diff --git "a/log.txt" "b/log.txt"
new file mode 100644--- /dev/null
+++ "b/log.txt"
@@ -0,0 +1,930 @@
+Running 1 job
+{
+    "type": "ui_trainer",
+    "training_folder": "/home/featurize/ai-toolkit/output",
+    "sqlite_db_path": "/home/featurize/ai-toolkit/aitk_db.db",
+    "device": "cuda",
+    "trigger_word": null,
+    "performance_log_every": 10,
+    "network": {
+        "type": "lora",
+        "linear": 32,
+        "linear_alpha": 32,
+        "conv": 16,
+        "conv_alpha": 16,
+        "lokr_full_rank": true,
+        "lokr_factor": -1,
+        "network_kwargs": {
+            "ignore_if_contains": []
+        }
+    },
+    "save": {
+        "dtype": "bf16",
+        "save_every": 250,
+        "max_step_saves_to_keep": 4000,
+        "save_format": "diffusers",
+        "push_to_hub": false
+    },
+    "datasets": [
+        {
+            "folder_path": "/home/featurize/ai-toolkit/datasets/Xiang_Handsome_Flux_SRPO_Keye_Captioned",
+            "control_path": null,
+            "mask_path": null,
+            "mask_min_value": 0.1,
+            "default_caption": "",
+            "caption_ext": "txt",
+            "caption_dropout_rate": 0.05,
+            "cache_latents_to_disk": false,
+            "is_reg": false,
+            "network_weight": 1,
+            "resolution": [
+                512,
+                768
+            ],
+            "controls": [],
+            "shrink_video_to_frames": true,
+            "num_frames": 1,
+            "do_i2v": true,
+            "flip_x": false,
+            "flip_y": false
+        }
+    ],
+    "train": {
+        "batch_size": 1,
+        "bypass_guidance_embedding": false,
+        "steps": 30000,
+        "gradient_accumulation": 1,
+        "train_unet": true,
+        "train_text_encoder": false,
+        "gradient_checkpointing": true,
+        "noise_scheduler": "flowmatch",
+        "optimizer": "adamw8bit",
+        "timestep_type": "linear",
+        "content_or_style": "balanced",
+        "optimizer_params": {
+            "weight_decay": 0.0001
+        },
+        "unload_text_encoder": false,
+        "cache_text_embeddings": true,
+        "lr": 0.0001,
+        "ema_config": {
+            "use_ema": false,
+            "ema_decay": 0.99
+        },
+        "skip_first_sample": true,
+        "force_first_sample": false,
+        "disable_sampling": true,
+        "dtype": "bf16",
+        "diff_output_preservation": false,
+        "diff_output_preservation_multiplier": 1,
+        "diff_output_preservation_class": "person",
+        "switch_boundary_every": 1
+    },
+    "model": {
+        "name_or_path": "/home/featurize/ai-toolkit/Wan2.2-T2V-A14B-Diffusers-bf16",
+        "quantize": true,
+        "qtype": "uint4|ostris/accuracy_recovery_adapters/wan22_14b_t2i_torchao_uint4.safetensors",
+        "quantize_te": true,
+        "qtype_te": "qfloat8",
+        "arch": "wan22_14b:t2v",
+        "low_vram": true,
+        "model_kwargs": {
+            "train_high_noise": true,
+            "train_low_noise": true
+        }
+    },
+    "sample": {
+        "sampler": "flowmatch",
+        "sample_every": 250,
+        "width": 1024,
+        "height": 1024,
+        "samples": [
+            {
+                "prompt": "woman with red hair, playing chess at the park, bomb going off in the background"
+            },
+            {
+                "prompt": "a woman holding a coffee cup, in a beanie, sitting at a cafe"
+            },
+            {
+                "prompt": "a horse is a DJ at a night club, fish eye lens, smoke machine, lazer lights, holding a martini"
+            },
+            {
+                "prompt": "a man showing off his cool new t shirt at the beach, a shark is jumping out of the water in the background"
+            },
+            {
+                "prompt": "a bear building a log cabin in the snow covered mountains"
+            },
+            {
+                "prompt": "woman playing the guitar, on stage, singing a song, laser lights, punk rocker"
+            },
+            {
+                "prompt": "hipster man with a beard, building a chair, in a wood shop"
+            },
+            {
+                "prompt": "photo of a man, white background, medium shot, modeling clothing, studio lighting, white backdrop"
+            },
+            {
+                "prompt": "a man holding a sign that says, 'this is a sign'"
+            },
+            {
+                "prompt": "a bulldog, in a post apocalyptic world, with a shotgun, in a leather jacket, in a desert, with a motorcycle"
+            }
+        ],
+        "neg": "",
+        "seed": 42,
+        "walk_seed": true,
+        "guidance_scale": 4,
+        "sample_steps": 25,
+        "num_frames": 41,
+        "fps": 16
+    }
+}
+Changing sample extention to animated webp
+Using SQLite database at /home/featurize/ai-toolkit/aitk_db.db
+Job ID: "89cb31e4-aeb9-47a6-91b8-d96bd8f03046"
+
+#############################################
+# Running job: Xiang_wan_2_2_lora
+#############################################
+
+
+Running  1 process
+Loading Wan model
+Loading transformer 1
+
Loading checkpoint shards:   0%|                                                                                 | 0/3 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|                                                                                 | 0/3 [00:00<?, ?it/s]
Loading checkpoint shards:  33%|########################3                                                | 1/3 [00:00<00:00,  4.96it/s]
Loading checkpoint shards:  33%|########################3                                                | 1/3 [00:00<00:00,  4.96it/s]
Loading checkpoint shards:  67%|################################################6                        | 2/3 [00:00<00:00,  5.05it/s]
Loading checkpoint shards:  67%|################################################6                        | 2/3 [00:00<00:00,  5.05it/s]
Loading checkpoint shards: 100%|#########################################################################| 3/3 [00:00<00:00,  5.72it/s]
Loading checkpoint shards: 100%|#########################################################################| 3/3 [00:00<00:00,  5.72it/s]
Loading checkpoint shards: 100%|#########################################################################| 3/3 [00:00<00:00,  5.51it/s]
Loading checkpoint shards: 100%|#########################################################################| 3/3 [00:00<00:00,  5.51it/s]
+
+Moving transformer 1 to CPU
+Loading transformer 2
+
Loading checkpoint shards:   0%|                                                                                 | 0/3 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|                                                                                 | 0/3 [00:00<?, ?it/s]
Loading checkpoint shards:  33%|########################3                                                | 1/3 [00:00<00:00,  5.19it/s]
Loading checkpoint shards:  33%|########################3                                                | 1/3 [00:00<00:00,  5.19it/s]
Loading checkpoint shards:  67%|################################################6                        | 2/3 [00:00<00:00,  5.67it/s]
Loading checkpoint shards:  67%|################################################6                        | 2/3 [00:00<00:00,  5.67it/s]
Loading checkpoint shards: 100%|#########################################################################| 3/3 [00:00<00:00,  6.14it/s]
Loading checkpoint shards: 100%|#########################################################################| 3/3 [00:00<00:00,  6.14it/s]
Loading checkpoint shards: 100%|#########################################################################| 3/3 [00:00<00:00,  5.94it/s]
Loading checkpoint shards: 100%|#########################################################################| 3/3 [00:00<00:00,  5.94it/s]
+
+Moving transformer 2 to CPU
+Creating DualWanTransformer3DModel
+Applying Accuracy Recovery Adapter to Transformers
+Grabbing lora from the hub: ostris/accuracy_recovery_adapters/wan22_14b_t2i_torchao_uint4.safetensors
+
wan22_14b_t2i_torchao_uint4.safetensors:   0%|                                                              | 0.00/312M [00:00<?, ?B/s]
wan22_14b_t2i_torchao_uint4.safetensors:   0%|                                                              | 0.00/312M [00:00<?, ?B/s]
wan22_14b_t2i_torchao_uint4.safetensors:  14%|#######4                                             | 43.6M/312M [00:06<00:38, 6.94MB/s]
wan22_14b_t2i_torchao_uint4.safetensors:  14%|#######4                                             | 43.6M/312M [00:06<00:38, 6.94MB/s]
wan22_14b_t2i_torchao_uint4.safetensors:  35%|###################1                                  | 111M/312M [00:08<00:12, 15.5MB/s]
wan22_14b_t2i_torchao_uint4.safetensors:  35%|###################1                                  | 111M/312M [00:08<00:12, 15.5MB/s]
wan22_14b_t2i_torchao_uint4.safetensors:  57%|##############################7                       | 178M/312M [00:09<00:06, 22.1MB/s]
wan22_14b_t2i_torchao_uint4.safetensors:  57%|##############################7                       | 178M/312M [00:09<00:06, 22.1MB/s]
wan22_14b_t2i_torchao_uint4.safetensors:  78%|##########################################3           | 245M/312M [00:10<00:01, 36.2MB/s]
wan22_14b_t2i_torchao_uint4.safetensors:  78%|##########################################3           | 245M/312M [00:10<00:01, 36.2MB/s]
wan22_14b_t2i_torchao_uint4.safetensors: 100%|######################################################| 312M/312M [00:10<00:00, 48.6MB/s]
wan22_14b_t2i_torchao_uint4.safetensors: 100%|######################################################| 312M/312M [00:10<00:00, 48.6MB/s]
wan22_14b_t2i_torchao_uint4.safetensors: 100%|######################################################| 312M/312M [00:10<00:00, 29.2MB/s]
wan22_14b_t2i_torchao_uint4.safetensors: 100%|######################################################| 312M/312M [00:10<00:00, 29.2MB/s]
+
+create LoRA network. base dim (rank): 16, alpha: 16
+neuron dropout: p=None, rank dropout: p=None, module dropout: p=None
+create LoRA for Text Encoder: 0 modules.
+create LoRA for U-Net: 812 modules.
+enable LoRA for U-Net
+Missing keys: []
+
Attaching quantization:   0%|                                                                                  | 0/812 [00:00<?, ?it/s]
Attaching quantization:   0%|                                                                                  | 0/812 [00:00<?, ?it/s]
Attaching quantization:   0%|2                                                                         | 3/812 [00:00<02:54,  4.64it/s]
Attaching quantization:   0%|2                                                                         | 3/812 [00:00<02:54,  4.64it/s]
Attaching quantization:   1%|4                                                                         | 5/812 [00:00<01:58,  6.83it/s]
Attaching quantization:   1%|4                                                                         | 5/812 [00:00<01:58,  6.83it/s]
Attaching quantization:   1%|6                                                                         | 7/812 [00:00<01:31,  8.82it/s]
Attaching quantization:   1%|6                                                                         | 7/812 [00:00<01:31,  8.82it/s]
Attaching quantization:   1%|8                                                                         | 9/812 [00:01<01:23,  9.62it/s]
Attaching quantization:   1%|8                                                                         | 9/812 [00:01<01:23,  9.62it/s]
Attaching quantization:   1%|9                                                                        | 11/812 [00:01<01:18, 10.21it/s]
Attaching quantization:   1%|9                                                                        | 11/812 [00:01<01:18, 10.21it/s]
Attaching quantization:   2%|#1                                                                       | 13/812 [00:01<01:12, 11.03it/s]
Attaching quantization:   2%|#1                                                                       | 13/812 [00:01<01:12, 11.03it/s]
Attaching quantization:   2%|#3                                                                       | 15/812 [00:01<01:51,  7.16it/s]
Attaching quantization:   2%|#3                                                                       | 15/812 [00:01<01:51,  7.16it/s]
Attaching quantization:   2%|#5                                                                       | 17/812 [00:02<01:35,  8.36it/s]
Attaching quantization:   2%|#5                                                                       | 17/812 [00:02<01:35,  8.36it/s]
Attaching quantization:   2%|#7                                                                       | 19/812 [00:02<01:30,  8.77it/s]
Attaching quantization:   2%|#7                                                                       | 19/812 [00:02<01:30,  8.77it/s]
Attaching quantization:   3%|#8                                                                       | 21/812 [00:02<01:23,  9.43it/s]
Attaching quantization:   3%|#8                                                                       | 21/812 [00:02<01:23,  9.43it/s]
Attaching quantization:   3%|##                                                                       | 23/812 [00:02<01:20,  9.86it/s]
Attaching quantization:   3%|##                                                                       | 23/812 [00:02<01:20,  9.86it/s]
Attaching quantization:   3%|##2                                                                      | 25/812 [00:03<01:56,  6.77it/s]
Attaching quantization:   3%|##2                                                                      | 25/812 [00:03<01:56,  6.77it/s]
Attaching quantization:   3%|##4                                                                      | 27/812 [00:03<01:44,  7.54it/s]
Attaching quantization:   3%|##4                                                                      | 27/812 [00:03<01:44,  7.54it/s]
Attaching quantization:   4%|##6                                                                      | 29/812 [00:03<01:32,  8.48it/s]
Attaching quantization:   4%|##6                                                                      | 29/812 [00:03<01:32,  8.48it/s]
Attaching quantization:   4%|##7                                                                      | 31/812 [00:03<01:26,  9.07it/s]
Attaching quantization:   4%|##7                                                                      | 31/812 [00:03<01:26,  9.07it/s]
Attaching quantization:   4%|##9                                                                      | 33/812 [00:03<01:22,  9.39it/s]
Attaching quantization:   4%|##9                                                                      | 33/812 [00:03<01:22,  9.39it/s]
Attaching quantization:   4%|###1                                                                     | 35/812 [00:04<02:04,  6.23it/s]
Attaching quantization:   4%|###1                                                                     | 35/812 [00:04<02:04,  6.23it/s]
Attaching quantization:   5%|###3                                                                     | 37/812 [00:04<01:43,  7.46it/s]
Attaching quantization:   5%|###3                                                                     | 37/812 [00:04<01:43,  7.46it/s]
Attaching quantization:   5%|###4                                                                     | 38/812 [00:04<01:39,  7.76it/s]
Attaching quantization:   5%|###4                                                                     | 38/812 [00:04<01:39,  7.76it/s]
Attaching quantization:   5%|###5                                                                     | 39/812 [00:04<01:37,  7.90it/s]
Attaching quantization:   5%|###5                                                                     | 39/812 [00:04<01:37,  7.90it/s]
Attaching quantization:   5%|###6                                                                     | 41/812 [00:05<01:28,  8.71it/s]
Attaching quantization:   5%|###6                                                                     | 41/812 [00:05<01:28,  8.71it/s]
Attaching quantization:   5%|###8                                                                     | 43/812 [00:05<01:19,  9.64it/s]
Attaching quantization:   5%|###8                                                                     | 43/812 [00:05<01:19,  9.64it/s]
Attaching quantization:   6%|####                                                                     | 45/812 [00:05<02:02,  6.26it/s]
Attaching quantization:   6%|####                                                                     | 45/812 [00:05<02:02,  6.26it/s]
Attaching quantization:   6%|####2                                                                    | 47/812 [00:05<01:42,  7.49it/s]
Attaching quantization:   6%|####2                                                                    | 47/812 [00:05<01:42,  7.49it/s]
Attaching quantization:   6%|####4                                                                    | 49/812 [00:06<01:25,  8.95it/s]
Attaching quantization:   6%|####4                                                                    | 49/812 [00:06<01:25,  8.95it/s]
Attaching quantization:   6%|####5                                                                    | 51/812 [00:06<01:19,  9.60it/s]
Attaching quantization:   6%|####5                                                                    | 51/812 [00:06<01:19,  9.60it/s]
Attaching quantization:   7%|####7                                                                    | 53/812 [00:06<01:17,  9.81it/s]
Attaching quantization:   7%|####7                                                                    | 53/812 [00:06<01:17,  9.81it/s]
Attaching quantization:   7%|####9                                                                    | 55/812 [00:06<02:01,  6.25it/s]
Attaching quantization:   7%|####9                                                                    | 55/812 [00:06<02:01,  6.25it/s]
Attaching quantization:   7%|#####1                                                                   | 57/812 [00:07<01:43,  7.27it/s]
Attaching quantization:   7%|#####1                                                                   | 57/812 [00:07<01:43,  7.27it/s]
Attaching quantization:   7%|#####3                                                                   | 59/812 [00:07<01:30,  8.35it/s]
Attaching quantization:   7%|#####3                                                                   | 59/812 [00:07<01:30,  8.35it/s]
Attaching quantization:   8%|#####4                                                                   | 61/812 [00:07<01:20,  9.30it/s]
Attaching quantization:   8%|#####4                                                                   | 61/812 [00:07<01:20,  9.30it/s]
Attaching quantization:   8%|#####6                                                                   | 63/812 [00:07<01:15,  9.95it/s]
Attaching quantization:   8%|#####6                                                                   | 63/812 [00:07<01:15,  9.95it/s]
Attaching quantization:   8%|#####8                                                                   | 65/812 [00:08<01:53,  6.61it/s]
Attaching quantization:   8%|#####8                                                                   | 65/812 [00:08<01:53,  6.61it/s]
Attaching quantization:   8%|######                                                                   | 67/812 [00:08<01:35,  7.77it/s]
Attaching quantization:   8%|######                                                                   | 67/812 [00:08<01:35,  7.77it/s]
Attaching quantization:   8%|######2                                                                  | 69/812 [00:08<01:28,  8.40it/s]
Attaching quantization:   8%|######2                                                                  | 69/812 [00:08<01:28,  8.40it/s]
Attaching quantization:   9%|######3                                                                  | 71/812 [00:08<01:20,  9.15it/s]
Attaching quantization:   9%|######3                                                                  | 71/812 [00:08<01:20,  9.15it/s]
Attaching quantization:   9%|######5                                                                  | 73/812 [00:08<01:20,  9.19it/s]
Attaching quantization:   9%|######5                                                                  | 73/812 [00:08<01:20,  9.19it/s]
Attaching quantization:   9%|######7                                                                  | 75/812 [00:09<02:07,  5.79it/s]
Attaching quantization:   9%|######7                                                                  | 75/812 [00:09<02:07,  5.79it/s]
Attaching quantization:   9%|######9                                                                  | 77/812 [00:09<01:49,  6.73it/s]
Attaching quantization:   9%|######9                                                                  | 77/812 [00:09<01:49,  6.73it/s]
Attaching quantization:  10%|#######                                                                  | 78/812 [00:09<01:44,  7.03it/s]
Attaching quantization:  10%|#######                                                                  | 78/812 [00:09<01:44,  7.03it/s]
Attaching quantization:  10%|#######1                                                                 | 79/812 [00:09<01:40,  7.26it/s]
Attaching quantization:  10%|#######1                                                                 | 79/812 [00:09<01:40,  7.26it/s]
Attaching quantization:  10%|#######2                                                                 | 81/812 [00:10<01:29,  8.14it/s]
Attaching quantization:  10%|#######2                                                                 | 81/812 [00:10<01:29,  8.14it/s]
Attaching quantization:  10%|#######3                                                                 | 82/812 [00:10<01:27,  8.37it/s]
Attaching quantization:  10%|#######3                                                                 | 82/812 [00:10<01:27,  8.37it/s]
Attaching quantization:  10%|#######4                                                                 | 83/812 [00:10<01:24,  8.62it/s]
Attaching quantization:  10%|#######4                                                                 | 83/812 [00:10<01:24,  8.62it/s]
Attaching quantization:  10%|#######5                                                                 | 84/812 [00:10<02:07,  5.72it/s]
Attaching quantization:  10%|#######5                                                                 | 84/812 [00:10<02:07,  5.72it/s]
Attaching quantization:  10%|#######6                                                                 | 85/812 [00:10<02:25,  5.01it/s]
Attaching quantization:  10%|#######6                                                                 | 85/812 [00:10<02:25,  5.01it/s]
Attaching quantization:  11%|#######7                                                                 | 86/812 [00:11<02:07,  5.71it/s]
Attaching quantization:  11%|#######7                                                                 | 86/812 [00:11<02:07,  5.71it/s]
Attaching quantization:  11%|#######9                                                                 | 88/812 [00:11<01:39,  7.25it/s]
Attaching quantization:  11%|#######9                                                                 | 88/812 [00:11<01:39,  7.25it/s]
Attaching quantization:  11%|########                                                                 | 89/812 [00:11<01:33,  7.74it/s]
Attaching quantization:  11%|########                                                                 | 89/812 [00:11<01:33,  7.74it/s]
Attaching quantization:  11%|########1                                                                | 91/812 [00:11<01:24,  8.55it/s]
Attaching quantization:  11%|########1                                                                | 91/812 [00:11<01:24,  8.55it/s]
Attaching quantization:  11%|########3                                                                | 93/812 [00:11<01:17,  9.29it/s]
Attaching quantization:  11%|########3                                                                | 93/812 [00:11<01:17,  9.29it/s]
Attaching quantization:  12%|########4                                                                | 94/812 [00:12<01:45,  6.83it/s]
Attaching quantization:  12%|########4                                                                | 94/812 [00:12<01:45,  6.83it/s]
Attaching quantization:  12%|########5                                                                | 95/812 [00:12<02:05,  5.70it/s]
Attaching quantization:  12%|########5                                                                | 95/812 [00:12<02:05,  5.70it/s]
Attaching quantization:  12%|########7                                                                | 97/812 [00:12<01:37,  7.32it/s]
Attaching quantization:  12%|########7                                                                | 97/812 [00:12<01:37,  7.32it/s]
Attaching quantization:  12%|########8                                                                | 98/812 [00:12<01:36,  7.37it/s]
Attaching quantization:  12%|########8                                                                | 98/812 [00:12<01:36,  7.37it/s]
Attaching quantization:  12%|########8                                                               | 100/812 [00:12<01:24,  8.40it/s]
Attaching quantization:  12%|########8                                                               | 100/812 [00:12<01:24,  8.40it/s]
Attaching quantization:  12%|########9                                                               | 101/812 [00:12<01:22,  8.59it/s]
Attaching quantization:  12%|########9                                                               | 101/812 [00:12<01:22,  8.59it/s]
Attaching quantization:  13%|#########1                                                              | 103/812 [00:13<01:11,  9.88it/s]
Attaching quantization:  13%|#########1                                                              | 103/812 [00:13<01:11,  9.88it/s]
Attaching quantization:  13%|#########3                                                              | 105/812 [00:13<01:56,  6.05it/s]
Attaching quantization:  13%|#########3                                                              | 105/812 [00:13<01:56,  6.05it/s]
Attaching quantization:  13%|#########4                                                              | 107/812 [00:13<01:37,  7.24it/s]
Attaching quantization:  13%|#########4                                                              | 107/812 [00:13<01:37,  7.24it/s]
Attaching quantization:  13%|#########6                                                              | 109/812 [00:13<01:26,  8.15it/s]
Attaching quantization:  13%|#########6                                                              | 109/812 [00:13<01:26,  8.15it/s]
Attaching quantization:  14%|#########8                                                              | 111/812 [00:14<01:18,  8.98it/s]
Attaching quantization:  14%|#########8                                                              | 111/812 [00:14<01:18,  8.98it/s]
Attaching quantization:  14%|##########                                                              | 113/812 [00:14<01:20,  8.63it/s]
Attaching quantization:  14%|##########                                                              | 113/812 [00:14<01:20,  8.63it/s]
Attaching quantization:  14%|##########1                                                             | 114/812 [00:14<01:42,  6.80it/s]
Attaching quantization:  14%|##########1                                                             | 114/812 [00:14<01:42,  6.80it/s]
Attaching quantization:  14%|##########1                                                             | 115/812 [00:14<01:59,  5.85it/s]
Attaching quantization:  14%|##########1                                                             | 115/812 [00:14<01:59,  5.85it/s]
Attaching quantization:  14%|##########3                                                             | 117/812 [00:15<01:36,  7.22it/s]
Attaching quantization:  14%|##########3                                                             | 117/812 [00:15<01:36,  7.22it/s]
Attaching quantization:  15%|##########5                                                             | 119/812 [00:15<01:22,  8.40it/s]
Attaching quantization:  15%|##########5                                                             | 119/812 [00:15<01:22,  8.40it/s]
Attaching quantization:  15%|##########6                                                             | 120/812 [00:15<01:20,  8.62it/s]
Attaching quantization:  15%|##########6                                                             | 120/812 [00:15<01:20,  8.62it/s]
Attaching quantization:  15%|##########7                                                             | 121/812 [00:15<01:18,  8.78it/s]
Attaching quantization:  15%|##########7                                                             | 121/812 [00:15<01:18,  8.78it/s]
Attaching quantization:  15%|##########9                                                             | 123/812 [00:15<01:12,  9.48it/s]
Attaching quantization:  15%|##########9                                                             | 123/812 [00:15<01:12,  9.48it/s]
Attaching quantization:  15%|##########9                                                             | 124/812 [00:15<01:37,  7.03it/s]
Attaching quantization:  15%|##########9                                                             | 124/812 [00:15<01:37,  7.03it/s]
Attaching quantization:  15%|###########                                                             | 125/812 [00:16<01:56,  5.90it/s]
Attaching quantization:  15%|###########                                                             | 125/812 [00:16<01:56,  5.90it/s]
Attaching quantization:  16%|###########2                                                            | 127/812 [00:16<01:31,  7.51it/s]
Attaching quantization:  16%|###########2                                                            | 127/812 [00:16<01:31,  7.51it/s]
Attaching quantization:  16%|###########4                                                            | 129/812 [00:16<01:17,  8.81it/s]
Attaching quantization:  16%|###########4                                                            | 129/812 [00:16<01:17,  8.81it/s]
Attaching quantization:  16%|###########5                                                            | 130/812 [00:16<01:17,  8.78it/s]
Attaching quantization:  16%|###########5                                                            | 130/812 [00:16<01:17,  8.78it/s]
Attaching quantization:  16%|###########7                                                            | 132/812 [00:16<01:10,  9.65it/s]
Attaching quantization:  16%|###########7                                                            | 132/812 [00:16<01:10,  9.65it/s]
Attaching quantization:  17%|###########8                                                            | 134/812 [00:17<01:33,  7.27it/s]
Attaching quantization:  17%|###########8                                                            | 134/812 [00:17<01:33,  7.27it/s]
Attaching quantization:  17%|###########9                                                            | 135/812 [00:17<01:49,  6.18it/s]
Attaching quantization:  17%|###########9                                                            | 135/812 [00:17<01:49,  6.18it/s]
Attaching quantization:  17%|############1                                                           | 137/812 [00:17<01:33,  7.19it/s]
Attaching quantization:  17%|############1                                                           | 137/812 [00:17<01:33,  7.19it/s]
Attaching quantization:  17%|############2                                                           | 138/812 [00:17<01:30,  7.43it/s]
Attaching quantization:  17%|############2                                                           | 138/812 [00:17<01:30,  7.43it/s]
Attaching quantization:  17%|############4                                                           | 140/812 [00:17<01:19,  8.50it/s]
Attaching quantization:  17%|############4                                                           | 140/812 [00:17<01:19,  8.50it/s]
Attaching quantization:  17%|############5                                                           | 142/812 [00:18<01:10,  9.48it/s]
Attaching quantization:  17%|############5                                                           | 142/812 [00:18<01:10,  9.48it/s]
Attaching quantization:  18%|############7                                                           | 144/812 [00:18<01:31,  7.34it/s]
Attaching quantization:  18%|############7                                                           | 144/812 [00:18<01:31,  7.34it/s]
Attaching quantization:  18%|############8                                                           | 145/812 [00:18<01:45,  6.35it/s]
Attaching quantization:  18%|############8                                                           | 145/812 [00:18<01:45,  6.35it/s]
Attaching quantization:  18%|############9                                                           | 146/812 [00:18<01:40,  6.63it/s]
Attaching quantization:  18%|############9                                                           | 146/812 [00:18<01:40,  6.63it/s]
Attaching quantization:  18%|#############1                                                          | 148/812 [00:19<01:23,  7.93it/s]
Attaching quantization:  18%|#############1                                                          | 148/812 [00:19<01:23,  7.93it/s]
Attaching quantization:  18%|#############3                                                          | 150/812 [00:19<01:14,  8.84it/s]
Attaching quantization:  18%|#############3                                                          | 150/812 [00:19<01:14,  8.84it/s]
Attaching quantization:  19%|#############4                                                          | 152/812 [00:19<01:09,  9.51it/s]
Attaching quantization:  19%|#############4                                                          | 152/812 [00:19<01:09,  9.51it/s]
Attaching quantization:  19%|#############5                                                          | 153/812 [00:19<01:08,  9.58it/s]
Attaching quantization:  19%|#############5                                                          | 153/812 [00:19<01:08,  9.58it/s]
Attaching quantization:  19%|#############6                                                          | 154/812 [00:19<01:38,  6.66it/s]
Attaching quantization:  19%|#############6                                                          | 154/812 [00:19<01:38,  6.66it/s]
Attaching quantization:  19%|#############7                                                          | 155/812 [00:20<02:09,  5.05it/s]
Attaching quantization:  19%|#############7                                                          | 155/812 [00:20<02:09,  5.05it/s]
Attaching quantization:  19%|#############9                                                          | 157/812 [00:20<01:39,  6.56it/s]
Attaching quantization:  19%|#############9                                                          | 157/812 [00:20<01:39,  6.56it/s]
Attaching quantization:  20%|##############                                                          | 159/812 [00:20<01:21,  7.97it/s]
Attaching quantization:  20%|##############                                                          | 159/812 [00:20<01:21,  7.97it/s]
Attaching quantization:  20%|##############2                                                         | 161/812 [00:20<01:11,  9.12it/s]
Attaching quantization:  20%|##############2                                                         | 161/812 [00:20<01:11,  9.12it/s]
Attaching quantization:  20%|##############4                                                         | 163/812 [00:20<01:10,  9.25it/s]
Attaching quantization:  20%|##############4                                                         | 163/812 [00:20<01:10,  9.25it/s]
Attaching quantization:  20%|##############6                                                         | 165/812 [00:21<01:49,  5.93it/s]
Attaching quantization:  20%|##############6                                                         | 165/812 [00:21<01:49,  5.93it/s]
Attaching quantization:  21%|##############8                                                         | 167/812 [00:21<01:31,  7.05it/s]
Attaching quantization:  21%|##############8                                                         | 167/812 [00:21<01:31,  7.05it/s]
Attaching quantization:  21%|##############9                                                         | 169/812 [00:21<01:19,  8.10it/s]
Attaching quantization:  21%|##############9                                                         | 169/812 [00:21<01:19,  8.10it/s]
Attaching quantization:  21%|###############1                                                        | 171/812 [00:22<01:16,  8.42it/s]
Attaching quantization:  21%|###############1                                                        | 171/812 [00:22<01:16,  8.42it/s]
Attaching quantization:  21%|###############3                                                        | 173/812 [00:22<01:14,  8.55it/s]
Attaching quantization:  21%|###############3                                                        | 173/812 [00:22<01:14,  8.55it/s]
Attaching quantization:  21%|###############4                                                        | 174/812 [00:22<01:38,  6.47it/s]
Attaching quantization:  21%|###############4                                                        | 174/812 [00:22<01:38,  6.47it/s]
Attaching quantization:  22%|###############5                                                        | 175/812 [00:22<01:50,  5.76it/s]
Attaching quantization:  22%|###############5                                                        | 175/812 [00:22<01:50,  5.76it/s]
Attaching quantization:  22%|###############6                                                        | 177/812 [00:22<01:29,  7.13it/s]
Attaching quantization:  22%|###############6                                                        | 177/812 [00:22<01:29,  7.13it/s]
Attaching quantization:  22%|###############8                                                        | 179/812 [00:23<01:17,  8.13it/s]
Attaching quantization:  22%|###############8                                                        | 179/812 [00:23<01:17,  8.13it/s]
Attaching quantization:  22%|###############9                                                        | 180/812 [00:23<01:17,  8.18it/s]
Attaching quantization:  22%|###############9                                                        | 180/812 [00:23<01:17,  8.18it/s]
Attaching quantization:  22%|################                                                        | 181/812 [00:23<01:19,  7.99it/s]
Attaching quantization:  22%|################                                                        | 181/812 [00:23<01:19,  7.99it/s]
Attaching quantization:  23%|################2                                                       | 183/812 [00:23<01:14,  8.44it/s]
Attaching quantization:  23%|################2                                                       | 183/812 [00:23<01:14,  8.44it/s]
Attaching quantization:  23%|################3                                                       | 184/812 [00:23<01:41,  6.18it/s]
Attaching quantization:  23%|################3                                                       | 184/812 [00:23<01:41,  6.18it/s]
Attaching quantization:  23%|################4                                                       | 185/812 [00:24<01:54,  5.46it/s]
Attaching quantization:  23%|################4                                                       | 185/812 [00:24<01:54,  5.46it/s]
Attaching quantization:  23%|################5                                                       | 187/812 [00:24<01:26,  7.22it/s]
Attaching quantization:  23%|################5                                                       | 187/812 [00:24<01:26,  7.22it/s]
Attaching quantization:  23%|################7                                                       | 189/812 [00:24<01:17,  8.06it/s]
Attaching quantization:  23%|################7                                                       | 189/812 [00:24<01:17,  8.06it/s]
Attaching quantization:  24%|################9                                                       | 191/812 [00:24<01:12,  8.59it/s]
Attaching quantization:  24%|################9                                                       | 191/812 [00:24<01:12,  8.59it/s]
Attaching quantization:  24%|#################1                                                      | 193/812 [00:24<01:08,  9.10it/s]
Attaching quantization:  24%|#################1                                                      | 193/812 [00:24<01:08,  9.10it/s]
Attaching quantization:  24%|#################2                                                      | 194/812 [00:25<01:31,  6.75it/s]
Attaching quantization:  24%|#################2                                                      | 194/812 [00:25<01:31,  6.75it/s]
Attaching quantization:  24%|#################2                                                      | 195/812 [00:25<01:42,  6.00it/s]
Attaching quantization:  24%|#################2                                                      | 195/812 [00:25<01:42,  6.00it/s]
Attaching quantization:  24%|#################4                                                      | 197/812 [00:25<01:26,  7.10it/s]
Attaching quantization:  24%|#################4                                                      | 197/812 [00:25<01:26,  7.10it/s]
Attaching quantization:  25%|#################6                                                      | 199/812 [00:25<01:17,  7.95it/s]
Attaching quantization:  25%|#################6                                                      | 199/812 [00:25<01:17,  7.95it/s]
Attaching quantization:  25%|#################8                                                      | 201/812 [00:26<01:08,  8.88it/s]
Attaching quantization:  25%|#################8                                                      | 201/812 [00:26<01:08,  8.88it/s]
Attaching quantization:  25%|##################                                                      | 203/812 [00:26<01:03,  9.67it/s]
Attaching quantization:  25%|##################                                                      | 203/812 [00:26<01:03,  9.67it/s]
Attaching quantization:  25%|##################1                                                     | 205/812 [00:26<01:35,  6.36it/s]
Attaching quantization:  25%|##################1                                                     | 205/812 [00:26<01:35,  6.36it/s]
Attaching quantization:  25%|##################2                                                     | 206/812 [00:26<01:29,  6.80it/s]
Attaching quantization:  25%|##################2                                                     | 206/812 [00:26<01:29,  6.80it/s]
Attaching quantization:  26%|##################4                                                     | 208/812 [00:27<01:17,  7.78it/s]
Attaching quantization:  26%|##################4                                                     | 208/812 [00:27<01:17,  7.78it/s]
Attaching quantization:  26%|##################6                                                     | 210/812 [00:27<01:06,  9.03it/s]
Attaching quantization:  26%|##################6                                                     | 210/812 [00:27<01:06,  9.03it/s]
Attaching quantization:  26%|##################7                                                     | 212/812 [00:27<01:03,  9.45it/s]
Attaching quantization:  26%|##################7                                                     | 212/812 [00:27<01:03,  9.45it/s]
Attaching quantization:  26%|##################9                                                     | 214/812 [00:27<01:28,  6.73it/s]
Attaching quantization:  26%|##################9                                                     | 214/812 [00:27<01:28,  6.73it/s]
Attaching quantization:  26%|###################                                                     | 215/812 [00:28<01:38,  6.03it/s]
Attaching quantization:  26%|###################                                                     | 215/812 [00:28<01:38,  6.03it/s]
Attaching quantization:  27%|###################2                                                    | 217/812 [00:28<01:19,  7.51it/s]
Attaching quantization:  27%|###################2                                                    | 217/812 [00:28<01:19,  7.51it/s]
Attaching quantization:  27%|###################4                                                    | 219/812 [00:28<01:08,  8.71it/s]
Attaching quantization:  27%|###################4                                                    | 219/812 [00:28<01:08,  8.71it/s]
Attaching quantization:  27%|###################5                                                    | 221/812 [00:28<01:03,  9.29it/s]
Attaching quantization:  27%|###################5                                                    | 221/812 [00:28<01:03,  9.29it/s]
Attaching quantization:  27%|###################7                                                    | 223/812 [00:28<00:58, 10.15it/s]
Attaching quantization:  27%|###################7                                                    | 223/812 [00:28<00:58, 10.15it/s]
Attaching quantization:  28%|###################9                                                    | 225/812 [00:29<01:33,  6.27it/s]
Attaching quantization:  28%|###################9                                                    | 225/812 [00:29<01:33,  6.27it/s]
Attaching quantization:  28%|####################1                                                   | 227/812 [00:29<01:16,  7.62it/s]
Attaching quantization:  28%|####################1                                                   | 227/812 [00:29<01:16,  7.62it/s]
Attaching quantization:  28%|####################3                                                   | 229/812 [00:29<01:06,  8.79it/s]
Attaching quantization:  28%|####################3                                                   | 229/812 [00:29<01:06,  8.79it/s]
Attaching quantization:  28%|####################4                                                   | 231/812 [00:29<01:06,  8.70it/s]
Attaching quantization:  28%|####################4                                                   | 231/812 [00:29<01:06,  8.70it/s]
Attaching quantization:  29%|####################6                                                   | 233/812 [00:30<01:05,  8.90it/s]
Attaching quantization:  29%|####################6                                                   | 233/812 [00:30<01:05,  8.90it/s]
Attaching quantization:  29%|####################8                                                   | 235/812 [00:30<01:33,  6.18it/s]
Attaching quantization:  29%|####################8                                                   | 235/812 [00:30<01:33,  6.18it/s]
Attaching quantization:  29%|#####################                                                   | 237/812 [00:30<01:18,  7.29it/s]
Attaching quantization:  29%|#####################                                                   | 237/812 [00:30<01:18,  7.29it/s]
Attaching quantization:  29%|#####################1                                                  | 239/812 [00:31<01:12,  7.87it/s]
Attaching quantization:  29%|#####################1                                                  | 239/812 [00:31<01:12,  7.87it/s]
Attaching quantization:  30%|#####################2                                                  | 240/812 [00:31<01:37,  5.86it/s]
Attaching quantization:  30%|#####################2                                                  | 240/812 [00:31<01:37,  5.86it/s]
Attaching quantization:  30%|#####################3                                                  | 241/812 [00:32<02:42,  3.51it/s]
Attaching quantization:  30%|#####################3                                                  | 241/812 [00:32<02:42,  3.51it/s]
Attaching quantization:  30%|#####################4                                                  | 242/812 [00:32<02:25,  3.92it/s]
Attaching quantization:  30%|#####################4                                                  | 242/812 [00:32<02:25,  3.92it/s]
Attaching quantization:  30%|#####################6                                                  | 244/812 [00:32<02:13,  4.27it/s]
Attaching quantization:  30%|#####################6                                                  | 244/812 [00:32<02:13,  4.27it/s]
Attaching quantization:  30%|#####################7                                                  | 245/812 [00:32<02:17,  4.13it/s]
Attaching quantization:  30%|#####################7                                                  | 245/812 [00:32<02:17,  4.13it/s]
Attaching quantization:  30%|#####################8                                                  | 246/812 [00:33<02:08,  4.39it/s]
Attaching quantization:  30%|#####################8                                                  | 246/812 [00:33<02:08,  4.39it/s]
Attaching quantization:  31%|#####################9                                                  | 248/812 [00:33<01:37,  5.79it/s]
Attaching quantization:  31%|#####################9                                                  | 248/812 [00:33<01:37,  5.79it/s]
Attaching quantization:  31%|######################1                                                 | 250/812 [00:33<01:16,  7.34it/s]
Attaching quantization:  31%|######################1                                                 | 250/812 [00:33<01:16,  7.34it/s]
Attaching quantization:  31%|######################2                                                 | 251/812 [00:33<01:13,  7.67it/s]
Attaching quantization:  31%|######################2                                                 | 251/812 [00:33<01:13,  7.67it/s]
Attaching quantization:  31%|######################4                                                 | 253/812 [00:33<01:02,  8.94it/s]
Attaching quantization:  31%|######################4                                                 | 253/812 [00:33<01:02,  8.94it/s]
Attaching quantization:  31%|######################6                                                 | 255/812 [00:34<01:29,  6.25it/s]
Attaching quantization:  31%|######################6                                                 | 255/812 [00:34<01:29,  6.25it/s]
Attaching quantization:  32%|######################7                                                 | 257/812 [00:34<01:12,  7.63it/s]
Attaching quantization:  32%|######################7                                                 | 257/812 [00:34<01:12,  7.63it/s]
Attaching quantization:  32%|######################9                                                 | 259/812 [00:34<01:02,  8.88it/s]
Attaching quantization:  32%|######################9                                                 | 259/812 [00:34<01:02,  8.88it/s]
Attaching quantization:  32%|#######################1                                                | 261/812 [00:34<00:58,  9.41it/s]
Attaching quantization:  32%|#######################1                                                | 261/812 [00:34<00:58,  9.41it/s]
Attaching quantization:  32%|#######################3                                                | 263/812 [00:34<00:53, 10.26it/s]
Attaching quantization:  32%|#######################3                                                | 263/812 [00:34<00:53, 10.26it/s]
Attaching quantization:  33%|#######################4                                                | 265/812 [00:35<01:18,  6.98it/s]
Attaching quantization:  33%|#######################4                                                | 265/812 [00:35<01:18,  6.98it/s]
Attaching quantization:  33%|#######################6                                                | 267/812 [00:35<01:06,  8.23it/s]
Attaching quantization:  33%|#######################6                                                | 267/812 [00:35<01:06,  8.23it/s]
Attaching quantization:  33%|#######################8                                                | 269/812 [00:35<00:58,  9.33it/s]
Attaching quantization:  33%|#######################8                                                | 269/812 [00:35<00:58,  9.33it/s]
Attaching quantization:  33%|########################                                                | 271/812 [00:35<00:55,  9.79it/s]
Attaching quantization:  33%|########################                                                | 271/812 [00:35<00:55,  9.79it/s]
Attaching quantization:  34%|########################2                                               | 273/812 [00:35<00:51, 10.51it/s]
Attaching quantization:  34%|########################2                                               | 273/812 [00:35<00:51, 10.51it/s]
Attaching quantization:  34%|########################3                                               | 275/812 [00:36<01:13,  7.31it/s]
Attaching quantization:  34%|########################3                                               | 275/812 [00:36<01:13,  7.31it/s]
Attaching quantization:  34%|########################5                                               | 277/812 [00:36<01:02,  8.49it/s]
Attaching quantization:  34%|########################5                                               | 277/812 [00:36<01:02,  8.49it/s]
Attaching quantization:  34%|########################7                                               | 279/812 [00:36<00:55,  9.55it/s]
Attaching quantization:  34%|########################7                                               | 279/812 [00:36<00:55,  9.55it/s]
Attaching quantization:  35%|########################9                                               | 281/812 [00:36<00:52, 10.05it/s]
Attaching quantization:  35%|########################9                                               | 281/812 [00:36<00:52, 10.05it/s]
Attaching quantization:  35%|#########################                                               | 283/812 [00:37<00:50, 10.47it/s]
Attaching quantization:  35%|#########################                                               | 283/812 [00:37<00:50, 10.47it/s]
Attaching quantization:  35%|#########################2                                              | 285/812 [00:37<01:12,  7.31it/s]
Attaching quantization:  35%|#########################2                                              | 285/812 [00:37<01:12,  7.31it/s]
Attaching quantization:  35%|#########################4                                              | 287/812 [00:37<01:01,  8.48it/s]
Attaching quantization:  35%|#########################4                                              | 287/812 [00:37<01:01,  8.48it/s]
Attaching quantization:  36%|#########################6                                              | 289/812 [00:37<00:54,  9.65it/s]
Attaching quantization:  36%|#########################6                                              | 289/812 [00:37<00:54,  9.65it/s]
Attaching quantization:  36%|#########################8                                              | 291/812 [00:38<00:52,  9.96it/s]
Attaching quantization:  36%|#########################8                                              | 291/812 [00:38<00:52,  9.96it/s]
Attaching quantization:  36%|#########################9                                              | 293/812 [00:38<00:47, 10.97it/s]
Attaching quantization:  36%|#########################9                                              | 293/812 [00:38<00:47, 10.97it/s]
Attaching quantization:  36%|##########################1                                             | 295/812 [00:38<01:07,  7.63it/s]
Attaching quantization:  36%|##########################1                                             | 295/812 [00:38<01:07,  7.63it/s]
Attaching quantization:  37%|##########################3                                             | 297/812 [00:38<00:58,  8.77it/s]
Attaching quantization:  37%|##########################3                                             | 297/812 [00:38<00:58,  8.77it/s]
Attaching quantization:  37%|##########################5                                             | 299/812 [00:38<00:51,  9.88it/s]
Attaching quantization:  37%|##########################5                                             | 299/812 [00:38<00:51,  9.88it/s]
Attaching quantization:  37%|##########################6                                             | 301/812 [00:39<00:47, 10.74it/s]
Attaching quantization:  37%|##########################6                                             | 301/812 [00:39<00:47, 10.74it/s]
Attaching quantization:  37%|##########################8                                             | 303/812 [00:39<00:47, 10.67it/s]
Attaching quantization:  37%|##########################8                                             | 303/812 [00:39<00:47, 10.67it/s]
Attaching quantization:  38%|###########################                                             | 305/812 [00:39<01:08,  7.40it/s]
Attaching quantization:  38%|###########################                                             | 305/812 [00:39<01:08,  7.40it/s]
Attaching quantization:  38%|###########################2                                            | 307/812 [00:39<00:59,  8.43it/s]
Attaching quantization:  38%|###########################2                                            | 307/812 [00:39<00:59,  8.43it/s]
Attaching quantization:  38%|###########################3                                            | 309/812 [00:40<00:52,  9.59it/s]
Attaching quantization:  38%|###########################3                                            | 309/812 [00:40<00:52,  9.59it/s]
Attaching quantization:  38%|###########################5                                            | 311/812 [00:40<00:47, 10.45it/s]
Attaching quantization:  38%|###########################5                                            | 311/812 [00:40<00:47, 10.45it/s]
Attaching quantization:  39%|###########################7                                            | 313/812 [00:40<00:46, 10.69it/s]
Attaching quantization:  39%|###########################7                                            | 313/812 [00:40<00:46, 10.69it/s]
Attaching quantization:  39%|###########################9                                            | 315/812 [00:40<01:07,  7.41it/s]
Attaching quantization:  39%|###########################9                                            | 315/812 [00:40<01:07,  7.41it/s]
Attaching quantization:  39%|############################1                                           | 317/812 [00:40<00:57,  8.54it/s]
Attaching quantization:  39%|############################1                                           | 317/812 [00:40<00:57,  8.54it/s]
Attaching quantization:  39%|############################2                                           | 319/812 [00:41<01:04,  7.65it/s]
Attaching quantization:  39%|############################2                                           | 319/812 [00:41<01:04,  7.65it/s]
Attaching quantization:  39%|############################3                                           | 320/812 [00:41<01:12,  6.74it/s]
Attaching quantization:  39%|############################3                                           | 320/812 [00:41<01:12,  6.74it/s]
Attaching quantization:  40%|############################5                                           | 322/812 [00:41<01:06,  7.38it/s]
Attaching quantization:  40%|############################5                                           | 322/812 [00:41<01:06,  7.38it/s]
Attaching quantization:  40%|############################7                                           | 324/812 [00:42<01:17,  6.27it/s]
Attaching quantization:  40%|############################7                                           | 324/812 [00:42<01:17,  6.27it/s]
Attaching quantization:  40%|############################8                                           | 325/812 [00:42<01:38,  4.95it/s]
Attaching quantization:  40%|############################8                                           | 325/812 [00:42<01:38,  4.95it/s]
Attaching quantization:  40%|############################9                                           | 327/812 [00:42<01:17,  6.27it/s]
Attaching quantization:  40%|############################9                                           | 327/812 [00:42<01:17,  6.27it/s]
Attaching quantization:  41%|#############################1                                          | 329/812 [00:42<01:14,  6.48it/s]
Attaching quantization:  41%|#############################1                                          | 329/812 [00:42<01:14,  6.48it/s]
Attaching quantization:  41%|#############################3                                          | 331/812 [00:43<01:01,  7.81it/s]
Attaching quantization:  41%|#############################3                                          | 331/812 [00:43<01:01,  7.81it/s]
Attaching quantization:  41%|#############################5                                          | 333/812 [00:43<00:52,  9.17it/s]
Attaching quantization:  41%|#############################5                                          | 333/812 [00:43<00:52,  9.17it/s]
Attaching quantization:  41%|#############################7                                          | 335/812 [00:44<01:56,  4.11it/s]
Attaching quantization:  41%|#############################7                                          | 335/812 [00:44<01:56,  4.11it/s]
Attaching quantization:  42%|#############################8                                          | 337/812 [00:44<01:30,  5.24it/s]
Attaching quantization:  42%|#############################8                                          | 337/812 [00:44<01:30,  5.24it/s]
Attaching quantization:  42%|##############################                                          | 339/812 [00:44<01:12,  6.49it/s]
Attaching quantization:  42%|##############################                                          | 339/812 [00:44<01:12,  6.49it/s]
Attaching quantization:  42%|##############################2                                         | 341/812 [00:44<01:02,  7.50it/s]
Attaching quantization:  42%|##############################2                                         | 341/812 [00:44<01:02,  7.50it/s]
Attaching quantization:  42%|##############################4                                         | 343/812 [00:44<00:54,  8.65it/s]
Attaching quantization:  42%|##############################4                                         | 343/812 [00:44<00:54,  8.65it/s]
Attaching quantization:  42%|##############################5                                         | 345/812 [00:45<01:40,  4.65it/s]
Attaching quantization:  42%|##############################5                                         | 345/812 [00:45<01:40,  4.65it/s]
Attaching quantization:  43%|##############################7                                         | 347/812 [00:46<01:20,  5.79it/s]
Attaching quantization:  43%|##############################7                                         | 347/812 [00:46<01:20,  5.79it/s]
Attaching quantization:  43%|##############################9                                         | 349/812 [00:46<01:12,  6.42it/s]
Attaching quantization:  43%|##############################9                                         | 349/812 [00:46<01:12,  6.42it/s]
Attaching quantization:  43%|###############################                                         | 350/812 [00:46<01:11,  6.43it/s]
Attaching quantization:  43%|###############################                                         | 350/812 [00:46<01:11,  6.43it/s]
Attaching quantization:  43%|###############################1                                        | 351/812 [00:46<01:08,  6.71it/s]
Attaching quantization:  43%|###############################1                                        | 351/812 [00:46<01:08,  6.71it/s]
Attaching quantization:  43%|###############################2                                        | 352/812 [00:48<03:30,  2.18it/s]
Attaching quantization:  43%|###############################2                                        | 352/812 [00:48<03:30,  2.18it/s]
Attaching quantization:  43%|###############################3                                        | 353/812 [00:48<04:20,  1.76it/s]
Attaching quantization:  43%|###############################3                                        | 353/812 [00:48<04:20,  1.76it/s]
Attaching quantization:  44%|###############################3                                        | 354/812 [00:49<05:11,  1.47it/s]
Attaching quantization:  44%|###############################3                                        | 354/812 [00:49<05:11,  1.47it/s]
Attaching quantization:  44%|###############################4                                        | 355/812 [00:50<05:43,  1.33it/s]
Attaching quantization:  44%|###############################4                                        | 355/812 [00:50<05:43,  1.33it/s]
Attaching quantization:  44%|###############################5                                        | 356/812 [00:51<05:52,  1.29it/s]
Attaching quantization:  44%|###############################5                                        | 356/812 [00:51<05:52,  1.29it/s]
Attaching quantization:  44%|###############################6                                        | 357/812 [00:52<06:10,  1.23it/s]
Attaching quantization:  44%|###############################6                                        | 357/812 [00:52<06:10,  1.23it/s]
Attaching quantization:  44%|###############################7                                        | 358/812 [00:53<06:20,  1.19it/s]
Attaching quantization:  44%|###############################7                                        | 358/812 [00:53<06:20,  1.19it/s]
Attaching quantization:  44%|###############################8                                        | 359/812 [00:54<06:24,  1.18it/s]
Attaching quantization:  44%|###############################8                                        | 359/812 [00:54<06:24,  1.18it/s]
Attaching quantization:  44%|###############################9                                        | 360/812 [00:55<06:29,  1.16it/s]
Attaching quantization:  44%|###############################9                                        | 360/812 [00:55<06:29,  1.16it/s]
Attaching quantization:  44%|################################                                        | 361/812 [00:56<06:27,  1.16it/s]
Attaching quantization:  44%|################################                                        | 361/812 [00:56<06:27,  1.16it/s]
Attaching quantization:  45%|################################                                        | 362/812 [00:57<06:23,  1.17it/s]
Attaching quantization:  45%|################################                                        | 362/812 [00:57<06:23,  1.17it/s]
Attaching quantization:  45%|################################1                                       | 363/812 [00:57<06:05,  1.23it/s]
Attaching quantization:  45%|################################1                                       | 363/812 [00:57<06:05,  1.23it/s]
Attaching quantization:  45%|################################2                                       | 364/812 [00:58<06:38,  1.12it/s]
Attaching quantization:  45%|################################2                                       | 364/812 [00:58<06:38,  1.12it/s]
Attaching quantization:  45%|################################3                                       | 365/812 [00:59<05:52,  1.27it/s]
Attaching quantization:  45%|################################3                                       | 365/812 [00:59<05:52,  1.27it/s]
Attaching quantization:  45%|################################5                                       | 367/812 [00:59<03:24,  2.18it/s]
Attaching quantization:  45%|################################5                                       | 367/812 [00:59<03:24,  2.18it/s]
Attaching quantization:  45%|################################7                                       | 369/812 [00:59<02:18,  3.20it/s]
Attaching quantization:  45%|################################7                                       | 369/812 [00:59<02:18,  3.20it/s]
Attaching quantization:  46%|################################8                                       | 370/812 [00:59<01:57,  3.76it/s]
Attaching quantization:  46%|################################8                                       | 370/812 [00:59<01:57,  3.76it/s]
Attaching quantization:  46%|################################9                                       | 372/812 [00:59<01:24,  5.22it/s]
Attaching quantization:  46%|################################9                                       | 372/812 [00:59<01:24,  5.22it/s]
Attaching quantization:  46%|#################################1                                      | 374/812 [01:00<01:16,  5.72it/s]
Attaching quantization:  46%|#################################1                                      | 374/812 [01:00<01:16,  5.72it/s]
Attaching quantization:  46%|#################################2                                      | 375/812 [01:00<01:22,  5.32it/s]
Attaching quantization:  46%|#################################2                                      | 375/812 [01:00<01:22,  5.32it/s]
Attaching quantization:  46%|#################################4                                      | 377/812 [01:00<01:03,  6.83it/s]
Attaching quantization:  46%|#################################4                                      | 377/812 [01:00<01:03,  6.83it/s]
Attaching quantization:  47%|#################################6                                      | 379/812 [01:00<00:55,  7.87it/s]
Attaching quantization:  47%|#################################6                                      | 379/812 [01:00<00:55,  7.87it/s]
Attaching quantization:  47%|#################################7                                      | 381/812 [01:00<00:47,  9.08it/s]
Attaching quantization:  47%|#################################7                                      | 381/812 [01:00<00:47,  9.08it/s]
Attaching quantization:  47%|#################################9                                      | 383/812 [01:01<00:41, 10.22it/s]
Attaching quantization:  47%|#################################9                                      | 383/812 [01:01<00:41, 10.22it/s]
Attaching quantization:  47%|##################################1                                     | 385/812 [01:01<00:58,  7.26it/s]
Attaching quantization:  47%|##################################1                                     | 385/812 [01:01<00:58,  7.26it/s]
Attaching quantization:  48%|##################################3                                     | 387/812 [01:01<00:49,  8.57it/s]
Attaching quantization:  48%|##################################3                                     | 387/812 [01:01<00:49,  8.57it/s]
Attaching quantization:  48%|##################################4                                     | 389/812 [01:01<00:44,  9.61it/s]
Attaching quantization:  48%|##################################4                                     | 389/812 [01:01<00:44,  9.61it/s]
Attaching quantization:  48%|##################################6                                     | 391/812 [01:01<00:41, 10.10it/s]
Attaching quantization:  48%|##################################6                                     | 391/812 [01:01<00:41, 10.10it/s]
Attaching quantization:  48%|##################################8                                     | 393/812 [01:02<00:38, 11.02it/s]
Attaching quantization:  48%|##################################8                                     | 393/812 [01:02<00:38, 11.02it/s]
Attaching quantization:  49%|###################################                                     | 395/812 [01:02<00:55,  7.53it/s]
Attaching quantization:  49%|###################################                                     | 395/812 [01:02<00:55,  7.53it/s]
Attaching quantization:  49%|###################################2                                    | 397/812 [01:02<00:47,  8.71it/s]
Attaching quantization:  49%|###################################2                                    | 397/812 [01:02<00:47,  8.71it/s]
Attaching quantization:  49%|###################################3                                    | 399/812 [01:02<00:42,  9.78it/s]
Attaching quantization:  49%|###################################3                                    | 399/812 [01:02<00:42,  9.78it/s]
Attaching quantization:  49%|###################################5                                    | 401/812 [01:03<00:40, 10.21it/s]
Attaching quantization:  49%|###################################5                                    | 401/812 [01:03<00:40, 10.21it/s]
Attaching quantization:  50%|###################################7                                    | 403/812 [01:03<00:37, 10.98it/s]
Attaching quantization:  50%|###################################7                                    | 403/812 [01:03<00:37, 10.98it/s]
Attaching quantization:  50%|###################################9                                    | 405/812 [01:03<00:54,  7.41it/s]
Attaching quantization:  50%|###################################9                                    | 405/812 [01:03<00:54,  7.41it/s]
Attaching quantization:  50%|####################################2                                   | 409/812 [01:04<00:56,  7.17it/s]
Attaching quantization:  50%|####################################2                                   | 409/812 [01:04<00:56,  7.17it/s]
Attaching quantization:  51%|####################################4                                   | 411/812 [01:04<00:48,  8.22it/s]
Attaching quantization:  51%|####################################4                                   | 411/812 [01:04<00:48,  8.22it/s]
Attaching quantization:  51%|####################################6                                   | 413/812 [01:04<00:43,  9.20it/s]
Attaching quantization:  51%|####################################6                                   | 413/812 [01:04<00:43,  9.20it/s]
Attaching quantization:  51%|####################################7                                   | 415/812 [01:04<00:40,  9.72it/s]
Attaching quantization:  51%|####################################7                                   | 415/812 [01:04<00:40,  9.72it/s]
Attaching quantization:  51%|####################################9                                   | 417/812 [01:04<00:37, 10.41it/s]
Attaching quantization:  51%|####################################9                                   | 417/812 [01:04<00:37, 10.41it/s]
Attaching quantization:  52%|#####################################1                                  | 419/812 [01:05<00:35, 11.22it/s]
Attaching quantization:  52%|#####################################1                                  | 419/812 [01:05<00:35, 11.22it/s]
Attaching quantization:  52%|#####################################3                                  | 421/812 [01:05<00:51,  7.63it/s]
Attaching quantization:  52%|#####################################3                                  | 421/812 [01:05<00:51,  7.63it/s]
Attaching quantization:  52%|#####################################5                                  | 423/812 [01:05<00:44,  8.77it/s]
Attaching quantization:  52%|#####################################5                                  | 423/812 [01:05<00:44,  8.77it/s]
Attaching quantization:  52%|#####################################6                                  | 425/812 [01:05<00:39,  9.88it/s]
Attaching quantization:  52%|#####################################6                                  | 425/812 [01:05<00:39,  9.88it/s]
Attaching quantization:  53%|#####################################8                                  | 427/812 [01:05<00:35, 10.80it/s]
Attaching quantization:  53%|#####################################8                                  | 427/812 [01:05<00:35, 10.80it/s]
Attaching quantization:  53%|######################################                                  | 429/812 [01:06<00:32, 11.70it/s]
Attaching quantization:  53%|######################################                                  | 429/812 [01:06<00:32, 11.70it/s]
Attaching quantization:  53%|######################################2                                 | 431/812 [01:06<00:50,  7.52it/s]
Attaching quantization:  53%|######################################2                                 | 431/812 [01:06<00:50,  7.52it/s]
Attaching quantization:  53%|######################################3                                 | 433/812 [01:06<00:43,  8.66it/s]
Attaching quantization:  53%|######################################3                                 | 433/812 [01:06<00:43,  8.66it/s]
Attaching quantization:  54%|######################################5                                 | 435/812 [01:06<00:38,  9.70it/s]
Attaching quantization:  54%|######################################5                                 | 435/812 [01:06<00:38,  9.70it/s]
Attaching quantization:  54%|######################################7                                 | 437/812 [01:07<00:36, 10.35it/s]
Attaching quantization:  54%|######################################7                                 | 437/812 [01:07<00:36, 10.35it/s]
Attaching quantization:  54%|######################################9                                 | 439/812 [01:07<00:33, 11.29it/s]
Attaching quantization:  54%|######################################9                                 | 439/812 [01:07<00:33, 11.29it/s]
Attaching quantization:  54%|#######################################1                                | 441/812 [01:07<00:47,  7.86it/s]
Attaching quantization:  54%|#######################################1                                | 441/812 [01:07<00:47,  7.86it/s]
Attaching quantization:  55%|#######################################2                                | 443/812 [01:07<00:41,  8.97it/s]
Attaching quantization:  55%|#######################################2                                | 443/812 [01:07<00:41,  8.97it/s]
Attaching quantization:  55%|#######################################4                                | 445/812 [01:07<00:36, 10.04it/s]
Attaching quantization:  55%|#######################################4                                | 445/812 [01:07<00:36, 10.04it/s]
Attaching quantization:  55%|#######################################6                                | 447/812 [01:08<00:32, 11.15it/s]
Attaching quantization:  55%|#######################################6                                | 447/812 [01:08<00:32, 11.15it/s]
Attaching quantization:  55%|#######################################8                                | 449/812 [01:08<00:30, 12.02it/s]
Attaching quantization:  55%|#######################################8                                | 449/812 [01:08<00:30, 12.02it/s]
Attaching quantization:  56%|#######################################9                                | 451/812 [01:08<00:44,  8.11it/s]
Attaching quantization:  56%|#######################################9                                | 451/812 [01:08<00:44,  8.11it/s]
Attaching quantization:  56%|########################################1                               | 453/812 [01:08<00:40,  8.96it/s]
Attaching quantization:  56%|########################################1                               | 453/812 [01:08<00:40,  8.96it/s]
Attaching quantization:  56%|########################################3                               | 455/812 [01:08<00:35, 10.06it/s]
Attaching quantization:  56%|########################################3                               | 455/812 [01:08<00:35, 10.06it/s]
Attaching quantization:  56%|########################################5                               | 457/812 [01:09<00:32, 11.05it/s]
Attaching quantization:  56%|########################################5                               | 457/812 [01:09<00:32, 11.05it/s]
Attaching quantization:  57%|########################################6                               | 459/812 [01:09<00:29, 11.85it/s]
Attaching quantization:  57%|########################################6                               | 459/812 [01:09<00:29, 11.85it/s]
Attaching quantization:  57%|########################################8                               | 461/812 [01:09<00:43,  8.04it/s]
Attaching quantization:  57%|########################################8                               | 461/812 [01:09<00:43,  8.04it/s]
Attaching quantization:  57%|#########################################                               | 463/812 [01:09<00:37,  9.27it/s]
Attaching quantization:  57%|#########################################                               | 463/812 [01:09<00:37,  9.27it/s]
Attaching quantization:  57%|#########################################2                              | 465/812 [01:09<00:35,  9.81it/s]
Attaching quantization:  57%|#########################################2                              | 465/812 [01:09<00:35,  9.81it/s]
Attaching quantization:  58%|#########################################4                              | 467/812 [01:10<00:31, 10.92it/s]
Attaching quantization:  58%|#########################################4                              | 467/812 [01:10<00:31, 10.92it/s]
Attaching quantization:  58%|#########################################5                              | 469/812 [01:10<00:28, 11.83it/s]
Attaching quantization:  58%|#########################################5                              | 469/812 [01:10<00:28, 11.83it/s]
Attaching quantization:  58%|#########################################7                              | 471/812 [01:10<00:42,  8.04it/s]
Attaching quantization:  58%|#########################################7                              | 471/812 [01:10<00:42,  8.04it/s]
Attaching quantization:  58%|#########################################9                              | 473/812 [01:10<00:36,  9.33it/s]
Attaching quantization:  58%|#########################################9                              | 473/812 [01:10<00:36,  9.33it/s]
Attaching quantization:  58%|##########################################1                             | 475/812 [01:10<00:32, 10.26it/s]
Attaching quantization:  58%|##########################################1                             | 475/812 [01:10<00:32, 10.26it/s]
Attaching quantization:  59%|##########################################2                             | 477/812 [01:11<00:31, 10.79it/s]
Attaching quantization:  59%|##########################################2                             | 477/812 [01:11<00:31, 10.79it/s]
Attaching quantization:  59%|##########################################4                             | 479/812 [01:11<00:28, 11.70it/s]
Attaching quantization:  59%|##########################################4                             | 479/812 [01:11<00:28, 11.70it/s]
Attaching quantization:  59%|##########################################6                             | 481/812 [01:11<00:41,  8.06it/s]
Attaching quantization:  59%|##########################################6                             | 481/812 [01:11<00:41,  8.06it/s]
Attaching quantization:  59%|##########################################8                             | 483/812 [01:11<00:35,  9.25it/s]
Attaching quantization:  59%|##########################################8                             | 483/812 [01:11<00:35,  9.25it/s]
Attaching quantization:  60%|###########################################                             | 485/812 [01:11<00:31, 10.38it/s]
Attaching quantization:  60%|###########################################                             | 485/812 [01:11<00:31, 10.38it/s]
Attaching quantization:  60%|###########################################1                            | 487/812 [01:12<00:29, 11.03it/s]
Attaching quantization:  60%|###########################################1                            | 487/812 [01:12<00:29, 11.03it/s]
Attaching quantization:  60%|###########################################3                            | 489/812 [01:12<00:28, 11.40it/s]
Attaching quantization:  60%|###########################################3                            | 489/812 [01:12<00:28, 11.40it/s]
Attaching quantization:  60%|###########################################5                            | 491/812 [01:12<00:41,  7.75it/s]
Attaching quantization:  60%|###########################################5                            | 491/812 [01:12<00:41,  7.75it/s]
Attaching quantization:  61%|###########################################7                            | 493/812 [01:12<00:35,  8.93it/s]
Attaching quantization:  61%|###########################################7                            | 493/812 [01:12<00:35,  8.93it/s]
Attaching quantization:  61%|###########################################8                            | 495/812 [01:12<00:31, 10.00it/s]
Attaching quantization:  61%|###########################################8                            | 495/812 [01:12<00:31, 10.00it/s]
Attaching quantization:  61%|############################################                            | 497/812 [01:13<00:28, 10.90it/s]
Attaching quantization:  61%|############################################                            | 497/812 [01:13<00:28, 10.90it/s]
Attaching quantization:  61%|############################################2                           | 499/812 [01:13<00:28, 10.90it/s]
Attaching quantization:  61%|############################################2                           | 499/812 [01:13<00:28, 10.90it/s]
Attaching quantization:  62%|############################################4                           | 501/812 [01:13<00:40,  7.61it/s]
Attaching quantization:  62%|############################################4                           | 501/812 [01:13<00:40,  7.61it/s]
Attaching quantization:  62%|############################################6                           | 503/812 [01:13<00:34,  8.85it/s]
Attaching quantization:  62%|############################################6                           | 503/812 [01:13<00:34,  8.85it/s]
Attaching quantization:  62%|############################################7                           | 505/812 [01:14<00:31,  9.85it/s]
Attaching quantization:  62%|############################################7                           | 505/812 [01:14<00:31,  9.85it/s]
Attaching quantization:  62%|############################################9                           | 507/812 [01:14<00:28, 10.77it/s]
Attaching quantization:  62%|############################################9                           | 507/812 [01:14<00:28, 10.77it/s]
Attaching quantization:  63%|#############################################1                          | 509/812 [01:14<00:26, 11.44it/s]
Attaching quantization:  63%|#############################################1                          | 509/812 [01:14<00:26, 11.44it/s]
Attaching quantization:  63%|#############################################3                          | 511/812 [01:14<00:40,  7.45it/s]
Attaching quantization:  63%|#############################################3                          | 511/812 [01:14<00:40,  7.45it/s]
Attaching quantization:  63%|#############################################4                          | 513/812 [01:14<00:34,  8.67it/s]
Attaching quantization:  63%|#############################################4                          | 513/812 [01:14<00:34,  8.67it/s]
Attaching quantization:  63%|#############################################6                          | 515/812 [01:15<00:30,  9.67it/s]
Attaching quantization:  63%|#############################################6                          | 515/812 [01:15<00:30,  9.67it/s]
Attaching quantization:  64%|#############################################8                          | 517/812 [01:15<00:27, 10.55it/s]
Attaching quantization:  64%|#############################################8                          | 517/812 [01:15<00:27, 10.55it/s]
Attaching quantization:  64%|##############################################                          | 519/812 [01:15<00:26, 11.23it/s]
Attaching quantization:  64%|##############################################                          | 519/812 [01:15<00:26, 11.23it/s]
Attaching quantization:  64%|##############################################1                         | 521/812 [01:15<00:38,  7.48it/s]
Attaching quantization:  64%|##############################################1                         | 521/812 [01:15<00:38,  7.48it/s]
Attaching quantization:  64%|##############################################3                         | 523/812 [01:16<00:33,  8.52it/s]
Attaching quantization:  64%|##############################################3                         | 523/812 [01:16<00:33,  8.52it/s]
Attaching quantization:  65%|##############################################5                         | 525/812 [01:16<00:29,  9.62it/s]
Attaching quantization:  65%|##############################################5                         | 525/812 [01:16<00:29,  9.62it/s]
Attaching quantization:  65%|##############################################7                         | 527/812 [01:16<00:27, 10.50it/s]
Attaching quantization:  65%|##############################################7                         | 527/812 [01:16<00:27, 10.50it/s]
Attaching quantization:  65%|##############################################9                         | 529/812 [01:16<00:25, 11.24it/s]
Attaching quantization:  65%|##############################################9                         | 529/812 [01:16<00:25, 11.24it/s]
Attaching quantization:  65%|###############################################                         | 531/812 [01:16<00:38,  7.34it/s]
Attaching quantization:  65%|###############################################                         | 531/812 [01:16<00:38,  7.34it/s]
Attaching quantization:  66%|###############################################2                        | 533/812 [01:17<00:32,  8.47it/s]
Attaching quantization:  66%|###############################################2                        | 533/812 [01:17<00:32,  8.47it/s]
Attaching quantization:  66%|###############################################4                        | 535/812 [01:17<00:29,  9.51it/s]
Attaching quantization:  66%|###############################################4                        | 535/812 [01:17<00:29,  9.51it/s]
Attaching quantization:  66%|###############################################6                        | 537/812 [01:17<00:26, 10.52it/s]
Attaching quantization:  66%|###############################################6                        | 537/812 [01:17<00:26, 10.52it/s]
Attaching quantization:  66%|###############################################7                        | 539/812 [01:17<00:24, 11.21it/s]
Attaching quantization:  66%|###############################################7                        | 539/812 [01:17<00:24, 11.21it/s]
Attaching quantization:  67%|###############################################9                        | 541/812 [01:18<00:40,  6.75it/s]
Attaching quantization:  67%|###############################################9                        | 541/812 [01:18<00:40,  6.75it/s]
Attaching quantization:  67%|################################################1                       | 543/812 [01:18<00:33,  7.97it/s]
Attaching quantization:  67%|################################################1                       | 543/812 [01:18<00:33,  7.97it/s]
Attaching quantization:  67%|################################################3                       | 545/812 [01:18<00:30,  8.66it/s]
Attaching quantization:  67%|################################################3                       | 545/812 [01:18<00:30,  8.66it/s]
Attaching quantization:  67%|################################################5                       | 547/812 [01:18<00:27,  9.50it/s]
Attaching quantization:  67%|################################################5                       | 547/812 [01:18<00:27,  9.50it/s]
Attaching quantization:  68%|################################################6                       | 549/812 [01:18<00:26,  9.75it/s]
Attaching quantization:  68%|################################################6                       | 549/812 [01:18<00:26,  9.75it/s]
Attaching quantization:  68%|################################################8                       | 551/812 [01:19<00:44,  5.83it/s]
Attaching quantization:  68%|################################################8                       | 551/812 [01:19<00:44,  5.83it/s]
Attaching quantization:  68%|#################################################                       | 553/812 [01:19<00:36,  7.03it/s]
Attaching quantization:  68%|#################################################                       | 553/812 [01:19<00:36,  7.03it/s]
Attaching quantization:  68%|#################################################2                      | 555/812 [01:19<00:33,  7.58it/s]
Attaching quantization:  68%|#################################################2                      | 555/812 [01:19<00:33,  7.58it/s]
Attaching quantization:  69%|#################################################3                      | 557/812 [01:20<00:33,  7.71it/s]
Attaching quantization:  69%|#################################################3                      | 557/812 [01:20<00:33,  7.71it/s]
Attaching quantization:  69%|#################################################5                      | 559/812 [01:20<00:30,  8.41it/s]
Attaching quantization:  69%|#################################################5                      | 559/812 [01:20<00:30,  8.41it/s]
Attaching quantization:  69%|#################################################6                      | 560/812 [01:20<00:41,  6.05it/s]
Attaching quantization:  69%|#################################################6                      | 560/812 [01:20<00:41,  6.05it/s]
Attaching quantization:  69%|#################################################7                      | 561/812 [01:20<00:48,  5.16it/s]
Attaching quantization:  69%|#################################################7                      | 561/812 [01:20<00:48,  5.16it/s]
Attaching quantization:  69%|#################################################9                      | 563/812 [01:21<00:38,  6.52it/s]
Attaching quantization:  69%|#################################################9                      | 563/812 [01:21<00:38,  6.52it/s]
Attaching quantization:  70%|##################################################                      | 565/812 [01:21<00:31,  7.72it/s]
Attaching quantization:  70%|##################################################                      | 565/812 [01:21<00:31,  7.72it/s]
Attaching quantization:  70%|##################################################2                     | 567/812 [01:21<00:31,  7.71it/s]
Attaching quantization:  70%|##################################################2                     | 567/812 [01:21<00:31,  7.71it/s]
Attaching quantization:  70%|##################################################4                     | 569/812 [01:21<00:27,  8.96it/s]
Attaching quantization:  70%|##################################################4                     | 569/812 [01:21<00:27,  8.96it/s]
Attaching quantization:  70%|##################################################6                     | 571/812 [01:22<00:44,  5.40it/s]
Attaching quantization:  70%|##################################################6                     | 571/812 [01:22<00:44,  5.40it/s]
Attaching quantization:  71%|##################################################8                     | 573/812 [01:22<00:36,  6.61it/s]
Attaching quantization:  71%|##################################################8                     | 573/812 [01:22<00:36,  6.61it/s]
Attaching quantization:  71%|##################################################9                     | 575/812 [01:22<00:31,  7.63it/s]
Attaching quantization:  71%|##################################################9                     | 575/812 [01:22<00:31,  7.63it/s]
Attaching quantization:  71%|###################################################1                    | 577/812 [01:22<00:27,  8.40it/s]
Attaching quantization:  71%|###################################################1                    | 577/812 [01:22<00:27,  8.40it/s]
Attaching quantization:  71%|###################################################3                    | 579/812 [01:23<00:32,  7.08it/s]
Attaching quantization:  71%|###################################################3                    | 579/812 [01:23<00:32,  7.08it/s]
Attaching quantization:  71%|###################################################4                    | 580/812 [01:23<00:42,  5.46it/s]
Attaching quantization:  71%|###################################################4                    | 580/812 [01:23<00:42,  5.46it/s]
Attaching quantization:  72%|###################################################5                    | 581/812 [01:24<00:47,  4.82it/s]
Attaching quantization:  72%|###################################################5                    | 581/812 [01:24<00:47,  4.82it/s]
Attaching quantization:  72%|###################################################6                    | 583/812 [01:24<00:36,  6.31it/s]
Attaching quantization:  72%|###################################################6                    | 583/812 [01:24<00:36,  6.31it/s]
Attaching quantization:  72%|###################################################8                    | 585/812 [01:24<00:30,  7.38it/s]
Attaching quantization:  72%|###################################################8                    | 585/812 [01:24<00:30,  7.38it/s]
Attaching quantization:  72%|####################################################                    | 587/812 [01:24<00:27,  8.09it/s]
Attaching quantization:  72%|####################################################                    | 587/812 [01:24<00:27,  8.09it/s]
Attaching quantization:  73%|####################################################2                   | 589/812 [01:24<00:28,  7.71it/s]
Attaching quantization:  73%|####################################################2                   | 589/812 [01:24<00:28,  7.71it/s]
Attaching quantization:  73%|####################################################3                   | 590/812 [01:25<00:37,  5.97it/s]
Attaching quantization:  73%|####################################################3                   | 590/812 [01:25<00:37,  5.97it/s]
Attaching quantization:  73%|####################################################4                   | 591/812 [01:25<00:44,  4.99it/s]
Attaching quantization:  73%|####################################################4                   | 591/812 [01:25<00:44,  4.99it/s]
Attaching quantization:  73%|####################################################5                   | 593/812 [01:25<00:33,  6.54it/s]
Attaching quantization:  73%|####################################################5                   | 593/812 [01:25<00:33,  6.54it/s]
Attaching quantization:  73%|####################################################7                   | 595/812 [01:25<00:29,  7.26it/s]
Attaching quantization:  73%|####################################################7                   | 595/812 [01:25<00:29,  7.26it/s]
Attaching quantization:  74%|####################################################9                   | 597/812 [01:26<00:26,  8.18it/s]
Attaching quantization:  74%|####################################################9                   | 597/812 [01:26<00:26,  8.18it/s]
Attaching quantization:  74%|#####################################################                   | 598/812 [01:26<00:25,  8.47it/s]
Attaching quantization:  74%|#####################################################                   | 598/812 [01:26<00:25,  8.47it/s]
Attaching quantization:  74%|#####################################################1                  | 599/812 [01:26<00:24,  8.71it/s]
Attaching quantization:  74%|#####################################################1                  | 599/812 [01:26<00:24,  8.71it/s]
Attaching quantization:  74%|#####################################################2                  | 600/812 [01:26<00:36,  5.83it/s]
Attaching quantization:  74%|#####################################################2                  | 600/812 [01:26<00:36,  5.83it/s]
Attaching quantization:  74%|#####################################################2                  | 601/812 [01:26<00:47,  4.48it/s]
Attaching quantization:  74%|#####################################################2                  | 601/812 [01:26<00:47,  4.48it/s]
Attaching quantization:  74%|#####################################################4                  | 603/812 [01:27<00:34,  6.09it/s]
Attaching quantization:  74%|#####################################################4                  | 603/812 [01:27<00:34,  6.09it/s]
Attaching quantization:  75%|#####################################################6                  | 605/812 [01:27<00:28,  7.28it/s]
Attaching quantization:  75%|#####################################################6                  | 605/812 [01:27<00:28,  7.28it/s]
Attaching quantization:  75%|#####################################################8                  | 607/812 [01:27<00:25,  7.98it/s]
Attaching quantization:  75%|#####################################################8                  | 607/812 [01:27<00:25,  7.98it/s]
Attaching quantization:  75%|######################################################                  | 609/812 [01:27<00:23,  8.75it/s]
Attaching quantization:  75%|######################################################                  | 609/812 [01:27<00:23,  8.75it/s]
Attaching quantization:  75%|######################################################                  | 610/812 [01:28<00:35,  5.68it/s]
Attaching quantization:  75%|######################################################                  | 610/812 [01:28<00:35,  5.68it/s]
Attaching quantization:  75%|######################################################1                 | 611/812 [01:28<00:41,  4.85it/s]
Attaching quantization:  75%|######################################################1                 | 611/812 [01:28<00:41,  4.85it/s]
Attaching quantization:  75%|######################################################3                 | 613/812 [01:28<00:31,  6.38it/s]
Attaching quantization:  75%|######################################################3                 | 613/812 [01:28<00:31,  6.38it/s]
Attaching quantization:  76%|######################################################5                 | 615/812 [01:28<00:26,  7.45it/s]
Attaching quantization:  76%|######################################################5                 | 615/812 [01:28<00:26,  7.45it/s]
Attaching quantization:  76%|######################################################7                 | 617/812 [01:28<00:23,  8.34it/s]
Attaching quantization:  76%|######################################################7                 | 617/812 [01:28<00:23,  8.34it/s]
Attaching quantization:  76%|######################################################8                 | 619/812 [01:29<00:25,  7.65it/s]
Attaching quantization:  76%|######################################################8                 | 619/812 [01:29<00:25,  7.65it/s]
Attaching quantization:  76%|######################################################9                 | 620/812 [01:29<00:30,  6.33it/s]
Attaching quantization:  76%|######################################################9                 | 620/812 [01:29<00:30,  6.33it/s]
Attaching quantization:  76%|#######################################################                 | 621/812 [01:29<00:36,  5.22it/s]
Attaching quantization:  76%|#######################################################                 | 621/812 [01:29<00:36,  5.22it/s]
Attaching quantization:  77%|#######################################################2                | 623/812 [01:30<00:27,  6.77it/s]
Attaching quantization:  77%|#######################################################2                | 623/812 [01:30<00:27,  6.77it/s]
Attaching quantization:  77%|#######################################################4                | 625/812 [01:30<00:23,  7.97it/s]
Attaching quantization:  77%|#######################################################4                | 625/812 [01:30<00:23,  7.97it/s]
Attaching quantization:  77%|#######################################################5                | 627/812 [01:30<00:23,  7.73it/s]
Attaching quantization:  77%|#######################################################5                | 627/812 [01:30<00:23,  7.73it/s]
Attaching quantization:  77%|#######################################################7                | 629/812 [01:30<00:20,  8.99it/s]
Attaching quantization:  77%|#######################################################7                | 629/812 [01:30<00:20,  8.99it/s]
Attaching quantization:  78%|#######################################################9                | 631/812 [01:31<00:32,  5.58it/s]
Attaching quantization:  78%|#######################################################9                | 631/812 [01:31<00:32,  5.58it/s]
Attaching quantization:  78%|########################################################1               | 633/812 [01:31<00:26,  6.85it/s]
Attaching quantization:  78%|########################################################1               | 633/812 [01:31<00:26,  6.85it/s]
Attaching quantization:  78%|########################################################3               | 635/812 [01:31<00:22,  7.75it/s]
Attaching quantization:  78%|########################################################3               | 635/812 [01:31<00:22,  7.75it/s]
Attaching quantization:  78%|########################################################4               | 637/812 [01:31<00:20,  8.49it/s]
Attaching quantization:  78%|########################################################4               | 637/812 [01:31<00:20,  8.49it/s]
Attaching quantization:  79%|########################################################6               | 639/812 [01:31<00:19,  8.81it/s]
Attaching quantization:  79%|########################################################6               | 639/812 [01:31<00:19,  8.81it/s]
Attaching quantization:  79%|########################################################8               | 641/812 [01:32<00:31,  5.39it/s]
Attaching quantization:  79%|########################################################8               | 641/812 [01:32<00:31,  5.39it/s]
Attaching quantization:  79%|#########################################################               | 643/812 [01:32<00:25,  6.59it/s]
Attaching quantization:  79%|#########################################################               | 643/812 [01:32<00:25,  6.59it/s]
Attaching quantization:  79%|#########################################################1              | 645/812 [01:33<00:21,  7.60it/s]
Attaching quantization:  79%|#########################################################1              | 645/812 [01:33<00:21,  7.60it/s]
Attaching quantization:  80%|#########################################################3              | 647/812 [01:33<00:19,  8.35it/s]
Attaching quantization:  80%|#########################################################3              | 647/812 [01:33<00:19,  8.35it/s]
Attaching quantization:  80%|#########################################################5              | 649/812 [01:33<00:18,  8.65it/s]
Attaching quantization:  80%|#########################################################5              | 649/812 [01:33<00:18,  8.65it/s]
Attaching quantization:  80%|#########################################################7              | 651/812 [01:34<00:30,  5.28it/s]
Attaching quantization:  80%|#########################################################7              | 651/812 [01:34<00:30,  5.28it/s]
Attaching quantization:  80%|#########################################################9              | 653/812 [01:34<00:24,  6.46it/s]
Attaching quantization:  80%|#########################################################9              | 653/812 [01:34<00:24,  6.46it/s]
Attaching quantization:  81%|##########################################################              | 655/812 [01:34<00:22,  7.10it/s]
Attaching quantization:  81%|##########################################################              | 655/812 [01:34<00:22,  7.10it/s]
Attaching quantization:  81%|##########################################################2             | 657/812 [01:34<00:19,  8.11it/s]
Attaching quantization:  81%|##########################################################2             | 657/812 [01:34<00:19,  8.11it/s]
Attaching quantization:  81%|##########################################################4             | 659/812 [01:34<00:19,  8.04it/s]
Attaching quantization:  81%|##########################################################4             | 659/812 [01:34<00:19,  8.04it/s]
Attaching quantization:  81%|##########################################################5             | 660/812 [01:35<00:24,  6.10it/s]
Attaching quantization:  81%|##########################################################5             | 660/812 [01:35<00:24,  6.10it/s]
Attaching quantization:  81%|##########################################################6             | 661/812 [01:35<00:28,  5.31it/s]
Attaching quantization:  81%|##########################################################6             | 661/812 [01:35<00:28,  5.31it/s]
Attaching quantization:  82%|##########################################################7             | 663/812 [01:35<00:21,  6.81it/s]
Attaching quantization:  82%|##########################################################7             | 663/812 [01:35<00:21,  6.81it/s]
Attaching quantization:  82%|##########################################################8             | 664/812 [01:35<00:20,  7.12it/s]
Attaching quantization:  82%|##########################################################8             | 664/812 [01:35<00:20,  7.12it/s]
Attaching quantization:  82%|###########################################################             | 666/812 [01:36<00:19,  7.66it/s]
Attaching quantization:  82%|###########################################################             | 666/812 [01:36<00:19,  7.66it/s]
Attaching quantization:  82%|###########################################################2            | 668/812 [01:36<00:15,  9.02it/s]
Attaching quantization:  82%|###########################################################2            | 668/812 [01:36<00:15,  9.02it/s]
Attaching quantization:  83%|###########################################################4            | 670/812 [01:36<00:20,  6.81it/s]
Attaching quantization:  83%|###########################################################4            | 670/812 [01:36<00:20,  6.81it/s]
Attaching quantization:  83%|###########################################################4            | 671/812 [01:36<00:23,  5.92it/s]
Attaching quantization:  83%|###########################################################4            | 671/812 [01:36<00:23,  5.92it/s]
Attaching quantization:  83%|###########################################################6            | 673/812 [01:37<00:18,  7.42it/s]
Attaching quantization:  83%|###########################################################6            | 673/812 [01:37<00:18,  7.42it/s]
Attaching quantization:  83%|###########################################################7            | 674/812 [01:37<00:19,  7.25it/s]
Attaching quantization:  83%|###########################################################7            | 674/812 [01:37<00:19,  7.25it/s]
Attaching quantization:  83%|###########################################################9            | 676/812 [01:37<00:15,  8.83it/s]
Attaching quantization:  83%|###########################################################9            | 676/812 [01:37<00:15,  8.83it/s]
Attaching quantization:  83%|############################################################1           | 678/812 [01:37<00:13,  9.94it/s]
Attaching quantization:  83%|############################################################1           | 678/812 [01:37<00:13,  9.94it/s]
Attaching quantization:  84%|############################################################2           | 680/812 [01:37<00:18,  7.17it/s]
Attaching quantization:  84%|############################################################2           | 680/812 [01:37<00:18,  7.17it/s]
Attaching quantization:  84%|############################################################3           | 681/812 [01:38<00:23,  5.48it/s]
Attaching quantization:  84%|############################################################3           | 681/812 [01:38<00:23,  5.48it/s]
Attaching quantization:  84%|############################################################5           | 683/812 [01:38<00:18,  6.95it/s]
Attaching quantization:  84%|############################################################5           | 683/812 [01:38<00:18,  6.95it/s]
Attaching quantization:  84%|############################################################7           | 685/812 [01:38<00:16,  7.92it/s]
Attaching quantization:  84%|############################################################7           | 685/812 [01:38<00:16,  7.92it/s]
Attaching quantization:  85%|############################################################9           | 687/812 [01:38<00:14,  8.34it/s]
Attaching quantization:  85%|############################################################9           | 687/812 [01:38<00:14,  8.34it/s]
Attaching quantization:  85%|#############################################################           | 689/812 [01:38<00:13,  9.03it/s]
Attaching quantization:  85%|#############################################################           | 689/812 [01:38<00:13,  9.03it/s]
Attaching quantization:  85%|#############################################################2          | 691/812 [01:39<00:23,  5.19it/s]
Attaching quantization:  85%|#############################################################2          | 691/812 [01:39<00:23,  5.19it/s]
Attaching quantization:  85%|#############################################################4          | 693/812 [01:39<00:18,  6.33it/s]
Attaching quantization:  85%|#############################################################4          | 693/812 [01:39<00:18,  6.33it/s]
Attaching quantization:  86%|#############################################################6          | 695/812 [01:40<00:16,  7.24it/s]
Attaching quantization:  86%|#############################################################6          | 695/812 [01:40<00:16,  7.24it/s]
Attaching quantization:  86%|#############################################################7          | 696/812 [01:40<00:15,  7.32it/s]
Attaching quantization:  86%|#############################################################7          | 696/812 [01:40<00:15,  7.32it/s]
Attaching quantization:  86%|#############################################################8          | 698/812 [01:40<00:13,  8.56it/s]
Attaching quantization:  86%|#############################################################8          | 698/812 [01:40<00:13,  8.56it/s]
Attaching quantization:  86%|##############################################################          | 700/812 [01:40<00:17,  6.51it/s]
Attaching quantization:  86%|##############################################################          | 700/812 [01:40<00:17,  6.51it/s]
Attaching quantization:  86%|##############################################################1         | 701/812 [01:41<00:19,  5.62it/s]
Attaching quantization:  86%|##############################################################1         | 701/812 [01:41<00:19,  5.62it/s]
Attaching quantization:  87%|##############################################################3         | 703/812 [01:41<00:15,  7.04it/s]
Attaching quantization:  87%|##############################################################3         | 703/812 [01:41<00:15,  7.04it/s]
Attaching quantization:  87%|##############################################################5         | 705/812 [01:41<00:13,  8.04it/s]
Attaching quantization:  87%|##############################################################5         | 705/812 [01:41<00:13,  8.04it/s]
Attaching quantization:  87%|##############################################################6         | 707/812 [01:41<00:12,  8.32it/s]
Attaching quantization:  87%|##############################################################6         | 707/812 [01:41<00:12,  8.32it/s]
Attaching quantization:  87%|##############################################################8         | 709/812 [01:41<00:11,  9.04it/s]
Attaching quantization:  87%|##############################################################8         | 709/812 [01:41<00:11,  9.04it/s]
Attaching quantization:  87%|##############################################################9         | 710/812 [01:42<00:16,  6.28it/s]
Attaching quantization:  87%|##############################################################9         | 710/812 [01:42<00:16,  6.28it/s]
Attaching quantization:  88%|###############################################################         | 711/812 [01:42<00:18,  5.42it/s]
Attaching quantization:  88%|###############################################################         | 711/812 [01:42<00:18,  5.42it/s]
Attaching quantization:  88%|###############################################################2        | 713/812 [01:42<00:14,  6.99it/s]
Attaching quantization:  88%|###############################################################2        | 713/812 [01:42<00:14,  6.99it/s]
Attaching quantization:  88%|###############################################################3        | 714/812 [01:42<00:13,  7.32it/s]
Attaching quantization:  88%|###############################################################3        | 714/812 [01:42<00:13,  7.32it/s]
Attaching quantization:  88%|###############################################################4        | 716/812 [01:42<00:11,  8.48it/s]
Attaching quantization:  88%|###############################################################4        | 716/812 [01:42<00:11,  8.48it/s]
Attaching quantization:  88%|###############################################################6        | 718/812 [01:43<00:10,  8.97it/s]
Attaching quantization:  88%|###############################################################6        | 718/812 [01:43<00:10,  8.97it/s]
Attaching quantization:  89%|###############################################################8        | 720/812 [01:43<00:13,  6.94it/s]
Attaching quantization:  89%|###############################################################8        | 720/812 [01:43<00:13,  6.94it/s]
Attaching quantization:  89%|###############################################################9        | 721/812 [01:43<00:16,  5.37it/s]
Attaching quantization:  89%|###############################################################9        | 721/812 [01:43<00:16,  5.37it/s]
Attaching quantization:  89%|################################################################1       | 723/812 [01:44<00:13,  6.84it/s]
Attaching quantization:  89%|################################################################1       | 723/812 [01:44<00:13,  6.84it/s]
Attaching quantization:  89%|################################################################2       | 725/812 [01:44<00:11,  7.88it/s]
Attaching quantization:  89%|################################################################2       | 725/812 [01:44<00:11,  7.88it/s]
Attaching quantization:  90%|################################################################4       | 727/812 [01:44<00:09,  8.73it/s]
Attaching quantization:  90%|################################################################4       | 727/812 [01:44<00:09,  8.73it/s]
Attaching quantization:  90%|################################################################6       | 729/812 [01:44<00:09,  9.19it/s]
Attaching quantization:  90%|################################################################6       | 729/812 [01:44<00:09,  9.19it/s]
Attaching quantization:  90%|################################################################8       | 731/812 [01:45<00:15,  5.33it/s]
Attaching quantization:  90%|################################################################8       | 731/812 [01:45<00:15,  5.33it/s]
Attaching quantization:  90%|################################################################9       | 733/812 [01:45<00:11,  6.59it/s]
Attaching quantization:  90%|################################################################9       | 733/812 [01:45<00:11,  6.59it/s]
Attaching quantization:  90%|#################################################################       | 734/812 [01:45<00:11,  7.00it/s]
Attaching quantization:  90%|#################################################################       | 734/812 [01:45<00:11,  7.00it/s]
Attaching quantization:  91%|#################################################################2      | 736/812 [01:45<00:09,  8.01it/s]
Attaching quantization:  91%|#################################################################2      | 736/812 [01:45<00:09,  8.01it/s]
Attaching quantization:  91%|#################################################################4      | 738/812 [01:45<00:08,  8.61it/s]
Attaching quantization:  91%|#################################################################4      | 738/812 [01:45<00:08,  8.61it/s]
Attaching quantization:  91%|#################################################################6      | 740/812 [01:46<00:11,  6.16it/s]
Attaching quantization:  91%|#################################################################6      | 740/812 [01:46<00:11,  6.16it/s]
Attaching quantization:  91%|#################################################################7      | 741/812 [01:46<00:12,  5.50it/s]
Attaching quantization:  91%|#################################################################7      | 741/812 [01:46<00:12,  5.50it/s]
Attaching quantization:  92%|#################################################################8      | 743/812 [01:46<00:09,  6.90it/s]
Attaching quantization:  92%|#################################################################8      | 743/812 [01:46<00:09,  6.90it/s]
Attaching quantization:  92%|##################################################################      | 745/812 [01:47<00:08,  7.80it/s]
Attaching quantization:  92%|##################################################################      | 745/812 [01:47<00:08,  7.80it/s]
Attaching quantization:  92%|##################################################################1     | 746/812 [01:47<00:09,  7.10it/s]
Attaching quantization:  92%|##################################################################1     | 746/812 [01:47<00:09,  7.10it/s]
Attaching quantization:  92%|##################################################################3     | 748/812 [01:47<00:08,  7.87it/s]
Attaching quantization:  92%|##################################################################3     | 748/812 [01:47<00:08,  7.87it/s]
Attaching quantization:  92%|##################################################################5     | 750/812 [01:48<00:15,  3.96it/s]
Attaching quantization:  92%|##################################################################5     | 750/812 [01:48<00:15,  3.96it/s]
Attaching quantization:  92%|##################################################################5     | 751/812 [01:49<00:21,  2.86it/s]
Attaching quantization:  92%|##################################################################5     | 751/812 [01:49<00:21,  2.86it/s]
Attaching quantization:  93%|##################################################################7     | 753/812 [01:49<00:14,  3.97it/s]
Attaching quantization:  93%|##################################################################7     | 753/812 [01:49<00:14,  3.97it/s]
Attaching quantization:  93%|##################################################################9     | 755/812 [01:49<00:10,  5.24it/s]
Attaching quantization:  93%|##################################################################9     | 755/812 [01:49<00:10,  5.24it/s]
Attaching quantization:  93%|###################################################################     | 756/812 [01:49<00:10,  5.21it/s]
Attaching quantization:  93%|###################################################################     | 756/812 [01:49<00:10,  5.21it/s]
Attaching quantization:  93%|###################################################################2    | 758/812 [01:49<00:08,  6.54it/s]
Attaching quantization:  93%|###################################################################2    | 758/812 [01:49<00:08,  6.54it/s]
Attaching quantization:  93%|###################################################################3    | 759/812 [01:50<00:07,  6.84it/s]
Attaching quantization:  93%|###################################################################3    | 759/812 [01:50<00:07,  6.84it/s]
Attaching quantization:  94%|###################################################################3    | 760/812 [01:50<00:10,  4.81it/s]
Attaching quantization:  94%|###################################################################3    | 760/812 [01:50<00:10,  4.81it/s]
Attaching quantization:  94%|###################################################################4    | 761/812 [01:50<00:12,  4.25it/s]
Attaching quantization:  94%|###################################################################4    | 761/812 [01:50<00:12,  4.25it/s]
Attaching quantization:  94%|###################################################################6    | 763/812 [01:50<00:08,  5.86it/s]
Attaching quantization:  94%|###################################################################6    | 763/812 [01:50<00:08,  5.86it/s]
Attaching quantization:  94%|###################################################################8    | 765/812 [01:51<00:06,  7.22it/s]
Attaching quantization:  94%|###################################################################8    | 765/812 [01:51<00:06,  7.22it/s]
Attaching quantization:  94%|###################################################################9    | 766/812 [01:51<00:07,  5.75it/s]
Attaching quantization:  94%|###################################################################9    | 766/812 [01:51<00:07,  5.75it/s]
Attaching quantization:  94%|####################################################################    | 767/812 [01:51<00:09,  4.61it/s]
Attaching quantization:  94%|####################################################################    | 767/812 [01:51<00:09,  4.61it/s]
Attaching quantization:  95%|####################################################################1   | 769/812 [01:51<00:07,  5.62it/s]
Attaching quantization:  95%|####################################################################1   | 769/812 [01:51<00:07,  5.62it/s]
Attaching quantization:  95%|####################################################################2   | 770/812 [01:52<00:08,  4.88it/s]
Attaching quantization:  95%|####################################################################2   | 770/812 [01:52<00:08,  4.88it/s]
Attaching quantization:  95%|####################################################################3   | 771/812 [01:52<00:11,  3.58it/s]
Attaching quantization:  95%|####################################################################3   | 771/812 [01:52<00:11,  3.58it/s]
Attaching quantization:  95%|####################################################################5   | 773/812 [01:52<00:07,  4.93it/s]
Attaching quantization:  95%|####################################################################5   | 773/812 [01:52<00:07,  4.93it/s]
Attaching quantization:  95%|####################################################################7   | 775/812 [01:53<00:05,  6.47it/s]
Attaching quantization:  95%|####################################################################7   | 775/812 [01:53<00:05,  6.47it/s]
Attaching quantization:  96%|####################################################################8   | 777/812 [01:53<00:04,  7.51it/s]
Attaching quantization:  96%|####################################################################8   | 777/812 [01:53<00:04,  7.51it/s]
Attaching quantization:  96%|####################################################################9   | 778/812 [01:53<00:04,  6.96it/s]
Attaching quantization:  96%|####################################################################9   | 778/812 [01:53<00:04,  6.96it/s]
Attaching quantization:  96%|#####################################################################   | 779/812 [01:53<00:04,  7.34it/s]
Attaching quantization:  96%|#####################################################################   | 779/812 [01:53<00:04,  7.34it/s]
Attaching quantization:  96%|#####################################################################1  | 780/812 [01:54<00:08,  3.85it/s]
Attaching quantization:  96%|#####################################################################1  | 780/812 [01:54<00:08,  3.85it/s]
Attaching quantization:  96%|#####################################################################2  | 781/812 [01:54<00:08,  3.85it/s]
Attaching quantization:  96%|#####################################################################2  | 781/812 [01:54<00:08,  3.85it/s]
Attaching quantization:  96%|#####################################################################4  | 783/812 [01:54<00:05,  5.47it/s]
Attaching quantization:  96%|#####################################################################4  | 783/812 [01:54<00:05,  5.47it/s]
Attaching quantization:  97%|#####################################################################5  | 784/812 [01:54<00:04,  5.78it/s]
Attaching quantization:  97%|#####################################################################5  | 784/812 [01:54<00:04,  5.78it/s]
Attaching quantization:  97%|#####################################################################6  | 786/812 [01:55<00:03,  6.82it/s]
Attaching quantization:  97%|#####################################################################6  | 786/812 [01:55<00:03,  6.82it/s]
Attaching quantization:  97%|#####################################################################8  | 788/812 [01:55<00:03,  7.39it/s]
Attaching quantization:  97%|#####################################################################8  | 788/812 [01:55<00:03,  7.39it/s]
Attaching quantization:  97%|######################################################################  | 790/812 [01:55<00:03,  6.03it/s]
Attaching quantization:  97%|######################################################################  | 790/812 [01:55<00:03,  6.03it/s]
Attaching quantization:  97%|######################################################################1 | 791/812 [01:56<00:04,  5.20it/s]
Attaching quantization:  97%|######################################################################1 | 791/812 [01:56<00:04,  5.20it/s]
Attaching quantization:  98%|######################################################################3 | 793/812 [01:56<00:02,  6.67it/s]
Attaching quantization:  98%|######################################################################3 | 793/812 [01:56<00:02,  6.67it/s]
Attaching quantization:  98%|######################################################################4 | 795/812 [01:56<00:02,  7.40it/s]
Attaching quantization:  98%|######################################################################4 | 795/812 [01:56<00:02,  7.40it/s]
Attaching quantization:  98%|######################################################################6 | 797/812 [01:56<00:01,  8.73it/s]
Attaching quantization:  98%|######################################################################6 | 797/812 [01:56<00:01,  8.73it/s]
Attaching quantization:  98%|######################################################################8 | 799/812 [01:56<00:01,  8.96it/s]
Attaching quantization:  98%|######################################################################8 | 799/812 [01:56<00:01,  8.96it/s]
Attaching quantization:  99%|######################################################################9 | 800/812 [01:57<00:01,  6.39it/s]
Attaching quantization:  99%|######################################################################9 | 800/812 [01:57<00:01,  6.39it/s]
Attaching quantization:  99%|####################################################################### | 801/812 [01:57<00:01,  5.60it/s]
Attaching quantization:  99%|####################################################################### | 801/812 [01:57<00:01,  5.60it/s]
Attaching quantization:  99%|#######################################################################2| 803/812 [01:57<00:01,  6.92it/s]
Attaching quantization:  99%|#######################################################################2| 803/812 [01:57<00:01,  6.92it/s]
Attaching quantization:  99%|#######################################################################3| 805/812 [01:57<00:00,  7.82it/s]
Attaching quantization:  99%|#######################################################################3| 805/812 [01:57<00:00,  7.82it/s]
Attaching quantization:  99%|#######################################################################5| 807/812 [01:57<00:00,  9.08it/s]
Attaching quantization:  99%|#######################################################################5| 807/812 [01:57<00:00,  9.08it/s]
Attaching quantization: 100%|#######################################################################7| 809/812 [01:58<00:00,  9.29it/s]
Attaching quantization: 100%|#######################################################################7| 809/812 [01:58<00:00,  9.29it/s]
Attaching quantization: 100%|#######################################################################9| 811/812 [01:58<00:00,  5.67it/s]
Attaching quantization: 100%|#######################################################################9| 811/812 [01:58<00:00,  5.67it/s]
Attaching quantization: 100%|########################################################################| 812/812 [01:58<00:00,  6.84it/s]
Attaching quantization: 100%|########################################################################| 812/812 [01:58<00:00,  6.84it/s]
+
+ - quantizing additional layers
+Loading UMT5EncoderModel
+
tokenizer_config.json: 0.00B [00:00, ?B/s]
tokenizer_config.json: 0.00B [00:00, ?B/s]
tokenizer_config.json: 61.8kB [00:00, 28.7MB/s]
tokenizer_config.json: 61.8kB [00:00, 28.7MB/s]
+
+
tokenizer/spiece.model:   0%|                                                                              | 0.00/4.55M [00:00<?, ?B/s]
tokenizer/spiece.model:   0%|                                                                              | 0.00/4.55M [00:00<?, ?B/s]
tokenizer/spiece.model:   3%|##                                                                    | 130k/4.55M [00:02<01:22, 53.8kB/s]
tokenizer/spiece.model:   3%|##                                                                    | 130k/4.55M [00:02<01:22, 53.8kB/s]
tokenizer/spiece.model: 100%|#####################################################################| 4.55M/4.55M [00:03<00:00, 1.40MB/s]
tokenizer/spiece.model: 100%|#####################################################################| 4.55M/4.55M [00:03<00:00, 1.40MB/s]
tokenizer/spiece.model: 100%|#####################################################################| 4.55M/4.55M [00:03<00:00, 1.15MB/s]
tokenizer/spiece.model: 100%|#####################################################################| 4.55M/4.55M [00:03<00:00, 1.15MB/s]
+
+
added_tokens.json: 0.00B [00:00, ?B/s]
added_tokens.json: 0.00B [00:00, ?B/s]
added_tokens.json: 8.29kB [00:00, 5.56MB/s]
added_tokens.json: 8.29kB [00:00, 5.56MB/s]
+
+
special_tokens_map.json: 0.00B [00:00, ?B/s]
special_tokens_map.json: 0.00B [00:00, ?B/s]
special_tokens_map.json: 7.08kB [00:00, 23.5MB/s]
special_tokens_map.json: 7.08kB [00:00, 23.5MB/s]
+
+You set `add_prefix_space`. The tokenizer needs to be converted from the slow tokenizers
+You set `add_prefix_space`. The tokenizer needs to be converted from the slow tokenizers
+Using ai-toolkit/umt5_xxl_encoder for UMT5 encoder.
+
config.json:   0%|                                                                                           | 0.00/812 [00:00<?, ?B/s]
config.json:   0%|                                                                                           | 0.00/812 [00:00<?, ?B/s]
config.json: 100%|####################################################################################| 812/812 [00:00<00:00, 3.44MB/s]
config.json: 100%|####################################################################################| 812/812 [00:00<00:00, 3.44MB/s]
+
+
model.safetensors.index.json: 0.00B [00:00, ?B/s]
model.safetensors.index.json: 0.00B [00:00, ?B/s]
model.safetensors.index.json: 22.5kB [00:00, 12.6MB/s]
model.safetensors.index.json: 22.5kB [00:00, 12.6MB/s]
+
+
Fetching 3 files:   0%|                                                                                          | 0/3 [00:00<?, ?it/s]
Fetching 3 files:   0%|                                                                                          | 0/3 [00:00<?, ?it/s]
+
+
text_encoder/model-00003-of-00003.safete(…):   0%|                                                         | 0.00/1.44G [00:00<?, ?B/s]
text_encoder/model-00003-of-00003.safete(…):   0%|                                                         | 0.00/1.44G [00:00<?, ?B/s]
+
+
+
+
text_encoder/model-00002-of-00003.safete(…):   0%|                                                         | 0.00/4.98G [00:00<?, ?B/s]
text_encoder/model-00002-of-00003.safete(…):   0%|                                                         | 0.00/4.98G [00:00<?, ?B/s]
+
+
+
+
+
+
text_encoder/model-00001-of-00003.safete(…):   0%|                                                         | 0.00/4.94G [00:00<?, ?B/s]
text_encoder/model-00001-of-00003.safete(…):   0%|                                                         | 0.00/4.94G [00:00<?, ?B/s]
+
+
text_encoder/model-00003-of-00003.safete(…):   0%|                                                | 560k/1.44G [00:02<1:32:17, 260kB/s]
text_encoder/model-00003-of-00003.safete(…):   0%|                                                | 560k/1.44G [00:02<1:32:17, 260kB/s]
+
+
+
+
text_encoder/model-00002-of-00003.safete(…):   0%|                                                | 710k/4.98G [00:03<7:04:20, 196kB/s]
text_encoder/model-00002-of-00003.safete(…):   0%|                                                | 710k/4.98G [00:03<7:04:20, 196kB/s]
+
+
text_encoder/model-00003-of-00003.safete(…):   0%|                                                | 560k/1.44G [00:14<1:32:17, 260kB/s]
text_encoder/model-00003-of-00003.safete(…):   0%|                                                | 560k/1.44G [00:14<1:32:17, 260kB/s]
+
+
+
+
text_encoder/model-00002-of-00003.safete(…):   0%|                                                | 710k/4.98G [00:14<7:04:20, 196kB/s]
text_encoder/model-00002-of-00003.safete(…):   0%|                                                | 710k/4.98G [00:14<7:04:20, 196kB/s]
+
+
text_encoder/model-00003-of-00003.safete(…):   2%|#1                                              | 35.1M/1.44G [00:15<09:50, 2.38MB/s]
text_encoder/model-00003-of-00003.safete(…):   2%|#1                                              | 35.1M/1.44G [00:15<09:50, 2.38MB/s]
+
+
+
+
+
+
text_encoder/model-00001-of-00003.safete(…):   0%|                                              | 764k/4.94G [00:15<27:19:56, 50.2kB/s]
text_encoder/model-00001-of-00003.safete(…):   0%|                                              | 764k/4.94G [00:15<27:19:56, 50.2kB/s]
+
+
text_encoder/model-00003-of-00003.safete(…):   7%|###4                                             | 102M/1.44G [00:23<04:23, 5.10MB/s]
text_encoder/model-00003-of-00003.safete(…):   7%|###4                                             | 102M/1.44G [00:23<04:23, 5.10MB/s]
+
+
text_encoder/model-00003-of-00003.safete(…):  12%|#####7                                           | 169M/1.44G [00:23<02:03, 10.3MB/s]
text_encoder/model-00003-of-00003.safete(…):  12%|#####7                                           | 169M/1.44G [00:23<02:03, 10.3MB/s]
+
+
text_encoder/model-00003-of-00003.safete(…):  16%|########                                         | 236M/1.44G [00:23<01:13, 16.3MB/s]
text_encoder/model-00003-of-00003.safete(…):  16%|########                                         | 236M/1.44G [00:23<01:13, 16.3MB/s]
+
+
text_encoder/model-00003-of-00003.safete(…):  21%|##########3                                      | 304M/1.44G [00:24<00:44, 25.6MB/s]
text_encoder/model-00003-of-00003.safete(…):  21%|##########3                                      | 304M/1.44G [00:24<00:44, 25.6MB/s]
+
+
text_encoder/model-00003-of-00003.safete(…):  26%|############5                                    | 371M/1.44G [00:24<00:28, 37.6MB/s]
text_encoder/model-00003-of-00003.safete(…):  26%|############5                                    | 371M/1.44G [00:24<00:28, 37.6MB/s]
+
+
+
+
text_encoder/model-00002-of-00003.safete(…):   1%|6                                               | 67.8M/4.98G [00:26<30:13, 2.71MB/s]
text_encoder/model-00002-of-00003.safete(…):   1%|6                                               | 67.8M/4.98G [00:26<30:13, 2.71MB/s]
+
+
text_encoder/model-00003-of-00003.safete(…):  30%|##############8                                  | 438M/1.44G [00:26<00:30, 33.2MB/s]
text_encoder/model-00003-of-00003.safete(…):  30%|##############8                                  | 438M/1.44G [00:26<00:30, 33.2MB/s]
+
+
text_encoder/model-00003-of-00003.safete(…):  35%|#################1                               | 505M/1.44G [00:27<00:21, 44.5MB/s]
text_encoder/model-00003-of-00003.safete(…):  35%|#################1                               | 505M/1.44G [00:27<00:21, 44.5MB/s]
+
+
+
+
text_encoder/model-00002-of-00003.safete(…):   4%|#9                                               | 202M/4.98G [00:27<08:12, 9.70MB/s]
text_encoder/model-00002-of-00003.safete(…):   4%|#9                                               | 202M/4.98G [00:27<08:12, 9.70MB/s]
+
+
+
+
text_encoder/model-00002-of-00003.safete(…):   7%|###3                                             | 336M/4.98G [00:27<03:56, 19.7MB/s]
text_encoder/model-00002-of-00003.safete(…):   7%|###3                                             | 336M/4.98G [00:27<03:56, 19.7MB/s]
+
+
+
+
text_encoder/model-00002-of-00003.safete(…):  11%|#####2                                           | 537M/4.98G [00:27<01:49, 40.6MB/s]
text_encoder/model-00002-of-00003.safete(…):  11%|#####2                                           | 537M/4.98G [00:27<01:49, 40.6MB/s]
+
+
+
+
+
+
text_encoder/model-00001-of-00003.safete(…):   0%|                                              | 764k/4.94G [00:28<27:19:56, 50.2kB/s]
text_encoder/model-00001-of-00003.safete(…):   0%|                                              | 764k/4.94G [00:28<27:19:56, 50.2kB/s]
+
+
+
+
+
+
text_encoder/model-00001-of-00003.safete(…):   1%|6                                               | 67.8M/4.94G [00:28<28:48, 2.82MB/s]
text_encoder/model-00001-of-00003.safete(…):   1%|6                                               | 67.8M/4.94G [00:28<28:48, 2.82MB/s]
+
+
text_encoder/model-00003-of-00003.safete(…):  40%|###################4                             | 572M/1.44G [00:29<00:21, 40.8MB/s]
text_encoder/model-00003-of-00003.safete(…):  40%|###################4                             | 572M/1.44G [00:29<00:21, 40.8MB/s]
+
+
+
+
+
+
text_encoder/model-00001-of-00003.safete(…):   3%|#3                                               | 135M/4.94G [00:29<12:09, 6.59MB/s]
text_encoder/model-00001-of-00003.safete(…):   3%|#3                                               | 135M/4.94G [00:29<12:09, 6.59MB/s]
+
+
text_encoder/model-00003-of-00003.safete(…):  49%|#######################9                         | 706M/1.44G [00:29<00:09, 75.5MB/s]
text_encoder/model-00003-of-00003.safete(…):  49%|#######################9                         | 706M/1.44G [00:29<00:09, 75.5MB/s]
+
+
text_encoder/model-00003-of-00003.safete(…):  54%|##########################2                      | 773M/1.44G [00:29<00:07, 89.2MB/s]
text_encoder/model-00003-of-00003.safete(…):  54%|##########################2                      | 773M/1.44G [00:29<00:07, 89.2MB/s]
+
+
text_encoder/model-00003-of-00003.safete(…):  58%|############################5                    | 840M/1.44G [00:30<00:06, 91.4MB/s]
text_encoder/model-00003-of-00003.safete(…):  58%|############################5                    | 840M/1.44G [00:30<00:06, 91.4MB/s]
+
+
+
+
text_encoder/model-00002-of-00003.safete(…):  14%|######6                                          | 673M/4.98G [00:30<01:39, 43.5MB/s]
text_encoder/model-00002-of-00003.safete(…):  14%|######6                                          | 673M/4.98G [00:30<01:39, 43.5MB/s]
+
+
text_encoder/model-00003-of-00003.safete(…):  63%|###############################4                  | 907M/1.44G [00:30<00:05, 107MB/s]
text_encoder/model-00003-of-00003.safete(…):  63%|###############################4                  | 907M/1.44G [00:30<00:05, 107MB/s]
+
+
+
+
+
+
text_encoder/model-00001-of-00003.safete(…):   4%|##                                               | 202M/4.94G [00:30<07:20, 10.8MB/s]
text_encoder/model-00001-of-00003.safete(…):   4%|##                                               | 202M/4.94G [00:30<07:20, 10.8MB/s]
+
+
+
+
+
+
text_encoder/model-00001-of-00003.safete(…):   5%|##6                                              | 269M/4.94G [00:30<04:28, 17.4MB/s]
text_encoder/model-00001-of-00003.safete(…):   5%|##6                                              | 269M/4.94G [00:30<04:28, 17.4MB/s]
+
+
text_encoder/model-00003-of-00003.safete(…):  72%|###################################3             | 1.04G/1.44G [00:31<00:03, 106MB/s]
text_encoder/model-00003-of-00003.safete(…):  72%|###################################3             | 1.04G/1.44G [00:31<00:03, 106MB/s]
+
+
+
+
text_encoder/model-00002-of-00003.safete(…):  15%|#######2                                         | 740M/4.98G [00:31<01:39, 42.7MB/s]
text_encoder/model-00002-of-00003.safete(…):  15%|#######2                                         | 740M/4.98G [00:31<01:39, 42.7MB/s]
+
+
+
+
+
+
text_encoder/model-00001-of-00003.safete(…):   7%|###3                                             | 336M/4.94G [00:32<03:17, 23.3MB/s]
text_encoder/model-00001-of-00003.safete(…):   7%|###3                                             | 336M/4.94G [00:32<03:17, 23.3MB/s]
+
+
text_encoder/model-00003-of-00003.safete(…):  77%|#####################################6           | 1.11G/1.44G [00:32<00:02, 129MB/s]
text_encoder/model-00003-of-00003.safete(…):  77%|#####################################6           | 1.11G/1.44G [00:32<00:02, 129MB/s]
+
+
text_encoder/model-00003-of-00003.safete(…):  86%|##########################################1      | 1.24G/1.44G [00:32<00:00, 202MB/s]
text_encoder/model-00003-of-00003.safete(…):  86%|##########################################1      | 1.24G/1.44G [00:32<00:00, 202MB/s]
+
+
+
+
+
+
text_encoder/model-00001-of-00003.safete(…):   8%|####                                             | 403M/4.94G [00:36<03:58, 19.0MB/s]
text_encoder/model-00001-of-00003.safete(…):   8%|####                                             | 403M/4.94G [00:36<03:58, 19.0MB/s]
+
+
+
+
+
+
text_encoder/model-00001-of-00003.safete(…):  10%|####6                                            | 472M/4.94G [00:38<03:10, 23.5MB/s]
text_encoder/model-00001-of-00003.safete(…):  10%|####6                                            | 472M/4.94G [00:38<03:10, 23.5MB/s]
+
+
+
+
text_encoder/model-00002-of-00003.safete(…):  18%|########5                                        | 874M/4.98G [00:40<02:29, 27.5MB/s]
text_encoder/model-00002-of-00003.safete(…):  18%|########5                                        | 874M/4.98G [00:40<02:29, 27.5MB/s]
+
+
text_encoder/model-00003-of-00003.safete(…):  95%|#############################################7  | 1.38G/1.44G [00:40<00:01, 40.5MB/s]
text_encoder/model-00003-of-00003.safete(…):  95%|#############################################7  | 1.38G/1.44G [00:40<00:01, 40.5MB/s]
+
+
+
+
+
+
text_encoder/model-00001-of-00003.safete(…):  11%|#####3                                           | 539M/4.94G [00:40<02:43, 26.9MB/s]
text_encoder/model-00001-of-00003.safete(…):  11%|#####3                                           | 539M/4.94G [00:40<02:43, 26.9MB/s]
+
+
+
+
text_encoder/model-00002-of-00003.safete(…):  19%|#########2                                       | 941M/4.98G [00:42<02:24, 27.9MB/s]
text_encoder/model-00002-of-00003.safete(…):  19%|#########2                                       | 941M/4.98G [00:42<02:24, 27.9MB/s]
+
+
+
+
+
+
text_encoder/model-00001-of-00003.safete(…):  12%|######                                           | 606M/4.94G [00:42<02:42, 26.7MB/s]
text_encoder/model-00001-of-00003.safete(…):  12%|######                                           | 606M/4.94G [00:42<02:42, 26.7MB/s]
+
+
+
+
text_encoder/model-00002-of-00003.safete(…):  20%|#########7                                      | 1.01G/4.98G [00:44<02:19, 28.4MB/s]
text_encoder/model-00002-of-00003.safete(…):  20%|#########7                                      | 1.01G/4.98G [00:44<02:19, 28.4MB/s]
+
+
text_encoder/model-00003-of-00003.safete(…): 100%|################################################| 1.44G/1.44G [00:44<00:00, 29.9MB/s]
text_encoder/model-00003-of-00003.safete(…): 100%|################################################| 1.44G/1.44G [00:44<00:00, 29.9MB/s]
text_encoder/model-00003-of-00003.safete(…): 100%|################################################| 1.44G/1.44G [00:44<00:00, 32.4MB/s]
text_encoder/model-00003-of-00003.safete(…): 100%|################################################| 1.44G/1.44G [00:44<00:00, 32.4MB/s]
+
+
+
+
+
+
text_encoder/model-00002-of-00003.safete(…):  22%|##########3                                     | 1.08G/4.98G [00:45<01:59, 32.7MB/s]
text_encoder/model-00002-of-00003.safete(…):  22%|##########3                                     | 1.08G/4.98G [00:45<01:59, 32.7MB/s]
+
+
+
+
text_encoder/model-00002-of-00003.safete(…):  23%|###########                                     | 1.14G/4.98G [00:47<01:52, 34.1MB/s]
text_encoder/model-00002-of-00003.safete(…):  23%|###########                                     | 1.14G/4.98G [00:47<01:52, 34.1MB/s]
+
+
+
+
+
+
text_encoder/model-00001-of-00003.safete(…):  14%|######6                                          | 673M/4.94G [00:47<03:23, 20.9MB/s]
text_encoder/model-00001-of-00003.safete(…):  14%|######6                                          | 673M/4.94G [00:47<03:23, 20.9MB/s]
+
+
+
+
text_encoder/model-00002-of-00003.safete(…):  24%|###########6                                    | 1.21G/4.98G [00:49<01:47, 35.0MB/s]
text_encoder/model-00002-of-00003.safete(…):  24%|###########6                                    | 1.21G/4.98G [00:49<01:47, 35.0MB/s]
+
+
+
+
text_encoder/model-00002-of-00003.safete(…):  26%|############2                                   | 1.28G/4.98G [00:51<01:58, 31.2MB/s]
text_encoder/model-00002-of-00003.safete(…):  26%|############2                                   | 1.28G/4.98G [00:51<01:58, 31.2MB/s]
+
+
+
+
+
+
text_encoder/model-00001-of-00003.safete(…):  15%|#######3                                         | 740M/4.94G [00:52<03:48, 18.4MB/s]
text_encoder/model-00001-of-00003.safete(…):  15%|#######3                                         | 740M/4.94G [00:52<03:48, 18.4MB/s]
+
+
+
+
text_encoder/model-00002-of-00003.safete(…):  27%|############9                                   | 1.34G/4.98G [00:53<01:47, 33.9MB/s]
text_encoder/model-00002-of-00003.safete(…):  27%|############9                                   | 1.34G/4.98G [00:53<01:47, 33.9MB/s]
+
+
+
+
+
+
text_encoder/model-00001-of-00003.safete(…):  18%|########6                                        | 874M/4.94G [00:53<02:17, 29.4MB/s]
text_encoder/model-00001-of-00003.safete(…):  18%|########6                                        | 874M/4.94G [00:53<02:17, 29.4MB/s]
+
+
+
+
text_encoder/model-00002-of-00003.safete(…):  28%|#############5                                  | 1.41G/4.98G [00:54<01:29, 40.0MB/s]
text_encoder/model-00002-of-00003.safete(…):  28%|#############5                                  | 1.41G/4.98G [00:54<01:29, 40.0MB/s]
+
+
+
+
+
+
text_encoder/model-00001-of-00003.safete(…):  19%|#########3                                       | 940M/4.94G [00:54<02:00, 33.1MB/s]
text_encoder/model-00001-of-00003.safete(…):  19%|#########3                                       | 940M/4.94G [00:54<02:00, 33.1MB/s]
+
+
+
+
+
+
text_encoder/model-00001-of-00003.safete(…):  20%|#########7                                      | 1.01G/4.94G [00:54<01:31, 42.8MB/s]
text_encoder/model-00001-of-00003.safete(…):  20%|#########7                                      | 1.01G/4.94G [00:54<01:31, 42.8MB/s]
+
+
+
+
text_encoder/model-00002-of-00003.safete(…):  30%|##############2                                 | 1.48G/4.98G [00:55<01:12, 48.2MB/s]
text_encoder/model-00002-of-00003.safete(…):  30%|##############2                                 | 1.48G/4.98G [00:55<01:12, 48.2MB/s]
+
+
+
+
text_encoder/model-00002-of-00003.safete(…):  34%|################5                                | 1.68G/4.98G [00:55<00:31, 104MB/s]
text_encoder/model-00002-of-00003.safete(…):  34%|################5                                | 1.68G/4.98G [00:55<00:31, 104MB/s]
+
+
+
+
+
+
text_encoder/model-00001-of-00003.safete(…):  23%|###########1                                    | 1.14G/4.94G [00:56<01:09, 54.9MB/s]
text_encoder/model-00001-of-00003.safete(…):  23%|###########1                                    | 1.14G/4.94G [00:56<01:09, 54.9MB/s]
+
+
+
+
text_encoder/model-00002-of-00003.safete(…):  35%|################8                               | 1.75G/4.98G [00:56<00:39, 81.7MB/s]
text_encoder/model-00002-of-00003.safete(…):  35%|################8                               | 1.75G/4.98G [00:56<00:39, 81.7MB/s]
+
+
+
+
+
+
text_encoder/model-00001-of-00003.safete(…):  27%|#############                                   | 1.34G/4.94G [00:56<00:36, 97.5MB/s]
text_encoder/model-00001-of-00003.safete(…):  27%|#############                                   | 1.34G/4.94G [00:56<00:36, 97.5MB/s]
+
+
+
+
text_encoder/model-00002-of-00003.safete(…):  36%|#################4                              | 1.81G/4.98G [00:56<00:31, 99.6MB/s]
text_encoder/model-00002-of-00003.safete(…):  36%|#################4                              | 1.81G/4.98G [00:56<00:31, 99.6MB/s]
+
+
+
+
+
+
text_encoder/model-00001-of-00003.safete(…):  30%|##############6                                  | 1.48G/4.94G [00:57<00:31, 110MB/s]
text_encoder/model-00001-of-00003.safete(…):  30%|##############6                                  | 1.48G/4.94G [00:57<00:31, 110MB/s]
+
+
+
+
+
+
text_encoder/model-00001-of-00003.safete(…):  31%|###############3                                 | 1.54G/4.94G [00:58<00:28, 118MB/s]
text_encoder/model-00001-of-00003.safete(…):  31%|###############3                                 | 1.54G/4.94G [00:58<00:28, 118MB/s]
+
+
+
+
text_encoder/model-00002-of-00003.safete(…):  38%|##################1                             | 1.88G/4.98G [00:58<00:38, 80.3MB/s]
text_encoder/model-00002-of-00003.safete(…):  38%|##################1                             | 1.88G/4.98G [00:58<00:38, 80.3MB/s]
+
+
+
+
+
+
text_encoder/model-00001-of-00003.safete(…):  33%|###############6                                | 1.61G/4.94G [00:59<00:35, 92.8MB/s]
text_encoder/model-00001-of-00003.safete(…):  33%|###############6                                | 1.61G/4.94G [00:59<00:35, 92.8MB/s]
+
+
+
+
text_encoder/model-00002-of-00003.safete(…):  39%|##################7                             | 1.95G/4.98G [00:59<00:42, 71.3MB/s]
text_encoder/model-00002-of-00003.safete(…):  39%|##################7                             | 1.95G/4.98G [00:59<00:42, 71.3MB/s]
+
+
+
+
text_encoder/model-00002-of-00003.safete(…):  40%|###################4                            | 2.02G/4.98G [00:59<00:36, 81.8MB/s]
text_encoder/model-00002-of-00003.safete(…):  40%|###################4                            | 2.02G/4.98G [00:59<00:36, 81.8MB/s]
+
+
+
+
text_encoder/model-00002-of-00003.safete(…):  43%|####################7                           | 2.15G/4.98G [01:01<00:29, 96.0MB/s]
text_encoder/model-00002-of-00003.safete(…):  43%|####################7                           | 2.15G/4.98G [01:01<00:29, 96.0MB/s]
+
+
+
+
+
+
text_encoder/model-00001-of-00003.safete(…):  34%|################3                               | 1.68G/4.94G [01:01<00:46, 70.5MB/s]
text_encoder/model-00001-of-00003.safete(…):  34%|################3                               | 1.68G/4.94G [01:01<00:46, 70.5MB/s]
+
+
+
+
+
+
text_encoder/model-00001-of-00003.safete(…):  37%|#################9                               | 1.81G/4.94G [01:01<00:28, 111MB/s]
text_encoder/model-00001-of-00003.safete(…):  37%|#################9                               | 1.81G/4.94G [01:01<00:28, 111MB/s]
+
+
+
+
text_encoder/model-00002-of-00003.safete(…):  44%|#####################3                          | 2.22G/4.98G [01:02<00:38, 71.6MB/s]
text_encoder/model-00002-of-00003.safete(…):  44%|#####################3                          | 2.22G/4.98G [01:02<00:38, 71.6MB/s]
+
+
+
+
text_encoder/model-00002-of-00003.safete(…):  46%|######################                          | 2.28G/4.98G [01:02<00:29, 92.3MB/s]
text_encoder/model-00002-of-00003.safete(…):  46%|######################                          | 2.28G/4.98G [01:02<00:29, 92.3MB/s]
+
+
+
+
+
+
text_encoder/model-00001-of-00003.safete(…):  39%|##################9                             | 1.95G/4.94G [01:03<00:37, 79.0MB/s]
text_encoder/model-00001-of-00003.safete(…):  39%|##################9                             | 1.95G/4.94G [01:03<00:37, 79.0MB/s]
+
+
+
+
text_encoder/model-00002-of-00003.safete(…):  47%|######################6                         | 2.35G/4.98G [01:04<00:33, 77.9MB/s]
text_encoder/model-00002-of-00003.safete(…):  47%|######################6                         | 2.35G/4.98G [01:04<00:33, 77.9MB/s]
+
+
+
+
+
+
text_encoder/model-00001-of-00003.safete(…):  41%|###################5                            | 2.01G/4.94G [01:08<01:12, 40.2MB/s]
text_encoder/model-00001-of-00003.safete(…):  41%|###################5                            | 2.01G/4.94G [01:08<01:12, 40.2MB/s]
+
+
+
+
+
+
text_encoder/model-00001-of-00003.safete(…):  44%|####################8                           | 2.15G/4.94G [01:10<00:57, 48.6MB/s]
text_encoder/model-00001-of-00003.safete(…):  44%|####################8                           | 2.15G/4.94G [01:10<00:57, 48.6MB/s]
+
+
+
+
text_encoder/model-00002-of-00003.safete(…):  49%|#######################2                        | 2.42G/4.98G [01:10<01:31, 27.9MB/s]
text_encoder/model-00002-of-00003.safete(…):  49%|#######################2                        | 2.42G/4.98G [01:10<01:31, 27.9MB/s]
+
+
+
+
+
+
text_encoder/model-00001-of-00003.safete(…):  45%|#####################5                          | 2.21G/4.94G [01:11<00:54, 49.6MB/s]
text_encoder/model-00001-of-00003.safete(…):  45%|#####################5                          | 2.21G/4.94G [01:11<00:54, 49.6MB/s]
+
+
+
+
text_encoder/model-00002-of-00003.safete(…):  51%|########################5                       | 2.55G/4.98G [01:14<01:19, 30.6MB/s]
text_encoder/model-00002-of-00003.safete(…):  51%|########################5                       | 2.55G/4.98G [01:14<01:19, 30.6MB/s]
+
+
+
+
text_encoder/model-00002-of-00003.safete(…):  54%|#########################8                      | 2.69G/4.98G [01:15<00:51, 44.8MB/s]
text_encoder/model-00002-of-00003.safete(…):  54%|#########################8                      | 2.69G/4.98G [01:15<00:51, 44.8MB/s]
+
+
+
+
+
+
text_encoder/model-00001-of-00003.safete(…):  46%|######################2                         | 2.28G/4.94G [01:16<01:27, 30.4MB/s]
text_encoder/model-00001-of-00003.safete(…):  46%|######################2                         | 2.28G/4.94G [01:16<01:27, 30.4MB/s]
+
+
+
+
text_encoder/model-00002-of-00003.safete(…):  55%|##########################5                     | 2.75G/4.98G [01:16<00:51, 43.0MB/s]
text_encoder/model-00002-of-00003.safete(…):  55%|##########################5                     | 2.75G/4.98G [01:16<00:51, 43.0MB/s]
+
+
+
+
+
+
text_encoder/model-00001-of-00003.safete(…):  48%|######################8                         | 2.35G/4.94G [01:16<01:05, 39.6MB/s]
text_encoder/model-00001-of-00003.safete(…):  48%|######################8                         | 2.35G/4.94G [01:16<01:05, 39.6MB/s]
+
+
+
+
text_encoder/model-00002-of-00003.safete(…):  57%|###########################1                    | 2.82G/4.98G [01:18<00:47, 45.4MB/s]
text_encoder/model-00002-of-00003.safete(…):  57%|###########################1                    | 2.82G/4.98G [01:18<00:47, 45.4MB/s]
+
+
+
+
text_encoder/model-00002-of-00003.safete(…):  58%|############################                    | 2.91G/4.98G [01:18<00:37, 54.8MB/s]
text_encoder/model-00002-of-00003.safete(…):  58%|############################                    | 2.91G/4.98G [01:18<00:37, 54.8MB/s]
+
+
+
+
text_encoder/model-00002-of-00003.safete(…):  61%|#############################3                  | 3.04G/4.98G [01:19<00:24, 77.7MB/s]
text_encoder/model-00002-of-00003.safete(…):  61%|#############################3                  | 3.04G/4.98G [01:19<00:24, 77.7MB/s]
+
+
+
+
+
+
text_encoder/model-00001-of-00003.safete(…):  49%|#######################5                        | 2.42G/4.94G [01:19<01:13, 34.2MB/s]
text_encoder/model-00001-of-00003.safete(…):  49%|#######################5                        | 2.42G/4.94G [01:19<01:13, 34.2MB/s]
+
+
+
+
text_encoder/model-00002-of-00003.safete(…):  64%|##############################5                 | 3.18G/4.98G [01:21<00:22, 80.7MB/s]
text_encoder/model-00002-of-00003.safete(…):  64%|##############################5                 | 3.18G/4.98G [01:21<00:22, 80.7MB/s]
+
+
+
+
+
+
text_encoder/model-00001-of-00003.safete(…):  52%|########################8                       | 2.55G/4.94G [01:22<01:00, 39.7MB/s]
text_encoder/model-00001-of-00003.safete(…):  52%|########################8                       | 2.55G/4.94G [01:22<01:00, 39.7MB/s]
+
+
+
+
+
+
text_encoder/model-00001-of-00003.safete(…):  53%|#########################4                      | 2.62G/4.94G [01:22<00:46, 49.7MB/s]
text_encoder/model-00001-of-00003.safete(…):  53%|#########################4                      | 2.62G/4.94G [01:22<00:46, 49.7MB/s]
+
+
+
+
+
+
text_encoder/model-00001-of-00003.safete(…):  56%|##########################7                     | 2.75G/4.94G [01:24<00:37, 58.4MB/s]
text_encoder/model-00001-of-00003.safete(…):  56%|##########################7                     | 2.75G/4.94G [01:24<00:37, 58.4MB/s]
+
+
+
+
text_encoder/model-00002-of-00003.safete(…):  65%|###############################2                | 3.24G/4.98G [01:24<00:34, 51.1MB/s]
text_encoder/model-00002-of-00003.safete(…):  65%|###############################2                | 3.24G/4.98G [01:24<00:34, 51.1MB/s]
+
+
+
+
text_encoder/model-00002-of-00003.safete(…):  66%|###############################8                | 3.31G/4.98G [01:24<00:26, 63.8MB/s]
text_encoder/model-00002-of-00003.safete(…):  66%|###############################8                | 3.31G/4.98G [01:24<00:26, 63.8MB/s]
+
+
+
+
text_encoder/model-00002-of-00003.safete(…):  68%|################################5               | 3.38G/4.98G [01:25<00:26, 60.4MB/s]
text_encoder/model-00002-of-00003.safete(…):  68%|################################5               | 3.38G/4.98G [01:25<00:26, 60.4MB/s]
+
+
+
+
+
+
text_encoder/model-00001-of-00003.safete(…):  59%|############################                    | 2.89G/4.94G [01:25<00:30, 66.2MB/s]
text_encoder/model-00001-of-00003.safete(…):  59%|############################                    | 2.89G/4.94G [01:25<00:30, 66.2MB/s]
+
+
+
+
text_encoder/model-00002-of-00003.safete(…):  70%|#################################8              | 3.51G/4.98G [01:26<00:15, 95.7MB/s]
text_encoder/model-00002-of-00003.safete(…):  70%|#################################8              | 3.51G/4.98G [01:26<00:15, 95.7MB/s]
+
+
+
+
text_encoder/model-00002-of-00003.safete(…):  73%|###################################8             | 3.64G/4.98G [01:26<00:09, 145MB/s]
text_encoder/model-00002-of-00003.safete(…):  73%|###################################8             | 3.64G/4.98G [01:26<00:09, 145MB/s]
+
+
+
+
+
+
text_encoder/model-00001-of-00003.safete(…):  60%|############################7                   | 2.95G/4.94G [01:26<00:26, 75.3MB/s]
text_encoder/model-00001-of-00003.safete(…):  60%|############################7                   | 2.95G/4.94G [01:26<00:26, 75.3MB/s]
+
+
+
+
text_encoder/model-00002-of-00003.safete(…):  76%|#####################################1           | 3.78G/4.98G [01:26<00:05, 203MB/s]
text_encoder/model-00002-of-00003.safete(…):  76%|#####################################1           | 3.78G/4.98G [01:26<00:05, 203MB/s]
+
+
+
+
+
+
text_encoder/model-00001-of-00003.safete(…):  61%|#############################3                  | 3.02G/4.94G [01:27<00:28, 66.6MB/s]
text_encoder/model-00001-of-00003.safete(…):  61%|#############################3                  | 3.02G/4.94G [01:27<00:28, 66.6MB/s]
+
+
+
+
+
+
text_encoder/model-00001-of-00003.safete(…):  63%|##############################                  | 3.09G/4.94G [01:28<00:28, 63.9MB/s]
text_encoder/model-00001-of-00003.safete(…):  63%|##############################                  | 3.09G/4.94G [01:28<00:28, 63.9MB/s]
+
+
+
+
text_encoder/model-00002-of-00003.safete(…):  77%|#####################################           | 3.85G/4.98G [01:29<00:13, 81.5MB/s]
text_encoder/model-00002-of-00003.safete(…):  77%|#####################################           | 3.85G/4.98G [01:29<00:13, 81.5MB/s]
+
+
+
+
+
+
text_encoder/model-00001-of-00003.safete(…):  64%|##############################6                 | 3.16G/4.94G [01:29<00:22, 78.4MB/s]
text_encoder/model-00001-of-00003.safete(…):  64%|##############################6                 | 3.16G/4.94G [01:29<00:22, 78.4MB/s]
+
+
+
+
+
+
text_encoder/model-00001-of-00003.safete(…):  67%|###############################9                | 3.29G/4.94G [01:30<00:17, 94.6MB/s]
text_encoder/model-00001-of-00003.safete(…):  67%|###############################9                | 3.29G/4.94G [01:30<00:17, 94.6MB/s]
+
+
+
+
+
+
text_encoder/model-00001-of-00003.safete(…):  68%|################################6               | 3.36G/4.94G [01:31<00:20, 76.7MB/s]
text_encoder/model-00001-of-00003.safete(…):  68%|################################6               | 3.36G/4.94G [01:31<00:20, 76.7MB/s]
+
+
+
+
text_encoder/model-00002-of-00003.safete(…):  79%|#####################################6          | 3.91G/4.98G [01:32<00:23, 44.7MB/s]
text_encoder/model-00002-of-00003.safete(…):  79%|#####################################6          | 3.91G/4.98G [01:32<00:23, 44.7MB/s]
+
+
+
+
+
+
text_encoder/model-00001-of-00003.safete(…):  69%|#################################3              | 3.42G/4.94G [01:33<00:23, 64.6MB/s]
text_encoder/model-00001-of-00003.safete(…):  69%|#################################3              | 3.42G/4.94G [01:33<00:23, 64.6MB/s]
+
+
+
+
text_encoder/model-00002-of-00003.safete(…):  80%|######################################3         | 3.98G/4.98G [01:33<00:17, 56.0MB/s]
text_encoder/model-00002-of-00003.safete(…):  80%|######################################3         | 3.98G/4.98G [01:33<00:17, 56.0MB/s]
+
+
+
+
text_encoder/model-00002-of-00003.safete(…):  81%|######################################9         | 4.05G/4.98G [01:34<00:15, 59.3MB/s]
text_encoder/model-00002-of-00003.safete(…):  81%|######################################9         | 4.05G/4.98G [01:34<00:15, 59.3MB/s]
+
+
+
+
text_encoder/model-00002-of-00003.safete(…):  83%|#######################################6        | 4.11G/4.98G [01:37<00:21, 41.2MB/s]
text_encoder/model-00002-of-00003.safete(…):  83%|#######################################6        | 4.11G/4.98G [01:37<00:21, 41.2MB/s]
+
+
+
+
+
+
text_encoder/model-00001-of-00003.safete(…):  71%|#################################9              | 3.49G/4.94G [01:37<00:40, 35.4MB/s]
text_encoder/model-00001-of-00003.safete(…):  71%|#################################9              | 3.49G/4.94G [01:37<00:40, 35.4MB/s]
+
+
+
+
+
+
text_encoder/model-00001-of-00003.safete(…):  73%|###################################2            | 3.63G/4.94G [01:38<00:26, 50.3MB/s]
text_encoder/model-00001-of-00003.safete(…):  73%|###################################2            | 3.63G/4.94G [01:38<00:26, 50.3MB/s]
+
+
+
+
+
+
text_encoder/model-00001-of-00003.safete(…):  76%|####################################5           | 3.76G/4.94G [01:38<00:15, 78.2MB/s]
text_encoder/model-00001-of-00003.safete(…):  76%|####################################5           | 3.76G/4.94G [01:38<00:15, 78.2MB/s]
+
+
+
+
+
+
text_encoder/model-00001-of-00003.safete(…):  78%|#####################################2          | 3.83G/4.94G [01:38<00:12, 92.3MB/s]
text_encoder/model-00001-of-00003.safete(…):  78%|#####################################2          | 3.83G/4.94G [01:38<00:12, 92.3MB/s]
+
+
+
+
text_encoder/model-00002-of-00003.safete(…):  85%|########################################9       | 4.25G/4.98G [01:40<00:17, 41.4MB/s]
text_encoder/model-00002-of-00003.safete(…):  85%|########################################9       | 4.25G/4.98G [01:40<00:17, 41.4MB/s]
+
+
+
+
+
+
text_encoder/model-00001-of-00003.safete(…):  79%|#####################################8          | 3.89G/4.94G [01:40<00:13, 75.2MB/s]
text_encoder/model-00001-of-00003.safete(…):  79%|#####################################8          | 3.89G/4.94G [01:40<00:13, 75.2MB/s]
+
+
+
+
text_encoder/model-00002-of-00003.safete(…):  87%|#########################################5      | 4.32G/4.98G [01:40<00:12, 52.9MB/s]
text_encoder/model-00002-of-00003.safete(…):  87%|#########################################5      | 4.32G/4.98G [01:40<00:12, 52.9MB/s]
+
+
+
+
+
+
text_encoder/model-00001-of-00003.safete(…):  80%|######################################2         | 3.93G/4.94G [01:41<00:15, 63.7MB/s]
text_encoder/model-00001-of-00003.safete(…):  80%|######################################2         | 3.93G/4.94G [01:41<00:15, 63.7MB/s]
+
+
+
+
text_encoder/model-00002-of-00003.safete(…):  88%|##########################################1     | 4.38G/4.98G [01:41<00:11, 52.3MB/s]
text_encoder/model-00002-of-00003.safete(…):  88%|##########################################1     | 4.38G/4.98G [01:41<00:11, 52.3MB/s]
+
+
+
+
+
+
text_encoder/model-00001-of-00003.safete(…):  81%|######################################9         | 4.00G/4.94G [01:41<00:11, 79.8MB/s]
text_encoder/model-00001-of-00003.safete(…):  81%|######################################9         | 4.00G/4.94G [01:41<00:11, 79.8MB/s]
+
+
+
+
text_encoder/model-00002-of-00003.safete(…):  89%|##########################################8     | 4.45G/4.98G [01:41<00:07, 68.8MB/s]
text_encoder/model-00002-of-00003.safete(…):  89%|##########################################8     | 4.45G/4.98G [01:41<00:07, 68.8MB/s]
+
+
+
+
+
+
text_encoder/model-00001-of-00003.safete(…):  82%|#######################################5        | 4.07G/4.94G [01:42<00:08, 97.8MB/s]
text_encoder/model-00001-of-00003.safete(…):  82%|#######################################5        | 4.07G/4.94G [01:42<00:08, 97.8MB/s]
+
+
+
+
text_encoder/model-00002-of-00003.safete(…):  91%|###########################################4    | 4.51G/4.98G [01:42<00:06, 74.2MB/s]
text_encoder/model-00002-of-00003.safete(…):  91%|###########################################4    | 4.51G/4.98G [01:42<00:06, 74.2MB/s]
+
+
+
+
+
+
text_encoder/model-00001-of-00003.safete(…):  85%|#########################################7       | 4.20G/4.94G [01:43<00:06, 111MB/s]
text_encoder/model-00001-of-00003.safete(…):  85%|#########################################7       | 4.20G/4.94G [01:43<00:06, 111MB/s]
+
+
+
+
text_encoder/model-00002-of-00003.safete(…):  92%|############################################1   | 4.58G/4.98G [01:43<00:04, 82.5MB/s]
text_encoder/model-00002-of-00003.safete(…):  92%|############################################1   | 4.58G/4.98G [01:43<00:04, 82.5MB/s]
+
+
+
+
+
+
text_encoder/model-00001-of-00003.safete(…):  86%|##########################################3      | 4.27G/4.94G [01:43<00:04, 136MB/s]
text_encoder/model-00001-of-00003.safete(…):  86%|##########################################3      | 4.27G/4.94G [01:43<00:04, 136MB/s]
+
+
+
+
text_encoder/model-00002-of-00003.safete(…):  96%|###############################################  | 4.78G/4.98G [01:43<00:01, 162MB/s]
text_encoder/model-00002-of-00003.safete(…):  96%|###############################################  | 4.78G/4.98G [01:43<00:01, 162MB/s]
+
+
+
+
+
+
text_encoder/model-00001-of-00003.safete(…):  89%|###########################################6     | 4.40G/4.94G [01:43<00:02, 203MB/s]
text_encoder/model-00001-of-00003.safete(…):  89%|###########################################6     | 4.40G/4.94G [01:43<00:02, 203MB/s]
+
+
+
+
+
+
text_encoder/model-00001-of-00003.safete(…):  91%|############################################3    | 4.47G/4.94G [01:44<00:03, 151MB/s]
text_encoder/model-00001-of-00003.safete(…):  91%|############################################3    | 4.47G/4.94G [01:44<00:03, 151MB/s]
+
+
+
+
text_encoder/model-00002-of-00003.safete(…):  99%|################################################3| 4.92G/4.98G [01:44<00:00, 152MB/s]
text_encoder/model-00002-of-00003.safete(…):  99%|################################################3| 4.92G/4.98G [01:44<00:00, 152MB/s]
+
+
+
+
+
+
text_encoder/model-00001-of-00003.safete(…):  92%|#############################################    | 4.53G/4.94G [01:44<00:02, 169MB/s]
text_encoder/model-00001-of-00003.safete(…):  92%|#############################################    | 4.53G/4.94G [01:44<00:02, 169MB/s]
+
+
+
+
text_encoder/model-00002-of-00003.safete(…): 100%|#################################################| 4.98G/4.98G [01:44<00:00, 169MB/s]
text_encoder/model-00002-of-00003.safete(…): 100%|#################################################| 4.98G/4.98G [01:44<00:00, 169MB/s]
text_encoder/model-00002-of-00003.safete(…): 100%|################################################| 4.98G/4.98G [01:44<00:00, 47.6MB/s]
text_encoder/model-00002-of-00003.safete(…): 100%|################################################| 4.98G/4.98G [01:44<00:00, 47.6MB/s]
+
+
+
+
+
+
+
+
text_encoder/model-00001-of-00003.safete(…):  93%|#############################################6   | 4.60G/4.94G [01:44<00:01, 174MB/s]
text_encoder/model-00001-of-00003.safete(…):  93%|#############################################6   | 4.60G/4.94G [01:44<00:01, 174MB/s]
+
+
+
+
+
+
text_encoder/model-00001-of-00003.safete(…):  96%|###############################################  | 4.74G/4.94G [01:45<00:00, 209MB/s]
text_encoder/model-00001-of-00003.safete(…):  96%|###############################################  | 4.74G/4.94G [01:45<00:00, 209MB/s]
+
+
+
+
+
+
text_encoder/model-00001-of-00003.safete(…):  99%|################################################3| 4.87G/4.94G [01:46<00:00, 187MB/s]
text_encoder/model-00001-of-00003.safete(…):  99%|################################################3| 4.87G/4.94G [01:46<00:00, 187MB/s]
+
+
+
+
+
+
text_encoder/model-00001-of-00003.safete(…): 100%|#################################################| 4.94G/4.94G [01:46<00:00, 165MB/s]
text_encoder/model-00001-of-00003.safete(…): 100%|#################################################| 4.94G/4.94G [01:46<00:00, 165MB/s]
text_encoder/model-00001-of-00003.safete(…): 100%|################################################| 4.94G/4.94G [01:46<00:00, 46.2MB/s]
text_encoder/model-00001-of-00003.safete(…): 100%|################################################| 4.94G/4.94G [01:46<00:00, 46.2MB/s]
+
+
Fetching 3 files:  33%|###########################                                                      | 1/3 [01:47<03:35, 107.75s/it]
Fetching 3 files:  33%|###########################                                                      | 1/3 [01:47<03:35, 107.75s/it]
Fetching 3 files: 100%|##################################################################################| 3/3 [01:47<00:00, 35.92s/it]
Fetching 3 files: 100%|##################################################################################| 3/3 [01:47<00:00, 35.92s/it]
+
+
Loading checkpoint shards:   0%|                                                                                 | 0/3 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|                                                                                 | 0/3 [00:00<?, ?it/s]
Loading checkpoint shards: 100%|#########################################################################| 3/3 [00:00<00:00, 50.52it/s]
Loading checkpoint shards: 100%|#########################################################################| 3/3 [00:00<00:00, 50.52it/s]
+
+Quantizing UMT5EncoderModel
+Moving transformer back to GPU
+Loading VAE
+
config.json:   0%|                                                                                           | 0.00/953 [00:00<?, ?B/s]
config.json:   0%|                                                                                           | 0.00/953 [00:00<?, ?B/s]
config.json: 100%|####################################################################################| 953/953 [00:00<00:00, 5.29MB/s]
config.json: 100%|####################################################################################| 953/953 [00:00<00:00, 5.29MB/s]
+
+
diffusion_pytorch_model.safetensors:   0%|                                                                  | 0.00/254M [00:00<?, ?B/s]
diffusion_pytorch_model.safetensors:   0%|                                                                  | 0.00/254M [00:00<?, ?B/s]
diffusion_pytorch_model.safetensors:   0%|                                                        | 131k/254M [00:03<2:05:28, 33.7kB/s]
diffusion_pytorch_model.safetensors:   0%|                                                        | 131k/254M [00:03<2:05:28, 33.7kB/s]
diffusion_pytorch_model.safetensors:   0%|2                                                         | 1.25M/254M [00:04<09:50, 427kB/s]
diffusion_pytorch_model.safetensors:   0%|2                                                         | 1.25M/254M [00:04<09:50, 427kB/s]
diffusion_pytorch_model.safetensors:   2%|9                                                        | 4.20M/254M [00:04<02:24, 1.73MB/s]
diffusion_pytorch_model.safetensors:   2%|9                                                        | 4.20M/254M [00:04<02:24, 1.73MB/s]
diffusion_pytorch_model.safetensors:   2%|#2                                                       | 5.38M/254M [00:04<02:02, 2.03MB/s]
diffusion_pytorch_model.safetensors:   2%|#2                                                       | 5.38M/254M [00:04<02:02, 2.03MB/s]
diffusion_pytorch_model.safetensors:   3%|#8                                                       | 8.36M/254M [00:04<01:04, 3.79MB/s]
diffusion_pytorch_model.safetensors:   3%|#8                                                       | 8.36M/254M [00:04<01:04, 3.79MB/s]
diffusion_pytorch_model.safetensors:   5%|##6                                                      | 12.0M/254M [00:04<00:38, 6.27MB/s]
diffusion_pytorch_model.safetensors:   5%|##6                                                      | 12.0M/254M [00:04<00:38, 6.27MB/s]
diffusion_pytorch_model.safetensors:   6%|###2                                                     | 14.5M/254M [00:05<00:30, 7.90MB/s]
diffusion_pytorch_model.safetensors:   6%|###2                                                     | 14.5M/254M [00:05<00:30, 7.90MB/s]
diffusion_pytorch_model.safetensors:   8%|####3                                                    | 19.3M/254M [00:05<00:21, 10.8MB/s]
diffusion_pytorch_model.safetensors:   8%|####3                                                    | 19.3M/254M [00:05<00:21, 10.8MB/s]
diffusion_pytorch_model.safetensors:   9%|####9                                                    | 22.1M/254M [00:05<00:22, 10.5MB/s]
diffusion_pytorch_model.safetensors:   9%|####9                                                    | 22.1M/254M [00:05<00:22, 10.5MB/s]
diffusion_pytorch_model.safetensors:  10%|#####5                                                   | 24.7M/254M [00:05<00:20, 11.0MB/s]
diffusion_pytorch_model.safetensors:  10%|#####5                                                   | 24.7M/254M [00:05<00:20, 11.0MB/s]
diffusion_pytorch_model.safetensors:  26%|###############                                          | 67.0M/254M [00:05<00:02, 64.4MB/s]
diffusion_pytorch_model.safetensors:  26%|###############                                          | 67.0M/254M [00:05<00:02, 64.4MB/s]
diffusion_pytorch_model.safetensors:  31%|#################6                                       | 78.6M/254M [00:06<00:04, 43.1MB/s]
diffusion_pytorch_model.safetensors:  31%|#################6                                       | 78.6M/254M [00:06<00:04, 43.1MB/s]
diffusion_pytorch_model.safetensors:  34%|###################4                                     | 86.4M/254M [00:06<00:03, 42.8MB/s]
diffusion_pytorch_model.safetensors:  34%|###################4                                     | 86.4M/254M [00:06<00:03, 42.8MB/s]
diffusion_pytorch_model.safetensors:  42%|########################2                                 | 106M/254M [00:06<00:02, 58.9MB/s]
diffusion_pytorch_model.safetensors:  42%|########################2                                 | 106M/254M [00:06<00:02, 58.9MB/s]
diffusion_pytorch_model.safetensors:  51%|#############################5                            | 129M/254M [00:07<00:02, 58.4MB/s]
diffusion_pytorch_model.safetensors:  51%|#############################5                            | 129M/254M [00:07<00:02, 58.4MB/s]
diffusion_pytorch_model.safetensors:  66%|#######################################2                   | 169M/254M [00:07<00:00, 101MB/s]
diffusion_pytorch_model.safetensors:  66%|#######################################2                   | 169M/254M [00:07<00:00, 101MB/s]
diffusion_pytorch_model.safetensors:  79%|##############################################5            | 200M/254M [00:07<00:00, 130MB/s]
diffusion_pytorch_model.safetensors:  79%|##############################################5            | 200M/254M [00:07<00:00, 130MB/s]
diffusion_pytorch_model.safetensors:  88%|###################################################7       | 223M/254M [00:07<00:00, 141MB/s]
diffusion_pytorch_model.safetensors:  88%|###################################################7       | 223M/254M [00:07<00:00, 141MB/s]
diffusion_pytorch_model.safetensors:  97%|########################################################2 | 246M/254M [00:08<00:00, 69.5MB/s]
diffusion_pytorch_model.safetensors:  97%|########################################################2 | 246M/254M [00:08<00:00, 69.5MB/s]
diffusion_pytorch_model.safetensors: 100%|##########################################################| 254M/254M [00:08<00:00, 30.1MB/s]
diffusion_pytorch_model.safetensors: 100%|##########################################################| 254M/254M [00:08<00:00, 30.1MB/s]
+
+Making pipe
+Preparing Model
+create LoRA network. base dim (rank): 32, alpha: 32
+neuron dropout: p=None, rank dropout: p=None, module dropout: p=None
+apply LoRA to Conv2d with kernel size (3,3). dim (rank): 16, alpha: 16
+create LoRA for Text Encoder: 0 modules.
