<19:39, 5.11it/s] 87%|████████▋ | 41977/48008 [2:24:12<20:56, 4.80it/s] 87%|████████▋ | 41978/48008 [2:24:12<20:17, 4.95it/s] 87%|████████▋ | 41979/48008 [2:24:12<19:55, 5.04it/s] 87%|████████▋ | 41980/48008 [2:24:13<19:42, 5.10it/s] 87%|████████▋ | 41981/48008 [2:24:13<19:23, 5.18it/s] 87%|████████▋ | 41982/48008 [2:24:13<19:13, 5.23it/s] 87%|████████▋ | 41983/48008 [2:24:13<19:04, 5.27it/s] 87%|████████▋ | 41984/48008 [2:24:13<19:13, 5.22it/s] 87%|████████▋ | 41985/48008 [2:24:14<20:34, 4.88it/s] 87%|████████▋ | 41986/48008 [2:24:14<20:05, 4.99it/s] 87%|████████▋ | 41987/48008 [2:24:14<19:43, 5.09it/s] 87%|████████▋ | 41988/48008 [2:24:14<19:44, 5.08it/s] 87%|████████▋ | 41989/48008 [2:24:14<19:31, 5.14it/s] 87%|████████▋ | 41990/48008 [2:24:15<20:47, 4.83it/s] 87%|████████▋ | 41991/48008 [2:24:15<20:10, 4.97it/s] 87%|████████▋ | 41992/48008 [2:24:15<19:59, 5.02it/s] 87%|████████▋ | 41993/48008 [2:24:15<19:37, 5.11it/s] 87%|████████▋ | 41994/48008 [2:24:15<19:30, 5.14it/s] 87%|████████▋ | 41995/48008 [2:24:16<19:16, 5.20it/s] 87%|████████▋ | 41996/48008 [2:24:16<19:10, 5.23it/s] 87%|████████▋ | 41997/48008 [2:24:16<19:00, 5.27it/s] 87%|████████▋ | 41998/48008 [2:24:16<19:17, 5.19it/s] 87%|████████▋ | 41999/48008 [2:24:16<19:21, 5.17it/s] 87%|████████▋ | 42000/48008 [2:24:17<20:45, 4.82it/s] {'loss': 0.2982, 'grad_norm': 0.1533203125, 'learning_rate': 2.5033327778703552e-05, 'epoch': 0.87} 87%|████████▋ | 42000/48008 [2:24:17<20:45, 4.82it/s] 87%|████████▋ | 42001/48008 [2:24:17<20:21, 4.92it/s] 87%|████████▋ | 42002/48008 [2:24:17<20:10, 4.96it/s] 87%|████████▋ | 42003/48008 [2:24:17<19:42, 5.08it/s] 87%|████████▋ | 42004/48008 [2:24:17<19:47, 5.06it/s] 87%|████████▋ | 42005/48008 [2:24:17<19:32, 5.12it/s] 87%|████████▋ | 42006/48008 [2:24:18<19:36, 5.10it/s] 88%|████████▊ | 42007/48008 [2:24:18<19:23, 5.16it/s] 88%|████████▊ | 42008/48008 [2:24:18<19:36, 5.10it/s] 88%|████████▊ | 42009/48008 [2:24:18<19:44, 5.06it/s] 88%|████████▊ | 42010/48008 [2:24:18<19:33, 5.11it/s] 88%|████████▊ | 42011/48008 [2:24:19<19:22, 5.16it/s] 88%|████████▊ | 42012/48008 [2:24:19<19:19, 5.17it/s] 88%|████████▊ | 42013/48008 [2:24:19<19:36, 5.10it/s] 88%|████████▊ | 42014/48008 [2:24:19<19:23, 5.15it/s] 88%|████████▊ | 42015/48008 [2:24:19<19:21, 5.16it/s] 88%|████████▊ | 42016/48008 [2:24:20<20:41, 4.83it/s] 88%|████████▊ | 42017/48008 [2:24:20<20:03, 4.98it/s] 88%|████████▊ | 42018/48008 [2:24:20<19:41, 5.07it/s] 88%|████████▊ | 42019/48008 [2:24:20<19:38, 5.08it/s] 88%|████████▊ | 42020/48008 [2:24:20<19:34, 5.10it/s] 88%|████████▊ | 42021/48008 [2:24:21<19:21, 5.16it/s] 88%|████████▊ | 42022/48008 [2:24:21<19:17, 5.17it/s] 88%|████████▊ | 42023/48008 [2:24:21<19:14, 5.18it/s] 88%|████████▊ | 42024/48008 [2:24:21<19:05, 5.23it/s] 88%|████████▊ | 42025/48008 [2:24:21<18:54, 5.27it/s] 88%|████████▊ | 42026/48008 [2:24:22<18:57, 5.26it/s] 88%|████████▊ | 42027/48008 [2:24:22<19:32, 5.10it/s] 88%|████████▊ | 42028/48008 [2:24:22<25:15, 3.95it/s] 88%|████████▊ | 42029/48008 [2:24:22<23:26, 4.25it/s] 88%|████████▊ | 42030/48008 [2:24:23<22:05, 4.51it/s] 88%|████████▊ | 42031/48008 [2:24:23<21:00, 4.74it/s] 88%|████████▊ | 42032/48008 [2:24:23<20:44, 4.80it/s] 88%|████████▊ | 42033/48008 [2:24:23<20:12, 4.93it/s] 88%|████████▊ | 42034/48008 [2:24:23<19:47, 5.03it/s] 88%|████████▊ | 42035/48008 [2:24:24<25:22, 3.92it/s] 88%|████████▊ | 42036/48008 [2:24:24<23:32, 4.23it/s] 88%|████████▊ | 42037/48008 [2:24:24<22:03, 4.51it/s] 88%|████████▊ | 42038/48008 [2:24:24<26:56, 3.69it/s] 88%|████████▊ | 42039/48008 [2:24:25<24:44, 4.02it/s] 88%|████████▊ | 42040/48008 [2:24:25<23:23, 4.25it/s] 88%|████████▊ | 42041/48008 [2:24:25<22:16, 4.46it/s] 88%|████████▊ | 42042/48008 [2:24:25<21:25, 4.64it/s] 88%|████████▊ | 42043/48008 [2:24:25<21:11, 4.69it/s] 88%|████████▊ | 42044/48008 [2:24:26<20:33, 4.84it/s] 88%|████████▊ | 42045/48008 [2:24:26<20:38, 4.81it/s] 88%|████████▊ | 42046/48008 [2:24:26<20:02, 4.96it/s] 88%|████████▊ | 42047/48008 [2:24:26<19:38, 5.06it/s] 88%|████████▊ | 42048/48008 [2:24:26<19:52, 5.00it/s] 88%|████████▊ | 42049/48008 [2:24:27<19:38, 5.05it/s] 88%|████████▊ | 42050/48008 [2:24:27<22:51, 4.35it/s] {'loss': 0.2978, 'grad_norm': 0.435546875, 'learning_rate': 2.482502916180637e-05, 'epoch': 0.88} 88%|████████▊ | 42050/48008 [2:24:27<22:51, 4.35it/s] 88%|████████▊ | 42051/48008 [2:24:27<21:33, 4.60it/s] 88%|████████▊ | 42052/48008 [2:24:27<20:50, 4.76it/s] 88%|████████▊ | 42053/48008 [2:24:28<20:15, 4.90it/s] 88%|████████▊ | 42054/48008 [2:24:28<19:50, 5.00it/s] 88%|████████▊ | 42055/48008 [2:24:28<23:03, 4.30it/s] 88%|████████▊ | 42056/48008 [2:24:28<22:00, 4.51it/s] 88%|████████▊ | 42057/48008 [2:24:29<50:28, 1.96it/s] 88%|████████▊ | 42058/48008 [2:24:30<41:09, 2.41it/s] 88%|████████▊ | 42059/48008 [2:24:30<34:42, 2.86it/s] 88%|████████▊ | 42060/48008 [2:24:30<29:54, 3.31it/s] 88%|████████▊ | 42061/48008 [2:24:30<26:28, 3.74it/s] 88%|████████▊ | 42062/48008 [2:24:30<25:33, 3.88it/s] 88%|████████▊ | 42063/48008 [2:24:31<23:22, 4.24it/s] 88%|████████▊ | 42064/48008 [2:24:31<21:59, 4.51it/s] 88%|████████▊ | 42065/48008 [2:24:31<21:01, 4.71it/s] 88%|████████▊ | 42066/48008 [2:24:31<20:15, 4.89it/s] 88%|████████▊ | 42067/48008 [2:24:31<20:09, 4.91it/s] 88%|████████▊ | 42068/48008 [2:24:32<19:44, 5.02it/s] 88%|████████▊ | 42069/48008 [2:24:32<19:12, 5.15it/s] 88%|████████▊ | 42070/48008 [2:24:32<22:36, 4.38it/s] 88%|████████▊ | 42071/48008 [2:24:32<21:32, 4.59it/s] 88%|████████▊ | 42072/48008 [2:24:32<20:42, 4.78it/s] 88%|████████▊ | 42073/48008 [2:24:33<20:11, 4.90it/s] 88%|████████▊ | 42074/48008 [2:24:33<20:18, 4.87it/s] 88%|████████▊ | 42075/48008 [2:24:33<19:42, 5.02it/s] 88%|████████▊ | 42076/48008 [2:24:33<19:28, 5.08it/s] 88%|████████▊ | 42077/48008 [2:24:33<19:28, 5.07it/s] 88%|████████▊ | 42078/48008 [2:24:34<19:11, 5.15it/s] 88%|████████▊ | 42079/48008 [2:24:34<19:17, 5.12it/s] 88%|████████▊ | 42080/48008 [2:24:34<19:03, 5.18it/s] 88%|████████▊ | 42081/48008 [2:24:34<22:24, 4.41it/s] 88%|████████▊ | 42082/48008 [2:24:34<21:35, 4.57it/s] 88%|████████▊ | 42083/48008 [2:24:35<20:55, 4.72it/s] 88%|████████▊ | 42084/48008 [2:24:35<19:58, 4.94it/s] 88%|████████▊ | 42085/48008 [2:24:35<19:57, 4.95it/s] 88%|████████▊ | 42086/48008 [2:24:35<19:34, 5.04it/s] 88%|████████▊ | 42087/48008 [2:24:35<19:44, 5.00it/s] 88%|████████▊ | 42088/48008 [2:24:36<19:27, 5.07it/s] 88%|████████▊ | 42089/48008 [2:24:36<19:27, 5.07it/s] 88%|████████▊ | 42090/48008 [2:24:36<19:32, 5.05it/s] 88%|████████▊ | 42091/48008 [2:24:36<19:41, 5.01it/s] 88%|████████▊ | 42092/48008 [2:24:36<19:31, 5.05it/s] 88%|████████▊ | 42093/48008 [2:24:37<19:22, 5.09it/s] 88%|████████▊ | 42094/48008 [2:24:37<24:57, 3.95it/s] 88%|████████▊ | 42095/48008 [2:24:37<23:13, 4.24it/s] 88%|████████▊ | 42096/48008 [2:24:37<21:55, 4.50it/s] 88%|████████▊ | 42097/48008 [2:24:38<21:18, 4.62it/s] 88%|████████▊ | 42098/48008 [2:24:38<20:24, 4.83it/s] 88%|████████▊ | 42099/48008 [2:24:38<19:57, 4.93it/s] 88%|████████▊ | 42100/48008 [2:24:38<19:49, 4.97it/s] 88%|████████▊ | 42100/48008 [2:24:38<19:49, 4.97it/s]{'loss': 0.2995, 'grad_norm': 0.2236328125, 'learning_rate': 2.4616730544909183e-05, 'epoch': 0.88} 88%|████████▊ | 42101/48008 [2:24:38<19:43, 4.99it/s] 88%|████████▊ | 42102/48008 [2:24:39<19:27, 5.06it/s] 88%|████████▊ | 42103/48008 [2:24:39<19:15, 5.11it/s] 88%|████████▊ | 42104/48008 [2:24:39<19:20, 5.09it/s] 88%|████████▊ | 42105/48008 [2:24:39<19:13, 5.12it/s] 88%|████████▊ | 42106/48008 [2:24:39<18:53, 5.21it/s] 88%|████████▊ | 42107/48008 [2:24:40<18:50, 5.22it/s] 88%|████████▊ | 42108/48008 [2:24:40<18:47, 5.23it/s] 88%|████████▊ | 42109/48008 [2:24:40<18:44, 5.25it/s] 88%|████████▊ | 42110/48008 [2:24:40<18:47, 5.23it/s] 88%|████████▊ | 42111/48008 [2:24:40<18:44, 5.24it/s] 88%|████████▊ | 42112/48008 [2:24:40<18:33, 5.30it/s] 88%|████████▊ | 42113/48008 [2:24:41<18:56, 5.19it/s] 88%|████████▊ | 42114/48008 [2:24:41<18:53, 5.20it/s] 88%|████████▊ | 42115/48008 [2:24:41<19:04, 5.15it/s] 88%|████████▊ | 42116/48008 [2:24:41<19:00, 5.16it/s] 88%|████████▊ | 42117/48008 [2:24:41<19:02, 5.15it/s] 88%|████████▊ | 42118/48008 [2:24:42<19:31, 5.03it/s] 88%|████████▊ | 42119/48008 [2:24:42<19:13, 5.11it/s] 88%|████████▊ | 42120/48008 [2:24:42<19:35, 5.01it/s] 88%|████████▊ | 42121/48008 [2:24:42<19:40, 4.99it/s] 88%|████████▊ | 42122/48008 [2:24:42<19:36, 5.00it/s] 88%|████████▊ | 42123/48008 [2:24:43<19:18, 5.08it/s] 88%|████████▊ | 42124/48008 [2:24:43<19:30, 5.03it/s] 88%|████████▊ | 42125/48008 [2:24:43<19:49, 4.95it/s] 88%|████████▊ | 42126/48008 [2:24:43<19:26, 5.04it/s] 88%|████████▊ | 42127/48008 [2:24:43<19:13, 5.10it/s] 88%|████████▊ | 42128/48008 [2:24:44<19:03, 5.14it/s] 88%|████████▊ | 42129/48008 [2:24:44<19:30, 5.02it/s] 88%|████████▊ | 42130/48008 [2:24:44<19:08, 5.12it/s] 88%|████████▊ | 42131/48008 [2:24:44<19:07, 5.12it/s] 88%|████████▊ | 42132/48008 [2:24:44<19:06, 5.12it/s] 88%|████████▊ | 42133/48008 [2:24:45<19:16, 5.08it/s] 88%|████████▊ | 42134/48008 [2:24:45<19:08, 5.12it/s] 88%|████████▊ | 42135/48008 [2:24:45<19:06, 5.12it/s] 88%|████████▊ | 42136/48008 [2:24:45<18:56, 5.16it/s] 88%|████████▊ | 42137/48008 [2:24:45<18:53, 5.18it/s] 88%|████████▊ | 42138/48008 [2:24:46<22:12, 4.41it/s] 88%|████████▊ | 42139/48008 [2:24:46<21:12, 4.61it/s] 88%|████████▊ | 42140/48008 [2:24:46<20:31, 4.76it/s] 88%|████████▊ | 42141/48008 [2:24:47<40:06, 2.44it/s] 88%|████████▊ | 42142/48008 [2:24:47<33:40, 2.90it/s] 88%|████████▊ | 42143/48008 [2:24:47<29:11, 3.35it/s] 88%|████████▊ | 42144/48008 [2:24:48<25:55, 3.77it/s] 88%|████████▊ | 42145/48008 [2:24:48<23:41, 4.12it/s] 88%|████████▊ | 42146/48008 [2:24:48<22:02, 4.43it/s] 88%|████████▊ | 42147/48008 [2:24:48<21:02, 4.64it/s] 88%|████████▊ | 42148/48008 [2:24:48<20:10, 4.84it/s] 88%|████████▊ | 42149/48008 [2:24:48<19:46, 4.94it/s] 88%|████████▊ | 42150/48008 [2:24:49<19:32, 5.00it/s] 88%|████████▊ | 42150/48008 [2:24:49<19:32, 5.00it/s]{'loss': 0.2988, 'grad_norm': 0.318359375, 'learning_rate': 2.4408431928012e-05, 'epoch': 0.88} 88%|████████▊ | 42151/48008 [2:24:49<19:15, 5.07it/s] 88%|████████▊ | 42152/48008 [2:24:49<19:08, 5.10it/s] 88%|████████▊ | 42153/48008 [2:24:49<19:00, 5.14it/s] 88%|████████▊ | 42154/48008 [2:24:49<18:54, 5.16it/s] 88%|████████▊ | 42155/48008 [2:24:50<18:53, 5.16it/s] 88%|████████▊ | 42156/48008 [2:24:50<18:59, 5.13it/s] 88%|████████▊ | 42157/48008 [2:24:50<19:14, 5.07it/s] 88%|████████▊ | 42158/48008 [2:24:50<19:13, 5.07it/s] 88%|████████▊ | 42159/48008 [2:24:50<19:04, 5.11it/s] 88%|████████▊ | 42160/48008 [2:24:51<19:00, 5.13it/s] 88%|████████▊ | 42161/48008 [2:24:51<18:48, 5.18it/s] 88%|████████▊ | 42162/48008 [2:24:51<18:37, 5.23it/s] 88%|████████▊ | 42163/48008 [2:24:51<18:48, 5.18it/s] 88%|████████▊ | 42164/48008 [2:24:51<18:36, 5.23it/s] 88%|████████▊ | 42165/48008 [2:24:52<18:39, 5.22it/s] 88%|████████▊ | 42166/48008 [2:24:52<18:43, 5.20it/s] 88%|████████▊ | 42167/48008 [2:24:52<20:02, 4.86it/s] 88%|████████▊ | 42168/48008 [2:24:52<25:17, 3.85it/s] 88%|████████▊ | 42169/48008 [2:24:53<23:21, 4.17it/s] 88%|████████▊ | 42170/48008 [2:24:53<22:18, 4.36it/s] 88%|████████▊ | 42171/48008 [2:24:53<22:36, 4.30it/s] 88%|████████▊ | 42172/48008 [2:24:53<21:22, 4.55it/s] 88%|████████▊ | 42173/48008 [2:24:53<20:24, 4.77it/s] 88%|████████▊ | 42174/48008 [2:24:54<23:14, 4.18it/s] 88%|████████▊ | 42175/48008 [2:24:54<21:53, 4.44it/s] 88%|████████▊ | 42176/48008 [2:24:54<21:00, 4.63it/s] 88%|████████▊ | 42177/48008 [2:24:54<20:18, 4.78it/s] 88%|████████▊ | 42178/48008 [2:24:54<19:43, 4.93it/s] 88%|████████▊ | 42179/48008 [2:24:55<19:23, 5.01it/s] 88%|████████▊ | 42180/48008 [2:24:55<19:29, 4.98it/s] 88%|████████▊ | 42181/48008 [2:24:55<19:14, 5.05it/s] 88%|████████▊ | 42182/48008 [2:24:55<19:10, 5.06it/s] 88%|████████▊ | 42183/48008 [2:24:55<19:20, 5.02it/s] 88%|████████▊ | 42184/48008 [2:24:56<19:02, 5.10it/s] 88%|████████▊ | 42185/48008 [2:24:56<18:49, 5.15it/s] 88%|████████▊ | 42186/48008 [2:24:56<19:05, 5.08it/s] 88%|████████▊ | 42187/48008 [2:24:56<18:54, 5.13it/s] 88%|████████▊ | 42188/48008 [2:24:56<18:45, 5.17it/s] 88%|████████▊ | 42189/48008 [2:24:57<18:36, 5.21it/s] 88%|████████▊ | 42190/48008 [2:24:57<18:55, 5.12it/s] 88%|████████▊ | 42191/48008 [2:24:57<19:08, 5.07it/s] 88%|████████▊ | 42192/48008 [2:24:57<19:00, 5.10it/s] 88%|████████▊ | 42193/48008 [2:24:57<18:52, 5.13it/s] 88%|████████▊ | 42194/48008 [2:24:58<18:48, 5.15it/s] 88%|████████▊ | 42195/48008 [2:24:58<18:47, 5.16it/s] 88%|████████▊ | 42196/48008 [2:24:58<22:03, 4.39it/s] 88%|████████▊ | 42197/48008 [2:24:58<21:13, 4.56it/s] 88%|████████▊ | 42198/48008 [2:24:58<20:19, 4.76it/s] 88%|████████▊ | 42199/48008 [2:24:59<19:40, 4.92it/s] 88%|████████▊ | 42200/48008 [2:24:59<19:21, 5.00it/s] {'loss': 0.3001, 'grad_norm': 0.1943359375, 'learning_rate': 2.4200133311114814e-05, 'epoch': 0.88} 88%|████████▊ | 42200/48008 [2:24:59<19:21, 5.00it/s] 88%|████████▊ | 42201/48008 [2:24:59<19:20, 5.00it/s] 88%|████████▊ | 42202/48008 [2:24:59<19:22, 4.99it/s] 88%|████████▊ | 42203/48008 [2:24:59<18:57, 5.10it/s] 88%|████████▊ | 42204/48008 [2:25:00<19:19, 5.01it/s] 88%|████████▊ | 42205/48008 [2:25:00<20:25, 4.74it/s] 88%|████████▊ | 42206/48008 [2:25:00<19:47, 4.89it/s] 88%|████████▊ | 42207/48008 [2:25:00<19:20, 5.00it/s] 88%|████████▊ | 42208/48008 [2:25:00<19:09, 5.04it/s] 88%|████████▊ | 42209/48008 [2:25:01<20:16, 4.77it/s] 88%|████████▊ | 42210/48008 [2:25:01<19:41, 4.91it/s] 88%|████████▊ | 42211/48008 [2:25:01<19:47, 4.88it/s] 88%|████████▊ | 42212/48008 [2:25:01<19:19, 5.00it/s] 88%|████████▊ | 42213/48008 [2:25:01<18:56, 5.10it/s] 88%|████████▊ | 42214/48008 [2:25:02<18:58, 5.09it/s] 88%|████████▊ | 42215/48008 [2:25:02<19:09, 5.04it/s] 88%|████████▊ | 42216/48008 [2:25:02<19:00, 5.08it/s] 88%|████████▊ | 42217/48008 [2:25:02<20:11, 4.78it/s] 88%|████████▊ | 42218/48008 [2:25:02<19:36, 4.92it/s] 88%|████████▊ | 42219/48008 [2:25:03<19:16, 5.00it/s] 88%|████████▊ | 42220/48008 [2:25:03<18:59, 5.08it/s] 88%|████████▊ | 42221/48008 [2:25:03<19:09, 5.04it/s] 88%|████████▊ | 42222/48008 [2:25:03<19:27, 4.95it/s] 88%|████████▊ | 42223/48008 [2:25:03<19:08, 5.04it/s] 88%|████████▊ | 42224/48008 [2:25:04<18:56, 5.09it/s] 88%|████████▊ | 42225/48008 [2:25:04<18:41, 5.16it/s] 88%|████████▊ | 42226/48008 [2:25:04<18:31, 5.20it/s] 88%|████████▊ | 42227/48008 [2:25:04<18:23, 5.24it/s] 88%|████████▊ | 42228/48008 [2:25:04<18:24, 5.23it/s] 88%|████████▊ | 42229/48008 [2:25:05<18:16, 5.27it/s] 88%|████████▊ | 42230/48008 [2:25:05<18:50, 5.11it/s] 88%|████████▊ | 42231/48008 [2:25:05<18:41, 5.15it/s] 88%|████████▊ | 42232/48008 [2:25:05<18:43, 5.14it/s] 88%|████████▊ | 42233/48008 [2:25:05<18:37, 5.17it/s] 88%|████████▊ | 42234/48008 [2:25:06<18:51, 5.10it/s] 88%|████████▊ | 42235/48008 [2:25:06<22:02, 4.37it/s] 88%|████████▊ | 42236/48008 [2:25:06<20:52, 4.61it/s] 88%|████████▊ | 42237/48008 [2:25:06<20:07, 4.78it/s] 88%|████████▊ | 42238/48008 [2:25:06<19:42, 4.88it/s] 88%|████████▊ | 42239/48008 [2:25:07<19:18, 4.98it/s] 88%|████████▊ | 42240/48008 [2:25:07<19:04, 5.04it/s] 88%|████████▊ | 42241/48008 [2:25:07<19:01, 5.05it/s] 88%|████████▊ | 42242/48008 [2:25:07<18:48, 5.11it/s] 88%|████████▊ | 42243/48008 [2:25:07<18:35, 5.17it/s] 88%|████████▊ | 42244/48008 [2:25:08<18:23, 5.22it/s] 88%|████████▊ | 42245/48008 [2:25:08<19:43, 4.87it/s] 88%|████████▊ | 42246/48008 [2:25:08<19:15, 4.99it/s] 88%|████████▊ | 42247/48008 [2:25:08<22:16, 4.31it/s] 88%|████████▊ | 42248/48008 [2:25:09<24:25, 3.93it/s] 88%|████████▊ | 42249/48008 [2:25:09<22:30, 4.27it/s] 88%|████████▊ | 42250/48008 [2:25:09<21:14, 4.52it/s] {'loss': 0.301, 'grad_norm': 0.2451171875, 'learning_rate': 2.3991834694217632e-05, 'epoch': 0.88} 88%|████████▊ | 42250/48008 [2:25:09<21:14, 4.52it/s] 88%|████████▊ | 42251/48008 [2:25:09<20:22, 4.71it/s] 88%|████████▊ | 42252/48008 [2:25:10<25:22, 3.78it/s] 88%|████████▊ | 42253/48008 [2:25:10<23:50, 4.02it/s] 88%|████████▊ | 42254/48008 [2:25:10<23:34, 4.07it/s] 88%|████████▊ | 42255/48008 [2:25:10<22:21, 4.29it/s] 88%|████████▊ | 42256/48008 [2:25:10<21:10, 4.53it/s] 88%|████████▊ | 42257/48008 [2:25:11<20:16, 4.73it/s] 88%|████████▊ | 42258/48008 [2:25:11<19:38, 4.88it/s] 88%|████████▊ | 42259/48008 [2:25:11<19:51, 4.83it/s] 88%|████████▊ | 42260/48008 [2:25:11<19:28, 4.92it/s] 88%|████████▊ | 42261/48008 [2:25:11<19:06, 5.01it/s] 88%|████████▊ | 42262/48008 [2:25:12<18:46, 5.10it/s] 88%|████████▊ | 42263/48008 [2:25:12<18:31, 5.17it/s] 88%|████████▊ | 42264/48008 [2:25:12<18:45, 5.10it/s] 88%|████████▊ | 42265/48008 [2:25:12<18:50, 5.08it/s] 88%|████████▊ | 42266/48008 [2:25:12<18:39, 5.13it/s] 88%|████████▊ | 42267/48008 [2:25:13<18:34, 5.15it/s] 88%|████████▊ | 42268/48008 [2:25:13<18:52, 5.07it/s] 88%|████████▊ | 42269/48008 [2:25:13<18:41, 5.12it/s] 88%|████████▊ | 42270/48008 [2:25:13<18:57, 5.04it/s] 88%|████████▊ | 42271/48008 [2:25:13<18:36, 5.14it/s] 88%|████████▊ | 42272/48008 [2:25:14<18:36, 5.14it/s] 88%|████████▊ | 42273/48008 [2:25:14<19:50, 4.82it/s] 88%|████████▊ | 42274/48008 [2:25:14<19:20, 4.94it/s] 88%|████████▊ | 42275/48008 [2:25:14<22:16, 4.29it/s] 88%|████████▊ | 42276/48008 [2:25:15<21:18, 4.48it/s] 88%|████████▊ | 42277/48008 [2:25:15<20:19, 4.70it/s] 88%|████████▊ | 42278/48008 [2:25:15<19:47, 4.82it/s] 88%|████████▊ | 42279/48008 [2:25:15<24:53, 3.84it/s] 88%|████████▊ | 42280/48008 [2:25:15<22:54, 4.17it/s] 88%|████████▊ | 42281/48008 [2:25:16<21:54, 4.36it/s] 88%|████████▊ | 42282/48008 [2:25:16<20:47, 4.59it/s] 88%|████████▊ | 42283/48008 [2:25:16<23:22, 4.08it/s] 88%|████████▊ | 42284/48008 [2:25:16<22:20, 4.27it/s] 88%|████████▊ | 42285/48008 [2:25:17<21:06, 4.52it/s] 88%|████████▊ | 42286/48008 [2:25:17<20:26, 4.67it/s] 88%|████████▊ | 42287/48008 [2:25:17<19:44, 4.83it/s] 88%|████████▊ | 42288/48008 [2:25:17<19:48, 4.81it/s] 88%|████████▊ | 42289/48008 [2:25:17<19:22, 4.92it/s] 88%|████████▊ | 42290/48008 [2:25:18<18:59, 5.02it/s] 88%|████████▊ | 42291/48008 [2:25:18<20:07, 4.73it/s] 88%|████████▊ | 42292/48008 [2:25:18<20:05, 4.74it/s] 88%|████████▊ | 42293/48008 [2:25:18<19:30, 4.88it/s] 88%|████████▊ | 42294/48008 [2:25:18<19:15, 4.94it/s] 88%|████████▊ | 42295/48008 [2:25:19<20:20, 4.68it/s] 88%|████████▊ | 42296/48008 [2:25:19<20:06, 4.74it/s] 88%|████████▊ | 42297/48008 [2:25:19<19:28, 4.89it/s] 88%|████████▊ | 42298/48008 [2:25:19<19:05, 4.99it/s] 88%|████████▊ | 42299/48008 [2:25:19<18:43, 5.08it/s] 88%|████████▊ | 42300/48008 [2:25:20<19:50, 4.79it/s] {'loss': 0.2963, 'grad_norm': 0.302734375, 'learning_rate': 2.378353607732045e-05, 'epoch': 0.88} 88%|████████▊ | 42300/48008 [2:25:20<19:50, 4.79it/s] 88%|████████▊ | 42301/48008 [2:25:20<19:21, 4.91it/s] 88%|████████▊ | 42302/48008 [2:25:20<20:20, 4.67it/s] 88%|████████▊ | 42303/48008 [2:25:20<19:43, 4.82it/s] 88%|████████▊ | 42304/48008 [2:25:20<19:31, 4.87it/s] 88%|████████▊ | 42305/48008 [2:25:21<19:03, 4.99it/s] 88%|████████▊ | 42306/48008 [2:25:21<18:40, 5.09it/s] 88%|████████▊ | 42307/48008 [2:25:21<18:21, 5.18it/s] 88%|████████▊ | 42308/48008 [2:25:21<18:23, 5.17it/s] 88%|████████▊ | 42309/48008 [2:25:21<18:27, 5.14it/s] 88%|████████▊ | 42310/48008 [2:25:22<18:24, 5.16it/s] 88%|████████▊ | 42311/48008 [2:25:22<18:55, 5.02it/s] 88%|████████▊ | 42312/48008 [2:25:22<18:47, 5.05it/s] 88%|████████▊ | 42313/48008 [2:25:22<18:53, 5.02it/s] 88%|████████▊ | 42314/48008 [2:25:22<19:03, 4.98it/s] 88%|████████▊ | 42315/48008 [2:25:23<18:47, 5.05it/s] 88%|████████▊ | 42316/48008 [2:25:23<18:49, 5.04it/s] 88%|████████▊ | 42317/48008 [2:25:23<18:46, 5.05it/s] 88%|████████▊ | 42318/48008 [2:25:23<18:36, 5.09it/s] 88%|████████▊ | 42319/48008 [2:25:23<18:23, 5.16it/s] 88%|████████▊ | 42320/48008 [2:25:24<18:11, 5.21it/s] 88%|████████▊ | 42321/48008 [2:25:24<18:02, 5.26it/s] 88%|████████▊ | 42322/48008 [2:25:24<18:05, 5.24it/s] 88%|████████▊ | 42323/48008 [2:25:24<18:02, 5.25it/s] 88%|████████▊ | 42324/48008 [2:25:24<17:54, 5.29it/s] 88%|████████▊ | 42325/48008 [2:25:25<18:05, 5.24it/s] 88%|████████▊ | 42326/48008 [2:25:25<18:02, 5.25it/s] 88%|████████▊ | 42327/48008 [2:25:25<18:23, 5.15it/s] 88%|████████▊ | 42328/48008 [2:25:25<18:29, 5.12it/s] 88%|████████▊ | 42329/48008 [2:25:25<18:30, 5.12it/s] 88%|████████▊ | 42330/48008 [2:25:26<18:36, 5.09it/s] 88%|████████▊ | 42331/48008 [2:25:26<18:24, 5.14it/s] 88%|████████▊ | 42332/48008 [2:25:26<18:19, 5.16it/s] 88%|████████▊ | 42333/48008 [2:25:26<18:06, 5.22it/s] 88%|████████▊ | 42334/48008 [2:25:26<18:06, 5.22it/s] 88%|████████▊ | 42335/48008 [2:25:26<18:02, 5.24it/s] 88%|████████▊ | 42336/48008 [2:25:27<17:54, 5.28it/s] 88%|████████▊ | 42337/48008 [2:25:27<17:57, 5.26it/s] 88%|████████▊ | 42338/48008 [2:25:27<18:00, 5.25it/s] 88%|████████▊ | 42339/48008 [2:25:27<18:02, 5.24it/s] 88%|████████▊ | 42340/48008 [2:25:27<18:07, 5.21it/s] 88%|████████▊ | 42341/48008 [2:25:28<18:00, 5.25it/s] 88%|████████▊ | 42342/48008 [2:25:28<17:51, 5.29it/s] 88%|████████▊ | 42343/48008 [2:25:28<18:03, 5.23it/s] 88%|████████▊ | 42344/48008 [2:25:28<18:14, 5.18it/s] 88%|████████▊ | 42345/48008 [2:25:28<18:09, 5.20it/s] 88%|████████▊ | 42346/48008 [2:25:29<18:14, 5.17it/s] 88%|████████▊ | 42347/48008 [2:25:29<18:11, 5.19it/s] 88%|████████▊ | 42348/48008 [2:25:29<18:26, 5.12it/s] 88%|████████▊ | 42349/48008 [2:25:29<18:20, 5.14it/s] 88%|████████▊ | 42350/48008 [2:25:29<18:21, 5.14it/s] {'loss': 0.3018, 'grad_norm': 0.1806640625, 'learning_rate': 2.3575237460423263e-05, 'epoch': 0.88} 88%|████████▊ | 42350/48008 [2:25:29<18:21, 5.14it/s] 88%|████████▊ | 42351/48008 [2:25:30<18:08, 5.20it/s] 88%|████████▊ | 42352/48008 [2:25:30<18:05, 5.21it/s] 88%|████████▊ | 42353/48008 [2:25:30<17:59, 5.24it/s] 88%|████████▊ | 42354/48008 [2:25:30<18:00, 5.23it/s] 88%|████████▊ | 42355/48008 [2:25:30<17:52, 5.27it/s] 88%|████████▊ | 42356/48008 [2:25:30<17:46, 5.30it/s] 88%|████████▊ | 42357/48008 [2:25:31<17:47, 5.29it/s] 88%|████████▊ | 42358/48008 [2:25:31<17:53, 5.26it/s] 88%|████████▊ | 42359/48008 [2:25:31<17:52, 5.27it/s] 88%|████████▊ | 42360/48008 [2:25:31<23:21, 4.03it/s] 88%|████████▊ | 42361/48008 [2:25:32<22:14, 4.23it/s] 88%|████████▊ | 42362/48008 [2:25:32<20:55, 4.50it/s] 88%|████████▊ | 42363/48008 [2:25:32<19:59, 4.71it/s] 88%|████████▊ | 42364/48008 [2:25:32<19:25, 4.84it/s] 88%|████████▊ | 42365/48008 [2:25:32<19:01, 4.94it/s] 88%|████████▊ | 42366/48008 [2:25:33<18:38, 5.04it/s] 88%|████████▊ | 42367/48008 [2:25:33<18:20, 5.13it/s] 88%|████████▊ | 42368/48008 [2:25:33<18:24, 5.11it/s] 88%|████████▊ | 42369/48008 [2:25:33<18:47, 5.00it/s] 88%|████████▊ | 42370/48008 [2:25:33<18:39, 5.04it/s] 88%|████████▊ | 42371/48008 [2:25:34<18:26, 5.10it/s] 88%|████████▊ | 42372/48008 [2:25:34<18:14, 5.15it/s] 88%|████████▊ | 42373/48008 [2:25:34<18:05, 5.19it/s] 88%|████████▊ | 42374/48008 [2:25:34<18:00, 5.21it/s] 88%|████████▊ | 42375/48008 [2:25:34<18:00, 5.21it/s] 88%|████████▊ | 42376/48008 [2:25:35<18:21, 5.11it/s] 88%|████████▊ | 42377/48008 [2:25:35<18:26, 5.09it/s] 88%|████████▊ | 42378/48008 [2:25:35<18:26, 5.09it/s] 88%|████████▊ | 42379/48008 [2:25:35<18:09, 5.17it/s] 88%|████████▊ | 42380/48008 [2:25:35<18:08, 5.17it/s] 88%|████████▊ | 42381/48008 [2:25:36<18:08, 5.17it/s] 88%|████████▊ | 42382/48008 [2:25:36<17:56, 5.23it/s] 88%|████████▊ | 42383/48008 [2:25:36<19:11, 4.88it/s] 88%|████████▊ | 42384/48008 [2:25:36<18:41, 5.01it/s] 88%|████████▊ | 42385/48008 [2:25:36<18:24, 5.09it/s] 88%|████████▊ | 42386/48008 [2:25:37<18:16, 5.13it/s] 88%|████████▊ | 42387/48008 [2:25:37<18:19, 5.11it/s] 88%|████████▊ | 42388/48008 [2:25:37<21:25, 4.37it/s] 88%|████████▊ | 42389/48008 [2:25:37<21:38, 4.33it/s] 88%|████████▊ | 42390/48008 [2:25:37<20:31, 4.56it/s] 88%|████████▊ | 42391/48008 [2:25:38<19:33, 4.78it/s] 88%|████████▊ | 42392/48008 [2:25:38<19:03, 4.91it/s] 88%|████████▊ | 42393/48008 [2:25:38<18:42, 5.00it/s] 88%|████████▊ | 42394/48008 [2:25:38<18:19, 5.11it/s] 88%|████████▊ | 42395/48008 [2:25:38<18:14, 5.13it/s] 88%|████████▊ | 42396/48008 [2:25:39<18:15, 5.12it/s] 88%|████████▊ | 42397/48008 [2:25:39<18:08, 5.15it/s] 88%|████████▊ | 42398/48008 [2:25:39<18:01, 5.19it/s] 88%|████████▊ | 42399/48008 [2:25:39<17:58, 5.20it/s] 88%|████████▊ | 42400/48008 [2:25:39<18:21, 5.09it/s] 88%|████████▊ | 42400/48008 [2:25:39<18:21, 5.09it/s]{'loss': 0.2985, 'grad_norm': 0.24609375, 'learning_rate': 2.336693884352608e-05, 'epoch': 0.88} 88%|████████▊ | 42401/48008 [2:25:40<18:08, 5.15it/s] 88%|████████▊ | 42402/48008 [2:25:40<17:52, 5.23it/s] 88%|████████▊ | 42403/48008 [2:25:40<17:38, 5.29it/s] 88%|████████▊ | 42404/48008 [2:25:40<17:42, 5.27it/s] 88%|████████▊ | 42405/48008 [2:25:40<18:02, 5.17it/s] 88%|████████▊ | 42406/48008 [2:25:40<17:53, 5.22it/s] 88%|████████▊ | 42407/48008 [2:25:41<18:24, 5.07it/s] 88%|████████▊ | 42408/48008 [2:25:41<18:30, 5.04it/s] 88%|████████▊ | 42409/48008 [2:25:41<18:15, 5.11it/s] 88%|████████▊ | 42410/48008 [2:25:41<18:22, 5.08it/s] 88%|████████▊ | 42411/48008 [2:25:41<18:12, 5.12it/s] 88%|████████▊ | 42412/48008 [2:25:42<18:03, 5.16it/s] 88%|████████▊ | 42413/48008 [2:25:42<18:11, 5.13it/s] 88%|████████▊ | 42414/48008 [2:25:42<19:21, 4.82it/s] 88%|████████▊ | 42415/48008 [2:25:42<19:12, 4.85it/s] 88%|████████▊ | 42416/48008 [2:25:43<18:47, 4.96it/s] 88%|████████▊ | 42417/48008 [2:25:43<18:36, 5.01it/s] 88%|████████▊ | 42418/48008 [2:25:43<18:37, 5.00it/s] 88%|████████▊ | 42419/48008 [2:25:43<18:22, 5.07it/s] 88%|████████▊ | 42420/48008 [2:25:43<18:11, 5.12it/s] 88%|████████▊ | 42421/48008 [2:25:43<18:12, 5.11it/s] 88%|████████▊ | 42422/48008 [2:25:44<18:25, 5.05it/s] 88%|████████▊ | 42423/48008 [2:25:44<18:10, 5.12it/s] 88%|████████▊ | 42424/48008 [2:25:44<17:58, 5.18it/s] 88%|████████▊ | 42425/48008 [2:25:44<18:18, 5.08it/s] 88%|████████▊ | 42426/48008 [2:25:44<18:16, 5.09it/s] 88%|████████▊ | 42427/48008 [2:25:45<18:16, 5.09it/s] 88%|████████▊ | 42428/48008 [2:25:45<18:06, 5.14it/s] 88%|████████▊ | 42429/48008 [2:25:45<18:21, 5.06it/s] 88%|████████▊ | 42430/48008 [2:25:45<18:02, 5.15it/s] 88%|████████▊ | 42431/48008 [2:25:45<18:18, 5.08it/s] 88%|████████▊ | 42432/48008 [2:25:46<18:19, 5.07it/s] 88%|████████▊ | 42433/48008 [2:25:46<18:12, 5.10it/s] 88%|████████▊ | 42434/48008 [2:25:46<18:01, 5.15it/s] 88%|████████▊ | 42435/48008 [2:25:46<17:56, 5.18it/s] 88%|████████▊ | 42436/48008 [2:25:46<18:14, 5.09it/s] 88%|████████▊ | 42437/48008 [2:25:47<18:09, 5.11it/s] 88%|████████▊ | 42438/48008 [2:25:47<18:08, 5.12it/s] 88%|████████▊ | 42439/48008 [2:25:47<18:12, 5.10it/s] 88%|████████▊ | 42440/48008 [2:25:47<18:03, 5.14it/s] 88%|████████▊ | 42441/48008 [2:25:47<17:54, 5.18it/s] 88%|████████▊ | 42442/48008 [2:25:48<17:52, 5.19it/s] 88%|████████▊ | 42443/48008 [2:25:48<17:45, 5.22it/s] 88%|████████▊ | 42444/48008 [2:25:48<17:38, 5.25it/s] 88%|████████▊ | 42445/48008 [2:25:48<17:41, 5.24it/s] 88%|████████▊ | 42446/48008 [2:25:48<18:58, 4.88it/s] 88%|████████▊ | 42447/48008 [2:25:49<18:37, 4.98it/s] 88%|████████▊ | 42448/48008 [2:25:49<18:18, 5.06it/s] 88%|████████▊ | 42449/48008 [2:25:49<18:34, 4.99it/s] 88%|████████▊ | 42450/48008 [2:25:49<18:34, 4.99it/s] {'loss': 0.3019, 'grad_norm': 0.490234375, 'learning_rate': 2.3158640226628898e-05, 'epoch': 0.88} 88%|████████▊ | 42450/48008 [2:25:49<18:34, 4.99it/s] 88%|████████▊ | 42451/48008 [2:25:49<18:36, 4.98it/s] 88%|████████▊ | 42452/48008 [2:25:50<18:22, 5.04it/s] 88%|████████▊ | 42453/48008 [2:25:50<23:36, 3.92it/s] 88%|████████▊ | 42454/48008 [2:25:50<21:45, 4.25it/s] 88%|████████▊ | 42455/48008 [2:25:50<20:30, 4.51it/s] 88%|████████▊ | 42456/48008 [2:25:51<19:36, 4.72it/s] 88%|████████▊ | 42457/48008 [2:25:51<19:22, 4.77it/s] 88%|████████▊ | 42458/48008 [2:25:51<19:14, 4.81it/s] 88%|████████▊ | 42459/48008 [2:25:51<18:37, 4.97it/s] 88%|████████▊ | 42460/48008 [2:25:51<18:25, 5.02it/s] 88%|████████▊ | 42461/48008 [2:25:52<23:37, 3.91it/s] 88%|████████▊ | 42462/48008 [2:25:52<21:39, 4.27it/s] 88%|████████▊ | 42463/48008 [2:25:52<20:18, 4.55it/s] 88%|████████▊ | 42464/48008 [2:25:52<19:27, 4.75it/s] 88%|████████▊ | 42465/48008 [2:25:52<19:04, 4.84it/s] 88%|████████▊ | 42466/48008 [2:25:53<18:27, 5.00it/s] 88%|████████▊ | 42467/48008 [2:25:53<18:13, 5.07it/s] 88%|████████▊ | 42468/48008 [2:25:53<18:05, 5.10it/s] 88%|████████▊ | 42469/48008 [2:25:53<18:25, 5.01it/s] 88%|████████▊ | 42470/48008 [2:25:53<18:12, 5.07it/s] 88%|████████▊ | 42471/48008 [2:25:54<23:25, 3.94it/s] 88%|████████▊ | 42472/48008 [2:25:54<21:24, 4.31it/s] 88%|████████▊ | 42473/48008 [2:25:54<20:15, 4.55it/s] 88%|████████▊ | 42474/48008 [2:25:54<19:18, 4.78it/s] 88%|████████▊ | 42475/48008 [2:25:55<18:48, 4.90it/s] 88%|████████▊ | 42476/48008 [2:25:55<18:26, 5.00it/s] 88%|████████▊ | 42477/48008 [2:25:55<18:30, 4.98it/s] 88%|████████▊ | 42478/48008 [2:25:55<18:25, 5.00it/s] 88%|████████▊ | 42479/48008 [2:25:55<18:12, 5.06it/s] 88%|████████▊ | 42480/48008 [2:25:56<18:12, 5.06it/s] 88%|████████▊ | 42481/48008 [2:25:56<18:03, 5.10it/s] 88%|████████▊ | 42482/48008 [2:25:56<17:55, 5.14it/s] 88%|████████▊ | 42483/48008 [2:25:56<18:05, 5.09it/s] 88%|████████▊ | 42484/48008 [2:25:56<17:50, 5.16it/s] 88%|████████▊ | 42485/48008 [2:25:57<17:58, 5.12it/s] 88%|████████▊ | 42486/48008 [2:25:57<18:02, 5.10it/s] 88%|████████▊ | 42487/48008 [2:25:57<17:45, 5.18it/s] 89%|████████▊ | 42488/48008 [2:25:57<17:46, 5.18it/s] 89%|████████▊ | 42489/48008 [2:25:57<17:37, 5.22it/s] 89%|████████▊ | 42490/48008 [2:25:57<17:50, 5.16it/s] 89%|████████▊ | 42491/48008 [2:25:58<17:45, 5.18it/s] 89%|████████▊ | 42492/48008 [2:25:58<17:25, 5.28it/s] 89%|████████▊ | 42493/48008 [2:25:58<17:26, 5.27it/s] 89%|████████▊ | 42494/48008 [2:25:58<17:58, 5.11it/s] 89%|████████▊ | 42495/48008 [2:25:58<17:38, 5.21it/s] 89%|████████▊ | 42496/48008 [2:25:59<17:35, 5.22it/s] 89%|████████▊ | 42497/48008 [2:25:59<17:33, 5.23it/s] 89%|████████▊ | 42498/48008 [2:25:59<17:34, 5.23it/s] 89%|████████▊ | 42499/48008 [2:25:59<17:27, 5.26it/s] 89%|████████▊ | 42500/48008 [2:25:59<17:35, 5.22it/s] {'loss': 0.3036, 'grad_norm': 0.1787109375, 'learning_rate': 2.2950341609731712e-05, 'epoch': 0.89} 89%|████████▊ | 42500/48008 [2:25:59<17:35, 5.22it/s] 89%|████████▊ | 42501/48008 [2:26:00<17:44, 5.17it/s] 89%|████████▊ | 42502/48008 [2:26:00<17:34, 5.22it/s] 89%|████████▊ | 42503/48008 [2:26:00<18:04, 5.08it/s] 89%|████████▊ | 42504/48008 [2:26:00<21:03, 4.36it/s] 89%|████████▊ | 42505/48008 [2:26:01<23:10, 3.96it/s] 89%|████████▊ | 42506/48008 [2:26:01<21:24, 4.28it/s] 89%|████████▊ | 42507/48008 [2:26:01<20:09, 4.55it/s] 89%|████████▊ | 42508/48008 [2:26:01<19:21, 4.73it/s] 89%|████████▊ | 42509/48008 [2:26:01<19:00, 4.82it/s] 89%|████████▊ | 42510/48008 [2:26:02<21:44, 4.21it/s] 89%|████████▊ | 42511/48008 [2:26:02<20:40, 4.43it/s] 89%|████████▊ | 42512/48008 [2:26:02<19:34, 4.68it/s] 89%|████████▊ | 42513/48008 [2:26:02<18:52, 4.85it/s] 89%|████████▊ | 42514/48008 [2:26:03<23:50, 3.84it/s] 89%|████████▊ | 42515/48008 [2:26:03<21:55, 4.18it/s] 89%|████████▊ | 42516/48008 [2:26:03<20:37, 4.44it/s] 89%|████████▊ | 42517/48008 [2:26:03<19:42, 4.64it/s] 89%|████████▊ | 42518/48008 [2:26:03<19:14, 4.76it/s] 89%|████████▊ | 42519/48008 [2:26:04<18:40, 4.90it/s] 89%|████████▊ | 42520/48008 [2:26:04<18:49, 4.86it/s] 89%|████████▊ | 42521/48008 [2:26:04<18:30, 4.94it/s] 89%|████████▊ | 42522/48008 [2:26:04<18:23, 4.97it/s] 89%|████████▊ | 42523/48008 [2:26:04<18:08, 5.04it/s] 89%|████████▊ | 42524/48008 [2:26:05<17:54, 5.10it/s] 89%|████████▊ | 42525/48008 [2:26:05<17:41, 5.16it/s] 89%|████████▊ | 42526/48008 [2:26:05<17:42, 5.16it/s] 89%|████████▊ | 42527/48008 [2:26:05<17:36, 5.19it/s] 89%|████████▊ | 42528/48008 [2:26:05<17:34, 5.20it/s] 89%|████████▊ | 42529/48008 [2:26:06<17:28, 5.22it/s] 89%|████████▊ | 42530/48008 [2:26:06<17:24, 5.24it/s] 89%|████████▊ | 42531/48008 [2:26:06<17:22, 5.25it/s] 89%|████████▊ | 42532/48008 [2:26:06<17:28, 5.22it/s] 89%|████████▊ | 42533/48008 [2:26:06<17:27, 5.23it/s] 89%|████████▊ | 42534/48008 [2:26:06<17:33, 5.20it/s] 89%|████████▊ | 42535/48008 [2:26:07<17:37, 5.17it/s] 89%|████████▊ | 42536/48008 [2:26:07<17:35, 5.18it/s] 89%|████████▊ | 42537/48008 [2:26:07<17:27, 5.22it/s] 89%|████████▊ | 42538/48008 [2:26:07<17:29, 5.21it/s] 89%|████████▊ | 42539/48008 [2:26:07<17:29, 5.21it/s] 89%|████████▊ | 42540/48008 [2:26:08<17:30, 5.20it/s] 89%|████████▊ | 42541/48008 [2:26:08<17:47, 5.12it/s] 89%|████████▊ | 42542/48008 [2:26:08<20:50, 4.37it/s] 89%|████████▊ | 42543/48008 [2:26:08<19:44, 4.62it/s] 89%|████████▊ | 42544/48008 [2:26:09<18:59, 4.80it/s] 89%|████████▊ | 42545/48008 [2:26:09<18:25, 4.94it/s] 89%|████████▊ | 42546/48008 [2:26:09<18:07, 5.02it/s] 89%|████████▊ | 42547/48008 [2:26:09<17:58, 5.06it/s] 89%|████████▊ | 42548/48008 [2:26:09<17:44, 5.13it/s] 89%|████████▊ | 42549/48008 [2:26:09<17:54, 5.08it/s] 89%|████████▊ | 42550/48008 [2:26:10<17:46, 5.12it/s] {'loss': 0.3004, 'grad_norm': 0.330078125, 'learning_rate': 2.274204299283453e-05, 'epoch': 0.89} 89%|████████▊ | 42550/48008 [2:26:10<17:46, 5.12it/s] 89%|████████▊ | 42551/48008 [2:26:10<17:39, 5.15it/s] 89%|████████▊ | 42552/48008 [2:26:10<17:37, 5.16it/s] 89%|████████▊ | 42553/48008 [2:26:10<17:55, 5.07it/s] 89%|████████▊ | 42554/48008 [2:26:10<17:39, 5.15it/s] 89%|████████▊ | 42555/48008 [2:26:11<17:27, 5.21it/s] 89%|████████▊ | 42556/48008 [2:26:11<17:29, 5.19it/s] 89%|████████▊ | 42557/48008 [2:26:11<17:15, 5.26it/s] 89%|████████▊ | 42558/48008 [2:26:11<17:36, 5.16it/s] 89%|████████▊ | 42559/48008 [2:26:11<17:59, 5.05it/s] 89%|████████▊ | 42560/48008 [2:26:12<18:04, 5.02it/s] 89%|████████▊ | 42561/48008 [2:26:12<18:06, 5.01it/s] 89%|████████▊ | 42562/48008 [2:26:12<17:44, 5.11it/s] 89%|████████▊ | 42563/48008 [2:26:12<17:38, 5.14it/s] 89%|████████▊ | 42564/48008 [2:26:12<17:34, 5.16it/s] 89%|████████▊ | 42565/48008 [2:26:13<17:33, 5.17it/s] 89%|████████▊ | 42566/48008 [2:26:13<17:29, 5.18it/s] 89%|████████▊ | 42567/48008 [2:26:13<17:46, 5.10it/s] 89%|████████▊ | 42568/48008 [2:26:13<17:43, 5.11it/s] 89%|████████▊ | 42569/48008 [2:26:13<17:48, 5.09it/s] 89%|████████▊ | 42570/48008 [2:26:14<23:00, 3.94it/s] 89%|████████▊ | 42571/48008 [2:26:14<21:35, 4.20it/s] 89%|████████▊ | 42572/48008 [2:26:14<20:30, 4.42it/s] 89%|████████▊ | 42573/48008 [2:26:14<19:27, 4.65it/s] 89%|████████▊ | 42574/48008 [2:26:15<38:00, 2.38it/s] 89%|████████▊ | 42575/48008 [2:26:15<31:45, 2.85it/s] 89%|████████▊ | 42576/48008 [2:26:16<27:25, 3.30it/s] 89%|████████▊ | 42577/48008 [2:26:16<24:40, 3.67it/s] 89%|████████▊ | 42578/48008 [2:26:16<22:24, 4.04it/s] 89%|████████▊ | 42579/48008 [2:26:16<20:45, 4.36it/s] 89%|████████▊ | 42580/48008 [2:26:16<19:44, 4.58it/s] 89%|████████▊ | 42581/48008 [2:26:17<19:03, 4.75it/s] 89%|████████▊ | 42582/48008 [2:26:17<18:31, 4.88it/s] 89%|████████▊ | 42583/48008 [2:26:17<18:36, 4.86it/s] 89%|████████▊ | 42584/48008 [2:26:17<18:17, 4.94it/s] 89%|████████▊ | 42585/48008 [2:26:17<17:54, 5.05it/s] 89%|████████▊ | 42586/48008 [2:26:18<34:59, 2.58it/s] 89%|████████▊ | 42587/48008 [2:26:18<29:49, 3.03it/s] 89%|████████▊ | 42588/48008 [2:26:19<26:00, 3.47it/s] 89%|████████▊ | 42589/48008 [2:26:19<23:30, 3.84it/s] 89%|████████▊ | 42590/48008 [2:26:19<21:32, 4.19it/s] 89%|████████▊ | 42591/48008 [2:26:19<20:16, 4.45it/s] 89%|████████▊ | 42592/48008 [2:26:20<42:05, 2.14it/s] 89%|████████▊ | 42593/48008 [2:26:20<34:27, 2.62it/s] 89%|████████▊ | 42594/48008 [2:26:21<29:17, 3.08it/s] 89%|████████▊ | 42595/48008 [2:26:21<25:35, 3.53it/s] 89%|████████▊ | 42596/48008 [2:26:21<23:04, 3.91it/s] 89%|████████▊ | 42597/48008 [2:26:21<21:14, 4.25it/s] 89%|████████▊ | 42598/48008 [2:26:22<46:58, 1.92it/s] 89%|████████▊ | 42599/48008 [2:26:23<37:53, 2.38it/s] 89%|████████▊ | 42600/48008 [2:26:23<31:44, 2.84it/s] {'loss': 0.2994, 'grad_norm': 0.1943359375, 'learning_rate': 2.2533744375937347e-05, 'epoch': 0.89} 89%|████████▊ | 42600/48008 [2:26:23<31:44, 2.84it/s] 89%|████████▊ | 42601/48008 [2:26:23<27:41, 3.26it/s] 89%|████████▊ | 42602/48008 [2:26:23<24:30, 3.68it/s] 89%|████████▊ | 42603/48008 [2:26:23<22:38, 3.98it/s] 89%|████████▊ | 42604/48008 [2:26:24<21:16, 4.23it/s] 89%|████████▊ | 42605/48008 [2:26:24<20:04, 4.49it/s] 89%|████████▊ | 42606/48008 [2:26:24<19:03, 4.73it/s] 89%|████████▊ | 42607/48008 [2:26:24<19:43, 4.56it/s] 89%|████████▉ | 42608/48008 [2:26:24<18:48, 4.79it/s] 89%|████████▉ | 42609/48008 [2:26:25<21:26, 4.20it/s] 89%|████████▉ | 42610/48008 [2:26:25<20:05, 4.48it/s] 89%|████████▉ | 42611/48008 [2:26:25<19:08, 4.70it/s] 89%|████████▉ | 42612/48008 [2:26:25<19:03, 4.72it/s] 89%|████████▉ | 42613/48008 [2:26:25<18:27, 4.87it/s] 89%|████████▉ | 42614/48008 [2:26:26<18:08, 4.95it/s] 89%|████████▉ | 42615/48008 [2:26:26<17:47, 5.05it/s] 89%|████████▉ | 42616/48008 [2:26:26<17:58, 5.00it/s] 89%|████████▉ | 42617/48008 [2:26:26<17:39, 5.09it/s] 89%|████████▉ | 42618/48008 [2:26:26<17:30, 5.13it/s] 89%|████████▉ | 42619/48008 [2:26:27<17:21, 5.18it/s] 89%|████████▉ | 42620/48008 [2:26:27<17:31, 5.13it/s] 89%|████████▉ | 42621/48008 [2:26:27<17:27, 5.14it/s] 89%|████████▉ | 42622/48008 [2:26:27<17:22, 5.17it/s] 89%|████████▉ | 42623/48008 [2:26:27<17:14, 5.20it/s] 89%|████████▉ | 42624/48008 [2:26:28<17:22, 5.16it/s] 89%|████████▉ | 42625/48008 [2:26:28<17:42, 5.07it/s] 89%|████████▉ | 42626/48008 [2:26:28<17:31, 5.12it/s] 89%|████████▉ | 42627/48008 [2:26:28<17:45, 5.05it/s] 89%|████████▉ | 42628/48008 [2:26:28<17:24, 5.15it/s] 89%|████████▉ | 42629/48008 [2:26:29<20:24, 4.39it/s] 89%|████████▉ | 42630/48008 [2:26:29<19:22, 4.63it/s] 89%|████████▉ | 42631/48008 [2:26:29<18:37, 4.81it/s] 89%|████████▉ | 42632/48008 [2:26:29<18:07, 4.94it/s] 89%|████████▉ | 42633/48008 [2:26:29<17:50, 5.02it/s] 89%|████████▉ | 42634/48008 [2:26:30<17:32, 5.11it/s] 89%|████████▉ | 42635/48008 [2:26:30<17:37, 5.08it/s] 89%|████████▉ | 42636/48008 [2:26:30<17:32, 5.10it/s] 89%|████████▉ | 42637/48008 [2:26:30<17:22, 5.15it/s] 89%|████████▉ | 42638/48008 [2:26:30<17:11, 5.20it/s] 89%|████████▉ | 42639/48008 [2:26:31<17:01, 5.25it/s] 89%|████████▉ | 42640/48008 [2:26:31<18:15, 4.90it/s] 89%|████████▉ | 42641/48008 [2:26:31<17:55, 4.99it/s] 89%|████████▉ | 42642/48008 [2:26:31<17:36, 5.08it/s] 89%|████████▉ | 42643/48008 [2:26:31<17:14, 5.19it/s] 89%|████████▉ | 42644/48008 [2:26:32<17:11, 5.20it/s] 89%|████████▉ | 42645/48008 [2:26:32<17:07, 5.22it/s] 89%|████████▉ | 42646/48008 [2:26:32<17:28, 5.11it/s] 89%|████████▉ | 42647/48008 [2:26:32<17:18, 5.16it/s] 89%|████████▉ | 42648/48008 [2:26:32<17:17, 5.17it/s] 89%|████████▉ | 42649/48008 [2:26:32<17:19, 5.15it/s] 89%|████████▉ | 42650/48008 [2:26:33<17:37, 5.07it/s] {'loss': 0.2999, 'grad_norm': 0.203125, 'learning_rate': 2.232544575904016e-05, 'epoch': 0.89} 89%|████████▉ | 42650/48008 [2:26:33<17:37, 5.07it/s] 89%|████████▉ | 42651/48008 [2:26:33<17:30, 5.10it/s] 89%|████████▉ | 42652/48008 [2:26:34<34:29, 2.59it/s] 89%|████████▉ | 42653/48008 [2:26:34<29:22, 3.04it/s] 89%|████████▉ | 42654/48008 [2:26:34<25:36, 3.48it/s] 89%|████████▉ | 42655/48008 [2:26:34<22:58, 3.88it/s] 89%|████████▉ | 42656/48008 [2:26:34<21:11, 4.21it/s] 89%|████████▉ | 42657/48008 [2:26:35<20:07, 4.43it/s] 89%|████████▉ | 42658/48008 [2:26:35<20:23, 4.37it/s] 89%|████████▉ | 42659/48008 [2:26:35<19:33, 4.56it/s] 89%|████████▉ | 42660/48008 [2:26:35<20:03, 4.44it/s] 89%|████████▉ | 42661/48008 [2:26:36<19:14, 4.63it/s] 89%|████████▉ | 42662/48008 [2:26:36<18:21, 4.85it/s] 89%|████████▉ | 42663/48008 [2:26:36<18:05, 4.93it/s] 89%|████████▉ | 42664/48008 [2:26:36<17:40, 5.04it/s] 89%|████████▉ | 42665/48008 [2:26:36<17:26, 5.11it/s] 89%|████████▉ | 42666/48008 [2:26:36<17:16, 5.16it/s] 89%|████████▉ | 42667/48008 [2:26:37<17:03, 5.22it/s] 89%|████████▉ | 42668/48008 [2:26:37<17:16, 5.15it/s] 89%|████████▉ | 42669/48008 [2:26:37<17:32, 5.07it/s] 89%|████████▉ | 42670/48008 [2:26:37<17:23, 5.12it/s] 89%|████████▉ | 42671/48008 [2:26:37<17:32, 5.07it/s] 89%|████████▉ | 42672/48008 [2:26:38<17:20, 5.13it/s] 89%|████████▉ | 42673/48008 [2:26:38<17:43, 5.02it/s] 89%|████████▉ | 42674/48008 [2:26:38<17:20, 5.13it/s] 89%|████████▉ | 42675/48008 [2:26:38<17:02, 5.21it/s] 89%|████████▉ | 42676/48008 [2:26:38<16:58, 5.24it/s] 89%|████████▉ | 42677/48008 [2:26:39<16:58, 5.24it/s] 89%|████████▉ | 42678/48008 [2:26:39<17:06, 5.19it/s] 89%|████████▉ | 42679/48008 [2:26:39<17:23, 5.11it/s] 89%|████████▉ | 42680/48008 [2:26:39<17:13, 5.15it/s] 89%|████████▉ | 42681/48008 [2:26:39<17:01, 5.21it/s] 89%|████████▉ | 42682/48008 [2:26:40<16:46, 5.29it/s] 89%|████████▉ | 42683/48008 [2:26:40<16:49, 5.27it/s] 89%|████████▉ | 42684/48008 [2:26:40<16:53, 5.25it/s] 89%|████████▉ | 42685/48008 [2:26:40<17:00, 5.22it/s] 89%|████████▉ | 42686/48008 [2:26:40<17:15, 5.14it/s] 89%|████████▉ | 42687/48008 [2:26:41<17:09, 5.17it/s] 89%|████████▉ | 42688/48008 [2:26:41<17:15, 5.14it/s] 89%|████████▉ | 42689/48008 [2:26:41<17:02, 5.20it/s] 89%|████████▉ | 42690/48008 [2:26:41<17:06, 5.18it/s] 89%|████████▉ | 42691/48008 [2:26:41<17:10, 5.16it/s] 89%|████████▉ | 42692/48008 [2:26:42<17:07, 5.17it/s] 89%|████████▉ | 42693/48008 [2:26:42<17:21, 5.11it/s] 89%|████████▉ | 42694/48008 [2:26:42<17:19, 5.11it/s] 89%|████████▉ | 42695/48008 [2:26:42<17:25, 5.08it/s] 89%|████████▉ | 42696/48008 [2:26:42<17:12, 5.14it/s] 89%|████████▉ | 42697/48008 [2:26:43<22:14, 3.98it/s] 89%|████████▉ | 42698/48008 [2:26:43<20:43, 4.27it/s] 89%|████████▉ | 42699/48008 [2:26:43<19:30, 4.54it/s] 89%|████████▉ | 42700/48008 [2:26:43<18:41, 4.73it/s] {'loss': 0.3014, 'grad_norm': 0.279296875, 'learning_rate': 2.2117147142142978e-05, 'epoch': 0.89} 89%|████████▉ | 42700/48008 [2:26:43<18:41, 4.73it/s] 89%|████████▉ | 42701/48008 [2:26:43<18:34, 4.76it/s] 89%|████████▉ | 42702/48008 [2:26:44<18:13, 4.85it/s] 89%|████████▉ | 42703/48008 [2:26:44<17:40, 5.00it/s] 89%|████████▉ | 42704/48008 [2:26:44<17:25, 5.07it/s] 89%|████████▉ | 42705/48008 [2:26:44<17:22, 5.09it/s] 89%|████████▉ | 42706/48008 [2:26:44<17:11, 5.14it/s] 89%|████████▉ | 42707/48008 [2:26:45<17:01, 5.19it/s] 89%|████████▉ | 42708/48008 [2:26:45<17:03, 5.18it/s] 89%|████████▉ | 42709/48008 [2:26:45<16:59, 5.20it/s] 89%|████████▉ | 42710/48008 [2:26:45<17:13, 5.13it/s] 89%|████████▉ | 42711/48008 [2:26:45<17:44, 4.98it/s] 89%|████████▉ | 42712/48008 [2:26:46<22:41, 3.89it/s] 89%|████████▉ | 42713/48008 [2:26:46<20:59, 4.20it/s] 89%|████████▉ | 42714/48008 [2:26:46<19:45, 4.47it/s] 89%|████████▉ | 42715/48008 [2:26:46<18:56, 4.66it/s] 89%|████████▉ | 42716/48008 [2:26:47<18:33, 4.75it/s] 89%|████████▉ | 42717/48008 [2:26:47<17:58, 4.90it/s] 89%|████████▉ | 42718/48008 [2:26:47<17:30, 5.03it/s] 89%|████████▉ | 42719/48008 [2:26:47<17:15, 5.11it/s] 89%|████████▉ | 42720/48008 [2:26:47<18:21, 4.80it/s] 89%|████████▉ | 42721/48008 [2:26:48<17:50, 4.94it/s] 89%|████████▉ | 42722/48008 [2:26:48<17:30, 5.03it/s] 89%|████████▉ | 42723/48008 [2:26:48<17:08, 5.14it/s] 89%|████████▉ | 42724/48008 [2:26:48<16:58, 5.19it/s] 89%|████████▉ | 42725/48008 [2:26:48<16:51, 5.23it/s] 89%|████████▉ | 42726/48008 [2:26:49<16:57, 5.19it/s] 89%|████████▉ | 42727/48008 [2:26:49<16:54, 5.21it/s] 89%|████████▉ | 42728/48008 [2:26:49<17:01, 5.17it/s] 89%|████████▉ | 42729/48008 [2:26:49<17:00, 5.17it/s] 89%|████████▉ | 42730/48008 [2:26:49<17:03, 5.16it/s] 89%|████████▉ | 42731/48008 [2:26:49<16:57, 5.19it/s] 