+create LoRA for U-Net: 800 modules.
+enable LoRA for U-Net
+Dataset: /home/featurize/ai-toolkit/datasets/Xiang_Handsome_Flux_SRPO_Keye_Captioned
+  -  Preprocessing image dimensions
+
  0%|                                                                                                          | 0/193 [00:00<?, ?it/s]
  0%|                                                                                                          | 0/193 [00:00<?, ?it/s]
  1%|5                                                                                                 | 1/193 [00:00<00:47,  4.01it/s]
  1%|5                                                                                                 | 1/193 [00:00<00:47,  4.01it/s]
  2%|##                                                                                                | 4/193 [00:00<00:14, 13.30it/s]
  2%|##                                                                                                | 4/193 [00:00<00:14, 13.30it/s]
  4%|###5                                                                                              | 7/193 [00:00<00:09, 18.94it/s]
  4%|###5                                                                                              | 7/193 [00:00<00:09, 18.94it/s]
  5%|#####                                                                                            | 10/193 [00:00<00:08, 22.45it/s]
  5%|#####                                                                                            | 10/193 [00:00<00:08, 22.45it/s]
  7%|#######                                                                                          | 14/193 [00:00<00:07, 25.23it/s]
  7%|#######                                                                                          | 14/193 [00:00<00:07, 25.23it/s]
  9%|########5                                                                                        | 17/193 [00:00<00:06, 26.45it/s]
  9%|########5                                                                                        | 17/193 [00:00<00:06, 26.45it/s]
 11%|##########5                                                                                      | 21/193 [00:00<00:06, 28.22it/s]
 11%|##########5                                                                                      | 21/193 [00:00<00:06, 28.22it/s]
 13%|############5                                                                                    | 25/193 [00:01<00:05, 29.43it/s]
 13%|############5                                                                                    | 25/193 [00:01<00:05, 29.43it/s]
 15%|##############5                                                                                  | 29/193 [00:01<00:05, 29.81it/s]
 15%|##############5                                                                                  | 29/193 [00:01<00:05, 29.81it/s]
 17%|################5                                                                                | 33/193 [00:01<00:05, 30.24it/s]
 17%|################5                                                                                | 33/193 [00:01<00:05, 30.24it/s]
 19%|##################5                                                                              | 37/193 [00:01<00:05, 30.14it/s]
 19%|##################5                                                                              | 37/193 [00:01<00:05, 30.14it/s]
 21%|####################6                                                                            | 41/193 [00:01<00:05, 30.23it/s]
 21%|####################6                                                                            | 41/193 [00:01<00:05, 30.23it/s]
 23%|######################6                                                                          | 45/193 [00:01<00:04, 30.32it/s]
 23%|######################6                                                                          | 45/193 [00:01<00:04, 30.32it/s]
 25%|########################6                                                                        | 49/193 [00:01<00:04, 30.38it/s]
 25%|########################6                                                                        | 49/193 [00:01<00:04, 30.38it/s]
 27%|##########################6                                                                      | 53/193 [00:01<00:04, 30.25it/s]
 27%|##########################6                                                                      | 53/193 [00:01<00:04, 30.25it/s]
 30%|############################6                                                                    | 57/193 [00:02<00:04, 29.97it/s]
 30%|############################6                                                                    | 57/193 [00:02<00:04, 29.97it/s]
 32%|##############################6                                                                  | 61/193 [00:02<00:04, 30.28it/s]
 32%|##############################6                                                                  | 61/193 [00:02<00:04, 30.28it/s]
 34%|################################6                                                                | 65/193 [00:02<00:04, 30.45it/s]
 34%|################################6                                                                | 65/193 [00:02<00:04, 30.45it/s]
 36%|##################################6                                                              | 69/193 [00:02<00:03, 31.31it/s]
 36%|##################################6                                                              | 69/193 [00:02<00:03, 31.31it/s]
 38%|####################################6                                                            | 73/193 [00:02<00:03, 31.36it/s]
 38%|####################################6                                                            | 73/193 [00:02<00:03, 31.36it/s]
 40%|######################################6                                                          | 77/193 [00:02<00:03, 31.18it/s]
 40%|######################################6                                                          | 77/193 [00:02<00:03, 31.18it/s]
 42%|########################################7                                                        | 81/193 [00:02<00:03, 30.44it/s]
 42%|########################################7                                                        | 81/193 [00:02<00:03, 30.44it/s]
 44%|##########################################7                                                      | 85/193 [00:03<00:03, 30.10it/s]
 44%|##########################################7                                                      | 85/193 [00:03<00:03, 30.10it/s]
 46%|############################################7                                                    | 89/193 [00:03<00:03, 30.26it/s]
 46%|############################################7                                                    | 89/193 [00:03<00:03, 30.26it/s]
 48%|##############################################7                                                  | 93/193 [00:03<00:03, 30.38it/s]
 48%|##############################################7                                                  | 93/193 [00:03<00:03, 30.38it/s]
 50%|################################################7                                                | 97/193 [00:03<00:03, 31.05it/s]
 50%|################################################7                                                | 97/193 [00:03<00:03, 31.05it/s]
 52%|##################################################2                                             | 101/193 [00:03<00:03, 30.62it/s]
 52%|##################################################2                                             | 101/193 [00:03<00:03, 30.62it/s]
 54%|####################################################2                                           | 105/193 [00:03<00:02, 30.27it/s]
 54%|####################################################2                                           | 105/193 [00:03<00:02, 30.27it/s]
 56%|######################################################2                                         | 109/193 [00:03<00:02, 30.80it/s]
 56%|######################################################2                                         | 109/193 [00:03<00:02, 30.80it/s]
 59%|########################################################2                                       | 113/193 [00:03<00:02, 30.71it/s]
 59%|########################################################2                                       | 113/193 [00:03<00:02, 30.71it/s]
 61%|##########################################################1                                     | 117/193 [00:04<00:02, 30.88it/s]
 61%|##########################################################1                                     | 117/193 [00:04<00:02, 30.88it/s]
 63%|############################################################1                                   | 121/193 [00:04<00:02, 30.70it/s]
 63%|############################################################1                                   | 121/193 [00:04<00:02, 30.70it/s]
 65%|##############################################################1                                 | 125/193 [00:04<00:02, 30.57it/s]
 65%|##############################################################1                                 | 125/193 [00:04<00:02, 30.57it/s]
 67%|################################################################1                               | 129/193 [00:04<00:02, 30.66it/s]
 67%|################################################################1                               | 129/193 [00:04<00:02, 30.66it/s]
 69%|##################################################################1                             | 133/193 [00:04<00:01, 30.99it/s]
 69%|##################################################################1                             | 133/193 [00:04<00:01, 30.99it/s]
 71%|####################################################################1                           | 137/193 [00:04<00:01, 30.71it/s]
 71%|####################################################################1                           | 137/193 [00:04<00:01, 30.71it/s]
 73%|######################################################################1                         | 141/193 [00:04<00:01, 30.30it/s]
 73%|######################################################################1                         | 141/193 [00:04<00:01, 30.30it/s]
 75%|########################################################################1                       | 145/193 [00:04<00:01, 30.64it/s]
 75%|########################################################################1                       | 145/193 [00:04<00:01, 30.64it/s]
 77%|##########################################################################1                     | 149/193 [00:05<00:01, 30.73it/s]
 77%|##########################################################################1                     | 149/193 [00:05<00:01, 30.73it/s]
 79%|############################################################################1                   | 153/193 [00:05<00:01, 30.58it/s]
 79%|############################################################################1                   | 153/193 [00:05<00:01, 30.58it/s]
 81%|##############################################################################                  | 157/193 [00:05<00:01, 30.46it/s]
 81%|##############################################################################                  | 157/193 [00:05<00:01, 30.46it/s]
 83%|################################################################################                | 161/193 [00:05<00:01, 30.39it/s]
 83%|################################################################################                | 161/193 [00:05<00:01, 30.39it/s]
 85%|##################################################################################              | 165/193 [00:05<00:00, 30.33it/s]
 85%|##################################################################################              | 165/193 [00:05<00:00, 30.33it/s]
 88%|####################################################################################            | 169/193 [00:05<00:00, 30.48it/s]
 88%|####################################################################################            | 169/193 [00:05<00:00, 30.48it/s]
 90%|######################################################################################          | 173/193 [00:05<00:00, 30.47it/s]
 90%|######################################################################################          | 173/193 [00:05<00:00, 30.47it/s]
 92%|########################################################################################        | 177/193 [00:05<00:00, 31.56it/s]
 92%|########################################################################################        | 177/193 [00:05<00:00, 31.56it/s]
 94%|##########################################################################################      | 181/193 [00:06<00:00, 31.08it/s]
 94%|##########################################################################################      | 181/193 [00:06<00:00, 31.08it/s]
 96%|############################################################################################    | 185/193 [00:06<00:00, 24.21it/s]
 96%|############################################################################################    | 185/193 [00:06<00:00, 24.21it/s]
 97%|#############################################################################################5  | 188/193 [00:06<00:00, 21.92it/s]
 97%|#############################################################################################5  | 188/193 [00:06<00:00, 21.92it/s]
 99%|############################################################################################### | 191/193 [00:06<00:00, 22.83it/s]
 99%|############################################################################################### | 191/193 [00:06<00:00, 22.83it/s]
100%|################################################################################################| 193/193 [00:06<00:00, 28.62it/s]
100%|################################################################################################| 193/193 [00:06<00:00, 28.62it/s]
+
+  -  Found 193 images
+Bucket sizes for /home/featurize/ai-toolkit/datasets/Xiang_Handsome_Flux_SRPO_Keye_Captioned:
+512x512: 193 files
+1 buckets made
+Caching text_embeddings for /home/featurize/ai-toolkit/datasets/Xiang_Handsome_Flux_SRPO_Keye_Captioned
+ - Saving text embeddings to disk
+
Caching text embeddings to disk:   0%|                                                                         | 0/193 [00:00<?, ?it/s]
Caching text embeddings to disk:   0%|                                                                         | 0/193 [00:00<?, ?it/s]
Caching text embeddings to disk:   1%|3                                                                | 1/193 [00:04<13:40,  4.27s/it]
Caching text embeddings to disk:   1%|3                                                                | 1/193 [00:04<13:40,  4.27s/it]
Caching text embeddings to disk:   1%|6                                                                | 2/193 [00:04<06:17,  1.98s/it]
Caching text embeddings to disk:   1%|6                                                                | 2/193 [00:04<06:17,  1.98s/it]
Caching text embeddings to disk:   2%|#                                                                | 3/193 [00:05<03:56,  1.24s/it]
Caching text embeddings to disk:   2%|#                                                                | 3/193 [00:05<03:56,  1.24s/it]
Caching text embeddings to disk:   2%|#3                                                               | 4/193 [00:05<02:55,  1.07it/s]
Caching text embeddings to disk:   2%|#3                                                               | 4/193 [00:05<02:55,  1.07it/s]
Caching text embeddings to disk:   3%|#6                                                               | 5/193 [00:05<02:16,  1.37it/s]
Caching text embeddings to disk:   3%|#6                                                               | 5/193 [00:05<02:16,  1.37it/s]
Caching text embeddings to disk:   3%|##                                                               | 6/193 [00:06<01:53,  1.65it/s]
Caching text embeddings to disk:   3%|##                                                               | 6/193 [00:06<01:53,  1.65it/s]
Caching text embeddings to disk:   4%|##3                                                              | 7/193 [00:06<01:38,  1.89it/s]
Caching text embeddings to disk:   4%|##3                                                              | 7/193 [00:06<01:38,  1.89it/s]
Caching text embeddings to disk:   4%|##6                                                              | 8/193 [00:06<01:28,  2.09it/s]
Caching text embeddings to disk:   4%|##6                                                              | 8/193 [00:06<01:28,  2.09it/s]
Caching text embeddings to disk:   5%|###                                                              | 9/193 [00:07<01:21,  2.25it/s]
Caching text embeddings to disk:   5%|###                                                              | 9/193 [00:07<01:21,  2.25it/s]
Caching text embeddings to disk:   5%|###3                                                            | 10/193 [00:07<01:17,  2.38it/s]
Caching text embeddings to disk:   5%|###3                                                            | 10/193 [00:07<01:17,  2.38it/s]
Caching text embeddings to disk:   6%|###6                                                            | 11/193 [00:08<01:13,  2.47it/s]
Caching text embeddings to disk:   6%|###6                                                            | 11/193 [00:08<01:13,  2.47it/s]
Caching text embeddings to disk:   6%|###9                                                            | 12/193 [00:08<01:11,  2.53it/s]
Caching text embeddings to disk:   6%|###9                                                            | 12/193 [00:08<01:11,  2.53it/s]
Caching text embeddings to disk:   7%|####3                                                           | 13/193 [00:08<01:09,  2.58it/s]
Caching text embeddings to disk:   7%|####3                                                           | 13/193 [00:08<01:09,  2.58it/s]
Caching text embeddings to disk:   7%|####6                                                           | 14/193 [00:09<01:08,  2.60it/s]
Caching text embeddings to disk:   7%|####6                                                           | 14/193 [00:09<01:08,  2.60it/s]
Caching text embeddings to disk:   8%|####9                                                           | 15/193 [00:09<01:07,  2.63it/s]
Caching text embeddings to disk:   8%|####9                                                           | 15/193 [00:09<01:07,  2.63it/s]
Caching text embeddings to disk:   8%|#####3                                                          | 16/193 [00:09<01:06,  2.65it/s]
Caching text embeddings to disk:   8%|#####3                                                          | 16/193 [00:09<01:06,  2.65it/s]
Caching text embeddings to disk:   9%|#####6                                                          | 17/193 [00:10<01:06,  2.66it/s]
Caching text embeddings to disk:   9%|#####6                                                          | 17/193 [00:10<01:06,  2.66it/s]
Caching text embeddings to disk:   9%|#####9                                                          | 18/193 [00:10<01:07,  2.59it/s]
Caching text embeddings to disk:   9%|#####9                                                          | 18/193 [00:10<01:07,  2.59it/s]
Caching text embeddings to disk:  10%|######3                                                         | 19/193 [00:11<01:06,  2.62it/s]
Caching text embeddings to disk:  10%|######3                                                         | 19/193 [00:11<01:06,  2.62it/s]
Caching text embeddings to disk:  10%|######6                                                         | 20/193 [00:11<01:05,  2.63it/s]
Caching text embeddings to disk:  10%|######6                                                         | 20/193 [00:11<01:05,  2.63it/s]
Caching text embeddings to disk:  11%|######9                                                         | 21/193 [00:11<01:05,  2.65it/s]
Caching text embeddings to disk:  11%|######9                                                         | 21/193 [00:11<01:05,  2.65it/s]
Caching text embeddings to disk:  11%|#######2                                                        | 22/193 [00:12<01:04,  2.66it/s]
Caching text embeddings to disk:  11%|#######2                                                        | 22/193 [00:12<01:04,  2.66it/s]
Caching text embeddings to disk:  12%|#######6                                                        | 23/193 [00:12<01:03,  2.67it/s]
Caching text embeddings to disk:  12%|#######6                                                        | 23/193 [00:12<01:03,  2.67it/s]
Caching text embeddings to disk:  12%|#######9                                                        | 24/193 [00:12<01:06,  2.55it/s]
Caching text embeddings to disk:  12%|#######9                                                        | 24/193 [00:12<01:06,  2.55it/s]
Caching text embeddings to disk:  13%|########2                                                       | 25/193 [00:13<01:04,  2.59it/s]
Caching text embeddings to disk:  13%|########2                                                       | 25/193 [00:13<01:04,  2.59it/s]
Caching text embeddings to disk:  13%|########6                                                       | 26/193 [00:13<01:03,  2.63it/s]
Caching text embeddings to disk:  13%|########6                                                       | 26/193 [00:13<01:03,  2.63it/s]
Caching text embeddings to disk:  14%|########9                                                       | 27/193 [00:14<01:02,  2.64it/s]
Caching text embeddings to disk:  14%|########9                                                       | 27/193 [00:14<01:02,  2.64it/s]
Caching text embeddings to disk:  15%|#########2                                                      | 28/193 [00:14<01:02,  2.66it/s]
Caching text embeddings to disk:  15%|#########2                                                      | 28/193 [00:14<01:02,  2.66it/s]
Caching text embeddings to disk:  15%|#########6                                                      | 29/193 [00:14<01:01,  2.67it/s]
Caching text embeddings to disk:  15%|#########6                                                      | 29/193 [00:14<01:01,  2.67it/s]
Caching text embeddings to disk:  16%|#########9                                                      | 30/193 [00:15<01:01,  2.67it/s]
Caching text embeddings to disk:  16%|#########9                                                      | 30/193 [00:15<01:01,  2.67it/s]
Caching text embeddings to disk:  16%|##########2                                                     | 31/193 [00:15<01:00,  2.68it/s]
Caching text embeddings to disk:  16%|##########2                                                     | 31/193 [00:15<01:00,  2.68it/s]
Caching text embeddings to disk:  17%|##########6                                                     | 32/193 [00:15<01:00,  2.68it/s]
Caching text embeddings to disk:  17%|##########6                                                     | 32/193 [00:15<01:00,  2.68it/s]
Caching text embeddings to disk:  17%|##########9                                                     | 33/193 [00:16<01:00,  2.64it/s]
Caching text embeddings to disk:  17%|##########9                                                     | 33/193 [00:16<01:00,  2.64it/s]
Caching text embeddings to disk:  18%|###########2                                                    | 34/193 [00:16<00:59,  2.66it/s]
Caching text embeddings to disk:  18%|###########2                                                    | 34/193 [00:16<00:59,  2.66it/s]
Caching text embeddings to disk:  18%|###########6                                                    | 35/193 [00:17<00:59,  2.67it/s]
Caching text embeddings to disk:  18%|###########6                                                    | 35/193 [00:17<00:59,  2.67it/s]
Caching text embeddings to disk:  19%|###########9                                                    | 36/193 [00:17<00:58,  2.68it/s]
Caching text embeddings to disk:  19%|###########9                                                    | 36/193 [00:17<00:58,  2.68it/s]
Caching text embeddings to disk:  19%|############2                                                   | 37/193 [00:17<00:57,  2.69it/s]
Caching text embeddings to disk:  19%|############2                                                   | 37/193 [00:17<00:57,  2.69it/s]
Caching text embeddings to disk:  20%|############6                                                   | 38/193 [00:18<00:57,  2.69it/s]
Caching text embeddings to disk:  20%|############6                                                   | 38/193 [00:18<00:57,  2.69it/s]
Caching text embeddings to disk:  20%|############9                                                   | 39/193 [00:18<00:57,  2.69it/s]
Caching text embeddings to disk:  20%|############9                                                   | 39/193 [00:18<00:57,  2.69it/s]
Caching text embeddings to disk:  21%|#############2                                                  | 40/193 [00:18<00:56,  2.70it/s]
Caching text embeddings to disk:  21%|#############2                                                  | 40/193 [00:18<00:56,  2.70it/s]
Caching text embeddings to disk:  21%|#############5                                                  | 41/193 [00:19<00:56,  2.70it/s]
Caching text embeddings to disk:  21%|#############5                                                  | 41/193 [00:19<00:56,  2.70it/s]
Caching text embeddings to disk:  22%|#############9                                                  | 42/193 [00:19<00:55,  2.70it/s]
Caching text embeddings to disk:  22%|#############9                                                  | 42/193 [00:19<00:55,  2.70it/s]
Caching text embeddings to disk:  22%|##############2                                                 | 43/193 [00:20<00:55,  2.70it/s]
Caching text embeddings to disk:  22%|##############2                                                 | 43/193 [00:20<00:55,  2.70it/s]
Caching text embeddings to disk:  23%|##############5                                                 | 44/193 [00:20<00:55,  2.70it/s]
Caching text embeddings to disk:  23%|##############5                                                 | 44/193 [00:20<00:55,  2.70it/s]
Caching text embeddings to disk:  23%|##############9                                                 | 45/193 [00:20<00:54,  2.69it/s]
Caching text embeddings to disk:  23%|##############9                                                 | 45/193 [00:20<00:54,  2.69it/s]
Caching text embeddings to disk:  24%|###############2                                                | 46/193 [00:21<00:54,  2.69it/s]
Caching text embeddings to disk:  24%|###############2                                                | 46/193 [00:21<00:54,  2.69it/s]
Caching text embeddings to disk:  24%|###############5                                                | 47/193 [00:21<00:55,  2.65it/s]
Caching text embeddings to disk:  24%|###############5                                                | 47/193 [00:21<00:55,  2.65it/s]
Caching text embeddings to disk:  25%|###############9                                                | 48/193 [00:21<00:54,  2.66it/s]
Caching text embeddings to disk:  25%|###############9                                                | 48/193 [00:21<00:54,  2.66it/s]
Caching text embeddings to disk:  25%|################2                                               | 49/193 [00:22<00:53,  2.67it/s]
Caching text embeddings to disk:  25%|################2                                               | 49/193 [00:22<00:53,  2.67it/s]
Caching text embeddings to disk:  26%|################5                                               | 50/193 [00:22<00:53,  2.68it/s]
Caching text embeddings to disk:  26%|################5                                               | 50/193 [00:22<00:53,  2.68it/s]
Caching text embeddings to disk:  26%|################9                                               | 51/193 [00:23<00:52,  2.68it/s]
Caching text embeddings to disk:  26%|################9                                               | 51/193 [00:23<00:52,  2.68it/s]
Caching text embeddings to disk:  27%|#################2                                              | 52/193 [00:23<00:52,  2.68it/s]
Caching text embeddings to disk:  27%|#################2                                              | 52/193 [00:23<00:52,  2.68it/s]
Caching text embeddings to disk:  27%|#################5                                              | 53/193 [00:23<00:52,  2.69it/s]
Caching text embeddings to disk:  27%|#################5                                              | 53/193 [00:23<00:52,  2.69it/s]
Caching text embeddings to disk:  28%|#################9                                              | 54/193 [00:24<00:51,  2.69it/s]
Caching text embeddings to disk:  28%|#################9                                              | 54/193 [00:24<00:51,  2.69it/s]
Caching text embeddings to disk:  28%|##################2                                             | 55/193 [00:24<00:51,  2.69it/s]
Caching text embeddings to disk:  28%|##################2                                             | 55/193 [00:24<00:51,  2.69it/s]
Caching text embeddings to disk:  29%|##################5                                             | 56/193 [00:24<00:50,  2.69it/s]
Caching text embeddings to disk:  29%|##################5                                             | 56/193 [00:24<00:50,  2.69it/s]
Caching text embeddings to disk:  30%|##################9                                             | 57/193 [00:25<00:50,  2.70it/s]
Caching text embeddings to disk:  30%|##################9                                             | 57/193 [00:25<00:50,  2.70it/s]
Caching text embeddings to disk:  30%|###################2                                            | 58/193 [00:25<00:50,  2.69it/s]
Caching text embeddings to disk:  30%|###################2                                            | 58/193 [00:25<00:50,  2.69it/s]
Caching text embeddings to disk:  31%|###################5                                            | 59/193 [00:26<00:49,  2.69it/s]
Caching text embeddings to disk:  31%|###################5                                            | 59/193 [00:26<00:49,  2.69it/s]
Caching text embeddings to disk:  31%|###################8                                            | 60/193 [00:26<00:49,  2.69it/s]
Caching text embeddings to disk:  31%|###################8                                            | 60/193 [00:26<00:49,  2.69it/s]
Caching text embeddings to disk:  32%|####################2                                           | 61/193 [00:26<00:49,  2.69it/s]
Caching text embeddings to disk:  32%|####################2                                           | 61/193 [00:26<00:49,  2.69it/s]
Caching text embeddings to disk:  32%|####################5                                           | 62/193 [00:27<00:48,  2.69it/s]
Caching text embeddings to disk:  32%|####################5                                           | 62/193 [00:27<00:48,  2.69it/s]
Caching text embeddings to disk:  33%|####################8                                           | 63/193 [00:27<00:48,  2.69it/s]
Caching text embeddings to disk:  33%|####################8                                           | 63/193 [00:27<00:48,  2.69it/s]
Caching text embeddings to disk:  33%|#####################2                                          | 64/193 [00:27<00:47,  2.69it/s]
Caching text embeddings to disk:  33%|#####################2                                          | 64/193 [00:27<00:47,  2.69it/s]
Caching text embeddings to disk:  34%|#####################5                                          | 65/193 [00:28<00:47,  2.69it/s]
Caching text embeddings to disk:  34%|#####################5                                          | 65/193 [00:28<00:47,  2.69it/s]
Caching text embeddings to disk:  34%|#####################8                                          | 66/193 [00:28<00:47,  2.69it/s]
Caching text embeddings to disk:  34%|#####################8                                          | 66/193 [00:28<00:47,  2.69it/s]
Caching text embeddings to disk:  35%|######################2                                         | 67/193 [00:28<00:46,  2.69it/s]
Caching text embeddings to disk:  35%|######################2                                         | 67/193 [00:28<00:46,  2.69it/s]
Caching text embeddings to disk:  35%|######################5                                         | 68/193 [00:29<00:46,  2.69it/s]
Caching text embeddings to disk:  35%|######################5                                         | 68/193 [00:29<00:46,  2.69it/s]
Caching text embeddings to disk:  36%|######################8                                         | 69/193 [00:29<00:46,  2.69it/s]
Caching text embeddings to disk:  36%|######################8                                         | 69/193 [00:29<00:46,  2.69it/s]
Caching text embeddings to disk:  36%|#######################2                                        | 70/193 [00:30<00:45,  2.69it/s]
Caching text embeddings to disk:  36%|#######################2                                        | 70/193 [00:30<00:45,  2.69it/s]
Caching text embeddings to disk:  37%|#######################5                                        | 71/193 [00:30<00:45,  2.69it/s]
Caching text embeddings to disk:  37%|#######################5                                        | 71/193 [00:30<00:45,  2.69it/s]
Caching text embeddings to disk:  37%|#######################8                                        | 72/193 [00:30<00:44,  2.69it/s]
Caching text embeddings to disk:  37%|#######################8                                        | 72/193 [00:30<00:44,  2.69it/s]
Caching text embeddings to disk:  38%|########################2                                       | 73/193 [00:31<00:44,  2.69it/s]
Caching text embeddings to disk:  38%|########################2                                       | 73/193 [00:31<00:44,  2.69it/s]
Caching text embeddings to disk:  38%|########################5                                       | 74/193 [00:31<00:44,  2.69it/s]
Caching text embeddings to disk:  38%|########################5                                       | 74/193 [00:31<00:44,  2.69it/s]
Caching text embeddings to disk:  39%|########################8                                       | 75/193 [00:31<00:43,  2.69it/s]
Caching text embeddings to disk:  39%|########################8                                       | 75/193 [00:31<00:43,  2.69it/s]
Caching text embeddings to disk:  39%|#########################2                                      | 76/193 [00:32<00:43,  2.69it/s]
Caching text embeddings to disk:  39%|#########################2                                      | 76/193 [00:32<00:43,  2.69it/s]
Caching text embeddings to disk:  40%|#########################5                                      | 77/193 [00:32<00:43,  2.68it/s]
Caching text embeddings to disk:  40%|#########################5                                      | 77/193 [00:32<00:43,  2.68it/s]
Caching text embeddings to disk:  40%|#########################8                                      | 78/193 [00:33<00:42,  2.68it/s]
Caching text embeddings to disk:  40%|#########################8                                      | 78/193 [00:33<00:42,  2.68it/s]
Caching text embeddings to disk:  41%|##########################1                                     | 79/193 [00:33<00:42,  2.69it/s]
Caching text embeddings to disk:  41%|##########################1                                     | 79/193 [00:33<00:42,  2.69it/s]
Caching text embeddings to disk:  41%|##########################5                                     | 80/193 [00:33<00:42,  2.68it/s]
Caching text embeddings to disk:  41%|##########################5                                     | 80/193 [00:33<00:42,  2.68it/s]
Caching text embeddings to disk:  42%|##########################8                                     | 81/193 [00:34<00:41,  2.68it/s]
Caching text embeddings to disk:  42%|##########################8                                     | 81/193 [00:34<00:41,  2.68it/s]
Caching text embeddings to disk:  42%|###########################1                                    | 82/193 [00:34<00:41,  2.68it/s]
Caching text embeddings to disk:  42%|###########################1                                    | 82/193 [00:34<00:41,  2.68it/s]
Caching text embeddings to disk:  43%|###########################5                                    | 83/193 [00:34<00:40,  2.69it/s]
Caching text embeddings to disk:  43%|###########################5                                    | 83/193 [00:34<00:40,  2.69it/s]
Caching text embeddings to disk:  44%|###########################8                                    | 84/193 [00:35<00:40,  2.69it/s]
Caching text embeddings to disk:  44%|###########################8                                    | 84/193 [00:35<00:40,  2.69it/s]
Caching text embeddings to disk:  44%|############################1                                   | 85/193 [00:35<00:40,  2.69it/s]
Caching text embeddings to disk:  44%|############################1                                   | 85/193 [00:35<00:40,  2.69it/s]
Caching text embeddings to disk:  45%|############################5                                   | 86/193 [00:36<00:39,  2.69it/s]
Caching text embeddings to disk:  45%|############################5                                   | 86/193 [00:36<00:39,  2.69it/s]
Caching text embeddings to disk:  45%|############################8                                   | 87/193 [00:36<00:39,  2.68it/s]
Caching text embeddings to disk:  45%|############################8                                   | 87/193 [00:36<00:39,  2.68it/s]
Caching text embeddings to disk:  46%|#############################1                                  | 88/193 [00:36<00:39,  2.68it/s]
Caching text embeddings to disk:  46%|#############################1                                  | 88/193 [00:36<00:39,  2.68it/s]
Caching text embeddings to disk:  46%|#############################5                                  | 89/193 [00:37<00:38,  2.68it/s]
Caching text embeddings to disk:  46%|#############################5                                  | 89/193 [00:37<00:38,  2.68it/s]
Caching text embeddings to disk:  47%|#############################8                                  | 90/193 [00:37<00:38,  2.68it/s]
Caching text embeddings to disk:  47%|#############################8                                  | 90/193 [00:37<00:38,  2.68it/s]
Caching text embeddings to disk:  47%|##############################1                                 | 91/193 [00:37<00:38,  2.68it/s]
Caching text embeddings to disk:  47%|##############################1                                 | 91/193 [00:37<00:38,  2.68it/s]
Caching text embeddings to disk:  48%|##############################5                                 | 92/193 [00:38<00:37,  2.68it/s]
Caching text embeddings to disk:  48%|##############################5                                 | 92/193 [00:38<00:37,  2.68it/s]
Caching text embeddings to disk:  48%|##############################8                                 | 93/193 [00:38<00:37,  2.68it/s]
Caching text embeddings to disk:  48%|##############################8                                 | 93/193 [00:38<00:37,  2.68it/s]
Caching text embeddings to disk:  49%|###############################1                                | 94/193 [00:39<00:36,  2.69it/s]
Caching text embeddings to disk:  49%|###############################1                                | 94/193 [00:39<00:36,  2.69it/s]
Caching text embeddings to disk:  49%|###############################5                                | 95/193 [00:39<00:36,  2.69it/s]
Caching text embeddings to disk:  49%|###############################5                                | 95/193 [00:39<00:36,  2.69it/s]
Caching text embeddings to disk:  50%|###############################8                                | 96/193 [00:39<00:36,  2.69it/s]
Caching text embeddings to disk:  50%|###############################8                                | 96/193 [00:39<00:36,  2.69it/s]
Caching text embeddings to disk:  50%|################################1                               | 97/193 [00:40<00:35,  2.69it/s]
Caching text embeddings to disk:  50%|################################1                               | 97/193 [00:40<00:35,  2.69it/s]
Caching text embeddings to disk:  51%|################################4                               | 98/193 [00:40<00:35,  2.69it/s]
Caching text embeddings to disk:  51%|################################4                               | 98/193 [00:40<00:35,  2.69it/s]
Caching text embeddings to disk:  51%|################################8                               | 99/193 [00:40<00:34,  2.69it/s]
Caching text embeddings to disk:  51%|################################8                               | 99/193 [00:40<00:34,  2.69it/s]
Caching text embeddings to disk:  52%|################################6                              | 100/193 [00:41<00:34,  2.69it/s]
Caching text embeddings to disk:  52%|################################6                              | 100/193 [00:41<00:34,  2.69it/s]
Caching text embeddings to disk:  52%|################################9                              | 101/193 [00:41<00:34,  2.69it/s]
Caching text embeddings to disk:  52%|################################9                              | 101/193 [00:41<00:34,  2.69it/s]
Caching text embeddings to disk:  53%|#################################2                             | 102/193 [00:42<00:33,  2.69it/s]
Caching text embeddings to disk:  53%|#################################2                             | 102/193 [00:42<00:33,  2.69it/s]
Caching text embeddings to disk:  53%|#################################6                             | 103/193 [00:42<00:33,  2.69it/s]
Caching text embeddings to disk:  53%|#################################6                             | 103/193 [00:42<00:33,  2.69it/s]
Caching text embeddings to disk:  54%|#################################9                             | 104/193 [00:42<00:33,  2.69it/s]
Caching text embeddings to disk:  54%|#################################9                             | 104/193 [00:42<00:33,  2.69it/s]
Caching text embeddings to disk:  54%|##################################2                            | 105/193 [00:43<00:32,  2.69it/s]
Caching text embeddings to disk:  54%|##################################2                            | 105/193 [00:43<00:32,  2.69it/s]
Caching text embeddings to disk:  55%|##################################6                            | 106/193 [00:43<00:32,  2.69it/s]
Caching text embeddings to disk:  55%|##################################6                            | 106/193 [00:43<00:32,  2.69it/s]
Caching text embeddings to disk:  55%|##################################9                            | 107/193 [00:43<00:32,  2.66it/s]
Caching text embeddings to disk:  55%|##################################9                            | 107/193 [00:43<00:32,  2.66it/s]
Caching text embeddings to disk:  56%|###################################2                           | 108/193 [00:44<00:31,  2.67it/s]
Caching text embeddings to disk:  56%|###################################2                           | 108/193 [00:44<00:31,  2.67it/s]
Caching text embeddings to disk:  56%|###################################5                           | 109/193 [00:44<00:31,  2.67it/s]
Caching text embeddings to disk:  56%|###################################5                           | 109/193 [00:44<00:31,  2.67it/s]
Caching text embeddings to disk:  57%|###################################9                           | 110/193 [00:45<00:31,  2.68it/s]
Caching text embeddings to disk:  57%|###################################9                           | 110/193 [00:45<00:31,  2.68it/s]
Caching text embeddings to disk:  58%|####################################2                          | 111/193 [00:45<00:30,  2.68it/s]
Caching text embeddings to disk:  58%|####################################2                          | 111/193 [00:45<00:30,  2.68it/s]
Caching text embeddings to disk:  58%|####################################5                          | 112/193 [00:45<00:30,  2.68it/s]
Caching text embeddings to disk:  58%|####################################5                          | 112/193 [00:45<00:30,  2.68it/s]
Caching text embeddings to disk:  59%|####################################8                          | 113/193 [00:46<00:29,  2.68it/s]
Caching text embeddings to disk:  59%|####################################8                          | 113/193 [00:46<00:29,  2.68it/s]
Caching text embeddings to disk:  59%|#####################################2                         | 114/193 [00:46<00:29,  2.68it/s]
Caching text embeddings to disk:  59%|#####################################2                         | 114/193 [00:46<00:29,  2.68it/s]
Caching text embeddings to disk:  60%|#####################################5                         | 115/193 [00:46<00:29,  2.68it/s]
Caching text embeddings to disk:  60%|#####################################5                         | 115/193 [00:46<00:29,  2.68it/s]
Caching text embeddings to disk:  60%|#####################################8                         | 116/193 [00:47<00:28,  2.68it/s]
Caching text embeddings to disk:  60%|#####################################8                         | 116/193 [00:47<00:28,  2.68it/s]
Caching text embeddings to disk:  61%|######################################1                        | 117/193 [00:47<00:28,  2.68it/s]
Caching text embeddings to disk:  61%|######################################1                        | 117/193 [00:47<00:28,  2.68it/s]
Caching text embeddings to disk:  61%|######################################5                        | 118/193 [00:47<00:27,  2.68it/s]
Caching text embeddings to disk:  61%|######################################5                        | 118/193 [00:47<00:27,  2.68it/s]
Caching text embeddings to disk:  62%|######################################8                        | 119/193 [00:48<00:27,  2.68it/s]
Caching text embeddings to disk:  62%|######################################8                        | 119/193 [00:48<00:27,  2.68it/s]
Caching text embeddings to disk:  62%|#######################################1                       | 120/193 [00:48<00:27,  2.68it/s]
Caching text embeddings to disk:  62%|#######################################1                       | 120/193 [00:48<00:27,  2.68it/s]
Caching text embeddings to disk:  63%|#######################################4                       | 121/193 [00:49<00:27,  2.64it/s]
Caching text embeddings to disk:  63%|#######################################4                       | 121/193 [00:49<00:27,  2.64it/s]
Caching text embeddings to disk:  63%|#######################################8                       | 122/193 [00:49<00:26,  2.64it/s]
Caching text embeddings to disk:  63%|#######################################8                       | 122/193 [00:49<00:26,  2.64it/s]
Caching text embeddings to disk:  64%|########################################1                      | 123/193 [00:49<00:26,  2.65it/s]
Caching text embeddings to disk:  64%|########################################1                      | 123/193 [00:49<00:26,  2.65it/s]
Caching text embeddings to disk:  64%|########################################4                      | 124/193 [00:50<00:25,  2.66it/s]
Caching text embeddings to disk:  64%|########################################4                      | 124/193 [00:50<00:25,  2.66it/s]
Caching text embeddings to disk:  65%|########################################8                      | 125/193 [00:50<00:25,  2.67it/s]
Caching text embeddings to disk:  65%|########################################8                      | 125/193 [00:50<00:25,  2.67it/s]
Caching text embeddings to disk:  65%|#########################################1                     | 126/193 [00:50<00:25,  2.67it/s]
Caching text embeddings to disk:  65%|#########################################1                     | 126/193 [00:50<00:25,  2.67it/s]
Caching text embeddings to disk:  66%|#########################################4                     | 127/193 [00:51<00:24,  2.67it/s]
Caching text embeddings to disk:  66%|#########################################4                     | 127/193 [00:51<00:24,  2.67it/s]
Caching text embeddings to disk:  66%|#########################################7                     | 128/193 [00:51<00:24,  2.67it/s]
Caching text embeddings to disk:  66%|#########################################7                     | 128/193 [00:51<00:24,  2.67it/s]
Caching text embeddings to disk:  67%|##########################################1                    | 129/193 [00:52<00:23,  2.67it/s]
Caching text embeddings to disk:  67%|##########################################1                    | 129/193 [00:52<00:23,  2.67it/s]
Caching text embeddings to disk:  67%|##########################################4                    | 130/193 [00:52<00:23,  2.67it/s]
Caching text embeddings to disk:  67%|##########################################4                    | 130/193 [00:52<00:23,  2.67it/s]
Caching text embeddings to disk:  68%|##########################################7                    | 131/193 [00:52<00:23,  2.67it/s]
Caching text embeddings to disk:  68%|##########################################7                    | 131/193 [00:52<00:23,  2.67it/s]
Caching text embeddings to disk:  68%|###########################################                    | 132/193 [00:53<00:22,  2.67it/s]
Caching text embeddings to disk:  68%|###########################################                    | 132/193 [00:53<00:22,  2.67it/s]
Caching text embeddings to disk:  69%|###########################################4                   | 133/193 [00:53<00:22,  2.67it/s]
Caching text embeddings to disk:  69%|###########################################4                   | 133/193 [00:53<00:22,  2.67it/s]
Caching text embeddings to disk:  69%|###########################################7                   | 134/193 [00:53<00:22,  2.67it/s]
Caching text embeddings to disk:  69%|###########################################7                   | 134/193 [00:53<00:22,  2.67it/s]
Caching text embeddings to disk:  70%|############################################                   | 135/193 [00:54<00:21,  2.67it/s]
Caching text embeddings to disk:  70%|############################################                   | 135/193 [00:54<00:21,  2.67it/s]
Caching text embeddings to disk:  70%|############################################3                  | 136/193 [00:54<00:21,  2.67it/s]
Caching text embeddings to disk:  70%|############################################3                  | 136/193 [00:54<00:21,  2.67it/s]
Caching text embeddings to disk:  71%|############################################7                  | 137/193 [00:55<00:21,  2.67it/s]
Caching text embeddings to disk:  71%|############################################7                  | 137/193 [00:55<00:21,  2.67it/s]
Caching text embeddings to disk:  72%|#############################################                  | 138/193 [00:55<00:20,  2.65it/s]
Caching text embeddings to disk:  72%|#############################################                  | 138/193 [00:55<00:20,  2.65it/s]
Caching text embeddings to disk:  72%|#############################################3                 | 139/193 [00:55<00:20,  2.66it/s]
Caching text embeddings to disk:  72%|#############################################3                 | 139/193 [00:55<00:20,  2.66it/s]
Caching text embeddings to disk:  73%|#############################################6                 | 140/193 [00:56<00:19,  2.66it/s]
Caching text embeddings to disk:  73%|#############################################6                 | 140/193 [00:56<00:19,  2.66it/s]
Caching text embeddings to disk:  73%|##############################################                 | 141/193 [00:56<00:19,  2.67it/s]
Caching text embeddings to disk:  73%|##############################################                 | 141/193 [00:56<00:19,  2.67it/s]
Caching text embeddings to disk:  74%|##############################################3                | 142/193 [00:56<00:19,  2.67it/s]
Caching text embeddings to disk:  74%|##############################################3                | 142/193 [00:56<00:19,  2.67it/s]
Caching text embeddings to disk:  74%|##############################################6                | 143/193 [00:57<00:18,  2.67it/s]
Caching text embeddings to disk:  74%|##############################################6                | 143/193 [00:57<00:18,  2.67it/s]
Caching text embeddings to disk:  75%|###############################################                | 144/193 [00:57<00:18,  2.67it/s]
Caching text embeddings to disk:  75%|###############################################                | 144/193 [00:57<00:18,  2.67it/s]
Caching text embeddings to disk:  75%|###############################################3               | 145/193 [00:58<00:17,  2.67it/s]
Caching text embeddings to disk:  75%|###############################################3               | 145/193 [00:58<00:17,  2.67it/s]
Caching text embeddings to disk:  76%|###############################################6               | 146/193 [00:58<00:17,  2.68it/s]
Caching text embeddings to disk:  76%|###############################################6               | 146/193 [00:58<00:17,  2.68it/s]
Caching text embeddings to disk:  76%|###############################################9               | 147/193 [00:58<00:17,  2.68it/s]
Caching text embeddings to disk:  76%|###############################################9               | 147/193 [00:58<00:17,  2.68it/s]
Caching text embeddings to disk:  77%|################################################3              | 148/193 [00:59<00:16,  2.67it/s]
Caching text embeddings to disk:  77%|################################################3              | 148/193 [00:59<00:16,  2.67it/s]
Caching text embeddings to disk:  77%|################################################6              | 149/193 [00:59<00:16,  2.68it/s]
Caching text embeddings to disk:  77%|################################################6              | 149/193 [00:59<00:16,  2.68it/s]
Caching text embeddings to disk:  78%|################################################9              | 150/193 [00:59<00:16,  2.68it/s]
Caching text embeddings to disk:  78%|################################################9              | 150/193 [00:59<00:16,  2.68it/s]
Caching text embeddings to disk:  78%|#################################################2             | 151/193 [01:00<00:15,  2.68it/s]
Caching text embeddings to disk:  78%|#################################################2             | 151/193 [01:00<00:15,  2.68it/s]
Caching text embeddings to disk:  79%|#################################################6             | 152/193 [01:00<00:15,  2.68it/s]
Caching text embeddings to disk:  79%|#################################################6             | 152/193 [01:00<00:15,  2.68it/s]
Caching text embeddings to disk:  79%|#################################################9             | 153/193 [01:01<00:14,  2.68it/s]
Caching text embeddings to disk:  79%|#################################################9             | 153/193 [01:01<00:14,  2.68it/s]
Caching text embeddings to disk:  80%|##################################################2            | 154/193 [01:01<00:14,  2.68it/s]
Caching text embeddings to disk:  80%|##################################################2            | 154/193 [01:01<00:14,  2.68it/s]
Caching text embeddings to disk:  80%|##################################################5            | 155/193 [01:01<00:14,  2.68it/s]
Caching text embeddings to disk:  80%|##################################################5            | 155/193 [01:01<00:14,  2.68it/s]
Caching text embeddings to disk:  81%|##################################################9            | 156/193 [01:02<00:13,  2.68it/s]
Caching text embeddings to disk:  81%|##################################################9            | 156/193 [01:02<00:13,  2.68it/s]
Caching text embeddings to disk:  81%|###################################################2           | 157/193 [01:02<00:13,  2.68it/s]
Caching text embeddings to disk:  81%|###################################################2           | 157/193 [01:02<00:13,  2.68it/s]
Caching text embeddings to disk:  82%|###################################################5           | 158/193 [01:02<00:13,  2.68it/s]
Caching text embeddings to disk:  82%|###################################################5           | 158/193 [01:02<00:13,  2.68it/s]
Caching text embeddings to disk:  82%|###################################################9           | 159/193 [01:03<00:12,  2.68it/s]
Caching text embeddings to disk:  82%|###################################################9           | 159/193 [01:03<00:12,  2.68it/s]
Caching text embeddings to disk:  83%|####################################################2          | 160/193 [01:03<00:12,  2.68it/s]
Caching text embeddings to disk:  83%|####################################################2          | 160/193 [01:03<00:12,  2.68it/s]
Caching text embeddings to disk:  83%|####################################################5          | 161/193 [01:04<00:11,  2.67it/s]
Caching text embeddings to disk:  83%|####################################################5          | 161/193 [01:04<00:11,  2.67it/s]
Caching text embeddings to disk:  84%|####################################################8          | 162/193 [01:04<00:11,  2.67it/s]
Caching text embeddings to disk:  84%|####################################################8          | 162/193 [01:04<00:11,  2.67it/s]
Caching text embeddings to disk:  84%|#####################################################2         | 163/193 [01:04<00:11,  2.67it/s]
Caching text embeddings to disk:  84%|#####################################################2         | 163/193 [01:04<00:11,  2.67it/s]
Caching text embeddings to disk:  85%|#####################################################5         | 164/193 [01:05<00:10,  2.67it/s]
Caching text embeddings to disk:  85%|#####################################################5         | 164/193 [01:05<00:10,  2.67it/s]
Caching text embeddings to disk:  85%|#####################################################8         | 165/193 [01:05<00:10,  2.67it/s]
Caching text embeddings to disk:  85%|#####################################################8         | 165/193 [01:05<00:10,  2.67it/s]
Caching text embeddings to disk:  86%|######################################################1        | 166/193 [01:05<00:10,  2.68it/s]
Caching text embeddings to disk:  86%|######################################################1        | 166/193 [01:05<00:10,  2.68it/s]
Caching text embeddings to disk:  87%|######################################################5        | 167/193 [01:06<00:09,  2.67it/s]
Caching text embeddings to disk:  87%|######################################################5        | 167/193 [01:06<00:09,  2.67it/s]
Caching text embeddings to disk:  87%|######################################################8        | 168/193 [01:06<00:09,  2.67it/s]
Caching text embeddings to disk:  87%|######################################################8        | 168/193 [01:06<00:09,  2.67it/s]
Caching text embeddings to disk:  88%|#######################################################1       | 169/193 [01:07<00:08,  2.67it/s]
Caching text embeddings to disk:  88%|#######################################################1       | 169/193 [01:07<00:08,  2.67it/s]
Caching text embeddings to disk:  88%|#######################################################4       | 170/193 [01:07<00:08,  2.67it/s]
Caching text embeddings to disk:  88%|#######################################################4       | 170/193 [01:07<00:08,  2.67it/s]
Caching text embeddings to disk:  89%|#######################################################8       | 171/193 [01:07<00:08,  2.67it/s]
Caching text embeddings to disk:  89%|#######################################################8       | 171/193 [01:07<00:08,  2.67it/s]
Caching text embeddings to disk:  89%|########################################################1      | 172/193 [01:08<00:07,  2.67it/s]
Caching text embeddings to disk:  89%|########################################################1      | 172/193 [01:08<00:07,  2.67it/s]
Caching text embeddings to disk:  90%|########################################################4      | 173/193 [01:08<00:07,  2.67it/s]
Caching text embeddings to disk:  90%|########################################################4      | 173/193 [01:08<00:07,  2.67it/s]
Caching text embeddings to disk:  90%|########################################################7      | 174/193 [01:08<00:07,  2.67it/s]
Caching text embeddings to disk:  90%|########################################################7      | 174/193 [01:08<00:07,  2.67it/s]
Caching text embeddings to disk:  91%|#########################################################1     | 175/193 [01:09<00:06,  2.67it/s]
Caching text embeddings to disk:  91%|#########################################################1     | 175/193 [01:09<00:06,  2.67it/s]
Caching text embeddings to disk:  91%|#########################################################4     | 176/193 [01:09<00:06,  2.67it/s]
Caching text embeddings to disk:  91%|#########################################################4     | 176/193 [01:09<00:06,  2.67it/s]
Caching text embeddings to disk:  92%|#########################################################7     | 177/193 [01:10<00:05,  2.67it/s]
Caching text embeddings to disk:  92%|#########################################################7     | 177/193 [01:10<00:05,  2.67it/s]
Caching text embeddings to disk:  92%|##########################################################1    | 178/193 [01:10<00:05,  2.67it/s]
Caching text embeddings to disk:  92%|##########################################################1    | 178/193 [01:10<00:05,  2.67it/s]
Caching text embeddings to disk:  93%|##########################################################4    | 179/193 [01:10<00:05,  2.67it/s]
Caching text embeddings to disk:  93%|##########################################################4    | 179/193 [01:10<00:05,  2.67it/s]
Caching text embeddings to disk:  93%|##########################################################7    | 180/193 [01:11<00:04,  2.67it/s]
Caching text embeddings to disk:  93%|##########################################################7    | 180/193 [01:11<00:04,  2.67it/s]
Caching text embeddings to disk:  94%|###########################################################    | 181/193 [01:11<00:04,  2.64it/s]
Caching text embeddings to disk:  94%|###########################################################    | 181/193 [01:11<00:04,  2.64it/s]
Caching text embeddings to disk:  94%|###########################################################4   | 182/193 [01:11<00:04,  2.65it/s]
Caching text embeddings to disk:  94%|###########################################################4   | 182/193 [01:11<00:04,  2.65it/s]
Caching text embeddings to disk:  95%|###########################################################7   | 183/193 [01:12<00:03,  2.65it/s]
Caching text embeddings to disk:  95%|###########################################################7   | 183/193 [01:12<00:03,  2.65it/s]
Caching text embeddings to disk:  95%|############################################################   | 184/193 [01:12<00:03,  2.66it/s]
Caching text embeddings to disk:  95%|############################################################   | 184/193 [01:12<00:03,  2.66it/s]
Caching text embeddings to disk:  96%|############################################################3  | 185/193 [01:13<00:03,  2.66it/s]
Caching text embeddings to disk:  96%|############################################################3  | 185/193 [01:13<00:03,  2.66it/s]
Caching text embeddings to disk:  96%|############################################################7  | 186/193 [01:13<00:02,  2.66it/s]
Caching text embeddings to disk:  96%|############################################################7  | 186/193 [01:13<00:02,  2.66it/s]
Caching text embeddings to disk:  97%|#############################################################  | 187/193 [01:13<00:02,  2.66it/s]
Caching text embeddings to disk:  97%|#############################################################  | 187/193 [01:13<00:02,  2.66it/s]
Caching text embeddings to disk:  97%|#############################################################3 | 188/193 [01:14<00:01,  2.66it/s]
Caching text embeddings to disk:  97%|#############################################################3 | 188/193 [01:14<00:01,  2.66it/s]
Caching text embeddings to disk:  98%|#############################################################6 | 189/193 [01:14<00:01,  2.66it/s]
Caching text embeddings to disk:  98%|#############################################################6 | 189/193 [01:14<00:01,  2.66it/s]
Caching text embeddings to disk:  98%|############################################################## | 190/193 [01:14<00:01,  2.66it/s]
Caching text embeddings to disk:  98%|############################################################## | 190/193 [01:14<00:01,  2.66it/s]
Caching text embeddings to disk:  99%|##############################################################3| 191/193 [01:15<00:00,  2.66it/s]
Caching text embeddings to disk:  99%|##############################################################3| 191/193 [01:15<00:00,  2.66it/s]
Caching text embeddings to disk:  99%|##############################################################6| 192/193 [01:15<00:00,  2.66it/s]
Caching text embeddings to disk:  99%|##############################################################6| 192/193 [01:15<00:00,  2.66it/s]
Caching text embeddings to disk: 100%|###############################################################| 193/193 [01:16<00:00,  2.66it/s]
Caching text embeddings to disk: 100%|###############################################################| 193/193 [01:16<00:00,  2.66it/s]
Caching text embeddings to disk: 100%|###############################################################| 193/193 [01:16<00:00,  2.54it/s]
Caching text embeddings to disk: 100%|###############################################################| 193/193 [01:16<00:00,  2.54it/s]
+
+Dataset: /home/featurize/ai-toolkit/datasets/Xiang_Handsome_Flux_SRPO_Keye_Captioned
+  -  Preprocessing image dimensions
+
  0%|                                                                                                          | 0/193 [00:00<?, ?it/s]
  0%|                                                                                                          | 0/193 [00:00<?, ?it/s]
100%|#############################################################################################| 193/193 [00:00<00:00, 39757.41it/s]
100%|#############################################################################################| 193/193 [00:00<00:00, 39757.41it/s]
+
+  -  Found 193 images
+Bucket sizes for /home/featurize/ai-toolkit/datasets/Xiang_Handsome_Flux_SRPO_Keye_Captioned:
+768x768: 193 files
+1 buckets made
+Caching text_embeddings for /home/featurize/ai-toolkit/datasets/Xiang_Handsome_Flux_SRPO_Keye_Captioned
+ - Saving text embeddings to disk
+
Caching text embeddings to disk:   0%|                                                                         | 0/193 [00:00<?, ?it/s]
Caching text embeddings to disk:   0%|                                                                         | 0/193 [00:00<?, ?it/s]
Caching text embeddings to disk:   1%|3                                                                | 1/193 [00:02<08:38,  2.70s/it]
Caching text embeddings to disk:   1%|3                                                                | 1/193 [00:02<08:38,  2.70s/it]
Caching text embeddings to disk:  12%|#######9                                                        | 24/193 [00:03<00:16, 10.47it/s]
Caching text embeddings to disk:  12%|#######9                                                        | 24/193 [00:03<00:16, 10.47it/s]
Caching text embeddings to disk:  17%|##########9                                                     | 33/193 [00:03<00:12, 13.06it/s]
Caching text embeddings to disk:  17%|##########9                                                     | 33/193 [00:03<00:12, 13.06it/s]
Caching text embeddings to disk:  19%|############2                                                   | 37/193 [00:03<00:12, 12.56it/s]
Caching text embeddings to disk:  19%|############2                                                   | 37/193 [00:03<00:12, 12.56it/s]
Caching text embeddings to disk:  22%|#############9                                                  | 42/193 [00:04<00:11, 12.74it/s]
Caching text embeddings to disk:  22%|#############9                                                  | 42/193 [00:04<00:11, 12.74it/s]
Caching text embeddings to disk:  27%|#################5                                              | 53/193 [00:04<00:08, 16.91it/s]
Caching text embeddings to disk:  27%|#################5                                              | 53/193 [00:04<00:08, 16.91it/s]
Caching text embeddings to disk:  30%|###################2                                            | 58/193 [00:04<00:08, 15.99it/s]
Caching text embeddings to disk:  30%|###################2                                            | 58/193 [00:04<00:08, 15.99it/s]
Caching text embeddings to disk:  36%|######################8                                         | 69/193 [00:05<00:06, 19.68it/s]
Caching text embeddings to disk:  36%|######################8                                         | 69/193 [00:05<00:06, 19.68it/s]
Caching text embeddings to disk:  55%|##################################6                            | 106/193 [00:05<00:02, 41.92it/s]
Caching text embeddings to disk:  55%|##################################6                            | 106/193 [00:05<00:02, 41.92it/s]
Caching text embeddings to disk:  62%|######################################8                        | 119/193 [00:06<00:01, 39.83it/s]
Caching text embeddings to disk:  62%|######################################8                        | 119/193 [00:06<00:01, 39.83it/s]
Caching text embeddings to disk:  64%|########################################4                      | 124/193 [00:06<00:02, 32.19it/s]
Caching text embeddings to disk:  64%|########################################4                      | 124/193 [00:06<00:02, 32.19it/s]
Caching text embeddings to disk:  80%|##################################################5            | 155/193 [00:06<00:00, 46.99it/s]
Caching text embeddings to disk:  80%|##################################################5            | 155/193 [00:06<00:00, 46.99it/s]
Caching text embeddings to disk:  84%|####################################################8          | 162/193 [00:07<00:00, 38.65it/s]
Caching text embeddings to disk:  84%|####################################################8          | 162/193 [00:07<00:00, 38.65it/s]
Caching text embeddings to disk:  86%|######################################################1        | 166/193 [00:07<00:00, 30.36it/s]
Caching text embeddings to disk:  86%|######################################################1        | 166/193 [00:07<00:00, 30.36it/s]
Caching text embeddings to disk:  99%|##############################################################3| 191/193 [00:07<00:00, 41.16it/s]
Caching text embeddings to disk:  99%|##############################################################3| 191/193 [00:07<00:00, 41.16it/s]
Caching text embeddings to disk: 100%|###############################################################| 193/193 [00:07<00:00, 24.30it/s]
Caching text embeddings to disk: 100%|###############################################################| 193/193 [00:07<00:00, 24.30it/s]
+
+
+***** UNLOADING TEXT ENCODER *****
+Embeddings cached to disk. We dont need the text encoder anymore
+***********************************
+
+Skipping first sample due to config setting
+
Xiang_wan_2_2_lora:   0%|                                                                                    | 0/30000 [00:00<?, ?it/s]
Xiang_wan_2_2_lora:   0%|                                                                                    | 0/30000 [00:00<?, ?it/s]
Xiang_wan_2_2_lora:   0%|                                                       | 0/30000 [00:06<?, ?it/s, lr: 1.0e-04 loss: 1.145e-01]
Xiang_wan_2_2_lora:   0%|                                                       | 0/30000 [00:06<?, ?it/s, lr: 1.0e-04 loss: 1.145e-01]
Xiang_wan_2_2_lora:   0%|                                                       | 0/30000 [00:06<?, ?it/s, lr: 1.0e-04 loss: 1.145e-01]
Xiang_wan_2_2_lora:   0%|                                                       | 0/30000 [00:06<?, ?it/s, lr: 1.0e-04 loss: 1.145e-01]
Xiang_wan_2_2_lora:   0%|                                                       | 0/30000 [00:16<?, ?it/s, lr: 1.0e-04 loss: 1.105e-01]
Xiang_wan_2_2_lora:   0%|                                                       | 0/30000 [00:16<?, ?it/s, lr: 1.0e-04 loss: 1.105e-01]
Xiang_wan_2_2_lora:   0%|                                            | 1/30000 [00:16<80:58:27,  9.72s/it, lr: 1.0e-04 loss: 1.105e-01]
Xiang_wan_2_2_lora:   0%|                                            | 1/30000 [00:16<80:58:27,  9.72s/it, lr: 1.0e-04 loss: 1.105e-01]
Xiang_wan_2_2_lora:   0%|                                            | 1/30000 [00:28<80:58:27,  9.72s/it, lr: 1.0e-04 loss: 9.129e-02]
Xiang_wan_2_2_lora:   0%|                                            | 1/30000 [00:28<80:58:27,  9.72s/it, lr: 1.0e-04 loss: 9.129e-02]
Xiang_wan_2_2_lora:   0%|                                            | 2/30000 [00:28<93:49:31, 11.26s/it, lr: 1.0e-04 loss: 9.129e-02]
Xiang_wan_2_2_lora:   0%|                                            | 2/30000 [00:28<93:49:31, 11.26s/it, lr: 1.0e-04 loss: 9.129e-02]
Xiang_wan_2_2_lora:   0%|                                            | 2/30000 [00:37<93:49:31, 11.26s/it, lr: 1.0e-04 loss: 6.329e-02]
Xiang_wan_2_2_lora:   0%|                                            | 2/30000 [00:37<93:49:31, 11.26s/it, lr: 1.0e-04 loss: 6.329e-02]
Xiang_wan_2_2_lora:   0%|                                            | 3/30000 [00:37<82:32:01,  9.91s/it, lr: 1.0e-04 loss: 6.329e-02]
Xiang_wan_2_2_lora:   0%|                                            | 3/30000 [00:37<82:32:01,  9.91s/it, lr: 1.0e-04 loss: 6.329e-02]
Xiang_wan_2_2_lora:   0%|                                            | 3/30000 [00:47<82:32:01,  9.91s/it, lr: 1.0e-04 loss: 2.388e-01]
Xiang_wan_2_2_lora:   0%|                                            | 3/30000 [00:47<82:32:01,  9.91s/it, lr: 1.0e-04 loss: 2.388e-01]
Xiang_wan_2_2_lora:   0%|                                            | 4/30000 [00:47<84:34:33, 10.15s/it, lr: 1.0e-04 loss: 2.388e-01]
Xiang_wan_2_2_lora:   0%|                                            | 4/30000 [00:47<84:34:33, 10.15s/it, lr: 1.0e-04 loss: 2.388e-01]
Xiang_wan_2_2_lora:   0%|                                            | 4/30000 [00:58<84:34:33, 10.15s/it, lr: 1.0e-04 loss: 4.219e-02]
Xiang_wan_2_2_lora:   0%|                                            | 4/30000 [00:58<84:34:33, 10.15s/it, lr: 1.0e-04 loss: 4.219e-02]
Xiang_wan_2_2_lora:   0%|                                            | 5/30000 [00:58<85:44:16, 10.29s/it, lr: 1.0e-04 loss: 4.219e-02]
Xiang_wan_2_2_lora:   0%|                                            | 5/30000 [00:58<85:44:16, 10.29s/it, lr: 1.0e-04 loss: 4.219e-02]
Xiang_wan_2_2_lora:   0%|                                            | 5/30000 [01:05<85:44:16, 10.29s/it, lr: 1.0e-04 loss: 1.056e-01]
Xiang_wan_2_2_lora:   0%|                                            | 5/30000 [01:05<85:44:16, 10.29s/it, lr: 1.0e-04 loss: 1.056e-01]
Xiang_wan_2_2_lora:   0%|                                            | 6/30000 [01:05<77:12:13,  9.27s/it, lr: 1.0e-04 loss: 1.056e-01]
Xiang_wan_2_2_lora:   0%|                                            | 6/30000 [01:05<77:12:13,  9.27s/it, lr: 1.0e-04 loss: 1.056e-01]
Xiang_wan_2_2_lora:   0%|                                            | 6/30000 [01:13<77:12:13,  9.27s/it, lr: 1.0e-04 loss: 9.483e-02]
Xiang_wan_2_2_lora:   0%|                                            | 6/30000 [01:13<77:12:13,  9.27s/it, lr: 1.0e-04 loss: 9.483e-02]
Xiang_wan_2_2_lora:   0%|                                            | 7/30000 [01:13<73:25:55,  8.81s/it, lr: 1.0e-04 loss: 9.483e-02]
Xiang_wan_2_2_lora:   0%|                                            | 7/30000 [01:13<73:25:55,  8.81s/it, lr: 1.0e-04 loss: 9.483e-02]
Xiang_wan_2_2_lora:   0%|                                            | 7/30000 [01:20<73:25:55,  8.81s/it, lr: 1.0e-04 loss: 1.978e-01]
Xiang_wan_2_2_lora:   0%|                                            | 7/30000 [01:20<73:25:55,  8.81s/it, lr: 1.0e-04 loss: 1.978e-01]
Xiang_wan_2_2_lora:   0%|                                            | 8/30000 [01:20<69:15:36,  8.31s/it, lr: 1.0e-04 loss: 1.978e-01]
Xiang_wan_2_2_lora:   0%|                                            | 8/30000 [01:20<69:15:36,  8.31s/it, lr: 1.0e-04 loss: 1.978e-01]
Xiang_wan_2_2_lora:   0%|                                            | 8/30000 [01:31<69:15:36,  8.31s/it, lr: 1.0e-04 loss: 5.385e-02]
Xiang_wan_2_2_lora:   0%|                                            | 8/30000 [01:31<69:15:36,  8.31s/it, lr: 1.0e-04 loss: 5.385e-02]
Xiang_wan_2_2_lora:   0%|                                            | 9/30000 [01:31<74:26:57,  8.94s/it, lr: 1.0e-04 loss: 5.385e-02]
Xiang_wan_2_2_lora:   0%|                                            | 9/30000 [01:31<74:26:57,  8.94s/it, lr: 1.0e-04 loss: 5.385e-02]
Xiang_wan_2_2_lora:   0%|                                            | 9/30000 [01:38<74:26:57,  8.94s/it, lr: 1.0e-04 loss: 2.412e-01]
Xiang_wan_2_2_lora:   0%|                                            | 9/30000 [01:38<74:26:57,  8.94s/it, lr: 1.0e-04 loss: 2.412e-01]
Xiang_wan_2_2_lora:   0%|                                           | 10/30000 [01:45<74:26:48,  8.94s/it, lr: 1.0e-04 loss: 6.097e-01]
Xiang_wan_2_2_lora:   0%|                                           | 10/30000 [01:45<74:26:48,  8.94s/it, lr: 1.0e-04 loss: 6.097e-01]
Xiang_wan_2_2_lora:   0%|                                           | 11/30000 [01:45<53:46:44,  6.46s/it, lr: 1.0e-04 loss: 6.097e-01]
Xiang_wan_2_2_lora:   0%|                                           | 11/30000 [01:45<53:46:44,  6.46s/it, lr: 1.0e-04 loss: 6.097e-01]
Xiang_wan_2_2_lora:   0%|                                           | 11/30000 [01:56<53:46:44,  6.46s/it, lr: 1.0e-04 loss: 1.320e-01]
Xiang_wan_2_2_lora:   0%|                                           | 11/30000 [01:56<53:46:44,  6.46s/it, lr: 1.0e-04 loss: 1.320e-01]
Xiang_wan_2_2_lora:   0%|                                           | 12/30000 [01:56<61:36:05,  7.40s/it, lr: 1.0e-04 loss: 1.320e-01]
Xiang_wan_2_2_lora:   0%|                                           | 12/30000 [01:56<61:36:05,  7.40s/it, lr: 1.0e-04 loss: 1.320e-01]
Xiang_wan_2_2_lora:   0%|                                           | 12/30000 [02:06<61:36:05,  7.40s/it, lr: 1.0e-04 loss: 5.095e-02]
Xiang_wan_2_2_lora:   0%|                                           | 12/30000 [02:06<61:36:05,  7.40s/it, lr: 1.0e-04 loss: 5.095e-02]
Xiang_wan_2_2_lora:   0%|                                           | 13/30000 [02:06<67:25:43,  8.09s/it, lr: 1.0e-04 loss: 5.095e-02]
Xiang_wan_2_2_lora:   0%|                                           | 13/30000 [02:06<67:25:43,  8.09s/it, lr: 1.0e-04 loss: 5.095e-02]
Xiang_wan_2_2_lora:   0%|                                           | 13/30000 [02:16<67:25:43,  8.09s/it, lr: 1.0e-04 loss: 1.433e-01]
Xiang_wan_2_2_lora:   0%|                                           | 13/30000 [02:16<67:25:43,  8.09s/it, lr: 1.0e-04 loss: 1.433e-01]
Xiang_wan_2_2_lora:   0%|                                           | 14/30000 [02:16<72:14:40,  8.67s/it, lr: 1.0e-04 loss: 1.433e-01]
Xiang_wan_2_2_lora:   0%|                                           | 14/30000 [02:16<72:14:40,  8.67s/it, lr: 1.0e-04 loss: 1.433e-01]
Xiang_wan_2_2_lora:   0%|                                           | 14/30000 [02:26<72:14:40,  8.67s/it, lr: 1.0e-04 loss: 1.411e-01]
Xiang_wan_2_2_lora:   0%|                                           | 14/30000 [02:26<72:14:40,  8.67s/it, lr: 1.0e-04 loss: 1.411e-01]
Xiang_wan_2_2_lora:   0%|                                           | 15/30000 [02:26<76:28:21,  9.18s/it, lr: 1.0e-04 loss: 1.411e-01]
Xiang_wan_2_2_lora:   0%|                                           | 15/30000 [02:26<76:28:21,  9.18s/it, lr: 1.0e-04 loss: 1.411e-01]
Xiang_wan_2_2_lora:   0%|                                           | 15/30000 [02:33<76:28:21,  9.18s/it, lr: 1.0e-04 loss: 1.055e-01]
Xiang_wan_2_2_lora:   0%|                                           | 15/30000 [02:33<76:28:21,  9.18s/it, lr: 1.0e-04 loss: 1.055e-01]
Xiang_wan_2_2_lora:   0%|                                           | 16/30000 [02:33<71:33:48,  8.59s/it, lr: 1.0e-04 loss: 1.055e-01]
Xiang_wan_2_2_lora:   0%|                                           | 16/30000 [02:33<71:33:48,  8.59s/it, lr: 1.0e-04 loss: 1.055e-01]
Xiang_wan_2_2_lora:   0%|                                           | 16/30000 [02:43<71:33:48,  8.59s/it, lr: 1.0e-04 loss: 7.419e-01]
Xiang_wan_2_2_lora:   0%|                                           | 16/30000 [02:43<71:33:48,  8.59s/it, lr: 1.0e-04 loss: 7.419e-01]
Xiang_wan_2_2_lora:   0%|                                           | 17/30000 [02:43<74:57:58,  9.00s/it, lr: 1.0e-04 loss: 7.419e-01]
Xiang_wan_2_2_lora:   0%|                                           | 17/30000 [02:43<74:57:58,  9.00s/it, lr: 1.0e-04 loss: 7.419e-01]
Xiang_wan_2_2_lora:   0%|                                           | 17/30000 [02:50<74:57:58,  9.00s/it, lr: 1.0e-04 loss: 1.205e-01]
Xiang_wan_2_2_lora:   0%|                                           | 17/30000 [02:50<74:57:58,  9.00s/it, lr: 1.0e-04 loss: 1.205e-01]
Xiang_wan_2_2_lora:   0%|                                           | 18/30000 [02:50<69:37:34,  8.36s/it, lr: 1.0e-04 loss: 1.205e-01]
Xiang_wan_2_2_lora:   0%|                                           | 18/30000 [02:50<69:37:34,  8.36s/it, lr: 1.0e-04 loss: 1.205e-01]
Xiang_wan_2_2_lora:   0%|                                           | 18/30000 [03:00<69:37:34,  8.36s/it, lr: 1.0e-04 loss: 3.269e-02]
Xiang_wan_2_2_lora:   0%|                                           | 18/30000 [03:00<69:37:34,  8.36s/it, lr: 1.0e-04 loss: 3.269e-02]
Xiang_wan_2_2_lora:   0%|                                           | 19/30000 [03:00<73:16:40,  8.80s/it, lr: 1.0e-04 loss: 3.269e-02]
Xiang_wan_2_2_lora:   0%|                                           | 19/30000 [03:00<73:16:40,  8.80s/it, lr: 1.0e-04 loss: 3.269e-02]
Xiang_wan_2_2_lora:   0%|                                           | 19/30000 [03:07<73:16:40,  8.80s/it, lr: 1.0e-04 loss: 9.606e-02]
Xiang_wan_2_2_lora:   0%|                                           | 19/30000 [03:07<73:16:40,  8.80s/it, lr: 1.0e-04 loss: 9.606e-02]
Xiang_wan_2_2_lora:   0%|                                           | 20/30000 [03:15<73:16:31,  8.80s/it, lr: 1.0e-04 loss: 7.319e-02]
Xiang_wan_2_2_lora:   0%|                                           | 20/30000 [03:15<73:16:31,  8.80s/it, lr: 1.0e-04 loss: 7.319e-02]
Xiang_wan_2_2_lora:   0%|                                           | 21/30000 [03:15<53:48:23,  6.46s/it, lr: 1.0e-04 loss: 7.319e-02]
Xiang_wan_2_2_lora:   0%|                                           | 21/30000 [03:15<53:48:23,  6.46s/it, lr: 1.0e-04 loss: 7.319e-02]
Xiang_wan_2_2_lora:   0%|                                           | 21/30000 [03:22<53:48:23,  6.46s/it, lr: 1.0e-04 loss: 1.345e-01]
Xiang_wan_2_2_lora:   0%|                                           | 21/30000 [03:22<53:48:23,  6.46s/it, lr: 1.0e-04 loss: 1.345e-01]
Xiang_wan_2_2_lora:   0%|                                           | 22/30000 [03:22<56:25:07,  6.78s/it, lr: 1.0e-04 loss: 1.345e-01]
Xiang_wan_2_2_lora:   0%|                                           | 22/30000 [03:22<56:25:07,  6.78s/it, lr: 1.0e-04 loss: 1.345e-01]
Xiang_wan_2_2_lora:   0%|                                           | 22/30000 [03:30<56:25:07,  6.78s/it, lr: 1.0e-04 loss: 4.474e-01]
Xiang_wan_2_2_lora:   0%|                                           | 22/30000 [03:30<56:25:07,  6.78s/it, lr: 1.0e-04 loss: 4.474e-01]
Xiang_wan_2_2_lora:   0%|                                           | 23/30000 [03:30<58:06:08,  6.98s/it, lr: 1.0e-04 loss: 4.474e-01]
Xiang_wan_2_2_lora:   0%|                                           | 23/30000 [03:30<58:06:08,  6.98s/it, lr: 1.0e-04 loss: 4.474e-01]
Xiang_wan_2_2_lora:   0%|                                           | 23/30000 [03:37<58:06:08,  6.98s/it, lr: 1.0e-04 loss: 1.491e-01]
Xiang_wan_2_2_lora:   0%|                                           | 23/30000 [03:37<58:06:08,  6.98s/it, lr: 1.0e-04 loss: 1.491e-01]
Xiang_wan_2_2_lora:   0%|                                           | 24/30000 [03:37<58:30:53,  7.03s/it, lr: 1.0e-04 loss: 1.491e-01]
Xiang_wan_2_2_lora:   0%|                                           | 24/30000 [03:37<58:30:53,  7.03s/it, lr: 1.0e-04 loss: 1.491e-01]
Xiang_wan_2_2_lora:   0%|                                           | 24/30000 [03:47<58:30:53,  7.03s/it, lr: 1.0e-04 loss: 9.066e-01]
Xiang_wan_2_2_lora:   0%|                                           | 24/30000 [03:47<58:30:53,  7.03s/it, lr: 1.0e-04 loss: 9.066e-01]
Xiang_wan_2_2_lora:   0%|                                           | 25/30000 [03:47<65:16:07,  7.84s/it, lr: 1.0e-04 loss: 9.066e-01]
Xiang_wan_2_2_lora:   0%|                                           | 25/30000 [03:47<65:16:07,  7.84s/it, lr: 1.0e-04 loss: 9.066e-01]
Xiang_wan_2_2_lora:   0%|                                           | 25/30000 [03:57<65:16:07,  7.84s/it, lr: 1.0e-04 loss: 1.169e-01]
Xiang_wan_2_2_lora:   0%|                                           | 25/30000 [03:57<65:16:07,  7.84s/it, lr: 1.0e-04 loss: 1.169e-01]
Xiang_wan_2_2_lora:   0%|                                           | 26/30000 [03:57<69:38:34,  8.36s/it, lr: 1.0e-04 loss: 1.169e-01]
Xiang_wan_2_2_lora:   0%|                                           | 26/30000 [03:57<69:38:34,  8.36s/it, lr: 1.0e-04 loss: 1.169e-01]
Xiang_wan_2_2_lora:   0%|                                           | 26/30000 [04:07<69:38:34,  8.36s/it, lr: 1.0e-04 loss: 5.878e-02]
Xiang_wan_2_2_lora:   0%|                                           | 26/30000 [04:07<69:38:34,  8.36s/it, lr: 1.0e-04 loss: 5.878e-02]
Xiang_wan_2_2_lora:   0%|                                           | 27/30000 [04:07<73:57:01,  8.88s/it, lr: 1.0e-04 loss: 5.878e-02]
Xiang_wan_2_2_lora:   0%|                                           | 27/30000 [04:07<73:57:01,  8.88s/it, lr: 1.0e-04 loss: 5.878e-02]
Xiang_wan_2_2_lora:   0%|                                           | 27/30000 [04:14<73:57:01,  8.88s/it, lr: 1.0e-04 loss: 1.083e-01]
Xiang_wan_2_2_lora:   0%|                                           | 27/30000 [04:14<73:57:01,  8.88s/it, lr: 1.0e-04 loss: 1.083e-01]
Xiang_wan_2_2_lora:   0%|                                           | 28/30000 [04:14<70:06:06,  8.42s/it, lr: 1.0e-04 loss: 1.083e-01]
Xiang_wan_2_2_lora:   0%|                                           | 28/30000 [04:14<70:06:06,  8.42s/it, lr: 1.0e-04 loss: 1.083e-01]
Xiang_wan_2_2_lora:   0%|                                           | 28/30000 [04:24<70:06:06,  8.42s/it, lr: 1.0e-04 loss: 5.152e-02]
Xiang_wan_2_2_lora:   0%|                                           | 28/30000 [04:24<70:06:06,  8.42s/it, lr: 1.0e-04 loss: 5.152e-02]
Xiang_wan_2_2_lora:   0%|                                           | 29/30000 [04:24<74:56:05,  9.00s/it, lr: 1.0e-04 loss: 5.152e-02]
Xiang_wan_2_2_lora:   0%|                                           | 29/30000 [04:24<74:56:05,  9.00s/it, lr: 1.0e-04 loss: 5.152e-02]
Xiang_wan_2_2_lora:   0%|                                           | 29/30000 [04:32<74:56:05,  9.00s/it, lr: 1.0e-04 loss: 1.198e-01]
Xiang_wan_2_2_lora:   0%|                                           | 29/30000 [04:32<74:56:05,  9.00s/it, lr: 1.0e-04 loss: 1.198e-01]
Xiang_wan_2_2_lora:   0%|                                           | 30/30000 [04:39<74:55:56,  9.00s/it, lr: 1.0e-04 loss: 7.922e-02]
Xiang_wan_2_2_lora:   0%|                                           | 30/30000 [04:39<74:55:56,  9.00s/it, lr: 1.0e-04 loss: 7.922e-02]
Xiang_wan_2_2_lora:   0%|                                           | 31/30000 [04:39<54:39:57,  6.57s/it, lr: 1.0e-04 loss: 7.922e-02]
Xiang_wan_2_2_lora:   0%|                                           | 31/30000 [04:39<54:39:57,  6.57s/it, lr: 1.0e-04 loss: 7.922e-02]
Xiang_wan_2_2_lora:   0%|                                           | 31/30000 [04:46<54:39:57,  6.57s/it, lr: 1.0e-04 loss: 1.100e-01]
Xiang_wan_2_2_lora:   0%|                                           | 31/30000 [04:46<54:39:57,  6.57s/it, lr: 1.0e-04 loss: 1.100e-01]
Xiang_wan_2_2_lora:   0%|                                           | 32/30000 [04:46<55:21:35,  6.65s/it, lr: 1.0e-04 loss: 1.100e-01]
Xiang_wan_2_2_lora:   0%|                                           | 32/30000 [04:46<55:21:35,  6.65s/it, lr: 1.0e-04 loss: 1.100e-01]
Xiang_wan_2_2_lora:   0%|                                           | 32/30000 [04:56<55:21:35,  6.65s/it, lr: 1.0e-04 loss: 6.767e-02]
Xiang_wan_2_2_lora:   0%|                                           | 32/30000 [04:56<55:21:35,  6.65s/it, lr: 1.0e-04 loss: 6.767e-02]
Xiang_wan_2_2_lora:   0%|                                           | 33/30000 [04:56<63:02:17,  7.57s/it, lr: 1.0e-04 loss: 6.767e-02]
Xiang_wan_2_2_lora:   0%|                                           | 33/30000 [04:56<63:02:17,  7.57s/it, lr: 1.0e-04 loss: 6.767e-02]
Xiang_wan_2_2_lora:   0%|                                           | 33/30000 [05:03<63:02:17,  7.57s/it, lr: 1.0e-04 loss: 1.345e-01]
Xiang_wan_2_2_lora:   0%|                                           | 33/30000 [05:03<63:02:17,  7.57s/it, lr: 1.0e-04 loss: 1.345e-01]
Xiang_wan_2_2_lora:   0%|                                           | 34/30000 [05:03<61:12:37,  7.35s/it, lr: 1.0e-04 loss: 1.345e-01]
Xiang_wan_2_2_lora:   0%|                                           | 34/30000 [05:03<61:12:37,  7.35s/it, lr: 1.0e-04 loss: 1.345e-01]
Xiang_wan_2_2_lora:   0%|                                           | 34/30000 [05:12<61:12:37,  7.35s/it, lr: 1.0e-04 loss: 5.635e-02]
Xiang_wan_2_2_lora:   0%|                                           | 34/30000 [05:12<61:12:37,  7.35s/it, lr: 1.0e-04 loss: 5.635e-02]
Xiang_wan_2_2_lora:   0%|                                           | 35/30000 [05:12<65:59:28,  7.93s/it, lr: 1.0e-04 loss: 5.635e-02]
Xiang_wan_2_2_lora:   0%|                                           | 35/30000 [05:12<65:59:28,  7.93s/it, lr: 1.0e-04 loss: 5.635e-02]
Xiang_wan_2_2_lora:   0%|                                           | 35/30000 [05:22<65:59:28,  7.93s/it, lr: 1.0e-04 loss: 1.037e-01]
Xiang_wan_2_2_lora:   0%|                                           | 35/30000 [05:22<65:59:28,  7.93s/it, lr: 1.0e-04 loss: 1.037e-01]
Xiang_wan_2_2_lora:   0%|                                           | 36/30000 [05:22<71:01:56,  8.53s/it, lr: 1.0e-04 loss: 1.037e-01]
Xiang_wan_2_2_lora:   0%|                                           | 36/30000 [05:22<71:01:56,  8.53s/it, lr: 1.0e-04 loss: 1.037e-01]
Xiang_wan_2_2_lora:   0%|                                           | 36/30000 [05:32<71:01:56,  8.53s/it, lr: 1.0e-04 loss: 1.075e-01]
Xiang_wan_2_2_lora:   0%|                                           | 36/30000 [05:32<71:01:56,  8.53s/it, lr: 1.0e-04 loss: 1.075e-01]
Xiang_wan_2_2_lora:   0%|                                           | 37/30000 [05:32<74:44:21,  8.98s/it, lr: 1.0e-04 loss: 1.075e-01]
Xiang_wan_2_2_lora:   0%|                                           | 37/30000 [05:32<74:44:21,  8.98s/it, lr: 1.0e-04 loss: 1.075e-01]
Xiang_wan_2_2_lora:   0%|                                           | 37/30000 [05:42<74:44:21,  8.98s/it, lr: 1.0e-04 loss: 9.238e-02]
Xiang_wan_2_2_lora:   0%|                                           | 37/30000 [05:42<74:44:21,  8.98s/it, lr: 1.0e-04 loss: 9.238e-02]
Xiang_wan_2_2_lora:   0%|                                           | 38/30000 [05:42<77:06:55,  9.27s/it, lr: 1.0e-04 loss: 9.238e-02]
Xiang_wan_2_2_lora:   0%|                                           | 38/30000 [05:42<77:06:55,  9.27s/it, lr: 1.0e-04 loss: 9.238e-02]
Xiang_wan_2_2_lora:   0%|                                           | 38/30000 [05:52<77:06:55,  9.27s/it, lr: 1.0e-04 loss: 4.581e-02]
Xiang_wan_2_2_lora:   0%|                                           | 38/30000 [05:52<77:06:55,  9.27s/it, lr: 1.0e-04 loss: 4.581e-02]
Xiang_wan_2_2_lora:   0%|                                           | 39/30000 [05:52<79:33:43,  9.56s/it, lr: 1.0e-04 loss: 4.581e-02]
Xiang_wan_2_2_lora:   0%|                                           | 39/30000 [05:52<79:33:43,  9.56s/it, lr: 1.0e-04 loss: 4.581e-02]
Xiang_wan_2_2_lora:   0%|                                           | 39/30000 [06:03<79:33:43,  9.56s/it, lr: 1.0e-04 loss: 1.172e-01]
Xiang_wan_2_2_lora:   0%|                                           | 39/30000 [06:03<79:33:43,  9.56s/it, lr: 1.0e-04 loss: 1.172e-01]
Xiang_wan_2_2_lora:   0%|                                           | 40/30000 [06:10<79:33:33,  9.56s/it, lr: 1.0e-04 loss: 8.461e-02]
Xiang_wan_2_2_lora:   0%|                                           | 40/30000 [06:10<79:33:33,  9.56s/it, lr: 1.0e-04 loss: 8.461e-02]
Xiang_wan_2_2_lora:   0%|                                           | 41/30000 [06:10<57:07:52,  6.87s/it, lr: 1.0e-04 loss: 8.461e-02]
Xiang_wan_2_2_lora:   0%|                                           | 41/30000 [06:10<57:07:52,  6.87s/it, lr: 1.0e-04 loss: 8.461e-02]
Xiang_wan_2_2_lora:   0%|                                           | 41/30000 [06:17<57:07:52,  6.87s/it, lr: 1.0e-04 loss: 9.328e-02]
Xiang_wan_2_2_lora:   0%|                                           | 41/30000 [06:17<57:07:52,  6.87s/it, lr: 1.0e-04 loss: 9.328e-02]
Xiang_wan_2_2_lora:   0%|                                           | 42/30000 [06:17<58:04:16,  6.98s/it, lr: 1.0e-04 loss: 9.328e-02]
Xiang_wan_2_2_lora:   0%|                                           | 42/30000 [06:17<58:04:16,  6.98s/it, lr: 1.0e-04 loss: 9.328e-02]
Xiang_wan_2_2_lora:   0%|                                           | 42/30000 [06:28<58:04:16,  6.98s/it, lr: 1.0e-04 loss: 6.702e-02]
Xiang_wan_2_2_lora:   0%|                                           | 42/30000 [06:28<58:04:16,  6.98s/it, lr: 1.0e-04 loss: 6.702e-02]
Xiang_wan_2_2_lora:   0%|                                           | 43/30000 [06:28<65:25:33,  7.86s/it, lr: 1.0e-04 loss: 6.702e-02]
Xiang_wan_2_2_lora:   0%|                                           | 43/30000 [06:28<65:25:33,  7.86s/it, lr: 1.0e-04 loss: 6.702e-02]
Xiang_wan_2_2_lora:   0%|                                           | 43/30000 [06:35<65:25:33,  7.86s/it, lr: 1.0e-04 loss: 1.280e-01]
Xiang_wan_2_2_lora:   0%|                                           | 43/30000 [06:35<65:25:33,  7.86s/it, lr: 1.0e-04 loss: 1.280e-01]
Xiang_wan_2_2_lora:   0%|                                           | 44/30000 [06:35<64:42:02,  7.78s/it, lr: 1.0e-04 loss: 1.280e-01]
Xiang_wan_2_2_lora:   0%|                                           | 44/30000 [06:35<64:42:02,  7.78s/it, lr: 1.0e-04 loss: 1.280e-01]
Xiang_wan_2_2_lora:   0%|                                           | 44/30000 [06:43<64:42:02,  7.78s/it, lr: 1.0e-04 loss: 1.327e-01]
Xiang_wan_2_2_lora:   0%|                                           | 44/30000 [06:43<64:42:02,  7.78s/it, lr: 1.0e-04 loss: 1.327e-01]
Xiang_wan_2_2_lora:   0%|                                           | 45/30000 [06:43<63:37:01,  7.65s/it, lr: 1.0e-04 loss: 1.327e-01]
Xiang_wan_2_2_lora:   0%|                                           | 45/30000 [06:43<63:37:01,  7.65s/it, lr: 1.0e-04 loss: 1.327e-01]
Xiang_wan_2_2_lora:   0%|                                           | 45/30000 [06:50<63:37:01,  7.65s/it, lr: 1.0e-04 loss: 1.095e-01]
Xiang_wan_2_2_lora:   0%|                                           | 45/30000 [06:50<63:37:01,  7.65s/it, lr: 1.0e-04 loss: 1.095e-01]
Xiang_wan_2_2_lora:   0%|                                           | 46/30000 [06:50<62:52:47,  7.56s/it, lr: 1.0e-04 loss: 1.095e-01]
Xiang_wan_2_2_lora:   0%|                                           | 46/30000 [06:50<62:52:47,  7.56s/it, lr: 1.0e-04 loss: 1.095e-01]
Xiang_wan_2_2_lora:   0%|                                           | 46/30000 [07:00<62:52:47,  7.56s/it, lr: 1.0e-04 loss: 1.312e-01]
Xiang_wan_2_2_lora:   0%|                                           | 46/30000 [07:00<62:52:47,  7.56s/it, lr: 1.0e-04 loss: 1.312e-01]
Xiang_wan_2_2_lora:   0%|                                           | 47/30000 [07:00<68:00:48,  8.17s/it, lr: 1.0e-04 loss: 1.312e-01]
Xiang_wan_2_2_lora:   0%|                                           | 47/30000 [07:00<68:00:48,  8.17s/it, lr: 1.0e-04 loss: 1.312e-01]
Xiang_wan_2_2_lora:   0%|                                           | 47/30000 [07:07<68:00:48,  8.17s/it, lr: 1.0e-04 loss: 1.034e-01]
Xiang_wan_2_2_lora:   0%|                                           | 47/30000 [07:07<68:00:48,  8.17s/it, lr: 1.0e-04 loss: 1.034e-01]
Xiang_wan_2_2_lora:   0%|                                           | 48/30000 [07:07<66:02:30,  7.94s/it, lr: 1.0e-04 loss: 1.034e-01]
Xiang_wan_2_2_lora:   0%|                                           | 48/30000 [07:07<66:02:30,  7.94s/it, lr: 1.0e-04 loss: 1.034e-01]
Xiang_wan_2_2_lora:   0%|                                           | 48/30000 [07:17<66:02:30,  7.94s/it, lr: 1.0e-04 loss: 8.438e-02]
Xiang_wan_2_2_lora:   0%|                                           | 48/30000 [07:17<66:02:30,  7.94s/it, lr: 1.0e-04 loss: 8.438e-02]
Xiang_wan_2_2_lora:   0%|                                           | 49/30000 [07:17<70:40:22,  8.49s/it, lr: 1.0e-04 loss: 8.438e-02]
Xiang_wan_2_2_lora:   0%|                                           | 49/30000 [07:17<70:40:22,  8.49s/it, lr: 1.0e-04 loss: 8.438e-02]
Xiang_wan_2_2_lora:   0%|                                           | 49/30000 [07:24<70:40:22,  8.49s/it, lr: 1.0e-04 loss: 1.044e-01]
Xiang_wan_2_2_lora:   0%|                                           | 49/30000 [07:24<70:40:22,  8.49s/it, lr: 1.0e-04 loss: 1.044e-01]
Xiang_wan_2_2_lora:   0%|                                           | 50/30000 [07:31<70:40:13,  8.49s/it, lr: 1.0e-04 loss: 2.147e-01]
Xiang_wan_2_2_lora:   0%|                                           | 50/30000 [07:31<70:40:13,  8.49s/it, lr: 1.0e-04 loss: 2.147e-01]
Xiang_wan_2_2_lora:   0%|                                           | 51/30000 [07:31<51:07:23,  6.15s/it, lr: 1.0e-04 loss: 2.147e-01]
Xiang_wan_2_2_lora:   0%|                                           | 51/30000 [07:31<51:07:23,  6.15s/it, lr: 1.0e-04 loss: 2.147e-01]
Xiang_wan_2_2_lora:   0%|                                           | 51/30000 [07:41<51:07:23,  6.15s/it, lr: 1.0e-04 loss: 9.208e-02]
Xiang_wan_2_2_lora:   0%|                                           | 51/30000 [07:41<51:07:23,  6.15s/it, lr: 1.0e-04 loss: 9.208e-02]
Xiang_wan_2_2_lora:   0%|                                           | 52/30000 [07:41<59:21:46,  7.14s/it, lr: 1.0e-04 loss: 9.208e-02]
Xiang_wan_2_2_lora:   0%|                                           | 52/30000 [07:41<59:21:46,  7.14s/it, lr: 1.0e-04 loss: 9.208e-02]
Xiang_wan_2_2_lora:   0%|                                           | 52/30000 [07:48<59:21:46,  7.14s/it, lr: 1.0e-04 loss: 1.224e-01]
Xiang_wan_2_2_lora:   0%|                                           | 52/30000 [07:48<59:21:46,  7.14s/it, lr: 1.0e-04 loss: 1.224e-01]
Xiang_wan_2_2_lora:   0%|                                           | 53/30000 [07:48<59:47:43,  7.19s/it, lr: 1.0e-04 loss: 1.224e-01]
Xiang_wan_2_2_lora:   0%|                                           | 53/30000 [07:48<59:47:43,  7.19s/it, lr: 1.0e-04 loss: 1.224e-01]
Xiang_wan_2_2_lora:   0%|                                           | 53/30000 [07:55<59:47:43,  7.19s/it, lr: 1.0e-04 loss: 9.653e-02]
Xiang_wan_2_2_lora:   0%|                                           | 53/30000 [07:55<59:47:43,  7.19s/it, lr: 1.0e-04 loss: 9.653e-02]
Xiang_wan_2_2_lora:   0%|                                           | 54/30000 [07:55<60:02:41,  7.22s/it, lr: 1.0e-04 loss: 9.653e-02]
Xiang_wan_2_2_lora:   0%|                                           | 54/30000 [07:55<60:02:41,  7.22s/it, lr: 1.0e-04 loss: 9.653e-02]
Xiang_wan_2_2_lora:   0%|                                           | 54/30000 [08:05<60:02:41,  7.22s/it, lr: 1.0e-04 loss: 6.706e-02]
Xiang_wan_2_2_lora:   0%|                                           | 54/30000 [08:05<60:02:41,  7.22s/it, lr: 1.0e-04 loss: 6.706e-02]
Xiang_wan_2_2_lora:   0%|                                           | 55/30000 [08:05<65:11:38,  7.84s/it, lr: 1.0e-04 loss: 6.706e-02]
Xiang_wan_2_2_lora:   0%|                                           | 55/30000 [08:05<65:11:38,  7.84s/it, lr: 1.0e-04 loss: 6.706e-02]
Xiang_wan_2_2_lora:   0%|                                           | 55/30000 [08:12<65:11:38,  7.84s/it, lr: 1.0e-04 loss: 1.041e-01]
Xiang_wan_2_2_lora:   0%|                                           | 55/30000 [08:12<65:11:38,  7.84s/it, lr: 1.0e-04 loss: 1.041e-01]
Xiang_wan_2_2_lora:   0%|                                           | 56/30000 [08:12<63:54:03,  7.68s/it, lr: 1.0e-04 loss: 1.041e-01]
Xiang_wan_2_2_lora:   0%|                                           | 56/30000 [08:12<63:54:03,  7.68s/it, lr: 1.0e-04 loss: 1.041e-01]
Xiang_wan_2_2_lora:   0%|                                           | 56/30000 [08:19<63:54:03,  7.68s/it, lr: 1.0e-04 loss: 2.031e-01]
Xiang_wan_2_2_lora:   0%|                                           | 56/30000 [08:19<63:54:03,  7.68s/it, lr: 1.0e-04 loss: 2.031e-01]
Xiang_wan_2_2_lora:   0%|                                           | 57/30000 [08:19<63:00:34,  7.58s/it, lr: 1.0e-04 loss: 2.031e-01]
Xiang_wan_2_2_lora:   0%|                                           | 57/30000 [08:19<63:00:34,  7.58s/it, lr: 1.0e-04 loss: 2.031e-01]
Xiang_wan_2_2_lora:   0%|                                           | 57/30000 [08:27<63:00:34,  7.58s/it, lr: 1.0e-04 loss: 1.388e-01]
Xiang_wan_2_2_lora:   0%|                                           | 57/30000 [08:27<63:00:34,  7.58s/it, lr: 1.0e-04 loss: 1.388e-01]
Xiang_wan_2_2_lora:   0%|                                           | 58/30000 [08:27<62:48:20,  7.55s/it, lr: 1.0e-04 loss: 1.388e-01]
Xiang_wan_2_2_lora:   0%|                                           | 58/30000 [08:27<62:48:20,  7.55s/it, lr: 1.0e-04 loss: 1.388e-01]
Xiang_wan_2_2_lora:   0%|                                           | 58/30000 [08:37<62:48:20,  7.55s/it, lr: 1.0e-04 loss: 4.605e-02]
Xiang_wan_2_2_lora:   0%|                                           | 58/30000 [08:37<62:48:20,  7.55s/it, lr: 1.0e-04 loss: 4.605e-02]
Xiang_wan_2_2_lora:   0%|                                           | 59/30000 [08:37<68:50:35,  8.28s/it, lr: 1.0e-04 loss: 4.605e-02]
Xiang_wan_2_2_lora:   0%|                                           | 59/30000 [08:37<68:50:35,  8.28s/it, lr: 1.0e-04 loss: 4.605e-02]
Xiang_wan_2_2_lora:   0%|                                           | 59/30000 [08:44<68:50:35,  8.28s/it, lr: 1.0e-04 loss: 1.339e-01]
Xiang_wan_2_2_lora:   0%|                                           | 59/30000 [08:44<68:50:35,  8.28s/it, lr: 1.0e-04 loss: 1.339e-01]
Xiang_wan_2_2_lora:   0%|                                           | 60/30000 [08:53<68:50:27,  8.28s/it, lr: 1.0e-04 loss: 7.372e-02]
Xiang_wan_2_2_lora:   0%|                                           | 60/30000 [08:53<68:50:27,  8.28s/it, lr: 1.0e-04 loss: 7.372e-02]
Xiang_wan_2_2_lora:   0%|                                           | 61/30000 [08:53<55:43:32,  6.70s/it, lr: 1.0e-04 loss: 7.372e-02]
Xiang_wan_2_2_lora:   0%|                                           | 61/30000 [08:53<55:43:32,  6.70s/it, lr: 1.0e-04 loss: 7.372e-02]
Xiang_wan_2_2_lora:   0%|                                           | 61/30000 [09:03<55:43:32,  6.70s/it, lr: 1.0e-04 loss: 1.004e-01]
Xiang_wan_2_2_lora:   0%|                                           | 61/30000 [09:03<55:43:32,  6.70s/it, lr: 1.0e-04 loss: 1.004e-01]
Xiang_wan_2_2_lora:   0%|                                           | 62/30000 [09:03<61:14:48,  7.36s/it, lr: 1.0e-04 loss: 1.004e-01]
Xiang_wan_2_2_lora:   0%|                                           | 62/30000 [09:03<61:14:48,  7.36s/it, lr: 1.0e-04 loss: 1.004e-01]
Xiang_wan_2_2_lora:   0%|                                           | 62/30000 [09:12<61:14:48,  7.36s/it, lr: 1.0e-04 loss: 3.678e-02]
Xiang_wan_2_2_lora:   0%|                                           | 62/30000 [09:12<61:14:48,  7.36s/it, lr: 1.0e-04 loss: 3.678e-02]
Xiang_wan_2_2_lora:   0%|                                           | 63/30000 [09:12<65:43:47,  7.90s/it, lr: 1.0e-04 loss: 3.678e-02]
Xiang_wan_2_2_lora:   0%|                                           | 63/30000 [09:12<65:43:47,  7.90s/it, lr: 1.0e-04 loss: 3.678e-02]
Xiang_wan_2_2_lora:   0%|                                           | 63/30000 [09:22<65:43:47,  7.90s/it, lr: 1.0e-04 loss: 7.154e-02]
Xiang_wan_2_2_lora:   0%|                                           | 63/30000 [09:22<65:43:47,  7.90s/it, lr: 1.0e-04 loss: 7.154e-02]
Xiang_wan_2_2_lora:   0%|                                           | 64/30000 [09:22<70:38:40,  8.50s/it, lr: 1.0e-04 loss: 7.154e-02]
Xiang_wan_2_2_lora:   0%|                                           | 64/30000 [09:22<70:38:40,  8.50s/it, lr: 1.0e-04 loss: 7.154e-02]
Xiang_wan_2_2_lora:   0%|                                           | 64/30000 [09:32<70:38:40,  8.50s/it, lr: 1.0e-04 loss: 4.639e-02]
Xiang_wan_2_2_lora:   0%|                                           | 64/30000 [09:32<70:38:40,  8.50s/it, lr: 1.0e-04 loss: 4.639e-02]
Xiang_wan_2_2_lora:   0%|                                           | 65/30000 [09:32<72:52:47,  8.76s/it, lr: 1.0e-04 loss: 4.639e-02]
Xiang_wan_2_2_lora:   0%|                                           | 65/30000 [09:32<72:52:47,  8.76s/it, lr: 1.0e-04 loss: 4.639e-02]
Xiang_wan_2_2_lora:   0%|                                           | 65/30000 [09:41<72:52:47,  8.76s/it, lr: 1.0e-04 loss: 1.128e-01]
Xiang_wan_2_2_lora:   0%|                                           | 65/30000 [09:41<72:52:47,  8.76s/it, lr: 1.0e-04 loss: 1.128e-01]
Xiang_wan_2_2_lora:   0%|                                           | 66/30000 [09:41<74:28:01,  8.96s/it, lr: 1.0e-04 loss: 1.128e-01]
Xiang_wan_2_2_lora:   0%|                                           | 66/30000 [09:41<74:28:01,  8.96s/it, lr: 1.0e-04 loss: 1.128e-01]
Xiang_wan_2_2_lora:   0%|                                           | 66/30000 [09:48<74:28:01,  8.96s/it, lr: 1.0e-04 loss: 7.956e-02]
Xiang_wan_2_2_lora:   0%|                                           | 66/30000 [09:48<74:28:01,  8.96s/it, lr: 1.0e-04 loss: 7.956e-02]
Xiang_wan_2_2_lora:   0%|                                           | 67/30000 [09:48<69:00:50,  8.30s/it, lr: 1.0e-04 loss: 7.956e-02]
Xiang_wan_2_2_lora:   0%|                                           | 67/30000 [09:48<69:00:50,  8.30s/it, lr: 1.0e-04 loss: 7.956e-02]
Xiang_wan_2_2_lora:   0%|                                           | 67/30000 [09:54<69:00:50,  8.30s/it, lr: 1.0e-04 loss: 1.358e-01]
Xiang_wan_2_2_lora:   0%|                                           | 67/30000 [09:54<69:00:50,  8.30s/it, lr: 1.0e-04 loss: 1.358e-01]
Xiang_wan_2_2_lora:   0%|                                           | 68/30000 [09:54<64:57:14,  7.81s/it, lr: 1.0e-04 loss: 1.358e-01]
Xiang_wan_2_2_lora:   0%|                                           | 68/30000 [09:54<64:57:14,  7.81s/it, lr: 1.0e-04 loss: 1.358e-01]
Xiang_wan_2_2_lora:   0%|                                           | 68/30000 [10:01<64:57:14,  7.81s/it, lr: 1.0e-04 loss: 6.264e-01]
Xiang_wan_2_2_lora:   0%|                                           | 68/30000 [10:01<64:57:14,  7.81s/it, lr: 1.0e-04 loss: 6.264e-01]
Xiang_wan_2_2_lora:   0%|                                           | 69/30000 [10:01<62:02:33,  7.46s/it, lr: 1.0e-04 loss: 6.264e-01]
Xiang_wan_2_2_lora:   0%|                                           | 69/30000 [10:01<62:02:33,  7.46s/it, lr: 1.0e-04 loss: 6.264e-01]
Xiang_wan_2_2_lora:   0%|                                           | 69/30000 [10:08<62:02:33,  7.46s/it, lr: 1.0e-04 loss: 1.878e-01]
Xiang_wan_2_2_lora:   0%|                                           | 69/30000 [10:08<62:02:33,  7.46s/it, lr: 1.0e-04 loss: 1.878e-01]
Xiang_wan_2_2_lora:   0%|1                                          | 70/30000 [10:15<62:02:25,  7.46s/it, lr: 1.0e-04 loss: 9.264e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 70/30000 [10:15<62:02:25,  7.46s/it, lr: 1.0e-04 loss: 9.264e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 71/30000 [10:15<46:42:22,  5.62s/it, lr: 1.0e-04 loss: 9.264e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 71/30000 [10:15<46:42:22,  5.62s/it, lr: 1.0e-04 loss: 9.264e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 71/30000 [10:21<46:42:22,  5.62s/it, lr: 1.0e-04 loss: 1.235e-01]
Xiang_wan_2_2_lora:   0%|1                                          | 71/30000 [10:21<46:42:22,  5.62s/it, lr: 1.0e-04 loss: 1.235e-01]
Xiang_wan_2_2_lora:   0%|1                                          | 72/30000 [10:21<48:45:38,  5.87s/it, lr: 1.0e-04 loss: 1.235e-01]
Xiang_wan_2_2_lora:   0%|1                                          | 72/30000 [10:21<48:45:38,  5.87s/it, lr: 1.0e-04 loss: 1.235e-01]
Xiang_wan_2_2_lora:   0%|1                                          | 72/30000 [10:28<48:45:38,  5.87s/it, lr: 1.0e-04 loss: 1.256e-01]
Xiang_wan_2_2_lora:   0%|1                                          | 72/30000 [10:28<48:45:38,  5.87s/it, lr: 1.0e-04 loss: 1.256e-01]
Xiang_wan_2_2_lora:   0%|1                                          | 73/30000 [10:28<50:28:31,  6.07s/it, lr: 1.0e-04 loss: 1.256e-01]
Xiang_wan_2_2_lora:   0%|1                                          | 73/30000 [10:28<50:28:31,  6.07s/it, lr: 1.0e-04 loss: 1.256e-01]
Xiang_wan_2_2_lora:   0%|1                                          | 73/30000 [10:35<50:28:31,  6.07s/it, lr: 1.0e-04 loss: 8.755e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 73/30000 [10:35<50:28:31,  6.07s/it, lr: 1.0e-04 loss: 8.755e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 74/30000 [10:35<51:48:48,  6.23s/it, lr: 1.0e-04 loss: 8.755e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 74/30000 [10:35<51:48:48,  6.23s/it, lr: 1.0e-04 loss: 8.755e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 74/30000 [10:44<51:48:48,  6.23s/it, lr: 1.0e-04 loss: 4.560e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 74/30000 [10:44<51:48:48,  6.23s/it, lr: 1.0e-04 loss: 4.560e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 75/30000 [10:44<59:30:44,  7.16s/it, lr: 1.0e-04 loss: 4.560e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 75/30000 [10:44<59:30:44,  7.16s/it, lr: 1.0e-04 loss: 4.560e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 75/30000 [10:51<59:30:44,  7.16s/it, lr: 1.0e-04 loss: 1.673e-01]
Xiang_wan_2_2_lora:   0%|1                                          | 75/30000 [10:51<59:30:44,  7.16s/it, lr: 1.0e-04 loss: 1.673e-01]
Xiang_wan_2_2_lora:   0%|1                                          | 76/30000 [10:51<58:25:44,  7.03s/it, lr: 1.0e-04 loss: 1.673e-01]
Xiang_wan_2_2_lora:   0%|1                                          | 76/30000 [10:51<58:25:44,  7.03s/it, lr: 1.0e-04 loss: 1.673e-01]
Xiang_wan_2_2_lora:   0%|1                                          | 76/30000 [11:00<58:25:44,  7.03s/it, lr: 1.0e-04 loss: 2.610e-01]
Xiang_wan_2_2_lora:   0%|1                                          | 76/30000 [11:00<58:25:44,  7.03s/it, lr: 1.0e-04 loss: 2.610e-01]
Xiang_wan_2_2_lora:   0%|1                                          | 77/30000 [11:00<64:10:43,  7.72s/it, lr: 1.0e-04 loss: 2.610e-01]
Xiang_wan_2_2_lora:   0%|1                                          | 77/30000 [11:00<64:10:43,  7.72s/it, lr: 1.0e-04 loss: 2.610e-01]
Xiang_wan_2_2_lora:   0%|1                                          | 77/30000 [11:10<64:10:43,  7.72s/it, lr: 1.0e-04 loss: 9.732e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 77/30000 [11:10<64:10:43,  7.72s/it, lr: 1.0e-04 loss: 9.732e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 78/30000 [11:10<68:49:07,  8.28s/it, lr: 1.0e-04 loss: 9.732e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 78/30000 [11:10<68:49:07,  8.28s/it, lr: 1.0e-04 loss: 9.732e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 78/30000 [11:19<68:49:07,  8.28s/it, lr: 1.0e-04 loss: 5.117e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 78/30000 [11:19<68:49:07,  8.28s/it, lr: 1.0e-04 loss: 5.117e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 79/30000 [11:19<71:39:48,  8.62s/it, lr: 1.0e-04 loss: 5.117e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 79/30000 [11:19<71:39:48,  8.62s/it, lr: 1.0e-04 loss: 5.117e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 79/30000 [11:26<71:39:48,  8.62s/it, lr: 1.0e-04 loss: 1.322e-01]
Xiang_wan_2_2_lora:   0%|1                                          | 79/30000 [11:26<71:39:48,  8.62s/it, lr: 1.0e-04 loss: 1.322e-01]
Xiang_wan_2_2_lora:   0%|1                                          | 80/30000 [11:33<71:39:39,  8.62s/it, lr: 1.0e-04 loss: 9.257e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 80/30000 [11:33<71:39:39,  8.62s/it, lr: 1.0e-04 loss: 9.257e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 81/30000 [11:33<52:35:51,  6.33s/it, lr: 1.0e-04 loss: 9.257e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 81/30000 [11:33<52:35:51,  6.33s/it, lr: 1.0e-04 loss: 9.257e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 81/30000 [11:43<52:35:51,  6.33s/it, lr: 1.0e-04 loss: 1.067e-01]
Xiang_wan_2_2_lora:   0%|1                                          | 81/30000 [11:43<52:35:51,  6.33s/it, lr: 1.0e-04 loss: 1.067e-01]
Xiang_wan_2_2_lora:   0%|1                                          | 82/30000 [11:43<59:01:15,  7.10s/it, lr: 1.0e-04 loss: 1.067e-01]
Xiang_wan_2_2_lora:   0%|1                                          | 82/30000 [11:43<59:01:15,  7.10s/it, lr: 1.0e-04 loss: 1.067e-01]
Xiang_wan_2_2_lora:   0%|1                                          | 82/30000 [11:50<59:01:15,  7.10s/it, lr: 1.0e-04 loss: 9.901e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 82/30000 [11:50<59:01:15,  7.10s/it, lr: 1.0e-04 loss: 9.901e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 83/30000 [11:50<58:56:36,  7.09s/it, lr: 1.0e-04 loss: 9.901e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 83/30000 [11:50<58:56:36,  7.09s/it, lr: 1.0e-04 loss: 9.901e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 83/30000 [11:59<58:56:36,  7.09s/it, lr: 1.0e-04 loss: 8.821e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 83/30000 [11:59<58:56:36,  7.09s/it, lr: 1.0e-04 loss: 8.821e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 84/30000 [11:59<64:04:20,  7.71s/it, lr: 1.0e-04 loss: 8.821e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 84/30000 [11:59<64:04:20,  7.71s/it, lr: 1.0e-04 loss: 8.821e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 84/30000 [12:06<64:04:20,  7.71s/it, lr: 1.0e-04 loss: 6.655e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 84/30000 [12:06<64:04:20,  7.71s/it, lr: 1.0e-04 loss: 6.655e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 85/30000 [12:06<62:21:09,  7.50s/it, lr: 1.0e-04 loss: 6.655e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 85/30000 [12:06<62:21:09,  7.50s/it, lr: 1.0e-04 loss: 6.655e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 85/30000 [12:13<62:21:09,  7.50s/it, lr: 1.0e-04 loss: 1.104e-01]
Xiang_wan_2_2_lora:   0%|1                                          | 85/30000 [12:13<62:21:09,  7.50s/it, lr: 1.0e-04 loss: 1.104e-01]
Xiang_wan_2_2_lora:   0%|1                                          | 86/30000 [12:13<60:22:01,  7.26s/it, lr: 1.0e-04 loss: 1.104e-01]
Xiang_wan_2_2_lora:   0%|1                                          | 86/30000 [12:13<60:22:01,  7.26s/it, lr: 1.0e-04 loss: 1.104e-01]
Xiang_wan_2_2_lora:   0%|1                                          | 86/30000 [12:22<60:22:01,  7.26s/it, lr: 1.0e-04 loss: 5.741e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 86/30000 [12:22<60:22:01,  7.26s/it, lr: 1.0e-04 loss: 5.741e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 87/30000 [12:22<65:46:57,  7.92s/it, lr: 1.0e-04 loss: 5.741e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 87/30000 [12:22<65:46:57,  7.92s/it, lr: 1.0e-04 loss: 5.741e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 87/30000 [12:29<65:46:57,  7.92s/it, lr: 1.0e-04 loss: 1.279e-01]
Xiang_wan_2_2_lora:   0%|1                                          | 87/30000 [12:29<65:46:57,  7.92s/it, lr: 1.0e-04 loss: 1.279e-01]
Xiang_wan_2_2_lora:   0%|1                                          | 88/30000 [12:29<63:44:45,  7.67s/it, lr: 1.0e-04 loss: 1.279e-01]
Xiang_wan_2_2_lora:   0%|1                                          | 88/30000 [12:29<63:44:45,  7.67s/it, lr: 1.0e-04 loss: 1.279e-01]
Xiang_wan_2_2_lora:   0%|1                                          | 88/30000 [12:36<63:44:45,  7.67s/it, lr: 1.0e-04 loss: 8.218e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 88/30000 [12:36<63:44:45,  7.67s/it, lr: 1.0e-04 loss: 8.218e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 89/30000 [12:36<61:34:52,  7.41s/it, lr: 1.0e-04 loss: 8.218e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 89/30000 [12:36<61:34:52,  7.41s/it, lr: 1.0e-04 loss: 8.218e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 89/30000 [12:43<61:34:52,  7.41s/it, lr: 1.0e-04 loss: 1.696e-01]
Xiang_wan_2_2_lora:   0%|1                                          | 89/30000 [12:43<61:34:52,  7.41s/it, lr: 1.0e-04 loss: 1.696e-01]
Xiang_wan_2_2_lora:   0%|1                                          | 90/30000 [12:52<61:34:45,  7.41s/it, lr: 1.0e-04 loss: 5.662e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 90/30000 [12:52<61:34:45,  7.41s/it, lr: 1.0e-04 loss: 5.662e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 91/30000 [12:52<51:20:16,  6.18s/it, lr: 1.0e-04 loss: 5.662e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 91/30000 [12:52<51:20:16,  6.18s/it, lr: 1.0e-04 loss: 5.662e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 91/30000 [13:02<51:20:16,  6.18s/it, lr: 1.0e-04 loss: 7.027e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 91/30000 [13:02<51:20:16,  6.18s/it, lr: 1.0e-04 loss: 7.027e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 92/30000 [13:02<59:08:30,  7.12s/it, lr: 1.0e-04 loss: 7.027e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 92/30000 [13:02<59:08:30,  7.12s/it, lr: 1.0e-04 loss: 7.027e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 92/30000 [13:12<59:08:30,  7.12s/it, lr: 1.0e-04 loss: 1.462e-01]
Xiang_wan_2_2_lora:   0%|1                                          | 92/30000 [13:12<59:08:30,  7.12s/it, lr: 1.0e-04 loss: 1.462e-01]
Xiang_wan_2_2_lora:   0%|1                                          | 93/30000 [13:12<65:13:53,  7.85s/it, lr: 1.0e-04 loss: 1.462e-01]
Xiang_wan_2_2_lora:   0%|1                                          | 93/30000 [13:12<65:13:53,  7.85s/it, lr: 1.0e-04 loss: 1.462e-01]
Xiang_wan_2_2_lora:   0%|1                                          | 93/30000 [13:19<65:13:53,  7.85s/it, lr: 1.0e-04 loss: 9.835e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 93/30000 [13:19<65:13:53,  7.85s/it, lr: 1.0e-04 loss: 9.835e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 94/30000 [13:19<63:53:52,  7.69s/it, lr: 1.0e-04 loss: 9.835e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 94/30000 [13:19<63:53:52,  7.69s/it, lr: 1.0e-04 loss: 9.835e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 94/30000 [13:29<63:53:52,  7.69s/it, lr: 1.0e-04 loss: 4.862e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 94/30000 [13:29<63:53:52,  7.69s/it, lr: 1.0e-04 loss: 4.862e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 95/30000 [13:29<68:31:56,  8.25s/it, lr: 1.0e-04 loss: 4.862e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 95/30000 [13:29<68:31:56,  8.25s/it, lr: 1.0e-04 loss: 4.862e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 95/30000 [13:36<68:31:56,  8.25s/it, lr: 1.0e-04 loss: 1.086e-01]
Xiang_wan_2_2_lora:   0%|1                                          | 95/30000 [13:36<68:31:56,  8.25s/it, lr: 1.0e-04 loss: 1.086e-01]
Xiang_wan_2_2_lora:   0%|1                                          | 96/30000 [13:36<65:34:07,  7.89s/it, lr: 1.0e-04 loss: 1.086e-01]
Xiang_wan_2_2_lora:   0%|1                                          | 96/30000 [13:36<65:34:07,  7.89s/it, lr: 1.0e-04 loss: 1.086e-01]
Xiang_wan_2_2_lora:   0%|1                                          | 96/30000 [13:43<65:34:07,  7.89s/it, lr: 1.0e-04 loss: 8.065e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 96/30000 [13:43<65:34:07,  7.89s/it, lr: 1.0e-04 loss: 8.065e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 97/30000 [13:43<63:00:29,  7.59s/it, lr: 1.0e-04 loss: 8.065e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 97/30000 [13:43<63:00:29,  7.59s/it, lr: 1.0e-04 loss: 8.065e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 97/30000 [13:53<63:00:29,  7.59s/it, lr: 1.0e-04 loss: 8.114e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 97/30000 [13:53<63:00:29,  7.59s/it, lr: 1.0e-04 loss: 8.114e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 98/30000 [13:53<69:13:58,  8.34s/it, lr: 1.0e-04 loss: 8.114e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 98/30000 [13:53<69:13:58,  8.34s/it, lr: 1.0e-04 loss: 8.114e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 98/30000 [14:03<69:13:58,  8.34s/it, lr: 1.0e-04 loss: 8.892e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 98/30000 [14:03<69:13:58,  8.34s/it, lr: 1.0e-04 loss: 8.892e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 99/30000 [14:03<73:50:22,  8.89s/it, lr: 1.0e-04 loss: 8.892e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 99/30000 [14:03<73:50:22,  8.89s/it, lr: 1.0e-04 loss: 8.892e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 99/30000 [14:10<73:50:22,  8.89s/it, lr: 1.0e-04 loss: 8.893e-02]
Xiang_wan_2_2_lora:   0%|1                                          | 99/30000 [14:10<73:50:22,  8.89s/it, lr: 1.0e-04 loss: 8.893e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 100/30000 [14:17<73:50:13,  8.89s/it, lr: 1.0e-04 loss: 1.002e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 100/30000 [14:17<73:50:13,  8.89s/it, lr: 1.0e-04 loss: 1.002e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 101/30000 [14:17<53:03:17,  6.39s/it, lr: 1.0e-04 loss: 1.002e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 101/30000 [14:17<53:03:17,  6.39s/it, lr: 1.0e-04 loss: 1.002e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 101/30000 [14:27<53:03:17,  6.39s/it, lr: 1.0e-04 loss: 8.951e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 101/30000 [14:27<53:03:17,  6.39s/it, lr: 1.0e-04 loss: 8.951e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 102/30000 [14:27<59:15:42,  7.14s/it, lr: 1.0e-04 loss: 8.951e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 102/30000 [14:27<59:15:42,  7.14s/it, lr: 1.0e-04 loss: 8.951e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 102/30000 [14:36<59:15:42,  7.14s/it, lr: 1.0e-04 loss: 3.826e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 102/30000 [14:36<59:15:42,  7.14s/it, lr: 1.0e-04 loss: 3.826e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 103/30000 [14:36<64:37:43,  7.78s/it, lr: 1.0e-04 loss: 3.826e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 103/30000 [14:36<64:37:43,  7.78s/it, lr: 1.0e-04 loss: 3.826e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 103/30000 [14:46<64:37:43,  7.78s/it, lr: 1.0e-04 loss: 8.573e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 103/30000 [14:46<64:37:43,  7.78s/it, lr: 1.0e-04 loss: 8.573e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 104/30000 [14:46<69:59:20,  8.43s/it, lr: 1.0e-04 loss: 8.573e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 104/30000 [14:46<69:59:20,  8.43s/it, lr: 1.0e-04 loss: 8.573e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 104/30000 [14:54<69:59:20,  8.43s/it, lr: 1.0e-04 loss: 9.331e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 104/30000 [14:54<69:59:20,  8.43s/it, lr: 1.0e-04 loss: 9.331e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 105/30000 [14:54<67:19:42,  8.11s/it, lr: 1.0e-04 loss: 9.331e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 105/30000 [14:54<67:19:42,  8.11s/it, lr: 1.0e-04 loss: 9.331e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 105/30000 [15:04<67:19:42,  8.11s/it, lr: 1.0e-04 loss: 1.573e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 105/30000 [15:04<67:19:42,  8.11s/it, lr: 1.0e-04 loss: 1.573e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 106/30000 [15:04<72:50:59,  8.77s/it, lr: 1.0e-04 loss: 1.573e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 106/30000 [15:04<72:50:59,  8.77s/it, lr: 1.0e-04 loss: 1.573e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 106/30000 [15:11<72:50:59,  8.77s/it, lr: 1.0e-04 loss: 6.357e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 106/30000 [15:11<72:50:59,  8.77s/it, lr: 1.0e-04 loss: 6.357e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 107/30000 [15:11<68:56:16,  8.30s/it, lr: 1.0e-04 loss: 6.357e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 107/30000 [15:11<68:56:16,  8.30s/it, lr: 1.0e-04 loss: 6.357e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 107/30000 [15:21<68:56:16,  8.30s/it, lr: 1.0e-04 loss: 9.675e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 107/30000 [15:21<68:56:16,  8.30s/it, lr: 1.0e-04 loss: 9.675e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 108/30000 [15:21<72:28:20,  8.73s/it, lr: 1.0e-04 loss: 9.675e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 108/30000 [15:21<72:28:20,  8.73s/it, lr: 1.0e-04 loss: 9.675e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 108/30000 [15:31<72:28:20,  8.73s/it, lr: 1.0e-04 loss: 4.894e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 108/30000 [15:31<72:28:20,  8.73s/it, lr: 1.0e-04 loss: 4.894e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 109/30000 [15:31<75:13:54,  9.06s/it, lr: 1.0e-04 loss: 4.894e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 109/30000 [15:31<75:13:54,  9.06s/it, lr: 1.0e-04 loss: 4.894e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 109/30000 [15:39<75:13:54,  9.06s/it, lr: 1.0e-04 loss: 1.002e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 109/30000 [15:39<75:13:54,  9.06s/it, lr: 1.0e-04 loss: 1.002e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 110/30000 [15:46<75:13:45,  9.06s/it, lr: 1.0e-04 loss: 6.120e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 110/30000 [15:46<75:13:45,  9.06s/it, lr: 1.0e-04 loss: 6.120e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 111/30000 [15:46<54:36:07,  6.58s/it, lr: 1.0e-04 loss: 6.120e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 111/30000 [15:46<54:36:07,  6.58s/it, lr: 1.0e-04 loss: 6.120e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 111/30000 [15:55<54:36:07,  6.58s/it, lr: 1.0e-04 loss: 1.131e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 111/30000 [15:55<54:36:07,  6.58s/it, lr: 1.0e-04 loss: 1.131e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 112/30000 [15:55<60:39:09,  7.31s/it, lr: 1.0e-04 loss: 1.131e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 112/30000 [15:55<60:39:09,  7.31s/it, lr: 1.0e-04 loss: 1.131e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 112/30000 [16:03<60:39:09,  7.31s/it, lr: 1.0e-04 loss: 5.742e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 112/30000 [16:03<60:39:09,  7.31s/it, lr: 1.0e-04 loss: 5.742e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 113/30000 [16:03<60:45:50,  7.32s/it, lr: 1.0e-04 loss: 5.742e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 113/30000 [16:03<60:45:50,  7.32s/it, lr: 1.0e-04 loss: 5.742e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 113/30000 [16:12<60:45:50,  7.32s/it, lr: 1.0e-04 loss: 8.264e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 113/30000 [16:12<60:45:50,  7.32s/it, lr: 1.0e-04 loss: 8.264e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 114/30000 [16:12<65:34:48,  7.90s/it, lr: 1.0e-04 loss: 8.264e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 114/30000 [16:12<65:34:48,  7.90s/it, lr: 1.0e-04 loss: 8.264e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 114/30000 [16:19<65:34:48,  7.90s/it, lr: 1.0e-04 loss: 1.016e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 114/30000 [16:19<65:34:48,  7.90s/it, lr: 1.0e-04 loss: 1.016e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 115/30000 [16:19<62:51:18,  7.57s/it, lr: 1.0e-04 loss: 1.016e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 115/30000 [16:19<62:51:18,  7.57s/it, lr: 1.0e-04 loss: 1.016e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 115/30000 [16:28<62:51:18,  7.57s/it, lr: 1.0e-04 loss: 9.193e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 115/30000 [16:28<62:51:18,  7.57s/it, lr: 1.0e-04 loss: 9.193e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 116/30000 [16:28<67:09:19,  8.09s/it, lr: 1.0e-04 loss: 9.193e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 116/30000 [16:28<67:09:19,  8.09s/it, lr: 1.0e-04 loss: 9.193e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 116/30000 [16:35<67:09:19,  8.09s/it, lr: 1.0e-04 loss: 3.027e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 116/30000 [16:35<67:09:19,  8.09s/it, lr: 1.0e-04 loss: 3.027e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 117/30000 [16:35<64:31:19,  7.77s/it, lr: 1.0e-04 loss: 3.027e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 117/30000 [16:35<64:31:19,  7.77s/it, lr: 1.0e-04 loss: 3.027e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 117/30000 [16:42<64:31:19,  7.77s/it, lr: 1.0e-04 loss: 1.006e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 117/30000 [16:42<64:31:19,  7.77s/it, lr: 1.0e-04 loss: 1.006e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 118/30000 [16:42<62:10:53,  7.49s/it, lr: 1.0e-04 loss: 1.006e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 118/30000 [16:42<62:10:53,  7.49s/it, lr: 1.0e-04 loss: 1.006e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 118/30000 [16:49<62:10:53,  7.49s/it, lr: 1.0e-04 loss: 1.217e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 118/30000 [16:49<62:10:53,  7.49s/it, lr: 1.0e-04 loss: 1.217e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 119/30000 [16:49<60:17:41,  7.26s/it, lr: 1.0e-04 loss: 1.217e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 119/30000 [16:49<60:17:41,  7.26s/it, lr: 1.0e-04 loss: 1.217e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 119/30000 [16:58<60:17:41,  7.26s/it, lr: 1.0e-04 loss: 6.942e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 119/30000 [16:58<60:17:41,  7.26s/it, lr: 1.0e-04 loss: 6.942e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 120/30000 [17:08<60:17:34,  7.26s/it, lr: 1.0e-04 loss: 4.027e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 120/30000 [17:08<60:17:34,  7.26s/it, lr: 1.0e-04 loss: 4.027e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 121/30000 [17:08<51:02:38,  6.15s/it, lr: 1.0e-04 loss: 4.027e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 121/30000 [17:08<51:02:38,  6.15s/it, lr: 1.0e-04 loss: 4.027e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 121/30000 [17:18<51:02:38,  6.15s/it, lr: 1.0e-04 loss: 7.272e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 121/30000 [17:18<51:02:38,  6.15s/it, lr: 1.0e-04 loss: 7.272e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 122/30000 [17:18<59:45:35,  7.20s/it, lr: 1.0e-04 loss: 7.272e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 122/30000 [17:18<59:45:35,  7.20s/it, lr: 1.0e-04 loss: 7.272e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 122/30000 [17:28<59:45:35,  7.20s/it, lr: 1.0e-04 loss: 1.336e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 122/30000 [17:28<59:45:35,  7.20s/it, lr: 1.0e-04 loss: 1.336e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 123/30000 [17:28<65:21:28,  7.88s/it, lr: 1.0e-04 loss: 1.336e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 123/30000 [17:28<65:21:28,  7.88s/it, lr: 1.0e-04 loss: 1.336e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 123/30000 [17:38<65:21:28,  7.88s/it, lr: 1.0e-04 loss: 1.010e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 123/30000 [17:38<65:21:28,  7.88s/it, lr: 1.0e-04 loss: 1.010e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 124/30000 [17:38<68:52:29,  8.30s/it, lr: 1.0e-04 loss: 1.010e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 124/30000 [17:38<68:52:29,  8.30s/it, lr: 1.0e-04 loss: 1.010e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 124/30000 [17:44<68:52:29,  8.30s/it, lr: 1.0e-04 loss: 1.120e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 124/30000 [17:44<68:52:29,  8.30s/it, lr: 1.0e-04 loss: 1.120e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 125/30000 [17:44<65:16:26,  7.87s/it, lr: 1.0e-04 loss: 1.120e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 125/30000 [17:44<65:16:26,  7.87s/it, lr: 1.0e-04 loss: 1.120e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 125/30000 [17:54<65:16:26,  7.87s/it, lr: 1.0e-04 loss: 1.048e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 125/30000 [17:54<65:16:26,  7.87s/it, lr: 1.0e-04 loss: 1.048e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 126/30000 [17:54<68:51:42,  8.30s/it, lr: 1.0e-04 loss: 1.048e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 126/30000 [17:54<68:51:42,  8.30s/it, lr: 1.0e-04 loss: 1.048e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 126/30000 [18:04<68:51:42,  8.30s/it, lr: 1.0e-04 loss: 8.676e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 126/30000 [18:04<68:51:42,  8.30s/it, lr: 1.0e-04 loss: 8.676e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 127/30000 [18:04<72:45:21,  8.77s/it, lr: 1.0e-04 loss: 8.676e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 127/30000 [18:04<72:45:21,  8.77s/it, lr: 1.0e-04 loss: 8.676e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 127/30000 [18:14<72:45:21,  8.77s/it, lr: 1.0e-04 loss: 8.820e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 127/30000 [18:14<72:45:21,  8.77s/it, lr: 1.0e-04 loss: 8.820e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 128/30000 [18:14<75:20:47,  9.08s/it, lr: 1.0e-04 loss: 8.820e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 128/30000 [18:14<75:20:47,  9.08s/it, lr: 1.0e-04 loss: 8.820e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 128/30000 [18:23<75:20:47,  9.08s/it, lr: 1.0e-04 loss: 2.616e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 128/30000 [18:23<75:20:47,  9.08s/it, lr: 1.0e-04 loss: 2.616e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 129/30000 [18:23<76:34:31,  9.23s/it, lr: 1.0e-04 loss: 2.616e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 129/30000 [18:23<76:34:31,  9.23s/it, lr: 1.0e-04 loss: 2.616e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 129/30000 [18:31<76:34:31,  9.23s/it, lr: 1.0e-04 loss: 1.182e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 129/30000 [18:31<76:34:31,  9.23s/it, lr: 1.0e-04 loss: 1.182e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 130/30000 [18:41<76:34:22,  9.23s/it, lr: 1.0e-04 loss: 5.812e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 130/30000 [18:41<76:34:22,  9.23s/it, lr: 1.0e-04 loss: 5.812e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 131/30000 [18:41<60:46:38,  7.33s/it, lr: 1.0e-04 loss: 5.812e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 131/30000 [18:41<60:46:38,  7.33s/it, lr: 1.0e-04 loss: 5.812e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 131/30000 [18:48<60:46:38,  7.33s/it, lr: 1.0e-04 loss: 1.101e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 131/30000 [18:48<60:46:38,  7.33s/it, lr: 1.0e-04 loss: 1.101e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 132/30000 [18:48<61:14:59,  7.38s/it, lr: 1.0e-04 loss: 1.101e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 132/30000 [18:48<61:14:59,  7.38s/it, lr: 1.0e-04 loss: 1.101e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 132/30000 [18:56<61:14:59,  7.38s/it, lr: 1.0e-04 loss: 5.937e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 132/30000 [18:56<61:14:59,  7.38s/it, lr: 1.0e-04 loss: 5.937e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 133/30000 [18:56<61:02:43,  7.36s/it, lr: 1.0e-04 loss: 5.937e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 133/30000 [18:56<61:02:43,  7.36s/it, lr: 1.0e-04 loss: 5.937e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 133/30000 [19:06<61:02:43,  7.36s/it, lr: 1.0e-04 loss: 9.121e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 133/30000 [19:06<61:02:43,  7.36s/it, lr: 1.0e-04 loss: 9.121e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 134/30000 [19:06<67:55:56,  8.19s/it, lr: 1.0e-04 loss: 9.121e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 134/30000 [19:06<67:55:56,  8.19s/it, lr: 1.0e-04 loss: 9.121e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 134/30000 [19:13<67:55:56,  8.19s/it, lr: 1.0e-04 loss: 8.193e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 134/30000 [19:13<67:55:56,  8.19s/it, lr: 1.0e-04 loss: 8.193e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 135/30000 [19:13<64:53:33,  7.82s/it, lr: 1.0e-04 loss: 8.193e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 135/30000 [19:13<64:53:33,  7.82s/it, lr: 1.0e-04 loss: 8.193e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 135/30000 [19:20<64:53:33,  7.82s/it, lr: 1.0e-04 loss: 9.837e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 135/30000 [19:20<64:53:33,  7.82s/it, lr: 1.0e-04 loss: 9.837e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 136/30000 [19:20<62:22:14,  7.52s/it, lr: 1.0e-04 loss: 9.837e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 136/30000 [19:20<62:22:14,  7.52s/it, lr: 1.0e-04 loss: 9.837e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 136/30000 [19:27<62:22:14,  7.52s/it, lr: 1.0e-04 loss: 1.158e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 136/30000 [19:27<62:22:14,  7.52s/it, lr: 1.0e-04 loss: 1.158e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 137/30000 [19:27<62:01:40,  7.48s/it, lr: 1.0e-04 loss: 1.158e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 137/30000 [19:27<62:01:40,  7.48s/it, lr: 1.0e-04 loss: 1.158e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 137/30000 [19:37<62:01:40,  7.48s/it, lr: 1.0e-04 loss: 8.062e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 137/30000 [19:37<62:01:40,  7.48s/it, lr: 1.0e-04 loss: 8.062e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 138/30000 [19:37<68:42:26,  8.28s/it, lr: 1.0e-04 loss: 8.062e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 138/30000 [19:37<68:42:26,  8.28s/it, lr: 1.0e-04 loss: 8.062e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 138/30000 [19:44<68:42:26,  8.28s/it, lr: 1.0e-04 loss: 9.397e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 138/30000 [19:44<68:42:26,  8.28s/it, lr: 1.0e-04 loss: 9.397e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 139/30000 [19:44<65:22:10,  7.88s/it, lr: 1.0e-04 loss: 9.397e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 139/30000 [19:44<65:22:10,  7.88s/it, lr: 1.0e-04 loss: 9.397e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 139/30000 [19:51<65:22:10,  7.88s/it, lr: 1.0e-04 loss: 1.071e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 139/30000 [19:51<65:22:10,  7.88s/it, lr: 1.0e-04 loss: 1.071e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 140/30000 [20:01<65:22:02,  7.88s/it, lr: 1.0e-04 loss: 1.769e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 140/30000 [20:01<65:22:02,  7.88s/it, lr: 1.0e-04 loss: 1.769e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 141/30000 [20:01<54:35:42,  6.58s/it, lr: 1.0e-04 loss: 1.769e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 141/30000 [20:01<54:35:42,  6.58s/it, lr: 1.0e-04 loss: 1.769e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 141/30000 [20:09<54:35:42,  6.58s/it, lr: 1.0e-04 loss: 1.038e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 141/30000 [20:09<54:35:42,  6.58s/it, lr: 1.0e-04 loss: 1.038e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 142/30000 [20:09<56:22:16,  6.80s/it, lr: 1.0e-04 loss: 1.038e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 142/30000 [20:09<56:22:16,  6.80s/it, lr: 1.0e-04 loss: 1.038e-01]
Xiang_wan_2_2_lora:   0%|1                                         | 142/30000 [20:19<56:22:16,  6.80s/it, lr: 1.0e-04 loss: 3.923e-02]
Xiang_wan_2_2_lora:   0%|1                                         | 142/30000 [20:19<56:22:16,  6.80s/it, lr: 1.0e-04 loss: 3.923e-02]
Xiang_wan_2_2_lora:   0%|2                                         | 143/30000 [20:19<63:03:49,  7.60s/it, lr: 1.0e-04 loss: 3.923e-02]
Xiang_wan_2_2_lora:   0%|2                                         | 143/30000 [20:19<63:03:49,  7.60s/it, lr: 1.0e-04 loss: 3.923e-02]
Xiang_wan_2_2_lora:   0%|2                                         | 143/30000 [20:25<63:03:49,  7.60s/it, lr: 1.0e-04 loss: 1.003e-01]
Xiang_wan_2_2_lora:   0%|2                                         | 143/30000 [20:25<63:03:49,  7.60s/it, lr: 1.0e-04 loss: 1.003e-01]
Xiang_wan_2_2_lora:   0%|2                                         | 144/30000 [20:25<60:54:11,  7.34s/it, lr: 1.0e-04 loss: 1.003e-01]
Xiang_wan_2_2_lora:   0%|2                                         | 144/30000 [20:25<60:54:11,  7.34s/it, lr: 1.0e-04 loss: 1.003e-01]
Xiang_wan_2_2_lora:   0%|2                                         | 144/30000 [20:33<60:54:11,  7.34s/it, lr: 1.0e-04 loss: 7.049e-02]
Xiang_wan_2_2_lora:   0%|2                                         | 144/30000 [20:33<60:54:11,  7.34s/it, lr: 1.0e-04 loss: 7.049e-02]
Xiang_wan_2_2_lora:   0%|2                                         | 145/30000 [20:33<60:35:54,  7.31s/it, lr: 1.0e-04 loss: 7.049e-02]
Xiang_wan_2_2_lora:   0%|2                                         | 145/30000 [20:33<60:35:54,  7.31s/it, lr: 1.0e-04 loss: 7.049e-02]
Xiang_wan_2_2_lora:   0%|2                                         | 145/30000 [20:40<60:35:54,  7.31s/it, lr: 1.0e-04 loss: 1.175e-01]
Xiang_wan_2_2_lora:   0%|2                                         | 145/30000 [20:40<60:35:54,  7.31s/it, lr: 1.0e-04 loss: 1.175e-01]
Xiang_wan_2_2_lora:   0%|2                                         | 146/30000 [20:40<60:38:32,  7.31s/it, lr: 1.0e-04 loss: 1.175e-01]
Xiang_wan_2_2_lora:   0%|2                                         | 146/30000 [20:40<60:38:32,  7.31s/it, lr: 1.0e-04 loss: 1.175e-01]
Xiang_wan_2_2_lora:   0%|2                                         | 146/30000 [20:47<60:38:32,  7.31s/it, lr: 1.0e-04 loss: 1.006e-01]
Xiang_wan_2_2_lora:   0%|2                                         | 146/30000 [20:47<60:38:32,  7.31s/it, lr: 1.0e-04 loss: 1.006e-01]
Xiang_wan_2_2_lora:   0%|2                                         | 147/30000 [20:47<59:08:23,  7.13s/it, lr: 1.0e-04 loss: 1.006e-01]
Xiang_wan_2_2_lora:   0%|2                                         | 147/30000 [20:47<59:08:23,  7.13s/it, lr: 1.0e-04 loss: 1.006e-01]
Xiang_wan_2_2_lora:   0%|2                                         | 147/30000 [20:56<59:08:23,  7.13s/it, lr: 1.0e-04 loss: 8.595e-02]
Xiang_wan_2_2_lora:   0%|2                                         | 147/30000 [20:56<59:08:23,  7.13s/it, lr: 1.0e-04 loss: 8.595e-02]
Xiang_wan_2_2_lora:   0%|2                                         | 148/30000 [20:56<65:34:36,  7.91s/it, lr: 1.0e-04 loss: 8.595e-02]
Xiang_wan_2_2_lora:   0%|2                                         | 148/30000 [20:56<65:34:36,  7.91s/it, lr: 1.0e-04 loss: 8.595e-02]
Xiang_wan_2_2_lora:   0%|2                                         | 148/30000 [21:06<65:34:36,  7.91s/it, lr: 1.0e-04 loss: 4.259e-02]
Xiang_wan_2_2_lora:   0%|2                                         | 148/30000 [21:06<65:34:36,  7.91s/it, lr: 1.0e-04 loss: 4.259e-02]
Xiang_wan_2_2_lora:   0%|2                                         | 149/30000 [21:06<70:21:09,  8.48s/it, lr: 1.0e-04 loss: 4.259e-02]
Xiang_wan_2_2_lora:   0%|2                                         | 149/30000 [21:06<70:21:09,  8.48s/it, lr: 1.0e-04 loss: 4.259e-02]
Xiang_wan_2_2_lora:   0%|2                                         | 149/30000 [21:13<70:21:09,  8.48s/it, lr: 1.0e-04 loss: 8.310e-02]
Xiang_wan_2_2_lora:   0%|2                                         | 149/30000 [21:13<70:21:09,  8.48s/it, lr: 1.0e-04 loss: 8.310e-02]
Xiang_wan_2_2_lora:   0%|2                                         | 150/30000 [21:23<70:21:00,  8.48s/it, lr: 1.0e-04 loss: 4.410e-02]
Xiang_wan_2_2_lora:   0%|2                                         | 150/30000 [21:23<70:21:00,  8.48s/it, lr: 1.0e-04 loss: 4.410e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 151/30000 [21:23<56:49:08,  6.85s/it, lr: 1.0e-04 loss: 4.410e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 151/30000 [21:23<56:49:08,  6.85s/it, lr: 1.0e-04 loss: 4.410e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 151/30000 [21:33<56:49:08,  6.85s/it, lr: 1.0e-04 loss: 7.677e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 151/30000 [21:33<56:49:08,  6.85s/it, lr: 1.0e-04 loss: 7.677e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 152/30000 [21:33<63:33:10,  7.67s/it, lr: 1.0e-04 loss: 7.677e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 152/30000 [21:33<63:33:10,  7.67s/it, lr: 1.0e-04 loss: 7.677e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 152/30000 [21:40<63:33:10,  7.67s/it, lr: 1.0e-04 loss: 2.556e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 152/30000 [21:40<63:33:10,  7.67s/it, lr: 1.0e-04 loss: 2.556e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 153/30000 [21:40<61:26:49,  7.41s/it, lr: 1.0e-04 loss: 2.556e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 153/30000 [21:40<61:26:49,  7.41s/it, lr: 1.0e-04 loss: 2.556e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 153/30000 [21:47<61:26:49,  7.41s/it, lr: 1.0e-04 loss: 1.349e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 153/30000 [21:47<61:26:49,  7.41s/it, lr: 1.0e-04 loss: 1.349e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 154/30000 [21:47<61:12:50,  7.38s/it, lr: 1.0e-04 loss: 1.349e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 154/30000 [21:47<61:12:50,  7.38s/it, lr: 1.0e-04 loss: 1.349e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 154/30000 [21:57<61:12:50,  7.38s/it, lr: 1.0e-04 loss: 1.911e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 154/30000 [21:57<61:12:50,  7.38s/it, lr: 1.0e-04 loss: 1.911e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 155/30000 [21:57<67:16:27,  8.11s/it, lr: 1.0e-04 loss: 1.911e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 155/30000 [21:57<67:16:27,  8.11s/it, lr: 1.0e-04 loss: 1.911e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 155/30000 [22:04<67:16:27,  8.11s/it, lr: 1.0e-04 loss: 6.149e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 155/30000 [22:04<67:16:27,  8.11s/it, lr: 1.0e-04 loss: 6.149e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 156/30000 [22:04<65:38:51,  7.92s/it, lr: 1.0e-04 loss: 6.149e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 156/30000 [22:04<65:38:51,  7.92s/it, lr: 1.0e-04 loss: 6.149e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 156/30000 [22:11<65:38:51,  7.92s/it, lr: 1.0e-04 loss: 1.139e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 156/30000 [22:11<65:38:51,  7.92s/it, lr: 1.0e-04 loss: 1.139e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 157/30000 [22:11<63:07:09,  7.61s/it, lr: 1.0e-04 loss: 1.139e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 157/30000 [22:11<63:07:09,  7.61s/it, lr: 1.0e-04 loss: 1.139e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 157/30000 [22:21<63:07:09,  7.61s/it, lr: 1.0e-04 loss: 1.235e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 157/30000 [22:21<63:07:09,  7.61s/it, lr: 1.0e-04 loss: 1.235e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 158/30000 [22:21<67:28:34,  8.14s/it, lr: 1.0e-04 loss: 1.235e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 158/30000 [22:21<67:28:34,  8.14s/it, lr: 1.0e-04 loss: 1.235e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 158/30000 [22:30<67:28:34,  8.14s/it, lr: 1.0e-04 loss: 5.870e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 158/30000 [22:30<67:28:34,  8.14s/it, lr: 1.0e-04 loss: 5.870e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 159/30000 [22:30<71:12:26,  8.59s/it, lr: 1.0e-04 loss: 5.870e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 159/30000 [22:30<71:12:26,  8.59s/it, lr: 1.0e-04 loss: 5.870e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 159/30000 [22:40<71:12:26,  8.59s/it, lr: 1.0e-04 loss: 8.955e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 159/30000 [22:40<71:12:26,  8.59s/it, lr: 1.0e-04 loss: 8.955e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 160/30000 [22:47<71:12:17,  8.59s/it, lr: 1.0e-04 loss: 7.498e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 160/30000 [22:47<71:12:17,  8.59s/it, lr: 1.0e-04 loss: 7.498e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 161/30000 [22:47<52:06:55,  6.29s/it, lr: 1.0e-04 loss: 7.498e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 161/30000 [22:47<52:06:55,  6.29s/it, lr: 1.0e-04 loss: 7.498e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 161/30000 [22:57<52:06:55,  6.29s/it, lr: 1.0e-04 loss: 7.700e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 161/30000 [22:57<52:06:55,  6.29s/it, lr: 1.0e-04 loss: 7.700e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 162/30000 [22:57<60:06:18,  7.25s/it, lr: 1.0e-04 loss: 7.700e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 162/30000 [22:57<60:06:18,  7.25s/it, lr: 1.0e-04 loss: 7.700e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 162/30000 [23:05<60:06:18,  7.25s/it, lr: 1.0e-04 loss: 9.146e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 162/30000 [23:05<60:06:18,  7.25s/it, lr: 1.0e-04 loss: 9.146e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 163/30000 [23:05<59:55:55,  7.23s/it, lr: 1.0e-04 loss: 9.146e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 163/30000 [23:05<59:55:55,  7.23s/it, lr: 1.0e-04 loss: 9.146e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 163/30000 [23:11<59:55:55,  7.23s/it, lr: 1.0e-04 loss: 9.211e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 163/30000 [23:11<59:55:55,  7.23s/it, lr: 1.0e-04 loss: 9.211e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 164/30000 [23:11<58:40:37,  7.08s/it, lr: 1.0e-04 loss: 9.211e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 164/30000 [23:11<58:40:37,  7.08s/it, lr: 1.0e-04 loss: 9.211e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 164/30000 [23:18<58:40:37,  7.08s/it, lr: 1.0e-04 loss: 7.550e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 164/30000 [23:18<58:40:37,  7.08s/it, lr: 1.0e-04 loss: 7.550e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 165/30000 [23:18<57:33:30,  6.95s/it, lr: 1.0e-04 loss: 7.550e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 165/30000 [23:18<57:33:30,  6.95s/it, lr: 1.0e-04 loss: 7.550e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 165/30000 [23:24<57:33:30,  6.95s/it, lr: 1.0e-04 loss: 9.819e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 165/30000 [23:24<57:33:30,  6.95s/it, lr: 1.0e-04 loss: 9.819e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 166/30000 [23:24<56:41:22,  6.84s/it, lr: 1.0e-04 loss: 9.819e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 166/30000 [23:24<56:41:22,  6.84s/it, lr: 1.0e-04 loss: 9.819e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 166/30000 [23:34<56:41:22,  6.84s/it, lr: 1.0e-04 loss: 7.279e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 166/30000 [23:34<56:41:22,  6.84s/it, lr: 1.0e-04 loss: 7.279e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 167/30000 [23:34<64:22:55,  7.77s/it, lr: 1.0e-04 loss: 7.279e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 167/30000 [23:34<64:22:55,  7.77s/it, lr: 1.0e-04 loss: 7.279e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 167/30000 [23:44<64:22:55,  7.77s/it, lr: 1.0e-04 loss: 7.068e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 167/30000 [23:44<64:22:55,  7.77s/it, lr: 1.0e-04 loss: 7.068e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 168/30000 [23:44<68:24:55,  8.26s/it, lr: 1.0e-04 loss: 7.068e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 168/30000 [23:44<68:24:55,  8.26s/it, lr: 1.0e-04 loss: 7.068e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 168/30000 [23:54<68:24:55,  8.26s/it, lr: 1.0e-04 loss: 3.421e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 168/30000 [23:54<68:24:55,  8.26s/it, lr: 1.0e-04 loss: 3.421e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 169/30000 [23:54<72:40:28,  8.77s/it, lr: 1.0e-04 loss: 3.421e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 169/30000 [23:54<72:40:28,  8.77s/it, lr: 1.0e-04 loss: 3.421e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 169/30000 [24:04<72:40:28,  8.77s/it, lr: 1.0e-04 loss: 7.268e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 169/30000 [24:04<72:40:28,  8.77s/it, lr: 1.0e-04 loss: 7.268e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 170/30000 [24:11<72:40:20,  8.77s/it, lr: 1.0e-04 loss: 7.930e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 170/30000 [24:11<72:40:20,  8.77s/it, lr: 1.0e-04 loss: 7.930e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 171/30000 [24:11<52:06:10,  6.29s/it, lr: 1.0e-04 loss: 7.930e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 171/30000 [24:11<52:06:10,  6.29s/it, lr: 1.0e-04 loss: 7.930e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 171/30000 [24:17<52:06:10,  6.29s/it, lr: 1.0e-04 loss: 1.160e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 171/30000 [24:17<52:06:10,  6.29s/it, lr: 1.0e-04 loss: 1.160e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 172/30000 [24:17<52:45:34,  6.37s/it, lr: 1.0e-04 loss: 1.160e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 172/30000 [24:17<52:45:34,  6.37s/it, lr: 1.0e-04 loss: 1.160e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 172/30000 [24:24<52:45:34,  6.37s/it, lr: 1.0e-04 loss: 7.029e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 172/30000 [24:24<52:45:34,  6.37s/it, lr: 1.0e-04 loss: 7.029e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 173/30000 [24:24<54:26:20,  6.57s/it, lr: 1.0e-04 loss: 7.029e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 173/30000 [24:24<54:26:20,  6.57s/it, lr: 1.0e-04 loss: 7.029e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 173/30000 [24:31<54:26:20,  6.57s/it, lr: 1.0e-04 loss: 1.203e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 173/30000 [24:31<54:26:20,  6.57s/it, lr: 1.0e-04 loss: 1.203e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 174/30000 [24:31<54:31:29,  6.58s/it, lr: 1.0e-04 loss: 1.203e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 174/30000 [24:31<54:31:29,  6.58s/it, lr: 1.0e-04 loss: 1.203e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 174/30000 [24:38<54:31:29,  6.58s/it, lr: 1.0e-04 loss: 6.311e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 174/30000 [24:38<54:31:29,  6.58s/it, lr: 1.0e-04 loss: 6.311e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 175/30000 [24:38<55:26:52,  6.69s/it, lr: 1.0e-04 loss: 6.311e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 175/30000 [24:38<55:26:52,  6.69s/it, lr: 1.0e-04 loss: 6.311e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 175/30000 [24:48<55:26:52,  6.69s/it, lr: 1.0e-04 loss: 8.359e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 175/30000 [24:48<55:26:52,  6.69s/it, lr: 1.0e-04 loss: 8.359e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 176/30000 [24:48<63:34:27,  7.67s/it, lr: 1.0e-04 loss: 8.359e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 176/30000 [24:48<63:34:27,  7.67s/it, lr: 1.0e-04 loss: 8.359e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 176/30000 [24:55<63:34:27,  7.67s/it, lr: 1.0e-04 loss: 6.599e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 176/30000 [24:55<63:34:27,  7.67s/it, lr: 1.0e-04 loss: 6.599e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 177/30000 [24:55<62:11:03,  7.51s/it, lr: 1.0e-04 loss: 6.599e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 177/30000 [24:55<62:11:03,  7.51s/it, lr: 1.0e-04 loss: 6.599e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 177/30000 [25:05<62:11:03,  7.51s/it, lr: 1.0e-04 loss: 9.183e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 177/30000 [25:05<62:11:03,  7.51s/it, lr: 1.0e-04 loss: 9.183e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 178/30000 [25:05<67:57:05,  8.20s/it, lr: 1.0e-04 loss: 9.183e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 178/30000 [25:05<67:57:05,  8.20s/it, lr: 1.0e-04 loss: 9.183e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 178/30000 [25:12<67:57:05,  8.20s/it, lr: 1.0e-04 loss: 8.163e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 178/30000 [25:12<67:57:05,  8.20s/it, lr: 1.0e-04 loss: 8.163e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 179/30000 [25:12<64:18:31,  7.76s/it, lr: 1.0e-04 loss: 8.163e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 179/30000 [25:12<64:18:31,  7.76s/it, lr: 1.0e-04 loss: 8.163e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 179/30000 [25:19<64:18:31,  7.76s/it, lr: 1.0e-04 loss: 8.642e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 179/30000 [25:19<64:18:31,  7.76s/it, lr: 1.0e-04 loss: 8.642e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 180/30000 [25:29<64:18:23,  7.76s/it, lr: 1.0e-04 loss: 6.620e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 180/30000 [25:29<64:18:23,  7.76s/it, lr: 1.0e-04 loss: 6.620e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 181/30000 [25:29<53:56:05,  6.51s/it, lr: 1.0e-04 loss: 6.620e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 181/30000 [25:29<53:56:05,  6.51s/it, lr: 1.0e-04 loss: 6.620e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 181/30000 [25:36<53:56:05,  6.51s/it, lr: 1.0e-04 loss: 1.045e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 181/30000 [25:36<53:56:05,  6.51s/it, lr: 1.0e-04 loss: 1.045e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 182/30000 [25:36<54:52:28,  6.63s/it, lr: 1.0e-04 loss: 1.045e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 182/30000 [25:36<54:52:28,  6.63s/it, lr: 1.0e-04 loss: 1.045e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 182/30000 [25:46<54:52:28,  6.63s/it, lr: 1.0e-04 loss: 4.595e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 182/30000 [25:46<54:52:28,  6.63s/it, lr: 1.0e-04 loss: 4.595e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 183/30000 [25:46<62:20:12,  7.53s/it, lr: 1.0e-04 loss: 4.595e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 183/30000 [25:46<62:20:12,  7.53s/it, lr: 1.0e-04 loss: 4.595e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 183/30000 [25:56<62:20:12,  7.53s/it, lr: 1.0e-04 loss: 9.017e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 183/30000 [25:56<62:20:12,  7.53s/it, lr: 1.0e-04 loss: 9.017e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 184/30000 [25:56<67:53:22,  8.20s/it, lr: 1.0e-04 loss: 9.017e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 184/30000 [25:56<67:53:22,  8.20s/it, lr: 1.0e-04 loss: 9.017e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 184/30000 [26:03<67:53:22,  8.20s/it, lr: 1.0e-04 loss: 6.799e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 184/30000 [26:03<67:53:22,  8.20s/it, lr: 1.0e-04 loss: 6.799e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 185/30000 [26:03<65:09:21,  7.87s/it, lr: 1.0e-04 loss: 6.799e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 185/30000 [26:03<65:09:21,  7.87s/it, lr: 1.0e-04 loss: 6.799e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 185/30000 [26:13<65:09:21,  7.87s/it, lr: 1.0e-04 loss: 4.035e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 185/30000 [26:13<65:09:21,  7.87s/it, lr: 1.0e-04 loss: 4.035e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 186/30000 [26:13<70:28:24,  8.51s/it, lr: 1.0e-04 loss: 4.035e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 186/30000 [26:13<70:28:24,  8.51s/it, lr: 1.0e-04 loss: 4.035e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 186/30000 [26:20<70:28:24,  8.51s/it, lr: 1.0e-04 loss: 7.693e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 186/30000 [26:20<70:28:24,  8.51s/it, lr: 1.0e-04 loss: 7.693e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 187/30000 [26:20<66:12:57,  8.00s/it, lr: 1.0e-04 loss: 7.693e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 187/30000 [26:20<66:12:57,  8.00s/it, lr: 1.0e-04 loss: 7.693e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 187/30000 [26:30<66:12:57,  8.00s/it, lr: 1.0e-04 loss: 6.874e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 187/30000 [26:30<66:12:57,  8.00s/it, lr: 1.0e-04 loss: 6.874e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 188/30000 [26:30<71:15:49,  8.61s/it, lr: 1.0e-04 loss: 6.874e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 188/30000 [26:30<71:15:49,  8.61s/it, lr: 1.0e-04 loss: 6.874e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 188/30000 [26:37<71:15:49,  8.61s/it, lr: 1.0e-04 loss: 1.922e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 188/30000 [26:37<71:15:49,  8.61s/it, lr: 1.0e-04 loss: 1.922e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 189/30000 [26:37<66:42:28,  8.06s/it, lr: 1.0e-04 loss: 1.922e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 189/30000 [26:37<66:42:28,  8.06s/it, lr: 1.0e-04 loss: 1.922e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 189/30000 [26:44<66:42:28,  8.06s/it, lr: 1.0e-04 loss: 1.189e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 189/30000 [26:44<66:42:28,  8.06s/it, lr: 1.0e-04 loss: 1.189e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 190/30000 [26:51<66:42:20,  8.06s/it, lr: 1.0e-04 loss: 1.015e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 190/30000 [26:51<66:42:20,  8.06s/it, lr: 1.0e-04 loss: 1.015e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 191/30000 [26:51<49:52:36,  6.02s/it, lr: 1.0e-04 loss: 1.015e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 191/30000 [26:51<49:52:36,  6.02s/it, lr: 1.0e-04 loss: 1.015e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 191/30000 [27:01<49:52:36,  6.02s/it, lr: 1.0e-04 loss: 7.793e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 191/30000 [27:01<49:52:36,  6.02s/it, lr: 1.0e-04 loss: 7.793e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 192/30000 [27:01<57:51:28,  6.99s/it, lr: 1.0e-04 loss: 7.793e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 192/30000 [27:01<57:51:28,  6.99s/it, lr: 1.0e-04 loss: 7.793e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 192/30000 [27:11<57:51:28,  6.99s/it, lr: 1.0e-04 loss: 1.206e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 192/30000 [27:11<57:51:28,  6.99s/it, lr: 1.0e-04 loss: 1.206e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 193/30000 [27:11<63:35:05,  7.68s/it, lr: 1.0e-04 loss: 1.206e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 193/30000 [27:11<63:35:05,  7.68s/it, lr: 1.0e-04 loss: 1.206e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 193/30000 [27:17<63:35:05,  7.68s/it, lr: 1.0e-04 loss: 9.619e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 193/30000 [27:17<63:35:05,  7.68s/it, lr: 1.0e-04 loss: 9.619e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 194/30000 [27:17<61:12:57,  7.39s/it, lr: 1.0e-04 loss: 9.619e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 194/30000 [27:17<61:12:57,  7.39s/it, lr: 1.0e-04 loss: 9.619e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 194/30000 [27:24<61:12:57,  7.39s/it, lr: 1.0e-04 loss: 6.313e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 194/30000 [27:24<61:12:57,  7.39s/it, lr: 1.0e-04 loss: 6.313e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 195/30000 [27:24<59:26:52,  7.18s/it, lr: 1.0e-04 loss: 6.313e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 195/30000 [27:24<59:26:52,  7.18s/it, lr: 1.0e-04 loss: 6.313e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 195/30000 [27:31<59:26:52,  7.18s/it, lr: 1.0e-04 loss: 1.077e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 195/30000 [27:31<59:26:52,  7.18s/it, lr: 1.0e-04 loss: 1.077e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 196/30000 [27:31<58:11:05,  7.03s/it, lr: 1.0e-04 loss: 1.077e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 196/30000 [27:31<58:11:05,  7.03s/it, lr: 1.0e-04 loss: 1.077e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 196/30000 [27:38<58:11:05,  7.03s/it, lr: 1.0e-04 loss: 1.712e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 196/30000 [27:38<58:11:05,  7.03s/it, lr: 1.0e-04 loss: 1.712e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 197/30000 [27:38<59:36:43,  7.20s/it, lr: 1.0e-04 loss: 1.712e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 197/30000 [27:38<59:36:43,  7.20s/it, lr: 1.0e-04 loss: 1.712e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 197/30000 [27:48<59:36:43,  7.20s/it, lr: 1.0e-04 loss: 8.458e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 197/30000 [27:48<59:36:43,  7.20s/it, lr: 1.0e-04 loss: 8.458e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 198/30000 [27:48<64:56:43,  7.85s/it, lr: 1.0e-04 loss: 8.458e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 198/30000 [27:48<64:56:43,  7.85s/it, lr: 1.0e-04 loss: 8.458e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 198/30000 [27:58<64:56:43,  7.85s/it, lr: 1.0e-04 loss: 5.091e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 198/30000 [27:58<64:56:43,  7.85s/it, lr: 1.0e-04 loss: 5.091e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 199/30000 [27:58<70:02:26,  8.46s/it, lr: 1.0e-04 loss: 5.091e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 199/30000 [27:58<70:02:26,  8.46s/it, lr: 1.0e-04 loss: 5.091e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 199/30000 [28:08<70:02:26,  8.46s/it, lr: 1.0e-04 loss: 8.999e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 199/30000 [28:08<70:02:26,  8.46s/it, lr: 1.0e-04 loss: 8.999e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 200/30000 [28:18<70:02:17,  8.46s/it, lr: 1.0e-04 loss: 9.904e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 200/30000 [28:18<70:02:17,  8.46s/it, lr: 1.0e-04 loss: 9.904e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 201/30000 [28:18<56:49:48,  6.87s/it, lr: 1.0e-04 loss: 9.904e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 201/30000 [28:18<56:49:48,  6.87s/it, lr: 1.0e-04 loss: 9.904e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 201/30000 [28:28<56:49:48,  6.87s/it, lr: 1.0e-04 loss: 1.019e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 201/30000 [28:28<56:49:48,  6.87s/it, lr: 1.0e-04 loss: 1.019e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 202/30000 [28:28<63:25:12,  7.66s/it, lr: 1.0e-04 loss: 1.019e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 202/30000 [28:28<63:25:12,  7.66s/it, lr: 1.0e-04 loss: 1.019e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 202/30000 [28:35<63:25:12,  7.66s/it, lr: 1.0e-04 loss: 8.916e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 202/30000 [28:35<63:25:12,  7.66s/it, lr: 1.0e-04 loss: 8.916e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 203/30000 [28:35<62:37:52,  7.57s/it, lr: 1.0e-04 loss: 8.916e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 203/30000 [28:35<62:37:52,  7.57s/it, lr: 1.0e-04 loss: 8.916e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 203/30000 [28:45<62:37:52,  7.57s/it, lr: 1.0e-04 loss: 7.850e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 203/30000 [28:45<62:37:52,  7.57s/it, lr: 1.0e-04 loss: 7.850e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 204/30000 [28:45<66:59:46,  8.09s/it, lr: 1.0e-04 loss: 7.850e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 204/30000 [28:45<66:59:46,  8.09s/it, lr: 1.0e-04 loss: 7.850e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 204/30000 [28:54<66:59:46,  8.09s/it, lr: 1.0e-04 loss: 1.729e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 204/30000 [28:54<66:59:46,  8.09s/it, lr: 1.0e-04 loss: 1.729e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 205/30000 [28:54<70:06:10,  8.47s/it, lr: 1.0e-04 loss: 1.729e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 205/30000 [28:54<70:06:10,  8.47s/it, lr: 1.0e-04 loss: 1.729e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 205/30000 [29:01<70:06:10,  8.47s/it, lr: 1.0e-04 loss: 1.384e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 205/30000 [29:01<70:06:10,  8.47s/it, lr: 1.0e-04 loss: 1.384e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 206/30000 [29:01<67:38:55,  8.17s/it, lr: 1.0e-04 loss: 1.384e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 206/30000 [29:01<67:38:55,  8.17s/it, lr: 1.0e-04 loss: 1.384e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 206/30000 [29:08<67:38:55,  8.17s/it, lr: 1.0e-04 loss: 1.094e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 206/30000 [29:08<67:38:55,  8.17s/it, lr: 1.0e-04 loss: 1.094e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 207/30000 [29:08<64:02:34,  7.74s/it, lr: 1.0e-04 loss: 1.094e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 207/30000 [29:08<64:02:34,  7.74s/it, lr: 1.0e-04 loss: 1.094e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 207/30000 [29:15<64:02:34,  7.74s/it, lr: 1.0e-04 loss: 1.171e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 207/30000 [29:15<64:02:34,  7.74s/it, lr: 1.0e-04 loss: 1.171e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 208/30000 [29:15<62:58:13,  7.61s/it, lr: 1.0e-04 loss: 1.171e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 208/30000 [29:15<62:58:13,  7.61s/it, lr: 1.0e-04 loss: 1.171e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 208/30000 [29:25<62:58:13,  7.61s/it, lr: 1.0e-04 loss: 9.589e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 208/30000 [29:25<62:58:13,  7.61s/it, lr: 1.0e-04 loss: 9.589e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 209/30000 [29:25<67:42:58,  8.18s/it, lr: 1.0e-04 loss: 9.589e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 209/30000 [29:25<67:42:58,  8.18s/it, lr: 1.0e-04 loss: 9.589e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 209/30000 [29:35<67:42:58,  8.18s/it, lr: 1.0e-04 loss: 8.174e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 209/30000 [29:35<67:42:58,  8.18s/it, lr: 1.0e-04 loss: 8.174e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 210/30000 [29:43<67:42:50,  8.18s/it, lr: 1.0e-04 loss: 4.992e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 210/30000 [29:43<67:42:50,  8.18s/it, lr: 1.0e-04 loss: 4.992e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 211/30000 [29:43<50:50:19,  6.14s/it, lr: 1.0e-04 loss: 4.992e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 211/30000 [29:43<50:50:19,  6.14s/it, lr: 1.0e-04 loss: 4.992e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 211/30000 [29:53<50:50:19,  6.14s/it, lr: 1.0e-04 loss: 8.657e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 211/30000 [29:53<50:50:19,  6.14s/it, lr: 1.0e-04 loss: 8.657e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 212/30000 [29:53<58:59:00,  7.13s/it, lr: 1.0e-04 loss: 8.657e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 212/30000 [29:53<58:59:00,  7.13s/it, lr: 1.0e-04 loss: 8.657e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 212/30000 [30:00<58:59:00,  7.13s/it, lr: 1.0e-04 loss: 7.459e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 212/30000 [30:00<58:59:00,  7.13s/it, lr: 1.0e-04 loss: 7.459e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 213/30000 [30:00<58:51:25,  7.11s/it, lr: 1.0e-04 loss: 7.459e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 213/30000 [30:00<58:51:25,  7.11s/it, lr: 1.0e-04 loss: 7.459e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 213/30000 [30:07<58:51:25,  7.11s/it, lr: 1.0e-04 loss: 9.144e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 213/30000 [30:07<58:51:25,  7.11s/it, lr: 1.0e-04 loss: 9.144e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 214/30000 [30:07<57:49:48,  6.99s/it, lr: 1.0e-04 loss: 9.144e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 214/30000 [30:07<57:49:48,  6.99s/it, lr: 1.0e-04 loss: 9.144e-02]
Xiang_wan_2_2_lora:   1%|2                                         | 214/30000 [30:13<57:49:48,  6.99s/it, lr: 1.0e-04 loss: 4.842e-01]
Xiang_wan_2_2_lora:   1%|2                                         | 214/30000 [30:13<57:49:48,  6.99s/it, lr: 1.0e-04 loss: 4.842e-01]
Xiang_wan_2_2_lora:   1%|3                                         | 215/30000 [30:13<56:56:01,  6.88s/it, lr: 1.0e-04 loss: 4.842e-01]
Xiang_wan_2_2_lora:   1%|3                                         | 215/30000 [30:13<56:56:01,  6.88s/it, lr: 1.0e-04 loss: 4.842e-01]
Xiang_wan_2_2_lora:   1%|3                                         | 215/30000 [30:20<56:56:01,  6.88s/it, lr: 1.0e-04 loss: 8.660e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 215/30000 [30:20<56:56:01,  6.88s/it, lr: 1.0e-04 loss: 8.660e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 216/30000 [30:20<56:43:56,  6.86s/it, lr: 1.0e-04 loss: 8.660e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 216/30000 [30:20<56:43:56,  6.86s/it, lr: 1.0e-04 loss: 8.660e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 216/30000 [30:30<56:43:56,  6.86s/it, lr: 1.0e-04 loss: 7.304e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 216/30000 [30:30<56:43:56,  6.86s/it, lr: 1.0e-04 loss: 7.304e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 217/30000 [30:30<63:40:59,  7.70s/it, lr: 1.0e-04 loss: 7.304e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 217/30000 [30:30<63:40:59,  7.70s/it, lr: 1.0e-04 loss: 7.304e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 217/30000 [30:40<63:40:59,  7.70s/it, lr: 1.0e-04 loss: 1.187e-01]
Xiang_wan_2_2_lora:   1%|3                                         | 217/30000 [30:40<63:40:59,  7.70s/it, lr: 1.0e-04 loss: 1.187e-01]
Xiang_wan_2_2_lora:   1%|3                                         | 218/30000 [30:40<69:17:50,  8.38s/it, lr: 1.0e-04 loss: 1.187e-01]
Xiang_wan_2_2_lora:   1%|3                                         | 218/30000 [30:40<69:17:50,  8.38s/it, lr: 1.0e-04 loss: 1.187e-01]
Xiang_wan_2_2_lora:   1%|3                                         | 218/30000 [30:47<69:17:50,  8.38s/it, lr: 1.0e-04 loss: 7.596e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 218/30000 [30:47<69:17:50,  8.38s/it, lr: 1.0e-04 loss: 7.596e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 219/30000 [30:47<66:40:42,  8.06s/it, lr: 1.0e-04 loss: 7.596e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 219/30000 [30:47<66:40:42,  8.06s/it, lr: 1.0e-04 loss: 7.596e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 219/30000 [30:57<66:40:42,  8.06s/it, lr: 1.0e-04 loss: 5.861e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 219/30000 [30:57<66:40:42,  8.06s/it, lr: 1.0e-04 loss: 5.861e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 220/30000 [31:04<66:40:34,  8.06s/it, lr: 1.0e-04 loss: 6.616e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 220/30000 [31:04<66:40:34,  8.06s/it, lr: 1.0e-04 loss: 6.616e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 221/30000 [31:04<49:59:55,  6.04s/it, lr: 1.0e-04 loss: 6.616e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 221/30000 [31:04<49:59:55,  6.04s/it, lr: 1.0e-04 loss: 6.616e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 221/30000 [31:14<49:59:55,  6.04s/it, lr: 1.0e-04 loss: 1.066e-01]
Xiang_wan_2_2_lora:   1%|3                                         | 221/30000 [31:14<49:59:55,  6.04s/it, lr: 1.0e-04 loss: 1.066e-01]
Xiang_wan_2_2_lora:   1%|3                                         | 222/30000 [31:14<58:13:19,  7.04s/it, lr: 1.0e-04 loss: 1.066e-01]
Xiang_wan_2_2_lora:   1%|3                                         | 222/30000 [31:14<58:13:19,  7.04s/it, lr: 1.0e-04 loss: 1.066e-01]
Xiang_wan_2_2_lora:   1%|3                                         | 222/30000 [31:22<58:13:19,  7.04s/it, lr: 1.0e-04 loss: 5.291e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 222/30000 [31:22<58:13:19,  7.04s/it, lr: 1.0e-04 loss: 5.291e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 223/30000 [31:22<59:17:56,  7.17s/it, lr: 1.0e-04 loss: 5.291e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 223/30000 [31:22<59:17:56,  7.17s/it, lr: 1.0e-04 loss: 5.291e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 223/30000 [31:29<59:17:56,  7.17s/it, lr: 1.0e-04 loss: 8.095e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 223/30000 [31:29<59:17:56,  7.17s/it, lr: 1.0e-04 loss: 8.095e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 224/30000 [31:29<59:09:02,  7.15s/it, lr: 1.0e-04 loss: 8.095e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 224/30000 [31:29<59:09:02,  7.15s/it, lr: 1.0e-04 loss: 8.095e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 224/30000 [31:39<59:09:02,  7.15s/it, lr: 1.0e-04 loss: 6.497e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 224/30000 [31:39<59:09:02,  7.15s/it, lr: 1.0e-04 loss: 6.497e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 225/30000 [31:39<66:01:41,  7.98s/it, lr: 1.0e-04 loss: 6.497e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 225/30000 [31:39<66:01:41,  7.98s/it, lr: 1.0e-04 loss: 6.497e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 225/30000 [31:46<66:01:41,  7.98s/it, lr: 1.0e-04 loss: 7.636e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 225/30000 [31:46<66:01:41,  7.98s/it, lr: 1.0e-04 loss: 7.636e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 226/30000 [31:46<64:42:34,  7.82s/it, lr: 1.0e-04 loss: 7.636e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 226/30000 [31:46<64:42:34,  7.82s/it, lr: 1.0e-04 loss: 7.636e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 226/30000 [31:56<64:42:34,  7.82s/it, lr: 1.0e-04 loss: 8.664e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 226/30000 [31:56<64:42:34,  7.82s/it, lr: 1.0e-04 loss: 8.664e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 227/30000 [31:56<69:20:14,  8.38s/it, lr: 1.0e-04 loss: 8.664e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 227/30000 [31:56<69:20:14,  8.38s/it, lr: 1.0e-04 loss: 8.664e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 227/30000 [32:03<69:20:14,  8.38s/it, lr: 1.0e-04 loss: 8.557e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 227/30000 [32:03<69:20:14,  8.38s/it, lr: 1.0e-04 loss: 8.557e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 228/30000 [32:03<65:16:33,  7.89s/it, lr: 1.0e-04 loss: 8.557e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 228/30000 [32:03<65:16:33,  7.89s/it, lr: 1.0e-04 loss: 8.557e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 228/30000 [32:10<65:16:33,  7.89s/it, lr: 1.0e-04 loss: 1.167e-01]
Xiang_wan_2_2_lora:   1%|3                                         | 228/30000 [32:10<65:16:33,  7.89s/it, lr: 1.0e-04 loss: 1.167e-01]
Xiang_wan_2_2_lora:   1%|3                                         | 229/30000 [32:10<62:51:19,  7.60s/it, lr: 1.0e-04 loss: 1.167e-01]
Xiang_wan_2_2_lora:   1%|3                                         | 229/30000 [32:10<62:51:19,  7.60s/it, lr: 1.0e-04 loss: 1.167e-01]
Xiang_wan_2_2_lora:   1%|3                                         | 229/30000 [32:17<62:51:19,  7.60s/it, lr: 1.0e-04 loss: 1.151e-01]
Xiang_wan_2_2_lora:   1%|3                                         | 229/30000 [32:17<62:51:19,  7.60s/it, lr: 1.0e-04 loss: 1.151e-01]
Xiang_wan_2_2_lora:   1%|3                                         | 230/30000 [32:24<62:51:12,  7.60s/it, lr: 1.0e-04 loss: 6.607e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 230/30000 [32:24<62:51:12,  7.60s/it, lr: 1.0e-04 loss: 6.607e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 231/30000 [32:24<46:48:47,  5.66s/it, lr: 1.0e-04 loss: 6.607e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 231/30000 [32:24<46:48:47,  5.66s/it, lr: 1.0e-04 loss: 6.607e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 231/30000 [32:33<46:48:47,  5.66s/it, lr: 1.0e-04 loss: 1.173e-01]
Xiang_wan_2_2_lora:   1%|3                                         | 231/30000 [32:33<46:48:47,  5.66s/it, lr: 1.0e-04 loss: 1.173e-01]
Xiang_wan_2_2_lora:   1%|3                                         | 232/30000 [32:33<55:34:07,  6.72s/it, lr: 1.0e-04 loss: 1.173e-01]
Xiang_wan_2_2_lora:   1%|3                                         | 232/30000 [32:33<55:34:07,  6.72s/it, lr: 1.0e-04 loss: 1.173e-01]
Xiang_wan_2_2_lora:   1%|3                                         | 232/30000 [32:40<55:34:07,  6.72s/it, lr: 1.0e-04 loss: 4.398e-01]
Xiang_wan_2_2_lora:   1%|3                                         | 232/30000 [32:40<55:34:07,  6.72s/it, lr: 1.0e-04 loss: 4.398e-01]
Xiang_wan_2_2_lora:   1%|3                                         | 233/30000 [32:40<55:41:24,  6.74s/it, lr: 1.0e-04 loss: 4.398e-01]
Xiang_wan_2_2_lora:   1%|3                                         | 233/30000 [32:40<55:41:24,  6.74s/it, lr: 1.0e-04 loss: 4.398e-01]
Xiang_wan_2_2_lora:   1%|3                                         | 233/30000 [32:47<55:41:24,  6.74s/it, lr: 1.0e-04 loss: 9.609e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 233/30000 [32:47<55:41:24,  6.74s/it, lr: 1.0e-04 loss: 9.609e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 234/30000 [32:47<55:26:13,  6.70s/it, lr: 1.0e-04 loss: 9.609e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 234/30000 [32:47<55:26:13,  6.70s/it, lr: 1.0e-04 loss: 9.609e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 234/30000 [32:56<55:26:13,  6.70s/it, lr: 1.0e-04 loss: 3.655e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 234/30000 [32:56<55:26:13,  6.70s/it, lr: 1.0e-04 loss: 3.655e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 235/30000 [32:56<61:36:23,  7.45s/it, lr: 1.0e-04 loss: 3.655e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 235/30000 [32:56<61:36:23,  7.45s/it, lr: 1.0e-04 loss: 3.655e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 235/30000 [33:06<61:36:23,  7.45s/it, lr: 1.0e-04 loss: 6.898e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 235/30000 [33:06<61:36:23,  7.45s/it, lr: 1.0e-04 loss: 6.898e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 236/30000 [33:06<66:37:48,  8.06s/it, lr: 1.0e-04 loss: 6.898e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 236/30000 [33:06<66:37:48,  8.06s/it, lr: 1.0e-04 loss: 6.898e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 236/30000 [33:13<66:37:48,  8.06s/it, lr: 1.0e-04 loss: 3.007e-01]
Xiang_wan_2_2_lora:   1%|3                                         | 236/30000 [33:13<66:37:48,  8.06s/it, lr: 1.0e-04 loss: 3.007e-01]
Xiang_wan_2_2_lora:   1%|3                                         | 237/30000 [33:13<65:12:32,  7.89s/it, lr: 1.0e-04 loss: 3.007e-01]
Xiang_wan_2_2_lora:   1%|3                                         | 237/30000 [33:13<65:12:32,  7.89s/it, lr: 1.0e-04 loss: 3.007e-01]
Xiang_wan_2_2_lora:   1%|3                                         | 237/30000 [33:23<65:12:32,  7.89s/it, lr: 1.0e-04 loss: 8.919e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 237/30000 [33:23<65:12:32,  7.89s/it, lr: 1.0e-04 loss: 8.919e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 238/30000 [33:23<68:51:14,  8.33s/it, lr: 1.0e-04 loss: 8.919e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 238/30000 [33:23<68:51:14,  8.33s/it, lr: 1.0e-04 loss: 8.919e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 238/30000 [33:30<68:51:14,  8.33s/it, lr: 1.0e-04 loss: 8.624e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 238/30000 [33:30<68:51:14,  8.33s/it, lr: 1.0e-04 loss: 8.624e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 239/30000 [33:30<67:19:38,  8.14s/it, lr: 1.0e-04 loss: 8.624e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 239/30000 [33:30<67:19:38,  8.14s/it, lr: 1.0e-04 loss: 8.624e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 239/30000 [33:40<67:19:38,  8.14s/it, lr: 1.0e-04 loss: 9.059e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 239/30000 [33:40<67:19:38,  8.14s/it, lr: 1.0e-04 loss: 9.059e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 240/30000 [33:49<67:19:30,  8.14s/it, lr: 1.0e-04 loss: 5.153e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 240/30000 [33:49<67:19:30,  8.14s/it, lr: 1.0e-04 loss: 5.153e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 241/30000 [33:49<54:21:25,  6.58s/it, lr: 1.0e-04 loss: 5.153e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 241/30000 [33:49<54:21:25,  6.58s/it, lr: 1.0e-04 loss: 5.153e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 241/30000 [33:56<54:21:25,  6.58s/it, lr: 1.0e-04 loss: 9.815e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 241/30000 [33:56<54:21:25,  6.58s/it, lr: 1.0e-04 loss: 9.815e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 242/30000 [33:56<54:36:12,  6.61s/it, lr: 1.0e-04 loss: 9.815e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 242/30000 [33:56<54:36:12,  6.61s/it, lr: 1.0e-04 loss: 9.815e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 242/30000 [34:03<54:36:12,  6.61s/it, lr: 1.0e-04 loss: 5.217e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 242/30000 [34:03<54:36:12,  6.61s/it, lr: 1.0e-04 loss: 5.217e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 243/30000 [34:03<54:55:54,  6.65s/it, lr: 1.0e-04 loss: 5.217e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 243/30000 [34:03<54:55:54,  6.65s/it, lr: 1.0e-04 loss: 5.217e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 243/30000 [34:09<54:55:54,  6.65s/it, lr: 1.0e-04 loss: 9.872e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 243/30000 [34:09<54:55:54,  6.65s/it, lr: 1.0e-04 loss: 9.872e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 244/30000 [34:09<55:01:28,  6.66s/it, lr: 1.0e-04 loss: 9.872e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 244/30000 [34:09<55:01:28,  6.66s/it, lr: 1.0e-04 loss: 9.872e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 244/30000 [34:16<55:01:28,  6.66s/it, lr: 1.0e-04 loss: 4.428e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 244/30000 [34:16<55:01:28,  6.66s/it, lr: 1.0e-04 loss: 4.428e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 245/30000 [34:16<55:54:44,  6.76s/it, lr: 1.0e-04 loss: 4.428e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 245/30000 [34:16<55:54:44,  6.76s/it, lr: 1.0e-04 loss: 4.428e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 245/30000 [34:27<55:54:44,  6.76s/it, lr: 1.0e-04 loss: 7.546e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 245/30000 [34:27<55:54:44,  6.76s/it, lr: 1.0e-04 loss: 7.546e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 246/30000 [34:27<63:53:39,  7.73s/it, lr: 1.0e-04 loss: 7.546e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 246/30000 [34:27<63:53:39,  7.73s/it, lr: 1.0e-04 loss: 7.546e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 246/30000 [34:36<63:53:39,  7.73s/it, lr: 1.0e-04 loss: 4.278e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 246/30000 [34:36<63:53:39,  7.73s/it, lr: 1.0e-04 loss: 4.278e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 247/30000 [34:36<69:03:17,  8.36s/it, lr: 1.0e-04 loss: 4.278e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 247/30000 [34:36<69:03:17,  8.36s/it, lr: 1.0e-04 loss: 4.278e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 247/30000 [34:46<69:03:17,  8.36s/it, lr: 1.0e-04 loss: 7.840e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 247/30000 [34:46<69:03:17,  8.36s/it, lr: 1.0e-04 loss: 7.840e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 248/30000 [34:46<71:43:09,  8.68s/it, lr: 1.0e-04 loss: 7.840e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 248/30000 [34:46<71:43:09,  8.68s/it, lr: 1.0e-04 loss: 7.840e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 248/30000 [34:55<71:43:09,  8.68s/it, lr: 1.0e-04 loss: 4.184e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 248/30000 [34:55<71:43:09,  8.68s/it, lr: 1.0e-04 loss: 4.184e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 249/30000 [34:55<73:43:35,  8.92s/it, lr: 1.0e-04 loss: 4.184e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 249/30000 [34:55<73:43:35,  8.92s/it, lr: 1.0e-04 loss: 4.184e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 249/30000 [35:02<73:43:35,  8.92s/it, lr: 1.0e-04 loss: 9.515e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 249/30000 [35:02<73:43:35,  8.92s/it, lr: 1.0e-04 loss: 9.515e-02]
+Saving at step 250
+Saved checkpoint to /home/featurize/ai-toolkit/output/Xiang_wan_2_2_lora/Xiang_wan_2_2_lora_000000250.safetensors
+Saved optimizer to /home/featurize/ai-toolkit/output/Xiang_wan_2_2_lora/optimizer.pt
+
Xiang_wan_2_2_lora:   1%|3                                         | 250/30000 [35:10<73:43:26,  8.92s/it, lr: 1.0e-04 loss: 1.065e-01]
Xiang_wan_2_2_lora:   1%|3                                         | 250/30000 [35:10<73:43:26,  8.92s/it, lr: 1.0e-04 loss: 1.065e-01]
Xiang_wan_2_2_lora:   1%|3                                         | 251/30000 [35:10<53:41:16,  6.50s/it, lr: 1.0e-04 loss: 1.065e-01]
Xiang_wan_2_2_lora:   1%|3                                         | 251/30000 [35:10<53:41:16,  6.50s/it, lr: 1.0e-04 loss: 1.065e-01]
Xiang_wan_2_2_lora:   1%|3                                         | 251/30000 [35:17<53:41:16,  6.50s/it, lr: 1.0e-04 loss: 1.061e-01]
Xiang_wan_2_2_lora:   1%|3                                         | 251/30000 [35:17<53:41:16,  6.50s/it, lr: 1.0e-04 loss: 1.061e-01]
Xiang_wan_2_2_lora:   1%|3                                         | 252/30000 [35:17<54:47:20,  6.63s/it, lr: 1.0e-04 loss: 1.061e-01]
Xiang_wan_2_2_lora:   1%|3                                         | 252/30000 [35:17<54:47:20,  6.63s/it, lr: 1.0e-04 loss: 1.061e-01]
Xiang_wan_2_2_lora:   1%|3                                         | 252/30000 [35:23<54:47:20,  6.63s/it, lr: 1.0e-04 loss: 7.903e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 252/30000 [35:23<54:47:20,  6.63s/it, lr: 1.0e-04 loss: 7.903e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 253/30000 [35:23<54:52:59,  6.64s/it, lr: 1.0e-04 loss: 7.903e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 253/30000 [35:23<54:52:59,  6.64s/it, lr: 1.0e-04 loss: 7.903e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 253/30000 [35:33<54:52:59,  6.64s/it, lr: 1.0e-04 loss: 9.807e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 253/30000 [35:33<54:52:59,  6.64s/it, lr: 1.0e-04 loss: 9.807e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 254/30000 [35:33<61:18:20,  7.42s/it, lr: 1.0e-04 loss: 9.807e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 254/30000 [35:33<61:18:20,  7.42s/it, lr: 1.0e-04 loss: 9.807e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 254/30000 [35:43<61:18:20,  7.42s/it, lr: 1.0e-04 loss: 4.745e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 254/30000 [35:43<61:18:20,  7.42s/it, lr: 1.0e-04 loss: 4.745e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 255/30000 [35:43<67:35:50,  8.18s/it, lr: 1.0e-04 loss: 4.745e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 255/30000 [35:43<67:35:50,  8.18s/it, lr: 1.0e-04 loss: 4.745e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 255/30000 [35:53<67:35:50,  8.18s/it, lr: 1.0e-04 loss: 7.426e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 255/30000 [35:53<67:35:50,  8.18s/it, lr: 1.0e-04 loss: 7.426e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 256/30000 [35:53<71:29:39,  8.65s/it, lr: 1.0e-04 loss: 7.426e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 256/30000 [35:53<71:29:39,  8.65s/it, lr: 1.0e-04 loss: 7.426e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 256/30000 [36:03<71:29:39,  8.65s/it, lr: 1.0e-04 loss: 8.512e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 256/30000 [36:03<71:29:39,  8.65s/it, lr: 1.0e-04 loss: 8.512e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 257/30000 [36:03<75:02:36,  9.08s/it, lr: 1.0e-04 loss: 8.512e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 257/30000 [36:03<75:02:36,  9.08s/it, lr: 1.0e-04 loss: 8.512e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 257/30000 [36:13<75:02:36,  9.08s/it, lr: 1.0e-04 loss: 5.288e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 257/30000 [36:13<75:02:36,  9.08s/it, lr: 1.0e-04 loss: 5.288e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 258/30000 [36:13<77:34:33,  9.39s/it, lr: 1.0e-04 loss: 5.288e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 258/30000 [36:13<77:34:33,  9.39s/it, lr: 1.0e-04 loss: 5.288e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 258/30000 [36:20<77:34:33,  9.39s/it, lr: 1.0e-04 loss: 8.198e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 258/30000 [36:20<77:34:33,  9.39s/it, lr: 1.0e-04 loss: 8.198e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 259/30000 [36:20<71:06:08,  8.61s/it, lr: 1.0e-04 loss: 8.198e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 259/30000 [36:20<71:06:08,  8.61s/it, lr: 1.0e-04 loss: 8.198e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 259/30000 [36:29<71:06:08,  8.61s/it, lr: 1.0e-04 loss: 7.791e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 259/30000 [36:29<71:06:08,  8.61s/it, lr: 1.0e-04 loss: 7.791e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 260/30000 [36:36<71:05:59,  8.61s/it, lr: 1.0e-04 loss: 8.434e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 260/30000 [36:36<71:05:59,  8.61s/it, lr: 1.0e-04 loss: 8.434e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 261/30000 [36:36<51:08:06,  6.19s/it, lr: 1.0e-04 loss: 8.434e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 261/30000 [36:36<51:08:06,  6.19s/it, lr: 1.0e-04 loss: 8.434e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 261/30000 [36:43<51:08:06,  6.19s/it, lr: 1.0e-04 loss: 9.449e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 261/30000 [36:43<51:08:06,  6.19s/it, lr: 1.0e-04 loss: 9.449e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 262/30000 [36:43<51:58:21,  6.29s/it, lr: 1.0e-04 loss: 9.449e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 262/30000 [36:43<51:58:21,  6.29s/it, lr: 1.0e-04 loss: 9.449e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 262/30000 [36:52<51:58:21,  6.29s/it, lr: 1.0e-04 loss: 6.817e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 262/30000 [36:52<51:58:21,  6.29s/it, lr: 1.0e-04 loss: 6.817e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 263/30000 [36:52<58:36:17,  7.09s/it, lr: 1.0e-04 loss: 6.817e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 263/30000 [36:52<58:36:17,  7.09s/it, lr: 1.0e-04 loss: 6.817e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 263/30000 [36:59<58:36:17,  7.09s/it, lr: 1.0e-04 loss: 9.457e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 263/30000 [36:59<58:36:17,  7.09s/it, lr: 1.0e-04 loss: 9.457e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 264/30000 [36:59<57:37:41,  6.98s/it, lr: 1.0e-04 loss: 9.457e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 264/30000 [36:59<57:37:41,  6.98s/it, lr: 1.0e-04 loss: 9.457e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 264/30000 [37:08<57:37:41,  6.98s/it, lr: 1.0e-04 loss: 5.883e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 264/30000 [37:08<57:37:41,  6.98s/it, lr: 1.0e-04 loss: 5.883e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 265/30000 [37:08<63:34:18,  7.70s/it, lr: 1.0e-04 loss: 5.883e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 265/30000 [37:08<63:34:18,  7.70s/it, lr: 1.0e-04 loss: 5.883e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 265/30000 [37:18<63:34:18,  7.70s/it, lr: 1.0e-04 loss: 6.230e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 265/30000 [37:18<63:34:18,  7.70s/it, lr: 1.0e-04 loss: 6.230e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 266/30000 [37:18<68:56:28,  8.35s/it, lr: 1.0e-04 loss: 6.230e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 266/30000 [37:18<68:56:28,  8.35s/it, lr: 1.0e-04 loss: 6.230e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 266/30000 [37:28<68:56:28,  8.35s/it, lr: 1.0e-04 loss: 6.369e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 266/30000 [37:28<68:56:28,  8.35s/it, lr: 1.0e-04 loss: 6.369e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 267/30000 [37:28<71:25:46,  8.65s/it, lr: 1.0e-04 loss: 6.369e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 267/30000 [37:28<71:25:46,  8.65s/it, lr: 1.0e-04 loss: 6.369e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 267/30000 [37:37<71:25:46,  8.65s/it, lr: 1.0e-04 loss: 5.186e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 267/30000 [37:37<71:25:46,  8.65s/it, lr: 1.0e-04 loss: 5.186e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 268/30000 [37:37<73:17:50,  8.87s/it, lr: 1.0e-04 loss: 5.186e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 268/30000 [37:37<73:17:50,  8.87s/it, lr: 1.0e-04 loss: 5.186e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 268/30000 [37:47<73:17:50,  8.87s/it, lr: 1.0e-04 loss: 6.804e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 268/30000 [37:47<73:17:50,  8.87s/it, lr: 1.0e-04 loss: 6.804e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 269/30000 [37:47<76:09:54,  9.22s/it, lr: 1.0e-04 loss: 6.804e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 269/30000 [37:47<76:09:54,  9.22s/it, lr: 1.0e-04 loss: 6.804e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 269/30000 [37:54<76:09:54,  9.22s/it, lr: 1.0e-04 loss: 9.992e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 269/30000 [37:54<76:09:54,  9.22s/it, lr: 1.0e-04 loss: 9.992e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 270/30000 [38:04<76:09:45,  9.22s/it, lr: 1.0e-04 loss: 3.462e-01]
Xiang_wan_2_2_lora:   1%|3                                         | 270/30000 [38:04<76:09:45,  9.22s/it, lr: 1.0e-04 loss: 3.462e-01]
Xiang_wan_2_2_lora:   1%|3                                         | 271/30000 [38:04<59:59:13,  7.26s/it, lr: 1.0e-04 loss: 3.462e-01]
Xiang_wan_2_2_lora:   1%|3                                         | 271/30000 [38:04<59:59:13,  7.26s/it, lr: 1.0e-04 loss: 3.462e-01]
Xiang_wan_2_2_lora:   1%|3                                         | 271/30000 [38:13<59:59:13,  7.26s/it, lr: 1.0e-04 loss: 5.723e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 271/30000 [38:13<59:59:13,  7.26s/it, lr: 1.0e-04 loss: 5.723e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 272/30000 [38:13<64:45:05,  7.84s/it, lr: 1.0e-04 loss: 5.723e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 272/30000 [38:13<64:45:05,  7.84s/it, lr: 1.0e-04 loss: 5.723e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 272/30000 [38:20<64:45:05,  7.84s/it, lr: 1.0e-04 loss: 7.426e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 272/30000 [38:20<64:45:05,  7.84s/it, lr: 1.0e-04 loss: 7.426e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 273/30000 [38:20<62:11:56,  7.53s/it, lr: 1.0e-04 loss: 7.426e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 273/30000 [38:20<62:11:56,  7.53s/it, lr: 1.0e-04 loss: 7.426e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 273/30000 [38:26<62:11:56,  7.53s/it, lr: 1.0e-04 loss: 1.142e-01]
Xiang_wan_2_2_lora:   1%|3                                         | 273/30000 [38:26<62:11:56,  7.53s/it, lr: 1.0e-04 loss: 1.142e-01]
Xiang_wan_2_2_lora:   1%|3                                         | 274/30000 [38:26<60:18:31,  7.30s/it, lr: 1.0e-04 loss: 1.142e-01]
Xiang_wan_2_2_lora:   1%|3                                         | 274/30000 [38:26<60:18:31,  7.30s/it, lr: 1.0e-04 loss: 1.142e-01]
Xiang_wan_2_2_lora:   1%|3                                         | 274/30000 [38:33<60:18:31,  7.30s/it, lr: 1.0e-04 loss: 1.946e-01]
Xiang_wan_2_2_lora:   1%|3                                         | 274/30000 [38:33<60:18:31,  7.30s/it, lr: 1.0e-04 loss: 1.946e-01]
Xiang_wan_2_2_lora:   1%|3                                         | 275/30000 [38:33<59:00:42,  7.15s/it, lr: 1.0e-04 loss: 1.946e-01]
Xiang_wan_2_2_lora:   1%|3                                         | 275/30000 [38:33<59:00:42,  7.15s/it, lr: 1.0e-04 loss: 1.946e-01]
Xiang_wan_2_2_lora:   1%|3                                         | 275/30000 [38:40<59:00:42,  7.15s/it, lr: 1.0e-04 loss: 1.155e-01]
Xiang_wan_2_2_lora:   1%|3                                         | 275/30000 [38:40<59:00:42,  7.15s/it, lr: 1.0e-04 loss: 1.155e-01]
Xiang_wan_2_2_lora:   1%|3                                         | 276/30000 [38:40<57:44:48,  6.99s/it, lr: 1.0e-04 loss: 1.155e-01]
Xiang_wan_2_2_lora:   1%|3                                         | 276/30000 [38:40<57:44:48,  6.99s/it, lr: 1.0e-04 loss: 1.155e-01]
Xiang_wan_2_2_lora:   1%|3                                         | 276/30000 [38:47<57:44:48,  6.99s/it, lr: 1.0e-04 loss: 7.041e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 276/30000 [38:47<57:44:48,  6.99s/it, lr: 1.0e-04 loss: 7.041e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 277/30000 [38:47<57:00:38,  6.91s/it, lr: 1.0e-04 loss: 7.041e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 277/30000 [38:47<57:00:38,  6.91s/it, lr: 1.0e-04 loss: 7.041e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 277/30000 [38:56<57:00:38,  6.91s/it, lr: 1.0e-04 loss: 6.350e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 277/30000 [38:56<57:00:38,  6.91s/it, lr: 1.0e-04 loss: 6.350e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 278/30000 [38:56<62:57:17,  7.63s/it, lr: 1.0e-04 loss: 6.350e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 278/30000 [38:56<62:57:17,  7.63s/it, lr: 1.0e-04 loss: 6.350e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 278/30000 [39:05<62:57:17,  7.63s/it, lr: 1.0e-04 loss: 4.732e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 278/30000 [39:05<62:57:17,  7.63s/it, lr: 1.0e-04 loss: 4.732e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 279/30000 [39:05<67:19:06,  8.15s/it, lr: 1.0e-04 loss: 4.732e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 279/30000 [39:05<67:19:06,  8.15s/it, lr: 1.0e-04 loss: 4.732e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 279/30000 [39:12<67:19:06,  8.15s/it, lr: 1.0e-04 loss: 8.380e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 279/30000 [39:12<67:19:06,  8.15s/it, lr: 1.0e-04 loss: 8.380e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 280/30000 [39:22<67:18:58,  8.15s/it, lr: 1.0e-04 loss: 5.467e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 280/30000 [39:22<67:18:58,  8.15s/it, lr: 1.0e-04 loss: 5.467e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 281/30000 [39:22<55:15:32,  6.69s/it, lr: 1.0e-04 loss: 5.467e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 281/30000 [39:22<55:15:32,  6.69s/it, lr: 1.0e-04 loss: 5.467e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 281/30000 [39:29<55:15:32,  6.69s/it, lr: 1.0e-04 loss: 1.091e-01]
Xiang_wan_2_2_lora:   1%|3                                         | 281/30000 [39:29<55:15:32,  6.69s/it, lr: 1.0e-04 loss: 1.091e-01]
Xiang_wan_2_2_lora:   1%|3                                         | 282/30000 [39:29<56:28:31,  6.84s/it, lr: 1.0e-04 loss: 1.091e-01]
Xiang_wan_2_2_lora:   1%|3                                         | 282/30000 [39:29<56:28:31,  6.84s/it, lr: 1.0e-04 loss: 1.091e-01]
Xiang_wan_2_2_lora:   1%|3                                         | 282/30000 [39:39<56:28:31,  6.84s/it, lr: 1.0e-04 loss: 6.055e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 282/30000 [39:39<56:28:31,  6.84s/it, lr: 1.0e-04 loss: 6.055e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 283/30000 [39:39<62:00:17,  7.51s/it, lr: 1.0e-04 loss: 6.055e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 283/30000 [39:39<62:00:17,  7.51s/it, lr: 1.0e-04 loss: 6.055e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 283/30000 [39:46<62:00:17,  7.51s/it, lr: 1.0e-04 loss: 9.979e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 283/30000 [39:46<62:00:17,  7.51s/it, lr: 1.0e-04 loss: 9.979e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 284/30000 [39:46<60:47:20,  7.36s/it, lr: 1.0e-04 loss: 9.979e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 284/30000 [39:46<60:47:20,  7.36s/it, lr: 1.0e-04 loss: 9.979e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 284/30000 [39:52<60:47:20,  7.36s/it, lr: 1.0e-04 loss: 1.120e-01]
Xiang_wan_2_2_lora:   1%|3                                         | 284/30000 [39:52<60:47:20,  7.36s/it, lr: 1.0e-04 loss: 1.120e-01]
Xiang_wan_2_2_lora:   1%|3                                         | 285/30000 [39:52<59:05:51,  7.16s/it, lr: 1.0e-04 loss: 1.120e-01]
Xiang_wan_2_2_lora:   1%|3                                         | 285/30000 [39:52<59:05:51,  7.16s/it, lr: 1.0e-04 loss: 1.120e-01]
Xiang_wan_2_2_lora:   1%|3                                         | 285/30000 [40:00<59:05:51,  7.16s/it, lr: 1.0e-04 loss: 9.150e-02]
Xiang_wan_2_2_lora:   1%|3                                         | 285/30000 [40:00<59:05:51,  7.16s/it, lr: 1.0e-04 loss: 9.150e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 286/30000 [40:00<59:29:24,  7.21s/it, lr: 1.0e-04 loss: 9.150e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 286/30000 [40:00<59:29:24,  7.21s/it, lr: 1.0e-04 loss: 9.150e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 286/30000 [40:07<59:29:24,  7.21s/it, lr: 1.0e-04 loss: 7.917e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 286/30000 [40:07<59:29:24,  7.21s/it, lr: 1.0e-04 loss: 7.917e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 287/30000 [40:07<60:00:29,  7.27s/it, lr: 1.0e-04 loss: 7.917e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 287/30000 [40:07<60:00:29,  7.27s/it, lr: 1.0e-04 loss: 7.917e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 287/30000 [40:14<60:00:29,  7.27s/it, lr: 1.0e-04 loss: 1.312e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 287/30000 [40:14<60:00:29,  7.27s/it, lr: 1.0e-04 loss: 1.312e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 288/30000 [40:14<58:33:52,  7.10s/it, lr: 1.0e-04 loss: 1.312e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 288/30000 [40:14<58:33:52,  7.10s/it, lr: 1.0e-04 loss: 1.312e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 288/30000 [40:23<58:33:52,  7.10s/it, lr: 1.0e-04 loss: 4.537e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 288/30000 [40:23<58:33:52,  7.10s/it, lr: 1.0e-04 loss: 4.537e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 289/30000 [40:23<64:04:20,  7.76s/it, lr: 1.0e-04 loss: 4.537e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 289/30000 [40:23<64:04:20,  7.76s/it, lr: 1.0e-04 loss: 4.537e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 289/30000 [40:33<64:04:20,  7.76s/it, lr: 1.0e-04 loss: 1.007e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 289/30000 [40:33<64:04:20,  7.76s/it, lr: 1.0e-04 loss: 1.007e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 290/30000 [40:40<64:04:12,  7.76s/it, lr: 1.0e-04 loss: 8.379e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 290/30000 [40:40<64:04:12,  7.76s/it, lr: 1.0e-04 loss: 8.379e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 291/30000 [40:40<47:49:01,  5.79s/it, lr: 1.0e-04 loss: 8.379e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 291/30000 [40:40<47:49:01,  5.79s/it, lr: 1.0e-04 loss: 8.379e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 291/30000 [40:50<47:49:01,  5.79s/it, lr: 1.0e-04 loss: 7.496e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 291/30000 [40:50<47:49:01,  5.79s/it, lr: 1.0e-04 loss: 7.496e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 292/30000 [40:50<55:32:04,  6.73s/it, lr: 1.0e-04 loss: 7.496e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 292/30000 [40:50<55:32:04,  6.73s/it, lr: 1.0e-04 loss: 7.496e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 292/30000 [40:59<55:32:04,  6.73s/it, lr: 1.0e-04 loss: 6.143e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 292/30000 [40:59<55:32:04,  6.73s/it, lr: 1.0e-04 loss: 6.143e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 293/30000 [40:59<61:16:26,  7.43s/it, lr: 1.0e-04 loss: 6.143e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 293/30000 [40:59<61:16:26,  7.43s/it, lr: 1.0e-04 loss: 6.143e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 293/30000 [41:09<61:16:26,  7.43s/it, lr: 1.0e-04 loss: 6.431e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 293/30000 [41:09<61:16:26,  7.43s/it, lr: 1.0e-04 loss: 6.431e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 294/30000 [41:09<65:47:16,  7.97s/it, lr: 1.0e-04 loss: 6.431e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 294/30000 [41:09<65:47:16,  7.97s/it, lr: 1.0e-04 loss: 6.431e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 294/30000 [41:19<65:47:16,  7.97s/it, lr: 1.0e-04 loss: 6.931e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 294/30000 [41:19<65:47:16,  7.97s/it, lr: 1.0e-04 loss: 6.931e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 295/30000 [41:19<71:05:06,  8.61s/it, lr: 1.0e-04 loss: 6.931e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 295/30000 [41:19<71:05:06,  8.61s/it, lr: 1.0e-04 loss: 6.931e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 295/30000 [41:28<71:05:06,  8.61s/it, lr: 1.0e-04 loss: 4.888e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 295/30000 [41:28<71:05:06,  8.61s/it, lr: 1.0e-04 loss: 4.888e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 296/30000 [41:28<73:09:41,  8.87s/it, lr: 1.0e-04 loss: 4.888e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 296/30000 [41:28<73:09:41,  8.87s/it, lr: 1.0e-04 loss: 4.888e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 296/30000 [41:36<73:09:41,  8.87s/it, lr: 1.0e-04 loss: 5.448e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 296/30000 [41:36<73:09:41,  8.87s/it, lr: 1.0e-04 loss: 5.448e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 297/30000 [41:36<69:14:52,  8.39s/it, lr: 1.0e-04 loss: 5.448e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 297/30000 [41:36<69:14:52,  8.39s/it, lr: 1.0e-04 loss: 5.448e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 297/30000 [41:42<69:14:52,  8.39s/it, lr: 1.0e-04 loss: 9.506e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 297/30000 [41:42<69:14:52,  8.39s/it, lr: 1.0e-04 loss: 9.506e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 298/30000 [41:42<64:56:25,  7.87s/it, lr: 1.0e-04 loss: 9.506e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 298/30000 [41:42<64:56:25,  7.87s/it, lr: 1.0e-04 loss: 9.506e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 298/30000 [41:51<64:56:25,  7.87s/it, lr: 1.0e-04 loss: 9.443e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 298/30000 [41:51<64:56:25,  7.87s/it, lr: 1.0e-04 loss: 9.443e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 299/30000 [41:51<68:32:10,  8.31s/it, lr: 1.0e-04 loss: 9.443e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 299/30000 [41:51<68:32:10,  8.31s/it, lr: 1.0e-04 loss: 9.443e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 299/30000 [41:58<68:32:10,  8.31s/it, lr: 1.0e-04 loss: 1.217e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 299/30000 [41:58<68:32:10,  8.31s/it, lr: 1.0e-04 loss: 1.217e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 300/30000 [42:08<68:32:01,  8.31s/it, lr: 1.0e-04 loss: 1.809e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 300/30000 [42:08<68:32:01,  8.31s/it, lr: 1.0e-04 loss: 1.809e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 301/30000 [42:08<56:06:22,  6.80s/it, lr: 1.0e-04 loss: 1.809e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 301/30000 [42:08<56:06:22,  6.80s/it, lr: 1.0e-04 loss: 1.809e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 301/30000 [42:16<56:06:22,  6.80s/it, lr: 1.0e-04 loss: 9.420e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 301/30000 [42:16<56:06:22,  6.80s/it, lr: 1.0e-04 loss: 9.420e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 302/30000 [42:16<57:10:39,  6.93s/it, lr: 1.0e-04 loss: 9.420e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 302/30000 [42:16<57:10:39,  6.93s/it, lr: 1.0e-04 loss: 9.420e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 302/30000 [42:25<57:10:39,  6.93s/it, lr: 1.0e-04 loss: 7.120e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 302/30000 [42:25<57:10:39,  6.93s/it, lr: 1.0e-04 loss: 7.120e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 303/30000 [42:25<62:29:39,  7.58s/it, lr: 1.0e-04 loss: 7.120e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 303/30000 [42:25<62:29:39,  7.58s/it, lr: 1.0e-04 loss: 7.120e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 303/30000 [42:35<62:29:39,  7.58s/it, lr: 1.0e-04 loss: 8.564e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 303/30000 [42:35<62:29:39,  7.58s/it, lr: 1.0e-04 loss: 8.564e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 304/30000 [42:35<67:07:40,  8.14s/it, lr: 1.0e-04 loss: 8.564e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 304/30000 [42:35<67:07:40,  8.14s/it, lr: 1.0e-04 loss: 8.564e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 304/30000 [42:45<67:07:40,  8.14s/it, lr: 1.0e-04 loss: 1.636e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 304/30000 [42:45<67:07:40,  8.14s/it, lr: 1.0e-04 loss: 1.636e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 305/30000 [42:45<71:48:05,  8.70s/it, lr: 1.0e-04 loss: 1.636e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 305/30000 [42:45<71:48:05,  8.70s/it, lr: 1.0e-04 loss: 1.636e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 305/30000 [42:54<71:48:05,  8.70s/it, lr: 1.0e-04 loss: 6.531e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 305/30000 [42:54<71:48:05,  8.70s/it, lr: 1.0e-04 loss: 6.531e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 306/30000 [42:54<73:27:18,  8.91s/it, lr: 1.0e-04 loss: 6.531e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 306/30000 [42:54<73:27:18,  8.91s/it, lr: 1.0e-04 loss: 6.531e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 306/30000 [43:04<73:27:18,  8.91s/it, lr: 1.0e-04 loss: 4.919e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 306/30000 [43:04<73:27:18,  8.91s/it, lr: 1.0e-04 loss: 4.919e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 307/30000 [43:04<74:42:00,  9.06s/it, lr: 1.0e-04 loss: 4.919e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 307/30000 [43:04<74:42:00,  9.06s/it, lr: 1.0e-04 loss: 4.919e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 307/30000 [43:11<74:42:00,  9.06s/it, lr: 1.0e-04 loss: 1.005e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 307/30000 [43:11<74:42:00,  9.06s/it, lr: 1.0e-04 loss: 1.005e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 308/30000 [43:11<69:30:35,  8.43s/it, lr: 1.0e-04 loss: 1.005e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 308/30000 [43:11<69:30:35,  8.43s/it, lr: 1.0e-04 loss: 1.005e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 308/30000 [43:21<69:30:35,  8.43s/it, lr: 1.0e-04 loss: 3.157e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 308/30000 [43:21<69:30:35,  8.43s/it, lr: 1.0e-04 loss: 3.157e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 309/30000 [43:21<74:00:10,  8.97s/it, lr: 1.0e-04 loss: 3.157e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 309/30000 [43:21<74:00:10,  8.97s/it, lr: 1.0e-04 loss: 3.157e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 309/30000 [43:31<74:00:10,  8.97s/it, lr: 1.0e-04 loss: 1.040e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 309/30000 [43:31<74:00:10,  8.97s/it, lr: 1.0e-04 loss: 1.040e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 310/30000 [43:40<74:00:01,  8.97s/it, lr: 1.0e-04 loss: 9.278e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 310/30000 [43:40<74:00:01,  8.97s/it, lr: 1.0e-04 loss: 9.278e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 311/30000 [43:40<58:07:49,  7.05s/it, lr: 1.0e-04 loss: 9.278e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 311/30000 [43:40<58:07:49,  7.05s/it, lr: 1.0e-04 loss: 9.278e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 311/30000 [43:47<58:07:49,  7.05s/it, lr: 1.0e-04 loss: 1.146e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 311/30000 [43:47<58:07:49,  7.05s/it, lr: 1.0e-04 loss: 1.146e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 312/30000 [43:47<57:23:23,  6.96s/it, lr: 1.0e-04 loss: 1.146e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 312/30000 [43:47<57:23:23,  6.96s/it, lr: 1.0e-04 loss: 1.146e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 312/30000 [43:57<57:23:23,  6.96s/it, lr: 1.0e-04 loss: 4.190e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 312/30000 [43:57<57:23:23,  6.96s/it, lr: 1.0e-04 loss: 4.190e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 313/30000 [43:57<62:49:05,  7.62s/it, lr: 1.0e-04 loss: 4.190e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 313/30000 [43:57<62:49:05,  7.62s/it, lr: 1.0e-04 loss: 4.190e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 313/30000 [44:06<62:49:05,  7.62s/it, lr: 1.0e-04 loss: 7.985e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 313/30000 [44:06<62:49:05,  7.62s/it, lr: 1.0e-04 loss: 7.985e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 314/30000 [44:06<67:52:22,  8.23s/it, lr: 1.0e-04 loss: 7.985e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 314/30000 [44:06<67:52:22,  8.23s/it, lr: 1.0e-04 loss: 7.985e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 314/30000 [44:16<67:52:22,  8.23s/it, lr: 1.0e-04 loss: 4.601e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 314/30000 [44:16<67:52:22,  8.23s/it, lr: 1.0e-04 loss: 4.601e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 315/30000 [44:16<71:23:46,  8.66s/it, lr: 1.0e-04 loss: 4.601e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 315/30000 [44:16<71:23:46,  8.66s/it, lr: 1.0e-04 loss: 4.601e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 315/30000 [44:26<71:23:46,  8.66s/it, lr: 1.0e-04 loss: 1.248e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 315/30000 [44:26<71:23:46,  8.66s/it, lr: 1.0e-04 loss: 1.248e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 316/30000 [44:26<74:22:38,  9.02s/it, lr: 1.0e-04 loss: 1.248e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 316/30000 [44:26<74:22:38,  9.02s/it, lr: 1.0e-04 loss: 1.248e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 316/30000 [44:34<74:22:38,  9.02s/it, lr: 1.0e-04 loss: 8.042e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 316/30000 [44:34<74:22:38,  9.02s/it, lr: 1.0e-04 loss: 8.042e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 317/30000 [44:34<71:10:33,  8.63s/it, lr: 1.0e-04 loss: 8.042e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 317/30000 [44:34<71:10:33,  8.63s/it, lr: 1.0e-04 loss: 8.042e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 317/30000 [44:40<71:10:33,  8.63s/it, lr: 1.0e-04 loss: 2.428e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 317/30000 [44:40<71:10:33,  8.63s/it, lr: 1.0e-04 loss: 2.428e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 318/30000 [44:40<66:23:44,  8.05s/it, lr: 1.0e-04 loss: 2.428e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 318/30000 [44:40<66:23:44,  8.05s/it, lr: 1.0e-04 loss: 2.428e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 318/30000 [44:47<66:23:44,  8.05s/it, lr: 1.0e-04 loss: 6.667e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 318/30000 [44:47<66:23:44,  8.05s/it, lr: 1.0e-04 loss: 6.667e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 319/30000 [44:47<62:54:49,  7.63s/it, lr: 1.0e-04 loss: 6.667e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 319/30000 [44:47<62:54:49,  7.63s/it, lr: 1.0e-04 loss: 6.667e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 319/30000 [44:57<62:54:49,  7.63s/it, lr: 1.0e-04 loss: 7.099e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 319/30000 [44:57<62:54:49,  7.63s/it, lr: 1.0e-04 loss: 7.099e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 320/30000 [45:07<62:54:42,  7.63s/it, lr: 1.0e-04 loss: 4.512e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 320/30000 [45:07<62:54:42,  7.63s/it, lr: 1.0e-04 loss: 4.512e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 321/30000 [45:07<52:40:54,  6.39s/it, lr: 1.0e-04 loss: 4.512e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 321/30000 [45:07<52:40:54,  6.39s/it, lr: 1.0e-04 loss: 4.512e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 321/30000 [45:16<52:40:54,  6.39s/it, lr: 1.0e-04 loss: 1.123e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 321/30000 [45:16<52:40:54,  6.39s/it, lr: 1.0e-04 loss: 1.123e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 322/30000 [45:16<59:01:28,  7.16s/it, lr: 1.0e-04 loss: 1.123e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 322/30000 [45:16<59:01:28,  7.16s/it, lr: 1.0e-04 loss: 1.123e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 322/30000 [45:23<59:01:28,  7.16s/it, lr: 1.0e-04 loss: 5.575e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 322/30000 [45:23<59:01:28,  7.16s/it, lr: 1.0e-04 loss: 5.575e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 323/30000 [45:23<58:04:57,  7.05s/it, lr: 1.0e-04 loss: 5.575e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 323/30000 [45:23<58:04:57,  7.05s/it, lr: 1.0e-04 loss: 5.575e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 323/30000 [45:33<58:04:57,  7.05s/it, lr: 1.0e-04 loss: 8.375e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 323/30000 [45:33<58:04:57,  7.05s/it, lr: 1.0e-04 loss: 8.375e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 324/30000 [45:33<64:50:26,  7.87s/it, lr: 1.0e-04 loss: 8.375e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 324/30000 [45:33<64:50:26,  7.87s/it, lr: 1.0e-04 loss: 8.375e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 324/30000 [45:41<64:50:26,  7.87s/it, lr: 1.0e-04 loss: 5.288e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 324/30000 [45:41<64:50:26,  7.87s/it, lr: 1.0e-04 loss: 5.288e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 325/30000 [45:41<63:45:58,  7.74s/it, lr: 1.0e-04 loss: 5.288e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 325/30000 [45:41<63:45:58,  7.74s/it, lr: 1.0e-04 loss: 5.288e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 325/30000 [45:50<63:45:58,  7.74s/it, lr: 1.0e-04 loss: 8.317e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 325/30000 [45:50<63:45:58,  7.74s/it, lr: 1.0e-04 loss: 8.317e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 326/30000 [45:50<68:57:09,  8.37s/it, lr: 1.0e-04 loss: 8.317e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 326/30000 [45:50<68:57:09,  8.37s/it, lr: 1.0e-04 loss: 8.317e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 326/30000 [46:00<68:57:09,  8.37s/it, lr: 1.0e-04 loss: 2.156e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 326/30000 [46:00<68:57:09,  8.37s/it, lr: 1.0e-04 loss: 2.156e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 327/30000 [46:00<72:27:52,  8.79s/it, lr: 1.0e-04 loss: 2.156e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 327/30000 [46:00<72:27:52,  8.79s/it, lr: 1.0e-04 loss: 2.156e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 327/30000 [46:10<72:27:52,  8.79s/it, lr: 1.0e-04 loss: 9.486e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 327/30000 [46:10<72:27:52,  8.79s/it, lr: 1.0e-04 loss: 9.486e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 328/30000 [46:10<75:20:36,  9.14s/it, lr: 1.0e-04 loss: 9.486e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 328/30000 [46:10<75:20:36,  9.14s/it, lr: 1.0e-04 loss: 9.486e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 328/30000 [46:20<75:20:36,  9.14s/it, lr: 1.0e-04 loss: 4.258e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 328/30000 [46:20<75:20:36,  9.14s/it, lr: 1.0e-04 loss: 4.258e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 329/30000 [46:20<76:55:40,  9.33s/it, lr: 1.0e-04 loss: 4.258e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 329/30000 [46:20<76:55:40,  9.33s/it, lr: 1.0e-04 loss: 4.258e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 329/30000 [46:30<76:55:40,  9.33s/it, lr: 1.0e-04 loss: 1.059e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 329/30000 [46:30<76:55:40,  9.33s/it, lr: 1.0e-04 loss: 1.059e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 330/30000 [46:39<76:55:31,  9.33s/it, lr: 1.0e-04 loss: 5.310e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 330/30000 [46:39<76:55:31,  9.33s/it, lr: 1.0e-04 loss: 5.310e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 331/30000 [46:39<59:55:49,  7.27s/it, lr: 1.0e-04 loss: 5.310e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 331/30000 [46:39<59:55:49,  7.27s/it, lr: 1.0e-04 loss: 5.310e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 331/30000 [46:49<59:55:49,  7.27s/it, lr: 1.0e-04 loss: 8.188e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 331/30000 [46:49<59:55:49,  7.27s/it, lr: 1.0e-04 loss: 8.188e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 332/30000 [46:49<64:13:44,  7.79s/it, lr: 1.0e-04 loss: 8.188e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 332/30000 [46:49<64:13:44,  7.79s/it, lr: 1.0e-04 loss: 8.188e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 332/30000 [46:56<64:13:44,  7.79s/it, lr: 1.0e-04 loss: 1.266e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 332/30000 [46:56<64:13:44,  7.79s/it, lr: 1.0e-04 loss: 1.266e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 333/30000 [46:56<61:44:04,  7.49s/it, lr: 1.0e-04 loss: 1.266e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 333/30000 [46:56<61:44:04,  7.49s/it, lr: 1.0e-04 loss: 1.266e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 333/30000 [47:02<61:44:04,  7.49s/it, lr: 1.0e-04 loss: 8.364e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 333/30000 [47:02<61:44:04,  7.49s/it, lr: 1.0e-04 loss: 8.364e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 334/30000 [47:02<59:50:49,  7.26s/it, lr: 1.0e-04 loss: 8.364e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 334/30000 [47:02<59:50:49,  7.26s/it, lr: 1.0e-04 loss: 8.364e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 334/30000 [47:12<59:50:49,  7.26s/it, lr: 1.0e-04 loss: 9.033e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 334/30000 [47:12<59:50:49,  7.26s/it, lr: 1.0e-04 loss: 9.033e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 335/30000 [47:12<65:56:52,  8.00s/it, lr: 1.0e-04 loss: 9.033e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 335/30000 [47:12<65:56:52,  8.00s/it, lr: 1.0e-04 loss: 9.033e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 335/30000 [47:22<65:56:52,  8.00s/it, lr: 1.0e-04 loss: 7.296e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 335/30000 [47:22<65:56:52,  8.00s/it, lr: 1.0e-04 loss: 7.296e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 336/30000 [47:22<69:22:02,  8.42s/it, lr: 1.0e-04 loss: 7.296e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 336/30000 [47:22<69:22:02,  8.42s/it, lr: 1.0e-04 loss: 7.296e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 336/30000 [47:31<69:22:02,  8.42s/it, lr: 1.0e-04 loss: 7.646e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 336/30000 [47:31<69:22:02,  8.42s/it, lr: 1.0e-04 loss: 7.646e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 337/30000 [47:31<72:01:01,  8.74s/it, lr: 1.0e-04 loss: 7.646e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 337/30000 [47:31<72:01:01,  8.74s/it, lr: 1.0e-04 loss: 7.646e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 337/30000 [47:41<72:01:01,  8.74s/it, lr: 1.0e-04 loss: 6.387e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 337/30000 [47:41<72:01:01,  8.74s/it, lr: 1.0e-04 loss: 6.387e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 338/30000 [47:41<74:31:16,  9.04s/it, lr: 1.0e-04 loss: 6.387e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 338/30000 [47:41<74:31:16,  9.04s/it, lr: 1.0e-04 loss: 6.387e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 338/30000 [47:50<74:31:16,  9.04s/it, lr: 1.0e-04 loss: 8.066e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 338/30000 [47:50<74:31:16,  9.04s/it, lr: 1.0e-04 loss: 8.066e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 339/30000 [47:50<75:28:42,  9.16s/it, lr: 1.0e-04 loss: 8.066e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 339/30000 [47:50<75:28:42,  9.16s/it, lr: 1.0e-04 loss: 8.066e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 339/30000 [47:57<75:28:42,  9.16s/it, lr: 1.0e-04 loss: 8.673e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 339/30000 [47:57<75:28:42,  9.16s/it, lr: 1.0e-04 loss: 8.673e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 340/30000 [48:07<75:28:33,  9.16s/it, lr: 1.0e-04 loss: 5.867e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 340/30000 [48:07<75:28:33,  9.16s/it, lr: 1.0e-04 loss: 5.867e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 341/30000 [48:07<58:56:26,  7.15s/it, lr: 1.0e-04 loss: 5.867e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 341/30000 [48:07<58:56:26,  7.15s/it, lr: 1.0e-04 loss: 5.867e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 341/30000 [48:14<58:56:26,  7.15s/it, lr: 1.0e-04 loss: 9.967e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 341/30000 [48:14<58:56:26,  7.15s/it, lr: 1.0e-04 loss: 9.967e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 342/30000 [48:14<59:18:50,  7.20s/it, lr: 1.0e-04 loss: 9.967e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 342/30000 [48:14<59:18:50,  7.20s/it, lr: 1.0e-04 loss: 9.967e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 342/30000 [48:24<59:18:50,  7.20s/it, lr: 1.0e-04 loss: 5.419e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 342/30000 [48:24<59:18:50,  7.20s/it, lr: 1.0e-04 loss: 5.419e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 343/30000 [48:24<65:44:19,  7.98s/it, lr: 1.0e-04 loss: 5.419e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 343/30000 [48:24<65:44:19,  7.98s/it, lr: 1.0e-04 loss: 5.419e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 343/30000 [48:35<65:44:19,  7.98s/it, lr: 1.0e-04 loss: 9.725e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 343/30000 [48:35<65:44:19,  7.98s/it, lr: 1.0e-04 loss: 9.725e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 344/30000 [48:35<70:31:23,  8.56s/it, lr: 1.0e-04 loss: 9.725e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 344/30000 [48:35<70:31:23,  8.56s/it, lr: 1.0e-04 loss: 9.725e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 344/30000 [48:41<70:31:23,  8.56s/it, lr: 1.0e-04 loss: 7.238e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 344/30000 [48:41<70:31:23,  8.56s/it, lr: 1.0e-04 loss: 7.238e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 345/30000 [48:41<66:17:29,  8.05s/it, lr: 1.0e-04 loss: 7.238e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 345/30000 [48:41<66:17:29,  8.05s/it, lr: 1.0e-04 loss: 7.238e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 345/30000 [48:48<66:17:29,  8.05s/it, lr: 1.0e-04 loss: 9.631e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 345/30000 [48:48<66:17:29,  8.05s/it, lr: 1.0e-04 loss: 9.631e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 346/30000 [48:48<63:41:36,  7.73s/it, lr: 1.0e-04 loss: 9.631e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 346/30000 [48:48<63:41:36,  7.73s/it, lr: 1.0e-04 loss: 9.631e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 346/30000 [48:58<63:41:36,  7.73s/it, lr: 1.0e-04 loss: 5.581e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 346/30000 [48:58<63:41:36,  7.73s/it, lr: 1.0e-04 loss: 5.581e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 347/30000 [48:58<68:15:50,  8.29s/it, lr: 1.0e-04 loss: 5.581e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 347/30000 [48:58<68:15:50,  8.29s/it, lr: 1.0e-04 loss: 5.581e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 347/30000 [49:05<68:15:50,  8.29s/it, lr: 1.0e-04 loss: 1.240e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 347/30000 [49:05<68:15:50,  8.29s/it, lr: 1.0e-04 loss: 1.240e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 348/30000 [49:05<64:26:46,  7.82s/it, lr: 1.0e-04 loss: 1.240e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 348/30000 [49:05<64:26:46,  7.82s/it, lr: 1.0e-04 loss: 1.240e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 348/30000 [49:11<64:26:46,  7.82s/it, lr: 1.0e-04 loss: 8.021e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 348/30000 [49:11<64:26:46,  7.82s/it, lr: 1.0e-04 loss: 8.021e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 349/30000 [49:11<61:37:42,  7.48s/it, lr: 1.0e-04 loss: 8.021e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 349/30000 [49:11<61:37:42,  7.48s/it, lr: 1.0e-04 loss: 8.021e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 349/30000 [49:18<61:37:42,  7.48s/it, lr: 1.0e-04 loss: 9.607e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 349/30000 [49:18<61:37:42,  7.48s/it, lr: 1.0e-04 loss: 9.607e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 350/30000 [49:29<61:37:35,  7.48s/it, lr: 1.0e-04 loss: 5.824e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 350/30000 [49:29<61:37:35,  7.48s/it, lr: 1.0e-04 loss: 5.824e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 351/30000 [49:29<52:23:47,  6.36s/it, lr: 1.0e-04 loss: 5.824e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 351/30000 [49:29<52:23:47,  6.36s/it, lr: 1.0e-04 loss: 5.824e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 351/30000 [49:38<52:23:47,  6.36s/it, lr: 1.0e-04 loss: 8.709e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 351/30000 [49:38<52:23:47,  6.36s/it, lr: 1.0e-04 loss: 8.709e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 352/30000 [49:38<58:33:39,  7.11s/it, lr: 1.0e-04 loss: 8.709e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 352/30000 [49:38<58:33:39,  7.11s/it, lr: 1.0e-04 loss: 8.709e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 352/30000 [49:48<58:33:39,  7.11s/it, lr: 1.0e-04 loss: 4.614e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 352/30000 [49:48<58:33:39,  7.11s/it, lr: 1.0e-04 loss: 4.614e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 353/30000 [49:48<63:57:43,  7.77s/it, lr: 1.0e-04 loss: 4.614e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 353/30000 [49:48<63:57:43,  7.77s/it, lr: 1.0e-04 loss: 4.614e-02]
Xiang_wan_2_2_lora:   1%|4                                         | 353/30000 [49:57<63:57:43,  7.77s/it, lr: 1.0e-04 loss: 1.031e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 353/30000 [49:57<63:57:43,  7.77s/it, lr: 1.0e-04 loss: 1.031e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 354/30000 [49:57<68:23:02,  8.30s/it, lr: 1.0e-04 loss: 1.031e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 354/30000 [49:57<68:23:02,  8.30s/it, lr: 1.0e-04 loss: 1.031e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 354/30000 [50:04<68:23:02,  8.30s/it, lr: 1.0e-04 loss: 1.859e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 354/30000 [50:04<68:23:02,  8.30s/it, lr: 1.0e-04 loss: 1.859e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 355/30000 [50:04<64:40:54,  7.85s/it, lr: 1.0e-04 loss: 1.859e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 355/30000 [50:04<64:40:54,  7.85s/it, lr: 1.0e-04 loss: 1.859e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 355/30000 [50:11<64:40:54,  7.85s/it, lr: 1.0e-04 loss: 1.221e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 355/30000 [50:11<64:40:54,  7.85s/it, lr: 1.0e-04 loss: 1.221e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 356/30000 [50:11<62:36:02,  7.60s/it, lr: 1.0e-04 loss: 1.221e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 356/30000 [50:11<62:36:02,  7.60s/it, lr: 1.0e-04 loss: 1.221e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 356/30000 [50:18<62:36:02,  7.60s/it, lr: 1.0e-04 loss: 1.345e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 356/30000 [50:18<62:36:02,  7.60s/it, lr: 1.0e-04 loss: 1.345e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 357/30000 [50:18<61:56:00,  7.52s/it, lr: 1.0e-04 loss: 1.345e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 357/30000 [50:18<61:56:00,  7.52s/it, lr: 1.0e-04 loss: 1.345e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 357/30000 [50:25<61:56:00,  7.52s/it, lr: 1.0e-04 loss: 1.308e-01]
Xiang_wan_2_2_lora:   1%|4                                         | 357/30000 [50:25<61:56:00,  7.52s/it, lr: 1.0e-04 loss: 1.308e-01]
Xiang_wan_2_2_lora:   1%|5                                         | 358/30000 [50:25<60:32:57,  7.35s/it, lr: 1.0e-04 loss: 1.308e-01]
Xiang_wan_2_2_lora:   1%|5                                         | 358/30000 [50:25<60:32:57,  7.35s/it, lr: 1.0e-04 loss: 1.308e-01]
Xiang_wan_2_2_lora:   1%|5                                         | 358/30000 [50:35<60:32:57,  7.35s/it, lr: 1.0e-04 loss: 5.846e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 358/30000 [50:35<60:32:57,  7.35s/it, lr: 1.0e-04 loss: 5.846e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 359/30000 [50:35<67:05:18,  8.15s/it, lr: 1.0e-04 loss: 5.846e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 359/30000 [50:35<67:05:18,  8.15s/it, lr: 1.0e-04 loss: 5.846e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 359/30000 [50:45<67:05:18,  8.15s/it, lr: 1.0e-04 loss: 8.954e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 359/30000 [50:45<67:05:18,  8.15s/it, lr: 1.0e-04 loss: 8.954e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 360/30000 [50:52<67:05:10,  8.15s/it, lr: 1.0e-04 loss: 6.221e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 360/30000 [50:52<67:05:10,  8.15s/it, lr: 1.0e-04 loss: 6.221e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 361/30000 [50:52<50:04:29,  6.08s/it, lr: 1.0e-04 loss: 6.221e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 361/30000 [50:52<50:04:29,  6.08s/it, lr: 1.0e-04 loss: 6.221e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 361/30000 [50:59<50:04:29,  6.08s/it, lr: 1.0e-04 loss: 1.005e-01]
Xiang_wan_2_2_lora:   1%|5                                         | 361/30000 [50:59<50:04:29,  6.08s/it, lr: 1.0e-04 loss: 1.005e-01]
Xiang_wan_2_2_lora:   1%|5                                         | 362/30000 [50:59<51:50:15,  6.30s/it, lr: 1.0e-04 loss: 1.005e-01]
Xiang_wan_2_2_lora:   1%|5                                         | 362/30000 [50:59<51:50:15,  6.30s/it, lr: 1.0e-04 loss: 1.005e-01]
Xiang_wan_2_2_lora:   1%|5                                         | 362/30000 [51:07<51:50:15,  6.30s/it, lr: 1.0e-04 loss: 9.046e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 362/30000 [51:07<51:50:15,  6.30s/it, lr: 1.0e-04 loss: 9.046e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 363/30000 [51:07<53:49:03,  6.54s/it, lr: 1.0e-04 loss: 9.046e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 363/30000 [51:07<53:49:03,  6.54s/it, lr: 1.0e-04 loss: 9.046e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 363/30000 [51:13<53:49:03,  6.54s/it, lr: 1.0e-04 loss: 1.097e-01]
Xiang_wan_2_2_lora:   1%|5                                         | 363/30000 [51:13<53:49:03,  6.54s/it, lr: 1.0e-04 loss: 1.097e-01]
Xiang_wan_2_2_lora:   1%|5                                         | 364/30000 [51:13<54:11:51,  6.58s/it, lr: 1.0e-04 loss: 1.097e-01]
Xiang_wan_2_2_lora:   1%|5                                         | 364/30000 [51:13<54:11:51,  6.58s/it, lr: 1.0e-04 loss: 1.097e-01]
Xiang_wan_2_2_lora:   1%|5                                         | 364/30000 [51:20<54:11:51,  6.58s/it, lr: 1.0e-04 loss: 7.578e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 364/30000 [51:20<54:11:51,  6.58s/it, lr: 1.0e-04 loss: 7.578e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 365/30000 [51:20<54:48:37,  6.66s/it, lr: 1.0e-04 loss: 7.578e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 365/30000 [51:20<54:48:37,  6.66s/it, lr: 1.0e-04 loss: 7.578e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 365/30000 [51:30<54:48:37,  6.66s/it, lr: 1.0e-04 loss: 6.272e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 365/30000 [51:30<54:48:37,  6.66s/it, lr: 1.0e-04 loss: 6.272e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 366/30000 [51:30<61:59:23,  7.53s/it, lr: 1.0e-04 loss: 6.272e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 366/30000 [51:30<61:59:23,  7.53s/it, lr: 1.0e-04 loss: 6.272e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 366/30000 [51:40<61:59:23,  7.53s/it, lr: 1.0e-04 loss: 3.799e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 366/30000 [51:40<61:59:23,  7.53s/it, lr: 1.0e-04 loss: 3.799e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 367/30000 [51:40<67:56:24,  8.25s/it, lr: 1.0e-04 loss: 3.799e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 367/30000 [51:40<67:56:24,  8.25s/it, lr: 1.0e-04 loss: 3.799e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 367/30000 [51:50<67:56:24,  8.25s/it, lr: 1.0e-04 loss: 7.846e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 367/30000 [51:50<67:56:24,  8.25s/it, lr: 1.0e-04 loss: 7.846e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 368/30000 [51:50<71:48:20,  8.72s/it, lr: 1.0e-04 loss: 7.846e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 368/30000 [51:50<71:48:20,  8.72s/it, lr: 1.0e-04 loss: 7.846e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 368/30000 [51:59<71:48:20,  8.72s/it, lr: 1.0e-04 loss: 9.414e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 368/30000 [51:59<71:48:20,  8.72s/it, lr: 1.0e-04 loss: 9.414e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 369/30000 [51:59<73:39:45,  8.95s/it, lr: 1.0e-04 loss: 9.414e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 369/30000 [51:59<73:39:45,  8.95s/it, lr: 1.0e-04 loss: 9.414e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 369/30000 [52:06<73:39:45,  8.95s/it, lr: 1.0e-04 loss: 9.642e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 369/30000 [52:06<73:39:45,  8.95s/it, lr: 1.0e-04 loss: 9.642e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 370/30000 [52:15<73:39:36,  8.95s/it, lr: 1.0e-04 loss: 6.229e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 370/30000 [52:15<73:39:36,  8.95s/it, lr: 1.0e-04 loss: 6.229e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 371/30000 [52:15<57:39:51,  7.01s/it, lr: 1.0e-04 loss: 6.229e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 371/30000 [52:15<57:39:51,  7.01s/it, lr: 1.0e-04 loss: 6.229e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 371/30000 [52:25<57:39:51,  7.01s/it, lr: 1.0e-04 loss: 5.639e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 371/30000 [52:25<57:39:51,  7.01s/it, lr: 1.0e-04 loss: 5.639e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 372/30000 [52:25<63:15:00,  7.69s/it, lr: 1.0e-04 loss: 5.639e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 372/30000 [52:25<63:15:00,  7.69s/it, lr: 1.0e-04 loss: 5.639e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 372/30000 [52:33<63:15:00,  7.69s/it, lr: 1.0e-04 loss: 6.578e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 372/30000 [52:33<63:15:00,  7.69s/it, lr: 1.0e-04 loss: 6.578e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 373/30000 [52:33<62:36:19,  7.61s/it, lr: 1.0e-04 loss: 6.578e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 373/30000 [52:33<62:36:19,  7.61s/it, lr: 1.0e-04 loss: 6.578e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 373/30000 [52:42<62:36:19,  7.61s/it, lr: 1.0e-04 loss: 8.453e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 373/30000 [52:42<62:36:19,  7.61s/it, lr: 1.0e-04 loss: 8.453e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 374/30000 [52:42<66:41:11,  8.10s/it, lr: 1.0e-04 loss: 8.453e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 374/30000 [52:42<66:41:11,  8.10s/it, lr: 1.0e-04 loss: 8.453e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 374/30000 [52:52<66:41:11,  8.10s/it, lr: 1.0e-04 loss: 7.524e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 374/30000 [52:52<66:41:11,  8.10s/it, lr: 1.0e-04 loss: 7.524e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 375/30000 [52:52<70:05:44,  8.52s/it, lr: 1.0e-04 loss: 7.524e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 375/30000 [52:52<70:05:44,  8.52s/it, lr: 1.0e-04 loss: 7.524e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 375/30000 [52:58<70:05:44,  8.52s/it, lr: 1.0e-04 loss: 1.157e-01]
Xiang_wan_2_2_lora:   1%|5                                         | 375/30000 [52:58<70:05:44,  8.52s/it, lr: 1.0e-04 loss: 1.157e-01]
Xiang_wan_2_2_lora:   1%|5                                         | 376/30000 [52:58<65:45:22,  7.99s/it, lr: 1.0e-04 loss: 1.157e-01]
Xiang_wan_2_2_lora:   1%|5                                         | 376/30000 [52:58<65:45:22,  7.99s/it, lr: 1.0e-04 loss: 1.157e-01]
Xiang_wan_2_2_lora:   1%|5                                         | 376/30000 [53:08<65:45:22,  7.99s/it, lr: 1.0e-04 loss: 3.783e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 376/30000 [53:08<65:45:22,  7.99s/it, lr: 1.0e-04 loss: 3.783e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 377/30000 [53:08<68:57:10,  8.38s/it, lr: 1.0e-04 loss: 3.783e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 377/30000 [53:08<68:57:10,  8.38s/it, lr: 1.0e-04 loss: 3.783e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 377/30000 [53:14<68:57:10,  8.38s/it, lr: 1.0e-04 loss: 1.379e-01]
Xiang_wan_2_2_lora:   1%|5                                         | 377/30000 [53:14<68:57:10,  8.38s/it, lr: 1.0e-04 loss: 1.379e-01]
Xiang_wan_2_2_lora:   1%|5                                         | 378/30000 [53:14<65:02:50,  7.91s/it, lr: 1.0e-04 loss: 1.379e-01]
Xiang_wan_2_2_lora:   1%|5                                         | 378/30000 [53:14<65:02:50,  7.91s/it, lr: 1.0e-04 loss: 1.379e-01]
Xiang_wan_2_2_lora:   1%|5                                         | 378/30000 [53:21<65:02:50,  7.91s/it, lr: 1.0e-04 loss: 9.904e-01]
Xiang_wan_2_2_lora:   1%|5                                         | 378/30000 [53:21<65:02:50,  7.91s/it, lr: 1.0e-04 loss: 9.904e-01]
Xiang_wan_2_2_lora:   1%|5                                         | 379/30000 [53:21<61:57:33,  7.53s/it, lr: 1.0e-04 loss: 9.904e-01]
Xiang_wan_2_2_lora:   1%|5                                         | 379/30000 [53:21<61:57:33,  7.53s/it, lr: 1.0e-04 loss: 9.904e-01]
Xiang_wan_2_2_lora:   1%|5                                         | 379/30000 [53:30<61:57:33,  7.53s/it, lr: 1.0e-04 loss: 6.794e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 379/30000 [53:30<61:57:33,  7.53s/it, lr: 1.0e-04 loss: 6.794e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 380/30000 [53:40<61:57:26,  7.53s/it, lr: 1.0e-04 loss: 4.819e-01]
Xiang_wan_2_2_lora:   1%|5                                         | 380/30000 [53:40<61:57:26,  7.53s/it, lr: 1.0e-04 loss: 4.819e-01]
Xiang_wan_2_2_lora:   1%|5                                         | 381/30000 [53:40<51:23:14,  6.25s/it, lr: 1.0e-04 loss: 4.819e-01]
Xiang_wan_2_2_lora:   1%|5                                         | 381/30000 [53:40<51:23:14,  6.25s/it, lr: 1.0e-04 loss: 4.819e-01]
Xiang_wan_2_2_lora:   1%|5                                         | 381/30000 [53:47<51:23:14,  6.25s/it, lr: 1.0e-04 loss: 8.871e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 381/30000 [53:47<51:23:14,  6.25s/it, lr: 1.0e-04 loss: 8.871e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 382/30000 [53:47<52:43:18,  6.41s/it, lr: 1.0e-04 loss: 8.871e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 382/30000 [53:47<52:43:18,  6.41s/it, lr: 1.0e-04 loss: 8.871e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 382/30000 [53:53<52:43:18,  6.41s/it, lr: 1.0e-04 loss: 6.034e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 382/30000 [53:53<52:43:18,  6.41s/it, lr: 1.0e-04 loss: 6.034e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 383/30000 [53:53<53:06:45,  6.46s/it, lr: 1.0e-04 loss: 6.034e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 383/30000 [53:53<53:06:45,  6.46s/it, lr: 1.0e-04 loss: 6.034e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 383/30000 [54:00<53:06:45,  6.46s/it, lr: 1.0e-04 loss: 7.971e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 383/30000 [54:00<53:06:45,  6.46s/it, lr: 1.0e-04 loss: 7.971e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 384/30000 [54:00<53:34:15,  6.51s/it, lr: 1.0e-04 loss: 7.971e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 384/30000 [54:00<53:34:15,  6.51s/it, lr: 1.0e-04 loss: 7.971e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 384/30000 [54:07<53:34:15,  6.51s/it, lr: 1.0e-04 loss: 1.487e-01]
Xiang_wan_2_2_lora:   1%|5                                         | 384/30000 [54:07<53:34:15,  6.51s/it, lr: 1.0e-04 loss: 1.487e-01]
Xiang_wan_2_2_lora:   1%|5                                         | 385/30000 [54:07<53:46:38,  6.54s/it, lr: 1.0e-04 loss: 1.487e-01]
Xiang_wan_2_2_lora:   1%|5                                         | 385/30000 [54:07<53:46:38,  6.54s/it, lr: 1.0e-04 loss: 1.487e-01]
Xiang_wan_2_2_lora:   1%|5                                         | 385/30000 [54:20<53:46:38,  6.54s/it, lr: 1.0e-04 loss: 9.327e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 385/30000 [54:20<53:46:38,  6.54s/it, lr: 1.0e-04 loss: 9.327e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 386/30000 [54:20<67:34:29,  8.21s/it, lr: 1.0e-04 loss: 9.327e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 386/30000 [54:20<67:34:29,  8.21s/it, lr: 1.0e-04 loss: 9.327e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 386/30000 [54:33<67:34:29,  8.21s/it, lr: 1.0e-04 loss: 6.706e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 386/30000 [54:33<67:34:29,  8.21s/it, lr: 1.0e-04 loss: 6.706e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 387/30000 [54:33<76:55:26,  9.35s/it, lr: 1.0e-04 loss: 6.706e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 387/30000 [54:33<76:55:26,  9.35s/it, lr: 1.0e-04 loss: 6.706e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 387/30000 [54:40<76:55:26,  9.35s/it, lr: 1.0e-04 loss: 9.597e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 387/30000 [54:40<76:55:26,  9.35s/it, lr: 1.0e-04 loss: 9.597e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 388/30000 [54:40<72:30:42,  8.82s/it, lr: 1.0e-04 loss: 9.597e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 388/30000 [54:40<72:30:42,  8.82s/it, lr: 1.0e-04 loss: 9.597e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 388/30000 [54:50<72:30:42,  8.82s/it, lr: 1.0e-04 loss: 5.272e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 388/30000 [54:50<72:30:42,  8.82s/it, lr: 1.0e-04 loss: 5.272e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 389/30000 [54:50<75:53:25,  9.23s/it, lr: 1.0e-04 loss: 5.272e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 389/30000 [54:50<75:53:25,  9.23s/it, lr: 1.0e-04 loss: 5.272e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 389/30000 [54:57<75:53:25,  9.23s/it, lr: 1.0e-04 loss: 1.033e-01]
Xiang_wan_2_2_lora:   1%|5                                         | 389/30000 [54:57<75:53:25,  9.23s/it, lr: 1.0e-04 loss: 1.033e-01]
Xiang_wan_2_2_lora:   1%|5                                         | 390/30000 [55:07<75:53:15,  9.23s/it, lr: 1.0e-04 loss: 7.283e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 390/30000 [55:07<75:53:15,  9.23s/it, lr: 1.0e-04 loss: 7.283e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 391/30000 [55:07<59:07:14,  7.19s/it, lr: 1.0e-04 loss: 7.283e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 391/30000 [55:07<59:07:14,  7.19s/it, lr: 1.0e-04 loss: 7.283e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 391/30000 [55:13<59:07:14,  7.19s/it, lr: 1.0e-04 loss: 9.877e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 391/30000 [55:13<59:07:14,  7.19s/it, lr: 1.0e-04 loss: 9.877e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 392/30000 [55:13<58:09:37,  7.07s/it, lr: 1.0e-04 loss: 9.877e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 392/30000 [55:13<58:09:37,  7.07s/it, lr: 1.0e-04 loss: 9.877e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 392/30000 [55:20<58:09:37,  7.07s/it, lr: 1.0e-04 loss: 9.810e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 392/30000 [55:20<58:09:37,  7.07s/it, lr: 1.0e-04 loss: 9.810e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 393/30000 [55:20<57:04:56,  6.94s/it, lr: 1.0e-04 loss: 9.810e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 393/30000 [55:20<57:04:56,  6.94s/it, lr: 1.0e-04 loss: 9.810e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 393/30000 [55:29<57:04:56,  6.94s/it, lr: 1.0e-04 loss: 5.672e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 393/30000 [55:29<57:04:56,  6.94s/it, lr: 1.0e-04 loss: 5.672e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 394/30000 [55:29<62:36:22,  7.61s/it, lr: 1.0e-04 loss: 5.672e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 394/30000 [55:29<62:36:22,  7.61s/it, lr: 1.0e-04 loss: 5.672e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 394/30000 [55:36<62:36:22,  7.61s/it, lr: 1.0e-04 loss: 9.378e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 394/30000 [55:36<62:36:22,  7.61s/it, lr: 1.0e-04 loss: 9.378e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 395/30000 [55:36<60:23:00,  7.34s/it, lr: 1.0e-04 loss: 9.378e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 395/30000 [55:36<60:23:00,  7.34s/it, lr: 1.0e-04 loss: 9.378e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 395/30000 [55:46<60:23:00,  7.34s/it, lr: 1.0e-04 loss: 8.787e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 395/30000 [55:46<60:23:00,  7.34s/it, lr: 1.0e-04 loss: 8.787e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 396/30000 [55:46<65:32:00,  7.97s/it, lr: 1.0e-04 loss: 8.787e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 396/30000 [55:46<65:32:00,  7.97s/it, lr: 1.0e-04 loss: 8.787e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 396/30000 [55:55<65:32:00,  7.97s/it, lr: 1.0e-04 loss: 4.402e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 396/30000 [55:55<65:32:00,  7.97s/it, lr: 1.0e-04 loss: 4.402e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 397/30000 [55:55<68:51:56,  8.37s/it, lr: 1.0e-04 loss: 4.402e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 397/30000 [55:55<68:51:56,  8.37s/it, lr: 1.0e-04 loss: 4.402e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 397/30000 [56:04<68:51:56,  8.37s/it, lr: 1.0e-04 loss: 1.000e-01]
Xiang_wan_2_2_lora:   1%|5                                         | 397/30000 [56:04<68:51:56,  8.37s/it, lr: 1.0e-04 loss: 1.000e-01]
Xiang_wan_2_2_lora:   1%|5                                         | 398/30000 [56:04<71:15:45,  8.67s/it, lr: 1.0e-04 loss: 1.000e-01]
Xiang_wan_2_2_lora:   1%|5                                         | 398/30000 [56:04<71:15:45,  8.67s/it, lr: 1.0e-04 loss: 1.000e-01]
Xiang_wan_2_2_lora:   1%|5                                         | 398/30000 [56:14<71:15:45,  8.67s/it, lr: 1.0e-04 loss: 4.246e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 398/30000 [56:14<71:15:45,  8.67s/it, lr: 1.0e-04 loss: 4.246e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 399/30000 [56:14<72:53:33,  8.87s/it, lr: 1.0e-04 loss: 4.246e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 399/30000 [56:14<72:53:33,  8.87s/it, lr: 1.0e-04 loss: 4.246e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 399/30000 [56:23<72:53:33,  8.87s/it, lr: 1.0e-04 loss: 8.676e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 399/30000 [56:23<72:53:33,  8.87s/it, lr: 1.0e-04 loss: 8.676e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 400/30000 [56:30<72:53:24,  8.87s/it, lr: 1.0e-04 loss: 7.479e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 400/30000 [56:30<72:53:24,  8.87s/it, lr: 1.0e-04 loss: 7.479e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 401/30000 [56:30<52:00:02,  6.32s/it, lr: 1.0e-04 loss: 7.479e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 401/30000 [56:30<52:00:02,  6.32s/it, lr: 1.0e-04 loss: 7.479e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 401/30000 [56:36<52:00:02,  6.32s/it, lr: 1.0e-04 loss: 9.394e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 401/30000 [56:36<52:00:02,  6.32s/it, lr: 1.0e-04 loss: 9.394e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 402/30000 [56:36<52:31:25,  6.39s/it, lr: 1.0e-04 loss: 9.394e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 402/30000 [56:36<52:31:25,  6.39s/it, lr: 1.0e-04 loss: 9.394e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 402/30000 [56:46<52:31:25,  6.39s/it, lr: 1.0e-04 loss: 3.961e-01]
Xiang_wan_2_2_lora:   1%|5                                         | 402/30000 [56:46<52:31:25,  6.39s/it, lr: 1.0e-04 loss: 3.961e-01]
Xiang_wan_2_2_lora:   1%|5                                         | 403/30000 [56:46<58:53:03,  7.16s/it, lr: 1.0e-04 loss: 3.961e-01]
Xiang_wan_2_2_lora:   1%|5                                         | 403/30000 [56:46<58:53:03,  7.16s/it, lr: 1.0e-04 loss: 3.961e-01]
Xiang_wan_2_2_lora:   1%|5                                         | 403/30000 [56:52<58:53:03,  7.16s/it, lr: 1.0e-04 loss: 7.868e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 403/30000 [56:52<58:53:03,  7.16s/it, lr: 1.0e-04 loss: 7.868e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 404/30000 [56:52<58:17:54,  7.09s/it, lr: 1.0e-04 loss: 7.868e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 404/30000 [56:52<58:17:54,  7.09s/it, lr: 1.0e-04 loss: 7.868e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 404/30000 [56:59<58:17:54,  7.09s/it, lr: 1.0e-04 loss: 5.581e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 404/30000 [56:59<58:17:54,  7.09s/it, lr: 1.0e-04 loss: 5.581e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 405/30000 [56:59<57:10:05,  6.95s/it, lr: 1.0e-04 loss: 5.581e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 405/30000 [56:59<57:10:05,  6.95s/it, lr: 1.0e-04 loss: 5.581e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 405/30000 [57:08<57:10:05,  6.95s/it, lr: 1.0e-04 loss: 6.835e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 405/30000 [57:08<57:10:05,  6.95s/it, lr: 1.0e-04 loss: 6.835e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 406/30000 [57:08<62:47:29,  7.64s/it, lr: 1.0e-04 loss: 6.835e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 406/30000 [57:08<62:47:29,  7.64s/it, lr: 1.0e-04 loss: 6.835e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 406/30000 [57:15<62:47:29,  7.64s/it, lr: 1.0e-04 loss: 7.040e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 406/30000 [57:15<62:47:29,  7.64s/it, lr: 1.0e-04 loss: 7.040e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 407/30000 [57:15<60:16:44,  7.33s/it, lr: 1.0e-04 loss: 7.040e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 407/30000 [57:15<60:16:44,  7.33s/it, lr: 1.0e-04 loss: 7.040e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 407/30000 [57:22<60:16:44,  7.33s/it, lr: 1.0e-04 loss: 9.331e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 407/30000 [57:22<60:16:44,  7.33s/it, lr: 1.0e-04 loss: 9.331e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 408/30000 [57:22<58:26:01,  7.11s/it, lr: 1.0e-04 loss: 9.331e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 408/30000 [57:22<58:26:01,  7.11s/it, lr: 1.0e-04 loss: 9.331e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 408/30000 [57:28<58:26:01,  7.11s/it, lr: 1.0e-04 loss: 6.432e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 408/30000 [57:28<58:26:01,  7.11s/it, lr: 1.0e-04 loss: 6.432e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 409/30000 [57:28<57:02:47,  6.94s/it, lr: 1.0e-04 loss: 6.432e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 409/30000 [57:28<57:02:47,  6.94s/it, lr: 1.0e-04 loss: 6.432e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 409/30000 [57:35<57:02:47,  6.94s/it, lr: 1.0e-04 loss: 1.055e-01]
Xiang_wan_2_2_lora:   1%|5                                         | 409/30000 [57:35<57:02:47,  6.94s/it, lr: 1.0e-04 loss: 1.055e-01]
Xiang_wan_2_2_lora:   1%|5                                         | 410/30000 [57:41<57:02:40,  6.94s/it, lr: 1.0e-04 loss: 5.271e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 410/30000 [57:41<57:02:40,  6.94s/it, lr: 1.0e-04 loss: 5.271e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 411/30000 [57:41<43:40:18,  5.31s/it, lr: 1.0e-04 loss: 5.271e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 411/30000 [57:41<43:40:18,  5.31s/it, lr: 1.0e-04 loss: 5.271e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 411/30000 [57:48<43:40:18,  5.31s/it, lr: 1.0e-04 loss: 1.955e-01]
Xiang_wan_2_2_lora:   1%|5                                         | 411/30000 [57:48<43:40:18,  5.31s/it, lr: 1.0e-04 loss: 1.955e-01]
Xiang_wan_2_2_lora:   1%|5                                         | 412/30000 [57:48<46:12:21,  5.62s/it, lr: 1.0e-04 loss: 1.955e-01]
Xiang_wan_2_2_lora:   1%|5                                         | 412/30000 [57:48<46:12:21,  5.62s/it, lr: 1.0e-04 loss: 1.955e-01]
Xiang_wan_2_2_lora:   1%|5                                         | 412/30000 [57:55<46:12:21,  5.62s/it, lr: 1.0e-04 loss: 6.532e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 412/30000 [57:55<46:12:21,  5.62s/it, lr: 1.0e-04 loss: 6.532e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 413/30000 [57:55<48:14:51,  5.87s/it, lr: 1.0e-04 loss: 6.532e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 413/30000 [57:55<48:14:51,  5.87s/it, lr: 1.0e-04 loss: 6.532e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 413/30000 [58:01<48:14:51,  5.87s/it, lr: 1.0e-04 loss: 8.191e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 413/30000 [58:01<48:14:51,  5.87s/it, lr: 1.0e-04 loss: 8.191e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 414/30000 [58:01<49:50:37,  6.06s/it, lr: 1.0e-04 loss: 8.191e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 414/30000 [58:01<49:50:37,  6.06s/it, lr: 1.0e-04 loss: 8.191e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 414/30000 [58:11<49:50:37,  6.06s/it, lr: 1.0e-04 loss: 3.170e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 414/30000 [58:11<49:50:37,  6.06s/it, lr: 1.0e-04 loss: 3.170e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 415/30000 [58:11<57:24:15,  6.99s/it, lr: 1.0e-04 loss: 3.170e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 415/30000 [58:11<57:24:15,  6.99s/it, lr: 1.0e-04 loss: 3.170e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 415/30000 [58:20<57:24:15,  6.99s/it, lr: 1.0e-04 loss: 1.285e-01]
Xiang_wan_2_2_lora:   1%|5                                         | 415/30000 [58:20<57:24:15,  6.99s/it, lr: 1.0e-04 loss: 1.285e-01]
Xiang_wan_2_2_lora:   1%|5                                         | 416/30000 [58:20<63:07:18,  7.68s/it, lr: 1.0e-04 loss: 1.285e-01]
Xiang_wan_2_2_lora:   1%|5                                         | 416/30000 [58:20<63:07:18,  7.68s/it, lr: 1.0e-04 loss: 1.285e-01]
Xiang_wan_2_2_lora:   1%|5                                         | 416/30000 [58:27<63:07:18,  7.68s/it, lr: 1.0e-04 loss: 4.524e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 416/30000 [58:27<63:07:18,  7.68s/it, lr: 1.0e-04 loss: 4.524e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 417/30000 [58:27<60:31:17,  7.36s/it, lr: 1.0e-04 loss: 4.524e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 417/30000 [58:27<60:31:17,  7.36s/it, lr: 1.0e-04 loss: 4.524e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 417/30000 [58:36<60:31:17,  7.36s/it, lr: 1.0e-04 loss: 6.082e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 417/30000 [58:36<60:31:17,  7.36s/it, lr: 1.0e-04 loss: 6.082e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 418/30000 [58:36<65:49:56,  8.01s/it, lr: 1.0e-04 loss: 6.082e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 418/30000 [58:36<65:49:56,  8.01s/it, lr: 1.0e-04 loss: 6.082e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 418/30000 [58:43<65:49:56,  8.01s/it, lr: 1.0e-04 loss: 6.780e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 418/30000 [58:43<65:49:56,  8.01s/it, lr: 1.0e-04 loss: 6.780e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 419/30000 [58:43<62:36:19,  7.62s/it, lr: 1.0e-04 loss: 6.780e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 419/30000 [58:43<62:36:19,  7.62s/it, lr: 1.0e-04 loss: 6.780e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 419/30000 [58:50<62:36:19,  7.62s/it, lr: 1.0e-04 loss: 1.127e-01]
Xiang_wan_2_2_lora:   1%|5                                         | 419/30000 [58:50<62:36:19,  7.62s/it, lr: 1.0e-04 loss: 1.127e-01]
Xiang_wan_2_2_lora:   1%|5                                         | 420/30000 [58:59<62:36:11,  7.62s/it, lr: 1.0e-04 loss: 6.428e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 420/30000 [58:59<62:36:11,  7.62s/it, lr: 1.0e-04 loss: 6.428e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 421/30000 [58:59<51:28:14,  6.26s/it, lr: 1.0e-04 loss: 6.428e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 421/30000 [58:59<51:28:14,  6.26s/it, lr: 1.0e-04 loss: 6.428e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 421/30000 [59:08<51:28:14,  6.26s/it, lr: 1.0e-04 loss: 8.939e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 421/30000 [59:08<51:28:14,  6.26s/it, lr: 1.0e-04 loss: 8.939e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 422/30000 [59:08<57:54:30,  7.05s/it, lr: 1.0e-04 loss: 8.939e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 422/30000 [59:08<57:54:30,  7.05s/it, lr: 1.0e-04 loss: 8.939e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 422/30000 [59:18<57:54:30,  7.05s/it, lr: 1.0e-04 loss: 5.191e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 422/30000 [59:18<57:54:30,  7.05s/it, lr: 1.0e-04 loss: 5.191e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 423/30000 [59:18<62:54:50,  7.66s/it, lr: 1.0e-04 loss: 5.191e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 423/30000 [59:18<62:54:50,  7.66s/it, lr: 1.0e-04 loss: 5.191e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 423/30000 [59:27<62:54:50,  7.66s/it, lr: 1.0e-04 loss: 9.721e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 423/30000 [59:27<62:54:50,  7.66s/it, lr: 1.0e-04 loss: 9.721e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 424/30000 [59:27<66:49:51,  8.13s/it, lr: 1.0e-04 loss: 9.721e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 424/30000 [59:27<66:49:51,  8.13s/it, lr: 1.0e-04 loss: 9.721e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 424/30000 [59:34<66:49:51,  8.13s/it, lr: 1.0e-04 loss: 6.432e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 424/30000 [59:34<66:49:51,  8.13s/it, lr: 1.0e-04 loss: 6.432e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 425/30000 [59:34<64:02:40,  7.80s/it, lr: 1.0e-04 loss: 6.432e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 425/30000 [59:34<64:02:40,  7.80s/it, lr: 1.0e-04 loss: 6.432e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 425/30000 [59:41<64:02:40,  7.80s/it, lr: 1.0e-04 loss: 1.355e-01]
Xiang_wan_2_2_lora:   1%|5                                         | 425/30000 [59:41<64:02:40,  7.80s/it, lr: 1.0e-04 loss: 1.355e-01]
Xiang_wan_2_2_lora:   1%|5                                         | 426/30000 [59:41<61:19:27,  7.46s/it, lr: 1.0e-04 loss: 1.355e-01]
Xiang_wan_2_2_lora:   1%|5                                         | 426/30000 [59:41<61:19:27,  7.46s/it, lr: 1.0e-04 loss: 1.355e-01]
Xiang_wan_2_2_lora:   1%|5                                         | 426/30000 [59:50<61:19:27,  7.46s/it, lr: 1.0e-04 loss: 6.021e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 426/30000 [59:50<61:19:27,  7.46s/it, lr: 1.0e-04 loss: 6.021e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 427/30000 [59:50<65:46:46,  8.01s/it, lr: 1.0e-04 loss: 6.021e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 427/30000 [59:50<65:46:46,  8.01s/it, lr: 1.0e-04 loss: 6.021e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 427/30000 [59:59<65:46:46,  8.01s/it, lr: 1.0e-04 loss: 8.726e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 427/30000 [59:59<65:46:46,  8.01s/it, lr: 1.0e-04 loss: 8.726e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 428/30000 [59:59<69:16:20,  8.43s/it, lr: 1.0e-04 loss: 8.726e-02]
Xiang_wan_2_2_lora:   1%|5                                         | 428/30000 [59:59<69:16:20,  8.43s/it, lr: 1.0e-04 loss: 8.726e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 428/30000 [1:00:09<69:16:20,  8.43s/it, lr: 1.0e-04 loss: 6.117e-01]
Xiang_wan_2_2_lora:   1%|5                                       | 428/30000 [1:00:09<69:16:20,  8.43s/it, lr: 1.0e-04 loss: 6.117e-01]
Xiang_wan_2_2_lora:   1%|5                                       | 429/30000 [1:00:09<71:31:07,  8.71s/it, lr: 1.0e-04 loss: 6.117e-01]
Xiang_wan_2_2_lora:   1%|5                                       | 429/30000 [1:00:09<71:31:07,  8.71s/it, lr: 1.0e-04 loss: 6.117e-01]
Xiang_wan_2_2_lora:   1%|5                                       | 429/30000 [1:00:18<71:31:07,  8.71s/it, lr: 1.0e-04 loss: 7.000e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 429/30000 [1:00:18<71:31:07,  8.71s/it, lr: 1.0e-04 loss: 7.000e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 430/30000 [1:00:28<71:30:58,  8.71s/it, lr: 1.0e-04 loss: 6.704e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 430/30000 [1:00:28<71:30:58,  8.71s/it, lr: 1.0e-04 loss: 6.704e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 431/30000 [1:00:28<56:44:22,  6.91s/it, lr: 1.0e-04 loss: 6.704e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 431/30000 [1:00:28<56:44:22,  6.91s/it, lr: 1.0e-04 loss: 6.704e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 431/30000 [1:00:34<56:44:22,  6.91s/it, lr: 1.0e-04 loss: 6.820e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 431/30000 [1:00:34<56:44:22,  6.91s/it, lr: 1.0e-04 loss: 6.820e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 432/30000 [1:00:34<56:09:59,  6.84s/it, lr: 1.0e-04 loss: 6.820e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 432/30000 [1:00:34<56:09:59,  6.84s/it, lr: 1.0e-04 loss: 6.820e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 432/30000 [1:00:41<56:09:59,  6.84s/it, lr: 1.0e-04 loss: 6.050e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 432/30000 [1:00:41<56:09:59,  6.84s/it, lr: 1.0e-04 loss: 6.050e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 433/30000 [1:00:41<56:07:37,  6.83s/it, lr: 1.0e-04 loss: 6.050e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 433/30000 [1:00:41<56:07:37,  6.83s/it, lr: 1.0e-04 loss: 6.050e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 433/30000 [1:00:51<56:07:37,  6.83s/it, lr: 1.0e-04 loss: 6.912e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 433/30000 [1:00:51<56:07:37,  6.83s/it, lr: 1.0e-04 loss: 6.912e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 434/30000 [1:00:51<61:45:22,  7.52s/it, lr: 1.0e-04 loss: 6.912e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 434/30000 [1:00:51<61:45:22,  7.52s/it, lr: 1.0e-04 loss: 6.912e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 434/30000 [1:01:00<61:45:22,  7.52s/it, lr: 1.0e-04 loss: 4.787e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 434/30000 [1:01:00<61:45:22,  7.52s/it, lr: 1.0e-04 loss: 4.787e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 435/30000 [1:01:00<66:06:00,  8.05s/it, lr: 1.0e-04 loss: 4.787e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 435/30000 [1:01:00<66:06:00,  8.05s/it, lr: 1.0e-04 loss: 4.787e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 435/30000 [1:01:09<66:06:00,  8.05s/it, lr: 1.0e-04 loss: 9.634e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 435/30000 [1:01:09<66:06:00,  8.05s/it, lr: 1.0e-04 loss: 9.634e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 436/30000 [1:01:09<69:08:44,  8.42s/it, lr: 1.0e-04 loss: 9.634e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 436/30000 [1:01:09<69:08:44,  8.42s/it, lr: 1.0e-04 loss: 9.634e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 436/30000 [1:01:19<69:08:44,  8.42s/it, lr: 1.0e-04 loss: 4.910e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 436/30000 [1:01:19<69:08:44,  8.42s/it, lr: 1.0e-04 loss: 4.910e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 437/30000 [1:01:19<71:24:48,  8.70s/it, lr: 1.0e-04 loss: 4.910e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 437/30000 [1:01:19<71:24:48,  8.70s/it, lr: 1.0e-04 loss: 4.910e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 437/30000 [1:01:25<71:24:48,  8.70s/it, lr: 1.0e-04 loss: 9.424e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 437/30000 [1:01:25<71:24:48,  8.70s/it, lr: 1.0e-04 loss: 9.424e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 438/30000 [1:01:25<66:18:14,  8.07s/it, lr: 1.0e-04 loss: 9.424e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 438/30000 [1:01:25<66:18:14,  8.07s/it, lr: 1.0e-04 loss: 9.424e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 438/30000 [1:01:32<66:18:14,  8.07s/it, lr: 1.0e-04 loss: 7.429e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 438/30000 [1:01:32<66:18:14,  8.07s/it, lr: 1.0e-04 loss: 7.429e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 439/30000 [1:01:32<62:45:48,  7.64s/it, lr: 1.0e-04 loss: 7.429e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 439/30000 [1:01:32<62:45:48,  7.64s/it, lr: 1.0e-04 loss: 7.429e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 439/30000 [1:01:39<62:45:48,  7.64s/it, lr: 1.0e-04 loss: 8.232e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 439/30000 [1:01:39<62:45:48,  7.64s/it, lr: 1.0e-04 loss: 8.232e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 440/30000 [1:01:45<62:45:41,  7.64s/it, lr: 1.0e-04 loss: 6.342e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 440/30000 [1:01:45<62:45:41,  7.64s/it, lr: 1.0e-04 loss: 6.342e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 441/30000 [1:01:45<46:31:11,  5.67s/it, lr: 1.0e-04 loss: 6.342e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 441/30000 [1:01:45<46:31:11,  5.67s/it, lr: 1.0e-04 loss: 6.342e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 441/30000 [1:01:52<46:31:11,  5.67s/it, lr: 1.0e-04 loss: 1.182e-01]
Xiang_wan_2_2_lora:   1%|5                                       | 441/30000 [1:01:52<46:31:11,  5.67s/it, lr: 1.0e-04 loss: 1.182e-01]
Xiang_wan_2_2_lora:   1%|5                                       | 442/30000 [1:01:52<48:23:30,  5.89s/it, lr: 1.0e-04 loss: 1.182e-01]
Xiang_wan_2_2_lora:   1%|5                                       | 442/30000 [1:01:52<48:23:30,  5.89s/it, lr: 1.0e-04 loss: 1.182e-01]
Xiang_wan_2_2_lora:   1%|5                                       | 442/30000 [1:02:01<48:23:30,  5.89s/it, lr: 1.0e-04 loss: 9.756e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 442/30000 [1:02:01<48:23:30,  5.89s/it, lr: 1.0e-04 loss: 9.756e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 443/30000 [1:02:01<55:46:54,  6.79s/it, lr: 1.0e-04 loss: 9.756e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 443/30000 [1:02:01<55:46:54,  6.79s/it, lr: 1.0e-04 loss: 9.756e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 443/30000 [1:02:08<55:46:54,  6.79s/it, lr: 1.0e-04 loss: 1.014e-01]
Xiang_wan_2_2_lora:   1%|5                                       | 443/30000 [1:02:08<55:46:54,  6.79s/it, lr: 1.0e-04 loss: 1.014e-01]
Xiang_wan_2_2_lora:   1%|5                                       | 444/30000 [1:02:08<55:25:16,  6.75s/it, lr: 1.0e-04 loss: 1.014e-01]
Xiang_wan_2_2_lora:   1%|5                                       | 444/30000 [1:02:08<55:25:16,  6.75s/it, lr: 1.0e-04 loss: 1.014e-01]
Xiang_wan_2_2_lora:   1%|5                                       | 444/30000 [1:02:17<55:25:16,  6.75s/it, lr: 1.0e-04 loss: 2.628e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 444/30000 [1:02:17<55:25:16,  6.75s/it, lr: 1.0e-04 loss: 2.628e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 445/30000 [1:02:17<61:26:34,  7.48s/it, lr: 1.0e-04 loss: 2.628e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 445/30000 [1:02:17<61:26:34,  7.48s/it, lr: 1.0e-04 loss: 2.628e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 445/30000 [1:02:24<61:26:34,  7.48s/it, lr: 1.0e-04 loss: 1.096e-01]
Xiang_wan_2_2_lora:   1%|5                                       | 445/30000 [1:02:24<61:26:34,  7.48s/it, lr: 1.0e-04 loss: 1.096e-01]
Xiang_wan_2_2_lora:   1%|5                                       | 446/30000 [1:02:24<59:17:56,  7.22s/it, lr: 1.0e-04 loss: 1.096e-01]
Xiang_wan_2_2_lora:   1%|5                                       | 446/30000 [1:02:24<59:17:56,  7.22s/it, lr: 1.0e-04 loss: 1.096e-01]
Xiang_wan_2_2_lora:   1%|5                                       | 446/30000 [1:02:31<59:17:56,  7.22s/it, lr: 1.0e-04 loss: 3.743e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 446/30000 [1:02:31<59:17:56,  7.22s/it, lr: 1.0e-04 loss: 3.743e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 447/30000 [1:02:31<58:25:41,  7.12s/it, lr: 1.0e-04 loss: 3.743e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 447/30000 [1:02:31<58:25:41,  7.12s/it, lr: 1.0e-04 loss: 3.743e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 447/30000 [1:02:40<58:25:41,  7.12s/it, lr: 1.0e-04 loss: 6.390e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 447/30000 [1:02:40<58:25:41,  7.12s/it, lr: 1.0e-04 loss: 6.390e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 448/30000 [1:02:40<63:42:16,  7.76s/it, lr: 1.0e-04 loss: 6.390e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 448/30000 [1:02:40<63:42:16,  7.76s/it, lr: 1.0e-04 loss: 6.390e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 448/30000 [1:02:47<63:42:16,  7.76s/it, lr: 1.0e-04 loss: 5.067e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 448/30000 [1:02:47<63:42:16,  7.76s/it, lr: 1.0e-04 loss: 5.067e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 449/30000 [1:02:47<61:02:27,  7.44s/it, lr: 1.0e-04 loss: 5.067e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 449/30000 [1:02:47<61:02:27,  7.44s/it, lr: 1.0e-04 loss: 5.067e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 449/30000 [1:02:53<61:02:27,  7.44s/it, lr: 1.0e-04 loss: 9.163e-02]