89%|████████▉ | 42732/48008 [2:26:50<22:04, 3.98it/s] 89%|████████▉ | 42733/48008 [2:26:50<20:37, 4.26it/s] 89%|████████▉ | 42734/48008 [2:26:50<19:28, 4.51it/s] 89%|████████▉ | 42735/48008 [2:26:50<18:49, 4.67it/s] 89%|████████▉ | 42736/48008 [2:26:51<18:30, 4.75it/s] 89%|████████▉ | 42737/48008 [2:26:51<17:58, 4.89it/s] 89%|████████▉ | 42738/48008 [2:26:51<17:30, 5.02it/s] 89%|████████▉ | 42739/48008 [2:26:51<17:19, 5.07it/s] 89%|████████▉ | 42740/48008 [2:26:51<17:00, 5.16it/s] 89%|████████▉ | 42741/48008 [2:26:52<16:56, 5.18it/s] 89%|████████▉ | 42742/48008 [2:26:52<16:58, 5.17it/s] 89%|████████▉ | 42743/48008 [2:26:52<17:03, 5.14it/s] 89%|████████▉ | 42744/48008 [2:26:52<17:01, 5.15it/s] 89%|████████▉ | 42745/48008 [2:26:52<16:51, 5.21it/s] 89%|████████▉ | 42746/48008 [2:26:53<17:18, 5.07it/s] 89%|████████▉ | 42747/48008 [2:26:53<17:02, 5.14it/s] 89%|████████▉ | 42748/48008 [2:26:53<16:56, 5.18it/s] 89%|████████▉ | 42749/48008 [2:26:53<16:56, 5.17it/s] 89%|████████▉ | 42750/48008 [2:26:53<17:11, 5.10it/s] 89%|████████▉ | 42750/48008 [2:26:53<17:11, 5.10it/s]{'loss': 0.2987, 'grad_norm': 0.345703125, 'learning_rate': 2.1908848525245792e-05, 'epoch': 0.89} 89%|████████▉ | 42751/48008 [2:26:54<17:22, 5.04it/s] 89%|████████▉ | 42752/48008 [2:26:54<17:29, 5.01it/s] 89%|████████▉ | 42753/48008 [2:26:54<17:16, 5.07it/s] 89%|████████▉ | 42754/48008 [2:26:54<17:36, 4.97it/s] 89%|████████▉ | 42755/48008 [2:26:54<17:39, 4.96it/s] 89%|████████▉ | 42756/48008 [2:26:55<17:07, 5.11it/s] 89%|████████▉ | 42757/48008 [2:26:55<16:54, 5.18it/s] 89%|████████▉ | 42758/48008 [2:26:55<16:56, 5.16it/s] 89%|████████▉ | 42759/48008 [2:26:55<16:52, 5.18it/s] 89%|████████▉ | 42760/48008 [2:26:55<16:34, 5.28it/s] 89%|████████▉ | 42761/48008 [2:26:55<16:46, 5.21it/s] 89%|████████▉ | 42762/48008 [2:26:56<16:53, 5.17it/s] 89%|████████▉ | 42763/48008 [2:26:56<17:07, 5.11it/s] 89%|████████▉ | 42764/48008 [2:26:56<16:51, 5.18it/s] 89%|████████▉ | 42765/48008 [2:26:56<17:10, 5.09it/s] 89%|████████▉ | 42766/48008 [2:26:56<16:57, 5.15it/s] 89%|████████▉ | 42767/48008 [2:26:57<16:53, 5.17it/s] 89%|████████▉ | 42768/48008 [2:26:57<16:45, 5.21it/s] 89%|████████▉ | 42769/48008 [2:26:57<16:46, 5.21it/s] 89%|████████▉ | 42770/48008 [2:26:57<17:17, 5.05it/s] 89%|████████▉ | 42771/48008 [2:26:57<17:25, 5.01it/s] 89%|████████▉ | 42772/48008 [2:26:58<17:08, 5.09it/s] 89%|████████▉ | 42773/48008 [2:26:58<17:04, 5.11it/s] 89%|████████▉ | 42774/48008 [2:26:58<17:19, 5.03it/s] 89%|████████▉ | 42775/48008 [2:26:58<17:07, 5.09it/s] 89%|████████▉ | 42776/48008 [2:26:58<16:57, 5.14it/s] 89%|████████▉ | 42777/48008 [2:26:59<17:11, 5.07it/s] 89%|████████▉ | 42778/48008 [2:26:59<16:57, 5.14it/s] 89%|████████▉ | 42779/48008 [2:26:59<16:57, 5.14it/s] 89%|████████▉ | 42780/48008 [2:26:59<16:48, 5.18it/s] 89%|████████▉ | 42781/48008 [2:26:59<16:49, 5.18it/s] 89%|████████▉ | 42782/48008 [2:27:00<17:01, 5.12it/s] 89%|████████▉ | 42783/48008 [2:27:00<16:46, 5.19it/s] 89%|████████▉ | 42784/48008 [2:27:00<16:38, 5.23it/s] 89%|████████▉ | 42785/48008 [2:27:00<17:51, 4.88it/s] 89%|████████▉ | 42786/48008 [2:27:00<17:27, 4.99it/s] 89%|████████▉ | 42787/48008 [2:27:01<17:20, 5.02it/s] 89%|████████▉ | 42788/48008 [2:27:01<17:07, 5.08it/s] 89%|████████▉ | 42789/48008 [2:27:01<17:00, 5.12it/s] 89%|████████▉ | 42790/48008 [2:27:01<16:47, 5.18it/s] 89%|████████▉ | 42791/48008 [2:27:01<16:46, 5.18it/s] 89%|████████▉ | 42792/48008 [2:27:02<16:41, 5.21it/s] 89%|████████▉ | 42793/48008 [2:27:02<17:00, 5.11it/s] 89%|████████▉ | 42794/48008 [2:27:02<16:49, 5.17it/s] 89%|████████▉ | 42795/48008 [2:27:02<17:03, 5.09it/s] 89%|████████▉ | 42796/48008 [2:27:02<16:54, 5.14it/s] 89%|████████▉ | 42797/48008 [2:27:03<16:40, 5.21it/s] 89%|████████▉ | 42798/48008 [2:27:03<16:42, 5.20it/s] 89%|████████▉ | 42799/48008 [2:27:03<17:06, 5.08it/s] 89%|████████▉ | 42800/48008 [2:27:03<16:56, 5.12it/s] {'loss': 0.3002, 'grad_norm': 0.203125, 'learning_rate': 2.170054990834861e-05, 'epoch': 0.89} 89%|████████▉ | 42800/48008 [2:27:03<16:56, 5.12it/s] 89%|████████▉ | 42801/48008 [2:27:03<16:54, 5.13it/s] 89%|████████▉ | 42802/48008 [2:27:03<16:46, 5.17it/s] 89%|████████▉ | 42803/48008 [2:27:04<17:00, 5.10it/s] 89%|████████▉ | 42804/48008 [2:27:04<17:11, 5.05it/s] 89%|████████▉ | 42805/48008 [2:27:04<16:55, 5.13it/s] 89%|████████▉ | 42806/48008 [2:27:04<16:47, 5.16it/s] 89%|████████▉ | 42807/48008 [2:27:04<16:44, 5.18it/s] 89%|████████▉ | 42808/48008 [2:27:05<16:47, 5.16it/s] 89%|████████▉ | 42809/48008 [2:27:05<16:44, 5.18it/s] 89%|████████▉ | 42810/48008 [2:27:05<16:33, 5.23it/s] 89%|████████▉ | 42811/48008 [2:27:05<16:31, 5.24it/s] 89%|████████▉ | 42812/48008 [2:27:05<16:36, 5.22it/s] 89%|████████▉ | 42813/48008 [2:27:06<16:35, 5.22it/s] 89%|████████▉ | 42814/48008 [2:27:06<16:35, 5.22it/s] 89%|████████▉ | 42815/48008 [2:27:06<16:27, 5.26it/s] 89%|████████▉ | 42816/48008 [2:27:06<17:41, 4.89it/s] 89%|████████▉ | 42817/48008 [2:27:06<17:39, 4.90it/s] 89%|████████▉ | 42818/48008 [2:27:07<17:31, 4.94it/s] 89%|████████▉ | 42819/48008 [2:27:07<17:30, 4.94it/s] 89%|████████▉ | 42820/48008 [2:27:07<17:31, 4.94it/s] 89%|████████▉ | 42821/48008 [2:27:07<17:12, 5.02it/s] 89%|████████▉ | 42822/48008 [2:27:07<16:58, 5.09it/s] 89%|████████▉ | 42823/48008 [2:27:08<16:42, 5.17it/s] 89%|████████▉ | 42824/48008 [2:27:08<16:35, 5.21it/s] 89%|████████▉ | 42825/48008 [2:27:08<16:34, 5.21it/s] 89%|████████▉ | 42826/48008 [2:27:08<16:23, 5.27it/s] 89%|████████▉ | 42827/48008 [2:27:08<16:36, 5.20it/s] 89%|████████▉ | 42828/48008 [2:27:09<16:32, 5.22it/s] 89%|████████▉ | 42829/48008 [2:27:09<16:55, 5.10it/s] 89%|████████▉ | 42830/48008 [2:27:09<17:15, 5.00it/s] 89%|████████▉ | 42831/48008 [2:27:09<16:55, 5.10it/s] 89%|████████▉ | 42832/48008 [2:27:09<16:46, 5.14it/s] 89%|████████▉ | 42833/48008 [2:27:10<16:33, 5.21it/s] 89%|████████▉ | 42834/48008 [2:27:10<16:25, 5.25it/s] 89%|████████▉ | 42835/48008 [2:27:10<16:26, 5.25it/s] 89%|████████▉ | 42836/48008 [2:27:10<16:39, 5.17it/s] 89%|████████▉ | 42837/48008 [2:27:10<16:35, 5.20it/s] 89%|████████▉ | 42838/48008 [2:27:11<16:50, 5.12it/s] 89%|████████▉ | 42839/48008 [2:27:11<16:44, 5.14it/s] 89%|████████▉ | 42840/48008 [2:27:11<16:47, 5.13it/s] 89%|████████▉ | 42841/48008 [2:27:11<16:39, 5.17it/s] 89%|████████▉ | 42842/48008 [2:27:11<16:33, 5.20it/s] 89%|████████▉ | 42843/48008 [2:27:11<16:31, 5.21it/s] 89%|████████▉ | 42844/48008 [2:27:12<16:49, 5.12it/s] 89%|████████▉ | 42845/48008 [2:27:12<16:45, 5.13it/s] 89%|████████▉ | 42846/48008 [2:27:12<16:46, 5.13it/s] 89%|████████▉ | 42847/48008 [2:27:12<21:40, 3.97it/s] 89%|████████▉ | 42848/48008 [2:27:13<20:08, 4.27it/s] 89%|████████▉ | 42849/48008 [2:27:14<37:03, 2.32it/s] 89%|████████▉ | 42850/48008 [2:27:14<32:00, 2.69it/s] {'loss': 0.3002, 'grad_norm': 0.244140625, 'learning_rate': 2.1492251291451427e-05, 'epoch': 0.89} 89%|████████▉ | 42850/48008 [2:27:14<32:00, 2.69it/s] 89%|████████▉ | 42851/48008 [2:27:14<27:28, 3.13it/s] 89%|████████▉ | 42852/48008 [2:27:14<24:05, 3.57it/s] 89%|████████▉ | 42853/48008 [2:27:14<21:41, 3.96it/s] 89%|████████▉ | 42854/48008 [2:27:15<19:58, 4.30it/s] 89%|████████▉ | 42855/48008 [2:27:15<18:57, 4.53it/s] 89%|████████▉ | 42856/48008 [2:27:15<18:20, 4.68it/s] 89%|████████▉ | 42857/48008 [2:27:15<17:44, 4.84it/s] 89%|████████▉ | 42858/48008 [2:27:15<17:24, 4.93it/s] 89%|████████▉ | 42859/48008 [2:27:15<17:12, 4.99it/s] 89%|████████▉ | 42860/48008 [2:27:16<17:15, 4.97it/s] 89%|████████▉ | 42861/48008 [2:27:16<17:11, 4.99it/s] 89%|████████▉ | 42862/48008 [2:27:16<16:55, 5.07it/s] 89%|████████▉ | 42863/48008 [2:27:16<16:42, 5.13it/s] 89%|████████▉ | 42864/48008 [2:27:16<16:55, 5.06it/s] 89%|████████▉ | 42865/48008 [2:27:17<16:38, 5.15it/s] 89%|████████▉ | 42866/48008 [2:27:17<16:29, 5.20it/s] 89%|████████▉ | 42867/48008 [2:27:17<16:58, 5.05it/s] 89%|████████▉ | 42868/48008 [2:27:17<16:54, 5.07it/s] 89%|████████▉ | 42869/48008 [2:27:17<16:42, 5.13it/s] 89%|████████▉ | 42870/48008 [2:27:18<16:34, 5.17it/s] 89%|████████▉ | 42871/48008 [2:27:18<16:27, 5.20it/s] 89%|████████▉ | 42872/48008 [2:27:18<16:18, 5.25it/s] 89%|████████▉ | 42873/48008 [2:27:18<16:36, 5.15it/s] 89%|████████▉ | 42874/48008 [2:27:18<16:27, 5.20it/s] 89%|████████▉ | 42875/48008 [2:27:19<16:16, 5.25it/s] 89%|████████▉ | 42876/48008 [2:27:19<16:40, 5.13it/s] 89%|████████▉ | 42877/48008 [2:27:19<16:30, 5.18it/s] 89%|████████▉ | 42878/48008 [2:27:19<16:36, 5.15it/s] 89%|████████▉ | 42879/48008 [2:27:19<16:53, 5.06it/s] 89%|████████▉ | 42880/48008 [2:27:20<16:42, 5.11it/s] 89%|████████▉ | 42881/48008 [2:27:20<16:52, 5.06it/s] 89%|████████▉ | 42882/48008 [2:27:20<16:35, 5.15it/s] 89%|████████▉ | 42883/48008 [2:27:20<16:29, 5.18it/s] 89%|████████▉ | 42884/48008 [2:27:20<16:20, 5.23it/s] 89%|████████▉ | 42885/48008 [2:27:21<16:36, 5.14it/s] 89%|████████▉ | 42886/48008 [2:27:21<16:31, 5.16it/s] 89%|████████▉ | 42887/48008 [2:27:21<16:33, 5.15it/s] 89%|████████▉ | 42888/48008 [2:27:21<16:29, 5.17it/s] 89%|████████▉ | 42889/48008 [2:27:21<16:14, 5.25it/s] 89%|████████▉ | 42890/48008 [2:27:22<16:13, 5.26it/s] 89%|████████▉ | 42891/48008 [2:27:22<21:12, 4.02it/s] 89%|████████▉ | 42892/48008 [2:27:22<19:44, 4.32it/s] 89%|████████▉ | 42893/48008 [2:27:22<18:53, 4.51it/s] 89%|████████▉ | 42894/48008 [2:27:22<18:04, 4.72it/s] 89%|████████▉ | 42895/48008 [2:27:23<17:32, 4.86it/s] 89%|████████▉ | 42896/48008 [2:27:23<17:15, 4.94it/s] 89%|████████▉ | 42897/48008 [2:27:23<16:54, 5.04it/s] 89%|████████▉ | 42898/48008 [2:27:23<16:43, 5.09it/s] 89%|████████▉ | 42899/48008 [2:27:23<16:47, 5.07it/s] 89%|████████▉ | 42900/48008 [2:27:24<16:38, 5.12it/s] {'loss': 0.3012, 'grad_norm': 0.271484375, 'learning_rate': 2.128395267455424e-05, 'epoch': 0.89} 89%|████████▉ | 42900/48008 [2:27:24<16:38, 5.12it/s] 89%|████████▉ | 42901/48008 [2:27:24<16:30, 5.16it/s] 89%|████████▉ | 42902/48008 [2:27:24<16:41, 5.10it/s] 89%|████████▉ | 42903/48008 [2:27:24<16:36, 5.12it/s] 89%|████████▉ | 42904/48008 [2:27:24<16:24, 5.18it/s] 89%|████████▉ | 42905/48008 [2:27:25<16:19, 5.21it/s] 89%|████████▉ | 42906/48008 [2:27:25<16:18, 5.22it/s] 89%|████████▉ | 42907/48008 [2:27:25<16:08, 5.27it/s] 89%|████████▉ | 42908/48008 [2:27:25<16:07, 5.27it/s] 89%|████████▉ | 42909/48008 [2:27:25<16:06, 5.28it/s] 89%|████████▉ | 42910/48008 [2:27:26<19:05, 4.45it/s] 89%|████████▉ | 42911/48008 [2:27:26<21:08, 4.02it/s] 89%|████████▉ | 42912/48008 [2:27:26<19:47, 4.29it/s] 89%|████████▉ | 42913/48008 [2:27:26<18:43, 4.54it/s] 89%|████████▉ | 42914/48008 [2:27:27<17:56, 4.73it/s] 89%|████████▉ | 42915/48008 [2:27:27<17:22, 4.88it/s] 89%|████████▉ | 42916/48008 [2:27:27<17:08, 4.95it/s] 89%|████████▉ | 42917/48008 [2:27:27<16:52, 5.03it/s] 89%|████████▉ | 42918/48008 [2:27:27<16:57, 5.00it/s] 89%|████████▉ | 42919/48008 [2:27:28<17:02, 4.98it/s] 89%|████████▉ | 42920/48008 [2:27:28<17:04, 4.97it/s] 89%|████████▉ | 42921/48008 [2:27:28<16:45, 5.06it/s] 89%|████████▉ | 42922/48008 [2:27:28<16:45, 5.06it/s] 89%|████████▉ | 42923/48008 [2:27:28<16:26, 5.16it/s] 89%|████████▉ | 42924/48008 [2:27:28<16:16, 5.21it/s] 89%|████████▉ | 42925/48008 [2:27:29<16:16, 5.20it/s] 89%|████████▉ | 42926/48008 [2:27:29<16:15, 5.21it/s] 89%|████████▉ | 42927/48008 [2:27:29<16:25, 5.16it/s] 89%|████████▉ | 42928/48008 [2:27:29<16:23, 5.17it/s] 89%|████████▉ | 42929/48008 [2:27:29<16:18, 5.19it/s] 89%|████████▉ | 42930/48008 [2:27:30<16:41, 5.07it/s] 89%|████████▉ | 42931/48008 [2:27:30<16:26, 5.15it/s] 89%|████████▉ | 42932/48008 [2:27:30<16:16, 5.20it/s] 89%|████████▉ | 42933/48008 [2:27:30<16:07, 5.24it/s] 89%|████████▉ | 42934/48008 [2:27:30<16:08, 5.24it/s] 89%|████████▉ | 42935/48008 [2:27:31<19:03, 4.43it/s] 89%|████████▉ | 42936/48008 [2:27:31<18:40, 4.53it/s] 89%|████████▉ | 42937/48008 [2:27:31<17:58, 4.70it/s] 89%|████████▉ | 42938/48008 [2:27:31<17:51, 4.73it/s] 89%|████████▉ | 42939/48008 [2:27:32<17:16, 4.89it/s] 89%|████████▉ | 42940/48008 [2:27:32<16:58, 4.98it/s] 89%|████████▉ | 42941/48008 [2:27:32<17:01, 4.96it/s] 89%|████████▉ | 42942/48008 [2:27:32<21:44, 3.88it/s] 89%|████████▉ | 42943/48008 [2:27:32<20:01, 4.22it/s] 89%|████████▉ | 42944/48008 [2:27:33<18:48, 4.49it/s] 89%|████████▉ | 42945/48008 [2:27:33<18:00, 4.69it/s] 89%|████████▉ | 42946/48008 [2:27:33<17:22, 4.86it/s] 89%|████████▉ | 42947/48008 [2:27:33<16:59, 4.97it/s] 89%|████████▉ | 42948/48008 [2:27:33<16:47, 5.02it/s] 89%|████████▉ | 42949/48008 [2:27:34<16:47, 5.02it/s] 89%|████████▉ | 42950/48008 [2:27:34<21:36, 3.90it/s] {'loss': 0.2998, 'grad_norm': 0.3515625, 'learning_rate': 2.1075654057657058e-05, 'epoch': 0.89} 89%|████████▉ | 42950/48008 [2:27:34<21:36, 3.90it/s] 89%|████████▉ | 42951/48008 [2:27:34<20:05, 4.20it/s] 89%|████████▉ | 42952/48008 [2:27:34<19:11, 4.39it/s] 89%|████████▉ | 42953/48008 [2:27:35<18:17, 4.61it/s] 89%|████████▉ | 42954/48008 [2:27:35<17:55, 4.70it/s] 89%|████████▉ | 42955/48008 [2:27:35<17:19, 4.86it/s] 89%|████████▉ | 42956/48008 [2:27:35<16:55, 4.97it/s] 89%|████████▉ | 42957/48008 [2:27:35<16:34, 5.08it/s] 89%|████████▉ | 42958/48008 [2:27:36<16:42, 5.04it/s] 89%|████████▉ | 42959/48008 [2:27:36<16:29, 5.10it/s] 89%|████████▉ | 42960/48008 [2:27:36<16:13, 5.19it/s] 89%|████████▉ | 42961/48008 [2:27:36<16:12, 5.19it/s] 89%|████████▉ | 42962/48008 [2:27:36<16:08, 5.21it/s] 89%|████████▉ | 42963/48008 [2:27:37<16:21, 5.14it/s] 89%|████████▉ | 42964/48008 [2:27:37<16:11, 5.19it/s] 89%|████████▉ | 42965/48008 [2:27:37<16:08, 5.21it/s] 89%|████████▉ | 42966/48008 [2:27:37<16:04, 5.23it/s] 89%|████████▉ | 42967/48008 [2:27:37<16:15, 5.17it/s] 90%|████████▉ | 42968/48008 [2:27:38<16:22, 5.13it/s] 90%|████████▉ | 42969/48008 [2:27:38<16:27, 5.10it/s] 90%|████████▉ | 42970/48008 [2:27:38<16:20, 5.14it/s] 90%|████████▉ | 42971/48008 [2:27:38<16:21, 5.13it/s] 90%|████████▉ | 42972/48008 [2:27:38<16:14, 5.17it/s] 90%|████████▉ | 42973/48008 [2:27:38<16:17, 5.15it/s] 90%|████████▉ | 42974/48008 [2:27:39<16:25, 5.11it/s] 90%|████████▉ | 42975/48008 [2:27:39<16:28, 5.09it/s] 90%|████████▉ | 42976/48008 [2:27:39<16:29, 5.09it/s] 90%|████████▉ | 42977/48008 [2:27:39<16:17, 5.14it/s] 90%|████████▉ | 42978/48008 [2:27:39<16:12, 5.17it/s] 90%|████████▉ | 42979/48008 [2:27:40<16:05, 5.21it/s] 90%|████████▉ | 42980/48008 [2:27:40<15:56, 5.26it/s] 90%|████████▉ | 42981/48008 [2:27:40<15:55, 5.26it/s] 90%|████████▉ | 42982/48008 [2:27:40<15:48, 5.30it/s] 90%|████████▉ | 42983/48008 [2:27:40<15:45, 5.31it/s] 90%|████████▉ | 42984/48008 [2:27:41<15:46, 5.31it/s] 90%|████████▉ | 42985/48008 [2:27:41<16:07, 5.19it/s] 90%|████████▉ | 42986/48008 [2:27:41<17:14, 4.85it/s] 90%|████████▉ | 42987/48008 [2:27:41<16:54, 4.95it/s] 90%|████████▉ | 42988/48008 [2:27:41<16:38, 5.03it/s] 90%|████████▉ | 42989/48008 [2:27:42<16:32, 5.06it/s] 90%|████████▉ | 42990/48008 [2:27:42<16:22, 5.11it/s] 90%|████████▉ | 42991/48008 [2:27:42<16:18, 5.13it/s] 90%|████████▉ | 42992/48008 [2:27:42<16:08, 5.18it/s] 90%|████████▉ | 42993/48008 [2:27:42<18:57, 4.41it/s] 90%|████████▉ | 42994/48008 [2:27:43<17:50, 4.69it/s] 90%|████████▉ | 42995/48008 [2:27:43<22:12, 3.76it/s] 90%|████████▉ | 42996/48008 [2:27:43<20:35, 4.06it/s] 90%|████████▉ | 42997/48008 [2:27:43<19:36, 4.26it/s] 90%|████████▉ | 42998/48008 [2:27:44<18:43, 4.46it/s] 90%|████████▉ | 42999/48008 [2:27:44<17:56, 4.65it/s] 90%|████████▉ | 43000/48008 [2:27:44<17:24, 4.80it/s] {'loss': 0.3009, 'grad_norm': 0.2265625, 'learning_rate': 2.0867355440759875e-05, 'epoch': 0.9} 90%|████████▉ | 43000/48008 [2:27:44<17:24, 4.80it/s] 90%|████████▉ | 43001/48008 [2:27:44<16:53, 4.94it/s] 90%|████████▉ | 43002/48008 [2:27:44<16:31, 5.05it/s] 90%|████████▉ | 43003/48008 [2:27:45<16:27, 5.07it/s] 90%|████████▉ | 43004/48008 [2:27:45<16:07, 5.17it/s] 90%|████████▉ | 43005/48008 [2:27:45<16:22, 5.09it/s] 90%|████████▉ | 43006/48008 [2:27:45<16:19, 5.11it/s] 90%|████████▉ | 43007/48008 [2:27:45<16:13, 5.14it/s] 90%|████████▉ | 43008/48008 [2:27:46<16:35, 5.02it/s] 90%|████████▉ | 43009/48008 [2:27:46<16:15, 5.12it/s] 90%|████████▉ | 43010/48008 [2:27:46<16:21, 5.09it/s] 90%|████████▉ | 43011/48008 [2:27:46<17:24, 4.78it/s] 90%|████████▉ | 43012/48008 [2:27:46<17:23, 4.79it/s] 90%|████████▉ | 43013/48008 [2:27:47<16:50, 4.95it/s] 90%|████████▉ | 43014/48008 [2:27:47<16:30, 5.04it/s] 90%|████████▉ | 43015/48008 [2:27:47<16:47, 4.96it/s] 90%|████████▉ | 43016/48008 [2:27:47<16:34, 5.02it/s] 90%|████████▉ | 43017/48008 [2:27:47<16:23, 5.08it/s] 90%|████████▉ | 43018/48008 [2:27:48<16:14, 5.12it/s] 90%|████████▉ | 43019/48008 [2:27:48<16:10, 5.14it/s] 90%|████████▉ | 43020/48008 [2:27:48<16:02, 5.18it/s] 90%|████████▉ | 43021/48008 [2:27:48<16:02, 5.18it/s] 90%|████████▉ | 43022/48008 [2:27:48<16:09, 5.14it/s] 90%|████████▉ | 43023/48008 [2:27:49<19:00, 4.37it/s] 90%|████████▉ | 43024/48008 [2:27:49<18:02, 4.61it/s] 90%|████████▉ | 43025/48008 [2:27:49<17:25, 4.77it/s] 90%|████████▉ | 43026/48008 [2:27:49<16:45, 4.95it/s] 90%|████████▉ | 43027/48008 [2:27:49<16:28, 5.04it/s] 90%|████████▉ | 43028/48008 [2:27:50<16:43, 4.96it/s] 90%|████████▉ | 43029/48008 [2:27:50<16:20, 5.08it/s] 90%|████████▉ | 43030/48008 [2:27:50<16:12, 5.12it/s] 90%|████████▉ | 43031/48008 [2:27:50<16:04, 5.16it/s] 90%|████████▉ | 43032/48008 [2:27:50<15:48, 5.25it/s] 90%|████████▉ | 43033/48008 [2:27:51<16:06, 5.15it/s] 90%|████████▉ | 43034/48008 [2:27:51<15:50, 5.24it/s] 90%|████████▉ | 43035/48008 [2:27:51<16:00, 5.18it/s] 90%|████████▉ | 43036/48008 [2:27:51<15:58, 5.19it/s] 90%|████████▉ | 43037/48008 [2:27:51<16:13, 5.10it/s] 90%|████████▉ | 43038/48008 [2:27:52<16:11, 5.11it/s] 90%|████████▉ | 43039/48008 [2:27:52<16:20, 5.07it/s] 90%|████████▉ | 43040/48008 [2:27:52<19:04, 4.34it/s] 90%|████████▉ | 43041/48008 [2:27:52<18:04, 4.58it/s] 90%|████████▉ | 43042/48008 [2:27:52<17:37, 4.69it/s] 90%|████████▉ | 43043/48008 [2:27:53<17:14, 4.80it/s] 90%|████████▉ | 43044/48008 [2:27:53<17:03, 4.85it/s] 90%|████████▉ | 43045/48008 [2:27:53<16:50, 4.91it/s] 90%|████████▉ | 43046/48008 [2:27:53<16:35, 4.98it/s] 90%|████████▉ | 43047/48008 [2:27:53<16:12, 5.10it/s] 90%|████████▉ | 43048/48008 [2:27:54<16:06, 5.13it/s] 90%|████████▉ | 43049/48008 [2:27:54<16:13, 5.09it/s] 90%|████████▉ | 43050/48008 [2:27:54<16:24, 5.04it/s] 90%|████████▉ | 43050/48008 [2:27:54<16:24, 5.04it/s]{'loss': 0.3013, 'grad_norm': 0.1513671875, 'learning_rate': 2.065905682386269e-05, 'epoch': 0.9} 90%|████████▉ | 43051/48008 [2:27:54<16:22, 5.05it/s] 90%|████████▉ | 43052/48008 [2:27:54<16:16, 5.08it/s] 90%|████████▉ | 43053/48008 [2:27:55<16:02, 5.15it/s] 90%|████████▉ | 43054/48008 [2:27:55<16:14, 5.08it/s] 90%|████████▉ | 43055/48008 [2:27:55<16:07, 5.12it/s] 90%|████████▉ | 43056/48008 [2:27:55<16:24, 5.03it/s] 90%|████████▉ | 43057/48008 [2:27:55<16:23, 5.03it/s] 90%|████████▉ | 43058/48008 [2:27:56<16:23, 5.03it/s] 90%|████████▉ | 43059/48008 [2:27:56<16:15, 5.07it/s] 90%|████████▉ | 43060/48008 [2:27:56<16:08, 5.11it/s] 90%|████████▉ | 43061/48008 [2:27:56<15:56, 5.17it/s] 90%|████████▉ | 43062/48008 [2:27:56<15:49, 5.21it/s] 90%|████████▉ | 43063/48008 [2:27:57<15:51, 5.20it/s] 90%|████████▉ | 43064/48008 [2:27:57<15:46, 5.22it/s] 90%|████████▉ | 43065/48008 [2:27:57<15:57, 5.16it/s] 90%|████████▉ | 43066/48008 [2:27:57<15:50, 5.20it/s] 90%|████████▉ | 43067/48008 [2:27:57<16:17, 5.05it/s] 90%|████████▉ | 43068/48008 [2:27:58<16:05, 5.12it/s] 90%|████████▉ | 43069/48008 [2:27:58<15:57, 5.16it/s] 90%|████████▉ | 43070/48008 [2:27:58<16:22, 5.02it/s] 90%|████████▉ | 43071/48008 [2:27:58<16:14, 5.07it/s] 90%|████████▉ | 43072/48008 [2:27:58<16:09, 5.09it/s] 90%|████████▉ | 43073/48008 [2:27:59<20:51, 3.94it/s] 90%|████████▉ | 43074/48008 [2:27:59<19:36, 4.19it/s] 90%|████████▉ | 43075/48008 [2:27:59<18:28, 4.45it/s] 90%|████████▉ | 43076/48008 [2:27:59<17:42, 4.64it/s] 90%|████████▉ | 43077/48008 [2:28:00<17:27, 4.71it/s] 90%|████████▉ | 43078/48008 [2:28:00<17:08, 4.79it/s] 90%|████████▉ | 43079/48008 [2:28:00<16:53, 4.86it/s] 90%|████████▉ | 43080/48008 [2:28:00<16:29, 4.98it/s] 90%|████████▉ | 43081/48008 [2:28:00<16:16, 5.05it/s] 90%|████████▉ | 43082/48008 [2:28:00<16:06, 5.10it/s] 90%|████████▉ | 43083/48008 [2:28:01<17:07, 4.79it/s] 90%|████████▉ | 43084/48008 [2:28:01<17:07, 4.79it/s] 90%|████████▉ | 43085/48008 [2:28:01<16:44, 4.90it/s] 90%|████████▉ | 43086/48008 [2:28:01<16:23, 5.01it/s] 90%|████████▉ | 43087/48008 [2:28:02<16:13, 5.06it/s] 90%|████████▉ | 43088/48008 [2:28:02<16:01, 5.12it/s] 90%|████████▉ | 43089/48008 [2:28:02<15:48, 5.19it/s] 90%|████████▉ | 43090/48008 [2:28:02<15:43, 5.21it/s] 90%|████████▉ | 43091/48008 [2:28:02<15:50, 5.17it/s] 90%|████████▉ | 43092/48008 [2:28:02<15:48, 5.18it/s] 90%|████████▉ | 43093/48008 [2:28:03<15:50, 5.17it/s] 90%|████████▉ | 43094/48008 [2:28:03<15:44, 5.20it/s] 90%|████████▉ | 43095/48008 [2:28:03<15:43, 5.21it/s] 90%|████████▉ | 43096/48008 [2:28:03<15:29, 5.28it/s] 90%|████████▉ | 43097/48008 [2:28:03<15:53, 5.15it/s] 90%|████████▉ | 43098/48008 [2:28:04<15:47, 5.18it/s] 90%|████████▉ | 43099/48008 [2:28:04<15:40, 5.22it/s] 90%|████████▉ | 43100/48008 [2:28:04<15:36, 5.24it/s] {'loss': 0.3006, 'grad_norm': 0.240234375, 'learning_rate': 2.0450758206965507e-05, 'epoch': 0.9} 90%|████████▉ | 43100/48008 [2:28:04<15:36, 5.24it/s] 90%|████████▉ | 43101/48008 [2:28:04<16:45, 4.88it/s] 90%|████████▉ | 43102/48008 [2:28:04<16:45, 4.88it/s] 90%|████████▉ | 43103/48008 [2:28:05<16:22, 4.99it/s] 90%|████████▉ | 43104/48008 [2:28:05<16:26, 4.97it/s] 90%|████████▉ | 43105/48008 [2:28:05<19:01, 4.29it/s] 90%|████████▉ | 43106/48008 [2:28:05<17:54, 4.56it/s] 90%|████████▉ | 43107/48008 [2:28:06<17:28, 4.68it/s] 90%|████████▉ | 43108/48008 [2:28:06<16:54, 4.83it/s] 90%|████████▉ | 43109/48008 [2:28:06<16:24, 4.98it/s] 90%|████████▉ | 43110/48008 [2:28:06<17:18, 4.72it/s] 90%|████████▉ | 43111/48008 [2:28:06<17:02, 4.79it/s] 90%|████████▉ | 43112/48008 [2:28:07<16:34, 4.92it/s] 90%|████████▉ | 43113/48008 [2:28:07<16:15, 5.02it/s] 90%|████████▉ | 43114/48008 [2:28:07<16:02, 5.09it/s] 90%|████████▉ | 43115/48008 [2:28:07<15:59, 5.10it/s] 90%|████████▉ | 43116/48008 [2:28:07<15:51, 5.14it/s] 90%|████████▉ | 43117/48008 [2:28:08<15:56, 5.11it/s] 90%|████████▉ | 43118/48008 [2:28:08<15:52, 5.13it/s] 90%|████████▉ | 43119/48008 [2:28:08<16:05, 5.06it/s] 90%|████████▉ | 43120/48008 [2:28:08<16:02, 5.08it/s] 90%|████████▉ | 43121/48008 [2:28:08<17:04, 4.77it/s] 90%|████████▉ | 43122/48008 [2:28:09<16:47, 4.85it/s] 90%|████████▉ | 43123/48008 [2:28:09<16:23, 4.97it/s] 90%|████████▉ | 43124/48008 [2:28:09<16:09, 5.04it/s] 90%|████████▉ | 43125/48008 [2:28:09<16:07, 5.04it/s] 90%|████████▉ | 43126/48008 [2:28:09<16:05, 5.05it/s] 90%|████████▉ | 43127/48008 [2:28:10<16:05, 5.06it/s] 90%|████████▉ | 43128/48008 [2:28:10<16:03, 5.07it/s] 90%|████████▉ | 43129/48008 [2:28:10<15:57, 5.09it/s] 90%|████████▉ | 43130/48008 [2:28:10<15:46, 5.15it/s] 90%|████████▉ | 43131/48008 [2:28:10<18:30, 4.39it/s] 90%|████████▉ | 43132/48008 [2:28:11<18:03, 4.50it/s] 90%|████████▉ | 43133/48008 [2:28:11<17:09, 4.74it/s] 90%|████████▉ | 43134/48008 [2:28:11<16:44, 4.85it/s] 90%|████████▉ | 43135/48008 [2:28:11<16:25, 4.95it/s] 90%|████████▉ | 43136/48008 [2:28:11<16:10, 5.02it/s] 90%|████████▉ | 43137/48008 [2:28:12<15:59, 5.08it/s] 90%|████████▉ | 43138/48008 [2:28:12<15:43, 5.16it/s] 90%|████████▉ | 43139/48008 [2:28:12<15:54, 5.10it/s] 90%|████████▉ | 43140/48008 [2:28:12<15:47, 5.14it/s] 90%|████████▉ | 43141/48008 [2:28:12<15:59, 5.07it/s] 90%|████████▉ | 43142/48008 [2:28:13<16:08, 5.03it/s] 90%|████████▉ | 43143/48008 [2:28:13<16:09, 5.02it/s] 90%|████████▉ | 43144/48008 [2:28:13<16:01, 5.06it/s] 90%|████████▉ | 43145/48008 [2:28:13<15:47, 5.13it/s] 90%|████████▉ | 43146/48008 [2:28:13<15:40, 5.17it/s] 90%|████████▉ | 43147/48008 [2:28:14<15:38, 5.18it/s] 90%|████████▉ | 43148/48008 [2:28:14<15:36, 5.19it/s] 90%|████████▉ | 43149/48008 [2:28:14<15:39, 5.17it/s] 90%|████████▉ | 43150/48008 [2:28:14<15:38, 5.18it/s] {'loss': 0.2983, 'grad_norm': 0.296875, 'learning_rate': 2.024245959006832e-05, 'epoch': 0.9} 90%|████████▉ | 43150/48008 [2:28:14<15:38, 5.18it/s] 90%|████████▉ | 43151/48008 [2:28:14<15:35, 5.19it/s] 90%|████████▉ | 43152/48008 [2:28:14<15:34, 5.19it/s] 90%|████████▉ | 43153/48008 [2:28:15<15:33, 5.20it/s] 90%|████████▉ | 43154/48008 [2:28:15<15:30, 5.22it/s] 90%|████████▉ | 43155/48008 [2:28:15<15:31, 5.21it/s] 90%|████████▉ | 43156/48008 [2:28:15<15:36, 5.18it/s] 90%|████████▉ | 43157/48008 [2:28:15<15:32, 5.20it/s] 90%|████████▉ | 43158/48008 [2:28:16<15:31, 5.21it/s] 90%|████████▉ | 43159/48008 [2:28:16<15:57, 5.07it/s] 90%|████████▉ | 43160/48008 [2:28:16<15:56, 5.07it/s] 90%|████████▉ | 43161/48008 [2:28:16<15:48, 5.11it/s] 90%|████████▉ | 43162/48008 [2:28:16<15:46, 5.12it/s] 90%|████████▉ | 43163/48008 [2:28:17<15:34, 5.19it/s] 90%|████████▉ | 43164/48008 [2:28:17<15:57, 5.06it/s] 90%|████████▉ | 43165/48008 [2:28:17<15:45, 5.12it/s] 90%|████████▉ | 43166/48008 [2:28:17<15:39, 5.15it/s] 90%|████████▉ | 43167/48008 [2:28:17<15:27, 5.22it/s] 90%|████████▉ | 43168/48008 [2:28:18<15:16, 5.28it/s] 90%|████████▉ | 43169/48008 [2:28:18<15:26, 5.22it/s] 90%|████████▉ | 43170/48008 [2:28:18<15:30, 5.20it/s] 90%|████████▉ | 43171/48008 [2:28:18<18:16, 4.41it/s] 90%|████████▉ | 43172/48008 [2:28:18<17:33, 4.59it/s] 90%|████████▉ | 43173/48008 [2:28:20<41:06, 1.96it/s] 90%|████████▉ | 43174/48008 [2:28:20<33:33, 2.40it/s] 90%|████████▉ | 43175/48008 [2:28:20<32:48, 2.45it/s] 90%|████████▉ | 43176/48008 [2:28:20<27:44, 2.90it/s] 90%|████████▉ | 43177/48008 [2:28:21<24:00, 3.35it/s] 90%|████████▉ | 43178/48008 [2:28:21<21:38, 3.72it/s] 90%|████████▉ | 43179/48008 [2:28:21<20:00, 4.02it/s] 90%|████████▉ | 43180/48008 [2:28:21<18:31, 4.35it/s] 90%|████████▉ | 43181/48008 [2:28:21<17:28, 4.60it/s] 90%|████████▉ | 43182/48008 [2:28:22<17:04, 4.71it/s] 90%|████████▉ | 43183/48008 [2:28:22<16:35, 4.85it/s] 90%|████████▉ | 43184/48008 [2:28:22<16:07, 4.98it/s] 90%|████████▉ | 43185/48008 [2:28:22<15:50, 5.08it/s] 90%|████████▉ | 43186/48008 [2:28:23<20:22, 3.94it/s] 90%|████████▉ | 43187/48008 [2:28:23<35:46, 2.25it/s] 90%|████████▉ | 43188/48008 [2:28:24<29:39, 2.71it/s] 90%|████████▉ | 43189/48008 [2:28:24<25:16, 3.18it/s] 90%|████████▉ | 43190/48008 [2:28:24<22:21, 3.59it/s] 90%|████████▉ | 43191/48008 [2:28:24<20:23, 3.94it/s] 90%|████████▉ | 43192/48008 [2:28:24<18:47, 4.27it/s] 90%|████████▉ | 43193/48008 [2:28:25<17:45, 4.52it/s] 90%|████████▉ | 43194/48008 [2:28:25<16:51, 4.76it/s] 90%|████████▉ | 43195/48008 [2:28:25<16:15, 4.94it/s] 90%|████████▉ | 43196/48008 [2:28:25<15:59, 5.02it/s] 90%|████████▉ | 43197/48008 [2:28:25<15:42, 5.10it/s] 90%|████████▉ | 43198/48008 [2:28:26<15:35, 5.14it/s] 90%|████████▉ | 43199/48008 [2:28:26<30:44, 2.61it/s] 90%|████████▉ | 43200/48008 [2:28:27<26:28, 3.03it/s] {'loss': 0.302, 'grad_norm': 0.255859375, 'learning_rate': 2.003416097317114e-05, 'epoch': 0.9} 90%|████████▉ | 43200/48008 [2:28:27<26:28, 3.03it/s] 90%|████████▉ | 43201/48008 [2:28:27<23:09, 3.46it/s] 90%|████████▉ | 43202/48008 [2:28:27<21:13, 3.77it/s] 90%|████████▉ | 43203/48008 [2:28:27<19:32, 4.10it/s] 90%|████████▉ | 43204/48008 [2:28:27<18:16, 4.38it/s] 90%|████████▉ | 43205/48008 [2:28:28<36:12, 2.21it/s] 90%|████████▉ | 43206/48008 [2:28:29<29:57, 2.67it/s] 90%|████████▉ | 43207/48008 [2:28:29<25:19, 3.16it/s] 90%|█████████ | 43208/48008 [2:28:29<22:36, 3.54it/s] 90%|█████████ | 43209/48008 [2:28:29<20:28, 3.91it/s] 90%|█████████ | 43210/48008 [2:28:29<18:58, 4.21it/s] 90%|█████████ | 43211/48008 [2:28:30<41:50, 1.91it/s] 90%|█████████ | 43212/48008 [2:28:31<34:05, 2.35it/s] 90%|█████████ | 43213/48008 [2:28:31<28:19, 2.82it/s] 90%|█████████ | 43214/48008 [2:28:31<24:41, 3.24it/s] 90%|█████████ | 43215/48008 [2:28:31<21:47, 3.66it/s] 90%|█████████ | 43216/48008 [2:28:31<19:53, 4.01it/s] 90%|█████████ | 43217/48008 [2:28:32<18:24, 4.34it/s] 90%|█████████ | 43218/48008 [2:28:32<17:20, 4.60it/s] 90%|█████████ | 43219/48008 [2:28:32<16:51, 4.74it/s] 90%|█████████ | 43220/48008 [2:28:32<17:27, 4.57it/s] 90%|█████████ | 43221/48008 [2:28:32<16:54, 4.72it/s] 90%|█████████ | 43222/48008 [2:28:33<16:39, 4.79it/s] 90%|█████████ | 43223/48008 [2:28:33<16:19, 4.89it/s] 90%|█████████ | 43224/48008 [2:28:33<16:20, 4.88it/s] 90%|█████████ | 43225/48008 [2:28:33<15:57, 4.99it/s] 90%|█████████ | 43226/48008 [2:28:33<16:51, 4.73it/s] 90%|█████████ | 43227/48008 [2:28:34<16:37, 4.79it/s] 90%|█████████ | 43228/48008 [2:28:34<16:10, 4.92it/s] 90%|█████████ | 43229/48008 [2:28:34<15:46, 5.05it/s] 90%|█████████ | 43230/48008 [2:28:34<15:54, 5.00it/s] 90%|█████████ | 43231/48008 [2:28:35<16:47, 4.74it/s] 90%|█████████ | 43232/48008 [2:28:35<16:21, 4.87it/s] 90%|█████████ | 43233/48008 [2:28:35<16:00, 4.97it/s] 90%|█████████ | 43234/48008 [2:28:35<15:47, 5.04it/s] 90%|█████████ | 43235/48008 [2:28:35<15:30, 5.13it/s] 90%|█████████ | 43236/48008 [2:28:35<15:41, 5.07it/s] 90%|█████████ | 43237/48008 [2:28:36<15:35, 5.10it/s] 90%|█████████ | 43238/48008 [2:28:36<15:28, 5.14it/s] 90%|█████████ | 43239/48008 [2:28:36<15:22, 5.17it/s] 90%|█████████ | 43240/48008 [2:28:36<15:28, 5.13it/s] 90%|█████████ | 43241/48008 [2:28:36<15:23, 5.16it/s] 90%|█████████ | 43242/48008 [2:28:37<15:32, 5.11it/s] 90%|█████████ | 43243/48008 [2:28:37<15:28, 5.13it/s] 90%|█████████ | 43244/48008 [2:28:37<15:22, 5.16it/s] 90%|█████████ | 43245/48008 [2:28:37<15:24, 5.15it/s] 90%|█████████ | 43246/48008 [2:28:37<15:17, 5.19it/s] 90%|█████████ | 43247/48008 [2:28:38<15:14, 5.21it/s] 90%|█████████ | 43248/48008 [2:28:38<15:13, 5.21it/s] 90%|█████████ | 43249/48008 [2:28:38<17:54, 4.43it/s] 90%|█████████ | 43250/48008 [2:28:38<17:02, 4.65it/s] {'loss': 0.2988, 'grad_norm': 0.447265625, 'learning_rate': 1.9825862356273955e-05, 'epoch': 0.9} 90%|█████████ | 43250/48008 [2:28:38<17:02, 4.65it/s] 90%|█████████ | 43251/48008 [2:28:38<16:44, 4.73it/s] 90%|█████████ | 43252/48008 [2:28:39<16:47, 4.72it/s] 90%|█████████ | 43253/48008 [2:28:39<19:04, 4.15it/s] 90%|█████████ | 43254/48008 [2:28:39<17:54, 4.42it/s] 90%|█████████ | 43255/48008 [2:28:40<21:41, 3.65it/s] 90%|█████████ | 43256/48008 [2:28:40<19:54, 3.98it/s] 90%|█████████ | 43257/48008 [2:28:40<18:44, 4.22it/s] 90%|█████████ | 43258/48008 [2:28:40<17:38, 4.49it/s] 90%|█████████ | 43259/48008 [2:28:40<16:48, 4.71it/s] 90%|█████████ | 43260/48008 [2:28:41<16:34, 4.77it/s] 90%|█████████ | 43261/48008 [2:28:41<16:11, 4.88it/s] 90%|█████████ | 43262/48008 [2:28:41<15:54, 4.97it/s] 90%|█████████ | 43263/48008 [2:28:41<16:47, 4.71it/s] 90%|█████████ | 43264/48008 [2:28:41<16:33, 4.77it/s] 90%|█████████ | 43265/48008 [2:28:42<31:04, 2.54it/s] 90%|█████████ | 43266/48008 [2:28:42<26:36, 2.97it/s] 90%|█████████ | 43267/48008 [2:28:43<23:03, 3.43it/s] 90%|█████████ | 43268/48008 [2:28:43<20:34, 3.84it/s] 90%|█████████ | 43269/48008 [2:28:43<18:58, 4.16it/s] 90%|█████████ | 43270/48008 [2:28:43<17:51, 4.42it/s] 90%|█████████ | 43271/48008 [2:28:43<17:00, 4.64it/s] 90%|█████████ | 43272/48008 [2:28:44<16:50, 4.69it/s] 90%|█████████ | 43273/48008 [2:28:44<16:17, 4.85it/s] 90%|█████████ | 43274/48008 [2:28:44<16:06, 4.90it/s] 90%|█████████ | 43275/48008 [2:28:44<15:49, 4.98it/s] 90%|█████████ | 43276/48008 [2:28:44<15:27, 5.10it/s] 90%|█████████ | 43277/48008 [2:28:45<15:21, 5.14it/s] 90%|█████████ | 43278/48008 [2:28:45<15:06, 5.22it/s] 90%|█████████ | 43279/48008 [2:28:45<14:59, 5.26it/s] 90%|█████████ | 43280/48008 [2:28:45<15:00, 5.25it/s] 90%|█████████ | 43281/48008 [2:28:45<16:08, 4.88it/s] 90%|█████████ | 43282/48008 [2:28:46<15:51, 4.97it/s] 90%|█████████ | 43283/48008 [2:28:46<15:38, 5.04it/s] 90%|█████████ | 43284/48008 [2:28:46<15:30, 5.08it/s] 90%|█████████ | 43285/48008 [2:28:46<15:33, 5.06it/s] 90%|█████████ | 43286/48008 [2:28:46<15:36, 5.04it/s] 90%|█████████ | 43287/48008 [2:28:47<15:25, 5.10it/s] 90%|█████████ | 43288/48008 [2:28:47<15:13, 5.17it/s] 90%|█████████ | 43289/48008 [2:28:47<15:15, 5.16it/s] 90%|█████████ | 43290/48008 [2:28:47<15:03, 5.22it/s] 90%|█████████ | 43291/48008 [2:28:47<15:05, 5.21it/s] 90%|█████████ | 43292/48008 [2:28:47<14:56, 5.26it/s] 90%|█████████ | 43293/48008 [2:28:48<15:01, 5.23it/s] 90%|█████████ | 43294/48008 [2:28:48<14:59, 5.24it/s] 90%|█████████ | 43295/48008 [2:28:48<14:57, 5.25it/s] 90%|█████████ | 43296/48008 [2:28:48<14:55, 5.26it/s] 90%|█████████ | 43297/48008 [2:28:48<15:11, 5.17it/s] 90%|█████████ | 43298/48008 [2:28:49<15:08, 5.19it/s] 90%|█████████ | 43299/48008 [2:28:49<15:30, 5.06it/s] 90%|█████████ | 43300/48008 [2:28:49<15:41, 5.00it/s] {'loss': 0.3009, 'grad_norm': 0.271484375, 'learning_rate': 1.961756373937677e-05, 'epoch': 0.9} 90%|█████████ | 43300/48008 [2:28:49<15:41, 5.00it/s] 90%|█████████ | 43301/48008 [2:28:49<15:32, 5.05it/s] 90%|█████████ | 43302/48008 [2:28:49<15:39, 5.01it/s] 90%|█████████ | 43303/48008 [2:28:50<15:39, 5.01it/s] 90%|█████████ | 43304/48008 [2:28:50<16:31, 4.74it/s] 90%|█████████ | 43305/48008 [2:28:50<16:08, 4.86it/s] 90%|█████████ | 43306/48008 [2:28:50<15:40, 5.00it/s] 90%|█████████ | 43307/48008 [2:28:50<15:26, 5.07it/s] 90%|█████████ | 43308/48008 [2:28:51<15:39, 5.00it/s] 90%|█████████ | 43309/48008 [2:28:51<15:36, 5.02it/s] 90%|█████████ | 43310/48008 [2:28:51<15:27, 5.07it/s] 90%|█████████ | 43311/48008 [2:28:51<15:34, 5.03it/s] 90%|█████████ | 43312/48008 [2:28:51<15:16, 5.12it/s] 90%|█████████ | 43313/48008 [2:28:52<15:19, 5.11it/s] 90%|█████████ | 43314/48008 [2:28:52<15:09, 5.16it/s] 90%|█████████ | 43315/48008 [2:28:52<15:06, 5.18it/s] 90%|█████████ | 43316/48008 [2:28:52<15:15, 5.13it/s] 90%|█████████ | 43317/48008 [2:28:52<15:07, 5.17it/s] 90%|█████████ | 43318/48008 [2:28:53<14:56, 5.23it/s] 90%|█████████ | 43319/48008 [2:28:53<14:53, 5.25it/s] 90%|█████████ | 43320/48008 [2:28:53<14:49, 5.27it/s] 90%|█████████ | 43321/48008 [2:28:53<15:55, 4.90it/s] 90%|█████████ | 43322/48008 [2:28:53<15:32, 5.02it/s] 90%|█████████ | 43323/48008 [2:28:54<15:21, 5.08it/s] 90%|█████████ | 43324/48008 [2:28:54<15:12, 5.13it/s] 90%|█████████ | 43325/48008 [2:28:54<15:22, 5.08it/s] 90%|█████████ | 43326/48008 [2:28:54<15:16, 5.11it/s] 90%|█████████ | 43327/48008 [2:28:54<15:18, 5.09it/s] 90%|█████████ | 43328/48008 [2:28:55<15:23, 5.07it/s] 90%|█████████ | 43329/48008 [2:28:55<15:10, 5.14it/s] 90%|█████████ | 43330/48008 [2:28:55<15:14, 5.12it/s] 90%|█████████ | 43331/48008 [2:28:55<15:00, 5.19it/s] 90%|█████████ | 43332/48008 [2:28:55<15:05, 5.17it/s] 90%|█████████ | 43333/48008 [2:28:56<15:16, 5.10it/s] 90%|█████████ | 43334/48008 [2:28:56<15:05, 5.16it/s] 90%|█████████ | 43335/48008 [2:28:56<15:01, 5.18it/s] 90%|█████████ | 43336/48008 [2:28:56<15:01, 5.18it/s] 90%|█████████ | 43337/48008 [2:28:56<15:01, 5.18it/s] 90%|█████████ | 43338/48008 [2:28:56<14:52, 5.23it/s] 90%|█████████ | 43339/48008 [2:28:57<15:04, 5.16it/s] 90%|█████████ | 43340/48008 [2:28:57<15:01, 5.18it/s] 90%|█████████ | 43341/48008 [2:28:57<14:52, 5.23it/s] 90%|█████████ | 43342/48008 [2:28:57<16:02, 4.85it/s] 90%|█████████ | 43343/48008 [2:28:57<15:50, 4.91it/s] 90%|█████████ | 43344/48008 [2:28:58<15:25, 5.04it/s] 90%|█████████ | 43345/48008 [2:28:58<15:16, 5.09it/s] 90%|█████████ | 43346/48008 [2:28:58<16:13, 4.79it/s] 90%|█████████ | 43347/48008 [2:28:58<16:04, 4.83it/s] 90%|█████████ | 43348/48008 [2:28:59<15:49, 4.91it/s] 90%|█████████ | 43349/48008 [2:28:59<15:46, 4.92it/s] 90%|█████████ | 43350/48008 [2:28:59<15:43, 4.94it/s] {'loss': 0.3008, 'grad_norm': 0.1650390625, 'learning_rate': 1.9409265122479587e-05, 'epoch': 0.9} 90%|█████████ | 43350/48008 [2:28:59<15:43, 4.94it/s] 90%|█████████ | 43351/48008 [2:28:59<15:30, 5.01it/s] 90%|█████████ | 43352/48008 [2:28:59<16:25, 4.72it/s] 90%|█████████ | 43353/48008 [2:29:00<15:57, 4.86it/s] 90%|█████████ | 43354/48008 [2:29:00<15:52, 4.89it/s] 90%|█████████ | 43355/48008 [2:29:00<15:52, 4.88it/s] 90%|█████████ | 43356/48008 [2:29:00<15:32, 4.99it/s] 90%|█████████ | 43357/48008 [2:29:00<15:18, 5.06it/s] 90%|█████████ | 43358/48008 [2:29:01<15:11, 5.10it/s] 90%|█████████ | 43359/48008 [2:29:01<15:07, 5.12it/s] 90%|█████████ | 43360/48008 [2:29:01<17:53, 4.33it/s] 90%|█████████ | 43361/48008 [2:29:01<17:00, 4.55it/s] 90%|█████████ | 43362/48008 [2:29:01<16:20, 4.74it/s] 90%|█████████ | 43363/48008 [2:29:02<16:01, 4.83it/s] 90%|█████████ | 43364/48008 [2:29:02<15:45, 4.91it/s] 90%|█████████ | 43365/48008 [2:29:02<15:14, 5.08it/s] 90%|█████████ | 43366/48008 [2:29:02<15:03, 5.14it/s] 90%|█████████ | 43367/48008 [2:29:02<15:00, 5.15it/s] 90%|█████████ | 43368/48008 [2:29:03<15:10, 5.10it/s] 90%|█████████ | 43369/48008 [2:29:03<19:36, 3.94it/s] 90%|█████████ | 43370/48008 [2:29:03<18:00, 4.29it/s] 90%|█████████ | 43371/48008 [2:29:03<16:54, 4.57it/s] 90%|█████████ | 43372/48008 [2:29:04<16:18, 4.74it/s] 90%|█████████ | 43373/48008 [2:29:04<15:51, 4.87it/s] 90%|█████████ | 43374/48008 [2:29:04<15:31, 4.97it/s] 90%|█████████ | 43375/48008 [2:29:04<16:23, 4.71it/s] 90%|█████████ | 43376/48008 [2:29:04<15:55, 4.85it/s] 90%|█████████ | 43377/48008 [2:29:05<15:33, 4.96it/s] 90%|█████████ | 43378/48008 [2:29:05<19:50, 3.89it/s] 90%|█████████ | 43379/48008 [2:29:05<18:20, 4.21it/s] 90%|█████████ | 43380/48008 [2:29:05<17:12, 4.48it/s] 90%|█████████ | 43381/48008 [2:29:05<16:25, 4.69it/s] 90%|█████████ | 43382/48008 [2:29:06<15:52, 4.85it/s] 90%|█████████ | 43383/48008 [2:29:06<15:33, 4.95it/s] 90%|█████████ | 43384/48008 [2:29:06<15:35, 4.94it/s] 90%|█████████ | 43385/48008 [2:29:06<15:13, 5.06it/s] 90%|█████████ | 43386/48008 [2:29:06<15:06, 5.10it/s] 90%|█████████ | 43387/48008 [2:29:07<15:00, 5.13it/s] 90%|█████████ | 43388/48008 [2:29:07<14:50, 5.19it/s] 90%|█████████ | 43389/48008 [2:29:07<14:53, 5.17it/s] 90%|█████████ | 43390/48008 [2:29:07<15:08, 5.08it/s] 90%|█████████ | 43391/48008 [2:29:08<19:30, 3.94it/s] 90%|█████████ | 43392/48008 [2:29:08<19:07, 4.02it/s] 90%|█████████ | 43393/48008 [2:29:08<17:48, 4.32it/s] 90%|█████████ | 43394/48008 [2:29:08<16:54, 4.55it/s] 90%|█████████ | 43395/48008 [2:29:08<16:12, 4.74it/s] 90%|█████████ | 43396/48008 [2:29:09<15:57, 4.81it/s] 90%|█████████ | 43397/48008 [2:29:09<15:23, 4.99it/s] 90%|█████████ | 43398/48008 [2:29:09<15:09, 5.07it/s] 90%|█████████ | 43399/48008 [2:29:09<15:27, 4.97it/s] 90%|█████████ | 43400/48008 [2:29:09<15:11, 5.05it/s] {'loss': 0.2987, 'grad_norm': 0.2451171875, 'learning_rate': 1.9200966505582404e-05, 'epoch': 0.9} 90%|█████████ | 43400/48008 [2:29:09<15:11, 5.05it/s] 90%|█████████ | 43401/48008 [2:29:10<14:59, 5.12it/s] 90%|█████████ | 43402/48008 [2:29:10<15:08, 5.07it/s] 90%|█████████ | 43403/48008 [2:29:10<14:59, 5.12it/s] 90%|█████████ | 43404/48008 [2:29:10<14:51, 5.16it/s] 90%|█████████ | 43405/48008 [2:29:10<14:44, 5.20it/s] 90%|█████████ | 43406/48008 [2:29:11<15:46, 4.86it/s] 90%|█████████ | 43407/48008 [2:29:11<15:25, 4.97it/s] 90%|█████████ | 43408/48008 [2:29:11<15:09, 5.06it/s] 90%|█████████ | 43409/48008 [2:29:11<15:15, 5.02it/s] 90%|█████████ | 43410/48008 [2:29:11<15:05, 5.08it/s] 90%|█████████ | 43411/48008 [2:29:12<14:53, 5.14it/s] 90%|█████████ | 43412/48008 [2:29:12<14:49, 5.16it/s] 90%|█████████ | 43413/48008 [2:29:12<15:01, 5.10it/s] 90%|█████████ | 43414/48008 [2:29:12<14:53, 5.14it/s] 90%|█████████ | 43415/48008 [2:29:12<15:00, 5.10it/s] 90%|█████████ | 43416/48008 [2:29:13<14:52, 5.15it/s] 90%|█████████ | 43417/48008 [2:29:13<14:54, 5.13it/s] 90%|█████████ | 43418/48008 [2:29:13<14:47, 5.17it/s] 90%|█████████ | 43419/48008 [2:29:13<15:08, 5.05it/s] 90%|█████████ | 43420/48008 [2:29:13<15:00, 5.09it/s] 90%|█████████ | 43421/48008 [2:29:14<14:54, 5.13it/s] 90%|█████████ | 43422/48008 [2:29:14<15:06, 5.06it/s] 90%|█████████ | 43423/48008 [2:29:14<14:56, 5.11it/s] 90%|█████████ | 43424/48008 [2:29:14<14:49, 5.15it/s] 90%|█████████ | 43425/48008 [2:29:14<15:15, 5.00it/s] 90%|█████████ | 43426/48008 [2:29:14<15:03, 5.07it/s] 90%|█████████ | 43427/48008 [2:29:15<14:53, 5.13it/s] 90%|█████████ | 43428/48008 [2:29:15<14:46, 5.17it/s] 90%|█████████ | 43429/48008 [2:29:15<14:43, 5.18it/s] 90%|█████████ | 43430/48008 [2:29:15<14:44, 5.18it/s] 90%|█████████ | 43431/48008 [2:29:15<14:38, 5.21it/s] 90%|█████████ | 43432/48008 [2:29:16<14:30, 5.26it/s] 90%|█████████ | 43433/48008 [2:29:16<14:33, 5.24it/s] 90%|█████████ | 43434/48008 [2:29:16<14:53, 5.12it/s] 90%|█████████ | 43435/48008 [2:29:16<14:43, 5.18it/s] 90%|█████████ | 43436/48008 [2:29:16<14:41, 5.19it/s] 90%|█████████ | 43437/48008 [2:29:17<15:04, 5.05it/s] 90%|█████████ | 43438/48008 [2:29:17<15:15, 4.99it/s] 90%|█████████ | 43439/48008 [2:29:17<15:32, 4.90it/s] 90%|█████████ | 43440/48008 [2:29:17<15:16, 4.99it/s] 90%|█████████ | 43441/48008 [2:29:17<15:08, 5.03it/s] 90%|█████████ | 43442/48008 [2:29:18<16:03, 4.74it/s] 90%|█████████ | 43443/48008 [2:29:18<20:06, 3.78it/s] 90%|█████████ | 43444/48008 [2:29:18<18:17, 4.16it/s] 90%|█████████ | 43445/48008 [2:29:18<17:08, 4.44it/s] 90%|█████████ | 43446/48008 [2:29:19<16:22, 4.65it/s] 90%|█████████ | 43447/48008 [2:29:19<16:01, 4.74it/s] 91%|█████████ | 43448/48008 [2:29:19<20:01, 3.79it/s] 91%|█████████ | 43449/48008 [2:29:19<18:27, 4.12it/s] 91%|█████████ | 43450/48008 [2:29:20<17:32, 4.33it/s] {'loss': 0.2976, 'grad_norm': 0.25, 'learning_rate': 1.899266788868522e-05, 'epoch': 0.91} 91%|█████████ | 43450/48008 [2:29:20<17:32, 4.33it/s] 91%|█████████ | 43451/48008 [2:29:20<16:33, 4.59it/s] 91%|█████████ | 43452/48008 [2:29:20<15:58, 4.76it/s] 91%|█████████ | 43453/48008 [2:29:20<15:37, 4.86it/s] 91%|█████████ | 43454/48008 [2:29:20<15:15, 4.98it/s] 91%|█████████ | 43455/48008 [2:29:21<15:10, 5.00it/s] 91%|█████████ | 43456/48008 [2:29:21<19:28, 3.89it/s] 91%|█████████ | 43457/48008 [2:29:21<18:02, 4.21it/s] 91%|█████████ | 43458/48008 [2:29:21<17:08, 4.42it/s] 91%|█████████ | 43459/48008 [2:29:22<16:33, 4.58it/s] 91%|█████████ | 43460/48008 [2:29:22<15:55, 4.76it/s] 91%|█████████ | 43461/48008 [2:29:22<15:44, 4.82it/s] 91%|█████████ | 43462/48008 [2:29:22<15:16, 4.96it/s] 91%|█████████ | 43463/48008 [2:29:22<15:19, 4.95it/s] 91%|█████████ | 43464/48008 [2:29:23<15:15, 4.96it/s] 91%|█████████ | 43465/48008 [2:29:23<15:08, 5.00it/s] 91%|█████████ | 43466/48008 [2:29:23<14:49, 5.10it/s] 91%|█████████ | 43467/48008 [2:29:23<14:43, 5.14it/s] 91%|█████████ | 43468/48008 [2:29:23<14:40, 5.15it/s] 91%|█████████ | 43469/48008 [2:29:24<15:44, 4.80it/s] 91%|█████████ | 43470/48008 [2:29:24<15:15, 4.95it/s] 91%|█████████ | 43471/48008 [2:29:24<15:05, 5.01it/s] 91%|█████████ | 43472/48008 [2:29:24<14:55, 5.07it/s] 91%|█████████ | 43473/48008 [2:29:24<14:38, 5.16it/s] 91%|█████████ | 43474/48008 [2:29:24<14:40, 5.15it/s] 91%|█████████ | 43475/48008 [2:29:25<14:35, 5.18it/s] 91%|█████████ | 43476/48008 [2:29:25<14:34, 5.18it/s] 91%|█████████ | 43477/48008 [2:29:25<14:48, 5.10it/s] 91%|█████████ | 43478/48008 [2:29:25<14:45, 5.12it/s] 91%|█████████ | 43479/48008 [2:29:25<14:42, 5.13it/s] 91%|█████████ | 43480/48008 [2:29:26<14:38, 5.16it/s] 91%|█████████ | 43481/48008 [2:29:26<15:00, 5.02it/s] 91%|█████████ | 43482/48008 [2:29:26<19:18, 3.91it/s] 91%|█████████ | 43483/48008 [2:29:26<17:41, 4.26it/s] 91%|█████████ | 43484/48008 [2:29:27<17:07, 4.40it/s] 91%|█████████ | 43485/48008 [2:29:27<16:48, 4.48it/s] 91%|█████████ | 43486/48008 [2:29:27<17:12, 4.38it/s] 91%|█████████ | 43487/48008 [2:29:27<16:33, 