Xiang_wan_2_2_lora:   1%|5                                       | 449/30000 [1:02:53<61:02:27,  7.44s/it, lr: 1.0e-04 loss: 9.163e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 450/30000 [1:03:03<61:02:19,  7.44s/it, lr: 1.0e-04 loss: 5.171e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 450/30000 [1:03:03<61:02:19,  7.44s/it, lr: 1.0e-04 loss: 5.171e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 451/30000 [1:03:03<50:38:03,  6.17s/it, lr: 1.0e-04 loss: 5.171e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 451/30000 [1:03:03<50:38:03,  6.17s/it, lr: 1.0e-04 loss: 5.171e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 451/30000 [1:03:09<50:38:03,  6.17s/it, lr: 1.0e-04 loss: 7.510e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 451/30000 [1:03:09<50:38:03,  6.17s/it, lr: 1.0e-04 loss: 7.510e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 452/30000 [1:03:09<51:28:14,  6.27s/it, lr: 1.0e-04 loss: 7.510e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 452/30000 [1:03:09<51:28:14,  6.27s/it, lr: 1.0e-04 loss: 7.510e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 452/30000 [1:03:16<51:28:14,  6.27s/it, lr: 1.0e-04 loss: 1.835e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 452/30000 [1:03:16<51:28:14,  6.27s/it, lr: 1.0e-04 loss: 1.835e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 453/30000 [1:03:16<52:11:35,  6.36s/it, lr: 1.0e-04 loss: 1.835e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 453/30000 [1:03:16<52:11:35,  6.36s/it, lr: 1.0e-04 loss: 1.835e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 453/30000 [1:03:25<52:11:35,  6.36s/it, lr: 1.0e-04 loss: 6.779e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 453/30000 [1:03:25<52:11:35,  6.36s/it, lr: 1.0e-04 loss: 6.779e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 454/30000 [1:03:25<59:24:09,  7.24s/it, lr: 1.0e-04 loss: 6.779e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 454/30000 [1:03:25<59:24:09,  7.24s/it, lr: 1.0e-04 loss: 6.779e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 454/30000 [1:03:32<59:24:09,  7.24s/it, lr: 1.0e-04 loss: 7.261e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 454/30000 [1:03:32<59:24:09,  7.24s/it, lr: 1.0e-04 loss: 7.261e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 455/30000 [1:03:32<58:08:30,  7.08s/it, lr: 1.0e-04 loss: 7.261e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 455/30000 [1:03:32<58:08:30,  7.08s/it, lr: 1.0e-04 loss: 7.261e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 455/30000 [1:03:39<58:08:30,  7.08s/it, lr: 1.0e-04 loss: 1.123e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 455/30000 [1:03:39<58:08:30,  7.08s/it, lr: 1.0e-04 loss: 1.123e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 456/30000 [1:03:39<56:52:29,  6.93s/it, lr: 1.0e-04 loss: 1.123e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 456/30000 [1:03:39<56:52:29,  6.93s/it, lr: 1.0e-04 loss: 1.123e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 456/30000 [1:03:45<56:52:29,  6.93s/it, lr: 1.0e-04 loss: 7.509e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 456/30000 [1:03:45<56:52:29,  6.93s/it, lr: 1.0e-04 loss: 7.509e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 457/30000 [1:03:45<56:43:00,  6.91s/it, lr: 1.0e-04 loss: 7.509e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 457/30000 [1:03:45<56:43:00,  6.91s/it, lr: 1.0e-04 loss: 7.509e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 457/30000 [1:03:52<56:43:00,  6.91s/it, lr: 1.0e-04 loss: 1.193e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 457/30000 [1:03:52<56:43:00,  6.91s/it, lr: 1.0e-04 loss: 1.193e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 458/30000 [1:03:52<55:58:04,  6.82s/it, lr: 1.0e-04 loss: 1.193e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 458/30000 [1:03:52<55:58:04,  6.82s/it, lr: 1.0e-04 loss: 1.193e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 458/30000 [1:04:01<55:58:04,  6.82s/it, lr: 1.0e-04 loss: 5.853e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 458/30000 [1:04:01<55:58:04,  6.82s/it, lr: 1.0e-04 loss: 5.853e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 459/30000 [1:04:01<62:10:37,  7.58s/it, lr: 1.0e-04 loss: 5.853e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 459/30000 [1:04:01<62:10:37,  7.58s/it, lr: 1.0e-04 loss: 5.853e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 459/30000 [1:04:11<62:10:37,  7.58s/it, lr: 1.0e-04 loss: 7.234e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 459/30000 [1:04:11<62:10:37,  7.58s/it, lr: 1.0e-04 loss: 7.234e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 460/30000 [1:04:20<62:10:30,  7.58s/it, lr: 1.0e-04 loss: 6.721e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 460/30000 [1:04:20<62:10:30,  7.58s/it, lr: 1.0e-04 loss: 6.721e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 461/30000 [1:04:20<51:17:14,  6.25s/it, lr: 1.0e-04 loss: 6.721e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 461/30000 [1:04:20<51:17:14,  6.25s/it, lr: 1.0e-04 loss: 6.721e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 461/30000 [1:04:27<51:17:14,  6.25s/it, lr: 1.0e-04 loss: 9.273e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 461/30000 [1:04:27<51:17:14,  6.25s/it, lr: 1.0e-04 loss: 9.273e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 462/30000 [1:04:27<52:29:32,  6.40s/it, lr: 1.0e-04 loss: 9.273e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 462/30000 [1:04:27<52:29:32,  6.40s/it, lr: 1.0e-04 loss: 9.273e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 462/30000 [1:04:34<52:29:32,  6.40s/it, lr: 1.0e-04 loss: 8.551e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 462/30000 [1:04:34<52:29:32,  6.40s/it, lr: 1.0e-04 loss: 8.551e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 463/30000 [1:04:34<53:05:57,  6.47s/it, lr: 1.0e-04 loss: 8.551e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 463/30000 [1:04:34<53:05:57,  6.47s/it, lr: 1.0e-04 loss: 8.551e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 463/30000 [1:04:43<53:05:57,  6.47s/it, lr: 1.0e-04 loss: 7.542e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 463/30000 [1:04:43<53:05:57,  6.47s/it, lr: 1.0e-04 loss: 7.542e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 464/30000 [1:04:43<59:16:09,  7.22s/it, lr: 1.0e-04 loss: 7.542e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 464/30000 [1:04:43<59:16:09,  7.22s/it, lr: 1.0e-04 loss: 7.542e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 464/30000 [1:04:50<59:16:09,  7.22s/it, lr: 1.0e-04 loss: 6.269e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 464/30000 [1:04:50<59:16:09,  7.22s/it, lr: 1.0e-04 loss: 6.269e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 465/30000 [1:04:50<57:51:47,  7.05s/it, lr: 1.0e-04 loss: 6.269e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 465/30000 [1:04:50<57:51:47,  7.05s/it, lr: 1.0e-04 loss: 6.269e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 465/30000 [1:04:59<57:51:47,  7.05s/it, lr: 1.0e-04 loss: 9.389e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 465/30000 [1:04:59<57:51:47,  7.05s/it, lr: 1.0e-04 loss: 9.389e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 466/30000 [1:04:59<63:06:38,  7.69s/it, lr: 1.0e-04 loss: 9.389e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 466/30000 [1:04:59<63:06:38,  7.69s/it, lr: 1.0e-04 loss: 9.389e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 466/30000 [1:05:05<63:06:38,  7.69s/it, lr: 1.0e-04 loss: 6.966e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 466/30000 [1:05:05<63:06:38,  7.69s/it, lr: 1.0e-04 loss: 6.966e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 467/30000 [1:05:05<60:35:26,  7.39s/it, lr: 1.0e-04 loss: 6.966e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 467/30000 [1:05:05<60:35:26,  7.39s/it, lr: 1.0e-04 loss: 6.966e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 467/30000 [1:05:12<60:35:26,  7.39s/it, lr: 1.0e-04 loss: 9.293e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 467/30000 [1:05:12<60:35:26,  7.39s/it, lr: 1.0e-04 loss: 9.293e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 468/30000 [1:05:12<58:27:08,  7.13s/it, lr: 1.0e-04 loss: 9.293e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 468/30000 [1:05:12<58:27:08,  7.13s/it, lr: 1.0e-04 loss: 9.293e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 468/30000 [1:05:22<58:27:08,  7.13s/it, lr: 1.0e-04 loss: 4.254e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 468/30000 [1:05:22<58:27:08,  7.13s/it, lr: 1.0e-04 loss: 4.254e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 469/30000 [1:05:22<64:25:25,  7.85s/it, lr: 1.0e-04 loss: 4.254e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 469/30000 [1:05:22<64:25:25,  7.85s/it, lr: 1.0e-04 loss: 4.254e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 469/30000 [1:05:31<64:25:25,  7.85s/it, lr: 1.0e-04 loss: 2.204e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 469/30000 [1:05:31<64:25:25,  7.85s/it, lr: 1.0e-04 loss: 2.204e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 470/30000 [1:05:38<64:25:17,  7.85s/it, lr: 1.0e-04 loss: 3.472e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 470/30000 [1:05:38<64:25:17,  7.85s/it, lr: 1.0e-04 loss: 3.472e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 471/30000 [1:05:38<47:25:23,  5.78s/it, lr: 1.0e-04 loss: 3.472e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 471/30000 [1:05:38<47:25:23,  5.78s/it, lr: 1.0e-04 loss: 3.472e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 471/30000 [1:05:44<47:25:23,  5.78s/it, lr: 1.0e-04 loss: 1.107e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 471/30000 [1:05:44<47:25:23,  5.78s/it, lr: 1.0e-04 loss: 1.107e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 472/30000 [1:05:44<48:56:49,  5.97s/it, lr: 1.0e-04 loss: 1.107e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 472/30000 [1:05:44<48:56:49,  5.97s/it, lr: 1.0e-04 loss: 1.107e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 472/30000 [1:05:54<48:56:49,  5.97s/it, lr: 1.0e-04 loss: 7.416e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 472/30000 [1:05:54<48:56:49,  5.97s/it, lr: 1.0e-04 loss: 7.416e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 473/30000 [1:05:54<56:13:59,  6.86s/it, lr: 1.0e-04 loss: 7.416e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 473/30000 [1:05:54<56:13:59,  6.86s/it, lr: 1.0e-04 loss: 7.416e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 473/30000 [1:06:03<56:13:59,  6.86s/it, lr: 1.0e-04 loss: 7.437e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 473/30000 [1:06:03<56:13:59,  6.86s/it, lr: 1.0e-04 loss: 7.437e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 474/30000 [1:06:03<62:02:49,  7.57s/it, lr: 1.0e-04 loss: 7.437e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 474/30000 [1:06:03<62:02:49,  7.57s/it, lr: 1.0e-04 loss: 7.437e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 474/30000 [1:06:10<62:02:49,  7.57s/it, lr: 1.0e-04 loss: 7.875e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 474/30000 [1:06:10<62:02:49,  7.57s/it, lr: 1.0e-04 loss: 7.875e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 475/30000 [1:06:10<59:54:11,  7.30s/it, lr: 1.0e-04 loss: 7.875e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 475/30000 [1:06:10<59:54:11,  7.30s/it, lr: 1.0e-04 loss: 7.875e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 475/30000 [1:06:17<59:54:11,  7.30s/it, lr: 1.0e-04 loss: 8.385e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 475/30000 [1:06:17<59:54:11,  7.30s/it, lr: 1.0e-04 loss: 8.385e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 476/30000 [1:06:17<58:54:52,  7.18s/it, lr: 1.0e-04 loss: 8.385e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 476/30000 [1:06:17<58:54:52,  7.18s/it, lr: 1.0e-04 loss: 8.385e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 476/30000 [1:06:26<58:54:52,  7.18s/it, lr: 1.0e-04 loss: 1.813e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 476/30000 [1:06:26<58:54:52,  7.18s/it, lr: 1.0e-04 loss: 1.813e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 477/30000 [1:06:26<64:05:15,  7.81s/it, lr: 1.0e-04 loss: 1.813e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 477/30000 [1:06:26<64:05:15,  7.81s/it, lr: 1.0e-04 loss: 1.813e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 477/30000 [1:06:33<64:05:15,  7.81s/it, lr: 1.0e-04 loss: 8.130e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 477/30000 [1:06:33<64:05:15,  7.81s/it, lr: 1.0e-04 loss: 8.130e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 478/30000 [1:06:33<61:08:56,  7.46s/it, lr: 1.0e-04 loss: 8.130e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 478/30000 [1:06:33<61:08:56,  7.46s/it, lr: 1.0e-04 loss: 8.130e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 478/30000 [1:06:42<61:08:56,  7.46s/it, lr: 1.0e-04 loss: 4.308e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 478/30000 [1:06:42<61:08:56,  7.46s/it, lr: 1.0e-04 loss: 4.308e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 479/30000 [1:06:42<65:42:02,  8.01s/it, lr: 1.0e-04 loss: 4.308e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 479/30000 [1:06:42<65:42:02,  8.01s/it, lr: 1.0e-04 loss: 4.308e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 479/30000 [1:06:51<65:42:02,  8.01s/it, lr: 1.0e-04 loss: 8.880e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 479/30000 [1:06:51<65:42:02,  8.01s/it, lr: 1.0e-04 loss: 8.880e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 480/30000 [1:06:58<65:41:54,  8.01s/it, lr: 1.0e-04 loss: 2.242e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 480/30000 [1:06:58<65:41:54,  8.01s/it, lr: 1.0e-04 loss: 2.242e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 481/30000 [1:06:58<47:58:07,  5.85s/it, lr: 1.0e-04 loss: 2.242e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 481/30000 [1:06:58<47:58:07,  5.85s/it, lr: 1.0e-04 loss: 2.242e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 481/30000 [1:07:07<47:58:07,  5.85s/it, lr: 1.0e-04 loss: 7.005e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 481/30000 [1:07:07<47:58:07,  5.85s/it, lr: 1.0e-04 loss: 7.005e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 482/30000 [1:07:07<55:17:29,  6.74s/it, lr: 1.0e-04 loss: 7.005e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 482/30000 [1:07:07<55:17:29,  6.74s/it, lr: 1.0e-04 loss: 7.005e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 482/30000 [1:07:17<55:17:29,  6.74s/it, lr: 1.0e-04 loss: 5.704e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 482/30000 [1:07:17<55:17:29,  6.74s/it, lr: 1.0e-04 loss: 5.704e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 483/30000 [1:07:17<61:22:12,  7.48s/it, lr: 1.0e-04 loss: 5.704e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 483/30000 [1:07:17<61:22:12,  7.48s/it, lr: 1.0e-04 loss: 5.704e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 483/30000 [1:07:24<61:22:12,  7.48s/it, lr: 1.0e-04 loss: 7.575e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 483/30000 [1:07:24<61:22:12,  7.48s/it, lr: 1.0e-04 loss: 7.575e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 484/30000 [1:07:24<59:59:20,  7.32s/it, lr: 1.0e-04 loss: 7.575e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 484/30000 [1:07:24<59:59:20,  7.32s/it, lr: 1.0e-04 loss: 7.575e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 484/30000 [1:07:33<59:59:20,  7.32s/it, lr: 1.0e-04 loss: 1.132e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 484/30000 [1:07:33<59:59:20,  7.32s/it, lr: 1.0e-04 loss: 1.132e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 485/30000 [1:07:33<64:34:36,  7.88s/it, lr: 1.0e-04 loss: 1.132e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 485/30000 [1:07:33<64:34:36,  7.88s/it, lr: 1.0e-04 loss: 1.132e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 485/30000 [1:07:40<64:34:36,  7.88s/it, lr: 1.0e-04 loss: 9.318e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 485/30000 [1:07:40<64:34:36,  7.88s/it, lr: 1.0e-04 loss: 9.318e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 486/30000 [1:07:40<61:27:46,  7.50s/it, lr: 1.0e-04 loss: 9.318e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 486/30000 [1:07:40<61:27:46,  7.50s/it, lr: 1.0e-04 loss: 9.318e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 486/30000 [1:07:46<61:27:46,  7.50s/it, lr: 1.0e-04 loss: 2.869e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 486/30000 [1:07:46<61:27:46,  7.50s/it, lr: 1.0e-04 loss: 2.869e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 487/30000 [1:07:46<59:09:53,  7.22s/it, lr: 1.0e-04 loss: 2.869e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 487/30000 [1:07:46<59:09:53,  7.22s/it, lr: 1.0e-04 loss: 2.869e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 487/30000 [1:07:53<59:09:53,  7.22s/it, lr: 1.0e-04 loss: 7.477e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 487/30000 [1:07:53<59:09:53,  7.22s/it, lr: 1.0e-04 loss: 7.477e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 488/30000 [1:07:53<57:42:16,  7.04s/it, lr: 1.0e-04 loss: 7.477e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 488/30000 [1:07:53<57:42:16,  7.04s/it, lr: 1.0e-04 loss: 7.477e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 488/30000 [1:08:02<57:42:16,  7.04s/it, lr: 1.0e-04 loss: 9.879e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 488/30000 [1:08:02<57:42:16,  7.04s/it, lr: 1.0e-04 loss: 9.879e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 489/30000 [1:08:02<63:53:22,  7.79s/it, lr: 1.0e-04 loss: 9.879e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 489/30000 [1:08:02<63:53:22,  7.79s/it, lr: 1.0e-04 loss: 9.879e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 489/30000 [1:08:12<63:53:22,  7.79s/it, lr: 1.0e-04 loss: 9.996e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 489/30000 [1:08:12<63:53:22,  7.79s/it, lr: 1.0e-04 loss: 9.996e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 490/30000 [1:08:19<63:53:14,  7.79s/it, lr: 1.0e-04 loss: 6.605e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 490/30000 [1:08:19<63:53:14,  7.79s/it, lr: 1.0e-04 loss: 6.605e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 491/30000 [1:08:19<47:51:33,  5.84s/it, lr: 1.0e-04 loss: 6.605e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 491/30000 [1:08:19<47:51:33,  5.84s/it, lr: 1.0e-04 loss: 6.605e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 491/30000 [1:08:28<47:51:33,  5.84s/it, lr: 1.0e-04 loss: 7.242e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 491/30000 [1:08:28<47:51:33,  5.84s/it, lr: 1.0e-04 loss: 7.242e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 492/30000 [1:08:28<54:46:20,  6.68s/it, lr: 1.0e-04 loss: 7.242e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 492/30000 [1:08:28<54:46:20,  6.68s/it, lr: 1.0e-04 loss: 7.242e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 492/30000 [1:08:37<54:46:20,  6.68s/it, lr: 1.0e-04 loss: 4.787e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 492/30000 [1:08:37<54:46:20,  6.68s/it, lr: 1.0e-04 loss: 4.787e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 493/30000 [1:08:37<60:29:56,  7.38s/it, lr: 1.0e-04 loss: 4.787e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 493/30000 [1:08:37<60:29:56,  7.38s/it, lr: 1.0e-04 loss: 4.787e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 493/30000 [1:08:44<60:29:56,  7.38s/it, lr: 1.0e-04 loss: 1.035e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 493/30000 [1:08:44<60:29:56,  7.38s/it, lr: 1.0e-04 loss: 1.035e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 494/30000 [1:08:44<58:47:43,  7.17s/it, lr: 1.0e-04 loss: 1.035e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 494/30000 [1:08:44<58:47:43,  7.17s/it, lr: 1.0e-04 loss: 1.035e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 494/30000 [1:08:53<58:47:43,  7.17s/it, lr: 1.0e-04 loss: 2.283e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 494/30000 [1:08:53<58:47:43,  7.17s/it, lr: 1.0e-04 loss: 2.283e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 495/30000 [1:08:53<63:51:21,  7.79s/it, lr: 1.0e-04 loss: 2.283e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 495/30000 [1:08:53<63:51:21,  7.79s/it, lr: 1.0e-04 loss: 2.283e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 495/30000 [1:09:00<63:51:21,  7.79s/it, lr: 1.0e-04 loss: 1.029e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 495/30000 [1:09:00<63:51:21,  7.79s/it, lr: 1.0e-04 loss: 1.029e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 496/30000 [1:09:00<61:06:50,  7.46s/it, lr: 1.0e-04 loss: 1.029e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 496/30000 [1:09:00<61:06:50,  7.46s/it, lr: 1.0e-04 loss: 1.029e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 496/30000 [1:09:09<61:06:50,  7.46s/it, lr: 1.0e-04 loss: 5.033e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 496/30000 [1:09:09<61:06:50,  7.46s/it, lr: 1.0e-04 loss: 5.033e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 497/30000 [1:09:09<65:44:54,  8.02s/it, lr: 1.0e-04 loss: 5.033e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 497/30000 [1:09:09<65:44:54,  8.02s/it, lr: 1.0e-04 loss: 5.033e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 497/30000 [1:09:19<65:44:54,  8.02s/it, lr: 1.0e-04 loss: 5.515e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 497/30000 [1:09:19<65:44:54,  8.02s/it, lr: 1.0e-04 loss: 5.515e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 498/30000 [1:09:19<68:52:34,  8.40s/it, lr: 1.0e-04 loss: 5.515e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 498/30000 [1:09:19<68:52:34,  8.40s/it, lr: 1.0e-04 loss: 5.515e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 498/30000 [1:09:28<68:52:34,  8.40s/it, lr: 1.0e-04 loss: 5.507e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 498/30000 [1:09:28<68:52:34,  8.40s/it, lr: 1.0e-04 loss: 5.507e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 499/30000 [1:09:28<71:53:04,  8.77s/it, lr: 1.0e-04 loss: 5.507e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 499/30000 [1:09:28<71:53:04,  8.77s/it, lr: 1.0e-04 loss: 5.507e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 499/30000 [1:09:38<71:53:04,  8.77s/it, lr: 1.0e-04 loss: 9.336e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 499/30000 [1:09:38<71:53:04,  8.77s/it, lr: 1.0e-04 loss: 9.336e-02]
+Saving at step 500
+Saved checkpoint to /home/featurize/ai-toolkit/output/Xiang_wan_2_2_lora/Xiang_wan_2_2_lora_000000500.safetensors
+Saved optimizer to /home/featurize/ai-toolkit/output/Xiang_wan_2_2_lora/optimizer.pt
+
Xiang_wan_2_2_lora:   2%|6                                       | 500/30000 [1:09:45<71:52:55,  8.77s/it, lr: 1.0e-04 loss: 5.786e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 500/30000 [1:09:45<71:52:55,  8.77s/it, lr: 1.0e-04 loss: 5.786e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 501/30000 [1:09:45<52:07:48,  6.36s/it, lr: 1.0e-04 loss: 5.786e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 501/30000 [1:09:45<52:07:48,  6.36s/it, lr: 1.0e-04 loss: 5.786e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 501/30000 [1:09:52<52:07:48,  6.36s/it, lr: 1.0e-04 loss: 9.630e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 501/30000 [1:09:52<52:07:48,  6.36s/it, lr: 1.0e-04 loss: 9.630e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 502/30000 [1:09:52<52:32:34,  6.41s/it, lr: 1.0e-04 loss: 9.630e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 502/30000 [1:09:52<52:32:34,  6.41s/it, lr: 1.0e-04 loss: 9.630e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 502/30000 [1:10:01<52:32:34,  6.41s/it, lr: 1.0e-04 loss: 3.462e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 502/30000 [1:10:01<52:32:34,  6.41s/it, lr: 1.0e-04 loss: 3.462e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 503/30000 [1:10:01<58:53:45,  7.19s/it, lr: 1.0e-04 loss: 3.462e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 503/30000 [1:10:01<58:53:45,  7.19s/it, lr: 1.0e-04 loss: 3.462e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 503/30000 [1:10:08<58:53:45,  7.19s/it, lr: 1.0e-04 loss: 7.999e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 503/30000 [1:10:08<58:53:45,  7.19s/it, lr: 1.0e-04 loss: 7.999e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 504/30000 [1:10:08<57:46:28,  7.05s/it, lr: 1.0e-04 loss: 7.999e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 504/30000 [1:10:08<57:46:28,  7.05s/it, lr: 1.0e-04 loss: 7.999e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 504/30000 [1:10:15<57:46:28,  7.05s/it, lr: 1.0e-04 loss: 1.808e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 504/30000 [1:10:15<57:46:28,  7.05s/it, lr: 1.0e-04 loss: 1.808e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 505/30000 [1:10:15<56:56:40,  6.95s/it, lr: 1.0e-04 loss: 1.808e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 505/30000 [1:10:15<56:56:40,  6.95s/it, lr: 1.0e-04 loss: 1.808e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 505/30000 [1:10:24<56:56:40,  6.95s/it, lr: 1.0e-04 loss: 6.606e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 505/30000 [1:10:24<56:56:40,  6.95s/it, lr: 1.0e-04 loss: 6.606e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 506/30000 [1:10:24<62:50:50,  7.67s/it, lr: 1.0e-04 loss: 6.606e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 506/30000 [1:10:24<62:50:50,  7.67s/it, lr: 1.0e-04 loss: 6.606e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 506/30000 [1:10:31<62:50:50,  7.67s/it, lr: 1.0e-04 loss: 7.087e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 506/30000 [1:10:31<62:50:50,  7.67s/it, lr: 1.0e-04 loss: 7.087e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 507/30000 [1:10:31<60:39:53,  7.40s/it, lr: 1.0e-04 loss: 7.087e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 507/30000 [1:10:31<60:39:53,  7.40s/it, lr: 1.0e-04 loss: 7.087e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 507/30000 [1:10:40<60:39:53,  7.40s/it, lr: 1.0e-04 loss: 8.178e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 507/30000 [1:10:40<60:39:53,  7.40s/it, lr: 1.0e-04 loss: 8.178e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 508/30000 [1:10:40<65:40:17,  8.02s/it, lr: 1.0e-04 loss: 8.178e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 508/30000 [1:10:40<65:40:17,  8.02s/it, lr: 1.0e-04 loss: 8.178e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 508/30000 [1:10:50<65:40:17,  8.02s/it, lr: 1.0e-04 loss: 8.358e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 508/30000 [1:10:50<65:40:17,  8.02s/it, lr: 1.0e-04 loss: 8.358e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 509/30000 [1:10:50<69:04:13,  8.43s/it, lr: 1.0e-04 loss: 8.358e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 509/30000 [1:10:50<69:04:13,  8.43s/it, lr: 1.0e-04 loss: 8.358e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 509/30000 [1:10:56<69:04:13,  8.43s/it, lr: 1.0e-04 loss: 1.261e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 509/30000 [1:10:56<69:04:13,  8.43s/it, lr: 1.0e-04 loss: 1.261e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 510/30000 [1:11:06<69:04:04,  8.43s/it, lr: 1.0e-04 loss: 2.054e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 510/30000 [1:11:06<69:04:04,  8.43s/it, lr: 1.0e-04 loss: 2.054e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 511/30000 [1:11:06<55:08:21,  6.73s/it, lr: 1.0e-04 loss: 2.054e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 511/30000 [1:11:06<55:08:21,  6.73s/it, lr: 1.0e-04 loss: 2.054e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 511/30000 [1:11:12<55:08:21,  6.73s/it, lr: 1.0e-04 loss: 1.541e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 511/30000 [1:11:12<55:08:21,  6.73s/it, lr: 1.0e-04 loss: 1.541e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 512/30000 [1:11:12<54:56:11,  6.71s/it, lr: 1.0e-04 loss: 1.541e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 512/30000 [1:11:12<54:56:11,  6.71s/it, lr: 1.0e-04 loss: 1.541e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 512/30000 [1:11:19<54:56:11,  6.71s/it, lr: 1.0e-04 loss: 4.711e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 512/30000 [1:11:19<54:56:11,  6.71s/it, lr: 1.0e-04 loss: 4.711e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 513/30000 [1:11:19<54:42:36,  6.68s/it, lr: 1.0e-04 loss: 4.711e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 513/30000 [1:11:19<54:42:36,  6.68s/it, lr: 1.0e-04 loss: 4.711e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 513/30000 [1:11:26<54:42:36,  6.68s/it, lr: 1.0e-04 loss: 8.927e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 513/30000 [1:11:26<54:42:36,  6.68s/it, lr: 1.0e-04 loss: 8.927e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 514/30000 [1:11:26<54:35:13,  6.66s/it, lr: 1.0e-04 loss: 8.927e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 514/30000 [1:11:26<54:35:13,  6.66s/it, lr: 1.0e-04 loss: 8.927e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 514/30000 [1:11:35<54:35:13,  6.66s/it, lr: 1.0e-04 loss: 4.320e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 514/30000 [1:11:35<54:35:13,  6.66s/it, lr: 1.0e-04 loss: 4.320e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 515/30000 [1:11:35<61:10:58,  7.47s/it, lr: 1.0e-04 loss: 4.320e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 515/30000 [1:11:35<61:10:58,  7.47s/it, lr: 1.0e-04 loss: 4.320e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 515/30000 [1:11:42<61:10:58,  7.47s/it, lr: 1.0e-04 loss: 9.089e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 515/30000 [1:11:42<61:10:58,  7.47s/it, lr: 1.0e-04 loss: 9.089e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 516/30000 [1:11:42<59:17:31,  7.24s/it, lr: 1.0e-04 loss: 9.089e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 516/30000 [1:11:42<59:17:31,  7.24s/it, lr: 1.0e-04 loss: 9.089e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 516/30000 [1:11:48<59:17:31,  7.24s/it, lr: 1.0e-04 loss: 4.537e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 516/30000 [1:11:48<59:17:31,  7.24s/it, lr: 1.0e-04 loss: 4.537e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 517/30000 [1:11:48<57:48:56,  7.06s/it, lr: 1.0e-04 loss: 4.537e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 517/30000 [1:11:48<57:48:56,  7.06s/it, lr: 1.0e-04 loss: 4.537e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 517/30000 [1:11:58<57:48:56,  7.06s/it, lr: 1.0e-04 loss: 1.644e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 517/30000 [1:11:58<57:48:56,  7.06s/it, lr: 1.0e-04 loss: 1.644e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 518/30000 [1:11:58<63:15:08,  7.72s/it, lr: 1.0e-04 loss: 1.644e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 518/30000 [1:11:58<63:15:08,  7.72s/it, lr: 1.0e-04 loss: 1.644e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 518/30000 [1:12:04<63:15:08,  7.72s/it, lr: 1.0e-04 loss: 1.249e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 518/30000 [1:12:04<63:15:08,  7.72s/it, lr: 1.0e-04 loss: 1.249e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 519/30000 [1:12:04<60:43:29,  7.42s/it, lr: 1.0e-04 loss: 1.249e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 519/30000 [1:12:04<60:43:29,  7.42s/it, lr: 1.0e-04 loss: 1.249e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 519/30000 [1:12:14<60:43:29,  7.42s/it, lr: 1.0e-04 loss: 8.968e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 519/30000 [1:12:14<60:43:29,  7.42s/it, lr: 1.0e-04 loss: 8.968e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 520/30000 [1:12:23<60:43:21,  7.42s/it, lr: 1.0e-04 loss: 4.587e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 520/30000 [1:12:23<60:43:21,  7.42s/it, lr: 1.0e-04 loss: 4.587e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 521/30000 [1:12:23<50:38:27,  6.18s/it, lr: 1.0e-04 loss: 4.587e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 521/30000 [1:12:23<50:38:27,  6.18s/it, lr: 1.0e-04 loss: 4.587e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 521/30000 [1:12:33<50:38:27,  6.18s/it, lr: 1.0e-04 loss: 8.268e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 521/30000 [1:12:33<50:38:27,  6.18s/it, lr: 1.0e-04 loss: 8.268e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 522/30000 [1:12:33<57:31:37,  7.03s/it, lr: 1.0e-04 loss: 8.268e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 522/30000 [1:12:33<57:31:37,  7.03s/it, lr: 1.0e-04 loss: 8.268e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 522/30000 [1:12:40<57:31:37,  7.03s/it, lr: 1.0e-04 loss: 1.516e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 522/30000 [1:12:40<57:31:37,  7.03s/it, lr: 1.0e-04 loss: 1.516e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 523/30000 [1:12:40<57:15:34,  6.99s/it, lr: 1.0e-04 loss: 1.516e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 523/30000 [1:12:40<57:15:34,  6.99s/it, lr: 1.0e-04 loss: 1.516e-01]
Xiang_wan_2_2_lora:   2%|6                                       | 523/30000 [1:12:49<57:15:34,  6.99s/it, lr: 1.0e-04 loss: 7.445e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 523/30000 [1:12:49<57:15:34,  6.99s/it, lr: 1.0e-04 loss: 7.445e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 524/30000 [1:12:49<63:03:47,  7.70s/it, lr: 1.0e-04 loss: 7.445e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 524/30000 [1:12:49<63:03:47,  7.70s/it, lr: 1.0e-04 loss: 7.445e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 524/30000 [1:12:59<63:03:47,  7.70s/it, lr: 1.0e-04 loss: 4.867e-02]
Xiang_wan_2_2_lora:   2%|6                                       | 524/30000 [1:12:59<63:03:47,  7.70s/it, lr: 1.0e-04 loss: 4.867e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 525/30000 [1:12:59<66:56:20,  8.18s/it, lr: 1.0e-04 loss: 4.867e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 525/30000 [1:12:59<66:56:20,  8.18s/it, lr: 1.0e-04 loss: 4.867e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 525/30000 [1:13:05<66:56:20,  8.18s/it, lr: 1.0e-04 loss: 1.032e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 525/30000 [1:13:05<66:56:20,  8.18s/it, lr: 1.0e-04 loss: 1.032e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 526/30000 [1:13:05<63:20:20,  7.74s/it, lr: 1.0e-04 loss: 1.032e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 526/30000 [1:13:05<63:20:20,  7.74s/it, lr: 1.0e-04 loss: 1.032e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 526/30000 [1:13:12<63:20:20,  7.74s/it, lr: 1.0e-04 loss: 6.283e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 526/30000 [1:13:12<63:20:20,  7.74s/it, lr: 1.0e-04 loss: 6.283e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 527/30000 [1:13:12<61:07:07,  7.47s/it, lr: 1.0e-04 loss: 6.283e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 527/30000 [1:13:12<61:07:07,  7.47s/it, lr: 1.0e-04 loss: 6.283e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 527/30000 [1:13:19<61:07:07,  7.47s/it, lr: 1.0e-04 loss: 1.094e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 527/30000 [1:13:19<61:07:07,  7.47s/it, lr: 1.0e-04 loss: 1.094e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 528/30000 [1:13:19<59:41:03,  7.29s/it, lr: 1.0e-04 loss: 1.094e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 528/30000 [1:13:19<59:41:03,  7.29s/it, lr: 1.0e-04 loss: 1.094e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 528/30000 [1:13:29<59:41:03,  7.29s/it, lr: 1.0e-04 loss: 5.172e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 528/30000 [1:13:29<59:41:03,  7.29s/it, lr: 1.0e-04 loss: 5.172e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 529/30000 [1:13:29<64:47:09,  7.91s/it, lr: 1.0e-04 loss: 5.172e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 529/30000 [1:13:29<64:47:09,  7.91s/it, lr: 1.0e-04 loss: 5.172e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 529/30000 [1:13:35<64:47:09,  7.91s/it, lr: 1.0e-04 loss: 8.917e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 529/30000 [1:13:35<64:47:09,  7.91s/it, lr: 1.0e-04 loss: 8.917e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 530/30000 [1:13:42<64:47:01,  7.91s/it, lr: 1.0e-04 loss: 2.050e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 530/30000 [1:13:42<64:47:01,  7.91s/it, lr: 1.0e-04 loss: 2.050e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 531/30000 [1:13:42<47:34:34,  5.81s/it, lr: 1.0e-04 loss: 2.050e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 531/30000 [1:13:42<47:34:34,  5.81s/it, lr: 1.0e-04 loss: 2.050e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 531/30000 [1:13:51<47:34:34,  5.81s/it, lr: 1.0e-04 loss: 1.653e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 531/30000 [1:13:51<47:34:34,  5.81s/it, lr: 1.0e-04 loss: 1.653e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 532/30000 [1:13:51<54:43:12,  6.68s/it, lr: 1.0e-04 loss: 1.653e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 532/30000 [1:13:51<54:43:12,  6.68s/it, lr: 1.0e-04 loss: 1.653e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 532/30000 [1:14:01<54:43:12,  6.68s/it, lr: 1.0e-04 loss: 5.616e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 532/30000 [1:14:01<54:43:12,  6.68s/it, lr: 1.0e-04 loss: 5.616e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 533/30000 [1:14:01<60:29:04,  7.39s/it, lr: 1.0e-04 loss: 5.616e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 533/30000 [1:14:01<60:29:04,  7.39s/it, lr: 1.0e-04 loss: 5.616e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 533/30000 [1:14:07<60:29:04,  7.39s/it, lr: 1.0e-04 loss: 9.821e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 533/30000 [1:14:07<60:29:04,  7.39s/it, lr: 1.0e-04 loss: 9.821e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 534/30000 [1:14:07<58:42:45,  7.17s/it, lr: 1.0e-04 loss: 9.821e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 534/30000 [1:14:07<58:42:45,  7.17s/it, lr: 1.0e-04 loss: 9.821e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 534/30000 [1:14:17<58:42:45,  7.17s/it, lr: 1.0e-04 loss: 7.801e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 534/30000 [1:14:17<58:42:45,  7.17s/it, lr: 1.0e-04 loss: 7.801e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 535/30000 [1:14:17<63:43:00,  7.78s/it, lr: 1.0e-04 loss: 7.801e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 535/30000 [1:14:17<63:43:00,  7.78s/it, lr: 1.0e-04 loss: 7.801e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 535/30000 [1:14:26<63:43:00,  7.78s/it, lr: 1.0e-04 loss: 8.660e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 535/30000 [1:14:26<63:43:00,  7.78s/it, lr: 1.0e-04 loss: 8.660e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 536/30000 [1:14:26<67:26:08,  8.24s/it, lr: 1.0e-04 loss: 8.660e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 536/30000 [1:14:26<67:26:08,  8.24s/it, lr: 1.0e-04 loss: 8.660e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 536/30000 [1:14:36<67:26:08,  8.24s/it, lr: 1.0e-04 loss: 3.415e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 536/30000 [1:14:36<67:26:08,  8.24s/it, lr: 1.0e-04 loss: 3.415e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 537/30000 [1:14:36<70:38:21,  8.63s/it, lr: 1.0e-04 loss: 3.415e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 537/30000 [1:14:36<70:38:21,  8.63s/it, lr: 1.0e-04 loss: 3.415e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 537/30000 [1:14:45<70:38:21,  8.63s/it, lr: 1.0e-04 loss: 1.191e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 537/30000 [1:14:45<70:38:21,  8.63s/it, lr: 1.0e-04 loss: 1.191e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 538/30000 [1:14:45<72:24:51,  8.85s/it, lr: 1.0e-04 loss: 1.191e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 538/30000 [1:14:45<72:24:51,  8.85s/it, lr: 1.0e-04 loss: 1.191e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 538/30000 [1:14:52<72:24:51,  8.85s/it, lr: 1.0e-04 loss: 1.302e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 538/30000 [1:14:52<72:24:51,  8.85s/it, lr: 1.0e-04 loss: 1.302e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 539/30000 [1:14:52<67:03:08,  8.19s/it, lr: 1.0e-04 loss: 1.302e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 539/30000 [1:14:52<67:03:08,  8.19s/it, lr: 1.0e-04 loss: 1.302e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 539/30000 [1:14:58<67:03:08,  8.19s/it, lr: 1.0e-04 loss: 9.444e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 539/30000 [1:14:58<67:03:08,  8.19s/it, lr: 1.0e-04 loss: 9.444e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 540/30000 [1:15:08<67:03:00,  8.19s/it, lr: 1.0e-04 loss: 4.785e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 540/30000 [1:15:08<67:03:00,  8.19s/it, lr: 1.0e-04 loss: 4.785e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 541/30000 [1:15:08<54:10:18,  6.62s/it, lr: 1.0e-04 loss: 4.785e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 541/30000 [1:15:08<54:10:18,  6.62s/it, lr: 1.0e-04 loss: 4.785e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 541/30000 [1:15:14<54:10:18,  6.62s/it, lr: 1.0e-04 loss: 1.003e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 541/30000 [1:15:14<54:10:18,  6.62s/it, lr: 1.0e-04 loss: 1.003e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 542/30000 [1:15:14<54:09:53,  6.62s/it, lr: 1.0e-04 loss: 1.003e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 542/30000 [1:15:14<54:09:53,  6.62s/it, lr: 1.0e-04 loss: 1.003e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 542/30000 [1:15:21<54:09:53,  6.62s/it, lr: 1.0e-04 loss: 8.282e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 542/30000 [1:15:21<54:09:53,  6.62s/it, lr: 1.0e-04 loss: 8.282e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 543/30000 [1:15:21<54:26:34,  6.65s/it, lr: 1.0e-04 loss: 8.282e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 543/30000 [1:15:21<54:26:34,  6.65s/it, lr: 1.0e-04 loss: 8.282e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 543/30000 [1:15:28<54:26:34,  6.65s/it, lr: 1.0e-04 loss: 1.131e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 543/30000 [1:15:28<54:26:34,  6.65s/it, lr: 1.0e-04 loss: 1.131e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 544/30000 [1:15:28<54:13:57,  6.63s/it, lr: 1.0e-04 loss: 1.131e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 544/30000 [1:15:28<54:13:57,  6.63s/it, lr: 1.0e-04 loss: 1.131e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 544/30000 [1:15:37<54:13:57,  6.63s/it, lr: 1.0e-04 loss: 3.846e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 544/30000 [1:15:37<54:13:57,  6.63s/it, lr: 1.0e-04 loss: 3.846e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 545/30000 [1:15:37<61:12:14,  7.48s/it, lr: 1.0e-04 loss: 3.846e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 545/30000 [1:15:37<61:12:14,  7.48s/it, lr: 1.0e-04 loss: 3.846e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 545/30000 [1:15:47<61:12:14,  7.48s/it, lr: 1.0e-04 loss: 6.122e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 545/30000 [1:15:47<61:12:14,  7.48s/it, lr: 1.0e-04 loss: 6.122e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 546/30000 [1:15:47<65:45:56,  8.04s/it, lr: 1.0e-04 loss: 6.122e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 546/30000 [1:15:47<65:45:56,  8.04s/it, lr: 1.0e-04 loss: 6.122e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 546/30000 [1:15:54<65:45:56,  8.04s/it, lr: 1.0e-04 loss: 4.918e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 546/30000 [1:15:54<65:45:56,  8.04s/it, lr: 1.0e-04 loss: 4.918e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 547/30000 [1:15:54<62:28:42,  7.64s/it, lr: 1.0e-04 loss: 4.918e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 547/30000 [1:15:54<62:28:42,  7.64s/it, lr: 1.0e-04 loss: 4.918e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 547/30000 [1:16:03<62:28:42,  7.64s/it, lr: 1.0e-04 loss: 6.962e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 547/30000 [1:16:03<62:28:42,  7.64s/it, lr: 1.0e-04 loss: 6.962e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 548/30000 [1:16:03<67:03:50,  8.20s/it, lr: 1.0e-04 loss: 6.962e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 548/30000 [1:16:03<67:03:50,  8.20s/it, lr: 1.0e-04 loss: 6.962e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 548/30000 [1:16:12<67:03:50,  8.20s/it, lr: 1.0e-04 loss: 9.991e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 548/30000 [1:16:12<67:03:50,  8.20s/it, lr: 1.0e-04 loss: 9.991e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 549/30000 [1:16:12<69:51:00,  8.54s/it, lr: 1.0e-04 loss: 9.991e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 549/30000 [1:16:12<69:51:00,  8.54s/it, lr: 1.0e-04 loss: 9.991e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 549/30000 [1:16:19<69:51:00,  8.54s/it, lr: 1.0e-04 loss: 8.559e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 549/30000 [1:16:19<69:51:00,  8.54s/it, lr: 1.0e-04 loss: 8.559e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 550/30000 [1:16:28<69:50:51,  8.54s/it, lr: 1.0e-04 loss: 4.366e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 550/30000 [1:16:28<69:50:51,  8.54s/it, lr: 1.0e-04 loss: 4.366e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 551/30000 [1:16:28<55:20:13,  6.76s/it, lr: 1.0e-04 loss: 4.366e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 551/30000 [1:16:28<55:20:13,  6.76s/it, lr: 1.0e-04 loss: 4.366e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 551/30000 [1:16:35<55:20:13,  6.76s/it, lr: 1.0e-04 loss: 9.271e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 551/30000 [1:16:35<55:20:13,  6.76s/it, lr: 1.0e-04 loss: 9.271e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 552/30000 [1:16:35<55:39:15,  6.80s/it, lr: 1.0e-04 loss: 9.271e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 552/30000 [1:16:35<55:39:15,  6.80s/it, lr: 1.0e-04 loss: 9.271e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 552/30000 [1:16:42<55:39:15,  6.80s/it, lr: 1.0e-04 loss: 8.272e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 552/30000 [1:16:42<55:39:15,  6.80s/it, lr: 1.0e-04 loss: 8.272e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 553/30000 [1:16:42<55:07:54,  6.74s/it, lr: 1.0e-04 loss: 8.272e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 553/30000 [1:16:42<55:07:54,  6.74s/it, lr: 1.0e-04 loss: 8.272e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 553/30000 [1:16:51<55:07:54,  6.74s/it, lr: 1.0e-04 loss: 5.153e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 553/30000 [1:16:51<55:07:54,  6.74s/it, lr: 1.0e-04 loss: 5.153e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 554/30000 [1:16:51<60:59:22,  7.46s/it, lr: 1.0e-04 loss: 5.153e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 554/30000 [1:16:51<60:59:22,  7.46s/it, lr: 1.0e-04 loss: 5.153e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 554/30000 [1:16:58<60:59:22,  7.46s/it, lr: 1.0e-04 loss: 5.606e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 554/30000 [1:16:58<60:59:22,  7.46s/it, lr: 1.0e-04 loss: 5.606e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 555/30000 [1:16:58<59:08:17,  7.23s/it, lr: 1.0e-04 loss: 5.606e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 555/30000 [1:16:58<59:08:17,  7.23s/it, lr: 1.0e-04 loss: 5.606e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 555/30000 [1:17:08<59:08:17,  7.23s/it, lr: 1.0e-04 loss: 5.900e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 555/30000 [1:17:08<59:08:17,  7.23s/it, lr: 1.0e-04 loss: 5.900e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 556/30000 [1:17:08<64:46:47,  7.92s/it, lr: 1.0e-04 loss: 5.900e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 556/30000 [1:17:08<64:46:47,  7.92s/it, lr: 1.0e-04 loss: 5.900e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 556/30000 [1:17:14<64:46:47,  7.92s/it, lr: 1.0e-04 loss: 7.453e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 556/30000 [1:17:14<64:46:47,  7.92s/it, lr: 1.0e-04 loss: 7.453e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 557/30000 [1:17:14<61:41:47,  7.54s/it, lr: 1.0e-04 loss: 7.453e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 557/30000 [1:17:14<61:41:47,  7.54s/it, lr: 1.0e-04 loss: 7.453e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 557/30000 [1:17:21<61:41:47,  7.54s/it, lr: 1.0e-04 loss: 1.007e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 557/30000 [1:17:21<61:41:47,  7.54s/it, lr: 1.0e-04 loss: 1.007e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 558/30000 [1:17:21<59:19:57,  7.25s/it, lr: 1.0e-04 loss: 1.007e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 558/30000 [1:17:21<59:19:57,  7.25s/it, lr: 1.0e-04 loss: 1.007e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 558/30000 [1:17:28<59:19:57,  7.25s/it, lr: 1.0e-04 loss: 8.459e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 558/30000 [1:17:28<59:19:57,  7.25s/it, lr: 1.0e-04 loss: 8.459e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 559/30000 [1:17:28<58:20:16,  7.13s/it, lr: 1.0e-04 loss: 8.459e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 559/30000 [1:17:28<58:20:16,  7.13s/it, lr: 1.0e-04 loss: 8.459e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 559/30000 [1:17:34<58:20:16,  7.13s/it, lr: 1.0e-04 loss: 1.743e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 559/30000 [1:17:34<58:20:16,  7.13s/it, lr: 1.0e-04 loss: 1.743e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 560/30000 [1:17:41<58:20:09,  7.13s/it, lr: 1.0e-04 loss: 1.013e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 560/30000 [1:17:41<58:20:09,  7.13s/it, lr: 1.0e-04 loss: 1.013e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 561/30000 [1:17:41<43:58:33,  5.38s/it, lr: 1.0e-04 loss: 1.013e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 561/30000 [1:17:41<43:58:33,  5.38s/it, lr: 1.0e-04 loss: 1.013e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 561/30000 [1:17:50<43:58:33,  5.38s/it, lr: 1.0e-04 loss: 7.526e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 561/30000 [1:17:50<43:58:33,  5.38s/it, lr: 1.0e-04 loss: 7.526e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 562/30000 [1:17:50<51:58:49,  6.36s/it, lr: 1.0e-04 loss: 7.526e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 562/30000 [1:17:50<51:58:49,  6.36s/it, lr: 1.0e-04 loss: 7.526e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 562/30000 [1:18:00<51:58:49,  6.36s/it, lr: 1.0e-04 loss: 5.625e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 562/30000 [1:18:00<51:58:49,  6.36s/it, lr: 1.0e-04 loss: 5.625e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 563/30000 [1:18:00<58:39:13,  7.17s/it, lr: 1.0e-04 loss: 5.625e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 563/30000 [1:18:00<58:39:13,  7.17s/it, lr: 1.0e-04 loss: 5.625e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 563/30000 [1:18:06<58:39:13,  7.17s/it, lr: 1.0e-04 loss: 1.203e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 563/30000 [1:18:06<58:39:13,  7.17s/it, lr: 1.0e-04 loss: 1.203e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 564/30000 [1:18:06<57:28:11,  7.03s/it, lr: 1.0e-04 loss: 1.203e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 564/30000 [1:18:06<57:28:11,  7.03s/it, lr: 1.0e-04 loss: 1.203e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 564/30000 [1:18:16<57:28:11,  7.03s/it, lr: 1.0e-04 loss: 1.379e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 564/30000 [1:18:16<57:28:11,  7.03s/it, lr: 1.0e-04 loss: 1.379e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 565/30000 [1:18:16<62:40:46,  7.67s/it, lr: 1.0e-04 loss: 1.379e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 565/30000 [1:18:16<62:40:46,  7.67s/it, lr: 1.0e-04 loss: 1.379e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 565/30000 [1:18:25<62:40:46,  7.67s/it, lr: 1.0e-04 loss: 6.950e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 565/30000 [1:18:25<62:40:46,  7.67s/it, lr: 1.0e-04 loss: 6.950e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 566/30000 [1:18:25<66:38:46,  8.15s/it, lr: 1.0e-04 loss: 6.950e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 566/30000 [1:18:25<66:38:46,  8.15s/it, lr: 1.0e-04 loss: 6.950e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 566/30000 [1:18:35<66:38:46,  8.15s/it, lr: 1.0e-04 loss: 1.238e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 566/30000 [1:18:35<66:38:46,  8.15s/it, lr: 1.0e-04 loss: 1.238e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 567/30000 [1:18:35<70:36:30,  8.64s/it, lr: 1.0e-04 loss: 1.238e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 567/30000 [1:18:35<70:36:30,  8.64s/it, lr: 1.0e-04 loss: 1.238e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 567/30000 [1:18:44<70:36:30,  8.64s/it, lr: 1.0e-04 loss: 7.878e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 567/30000 [1:18:44<70:36:30,  8.64s/it, lr: 1.0e-04 loss: 7.878e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 568/30000 [1:18:44<72:26:06,  8.86s/it, lr: 1.0e-04 loss: 7.878e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 568/30000 [1:18:44<72:26:06,  8.86s/it, lr: 1.0e-04 loss: 7.878e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 568/30000 [1:18:54<72:26:06,  8.86s/it, lr: 1.0e-04 loss: 1.614e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 568/30000 [1:18:54<72:26:06,  8.86s/it, lr: 1.0e-04 loss: 1.614e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 569/30000 [1:18:54<73:46:22,  9.02s/it, lr: 1.0e-04 loss: 1.614e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 569/30000 [1:18:54<73:46:22,  9.02s/it, lr: 1.0e-04 loss: 1.614e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 569/30000 [1:19:03<73:46:22,  9.02s/it, lr: 1.0e-04 loss: 9.317e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 569/30000 [1:19:03<73:46:22,  9.02s/it, lr: 1.0e-04 loss: 9.317e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 570/30000 [1:19:12<73:46:13,  9.02s/it, lr: 1.0e-04 loss: 8.898e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 570/30000 [1:19:12<73:46:13,  9.02s/it, lr: 1.0e-04 loss: 8.898e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 571/30000 [1:19:12<57:30:46,  7.04s/it, lr: 1.0e-04 loss: 8.898e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 571/30000 [1:19:12<57:30:46,  7.04s/it, lr: 1.0e-04 loss: 8.898e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 571/30000 [1:19:19<57:30:46,  7.04s/it, lr: 1.0e-04 loss: 8.905e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 571/30000 [1:19:19<57:30:46,  7.04s/it, lr: 1.0e-04 loss: 8.905e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 572/30000 [1:19:19<57:04:33,  6.98s/it, lr: 1.0e-04 loss: 8.905e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 572/30000 [1:19:19<57:04:33,  6.98s/it, lr: 1.0e-04 loss: 8.905e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 572/30000 [1:19:26<57:04:33,  6.98s/it, lr: 1.0e-04 loss: 2.530e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 572/30000 [1:19:26<57:04:33,  6.98s/it, lr: 1.0e-04 loss: 2.530e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 573/30000 [1:19:26<56:14:39,  6.88s/it, lr: 1.0e-04 loss: 2.530e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 573/30000 [1:19:26<56:14:39,  6.88s/it, lr: 1.0e-04 loss: 2.530e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 573/30000 [1:19:33<56:14:39,  6.88s/it, lr: 1.0e-04 loss: 8.848e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 573/30000 [1:19:33<56:14:39,  6.88s/it, lr: 1.0e-04 loss: 8.848e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 574/30000 [1:19:33<56:19:45,  6.89s/it, lr: 1.0e-04 loss: 8.848e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 574/30000 [1:19:33<56:19:45,  6.89s/it, lr: 1.0e-04 loss: 8.848e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 574/30000 [1:19:42<56:19:45,  6.89s/it, lr: 1.0e-04 loss: 2.406e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 574/30000 [1:19:42<56:19:45,  6.89s/it, lr: 1.0e-04 loss: 2.406e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 575/30000 [1:19:42<62:07:34,  7.60s/it, lr: 1.0e-04 loss: 2.406e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 575/30000 [1:19:42<62:07:34,  7.60s/it, lr: 1.0e-04 loss: 2.406e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 575/30000 [1:19:49<62:07:34,  7.60s/it, lr: 1.0e-04 loss: 1.099e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 575/30000 [1:19:49<62:07:34,  7.60s/it, lr: 1.0e-04 loss: 1.099e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 576/30000 [1:19:49<59:54:59,  7.33s/it, lr: 1.0e-04 loss: 1.099e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 576/30000 [1:19:49<59:54:59,  7.33s/it, lr: 1.0e-04 loss: 1.099e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 576/30000 [1:19:56<59:54:59,  7.33s/it, lr: 1.0e-04 loss: 5.084e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 576/30000 [1:19:56<59:54:59,  7.33s/it, lr: 1.0e-04 loss: 5.084e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 577/30000 [1:19:56<58:41:58,  7.18s/it, lr: 1.0e-04 loss: 5.084e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 577/30000 [1:19:56<58:41:58,  7.18s/it, lr: 1.0e-04 loss: 5.084e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 577/30000 [1:20:05<58:41:58,  7.18s/it, lr: 1.0e-04 loss: 9.902e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 577/30000 [1:20:05<58:41:58,  7.18s/it, lr: 1.0e-04 loss: 9.902e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 578/30000 [1:20:05<64:00:07,  7.83s/it, lr: 1.0e-04 loss: 9.902e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 578/30000 [1:20:05<64:00:07,  7.83s/it, lr: 1.0e-04 loss: 9.902e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 578/30000 [1:20:14<64:00:07,  7.83s/it, lr: 1.0e-04 loss: 5.560e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 578/30000 [1:20:14<64:00:07,  7.83s/it, lr: 1.0e-04 loss: 5.560e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 579/30000 [1:20:14<67:40:34,  8.28s/it, lr: 1.0e-04 loss: 5.560e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 579/30000 [1:20:14<67:40:34,  8.28s/it, lr: 1.0e-04 loss: 5.560e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 579/30000 [1:20:24<67:40:34,  8.28s/it, lr: 1.0e-04 loss: 9.174e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 579/30000 [1:20:24<67:40:34,  8.28s/it, lr: 1.0e-04 loss: 9.174e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 580/30000 [1:20:31<67:40:26,  8.28s/it, lr: 1.0e-04 loss: 9.734e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 580/30000 [1:20:31<67:40:26,  8.28s/it, lr: 1.0e-04 loss: 9.734e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 581/30000 [1:20:31<49:31:52,  6.06s/it, lr: 1.0e-04 loss: 9.734e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 581/30000 [1:20:31<49:31:52,  6.06s/it, lr: 1.0e-04 loss: 9.734e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 581/30000 [1:20:37<49:31:52,  6.06s/it, lr: 1.0e-04 loss: 8.968e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 581/30000 [1:20:37<49:31:52,  6.06s/it, lr: 1.0e-04 loss: 8.968e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 582/30000 [1:20:37<50:47:05,  6.21s/it, lr: 1.0e-04 loss: 8.968e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 582/30000 [1:20:37<50:47:05,  6.21s/it, lr: 1.0e-04 loss: 8.968e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 582/30000 [1:20:47<50:47:05,  6.21s/it, lr: 1.0e-04 loss: 2.168e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 582/30000 [1:20:47<50:47:05,  6.21s/it, lr: 1.0e-04 loss: 2.168e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 583/30000 [1:20:47<57:35:20,  7.05s/it, lr: 1.0e-04 loss: 2.168e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 583/30000 [1:20:47<57:35:20,  7.05s/it, lr: 1.0e-04 loss: 2.168e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 583/30000 [1:20:56<57:35:20,  7.05s/it, lr: 1.0e-04 loss: 1.345e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 583/30000 [1:20:56<57:35:20,  7.05s/it, lr: 1.0e-04 loss: 1.345e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 584/30000 [1:20:56<62:50:00,  7.69s/it, lr: 1.0e-04 loss: 1.345e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 584/30000 [1:20:56<62:50:00,  7.69s/it, lr: 1.0e-04 loss: 1.345e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 584/30000 [1:21:06<62:50:00,  7.69s/it, lr: 1.0e-04 loss: 6.618e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 584/30000 [1:21:06<62:50:00,  7.69s/it, lr: 1.0e-04 loss: 6.618e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 585/30000 [1:21:06<67:21:32,  8.24s/it, lr: 1.0e-04 loss: 6.618e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 585/30000 [1:21:06<67:21:32,  8.24s/it, lr: 1.0e-04 loss: 6.618e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 585/30000 [1:21:15<67:21:32,  8.24s/it, lr: 1.0e-04 loss: 8.814e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 585/30000 [1:21:15<67:21:32,  8.24s/it, lr: 1.0e-04 loss: 8.814e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 586/30000 [1:21:15<70:11:42,  8.59s/it, lr: 1.0e-04 loss: 8.814e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 586/30000 [1:21:15<70:11:42,  8.59s/it, lr: 1.0e-04 loss: 8.814e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 586/30000 [1:21:25<70:11:42,  8.59s/it, lr: 1.0e-04 loss: 4.354e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 586/30000 [1:21:25<70:11:42,  8.59s/it, lr: 1.0e-04 loss: 4.354e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 587/30000 [1:21:25<72:09:01,  8.83s/it, lr: 1.0e-04 loss: 4.354e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 587/30000 [1:21:25<72:09:01,  8.83s/it, lr: 1.0e-04 loss: 4.354e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 587/30000 [1:21:32<72:09:01,  8.83s/it, lr: 1.0e-04 loss: 9.263e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 587/30000 [1:21:32<72:09:01,  8.83s/it, lr: 1.0e-04 loss: 9.263e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 588/30000 [1:21:32<66:57:41,  8.20s/it, lr: 1.0e-04 loss: 9.263e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 588/30000 [1:21:32<66:57:41,  8.20s/it, lr: 1.0e-04 loss: 9.263e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 588/30000 [1:21:38<66:57:41,  8.20s/it, lr: 1.0e-04 loss: 6.845e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 588/30000 [1:21:38<66:57:41,  8.20s/it, lr: 1.0e-04 loss: 6.845e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 589/30000 [1:21:38<63:55:38,  7.82s/it, lr: 1.0e-04 loss: 6.845e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 589/30000 [1:21:38<63:55:38,  7.82s/it, lr: 1.0e-04 loss: 6.845e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 589/30000 [1:21:48<63:55:38,  7.82s/it, lr: 1.0e-04 loss: 6.671e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 589/30000 [1:21:48<63:55:38,  7.82s/it, lr: 1.0e-04 loss: 6.671e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 590/30000 [1:21:55<63:55:30,  7.82s/it, lr: 1.0e-04 loss: 6.278e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 590/30000 [1:21:55<63:55:30,  7.82s/it, lr: 1.0e-04 loss: 6.278e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 591/30000 [1:21:55<47:09:33,  5.77s/it, lr: 1.0e-04 loss: 6.278e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 591/30000 [1:21:55<47:09:33,  5.77s/it, lr: 1.0e-04 loss: 6.278e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 591/30000 [1:22:01<47:09:33,  5.77s/it, lr: 1.0e-04 loss: 1.035e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 591/30000 [1:22:01<47:09:33,  5.77s/it, lr: 1.0e-04 loss: 1.035e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 592/30000 [1:22:01<48:48:27,  5.97s/it, lr: 1.0e-04 loss: 1.035e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 592/30000 [1:22:01<48:48:27,  5.97s/it, lr: 1.0e-04 loss: 1.035e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 592/30000 [1:22:08<48:48:27,  5.97s/it, lr: 1.0e-04 loss: 6.154e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 592/30000 [1:22:08<48:48:27,  5.97s/it, lr: 1.0e-04 loss: 6.154e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 593/30000 [1:22:08<50:06:01,  6.13s/it, lr: 1.0e-04 loss: 6.154e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 593/30000 [1:22:08<50:06:01,  6.13s/it, lr: 1.0e-04 loss: 6.154e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 593/30000 [1:22:17<50:06:01,  6.13s/it, lr: 1.0e-04 loss: 8.966e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 593/30000 [1:22:17<50:06:01,  6.13s/it, lr: 1.0e-04 loss: 8.966e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 594/30000 [1:22:17<57:23:47,  7.03s/it, lr: 1.0e-04 loss: 8.966e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 594/30000 [1:22:17<57:23:47,  7.03s/it, lr: 1.0e-04 loss: 8.966e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 594/30000 [1:22:24<57:23:47,  7.03s/it, lr: 1.0e-04 loss: 3.674e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 594/30000 [1:22:24<57:23:47,  7.03s/it, lr: 1.0e-04 loss: 3.674e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 595/30000 [1:22:24<56:37:22,  6.93s/it, lr: 1.0e-04 loss: 3.674e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 595/30000 [1:22:24<56:37:22,  6.93s/it, lr: 1.0e-04 loss: 3.674e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 595/30000 [1:22:34<56:37:22,  6.93s/it, lr: 1.0e-04 loss: 7.076e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 595/30000 [1:22:34<56:37:22,  6.93s/it, lr: 1.0e-04 loss: 7.076e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 596/30000 [1:22:34<63:27:56,  7.77s/it, lr: 1.0e-04 loss: 7.076e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 596/30000 [1:22:34<63:27:56,  7.77s/it, lr: 1.0e-04 loss: 7.076e-02]
Xiang_wan_2_2_lora:   2%|7                                       | 596/30000 [1:22:40<63:27:56,  7.77s/it, lr: 1.0e-04 loss: 1.758e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 596/30000 [1:22:40<63:27:56,  7.77s/it, lr: 1.0e-04 loss: 1.758e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 597/30000 [1:22:40<60:52:41,  7.45s/it, lr: 1.0e-04 loss: 1.758e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 597/30000 [1:22:40<60:52:41,  7.45s/it, lr: 1.0e-04 loss: 1.758e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 597/30000 [1:22:47<60:52:41,  7.45s/it, lr: 1.0e-04 loss: 1.155e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 597/30000 [1:22:47<60:52:41,  7.45s/it, lr: 1.0e-04 loss: 1.155e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 598/30000 [1:22:47<59:09:27,  7.24s/it, lr: 1.0e-04 loss: 1.155e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 598/30000 [1:22:47<59:09:27,  7.24s/it, lr: 1.0e-04 loss: 1.155e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 598/30000 [1:22:54<59:09:27,  7.24s/it, lr: 1.0e-04 loss: 7.075e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 598/30000 [1:22:54<59:09:27,  7.24s/it, lr: 1.0e-04 loss: 7.075e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 599/30000 [1:22:54<57:31:20,  7.04s/it, lr: 1.0e-04 loss: 7.075e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 599/30000 [1:22:54<57:31:20,  7.04s/it, lr: 1.0e-04 loss: 7.075e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 599/30000 [1:23:00<57:31:20,  7.04s/it, lr: 1.0e-04 loss: 1.320e-01]
Xiang_wan_2_2_lora:   2%|7                                       | 599/30000 [1:23:00<57:31:20,  7.04s/it, lr: 1.0e-04 loss: 1.320e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 600/30000 [1:23:10<57:31:13,  7.04s/it, lr: 1.0e-04 loss: 4.498e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 600/30000 [1:23:10<57:31:13,  7.04s/it, lr: 1.0e-04 loss: 4.498e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 601/30000 [1:23:10<48:49:52,  5.98s/it, lr: 1.0e-04 loss: 4.498e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 601/30000 [1:23:10<48:49:52,  5.98s/it, lr: 1.0e-04 loss: 4.498e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 601/30000 [1:23:19<48:49:52,  5.98s/it, lr: 1.0e-04 loss: 8.713e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 601/30000 [1:23:19<48:49:52,  5.98s/it, lr: 1.0e-04 loss: 8.713e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 602/30000 [1:23:19<55:44:34,  6.83s/it, lr: 1.0e-04 loss: 8.713e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 602/30000 [1:23:19<55:44:34,  6.83s/it, lr: 1.0e-04 loss: 8.713e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 602/30000 [1:23:29<55:44:34,  6.83s/it, lr: 1.0e-04 loss: 7.188e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 602/30000 [1:23:29<55:44:34,  6.83s/it, lr: 1.0e-04 loss: 7.188e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 603/30000 [1:23:29<61:39:00,  7.55s/it, lr: 1.0e-04 loss: 7.188e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 603/30000 [1:23:29<61:39:00,  7.55s/it, lr: 1.0e-04 loss: 7.188e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 603/30000 [1:23:38<61:39:00,  7.55s/it, lr: 1.0e-04 loss: 1.039e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 603/30000 [1:23:38<61:39:00,  7.55s/it, lr: 1.0e-04 loss: 1.039e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 604/30000 [1:23:38<66:12:26,  8.11s/it, lr: 1.0e-04 loss: 1.039e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 604/30000 [1:23:38<66:12:26,  8.11s/it, lr: 1.0e-04 loss: 1.039e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 604/30000 [1:23:48<66:12:26,  8.11s/it, lr: 1.0e-04 loss: 2.626e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 604/30000 [1:23:48<66:12:26,  8.11s/it, lr: 1.0e-04 loss: 2.626e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 605/30000 [1:23:48<69:04:40,  8.46s/it, lr: 1.0e-04 loss: 2.626e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 605/30000 [1:23:48<69:04:40,  8.46s/it, lr: 1.0e-04 loss: 2.626e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 605/30000 [1:23:57<69:04:40,  8.46s/it, lr: 1.0e-04 loss: 8.376e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 605/30000 [1:23:57<69:04:40,  8.46s/it, lr: 1.0e-04 loss: 8.376e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 606/30000 [1:23:57<71:20:41,  8.74s/it, lr: 1.0e-04 loss: 8.376e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 606/30000 [1:23:57<71:20:41,  8.74s/it, lr: 1.0e-04 loss: 8.376e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 606/30000 [1:24:07<71:20:41,  8.74s/it, lr: 1.0e-04 loss: 4.026e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 606/30000 [1:24:07<71:20:41,  8.74s/it, lr: 1.0e-04 loss: 4.026e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 607/30000 [1:24:07<73:25:24,  8.99s/it, lr: 1.0e-04 loss: 4.026e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 607/30000 [1:24:07<73:25:24,  8.99s/it, lr: 1.0e-04 loss: 4.026e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 607/30000 [1:24:14<73:25:24,  8.99s/it, lr: 1.0e-04 loss: 1.255e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 607/30000 [1:24:14<73:25:24,  8.99s/it, lr: 1.0e-04 loss: 1.255e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 608/30000 [1:24:14<68:00:26,  8.33s/it, lr: 1.0e-04 loss: 1.255e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 608/30000 [1:24:14<68:00:26,  8.33s/it, lr: 1.0e-04 loss: 1.255e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 608/30000 [1:24:20<68:00:26,  8.33s/it, lr: 1.0e-04 loss: 5.873e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 608/30000 [1:24:20<68:00:26,  8.33s/it, lr: 1.0e-04 loss: 5.873e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 609/30000 [1:24:20<63:57:09,  7.83s/it, lr: 1.0e-04 loss: 5.873e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 609/30000 [1:24:20<63:57:09,  7.83s/it, lr: 1.0e-04 loss: 5.873e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 609/30000 [1:24:27<63:57:09,  7.83s/it, lr: 1.0e-04 loss: 8.284e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 609/30000 [1:24:27<63:57:09,  7.83s/it, lr: 1.0e-04 loss: 8.284e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 610/30000 [1:24:34<63:57:01,  7.83s/it, lr: 1.0e-04 loss: 2.837e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 610/30000 [1:24:34<63:57:01,  7.83s/it, lr: 1.0e-04 loss: 2.837e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 611/30000 [1:24:34<47:26:05,  5.81s/it, lr: 1.0e-04 loss: 2.837e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 611/30000 [1:24:34<47:26:05,  5.81s/it, lr: 1.0e-04 loss: 2.837e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 611/30000 [1:24:41<47:26:05,  5.81s/it, lr: 1.0e-04 loss: 1.181e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 611/30000 [1:24:41<47:26:05,  5.81s/it, lr: 1.0e-04 loss: 1.181e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 612/30000 [1:24:41<49:06:51,  6.02s/it, lr: 1.0e-04 loss: 1.181e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 612/30000 [1:24:41<49:06:51,  6.02s/it, lr: 1.0e-04 loss: 1.181e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 612/30000 [1:24:47<49:06:51,  6.02s/it, lr: 1.0e-04 loss: 6.312e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 612/30000 [1:24:47<49:06:51,  6.02s/it, lr: 1.0e-04 loss: 6.312e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 613/30000 [1:24:47<50:25:05,  6.18s/it, lr: 1.0e-04 loss: 6.312e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 613/30000 [1:24:47<50:25:05,  6.18s/it, lr: 1.0e-04 loss: 6.312e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 613/30000 [1:24:57<50:25:05,  6.18s/it, lr: 1.0e-04 loss: 7.449e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 613/30000 [1:24:57<50:25:05,  6.18s/it, lr: 1.0e-04 loss: 7.449e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 614/30000 [1:24:57<57:23:41,  7.03s/it, lr: 1.0e-04 loss: 7.449e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 614/30000 [1:24:57<57:23:41,  7.03s/it, lr: 1.0e-04 loss: 7.449e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 614/30000 [1:25:03<57:23:41,  7.03s/it, lr: 1.0e-04 loss: 5.272e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 614/30000 [1:25:03<57:23:41,  7.03s/it, lr: 1.0e-04 loss: 5.272e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 615/30000 [1:25:03<56:33:45,  6.93s/it, lr: 1.0e-04 loss: 5.272e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 615/30000 [1:25:03<56:33:45,  6.93s/it, lr: 1.0e-04 loss: 5.272e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 615/30000 [1:25:13<56:33:45,  6.93s/it, lr: 1.0e-04 loss: 1.009e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 615/30000 [1:25:13<56:33:45,  6.93s/it, lr: 1.0e-04 loss: 1.009e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 616/30000 [1:25:13<62:11:36,  7.62s/it, lr: 1.0e-04 loss: 1.009e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 616/30000 [1:25:13<62:11:36,  7.62s/it, lr: 1.0e-04 loss: 1.009e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 616/30000 [1:25:22<62:11:36,  7.62s/it, lr: 1.0e-04 loss: 9.701e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 616/30000 [1:25:22<62:11:36,  7.62s/it, lr: 1.0e-04 loss: 9.701e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 617/30000 [1:25:22<66:25:22,  8.14s/it, lr: 1.0e-04 loss: 9.701e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 617/30000 [1:25:22<66:25:22,  8.14s/it, lr: 1.0e-04 loss: 9.701e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 617/30000 [1:25:29<66:25:22,  8.14s/it, lr: 1.0e-04 loss: 1.238e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 617/30000 [1:25:29<66:25:22,  8.14s/it, lr: 1.0e-04 loss: 1.238e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 618/30000 [1:25:29<63:22:48,  7.77s/it, lr: 1.0e-04 loss: 1.238e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 618/30000 [1:25:29<63:22:48,  7.77s/it, lr: 1.0e-04 loss: 1.238e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 618/30000 [1:25:38<63:22:48,  7.77s/it, lr: 1.0e-04 loss: 7.082e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 618/30000 [1:25:38<63:22:48,  7.77s/it, lr: 1.0e-04 loss: 7.082e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 619/30000 [1:25:38<67:19:09,  8.25s/it, lr: 1.0e-04 loss: 7.082e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 619/30000 [1:25:38<67:19:09,  8.25s/it, lr: 1.0e-04 loss: 7.082e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 619/30000 [1:25:45<67:19:09,  8.25s/it, lr: 1.0e-04 loss: 1.017e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 619/30000 [1:25:45<67:19:09,  8.25s/it, lr: 1.0e-04 loss: 1.017e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 620/30000 [1:25:55<67:19:01,  8.25s/it, lr: 1.0e-04 loss: 2.374e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 620/30000 [1:25:55<67:19:01,  8.25s/it, lr: 1.0e-04 loss: 2.374e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 621/30000 [1:25:55<54:03:21,  6.62s/it, lr: 1.0e-04 loss: 2.374e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 621/30000 [1:25:55<54:03:21,  6.62s/it, lr: 1.0e-04 loss: 2.374e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 621/30000 [1:26:01<54:03:21,  6.62s/it, lr: 1.0e-04 loss: 1.298e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 621/30000 [1:26:01<54:03:21,  6.62s/it, lr: 1.0e-04 loss: 1.298e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 622/30000 [1:26:01<54:04:54,  6.63s/it, lr: 1.0e-04 loss: 1.298e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 622/30000 [1:26:01<54:04:54,  6.63s/it, lr: 1.0e-04 loss: 1.298e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 622/30000 [1:26:11<54:04:54,  6.63s/it, lr: 1.0e-04 loss: 5.252e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 622/30000 [1:26:11<54:04:54,  6.63s/it, lr: 1.0e-04 loss: 5.252e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 623/30000 [1:26:11<60:12:20,  7.38s/it, lr: 1.0e-04 loss: 5.252e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 623/30000 [1:26:11<60:12:20,  7.38s/it, lr: 1.0e-04 loss: 5.252e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 623/30000 [1:26:18<60:12:20,  7.38s/it, lr: 1.0e-04 loss: 9.887e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 623/30000 [1:26:18<60:12:20,  7.38s/it, lr: 1.0e-04 loss: 9.887e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 624/30000 [1:26:18<59:05:23,  7.24s/it, lr: 1.0e-04 loss: 9.887e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 624/30000 [1:26:18<59:05:23,  7.24s/it, lr: 1.0e-04 loss: 9.887e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 624/30000 [1:26:25<59:05:23,  7.24s/it, lr: 1.0e-04 loss: 5.706e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 624/30000 [1:26:25<59:05:23,  7.24s/it, lr: 1.0e-04 loss: 5.706e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 625/30000 [1:26:25<59:14:04,  7.26s/it, lr: 1.0e-04 loss: 5.706e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 625/30000 [1:26:25<59:14:04,  7.26s/it, lr: 1.0e-04 loss: 5.706e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 625/30000 [1:26:34<59:14:04,  7.26s/it, lr: 1.0e-04 loss: 6.443e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 625/30000 [1:26:34<59:14:04,  7.26s/it, lr: 1.0e-04 loss: 6.443e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 626/30000 [1:26:34<64:10:25,  7.86s/it, lr: 1.0e-04 loss: 6.443e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 626/30000 [1:26:34<64:10:25,  7.86s/it, lr: 1.0e-04 loss: 6.443e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 626/30000 [1:26:44<64:10:25,  7.86s/it, lr: 1.0e-04 loss: 4.330e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 626/30000 [1:26:44<64:10:25,  7.86s/it, lr: 1.0e-04 loss: 4.330e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 627/30000 [1:26:44<67:47:22,  8.31s/it, lr: 1.0e-04 loss: 4.330e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 627/30000 [1:26:44<67:47:22,  8.31s/it, lr: 1.0e-04 loss: 4.330e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 627/30000 [1:26:50<67:47:22,  8.31s/it, lr: 1.0e-04 loss: 9.819e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 627/30000 [1:26:50<67:47:22,  8.31s/it, lr: 1.0e-04 loss: 9.819e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 628/30000 [1:26:50<63:50:09,  7.82s/it, lr: 1.0e-04 loss: 9.819e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 628/30000 [1:26:50<63:50:09,  7.82s/it, lr: 1.0e-04 loss: 9.819e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 628/30000 [1:27:00<63:50:09,  7.82s/it, lr: 1.0e-04 loss: 5.350e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 628/30000 [1:27:00<63:50:09,  7.82s/it, lr: 1.0e-04 loss: 5.350e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 629/30000 [1:27:00<67:30:31,  8.27s/it, lr: 1.0e-04 loss: 5.350e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 629/30000 [1:27:00<67:30:31,  8.27s/it, lr: 1.0e-04 loss: 5.350e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 629/30000 [1:27:06<67:30:31,  8.27s/it, lr: 1.0e-04 loss: 9.845e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 629/30000 [1:27:06<67:30:31,  8.27s/it, lr: 1.0e-04 loss: 9.845e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 630/30000 [1:27:16<67:30:22,  8.27s/it, lr: 1.0e-04 loss: 4.220e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 630/30000 [1:27:16<67:30:22,  8.27s/it, lr: 1.0e-04 loss: 4.220e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 631/30000 [1:27:16<54:12:22,  6.64s/it, lr: 1.0e-04 loss: 4.220e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 631/30000 [1:27:16<54:12:22,  6.64s/it, lr: 1.0e-04 loss: 4.220e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 631/30000 [1:27:23<54:12:22,  6.64s/it, lr: 1.0e-04 loss: 7.982e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 631/30000 [1:27:23<54:12:22,  6.64s/it, lr: 1.0e-04 loss: 7.982e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 632/30000 [1:27:23<54:44:02,  6.71s/it, lr: 1.0e-04 loss: 7.982e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 632/30000 [1:27:23<54:44:02,  6.71s/it, lr: 1.0e-04 loss: 7.982e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 632/30000 [1:27:29<54:44:02,  6.71s/it, lr: 1.0e-04 loss: 6.000e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 632/30000 [1:27:29<54:44:02,  6.71s/it, lr: 1.0e-04 loss: 6.000e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 633/30000 [1:27:29<54:27:52,  6.68s/it, lr: 1.0e-04 loss: 6.000e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 633/30000 [1:27:29<54:27:52,  6.68s/it, lr: 1.0e-04 loss: 6.000e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 633/30000 [1:27:39<54:27:52,  6.68s/it, lr: 1.0e-04 loss: 6.693e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 633/30000 [1:27:39<54:27:52,  6.68s/it, lr: 1.0e-04 loss: 6.693e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 634/30000 [1:27:39<60:21:46,  7.40s/it, lr: 1.0e-04 loss: 6.693e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 634/30000 [1:27:39<60:21:46,  7.40s/it, lr: 1.0e-04 loss: 6.693e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 634/30000 [1:27:45<60:21:46,  7.40s/it, lr: 1.0e-04 loss: 8.159e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 634/30000 [1:27:45<60:21:46,  7.40s/it, lr: 1.0e-04 loss: 8.159e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 635/30000 [1:27:45<58:33:08,  7.18s/it, lr: 1.0e-04 loss: 8.159e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 635/30000 [1:27:45<58:33:08,  7.18s/it, lr: 1.0e-04 loss: 8.159e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 635/30000 [1:27:55<58:33:08,  7.18s/it, lr: 1.0e-04 loss: 8.054e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 635/30000 [1:27:55<58:33:08,  7.18s/it, lr: 1.0e-04 loss: 8.054e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 636/30000 [1:27:55<63:30:42,  7.79s/it, lr: 1.0e-04 loss: 8.054e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 636/30000 [1:27:55<63:30:42,  7.79s/it, lr: 1.0e-04 loss: 8.054e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 636/30000 [1:28:04<63:30:42,  7.79s/it, lr: 1.0e-04 loss: 7.316e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 636/30000 [1:28:04<63:30:42,  7.79s/it, lr: 1.0e-04 loss: 7.316e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 637/30000 [1:28:04<67:11:55,  8.24s/it, lr: 1.0e-04 loss: 7.316e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 637/30000 [1:28:04<67:11:55,  8.24s/it, lr: 1.0e-04 loss: 7.316e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 637/30000 [1:28:13<67:11:55,  8.24s/it, lr: 1.0e-04 loss: 7.398e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 637/30000 [1:28:13<67:11:55,  8.24s/it, lr: 1.0e-04 loss: 7.398e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 638/30000 [1:28:13<69:48:10,  8.56s/it, lr: 1.0e-04 loss: 7.398e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 638/30000 [1:28:13<69:48:10,  8.56s/it, lr: 1.0e-04 loss: 7.398e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 638/30000 [1:28:23<69:48:10,  8.56s/it, lr: 1.0e-04 loss: 4.222e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 638/30000 [1:28:23<69:48:10,  8.56s/it, lr: 1.0e-04 loss: 4.222e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 639/30000 [1:28:23<72:21:14,  8.87s/it, lr: 1.0e-04 loss: 4.222e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 639/30000 [1:28:23<72:21:14,  8.87s/it, lr: 1.0e-04 loss: 4.222e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 639/30000 [1:28:32<72:21:14,  8.87s/it, lr: 1.0e-04 loss: 6.948e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 639/30000 [1:28:32<72:21:14,  8.87s/it, lr: 1.0e-04 loss: 6.948e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 640/30000 [1:28:39<72:21:05,  8.87s/it, lr: 1.0e-04 loss: 5.398e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 640/30000 [1:28:39<72:21:05,  8.87s/it, lr: 1.0e-04 loss: 5.398e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 641/30000 [1:28:39<51:43:30,  6.34s/it, lr: 1.0e-04 loss: 5.398e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 641/30000 [1:28:39<51:43:30,  6.34s/it, lr: 1.0e-04 loss: 5.398e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 641/30000 [1:28:48<51:43:30,  6.34s/it, lr: 1.0e-04 loss: 9.184e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 641/30000 [1:28:48<51:43:30,  6.34s/it, lr: 1.0e-04 loss: 9.184e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 642/30000 [1:28:48<57:43:31,  7.08s/it, lr: 1.0e-04 loss: 9.184e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 642/30000 [1:28:48<57:43:31,  7.08s/it, lr: 1.0e-04 loss: 9.184e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 642/30000 [1:28:58<57:43:31,  7.08s/it, lr: 1.0e-04 loss: 8.631e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 642/30000 [1:28:58<57:43:31,  7.08s/it, lr: 1.0e-04 loss: 8.631e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 643/30000 [1:28:58<62:38:29,  7.68s/it, lr: 1.0e-04 loss: 8.631e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 643/30000 [1:28:58<62:38:29,  7.68s/it, lr: 1.0e-04 loss: 8.631e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 643/30000 [1:29:04<62:38:29,  7.68s/it, lr: 1.0e-04 loss: 1.218e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 643/30000 [1:29:04<62:38:29,  7.68s/it, lr: 1.0e-04 loss: 1.218e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 644/30000 [1:29:04<60:23:11,  7.41s/it, lr: 1.0e-04 loss: 1.218e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 644/30000 [1:29:04<60:23:11,  7.41s/it, lr: 1.0e-04 loss: 1.218e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 644/30000 [1:29:14<60:23:11,  7.41s/it, lr: 1.0e-04 loss: 6.352e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 644/30000 [1:29:14<60:23:11,  7.41s/it, lr: 1.0e-04 loss: 6.352e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 645/30000 [1:29:14<64:50:46,  7.95s/it, lr: 1.0e-04 loss: 6.352e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 645/30000 [1:29:14<64:50:46,  7.95s/it, lr: 1.0e-04 loss: 6.352e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 645/30000 [1:29:23<64:50:46,  7.95s/it, lr: 1.0e-04 loss: 5.875e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 645/30000 [1:29:23<64:50:46,  7.95s/it, lr: 1.0e-04 loss: 5.875e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 646/30000 [1:29:23<68:09:19,  8.36s/it, lr: 1.0e-04 loss: 5.875e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 646/30000 [1:29:23<68:09:19,  8.36s/it, lr: 1.0e-04 loss: 5.875e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 646/30000 [1:29:33<68:09:19,  8.36s/it, lr: 1.0e-04 loss: 8.700e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 646/30000 [1:29:33<68:09:19,  8.36s/it, lr: 1.0e-04 loss: 8.700e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 647/30000 [1:29:33<71:05:42,  8.72s/it, lr: 1.0e-04 loss: 8.700e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 647/30000 [1:29:33<71:05:42,  8.72s/it, lr: 1.0e-04 loss: 8.700e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 647/30000 [1:29:42<71:05:42,  8.72s/it, lr: 1.0e-04 loss: 8.996e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 647/30000 [1:29:42<71:05:42,  8.72s/it, lr: 1.0e-04 loss: 8.996e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 648/30000 [1:29:42<72:40:19,  8.91s/it, lr: 1.0e-04 loss: 8.996e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 648/30000 [1:29:42<72:40:19,  8.91s/it, lr: 1.0e-04 loss: 8.996e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 648/30000 [1:29:49<72:40:19,  8.91s/it, lr: 1.0e-04 loss: 1.536e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 648/30000 [1:29:49<72:40:19,  8.91s/it, lr: 1.0e-04 loss: 1.536e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 649/30000 [1:29:49<67:05:06,  8.23s/it, lr: 1.0e-04 loss: 1.536e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 649/30000 [1:29:49<67:05:06,  8.23s/it, lr: 1.0e-04 loss: 1.536e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 649/30000 [1:29:58<67:05:06,  8.23s/it, lr: 1.0e-04 loss: 6.911e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 649/30000 [1:29:58<67:05:06,  8.23s/it, lr: 1.0e-04 loss: 6.911e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 650/30000 [1:30:05<67:04:57,  8.23s/it, lr: 1.0e-04 loss: 6.955e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 650/30000 [1:30:05<67:04:57,  8.23s/it, lr: 1.0e-04 loss: 6.955e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 651/30000 [1:30:05<48:47:42,  5.99s/it, lr: 1.0e-04 loss: 6.955e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 651/30000 [1:30:05<48:47:42,  5.99s/it, lr: 1.0e-04 loss: 6.955e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 651/30000 [1:30:14<48:47:42,  5.99s/it, lr: 1.0e-04 loss: 6.907e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 651/30000 [1:30:14<48:47:42,  5.99s/it, lr: 1.0e-04 loss: 6.907e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 652/30000 [1:30:14<55:33:22,  6.81s/it, lr: 1.0e-04 loss: 6.907e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 652/30000 [1:30:14<55:33:22,  6.81s/it, lr: 1.0e-04 loss: 6.907e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 652/30000 [1:30:24<55:33:22,  6.81s/it, lr: 1.0e-04 loss: 4.594e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 652/30000 [1:30:24<55:33:22,  6.81s/it, lr: 1.0e-04 loss: 4.594e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 653/30000 [1:30:24<61:27:37,  7.54s/it, lr: 1.0e-04 loss: 4.594e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 653/30000 [1:30:24<61:27:37,  7.54s/it, lr: 1.0e-04 loss: 4.594e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 653/30000 [1:30:33<61:27:37,  7.54s/it, lr: 1.0e-04 loss: 7.113e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 653/30000 [1:30:33<61:27:37,  7.54s/it, lr: 1.0e-04 loss: 7.113e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 654/30000 [1:30:33<65:55:57,  8.09s/it, lr: 1.0e-04 loss: 7.113e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 654/30000 [1:30:33<65:55:57,  8.09s/it, lr: 1.0e-04 loss: 7.113e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 654/30000 [1:30:40<65:55:57,  8.09s/it, lr: 1.0e-04 loss: 1.525e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 654/30000 [1:30:40<65:55:57,  8.09s/it, lr: 1.0e-04 loss: 1.525e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 655/30000 [1:30:40<62:40:34,  7.69s/it, lr: 1.0e-04 loss: 1.525e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 655/30000 [1:30:40<62:40:34,  7.69s/it, lr: 1.0e-04 loss: 1.525e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 655/30000 [1:30:49<62:40:34,  7.69s/it, lr: 1.0e-04 loss: 8.190e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 655/30000 [1:30:49<62:40:34,  7.69s/it, lr: 1.0e-04 loss: 8.190e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 656/30000 [1:30:49<66:33:25,  8.17s/it, lr: 1.0e-04 loss: 8.190e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 656/30000 [1:30:49<66:33:25,  8.17s/it, lr: 1.0e-04 loss: 8.190e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 656/30000 [1:30:56<66:33:25,  8.17s/it, lr: 1.0e-04 loss: 7.075e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 656/30000 [1:30:56<66:33:25,  8.17s/it, lr: 1.0e-04 loss: 7.075e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 657/30000 [1:30:56<62:59:09,  7.73s/it, lr: 1.0e-04 loss: 7.075e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 657/30000 [1:30:56<62:59:09,  7.73s/it, lr: 1.0e-04 loss: 7.075e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 657/30000 [1:31:05<62:59:09,  7.73s/it, lr: 1.0e-04 loss: 7.481e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 657/30000 [1:31:05<62:59:09,  7.73s/it, lr: 1.0e-04 loss: 7.481e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 658/30000 [1:31:05<66:49:14,  8.20s/it, lr: 1.0e-04 loss: 7.481e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 658/30000 [1:31:05<66:49:14,  8.20s/it, lr: 1.0e-04 loss: 7.481e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 658/30000 [1:31:15<66:49:14,  8.20s/it, lr: 1.0e-04 loss: 5.094e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 658/30000 [1:31:15<66:49:14,  8.20s/it, lr: 1.0e-04 loss: 5.094e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 659/30000 [1:31:15<69:41:12,  8.55s/it, lr: 1.0e-04 loss: 5.094e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 659/30000 [1:31:15<69:41:12,  8.55s/it, lr: 1.0e-04 loss: 5.094e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 659/30000 [1:31:24<69:41:12,  8.55s/it, lr: 1.0e-04 loss: 7.895e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 659/30000 [1:31:24<69:41:12,  8.55s/it, lr: 1.0e-04 loss: 7.895e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 660/30000 [1:31:34<69:41:04,  8.55s/it, lr: 1.0e-04 loss: 7.855e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 660/30000 [1:31:34<69:41:04,  8.55s/it, lr: 1.0e-04 loss: 7.855e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 661/30000 [1:31:34<55:38:51,  6.83s/it, lr: 1.0e-04 loss: 7.855e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 661/30000 [1:31:34<55:38:51,  6.83s/it, lr: 1.0e-04 loss: 7.855e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 661/30000 [1:31:43<55:38:51,  6.83s/it, lr: 1.0e-04 loss: 8.904e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 661/30000 [1:31:43<55:38:51,  6.83s/it, lr: 1.0e-04 loss: 8.904e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 662/30000 [1:31:43<60:41:45,  7.45s/it, lr: 1.0e-04 loss: 8.904e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 662/30000 [1:31:43<60:41:45,  7.45s/it, lr: 1.0e-04 loss: 8.904e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 662/30000 [1:31:49<60:41:45,  7.45s/it, lr: 1.0e-04 loss: 9.089e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 662/30000 [1:31:49<60:41:45,  7.45s/it, lr: 1.0e-04 loss: 9.089e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 663/30000 [1:31:49<58:56:17,  7.23s/it, lr: 1.0e-04 loss: 9.089e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 663/30000 [1:31:49<58:56:17,  7.23s/it, lr: 1.0e-04 loss: 9.089e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 663/30000 [1:31:56<58:56:17,  7.23s/it, lr: 1.0e-04 loss: 7.758e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 663/30000 [1:31:56<58:56:17,  7.23s/it, lr: 1.0e-04 loss: 7.758e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 664/30000 [1:31:56<57:31:29,  7.06s/it, lr: 1.0e-04 loss: 7.758e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 664/30000 [1:31:56<57:31:29,  7.06s/it, lr: 1.0e-04 loss: 7.758e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 664/30000 [1:32:03<57:31:29,  7.06s/it, lr: 1.0e-04 loss: 6.855e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 664/30000 [1:32:03<57:31:29,  7.06s/it, lr: 1.0e-04 loss: 6.855e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 665/30000 [1:32:03<56:22:01,  6.92s/it, lr: 1.0e-04 loss: 6.855e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 665/30000 [1:32:03<56:22:01,  6.92s/it, lr: 1.0e-04 loss: 6.855e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 665/30000 [1:32:09<56:22:01,  6.92s/it, lr: 1.0e-04 loss: 8.569e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 665/30000 [1:32:09<56:22:01,  6.92s/it, lr: 1.0e-04 loss: 8.569e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 666/30000 [1:32:09<55:47:17,  6.85s/it, lr: 1.0e-04 loss: 8.569e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 666/30000 [1:32:09<55:47:17,  6.85s/it, lr: 1.0e-04 loss: 8.569e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 666/30000 [1:32:19<55:47:17,  6.85s/it, lr: 1.0e-04 loss: 3.877e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 666/30000 [1:32:19<55:47:17,  6.85s/it, lr: 1.0e-04 loss: 3.877e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 667/30000 [1:32:19<61:46:22,  7.58s/it, lr: 1.0e-04 loss: 3.877e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 667/30000 [1:32:19<61:46:22,  7.58s/it, lr: 1.0e-04 loss: 3.877e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 667/30000 [1:32:26<61:46:22,  7.58s/it, lr: 1.0e-04 loss: 1.207e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 667/30000 [1:32:26<61:46:22,  7.58s/it, lr: 1.0e-04 loss: 1.207e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 668/30000 [1:32:26<59:59:03,  7.36s/it, lr: 1.0e-04 loss: 1.207e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 668/30000 [1:32:26<59:59:03,  7.36s/it, lr: 1.0e-04 loss: 1.207e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 668/30000 [1:32:32<59:59:03,  7.36s/it, lr: 1.0e-04 loss: 6.296e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 668/30000 [1:32:32<59:59:03,  7.36s/it, lr: 1.0e-04 loss: 6.296e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 669/30000 [1:32:32<58:08:12,  7.14s/it, lr: 1.0e-04 loss: 6.296e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 669/30000 [1:32:32<58:08:12,  7.14s/it, lr: 1.0e-04 loss: 6.296e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 669/30000 [1:32:41<58:08:12,  7.14s/it, lr: 1.0e-04 loss: 1.114e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 669/30000 [1:32:41<58:08:12,  7.14s/it, lr: 1.0e-04 loss: 1.114e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 670/30000 [1:32:48<58:08:05,  7.14s/it, lr: 1.0e-04 loss: 4.228e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 670/30000 [1:32:48<58:08:05,  7.14s/it, lr: 1.0e-04 loss: 4.228e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 671/30000 [1:32:48<44:19:26,  5.44s/it, lr: 1.0e-04 loss: 4.228e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 671/30000 [1:32:48<44:19:26,  5.44s/it, lr: 1.0e-04 loss: 4.228e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 671/30000 [1:32:58<44:19:26,  5.44s/it, lr: 1.0e-04 loss: 8.899e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 671/30000 [1:32:58<44:19:26,  5.44s/it, lr: 1.0e-04 loss: 8.899e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 672/30000 [1:32:58<52:15:37,  6.41s/it, lr: 1.0e-04 loss: 8.899e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 672/30000 [1:32:58<52:15:37,  6.41s/it, lr: 1.0e-04 loss: 8.899e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 672/30000 [1:33:04<52:15:37,  6.41s/it, lr: 1.0e-04 loss: 5.403e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 672/30000 [1:33:04<52:15:37,  6.41s/it, lr: 1.0e-04 loss: 5.403e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 673/30000 [1:33:04<52:42:31,  6.47s/it, lr: 1.0e-04 loss: 5.403e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 673/30000 [1:33:04<52:42:31,  6.47s/it, lr: 1.0e-04 loss: 5.403e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 673/30000 [1:33:11<52:42:31,  6.47s/it, lr: 1.0e-04 loss: 1.456e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 673/30000 [1:33:11<52:42:31,  6.47s/it, lr: 1.0e-04 loss: 1.456e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 674/30000 [1:33:11<52:51:51,  6.49s/it, lr: 1.0e-04 loss: 1.456e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 674/30000 [1:33:11<52:51:51,  6.49s/it, lr: 1.0e-04 loss: 1.456e-01]
Xiang_wan_2_2_lora:   2%|8                                       | 674/30000 [1:33:18<52:51:51,  6.49s/it, lr: 1.0e-04 loss: 6.619e-02]
Xiang_wan_2_2_lora:   2%|8                                       | 674/30000 [1:33:18<52:51:51,  6.49s/it, lr: 1.0e-04 loss: 6.619e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 675/30000 [1:33:18<53:20:32,  6.55s/it, lr: 1.0e-04 loss: 6.619e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 675/30000 [1:33:18<53:20:32,  6.55s/it, lr: 1.0e-04 loss: 6.619e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 675/30000 [1:33:24<53:20:32,  6.55s/it, lr: 1.0e-04 loss: 1.100e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 675/30000 [1:33:24<53:20:32,  6.55s/it, lr: 1.0e-04 loss: 1.100e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 676/30000 [1:33:24<54:05:52,  6.64s/it, lr: 1.0e-04 loss: 1.100e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 676/30000 [1:33:24<54:05:52,  6.64s/it, lr: 1.0e-04 loss: 1.100e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 676/30000 [1:33:34<54:05:52,  6.64s/it, lr: 1.0e-04 loss: 4.764e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 676/30000 [1:33:34<54:05:52,  6.64s/it, lr: 1.0e-04 loss: 4.764e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 677/30000 [1:33:34<61:07:06,  7.50s/it, lr: 1.0e-04 loss: 4.764e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 677/30000 [1:33:34<61:07:06,  7.50s/it, lr: 1.0e-04 loss: 4.764e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 677/30000 [1:33:41<61:07:06,  7.50s/it, lr: 1.0e-04 loss: 1.246e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 677/30000 [1:33:41<61:07:06,  7.50s/it, lr: 1.0e-04 loss: 1.246e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 678/30000 [1:33:41<59:07:54,  7.26s/it, lr: 1.0e-04 loss: 1.246e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 678/30000 [1:33:41<59:07:54,  7.26s/it, lr: 1.0e-04 loss: 1.246e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 678/30000 [1:33:47<59:07:54,  7.26s/it, lr: 1.0e-04 loss: 3.396e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 678/30000 [1:33:47<59:07:54,  7.26s/it, lr: 1.0e-04 loss: 3.396e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 679/30000 [1:33:47<57:42:59,  7.09s/it, lr: 1.0e-04 loss: 3.396e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 679/30000 [1:33:47<57:42:59,  7.09s/it, lr: 1.0e-04 loss: 3.396e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 679/30000 [1:33:54<57:42:59,  7.09s/it, lr: 1.0e-04 loss: 9.461e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 679/30000 [1:33:54<57:42:59,  7.09s/it, lr: 1.0e-04 loss: 9.461e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 680/30000 [1:34:01<57:42:52,  7.09s/it, lr: 1.0e-04 loss: 2.842e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 680/30000 [1:34:01<57:42:52,  7.09s/it, lr: 1.0e-04 loss: 2.842e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 681/30000 [1:34:01<43:36:23,  5.35s/it, lr: 1.0e-04 loss: 2.842e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 681/30000 [1:34:01<43:36:23,  5.35s/it, lr: 1.0e-04 loss: 2.842e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 681/30000 [1:34:07<43:36:23,  5.35s/it, lr: 1.0e-04 loss: 9.397e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 681/30000 [1:34:07<43:36:23,  5.35s/it, lr: 1.0e-04 loss: 9.397e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 682/30000 [1:34:07<46:04:32,  5.66s/it, lr: 1.0e-04 loss: 9.397e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 682/30000 [1:34:07<46:04:32,  5.66s/it, lr: 1.0e-04 loss: 9.397e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 682/30000 [1:34:17<46:04:32,  5.66s/it, lr: 1.0e-04 loss: 2.772e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 682/30000 [1:34:17<46:04:32,  5.66s/it, lr: 1.0e-04 loss: 2.772e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 683/30000 [1:34:17<54:18:46,  6.67s/it, lr: 1.0e-04 loss: 2.772e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 683/30000 [1:34:17<54:18:46,  6.67s/it, lr: 1.0e-04 loss: 2.772e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 683/30000 [1:34:23<54:18:46,  6.67s/it, lr: 1.0e-04 loss: 1.017e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 683/30000 [1:34:23<54:18:46,  6.67s/it, lr: 1.0e-04 loss: 1.017e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 684/30000 [1:34:23<54:32:48,  6.70s/it, lr: 1.0e-04 loss: 1.017e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 684/30000 [1:34:23<54:32:48,  6.70s/it, lr: 1.0e-04 loss: 1.017e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 684/30000 [1:34:30<54:32:48,  6.70s/it, lr: 1.0e-04 loss: 8.390e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 684/30000 [1:34:30<54:32:48,  6.70s/it, lr: 1.0e-04 loss: 8.390e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 685/30000 [1:34:30<54:24:42,  6.68s/it, lr: 1.0e-04 loss: 8.390e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 685/30000 [1:34:30<54:24:42,  6.68s/it, lr: 1.0e-04 loss: 8.390e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 685/30000 [1:34:40<54:24:42,  6.68s/it, lr: 1.0e-04 loss: 7.625e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 685/30000 [1:34:40<54:24:42,  6.68s/it, lr: 1.0e-04 loss: 7.625e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 686/30000 [1:34:40<60:41:48,  7.45s/it, lr: 1.0e-04 loss: 7.625e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 686/30000 [1:34:40<60:41:48,  7.45s/it, lr: 1.0e-04 loss: 7.625e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 686/30000 [1:34:49<60:41:48,  7.45s/it, lr: 1.0e-04 loss: 4.862e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 686/30000 [1:34:49<60:41:48,  7.45s/it, lr: 1.0e-04 loss: 4.862e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 687/30000 [1:34:49<65:29:08,  8.04s/it, lr: 1.0e-04 loss: 4.862e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 687/30000 [1:34:49<65:29:08,  8.04s/it, lr: 1.0e-04 loss: 4.862e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 687/30000 [1:34:58<65:29:08,  8.04s/it, lr: 1.0e-04 loss: 6.655e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 687/30000 [1:34:58<65:29:08,  8.04s/it, lr: 1.0e-04 loss: 6.655e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 688/30000 [1:34:58<68:47:30,  8.45s/it, lr: 1.0e-04 loss: 6.655e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 688/30000 [1:34:58<68:47:30,  8.45s/it, lr: 1.0e-04 loss: 6.655e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 688/30000 [1:35:08<68:47:30,  8.45s/it, lr: 1.0e-04 loss: 8.396e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 688/30000 [1:35:08<68:47:30,  8.45s/it, lr: 1.0e-04 loss: 8.396e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 689/30000 [1:35:08<70:53:49,  8.71s/it, lr: 1.0e-04 loss: 8.396e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 689/30000 [1:35:08<70:53:49,  8.71s/it, lr: 1.0e-04 loss: 8.396e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 689/30000 [1:35:18<70:53:49,  8.71s/it, lr: 1.0e-04 loss: 8.450e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 689/30000 [1:35:18<70:53:49,  8.71s/it, lr: 1.0e-04 loss: 8.450e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 690/30000 [1:35:27<70:53:40,  8.71s/it, lr: 1.0e-04 loss: 4.687e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 690/30000 [1:35:27<70:53:40,  8.71s/it, lr: 1.0e-04 loss: 4.687e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 691/30000 [1:35:27<55:58:43,  6.88s/it, lr: 1.0e-04 loss: 4.687e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 691/30000 [1:35:27<55:58:43,  6.88s/it, lr: 1.0e-04 loss: 4.687e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 691/30000 [1:35:34<55:58:43,  6.88s/it, lr: 1.0e-04 loss: 1.055e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 691/30000 [1:35:34<55:58:43,  6.88s/it, lr: 1.0e-04 loss: 1.055e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 692/30000 [1:35:34<55:31:25,  6.82s/it, lr: 1.0e-04 loss: 1.055e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 692/30000 [1:35:34<55:31:25,  6.82s/it, lr: 1.0e-04 loss: 1.055e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 692/30000 [1:35:40<55:31:25,  6.82s/it, lr: 1.0e-04 loss: 1.567e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 692/30000 [1:35:40<55:31:25,  6.82s/it, lr: 1.0e-04 loss: 1.567e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 693/30000 [1:35:40<55:03:05,  6.76s/it, lr: 1.0e-04 loss: 1.567e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 693/30000 [1:35:40<55:03:05,  6.76s/it, lr: 1.0e-04 loss: 1.567e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 693/30000 [1:35:47<55:03:05,  6.76s/it, lr: 1.0e-04 loss: 6.041e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 693/30000 [1:35:47<55:03:05,  6.76s/it, lr: 1.0e-04 loss: 6.041e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 694/30000 [1:35:47<54:44:26,  6.72s/it, lr: 1.0e-04 loss: 6.041e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 694/30000 [1:35:47<54:44:26,  6.72s/it, lr: 1.0e-04 loss: 6.041e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 694/30000 [1:35:56<54:44:26,  6.72s/it, lr: 1.0e-04 loss: 2.603e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 694/30000 [1:35:56<54:44:26,  6.72s/it, lr: 1.0e-04 loss: 2.603e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 695/30000 [1:35:56<60:39:29,  7.45s/it, lr: 1.0e-04 loss: 2.603e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 695/30000 [1:35:56<60:39:29,  7.45s/it, lr: 1.0e-04 loss: 2.603e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 695/30000 [1:36:06<60:39:29,  7.45s/it, lr: 1.0e-04 loss: 6.391e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 695/30000 [1:36:06<60:39:29,  7.45s/it, lr: 1.0e-04 loss: 6.391e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 696/30000 [1:36:06<65:10:16,  8.01s/it, lr: 1.0e-04 loss: 6.391e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 696/30000 [1:36:06<65:10:16,  8.01s/it, lr: 1.0e-04 loss: 6.391e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 696/30000 [1:36:13<65:10:16,  8.01s/it, lr: 1.0e-04 loss: 5.440e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 696/30000 [1:36:13<65:10:16,  8.01s/it, lr: 1.0e-04 loss: 5.440e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 697/30000 [1:36:13<63:07:20,  7.75s/it, lr: 1.0e-04 loss: 5.440e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 697/30000 [1:36:13<63:07:20,  7.75s/it, lr: 1.0e-04 loss: 5.440e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 697/30000 [1:36:19<63:07:20,  7.75s/it, lr: 1.0e-04 loss: 1.081e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 697/30000 [1:36:19<63:07:20,  7.75s/it, lr: 1.0e-04 loss: 1.081e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 698/30000 [1:36:19<60:31:58,  7.44s/it, lr: 1.0e-04 loss: 1.081e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 698/30000 [1:36:19<60:31:58,  7.44s/it, lr: 1.0e-04 loss: 1.081e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 698/30000 [1:36:29<60:31:58,  7.44s/it, lr: 1.0e-04 loss: 3.883e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 698/30000 [1:36:29<60:31:58,  7.44s/it, lr: 1.0e-04 loss: 3.883e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 699/30000 [1:36:29<65:37:10,  8.06s/it, lr: 1.0e-04 loss: 3.883e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 699/30000 [1:36:29<65:37:10,  8.06s/it, lr: 1.0e-04 loss: 3.883e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 699/30000 [1:36:38<65:37:10,  8.06s/it, lr: 1.0e-04 loss: 5.699e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 699/30000 [1:36:38<65:37:10,  8.06s/it, lr: 1.0e-04 loss: 5.699e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 700/30000 [1:36:45<65:37:02,  8.06s/it, lr: 1.0e-04 loss: 2.068e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 700/30000 [1:36:45<65:37:02,  8.06s/it, lr: 1.0e-04 loss: 2.068e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 701/30000 [1:36:45<48:12:44,  5.92s/it, lr: 1.0e-04 loss: 2.068e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 701/30000 [1:36:45<48:12:44,  5.92s/it, lr: 1.0e-04 loss: 2.068e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 701/30000 [1:36:55<48:12:44,  5.92s/it, lr: 1.0e-04 loss: 8.842e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 701/30000 [1:36:55<48:12:44,  5.92s/it, lr: 1.0e-04 loss: 8.842e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 702/30000 [1:36:55<55:07:36,  6.77s/it, lr: 1.0e-04 loss: 8.842e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 702/30000 [1:36:55<55:07:36,  6.77s/it, lr: 1.0e-04 loss: 8.842e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 702/30000 [1:37:01<55:07:36,  6.77s/it, lr: 1.0e-04 loss: 7.594e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 702/30000 [1:37:01<55:07:36,  6.77s/it, lr: 1.0e-04 loss: 7.594e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 703/30000 [1:37:01<54:56:28,  6.75s/it, lr: 1.0e-04 loss: 7.594e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 703/30000 [1:37:01<54:56:28,  6.75s/it, lr: 1.0e-04 loss: 7.594e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 703/30000 [1:37:11<54:56:28,  6.75s/it, lr: 1.0e-04 loss: 7.804e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 703/30000 [1:37:11<54:56:28,  6.75s/it, lr: 1.0e-04 loss: 7.804e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 704/30000 [1:37:11<60:40:35,  7.46s/it, lr: 1.0e-04 loss: 7.804e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 704/30000 [1:37:11<60:40:35,  7.46s/it, lr: 1.0e-04 loss: 7.804e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 704/30000 [1:37:18<60:40:35,  7.46s/it, lr: 1.0e-04 loss: 7.319e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 704/30000 [1:37:18<60:40:35,  7.46s/it, lr: 1.0e-04 loss: 7.319e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 705/30000 [1:37:18<60:22:51,  7.42s/it, lr: 1.0e-04 loss: 7.319e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 705/30000 [1:37:18<60:22:51,  7.42s/it, lr: 1.0e-04 loss: 7.319e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 705/30000 [1:37:25<60:22:51,  7.42s/it, lr: 1.0e-04 loss: 1.066e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 705/30000 [1:37:25<60:22:51,  7.42s/it, lr: 1.0e-04 loss: 1.066e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 706/30000 [1:37:25<58:37:41,  7.20s/it, lr: 1.0e-04 loss: 1.066e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 706/30000 [1:37:25<58:37:41,  7.20s/it, lr: 1.0e-04 loss: 1.066e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 706/30000 [1:37:34<58:37:41,  7.20s/it, lr: 1.0e-04 loss: 3.599e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 706/30000 [1:37:34<58:37:41,  7.20s/it, lr: 1.0e-04 loss: 3.599e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 707/30000 [1:37:34<63:43:20,  7.83s/it, lr: 1.0e-04 loss: 3.599e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 707/30000 [1:37:34<63:43:20,  7.83s/it, lr: 1.0e-04 loss: 3.599e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 707/30000 [1:37:41<63:43:20,  7.83s/it, lr: 1.0e-04 loss: 7.967e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 707/30000 [1:37:41<63:43:20,  7.83s/it, lr: 1.0e-04 loss: 7.967e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 708/30000 [1:37:41<60:53:37,  7.48s/it, lr: 1.0e-04 loss: 7.967e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 708/30000 [1:37:41<60:53:37,  7.48s/it, lr: 1.0e-04 loss: 7.967e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 708/30000 [1:37:50<60:53:37,  7.48s/it, lr: 1.0e-04 loss: 3.485e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 708/30000 [1:37:50<60:53:37,  7.48s/it, lr: 1.0e-04 loss: 3.485e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 709/30000 [1:37:50<65:23:48,  8.04s/it, lr: 1.0e-04 loss: 3.485e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 709/30000 [1:37:50<65:23:48,  8.04s/it, lr: 1.0e-04 loss: 3.485e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 709/30000 [1:37:57<65:23:48,  8.04s/it, lr: 1.0e-04 loss: 9.987e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 709/30000 [1:37:57<65:23:48,  8.04s/it, lr: 1.0e-04 loss: 9.987e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 710/30000 [1:38:06<65:23:40,  8.04s/it, lr: 1.0e-04 loss: 1.649e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 710/30000 [1:38:06<65:23:40,  8.04s/it, lr: 1.0e-04 loss: 1.649e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 711/30000 [1:38:06<53:07:47,  6.53s/it, lr: 1.0e-04 loss: 1.649e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 711/30000 [1:38:06<53:07:47,  6.53s/it, lr: 1.0e-04 loss: 1.649e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 711/30000 [1:38:13<53:07:47,  6.53s/it, lr: 1.0e-04 loss: 8.204e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 711/30000 [1:38:13<53:07:47,  6.53s/it, lr: 1.0e-04 loss: 8.204e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 712/30000 [1:38:13<53:51:37,  6.62s/it, lr: 1.0e-04 loss: 8.204e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 712/30000 [1:38:13<53:51:37,  6.62s/it, lr: 1.0e-04 loss: 8.204e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 712/30000 [1:38:20<53:51:37,  6.62s/it, lr: 1.0e-04 loss: 8.280e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 712/30000 [1:38:20<53:51:37,  6.62s/it, lr: 1.0e-04 loss: 8.280e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 713/30000 [1:38:20<54:01:04,  6.64s/it, lr: 1.0e-04 loss: 8.280e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 713/30000 [1:38:20<54:01:04,  6.64s/it, lr: 1.0e-04 loss: 8.280e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 713/30000 [1:38:29<54:01:04,  6.64s/it, lr: 1.0e-04 loss: 7.785e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 713/30000 [1:38:29<54:01:04,  6.64s/it, lr: 1.0e-04 loss: 7.785e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 714/30000 [1:38:29<60:14:15,  7.40s/it, lr: 1.0e-04 loss: 7.785e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 714/30000 [1:38:29<60:14:15,  7.40s/it, lr: 1.0e-04 loss: 7.785e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 714/30000 [1:38:39<60:14:15,  7.40s/it, lr: 1.0e-04 loss: 4.438e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 714/30000 [1:38:39<60:14:15,  7.40s/it, lr: 1.0e-04 loss: 4.438e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 715/30000 [1:38:39<64:53:18,  7.98s/it, lr: 1.0e-04 loss: 4.438e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 715/30000 [1:38:39<64:53:18,  7.98s/it, lr: 1.0e-04 loss: 4.438e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 715/30000 [1:38:48<64:53:18,  7.98s/it, lr: 1.0e-04 loss: 7.515e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 715/30000 [1:38:48<64:53:18,  7.98s/it, lr: 1.0e-04 loss: 7.515e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 716/30000 [1:38:48<68:23:30,  8.41s/it, lr: 1.0e-04 loss: 7.515e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 716/30000 [1:38:48<68:23:30,  8.41s/it, lr: 1.0e-04 loss: 7.515e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 716/30000 [1:38:58<68:23:30,  8.41s/it, lr: 1.0e-04 loss: 4.130e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 716/30000 [1:38:58<68:23:30,  8.41s/it, lr: 1.0e-04 loss: 4.130e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 717/30000 [1:38:58<70:55:51,  8.72s/it, lr: 1.0e-04 loss: 4.130e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 717/30000 [1:38:58<70:55:51,  8.72s/it, lr: 1.0e-04 loss: 4.130e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 717/30000 [1:39:07<70:55:51,  8.72s/it, lr: 1.0e-04 loss: 9.354e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 717/30000 [1:39:07<70:55:51,  8.72s/it, lr: 1.0e-04 loss: 9.354e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 718/30000 [1:39:07<73:19:03,  9.01s/it, lr: 1.0e-04 loss: 9.354e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 718/30000 [1:39:07<73:19:03,  9.01s/it, lr: 1.0e-04 loss: 9.354e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 718/30000 [1:39:17<73:19:03,  9.01s/it, lr: 1.0e-04 loss: 6.589e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 718/30000 [1:39:17<73:19:03,  9.01s/it, lr: 1.0e-04 loss: 6.589e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 719/30000 [1:39:17<74:49:50,  9.20s/it, lr: 1.0e-04 loss: 6.589e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 719/30000 [1:39:17<74:49:50,  9.20s/it, lr: 1.0e-04 loss: 6.589e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 719/30000 [1:39:26<74:49:50,  9.20s/it, lr: 1.0e-04 loss: 6.361e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 719/30000 [1:39:26<74:49:50,  9.20s/it, lr: 1.0e-04 loss: 6.361e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 720/30000 [1:39:33<74:49:41,  9.20s/it, lr: 1.0e-04 loss: 5.401e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 720/30000 [1:39:33<74:49:41,  9.20s/it, lr: 1.0e-04 loss: 5.401e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 721/30000 [1:39:33<52:55:24,  6.51s/it, lr: 1.0e-04 loss: 5.401e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 721/30000 [1:39:33<52:55:24,  6.51s/it, lr: 1.0e-04 loss: 5.401e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 721/30000 [1:39:42<52:55:24,  6.51s/it, lr: 1.0e-04 loss: 5.332e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 721/30000 [1:39:42<52:55:24,  6.51s/it, lr: 1.0e-04 loss: 5.332e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 722/30000 [1:39:42<58:41:15,  7.22s/it, lr: 1.0e-04 loss: 5.332e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 722/30000 [1:39:42<58:41:15,  7.22s/it, lr: 1.0e-04 loss: 5.332e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 722/30000 [1:39:52<58:41:15,  7.22s/it, lr: 1.0e-04 loss: 2.955e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 722/30000 [1:39:52<58:41:15,  7.22s/it, lr: 1.0e-04 loss: 2.955e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 723/30000 [1:39:52<63:38:21,  7.83s/it, lr: 1.0e-04 loss: 2.955e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 723/30000 [1:39:52<63:38:21,  7.83s/it, lr: 1.0e-04 loss: 2.955e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 723/30000 [1:39:59<63:38:21,  7.83s/it, lr: 1.0e-04 loss: 7.946e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 723/30000 [1:39:59<63:38:21,  7.83s/it, lr: 1.0e-04 loss: 7.946e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 724/30000 [1:39:59<60:57:27,  7.50s/it, lr: 1.0e-04 loss: 7.946e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 724/30000 [1:39:59<60:57:27,  7.50s/it, lr: 1.0e-04 loss: 7.946e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 724/30000 [1:40:08<60:57:27,  7.50s/it, lr: 1.0e-04 loss: 1.032e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 724/30000 [1:40:08<60:57:27,  7.50s/it, lr: 1.0e-04 loss: 1.032e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 725/30000 [1:40:08<65:26:02,  8.05s/it, lr: 1.0e-04 loss: 1.032e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 725/30000 [1:40:08<65:26:02,  8.05s/it, lr: 1.0e-04 loss: 1.032e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 725/30000 [1:40:18<65:26:02,  8.05s/it, lr: 1.0e-04 loss: 8.515e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 725/30000 [1:40:18<65:26:02,  8.05s/it, lr: 1.0e-04 loss: 8.515e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 726/30000 [1:40:18<68:44:26,  8.45s/it, lr: 1.0e-04 loss: 8.515e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 726/30000 [1:40:18<68:44:26,  8.45s/it, lr: 1.0e-04 loss: 8.515e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 726/30000 [1:40:24<68:44:26,  8.45s/it, lr: 1.0e-04 loss: 4.881e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 726/30000 [1:40:24<68:44:26,  8.45s/it, lr: 1.0e-04 loss: 4.881e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 727/30000 [1:40:24<64:58:23,  7.99s/it, lr: 1.0e-04 loss: 4.881e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 727/30000 [1:40:24<64:58:23,  7.99s/it, lr: 1.0e-04 loss: 4.881e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 727/30000 [1:40:31<64:58:23,  7.99s/it, lr: 1.0e-04 loss: 1.040e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 727/30000 [1:40:31<64:58:23,  7.99s/it, lr: 1.0e-04 loss: 1.040e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 728/30000 [1:40:31<61:41:25,  7.59s/it, lr: 1.0e-04 loss: 1.040e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 728/30000 [1:40:31<61:41:25,  7.59s/it, lr: 1.0e-04 loss: 1.040e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 728/30000 [1:40:41<61:41:25,  7.59s/it, lr: 1.0e-04 loss: 1.113e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 728/30000 [1:40:41<61:41:25,  7.59s/it, lr: 1.0e-04 loss: 1.113e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 729/30000 [1:40:41<66:15:38,  8.15s/it, lr: 1.0e-04 loss: 1.113e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 729/30000 [1:40:41<66:15:38,  8.15s/it, lr: 1.0e-04 loss: 1.113e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 729/30000 [1:40:47<66:15:38,  8.15s/it, lr: 1.0e-04 loss: 9.155e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 729/30000 [1:40:47<66:15:38,  8.15s/it, lr: 1.0e-04 loss: 9.155e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 730/30000 [1:40:57<66:15:30,  8.15s/it, lr: 1.0e-04 loss: 8.441e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 730/30000 [1:40:57<66:15:30,  8.15s/it, lr: 1.0e-04 loss: 8.441e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 731/30000 [1:40:57<53:17:35,  6.55s/it, lr: 1.0e-04 loss: 8.441e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 731/30000 [1:40:57<53:17:35,  6.55s/it, lr: 1.0e-04 loss: 8.441e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 731/30000 [1:41:06<53:17:35,  6.55s/it, lr: 1.0e-04 loss: 6.555e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 731/30000 [1:41:06<53:17:35,  6.55s/it, lr: 1.0e-04 loss: 6.555e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 732/30000 [1:41:06<58:53:04,  7.24s/it, lr: 1.0e-04 loss: 6.555e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 732/30000 [1:41:06<58:53:04,  7.24s/it, lr: 1.0e-04 loss: 6.555e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 732/30000 [1:41:15<58:53:04,  7.24s/it, lr: 1.0e-04 loss: 2.638e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 732/30000 [1:41:15<58:53:04,  7.24s/it, lr: 1.0e-04 loss: 2.638e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 733/30000 [1:41:15<63:35:16,  7.82s/it, lr: 1.0e-04 loss: 2.638e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 733/30000 [1:41:15<63:35:16,  7.82s/it, lr: 1.0e-04 loss: 2.638e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 733/30000 [1:41:23<63:35:16,  7.82s/it, lr: 1.0e-04 loss: 1.182e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 733/30000 [1:41:23<63:35:16,  7.82s/it, lr: 1.0e-04 loss: 1.182e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 734/30000 [1:41:23<62:21:01,  7.67s/it, lr: 1.0e-04 loss: 1.182e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 734/30000 [1:41:23<62:21:01,  7.67s/it, lr: 1.0e-04 loss: 1.182e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 734/30000 [1:41:32<62:21:01,  7.67s/it, lr: 1.0e-04 loss: 7.500e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 734/30000 [1:41:32<62:21:01,  7.67s/it, lr: 1.0e-04 loss: 7.500e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 735/30000 [1:41:32<66:07:54,  8.14s/it, lr: 1.0e-04 loss: 7.500e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 735/30000 [1:41:32<66:07:54,  8.14s/it, lr: 1.0e-04 loss: 7.500e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 735/30000 [1:41:41<66:07:54,  8.14s/it, lr: 1.0e-04 loss: 1.197e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 735/30000 [1:41:41<66:07:54,  8.14s/it, lr: 1.0e-04 loss: 1.197e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 736/30000 [1:41:41<68:54:42,  8.48s/it, lr: 1.0e-04 loss: 1.197e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 736/30000 [1:41:41<68:54:42,  8.48s/it, lr: 1.0e-04 loss: 1.197e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 736/30000 [1:41:48<68:54:42,  8.48s/it, lr: 1.0e-04 loss: 4.748e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 736/30000 [1:41:48<68:54:42,  8.48s/it, lr: 1.0e-04 loss: 4.748e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 737/30000 [1:41:48<64:47:59,  7.97s/it, lr: 1.0e-04 loss: 4.748e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 737/30000 [1:41:48<64:47:59,  7.97s/it, lr: 1.0e-04 loss: 4.748e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 737/30000 [1:41:55<64:47:59,  7.97s/it, lr: 1.0e-04 loss: 1.094e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 737/30000 [1:41:55<64:47:59,  7.97s/it, lr: 1.0e-04 loss: 1.094e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 738/30000 [1:41:55<61:24:22,  7.55s/it, lr: 1.0e-04 loss: 1.094e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 738/30000 [1:41:55<61:24:22,  7.55s/it, lr: 1.0e-04 loss: 1.094e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 738/30000 [1:42:01<61:24:22,  7.55s/it, lr: 1.0e-04 loss: 5.348e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 738/30000 [1:42:01<61:24:22,  7.55s/it, lr: 1.0e-04 loss: 5.348e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 739/30000 [1:42:01<59:09:45,  7.28s/it, lr: 1.0e-04 loss: 5.348e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 739/30000 [1:42:01<59:09:45,  7.28s/it, lr: 1.0e-04 loss: 5.348e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 739/30000 [1:42:11<59:09:45,  7.28s/it, lr: 1.0e-04 loss: 5.612e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 739/30000 [1:42:11<59:09:45,  7.28s/it, lr: 1.0e-04 loss: 5.612e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 740/30000 [1:42:20<59:09:38,  7.28s/it, lr: 1.0e-04 loss: 4.923e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 740/30000 [1:42:20<59:09:38,  7.28s/it, lr: 1.0e-04 loss: 4.923e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 741/30000 [1:42:20<49:38:37,  6.11s/it, lr: 1.0e-04 loss: 4.923e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 741/30000 [1:42:20<49:38:37,  6.11s/it, lr: 1.0e-04 loss: 4.923e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 741/30000 [1:42:27<49:38:37,  6.11s/it, lr: 1.0e-04 loss: 8.656e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 741/30000 [1:42:27<49:38:37,  6.11s/it, lr: 1.0e-04 loss: 8.656e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 742/30000 [1:42:27<51:17:59,  6.31s/it, lr: 1.0e-04 loss: 8.656e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 742/30000 [1:42:27<51:17:59,  6.31s/it, lr: 1.0e-04 loss: 8.656e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 742/30000 [1:42:34<51:17:59,  6.31s/it, lr: 1.0e-04 loss: 8.570e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 742/30000 [1:42:34<51:17:59,  6.31s/it, lr: 1.0e-04 loss: 8.570e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 743/30000 [1:42:34<51:54:29,  6.39s/it, lr: 1.0e-04 loss: 8.570e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 743/30000 [1:42:34<51:54:29,  6.39s/it, lr: 1.0e-04 loss: 8.570e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 743/30000 [1:42:43<51:54:29,  6.39s/it, lr: 1.0e-04 loss: 7.894e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 743/30000 [1:42:43<51:54:29,  6.39s/it, lr: 1.0e-04 loss: 7.894e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 744/30000 [1:42:43<58:48:31,  7.24s/it, lr: 1.0e-04 loss: 7.894e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 744/30000 [1:42:43<58:48:31,  7.24s/it, lr: 1.0e-04 loss: 7.894e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 744/30000 [1:42:50<58:48:31,  7.24s/it, lr: 1.0e-04 loss: 7.196e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 744/30000 [1:42:50<58:48:31,  7.24s/it, lr: 1.0e-04 loss: 7.196e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 745/30000 [1:42:50<57:26:30,  7.07s/it, lr: 1.0e-04 loss: 7.196e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 745/30000 [1:42:50<57:26:30,  7.07s/it, lr: 1.0e-04 loss: 7.196e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 745/30000 [1:42:57<57:26:30,  7.07s/it, lr: 1.0e-04 loss: 6.866e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 745/30000 [1:42:57<57:26:30,  7.07s/it, lr: 1.0e-04 loss: 6.866e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 746/30000 [1:42:57<56:16:28,  6.93s/it, lr: 1.0e-04 loss: 6.866e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 746/30000 [1:42:57<56:16:28,  6.93s/it, lr: 1.0e-04 loss: 6.866e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 746/30000 [1:43:03<56:16:28,  6.93s/it, lr: 1.0e-04 loss: 4.978e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 746/30000 [1:43:03<56:16:28,  6.93s/it, lr: 1.0e-04 loss: 4.978e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 747/30000 [1:43:03<55:18:54,  6.81s/it, lr: 1.0e-04 loss: 4.978e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 747/30000 [1:43:03<55:18:54,  6.81s/it, lr: 1.0e-04 loss: 4.978e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 747/30000 [1:43:10<55:18:54,  6.81s/it, lr: 1.0e-04 loss: 1.057e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 747/30000 [1:43:10<55:18:54,  6.81s/it, lr: 1.0e-04 loss: 1.057e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 748/30000 [1:43:10<54:39:55,  6.73s/it, lr: 1.0e-04 loss: 1.057e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 748/30000 [1:43:10<54:39:55,  6.73s/it, lr: 1.0e-04 loss: 1.057e-01]
Xiang_wan_2_2_lora:   2%|9                                       | 748/30000 [1:43:17<54:39:55,  6.73s/it, lr: 1.0e-04 loss: 8.530e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 748/30000 [1:43:17<54:39:55,  6.73s/it, lr: 1.0e-04 loss: 8.530e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 749/30000 [1:43:17<56:13:04,  6.92s/it, lr: 1.0e-04 loss: 8.530e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 749/30000 [1:43:17<56:13:04,  6.92s/it, lr: 1.0e-04 loss: 8.530e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 749/30000 [1:43:26<56:13:04,  6.92s/it, lr: 1.0e-04 loss: 8.473e-02]
Xiang_wan_2_2_lora:   2%|9                                       | 749/30000 [1:43:26<56:13:04,  6.92s/it, lr: 1.0e-04 loss: 8.473e-02]
+Saving at step 750
+Saved checkpoint to /home/featurize/ai-toolkit/output/Xiang_wan_2_2_lora/Xiang_wan_2_2_lora_000000750.safetensors
+Saved optimizer to /home/featurize/ai-toolkit/output/Xiang_wan_2_2_lora/optimizer.pt
+
Xiang_wan_2_2_lora:   2%|#                                       | 750/30000 [1:43:36<56:12:57,  6.92s/it, lr: 1.0e-04 loss: 5.210e-02]
Xiang_wan_2_2_lora:   2%|#                                       | 750/30000 [1:43:36<56:12:57,  6.92s/it, lr: 1.0e-04 loss: 5.210e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 751/30000 [1:43:36<48:24:07,  5.96s/it, lr: 1.0e-04 loss: 5.210e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 751/30000 [1:43:36<48:24:07,  5.96s/it, lr: 1.0e-04 loss: 5.210e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 751/30000 [1:43:46<48:24:07,  5.96s/it, lr: 1.0e-04 loss: 5.737e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 751/30000 [1:43:46<48:24:07,  5.96s/it, lr: 1.0e-04 loss: 5.737e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 752/30000 [1:43:46<55:20:46,  6.81s/it, lr: 1.0e-04 loss: 5.737e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 752/30000 [1:43:46<55:20:46,  6.81s/it, lr: 1.0e-04 loss: 5.737e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 752/30000 [1:43:55<55:20:46,  6.81s/it, lr: 1.0e-04 loss: 7.231e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 752/30000 [1:43:55<55:20:46,  6.81s/it, lr: 1.0e-04 loss: 7.231e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 753/30000 [1:43:55<60:48:11,  7.48s/it, lr: 1.0e-04 loss: 7.231e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 753/30000 [1:43:55<60:48:11,  7.48s/it, lr: 1.0e-04 loss: 7.231e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 753/30000 [1:44:02<60:48:11,  7.48s/it, lr: 1.0e-04 loss: 1.196e-01]
Xiang_wan_2_2_lora:   3%|#                                       | 753/30000 [1:44:02<60:48:11,  7.48s/it, lr: 1.0e-04 loss: 1.196e-01]
Xiang_wan_2_2_lora:   3%|#                                       | 754/30000 [1:44:02<58:50:24,  7.24s/it, lr: 1.0e-04 loss: 1.196e-01]
Xiang_wan_2_2_lora:   3%|#                                       | 754/30000 [1:44:02<58:50:24,  7.24s/it, lr: 1.0e-04 loss: 1.196e-01]
Xiang_wan_2_2_lora:   3%|#                                       | 754/30000 [1:44:08<58:50:24,  7.24s/it, lr: 1.0e-04 loss: 7.041e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 754/30000 [1:44:08<58:50:24,  7.24s/it, lr: 1.0e-04 loss: 7.041e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 755/30000 [1:44:08<57:19:54,  7.06s/it, lr: 1.0e-04 loss: 7.041e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 755/30000 [1:44:08<57:19:54,  7.06s/it, lr: 1.0e-04 loss: 7.041e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 755/30000 [1:44:15<57:19:54,  7.06s/it, lr: 1.0e-04 loss: 1.037e-01]
Xiang_wan_2_2_lora:   3%|#                                       | 755/30000 [1:44:15<57:19:54,  7.06s/it, lr: 1.0e-04 loss: 1.037e-01]
Xiang_wan_2_2_lora:   3%|#                                       | 756/30000 [1:44:15<56:06:45,  6.91s/it, lr: 1.0e-04 loss: 1.037e-01]
Xiang_wan_2_2_lora:   3%|#                                       | 756/30000 [1:44:15<56:06:45,  6.91s/it, lr: 1.0e-04 loss: 1.037e-01]
Xiang_wan_2_2_lora:   3%|#                                       | 756/30000 [1:44:21<56:06:45,  6.91s/it, lr: 1.0e-04 loss: 9.400e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 756/30000 [1:44:21<56:06:45,  6.91s/it, lr: 1.0e-04 loss: 9.400e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 757/30000 [1:44:21<55:35:04,  6.84s/it, lr: 1.0e-04 loss: 9.400e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 757/30000 [1:44:21<55:35:04,  6.84s/it, lr: 1.0e-04 loss: 9.400e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 757/30000 [1:44:29<55:35:04,  6.84s/it, lr: 1.0e-04 loss: 9.650e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 757/30000 [1:44:29<55:35:04,  6.84s/it, lr: 1.0e-04 loss: 9.650e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 758/30000 [1:44:29<57:14:00,  7.05s/it, lr: 1.0e-04 loss: 9.650e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 758/30000 [1:44:29<57:14:00,  7.05s/it, lr: 1.0e-04 loss: 9.650e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 758/30000 [1:44:38<57:14:00,  7.05s/it, lr: 1.0e-04 loss: 4.441e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 758/30000 [1:44:38<57:14:00,  7.05s/it, lr: 1.0e-04 loss: 4.441e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 759/30000 [1:44:38<63:22:38,  7.80s/it, lr: 1.0e-04 loss: 4.441e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 759/30000 [1:44:38<63:22:38,  7.80s/it, lr: 1.0e-04 loss: 4.441e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 759/30000 [1:44:48<63:22:38,  7.80s/it, lr: 1.0e-04 loss: 6.845e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 759/30000 [1:44:48<63:22:38,  7.80s/it, lr: 1.0e-04 loss: 6.845e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 760/30000 [1:44:57<63:22:30,  7.80s/it, lr: 1.0e-04 loss: 3.696e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 760/30000 [1:44:57<63:22:30,  7.80s/it, lr: 1.0e-04 loss: 3.696e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 761/30000 [1:44:57<51:53:38,  6.39s/it, lr: 1.0e-04 loss: 3.696e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 761/30000 [1:44:57<51:53:38,  6.39s/it, lr: 1.0e-04 loss: 3.696e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 761/30000 [1:45:04<51:53:38,  6.39s/it, lr: 1.0e-04 loss: 7.555e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 761/30000 [1:45:04<51:53:38,  6.39s/it, lr: 1.0e-04 loss: 7.555e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 762/30000 [1:45:04<52:24:22,  6.45s/it, lr: 1.0e-04 loss: 7.555e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 762/30000 [1:45:04<52:24:22,  6.45s/it, lr: 1.0e-04 loss: 7.555e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 762/30000 [1:45:11<52:24:22,  6.45s/it, lr: 1.0e-04 loss: 9.087e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 762/30000 [1:45:11<52:24:22,  6.45s/it, lr: 1.0e-04 loss: 9.087e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 763/30000 [1:45:11<53:05:04,  6.54s/it, lr: 1.0e-04 loss: 9.087e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 763/30000 [1:45:11<53:05:04,  6.54s/it, lr: 1.0e-04 loss: 9.087e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 763/30000 [1:45:20<53:05:04,  6.54s/it, lr: 1.0e-04 loss: 8.524e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 763/30000 [1:45:20<53:05:04,  6.54s/it, lr: 1.0e-04 loss: 8.524e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 764/30000 [1:45:20<59:23:09,  7.31s/it, lr: 1.0e-04 loss: 8.524e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 764/30000 [1:45:20<59:23:09,  7.31s/it, lr: 1.0e-04 loss: 8.524e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 764/30000 [1:45:27<59:23:09,  7.31s/it, lr: 1.0e-04 loss: 8.059e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 764/30000 [1:45:27<59:23:09,  7.31s/it, lr: 1.0e-04 loss: 8.059e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 765/30000 [1:45:27<59:27:37,  7.32s/it, lr: 1.0e-04 loss: 8.059e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 765/30000 [1:45:27<59:27:37,  7.32s/it, lr: 1.0e-04 loss: 8.059e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 765/30000 [1:45:37<59:27:37,  7.32s/it, lr: 1.0e-04 loss: 9.136e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 765/30000 [1:45:37<59:27:37,  7.32s/it, lr: 1.0e-04 loss: 9.136e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 766/30000 [1:45:37<64:17:22,  7.92s/it, lr: 1.0e-04 loss: 9.136e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 766/30000 [1:45:37<64:17:22,  7.92s/it, lr: 1.0e-04 loss: 9.136e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 766/30000 [1:45:44<64:17:22,  7.92s/it, lr: 1.0e-04 loss: 6.240e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 766/30000 [1:45:44<64:17:22,  7.92s/it, lr: 1.0e-04 loss: 6.240e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 767/30000 [1:45:44<61:24:23,  7.56s/it, lr: 1.0e-04 loss: 6.240e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 767/30000 [1:45:44<61:24:23,  7.56s/it, lr: 1.0e-04 loss: 6.240e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 767/30000 [1:45:50<61:24:23,  7.56s/it, lr: 1.0e-04 loss: 9.629e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 767/30000 [1:45:50<61:24:23,  7.56s/it, lr: 1.0e-04 loss: 9.629e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 768/30000 [1:45:50<59:24:37,  7.32s/it, lr: 1.0e-04 loss: 9.629e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 768/30000 [1:45:50<59:24:37,  7.32s/it, lr: 1.0e-04 loss: 9.629e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 768/30000 [1:45:57<59:24:37,  7.32s/it, lr: 1.0e-04 loss: 9.388e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 768/30000 [1:45:57<59:24:37,  7.32s/it, lr: 1.0e-04 loss: 9.388e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 769/30000 [1:45:57<57:45:52,  7.11s/it, lr: 1.0e-04 loss: 9.388e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 769/30000 [1:45:57<57:45:52,  7.11s/it, lr: 1.0e-04 loss: 9.388e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 769/30000 [1:46:03<57:45:52,  7.11s/it, lr: 1.0e-04 loss: 9.206e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 769/30000 [1:46:03<57:45:52,  7.11s/it, lr: 1.0e-04 loss: 9.206e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 770/30000 [1:46:10<57:45:45,  7.11s/it, lr: 1.0e-04 loss: 9.282e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 770/30000 [1:46:10<57:45:45,  7.11s/it, lr: 1.0e-04 loss: 9.282e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 771/30000 [1:46:10<43:43:06,  5.38s/it, lr: 1.0e-04 loss: 9.282e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 771/30000 [1:46:10<43:43:06,  5.38s/it, lr: 1.0e-04 loss: 9.282e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 771/30000 [1:46:22<43:43:06,  5.38s/it, lr: 1.0e-04 loss: 9.689e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 771/30000 [1:46:22<43:43:06,  5.38s/it, lr: 1.0e-04 loss: 9.689e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 772/30000 [1:46:22<53:57:38,  6.65s/it, lr: 1.0e-04 loss: 9.689e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 772/30000 [1:46:22<53:57:38,  6.65s/it, lr: 1.0e-04 loss: 9.689e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 772/30000 [1:46:31<53:57:38,  6.65s/it, lr: 1.0e-04 loss: 7.355e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 772/30000 [1:46:31<53:57:38,  6.65s/it, lr: 1.0e-04 loss: 7.355e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 773/30000 [1:46:31<59:30:18,  7.33s/it, lr: 1.0e-04 loss: 7.355e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 773/30000 [1:46:31<59:30:18,  7.33s/it, lr: 1.0e-04 loss: 7.355e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 773/30000 [1:46:39<59:30:18,  7.33s/it, lr: 1.0e-04 loss: 8.008e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 773/30000 [1:46:39<59:30:18,  7.33s/it, lr: 1.0e-04 loss: 8.008e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 774/30000 [1:46:39<59:36:07,  7.34s/it, lr: 1.0e-04 loss: 8.008e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 774/30000 [1:46:39<59:36:07,  7.34s/it, lr: 1.0e-04 loss: 8.008e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 774/30000 [1:46:46<59:36:07,  7.34s/it, lr: 1.0e-04 loss: 1.212e-01]
Xiang_wan_2_2_lora:   3%|#                                       | 774/30000 [1:46:46<59:36:07,  7.34s/it, lr: 1.0e-04 loss: 1.212e-01]
Xiang_wan_2_2_lora:   3%|#                                       | 775/30000 [1:46:46<59:17:50,  7.30s/it, lr: 1.0e-04 loss: 1.212e-01]
Xiang_wan_2_2_lora:   3%|#                                       | 775/30000 [1:46:46<59:17:50,  7.30s/it, lr: 1.0e-04 loss: 1.212e-01]
Xiang_wan_2_2_lora:   3%|#                                       | 775/30000 [1:46:53<59:17:50,  7.30s/it, lr: 1.0e-04 loss: 1.195e-01]
Xiang_wan_2_2_lora:   3%|#                                       | 775/30000 [1:46:53<59:17:50,  7.30s/it, lr: 1.0e-04 loss: 1.195e-01]
Xiang_wan_2_2_lora:   3%|#                                       | 776/30000 [1:46:53<58:12:51,  7.17s/it, lr: 1.0e-04 loss: 1.195e-01]
Xiang_wan_2_2_lora:   3%|#                                       | 776/30000 [1:46:53<58:12:51,  7.17s/it, lr: 1.0e-04 loss: 1.195e-01]
Xiang_wan_2_2_lora:   3%|#                                       | 776/30000 [1:47:02<58:12:51,  7.17s/it, lr: 1.0e-04 loss: 4.168e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 776/30000 [1:47:02<58:12:51,  7.17s/it, lr: 1.0e-04 loss: 4.168e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 777/30000 [1:47:02<63:41:38,  7.85s/it, lr: 1.0e-04 loss: 4.168e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 777/30000 [1:47:02<63:41:38,  7.85s/it, lr: 1.0e-04 loss: 4.168e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 777/30000 [1:47:09<63:41:38,  7.85s/it, lr: 1.0e-04 loss: 9.045e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 777/30000 [1:47:09<63:41:38,  7.85s/it, lr: 1.0e-04 loss: 9.045e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 778/30000 [1:47:09<60:51:21,  7.50s/it, lr: 1.0e-04 loss: 9.045e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 778/30000 [1:47:09<60:51:21,  7.50s/it, lr: 1.0e-04 loss: 9.045e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 778/30000 [1:47:16<60:51:21,  7.50s/it, lr: 1.0e-04 loss: 7.556e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 778/30000 [1:47:16<60:51:21,  7.50s/it, lr: 1.0e-04 loss: 7.556e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 779/30000 [1:47:16<58:40:56,  7.23s/it, lr: 1.0e-04 loss: 7.556e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 779/30000 [1:47:16<58:40:56,  7.23s/it, lr: 1.0e-04 loss: 7.556e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 779/30000 [1:47:25<58:40:56,  7.23s/it, lr: 1.0e-04 loss: 4.744e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 779/30000 [1:47:25<58:40:56,  7.23s/it, lr: 1.0e-04 loss: 4.744e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 780/30000 [1:47:32<58:40:49,  7.23s/it, lr: 1.0e-04 loss: 1.149e-01]
Xiang_wan_2_2_lora:   3%|#                                       | 780/30000 [1:47:32<58:40:49,  7.23s/it, lr: 1.0e-04 loss: 1.149e-01]
Xiang_wan_2_2_lora:   3%|#                                       | 781/30000 [1:47:32<44:14:49,  5.45s/it, lr: 1.0e-04 loss: 1.149e-01]
Xiang_wan_2_2_lora:   3%|#                                       | 781/30000 [1:47:32<44:14:49,  5.45s/it, lr: 1.0e-04 loss: 1.149e-01]
Xiang_wan_2_2_lora:   3%|#                                       | 781/30000 [1:47:41<44:14:49,  5.45s/it, lr: 1.0e-04 loss: 1.160e-01]
Xiang_wan_2_2_lora:   3%|#                                       | 781/30000 [1:47:41<44:14:49,  5.45s/it, lr: 1.0e-04 loss: 1.160e-01]
Xiang_wan_2_2_lora:   3%|#                                       | 782/30000 [1:47:41<51:58:29,  6.40s/it, lr: 1.0e-04 loss: 1.160e-01]
Xiang_wan_2_2_lora:   3%|#                                       | 782/30000 [1:47:41<51:58:29,  6.40s/it, lr: 1.0e-04 loss: 1.160e-01]
Xiang_wan_2_2_lora:   3%|#                                       | 782/30000 [1:47:48<51:58:29,  6.40s/it, lr: 1.0e-04 loss: 6.557e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 782/30000 [1:47:48<51:58:29,  6.40s/it, lr: 1.0e-04 loss: 6.557e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 783/30000 [1:47:48<52:22:36,  6.45s/it, lr: 1.0e-04 loss: 6.557e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 783/30000 [1:47:48<52:22:36,  6.45s/it, lr: 1.0e-04 loss: 6.557e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 783/30000 [1:47:57<52:22:36,  6.45s/it, lr: 1.0e-04 loss: 9.899e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 783/30000 [1:47:57<52:22:36,  6.45s/it, lr: 1.0e-04 loss: 9.899e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 784/30000 [1:47:57<58:43:26,  7.24s/it, lr: 1.0e-04 loss: 9.899e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 784/30000 [1:47:57<58:43:26,  7.24s/it, lr: 1.0e-04 loss: 9.899e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 784/30000 [1:48:04<58:43:26,  7.24s/it, lr: 1.0e-04 loss: 5.525e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 784/30000 [1:48:04<58:43:26,  7.24s/it, lr: 1.0e-04 loss: 5.525e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 785/30000 [1:48:04<57:26:10,  7.08s/it, lr: 1.0e-04 loss: 5.525e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 785/30000 [1:48:04<57:26:10,  7.08s/it, lr: 1.0e-04 loss: 5.525e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 785/30000 [1:48:10<57:26:10,  7.08s/it, lr: 1.0e-04 loss: 7.065e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 785/30000 [1:48:10<57:26:10,  7.08s/it, lr: 1.0e-04 loss: 7.065e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 786/30000 [1:48:10<56:07:52,  6.92s/it, lr: 1.0e-04 loss: 7.065e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 786/30000 [1:48:10<56:07:52,  6.92s/it, lr: 1.0e-04 loss: 7.065e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 786/30000 [1:48:17<56:07:52,  6.92s/it, lr: 1.0e-04 loss: 3.878e-01]
Xiang_wan_2_2_lora:   3%|#                                       | 786/30000 [1:48:17<56:07:52,  6.92s/it, lr: 1.0e-04 loss: 3.878e-01]
Xiang_wan_2_2_lora:   3%|#                                       | 787/30000 [1:48:17<56:02:28,  6.91s/it, lr: 1.0e-04 loss: 3.878e-01]
Xiang_wan_2_2_lora:   3%|#                                       | 787/30000 [1:48:17<56:02:28,  6.91s/it, lr: 1.0e-04 loss: 3.878e-01]
Xiang_wan_2_2_lora:   3%|#                                       | 787/30000 [1:48:26<56:02:28,  6.91s/it, lr: 1.0e-04 loss: 6.986e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 787/30000 [1:48:26<56:02:28,  6.91s/it, lr: 1.0e-04 loss: 6.986e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 788/30000 [1:48:26<61:42:38,  7.61s/it, lr: 1.0e-04 loss: 6.986e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 788/30000 [1:48:26<61:42:38,  7.61s/it, lr: 1.0e-04 loss: 6.986e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 788/30000 [1:48:33<61:42:38,  7.61s/it, lr: 1.0e-04 loss: 5.686e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 788/30000 [1:48:33<61:42:38,  7.61s/it, lr: 1.0e-04 loss: 5.686e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 789/30000 [1:48:33<59:13:46,  7.30s/it, lr: 1.0e-04 loss: 5.686e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 789/30000 [1:48:33<59:13:46,  7.30s/it, lr: 1.0e-04 loss: 5.686e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 789/30000 [1:48:39<59:13:46,  7.30s/it, lr: 1.0e-04 loss: 8.589e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 789/30000 [1:48:39<59:13:46,  7.30s/it, lr: 1.0e-04 loss: 8.589e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 790/30000 [1:48:46<59:13:39,  7.30s/it, lr: 1.0e-04 loss: 4.531e-01]
Xiang_wan_2_2_lora:   3%|#                                       | 790/30000 [1:48:46<59:13:39,  7.30s/it, lr: 1.0e-04 loss: 4.531e-01]
Xiang_wan_2_2_lora:   3%|#                                       | 791/30000 [1:48:46<44:18:06,  5.46s/it, lr: 1.0e-04 loss: 4.531e-01]
Xiang_wan_2_2_lora:   3%|#                                       | 791/30000 [1:48:46<44:18:06,  5.46s/it, lr: 1.0e-04 loss: 4.531e-01]
Xiang_wan_2_2_lora:   3%|#                                       | 791/30000 [1:48:53<44:18:06,  5.46s/it, lr: 1.0e-04 loss: 9.088e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 791/30000 [1:48:53<44:18:06,  5.46s/it, lr: 1.0e-04 loss: 9.088e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 792/30000 [1:48:53<46:34:24,  5.74s/it, lr: 1.0e-04 loss: 9.088e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 792/30000 [1:48:53<46:34:24,  5.74s/it, lr: 1.0e-04 loss: 9.088e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 792/30000 [1:48:59<46:34:24,  5.74s/it, lr: 1.0e-04 loss: 2.000e-01]
Xiang_wan_2_2_lora:   3%|#                                       | 792/30000 [1:48:59<46:34:24,  5.74s/it, lr: 1.0e-04 loss: 2.000e-01]
Xiang_wan_2_2_lora:   3%|#                                       | 793/30000 [1:48:59<48:19:51,  5.96s/it, lr: 1.0e-04 loss: 2.000e-01]
Xiang_wan_2_2_lora:   3%|#                                       | 793/30000 [1:48:59<48:19:51,  5.96s/it, lr: 1.0e-04 loss: 2.000e-01]
Xiang_wan_2_2_lora:   3%|#                                       | 793/30000 [1:49:08<48:19:51,  5.96s/it, lr: 1.0e-04 loss: 5.142e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 793/30000 [1:49:08<48:19:51,  5.96s/it, lr: 1.0e-04 loss: 5.142e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 794/30000 [1:49:08<55:44:35,  6.87s/it, lr: 1.0e-04 loss: 5.142e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 794/30000 [1:49:08<55:44:35,  6.87s/it, lr: 1.0e-04 loss: 5.142e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 794/30000 [1:49:15<55:44:35,  6.87s/it, lr: 1.0e-04 loss: 1.004e-01]
Xiang_wan_2_2_lora:   3%|#                                       | 794/30000 [1:49:15<55:44:35,  6.87s/it, lr: 1.0e-04 loss: 1.004e-01]
Xiang_wan_2_2_lora:   3%|#                                       | 795/30000 [1:49:15<55:49:08,  6.88s/it, lr: 1.0e-04 loss: 1.004e-01]
Xiang_wan_2_2_lora:   3%|#                                       | 795/30000 [1:49:15<55:49:08,  6.88s/it, lr: 1.0e-04 loss: 1.004e-01]
Xiang_wan_2_2_lora:   3%|#                                       | 795/30000 [1:49:22<55:49:08,  6.88s/it, lr: 1.0e-04 loss: 1.445e-01]
Xiang_wan_2_2_lora:   3%|#                                       | 795/30000 [1:49:22<55:49:08,  6.88s/it, lr: 1.0e-04 loss: 1.445e-01]
Xiang_wan_2_2_lora:   3%|#                                       | 796/30000 [1:49:22<54:58:12,  6.78s/it, lr: 1.0e-04 loss: 1.445e-01]
Xiang_wan_2_2_lora:   3%|#                                       | 796/30000 [1:49:22<54:58:12,  6.78s/it, lr: 1.0e-04 loss: 1.445e-01]
Xiang_wan_2_2_lora:   3%|#                                       | 796/30000 [1:49:28<54:58:12,  6.78s/it, lr: 1.0e-04 loss: 7.982e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 796/30000 [1:49:28<54:58:12,  6.78s/it, lr: 1.0e-04 loss: 7.982e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 797/30000 [1:49:28<54:27:02,  6.71s/it, lr: 1.0e-04 loss: 7.982e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 797/30000 [1:49:28<54:27:02,  6.71s/it, lr: 1.0e-04 loss: 7.982e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 797/30000 [1:49:35<54:27:02,  6.71s/it, lr: 1.0e-04 loss: 9.233e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 797/30000 [1:49:35<54:27:02,  6.71s/it, lr: 1.0e-04 loss: 9.233e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 798/30000 [1:49:35<54:17:23,  6.69s/it, lr: 1.0e-04 loss: 9.233e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 798/30000 [1:49:35<54:17:23,  6.69s/it, lr: 1.0e-04 loss: 9.233e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 798/30000 [1:49:42<54:17:23,  6.69s/it, lr: 1.0e-04 loss: 6.837e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 798/30000 [1:49:42<54:17:23,  6.69s/it, lr: 1.0e-04 loss: 6.837e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 799/30000 [1:49:42<54:01:30,  6.66s/it, lr: 1.0e-04 loss: 6.837e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 799/30000 [1:49:42<54:01:30,  6.66s/it, lr: 1.0e-04 loss: 6.837e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 799/30000 [1:49:48<54:01:30,  6.66s/it, lr: 1.0e-04 loss: 1.100e-01]