4.55it/s] 91%|█████████ | 43488/48008 [2:29:28<16:04, 4.69it/s] 91%|█████████ | 43489/48008 [2:29:28<15:55, 4.73it/s] 91%|█████████ | 43490/48008 [2:29:28<15:20, 4.91it/s] 91%|█████████ | 43491/48008 [2:29:28<15:02, 5.00it/s] 91%|█████████ | 43492/48008 [2:29:28<14:44, 5.10it/s] 91%|█████████ | 43493/48008 [2:29:28<14:36, 5.15it/s] 91%|█████████ | 43494/48008 [2:29:29<14:37, 5.15it/s] 91%|█████████ | 43495/48008 [2:29:29<14:36, 5.15it/s] 91%|█████████ | 43496/48008 [2:29:29<14:34, 5.16it/s] 91%|█████████ | 43497/48008 [2:29:29<14:48, 5.08it/s] 91%|█████████ | 43498/48008 [2:29:29<14:40, 5.12it/s] 91%|█████████ | 43499/48008 [2:29:30<14:49, 5.07it/s] 91%|█████████ | 43500/48008 [2:29:30<14:41, 5.12it/s] {'loss': 0.299, 'grad_norm': 0.279296875, 'learning_rate': 1.8784369271788036e-05, 'epoch': 0.91} 91%|█████████ | 43500/48008 [2:29:30<14:41, 5.12it/s] 91%|█████████ | 43501/48008 [2:29:30<14:50, 5.06it/s] 91%|█████████ | 43502/48008 [2:29:30<14:39, 5.12it/s] 91%|█████████ | 43503/48008 [2:29:30<14:58, 5.01it/s] 91%|█████████ | 43504/48008 [2:29:31<14:56, 5.02it/s] 91%|█████████ | 43505/48008 [2:29:31<14:44, 5.09it/s] 91%|█████████ | 43506/48008 [2:29:31<14:45, 5.08it/s] 91%|█████████ | 43507/48008 [2:29:31<14:47, 5.07it/s] 91%|█████████ | 43508/48008 [2:29:31<14:37, 5.13it/s] 91%|█████████ | 43509/48008 [2:29:32<14:26, 5.19it/s] 91%|█████████ | 43510/48008 [2:29:32<14:38, 5.12it/s] 91%|█████████ | 43511/48008 [2:29:32<14:35, 5.14it/s] 91%|█████████ | 43512/48008 [2:29:32<14:47, 5.07it/s] 91%|█████████ | 43513/48008 [2:29:32<14:58, 5.00it/s] 91%|█████████ | 43514/48008 [2:29:33<14:52, 5.03it/s] 91%|█████████ | 43515/48008 [2:29:33<15:00, 4.99it/s] 91%|█████████ | 43516/48008 [2:29:33<14:46, 5.07it/s] 91%|█████████ | 43517/48008 [2:29:33<14:35, 5.13it/s] 91%|█████████ | 43518/48008 [2:29:33<14:32, 5.15it/s] 91%|█████████ | 43519/48008 [2:29:34<15:29, 4.83it/s] 91%|█████████ | 43520/48008 [2:29:34<15:21, 4.87it/s] 91%|█████████ | 43521/48008 [2:29:34<15:16, 4.90it/s] 91%|█████████ | 43522/48008 [2:29:34<14:55, 5.01it/s] 91%|█████████ | 43523/48008 [2:29:34<14:45, 5.06it/s] 91%|█████████ | 43524/48008 [2:29:35<14:49, 5.04it/s] 91%|█████████ | 43525/48008 [2:29:35<14:41, 5.08it/s] 91%|█████████ | 43526/48008 [2:29:35<14:45, 5.06it/s] 91%|█████████ | 43527/48008 [2:29:35<14:38, 5.10it/s] 91%|█████████ | 43528/48008 [2:29:36<18:55, 3.94it/s] 91%|█████████ | 43529/48008 [2:29:36<17:35, 4.24it/s] 91%|█████████ | 43530/48008 [2:29:36<16:52, 4.42it/s] 91%|█████████ | 43531/48008 [2:29:36<17:10, 4.35it/s] 91%|█████████ | 43532/48008 [2:29:37<18:54, 3.94it/s] 91%|█████████ | 43533/48008 [2:29:37<17:27, 4.27it/s] 91%|█████████ | 43534/48008 [2:29:37<16:26, 4.54it/s] 91%|█████████ | 43535/48008 [2:29:37<15:45, 4.73it/s] 91%|█████████ | 43536/48008 [2:29:37<15:22, 4.85it/s] 91%|█████████ | 43537/48008 [2:29:37<15:04, 4.94it/s] 91%|█████████ | 43538/48008 [2:29:38<14:51, 5.01it/s] 91%|█████████ | 43539/48008 [2:29:38<14:42, 5.06it/s] 91%|█████████ | 43540/48008 [2:29:38<14:50, 5.02it/s] 91%|█████████ | 43541/48008 [2:29:38<14:35, 5.10it/s] 91%|█████████ | 43542/48008 [2:29:39<17:05, 4.35it/s] 91%|█████████ | 43543/48008 [2:29:39<16:12, 4.59it/s] 91%|█████████ | 43544/48008 [2:29:39<15:36, 4.77it/s] 91%|█████████ | 43545/48008 [2:29:39<15:26, 4.82it/s] 91%|█████████ | 43546/48008 [2:29:39<16:07, 4.61it/s] 91%|█████████ | 43547/48008 [2:29:40<15:36, 4.76it/s] 91%|█████████ | 43548/48008 [2:29:40<15:14, 4.88it/s] 91%|█████████ | 43549/48008 [2:29:40<14:57, 4.97it/s] 91%|█████████ | 43550/48008 [2:29:40<19:04, 3.89it/s] 91%|█████████ | 43550/48008 [2:29:40<19:04, 3.89it/s]{'loss': 0.299, 'grad_norm': 0.671875, 'learning_rate': 1.8576070654890853e-05, 'epoch': 0.91} 91%|█████████ | 43551/48008 [2:29:41<21:58, 3.38it/s] 91%|█████████ | 43552/48008 [2:29:41<19:35, 3.79it/s] 91%|█████████ | 43553/48008 [2:29:41<18:16, 4.06it/s] 91%|█████████ | 43554/48008 [2:29:41<17:11, 4.32it/s] 91%|█████████ | 43555/48008 [2:29:42<16:19, 4.55it/s] 91%|█████████ | 43556/48008 [2:29:42<30:45, 2.41it/s] 91%|█████████ | 43557/48008 [2:29:43<26:09, 2.84it/s] 91%|█████████ | 43558/48008 [2:29:43<22:28, 3.30it/s] 91%|█████████ | 43559/48008 [2:29:43<19:58, 3.71it/s] 91%|█████████ | 43560/48008 [2:29:43<19:15, 3.85it/s] 91%|█████████ | 43561/48008 [2:29:43<17:46, 4.17it/s] 91%|█████████ | 43562/48008 [2:29:44<16:38, 4.45it/s] 91%|█████████ | 43563/48008 [2:29:44<15:52, 4.67it/s] 91%|█████████ | 43564/48008 [2:29:44<15:29, 4.78it/s] 91%|█████████ | 43565/48008 [2:29:44<15:24, 4.81it/s] 91%|█████████ | 43566/48008 [2:29:44<14:56, 4.96it/s] 91%|█████████ | 43567/48008 [2:29:45<14:52, 4.98it/s] 91%|█████████ | 43568/48008 [2:29:45<15:41, 4.71it/s] 91%|█████████ | 43569/48008 [2:29:45<15:03, 4.91it/s] 91%|█████████ | 43570/48008 [2:29:45<14:53, 4.97it/s] 91%|█████████ | 43571/48008 [2:29:45<14:35, 5.07it/s] 91%|█████████ | 43572/48008 [2:29:46<18:49, 3.93it/s] 91%|█████████ | 43573/48008 [2:29:46<17:23, 4.25it/s] 91%|█████████ | 43574/48008 [2:29:46<16:23, 4.51it/s] 91%|█████████ | 43575/48008 [2:29:46<15:44, 4.69it/s] 91%|█████████ | 43576/48008 [2:29:47<15:13, 4.85it/s] 91%|█████████ | 43577/48008 [2:29:47<14:51, 4.97it/s] 91%|█████████ | 43578/48008 [2:29:47<14:42, 5.02it/s] 91%|█████████ | 43579/48008 [2:29:47<18:51, 3.91it/s] 91%|█████████ | 43580/48008 [2:29:47<17:26, 4.23it/s] 91%|█████████ | 43581/48008 [2:29:48<16:23, 4.50it/s] 91%|█████████ | 43582/48008 [2:29:48<15:44, 4.69it/s] 91%|█████████ | 43583/48008 [2:29:48<15:15, 4.83it/s] 91%|█████████ | 43584/48008 [2:29:48<14:46, 4.99it/s] 91%|█████████ | 43585/48008 [2:29:48<14:37, 5.04it/s] 91%|█████████ | 43586/48008 [2:29:49<15:29, 4.76it/s] 91%|█████████ | 43587/48008 [2:29:49<14:57, 4.93it/s] 91%|█████████ | 43588/48008 [2:29:49<14:36, 5.04it/s] 91%|█████████ | 43589/48008 [2:29:49<14:33, 5.06it/s] 91%|█████████ | 43590/48008 [2:29:49<14:27, 5.09it/s] 91%|█████████ | 43591/48008 [2:29:50<14:45, 4.99it/s] 91%|█████████ | 43592/48008 [2:29:50<14:36, 5.04it/s] 91%|█████████ | 43593/48008 [2:29:50<14:28, 5.08it/s] 91%|█████████ | 43594/48008 [2:29:50<14:18, 5.14it/s] 91%|█████████ | 43595/48008 [2:29:50<14:13, 5.17it/s] 91%|█████████ | 43596/48008 [2:29:51<14:05, 5.22it/s] 91%|█████████ | 43597/48008 [2:29:51<14:09, 5.19it/s] 91%|█████████ | 43598/48008 [2:29:51<14:08, 5.20it/s] 91%|█████████ | 43599/48008 [2:29:51<14:10, 5.18it/s] 91%|█████████ | 43600/48008 [2:29:51<14:08, 5.19it/s] {'loss': 0.2995, 'grad_norm': 0.375, 'learning_rate': 1.836777203799367e-05, 'epoch': 0.91} 91%|█████████ | 43600/48008 [2:29:51<14:08, 5.19it/s] 91%|█████████ | 43601/48008 [2:29:52<14:07, 5.20it/s] 91%|█████████ | 43602/48008 [2:29:52<14:07, 5.20it/s] 91%|█████████ | 43603/48008 [2:29:52<15:07, 4.86it/s] 91%|█████████ | 43604/48008 [2:29:52<14:52, 4.93it/s] 91%|█████████ | 43605/48008 [2:29:52<14:37, 5.02it/s] 91%|█████████ | 43606/48008 [2:29:53<14:26, 5.08it/s] 91%|█████████ | 43607/48008 [2:29:53<14:19, 5.12it/s] 91%|█████████ | 43608/48008 [2:29:53<14:11, 5.17it/s] 91%|█████████ | 43609/48008 [2:29:53<14:36, 5.02it/s] 91%|█████████ | 43610/48008 [2:29:53<14:29, 5.06it/s] 91%|█████████ | 43611/48008 [2:29:54<14:16, 5.14it/s] 91%|█████████ | 43612/48008 [2:29:54<14:12, 5.16it/s] 91%|█████████ | 43613/48008 [2:29:54<14:15, 5.14it/s] 91%|█████████ | 43614/48008 [2:29:54<14:32, 5.04it/s] 91%|█████████ | 43615/48008 [2:29:54<15:26, 4.74it/s] 91%|█████████ | 43616/48008 [2:29:55<14:59, 4.88it/s] 91%|█████████ | 43617/48008 [2:29:55<14:57, 4.89it/s] 91%|█████████ | 43618/48008 [2:29:55<15:41, 4.66it/s] 91%|█████████ | 43619/48008 [2:29:55<15:13, 4.80it/s] 91%|█████████ | 43620/48008 [2:29:55<14:42, 4.97it/s] 91%|█████████ | 43621/48008 [2:29:56<14:30, 5.04it/s] 91%|█████████ | 43622/48008 [2:29:56<14:24, 5.07it/s] 91%|█████████ | 43623/48008 [2:29:56<14:17, 5.11it/s] 91%|█████████ | 43624/48008 [2:29:56<14:11, 5.15it/s] 91%|█████████ | 43625/48008 [2:29:56<14:02, 5.21it/s] 91%|█████████ | 43626/48008 [2:29:57<13:59, 5.22it/s] 91%|█████████ | 43627/48008 [2:29:57<14:05, 5.18it/s] 91%|█████████ | 43628/48008 [2:29:57<14:01, 5.21it/s] 91%|█████████ | 43629/48008 [2:29:57<13:58, 5.22it/s] 91%|█████████ | 43630/48008 [2:29:57<13:58, 5.22it/s] 91%|█████████ | 43631/48008 [2:29:58<13:58, 5.22it/s] 91%|█████████ | 43632/48008 [2:29:58<14:15, 5.11it/s] 91%|█████████ | 43633/48008 [2:29:58<14:25, 5.06it/s] 91%|█████████ | 43634/48008 [2:29:58<14:23, 5.07it/s] 91%|█████████ | 43635/48008 [2:29:58<14:14, 5.12it/s] 91%|█████████ | 43636/48008 [2:29:59<14:26, 5.04it/s] 91%|█████████ | 43637/48008 [2:29:59<14:11, 5.13it/s] 91%|█████████ | 43638/48008 [2:29:59<14:23, 5.06it/s] 91%|█████████ | 43639/48008 [2:29:59<14:18, 5.09it/s] 91%|█████████ | 43640/48008 [2:29:59<14:13, 5.12it/s] 91%|█████████ | 43641/48008 [2:29:59<14:09, 5.14it/s] 91%|█████████ | 43642/48008 [2:30:00<14:05, 5.17it/s] 91%|█████████ | 43643/48008 [2:30:00<14:19, 5.08it/s] 91%|█████████ | 43644/48008 [2:30:00<14:12, 5.12it/s] 91%|█████████ | 43645/48008 [2:30:00<14:08, 5.14it/s] 91%|█████████ | 43646/48008 [2:30:00<14:28, 5.02it/s] 91%|█████████ | 43647/48008 [2:30:01<14:33, 4.99it/s] 91%|█████████ | 43648/48008 [2:30:01<14:19, 5.07it/s] 91%|█████████ | 43649/48008 [2:30:01<14:43, 4.93it/s] 91%|█████████ | 43650/48008 [2:30:01<14:25, 5.04it/s] 91%|█████████ | 43650/48008 [2:30:01<14:25, 5.04it/s]{'loss': 0.2987, 'grad_norm': 0.1455078125, 'learning_rate': 1.8159473421096484e-05, 'epoch': 0.91} 91%|█████████ | 43651/48008 [2:30:01<14:14, 5.10it/s] 91%|█████████ | 43652/48008 [2:30:02<14:17, 5.08it/s] 91%|█████████ | 43653/48008 [2:30:02<14:12, 5.11it/s] 91%|█████████ | 43654/48008 [2:30:02<15:07, 4.80it/s] 91%|█████████ | 43655/48008 [2:30:02<14:43, 4.93it/s] 91%|█████████ | 43656/48008 [2:30:02<14:25, 5.03it/s] 91%|█████████ | 43657/48008 [2:30:03<14:40, 4.94it/s] 91%|█████████ | 43658/48008 [2:30:03<14:31, 4.99it/s] 91%|█████████ | 43659/48008 [2:30:03<14:19, 5.06it/s] 91%|█████████ | 43660/48008 [2:30:03<15:13, 4.76it/s] 91%|█████████ | 43661/48008 [2:30:03<14:49, 4.89it/s] 91%|█████████ | 43662/48008 [2:30:04<14:36, 4.96it/s] 91%|█████████ | 43663/48008 [2:30:04<14:30, 4.99it/s] 91%|█████████ | 43664/48008 [2:30:04<14:20, 5.05it/s] 91%|█████████ | 43665/48008 [2:30:04<14:13, 5.09it/s] 91%|█████████ | 43666/48008 [2:30:04<14:12, 5.09it/s] 91%|█████████ | 43667/48008 [2:30:05<14:13, 5.09it/s] 91%|█████████ | 43668/48008 [2:30:05<14:12, 5.09it/s] 91%|█████████ | 43669/48008 [2:30:05<18:18, 3.95it/s] 91%|█████████ | 43670/48008 [2:30:05<17:59, 4.02it/s] 91%|█████████ | 43671/48008 [2:30:06<17:08, 4.22it/s] 91%|█████████ | 43672/48008 [2:30:06<16:05, 4.49it/s] 91%|█████████ | 43673/48008 [2:30:06<15:38, 4.62it/s] 91%|█████████ | 43674/48008 [2:30:06<15:06, 4.78it/s] 91%|█████████ | 43675/48008 [2:30:06<14:39, 4.93it/s] 91%|█████████ | 43676/48008 [2:30:07<14:48, 4.88it/s] 91%|█████████ | 43677/48008 [2:30:07<14:27, 4.99it/s] 91%|█████████ | 43678/48008 [2:30:07<14:18, 5.04it/s] 91%|█████████ | 43679/48008 [2:30:07<14:14, 5.07it/s] 91%|█████████ | 43680/48008 [2:30:07<14:12, 5.08it/s] 91%|█████████ | 43681/48008 [2:30:08<14:01, 5.14it/s] 91%|█████████ | 43682/48008 [2:30:08<16:27, 4.38it/s] 91%|█████████ | 43683/48008 [2:30:08<15:40, 4.60it/s] 91%|█████████ | 43684/48008 [2:30:08<15:05, 4.77it/s] 91%|█████████ | 43685/48008 [2:30:09<14:55, 4.83it/s] 91%|█████████ | 43686/48008 [2:30:09<14:39, 4.91it/s] 91%|█████████ | 43687/48008 [2:30:09<14:19, 5.03it/s] 91%|█████████ | 43688/48008 [2:30:09<14:23, 5.00it/s] 91%|█████████ | 43689/48008 [2:30:09<14:12, 5.07it/s] 91%|█████████ | 43690/48008 [2:30:10<14:05, 5.11it/s] 91%|█████████ | 43691/48008 [2:30:10<14:26, 4.98it/s] 91%|█████████ | 43692/48008 [2:30:10<14:19, 5.02it/s] 91%|█████████ | 43693/48008 [2:30:10<14:21, 5.01it/s] 91%|█████████ | 43694/48008 [2:30:10<14:23, 5.00it/s] 91%|█████████ | 43695/48008 [2:30:11<14:21, 5.01it/s] 91%|█████████ | 43696/48008 [2:30:11<14:14, 5.05it/s] 91%|█████████ | 43697/48008 [2:30:11<14:21, 5.00it/s] 91%|█████████ | 43698/48008 [2:30:11<14:23, 4.99it/s] 91%|█████████ | 43699/48008 [2:30:11<14:10, 5.06it/s] 91%|█████████ | 43700/48008 [2:30:12<18:15, 3.93it/s] {'loss': 0.2957, 'grad_norm': 0.4453125, 'learning_rate': 1.79511748041993e-05, 'epoch': 0.91} 91%|█████████ | 43700/48008 [2:30:12<18:15, 3.93it/s] 91%|█████████ | 43701/48008 [2:30:12<17:00, 4.22it/s] 91%|█████████ | 43702/48008 [2:30:12<16:06, 4.45it/s] 91%|█████████ | 43703/48008 [2:30:12<15:26, 4.65it/s] 91%|█████████ | 43704/48008 [2:30:12<15:08, 4.74it/s] 91%|█████████ | 43705/48008 [2:30:13<15:46, 4.55it/s] 91%|█████████ | 43706/48008 [2:30:13<15:08, 4.74it/s] 91%|█████████ | 43707/48008 [2:30:13<14:44, 4.87it/s] 91%|█████████ | 43708/48008 [2:30:13<14:36, 4.91it/s] 91%|█████████ | 43709/48008 [2:30:13<14:15, 5.03it/s] 91%|█████████ | 43710/48008 [2:30:14<14:17, 5.01it/s] 91%|█████████ | 43711/48008 [2:30:14<14:21, 4.99it/s] 91%|█████████ | 43712/48008 [2:30:14<14:07, 5.07it/s] 91%|█████████ | 43713/48008 [2:30:14<14:00, 5.11it/s] 91%|█████████ | 43714/48008 [2:30:15<16:23, 4.37it/s] 91%|█████████ | 43715/48008 [2:30:15<15:44, 4.55it/s] 91%|█████████ | 43716/48008 [2:30:15<15:08, 4.72it/s] 91%|█████████ | 43717/48008 [2:30:15<14:47, 4.83it/s] 91%|█████████ | 43718/48008 [2:30:15<14:32, 4.92it/s] 91%|█████████ | 43719/48008 [2:30:16<16:42, 4.28it/s] 91%|█████████ | 43720/48008 [2:30:16<15:50, 4.51it/s] 91%|█████████ | 43721/48008 [2:30:16<15:18, 4.66it/s] 91%|█████████ | 43722/48008 [2:30:16<14:48, 4.83it/s] 91%|█████████ | 43723/48008 [2:30:16<14:46, 4.84it/s] 91%|█████████ | 43724/48008 [2:30:17<14:43, 4.85it/s] 91%|█████████ | 43725/48008 [2:30:17<14:42, 4.85it/s] 91%|█████████ | 43726/48008 [2:30:17<14:22, 4.97it/s] 91%|█████████ | 43727/48008 [2:30:17<14:33, 4.90it/s] 91%|█████████ | 43728/48008 [2:30:17<14:10, 5.03it/s] 91%|█████████ | 43729/48008 [2:30:18<14:01, 5.08it/s] 91%|█████████ | 43730/48008 [2:30:18<13:55, 5.12it/s] 91%|█████████ | 43731/48008 [2:30:18<13:46, 5.18it/s] 91%|█████████ | 43732/48008 [2:30:18<13:41, 5.20it/s] 91%|█████████ | 43733/48008 [2:30:18<13:37, 5.23it/s] 91%|█████████ | 43734/48008 [2:30:19<13:35, 5.24it/s] 91%|█████████ | 43735/48008 [2:30:19<13:37, 5.23it/s] 91%|█████████ | 43736/48008 [2:30:19<13:37, 5.23it/s] 91%|█████████ | 43737/48008 [2:30:19<13:39, 5.21it/s] 91%|█████████ | 43738/48008 [2:30:19<13:35, 5.23it/s] 91%|█████████ | 43739/48008 [2:30:20<13:52, 5.13it/s] 91%|█████████ | 43740/48008 [2:30:20<13:39, 5.21it/s] 91%|█████████ | 43741/48008 [2:30:20<13:32, 5.25it/s] 91%|█████████ | 43742/48008 [2:30:20<13:33, 5.24it/s] 91%|█████████ | 43743/48008 [2:30:20<13:30, 5.26it/s] 91%|█████████ | 43744/48008 [2:30:21<13:33, 5.24it/s] 91%|█████████ | 43745/48008 [2:30:21<16:00, 4.44it/s] 91%|█████████ | 43746/48008 [2:30:21<15:44, 4.51it/s] 91%|█████████ | 43747/48008 [2:30:21<16:06, 4.41it/s] 91%|█████████ | 43748/48008 [2:30:21<15:21, 4.62it/s] 91%|█████████ | 43749/48008 [2:30:22<14:43, 4.82it/s] 91%|█████████ | 43750/48008 [2:30:22<14:26, 4.91it/s] {'loss': 0.3008, 'grad_norm': 0.197265625, 'learning_rate': 1.774287618730212e-05, 'epoch': 0.91} 91%|█████████ | 43750/48008 [2:30:22<14:26, 4.91it/s] 91%|█████████ | 43751/48008 [2:30:22<14:15, 4.97it/s] 91%|█████████ | 43752/48008 [2:30:22<14:02, 5.05it/s] 91%|█████████ | 43753/48008 [2:30:22<13:54, 5.10it/s] 91%|█████████ | 43754/48008 [2:30:23<13:38, 5.20it/s] 91%|█████████ | 43755/48008 [2:30:23<13:39, 5.19it/s] 91%|█████████ | 43756/48008 [2:30:23<13:38, 5.19it/s] 91%|█████████ | 43757/48008 [2:30:23<13:38, 5.20it/s] 91%|█████████ | 43758/48008 [2:30:23<13:57, 5.08it/s] 91%|█████████ | 43759/48008 [2:30:24<13:48, 5.13it/s] 91%|█████████ | 43760/48008 [2:30:24<13:37, 5.20it/s] 91%|█████████ | 43761/48008 [2:30:24<13:41, 5.17it/s] 91%|█████████ | 43762/48008 [2:30:24<13:40, 5.17it/s] 91%|█████████ | 43763/48008 [2:30:24<13:47, 5.13it/s] 91%|█████████ | 43764/48008 [2:30:25<13:39, 5.18it/s] 91%|█████████ | 43765/48008 [2:30:25<13:36, 5.19it/s] 91%|█████████ | 43766/48008 [2:30:25<13:37, 5.19it/s] 91%|█████████ | 43767/48008 [2:30:25<13:28, 5.24it/s] 91%|█████████ | 43768/48008 [2:30:25<13:29, 5.24it/s] 91%|█████████ | 43769/48008 [2:30:25<13:30, 5.23it/s] 91%|█████████ | 43770/48008 [2:30:26<13:28, 5.24it/s] 91%|█████████ | 43771/48008 [2:30:26<13:45, 5.13it/s] 91%|█████████ | 43772/48008 [2:30:26<13:35, 5.19it/s] 91%|█████████ | 43773/48008 [2:30:26<14:38, 4.82it/s] 91%|█████████ | 43774/48008 [2:30:27<14:31, 4.86it/s] 91%|█████████ | 43775/48008 [2:30:27<14:09, 4.99it/s] 91%|█████████ | 43776/48008 [2:30:27<14:10, 4.98it/s] 91%|█████████ | 43777/48008 [2:30:27<13:56, 5.06it/s] 91%|█████████ | 43778/48008 [2:30:27<13:51, 5.09it/s] 91%|█████████ | 43779/48008 [2:30:27<13:40, 5.15it/s] 91%|█████████ | 43780/48008 [2:30:28<13:42, 5.14it/s] 91%|█████████ | 43781/48008 [2:30:28<17:47, 3.96it/s] 91%|█████████ | 43782/48008 [2:30:28<16:45, 4.20it/s] 91%|█████████ | 43783/48008 [2:30:28<16:02, 4.39it/s] 91%|█████████ | 43784/48008 [2:30:29<15:47, 4.46it/s] 91%|█████████ | 43785/48008 [2:30:29<15:20, 4.59it/s] 91%|█████████ | 43786/48008 [2:30:30<36:04, 1.95it/s] 91%|█████████ | 43787/48008 [2:30:30<29:12, 2.41it/s] 91%|█████████ | 43788/48008 [2:30:30<24:33, 2.86it/s] 91%|█████████ | 43789/48008 [2:30:31<21:11, 3.32it/s] 91%|█████████ | 43790/48008 [2:30:31<18:50, 3.73it/s] 91%|█████████ | 43791/48008 [2:30:31<17:15, 4.07it/s] 91%|█████████ | 43792/48008 [2:30:31<16:28, 4.26it/s] 91%|█████████ | 43793/48008 [2:30:31<15:36, 4.50it/s] 91%|█████████ | 43794/48008 [2:30:32<14:57, 4.70it/s] 91%|█████████ | 43795/48008 [2:30:32<14:27, 4.85it/s] 91%|█████████ | 43796/48008 [2:30:32<14:06, 4.98it/s] 91%|█████████ | 43797/48008 [2:30:32<13:52, 5.06it/s] 91%|█████████ | 43798/48008 [2:30:32<13:43, 5.11it/s] 91%|█████████ | 43799/48008 [2:30:33<13:35, 5.16it/s] 91%|█████████ | 43800/48008 [2:30:33<14:30, 4.83it/s] {'loss': 0.3006, 'grad_norm': 0.1875, 'learning_rate': 1.7534577570404933e-05, 'epoch': 0.91} 91%|█████████ | 43800/48008 [2:30:33<14:30, 4.83it/s] 91%|█████████ | 43801/48008 [2:30:33<14:17, 4.90it/s] 91%|█████████ | 43802/48008 [2:30:33<14:03, 4.98it/s] 91%|█████████ | 43803/48008 [2:30:33<13:52, 5.05it/s] 91%|█████████ | 43804/48008 [2:30:34<13:46, 5.09it/s] 91%|█████████ | 43805/48008 [2:30:34<13:38, 5.14it/s] 91%|█████████ | 43806/48008 [2:30:34<13:40, 5.12it/s] 91%|█████████ | 43807/48008 [2:30:34<13:35, 5.15it/s] 91%|█████████▏| 43808/48008 [2:30:34<13:31, 5.18it/s] 91%|█████████▏| 43809/48008 [2:30:35<13:44, 5.09it/s] 91%|█████████▏| 43810/48008 [2:30:35<13:39, 5.12it/s] 91%|█████████▏| 43811/48008 [2:30:35<13:33, 5.16it/s] 91%|█████████▏| 43812/48008 [2:30:35<13:53, 5.03it/s] 91%|█████████▏| 43813/48008 [2:30:35<13:35, 5.14it/s] 91%|█████████▏| 43814/48008 [2:30:36<13:39, 5.12it/s] 91%|█████████▏| 43815/48008 [2:30:36<13:33, 5.16it/s] 91%|█████████▏| 43816/48008 [2:30:36<13:29, 5.18it/s] 91%|█████████▏| 43817/48008 [2:30:36<13:31, 5.16it/s] 91%|█████████▏| 43818/48008 [2:30:36<13:29, 5.18it/s] 91%|█████████▏| 43819/48008 [2:30:37<13:34, 5.15it/s] 91%|█████████▏| 43820/48008 [2:30:37<13:28, 5.18it/s] 91%|█████████▏| 43821/48008 [2:30:37<13:20, 5.23it/s] 91%|█████████▏| 43822/48008 [2:30:37<13:17, 5.25it/s] 91%|█████████▏| 43823/48008 [2:30:37<13:10, 5.30it/s] 91%|█████████▏| 43824/48008 [2:30:37<13:12, 5.28it/s] 91%|█████████▏| 43825/48008 [2:30:38<13:16, 5.25it/s] 91%|█████████▏| 43826/48008 [2:30:38<13:19, 5.23it/s] 91%|█████████▏| 43827/48008 [2:30:38<13:15, 5.26it/s] 91%|█████████▏| 43828/48008 [2:30:38<13:21, 5.21it/s] 91%|█████████▏| 43829/48008 [2:30:38<13:15, 5.25it/s] 91%|█████████▏| 43830/48008 [2:30:39<13:16, 5.24it/s] 91%|█████████▏| 43831/48008 [2:30:39<13:44, 5.07it/s] 91%|█████████▏| 43832/48008 [2:30:39<13:37, 5.11it/s] 91%|█████████▏| 43833/48008 [2:30:39<13:34, 5.13it/s] 91%|█████████▏| 43834/48008 [2:30:39<13:28, 5.17it/s] 91%|█████████▏| 43835/48008 [2:30:40<14:23, 4.83it/s] 91%|█████████▏| 43836/48008 [2:30:40<13:57, 4.98it/s] 91%|█████████▏| 43837/48008 [2:30:40<13:57, 4.98it/s] 91%|█████████▏| 43838/48008 [2:30:40<13:45, 5.05it/s] 91%|█████████▏| 43839/48008 [2:30:40<13:34, 5.12it/s] 91%|█████████▏| 43840/48008 [2:30:41<13:46, 5.04it/s] 91%|█████████▏| 43841/48008 [2:30:41<13:46, 5.04it/s] 91%|█████████▏| 43842/48008 [2:30:41<13:35, 5.11it/s] 91%|█████████▏| 43843/48008 [2:30:41<13:33, 5.12it/s] 91%|█████████▏| 43844/48008 [2:30:41<13:31, 5.13it/s] 91%|█████████▏| 43845/48008 [2:30:42<13:30, 5.13it/s] 91%|█████████▏| 43846/48008 [2:30:42<13:28, 5.15it/s] 91%|█████████▏| 43847/48008 [2:30:42<13:25, 5.16it/s] 91%|█████████▏| 43848/48008 [2:30:42<13:31, 5.12it/s] 91%|█████████▏| 43849/48008 [2:30:42<13:21, 5.19it/s] 91%|█████████▏| 43850/48008 [2:30:43<13:15, 5.23it/s] {'loss': 0.3021, 'grad_norm': 0.162109375, 'learning_rate': 1.732627895350775e-05, 'epoch': 0.91} 91%|█████████▏| 43850/48008 [2:30:43<13:15, 5.23it/s] 91%|█████████▏| 43851/48008 [2:30:43<13:26, 5.16it/s] 91%|█████████▏| 43852/48008 [2:30:43<13:22, 5.18it/s] 91%|█████████▏| 43853/48008 [2:30:43<17:24, 3.98it/s] 91%|█████████▏| 43854/48008 [2:30:44<16:18, 4.24it/s] 91%|█████████▏| 43855/48008 [2:30:44<15:22, 4.50it/s] 91%|█████████▏| 43856/48008 [2:30:44<14:48, 4.67it/s] 91%|█████████▏| 43857/48008 [2:30:44<14:29, 4.78it/s] 91%|█████████▏| 43858/48008 [2:30:44<14:06, 4.91it/s] 91%|█████████▏| 43859/48008 [2:30:44<13:58, 4.95it/s] 91%|█████████▏| 43860/48008 [2:30:45<14:44, 4.69it/s] 91%|█████████▏| 43861/48008 [2:30:45<14:09, 4.88it/s] 91%|█████████▏| 43862/48008 [2:30:45<13:53, 4.97it/s] 91%|█████████▏| 43863/48008 [2:30:45<13:52, 4.98it/s] 91%|█████████▏| 43864/48008 [2:30:45<13:36, 5.08it/s] 91%|█████████▏| 43865/48008 [2:30:46<13:26, 5.13it/s] 91%|█████████▏| 43866/48008 [2:30:46<13:27, 5.13it/s] 91%|█████████▏| 43867/48008 [2:30:46<13:25, 5.14it/s] 91%|█████████▏| 43868/48008 [2:30:46<13:25, 5.14it/s] 91%|█████████▏| 43869/48008 [2:30:46<13:24, 5.14it/s] 91%|█████████▏| 43870/48008 [2:30:47<13:21, 5.16it/s] 91%|█████████▏| 43871/48008 [2:30:47<13:20, 5.17it/s] 91%|█████████▏| 43872/48008 [2:30:47<13:26, 5.13it/s] 91%|█████████▏| 43873/48008 [2:30:47<13:36, 5.07it/s] 91%|█████████▏| 43874/48008 [2:30:47<13:25, 5.13it/s] 91%|█████████▏| 43875/48008 [2:30:48<13:37, 5.06it/s] 91%|█████████▏| 43876/48008 [2:30:48<13:26, 5.12it/s] 91%|█████████▏| 43877/48008 [2:30:48<13:19, 5.16it/s] 91%|█████████▏| 43878/48008 [2:30:48<13:19, 5.16it/s] 91%|█████████▏| 43879/48008 [2:30:48<13:42, 5.02it/s] 91%|█████████▏| 43880/48008 [2:30:49<13:34, 5.07it/s] 91%|█████████▏| 43881/48008 [2:30:49<13:30, 5.09it/s] 91%|█████████▏| 43882/48008 [2:30:49<13:48, 4.98it/s] 91%|█████████▏| 43883/48008 [2:30:49<13:38, 5.04it/s] 91%|█████████▏| 43884/48008 [2:30:49<13:24, 5.13it/s] 91%|█████████▏| 43885/48008 [2:30:50<13:18, 5.16it/s] 91%|█████████▏| 43886/48008 [2:30:50<13:24, 5.12it/s] 91%|█████████▏| 43887/48008 [2:30:50<13:42, 5.01it/s] 91%|█████████▏| 43888/48008 [2:30:50<13:26, 5.11it/s] 91%|█████████▏| 43889/48008 [2:30:50<13:23, 5.13it/s] 91%|█████████▏| 43890/48008 [2:30:51<13:21, 5.14it/s] 91%|█████████▏| 43891/48008 [2:30:51<13:19, 5.15it/s] 91%|█████████▏| 43892/48008 [2:30:51<13:23, 5.13it/s] 91%|█████████▏| 43893/48008 [2:30:51<13:33, 5.06it/s] 91%|█████████▏| 43894/48008 [2:30:51<13:22, 5.13it/s] 91%|█████████▏| 43895/48008 [2:30:52<13:07, 5.22it/s] 91%|█████████▏| 43896/48008 [2:30:52<13:08, 5.22it/s] 91%|█████████▏| 43897/48008 [2:30:52<13:08, 5.21it/s] 91%|█████████▏| 43898/48008 [2:30:52<13:12, 5.19it/s] 91%|█████████▏| 43899/48008 [2:30:52<13:21, 5.13it/s] 91%|█████████▏| 43900/48008 [2:30:53<13:14, 5.17it/s] {'loss': 0.2999, 'grad_norm': 0.1806640625, 'learning_rate': 1.7117980336610564e-05, 'epoch': 0.91} 91%|█████████▏| 43900/48008 [2:30:53<13:14, 5.17it/s] 91%|█████████▏| 43901/48008 [2:30:53<13:14, 5.17it/s] 91%|█████████▏| 43902/48008 [2:30:53<13:25, 5.10it/s] 91%|█████████▏| 43903/48008 [2:30:53<13:18, 5.14it/s] 91%|█████████▏| 43904/48008 [2:30:53<13:22, 5.11it/s] 91%|█████████▏| 43905/48008 [2:30:53<13:30, 5.06it/s] 91%|█████████▏| 43906/48008 [2:30:54<13:21, 5.12it/s] 91%|█████████▏| 43907/48008 [2:30:54<13:11, 5.18it/s] 91%|█████████▏| 43908/48008 [2:30:54<17:12, 3.97it/s] 91%|█████████▏| 43909/48008 [2:30:54<16:05, 4.25it/s] 91%|█████████▏| 43910/48008 [2:30:55<15:16, 4.47it/s] 91%|█████████▏| 43911/48008 [2:30:55<14:37, 4.67it/s] 91%|█████████▏| 43912/48008 [2:30:55<14:09, 4.82it/s] 91%|█████████▏| 43913/48008 [2:30:55<13:52, 4.92it/s] 91%|█████████▏| 43914/48008 [2:30:55<13:40, 4.99it/s] 91%|█████████▏| 43915/48008 [2:30:56<13:33, 5.03it/s] 91%|█████████▏| 43916/48008 [2:30:56<13:45, 4.96it/s] 91%|█████████▏| 43917/48008 [2:30:56<13:29, 5.05it/s] 91%|█████████▏| 43918/48008 [2:30:56<13:27, 5.06it/s] 91%|█████████▏| 43919/48008 [2:30:56<13:22, 5.10it/s] 91%|█████████▏| 43920/48008 [2:30:57<13:18, 5.12it/s] 91%|█████████▏| 43921/48008 [2:30:57<13:16, 5.13it/s] 91%|█████████▏| 43922/48008 [2:30:57<13:25, 5.07it/s] 91%|█████████▏| 43923/48008 [2:30:57<13:14, 5.14it/s] 91%|█████████▏| 43924/48008 [2:30:57<13:13, 5.14it/s] 91%|█████████▏| 43925/48008 [2:30:58<13:10, 5.17it/s] 91%|█████████▏| 43926/48008 [2:30:58<17:11, 3.96it/s] 91%|█████████▏| 43927/48008 [2:30:58<19:58, 3.41it/s] 92%|█████████▏| 43928/48008 [2:30:59<17:51, 3.81it/s] 92%|█████████▏| 43929/48008 [2:30:59<16:19, 4.16it/s] 92%|█████████▏| 43930/48008 [2:30:59<15:21, 4.43it/s] 92%|█████████▏| 43931/48008 [2:30:59<14:42, 4.62it/s] 92%|█████████▏| 43932/48008 [2:31:00<18:10, 3.74it/s] 92%|█████████▏| 43933/48008 [2:31:00<16:45, 4.05it/s] 92%|█████████▏| 43934/48008 [2:31:00<15:44, 4.31it/s] 92%|█████████▏| 43935/48008 [2:31:00<14:56, 4.54it/s] 92%|█████████▏| 43936/48008 [2:31:00<14:46, 4.59it/s] 92%|█████████▏| 43937/48008 [2:31:00<14:14, 4.77it/s] 92%|█████████▏| 43938/48008 [2:31:01<14:09, 4.79it/s] 92%|█████████▏| 43939/48008 [2:31:01<14:48, 4.58it/s] 92%|█████████▏| 43940/48008 [2:31:01<14:16, 4.75it/s] 92%|█████████▏| 43941/48008 [2:31:01<13:53, 4.88it/s] 92%|█████████▏| 43942/48008 [2:31:02<13:31, 5.01it/s] 92%|█████████▏| 43943/48008 [2:31:02<13:09, 5.15it/s] 92%|█████████▏| 43944/48008 [2:31:02<13:15, 5.11it/s] 92%|█████████▏| 43945/48008 [2:31:02<13:16, 5.10it/s] 92%|█████████▏| 43946/48008 [2:31:02<13:16, 5.10it/s] 92%|█████████▏| 43947/48008 [2:31:02<13:11, 5.13it/s] 92%|█████████▏| 43948/48008 [2:31:03<13:15, 5.10it/s] 92%|█████████▏| 43949/48008 [2:31:03<13:25, 5.04it/s] 92%|█████████▏| 43950/48008 [2:31:03<13:13, 5.11it/s] {'loss': 0.2985, 'grad_norm': 0.1513671875, 'learning_rate': 1.690968171971338e-05, 'epoch': 0.92} 92%|█████████▏| 43950/48008 [2:31:03<13:13, 5.11it/s] 92%|█████████▏| 43951/48008 [2:31:03<13:16, 5.10it/s] 92%|█████████▏| 43952/48008 [2:31:03<13:07, 5.15it/s] 92%|█████████▏| 43953/48008 [2:31:04<13:07, 5.15it/s] 92%|█████████▏| 43954/48008 [2:31:04<13:03, 5.17it/s] 92%|█████████▏| 43955/48008 [2:31:04<12:58, 5.20it/s] 92%|█████████▏| 43956/48008 [2:31:04<12:53, 5.24it/s] 92%|█████████▏| 43957/48008 [2:31:04<12:54, 5.23it/s] 92%|█████████▏| 43958/48008 [2:31:05<12:51, 5.25it/s] 92%|█████████▏| 43959/48008 [2:31:05<13:00, 5.19it/s] 92%|█████████▏| 43960/48008 [2:31:05<13:00, 5.19it/s] 92%|█████████▏| 43961/48008 [2:31:05<13:12, 5.11it/s] 92%|█████████▏| 43962/48008 [2:31:05<13:10, 5.12it/s] 92%|█████████▏| 43963/48008 [2:31:06<17:02, 3.96it/s] 92%|█████████▏| 43964/48008 [2:31:06<16:01, 4.21it/s] 92%|█████████▏| 43965/48008 [2:31:06<15:05, 4.47it/s] 92%|█████████▏| 43966/48008 [2:31:06<14:25, 4.67it/s] 92%|█████████▏| 43967/48008 [2:31:07<13:56, 4.83it/s] 92%|█████████▏| 43968/48008 [2:31:07<13:47, 4.88it/s] 92%|█████████▏| 43969/48008 [2:31:07<13:42, 4.91it/s] 92%|█████████▏| 43970/48008 [2:31:07<13:22, 5.03it/s] 92%|█████████▏| 43971/48008 [2:31:07<13:14, 5.08it/s] 92%|█████████▏| 43972/48008 [2:31:08<15:29, 4.34it/s] 92%|█████████▏| 43973/48008 [2:31:08<14:46, 4.55it/s] 92%|█████████▏| 43974/48008 [2:31:08<14:05, 4.77it/s] 92%|█████████▏| 43975/48008 [2:31:08<14:06, 4.77it/s] 92%|█████████▏| 43976/48008 [2:31:08<13:49, 4.86it/s] 92%|█████████▏| 43977/48008 [2:31:09<13:33, 4.96it/s] 92%|█████████▏| 43978/48008 [2:31:09<13:20, 5.03it/s] 92%|█████████▏| 43979/48008 [2:31:09<13:20, 5.03it/s] 92%|█████████▏| 43980/48008 [2:31:09<13:12, 5.08it/s] 92%|█████████▏| 43981/48008 [2:31:09<13:19, 5.04it/s] 92%|█████████▏| 43982/48008 [2:31:10<13:05, 5.13it/s] 92%|█████████▏| 43983/48008 [2:31:10<13:01, 5.15it/s] 92%|█████████▏| 43984/48008 [2:31:10<12:59, 5.16it/s] 92%|█████████▏| 43985/48008 [2:31:10<13:12, 5.08it/s] 92%|█████████▏| 43986/48008 [2:31:10<13:06, 5.11it/s] 92%|█████████▏| 43987/48008 [2:31:11<12:59, 5.16it/s] 92%|█████████▏| 43988/48008 [2:31:11<12:55, 5.19it/s] 92%|█████████▏| 43989/48008 [2:31:11<12:53, 5.20it/s] 92%|█████████▏| 43990/48008 [2:31:11<12:50, 5.21it/s] 92%|█████████▏| 43991/48008 [2:31:11<13:02, 5.13it/s] 92%|█████████▏| 43992/48008 [2:31:12<12:50, 5.21it/s] 92%|█████████▏| 43993/48008 [2:31:12<12:48, 5.22it/s] 92%|█████████▏| 43994/48008 [2:31:12<12:53, 5.19it/s] 92%|█████████▏| 43995/48008 [2:31:12<13:04, 5.12it/s] 92%|█████████▏| 43996/48008 [2:31:12<12:53, 5.18it/s] 92%|█████████▏| 43997/48008 [2:31:12<12:51, 5.20it/s] 92%|█████████▏| 43998/48008 [2:31:13<12:48, 5.22it/s] 92%|█████████▏| 43999/48008 [2:31:13<12:50, 5.20it/s] 92%|█████████▏| 44000/48008 [2:31:13<13:11, 5.06it/s] {'loss': 0.3018, 'grad_norm': 0.251953125, 'learning_rate': 1.67013831028162e-05, 'epoch': 0.92} 92%|█████████▏| 44000/48008 [2:31:13<13:11, 5.06it/s] 92%|█████████▏| 44001/48008 [2:31:13<13:03, 5.12it/s] 92%|█████████▏| 44002/48008 [2:31:13<13:02, 5.12it/s] 92%|█████████▏| 44003/48008 [2:31:14<12:57, 5.15it/s] 92%|█████████▏| 44004/48008 [2:31:14<12:54, 5.17it/s] 92%|█████████▏| 44005/48008 [2:31:14<13:50, 4.82it/s] 92%|█████████▏| 44006/48008 [2:31:14<13:52, 4.81it/s] 92%|█████████▏| 44007/48008 [2:31:15<13:54, 4.79it/s] 92%|█████████▏| 44008/48008 [2:31:15<13:46, 4.84it/s] 92%|█████████▏| 44009/48008 [2:31:15<13:36, 4.90it/s] 92%|█████████▏| 44010/48008 [2:31:15<13:22, 4.98it/s] 92%|█████████▏| 44011/48008 [2:31:15<13:24, 4.97it/s] 92%|█████████▏| 44012/48008 [2:31:15<13:11, 5.05it/s] 92%|█████████▏| 44013/48008 [2:31:16<13:03, 5.10it/s] 92%|█████████▏| 44014/48008 [2:31:16<13:05, 5.09it/s] 92%|█████████▏| 44015/48008 [2:31:16<13:13, 5.03it/s] 92%|█████████▏| 44016/48008 [2:31:16<13:21, 4.98it/s] 92%|█████████▏| 44017/48008 [2:31:16<13:12, 5.04it/s] 92%|█████████▏| 44018/48008 [2:31:17<13:18, 5.00it/s] 92%|█████████▏| 44019/48008 [2:31:17<13:10, 5.05it/s] 92%|█████████▏| 44020/48008 [2:31:17<13:06, 5.07it/s] 92%|█████████▏| 44021/48008 [2:31:17<12:57, 5.13it/s] 92%|█████████▏| 44022/48008 [2:31:17<12:47, 5.19it/s] 92%|█████████▏| 44023/48008 [2:31:18<12:48, 5.18it/s] 92%|█████████▏| 44024/48008 [2:31:18<12:48, 5.19it/s] 92%|█████████▏| 44025/48008 [2:31:18<13:14, 5.01it/s] 92%|█████████▏| 44026/48008 [2:31:18<13:02, 5.09it/s] 92%|█████████▏| 44027/48008 [2:31:18<12:57, 5.12it/s] 92%|█████████▏| 44028/48008 [2:31:19<12:56, 5.12it/s] 92%|█████████▏| 44029/48008 [2:31:19<12:51, 5.15it/s] 92%|█████████▏| 44030/48008 [2:31:19<12:47, 5.18it/s] 92%|█████████▏| 44031/48008 [2:31:19<16:37, 3.99it/s] 92%|█████████▏| 44032/48008 [2:31:20<15:38, 4.23it/s] 92%|█████████▏| 44033/48008 [2:31:20<14:50, 4.46it/s] 92%|█████████▏| 44034/48008 [2:31:20<14:14, 4.65it/s] 92%|█████████▏| 44035/48008 [2:31:20<14:04, 4.71it/s] 92%|█████████▏| 44036/48008 [2:31:21<15:58, 4.14it/s] 92%|█████████▏| 44037/48008 [2:31:21<14:57, 4.42it/s] 92%|█████████▏| 44038/48008 [2:31:21<14:34, 4.54it/s] 92%|█████████▏| 44039/48008 [2:31:21<14:02, 4.71it/s] 92%|█████████▏| 44040/48008 [2:31:21<13:36, 4.86it/s] 92%|█████████▏| 44041/48008 [2:31:21<13:31, 4.89it/s] 92%|█████████▏| 44042/48008 [2:31:22<13:26, 4.92it/s] 92%|█████████▏| 44043/48008 [2:31:22<13:13, 5.00it/s] 92%|█████████▏| 44044/48008 [2:31:22<13:12, 5.00it/s] 92%|█████████▏| 44045/48008 [2:31:22<12:56, 5.10it/s] 92%|█████████▏| 44046/48008 [2:31:22<13:15, 4.98it/s] 92%|█████████▏| 44047/48008 [2:31:23<13:18, 4.96it/s] 92%|█████████▏| 44048/48008 [2:31:23<13:15, 4.98it/s] 92%|█████████▏| 44049/48008 [2:31:23<13:11, 5.00it/s] 92%|█████████▏| 44050/48008 [2:31:23<13:02, 5.06it/s] {'loss': 0.2999, 'grad_norm': 0.2421875, 'learning_rate': 1.6493084485919013e-05, 'epoch': 0.92} 92%|█████████▏| 44050/48008 [2:31:23<13:02, 5.06it/s] 92%|█████████▏| 44051/48008 [2:31:23<13:15, 4.97it/s] 92%|█████████▏| 44052/48008 [2:31:24<13:12, 4.99it/s] 92%|█████████▏| 44053/48008 [2:31:24<13:00, 5.06it/s] 92%|█████████▏| 44054/48008 [2:31:24<13:11, 5.00it/s] 92%|█████████▏| 44055/48008 [2:31:24<13:07, 5.02it/s] 92%|█████████▏| 44056/48008 [2:31:24<13:08, 5.01it/s] 92%|█████████▏| 44057/48008 [2:31:25<12:53, 5.11it/s] 92%|█████████▏| 44058/48008 [2:31:25<12:40, 5.19it/s] 92%|█████████▏| 44059/48008 [2:31:25<12:54, 5.10it/s] 92%|█████████▏| 44060/48008 [2:31:25<13:11, 4.99it/s] 92%|█████████▏| 44061/48008 [2:31:25<13:14, 4.97it/s] 92%|█████████▏| 44062/48008 [2:31:26<13:24, 4.90it/s] 92%|█████████▏| 44063/48008 [2:31:26<13:10, 4.99it/s] 92%|█████████▏| 44064/48008 [2:31:26<13:01, 5.05it/s] 92%|█████████▏| 44065/48008 [2:31:26<15:10, 4.33it/s] 92%|█████████▏| 44066/48008 [2:31:27<14:23, 4.57it/s] 92%|█████████▏| 44067/48008 [2:31:27<13:51, 4.74it/s] 92%|█████████▏| 44068/48008 [2:31:27<13:31, 4.86it/s] 92%|█████████▏| 44069/48008 [2:31:27<13:12, 4.97it/s] 92%|█████████▏| 44070/48008 [2:31:27<13:10, 4.98it/s] 92%|█████████▏| 44071/48008 [2:31:28<13:08, 4.99it/s] 92%|█████████▏| 44072/48008 [2:31:28<13:00, 5.04it/s] 92%|█████████▏| 44073/48008 [2:31:28<12:46, 5.13it/s] 92%|█████████▏| 44074/48008 [2:31:28<12:42, 5.16it/s] 92%|█████████▏| 44075/48008 [2:31:28<12:40, 5.17it/s] 92%|█████████▏| 44076/48008 [2:31:29<12:47, 5.13it/s] 92%|█████████▏| 44077/48008 [2:31:29<12:56, 5.06it/s] 92%|█████████▏| 44078/48008 [2:31:29<13:45, 4.76it/s] 92%|█████████▏| 44079/48008 [2:31:29<13:27, 4.86it/s] 92%|█████████▏| 44080/48008 [2:31:29<13:05, 5.00it/s] 92%|█████████▏| 44081/48008 [2:31:30<12:56, 5.06it/s] 92%|█████████▏| 44082/48008 [2:31:30<12:44, 5.14it/s] 92%|█████████▏| 44083/48008 [2:31:30<12:46, 5.12it/s] 92%|█████████▏| 44084/48008 [2:31:30<12:47, 5.11it/s] 92%|█████████▏| 44085/48008 [2:31:30<12:59, 5.03it/s] 92%|█████████▏| 44086/48008 [2:31:31<12:54, 5.06it/s] 92%|█████████▏| 44087/48008 [2:31:31<12:48, 5.11it/s] 92%|█████████▏| 44088/48008 [2:31:31<12:56, 5.05it/s] 92%|█████████▏| 44089/48008 [2:31:31<12:47, 5.10it/s] 92%|█████████▏| 44090/48008 [2:31:31<12:41, 5.14it/s] 92%|█████████▏| 44091/48008 [2:31:31<12:31, 5.21it/s] 92%|█████████▏| 44092/48008 [2:31:32<12:31, 5.21it/s] 92%|█████████▏| 44093/48008 [2:31:32<12:33, 5.19it/s] 92%|█████████▏| 44094/48008 [2:31:32<13:30, 4.83it/s] 92%|█████████▏| 44095/48008 [2:31:32<13:13, 4.93it/s] 92%|█████████▏| 44096/48008 [2:31:32<12:57, 5.03it/s] 92%|█████████▏| 44097/48008 [2:31:33<12:43, 5.12it/s] 92%|█████████▏| 44098/48008 [2:31:33<12:36, 5.17it/s] 92%|█████████▏| 44099/48008 [2:31:33<12:31, 5.20it/s] 92%|█████████▏| 44100/48008 [2:31:33<12:24, 5.25it/s] {'loss': 0.303, 'grad_norm': 0.328125, 'learning_rate': 1.628478586902183e-05, 'epoch': 0.92} 92%|█████████▏| 44100/48008 [2:31:33<12:24, 5.25it/s] 92%|█████████▏| 44101/48008 [2:31:33<12:28, 5.22it/s] 92%|█████████▏| 44102/48008 [2:31:34<12:40, 5.14it/s] 92%|█████████▏| 44103/48008 [2:31:34<12:42, 5.12it/s] 92%|█████████▏| 44104/48008 [2:31:34<12:45, 5.10it/s] 92%|█████████▏| 44105/48008 [2:31:34<12:34, 5.17it/s] 92%|█████████▏| 44106/48008 [2:31:35<16:24, 3.97it/s] 92%|█████████▏| 44107/48008 [2:31:35<15:28, 4.20it/s] 92%|█████████▏| 44108/48008 [2:31:35<14:46, 4.40it/s] 92%|█████████▏| 44109/48008 [2:31:35<14:19, 4.54it/s] 92%|█████████▏| 44110/48008 [2:31:35<13:49, 4.70it/s] 92%|█████████▏| 44111/48008 [2:31:36<13:27, 4.82it/s] 92%|█████████▏| 44112/48008 [2:31:36<13:25, 4.84it/s] 92%|█████████▏| 44113/48008 [2:31:36<13:13, 4.91it/s] 92%|█████████▏| 44114/48008 [2:31:36<13:03, 4.97it/s] 92%|█████████▏| 44115/48008 [2:31:36<12:55, 5.02it/s] 92%|█████████▏| 44116/48008 [2:31:37<12:53, 5.03it/s] 92%|█████████▏| 44117/48008 [2:31:37<12:37, 5.13it/s] 92%|█████████▏| 44118/48008 [2:31:37<12:37, 5.14it/s] 92%|█████████▏| 44119/48008 [2:31:37<12:35, 5.15it/s] 92%|█████████▏| 44120/48008 [2:31:37<12:45, 5.08it/s] 92%|█████████▏| 44121/48008 [2:31:38<12:57, 5.00it/s] 92%|█████████▏| 44122/48008 [2:31:38<12:44, 5.09it/s] 92%|█████████▏| 44123/48008 [2:31:38<12:39, 5.12it/s] 92%|█████████▏| 44124/48008 [2:31:38<12:31, 5.17it/s] 92%|█████████▏| 44125/48008 [2:31:38<12:30, 5.17it/s] 92%|█████████▏| 44126/48008 [2:31:39<14:42, 4.40it/s] 92%|█████████▏| 44127/48008 [2:31:39<13:59, 4.62it/s] 92%|█████████▏| 44128/48008 [2:31:39<13:26, 4.81it/s] 92%|█████████▏| 44129/48008 [2:31:39<12:59, 4.97it/s] 92%|█████████▏| 44130/48008 [2:31:39<12:49, 5.04it/s] 92%|█████████▏| 44131/48008 [2:31:40<12:43, 5.08it/s] 92%|█████████▏| 44132/48008 [2:31:40<13:01, 4.96it/s] 92%|█████████▏| 44133/48008 [2:31:40<12:55, 5.00it/s] 92%|█████████▏| 44134/48008 [2:31:40<12:53, 5.01it/s] 92%|█████████▏| 44135/48008 [2:31:40<12:44, 5.07it/s] 92%|█████████▏| 44136/48008 [2:31:41<12:33, 5.14it/s] 92%|█████████▏| 44137/48008 [2:31:41<12:33, 5.14it/s] 92%|█████████▏| 44138/48008 [2:31:41<12:46, 5.05it/s] 92%|█████████▏| 44139/48008 [2:31:41<16:32, 3.90it/s] 92%|█████████▏| 44140/48008 [2:31:42<15:20, 4.20it/s] 92%|█████████▏| 44141/48008 [2:31:42<14:49, 4.35it/s] 92%|█████████▏| 44142/48008 [2:31:42<14:00, 4.60it/s] 92%|█████████▏| 44143/48008 [2:31:42<13:39, 4.71it/s] 92%|█████████▏| 44144/48008 [2:31:42<13:14, 4.86it/s] 92%|█████████▏| 44145/48008 [2:31:43<13:14, 4.86it/s] 92%|█████████▏| 44146/48008 [2:31:43<13:52, 4.64it/s] 92%|█████████▏| 44147/48008 [2:31:43<13:33, 4.75it/s] 92%|█████████▏| 44148/48008 [2:31:43<13:10, 4.88it/s] 92%|█████████▏| 44149/48008 [2:31:43<12:55, 4.98it/s] 92%|█████████▏| 44150/48008 [2:31:44<12:43, 5.05it/s] 92%|█████████▏| 44150/48008 [2:31:44<12:43, 5.05it/s]{'loss': 0.3013, 'grad_norm': 0.17578125, 'learning_rate': 1.6076487252124648e-05, 'epoch': 0.92} 92%|█████████▏| 44151/48008 [2:31:44<12:26, 5.17it/s] 92%|█████████▏| 44152/48008 [2:31:44<12:19, 5.22it/s] 92%|█████████▏| 44153/48008 [2:31:44<12:17, 5.23it/s] 92%|█████████▏| 44154/48008 [2:31:44<12:18, 5.22it/s] 92%|█████████▏| 44155/48008 [2:31:45<12:32, 5.12it/s] 92%|█████████▏| 44156/48008 [2:31:45<12:32, 5.12it/s] 92%|█████████▏| 44157/48008 [2:31:45<12:32, 5.12it/s] 92%|█████████▏| 44158/48008 [2:31:45<12:48, 5.01it/s] 92%|█████████▏| 44159/48008 [2:31:45<12:52, 4.98it/s] 92%|█████████▏| 44160/48008 [2:31:46<12:44, 5.04it/s] 92%|█████████▏| 44161/48008 [2:31:46<12:35, 5.10it/s] 92%|█████████▏| 44162/48008 [2:31:46<12:22, 5.18it/s] 92%|█████████▏| 44163/48008 [2:31:46<12:15, 5.23it/s] 92%|█████████▏| 44164/48008 [2:31:46<12:15, 5.22it/s] 92%|█████████▏| 44165/48008 [2:31:46<12:18, 5.20it/s] 92%|█████████▏| 44166/48008 [2:31:47<12:24, 5.16it/s] 92%|█████████▏| 44167/48008 [2:31:47<12:23, 5.17it/s] 92%|█████████▏| 44168/48008 [2:31:47<12:15, 5.22it/s] 92%|█████████▏| 44169/48008 [2:31:47<12:37, 5.07it/s] 92%|█████████▏| 44170/48008 [2:31:47<12:31, 5.11it/s] 92%|█████████▏| 44171/48008 [2:31:48<13:19, 4.80it/s] 92%|█████████▏| 44172/48008 [2:31:48<13:12, 4.84it/s] 92%|█████████▏| 44173/48008 [2:31:48<13:17, 4.81it/s] 92%|█████████▏| 44174/48008 [2:31:48<13:20, 4.79it/s] 92%|█████████▏| 44175/48008 [2:31:49<13:04, 4.88it/s] 92%|█████████▏| 44176/48008 [2:31:49<12:48, 4.99it/s] 92%|█████████▏| 44177/48008 [2:31:49<12:34, 5.08it/s] 92%|█████████▏| 44178/48008 [2:31:49<12:28, 5.12it/s] 92%|█████████▏| 44179/48008 [2:31:49<12:28, 5.12it/s] 92%|█████████▏| 44180/48008 [2:31:49<12:31, 5.10it/s] 92%|█████████▏| 44181/48008 [2:31:50<12:22, 5.15it/s] 92%|█████████▏| 44182/48008 [2:31:50<12:31, 5.09it/s] 92%|█████████▏| 44183/48008 [2:31:50<12:30, 5.10it/s] 92%|█████████▏| 44184/48008 [2:31:50<12:23, 5.14it/s] 92%|█████████▏| 44185/48008 [2:31:50<12:19, 5.17it/s] 92%|█████████▏| 44186/48008 [2:31:51<12:21, 5.15it/s] 92%|█████████▏| 44187/48008 [2:31:51<12:27, 5.11it/s] 92%|█████████▏| 44188/48008 [2:31:51<12:32, 5.07it/s] 92%|█████████▏| 44189/48008 [2:31:51<12:25, 5.12it/s] 92%|█████████▏| 44190/48008 [2:31:51<12:19, 5.16it/s] 92%|█████████▏| 44191/48008 [2:31:52<12:19, 5.16it/s] 92%|█████████▏| 44192/48008 [2:31:52<12:21, 5.15it/s] 92%|█████████▏| 44193/48008 [2:31:52<12:26, 5.11it/s] 92%|█████████▏| 44194/48008 [2:31:52<12:21, 5.15it/s] 92%|█████████▏| 44195/48008 [2:31:52<12:28, 5.10it/s] 92%|█████████▏| 44196/48008 [2:31:53<12:27, 5.10it/s] 92%|█████████▏| 44197/48008 [2:31:53<12:25, 5.11it/s] 92%|█████████▏| 44198/48008 [2:31:53<12:15, 5.18it/s] 92%|█████████▏| 44199/48008 [2:31:53<12:19, 5.15it/s] 92%|█████████▏| 44200/48008 [2:31:53<12:20, 5.14it/s] {'loss': 0.2987, 'grad_norm': 0.1923828125, 'learning_rate': 1.5868188635227462e-05, 'epoch': 0.92} 92%|█████████▏| 44200/48008 [2:31:53<12:20, 5.14it/s] 92%|█████████▏| 44201/48008 [2:31:54<12:22, 5.13it/s] 92%|█████████▏| 44202/48008 [2:31:54<12:24, 5.11it/s] 92%|█████████▏| 44203/48008 [2:31:54<14:33, 4.36it/s] 92%|█████████▏| 44204/48008 [2:31:54<13:50, 4.58it/s] 92%|█████████▏| 44205/48008 [2:31:54<13:26, 4.72it/s] 92%|█████████▏| 44206/48008 [2:31:55<13:23, 4.73it/s] 92%|█████████▏| 44207/48008 [2:31:55<13:05, 4.84it/s] 92%|█████████▏| 44208/48008 [2:31:55<13:01, 4.86it/s] 92%|█████████▏| 44209/48008 [2:31:55<13:01, 4.86it/s] 92%|█████████▏| 44210/48008 [2:31:55<12:52, 4.91it/s] 92%|█████████▏| 44211/48008 [2:31:56<12:53, 4.91it/s] 92%|█████████▏| 44212/48008 [2:31:56<12:51, 4.92it/s] 92%|█████████▏| 44213/48008 [2:31:56<12:33, 5.04it/s] 92%|█████████▏| 44214/48008 [2:31:56<12:22, 5.11it/s] 92%|█████████▏| 44215/48008 [2:31:56<12:19, 5.13it/s] 92%|█████████▏| 44216/48008 [2:31:57<12:14, 5.16it/s] 92%|█████████▏| 44217/48008 [2:31:57<12:16, 5.15it/s] 92%|█████████▏| 44218/48008 [2:31:57<12:14, 5.16it/s] 92%|█████████▏| 44219/48008 [2:31:57<15:56, 3.96it/s] 92%|█████████▏| 44220/48008 [2:31:58<14:45, 4.28it/s] 92%|█████████▏| 44221/48008 [2:31:58<13:57, 4.52it/s] 92%|█████████▏| 44222/48008 [2:31:58<13:39, 4.62it/s] 92%|█████████▏| 44223/48008 [2:31:58<13:23, 4.71it/s] 92%|█████████▏| 44224/48008 [2:31:58<12:57, 4.87it/s] 92%|█████████▏| 44225/48008 [2:31:59<12:58, 4.86it/s] 92%|█████████▏| 44226/48008 [2:31:59<16:26, 3.83it/s] 92%|█████████▏| 44227/48008 [2:31:59<15:12, 4.14it/s] 92%|█████████▏| 44228/48008 [2:31:59<14:21, 4.39it/s] 92%|█████████▏| 44229/48008 [2:32:00<13:39, 4.61it/s] 92%|█████████▏| 44230/48008 [2:32:00<13:18, 4.73it/s] 92%|█████████▏| 44231/48008 [2:32:00<12:55, 4.87it/s] 92%|█████████▏| 44232/48008 [2:32:00<12:38, 4.98it/s] 92%|█████████▏| 44233/48008 [2:32:00<12:29, 5.04it/s] 92%|█████████▏| 44234/48008 [2:32:01<12:17, 5.12it/s] 92%|█████████▏| 44235/48008 [2:32:01<12:34, 5.00it/s] 92%|█████████▏| 44236/48008 [2:32:01<12:28, 5.04it/s] 92%|█████████▏| 44237/48008 [2:32:01<12:17, 5.11it/s] 92%|█████████▏| 44238/48008 [2:32:01<12:29, 5.03it/s] 92%|█████████▏| 44239/48008 [2:32:02<12:24, 5.06it/s] 92%|█████████▏| 44240/48008 [2:32:02<12:25, 5.06it/s] 92%|█████████▏| 44241/48008 [2:32:02<12:10, 5.15it/s] 92%|█████████▏| 44242/48008 [2:32:02<12:13, 5.13it/s] 92%|█████████▏| 44243/48008 [2:32:02<12:24, 5.06it/s] 92%|█████████▏| 44244/48008 [2:32:03<12:15, 5.11it/s] 92%|█████████▏| 44245/48008 [2:32:03<12:21, 5.08it/s] 92%|█████████▏| 44246/48008 [2:32:03<12:14, 5.12it/s] 92%|█████████▏| 44247/48008 [2:32:03<12:13, 5.13it/s] 92%|█████████▏| 44248/48008 [2:32:03<13:01, 4.81it/s] 92%|█████████▏| 44249/48008 [2:32:04<12:57, 4.83it/s] 92%|█████████▏| 44250/48008 [2:32:04<12:36, 4.97it/s] {'loss': 0.2983, 'grad_norm': 0.61328125, 'learning_rate': 1.565989001833028e-05, 'epoch': 0.92} 92%|█████████▏| 44250/48008 [2:32:04<12:36, 4.97it/s] 92%|█████████▏| 44251/48008 [2:32:04<12:25, 5.04it/s] 92%|█████████▏| 44252/48008 [2:32:04<12:17, 5.09it/s] 92%|█████████▏| 44253/48008 [2:32:04<12:06, 5.17it/s] 92%|█████████▏| 44254/48008 [2:32:04<12:11, 5.14it/s] 92%|█████████▏| 44255/48008 [2:32:05<12:06, 5.16it/s] 92%|█████████▏| 44256/48008 [2:32:05<12:15, 5.10it/s] 92%|█████████▏| 44257/48008 [2:32:05<12:10, 5.14it/s] 92%|█████████▏| 44258/48008 [2:32:05<12:09, 5.14it/s] 92%|█████████▏| 44259/48008 [2:32:05<12:05, 5.17it/s] 92%|█████████▏| 44260/48008 [2:32:06<12:04, 5.18it/s] 92%|█████████▏| 44261/48008 [2:32:06<15:39, 3.99it/s] 92%|█████████▏| 44262/48008 [2:32:06<14:54, 4.19it/s] 92%|█████████▏| 44263/48008 [2:32:06<13:58, 4.46it/s] 92%|█████████▏| 44264/48008 [2:32:07<13:21, 4.67it/s] 92%|█████████▏| 44265/48008 [2:32:07<12:55, 4.82it/s] 92%|█████████▏| 44266/48008 [2:32:07<12:37, 4.94it/s] 92%|█████████▏| 44267/48008 [2:32:07<12:31, 4.98it/s] 92%|█████████▏| 44268/48008 [2:32:07<12:18, 5.07it/s] 92%|█████████▏| 44269/48008 [2:32:08<12:07, 5.14it/s] 92%|█████████▏| 44270/48008 [2:32:08<12:24, 5.02it/s] 92%|█████████▏| 44271/48008 [2:32:08<12:16, 5.08it/s] 92%|█████████▏| 44272/48008 [2:32:08<12:14, 5.08it/s] 92%|█████████▏| 44273/48008 [2:32:08<12:01, 5.18it/s] 92%|█████████▏| 44274/48008 [2:32:09<12:00, 5.18it/s] 92%|█████████▏| 44275/48008 [2:32:09<11:56, 5.21it/s] 92%|█████████▏| 44276/48008 [2:32:09<12:02, 5.17it/s] 92%|█████████▏| 44277/48008 [2:32:09<11:59, 5.18it/s] 92%|█████████▏| 44278/48008 [2:32:09<11:59, 5.19it/s] 92%|█████████▏| 44279/48008 [2:32:10<11:58, 5.19it/s] 92%|█████████▏| 44280/48008 [2:32:10<11:48, 5.26it/s] 92%|█████████▏| 44281/48008 [2:32:10<12:10, 5.10it/s] 92%|█████████▏| 44282/48008 [2:32:10<12:07, 5.12it/s] 92%|█████████▏| 44283/48008 [2:32:10<12:10, 5.10it/s] 92%|█████████▏| 44284/48008 [2:32:11<12:09, 5.10it/s] 92%|█████████▏| 44285/48008 [2:32:11<12:07, 5.12it/s] 92%|█████████▏| 44286/48008 [2:32:11<12:00, 5.17it/s] 92%|█████████▏| 44287/48008 [2:32:11<11:58, 5.18it/s] 92%|█████████▏| 44288/48008 [2:32:11<12:03, 5.14it/s] 92%|█████████▏| 44289/48008 [2:32:11<11:59, 5.17it/s] 92%|█████████▏| 44290/48008 [2:32:12<12:48, 4.84it/s] 92%|█████████▏| 44291/48008 [2:32:12<12:34, 4.93it/s] 92%|█████████▏| 44292/48008 [2:32:12<12:39, 4.89it/s] 92%|█████████▏| 44293/48008 [2:32:12<14:34, 4.25it/s] 92%|█████████▏| 44294/48008 [2:32:13<13:37, 4.55it/s] 92%|█████████▏| 44295/48008 [2:32:13<12:59, 4.76it/s] 92%|█████████▏| 44296/48008 [2:32:13<12:40, 4.88it/s] 92%|█████████▏| 44297/48008 [2:32:13<12:38, 4.89it/s] 92%|█████████▏| 44298/48008 [2:32:13<12:16, 5.04it/s] 92%|█████████▏| 44299/48008 [2:32:14<12:17, 5.03it/s] 92%|█████████▏| 44300/48008 [2:32:14<12:12, 5.06it/s] {'loss': 0.3004, 'grad_norm': 0.251953125, 'learning_rate': 1.5451591401433096e-05, 'epoch': 0.92} 92%|█████████▏| 44300/48008 [2:32:14<12:12, 5.06it/s] 92%|█████████▏| 44301/48008 [2:32:14<12:04, 5.12it/s] 92%|█████████▏| 44302/48008 [2:32:14<11:53, 5.19it/s] 92%|█████████▏| 44303/48008 [2:32:14<11:49, 5.22it/s] 92%|█████████▏| 44304/48008 [2:32:15<12:08, 5.08it/s] 92%|█████████▏| 44305/48008 [2:32:15<11:56, 5.17it/s] 92%|█████████▏| 44306/48008 [2:32:15<11:49, 5.22it/s] 92%|█████████▏| 44307/48008 [2:32:15<11:46, 5.24it/s] 92%|█████████▏| 44308/48008 [2:32:15<11:47, 5.23it/s] 92%|█████████▏| 44309/48008 [2:32:16<29:39, 2.08it/s] 92%|█████████▏| 44310/48008 [2:32:17<24:19, 2.53it/s] 92%|█████████▏| 44311/48008 [2:32:17<20:45, 2.97it/s] 92%|█████████▏| 44312/48008 [2:32:17<18:02, 3.41it/s] 92%|█████████▏| 44313/48008 [2:32:17<17:01, 3.62it/s] 92%|█████████▏| 44314/48008 [2:32:17<15:24, 3.99it/s] 92%|█████████▏| 44315/48008 [2:32:18<14:15, 4.32it/s] 92%|█████████▏| 44316/48008 [2:32:18<13:32, 4.54it/s] 92%|█████████▏| 44317/48008 [2:32:18<12:58, 4.74it/s] 92%|█████████▏| 44318/48008 [2:32:18<12:33, 4.90it/s] 92%|█████████▏| 44319/48008 [2:32:18<12:12, 5.03it/s] 92%|█████████▏| 44320/48008 [2:32:19<12:04, 5.09it/s] 92%|█████████▏| 44321/48008 [2:32:20<26:30, 2.32it/s] 92%|█████████▏| 44322/48008 [2:32:20<22:04, 2.78it/s] 92%|█████████▏| 44323/48008 [2:32:20<18:57, 3.24it/s] 92%|█████████▏| 44324/48008 [2:32:20<17:00, 3.61it/s] 92%|█████████▏| 44325/48008 [2:32:20<15:30, 3.96it/s] 92%|█████████▏| 44326/48008 [2:32:21<14:24, 4.26it/s] 92%|█████████▏| 44327/48008 [2:32:22<31:36, 1.94it/s] 92%|█████████▏| 44328/48008 [2:32:22<25:38, 2.39it/s] 92%|█████████▏| 44329/48008 [2:32:22<21:35, 2.84it/s] 92%|█████████▏| 44330/48008 [2:32:22<18:38, 3.29it/s] 92%|█████████▏| 44331/48008 [2:32:23<16:36, 3.69it/s] 92%|█████████▏| 44332/48008 [2:32:23<15:12, 4.03it/s] 92%|█████████▏| 44333/48008 [2:32:23<14:04, 4.35it/s] 92%|█████████▏| 44334/48008 [2:32:23<13:36, 4.50it/s] 92%|█████████▏| 44335/48008 [2:32:23<13:09, 4.65it/s] 92%|█████████▏| 44336/48008 [2:32:23<12:39, 4.83it/s] 92%|█████████▏| 44337/48008 [2:32:24<13:13, 4.62it/s] 92%|█████████▏| 44338/48008 [2:32:24<12:57, 4.72it/s] 92%|█████████▏| 44339/48008 [2:32:24<12:32, 4.88it/s] 92%|█████████▏| 44340/48008 [2:32:24<12:15, 4.99it/s] 92%|█████████▏| 44341/48008 [2:32:24<12:10, 5.02it/s] 92%|█████████▏| 44342/48008 [2:32:25<12:21, 4.94it/s] 92%|█████████▏| 44343/48008 [2:32:25<12:02, 5.07it/s] 92%|█████████▏| 44344/48008 [2:32:25<12:08, 5.03it/s] 92%|█████████▏| 44345/48008 [2:32:25<12:00, 5.08it/s] 92%|█████████▏| 44346/48008 [2:32:25<11:48, 5.17it/s] 92%|█████████▏| 44347/48008 [2:32:26<15:19, 3.98it/s] 92%|█████████▏| 44348/48008 [2:32:26<14:17, 4.27it/s] 92%|█████████▏| 44349/48008 [2:32:26<13:42, 4.45it/s] 92%|█████████▏| 44350/48008 [2:32:26<13:06, 4.65it/s] 92%|█████████▏| 44350/48008 [2:32:26<13:06, 4.65it/s]{'loss': 0.301, 'grad_norm': 0.1533203125, 'learning_rate': 1.5243292784535912e-05, 'epoch': 0.92} 92%|█████████▏| 44351/48008 [2:32:27<12:53, 4.73it/s] 92%|█████████▏| 44352/48008 [2:32:27<12:33, 4.85it/s] 92%|█████████▏| 44353/48008 [2:32:27<12:33, 4.85it/s] 92%|█████████▏| 44354/48008 [2:32:27<12:15, 4.97it/s] 92%|█████████▏| 44355/48008 [2:32:27<11:59, 5.08it/s] 92%|█████████▏| 44356/48008 [2:32:28<12:06, 5.02it/s] 92%|█████████▏| 44357/48008 [2:32:28<11:59, 5.07it/s] 92%|█████████▏| 44358/48008 [2:32:28<12:08, 5.01it/s] 92%|█████████▏| 44359/48008 [2:32:28<12:10, 5.00it/s] 92%|█████████▏| 44360/48008 [2:32:28<12:06, 5.02it/s] 92%|█████████▏| 44361/48008 [2:32:29<11:57, 5.08it/s] 92%|█████████▏| 44362/48008 [2:32:29<12:43, 4.78it/s] 92%|█████████▏| 44363/48008 [2:32:29<12:25, 4.89it/s] 92%|█████████▏| 44364/48008 [2:32:30<25:18, 2.40it/s] 92%|█████████▏| 44365/48008 [2:32:30<21:31, 2.82it/s] 92%|█████████▏| 44366/48008 [2:32:30<18:28, 3.29it/s] 92%|█████████▏| 44367/48008 [2:32:31<16:31, 3.67it/s] 92%|█████████▏| 44368/48008 [2:32:31<15:04, 4.02it/s] 92%|█████████▏| 44369/48008 [2:32:31<14:16, 4.25it/s] 92%|█████████▏| 44370/48008 [2:32:31<13:41, 4.43it/s] 92%|█████████▏| 44371/48008 [2:32:31<13:03, 4.64it/s] 92%|█████████▏| 44372/48008 [2:32:32<12:33, 4.83it/s] 92%|█████████▏| 44373/48008 [2:32:32<12:22, 4.89it/s] 92%|█████████▏| 44374/48008 [2:32:32<12:06, 5.00it/s] 92%|█████████▏| 44375/48008 [2:32:32<12:07, 5.00it/s] 92%|█████████▏| 44376/48008 [2:32:32<12:10, 4.97it/s] 92%|█████████▏| 44377/48008 [2:32:33<11:58, 5.06it/s] 92%|█████████▏| 44378/48008 [2:32:33<11:53, 5.09it/s] 92%|█████████▏| 44379/48008 [2:32:33<11:48, 5.12it/s] 92%|█████████▏| 44380/48008 [2:32:33<11:43, 5.16it/s] 92%|█████████▏| 44381/48008 [2:32:33<11:34, 5.22it/s] 92%|█████████▏| 44382/48008 [2:32:33<11:30, 5.25it/s] 92%|█████████▏| 44383/48008 [2:32:34<11:32, 5.23it/s] 92%|█████████▏| 44384/48008 [2:32:34<11:29, 5.26it/s] 92%|█████████▏| 44385/48008 [2:32:34<11:51, 5.09it/s] 92%|█████████▏| 44386/48008 [2:32:34<11:43, 5.15it/s] 92%|█████████▏| 44387/48008 [2:32:35<22:58, 2.63it/s] 92%|█████████▏| 44388/48008 [2:32:35<19:32, 3.09it/s] 92%|█████████▏| 44389/48008 [2:32:35<17:09, 3.52it/s] 92%|█████████▏| 44390/48008 [2:32:36<15:26, 3.91it/s] 92%|█████████▏| 44391/48008 [2:32:36<14:25, 4.18it/s] 92%|█████████▏| 44392/48008 [2:32:36<13:32, 4.45it/s] 92%|█████████▏| 44393/48008 [2:32:36<13:01, 4.63it/s] 92%|█████████▏| 44394/48008 [2:32:36<12:31, 4.81it/s] 92%|█████████▏| 44395/48008 [2:32:37<12:12, 4.93it/s] 92%|█████████▏| 44396/48008 [2:32:37<12:03, 4.99it/s] 92%|█████████▏| 44397/48008 [2:32:37<12:02, 5.00it/s] 92%|█████████▏| 44398/48008 [2:32:37<11:58, 5.02it/s] 92%|█████████▏| 44399/48008 [2:32:37<12:04, 4.98it/s] 92%|█████████▏| 44400/48008 [2:32:38<11:56, 5.04it/s] {'loss': 0.2987, 'grad_norm': 0.29296875, 'learning_rate': 1.5034994167638728e-05, 'epoch': 0.92} 92%|█████████▏| 44400/48008 [2:32:38<11:56, 5.04it/s] 92%|█████████▏| 44401/48008 [2:32:38<11:49, 5.08it/s] 92%|█████████▏| 44402/48008 [2:32:38<11:42, 5.13it/s] 92%|█████████▏| 44403/48008 [2:32:38<11:36, 5.17it/s] 92%|█████████▏| 44404/48008 [2:32:38<11:41, 5.14it/s] 92%|█████████▏| 44405/48008 [2:32:39<11:56, 5.03it/s] 92%|█████████▏| 44406/48008 [2:32:39<13:51, 4.33it/s] 92%|█████████▏| 44407/48008 [2:32:39<13:23, 4.48it/s] 93%|█████████▎| 44408/48008 [2:32:39<12:57, 4.63it/s] 93%|█████████▎| 44409/48008 [2:32:39<12:33, 4.78it/s] 93%|█████████▎| 44410/48008 [2:32:40<12:24, 4.83it/s] 93%|█████████▎| 44411/48008 [2:32:40<12:04, 4.97it/s] 93%|█████████▎| 44412/48008 [2:32:40<11:46, 5.09it/s] 93%|█████████▎| 44413/48008 [2:32:40<11:49, 5.07it/s] 93%|█████████▎| 44414/48008 [2:32:40<11:56, 5.02it/s] 93%|█████████▎| 44415/48008 [2:32:41<11:38, 5.14it/s] 93%|█████████▎| 44416/48008 [2:32:41<11:36, 5.16it/s] 93%|█████████▎| 44417/48008 [2:32:41<11:33, 5.18it/s] 93%|█████████▎| 44418/48008 [2:32:41<11:30, 5.20it/s] 93%|█████████▎| 44419/48008 [2:32:41<11:33, 5.18it/s] 93%|█████████▎| 44420/48008 [2:32:42<12:21, 4.84it/s] 93%|█████████▎| 44421/48008 [2:32:42<12:06, 4.94it/s] 93%|█████████▎| 44422/48008 [2:32:42<11:56, 5.01it/s] 93%|█████████▎| 44423/48008 [2:32:42<11:48, 5.06it/s] 93%|█████████▎| 44424/48008 [2:32:42<11:40, 5.11it/s] 93%|█████████▎| 44425/48008 [2:32:43<11:35, 5.15it/s] 93%|█████████▎| 44426/48008 [2:32:43<11:38, 5.13it/s] 93%|█████████▎| 44427/48008 [2:32:43<11:29, 5.20it/s] 93%|█████████▎| 44428/48008 [2:32:43<11:30, 5.18it/s] 93%|█████████▎| 44429/48008 [2:32:43<11:30, 5.18it/s] 93%|█████████▎| 44430/48008 [2:32:44<14:58, 3.98it/s] 93%|█████████▎| 44431/48008 [2:32:44<14:00, 4.25it/s] 93%|█████████▎| 44432/48008 [2:32:44<13:15, 4.50it/s] 93%|█████████▎| 44433/48008 [2:32:44<12:35, 4.73it/s] 93%|█████████▎| 44434/48008 [2:32:45<12:09, 4.90it/s] 93%|█████████▎| 44435/48008 [2:32:45<12:04, 4.93it/s] 93%|█████████▎| 44436/48008 [2:32:45<11:52, 5.02it/s] 93%|█████████▎| 44437/48008 [2:32:45<11:39, 5.11it/s] 93%|█████████▎| 44438/48008 [2:32:45<11:30, 5.17it/s] 93%|█████████▎| 44439/48008 [2:32:45<11:42, 5.08it/s] 93%|█████████▎| 44440/48008 [2:32:46<11:33, 5.15it/s] 93%|█████████▎| 44441/48008 [2:32:46<11:30, 5.17it/s] 93%|█████████▎| 44442/48008 [2:32:46<11:27, 5.18it/s] 93%|█████████▎| 44443/48008 [2:32:46<11:26, 5.19it/s] 93%|█████████▎| 44444/48008 [2:32:46<12:14, 4.85it/s] 93%|█████████▎| 44445/48008 [2:32:47<15:25, 3.85it/s] 93%|█████████▎| 44446/48008 [2:32:47<14:14, 4.17it/s] 93%|█████████▎| 44447/48008 [2:32:47<13:26, 4.41it/s] 93%|█████████▎| 44448/48008 [2:32:47<13:01, 4.56it/s] 93%|█████████▎| 44449/48008 [2:32:48<12:31, 4.74it/s] 93%|█████████▎| 44450/48008 [2:32:48<12:16, 4.83it/s] {'loss': 0.2983, 'grad_norm': 0.26953125, 'learning_rate': 1.4826695550741543e-05, 'epoch': 0.93} 93%|█████████▎| 44450/48008 [2:32:48<12:16, 4.83it/s] 93%|█████████▎| 44451/48008 [2:32:48<12:23, 4.78it/s] 93%|█████████▎| 44452/48008 [2:32:48<12:04, 4.91it/s] 93%|█████████▎| 44453/48008 [2:32:48<12:01, 4.93it/s] 93%|█████████▎| 44454/48008 [2:32:49<11:48, 5.01it/s] 93%|█████████▎| 44455/48008 [2:32:49<12:28, 4.74it/s] 93%|█████████▎| 44456/48008 [2:32:49<12:16, 4.82it/s] 93%|█████████▎| 44457/48008 [2:32:49<12:05, 4.89it/s] 93%|█████████▎| 44458/48008 [2:32:49<11:50, 5.00it/s] 93%|█████████▎| 44459/48008 [2:32:50<11:43, 5.05it/s] 93%|█████████▎| 44460/48008 [2:32:50<11:31, 5.13it/s] 93%|█████████▎| 44461/48008 [2:32:50<11:24, 5.18it/s] 93%|█████████▎| 44462/48008 [2:32:50<11:30, 5.13it/s] 93%|█████████▎| 44463/48008 [2:32:50<11:25, 5.17it/s] 93%|█████████▎| 44464/48008 [2:32:51<11:24, 5.18it/s] 93%|█████████▎| 44465/48008 [2:32:51<11:41, 5.05it/s] 93%|█████████▎| 44466/48008 [2:32:51<12:24, 4.76it/s] 93%|█████████▎| 44467/48008 [2:32:51<14:08, 4.17it/s] 93%|█████████▎| 44468/48008 [2:32:52<13:37, 4.33it/s] 93%|█████████▎| 44469/48008 [2:32:52<13:14, 4.46it/s] 93%|█████████▎| 44470/48008 [2:32:52<12:37, 4.67it/s] 93%|█████████▎| 44471/48008 [2:32:52<12:06, 4.87it/s] 93%|█████████▎| 44472/48008 [2:32:52<11:46, 5.00it/s] 93%|█████████▎| 44473/48008 [2:32:53<15:06, 3.90it/s] 93%|█████████▎| 44474/48008 [2:32:53<13:57, 4.22it/s] 93%|█████████▎| 44475/48008 [2:32:53<13:08, 4.48it/s] 93%|█████████▎| 44476/48008 [2:32:53<12:39, 4.65it/s] 93%|█████████▎| 44477/48008 [2:32:54<12:17, 4.79it/s] 93%|█████████▎| 44478/48008 [2:32:54<11:57, 4.92it/s] 93%|█████████▎| 44479/48008 [2:32:54<11:41, 5.03it/s] 93%|█████████▎| 44480/48008 [2:32:54<11:32, 5.09it/s] 93%|█████████▎| 44481/48008 [2:32:54<11:17, 5.21it/s] 93%|█████████▎| 44482/48008 [2:32:54<11:23, 5.16it/s] 93%|█████████▎| 44483/48008 [2:32:55<11:19, 5.19it/s] 93%|█████████▎| 44484/48008 [2:32:55<11:20, 5.18it/s] 93%|█████████▎| 44485/48008 [2:32:55<11:16, 5.21it/s] 93%|█████████▎| 44486/48008 [2:32:55<11:18, 5.19it/s] 93%|█████████▎| 44487/48008 [2:32:55<11:16, 5.20it/s] 93%|█████████▎| 44488/48008 [2:32:56<11:19, 5.18it/s] 93%|█████████▎| 44489/48008 [2:32:56<11:16, 5.20it/s] 93%|█████████▎| 44490/48008 [2:32:56<11:16, 5.20it/s] 93%|█████████▎| 44491/48008 [2:32:56<11:14, 5.21it/s] 93%|█████████▎| 44492/48008 [2:32:56<11:11, 5.24it/s] 93%|█████████▎| 44493/48008 [2:32:57<13:18, 4.40it/s] 93%|█████████▎| 44494/48008 [2:32:57<12:39, 4.62it/s] 93%|█████████▎| 44495/48008 [2:32:57<12:13, 4.79it/s] 93%|█████████▎| 44496/48008 [2:32:57<11:56, 4.90it/s] 93%|█████████▎| 44497/48008 [2:32:57<11:49, 4.95it/s] 93%|█████████▎| 44498/48008 [2:32:58<11:37, 5.03it/s] 93%|█████████▎| 44499/48008 [2:32:58<11:44, 4.98it/s] 93%|█████████▎| 44500/48008 [2:32:58<11:28, 5.10it/s] {'loss': 0.3012, 'grad_norm': 0.365234375, 'learning_rate': 1.4618396933844361e-05, 'epoch': 0.93} 93%|█████████▎| 44500/48008 [2:32:58<11:28, 5.10it/s] 93%|█████████▎| 44501/48008 [2:32:58<11:36, 5.04it/s] 93%|█████████▎| 44502/48008 [2:32:58<11:28, 5.10it/s] 93%|█████████▎| 44503/48008 [2:32:59<11:17, 5.18it/s] 93%|█████████▎| 44504/48008 [2:32:59<11:15, 5.19it/s] 93%|█████████▎| 44505/48008 [2:32:59<11:16, 5.18it/s] 93%|█████████▎| 44506/48008 [2:32:59<11:14, 5.19it/s] 93%|█████████▎| 44507/48008 [2:32:59<11:23, 5.12it/s] 93%|█████████▎| 44508/48008 [2:33:00<11:16, 5.17it/s] 93%|█████████▎| 44509/48008 [2:33:00<11:16, 5.18it/s] 93%|█████████▎| 44510/48008 [2:33:00<11:16, 5.17it/s] 93%|█████████▎| 44511/48008 [2:33:00<11:13, 5.20it/s] 93%|█████████▎| 44512/48008 [2:33:00<11:06, 5.25it/s] 93%|█████████▎| 44513/48008 [2:33:01<11:07, 5.24it/s] 93%|█████████▎| 44514/48008 [2:33:01<11:11, 5.21it/s] 93%|█████████▎| 44515/48008 [2:33:01<11:21, 5.13it/s] 93%|█████████▎| 44516/48008 [2:33:01<11:16, 5.16it/s] 93%|█████████▎| 44517/48008 [2:33:01<11:14, 5.17it/s] 93%|█████████▎| 44518/48008 [2:33:02<14:35, 3.99it/s] 93%|█████████▎| 44519/48008 [2:33:02<13:30, 4.31it/s] 93%|█████████▎| 44520/48008 [2:33:02<12:45, 4.56it/s] 93%|█████████▎| 44521/48008 [2:33:02<12:13, 4.75it/s] 93%|█████████▎| 44522/48008 [2:33:02<11:48, 4.92it/s] 93%|█████████▎| 44523/48008 [2:33:03<11:30, 5.04it/s] 93%|█████████▎| 44524/48008 [2:33:03<11:31, 5.04it/s] 93%|█████████▎| 44525/48008 [2:33:03<11:29, 5.05it/s] 93%|█████████▎| 44526/48008 [2:33:03<11:22, 5.10it/s] 93%|█████████▎| 44527/48008 [2:33:03<11:28, 5.06it/s] 93%|█████████▎| 44528/48008 [2:33:04<11:31, 5.03it/s] 93%|█████████▎| 44529/48008 [2:33:04<11:17, 5.14it/s] 93%|█████████▎| 44530/48008 [2:33:04<11:27, 5.06it/s] 93%|█████████▎| 44531/48008 [2:33:04<11:20, 5.11it/s] 93%|█████████▎| 44532/48008 [2:33:04<11:11, 5.18it/s] 93%|█████████▎| 44533/48008 [2:33:05<11:08, 5.20it/s] 93%|█████████▎| 44534/48008 [2:33:05<11:06, 5.21it/s] 93%|█████████▎| 44535/48008 [2:33:05<10:58, 5.27it/s] 93%|█████████▎| 44536/48008 [2:33:05<11:01, 5.25it/s] 93%|█████████▎| 44537/48008 [2:33:05<11:00, 5.25it/s] 93%|█████████▎| 44538/48008 [2:33:06<10:55, 5.29it/s] 93%|█████████▎| 44539/48008 [2:33:06<11:00, 5.25it/s] 93%|█████████▎| 44540/48008 [2:33:06<11:07, 5.19it/s] 93%|█████████▎| 44541/48008 [2:33:06<11:17, 5.12it/s] 93%|█████████▎| 44542/48008 [2:33:06<11:10, 5.17it/s] 93%|█████████▎| 44543/48008 [2:33:07<11:06, 5.20it/s] 93%|█████████▎| 44544/48008 [2:33:07<11:04, 5.21it/s] 93%|█████████▎| 44545/48008 [2:33:07<11:03, 5.22it/s] 93%|█████████▎| 44546/48008 [2:33:07<11:15, 5.13it/s] 93%|█████████▎| 44547/48008 [2:33:07<11:08, 5.18it/s] 93%|█████████▎| 44548/48008 [2:33:07<11:01, 5.23it/s] 93%|█████████▎| 44549/48008 [2:33:08<11:00, 5.24it/s] 93%|█████████▎| 44550/48008 [2:33:08<11:01, 5.23it/s] {'loss': 0.3025, 'grad_norm': 0.50390625, 'learning_rate': 1.4410098316947177e-05, 'epoch': 0.93} 93%|█████████▎| 44550/48008 [2:33:08<11:01, 5.23it/s] 93%|█████████▎| 44551/48008 [2:33:08<11:04, 5.20it/s] 93%|█████████▎| 44552/48008 [2:33:08<11:08, 5.17it/s] 93%|█████████▎| 44553/48008 [2:33:08<11:17, 5.10it/s] 93%|█████████▎| 44554/48008 [2:33:09<11:13, 5.13it/s] 93%|█████████▎| 44555/48008 [2:33:09<11:03, 5.20it/s] 93%|█████████▎| 44556/48008 [2:33:09<11:02, 5.21it/s] 93%|█████████▎| 44557/48008 [2:33:09<11:03, 5.20it/s] 93%|█████████▎| 44558/48008 [2:33:09<10:55, 5.26it/s] 93%|█████████▎| 44559/48008 [2:33:10<11:09, 5.15it/s] 93%|█████████▎| 44560/48008 [2:33:10<11:04, 5.19it/s] 93%|█████████▎| 44561/48008 [2:33:10<11:05, 5.18it/s] 93%|█████████▎| 44562/48008 [2:33:10<11:06, 5.17it/s] 93%|█████████▎| 44563/48008 [2:33:10<11:03, 5.19it/s] 93%|█████████▎| 44564/48008 [2:33:11<11:07, 5.16it/s] 93%|█████████▎| 44565/48008 [2:33:11<10:57, 5.24it/s] 93%|█████████▎| 44566/48008 [2:33:11<10:57, 5.23it/s] 93%|█████████▎| 44567/48008 [2:33:11<10:58, 5.22it/s] 93%|█████████▎| 44568/48008 [2:33:11<11:02, 5.19it/s] 93%|█████████▎| 44569/48008 [2:33:12<10:54, 5.25it/s] 93%|█████████▎| 44570/48008 [2:33:12<10:50, 5.29it/s] 93%|█████████▎| 44571/48008 [2:33:12<11:03, 5.18it/s] 93%|█████████▎| 44572/48008 [2:33:12<10:55, 5.24it/s] 93%|█████████▎| 44573/48008 [2:33:12<10:51, 5.28it/s] 93%|█████████▎| 44574/48008 [2:33:13<11:39, 4.91it/s] 93%|█████████▎| 44575/48008 [2:33:13<11:27, 4.99it/s] 93%|█████████▎| 44576/48008 [2:33:13<11:18, 5.06it/s] 93%|█████████▎| 44577/48008 [2:33:13<11:09, 5.12it/s] 93%|█████████▎| 44578/48008 [2:33:13<11:17, 5.06it/s] 93%|█████████▎| 44579/48008 [2:33:14<11:15, 5.08it/s] 93%|█████████▎| 44580/48008 [2:33:14<11:16, 5.07it/s] 93%|█████████▎| 44581/48008 [2:33:14<11:10, 5.11it/s] 93%|█████████▎| 44582/48008 [2:33:14<11:09, 5.12it/s] 93%|█████████▎| 44583/48008 [2:33:14<11:08, 5.12it/s] 93%|█████████▎| 44584/48008 [2:33:14<11:09, 5.11it/s] 93%|█████████▎| 44585/48008 [2:33:15<11:04, 5.15it/s] 93%|█████████▎| 44586/48008 [2:33:15<10:59, 5.19it/s] 93%|█████████▎| 44587/48008 [2:33:15<11:09, 5.11it/s] 93%|█████████▎| 44588/48008 [2:33:15<11:08, 5.11it/s] 93%|█████████▎| 44589/48008 [2:33:15<11:17, 5.05it/s] 93%|█████████▎| 44590/48008 [2:33:16<11:14, 5.06it/s] 93%|█████████▎| 44591/48008 [2:33:16<11:20, 5.02it/s] 93%|█████████▎| 44592/48008 [2:33:16<11:16, 5.05it/s] 93%|█████████▎| 44593/48008 [2:33:16<11:21, 5.01it/s] 93%|█████████▎| 44594/48008 [2:33:16<11:12, 5.08it/s] 93%|█████████▎| 44595/48008 [2:33:17<11:10, 5.09it/s] 93%|█████████▎| 44596/48008 [2:33:17<11:06, 5.12it/s] 93%|█████████▎| 44597/48008 [2:33:17<11:00, 5.17it/s] 93%|█████████▎| 44598/48008 [2:33:17<10:57, 5.19it/s] 93%|█████████▎| 44599/48008 [2:33:18<14:14, 3.99it/s] 93%|█████████▎| 44600/48008 [2:33:18<13:18, 4.27it/s] {'loss': 0.3006, 'grad_norm': 0.357421875, 'learning_rate': 1.4201799700049992e-05, 'epoch': 0.93} 93%|█████████▎| 44600/48008 [2:33:18<13:18, 4.27it/s] 93%|█████████▎| 44601/48008 [2:33:18<12:31, 4.53it/s] 93%|█████████▎| 44602/48008 [2:33:18<11:59, 4.73it/s] 93%|█████████▎| 44603/48008 [2:33:18<11:39, 4.87it/s] 93%|█████████▎| 44604/48008 [2:33:19<11:24, 4.97it/s] 93%|█████████▎| 44605/48008 [2:33:19<11:14, 5.04it/s] 93%|█████████▎| 44606/48008 [2:33:19<11:09, 5.08it/s] 93%|█████████▎| 44607/48008 [2:33:19<11:14, 5.04it/s] 93%|█████████▎| 44608/48008 [2:33:19<11:10, 5.07it/s] 93%|█████████▎| 44609/48008 [2:33:20<11:05, 5.11it/s] 93%|█████████▎| 44610/48008 [2:33:20<11:02, 5.13it/s] 93%|█████████▎| 44611/48008 [2:33:20<11:05, 5.10it/s] 93%|█████████▎| 44612/48008 [2:33:20<10:58, 5.16it/s] 93%|█████████▎| 44613/48008 [2:33:20<10:54, 5.19it/s] 93%|█████████▎| 44614/48008 [2:33:20<10:49, 5.23it/s] 93%|█████████▎| 44615/48008 [2:33:21<11:00, 5.14it/s] 93%|█████████▎| 44616/48008 [2:33:21<10:51, 5.20it/s] 93%|█████████▎| 44617/48008 [2:33:21<10:59, 5.14it/s] 93%|█████████▎| 44618/48008 [2:33:21<10:54, 5.18it/s] 93%|█████████▎| 44619/48008 [2:33:21<11:10, 5.05it/s] 93%|█████████▎| 44620/48008 [2:33:22<11:07, 5.07it/s] 93%|█████████▎| 44621/48008 [2:33:22<11:02, 5.11it/s] 93%|█████████▎| 44622/48008 [2:33:22<10:59, 5.13it/s] 93%|█████████▎| 44623/48008 [2:33:22<10:59, 5.13it/s] 93%|█████████▎| 44624/48008 [2:33:22<10:57, 5.14it/s] 93%|█████████▎| 44625/48008 [2:33:23<11:07, 5.07it/s] 93%|█████████▎| 44626/48008 [2:33:23<11:01, 5.11it/s] 93%|█████████▎| 44627/48008 [2:33:23<12:55, 4.36it/s] 93%|█████████▎| 44628/48008 [2:33:23<12:29, 4.51it/s] 93%|█████████▎| 44629/48008 [2:33:24<12:04, 4.66it/s] 93%|█████████▎| 44630/48008 [2:33:24<11:35, 4.85it/s] 93%|█████████▎| 44631/48008 [2:33:24<11:21, 4.96it/s] 93%|█████████▎| 44632/48008 [2:33:24<11:11, 5.03it/s] 93%|█████████▎| 44633/48008 [2:33:24<11:06, 5.07it/s] 93%|█████████▎| 44634/48008 [2:33:25<11:00, 5.10it/s] 93%|█████████▎| 44635/48008 [2:33:25<11:07, 5.05it/s] 93%|█████████▎| 44636/48008 [2:33:25<11:02, 5.09it/s] 93%|█████████▎| 44637/48008 [2:33:25<10:53, 5.16it/s] 93%|█████████▎| 44638/48008 [2:33:25<11:09, 5.03it/s] 93%|█████████▎| 44639/48008 [2:33:26<11:06, 5.06it/s] 93%|█████████▎| 44640/48008 [2:33:26<11:07, 5.04it/s] 93%|█████████▎| 44641/48008 [2:33:26<11:02, 5.09it/s] 93%|█████████▎| 44642/48008 [2:33:26<11:09, 5.03it/s] 93%|█████████▎| 44643/48008 [2:33:26<11:13, 4.99it/s] 93%|█████████▎| 44644/48008 [2:33:27<11:07, 5.04it/s] 93%|█████████▎| 44645/48008 [2:33:27<11:46, 4.76it/s] 93%|█████████▎| 44646/48008 [2:33:27<11:26, 4.89it/s] 93%|█████████▎| 44647/48008 [2:33:27<11:14, 4.98it/s] 93%|█████████▎| 44648/48008 [2:33:28<14:21, 3.90it/s] 93%|█████████▎| 44649/48008 [2:33:28<13:19, 4.20it/s] 93%|█████████▎| 44650/48008 [2:33:28<12:35, 4.45it/s] {'loss': 0.3005, 'grad_norm': 0.453125, 'learning_rate': 1.3993501083152808e-05, 'epoch': 0.93} 93%|█████████▎| 44650/48008 [2:33:28<12:35, 4.45it/s] 93%|█████████▎| 44651/48008 [2:33:28<12:04, 4.64it/s] 93%|█████████▎| 44652/48008 [2:33:28<11:44, 4.76it/s] 93%|█████████▎| 44653/48008 [2:33:28<11:30, 4.86it/s] 93%|█████████▎| 44654/48008 [2:33:29<11:14, 4.97it/s] 93%|█████████▎| 44655/48008 [2:33:29<11:07, 5.03it/s] 93%|█████████▎| 44656/48008 [2:33:29<11:05, 5.03it/s] 93%|█████████▎| 44657/48008 [2:33:29<11:06, 5.03it/s] 93%|█████████▎| 44658/48008 [2:33:29<11:00, 5.07it/s] 93%|█████████▎| 44659/48008 [2:33:30<14:10, 3.94it/s] 93%|█████████▎| 44660/48008 [2:33:30<13:18, 4.19it/s] 93%|█████████▎| 44661/48008 [2:33:30<12:33, 4.44it/s] 93%|█████████▎| 44662/48008 [2:33:30<12:03, 4.63it/s] 93%|█████████▎| 44663/48008 [2:33:31<11:40, 4.78it/s] 93%|█████████▎| 44664/48008 [2:33:31<13:16, 4.20it/s] 93%|█████████▎| 44665/48008 [2:33:31<12:31, 4.45it/s] 93%|█████████▎| 44666/48008 [2:33:31<13:54, 4.00it/s] 93%|█████████▎| 44667/48008 [2:33:32<13:42, 4.06it/s] 93%|█████████▎| 44668/48008 [2:33:32<12:48, 4.35it/s] 93%|█████████▎| 44669/48008 [2:33:32<14:05, 3.95it/s] 93%|█████████▎| 44670/48008 [2:33:32<13:12, 4.21it/s] 93%|█████████▎| 44671/48008 [2:33:33<12:36, 4.41it/s] 93%|█████████▎| 44672/48008 [2:33:33<12:02, 4.61it/s] 93%|█████████▎| 44673/48008 [2:33:33<11:33, 4.81it/s] 93%|█████████▎| 44674/48008 [2:33:33<11:18, 4.91it/s] 93%|█████████▎| 44675/48008 [2:33:33<11:02, 5.03it/s] 93%|█████████▎| 44676/48008 [2:33:34<10:53, 5.10it/s] 93%|█████████▎| 44677/48008 [2:33:34<10:47, 5.14it/s] 93%|█████████▎| 44678/48008 [2:33:34<10:42, 5.18it/s] 93%|█████████▎| 44679/48008 [2:33:34<10:40, 5.19it/s] 93%|█████████▎| 44680/48008 [2:33:34<10:42, 5.18it/s] 93%|█████████▎| 44681/48008 [2:33:34<10:45, 5.16it/s] 93%|█████████▎| 44682/48008 [2:33:35<10:47, 5.14it/s] 93%|█████████▎| 44683/48008 [2:33:35<10:46, 5.14it/s] 93%|█████████▎| 44684/48008 [2:33:35<10:38, 5.21it/s] 93%|█████████▎| 44685/48008 [2:33:35<10:35, 5.23it/s] 93%|█████████▎| 44686/48008 [2:33:35<10:42, 5.17it/s] 93%|█████████▎| 44687/48008 [2:33:36<10:50, 5.10it/s] 93%|█████████▎| 44688/48008 [2:33:36<10:51, 5.10it/s] 93%|█████████▎| 44689/48008 [2:33:36<10:45, 5.14it/s] 93%|█████████▎| 44690/48008 [2:33:36<10:41, 5.17it/s] 93%|█████████▎| 44691/48008 [2:33:36<10:49, 5.10it/s] 93%|█████████▎| 44692/48008 [2:33:37<10:59, 5.03it/s] 93%|█████████▎| 44693/48008 [2:33:37<10:48, 5.11it/s] 93%|█████████▎| 44694/48008 [2:33:37<10:44, 5.14it/s] 93%|█████████▎| 44695/48008 [2:33:37<11:26, 4.82it/s] 93%|█████████▎| 44696/48008 [2:33:37<11:23, 4.85it/s] 93%|█████████▎| 44697/48008 [2:33:38<11:03, 4.99it/s] 93%|█████████▎| 44698/48008 [2:33:38<10:56, 5.04it/s] 93%|█████████▎| 44699/48008 [2:33:38<10:54, 5.06it/s] 93%|█████████▎| 44700/48008 [2:33:38<10:46, 5.11it/s] 93%|█████████▎| 44700/48008 [2:33:38<10:46, 5.11it/s]{'loss': 0.3006, 'grad_norm': 0.25390625, 'learning_rate': 1.3785202466255625e-05, 'epoch': 0.93} 93%|█████████▎| 44701/48008 [2:33:38<10:44, 5.13it/s] 93%|█████████▎| 44702/48008 [2:33:39<10:56, 5.04it/s] 93%|█████████▎| 44703/48008 [2:33:39<10:50, 5.08it/s] 93%|█████████▎| 44704/48008 [2:33:39<10:58, 5.02it/s] 93%|█████████▎| 44705/48008 [2:33:39<11:03, 4.98it/s] 93%|█████████▎| 44706/48008 [2:33:39<10:52, 5.06it/s] 93%|█████████▎| 44707/48008 [2:33:40<10:51, 5.07it/s] 93%|█████████▎| 44708/48008 [2:33:40<10:41, 5.14it/s] 93%|█████████▎| 44709/48008 [2:33:40<13:50, 3.97it/s] 93%|█████████▎| 44710/48008 [2:33:40<13:03, 4.21it/s] 93%|█████████▎| 44711/48008 [2:33:41<12:20, 4.45it/s] 93%|█████████▎| 44712/48008 [2:33:41<11:50, 4.64it/s] 93%|█████████▎| 44713/48008 [2:33:41<11:34, 4.75it/s] 93%|█████████▎| 44714/48008 [2:33:41<11:16, 4.87it/s] 93%|█████████▎| 44715/48008 [2:33:41<11:03, 4.96it/s] 93%|█████████▎| 44716/48008 [2:33:42<10:54, 5.03it/s] 93%|█████████▎| 44717/48008 [2:33:42<10:51, 5.05it/s] 93%|█████████▎| 44718/48008 [2:33:42<13:57, 3.93it/s] 93%|█████████▎| 44719/48008 [2:33:42<13:07, 4.18it/s] 93%|█████████▎| 44720/48008 [2:33:43<12:20, 4.44it/s] 93%|█████████▎| 44721/48008 [2:33:43<11:46, 4.65it/s] 93%|█████████▎| 44722/48008 [2:33:43<11:27, 4.78it/s] 93%|█████████▎| 44723/48008 [2:33:43<11:11, 4.90it/s] 93%|█████████▎| 44724/48008 [2:33:43<11:44, 4.66it/s] 93%|█████████▎| 44725/48008 [2:33:44<11:23, 4.81it/s] 93%|█████████▎| 44726/48008 [2:33:44<11:05, 4.93it/s] 93%|█████████▎| 44727/48008 [2:33:44<10:57, 4.99it/s] 93%|█████████▎| 44728/48008 [2:33:44<10:55, 5.01it/s] 93%|█████████▎| 44729/48008 [2:33:44<10:46, 5.07it/s] 93%|█████████▎| 44730/48008 [2:33:45<10:52, 5.02it/s] 93%|█████████▎| 44731/48008 [2:33:45<10:45, 5.07it/s] 93%|█████████▎| 44732/48008 [2:33:45<10:41, 5.11it/s] 93%|█████████▎| 44733/48008 [2:33:45<10:38, 5.13it/s] 93%|█████████▎| 44734/48008 [2:33:45<10:40, 5.11it/s] 93%|█████████▎| 44735/48008 [2:33:46<11:26, 4.77it/s] 93%|█████████▎| 44736/48008 [2:33:46<11:06, 4.91it/s] 93%|█████████▎| 44737/48008 [2:33:46<10:53, 5.01it/s] 93%|█████████▎| 44738/48008 [2:33:46<10:54, 5.00it/s] 93%|█████████▎| 44739/48008 [2:33:46<10:44, 5.07it/s] 93%|█████████▎| 44740/48008 [2:33:47<10:38, 5.12it/s] 93%|█████████▎| 44741/48008 [2:33:47<10:32, 5.17it/s] 93%|█████████▎| 44742/48008 [2:33:47<10:28, 5.19it/s] 93%|█████████▎| 44743/48008 [2:33:47<10:24, 5.22it/s] 93%|█████████▎| 44744/48008 [2:33:47<10:22, 5.24it/s] 93%|█████████▎| 44745/48008 [2:33:47<10:24, 5.22it/s] 93%|█████████▎| 44746/48008 [2:33:48<10:25, 5.21it/s] 93%|█████████▎| 44747/48008 [2:33:48<10:22, 5.24it/s] 93%|█████████▎| 44748/48008 [2:33:48<10:24, 5.22it/s] 93%|█████████▎| 44749/48008 [2:33:48<10:35, 5.13it/s] 93%|█████████▎| 44750/48008 [2:33:48<10:31, 5.16it/s] {'loss': 0.2983, 'grad_norm': 0.2060546875, 'learning_rate': 1.3576903849358441e-05, 'epoch': 0.93} 93%|█████████▎| 44750/48008 [2:33:48<10:31, 5.16it/s] 93%|█████████▎| 44751/48008 [2:33:49<10:31, 5.16it/s] 93%|█████████▎| 44752/48008 [2:33:49<11:15, 4.82it/s] 93%|█████████▎| 44753/48008 [2:33:49<10:59, 4.93it/s] 93%|█████████▎| 44754/48008 [2:33:49<10:42, 5.07it/s] 93%|█████████▎| 44755/48008 [2:33:49<10:38, 5.09it/s] 93%|█████████▎| 44756/48008 [2:33:50<10:35, 5.11it/s] 93%|█████████▎| 44757/48008 [2:33:50<10:28, 5.18it/s] 93%|█████████▎| 44758/48008 [2:33:50<10:24, 5.21it/s] 93%|█████████▎| 44759/48008 [2:33:50<10:26, 5.18it/s] 93%|█████████▎| 44760/48008 [2:33:50<11:11, 4.84it/s] 93%|█████████▎| 44761/48008 [2:33:51<11:14, 4.81it/s] 93%|█████████▎| 44762/48008 [2:33:51<10:56, 4.95it/s] 93%|█████████▎| 44763/48008 [2:33:51<10:43, 5.04it/s] 93%|█████████▎| 44764/48008 [2:33:51<10:38, 5.08it/s] 93%|█████████▎| 44765/48008 [2:33:51<10:44, 5.03it/s] 93%|█████████▎| 44766/48008 [2:33:52<10:47, 5.01it/s] 93%|█████████▎| 44767/48008 [2:33:52<10:38, 5.08it/s] 93%|█████████▎| 44768/48008 [2:33:52<10:35, 5.10it/s] 93%|█████████▎| 44769/48008 [2:33:52<10:31, 5.13it/s] 93%|█████████▎| 44770/48008 [2:33:53<13:37, 3.96it/s] 93%|█████████▎| 44771/48008 [2:33:53<12:39, 4.26it/s] 93%|█████████▎| 44772/48008 [2:33:53<12:05, 4.46it/s] 93%|█████████▎| 44773/48008 [2:33:53<12:16, 4.39it/s] 93%|█████████▎| 44774/48008 [2:33:53<11:47, 4.57it/s] 93%|█████████▎| 44775/48008 [2:33:54<11:22, 4.74it/s] 93%|█████████▎| 44776/48008 [2:33:54<11:01, 4.89it/s] 93%|█████████▎| 44777/48008 [2:33:54<11:02, 4.88it/s] 93%|█████████▎| 44778/48008 [2:33:54<10:49, 4.97it/s] 93%|█████████▎| 44779/48008 [2:33:54<10:39, 5.05it/s] 93%|█████████▎| 44780/48008 [2:33:55<10:32, 5.11it/s] 93%|█████████▎| 44781/48008 [2:33:55<10:45, 5.00it/s] 93%|█████████▎| 44782/48008 [2:33:55<10:43, 5.01it/s] 93%|█████████▎| 44783/48008 [2:33:55<10:41, 5.03it/s] 93%|█████████▎| 44784/48008 [2:33:55<10:34, 5.08it/s] 93%|█████████▎| 44785/48008 [2:33:56<10:36, 5.06it/s] 93%|█████████▎| 44786/48008 [2:33:56<10:28, 5.12it/s] 93%|█████████▎| 44787/48008 [2:33:56<10:36, 5.06it/s] 93%|█████████▎| 44788/48008 [2:33:56<10:24, 5.16it/s] 93%|█████████▎| 44789/48008 [2:33:56<10:21, 5.18it/s] 93%|█████████▎| 44790/48008 [2:33:57<11:04, 4.84it/s] 93%|█████████▎| 44791/48008 [2:33:57<10:50, 4.95it/s] 93%|█████████▎| 44792/48008 [2:33:57<10:53, 4.92it/s] 93%|█████████▎| 44793/48008 [2:33:57<10:44, 4.99it/s] 93%|█████████▎| 44794/48008 [2:33:57<10:35, 5.06it/s] 93%|█████████▎| 44795/48008 [2:33:58<10:32, 5.08it/s] 93%|█████████▎| 44796/48008 [2:33:58<10:28, 5.11it/s] 93%|█████████▎| 44797/48008 [2:33:58<10:27, 5.12it/s] 93%|█████████▎| 44798/48008 [2:33:58<10:24, 5.14it/s] 93%|█████████▎| 44799/48008 [2:33:58<10:21, 5.17it/s] 93%|█████████▎| 44800/48008 [2:33:59<10:22, 5.16it/s] {'loss': 0.2985, 'grad_norm': 0.33203125, 'learning_rate': 1.3368605232461257e-05, 'epoch': 0.93} 93%|█████████▎| 44800/48008 [2:33:59<10:22, 5.16it/s] 93%|█████████▎| 44801/48008 [2:33:59<10:22, 5.15it/s] 93%|█████████▎| 44802/48008 [2:33:59<10:25, 5.13it/s] 93%|█████████▎| 44803/48008 [2:33:59<10:21, 5.16it/s] 93%|█████████▎| 44804/48008 [2:33:59<10:29, 5.09it/s] 93%|█████████▎| 44805/48008 [2:34:00<12:16, 4.35it/s] 93%|█████████▎| 44806/48008 [2:34:00<11:39, 4.58it/s] 93%|█████████▎| 44807/48008 [2:34:00<11:19, 4.71it/s] 93%|█████████▎| 44808/48008 [2:34:00<10:55, 4.88it/s] 93%|█████████▎| 44809/48008 [2:34:00<10:56, 4.87it/s] 93%|█████████▎| 44810/48008 [2:34:01<10:54, 4.89it/s] 93%|█████████▎| 44811/48008 [2:34:01<10:47, 4.94it/s] 93%|█████████▎| 44812/48008 [2:34:01<10:41, 4.98it/s] 93%|█████████▎| 44813/48008 [2:34:01<10:31, 5.06it/s] 93%|█████████▎| 44814/48008 [2:34:01<10:25, 5.10it/s] 93%|█████████▎| 44815/48008 [2:34:02<10:23, 5.12it/s] 93%|█████████▎| 44816/48008 [2:34:02<10:21, 5.14it/s] 93%|█████████▎| 44817/48008 [2:34:02<10:20, 5.14it/s] 93%|█████████▎| 44818/48008 [2:34:02<10:18, 5.16it/s] 93%|█████████▎| 44819/48008 [2:34:02<10:34, 5.03it/s] 93%|█████████▎| 44820/48008 [2:34:03<10:25, 5.09it/s] 93%|█████████▎| 44821/48008 [2:34:03<10:15, 5.18it/s] 93%|█████████▎| 44822/48008 [2:34:03<10:20, 5.13it/s] 93%|█████████▎| 44823/48008 [2:34:03<10:09, 5.22it/s] 93%|█████████▎| 44824/48008 [2:34:03<10:08, 5.23it/s] 93%|█████████▎| 44825/48008 [2:34:04<10:05, 5.25it/s] 93%|█████████▎| 44826/48008 [2:34:04<10:12, 5.19it/s] 93%|█████████▎| 44827/48008 [2:34:04<10:14, 5.18it/s] 93%|█████████▎| 44828/48008 [2:34:04<10:13, 5.18it/s] 93%|█████████▎| 44829/48008 [2:34:04<10:10, 5.21it/s] 93%|█████████▎| 44830/48008 [2:34:04<10:07, 5.23it/s] 93%|█████████▎| 44831/48008 [2:34:05<10:18, 5.13it/s] 93%|█████████▎| 44832/48008 [2:34:05<10:14, 5.17it/s] 93%|█████████▎| 44833/48008 [2:34:05<10:08, 5.22it/s] 93%|█████████▎| 44834/48008 [2:34:05<10:19, 5.12it/s] 93%|█████████▎| 44835/48008 [2:34:05<10:17, 5.14it/s] 93%|█████████▎| 44836/48008 [2:34:06<10:15, 5.15it/s] 93%|█████████▎| 44837/48008 [2:34:06<10:24, 5.07it/s] 93%|█████████▎| 44838/48008 [2:34:06<10:21, 5.10it/s] 93%|█████████▎| 44839/48008 [2:34:06<10:16, 5.14it/s] 93%|█████████▎| 44840/48008 [2:34:07<13:19, 3.96it/s] 93%|█████████▎| 44841/48008 [2:34:07<12:21, 4.27it/s] 93%|█████████▎| 44842/48008 [2:34:07<11:32, 4.57it/s] 93%|█████████▎| 44843/48008 [2:34:07<11:17, 4.67it/s] 93%|█████████▎| 44844/48008 [2:34:07<10:51, 4.86it/s] 93%|█████████▎| 44845/48008 [2:34:08<10:36, 4.97it/s] 93%|█████████▎| 44846/48008 [2:34:08<10:21, 5.08it/s] 93%|█████████▎| 44847/48008 [2:34:08<12:06, 4.35it/s] 93%|█████████▎| 44848/48008 [2:34:08<11:31, 4.57it/s] 93%|█████████▎| 44849/48008 [2:34:08<11:09, 4.72it/s] 93%|█████████▎| 44850/48008 [2:34:09<11:01, 4.78it/s] {'loss': 0.302, 'grad_norm': 0.14453125, 'learning_rate': 1.3160306615564072e-05, 'epoch': 0.93} 93%|█████████▎| 44850/48008 [2:34:09<11:01, 4.78it/s] 93%|█████████▎| 44851/48008 [2:34:09<10:49, 4.86it/s] 93%|█████████▎| 44852/48008 [2:34:09<10:41, 4.92it/s] 93%|█████████▎| 44853/48008 [2:34:09<10:29, 5.01it/s] 93%|█████████▎| 44854/48008 [2:34:09<10:25, 5.04it/s] 93%|█████████▎| 44855/48008 [2:34:10<10:15, 5.12it/s] 93%|█████████▎| 44856/48008 [2:34:10<10:06, 5.19it/s] 93%|█████████▎| 44857/48008 [2:34:10<10:07, 5.18it/s] 93%|█████████▎| 44858/48008 [2:34:10<10:22, 5.06it/s] 93%|█████████▎| 44859/48008 [2:34:10<10:15, 5.12it/s] 93%|█████████▎| 44860/48008 [2:34:11<10:25, 5.03it/s] 93%|█████████▎| 44861/48008 [2:34:11<10:27, 5.02it/s] 93%|█████████▎| 44862/48008 [2:34:11<10:18, 5.09it/s] 93%|█████████▎| 44863/48008 [2:34:11<13:19, 3.94it/s] 93%|█████████▎| 44864/48008 [2:34:12<12:22, 4.23it/s] 93%|█████████▎| 44865/48008 [2:34:12<11:37, 4.51it/s] 93%|█████████▎| 44866/48008 [2:34:12<11:10, 4.69it/s] 93%|█████████▎| 44867/48008 [2:34:12<10:59, 4.76it/s] 93%|█████████▎| 44868/48008 [2:34:12<10:46, 4.86it/s] 93%|█████████▎| 44869/48008 [2:34:13<10:35, 4.94it/s] 93%|█████████▎| 44870/48008 [2:34:13<10:25, 5.02it/s] 93%|█████████▎| 44871/48008 [2:34:13<12:06, 4.32it/s] 93%|█████████▎| 44872/48008 [2:34:13<11:29, 4.55it/s] 93%|█████████▎| 44873/48008 [2:34:13<11:00, 4.74it/s] 93%|█████████▎| 44874/48008 [2:34:14<10:42, 4.87it/s] 93%|█████████▎| 44875/48008 [2:34:14<10:28, 4.98it/s] 93%|█████████▎| 44876/48008 [2:34:14<10:20, 5.05it/s] 93%|█████████▎| 44877/48008 [2:34:14<10:20, 5.05it/s] 93%|█████████▎| 44878/48008 [2:34:14<10:12, 5.11it/s] 93%|█████████▎| 44879/48008 [2:34:15<10:02, 5.19it/s] 93%|█████████▎| 44880/48008 [2:34:15<10:00, 5.21it/s] 93%|█████████▎| 44881/48008 [2:34:15<10:09, 5.13it/s] 93%|█████████▎| 44882/48008 [2:34:15<10:16, 5.07it/s] 93%|█████████▎| 44883/48008 [2:34:15<10:31, 4.95it/s] 93%|█████████▎| 44884/48008 [2:34:16<10:13, 5.09it/s] 93%|█████████▎| 44885/48008 [2:34:16<10:20, 5.03it/s] 93%|█████████▎| 44886/48008 [2:34:16<10:10, 5.11it/s] 93%|█████████▎| 44887/48008 [2:34:16<10:50, 4.80it/s] 94%|█████████▎| 44888/48008 [2:34:16<10:33, 4.92it/s] 94%|█████████▎| 44889/48008 [2:34:17<10:16, 5.06it/s] 94%|█████████▎| 44890/48008 [2:34:17<10:30, 4.94it/s] 94%|█████████▎| 44891/48008 [2:34:17<10:20, 5.02it/s] 94%|█████████▎| 44892/48008 [2:34:17<10:15, 5.06it/s] 94%|█████████▎| 44893/48008 [2:34:17<10:22, 5.00it/s] 94%|█████████▎| 44894/48008 [2:34:18<10:11, 5.09it/s] 94%|█████████▎| 44895/48008 [2:34:18<10:07, 5.12it/s] 94%|█████████▎| 44896/48008 [2:34:18<10:02, 5.17it/s] 94%|█████████▎| 44897/48008 [2:34:18<10:01, 5.17it/s] 94%|█████████▎| 44898/48008 [2:34:18<09:55, 5.22it/s] 94%|█████████▎| 44899/48008 [2:34:19<10:04, 5.14it/s] 94%|█████████▎| 44900/48008 [2:34:19<10:19, 5.02it/s] {'loss': 0.299, 'grad_norm': 0.15234375, 'learning_rate': 1.295200799866689e-05, 'epoch': 0.94} 94%|█████████▎| 44900/48008 [2:34:19<10:19, 5.02it/s] 94%|█████████▎| 44901/48008 [2:34:19<10:13, 5.07it/s] 94%|█████████▎| 44902/48008 [2:34:20<25:42, 2.01it/s] 94%|█████████▎| 44903/48008 [2:34:20<21:09, 2.45it/s] 94%|█████████▎| 44904/48008 [2:34:21<17:47, 2.91it/s] 94%|█████████▎| 44905/48008 [2:34:21<15:26, 3.35it/s] 94%|█████████▎| 44906/48008 [2:34:21<13:45, 3.76it/s] 94%|█████████▎| 44907/48008 [2:34:21<12:34, 4.11it/s] 94%|█████████▎| 44908/48008 [2:34:21<13:32, 3.81it/s] 94%|█████████▎| 44909/48008 [2:34:22<13:10, 3.92it/s] 94%|█████████▎| 44910/48008 [2:34:22<13:59, 3.69it/s] 94%|█████████▎| 44911/48008 [2:34:22<12:47, 4.04it/s] 94%|█████████▎| 44912/48008 [2:34:22<13:41, 3.77it/s] 94%|█████████▎| 44913/48008 [2:34:23<12:36, 4.09it/s] 94%|█████████▎| 44914/48008 [2:34:23<11:46, 4.38it/s] 94%|█████████▎| 44915/48008 [2:34:23<11:23, 4.52it/s] 94%|█████████▎| 44916/48008 [2:34:23<10:57, 4.70it/s] 94%|█████████▎| 44917/48008 [2:34:23<10:42, 4.81it/s] 94%|█████████▎| 44918/48008 [2:34:24<10:26, 4.93it/s] 94%|█████████▎| 44919/48008 [2:34:24<10:15, 5.02it/s] 94%|█████████▎| 44920/48008 [2:34:24<10:02, 5.12it/s] 94%|█████████▎| 44921/48008 [2:34:24<09:59, 5.15it/s] 94%|█████████▎| 44922/48008 [2:34:25<24:37, 2.09it/s] 94%|█████████▎| 44923/48008 [2:34:26<20:15, 2.54it/s] 94%|█████████▎| 44924/48008 [2:34:26<17:48, 2.89it/s] 94%|█████████▎| 44925/48008 [2:34:26<15:26, 3.33it/s] 94%|█████████▎| 44926/48008 [2:34:26<13:39, 3.76it/s] 94%|█████████▎| 44927/48008 [2:34:26<12:30, 4.11it/s] 94%|█████████▎| 44928/48008 [2:34:27<11:38, 4.41it/s] 94%|█████████▎| 44929/48008 [2:34:27<11:02, 4.65it/s] 94%|█████████▎| 44930/48008 [2:34:27<10:43, 4.78it/s] 94%|█████████▎| 44931/48008 [2:34:27<10:21, 4.95it/s] 94%|█████████▎| 44932/48008 [2:34:27<10:52, 4.71it/s] 94%|█████████▎| 44933/48008 [2:34:28<10:43, 4.78it/s] 94%|█████████▎| 44934/48008 [2:34:29<22:29, 2.28it/s] 94%|█████████▎| 44935/48008 [2:34:29<18:43, 2.73it/s] 94%|█████████▎| 44936/48008 [2:34:29<15:57, 3.21it/s] 94%|█████████▎| 44937/48008 [2:34:29<14:06, 3.63it/s] 94%|█████████▎| 44938/48008 [2:34:29<12:59, 3.94it/s] 94%|█████████▎| 44939/48008 [2:34:29<12:04, 4.24it/s] 94%|█████████▎| 44940/48008 [2:34:31<26:31, 1.93it/s] 94%|█████████▎| 44941/48008 [2:34:31<21:29, 2.38it/s] 94%|█████████▎| 44942/48008 [2:34:31<17:54, 2.85it/s] 94%|█████████▎| 44943/48008 [2:34:31<18:29, 2.76it/s] 94%|█████████▎| 44944/48008 [2:34:32<15:47, 3.23it/s] 94%|█████████▎| 44945/48008 [2:34:32<13:57, 3.66it/s] 94%|█████████▎| 44946/48008 [2:34:32<12:39, 4.03it/s] 94%|█████████▎| 44947/48008 [2:34:32<11:47, 4.32it/s] 94%|█████████▎| 44948/48008 [2:34:32<11:11, 4.55it/s] 94%|█████████▎| 44949/48008 [2:34:33<10:48, 4.72it/s] 94%|█████████▎| 44950/48008 [2:34:33<10:41, 4.77it/s] {'loss': 0.2991, 'grad_norm': 0.2890625, 'learning_rate': 1.2743709381769705e-05, 'epoch': 0.94} 94%|█████████▎| 44950/48008 [2:34:33<10:41, 4.77it/s] 94%|█████████▎| 44951/48008 [2:34:33<10:22, 4.91it/s] 94%|█████████▎| 44952/48008 [2:34:33<10:12, 4.99it/s] 94%|█████████▎| 44953/48008 [2:34:33<10:05, 5.04it/s] 94%|█████████▎| 44954/48008 [2:34:34<09:58, 5.10it/s] 94%|█████████▎| 44955/48008 [2:34:34<09:50, 5.17it/s] 94%|█████████▎| 44956/48008 [2:34:34<09:48, 5.18it/s] 94%|█████████▎| 44957/48008 [2:34:34<09:44, 5.22it/s] 94%|█████████▎| 44958/48008 [2:34:34<09:44, 5.22it/s] 94%|█████████▎| 44959/48008 [2:34:34<09:45, 5.21it/s] 94%|█████████▎| 44960/48008 [2:34:35<09:47, 5.18it/s] 94%|█████████▎| 44961/48008 [2:34:35<09:47, 5.18it/s] 94%|█████████▎| 44962/48008 [2:34:35<09:46, 5.19it/s] 94%|█████████▎| 44963/48008 [2:34:35<09:47, 5.18it/s] 94%|█████████▎| 44964/48008 [2:34:35<09:45, 5.20it/s] 94%|█████████▎| 44965/48008 [2:34:36<09:42, 5.22it/s] 94%|█████████▎| 44966/48008 [2:34:36<09:47, 5.18it/s] 94%|█████████▎| 44967/48008 [2:34:36<12:43, 3.99it/s] 94%|█████████▎| 44968/48008 [2:34:37<14:46, 3.43it/s] 94%|█████████▎| 44969/48008 [2:34:37<13:20, 3.80it/s] 94%|█████████▎| 44970/48008 [2:34:37<12:12, 4.14it/s] 94%|█████████▎| 44971/48008 [2:34:37<11:27, 4.42it/s] 94%|█████████▎| 44972/48008 [2:34:37<10:58, 4.61it/s] 94%|█████████▎| 44973/48008 [2:34:38<11:20, 4.46it/s] 94%|█████████▎| 44974/48008 [2:34:38<10:45, 4.70it/s] 94%|█████████▎| 44975/48008 [2:34:38<12:11, 4.14it/s] 94%|█████████▎| 44976/48008 [2:34:38<11:37, 4.35it/s] 94%|█████████▎| 44977/48008 [2:34:39<10:57, 4.61it/s] 94%|█████████▎| 44978/48008 [2:34:39<10:33, 4.78it/s] 94%|█████████▎| 44979/48008 [2:34:39<13:13, 3.82it/s] 94%|█████████▎| 44980/48008 [2:34:39<12:21, 4.09it/s] 94%|█████████▎| 44981/48008 [2:34:39<11:40, 4.32it/s] 94%|█████████▎| 44982/48008 [2:34:40<11:06, 4.54it/s] 94%|█████████▎| 44983/48008 [2:34:40<12:25, 4.06it/s] 94%|█████████▎| 44984/48008 [2:34:40<11:33, 4.36it/s] 94%|█████████▎| 44985/48008 [2:34:40<11:06, 4.54it/s] 94%|█████████▎| 44986/48008 [2:34:41<10:39, 4.72it/s] 94%|█████████▎| 44987/48008 [2:34:41<10:22, 4.85it/s] 94%|█████████▎| 44988/48008 [2:34:41<10:08, 4.96it/s] 94%|█████████▎| 44989/48008 [2:34:41<09:59, 5.04it/s] 94%|█████████▎| 44990/48008 [2:34:41<09:49, 5.12it/s] 94%|█████████▎| 44991/48008 [2:34:42<09:44, 5.16it/s] 94%|█████████▎| 44992/48008 [2:34:42<10:25, 4.82it/s] 94%|█████████▎| 44993/48008 [2:34:42<10:06, 4.97it/s] 94%|█████████▎| 44994/48008 [2:34:42<09:54, 5.07it/s] 94%|█████████▎| 44995/48008 [2:34:42<10:00, 5.02it/s] 94%|█████████▎| 44996/48008 [2:34:43<09:58, 5.03it/s] 94%|█████████▎| 44997/48008 [2:34:43<10:01, 5.01it/s] 94%|█████████▎| 44998/48008 [2:34:43<09:48, 5.12it/s] 94%|█████████▎| 