Xiang_wan_2_2_lora:   3%|#                                       | 799/30000 [1:49:48<54:01:30,  6.66s/it, lr: 1.0e-04 loss: 1.100e-01]
Xiang_wan_2_2_lora:   3%|#                                       | 800/30000 [1:49:58<54:01:23,  6.66s/it, lr: 1.0e-04 loss: 7.531e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 800/30000 [1:49:58<54:01:23,  6.66s/it, lr: 1.0e-04 loss: 7.531e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 801/30000 [1:49:58<46:36:46,  5.75s/it, lr: 1.0e-04 loss: 7.531e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 801/30000 [1:49:58<46:36:46,  5.75s/it, lr: 1.0e-04 loss: 7.531e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 801/30000 [1:50:04<46:36:46,  5.75s/it, lr: 1.0e-04 loss: 9.189e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 801/30000 [1:50:04<46:36:46,  5.75s/it, lr: 1.0e-04 loss: 9.189e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 802/30000 [1:50:04<48:57:16,  6.04s/it, lr: 1.0e-04 loss: 9.189e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 802/30000 [1:50:04<48:57:16,  6.04s/it, lr: 1.0e-04 loss: 9.189e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 802/30000 [1:50:14<48:57:16,  6.04s/it, lr: 1.0e-04 loss: 4.075e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 802/30000 [1:50:14<48:57:16,  6.04s/it, lr: 1.0e-04 loss: 4.075e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 803/30000 [1:50:14<55:58:26,  6.90s/it, lr: 1.0e-04 loss: 4.075e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 803/30000 [1:50:14<55:58:26,  6.90s/it, lr: 1.0e-04 loss: 4.075e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 803/30000 [1:50:20<55:58:26,  6.90s/it, lr: 1.0e-04 loss: 7.243e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 803/30000 [1:50:20<55:58:26,  6.90s/it, lr: 1.0e-04 loss: 7.243e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 804/30000 [1:50:20<55:22:44,  6.83s/it, lr: 1.0e-04 loss: 7.243e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 804/30000 [1:50:20<55:22:44,  6.83s/it, lr: 1.0e-04 loss: 7.243e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 804/30000 [1:50:27<55:22:44,  6.83s/it, lr: 1.0e-04 loss: 7.979e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 804/30000 [1:50:27<55:22:44,  6.83s/it, lr: 1.0e-04 loss: 7.979e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 805/30000 [1:50:27<54:55:45,  6.77s/it, lr: 1.0e-04 loss: 7.979e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 805/30000 [1:50:27<54:55:45,  6.77s/it, lr: 1.0e-04 loss: 7.979e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 805/30000 [1:50:36<54:55:45,  6.77s/it, lr: 1.0e-04 loss: 6.489e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 805/30000 [1:50:36<54:55:45,  6.77s/it, lr: 1.0e-04 loss: 6.489e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 806/30000 [1:50:36<60:56:34,  7.52s/it, lr: 1.0e-04 loss: 6.489e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 806/30000 [1:50:36<60:56:34,  7.52s/it, lr: 1.0e-04 loss: 6.489e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 806/30000 [1:50:46<60:56:34,  7.52s/it, lr: 1.0e-04 loss: 4.591e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 806/30000 [1:50:46<60:56:34,  7.52s/it, lr: 1.0e-04 loss: 4.591e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 807/30000 [1:50:46<65:27:38,  8.07s/it, lr: 1.0e-04 loss: 4.591e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 807/30000 [1:50:46<65:27:38,  8.07s/it, lr: 1.0e-04 loss: 4.591e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 807/30000 [1:50:55<65:27:38,  8.07s/it, lr: 1.0e-04 loss: 9.266e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 807/30000 [1:50:55<65:27:38,  8.07s/it, lr: 1.0e-04 loss: 9.266e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 808/30000 [1:50:55<68:28:28,  8.44s/it, lr: 1.0e-04 loss: 9.266e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 808/30000 [1:50:55<68:28:28,  8.44s/it, lr: 1.0e-04 loss: 9.266e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 808/30000 [1:51:02<68:28:28,  8.44s/it, lr: 1.0e-04 loss: 4.069e-01]
Xiang_wan_2_2_lora:   3%|#                                       | 808/30000 [1:51:02<68:28:28,  8.44s/it, lr: 1.0e-04 loss: 4.069e-01]
Xiang_wan_2_2_lora:   3%|#                                       | 809/30000 [1:51:02<64:53:49,  8.00s/it, lr: 1.0e-04 loss: 4.069e-01]
Xiang_wan_2_2_lora:   3%|#                                       | 809/30000 [1:51:02<64:53:49,  8.00s/it, lr: 1.0e-04 loss: 4.069e-01]
Xiang_wan_2_2_lora:   3%|#                                       | 809/30000 [1:51:09<64:53:49,  8.00s/it, lr: 1.0e-04 loss: 8.196e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 809/30000 [1:51:09<64:53:49,  8.00s/it, lr: 1.0e-04 loss: 8.196e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 810/30000 [1:51:15<64:53:41,  8.00s/it, lr: 1.0e-04 loss: 8.423e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 810/30000 [1:51:15<64:53:41,  8.00s/it, lr: 1.0e-04 loss: 8.423e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 811/30000 [1:51:15<47:27:25,  5.85s/it, lr: 1.0e-04 loss: 8.423e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 811/30000 [1:51:15<47:27:25,  5.85s/it, lr: 1.0e-04 loss: 8.423e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 811/30000 [1:51:22<47:27:25,  5.85s/it, lr: 1.0e-04 loss: 1.009e-01]
Xiang_wan_2_2_lora:   3%|#                                       | 811/30000 [1:51:22<47:27:25,  5.85s/it, lr: 1.0e-04 loss: 1.009e-01]
Xiang_wan_2_2_lora:   3%|#                                       | 812/30000 [1:51:22<48:56:52,  6.04s/it, lr: 1.0e-04 loss: 1.009e-01]
Xiang_wan_2_2_lora:   3%|#                                       | 812/30000 [1:51:22<48:56:52,  6.04s/it, lr: 1.0e-04 loss: 1.009e-01]
Xiang_wan_2_2_lora:   3%|#                                       | 812/30000 [1:51:29<48:56:52,  6.04s/it, lr: 1.0e-04 loss: 3.837e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 812/30000 [1:51:29<48:56:52,  6.04s/it, lr: 1.0e-04 loss: 3.837e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 813/30000 [1:51:29<50:05:14,  6.18s/it, lr: 1.0e-04 loss: 3.837e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 813/30000 [1:51:29<50:05:14,  6.18s/it, lr: 1.0e-04 loss: 3.837e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 813/30000 [1:51:35<50:05:14,  6.18s/it, lr: 1.0e-04 loss: 9.920e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 813/30000 [1:51:35<50:05:14,  6.18s/it, lr: 1.0e-04 loss: 9.920e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 814/30000 [1:51:35<51:03:14,  6.30s/it, lr: 1.0e-04 loss: 9.920e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 814/30000 [1:51:35<51:03:14,  6.30s/it, lr: 1.0e-04 loss: 9.920e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 814/30000 [1:51:45<51:03:14,  6.30s/it, lr: 1.0e-04 loss: 2.669e-01]
Xiang_wan_2_2_lora:   3%|#                                       | 814/30000 [1:51:45<51:03:14,  6.30s/it, lr: 1.0e-04 loss: 2.669e-01]
Xiang_wan_2_2_lora:   3%|#                                       | 815/30000 [1:51:45<57:59:34,  7.15s/it, lr: 1.0e-04 loss: 2.669e-01]
Xiang_wan_2_2_lora:   3%|#                                       | 815/30000 [1:51:45<57:59:34,  7.15s/it, lr: 1.0e-04 loss: 2.669e-01]
Xiang_wan_2_2_lora:   3%|#                                       | 815/30000 [1:51:51<57:59:34,  7.15s/it, lr: 1.0e-04 loss: 7.808e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 815/30000 [1:51:51<57:59:34,  7.15s/it, lr: 1.0e-04 loss: 7.808e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 816/30000 [1:51:51<56:46:14,  7.00s/it, lr: 1.0e-04 loss: 7.808e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 816/30000 [1:51:51<56:46:14,  7.00s/it, lr: 1.0e-04 loss: 7.808e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 816/30000 [1:51:58<56:46:14,  7.00s/it, lr: 1.0e-04 loss: 7.448e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 816/30000 [1:51:58<56:46:14,  7.00s/it, lr: 1.0e-04 loss: 7.448e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 817/30000 [1:51:58<56:29:49,  6.97s/it, lr: 1.0e-04 loss: 7.448e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 817/30000 [1:51:58<56:29:49,  6.97s/it, lr: 1.0e-04 loss: 7.448e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 817/30000 [1:52:05<56:29:49,  6.97s/it, lr: 1.0e-04 loss: 6.130e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 817/30000 [1:52:05<56:29:49,  6.97s/it, lr: 1.0e-04 loss: 6.130e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 818/30000 [1:52:05<55:33:12,  6.85s/it, lr: 1.0e-04 loss: 6.130e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 818/30000 [1:52:05<55:33:12,  6.85s/it, lr: 1.0e-04 loss: 6.130e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 818/30000 [1:52:11<55:33:12,  6.85s/it, lr: 1.0e-04 loss: 5.435e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 818/30000 [1:52:11<55:33:12,  6.85s/it, lr: 1.0e-04 loss: 5.435e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 819/30000 [1:52:11<55:02:49,  6.79s/it, lr: 1.0e-04 loss: 5.435e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 819/30000 [1:52:11<55:02:49,  6.79s/it, lr: 1.0e-04 loss: 5.435e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 819/30000 [1:52:18<55:02:49,  6.79s/it, lr: 1.0e-04 loss: 8.668e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 819/30000 [1:52:18<55:02:49,  6.79s/it, lr: 1.0e-04 loss: 8.668e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 820/30000 [1:52:24<55:02:42,  6.79s/it, lr: 1.0e-04 loss: 6.362e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 820/30000 [1:52:24<55:02:42,  6.79s/it, lr: 1.0e-04 loss: 6.362e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 821/30000 [1:52:24<42:05:24,  5.19s/it, lr: 1.0e-04 loss: 6.362e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 821/30000 [1:52:24<42:05:24,  5.19s/it, lr: 1.0e-04 loss: 6.362e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 821/30000 [1:52:31<42:05:24,  5.19s/it, lr: 1.0e-04 loss: 7.928e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 821/30000 [1:52:31<42:05:24,  5.19s/it, lr: 1.0e-04 loss: 7.928e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 822/30000 [1:52:31<44:45:58,  5.52s/it, lr: 1.0e-04 loss: 7.928e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 822/30000 [1:52:31<44:45:58,  5.52s/it, lr: 1.0e-04 loss: 7.928e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 822/30000 [1:52:40<44:45:58,  5.52s/it, lr: 1.0e-04 loss: 4.067e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 822/30000 [1:52:40<44:45:58,  5.52s/it, lr: 1.0e-04 loss: 4.067e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 823/30000 [1:52:40<52:48:05,  6.51s/it, lr: 1.0e-04 loss: 4.067e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 823/30000 [1:52:40<52:48:05,  6.51s/it, lr: 1.0e-04 loss: 4.067e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 823/30000 [1:52:50<52:48:05,  6.51s/it, lr: 1.0e-04 loss: 6.570e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 823/30000 [1:52:50<52:48:05,  6.51s/it, lr: 1.0e-04 loss: 6.570e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 824/30000 [1:52:50<59:20:25,  7.32s/it, lr: 1.0e-04 loss: 6.570e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 824/30000 [1:52:50<59:20:25,  7.32s/it, lr: 1.0e-04 loss: 6.570e-02]
Xiang_wan_2_2_lora:   3%|#                                       | 824/30000 [1:52:56<59:20:25,  7.32s/it, lr: 1.0e-04 loss: 1.556e-01]
Xiang_wan_2_2_lora:   3%|#                                       | 824/30000 [1:52:56<59:20:25,  7.32s/it, lr: 1.0e-04 loss: 1.556e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 825/30000 [1:52:56<57:42:32,  7.12s/it, lr: 1.0e-04 loss: 1.556e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 825/30000 [1:52:56<57:42:32,  7.12s/it, lr: 1.0e-04 loss: 1.556e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 825/30000 [1:53:06<57:42:32,  7.12s/it, lr: 1.0e-04 loss: 6.465e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 825/30000 [1:53:06<57:42:32,  7.12s/it, lr: 1.0e-04 loss: 6.465e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 826/30000 [1:53:06<62:47:58,  7.75s/it, lr: 1.0e-04 loss: 6.465e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 826/30000 [1:53:06<62:47:58,  7.75s/it, lr: 1.0e-04 loss: 6.465e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 826/30000 [1:53:15<62:47:58,  7.75s/it, lr: 1.0e-04 loss: 1.084e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 826/30000 [1:53:15<62:47:58,  7.75s/it, lr: 1.0e-04 loss: 1.084e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 827/30000 [1:53:15<66:34:49,  8.22s/it, lr: 1.0e-04 loss: 1.084e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 827/30000 [1:53:15<66:34:49,  8.22s/it, lr: 1.0e-04 loss: 1.084e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 827/30000 [1:53:22<66:34:49,  8.22s/it, lr: 1.0e-04 loss: 7.526e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 827/30000 [1:53:22<66:34:49,  8.22s/it, lr: 1.0e-04 loss: 7.526e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 828/30000 [1:53:22<62:52:34,  7.76s/it, lr: 1.0e-04 loss: 7.526e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 828/30000 [1:53:22<62:52:34,  7.76s/it, lr: 1.0e-04 loss: 7.526e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 828/30000 [1:53:28<62:52:34,  7.76s/it, lr: 1.0e-04 loss: 2.043e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 828/30000 [1:53:28<62:52:34,  7.76s/it, lr: 1.0e-04 loss: 2.043e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 829/30000 [1:53:28<59:57:59,  7.40s/it, lr: 1.0e-04 loss: 2.043e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 829/30000 [1:53:28<59:57:59,  7.40s/it, lr: 1.0e-04 loss: 2.043e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 829/30000 [1:53:38<59:57:59,  7.40s/it, lr: 1.0e-04 loss: 6.903e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 829/30000 [1:53:38<59:57:59,  7.40s/it, lr: 1.0e-04 loss: 6.903e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 830/30000 [1:53:47<59:57:51,  7.40s/it, lr: 1.0e-04 loss: 4.425e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 830/30000 [1:53:47<59:57:51,  7.40s/it, lr: 1.0e-04 loss: 4.425e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 831/30000 [1:53:47<50:14:05,  6.20s/it, lr: 1.0e-04 loss: 4.425e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 831/30000 [1:53:47<50:14:05,  6.20s/it, lr: 1.0e-04 loss: 4.425e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 831/30000 [1:53:54<50:14:05,  6.20s/it, lr: 1.0e-04 loss: 9.417e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 831/30000 [1:53:54<50:14:05,  6.20s/it, lr: 1.0e-04 loss: 9.417e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 832/30000 [1:53:54<51:02:01,  6.30s/it, lr: 1.0e-04 loss: 9.417e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 832/30000 [1:53:54<51:02:01,  6.30s/it, lr: 1.0e-04 loss: 9.417e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 832/30000 [1:54:03<51:02:01,  6.30s/it, lr: 1.0e-04 loss: 4.237e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 832/30000 [1:54:03<51:02:01,  6.30s/it, lr: 1.0e-04 loss: 4.237e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 833/30000 [1:54:03<57:23:35,  7.08s/it, lr: 1.0e-04 loss: 4.237e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 833/30000 [1:54:03<57:23:35,  7.08s/it, lr: 1.0e-04 loss: 4.237e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 833/30000 [1:54:12<57:23:35,  7.08s/it, lr: 1.0e-04 loss: 4.127e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 833/30000 [1:54:12<57:23:35,  7.08s/it, lr: 1.0e-04 loss: 4.127e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 834/30000 [1:54:12<62:22:28,  7.70s/it, lr: 1.0e-04 loss: 4.127e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 834/30000 [1:54:12<62:22:28,  7.70s/it, lr: 1.0e-04 loss: 4.127e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 834/30000 [1:54:19<62:22:28,  7.70s/it, lr: 1.0e-04 loss: 2.002e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 834/30000 [1:54:19<62:22:28,  7.70s/it, lr: 1.0e-04 loss: 2.002e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 835/30000 [1:54:19<59:57:02,  7.40s/it, lr: 1.0e-04 loss: 2.002e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 835/30000 [1:54:19<59:57:02,  7.40s/it, lr: 1.0e-04 loss: 2.002e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 835/30000 [1:54:28<59:57:02,  7.40s/it, lr: 1.0e-04 loss: 7.580e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 835/30000 [1:54:28<59:57:02,  7.40s/it, lr: 1.0e-04 loss: 7.580e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 836/30000 [1:54:28<64:24:36,  7.95s/it, lr: 1.0e-04 loss: 7.580e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 836/30000 [1:54:28<64:24:36,  7.95s/it, lr: 1.0e-04 loss: 7.580e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 836/30000 [1:54:35<64:24:36,  7.95s/it, lr: 1.0e-04 loss: 1.265e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 836/30000 [1:54:35<64:24:36,  7.95s/it, lr: 1.0e-04 loss: 1.265e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 837/30000 [1:54:35<61:13:52,  7.56s/it, lr: 1.0e-04 loss: 1.265e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 837/30000 [1:54:35<61:13:52,  7.56s/it, lr: 1.0e-04 loss: 1.265e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 837/30000 [1:54:44<61:13:52,  7.56s/it, lr: 1.0e-04 loss: 7.920e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 837/30000 [1:54:44<61:13:52,  7.56s/it, lr: 1.0e-04 loss: 7.920e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 838/30000 [1:54:44<65:26:56,  8.08s/it, lr: 1.0e-04 loss: 7.920e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 838/30000 [1:54:44<65:26:56,  8.08s/it, lr: 1.0e-04 loss: 7.920e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 838/30000 [1:54:51<65:26:56,  8.08s/it, lr: 1.0e-04 loss: 5.426e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 838/30000 [1:54:51<65:26:56,  8.08s/it, lr: 1.0e-04 loss: 5.426e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 839/30000 [1:54:51<62:36:06,  7.73s/it, lr: 1.0e-04 loss: 5.426e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 839/30000 [1:54:51<62:36:06,  7.73s/it, lr: 1.0e-04 loss: 5.426e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 839/30000 [1:55:00<62:36:06,  7.73s/it, lr: 1.0e-04 loss: 7.007e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 839/30000 [1:55:00<62:36:06,  7.73s/it, lr: 1.0e-04 loss: 7.007e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 840/30000 [1:55:07<62:35:58,  7.73s/it, lr: 1.0e-04 loss: 1.379e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 840/30000 [1:55:07<62:35:58,  7.73s/it, lr: 1.0e-04 loss: 1.379e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 841/30000 [1:55:07<46:12:49,  5.71s/it, lr: 1.0e-04 loss: 1.379e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 841/30000 [1:55:07<46:12:49,  5.71s/it, lr: 1.0e-04 loss: 1.379e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 841/30000 [1:55:14<46:12:49,  5.71s/it, lr: 1.0e-04 loss: 7.289e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 841/30000 [1:55:14<46:12:49,  5.71s/it, lr: 1.0e-04 loss: 7.289e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 842/30000 [1:55:14<47:58:34,  5.92s/it, lr: 1.0e-04 loss: 7.289e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 842/30000 [1:55:14<47:58:34,  5.92s/it, lr: 1.0e-04 loss: 7.289e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 842/30000 [1:55:23<47:58:34,  5.92s/it, lr: 1.0e-04 loss: 4.514e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 842/30000 [1:55:23<47:58:34,  5.92s/it, lr: 1.0e-04 loss: 4.514e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 843/30000 [1:55:23<55:13:17,  6.82s/it, lr: 1.0e-04 loss: 4.514e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 843/30000 [1:55:23<55:13:17,  6.82s/it, lr: 1.0e-04 loss: 4.514e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 843/30000 [1:55:30<55:13:17,  6.82s/it, lr: 1.0e-04 loss: 1.383e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 843/30000 [1:55:30<55:13:17,  6.82s/it, lr: 1.0e-04 loss: 1.383e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 844/30000 [1:55:30<54:45:59,  6.76s/it, lr: 1.0e-04 loss: 1.383e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 844/30000 [1:55:30<54:45:59,  6.76s/it, lr: 1.0e-04 loss: 1.383e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 844/30000 [1:55:36<54:45:59,  6.76s/it, lr: 1.0e-04 loss: 9.132e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 844/30000 [1:55:36<54:45:59,  6.76s/it, lr: 1.0e-04 loss: 9.132e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 845/30000 [1:55:36<54:26:12,  6.72s/it, lr: 1.0e-04 loss: 9.132e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 845/30000 [1:55:36<54:26:12,  6.72s/it, lr: 1.0e-04 loss: 9.132e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 845/30000 [1:55:43<54:26:12,  6.72s/it, lr: 1.0e-04 loss: 8.157e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 845/30000 [1:55:43<54:26:12,  6.72s/it, lr: 1.0e-04 loss: 8.157e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 846/30000 [1:55:43<54:42:05,  6.75s/it, lr: 1.0e-04 loss: 8.157e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 846/30000 [1:55:43<54:42:05,  6.75s/it, lr: 1.0e-04 loss: 8.157e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 846/30000 [1:55:52<54:42:05,  6.75s/it, lr: 1.0e-04 loss: 5.539e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 846/30000 [1:55:52<54:42:05,  6.75s/it, lr: 1.0e-04 loss: 5.539e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 847/30000 [1:55:52<60:50:41,  7.51s/it, lr: 1.0e-04 loss: 5.539e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 847/30000 [1:55:52<60:50:41,  7.51s/it, lr: 1.0e-04 loss: 5.539e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 847/30000 [1:56:02<60:50:41,  7.51s/it, lr: 1.0e-04 loss: 8.573e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 847/30000 [1:56:02<60:50:41,  7.51s/it, lr: 1.0e-04 loss: 8.573e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 848/30000 [1:56:02<65:10:52,  8.05s/it, lr: 1.0e-04 loss: 8.573e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 848/30000 [1:56:02<65:10:52,  8.05s/it, lr: 1.0e-04 loss: 8.573e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 848/30000 [1:56:08<65:10:52,  8.05s/it, lr: 1.0e-04 loss: 7.330e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 848/30000 [1:56:08<65:10:52,  8.05s/it, lr: 1.0e-04 loss: 7.330e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 849/30000 [1:56:08<61:56:04,  7.65s/it, lr: 1.0e-04 loss: 7.330e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 849/30000 [1:56:08<61:56:04,  7.65s/it, lr: 1.0e-04 loss: 7.330e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 849/30000 [1:56:18<61:56:04,  7.65s/it, lr: 1.0e-04 loss: 5.680e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 849/30000 [1:56:18<61:56:04,  7.65s/it, lr: 1.0e-04 loss: 5.680e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 850/30000 [1:56:27<61:55:56,  7.65s/it, lr: 1.0e-04 loss: 4.655e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 850/30000 [1:56:27<61:55:56,  7.65s/it, lr: 1.0e-04 loss: 4.655e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 851/30000 [1:56:27<51:03:29,  6.31s/it, lr: 1.0e-04 loss: 4.655e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 851/30000 [1:56:27<51:03:29,  6.31s/it, lr: 1.0e-04 loss: 4.655e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 851/30000 [1:56:37<51:03:29,  6.31s/it, lr: 1.0e-04 loss: 6.806e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 851/30000 [1:56:37<51:03:29,  6.31s/it, lr: 1.0e-04 loss: 6.806e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 852/30000 [1:56:37<57:07:00,  7.05s/it, lr: 1.0e-04 loss: 6.806e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 852/30000 [1:56:37<57:07:00,  7.05s/it, lr: 1.0e-04 loss: 6.806e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 852/30000 [1:56:44<57:07:00,  7.05s/it, lr: 1.0e-04 loss: 8.537e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 852/30000 [1:56:44<57:07:00,  7.05s/it, lr: 1.0e-04 loss: 8.537e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 853/30000 [1:56:44<56:54:08,  7.03s/it, lr: 1.0e-04 loss: 8.537e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 853/30000 [1:56:44<56:54:08,  7.03s/it, lr: 1.0e-04 loss: 8.537e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 853/30000 [1:56:53<56:54:08,  7.03s/it, lr: 1.0e-04 loss: 6.310e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 853/30000 [1:56:53<56:54:08,  7.03s/it, lr: 1.0e-04 loss: 6.310e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 854/30000 [1:56:53<61:54:51,  7.65s/it, lr: 1.0e-04 loss: 6.310e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 854/30000 [1:56:53<61:54:51,  7.65s/it, lr: 1.0e-04 loss: 6.310e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 854/30000 [1:57:00<61:54:51,  7.65s/it, lr: 1.0e-04 loss: 8.571e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 854/30000 [1:57:00<61:54:51,  7.65s/it, lr: 1.0e-04 loss: 8.571e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 855/30000 [1:57:00<59:35:07,  7.36s/it, lr: 1.0e-04 loss: 8.571e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 855/30000 [1:57:00<59:35:07,  7.36s/it, lr: 1.0e-04 loss: 8.571e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 855/30000 [1:57:06<59:35:07,  7.36s/it, lr: 1.0e-04 loss: 1.029e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 855/30000 [1:57:06<59:35:07,  7.36s/it, lr: 1.0e-04 loss: 1.029e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 856/30000 [1:57:06<57:57:04,  7.16s/it, lr: 1.0e-04 loss: 1.029e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 856/30000 [1:57:06<57:57:04,  7.16s/it, lr: 1.0e-04 loss: 1.029e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 856/30000 [1:57:13<57:57:04,  7.16s/it, lr: 1.0e-04 loss: 7.500e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 856/30000 [1:57:13<57:57:04,  7.16s/it, lr: 1.0e-04 loss: 7.500e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 857/30000 [1:57:13<56:45:41,  7.01s/it, lr: 1.0e-04 loss: 7.500e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 857/30000 [1:57:13<56:45:41,  7.01s/it, lr: 1.0e-04 loss: 7.500e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 857/30000 [1:57:22<56:45:41,  7.01s/it, lr: 1.0e-04 loss: 8.815e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 857/30000 [1:57:22<56:45:41,  7.01s/it, lr: 1.0e-04 loss: 8.815e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 858/30000 [1:57:22<62:13:39,  7.69s/it, lr: 1.0e-04 loss: 8.815e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 858/30000 [1:57:22<62:13:39,  7.69s/it, lr: 1.0e-04 loss: 8.815e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 858/30000 [1:57:32<62:13:39,  7.69s/it, lr: 1.0e-04 loss: 7.121e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 858/30000 [1:57:32<62:13:39,  7.69s/it, lr: 1.0e-04 loss: 7.121e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 859/30000 [1:57:32<66:15:41,  8.19s/it, lr: 1.0e-04 loss: 7.121e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 859/30000 [1:57:32<66:15:41,  8.19s/it, lr: 1.0e-04 loss: 7.121e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 859/30000 [1:57:38<66:15:41,  8.19s/it, lr: 1.0e-04 loss: 8.719e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 859/30000 [1:57:38<66:15:41,  8.19s/it, lr: 1.0e-04 loss: 8.719e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 860/30000 [1:57:48<66:15:33,  8.19s/it, lr: 1.0e-04 loss: 3.814e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 860/30000 [1:57:48<66:15:33,  8.19s/it, lr: 1.0e-04 loss: 3.814e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 861/30000 [1:57:48<53:15:11,  6.58s/it, lr: 1.0e-04 loss: 3.814e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 861/30000 [1:57:48<53:15:11,  6.58s/it, lr: 1.0e-04 loss: 3.814e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 861/30000 [1:57:54<53:15:11,  6.58s/it, lr: 1.0e-04 loss: 9.017e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 861/30000 [1:57:54<53:15:11,  6.58s/it, lr: 1.0e-04 loss: 9.017e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 862/30000 [1:57:54<53:21:35,  6.59s/it, lr: 1.0e-04 loss: 9.017e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 862/30000 [1:57:54<53:21:35,  6.59s/it, lr: 1.0e-04 loss: 9.017e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 862/30000 [1:58:04<53:21:35,  6.59s/it, lr: 1.0e-04 loss: 5.290e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 862/30000 [1:58:04<53:21:35,  6.59s/it, lr: 1.0e-04 loss: 5.290e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 863/30000 [1:58:04<59:15:26,  7.32s/it, lr: 1.0e-04 loss: 5.290e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 863/30000 [1:58:04<59:15:26,  7.32s/it, lr: 1.0e-04 loss: 5.290e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 863/30000 [1:58:10<59:15:26,  7.32s/it, lr: 1.0e-04 loss: 1.319e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 863/30000 [1:58:10<59:15:26,  7.32s/it, lr: 1.0e-04 loss: 1.319e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 864/30000 [1:58:10<57:38:45,  7.12s/it, lr: 1.0e-04 loss: 1.319e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 864/30000 [1:58:10<57:38:45,  7.12s/it, lr: 1.0e-04 loss: 1.319e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 864/30000 [1:58:17<57:38:45,  7.12s/it, lr: 1.0e-04 loss: 4.959e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 864/30000 [1:58:17<57:38:45,  7.12s/it, lr: 1.0e-04 loss: 4.959e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 865/30000 [1:58:17<56:36:05,  6.99s/it, lr: 1.0e-04 loss: 4.959e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 865/30000 [1:58:17<56:36:05,  6.99s/it, lr: 1.0e-04 loss: 4.959e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 865/30000 [1:58:26<56:36:05,  6.99s/it, lr: 1.0e-04 loss: 7.877e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 865/30000 [1:58:26<56:36:05,  6.99s/it, lr: 1.0e-04 loss: 7.877e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 866/30000 [1:58:26<61:55:37,  7.65s/it, lr: 1.0e-04 loss: 7.877e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 866/30000 [1:58:26<61:55:37,  7.65s/it, lr: 1.0e-04 loss: 7.877e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 866/30000 [1:58:36<61:55:37,  7.65s/it, lr: 1.0e-04 loss: 7.667e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 866/30000 [1:58:36<61:55:37,  7.65s/it, lr: 1.0e-04 loss: 7.667e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 867/30000 [1:58:36<66:00:26,  8.16s/it, lr: 1.0e-04 loss: 7.667e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 867/30000 [1:58:36<66:00:26,  8.16s/it, lr: 1.0e-04 loss: 7.667e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 867/30000 [1:58:45<66:00:26,  8.16s/it, lr: 1.0e-04 loss: 7.372e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 867/30000 [1:58:45<66:00:26,  8.16s/it, lr: 1.0e-04 loss: 7.372e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 868/30000 [1:58:45<69:19:54,  8.57s/it, lr: 1.0e-04 loss: 7.372e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 868/30000 [1:58:45<69:19:54,  8.57s/it, lr: 1.0e-04 loss: 7.372e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 868/30000 [1:58:52<69:19:54,  8.57s/it, lr: 1.0e-04 loss: 1.285e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 868/30000 [1:58:52<69:19:54,  8.57s/it, lr: 1.0e-04 loss: 1.285e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 869/30000 [1:58:52<64:49:36,  8.01s/it, lr: 1.0e-04 loss: 1.285e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 869/30000 [1:58:52<64:49:36,  8.01s/it, lr: 1.0e-04 loss: 1.285e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 869/30000 [1:59:01<64:49:36,  8.01s/it, lr: 1.0e-04 loss: 8.652e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 869/30000 [1:59:01<64:49:36,  8.01s/it, lr: 1.0e-04 loss: 8.652e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 870/30000 [1:59:11<64:49:28,  8.01s/it, lr: 1.0e-04 loss: 3.829e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 870/30000 [1:59:11<64:49:28,  8.01s/it, lr: 1.0e-04 loss: 3.829e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 871/30000 [1:59:11<52:38:01,  6.50s/it, lr: 1.0e-04 loss: 3.829e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 871/30000 [1:59:11<52:38:01,  6.50s/it, lr: 1.0e-04 loss: 3.829e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 871/30000 [1:59:20<52:38:01,  6.50s/it, lr: 1.0e-04 loss: 1.037e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 871/30000 [1:59:20<52:38:01,  6.50s/it, lr: 1.0e-04 loss: 1.037e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 872/30000 [1:59:20<58:16:33,  7.20s/it, lr: 1.0e-04 loss: 1.037e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 872/30000 [1:59:20<58:16:33,  7.20s/it, lr: 1.0e-04 loss: 1.037e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 872/30000 [1:59:27<58:16:33,  7.20s/it, lr: 1.0e-04 loss: 7.287e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 872/30000 [1:59:27<58:16:33,  7.20s/it, lr: 1.0e-04 loss: 7.287e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 873/30000 [1:59:27<57:10:47,  7.07s/it, lr: 1.0e-04 loss: 7.287e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 873/30000 [1:59:27<57:10:47,  7.07s/it, lr: 1.0e-04 loss: 7.287e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 873/30000 [1:59:33<57:10:47,  7.07s/it, lr: 1.0e-04 loss: 9.773e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 873/30000 [1:59:33<57:10:47,  7.07s/it, lr: 1.0e-04 loss: 9.773e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 874/30000 [1:59:33<56:04:59,  6.93s/it, lr: 1.0e-04 loss: 9.773e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 874/30000 [1:59:33<56:04:59,  6.93s/it, lr: 1.0e-04 loss: 9.773e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 874/30000 [1:59:40<56:04:59,  6.93s/it, lr: 1.0e-04 loss: 6.279e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 874/30000 [1:59:40<56:04:59,  6.93s/it, lr: 1.0e-04 loss: 6.279e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 875/30000 [1:59:40<55:52:53,  6.91s/it, lr: 1.0e-04 loss: 6.279e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 875/30000 [1:59:40<55:52:53,  6.91s/it, lr: 1.0e-04 loss: 6.279e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 875/30000 [1:59:47<55:52:53,  6.91s/it, lr: 1.0e-04 loss: 7.580e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 875/30000 [1:59:47<55:52:53,  6.91s/it, lr: 1.0e-04 loss: 7.580e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 876/30000 [1:59:47<55:06:36,  6.81s/it, lr: 1.0e-04 loss: 7.580e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 876/30000 [1:59:47<55:06:36,  6.81s/it, lr: 1.0e-04 loss: 7.580e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 876/30000 [1:59:56<55:06:36,  6.81s/it, lr: 1.0e-04 loss: 4.457e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 876/30000 [1:59:56<55:06:36,  6.81s/it, lr: 1.0e-04 loss: 4.457e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 877/30000 [1:59:56<61:08:21,  7.56s/it, lr: 1.0e-04 loss: 4.457e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 877/30000 [1:59:56<61:08:21,  7.56s/it, lr: 1.0e-04 loss: 4.457e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 877/30000 [2:00:03<61:08:21,  7.56s/it, lr: 1.0e-04 loss: 9.551e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 877/30000 [2:00:03<61:08:21,  7.56s/it, lr: 1.0e-04 loss: 9.551e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 878/30000 [2:00:03<59:00:38,  7.29s/it, lr: 1.0e-04 loss: 9.551e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 878/30000 [2:00:03<59:00:38,  7.29s/it, lr: 1.0e-04 loss: 9.551e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 878/30000 [2:00:09<59:00:38,  7.29s/it, lr: 1.0e-04 loss: 1.657e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 878/30000 [2:00:09<59:00:38,  7.29s/it, lr: 1.0e-04 loss: 1.657e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 879/30000 [2:00:09<57:29:10,  7.11s/it, lr: 1.0e-04 loss: 1.657e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 879/30000 [2:00:09<57:29:10,  7.11s/it, lr: 1.0e-04 loss: 1.657e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 879/30000 [2:00:16<57:29:10,  7.11s/it, lr: 1.0e-04 loss: 9.562e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 879/30000 [2:00:16<57:29:10,  7.11s/it, lr: 1.0e-04 loss: 9.562e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 880/30000 [2:00:25<57:29:03,  7.11s/it, lr: 1.0e-04 loss: 5.529e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 880/30000 [2:00:25<57:29:03,  7.11s/it, lr: 1.0e-04 loss: 5.529e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 881/30000 [2:00:25<48:32:16,  6.00s/it, lr: 1.0e-04 loss: 5.529e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 881/30000 [2:00:25<48:32:16,  6.00s/it, lr: 1.0e-04 loss: 5.529e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 881/30000 [2:00:32<48:32:16,  6.00s/it, lr: 1.0e-04 loss: 8.647e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 881/30000 [2:00:32<48:32:16,  6.00s/it, lr: 1.0e-04 loss: 8.647e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 882/30000 [2:00:32<50:11:40,  6.21s/it, lr: 1.0e-04 loss: 8.647e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 882/30000 [2:00:32<50:11:40,  6.21s/it, lr: 1.0e-04 loss: 8.647e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 882/30000 [2:00:42<50:11:40,  6.21s/it, lr: 1.0e-04 loss: 4.942e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 882/30000 [2:00:42<50:11:40,  6.21s/it, lr: 1.0e-04 loss: 4.942e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 883/30000 [2:00:42<56:55:11,  7.04s/it, lr: 1.0e-04 loss: 4.942e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 883/30000 [2:00:42<56:55:11,  7.04s/it, lr: 1.0e-04 loss: 4.942e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 883/30000 [2:00:48<56:55:11,  7.04s/it, lr: 1.0e-04 loss: 7.713e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 883/30000 [2:00:48<56:55:11,  7.04s/it, lr: 1.0e-04 loss: 7.713e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 884/30000 [2:00:48<55:47:20,  6.90s/it, lr: 1.0e-04 loss: 7.713e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 884/30000 [2:00:48<55:47:20,  6.90s/it, lr: 1.0e-04 loss: 7.713e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 884/30000 [2:00:55<55:47:20,  6.90s/it, lr: 1.0e-04 loss: 1.394e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 884/30000 [2:00:55<55:47:20,  6.90s/it, lr: 1.0e-04 loss: 1.394e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 885/30000 [2:00:55<55:20:18,  6.84s/it, lr: 1.0e-04 loss: 1.394e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 885/30000 [2:00:55<55:20:18,  6.84s/it, lr: 1.0e-04 loss: 1.394e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 885/30000 [2:01:01<55:20:18,  6.84s/it, lr: 1.0e-04 loss: 7.742e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 885/30000 [2:01:01<55:20:18,  6.84s/it, lr: 1.0e-04 loss: 7.742e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 886/30000 [2:01:01<54:40:38,  6.76s/it, lr: 1.0e-04 loss: 7.742e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 886/30000 [2:01:01<54:40:38,  6.76s/it, lr: 1.0e-04 loss: 7.742e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 886/30000 [2:01:08<54:40:38,  6.76s/it, lr: 1.0e-04 loss: 5.502e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 886/30000 [2:01:08<54:40:38,  6.76s/it, lr: 1.0e-04 loss: 5.502e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 887/30000 [2:01:08<54:18:21,  6.72s/it, lr: 1.0e-04 loss: 5.502e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 887/30000 [2:01:08<54:18:21,  6.72s/it, lr: 1.0e-04 loss: 5.502e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 887/30000 [2:01:15<54:18:21,  6.72s/it, lr: 1.0e-04 loss: 7.731e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 887/30000 [2:01:15<54:18:21,  6.72s/it, lr: 1.0e-04 loss: 7.731e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 888/30000 [2:01:15<54:21:42,  6.72s/it, lr: 1.0e-04 loss: 7.731e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 888/30000 [2:01:15<54:21:42,  6.72s/it, lr: 1.0e-04 loss: 7.731e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 888/30000 [2:01:24<54:21:42,  6.72s/it, lr: 1.0e-04 loss: 1.139e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 888/30000 [2:01:24<54:21:42,  6.72s/it, lr: 1.0e-04 loss: 1.139e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 889/30000 [2:01:24<61:12:38,  7.57s/it, lr: 1.0e-04 loss: 1.139e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 889/30000 [2:01:24<61:12:38,  7.57s/it, lr: 1.0e-04 loss: 1.139e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 889/30000 [2:01:34<61:12:38,  7.57s/it, lr: 1.0e-04 loss: 7.202e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 889/30000 [2:01:34<61:12:38,  7.57s/it, lr: 1.0e-04 loss: 7.202e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 890/30000 [2:01:40<61:12:31,  7.57s/it, lr: 1.0e-04 loss: 6.958e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 890/30000 [2:01:40<61:12:31,  7.57s/it, lr: 1.0e-04 loss: 6.958e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 891/30000 [2:01:40<45:28:33,  5.62s/it, lr: 1.0e-04 loss: 6.958e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 891/30000 [2:01:40<45:28:33,  5.62s/it, lr: 1.0e-04 loss: 6.958e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 891/30000 [2:01:47<45:28:33,  5.62s/it, lr: 1.0e-04 loss: 8.290e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 891/30000 [2:01:47<45:28:33,  5.62s/it, lr: 1.0e-04 loss: 8.290e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 892/30000 [2:01:47<47:13:55,  5.84s/it, lr: 1.0e-04 loss: 8.290e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 892/30000 [2:01:47<47:13:55,  5.84s/it, lr: 1.0e-04 loss: 8.290e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 892/30000 [2:01:53<47:13:55,  5.84s/it, lr: 1.0e-04 loss: 2.268e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 892/30000 [2:01:53<47:13:55,  5.84s/it, lr: 1.0e-04 loss: 2.268e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 893/30000 [2:01:53<48:46:41,  6.03s/it, lr: 1.0e-04 loss: 2.268e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 893/30000 [2:01:53<48:46:41,  6.03s/it, lr: 1.0e-04 loss: 2.268e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 893/30000 [2:02:03<48:46:41,  6.03s/it, lr: 1.0e-04 loss: 6.048e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 893/30000 [2:02:03<48:46:41,  6.03s/it, lr: 1.0e-04 loss: 6.048e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 894/30000 [2:02:03<55:58:17,  6.92s/it, lr: 1.0e-04 loss: 6.048e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 894/30000 [2:02:03<55:58:17,  6.92s/it, lr: 1.0e-04 loss: 6.048e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 894/30000 [2:02:09<55:58:17,  6.92s/it, lr: 1.0e-04 loss: 8.136e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 894/30000 [2:02:09<55:58:17,  6.92s/it, lr: 1.0e-04 loss: 8.136e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 895/30000 [2:02:09<55:24:25,  6.85s/it, lr: 1.0e-04 loss: 8.136e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 895/30000 [2:02:09<55:24:25,  6.85s/it, lr: 1.0e-04 loss: 8.136e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 895/30000 [2:02:16<55:24:25,  6.85s/it, lr: 1.0e-04 loss: 9.810e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 895/30000 [2:02:16<55:24:25,  6.85s/it, lr: 1.0e-04 loss: 9.810e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 896/30000 [2:02:16<54:44:52,  6.77s/it, lr: 1.0e-04 loss: 9.810e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 896/30000 [2:02:16<54:44:52,  6.77s/it, lr: 1.0e-04 loss: 9.810e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 896/30000 [2:02:23<54:44:52,  6.77s/it, lr: 1.0e-04 loss: 7.577e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 896/30000 [2:02:23<54:44:52,  6.77s/it, lr: 1.0e-04 loss: 7.577e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 897/30000 [2:02:23<54:56:52,  6.80s/it, lr: 1.0e-04 loss: 7.577e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 897/30000 [2:02:23<54:56:52,  6.80s/it, lr: 1.0e-04 loss: 7.577e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 897/30000 [2:02:29<54:56:52,  6.80s/it, lr: 1.0e-04 loss: 8.844e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 897/30000 [2:02:29<54:56:52,  6.80s/it, lr: 1.0e-04 loss: 8.844e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 898/30000 [2:02:29<54:20:05,  6.72s/it, lr: 1.0e-04 loss: 8.844e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 898/30000 [2:02:29<54:20:05,  6.72s/it, lr: 1.0e-04 loss: 8.844e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 898/30000 [2:02:36<54:20:05,  6.72s/it, lr: 1.0e-04 loss: 1.465e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 898/30000 [2:02:36<54:20:05,  6.72s/it, lr: 1.0e-04 loss: 1.465e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 899/30000 [2:02:36<54:06:15,  6.69s/it, lr: 1.0e-04 loss: 1.465e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 899/30000 [2:02:36<54:06:15,  6.69s/it, lr: 1.0e-04 loss: 1.465e-01]
Xiang_wan_2_2_lora:   3%|#1                                      | 899/30000 [2:02:45<54:06:15,  6.69s/it, lr: 1.0e-04 loss: 5.512e-02]
Xiang_wan_2_2_lora:   3%|#1                                      | 899/30000 [2:02:45<54:06:15,  6.69s/it, lr: 1.0e-04 loss: 5.512e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 900/30000 [2:02:55<54:06:08,  6.69s/it, lr: 1.0e-04 loss: 5.371e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 900/30000 [2:02:55<54:06:08,  6.69s/it, lr: 1.0e-04 loss: 5.371e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 901/30000 [2:02:55<46:34:34,  5.76s/it, lr: 1.0e-04 loss: 5.371e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 901/30000 [2:02:55<46:34:34,  5.76s/it, lr: 1.0e-04 loss: 5.371e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 901/30000 [2:03:01<46:34:34,  5.76s/it, lr: 1.0e-04 loss: 6.640e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 901/30000 [2:03:01<46:34:34,  5.76s/it, lr: 1.0e-04 loss: 6.640e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 902/30000 [2:03:01<48:16:23,  5.97s/it, lr: 1.0e-04 loss: 6.640e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 902/30000 [2:03:01<48:16:23,  5.97s/it, lr: 1.0e-04 loss: 6.640e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 902/30000 [2:03:10<48:16:23,  5.97s/it, lr: 1.0e-04 loss: 5.624e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 902/30000 [2:03:10<48:16:23,  5.97s/it, lr: 1.0e-04 loss: 5.624e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 903/30000 [2:03:10<55:13:37,  6.83s/it, lr: 1.0e-04 loss: 5.624e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 903/30000 [2:03:10<55:13:37,  6.83s/it, lr: 1.0e-04 loss: 5.624e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 903/30000 [2:03:17<55:13:37,  6.83s/it, lr: 1.0e-04 loss: 8.194e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 903/30000 [2:03:17<55:13:37,  6.83s/it, lr: 1.0e-04 loss: 8.194e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 904/30000 [2:03:17<55:13:10,  6.83s/it, lr: 1.0e-04 loss: 8.194e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 904/30000 [2:03:17<55:13:10,  6.83s/it, lr: 1.0e-04 loss: 8.194e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 904/30000 [2:03:27<55:13:10,  6.83s/it, lr: 1.0e-04 loss: 7.262e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 904/30000 [2:03:27<55:13:10,  6.83s/it, lr: 1.0e-04 loss: 7.262e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 905/30000 [2:03:27<60:54:08,  7.54s/it, lr: 1.0e-04 loss: 7.262e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 905/30000 [2:03:27<60:54:08,  7.54s/it, lr: 1.0e-04 loss: 7.262e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 905/30000 [2:03:33<60:54:08,  7.54s/it, lr: 1.0e-04 loss: 1.228e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 905/30000 [2:03:33<60:54:08,  7.54s/it, lr: 1.0e-04 loss: 1.228e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 906/30000 [2:03:33<59:18:03,  7.34s/it, lr: 1.0e-04 loss: 1.228e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 906/30000 [2:03:33<59:18:03,  7.34s/it, lr: 1.0e-04 loss: 1.228e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 906/30000 [2:03:40<59:18:03,  7.34s/it, lr: 1.0e-04 loss: 5.624e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 906/30000 [2:03:40<59:18:03,  7.34s/it, lr: 1.0e-04 loss: 5.624e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 907/30000 [2:03:40<57:27:27,  7.11s/it, lr: 1.0e-04 loss: 5.624e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 907/30000 [2:03:40<57:27:27,  7.11s/it, lr: 1.0e-04 loss: 5.624e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 907/30000 [2:03:47<57:27:27,  7.11s/it, lr: 1.0e-04 loss: 1.121e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 907/30000 [2:03:47<57:27:27,  7.11s/it, lr: 1.0e-04 loss: 1.121e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 908/30000 [2:03:47<56:12:43,  6.96s/it, lr: 1.0e-04 loss: 1.121e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 908/30000 [2:03:47<56:12:43,  6.96s/it, lr: 1.0e-04 loss: 1.121e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 908/30000 [2:03:53<56:12:43,  6.96s/it, lr: 1.0e-04 loss: 7.248e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 908/30000 [2:03:53<56:12:43,  6.96s/it, lr: 1.0e-04 loss: 7.248e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 909/30000 [2:03:53<55:14:41,  6.84s/it, lr: 1.0e-04 loss: 7.248e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 909/30000 [2:03:53<55:14:41,  6.84s/it, lr: 1.0e-04 loss: 7.248e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 909/30000 [2:04:00<55:14:41,  6.84s/it, lr: 1.0e-04 loss: 1.499e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 909/30000 [2:04:00<55:14:41,  6.84s/it, lr: 1.0e-04 loss: 1.499e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 910/30000 [2:04:09<55:14:34,  6.84s/it, lr: 1.0e-04 loss: 9.656e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 910/30000 [2:04:09<55:14:34,  6.84s/it, lr: 1.0e-04 loss: 9.656e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 911/30000 [2:04:09<47:34:20,  5.89s/it, lr: 1.0e-04 loss: 9.656e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 911/30000 [2:04:09<47:34:20,  5.89s/it, lr: 1.0e-04 loss: 9.656e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 911/30000 [2:04:19<47:34:20,  5.89s/it, lr: 1.0e-04 loss: 7.823e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 911/30000 [2:04:19<47:34:20,  5.89s/it, lr: 1.0e-04 loss: 7.823e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 912/30000 [2:04:19<54:27:42,  6.74s/it, lr: 1.0e-04 loss: 7.823e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 912/30000 [2:04:19<54:27:42,  6.74s/it, lr: 1.0e-04 loss: 7.823e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 912/30000 [2:04:25<54:27:42,  6.74s/it, lr: 1.0e-04 loss: 9.637e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 912/30000 [2:04:25<54:27:42,  6.74s/it, lr: 1.0e-04 loss: 9.637e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 913/30000 [2:04:25<54:14:03,  6.71s/it, lr: 1.0e-04 loss: 9.637e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 913/30000 [2:04:25<54:14:03,  6.71s/it, lr: 1.0e-04 loss: 9.637e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 913/30000 [2:04:32<54:14:03,  6.71s/it, lr: 1.0e-04 loss: 1.121e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 913/30000 [2:04:32<54:14:03,  6.71s/it, lr: 1.0e-04 loss: 1.121e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 914/30000 [2:04:32<53:56:12,  6.68s/it, lr: 1.0e-04 loss: 1.121e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 914/30000 [2:04:32<53:56:12,  6.68s/it, lr: 1.0e-04 loss: 1.121e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 914/30000 [2:04:38<53:56:12,  6.68s/it, lr: 1.0e-04 loss: 2.595e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 914/30000 [2:04:38<53:56:12,  6.68s/it, lr: 1.0e-04 loss: 2.595e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 915/30000 [2:04:38<53:45:40,  6.65s/it, lr: 1.0e-04 loss: 2.595e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 915/30000 [2:04:38<53:45:40,  6.65s/it, lr: 1.0e-04 loss: 2.595e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 915/30000 [2:04:45<53:45:40,  6.65s/it, lr: 1.0e-04 loss: 1.754e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 915/30000 [2:04:45<53:45:40,  6.65s/it, lr: 1.0e-04 loss: 1.754e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 916/30000 [2:04:45<53:43:33,  6.65s/it, lr: 1.0e-04 loss: 1.754e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 916/30000 [2:04:45<53:43:33,  6.65s/it, lr: 1.0e-04 loss: 1.754e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 916/30000 [2:04:52<53:43:33,  6.65s/it, lr: 1.0e-04 loss: 7.323e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 916/30000 [2:04:52<53:43:33,  6.65s/it, lr: 1.0e-04 loss: 7.323e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 917/30000 [2:04:52<53:21:51,  6.61s/it, lr: 1.0e-04 loss: 7.323e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 917/30000 [2:04:52<53:21:51,  6.61s/it, lr: 1.0e-04 loss: 7.323e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 917/30000 [2:04:58<53:21:51,  6.61s/it, lr: 1.0e-04 loss: 1.120e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 917/30000 [2:04:58<53:21:51,  6.61s/it, lr: 1.0e-04 loss: 1.120e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 918/30000 [2:04:58<53:55:08,  6.67s/it, lr: 1.0e-04 loss: 1.120e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 918/30000 [2:04:58<53:55:08,  6.67s/it, lr: 1.0e-04 loss: 1.120e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 918/30000 [2:05:05<53:55:08,  6.67s/it, lr: 1.0e-04 loss: 3.230e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 918/30000 [2:05:05<53:55:08,  6.67s/it, lr: 1.0e-04 loss: 3.230e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 919/30000 [2:05:05<53:32:50,  6.63s/it, lr: 1.0e-04 loss: 3.230e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 919/30000 [2:05:05<53:32:50,  6.63s/it, lr: 1.0e-04 loss: 3.230e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 919/30000 [2:05:14<53:32:50,  6.63s/it, lr: 1.0e-04 loss: 9.699e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 919/30000 [2:05:14<53:32:50,  6.63s/it, lr: 1.0e-04 loss: 9.699e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 920/30000 [2:05:24<53:32:43,  6.63s/it, lr: 1.0e-04 loss: 4.075e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 920/30000 [2:05:24<53:32:43,  6.63s/it, lr: 1.0e-04 loss: 4.075e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 921/30000 [2:05:24<46:21:11,  5.74s/it, lr: 1.0e-04 loss: 4.075e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 921/30000 [2:05:24<46:21:11,  5.74s/it, lr: 1.0e-04 loss: 4.075e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 921/30000 [2:05:33<46:21:11,  5.74s/it, lr: 1.0e-04 loss: 8.587e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 921/30000 [2:05:33<46:21:11,  5.74s/it, lr: 1.0e-04 loss: 8.587e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 922/30000 [2:05:33<53:32:32,  6.63s/it, lr: 1.0e-04 loss: 8.587e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 922/30000 [2:05:33<53:32:32,  6.63s/it, lr: 1.0e-04 loss: 8.587e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 922/30000 [2:05:40<53:32:32,  6.63s/it, lr: 1.0e-04 loss: 8.408e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 922/30000 [2:05:40<53:32:32,  6.63s/it, lr: 1.0e-04 loss: 8.408e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 923/30000 [2:05:40<53:34:04,  6.63s/it, lr: 1.0e-04 loss: 8.408e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 923/30000 [2:05:40<53:34:04,  6.63s/it, lr: 1.0e-04 loss: 8.408e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 923/30000 [2:05:46<53:34:04,  6.63s/it, lr: 1.0e-04 loss: 9.796e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 923/30000 [2:05:46<53:34:04,  6.63s/it, lr: 1.0e-04 loss: 9.796e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 924/30000 [2:05:46<53:29:46,  6.62s/it, lr: 1.0e-04 loss: 9.796e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 924/30000 [2:05:46<53:29:46,  6.62s/it, lr: 1.0e-04 loss: 9.796e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 924/30000 [2:05:53<53:29:46,  6.62s/it, lr: 1.0e-04 loss: 9.167e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 924/30000 [2:05:53<53:29:46,  6.62s/it, lr: 1.0e-04 loss: 9.167e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 925/30000 [2:05:53<53:27:39,  6.62s/it, lr: 1.0e-04 loss: 9.167e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 925/30000 [2:05:53<53:27:39,  6.62s/it, lr: 1.0e-04 loss: 9.167e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 925/30000 [2:06:00<53:27:39,  6.62s/it, lr: 1.0e-04 loss: 8.937e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 925/30000 [2:06:00<53:27:39,  6.62s/it, lr: 1.0e-04 loss: 8.937e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 926/30000 [2:06:00<53:55:17,  6.68s/it, lr: 1.0e-04 loss: 8.937e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 926/30000 [2:06:00<53:55:17,  6.68s/it, lr: 1.0e-04 loss: 8.937e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 926/30000 [2:06:09<53:55:17,  6.68s/it, lr: 1.0e-04 loss: 1.648e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 926/30000 [2:06:09<53:55:17,  6.68s/it, lr: 1.0e-04 loss: 1.648e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 927/30000 [2:06:09<60:24:51,  7.48s/it, lr: 1.0e-04 loss: 1.648e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 927/30000 [2:06:09<60:24:51,  7.48s/it, lr: 1.0e-04 loss: 1.648e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 927/30000 [2:06:16<60:24:51,  7.48s/it, lr: 1.0e-04 loss: 8.933e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 927/30000 [2:06:16<60:24:51,  7.48s/it, lr: 1.0e-04 loss: 8.933e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 928/30000 [2:06:16<58:40:09,  7.27s/it, lr: 1.0e-04 loss: 8.933e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 928/30000 [2:06:16<58:40:09,  7.27s/it, lr: 1.0e-04 loss: 8.933e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 928/30000 [2:06:22<58:40:09,  7.27s/it, lr: 1.0e-04 loss: 5.299e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 928/30000 [2:06:22<58:40:09,  7.27s/it, lr: 1.0e-04 loss: 5.299e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 929/30000 [2:06:22<56:58:12,  7.05s/it, lr: 1.0e-04 loss: 5.299e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 929/30000 [2:06:22<56:58:12,  7.05s/it, lr: 1.0e-04 loss: 5.299e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 929/30000 [2:06:29<56:58:12,  7.05s/it, lr: 1.0e-04 loss: 7.696e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 929/30000 [2:06:29<56:58:12,  7.05s/it, lr: 1.0e-04 loss: 7.696e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 930/30000 [2:06:38<56:58:05,  7.05s/it, lr: 1.0e-04 loss: 8.857e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 930/30000 [2:06:38<56:58:05,  7.05s/it, lr: 1.0e-04 loss: 8.857e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 931/30000 [2:06:38<48:00:50,  5.95s/it, lr: 1.0e-04 loss: 8.857e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 931/30000 [2:06:38<48:00:50,  5.95s/it, lr: 1.0e-04 loss: 8.857e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 931/30000 [2:06:48<48:00:50,  5.95s/it, lr: 1.0e-04 loss: 7.593e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 931/30000 [2:06:48<48:00:50,  5.95s/it, lr: 1.0e-04 loss: 7.593e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 932/30000 [2:06:48<54:57:55,  6.81s/it, lr: 1.0e-04 loss: 7.593e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 932/30000 [2:06:48<54:57:55,  6.81s/it, lr: 1.0e-04 loss: 7.593e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 932/30000 [2:06:55<54:57:55,  6.81s/it, lr: 1.0e-04 loss: 6.149e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 932/30000 [2:06:55<54:57:55,  6.81s/it, lr: 1.0e-04 loss: 6.149e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 933/30000 [2:06:55<55:07:00,  6.83s/it, lr: 1.0e-04 loss: 6.149e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 933/30000 [2:06:55<55:07:00,  6.83s/it, lr: 1.0e-04 loss: 6.149e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 933/30000 [2:07:04<55:07:00,  6.83s/it, lr: 1.0e-04 loss: 9.424e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 933/30000 [2:07:04<55:07:00,  6.83s/it, lr: 1.0e-04 loss: 9.424e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 934/30000 [2:07:04<60:41:57,  7.52s/it, lr: 1.0e-04 loss: 9.424e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 934/30000 [2:07:04<60:41:57,  7.52s/it, lr: 1.0e-04 loss: 9.424e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 934/30000 [2:07:13<60:41:57,  7.52s/it, lr: 1.0e-04 loss: 3.411e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 934/30000 [2:07:13<60:41:57,  7.52s/it, lr: 1.0e-04 loss: 3.411e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 935/30000 [2:07:13<64:49:31,  8.03s/it, lr: 1.0e-04 loss: 3.411e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 935/30000 [2:07:13<64:49:31,  8.03s/it, lr: 1.0e-04 loss: 3.411e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 935/30000 [2:07:20<64:49:31,  8.03s/it, lr: 1.0e-04 loss: 1.100e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 935/30000 [2:07:20<64:49:31,  8.03s/it, lr: 1.0e-04 loss: 1.100e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 936/30000 [2:07:20<61:33:49,  7.63s/it, lr: 1.0e-04 loss: 1.100e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 936/30000 [2:07:20<61:33:49,  7.63s/it, lr: 1.0e-04 loss: 1.100e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 936/30000 [2:07:29<61:33:49,  7.63s/it, lr: 1.0e-04 loss: 6.012e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 936/30000 [2:07:29<61:33:49,  7.63s/it, lr: 1.0e-04 loss: 6.012e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 937/30000 [2:07:29<65:25:59,  8.11s/it, lr: 1.0e-04 loss: 6.012e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 937/30000 [2:07:29<65:25:59,  8.11s/it, lr: 1.0e-04 loss: 6.012e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 937/30000 [2:07:39<65:25:59,  8.11s/it, lr: 1.0e-04 loss: 9.875e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 937/30000 [2:07:39<65:25:59,  8.11s/it, lr: 1.0e-04 loss: 9.875e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 938/30000 [2:07:39<68:23:31,  8.47s/it, lr: 1.0e-04 loss: 9.875e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 938/30000 [2:07:39<68:23:31,  8.47s/it, lr: 1.0e-04 loss: 9.875e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 938/30000 [2:07:48<68:23:31,  8.47s/it, lr: 1.0e-04 loss: 3.899e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 938/30000 [2:07:48<68:23:31,  8.47s/it, lr: 1.0e-04 loss: 3.899e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 939/30000 [2:07:48<70:29:17,  8.73s/it, lr: 1.0e-04 loss: 3.899e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 939/30000 [2:07:48<70:29:17,  8.73s/it, lr: 1.0e-04 loss: 3.899e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 939/30000 [2:07:55<70:29:17,  8.73s/it, lr: 1.0e-04 loss: 8.624e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 939/30000 [2:07:55<70:29:17,  8.73s/it, lr: 1.0e-04 loss: 8.624e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 940/30000 [2:08:04<70:29:08,  8.73s/it, lr: 1.0e-04 loss: 3.582e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 940/30000 [2:08:04<70:29:08,  8.73s/it, lr: 1.0e-04 loss: 3.582e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 941/30000 [2:08:04<55:24:33,  6.86s/it, lr: 1.0e-04 loss: 3.582e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 941/30000 [2:08:04<55:24:33,  6.86s/it, lr: 1.0e-04 loss: 3.582e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 941/30000 [2:08:11<55:24:33,  6.86s/it, lr: 1.0e-04 loss: 1.431e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 941/30000 [2:08:11<55:24:33,  6.86s/it, lr: 1.0e-04 loss: 1.431e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 942/30000 [2:08:11<55:05:24,  6.83s/it, lr: 1.0e-04 loss: 1.431e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 942/30000 [2:08:11<55:05:24,  6.83s/it, lr: 1.0e-04 loss: 1.431e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 942/30000 [2:08:20<55:05:24,  6.83s/it, lr: 1.0e-04 loss: 4.417e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 942/30000 [2:08:20<55:05:24,  6.83s/it, lr: 1.0e-04 loss: 4.417e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 943/30000 [2:08:20<60:20:29,  7.48s/it, lr: 1.0e-04 loss: 4.417e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 943/30000 [2:08:20<60:20:29,  7.48s/it, lr: 1.0e-04 loss: 4.417e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 943/30000 [2:08:30<60:20:29,  7.48s/it, lr: 1.0e-04 loss: 7.705e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 943/30000 [2:08:30<60:20:29,  7.48s/it, lr: 1.0e-04 loss: 7.705e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 944/30000 [2:08:30<64:32:53,  8.00s/it, lr: 1.0e-04 loss: 7.705e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 944/30000 [2:08:30<64:32:53,  8.00s/it, lr: 1.0e-04 loss: 7.705e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 944/30000 [2:08:39<64:32:53,  8.00s/it, lr: 1.0e-04 loss: 8.216e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 944/30000 [2:08:39<64:32:53,  8.00s/it, lr: 1.0e-04 loss: 8.216e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 945/30000 [2:08:39<67:38:33,  8.38s/it, lr: 1.0e-04 loss: 8.216e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 945/30000 [2:08:39<67:38:33,  8.38s/it, lr: 1.0e-04 loss: 8.216e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 945/30000 [2:08:46<67:38:33,  8.38s/it, lr: 1.0e-04 loss: 1.056e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 945/30000 [2:08:46<67:38:33,  8.38s/it, lr: 1.0e-04 loss: 1.056e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 946/30000 [2:08:46<63:33:44,  7.88s/it, lr: 1.0e-04 loss: 1.056e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 946/30000 [2:08:46<63:33:44,  7.88s/it, lr: 1.0e-04 loss: 1.056e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 946/30000 [2:08:55<63:33:44,  7.88s/it, lr: 1.0e-04 loss: 4.502e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 946/30000 [2:08:55<63:33:44,  7.88s/it, lr: 1.0e-04 loss: 4.502e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 947/30000 [2:08:55<67:38:28,  8.38s/it, lr: 1.0e-04 loss: 4.502e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 947/30000 [2:08:55<67:38:28,  8.38s/it, lr: 1.0e-04 loss: 4.502e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 947/30000 [2:09:05<67:38:28,  8.38s/it, lr: 1.0e-04 loss: 6.259e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 947/30000 [2:09:05<67:38:28,  8.38s/it, lr: 1.0e-04 loss: 6.259e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 948/30000 [2:09:05<69:56:15,  8.67s/it, lr: 1.0e-04 loss: 6.259e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 948/30000 [2:09:05<69:56:15,  8.67s/it, lr: 1.0e-04 loss: 6.259e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 948/30000 [2:09:11<69:56:15,  8.67s/it, lr: 1.0e-04 loss: 1.496e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 948/30000 [2:09:11<69:56:15,  8.67s/it, lr: 1.0e-04 loss: 1.496e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 949/30000 [2:09:11<64:56:50,  8.05s/it, lr: 1.0e-04 loss: 1.496e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 949/30000 [2:09:11<64:56:50,  8.05s/it, lr: 1.0e-04 loss: 1.496e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 949/30000 [2:09:18<64:56:50,  8.05s/it, lr: 1.0e-04 loss: 8.837e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 949/30000 [2:09:18<64:56:50,  8.05s/it, lr: 1.0e-04 loss: 8.837e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 950/30000 [2:09:24<64:56:42,  8.05s/it, lr: 1.0e-04 loss: 5.386e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 950/30000 [2:09:24<64:56:42,  8.05s/it, lr: 1.0e-04 loss: 5.386e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 951/30000 [2:09:24<47:11:32,  5.85s/it, lr: 1.0e-04 loss: 5.386e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 951/30000 [2:09:24<47:11:32,  5.85s/it, lr: 1.0e-04 loss: 5.386e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 951/30000 [2:09:33<47:11:32,  5.85s/it, lr: 1.0e-04 loss: 1.143e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 951/30000 [2:09:33<47:11:32,  5.85s/it, lr: 1.0e-04 loss: 1.143e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 952/30000 [2:09:34<54:08:04,  6.71s/it, lr: 1.0e-04 loss: 1.143e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 952/30000 [2:09:34<54:08:04,  6.71s/it, lr: 1.0e-04 loss: 1.143e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 952/30000 [2:09:43<54:08:04,  6.71s/it, lr: 1.0e-04 loss: 8.933e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 952/30000 [2:09:43<54:08:04,  6.71s/it, lr: 1.0e-04 loss: 8.933e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 953/30000 [2:09:43<59:44:29,  7.40s/it, lr: 1.0e-04 loss: 8.933e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 953/30000 [2:09:43<59:44:29,  7.40s/it, lr: 1.0e-04 loss: 8.933e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 953/30000 [2:09:52<59:44:29,  7.40s/it, lr: 1.0e-04 loss: 1.320e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 953/30000 [2:09:52<59:44:29,  7.40s/it, lr: 1.0e-04 loss: 1.320e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 954/30000 [2:09:52<64:14:52,  7.96s/it, lr: 1.0e-04 loss: 1.320e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 954/30000 [2:09:52<64:14:52,  7.96s/it, lr: 1.0e-04 loss: 1.320e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 954/30000 [2:10:02<64:14:52,  7.96s/it, lr: 1.0e-04 loss: 3.975e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 954/30000 [2:10:02<64:14:52,  7.96s/it, lr: 1.0e-04 loss: 3.975e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 955/30000 [2:10:02<67:48:53,  8.41s/it, lr: 1.0e-04 loss: 3.975e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 955/30000 [2:10:02<67:48:53,  8.41s/it, lr: 1.0e-04 loss: 3.975e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 955/30000 [2:10:09<67:48:53,  8.41s/it, lr: 1.0e-04 loss: 9.988e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 955/30000 [2:10:09<67:48:53,  8.41s/it, lr: 1.0e-04 loss: 9.988e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 956/30000 [2:10:09<63:47:56,  7.91s/it, lr: 1.0e-04 loss: 9.988e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 956/30000 [2:10:09<63:47:56,  7.91s/it, lr: 1.0e-04 loss: 9.988e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 956/30000 [2:10:18<63:47:56,  7.91s/it, lr: 1.0e-04 loss: 5.605e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 956/30000 [2:10:18<63:47:56,  7.91s/it, lr: 1.0e-04 loss: 5.605e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 957/30000 [2:10:18<67:02:31,  8.31s/it, lr: 1.0e-04 loss: 5.605e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 957/30000 [2:10:18<67:02:31,  8.31s/it, lr: 1.0e-04 loss: 5.605e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 957/30000 [2:10:27<67:02:31,  8.31s/it, lr: 1.0e-04 loss: 6.566e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 957/30000 [2:10:27<67:02:31,  8.31s/it, lr: 1.0e-04 loss: 6.566e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 958/30000 [2:10:27<69:28:46,  8.61s/it, lr: 1.0e-04 loss: 6.566e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 958/30000 [2:10:27<69:28:46,  8.61s/it, lr: 1.0e-04 loss: 6.566e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 958/30000 [2:10:34<69:28:46,  8.61s/it, lr: 1.0e-04 loss: 7.540e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 958/30000 [2:10:34<69:28:46,  8.61s/it, lr: 1.0e-04 loss: 7.540e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 959/30000 [2:10:34<64:50:52,  8.04s/it, lr: 1.0e-04 loss: 7.540e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 959/30000 [2:10:34<64:50:52,  8.04s/it, lr: 1.0e-04 loss: 7.540e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 959/30000 [2:10:40<64:50:52,  8.04s/it, lr: 1.0e-04 loss: 5.820e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 959/30000 [2:10:40<64:50:52,  8.04s/it, lr: 1.0e-04 loss: 5.820e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 960/30000 [2:10:50<64:50:44,  8.04s/it, lr: 1.0e-04 loss: 3.403e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 960/30000 [2:10:50<64:50:44,  8.04s/it, lr: 1.0e-04 loss: 3.403e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 961/30000 [2:10:50<52:21:08,  6.49s/it, lr: 1.0e-04 loss: 3.403e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 961/30000 [2:10:50<52:21:08,  6.49s/it, lr: 1.0e-04 loss: 3.403e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 961/30000 [2:10:57<52:21:08,  6.49s/it, lr: 1.0e-04 loss: 9.321e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 961/30000 [2:10:57<52:21:08,  6.49s/it, lr: 1.0e-04 loss: 9.321e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 962/30000 [2:10:57<53:05:51,  6.58s/it, lr: 1.0e-04 loss: 9.321e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 962/30000 [2:10:57<53:05:51,  6.58s/it, lr: 1.0e-04 loss: 9.321e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 962/30000 [2:11:06<53:05:51,  6.58s/it, lr: 1.0e-04 loss: 1.219e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 962/30000 [2:11:06<53:05:51,  6.58s/it, lr: 1.0e-04 loss: 1.219e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 963/30000 [2:11:06<58:50:26,  7.30s/it, lr: 1.0e-04 loss: 1.219e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 963/30000 [2:11:06<58:50:26,  7.30s/it, lr: 1.0e-04 loss: 1.219e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 963/30000 [2:11:13<58:50:26,  7.30s/it, lr: 1.0e-04 loss: 7.847e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 963/30000 [2:11:13<58:50:26,  7.30s/it, lr: 1.0e-04 loss: 7.847e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 964/30000 [2:11:13<57:26:37,  7.12s/it, lr: 1.0e-04 loss: 7.847e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 964/30000 [2:11:13<57:26:37,  7.12s/it, lr: 1.0e-04 loss: 7.847e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 964/30000 [2:11:19<57:26:37,  7.12s/it, lr: 1.0e-04 loss: 5.386e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 964/30000 [2:11:19<57:26:37,  7.12s/it, lr: 1.0e-04 loss: 5.386e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 965/30000 [2:11:19<56:24:17,  6.99s/it, lr: 1.0e-04 loss: 5.386e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 965/30000 [2:11:19<56:24:17,  6.99s/it, lr: 1.0e-04 loss: 5.386e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 965/30000 [2:11:29<56:24:17,  6.99s/it, lr: 1.0e-04 loss: 6.436e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 965/30000 [2:11:29<56:24:17,  6.99s/it, lr: 1.0e-04 loss: 6.436e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 966/30000 [2:11:29<61:54:45,  7.68s/it, lr: 1.0e-04 loss: 6.436e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 966/30000 [2:11:29<61:54:45,  7.68s/it, lr: 1.0e-04 loss: 6.436e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 966/30000 [2:11:38<61:54:45,  7.68s/it, lr: 1.0e-04 loss: 5.072e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 966/30000 [2:11:38<61:54:45,  7.68s/it, lr: 1.0e-04 loss: 5.072e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 967/30000 [2:11:38<65:56:30,  8.18s/it, lr: 1.0e-04 loss: 5.072e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 967/30000 [2:11:38<65:56:30,  8.18s/it, lr: 1.0e-04 loss: 5.072e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 967/30000 [2:11:47<65:56:30,  8.18s/it, lr: 1.0e-04 loss: 9.034e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 967/30000 [2:11:47<65:56:30,  8.18s/it, lr: 1.0e-04 loss: 9.034e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 968/30000 [2:11:47<68:47:43,  8.53s/it, lr: 1.0e-04 loss: 9.034e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 968/30000 [2:11:47<68:47:43,  8.53s/it, lr: 1.0e-04 loss: 9.034e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 968/30000 [2:11:54<68:47:43,  8.53s/it, lr: 1.0e-04 loss: 1.126e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 968/30000 [2:11:54<68:47:43,  8.53s/it, lr: 1.0e-04 loss: 1.126e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 969/30000 [2:11:54<64:59:36,  8.06s/it, lr: 1.0e-04 loss: 1.126e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 969/30000 [2:11:54<64:59:36,  8.06s/it, lr: 1.0e-04 loss: 1.126e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 969/30000 [2:12:01<64:59:36,  8.06s/it, lr: 1.0e-04 loss: 7.683e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 969/30000 [2:12:01<64:59:36,  8.06s/it, lr: 1.0e-04 loss: 7.683e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 970/30000 [2:12:10<64:59:28,  8.06s/it, lr: 1.0e-04 loss: 1.145e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 970/30000 [2:12:10<64:59:28,  8.06s/it, lr: 1.0e-04 loss: 1.145e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 971/30000 [2:12:10<52:32:59,  6.52s/it, lr: 1.0e-04 loss: 1.145e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 971/30000 [2:12:10<52:32:59,  6.52s/it, lr: 1.0e-04 loss: 1.145e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 971/30000 [2:12:20<52:32:59,  6.52s/it, lr: 1.0e-04 loss: 6.981e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 971/30000 [2:12:20<52:32:59,  6.52s/it, lr: 1.0e-04 loss: 6.981e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 972/30000 [2:12:20<58:18:35,  7.23s/it, lr: 1.0e-04 loss: 6.981e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 972/30000 [2:12:20<58:18:35,  7.23s/it, lr: 1.0e-04 loss: 6.981e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 972/30000 [2:12:29<58:18:35,  7.23s/it, lr: 1.0e-04 loss: 3.765e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 972/30000 [2:12:29<58:18:35,  7.23s/it, lr: 1.0e-04 loss: 3.765e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 973/30000 [2:12:29<62:46:04,  7.78s/it, lr: 1.0e-04 loss: 3.765e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 973/30000 [2:12:29<62:46:04,  7.78s/it, lr: 1.0e-04 loss: 3.765e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 973/30000 [2:12:36<62:46:04,  7.78s/it, lr: 1.0e-04 loss: 8.719e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 973/30000 [2:12:36<62:46:04,  7.78s/it, lr: 1.0e-04 loss: 8.719e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 974/30000 [2:12:36<60:19:43,  7.48s/it, lr: 1.0e-04 loss: 8.719e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 974/30000 [2:12:36<60:19:43,  7.48s/it, lr: 1.0e-04 loss: 8.719e-02]
Xiang_wan_2_2_lora:   3%|#2                                      | 974/30000 [2:12:45<60:19:43,  7.48s/it, lr: 1.0e-04 loss: 1.263e-01]
Xiang_wan_2_2_lora:   3%|#2                                      | 974/30000 [2:12:45<60:19:43,  7.48s/it, lr: 1.0e-04 loss: 1.263e-01]
Xiang_wan_2_2_lora:   3%|#3                                      | 975/30000 [2:12:45<64:30:20,  8.00s/it, lr: 1.0e-04 loss: 1.263e-01]
Xiang_wan_2_2_lora:   3%|#3                                      | 975/30000 [2:12:45<64:30:20,  8.00s/it, lr: 1.0e-04 loss: 1.263e-01]
Xiang_wan_2_2_lora:   3%|#3                                      | 975/30000 [2:12:55<64:30:20,  8.00s/it, lr: 1.0e-04 loss: 7.251e-02]
Xiang_wan_2_2_lora:   3%|#3                                      | 975/30000 [2:12:55<64:30:20,  8.00s/it, lr: 1.0e-04 loss: 7.251e-02]
Xiang_wan_2_2_lora:   3%|#3                                      | 976/30000 [2:12:55<67:42:36,  8.40s/it, lr: 1.0e-04 loss: 7.251e-02]
Xiang_wan_2_2_lora:   3%|#3                                      | 976/30000 [2:12:55<67:42:36,  8.40s/it, lr: 1.0e-04 loss: 7.251e-02]
Xiang_wan_2_2_lora:   3%|#3                                      | 976/30000 [2:13:04<67:42:36,  8.40s/it, lr: 1.0e-04 loss: 2.191e-01]
Xiang_wan_2_2_lora:   3%|#3                                      | 976/30000 [2:13:04<67:42:36,  8.40s/it, lr: 1.0e-04 loss: 2.191e-01]
Xiang_wan_2_2_lora:   3%|#3                                      | 977/30000 [2:13:04<70:23:51,  8.73s/it, lr: 1.0e-04 loss: 2.191e-01]
Xiang_wan_2_2_lora:   3%|#3                                      | 977/30000 [2:13:04<70:23:51,  8.73s/it, lr: 1.0e-04 loss: 2.191e-01]
Xiang_wan_2_2_lora:   3%|#3                                      | 977/30000 [2:13:11<70:23:51,  8.73s/it, lr: 1.0e-04 loss: 1.290e-01]
Xiang_wan_2_2_lora:   3%|#3                                      | 977/30000 [2:13:11<70:23:51,  8.73s/it, lr: 1.0e-04 loss: 1.290e-01]
Xiang_wan_2_2_lora:   3%|#3                                      | 978/30000 [2:13:11<65:22:34,  8.11s/it, lr: 1.0e-04 loss: 1.290e-01]
Xiang_wan_2_2_lora:   3%|#3                                      | 978/30000 [2:13:11<65:22:34,  8.11s/it, lr: 1.0e-04 loss: 1.290e-01]
Xiang_wan_2_2_lora:   3%|#3                                      | 978/30000 [2:13:20<65:22:34,  8.11s/it, lr: 1.0e-04 loss: 5.314e-02]
Xiang_wan_2_2_lora:   3%|#3                                      | 978/30000 [2:13:20<65:22:34,  8.11s/it, lr: 1.0e-04 loss: 5.314e-02]
Xiang_wan_2_2_lora:   3%|#3                                      | 979/30000 [2:13:20<68:23:44,  8.48s/it, lr: 1.0e-04 loss: 5.314e-02]
Xiang_wan_2_2_lora:   3%|#3                                      | 979/30000 [2:13:20<68:23:44,  8.48s/it, lr: 1.0e-04 loss: 5.314e-02]
Xiang_wan_2_2_lora:   3%|#3                                      | 979/30000 [2:13:29<68:23:44,  8.48s/it, lr: 1.0e-04 loss: 1.587e-01]
Xiang_wan_2_2_lora:   3%|#3                                      | 979/30000 [2:13:29<68:23:44,  8.48s/it, lr: 1.0e-04 loss: 1.587e-01]
Xiang_wan_2_2_lora:   3%|#3                                      | 980/30000 [2:13:36<68:23:36,  8.48s/it, lr: 1.0e-04 loss: 6.870e-02]
Xiang_wan_2_2_lora:   3%|#3                                      | 980/30000 [2:13:36<68:23:36,  8.48s/it, lr: 1.0e-04 loss: 6.870e-02]
Xiang_wan_2_2_lora:   3%|#3                                      | 981/30000 [2:13:36<49:16:13,  6.11s/it, lr: 1.0e-04 loss: 6.870e-02]
Xiang_wan_2_2_lora:   3%|#3                                      | 981/30000 [2:13:36<49:16:13,  6.11s/it, lr: 1.0e-04 loss: 6.870e-02]
Xiang_wan_2_2_lora:   3%|#3                                      | 981/30000 [2:13:43<49:16:13,  6.11s/it, lr: 1.0e-04 loss: 9.564e-02]
Xiang_wan_2_2_lora:   3%|#3                                      | 981/30000 [2:13:43<49:16:13,  6.11s/it, lr: 1.0e-04 loss: 9.564e-02]
Xiang_wan_2_2_lora:   3%|#3                                      | 982/30000 [2:13:43<50:13:02,  6.23s/it, lr: 1.0e-04 loss: 9.564e-02]
Xiang_wan_2_2_lora:   3%|#3                                      | 982/30000 [2:13:43<50:13:02,  6.23s/it, lr: 1.0e-04 loss: 9.564e-02]
Xiang_wan_2_2_lora:   3%|#3                                      | 982/30000 [2:13:52<50:13:02,  6.23s/it, lr: 1.0e-04 loss: 6.980e-02]
Xiang_wan_2_2_lora:   3%|#3                                      | 982/30000 [2:13:52<50:13:02,  6.23s/it, lr: 1.0e-04 loss: 6.980e-02]
Xiang_wan_2_2_lora:   3%|#3                                      | 983/30000 [2:13:52<56:42:03,  7.03s/it, lr: 1.0e-04 loss: 6.980e-02]
Xiang_wan_2_2_lora:   3%|#3                                      | 983/30000 [2:13:52<56:42:03,  7.03s/it, lr: 1.0e-04 loss: 6.980e-02]
Xiang_wan_2_2_lora:   3%|#3                                      | 983/30000 [2:14:02<56:42:03,  7.03s/it, lr: 1.0e-04 loss: 6.706e-02]
Xiang_wan_2_2_lora:   3%|#3                                      | 983/30000 [2:14:02<56:42:03,  7.03s/it, lr: 1.0e-04 loss: 6.706e-02]
Xiang_wan_2_2_lora:   3%|#3                                      | 984/30000 [2:14:02<62:21:07,  7.74s/it, lr: 1.0e-04 loss: 6.706e-02]
Xiang_wan_2_2_lora:   3%|#3                                      | 984/30000 [2:14:02<62:21:07,  7.74s/it, lr: 1.0e-04 loss: 6.706e-02]
Xiang_wan_2_2_lora:   3%|#3                                      | 984/30000 [2:14:11<62:21:07,  7.74s/it, lr: 1.0e-04 loss: 1.258e-01]
Xiang_wan_2_2_lora:   3%|#3                                      | 984/30000 [2:14:11<62:21:07,  7.74s/it, lr: 1.0e-04 loss: 1.258e-01]
Xiang_wan_2_2_lora:   3%|#3                                      | 985/30000 [2:14:11<66:06:41,  8.20s/it, lr: 1.0e-04 loss: 1.258e-01]
Xiang_wan_2_2_lora:   3%|#3                                      | 985/30000 [2:14:11<66:06:41,  8.20s/it, lr: 1.0e-04 loss: 1.258e-01]
Xiang_wan_2_2_lora:   3%|#3                                      | 985/30000 [2:14:18<66:06:41,  8.20s/it, lr: 1.0e-04 loss: 9.450e-02]
Xiang_wan_2_2_lora:   3%|#3                                      | 985/30000 [2:14:18<66:06:41,  8.20s/it, lr: 1.0e-04 loss: 9.450e-02]
Xiang_wan_2_2_lora:   3%|#3                                      | 986/30000 [2:14:18<62:29:05,  7.75s/it, lr: 1.0e-04 loss: 9.450e-02]
Xiang_wan_2_2_lora:   3%|#3                                      | 986/30000 [2:14:18<62:29:05,  7.75s/it, lr: 1.0e-04 loss: 9.450e-02]
Xiang_wan_2_2_lora:   3%|#3                                      | 986/30000 [2:14:24<62:29:05,  7.75s/it, lr: 1.0e-04 loss: 6.702e-01]
Xiang_wan_2_2_lora:   3%|#3                                      | 986/30000 [2:14:24<62:29:05,  7.75s/it, lr: 1.0e-04 loss: 6.702e-01]
Xiang_wan_2_2_lora:   3%|#3                                      | 987/30000 [2:14:24<59:47:10,  7.42s/it, lr: 1.0e-04 loss: 6.702e-01]
Xiang_wan_2_2_lora:   3%|#3                                      | 987/30000 [2:14:24<59:47:10,  7.42s/it, lr: 1.0e-04 loss: 6.702e-01]
Xiang_wan_2_2_lora:   3%|#3                                      | 987/30000 [2:14:31<59:47:10,  7.42s/it, lr: 1.0e-04 loss: 1.057e-01]
Xiang_wan_2_2_lora:   3%|#3                                      | 987/30000 [2:14:31<59:47:10,  7.42s/it, lr: 1.0e-04 loss: 1.057e-01]
Xiang_wan_2_2_lora:   3%|#3                                      | 988/30000 [2:14:31<57:45:54,  7.17s/it, lr: 1.0e-04 loss: 1.057e-01]
Xiang_wan_2_2_lora:   3%|#3                                      | 988/30000 [2:14:31<57:45:54,  7.17s/it, lr: 1.0e-04 loss: 1.057e-01]
Xiang_wan_2_2_lora:   3%|#3                                      | 988/30000 [2:14:37<57:45:54,  7.17s/it, lr: 1.0e-04 loss: 7.606e-02]
Xiang_wan_2_2_lora:   3%|#3                                      | 988/30000 [2:14:37<57:45:54,  7.17s/it, lr: 1.0e-04 loss: 7.606e-02]
Xiang_wan_2_2_lora:   3%|#3                                      | 989/30000 [2:14:37<56:31:57,  7.02s/it, lr: 1.0e-04 loss: 7.606e-02]
Xiang_wan_2_2_lora:   3%|#3                                      | 989/30000 [2:14:37<56:31:57,  7.02s/it, lr: 1.0e-04 loss: 7.606e-02]
Xiang_wan_2_2_lora:   3%|#3                                      | 989/30000 [2:14:47<56:31:57,  7.02s/it, lr: 1.0e-04 loss: 8.640e-02]
Xiang_wan_2_2_lora:   3%|#3                                      | 989/30000 [2:14:47<56:31:57,  7.02s/it, lr: 1.0e-04 loss: 8.640e-02]
Xiang_wan_2_2_lora:   3%|#3                                      | 990/30000 [2:14:53<56:31:50,  7.02s/it, lr: 1.0e-04 loss: 1.944e-01]
Xiang_wan_2_2_lora:   3%|#3                                      | 990/30000 [2:14:53<56:31:50,  7.02s/it, lr: 1.0e-04 loss: 1.944e-01]
Xiang_wan_2_2_lora:   3%|#3                                      | 991/30000 [2:14:53<42:56:56,  5.33s/it, lr: 1.0e-04 loss: 1.944e-01]
Xiang_wan_2_2_lora:   3%|#3                                      | 991/30000 [2:14:53<42:56:56,  5.33s/it, lr: 1.0e-04 loss: 1.944e-01]
Xiang_wan_2_2_lora:   3%|#3                                      | 991/30000 [2:15:00<42:56:56,  5.33s/it, lr: 1.0e-04 loss: 9.743e-02]
Xiang_wan_2_2_lora:   3%|#3                                      | 991/30000 [2:15:00<42:56:56,  5.33s/it, lr: 1.0e-04 loss: 9.743e-02]
Xiang_wan_2_2_lora:   3%|#3                                      | 992/30000 [2:15:00<45:57:11,  5.70s/it, lr: 1.0e-04 loss: 9.743e-02]
Xiang_wan_2_2_lora:   3%|#3                                      | 992/30000 [2:15:00<45:57:11,  5.70s/it, lr: 1.0e-04 loss: 9.743e-02]
Xiang_wan_2_2_lora:   3%|#3                                      | 992/30000 [2:15:10<45:57:11,  5.70s/it, lr: 1.0e-04 loss: 8.319e-02]
Xiang_wan_2_2_lora:   3%|#3                                      | 992/30000 [2:15:10<45:57:11,  5.70s/it, lr: 1.0e-04 loss: 8.319e-02]
Xiang_wan_2_2_lora:   3%|#3                                      | 993/30000 [2:15:10<53:38:47,  6.66s/it, lr: 1.0e-04 loss: 8.319e-02]
Xiang_wan_2_2_lora:   3%|#3                                      | 993/30000 [2:15:10<53:38:47,  6.66s/it, lr: 1.0e-04 loss: 8.319e-02]
Xiang_wan_2_2_lora:   3%|#3                                      | 993/30000 [2:15:16<53:38:47,  6.66s/it, lr: 1.0e-04 loss: 1.038e-01]
Xiang_wan_2_2_lora:   3%|#3                                      | 993/30000 [2:15:16<53:38:47,  6.66s/it, lr: 1.0e-04 loss: 1.038e-01]
Xiang_wan_2_2_lora:   3%|#3                                      | 994/30000 [2:15:16<53:26:31,  6.63s/it, lr: 1.0e-04 loss: 1.038e-01]
Xiang_wan_2_2_lora:   3%|#3                                      | 994/30000 [2:15:16<53:26:31,  6.63s/it, lr: 1.0e-04 loss: 1.038e-01]
Xiang_wan_2_2_lora:   3%|#3                                      | 994/30000 [2:15:26<53:26:31,  6.63s/it, lr: 1.0e-04 loss: 5.697e-02]
Xiang_wan_2_2_lora:   3%|#3                                      | 994/30000 [2:15:26<53:26:31,  6.63s/it, lr: 1.0e-04 loss: 5.697e-02]
Xiang_wan_2_2_lora:   3%|#3                                      | 995/30000 [2:15:26<59:42:55,  7.41s/it, lr: 1.0e-04 loss: 5.697e-02]
Xiang_wan_2_2_lora:   3%|#3                                      | 995/30000 [2:15:26<59:42:55,  7.41s/it, lr: 1.0e-04 loss: 5.697e-02]
Xiang_wan_2_2_lora:   3%|#3                                      | 995/30000 [2:15:35<59:42:55,  7.41s/it, lr: 1.0e-04 loss: 4.989e-02]
Xiang_wan_2_2_lora:   3%|#3                                      | 995/30000 [2:15:35<59:42:55,  7.41s/it, lr: 1.0e-04 loss: 4.989e-02]
Xiang_wan_2_2_lora:   3%|#3                                      | 996/30000 [2:15:35<64:09:29,  7.96s/it, lr: 1.0e-04 loss: 4.989e-02]
Xiang_wan_2_2_lora:   3%|#3                                      | 996/30000 [2:15:35<64:09:29,  7.96s/it, lr: 1.0e-04 loss: 4.989e-02]
Xiang_wan_2_2_lora:   3%|#3                                      | 996/30000 [2:15:44<64:09:29,  7.96s/it, lr: 1.0e-04 loss: 5.725e-02]
Xiang_wan_2_2_lora:   3%|#3                                      | 996/30000 [2:15:44<64:09:29,  7.96s/it, lr: 1.0e-04 loss: 5.725e-02]
Xiang_wan_2_2_lora:   3%|#3                                      | 997/30000 [2:15:44<67:40:05,  8.40s/it, lr: 1.0e-04 loss: 5.725e-02]
Xiang_wan_2_2_lora:   3%|#3                                      | 997/30000 [2:15:44<67:40:05,  8.40s/it, lr: 1.0e-04 loss: 5.725e-02]
Xiang_wan_2_2_lora:   3%|#3                                      | 997/30000 [2:15:51<67:40:05,  8.40s/it, lr: 1.0e-04 loss: 1.043e-01]
Xiang_wan_2_2_lora:   3%|#3                                      | 997/30000 [2:15:51<67:40:05,  8.40s/it, lr: 1.0e-04 loss: 1.043e-01]
Xiang_wan_2_2_lora:   3%|#3                                      | 998/30000 [2:15:51<63:28:11,  7.88s/it, lr: 1.0e-04 loss: 1.043e-01]
Xiang_wan_2_2_lora:   3%|#3                                      | 998/30000 [2:15:51<63:28:11,  7.88s/it, lr: 1.0e-04 loss: 1.043e-01]
Xiang_wan_2_2_lora:   3%|#3                                      | 998/30000 [2:15:58<63:28:11,  7.88s/it, lr: 1.0e-04 loss: 4.521e-02]
Xiang_wan_2_2_lora:   3%|#3                                      | 998/30000 [2:15:58<63:28:11,  7.88s/it, lr: 1.0e-04 loss: 4.521e-02]
Xiang_wan_2_2_lora:   3%|#3                                      | 999/30000 [2:15:58<61:10:24,  7.59s/it, lr: 1.0e-04 loss: 4.521e-02]
Xiang_wan_2_2_lora:   3%|#3                                      | 999/30000 [2:15:58<61:10:24,  7.59s/it, lr: 1.0e-04 loss: 4.521e-02]
Xiang_wan_2_2_lora:   3%|#3                                      | 999/30000 [2:16:07<61:10:24,  7.59s/it, lr: 1.0e-04 loss: 7.729e-02]
Xiang_wan_2_2_lora:   3%|#3                                      | 999/30000 [2:16:07<61:10:24,  7.59s/it, lr: 1.0e-04 loss: 7.729e-02]
+Saving at step 1000
+Saved checkpoint to /home/featurize/ai-toolkit/output/Xiang_wan_2_2_lora/Xiang_wan_2_2_lora_000001000.safetensors
+Saved optimizer to /home/featurize/ai-toolkit/output/Xiang_wan_2_2_lora/optimizer.pt
+
Xiang_wan_2_2_lora:   3%|#3                                     | 1000/30000 [2:16:14<61:10:17,  7.59s/it, lr: 1.0e-04 loss: 5.575e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1000/30000 [2:16:14<61:10:17,  7.59s/it, lr: 1.0e-04 loss: 5.575e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1001/30000 [2:16:14<45:55:43,  5.70s/it, lr: 1.0e-04 loss: 5.575e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1001/30000 [2:16:14<45:55:43,  5.70s/it, lr: 1.0e-04 loss: 5.575e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1001/30000 [2:16:21<45:55:43,  5.70s/it, lr: 1.0e-04 loss: 1.126e-01]
Xiang_wan_2_2_lora:   3%|#3                                     | 1001/30000 [2:16:21<45:55:43,  5.70s/it, lr: 1.0e-04 loss: 1.126e-01]
Xiang_wan_2_2_lora:   3%|#3                                     | 1002/30000 [2:16:21<47:38:01,  5.91s/it, lr: 1.0e-04 loss: 1.126e-01]
Xiang_wan_2_2_lora:   3%|#3                                     | 1002/30000 [2:16:21<47:38:01,  5.91s/it, lr: 1.0e-04 loss: 1.126e-01]
Xiang_wan_2_2_lora:   3%|#3                                     | 1002/30000 [2:16:30<47:38:01,  5.91s/it, lr: 1.0e-04 loss: 5.206e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1002/30000 [2:16:30<47:38:01,  5.91s/it, lr: 1.0e-04 loss: 5.206e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1003/30000 [2:16:30<54:50:25,  6.81s/it, lr: 1.0e-04 loss: 5.206e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1003/30000 [2:16:30<54:50:25,  6.81s/it, lr: 1.0e-04 loss: 5.206e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1003/30000 [2:16:39<54:50:25,  6.81s/it, lr: 1.0e-04 loss: 6.096e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1003/30000 [2:16:39<54:50:25,  6.81s/it, lr: 1.0e-04 loss: 6.096e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1004/30000 [2:16:39<60:17:56,  7.49s/it, lr: 1.0e-04 loss: 6.096e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1004/30000 [2:16:39<60:17:56,  7.49s/it, lr: 1.0e-04 loss: 6.096e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1004/30000 [2:16:46<60:17:56,  7.49s/it, lr: 1.0e-04 loss: 6.165e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1004/30000 [2:16:46<60:17:56,  7.49s/it, lr: 1.0e-04 loss: 6.165e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1005/30000 [2:16:46<58:22:53,  7.25s/it, lr: 1.0e-04 loss: 6.165e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1005/30000 [2:16:46<58:22:53,  7.25s/it, lr: 1.0e-04 loss: 6.165e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1005/30000 [2:16:53<58:22:53,  7.25s/it, lr: 1.0e-04 loss: 1.066e-01]
Xiang_wan_2_2_lora:   3%|#3                                     | 1005/30000 [2:16:53<58:22:53,  7.25s/it, lr: 1.0e-04 loss: 1.066e-01]
Xiang_wan_2_2_lora:   3%|#3                                     | 1006/30000 [2:16:53<56:57:13,  7.07s/it, lr: 1.0e-04 loss: 1.066e-01]
Xiang_wan_2_2_lora:   3%|#3                                     | 1006/30000 [2:16:53<56:57:13,  7.07s/it, lr: 1.0e-04 loss: 1.066e-01]
Xiang_wan_2_2_lora:   3%|#3                                     | 1006/30000 [2:16:59<56:57:13,  7.07s/it, lr: 1.0e-04 loss: 5.118e-01]
Xiang_wan_2_2_lora:   3%|#3                                     | 1006/30000 [2:16:59<56:57:13,  7.07s/it, lr: 1.0e-04 loss: 5.118e-01]
Xiang_wan_2_2_lora:   3%|#3                                     | 1007/30000 [2:16:59<56:02:53,  6.96s/it, lr: 1.0e-04 loss: 5.118e-01]
Xiang_wan_2_2_lora:   3%|#3                                     | 1007/30000 [2:16:59<56:02:53,  6.96s/it, lr: 1.0e-04 loss: 5.118e-01]
Xiang_wan_2_2_lora:   3%|#3                                     | 1007/30000 [2:17:09<56:02:53,  6.96s/it, lr: 1.0e-04 loss: 9.266e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1007/30000 [2:17:09<56:02:53,  6.96s/it, lr: 1.0e-04 loss: 9.266e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1008/30000 [2:17:09<62:02:07,  7.70s/it, lr: 1.0e-04 loss: 9.266e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1008/30000 [2:17:09<62:02:07,  7.70s/it, lr: 1.0e-04 loss: 9.266e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1008/30000 [2:17:15<62:02:07,  7.70s/it, lr: 1.0e-04 loss: 7.416e-01]
Xiang_wan_2_2_lora:   3%|#3                                     | 1008/30000 [2:17:15<62:02:07,  7.70s/it, lr: 1.0e-04 loss: 7.416e-01]
Xiang_wan_2_2_lora:   3%|#3                                     | 1009/30000 [2:17:16<59:30:33,  7.39s/it, lr: 1.0e-04 loss: 7.416e-01]
Xiang_wan_2_2_lora:   3%|#3                                     | 1009/30000 [2:17:16<59:30:33,  7.39s/it, lr: 1.0e-04 loss: 7.416e-01]
Xiang_wan_2_2_lora:   3%|#3                                     | 1009/30000 [2:17:25<59:30:33,  7.39s/it, lr: 1.0e-04 loss: 1.066e-01]
Xiang_wan_2_2_lora:   3%|#3                                     | 1009/30000 [2:17:25<59:30:33,  7.39s/it, lr: 1.0e-04 loss: 1.066e-01]
Xiang_wan_2_2_lora:   3%|#3                                     | 1010/30000 [2:17:34<59:30:26,  7.39s/it, lr: 1.0e-04 loss: 4.487e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1010/30000 [2:17:34<59:30:26,  7.39s/it, lr: 1.0e-04 loss: 4.487e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1011/30000 [2:17:34<49:37:11,  6.16s/it, lr: 1.0e-04 loss: 4.487e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1011/30000 [2:17:34<49:37:11,  6.16s/it, lr: 1.0e-04 loss: 4.487e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1011/30000 [2:17:44<49:37:11,  6.16s/it, lr: 1.0e-04 loss: 7.612e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1011/30000 [2:17:44<49:37:11,  6.16s/it, lr: 1.0e-04 loss: 7.612e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1012/30000 [2:17:44<55:57:42,  6.95s/it, lr: 1.0e-04 loss: 7.612e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1012/30000 [2:17:44<55:57:42,  6.95s/it, lr: 1.0e-04 loss: 7.612e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1012/30000 [2:17:53<55:57:42,  6.95s/it, lr: 1.0e-04 loss: 3.708e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1012/30000 [2:17:53<55:57:42,  6.95s/it, lr: 1.0e-04 loss: 3.708e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1013/30000 [2:17:53<61:07:42,  7.59s/it, lr: 1.0e-04 loss: 3.708e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1013/30000 [2:17:53<61:07:42,  7.59s/it, lr: 1.0e-04 loss: 3.708e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1013/30000 [2:18:02<61:07:42,  7.59s/it, lr: 1.0e-04 loss: 6.221e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1013/30000 [2:18:02<61:07:42,  7.59s/it, lr: 1.0e-04 loss: 6.221e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1014/30000 [2:18:02<65:06:33,  8.09s/it, lr: 1.0e-04 loss: 6.221e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1014/30000 [2:18:02<65:06:33,  8.09s/it, lr: 1.0e-04 loss: 6.221e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1014/30000 [2:18:12<65:06:33,  8.09s/it, lr: 1.0e-04 loss: 1.782e-01]
Xiang_wan_2_2_lora:   3%|#3                                     | 1014/30000 [2:18:12<65:06:33,  8.09s/it, lr: 1.0e-04 loss: 1.782e-01]
Xiang_wan_2_2_lora:   3%|#3                                     | 1015/30000 [2:18:12<68:23:29,  8.49s/it, lr: 1.0e-04 loss: 1.782e-01]
Xiang_wan_2_2_lora:   3%|#3                                     | 1015/30000 [2:18:12<68:23:29,  8.49s/it, lr: 1.0e-04 loss: 1.782e-01]
Xiang_wan_2_2_lora:   3%|#3                                     | 1015/30000 [2:18:19<68:23:29,  8.49s/it, lr: 1.0e-04 loss: 7.863e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1015/30000 [2:18:19<68:23:29,  8.49s/it, lr: 1.0e-04 loss: 7.863e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1016/30000 [2:18:19<64:16:24,  7.98s/it, lr: 1.0e-04 loss: 7.863e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1016/30000 [2:18:19<64:16:24,  7.98s/it, lr: 1.0e-04 loss: 7.863e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1016/30000 [2:18:28<64:16:24,  7.98s/it, lr: 1.0e-04 loss: 4.170e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1016/30000 [2:18:28<64:16:24,  7.98s/it, lr: 1.0e-04 loss: 4.170e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1017/30000 [2:18:28<67:22:26,  8.37s/it, lr: 1.0e-04 loss: 4.170e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1017/30000 [2:18:28<67:22:26,  8.37s/it, lr: 1.0e-04 loss: 4.170e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1017/30000 [2:18:37<67:22:26,  8.37s/it, lr: 1.0e-04 loss: 9.550e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1017/30000 [2:18:37<67:22:26,  8.37s/it, lr: 1.0e-04 loss: 9.550e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1018/30000 [2:18:37<69:48:41,  8.67s/it, lr: 1.0e-04 loss: 9.550e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1018/30000 [2:18:37<69:48:41,  8.67s/it, lr: 1.0e-04 loss: 9.550e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1018/30000 [2:18:44<69:48:41,  8.67s/it, lr: 1.0e-04 loss: 7.324e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1018/30000 [2:18:44<69:48:41,  8.67s/it, lr: 1.0e-04 loss: 7.324e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1019/30000 [2:18:44<64:49:00,  8.05s/it, lr: 1.0e-04 loss: 7.324e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1019/30000 [2:18:44<64:49:00,  8.05s/it, lr: 1.0e-04 loss: 7.324e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1019/30000 [2:18:53<64:49:00,  8.05s/it, lr: 1.0e-04 loss: 7.805e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1019/30000 [2:18:53<64:49:00,  8.05s/it, lr: 1.0e-04 loss: 7.805e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1020/30000 [2:19:03<64:48:52,  8.05s/it, lr: 1.0e-04 loss: 5.723e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1020/30000 [2:19:03<64:48:52,  8.05s/it, lr: 1.0e-04 loss: 5.723e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1021/30000 [2:19:03<52:36:50,  6.54s/it, lr: 1.0e-04 loss: 5.723e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1021/30000 [2:19:03<52:36:50,  6.54s/it, lr: 1.0e-04 loss: 5.723e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1021/30000 [2:19:10<52:36:50,  6.54s/it, lr: 1.0e-04 loss: 8.875e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1021/30000 [2:19:10<52:36:50,  6.54s/it, lr: 1.0e-04 loss: 8.875e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1022/30000 [2:19:10<53:28:56,  6.64s/it, lr: 1.0e-04 loss: 8.875e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1022/30000 [2:19:10<53:28:56,  6.64s/it, lr: 1.0e-04 loss: 8.875e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1022/30000 [2:19:19<53:28:56,  6.64s/it, lr: 1.0e-04 loss: 3.788e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1022/30000 [2:19:19<53:28:56,  6.64s/it, lr: 1.0e-04 loss: 3.788e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1023/30000 [2:19:19<59:01:56,  7.33s/it, lr: 1.0e-04 loss: 3.788e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1023/30000 [2:19:19<59:01:56,  7.33s/it, lr: 1.0e-04 loss: 3.788e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1023/30000 [2:19:28<59:01:56,  7.33s/it, lr: 1.0e-04 loss: 6.207e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1023/30000 [2:19:28<59:01:56,  7.33s/it, lr: 1.0e-04 loss: 6.207e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1024/30000 [2:19:28<63:29:40,  7.89s/it, lr: 1.0e-04 loss: 6.207e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1024/30000 [2:19:28<63:29:40,  7.89s/it, lr: 1.0e-04 loss: 6.207e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1024/30000 [2:19:38<63:29:40,  7.89s/it, lr: 1.0e-04 loss: 5.119e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1024/30000 [2:19:38<63:29:40,  7.89s/it, lr: 1.0e-04 loss: 5.119e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1025/30000 [2:19:38<66:59:23,  8.32s/it, lr: 1.0e-04 loss: 5.119e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1025/30000 [2:19:38<66:59:23,  8.32s/it, lr: 1.0e-04 loss: 5.119e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1025/30000 [2:19:45<66:59:23,  8.32s/it, lr: 1.0e-04 loss: 8.564e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1025/30000 [2:19:45<66:59:23,  8.32s/it, lr: 1.0e-04 loss: 8.564e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1026/30000 [2:19:45<63:00:00,  7.83s/it, lr: 1.0e-04 loss: 8.564e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1026/30000 [2:19:45<63:00:00,  7.83s/it, lr: 1.0e-04 loss: 8.564e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1026/30000 [2:19:51<63:00:00,  7.83s/it, lr: 1.0e-04 loss: 1.605e-01]
Xiang_wan_2_2_lora:   3%|#3                                     | 1026/30000 [2:19:51<63:00:00,  7.83s/it, lr: 1.0e-04 loss: 1.605e-01]
Xiang_wan_2_2_lora:   3%|#3                                     | 1027/30000 [2:19:51<60:07:10,  7.47s/it, lr: 1.0e-04 loss: 1.605e-01]
Xiang_wan_2_2_lora:   3%|#3                                     | 1027/30000 [2:19:51<60:07:10,  7.47s/it, lr: 1.0e-04 loss: 1.605e-01]
Xiang_wan_2_2_lora:   3%|#3                                     | 1027/30000 [2:20:00<60:07:10,  7.47s/it, lr: 1.0e-04 loss: 5.479e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1027/30000 [2:20:00<60:07:10,  7.47s/it, lr: 1.0e-04 loss: 5.479e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1028/30000 [2:20:00<64:24:20,  8.00s/it, lr: 1.0e-04 loss: 5.479e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1028/30000 [2:20:00<64:24:20,  8.00s/it, lr: 1.0e-04 loss: 5.479e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1028/30000 [2:20:07<64:24:20,  8.00s/it, lr: 1.0e-04 loss: 6.743e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1028/30000 [2:20:07<64:24:20,  8.00s/it, lr: 1.0e-04 loss: 6.743e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1029/30000 [2:20:07<61:03:37,  7.59s/it, lr: 1.0e-04 loss: 6.743e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1029/30000 [2:20:07<61:03:37,  7.59s/it, lr: 1.0e-04 loss: 6.743e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1029/30000 [2:20:16<61:03:37,  7.59s/it, lr: 1.0e-04 loss: 9.523e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1029/30000 [2:20:16<61:03:37,  7.59s/it, lr: 1.0e-04 loss: 9.523e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1030/30000 [2:20:23<61:03:30,  7.59s/it, lr: 1.0e-04 loss: 6.491e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1030/30000 [2:20:23<61:03:30,  7.59s/it, lr: 1.0e-04 loss: 6.491e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1031/30000 [2:20:23<45:20:33,  5.63s/it, lr: 1.0e-04 loss: 6.491e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1031/30000 [2:20:23<45:20:33,  5.63s/it, lr: 1.0e-04 loss: 6.491e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1031/30000 [2:20:30<45:20:33,  5.63s/it, lr: 1.0e-04 loss: 7.300e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1031/30000 [2:20:30<45:20:33,  5.63s/it, lr: 1.0e-04 loss: 7.300e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1032/30000 [2:20:30<47:14:30,  5.87s/it, lr: 1.0e-04 loss: 7.300e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1032/30000 [2:20:30<47:14:30,  5.87s/it, lr: 1.0e-04 loss: 7.300e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1032/30000 [2:20:39<47:14:30,  5.87s/it, lr: 1.0e-04 loss: 2.333e-01]
Xiang_wan_2_2_lora:   3%|#3                                     | 1032/30000 [2:20:39<47:14:30,  5.87s/it, lr: 1.0e-04 loss: 2.333e-01]
Xiang_wan_2_2_lora:   3%|#3                                     | 1033/30000 [2:20:39<54:32:08,  6.78s/it, lr: 1.0e-04 loss: 2.333e-01]
Xiang_wan_2_2_lora:   3%|#3                                     | 1033/30000 [2:20:39<54:32:08,  6.78s/it, lr: 1.0e-04 loss: 2.333e-01]
Xiang_wan_2_2_lora:   3%|#3                                     | 1033/30000 [2:20:48<54:32:08,  6.78s/it, lr: 1.0e-04 loss: 1.205e-01]
Xiang_wan_2_2_lora:   3%|#3                                     | 1033/30000 [2:20:48<54:32:08,  6.78s/it, lr: 1.0e-04 loss: 1.205e-01]
Xiang_wan_2_2_lora:   3%|#3                                     | 1034/30000 [2:20:48<60:12:29,  7.48s/it, lr: 1.0e-04 loss: 1.205e-01]
Xiang_wan_2_2_lora:   3%|#3                                     | 1034/30000 [2:20:48<60:12:29,  7.48s/it, lr: 1.0e-04 loss: 1.205e-01]
Xiang_wan_2_2_lora:   3%|#3                                     | 1034/30000 [2:20:58<60:12:29,  7.48s/it, lr: 1.0e-04 loss: 3.477e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1034/30000 [2:20:58<60:12:29,  7.48s/it, lr: 1.0e-04 loss: 3.477e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1035/30000 [2:20:58<64:26:44,  8.01s/it, lr: 1.0e-04 loss: 3.477e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1035/30000 [2:20:58<64:26:44,  8.01s/it, lr: 1.0e-04 loss: 3.477e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1035/30000 [2:21:07<64:26:44,  8.01s/it, lr: 1.0e-04 loss: 7.461e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1035/30000 [2:21:07<64:26:44,  8.01s/it, lr: 1.0e-04 loss: 7.461e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1036/30000 [2:21:07<67:34:11,  8.40s/it, lr: 1.0e-04 loss: 7.461e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1036/30000 [2:21:07<67:34:11,  8.40s/it, lr: 1.0e-04 loss: 7.461e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1036/30000 [2:21:17<67:34:11,  8.40s/it, lr: 1.0e-04 loss: 4.996e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1036/30000 [2:21:17<67:34:11,  8.40s/it, lr: 1.0e-04 loss: 4.996e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1037/30000 [2:21:17<70:15:27,  8.73s/it, lr: 1.0e-04 loss: 4.996e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1037/30000 [2:21:17<70:15:27,  8.73s/it, lr: 1.0e-04 loss: 4.996e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1037/30000 [2:21:23<70:15:27,  8.73s/it, lr: 1.0e-04 loss: 1.045e-01]
Xiang_wan_2_2_lora:   3%|#3                                     | 1037/30000 [2:21:23<70:15:27,  8.73s/it, lr: 1.0e-04 loss: 1.045e-01]
Xiang_wan_2_2_lora:   3%|#3                                     | 1038/30000 [2:21:23<65:14:31,  8.11s/it, lr: 1.0e-04 loss: 1.045e-01]
Xiang_wan_2_2_lora:   3%|#3                                     | 1038/30000 [2:21:23<65:14:31,  8.11s/it, lr: 1.0e-04 loss: 1.045e-01]
Xiang_wan_2_2_lora:   3%|#3                                     | 1038/30000 [2:21:33<65:14:31,  8.11s/it, lr: 1.0e-04 loss: 2.948e-01]
Xiang_wan_2_2_lora:   3%|#3                                     | 1038/30000 [2:21:33<65:14:31,  8.11s/it, lr: 1.0e-04 loss: 2.948e-01]
Xiang_wan_2_2_lora:   3%|#3                                     | 1039/30000 [2:21:33<68:13:46,  8.48s/it, lr: 1.0e-04 loss: 2.948e-01]
Xiang_wan_2_2_lora:   3%|#3                                     | 1039/30000 [2:21:33<68:13:46,  8.48s/it, lr: 1.0e-04 loss: 2.948e-01]
Xiang_wan_2_2_lora:   3%|#3                                     | 1039/30000 [2:21:42<68:13:46,  8.48s/it, lr: 1.0e-04 loss: 7.689e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1039/30000 [2:21:42<68:13:46,  8.48s/it, lr: 1.0e-04 loss: 7.689e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1040/30000 [2:21:51<68:13:38,  8.48s/it, lr: 1.0e-04 loss: 5.206e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1040/30000 [2:21:51<68:13:38,  8.48s/it, lr: 1.0e-04 loss: 5.206e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1041/30000 [2:21:51<54:15:49,  6.75s/it, lr: 1.0e-04 loss: 5.206e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1041/30000 [2:21:51<54:15:49,  6.75s/it, lr: 1.0e-04 loss: 5.206e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1041/30000 [2:22:01<54:15:49,  6.75s/it, lr: 1.0e-04 loss: 6.763e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1041/30000 [2:22:01<54:15:49,  6.75s/it, lr: 1.0e-04 loss: 6.763e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1042/30000 [2:22:01<59:24:05,  7.38s/it, lr: 1.0e-04 loss: 6.763e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1042/30000 [2:22:01<59:24:05,  7.38s/it, lr: 1.0e-04 loss: 6.763e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1042/30000 [2:22:10<59:24:05,  7.38s/it, lr: 1.0e-04 loss: 7.219e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1042/30000 [2:22:10<59:24:05,  7.38s/it, lr: 1.0e-04 loss: 7.219e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1043/30000 [2:22:10<63:44:05,  7.92s/it, lr: 1.0e-04 loss: 7.219e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1043/30000 [2:22:10<63:44:05,  7.92s/it, lr: 1.0e-04 loss: 7.219e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1043/30000 [2:22:17<63:44:05,  7.92s/it, lr: 1.0e-04 loss: 7.085e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1043/30000 [2:22:17<63:44:05,  7.92s/it, lr: 1.0e-04 loss: 7.085e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1044/30000 [2:22:17<60:48:52,  7.56s/it, lr: 1.0e-04 loss: 7.085e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1044/30000 [2:22:17<60:48:52,  7.56s/it, lr: 1.0e-04 loss: 7.085e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1044/30000 [2:22:24<60:48:52,  7.56s/it, lr: 1.0e-04 loss: 5.540e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1044/30000 [2:22:24<60:48:52,  7.56s/it, lr: 1.0e-04 loss: 5.540e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1045/30000 [2:22:24<59:20:46,  7.38s/it, lr: 1.0e-04 loss: 5.540e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1045/30000 [2:22:24<59:20:46,  7.38s/it, lr: 1.0e-04 loss: 5.540e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1045/30000 [2:22:33<59:20:46,  7.38s/it, lr: 1.0e-04 loss: 8.496e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1045/30000 [2:22:33<59:20:46,  7.38s/it, lr: 1.0e-04 loss: 8.496e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1046/30000 [2:22:33<63:49:53,  7.94s/it, lr: 1.0e-04 loss: 8.496e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1046/30000 [2:22:33<63:49:53,  7.94s/it, lr: 1.0e-04 loss: 8.496e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1046/30000 [2:22:40<63:49:53,  7.94s/it, lr: 1.0e-04 loss: 5.193e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1046/30000 [2:22:40<63:49:53,  7.94s/it, lr: 1.0e-04 loss: 5.193e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1047/30000 [2:22:40<60:53:28,  7.57s/it, lr: 1.0e-04 loss: 5.193e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1047/30000 [2:22:40<60:53:28,  7.57s/it, lr: 1.0e-04 loss: 5.193e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1047/30000 [2:22:49<60:53:28,  7.57s/it, lr: 1.0e-04 loss: 8.879e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1047/30000 [2:22:49<60:53:28,  7.57s/it, lr: 1.0e-04 loss: 8.879e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1048/30000 [2:22:49<65:01:32,  8.09s/it, lr: 1.0e-04 loss: 8.879e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1048/30000 [2:22:49<65:01:32,  8.09s/it, lr: 1.0e-04 loss: 8.879e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1048/30000 [2:22:58<65:01:32,  8.09s/it, lr: 1.0e-04 loss: 4.621e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1048/30000 [2:22:58<65:01:32,  8.09s/it, lr: 1.0e-04 loss: 4.621e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1049/30000 [2:22:58<68:08:07,  8.47s/it, lr: 1.0e-04 loss: 4.621e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1049/30000 [2:22:58<68:08:07,  8.47s/it, lr: 1.0e-04 loss: 4.621e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1049/30000 [2:23:08<68:08:07,  8.47s/it, lr: 1.0e-04 loss: 8.688e-02]
Xiang_wan_2_2_lora:   3%|#3                                     | 1049/30000 [2:23:08<68:08:07,  8.47s/it, lr: 1.0e-04 loss: 8.688e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1050/30000 [2:23:17<68:07:58,  8.47s/it, lr: 1.0e-04 loss: 5.488e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1050/30000 [2:23:17<68:07:58,  8.47s/it, lr: 1.0e-04 loss: 5.488e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1051/30000 [2:23:17<54:13:07,  6.74s/it, lr: 1.0e-04 loss: 5.488e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1051/30000 [2:23:17<54:13:07,  6.74s/it, lr: 1.0e-04 loss: 5.488e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1051/30000 [2:23:24<54:13:07,  6.74s/it, lr: 1.0e-04 loss: 1.557e-01]
Xiang_wan_2_2_lora:   4%|#3                                     | 1051/30000 [2:23:24<54:13:07,  6.74s/it, lr: 1.0e-04 loss: 1.557e-01]
Xiang_wan_2_2_lora:   4%|#3                                     | 1052/30000 [2:23:24<54:20:44,  6.76s/it, lr: 1.0e-04 loss: 1.557e-01]
Xiang_wan_2_2_lora:   4%|#3                                     | 1052/30000 [2:23:24<54:20:44,  6.76s/it, lr: 1.0e-04 loss: 1.557e-01]
Xiang_wan_2_2_lora:   4%|#3                                     | 1052/30000 [2:23:33<54:20:44,  6.76s/it, lr: 1.0e-04 loss: 4.916e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1052/30000 [2:23:33<54:20:44,  6.76s/it, lr: 1.0e-04 loss: 4.916e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1053/30000 [2:23:33<59:46:02,  7.43s/it, lr: 1.0e-04 loss: 4.916e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1053/30000 [2:23:33<59:46:02,  7.43s/it, lr: 1.0e-04 loss: 4.916e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1053/30000 [2:23:40<59:46:02,  7.43s/it, lr: 1.0e-04 loss: 1.569e-01]
Xiang_wan_2_2_lora:   4%|#3                                     | 1053/30000 [2:23:40<59:46:02,  7.43s/it, lr: 1.0e-04 loss: 1.569e-01]
Xiang_wan_2_2_lora:   4%|#3                                     | 1054/30000 [2:23:40<57:59:50,  7.21s/it, lr: 1.0e-04 loss: 1.569e-01]
Xiang_wan_2_2_lora:   4%|#3                                     | 1054/30000 [2:23:40<57:59:50,  7.21s/it, lr: 1.0e-04 loss: 1.569e-01]
Xiang_wan_2_2_lora:   4%|#3                                     | 1054/30000 [2:23:49<57:59:50,  7.21s/it, lr: 1.0e-04 loss: 4.217e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1054/30000 [2:23:49<57:59:50,  7.21s/it, lr: 1.0e-04 loss: 4.217e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1055/30000 [2:23:49<62:52:42,  7.82s/it, lr: 1.0e-04 loss: 4.217e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1055/30000 [2:23:49<62:52:42,  7.82s/it, lr: 1.0e-04 loss: 4.217e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1055/30000 [2:23:59<62:52:42,  7.82s/it, lr: 1.0e-04 loss: 7.489e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1055/30000 [2:23:59<62:52:42,  7.82s/it, lr: 1.0e-04 loss: 7.489e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1056/30000 [2:23:59<66:30:06,  8.27s/it, lr: 1.0e-04 loss: 7.489e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1056/30000 [2:23:59<66:30:06,  8.27s/it, lr: 1.0e-04 loss: 7.489e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1056/30000 [2:24:05<66:30:06,  8.27s/it, lr: 1.0e-04 loss: 5.067e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1056/30000 [2:24:05<66:30:06,  8.27s/it, lr: 1.0e-04 loss: 5.067e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1057/30000 [2:24:05<62:42:47,  7.80s/it, lr: 1.0e-04 loss: 5.067e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1057/30000 [2:24:05<62:42:47,  7.80s/it, lr: 1.0e-04 loss: 5.067e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1057/30000 [2:24:15<62:42:47,  7.80s/it, lr: 1.0e-04 loss: 9.370e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1057/30000 [2:24:15<62:42:47,  7.80s/it, lr: 1.0e-04 loss: 9.370e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1058/30000 [2:24:15<66:15:26,  8.24s/it, lr: 1.0e-04 loss: 9.370e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1058/30000 [2:24:15<66:15:26,  8.24s/it, lr: 1.0e-04 loss: 9.370e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1058/30000 [2:24:24<66:15:26,  8.24s/it, lr: 1.0e-04 loss: 5.574e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1058/30000 [2:24:24<66:15:26,  8.24s/it, lr: 1.0e-04 loss: 5.574e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1059/30000 [2:24:24<69:25:24,  8.64s/it, lr: 1.0e-04 loss: 5.574e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1059/30000 [2:24:24<69:25:24,  8.64s/it, lr: 1.0e-04 loss: 5.574e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1059/30000 [2:24:34<69:25:24,  8.64s/it, lr: 1.0e-04 loss: 7.722e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1059/30000 [2:24:34<69:25:24,  8.64s/it, lr: 1.0e-04 loss: 7.722e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1060/30000 [2:24:40<69:25:16,  8.64s/it, lr: 1.0e-04 loss: 7.420e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1060/30000 [2:24:40<69:25:16,  8.64s/it, lr: 1.0e-04 loss: 7.420e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1061/30000 [2:24:40<49:48:09,  6.20s/it, lr: 1.0e-04 loss: 7.420e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1061/30000 [2:24:40<49:48:09,  6.20s/it, lr: 1.0e-04 loss: 7.420e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1061/30000 [2:24:50<49:48:09,  6.20s/it, lr: 1.0e-04 loss: 5.849e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1061/30000 [2:24:50<49:48:09,  6.20s/it, lr: 1.0e-04 loss: 5.849e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1062/30000 [2:24:50<56:01:46,  6.97s/it, lr: 1.0e-04 loss: 5.849e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1062/30000 [2:24:50<56:01:46,  6.97s/it, lr: 1.0e-04 loss: 5.849e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1062/30000 [2:24:59<56:01:46,  6.97s/it, lr: 1.0e-04 loss: 5.508e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1062/30000 [2:24:59<56:01:46,  6.97s/it, lr: 1.0e-04 loss: 5.508e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1063/30000 [2:24:59<61:00:43,  7.59s/it, lr: 1.0e-04 loss: 5.508e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1063/30000 [2:24:59<61:00:43,  7.59s/it, lr: 1.0e-04 loss: 5.508e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1063/30000 [2:25:06<61:00:43,  7.59s/it, lr: 1.0e-04 loss: 1.187e-01]
Xiang_wan_2_2_lora:   4%|#3                                     | 1063/30000 [2:25:06<61:00:43,  7.59s/it, lr: 1.0e-04 loss: 1.187e-01]
Xiang_wan_2_2_lora:   4%|#3                                     | 1064/30000 [2:25:06<58:55:36,  7.33s/it, lr: 1.0e-04 loss: 1.187e-01]
Xiang_wan_2_2_lora:   4%|#3                                     | 1064/30000 [2:25:06<58:55:36,  7.33s/it, lr: 1.0e-04 loss: 1.187e-01]
Xiang_wan_2_2_lora:   4%|#3                                     | 1064/30000 [2:25:12<58:55:36,  7.33s/it, lr: 1.0e-04 loss: 4.283e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1064/30000 [2:25:12<58:55:36,  7.33s/it, lr: 1.0e-04 loss: 4.283e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1065/30000 [2:25:12<57:15:39,  7.12s/it, lr: 1.0e-04 loss: 4.283e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1065/30000 [2:25:12<57:15:39,  7.12s/it, lr: 1.0e-04 loss: 4.283e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1065/30000 [2:25:22<57:15:39,  7.12s/it, lr: 1.0e-04 loss: 7.910e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1065/30000 [2:25:22<57:15:39,  7.12s/it, lr: 1.0e-04 loss: 7.910e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1066/30000 [2:25:22<62:27:34,  7.77s/it, lr: 1.0e-04 loss: 7.910e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1066/30000 [2:25:22<62:27:34,  7.77s/it, lr: 1.0e-04 loss: 7.910e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1066/30000 [2:25:31<62:27:34,  7.77s/it, lr: 1.0e-04 loss: 5.015e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1066/30000 [2:25:31<62:27:34,  7.77s/it, lr: 1.0e-04 loss: 5.015e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1067/30000 [2:25:31<67:14:26,  8.37s/it, lr: 1.0e-04 loss: 5.015e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1067/30000 [2:25:31<67:14:26,  8.37s/it, lr: 1.0e-04 loss: 5.015e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1067/30000 [2:25:41<67:14:26,  8.37s/it, lr: 1.0e-04 loss: 8.842e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1067/30000 [2:25:41<67:14:26,  8.37s/it, lr: 1.0e-04 loss: 8.842e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1068/30000 [2:25:41<69:37:39,  8.66s/it, lr: 1.0e-04 loss: 8.842e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1068/30000 [2:25:41<69:37:39,  8.66s/it, lr: 1.0e-04 loss: 8.842e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1068/30000 [2:25:47<69:37:39,  8.66s/it, lr: 1.0e-04 loss: 1.340e-01]
Xiang_wan_2_2_lora:   4%|#3                                     | 1068/30000 [2:25:47<69:37:39,  8.66s/it, lr: 1.0e-04 loss: 1.340e-01]
Xiang_wan_2_2_lora:   4%|#3                                     | 1069/30000 [2:25:47<64:33:23,  8.03s/it, lr: 1.0e-04 loss: 1.340e-01]
Xiang_wan_2_2_lora:   4%|#3                                     | 1069/30000 [2:25:47<64:33:23,  8.03s/it, lr: 1.0e-04 loss: 1.340e-01]
Xiang_wan_2_2_lora:   4%|#3                                     | 1069/30000 [2:25:54<64:33:23,  8.03s/it, lr: 1.0e-04 loss: 9.844e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1069/30000 [2:25:54<64:33:23,  8.03s/it, lr: 1.0e-04 loss: 9.844e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1070/30000 [2:26:01<64:33:15,  8.03s/it, lr: 1.0e-04 loss: 1.603e-01]
Xiang_wan_2_2_lora:   4%|#3                                     | 1070/30000 [2:26:01<64:33:15,  8.03s/it, lr: 1.0e-04 loss: 1.603e-01]
Xiang_wan_2_2_lora:   4%|#3                                     | 1071/30000 [2:26:01<47:05:19,  5.86s/it, lr: 1.0e-04 loss: 1.603e-01]
Xiang_wan_2_2_lora:   4%|#3                                     | 1071/30000 [2:26:01<47:05:19,  5.86s/it, lr: 1.0e-04 loss: 1.603e-01]
Xiang_wan_2_2_lora:   4%|#3                                     | 1071/30000 [2:26:07<47:05:19,  5.86s/it, lr: 1.0e-04 loss: 1.047e-01]
Xiang_wan_2_2_lora:   4%|#3                                     | 1071/30000 [2:26:07<47:05:19,  5.86s/it, lr: 1.0e-04 loss: 1.047e-01]
Xiang_wan_2_2_lora:   4%|#3                                     | 1072/30000 [2:26:07<48:24:54,  6.03s/it, lr: 1.0e-04 loss: 1.047e-01]
Xiang_wan_2_2_lora:   4%|#3                                     | 1072/30000 [2:26:07<48:24:54,  6.03s/it, lr: 1.0e-04 loss: 1.047e-01]
Xiang_wan_2_2_lora:   4%|#3                                     | 1072/30000 [2:26:16<48:24:54,  6.03s/it, lr: 1.0e-04 loss: 9.840e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1072/30000 [2:26:16<48:24:54,  6.03s/it, lr: 1.0e-04 loss: 9.840e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1073/30000 [2:26:16<55:22:30,  6.89s/it, lr: 1.0e-04 loss: 9.840e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1073/30000 [2:26:16<55:22:30,  6.89s/it, lr: 1.0e-04 loss: 9.840e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1073/30000 [2:26:23<55:22:30,  6.89s/it, lr: 1.0e-04 loss: 9.036e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1073/30000 [2:26:23<55:22:30,  6.89s/it, lr: 1.0e-04 loss: 9.036e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1074/30000 [2:26:23<55:09:33,  6.86s/it, lr: 1.0e-04 loss: 9.036e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1074/30000 [2:26:23<55:09:33,  6.86s/it, lr: 1.0e-04 loss: 9.036e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1074/30000 [2:26:30<55:09:33,  6.86s/it, lr: 1.0e-04 loss: 7.801e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1074/30000 [2:26:30<55:09:33,  6.86s/it, lr: 1.0e-04 loss: 7.801e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1075/30000 [2:26:30<54:23:03,  6.77s/it, lr: 1.0e-04 loss: 7.801e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1075/30000 [2:26:30<54:23:03,  6.77s/it, lr: 1.0e-04 loss: 7.801e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1075/30000 [2:26:39<54:23:03,  6.77s/it, lr: 1.0e-04 loss: 8.564e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1075/30000 [2:26:39<54:23:03,  6.77s/it, lr: 1.0e-04 loss: 8.564e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1076/30000 [2:26:39<60:13:01,  7.49s/it, lr: 1.0e-04 loss: 8.564e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1076/30000 [2:26:39<60:13:01,  7.49s/it, lr: 1.0e-04 loss: 8.564e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1076/30000 [2:26:46<60:13:01,  7.49s/it, lr: 1.0e-04 loss: 6.753e-02]
Xiang_wan_2_2_lora:   4%|#3                                     | 1076/30000 [2:26:46<60:13:01,  7.49s/it, lr: 1.0e-04 loss: 6.753e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1077/30000 [2:26:46<58:01:15,  7.22s/it, lr: 1.0e-04 loss: 6.753e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1077/30000 [2:26:46<58:01:15,  7.22s/it, lr: 1.0e-04 loss: 6.753e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1077/30000 [2:26:52<58:01:15,  7.22s/it, lr: 1.0e-04 loss: 9.210e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1077/30000 [2:26:52<58:01:15,  7.22s/it, lr: 1.0e-04 loss: 9.210e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1078/30000 [2:26:52<56:26:47,  7.03s/it, lr: 1.0e-04 loss: 9.210e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1078/30000 [2:26:52<56:26:47,  7.03s/it, lr: 1.0e-04 loss: 9.210e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1078/30000 [2:26:59<56:26:47,  7.03s/it, lr: 1.0e-04 loss: 8.933e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1078/30000 [2:26:59<56:26:47,  7.03s/it, lr: 1.0e-04 loss: 8.933e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1079/30000 [2:26:59<55:32:51,  6.91s/it, lr: 1.0e-04 loss: 8.933e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1079/30000 [2:26:59<55:32:51,  6.91s/it, lr: 1.0e-04 loss: 8.933e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1079/30000 [2:27:08<55:32:51,  6.91s/it, lr: 1.0e-04 loss: 7.220e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1079/30000 [2:27:08<55:32:51,  6.91s/it, lr: 1.0e-04 loss: 7.220e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1080/30000 [2:27:15<55:32:45,  6.91s/it, lr: 1.0e-04 loss: 8.085e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1080/30000 [2:27:15<55:32:45,  6.91s/it, lr: 1.0e-04 loss: 8.085e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1081/30000 [2:27:15<42:50:02,  5.33s/it, lr: 1.0e-04 loss: 8.085e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1081/30000 [2:27:15<42:50:02,  5.33s/it, lr: 1.0e-04 loss: 8.085e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1081/30000 [2:27:24<42:50:02,  5.33s/it, lr: 1.0e-04 loss: 7.542e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1081/30000 [2:27:24<42:50:02,  5.33s/it, lr: 1.0e-04 loss: 7.542e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1082/30000 [2:27:24<50:43:16,  6.31s/it, lr: 1.0e-04 loss: 7.542e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1082/30000 [2:27:24<50:43:16,  6.31s/it, lr: 1.0e-04 loss: 7.542e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1082/30000 [2:27:34<50:43:16,  6.31s/it, lr: 1.0e-04 loss: 8.976e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1082/30000 [2:27:34<50:43:16,  6.31s/it, lr: 1.0e-04 loss: 8.976e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1083/30000 [2:27:34<57:09:26,  7.12s/it, lr: 1.0e-04 loss: 8.976e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1083/30000 [2:27:34<57:09:26,  7.12s/it, lr: 1.0e-04 loss: 8.976e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1083/30000 [2:27:40<57:09:26,  7.12s/it, lr: 1.0e-04 loss: 9.429e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1083/30000 [2:27:40<57:09:26,  7.12s/it, lr: 1.0e-04 loss: 9.429e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1084/30000 [2:27:40<56:06:34,  6.99s/it, lr: 1.0e-04 loss: 9.429e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1084/30000 [2:27:40<56:06:34,  6.99s/it, lr: 1.0e-04 loss: 9.429e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1084/30000 [2:27:50<56:06:34,  6.99s/it, lr: 1.0e-04 loss: 8.846e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1084/30000 [2:27:50<56:06:34,  6.99s/it, lr: 1.0e-04 loss: 8.846e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1085/30000 [2:27:50<61:23:33,  7.64s/it, lr: 1.0e-04 loss: 8.846e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1085/30000 [2:27:50<61:23:33,  7.64s/it, lr: 1.0e-04 loss: 8.846e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1085/30000 [2:27:59<61:23:33,  7.64s/it, lr: 1.0e-04 loss: 6.848e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1085/30000 [2:27:59<61:23:33,  7.64s/it, lr: 1.0e-04 loss: 6.848e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1086/30000 [2:27:59<65:26:54,  8.15s/it, lr: 1.0e-04 loss: 6.848e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1086/30000 [2:27:59<65:26:54,  8.15s/it, lr: 1.0e-04 loss: 6.848e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1086/30000 [2:28:06<65:26:54,  8.15s/it, lr: 1.0e-04 loss: 1.372e-01]
Xiang_wan_2_2_lora:   4%|#4                                     | 1086/30000 [2:28:06<65:26:54,  8.15s/it, lr: 1.0e-04 loss: 1.372e-01]
Xiang_wan_2_2_lora:   4%|#4                                     | 1087/30000 [2:28:06<61:49:07,  7.70s/it, lr: 1.0e-04 loss: 1.372e-01]
Xiang_wan_2_2_lora:   4%|#4                                     | 1087/30000 [2:28:06<61:49:07,  7.70s/it, lr: 1.0e-04 loss: 1.372e-01]
Xiang_wan_2_2_lora:   4%|#4                                     | 1087/30000 [2:28:15<61:49:07,  7.70s/it, lr: 1.0e-04 loss: 1.065e-01]
Xiang_wan_2_2_lora:   4%|#4                                     | 1087/30000 [2:28:15<61:49:07,  7.70s/it, lr: 1.0e-04 loss: 1.065e-01]
Xiang_wan_2_2_lora:   4%|#4                                     | 1088/30000 [2:28:15<65:35:21,  8.17s/it, lr: 1.0e-04 loss: 1.065e-01]
Xiang_wan_2_2_lora:   4%|#4                                     | 1088/30000 [2:28:15<65:35:21,  8.17s/it, lr: 1.0e-04 loss: 1.065e-01]
Xiang_wan_2_2_lora:   4%|#4                                     | 1088/30000 [2:28:25<65:35:21,  8.17s/it, lr: 1.0e-04 loss: 7.371e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1088/30000 [2:28:25<65:35:21,  8.17s/it, lr: 1.0e-04 loss: 7.371e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1089/30000 [2:28:25<68:56:32,  8.58s/it, lr: 1.0e-04 loss: 7.371e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1089/30000 [2:28:25<68:56:32,  8.58s/it, lr: 1.0e-04 loss: 7.371e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1089/30000 [2:28:34<68:56:32,  8.58s/it, lr: 1.0e-04 loss: 6.887e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1089/30000 [2:28:34<68:56:32,  8.58s/it, lr: 1.0e-04 loss: 6.887e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1090/30000 [2:28:43<68:56:23,  8.58s/it, lr: 1.0e-04 loss: 4.801e-01]
Xiang_wan_2_2_lora:   4%|#4                                     | 1090/30000 [2:28:43<68:56:23,  8.58s/it, lr: 1.0e-04 loss: 4.801e-01]
Xiang_wan_2_2_lora:   4%|#4                                     | 1091/30000 [2:28:43<54:34:05,  6.80s/it, lr: 1.0e-04 loss: 4.801e-01]
Xiang_wan_2_2_lora:   4%|#4                                     | 1091/30000 [2:28:43<54:34:05,  6.80s/it, lr: 1.0e-04 loss: 4.801e-01]
Xiang_wan_2_2_lora:   4%|#4                                     | 1091/30000 [2:28:53<54:34:05,  6.80s/it, lr: 1.0e-04 loss: 7.353e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1091/30000 [2:28:53<54:34:05,  6.80s/it, lr: 1.0e-04 loss: 7.353e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1092/30000 [2:28:53<59:36:23,  7.42s/it, lr: 1.0e-04 loss: 7.353e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1092/30000 [2:28:53<59:36:23,  7.42s/it, lr: 1.0e-04 loss: 7.353e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1092/30000 [2:29:02<59:36:23,  7.42s/it, lr: 1.0e-04 loss: 4.532e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1092/30000 [2:29:02<59:36:23,  7.42s/it, lr: 1.0e-04 loss: 4.532e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1093/30000 [2:29:02<63:40:15,  7.93s/it, lr: 1.0e-04 loss: 4.532e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1093/30000 [2:29:02<63:40:15,  7.93s/it, lr: 1.0e-04 loss: 4.532e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1093/30000 [2:29:09<63:40:15,  7.93s/it, lr: 1.0e-04 loss: 1.021e-01]
Xiang_wan_2_2_lora:   4%|#4                                     | 1093/30000 [2:29:09<63:40:15,  7.93s/it, lr: 1.0e-04 loss: 1.021e-01]
Xiang_wan_2_2_lora:   4%|#4                                     | 1094/30000 [2:29:09<60:56:24,  7.59s/it, lr: 1.0e-04 loss: 1.021e-01]
Xiang_wan_2_2_lora:   4%|#4                                     | 1094/30000 [2:29:09<60:56:24,  7.59s/it, lr: 1.0e-04 loss: 1.021e-01]
Xiang_wan_2_2_lora:   4%|#4                                     | 1094/30000 [2:29:18<60:56:24,  7.59s/it, lr: 1.0e-04 loss: 3.280e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1094/30000 [2:29:18<60:56:24,  7.59s/it, lr: 1.0e-04 loss: 3.280e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1095/30000 [2:29:18<64:57:25,  8.09s/it, lr: 1.0e-04 loss: 3.280e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1095/30000 [2:29:18<64:57:25,  8.09s/it, lr: 1.0e-04 loss: 3.280e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1095/30000 [2:29:27<64:57:25,  8.09s/it, lr: 1.0e-04 loss: 7.007e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1095/30000 [2:29:27<64:57:25,  8.09s/it, lr: 1.0e-04 loss: 7.007e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1096/30000 [2:29:27<68:16:23,  8.50s/it, lr: 1.0e-04 loss: 7.007e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1096/30000 [2:29:27<68:16:23,  8.50s/it, lr: 1.0e-04 loss: 7.007e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1096/30000 [2:29:37<68:16:23,  8.50s/it, lr: 1.0e-04 loss: 2.440e-01]
Xiang_wan_2_2_lora:   4%|#4                                     | 1096/30000 [2:29:37<68:16:23,  8.50s/it, lr: 1.0e-04 loss: 2.440e-01]
Xiang_wan_2_2_lora:   4%|#4                                     | 1097/30000 [2:29:37<70:24:46,  8.77s/it, lr: 1.0e-04 loss: 2.440e-01]
Xiang_wan_2_2_lora:   4%|#4                                     | 1097/30000 [2:29:37<70:24:46,  8.77s/it, lr: 1.0e-04 loss: 2.440e-01]
Xiang_wan_2_2_lora:   4%|#4                                     | 1097/30000 [2:29:46<70:24:46,  8.77s/it, lr: 1.0e-04 loss: 6.859e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1097/30000 [2:29:46<70:24:46,  8.77s/it, lr: 1.0e-04 loss: 6.859e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1098/30000 [2:29:46<71:49:01,  8.95s/it, lr: 1.0e-04 loss: 6.859e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1098/30000 [2:29:46<71:49:01,  8.95s/it, lr: 1.0e-04 loss: 6.859e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1098/30000 [2:29:56<71:49:01,  8.95s/it, lr: 1.0e-04 loss: 4.625e-01]
Xiang_wan_2_2_lora:   4%|#4                                     | 1098/30000 [2:29:56<71:49:01,  8.95s/it, lr: 1.0e-04 loss: 4.625e-01]
Xiang_wan_2_2_lora:   4%|#4                                     | 1099/30000 [2:29:56<72:56:41,  9.09s/it, lr: 1.0e-04 loss: 4.625e-01]
Xiang_wan_2_2_lora:   4%|#4                                     | 1099/30000 [2:29:56<72:56:41,  9.09s/it, lr: 1.0e-04 loss: 4.625e-01]
Xiang_wan_2_2_lora:   4%|#4                                     | 1099/30000 [2:30:02<72:56:41,  9.09s/it, lr: 1.0e-04 loss: 7.040e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1099/30000 [2:30:02<72:56:41,  9.09s/it, lr: 1.0e-04 loss: 7.040e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1100/30000 [2:30:12<72:56:32,  9.09s/it, lr: 1.0e-04 loss: 2.639e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1100/30000 [2:30:12<72:56:32,  9.09s/it, lr: 1.0e-04 loss: 2.639e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1101/30000 [2:30:12<56:47:47,  7.08s/it, lr: 1.0e-04 loss: 2.639e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1101/30000 [2:30:12<56:47:47,  7.08s/it, lr: 1.0e-04 loss: 2.639e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1101/30000 [2:30:18<56:47:47,  7.08s/it, lr: 1.0e-04 loss: 1.046e-01]
Xiang_wan_2_2_lora:   4%|#4                                     | 1101/30000 [2:30:18<56:47:47,  7.08s/it, lr: 1.0e-04 loss: 1.046e-01]
Xiang_wan_2_2_lora:   4%|#4                                     | 1102/30000 [2:30:18<55:47:29,  6.95s/it, lr: 1.0e-04 loss: 1.046e-01]
Xiang_wan_2_2_lora:   4%|#4                                     | 1102/30000 [2:30:18<55:47:29,  6.95s/it, lr: 1.0e-04 loss: 1.046e-01]
Xiang_wan_2_2_lora:   4%|#4                                     | 1102/30000 [2:30:28<55:47:29,  6.95s/it, lr: 1.0e-04 loss: 8.199e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1102/30000 [2:30:28<55:47:29,  6.95s/it, lr: 1.0e-04 loss: 8.199e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1103/30000 [2:30:28<61:17:01,  7.63s/it, lr: 1.0e-04 loss: 8.199e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1103/30000 [2:30:28<61:17:01,  7.63s/it, lr: 1.0e-04 loss: 8.199e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1103/30000 [2:30:37<61:17:01,  7.63s/it, lr: 1.0e-04 loss: 7.186e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1103/30000 [2:30:37<61:17:01,  7.63s/it, lr: 1.0e-04 loss: 7.186e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1104/30000 [2:30:37<65:04:40,  8.11s/it, lr: 1.0e-04 loss: 7.186e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1104/30000 [2:30:37<65:04:40,  8.11s/it, lr: 1.0e-04 loss: 7.186e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1104/30000 [2:30:44<65:04:40,  8.11s/it, lr: 1.0e-04 loss: 1.192e-01]
Xiang_wan_2_2_lora:   4%|#4                                     | 1104/30000 [2:30:44<65:04:40,  8.11s/it, lr: 1.0e-04 loss: 1.192e-01]
Xiang_wan_2_2_lora:   4%|#4                                     | 1105/30000 [2:30:44<61:52:01,  7.71s/it, lr: 1.0e-04 loss: 1.192e-01]
Xiang_wan_2_2_lora:   4%|#4                                     | 1105/30000 [2:30:44<61:52:01,  7.71s/it, lr: 1.0e-04 loss: 1.192e-01]