44999/48008 [2:34:43<09:44, 5.15it/s] 94%|█████████▎| 45000/48008 [2:34:44<19:28, 2.58it/s] {'loss': 0.2988, 'grad_norm': 0.373046875, 'learning_rate': 1.2535410764872521e-05, 'epoch': 0.94} 94%|█████████▎| 45000/48008 [2:34:44<19:28, 2.58it/s] 94%|█████████▎| 45001/48008 [2:34:44<16:35, 3.02it/s] 94%|█████████▎| 45002/48008 [2:34:44<14:47, 3.39it/s] 94%|█████████▎| 45003/48008 [2:34:45<13:12, 3.79it/s] 94%|█████████▎| 45004/48008 [2:34:45<12:04, 4.15it/s] 94%|█████████▎| 45005/48008 [2:34:45<11:23, 4.39it/s] 94%|█████████▎| 45006/48008 [2:34:45<10:51, 4.61it/s] 94%|█████████▎| 45007/48008 [2:34:45<10:34, 4.73it/s] 94%|█████████▍| 45008/48008 [2:34:46<10:25, 4.80it/s] 94%|█████████▍| 45009/48008 [2:34:46<10:05, 4.95it/s] 94%|█████████▍| 45010/48008 [2:34:46<10:38, 4.70it/s] 94%|█████████▍| 45011/48008 [2:34:46<10:34, 4.73it/s] 94%|█████████▍| 45012/48008 [2:34:46<10:22, 4.82it/s] 94%|█████████▍| 45013/48008 [2:34:47<10:20, 4.82it/s] 94%|█████████▍| 45014/48008 [2:34:47<10:04, 4.96it/s] 94%|█████████▍| 45015/48008 [2:34:47<09:51, 5.06it/s] 94%|█████████▍| 45016/48008 [2:34:47<10:28, 4.76it/s] 94%|█████████▍| 45017/48008 [2:34:47<10:10, 4.90it/s] 94%|█████████▍| 45018/48008 [2:34:48<09:54, 5.03it/s] 94%|█████████▍| 45019/48008 [2:34:48<09:48, 5.08it/s] 94%|█████████▍| 45020/48008 [2:34:48<09:54, 5.03it/s] 94%|█████████▍| 45021/48008 [2:34:48<09:44, 5.11it/s] 94%|█████████▍| 45022/48008 [2:34:48<09:39, 5.15it/s] 94%|█████████▍| 45023/48008 [2:34:49<09:42, 5.13it/s] 94%|█████████▍| 45024/48008 [2:34:49<09:39, 5.15it/s] 94%|█████████▍| 45025/48008 [2:34:49<10:17, 4.83it/s] 94%|█████████▍| 45026/48008 [2:34:49<09:54, 5.02it/s] 94%|█████████▍| 45027/48008 [2:34:50<12:40, 3.92it/s] 94%|█████████▍| 45028/48008 [2:34:50<11:42, 4.24it/s] 94%|█████████▍| 45029/48008 [2:34:50<11:00, 4.51it/s] 94%|█████████▍| 45030/48008 [2:34:50<10:34, 4.69it/s] 94%|█████████▍| 45031/48008 [2:34:50<10:20, 4.80it/s] 94%|█████████▍| 45032/48008 [2:34:51<10:45, 4.61it/s] 94%|█████████▍| 45033/48008 [2:34:51<10:27, 4.74it/s] 94%|█████████▍| 45034/48008 [2:34:51<10:06, 4.91it/s] 94%|█████████▍| 45035/48008 [2:34:51<09:57, 4.98it/s] 94%|█████████▍| 45036/48008 [2:34:51<09:47, 5.06it/s] 94%|█████████▍| 45037/48008 [2:34:52<09:38, 5.14it/s] 94%|█████████▍| 45038/48008 [2:34:52<09:34, 5.17it/s] 94%|█████████▍| 45039/48008 [2:34:52<09:39, 5.12it/s] 94%|█████████▍| 45040/48008 [2:34:52<09:39, 5.12it/s] 94%|█████████▍| 45041/48008 [2:34:52<09:39, 5.12it/s] 94%|█████████▍| 45042/48008 [2:34:52<09:46, 5.06it/s] 94%|█████████▍| 45043/48008 [2:34:53<09:35, 5.15it/s] 94%|█████████▍| 45044/48008 [2:34:53<09:32, 5.18it/s] 94%|█████████▍| 45045/48008 [2:34:53<09:27, 5.23it/s] 94%|█████████▍| 45046/48008 [2:34:53<09:24, 5.24it/s] 94%|█████████▍| 45047/48008 [2:34:53<09:27, 5.22it/s] 94%|█████████▍| 45048/48008 [2:34:54<09:29, 5.19it/s] 94%|█████████▍| 45049/48008 [2:34:54<09:44, 5.06it/s] 94%|█████████▍| 45050/48008 [2:34:54<09:54, 4.97it/s] {'loss': 0.2991, 'grad_norm': 0.2177734375, 'learning_rate': 1.2327112147975338e-05, 'epoch': 0.94} 94%|█████████▍| 45050/48008 [2:34:54<09:54, 4.97it/s] 94%|█████████▍| 45051/48008 [2:34:54<09:49, 5.02it/s] 94%|█████████▍| 45052/48008 [2:34:54<09:52, 4.99it/s] 94%|█████████▍| 45053/48008 [2:34:55<09:37, 5.12it/s] 94%|█████████▍| 45054/48008 [2:34:55<09:33, 5.15it/s] 94%|█████████▍| 45055/48008 [2:34:55<10:11, 4.83it/s] 94%|█████████▍| 45056/48008 [2:34:55<09:58, 4.93it/s] 94%|█████████▍| 45057/48008 [2:34:55<09:42, 5.07it/s] 94%|█████████▍| 45058/48008 [2:34:56<10:17, 4.78it/s] 94%|█████████▍| 45059/48008 [2:34:56<10:02, 4.89it/s] 94%|█████████▍| 45060/48008 [2:34:56<10:08, 4.85it/s] 94%|█████████▍| 45061/48008 [2:34:56<09:57, 4.94it/s] 94%|█████████▍| 45062/48008 [2:34:56<09:43, 5.05it/s] 94%|█████████▍| 45063/48008 [2:34:57<09:36, 5.11it/s] 94%|█████████▍| 45064/48008 [2:34:57<09:27, 5.19it/s] 94%|█████████▍| 45065/48008 [2:34:57<09:26, 5.20it/s] 94%|█████████▍| 45066/48008 [2:34:57<09:43, 5.05it/s] 94%|█████████▍| 45067/48008 [2:34:57<09:35, 5.11it/s] 94%|█████████▍| 45068/48008 [2:34:58<09:27, 5.18it/s] 94%|█████████▍| 45069/48008 [2:34:58<09:24, 5.21it/s] 94%|█████████▍| 45070/48008 [2:34:58<09:22, 5.22it/s] 94%|█████████▍| 45071/48008 [2:34:58<09:19, 5.25it/s] 94%|█████████▍| 45072/48008 [2:34:58<09:30, 5.15it/s] 94%|█████████▍| 45073/48008 [2:34:59<09:36, 5.09it/s] 94%|█████████▍| 45074/48008 [2:34:59<09:43, 5.02it/s] 94%|█████████▍| 45075/48008 [2:34:59<09:47, 4.99it/s] 94%|█████████▍| 45076/48008 [2:34:59<09:50, 4.97it/s] 94%|█████████▍| 45077/48008 [2:34:59<09:39, 5.05it/s] 94%|█████████▍| 45078/48008 [2:35:00<09:29, 5.15it/s] 94%|█████████▍| 45079/48008 [2:35:00<09:26, 5.17it/s] 94%|█████████▍| 45080/48008 [2:35:00<09:23, 5.20it/s] 94%|█████████▍| 45081/48008 [2:35:00<09:22, 5.20it/s] 94%|█████████▍| 45082/48008 [2:35:00<09:23, 5.19it/s] 94%|█████████▍| 45083/48008 [2:35:01<09:25, 5.17it/s] 94%|█████████▍| 45084/48008 [2:35:01<09:29, 5.14it/s] 94%|█████████▍| 45085/48008 [2:35:01<09:25, 5.17it/s] 94%|█████████▍| 45086/48008 [2:35:01<09:24, 5.18it/s] 94%|█████████▍| 45087/48008 [2:35:01<09:23, 5.18it/s] 94%|█████████▍| 45088/48008 [2:35:02<09:25, 5.16it/s] 94%|█████████▍| 45089/48008 [2:35:02<09:18, 5.22it/s] 94%|█████████▍| 45090/48008 [2:35:02<09:16, 5.25it/s] 94%|█████████▍| 45091/48008 [2:35:02<09:20, 5.21it/s] 94%|█████████▍| 45092/48008 [2:35:02<09:20, 5.21it/s] 94%|█████████▍| 45093/48008 [2:35:02<09:20, 5.20it/s] 94%|█████████▍| 45094/48008 [2:35:03<09:21, 5.19it/s] 94%|█████████▍| 45095/48008 [2:35:03<09:19, 5.21it/s] 94%|█████████▍| 45096/48008 [2:35:03<09:23, 5.17it/s] 94%|█████████▍| 45097/48008 [2:35:03<09:18, 5.22it/s] 94%|█████████▍| 45098/48008 [2:35:03<09:19, 5.20it/s] 94%|█████████▍| 45099/48008 [2:35:04<09:20, 5.19it/s] 94%|█████████▍| 45100/48008 [2:35:04<09:18, 5.21it/s] 94%|█████████▍| 45100/48008 [2:35:04<09:18, 5.21it/s]{'loss': 0.3013, 'grad_norm': 0.287109375, 'learning_rate': 1.2118813531078154e-05, 'epoch': 0.94} 94%|█████████▍| 45101/48008 [2:35:04<09:30, 5.10it/s] 94%|█████████▍| 45102/48008 [2:35:04<09:24, 5.14it/s] 94%|█████████▍| 45103/48008 [2:35:04<09:22, 5.16it/s] 94%|█████████▍| 45104/48008 [2:35:05<09:20, 5.18it/s] 94%|█████████▍| 45105/48008 [2:35:05<09:21, 5.17it/s] 94%|█████████▍| 45106/48008 [2:35:05<12:10, 3.97it/s] 94%|█████████▍| 45107/48008 [2:35:05<11:16, 4.29it/s] 94%|█████████▍| 45108/48008 [2:35:06<10:39, 4.54it/s] 94%|█████████▍| 45109/48008 [2:35:06<10:18, 4.69it/s] 94%|█████████▍| 45110/48008 [2:35:06<10:11, 4.74it/s] 94%|█████████▍| 45111/48008 [2:35:06<10:02, 4.81it/s] 94%|█████████▍| 45112/48008 [2:35:06<09:47, 4.93it/s] 94%|█████████▍| 45113/48008 [2:35:07<10:19, 4.68it/s] 94%|█████████▍| 45114/48008 [2:35:07<10:01, 4.81it/s] 94%|█████████▍| 45115/48008 [2:35:07<10:02, 4.80it/s] 94%|█████████▍| 45116/48008 [2:35:07<09:47, 4.92it/s] 94%|█████████▍| 45117/48008 [2:35:07<09:42, 4.96it/s] 94%|█████████▍| 45118/48008 [2:35:08<09:37, 5.01it/s] 94%|█████████▍| 45119/48008 [2:35:08<09:33, 5.04it/s] 94%|█████████▍| 45120/48008 [2:35:08<09:25, 5.11it/s] 94%|█████████▍| 45121/48008 [2:35:08<09:20, 5.16it/s] 94%|█████████▍| 45122/48008 [2:35:08<09:17, 5.17it/s] 94%|█████████▍| 45123/48008 [2:35:09<09:21, 5.14it/s] 94%|█████████▍| 45124/48008 [2:35:09<09:17, 5.17it/s] 94%|█████████▍| 45125/48008 [2:35:09<09:15, 5.19it/s] 94%|█████████▍| 45126/48008 [2:35:09<09:25, 5.09it/s] 94%|█████████▍| 45127/48008 [2:35:09<09:20, 5.14it/s] 94%|█████████▍| 45128/48008 [2:35:10<09:13, 5.21it/s] 94%|█████████▍| 45129/48008 [2:35:10<09:09, 5.24it/s] 94%|█████████▍| 45130/48008 [2:35:10<09:13, 5.20it/s] 94%|█████████▍| 45131/48008 [2:35:10<09:14, 5.19it/s] 94%|█████████▍| 45132/48008 [2:35:10<09:08, 5.24it/s] 94%|█████████▍| 45133/48008 [2:35:10<09:07, 5.26it/s] 94%|█████████▍| 45134/48008 [2:35:11<09:07, 5.25it/s] 94%|█████████▍| 45135/48008 [2:35:11<09:15, 5.17it/s] 94%|█████████▍| 45136/48008 [2:35:11<09:16, 5.16it/s] 94%|█████████▍| 45137/48008 [2:35:11<09:15, 5.17it/s] 94%|█████████▍| 45138/48008 [2:35:11<09:27, 5.06it/s] 94%|█████████▍| 45139/48008 [2:35:12<10:02, 4.76it/s] 94%|█████████▍| 45140/48008 [2:35:12<09:54, 4.82it/s] 94%|█████████▍| 45141/48008 [2:35:12<09:56, 4.81it/s] 94%|█████████▍| 45142/48008 [2:35:12<09:40, 4.94it/s] 94%|█████████▍| 45143/48008 [2:35:12<09:30, 5.02it/s] 94%|█████████▍| 45144/48008 [2:35:13<09:23, 5.08it/s] 94%|█████████▍| 45145/48008 [2:35:13<10:00, 4.77it/s] 94%|█████████▍| 45146/48008 [2:35:13<09:49, 4.86it/s] 94%|█████████▍| 45147/48008 [2:35:13<10:16, 4.64it/s] 94%|█████████▍| 45148/48008 [2:35:14<09:53, 4.82it/s] 94%|█████████▍| 45149/48008 [2:35:14<10:19, 4.62it/s] 94%|█████████▍| 45150/48008 [2:35:14<09:59, 4.77it/s] 94%|█████████▍| 45150/48008 [2:35:14<09:59, 4.77it/s]{'loss': 0.2998, 'grad_norm': 0.189453125, 'learning_rate': 1.191051491418097e-05, 'epoch': 0.94} 94%|█████████▍| 45151/48008 [2:35:14<09:47, 4.86it/s] 94%|█████████▍| 45152/48008 [2:35:14<09:36, 4.95it/s] 94%|█████████▍| 45153/48008 [2:35:15<09:36, 4.95it/s] 94%|█████████▍| 45154/48008 [2:35:15<09:34, 4.97it/s] 94%|█████████▍| 45155/48008 [2:35:15<09:29, 5.01it/s] 94%|█████████▍| 45156/48008 [2:35:15<09:28, 5.02it/s] 94%|█████████▍| 45157/48008 [2:35:15<09:21, 5.08it/s] 94%|█████████▍| 45158/48008 [2:35:16<09:19, 5.10it/s] 94%|█████████▍| 45159/48008 [2:35:16<09:54, 4.80it/s] 94%|█████████▍| 45160/48008 [2:35:16<09:39, 4.92it/s] 94%|█████████▍| 45161/48008 [2:35:16<09:29, 5.00it/s] 94%|█████████▍| 45162/48008 [2:35:16<09:18, 5.09it/s] 94%|█████████▍| 45163/48008 [2:35:17<09:18, 5.10it/s] 94%|█████████▍| 45164/48008 [2:35:17<09:12, 5.15it/s] 94%|█████████▍| 45165/48008 [2:35:17<10:48, 4.38it/s] 94%|█████████▍| 45166/48008 [2:35:17<13:03, 3.63it/s] 94%|█████████▍| 45167/48008 [2:35:18<13:30, 3.50it/s] 94%|█████████▍| 45168/48008 [2:35:18<12:21, 3.83it/s] 94%|█████████▍| 45169/48008 [2:35:18<11:27, 4.13it/s] 94%|█████████▍| 45170/48008 [2:35:18<10:41, 4.42it/s] 94%|█████████▍| 45171/48008 [2:35:19<10:07, 4.67it/s] 94%|█████████▍| 45172/48008 [2:35:19<09:47, 4.83it/s] 94%|█████████▍| 45173/48008 [2:35:19<09:35, 4.93it/s] 94%|█████████▍| 45174/48008 [2:35:19<09:24, 5.02it/s] 94%|█████████▍| 45175/48008 [2:35:19<09:27, 4.99it/s] 94%|█████████▍| 45176/48008 [2:35:19<09:20, 5.05it/s] 94%|█████████▍| 45177/48008 [2:35:20<09:21, 5.04it/s] 94%|█████████▍| 45178/48008 [2:35:20<09:21, 5.04it/s] 94%|█████████▍| 45179/48008 [2:35:20<09:13, 5.11it/s] 94%|█████████▍| 45180/48008 [2:35:20<09:08, 5.16it/s] 94%|█████████▍| 45181/48008 [2:35:20<09:10, 5.14it/s] 94%|█████████▍| 45182/48008 [2:35:21<09:16, 5.08it/s] 94%|█████████▍| 45183/48008 [2:35:21<09:12, 5.12it/s] 94%|█████████▍| 45184/48008 [2:35:21<09:09, 5.14it/s] 94%|█████████▍| 45185/48008 [2:35:21<09:16, 5.07it/s] 94%|█████████▍| 45186/48008 [2:35:22<11:56, 3.94it/s] 94%|█████████▍| 45187/48008 [2:35:22<11:07, 4.22it/s] 94%|█████████▍| 45188/48008 [2:35:22<10:33, 4.45it/s] 94%|█████████▍| 45189/48008 [2:35:22<11:43, 4.01it/s] 94%|█████████▍| 45190/48008 [2:35:23<12:31, 3.75it/s] 94%|█████████▍| 45191/48008 [2:35:23<11:28, 4.09it/s] 94%|█████████▍| 45192/48008 [2:35:23<10:42, 4.38it/s] 94%|█████████▍| 45193/48008 [2:35:23<11:48, 3.97it/s] 94%|█████████▍| 45194/48008 [2:35:24<11:08, 4.21it/s] 94%|█████████▍| 45195/48008 [2:35:24<11:07, 4.21it/s] 94%|█████████▍| 45196/48008 [2:35:24<10:27, 4.48it/s] 94%|█████████▍| 45197/48008 [2:35:24<11:37, 4.03it/s] 94%|█████████▍| 45198/48008 [2:35:24<10:48, 4.33it/s] 94%|█████████▍| 45199/48008 [2:35:25<10:14, 4.57it/s] 94%|█████████▍| 45200/48008 [2:35:25<09:49, 4.76it/s] {'loss': 0.2956, 'grad_norm': 0.2578125, 'learning_rate': 1.1702216297283787e-05, 'epoch': 0.94} 94%|█████████▍| 45200/48008 [2:35:25<09:49, 4.76it/s] 94%|█████████▍| 45201/48008 [2:35:25<09:32, 4.90it/s] 94%|█████████▍| 45202/48008 [2:35:25<09:20, 5.01it/s] 94%|█████████▍| 45203/48008 [2:35:25<09:12, 5.08it/s] 94%|█████████▍| 45204/48008 [2:35:26<09:07, 5.12it/s] 94%|█████████▍| 45205/48008 [2:35:26<10:40, 4.38it/s] 94%|█████████▍| 45206/48008 [2:35:26<10:12, 4.57it/s] 94%|█████████▍| 45207/48008 [2:35:26<09:52, 4.73it/s] 94%|█████████▍| 45208/48008 [2:35:26<09:41, 4.82it/s] 94%|█████████▍| 45209/48008 [2:35:27<09:22, 4.98it/s] 94%|█████████▍| 45210/48008 [2:35:27<11:56, 3.90it/s] 94%|█████████▍| 45211/48008 [2:35:27<10:59, 4.24it/s] 94%|█████████▍| 45212/48008 [2:35:27<10:26, 4.46it/s] 94%|█████████▍| 45213/48008 [2:35:28<09:59, 4.66it/s] 94%|█████████▍| 45214/48008 [2:35:28<11:16, 4.13it/s] 94%|█████████▍| 45215/48008 [2:35:28<10:47, 4.31it/s] 94%|█████████▍| 45216/48008 [2:35:28<10:14, 4.54it/s] 94%|█████████▍| 45217/48008 [2:35:29<09:49, 4.73it/s] 94%|█████████▍| 45218/48008 [2:35:29<09:42, 4.79it/s] 94%|█████████▍| 45219/48008 [2:35:29<09:24, 4.94it/s] 94%|█████████▍| 45220/48008 [2:35:29<09:12, 5.05it/s] 94%|█████████▍| 45221/48008 [2:35:29<09:06, 5.10it/s] 94%|█████████▍| 45222/48008 [2:35:30<09:12, 5.04it/s] 94%|█████████▍| 45223/48008 [2:35:30<09:04, 5.12it/s] 94%|█████████▍| 45224/48008 [2:35:30<09:01, 5.14it/s] 94%|█████████▍| 45225/48008 [2:35:30<09:08, 5.08it/s] 94%|█████████▍| 45226/48008 [2:35:30<09:01, 5.13it/s] 94%|█████████▍| 45227/48008 [2:35:30<08:57, 5.17it/s] 94%|█████████▍| 45228/48008 [2:35:31<08:55, 5.20it/s] 94%|█████████▍| 45229/48008 [2:35:31<08:55, 5.19it/s] 94%|█████████▍| 45230/48008 [2:35:31<08:54, 5.20it/s] 94%|█████████▍| 45231/48008 [2:35:31<08:58, 5.15it/s] 94%|█████████▍| 45232/48008 [2:35:31<09:06, 5.08it/s] 94%|█████████▍| 45233/48008 [2:35:32<09:19, 4.96it/s] 94%|█████████▍| 45234/48008 [2:35:32<09:22, 4.93it/s] 94%|█████████▍| 45235/48008 [2:35:32<09:21, 4.94it/s] 94%|█████████▍| 45236/48008 [2:35:32<09:08, 5.05it/s] 94%|█████████▍| 45237/48008 [2:35:32<09:09, 5.05it/s] 94%|█████████▍| 45238/48008 [2:35:33<09:17, 4.97it/s] 94%|█████████▍| 45239/48008 [2:35:33<09:14, 4.99it/s] 94%|█████████▍| 45240/48008 [2:35:33<09:01, 5.11it/s] 94%|█████████▍| 45241/48008 [2:35:33<08:57, 5.15it/s] 94%|█████████▍| 45242/48008 [2:35:33<08:57, 5.14it/s] 94%|█████████▍| 45243/48008 [2:35:34<09:02, 5.10it/s] 94%|█████████▍| 45244/48008 [2:35:34<09:03, 5.08it/s] 94%|█████████▍| 45245/48008 [2:35:34<08:58, 5.13it/s] 94%|█████████▍| 45246/48008 [2:35:34<08:54, 5.17it/s] 94%|█████████▍| 45247/48008 [2:35:34<08:52, 5.18it/s] 94%|█████████▍| 45248/48008 [2:35:35<08:50, 5.20it/s] 94%|█████████▍| 45249/48008 [2:35:35<09:01, 5.10it/s] 94%|█████████▍| 45250/48008 [2:35:35<09:06, 5.05it/s] {'loss': 0.2981, 'grad_norm': 0.1533203125, 'learning_rate': 1.1493917680386603e-05, 'epoch': 0.94} 94%|█████████▍| 45250/48008 [2:35:35<09:06, 5.05it/s] 94%|█████████▍| 45251/48008 [2:35:35<09:02, 5.08it/s] 94%|█████████▍| 45252/48008 [2:35:35<08:56, 5.14it/s] 94%|█████████▍| 45253/48008 [2:35:36<08:56, 5.13it/s] 94%|█████████▍| 45254/48008 [2:35:36<08:47, 5.22it/s] 94%|█████████▍| 45255/48008 [2:35:36<08:46, 5.23it/s] 94%|█████████▍| 45256/48008 [2:35:36<08:48, 5.21it/s] 94%|█████████▍| 45257/48008 [2:35:36<08:51, 5.18it/s] 94%|█████████▍| 45258/48008 [2:35:37<08:45, 5.23it/s] 94%|█████████▍| 45259/48008 [2:35:37<08:43, 5.25it/s] 94%|█████████▍| 45260/48008 [2:35:37<08:41, 5.27it/s] 94%|█████████▍| 45261/48008 [2:35:37<08:52, 5.16it/s] 94%|█████████▍| 45262/48008 [2:35:37<08:47, 5.21it/s] 94%|█████████▍| 45263/48008 [2:35:37<08:48, 5.19it/s] 94%|█████████▍| 45264/48008 [2:35:38<08:53, 5.15it/s] 94%|█████████▍| 45265/48008 [2:35:38<08:49, 5.18it/s] 94%|█████████▍| 45266/48008 [2:35:38<09:04, 5.03it/s] 94%|█████████▍| 45267/48008 [2:35:38<08:54, 5.13it/s] 94%|█████████▍| 45268/48008 [2:35:38<08:51, 5.15it/s] 94%|█████████▍| 45269/48008 [2:35:39<09:27, 4.83it/s] 94%|█████████▍| 45270/48008 [2:35:39<09:13, 4.95it/s] 94%|█████████▍| 45271/48008 [2:35:39<09:03, 5.03it/s] 94%|█████████▍| 45272/48008 [2:35:39<08:57, 5.09it/s] 94%|█████████▍| 45273/48008 [2:35:39<08:49, 5.16it/s] 94%|█████████▍| 45274/48008 [2:35:40<10:23, 4.39it/s] 94%|█████████▍| 45275/48008 [2:35:40<09:56, 4.58it/s] 94%|█████████▍| 45276/48008 [2:35:40<11:08, 4.09it/s] 94%|█████████▍| 45277/48008 [2:35:40<10:26, 4.36it/s] 94%|█████████▍| 45278/48008 [2:35:41<10:04, 4.51it/s] 94%|█████████▍| 45279/48008 [2:35:41<09:41, 4.69it/s] 94%|█████████▍| 45280/48008 [2:35:41<09:31, 4.77it/s] 94%|█████████▍| 45281/48008 [2:35:41<09:15, 4.91it/s] 94%|█████████▍| 45282/48008 [2:35:41<09:06, 4.99it/s] 94%|█████████▍| 45283/48008 [2:35:42<09:00, 5.04it/s] 94%|█████████▍| 45284/48008 [2:35:42<08:57, 5.07it/s] 94%|█████████▍| 45285/48008 [2:35:42<08:57, 5.07it/s] 94%|█████████▍| 45286/48008 [2:35:42<08:55, 5.08it/s] 94%|█████████▍| 45287/48008 [2:35:42<08:49, 5.14it/s] 94%|█████████▍| 45288/48008 [2:35:43<08:50, 5.13it/s] 94%|█████████▍| 45289/48008 [2:35:43<08:46, 5.17it/s] 94%|█████████▍| 45290/48008 [2:35:43<08:42, 5.20it/s] 94%|█████████▍| 45291/48008 [2:35:43<08:49, 5.13it/s] 94%|█████████▍| 45292/48008 [2:35:43<08:49, 5.13it/s] 94%|█████████▍| 45293/48008 [2:35:44<08:51, 5.11it/s] 94%|█████████▍| 45294/48008 [2:35:44<08:44, 5.18it/s] 94%|█████████▍| 45295/48008 [2:35:44<08:52, 5.10it/s] 94%|█████████▍| 45296/48008 [2:35:44<08:41, 5.20it/s] 94%|█████████▍| 45297/48008 [2:35:44<08:50, 5.11it/s] 94%|█████████▍| 45298/48008 [2:35:45<08:54, 5.07it/s] 94%|█████████▍| 45299/48008 [2:35:45<08:44, 5.17it/s] 94%|█████████▍| 45300/48008 [2:35:45<08:53, 5.07it/s] {'loss': 0.3003, 'grad_norm': 0.341796875, 'learning_rate': 1.128561906348942e-05, 'epoch': 0.94} 94%|█████████▍| 45300/48008 [2:35:45<08:53, 5.07it/s] 94%|█████████▍| 45301/48008 [2:35:45<08:53, 5.07it/s] 94%|█████████▍| 45302/48008 [2:35:45<08:49, 5.11it/s] 94%|█████████▍| 45303/48008 [2:35:46<08:47, 5.13it/s] 94%|█████████▍| 45304/48008 [2:35:46<08:50, 5.10it/s] 94%|█████████▍| 45305/48008 [2:35:46<08:42, 5.17it/s] 94%|█████████▍| 45306/48008 [2:35:46<08:46, 5.14it/s] 94%|█████████▍| 45307/48008 [2:35:46<08:37, 5.21it/s] 94%|█████████▍| 45308/48008 [2:35:47<08:42, 5.17it/s] 94%|█████████▍| 45309/48008 [2:35:47<08:40, 5.19it/s] 94%|█████████▍| 45310/48008 [2:35:47<08:35, 5.24it/s] 94%|█████████▍| 45311/48008 [2:35:47<08:30, 5.28it/s] 94%|█████████▍| 45312/48008 [2:35:47<08:35, 5.23it/s] 94%|█████████▍| 45313/48008 [2:35:47<09:13, 4.87it/s] 94%|█████████▍| 45314/48008 [2:35:48<09:40, 4.64it/s] 94%|█████████▍| 45315/48008 [2:35:48<09:25, 4.76it/s] 94%|█████████▍| 45316/48008 [2:35:48<09:25, 4.76it/s] 94%|█████████▍| 45317/48008 [2:35:48<09:20, 4.80it/s] 94%|█████████▍| 45318/48008 [2:35:49<09:07, 4.91it/s] 94%|█████████▍| 45319/48008 [2:35:49<08:56, 5.01it/s] 94%|█████████▍| 45320/48008 [2:35:49<08:49, 5.08it/s] 94%|█████████▍| 45321/48008 [2:35:49<08:50, 5.06it/s] 94%|█████████▍| 45322/48008 [2:35:49<08:51, 5.05it/s] 94%|█████████▍| 45323/48008 [2:35:50<08:46, 5.10it/s] 94%|█████████▍| 45324/48008 [2:35:50<08:47, 5.08it/s] 94%|█████████▍| 45325/48008 [2:35:50<10:16, 4.35it/s] 94%|█████████▍| 45326/48008 [2:35:50<12:22, 3.61it/s] 94%|█████████▍| 45327/48008 [2:35:51<11:22, 3.93it/s] 94%|█████████▍| 45328/48008 [2:35:51<10:29, 4.26it/s] 94%|█████████▍| 45329/48008 [2:35:52<19:16, 2.32it/s] 94%|█████████▍| 45330/48008 [2:35:52<16:03, 2.78it/s] 94%|█████████▍| 45331/48008 [2:35:52<13:50, 3.22it/s] 94%|█████████▍| 45332/48008 [2:35:52<12:25, 3.59it/s] 94%|█████████▍| 45333/48008 [2:35:52<11:14, 3.97it/s] 94%|█████████▍| 45334/48008 [2:35:53<10:21, 4.30it/s] 94%|█████████▍| 45335/48008 [2:35:53<09:53, 4.50it/s] 94%|█████████▍| 45336/48008 [2:35:53<09:28, 4.70it/s] 94%|█████████▍| 45337/48008 [2:35:53<09:08, 4.87it/s] 94%|█████████▍| 45338/48008 [2:35:53<09:05, 4.89it/s] 94%|█████████▍| 45339/48008 [2:35:54<08:55, 4.99it/s] 94%|█████████▍| 45340/48008 [2:35:54<08:53, 5.00it/s] 94%|█████████▍| 45341/48008 [2:35:54<08:42, 5.10it/s] 94%|█████████▍| 45342/48008 [2:35:54<08:36, 5.16it/s] 94%|█████████▍| 45343/48008 [2:35:54<08:34, 5.18it/s] 94%|█████████▍| 45344/48008 [2:35:55<08:40, 5.12it/s] 94%|█████████▍| 45345/48008 [2:35:55<08:50, 5.02it/s] 94%|█████████▍| 45346/48008 [2:35:55<08:53, 4.99it/s] 94%|█████████▍| 45347/48008 [2:35:55<11:22, 3.90it/s] 94%|█████████▍| 45348/48008 [2:35:56<10:29, 4.22it/s] 94%|█████████▍| 45349/48008 [2:35:56<09:48, 4.52it/s] 94%|█████████▍| 45350/48008 [2:35:56<09:24, 4.71it/s] 94%|█████████▍| 45350/48008 [2:35:56<09:24, 4.71it/s]{'loss': 0.2983, 'grad_norm': 0.177734375, 'learning_rate': 1.1077320446592234e-05, 'epoch': 0.94} 94%|█████████▍| 45351/48008 [2:35:56<09:10, 4.82it/s] 94%|█████████▍| 45352/48008 [2:35:56<08:59, 4.92it/s] 94%|█████████▍| 45353/48008 [2:35:57<08:52, 4.99it/s] 94%|█████████▍| 45354/48008 [2:35:57<08:46, 5.04it/s] 94%|█████████▍| 45355/48008 [2:35:57<08:40, 5.10it/s] 94%|█████████▍| 45356/48008 [2:35:57<08:40, 5.09it/s] 94%|█████████▍| 45357/48008 [2:35:57<08:41, 5.08it/s] 94%|█████████▍| 45358/48008 [2:35:58<08:35, 5.14it/s] 94%|█████████▍| 45359/48008 [2:35:58<08:32, 5.17it/s] 94%|█████████▍| 45360/48008 [2:35:58<08:38, 5.10it/s] 94%|█████████▍| 45361/48008 [2:35:58<08:31, 5.17it/s] 94%|█████████▍| 45362/48008 [2:35:58<08:30, 5.18it/s] 94%|█████████▍| 45363/48008 [2:35:58<08:39, 5.09it/s] 94%|█████████▍| 45364/48008 [2:35:59<08:38, 5.10it/s] 94%|█████████▍| 45365/48008 [2:35:59<08:36, 5.11it/s] 94%|█████████▍| 45366/48008 [2:35:59<08:36, 5.11it/s] 94%|█████████▍| 45367/48008 [2:35:59<08:46, 5.02it/s] 95%|█████████▍| 45368/48008 [2:35:59<08:42, 5.06it/s] 95%|█████████▍| 45369/48008 [2:36:00<08:38, 5.09it/s] 95%|█████████▍| 45370/48008 [2:36:00<08:34, 5.12it/s] 95%|█████████▍| 45371/48008 [2:36:00<08:41, 5.06it/s] 95%|█████████▍| 45372/48008 [2:36:00<08:37, 5.10it/s] 95%|█████████▍| 45373/48008 [2:36:00<08:38, 5.08it/s] 95%|█████████▍| 45374/48008 [2:36:01<08:53, 4.94it/s] 95%|█████████▍| 45375/48008 [2:36:01<08:40, 5.06it/s] 95%|█████████▍| 45376/48008 [2:36:01<08:45, 5.01it/s] 95%|█████████▍| 45377/48008 [2:36:01<10:08, 4.32it/s] 95%|█████████▍| 45378/48008 [2:36:02<12:11, 3.59it/s] 95%|█████████▍| 45379/48008 [2:36:02<11:07, 3.94it/s] 95%|█████████▍| 45380/48008 [2:36:02<10:16, 4.26it/s] 95%|█████████▍| 45381/48008 [2:36:02<09:43, 4.50it/s] 95%|█████████▍| 45382/48008 [2:36:03<09:15, 4.73it/s] 95%|█████████▍| 45383/48008 [2:36:03<09:12, 4.75it/s] 95%|█████████▍| 45384/48008 [2:36:03<08:56, 4.89it/s] 95%|█████████▍| 45385/48008 [2:36:03<08:45, 4.99it/s] 95%|█████████▍| 45386/48008 [2:36:03<08:39, 5.05it/s] 95%|█████████▍| 45387/48008 [2:36:04<08:42, 5.02it/s] 95%|█████████▍| 45388/48008 [2:36:04<08:34, 5.09it/s] 95%|█████████▍| 45389/48008 [2:36:04<08:27, 5.16it/s] 95%|█████████▍| 45390/48008 [2:36:04<08:21, 5.22it/s] 95%|█████████▍| 45391/48008 [2:36:04<08:23, 5.20it/s] 95%|█████████▍| 45392/48008 [2:36:04<08:22, 5.21it/s] 95%|█████████▍| 45393/48008 [2:36:05<08:25, 5.17it/s] 95%|█████████▍| 45394/48008 [2:36:05<09:53, 4.40it/s] 95%|█████████▍| 45395/48008 [2:36:05<09:25, 4.62it/s] 95%|█████████▍| 45396/48008 [2:36:05<09:06, 4.78it/s] 95%|█████████▍| 45397/48008 [2:36:06<08:51, 4.92it/s] 95%|█████████▍| 45398/48008 [2:36:06<08:46, 4.96it/s] 95%|█████████▍| 45399/48008 [2:36:06<08:37, 5.04it/s] 95%|█████████▍| 45400/48008 [2:36:06<08:39, 5.02it/s] {'loss': 0.2981, 'grad_norm': 0.267578125, 'learning_rate': 1.0869021829695051e-05, 'epoch': 0.95} 95%|█████████▍| 45400/48008 [2:36:06<08:39, 5.02it/s] 95%|█████████▍| 45401/48008 [2:36:06<08:49, 4.92it/s] 95%|█████████▍| 45402/48008 [2:36:07<08:34, 5.06it/s] 95%|█████████▍| 45403/48008 [2:36:07<08:29, 5.11it/s] 95%|█████████▍| 45404/48008 [2:36:07<10:59, 3.95it/s] 95%|█████████▍| 45405/48008 [2:36:07<10:17, 4.22it/s] 95%|█████████▍| 45406/48008 [2:36:07<09:42, 4.47it/s] 95%|█████████▍| 45407/48008 [2:36:08<09:16, 4.67it/s] 95%|█████████▍| 45408/48008 [2:36:08<08:59, 4.82it/s] 95%|█████████▍| 45409/48008 [2:36:08<08:56, 4.85it/s] 95%|█████████▍| 45410/48008 [2:36:08<08:41, 4.98it/s] 95%|█████████▍| 45411/48008 [2:36:08<08:32, 5.06it/s] 95%|█████████▍| 45412/48008 [2:36:09<08:28, 5.10it/s] 95%|█████████▍| 45413/48008 [2:36:09<08:26, 5.12it/s] 95%|█████████▍| 45414/48008 [2:36:09<08:20, 5.18it/s] 95%|█████████▍| 45415/48008 [2:36:09<08:18, 5.20it/s] 95%|█████████▍| 45416/48008 [2:36:09<08:18, 5.20it/s] 95%|█████████▍| 45417/48008 [2:36:10<08:25, 5.13it/s] 95%|█████████▍| 45418/48008 [2:36:10<08:59, 4.80it/s] 95%|█████████▍| 45419/48008 [2:36:10<08:46, 4.92it/s] 95%|█████████▍| 45420/48008 [2:36:10<09:12, 4.68it/s] 95%|█████████▍| 45421/48008 [2:36:10<09:03, 4.76it/s] 95%|█████████▍| 45422/48008 [2:36:11<08:54, 4.84it/s] 95%|█████████▍| 45423/48008 [2:36:11<08:41, 4.95it/s] 95%|█████████▍| 45424/48008 [2:36:11<08:41, 4.95it/s] 95%|█████████▍| 45425/48008 [2:36:11<08:32, 5.04it/s] 95%|█████████▍| 45426/48008 [2:36:11<08:36, 5.00it/s] 95%|█████████▍| 45427/48008 [2:36:12<08:27, 5.08it/s] 95%|█████████▍| 45428/48008 [2:36:12<08:31, 5.04it/s] 95%|█████████▍| 45429/48008 [2:36:12<08:23, 5.13it/s] 95%|█████████▍| 45430/48008 [2:36:12<08:18, 5.17it/s] 95%|█████████▍| 45431/48008 [2:36:12<08:17, 5.18it/s] 95%|█████████▍| 45432/48008 [2:36:13<08:11, 5.24it/s] 95%|█████████▍| 45433/48008 [2:36:13<08:12, 5.23it/s] 95%|█████████▍| 45434/48008 [2:36:13<08:13, 5.22it/s] 95%|█████████▍| 45435/48008 [2:36:13<08:08, 5.26it/s] 95%|█████████▍| 45436/48008 [2:36:13<08:10, 5.25it/s] 95%|█████████▍| 45437/48008 [2:36:14<08:07, 5.28it/s] 95%|█████████▍| 45438/48008 [2:36:14<08:05, 5.30it/s] 95%|█████████▍| 45439/48008 [2:36:14<08:21, 5.12it/s] 95%|█████████▍| 45440/48008 [2:36:14<09:46, 4.38it/s] 95%|█████████▍| 45441/48008 [2:36:14<09:23, 4.56it/s] 95%|█████████▍| 45442/48008 [2:36:15<09:00, 4.75it/s] 95%|█████████▍| 45443/48008 [2:36:15<08:46, 4.87it/s] 95%|█████████▍| 45444/48008 [2:36:15<08:33, 5.00it/s] 95%|█████████▍| 45445/48008 [2:36:15<08:36, 4.96it/s] 95%|█████████▍| 45446/48008 [2:36:15<08:32, 5.00it/s] 95%|█████████▍| 45447/48008 [2:36:16<10:53, 3.92it/s] 95%|█████████▍| 45448/48008 [2:36:16<10:41, 3.99it/s] 95%|█████████▍| 45449/48008 [2:36:16<09:52, 4.32it/s] 95%|█████████▍| 45450/48008 [2:36:16<09:17, 4.59it/s] {'loss': 0.3005, 'grad_norm': 0.1787109375, 'learning_rate': 1.0660723212797867e-05, 'epoch': 0.95} 95%|█████████▍| 45450/48008 [2:36:16<09:17, 4.59it/s] 95%|█████████▍| 45451/48008 [2:36:17<08:58, 4.74it/s] 95%|█████████▍| 45452/48008 [2:36:17<08:44, 4.87it/s] 95%|█████████▍| 45453/48008 [2:36:17<08:30, 5.01it/s] 95%|█████████▍| 45454/48008 [2:36:17<08:19, 5.11it/s] 95%|█████████▍| 45455/48008 [2:36:17<08:11, 5.19it/s] 95%|█████████▍| 45456/48008 [2:36:18<08:09, 5.21it/s] 95%|█████████▍| 45457/48008 [2:36:18<08:20, 5.10it/s] 95%|█████████▍| 45458/48008 [2:36:18<10:44, 3.96it/s] 95%|█████████▍| 45459/48008 [2:36:18<10:32, 4.03it/s] 95%|█████████▍| 45460/48008 [2:36:19<10:24, 4.08it/s] 95%|█████████▍| 45461/48008 [2:36:19<09:43, 4.36it/s] 95%|█████████▍| 45462/48008 [2:36:19<09:23, 4.52it/s] 95%|█████████▍| 45463/48008 [2:36:19<09:08, 4.64it/s] 95%|█████████▍| 45464/48008 [2:36:19<08:46, 4.83it/s] 95%|█████████▍| 45465/48008 [2:36:20<08:34, 4.94it/s] 95%|█████████▍| 45466/48008 [2:36:20<08:25, 5.03it/s] 95%|█████████▍| 45467/48008 [2:36:20<08:18, 5.10it/s] 95%|█████████▍| 45468/48008 [2:36:20<08:50, 4.78it/s] 95%|█████████▍| 45469/48008 [2:36:20<08:37, 4.90it/s] 95%|█████████▍| 45470/48008 [2:36:21<08:28, 4.99it/s] 95%|█████████▍| 45471/48008 [2:36:21<08:57, 4.72it/s] 95%|█████████▍| 45472/48008 [2:36:21<08:51, 4.77it/s] 95%|█████████▍| 45473/48008 [2:36:21<08:39, 4.88it/s] 95%|█████████▍| 45474/48008 [2:36:21<08:23, 5.04it/s] 95%|█████████▍| 45475/48008 [2:36:22<08:52, 4.75it/s] 95%|█████████▍| 45476/48008 [2:36:22<08:38, 4.88it/s] 95%|█████████▍| 45477/48008 [2:36:22<08:28, 4.98it/s] 95%|█████████▍| 45478/48008 [2:36:22<08:21, 5.05it/s] 95%|█████████▍| 45479/48008 [2:36:22<08:24, 5.02it/s] 95%|█████████▍| 45480/48008 [2:36:23<08:16, 5.09it/s] 95%|█████████▍| 45481/48008 [2:36:23<08:16, 5.09it/s] 95%|█████████▍| 45482/48008 [2:36:23<08:12, 5.13it/s] 95%|█████████▍| 45483/48008 [2:36:23<08:05, 5.21it/s] 95%|█████████▍| 45484/48008 [2:36:23<08:05, 5.19it/s] 95%|█████████▍| 45485/48008 [2:36:24<08:05, 5.20it/s] 95%|█████████▍| 45486/48008 [2:36:24<08:07, 5.17it/s] 95%|█████████▍| 45487/48008 [2:36:24<08:04, 5.20it/s] 95%|█████████▍| 45488/48008 [2:36:24<08:00, 5.24it/s] 95%|█████████▍| 45489/48008 [2:36:24<08:09, 5.15it/s] 95%|█████████▍| 45490/48008 [2:36:25<08:16, 5.07it/s] 95%|█████████▍| 45491/48008 [2:36:25<08:05, 5.19it/s] 95%|█████████▍| 45492/48008 [2:36:25<08:04, 5.19it/s] 95%|█████████▍| 45493/48008 [2:36:25<08:02, 5.22it/s] 95%|█████████▍| 45494/48008 [2:36:25<08:11, 5.12it/s] 95%|█████████▍| 45495/48008 [2:36:26<08:42, 4.81it/s] 95%|█████████▍| 45496/48008 [2:36:26<08:30, 4.92it/s] 95%|█████████▍| 45497/48008 [2:36:26<08:22, 5.00it/s] 95%|█████████▍| 45498/48008 [2:36:26<08:16, 5.05it/s] 95%|█████████▍| 45499/48008 [2:36:26<08:12, 5.09it/s] 95%|█████████▍| 45500/48008 [2:36:27<08:09, 5.13it/s] {'loss': 0.3012, 'grad_norm': 0.2431640625, 'learning_rate': 1.0452424595900684e-05, 'epoch': 0.95} 95%|█████████▍| 45500/48008 [2:36:27<08:09, 5.13it/s] 95%|█████████▍| 45501/48008 [2:36:27<08:06, 5.16it/s] 95%|█████████▍| 45502/48008 [2:36:27<08:04, 5.17it/s] 95%|█████████▍| 45503/48008 [2:36:27<07:59, 5.22it/s] 95%|█████████▍| 45504/48008 [2:36:27<07:58, 5.24it/s] 95%|█████████▍| 45505/48008 [2:36:27<07:59, 5.22it/s] 95%|█████████▍| 45506/48008 [2:36:28<08:01, 5.19it/s] 95%|█████████▍| 45507/48008 [2:36:28<08:07, 5.13it/s] 95%|█████████▍| 45508/48008 [2:36:28<08:05, 5.14it/s] 95%|█████████▍| 45509/48008 [2:36:28<08:05, 5.15it/s] 95%|█████████▍| 45510/48008 [2:36:29<10:27, 3.98it/s] 95%|█████████▍| 45511/48008 [2:36:29<09:55, 4.19it/s] 95%|█████████▍| 45512/48008 [2:36:29<09:20, 4.45it/s] 95%|█████████▍| 45513/48008 [2:36:29<09:03, 4.59it/s] 95%|█████████▍| 45514/48008 [2:36:29<08:40, 4.79it/s] 95%|█████████▍| 45515/48008 [2:36:30<08:26, 4.93it/s] 95%|█████████▍| 45516/48008 [2:36:30<08:18, 5.00it/s] 95%|█████████▍| 45517/48008 [2:36:30<08:10, 5.08it/s] 95%|█████████▍| 45518/48008 [2:36:30<08:12, 5.05it/s] 95%|█████████▍| 45519/48008 [2:36:30<08:07, 5.11it/s] 95%|█████████▍| 45520/48008 [2:36:31<08:02, 5.16it/s] 95%|█████████▍| 45521/48008 [2:36:31<07:57, 5.21it/s] 95%|█████████▍| 45522/48008 [2:36:31<09:21, 4.43it/s] 95%|█████████▍| 45523/48008 [2:36:31<08:54, 4.65it/s] 95%|█████████▍| 45524/48008 [2:36:31<08:36, 4.81it/s] 95%|█████████▍| 45525/48008 [2:36:32<08:23, 4.94it/s] 95%|█████████▍| 45526/48008 [2:36:32<08:25, 4.91it/s] 95%|█████████▍| 45527/48008 [2:36:32<08:17, 4.99it/s] 95%|█████████▍| 45528/48008 [2:36:32<08:10, 5.05it/s] 95%|█████████▍| 45529/48008 [2:36:32<08:03, 5.13it/s] 95%|█████████▍| 45530/48008 [2:36:33<07:56, 5.20it/s] 95%|█████████▍| 45531/48008 [2:36:33<07:55, 5.21it/s] 95%|█████████▍| 45532/48008 [2:36:33<07:53, 5.23it/s] 95%|█████████▍| 45533/48008 [2:36:33<07:53, 5.23it/s] 95%|█████████▍| 45534/48008 [2:36:33<07:57, 5.18it/s] 95%|█████████▍| 45535/48008 [2:36:34<07:56, 5.19it/s] 95%|█████████▍| 45536/48008 [2:36:34<07:51, 5.24it/s] 95%|█████████▍| 45537/48008 [2:36:34<08:07, 5.07it/s] 95%|█████████▍| 45538/48008 [2:36:34<08:00, 5.14it/s] 95%|█████████▍| 45539/48008 [2:36:34<07:59, 5.15it/s] 95%|█████████▍| 45540/48008 [2:36:35<07:58, 5.16it/s] 95%|█████████▍| 45541/48008 [2:36:35<07:56, 5.18it/s] 95%|█████████▍| 45542/48008 [2:36:35<09:21, 4.39it/s] 95%|█████████▍| 45543/48008 [2:36:35<08:52, 4.63it/s] 95%|█████████▍| 45544/48008 [2:36:35<08:39, 4.74it/s] 95%|█████████▍| 45545/48008 [2:36:36<08:29, 4.83it/s] 95%|█████████▍| 45546/48008 [2:36:36<08:16, 4.96it/s] 95%|█████████▍| 45547/48008 [2:36:36<08:08, 5.04it/s] 95%|█████████▍| 45548/48008 [2:36:36<08:02, 5.10it/s] 95%|█████████▍| 45549/48008 [2:36:36<07:57, 5.15it/s] 95%|█████████▍| 45550/48008 [2:36:37<07:53, 5.19it/s] {'loss': 0.3005, 'grad_norm': 0.32421875, 'learning_rate': 1.02441259790035e-05, 'epoch': 0.95} 95%|█████████▍| 45550/48008 [2:36:37<07:53, 5.19it/s] 95%|█████████▍| 45551/48008 [2:36:37<07:49, 5.23it/s] 95%|█████████▍| 45552/48008 [2:36:37<07:52, 5.20it/s] 95%|█████████▍| 45553/48008 [2:36:37<08:01, 5.10it/s] 95%|█████████▍| 45554/48008 [2:36:37<07:57, 5.14it/s] 95%|█████████▍| 45555/48008 [2:36:38<07:51, 5.20it/s] 95%|█████████▍| 45556/48008 [2:36:38<07:50, 5.21it/s] 95%|█████████▍| 45557/48008 [2:36:38<07:50, 5.21it/s] 95%|█████████▍| 45558/48008 [2:36:38<07:50, 5.20it/s] 95%|█████████▍| 45559/48008 [2:36:38<07:50, 5.20it/s] 95%|█████████▍| 45560/48008 [2:36:39<07:52, 5.18it/s] 95%|█████████▍| 45561/48008 [2:36:39<07:53, 5.16it/s] 95%|█████████▍| 45562/48008 [2:36:39<07:59, 5.10it/s] 95%|█████████▍| 45563/48008 [2:36:39<08:28, 4.81it/s] 95%|█████████▍| 45564/48008 [2:36:39<08:18, 4.90it/s] 95%|█████████▍| 45565/48008 [2:36:40<08:43, 4.67it/s] 95%|█████████▍| 45566/48008 [2:36:40<08:31, 4.77it/s] 95%|█████████▍| 45567/48008 [2:36:40<08:17, 4.90it/s] 95%|█████████▍| 45568/48008 [2:36:40<08:12, 4.95it/s] 95%|█████████▍| 45569/48008 [2:36:40<08:05, 5.02it/s] 95%|█████████▍| 45570/48008 [2:36:41<07:57, 5.11it/s] 95%|█████████▍| 45571/48008 [2:36:41<07:51, 5.16it/s] 95%|█████████▍| 45572/48008 [2:36:41<07:58, 5.09it/s] 95%|█████████▍| 45573/48008 [2:36:41<08:03, 5.04it/s] 95%|█████████▍| 45574/48008 [2:36:41<07:56, 5.11it/s] 95%|█████████▍| 45575/48008 [2:36:42<07:50, 5.17it/s] 95%|█████████▍| 45576/48008 [2:36:42<07:48, 5.19it/s] 95%|█████████▍| 45577/48008 [2:36:42<10:11, 3.98it/s] 95%|█████████▍| 45578/48008 [2:36:42<09:28, 4.28it/s] 95%|█████████▍| 45579/48008 [2:36:43<09:02, 4.48it/s] 95%|█████████▍| 45580/48008 [2:36:43<08:35, 4.71it/s] 95%|█████████▍| 45581/48008 [2:36:43<08:26, 4.79it/s] 95%|█████████▍| 45582/48008 [2:36:43<08:16, 4.88it/s] 95%|█████████▍| 45583/48008 [2:36:43<08:11, 4.93it/s] 95%|█████████▍| 45584/48008 [2:36:43<08:03, 5.01it/s] 95%|█████████▍| 45585/48008 [2:36:44<07:58, 5.07it/s] 95%|█████████▍| 45586/48008 [2:36:44<07:52, 5.13it/s] 95%|█████████▍| 45587/48008 [2:36:44<07:49, 5.15it/s] 95%|█████████▍| 45588/48008 [2:36:44<07:45, 5.20it/s] 95%|█████████▍| 45589/48008 [2:36:44<07:45, 5.19it/s] 95%|█████████▍| 45590/48008 [2:36:45<07:41, 5.24it/s] 95%|█████████▍| 45591/48008 [2:36:45<07:47, 5.17it/s] 95%|█████████▍| 45592/48008 [2:36:45<07:43, 5.21it/s] 95%|█████████▍| 45593/48008 [2:36:45<07:40, 5.24it/s] 95%|█████████▍| 45594/48008 [2:36:45<07:40, 5.24it/s] 95%|█████████▍| 45595/48008 [2:36:46<07:39, 5.25it/s] 95%|█████████▍| 45596/48008 [2:36:46<07:40, 5.23it/s] 95%|█████████▍| 45597/48008 [2:36:46<07:41, 5.23it/s] 95%|█████████▍| 45598/48008 [2:36:46<07:43, 5.20it/s] 95%|█████████▍| 45599/48008 [2:36:46<07:41, 5.22it/s] 95%|█████████▍| 45600/48008 [2:36:47<07:41, 5.22it/s] {'loss': 0.3002, 'grad_norm': 0.3671875, 'learning_rate': 1.0035827362106316e-05, 'epoch': 0.95} 95%|█████████▍| 45600/48008 [2:36:47<07:41, 5.22it/s] 95%|█████████▍| 45601/48008 [2:36:47<07:38, 5.24it/s] 95%|█████████▍| 45602/48008 [2:36:47<07:39, 5.24it/s] 95%|█████████▍| 45603/48008 [2:36:47<07:40, 5.22it/s] 95%|█████████▍| 45604/48008 [2:36:47<07:40, 5.22it/s] 95%|█████████▍| 45605/48008 [2:36:48<10:00, 4.00it/s] 95%|█████████▍| 45606/48008 [2:36:48<09:21, 4.28it/s] 95%|█████████▍| 45607/48008 [2:36:48<09:26, 4.24it/s] 95%|█████████▌| 45608/48008 [2:36:48<08:58, 4.46it/s] 95%|█████████▌| 45609/48008 [2:36:49<08:43, 4.59it/s] 95%|█████████▌| 45610/48008 [2:36:49<08:20, 4.79it/s] 95%|█████████▌| 45611/48008 [2:36:49<08:07, 4.92it/s] 95%|█████████▌| 45612/48008 [2:36:49<08:02, 4.96it/s] 95%|█████████▌| 45613/48008 [2:36:49<07:51, 5.08it/s] 95%|█████████▌| 45614/48008 [2:36:49<08:00, 4.98it/s] 95%|█████████▌| 45615/48008 [2:36:50<07:49, 5.10it/s] 95%|█████████▌| 45616/48008 [2:36:50<07:55, 5.03it/s] 95%|█████████▌| 45617/48008 [2:36:50<07:49, 5.09it/s] 95%|█████████▌| 45618/48008 [2:36:50<07:46, 5.13it/s] 95%|█████████▌| 45619/48008 [2:36:50<07:44, 5.15it/s] 95%|█████████▌| 45620/48008 [2:36:51<07:41, 5.18it/s] 95%|█████████▌| 45621/48008 [2:36:51<07:37, 5.22it/s] 95%|█████████▌| 45622/48008 [2:36:51<07:33, 5.26it/s] 95%|█████████▌| 45623/48008 [2:36:51<07:43, 5.15it/s] 95%|█████████▌| 45624/48008 [2:36:51<07:45, 5.13it/s] 95%|█████████▌| 45625/48008 [2:36:52<07:48, 5.08it/s] 95%|█████████▌| 45626/48008 [2:36:52<10:04, 3.94it/s] 95%|█████████▌| 45627/48008 [2:36:52<09:25, 4.21it/s] 95%|█████████▌| 45628/48008 [2:36:52<08:53, 4.46it/s] 95%|█████████▌| 45629/48008 [2:36:53<08:36, 4.60it/s] 95%|█████████▌| 45630/48008 [2:36:53<08:14, 4.81it/s] 95%|█████████▌| 45631/48008 [2:36:53<08:03, 4.91it/s] 95%|█████████▌| 45632/48008 [2:36:53<07:51, 5.04it/s] 95%|█████████▌| 45633/48008 [2:36:53<07:55, 4.99it/s] 95%|█████████▌| 45634/48008 [2:36:54<07:53, 5.02it/s] 95%|█████████▌| 45635/48008 [2:36:54<07:53, 5.01it/s] 95%|█████████▌| 45636/48008 [2:36:54<07:51, 5.04it/s] 95%|█████████▌| 45637/48008 [2:36:54<07:44, 5.11it/s] 95%|█████████▌| 45638/48008 [2:36:54<07:37, 5.19it/s] 95%|█████████▌| 45639/48008 [2:36:55<07:36, 5.19it/s] 95%|█████████▌| 45640/48008 [2:36:55<07:38, 5.17it/s] 95%|█████████▌| 45641/48008 [2:36:55<07:44, 5.09it/s] 95%|█████████▌| 45642/48008 [2:36:55<07:45, 5.08it/s] 95%|█████████▌| 45643/48008 [2:36:55<07:44, 5.09it/s] 95%|█████████▌| 45644/48008 [2:36:56<07:42, 5.12it/s] 95%|█████████▌| 45645/48008 [2:36:56<09:01, 4.37it/s] 95%|█████████▌| 45646/48008 [2:36:56<08:32, 4.61it/s] 95%|█████████▌| 45647/48008 [2:36:56<08:47, 4.48it/s] 95%|█████████▌| 45648/48008 [2:36:56<08:21, 4.70it/s] 95%|█████████▌| 45649/48008 [2:36:57<08:18, 4.73it/s] 95%|█████████▌| 45650/48008 [2:36:57<08:36, 4.56it/s] {'loss': 0.3024, 'grad_norm': 0.291015625, 'learning_rate': 9.827528745209131e-06, 'epoch': 0.95} 95%|█████████▌| 45650/48008 [2:36:57<08:36, 4.56it/s] 95%|█████████▌| 45651/48008 [2:36:57<08:23, 4.68it/s] 95%|█████████▌| 45652/48008 [2:36:57<08:07, 4.84it/s] 95%|█████████▌| 45653/48008 [2:36:57<07:58, 4.92it/s] 95%|█████████▌| 45654/48008 [2:36:58<07:49, 5.01it/s] 95%|█████████▌| 45655/48008 [2:36:58<07:49, 5.02it/s] 95%|█████████▌| 45656/48008 [2:36:58<07:46, 5.05it/s] 95%|█████████▌| 45657/48008 [2:36:58<07:38, 5.13it/s] 95%|█████████▌| 45658/48008 [2:36:59<08:57, 4.37it/s] 95%|█████████▌| 45659/48008 [2:36:59<08:32, 4.58it/s] 95%|█████████▌| 45660/48008 [2:36:59<08:30, 4.60it/s] 95%|█████████▌| 45661/48008 [2:36:59<08:10, 4.79it/s] 95%|█████████▌| 45662/48008 [2:36:59<08:04, 4.84it/s] 95%|█████████▌| 45663/48008 [2:37:00<10:13, 3.82it/s] 95%|█████████▌| 45664/48008 [2:37:00<09:24, 4.15it/s] 95%|█████████▌| 45665/48008 [2:37:00<08:52, 4.40it/s] 95%|█████████▌| 45666/48008 [2:37:00<08:28, 4.61it/s] 95%|█████████▌| 45667/48008 [2:37:01<08:20, 4.68it/s] 95%|█████████▌| 45668/48008 [2:37:01<08:12, 4.75it/s] 95%|█████████▌| 45669/48008 [2:37:01<07:58, 4.89it/s] 95%|█████████▌| 45670/48008 [2:37:01<07:50, 4.97it/s] 95%|█████████▌| 45671/48008 [2:37:01<07:49, 4.98it/s] 95%|█████████▌| 45672/48008 [2:37:02<07:42, 5.05it/s] 95%|█████████▌| 45673/48008 [2:37:02<07:50, 4.96it/s] 95%|█████████▌| 45674/48008 [2:37:02<07:40, 5.07it/s] 95%|█████████▌| 45675/48008 [2:37:02<08:12, 4.74it/s] 95%|█████████▌| 45676/48008 [2:37:02<07:59, 4.86it/s] 95%|█████████▌| 45677/48008 [2:37:03<07:47, 4.98it/s] 95%|█████████▌| 45678/48008 [2:37:03<07:42, 5.04it/s] 95%|█████████▌| 45679/48008 [2:37:03<07:39, 5.07it/s] 95%|█████████▌| 45680/48008 [2:37:03<07:38, 5.08it/s] 95%|█████████▌| 45681/48008 [2:37:03<07:37, 5.09it/s] 95%|█████████▌| 45682/48008 [2:37:04<07:32, 5.14it/s] 95%|█████████▌| 45683/48008 [2:37:04<07:34, 5.12it/s] 95%|█████████▌| 45684/48008 [2:37:04<07:28, 5.18it/s] 95%|█████████▌| 45685/48008 [2:37:04<07:31, 5.14it/s] 95%|█████████▌| 45686/48008 [2:37:04<09:43, 3.98it/s] 95%|█████████▌| 45687/48008 [2:37:05<09:00, 4.29it/s] 95%|█████████▌| 45688/48008 [2:37:05<08:35, 4.50it/s] 95%|█████████▌| 45689/48008 [2:37:05<08:09, 4.73it/s] 95%|█████████▌| 45690/48008 [2:37:05<07:55, 4.87it/s] 95%|█████████▌| 45691/48008 [2:37:05<07:50, 4.92it/s] 95%|█████████▌| 45692/48008 [2:37:06<07:43, 4.99it/s] 95%|█████████▌| 45693/48008 [2:37:06<07:36, 5.07it/s] 95%|█████████▌| 45694/48008 [2:37:06<07:34, 5.09it/s] 95%|█████████▌| 45695/48008 [2:37:06<07:31, 5.12it/s] 95%|█████████▌| 45696/48008 [2:37:06<07:35, 5.08it/s] 95%|█████████▌| 45697/48008 [2:37:07<07:30, 5.12it/s] 95%|█████████▌| 45698/48008 [2:37:07<07:25, 5.18it/s] 95%|█████████▌| 45699/48008 [2:37:07<07:29, 5.13it/s] 95%|█████████▌| 45700/48008 [2:37:07<07:27, 5.16it/s] {'loss': 0.301, 'grad_norm': 0.1669921875, 'learning_rate': 9.619230128311949e-06, 'epoch': 0.95} 95%|█████████▌| 45700/48008 [2:37:07<07:27, 5.16it/s] 95%|█████████▌| 45701/48008 [2:37:07<07:57, 4.83it/s] 95%|█████████▌| 45702/48008 [2:37:08<07:44, 4.97it/s] 95%|█████████▌| 45703/48008 [2:37:08<07:35, 5.06it/s] 95%|█████████▌| 45704/48008 [2:37:08<08:03, 4.77it/s] 95%|█████████▌| 45705/48008 [2:37:08<07:52, 4.88it/s] 95%|█████████▌| 45706/48008 [2:37:08<07:43, 4.96it/s] 95%|█████████▌| 45707/48008 [2:37:09<07:37, 5.03it/s] 95%|█████████▌| 45708/48008 [2:37:09<07:33, 5.07it/s] 95%|█████████▌| 45709/48008 [2:37:09<07:26, 5.15it/s] 95%|█████████▌| 45710/48008 [2:37:09<07:23, 5.18it/s] 95%|█████████▌| 45711/48008 [2:37:09<07:21, 5.20it/s] 95%|█████████▌| 45712/48008 [2:37:10<07:17, 5.24it/s] 95%|█████████▌| 45713/48008 [2:37:10<07:19, 5.22it/s] 95%|█████████▌| 45714/48008 [2:37:10<07:15, 5.26it/s] 95%|█████████▌| 45715/48008 [2:37:10<07:21, 5.19it/s] 95%|█████████▌| 45716/48008 [2:37:10<07:20, 5.20it/s] 95%|█████████▌| 45717/48008 [2:37:11<07:31, 5.08it/s] 95%|█████████▌| 45718/48008 [2:37:11<07:32, 5.07it/s] 95%|█████████▌| 45719/48008 [2:37:11<07:28, 5.10it/s] 95%|█████████▌| 45720/48008 [2:37:11<07:37, 5.00it/s] 95%|█████████▌| 45721/48008 [2:37:11<07:37, 5.00it/s] 95%|█████████▌| 45722/48008 [2:37:12<07:27, 5.10it/s] 95%|█████████▌| 45723/48008 [2:37:12<07:22, 5.16it/s] 95%|█████████▌| 45724/48008 [2:37:12<07:28, 5.09it/s] 95%|█████████▌| 45725/48008 [2:37:12<07:38, 4.97it/s] 95%|█████████▌| 45726/48008 [2:37:12<07:32, 5.04it/s] 95%|█████████▌| 45727/48008 [2:37:13<07:27, 5.09it/s] 95%|█████████▌| 45728/48008 [2:37:13<07:24, 5.13it/s] 95%|█████████▌| 45729/48008 [2:37:13<07:21, 5.16it/s] 95%|█████████▌| 45730/48008 [2:37:13<07:20, 5.17it/s] 95%|█████████▌| 45731/48008 [2:37:13<07:31, 5.04it/s] 95%|█████████▌| 45732/48008 [2:37:13<07:23, 5.13it/s] 95%|█████████▌| 45733/48008 [2:37:14<08:39, 4.38it/s] 95%|█████████▌| 45734/48008 [2:37:14<08:14, 4.60it/s] 95%|█████████▌| 45735/48008 [2:37:14<07:54, 4.80it/s] 95%|█████████▌| 45736/48008 [2:37:14<07:39, 4.95it/s] 95%|█████████▌| 45737/48008 [2:37:15<07:33, 5.01it/s] 95%|█████████▌| 45738/48008 [2:37:15<07:36, 4.97it/s] 95%|█████████▌| 45739/48008 [2:37:15<07:30, 5.03it/s] 95%|█████████▌| 45740/48008 [2:37:15<07:27, 5.07it/s] 95%|█████████▌| 45741/48008 [2:37:15<07:21, 5.13it/s] 95%|█████████▌| 45742/48008 [2:37:16<07:19, 5.16it/s] 95%|█████████▌| 45743/48008 [2:37:16<07:19, 5.15it/s] 95%|█████████▌| 45744/48008 [2:37:16<07:26, 5.07it/s] 95%|█████████▌| 45745/48008 [2:37:16<07:31, 5.01it/s] 95%|█████████▌| 45746/48008 [2:37:16<07:22, 5.11it/s] 95%|█████████▌| 45747/48008 [2:37:17<07:18, 5.16it/s] 95%|█████████▌| 45748/48008 [2:37:17<07:16, 5.18it/s] 95%|█████████▌| 45749/48008 [2:37:17<07:18, 5.15it/s] 95%|█████████▌| 45750/48008 [2:37:17<07:24, 5.08it/s] {'loss': 0.2985, 'grad_norm': 0.216796875, 'learning_rate': 9.410931511414765e-06, 'epoch': 0.95} 95%|█████████▌| 45750/48008 [2:37:17<07:24, 5.08it/s] 95%|█████████▌| 45751/48008 [2:37:17<07:22, 5.10it/s] 95%|█████████▌| 45752/48008 [2:37:17<07:16, 5.17it/s] 95%|█████████▌| 45753/48008 [2:37:18<07:15, 5.18it/s] 95%|█████████▌| 45754/48008 [2:37:18<07:26, 5.05it/s] 95%|█████████▌| 45755/48008 [2:37:18<07:19, 5.13it/s] 95%|█████████▌| 45756/48008 [2:37:18<07:16, 5.16it/s] 95%|█████████▌| 45757/48008 [2:37:18<07:15, 5.17it/s] 95%|█████████▌| 45758/48008 [2:37:19<07:12, 5.20it/s] 95%|█████████▌| 45759/48008 [2:37:19<07:13, 5.19it/s] 95%|█████████▌| 45760/48008 [2:37:19<07:19, 5.12it/s] 95%|█████████▌| 45761/48008 [2:37:19<07:16, 5.15it/s] 95%|█████████▌| 45762/48008 [2:37:19<07:47, 4.81it/s] 95%|█████████▌| 45763/48008 [2:37:20<07:36, 4.91it/s] 95%|█████████▌| 45764/48008 [2:37:20<07:32, 4.96it/s] 95%|█████████▌| 45765/48008 [2:37:20<07:38, 4.90it/s] 95%|█████████▌| 45766/48008 [2:37:20<07:26, 5.02it/s] 95%|█████████▌| 45767/48008 [2:37:20<07:18, 5.11it/s] 95%|█████████▌| 45768/48008 [2:37:21<07:12, 5.18it/s] 95%|█████████▌| 45769/48008 [2:37:21<07:11, 5.19it/s] 95%|█████████▌| 45770/48008 [2:37:21<07:06, 5.25it/s] 95%|█████████▌| 45771/48008 [2:37:21<07:04, 5.27it/s] 95%|█████████▌| 45772/48008 [2:37:21<07:04, 5.26it/s] 95%|█████████▌| 45773/48008 [2:37:22<07:17, 5.11it/s] 95%|█████████▌| 45774/48008 [2:37:22<07:14, 5.14it/s] 95%|█████████▌| 45775/48008 [2:37:22<07:20, 5.07it/s] 95%|█████████▌| 45776/48008 [2:37:22<07:24, 5.02it/s] 95%|█████████▌| 45777/48008 [2:37:22<07:27, 4.99it/s] 95%|█████████▌| 45778/48008 [2:37:23<07:28, 4.97it/s] 95%|█████████▌| 45779/48008 [2:37:23<07:20, 5.06it/s] 95%|█████████▌| 45780/48008 [2:37:23<07:13, 5.14it/s] 95%|█████████▌| 45781/48008 [2:37:23<07:11, 5.16it/s] 95%|█████████▌| 45782/48008 [2:37:23<07:10, 5.17it/s] 95%|█████████▌| 45783/48008 [2:37:24<07:40, 4.83it/s] 95%|█████████▌| 45784/48008 [2:37:24<07:27, 4.97it/s] 95%|█████████▌| 45785/48008 [2:37:24<07:23, 5.02it/s] 95%|█████████▌| 45786/48008 [2:37:24<07:22, 5.02it/s] 95%|█████████▌| 45787/48008 [2:37:24<07:11, 5.15it/s] 95%|█████████▌| 45788/48008 [2:37:25<07:16, 5.08it/s] 95%|█████████▌| 45789/48008 [2:37:25<07:13, 5.11it/s] 95%|█████████▌| 45790/48008 [2:37:25<07:15, 5.09it/s] 95%|█████████▌| 45791/48008 [2:37:25<07:12, 5.13it/s] 95%|█████████▌| 45792/48008 [2:37:25<07:19, 5.05it/s] 95%|█████████▌| 45793/48008 [2:37:26<07:20, 5.03it/s] 95%|█████████▌| 45794/48008 [2:37:26<07:25, 4.97it/s] 95%|█████████▌| 45795/48008 [2:37:26<07:17, 5.06it/s] 95%|█████████▌| 45796/48008 [2:37:26<07:13, 5.11it/s] 95%|█████████▌| 45797/48008 [2:37:26<07:08, 5.15it/s] 95%|█████████▌| 45798/48008 [2:37:27<07:07, 5.17it/s] 95%|█████████▌| 45799/48008 [2:37:27<07:08, 5.16it/s] 95%|█████████▌| 45800/48008 [2:37:27<07:00, 5.24it/s] {'loss': 0.3015, 'grad_norm': 0.5859375, 'learning_rate': 9.20263289451758e-06, 'epoch': 0.95} 95%|█████████▌| 45800/48008 [2:37:27<07:00, 5.24it/s] 95%|█████████▌| 45801/48008 [2:37:27<07:06, 5.17it/s] 95%|█████████▌| 45802/48008 [2:37:27<07:00, 5.24it/s] 95%|█████████▌| 45803/48008 [2:37:27<06:58, 5.26it/s] 95%|█████████▌| 45804/48008 [2:37:28<06:59, 5.26it/s] 95%|█████████▌| 45805/48008 [2:37:28<06:55, 5.30it/s] 95%|█████████▌| 45806/48008 [2:37:28<06:58, 5.26it/s] 95%|█████████▌| 45807/48008 [2:37:28<06:58, 5.26it/s] 95%|█████████▌| 45808/48008 [2:37:28<07:00, 5.23it/s] 95%|█████████▌| 45809/48008 [2:37:29<07:11, 5.10it/s] 95%|█████████▌| 45810/48008 [2:37:29<07:21, 4.98it/s] 95%|█████████▌| 45811/48008 [2:37:29<07:15, 5.04it/s] 95%|█████████▌| 45812/48008 [2:37:29<08:25, 4.34it/s] 95%|█████████▌| 45813/48008 [2:37:30<08:05, 4.52it/s] 95%|█████████▌| 45814/48008 [2:37:30<07:45, 4.72it/s] 95%|█████████▌| 45815/48008 [2:37:30<07:32, 4.85it/s] 95%|█████████▌| 45816/48008 [2:37:30<07:22, 4.95it/s] 95%|█████████▌| 45817/48008 [2:37:30<07:14, 5.04it/s] 95%|█████████▌| 45818/48008 [2:37:31<07:09, 5.10it/s] 95%|█████████▌| 45819/48008 [2:37:31<07:14, 5.03it/s] 95%|█████████▌| 45820/48008 [2:37:31<07:21, 4.96it/s] 95%|█████████▌| 45821/48008 [2:37:31<07:27, 4.89it/s] 95%|█████████▌| 45822/48008 [2:37:31<07:19, 4.98it/s] 95%|█████████▌| 45823/48008 [2:37:32<08:30, 4.28it/s] 95%|█████████▌| 45824/48008 [2:37:32<07:56, 4.58it/s] 95%|█████████▌| 45825/48008 [2:37:32<07:50, 4.64it/s] 95%|█████████▌| 45826/48008 [2:37:32<07:35, 4.79it/s] 95%|█████████▌| 45827/48008 [2:37:32<07:36, 4.78it/s] 95%|█████████▌| 45828/48008 [2:37:33<07:25, 4.89it/s] 95%|█████████▌| 45829/48008 [2:37:33<07:17, 4.99it/s] 95%|█████████▌| 45830/48008 [2:37:33<07:23, 4.91it/s] 95%|█████████▌| 45831/48008 [2:37:33<07:17, 4.98it/s] 95%|█████████▌| 45832/48008 [2:37:33<07:11, 5.05it/s] 95%|█████████▌| 45833/48008 [2:37:34<07:07, 5.08it/s] 95%|█████████▌| 45834/48008 [2:37:34<07:03, 5.14it/s] 95%|█████████▌| 45835/48008 [2:37:34<07:30, 4.82it/s] 95%|█████████▌| 45836/48008 [2:37:34<07:30, 4.83it/s] 95%|█████████▌| 45837/48008 [2:37:34<07:18, 4.95it/s] 95%|█████████▌| 45838/48008 [2:37:35<07:18, 4.94it/s] 95%|█████████▌| 45839/48008 [2:37:35<07:41, 4.70it/s] 95%|█████████▌| 45840/48008 [2:37:35<07:26, 4.85it/s] 95%|█████████▌| 45841/48008 [2:37:35<07:24, 4.87it/s] 95%|█████████▌| 45842/48008 [2:37:35<07:15, 4.98it/s] 95%|█████████▌| 45843/48008 [2:37:36<07:14, 4.98it/s] 95%|█████████▌| 45844/48008 [2:37:37<17:53, 2.02it/s] 95%|█████████▌| 45845/48008 [2:37:37<14:39, 2.46it/s] 95%|█████████▌| 45846/48008 [2:37:37<12:21, 2.92it/s] 95%|█████████▌| 45847/48008 [2:37:37<10:42, 3.36it/s] 96%|█████████▌| 45848/48008 [2:37:38<09:37, 3.74it/s] 96%|█████████▌| 45849/48008 [2:37:38<08:44, 4.12it/s] 96%|█████████▌| 45850/48008 [2:37:38<10:16, 3.50it/s] {'loss': 0.298, 'grad_norm': 0.56640625, 'learning_rate': 8.994334277620398e-06, 'epoch': 0.96} 96%|█████████▌| 45850/48008 [2:37:38<10:16, 3.50it/s] 96%|█████████▌| 45851/48008 [2:37:38<09:17, 3.87it/s] 96%|█████████▌| 45852/48008 [2:37:39<08:30, 4.22it/s] 96%|█████████▌| 45853/48008 [2:37:39<08:02, 4.46it/s] 96%|█████████▌| 45854/48008 [2:37:39<07:40, 4.68it/s] 96%|█████████▌| 45855/48008 [2:37:39<07:24, 4.84it/s] 96%|█████████▌| 45856/48008 [2:37:39<07:11, 4.99it/s] 96%|█████████▌| 45857/48008 [2:37:40<07:02, 5.09it/s] 96%|█████████▌| 45858/48008 [2:37:40<06:59, 5.13it/s] 96%|█████████▌| 45859/48008 [2:37:40<07:28, 4.79it/s] 96%|█████████▌| 45860/48008 [2:37:40<07:21, 4.87it/s] 96%|█████████▌| 45861/48008 [2:37:40<07:12, 4.96it/s] 96%|█████████▌| 45862/48008 [2:37:41<07:07, 5.02it/s] 96%|█████████▌| 45863/48008 [2:37:41<08:16, 4.32it/s] 96%|█████████▌| 45864/48008 [2:37:41<07:46, 4.59it/s] 96%|█████████▌| 45865/48008 [2:37:41<07:29, 4.77it/s] 96%|█████████▌| 45866/48008 [2:37:41<07:15, 4.92it/s] 96%|█████████▌| 45867/48008 [2:37:42<08:22, 4.26it/s] 96%|█████████▌| 45868/48008 [2:37:42<07:53, 4.52it/s] 96%|█████████▌| 45869/48008 [2:37:42<07:40, 4.64it/s] 96%|█████████▌| 45870/48008 [2:37:42<07:28, 4.76it/s] 96%|█████████▌| 45871/48008 [2:37:42<07:13, 4.93it/s] 96%|█████████▌| 45872/48008 [2:37:43<07:10, 4.96it/s] 96%|█████████▌| 45873/48008 [2:37:43<07:01, 5.06it/s] 96%|█████████▌| 45874/48008 [2:37:43<07:27, 4.77it/s] 96%|█████████▌| 45875/48008 [2:37:43<07:20, 4.85it/s] 96%|█████████▌| 45876/48008 [2:37:43<07:08, 4.97it/s] 96%|█████████▌| 45877/48008 [2:37:44<07:09, 4.96it/s] 96%|█████████▌| 45878/48008 [2:37:44<07:35, 4.67it/s] 96%|█████████▌| 45879/48008 [2:37:44<07:19, 4.84it/s] 96%|█████████▌| 45880/48008 [2:37:44<07:06, 4.99it/s] 96%|█████████▌| 45881/48008 [2:37:45<07:01, 5.05it/s] 96%|█████████▌| 45882/48008 [2:37:45<06:56, 5.11it/s] 96%|█████████▌| 45883/48008 [2:37:45<06:51, 5.16it/s] 96%|█████████▌| 45884/48008 [2:37:45<06:54, 5.13it/s] 96%|█████████▌| 45885/48008 [2:37:45<06:54, 5.13it/s] 96%|█████████▌| 45886/48008 [2:37:45<06:52, 5.15it/s] 96%|█████████▌| 45887/48008 [2:37:46<06:55, 5.11it/s] 96%|█████████▌| 45888/48008 [2:37:46<07:21, 4.80it/s] 96%|█████████▌| 45889/48008 [2:37:46<07:10, 4.92it/s] 96%|█████████▌| 45890/48008 [2:37:46<07:04, 4.99it/s] 96%|█████████▌| 45891/48008 [2:37:46<07:00, 5.03it/s] 96%|█████████▌| 45892/48008 [2:37:47<06:56, 5.08it/s] 96%|█████████▌| 45893/48008 [2:37:47<07:20, 4.80it/s] 96%|█████████▌| 45894/48008 [2:37:47<07:16, 4.85it/s] 96%|█████████▌| 45895/48008 [2:37:47<07:06, 4.95it/s] 96%|█████████▌| 45896/48008 [2:37:48<07:00, 5.02it/s] 96%|█████████▌| 45897/48008 [2:37:48<07:03, 4.98it/s] 96%|█████████▌| 45898/48008 [2:37:48<07:02, 4.99it/s] 96%|█████████▌| 45899/48008 [2:37:48<07:00, 5.02it/s] 96%|█████████▌| 45900/48008 [2:37:48<06:52, 5.11it/s] {'loss': 0.2999, 'grad_norm': 0.37890625, 'learning_rate': 8.786035660723213e-06, 'epoch': 0.96} 96%|█████████▌| 45900/48008 [2:37:48<06:52, 5.11it/s] 96%|█████████▌| 45901/48008 [2:37:48<06:48, 5.16it/s] 96%|█████████▌| 45902/48008 [2:37:49<06:48, 5.15it/s] 96%|█████████▌| 45903/48008 [2:37:49<06:48, 5.16it/s] 96%|█████████▌| 45904/48008 [2:37:49<06:55, 5.07it/s] 96%|█████████▌| 45905/48008 [2:37:49<06:49, 5.14it/s] 96%|█████████▌| 45906/48008 [2:37:49<06:55, 5.06it/s] 96%|█████████▌| 45907/48008 [2:37:50<06:54, 5.07it/s] 96%|█████████▌| 45908/48008 [2:37:50<06:50, 5.12it/s] 96%|█████████▌| 45909/48008 [2:37:50<06:53, 5.08it/s] 96%|█████████▌| 45910/48008 [2:37:50<06:48, 5.13it/s] 96%|█████████▌| 45911/48008 [2:37:50<06:46, 5.16it/s] 96%|█████████▌| 45912/48008 [2:37:51<06:45, 5.17it/s] 96%|█████████▌| 45913/48008 [2:37:51<06:42, 5.21it/s] 96%|█████████▌| 45914/48008 [2:37:51<06:40, 5.23it/s] 96%|█████████▌| 45915/48008 [2:37:51<06:39, 5.24it/s] 96%|█████████▌| 45916/48008 [2:37:51<06:37, 5.26it/s] 96%|█████████▌| 45917/48008 [2:37:52<06:46, 5.14it/s] 96%|█████████▌| 45918/48008 [2:37:52<06:44, 5.16it/s] 96%|█████████▌| 45919/48008 [2:37:52<06:41, 5.21it/s] 96%|█████████▌| 45920/48008 [2:37:52<06:37, 5.26it/s] 96%|█████████▌| 45921/48008 [2:37:52<07:09, 4.86it/s] 96%|█████████▌| 45922/48008 [2:37:53<06:58, 4.98it/s] 96%|█████████▌| 45923/48008 [2:37:53<06:52, 5.06it/s] 96%|█████████▌| 45924/48008 [2:37:53<07:18, 4.76it/s] 96%|█████████▌| 45925/48008 [2:37:53<07:13, 4.80it/s] 96%|█████████▌| 45926/48008 [2:37:53<07:02, 4.92it/s] 96%|█████████▌| 45927/48008 [2:37:54<06:54, 5.01it/s] 96%|█████████▌| 45928/48008 [2:37:54<06:55, 5.00it/s] 96%|█████████▌| 45929/48008 [2:37:54<06:53, 5.03it/s] 96%|█████████▌| 45930/48008 [2:37:54<08:50, 3.92it/s] 96%|█████████▌| 45931/48008 [2:37:55<08:10, 4.24it/s] 96%|█████████▌| 45932/48008 [2:37:55<07:47, 4.44it/s] 96%|█████████▌| 45933/48008 [2:37:55<07:35, 4.55it/s] 96%|█████████▌| 45934/48008 [2:37:55<07:22, 4.69it/s] 96%|█████████▌| 45935/48008 [2:37:55<08:21, 4.13it/s] 96%|█████████▌| 45936/48008 [2:37:56<07:49, 4.41it/s] 96%|█████████▌| 45937/48008 [2:37:56<07:34, 4.55it/s] 96%|█████████▌| 45938/48008 [2:37:56<07:16, 4.74it/s] 96%|█████████▌| 45939/48008 [2:37:56<09:04, 3.80it/s] 96%|█████████▌| 45940/48008 [2:37:57<08:26, 4.08it/s] 96%|█████████▌| 45941/48008 [2:37:57<07:52, 4.37it/s] 96%|█████████▌| 45942/48008 [2:37:58<15:34, 2.21it/s] 96%|█████████▌| 45943/48008 [2:37:58<12:53, 2.67it/s] 96%|█████████▌| 45944/48008 [2:37:58<10:58, 3.13it/s] 96%|█████████▌| 45945/48008 [2:37:58<09:40, 3.55it/s] 96%|█████████▌| 45946/48008 [2:37:59<08:43, 3.94it/s] 96%|█████████▌| 45947/48008 [2:37:59<08:09, 4.21it/s] 96%|█████████▌| 45948/48008 [2:37:59<07:44, 4.43it/s] 96%|█████████▌| 45949/48008 [2:37:59<07:28, 4.59it/s] 96%|█████████▌| 45950/48008 [2:37:59<07:15, 4.73it/s] {'loss': 0.2996, 'grad_norm': 0.240234375, 'learning_rate': 8.57773704382603e-06, 'epoch': 0.96} 96%|█████████▌| 45950/48008 [2:37:59<07:15, 4.73it/s] 96%|█████████▌| 45951/48008 [2:38:00<07:31, 4.56it/s] 96%|█████████▌| 45952/48008 [2:38:00<07:19, 4.68it/s] 96%|█████████▌| 45953/48008 [2:38:00<07:15, 4.72it/s] 96%|█████████▌| 45954/48008 [2:38:00<07:10, 4.77it/s] 96%|█████████▌| 45955/48008 [2:38:00<07:05, 4.83it/s] 96%|█████████▌| 45956/48008 [2:38:01<06:54, 4.95it/s] 96%|█████████▌| 45957/48008 [2:38:01<06:54, 4.94it/s] 96%|█████████▌| 45958/48008 [2:38:01<06:50, 4.99it/s] 96%|█████████▌| 45959/48008 [2:38:01<06:44, 5.06it/s] 96%|█████████▌| 45960/48008 [2:38:01<06:54, 4.93it/s] 96%|█████████▌| 45961/48008 [2:38:02<07:17, 4.68it/s] 96%|█████████▌| 45962/48008 [2:38:02<07:07, 4.78it/s] 96%|█████████▌| 45963/48008 [2:38:02<06:59, 4.87it/s] 96%|█████████▌| 45964/48008 [2:38:02<06:49, 5.00it/s] 96%|█████████▌| 45965/48008 [2:38:02<06:41, 5.09it/s] 96%|█████████▌| 45966/48008 [2:38:03<06:38, 5.13it/s] 96%|█████████▌| 45967/48008 [2:38:03<06:35, 5.16it/s] 96%|█████████▌| 45968/48008 [2:38:03<06:41, 5.08it/s] 96%|█████████▌| 45969/48008 [2:38:03<06:40, 5.10it/s] 96%|█████████▌| 45970/48008 [2:38:04<08:37, 3.94it/s] 96%|█████████▌| 45971/48008 [2:38:04<07:59, 4.25it/s] 96%|█████████▌| 45972/48008 [2:38:04<07:40, 4.42it/s] 96%|█████████▌| 45973/48008 [2:38:04<07:17, 4.65it/s] 96%|█████████▌| 45974/48008 [2:38:04<07:08, 4.74it/s] 96%|█████████▌| 45975/48008 [2:38:05<06:56, 4.88it/s] 96%|█████████▌| 45976/48008 [2:38:05<06:55, 4.89it/s] 96%|█████████▌| 45977/48008 [2:38:05<06:58, 4.85it/s] 96%|█████████▌| 45978/48008 [2:38:05<06:50, 4.95it/s] 96%|█████████▌| 45979/48008 [2:38:05<06:49, 4.96it/s] 96%|█████████▌| 45980/48008 [2:38:06<06:43, 5.03it/s] 96%|█████████▌| 45981/48008 [2:38:06<07:05, 4.76it/s] 96%|█████████▌| 45982/48008 [2:38:06<06:52, 4.91it/s] 96%|█████████▌| 45983/48008 [2:38:06<06:46, 4.98it/s] 96%|█████████▌| 45984/48008 [2:38:06<06:38, 5.08it/s] 96%|█████████▌| 45985/48008 [2:38:07<06:35, 5.11it/s] 96%|█████████▌| 45986/48008 [2:38:07<06:34, 5.13it/s] 96%|█████████▌| 45987/48008 [2:38:07<06:31, 5.16it/s] 96%|█████████▌| 45988/48008 [2:38:07<06:27, 5.21it/s] 96%|█████████▌| 45989/48008 [2:38:07<06:25, 5.24it/s] 96%|█████████▌| 45990/48008 [2:38:08<06:24, 5.25it/s] 96%|█████████▌| 45991/48008 [2:38:08<06:53, 4.88it/s] 96%|█████████▌| 45992/48008 [2:38:08<06:56, 4.84it/s] 96%|█████████▌| 45993/48008 [2:38:08<06:47, 4.95it/s] 96%|█████████▌| 45994/48008 [2:38:09<08:37, 3.89it/s] 96%|█████████▌| 45995/48008 [2:38:09<08:04, 4.16it/s] 96%|█████████▌| 45996/48008 [2:38:09<07:34, 4.43it/s] 96%|█████████▌| 45997/48008 [2:38:09<07:23, 4.54it/s] 96%|█████████▌| 45998/48008 [2:38:09<07:10, 4.67it/s] 96%|█████████▌| 45999/48008 [2:38:10<06:53, 4.86it/s] 96%|█████████▌| 46000/48008 [2:38:10<06:44, 4.97it/s] {'loss': 0.2972, 'grad_norm': 0.1826171875, 'learning_rate': 8.369438426928845e-06, 'epoch': 0.96} 96%|█████████▌| 46000/48008 [2:38:10<06:44, 4.97it/s] 96%|█████████▌| 46001/48008 [2:38:10<06:39, 5.03it/s] 96%|█████████▌| 46002/48008 [2:38:10<06:35, 5.07it/s] 96%|█████████▌| 46003/48008 [2:38:11<08:28, 3.94it/s] 96%|█████████▌| 46004/48008 [2:38:11<07:50, 4.26it/s] 96%|█████████▌| 46005/48008 [2:38:11<07:27, 4.48it/s] 96%|█████████▌| 46006/48008 [2:38:11<07:13, 4.62it/s] 96%|█████████▌| 46007/48008 [2:38:11<06:59, 4.77it/s] 96%|█████████▌| 46008/48008 [2:38:11<06:48, 4.90it/s] 96%|█████████▌| 46009/48008 [2:38:12<06:44, 4.94it/s] 96%|█████████▌| 46010/48008 [2:38:12<06:42, 4.96it/s] 96%|█████████▌| 46011/48008 [2:38:12<06:36, 5.03it/s] 96%|█████████▌| 46012/48008 [2:38:12<06:28, 5.13it/s] 96%|█████████▌| 46013/48008 [2:38:12<06:24, 5.19it/s] 96%|█████████▌| 46014/48008 [2:38:13<06:21, 5.23it/s] 96%|█████████▌| 46015/48008 [2:38:13<06:23, 5.20it/s] 96%|█████████▌| 46016/48008 [2:38:13<06:31, 5.08it/s] 96%|█████████▌| 46017/48008 [2:38:13<07:37, 4.35it/s] 96%|█████████▌| 46018/48008 [2:38:14<07:12, 4.60it/s] 96%|█████████▌| 46019/48008 [2:38:14<06:57, 4.76it/s] 96%|█████████▌| 46020/48008 [2:38:14<06:53, 4.81it/s] 96%|█████████▌| 46021/48008 [2:38:14<06:47, 4.88it/s] 96%|█████████▌| 46022/48008 [2:38:14<06:36, 5.01it/s] 96%|█████████▌| 46023/48008 [2:38:15<08:27, 3.91it/s] 96%|█████████▌| 46024/48008 [2:38:15<07:49, 4.23it/s] 96%|█████████▌| 46025/48008 [2:38:15<07:20, 4.50it/s] 96%|█████████▌| 46026/48008 [2:38:15<07:00, 4.72it/s] 96%|█████████▌| 46027/48008 [2:38:15<06:48, 4.85it/s] 96%|█████████▌| 46028/48008 [2:38:16<06:45, 4.89it/s] 96%|█████████▌| 46029/48008 [2:38:16<06:35, 5.00it/s] 96%|█████████▌| 46030/48008 [2:38:16<06:30, 5.07it/s] 96%|█████████▌| 46031/48008 [2:38:16<06:25, 5.12it/s] 96%|█████████▌| 46032/48008 [2:38:16<06:32, 5.03it/s] 96%|█████████▌| 46033/48008 [2:38:17<06:28, 5.08it/s] 96%|█████████▌| 46034/48008 [2:38:17<06:25, 5.12it/s] 96%|█████████▌| 46035/48008 [2:38:17<06:23, 5.15it/s] 96%|█████████▌| 46036/48008 [2:38:17<06:22, 5.16it/s] 96%|█████████▌| 46037/48008 [2:38:17<06:18, 5.20it/s] 96%|█████████▌| 46038/48008 [2:38:19<15:19, 2.14it/s] 96%|█████████▌| 46039/48008 [2:38:19<12:34, 2.61it/s] 96%|█████████▌| 46040/48008 [2:38:19<10:38, 3.08it/s] 96%|█████████▌| 46041/48008 [2:38:19<09:27, 3.47it/s] 96%|█████████▌| 46042/48008 [2:38:19<08:56, 3.67it/s] 96%|█████████▌| 46043/48008 [2:38:20<08:09, 4.01it/s] 96%|█████████▌| 46044/48008 [2:38:20<07:34, 4.32it/s] 96%|█████████▌| 46045/48008 [2:38:20<07:09, 4.57it/s] 96%|█████████▌| 46046/48008 [2:38:20<06:53, 4.75it/s] 96%|█████████▌| 46047/48008 [2:38:20<06:45, 4.84it/s] 96%|█████████▌| 46048/48008 [2:38:21<07:02, 4.64it/s] 96%|█████████▌| 46049/48008 [2:38:21<06:55, 4.72it/s] 96%|█████████▌| 46050/48008 [2:38:22<14:30, 2.25it/s] {'loss': 0.299, 'grad_norm': 0.154296875, 'learning_rate': 8.161139810031662e-06, 'epoch': 0.96} 96%|█████████▌| 46050/48008 [2:38:22<14:30, 2.25it/s] 96%|█████████▌| 46051/48008 [2:38:22<12:07, 2.69it/s] 96%|█████████▌| 46052/48008 [2:38:22<10:20, 3.15it/s] 96%|█████████▌| 46053/48008 [2:38:22<09:04, 3.59it/s] 96%|█████████▌| 46054/48008 [2:38:22<08:11, 3.98it/s] 96%|█████████▌| 46055/48008 [2:38:23<07:47, 4.18it/s] 96%|█████████▌| 46056/48008 [2:38:24<16:56, 1.92it/s] 96%|█████████▌| 46057/48008 [2:38:24<13:44, 2.37it/s] 96%|█████████▌| 46058/48008 [2:38:24<11:28, 2.83it/s] 96%|█████████▌| 46059/48008 [2:38:24<09:53, 3.28it/s] 96%|█████████▌| 46060/48008 [2:38:25<08:48, 3.69it/s] 96%|█████████▌| 46061/48008 [2:38:25<07:59, 4.06it/s] 96%|█████████▌| 46062/48008 [2:38:25<07:28, 4.34it/s] 96%|█████████▌| 46063/48008 [2:38:25<07:05, 4.57it/s] 96%|█████████▌| 46064/48008 [2:38:25<07:17, 4.45it/s] 96%|█████████▌| 46065/48008 [2:38:26<07:24, 4.37it/s] 96%|█████████▌| 46066/48008 [2:38:26<07:29, 4.32it/s] 96%|█████████▌| 46067/48008 [2:38:26<07:06, 4.55it/s] 96%|█████████▌| 46068/48008 [2:38:26<06:47, 4.76it/s] 96%|█████████▌| 46069/48008 [2:38:27<06:39, 4.85it/s] 96%|█████████▌| 46070/48008 [2:38:27<06:29, 4.98it/s] 96%|█████████▌| 46071/48008 [2:38:27<06:25, 5.02it/s] 96%|█████████▌| 46072/48008 [2:38:27<06:25, 5.02it/s] 96%|█████████▌| 46073/48008 [2:38:27<06:18, 5.11it/s] 96%|█████████▌| 46074/48008 [2:38:27<06:15, 5.15it/s] 96%|█████████▌| 46075/48008 [2:38:28<06:12, 5.19it/s] 96%|█████████▌| 46076/48008 [2:38:28<06:10, 5.21it/s] 96%|█████████▌| 46077/48008 [2:38:28<07:16, 4.43it/s] 96%|█████████▌| 46078/48008 [2:38:28<07:01, 4.58it/s] 96%|█████████▌| 46079/48008 [2:38:29<06:51, 4.69it/s] 96%|█████████▌| 46080/48008 [2:38:29<06:41, 4.80it/s] 96%|█████████▌| 46081/48008 [2:38:29<06:59, 4.60it/s] 96%|█████████▌| 46082/48008 [2:38:29<06:43, 4.77it/s] 96%|█████████▌| 46083/48008 [2:38:29<06:37, 4.85it/s] 96%|█████████▌| 46084/48008 [2:38:30<06:29, 4.94it/s] 96%|█████████▌| 46085/48008 [2:38:30<06:29, 4.93it/s] 96%|█████████▌| 46086/48008 [2:38:30<06:22, 5.03it/s] 96%|█████████▌| 46087/48008 [2:38:30<06:17, 5.09it/s] 96%|█████████▌| 46088/48008 [2:38:30<06:14, 5.13it/s] 96%|█████████▌| 46089/48008 [2:38:31<06:16, 5.10it/s] 96%|█████████▌| 46090/48008 [2:38:31<06:11, 5.16it/s] 96%|█████████▌| 46091/48008 [2:38:31<06:09, 5.19it/s] 96%|█████████▌| 46092/48008 [2:38:31<06:08, 5.21it/s] 96%|█████████▌| 46093/48008 [2:38:31<06:05, 5.24it/s] 96%|█████████▌| 46094/48008 [2:38:31<06:03, 5.27it/s] 96%|█████████▌| 46095/48008 [2:38:32<06:30, 4.90it/s] 96%|█████████▌| 46096/48008 [2:38:32<06:26, 4.95it/s] 96%|█████████▌| 46097/48008 [2:38:32<06:20, 5.03it/s] 96%|█████████▌| 46098/48008 [2:38:32<06:15, 5.09it/s] 96%|█████████▌| 46099/48008 [2:38:33<06:12, 5.13it/s] 96%|█████████▌| 46100/48008 [2:38:33<06:09, 5.17it/s] {'loss': 0.299, 'grad_norm': 0.25390625, 'learning_rate': 7.952841193134478e-06, 'epoch': 0.96} 96%|█████████▌| 46100/48008 [2:38:33<06:09, 5.17it/s] 96%|█████████▌| 46101/48008 [2:38:33<06:16, 5.06it/s] 96%|█████████▌| 46102/48008 [2:38:33<06:17, 5.04it/s] 96%|█████████▌| 46103/48008 [2:38:33<06:14, 5.08it/s] 96%|█████████▌| 46104/48008 [2:38:33<06:10, 5.14it/s] 96%|█████████▌| 46105/48008 [2:38:34<06:07, 5.17it/s] 96%|█████████▌| 46106/48008 [2:38:34<06:04, 5.21it/s] 96%|█████████▌| 46107/48008 [2:38:34<06:30, 4.87it/s] 96%|█████████▌| 46108/48008 [2:38:34<06:29, 4.88it/s] 96%|█████████▌| 46109/48008 [2:38:34<06:24, 4.95it/s] 96%|█████████▌| 46110/48008 [2:38:35<06:17, 5.03it/s] 96%|█████████▌| 46111/48008 [2:38:35<06:10, 5.12it/s] 96%|█████████▌| 46112/48008 [2:38:35<06:07, 5.16it/s] 96%|█████████▌| 46113/48008 [2:38:35<06:08, 5.15it/s] 96%|█████████▌| 46114/48008 [2:38:35<06:07, 5.15it/s] 96%|█████████▌| 46115/48008 [2:38:36<06:02, 5.22it/s] 96%|█████████▌| 46116/48008 [2:38:36<06:05, 5.18it/s] 96%|█████████▌| 46117/48008 [2:38:36<06:00, 5.25it/s] 96%|█████████▌| 46118/48008 [2:38:36<06:08, 5.13it/s] 96%|█████████▌| 46119/48008 [2:38:36<06:04, 5.18it/s] 96%|█████████▌| 46120/48008 [2:38:37<06:01, 5.22it/s] 96%|█████████▌| 46121/48008 [2:38:37<06:09, 5.11it/s] 96%|█████████▌| 46122/48008 [2:38:38<12:09, 2.59it/s] 96%|█████████▌| 46123/48008 [2:38:38<10:12, 3.08it/s] 96%|█████████▌| 46124/48008 [2:38:38<09:03, 3.47it/s] 96%|█████████▌| 46125/48008 [2:38:38<08:08, 3.85it/s] 96%|█████████▌| 46126/48008 [2:38:38<07:28, 4.19it/s] 96%|█████████▌| 46127/48008 [2:38:39<07:02, 4.45it/s] 96%|█████████▌| 46128/48008 [2:38:39<06:42, 4.68it/s] 96%|█████████▌| 46129/48008 [2:38:39<06:33, 4.78it/s] 96%|█████████▌| 46130/48008 [2:38:39<06:24, 4.89it/s] 96%|█████████▌| 46131/48008 [2:38:39<06:15, 5.00it/s] 96%|█████████▌| 46132/48008 [2:38:40<06:16, 4.98it/s] 96%|█████████▌| 46133/48008 [2:38:40<06:09, 5.07it/s] 96%|█████████▌| 46134/48008 [2:38:40<06:07, 5.10it/s] 96%|█████████▌| 46135/48008 [2:38:40<06:10, 5.06it/s] 96%|█████████▌| 46136/48008 [2:38:40<06:06, 5.10it/s] 96%|█████████▌| 46137/48008 [2:38:41<06:03, 5.15it/s] 96%|█████████▌| 46138/48008 [2:38:41<06:01, 5.17it/s] 96%|█████████▌| 46139/48008 [2:38:41<06:26, 4.83it/s] 96%|█████████▌| 46140/48008 [2:38:41<06:23, 4.88it/s] 96%|█████████▌| 46141/48008 [2:38:41<06:15, 4.98it/s] 96%|█████████▌| 46142/48008 [2:38:42<06:06, 5.10it/s] 96%|█████████▌| 46143/48008 [2:38:42<06:01, 5.15it/s] 96%|█████████▌| 46144/48008 [2:38:42<05:58, 5.20it/s] 96%|█████████▌| 46145/48008 [2:38:42<05:57, 5.21it/s] 96%|█████████▌| 46146/48008 [2:38:42<05:57, 5.21it/s] 96%|█████████▌| 46147/48008 [2:38:43<05:58, 5.19it/s] 96%|█████████▌| 46148/48008 [2:38:43<05:58, 5.19it/s] 96%|█████████▌| 46149/48008 [2:38:43<05:57, 5.20it/s] 96%|█████████▌| 46150/48008 [2:38:43<05:53, 5.25it/s] {'loss': 0.3018, 'grad_norm': 0.24609375, 'learning_rate': 7.744542576237295e-06, 'epoch': 0.96} 96%|█████████▌| 46150/48008 [2:38:43<05:53, 5.25it/s] 96%|█████████▌| 46151/48008 [2:38:43<06:04, 5.10it/s] 96%|█████████▌| 46152/48008 [2:38:43<06:07, 5.06it/s] 96%|█████████▌| 46153/48008 [2:38:44<06:05, 5.08it/s] 96%|█████████▌| 46154/48008 [2:38:44<07:07, 4.34it/s] 96%|█████████▌| 46155/48008 [2:38:44<06:44, 4.58it/s] 96%|█████████▌| 46156/48008 [2:38:44<06:33, 4.71it/s] 96%|█████████▌| 46157/48008 [2:38:45<06:21, 4.85it/s] 96%|█████████▌| 46158/48008 [2:38:45<06:16, 4.92it/s] 96%|█████████▌| 46159/48008 [2:38:45<06:09, 5.00it/s] 96%|█████████▌| 46160/48008 [2:38:45<06:03, 5.08it/s] 96%|█████████▌| 46161/48008 [2:38:45<06:01, 5.11it/s] 96%|█████████▌| 46162/48008 [2:38:46<07:45, 3.96it/s] 96%|█████████▌| 46163/48008 [2:38:46<07:12, 4.26it/s] 96%|█████████▌| 46164/48008 [2:38:46<06:50, 4.50it/s] 96%|█████████▌| 46165/48008 [2:38:46<06:35, 4.66it/s] 96%|█████████▌| 46166/48008 [2:38:46<06:21, 4.83it/s] 96%|█████████▌| 46167/48008 [2:38:47<06:14, 4.92it/s] 96%|█████████▌| 46168/48008 [2:38:47<06:16, 4.88it/s] 96%|█████████▌| 46169/48008 [2:38:47<06:08, 4.99it/s] 96%|█████████▌| 46170/48008 [2:38:47<06:01, 5.09it/s] 96%|█████████▌| 46171/48008 [2:38:47<06:00, 5.09it/s] 96%|█████████▌| 46172/48008 [2:38:48<05:59, 5.11it/s] 96%|█████████▌| 46173/48008 [2:38:48<05:55, 5.16it/s] 96%|█████████▌| 46174/48008 [2:38:48<05:55, 5.16it/s] 96%|█████████▌| 46175/48008 [2:38:48<05:53, 5.19it/s] 96%|█████████▌| 46176/48008 [2:38:49<06:55, 4.41it/s] 96%|█████████▌| 46177/48008 [2:38:49<06:46, 4.51it/s] 96%|█████████▌| 46178/48008 [2:38:49<06:34, 4.64it/s] 96%|█████████▌| 46179/48008 [2:38:49<06:20, 4.81it/s] 96%|█████████▌| 46180/48008 [2:38:49<06:14, 4.89it/s] 96%|█████████▌| 46181/48008 [2:38:50<06:08, 4.96it/s] 96%|█████████▌| 46182/48008 [2:38:50<06:02, 5.03it/s] 96%|█████████▌| 46183/48008 [2:38:50<05:56, 5.11it/s] 96%|█████████▌| 46184/48008 [2:38:50<05:52, 5.17it/s] 96%|█████████▌| 46185/48008 [2:38:50<07:37, 3.99it/s] 96%|█████████▌| 46186/48008 [2:38:51<07:16, 4.18it/s] 96%|█████████▌| 46187/48008 [2:38:51<06:47, 4.46it/s] 96%|█████████▌| 46188/48008 [2:38:51<06:31, 4.65it/s] 96%|█████████▌| 46189/48008 [2:38:51<06:17, 4.82it/s] 96%|█████████▌| 46190/48008 [2:38:51<06:07, 4.94it/s] 96%|█████████▌| 46191/48008 [2:38:52<06:01, 5.03it/s] 96%|█████████▌| 46192/48008 [2:38:52<06:21, 4.76it/s] 96%|█████████▌| 46193/48008 [2:38:52<06:17, 4.80it/s] 96%|█████████▌| 46194/48008 [2:38:52<06:06, 4.96it/s] 96%|█████████▌| 46195/48008 [2:38:52<06:05, 4.96it/s] 96%|█████████▌| 46196/48008 [2:38:53<06:05, 4.96it/s] 96%|█████████▌| 46197/48008 [2:38:53<05:59, 5.04it/s] 96%|█████████▌| 46198/48008 [2:38:53<05:54, 5.11it/s] 96%|█████████▌| 46199/48008 [2:38:53<05:53, 5.12it/s] 96%|█████████▌| 46200/48008 [2:38:53<05:55, 5.09it/s] {'loss': 0.2966, 'grad_norm': 0.14453125, 'learning_rate': 7.53624395934011e-06, 'epoch': 0.96} 96%|█████████▌| 46200/48008 [2:38:53<05:55, 5.09it/s] 96%|█████████▌| 46201/48008 [2:38:54<05:56, 5.08it/s] 96%|█████████▌| 46202/48008 [2:38:54<05:52, 5.12it/s] 96%|█████████▌| 46203/48008 [2:38:54<05:51, 5.14it/s] 96%|█████████▌| 46204/48008 [2:38:54<05:55, 5.08it/s] 96%|█████████▌| 46205/48008 [2:38:54<05:50, 5.15it/s] 96%|█████████▌| 46206/48008 [2:38:55<05:48, 5.17it/s] 96%|█████████▌| 46207/48008 [2:38:55<05:46, 5.20it/s] 96%|█████████▋| 46208/48008 [2:38:55<05:43, 5.23it/s] 96%|█████████▋| 46209/48008 [2:38:55<05:43, 5.24it/s] 96%|█████████▋| 46210/48008 [2:38:55<05:43, 5.24it/s] 96%|█████████▋| 46211/48008 [2:38:56<05:49, 5.15it/s] 96%|█████████▋| 46212/48008 [2:38:56<05:55, 5.05it/s] 96%|█████████▋| 46213/48008 [2:38:56<05:49, 5.13it/s] 96%|█████████▋| 46214/48008 [2:38:56<05:48, 5.15it/s] 96%|█████████▋| 46215/48008 [2:38:56<05:41, 5.25it/s] 96%|█████████▋| 46216/48008 [2:38:57<05:48, 5.14it/s] 96%|█████████▋| 46217/48008 [2:38:57<05:44, 5.19it/s] 96%|█████████▋| 46218/48008 [2:38:57<05:49, 5.12it/s] 96%|█████████▋| 46219/48008 [2:38:57<05:51, 5.10it/s] 96%|█████████▋| 46220/48008 [2:38:57<05:53, 5.06it/s] 96%|█████████▋| 46221/48008 [2:38:58<05:49, 5.11it/s] 96%|█████████▋| 46222/48008 [2:38:58<05:47, 5.15it/s] 96%|█████████▋| 46223/48008 [2:38:58<05:44, 5.18it/s] 96%|█████████▋| 46224/48008 [2:38:58<05:53, 5.05it/s] 96%|█████████▋| 46225/48008 [2:38:58<05:50, 5.08it/s] 96%|█████████▋| 46226/48008 [2:38:59<05:51, 5.07it/s] 96%|█████████▋| 46227/48008 [2:38:59<05:49, 5.09it/s] 96%|█████████▋| 46228/48008 [2:38:59<05:44, 5.17it/s] 96%|█████████▋| 46229/48008 [2:38:59<05:45, 5.15it/s] 96%|█████████▋| 46230/48008 [2:38:59<05:51, 5.06it/s] 96%|█████████▋| 46231/48008 [2:39:00<05:54, 5.02it/s] 96%|█████████▋| 46232/48008 [2:39:00<05:54, 5.00it/s] 96%|█████████▋| 46233/48008 [2:39:00<05:51, 5.05it/s] 96%|█████████▋| 46234/48008 [2:39:00<05:46, 5.13it/s] 96%|█████████▋| 46235/48008 [2:39:00<05:42, 5.17it/s] 96%|█████████▋| 46236/48008 [2:39:00<05:41, 5.19it/s] 96%|█████████▋| 46237/48008 [2:39:01<05:40, 5.21it/s] 96%|█████████▋| 46238/48008 [2:39:01<05:41, 5.18it/s] 96%|█████████▋| 46239/48008 [2:39:01<05:37, 5.23it/s] 96%|█████████▋| 46240/48008 [2:39:01<05:32, 5.32it/s] 96%|█████████▋| 46241/48008 [2:39:01<05:31, 5.33it/s] 96%|█████████▋| 46242/48008 [2:39:02<05:32, 5.32it/s] 96%|█████████▋| 46243/48008 [2:39:02<05:33, 5.29it/s] 96%|█████████▋| 46244/48008 [2:39:02<05:45, 5.10it/s] 96%|█████████▋| 46245/48008 [2:39:02<06:07, 4.80it/s] 96%|█████████▋| 46246/48008 [2:39:02<06:00, 4.88it/s] 96%|█████████▋| 46247/48008 [2:39:03<05:53, 4.98it/s] 96%|█████████▋| 46248/48008 [2:39:03<05:51, 5.00it/s] 96%|█████████▋| 46249/48008 [2:39:03<05:45, 5.09it/s] 96%|█████████▋| 46250/48008 [2:39:03<05:41, 5.15it/s] {'loss': 0.3022, 'grad_norm': 0.365234375, 'learning_rate': 7.327945342442926e-06, 'epoch': 0.96} 96%|█████████▋| 46250/48008 [2:39:03<05:41, 5.15it/s] 96%|█████████▋| 46251/48008 [2:39:03<05:38, 5.19it/s] 96%|█████████▋| 46252/48008 [2:39:04<05:33, 5.27it/s] 96%|█████████▋| 46253/48008 [2:39:04<05:30, 5.31it/s] 96%|█████████▋| 46254/48008 [2:39:04<05:29, 5.32it/s] 96%|█████████▋| 46255/48008 [2:39:04<05:32, 5.28it/s] 96%|█████████▋| 46256/48008 [2:39:04<05:35, 5.22it/s] 96%|█████████▋| 46257/48008 [2:39:05<05:35, 5.22it/s] 96%|█████████▋| 46258/48008 [2:39:05<05:33, 5.24it/s] 96%|█████████▋| 46259/48008 [2:39:05<05:33, 5.25it/s] 96%|█████████▋| 46260/48008 [2:39:05<05:38, 5.17it/s] 96%|█████████▋| 46261/48008 [2:39:05<05:39, 5.15it/s] 96%|█████████▋| 46262/48008 [2:39:06<05:39, 5.14it/s] 96%|█████████▋| 46263/48008 [2:39:06<05:39, 5.14it/s] 96%|█████████▋| 46264/48008 [2:39:06<06:01, 4.82it/s] 96%|█████████▋| 46265/48008 [2:39:06<05:50, 4.97it/s] 96%|█████████▋| 46266/48008 [2:39:06<05:45, 5.05it/s] 96%|█████████▋| 46267/48008 [2:39:06<05:39, 5.13it/s] 96%|█████████▋| 46268/48008 [2:39:07<05:41, 5.10it/s] 96%|█████████▋| 46269/48008 [2:39:07<05:38, 5.14it/s] 96%|█████████▋| 46270/48008 [2:39:07<05:36, 5.16it/s] 96%|█████████▋| 46271/48008 [2:39:07<05:41, 5.08it/s] 96%|█████████▋| 46272/48008 [2:39:07<05:41, 5.08it/s] 96%|█████████▋| 46273/48008 [2:39:08<06:05, 4.74it/s] 96%|█████████▋| 46274/48008 [2:39:08<05:56, 4.87it/s] 96%|█████████▋| 46275/48008 [2:39:08<05:46, 5.00it/s] 96%|█████████▋| 46276/48008 [2:39:08<05:46, 5.00it/s] 96%|█████████▋| 46277/48008 [2:39:08<05:41, 5.07it/s] 96%|█████████▋| 46278/48008 [2:39:09<06:03, 4.76it/s] 96%|█████████▋| 46279/48008 [2:39:09<05:50, 4.93it/s] 96%|█████████▋| 46280/48008 [2:39:09<05:41, 5.06it/s] 96%|█████████▋| 46281/48008 [2:39:09<05:37, 5.12it/s] 96%|█████████▋| 46282/48008 [2:39:09<05:31, 5.20it/s] 96%|█████████▋| 46283/48008 [2:39:10<05:31, 5.21it/s] 96%|█████████▋| 46284/48008 [2:39:10<05:29, 5.23it/s] 96%|█████████▋| 46285/48008 [2:39:10<05:32, 5.19it/s] 96%|█████████▋| 46286/48008 [2:39:10<05:30, 5.21it/s] 96%|█████████▋| 46287/48008 [2:39:10<05:29, 5.23it/s] 96%|█████████▋| 46288/48008 [2:39:11<05:32, 5.18it/s] 96%|█████████▋| 46289/48008 [2:39:11<05:29, 5.22it/s] 96%|█████████▋| 46290/48008 [2:39:11<05:26, 5.26it/s] 96%|█████████▋| 46291/48008 [2:39:11<05:27, 5.24it/s] 96%|█████████▋| 46292/48008 [2:39:11<05:26, 5.25it/s] 96%|█████████▋| 46293/48008 [2:39:12<05:23, 5.30it/s] 96%|█████████▋| 46294/48008 [2:39:12<05:30, 5.18it/s] 96%|█████████▋| 46295/48008 [2:39:12<05:29, 5.19it/s] 96%|█████████▋| 46296/48008 [2:39:12<05:28, 5.21it/s] 96%|█████████▋| 46297/48008 [2:39:12<05:26, 5.24it/s] 96%|█████████▋| 46298/48008 [2:39:13<05:27, 5.23it/s] 96%|█████████▋| 46299/48008 [2:39:13<05:26, 5.23it/s] 96%|█████████▋| 46300/48008 [2:39:13<05:33, 5.11it/s] {'loss': 0.3038, 'grad_norm': 0.7421875, 'learning_rate': 7.119646725545742e-06, 'epoch': 0.96} 96%|█████████▋| 46300/48008 [2:39:13<05:33, 5.11it/s] 96%|█████████▋| 46301/48008 [2:39:13<05:38, 5.05it/s] 96%|█████████▋| 46302/48008 [2:39:13<05:34, 5.11it/s] 96%|█████████▋| 46303/48008 [2:39:14<05:30, 5.16it/s] 96%|█████████▋| 46304/48008 [2:39:14<05:27, 5.21it/s] 96%|█████████▋| 46305/48008 [2:39:14<05:27, 5.21it/s] 96%|█████████▋| 46306/48008 [2:39:14<05:29, 5.17it/s] 96%|█████████▋| 46307/48008 [2:39:14<05:26, 5.20it/s] 96%|█████████▋| 46308/48008 [2:39:14<05:27, 5.20it/s] 96%|█████████▋| 46309/48008 [2:39:15<05:26, 5.20it/s] 96%|█████████▋| 46310/48008 [2:39:15<05:32, 5.11it/s] 96%|█████████▋| 46311/48008 [2:39:15<05:30, 5.13it/s] 96%|█████████▋| 46312/48008 [2:39:15<05:31, 5.11it/s] 96%|█████████▋| 46313/48008 [2:39:15<05:28, 5.16it/s] 96%|█████████▋| 46314/48008 [2:39:16<05:30, 5.13it/s] 96%|█████████▋| 46315/48008 [2:39:16<05:28, 5.15it/s] 96%|█████████▋| 46316/48008 [2:39:16<05:36, 5.03it/s] 96%|█████████▋| 46317/48008 [2:39:16<05:29, 5.14it/s] 96%|█████████▋| 46318/48008 [2:39:16<05:25, 5.20it/s] 96%|█████████▋| 46319/48008 [2:39:17<05:33, 5.06it/s] 96%|█████████▋| 46320/48008 [2:39:17<05:36, 5.02it/s] 96%|█████████▋| 46321/48008 [2:39:17<05:28, 5.14it/s] 96%|█████████▋| 46322/48008 [2:39:17<05:27, 5.14it/s] 96%|█████████▋| 46323/48008 [2:39:17<05:31, 5.08it/s] 96%|█████████▋| 46324/48008 [2:39:18<05:51, 4.79it/s] 96%|█████████▋| 46325/48008 [2:39:18<05:40, 4.94it/s] 96%|█████████▋| 46326/48008 [2:39:18<05:33, 5.04it/s] 96%|█████████▋| 46327/48008 [2:39:18<05:53, 4.75it/s] 97%|█████████▋| 46328/48008 [2:39:18<05:43, 4.89it/s] 97%|█████████▋| 46329/48008 [2:39:19<05:35, 5.01it/s] 97%|█████████▋| 46330/48008 [2:39:19<05:29, 5.09it/s] 97%|█████████▋| 46331/48008 [2:39:19<05:24, 5.16it/s] 97%|█████████▋| 46332/48008 [2:39:19<05:25, 5.14it/s] 97%|█████████▋| 46333/48008 [2:39:20<06:21, 4.39it/s] 97%|█████████▋| 46334/48008 [2:39:20<06:02, 4.61it/s] 97%|█████████▋| 46335/48008 [2:39:20<05:55, 4.70it/s] 97%|█████████▋| 46336/48008 [2:39:20<05:42, 4.89it/s] 97%|█████████▋| 46337/48008 [2:39:20<05:34, 4.99it/s] 97%|█████████▋| 46338/48008 [2:39:20<05:29, 5.07it/s] 97%|█████████▋| 46339/48008 [2:39:21<05:31, 5.03it/s] 97%|█████████▋| 46340/48008 [2:39:21<05:25, 5.12it/s] 97%|█████████▋| 46341/48008 [2:39:21<05:21, 5.18it/s] 97%|█████████▋| 46342/48008 [2:39:21<05:22, 5.17it/s] 97%|█████████▋| 46343/48008 [2:39:21<05:30, 5.03it/s] 97%|█████████▋| 46344/48008 [2:39:22<05:25, 5.11it/s] 97%|█████████▋| 46345/48008 [2:39:22<05:21, 5.17it/s] 97%|█████████▋| 46346/48008 [2:39:22<05:42, 4.85it/s] 97%|█████████▋| 46347/48008 [2:39:22<07:12, 3.84it/s] 97%|█████████▋| 46348/48008 [2:39:23<06:37, 4.17it/s] 97%|█████████▋| 46349/48008 [2:39:23<06:12, 4.46it/s] 97%|█████████▋| 46350/48008 [2:39:23<05:52, 4.70it/s] {'loss': 0.3011, 'grad_norm': 0.51953125, 'learning_rate': 6.911348108648559e-06, 'epoch': 0.97} 97%|█████████▋| 46350/48008 [2:39:23<05:52, 4.70it/s] 97%|█████████▋| 46351/48008 [2:39:23<05:41, 4.86it/s] 97%|█████████▋| 46352/48008 [2:39:23<05:33, 4.97it/s] 97%|█████████▋| 46353/48008 [2:39:24<05:27, 5.06it/s] 97%|█████████▋| 46354/48008 [2:39:24<05:23, 5.12it/s] 97%|█████████▋| 46355/48008 [2:39:24<05:20, 5.16it/s] 97%|█████████▋| 46356/48008 [2:39:24<05:22, 5.13it/s] 97%|█████████▋| 46357/48008 [2:39:24<05:42, 4.82it/s] 97%|█████████▋| 46358/48008 [2:39:25<05:32, 4.97it/s] 97%|█████████▋| 46359/48008 [2:39:25<05:26, 5.05it/s] 97%|█████████▋| 46360/48008 [2:39:25<05:28, 5.02it/s] 97%|█████████▋| 46361/48008 [2:39:25<05:26, 5.04it/s] 97%|█████████▋| 46362/48008 [2:39:25<05:23, 5.09it/s] 97%|█████████▋| 46363/48008 [2:39:26<05:23, 5.08it/s] 97%|█████████▋| 46364/48008 [2:39:26<05:18, 5.16it/s] 97%|█████████▋| 46365/48008 [2:39:26<05:23, 5.07it/s] 97%|█████████▋| 46366/48008 [2:39:26<05:21, 5.11it/s] 97%|█████████▋| 46367/48008 [2:39:26<05:24, 5.06it/s] 97%|█████████▋| 46368/48008 [2:39:27<05:31, 4.95it/s] 97%|█████████▋| 46369/48008 [2:39:27<05:27, 5.00it/s] 97%|█████████▋| 46370/48008 [2:39:27<05:22, 5.07it/s] 97%|█████████▋| 46371/48008 [2:39:27<05:19, 5.12it/s] 97%|█████████▋| 46372/48008 [2:39:27<05:17, 5.16it/s] 97%|█████████▋| 46373/48008 [2:39:28<05:16, 5.17it/s] 97%|█████████▋| 46374/48008 [2:39:28<05:15, 5.17it/s] 97%|█████████▋| 46375/48008 [2:39:28<05:13, 5.20it/s] 97%|█████████▋| 46376/48008 [2:39:28<05:12, 5.22it/s] 97%|█████████▋| 46377/48008 [2:39:28<05:10, 5.26it/s] 97%|█████████▋| 46378/48008 [2:39:28<05:08, 5.28it/s] 97%|█████████▋| 46379/48008 [2:39:29<05:15, 5.17it/s] 97%|█████████▋| 46380/48008 [2:39:29<06:09, 4.40it/s] 97%|█████████▋| 46381/48008 [2:39:29<05:50, 4.64it/s] 97%|█████████▋| 46382/48008 [2:39:29<05:40, 4.78it/s] 97%|█████████▋| 46383/48008 [2:39:30<05:31, 4.90it/s] 97%|█████████▋| 46384/48008 [2:39:30<05:25, 4.99it/s] 97%|█████████▋| 46385/48008 [2:39:30<05:29, 4.93it/s] 97%|█████████▋| 46386/48008 [2:39:30<05:22, 5.03it/s] 97%|█████████▋| 46387/48008 [2:39:30<05:17, 5.10it/s] 97%|█████████▋| 46388/48008 [2:39:31<05:20, 5.05it/s] 97%|█████████▋| 46389/48008 [2:39:31<05:38, 4.78it/s] 97%|█████████▋| 46390/48008 [2:39:31<05:32, 4.86it/s] 97%|█████████▋| 46391/48008 [2:39:31<05:24, 4.98it/s] 97%|█████████▋| 46392/48008 [2:39:31<05:23, 4.99it/s] 97%|█████████▋| 46393/48008 [2:39:32<05:40, 4.74it/s] 97%|█████████▋| 46394/48008 [2:39:32<05:52, 4.57it/s] 97%|█████████▋| 46395/48008 [2:39:32<05:37, 4.79it/s] 97%|█████████▋| 46396/48008 [2:39:32<05:50, 4.60it/s] 97%|█████████▋| 46397/48008 [2:39:32<05:43, 4.69it/s] 97%|█████████▋| 46398/48008 [2:39:33<05:38, 4.75it/s] 97%|█████████▋| 46399/48008 [2:39:33<05:31, 4.85it/s] 97%|█████████▋| 46400/48008 [2:39:33<05:23, 4.97it/s] {'loss': 0.2997, 'grad_norm': 0.15625, 'learning_rate': 6.703049491751374e-06, 'epoch': 0.97} 97%|█████████▋| 46400/48008 [2:39:33<05:23, 4.97it/s] 97%|█████████▋| 46401/48008 [2:39:33<05:16, 5.08it/s] 97%|█████████▋| 46402/48008 [2:39:33<05:15, 5.09it/s] 97%|█████████▋| 46403/48008 [2:39:34<05:12, 5.14it/s] 97%|█████████▋| 46404/48008 [2:39:34<05:14, 5.10it/s] 97%|█████████▋| 46405/48008 [2:39:34<05:12, 5.12it/s] 97%|█████████▋| 46406/48008 [2:39:34<05:12, 5.12it/s] 97%|█████████▋| 46407/48008 [2:39:35<06:44, 3.96it/s] 97%|█████████▋| 46408/48008 [2:39:35<06:18, 4.23it/s] 97%|█████████▋| 46409/48008 [2:39:35<05:57, 4.48it/s] 97%|█████████▋| 46410/48008 [2:39:35<05:44, 4.63it/s] 97%|█████████▋| 46411/48008 [2:39:35<05:38, 4.72it/s] 97%|█████████▋| 46412/48008 [2:39:36<05:34, 4.77it/s] 97%|█████████▋| 46413/48008 [2:39:36<05:30, 4.83it/s] 97%|█████████▋| 46414/48008 [2:39:36<05:24, 4.91it/s] 97%|█████████▋| 46415/48008 [2:39:36<05:27, 4.86it/s] 97%|█████████▋| 46416/48008 [2:39:36<05:26, 4.88it/s] 97%|█████████▋| 46417/48008 [2:39:37<06:13, 4.26it/s] 97%|█████████▋| 46418/48008 [2:39:37<05:52, 4.51it/s] 97%|█████████▋| 46419/48008 [2:39:37<05:44, 4.62it/s] 97%|█████████▋| 46420/48008 [2:39:38<07:04, 3.74it/s] 97%|█████████▋| 46421/48008 [2:39:38<06:29, 4.08it/s] 97%|█████████▋| 46422/48008 [2:39:38<06:02, 4.37it/s] 97%|█████████▋| 46423/48008 [2:39:38<05:44, 4.60it/s] 97%|█████████▋| 46424/48008 [2:39:38<05:31, 4.78it/s] 97%|█████████▋| 46425/48008 [2:39:38<05:21, 4.93it/s] 97%|█████████▋| 46426/48008 [2:39:39<05:13, 5.04it/s] 97%|█████████▋| 46427/48008 [2:39:39<05:12, 5.06it/s] 97%|█████████▋| 46428/48008 [2:39:39<05:18, 4.96it/s] 97%|█████████▋| 46429/48008 [2:39:39<05:15, 5.01it/s] 97%|█████████▋| 46430/48008 [2:39:39<05:16, 4.99it/s] 97%|█████████▋| 46431/48008 [2:39:40<05:11, 5.07it/s] 97%|█████████▋| 46432/48008 [2:39:40<05:11, 5.06it/s] 97%|█████████▋| 46433/48008 [2:39:40<05:10, 5.06it/s] 97%|█████████▋| 46434/48008 [2:39:40<05:08, 5.10it/s] 97%|█████████▋| 46435/48008 [2:39:40<05:05, 5.15it/s] 97%|█████████▋| 46436/48008 [2:39:41<05:02, 5.19it/s] 97%|█████████▋| 46437/48008 [2:39:41<05:23, 4.85it/s] 97%|█████████▋| 46438/48008 [2:39:41<05:17, 4.95it/s] 97%|█████████▋| 46439/48008 [2:39:41<05:12, 5.02it/s] 97%|█████████▋| 46440/48008 [2:39:41<05:13, 5.00it/s] 97%|█████████▋| 46441/48008 [2:39:42<05:06, 5.11it/s] 97%|█████████▋| 46442/48008 [2:39:42<05:10, 5.04it/s] 97%|█████████▋| 46443/48008 [2:39:42<05:05, 5.12it/s] 97%|█████████▋| 46444/48008 [2:39:42<05:03, 5.15it/s] 97%|█████████▋| 46445/48008 [2:39:42<05:02, 5.16it/s] 97%|█████████▋| 46446/48008 [2:39:43<05:23, 4.84it/s] 97%|█████████▋| 46447/48008 [2:39:43<05:13, 4.98it/s] 97%|█████████▋| 46448/48008 [2:39:43<05:10, 5.03it/s] 97%|█████████▋| 46449/48008 [2:39:43<05:07, 5.07it/s] 97%|█████████▋| 46450/48008 [2:39:43<05:10, 5.02it/s] {'loss': 0.2968, 'grad_norm': 0.1787109375, 'learning_rate': 6.494750874854192e-06, 'epoch': 0.97} 97%|█████████▋| 46450/48008 [2:39:43<05:10, 5.02it/s] 97%|█████████▋| 46451/48008 [2:39:44<05:07, 5.06it/s] 97%|█████████▋| 46452/48008 [2:39:44<05:04, 5.10it/s] 97%|█████████▋| 46453/48008 [2:39:44<05:04, 5.11it/s] 97%|█████████▋| 46454/48008 [2:39:44<05:02, 5.14it/s] 97%|█████████▋| 46455/48008 [2:39:44<04:58, 5.20it/s] 97%|█████████▋| 46456/48008 [2:39:45<05:01, 5.16it/s] 97%|█████████▋| 46457/48008 [2:39:46<12:49, 2.02it/s] 97%|█████████▋| 46458/48008 [2:39:46<10:27, 2.47it/s] 97%|█████████▋| 46459/48008 [2:39:46<08:46, 2.94it/s] 97%|█████████▋| 46460/48008 [2:39:46<07:38, 3.38it/s] 97%|█████████▋| 46461/48008 [2:39:47<06:49, 3.78it/s] 97%|█████████▋| 46462/48008 [2:39:47<06:13, 4.14it/s] 97%|█████████▋| 46463/48008 [2:39:47<05:52, 4.39it/s] 97%|█████████▋| 46464/48008 [2:39:47<05:34, 4.61it/s] 97%|█████████▋| 46465/48008 [2:39:47<05:28, 4.70it/s] 97%|█████████▋| 46466/48008 [2:39:48<05:24, 4.76it/s] 97%|█████████▋| 46467/48008 [2:39:48<05:14, 4.91it/s] 97%|█████████▋| 46468/48008 [2:39:48<05:09, 4.97it/s] 97%|█████████▋| 46469/48008 [2:39:48<05:05, 5.04it/s] 97%|█████████▋| 46470/48008 [2:39:48<05:01, 5.11it/s] 97%|█████████▋| 46471/48008 [2:39:48<05:00, 5.12it/s] 97%|█████████▋| 46472/48008 [2:39:49<04:59, 5.12it/s] 97%|█████████▋| 46473/48008 [2:39:49<04:57, 5.16it/s] 97%|█████████▋| 46474/48008 [2:39:49<05:01, 5.09it/s] 97%|█████████▋| 46475/48008 [2:39:49<04:59, 5.13it/s] 97%|█████████▋| 46476/48008 [2:39:49<05:05, 5.02it/s] 97%|█████████▋| 46477/48008 [2:39:50<06:31, 3.91it/s] 97%|█████████▋| 46478/48008 [2:39:50<06:01, 4.24it/s] 97%|█████████▋| 46479/48008 [2:39:50<05:37, 4.53it/s] 97%|█████████▋| 46480/48008 [2:39:50<05:22, 4.73it/s] 97%|█████████▋| 46481/48008 [2:39:51<05:13, 4.87it/s] 97%|█████████▋| 46482/48008 [2:39:51<05:05, 4.99it/s] 97%|█████████▋| 46483/48008 [2:39:51<05:02, 5.04it/s] 97%|█████████▋| 46484/48008 [2:39:51<05:01, 5.06it/s] 97%|█████████▋| 46485/48008 [2:39:51<04:59, 5.08it/s] 97%|█████████▋| 46486/48008 [2:39:52<04:56, 5.14it/s] 97%|█████████▋| 46487/48008 [2:39:52<04:52, 5.19it/s] 97%|█████████▋| 46488/48008 [2:39:52<05:00, 5.06it/s] 97%|█████████▋| 46489/48008 [2:39:52<05:02, 5.03it/s] 97%|█████████▋| 46490/48008 [2:39:52<04:57, 5.10it/s] 97%|█████████▋| 46491/48008 [2:39:53<04:53, 5.17it/s] 97%|█████████▋| 46492/48008 [2:39:53<04:57, 5.10it/s] 97%|█████████▋| 46493/48008 [2:39:53<04:55, 5.13it/s] 97%|█████████▋| 46494/48008 [2:39:53<04:50, 5.21it/s] 97%|█████████▋| 46495/48008 [2:39:53<04:49, 5.23it/s] 97%|█████████▋| 46496/48008 [2:39:54<04:48, 5.24it/s] 97%|█████████▋| 46497/48008 [2:39:54<05:41, 4.43it/s] 97%|█████████▋| 46498/48008 [2:39:54<05:27, 4.61it/s] 97%|█████████▋| 46499/48008 [2:39:54<05:16, 4.77it/s] 97%|█████████▋| 46500/48008 [2:39:54<05:12, 4.83it/s] {'loss': 0.3002, 'grad_norm': 0.2373046875, 'learning_rate': 6.286452257957008e-06, 'epoch': 0.97} 97%|█████████▋| 46500/48008 [2:39:54<05:12, 4.83it/s] 97%|█████████▋| 46501/48008 [2:39:55<05:10, 4.86it/s] 97%|█████████▋| 46502/48008 [2:39:55<05:02, 4.98it/s] 97%|█████████▋| 46503/48008 [2:39:55<04:57, 5.06it/s] 97%|█████████▋| 46504/48008 [2:39:55<05:00, 5.01it/s] 97%|█████████▋| 46505/48008 [2:39:55<04:52, 5.13it/s] 97%|█████████▋| 46506/48008 [2:39:56<04:50, 5.18it/s] 97%|█████████▋| 46507/48008 [2:39:56<04:49, 