Xiang_wan_2_2_lora:   4%|#4                                     | 1105/30000 [2:30:53<61:52:01,  7.71s/it, lr: 1.0e-04 loss: 6.432e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1105/30000 [2:30:53<61:52:01,  7.71s/it, lr: 1.0e-04 loss: 6.432e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1106/30000 [2:30:53<65:26:10,  8.15s/it, lr: 1.0e-04 loss: 6.432e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1106/30000 [2:30:53<65:26:10,  8.15s/it, lr: 1.0e-04 loss: 6.432e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1106/30000 [2:31:00<65:26:10,  8.15s/it, lr: 1.0e-04 loss: 4.896e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1106/30000 [2:31:00<65:26:10,  8.15s/it, lr: 1.0e-04 loss: 4.896e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1107/30000 [2:31:00<62:04:56,  7.74s/it, lr: 1.0e-04 loss: 4.896e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1107/30000 [2:31:00<62:04:56,  7.74s/it, lr: 1.0e-04 loss: 4.896e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1107/30000 [2:31:09<62:04:56,  7.74s/it, lr: 1.0e-04 loss: 7.365e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1107/30000 [2:31:09<62:04:56,  7.74s/it, lr: 1.0e-04 loss: 7.365e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1108/30000 [2:31:09<65:47:49,  8.20s/it, lr: 1.0e-04 loss: 7.365e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1108/30000 [2:31:09<65:47:49,  8.20s/it, lr: 1.0e-04 loss: 7.365e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1108/30000 [2:31:19<65:47:49,  8.20s/it, lr: 1.0e-04 loss: 4.924e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1108/30000 [2:31:19<65:47:49,  8.20s/it, lr: 1.0e-04 loss: 4.924e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1109/30000 [2:31:19<68:38:20,  8.55s/it, lr: 1.0e-04 loss: 4.924e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1109/30000 [2:31:19<68:38:20,  8.55s/it, lr: 1.0e-04 loss: 4.924e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1109/30000 [2:31:28<68:38:20,  8.55s/it, lr: 1.0e-04 loss: 9.595e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1109/30000 [2:31:28<68:38:20,  8.55s/it, lr: 1.0e-04 loss: 9.595e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1110/30000 [2:31:35<68:38:11,  8.55s/it, lr: 1.0e-04 loss: 6.378e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1110/30000 [2:31:35<68:38:11,  8.55s/it, lr: 1.0e-04 loss: 6.378e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1111/30000 [2:31:35<49:22:56,  6.15s/it, lr: 1.0e-04 loss: 6.378e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1111/30000 [2:31:35<49:22:56,  6.15s/it, lr: 1.0e-04 loss: 6.378e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1111/30000 [2:31:44<49:22:56,  6.15s/it, lr: 1.0e-04 loss: 9.805e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1111/30000 [2:31:44<49:22:56,  6.15s/it, lr: 1.0e-04 loss: 9.805e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1112/30000 [2:31:44<55:37:37,  6.93s/it, lr: 1.0e-04 loss: 9.805e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1112/30000 [2:31:44<55:37:37,  6.93s/it, lr: 1.0e-04 loss: 9.805e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1112/30000 [2:31:53<55:37:37,  6.93s/it, lr: 1.0e-04 loss: 5.827e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1112/30000 [2:31:53<55:37:37,  6.93s/it, lr: 1.0e-04 loss: 5.827e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1113/30000 [2:31:53<60:43:39,  7.57s/it, lr: 1.0e-04 loss: 5.827e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1113/30000 [2:31:53<60:43:39,  7.57s/it, lr: 1.0e-04 loss: 5.827e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1113/30000 [2:32:03<60:43:39,  7.57s/it, lr: 1.0e-04 loss: 7.231e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1113/30000 [2:32:03<60:43:39,  7.57s/it, lr: 1.0e-04 loss: 7.231e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1114/30000 [2:32:03<64:40:31,  8.06s/it, lr: 1.0e-04 loss: 7.231e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1114/30000 [2:32:03<64:40:31,  8.06s/it, lr: 1.0e-04 loss: 7.231e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1114/30000 [2:32:10<64:40:31,  8.06s/it, lr: 1.0e-04 loss: 4.992e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1114/30000 [2:32:10<64:40:31,  8.06s/it, lr: 1.0e-04 loss: 4.992e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1115/30000 [2:32:10<61:33:54,  7.67s/it, lr: 1.0e-04 loss: 4.992e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1115/30000 [2:32:10<61:33:54,  7.67s/it, lr: 1.0e-04 loss: 4.992e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1115/30000 [2:32:19<61:33:54,  7.67s/it, lr: 1.0e-04 loss: 7.696e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1115/30000 [2:32:19<61:33:54,  7.67s/it, lr: 1.0e-04 loss: 7.696e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1116/30000 [2:32:19<65:24:09,  8.15s/it, lr: 1.0e-04 loss: 7.696e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1116/30000 [2:32:19<65:24:09,  8.15s/it, lr: 1.0e-04 loss: 7.696e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1116/30000 [2:32:26<65:24:09,  8.15s/it, lr: 1.0e-04 loss: 8.711e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1116/30000 [2:32:26<65:24:09,  8.15s/it, lr: 1.0e-04 loss: 8.711e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1117/30000 [2:32:26<61:48:53,  7.70s/it, lr: 1.0e-04 loss: 8.711e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1117/30000 [2:32:26<61:48:53,  7.70s/it, lr: 1.0e-04 loss: 8.711e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1117/30000 [2:32:35<61:48:53,  7.70s/it, lr: 1.0e-04 loss: 8.093e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1117/30000 [2:32:35<61:48:53,  7.70s/it, lr: 1.0e-04 loss: 8.093e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1118/30000 [2:32:35<66:07:54,  8.24s/it, lr: 1.0e-04 loss: 8.093e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1118/30000 [2:32:35<66:07:54,  8.24s/it, lr: 1.0e-04 loss: 8.093e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1118/30000 [2:32:42<66:07:54,  8.24s/it, lr: 1.0e-04 loss: 8.416e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1118/30000 [2:32:42<66:07:54,  8.24s/it, lr: 1.0e-04 loss: 8.416e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1119/30000 [2:32:42<62:25:11,  7.78s/it, lr: 1.0e-04 loss: 8.416e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1119/30000 [2:32:42<62:25:11,  7.78s/it, lr: 1.0e-04 loss: 8.416e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1119/30000 [2:32:48<62:25:11,  7.78s/it, lr: 1.0e-04 loss: 8.363e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1119/30000 [2:32:48<62:25:11,  7.78s/it, lr: 1.0e-04 loss: 8.363e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1120/30000 [2:32:58<62:25:04,  7.78s/it, lr: 1.0e-04 loss: 5.056e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1120/30000 [2:32:58<62:25:04,  7.78s/it, lr: 1.0e-04 loss: 5.056e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1121/30000 [2:32:58<51:02:49,  6.36s/it, lr: 1.0e-04 loss: 5.056e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1121/30000 [2:32:58<51:02:49,  6.36s/it, lr: 1.0e-04 loss: 5.056e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1121/30000 [2:33:04<51:02:49,  6.36s/it, lr: 1.0e-04 loss: 8.846e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1121/30000 [2:33:04<51:02:49,  6.36s/it, lr: 1.0e-04 loss: 8.846e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1122/30000 [2:33:04<51:30:42,  6.42s/it, lr: 1.0e-04 loss: 8.846e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1122/30000 [2:33:04<51:30:42,  6.42s/it, lr: 1.0e-04 loss: 8.846e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1122/30000 [2:33:14<51:30:42,  6.42s/it, lr: 1.0e-04 loss: 4.563e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1122/30000 [2:33:14<51:30:42,  6.42s/it, lr: 1.0e-04 loss: 4.563e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1123/30000 [2:33:14<57:38:02,  7.19s/it, lr: 1.0e-04 loss: 4.563e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1123/30000 [2:33:14<57:38:02,  7.19s/it, lr: 1.0e-04 loss: 4.563e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1123/30000 [2:33:20<57:38:02,  7.19s/it, lr: 1.0e-04 loss: 9.316e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1123/30000 [2:33:20<57:38:02,  7.19s/it, lr: 1.0e-04 loss: 9.316e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1124/30000 [2:33:20<56:24:17,  7.03s/it, lr: 1.0e-04 loss: 9.316e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1124/30000 [2:33:20<56:24:17,  7.03s/it, lr: 1.0e-04 loss: 9.316e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1124/30000 [2:33:27<56:24:17,  7.03s/it, lr: 1.0e-04 loss: 5.757e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1124/30000 [2:33:27<56:24:17,  7.03s/it, lr: 1.0e-04 loss: 5.757e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1125/30000 [2:33:27<55:53:27,  6.97s/it, lr: 1.0e-04 loss: 5.757e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1125/30000 [2:33:27<55:53:27,  6.97s/it, lr: 1.0e-04 loss: 5.757e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1125/30000 [2:33:36<55:53:27,  6.97s/it, lr: 1.0e-04 loss: 8.901e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1125/30000 [2:33:36<55:53:27,  6.97s/it, lr: 1.0e-04 loss: 8.901e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1126/30000 [2:33:36<61:22:21,  7.65s/it, lr: 1.0e-04 loss: 8.901e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1126/30000 [2:33:36<61:22:21,  7.65s/it, lr: 1.0e-04 loss: 8.901e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1126/30000 [2:33:43<61:22:21,  7.65s/it, lr: 1.0e-04 loss: 5.898e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1126/30000 [2:33:43<61:22:21,  7.65s/it, lr: 1.0e-04 loss: 5.898e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1127/30000 [2:33:43<59:04:44,  7.37s/it, lr: 1.0e-04 loss: 5.898e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1127/30000 [2:33:43<59:04:44,  7.37s/it, lr: 1.0e-04 loss: 5.898e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1127/30000 [2:33:50<59:04:44,  7.37s/it, lr: 1.0e-04 loss: 1.028e-01]
Xiang_wan_2_2_lora:   4%|#4                                     | 1127/30000 [2:33:50<59:04:44,  7.37s/it, lr: 1.0e-04 loss: 1.028e-01]
Xiang_wan_2_2_lora:   4%|#4                                     | 1128/30000 [2:33:50<57:24:20,  7.16s/it, lr: 1.0e-04 loss: 1.028e-01]
Xiang_wan_2_2_lora:   4%|#4                                     | 1128/30000 [2:33:50<57:24:20,  7.16s/it, lr: 1.0e-04 loss: 1.028e-01]
Xiang_wan_2_2_lora:   4%|#4                                     | 1128/30000 [2:33:59<57:24:20,  7.16s/it, lr: 1.0e-04 loss: 1.042e-01]
Xiang_wan_2_2_lora:   4%|#4                                     | 1128/30000 [2:33:59<57:24:20,  7.16s/it, lr: 1.0e-04 loss: 1.042e-01]
Xiang_wan_2_2_lora:   4%|#4                                     | 1129/30000 [2:33:59<62:35:02,  7.80s/it, lr: 1.0e-04 loss: 1.042e-01]
Xiang_wan_2_2_lora:   4%|#4                                     | 1129/30000 [2:33:59<62:35:02,  7.80s/it, lr: 1.0e-04 loss: 1.042e-01]
Xiang_wan_2_2_lora:   4%|#4                                     | 1129/30000 [2:34:09<62:35:02,  7.80s/it, lr: 1.0e-04 loss: 7.484e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1129/30000 [2:34:09<62:35:02,  7.80s/it, lr: 1.0e-04 loss: 7.484e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1130/30000 [2:34:18<62:34:54,  7.80s/it, lr: 1.0e-04 loss: 2.195e-01]
Xiang_wan_2_2_lora:   4%|#4                                     | 1130/30000 [2:34:18<62:34:54,  7.80s/it, lr: 1.0e-04 loss: 2.195e-01]
Xiang_wan_2_2_lora:   4%|#4                                     | 1131/30000 [2:34:18<51:16:35,  6.39s/it, lr: 1.0e-04 loss: 2.195e-01]
Xiang_wan_2_2_lora:   4%|#4                                     | 1131/30000 [2:34:18<51:16:35,  6.39s/it, lr: 1.0e-04 loss: 2.195e-01]
Xiang_wan_2_2_lora:   4%|#4                                     | 1131/30000 [2:34:25<51:16:35,  6.39s/it, lr: 1.0e-04 loss: 1.113e-01]
Xiang_wan_2_2_lora:   4%|#4                                     | 1131/30000 [2:34:25<51:16:35,  6.39s/it, lr: 1.0e-04 loss: 1.113e-01]
Xiang_wan_2_2_lora:   4%|#4                                     | 1132/30000 [2:34:25<52:10:25,  6.51s/it, lr: 1.0e-04 loss: 1.113e-01]
Xiang_wan_2_2_lora:   4%|#4                                     | 1132/30000 [2:34:25<52:10:25,  6.51s/it, lr: 1.0e-04 loss: 1.113e-01]
Xiang_wan_2_2_lora:   4%|#4                                     | 1132/30000 [2:34:31<52:10:25,  6.51s/it, lr: 1.0e-04 loss: 7.483e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1132/30000 [2:34:31<52:10:25,  6.51s/it, lr: 1.0e-04 loss: 7.483e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1133/30000 [2:34:31<52:27:47,  6.54s/it, lr: 1.0e-04 loss: 7.483e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1133/30000 [2:34:31<52:27:47,  6.54s/it, lr: 1.0e-04 loss: 7.483e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1133/30000 [2:34:41<52:27:47,  6.54s/it, lr: 1.0e-04 loss: 6.116e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1133/30000 [2:34:41<52:27:47,  6.54s/it, lr: 1.0e-04 loss: 6.116e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1134/30000 [2:34:41<58:27:40,  7.29s/it, lr: 1.0e-04 loss: 6.116e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1134/30000 [2:34:41<58:27:40,  7.29s/it, lr: 1.0e-04 loss: 6.116e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1134/30000 [2:34:50<58:27:40,  7.29s/it, lr: 1.0e-04 loss: 9.807e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1134/30000 [2:34:50<58:27:40,  7.29s/it, lr: 1.0e-04 loss: 9.807e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1135/30000 [2:34:50<63:09:49,  7.88s/it, lr: 1.0e-04 loss: 9.807e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1135/30000 [2:34:50<63:09:49,  7.88s/it, lr: 1.0e-04 loss: 9.807e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1135/30000 [2:35:00<63:09:49,  7.88s/it, lr: 1.0e-04 loss: 7.550e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1135/30000 [2:35:00<63:09:49,  7.88s/it, lr: 1.0e-04 loss: 7.550e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1136/30000 [2:35:00<66:28:51,  8.29s/it, lr: 1.0e-04 loss: 7.550e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1136/30000 [2:35:00<66:28:51,  8.29s/it, lr: 1.0e-04 loss: 7.550e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1136/30000 [2:35:09<66:28:51,  8.29s/it, lr: 1.0e-04 loss: 3.679e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1136/30000 [2:35:09<66:28:51,  8.29s/it, lr: 1.0e-04 loss: 3.679e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1137/30000 [2:35:09<69:02:28,  8.61s/it, lr: 1.0e-04 loss: 3.679e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1137/30000 [2:35:09<69:02:28,  8.61s/it, lr: 1.0e-04 loss: 3.679e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1137/30000 [2:35:16<69:02:28,  8.61s/it, lr: 1.0e-04 loss: 9.551e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1137/30000 [2:35:16<69:02:28,  8.61s/it, lr: 1.0e-04 loss: 9.551e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1138/30000 [2:35:16<64:32:38,  8.05s/it, lr: 1.0e-04 loss: 9.551e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1138/30000 [2:35:16<64:32:38,  8.05s/it, lr: 1.0e-04 loss: 9.551e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1138/30000 [2:35:25<64:32:38,  8.05s/it, lr: 1.0e-04 loss: 2.496e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1138/30000 [2:35:25<64:32:38,  8.05s/it, lr: 1.0e-04 loss: 2.496e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1139/30000 [2:35:25<68:11:52,  8.51s/it, lr: 1.0e-04 loss: 2.496e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1139/30000 [2:35:25<68:11:52,  8.51s/it, lr: 1.0e-04 loss: 2.496e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1139/30000 [2:35:32<68:11:52,  8.51s/it, lr: 1.0e-04 loss: 9.917e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1139/30000 [2:35:32<68:11:52,  8.51s/it, lr: 1.0e-04 loss: 9.917e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1140/30000 [2:35:41<68:11:43,  8.51s/it, lr: 1.0e-04 loss: 4.965e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1140/30000 [2:35:41<68:11:43,  8.51s/it, lr: 1.0e-04 loss: 4.965e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1141/30000 [2:35:41<54:14:42,  6.77s/it, lr: 1.0e-04 loss: 4.965e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1141/30000 [2:35:41<54:14:42,  6.77s/it, lr: 1.0e-04 loss: 4.965e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1141/30000 [2:35:48<54:14:42,  6.77s/it, lr: 1.0e-04 loss: 7.975e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1141/30000 [2:35:48<54:14:42,  6.77s/it, lr: 1.0e-04 loss: 7.975e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1142/30000 [2:35:48<53:54:29,  6.72s/it, lr: 1.0e-04 loss: 7.975e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1142/30000 [2:35:48<53:54:29,  6.72s/it, lr: 1.0e-04 loss: 7.975e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1142/30000 [2:35:57<53:54:29,  6.72s/it, lr: 1.0e-04 loss: 1.369e-01]
Xiang_wan_2_2_lora:   4%|#4                                     | 1142/30000 [2:35:57<53:54:29,  6.72s/it, lr: 1.0e-04 loss: 1.369e-01]
Xiang_wan_2_2_lora:   4%|#4                                     | 1143/30000 [2:35:57<59:22:38,  7.41s/it, lr: 1.0e-04 loss: 1.369e-01]
Xiang_wan_2_2_lora:   4%|#4                                     | 1143/30000 [2:35:57<59:22:38,  7.41s/it, lr: 1.0e-04 loss: 1.369e-01]
Xiang_wan_2_2_lora:   4%|#4                                     | 1143/30000 [2:36:07<59:22:38,  7.41s/it, lr: 1.0e-04 loss: 6.841e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1143/30000 [2:36:07<59:22:38,  7.41s/it, lr: 1.0e-04 loss: 6.841e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1144/30000 [2:36:07<63:32:37,  7.93s/it, lr: 1.0e-04 loss: 6.841e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1144/30000 [2:36:07<63:32:37,  7.93s/it, lr: 1.0e-04 loss: 6.841e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1144/30000 [2:36:13<63:32:37,  7.93s/it, lr: 1.0e-04 loss: 6.776e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1144/30000 [2:36:13<63:32:37,  7.93s/it, lr: 1.0e-04 loss: 6.776e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1145/30000 [2:36:13<60:32:38,  7.55s/it, lr: 1.0e-04 loss: 6.776e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1145/30000 [2:36:13<60:32:38,  7.55s/it, lr: 1.0e-04 loss: 6.776e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1145/30000 [2:36:20<60:32:38,  7.55s/it, lr: 1.0e-04 loss: 9.083e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1145/30000 [2:36:20<60:32:38,  7.55s/it, lr: 1.0e-04 loss: 9.083e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1146/30000 [2:36:20<58:15:24,  7.27s/it, lr: 1.0e-04 loss: 9.083e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1146/30000 [2:36:20<58:15:24,  7.27s/it, lr: 1.0e-04 loss: 9.083e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1146/30000 [2:36:26<58:15:24,  7.27s/it, lr: 1.0e-04 loss: 7.056e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1146/30000 [2:36:26<58:15:24,  7.27s/it, lr: 1.0e-04 loss: 7.056e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1147/30000 [2:36:26<57:05:09,  7.12s/it, lr: 1.0e-04 loss: 7.056e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1147/30000 [2:36:26<57:05:09,  7.12s/it, lr: 1.0e-04 loss: 7.056e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1147/30000 [2:36:36<57:05:09,  7.12s/it, lr: 1.0e-04 loss: 8.648e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1147/30000 [2:36:36<57:05:09,  7.12s/it, lr: 1.0e-04 loss: 8.648e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1148/30000 [2:36:36<62:16:07,  7.77s/it, lr: 1.0e-04 loss: 8.648e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1148/30000 [2:36:36<62:16:07,  7.77s/it, lr: 1.0e-04 loss: 8.648e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1148/30000 [2:36:45<62:16:07,  7.77s/it, lr: 1.0e-04 loss: 1.023e-01]
Xiang_wan_2_2_lora:   4%|#4                                     | 1148/30000 [2:36:45<62:16:07,  7.77s/it, lr: 1.0e-04 loss: 1.023e-01]
Xiang_wan_2_2_lora:   4%|#4                                     | 1149/30000 [2:36:45<66:06:10,  8.25s/it, lr: 1.0e-04 loss: 1.023e-01]
Xiang_wan_2_2_lora:   4%|#4                                     | 1149/30000 [2:36:45<66:06:10,  8.25s/it, lr: 1.0e-04 loss: 1.023e-01]
Xiang_wan_2_2_lora:   4%|#4                                     | 1149/30000 [2:36:55<66:06:10,  8.25s/it, lr: 1.0e-04 loss: 8.309e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1149/30000 [2:36:55<66:06:10,  8.25s/it, lr: 1.0e-04 loss: 8.309e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1150/30000 [2:37:04<66:06:02,  8.25s/it, lr: 1.0e-04 loss: 7.392e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1150/30000 [2:37:04<66:06:02,  8.25s/it, lr: 1.0e-04 loss: 7.392e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1151/30000 [2:37:04<53:01:42,  6.62s/it, lr: 1.0e-04 loss: 7.392e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1151/30000 [2:37:04<53:01:42,  6.62s/it, lr: 1.0e-04 loss: 7.392e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1151/30000 [2:37:11<53:01:42,  6.62s/it, lr: 1.0e-04 loss: 8.524e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1151/30000 [2:37:11<53:01:42,  6.62s/it, lr: 1.0e-04 loss: 8.524e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1152/30000 [2:37:11<52:56:14,  6.61s/it, lr: 1.0e-04 loss: 8.524e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1152/30000 [2:37:11<52:56:14,  6.61s/it, lr: 1.0e-04 loss: 8.524e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1152/30000 [2:37:17<52:56:14,  6.61s/it, lr: 1.0e-04 loss: 6.197e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1152/30000 [2:37:17<52:56:14,  6.61s/it, lr: 1.0e-04 loss: 6.197e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1153/30000 [2:37:17<52:50:49,  6.60s/it, lr: 1.0e-04 loss: 6.197e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1153/30000 [2:37:17<52:50:49,  6.60s/it, lr: 1.0e-04 loss: 6.197e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1153/30000 [2:37:27<52:50:49,  6.60s/it, lr: 1.0e-04 loss: 8.409e-02]
Xiang_wan_2_2_lora:   4%|#4                                     | 1153/30000 [2:37:27<52:50:49,  6.60s/it, lr: 1.0e-04 loss: 8.409e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1154/30000 [2:37:27<59:09:20,  7.38s/it, lr: 1.0e-04 loss: 8.409e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1154/30000 [2:37:27<59:09:20,  7.38s/it, lr: 1.0e-04 loss: 8.409e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1154/30000 [2:37:36<59:09:20,  7.38s/it, lr: 1.0e-04 loss: 3.272e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1154/30000 [2:37:36<59:09:20,  7.38s/it, lr: 1.0e-04 loss: 3.272e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1155/30000 [2:37:36<63:33:04,  7.93s/it, lr: 1.0e-04 loss: 3.272e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1155/30000 [2:37:36<63:33:04,  7.93s/it, lr: 1.0e-04 loss: 3.272e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1155/30000 [2:37:45<63:33:04,  7.93s/it, lr: 1.0e-04 loss: 5.235e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1155/30000 [2:37:45<63:33:04,  7.93s/it, lr: 1.0e-04 loss: 5.235e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1156/30000 [2:37:45<66:45:17,  8.33s/it, lr: 1.0e-04 loss: 5.235e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1156/30000 [2:37:45<66:45:17,  8.33s/it, lr: 1.0e-04 loss: 5.235e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1156/30000 [2:37:55<66:45:17,  8.33s/it, lr: 1.0e-04 loss: 5.770e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1156/30000 [2:37:55<66:45:17,  8.33s/it, lr: 1.0e-04 loss: 5.770e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1157/30000 [2:37:55<69:13:36,  8.64s/it, lr: 1.0e-04 loss: 5.770e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1157/30000 [2:37:55<69:13:36,  8.64s/it, lr: 1.0e-04 loss: 5.770e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1157/30000 [2:38:09<69:13:36,  8.64s/it, lr: 1.0e-04 loss: 6.532e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1157/30000 [2:38:09<69:13:36,  8.64s/it, lr: 1.0e-04 loss: 6.532e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1158/30000 [2:38:09<80:06:26, 10.00s/it, lr: 1.0e-04 loss: 6.532e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1158/30000 [2:38:09<80:06:26, 10.00s/it, lr: 1.0e-04 loss: 6.532e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1158/30000 [2:38:21<80:06:26, 10.00s/it, lr: 1.0e-04 loss: 8.225e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1158/30000 [2:38:21<80:06:26, 10.00s/it, lr: 1.0e-04 loss: 8.225e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1159/30000 [2:38:21<84:53:56, 10.60s/it, lr: 1.0e-04 loss: 8.225e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1159/30000 [2:38:21<84:53:56, 10.60s/it, lr: 1.0e-04 loss: 8.225e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1159/30000 [2:38:29<84:53:56, 10.60s/it, lr: 1.0e-04 loss: 9.440e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1159/30000 [2:38:29<84:53:56, 10.60s/it, lr: 1.0e-04 loss: 9.440e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1160/30000 [2:38:39<84:53:46, 10.60s/it, lr: 1.0e-04 loss: 8.907e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1160/30000 [2:38:39<84:53:46, 10.60s/it, lr: 1.0e-04 loss: 8.907e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1161/30000 [2:38:39<64:44:02,  8.08s/it, lr: 1.0e-04 loss: 8.907e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1161/30000 [2:38:39<64:44:02,  8.08s/it, lr: 1.0e-04 loss: 8.907e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1161/30000 [2:38:46<64:44:02,  8.08s/it, lr: 1.0e-04 loss: 1.027e-01]
Xiang_wan_2_2_lora:   4%|#5                                     | 1161/30000 [2:38:46<64:44:02,  8.08s/it, lr: 1.0e-04 loss: 1.027e-01]
Xiang_wan_2_2_lora:   4%|#5                                     | 1162/30000 [2:38:46<62:06:40,  7.75s/it, lr: 1.0e-04 loss: 1.027e-01]
Xiang_wan_2_2_lora:   4%|#5                                     | 1162/30000 [2:38:46<62:06:40,  7.75s/it, lr: 1.0e-04 loss: 1.027e-01]
Xiang_wan_2_2_lora:   4%|#5                                     | 1162/30000 [2:38:52<62:06:40,  7.75s/it, lr: 1.0e-04 loss: 8.553e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1162/30000 [2:38:52<62:06:40,  7.75s/it, lr: 1.0e-04 loss: 8.553e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1163/30000 [2:38:52<60:10:37,  7.51s/it, lr: 1.0e-04 loss: 8.553e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1163/30000 [2:38:52<60:10:37,  7.51s/it, lr: 1.0e-04 loss: 8.553e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1163/30000 [2:39:02<60:10:37,  7.51s/it, lr: 1.0e-04 loss: 6.691e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1163/30000 [2:39:02<60:10:37,  7.51s/it, lr: 1.0e-04 loss: 6.691e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1164/30000 [2:39:02<64:11:11,  8.01s/it, lr: 1.0e-04 loss: 6.691e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1164/30000 [2:39:02<64:11:11,  8.01s/it, lr: 1.0e-04 loss: 6.691e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1164/30000 [2:39:08<64:11:11,  8.01s/it, lr: 1.0e-04 loss: 4.995e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1164/30000 [2:39:08<64:11:11,  8.01s/it, lr: 1.0e-04 loss: 4.995e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1165/30000 [2:39:08<61:10:38,  7.64s/it, lr: 1.0e-04 loss: 4.995e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1165/30000 [2:39:08<61:10:38,  7.64s/it, lr: 1.0e-04 loss: 4.995e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1165/30000 [2:39:18<61:10:38,  7.64s/it, lr: 1.0e-04 loss: 5.648e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1165/30000 [2:39:18<61:10:38,  7.64s/it, lr: 1.0e-04 loss: 5.648e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1166/30000 [2:39:18<64:51:37,  8.10s/it, lr: 1.0e-04 loss: 5.648e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1166/30000 [2:39:18<64:51:37,  8.10s/it, lr: 1.0e-04 loss: 5.648e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1166/30000 [2:39:27<64:51:37,  8.10s/it, lr: 1.0e-04 loss: 1.090e-01]
Xiang_wan_2_2_lora:   4%|#5                                     | 1166/30000 [2:39:27<64:51:37,  8.10s/it, lr: 1.0e-04 loss: 1.090e-01]
Xiang_wan_2_2_lora:   4%|#5                                     | 1167/30000 [2:39:27<67:49:57,  8.47s/it, lr: 1.0e-04 loss: 1.090e-01]
Xiang_wan_2_2_lora:   4%|#5                                     | 1167/30000 [2:39:27<67:49:57,  8.47s/it, lr: 1.0e-04 loss: 1.090e-01]
Xiang_wan_2_2_lora:   4%|#5                                     | 1167/30000 [2:39:37<67:49:57,  8.47s/it, lr: 1.0e-04 loss: 8.299e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1167/30000 [2:39:37<67:49:57,  8.47s/it, lr: 1.0e-04 loss: 8.299e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1168/30000 [2:39:37<70:18:04,  8.78s/it, lr: 1.0e-04 loss: 8.299e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1168/30000 [2:39:37<70:18:04,  8.78s/it, lr: 1.0e-04 loss: 8.299e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1168/30000 [2:39:46<70:18:04,  8.78s/it, lr: 1.0e-04 loss: 4.336e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1168/30000 [2:39:46<70:18:04,  8.78s/it, lr: 1.0e-04 loss: 4.336e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1169/30000 [2:39:46<71:37:53,  8.94s/it, lr: 1.0e-04 loss: 4.336e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1169/30000 [2:39:46<71:37:53,  8.94s/it, lr: 1.0e-04 loss: 4.336e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1169/30000 [2:39:55<71:37:53,  8.94s/it, lr: 1.0e-04 loss: 9.584e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1169/30000 [2:39:55<71:37:53,  8.94s/it, lr: 1.0e-04 loss: 9.584e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1170/30000 [2:40:05<71:37:44,  8.94s/it, lr: 1.0e-04 loss: 4.717e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1170/30000 [2:40:05<71:37:44,  8.94s/it, lr: 1.0e-04 loss: 4.717e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1171/30000 [2:40:05<55:59:15,  6.99s/it, lr: 1.0e-04 loss: 4.717e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1171/30000 [2:40:05<55:59:15,  6.99s/it, lr: 1.0e-04 loss: 4.717e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1171/30000 [2:40:11<55:59:15,  6.99s/it, lr: 1.0e-04 loss: 9.460e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1171/30000 [2:40:11<55:59:15,  6.99s/it, lr: 1.0e-04 loss: 9.460e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1172/30000 [2:40:11<55:06:14,  6.88s/it, lr: 1.0e-04 loss: 9.460e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1172/30000 [2:40:11<55:06:14,  6.88s/it, lr: 1.0e-04 loss: 9.460e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1172/30000 [2:40:21<55:06:14,  6.88s/it, lr: 1.0e-04 loss: 1.335e-01]
Xiang_wan_2_2_lora:   4%|#5                                     | 1172/30000 [2:40:21<55:06:14,  6.88s/it, lr: 1.0e-04 loss: 1.335e-01]
Xiang_wan_2_2_lora:   4%|#5                                     | 1173/30000 [2:40:21<60:06:49,  7.51s/it, lr: 1.0e-04 loss: 1.335e-01]
Xiang_wan_2_2_lora:   4%|#5                                     | 1173/30000 [2:40:21<60:06:49,  7.51s/it, lr: 1.0e-04 loss: 1.335e-01]
Xiang_wan_2_2_lora:   4%|#5                                     | 1173/30000 [2:40:30<60:06:49,  7.51s/it, lr: 1.0e-04 loss: 7.301e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1173/30000 [2:40:30<60:06:49,  7.51s/it, lr: 1.0e-04 loss: 7.301e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1174/30000 [2:40:30<64:05:48,  8.00s/it, lr: 1.0e-04 loss: 7.301e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1174/30000 [2:40:30<64:05:48,  8.00s/it, lr: 1.0e-04 loss: 7.301e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1174/30000 [2:40:37<64:05:48,  8.00s/it, lr: 1.0e-04 loss: 8.505e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1174/30000 [2:40:37<64:05:48,  8.00s/it, lr: 1.0e-04 loss: 8.505e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1175/30000 [2:40:37<61:05:41,  7.63s/it, lr: 1.0e-04 loss: 8.505e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1175/30000 [2:40:37<61:05:41,  7.63s/it, lr: 1.0e-04 loss: 8.505e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1175/30000 [2:40:43<61:05:41,  7.63s/it, lr: 1.0e-04 loss: 1.063e-01]
Xiang_wan_2_2_lora:   4%|#5                                     | 1175/30000 [2:40:43<61:05:41,  7.63s/it, lr: 1.0e-04 loss: 1.063e-01]
Xiang_wan_2_2_lora:   4%|#5                                     | 1176/30000 [2:40:43<59:19:27,  7.41s/it, lr: 1.0e-04 loss: 1.063e-01]
Xiang_wan_2_2_lora:   4%|#5                                     | 1176/30000 [2:40:43<59:19:27,  7.41s/it, lr: 1.0e-04 loss: 1.063e-01]
Xiang_wan_2_2_lora:   4%|#5                                     | 1176/30000 [2:40:50<59:19:27,  7.41s/it, lr: 1.0e-04 loss: 5.114e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1176/30000 [2:40:50<59:19:27,  7.41s/it, lr: 1.0e-04 loss: 5.114e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1177/30000 [2:40:50<57:13:04,  7.15s/it, lr: 1.0e-04 loss: 5.114e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1177/30000 [2:40:50<57:13:04,  7.15s/it, lr: 1.0e-04 loss: 5.114e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1177/30000 [2:40:56<57:13:04,  7.15s/it, lr: 1.0e-04 loss: 8.533e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1177/30000 [2:40:56<57:13:04,  7.15s/it, lr: 1.0e-04 loss: 8.533e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1178/30000 [2:40:56<55:48:22,  6.97s/it, lr: 1.0e-04 loss: 8.533e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1178/30000 [2:40:56<55:48:22,  6.97s/it, lr: 1.0e-04 loss: 8.533e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1178/30000 [2:41:06<55:48:22,  6.97s/it, lr: 1.0e-04 loss: 4.676e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1178/30000 [2:41:06<55:48:22,  6.97s/it, lr: 1.0e-04 loss: 4.676e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1179/30000 [2:41:06<61:21:08,  7.66s/it, lr: 1.0e-04 loss: 4.676e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1179/30000 [2:41:06<61:21:08,  7.66s/it, lr: 1.0e-04 loss: 4.676e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1179/30000 [2:41:15<61:21:08,  7.66s/it, lr: 1.0e-04 loss: 6.205e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1179/30000 [2:41:15<61:21:08,  7.66s/it, lr: 1.0e-04 loss: 6.205e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1180/30000 [2:41:25<61:21:00,  7.66s/it, lr: 1.0e-04 loss: 5.416e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1180/30000 [2:41:25<61:21:00,  7.66s/it, lr: 1.0e-04 loss: 5.416e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1181/30000 [2:41:25<50:32:09,  6.31s/it, lr: 1.0e-04 loss: 5.416e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1181/30000 [2:41:25<50:32:09,  6.31s/it, lr: 1.0e-04 loss: 5.416e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1181/30000 [2:41:34<50:32:09,  6.31s/it, lr: 1.0e-04 loss: 5.840e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1181/30000 [2:41:34<50:32:09,  6.31s/it, lr: 1.0e-04 loss: 5.840e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1182/30000 [2:41:34<56:33:25,  7.07s/it, lr: 1.0e-04 loss: 5.840e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1182/30000 [2:41:34<56:33:25,  7.07s/it, lr: 1.0e-04 loss: 5.840e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1182/30000 [2:41:43<56:33:25,  7.07s/it, lr: 1.0e-04 loss: 8.009e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1182/30000 [2:41:43<56:33:25,  7.07s/it, lr: 1.0e-04 loss: 8.009e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1183/30000 [2:41:43<61:43:50,  7.71s/it, lr: 1.0e-04 loss: 8.009e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1183/30000 [2:41:43<61:43:50,  7.71s/it, lr: 1.0e-04 loss: 8.009e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1183/30000 [2:41:53<61:43:50,  7.71s/it, lr: 1.0e-04 loss: 9.272e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1183/30000 [2:41:53<61:43:50,  7.71s/it, lr: 1.0e-04 loss: 9.272e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1184/30000 [2:41:53<65:17:25,  8.16s/it, lr: 1.0e-04 loss: 9.272e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1184/30000 [2:41:53<65:17:25,  8.16s/it, lr: 1.0e-04 loss: 9.272e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1184/30000 [2:41:59<65:17:25,  8.16s/it, lr: 1.0e-04 loss: 6.520e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1184/30000 [2:41:59<65:17:25,  8.16s/it, lr: 1.0e-04 loss: 6.520e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1185/30000 [2:41:59<61:50:25,  7.73s/it, lr: 1.0e-04 loss: 6.520e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1185/30000 [2:41:59<61:50:25,  7.73s/it, lr: 1.0e-04 loss: 6.520e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1185/30000 [2:42:06<61:50:25,  7.73s/it, lr: 1.0e-04 loss: 8.356e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1185/30000 [2:42:06<61:50:25,  7.73s/it, lr: 1.0e-04 loss: 8.356e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1186/30000 [2:42:06<59:05:13,  7.38s/it, lr: 1.0e-04 loss: 8.356e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1186/30000 [2:42:06<59:05:13,  7.38s/it, lr: 1.0e-04 loss: 8.356e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1186/30000 [2:42:13<59:05:13,  7.38s/it, lr: 1.0e-04 loss: 6.177e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1186/30000 [2:42:13<59:05:13,  7.38s/it, lr: 1.0e-04 loss: 6.177e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1187/30000 [2:42:13<57:17:39,  7.16s/it, lr: 1.0e-04 loss: 6.177e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1187/30000 [2:42:13<57:17:39,  7.16s/it, lr: 1.0e-04 loss: 6.177e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1187/30000 [2:42:22<57:17:39,  7.16s/it, lr: 1.0e-04 loss: 6.500e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1187/30000 [2:42:22<57:17:39,  7.16s/it, lr: 1.0e-04 loss: 6.500e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1188/30000 [2:42:22<62:19:34,  7.79s/it, lr: 1.0e-04 loss: 6.500e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1188/30000 [2:42:22<62:19:34,  7.79s/it, lr: 1.0e-04 loss: 6.500e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1188/30000 [2:42:31<62:19:34,  7.79s/it, lr: 1.0e-04 loss: 5.606e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1188/30000 [2:42:31<62:19:34,  7.79s/it, lr: 1.0e-04 loss: 5.606e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1189/30000 [2:42:31<66:07:31,  8.26s/it, lr: 1.0e-04 loss: 5.606e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1189/30000 [2:42:31<66:07:31,  8.26s/it, lr: 1.0e-04 loss: 5.606e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1189/30000 [2:42:38<66:07:31,  8.26s/it, lr: 1.0e-04 loss: 1.091e-01]
Xiang_wan_2_2_lora:   4%|#5                                     | 1189/30000 [2:42:38<66:07:31,  8.26s/it, lr: 1.0e-04 loss: 1.091e-01]
Xiang_wan_2_2_lora:   4%|#5                                     | 1190/30000 [2:42:47<66:07:23,  8.26s/it, lr: 1.0e-04 loss: 6.484e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1190/30000 [2:42:47<66:07:23,  8.26s/it, lr: 1.0e-04 loss: 6.484e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1191/30000 [2:42:47<52:59:07,  6.62s/it, lr: 1.0e-04 loss: 6.484e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1191/30000 [2:42:47<52:59:07,  6.62s/it, lr: 1.0e-04 loss: 6.484e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1191/30000 [2:42:54<52:59:07,  6.62s/it, lr: 1.0e-04 loss: 6.929e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1191/30000 [2:42:54<52:59:07,  6.62s/it, lr: 1.0e-04 loss: 6.929e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1192/30000 [2:42:54<52:51:37,  6.61s/it, lr: 1.0e-04 loss: 6.929e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1192/30000 [2:42:54<52:51:37,  6.61s/it, lr: 1.0e-04 loss: 6.929e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1192/30000 [2:43:01<52:51:37,  6.61s/it, lr: 1.0e-04 loss: 7.101e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1192/30000 [2:43:01<52:51:37,  6.61s/it, lr: 1.0e-04 loss: 7.101e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1193/30000 [2:43:01<53:00:41,  6.62s/it, lr: 1.0e-04 loss: 7.101e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1193/30000 [2:43:01<53:00:41,  6.62s/it, lr: 1.0e-04 loss: 7.101e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1193/30000 [2:43:07<53:00:41,  6.62s/it, lr: 1.0e-04 loss: 8.117e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1193/30000 [2:43:07<53:00:41,  6.62s/it, lr: 1.0e-04 loss: 8.117e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1194/30000 [2:43:07<52:52:58,  6.61s/it, lr: 1.0e-04 loss: 8.117e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1194/30000 [2:43:07<52:52:58,  6.61s/it, lr: 1.0e-04 loss: 8.117e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1194/30000 [2:43:14<52:52:58,  6.61s/it, lr: 1.0e-04 loss: 8.852e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1194/30000 [2:43:14<52:52:58,  6.61s/it, lr: 1.0e-04 loss: 8.852e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1195/30000 [2:43:14<52:55:53,  6.62s/it, lr: 1.0e-04 loss: 8.852e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1195/30000 [2:43:14<52:55:53,  6.62s/it, lr: 1.0e-04 loss: 8.852e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1195/30000 [2:43:23<52:55:53,  6.62s/it, lr: 1.0e-04 loss: 6.278e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1195/30000 [2:43:23<52:55:53,  6.62s/it, lr: 1.0e-04 loss: 6.278e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1196/30000 [2:43:23<58:59:38,  7.37s/it, lr: 1.0e-04 loss: 6.278e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1196/30000 [2:43:23<58:59:38,  7.37s/it, lr: 1.0e-04 loss: 6.278e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1196/30000 [2:43:33<58:59:38,  7.37s/it, lr: 1.0e-04 loss: 6.518e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1196/30000 [2:43:33<58:59:38,  7.37s/it, lr: 1.0e-04 loss: 6.518e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1197/30000 [2:43:33<64:19:30,  8.04s/it, lr: 1.0e-04 loss: 6.518e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1197/30000 [2:43:33<64:19:30,  8.04s/it, lr: 1.0e-04 loss: 6.518e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1197/30000 [2:43:42<64:19:30,  8.04s/it, lr: 1.0e-04 loss: 6.442e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1197/30000 [2:43:42<64:19:30,  8.04s/it, lr: 1.0e-04 loss: 6.442e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1198/30000 [2:43:42<67:20:23,  8.42s/it, lr: 1.0e-04 loss: 6.442e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1198/30000 [2:43:42<67:20:23,  8.42s/it, lr: 1.0e-04 loss: 6.442e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1198/30000 [2:43:49<67:20:23,  8.42s/it, lr: 1.0e-04 loss: 5.621e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1198/30000 [2:43:49<67:20:23,  8.42s/it, lr: 1.0e-04 loss: 5.621e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1199/30000 [2:43:49<63:03:18,  7.88s/it, lr: 1.0e-04 loss: 5.621e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1199/30000 [2:43:49<63:03:18,  7.88s/it, lr: 1.0e-04 loss: 5.621e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1199/30000 [2:43:58<63:03:18,  7.88s/it, lr: 1.0e-04 loss: 1.176e-01]
Xiang_wan_2_2_lora:   4%|#5                                     | 1199/30000 [2:43:58<63:03:18,  7.88s/it, lr: 1.0e-04 loss: 1.176e-01]
Xiang_wan_2_2_lora:   4%|#5                                     | 1200/30000 [2:44:07<63:03:10,  7.88s/it, lr: 1.0e-04 loss: 4.686e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1200/30000 [2:44:07<63:03:10,  7.88s/it, lr: 1.0e-04 loss: 4.686e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1201/30000 [2:44:07<51:24:00,  6.43s/it, lr: 1.0e-04 loss: 4.686e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1201/30000 [2:44:07<51:24:00,  6.43s/it, lr: 1.0e-04 loss: 4.686e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1201/30000 [2:44:14<51:24:00,  6.43s/it, lr: 1.0e-04 loss: 1.000e-01]
Xiang_wan_2_2_lora:   4%|#5                                     | 1201/30000 [2:44:14<51:24:00,  6.43s/it, lr: 1.0e-04 loss: 1.000e-01]
Xiang_wan_2_2_lora:   4%|#5                                     | 1202/30000 [2:44:14<51:44:21,  6.47s/it, lr: 1.0e-04 loss: 1.000e-01]
Xiang_wan_2_2_lora:   4%|#5                                     | 1202/30000 [2:44:14<51:44:21,  6.47s/it, lr: 1.0e-04 loss: 1.000e-01]
Xiang_wan_2_2_lora:   4%|#5                                     | 1202/30000 [2:44:21<51:44:21,  6.47s/it, lr: 1.0e-04 loss: 8.212e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1202/30000 [2:44:21<51:44:21,  6.47s/it, lr: 1.0e-04 loss: 8.212e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1203/30000 [2:44:21<52:00:59,  6.50s/it, lr: 1.0e-04 loss: 8.212e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1203/30000 [2:44:21<52:00:59,  6.50s/it, lr: 1.0e-04 loss: 8.212e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1203/30000 [2:44:27<52:00:59,  6.50s/it, lr: 1.0e-04 loss: 7.406e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1203/30000 [2:44:27<52:00:59,  6.50s/it, lr: 1.0e-04 loss: 7.406e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1204/30000 [2:44:27<52:08:09,  6.52s/it, lr: 1.0e-04 loss: 7.406e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1204/30000 [2:44:27<52:08:09,  6.52s/it, lr: 1.0e-04 loss: 7.406e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1204/30000 [2:44:37<52:08:09,  6.52s/it, lr: 1.0e-04 loss: 4.068e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1204/30000 [2:44:37<52:08:09,  6.52s/it, lr: 1.0e-04 loss: 4.068e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1205/30000 [2:44:37<58:52:47,  7.36s/it, lr: 1.0e-04 loss: 4.068e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1205/30000 [2:44:37<58:52:47,  7.36s/it, lr: 1.0e-04 loss: 4.068e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1205/30000 [2:44:46<58:52:47,  7.36s/it, lr: 1.0e-04 loss: 7.105e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1205/30000 [2:44:46<58:52:47,  7.36s/it, lr: 1.0e-04 loss: 7.105e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1206/30000 [2:44:46<63:18:03,  7.91s/it, lr: 1.0e-04 loss: 7.105e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1206/30000 [2:44:46<63:18:03,  7.91s/it, lr: 1.0e-04 loss: 7.105e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1206/30000 [2:44:55<63:18:03,  7.91s/it, lr: 1.0e-04 loss: 1.390e-01]
Xiang_wan_2_2_lora:   4%|#5                                     | 1206/30000 [2:44:55<63:18:03,  7.91s/it, lr: 1.0e-04 loss: 1.390e-01]
Xiang_wan_2_2_lora:   4%|#5                                     | 1207/30000 [2:44:55<66:38:32,  8.33s/it, lr: 1.0e-04 loss: 1.390e-01]
Xiang_wan_2_2_lora:   4%|#5                                     | 1207/30000 [2:44:55<66:38:32,  8.33s/it, lr: 1.0e-04 loss: 1.390e-01]
Xiang_wan_2_2_lora:   4%|#5                                     | 1207/30000 [2:45:02<66:38:32,  8.33s/it, lr: 1.0e-04 loss: 9.005e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1207/30000 [2:45:02<66:38:32,  8.33s/it, lr: 1.0e-04 loss: 9.005e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1208/30000 [2:45:02<62:37:58,  7.83s/it, lr: 1.0e-04 loss: 9.005e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1208/30000 [2:45:02<62:37:58,  7.83s/it, lr: 1.0e-04 loss: 9.005e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1208/30000 [2:45:11<62:37:58,  7.83s/it, lr: 1.0e-04 loss: 6.897e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1208/30000 [2:45:11<62:37:58,  7.83s/it, lr: 1.0e-04 loss: 6.897e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1209/30000 [2:45:11<66:06:55,  8.27s/it, lr: 1.0e-04 loss: 6.897e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1209/30000 [2:45:11<66:06:55,  8.27s/it, lr: 1.0e-04 loss: 6.897e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1209/30000 [2:45:18<66:06:55,  8.27s/it, lr: 1.0e-04 loss: 1.153e-01]
Xiang_wan_2_2_lora:   4%|#5                                     | 1209/30000 [2:45:18<66:06:55,  8.27s/it, lr: 1.0e-04 loss: 1.153e-01]
Xiang_wan_2_2_lora:   4%|#5                                     | 1210/30000 [2:45:27<66:06:47,  8.27s/it, lr: 1.0e-04 loss: 2.521e-01]
Xiang_wan_2_2_lora:   4%|#5                                     | 1210/30000 [2:45:27<66:06:47,  8.27s/it, lr: 1.0e-04 loss: 2.521e-01]
Xiang_wan_2_2_lora:   4%|#5                                     | 1211/30000 [2:45:27<52:53:42,  6.61s/it, lr: 1.0e-04 loss: 2.521e-01]
Xiang_wan_2_2_lora:   4%|#5                                     | 1211/30000 [2:45:27<52:53:42,  6.61s/it, lr: 1.0e-04 loss: 2.521e-01]
Xiang_wan_2_2_lora:   4%|#5                                     | 1211/30000 [2:45:37<52:53:42,  6.61s/it, lr: 1.0e-04 loss: 5.307e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1211/30000 [2:45:37<52:53:42,  6.61s/it, lr: 1.0e-04 loss: 5.307e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1212/30000 [2:45:37<58:51:40,  7.36s/it, lr: 1.0e-04 loss: 5.307e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1212/30000 [2:45:37<58:51:40,  7.36s/it, lr: 1.0e-04 loss: 5.307e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1212/30000 [2:45:46<58:51:40,  7.36s/it, lr: 1.0e-04 loss: 5.064e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1212/30000 [2:45:46<58:51:40,  7.36s/it, lr: 1.0e-04 loss: 5.064e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1213/30000 [2:45:46<62:55:58,  7.87s/it, lr: 1.0e-04 loss: 5.064e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1213/30000 [2:45:46<62:55:58,  7.87s/it, lr: 1.0e-04 loss: 5.064e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1213/30000 [2:45:53<62:55:58,  7.87s/it, lr: 1.0e-04 loss: 8.845e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1213/30000 [2:45:53<62:55:58,  7.87s/it, lr: 1.0e-04 loss: 8.845e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1214/30000 [2:45:53<60:06:26,  7.52s/it, lr: 1.0e-04 loss: 8.845e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1214/30000 [2:45:53<60:06:26,  7.52s/it, lr: 1.0e-04 loss: 8.845e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1214/30000 [2:45:59<60:06:26,  7.52s/it, lr: 1.0e-04 loss: 8.823e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1214/30000 [2:45:59<60:06:26,  7.52s/it, lr: 1.0e-04 loss: 8.823e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1215/30000 [2:45:59<57:53:29,  7.24s/it, lr: 1.0e-04 loss: 8.823e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1215/30000 [2:45:59<57:53:29,  7.24s/it, lr: 1.0e-04 loss: 8.823e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1215/30000 [2:46:09<57:53:29,  7.24s/it, lr: 1.0e-04 loss: 1.050e-01]
Xiang_wan_2_2_lora:   4%|#5                                     | 1215/30000 [2:46:09<57:53:29,  7.24s/it, lr: 1.0e-04 loss: 1.050e-01]
Xiang_wan_2_2_lora:   4%|#5                                     | 1216/30000 [2:46:09<62:41:48,  7.84s/it, lr: 1.0e-04 loss: 1.050e-01]
Xiang_wan_2_2_lora:   4%|#5                                     | 1216/30000 [2:46:09<62:41:48,  7.84s/it, lr: 1.0e-04 loss: 1.050e-01]
Xiang_wan_2_2_lora:   4%|#5                                     | 1216/30000 [2:46:18<62:41:48,  7.84s/it, lr: 1.0e-04 loss: 5.883e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1216/30000 [2:46:18<62:41:48,  7.84s/it, lr: 1.0e-04 loss: 5.883e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1217/30000 [2:46:18<66:13:15,  8.28s/it, lr: 1.0e-04 loss: 5.883e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1217/30000 [2:46:18<66:13:15,  8.28s/it, lr: 1.0e-04 loss: 5.883e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1217/30000 [2:46:25<66:13:15,  8.28s/it, lr: 1.0e-04 loss: 7.172e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1217/30000 [2:46:25<66:13:15,  8.28s/it, lr: 1.0e-04 loss: 7.172e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1218/30000 [2:46:25<62:16:33,  7.79s/it, lr: 1.0e-04 loss: 7.172e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1218/30000 [2:46:25<62:16:33,  7.79s/it, lr: 1.0e-04 loss: 7.172e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1218/30000 [2:46:31<62:16:33,  7.79s/it, lr: 1.0e-04 loss: 7.735e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1218/30000 [2:46:31<62:16:33,  7.79s/it, lr: 1.0e-04 loss: 7.735e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1219/30000 [2:46:31<59:57:58,  7.50s/it, lr: 1.0e-04 loss: 7.735e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1219/30000 [2:46:31<59:57:58,  7.50s/it, lr: 1.0e-04 loss: 7.735e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1219/30000 [2:46:41<59:57:58,  7.50s/it, lr: 1.0e-04 loss: 8.311e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1219/30000 [2:46:41<59:57:58,  7.50s/it, lr: 1.0e-04 loss: 8.311e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1220/30000 [2:46:50<59:57:50,  7.50s/it, lr: 1.0e-04 loss: 2.451e-01]
Xiang_wan_2_2_lora:   4%|#5                                     | 1220/30000 [2:46:50<59:57:50,  7.50s/it, lr: 1.0e-04 loss: 2.451e-01]
Xiang_wan_2_2_lora:   4%|#5                                     | 1221/30000 [2:46:50<49:37:00,  6.21s/it, lr: 1.0e-04 loss: 2.451e-01]
Xiang_wan_2_2_lora:   4%|#5                                     | 1221/30000 [2:46:50<49:37:00,  6.21s/it, lr: 1.0e-04 loss: 2.451e-01]
Xiang_wan_2_2_lora:   4%|#5                                     | 1221/30000 [2:46:57<49:37:00,  6.21s/it, lr: 1.0e-04 loss: 7.248e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1221/30000 [2:46:57<49:37:00,  6.21s/it, lr: 1.0e-04 loss: 7.248e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1222/30000 [2:46:57<50:31:00,  6.32s/it, lr: 1.0e-04 loss: 7.248e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1222/30000 [2:46:57<50:31:00,  6.32s/it, lr: 1.0e-04 loss: 7.248e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1222/30000 [2:47:06<50:31:00,  6.32s/it, lr: 1.0e-04 loss: 4.415e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1222/30000 [2:47:06<50:31:00,  6.32s/it, lr: 1.0e-04 loss: 4.415e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1223/30000 [2:47:06<56:43:58,  7.10s/it, lr: 1.0e-04 loss: 4.415e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1223/30000 [2:47:06<56:43:58,  7.10s/it, lr: 1.0e-04 loss: 4.415e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1223/30000 [2:47:15<56:43:58,  7.10s/it, lr: 1.0e-04 loss: 8.461e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1223/30000 [2:47:15<56:43:58,  7.10s/it, lr: 1.0e-04 loss: 8.461e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1224/30000 [2:47:15<61:37:37,  7.71s/it, lr: 1.0e-04 loss: 8.461e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1224/30000 [2:47:15<61:37:37,  7.71s/it, lr: 1.0e-04 loss: 8.461e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1224/30000 [2:47:25<61:37:37,  7.71s/it, lr: 1.0e-04 loss: 4.368e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1224/30000 [2:47:25<61:37:37,  7.71s/it, lr: 1.0e-04 loss: 4.368e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1225/30000 [2:47:25<65:12:08,  8.16s/it, lr: 1.0e-04 loss: 4.368e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1225/30000 [2:47:25<65:12:08,  8.16s/it, lr: 1.0e-04 loss: 4.368e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1225/30000 [2:47:34<65:12:08,  8.16s/it, lr: 1.0e-04 loss: 9.106e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1225/30000 [2:47:34<65:12:08,  8.16s/it, lr: 1.0e-04 loss: 9.106e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1226/30000 [2:47:34<67:57:10,  8.50s/it, lr: 1.0e-04 loss: 9.106e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1226/30000 [2:47:34<67:57:10,  8.50s/it, lr: 1.0e-04 loss: 9.106e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1226/30000 [2:47:44<67:57:10,  8.50s/it, lr: 1.0e-04 loss: 1.216e-01]
Xiang_wan_2_2_lora:   4%|#5                                     | 1226/30000 [2:47:44<67:57:10,  8.50s/it, lr: 1.0e-04 loss: 1.216e-01]
Xiang_wan_2_2_lora:   4%|#5                                     | 1227/30000 [2:47:44<70:19:41,  8.80s/it, lr: 1.0e-04 loss: 1.216e-01]
Xiang_wan_2_2_lora:   4%|#5                                     | 1227/30000 [2:47:44<70:19:41,  8.80s/it, lr: 1.0e-04 loss: 1.216e-01]
Xiang_wan_2_2_lora:   4%|#5                                     | 1227/30000 [2:47:53<70:19:41,  8.80s/it, lr: 1.0e-04 loss: 7.290e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1227/30000 [2:47:53<70:19:41,  8.80s/it, lr: 1.0e-04 loss: 7.290e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1228/30000 [2:47:53<71:43:03,  8.97s/it, lr: 1.0e-04 loss: 7.290e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1228/30000 [2:47:53<71:43:03,  8.97s/it, lr: 1.0e-04 loss: 7.290e-02]
Xiang_wan_2_2_lora:   4%|#5                                     | 1228/30000 [2:48:02<71:43:03,  8.97s/it, lr: 1.0e-04 loss: 1.002e-01]
Xiang_wan_2_2_lora:   4%|#5                                     | 1228/30000 [2:48:02<71:43:03,  8.97s/it, lr: 1.0e-04 loss: 1.002e-01]
Xiang_wan_2_2_lora:   4%|#5                                     | 1229/30000 [2:48:02<72:37:15,  9.09s/it, lr: 1.0e-04 loss: 1.002e-01]
Xiang_wan_2_2_lora:   4%|#5                                     | 1229/30000 [2:48:02<72:37:15,  9.09s/it, lr: 1.0e-04 loss: 1.002e-01]
Xiang_wan_2_2_lora:   4%|#5                                     | 1229/30000 [2:48:12<72:37:15,  9.09s/it, lr: 1.0e-04 loss: 1.017e-01]
Xiang_wan_2_2_lora:   4%|#5                                     | 1229/30000 [2:48:12<72:37:15,  9.09s/it, lr: 1.0e-04 loss: 1.017e-01]
Xiang_wan_2_2_lora:   4%|#5                                     | 1230/30000 [2:48:21<72:37:06,  9.09s/it, lr: 1.0e-04 loss: 2.490e-01]
Xiang_wan_2_2_lora:   4%|#5                                     | 1230/30000 [2:48:21<72:37:06,  9.09s/it, lr: 1.0e-04 loss: 2.490e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1231/30000 [2:48:21<56:24:31,  7.06s/it, lr: 1.0e-04 loss: 2.490e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1231/30000 [2:48:21<56:24:31,  7.06s/it, lr: 1.0e-04 loss: 2.490e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1231/30000 [2:48:30<56:24:31,  7.06s/it, lr: 1.0e-04 loss: 9.502e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1231/30000 [2:48:30<56:24:31,  7.06s/it, lr: 1.0e-04 loss: 9.502e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1232/30000 [2:48:30<61:00:15,  7.63s/it, lr: 1.0e-04 loss: 9.502e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1232/30000 [2:48:30<61:00:15,  7.63s/it, lr: 1.0e-04 loss: 9.502e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1232/30000 [2:48:40<61:00:15,  7.63s/it, lr: 1.0e-04 loss: 6.119e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1232/30000 [2:48:40<61:00:15,  7.63s/it, lr: 1.0e-04 loss: 6.119e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1233/30000 [2:48:40<64:36:51,  8.09s/it, lr: 1.0e-04 loss: 6.119e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1233/30000 [2:48:40<64:36:51,  8.09s/it, lr: 1.0e-04 loss: 6.119e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1233/30000 [2:48:47<64:36:51,  8.09s/it, lr: 1.0e-04 loss: 7.435e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1233/30000 [2:48:47<64:36:51,  8.09s/it, lr: 1.0e-04 loss: 7.435e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1234/30000 [2:48:47<62:05:18,  7.77s/it, lr: 1.0e-04 loss: 7.435e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1234/30000 [2:48:47<62:05:18,  7.77s/it, lr: 1.0e-04 loss: 7.435e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1234/30000 [2:48:53<62:05:18,  7.77s/it, lr: 1.0e-04 loss: 8.920e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1234/30000 [2:48:53<62:05:18,  7.77s/it, lr: 1.0e-04 loss: 8.920e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1235/30000 [2:48:53<59:22:45,  7.43s/it, lr: 1.0e-04 loss: 8.920e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1235/30000 [2:48:53<59:22:45,  7.43s/it, lr: 1.0e-04 loss: 8.920e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1235/30000 [2:49:03<59:22:45,  7.43s/it, lr: 1.0e-04 loss: 7.212e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1235/30000 [2:49:03<59:22:45,  7.43s/it, lr: 1.0e-04 loss: 7.212e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1236/30000 [2:49:03<63:45:52,  7.98s/it, lr: 1.0e-04 loss: 7.212e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1236/30000 [2:49:03<63:45:52,  7.98s/it, lr: 1.0e-04 loss: 7.212e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1236/30000 [2:49:09<63:45:52,  7.98s/it, lr: 1.0e-04 loss: 1.326e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1236/30000 [2:49:09<63:45:52,  7.98s/it, lr: 1.0e-04 loss: 1.326e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1237/30000 [2:49:09<60:31:30,  7.58s/it, lr: 1.0e-04 loss: 1.326e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1237/30000 [2:49:09<60:31:30,  7.58s/it, lr: 1.0e-04 loss: 1.326e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1237/30000 [2:49:19<60:31:30,  7.58s/it, lr: 1.0e-04 loss: 5.673e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1237/30000 [2:49:19<60:31:30,  7.58s/it, lr: 1.0e-04 loss: 5.673e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1238/30000 [2:49:19<64:29:48,  8.07s/it, lr: 1.0e-04 loss: 5.673e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1238/30000 [2:49:19<64:29:48,  8.07s/it, lr: 1.0e-04 loss: 5.673e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1238/30000 [2:49:28<64:29:48,  8.07s/it, lr: 1.0e-04 loss: 4.807e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1238/30000 [2:49:28<64:29:48,  8.07s/it, lr: 1.0e-04 loss: 4.807e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1239/30000 [2:49:28<67:33:50,  8.46s/it, lr: 1.0e-04 loss: 4.807e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1239/30000 [2:49:28<67:33:50,  8.46s/it, lr: 1.0e-04 loss: 4.807e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1239/30000 [2:49:37<67:33:50,  8.46s/it, lr: 1.0e-04 loss: 8.149e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1239/30000 [2:49:37<67:33:50,  8.46s/it, lr: 1.0e-04 loss: 8.149e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1240/30000 [2:49:44<67:33:42,  8.46s/it, lr: 1.0e-04 loss: 7.967e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1240/30000 [2:49:44<67:33:42,  8.46s/it, lr: 1.0e-04 loss: 7.967e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1241/30000 [2:49:44<48:45:26,  6.10s/it, lr: 1.0e-04 loss: 7.967e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1241/30000 [2:49:44<48:45:26,  6.10s/it, lr: 1.0e-04 loss: 7.967e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1241/30000 [2:49:51<48:45:26,  6.10s/it, lr: 1.0e-04 loss: 8.765e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1241/30000 [2:49:51<48:45:26,  6.10s/it, lr: 1.0e-04 loss: 8.765e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1242/30000 [2:49:51<50:16:53,  6.29s/it, lr: 1.0e-04 loss: 8.765e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1242/30000 [2:49:51<50:16:53,  6.29s/it, lr: 1.0e-04 loss: 8.765e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1242/30000 [2:49:57<50:16:53,  6.29s/it, lr: 1.0e-04 loss: 5.792e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1242/30000 [2:49:57<50:16:53,  6.29s/it, lr: 1.0e-04 loss: 5.792e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1243/30000 [2:49:57<51:01:06,  6.39s/it, lr: 1.0e-04 loss: 5.792e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1243/30000 [2:49:57<51:01:06,  6.39s/it, lr: 1.0e-04 loss: 5.792e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1243/30000 [2:50:04<51:01:06,  6.39s/it, lr: 1.0e-04 loss: 8.994e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1243/30000 [2:50:04<51:01:06,  6.39s/it, lr: 1.0e-04 loss: 8.994e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1244/30000 [2:50:04<51:19:04,  6.42s/it, lr: 1.0e-04 loss: 8.994e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1244/30000 [2:50:04<51:19:04,  6.42s/it, lr: 1.0e-04 loss: 8.994e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1244/30000 [2:50:13<51:19:04,  6.42s/it, lr: 1.0e-04 loss: 8.852e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1244/30000 [2:50:13<51:19:04,  6.42s/it, lr: 1.0e-04 loss: 8.852e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1245/30000 [2:50:13<57:54:18,  7.25s/it, lr: 1.0e-04 loss: 8.852e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1245/30000 [2:50:13<57:54:18,  7.25s/it, lr: 1.0e-04 loss: 8.852e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1245/30000 [2:50:23<57:54:18,  7.25s/it, lr: 1.0e-04 loss: 5.754e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1245/30000 [2:50:23<57:54:18,  7.25s/it, lr: 1.0e-04 loss: 5.754e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1246/30000 [2:50:23<62:41:45,  7.85s/it, lr: 1.0e-04 loss: 5.754e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1246/30000 [2:50:23<62:41:45,  7.85s/it, lr: 1.0e-04 loss: 5.754e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1246/30000 [2:50:29<62:41:45,  7.85s/it, lr: 1.0e-04 loss: 5.532e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1246/30000 [2:50:29<62:41:45,  7.85s/it, lr: 1.0e-04 loss: 5.532e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1247/30000 [2:50:29<59:53:26,  7.50s/it, lr: 1.0e-04 loss: 5.532e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1247/30000 [2:50:29<59:53:26,  7.50s/it, lr: 1.0e-04 loss: 5.532e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1247/30000 [2:50:39<59:53:26,  7.50s/it, lr: 1.0e-04 loss: 6.436e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1247/30000 [2:50:39<59:53:26,  7.50s/it, lr: 1.0e-04 loss: 6.436e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1248/30000 [2:50:39<64:08:55,  8.03s/it, lr: 1.0e-04 loss: 6.436e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1248/30000 [2:50:39<64:08:55,  8.03s/it, lr: 1.0e-04 loss: 6.436e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1248/30000 [2:50:48<64:08:55,  8.03s/it, lr: 1.0e-04 loss: 4.861e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1248/30000 [2:50:48<64:08:55,  8.03s/it, lr: 1.0e-04 loss: 4.861e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1249/30000 [2:50:48<67:50:08,  8.49s/it, lr: 1.0e-04 loss: 4.861e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1249/30000 [2:50:48<67:50:08,  8.49s/it, lr: 1.0e-04 loss: 4.861e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1249/30000 [2:50:58<67:50:08,  8.49s/it, lr: 1.0e-04 loss: 5.980e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1249/30000 [2:50:58<67:50:08,  8.49s/it, lr: 1.0e-04 loss: 5.980e-02]
+Saving at step 1250
+Saved checkpoint to /home/featurize/ai-toolkit/output/Xiang_wan_2_2_lora/Xiang_wan_2_2_lora_000001250.safetensors
+Saved optimizer to /home/featurize/ai-toolkit/output/Xiang_wan_2_2_lora/optimizer.pt
+
Xiang_wan_2_2_lora:   4%|#6                                     | 1250/30000 [2:51:07<67:49:59,  8.49s/it, lr: 1.0e-04 loss: 3.865e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1250/30000 [2:51:07<67:49:59,  8.49s/it, lr: 1.0e-04 loss: 3.865e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1251/30000 [2:51:07<54:30:43,  6.83s/it, lr: 1.0e-04 loss: 3.865e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1251/30000 [2:51:07<54:30:43,  6.83s/it, lr: 1.0e-04 loss: 3.865e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1251/30000 [2:51:14<54:30:43,  6.83s/it, lr: 1.0e-04 loss: 1.096e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1251/30000 [2:51:14<54:30:43,  6.83s/it, lr: 1.0e-04 loss: 1.096e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1252/30000 [2:51:14<54:12:12,  6.79s/it, lr: 1.0e-04 loss: 1.096e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1252/30000 [2:51:14<54:12:12,  6.79s/it, lr: 1.0e-04 loss: 1.096e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1252/30000 [2:51:23<54:12:12,  6.79s/it, lr: 1.0e-04 loss: 3.713e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1252/30000 [2:51:23<54:12:12,  6.79s/it, lr: 1.0e-04 loss: 3.713e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1253/30000 [2:51:23<59:30:44,  7.45s/it, lr: 1.0e-04 loss: 3.713e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1253/30000 [2:51:23<59:30:44,  7.45s/it, lr: 1.0e-04 loss: 3.713e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1253/30000 [2:51:30<59:30:44,  7.45s/it, lr: 1.0e-04 loss: 8.646e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1253/30000 [2:51:30<59:30:44,  7.45s/it, lr: 1.0e-04 loss: 8.646e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1254/30000 [2:51:30<57:37:31,  7.22s/it, lr: 1.0e-04 loss: 8.646e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1254/30000 [2:51:30<57:37:31,  7.22s/it, lr: 1.0e-04 loss: 8.646e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1254/30000 [2:51:36<57:37:31,  7.22s/it, lr: 1.0e-04 loss: 1.858e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1254/30000 [2:51:36<57:37:31,  7.22s/it, lr: 1.0e-04 loss: 1.858e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1255/30000 [2:51:36<56:14:44,  7.04s/it, lr: 1.0e-04 loss: 1.858e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1255/30000 [2:51:36<56:14:44,  7.04s/it, lr: 1.0e-04 loss: 1.858e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1255/30000 [2:51:43<56:14:44,  7.04s/it, lr: 1.0e-04 loss: 6.278e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1255/30000 [2:51:43<56:14:44,  7.04s/it, lr: 1.0e-04 loss: 6.278e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1256/30000 [2:51:43<55:09:23,  6.91s/it, lr: 1.0e-04 loss: 6.278e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1256/30000 [2:51:43<55:09:23,  6.91s/it, lr: 1.0e-04 loss: 6.278e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1256/30000 [2:51:52<55:09:23,  6.91s/it, lr: 1.0e-04 loss: 4.490e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1256/30000 [2:51:52<55:09:23,  6.91s/it, lr: 1.0e-04 loss: 4.490e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1257/30000 [2:51:52<60:51:19,  7.62s/it, lr: 1.0e-04 loss: 4.490e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1257/30000 [2:51:52<60:51:19,  7.62s/it, lr: 1.0e-04 loss: 4.490e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1257/30000 [2:51:59<60:51:19,  7.62s/it, lr: 1.0e-04 loss: 8.360e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1257/30000 [2:51:59<60:51:19,  7.62s/it, lr: 1.0e-04 loss: 8.360e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1258/30000 [2:51:59<59:02:13,  7.39s/it, lr: 1.0e-04 loss: 8.360e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1258/30000 [2:51:59<59:02:13,  7.39s/it, lr: 1.0e-04 loss: 8.360e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1258/30000 [2:52:09<59:02:13,  7.39s/it, lr: 1.0e-04 loss: 9.097e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1258/30000 [2:52:09<59:02:13,  7.39s/it, lr: 1.0e-04 loss: 9.097e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1259/30000 [2:52:09<63:33:10,  7.96s/it, lr: 1.0e-04 loss: 9.097e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1259/30000 [2:52:09<63:33:10,  7.96s/it, lr: 1.0e-04 loss: 9.097e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1259/30000 [2:52:15<63:33:10,  7.96s/it, lr: 1.0e-04 loss: 1.059e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1259/30000 [2:52:15<63:33:10,  7.96s/it, lr: 1.0e-04 loss: 1.059e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1260/30000 [2:52:22<63:33:02,  7.96s/it, lr: 1.0e-04 loss: 7.122e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1260/30000 [2:52:22<63:33:02,  7.96s/it, lr: 1.0e-04 loss: 7.122e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1261/30000 [2:52:22<46:30:13,  5.83s/it, lr: 1.0e-04 loss: 7.122e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1261/30000 [2:52:22<46:30:13,  5.83s/it, lr: 1.0e-04 loss: 7.122e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1261/30000 [2:52:28<46:30:13,  5.83s/it, lr: 1.0e-04 loss: 7.732e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1261/30000 [2:52:28<46:30:13,  5.83s/it, lr: 1.0e-04 loss: 7.732e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1262/30000 [2:52:28<48:02:57,  6.02s/it, lr: 1.0e-04 loss: 7.732e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1262/30000 [2:52:28<48:02:57,  6.02s/it, lr: 1.0e-04 loss: 7.732e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1262/30000 [2:52:35<48:02:57,  6.02s/it, lr: 1.0e-04 loss: 7.715e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1262/30000 [2:52:35<48:02:57,  6.02s/it, lr: 1.0e-04 loss: 7.715e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1263/30000 [2:52:35<49:22:00,  6.18s/it, lr: 1.0e-04 loss: 7.715e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1263/30000 [2:52:35<49:22:00,  6.18s/it, lr: 1.0e-04 loss: 7.715e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1263/30000 [2:52:42<49:22:00,  6.18s/it, lr: 1.0e-04 loss: 7.014e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1263/30000 [2:52:42<49:22:00,  6.18s/it, lr: 1.0e-04 loss: 7.014e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1264/30000 [2:52:42<50:15:44,  6.30s/it, lr: 1.0e-04 loss: 7.014e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1264/30000 [2:52:42<50:15:44,  6.30s/it, lr: 1.0e-04 loss: 7.014e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1264/30000 [2:52:48<50:15:44,  6.30s/it, lr: 1.0e-04 loss: 2.217e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1264/30000 [2:52:48<50:15:44,  6.30s/it, lr: 1.0e-04 loss: 2.217e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1265/30000 [2:52:48<51:21:37,  6.43s/it, lr: 1.0e-04 loss: 2.217e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1265/30000 [2:52:48<51:21:37,  6.43s/it, lr: 1.0e-04 loss: 2.217e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1265/30000 [2:52:55<51:21:37,  6.43s/it, lr: 1.0e-04 loss: 9.556e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1265/30000 [2:52:55<51:21:37,  6.43s/it, lr: 1.0e-04 loss: 9.556e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1266/30000 [2:52:55<51:38:59,  6.47s/it, lr: 1.0e-04 loss: 9.556e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1266/30000 [2:52:55<51:38:59,  6.47s/it, lr: 1.0e-04 loss: 9.556e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1266/30000 [2:53:01<51:38:59,  6.47s/it, lr: 1.0e-04 loss: 6.283e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1266/30000 [2:53:01<51:38:59,  6.47s/it, lr: 1.0e-04 loss: 6.283e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1267/30000 [2:53:01<51:48:58,  6.49s/it, lr: 1.0e-04 loss: 6.283e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1267/30000 [2:53:01<51:48:58,  6.49s/it, lr: 1.0e-04 loss: 6.283e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1267/30000 [2:53:11<51:48:58,  6.49s/it, lr: 1.0e-04 loss: 9.907e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1267/30000 [2:53:11<51:48:58,  6.49s/it, lr: 1.0e-04 loss: 9.907e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1268/30000 [2:53:11<58:19:32,  7.31s/it, lr: 1.0e-04 loss: 9.907e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1268/30000 [2:53:11<58:19:32,  7.31s/it, lr: 1.0e-04 loss: 9.907e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1268/30000 [2:53:17<58:19:32,  7.31s/it, lr: 1.0e-04 loss: 7.734e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1268/30000 [2:53:17<58:19:32,  7.31s/it, lr: 1.0e-04 loss: 7.734e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1269/30000 [2:53:17<56:42:14,  7.11s/it, lr: 1.0e-04 loss: 7.734e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1269/30000 [2:53:17<56:42:14,  7.11s/it, lr: 1.0e-04 loss: 7.734e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1269/30000 [2:53:27<56:42:14,  7.11s/it, lr: 1.0e-04 loss: 7.204e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1269/30000 [2:53:27<56:42:14,  7.11s/it, lr: 1.0e-04 loss: 7.204e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1270/30000 [2:53:33<56:42:07,  7.11s/it, lr: 1.0e-04 loss: 1.162e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1270/30000 [2:53:33<56:42:07,  7.11s/it, lr: 1.0e-04 loss: 1.162e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1271/30000 [2:53:33<42:46:33,  5.36s/it, lr: 1.0e-04 loss: 1.162e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1271/30000 [2:53:33<42:46:33,  5.36s/it, lr: 1.0e-04 loss: 1.162e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1271/30000 [2:53:40<42:46:33,  5.36s/it, lr: 1.0e-04 loss: 1.650e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1271/30000 [2:53:40<42:46:33,  5.36s/it, lr: 1.0e-04 loss: 1.650e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1272/30000 [2:53:40<45:36:42,  5.72s/it, lr: 1.0e-04 loss: 1.650e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1272/30000 [2:53:40<45:36:42,  5.72s/it, lr: 1.0e-04 loss: 1.650e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1272/30000 [2:53:47<45:36:42,  5.72s/it, lr: 1.0e-04 loss: 5.403e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1272/30000 [2:53:47<45:36:42,  5.72s/it, lr: 1.0e-04 loss: 5.403e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1273/30000 [2:53:47<47:26:02,  5.94s/it, lr: 1.0e-04 loss: 5.403e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1273/30000 [2:53:47<47:26:02,  5.94s/it, lr: 1.0e-04 loss: 5.403e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1273/30000 [2:53:53<47:26:02,  5.94s/it, lr: 1.0e-04 loss: 8.739e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1273/30000 [2:53:53<47:26:02,  5.94s/it, lr: 1.0e-04 loss: 8.739e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1274/30000 [2:53:53<48:45:44,  6.11s/it, lr: 1.0e-04 loss: 8.739e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1274/30000 [2:53:53<48:45:44,  6.11s/it, lr: 1.0e-04 loss: 8.739e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1274/30000 [2:54:00<48:45:44,  6.11s/it, lr: 1.0e-04 loss: 4.849e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1274/30000 [2:54:00<48:45:44,  6.11s/it, lr: 1.0e-04 loss: 4.849e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1275/30000 [2:54:00<49:46:27,  6.24s/it, lr: 1.0e-04 loss: 4.849e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1275/30000 [2:54:00<49:46:27,  6.24s/it, lr: 1.0e-04 loss: 4.849e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1275/30000 [2:54:09<49:46:27,  6.24s/it, lr: 1.0e-04 loss: 7.166e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1275/30000 [2:54:09<49:46:27,  6.24s/it, lr: 1.0e-04 loss: 7.166e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1276/30000 [2:54:09<56:43:50,  7.11s/it, lr: 1.0e-04 loss: 7.166e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1276/30000 [2:54:09<56:43:50,  7.11s/it, lr: 1.0e-04 loss: 7.166e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1276/30000 [2:54:16<56:43:50,  7.11s/it, lr: 1.0e-04 loss: 4.859e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1276/30000 [2:54:16<56:43:50,  7.11s/it, lr: 1.0e-04 loss: 4.859e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1277/30000 [2:54:16<55:33:29,  6.96s/it, lr: 1.0e-04 loss: 4.859e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1277/30000 [2:54:16<55:33:29,  6.96s/it, lr: 1.0e-04 loss: 4.859e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1277/30000 [2:54:25<55:33:29,  6.96s/it, lr: 1.0e-04 loss: 8.715e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1277/30000 [2:54:25<55:33:29,  6.96s/it, lr: 1.0e-04 loss: 8.715e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1278/30000 [2:54:25<61:02:22,  7.65s/it, lr: 1.0e-04 loss: 8.715e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1278/30000 [2:54:25<61:02:22,  7.65s/it, lr: 1.0e-04 loss: 8.715e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1278/30000 [2:54:31<61:02:22,  7.65s/it, lr: 1.0e-04 loss: 4.115e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1278/30000 [2:54:31<61:02:22,  7.65s/it, lr: 1.0e-04 loss: 4.115e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1279/30000 [2:54:31<58:32:02,  7.34s/it, lr: 1.0e-04 loss: 4.115e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1279/30000 [2:54:31<58:32:02,  7.34s/it, lr: 1.0e-04 loss: 4.115e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1279/30000 [2:54:41<58:32:02,  7.34s/it, lr: 1.0e-04 loss: 5.588e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1279/30000 [2:54:41<58:32:02,  7.34s/it, lr: 1.0e-04 loss: 5.588e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1280/30000 [2:54:50<58:31:54,  7.34s/it, lr: 1.0e-04 loss: 8.796e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1280/30000 [2:54:50<58:31:54,  7.34s/it, lr: 1.0e-04 loss: 8.796e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1281/30000 [2:54:50<48:48:41,  6.12s/it, lr: 1.0e-04 loss: 8.796e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1281/30000 [2:54:50<48:48:41,  6.12s/it, lr: 1.0e-04 loss: 8.796e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1281/30000 [2:55:00<48:48:41,  6.12s/it, lr: 1.0e-04 loss: 8.430e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1281/30000 [2:55:00<48:48:41,  6.12s/it, lr: 1.0e-04 loss: 8.430e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1282/30000 [2:55:00<55:07:56,  6.91s/it, lr: 1.0e-04 loss: 8.430e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1282/30000 [2:55:00<55:07:56,  6.91s/it, lr: 1.0e-04 loss: 8.430e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1282/30000 [2:55:06<55:07:56,  6.91s/it, lr: 1.0e-04 loss: 7.256e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1282/30000 [2:55:06<55:07:56,  6.91s/it, lr: 1.0e-04 loss: 7.256e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1283/30000 [2:55:06<54:33:48,  6.84s/it, lr: 1.0e-04 loss: 7.256e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1283/30000 [2:55:06<54:33:48,  6.84s/it, lr: 1.0e-04 loss: 7.256e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1283/30000 [2:55:13<54:33:48,  6.84s/it, lr: 1.0e-04 loss: 9.489e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1283/30000 [2:55:13<54:33:48,  6.84s/it, lr: 1.0e-04 loss: 9.489e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1284/30000 [2:55:13<53:59:16,  6.77s/it, lr: 1.0e-04 loss: 9.489e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1284/30000 [2:55:13<53:59:16,  6.77s/it, lr: 1.0e-04 loss: 9.489e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1284/30000 [2:55:22<53:59:16,  6.77s/it, lr: 1.0e-04 loss: 2.537e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1284/30000 [2:55:22<53:59:16,  6.77s/it, lr: 1.0e-04 loss: 2.537e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1285/30000 [2:55:22<59:46:12,  7.49s/it, lr: 1.0e-04 loss: 2.537e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1285/30000 [2:55:22<59:46:12,  7.49s/it, lr: 1.0e-04 loss: 2.537e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1285/30000 [2:55:29<59:46:12,  7.49s/it, lr: 1.0e-04 loss: 1.048e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1285/30000 [2:55:29<59:46:12,  7.49s/it, lr: 1.0e-04 loss: 1.048e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1286/30000 [2:55:29<57:55:00,  7.26s/it, lr: 1.0e-04 loss: 1.048e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1286/30000 [2:55:29<57:55:00,  7.26s/it, lr: 1.0e-04 loss: 1.048e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1286/30000 [2:55:36<57:55:00,  7.26s/it, lr: 1.0e-04 loss: 6.660e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1286/30000 [2:55:36<57:55:00,  7.26s/it, lr: 1.0e-04 loss: 6.660e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1287/30000 [2:55:36<57:03:34,  7.15s/it, lr: 1.0e-04 loss: 6.660e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1287/30000 [2:55:36<57:03:34,  7.15s/it, lr: 1.0e-04 loss: 6.660e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1287/30000 [2:55:42<57:03:34,  7.15s/it, lr: 1.0e-04 loss: 9.093e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1287/30000 [2:55:42<57:03:34,  7.15s/it, lr: 1.0e-04 loss: 9.093e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1288/30000 [2:55:42<55:43:23,  6.99s/it, lr: 1.0e-04 loss: 9.093e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1288/30000 [2:55:42<55:43:23,  6.99s/it, lr: 1.0e-04 loss: 9.093e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1288/30000 [2:55:49<55:43:23,  6.99s/it, lr: 1.0e-04 loss: 9.920e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1288/30000 [2:55:49<55:43:23,  6.99s/it, lr: 1.0e-04 loss: 9.920e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1289/30000 [2:55:49<54:58:05,  6.89s/it, lr: 1.0e-04 loss: 9.920e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1289/30000 [2:55:49<54:58:05,  6.89s/it, lr: 1.0e-04 loss: 9.920e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1289/30000 [2:55:56<54:58:05,  6.89s/it, lr: 1.0e-04 loss: 7.181e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1289/30000 [2:55:56<54:58:05,  6.89s/it, lr: 1.0e-04 loss: 7.181e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1290/30000 [2:56:02<54:57:58,  6.89s/it, lr: 1.0e-04 loss: 4.270e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1290/30000 [2:56:02<54:57:58,  6.89s/it, lr: 1.0e-04 loss: 4.270e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1291/30000 [2:56:02<41:48:02,  5.24s/it, lr: 1.0e-04 loss: 4.270e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1291/30000 [2:56:02<41:48:02,  5.24s/it, lr: 1.0e-04 loss: 4.270e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1291/30000 [2:56:09<41:48:02,  5.24s/it, lr: 1.0e-04 loss: 1.021e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1291/30000 [2:56:09<41:48:02,  5.24s/it, lr: 1.0e-04 loss: 1.021e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1292/30000 [2:56:09<44:33:53,  5.59s/it, lr: 1.0e-04 loss: 1.021e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1292/30000 [2:56:09<44:33:53,  5.59s/it, lr: 1.0e-04 loss: 1.021e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1292/30000 [2:56:15<44:33:53,  5.59s/it, lr: 1.0e-04 loss: 3.401e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1292/30000 [2:56:15<44:33:53,  5.59s/it, lr: 1.0e-04 loss: 3.401e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1293/30000 [2:56:15<46:45:27,  5.86s/it, lr: 1.0e-04 loss: 3.401e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1293/30000 [2:56:15<46:45:27,  5.86s/it, lr: 1.0e-04 loss: 3.401e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1293/30000 [2:56:22<46:45:27,  5.86s/it, lr: 1.0e-04 loss: 1.391e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1293/30000 [2:56:22<46:45:27,  5.86s/it, lr: 1.0e-04 loss: 1.391e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1294/30000 [2:56:22<48:23:53,  6.07s/it, lr: 1.0e-04 loss: 1.391e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1294/30000 [2:56:22<48:23:53,  6.07s/it, lr: 1.0e-04 loss: 1.391e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1294/30000 [2:56:29<48:23:53,  6.07s/it, lr: 1.0e-04 loss: 8.636e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1294/30000 [2:56:29<48:23:53,  6.07s/it, lr: 1.0e-04 loss: 8.636e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1295/30000 [2:56:29<50:21:13,  6.32s/it, lr: 1.0e-04 loss: 8.636e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1295/30000 [2:56:29<50:21:13,  6.32s/it, lr: 1.0e-04 loss: 8.636e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1295/30000 [2:56:38<50:21:13,  6.32s/it, lr: 1.0e-04 loss: 6.320e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1295/30000 [2:56:38<50:21:13,  6.32s/it, lr: 1.0e-04 loss: 6.320e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1296/30000 [2:56:38<57:18:10,  7.19s/it, lr: 1.0e-04 loss: 6.320e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1296/30000 [2:56:38<57:18:10,  7.19s/it, lr: 1.0e-04 loss: 6.320e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1296/30000 [2:56:45<57:18:10,  7.19s/it, lr: 1.0e-04 loss: 6.097e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1296/30000 [2:56:45<57:18:10,  7.19s/it, lr: 1.0e-04 loss: 6.097e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1297/30000 [2:56:45<56:15:44,  7.06s/it, lr: 1.0e-04 loss: 6.097e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1297/30000 [2:56:45<56:15:44,  7.06s/it, lr: 1.0e-04 loss: 6.097e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1297/30000 [2:56:52<56:15:44,  7.06s/it, lr: 1.0e-04 loss: 1.012e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1297/30000 [2:56:52<56:15:44,  7.06s/it, lr: 1.0e-04 loss: 1.012e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1298/30000 [2:56:52<55:11:56,  6.92s/it, lr: 1.0e-04 loss: 1.012e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1298/30000 [2:56:52<55:11:56,  6.92s/it, lr: 1.0e-04 loss: 1.012e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1298/30000 [2:57:01<55:11:56,  6.92s/it, lr: 1.0e-04 loss: 1.136e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1298/30000 [2:57:01<55:11:56,  6.92s/it, lr: 1.0e-04 loss: 1.136e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1299/30000 [2:57:01<60:56:22,  7.64s/it, lr: 1.0e-04 loss: 1.136e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1299/30000 [2:57:01<60:56:22,  7.64s/it, lr: 1.0e-04 loss: 1.136e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1299/30000 [2:57:08<60:56:22,  7.64s/it, lr: 1.0e-04 loss: 1.203e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1299/30000 [2:57:08<60:56:22,  7.64s/it, lr: 1.0e-04 loss: 1.203e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1300/30000 [2:57:14<60:56:14,  7.64s/it, lr: 1.0e-04 loss: 8.879e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1300/30000 [2:57:14<60:56:14,  7.64s/it, lr: 1.0e-04 loss: 8.879e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1301/30000 [2:57:14<45:05:24,  5.66s/it, lr: 1.0e-04 loss: 8.879e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1301/30000 [2:57:14<45:05:24,  5.66s/it, lr: 1.0e-04 loss: 8.879e-01]
Xiang_wan_2_2_lora:   4%|#6                                     | 1301/30000 [2:57:24<45:05:24,  5.66s/it, lr: 1.0e-04 loss: 6.805e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1301/30000 [2:57:24<45:05:24,  5.66s/it, lr: 1.0e-04 loss: 6.805e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1302/30000 [2:57:24<52:45:35,  6.62s/it, lr: 1.0e-04 loss: 6.805e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1302/30000 [2:57:24<52:45:35,  6.62s/it, lr: 1.0e-04 loss: 6.805e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1302/30000 [2:57:33<52:45:35,  6.62s/it, lr: 1.0e-04 loss: 6.791e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1302/30000 [2:57:33<52:45:35,  6.62s/it, lr: 1.0e-04 loss: 6.791e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1303/30000 [2:57:33<58:38:48,  7.36s/it, lr: 1.0e-04 loss: 6.791e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1303/30000 [2:57:33<58:38:48,  7.36s/it, lr: 1.0e-04 loss: 6.791e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1303/30000 [2:57:43<58:38:48,  7.36s/it, lr: 1.0e-04 loss: 7.410e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1303/30000 [2:57:43<58:38:48,  7.36s/it, lr: 1.0e-04 loss: 7.410e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1304/30000 [2:57:43<62:49:08,  7.88s/it, lr: 1.0e-04 loss: 7.410e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1304/30000 [2:57:43<62:49:08,  7.88s/it, lr: 1.0e-04 loss: 7.410e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1304/30000 [2:57:52<62:49:08,  7.88s/it, lr: 1.0e-04 loss: 5.400e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1304/30000 [2:57:52<62:49:08,  7.88s/it, lr: 1.0e-04 loss: 5.400e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1305/30000 [2:57:52<66:17:20,  8.32s/it, lr: 1.0e-04 loss: 5.400e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1305/30000 [2:57:52<66:17:20,  8.32s/it, lr: 1.0e-04 loss: 5.400e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1305/30000 [2:57:59<66:17:20,  8.32s/it, lr: 1.0e-04 loss: 9.133e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1305/30000 [2:57:59<66:17:20,  8.32s/it, lr: 1.0e-04 loss: 9.133e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1306/30000 [2:57:59<62:21:23,  7.82s/it, lr: 1.0e-04 loss: 9.133e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1306/30000 [2:57:59<62:21:23,  7.82s/it, lr: 1.0e-04 loss: 9.133e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1306/30000 [2:58:08<62:21:23,  7.82s/it, lr: 1.0e-04 loss: 3.740e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1306/30000 [2:58:08<62:21:23,  7.82s/it, lr: 1.0e-04 loss: 3.740e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1307/30000 [2:58:08<65:47:31,  8.25s/it, lr: 1.0e-04 loss: 3.740e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1307/30000 [2:58:08<65:47:31,  8.25s/it, lr: 1.0e-04 loss: 3.740e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1307/30000 [2:58:17<65:47:31,  8.25s/it, lr: 1.0e-04 loss: 5.778e-02]
Xiang_wan_2_2_lora:   4%|#6                                     | 1307/30000 [2:58:17<65:47:31,  8.25s/it, lr: 1.0e-04 loss: 5.778e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1308/30000 [2:58:17<68:26:15,  8.59s/it, lr: 1.0e-04 loss: 5.778e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1308/30000 [2:58:17<68:26:15,  8.59s/it, lr: 1.0e-04 loss: 5.778e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1308/30000 [2:58:27<68:26:15,  8.59s/it, lr: 1.0e-04 loss: 2.571e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1308/30000 [2:58:27<68:26:15,  8.59s/it, lr: 1.0e-04 loss: 2.571e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1309/30000 [2:58:27<70:43:14,  8.87s/it, lr: 1.0e-04 loss: 2.571e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1309/30000 [2:58:27<70:43:14,  8.87s/it, lr: 1.0e-04 loss: 2.571e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1309/30000 [2:58:36<70:43:14,  8.87s/it, lr: 1.0e-04 loss: 8.464e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1309/30000 [2:58:36<70:43:14,  8.87s/it, lr: 1.0e-04 loss: 8.464e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1310/30000 [2:58:46<70:43:05,  8.87s/it, lr: 1.0e-04 loss: 3.858e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1310/30000 [2:58:46<70:43:05,  8.87s/it, lr: 1.0e-04 loss: 3.858e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1311/30000 [2:58:46<55:30:29,  6.97s/it, lr: 1.0e-04 loss: 3.858e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1311/30000 [2:58:46<55:30:29,  6.97s/it, lr: 1.0e-04 loss: 3.858e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1311/30000 [2:58:52<55:30:29,  6.97s/it, lr: 1.0e-04 loss: 9.614e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1311/30000 [2:58:52<55:30:29,  6.97s/it, lr: 1.0e-04 loss: 9.614e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1312/30000 [2:58:52<54:53:42,  6.89s/it, lr: 1.0e-04 loss: 9.614e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1312/30000 [2:58:52<54:53:42,  6.89s/it, lr: 1.0e-04 loss: 9.614e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1312/30000 [2:58:59<54:53:42,  6.89s/it, lr: 1.0e-04 loss: 7.665e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1312/30000 [2:58:59<54:53:42,  6.89s/it, lr: 1.0e-04 loss: 7.665e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1313/30000 [2:58:59<54:09:11,  6.80s/it, lr: 1.0e-04 loss: 7.665e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1313/30000 [2:58:59<54:09:11,  6.80s/it, lr: 1.0e-04 loss: 7.665e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1313/30000 [2:59:05<54:09:11,  6.80s/it, lr: 1.0e-04 loss: 9.189e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1313/30000 [2:59:05<54:09:11,  6.80s/it, lr: 1.0e-04 loss: 9.189e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1314/30000 [2:59:05<53:35:23,  6.73s/it, lr: 1.0e-04 loss: 9.189e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1314/30000 [2:59:05<53:35:23,  6.73s/it, lr: 1.0e-04 loss: 9.189e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1314/30000 [2:59:12<53:35:23,  6.73s/it, lr: 1.0e-04 loss: 6.702e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1314/30000 [2:59:12<53:35:23,  6.73s/it, lr: 1.0e-04 loss: 6.702e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1315/30000 [2:59:12<53:14:27,  6.68s/it, lr: 1.0e-04 loss: 6.702e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1315/30000 [2:59:12<53:14:27,  6.68s/it, lr: 1.0e-04 loss: 6.702e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1315/30000 [2:59:21<53:14:27,  6.68s/it, lr: 1.0e-04 loss: 8.621e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1315/30000 [2:59:21<53:14:27,  6.68s/it, lr: 1.0e-04 loss: 8.621e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1316/30000 [2:59:21<59:12:02,  7.43s/it, lr: 1.0e-04 loss: 8.621e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1316/30000 [2:59:21<59:12:02,  7.43s/it, lr: 1.0e-04 loss: 8.621e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1316/30000 [2:59:28<59:12:02,  7.43s/it, lr: 1.0e-04 loss: 8.501e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1316/30000 [2:59:28<59:12:02,  7.43s/it, lr: 1.0e-04 loss: 8.501e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1317/30000 [2:59:28<57:52:39,  7.26s/it, lr: 1.0e-04 loss: 8.501e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1317/30000 [2:59:28<57:52:39,  7.26s/it, lr: 1.0e-04 loss: 8.501e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1317/30000 [2:59:35<57:52:39,  7.26s/it, lr: 1.0e-04 loss: 1.147e-01]
Xiang_wan_2_2_lora:   4%|#7                                     | 1317/30000 [2:59:35<57:52:39,  7.26s/it, lr: 1.0e-04 loss: 1.147e-01]
Xiang_wan_2_2_lora:   4%|#7                                     | 1318/30000 [2:59:35<56:16:28,  7.06s/it, lr: 1.0e-04 loss: 1.147e-01]
Xiang_wan_2_2_lora:   4%|#7                                     | 1318/30000 [2:59:35<56:16:28,  7.06s/it, lr: 1.0e-04 loss: 1.147e-01]
Xiang_wan_2_2_lora:   4%|#7                                     | 1318/30000 [2:59:44<56:16:28,  7.06s/it, lr: 1.0e-04 loss: 5.596e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1318/30000 [2:59:44<56:16:28,  7.06s/it, lr: 1.0e-04 loss: 5.596e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1319/30000 [2:59:44<61:27:54,  7.72s/it, lr: 1.0e-04 loss: 5.596e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1319/30000 [2:59:44<61:27:54,  7.72s/it, lr: 1.0e-04 loss: 5.596e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1319/30000 [2:59:51<61:27:54,  7.72s/it, lr: 1.0e-04 loss: 8.543e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1319/30000 [2:59:51<61:27:54,  7.72s/it, lr: 1.0e-04 loss: 8.543e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1320/30000 [3:00:00<61:27:46,  7.72s/it, lr: 1.0e-04 loss: 4.762e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1320/30000 [3:00:00<61:27:46,  7.72s/it, lr: 1.0e-04 loss: 4.762e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1321/30000 [3:00:00<50:19:00,  6.32s/it, lr: 1.0e-04 loss: 4.762e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1321/30000 [3:00:00<50:19:00,  6.32s/it, lr: 1.0e-04 loss: 4.762e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1321/30000 [3:00:07<50:19:00,  6.32s/it, lr: 1.0e-04 loss: 1.184e-01]
Xiang_wan_2_2_lora:   4%|#7                                     | 1321/30000 [3:00:07<50:19:00,  6.32s/it, lr: 1.0e-04 loss: 1.184e-01]
Xiang_wan_2_2_lora:   4%|#7                                     | 1322/30000 [3:00:07<51:00:45,  6.40s/it, lr: 1.0e-04 loss: 1.184e-01]
Xiang_wan_2_2_lora:   4%|#7                                     | 1322/30000 [3:00:07<51:00:45,  6.40s/it, lr: 1.0e-04 loss: 1.184e-01]
Xiang_wan_2_2_lora:   4%|#7                                     | 1322/30000 [3:00:13<51:00:45,  6.40s/it, lr: 1.0e-04 loss: 6.166e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1322/30000 [3:00:13<51:00:45,  6.40s/it, lr: 1.0e-04 loss: 6.166e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1323/30000 [3:00:13<51:14:05,  6.43s/it, lr: 1.0e-04 loss: 6.166e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1323/30000 [3:00:13<51:14:05,  6.43s/it, lr: 1.0e-04 loss: 6.166e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1323/30000 [3:00:20<51:14:05,  6.43s/it, lr: 1.0e-04 loss: 9.893e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1323/30000 [3:00:20<51:14:05,  6.43s/it, lr: 1.0e-04 loss: 9.893e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1324/30000 [3:00:20<52:05:54,  6.54s/it, lr: 1.0e-04 loss: 9.893e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1324/30000 [3:00:20<52:05:54,  6.54s/it, lr: 1.0e-04 loss: 9.893e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1324/30000 [3:00:27<52:05:54,  6.54s/it, lr: 1.0e-04 loss: 4.621e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1324/30000 [3:00:27<52:05:54,  6.54s/it, lr: 1.0e-04 loss: 4.621e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1325/30000 [3:00:27<52:18:29,  6.57s/it, lr: 1.0e-04 loss: 4.621e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1325/30000 [3:00:27<52:18:29,  6.57s/it, lr: 1.0e-04 loss: 4.621e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1325/30000 [3:00:33<52:18:29,  6.57s/it, lr: 1.0e-04 loss: 1.099e-01]
Xiang_wan_2_2_lora:   4%|#7                                     | 1325/30000 [3:00:33<52:18:29,  6.57s/it, lr: 1.0e-04 loss: 1.099e-01]
Xiang_wan_2_2_lora:   4%|#7                                     | 1326/30000 [3:00:33<52:20:53,  6.57s/it, lr: 1.0e-04 loss: 1.099e-01]
Xiang_wan_2_2_lora:   4%|#7                                     | 1326/30000 [3:00:33<52:20:53,  6.57s/it, lr: 1.0e-04 loss: 1.099e-01]
Xiang_wan_2_2_lora:   4%|#7                                     | 1326/30000 [3:00:40<52:20:53,  6.57s/it, lr: 1.0e-04 loss: 5.146e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1326/30000 [3:00:40<52:20:53,  6.57s/it, lr: 1.0e-04 loss: 5.146e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1327/30000 [3:00:40<52:29:23,  6.59s/it, lr: 1.0e-04 loss: 5.146e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1327/30000 [3:00:40<52:29:23,  6.59s/it, lr: 1.0e-04 loss: 5.146e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1327/30000 [3:00:46<52:29:23,  6.59s/it, lr: 1.0e-04 loss: 7.456e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1327/30000 [3:00:46<52:29:23,  6.59s/it, lr: 1.0e-04 loss: 7.456e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1328/30000 [3:00:46<52:30:27,  6.59s/it, lr: 1.0e-04 loss: 7.456e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1328/30000 [3:00:46<52:30:27,  6.59s/it, lr: 1.0e-04 loss: 7.456e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1328/30000 [3:00:56<52:30:27,  6.59s/it, lr: 1.0e-04 loss: 4.405e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1328/30000 [3:00:56<52:30:27,  6.59s/it, lr: 1.0e-04 loss: 4.405e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1329/30000 [3:00:56<58:56:17,  7.40s/it, lr: 1.0e-04 loss: 4.405e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1329/30000 [3:00:56<58:56:17,  7.40s/it, lr: 1.0e-04 loss: 4.405e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1329/30000 [3:01:02<58:56:17,  7.40s/it, lr: 1.0e-04 loss: 8.727e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1329/30000 [3:01:02<58:56:17,  7.40s/it, lr: 1.0e-04 loss: 8.727e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1330/30000 [3:01:09<58:56:09,  7.40s/it, lr: 1.0e-04 loss: 1.465e-01]
Xiang_wan_2_2_lora:   4%|#7                                     | 1330/30000 [3:01:09<58:56:09,  7.40s/it, lr: 1.0e-04 loss: 1.465e-01]
Xiang_wan_2_2_lora:   4%|#7                                     | 1331/30000 [3:01:09<44:28:39,  5.59s/it, lr: 1.0e-04 loss: 1.465e-01]
Xiang_wan_2_2_lora:   4%|#7                                     | 1331/30000 [3:01:09<44:28:39,  5.59s/it, lr: 1.0e-04 loss: 1.465e-01]
Xiang_wan_2_2_lora:   4%|#7                                     | 1331/30000 [3:01:18<44:28:39,  5.59s/it, lr: 1.0e-04 loss: 7.416e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1331/30000 [3:01:18<44:28:39,  5.59s/it, lr: 1.0e-04 loss: 7.416e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1332/30000 [3:01:18<51:47:02,  6.50s/it, lr: 1.0e-04 loss: 7.416e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1332/30000 [3:01:18<51:47:02,  6.50s/it, lr: 1.0e-04 loss: 7.416e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1332/30000 [3:01:25<51:47:02,  6.50s/it, lr: 1.0e-04 loss: 7.321e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1332/30000 [3:01:25<51:47:02,  6.50s/it, lr: 1.0e-04 loss: 7.321e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1333/30000 [3:01:25<52:12:40,  6.56s/it, lr: 1.0e-04 loss: 7.321e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1333/30000 [3:01:25<52:12:40,  6.56s/it, lr: 1.0e-04 loss: 7.321e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1333/30000 [3:01:35<52:12:40,  6.56s/it, lr: 1.0e-04 loss: 9.519e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1333/30000 [3:01:35<52:12:40,  6.56s/it, lr: 1.0e-04 loss: 9.519e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1334/30000 [3:01:35<58:15:32,  7.32s/it, lr: 1.0e-04 loss: 9.519e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1334/30000 [3:01:35<58:15:32,  7.32s/it, lr: 1.0e-04 loss: 9.519e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1334/30000 [3:01:41<58:15:32,  7.32s/it, lr: 1.0e-04 loss: 8.465e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1334/30000 [3:01:41<58:15:32,  7.32s/it, lr: 1.0e-04 loss: 8.465e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1335/30000 [3:01:41<56:50:10,  7.14s/it, lr: 1.0e-04 loss: 8.465e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1335/30000 [3:01:41<56:50:10,  7.14s/it, lr: 1.0e-04 loss: 8.465e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1335/30000 [3:01:48<56:50:10,  7.14s/it, lr: 1.0e-04 loss: 8.797e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1335/30000 [3:01:48<56:50:10,  7.14s/it, lr: 1.0e-04 loss: 8.797e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1336/30000 [3:01:48<55:45:39,  7.00s/it, lr: 1.0e-04 loss: 8.797e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1336/30000 [3:01:48<55:45:39,  7.00s/it, lr: 1.0e-04 loss: 8.797e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1336/30000 [3:01:57<55:45:39,  7.00s/it, lr: 1.0e-04 loss: 5.471e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1336/30000 [3:01:57<55:45:39,  7.00s/it, lr: 1.0e-04 loss: 5.471e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1337/30000 [3:01:57<61:24:27,  7.71s/it, lr: 1.0e-04 loss: 5.471e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1337/30000 [3:01:57<61:24:27,  7.71s/it, lr: 1.0e-04 loss: 5.471e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1337/30000 [3:02:07<61:24:27,  7.71s/it, lr: 1.0e-04 loss: 4.996e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1337/30000 [3:02:07<61:24:27,  7.71s/it, lr: 1.0e-04 loss: 4.996e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1338/30000 [3:02:07<65:19:49,  8.21s/it, lr: 1.0e-04 loss: 4.996e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1338/30000 [3:02:07<65:19:49,  8.21s/it, lr: 1.0e-04 loss: 4.996e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1338/30000 [3:02:14<65:19:49,  8.21s/it, lr: 1.0e-04 loss: 5.860e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1338/30000 [3:02:14<65:19:49,  8.21s/it, lr: 1.0e-04 loss: 5.860e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1339/30000 [3:02:14<62:21:31,  7.83s/it, lr: 1.0e-04 loss: 5.860e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1339/30000 [3:02:14<62:21:31,  7.83s/it, lr: 1.0e-04 loss: 5.860e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1339/30000 [3:02:20<62:21:31,  7.83s/it, lr: 1.0e-04 loss: 5.685e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1339/30000 [3:02:20<62:21:31,  7.83s/it, lr: 1.0e-04 loss: 5.685e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1340/30000 [3:02:27<62:21:23,  7.83s/it, lr: 1.0e-04 loss: 8.652e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1340/30000 [3:02:27<62:21:23,  7.83s/it, lr: 1.0e-04 loss: 8.652e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1341/30000 [3:02:27<45:48:28,  5.75s/it, lr: 1.0e-04 loss: 8.652e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1341/30000 [3:02:27<45:48:28,  5.75s/it, lr: 1.0e-04 loss: 8.652e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1341/30000 [3:02:33<45:48:28,  5.75s/it, lr: 1.0e-04 loss: 8.256e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1341/30000 [3:02:33<45:48:28,  5.75s/it, lr: 1.0e-04 loss: 8.256e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1342/30000 [3:02:33<47:33:18,  5.97s/it, lr: 1.0e-04 loss: 8.256e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1342/30000 [3:02:33<47:33:18,  5.97s/it, lr: 1.0e-04 loss: 8.256e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1342/30000 [3:02:43<47:33:18,  5.97s/it, lr: 1.0e-04 loss: 4.934e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1342/30000 [3:02:43<47:33:18,  5.97s/it, lr: 1.0e-04 loss: 4.934e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1343/30000 [3:02:43<54:36:54,  6.86s/it, lr: 1.0e-04 loss: 4.934e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1343/30000 [3:02:43<54:36:54,  6.86s/it, lr: 1.0e-04 loss: 4.934e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1343/30000 [3:02:52<54:36:54,  6.86s/it, lr: 1.0e-04 loss: 7.894e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1343/30000 [3:02:52<54:36:54,  6.86s/it, lr: 1.0e-04 loss: 7.894e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1344/30000 [3:02:52<60:08:40,  7.56s/it, lr: 1.0e-04 loss: 7.894e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1344/30000 [3:02:52<60:08:40,  7.56s/it, lr: 1.0e-04 loss: 7.894e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1344/30000 [3:03:02<60:08:40,  7.56s/it, lr: 1.0e-04 loss: 4.996e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1344/30000 [3:03:02<60:08:40,  7.56s/it, lr: 1.0e-04 loss: 4.996e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1345/30000 [3:03:02<64:09:20,  8.06s/it, lr: 1.0e-04 loss: 4.996e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1345/30000 [3:03:02<64:09:20,  8.06s/it, lr: 1.0e-04 loss: 4.996e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1345/30000 [3:03:09<64:09:20,  8.06s/it, lr: 1.0e-04 loss: 8.889e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1345/30000 [3:03:09<64:09:20,  8.06s/it, lr: 1.0e-04 loss: 8.889e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1346/30000 [3:03:09<61:27:00,  7.72s/it, lr: 1.0e-04 loss: 8.889e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1346/30000 [3:03:09<61:27:00,  7.72s/it, lr: 1.0e-04 loss: 8.889e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1346/30000 [3:03:18<61:27:00,  7.72s/it, lr: 1.0e-04 loss: 2.439e-01]
Xiang_wan_2_2_lora:   4%|#7                                     | 1346/30000 [3:03:18<61:27:00,  7.72s/it, lr: 1.0e-04 loss: 2.439e-01]
Xiang_wan_2_2_lora:   4%|#7                                     | 1347/30000 [3:03:18<65:14:29,  8.20s/it, lr: 1.0e-04 loss: 2.439e-01]
Xiang_wan_2_2_lora:   4%|#7                                     | 1347/30000 [3:03:18<65:14:29,  8.20s/it, lr: 1.0e-04 loss: 2.439e-01]
Xiang_wan_2_2_lora:   4%|#7                                     | 1347/30000 [3:03:27<65:14:29,  8.20s/it, lr: 1.0e-04 loss: 5.371e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1347/30000 [3:03:27<65:14:29,  8.20s/it, lr: 1.0e-04 loss: 5.371e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1348/30000 [3:03:27<68:04:51,  8.55s/it, lr: 1.0e-04 loss: 5.371e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1348/30000 [3:03:27<68:04:51,  8.55s/it, lr: 1.0e-04 loss: 5.371e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1348/30000 [3:03:37<68:04:51,  8.55s/it, lr: 1.0e-04 loss: 8.305e-01]
Xiang_wan_2_2_lora:   4%|#7                                     | 1348/30000 [3:03:37<68:04:51,  8.55s/it, lr: 1.0e-04 loss: 8.305e-01]
Xiang_wan_2_2_lora:   4%|#7                                     | 1349/30000 [3:03:37<69:56:17,  8.79s/it, lr: 1.0e-04 loss: 8.305e-01]
Xiang_wan_2_2_lora:   4%|#7                                     | 1349/30000 [3:03:37<69:56:17,  8.79s/it, lr: 1.0e-04 loss: 8.305e-01]
Xiang_wan_2_2_lora:   4%|#7                                     | 1349/30000 [3:03:46<69:56:17,  8.79s/it, lr: 1.0e-04 loss: 8.820e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1349/30000 [3:03:46<69:56:17,  8.79s/it, lr: 1.0e-04 loss: 8.820e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1350/30000 [3:03:55<69:56:08,  8.79s/it, lr: 1.0e-04 loss: 6.183e-02]
Xiang_wan_2_2_lora:   4%|#7                                     | 1350/30000 [3:03:55<69:56:08,  8.79s/it, lr: 1.0e-04 loss: 6.183e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1351/30000 [3:03:55<55:04:47,  6.92s/it, lr: 1.0e-04 loss: 6.183e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1351/30000 [3:03:55<55:04:47,  6.92s/it, lr: 1.0e-04 loss: 6.183e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1351/30000 [3:04:05<55:04:47,  6.92s/it, lr: 1.0e-04 loss: 6.628e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1351/30000 [3:04:05<55:04:47,  6.92s/it, lr: 1.0e-04 loss: 6.628e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1352/30000 [3:04:05<59:48:59,  7.52s/it, lr: 1.0e-04 loss: 6.628e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1352/30000 [3:04:05<59:48:59,  7.52s/it, lr: 1.0e-04 loss: 6.628e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1352/30000 [3:04:14<59:48:59,  7.52s/it, lr: 1.0e-04 loss: 5.761e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1352/30000 [3:04:14<59:48:59,  7.52s/it, lr: 1.0e-04 loss: 5.761e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1353/30000 [3:04:14<63:58:21,  8.04s/it, lr: 1.0e-04 loss: 5.761e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1353/30000 [3:04:14<63:58:21,  8.04s/it, lr: 1.0e-04 loss: 5.761e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1353/30000 [3:04:21<63:58:21,  8.04s/it, lr: 1.0e-04 loss: 1.135e-01]
Xiang_wan_2_2_lora:   5%|#7                                     | 1353/30000 [3:04:21<63:58:21,  8.04s/it, lr: 1.0e-04 loss: 1.135e-01]
Xiang_wan_2_2_lora:   5%|#7                                     | 1354/30000 [3:04:21<60:56:58,  7.66s/it, lr: 1.0e-04 loss: 1.135e-01]
Xiang_wan_2_2_lora:   5%|#7                                     | 1354/30000 [3:04:21<60:56:58,  7.66s/it, lr: 1.0e-04 loss: 1.135e-01]
Xiang_wan_2_2_lora:   5%|#7                                     | 1354/30000 [3:04:30<60:56:58,  7.66s/it, lr: 1.0e-04 loss: 8.200e-01]
Xiang_wan_2_2_lora:   5%|#7                                     | 1354/30000 [3:04:30<60:56:58,  7.66s/it, lr: 1.0e-04 loss: 8.200e-01]
Xiang_wan_2_2_lora:   5%|#7                                     | 1355/30000 [3:04:30<64:35:26,  8.12s/it, lr: 1.0e-04 loss: 8.200e-01]
Xiang_wan_2_2_lora:   5%|#7                                     | 1355/30000 [3:04:30<64:35:26,  8.12s/it, lr: 1.0e-04 loss: 8.200e-01]
Xiang_wan_2_2_lora:   5%|#7                                     | 1355/30000 [3:04:40<64:35:26,  8.12s/it, lr: 1.0e-04 loss: 9.096e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1355/30000 [3:04:40<64:35:26,  8.12s/it, lr: 1.0e-04 loss: 9.096e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1356/30000 [3:04:40<67:24:04,  8.47s/it, lr: 1.0e-04 loss: 9.096e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1356/30000 [3:04:40<67:24:04,  8.47s/it, lr: 1.0e-04 loss: 9.096e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1356/30000 [3:04:49<67:24:04,  8.47s/it, lr: 1.0e-04 loss: 6.310e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1356/30000 [3:04:49<67:24:04,  8.47s/it, lr: 1.0e-04 loss: 6.310e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1357/30000 [3:04:49<69:23:13,  8.72s/it, lr: 1.0e-04 loss: 6.310e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1357/30000 [3:04:49<69:23:13,  8.72s/it, lr: 1.0e-04 loss: 6.310e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1357/30000 [3:04:56<69:23:13,  8.72s/it, lr: 1.0e-04 loss: 8.489e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1357/30000 [3:04:56<69:23:13,  8.72s/it, lr: 1.0e-04 loss: 8.489e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1358/30000 [3:04:56<64:29:54,  8.11s/it, lr: 1.0e-04 loss: 8.489e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1358/30000 [3:04:56<64:29:54,  8.11s/it, lr: 1.0e-04 loss: 8.489e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1358/30000 [3:05:05<64:29:54,  8.11s/it, lr: 1.0e-04 loss: 1.259e-01]
Xiang_wan_2_2_lora:   5%|#7                                     | 1358/30000 [3:05:05<64:29:54,  8.11s/it, lr: 1.0e-04 loss: 1.259e-01]
Xiang_wan_2_2_lora:   5%|#7                                     | 1359/30000 [3:05:05<67:18:01,  8.46s/it, lr: 1.0e-04 loss: 1.259e-01]
Xiang_wan_2_2_lora:   5%|#7                                     | 1359/30000 [3:05:05<67:18:01,  8.46s/it, lr: 1.0e-04 loss: 1.259e-01]
Xiang_wan_2_2_lora:   5%|#7                                     | 1359/30000 [3:05:11<67:18:01,  8.46s/it, lr: 1.0e-04 loss: 9.113e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1359/30000 [3:05:11<67:18:01,  8.46s/it, lr: 1.0e-04 loss: 9.113e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1360/30000 [3:05:18<67:17:52,  8.46s/it, lr: 1.0e-04 loss: 5.781e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1360/30000 [3:05:18<67:17:52,  8.46s/it, lr: 1.0e-04 loss: 5.781e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1361/30000 [3:05:18<48:55:58,  6.15s/it, lr: 1.0e-04 loss: 5.781e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1361/30000 [3:05:18<48:55:58,  6.15s/it, lr: 1.0e-04 loss: 5.781e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1361/30000 [3:05:28<48:55:58,  6.15s/it, lr: 1.0e-04 loss: 6.627e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1361/30000 [3:05:28<48:55:58,  6.15s/it, lr: 1.0e-04 loss: 6.627e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1362/30000 [3:05:28<55:10:27,  6.94s/it, lr: 1.0e-04 loss: 6.627e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1362/30000 [3:05:28<55:10:27,  6.94s/it, lr: 1.0e-04 loss: 6.627e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1362/30000 [3:05:34<55:10:27,  6.94s/it, lr: 1.0e-04 loss: 4.460e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1362/30000 [3:05:34<55:10:27,  6.94s/it, lr: 1.0e-04 loss: 4.460e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1363/30000 [3:05:34<54:27:03,  6.85s/it, lr: 1.0e-04 loss: 4.460e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1363/30000 [3:05:34<54:27:03,  6.85s/it, lr: 1.0e-04 loss: 4.460e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1363/30000 [3:05:44<54:27:03,  6.85s/it, lr: 1.0e-04 loss: 5.595e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1363/30000 [3:05:44<54:27:03,  6.85s/it, lr: 1.0e-04 loss: 5.595e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1364/30000 [3:05:44<59:49:17,  7.52s/it, lr: 1.0e-04 loss: 5.595e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1364/30000 [3:05:44<59:49:17,  7.52s/it, lr: 1.0e-04 loss: 5.595e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1364/30000 [3:05:53<59:49:17,  7.52s/it, lr: 1.0e-04 loss: 4.064e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1364/30000 [3:05:53<59:49:17,  7.52s/it, lr: 1.0e-04 loss: 4.064e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1365/30000 [3:05:53<63:52:56,  8.03s/it, lr: 1.0e-04 loss: 4.064e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1365/30000 [3:05:53<63:52:56,  8.03s/it, lr: 1.0e-04 loss: 4.064e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1365/30000 [3:05:59<63:52:56,  8.03s/it, lr: 1.0e-04 loss: 6.890e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1365/30000 [3:05:59<63:52:56,  8.03s/it, lr: 1.0e-04 loss: 6.890e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1366/30000 [3:05:59<60:41:25,  7.63s/it, lr: 1.0e-04 loss: 6.890e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1366/30000 [3:05:59<60:41:25,  7.63s/it, lr: 1.0e-04 loss: 6.890e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1366/30000 [3:06:09<60:41:25,  7.63s/it, lr: 1.0e-04 loss: 4.681e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1366/30000 [3:06:09<60:41:25,  7.63s/it, lr: 1.0e-04 loss: 4.681e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1367/30000 [3:06:09<64:42:28,  8.14s/it, lr: 1.0e-04 loss: 4.681e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1367/30000 [3:06:09<64:42:28,  8.14s/it, lr: 1.0e-04 loss: 4.681e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1367/30000 [3:06:16<64:42:28,  8.14s/it, lr: 1.0e-04 loss: 1.263e-01]
Xiang_wan_2_2_lora:   5%|#7                                     | 1367/30000 [3:06:16<64:42:28,  8.14s/it, lr: 1.0e-04 loss: 1.263e-01]
Xiang_wan_2_2_lora:   5%|#7                                     | 1368/30000 [3:06:16<61:41:14,  7.76s/it, lr: 1.0e-04 loss: 1.263e-01]
Xiang_wan_2_2_lora:   5%|#7                                     | 1368/30000 [3:06:16<61:41:14,  7.76s/it, lr: 1.0e-04 loss: 1.263e-01]
Xiang_wan_2_2_lora:   5%|#7                                     | 1368/30000 [3:06:22<61:41:14,  7.76s/it, lr: 1.0e-04 loss: 6.089e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1368/30000 [3:06:22<61:41:14,  7.76s/it, lr: 1.0e-04 loss: 6.089e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1369/30000 [3:06:22<59:00:11,  7.42s/it, lr: 1.0e-04 loss: 6.089e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1369/30000 [3:06:22<59:00:11,  7.42s/it, lr: 1.0e-04 loss: 6.089e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1369/30000 [3:06:29<59:00:11,  7.42s/it, lr: 1.0e-04 loss: 8.882e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1369/30000 [3:06:29<59:00:11,  7.42s/it, lr: 1.0e-04 loss: 8.882e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1370/30000 [3:06:38<59:00:04,  7.42s/it, lr: 1.0e-04 loss: 3.224e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1370/30000 [3:06:38<59:00:04,  7.42s/it, lr: 1.0e-04 loss: 3.224e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1371/30000 [3:06:38<49:09:14,  6.18s/it, lr: 1.0e-04 loss: 3.224e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1371/30000 [3:06:38<49:09:14,  6.18s/it, lr: 1.0e-04 loss: 3.224e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1371/30000 [3:06:48<49:09:14,  6.18s/it, lr: 1.0e-04 loss: 1.038e-01]
Xiang_wan_2_2_lora:   5%|#7                                     | 1371/30000 [3:06:48<49:09:14,  6.18s/it, lr: 1.0e-04 loss: 1.038e-01]
Xiang_wan_2_2_lora:   5%|#7                                     | 1372/30000 [3:06:48<55:19:09,  6.96s/it, lr: 1.0e-04 loss: 1.038e-01]
Xiang_wan_2_2_lora:   5%|#7                                     | 1372/30000 [3:06:48<55:19:09,  6.96s/it, lr: 1.0e-04 loss: 1.038e-01]
Xiang_wan_2_2_lora:   5%|#7                                     | 1372/30000 [3:06:57<55:19:09,  6.96s/it, lr: 1.0e-04 loss: 5.635e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1372/30000 [3:06:57<55:19:09,  6.96s/it, lr: 1.0e-04 loss: 5.635e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1373/30000 [3:06:57<60:10:27,  7.57s/it, lr: 1.0e-04 loss: 5.635e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1373/30000 [3:06:57<60:10:27,  7.57s/it, lr: 1.0e-04 loss: 5.635e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1373/30000 [3:07:06<60:10:27,  7.57s/it, lr: 1.0e-04 loss: 7.707e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1373/30000 [3:07:06<60:10:27,  7.57s/it, lr: 1.0e-04 loss: 7.707e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1374/30000 [3:07:06<64:04:38,  8.06s/it, lr: 1.0e-04 loss: 7.707e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1374/30000 [3:07:06<64:04:38,  8.06s/it, lr: 1.0e-04 loss: 7.707e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1374/30000 [3:07:16<64:04:38,  8.06s/it, lr: 1.0e-04 loss: 7.669e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1374/30000 [3:07:16<64:04:38,  8.06s/it, lr: 1.0e-04 loss: 7.669e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1375/30000 [3:07:16<67:18:15,  8.46s/it, lr: 1.0e-04 loss: 7.669e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1375/30000 [3:07:16<67:18:15,  8.46s/it, lr: 1.0e-04 loss: 7.669e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1375/30000 [3:07:25<67:18:15,  8.46s/it, lr: 1.0e-04 loss: 6.453e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1375/30000 [3:07:25<67:18:15,  8.46s/it, lr: 1.0e-04 loss: 6.453e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1376/30000 [3:07:25<69:24:19,  8.73s/it, lr: 1.0e-04 loss: 6.453e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1376/30000 [3:07:25<69:24:19,  8.73s/it, lr: 1.0e-04 loss: 6.453e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1376/30000 [3:07:35<69:24:19,  8.73s/it, lr: 1.0e-04 loss: 4.796e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1376/30000 [3:07:35<69:24:19,  8.73s/it, lr: 1.0e-04 loss: 4.796e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1377/30000 [3:07:35<70:53:05,  8.92s/it, lr: 1.0e-04 loss: 4.796e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1377/30000 [3:07:35<70:53:05,  8.92s/it, lr: 1.0e-04 loss: 4.796e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1377/30000 [3:07:44<70:53:05,  8.92s/it, lr: 1.0e-04 loss: 6.316e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1377/30000 [3:07:44<70:53:05,  8.92s/it, lr: 1.0e-04 loss: 6.316e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1378/30000 [3:07:44<71:51:30,  9.04s/it, lr: 1.0e-04 loss: 6.316e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1378/30000 [3:07:44<71:51:30,  9.04s/it, lr: 1.0e-04 loss: 6.316e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1378/30000 [3:07:53<71:51:30,  9.04s/it, lr: 1.0e-04 loss: 3.491e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1378/30000 [3:07:53<71:51:30,  9.04s/it, lr: 1.0e-04 loss: 3.491e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1379/30000 [3:07:53<72:38:32,  9.14s/it, lr: 1.0e-04 loss: 3.491e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1379/30000 [3:07:53<72:38:32,  9.14s/it, lr: 1.0e-04 loss: 3.491e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1379/30000 [3:08:00<72:38:32,  9.14s/it, lr: 1.0e-04 loss: 7.338e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1379/30000 [3:08:00<72:38:32,  9.14s/it, lr: 1.0e-04 loss: 7.338e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1380/30000 [3:08:07<72:38:23,  9.14s/it, lr: 1.0e-04 loss: 2.175e-01]
Xiang_wan_2_2_lora:   5%|#7                                     | 1380/30000 [3:08:07<72:38:23,  9.14s/it, lr: 1.0e-04 loss: 2.175e-01]
Xiang_wan_2_2_lora:   5%|#7                                     | 1381/30000 [3:08:07<51:26:04,  6.47s/it, lr: 1.0e-04 loss: 2.175e-01]
Xiang_wan_2_2_lora:   5%|#7                                     | 1381/30000 [3:08:07<51:26:04,  6.47s/it, lr: 1.0e-04 loss: 2.175e-01]
Xiang_wan_2_2_lora:   5%|#7                                     | 1381/30000 [3:08:13<51:26:04,  6.47s/it, lr: 1.0e-04 loss: 7.214e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1381/30000 [3:08:13<51:26:04,  6.47s/it, lr: 1.0e-04 loss: 7.214e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1382/30000 [3:08:13<51:41:13,  6.50s/it, lr: 1.0e-04 loss: 7.214e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1382/30000 [3:08:13<51:41:13,  6.50s/it, lr: 1.0e-04 loss: 7.214e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1382/30000 [3:08:23<51:41:13,  6.50s/it, lr: 1.0e-04 loss: 1.355e-01]
Xiang_wan_2_2_lora:   5%|#7                                     | 1382/30000 [3:08:23<51:41:13,  6.50s/it, lr: 1.0e-04 loss: 1.355e-01]
Xiang_wan_2_2_lora:   5%|#7                                     | 1383/30000 [3:08:23<57:55:27,  7.29s/it, lr: 1.0e-04 loss: 1.355e-01]
Xiang_wan_2_2_lora:   5%|#7                                     | 1383/30000 [3:08:23<57:55:27,  7.29s/it, lr: 1.0e-04 loss: 1.355e-01]
Xiang_wan_2_2_lora:   5%|#7                                     | 1383/30000 [3:08:29<57:55:27,  7.29s/it, lr: 1.0e-04 loss: 8.770e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1383/30000 [3:08:29<57:55:27,  7.29s/it, lr: 1.0e-04 loss: 8.770e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1384/30000 [3:08:29<56:32:20,  7.11s/it, lr: 1.0e-04 loss: 8.770e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1384/30000 [3:08:29<56:32:20,  7.11s/it, lr: 1.0e-04 loss: 8.770e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1384/30000 [3:08:39<56:32:20,  7.11s/it, lr: 1.0e-04 loss: 3.637e-02]
Xiang_wan_2_2_lora:   5%|#7                                     | 1384/30000 [3:08:39<56:32:20,  7.11s/it, lr: 1.0e-04 loss: 3.637e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1385/30000 [3:08:39<61:27:13,  7.73s/it, lr: 1.0e-04 loss: 3.637e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1385/30000 [3:08:39<61:27:13,  7.73s/it, lr: 1.0e-04 loss: 3.637e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1385/30000 [3:08:45<61:27:13,  7.73s/it, lr: 1.0e-04 loss: 9.855e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1385/30000 [3:08:45<61:27:13,  7.73s/it, lr: 1.0e-04 loss: 9.855e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1386/30000 [3:08:45<58:53:05,  7.41s/it, lr: 1.0e-04 loss: 9.855e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1386/30000 [3:08:45<58:53:05,  7.41s/it, lr: 1.0e-04 loss: 9.855e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1386/30000 [3:08:55<58:53:05,  7.41s/it, lr: 1.0e-04 loss: 5.819e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1386/30000 [3:08:55<58:53:05,  7.41s/it, lr: 1.0e-04 loss: 5.819e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1387/30000 [3:08:55<63:16:36,  7.96s/it, lr: 1.0e-04 loss: 5.819e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1387/30000 [3:08:55<63:16:36,  7.96s/it, lr: 1.0e-04 loss: 5.819e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1387/30000 [3:09:01<63:16:36,  7.96s/it, lr: 1.0e-04 loss: 8.627e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1387/30000 [3:09:01<63:16:36,  7.96s/it, lr: 1.0e-04 loss: 8.627e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1388/30000 [3:09:01<60:11:14,  7.57s/it, lr: 1.0e-04 loss: 8.627e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1388/30000 [3:09:01<60:11:14,  7.57s/it, lr: 1.0e-04 loss: 8.627e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1388/30000 [3:09:11<60:11:14,  7.57s/it, lr: 1.0e-04 loss: 4.959e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1388/30000 [3:09:11<60:11:14,  7.57s/it, lr: 1.0e-04 loss: 4.959e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1389/30000 [3:09:11<64:25:45,  8.11s/it, lr: 1.0e-04 loss: 4.959e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1389/30000 [3:09:11<64:25:45,  8.11s/it, lr: 1.0e-04 loss: 4.959e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1389/30000 [3:09:20<64:25:45,  8.11s/it, lr: 1.0e-04 loss: 8.123e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1389/30000 [3:09:20<64:25:45,  8.11s/it, lr: 1.0e-04 loss: 8.123e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1390/30000 [3:09:30<64:25:37,  8.11s/it, lr: 1.0e-04 loss: 3.562e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1390/30000 [3:09:30<64:25:37,  8.11s/it, lr: 1.0e-04 loss: 3.562e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1391/30000 [3:09:30<51:54:28,  6.53s/it, lr: 1.0e-04 loss: 3.562e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1391/30000 [3:09:30<51:54:28,  6.53s/it, lr: 1.0e-04 loss: 3.562e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1391/30000 [3:09:39<51:54:28,  6.53s/it, lr: 1.0e-04 loss: 1.103e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1391/30000 [3:09:39<51:54:28,  6.53s/it, lr: 1.0e-04 loss: 1.103e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1392/30000 [3:09:39<57:32:11,  7.24s/it, lr: 1.0e-04 loss: 1.103e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1392/30000 [3:09:39<57:32:11,  7.24s/it, lr: 1.0e-04 loss: 1.103e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1392/30000 [3:09:48<57:32:11,  7.24s/it, lr: 1.0e-04 loss: 6.542e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1392/30000 [3:09:48<57:32:11,  7.24s/it, lr: 1.0e-04 loss: 6.542e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1393/30000 [3:09:48<61:50:55,  7.78s/it, lr: 1.0e-04 loss: 6.542e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1393/30000 [3:09:48<61:50:55,  7.78s/it, lr: 1.0e-04 loss: 6.542e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1393/30000 [3:09:55<61:50:55,  7.78s/it, lr: 1.0e-04 loss: 1.279e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1393/30000 [3:09:55<61:50:55,  7.78s/it, lr: 1.0e-04 loss: 1.279e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1394/30000 [3:09:55<59:20:20,  7.47s/it, lr: 1.0e-04 loss: 1.279e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1394/30000 [3:09:55<59:20:20,  7.47s/it, lr: 1.0e-04 loss: 1.279e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1394/30000 [3:10:01<59:20:20,  7.47s/it, lr: 1.0e-04 loss: 1.903e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1394/30000 [3:10:01<59:20:20,  7.47s/it, lr: 1.0e-04 loss: 1.903e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1395/30000 [3:10:01<57:19:18,  7.21s/it, lr: 1.0e-04 loss: 1.903e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1395/30000 [3:10:01<57:19:18,  7.21s/it, lr: 1.0e-04 loss: 1.903e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1395/30000 [3:10:08<57:19:18,  7.21s/it, lr: 1.0e-04 loss: 1.251e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1395/30000 [3:10:08<57:19:18,  7.21s/it, lr: 1.0e-04 loss: 1.251e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1396/30000 [3:10:08<55:49:46,  7.03s/it, lr: 1.0e-04 loss: 1.251e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1396/30000 [3:10:08<55:49:46,  7.03s/it, lr: 1.0e-04 loss: 1.251e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1396/30000 [3:10:15<55:49:46,  7.03s/it, lr: 1.0e-04 loss: 7.215e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1396/30000 [3:10:15<55:49:46,  7.03s/it, lr: 1.0e-04 loss: 7.215e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1397/30000 [3:10:15<55:22:22,  6.97s/it, lr: 1.0e-04 loss: 7.215e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1397/30000 [3:10:15<55:22:22,  6.97s/it, lr: 1.0e-04 loss: 7.215e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1397/30000 [3:10:24<55:22:22,  6.97s/it, lr: 1.0e-04 loss: 6.978e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1397/30000 [3:10:24<55:22:22,  6.97s/it, lr: 1.0e-04 loss: 6.978e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1398/30000 [3:10:24<60:52:22,  7.66s/it, lr: 1.0e-04 loss: 6.978e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1398/30000 [3:10:24<60:52:22,  7.66s/it, lr: 1.0e-04 loss: 6.978e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1398/30000 [3:10:33<60:52:22,  7.66s/it, lr: 1.0e-04 loss: 9.821e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1398/30000 [3:10:33<60:52:22,  7.66s/it, lr: 1.0e-04 loss: 9.821e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1399/30000 [3:10:33<64:44:55,  8.15s/it, lr: 1.0e-04 loss: 9.821e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1399/30000 [3:10:33<64:44:55,  8.15s/it, lr: 1.0e-04 loss: 9.821e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1399/30000 [3:10:43<64:44:55,  8.15s/it, lr: 1.0e-04 loss: 7.405e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1399/30000 [3:10:43<64:44:55,  8.15s/it, lr: 1.0e-04 loss: 7.405e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1400/30000 [3:10:52<64:44:47,  8.15s/it, lr: 1.0e-04 loss: 6.120e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1400/30000 [3:10:52<64:44:47,  8.15s/it, lr: 1.0e-04 loss: 6.120e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1401/30000 [3:10:52<52:04:41,  6.56s/it, lr: 1.0e-04 loss: 6.120e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1401/30000 [3:10:52<52:04:41,  6.56s/it, lr: 1.0e-04 loss: 6.120e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1401/30000 [3:11:02<52:04:41,  6.56s/it, lr: 1.0e-04 loss: 8.131e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1401/30000 [3:11:02<52:04:41,  6.56s/it, lr: 1.0e-04 loss: 8.131e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1402/30000 [3:11:02<57:36:53,  7.25s/it, lr: 1.0e-04 loss: 8.131e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1402/30000 [3:11:02<57:36:53,  7.25s/it, lr: 1.0e-04 loss: 8.131e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1402/30000 [3:11:11<57:36:53,  7.25s/it, lr: 1.0e-04 loss: 5.334e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1402/30000 [3:11:11<57:36:53,  7.25s/it, lr: 1.0e-04 loss: 5.334e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1403/30000 [3:11:11<62:01:20,  7.81s/it, lr: 1.0e-04 loss: 5.334e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1403/30000 [3:11:11<62:01:20,  7.81s/it, lr: 1.0e-04 loss: 5.334e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1403/30000 [3:11:21<62:01:20,  7.81s/it, lr: 1.0e-04 loss: 6.182e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1403/30000 [3:11:21<62:01:20,  7.81s/it, lr: 1.0e-04 loss: 6.182e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1404/30000 [3:11:21<65:47:12,  8.28s/it, lr: 1.0e-04 loss: 6.182e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1404/30000 [3:11:21<65:47:12,  8.28s/it, lr: 1.0e-04 loss: 6.182e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1404/30000 [3:11:27<65:47:12,  8.28s/it, lr: 1.0e-04 loss: 7.353e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1404/30000 [3:11:27<65:47:12,  8.28s/it, lr: 1.0e-04 loss: 7.353e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1405/30000 [3:11:27<62:01:36,  7.81s/it, lr: 1.0e-04 loss: 7.353e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1405/30000 [3:11:27<62:01:36,  7.81s/it, lr: 1.0e-04 loss: 7.353e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1405/30000 [3:11:34<62:01:36,  7.81s/it, lr: 1.0e-04 loss: 1.033e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1405/30000 [3:11:34<62:01:36,  7.81s/it, lr: 1.0e-04 loss: 1.033e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1406/30000 [3:11:34<59:23:22,  7.48s/it, lr: 1.0e-04 loss: 1.033e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1406/30000 [3:11:34<59:23:22,  7.48s/it, lr: 1.0e-04 loss: 1.033e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1406/30000 [3:11:43<59:23:22,  7.48s/it, lr: 1.0e-04 loss: 5.317e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1406/30000 [3:11:43<59:23:22,  7.48s/it, lr: 1.0e-04 loss: 5.317e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1407/30000 [3:11:43<63:32:51,  8.00s/it, lr: 1.0e-04 loss: 5.317e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1407/30000 [3:11:43<63:32:51,  8.00s/it, lr: 1.0e-04 loss: 5.317e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1407/30000 [3:11:50<63:32:51,  8.00s/it, lr: 1.0e-04 loss: 9.536e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1407/30000 [3:11:50<63:32:51,  8.00s/it, lr: 1.0e-04 loss: 9.536e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1408/30000 [3:11:50<60:15:24,  7.59s/it, lr: 1.0e-04 loss: 9.536e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1408/30000 [3:11:50<60:15:24,  7.59s/it, lr: 1.0e-04 loss: 9.536e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1408/30000 [3:11:56<60:15:24,  7.59s/it, lr: 1.0e-04 loss: 6.328e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1408/30000 [3:11:56<60:15:24,  7.59s/it, lr: 1.0e-04 loss: 6.328e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1409/30000 [3:11:56<57:38:30,  7.26s/it, lr: 1.0e-04 loss: 6.328e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1409/30000 [3:11:56<57:38:30,  7.26s/it, lr: 1.0e-04 loss: 6.328e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1409/30000 [3:12:05<57:38:30,  7.26s/it, lr: 1.0e-04 loss: 7.471e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1409/30000 [3:12:05<57:38:30,  7.26s/it, lr: 1.0e-04 loss: 7.471e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1410/30000 [3:12:12<57:38:23,  7.26s/it, lr: 1.0e-04 loss: 1.573e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1410/30000 [3:12:12<57:38:23,  7.26s/it, lr: 1.0e-04 loss: 1.573e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1411/30000 [3:12:12<43:19:58,  5.46s/it, lr: 1.0e-04 loss: 1.573e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1411/30000 [3:12:12<43:19:58,  5.46s/it, lr: 1.0e-04 loss: 1.573e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1411/30000 [3:12:19<43:19:58,  5.46s/it, lr: 1.0e-04 loss: 8.813e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1411/30000 [3:12:19<43:19:58,  5.46s/it, lr: 1.0e-04 loss: 8.813e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1412/30000 [3:12:19<45:50:19,  5.77s/it, lr: 1.0e-04 loss: 8.813e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1412/30000 [3:12:19<45:50:19,  5.77s/it, lr: 1.0e-04 loss: 8.813e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1412/30000 [3:12:25<45:50:19,  5.77s/it, lr: 1.0e-04 loss: 6.158e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1412/30000 [3:12:25<45:50:19,  5.77s/it, lr: 1.0e-04 loss: 6.158e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1413/30000 [3:12:25<47:28:00,  5.98s/it, lr: 1.0e-04 loss: 6.158e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1413/30000 [3:12:25<47:28:00,  5.98s/it, lr: 1.0e-04 loss: 6.158e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1413/30000 [3:12:32<47:28:00,  5.98s/it, lr: 1.0e-04 loss: 1.198e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1413/30000 [3:12:32<47:28:00,  5.98s/it, lr: 1.0e-04 loss: 1.198e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1414/30000 [3:12:32<48:53:27,  6.16s/it, lr: 1.0e-04 loss: 1.198e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1414/30000 [3:12:32<48:53:27,  6.16s/it, lr: 1.0e-04 loss: 1.198e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1414/30000 [3:12:41<48:53:27,  6.16s/it, lr: 1.0e-04 loss: 5.284e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1414/30000 [3:12:41<48:53:27,  6.16s/it, lr: 1.0e-04 loss: 5.284e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1415/30000 [3:12:41<55:45:19,  7.02s/it, lr: 1.0e-04 loss: 5.284e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1415/30000 [3:12:41<55:45:19,  7.02s/it, lr: 1.0e-04 loss: 5.284e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1415/30000 [3:12:48<55:45:19,  7.02s/it, lr: 1.0e-04 loss: 7.674e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1415/30000 [3:12:48<55:45:19,  7.02s/it, lr: 1.0e-04 loss: 7.674e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1416/30000 [3:12:48<54:48:32,  6.90s/it, lr: 1.0e-04 loss: 7.674e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1416/30000 [3:12:48<54:48:32,  6.90s/it, lr: 1.0e-04 loss: 7.674e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1416/30000 [3:12:54<54:48:32,  6.90s/it, lr: 1.0e-04 loss: 1.297e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1416/30000 [3:12:54<54:48:32,  6.90s/it, lr: 1.0e-04 loss: 1.297e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1417/30000 [3:12:54<53:53:53,  6.79s/it, lr: 1.0e-04 loss: 1.297e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1417/30000 [3:12:54<53:53:53,  6.79s/it, lr: 1.0e-04 loss: 1.297e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1417/30000 [3:13:04<53:53:53,  6.79s/it, lr: 1.0e-04 loss: 7.834e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1417/30000 [3:13:04<53:53:53,  6.79s/it, lr: 1.0e-04 loss: 7.834e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1418/30000 [3:13:04<59:50:29,  7.54s/it, lr: 1.0e-04 loss: 7.834e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1418/30000 [3:13:04<59:50:29,  7.54s/it, lr: 1.0e-04 loss: 7.834e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1418/30000 [3:13:11<59:50:29,  7.54s/it, lr: 1.0e-04 loss: 9.805e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1418/30000 [3:13:11<59:50:29,  7.54s/it, lr: 1.0e-04 loss: 9.805e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1419/30000 [3:13:11<58:10:52,  7.33s/it, lr: 1.0e-04 loss: 9.805e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1419/30000 [3:13:11<58:10:52,  7.33s/it, lr: 1.0e-04 loss: 9.805e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1419/30000 [3:13:20<58:10:52,  7.33s/it, lr: 1.0e-04 loss: 8.198e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1419/30000 [3:13:20<58:10:52,  7.33s/it, lr: 1.0e-04 loss: 8.198e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1420/30000 [3:13:26<58:10:45,  7.33s/it, lr: 1.0e-04 loss: 6.857e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1420/30000 [3:13:26<58:10:45,  7.33s/it, lr: 1.0e-04 loss: 6.857e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1421/30000 [3:13:26<43:30:53,  5.48s/it, lr: 1.0e-04 loss: 6.857e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1421/30000 [3:13:26<43:30:53,  5.48s/it, lr: 1.0e-04 loss: 6.857e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1421/30000 [3:13:36<43:30:53,  5.48s/it, lr: 1.0e-04 loss: 5.875e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1421/30000 [3:13:36<43:30:53,  5.48s/it, lr: 1.0e-04 loss: 5.875e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1422/30000 [3:13:36<51:04:23,  6.43s/it, lr: 1.0e-04 loss: 5.875e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1422/30000 [3:13:36<51:04:23,  6.43s/it, lr: 1.0e-04 loss: 5.875e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1422/30000 [3:13:42<51:04:23,  6.43s/it, lr: 1.0e-04 loss: 1.820e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1422/30000 [3:13:42<51:04:23,  6.43s/it, lr: 1.0e-04 loss: 1.820e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1423/30000 [3:13:42<51:19:15,  6.47s/it, lr: 1.0e-04 loss: 1.820e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1423/30000 [3:13:42<51:19:15,  6.47s/it, lr: 1.0e-04 loss: 1.820e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1423/30000 [3:13:49<51:19:15,  6.47s/it, lr: 1.0e-04 loss: 9.644e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1423/30000 [3:13:49<51:19:15,  6.47s/it, lr: 1.0e-04 loss: 9.644e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1424/30000 [3:13:49<51:29:32,  6.49s/it, lr: 1.0e-04 loss: 9.644e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1424/30000 [3:13:49<51:29:32,  6.49s/it, lr: 1.0e-04 loss: 9.644e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1424/30000 [3:13:58<51:29:32,  6.49s/it, lr: 1.0e-04 loss: 6.360e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1424/30000 [3:13:58<51:29:32,  6.49s/it, lr: 1.0e-04 loss: 6.360e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1425/30000 [3:13:58<57:43:37,  7.27s/it, lr: 1.0e-04 loss: 6.360e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1425/30000 [3:13:58<57:43:37,  7.27s/it, lr: 1.0e-04 loss: 6.360e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1425/30000 [3:14:08<57:43:37,  7.27s/it, lr: 1.0e-04 loss: 5.009e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1425/30000 [3:14:08<57:43:37,  7.27s/it, lr: 1.0e-04 loss: 5.009e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1426/30000 [3:14:08<62:52:11,  7.92s/it, lr: 1.0e-04 loss: 5.009e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1426/30000 [3:14:08<62:52:11,  7.92s/it, lr: 1.0e-04 loss: 5.009e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1426/30000 [3:14:17<62:52:11,  7.92s/it, lr: 1.0e-04 loss: 4.030e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1426/30000 [3:14:17<62:52:11,  7.92s/it, lr: 1.0e-04 loss: 4.030e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1427/30000 [3:14:17<66:06:52,  8.33s/it, lr: 1.0e-04 loss: 4.030e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1427/30000 [3:14:17<66:06:52,  8.33s/it, lr: 1.0e-04 loss: 4.030e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1427/30000 [3:14:24<66:06:52,  8.33s/it, lr: 1.0e-04 loss: 7.466e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1427/30000 [3:14:24<66:06:52,  8.33s/it, lr: 1.0e-04 loss: 7.466e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1428/30000 [3:14:24<62:20:31,  7.85s/it, lr: 1.0e-04 loss: 7.466e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1428/30000 [3:14:24<62:20:31,  7.85s/it, lr: 1.0e-04 loss: 7.466e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1428/30000 [3:14:30<62:20:31,  7.85s/it, lr: 1.0e-04 loss: 5.310e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1428/30000 [3:14:30<62:20:31,  7.85s/it, lr: 1.0e-04 loss: 5.310e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1429/30000 [3:14:30<59:07:04,  7.45s/it, lr: 1.0e-04 loss: 5.310e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1429/30000 [3:14:30<59:07:04,  7.45s/it, lr: 1.0e-04 loss: 5.310e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1429/30000 [3:14:37<59:07:04,  7.45s/it, lr: 1.0e-04 loss: 6.408e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1429/30000 [3:14:37<59:07:04,  7.45s/it, lr: 1.0e-04 loss: 6.408e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1430/30000 [3:14:43<59:06:56,  7.45s/it, lr: 1.0e-04 loss: 6.190e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1430/30000 [3:14:43<59:06:56,  7.45s/it, lr: 1.0e-04 loss: 6.190e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1431/30000 [3:14:43<43:56:48,  5.54s/it, lr: 1.0e-04 loss: 6.190e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1431/30000 [3:14:43<43:56:48,  5.54s/it, lr: 1.0e-04 loss: 6.190e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1431/30000 [3:14:50<43:56:48,  5.54s/it, lr: 1.0e-04 loss: 8.477e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1431/30000 [3:14:50<43:56:48,  5.54s/it, lr: 1.0e-04 loss: 8.477e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1432/30000 [3:14:50<46:00:44,  5.80s/it, lr: 1.0e-04 loss: 8.477e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1432/30000 [3:14:50<46:00:44,  5.80s/it, lr: 1.0e-04 loss: 8.477e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1432/30000 [3:15:00<46:00:44,  5.80s/it, lr: 1.0e-04 loss: 4.114e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1432/30000 [3:15:00<46:00:44,  5.80s/it, lr: 1.0e-04 loss: 4.114e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1433/30000 [3:15:00<53:42:58,  6.77s/it, lr: 1.0e-04 loss: 4.114e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1433/30000 [3:15:00<53:42:58,  6.77s/it, lr: 1.0e-04 loss: 4.114e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1433/30000 [3:15:09<53:42:58,  6.77s/it, lr: 1.0e-04 loss: 6.304e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1433/30000 [3:15:09<53:42:58,  6.77s/it, lr: 1.0e-04 loss: 6.304e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1434/30000 [3:15:09<59:18:31,  7.47s/it, lr: 1.0e-04 loss: 6.304e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1434/30000 [3:15:09<59:18:31,  7.47s/it, lr: 1.0e-04 loss: 6.304e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1434/30000 [3:15:18<59:18:31,  7.47s/it, lr: 1.0e-04 loss: 3.015e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1434/30000 [3:15:18<59:18:31,  7.47s/it, lr: 1.0e-04 loss: 3.015e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1435/30000 [3:15:18<63:28:14,  8.00s/it, lr: 1.0e-04 loss: 3.015e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1435/30000 [3:15:18<63:28:14,  8.00s/it, lr: 1.0e-04 loss: 3.015e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1435/30000 [3:15:28<63:28:14,  8.00s/it, lr: 1.0e-04 loss: 6.530e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1435/30000 [3:15:28<63:28:14,  8.00s/it, lr: 1.0e-04 loss: 6.530e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1436/30000 [3:15:28<66:27:21,  8.38s/it, lr: 1.0e-04 loss: 6.530e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1436/30000 [3:15:28<66:27:21,  8.38s/it, lr: 1.0e-04 loss: 6.530e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1436/30000 [3:15:37<66:27:21,  8.38s/it, lr: 1.0e-04 loss: 4.737e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1436/30000 [3:15:37<66:27:21,  8.38s/it, lr: 1.0e-04 loss: 4.737e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1437/30000 [3:15:37<68:47:29,  8.67s/it, lr: 1.0e-04 loss: 4.737e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1437/30000 [3:15:37<68:47:29,  8.67s/it, lr: 1.0e-04 loss: 4.737e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1437/30000 [3:15:46<68:47:29,  8.67s/it, lr: 1.0e-04 loss: 5.292e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1437/30000 [3:15:46<68:47:29,  8.67s/it, lr: 1.0e-04 loss: 5.292e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1438/30000 [3:15:46<70:24:33,  8.87s/it, lr: 1.0e-04 loss: 5.292e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1438/30000 [3:15:46<70:24:33,  8.87s/it, lr: 1.0e-04 loss: 5.292e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1438/30000 [3:15:53<70:24:33,  8.87s/it, lr: 1.0e-04 loss: 8.593e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1438/30000 [3:15:53<70:24:33,  8.87s/it, lr: 1.0e-04 loss: 8.593e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1439/30000 [3:15:53<65:02:49,  8.20s/it, lr: 1.0e-04 loss: 8.593e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1439/30000 [3:15:53<65:02:49,  8.20s/it, lr: 1.0e-04 loss: 8.593e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1439/30000 [3:16:02<65:02:49,  8.20s/it, lr: 1.0e-04 loss: 8.712e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1439/30000 [3:16:02<65:02:49,  8.20s/it, lr: 1.0e-04 loss: 8.712e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1440/30000 [3:16:12<65:02:41,  8.20s/it, lr: 1.0e-04 loss: 4.835e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1440/30000 [3:16:12<65:02:41,  8.20s/it, lr: 1.0e-04 loss: 4.835e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1441/30000 [3:16:12<52:41:19,  6.64s/it, lr: 1.0e-04 loss: 4.835e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1441/30000 [3:16:12<52:41:19,  6.64s/it, lr: 1.0e-04 loss: 4.835e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1441/30000 [3:16:18<52:41:19,  6.64s/it, lr: 1.0e-04 loss: 1.164e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1441/30000 [3:16:18<52:41:19,  6.64s/it, lr: 1.0e-04 loss: 1.164e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1442/30000 [3:16:18<52:31:19,  6.62s/it, lr: 1.0e-04 loss: 1.164e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1442/30000 [3:16:18<52:31:19,  6.62s/it, lr: 1.0e-04 loss: 1.164e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1442/30000 [3:16:25<52:31:19,  6.62s/it, lr: 1.0e-04 loss: 4.681e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1442/30000 [3:16:25<52:31:19,  6.62s/it, lr: 1.0e-04 loss: 4.681e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1443/30000 [3:16:25<52:32:29,  6.62s/it, lr: 1.0e-04 loss: 4.681e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1443/30000 [3:16:25<52:32:29,  6.62s/it, lr: 1.0e-04 loss: 4.681e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1443/30000 [3:16:32<52:32:29,  6.62s/it, lr: 1.0e-04 loss: 7.694e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1443/30000 [3:16:32<52:32:29,  6.62s/it, lr: 1.0e-04 loss: 7.694e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1444/30000 [3:16:32<52:27:54,  6.61s/it, lr: 1.0e-04 loss: 7.694e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1444/30000 [3:16:32<52:27:54,  6.61s/it, lr: 1.0e-04 loss: 7.694e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1444/30000 [3:16:38<52:27:54,  6.61s/it, lr: 1.0e-04 loss: 6.581e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1444/30000 [3:16:38<52:27:54,  6.61s/it, lr: 1.0e-04 loss: 6.581e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1445/30000 [3:16:38<52:28:58,  6.62s/it, lr: 1.0e-04 loss: 6.581e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1445/30000 [3:16:38<52:28:58,  6.62s/it, lr: 1.0e-04 loss: 6.581e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1445/30000 [3:16:45<52:28:58,  6.62s/it, lr: 1.0e-04 loss: 1.041e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1445/30000 [3:16:45<52:28:58,  6.62s/it, lr: 1.0e-04 loss: 1.041e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1446/30000 [3:16:45<52:29:51,  6.62s/it, lr: 1.0e-04 loss: 1.041e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1446/30000 [3:16:45<52:29:51,  6.62s/it, lr: 1.0e-04 loss: 1.041e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1446/30000 [3:16:54<52:29:51,  6.62s/it, lr: 1.0e-04 loss: 3.292e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1446/30000 [3:16:54<52:29:51,  6.62s/it, lr: 1.0e-04 loss: 3.292e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1447/30000 [3:16:54<58:38:55,  7.39s/it, lr: 1.0e-04 loss: 3.292e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1447/30000 [3:16:54<58:38:55,  7.39s/it, lr: 1.0e-04 loss: 3.292e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1447/30000 [3:17:04<58:38:55,  7.39s/it, lr: 1.0e-04 loss: 9.196e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1447/30000 [3:17:04<58:38:55,  7.39s/it, lr: 1.0e-04 loss: 9.196e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1448/30000 [3:17:04<63:43:07,  8.03s/it, lr: 1.0e-04 loss: 9.196e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1448/30000 [3:17:04<63:43:07,  8.03s/it, lr: 1.0e-04 loss: 9.196e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1448/30000 [3:17:13<63:43:07,  8.03s/it, lr: 1.0e-04 loss: 1.219e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1448/30000 [3:17:13<63:43:07,  8.03s/it, lr: 1.0e-04 loss: 1.219e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1449/30000 [3:17:13<66:46:14,  8.42s/it, lr: 1.0e-04 loss: 1.219e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1449/30000 [3:17:13<66:46:14,  8.42s/it, lr: 1.0e-04 loss: 1.219e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1449/30000 [3:17:22<66:46:14,  8.42s/it, lr: 1.0e-04 loss: 7.674e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1449/30000 [3:17:22<66:46:14,  8.42s/it, lr: 1.0e-04 loss: 7.674e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1450/30000 [3:17:29<66:46:06,  8.42s/it, lr: 1.0e-04 loss: 4.719e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1450/30000 [3:17:29<66:46:06,  8.42s/it, lr: 1.0e-04 loss: 4.719e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1451/30000 [3:17:29<48:11:21,  6.08s/it, lr: 1.0e-04 loss: 4.719e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1451/30000 [3:17:29<48:11:21,  6.08s/it, lr: 1.0e-04 loss: 4.719e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1451/30000 [3:17:38<48:11:21,  6.08s/it, lr: 1.0e-04 loss: 5.483e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1451/30000 [3:17:38<48:11:21,  6.08s/it, lr: 1.0e-04 loss: 5.483e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1452/30000 [3:17:38<54:33:33,  6.88s/it, lr: 1.0e-04 loss: 5.483e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1452/30000 [3:17:38<54:33:33,  6.88s/it, lr: 1.0e-04 loss: 5.483e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1452/30000 [3:17:45<54:33:33,  6.88s/it, lr: 1.0e-04 loss: 2.553e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1452/30000 [3:17:45<54:33:33,  6.88s/it, lr: 1.0e-04 loss: 2.553e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1453/30000 [3:17:45<54:29:56,  6.87s/it, lr: 1.0e-04 loss: 2.553e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1453/30000 [3:17:45<54:29:56,  6.87s/it, lr: 1.0e-04 loss: 2.553e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1453/30000 [3:17:52<54:29:56,  6.87s/it, lr: 1.0e-04 loss: 7.412e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1453/30000 [3:17:52<54:29:56,  6.87s/it, lr: 1.0e-04 loss: 7.412e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1454/30000 [3:17:52<53:48:45,  6.79s/it, lr: 1.0e-04 loss: 7.412e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1454/30000 [3:17:52<53:48:45,  6.79s/it, lr: 1.0e-04 loss: 7.412e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1454/30000 [3:17:59<53:48:45,  6.79s/it, lr: 1.0e-04 loss: 1.895e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1454/30000 [3:17:59<53:48:45,  6.79s/it, lr: 1.0e-04 loss: 1.895e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1455/30000 [3:17:59<54:08:21,  6.83s/it, lr: 1.0e-04 loss: 1.895e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1455/30000 [3:17:59<54:08:21,  6.83s/it, lr: 1.0e-04 loss: 1.895e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1455/30000 [3:18:05<54:08:21,  6.83s/it, lr: 1.0e-04 loss: 1.097e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1455/30000 [3:18:05<54:08:21,  6.83s/it, lr: 1.0e-04 loss: 1.097e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1456/30000 [3:18:05<53:39:59,  6.77s/it, lr: 1.0e-04 loss: 1.097e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1456/30000 [3:18:05<53:39:59,  6.77s/it, lr: 1.0e-04 loss: 1.097e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1456/30000 [3:18:15<53:39:59,  6.77s/it, lr: 1.0e-04 loss: 4.739e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1456/30000 [3:18:15<53:39:59,  6.77s/it, lr: 1.0e-04 loss: 4.739e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1457/30000 [3:18:15<59:34:11,  7.51s/it, lr: 1.0e-04 loss: 4.739e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1457/30000 [3:18:15<59:34:11,  7.51s/it, lr: 1.0e-04 loss: 4.739e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1457/30000 [3:18:21<59:34:11,  7.51s/it, lr: 1.0e-04 loss: 7.955e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1457/30000 [3:18:21<59:34:11,  7.51s/it, lr: 1.0e-04 loss: 7.955e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1458/30000 [3:18:21<57:25:20,  7.24s/it, lr: 1.0e-04 loss: 7.955e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1458/30000 [3:18:21<57:25:20,  7.24s/it, lr: 1.0e-04 loss: 7.955e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1458/30000 [3:18:31<57:25:20,  7.24s/it, lr: 1.0e-04 loss: 1.439e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1458/30000 [3:18:31<57:25:20,  7.24s/it, lr: 1.0e-04 loss: 1.439e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1459/30000 [3:18:31<62:25:24,  7.87s/it, lr: 1.0e-04 loss: 1.439e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1459/30000 [3:18:31<62:25:24,  7.87s/it, lr: 1.0e-04 loss: 1.439e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1459/30000 [3:18:37<62:25:24,  7.87s/it, lr: 1.0e-04 loss: 6.264e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1459/30000 [3:18:37<62:25:24,  7.87s/it, lr: 1.0e-04 loss: 6.264e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1460/30000 [3:18:44<62:25:17,  7.87s/it, lr: 1.0e-04 loss: 3.016e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1460/30000 [3:18:44<62:25:17,  7.87s/it, lr: 1.0e-04 loss: 3.016e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1461/30000 [3:18:44<45:54:37,  5.79s/it, lr: 1.0e-04 loss: 3.016e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1461/30000 [3:18:44<45:54:37,  5.79s/it, lr: 1.0e-04 loss: 3.016e-01]