5.19it/s] 97%|█████████▋| 46508/48008 [2:39:56<04:47, 5.22it/s] 97%|█████████▋| 46509/48008 [2:39:56<06:14, 4.00it/s] 97%|█████████▋| 46510/48008 [2:39:57<05:54, 4.23it/s] 97%|█████████▋| 46511/48008 [2:39:57<05:38, 4.42it/s] 97%|█████████▋| 46512/48008 [2:39:57<05:43, 4.36it/s] 97%|█████████▋| 46513/48008 [2:39:57<05:24, 4.60it/s] 97%|█████████▋| 46514/48008 [2:39:57<05:12, 4.78it/s] 97%|█████████▋| 46515/48008 [2:39:58<05:04, 4.90it/s] 97%|█████████▋| 46516/48008 [2:39:58<05:00, 4.96it/s] 97%|█████████▋| 46517/48008 [2:39:58<04:55, 5.05it/s] 97%|█████████▋| 46518/48008 [2:39:58<04:52, 5.09it/s] 97%|█████████▋| 46519/48008 [2:39:58<04:51, 5.10it/s] 97%|█████████▋| 46520/48008 [2:39:59<04:48, 5.16it/s] 97%|█████████▋| 46521/48008 [2:39:59<05:07, 4.84it/s] 97%|█████████▋| 46522/48008 [2:39:59<05:05, 4.86it/s] 97%|█████████▋| 46523/48008 [2:39:59<04:59, 4.97it/s] 97%|█████████▋| 46524/48008 [2:39:59<04:55, 5.02it/s] 97%|█████████▋| 46525/48008 [2:40:00<05:12, 4.74it/s] 97%|█████████▋| 46526/48008 [2:40:00<05:08, 4.80it/s] 97%|█████████▋| 46527/48008 [2:40:00<05:01, 4.91it/s] 97%|█████████▋| 46528/48008 [2:40:00<04:58, 4.96it/s] 97%|█████████▋| 46529/48008 [2:40:00<04:54, 5.03it/s] 97%|█████████▋| 46530/48008 [2:40:01<04:50, 5.09it/s] 97%|█████████▋| 46531/48008 [2:40:01<04:49, 5.11it/s] 97%|█████████▋| 46532/48008 [2:40:01<04:45, 5.18it/s] 97%|█████████▋| 46533/48008 [2:40:01<04:52, 5.04it/s] 97%|█████████▋| 46534/48008 [2:40:01<04:50, 5.08it/s] 97%|█████████▋| 46535/48008 [2:40:02<04:46, 5.14it/s] 97%|█████████▋| 46536/48008 [2:40:02<04:44, 5.18it/s] 97%|█████████▋| 46537/48008 [2:40:02<04:52, 5.04it/s] 97%|█████████▋| 46538/48008 [2:40:02<04:55, 4.97it/s] 97%|█████████▋| 46539/48008 [2:40:02<04:52, 5.03it/s] 97%|█████████▋| 46540/48008 [2:40:03<04:52, 5.01it/s] 97%|█████████▋| 46541/48008 [2:40:03<04:46, 5.12it/s] 97%|█████████▋| 46542/48008 [2:40:03<04:45, 5.13it/s] 97%|█████████▋| 46543/48008 [2:40:03<04:41, 5.20it/s] 97%|█████████▋| 46544/48008 [2:40:03<04:42, 5.17it/s] 97%|█████████▋| 46545/48008 [2:40:03<04:43, 5.16it/s] 97%|█████████▋| 46546/48008 [2:40:04<04:47, 5.09it/s] 97%|█████████▋| 46547/48008 [2:40:04<04:50, 5.03it/s] 97%|█████████▋| 46548/48008 [2:40:04<04:46, 5.09it/s] 97%|█████████▋| 46549/48008 [2:40:04<04:48, 5.06it/s] 97%|█████████▋| 46550/48008 [2:40:05<05:35, 4.34it/s] {'loss': 0.2976, 'grad_norm': 0.2041015625, 'learning_rate': 6.078153641059824e-06, 'epoch': 0.97} 97%|█████████▋| 46550/48008 [2:40:05<05:35, 4.34it/s] 97%|█████████▋| 46551/48008 [2:40:05<05:18, 4.57it/s] 97%|█████████▋| 46552/48008 [2:40:05<05:06, 4.75it/s] 97%|█████████▋| 46553/48008 [2:40:05<05:18, 4.57it/s] 97%|█████████▋| 46554/48008 [2:40:05<05:04, 4.77it/s] 97%|█████████▋| 46555/48008 [2:40:06<04:57, 4.88it/s] 97%|█████████▋| 46556/48008 [2:40:06<04:51, 4.99it/s] 97%|█████████▋| 46557/48008 [2:40:06<04:45, 5.09it/s] 97%|█████████▋| 46558/48008 [2:40:06<04:47, 5.04it/s] 97%|█████████▋| 46559/48008 [2:40:06<04:49, 5.01it/s] 97%|█████████▋| 46560/48008 [2:40:07<04:45, 5.06it/s] 97%|█████████▋| 46561/48008 [2:40:07<05:03, 4.76it/s] 97%|█████████▋| 46562/48008 [2:40:07<04:55, 4.90it/s] 97%|█████████▋| 46563/48008 [2:40:07<04:52, 4.95it/s] 97%|█████████▋| 46564/48008 [2:40:07<04:47, 5.02it/s] 97%|█████████▋| 46565/48008 [2:40:08<04:41, 5.13it/s] 97%|█████████▋| 46566/48008 [2:40:08<04:38, 5.17it/s] 97%|█████████▋| 46567/48008 [2:40:08<04:40, 5.13it/s] 97%|█████████▋| 46568/48008 [2:40:08<04:44, 5.06it/s] 97%|█████████▋| 46569/48008 [2:40:08<04:42, 5.10it/s] 97%|█████████▋| 46570/48008 [2:40:09<04:40, 5.13it/s] 97%|█████████▋| 46571/48008 [2:40:09<04:35, 5.21it/s] 97%|█████████▋| 46572/48008 [2:40:09<04:40, 5.11it/s] 97%|█████████▋| 46573/48008 [2:40:09<04:38, 5.15it/s] 97%|█████████▋| 46574/48008 [2:40:09<04:39, 5.13it/s] 97%|█████████▋| 46575/48008 [2:40:10<04:42, 5.07it/s] 97%|█████████▋| 46576/48008 [2:40:10<04:40, 5.11it/s] 97%|█████████▋| 46577/48008 [2:40:10<04:39, 5.12it/s] 97%|█████████▋| 46578/48008 [2:40:10<04:58, 4.79it/s] 97%|█████████▋| 46579/48008 [2:40:10<04:49, 4.94it/s] 97%|█████████▋| 46580/48008 [2:40:11<04:43, 5.04it/s] 97%|█████████▋| 46581/48008 [2:40:11<04:41, 5.07it/s] 97%|█████████▋| 46582/48008 [2:40:11<04:43, 5.03it/s] 97%|█████████▋| 46583/48008 [2:40:11<04:40, 5.09it/s] 97%|█████████▋| 46584/48008 [2:40:12<09:48, 2.42it/s] 97%|█████████▋| 46585/48008 [2:40:12<08:14, 2.88it/s] 97%|█████████▋| 46586/48008 [2:40:12<07:06, 3.34it/s] 97%|█████████▋| 46587/48008 [2:40:13<06:22, 3.72it/s] 97%|█████████▋| 46588/48008 [2:40:13<06:38, 3.57it/s] 97%|█████████▋| 46589/48008 [2:40:13<05:58, 3.96it/s] 97%|█████████▋| 46590/48008 [2:40:13<05:32, 4.26it/s] 97%|█████████▋| 46591/48008 [2:40:13<05:14, 4.51it/s] 97%|█████████▋| 46592/48008 [2:40:14<05:00, 4.71it/s] 97%|█████████▋| 46593/48008 [2:40:14<06:14, 3.78it/s] 97%|█████████▋| 46594/48008 [2:40:14<05:50, 4.03it/s] 97%|█████████▋| 46595/48008 [2:40:14<05:27, 4.32it/s] 97%|█████████▋| 46596/48008 [2:40:15<05:14, 4.49it/s] 97%|█████████▋| 46597/48008 [2:40:15<05:01, 4.69it/s] 97%|█████████▋| 46598/48008 [2:40:15<04:50, 4.85it/s] 97%|█████████▋| 46599/48008 [2:40:15<04:48, 4.88it/s] 97%|█████████▋| 46600/48008 [2:40:15<04:42, 4.98it/s] {'loss': 0.2999, 'grad_norm': 0.203125, 'learning_rate': 5.8698550241626395e-06, 'epoch': 0.97} 97%|█████████▋| 46600/48008 [2:40:15<04:42, 4.98it/s] 97%|█████████▋| 46601/48008 [2:40:16<04:37, 5.06it/s] 97%|█████████▋| 46602/48008 [2:40:16<04:36, 5.09it/s] 97%|█████████▋| 46603/48008 [2:40:16<04:38, 5.04it/s] 97%|█████████▋| 46604/48008 [2:40:16<04:40, 5.01it/s] 97%|█████████▋| 46605/48008 [2:40:16<04:40, 5.01it/s] 97%|█████████▋| 46606/48008 [2:40:17<04:35, 5.08it/s] 97%|█████████▋| 46607/48008 [2:40:17<04:39, 5.02it/s] 97%|█████████▋| 46608/48008 [2:40:17<04:37, 5.05it/s] 97%|█████████▋| 46609/48008 [2:40:17<04:38, 5.03it/s] 97%|█████████▋| 46610/48008 [2:40:17<04:39, 5.00it/s] 97%|█████████▋| 46611/48008 [2:40:18<04:35, 5.08it/s] 97%|█████████▋| 46612/48008 [2:40:18<04:31, 5.14it/s] 97%|█████████▋| 46613/48008 [2:40:18<04:28, 5.19it/s] 97%|█████████▋| 46614/48008 [2:40:18<04:36, 5.04it/s] 97%|█████████▋| 46615/48008 [2:40:18<04:33, 5.08it/s] 97%|█████████▋| 46616/48008 [2:40:19<04:32, 5.11it/s] 97%|█████████▋| 46617/48008 [2:40:19<04:29, 5.16it/s] 97%|█████████▋| 46618/48008 [2:40:19<04:34, 5.07it/s] 97%|█████████▋| 46619/48008 [2:40:19<04:36, 5.03it/s] 97%|█████████▋| 46620/48008 [2:40:19<04:35, 5.04it/s] 97%|█████████▋| 46621/48008 [2:40:20<04:32, 5.10it/s] 97%|█████████▋| 46622/48008 [2:40:20<04:29, 5.15it/s] 97%|█████████▋| 46623/48008 [2:40:20<05:48, 3.98it/s] 97%|█████████▋| 46624/48008 [2:40:20<05:24, 4.27it/s] 97%|█████████▋| 46625/48008 [2:40:21<05:09, 4.47it/s] 97%|█████████▋| 46626/48008 [2:40:21<04:55, 4.68it/s] 97%|█████████▋| 46627/48008 [2:40:21<04:46, 4.83it/s] 97%|█████████▋| 46628/48008 [2:40:21<04:39, 4.94it/s] 97%|█████████▋| 46629/48008 [2:40:21<04:39, 4.93it/s] 97%|█████████▋| 46630/48008 [2:40:22<04:39, 4.94it/s] 97%|█████████▋| 46631/48008 [2:40:22<04:38, 4.94it/s] 97%|█████████▋| 46632/48008 [2:40:22<04:32, 5.06it/s] 97%|█████████▋| 46633/48008 [2:40:22<04:29, 5.11it/s] 97%|█████████▋| 46634/48008 [2:40:22<04:27, 5.13it/s] 97%|█████████▋| 46635/48008 [2:40:23<04:28, 5.10it/s] 97%|█████████▋| 46636/48008 [2:40:23<04:34, 5.00it/s] 97%|█████████▋| 46637/48008 [2:40:23<04:29, 5.08it/s] 97%|█████████▋| 46638/48008 [2:40:23<04:29, 5.09it/s] 97%|█████████▋| 46639/48008 [2:40:23<04:26, 5.13it/s] 97%|█████████▋| 46640/48008 [2:40:23<04:25, 5.16it/s] 97%|█████████▋| 46641/48008 [2:40:24<04:23, 5.18it/s] 97%|█████████▋| 46642/48008 [2:40:24<04:22, 5.20it/s] 97%|█████████▋| 46643/48008 [2:40:24<04:21, 5.22it/s] 97%|█████████▋| 46644/48008 [2:40:24<04:19, 5.25it/s] 97%|█████████▋| 46645/48008 [2:40:24<04:23, 5.17it/s] 97%|█████████▋| 46646/48008 [2:40:25<04:22, 5.18it/s] 97%|█████████▋| 46647/48008 [2:40:25<04:21, 5.20it/s] 97%|█████████▋| 46648/48008 [2:40:25<04:20, 5.23it/s] 97%|█████████▋| 46649/48008 [2:40:25<04:19, 5.24it/s] 97%|█████████▋| 46650/48008 [2:40:25<04:23, 5.15it/s] {'loss': 0.2994, 'grad_norm': 0.1669921875, 'learning_rate': 5.661556407265456e-06, 'epoch': 0.97} 97%|█████████▋| 46650/48008 [2:40:25<04:23, 5.15it/s] 97%|█████████▋| 46651/48008 [2:40:27<10:47, 2.09it/s] 97%|█████████▋| 46652/48008 [2:40:27<08:50, 2.56it/s] 97%|█████████▋| 46653/48008 [2:40:27<07:30, 3.01it/s] 97%|█████████▋| 46654/48008 [2:40:27<06:52, 3.28it/s] 97%|█████████▋| 46655/48008 [2:40:27<06:06, 3.69it/s] 97%|█████████▋| 46656/48008 [2:40:28<05:38, 4.00it/s] 97%|█████████▋| 46657/48008 [2:40:28<05:16, 4.26it/s] 97%|█████████▋| 46658/48008 [2:40:28<04:58, 4.53it/s] 97%|█████████▋| 46659/48008 [2:40:28<06:04, 3.70it/s] 97%|█████████▋| 46660/48008 [2:40:29<05:32, 4.05it/s] 97%|█████████▋| 46661/48008 [2:40:29<05:11, 4.32it/s] 97%|█████████▋| 46662/48008 [2:40:29<04:55, 4.56it/s] 97%|█████████▋| 46663/48008 [2:40:30<10:09, 2.21it/s] 97%|█████████▋| 46664/48008 [2:40:30<08:22, 2.67it/s] 97%|█████████▋| 46665/48008 [2:40:30<07:09, 3.13it/s] 97%|█████████▋| 46666/48008 [2:40:30<06:19, 3.53it/s] 97%|█████████▋| 46667/48008 [2:40:31<05:42, 3.92it/s] 97%|█████████▋| 46668/48008 [2:40:31<05:15, 4.24it/s] 97%|█████████▋| 46669/48008 [2:40:32<12:01, 1.86it/s] 97%|█████████▋| 46670/48008 [2:40:32<10:28, 2.13it/s] 97%|█████████▋| 46671/48008 [2:40:33<08:36, 2.59it/s] 97%|█████████▋| 46672/48008 [2:40:33<07:17, 3.05it/s] 97%|█████████▋| 46673/48008 [2:40:33<06:24, 3.47it/s] 97%|█████████▋| 46674/48008 [2:40:33<05:50, 3.81it/s] 97%|█████████▋| 46675/48008 [2:40:33<05:23, 4.12it/s] 97%|█████████▋| 46676/48008 [2:40:34<05:01, 4.41it/s] 97%|█████████▋| 46677/48008 [2:40:34<04:51, 4.56it/s] 97%|█████████▋| 46678/48008 [2:40:34<04:46, 4.64it/s] 97%|█████████▋| 46679/48008 [2:40:34<04:36, 4.80it/s] 97%|█████████▋| 46680/48008 [2:40:34<04:27, 4.96it/s] 97%|█████████▋| 46681/48008 [2:40:35<04:23, 5.03it/s] 97%|█████████▋| 46682/48008 [2:40:35<04:23, 5.03it/s] 97%|█████████▋| 46683/48008 [2:40:35<04:20, 5.09it/s] 97%|█████████▋| 46684/48008 [2:40:35<04:16, 5.17it/s] 97%|█████████▋| 46685/48008 [2:40:35<04:19, 5.11it/s] 97%|█████████▋| 46686/48008 [2:40:36<04:17, 5.14it/s] 97%|█████████▋| 46687/48008 [2:40:36<04:14, 5.19it/s] 97%|█████████▋| 46688/48008 [2:40:36<04:16, 5.15it/s] 97%|█████████▋| 46689/48008 [2:40:36<04:13, 5.20it/s] 97%|█████████▋| 46690/48008 [2:40:36<04:14, 5.19it/s] 97%|█████████▋| 46691/48008 [2:40:37<04:13, 5.19it/s] 97%|█████████▋| 46692/48008 [2:40:37<04:10, 5.26it/s] 97%|█████████▋| 46693/48008 [2:40:37<04:56, 4.44it/s] 97%|█████████▋| 46694/48008 [2:40:37<04:45, 4.61it/s] 97%|█████████▋| 46695/48008 [2:40:37<04:40, 4.68it/s] 97%|█████████▋| 46696/48008 [2:40:38<04:30, 4.85it/s] 97%|█████████▋| 46697/48008 [2:40:38<04:22, 4.99it/s] 97%|█████████▋| 46698/48008 [2:40:38<04:18, 5.06it/s] 97%|█████████▋| 46699/48008 [2:40:38<04:16, 5.10it/s] 97%|█████████▋| 46700/48008 [2:40:38<04:17, 5.07it/s] {'loss': 0.298, 'grad_norm': 0.337890625, 'learning_rate': 5.453257790368272e-06, 'epoch': 0.97} 97%|█████████▋| 46700/48008 [2:40:38<04:17, 5.07it/s] 97%|█████████▋| 46701/48008 [2:40:39<04:15, 5.11it/s] 97%|█████████▋| 46702/48008 [2:40:39<04:12, 5.16it/s] 97%|█████████▋| 46703/48008 [2:40:39<04:10, 5.20it/s] 97%|█████████▋| 46704/48008 [2:40:39<04:08, 5.25it/s] 97%|█████████▋| 46705/48008 [2:40:39<04:10, 5.20it/s] 97%|█████████▋| 46706/48008 [2:40:40<04:13, 5.13it/s] 97%|█████████▋| 46707/48008 [2:40:40<04:12, 5.15it/s] 97%|█████████▋| 46708/48008 [2:40:40<04:09, 5.22it/s] 97%|█████████▋| 46709/48008 [2:40:40<04:13, 5.12it/s] 97%|█████████▋| 46710/48008 [2:40:40<04:11, 5.17it/s] 97%|█████████▋| 46711/48008 [2:40:40<04:17, 5.03it/s] 97%|█████████▋| 46712/48008 [2:40:41<04:15, 5.08it/s] 97%|█████████▋| 46713/48008 [2:40:41<05:28, 3.94it/s] 97%|█████████▋| 46714/48008 [2:40:41<05:05, 4.23it/s] 97%|█████████▋| 46715/48008 [2:40:41<04:51, 4.44it/s] 97%|█████████▋| 46716/48008 [2:40:42<04:38, 4.64it/s] 97%|█████████▋| 46717/48008 [2:40:42<04:36, 4.67it/s] 97%|█████████▋| 46718/48008 [2:40:42<04:27, 4.83it/s] 97%|█████████▋| 46719/48008 [2:40:42<04:17, 5.00it/s] 97%|█████████▋| 46720/48008 [2:40:42<04:14, 5.06it/s] 97%|█████████▋| 46721/48008 [2:40:43<04:11, 5.12it/s] 97%|█████████▋| 46722/48008 [2:40:43<04:26, 4.82it/s] 97%|█████████▋| 46723/48008 [2:40:43<04:18, 4.97it/s] 97%|█████████▋| 46724/48008 [2:40:43<04:15, 5.03it/s] 97%|█████████▋| 46725/48008 [2:40:43<04:10, 5.12it/s] 97%|█████████▋| 46726/48008 [2:40:44<04:15, 5.01it/s] 97%|█████████▋| 46727/48008 [2:40:44<04:12, 5.07it/s] 97%|█████████▋| 46728/48008 [2:40:44<04:10, 5.10it/s] 97%|█████████▋| 46729/48008 [2:40:44<04:12, 5.07it/s] 97%|█████████▋| 46730/48008 [2:40:44<04:09, 5.13it/s] 97%|█████████▋| 46731/48008 [2:40:45<04:10, 5.11it/s] 97%|█████████▋| 46732/48008 [2:40:45<04:12, 5.06it/s] 97%|█████████▋| 46733/48008 [2:40:45<04:10, 5.10it/s] 97%|█████████▋| 46734/48008 [2:40:45<04:08, 5.12it/s] 97%|█████████▋| 46735/48008 [2:40:46<08:17, 2.56it/s] 97%|█████████▋| 46736/48008 [2:40:46<07:00, 3.03it/s] 97%|█████████▋| 46737/48008 [2:40:46<06:24, 3.30it/s] 97%|█████████▋| 46738/48008 [2:40:47<05:42, 3.71it/s] 97%|█████████▋| 46739/48008 [2:40:47<05:12, 4.06it/s] 97%|█████████▋| 46740/48008 [2:40:47<04:53, 4.32it/s] 97%|█████████▋| 46741/48008 [2:40:47<04:37, 4.57it/s] 97%|█████████▋| 46742/48008 [2:40:47<04:44, 4.46it/s] 97%|█████████▋| 46743/48008 [2:40:48<04:31, 4.66it/s] 97%|█████████▋| 46744/48008 [2:40:48<04:23, 4.79it/s] 97%|█████████▋| 46745/48008 [2:40:48<04:34, 4.59it/s] 97%|█████████▋| 46746/48008 [2:40:48<04:24, 4.77it/s] 97%|█████████▋| 46747/48008 [2:40:48<04:17, 4.90it/s] 97%|█████████▋| 46748/48008 [2:40:49<04:10, 5.04it/s] 97%|█████████▋| 46749/48008 [2:40:49<04:12, 4.99it/s] 97%|█████████▋| 46750/48008 [2:40:49<04:09, 5.04it/s] {'loss': 0.2994, 'grad_norm': 0.2275390625, 'learning_rate': 5.244959173471088e-06, 'epoch': 0.97} 97%|█████████▋| 46750/48008 [2:40:49<04:09, 5.04it/s] 97%|█████████▋| 46751/48008 [2:40:49<04:05, 5.12it/s] 97%|█████████▋| 46752/48008 [2:40:49<04:04, 5.13it/s] 97%|█████████▋| 46753/48008 [2:40:50<03:59, 5.24it/s] 97%|█████████▋| 46754/48008 [2:40:50<04:16, 4.89it/s] 97%|█████████▋| 46755/48008 [2:40:50<04:14, 4.93it/s] 97%|█████████▋| 46756/48008 [2:40:50<04:12, 4.96it/s] 97%|█████████▋| 46757/48008 [2:40:50<04:08, 5.03it/s] 97%|█████████▋| 46758/48008 [2:40:51<04:04, 5.11it/s] 97%|█████████▋| 46759/48008 [2:40:51<04:07, 5.05it/s] 97%|█████████▋| 46760/48008 [2:40:51<04:12, 4.95it/s] 97%|█████████▋| 46761/48008 [2:40:51<04:13, 4.92it/s] 97%|█████████▋| 46762/48008 [2:40:51<04:09, 4.99it/s] 97%|█████████▋| 46763/48008 [2:40:52<04:09, 4.98it/s] 97%|█████████▋| 46764/48008 [2:40:52<04:04, 5.09it/s] 97%|█████████▋| 46765/48008 [2:40:52<04:45, 4.35it/s] 97%|█████████▋| 46766/48008 [2:40:52<04:33, 4.55it/s] 97%|█████████▋| 46767/48008 [2:40:53<04:26, 4.66it/s] 97%|█████████▋| 46768/48008 [2:40:53<04:17, 4.81it/s] 97%|█████████▋| 46769/48008 [2:40:53<04:13, 4.89it/s] 97%|█████████▋| 46770/48008 [2:40:53<04:25, 4.67it/s] 97%|█████████▋| 46771/48008 [2:40:53<04:17, 4.81it/s] 97%|█████████▋| 46772/48008 [2:40:54<04:09, 4.94it/s] 97%|█████████▋| 46773/48008 [2:40:54<04:04, 5.05it/s] 97%|█████████▋| 46774/48008 [2:40:54<03:59, 5.15it/s] 97%|█████████▋| 46775/48008 [2:40:54<04:15, 4.83it/s] 97%|█████████▋| 46776/48008 [2:40:55<05:21, 3.84it/s] 97%|█████████▋| 46777/48008 [2:40:55<04:55, 4.17it/s] 97%|█████████▋| 46778/48008 [2:40:55<04:37, 4.43it/s] 97%|█████████▋| 46779/48008 [2:40:55<04:25, 4.63it/s] 97%|█████████▋| 46780/48008 [2:40:55<04:17, 4.77it/s] 97%|█████████▋| 46781/48008 [2:40:56<04:53, 4.18it/s] 97%|█████████▋| 46782/48008 [2:40:56<04:52, 4.19it/s] 97%|█████████▋| 46783/48008 [2:40:56<04:35, 4.44it/s] 97%|█████████▋| 46784/48008 [2:40:56<04:27, 4.57it/s] 97%|█████████▋| 46785/48008 [2:40:56<04:18, 4.73it/s] 97%|█████████▋| 46786/48008 [2:40:57<04:28, 4.55it/s] 97%|█████████▋| 46787/48008 [2:40:57<04:18, 4.73it/s] 97%|█████████▋| 46788/48008 [2:40:57<04:14, 4.79it/s] 97%|█████████▋| 46789/48008 [2:40:57<04:08, 4.90it/s] 97%|█████████▋| 46790/48008 [2:40:58<04:07, 4.92it/s] 97%|█████████▋| 46791/48008 [2:40:58<04:02, 5.02it/s] 97%|█████████▋| 46792/48008 [2:40:58<04:01, 5.03it/s] 97%|█████████▋| 46793/48008 [2:40:58<03:56, 5.13it/s] 97%|█████████▋| 46794/48008 [2:40:58<04:12, 4.81it/s] 97%|█████████▋| 46795/48008 [2:40:59<04:06, 4.92it/s] 97%|█████████▋| 46796/48008 [2:40:59<04:01, 5.03it/s] 97%|█████████▋| 46797/48008 [2:40:59<03:58, 5.08it/s] 97%|█████████▋| 46798/48008 [2:40:59<04:00, 5.03it/s] 97%|█████████▋| 46799/48008 [2:40:59<03:57, 5.09it/s] 97%|█████████▋| 46800/48008 [2:40:59<03:55, 5.13it/s] 97%|█████████▋| 46800/48008 [2:40:59<03:55, 5.13it/s]{'loss': 0.298, 'grad_norm': 0.63671875, 'learning_rate': 5.036660556573905e-06, 'epoch': 0.97} 97%|█████████▋| 46801/48008 [2:41:00<03:54, 5.15it/s] 97%|█████████▋| 46802/48008 [2:41:00<03:54, 5.15it/s] 97%|█████████▋| 46803/48008 [2:41:00<03:57, 5.07it/s] 97%|█████████▋| 46804/48008 [2:41:00<03:53, 5.15it/s] 97%|█████████▋| 46805/48008 [2:41:00<04:10, 4.80it/s] 97%|█████████▋| 46806/48008 [2:41:01<04:04, 4.91it/s] 97%|█████████▋| 46807/48008 [2:41:01<03:59, 5.02it/s] 98%|█████████▊| 46808/48008 [2:41:01<03:56, 5.07it/s] 98%|█████████▊| 46809/48008 [2:41:01<03:55, 5.10it/s] 98%|█████████▊| 46810/48008 [2:41:01<03:52, 5.15it/s] 98%|█████████▊| 46811/48008 [2:41:02<03:53, 5.13it/s] 98%|█████████▊| 46812/48008 [2:41:02<03:52, 5.14it/s] 98%|█████████▊| 46813/48008 [2:41:02<03:50, 5.19it/s] 98%|█████████▊| 46814/48008 [2:41:02<03:49, 5.21it/s] 98%|█████████▊| 46815/48008 [2:41:02<03:52, 5.13it/s] 98%|█████████▊| 46816/48008 [2:41:03<03:51, 5.14it/s] 98%|█████████▊| 46817/48008 [2:41:03<03:51, 5.15it/s] 98%|█████████▊| 46818/48008 [2:41:03<03:50, 5.17it/s] 98%|█████████▊| 46819/48008 [2:41:03<03:51, 5.13it/s] 98%|█████████▊| 46820/48008 [2:41:03<03:49, 5.18it/s] 98%|█████████▊| 46821/48008 [2:41:04<03:48, 5.19it/s] 98%|█████████▊| 46822/48008 [2:41:04<03:53, 5.09it/s] 98%|█████████▊| 46823/48008 [2:41:04<05:01, 3.94it/s] 98%|█████████▊| 46824/48008 [2:41:04<04:40, 4.22it/s] 98%|█████████▊| 46825/48008 [2:41:05<04:41, 4.21it/s] 98%|█████████▊| 46826/48008 [2:41:05<04:22, 4.50it/s] 98%|█████████▊| 46827/48008 [2:41:05<04:08, 4.75it/s] 98%|█████████▊| 46828/48008 [2:41:05<03:59, 4.92it/s] 98%|█████████▊| 46829/48008 [2:41:05<03:56, 4.99it/s] 98%|█████████▊| 46830/48008 [2:41:06<03:53, 5.04it/s] 98%|█████████▊| 46831/48008 [2:41:06<03:51, 5.08it/s] 98%|█████████▊| 46832/48008 [2:41:06<03:53, 5.03it/s] 98%|█████████▊| 46833/48008 [2:41:06<03:50, 5.09it/s] 98%|█████████▊| 46834/48008 [2:41:06<03:48, 5.13it/s] 98%|█████████▊| 46835/48008 [2:41:07<03:45, 5.21it/s] 98%|█████████▊| 46836/48008 [2:41:07<04:25, 4.41it/s] 98%|█████████▊| 46837/48008 [2:41:07<04:12, 4.64it/s] 98%|█████████▊| 46838/48008 [2:41:07<04:04, 4.78it/s] 98%|█████████▊| 46839/48008 [2:41:07<04:14, 4.59it/s] 98%|█████████▊| 46840/48008 [2:41:08<04:05, 4.76it/s] 98%|█████████▊| 46841/48008 [2:41:08<04:03, 4.80it/s] 98%|█████████▊| 46842/48008 [2:41:08<03:57, 4.90it/s] 98%|█████████▊| 46843/48008 [2:41:08<03:55, 4.95it/s] 98%|█████████▊| 46844/48008 [2:41:08<03:56, 4.93it/s] 98%|█████████▊| 46845/48008 [2:41:09<03:50, 5.06it/s] 98%|█████████▊| 46846/48008 [2:41:09<03:55, 4.94it/s] 98%|█████████▊| 46847/48008 [2:41:09<03:51, 5.02it/s] 98%|█████████▊| 46848/48008 [2:41:09<03:48, 5.08it/s] 98%|█████████▊| 46849/48008 [2:41:09<03:44, 5.16it/s] 98%|█████████▊| 46850/48008 [2:41:10<03:43, 5.17it/s] {'loss': 0.3001, 'grad_norm': 0.2177734375, 'learning_rate': 4.82836193967672e-06, 'epoch': 0.98} 98%|█████████▊| 46850/48008 [2:41:10<03:43, 5.17it/s] 98%|█████████▊| 46851/48008 [2:41:10<03:43, 5.18it/s] 98%|█████████▊| 46852/48008 [2:41:10<03:43, 5.18it/s] 98%|█████████▊| 46853/48008 [2:41:10<03:46, 5.11it/s] 98%|█████████▊| 46854/48008 [2:41:10<03:44, 5.14it/s] 98%|█████████▊| 46855/48008 [2:41:11<03:43, 5.17it/s] 98%|█████████▊| 46856/48008 [2:41:11<03:44, 5.13it/s] 98%|█████████▊| 46857/48008 [2:41:11<03:40, 5.21it/s] 98%|█████████▊| 46858/48008 [2:41:11<03:40, 5.22it/s] 98%|█████████▊| 46859/48008 [2:41:11<03:40, 5.22it/s] 98%|█████████▊| 46860/48008 [2:41:12<03:40, 5.22it/s] 98%|█████████▊| 46861/48008 [2:41:12<03:40, 5.20it/s] 98%|█████████▊| 46862/48008 [2:41:12<03:45, 5.09it/s] 98%|█████████▊| 46863/48008 [2:41:12<03:42, 5.15it/s] 98%|█████████▊| 46864/48008 [2:41:12<03:41, 5.16it/s] 98%|█████████▊| 46865/48008 [2:41:13<03:41, 5.17it/s] 98%|█████████▊| 46866/48008 [2:41:13<03:39, 5.20it/s] 98%|█████████▊| 46867/48008 [2:41:13<03:37, 5.24it/s] 98%|█████████▊| 46868/48008 [2:41:13<03:36, 5.26it/s] 98%|█████████▊| 46869/48008 [2:41:13<03:36, 5.25it/s] 98%|█████████▊| 46870/48008 [2:41:13<03:36, 5.25it/s] 98%|█████████▊| 46871/48008 [2:41:14<03:37, 5.24it/s] 98%|█████████▊| 46872/48008 [2:41:14<03:37, 5.21it/s] 98%|█████████▊| 46873/48008 [2:41:14<03:40, 5.14it/s] 98%|█████████▊| 46874/48008 [2:41:14<03:43, 5.08it/s] 98%|█████████▊| 46875/48008 [2:41:14<03:40, 5.15it/s] 98%|█████████▊| 46876/48008 [2:41:15<03:38, 5.18it/s] 98%|█████████▊| 46877/48008 [2:41:15<03:38, 5.18it/s] 98%|█████████▊| 46878/48008 [2:41:15<03:52, 4.87it/s] 98%|█████████▊| 46879/48008 [2:41:15<03:46, 4.98it/s] 98%|█████████▊| 46880/48008 [2:41:15<03:46, 4.99it/s] 98%|█████████▊| 46881/48008 [2:41:16<03:42, 5.07it/s] 98%|█████████▊| 46882/48008 [2:41:16<03:40, 5.11it/s] 98%|█████████▊| 46883/48008 [2:41:16<03:39, 5.12it/s] 98%|█████████▊| 46884/48008 [2:41:16<03:38, 5.14it/s] 98%|█████████▊| 46885/48008 [2:41:16<03:37, 5.17it/s] 98%|█████████▊| 46886/48008 [2:41:17<03:36, 5.19it/s] 98%|█████████▊| 46887/48008 [2:41:17<03:35, 5.19it/s] 98%|█████████▊| 46888/48008 [2:41:17<03:38, 5.12it/s] 98%|█████████▊| 46889/48008 [2:41:17<03:36, 5.17it/s] 98%|█████████▊| 46890/48008 [2:41:17<03:38, 5.11it/s] 98%|█████████▊| 46891/48008 [2:41:18<03:36, 5.17it/s] 98%|█████████▊| 46892/48008 [2:41:18<03:34, 5.19it/s] 98%|█████████▊| 46893/48008 [2:41:18<03:34, 5.19it/s] 98%|█████████▊| 46894/48008 [2:41:18<03:33, 5.21it/s] 98%|█████████▊| 46895/48008 [2:41:18<03:34, 5.19it/s] 98%|█████████▊| 46896/48008 [2:41:19<03:33, 5.20it/s] 98%|█████████▊| 46897/48008 [2:41:19<03:31, 5.24it/s] 98%|█████████▊| 46898/48008 [2:41:19<03:34, 5.17it/s] 98%|█████████▊| 46899/48008 [2:41:19<03:32, 5.22it/s] 98%|█████████▊| 46900/48008 [2:41:19<03:32, 5.22it/s] 98%|█████████▊| 46900/48008 [2:41:19<03:32, 5.22it/s]{'loss': 0.3004, 'grad_norm': 0.2470703125, 'learning_rate': 4.620063322779537e-06, 'epoch': 0.98} 98%|█████████▊| 46901/48008 [2:41:19<03:30, 5.25it/s] 98%|█████████▊| 46902/48008 [2:41:20<03:33, 5.19it/s] 98%|█████████▊| 46903/48008 [2:41:20<03:34, 5.15it/s] 98%|█████████▊| 46904/48008 [2:41:20<04:37, 3.98it/s] 98%|█████████▊| 46905/48008 [2:41:20<04:32, 4.04it/s] 98%|█████████▊| 46906/48008 [2:41:21<04:14, 4.33it/s] 98%|█████████▊| 46907/48008 [2:41:21<04:04, 4.51it/s] 98%|█████████▊| 46908/48008 [2:41:21<04:00, 4.58it/s] 98%|█████████▊| 46909/48008 [2:41:21<03:49, 4.79it/s] 98%|█████████▊| 46910/48008 [2:41:21<03:42, 4.94it/s] 98%|█████████▊| 46911/48008 [2:41:22<03:38, 5.02it/s] 98%|█████████▊| 46912/48008 [2:41:22<03:32, 5.17it/s] 98%|█████████▊| 46913/48008 [2:41:22<03:31, 5.19it/s] 98%|█████████▊| 46914/48008 [2:41:22<04:08, 4.41it/s] 98%|█████████▊| 46915/48008 [2:41:23<03:55, 4.64it/s] 98%|█████████▊| 46916/48008 [2:41:23<03:48, 4.78it/s] 98%|█████████▊| 46917/48008 [2:41:23<03:44, 4.86it/s] 98%|█████████▊| 46918/48008 [2:41:23<03:40, 4.95it/s] 98%|█████████▊| 46919/48008 [2:41:23<03:40, 4.95it/s] 98%|█████████▊| 46920/48008 [2:41:24<03:36, 5.03it/s] 98%|█████████▊| 46921/48008 [2:41:24<03:33, 5.09it/s] 98%|█████████▊| 46922/48008 [2:41:24<03:34, 5.05it/s] 98%|█████████▊| 46923/48008 [2:41:24<03:31, 5.13it/s] 98%|█████████▊| 46924/48008 [2:41:24<03:30, 5.16it/s] 98%|█████████▊| 46925/48008 [2:41:24<03:29, 5.16it/s] 98%|█████████▊| 46926/48008 [2:41:25<03:33, 5.08it/s] 98%|█████████▊| 46927/48008 [2:41:25<03:30, 5.12it/s] 98%|█████████▊| 46928/48008 [2:41:25<03:32, 5.08it/s] 98%|█████████▊| 46929/48008 [2:41:25<03:29, 5.16it/s] 98%|█████████▊| 46930/48008 [2:41:25<03:27, 5.19it/s] 98%|█████████▊| 46931/48008 [2:41:26<03:29, 5.14it/s] 98%|█████████▊| 46932/48008 [2:41:26<03:31, 5.09it/s] 98%|█████████▊| 46933/48008 [2:41:26<03:30, 5.11it/s] 98%|█████████▊| 46934/48008 [2:41:26<03:27, 5.17it/s] 98%|█████████▊| 46935/48008 [2:41:26<03:27, 5.17it/s] 98%|█████████▊| 46936/48008 [2:41:27<03:27, 5.17it/s] 98%|█████████▊| 46937/48008 [2:41:27<03:25, 5.20it/s] 98%|█████████▊| 46938/48008 [2:41:27<03:24, 5.22it/s] 98%|█████████▊| 46939/48008 [2:41:27<03:39, 4.88it/s] 98%|█████████▊| 46940/48008 [2:41:27<03:33, 5.01it/s] 98%|█████████▊| 46941/48008 [2:41:28<03:26, 5.17it/s] 98%|█████████▊| 46942/48008 [2:41:28<03:24, 5.20it/s] 98%|█████████▊| 46943/48008 [2:41:28<03:24, 5.21it/s] 98%|█████████▊| 46944/48008 [2:41:28<03:24, 5.21it/s] 98%|█████████▊| 46945/48008 [2:41:28<03:25, 5.18it/s] 98%|█████████▊| 46946/48008 [2:41:29<03:29, 5.07it/s] 98%|█████████▊| 46947/48008 [2:41:29<03:26, 5.15it/s] 98%|█████████▊| 46948/48008 [2:41:29<03:28, 5.09it/s] 98%|█████████▊| 46949/48008 [2:41:29<03:26, 5.14it/s] 98%|█████████▊| 46950/48008 [2:41:29<03:29, 5.06it/s] 98%|█████████▊| 46950/48008 [2:41:29<03:29, 5.06it/s]{'loss': 0.3002, 'grad_norm': 0.2470703125, 'learning_rate': 4.411764705882353e-06, 'epoch': 0.98} 98%|█████████▊| 46951/48008 [2:41:30<03:28, 5.08it/s] 98%|█████████▊| 46952/48008 [2:41:30<03:30, 5.02it/s] 98%|█████████▊| 46953/48008 [2:41:30<03:27, 5.09it/s] 98%|█████████▊| 46954/48008 [2:41:30<03:24, 5.15it/s] 98%|█████████▊| 46955/48008 [2:41:30<03:29, 5.01it/s] 98%|█████████▊| 46956/48008 [2:41:31<03:29, 5.03it/s] 98%|█████████▊| 46957/48008 [2:41:31<03:26, 5.08it/s] 98%|█████████▊| 46958/48008 [2:41:31<03:24, 5.12it/s] 98%|█████████▊| 46959/48008 [2:41:31<03:22, 5.17it/s] 98%|█████████▊| 46960/48008 [2:41:31<03:23, 5.16it/s] 98%|█████████▊| 46961/48008 [2:41:32<03:25, 5.10it/s] 98%|█████████▊| 46962/48008 [2:41:32<03:23, 5.14it/s] 98%|█████████▊| 46963/48008 [2:41:32<03:22, 5.16it/s] 98%|█████████▊| 46964/48008 [2:41:32<03:22, 5.16it/s] 98%|█████████▊| 46965/48008 [2:41:32<03:20, 5.19it/s] 98%|█████████▊| 46966/48008 [2:41:32<03:19, 5.24it/s] 98%|█████████▊| 46967/48008 [2:41:33<03:18, 5.25it/s] 98%|█████████▊| 46968/48008 [2:41:33<03:33, 4.88it/s] 98%|█████████▊| 46969/48008 [2:41:33<03:28, 4.99it/s] 98%|█████████▊| 46970/48008 [2:41:33<03:25, 5.06it/s] 98%|█████████▊| 46971/48008 [2:41:33<03:30, 4.93it/s] 98%|█████████▊| 46972/48008 [2:41:34<03:24, 5.07it/s] 98%|█████████▊| 46973/48008 [2:41:34<03:22, 5.10it/s] 98%|█████████▊| 46974/48008 [2:41:34<03:23, 5.09it/s] 98%|█████████▊| 46975/48008 [2:41:34<03:27, 4.97it/s] 98%|█████████▊| 46976/48008 [2:41:35<04:24, 3.90it/s] 98%|█████████▊| 46977/48008 [2:41:35<04:05, 4.21it/s] 98%|█████████▊| 46978/48008 [2:41:35<03:53, 4.40it/s] 98%|█████████▊| 46979/48008 [2:41:35<03:43, 4.61it/s] 98%|█████████▊| 46980/48008 [2:41:35<03:34, 4.80it/s] 98%|█████████▊| 46981/48008 [2:41:36<03:30, 4.87it/s] 98%|█████████▊| 46982/48008 [2:41:36<03:30, 4.88it/s] 98%|█████████▊| 46983/48008 [2:41:36<03:23, 5.04it/s] 98%|█████████▊| 46984/48008 [2:41:36<03:35, 4.75it/s] 98%|█████████▊| 46985/48008 [2:41:36<03:30, 4.86it/s] 98%|█████████▊| 46986/48008 [2:41:37<03:28, 4.89it/s] 98%|█████████▊| 46987/48008 [2:41:37<03:27, 4.92it/s] 98%|█████████▊| 46988/48008 [2:41:37<03:25, 4.97it/s] 98%|█████████▊| 46989/48008 [2:41:37<03:27, 4.91it/s] 98%|█████████▊| 46990/48008 [2:41:37<03:22, 5.03it/s] 98%|█████████▊| 46991/48008 [2:41:38<03:33, 4.76it/s] 98%|█████████▊| 46992/48008 [2:41:38<03:28, 4.87it/s] 98%|█████████▊| 46993/48008 [2:41:38<03:31, 4.81it/s] 98%|█████████▊| 46994/48008 [2:41:38<03:28, 4.86it/s] 98%|█████████▊| 46995/48008 [2:41:38<03:25, 4.94it/s] 98%|█████████▊| 46996/48008 [2:41:39<03:21, 5.03it/s] 98%|█████████▊| 46997/48008 [2:41:39<03:18, 5.08it/s] 98%|█████████▊| 46998/48008 [2:41:39<03:17, 5.13it/s] 98%|█████████▊| 46999/48008 [2:41:39<03:15, 5.17it/s] 98%|█████████▊| 47000/48008 [2:41:39<03:14, 5.18it/s] {'loss': 0.2992, 'grad_norm': 0.380859375, 'learning_rate': 4.20346608898517e-06, 'epoch': 0.98} 98%|█████████▊| 47000/48008 [2:41:39<03:14, 5.18it/s] 98%|█████████▊| 47001/48008 [2:41:40<03:14, 5.17it/s] 98%|█████████▊| 47002/48008 [2:41:40<03:18, 5.07it/s] 98%|█████████▊| 47003/48008 [2:41:40<03:16, 5.12it/s] 98%|█████████▊| 47004/48008 [2:41:40<03:15, 5.12it/s] 98%|█████████▊| 47005/48008 [2:41:40<03:14, 5.17it/s] 98%|█████████▊| 47006/48008 [2:41:41<03:14, 5.16it/s] 98%|█████████▊| 47007/48008 [2:41:41<03:13, 5.18it/s] 98%|█████████▊| 47008/48008 [2:41:41<03:13, 5.17it/s] 98%|█████████▊| 47009/48008 [2:41:41<03:18, 5.04it/s] 98%|█████████▊| 47010/48008 [2:41:41<03:17, 5.06it/s] 98%|█████████▊| 47011/48008 [2:41:42<03:13, 5.14it/s] 98%|█████████▊| 47012/48008 [2:41:42<03:12, 5.18it/s] 98%|█████████▊| 47013/48008 [2:41:42<03:10, 5.22it/s] 98%|█████████▊| 47014/48008 [2:41:42<03:10, 5.21it/s] 98%|█████████▊| 47015/48008 [2:41:42<03:13, 5.13it/s] 98%|█████████▊| 47016/48008 [2:41:43<03:25, 4.82it/s] 98%|█████████▊| 47017/48008 [2:41:43<03:20, 4.95it/s] 98%|█████████▊| 47018/48008 [2:41:43<03:31, 4.69it/s] 98%|█████████▊| 47019/48008 [2:41:43<03:29, 4.72it/s] 98%|█████████▊| 47020/48008 [2:41:43<03:27, 4.76it/s] 98%|█████████▊| 47021/48008 [2:41:44<03:22, 4.88it/s] 98%|█████████▊| 47022/48008 [2:41:44<03:17, 5.00it/s] 98%|█████████▊| 47023/48008 [2:41:44<03:14, 5.07it/s] 98%|█████████▊| 47024/48008 [2:41:44<03:12, 5.11it/s] 98%|█████████▊| 47025/48008 [2:41:44<03:12, 5.12it/s] 98%|█████████▊| 47026/48008 [2:41:45<03:14, 5.06it/s] 98%|█████████▊| 47027/48008 [2:41:45<03:11, 5.14it/s] 98%|█████████▊| 47028/48008 [2:41:45<03:09, 5.18it/s] 98%|█████████▊| 47029/48008 [2:41:45<03:10, 5.15it/s] 98%|█████████▊| 47030/48008 [2:41:45<03:08, 5.20it/s] 98%|█████████▊| 47031/48008 [2:41:46<03:06, 5.24it/s] 98%|█████████▊| 47032/48008 [2:41:46<03:05, 5.27it/s] 98%|█████████▊| 47033/48008 [2:41:46<03:05, 5.27it/s] 98%|█████████▊| 47034/48008 [2:41:46<03:11, 5.10it/s] 98%|█████████▊| 47035/48008 [2:41:47<04:06, 3.95it/s] 98%|█████████▊| 47036/48008 [2:41:47<03:48, 4.26it/s] 98%|█████████▊| 47037/48008 [2:41:47<03:32, 4.56it/s] 98%|█████████▊| 47038/48008 [2:41:47<03:24, 4.73it/s] 98%|█████████▊| 47039/48008 [2:41:47<03:21, 4.82it/s] 98%|█████████▊| 47040/48008 [2:41:47<03:17, 4.91it/s] 98%|█████████▊| 47041/48008 [2:41:48<04:09, 3.87it/s] 98%|█████████▊| 47042/48008 [2:41:48<03:51, 4.18it/s] 98%|█████████▊| 47043/48008 [2:41:48<03:35, 4.47it/s] 98%|█████████▊| 47044/48008 [2:41:48<03:25, 4.69it/s] 98%|█████████▊| 47045/48008 [2:41:49<03:17, 4.87it/s] 98%|█████████▊| 47046/48008 [2:41:49<03:11, 5.02it/s] 98%|█████████▊| 47047/48008 [2:41:49<03:14, 4.95it/s] 98%|█████████▊| 47048/48008 [2:41:49<03:11, 5.03it/s] 98%|█████████▊| 47049/48008 [2:41:49<03:07, 5.12it/s] 98%|█████████▊| 47050/48008 [2:41:50<03:05, 5.15it/s] {'loss': 0.3005, 'grad_norm': 0.23046875, 'learning_rate': 3.995167472087986e-06, 'epoch': 0.98} 98%|█████████▊| 47050/48008 [2:41:50<03:05, 5.15it/s] 98%|█████████▊| 47051/48008 [2:41:50<03:08, 5.07it/s] 98%|█████████▊| 47052/48008 [2:41:50<03:06, 5.12it/s] 98%|█████████▊| 47053/48008 [2:41:50<03:09, 5.05it/s] 98%|█████████▊| 47054/48008 [2:41:50<03:10, 4.99it/s] 98%|█████████▊| 47055/48008 [2:41:51<03:09, 5.03it/s] 98%|█████████▊| 47056/48008 [2:41:51<03:08, 5.04it/s] 98%|█████████▊| 47057/48008 [2:41:51<03:10, 4.99it/s] 98%|█████████▊| 47058/48008 [2:41:51<03:09, 5.01it/s] 98%|█████████▊| 47059/48008 [2:41:51<03:10, 4.99it/s] 98%|█████████▊| 47060/48008 [2:41:52<03:12, 4.93it/s] 98%|█████████▊| 47061/48008 [2:41:52<03:22, 4.67it/s] 98%|█████████▊| 47062/48008 [2:41:52<03:19, 4.74it/s] 98%|█████████▊| 47063/48008 [2:41:52<03:19, 4.75it/s] 98%|█████████▊| 47064/48008 [2:41:52<03:16, 4.80it/s] 98%|█████████▊| 47065/48008 [2:41:53<03:12, 4.90it/s] 98%|█████████▊| 47066/48008 [2:41:53<03:11, 4.92it/s] 98%|█████████▊| 47067/48008 [2:41:53<03:07, 5.02it/s] 98%|█████████▊| 47068/48008 [2:41:53<03:17, 4.76it/s] 98%|█████████▊| 47069/48008 [2:41:53<03:12, 4.88it/s] 98%|█████████▊| 47070/48008 [2:41:54<03:07, 5.00it/s] 98%|█████████▊| 47071/48008 [2:41:54<03:07, 5.00it/s] 98%|█████████▊| 47072/48008 [2:41:54<03:04, 5.07it/s] 98%|█████████▊| 47073/48008 [2:41:54<03:02, 5.13it/s] 98%|█████████▊| 47074/48008 [2:41:54<03:00, 5.17it/s] 98%|█████████▊| 47075/48008 [2:41:55<03:02, 5.11it/s] 98%|█████████▊| 47076/48008 [2:41:55<03:00, 5.16it/s] 98%|█████████▊| 47077/48008 [2:41:55<02:58, 5.22it/s] 98%|█████████▊| 47078/48008 [2:41:55<02:57, 5.24it/s] 98%|█████████▊| 47079/48008 [2:41:55<02:57, 5.23it/s] 98%|█████████▊| 47080/48008 [2:41:56<02:57, 5.23it/s] 98%|█████████▊| 47081/48008 [2:41:56<02:59, 5.17it/s] 98%|█████████▊| 47082/48008 [2:41:56<02:58, 5.19it/s] 98%|█████████▊| 47083/48008 [2:41:56<02:57, 5.22it/s] 98%|█████████▊| 47084/48008 [2:41:56<02:56, 5.24it/s] 98%|█████████▊| 47085/48008 [2:41:57<02:59, 5.15it/s] 98%|█████████▊| 47086/48008 [2:41:57<02:57, 5.19it/s] 98%|█████████▊| 47087/48008 [2:41:57<02:57, 5.18it/s] 98%|█████████▊| 47088/48008 [2:41:57<02:57, 5.18it/s] 98%|█████████▊| 47089/48008 [2:41:57<02:57, 5.17it/s] 98%|█████████▊| 47090/48008 [2:41:58<02:56, 5.20it/s] 98%|█████████▊| 47091/48008 [2:41:58<03:00, 5.08it/s] 98%|█████████▊| 47092/48008 [2:41:58<02:59, 5.11it/s] 98%|█████████▊| 47093/48008 [2:41:58<03:01, 5.03it/s] 98%|█████████▊| 47094/48008 [2:41:59<03:52, 3.93it/s] 98%|█████████▊| 47095/48008 [2:41:59<03:35, 4.23it/s] 98%|█████████▊| 47096/48008 [2:41:59<03:26, 4.42it/s] 98%|█████████▊| 47097/48008 [2:41:59<03:16, 4.64it/s] 98%|█████████▊| 47098/48008 [2:41:59<03:09, 4.80it/s] 98%|█████████▊| 47099/48008 [2:41:59<03:09, 4.81it/s] 98%|█████████▊| 47100/48008 [2:42:00<03:02, 4.97it/s] 98%|█████████▊| 47100/48008 [2:42:00<03:02, 4.97it/s]{'loss': 0.2987, 'grad_norm': 0.30078125, 'learning_rate': 3.7868688551908017e-06, 'epoch': 0.98} 98%|█████████▊| 47101/48008 [2:42:00<02:59, 5.04it/s] 98%|█████████▊| 47102/48008 [2:42:00<02:57, 5.09it/s] 98%|█████████▊| 47103/48008 [2:42:00<03:02, 4.96it/s] 98%|█████████▊| 47104/48008 [2:42:00<02:58, 5.06it/s] 98%|█████████▊| 47105/48008 [2:42:01<03:01, 4.99it/s] 98%|█████████▊| 47106/48008 [2:42:01<03:01, 4.98it/s] 98%|█████████▊| 47107/48008 [2:42:01<03:03, 4.90it/s] 98%|█████████▊| 47108/48008 [2:42:01<03:31, 4.26it/s] 98%|█████████▊| 47109/48008 [2:42:02<03:23, 4.43it/s] 98%|█████████▊| 47110/48008 [2:42:02<03:19, 4.51it/s] 98%|█████████▊| 47111/48008 [2:42:02<03:10, 4.71it/s] 98%|█████████▊| 47112/48008 [2:42:02<03:06, 4.81it/s] 98%|█████████▊| 47113/48008 [2:42:02<03:02, 4.91it/s] 98%|█████████▊| 47114/48008 [2:42:03<03:11, 4.66it/s] 98%|█████████▊| 47115/48008 [2:42:03<03:05, 4.80it/s] 98%|█████████▊| 47116/48008 [2:42:03<03:01, 4.91it/s] 98%|█████████▊| 47117/48008 [2:42:03<02:59, 4.96it/s] 98%|█████████▊| 47118/48008 [2:42:03<02:56, 5.04it/s] 98%|█████████▊| 47119/48008 [2:42:04<02:54, 5.11it/s] 98%|█████████▊| 47120/48008 [2:42:04<03:05, 4.78it/s] 98%|█████████▊| 47121/48008 [2:42:04<03:01, 4.89it/s] 98%|█████████▊| 47122/48008 [2:42:04<02:57, 4.98it/s] 98%|█████████▊| 47123/48008 [2:42:04<02:54, 5.06it/s] 98%|█████████▊| 47124/48008 [2:42:05<02:53, 5.11it/s] 98%|█████████▊| 47125/48008 [2:42:05<02:50, 5.18it/s] 98%|█████████▊| 47126/48008 [2:42:05<02:51, 5.15it/s] 98%|█████████▊| 47127/48008 [2:42:05<02:50, 5.15it/s] 98%|█████████▊| 47128/48008 [2:42:05<02:50, 5.16it/s] 98%|█████████▊| 47129/48008 [2:42:06<02:49, 5.19it/s] 98%|█████████▊| 47130/48008 [2:42:06<02:48, 5.20it/s] 98%|█████████▊| 47131/48008 [2:42:06<02:51, 5.10it/s] 98%|█████████▊| 47132/48008 [2:42:06<02:53, 5.05it/s] 98%|█████████▊| 47133/48008 [2:42:06<02:53, 5.05it/s] 98%|█████████▊| 47134/48008 [2:42:07<02:51, 5.11it/s] 98%|█████████▊| 47135/48008 [2:42:07<02:52, 5.05it/s] 98%|█████████▊| 47136/48008 [2:42:07<02:52, 5.05it/s] 98%|█████████▊| 47137/48008 [2:42:07<02:50, 5.12it/s] 98%|█████████▊| 47138/48008 [2:42:07<02:52, 5.04it/s] 98%|█████████▊| 47139/48008 [2:42:08<02:51, 5.07it/s] 98%|█████████▊| 47140/48008 [2:42:08<02:50, 5.08it/s] 98%|█████████▊| 47141/48008 [2:42:08<02:48, 5.14it/s] 98%|█████████▊| 47142/48008 [2:42:08<02:47, 5.18it/s] 98%|█████████▊| 47143/48008 [2:42:08<02:47, 5.18it/s] 98%|█████████▊| 47144/48008 [2:42:09<02:48, 5.12it/s] 98%|█████████▊| 47145/48008 [2:42:09<02:47, 5.15it/s] 98%|█████████▊| 47146/48008 [2:42:09<02:47, 5.14it/s] 98%|█████████▊| 47147/48008 [2:42:09<02:48, 5.12it/s] 98%|█████████▊| 47148/48008 [2:42:09<02:45, 5.19it/s] 98%|█████████▊| 47149/48008 [2:42:09<02:44, 5.21it/s] 98%|█████████▊| 47150/48008 [2:42:10<02:47, 5.13it/s] 98%|█████████▊| 47150/48008 [2:42:10<02:47, 5.13it/s]{'loss': 0.3, 'grad_norm': 0.3671875, 'learning_rate': 3.5785702382936182e-06, 'epoch': 0.98} 98%|█████████▊| 47151/48008 [2:42:10<02:49, 5.06it/s] 98%|█████████▊| 47152/48008 [2:42:10<03:17, 4.33it/s] 98%|█████████▊| 47153/48008 [2:42:10<03:10, 4.49it/s] 98%|█████████▊| 47154/48008 [2:42:11<03:05, 4.61it/s] 98%|█████████▊| 47155/48008 [2:42:11<02:58, 4.78it/s] 98%|█████████▊| 47156/48008 [2:42:11<02:53, 4.90it/s] 98%|█████████▊| 47157/48008 [2:42:11<02:53, 4.92it/s] 98%|█████████▊| 47158/48008 [2:42:11<02:49, 5.03it/s] 98%|█████████▊| 47159/48008 [2:42:12<02:46, 5.09it/s] 98%|█████████▊| 47160/48008 [2:42:12<02:45, 5.13it/s] 98%|█████████▊| 47161/48008 [2:42:12<02:46, 5.09it/s] 98%|█████████▊| 47162/48008 [2:42:12<02:44, 5.15it/s] 98%|█████████▊| 47163/48008 [2:42:12<02:42, 5.20it/s] 98%|█████████▊| 47164/48008 [2:42:13<02:43, 5.15it/s] 98%|█████████▊| 47165/48008 [2:42:13<02:43, 5.16it/s] 98%|█████████▊| 47166/48008 [2:42:13<02:42, 5.18it/s] 98%|█████████▊| 47167/48008 [2:42:13<02:41, 5.22it/s] 98%|█████████▊| 47168/48008 [2:42:13<02:39, 5.26it/s] 98%|█████████▊| 47169/48008 [2:42:13<02:42, 5.15it/s] 98%|█████████▊| 47170/48008 [2:42:14<02:42, 5.17it/s] 98%|█████████▊| 47171/48008 [2:42:14<02:44, 5.09it/s] 98%|█████████▊| 47172/48008 [2:42:14<02:47, 4.99it/s] 98%|█████████▊| 47173/48008 [2:42:14<02:44, 5.07it/s] 98%|█████████▊| 47174/48008 [2:42:14<02:42, 5.13it/s] 98%|█████████▊| 47175/48008 [2:42:15<02:41, 5.16it/s] 98%|█████████▊| 47176/48008 [2:42:15<02:52, 4.83it/s] 98%|█████████▊| 47177/48008 [2:42:15<03:00, 4.61it/s] 98%|█████████▊| 47178/48008 [2:42:15<02:56, 4.70it/s] 98%|█████████▊| 47179/48008 [2:42:16<02:50, 4.86it/s] 98%|█████████▊| 47180/48008 [2:42:16<02:47, 4.95it/s] 98%|█████████▊| 47181/48008 [2:42:16<02:43, 5.05it/s] 98%|█████████▊| 47182/48008 [2:42:16<02:42, 5.09it/s] 98%|█████████▊| 47183/48008 [2:42:16<02:41, 5.10it/s] 98%|█████████▊| 47184/48008 [2:42:16<02:39, 5.17it/s] 98%|█████████▊| 47185/48008 [2:42:17<02:39, 5.17it/s] 98%|█████████▊| 47186/48008 [2:42:17<02:40, 5.13it/s] 98%|█████████▊| 47187/48008 [2:42:17<02:40, 5.13it/s] 98%|█████████▊| 47188/48008 [2:42:17<02:39, 5.15it/s] 98%|█████████▊| 47189/48008 [2:42:17<02:39, 5.14it/s] 98%|█████████▊| 47190/48008 [2:42:18<02:39, 5.13it/s] 98%|█████████▊| 47191/48008 [2:42:18<02:40, 5.10it/s] 98%|█████████▊| 47192/48008 [2:42:18<02:39, 5.13it/s] 98%|█████████▊| 47193/48008 [2:42:18<02:38, 5.13it/s] 98%|█████████▊| 47194/48008 [2:42:18<02:40, 5.07it/s] 98%|█████████▊| 47195/48008 [2:42:19<02:38, 5.14it/s] 98%|█████████▊| 47196/48008 [2:42:19<02:35, 5.22it/s] 98%|█████████▊| 47197/48008 [2:42:19<02:38, 5.12it/s] 98%|█████████▊| 47198/48008 [2:42:19<02:36, 5.16it/s] 98%|█████████▊| 47199/48008 [2:42:19<02:35, 5.21it/s] 98%|█████████▊| 47200/48008 [2:42:20<02:39, 5.07it/s] {'loss': 0.2992, 'grad_norm': 0.494140625, 'learning_rate': 3.3702716213964343e-06, 'epoch': 0.98} 98%|█████████▊| 47200/48008 [2:42:20<02:39, 5.07it/s] 98%|█████████▊| 47201/48008 [2:42:20<02:40, 5.02it/s] 98%|█████████▊| 47202/48008 [2:42:20<02:38, 5.08it/s] 98%|█████████▊| 47203/48008 [2:42:20<02:37, 5.12it/s] 98%|█████████▊| 47204/48008 [2:42:20<02:35, 5.16it/s] 98%|█████████▊| 47205/48008 [2:42:21<03:03, 4.38it/s] 98%|█████████▊| 47206/48008 [2:42:21<02:54, 4.59it/s] 98%|█████████▊| 47207/48008 [2:42:21<02:48, 4.75it/s] 98%|█████████▊| 47208/48008 [2:42:21<02:44, 4.86it/s] 98%|█████████▊| 47209/48008 [2:42:21<02:41, 4.94it/s] 98%|█████████▊| 47210/48008 [2:42:22<02:40, 4.97it/s] 98%|█████████▊| 47211/48008 [2:42:22<03:05, 4.30it/s] 98%|█████████▊| 47212/48008 [2:42:22<02:58, 4.47it/s] 98%|█████████▊| 47213/48008 [2:42:22<02:50, 4.67it/s] 98%|█████████▊| 47214/48008 [2:42:23<02:48, 4.70it/s] 98%|█████████▊| 47215/48008 [2:42:23<02:43, 4.84it/s] 98%|█████████▊| 47216/48008 [2:42:23<02:39, 4.95it/s] 98%|█████████▊| 47217/48008 [2:42:23<02:39, 4.97it/s] 98%|█████████▊| 47218/48008 [2:42:23<02:36, 5.04it/s] 98%|█████████▊| 47219/48008 [2:42:24<02:34, 5.09it/s] 98%|█████████▊| 47220/48008 [2:42:24<02:33, 5.12it/s] 98%|█████████▊| 47221/48008 [2:42:24<03:00, 4.37it/s] 98%|█████████▊| 47222/48008 [2:42:24<02:51, 4.58it/s] 98%|█████████▊| 47223/48008 [2:42:24<02:46, 4.70it/s] 98%|█████████▊| 47224/48008 [2:42:25<02:40, 4.87it/s] 98%|█████████▊| 47225/48008 [2:42:25<02:36, 5.01it/s] 98%|█████████▊| 47226/48008 [2:42:25<02:34, 5.07it/s] 98%|█████████▊| 47227/48008 [2:42:25<02:32, 5.11it/s] 98%|█████████▊| 47228/48008 [2:42:25<02:32, 5.10it/s] 98%|█████████▊| 47229/48008 [2:42:26<02:30, 5.17it/s] 98%|█████████▊| 47230/48008 [2:42:26<02:33, 5.08it/s] 98%|█████████▊| 47231/48008 [2:42:26<02:33, 5.07it/s] 98%|█████████▊| 47232/48008 [2:42:27<05:20, 2.42it/s] 98%|█████████▊| 47233/48008 [2:42:27<04:28, 2.89it/s] 98%|█████████▊| 47234/48008 [2:42:27<03:51, 3.35it/s] 98%|█████████▊| 47235/48008 [2:42:27<03:26, 3.75it/s] 98%|█████████▊| 47236/48008 [2:42:28<03:10, 4.05it/s] 98%|█████████▊| 47237/48008 [2:42:28<02:59, 4.29it/s] 98%|█████████▊| 47238/48008 [2:42:28<02:52, 4.47it/s] 98%|█████████▊| 47239/48008 [2:42:28<03:30, 3.66it/s] 98%|█████████▊| 47240/48008 [2:42:29<03:10, 4.02it/s] 98%|█████████▊| 47241/48008 [2:42:29<02:57, 4.32it/s] 98%|█████████▊| 47242/48008 [2:42:29<02:49, 4.52it/s] 98%|█████████▊| 47243/48008 [2:42:29<02:42, 4.70it/s] 98%|█████████▊| 47244/48008 [2:42:29<02:37, 4.85it/s] 98%|█████████▊| 47245/48008 [2:42:30<02:32, 4.99it/s] 98%|█████████▊| 47246/48008 [2:42:30<02:41, 4.71it/s] 98%|█████████▊| 47247/48008 [2:42:30<02:36, 4.87it/s] 98%|█████████▊| 47248/48008 [2:42:30<02:32, 4.99it/s] 98%|█████████▊| 47249/48008 [2:42:30<02:30, 5.06it/s] 98%|█████████▊| 47250/48008 [2:42:31<02:28, 5.11it/s] {'loss': 0.2989, 'grad_norm': 0.162109375, 'learning_rate': 3.1619730044992504e-06, 'epoch': 0.98} 98%|█████████▊| 47250/48008 [2:42:31<02:28, 5.11it/s] 98%|█████████▊| 47251/48008 [2:42:31<02:27, 5.15it/s] 98%|█████████▊| 47252/48008 [2:42:31<02:27, 5.14it/s] 98%|█████████▊| 47253/48008 [2:42:31<02:26, 5.16it/s] 98%|█████████▊| 47254/48008 [2:42:31<02:27, 5.11it/s] 98%|█████████▊| 47255/48008 [2:42:32<02:26, 5.15it/s] 98%|█████████▊| 47256/48008 [2:42:32<02:26, 5.15it/s] 98%|█████████▊| 47257/48008 [2:42:32<02:25, 5.18it/s] 98%|█████████▊| 47258/48008 [2:42:32<02:26, 5.13it/s] 98%|█████████▊| 47259/48008 [2:42:32<02:28, 5.05it/s] 98%|█████████▊| 47260/48008 [2:42:33<02:29, 5.02it/s] 98%|█████████▊| 47261/48008 [2:42:33<02:25, 5.13it/s] 98%|█████████▊| 47262/48008 [2:42:33<02:25, 5.14it/s] 98%|█████████▊| 47263/48008 [2:42:33<02:50, 4.38it/s] 98%|█████████▊| 47264/48008 [2:42:33<02:42, 4.59it/s] 98%|█████████▊| 47265/48008 [2:42:34<02:35, 4.77it/s] 98%|█████████▊| 47266/48008 [2:42:34<02:31, 4.91it/s] 98%|█████████▊| 47267/48008 [2:42:34<02:29, 4.96it/s] 98%|█████████▊| 47268/48008 [2:42:34<02:29, 4.96it/s] 98%|█████████▊| 47269/48008 [2:42:34<02:26, 5.04it/s] 98%|█████████▊| 47270/48008 [2:42:35<02:27, 4.99it/s] 98%|█████████▊| 47271/48008 [2:42:35<02:24, 5.10it/s] 98%|█████████▊| 47272/48008 [2:42:35<02:23, 5.14it/s] 98%|█████████▊| 47273/48008 [2:42:35<02:21, 5.19it/s] 98%|█████████▊| 47274/48008 [2:42:35<02:20, 5.21it/s] 98%|█████████▊| 47275/48008 [2:42:36<02:19, 5.24it/s] 98%|█████████▊| 47276/48008 [2:42:36<02:19, 5.24it/s] 98%|█████████▊| 47277/48008 [2:42:36<02:19, 5.23it/s] 98%|█████████▊| 47278/48008 [2:42:36<02:19, 5.25it/s] 98%|█████████▊| 47279/48008 [2:42:36<02:19, 5.23it/s] 98%|█████████▊| 47280/48008 [2:42:37<02:20, 5.17it/s] 98%|█████████▊| 47281/48008 [2:42:37<02:20, 5.16it/s] 98%|█████████▊| 47282/48008 [2:42:37<02:20, 5.18it/s] 98%|█████████▊| 47283/48008 [2:42:37<02:19, 5.19it/s] 98%|█████████▊| 47284/48008 [2:42:37<02:18, 5.22it/s] 98%|█████████▊| 47285/48008 [2:42:38<02:21, 5.12it/s] 98%|█████████▊| 47286/48008 [2:42:38<02:20, 5.15it/s] 98%|█████████▊| 47287/48008 [2:42:38<02:19, 5.18it/s] 99%|█████████▊| 47288/48008 [2:42:38<02:18, 5.21it/s] 99%|█████████▊| 47289/48008 [2:42:38<02:18, 5.21it/s] 99%|█████████▊| 47290/48008 [2:42:38<02:19, 5.16it/s] 99%|█████████▊| 47291/48008 [2:42:39<02:17, 5.22it/s] 99%|█████████▊| 47292/48008 [2:42:39<02:20, 5.09it/s] 99%|█████████▊| 47293/48008 [2:42:39<02:20, 5.10it/s] 99%|█████████▊| 47294/48008 [2:42:39<02:19, 5.12it/s] 99%|█████████▊| 47295/48008 [2:42:39<02:16, 5.21it/s] 99%|█████████▊| 47296/48008 [2:42:40<02:58, 3.99it/s] 99%|█████████▊| 47297/48008 [2:42:40<02:44, 4.31it/s] 99%|█████████▊| 47298/48008 [2:42:40<02:38, 4.49it/s] 99%|█████████▊| 47299/48008 [2:42:40<02:32, 4.66it/s] 99%|█████████▊| 47300/48008 [2:42:41<02:30, 4.72it/s] {'loss': 0.3032, 'grad_norm': 0.1572265625, 'learning_rate': 2.9536743876020665e-06, 'epoch': 0.99} 99%|█████████▊| 47300/48008 [2:42:41<02:30, 4.72it/s] 99%|█████████▊| 47301/48008 [2:42:41<02:29, 4.74it/s] 99%|█████████▊| 47302/48008 [2:42:41<02:24, 4.88it/s] 99%|█████████▊| 47303/48008 [2:42:41<02:21, 4.97it/s] 99%|█████████▊| 47304/48008 [2:42:41<02:20, 4.99it/s] 99%|█████████▊| 47305/48008 [2:42:42<02:21, 4.96it/s] 99%|█████████▊| 47306/48008 [2:42:42<02:19, 5.03it/s] 99%|█████████▊| 47307/48008 [2:42:42<02:17, 5.09it/s] 99%|█████████▊| 47308/48008 [2:42:42<02:15, 5.16it/s] 99%|█████████▊| 47309/48008 [2:42:42<02:14, 5.18it/s] 99%|█████████▊| 47310/48008 [2:42:43<02:17, 5.06it/s] 99%|█████████▊| 47311/48008 [2:42:43<02:17, 5.07it/s] 99%|█████████▊| 47312/48008 [2:42:43<02:16, 5.10it/s] 99%|█████████▊| 47313/48008 [2:42:43<02:15, 5.12it/s] 99%|█████████▊| 47314/48008 [2:42:43<02:15, 5.11it/s] 99%|█████████▊| 47315/48008 [2:42:44<02:15, 5.12it/s] 99%|█████████▊| 47316/48008 [2:42:44<02:15, 5.12it/s] 99%|█████████▊| 47317/48008 [2:42:44<02:13, 5.16it/s] 99%|█████████▊| 47318/48008 [2:42:44<02:15, 5.08it/s] 99%|█████████▊| 47319/48008 [2:42:44<02:14, 5.12it/s] 99%|█████████▊| 47320/48008 [2:42:45<02:13, 5.14it/s] 99%|█████████▊| 47321/48008 [2:42:45<02:12, 5.18it/s] 99%|█████████▊| 47322/48008 [2:42:45<02:12, 5.19it/s] 99%|█████████▊| 47323/48008 [2:42:45<02:12, 5.17it/s] 99%|█████████▊| 47324/48008 [2:42:45<02:12, 5.15it/s] 99%|█████████▊| 47325/48008 [2:42:46<02:15, 5.06it/s] 99%|█████████▊| 47326/48008 [2:42:46<02:14, 5.09it/s] 99%|█████████▊| 47327/48008 [2:42:46<02:12, 5.13it/s] 99%|█████████▊| 47328/48008 [2:42:46<02:14, 5.07it/s] 99%|█████████▊| 47329/48008 [2:42:46<02:14, 5.03it/s] 99%|█████████▊| 47330/48008 [2:42:46<02:14, 5.06it/s] 99%|█████████▊| 47331/48008 [2:42:47<02:12, 5.10it/s] 99%|█████████▊| 47332/48008 [2:42:47<02:12, 5.12it/s] 99%|█████████▊| 47333/48008 [2:42:47<02:10, 5.17it/s] 99%|█████████▊| 47334/48008 [2:42:47<02:49, 3.98it/s] 99%|█████████▊| 47335/48008 [2:42:48<02:36, 4.29it/s] 99%|█████████▊| 47336/48008 [2:42:48<02:30, 4.47it/s] 99%|█████████▊| 47337/48008 [2:42:48<02:24, 4.64it/s] 99%|█████████▊| 47338/48008 [2:42:48<02:21, 4.75it/s] 99%|█████████▊| 47339/48008 [2:42:48<02:15, 4.95it/s] 99%|█████████▊| 47340/48008 [2:42:49<02:13, 5.02it/s] 99%|█████████▊| 47341/48008 [2:42:49<02:12, 5.04it/s] 99%|█████████▊| 47342/48008 [2:42:49<02:12, 5.04it/s] 99%|█████████▊| 47343/48008 [2:42:49<02:10, 5.08it/s] 99%|█████████▊| 47344/48008 [2:42:49<02:10, 5.09it/s] 99%|█████████▊| 47345/48008 [2:42:50<02:10, 5.09it/s] 99%|█████████▊| 47346/48008 [2:42:50<02:11, 5.04it/s] 99%|█████████▊| 47347/48008 [2:42:50<02:09, 5.11it/s] 99%|█████████▊| 47348/48008 [2:42:50<02:11, 5.04it/s] 99%|█████████▊| 47349/48008 [2:42:50<02:18, 4.75it/s] 99%|█████████▊| 47350/48008 [2:42:51<02:14, 4.90it/s] {'loss': 0.3013, 'grad_norm': 0.2041015625, 'learning_rate': 2.7453757707048826e-06, 'epoch': 0.99} 99%|█████████▊| 47350/48008 [2:42:51<02:14, 4.90it/s] 99%|█████████▊| 47351/48008 [2:42:51<02:13, 4.93it/s] 99%|█████████▊| 47352/48008 [2:42:51<02:20, 4.67it/s] 99%|█████████▊| 47353/48008 [2:42:51<02:16, 4.81it/s] 99%|█████████▊| 47354/48008 [2:42:52<02:51, 3.82it/s] 99%|█████████▊| 47355/48008 [2:42:52<03:00, 3.63it/s] 99%|█████████▊| 47356/48008 [2:42:52<02:43, 3.98it/s] 99%|█████████▊| 47357/48008 [2:42:52<02:31, 4.29it/s] 99%|█████████▊| 47358/48008 [2:42:53<02:25, 4.47it/s] 99%|█████████▊| 47359/48008 [2:42:53<02:19, 4.65it/s] 99%|█████████▊| 47360/48008 [2:42:53<02:13, 4.85it/s] 99%|█████████▊| 47361/48008 [2:42:53<02:11, 4.90it/s] 99%|█████████▊| 47362/48008 [2:42:53<02:10, 4.94it/s] 99%|█████████▊| 47363/48008 [2:42:54<02:08, 5.01it/s] 99%|█████████▊| 47364/48008 [2:42:54<02:09, 4.98it/s] 99%|█████████▊| 47365/48008 [2:42:54<02:09, 4.95it/s] 99%|█████████▊| 47366/48008 [2:42:54<02:08, 5.00it/s] 99%|█████████▊| 47367/48008 [2:42:54<02:06, 5.07it/s] 99%|█████████▊| 47368/48008 [2:42:54<02:05, 5.10it/s] 99%|█████████▊| 47369/48008 [2:42:55<02:06, 5.04it/s] 99%|█████████▊| 47370/48008 [2:42:55<02:27, 4.32it/s] 99%|█████████▊| 47371/48008 [2:42:55<02:41, 3.94it/s] 99%|█████████▊| 47372/48008 [2:42:56<03:07, 3.40it/s] 99%|█████████▊| 47373/48008 [2:42:56<02:47, 3.79it/s] 99%|█████████▊| 47374/48008 [2:42:56<02:35, 4.07it/s] 99%|█████████▊| 47375/48008 [2:42:56<02:47, 3.79it/s] 99%|█████████▊| 47376/48008 [2:42:57<02:31, 4.16it/s] 99%|█████████▊| 47377/48008 [2:42:57<02:22, 4.44it/s] 99%|█████████▊| 47378/48008 [2:42:57<02:16, 4.63it/s] 99%|█████████▊| 47379/48008 [2:42:57<02:12, 4.74it/s] 99%|█████████▊| 47380/48008 [2:42:57<02:08, 4.88it/s] 99%|█████████▊| 47381/48008 [2:42:58<02:04, 5.03it/s] 99%|█████████▊| 47382/48008 [2:42:58<02:03, 5.08it/s] 99%|█████████▊| 47383/48008 [2:42:58<02:02, 5.11it/s] 99%|█████████▊| 47384/48008 [2:42:58<02:01, 5.12it/s] 99%|█████████▊| 47385/48008 [2:42:58<02:01, 5.12it/s] 99%|█████████▊| 47386/48008 [2:42:59<02:01, 5.13it/s] 99%|█████████▊| 47387/48008 [2:42:59<02:01, 5.12it/s] 99%|█████████▊| 47388/48008 [2:42:59<02:00, 5.14it/s] 99%|█████████▊| 47389/48008 [2:42:59<02:00, 5.15it/s] 99%|█████████▊| 47390/48008 [2:42:59<01:58, 5.20it/s] 99%|█████████▊| 47391/48008 [2:42:59<01:59, 5.18it/s] 99%|█████████▊| 47392/48008 [2:43:00<02:01, 5.08it/s] 99%|█████████▊| 47393/48008 [2:43:00<02:01, 5.05it/s] 99%|█████████▊| 47394/48008 [2:43:00<02:02, 5.00it/s] 99%|█████████▊| 47395/48008 [2:43:00<02:36, 3.91it/s] 99%|█████████▊| 47396/48008 [2:43:01<02:24, 4.24it/s] 99%|█████████▊| 47397/48008 [2:43:01<02:16, 4.46it/s] 99%|█████████▊| 47398/48008 [2:43:01<02:11, 4.66it/s] 99%|█████████▊| 47399/48008 [2:43:01<02:08, 4.75it/s] 99%|█████████▊| 47400/48008 [2:43:01<02:05, 4.85it/s] {'loss': 0.2994, 'grad_norm': 0.1796875, 'learning_rate': 2.5370771538076987e-06, 'epoch': 0.99} 99%|█████████▊| 47400/48008 [2:43:01<02:05, 4.85it/s] 99%|█████████▊| 47401/48008 [2:43:02<02:03, 4.93it/s] 99%|█████████▊| 47402/48008 [2:43:02<02:00, 5.01it/s] 99%|█████████▊| 47403/48008 [2:43:02<01:59, 5.07it/s] 99%|█████████▊| 47404/48008 [2:43:02<01:58, 5.09it/s] 99%|█████████▊| 47405/48008 [2:43:02<01:57, 5.11it/s] 99%|█████████▊| 47406/48008 [2:43:03<02:00, 4.99it/s] 99%|█████████▊| 47407/48008 [2:43:03<01:59, 5.02it/s] 99%|█████████▉| 47408/48008 [2:43:03<01:59, 5.03it/s] 99%|█████████▉| 47409/48008 [2:43:03<01:56, 5.14it/s] 99%|█████████▉| 47410/48008 [2:43:03<01:56, 5.13it/s] 99%|█████████▉| 47411/48008 [2:43:04<01:55, 5.17it/s] 99%|█████████▉| 47412/48008 [2:43:04<01:55, 5.18it/s] 99%|█████████▉| 47413/48008 [2:43:04<01:56, 5.13it/s] 99%|█████████▉| 47414/48008 [2:43:04<01:55, 5.13it/s] 99%|█████████▉| 47415/48008 [2:43:04<01:54, 5.16it/s] 99%|█████████▉| 47416/48008 [2:43:05<01:54, 5.18it/s] 99%|█████████▉| 47417/48008 [2:43:05<01:54, 5.15it/s] 99%|█████████▉| 47418/48008 [2:43:05<01:54, 5.16it/s] 99%|█████████▉| 47419/48008 [2:43:05<01:53, 5.17it/s] 99%|█████████▉| 47420/48008 [2:43:05<01:52, 5.20it/s] 99%|█████████▉| 47421/48008 [2:43:06<01:53, 5.17it/s] 99%|█████████▉| 47422/48008 [2:43:06<01:52, 5.19it/s] 99%|█████████▉| 47423/48008 [2:43:06<01:51, 5.25it/s] 99%|█████████▉| 47424/48008 [2:43:06<01:52, 5.21it/s] 99%|█████████▉| 47425/48008 [2:43:06<01:51, 5.22it/s] 99%|█████████▉| 47426/48008 [2:43:06<01:50, 5.25it/s] 99%|█████████▉| 47427/48008 [2:43:07<01:51, 5.23it/s] 99%|█████████▉| 47428/48008 [2:43:07<01:50, 5.25it/s] 99%|█████████▉| 47429/48008 [2:43:07<01:50, 5.22it/s] 99%|█████████▉| 47430/48008 [2:43:07<01:50, 5.24it/s] 99%|█████████▉| 47431/48008 [2:43:07<01:52, 5.15it/s] 99%|█████████▉| 47432/48008 [2:43:08<01:53, 5.07it/s] 99%|█████████▉| 47433/48008 [2:43:08<01:52, 5.12it/s] 99%|█████████▉| 47434/48008 [2:43:08<01:50, 5.18it/s] 99%|█████████▉| 47435/48008 [2:43:08<01:50, 5.20it/s] 99%|█████████▉| 47436/48008 [2:43:08<01:50, 5.19it/s] 99%|█████████▉| 47437/48008 [2:43:09<01:57, 4.86it/s] 99%|█████████▉| 47438/48008 [2:43:09<01:56, 4.88it/s] 99%|█████████▉| 47439/48008 [2:43:09<01:54, 4.97it/s] 99%|█████████▉| 47440/48008 [2:43:09<01:52, 5.04it/s] 99%|█████████▉| 47441/48008 [2:43:09<01:53, 4.98it/s] 99%|█████████▉| 47442/48008 [2:43:10<01:52, 5.01it/s] 99%|█████████▉| 47443/48008 [2:43:10<01:50, 5.10it/s] 99%|█████████▉| 47444/48008 [2:43:10<01:50, 5.08it/s] 99%|█████████▉| 47445/48008 [2:43:10<01:52, 5.00it/s] 99%|█████████▉| 47446/48008 [2:43:10<01:51, 5.06it/s] 99%|█████████▉| 47447/48008 [2:43:11<01:50, 5.07it/s] 99%|█████████▉| 47448/48008 [2:43:11<01:49, 5.12it/s] 99%|█████████▉| 47449/48008 [2:43:11<01:50, 5.07it/s] 99%|█████████▉| 47450/48008 [2:43:11<01:49, 5.09it/s] {'loss': 0.3028, 'grad_norm': 0.2265625, 'learning_rate': 2.328778536910515e-06, 'epoch': 0.99} 99%|█████████▉| 47450/48008 [2:43:11<01:49, 5.09it/s] 99%|█████████▉| 47451/48008 [2:43:11<01:48, 5.12it/s] 99%|█████████▉| 47452/48008 [2:43:12<01:47, 5.16it/s] 99%|█████████▉| 47453/48008 [2:43:12<01:46, 5.19it/s] 99%|█████████▉| 47454/48008 [2:43:12<01:46, 5.21it/s] 99%|█████████▉| 47455/48008 [2:43:12<01:46, 5.21it/s] 99%|█████████▉| 47456/48008 [2:43:12<01:46, 5.19it/s] 99%|█████████▉| 47457/48008 [2:43:13<01:47, 5.11it/s] 99%|█████████▉| 47458/48008 [2:43:13<01:46, 5.16it/s] 99%|█████████▉| 47459/48008 [2:43:13<01:46, 5.15it/s] 99%|█████████▉| 47460/48008 [2:43:13<01:45, 5.22it/s] 99%|█████████▉| 47461/48008 [2:43:13<01:45, 5.20it/s] 99%|█████████▉| 47462/48008 [2:43:14<01:45, 5.20it/s] 99%|█████████▉| 47463/48008 [2:43:14<01:45, 5.18it/s] 99%|█████████▉| 47464/48008 [2:43:14<01:52, 4.83it/s] 99%|█████████▉| 47465/48008 [2:43:14<01:50, 4.91it/s] 99%|█████████▉| 47466/48008 [2:43:14<01:48, 5.01it/s] 99%|█████████▉| 47467/48008 [2:43:15<01:47, 5.05it/s] 99%|█████████▉| 47468/48008 [2:43:15<01:46, 5.07it/s] 99%|█████████▉| 47469/48008 [2:43:15<01:45, 5.12it/s] 99%|█████████▉| 47470/48008 [2:43:15<01:44, 5.15it/s] 99%|█████████▉| 47471/48008 [2:43:15<01:43, 5.17it/s] 99%|█████████▉| 47472/48008 [2:43:15<01:42, 5.23it/s] 99%|█████████▉| 47473/48008 [2:43:16<01:42, 5.24it/s] 99%|█████████▉| 47474/48008 [2:43:16<01:42, 5.19it/s] 99%|█████████▉| 47475/48008 [2:43:16<01:43, 5.13it/s] 99%|█████████▉| 47476/48008 [2:43:16<01:51, 4.79it/s] 99%|█████████▉| 47477/48008 [2:43:17<01:48, 4.91it/s] 99%|█████████▉| 47478/48008 [2:43:17<01:48, 4.89it/s] 99%|█████████▉| 47479/48008 [2:43:17<01:45, 4.99it/s] 99%|█████████▉| 47480/48008 [2:43:17<01:44, 5.05it/s] 99%|█████████▉| 47481/48008 [2:43:17<01:45, 4.99it/s] 99%|█████████▉| 47482/48008 [2:43:18<01:43, 5.06it/s] 99%|█████████▉| 47483/48008 [2:43:18<01:42, 5.11it/s] 99%|█████████▉| 47484/48008 [2:43:18<01:41, 5.17it/s] 99%|█████████▉| 47485/48008 [2:43:18<01:40, 5.18it/s] 99%|█████████▉| 47486/48008 [2:43:18<01:41, 5.15it/s] 99%|█████████▉| 47487/48008 [2:43:18<01:40, 5.20it/s] 99%|█████████▉| 47488/48008 [2:43:19<01:43, 5.05it/s] 99%|█████████▉| 47489/48008 [2:43:19<01:41, 5.09it/s] 99%|█████████▉| 47490/48008 [2:43:19<01:40, 5.15it/s] 99%|█████████▉| 47491/48008 [2:43:19<01:39, 5.19it/s] 99%|█████████▉| 47492/48008 [2:43:19<01:40, 5.16it/s] 99%|█████████▉| 47493/48008 [2:43:20<01:39, 5.16it/s] 99%|█████████▉| 47494/48008 [2:43:20<01:38, 5.21it/s] 99%|█████████▉| 47495/48008 [2:43:20<01:40, 5.11it/s] 99%|█████████▉| 47496/48008 [2:43:20<01:40, 5.09it/s] 99%|█████████▉| 47497/48008 [2:43:20<01:40, 5.07it/s] 99%|█████████▉| 47498/48008 [2:43:21<01:39, 5.11it/s] 99%|█████████▉| 47499/48008 [2:43:21<01:39, 5.11it/s] 99%|█████████▉| 47500/48008 [2:43:21<01:38, 5.16it/s] {'loss': 0.3018, 'grad_norm': 0.314453125, 'learning_rate': 2.1204799200133313e-06, 'epoch': 0.99} 99%|█████████▉| 47500/48008 [2:43:21<01:38, 5.16it/s] 99%|█████████▉| 47501/48008 [2:43:21<01:40, 5.05it/s] 99%|█████████▉| 47502/48008 [2:43:21<01:38, 5.13it/s] 99%|█████████▉| 47503/48008 [2:43:22<01:39, 5.08it/s] 99%|█████████▉| 47504/48008 [2:43:22<01:38, 5.12it/s] 99%|█████████▉| 47505/48008 [2:43:22<01:39, 5.06it/s] 99%|█████████▉| 47506/48008 [2:43:22<01:40, 5.01it/s] 99%|█████████▉| 47507/48008 [2:43:22<01:41, 4.95it/s] 99%|█████████▉| 47508/48008 [2:43:23<01:39, 5.01it/s] 99%|█████████▉| 47509/48008 [2:43:23<01:38, 5.06it/s] 99%|█████████▉| 47510/48008 [2:43:23<01:37, 5.11it/s] 99%|█████████▉| 47511/48008 [2:43:23<01:37, 5.09it/s] 99%|█████████▉| 47512/48008 [2:43:23<01:37, 5.11it/s] 99%|█████████▉| 47513/48008 [2:43:24<01:37, 5.06it/s] 99%|█████████▉| 47514/48008 [2:43:24<01:36, 5.12it/s] 99%|█████████▉| 47515/48008 [2:43:24<01:37, 5.06it/s] 99%|█████████▉| 47516/48008 [2:43:24<01:39, 4.96it/s] 99%|█████████▉| 47517/48008 [2:43:24<01:36, 5.11it/s] 99%|█████████▉| 47518/48008 [2:43:25<01:35, 5.15it/s] 99%|█████████▉| 47519/48008 [2:43:25<01:34, 5.16it/s] 99%|█████████▉| 47520/48008 [2:43:25<01:34, 5.16it/s] 99%|█████████▉| 47521/48008 [2:43:25<01:33, 5.19it/s] 99%|█████████▉| 47522/48008 [2:43:25<01:33, 5.21it/s] 99%|█████████▉| 47523/48008 [2:43:26<01:35, 5.10it/s] 99%|█████████▉| 47524/48008 [2:43:26<01:35, 5.07it/s] 99%|█████████▉| 47525/48008 [2:43:26<01:34, 5.12it/s] 99%|█████████▉| 47526/48008 [2:43:26<01:35, 5.05it/s] 99%|█████████▉| 47527/48008 [2:43:26<01:37, 4.95it/s] 99%|█████████▉| 47528/48008 [2:43:27<01:35, 5.02it/s] 99%|█████████▉| 47529/48008 [2:43:27<01:36, 4.97it/s] 99%|█████████▉| 47530/48008 [2:43:27<01:34, 5.05it/s] 99%|█████████▉| 47531/48008 [2:43:27<01:33, 5.10it/s] 99%|█████████▉| 47532/48008 [2:43:27<01:33, 5.07it/s] 99%|█████████▉| 47533/48008 [2:43:28<01:34, 5.01it/s] 99%|█████████▉| 47534/48008 [2:43:28<01:33, 5.05it/s] 99%|█████████▉| 47535/48008 [2:43:28<01:33, 5.08it/s] 99%|█████████▉| 47536/48008 [2:43:28<01:33, 5.04it/s] 99%|█████████▉| 47537/48008 [2:43:28<01:32, 5.11it/s] 99%|█████████▉| 47538/48008 [2:43:28<01:32, 5.06it/s] 99%|█████████▉| 47539/48008 [2:43:29<01:31, 5.10it/s] 99%|█████████▉| 47540/48008 [2:43:29<01:30, 5.15it/s] 99%|█████████▉| 47541/48008 [2:43:29<01:37, 4.81it/s] 99%|█████████▉| 47542/48008 [2:43:29<01:35, 4.90it/s] 99%|█████████▉| 47543/48008 [2:43:30<01:39, 4.67it/s] 99%|█████████▉| 47544/48008 [2:43:30<01:36, 4.83it/s] 99%|█████████▉| 47545/48008 [2:43:30<01:34, 4.90it/s] 99%|█████████▉| 47546/48008 [2:43:30<01:33, 4.94it/s] 99%|█████████▉| 47547/48008 [2:43:30<01:34, 4.90it/s] 99%|█████████▉| 47548/48008 [2:43:31<01:32, 5.00it/s] 99%|█████████▉| 47549/48008 [2:43:31<01:36, 4.73it/s] 99%|█████████▉| 47550/48008 [2:43:31<01:34, 4.86it/s] {'loss': 0.3002, 'grad_norm': 0.361328125, 'learning_rate': 1.9121813031161474e-06, 'epoch': 0.99} 99%|█████████▉| 47550/48008 [2:43:31<01:34, 4.86it/s] 99%|█████████▉| 47551/48008 [2:43:31<01:31, 4.97it/s] 99%|█████████▉| 47552/48008 [2:43:31<01:30, 5.04it/s] 99%|█████████▉| 47553/48008 [2:43:32<01:30, 5.02it/s] 99%|█████████▉| 47554/48008 [2:43:32<01:30, 5.04it/s] 99%|█████████▉| 47555/48008 [2:43:32<01:31, 4.95it/s] 99%|█████████▉| 47556/48008 [2:43:32<01:30, 5.02it/s] 99%|█████████▉| 47557/48008 [2:43:32<01:28, 5.07it/s] 99%|█████████▉| 47558/48008 [2:43:33<01:27, 5.12it/s] 99%|█████████▉| 47559/48008 [2:43:33<01:27, 5.16it/s] 99%|█████████▉| 47560/48008 [2:43:33<01:25, 5.21it/s] 99%|█████████▉| 47561/48008 [2:43:33<01:27, 5.10it/s] 99%|█████████▉| 47562/48008 [2:43:33<01:26, 5.13it/s] 99%|█████████▉| 47563/48008 [2:43:34<01:26, 5.15it/s] 99%|█████████▉| 47564/48008 [2:43:34<01:26, 5.15it/s] 99%|█████████▉| 47565/48008 [2:43:34<01:25, 5.16it/s] 99%|█████████▉| 47566/48008 [2:43:34<01:26, 5.10it/s] 99%|█████████▉| 47567/48008 [2:43:34<01:26, 5.13it/s] 99%|█████████▉| 47568/48008 [2:43:34<01:26, 5.07it/s] 99%|█████████▉| 47569/48008 [2:43:35<01:32, 4.76it/s] 99%|█████████▉| 47570/48008 [2:43:35<01:30, 4.82it/s] 99%|█████████▉| 47571/48008 [2:43:35<01:28, 4.96it/s] 99%|█████████▉| 47572/48008 [2:43:35<01:26, 5.04it/s] 99%|█████████▉| 47573/48008 [2:43:37<03:41, 1.97it/s] 99%|█████████▉| 47574/48008 [2:43:37<02:59, 2.42it/s] 99%|█████████▉| 47575/48008 [2:43:37<02:29, 2.89it/s] 99%|█████████▉| 47576/48008 [2:43:37<02:09, 3.34it/s] 99%|█████████▉| 47577/48008 [2:43:37<01:56, 3.71it/s] 99%|█████████▉| 47578/48008 [2:43:38<01:46, 4.05it/s] 99%|█████████▉| 47579/48008 [2:43:38<01:40, 4.27it/s] 99%|█████████▉| 47580/48008 [2:43:38<01:35, 4.50it/s] 99%|█████████▉| 47581/48008 [2:43:38<01:55, 3.68it/s] 99%|█████████▉| 47582/48008 [2:43:38<01:47, 3.97it/s] 99%|█████████▉| 47583/48008 [2:43:39<01:40, 4.22it/s] 99%|█████████▉| 47584/48008 [2:43:39<01:35, 4.45it/s] 99%|█████████▉| 47585/48008 [2:43:39<01:31, 4.64it/s] 99%|█████████▉| 47586/48008 [2:43:39<01:28, 4.75it/s] 99%|█████████▉| 47587/48008 [2:43:39<01:28, 4.78it/s] 99%|█████████▉| 47588/48008 [2:43:40<01:26, 4.85it/s] 99%|█████████▉| 47589/48008 [2:43:40<01:24, 4.95it/s] 99%|█████████▉| 47590/48008 [2:43:40<01:23, 5.03it/s] 99%|█████████▉| 47591/48008 [2:43:40<01:21, 5.09it/s] 99%|█████████▉| 47592/48008 [2:43:40<01:21, 5.13it/s] 99%|█████████▉| 47593/48008 [2:43:41<01:20, 5.18it/s] 99%|█████████▉| 47594/48008 [2:43:41<01:20, 5.14it/s] 99%|█████████▉| 47595/48008 [2:43:41<01:20, 5.12it/s] 99%|█████████▉| 47596/48008 [2:43:41<01:20, 5.09it/s] 99%|█████████▉| 47597/48008 [2:43:41<01:21, 5.02it/s] 99%|█████████▉| 47598/48008 [2:43:42<01:21, 5.06it/s] 99%|█████████▉| 47599/48008 [2:43:42<01:20, 5.06it/s] 99%|█████████▉| 47600/48008 [2:43:42<01:19, 5.14it/s] 99%|█████████▉| 47600/48008 [2:43:42<01:19, 5.14it/s]{'loss': 0.2994, 'grad_norm': 0.373046875, 'learning_rate': 1.7038826862189637e-06, 'epoch': 0.99} 99%|█████████▉| 47601/48008 [2:43:42<01:19, 5.15it/s] 99%|█████████▉| 47602/48008 [2:43:42<01:18, 5.16it/s] 99%|█████████▉| 47603/48008 [2:43:43<01:17, 5.22it/s] 99%|█████████▉| 47604/48008 [2:43:43<01:17, 5.21it/s] 99%|█████████▉| 47605/48008 [2:43:43<01:22, 4.86it/s] 99%|█████████▉| 47606/48008 [2:43:43<01:20, 4.96it/s] 99%|█████████▉| 47607/48008 [2:43:44<01:33, 4.30it/s] 99%|█████████▉| 47608/48008 [2:43:44<01:28, 4.54it/s] 99%|█████████▉| 47609/48008 [2:43:44<01:23, 4.76it/s] 99%|█████████▉| 47610/48008 [2:43:44<01:21, 4.91it/s] 99%|█████████▉| 47611/48008 [2:43:44<01:19, 4.98it/s] 99%|█████████▉| 47612/48008 [2:43:44<01:17, 5.11it/s] 99%|█████████▉| 47613/48008 [2:43:45<01:16, 5.14it/s] 99%|█████████▉| 47614/48008 [2:43:45<01:16, 5.16it/s] 99%|█████████▉| 47615/48008 [2:43:45<01:16, 5.11it/s] 99%|█████████▉| 47616/48008 [2:43:45<01:17, 5.08it/s] 99%|█████████▉| 47617/48008 [2:43:45<01:16, 5.10it/s] 99%|█████████▉| 47618/48008 [2:43:46<01:16, 5.09it/s] 99%|█████████▉| 47619/48008 [2:43:46<01:15, 5.13it/s] 99%|█████████▉| 47620/48008 [2:43:46<01:15, 5.15it/s] 99%|█████████▉| 47621/48008 [2:43:46<01:14, 5.19it/s] 99%|█████████▉| 47622/48008 [2:43:46<01:15, 5.11it/s] 99%|█████████▉| 47623/48008 [2:43:47<01:14, 5.17it/s] 99%|█████████▉| 47624/48008 [2:43:47<01:13, 5.19it/s] 99%|█████████▉| 47625/48008 [2:43:47<01:13, 5.19it/s] 99%|█████████▉| 47626/48008 [2:43:47<01:13, 5.21it/s] 99%|█████████▉| 47627/48008 [2:43:47<01:14, 5.15it/s] 99%|█████████▉| 47628/48008 [2:43:48<01:12, 5.22it/s] 99%|█████████▉| 47629/48008 [2:43:48<01:12, 5.21it/s] 99%|█████████▉| 47630/48008 [2:43:48<01:12, 5.20it/s] 99%|█████████▉| 47631/48008 [2:43:48<01:13, 5.14it/s] 99%|█████████▉| 47632/48008 [2:43:48<01:12, 5.21it/s] 99%|█████████▉| 47633/48008 [2:43:49<01:11, 5.21it/s] 99%|█████████▉| 47634/48008 [2:43:49<01:11, 5.20it/s] 99%|█████████▉| 47635/48008 [2:43:49<01:11, 5.20it/s] 99%|█████████▉| 47636/48008 [2:43:49<01:11, 5.20it/s] 99%|█████████▉| 47637/48008 [2:43:49<01:11, 5.20it/s] 99%|█████████▉| 47638/48008 [2:43:50<01:16, 4.85it/s] 99%|█████████▉| 47639/48008 [2:43:50<01:15, 4.92it/s] 99%|█████████▉| 47640/48008 [2:43:50<01:15, 4.87it/s] 99%|█████████▉| 47641/48008 [2:43:50<01:15, 4.86it/s] 99%|█████████▉| 47642/48008 [2:43:50<01:14, 4.89it/s] 99%|█████████▉| 47643/48008 [2:43:51<01:13, 4.96it/s] 99%|█████████▉| 47644/48008 [2:43:51<01:17, 4.70it/s] 99%|█████████▉| 47645/48008 [2:43:51<01:15, 4.84it/s] 99%|█████████▉| 47646/48008 [2:43:51<01:15, 4.81it/s] 99%|█████████▉| 47647/48008 [2:43:51<01:14, 4.84it/s] 99%|█████████▉| 47648/48008 [2:43:52<01:14, 4.84it/s] 99%|█████████▉| 47649/48008 [2:43:52<01:12, 4.94it/s] 99%|█████████▉| 47650/48008 [2:43:52<01:10, 5.07it/s] {'loss': 0.2999, 'grad_norm': 0.435546875, 'learning_rate': 1.4955840693217798e-06, 'epoch': 0.99} 99%|█████████▉| 47650/48008 [2:43:52<01:10, 5.07it/s] 99%|█████████▉| 47651/48008 [2:43:52<01:09, 5.11it/s] 99%|█████████▉| 47652/48008 [2:43:52<01:08, 5.18it/s] 99%|█████████▉| 47653/48008 [2:43:53<01:08, 5.20it/s] 99%|█████████▉| 47654/48008 [2:43:53<01:07, 5.23it/s] 99%|█████████▉| 47655/48008 [2:43:53<01:07, 5.21it/s] 99%|█████████▉| 47656/48008 [2:43:53<01:19, 4.40it/s] 99%|█████████▉| 47657/48008 [2:43:53<01:16, 4.56it/s] 99%|█████████▉| 47658/48008 [2:43:54<01:26, 4.06it/s] 99%|█████████▉| 47659/48008 [2:43:54<01:20, 4.34it/s] 99%|█████████▉| 47660/48008 [2:43:54<01:17, 4.48it/s] 99%|█████████▉| 47661/48008 [2:43:54<01:14, 4.66it/s] 99%|█████████▉| 47662/48008 [2:43:55<01:12, 4.78it/s] 99%|█████████▉| 47663/48008 [2:43:55<01:11, 4.79it/s] 99%|█████████▉| 47664/48008 [2:43:55<01:09, 4.93it/s] 99%|█████████▉| 47665/48008 [2:43:55<01:10, 4.90it/s] 99%|█████████▉| 47666/48008 [2:43:55<01:08, 4.96it/s] 99%|█████████▉| 47667/48008 [2:43:56<01:07, 5.05it/s] 99%|█████████▉| 47668/48008 [2:43:56<01:08, 5.00it/s] 99%|█████████▉| 47669/48008 [2:43:56<01:07, 5.04it/s] 99%|█████████▉| 47670/48008 [2:43:56<01:06, 5.09it/s] 99%|█████████▉| 47671/48008 [2:43:56<01:05, 5.14it/s] 99%|█████████▉| 47672/48008 [2:43:57<01:04, 5.18it/s] 99%|█████████▉| 47673/48008 [2:43:57<01:04, 5.16it/s] 99%|█████████▉| 47674/48008 [2:43:57<01:05, 5.07it/s] 99%|█████████▉| 47675/48008 [2:43:57<01:05, 5.07it/s] 99%|█████████▉| 47676/48008 [2:43:57<01:04, 5.12it/s] 99%|█████████▉| 47677/48008 [2:43:57<01:04, 5.17it/s] 99%|█████████▉| 47678/48008 [2:43:58<01:04, 5.11it/s] 99%|█████████▉| 47679/48008 [2:43:58<01:03, 5.14it/s] 99%|█████████▉| 47680/48008 [2:43:58<01:04, 5.07it/s] 99%|█████████▉| 47681/48008 [2:43:58<01:04, 5.05it/s] 99%|█████████▉| 47682/48008 [2:43:58<01:03, 5.11it/s] 99%|█████████▉| 47683/48008 [2:43:59<01:04, 5.08it/s] 99%|█████████▉| 47684/48008 [2:43:59<01:04, 4.99it/s] 99%|█████████▉| 47685/48008 [2:43:59<01:03, 5.05it/s] 99%|█████████▉| 47686/48008 [2:43:59<01:07, 4.76it/s] 99%|█████████▉| 47687/48008 [2:44:00<01:07, 4.76it/s] 99%|█████████▉| 47688/48008 [2:44:00<01:05, 4.88it/s] 99%|█████████▉| 47689/48008 [2:44:00<01:04, 4.95it/s] 99%|█████████▉| 47690/48008 [2:44:00<01:03, 4.99it/s] 99%|█████████▉| 47691/48008 [2:44:00<01:02, 5.10it/s] 99%|█████████▉| 47692/48008 [2:44:00<01:02, 5.04it/s] 99%|█████████▉| 47693/48008 [2:44:01<01:01, 5.13it/s] 99%|█████████▉| 47694/48008 [2:44:01<01:01, 5.14it/s] 99%|█████████▉| 47695/48008 [2:44:01<01:00, 5.14it/s] 99%|█████████▉| 47696/48008 [2:44:01<01:00, 5.13it/s] 99%|█████████▉| 47697/48008 [2:44:01<01:01, 5.09it/s] 99%|█████████▉| 47698/48008 [2:44:02<01:01, 5.03it/s] 99%|█████████▉| 47699/48008 [2:44:02<01:00, 5.08it/s] 99%|█████████▉| 47700/48008 [2:44:02<00:59, 5.15it/s] {'loss': 0.3006, 'grad_norm': 0.1787109375, 'learning_rate': 1.287285452424596e-06, 'epoch': 0.99} 99%|█████████▉| 47700/48008 [2:44:02<00:59, 5.15it/s] 99%|█████████▉| 47701/48008 [2:44:02<00:58, 5.21it/s] 99%|█████████▉| 47702/48008 [2:44:02<00:58, 5.22it/s] 99%|█████████▉| 47703/48008 [2:44:03<00:58, 5.23it/s] 99%|█████████▉| 47704/48008 [2:44:03<00:58, 5.19it/s] 99%|█████████▉| 47705/48008 [2:44:03<00:59, 5.09it/s] 99%|█████████▉| 47706/48008 [2:44:03<01:03, 4.76it/s] 99%|█████████▉| 47707/48008 [2:44:03<01:02, 4.84it/s] 99%|█████████▉| 47708/48008 [2:44:04<01:00, 4.94it/s] 99%|█████████▉| 47709/48008 [2:44:04<00:59, 5.01it/s] 99%|█████████▉| 47710/48008 [2:44:04<01:00, 4.95it/s] 99%|█████████▉| 47711/48008 [2:44:04<00:58, 5.07it/s] 99%|█████████▉| 47712/48008 [2:44:04<00:58, 5.09it/s] 99%|█████████▉| 47713/48008 [2:44:05<00:58, 5.03it/s] 99%|█████████▉| 47714/48008 [2:44:05<01:01, 4.76it/s] 99%|█████████▉| 47715/48008 [2:44:05<01:00, 4.81it/s] 99%|█████████▉| 47716/48008 [2:44:05<00:59, 4.93it/s] 99%|█████████▉| 47717/48008 [2:44:05<00:58, 4.98it/s] 99%|█████████▉| 47718/48008 [2:44:06<00:58, 5.00it/s] 99%|█████████▉| 47719/48008 [2:44:06<00:57, 5.06it/s] 99%|█████████▉| 47720/48008 [2:44:06<00:56, 5.11it/s] 99%|█████████▉| 47721/48008 [2:44:06<00:55, 5.14it/s] 99%|█████████▉| 47722/48008 [2:44:06<00:55, 5.13it/s] 99%|█████████▉| 47723/48008 [2:44:07<00:54, 5.18it/s] 99%|█████████▉| 47724/48008 [2:44:07<00:54, 5.20it/s] 99%|█████████▉| 47725/48008 [2:44:07<00:54, 5.19it/s] 99%|█████████▉| 47726/48008 [2:44:07<01:04, 4.39it/s] 99%|█████████▉| 47727/48008 [2:44:08<01:00, 4.61it/s] 99%|█████████▉| 47728/48008 [2:44:08<00:58, 4.76it/s] 99%|█████████▉| 47729/48008 [2:44:08<00:57, 4.87it/s] 99%|█████████▉| 47730/48008 [2:44:08<01:12, 3.84it/s] 99%|█████████▉| 47731/48008 [2:44:08<01:06, 4.19it/s] 99%|█████████▉| 47732/48008 [2:44:09<01:01, 4.46it/s] 99%|█████████▉| 47733/48008 [2:44:09<01:00, 4.58it/s] 99%|█████████▉| 47734/48008 [2:44:09<00:58, 4.67it/s] 99%|█████████▉| 47735/48008 [2:44:09<00:56, 4.80it/s] 99%|█████████▉| 47736/48008 [2:44:09<00:54, 4.95it/s] 99%|█████████▉| 47737/48008 [2:44:10<00:54, 4.97it/s] 99%|█████████▉| 47738/48008 [2:44:10<00:54, 4.97it/s] 99%|█████████▉| 47739/48008 [2:44:10<00:53, 5.03it/s] 99%|█████████▉| 47740/48008 [2:44:10<00:53, 5.05it/s] 99%|█████████▉| 47741/48008 [2:44:10<00:52, 5.11it/s] 99%|█████████▉| 47742/48008 [2:44:11<00:51, 5.14it/s] 99%|█████████▉| 47743/48008 [2:44:11<00:51, 5.15it/s] 99%|█████████▉| 47744/48008 [2:44:11<00:50, 5.18it/s] 99%|█████████▉| 47745/48008 [2:44:11<00:51, 5.10it/s] 99%|█████████▉| 47746/48008 [2:44:11<00:51, 5.13it/s] 99%|█████████▉| 47747/48008 [2:44:12<00:51, 5.10it/s] 99%|█████████▉| 47748/48008 [2:44:12<00:50, 5.15it/s] 99%|█████████▉| 47749/48008 [2:44:12<00:50, 5.11it/s] 99%|█████████▉| 47750/48008 [2:44:12<00:50, 5.15it/s] {'loss': 0.302, 'grad_norm': 0.18359375, 'learning_rate': 1.078986835527412e-06, 'epoch': 0.99} 99%|█████████▉| 47750/48008 [2:44:12<00:50, 5.15it/s] 99%|█████████▉| 47751/48008 [2:44:12<00:50, 5.08it/s] 99%|█████████▉| 47752/48008 [2:44:13<00:50, 5.05it/s] 99%|█████████▉| 47753/48008 [2:44:13<00:50, 5.03it/s] 99%|█████████▉| 47754/48008 [2:44:13<00:53, 4.74it/s] 99%|█████████▉| 47755/48008 [2:44:13<00:52, 4.82it/s] 99%|█████████▉| 47756/48008 [2:44:13<00:51, 4.90it/s] 99%|█████████▉| 47757/48008 [2:44:14<00:50, 4.96it/s] 99%|█████████▉| 47758/48008 [2:44:14<00:50, 4.91it/s] 99%|█████████▉| 47759/48008 [2:44:14<00:49, 4.99it/s] 99%|█████████▉| 47760/48008 [2:44:14<01:03, 3.91it/s] 99%|█████████▉| 47761/48008 [2:44:15<00:58, 4.21it/s] 99%|█████████▉| 47762/48008 [2:44:15<00:56, 4.38it/s] 99%|█████████▉| 47763/48008 [2:44:15<00:53, 4.59it/s] 99%|█████████▉| 47764/48008 [2:44:15<01:05, 3.71it/s] 99%|█████████▉| 47765/48008 [2:44:16<00:59, 4.07it/s] 99%|█████████▉| 47766/48008 [2:44:16<00:55, 4.35it/s] 99%|█████████▉| 47767/48008 [2:44:17<02:01, 1.99it/s] 100%|█████████▉| 47768/48008 [2:44:17<01:38, 2.43it/s] 100%|█████████▉| 47769/48008 [2:44:17<01:22, 2.89it/s] 100%|█████████▉| 47770/48008 [2:44:17<01:11, 3.35it/s] 100%|█████████▉| 47771/48008 [2:44:18<01:06, 3.56it/s] 100%|█████████▉| 47772/48008 [2:44:18<01:00, 3.92it/s] 100%|█████████▉| 47773/48008 [2:44:18<00:57, 4.12it/s] 100%|█████████▉| 47774/48008 [2:44:18<00:53, 4.39it/s] 100%|█████████▉| 47775/48008 [2:44:19<00:51, 4.53it/s] 100%|█████████▉| 47776/48008 [2:44:19<00:49, 4.72it/s] 100%|█████████▉| 47777/48008 [2:44:19<00:47, 4.87it/s] 100%|█████████▉| 47778/48008 [2:44:19<00:46, 5.00it/s] 100%|█████████▉| 47779/48008 [2:44:20<01:28, 2.59it/s] 100%|█████████▉| 47780/48008 [2:44:20<01:15, 3.03it/s] 100%|█████████▉| 47781/48008 [2:44:20<01:06, 3.43it/s] 100%|█████████▉| 47782/48008 [2:44:21<00:59, 3.83it/s] 100%|█████████▉| 47783/48008 [2:44:21<00:53, 4.20it/s] 100%|█████████▉| 47784/48008 [2:44:21<00:51, 4.36it/s] 100%|█████████▉| 47785/48008 [2:44:22<01:41, 2.20it/s] 100%|█████████▉| 47786/48008 [2:44:22<01:24, 2.63it/s] 100%|█████████▉| 47787/48008 [2:44:22<01:14, 2.96it/s] 100%|█████████▉| 47788/48008 [2:44:23<01:05, 3.37it/s] 100%|█████████▉| 47789/48008 [2:44:23<00:59, 3.70it/s] 100%|█████████▉| 47790/48008 [2:44:23<00:53, 4.05it/s] 100%|█████████▉| 47791/48008 [2:44:24<01:54, 1.90it/s] 100%|█████████▉| 47792/48008 [2:44:24<01:32, 2.34it/s] 100%|█████████▉| 47793/48008 [2:44:25<01:17, 2.78it/s] 100%|█████████▉| 47794/48008 [2:44:25<01:07, 3.19it/s] 100%|█████████▉| 47795/48008 [2:44:25<00:59, 3.61it/s] 100%|█████████▉| 47796/48008 [2:44:25<00:54, 3.91it/s] 100%|█████████▉| 47797/48008 [2:44:25<00:50, 4.21it/s] 100%|█████████▉| 47798/48008 [2:44:26<00:46, 4.48it/s] 100%|█████████▉| 47799/48008 [2:44:26<00:45, 4.62it/s] 100%|█████████▉| 47800/48008 [2:44:26<00:43, 4.78it/s] {'loss': 0.299, 'grad_norm': 0.15234375, 'learning_rate': 8.706882186302284e-07, 'epoch': 1.0} 100%|█████████▉| 47800/48008 [2:44:26<00:43, 4.78it/s] 100%|█████████▉| 47801/48008 [2:44:26<00:42, 4.82it/s] 100%|█████████▉| 47802/48008 [2:44:26<00:41, 4.96it/s] 100%|█████████▉| 47803/48008 [2:44:26<00:40, 5.04it/s] 100%|█████████▉| 47804/48008 [2:44:27<00:39, 5.13it/s] 100%|█████████▉| 47805/48008 [2:44:27<00:39, 5.15it/s] 100%|█████████▉| 47806/48008 [2:44:27<00:38, 5.19it/s] 100%|█████████▉| 47807/48008 [2:44:27<00:38, 5.20it/s] 100%|█████████▉| 47808/48008 [2:44:27<00:38, 5.19it/s] 100%|█████████▉| 47809/48008 [2:44:28<00:38, 5.19it/s] 100%|█████████▉| 47810/48008 [2:44:28<00:38, 5.18it/s] 100%|█████████▉| 47811/48008 [2:44:28<00:49, 3.98it/s] 100%|█████████▉| 47812/48008 [2:44:28<00:45, 4.27it/s] 100%|█████████▉| 47813/48008 [2:44:29<00:43, 4.46it/s] 100%|█████████▉| 47814/48008 [2:44:29<00:53, 3.66it/s] 100%|█████████▉| 47815/48008 [2:44:29<00:48, 4.00it/s] 100%|█████████▉| 47816/48008 [2:44:29<00:45, 4.23it/s] 100%|█████████▉| 47817/48008 [2:44:30<00:43, 4.42it/s] 100%|█████████▉| 47818/48008 [2:44:30<00:40, 4.68it/s] 100%|█████████▉| 47819/48008 [2:44:30<00:39, 4.78it/s] 100%|█████████▉| 47820/48008 [2:44:30<00:39, 4.81it/s] 100%|█████████▉| 47821/48008 [2:44:30<00:38, 4.86it/s] 100%|█████████▉| 47822/48008 [2:44:31<00:37, 4.93it/s] 100%|█████████▉| 47823/48008 [2:44:31<00:37, 4.99it/s] 100%|█████████▉| 47824/48008 [2:44:31<00:36, 5.01it/s] 100%|█████████▉| 47825/48008 [2:44:31<00:36, 5.00it/s] 100%|█████████▉| 47826/48008 [2:44:31<00:35, 5.06it/s] 100%|█████████▉| 47827/48008 [2:44:32<00:35, 5.16it/s] 100%|█████████▉| 47828/48008 [2:44:32<00:34, 5.19it/s] 100%|█████████▉| 47829/48008 [2:44:32<00:34, 5.17it/s] 100%|█████████▉| 47830/48008 [2:44:32<00:34, 5.11it/s] 100%|█████████▉| 47831/48008 [2:44:32<00:34, 5.08it/s] 100%|█████████▉| 47832/48008 [2:44:33<00:34, 5.10it/s] 100%|█████████▉| 47833/48008 [2:44:33<00:34, 5.11it/s] 100%|█████████▉| 47834/48008 [2:44:33<00:33, 5.13it/s] 100%|█████████▉| 47835/48008 [2:44:33<00:34, 5.05it/s] 100%|█████████▉| 47836/48008 [2:44:33<00:33, 5.08it/s] 100%|█████████▉| 47837/48008 [2:44:34<00:33, 5.10it/s] 100%|█████████▉| 47838/48008 [2:44:34<00:33, 5.12it/s] 100%|█████████▉| 47839/48008 [2:44:34<00:32, 5.17it/s] 100%|█████████▉| 47840/48008 [2:44:34<00:32, 5.12it/s] 100%|█████████▉| 47841/48008 [2:44:34<00:32, 5.11it/s] 100%|█████████▉| 47842/48008 [2:44:34<00:33, 5.00it/s] 100%|█████████▉| 47843/48008 [2:44:35<00:32, 5.10it/s] 100%|█████████▉| 47844/48008 [2:44:35<00:31, 5.13it/s] 100%|█████████▉| 47845/48008 [2:44:35<00:31, 5.21it/s] 100%|█████████▉| 47846/48008 [2:44:35<00:31, 5.21it/s] 100%|█████████▉| 47847/48008 [2:44:35<00:31, 5.16it/s] 100%|█████████▉| 47848/48008 [2:44:36<00:30, 5.22it/s] 100%|█████████▉| 47849/48008 [2:44:36<00:30, 5.14it/s] 100%|█████████▉| 47850/48008 [2:44:36<00:30, 5.21it/s] {'loss': 0.301, 'grad_norm': 0.2109375, 'learning_rate': 6.623896017330445e-07, 'epoch': 1.0} 100%|█████████▉| 47850/48008 [2:44:36<00:30, 5.21it/s] 100%|█████████▉| 47851/48008 [2:44:36<00:29, 5.25it/s] 100%|█████████▉| 47852/48008 [2:44:36<00:29, 5.26it/s] 100%|█████████▉| 47853/48008 [2:44:37<00:29, 5.26it/s] 100%|█████████▉| 47854/48008 [2:44:37<00:29, 5.24it/s] 100%|█████████▉| 47855/48008 [2:44:37<00:29, 5.22it/s] 100%|█████████▉| 47856/48008 [2:44:37<00:29, 5.15it/s] 100%|█████████▉| 47857/48008 [2:44:38<00:58, 2.57it/s] 100%|█████████▉| 47858/48008 [2:44:38<00:51, 2.91it/s] 100%|█████████▉| 47859/48008 [2:44:38<00:44, 3.31it/s] 100%|█████████▉| 47860/48008 [2:44:39<00:39, 3.73it/s] 100%|█████████▉| 47861/48008 [2:44:39<00:36, 4.04it/s] 100%|█████████▉| 47862/48008 [2:44:39<00:34, 4.29it/s] 100%|█████████▉| 47863/48008 [2:44:39<00:32, 4.52it/s] 100%|█████████▉| 47864/48008 [2:44:39<00:30, 4.69it/s] 100%|█████████▉| 47865/48008 [2:44:40<00:29, 4.84it/s] 100%|█████████▉| 47866/48008 [2:44:40<00:29, 4.87it/s] 100%|█████████▉| 47867/48008 [2:44:40<00:29, 4.84it/s] 100%|█████████▉| 47868/48008 [2:44:40<00:28, 4.84it/s] 100%|█████████▉| 47869/48008 [2:44:40<00:27, 4.97it/s] 100%|█████████▉| 47870/48008 [2:44:41<00:27, 4.93it/s] 100%|█████████▉| 47871/48008 [2:44:41<00:27, 5.07it/s] 100%|█████████▉| 47872/48008 [2:44:41<00:26, 5.11it/s] 100%|█████████▉| 47873/48008 [2:44:41<00:26, 5.10it/s] 100%|█████████▉| 47874/48008 [2:44:41<00:26, 5.09it/s] 100%|█████████▉| 47875/48008 [2:44:42<00:26, 5.08it/s] 100%|█████████▉| 47876/48008 [2:44:42<00:25, 5.13it/s] 100%|█████████▉| 47877/48008 [2:44:42<00:25, 5.14it/s] 100%|█████████▉| 47878/48008 [2:44:42<00:25, 5.05it/s] 100%|█████████▉| 47879/48008 [2:44:42<00:25, 5.09it/s] 100%|█████████▉| 47880/48008 [2:44:43<00:24, 5.15it/s] 100%|█████████▉| 47881/48008 [2:44:43<00:24, 5.15it/s] 100%|█████████▉| 47882/48008 [2:44:43<00:24, 5.13it/s] 100%|█████████▉| 47883/48008 [2:44:43<00:24, 5.16it/s] 100%|█████████▉| 47884/48008 [2:44:43<00:23, 5.24it/s] 100%|█████████▉| 47885/48008 [2:44:44<00:27, 4.43it/s] 100%|█████████▉| 47886/48008 [2:44:44<00:26, 4.59it/s] 100%|█████████▉| 47887/48008 [2:44:44<00:25, 4.75it/s] 100%|█████████▉| 47888/48008 [2:44:44<00:24, 4.87it/s] 100%|█████████▉| 47889/48008 [2:44:44<00:23, 4.98it/s] 100%|█████████▉| 47890/48008 [2:44:45<00:23, 5.06it/s] 100%|█████████▉| 47891/48008 [2:44:45<00:22, 5.09it/s] 100%|█████████▉| 47892/48008 [2:44:45<00:24, 4.79it/s] 100%|█████████▉| 47893/48008 [2:44:45<00:23, 4.91it/s] 100%|█████████▉| 47894/48008 [2:44:45<00:22, 5.03it/s] 100%|█████████▉| 47895/48008 [2:44:46<00:22, 5.05it/s] 100%|█████████▉| 47896/48008 [2:44:46<00:22, 4.99it/s] 100%|█████████▉| 47897/48008 [2:44:46<00:21, 5.09it/s] 100%|█████████▉| 47898/48008 [2:44:46<00:21, 5.04it/s] 100%|█████████▉| 47899/48008 [2:44:46<00:21, 5.08it/s] 100%|█████████▉| 47900/48008 [2:44:47<00:21, 5.13it/s] {'loss': 0.301, 'grad_norm': 0.1669921875, 'learning_rate': 4.5409098483586067e-07, 'epoch': 1.0} 100%|█████████▉| 47900/48008 [2:44:47<00:21, 5.13it/s] 100%|█████████▉| 47901/48008 [2:44:47<00:20, 5.14it/s] 100%|█████████▉| 47902/48008 [2:44:47<00:20, 5.16it/s] 100%|█████████▉| 47903/48008 [2:44:47<00:20, 5.02it/s] 100%|█████████▉| 47904/48008 [2:44:47<00:20, 5.08it/s] 100%|█████████▉| 47905/48008 [2:44:48<00:20, 5.02it/s] 100%|█████████▉| 47906/48008 [2:44:48<00:23, 4.33it/s] 100%|█████████▉| 47907/48008 [2:44:48<00:22, 4.56it/s] 100%|█████████▉| 47908/48008 [2:44:48<00:21, 4.74it/s] 100%|█████████▉| 47909/48008 [2:44:48<00:20, 4.86it/s] 100%|█████████▉| 47910/48008 [2:44:49<00:19, 4.95it/s] 100%|█████████▉| 47911/48008 [2:44:49<00:19, 5.04it/s] 100%|█████████▉| 47912/48008 [2:44:49<00:18, 5.08it/s] 100%|█████████▉| 47913/48008 [2:44:49<00:18, 5.11it/s] 100%|█████████▉| 47914/48008 [2:44:49<00:18, 5.10it/s] 100%|█████████▉| 47915/48008 [2:44:50<00:18, 5.09it/s] 100%|█████████▉| 47916/48008 [2:44:50<00:18, 5.04it/s] 100%|█████████▉| 47917/48008 [2:44:50<00:17, 5.09it/s] 100%|█████████▉| 47918/48008 [2:44:50<00:17, 5.10it/s] 100%|█████████▉| 47919/48008 [2:44:50<00:17, 5.11it/s] 100%|█████████▉| 47920/48008 [2:44:51<00:17, 4.98it/s] 100%|█████████▉| 47921/48008 [2:44:51<00:17, 5.03it/s] 100%|█████████▉| 47922/48008 [2:44:51<00:16, 5.12it/s] 100%|█████████▉| 47923/48008 [2:44:51<00:16, 5.14it/s] 100%|█████████▉| 47924/48008 [2:44:51<00:16, 5.19it/s] 100%|█████████▉| 47925/48008 [2:44:52<00:16, 5.15it/s] 100%|█████████▉| 47926/48008 [2:44:52<00:15, 5.17it/s] 100%|█████████▉| 47927/48008 [2:44:52<00:15, 5.23it/s] 100%|█████████▉| 47928/48008 [2:44:52<00:15, 5.22it/s] 100%|█████████▉| 47929/48008 [2:44:53<00:19, 4.00it/s] 100%|█████████▉| 47930/48008 [2:44:53<00:18, 4.27it/s] 100%|█████████▉| 47931/48008 [2:44:53<00:17, 4.49it/s] 100%|█████████▉| 47932/48008 [2:44:53<00:16, 4.69it/s] 100%|█████████▉| 47933/48008 [2:44:54<00:19, 3.77it/s] 100%|█████████▉| 47934/48008 [2:44:54<00:18, 4.10it/s] 100%|█████████▉| 47935/48008 [2:44:54<00:16, 4.32it/s] 100%|█████████▉| 47936/48008 [2:44:54<00:16, 4.44it/s] 100%|█████████▉| 47937/48008 [2:44:54<00:15, 4.60it/s] 100%|█████████▉| 47938/48008 [2:44:55<00:14, 4.77it/s] 100%|█████████▉| 47939/48008 [2:44:55<00:14, 4.87it/s] 100%|█████████▉| 47940/48008 [2:44:55<00:13, 5.00it/s] 100%|█████████▉| 47941/48008 [2:44:55<00:13, 4.97it/s] 100%|█████████▉| 47942/48008 [2:44:55<00:13, 5.02it/s] 100%|█████████▉| 47943/48008 [2:44:56<00:12, 5.12it/s] 100%|█████████▉| 47944/48008 [2:44:56<00:12, 5.11it/s] 100%|█████████▉| 47945/48008 [2:44:56<00:12, 5.18it/s] 100%|█████████▉| 47946/48008 [2:44:56<00:11, 5.20it/s] 100%|█████████▉| 47947/48008 [2:44:56<00:11, 5.11it/s] 100%|█████████▉| 47948/48008 [2:44:56<00:11, 5.14it/s] 100%|█████████▉| 47949/48008 [2:44:57<00:11, 5.11it/s] 100%|█████████▉| 47950/48008 [2:44:57<00:11, 5.14it/s] {'loss': 0.2995, 'grad_norm': 0.486328125, 'learning_rate': 2.457923679386769e-07, 'epoch': 1.0} 100%|█████████▉| 47950/48008 [2:44:57<00:11, 5.14it/s] 100%|█████████▉| 47951/48008 [2:44:57<00:11, 5.15it/s] 100%|█████████▉| 47952/48008 [2:44:57<00:10, 5.19it/s] 100%|█████████▉| 47953/48008 [2:44:57<00:10, 5.18it/s] 100%|█████████▉| 47954/48008 [2:44:58<00:10, 5.27it/s] 100%|█████████▉| 47955/48008 [2:44:58<00:13, 4.02it/s] 100%|█████████▉| 47956/48008 [2:44:58<00:12, 4.29it/s] 100%|█████████▉| 47957/48008 [2:44:58<00:11, 4.54it/s] 100%|█████████▉| 47958/48008 [2:44:59<00:10, 4.73it/s] 100%|█████████▉| 47959/48008 [2:44:59<00:10, 4.85it/s] 100%|█████████▉| 47960/48008 [2:44:59<00:09, 4.97it/s] 100%|█████████▉| 47961/48008 [2:44:59<00:09, 4.97it/s] 100%|█████████▉| 47962/48008 [2:44:59<00:09, 4.96it/s] 100%|█████████▉| 47963/48008 [2:45:00<00:09, 4.95it/s] 100%|█████████▉| 47964/48008 [2:45:00<00:08, 5.05it/s] 100%|█████████▉| 47965/48008 [2:45:00<00:08, 5.07it/s] 100%|█████████▉| 47966/48008 [2:45:00<00:08, 5.14it/s] 100%|█████████▉| 47967/48008 [2:45:00<00:08, 5.12it/s] 100%|█████████▉| 47968/48008 [2:45:01<00:07, 5.18it/s] 100%|█████████▉| 47969/48008 [2:45:01<00:07, 5.18it/s] 100%|█████████▉| 47970/48008 [2:45:01<00:07, 5.14it/s] 100%|█████████▉| 47971/48008 [2:45:01<00:07, 5.18it/s] 100%|█████████▉| 47972/48008 [2:45:01<00:06, 5.19it/s] 100%|█████████▉| 47973/48008 [2:45:01<00:06, 5.20it/s] 100%|█████████▉| 47974/48008 [2:45:02<00:06, 5.11it/s] 100%|█████████▉| 47975/48008 [2:45:02<00:06, 5.11it/s] 100%|█████████▉| 47976/48008 [2:45:02<00:06, 5.07it/s] 100%|█████████▉| 47977/48008 [2:45:02<00:06, 5.09it/s] 100%|█████████▉| 47978/48008 [2:45:02<00:05, 5.03it/s] 100%|█████████▉| 47979/48008 [2:45:03<00:05, 5.03it/s] 100%|█████████▉| 47980/48008 [2:45:03<00:05, 5.06it/s] 100%|█████████▉| 47981/48008 [2:45:03<00:05, 5.09it/s] 100%|█████████▉| 47982/48008 [2:45:03<00:05, 5.18it/s] 100%|█████████▉| 47983/48008 [2:45:03<00:04, 5.20it/s] 100%|█████████▉| 47984/48008 [2:45:04<00:04, 5.13it/s] 100%|█████████▉| 47985/48008 [2:45:04<00:04, 5.15it/s] 100%|█████████▉| 47986/48008 [2:45:04<00:04, 5.14it/s] 100%|█████████▉| 47987/48008 [2:45:04<00:04, 5.20it/s] 100%|█████████▉| 47988/48008 [2:45:04<00:03, 5.06it/s] 100%|█████████▉| 47989/48008 [2:45:05<00:03, 5.08it/s] 100%|█████████▉| 47990/48008 [2:45:05<00:03, 4.98it/s] 100%|█████████▉| 47991/48008 [2:45:05<00:03, 5.05it/s] 100%|█████████▉| 47992/48008 [2:45:05<00:03, 5.06it/s] 100%|█████████▉| 47993/48008 [2:45:05<00:02, 5.10it/s] 100%|█████████▉| 47994/48008 [2:45:06<00:02, 5.08it/s] 100%|█████████▉| 47995/48008 [2:45:06<00:02, 5.09it/s] 100%|█████████▉| 47996/48008 [2:45:06<00:02, 5.12it/s] 100%|█████████▉| 47997/48008 [2:45:06<00:02, 5.14it/s] 100%|█████████▉| 47998/48008 [2:45:06<00:01, 5.20it/s] 100%|█████████▉| 47999/48008 [2:45:07<00:01, 5.06it/s] 100%|█████████▉| 48000/48008 [2:45:07<00:01, 5.14it/s] {'loss': 0.3027, 'grad_norm': 0.341796875, 'learning_rate': 3.749375104149309e-08, 'epoch': 1.0} 100%|█████████▉| 48000/48008 [2:45:07<00:01, 5.14it/s] 100%|█████████▉| 48001/48008 [2:45:07<00:01, 5.18it/s] 100%|█████████▉| 48002/48008 [2:45:07<00:01, 5.16it/s] 100%|█████████▉| 48003/48008 [2:45:07<00:00, 5.15it/s] 100%|█████████▉| 48004/48008 [2:45:08<00:00, 5.16it/s] 100%|█████████▉| 48005/48008 [2:45:08<00:00, 5.12it/s] 100%|█████████▉| 48006/48008 [2:45:08<00:00, 5.11it/s] 100%|█████████▉| 48007/48008 [2:45:09<00:00, 2.48it/s] 100%|██████████| 48008/48008 [2:45:09<00:00, 2.95it/s] {'train_runtime': 9912.1103, 'train_samples_per_second': 1239.897, 'train_steps_per_second': 4.843, 'train_loss': 0.3087104031829472, 'epoch': 1.0} 100%|██████████| 48008/48008 [2:45:09<00:00, 2.95it/s] 100%|██████████| 48008/48008 [2:45:09<00:00, 4.84it/s] Upload 3 LFS files: 0%| | 0/3 [00:00