Xiang_wan_2_2_lora:   5%|#8                                     | 1461/30000 [3:18:53<45:54:37,  5.79s/it, lr: 1.0e-04 loss: 8.687e-02]
Xiang_wan_2_2_lora:   5%|#8                                     | 1461/30000 [3:18:53<45:54:37,  5.79s/it, lr: 1.0e-04 loss: 8.687e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1462/30000 [3:18:53<52:47:37,  6.66s/it, lr: 1.0e-04 loss: 8.687e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1462/30000 [3:18:53<52:47:37,  6.66s/it, lr: 1.0e-04 loss: 8.687e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1462/30000 [3:19:00<52:47:37,  6.66s/it, lr: 1.0e-04 loss: 8.614e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1462/30000 [3:19:00<52:47:37,  6.66s/it, lr: 1.0e-04 loss: 8.614e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1463/30000 [3:19:00<53:12:00,  6.71s/it, lr: 1.0e-04 loss: 8.614e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1463/30000 [3:19:00<53:12:00,  6.71s/it, lr: 1.0e-04 loss: 8.614e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1463/30000 [3:19:07<53:12:00,  6.71s/it, lr: 1.0e-04 loss: 7.874e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1463/30000 [3:19:07<53:12:00,  6.71s/it, lr: 1.0e-04 loss: 7.874e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1464/30000 [3:19:07<52:53:09,  6.67s/it, lr: 1.0e-04 loss: 7.874e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1464/30000 [3:19:07<52:53:09,  6.67s/it, lr: 1.0e-04 loss: 7.874e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1464/30000 [3:19:16<52:53:09,  6.67s/it, lr: 1.0e-04 loss: 1.169e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1464/30000 [3:19:16<52:53:09,  6.67s/it, lr: 1.0e-04 loss: 1.169e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1465/30000 [3:19:16<58:45:18,  7.41s/it, lr: 1.0e-04 loss: 1.169e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1465/30000 [3:19:16<58:45:18,  7.41s/it, lr: 1.0e-04 loss: 1.169e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1465/30000 [3:19:23<58:45:18,  7.41s/it, lr: 1.0e-04 loss: 8.451e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1465/30000 [3:19:23<58:45:18,  7.41s/it, lr: 1.0e-04 loss: 8.451e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1466/30000 [3:19:23<57:02:11,  7.20s/it, lr: 1.0e-04 loss: 8.451e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1466/30000 [3:19:23<57:02:11,  7.20s/it, lr: 1.0e-04 loss: 8.451e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1466/30000 [3:19:32<57:02:11,  7.20s/it, lr: 1.0e-04 loss: 6.284e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1466/30000 [3:19:32<57:02:11,  7.20s/it, lr: 1.0e-04 loss: 6.284e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1467/30000 [3:19:32<62:01:39,  7.83s/it, lr: 1.0e-04 loss: 6.284e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1467/30000 [3:19:32<62:01:39,  7.83s/it, lr: 1.0e-04 loss: 6.284e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1467/30000 [3:19:41<62:01:39,  7.83s/it, lr: 1.0e-04 loss: 7.955e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1467/30000 [3:19:41<62:01:39,  7.83s/it, lr: 1.0e-04 loss: 7.955e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1468/30000 [3:19:41<65:39:42,  8.28s/it, lr: 1.0e-04 loss: 7.955e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1468/30000 [3:19:41<65:39:42,  8.28s/it, lr: 1.0e-04 loss: 7.955e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1468/30000 [3:19:48<65:39:42,  8.28s/it, lr: 1.0e-04 loss: 8.667e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1468/30000 [3:19:48<65:39:42,  8.28s/it, lr: 1.0e-04 loss: 8.667e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1469/30000 [3:19:48<61:53:24,  7.81s/it, lr: 1.0e-04 loss: 8.667e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1469/30000 [3:19:48<61:53:24,  7.81s/it, lr: 1.0e-04 loss: 8.667e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1469/30000 [3:19:55<61:53:24,  7.81s/it, lr: 1.0e-04 loss: 6.894e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1469/30000 [3:19:55<61:53:24,  7.81s/it, lr: 1.0e-04 loss: 6.894e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1470/30000 [3:20:01<61:53:16,  7.81s/it, lr: 1.0e-04 loss: 8.055e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1470/30000 [3:20:01<61:53:16,  7.81s/it, lr: 1.0e-04 loss: 8.055e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1471/30000 [3:20:01<45:31:16,  5.74s/it, lr: 1.0e-04 loss: 8.055e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1471/30000 [3:20:01<45:31:16,  5.74s/it, lr: 1.0e-04 loss: 8.055e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1471/30000 [3:20:08<45:31:16,  5.74s/it, lr: 1.0e-04 loss: 8.628e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1471/30000 [3:20:08<45:31:16,  5.74s/it, lr: 1.0e-04 loss: 8.628e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1472/30000 [3:20:08<47:04:11,  5.94s/it, lr: 1.0e-04 loss: 8.628e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1472/30000 [3:20:08<47:04:11,  5.94s/it, lr: 1.0e-04 loss: 8.628e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1472/30000 [3:20:17<47:04:11,  5.94s/it, lr: 1.0e-04 loss: 7.605e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1472/30000 [3:20:17<47:04:11,  5.94s/it, lr: 1.0e-04 loss: 7.605e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1473/30000 [3:20:17<53:59:47,  6.81s/it, lr: 1.0e-04 loss: 7.605e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1473/30000 [3:20:17<53:59:47,  6.81s/it, lr: 1.0e-04 loss: 7.605e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1473/30000 [3:20:27<53:59:47,  6.81s/it, lr: 1.0e-04 loss: 6.823e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1473/30000 [3:20:27<53:59:47,  6.81s/it, lr: 1.0e-04 loss: 6.823e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1474/30000 [3:20:27<59:20:15,  7.49s/it, lr: 1.0e-04 loss: 6.823e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1474/30000 [3:20:27<59:20:15,  7.49s/it, lr: 1.0e-04 loss: 6.823e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1474/30000 [3:20:36<59:20:15,  7.49s/it, lr: 1.0e-04 loss: 2.470e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1474/30000 [3:20:36<59:20:15,  7.49s/it, lr: 1.0e-04 loss: 2.470e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1475/30000 [3:20:36<63:26:10,  8.01s/it, lr: 1.0e-04 loss: 2.470e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1475/30000 [3:20:36<63:26:10,  8.01s/it, lr: 1.0e-04 loss: 2.470e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1475/30000 [3:20:45<63:26:10,  8.01s/it, lr: 1.0e-04 loss: 8.464e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1475/30000 [3:20:45<63:26:10,  8.01s/it, lr: 1.0e-04 loss: 8.464e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1476/30000 [3:20:45<66:26:13,  8.38s/it, lr: 1.0e-04 loss: 8.464e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1476/30000 [3:20:45<66:26:13,  8.38s/it, lr: 1.0e-04 loss: 8.464e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1476/30000 [3:20:52<66:26:13,  8.38s/it, lr: 1.0e-04 loss: 7.714e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1476/30000 [3:20:52<66:26:13,  8.38s/it, lr: 1.0e-04 loss: 7.714e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1477/30000 [3:20:52<62:59:51,  7.95s/it, lr: 1.0e-04 loss: 7.714e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1477/30000 [3:20:52<62:59:51,  7.95s/it, lr: 1.0e-04 loss: 7.714e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1477/30000 [3:20:59<62:59:51,  7.95s/it, lr: 1.0e-04 loss: 8.854e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1477/30000 [3:20:59<62:59:51,  7.95s/it, lr: 1.0e-04 loss: 8.854e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1478/30000 [3:20:59<59:42:05,  7.54s/it, lr: 1.0e-04 loss: 8.854e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1478/30000 [3:20:59<59:42:05,  7.54s/it, lr: 1.0e-04 loss: 8.854e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1478/30000 [3:21:08<59:42:05,  7.54s/it, lr: 1.0e-04 loss: 4.505e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1478/30000 [3:21:08<59:42:05,  7.54s/it, lr: 1.0e-04 loss: 4.505e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1479/30000 [3:21:08<63:45:39,  8.05s/it, lr: 1.0e-04 loss: 4.505e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1479/30000 [3:21:08<63:45:39,  8.05s/it, lr: 1.0e-04 loss: 4.505e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1479/30000 [3:21:15<63:45:39,  8.05s/it, lr: 1.0e-04 loss: 9.868e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1479/30000 [3:21:15<63:45:39,  8.05s/it, lr: 1.0e-04 loss: 9.868e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1480/30000 [3:21:21<63:45:31,  8.05s/it, lr: 1.0e-04 loss: 6.114e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1480/30000 [3:21:21<63:45:31,  8.05s/it, lr: 1.0e-04 loss: 6.114e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1481/30000 [3:21:21<46:33:42,  5.88s/it, lr: 1.0e-04 loss: 6.114e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1481/30000 [3:21:21<46:33:42,  5.88s/it, lr: 1.0e-04 loss: 6.114e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1481/30000 [3:21:28<46:33:42,  5.88s/it, lr: 1.0e-04 loss: 9.708e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1481/30000 [3:21:28<46:33:42,  5.88s/it, lr: 1.0e-04 loss: 9.708e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1482/30000 [3:21:28<47:57:25,  6.05s/it, lr: 1.0e-04 loss: 9.708e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1482/30000 [3:21:28<47:57:25,  6.05s/it, lr: 1.0e-04 loss: 9.708e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1482/30000 [3:21:37<47:57:25,  6.05s/it, lr: 1.0e-04 loss: 7.913e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1482/30000 [3:21:37<47:57:25,  6.05s/it, lr: 1.0e-04 loss: 7.913e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1483/30000 [3:21:37<54:30:43,  6.88s/it, lr: 1.0e-04 loss: 7.913e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1483/30000 [3:21:37<54:30:43,  6.88s/it, lr: 1.0e-04 loss: 7.913e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1483/30000 [3:21:46<54:30:43,  6.88s/it, lr: 1.0e-04 loss: 7.030e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1483/30000 [3:21:46<54:30:43,  6.88s/it, lr: 1.0e-04 loss: 7.030e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1484/30000 [3:21:46<59:47:23,  7.55s/it, lr: 1.0e-04 loss: 7.030e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1484/30000 [3:21:46<59:47:23,  7.55s/it, lr: 1.0e-04 loss: 7.030e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1484/30000 [3:21:53<59:47:23,  7.55s/it, lr: 1.0e-04 loss: 5.515e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1484/30000 [3:21:53<59:47:23,  7.55s/it, lr: 1.0e-04 loss: 5.515e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1485/30000 [3:21:53<58:13:21,  7.35s/it, lr: 1.0e-04 loss: 5.515e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1485/30000 [3:21:53<58:13:21,  7.35s/it, lr: 1.0e-04 loss: 5.515e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1485/30000 [3:22:00<58:13:21,  7.35s/it, lr: 1.0e-04 loss: 1.339e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1485/30000 [3:22:00<58:13:21,  7.35s/it, lr: 1.0e-04 loss: 1.339e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1486/30000 [3:22:00<56:20:17,  7.11s/it, lr: 1.0e-04 loss: 1.339e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1486/30000 [3:22:00<56:20:17,  7.11s/it, lr: 1.0e-04 loss: 1.339e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1486/30000 [3:22:09<56:20:17,  7.11s/it, lr: 1.0e-04 loss: 2.633e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1486/30000 [3:22:09<56:20:17,  7.11s/it, lr: 1.0e-04 loss: 2.633e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1487/30000 [3:22:09<61:18:38,  7.74s/it, lr: 1.0e-04 loss: 2.633e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1487/30000 [3:22:09<61:18:38,  7.74s/it, lr: 1.0e-04 loss: 2.633e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1487/30000 [3:22:16<61:18:38,  7.74s/it, lr: 1.0e-04 loss: 9.720e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1487/30000 [3:22:16<61:18:38,  7.74s/it, lr: 1.0e-04 loss: 9.720e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1488/30000 [3:22:16<58:49:47,  7.43s/it, lr: 1.0e-04 loss: 9.720e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1488/30000 [3:22:16<58:49:47,  7.43s/it, lr: 1.0e-04 loss: 9.720e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1488/30000 [3:22:25<58:49:47,  7.43s/it, lr: 1.0e-04 loss: 4.621e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1488/30000 [3:22:25<58:49:47,  7.43s/it, lr: 1.0e-04 loss: 4.621e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1489/30000 [3:22:25<63:22:33,  8.00s/it, lr: 1.0e-04 loss: 4.621e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1489/30000 [3:22:25<63:22:33,  8.00s/it, lr: 1.0e-04 loss: 4.621e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1489/30000 [3:22:32<63:22:33,  8.00s/it, lr: 1.0e-04 loss: 1.199e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1489/30000 [3:22:32<63:22:33,  8.00s/it, lr: 1.0e-04 loss: 1.199e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1490/30000 [3:22:41<63:22:25,  8.00s/it, lr: 1.0e-04 loss: 3.806e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1490/30000 [3:22:41<63:22:25,  8.00s/it, lr: 1.0e-04 loss: 3.806e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1491/30000 [3:22:41<51:17:34,  6.48s/it, lr: 1.0e-04 loss: 3.806e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1491/30000 [3:22:41<51:17:34,  6.48s/it, lr: 1.0e-04 loss: 3.806e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1491/30000 [3:22:51<51:17:34,  6.48s/it, lr: 1.0e-04 loss: 9.170e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1491/30000 [3:22:51<51:17:34,  6.48s/it, lr: 1.0e-04 loss: 9.170e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1492/30000 [3:22:51<57:25:14,  7.25s/it, lr: 1.0e-04 loss: 9.170e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1492/30000 [3:22:51<57:25:14,  7.25s/it, lr: 1.0e-04 loss: 9.170e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1492/30000 [3:22:57<57:25:14,  7.25s/it, lr: 1.0e-04 loss: 2.358e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1492/30000 [3:22:57<57:25:14,  7.25s/it, lr: 1.0e-04 loss: 2.358e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1493/30000 [3:22:57<55:59:10,  7.07s/it, lr: 1.0e-04 loss: 2.358e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1493/30000 [3:22:57<55:59:10,  7.07s/it, lr: 1.0e-04 loss: 2.358e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1493/30000 [3:23:04<55:59:10,  7.07s/it, lr: 1.0e-04 loss: 1.160e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1493/30000 [3:23:04<55:59:10,  7.07s/it, lr: 1.0e-04 loss: 1.160e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1494/30000 [3:23:04<55:01:55,  6.95s/it, lr: 1.0e-04 loss: 1.160e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1494/30000 [3:23:04<55:01:55,  6.95s/it, lr: 1.0e-04 loss: 1.160e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1494/30000 [3:23:10<55:01:55,  6.95s/it, lr: 1.0e-04 loss: 1.265e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1494/30000 [3:23:10<55:01:55,  6.95s/it, lr: 1.0e-04 loss: 1.265e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1495/30000 [3:23:10<54:09:43,  6.84s/it, lr: 1.0e-04 loss: 1.265e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1495/30000 [3:23:10<54:09:43,  6.84s/it, lr: 1.0e-04 loss: 1.265e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1495/30000 [3:23:17<54:09:43,  6.84s/it, lr: 1.0e-04 loss: 7.148e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1495/30000 [3:23:17<54:09:43,  6.84s/it, lr: 1.0e-04 loss: 7.148e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1496/30000 [3:23:17<53:39:44,  6.78s/it, lr: 1.0e-04 loss: 7.148e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1496/30000 [3:23:17<53:39:44,  6.78s/it, lr: 1.0e-04 loss: 7.148e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1496/30000 [3:23:24<53:39:44,  6.78s/it, lr: 1.0e-04 loss: 1.153e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1496/30000 [3:23:24<53:39:44,  6.78s/it, lr: 1.0e-04 loss: 1.153e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1497/30000 [3:23:24<53:05:36,  6.71s/it, lr: 1.0e-04 loss: 1.153e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1497/30000 [3:23:24<53:05:36,  6.71s/it, lr: 1.0e-04 loss: 1.153e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1497/30000 [3:23:33<53:05:36,  6.71s/it, lr: 1.0e-04 loss: 7.865e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1497/30000 [3:23:33<53:05:36,  6.71s/it, lr: 1.0e-04 loss: 7.865e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1498/30000 [3:23:33<59:12:09,  7.48s/it, lr: 1.0e-04 loss: 7.865e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1498/30000 [3:23:33<59:12:09,  7.48s/it, lr: 1.0e-04 loss: 7.865e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1498/30000 [3:23:40<59:12:09,  7.48s/it, lr: 1.0e-04 loss: 5.020e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1498/30000 [3:23:40<59:12:09,  7.48s/it, lr: 1.0e-04 loss: 5.020e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1499/30000 [3:23:40<57:09:21,  7.22s/it, lr: 1.0e-04 loss: 5.020e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1499/30000 [3:23:40<57:09:21,  7.22s/it, lr: 1.0e-04 loss: 5.020e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1499/30000 [3:23:49<57:09:21,  7.22s/it, lr: 1.0e-04 loss: 5.513e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1499/30000 [3:23:49<57:09:21,  7.22s/it, lr: 1.0e-04 loss: 5.513e-02]
+Saving at step 1500
+Saved checkpoint to /home/featurize/ai-toolkit/output/Xiang_wan_2_2_lora/Xiang_wan_2_2_lora_000001500.safetensors
+Saved optimizer to /home/featurize/ai-toolkit/output/Xiang_wan_2_2_lora/optimizer.pt
+
Xiang_wan_2_2_lora:   5%|#9                                     | 1500/30000 [3:23:59<57:09:13,  7.22s/it, lr: 1.0e-04 loss: 3.730e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1500/30000 [3:23:59<57:09:13,  7.22s/it, lr: 1.0e-04 loss: 3.730e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1501/30000 [3:23:59<48:29:16,  6.13s/it, lr: 1.0e-04 loss: 3.730e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1501/30000 [3:23:59<48:29:16,  6.13s/it, lr: 1.0e-04 loss: 3.730e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1501/30000 [3:24:05<48:29:16,  6.13s/it, lr: 1.0e-04 loss: 7.242e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1501/30000 [3:24:05<48:29:16,  6.13s/it, lr: 1.0e-04 loss: 7.242e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1502/30000 [3:24:05<49:20:22,  6.23s/it, lr: 1.0e-04 loss: 7.242e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1502/30000 [3:24:05<49:20:22,  6.23s/it, lr: 1.0e-04 loss: 7.242e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1502/30000 [3:24:12<49:20:22,  6.23s/it, lr: 1.0e-04 loss: 9.763e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1502/30000 [3:24:12<49:20:22,  6.23s/it, lr: 1.0e-04 loss: 9.763e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1503/30000 [3:24:12<49:53:39,  6.30s/it, lr: 1.0e-04 loss: 9.763e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1503/30000 [3:24:12<49:53:39,  6.30s/it, lr: 1.0e-04 loss: 9.763e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1503/30000 [3:24:21<49:53:39,  6.30s/it, lr: 1.0e-04 loss: 6.855e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1503/30000 [3:24:21<49:53:39,  6.30s/it, lr: 1.0e-04 loss: 6.855e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1504/30000 [3:24:21<56:22:16,  7.12s/it, lr: 1.0e-04 loss: 6.855e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1504/30000 [3:24:21<56:22:16,  7.12s/it, lr: 1.0e-04 loss: 6.855e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1504/30000 [3:24:31<56:22:16,  7.12s/it, lr: 1.0e-04 loss: 3.437e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1504/30000 [3:24:31<56:22:16,  7.12s/it, lr: 1.0e-04 loss: 3.437e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1505/30000 [3:24:31<61:20:26,  7.75s/it, lr: 1.0e-04 loss: 3.437e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1505/30000 [3:24:31<61:20:26,  7.75s/it, lr: 1.0e-04 loss: 3.437e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1505/30000 [3:24:40<61:20:26,  7.75s/it, lr: 1.0e-04 loss: 6.494e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1505/30000 [3:24:40<61:20:26,  7.75s/it, lr: 1.0e-04 loss: 6.494e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1506/30000 [3:24:40<64:57:02,  8.21s/it, lr: 1.0e-04 loss: 6.494e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1506/30000 [3:24:40<64:57:02,  8.21s/it, lr: 1.0e-04 loss: 6.494e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1506/30000 [3:24:49<64:57:02,  8.21s/it, lr: 1.0e-04 loss: 8.450e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1506/30000 [3:24:49<64:57:02,  8.21s/it, lr: 1.0e-04 loss: 8.450e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1507/30000 [3:24:49<67:38:35,  8.55s/it, lr: 1.0e-04 loss: 8.450e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1507/30000 [3:24:49<67:38:35,  8.55s/it, lr: 1.0e-04 loss: 8.450e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1507/30000 [3:24:56<67:38:35,  8.55s/it, lr: 1.0e-04 loss: 7.129e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1507/30000 [3:24:56<67:38:35,  8.55s/it, lr: 1.0e-04 loss: 7.129e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1508/30000 [3:24:56<63:50:55,  8.07s/it, lr: 1.0e-04 loss: 7.129e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1508/30000 [3:24:56<63:50:55,  8.07s/it, lr: 1.0e-04 loss: 7.129e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1508/30000 [3:25:06<63:50:55,  8.07s/it, lr: 1.0e-04 loss: 6.081e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1508/30000 [3:25:06<63:50:55,  8.07s/it, lr: 1.0e-04 loss: 6.081e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1509/30000 [3:25:06<67:02:14,  8.47s/it, lr: 1.0e-04 loss: 6.081e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1509/30000 [3:25:06<67:02:14,  8.47s/it, lr: 1.0e-04 loss: 6.081e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1509/30000 [3:25:15<67:02:14,  8.47s/it, lr: 1.0e-04 loss: 6.734e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1509/30000 [3:25:15<67:02:14,  8.47s/it, lr: 1.0e-04 loss: 6.734e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1510/30000 [3:25:24<67:02:05,  8.47s/it, lr: 1.0e-04 loss: 5.666e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1510/30000 [3:25:24<67:02:05,  8.47s/it, lr: 1.0e-04 loss: 5.666e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1511/30000 [3:25:24<53:17:51,  6.73s/it, lr: 1.0e-04 loss: 5.666e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1511/30000 [3:25:24<53:17:51,  6.73s/it, lr: 1.0e-04 loss: 5.666e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1511/30000 [3:25:34<53:17:51,  6.73s/it, lr: 1.0e-04 loss: 5.794e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1511/30000 [3:25:34<53:17:51,  6.73s/it, lr: 1.0e-04 loss: 5.794e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1512/30000 [3:25:34<58:21:53,  7.38s/it, lr: 1.0e-04 loss: 5.794e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1512/30000 [3:25:34<58:21:53,  7.38s/it, lr: 1.0e-04 loss: 5.794e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1512/30000 [3:25:43<58:21:53,  7.38s/it, lr: 1.0e-04 loss: 5.404e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1512/30000 [3:25:43<58:21:53,  7.38s/it, lr: 1.0e-04 loss: 5.404e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1513/30000 [3:25:43<62:27:56,  7.89s/it, lr: 1.0e-04 loss: 5.404e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1513/30000 [3:25:43<62:27:56,  7.89s/it, lr: 1.0e-04 loss: 5.404e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1513/30000 [3:25:50<62:27:56,  7.89s/it, lr: 1.0e-04 loss: 1.094e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1513/30000 [3:25:50<62:27:56,  7.89s/it, lr: 1.0e-04 loss: 1.094e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1514/30000 [3:25:50<59:37:29,  7.54s/it, lr: 1.0e-04 loss: 1.094e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1514/30000 [3:25:50<59:37:29,  7.54s/it, lr: 1.0e-04 loss: 1.094e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1514/30000 [3:25:56<59:37:29,  7.54s/it, lr: 1.0e-04 loss: 6.174e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1514/30000 [3:25:56<59:37:29,  7.54s/it, lr: 1.0e-04 loss: 6.174e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1515/30000 [3:25:56<58:00:43,  7.33s/it, lr: 1.0e-04 loss: 6.174e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1515/30000 [3:25:56<58:00:43,  7.33s/it, lr: 1.0e-04 loss: 6.174e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1515/30000 [3:26:06<58:00:43,  7.33s/it, lr: 1.0e-04 loss: 8.362e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1515/30000 [3:26:06<58:00:43,  7.33s/it, lr: 1.0e-04 loss: 8.362e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1516/30000 [3:26:06<62:31:04,  7.90s/it, lr: 1.0e-04 loss: 8.362e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1516/30000 [3:26:06<62:31:04,  7.90s/it, lr: 1.0e-04 loss: 8.362e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1516/30000 [3:26:12<62:31:04,  7.90s/it, lr: 1.0e-04 loss: 7.374e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1516/30000 [3:26:12<62:31:04,  7.90s/it, lr: 1.0e-04 loss: 7.374e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1517/30000 [3:26:12<59:32:03,  7.52s/it, lr: 1.0e-04 loss: 7.374e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1517/30000 [3:26:12<59:32:03,  7.52s/it, lr: 1.0e-04 loss: 7.374e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1517/30000 [3:26:22<59:32:03,  7.52s/it, lr: 1.0e-04 loss: 1.138e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1517/30000 [3:26:22<59:32:03,  7.52s/it, lr: 1.0e-04 loss: 1.138e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1518/30000 [3:26:22<63:32:34,  8.03s/it, lr: 1.0e-04 loss: 1.138e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1518/30000 [3:26:22<63:32:34,  8.03s/it, lr: 1.0e-04 loss: 1.138e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1518/30000 [3:26:31<63:32:34,  8.03s/it, lr: 1.0e-04 loss: 9.930e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1518/30000 [3:26:31<63:32:34,  8.03s/it, lr: 1.0e-04 loss: 9.930e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1519/30000 [3:26:31<66:41:28,  8.43s/it, lr: 1.0e-04 loss: 9.930e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1519/30000 [3:26:31<66:41:28,  8.43s/it, lr: 1.0e-04 loss: 9.930e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1519/30000 [3:26:38<66:41:28,  8.43s/it, lr: 1.0e-04 loss: 1.332e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1519/30000 [3:26:38<66:41:28,  8.43s/it, lr: 1.0e-04 loss: 1.332e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1520/30000 [3:26:47<66:41:20,  8.43s/it, lr: 1.0e-04 loss: 1.113e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1520/30000 [3:26:47<66:41:20,  8.43s/it, lr: 1.0e-04 loss: 1.113e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1521/30000 [3:26:47<53:06:23,  6.71s/it, lr: 1.0e-04 loss: 1.113e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1521/30000 [3:26:47<53:06:23,  6.71s/it, lr: 1.0e-04 loss: 1.113e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1521/30000 [3:26:54<53:06:23,  6.71s/it, lr: 1.0e-04 loss: 1.465e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1521/30000 [3:26:54<53:06:23,  6.71s/it, lr: 1.0e-04 loss: 1.465e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1522/30000 [3:26:54<53:27:37,  6.76s/it, lr: 1.0e-04 loss: 1.465e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1522/30000 [3:26:54<53:27:37,  6.76s/it, lr: 1.0e-04 loss: 1.465e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1522/30000 [3:27:00<53:27:37,  6.76s/it, lr: 1.0e-04 loss: 1.835e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1522/30000 [3:27:00<53:27:37,  6.76s/it, lr: 1.0e-04 loss: 1.835e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1523/30000 [3:27:00<53:03:33,  6.71s/it, lr: 1.0e-04 loss: 1.835e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1523/30000 [3:27:00<53:03:33,  6.71s/it, lr: 1.0e-04 loss: 1.835e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1523/30000 [3:27:07<53:03:33,  6.71s/it, lr: 1.0e-04 loss: 9.577e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1523/30000 [3:27:07<53:03:33,  6.71s/it, lr: 1.0e-04 loss: 9.577e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1524/30000 [3:27:07<52:50:10,  6.68s/it, lr: 1.0e-04 loss: 9.577e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1524/30000 [3:27:07<52:50:10,  6.68s/it, lr: 1.0e-04 loss: 9.577e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1524/30000 [3:27:14<52:50:10,  6.68s/it, lr: 1.0e-04 loss: 6.077e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1524/30000 [3:27:14<52:50:10,  6.68s/it, lr: 1.0e-04 loss: 6.077e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1525/30000 [3:27:14<52:37:54,  6.65s/it, lr: 1.0e-04 loss: 6.077e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1525/30000 [3:27:14<52:37:54,  6.65s/it, lr: 1.0e-04 loss: 6.077e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1525/30000 [3:27:23<52:37:54,  6.65s/it, lr: 1.0e-04 loss: 7.046e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1525/30000 [3:27:23<52:37:54,  6.65s/it, lr: 1.0e-04 loss: 7.046e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1526/30000 [3:27:23<58:36:04,  7.41s/it, lr: 1.0e-04 loss: 7.046e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1526/30000 [3:27:23<58:36:04,  7.41s/it, lr: 1.0e-04 loss: 7.046e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1526/30000 [3:27:32<58:36:04,  7.41s/it, lr: 1.0e-04 loss: 5.268e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1526/30000 [3:27:32<58:36:04,  7.41s/it, lr: 1.0e-04 loss: 5.268e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1527/30000 [3:27:32<63:07:08,  7.98s/it, lr: 1.0e-04 loss: 5.268e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1527/30000 [3:27:32<63:07:08,  7.98s/it, lr: 1.0e-04 loss: 5.268e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1527/30000 [3:27:42<63:07:08,  7.98s/it, lr: 1.0e-04 loss: 8.948e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1527/30000 [3:27:42<63:07:08,  7.98s/it, lr: 1.0e-04 loss: 8.948e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1528/30000 [3:27:42<66:23:20,  8.39s/it, lr: 1.0e-04 loss: 8.948e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1528/30000 [3:27:42<66:23:20,  8.39s/it, lr: 1.0e-04 loss: 8.948e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1528/30000 [3:27:48<66:23:20,  8.39s/it, lr: 1.0e-04 loss: 7.501e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1528/30000 [3:27:48<66:23:20,  8.39s/it, lr: 1.0e-04 loss: 7.501e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1529/30000 [3:27:48<62:15:36,  7.87s/it, lr: 1.0e-04 loss: 7.501e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1529/30000 [3:27:48<62:15:36,  7.87s/it, lr: 1.0e-04 loss: 7.501e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1529/30000 [3:27:55<62:15:36,  7.87s/it, lr: 1.0e-04 loss: 1.122e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1529/30000 [3:27:55<62:15:36,  7.87s/it, lr: 1.0e-04 loss: 1.122e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1530/30000 [3:28:02<62:15:28,  7.87s/it, lr: 1.0e-04 loss: 9.865e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1530/30000 [3:28:02<62:15:28,  7.87s/it, lr: 1.0e-04 loss: 9.865e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1531/30000 [3:28:02<45:37:49,  5.77s/it, lr: 1.0e-04 loss: 9.865e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1531/30000 [3:28:02<45:37:49,  5.77s/it, lr: 1.0e-04 loss: 9.865e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1531/30000 [3:28:08<45:37:49,  5.77s/it, lr: 1.0e-04 loss: 7.954e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1531/30000 [3:28:08<45:37:49,  5.77s/it, lr: 1.0e-04 loss: 7.954e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1532/30000 [3:28:08<47:10:03,  5.96s/it, lr: 1.0e-04 loss: 7.954e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1532/30000 [3:28:08<47:10:03,  5.96s/it, lr: 1.0e-04 loss: 7.954e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1532/30000 [3:28:15<47:10:03,  5.96s/it, lr: 1.0e-04 loss: 7.494e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1532/30000 [3:28:15<47:10:03,  5.96s/it, lr: 1.0e-04 loss: 7.494e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1533/30000 [3:28:15<48:14:22,  6.10s/it, lr: 1.0e-04 loss: 7.494e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1533/30000 [3:28:15<48:14:22,  6.10s/it, lr: 1.0e-04 loss: 7.494e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1533/30000 [3:28:24<48:14:22,  6.10s/it, lr: 1.0e-04 loss: 8.934e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1533/30000 [3:28:24<48:14:22,  6.10s/it, lr: 1.0e-04 loss: 8.934e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1534/30000 [3:28:24<55:22:56,  7.00s/it, lr: 1.0e-04 loss: 8.934e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1534/30000 [3:28:24<55:22:56,  7.00s/it, lr: 1.0e-04 loss: 8.934e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1534/30000 [3:28:31<55:22:56,  7.00s/it, lr: 1.0e-04 loss: 6.928e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1534/30000 [3:28:31<55:22:56,  7.00s/it, lr: 1.0e-04 loss: 6.928e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1535/30000 [3:28:31<54:23:12,  6.88s/it, lr: 1.0e-04 loss: 6.928e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1535/30000 [3:28:31<54:23:12,  6.88s/it, lr: 1.0e-04 loss: 6.928e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1535/30000 [3:28:37<54:23:12,  6.88s/it, lr: 1.0e-04 loss: 1.354e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1535/30000 [3:28:37<54:23:12,  6.88s/it, lr: 1.0e-04 loss: 1.354e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1536/30000 [3:28:37<53:44:23,  6.80s/it, lr: 1.0e-04 loss: 1.354e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1536/30000 [3:28:37<53:44:23,  6.80s/it, lr: 1.0e-04 loss: 1.354e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1536/30000 [3:28:47<53:44:23,  6.80s/it, lr: 1.0e-04 loss: 5.589e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1536/30000 [3:28:47<53:44:23,  6.80s/it, lr: 1.0e-04 loss: 5.589e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1537/30000 [3:28:47<59:55:15,  7.58s/it, lr: 1.0e-04 loss: 5.589e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1537/30000 [3:28:47<59:55:15,  7.58s/it, lr: 1.0e-04 loss: 5.589e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1537/30000 [3:28:56<59:55:15,  7.58s/it, lr: 1.0e-04 loss: 8.910e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1537/30000 [3:28:56<59:55:15,  7.58s/it, lr: 1.0e-04 loss: 8.910e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1538/30000 [3:28:56<64:03:33,  8.10s/it, lr: 1.0e-04 loss: 8.910e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1538/30000 [3:28:56<64:03:33,  8.10s/it, lr: 1.0e-04 loss: 8.910e-02]
Xiang_wan_2_2_lora:   5%|#9                                     | 1538/30000 [3:29:03<64:03:33,  8.10s/it, lr: 1.0e-04 loss: 2.531e-01]
Xiang_wan_2_2_lora:   5%|#9                                     | 1538/30000 [3:29:03<64:03:33,  8.10s/it, lr: 1.0e-04 loss: 2.531e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1539/30000 [3:29:03<60:26:02,  7.64s/it, lr: 1.0e-04 loss: 2.531e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1539/30000 [3:29:03<60:26:02,  7.64s/it, lr: 1.0e-04 loss: 2.531e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1539/30000 [3:29:09<60:26:02,  7.64s/it, lr: 1.0e-04 loss: 9.072e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1539/30000 [3:29:09<60:26:02,  7.64s/it, lr: 1.0e-04 loss: 9.072e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1540/30000 [3:29:19<60:25:54,  7.64s/it, lr: 1.0e-04 loss: 4.370e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1540/30000 [3:29:19<60:25:54,  7.64s/it, lr: 1.0e-04 loss: 4.370e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1541/30000 [3:29:19<49:33:52,  6.27s/it, lr: 1.0e-04 loss: 4.370e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1541/30000 [3:29:19<49:33:52,  6.27s/it, lr: 1.0e-04 loss: 4.370e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1541/30000 [3:29:25<49:33:52,  6.27s/it, lr: 1.0e-04 loss: 9.604e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1541/30000 [3:29:25<49:33:52,  6.27s/it, lr: 1.0e-04 loss: 9.604e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1542/30000 [3:29:25<50:06:47,  6.34s/it, lr: 1.0e-04 loss: 9.604e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1542/30000 [3:29:25<50:06:47,  6.34s/it, lr: 1.0e-04 loss: 9.604e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1542/30000 [3:29:34<50:06:47,  6.34s/it, lr: 1.0e-04 loss: 4.523e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1542/30000 [3:29:34<50:06:47,  6.34s/it, lr: 1.0e-04 loss: 4.523e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1543/30000 [3:29:34<56:12:58,  7.11s/it, lr: 1.0e-04 loss: 4.523e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1543/30000 [3:29:34<56:12:58,  7.11s/it, lr: 1.0e-04 loss: 4.523e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1543/30000 [3:29:49<56:12:58,  7.11s/it, lr: 1.0e-04 loss: 8.242e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1543/30000 [3:29:49<56:12:58,  7.11s/it, lr: 1.0e-04 loss: 8.242e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1544/30000 [3:29:49<69:33:27,  8.80s/it, lr: 1.0e-04 loss: 8.242e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1544/30000 [3:29:49<69:33:27,  8.80s/it, lr: 1.0e-04 loss: 8.242e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1544/30000 [3:30:02<69:33:27,  8.80s/it, lr: 1.0e-04 loss: 7.745e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1544/30000 [3:30:02<69:33:27,  8.80s/it, lr: 1.0e-04 loss: 7.745e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1545/30000 [3:30:02<77:29:51,  9.80s/it, lr: 1.0e-04 loss: 7.745e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1545/30000 [3:30:02<77:29:51,  9.80s/it, lr: 1.0e-04 loss: 7.745e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1545/30000 [3:30:09<77:29:51,  9.80s/it, lr: 1.0e-04 loss: 1.514e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1545/30000 [3:30:09<77:29:51,  9.80s/it, lr: 1.0e-04 loss: 1.514e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1546/30000 [3:30:09<72:03:47,  9.12s/it, lr: 1.0e-04 loss: 1.514e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1546/30000 [3:30:09<72:03:47,  9.12s/it, lr: 1.0e-04 loss: 1.514e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1546/30000 [3:30:19<72:03:47,  9.12s/it, lr: 1.0e-04 loss: 2.611e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1546/30000 [3:30:19<72:03:47,  9.12s/it, lr: 1.0e-04 loss: 2.611e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1547/30000 [3:30:19<74:10:21,  9.38s/it, lr: 1.0e-04 loss: 2.611e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1547/30000 [3:30:19<74:10:21,  9.38s/it, lr: 1.0e-04 loss: 2.611e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1547/30000 [3:30:29<74:10:21,  9.38s/it, lr: 1.0e-04 loss: 7.994e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1547/30000 [3:30:29<74:10:21,  9.38s/it, lr: 1.0e-04 loss: 7.994e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1548/30000 [3:30:29<74:39:50,  9.45s/it, lr: 1.0e-04 loss: 7.994e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1548/30000 [3:30:29<74:39:50,  9.45s/it, lr: 1.0e-04 loss: 7.994e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1548/30000 [3:30:35<74:39:50,  9.45s/it, lr: 1.0e-04 loss: 6.229e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1548/30000 [3:30:35<74:39:50,  9.45s/it, lr: 1.0e-04 loss: 6.229e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1549/30000 [3:30:35<68:31:10,  8.67s/it, lr: 1.0e-04 loss: 6.229e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1549/30000 [3:30:35<68:31:10,  8.67s/it, lr: 1.0e-04 loss: 6.229e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1549/30000 [3:30:45<68:31:10,  8.67s/it, lr: 1.0e-04 loss: 7.122e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1549/30000 [3:30:45<68:31:10,  8.67s/it, lr: 1.0e-04 loss: 7.122e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1550/30000 [3:30:52<68:31:01,  8.67s/it, lr: 1.0e-04 loss: 1.771e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1550/30000 [3:30:52<68:31:01,  8.67s/it, lr: 1.0e-04 loss: 1.771e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1551/30000 [3:30:52<49:13:10,  6.23s/it, lr: 1.0e-04 loss: 1.771e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1551/30000 [3:30:52<49:13:10,  6.23s/it, lr: 1.0e-04 loss: 1.771e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1551/30000 [3:30:58<49:13:10,  6.23s/it, lr: 1.0e-04 loss: 6.302e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1551/30000 [3:30:58<49:13:10,  6.23s/it, lr: 1.0e-04 loss: 6.302e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1552/30000 [3:30:58<50:23:25,  6.38s/it, lr: 1.0e-04 loss: 6.302e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1552/30000 [3:30:58<50:23:25,  6.38s/it, lr: 1.0e-04 loss: 6.302e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1552/30000 [3:31:08<50:23:25,  6.38s/it, lr: 1.0e-04 loss: 3.710e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1552/30000 [3:31:08<50:23:25,  6.38s/it, lr: 1.0e-04 loss: 3.710e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1553/30000 [3:31:08<56:31:18,  7.15s/it, lr: 1.0e-04 loss: 3.710e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1553/30000 [3:31:08<56:31:18,  7.15s/it, lr: 1.0e-04 loss: 3.710e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1553/30000 [3:31:17<56:31:18,  7.15s/it, lr: 1.0e-04 loss: 7.563e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1553/30000 [3:31:17<56:31:18,  7.15s/it, lr: 1.0e-04 loss: 7.563e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1554/30000 [3:31:17<61:11:55,  7.75s/it, lr: 1.0e-04 loss: 7.563e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1554/30000 [3:31:17<61:11:55,  7.75s/it, lr: 1.0e-04 loss: 7.563e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1554/30000 [3:31:24<61:11:55,  7.75s/it, lr: 1.0e-04 loss: 5.949e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1554/30000 [3:31:24<61:11:55,  7.75s/it, lr: 1.0e-04 loss: 5.949e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1555/30000 [3:31:24<58:40:08,  7.43s/it, lr: 1.0e-04 loss: 5.949e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1555/30000 [3:31:24<58:40:08,  7.43s/it, lr: 1.0e-04 loss: 5.949e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1555/30000 [3:31:30<58:40:08,  7.43s/it, lr: 1.0e-04 loss: 1.007e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1555/30000 [3:31:30<58:40:08,  7.43s/it, lr: 1.0e-04 loss: 1.007e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1556/30000 [3:31:30<56:45:24,  7.18s/it, lr: 1.0e-04 loss: 1.007e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1556/30000 [3:31:30<56:45:24,  7.18s/it, lr: 1.0e-04 loss: 1.007e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1556/30000 [3:31:40<56:45:24,  7.18s/it, lr: 1.0e-04 loss: 4.802e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1556/30000 [3:31:40<56:45:24,  7.18s/it, lr: 1.0e-04 loss: 4.802e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1557/30000 [3:31:40<61:28:26,  7.78s/it, lr: 1.0e-04 loss: 4.802e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1557/30000 [3:31:40<61:28:26,  7.78s/it, lr: 1.0e-04 loss: 4.802e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1557/30000 [3:31:49<61:28:26,  7.78s/it, lr: 1.0e-04 loss: 7.224e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1557/30000 [3:31:49<61:28:26,  7.78s/it, lr: 1.0e-04 loss: 7.224e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1558/30000 [3:31:49<65:05:30,  8.24s/it, lr: 1.0e-04 loss: 7.224e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1558/30000 [3:31:49<65:05:30,  8.24s/it, lr: 1.0e-04 loss: 7.224e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1558/30000 [3:31:58<65:05:30,  8.24s/it, lr: 1.0e-04 loss: 4.682e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1558/30000 [3:31:58<65:05:30,  8.24s/it, lr: 1.0e-04 loss: 4.682e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1559/30000 [3:31:58<68:12:12,  8.63s/it, lr: 1.0e-04 loss: 4.682e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1559/30000 [3:31:58<68:12:12,  8.63s/it, lr: 1.0e-04 loss: 4.682e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1559/30000 [3:32:08<68:12:12,  8.63s/it, lr: 1.0e-04 loss: 6.533e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1559/30000 [3:32:08<68:12:12,  8.63s/it, lr: 1.0e-04 loss: 6.533e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1560/30000 [3:32:14<68:12:03,  8.63s/it, lr: 1.0e-04 loss: 6.148e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1560/30000 [3:32:14<68:12:03,  8.63s/it, lr: 1.0e-04 loss: 6.148e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1561/30000 [3:32:14<48:57:28,  6.20s/it, lr: 1.0e-04 loss: 6.148e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1561/30000 [3:32:14<48:57:28,  6.20s/it, lr: 1.0e-04 loss: 6.148e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1561/30000 [3:32:21<48:57:28,  6.20s/it, lr: 1.0e-04 loss: 1.087e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1561/30000 [3:32:21<48:57:28,  6.20s/it, lr: 1.0e-04 loss: 1.087e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1562/30000 [3:32:21<49:39:24,  6.29s/it, lr: 1.0e-04 loss: 1.087e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1562/30000 [3:32:21<49:39:24,  6.29s/it, lr: 1.0e-04 loss: 1.087e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1562/30000 [3:32:30<49:39:24,  6.29s/it, lr: 1.0e-04 loss: 7.219e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1562/30000 [3:32:30<49:39:24,  6.29s/it, lr: 1.0e-04 loss: 7.219e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1563/30000 [3:32:30<55:47:57,  7.06s/it, lr: 1.0e-04 loss: 7.219e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1563/30000 [3:32:30<55:47:57,  7.06s/it, lr: 1.0e-04 loss: 7.219e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1563/30000 [3:32:40<55:47:57,  7.06s/it, lr: 1.0e-04 loss: 5.512e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1563/30000 [3:32:40<55:47:57,  7.06s/it, lr: 1.0e-04 loss: 5.512e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1564/30000 [3:32:40<60:41:09,  7.68s/it, lr: 1.0e-04 loss: 5.512e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1564/30000 [3:32:40<60:41:09,  7.68s/it, lr: 1.0e-04 loss: 5.512e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1564/30000 [3:32:49<60:41:09,  7.68s/it, lr: 1.0e-04 loss: 6.891e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1564/30000 [3:32:49<60:41:09,  7.68s/it, lr: 1.0e-04 loss: 6.891e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1565/30000 [3:32:49<64:12:59,  8.13s/it, lr: 1.0e-04 loss: 6.891e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1565/30000 [3:32:49<64:12:59,  8.13s/it, lr: 1.0e-04 loss: 6.891e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1565/30000 [3:32:56<64:12:59,  8.13s/it, lr: 1.0e-04 loss: 1.093e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1565/30000 [3:32:56<64:12:59,  8.13s/it, lr: 1.0e-04 loss: 1.093e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1566/30000 [3:32:56<60:53:09,  7.71s/it, lr: 1.0e-04 loss: 1.093e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1566/30000 [3:32:56<60:53:09,  7.71s/it, lr: 1.0e-04 loss: 1.093e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1566/30000 [3:33:02<60:53:09,  7.71s/it, lr: 1.0e-04 loss: 9.862e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1566/30000 [3:33:02<60:53:09,  7.71s/it, lr: 1.0e-04 loss: 9.862e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1567/30000 [3:33:02<58:47:01,  7.44s/it, lr: 1.0e-04 loss: 9.862e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1567/30000 [3:33:02<58:47:01,  7.44s/it, lr: 1.0e-04 loss: 9.862e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1567/30000 [3:33:12<58:47:01,  7.44s/it, lr: 1.0e-04 loss: 6.687e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1567/30000 [3:33:12<58:47:01,  7.44s/it, lr: 1.0e-04 loss: 6.687e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1568/30000 [3:33:12<63:05:37,  7.99s/it, lr: 1.0e-04 loss: 6.687e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1568/30000 [3:33:12<63:05:37,  7.99s/it, lr: 1.0e-04 loss: 6.687e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1568/30000 [3:33:18<63:05:37,  7.99s/it, lr: 1.0e-04 loss: 6.829e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1568/30000 [3:33:18<63:05:37,  7.99s/it, lr: 1.0e-04 loss: 6.829e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1569/30000 [3:33:18<59:48:34,  7.57s/it, lr: 1.0e-04 loss: 6.829e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1569/30000 [3:33:18<59:48:34,  7.57s/it, lr: 1.0e-04 loss: 6.829e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1569/30000 [3:33:27<59:48:34,  7.57s/it, lr: 1.0e-04 loss: 7.334e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1569/30000 [3:33:27<59:48:34,  7.57s/it, lr: 1.0e-04 loss: 7.334e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1570/30000 [3:33:37<59:48:27,  7.57s/it, lr: 1.0e-04 loss: 8.259e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1570/30000 [3:33:37<59:48:27,  7.57s/it, lr: 1.0e-04 loss: 8.259e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1571/30000 [3:33:37<49:20:13,  6.25s/it, lr: 1.0e-04 loss: 8.259e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1571/30000 [3:33:37<49:20:13,  6.25s/it, lr: 1.0e-04 loss: 8.259e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1571/30000 [3:33:43<49:20:13,  6.25s/it, lr: 1.0e-04 loss: 7.638e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1571/30000 [3:33:43<49:20:13,  6.25s/it, lr: 1.0e-04 loss: 7.638e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1572/30000 [3:33:43<49:53:26,  6.32s/it, lr: 1.0e-04 loss: 7.638e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1572/30000 [3:33:43<49:53:26,  6.32s/it, lr: 1.0e-04 loss: 7.638e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1572/30000 [3:33:50<49:53:26,  6.32s/it, lr: 1.0e-04 loss: 4.807e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1572/30000 [3:33:50<49:53:26,  6.32s/it, lr: 1.0e-04 loss: 4.807e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1573/30000 [3:33:50<50:23:42,  6.38s/it, lr: 1.0e-04 loss: 4.807e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1573/30000 [3:33:50<50:23:42,  6.38s/it, lr: 1.0e-04 loss: 4.807e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1573/30000 [3:33:57<50:23:42,  6.38s/it, lr: 1.0e-04 loss: 6.911e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1573/30000 [3:33:57<50:23:42,  6.38s/it, lr: 1.0e-04 loss: 6.911e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1574/30000 [3:33:57<51:18:09,  6.50s/it, lr: 1.0e-04 loss: 6.911e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1574/30000 [3:33:57<51:18:09,  6.50s/it, lr: 1.0e-04 loss: 6.911e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1574/30000 [3:34:06<51:18:09,  6.50s/it, lr: 1.0e-04 loss: 3.890e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1574/30000 [3:34:06<51:18:09,  6.50s/it, lr: 1.0e-04 loss: 3.890e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1575/30000 [3:34:06<57:27:05,  7.28s/it, lr: 1.0e-04 loss: 3.890e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1575/30000 [3:34:06<57:27:05,  7.28s/it, lr: 1.0e-04 loss: 3.890e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1575/30000 [3:34:15<57:27:05,  7.28s/it, lr: 1.0e-04 loss: 9.347e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1575/30000 [3:34:15<57:27:05,  7.28s/it, lr: 1.0e-04 loss: 9.347e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1576/30000 [3:34:15<62:07:12,  7.87s/it, lr: 1.0e-04 loss: 9.347e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1576/30000 [3:34:15<62:07:12,  7.87s/it, lr: 1.0e-04 loss: 9.347e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1576/30000 [3:34:25<62:07:12,  7.87s/it, lr: 1.0e-04 loss: 3.999e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1576/30000 [3:34:25<62:07:12,  7.87s/it, lr: 1.0e-04 loss: 3.999e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1577/30000 [3:34:25<65:24:50,  8.29s/it, lr: 1.0e-04 loss: 3.999e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1577/30000 [3:34:25<65:24:50,  8.29s/it, lr: 1.0e-04 loss: 3.999e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1577/30000 [3:34:31<65:24:50,  8.29s/it, lr: 1.0e-04 loss: 7.628e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1577/30000 [3:34:31<65:24:50,  8.29s/it, lr: 1.0e-04 loss: 7.628e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1578/30000 [3:34:31<61:40:14,  7.81s/it, lr: 1.0e-04 loss: 7.628e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1578/30000 [3:34:31<61:40:14,  7.81s/it, lr: 1.0e-04 loss: 7.628e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1578/30000 [3:34:38<61:40:14,  7.81s/it, lr: 1.0e-04 loss: 5.829e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1578/30000 [3:34:38<61:40:14,  7.81s/it, lr: 1.0e-04 loss: 5.829e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1579/30000 [3:34:38<58:36:54,  7.42s/it, lr: 1.0e-04 loss: 5.829e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1579/30000 [3:34:38<58:36:54,  7.42s/it, lr: 1.0e-04 loss: 5.829e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1579/30000 [3:34:44<58:36:54,  7.42s/it, lr: 1.0e-04 loss: 1.150e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1579/30000 [3:34:44<58:36:54,  7.42s/it, lr: 1.0e-04 loss: 1.150e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1580/30000 [3:34:54<58:36:46,  7.42s/it, lr: 1.0e-04 loss: 6.163e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1580/30000 [3:34:54<58:36:46,  7.42s/it, lr: 1.0e-04 loss: 6.163e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1581/30000 [3:34:54<48:58:25,  6.20s/it, lr: 1.0e-04 loss: 6.163e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1581/30000 [3:34:54<48:58:25,  6.20s/it, lr: 1.0e-04 loss: 6.163e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1581/30000 [3:35:03<48:58:25,  6.20s/it, lr: 1.0e-04 loss: 7.675e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1581/30000 [3:35:03<48:58:25,  6.20s/it, lr: 1.0e-04 loss: 7.675e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1582/30000 [3:35:03<55:04:36,  6.98s/it, lr: 1.0e-04 loss: 7.675e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1582/30000 [3:35:03<55:04:36,  6.98s/it, lr: 1.0e-04 loss: 7.675e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1582/30000 [3:35:13<55:04:36,  6.98s/it, lr: 1.0e-04 loss: 3.960e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1582/30000 [3:35:13<55:04:36,  6.98s/it, lr: 1.0e-04 loss: 3.960e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1583/30000 [3:35:13<59:49:38,  7.58s/it, lr: 1.0e-04 loss: 3.960e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1583/30000 [3:35:13<59:49:38,  7.58s/it, lr: 1.0e-04 loss: 3.960e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1583/30000 [3:35:22<59:49:38,  7.58s/it, lr: 1.0e-04 loss: 6.112e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1583/30000 [3:35:22<59:49:38,  7.58s/it, lr: 1.0e-04 loss: 6.112e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1584/30000 [3:35:22<63:47:59,  8.08s/it, lr: 1.0e-04 loss: 6.112e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1584/30000 [3:35:22<63:47:59,  8.08s/it, lr: 1.0e-04 loss: 6.112e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1584/30000 [3:35:31<63:47:59,  8.08s/it, lr: 1.0e-04 loss: 4.849e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1584/30000 [3:35:31<63:47:59,  8.08s/it, lr: 1.0e-04 loss: 4.849e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1585/30000 [3:35:31<66:36:27,  8.44s/it, lr: 1.0e-04 loss: 4.849e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1585/30000 [3:35:31<66:36:27,  8.44s/it, lr: 1.0e-04 loss: 4.849e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1585/30000 [3:35:41<66:36:27,  8.44s/it, lr: 1.0e-04 loss: 9.105e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1585/30000 [3:35:41<66:36:27,  8.44s/it, lr: 1.0e-04 loss: 9.105e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1586/30000 [3:35:41<68:38:23,  8.70s/it, lr: 1.0e-04 loss: 9.105e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1586/30000 [3:35:41<68:38:23,  8.70s/it, lr: 1.0e-04 loss: 9.105e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1586/30000 [3:35:47<68:38:23,  8.70s/it, lr: 1.0e-04 loss: 8.471e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1586/30000 [3:35:47<68:38:23,  8.70s/it, lr: 1.0e-04 loss: 8.471e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1587/30000 [3:35:47<63:58:02,  8.10s/it, lr: 1.0e-04 loss: 8.471e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1587/30000 [3:35:47<63:58:02,  8.10s/it, lr: 1.0e-04 loss: 8.471e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1587/30000 [3:35:57<63:58:02,  8.10s/it, lr: 1.0e-04 loss: 5.725e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1587/30000 [3:35:57<63:58:02,  8.10s/it, lr: 1.0e-04 loss: 5.725e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1588/30000 [3:35:57<66:45:29,  8.46s/it, lr: 1.0e-04 loss: 5.725e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1588/30000 [3:35:57<66:45:29,  8.46s/it, lr: 1.0e-04 loss: 5.725e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1588/30000 [3:36:06<66:45:29,  8.46s/it, lr: 1.0e-04 loss: 4.087e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1588/30000 [3:36:06<66:45:29,  8.46s/it, lr: 1.0e-04 loss: 4.087e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1589/30000 [3:36:06<69:28:47,  8.80s/it, lr: 1.0e-04 loss: 4.087e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1589/30000 [3:36:06<69:28:47,  8.80s/it, lr: 1.0e-04 loss: 4.087e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1589/30000 [3:36:13<69:28:47,  8.80s/it, lr: 1.0e-04 loss: 8.092e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1589/30000 [3:36:13<69:28:47,  8.80s/it, lr: 1.0e-04 loss: 8.092e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1590/30000 [3:36:22<69:28:38,  8.80s/it, lr: 1.0e-04 loss: 7.928e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1590/30000 [3:36:22<69:28:38,  8.80s/it, lr: 1.0e-04 loss: 7.928e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1591/30000 [3:36:22<54:30:17,  6.91s/it, lr: 1.0e-04 loss: 7.928e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1591/30000 [3:36:22<54:30:17,  6.91s/it, lr: 1.0e-04 loss: 7.928e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1591/30000 [3:36:32<54:30:17,  6.91s/it, lr: 1.0e-04 loss: 5.681e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1591/30000 [3:36:32<54:30:17,  6.91s/it, lr: 1.0e-04 loss: 5.681e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1592/30000 [3:36:32<59:24:09,  7.53s/it, lr: 1.0e-04 loss: 5.681e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1592/30000 [3:36:32<59:24:09,  7.53s/it, lr: 1.0e-04 loss: 5.681e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1592/30000 [3:36:41<59:24:09,  7.53s/it, lr: 1.0e-04 loss: 2.491e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1592/30000 [3:36:41<59:24:09,  7.53s/it, lr: 1.0e-04 loss: 2.491e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1593/30000 [3:36:41<63:20:30,  8.03s/it, lr: 1.0e-04 loss: 2.491e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1593/30000 [3:36:41<63:20:30,  8.03s/it, lr: 1.0e-04 loss: 2.491e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1593/30000 [3:36:48<63:20:30,  8.03s/it, lr: 1.0e-04 loss: 9.067e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1593/30000 [3:36:48<63:20:30,  8.03s/it, lr: 1.0e-04 loss: 9.067e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1594/30000 [3:36:48<60:32:07,  7.67s/it, lr: 1.0e-04 loss: 9.067e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1594/30000 [3:36:48<60:32:07,  7.67s/it, lr: 1.0e-04 loss: 9.067e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1594/30000 [3:36:57<60:32:07,  7.67s/it, lr: 1.0e-04 loss: 6.743e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1594/30000 [3:36:57<60:32:07,  7.67s/it, lr: 1.0e-04 loss: 6.743e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1595/30000 [3:36:57<64:16:43,  8.15s/it, lr: 1.0e-04 loss: 6.743e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1595/30000 [3:36:57<64:16:43,  8.15s/it, lr: 1.0e-04 loss: 6.743e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1595/30000 [3:37:04<64:16:43,  8.15s/it, lr: 1.0e-04 loss: 8.637e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1595/30000 [3:37:04<64:16:43,  8.15s/it, lr: 1.0e-04 loss: 8.637e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1596/30000 [3:37:04<61:36:20,  7.81s/it, lr: 1.0e-04 loss: 8.637e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1596/30000 [3:37:04<61:36:20,  7.81s/it, lr: 1.0e-04 loss: 8.637e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1596/30000 [3:37:14<61:36:20,  7.81s/it, lr: 1.0e-04 loss: 5.311e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1596/30000 [3:37:14<61:36:20,  7.81s/it, lr: 1.0e-04 loss: 5.311e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1597/30000 [3:37:14<65:24:41,  8.29s/it, lr: 1.0e-04 loss: 5.311e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1597/30000 [3:37:14<65:24:41,  8.29s/it, lr: 1.0e-04 loss: 5.311e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1597/30000 [3:37:23<65:24:41,  8.29s/it, lr: 1.0e-04 loss: 5.951e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1597/30000 [3:37:23<65:24:41,  8.29s/it, lr: 1.0e-04 loss: 5.951e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1598/30000 [3:37:23<67:53:37,  8.61s/it, lr: 1.0e-04 loss: 5.951e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1598/30000 [3:37:23<67:53:37,  8.61s/it, lr: 1.0e-04 loss: 5.951e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1598/30000 [3:37:30<67:53:37,  8.61s/it, lr: 1.0e-04 loss: 2.860e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1598/30000 [3:37:30<67:53:37,  8.61s/it, lr: 1.0e-04 loss: 2.860e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1599/30000 [3:37:30<63:19:00,  8.03s/it, lr: 1.0e-04 loss: 2.860e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1599/30000 [3:37:30<63:19:00,  8.03s/it, lr: 1.0e-04 loss: 2.860e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1599/30000 [3:37:36<63:19:00,  8.03s/it, lr: 1.0e-04 loss: 7.737e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1599/30000 [3:37:36<63:19:00,  8.03s/it, lr: 1.0e-04 loss: 7.737e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1600/30000 [3:37:43<63:18:52,  8.03s/it, lr: 1.0e-04 loss: 1.583e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1600/30000 [3:37:43<63:18:52,  8.03s/it, lr: 1.0e-04 loss: 1.583e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1601/30000 [3:37:43<46:15:33,  5.86s/it, lr: 1.0e-04 loss: 1.583e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1601/30000 [3:37:43<46:15:33,  5.86s/it, lr: 1.0e-04 loss: 1.583e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1601/30000 [3:37:49<46:15:33,  5.86s/it, lr: 1.0e-04 loss: 6.913e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1601/30000 [3:37:49<46:15:33,  5.86s/it, lr: 1.0e-04 loss: 6.913e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1602/30000 [3:37:49<47:40:19,  6.04s/it, lr: 1.0e-04 loss: 6.913e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1602/30000 [3:37:49<47:40:19,  6.04s/it, lr: 1.0e-04 loss: 6.913e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1602/30000 [3:37:59<47:40:19,  6.04s/it, lr: 1.0e-04 loss: 1.036e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1602/30000 [3:37:59<47:40:19,  6.04s/it, lr: 1.0e-04 loss: 1.036e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1603/30000 [3:37:59<55:01:31,  6.98s/it, lr: 1.0e-04 loss: 1.036e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1603/30000 [3:37:59<55:01:31,  6.98s/it, lr: 1.0e-04 loss: 1.036e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1603/30000 [3:38:06<55:01:31,  6.98s/it, lr: 1.0e-04 loss: 1.037e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1603/30000 [3:38:06<55:01:31,  6.98s/it, lr: 1.0e-04 loss: 1.037e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1604/30000 [3:38:06<54:18:30,  6.89s/it, lr: 1.0e-04 loss: 1.037e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1604/30000 [3:38:06<54:18:30,  6.89s/it, lr: 1.0e-04 loss: 1.037e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1604/30000 [3:38:12<54:18:30,  6.89s/it, lr: 1.0e-04 loss: 7.151e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1604/30000 [3:38:12<54:18:30,  6.89s/it, lr: 1.0e-04 loss: 7.151e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1605/30000 [3:38:12<53:40:41,  6.81s/it, lr: 1.0e-04 loss: 7.151e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1605/30000 [3:38:12<53:40:41,  6.81s/it, lr: 1.0e-04 loss: 7.151e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1605/30000 [3:38:19<53:40:41,  6.81s/it, lr: 1.0e-04 loss: 1.119e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1605/30000 [3:38:19<53:40:41,  6.81s/it, lr: 1.0e-04 loss: 1.119e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1606/30000 [3:38:19<53:23:27,  6.77s/it, lr: 1.0e-04 loss: 1.119e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1606/30000 [3:38:19<53:23:27,  6.77s/it, lr: 1.0e-04 loss: 1.119e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1606/30000 [3:38:28<53:23:27,  6.77s/it, lr: 1.0e-04 loss: 5.382e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1606/30000 [3:38:28<53:23:27,  6.77s/it, lr: 1.0e-04 loss: 5.382e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1607/30000 [3:38:28<59:05:49,  7.49s/it, lr: 1.0e-04 loss: 5.382e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1607/30000 [3:38:28<59:05:49,  7.49s/it, lr: 1.0e-04 loss: 5.382e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1607/30000 [3:38:35<59:05:49,  7.49s/it, lr: 1.0e-04 loss: 1.053e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1607/30000 [3:38:35<59:05:49,  7.49s/it, lr: 1.0e-04 loss: 1.053e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1608/30000 [3:38:35<57:16:20,  7.26s/it, lr: 1.0e-04 loss: 1.053e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1608/30000 [3:38:35<57:16:20,  7.26s/it, lr: 1.0e-04 loss: 1.053e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1608/30000 [3:38:42<57:16:20,  7.26s/it, lr: 1.0e-04 loss: 4.607e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1608/30000 [3:38:42<57:16:20,  7.26s/it, lr: 1.0e-04 loss: 4.607e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1609/30000 [3:38:42<55:37:28,  7.05s/it, lr: 1.0e-04 loss: 4.607e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1609/30000 [3:38:42<55:37:28,  7.05s/it, lr: 1.0e-04 loss: 4.607e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1609/30000 [3:38:48<55:37:28,  7.05s/it, lr: 1.0e-04 loss: 1.226e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1609/30000 [3:38:48<55:37:28,  7.05s/it, lr: 1.0e-04 loss: 1.226e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1610/30000 [3:38:55<55:37:21,  7.05s/it, lr: 1.0e-04 loss: 5.843e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1610/30000 [3:38:55<55:37:21,  7.05s/it, lr: 1.0e-04 loss: 5.843e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1611/30000 [3:38:55<42:26:40,  5.38s/it, lr: 1.0e-04 loss: 5.843e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1611/30000 [3:38:55<42:26:40,  5.38s/it, lr: 1.0e-04 loss: 5.843e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1611/30000 [3:39:02<42:26:40,  5.38s/it, lr: 1.0e-04 loss: 1.282e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1611/30000 [3:39:02<42:26:40,  5.38s/it, lr: 1.0e-04 loss: 1.282e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1612/30000 [3:39:02<44:47:41,  5.68s/it, lr: 1.0e-04 loss: 1.282e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1612/30000 [3:39:02<44:47:41,  5.68s/it, lr: 1.0e-04 loss: 1.282e-01]
Xiang_wan_2_2_lora:   5%|##                                     | 1612/30000 [3:39:08<44:47:41,  5.68s/it, lr: 1.0e-04 loss: 6.201e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1612/30000 [3:39:08<44:47:41,  5.68s/it, lr: 1.0e-04 loss: 6.201e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1613/30000 [3:39:08<46:34:12,  5.91s/it, lr: 1.0e-04 loss: 6.201e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1613/30000 [3:39:08<46:34:12,  5.91s/it, lr: 1.0e-04 loss: 6.201e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1613/30000 [3:39:17<46:34:12,  5.91s/it, lr: 1.0e-04 loss: 5.616e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1613/30000 [3:39:17<46:34:12,  5.91s/it, lr: 1.0e-04 loss: 5.616e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1614/30000 [3:39:17<53:57:44,  6.84s/it, lr: 1.0e-04 loss: 5.616e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1614/30000 [3:39:17<53:57:44,  6.84s/it, lr: 1.0e-04 loss: 5.616e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1614/30000 [3:39:24<53:57:44,  6.84s/it, lr: 1.0e-04 loss: 5.652e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1614/30000 [3:39:24<53:57:44,  6.84s/it, lr: 1.0e-04 loss: 5.652e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1615/30000 [3:39:24<53:18:39,  6.76s/it, lr: 1.0e-04 loss: 5.652e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1615/30000 [3:39:24<53:18:39,  6.76s/it, lr: 1.0e-04 loss: 5.652e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1615/30000 [3:39:33<53:18:39,  6.76s/it, lr: 1.0e-04 loss: 8.182e-02]
Xiang_wan_2_2_lora:   5%|##                                     | 1615/30000 [3:39:33<53:18:39,  6.76s/it, lr: 1.0e-04 loss: 8.182e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1616/30000 [3:39:33<59:12:42,  7.51s/it, lr: 1.0e-04 loss: 8.182e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1616/30000 [3:39:33<59:12:42,  7.51s/it, lr: 1.0e-04 loss: 8.182e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1616/30000 [3:39:40<59:12:42,  7.51s/it, lr: 1.0e-04 loss: 4.506e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1616/30000 [3:39:40<59:12:42,  7.51s/it, lr: 1.0e-04 loss: 4.506e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1617/30000 [3:39:40<57:00:15,  7.23s/it, lr: 1.0e-04 loss: 4.506e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1617/30000 [3:39:40<57:00:15,  7.23s/it, lr: 1.0e-04 loss: 4.506e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1617/30000 [3:39:47<57:00:15,  7.23s/it, lr: 1.0e-04 loss: 1.092e-01]
Xiang_wan_2_2_lora:   5%|##1                                    | 1617/30000 [3:39:47<57:00:15,  7.23s/it, lr: 1.0e-04 loss: 1.092e-01]
Xiang_wan_2_2_lora:   5%|##1                                    | 1618/30000 [3:39:47<56:15:11,  7.14s/it, lr: 1.0e-04 loss: 1.092e-01]
Xiang_wan_2_2_lora:   5%|##1                                    | 1618/30000 [3:39:47<56:15:11,  7.14s/it, lr: 1.0e-04 loss: 1.092e-01]
Xiang_wan_2_2_lora:   5%|##1                                    | 1618/30000 [3:39:56<56:15:11,  7.14s/it, lr: 1.0e-04 loss: 5.826e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1618/30000 [3:39:56<56:15:11,  7.14s/it, lr: 1.0e-04 loss: 5.826e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1619/30000 [3:39:56<61:22:04,  7.78s/it, lr: 1.0e-04 loss: 5.826e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1619/30000 [3:39:56<61:22:04,  7.78s/it, lr: 1.0e-04 loss: 5.826e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1619/30000 [3:40:03<61:22:04,  7.78s/it, lr: 1.0e-04 loss: 9.460e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1619/30000 [3:40:03<61:22:04,  7.78s/it, lr: 1.0e-04 loss: 9.460e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1620/30000 [3:40:09<61:21:56,  7.78s/it, lr: 1.0e-04 loss: 6.486e-01]
Xiang_wan_2_2_lora:   5%|##1                                    | 1620/30000 [3:40:09<61:21:56,  7.78s/it, lr: 1.0e-04 loss: 6.486e-01]
Xiang_wan_2_2_lora:   5%|##1                                    | 1621/30000 [3:40:09<45:01:53,  5.71s/it, lr: 1.0e-04 loss: 6.486e-01]
Xiang_wan_2_2_lora:   5%|##1                                    | 1621/30000 [3:40:09<45:01:53,  5.71s/it, lr: 1.0e-04 loss: 6.486e-01]
Xiang_wan_2_2_lora:   5%|##1                                    | 1621/30000 [3:40:16<45:01:53,  5.71s/it, lr: 1.0e-04 loss: 9.613e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1621/30000 [3:40:16<45:01:53,  5.71s/it, lr: 1.0e-04 loss: 9.613e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1622/30000 [3:40:16<46:47:57,  5.94s/it, lr: 1.0e-04 loss: 9.613e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1622/30000 [3:40:16<46:47:57,  5.94s/it, lr: 1.0e-04 loss: 9.613e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1622/30000 [3:40:23<46:47:57,  5.94s/it, lr: 1.0e-04 loss: 8.435e-01]
Xiang_wan_2_2_lora:   5%|##1                                    | 1622/30000 [3:40:23<46:47:57,  5.94s/it, lr: 1.0e-04 loss: 8.435e-01]
Xiang_wan_2_2_lora:   5%|##1                                    | 1623/30000 [3:40:23<48:29:59,  6.15s/it, lr: 1.0e-04 loss: 8.435e-01]
Xiang_wan_2_2_lora:   5%|##1                                    | 1623/30000 [3:40:23<48:29:59,  6.15s/it, lr: 1.0e-04 loss: 8.435e-01]
Xiang_wan_2_2_lora:   5%|##1                                    | 1623/30000 [3:40:32<48:29:59,  6.15s/it, lr: 1.0e-04 loss: 8.603e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1623/30000 [3:40:32<48:29:59,  6.15s/it, lr: 1.0e-04 loss: 8.603e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1624/30000 [3:40:32<55:23:06,  7.03s/it, lr: 1.0e-04 loss: 8.603e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1624/30000 [3:40:32<55:23:06,  7.03s/it, lr: 1.0e-04 loss: 8.603e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1624/30000 [3:40:39<55:23:06,  7.03s/it, lr: 1.0e-04 loss: 6.118e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1624/30000 [3:40:39<55:23:06,  7.03s/it, lr: 1.0e-04 loss: 6.118e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1625/30000 [3:40:39<55:00:01,  6.98s/it, lr: 1.0e-04 loss: 6.118e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1625/30000 [3:40:39<55:00:01,  6.98s/it, lr: 1.0e-04 loss: 6.118e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1625/30000 [3:40:48<55:00:01,  6.98s/it, lr: 1.0e-04 loss: 7.001e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1625/30000 [3:40:48<55:00:01,  6.98s/it, lr: 1.0e-04 loss: 7.001e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1626/30000 [3:40:48<60:19:51,  7.65s/it, lr: 1.0e-04 loss: 7.001e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1626/30000 [3:40:48<60:19:51,  7.65s/it, lr: 1.0e-04 loss: 7.001e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1626/30000 [3:40:55<60:19:51,  7.65s/it, lr: 1.0e-04 loss: 8.277e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1626/30000 [3:40:55<60:19:51,  7.65s/it, lr: 1.0e-04 loss: 8.277e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1627/30000 [3:40:55<57:48:10,  7.33s/it, lr: 1.0e-04 loss: 8.277e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1627/30000 [3:40:55<57:48:10,  7.33s/it, lr: 1.0e-04 loss: 8.277e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1627/30000 [3:41:01<57:48:10,  7.33s/it, lr: 1.0e-04 loss: 7.925e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1627/30000 [3:41:01<57:48:10,  7.33s/it, lr: 1.0e-04 loss: 7.925e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1628/30000 [3:41:01<56:11:44,  7.13s/it, lr: 1.0e-04 loss: 7.925e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1628/30000 [3:41:01<56:11:44,  7.13s/it, lr: 1.0e-04 loss: 7.925e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1628/30000 [3:41:11<56:11:44,  7.13s/it, lr: 1.0e-04 loss: 1.100e-01]
Xiang_wan_2_2_lora:   5%|##1                                    | 1628/30000 [3:41:11<56:11:44,  7.13s/it, lr: 1.0e-04 loss: 1.100e-01]
Xiang_wan_2_2_lora:   5%|##1                                    | 1629/30000 [3:41:11<61:13:36,  7.77s/it, lr: 1.0e-04 loss: 1.100e-01]
Xiang_wan_2_2_lora:   5%|##1                                    | 1629/30000 [3:41:11<61:13:36,  7.77s/it, lr: 1.0e-04 loss: 1.100e-01]
Xiang_wan_2_2_lora:   5%|##1                                    | 1629/30000 [3:41:17<61:13:36,  7.77s/it, lr: 1.0e-04 loss: 9.088e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1629/30000 [3:41:17<61:13:36,  7.77s/it, lr: 1.0e-04 loss: 9.088e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1630/30000 [3:41:24<61:13:28,  7.77s/it, lr: 1.0e-04 loss: 4.595e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1630/30000 [3:41:24<61:13:28,  7.77s/it, lr: 1.0e-04 loss: 4.595e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1631/30000 [3:41:24<45:02:13,  5.72s/it, lr: 1.0e-04 loss: 4.595e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1631/30000 [3:41:24<45:02:13,  5.72s/it, lr: 1.0e-04 loss: 4.595e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1631/30000 [3:41:33<45:02:13,  5.72s/it, lr: 1.0e-04 loss: 7.620e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1631/30000 [3:41:33<45:02:13,  5.72s/it, lr: 1.0e-04 loss: 7.620e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1632/30000 [3:41:33<52:02:57,  6.61s/it, lr: 1.0e-04 loss: 7.620e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1632/30000 [3:41:33<52:02:57,  6.61s/it, lr: 1.0e-04 loss: 7.620e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1632/30000 [3:41:43<52:02:57,  6.61s/it, lr: 1.0e-04 loss: 9.221e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1632/30000 [3:41:43<52:02:57,  6.61s/it, lr: 1.0e-04 loss: 9.221e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1633/30000 [3:41:43<58:02:29,  7.37s/it, lr: 1.0e-04 loss: 9.221e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1633/30000 [3:41:43<58:02:29,  7.37s/it, lr: 1.0e-04 loss: 9.221e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1633/30000 [3:41:49<58:02:29,  7.37s/it, lr: 1.0e-04 loss: 7.268e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1633/30000 [3:41:49<58:02:29,  7.37s/it, lr: 1.0e-04 loss: 7.268e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1634/30000 [3:41:49<56:29:48,  7.17s/it, lr: 1.0e-04 loss: 7.268e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1634/30000 [3:41:49<56:29:48,  7.17s/it, lr: 1.0e-04 loss: 7.268e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1634/30000 [3:41:56<56:29:48,  7.17s/it, lr: 1.0e-04 loss: 3.819e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1634/30000 [3:41:56<56:29:48,  7.17s/it, lr: 1.0e-04 loss: 3.819e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1635/30000 [3:41:56<55:07:17,  7.00s/it, lr: 1.0e-04 loss: 3.819e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1635/30000 [3:41:56<55:07:17,  7.00s/it, lr: 1.0e-04 loss: 3.819e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1635/30000 [3:42:05<55:07:17,  7.00s/it, lr: 1.0e-04 loss: 6.383e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1635/30000 [3:42:05<55:07:17,  7.00s/it, lr: 1.0e-04 loss: 6.383e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1636/30000 [3:42:05<60:22:16,  7.66s/it, lr: 1.0e-04 loss: 6.383e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1636/30000 [3:42:05<60:22:16,  7.66s/it, lr: 1.0e-04 loss: 6.383e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1636/30000 [3:42:12<60:22:16,  7.66s/it, lr: 1.0e-04 loss: 8.373e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1636/30000 [3:42:12<60:22:16,  7.66s/it, lr: 1.0e-04 loss: 8.373e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1637/30000 [3:42:12<58:01:12,  7.36s/it, lr: 1.0e-04 loss: 8.373e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1637/30000 [3:42:12<58:01:12,  7.36s/it, lr: 1.0e-04 loss: 8.373e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1637/30000 [3:42:21<58:01:12,  7.36s/it, lr: 1.0e-04 loss: 4.980e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1637/30000 [3:42:21<58:01:12,  7.36s/it, lr: 1.0e-04 loss: 4.980e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1638/30000 [3:42:21<62:29:55,  7.93s/it, lr: 1.0e-04 loss: 4.980e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1638/30000 [3:42:21<62:29:55,  7.93s/it, lr: 1.0e-04 loss: 4.980e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1638/30000 [3:42:28<62:29:55,  7.93s/it, lr: 1.0e-04 loss: 5.803e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1638/30000 [3:42:28<62:29:55,  7.93s/it, lr: 1.0e-04 loss: 5.803e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1639/30000 [3:42:28<59:25:52,  7.54s/it, lr: 1.0e-04 loss: 5.803e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1639/30000 [3:42:28<59:25:52,  7.54s/it, lr: 1.0e-04 loss: 5.803e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1639/30000 [3:42:37<59:25:52,  7.54s/it, lr: 1.0e-04 loss: 5.740e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1639/30000 [3:42:37<59:25:52,  7.54s/it, lr: 1.0e-04 loss: 5.740e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1640/30000 [3:42:44<59:25:45,  7.54s/it, lr: 1.0e-04 loss: 1.001e+00]
Xiang_wan_2_2_lora:   5%|##1                                    | 1640/30000 [3:42:44<59:25:45,  7.54s/it, lr: 1.0e-04 loss: 1.001e+00]
Xiang_wan_2_2_lora:   5%|##1                                    | 1641/30000 [3:42:44<44:05:04,  5.60s/it, lr: 1.0e-04 loss: 1.001e+00]
Xiang_wan_2_2_lora:   5%|##1                                    | 1641/30000 [3:42:44<44:05:04,  5.60s/it, lr: 1.0e-04 loss: 1.001e+00]
Xiang_wan_2_2_lora:   5%|##1                                    | 1641/30000 [3:42:50<44:05:04,  5.60s/it, lr: 1.0e-04 loss: 8.533e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1641/30000 [3:42:50<44:05:04,  5.60s/it, lr: 1.0e-04 loss: 8.533e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1642/30000 [3:42:50<46:02:27,  5.84s/it, lr: 1.0e-04 loss: 8.533e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1642/30000 [3:42:50<46:02:27,  5.84s/it, lr: 1.0e-04 loss: 8.533e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1642/30000 [3:43:00<46:02:27,  5.84s/it, lr: 1.0e-04 loss: 6.899e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1642/30000 [3:43:00<46:02:27,  5.84s/it, lr: 1.0e-04 loss: 6.899e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1643/30000 [3:43:00<53:07:50,  6.75s/it, lr: 1.0e-04 loss: 6.899e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1643/30000 [3:43:00<53:07:50,  6.75s/it, lr: 1.0e-04 loss: 6.899e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1643/30000 [3:43:07<53:07:50,  6.75s/it, lr: 1.0e-04 loss: 8.541e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1643/30000 [3:43:07<53:07:50,  6.75s/it, lr: 1.0e-04 loss: 8.541e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1644/30000 [3:43:07<53:07:17,  6.74s/it, lr: 1.0e-04 loss: 8.541e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1644/30000 [3:43:07<53:07:17,  6.74s/it, lr: 1.0e-04 loss: 8.541e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1644/30000 [3:43:16<53:07:17,  6.74s/it, lr: 1.0e-04 loss: 3.188e-01]
Xiang_wan_2_2_lora:   5%|##1                                    | 1644/30000 [3:43:16<53:07:17,  6.74s/it, lr: 1.0e-04 loss: 3.188e-01]
Xiang_wan_2_2_lora:   5%|##1                                    | 1645/30000 [3:43:16<58:54:57,  7.48s/it, lr: 1.0e-04 loss: 3.188e-01]
Xiang_wan_2_2_lora:   5%|##1                                    | 1645/30000 [3:43:16<58:54:57,  7.48s/it, lr: 1.0e-04 loss: 3.188e-01]
Xiang_wan_2_2_lora:   5%|##1                                    | 1645/30000 [3:43:23<58:54:57,  7.48s/it, lr: 1.0e-04 loss: 7.686e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1645/30000 [3:43:23<58:54:57,  7.48s/it, lr: 1.0e-04 loss: 7.686e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1646/30000 [3:43:23<56:58:53,  7.23s/it, lr: 1.0e-04 loss: 7.686e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1646/30000 [3:43:23<56:58:53,  7.23s/it, lr: 1.0e-04 loss: 7.686e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1646/30000 [3:43:29<56:58:53,  7.23s/it, lr: 1.0e-04 loss: 2.143e-01]
Xiang_wan_2_2_lora:   5%|##1                                    | 1646/30000 [3:43:29<56:58:53,  7.23s/it, lr: 1.0e-04 loss: 2.143e-01]
Xiang_wan_2_2_lora:   5%|##1                                    | 1647/30000 [3:43:29<55:59:45,  7.11s/it, lr: 1.0e-04 loss: 2.143e-01]
Xiang_wan_2_2_lora:   5%|##1                                    | 1647/30000 [3:43:29<55:59:45,  7.11s/it, lr: 1.0e-04 loss: 2.143e-01]
Xiang_wan_2_2_lora:   5%|##1                                    | 1647/30000 [3:43:36<55:59:45,  7.11s/it, lr: 1.0e-04 loss: 1.244e-01]
Xiang_wan_2_2_lora:   5%|##1                                    | 1647/30000 [3:43:36<55:59:45,  7.11s/it, lr: 1.0e-04 loss: 1.244e-01]
Xiang_wan_2_2_lora:   5%|##1                                    | 1648/30000 [3:43:36<54:54:36,  6.97s/it, lr: 1.0e-04 loss: 1.244e-01]
Xiang_wan_2_2_lora:   5%|##1                                    | 1648/30000 [3:43:36<54:54:36,  6.97s/it, lr: 1.0e-04 loss: 1.244e-01]
Xiang_wan_2_2_lora:   5%|##1                                    | 1648/30000 [3:43:43<54:54:36,  6.97s/it, lr: 1.0e-04 loss: 8.779e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1648/30000 [3:43:43<54:54:36,  6.97s/it, lr: 1.0e-04 loss: 8.779e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1649/30000 [3:43:43<53:55:04,  6.85s/it, lr: 1.0e-04 loss: 8.779e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1649/30000 [3:43:43<53:55:04,  6.85s/it, lr: 1.0e-04 loss: 8.779e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1649/30000 [3:43:49<53:55:04,  6.85s/it, lr: 1.0e-04 loss: 7.911e-02]
Xiang_wan_2_2_lora:   5%|##1                                    | 1649/30000 [3:43:49<53:55:04,  6.85s/it, lr: 1.0e-04 loss: 7.911e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1650/30000 [3:43:58<53:54:57,  6.85s/it, lr: 1.0e-04 loss: 5.447e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1650/30000 [3:43:58<53:54:57,  6.85s/it, lr: 1.0e-04 loss: 5.447e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1651/30000 [3:43:58<45:55:57,  5.83s/it, lr: 1.0e-04 loss: 5.447e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1651/30000 [3:43:58<45:55:57,  5.83s/it, lr: 1.0e-04 loss: 5.447e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1651/30000 [3:44:08<45:55:57,  5.83s/it, lr: 1.0e-04 loss: 7.312e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1651/30000 [3:44:08<45:55:57,  5.83s/it, lr: 1.0e-04 loss: 7.312e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1652/30000 [3:44:08<52:56:18,  6.72s/it, lr: 1.0e-04 loss: 7.312e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1652/30000 [3:44:08<52:56:18,  6.72s/it, lr: 1.0e-04 loss: 7.312e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1652/30000 [3:44:14<52:56:18,  6.72s/it, lr: 1.0e-04 loss: 8.159e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1652/30000 [3:44:14<52:56:18,  6.72s/it, lr: 1.0e-04 loss: 8.159e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1653/30000 [3:44:14<52:34:00,  6.68s/it, lr: 1.0e-04 loss: 8.159e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1653/30000 [3:44:14<52:34:00,  6.68s/it, lr: 1.0e-04 loss: 8.159e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1653/30000 [3:44:21<52:34:00,  6.68s/it, lr: 1.0e-04 loss: 1.393e-01]
Xiang_wan_2_2_lora:   6%|##1                                    | 1653/30000 [3:44:21<52:34:00,  6.68s/it, lr: 1.0e-04 loss: 1.393e-01]
Xiang_wan_2_2_lora:   6%|##1                                    | 1654/30000 [3:44:21<52:55:57,  6.72s/it, lr: 1.0e-04 loss: 1.393e-01]
Xiang_wan_2_2_lora:   6%|##1                                    | 1654/30000 [3:44:21<52:55:57,  6.72s/it, lr: 1.0e-04 loss: 1.393e-01]
Xiang_wan_2_2_lora:   6%|##1                                    | 1654/30000 [3:44:31<52:55:57,  6.72s/it, lr: 1.0e-04 loss: 8.265e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1654/30000 [3:44:31<52:55:57,  6.72s/it, lr: 1.0e-04 loss: 8.265e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1655/30000 [3:44:31<58:36:07,  7.44s/it, lr: 1.0e-04 loss: 8.265e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1655/30000 [3:44:31<58:36:07,  7.44s/it, lr: 1.0e-04 loss: 8.265e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1655/30000 [3:44:37<58:36:07,  7.44s/it, lr: 1.0e-04 loss: 9.643e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1655/30000 [3:44:37<58:36:07,  7.44s/it, lr: 1.0e-04 loss: 9.643e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1656/30000 [3:44:37<56:40:52,  7.20s/it, lr: 1.0e-04 loss: 9.643e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1656/30000 [3:44:37<56:40:52,  7.20s/it, lr: 1.0e-04 loss: 9.643e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1656/30000 [3:44:44<56:40:52,  7.20s/it, lr: 1.0e-04 loss: 9.506e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1656/30000 [3:44:44<56:40:52,  7.20s/it, lr: 1.0e-04 loss: 9.506e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1657/30000 [3:44:44<55:25:25,  7.04s/it, lr: 1.0e-04 loss: 9.506e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1657/30000 [3:44:44<55:25:25,  7.04s/it, lr: 1.0e-04 loss: 9.506e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1657/30000 [3:44:53<55:25:25,  7.04s/it, lr: 1.0e-04 loss: 6.211e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1657/30000 [3:44:53<55:25:25,  7.04s/it, lr: 1.0e-04 loss: 6.211e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1658/30000 [3:44:53<60:44:16,  7.71s/it, lr: 1.0e-04 loss: 6.211e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1658/30000 [3:44:53<60:44:16,  7.71s/it, lr: 1.0e-04 loss: 6.211e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1658/30000 [3:45:02<60:44:16,  7.71s/it, lr: 1.0e-04 loss: 6.261e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1658/30000 [3:45:02<60:44:16,  7.71s/it, lr: 1.0e-04 loss: 6.261e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1659/30000 [3:45:02<64:27:49,  8.19s/it, lr: 1.0e-04 loss: 6.261e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1659/30000 [3:45:02<64:27:49,  8.19s/it, lr: 1.0e-04 loss: 6.261e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1659/30000 [3:45:09<64:27:49,  8.19s/it, lr: 1.0e-04 loss: 7.764e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1659/30000 [3:45:09<64:27:49,  8.19s/it, lr: 1.0e-04 loss: 7.764e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1660/30000 [3:45:18<64:27:41,  8.19s/it, lr: 1.0e-04 loss: 9.626e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1660/30000 [3:45:18<64:27:41,  8.19s/it, lr: 1.0e-04 loss: 9.626e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1661/30000 [3:45:18<51:41:43,  6.57s/it, lr: 1.0e-04 loss: 9.626e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1661/30000 [3:45:18<51:41:43,  6.57s/it, lr: 1.0e-04 loss: 9.626e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1661/30000 [3:45:28<51:41:43,  6.57s/it, lr: 1.0e-04 loss: 7.072e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1661/30000 [3:45:28<51:41:43,  6.57s/it, lr: 1.0e-04 loss: 7.072e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1662/30000 [3:45:28<57:35:45,  7.32s/it, lr: 1.0e-04 loss: 7.072e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1662/30000 [3:45:28<57:35:45,  7.32s/it, lr: 1.0e-04 loss: 7.072e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1662/30000 [3:45:37<57:35:45,  7.32s/it, lr: 1.0e-04 loss: 5.505e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1662/30000 [3:45:37<57:35:45,  7.32s/it, lr: 1.0e-04 loss: 5.505e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1663/30000 [3:45:37<61:44:52,  7.84s/it, lr: 1.0e-04 loss: 5.505e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1663/30000 [3:45:37<61:44:52,  7.84s/it, lr: 1.0e-04 loss: 5.505e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1663/30000 [3:45:44<61:44:52,  7.84s/it, lr: 1.0e-04 loss: 1.244e-01]
Xiang_wan_2_2_lora:   6%|##1                                    | 1663/30000 [3:45:44<61:44:52,  7.84s/it, lr: 1.0e-04 loss: 1.244e-01]
Xiang_wan_2_2_lora:   6%|##1                                    | 1664/30000 [3:45:44<59:01:46,  7.50s/it, lr: 1.0e-04 loss: 1.244e-01]
Xiang_wan_2_2_lora:   6%|##1                                    | 1664/30000 [3:45:44<59:01:46,  7.50s/it, lr: 1.0e-04 loss: 1.244e-01]
Xiang_wan_2_2_lora:   6%|##1                                    | 1664/30000 [3:45:53<59:01:46,  7.50s/it, lr: 1.0e-04 loss: 2.241e-01]
Xiang_wan_2_2_lora:   6%|##1                                    | 1664/30000 [3:45:53<59:01:46,  7.50s/it, lr: 1.0e-04 loss: 2.241e-01]
Xiang_wan_2_2_lora:   6%|##1                                    | 1665/30000 [3:45:53<62:59:44,  8.00s/it, lr: 1.0e-04 loss: 2.241e-01]
Xiang_wan_2_2_lora:   6%|##1                                    | 1665/30000 [3:45:53<62:59:44,  8.00s/it, lr: 1.0e-04 loss: 2.241e-01]
Xiang_wan_2_2_lora:   6%|##1                                    | 1665/30000 [3:46:03<62:59:44,  8.00s/it, lr: 1.0e-04 loss: 5.951e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1665/30000 [3:46:03<62:59:44,  8.00s/it, lr: 1.0e-04 loss: 5.951e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1666/30000 [3:46:03<66:04:58,  8.40s/it, lr: 1.0e-04 loss: 5.951e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1666/30000 [3:46:03<66:04:58,  8.40s/it, lr: 1.0e-04 loss: 5.951e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1666/30000 [3:46:12<66:04:58,  8.40s/it, lr: 1.0e-04 loss: 7.771e-01]
Xiang_wan_2_2_lora:   6%|##1                                    | 1666/30000 [3:46:12<66:04:58,  8.40s/it, lr: 1.0e-04 loss: 7.771e-01]
Xiang_wan_2_2_lora:   6%|##1                                    | 1667/30000 [3:46:12<68:15:30,  8.67s/it, lr: 1.0e-04 loss: 7.771e-01]
Xiang_wan_2_2_lora:   6%|##1                                    | 1667/30000 [3:46:12<68:15:30,  8.67s/it, lr: 1.0e-04 loss: 7.771e-01]
Xiang_wan_2_2_lora:   6%|##1                                    | 1667/30000 [3:46:21<68:15:30,  8.67s/it, lr: 1.0e-04 loss: 7.415e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1667/30000 [3:46:21<68:15:30,  8.67s/it, lr: 1.0e-04 loss: 7.415e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1668/30000 [3:46:21<69:49:16,  8.87s/it, lr: 1.0e-04 loss: 7.415e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1668/30000 [3:46:21<69:49:16,  8.87s/it, lr: 1.0e-04 loss: 7.415e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1668/30000 [3:46:31<69:49:16,  8.87s/it, lr: 1.0e-04 loss: 6.021e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1668/30000 [3:46:31<69:49:16,  8.87s/it, lr: 1.0e-04 loss: 6.021e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1669/30000 [3:46:31<71:17:33,  9.06s/it, lr: 1.0e-04 loss: 6.021e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1669/30000 [3:46:31<71:17:33,  9.06s/it, lr: 1.0e-04 loss: 6.021e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1669/30000 [3:46:37<71:17:33,  9.06s/it, lr: 1.0e-04 loss: 8.777e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1669/30000 [3:46:37<71:17:33,  9.06s/it, lr: 1.0e-04 loss: 8.777e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1670/30000 [3:46:47<71:17:24,  9.06s/it, lr: 1.0e-04 loss: 5.155e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1670/30000 [3:46:47<71:17:24,  9.06s/it, lr: 1.0e-04 loss: 5.155e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1671/30000 [3:46:47<55:22:56,  7.04s/it, lr: 1.0e-04 loss: 5.155e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1671/30000 [3:46:47<55:22:56,  7.04s/it, lr: 1.0e-04 loss: 5.155e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1671/30000 [3:46:53<55:22:56,  7.04s/it, lr: 1.0e-04 loss: 1.096e-01]
Xiang_wan_2_2_lora:   6%|##1                                    | 1671/30000 [3:46:53<55:22:56,  7.04s/it, lr: 1.0e-04 loss: 1.096e-01]
Xiang_wan_2_2_lora:   6%|##1                                    | 1672/30000 [3:46:53<54:34:23,  6.94s/it, lr: 1.0e-04 loss: 1.096e-01]
Xiang_wan_2_2_lora:   6%|##1                                    | 1672/30000 [3:46:53<54:34:23,  6.94s/it, lr: 1.0e-04 loss: 1.096e-01]
Xiang_wan_2_2_lora:   6%|##1                                    | 1672/30000 [3:47:03<54:34:23,  6.94s/it, lr: 1.0e-04 loss: 4.506e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1672/30000 [3:47:03<54:34:23,  6.94s/it, lr: 1.0e-04 loss: 4.506e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1673/30000 [3:47:03<59:18:14,  7.54s/it, lr: 1.0e-04 loss: 4.506e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1673/30000 [3:47:03<59:18:14,  7.54s/it, lr: 1.0e-04 loss: 4.506e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1673/30000 [3:47:12<59:18:14,  7.54s/it, lr: 1.0e-04 loss: 5.921e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1673/30000 [3:47:12<59:18:14,  7.54s/it, lr: 1.0e-04 loss: 5.921e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1674/30000 [3:47:12<63:10:57,  8.03s/it, lr: 1.0e-04 loss: 5.921e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1674/30000 [3:47:12<63:10:57,  8.03s/it, lr: 1.0e-04 loss: 5.921e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1674/30000 [3:47:21<63:10:57,  8.03s/it, lr: 1.0e-04 loss: 2.550e-01]
Xiang_wan_2_2_lora:   6%|##1                                    | 1674/30000 [3:47:21<63:10:57,  8.03s/it, lr: 1.0e-04 loss: 2.550e-01]
Xiang_wan_2_2_lora:   6%|##1                                    | 1675/30000 [3:47:21<66:05:51,  8.40s/it, lr: 1.0e-04 loss: 2.550e-01]
Xiang_wan_2_2_lora:   6%|##1                                    | 1675/30000 [3:47:21<66:05:51,  8.40s/it, lr: 1.0e-04 loss: 2.550e-01]
Xiang_wan_2_2_lora:   6%|##1                                    | 1675/30000 [3:47:31<66:05:51,  8.40s/it, lr: 1.0e-04 loss: 5.663e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1675/30000 [3:47:31<66:05:51,  8.40s/it, lr: 1.0e-04 loss: 5.663e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1676/30000 [3:47:31<68:37:30,  8.72s/it, lr: 1.0e-04 loss: 5.663e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1676/30000 [3:47:31<68:37:30,  8.72s/it, lr: 1.0e-04 loss: 5.663e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1676/30000 [3:47:40<68:37:30,  8.72s/it, lr: 1.0e-04 loss: 4.037e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1676/30000 [3:47:40<68:37:30,  8.72s/it, lr: 1.0e-04 loss: 4.037e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1677/30000 [3:47:40<70:02:19,  8.90s/it, lr: 1.0e-04 loss: 4.037e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1677/30000 [3:47:40<70:02:19,  8.90s/it, lr: 1.0e-04 loss: 4.037e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1677/30000 [3:47:49<70:02:19,  8.90s/it, lr: 1.0e-04 loss: 7.829e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1677/30000 [3:47:49<70:02:19,  8.90s/it, lr: 1.0e-04 loss: 7.829e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1678/30000 [3:47:49<71:08:06,  9.04s/it, lr: 1.0e-04 loss: 7.829e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1678/30000 [3:47:49<71:08:06,  9.04s/it, lr: 1.0e-04 loss: 7.829e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1678/30000 [3:47:59<71:08:06,  9.04s/it, lr: 1.0e-04 loss: 4.680e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1678/30000 [3:47:59<71:08:06,  9.04s/it, lr: 1.0e-04 loss: 4.680e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1679/30000 [3:47:59<71:44:34,  9.12s/it, lr: 1.0e-04 loss: 4.680e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1679/30000 [3:47:59<71:44:34,  9.12s/it, lr: 1.0e-04 loss: 4.680e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1679/30000 [3:48:05<71:44:34,  9.12s/it, lr: 1.0e-04 loss: 8.573e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1679/30000 [3:48:05<71:44:34,  9.12s/it, lr: 1.0e-04 loss: 8.573e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1680/30000 [3:48:12<71:44:25,  9.12s/it, lr: 1.0e-04 loss: 4.363e-01]
Xiang_wan_2_2_lora:   6%|##1                                    | 1680/30000 [3:48:12<71:44:25,  9.12s/it, lr: 1.0e-04 loss: 4.363e-01]
Xiang_wan_2_2_lora:   6%|##1                                    | 1681/30000 [3:48:12<50:45:07,  6.45s/it, lr: 1.0e-04 loss: 4.363e-01]
Xiang_wan_2_2_lora:   6%|##1                                    | 1681/30000 [3:48:12<50:45:07,  6.45s/it, lr: 1.0e-04 loss: 4.363e-01]
Xiang_wan_2_2_lora:   6%|##1                                    | 1681/30000 [3:48:19<50:45:07,  6.45s/it, lr: 1.0e-04 loss: 7.518e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1681/30000 [3:48:19<50:45:07,  6.45s/it, lr: 1.0e-04 loss: 7.518e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1682/30000 [3:48:19<51:01:02,  6.49s/it, lr: 1.0e-04 loss: 7.518e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1682/30000 [3:48:19<51:01:02,  6.49s/it, lr: 1.0e-04 loss: 7.518e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1682/30000 [3:48:25<51:01:02,  6.49s/it, lr: 1.0e-04 loss: 5.310e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1682/30000 [3:48:25<51:01:02,  6.49s/it, lr: 1.0e-04 loss: 5.310e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1683/30000 [3:48:25<51:35:39,  6.56s/it, lr: 1.0e-04 loss: 5.310e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1683/30000 [3:48:25<51:35:39,  6.56s/it, lr: 1.0e-04 loss: 5.310e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1683/30000 [3:48:35<51:35:39,  6.56s/it, lr: 1.0e-04 loss: 6.453e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1683/30000 [3:48:35<51:35:39,  6.56s/it, lr: 1.0e-04 loss: 6.453e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1684/30000 [3:48:35<57:26:29,  7.30s/it, lr: 1.0e-04 loss: 6.453e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1684/30000 [3:48:35<57:26:29,  7.30s/it, lr: 1.0e-04 loss: 6.453e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1684/30000 [3:48:41<57:26:29,  7.30s/it, lr: 1.0e-04 loss: 5.763e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1684/30000 [3:48:41<57:26:29,  7.30s/it, lr: 1.0e-04 loss: 5.763e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1685/30000 [3:48:41<55:50:58,  7.10s/it, lr: 1.0e-04 loss: 5.763e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1685/30000 [3:48:41<55:50:58,  7.10s/it, lr: 1.0e-04 loss: 5.763e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1685/30000 [3:48:51<55:50:58,  7.10s/it, lr: 1.0e-04 loss: 7.281e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1685/30000 [3:48:51<55:50:58,  7.10s/it, lr: 1.0e-04 loss: 7.281e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1686/30000 [3:48:51<60:48:12,  7.73s/it, lr: 1.0e-04 loss: 7.281e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1686/30000 [3:48:51<60:48:12,  7.73s/it, lr: 1.0e-04 loss: 7.281e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1686/30000 [3:48:57<60:48:12,  7.73s/it, lr: 1.0e-04 loss: 6.913e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1686/30000 [3:48:57<60:48:12,  7.73s/it, lr: 1.0e-04 loss: 6.913e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1687/30000 [3:48:57<58:18:27,  7.41s/it, lr: 1.0e-04 loss: 6.913e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1687/30000 [3:48:57<58:18:27,  7.41s/it, lr: 1.0e-04 loss: 6.913e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1687/30000 [3:49:07<58:18:27,  7.41s/it, lr: 1.0e-04 loss: 4.925e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1687/30000 [3:49:07<58:18:27,  7.41s/it, lr: 1.0e-04 loss: 4.925e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1688/30000 [3:49:07<63:02:42,  8.02s/it, lr: 1.0e-04 loss: 4.925e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1688/30000 [3:49:07<63:02:42,  8.02s/it, lr: 1.0e-04 loss: 4.925e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1688/30000 [3:49:16<63:02:42,  8.02s/it, lr: 1.0e-04 loss: 4.487e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1688/30000 [3:49:16<63:02:42,  8.02s/it, lr: 1.0e-04 loss: 4.487e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1689/30000 [3:49:16<65:58:17,  8.39s/it, lr: 1.0e-04 loss: 4.487e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1689/30000 [3:49:16<65:58:17,  8.39s/it, lr: 1.0e-04 loss: 4.487e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1689/30000 [3:49:23<65:58:17,  8.39s/it, lr: 1.0e-04 loss: 1.237e-01]
Xiang_wan_2_2_lora:   6%|##1                                    | 1689/30000 [3:49:23<65:58:17,  8.39s/it, lr: 1.0e-04 loss: 1.237e-01]
Xiang_wan_2_2_lora:   6%|##1                                    | 1690/30000 [3:49:29<65:58:08,  8.39s/it, lr: 1.0e-04 loss: 5.586e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1690/30000 [3:49:29<65:58:08,  8.39s/it, lr: 1.0e-04 loss: 5.586e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1691/30000 [3:49:29<47:58:22,  6.10s/it, lr: 1.0e-04 loss: 5.586e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1691/30000 [3:49:29<47:58:22,  6.10s/it, lr: 1.0e-04 loss: 5.586e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1691/30000 [3:49:36<47:58:22,  6.10s/it, lr: 1.0e-04 loss: 9.340e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1691/30000 [3:49:36<47:58:22,  6.10s/it, lr: 1.0e-04 loss: 9.340e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1692/30000 [3:49:36<48:56:09,  6.22s/it, lr: 1.0e-04 loss: 9.340e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1692/30000 [3:49:36<48:56:09,  6.22s/it, lr: 1.0e-04 loss: 9.340e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1692/30000 [3:49:43<48:56:09,  6.22s/it, lr: 1.0e-04 loss: 6.430e-02]
Xiang_wan_2_2_lora:   6%|##1                                    | 1692/30000 [3:49:43<48:56:09,  6.22s/it, lr: 1.0e-04 loss: 6.430e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1693/30000 [3:49:43<49:34:54,  6.31s/it, lr: 1.0e-04 loss: 6.430e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1693/30000 [3:49:43<49:34:54,  6.31s/it, lr: 1.0e-04 loss: 6.430e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1693/30000 [3:49:49<49:34:54,  6.31s/it, lr: 1.0e-04 loss: 6.796e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1693/30000 [3:49:49<49:34:54,  6.31s/it, lr: 1.0e-04 loss: 6.796e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1694/30000 [3:49:49<50:17:11,  6.40s/it, lr: 1.0e-04 loss: 6.796e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1694/30000 [3:49:49<50:17:11,  6.40s/it, lr: 1.0e-04 loss: 6.796e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1694/30000 [3:49:58<50:17:11,  6.40s/it, lr: 1.0e-04 loss: 3.710e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1694/30000 [3:49:58<50:17:11,  6.40s/it, lr: 1.0e-04 loss: 3.710e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1695/30000 [3:49:58<56:37:14,  7.20s/it, lr: 1.0e-04 loss: 3.710e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1695/30000 [3:49:58<56:37:14,  7.20s/it, lr: 1.0e-04 loss: 3.710e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1695/30000 [3:50:05<56:37:14,  7.20s/it, lr: 1.0e-04 loss: 1.045e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1695/30000 [3:50:05<56:37:14,  7.20s/it, lr: 1.0e-04 loss: 1.045e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1696/30000 [3:50:05<55:19:59,  7.04s/it, lr: 1.0e-04 loss: 1.045e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1696/30000 [3:50:05<55:19:59,  7.04s/it, lr: 1.0e-04 loss: 1.045e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1696/30000 [3:50:14<55:19:59,  7.04s/it, lr: 1.0e-04 loss: 1.104e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1696/30000 [3:50:14<55:19:59,  7.04s/it, lr: 1.0e-04 loss: 1.104e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1697/30000 [3:50:14<60:30:54,  7.70s/it, lr: 1.0e-04 loss: 1.104e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1697/30000 [3:50:14<60:30:54,  7.70s/it, lr: 1.0e-04 loss: 1.104e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1697/30000 [3:50:24<60:30:54,  7.70s/it, lr: 1.0e-04 loss: 4.762e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1697/30000 [3:50:24<60:30:54,  7.70s/it, lr: 1.0e-04 loss: 4.762e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1698/30000 [3:50:24<64:47:58,  8.24s/it, lr: 1.0e-04 loss: 4.762e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1698/30000 [3:50:24<64:47:58,  8.24s/it, lr: 1.0e-04 loss: 4.762e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1698/30000 [3:50:31<64:47:58,  8.24s/it, lr: 1.0e-04 loss: 9.614e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1698/30000 [3:50:31<64:47:58,  8.24s/it, lr: 1.0e-04 loss: 9.614e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1699/30000 [3:50:31<60:58:50,  7.76s/it, lr: 1.0e-04 loss: 9.614e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1699/30000 [3:50:31<60:58:50,  7.76s/it, lr: 1.0e-04 loss: 9.614e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1699/30000 [3:50:40<60:58:50,  7.76s/it, lr: 1.0e-04 loss: 6.721e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1699/30000 [3:50:40<60:58:50,  7.76s/it, lr: 1.0e-04 loss: 6.721e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1700/30000 [3:50:49<60:58:42,  7.76s/it, lr: 1.0e-04 loss: 5.007e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1700/30000 [3:50:49<60:58:42,  7.76s/it, lr: 1.0e-04 loss: 5.007e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1701/30000 [3:50:49<49:55:09,  6.35s/it, lr: 1.0e-04 loss: 5.007e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1701/30000 [3:50:49<49:55:09,  6.35s/it, lr: 1.0e-04 loss: 5.007e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1701/30000 [3:50:59<49:55:09,  6.35s/it, lr: 1.0e-04 loss: 8.164e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1701/30000 [3:50:59<49:55:09,  6.35s/it, lr: 1.0e-04 loss: 8.164e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1702/30000 [3:50:59<55:44:23,  7.09s/it, lr: 1.0e-04 loss: 8.164e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1702/30000 [3:50:59<55:44:23,  7.09s/it, lr: 1.0e-04 loss: 8.164e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1702/30000 [3:51:05<55:44:23,  7.09s/it, lr: 1.0e-04 loss: 6.943e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1702/30000 [3:51:05<55:44:23,  7.09s/it, lr: 1.0e-04 loss: 6.943e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1703/30000 [3:51:05<54:38:53,  6.95s/it, lr: 1.0e-04 loss: 6.943e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1703/30000 [3:51:05<54:38:53,  6.95s/it, lr: 1.0e-04 loss: 6.943e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1703/30000 [3:51:12<54:38:53,  6.95s/it, lr: 1.0e-04 loss: 1.102e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1703/30000 [3:51:12<54:38:53,  6.95s/it, lr: 1.0e-04 loss: 1.102e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1704/30000 [3:51:12<53:56:50,  6.86s/it, lr: 1.0e-04 loss: 1.102e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1704/30000 [3:51:12<53:56:50,  6.86s/it, lr: 1.0e-04 loss: 1.102e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1704/30000 [3:51:19<53:56:50,  6.86s/it, lr: 1.0e-04 loss: 6.215e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1704/30000 [3:51:19<53:56:50,  6.86s/it, lr: 1.0e-04 loss: 6.215e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1705/30000 [3:51:19<53:59:03,  6.87s/it, lr: 1.0e-04 loss: 6.215e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1705/30000 [3:51:19<53:59:03,  6.87s/it, lr: 1.0e-04 loss: 6.215e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1705/30000 [3:51:28<53:59:03,  6.87s/it, lr: 1.0e-04 loss: 7.821e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1705/30000 [3:51:28<53:59:03,  6.87s/it, lr: 1.0e-04 loss: 7.821e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1706/30000 [3:51:28<59:25:31,  7.56s/it, lr: 1.0e-04 loss: 7.821e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1706/30000 [3:51:28<59:25:31,  7.56s/it, lr: 1.0e-04 loss: 7.821e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1706/30000 [3:51:35<59:25:31,  7.56s/it, lr: 1.0e-04 loss: 2.325e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1706/30000 [3:51:35<59:25:31,  7.56s/it, lr: 1.0e-04 loss: 2.325e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1707/30000 [3:51:35<57:18:38,  7.29s/it, lr: 1.0e-04 loss: 2.325e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1707/30000 [3:51:35<57:18:38,  7.29s/it, lr: 1.0e-04 loss: 2.325e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1707/30000 [3:51:41<57:18:38,  7.29s/it, lr: 1.0e-04 loss: 8.180e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1707/30000 [3:51:41<57:18:38,  7.29s/it, lr: 1.0e-04 loss: 8.180e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1708/30000 [3:51:41<55:42:55,  7.09s/it, lr: 1.0e-04 loss: 8.180e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1708/30000 [3:51:41<55:42:55,  7.09s/it, lr: 1.0e-04 loss: 8.180e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1708/30000 [3:51:51<55:42:55,  7.09s/it, lr: 1.0e-04 loss: 2.567e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1708/30000 [3:51:51<55:42:55,  7.09s/it, lr: 1.0e-04 loss: 2.567e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1709/30000 [3:51:51<60:51:57,  7.75s/it, lr: 1.0e-04 loss: 2.567e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1709/30000 [3:51:51<60:51:57,  7.75s/it, lr: 1.0e-04 loss: 2.567e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1709/30000 [3:52:00<60:51:57,  7.75s/it, lr: 1.0e-04 loss: 8.360e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1709/30000 [3:52:00<60:51:57,  7.75s/it, lr: 1.0e-04 loss: 8.360e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1710/30000 [3:52:07<60:51:49,  7.75s/it, lr: 1.0e-04 loss: 1.372e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1710/30000 [3:52:07<60:51:49,  7.75s/it, lr: 1.0e-04 loss: 1.372e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1711/30000 [3:52:07<45:05:18,  5.74s/it, lr: 1.0e-04 loss: 1.372e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1711/30000 [3:52:07<45:05:18,  5.74s/it, lr: 1.0e-04 loss: 1.372e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1711/30000 [3:52:13<45:05:18,  5.74s/it, lr: 1.0e-04 loss: 8.810e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1711/30000 [3:52:13<45:05:18,  5.74s/it, lr: 1.0e-04 loss: 8.810e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1712/30000 [3:52:13<46:50:35,  5.96s/it, lr: 1.0e-04 loss: 8.810e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1712/30000 [3:52:13<46:50:35,  5.96s/it, lr: 1.0e-04 loss: 8.810e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1712/30000 [3:52:23<46:50:35,  5.96s/it, lr: 1.0e-04 loss: 5.102e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1712/30000 [3:52:23<46:50:35,  5.96s/it, lr: 1.0e-04 loss: 5.102e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1713/30000 [3:52:23<55:02:50,  7.01s/it, lr: 1.0e-04 loss: 5.102e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1713/30000 [3:52:23<55:02:50,  7.01s/it, lr: 1.0e-04 loss: 5.102e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1713/30000 [3:52:33<55:02:50,  7.01s/it, lr: 1.0e-04 loss: 6.246e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1713/30000 [3:52:33<55:02:50,  7.01s/it, lr: 1.0e-04 loss: 6.246e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1714/30000 [3:52:33<60:06:25,  7.65s/it, lr: 1.0e-04 loss: 6.246e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1714/30000 [3:52:33<60:06:25,  7.65s/it, lr: 1.0e-04 loss: 6.246e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1714/30000 [3:52:42<60:06:25,  7.65s/it, lr: 1.0e-04 loss: 4.969e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1714/30000 [3:52:42<60:06:25,  7.65s/it, lr: 1.0e-04 loss: 4.969e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1715/30000 [3:52:42<64:01:34,  8.15s/it, lr: 1.0e-04 loss: 4.969e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1715/30000 [3:52:42<64:01:34,  8.15s/it, lr: 1.0e-04 loss: 4.969e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1715/30000 [3:52:52<64:01:34,  8.15s/it, lr: 1.0e-04 loss: 7.496e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1715/30000 [3:52:52<64:01:34,  8.15s/it, lr: 1.0e-04 loss: 7.496e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1716/30000 [3:52:52<66:51:27,  8.51s/it, lr: 1.0e-04 loss: 7.496e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1716/30000 [3:52:52<66:51:27,  8.51s/it, lr: 1.0e-04 loss: 7.496e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1716/30000 [3:53:01<66:51:27,  8.51s/it, lr: 1.0e-04 loss: 7.463e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1716/30000 [3:53:01<66:51:27,  8.51s/it, lr: 1.0e-04 loss: 7.463e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1717/30000 [3:53:01<68:50:23,  8.76s/it, lr: 1.0e-04 loss: 7.463e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1717/30000 [3:53:01<68:50:23,  8.76s/it, lr: 1.0e-04 loss: 7.463e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1717/30000 [3:53:08<68:50:23,  8.76s/it, lr: 1.0e-04 loss: 9.000e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1717/30000 [3:53:08<68:50:23,  8.76s/it, lr: 1.0e-04 loss: 9.000e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1718/30000 [3:53:08<63:43:25,  8.11s/it, lr: 1.0e-04 loss: 9.000e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1718/30000 [3:53:08<63:43:25,  8.11s/it, lr: 1.0e-04 loss: 9.000e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1718/30000 [3:53:17<63:43:25,  8.11s/it, lr: 1.0e-04 loss: 3.637e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1718/30000 [3:53:17<63:43:25,  8.11s/it, lr: 1.0e-04 loss: 3.637e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1719/30000 [3:53:17<66:42:18,  8.49s/it, lr: 1.0e-04 loss: 3.637e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1719/30000 [3:53:17<66:42:18,  8.49s/it, lr: 1.0e-04 loss: 3.637e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1719/30000 [3:53:26<66:42:18,  8.49s/it, lr: 1.0e-04 loss: 5.339e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1719/30000 [3:53:26<66:42:18,  8.49s/it, lr: 1.0e-04 loss: 5.339e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1720/30000 [3:53:36<66:42:10,  8.49s/it, lr: 1.0e-04 loss: 4.492e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1720/30000 [3:53:36<66:42:10,  8.49s/it, lr: 1.0e-04 loss: 4.492e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1721/30000 [3:53:36<53:03:40,  6.75s/it, lr: 1.0e-04 loss: 4.492e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1721/30000 [3:53:36<53:03:40,  6.75s/it, lr: 1.0e-04 loss: 4.492e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1721/30000 [3:53:42<53:03:40,  6.75s/it, lr: 1.0e-04 loss: 1.042e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1721/30000 [3:53:42<53:03:40,  6.75s/it, lr: 1.0e-04 loss: 1.042e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1722/30000 [3:53:42<52:46:01,  6.72s/it, lr: 1.0e-04 loss: 1.042e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1722/30000 [3:53:42<52:46:01,  6.72s/it, lr: 1.0e-04 loss: 1.042e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1722/30000 [3:53:49<52:46:01,  6.72s/it, lr: 1.0e-04 loss: 7.548e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1722/30000 [3:53:49<52:46:01,  6.72s/it, lr: 1.0e-04 loss: 7.548e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1723/30000 [3:53:49<52:26:27,  6.68s/it, lr: 1.0e-04 loss: 7.548e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1723/30000 [3:53:49<52:26:27,  6.68s/it, lr: 1.0e-04 loss: 7.548e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1723/30000 [3:53:56<52:26:27,  6.68s/it, lr: 1.0e-04 loss: 1.128e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1723/30000 [3:53:56<52:26:27,  6.68s/it, lr: 1.0e-04 loss: 1.128e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1724/30000 [3:53:56<52:11:02,  6.64s/it, lr: 1.0e-04 loss: 1.128e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1724/30000 [3:53:56<52:11:02,  6.64s/it, lr: 1.0e-04 loss: 1.128e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1724/30000 [3:54:02<52:11:02,  6.64s/it, lr: 1.0e-04 loss: 7.404e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1724/30000 [3:54:02<52:11:02,  6.64s/it, lr: 1.0e-04 loss: 7.404e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1725/30000 [3:54:02<52:07:15,  6.64s/it, lr: 1.0e-04 loss: 7.404e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1725/30000 [3:54:02<52:07:15,  6.64s/it, lr: 1.0e-04 loss: 7.404e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1725/30000 [3:54:12<52:07:15,  6.64s/it, lr: 1.0e-04 loss: 6.344e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1725/30000 [3:54:12<52:07:15,  6.64s/it, lr: 1.0e-04 loss: 6.344e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1726/30000 [3:54:12<58:14:26,  7.42s/it, lr: 1.0e-04 loss: 6.344e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1726/30000 [3:54:12<58:14:26,  7.42s/it, lr: 1.0e-04 loss: 6.344e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1726/30000 [3:54:18<58:14:26,  7.42s/it, lr: 1.0e-04 loss: 8.684e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1726/30000 [3:54:18<58:14:26,  7.42s/it, lr: 1.0e-04 loss: 8.684e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1727/30000 [3:54:18<56:17:02,  7.17s/it, lr: 1.0e-04 loss: 8.684e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1727/30000 [3:54:18<56:17:02,  7.17s/it, lr: 1.0e-04 loss: 8.684e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1727/30000 [3:54:25<56:17:02,  7.17s/it, lr: 1.0e-04 loss: 7.473e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1727/30000 [3:54:25<56:17:02,  7.17s/it, lr: 1.0e-04 loss: 7.473e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1728/30000 [3:54:25<55:25:54,  7.06s/it, lr: 1.0e-04 loss: 7.473e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1728/30000 [3:54:25<55:25:54,  7.06s/it, lr: 1.0e-04 loss: 7.473e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1728/30000 [3:54:32<55:25:54,  7.06s/it, lr: 1.0e-04 loss: 4.854e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1728/30000 [3:54:32<55:25:54,  7.06s/it, lr: 1.0e-04 loss: 4.854e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1729/30000 [3:54:32<54:17:04,  6.91s/it, lr: 1.0e-04 loss: 4.854e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1729/30000 [3:54:32<54:17:04,  6.91s/it, lr: 1.0e-04 loss: 4.854e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1729/30000 [3:54:38<54:17:04,  6.91s/it, lr: 1.0e-04 loss: 8.935e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1729/30000 [3:54:38<54:17:04,  6.91s/it, lr: 1.0e-04 loss: 8.935e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1730/30000 [3:54:45<54:16:57,  6.91s/it, lr: 1.0e-04 loss: 7.029e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1730/30000 [3:54:45<54:16:57,  6.91s/it, lr: 1.0e-04 loss: 7.029e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1731/30000 [3:54:45<41:07:06,  5.24s/it, lr: 1.0e-04 loss: 7.029e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1731/30000 [3:54:45<41:07:06,  5.24s/it, lr: 1.0e-04 loss: 7.029e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1731/30000 [3:54:54<41:07:06,  5.24s/it, lr: 1.0e-04 loss: 1.003e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1731/30000 [3:54:54<41:07:06,  5.24s/it, lr: 1.0e-04 loss: 1.003e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1732/30000 [3:54:54<48:55:12,  6.23s/it, lr: 1.0e-04 loss: 1.003e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1732/30000 [3:54:54<48:55:12,  6.23s/it, lr: 1.0e-04 loss: 1.003e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1732/30000 [3:55:00<48:55:12,  6.23s/it, lr: 1.0e-04 loss: 2.187e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1732/30000 [3:55:00<48:55:12,  6.23s/it, lr: 1.0e-04 loss: 2.187e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1733/30000 [3:55:00<49:37:41,  6.32s/it, lr: 1.0e-04 loss: 2.187e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1733/30000 [3:55:00<49:37:41,  6.32s/it, lr: 1.0e-04 loss: 2.187e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1733/30000 [3:55:10<49:37:41,  6.32s/it, lr: 1.0e-04 loss: 8.111e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1733/30000 [3:55:10<49:37:41,  6.32s/it, lr: 1.0e-04 loss: 8.111e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1734/30000 [3:55:10<56:10:32,  7.15s/it, lr: 1.0e-04 loss: 8.111e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1734/30000 [3:55:10<56:10:32,  7.15s/it, lr: 1.0e-04 loss: 8.111e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1734/30000 [3:55:17<56:10:32,  7.15s/it, lr: 1.0e-04 loss: 6.614e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1734/30000 [3:55:17<56:10:32,  7.15s/it, lr: 1.0e-04 loss: 6.614e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1735/30000 [3:55:17<55:28:40,  7.07s/it, lr: 1.0e-04 loss: 6.614e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1735/30000 [3:55:17<55:28:40,  7.07s/it, lr: 1.0e-04 loss: 6.614e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1735/30000 [3:55:23<55:28:40,  7.07s/it, lr: 1.0e-04 loss: 7.378e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1735/30000 [3:55:23<55:28:40,  7.07s/it, lr: 1.0e-04 loss: 7.378e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1736/30000 [3:55:23<54:25:29,  6.93s/it, lr: 1.0e-04 loss: 7.378e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1736/30000 [3:55:23<54:25:29,  6.93s/it, lr: 1.0e-04 loss: 7.378e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1736/30000 [3:55:30<54:25:29,  6.93s/it, lr: 1.0e-04 loss: 1.641e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1736/30000 [3:55:30<54:25:29,  6.93s/it, lr: 1.0e-04 loss: 1.641e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1737/30000 [3:55:30<53:38:02,  6.83s/it, lr: 1.0e-04 loss: 1.641e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1737/30000 [3:55:30<53:38:02,  6.83s/it, lr: 1.0e-04 loss: 1.641e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1737/30000 [3:55:36<53:38:02,  6.83s/it, lr: 1.0e-04 loss: 1.036e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1737/30000 [3:55:36<53:38:02,  6.83s/it, lr: 1.0e-04 loss: 1.036e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1738/30000 [3:55:36<53:09:31,  6.77s/it, lr: 1.0e-04 loss: 1.036e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1738/30000 [3:55:36<53:09:31,  6.77s/it, lr: 1.0e-04 loss: 1.036e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1738/30000 [3:55:46<53:09:31,  6.77s/it, lr: 1.0e-04 loss: 1.813e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1738/30000 [3:55:46<53:09:31,  6.77s/it, lr: 1.0e-04 loss: 1.813e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1739/30000 [3:55:46<59:07:37,  7.53s/it, lr: 1.0e-04 loss: 1.813e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1739/30000 [3:55:46<59:07:37,  7.53s/it, lr: 1.0e-04 loss: 1.813e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1739/30000 [3:55:55<59:07:37,  7.53s/it, lr: 1.0e-04 loss: 5.093e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1739/30000 [3:55:55<59:07:37,  7.53s/it, lr: 1.0e-04 loss: 5.093e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1740/30000 [3:56:05<59:07:30,  7.53s/it, lr: 1.0e-04 loss: 6.751e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1740/30000 [3:56:05<59:07:30,  7.53s/it, lr: 1.0e-04 loss: 6.751e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1741/30000 [3:56:05<48:49:21,  6.22s/it, lr: 1.0e-04 loss: 6.751e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1741/30000 [3:56:05<48:49:21,  6.22s/it, lr: 1.0e-04 loss: 6.751e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1741/30000 [3:56:11<48:49:21,  6.22s/it, lr: 1.0e-04 loss: 7.060e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1741/30000 [3:56:11<48:49:21,  6.22s/it, lr: 1.0e-04 loss: 7.060e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1742/30000 [3:56:11<49:39:02,  6.33s/it, lr: 1.0e-04 loss: 7.060e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1742/30000 [3:56:11<49:39:02,  6.33s/it, lr: 1.0e-04 loss: 7.060e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1742/30000 [3:56:21<49:39:02,  6.33s/it, lr: 1.0e-04 loss: 4.235e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1742/30000 [3:56:21<49:39:02,  6.33s/it, lr: 1.0e-04 loss: 4.235e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1743/30000 [3:56:21<56:15:28,  7.17s/it, lr: 1.0e-04 loss: 4.235e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1743/30000 [3:56:21<56:15:28,  7.17s/it, lr: 1.0e-04 loss: 4.235e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1743/30000 [3:56:27<56:15:28,  7.17s/it, lr: 1.0e-04 loss: 8.369e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1743/30000 [3:56:27<56:15:28,  7.17s/it, lr: 1.0e-04 loss: 8.369e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1744/30000 [3:56:27<55:05:41,  7.02s/it, lr: 1.0e-04 loss: 8.369e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1744/30000 [3:56:27<55:05:41,  7.02s/it, lr: 1.0e-04 loss: 8.369e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1744/30000 [3:56:34<55:05:41,  7.02s/it, lr: 1.0e-04 loss: 9.966e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1744/30000 [3:56:34<55:05:41,  7.02s/it, lr: 1.0e-04 loss: 9.966e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1745/30000 [3:56:34<54:12:03,  6.91s/it, lr: 1.0e-04 loss: 9.966e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1745/30000 [3:56:34<54:12:03,  6.91s/it, lr: 1.0e-04 loss: 9.966e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1745/30000 [3:56:43<54:12:03,  6.91s/it, lr: 1.0e-04 loss: 4.877e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1745/30000 [3:56:43<54:12:03,  6.91s/it, lr: 1.0e-04 loss: 4.877e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1746/30000 [3:56:43<59:37:32,  7.60s/it, lr: 1.0e-04 loss: 4.877e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1746/30000 [3:56:43<59:37:32,  7.60s/it, lr: 1.0e-04 loss: 4.877e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1746/30000 [3:56:50<59:37:32,  7.60s/it, lr: 1.0e-04 loss: 1.130e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1746/30000 [3:56:50<59:37:32,  7.60s/it, lr: 1.0e-04 loss: 1.130e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1747/30000 [3:56:50<57:27:00,  7.32s/it, lr: 1.0e-04 loss: 1.130e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1747/30000 [3:56:50<57:27:00,  7.32s/it, lr: 1.0e-04 loss: 1.130e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1747/30000 [3:56:59<57:27:00,  7.32s/it, lr: 1.0e-04 loss: 7.140e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1747/30000 [3:56:59<57:27:00,  7.32s/it, lr: 1.0e-04 loss: 7.140e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1748/30000 [3:56:59<62:06:31,  7.91s/it, lr: 1.0e-04 loss: 7.140e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1748/30000 [3:56:59<62:06:31,  7.91s/it, lr: 1.0e-04 loss: 7.140e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1748/30000 [3:57:06<62:06:31,  7.91s/it, lr: 1.0e-04 loss: 9.651e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1748/30000 [3:57:06<62:06:31,  7.91s/it, lr: 1.0e-04 loss: 9.651e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1749/30000 [3:57:06<59:01:22,  7.52s/it, lr: 1.0e-04 loss: 9.651e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1749/30000 [3:57:06<59:01:22,  7.52s/it, lr: 1.0e-04 loss: 9.651e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1749/30000 [3:57:13<59:01:22,  7.52s/it, lr: 1.0e-04 loss: 7.302e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1749/30000 [3:57:13<59:01:22,  7.52s/it, lr: 1.0e-04 loss: 7.302e-02]
+Saving at step 1750
+Saved checkpoint to /home/featurize/ai-toolkit/output/Xiang_wan_2_2_lora/Xiang_wan_2_2_lora_000001750.safetensors
+Saved optimizer to /home/featurize/ai-toolkit/output/Xiang_wan_2_2_lora/optimizer.pt
+
Xiang_wan_2_2_lora:   6%|##2                                    | 1750/30000 [3:57:21<59:01:14,  7.52s/it, lr: 1.0e-04 loss: 8.958e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1750/30000 [3:57:21<59:01:14,  7.52s/it, lr: 1.0e-04 loss: 8.958e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1751/30000 [3:57:21<46:52:17,  5.97s/it, lr: 1.0e-04 loss: 8.958e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1751/30000 [3:57:21<46:52:17,  5.97s/it, lr: 1.0e-04 loss: 8.958e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1751/30000 [3:57:28<46:52:17,  5.97s/it, lr: 1.0e-04 loss: 7.792e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1751/30000 [3:57:28<46:52:17,  5.97s/it, lr: 1.0e-04 loss: 7.792e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1752/30000 [3:57:28<48:17:53,  6.16s/it, lr: 1.0e-04 loss: 7.792e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1752/30000 [3:57:28<48:17:53,  6.16s/it, lr: 1.0e-04 loss: 7.792e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1752/30000 [3:57:34<48:17:53,  6.16s/it, lr: 1.0e-04 loss: 6.932e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1752/30000 [3:57:34<48:17:53,  6.16s/it, lr: 1.0e-04 loss: 6.932e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1753/30000 [3:57:34<49:17:57,  6.28s/it, lr: 1.0e-04 loss: 6.932e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1753/30000 [3:57:34<49:17:57,  6.28s/it, lr: 1.0e-04 loss: 6.932e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1753/30000 [3:57:44<49:17:57,  6.28s/it, lr: 1.0e-04 loss: 7.432e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1753/30000 [3:57:44<49:17:57,  6.28s/it, lr: 1.0e-04 loss: 7.432e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1754/30000 [3:57:44<55:53:23,  7.12s/it, lr: 1.0e-04 loss: 7.432e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1754/30000 [3:57:44<55:53:23,  7.12s/it, lr: 1.0e-04 loss: 7.432e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1754/30000 [3:57:50<55:53:23,  7.12s/it, lr: 1.0e-04 loss: 7.186e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1754/30000 [3:57:50<55:53:23,  7.12s/it, lr: 1.0e-04 loss: 7.186e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1755/30000 [3:57:50<54:46:40,  6.98s/it, lr: 1.0e-04 loss: 7.186e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1755/30000 [3:57:50<54:46:40,  6.98s/it, lr: 1.0e-04 loss: 7.186e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1755/30000 [3:58:00<54:46:40,  6.98s/it, lr: 1.0e-04 loss: 8.028e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1755/30000 [3:58:00<54:46:40,  6.98s/it, lr: 1.0e-04 loss: 8.028e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1756/30000 [3:58:00<60:04:54,  7.66s/it, lr: 1.0e-04 loss: 8.028e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1756/30000 [3:58:00<60:04:54,  7.66s/it, lr: 1.0e-04 loss: 8.028e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1756/30000 [3:58:06<60:04:54,  7.66s/it, lr: 1.0e-04 loss: 6.628e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1756/30000 [3:58:06<60:04:54,  7.66s/it, lr: 1.0e-04 loss: 6.628e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1757/30000 [3:58:06<57:49:08,  7.37s/it, lr: 1.0e-04 loss: 6.628e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1757/30000 [3:58:06<57:49:08,  7.37s/it, lr: 1.0e-04 loss: 6.628e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1757/30000 [3:58:16<57:49:08,  7.37s/it, lr: 1.0e-04 loss: 6.092e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1757/30000 [3:58:16<57:49:08,  7.37s/it, lr: 1.0e-04 loss: 6.092e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1758/30000 [3:58:16<62:48:30,  8.01s/it, lr: 1.0e-04 loss: 6.092e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1758/30000 [3:58:16<62:48:30,  8.01s/it, lr: 1.0e-04 loss: 6.092e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1758/30000 [3:58:23<62:48:30,  8.01s/it, lr: 1.0e-04 loss: 7.178e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1758/30000 [3:58:23<62:48:30,  8.01s/it, lr: 1.0e-04 loss: 7.178e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1759/30000 [3:58:23<59:41:13,  7.61s/it, lr: 1.0e-04 loss: 7.178e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1759/30000 [3:58:23<59:41:13,  7.61s/it, lr: 1.0e-04 loss: 7.178e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1759/30000 [3:58:32<59:41:13,  7.61s/it, lr: 1.0e-04 loss: 5.482e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1759/30000 [3:58:32<59:41:13,  7.61s/it, lr: 1.0e-04 loss: 5.482e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1760/30000 [3:58:41<59:41:06,  7.61s/it, lr: 1.0e-04 loss: 3.611e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1760/30000 [3:58:41<59:41:06,  7.61s/it, lr: 1.0e-04 loss: 3.611e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1761/30000 [3:58:41<49:13:42,  6.28s/it, lr: 1.0e-04 loss: 3.611e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1761/30000 [3:58:41<49:13:42,  6.28s/it, lr: 1.0e-04 loss: 3.611e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1761/30000 [3:58:51<49:13:42,  6.28s/it, lr: 1.0e-04 loss: 8.168e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1761/30000 [3:58:51<49:13:42,  6.28s/it, lr: 1.0e-04 loss: 8.168e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1762/30000 [3:58:51<55:25:43,  7.07s/it, lr: 1.0e-04 loss: 8.168e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1762/30000 [3:58:51<55:25:43,  7.07s/it, lr: 1.0e-04 loss: 8.168e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1762/30000 [3:58:58<55:25:43,  7.07s/it, lr: 1.0e-04 loss: 5.336e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1762/30000 [3:58:58<55:25:43,  7.07s/it, lr: 1.0e-04 loss: 5.336e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1763/30000 [3:58:58<54:52:09,  7.00s/it, lr: 1.0e-04 loss: 5.336e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1763/30000 [3:58:58<54:52:09,  7.00s/it, lr: 1.0e-04 loss: 5.336e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1763/30000 [3:59:07<54:52:09,  7.00s/it, lr: 1.0e-04 loss: 4.783e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1763/30000 [3:59:07<54:52:09,  7.00s/it, lr: 1.0e-04 loss: 4.783e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1764/30000 [3:59:07<59:47:14,  7.62s/it, lr: 1.0e-04 loss: 4.783e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1764/30000 [3:59:07<59:47:14,  7.62s/it, lr: 1.0e-04 loss: 4.783e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1764/30000 [3:59:16<59:47:14,  7.62s/it, lr: 1.0e-04 loss: 3.785e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1764/30000 [3:59:16<59:47:14,  7.62s/it, lr: 1.0e-04 loss: 3.785e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1765/30000 [3:59:16<63:58:21,  8.16s/it, lr: 1.0e-04 loss: 3.785e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1765/30000 [3:59:16<63:58:21,  8.16s/it, lr: 1.0e-04 loss: 3.785e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1765/30000 [3:59:26<63:58:21,  8.16s/it, lr: 1.0e-04 loss: 1.087e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1765/30000 [3:59:26<63:58:21,  8.16s/it, lr: 1.0e-04 loss: 1.087e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1766/30000 [3:59:26<66:36:43,  8.49s/it, lr: 1.0e-04 loss: 1.087e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1766/30000 [3:59:26<66:36:43,  8.49s/it, lr: 1.0e-04 loss: 1.087e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1766/30000 [3:59:35<66:36:43,  8.49s/it, lr: 1.0e-04 loss: 4.761e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1766/30000 [3:59:35<66:36:43,  8.49s/it, lr: 1.0e-04 loss: 4.761e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1767/30000 [3:59:35<68:37:59,  8.75s/it, lr: 1.0e-04 loss: 4.761e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1767/30000 [3:59:35<68:37:59,  8.75s/it, lr: 1.0e-04 loss: 4.761e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1767/30000 [3:59:42<68:37:59,  8.75s/it, lr: 1.0e-04 loss: 8.649e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1767/30000 [3:59:42<68:37:59,  8.75s/it, lr: 1.0e-04 loss: 8.649e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1768/30000 [3:59:42<63:44:07,  8.13s/it, lr: 1.0e-04 loss: 8.649e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1768/30000 [3:59:42<63:44:07,  8.13s/it, lr: 1.0e-04 loss: 8.649e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1768/30000 [3:59:48<63:44:07,  8.13s/it, lr: 1.0e-04 loss: 1.428e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1768/30000 [3:59:48<63:44:07,  8.13s/it, lr: 1.0e-04 loss: 1.428e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1769/30000 [3:59:48<60:06:34,  7.67s/it, lr: 1.0e-04 loss: 1.428e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1769/30000 [3:59:48<60:06:34,  7.67s/it, lr: 1.0e-04 loss: 1.428e-01]
Xiang_wan_2_2_lora:   6%|##2                                    | 1769/30000 [3:59:55<60:06:34,  7.67s/it, lr: 1.0e-04 loss: 8.073e-02]
Xiang_wan_2_2_lora:   6%|##2                                    | 1769/30000 [3:59:55<60:06:34,  7.67s/it, lr: 1.0e-04 loss: 8.073e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1770/30000 [4:00:04<60:06:27,  7.67s/it, lr: 1.0e-04 loss: 3.902e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1770/30000 [4:00:04<60:06:27,  7.67s/it, lr: 1.0e-04 loss: 3.902e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1771/30000 [4:00:04<49:19:33,  6.29s/it, lr: 1.0e-04 loss: 3.902e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1771/30000 [4:00:04<49:19:33,  6.29s/it, lr: 1.0e-04 loss: 3.902e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1771/30000 [4:00:14<49:19:33,  6.29s/it, lr: 1.0e-04 loss: 9.957e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1771/30000 [4:00:14<49:19:33,  6.29s/it, lr: 1.0e-04 loss: 9.957e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1772/30000 [4:00:14<55:17:10,  7.05s/it, lr: 1.0e-04 loss: 9.957e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1772/30000 [4:00:14<55:17:10,  7.05s/it, lr: 1.0e-04 loss: 9.957e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1772/30000 [4:00:20<55:17:10,  7.05s/it, lr: 1.0e-04 loss: 6.547e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1772/30000 [4:00:20<55:17:10,  7.05s/it, lr: 1.0e-04 loss: 6.547e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1773/30000 [4:00:20<54:52:33,  7.00s/it, lr: 1.0e-04 loss: 6.547e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1773/30000 [4:00:20<54:52:33,  7.00s/it, lr: 1.0e-04 loss: 6.547e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1773/30000 [4:00:30<54:52:33,  7.00s/it, lr: 1.0e-04 loss: 1.029e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1773/30000 [4:00:30<54:52:33,  7.00s/it, lr: 1.0e-04 loss: 1.029e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1774/30000 [4:00:30<59:56:31,  7.65s/it, lr: 1.0e-04 loss: 1.029e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1774/30000 [4:00:30<59:56:31,  7.65s/it, lr: 1.0e-04 loss: 1.029e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1774/30000 [4:00:39<59:56:31,  7.65s/it, lr: 1.0e-04 loss: 9.465e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1774/30000 [4:00:39<59:56:31,  7.65s/it, lr: 1.0e-04 loss: 9.465e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1775/30000 [4:00:39<63:44:21,  8.13s/it, lr: 1.0e-04 loss: 9.465e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1775/30000 [4:00:39<63:44:21,  8.13s/it, lr: 1.0e-04 loss: 9.465e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1775/30000 [4:00:49<63:44:21,  8.13s/it, lr: 1.0e-04 loss: 7.415e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1775/30000 [4:00:49<63:44:21,  8.13s/it, lr: 1.0e-04 loss: 7.415e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1776/30000 [4:00:49<66:32:06,  8.49s/it, lr: 1.0e-04 loss: 7.415e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1776/30000 [4:00:49<66:32:06,  8.49s/it, lr: 1.0e-04 loss: 7.415e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1776/30000 [4:00:58<66:32:06,  8.49s/it, lr: 1.0e-04 loss: 5.271e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1776/30000 [4:00:58<66:32:06,  8.49s/it, lr: 1.0e-04 loss: 5.271e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1777/30000 [4:00:58<68:30:48,  8.74s/it, lr: 1.0e-04 loss: 5.271e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1777/30000 [4:00:58<68:30:48,  8.74s/it, lr: 1.0e-04 loss: 5.271e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1777/30000 [4:01:05<68:30:48,  8.74s/it, lr: 1.0e-04 loss: 7.019e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1777/30000 [4:01:05<68:30:48,  8.74s/it, lr: 1.0e-04 loss: 7.019e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1778/30000 [4:01:05<63:33:43,  8.11s/it, lr: 1.0e-04 loss: 7.019e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1778/30000 [4:01:05<63:33:43,  8.11s/it, lr: 1.0e-04 loss: 7.019e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1778/30000 [4:01:14<63:33:43,  8.11s/it, lr: 1.0e-04 loss: 3.608e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1778/30000 [4:01:14<63:33:43,  8.11s/it, lr: 1.0e-04 loss: 3.608e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1779/30000 [4:01:14<66:24:37,  8.47s/it, lr: 1.0e-04 loss: 3.608e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1779/30000 [4:01:14<66:24:37,  8.47s/it, lr: 1.0e-04 loss: 3.608e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1779/30000 [4:01:21<66:24:37,  8.47s/it, lr: 1.0e-04 loss: 7.666e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1779/30000 [4:01:21<66:24:37,  8.47s/it, lr: 1.0e-04 loss: 7.666e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1780/30000 [4:01:27<66:24:29,  8.47s/it, lr: 1.0e-04 loss: 7.678e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1780/30000 [4:01:27<66:24:29,  8.47s/it, lr: 1.0e-04 loss: 7.678e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1781/30000 [4:01:27<47:55:36,  6.11s/it, lr: 1.0e-04 loss: 7.678e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1781/30000 [4:01:27<47:55:36,  6.11s/it, lr: 1.0e-04 loss: 7.678e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1781/30000 [4:01:34<47:55:36,  6.11s/it, lr: 1.0e-04 loss: 7.414e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1781/30000 [4:01:34<47:55:36,  6.11s/it, lr: 1.0e-04 loss: 7.414e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1782/30000 [4:01:34<48:53:55,  6.24s/it, lr: 1.0e-04 loss: 7.414e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1782/30000 [4:01:34<48:53:55,  6.24s/it, lr: 1.0e-04 loss: 7.414e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1782/30000 [4:01:43<48:53:55,  6.24s/it, lr: 1.0e-04 loss: 4.159e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1782/30000 [4:01:43<48:53:55,  6.24s/it, lr: 1.0e-04 loss: 4.159e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1783/30000 [4:01:43<55:11:08,  7.04s/it, lr: 1.0e-04 loss: 4.159e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1783/30000 [4:01:43<55:11:08,  7.04s/it, lr: 1.0e-04 loss: 4.159e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1783/30000 [4:01:50<55:11:08,  7.04s/it, lr: 1.0e-04 loss: 8.534e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1783/30000 [4:01:50<55:11:08,  7.04s/it, lr: 1.0e-04 loss: 8.534e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1784/30000 [4:01:50<54:13:00,  6.92s/it, lr: 1.0e-04 loss: 8.534e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1784/30000 [4:01:50<54:13:00,  6.92s/it, lr: 1.0e-04 loss: 8.534e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1784/30000 [4:01:59<54:13:00,  6.92s/it, lr: 1.0e-04 loss: 7.054e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1784/30000 [4:01:59<54:13:00,  6.92s/it, lr: 1.0e-04 loss: 7.054e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1785/30000 [4:01:59<59:26:28,  7.58s/it, lr: 1.0e-04 loss: 7.054e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1785/30000 [4:01:59<59:26:28,  7.58s/it, lr: 1.0e-04 loss: 7.054e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1785/30000 [4:02:08<59:26:28,  7.58s/it, lr: 1.0e-04 loss: 6.623e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1785/30000 [4:02:08<59:26:28,  7.58s/it, lr: 1.0e-04 loss: 6.623e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1786/30000 [4:02:08<63:20:25,  8.08s/it, lr: 1.0e-04 loss: 6.623e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1786/30000 [4:02:08<63:20:25,  8.08s/it, lr: 1.0e-04 loss: 6.623e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1786/30000 [4:02:18<63:20:25,  8.08s/it, lr: 1.0e-04 loss: 8.690e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1786/30000 [4:02:18<63:20:25,  8.08s/it, lr: 1.0e-04 loss: 8.690e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1787/30000 [4:02:18<66:24:30,  8.47s/it, lr: 1.0e-04 loss: 8.690e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1787/30000 [4:02:18<66:24:30,  8.47s/it, lr: 1.0e-04 loss: 8.690e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1787/30000 [4:02:27<66:24:30,  8.47s/it, lr: 1.0e-04 loss: 9.378e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1787/30000 [4:02:27<66:24:30,  8.47s/it, lr: 1.0e-04 loss: 9.378e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1788/30000 [4:02:27<68:52:54,  8.79s/it, lr: 1.0e-04 loss: 9.378e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1788/30000 [4:02:27<68:52:54,  8.79s/it, lr: 1.0e-04 loss: 9.378e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1788/30000 [4:02:37<68:52:54,  8.79s/it, lr: 1.0e-04 loss: 4.772e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1788/30000 [4:02:37<68:52:54,  8.79s/it, lr: 1.0e-04 loss: 4.772e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1789/30000 [4:02:37<70:07:31,  8.95s/it, lr: 1.0e-04 loss: 4.772e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1789/30000 [4:02:37<70:07:31,  8.95s/it, lr: 1.0e-04 loss: 4.772e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1789/30000 [4:02:46<70:07:31,  8.95s/it, lr: 1.0e-04 loss: 6.021e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1789/30000 [4:02:46<70:07:31,  8.95s/it, lr: 1.0e-04 loss: 6.021e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1790/30000 [4:02:56<70:07:22,  8.95s/it, lr: 1.0e-04 loss: 1.189e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1790/30000 [4:02:56<70:07:22,  8.95s/it, lr: 1.0e-04 loss: 1.189e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1791/30000 [4:02:56<54:56:44,  7.01s/it, lr: 1.0e-04 loss: 1.189e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1791/30000 [4:02:56<54:56:44,  7.01s/it, lr: 1.0e-04 loss: 1.189e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1791/30000 [4:03:05<54:56:44,  7.01s/it, lr: 1.0e-04 loss: 6.362e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1791/30000 [4:03:05<54:56:44,  7.01s/it, lr: 1.0e-04 loss: 6.362e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1792/30000 [4:03:05<59:28:21,  7.59s/it, lr: 1.0e-04 loss: 6.362e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1792/30000 [4:03:05<59:28:21,  7.59s/it, lr: 1.0e-04 loss: 6.362e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1792/30000 [4:03:14<59:28:21,  7.59s/it, lr: 1.0e-04 loss: 2.063e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1792/30000 [4:03:14<59:28:21,  7.59s/it, lr: 1.0e-04 loss: 2.063e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1793/30000 [4:03:14<63:05:08,  8.05s/it, lr: 1.0e-04 loss: 2.063e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1793/30000 [4:03:14<63:05:08,  8.05s/it, lr: 1.0e-04 loss: 2.063e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1793/30000 [4:03:21<63:05:08,  8.05s/it, lr: 1.0e-04 loss: 8.746e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1793/30000 [4:03:21<63:05:08,  8.05s/it, lr: 1.0e-04 loss: 8.746e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1794/30000 [4:03:21<60:01:04,  7.66s/it, lr: 1.0e-04 loss: 8.746e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1794/30000 [4:03:21<60:01:04,  7.66s/it, lr: 1.0e-04 loss: 8.746e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1794/30000 [4:03:30<60:01:04,  7.66s/it, lr: 1.0e-04 loss: 2.090e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1794/30000 [4:03:30<60:01:04,  7.66s/it, lr: 1.0e-04 loss: 2.090e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1795/30000 [4:03:30<64:13:16,  8.20s/it, lr: 1.0e-04 loss: 2.090e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1795/30000 [4:03:30<64:13:16,  8.20s/it, lr: 1.0e-04 loss: 2.090e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1795/30000 [4:03:37<64:13:16,  8.20s/it, lr: 1.0e-04 loss: 8.476e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1795/30000 [4:03:37<64:13:16,  8.20s/it, lr: 1.0e-04 loss: 8.476e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1796/30000 [4:03:37<60:32:43,  7.73s/it, lr: 1.0e-04 loss: 8.476e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1796/30000 [4:03:37<60:32:43,  7.73s/it, lr: 1.0e-04 loss: 8.476e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1796/30000 [4:03:46<60:32:43,  7.73s/it, lr: 1.0e-04 loss: 8.030e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1796/30000 [4:03:46<60:32:43,  7.73s/it, lr: 1.0e-04 loss: 8.030e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1797/30000 [4:03:46<64:08:42,  8.19s/it, lr: 1.0e-04 loss: 8.030e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1797/30000 [4:03:46<64:08:42,  8.19s/it, lr: 1.0e-04 loss: 8.030e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1797/30000 [4:03:53<64:08:42,  8.19s/it, lr: 1.0e-04 loss: 8.593e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1797/30000 [4:03:53<64:08:42,  8.19s/it, lr: 1.0e-04 loss: 8.593e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1798/30000 [4:03:53<60:32:02,  7.73s/it, lr: 1.0e-04 loss: 8.593e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1798/30000 [4:03:53<60:32:02,  7.73s/it, lr: 1.0e-04 loss: 8.593e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1798/30000 [4:03:59<60:32:02,  7.73s/it, lr: 1.0e-04 loss: 8.942e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1798/30000 [4:03:59<60:32:02,  7.73s/it, lr: 1.0e-04 loss: 8.942e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1799/30000 [4:03:59<57:51:33,  7.39s/it, lr: 1.0e-04 loss: 8.942e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1799/30000 [4:03:59<57:51:33,  7.39s/it, lr: 1.0e-04 loss: 8.942e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1799/30000 [4:04:09<57:51:33,  7.39s/it, lr: 1.0e-04 loss: 5.247e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1799/30000 [4:04:09<57:51:33,  7.39s/it, lr: 1.0e-04 loss: 5.247e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1800/30000 [4:04:18<57:51:25,  7.39s/it, lr: 1.0e-04 loss: 1.671e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1800/30000 [4:04:18<57:51:25,  7.39s/it, lr: 1.0e-04 loss: 1.671e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1801/30000 [4:04:18<48:10:15,  6.15s/it, lr: 1.0e-04 loss: 1.671e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1801/30000 [4:04:18<48:10:15,  6.15s/it, lr: 1.0e-04 loss: 1.671e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1801/30000 [4:04:25<48:10:15,  6.15s/it, lr: 1.0e-04 loss: 8.711e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1801/30000 [4:04:25<48:10:15,  6.15s/it, lr: 1.0e-04 loss: 8.711e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1802/30000 [4:04:25<48:56:33,  6.25s/it, lr: 1.0e-04 loss: 8.711e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1802/30000 [4:04:25<48:56:33,  6.25s/it, lr: 1.0e-04 loss: 8.711e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1802/30000 [4:04:32<48:56:33,  6.25s/it, lr: 1.0e-04 loss: 5.481e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1802/30000 [4:04:32<48:56:33,  6.25s/it, lr: 1.0e-04 loss: 5.481e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1803/30000 [4:04:32<50:05:43,  6.40s/it, lr: 1.0e-04 loss: 5.481e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1803/30000 [4:04:32<50:05:43,  6.40s/it, lr: 1.0e-04 loss: 5.481e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1803/30000 [4:04:38<50:05:43,  6.40s/it, lr: 1.0e-04 loss: 1.135e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1803/30000 [4:04:38<50:05:43,  6.40s/it, lr: 1.0e-04 loss: 1.135e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1804/30000 [4:04:38<50:29:52,  6.45s/it, lr: 1.0e-04 loss: 1.135e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1804/30000 [4:04:38<50:29:52,  6.45s/it, lr: 1.0e-04 loss: 1.135e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1804/30000 [4:04:45<50:29:52,  6.45s/it, lr: 1.0e-04 loss: 6.362e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1804/30000 [4:04:45<50:29:52,  6.45s/it, lr: 1.0e-04 loss: 6.362e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1805/30000 [4:04:45<50:47:27,  6.49s/it, lr: 1.0e-04 loss: 6.362e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1805/30000 [4:04:45<50:47:27,  6.49s/it, lr: 1.0e-04 loss: 6.362e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1805/30000 [4:04:54<50:47:27,  6.49s/it, lr: 1.0e-04 loss: 9.350e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1805/30000 [4:04:54<50:47:27,  6.49s/it, lr: 1.0e-04 loss: 9.350e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1806/30000 [4:04:54<57:06:17,  7.29s/it, lr: 1.0e-04 loss: 9.350e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1806/30000 [4:04:54<57:06:17,  7.29s/it, lr: 1.0e-04 loss: 9.350e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1806/30000 [4:05:03<57:06:17,  7.29s/it, lr: 1.0e-04 loss: 5.089e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1806/30000 [4:05:03<57:06:17,  7.29s/it, lr: 1.0e-04 loss: 5.089e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1807/30000 [4:05:03<61:44:04,  7.88s/it, lr: 1.0e-04 loss: 5.089e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1807/30000 [4:05:03<61:44:04,  7.88s/it, lr: 1.0e-04 loss: 5.089e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1807/30000 [4:05:13<61:44:04,  7.88s/it, lr: 1.0e-04 loss: 5.641e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1807/30000 [4:05:13<61:44:04,  7.88s/it, lr: 1.0e-04 loss: 5.641e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1808/30000 [4:05:13<64:55:09,  8.29s/it, lr: 1.0e-04 loss: 5.641e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1808/30000 [4:05:13<64:55:09,  8.29s/it, lr: 1.0e-04 loss: 5.641e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1808/30000 [4:05:22<64:55:09,  8.29s/it, lr: 1.0e-04 loss: 1.901e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1808/30000 [4:05:22<64:55:09,  8.29s/it, lr: 1.0e-04 loss: 1.901e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1809/30000 [4:05:22<67:27:03,  8.61s/it, lr: 1.0e-04 loss: 1.901e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1809/30000 [4:05:22<67:27:03,  8.61s/it, lr: 1.0e-04 loss: 1.901e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1809/30000 [4:05:29<67:27:03,  8.61s/it, lr: 1.0e-04 loss: 7.524e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1809/30000 [4:05:29<67:27:03,  8.61s/it, lr: 1.0e-04 loss: 7.524e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1810/30000 [4:05:36<67:26:54,  8.61s/it, lr: 1.0e-04 loss: 1.102e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1810/30000 [4:05:36<67:26:54,  8.61s/it, lr: 1.0e-04 loss: 1.102e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1811/30000 [4:05:36<48:27:27,  6.19s/it, lr: 1.0e-04 loss: 1.102e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1811/30000 [4:05:36<48:27:27,  6.19s/it, lr: 1.0e-04 loss: 1.102e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1811/30000 [4:05:45<48:27:27,  6.19s/it, lr: 1.0e-04 loss: 5.402e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1811/30000 [4:05:45<48:27:27,  6.19s/it, lr: 1.0e-04 loss: 5.402e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1812/30000 [4:05:45<54:29:10,  6.96s/it, lr: 1.0e-04 loss: 5.402e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1812/30000 [4:05:45<54:29:10,  6.96s/it, lr: 1.0e-04 loss: 5.402e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1812/30000 [4:05:51<54:29:10,  6.96s/it, lr: 1.0e-04 loss: 6.431e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1812/30000 [4:05:51<54:29:10,  6.96s/it, lr: 1.0e-04 loss: 6.431e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1813/30000 [4:05:51<53:45:12,  6.87s/it, lr: 1.0e-04 loss: 6.431e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1813/30000 [4:05:51<53:45:12,  6.87s/it, lr: 1.0e-04 loss: 6.431e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1813/30000 [4:05:58<53:45:12,  6.87s/it, lr: 1.0e-04 loss: 1.005e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1813/30000 [4:05:58<53:45:12,  6.87s/it, lr: 1.0e-04 loss: 1.005e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1814/30000 [4:05:58<53:08:15,  6.79s/it, lr: 1.0e-04 loss: 1.005e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1814/30000 [4:05:58<53:08:15,  6.79s/it, lr: 1.0e-04 loss: 1.005e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1814/30000 [4:06:07<53:08:15,  6.79s/it, lr: 1.0e-04 loss: 8.372e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1814/30000 [4:06:07<53:08:15,  6.79s/it, lr: 1.0e-04 loss: 8.372e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1815/30000 [4:06:07<58:45:40,  7.51s/it, lr: 1.0e-04 loss: 8.372e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1815/30000 [4:06:07<58:45:40,  7.51s/it, lr: 1.0e-04 loss: 8.372e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1815/30000 [4:06:14<58:45:40,  7.51s/it, lr: 1.0e-04 loss: 1.172e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1815/30000 [4:06:14<58:45:40,  7.51s/it, lr: 1.0e-04 loss: 1.172e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1816/30000 [4:06:14<56:35:01,  7.23s/it, lr: 1.0e-04 loss: 1.172e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1816/30000 [4:06:14<56:35:01,  7.23s/it, lr: 1.0e-04 loss: 1.172e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1816/30000 [4:06:21<56:35:01,  7.23s/it, lr: 1.0e-04 loss: 2.886e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1816/30000 [4:06:21<56:35:01,  7.23s/it, lr: 1.0e-04 loss: 2.886e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1817/30000 [4:06:21<55:47:42,  7.13s/it, lr: 1.0e-04 loss: 2.886e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1817/30000 [4:06:21<55:47:42,  7.13s/it, lr: 1.0e-04 loss: 2.886e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1817/30000 [4:06:30<55:47:42,  7.13s/it, lr: 1.0e-04 loss: 6.979e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1817/30000 [4:06:30<55:47:42,  7.13s/it, lr: 1.0e-04 loss: 6.979e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1818/30000 [4:06:30<60:43:19,  7.76s/it, lr: 1.0e-04 loss: 6.979e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1818/30000 [4:06:30<60:43:19,  7.76s/it, lr: 1.0e-04 loss: 6.979e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1818/30000 [4:06:39<60:43:19,  7.76s/it, lr: 1.0e-04 loss: 6.185e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1818/30000 [4:06:39<60:43:19,  7.76s/it, lr: 1.0e-04 loss: 6.185e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1819/30000 [4:06:39<64:29:41,  8.24s/it, lr: 1.0e-04 loss: 6.185e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1819/30000 [4:06:39<64:29:41,  8.24s/it, lr: 1.0e-04 loss: 6.185e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1819/30000 [4:06:49<64:29:41,  8.24s/it, lr: 1.0e-04 loss: 1.359e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1819/30000 [4:06:49<64:29:41,  8.24s/it, lr: 1.0e-04 loss: 1.359e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1820/30000 [4:06:55<64:29:33,  8.24s/it, lr: 1.0e-04 loss: 7.842e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1820/30000 [4:06:55<64:29:33,  8.24s/it, lr: 1.0e-04 loss: 7.842e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1821/30000 [4:06:55<46:51:16,  5.99s/it, lr: 1.0e-04 loss: 7.842e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1821/30000 [4:06:55<46:51:16,  5.99s/it, lr: 1.0e-04 loss: 7.842e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1821/30000 [4:07:02<46:51:16,  5.99s/it, lr: 1.0e-04 loss: 1.070e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1821/30000 [4:07:02<46:51:16,  5.99s/it, lr: 1.0e-04 loss: 1.070e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1822/30000 [4:07:02<48:48:53,  6.24s/it, lr: 1.0e-04 loss: 1.070e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1822/30000 [4:07:02<48:48:53,  6.24s/it, lr: 1.0e-04 loss: 1.070e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1822/30000 [4:07:09<48:48:53,  6.24s/it, lr: 1.0e-04 loss: 6.052e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1822/30000 [4:07:09<48:48:53,  6.24s/it, lr: 1.0e-04 loss: 6.052e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1823/30000 [4:07:09<49:31:42,  6.33s/it, lr: 1.0e-04 loss: 6.052e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1823/30000 [4:07:09<49:31:42,  6.33s/it, lr: 1.0e-04 loss: 6.052e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1823/30000 [4:07:16<49:31:42,  6.33s/it, lr: 1.0e-04 loss: 1.264e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1823/30000 [4:07:16<49:31:42,  6.33s/it, lr: 1.0e-04 loss: 1.264e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1824/30000 [4:07:16<49:58:03,  6.38s/it, lr: 1.0e-04 loss: 1.264e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1824/30000 [4:07:16<49:58:03,  6.38s/it, lr: 1.0e-04 loss: 1.264e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1824/30000 [4:07:22<49:58:03,  6.38s/it, lr: 1.0e-04 loss: 9.300e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1824/30000 [4:07:22<49:58:03,  6.38s/it, lr: 1.0e-04 loss: 9.300e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1825/30000 [4:07:22<51:00:42,  6.52s/it, lr: 1.0e-04 loss: 9.300e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1825/30000 [4:07:22<51:00:42,  6.52s/it, lr: 1.0e-04 loss: 9.300e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1825/30000 [4:07:29<51:00:42,  6.52s/it, lr: 1.0e-04 loss: 8.421e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1825/30000 [4:07:29<51:00:42,  6.52s/it, lr: 1.0e-04 loss: 8.421e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1826/30000 [4:07:29<51:10:35,  6.54s/it, lr: 1.0e-04 loss: 8.421e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1826/30000 [4:07:29<51:10:35,  6.54s/it, lr: 1.0e-04 loss: 8.421e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1826/30000 [4:07:38<51:10:35,  6.54s/it, lr: 1.0e-04 loss: 6.415e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1826/30000 [4:07:38<51:10:35,  6.54s/it, lr: 1.0e-04 loss: 6.415e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1827/30000 [4:07:38<57:28:00,  7.34s/it, lr: 1.0e-04 loss: 6.415e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1827/30000 [4:07:38<57:28:00,  7.34s/it, lr: 1.0e-04 loss: 6.415e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1827/30000 [4:07:48<57:28:00,  7.34s/it, lr: 1.0e-04 loss: 7.240e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1827/30000 [4:07:48<57:28:00,  7.34s/it, lr: 1.0e-04 loss: 7.240e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1828/30000 [4:07:48<62:03:40,  7.93s/it, lr: 1.0e-04 loss: 7.240e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1828/30000 [4:07:48<62:03:40,  7.93s/it, lr: 1.0e-04 loss: 7.240e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1828/30000 [4:07:54<62:03:40,  7.93s/it, lr: 1.0e-04 loss: 1.309e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1828/30000 [4:07:54<62:03:40,  7.93s/it, lr: 1.0e-04 loss: 1.309e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1829/30000 [4:07:54<58:56:50,  7.53s/it, lr: 1.0e-04 loss: 1.309e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1829/30000 [4:07:54<58:56:50,  7.53s/it, lr: 1.0e-04 loss: 1.309e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1829/30000 [4:08:04<58:56:50,  7.53s/it, lr: 1.0e-04 loss: 6.690e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1829/30000 [4:08:04<58:56:50,  7.53s/it, lr: 1.0e-04 loss: 6.690e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1830/30000 [4:08:13<58:56:42,  7.53s/it, lr: 1.0e-04 loss: 4.443e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1830/30000 [4:08:13<58:56:42,  7.53s/it, lr: 1.0e-04 loss: 4.443e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1831/30000 [4:08:13<48:47:21,  6.24s/it, lr: 1.0e-04 loss: 4.443e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1831/30000 [4:08:13<48:47:21,  6.24s/it, lr: 1.0e-04 loss: 4.443e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1831/30000 [4:08:22<48:47:21,  6.24s/it, lr: 1.0e-04 loss: 7.569e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1831/30000 [4:08:22<48:47:21,  6.24s/it, lr: 1.0e-04 loss: 7.569e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1832/30000 [4:08:22<55:10:38,  7.05s/it, lr: 1.0e-04 loss: 7.569e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1832/30000 [4:08:22<55:10:38,  7.05s/it, lr: 1.0e-04 loss: 7.569e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1832/30000 [4:08:29<55:10:38,  7.05s/it, lr: 1.0e-04 loss: 4.601e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1832/30000 [4:08:29<55:10:38,  7.05s/it, lr: 1.0e-04 loss: 4.601e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1833/30000 [4:08:29<54:12:07,  6.93s/it, lr: 1.0e-04 loss: 4.601e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1833/30000 [4:08:29<54:12:07,  6.93s/it, lr: 1.0e-04 loss: 4.601e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1833/30000 [4:08:38<54:12:07,  6.93s/it, lr: 1.0e-04 loss: 5.610e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1833/30000 [4:08:38<54:12:07,  6.93s/it, lr: 1.0e-04 loss: 5.610e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1834/30000 [4:08:38<59:12:55,  7.57s/it, lr: 1.0e-04 loss: 5.610e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1834/30000 [4:08:38<59:12:55,  7.57s/it, lr: 1.0e-04 loss: 5.610e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1834/30000 [4:08:48<59:12:55,  7.57s/it, lr: 1.0e-04 loss: 7.296e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1834/30000 [4:08:48<59:12:55,  7.57s/it, lr: 1.0e-04 loss: 7.296e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1835/30000 [4:08:48<63:14:57,  8.08s/it, lr: 1.0e-04 loss: 7.296e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1835/30000 [4:08:48<63:14:57,  8.08s/it, lr: 1.0e-04 loss: 7.296e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1835/30000 [4:08:54<63:14:57,  8.08s/it, lr: 1.0e-04 loss: 1.049e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1835/30000 [4:08:54<63:14:57,  8.08s/it, lr: 1.0e-04 loss: 1.049e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1836/30000 [4:08:54<59:47:43,  7.64s/it, lr: 1.0e-04 loss: 1.049e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1836/30000 [4:08:54<59:47:43,  7.64s/it, lr: 1.0e-04 loss: 1.049e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1836/30000 [4:09:01<59:47:43,  7.64s/it, lr: 1.0e-04 loss: 1.179e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1836/30000 [4:09:01<59:47:43,  7.64s/it, lr: 1.0e-04 loss: 1.179e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1837/30000 [4:09:01<57:17:16,  7.32s/it, lr: 1.0e-04 loss: 1.179e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1837/30000 [4:09:01<57:17:16,  7.32s/it, lr: 1.0e-04 loss: 1.179e-01]
Xiang_wan_2_2_lora:   6%|##3                                    | 1837/30000 [4:09:10<57:17:16,  7.32s/it, lr: 1.0e-04 loss: 8.527e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1837/30000 [4:09:10<57:17:16,  7.32s/it, lr: 1.0e-04 loss: 8.527e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1838/30000 [4:09:10<61:48:38,  7.90s/it, lr: 1.0e-04 loss: 8.527e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1838/30000 [4:09:10<61:48:38,  7.90s/it, lr: 1.0e-04 loss: 8.527e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1838/30000 [4:09:17<61:48:38,  7.90s/it, lr: 1.0e-04 loss: 7.245e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1838/30000 [4:09:17<61:48:38,  7.90s/it, lr: 1.0e-04 loss: 7.245e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1839/30000 [4:09:17<58:58:17,  7.54s/it, lr: 1.0e-04 loss: 7.245e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1839/30000 [4:09:17<58:58:17,  7.54s/it, lr: 1.0e-04 loss: 7.245e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1839/30000 [4:09:26<58:58:17,  7.54s/it, lr: 1.0e-04 loss: 7.600e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1839/30000 [4:09:26<58:58:17,  7.54s/it, lr: 1.0e-04 loss: 7.600e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1840/30000 [4:09:36<58:58:10,  7.54s/it, lr: 1.0e-04 loss: 4.084e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1840/30000 [4:09:36<58:58:10,  7.54s/it, lr: 1.0e-04 loss: 4.084e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1841/30000 [4:09:36<48:53:04,  6.25s/it, lr: 1.0e-04 loss: 4.084e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1841/30000 [4:09:36<48:53:04,  6.25s/it, lr: 1.0e-04 loss: 4.084e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1841/30000 [4:09:45<48:53:04,  6.25s/it, lr: 1.0e-04 loss: 6.202e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1841/30000 [4:09:45<48:53:04,  6.25s/it, lr: 1.0e-04 loss: 6.202e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1842/30000 [4:09:45<54:43:43,  7.00s/it, lr: 1.0e-04 loss: 6.202e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1842/30000 [4:09:45<54:43:43,  7.00s/it, lr: 1.0e-04 loss: 6.202e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1842/30000 [4:09:52<54:43:43,  7.00s/it, lr: 1.0e-04 loss: 7.181e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1842/30000 [4:09:52<54:43:43,  7.00s/it, lr: 1.0e-04 loss: 7.181e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1843/30000 [4:09:52<54:04:52,  6.91s/it, lr: 1.0e-04 loss: 7.181e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1843/30000 [4:09:52<54:04:52,  6.91s/it, lr: 1.0e-04 loss: 7.181e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1843/30000 [4:09:58<54:04:52,  6.91s/it, lr: 1.0e-04 loss: 6.728e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1843/30000 [4:09:58<54:04:52,  6.91s/it, lr: 1.0e-04 loss: 6.728e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1844/30000 [4:09:58<53:17:04,  6.81s/it, lr: 1.0e-04 loss: 6.728e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1844/30000 [4:09:58<53:17:04,  6.81s/it, lr: 1.0e-04 loss: 6.728e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1844/30000 [4:10:05<53:17:04,  6.81s/it, lr: 1.0e-04 loss: 4.836e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1844/30000 [4:10:05<53:17:04,  6.81s/it, lr: 1.0e-04 loss: 4.836e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1845/30000 [4:10:05<52:49:44,  6.75s/it, lr: 1.0e-04 loss: 4.836e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1845/30000 [4:10:05<52:49:44,  6.75s/it, lr: 1.0e-04 loss: 4.836e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1845/30000 [4:10:11<52:49:44,  6.75s/it, lr: 1.0e-04 loss: 9.082e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1845/30000 [4:10:11<52:49:44,  6.75s/it, lr: 1.0e-04 loss: 9.082e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1846/30000 [4:10:11<52:23:59,  6.70s/it, lr: 1.0e-04 loss: 9.082e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1846/30000 [4:10:11<52:23:59,  6.70s/it, lr: 1.0e-04 loss: 9.082e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1846/30000 [4:10:18<52:23:59,  6.70s/it, lr: 1.0e-04 loss: 9.839e-02]
Xiang_wan_2_2_lora:   6%|##3                                    | 1846/30000 [4:10:18<52:23:59,  6.70s/it, lr: 1.0e-04 loss: 9.839e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1847/30000 [4:10:18<52:45:03,  6.75s/it, lr: 1.0e-04 loss: 9.839e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1847/30000 [4:10:18<52:45:03,  6.75s/it, lr: 1.0e-04 loss: 9.839e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1847/30000 [4:10:28<52:45:03,  6.75s/it, lr: 1.0e-04 loss: 8.025e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1847/30000 [4:10:28<52:45:03,  6.75s/it, lr: 1.0e-04 loss: 8.025e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1848/30000 [4:10:28<58:39:58,  7.50s/it, lr: 1.0e-04 loss: 8.025e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1848/30000 [4:10:28<58:39:58,  7.50s/it, lr: 1.0e-04 loss: 8.025e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1848/30000 [4:10:37<58:39:58,  7.50s/it, lr: 1.0e-04 loss: 3.774e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1848/30000 [4:10:37<58:39:58,  7.50s/it, lr: 1.0e-04 loss: 3.774e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1849/30000 [4:10:37<63:03:37,  8.06s/it, lr: 1.0e-04 loss: 3.774e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1849/30000 [4:10:37<63:03:37,  8.06s/it, lr: 1.0e-04 loss: 3.774e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1849/30000 [4:10:46<63:03:37,  8.06s/it, lr: 1.0e-04 loss: 8.542e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1849/30000 [4:10:46<63:03:37,  8.06s/it, lr: 1.0e-04 loss: 8.542e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1850/30000 [4:10:53<63:03:29,  8.06s/it, lr: 1.0e-04 loss: 8.145e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1850/30000 [4:10:53<63:03:29,  8.06s/it, lr: 1.0e-04 loss: 8.145e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1851/30000 [4:10:53<46:02:21,  5.89s/it, lr: 1.0e-04 loss: 8.145e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1851/30000 [4:10:53<46:02:21,  5.89s/it, lr: 1.0e-04 loss: 8.145e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1851/30000 [4:11:00<46:02:21,  5.89s/it, lr: 1.0e-04 loss: 1.018e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1851/30000 [4:11:00<46:02:21,  5.89s/it, lr: 1.0e-04 loss: 1.018e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1852/30000 [4:11:00<47:20:24,  6.05s/it, lr: 1.0e-04 loss: 1.018e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1852/30000 [4:11:00<47:20:24,  6.05s/it, lr: 1.0e-04 loss: 1.018e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1852/30000 [4:11:06<47:20:24,  6.05s/it, lr: 1.0e-04 loss: 7.424e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1852/30000 [4:11:06<47:20:24,  6.05s/it, lr: 1.0e-04 loss: 7.424e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1853/30000 [4:11:06<48:16:34,  6.17s/it, lr: 1.0e-04 loss: 7.424e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1853/30000 [4:11:06<48:16:34,  6.17s/it, lr: 1.0e-04 loss: 7.424e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1853/30000 [4:11:13<48:16:34,  6.17s/it, lr: 1.0e-04 loss: 1.006e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1853/30000 [4:11:13<48:16:34,  6.17s/it, lr: 1.0e-04 loss: 1.006e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1854/30000 [4:11:13<49:44:34,  6.36s/it, lr: 1.0e-04 loss: 1.006e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1854/30000 [4:11:13<49:44:34,  6.36s/it, lr: 1.0e-04 loss: 1.006e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1854/30000 [4:11:22<49:44:34,  6.36s/it, lr: 1.0e-04 loss: 5.955e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1854/30000 [4:11:22<49:44:34,  6.36s/it, lr: 1.0e-04 loss: 5.955e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1855/30000 [4:11:22<56:15:03,  7.19s/it, lr: 1.0e-04 loss: 5.955e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1855/30000 [4:11:22<56:15:03,  7.19s/it, lr: 1.0e-04 loss: 5.955e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1855/30000 [4:11:29<56:15:03,  7.19s/it, lr: 1.0e-04 loss: 8.638e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1855/30000 [4:11:29<56:15:03,  7.19s/it, lr: 1.0e-04 loss: 8.638e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1856/30000 [4:11:29<55:00:20,  7.04s/it, lr: 1.0e-04 loss: 8.638e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1856/30000 [4:11:29<55:00:20,  7.04s/it, lr: 1.0e-04 loss: 8.638e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1856/30000 [4:11:38<55:00:20,  7.04s/it, lr: 1.0e-04 loss: 4.659e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1856/30000 [4:11:38<55:00:20,  7.04s/it, lr: 1.0e-04 loss: 4.659e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1857/30000 [4:11:38<60:14:44,  7.71s/it, lr: 1.0e-04 loss: 4.659e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1857/30000 [4:11:38<60:14:44,  7.71s/it, lr: 1.0e-04 loss: 4.659e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1857/30000 [4:11:45<60:14:44,  7.71s/it, lr: 1.0e-04 loss: 9.370e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1857/30000 [4:11:45<60:14:44,  7.71s/it, lr: 1.0e-04 loss: 9.370e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1858/30000 [4:11:45<57:46:29,  7.39s/it, lr: 1.0e-04 loss: 9.370e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1858/30000 [4:11:45<57:46:29,  7.39s/it, lr: 1.0e-04 loss: 9.370e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1858/30000 [4:11:51<57:46:29,  7.39s/it, lr: 1.0e-04 loss: 7.670e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1858/30000 [4:11:51<57:46:29,  7.39s/it, lr: 1.0e-04 loss: 7.670e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1859/30000 [4:11:51<55:45:00,  7.13s/it, lr: 1.0e-04 loss: 7.670e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1859/30000 [4:11:51<55:45:00,  7.13s/it, lr: 1.0e-04 loss: 7.670e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1859/30000 [4:11:58<55:45:00,  7.13s/it, lr: 1.0e-04 loss: 8.485e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1859/30000 [4:11:58<55:45:00,  7.13s/it, lr: 1.0e-04 loss: 8.485e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1860/30000 [4:12:07<55:44:53,  7.13s/it, lr: 1.0e-04 loss: 1.165e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1860/30000 [4:12:07<55:44:53,  7.13s/it, lr: 1.0e-04 loss: 1.165e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1861/30000 [4:12:07<46:54:46,  6.00s/it, lr: 1.0e-04 loss: 1.165e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1861/30000 [4:12:07<46:54:46,  6.00s/it, lr: 1.0e-04 loss: 1.165e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1861/30000 [4:12:17<46:54:46,  6.00s/it, lr: 1.0e-04 loss: 6.389e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1861/30000 [4:12:17<46:54:46,  6.00s/it, lr: 1.0e-04 loss: 6.389e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1862/30000 [4:12:17<53:40:36,  6.87s/it, lr: 1.0e-04 loss: 6.389e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1862/30000 [4:12:17<53:40:36,  6.87s/it, lr: 1.0e-04 loss: 6.389e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1862/30000 [4:12:23<53:40:36,  6.87s/it, lr: 1.0e-04 loss: 6.012e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1862/30000 [4:12:23<53:40:36,  6.87s/it, lr: 1.0e-04 loss: 6.012e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1863/30000 [4:12:23<53:06:36,  6.80s/it, lr: 1.0e-04 loss: 6.012e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1863/30000 [4:12:23<53:06:36,  6.80s/it, lr: 1.0e-04 loss: 6.012e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1863/30000 [4:12:30<53:06:36,  6.80s/it, lr: 1.0e-04 loss: 9.794e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1863/30000 [4:12:30<53:06:36,  6.80s/it, lr: 1.0e-04 loss: 9.794e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1864/30000 [4:12:30<52:43:11,  6.75s/it, lr: 1.0e-04 loss: 9.794e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1864/30000 [4:12:30<52:43:11,  6.75s/it, lr: 1.0e-04 loss: 9.794e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1864/30000 [4:12:39<52:43:11,  6.75s/it, lr: 1.0e-04 loss: 7.646e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1864/30000 [4:12:39<52:43:11,  6.75s/it, lr: 1.0e-04 loss: 7.646e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1865/30000 [4:12:39<58:27:42,  7.48s/it, lr: 1.0e-04 loss: 7.646e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1865/30000 [4:12:39<58:27:42,  7.48s/it, lr: 1.0e-04 loss: 7.646e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1865/30000 [4:12:46<58:27:42,  7.48s/it, lr: 1.0e-04 loss: 6.659e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1865/30000 [4:12:46<58:27:42,  7.48s/it, lr: 1.0e-04 loss: 6.659e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1866/30000 [4:12:46<56:22:29,  7.21s/it, lr: 1.0e-04 loss: 6.659e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1866/30000 [4:12:46<56:22:29,  7.21s/it, lr: 1.0e-04 loss: 6.659e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1866/30000 [4:12:53<56:22:29,  7.21s/it, lr: 1.0e-04 loss: 1.187e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1866/30000 [4:12:53<56:22:29,  7.21s/it, lr: 1.0e-04 loss: 1.187e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1867/30000 [4:12:53<54:52:04,  7.02s/it, lr: 1.0e-04 loss: 1.187e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1867/30000 [4:12:53<54:52:04,  7.02s/it, lr: 1.0e-04 loss: 1.187e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1867/30000 [4:13:02<54:52:04,  7.02s/it, lr: 1.0e-04 loss: 5.283e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1867/30000 [4:13:02<54:52:04,  7.02s/it, lr: 1.0e-04 loss: 5.283e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1868/30000 [4:13:02<60:13:20,  7.71s/it, lr: 1.0e-04 loss: 5.283e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1868/30000 [4:13:02<60:13:20,  7.71s/it, lr: 1.0e-04 loss: 5.283e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1868/30000 [4:13:11<60:13:20,  7.71s/it, lr: 1.0e-04 loss: 4.221e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1868/30000 [4:13:11<60:13:20,  7.71s/it, lr: 1.0e-04 loss: 4.221e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1869/30000 [4:13:11<64:28:12,  8.25s/it, lr: 1.0e-04 loss: 4.221e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1869/30000 [4:13:11<64:28:12,  8.25s/it, lr: 1.0e-04 loss: 4.221e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1869/30000 [4:13:21<64:28:12,  8.25s/it, lr: 1.0e-04 loss: 8.108e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1869/30000 [4:13:21<64:28:12,  8.25s/it, lr: 1.0e-04 loss: 8.108e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1870/30000 [4:13:30<64:28:04,  8.25s/it, lr: 1.0e-04 loss: 4.140e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1870/30000 [4:13:30<64:28:04,  8.25s/it, lr: 1.0e-04 loss: 4.140e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1871/30000 [4:13:30<51:41:12,  6.61s/it, lr: 1.0e-04 loss: 4.140e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1871/30000 [4:13:30<51:41:12,  6.61s/it, lr: 1.0e-04 loss: 4.140e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1871/30000 [4:13:40<51:41:12,  6.61s/it, lr: 1.0e-04 loss: 1.100e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1871/30000 [4:13:40<51:41:12,  6.61s/it, lr: 1.0e-04 loss: 1.100e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1872/30000 [4:13:40<57:03:00,  7.30s/it, lr: 1.0e-04 loss: 1.100e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1872/30000 [4:13:40<57:03:00,  7.30s/it, lr: 1.0e-04 loss: 1.100e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1872/30000 [4:13:46<57:03:00,  7.30s/it, lr: 1.0e-04 loss: 8.867e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1872/30000 [4:13:46<57:03:00,  7.30s/it, lr: 1.0e-04 loss: 8.867e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1873/30000 [4:13:46<55:49:19,  7.14s/it, lr: 1.0e-04 loss: 8.867e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1873/30000 [4:13:46<55:49:19,  7.14s/it, lr: 1.0e-04 loss: 8.867e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1873/30000 [4:13:53<55:49:19,  7.14s/it, lr: 1.0e-04 loss: 9.865e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1873/30000 [4:13:53<55:49:19,  7.14s/it, lr: 1.0e-04 loss: 9.865e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1874/30000 [4:13:53<54:30:30,  6.98s/it, lr: 1.0e-04 loss: 9.865e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1874/30000 [4:13:53<54:30:30,  6.98s/it, lr: 1.0e-04 loss: 9.865e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1874/30000 [4:13:59<54:30:30,  6.98s/it, lr: 1.0e-04 loss: 8.627e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1874/30000 [4:13:59<54:30:30,  6.98s/it, lr: 1.0e-04 loss: 8.627e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1875/30000 [4:13:59<53:41:13,  6.87s/it, lr: 1.0e-04 loss: 8.627e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1875/30000 [4:13:59<53:41:13,  6.87s/it, lr: 1.0e-04 loss: 8.627e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1875/30000 [4:14:06<53:41:13,  6.87s/it, lr: 1.0e-04 loss: 9.610e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1875/30000 [4:14:06<53:41:13,  6.87s/it, lr: 1.0e-04 loss: 9.610e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1876/30000 [4:14:06<53:32:22,  6.85s/it, lr: 1.0e-04 loss: 9.610e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1876/30000 [4:14:06<53:32:22,  6.85s/it, lr: 1.0e-04 loss: 9.610e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1876/30000 [4:14:13<53:32:22,  6.85s/it, lr: 1.0e-04 loss: 6.790e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1876/30000 [4:14:13<53:32:22,  6.85s/it, lr: 1.0e-04 loss: 6.790e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1877/30000 [4:14:13<52:57:41,  6.78s/it, lr: 1.0e-04 loss: 6.790e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1877/30000 [4:14:13<52:57:41,  6.78s/it, lr: 1.0e-04 loss: 6.790e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1877/30000 [4:14:22<52:57:41,  6.78s/it, lr: 1.0e-04 loss: 9.649e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1877/30000 [4:14:22<52:57:41,  6.78s/it, lr: 1.0e-04 loss: 9.649e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1878/30000 [4:14:22<58:39:52,  7.51s/it, lr: 1.0e-04 loss: 9.649e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1878/30000 [4:14:22<58:39:52,  7.51s/it, lr: 1.0e-04 loss: 9.649e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1878/30000 [4:14:29<58:39:52,  7.51s/it, lr: 1.0e-04 loss: 7.379e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1878/30000 [4:14:29<58:39:52,  7.51s/it, lr: 1.0e-04 loss: 7.379e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1879/30000 [4:14:29<56:36:43,  7.25s/it, lr: 1.0e-04 loss: 7.379e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1879/30000 [4:14:29<56:36:43,  7.25s/it, lr: 1.0e-04 loss: 7.379e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1879/30000 [4:14:38<56:36:43,  7.25s/it, lr: 1.0e-04 loss: 8.544e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1879/30000 [4:14:38<56:36:43,  7.25s/it, lr: 1.0e-04 loss: 8.544e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1880/30000 [4:14:45<56:36:35,  7.25s/it, lr: 1.0e-04 loss: 9.110e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1880/30000 [4:14:45<56:36:35,  7.25s/it, lr: 1.0e-04 loss: 9.110e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1881/30000 [4:14:45<42:28:58,  5.44s/it, lr: 1.0e-04 loss: 9.110e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1881/30000 [4:14:45<42:28:58,  5.44s/it, lr: 1.0e-04 loss: 9.110e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1881/30000 [4:14:54<42:28:58,  5.44s/it, lr: 1.0e-04 loss: 5.144e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1881/30000 [4:14:54<42:28:58,  5.44s/it, lr: 1.0e-04 loss: 5.144e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1882/30000 [4:14:54<49:52:32,  6.39s/it, lr: 1.0e-04 loss: 5.144e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1882/30000 [4:14:54<49:52:32,  6.39s/it, lr: 1.0e-04 loss: 5.144e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1882/30000 [4:15:01<49:52:32,  6.39s/it, lr: 1.0e-04 loss: 5.173e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1882/30000 [4:15:01<49:52:32,  6.39s/it, lr: 1.0e-04 loss: 5.173e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1883/30000 [4:15:01<50:22:02,  6.45s/it, lr: 1.0e-04 loss: 5.173e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1883/30000 [4:15:01<50:22:02,  6.45s/it, lr: 1.0e-04 loss: 5.173e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1883/30000 [4:15:10<50:22:02,  6.45s/it, lr: 1.0e-04 loss: 5.845e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1883/30000 [4:15:10<50:22:02,  6.45s/it, lr: 1.0e-04 loss: 5.845e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1884/30000 [4:15:10<56:45:07,  7.27s/it, lr: 1.0e-04 loss: 5.845e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1884/30000 [4:15:10<56:45:07,  7.27s/it, lr: 1.0e-04 loss: 5.845e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1884/30000 [4:15:19<56:45:07,  7.27s/it, lr: 1.0e-04 loss: 2.660e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1884/30000 [4:15:19<56:45:07,  7.27s/it, lr: 1.0e-04 loss: 2.660e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1885/30000 [4:15:19<61:11:47,  7.84s/it, lr: 1.0e-04 loss: 2.660e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1885/30000 [4:15:19<61:11:47,  7.84s/it, lr: 1.0e-04 loss: 2.660e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1885/30000 [4:15:29<61:11:47,  7.84s/it, lr: 1.0e-04 loss: 5.144e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1885/30000 [4:15:29<61:11:47,  7.84s/it, lr: 1.0e-04 loss: 5.144e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1886/30000 [4:15:29<64:28:27,  8.26s/it, lr: 1.0e-04 loss: 5.144e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1886/30000 [4:15:29<64:28:27,  8.26s/it, lr: 1.0e-04 loss: 5.144e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1886/30000 [4:15:38<64:28:27,  8.26s/it, lr: 1.0e-04 loss: 2.716e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1886/30000 [4:15:38<64:28:27,  8.26s/it, lr: 1.0e-04 loss: 2.716e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1887/30000 [4:15:38<66:50:08,  8.56s/it, lr: 1.0e-04 loss: 2.716e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1887/30000 [4:15:38<66:50:08,  8.56s/it, lr: 1.0e-04 loss: 2.716e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1887/30000 [4:15:47<66:50:08,  8.56s/it, lr: 1.0e-04 loss: 4.941e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1887/30000 [4:15:47<66:50:08,  8.56s/it, lr: 1.0e-04 loss: 4.941e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1888/30000 [4:15:47<68:37:44,  8.79s/it, lr: 1.0e-04 loss: 4.941e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1888/30000 [4:15:47<68:37:44,  8.79s/it, lr: 1.0e-04 loss: 4.941e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1888/30000 [4:15:57<68:37:44,  8.79s/it, lr: 1.0e-04 loss: 4.950e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1888/30000 [4:15:57<68:37:44,  8.79s/it, lr: 1.0e-04 loss: 4.950e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1889/30000 [4:15:57<70:03:06,  8.97s/it, lr: 1.0e-04 loss: 4.950e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1889/30000 [4:15:57<70:03:06,  8.97s/it, lr: 1.0e-04 loss: 4.950e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1889/30000 [4:16:06<70:03:06,  8.97s/it, lr: 1.0e-04 loss: 7.275e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1889/30000 [4:16:06<70:03:06,  8.97s/it, lr: 1.0e-04 loss: 7.275e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1890/30000 [4:16:15<70:02:57,  8.97s/it, lr: 1.0e-04 loss: 3.233e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1890/30000 [4:16:15<70:02:57,  8.97s/it, lr: 1.0e-04 loss: 3.233e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1891/30000 [4:16:15<54:59:05,  7.04s/it, lr: 1.0e-04 loss: 3.233e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1891/30000 [4:16:15<54:59:05,  7.04s/it, lr: 1.0e-04 loss: 3.233e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1891/30000 [4:16:25<54:59:05,  7.04s/it, lr: 1.0e-04 loss: 5.660e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1891/30000 [4:16:25<54:59:05,  7.04s/it, lr: 1.0e-04 loss: 5.660e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1892/30000 [4:16:25<59:22:27,  7.60s/it, lr: 1.0e-04 loss: 5.660e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1892/30000 [4:16:25<59:22:27,  7.60s/it, lr: 1.0e-04 loss: 5.660e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1892/30000 [4:16:34<59:22:27,  7.60s/it, lr: 1.0e-04 loss: 1.214e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1892/30000 [4:16:34<59:22:27,  7.60s/it, lr: 1.0e-04 loss: 1.214e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1893/30000 [4:16:34<62:48:39,  8.04s/it, lr: 1.0e-04 loss: 1.214e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1893/30000 [4:16:34<62:48:39,  8.04s/it, lr: 1.0e-04 loss: 1.214e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1893/30000 [4:16:41<62:48:39,  8.04s/it, lr: 1.0e-04 loss: 7.721e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1893/30000 [4:16:41<62:48:39,  8.04s/it, lr: 1.0e-04 loss: 7.721e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1894/30000 [4:16:41<59:46:03,  7.66s/it, lr: 1.0e-04 loss: 7.721e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1894/30000 [4:16:41<59:46:03,  7.66s/it, lr: 1.0e-04 loss: 7.721e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1894/30000 [4:16:47<59:46:03,  7.66s/it, lr: 1.0e-04 loss: 4.299e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1894/30000 [4:16:47<59:46:03,  7.66s/it, lr: 1.0e-04 loss: 4.299e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1895/30000 [4:16:47<57:21:11,  7.35s/it, lr: 1.0e-04 loss: 4.299e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1895/30000 [4:16:47<57:21:11,  7.35s/it, lr: 1.0e-04 loss: 4.299e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1895/30000 [4:16:54<57:21:11,  7.35s/it, lr: 1.0e-04 loss: 9.337e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1895/30000 [4:16:54<57:21:11,  7.35s/it, lr: 1.0e-04 loss: 9.337e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1896/30000 [4:16:54<55:30:47,  7.11s/it, lr: 1.0e-04 loss: 9.337e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1896/30000 [4:16:54<55:30:47,  7.11s/it, lr: 1.0e-04 loss: 9.337e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1896/30000 [4:17:00<55:30:47,  7.11s/it, lr: 1.0e-04 loss: 1.279e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1896/30000 [4:17:00<55:30:47,  7.11s/it, lr: 1.0e-04 loss: 1.279e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1897/30000 [4:17:00<54:04:03,  6.93s/it, lr: 1.0e-04 loss: 1.279e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1897/30000 [4:17:00<54:04:03,  6.93s/it, lr: 1.0e-04 loss: 1.279e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1897/30000 [4:17:07<54:04:03,  6.93s/it, lr: 1.0e-04 loss: 1.082e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1897/30000 [4:17:07<54:04:03,  6.93s/it, lr: 1.0e-04 loss: 1.082e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1898/30000 [4:17:07<53:47:46,  6.89s/it, lr: 1.0e-04 loss: 1.082e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1898/30000 [4:17:07<53:47:46,  6.89s/it, lr: 1.0e-04 loss: 1.082e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1898/30000 [4:17:16<53:47:46,  6.89s/it, lr: 1.0e-04 loss: 6.052e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1898/30000 [4:17:16<53:47:46,  6.89s/it, lr: 1.0e-04 loss: 6.052e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1899/30000 [4:17:16<59:16:34,  7.59s/it, lr: 1.0e-04 loss: 6.052e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1899/30000 [4:17:16<59:16:34,  7.59s/it, lr: 1.0e-04 loss: 6.052e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1899/30000 [4:17:26<59:16:34,  7.59s/it, lr: 1.0e-04 loss: 6.012e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1899/30000 [4:17:26<59:16:34,  7.59s/it, lr: 1.0e-04 loss: 6.012e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1900/30000 [4:17:32<59:16:27,  7.59s/it, lr: 1.0e-04 loss: 4.905e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1900/30000 [4:17:32<59:16:27,  7.59s/it, lr: 1.0e-04 loss: 4.905e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1901/30000 [4:17:32<43:54:25,  5.63s/it, lr: 1.0e-04 loss: 4.905e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1901/30000 [4:17:32<43:54:25,  5.63s/it, lr: 1.0e-04 loss: 4.905e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1901/30000 [4:17:39<43:54:25,  5.63s/it, lr: 1.0e-04 loss: 9.829e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1901/30000 [4:17:39<43:54:25,  5.63s/it, lr: 1.0e-04 loss: 9.829e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1902/30000 [4:17:39<45:44:16,  5.86s/it, lr: 1.0e-04 loss: 9.829e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1902/30000 [4:17:39<45:44:16,  5.86s/it, lr: 1.0e-04 loss: 9.829e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1902/30000 [4:17:45<45:44:16,  5.86s/it, lr: 1.0e-04 loss: 1.576e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1902/30000 [4:17:45<45:44:16,  5.86s/it, lr: 1.0e-04 loss: 1.576e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1903/30000 [4:17:45<47:16:15,  6.06s/it, lr: 1.0e-04 loss: 1.576e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1903/30000 [4:17:45<47:16:15,  6.06s/it, lr: 1.0e-04 loss: 1.576e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1903/30000 [4:17:55<47:16:15,  6.06s/it, lr: 1.0e-04 loss: 7.854e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1903/30000 [4:17:55<47:16:15,  6.06s/it, lr: 1.0e-04 loss: 7.854e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1904/30000 [4:17:55<54:17:26,  6.96s/it, lr: 1.0e-04 loss: 7.854e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1904/30000 [4:17:55<54:17:26,  6.96s/it, lr: 1.0e-04 loss: 7.854e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1904/30000 [4:18:01<54:17:26,  6.96s/it, lr: 1.0e-04 loss: 7.370e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1904/30000 [4:18:01<54:17:26,  6.96s/it, lr: 1.0e-04 loss: 7.370e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1905/30000 [4:18:01<53:22:31,  6.84s/it, lr: 1.0e-04 loss: 7.370e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1905/30000 [4:18:01<53:22:31,  6.84s/it, lr: 1.0e-04 loss: 7.370e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1905/30000 [4:18:11<53:22:31,  6.84s/it, lr: 1.0e-04 loss: 1.004e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1905/30000 [4:18:11<53:22:31,  6.84s/it, lr: 1.0e-04 loss: 1.004e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1906/30000 [4:18:11<59:20:03,  7.60s/it, lr: 1.0e-04 loss: 1.004e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1906/30000 [4:18:11<59:20:03,  7.60s/it, lr: 1.0e-04 loss: 1.004e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1906/30000 [4:18:17<59:20:03,  7.60s/it, lr: 1.0e-04 loss: 7.077e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1906/30000 [4:18:17<59:20:03,  7.60s/it, lr: 1.0e-04 loss: 7.077e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1907/30000 [4:18:17<57:07:08,  7.32s/it, lr: 1.0e-04 loss: 7.077e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1907/30000 [4:18:17<57:07:08,  7.32s/it, lr: 1.0e-04 loss: 7.077e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1907/30000 [4:18:24<57:07:08,  7.32s/it, lr: 1.0e-04 loss: 9.257e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1907/30000 [4:18:24<57:07:08,  7.32s/it, lr: 1.0e-04 loss: 9.257e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1908/30000 [4:18:24<55:29:19,  7.11s/it, lr: 1.0e-04 loss: 9.257e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1908/30000 [4:18:24<55:29:19,  7.11s/it, lr: 1.0e-04 loss: 9.257e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1908/30000 [4:18:33<55:29:19,  7.11s/it, lr: 1.0e-04 loss: 7.961e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1908/30000 [4:18:33<55:29:19,  7.11s/it, lr: 1.0e-04 loss: 7.961e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1909/30000 [4:18:33<60:34:46,  7.76s/it, lr: 1.0e-04 loss: 7.961e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1909/30000 [4:18:33<60:34:46,  7.76s/it, lr: 1.0e-04 loss: 7.961e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1909/30000 [4:18:43<60:34:46,  7.76s/it, lr: 1.0e-04 loss: 7.246e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1909/30000 [4:18:43<60:34:46,  7.76s/it, lr: 1.0e-04 loss: 7.246e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1910/30000 [4:18:49<60:34:38,  7.76s/it, lr: 1.0e-04 loss: 1.201e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1910/30000 [4:18:49<60:34:38,  7.76s/it, lr: 1.0e-04 loss: 1.201e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1911/30000 [4:18:49<44:46:09,  5.74s/it, lr: 1.0e-04 loss: 1.201e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1911/30000 [4:18:49<44:46:09,  5.74s/it, lr: 1.0e-04 loss: 1.201e-01]
Xiang_wan_2_2_lora:   6%|##4                                    | 1911/30000 [4:18:56<44:46:09,  5.74s/it, lr: 1.0e-04 loss: 7.010e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1911/30000 [4:18:56<44:46:09,  5.74s/it, lr: 1.0e-04 loss: 7.010e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1912/30000 [4:18:56<46:22:46,  5.94s/it, lr: 1.0e-04 loss: 7.010e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1912/30000 [4:18:56<46:22:46,  5.94s/it, lr: 1.0e-04 loss: 7.010e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1912/30000 [4:19:03<46:22:46,  5.94s/it, lr: 1.0e-04 loss: 7.022e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1912/30000 [4:19:03<46:22:46,  5.94s/it, lr: 1.0e-04 loss: 7.022e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1913/30000 [4:19:03<48:17:23,  6.19s/it, lr: 1.0e-04 loss: 7.022e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1913/30000 [4:19:03<48:17:23,  6.19s/it, lr: 1.0e-04 loss: 7.022e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1913/30000 [4:19:10<48:17:23,  6.19s/it, lr: 1.0e-04 loss: 8.483e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1913/30000 [4:19:10<48:17:23,  6.19s/it, lr: 1.0e-04 loss: 8.483e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1914/30000 [4:19:10<49:16:00,  6.31s/it, lr: 1.0e-04 loss: 8.483e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1914/30000 [4:19:10<49:16:00,  6.31s/it, lr: 1.0e-04 loss: 8.483e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1914/30000 [4:19:19<49:16:00,  6.31s/it, lr: 1.0e-04 loss: 5.746e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1914/30000 [4:19:19<49:16:00,  6.31s/it, lr: 1.0e-04 loss: 5.746e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1915/30000 [4:19:19<55:49:24,  7.16s/it, lr: 1.0e-04 loss: 5.746e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1915/30000 [4:19:19<55:49:24,  7.16s/it, lr: 1.0e-04 loss: 5.746e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1915/30000 [4:19:25<55:49:24,  7.16s/it, lr: 1.0e-04 loss: 7.875e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1915/30000 [4:19:25<55:49:24,  7.16s/it, lr: 1.0e-04 loss: 7.875e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1916/30000 [4:19:25<54:35:55,  7.00s/it, lr: 1.0e-04 loss: 7.875e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1916/30000 [4:19:25<54:35:55,  7.00s/it, lr: 1.0e-04 loss: 7.875e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1916/30000 [4:19:35<54:35:55,  7.00s/it, lr: 1.0e-04 loss: 4.966e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1916/30000 [4:19:35<54:35:55,  7.00s/it, lr: 1.0e-04 loss: 4.966e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1917/30000 [4:19:35<59:57:14,  7.69s/it, lr: 1.0e-04 loss: 4.966e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1917/30000 [4:19:35<59:57:14,  7.69s/it, lr: 1.0e-04 loss: 4.966e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1917/30000 [4:19:44<59:57:14,  7.69s/it, lr: 1.0e-04 loss: 8.487e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1917/30000 [4:19:44<59:57:14,  7.69s/it, lr: 1.0e-04 loss: 8.487e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1918/30000 [4:19:44<63:54:32,  8.19s/it, lr: 1.0e-04 loss: 8.487e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1918/30000 [4:19:44<63:54:32,  8.19s/it, lr: 1.0e-04 loss: 8.487e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1918/30000 [4:19:51<63:54:32,  8.19s/it, lr: 1.0e-04 loss: 6.731e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1918/30000 [4:19:51<63:54:32,  8.19s/it, lr: 1.0e-04 loss: 6.731e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1919/30000 [4:19:51<60:16:05,  7.73s/it, lr: 1.0e-04 loss: 6.731e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1919/30000 [4:19:51<60:16:05,  7.73s/it, lr: 1.0e-04 loss: 6.731e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1919/30000 [4:20:00<60:16:05,  7.73s/it, lr: 1.0e-04 loss: 7.833e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1919/30000 [4:20:00<60:16:05,  7.73s/it, lr: 1.0e-04 loss: 7.833e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1920/30000 [4:20:07<60:15:57,  7.73s/it, lr: 1.0e-04 loss: 4.942e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1920/30000 [4:20:07<60:15:57,  7.73s/it, lr: 1.0e-04 loss: 4.942e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1921/30000 [4:20:07<44:25:21,  5.70s/it, lr: 1.0e-04 loss: 4.942e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1921/30000 [4:20:07<44:25:21,  5.70s/it, lr: 1.0e-04 loss: 4.942e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1921/30000 [4:20:16<44:25:21,  5.70s/it, lr: 1.0e-04 loss: 9.702e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1921/30000 [4:20:16<44:25:21,  5.70s/it, lr: 1.0e-04 loss: 9.702e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1922/30000 [4:20:16<51:25:34,  6.59s/it, lr: 1.0e-04 loss: 9.702e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1922/30000 [4:20:16<51:25:34,  6.59s/it, lr: 1.0e-04 loss: 9.702e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1922/30000 [4:20:23<51:25:34,  6.59s/it, lr: 1.0e-04 loss: 9.425e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1922/30000 [4:20:23<51:25:34,  6.59s/it, lr: 1.0e-04 loss: 9.425e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1923/30000 [4:20:23<51:28:56,  6.60s/it, lr: 1.0e-04 loss: 9.425e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1923/30000 [4:20:23<51:28:56,  6.60s/it, lr: 1.0e-04 loss: 9.425e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1923/30000 [4:20:30<51:28:56,  6.60s/it, lr: 1.0e-04 loss: 7.687e-02]
Xiang_wan_2_2_lora:   6%|##4                                    | 1923/30000 [4:20:30<51:28:56,  6.60s/it, lr: 1.0e-04 loss: 7.687e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1924/30000 [4:20:30<51:29:03,  6.60s/it, lr: 1.0e-04 loss: 7.687e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1924/30000 [4:20:30<51:29:03,  6.60s/it, lr: 1.0e-04 loss: 7.687e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1924/30000 [4:20:39<51:29:03,  6.60s/it, lr: 1.0e-04 loss: 7.730e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1924/30000 [4:20:39<51:29:03,  6.60s/it, lr: 1.0e-04 loss: 7.730e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1925/30000 [4:20:39<57:23:39,  7.36s/it, lr: 1.0e-04 loss: 7.730e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1925/30000 [4:20:39<57:23:39,  7.36s/it, lr: 1.0e-04 loss: 7.730e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1925/30000 [4:20:48<57:23:39,  7.36s/it, lr: 1.0e-04 loss: 5.593e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1925/30000 [4:20:48<57:23:39,  7.36s/it, lr: 1.0e-04 loss: 5.593e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1926/30000 [4:20:48<61:40:45,  7.91s/it, lr: 1.0e-04 loss: 5.593e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1926/30000 [4:20:48<61:40:45,  7.91s/it, lr: 1.0e-04 loss: 5.593e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1926/30000 [4:20:58<61:40:45,  7.91s/it, lr: 1.0e-04 loss: 4.419e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1926/30000 [4:20:58<61:40:45,  7.91s/it, lr: 1.0e-04 loss: 4.419e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1927/30000 [4:20:58<64:58:22,  8.33s/it, lr: 1.0e-04 loss: 4.419e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1927/30000 [4:20:58<64:58:22,  8.33s/it, lr: 1.0e-04 loss: 4.419e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1927/30000 [4:21:07<64:58:22,  8.33s/it, lr: 1.0e-04 loss: 6.095e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1927/30000 [4:21:07<64:58:22,  8.33s/it, lr: 1.0e-04 loss: 6.095e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1928/30000 [4:21:07<67:48:15,  8.70s/it, lr: 1.0e-04 loss: 6.095e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1928/30000 [4:21:07<67:48:15,  8.70s/it, lr: 1.0e-04 loss: 6.095e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1928/30000 [4:21:16<67:48:15,  8.70s/it, lr: 1.0e-04 loss: 3.749e-01]
Xiang_wan_2_2_lora:   6%|##5                                    | 1928/30000 [4:21:16<67:48:15,  8.70s/it, lr: 1.0e-04 loss: 3.749e-01]
Xiang_wan_2_2_lora:   6%|##5                                    | 1929/30000 [4:21:16<69:16:37,  8.88s/it, lr: 1.0e-04 loss: 3.749e-01]
Xiang_wan_2_2_lora:   6%|##5                                    | 1929/30000 [4:21:16<69:16:37,  8.88s/it, lr: 1.0e-04 loss: 3.749e-01]
Xiang_wan_2_2_lora:   6%|##5                                    | 1929/30000 [4:21:31<69:16:37,  8.88s/it, lr: 1.0e-04 loss: 6.800e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1929/30000 [4:21:31<69:16:37,  8.88s/it, lr: 1.0e-04 loss: 6.800e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1930/30000 [4:21:40<69:16:28,  8.88s/it, lr: 1.0e-04 loss: 9.888e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1930/30000 [4:21:40<69:16:28,  8.88s/it, lr: 1.0e-04 loss: 9.888e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1931/30000 [4:21:40<54:14:15,  6.96s/it, lr: 1.0e-04 loss: 9.888e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1931/30000 [4:21:40<54:14:15,  6.96s/it, lr: 1.0e-04 loss: 9.888e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1931/30000 [4:21:51<54:14:15,  6.96s/it, lr: 1.0e-04 loss: 9.444e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1931/30000 [4:21:51<54:14:15,  6.96s/it, lr: 1.0e-04 loss: 9.444e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1932/30000 [4:21:51<60:26:04,  7.75s/it, lr: 1.0e-04 loss: 9.444e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1932/30000 [4:21:51<60:26:04,  7.75s/it, lr: 1.0e-04 loss: 9.444e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1932/30000 [4:22:01<60:26:04,  7.75s/it, lr: 1.0e-04 loss: 5.087e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1932/30000 [4:22:01<60:26:04,  7.75s/it, lr: 1.0e-04 loss: 5.087e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1933/30000 [4:22:01<64:59:10,  8.34s/it, lr: 1.0e-04 loss: 5.087e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1933/30000 [4:22:01<64:59:10,  8.34s/it, lr: 1.0e-04 loss: 5.087e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1933/30000 [4:22:10<64:59:10,  8.34s/it, lr: 1.0e-04 loss: 7.146e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1933/30000 [4:22:10<64:59:10,  8.34s/it, lr: 1.0e-04 loss: 7.146e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1934/30000 [4:22:10<67:40:23,  8.68s/it, lr: 1.0e-04 loss: 7.146e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1934/30000 [4:22:10<67:40:23,  8.68s/it, lr: 1.0e-04 loss: 7.146e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1934/30000 [4:22:20<67:40:23,  8.68s/it, lr: 1.0e-04 loss: 6.093e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1934/30000 [4:22:20<67:40:23,  8.68s/it, lr: 1.0e-04 loss: 6.093e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1935/30000 [4:22:20<69:58:52,  8.98s/it, lr: 1.0e-04 loss: 6.093e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1935/30000 [4:22:20<69:58:52,  8.98s/it, lr: 1.0e-04 loss: 6.093e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1935/30000 [4:22:27<69:58:52,  8.98s/it, lr: 1.0e-04 loss: 9.800e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1935/30000 [4:22:27<69:58:52,  8.98s/it, lr: 1.0e-04 loss: 9.800e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1936/30000 [4:22:27<65:05:46,  8.35s/it, lr: 1.0e-04 loss: 9.800e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1936/30000 [4:22:27<65:05:46,  8.35s/it, lr: 1.0e-04 loss: 9.800e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1936/30000 [4:22:36<65:05:46,  8.35s/it, lr: 1.0e-04 loss: 1.082e-01]
Xiang_wan_2_2_lora:   6%|##5                                    | 1936/30000 [4:22:36<65:05:46,  8.35s/it, lr: 1.0e-04 loss: 1.082e-01]
Xiang_wan_2_2_lora:   6%|##5                                    | 1937/30000 [4:22:36<67:31:16,  8.66s/it, lr: 1.0e-04 loss: 1.082e-01]
Xiang_wan_2_2_lora:   6%|##5                                    | 1937/30000 [4:22:36<67:31:16,  8.66s/it, lr: 1.0e-04 loss: 1.082e-01]
Xiang_wan_2_2_lora:   6%|##5                                    | 1937/30000 [4:22:43<67:31:16,  8.66s/it, lr: 1.0e-04 loss: 1.037e-01]
Xiang_wan_2_2_lora:   6%|##5                                    | 1937/30000 [4:22:43<67:31:16,  8.66s/it, lr: 1.0e-04 loss: 1.037e-01]
Xiang_wan_2_2_lora:   6%|##5                                    | 1938/30000 [4:22:43<63:07:47,  8.10s/it, lr: 1.0e-04 loss: 1.037e-01]
Xiang_wan_2_2_lora:   6%|##5                                    | 1938/30000 [4:22:43<63:07:47,  8.10s/it, lr: 1.0e-04 loss: 1.037e-01]
Xiang_wan_2_2_lora:   6%|##5                                    | 1938/30000 [4:22:52<63:07:47,  8.10s/it, lr: 1.0e-04 loss: 5.991e-01]
Xiang_wan_2_2_lora:   6%|##5                                    | 1938/30000 [4:22:52<63:07:47,  8.10s/it, lr: 1.0e-04 loss: 5.991e-01]
Xiang_wan_2_2_lora:   6%|##5                                    | 1939/30000 [4:22:52<66:05:18,  8.48s/it, lr: 1.0e-04 loss: 5.991e-01]
Xiang_wan_2_2_lora:   6%|##5                                    | 1939/30000 [4:22:52<66:05:18,  8.48s/it, lr: 1.0e-04 loss: 5.991e-01]
Xiang_wan_2_2_lora:   6%|##5                                    | 1939/30000 [4:22:59<66:05:18,  8.48s/it, lr: 1.0e-04 loss: 8.224e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1939/30000 [4:22:59<66:05:18,  8.48s/it, lr: 1.0e-04 loss: 8.224e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1940/30000 [4:23:08<66:05:09,  8.48s/it, lr: 1.0e-04 loss: 5.000e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1940/30000 [4:23:08<66:05:09,  8.48s/it, lr: 1.0e-04 loss: 5.000e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1941/30000 [4:23:08<52:26:25,  6.73s/it, lr: 1.0e-04 loss: 5.000e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1941/30000 [4:23:08<52:26:25,  6.73s/it, lr: 1.0e-04 loss: 5.000e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1941/30000 [4:23:15<52:26:25,  6.73s/it, lr: 1.0e-04 loss: 1.083e-01]
Xiang_wan_2_2_lora:   6%|##5                                    | 1941/30000 [4:23:15<52:26:25,  6.73s/it, lr: 1.0e-04 loss: 1.083e-01]
Xiang_wan_2_2_lora:   6%|##5                                    | 1942/30000 [4:23:15<52:48:32,  6.78s/it, lr: 1.0e-04 loss: 1.083e-01]
Xiang_wan_2_2_lora:   6%|##5                                    | 1942/30000 [4:23:15<52:48:32,  6.78s/it, lr: 1.0e-04 loss: 1.083e-01]
Xiang_wan_2_2_lora:   6%|##5                                    | 1942/30000 [4:23:25<52:48:32,  6.78s/it, lr: 1.0e-04 loss: 5.355e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1942/30000 [4:23:25<52:48:32,  6.78s/it, lr: 1.0e-04 loss: 5.355e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1943/30000 [4:23:25<58:01:11,  7.44s/it, lr: 1.0e-04 loss: 5.355e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1943/30000 [4:23:25<58:01:11,  7.44s/it, lr: 1.0e-04 loss: 5.355e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1943/30000 [4:23:31<58:01:11,  7.44s/it, lr: 1.0e-04 loss: 9.580e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1943/30000 [4:23:31<58:01:11,  7.44s/it, lr: 1.0e-04 loss: 9.580e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1944/30000 [4:23:31<56:10:46,  7.21s/it, lr: 1.0e-04 loss: 9.580e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1944/30000 [4:23:31<56:10:46,  7.21s/it, lr: 1.0e-04 loss: 9.580e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1944/30000 [4:23:40<56:10:46,  7.21s/it, lr: 1.0e-04 loss: 4.772e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1944/30000 [4:23:40<56:10:46,  7.21s/it, lr: 1.0e-04 loss: 4.772e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1945/30000 [4:23:40<60:38:06,  7.78s/it, lr: 1.0e-04 loss: 4.772e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1945/30000 [4:23:40<60:38:06,  7.78s/it, lr: 1.0e-04 loss: 4.772e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1945/30000 [4:23:47<60:38:06,  7.78s/it, lr: 1.0e-04 loss: 8.031e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1945/30000 [4:23:47<60:38:06,  7.78s/it, lr: 1.0e-04 loss: 8.031e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1946/30000 [4:23:47<57:57:03,  7.44s/it, lr: 1.0e-04 loss: 8.031e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1946/30000 [4:23:47<57:57:03,  7.44s/it, lr: 1.0e-04 loss: 8.031e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1946/30000 [4:23:54<57:57:03,  7.44s/it, lr: 1.0e-04 loss: 7.217e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1946/30000 [4:23:54<57:57:03,  7.44s/it, lr: 1.0e-04 loss: 7.217e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1947/30000 [4:23:54<56:02:33,  7.19s/it, lr: 1.0e-04 loss: 7.217e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1947/30000 [4:23:54<56:02:33,  7.19s/it, lr: 1.0e-04 loss: 7.217e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1947/30000 [4:24:03<56:02:33,  7.19s/it, lr: 1.0e-04 loss: 7.722e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1947/30000 [4:24:03<56:02:33,  7.19s/it, lr: 1.0e-04 loss: 7.722e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1948/30000 [4:24:03<60:49:00,  7.80s/it, lr: 1.0e-04 loss: 7.722e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1948/30000 [4:24:03<60:49:00,  7.80s/it, lr: 1.0e-04 loss: 7.722e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1948/30000 [4:24:12<60:49:00,  7.80s/it, lr: 1.0e-04 loss: 1.743e-01]
Xiang_wan_2_2_lora:   6%|##5                                    | 1948/30000 [4:24:12<60:49:00,  7.80s/it, lr: 1.0e-04 loss: 1.743e-01]
Xiang_wan_2_2_lora:   6%|##5                                    | 1949/30000 [4:24:12<64:17:15,  8.25s/it, lr: 1.0e-04 loss: 1.743e-01]
Xiang_wan_2_2_lora:   6%|##5                                    | 1949/30000 [4:24:12<64:17:15,  8.25s/it, lr: 1.0e-04 loss: 1.743e-01]
Xiang_wan_2_2_lora:   6%|##5                                    | 1949/30000 [4:24:22<64:17:15,  8.25s/it, lr: 1.0e-04 loss: 5.018e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1949/30000 [4:24:22<64:17:15,  8.25s/it, lr: 1.0e-04 loss: 5.018e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1950/30000 [4:24:31<64:17:06,  8.25s/it, lr: 1.0e-04 loss: 2.820e-02]
Xiang_wan_2_2_lora:   6%|##5                                    | 1950/30000 [4:24:31<64:17:06,  8.25s/it, lr: 1.0e-04 loss: 2.820e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1951/30000 [4:24:31<51:29:32,  6.61s/it, lr: 1.0e-04 loss: 2.820e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1951/30000 [4:24:31<51:29:32,  6.61s/it, lr: 1.0e-04 loss: 2.820e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1951/30000 [4:24:38<51:29:32,  6.61s/it, lr: 1.0e-04 loss: 7.189e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1951/30000 [4:24:38<51:29:32,  6.61s/it, lr: 1.0e-04 loss: 7.189e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1952/30000 [4:24:38<51:33:11,  6.62s/it, lr: 1.0e-04 loss: 7.189e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1952/30000 [4:24:38<51:33:11,  6.62s/it, lr: 1.0e-04 loss: 7.189e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1952/30000 [4:24:44<51:33:11,  6.62s/it, lr: 1.0e-04 loss: 5.647e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1952/30000 [4:24:44<51:33:11,  6.62s/it, lr: 1.0e-04 loss: 5.647e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1953/30000 [4:24:44<51:21:13,  6.59s/it, lr: 1.0e-04 loss: 5.647e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1953/30000 [4:24:44<51:21:13,  6.59s/it, lr: 1.0e-04 loss: 5.647e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1953/30000 [4:24:51<51:21:13,  6.59s/it, lr: 1.0e-04 loss: 9.545e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1953/30000 [4:24:51<51:21:13,  6.59s/it, lr: 1.0e-04 loss: 9.545e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1954/30000 [4:24:51<51:25:07,  6.60s/it, lr: 1.0e-04 loss: 9.545e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1954/30000 [4:24:51<51:25:07,  6.60s/it, lr: 1.0e-04 loss: 9.545e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1954/30000 [4:25:00<51:25:07,  6.60s/it, lr: 1.0e-04 loss: 3.985e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1954/30000 [4:25:00<51:25:07,  6.60s/it, lr: 1.0e-04 loss: 3.985e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1955/30000 [4:25:00<57:19:42,  7.36s/it, lr: 1.0e-04 loss: 3.985e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1955/30000 [4:25:00<57:19:42,  7.36s/it, lr: 1.0e-04 loss: 3.985e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1955/30000 [4:25:10<57:19:42,  7.36s/it, lr: 1.0e-04 loss: 6.952e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1955/30000 [4:25:10<57:19:42,  7.36s/it, lr: 1.0e-04 loss: 6.952e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1956/30000 [4:25:10<61:56:41,  7.95s/it, lr: 1.0e-04 loss: 6.952e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1956/30000 [4:25:10<61:56:41,  7.95s/it, lr: 1.0e-04 loss: 6.952e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1956/30000 [4:25:19<61:56:41,  7.95s/it, lr: 1.0e-04 loss: 4.723e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1956/30000 [4:25:19<61:56:41,  7.95s/it, lr: 1.0e-04 loss: 4.723e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1957/30000 [4:25:19<65:35:25,  8.42s/it, lr: 1.0e-04 loss: 4.723e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1957/30000 [4:25:19<65:35:25,  8.42s/it, lr: 1.0e-04 loss: 4.723e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1957/30000 [4:25:26<65:35:25,  8.42s/it, lr: 1.0e-04 loss: 8.027e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1957/30000 [4:25:26<65:35:25,  8.42s/it, lr: 1.0e-04 loss: 8.027e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1958/30000 [4:25:26<61:27:51,  7.89s/it, lr: 1.0e-04 loss: 8.027e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1958/30000 [4:25:26<61:27:51,  7.89s/it, lr: 1.0e-04 loss: 8.027e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1958/30000 [4:25:32<61:27:51,  7.89s/it, lr: 1.0e-04 loss: 2.793e-01]
Xiang_wan_2_2_lora:   7%|##5                                    | 1958/30000 [4:25:32<61:27:51,  7.89s/it, lr: 1.0e-04 loss: 2.793e-01]
Xiang_wan_2_2_lora:   7%|##5                                    | 1959/30000 [4:25:32<58:22:08,  7.49s/it, lr: 1.0e-04 loss: 2.793e-01]
Xiang_wan_2_2_lora:   7%|##5                                    | 1959/30000 [4:25:32<58:22:08,  7.49s/it, lr: 1.0e-04 loss: 2.793e-01]
Xiang_wan_2_2_lora:   7%|##5                                    | 1959/30000 [4:25:39<58:22:08,  7.49s/it, lr: 1.0e-04 loss: 1.152e-01]
Xiang_wan_2_2_lora:   7%|##5                                    | 1959/30000 [4:25:39<58:22:08,  7.49s/it, lr: 1.0e-04 loss: 1.152e-01]
Xiang_wan_2_2_lora:   7%|##5                                    | 1960/30000 [4:25:45<58:22:00,  7.49s/it, lr: 1.0e-04 loss: 5.366e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1960/30000 [4:25:45<58:22:00,  7.49s/it, lr: 1.0e-04 loss: 5.366e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1961/30000 [4:25:45<43:23:19,  5.57s/it, lr: 1.0e-04 loss: 5.366e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1961/30000 [4:25:45<43:23:19,  5.57s/it, lr: 1.0e-04 loss: 5.366e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1961/30000 [4:25:52<43:23:19,  5.57s/it, lr: 1.0e-04 loss: 9.159e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1961/30000 [4:25:52<43:23:19,  5.57s/it, lr: 1.0e-04 loss: 9.159e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1962/30000 [4:25:52<45:15:06,  5.81s/it, lr: 1.0e-04 loss: 9.159e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1962/30000 [4:25:52<45:15:06,  5.81s/it, lr: 1.0e-04 loss: 9.159e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1962/30000 [4:26:01<45:15:06,  5.81s/it, lr: 1.0e-04 loss: 9.306e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1962/30000 [4:26:01<45:15:06,  5.81s/it, lr: 1.0e-04 loss: 9.306e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1963/30000 [4:26:01<52:16:59,  6.71s/it, lr: 1.0e-04 loss: 9.306e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1963/30000 [4:26:01<52:16:59,  6.71s/it, lr: 1.0e-04 loss: 9.306e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1963/30000 [4:26:11<52:16:59,  6.71s/it, lr: 1.0e-04 loss: 6.563e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1963/30000 [4:26:11<52:16:59,  6.71s/it, lr: 1.0e-04 loss: 6.563e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1964/30000 [4:26:11<58:08:42,  7.47s/it, lr: 1.0e-04 loss: 6.563e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1964/30000 [4:26:11<58:08:42,  7.47s/it, lr: 1.0e-04 loss: 6.563e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1964/30000 [4:26:20<58:08:42,  7.47s/it, lr: 1.0e-04 loss: 5.806e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1964/30000 [4:26:20<58:08:42,  7.47s/it, lr: 1.0e-04 loss: 5.806e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1965/30000 [4:26:20<62:23:19,  8.01s/it, lr: 1.0e-04 loss: 5.806e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1965/30000 [4:26:20<62:23:19,  8.01s/it, lr: 1.0e-04 loss: 5.806e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1965/30000 [4:26:30<62:23:19,  8.01s/it, lr: 1.0e-04 loss: 7.240e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1965/30000 [4:26:30<62:23:19,  8.01s/it, lr: 1.0e-04 loss: 7.240e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1966/30000 [4:26:30<65:27:34,  8.41s/it, lr: 1.0e-04 loss: 7.240e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1966/30000 [4:26:30<65:27:34,  8.41s/it, lr: 1.0e-04 loss: 7.240e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1966/30000 [4:26:36<65:27:34,  8.41s/it, lr: 1.0e-04 loss: 8.195e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1966/30000 [4:26:36<65:27:34,  8.41s/it, lr: 1.0e-04 loss: 8.195e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1967/30000 [4:26:36<61:17:56,  7.87s/it, lr: 1.0e-04 loss: 8.195e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1967/30000 [4:26:36<61:17:56,  7.87s/it, lr: 1.0e-04 loss: 8.195e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1967/30000 [4:26:45<61:17:56,  7.87s/it, lr: 1.0e-04 loss: 8.573e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1967/30000 [4:26:45<61:17:56,  7.87s/it, lr: 1.0e-04 loss: 8.573e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1968/30000 [4:26:45<64:29:14,  8.28s/it, lr: 1.0e-04 loss: 8.573e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1968/30000 [4:26:45<64:29:14,  8.28s/it, lr: 1.0e-04 loss: 8.573e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1968/30000 [4:26:55<64:29:14,  8.28s/it, lr: 1.0e-04 loss: 4.781e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1968/30000 [4:26:55<64:29:14,  8.28s/it, lr: 1.0e-04 loss: 4.781e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1969/30000 [4:26:55<66:50:25,  8.58s/it, lr: 1.0e-04 loss: 4.781e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1969/30000 [4:26:55<66:50:25,  8.58s/it, lr: 1.0e-04 loss: 4.781e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1969/30000 [4:27:01<66:50:25,  8.58s/it, lr: 1.0e-04 loss: 9.542e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1969/30000 [4:27:01<66:50:25,  8.58s/it, lr: 1.0e-04 loss: 9.542e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1970/30000 [4:27:11<66:50:16,  8.58s/it, lr: 1.0e-04 loss: 5.084e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1970/30000 [4:27:11<66:50:16,  8.58s/it, lr: 1.0e-04 loss: 5.084e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1971/30000 [4:27:11<53:10:15,  6.83s/it, lr: 1.0e-04 loss: 5.084e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1971/30000 [4:27:11<53:10:15,  6.83s/it, lr: 1.0e-04 loss: 5.084e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1971/30000 [4:27:17<53:10:15,  6.83s/it, lr: 1.0e-04 loss: 9.411e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1971/30000 [4:27:17<53:10:15,  6.83s/it, lr: 1.0e-04 loss: 9.411e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1972/30000 [4:27:17<52:41:13,  6.77s/it, lr: 1.0e-04 loss: 9.411e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1972/30000 [4:27:17<52:41:13,  6.77s/it, lr: 1.0e-04 loss: 9.411e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1972/30000 [4:27:27<52:41:13,  6.77s/it, lr: 1.0e-04 loss: 6.183e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1972/30000 [4:27:27<52:41:13,  6.77s/it, lr: 1.0e-04 loss: 6.183e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1973/30000 [4:27:27<57:45:46,  7.42s/it, lr: 1.0e-04 loss: 6.183e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1973/30000 [4:27:27<57:45:46,  7.42s/it, lr: 1.0e-04 loss: 6.183e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1973/30000 [4:27:36<57:45:46,  7.42s/it, lr: 1.0e-04 loss: 7.893e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1973/30000 [4:27:36<57:45:46,  7.42s/it, lr: 1.0e-04 loss: 7.893e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1974/30000 [4:27:36<61:48:12,  7.94s/it, lr: 1.0e-04 loss: 7.893e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1974/30000 [4:27:36<61:48:12,  7.94s/it, lr: 1.0e-04 loss: 7.893e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1974/30000 [4:27:43<61:48:12,  7.94s/it, lr: 1.0e-04 loss: 8.180e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1974/30000 [4:27:43<61:48:12,  7.94s/it, lr: 1.0e-04 loss: 8.180e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1975/30000 [4:27:43<58:52:39,  7.56s/it, lr: 1.0e-04 loss: 8.180e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1975/30000 [4:27:43<58:52:39,  7.56s/it, lr: 1.0e-04 loss: 8.180e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1975/30000 [4:27:49<58:52:39,  7.56s/it, lr: 1.0e-04 loss: 1.116e-01]
Xiang_wan_2_2_lora:   7%|##5                                    | 1975/30000 [4:27:49<58:52:39,  7.56s/it, lr: 1.0e-04 loss: 1.116e-01]
Xiang_wan_2_2_lora:   7%|##5                                    | 1976/30000 [4:27:49<56:44:28,  7.29s/it, lr: 1.0e-04 loss: 1.116e-01]
Xiang_wan_2_2_lora:   7%|##5                                    | 1976/30000 [4:27:49<56:44:28,  7.29s/it, lr: 1.0e-04 loss: 1.116e-01]
Xiang_wan_2_2_lora:   7%|##5                                    | 1976/30000 [4:27:59<56:44:28,  7.29s/it, lr: 1.0e-04 loss: 4.348e-01]
Xiang_wan_2_2_lora:   7%|##5                                    | 1976/30000 [4:27:59<56:44:28,  7.29s/it, lr: 1.0e-04 loss: 4.348e-01]
Xiang_wan_2_2_lora:   7%|##5                                    | 1977/30000 [4:27:59<61:19:44,  7.88s/it, lr: 1.0e-04 loss: 4.348e-01]
Xiang_wan_2_2_lora:   7%|##5                                    | 1977/30000 [4:27:59<61:19:44,  7.88s/it, lr: 1.0e-04 loss: 4.348e-01]
Xiang_wan_2_2_lora:   7%|##5                                    | 1977/30000 [4:28:08<61:19:44,  7.88s/it, lr: 1.0e-04 loss: 6.582e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1977/30000 [4:28:08<61:19:44,  7.88s/it, lr: 1.0e-04 loss: 6.582e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1978/30000 [4:28:08<64:41:24,  8.31s/it, lr: 1.0e-04 loss: 6.582e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1978/30000 [4:28:08<64:41:24,  8.31s/it, lr: 1.0e-04 loss: 6.582e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1978/30000 [4:28:15<64:41:24,  8.31s/it, lr: 1.0e-04 loss: 4.426e-01]
Xiang_wan_2_2_lora:   7%|##5                                    | 1978/30000 [4:28:15<64:41:24,  8.31s/it, lr: 1.0e-04 loss: 4.426e-01]
Xiang_wan_2_2_lora:   7%|##5                                    | 1979/30000 [4:28:15<61:24:56,  7.89s/it, lr: 1.0e-04 loss: 4.426e-01]
Xiang_wan_2_2_lora:   7%|##5                                    | 1979/30000 [4:28:15<61:24:56,  7.89s/it, lr: 1.0e-04 loss: 4.426e-01]
Xiang_wan_2_2_lora:   7%|##5                                    | 1979/30000 [4:28:24<61:24:56,  7.89s/it, lr: 1.0e-04 loss: 8.230e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1979/30000 [4:28:24<61:24:56,  7.89s/it, lr: 1.0e-04 loss: 8.230e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1980/30000 [4:28:33<61:24:48,  7.89s/it, lr: 1.0e-04 loss: 8.764e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1980/30000 [4:28:33<61:24:48,  7.89s/it, lr: 1.0e-04 loss: 8.764e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1981/30000 [4:28:33<49:54:43,  6.41s/it, lr: 1.0e-04 loss: 8.764e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1981/30000 [4:28:33<49:54:43,  6.41s/it, lr: 1.0e-04 loss: 8.764e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1981/30000 [4:28:40<49:54:43,  6.41s/it, lr: 1.0e-04 loss: 1.080e-01]
Xiang_wan_2_2_lora:   7%|##5                                    | 1981/30000 [4:28:40<49:54:43,  6.41s/it, lr: 1.0e-04 loss: 1.080e-01]
Xiang_wan_2_2_lora:   7%|##5                                    | 1982/30000 [4:28:40<50:14:21,  6.46s/it, lr: 1.0e-04 loss: 1.080e-01]
Xiang_wan_2_2_lora:   7%|##5                                    | 1982/30000 [4:28:40<50:14:21,  6.46s/it, lr: 1.0e-04 loss: 1.080e-01]
Xiang_wan_2_2_lora:   7%|##5                                    | 1982/30000 [4:28:47<50:14:21,  6.46s/it, lr: 1.0e-04 loss: 7.248e-01]
Xiang_wan_2_2_lora:   7%|##5                                    | 1982/30000 [4:28:47<50:14:21,  6.46s/it, lr: 1.0e-04 loss: 7.248e-01]
Xiang_wan_2_2_lora:   7%|##5                                    | 1983/30000 [4:28:47<50:25:06,  6.48s/it, lr: 1.0e-04 loss: 7.248e-01]
Xiang_wan_2_2_lora:   7%|##5                                    | 1983/30000 [4:28:47<50:25:06,  6.48s/it, lr: 1.0e-04 loss: 7.248e-01]
Xiang_wan_2_2_lora:   7%|##5                                    | 1983/30000 [4:28:53<50:25:06,  6.48s/it, lr: 1.0e-04 loss: 9.460e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1983/30000 [4:28:53<50:25:06,  6.48s/it, lr: 1.0e-04 loss: 9.460e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1984/30000 [4:28:53<50:27:12,  6.48s/it, lr: 1.0e-04 loss: 9.460e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1984/30000 [4:28:53<50:27:12,  6.48s/it, lr: 1.0e-04 loss: 9.460e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1984/30000 [4:29:02<50:27:12,  6.48s/it, lr: 1.0e-04 loss: 7.568e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1984/30000 [4:29:02<50:27:12,  6.48s/it, lr: 1.0e-04 loss: 7.568e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1985/30000 [4:29:02<56:34:27,  7.27s/it, lr: 1.0e-04 loss: 7.568e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1985/30000 [4:29:02<56:34:27,  7.27s/it, lr: 1.0e-04 loss: 7.568e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1985/30000 [4:29:09<56:34:27,  7.27s/it, lr: 1.0e-04 loss: 1.141e-01]
Xiang_wan_2_2_lora:   7%|##5                                    | 1985/30000 [4:29:09<56:34:27,  7.27s/it, lr: 1.0e-04 loss: 1.141e-01]
Xiang_wan_2_2_lora:   7%|##5                                    | 1986/30000 [4:29:09<55:35:38,  7.14s/it, lr: 1.0e-04 loss: 1.141e-01]
Xiang_wan_2_2_lora:   7%|##5                                    | 1986/30000 [4:29:09<55:35:38,  7.14s/it, lr: 1.0e-04 loss: 1.141e-01]
Xiang_wan_2_2_lora:   7%|##5                                    | 1986/30000 [4:29:18<55:35:38,  7.14s/it, lr: 1.0e-04 loss: 1.536e-01]
Xiang_wan_2_2_lora:   7%|##5                                    | 1986/30000 [4:29:18<55:35:38,  7.14s/it, lr: 1.0e-04 loss: 1.536e-01]
Xiang_wan_2_2_lora:   7%|##5                                    | 1987/30000 [4:29:18<60:27:51,  7.77s/it, lr: 1.0e-04 loss: 1.536e-01]
Xiang_wan_2_2_lora:   7%|##5                                    | 1987/30000 [4:29:18<60:27:51,  7.77s/it, lr: 1.0e-04 loss: 1.536e-01]
Xiang_wan_2_2_lora:   7%|##5                                    | 1987/30000 [4:29:28<60:27:51,  7.77s/it, lr: 1.0e-04 loss: 8.297e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1987/30000 [4:29:28<60:27:51,  7.77s/it, lr: 1.0e-04 loss: 8.297e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1988/30000 [4:29:28<64:01:49,  8.23s/it, lr: 1.0e-04 loss: 8.297e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1988/30000 [4:29:28<64:01:49,  8.23s/it, lr: 1.0e-04 loss: 8.297e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1988/30000 [4:29:34<64:01:49,  8.23s/it, lr: 1.0e-04 loss: 7.676e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1988/30000 [4:29:34<64:01:49,  8.23s/it, lr: 1.0e-04 loss: 7.676e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1989/30000 [4:29:34<60:09:34,  7.73s/it, lr: 1.0e-04 loss: 7.676e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1989/30000 [4:29:34<60:09:34,  7.73s/it, lr: 1.0e-04 loss: 7.676e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1989/30000 [4:29:44<60:09:34,  7.73s/it, lr: 1.0e-04 loss: 7.068e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1989/30000 [4:29:44<60:09:34,  7.73s/it, lr: 1.0e-04 loss: 7.068e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1990/30000 [4:29:53<60:09:26,  7.73s/it, lr: 1.0e-04 loss: 5.160e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1990/30000 [4:29:53<60:09:26,  7.73s/it, lr: 1.0e-04 loss: 5.160e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1991/30000 [4:29:53<49:13:17,  6.33s/it, lr: 1.0e-04 loss: 5.160e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1991/30000 [4:29:53<49:13:17,  6.33s/it, lr: 1.0e-04 loss: 5.160e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1991/30000 [4:29:59<49:13:17,  6.33s/it, lr: 1.0e-04 loss: 8.267e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1991/30000 [4:29:59<49:13:17,  6.33s/it, lr: 1.0e-04 loss: 8.267e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1992/30000 [4:29:59<49:36:14,  6.38s/it, lr: 1.0e-04 loss: 8.267e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1992/30000 [4:29:59<49:36:14,  6.38s/it, lr: 1.0e-04 loss: 8.267e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1992/30000 [4:30:09<49:36:14,  6.38s/it, lr: 1.0e-04 loss: 1.799e-01]
Xiang_wan_2_2_lora:   7%|##5                                    | 1992/30000 [4:30:09<49:36:14,  6.38s/it, lr: 1.0e-04 loss: 1.799e-01]
Xiang_wan_2_2_lora:   7%|##5                                    | 1993/30000 [4:30:09<55:51:03,  7.18s/it, lr: 1.0e-04 loss: 1.799e-01]
Xiang_wan_2_2_lora:   7%|##5                                    | 1993/30000 [4:30:09<55:51:03,  7.18s/it, lr: 1.0e-04 loss: 1.799e-01]
Xiang_wan_2_2_lora:   7%|##5                                    | 1993/30000 [4:30:15<55:51:03,  7.18s/it, lr: 1.0e-04 loss: 8.408e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1993/30000 [4:30:15<55:51:03,  7.18s/it, lr: 1.0e-04 loss: 8.408e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1994/30000 [4:30:15<54:33:17,  7.01s/it, lr: 1.0e-04 loss: 8.408e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1994/30000 [4:30:15<54:33:17,  7.01s/it, lr: 1.0e-04 loss: 8.408e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1994/30000 [4:30:25<54:33:17,  7.01s/it, lr: 1.0e-04 loss: 1.361e-01]
Xiang_wan_2_2_lora:   7%|##5                                    | 1994/30000 [4:30:25<54:33:17,  7.01s/it, lr: 1.0e-04 loss: 1.361e-01]
Xiang_wan_2_2_lora:   7%|##5                                    | 1995/30000 [4:30:25<59:30:51,  7.65s/it, lr: 1.0e-04 loss: 1.361e-01]
Xiang_wan_2_2_lora:   7%|##5                                    | 1995/30000 [4:30:25<59:30:51,  7.65s/it, lr: 1.0e-04 loss: 1.361e-01]
Xiang_wan_2_2_lora:   7%|##5                                    | 1995/30000 [4:30:31<59:30:51,  7.65s/it, lr: 1.0e-04 loss: 8.706e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1995/30000 [4:30:31<59:30:51,  7.65s/it, lr: 1.0e-04 loss: 8.706e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1996/30000 [4:30:31<57:05:50,  7.34s/it, lr: 1.0e-04 loss: 8.706e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1996/30000 [4:30:31<57:05:50,  7.34s/it, lr: 1.0e-04 loss: 8.706e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1996/30000 [4:30:38<57:05:50,  7.34s/it, lr: 1.0e-04 loss: 5.040e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1996/30000 [4:30:38<57:05:50,  7.34s/it, lr: 1.0e-04 loss: 5.040e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1997/30000 [4:30:38<55:19:59,  7.11s/it, lr: 1.0e-04 loss: 5.040e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1997/30000 [4:30:38<55:19:59,  7.11s/it, lr: 1.0e-04 loss: 5.040e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1997/30000 [4:30:44<55:19:59,  7.11s/it, lr: 1.0e-04 loss: 9.516e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1997/30000 [4:30:44<55:19:59,  7.11s/it, lr: 1.0e-04 loss: 9.516e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1998/30000 [4:30:44<54:00:37,  6.94s/it, lr: 1.0e-04 loss: 9.516e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1998/30000 [4:30:44<54:00:37,  6.94s/it, lr: 1.0e-04 loss: 9.516e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1998/30000 [4:30:51<54:00:37,  6.94s/it, lr: 1.0e-04 loss: 1.121e-01]
Xiang_wan_2_2_lora:   7%|##5                                    | 1998/30000 [4:30:51<54:00:37,  6.94s/it, lr: 1.0e-04 loss: 1.121e-01]
Xiang_wan_2_2_lora:   7%|##5                                    | 1999/30000 [4:30:51<53:10:44,  6.84s/it, lr: 1.0e-04 loss: 1.121e-01]
Xiang_wan_2_2_lora:   7%|##5                                    | 1999/30000 [4:30:51<53:10:44,  6.84s/it, lr: 1.0e-04 loss: 1.121e-01]
Xiang_wan_2_2_lora:   7%|##5                                    | 1999/30000 [4:31:01<53:10:44,  6.84s/it, lr: 1.0e-04 loss: 7.380e-02]
Xiang_wan_2_2_lora:   7%|##5                                    | 1999/30000 [4:31:01<53:10:44,  6.84s/it, lr: 1.0e-04 loss: 7.380e-02]
+Saving at step 2000
+Saved checkpoint to /home/featurize/ai-toolkit/output/Xiang_wan_2_2_lora/Xiang_wan_2_2_lora_000002000.safetensors
+Saved optimizer to /home/featurize/ai-toolkit/output/Xiang_wan_2_2_lora/optimizer.pt
+
Xiang_wan_2_2_lora:   7%|##6                                    | 2000/30000 [4:31:07<53:10:37,  6.84s/it, lr: 1.0e-04 loss: 1.634e-01]
Xiang_wan_2_2_lora:   7%|##6                                    | 2000/30000 [4:31:07<53:10:37,  6.84s/it, lr: 1.0e-04 loss: 1.634e-01]
Xiang_wan_2_2_lora:   7%|##6                                    | 2001/30000 [4:31:07<41:11:08,  5.30s/it, lr: 1.0e-04 loss: 1.634e-01]
Xiang_wan_2_2_lora:   7%|##6                                    | 2001/30000 [4:31:07<41:11:08,  5.30s/it, lr: 1.0e-04 loss: 1.634e-01]
Xiang_wan_2_2_lora:   7%|##6                                    | 2001/30000 [4:31:14<41:11:08,  5.30s/it, lr: 1.0e-04 loss: 8.878e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2001/30000 [4:31:14<41:11:08,  5.30s/it, lr: 1.0e-04 loss: 8.878e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2002/30000 [4:31:14<43:44:04,  5.62s/it, lr: 1.0e-04 loss: 8.878e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2002/30000 [4:31:14<43:44:04,  5.62s/it, lr: 1.0e-04 loss: 8.878e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2002/30000 [4:31:23<43:44:04,  5.62s/it, lr: 1.0e-04 loss: 3.346e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2002/30000 [4:31:23<43:44:04,  5.62s/it, lr: 1.0e-04 loss: 3.346e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2003/30000 [4:31:23<51:11:33,  6.58s/it, lr: 1.0e-04 loss: 3.346e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2003/30000 [4:31:23<51:11:33,  6.58s/it, lr: 1.0e-04 loss: 3.346e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2003/30000 [4:31:30<51:11:33,  6.58s/it, lr: 1.0e-04 loss: 6.611e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2003/30000 [4:31:30<51:11:33,  6.58s/it, lr: 1.0e-04 loss: 6.611e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2004/30000 [4:31:30<51:16:23,  6.59s/it, lr: 1.0e-04 loss: 6.611e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2004/30000 [4:31:30<51:16:23,  6.59s/it, lr: 1.0e-04 loss: 6.611e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2004/30000 [4:31:37<51:16:23,  6.59s/it, lr: 1.0e-04 loss: 5.097e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2004/30000 [4:31:37<51:16:23,  6.59s/it, lr: 1.0e-04 loss: 5.097e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2005/30000 [4:31:37<51:10:52,  6.58s/it, lr: 1.0e-04 loss: 5.097e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2005/30000 [4:31:37<51:10:52,  6.58s/it, lr: 1.0e-04 loss: 5.097e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2005/30000 [4:31:47<51:10:52,  6.58s/it, lr: 1.0e-04 loss: 6.460e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2005/30000 [4:31:47<51:10:52,  6.58s/it, lr: 1.0e-04 loss: 6.460e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2006/30000 [4:31:47<59:29:42,  7.65s/it, lr: 1.0e-04 loss: 6.460e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2006/30000 [4:31:47<59:29:42,  7.65s/it, lr: 1.0e-04 loss: 6.460e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2006/30000 [4:31:57<59:29:42,  7.65s/it, lr: 1.0e-04 loss: 1.080e-01]
Xiang_wan_2_2_lora:   7%|##6                                    | 2006/30000 [4:31:57<59:29:42,  7.65s/it, lr: 1.0e-04 loss: 1.080e-01]
Xiang_wan_2_2_lora:   7%|##6                                    | 2007/30000 [4:31:57<65:30:54,  8.43s/it, lr: 1.0e-04 loss: 1.080e-01]
Xiang_wan_2_2_lora:   7%|##6                                    | 2007/30000 [4:31:57<65:30:54,  8.43s/it, lr: 1.0e-04 loss: 1.080e-01]
Xiang_wan_2_2_lora:   7%|##6                                    | 2007/30000 [4:32:05<65:30:54,  8.43s/it, lr: 1.0e-04 loss: 6.904e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2007/30000 [4:32:05<65:30:54,  8.43s/it, lr: 1.0e-04 loss: 6.904e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2008/30000 [4:32:05<64:55:43,  8.35s/it, lr: 1.0e-04 loss: 6.904e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2008/30000 [4:32:05<64:55:43,  8.35s/it, lr: 1.0e-04 loss: 6.904e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2008/30000 [4:32:16<64:55:43,  8.35s/it, lr: 1.0e-04 loss: 4.332e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2008/30000 [4:32:16<64:55:43,  8.35s/it, lr: 1.0e-04 loss: 4.332e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2009/30000 [4:32:16<69:15:20,  8.91s/it, lr: 1.0e-04 loss: 4.332e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2009/30000 [4:32:16<69:15:20,  8.91s/it, lr: 1.0e-04 loss: 4.332e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2009/30000 [4:32:26<69:15:20,  8.91s/it, lr: 1.0e-04 loss: 8.039e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2009/30000 [4:32:26<69:15:20,  8.91s/it, lr: 1.0e-04 loss: 8.039e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2010/30000 [4:32:36<69:15:11,  8.91s/it, lr: 1.0e-04 loss: 4.061e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2010/30000 [4:32:36<69:15:11,  8.91s/it, lr: 1.0e-04 loss: 4.061e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2011/30000 [4:32:36<55:52:01,  7.19s/it, lr: 1.0e-04 loss: 4.061e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2011/30000 [4:32:36<55:52:01,  7.19s/it, lr: 1.0e-04 loss: 4.061e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2011/30000 [4:32:44<55:52:01,  7.19s/it, lr: 1.0e-04 loss: 7.109e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2011/30000 [4:32:44<55:52:01,  7.19s/it, lr: 1.0e-04 loss: 7.109e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2012/30000 [4:32:44<56:33:56,  7.28s/it, lr: 1.0e-04 loss: 7.109e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2012/30000 [4:32:44<56:33:56,  7.28s/it, lr: 1.0e-04 loss: 7.109e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2012/30000 [4:32:51<56:33:56,  7.28s/it, lr: 1.0e-04 loss: 7.826e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2012/30000 [4:32:51<56:33:56,  7.28s/it, lr: 1.0e-04 loss: 7.826e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2013/30000 [4:32:51<57:05:25,  7.34s/it, lr: 1.0e-04 loss: 7.826e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2013/30000 [4:32:51<57:05:25,  7.34s/it, lr: 1.0e-04 loss: 7.826e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2013/30000 [4:33:02<57:05:25,  7.34s/it, lr: 1.0e-04 loss: 7.771e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2013/30000 [4:33:02<57:05:25,  7.34s/it, lr: 1.0e-04 loss: 7.771e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2014/30000 [4:33:02<63:19:48,  8.15s/it, lr: 1.0e-04 loss: 7.771e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2014/30000 [4:33:02<63:19:48,  8.15s/it, lr: 1.0e-04 loss: 7.771e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2014/30000 [4:33:10<63:19:48,  8.15s/it, lr: 1.0e-04 loss: 5.413e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2014/30000 [4:33:10<63:19:48,  8.15s/it, lr: 1.0e-04 loss: 5.413e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2015/30000 [4:33:10<63:29:20,  8.17s/it, lr: 1.0e-04 loss: 5.413e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2015/30000 [4:33:10<63:29:20,  8.17s/it, lr: 1.0e-04 loss: 5.413e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2015/30000 [4:33:20<63:29:20,  8.17s/it, lr: 1.0e-04 loss: 6.956e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2015/30000 [4:33:20<63:29:20,  8.17s/it, lr: 1.0e-04 loss: 6.956e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2016/30000 [4:33:20<68:08:14,  8.77s/it, lr: 1.0e-04 loss: 6.956e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2016/30000 [4:33:20<68:08:14,  8.77s/it, lr: 1.0e-04 loss: 6.956e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2016/30000 [4:33:27<68:08:14,  8.77s/it, lr: 1.0e-04 loss: 7.842e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2016/30000 [4:33:27<68:08:14,  8.77s/it, lr: 1.0e-04 loss: 7.842e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2017/30000 [4:33:27<64:56:03,  8.35s/it, lr: 1.0e-04 loss: 7.842e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2017/30000 [4:33:27<64:56:03,  8.35s/it, lr: 1.0e-04 loss: 7.842e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2017/30000 [4:33:37<64:56:03,  8.35s/it, lr: 1.0e-04 loss: 4.856e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2017/30000 [4:33:37<64:56:03,  8.35s/it, lr: 1.0e-04 loss: 4.856e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2018/30000 [4:33:37<67:14:32,  8.65s/it, lr: 1.0e-04 loss: 4.856e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2018/30000 [4:33:37<67:14:32,  8.65s/it, lr: 1.0e-04 loss: 4.856e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2018/30000 [4:33:43<67:14:32,  8.65s/it, lr: 1.0e-04 loss: 6.880e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2018/30000 [4:33:43<67:14:32,  8.65s/it, lr: 1.0e-04 loss: 6.880e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2019/30000 [4:33:43<62:32:22,  8.05s/it, lr: 1.0e-04 loss: 6.880e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2019/30000 [4:33:43<62:32:22,  8.05s/it, lr: 1.0e-04 loss: 6.880e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2019/30000 [4:33:50<62:32:22,  8.05s/it, lr: 1.0e-04 loss: 1.079e-01]
Xiang_wan_2_2_lora:   7%|##6                                    | 2019/30000 [4:33:50<62:32:22,  8.05s/it, lr: 1.0e-04 loss: 1.079e-01]
Xiang_wan_2_2_lora:   7%|##6                                    | 2020/30000 [4:33:59<62:32:14,  8.05s/it, lr: 1.0e-04 loss: 4.932e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2020/30000 [4:33:59<62:32:14,  8.05s/it, lr: 1.0e-04 loss: 4.932e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2021/30000 [4:33:59<50:30:27,  6.50s/it, lr: 1.0e-04 loss: 4.932e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2021/30000 [4:33:59<50:30:27,  6.50s/it, lr: 1.0e-04 loss: 4.932e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2021/30000 [4:34:06<50:30:27,  6.50s/it, lr: 1.0e-04 loss: 9.098e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2021/30000 [4:34:06<50:30:27,  6.50s/it, lr: 1.0e-04 loss: 9.098e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2022/30000 [4:34:06<50:39:58,  6.52s/it, lr: 1.0e-04 loss: 9.098e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2022/30000 [4:34:06<50:39:58,  6.52s/it, lr: 1.0e-04 loss: 9.098e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2022/30000 [4:34:15<50:39:58,  6.52s/it, lr: 1.0e-04 loss: 6.616e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2022/30000 [4:34:15<50:39:58,  6.52s/it, lr: 1.0e-04 loss: 6.616e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2023/30000 [4:34:15<56:40:22,  7.29s/it, lr: 1.0e-04 loss: 6.616e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2023/30000 [4:34:15<56:40:22,  7.29s/it, lr: 1.0e-04 loss: 6.616e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2023/30000 [4:34:25<56:40:22,  7.29s/it, lr: 1.0e-04 loss: 9.302e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2023/30000 [4:34:25<56:40:22,  7.29s/it, lr: 1.0e-04 loss: 9.302e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2024/30000 [4:34:25<60:56:11,  7.84s/it, lr: 1.0e-04 loss: 9.302e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2024/30000 [4:34:25<60:56:11,  7.84s/it, lr: 1.0e-04 loss: 9.302e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2024/30000 [4:34:31<60:56:11,  7.84s/it, lr: 1.0e-04 loss: 7.424e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2024/30000 [4:34:31<60:56:11,  7.84s/it, lr: 1.0e-04 loss: 7.424e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2025/30000 [4:34:31<58:18:55,  7.50s/it, lr: 1.0e-04 loss: 7.424e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2025/30000 [4:34:31<58:18:55,  7.50s/it, lr: 1.0e-04 loss: 7.424e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2025/30000 [4:34:38<58:18:55,  7.50s/it, lr: 1.0e-04 loss: 1.049e-01]
Xiang_wan_2_2_lora:   7%|##6                                    | 2025/30000 [4:34:38<58:18:55,  7.50s/it, lr: 1.0e-04 loss: 1.049e-01]
Xiang_wan_2_2_lora:   7%|##6                                    | 2026/30000 [4:34:38<56:05:48,  7.22s/it, lr: 1.0e-04 loss: 1.049e-01]
Xiang_wan_2_2_lora:   7%|##6                                    | 2026/30000 [4:34:38<56:05:48,  7.22s/it, lr: 1.0e-04 loss: 1.049e-01]
Xiang_wan_2_2_lora:   7%|##6                                    | 2026/30000 [4:34:47<56:05:48,  7.22s/it, lr: 1.0e-04 loss: 6.266e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2026/30000 [4:34:47<56:05:48,  7.22s/it, lr: 1.0e-04 loss: 6.266e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2027/30000 [4:34:47<60:53:34,  7.84s/it, lr: 1.0e-04 loss: 6.266e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2027/30000 [4:34:47<60:53:34,  7.84s/it, lr: 1.0e-04 loss: 6.266e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2027/30000 [4:34:56<60:53:34,  7.84s/it, lr: 1.0e-04 loss: 8.098e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2027/30000 [4:34:56<60:53:34,  7.84s/it, lr: 1.0e-04 loss: 8.098e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2028/30000 [4:34:56<64:10:05,  8.26s/it, lr: 1.0e-04 loss: 8.098e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2028/30000 [4:34:56<64:10:05,  8.26s/it, lr: 1.0e-04 loss: 8.098e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2028/30000 [4:35:03<64:10:05,  8.26s/it, lr: 1.0e-04 loss: 5.305e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2028/30000 [4:35:03<64:10:05,  8.26s/it, lr: 1.0e-04 loss: 5.305e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2029/30000 [4:35:03<60:29:41,  7.79s/it, lr: 1.0e-04 loss: 5.305e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2029/30000 [4:35:03<60:29:41,  7.79s/it, lr: 1.0e-04 loss: 5.305e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2029/30000 [4:35:10<60:29:41,  7.79s/it, lr: 1.0e-04 loss: 8.281e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2029/30000 [4:35:10<60:29:41,  7.79s/it, lr: 1.0e-04 loss: 8.281e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2030/30000 [4:35:19<60:29:33,  7.79s/it, lr: 1.0e-04 loss: 3.303e-01]
Xiang_wan_2_2_lora:   7%|##6                                    | 2030/30000 [4:35:19<60:29:33,  7.79s/it, lr: 1.0e-04 loss: 3.303e-01]
Xiang_wan_2_2_lora:   7%|##6                                    | 2031/30000 [4:35:19<49:28:29,  6.37s/it, lr: 1.0e-04 loss: 3.303e-01]
Xiang_wan_2_2_lora:   7%|##6                                    | 2031/30000 [4:35:19<49:28:29,  6.37s/it, lr: 1.0e-04 loss: 3.303e-01]
Xiang_wan_2_2_lora:   7%|##6                                    | 2031/30000 [4:35:26<49:28:29,  6.37s/it, lr: 1.0e-04 loss: 9.809e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2031/30000 [4:35:26<49:28:29,  6.37s/it, lr: 1.0e-04 loss: 9.809e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2032/30000 [4:35:26<49:51:46,  6.42s/it, lr: 1.0e-04 loss: 9.809e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2032/30000 [4:35:26<49:51:46,  6.42s/it, lr: 1.0e-04 loss: 9.809e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2032/30000 [4:35:32<49:51:46,  6.42s/it, lr: 1.0e-04 loss: 7.216e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2032/30000 [4:35:32<49:51:46,  6.42s/it, lr: 1.0e-04 loss: 7.216e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2033/30000 [4:35:32<50:06:42,  6.45s/it, lr: 1.0e-04 loss: 7.216e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2033/30000 [4:35:32<50:06:42,  6.45s/it, lr: 1.0e-04 loss: 7.216e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2033/30000 [4:35:42<50:06:42,  6.45s/it, lr: 1.0e-04 loss: 5.324e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2033/30000 [4:35:42<50:06:42,  6.45s/it, lr: 1.0e-04 loss: 5.324e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2034/30000 [4:35:42<56:03:04,  7.22s/it, lr: 1.0e-04 loss: 5.324e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2034/30000 [4:35:42<56:03:04,  7.22s/it, lr: 1.0e-04 loss: 5.324e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2034/30000 [4:35:51<56:03:04,  7.22s/it, lr: 1.0e-04 loss: 4.579e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2034/30000 [4:35:51<56:03:04,  7.22s/it, lr: 1.0e-04 loss: 4.579e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2035/30000 [4:35:51<60:41:43,  7.81s/it, lr: 1.0e-04 loss: 4.579e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2035/30000 [4:35:51<60:41:43,  7.81s/it, lr: 1.0e-04 loss: 4.579e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2035/30000 [4:35:57<60:41:43,  7.81s/it, lr: 1.0e-04 loss: 1.088e-01]
Xiang_wan_2_2_lora:   7%|##6                                    | 2035/30000 [4:35:57<60:41:43,  7.81s/it, lr: 1.0e-04 loss: 1.088e-01]
Xiang_wan_2_2_lora:   7%|##6                                    | 2036/30000 [4:35:57<57:50:29,  7.45s/it, lr: 1.0e-04 loss: 1.088e-01]
Xiang_wan_2_2_lora:   7%|##6                                    | 2036/30000 [4:35:57<57:50:29,  7.45s/it, lr: 1.0e-04 loss: 1.088e-01]
Xiang_wan_2_2_lora:   7%|##6                                    | 2036/30000 [4:36:04<57:50:29,  7.45s/it, lr: 1.0e-04 loss: 6.091e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2036/30000 [4:36:04<57:50:29,  7.45s/it, lr: 1.0e-04 loss: 6.091e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2037/30000 [4:36:04<56:04:30,  7.22s/it, lr: 1.0e-04 loss: 6.091e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2037/30000 [4:36:04<56:04:30,  7.22s/it, lr: 1.0e-04 loss: 6.091e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2037/30000 [4:36:11<56:04:30,  7.22s/it, lr: 1.0e-04 loss: 8.952e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2037/30000 [4:36:11<56:04:30,  7.22s/it, lr: 1.0e-04 loss: 8.952e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2038/30000 [4:36:11<55:14:30,  7.11s/it, lr: 1.0e-04 loss: 8.952e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2038/30000 [4:36:11<55:14:30,  7.11s/it, lr: 1.0e-04 loss: 8.952e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2038/30000 [4:36:20<55:14:30,  7.11s/it, lr: 1.0e-04 loss: 3.729e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2038/30000 [4:36:20<55:14:30,  7.11s/it, lr: 1.0e-04 loss: 3.729e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2039/30000 [4:36:20<60:28:20,  7.79s/it, lr: 1.0e-04 loss: 3.729e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2039/30000 [4:36:20<60:28:20,  7.79s/it, lr: 1.0e-04 loss: 3.729e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2039/30000 [4:36:30<60:28:20,  7.79s/it, lr: 1.0e-04 loss: 7.680e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2039/30000 [4:36:30<60:28:20,  7.79s/it, lr: 1.0e-04 loss: 7.680e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2040/30000 [4:36:36<60:28:12,  7.79s/it, lr: 1.0e-04 loss: 4.978e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2040/30000 [4:36:36<60:28:12,  7.79s/it, lr: 1.0e-04 loss: 4.978e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2041/30000 [4:36:36<44:35:22,  5.74s/it, lr: 1.0e-04 loss: 4.978e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2041/30000 [4:36:36<44:35:22,  5.74s/it, lr: 1.0e-04 loss: 4.978e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2041/30000 [4:36:46<44:35:22,  5.74s/it, lr: 1.0e-04 loss: 7.496e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2041/30000 [4:36:46<44:35:22,  5.74s/it, lr: 1.0e-04 loss: 7.496e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2042/30000 [4:36:46<51:23:39,  6.62s/it, lr: 1.0e-04 loss: 7.496e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2042/30000 [4:36:46<51:23:39,  6.62s/it, lr: 1.0e-04 loss: 7.496e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2042/30000 [4:36:52<51:23:39,  6.62s/it, lr: 1.0e-04 loss: 7.984e-01]
Xiang_wan_2_2_lora:   7%|##6                                    | 2042/30000 [4:36:52<51:23:39,  6.62s/it, lr: 1.0e-04 loss: 7.984e-01]
Xiang_wan_2_2_lora:   7%|##6                                    | 2043/30000 [4:36:52<51:19:32,  6.61s/it, lr: 1.0e-04 loss: 7.984e-01]
Xiang_wan_2_2_lora:   7%|##6                                    | 2043/30000 [4:36:52<51:19:32,  6.61s/it, lr: 1.0e-04 loss: 7.984e-01]
Xiang_wan_2_2_lora:   7%|##6                                    | 2043/30000 [4:37:02<51:19:32,  6.61s/it, lr: 1.0e-04 loss: 6.644e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2043/30000 [4:37:02<51:19:32,  6.61s/it, lr: 1.0e-04 loss: 6.644e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2044/30000 [4:37:02<57:01:46,  7.34s/it, lr: 1.0e-04 loss: 6.644e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2044/30000 [4:37:02<57:01:46,  7.34s/it, lr: 1.0e-04 loss: 6.644e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2044/30000 [4:37:11<57:01:46,  7.34s/it, lr: 1.0e-04 loss: 4.585e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2044/30000 [4:37:11<57:01:46,  7.34s/it, lr: 1.0e-04 loss: 4.585e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2045/30000 [4:37:11<61:54:11,  7.97s/it, lr: 1.0e-04 loss: 4.585e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2045/30000 [4:37:11<61:54:11,  7.97s/it, lr: 1.0e-04 loss: 4.585e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2045/30000 [4:37:21<61:54:11,  7.97s/it, lr: 1.0e-04 loss: 7.274e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2045/30000 [4:37:21<61:54:11,  7.97s/it, lr: 1.0e-04 loss: 7.274e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2046/30000 [4:37:21<64:58:29,  8.37s/it, lr: 1.0e-04 loss: 7.274e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2046/30000 [4:37:21<64:58:29,  8.37s/it, lr: 1.0e-04 loss: 7.274e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2046/30000 [4:37:27<64:58:29,  8.37s/it, lr: 1.0e-04 loss: 4.509e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2046/30000 [4:37:27<64:58:29,  8.37s/it, lr: 1.0e-04 loss: 4.509e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2047/30000 [4:37:27<61:02:00,  7.86s/it, lr: 1.0e-04 loss: 4.509e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2047/30000 [4:37:27<61:02:00,  7.86s/it, lr: 1.0e-04 loss: 4.509e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2047/30000 [4:37:36<61:02:00,  7.86s/it, lr: 1.0e-04 loss: 7.212e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2047/30000 [4:37:36<61:02:00,  7.86s/it, lr: 1.0e-04 loss: 7.212e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2048/30000 [4:37:36<64:20:45,  8.29s/it, lr: 1.0e-04 loss: 7.212e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2048/30000 [4:37:36<64:20:45,  8.29s/it, lr: 1.0e-04 loss: 7.212e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2048/30000 [4:37:46<64:20:45,  8.29s/it, lr: 1.0e-04 loss: 4.843e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2048/30000 [4:37:46<64:20:45,  8.29s/it, lr: 1.0e-04 loss: 4.843e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2049/30000 [4:37:46<66:50:59,  8.61s/it, lr: 1.0e-04 loss: 4.843e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2049/30000 [4:37:46<66:50:59,  8.61s/it, lr: 1.0e-04 loss: 4.843e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2049/30000 [4:37:53<66:50:59,  8.61s/it, lr: 1.0e-04 loss: 6.600e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2049/30000 [4:37:53<66:50:59,  8.61s/it, lr: 1.0e-04 loss: 6.600e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2050/30000 [4:37:59<66:50:50,  8.61s/it, lr: 1.0e-04 loss: 3.591e-01]
Xiang_wan_2_2_lora:   7%|##6                                    | 2050/30000 [4:37:59<66:50:50,  8.61s/it, lr: 1.0e-04 loss: 3.591e-01]
Xiang_wan_2_2_lora:   7%|##6                                    | 2051/30000 [4:37:59<47:54:11,  6.17s/it, lr: 1.0e-04 loss: 3.591e-01]
Xiang_wan_2_2_lora:   7%|##6                                    | 2051/30000 [4:37:59<47:54:11,  6.17s/it, lr: 1.0e-04 loss: 3.591e-01]
Xiang_wan_2_2_lora:   7%|##6                                    | 2051/30000 [4:38:06<47:54:11,  6.17s/it, lr: 1.0e-04 loss: 1.096e-01]
Xiang_wan_2_2_lora:   7%|##6                                    | 2051/30000 [4:38:06<47:54:11,  6.17s/it, lr: 1.0e-04 loss: 1.096e-01]
Xiang_wan_2_2_lora:   7%|##6                                    | 2052/30000 [4:38:06<49:17:44,  6.35s/it, lr: 1.0e-04 loss: 1.096e-01]
Xiang_wan_2_2_lora:   7%|##6                                    | 2052/30000 [4:38:06<49:17:44,  6.35s/it, lr: 1.0e-04 loss: 1.096e-01]
Xiang_wan_2_2_lora:   7%|##6                                    | 2052/30000 [4:38:13<49:17:44,  6.35s/it, lr: 1.0e-04 loss: 8.341e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2052/30000 [4:38:13<49:17:44,  6.35s/it, lr: 1.0e-04 loss: 8.341e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2053/30000 [4:38:13<49:49:12,  6.42s/it, lr: 1.0e-04 loss: 8.341e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2053/30000 [4:38:13<49:49:12,  6.42s/it, lr: 1.0e-04 loss: 8.341e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2053/30000 [4:38:22<49:49:12,  6.42s/it, lr: 1.0e-04 loss: 5.799e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2053/30000 [4:38:22<49:49:12,  6.42s/it, lr: 1.0e-04 loss: 5.799e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2054/30000 [4:38:22<55:56:33,  7.21s/it, lr: 1.0e-04 loss: 5.799e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2054/30000 [4:38:22<55:56:33,  7.21s/it, lr: 1.0e-04 loss: 5.799e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2054/30000 [4:38:29<55:56:33,  7.21s/it, lr: 1.0e-04 loss: 7.546e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2054/30000 [4:38:29<55:56:33,  7.21s/it, lr: 1.0e-04 loss: 7.546e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2055/30000 [4:38:29<54:40:13,  7.04s/it, lr: 1.0e-04 loss: 7.546e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2055/30000 [4:38:29<54:40:13,  7.04s/it, lr: 1.0e-04 loss: 7.546e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2055/30000 [4:38:38<54:40:13,  7.04s/it, lr: 1.0e-04 loss: 5.670e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2055/30000 [4:38:38<54:40:13,  7.04s/it, lr: 1.0e-04 loss: 5.670e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2056/30000 [4:38:38<59:39:34,  7.69s/it, lr: 1.0e-04 loss: 5.670e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2056/30000 [4:38:38<59:39:34,  7.69s/it, lr: 1.0e-04 loss: 5.670e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2056/30000 [4:38:44<59:39:34,  7.69s/it, lr: 1.0e-04 loss: 6.257e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2056/30000 [4:38:44<59:39:34,  7.69s/it, lr: 1.0e-04 loss: 6.257e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2057/30000 [4:38:44<57:18:12,  7.38s/it, lr: 1.0e-04 loss: 6.257e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2057/30000 [4:38:44<57:18:12,  7.38s/it, lr: 1.0e-04 loss: 6.257e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2057/30000 [4:38:54<57:18:12,  7.38s/it, lr: 1.0e-04 loss: 8.580e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2057/30000 [4:38:54<57:18:12,  7.38s/it, lr: 1.0e-04 loss: 8.580e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2058/30000 [4:38:54<61:48:07,  7.96s/it, lr: 1.0e-04 loss: 8.580e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2058/30000 [4:38:54<61:48:07,  7.96s/it, lr: 1.0e-04 loss: 8.580e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2058/30000 [4:39:03<61:48:07,  7.96s/it, lr: 1.0e-04 loss: 5.761e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2058/30000 [4:39:03<61:48:07,  7.96s/it, lr: 1.0e-04 loss: 5.761e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2059/30000 [4:39:03<65:05:07,  8.39s/it, lr: 1.0e-04 loss: 5.761e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2059/30000 [4:39:03<65:05:07,  8.39s/it, lr: 1.0e-04 loss: 5.761e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2059/30000 [4:39:10<65:05:07,  8.39s/it, lr: 1.0e-04 loss: 7.314e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2059/30000 [4:39:10<65:05:07,  8.39s/it, lr: 1.0e-04 loss: 7.314e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2060/30000 [4:39:17<65:04:59,  8.39s/it, lr: 1.0e-04 loss: 9.326e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2060/30000 [4:39:17<65:04:59,  8.39s/it, lr: 1.0e-04 loss: 9.326e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2061/30000 [4:39:17<47:02:06,  6.06s/it, lr: 1.0e-04 loss: 9.326e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2061/30000 [4:39:17<47:02:06,  6.06s/it, lr: 1.0e-04 loss: 9.326e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2061/30000 [4:39:23<47:02:06,  6.06s/it, lr: 1.0e-04 loss: 7.121e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2061/30000 [4:39:23<47:02:06,  6.06s/it, lr: 1.0e-04 loss: 7.121e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2062/30000 [4:39:23<47:58:55,  6.18s/it, lr: 1.0e-04 loss: 7.121e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2062/30000 [4:39:23<47:58:55,  6.18s/it, lr: 1.0e-04 loss: 7.121e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2062/30000 [4:39:30<47:58:55,  6.18s/it, lr: 1.0e-04 loss: 6.342e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2062/30000 [4:39:30<47:58:55,  6.18s/it, lr: 1.0e-04 loss: 6.342e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2063/30000 [4:39:30<48:49:17,  6.29s/it, lr: 1.0e-04 loss: 6.342e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2063/30000 [4:39:30<48:49:17,  6.29s/it, lr: 1.0e-04 loss: 6.342e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2063/30000 [4:39:36<48:49:17,  6.29s/it, lr: 1.0e-04 loss: 1.106e-01]
Xiang_wan_2_2_lora:   7%|##6                                    | 2063/30000 [4:39:36<48:49:17,  6.29s/it, lr: 1.0e-04 loss: 1.106e-01]
Xiang_wan_2_2_lora:   7%|##6                                    | 2064/30000 [4:39:36<49:26:04,  6.37s/it, lr: 1.0e-04 loss: 1.106e-01]
Xiang_wan_2_2_lora:   7%|##6                                    | 2064/30000 [4:39:36<49:26:04,  6.37s/it, lr: 1.0e-04 loss: 1.106e-01]
Xiang_wan_2_2_lora:   7%|##6                                    | 2064/30000 [4:39:43<49:26:04,  6.37s/it, lr: 1.0e-04 loss: 7.321e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2064/30000 [4:39:43<49:26:04,  6.37s/it, lr: 1.0e-04 loss: 7.321e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2065/30000 [4:39:43<49:52:32,  6.43s/it, lr: 1.0e-04 loss: 7.321e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2065/30000 [4:39:43<49:52:32,  6.43s/it, lr: 1.0e-04 loss: 7.321e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2065/30000 [4:39:50<49:52:32,  6.43s/it, lr: 1.0e-04 loss: 6.663e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2065/30000 [4:39:50<49:52:32,  6.43s/it, lr: 1.0e-04 loss: 6.663e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2066/30000 [4:39:50<50:19:43,  6.49s/it, lr: 1.0e-04 loss: 6.663e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2066/30000 [4:39:50<50:19:43,  6.49s/it, lr: 1.0e-04 loss: 6.663e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2066/30000 [4:39:57<50:19:43,  6.49s/it, lr: 1.0e-04 loss: 8.572e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2066/30000 [4:39:57<50:19:43,  6.49s/it, lr: 1.0e-04 loss: 8.572e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2067/30000 [4:39:57<51:22:14,  6.62s/it, lr: 1.0e-04 loss: 8.572e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2067/30000 [4:39:57<51:22:14,  6.62s/it, lr: 1.0e-04 loss: 8.572e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2067/30000 [4:40:06<51:22:14,  6.62s/it, lr: 1.0e-04 loss: 8.582e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2067/30000 [4:40:06<51:22:14,  6.62s/it, lr: 1.0e-04 loss: 8.582e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2068/30000 [4:40:06<57:27:10,  7.40s/it, lr: 1.0e-04 loss: 8.582e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2068/30000 [4:40:06<57:27:10,  7.40s/it, lr: 1.0e-04 loss: 8.582e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2068/30000 [4:40:15<57:27:10,  7.40s/it, lr: 1.0e-04 loss: 2.425e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2068/30000 [4:40:15<57:27:10,  7.40s/it, lr: 1.0e-04 loss: 2.425e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2069/30000 [4:40:15<62:00:51,  7.99s/it, lr: 1.0e-04 loss: 2.425e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2069/30000 [4:40:15<62:00:51,  7.99s/it, lr: 1.0e-04 loss: 2.425e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2069/30000 [4:40:25<62:00:51,  7.99s/it, lr: 1.0e-04 loss: 8.184e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2069/30000 [4:40:25<62:00:51,  7.99s/it, lr: 1.0e-04 loss: 8.184e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2070/30000 [4:40:34<62:00:43,  7.99s/it, lr: 1.0e-04 loss: 5.228e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2070/30000 [4:40:34<62:00:43,  7.99s/it, lr: 1.0e-04 loss: 5.228e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2071/30000 [4:40:34<50:14:39,  6.48s/it, lr: 1.0e-04 loss: 5.228e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2071/30000 [4:40:34<50:14:39,  6.48s/it, lr: 1.0e-04 loss: 5.228e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2071/30000 [4:40:44<50:14:39,  6.48s/it, lr: 1.0e-04 loss: 8.773e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2071/30000 [4:40:44<50:14:39,  6.48s/it, lr: 1.0e-04 loss: 8.773e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2072/30000 [4:40:44<55:49:09,  7.20s/it, lr: 1.0e-04 loss: 8.773e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2072/30000 [4:40:44<55:49:09,  7.20s/it, lr: 1.0e-04 loss: 8.773e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2072/30000 [4:40:53<55:49:09,  7.20s/it, lr: 1.0e-04 loss: 1.213e-01]
Xiang_wan_2_2_lora:   7%|##6                                    | 2072/30000 [4:40:53<55:49:09,  7.20s/it, lr: 1.0e-04 loss: 1.213e-01]
Xiang_wan_2_2_lora:   7%|##6                                    | 2073/30000 [4:40:53<60:21:00,  7.78s/it, lr: 1.0e-04 loss: 1.213e-01]
Xiang_wan_2_2_lora:   7%|##6                                    | 2073/30000 [4:40:53<60:21:00,  7.78s/it, lr: 1.0e-04 loss: 1.213e-01]
Xiang_wan_2_2_lora:   7%|##6                                    | 2073/30000 [4:41:02<60:21:00,  7.78s/it, lr: 1.0e-04 loss: 1.063e-01]
Xiang_wan_2_2_lora:   7%|##6                                    | 2073/30000 [4:41:02<60:21:00,  7.78s/it, lr: 1.0e-04 loss: 1.063e-01]
Xiang_wan_2_2_lora:   7%|##6                                    | 2074/30000 [4:41:02<63:42:11,  8.21s/it, lr: 1.0e-04 loss: 1.063e-01]
Xiang_wan_2_2_lora:   7%|##6                                    | 2074/30000 [4:41:02<63:42:11,  8.21s/it, lr: 1.0e-04 loss: 1.063e-01]
Xiang_wan_2_2_lora:   7%|##6                                    | 2074/30000 [4:41:09<63:42:11,  8.21s/it, lr: 1.0e-04 loss: 9.229e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2074/30000 [4:41:09<63:42:11,  8.21s/it, lr: 1.0e-04 loss: 9.229e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2075/30000 [4:41:09<60:52:17,  7.85s/it, lr: 1.0e-04 loss: 9.229e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2075/30000 [4:41:09<60:52:17,  7.85s/it, lr: 1.0e-04 loss: 9.229e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2075/30000 [4:41:19<60:52:17,  7.85s/it, lr: 1.0e-04 loss: 6.703e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2075/30000 [4:41:19<60:52:17,  7.85s/it, lr: 1.0e-04 loss: 6.703e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2076/30000 [4:41:19<64:07:54,  8.27s/it, lr: 1.0e-04 loss: 6.703e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2076/30000 [4:41:19<64:07:54,  8.27s/it, lr: 1.0e-04 loss: 6.703e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2076/30000 [4:41:28<64:07:54,  8.27s/it, lr: 1.0e-04 loss: 5.642e-02]
Xiang_wan_2_2_lora:   7%|##6                                    | 2076/30000 [4:41:28<64:07:54,  8.27s/it, lr: 1.0e-04 loss: 5.642e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2077/30000 [4:41:28<66:33:43,  8.58s/it, lr: 1.0e-04 loss: 5.642e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2077/30000 [4:41:28<66:33:43,  8.58s/it, lr: 1.0e-04 loss: 5.642e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2077/30000 [4:41:37<66:33:43,  8.58s/it, lr: 1.0e-04 loss: 5.251e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2077/30000 [4:41:37<66:33:43,  8.58s/it, lr: 1.0e-04 loss: 5.251e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2078/30000 [4:41:37<68:17:59,  8.81s/it, lr: 1.0e-04 loss: 5.251e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2078/30000 [4:41:37<68:17:59,  8.81s/it, lr: 1.0e-04 loss: 5.251e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2078/30000 [4:41:47<68:17:59,  8.81s/it, lr: 1.0e-04 loss: 2.187e-01]
Xiang_wan_2_2_lora:   7%|##7                                    | 2078/30000 [4:41:47<68:17:59,  8.81s/it, lr: 1.0e-04 loss: 2.187e-01]
Xiang_wan_2_2_lora:   7%|##7                                    | 2079/30000 [4:41:47<69:39:58,  8.98s/it, lr: 1.0e-04 loss: 2.187e-01]
Xiang_wan_2_2_lora:   7%|##7                                    | 2079/30000 [4:41:47<69:39:58,  8.98s/it, lr: 1.0e-04 loss: 2.187e-01]
Xiang_wan_2_2_lora:   7%|##7                                    | 2079/30000 [4:41:56<69:39:58,  8.98s/it, lr: 1.0e-04 loss: 8.831e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2079/30000 [4:41:56<69:39:58,  8.98s/it, lr: 1.0e-04 loss: 8.831e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2080/30000 [4:42:05<69:39:49,  8.98s/it, lr: 1.0e-04 loss: 3.072e-01]
Xiang_wan_2_2_lora:   7%|##7                                    | 2080/30000 [4:42:05<69:39:49,  8.98s/it, lr: 1.0e-04 loss: 3.072e-01]
Xiang_wan_2_2_lora:   7%|##7                                    | 2081/30000 [4:42:05<54:24:49,  7.02s/it, lr: 1.0e-04 loss: 3.072e-01]
Xiang_wan_2_2_lora:   7%|##7                                    | 2081/30000 [4:42:05<54:24:49,  7.02s/it, lr: 1.0e-04 loss: 3.072e-01]
Xiang_wan_2_2_lora:   7%|##7                                    | 2081/30000 [4:42:15<54:24:49,  7.02s/it, lr: 1.0e-04 loss: 5.222e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2081/30000 [4:42:15<54:24:49,  7.02s/it, lr: 1.0e-04 loss: 5.222e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2082/30000 [4:42:15<59:17:40,  7.65s/it, lr: 1.0e-04 loss: 5.222e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2082/30000 [4:42:15<59:17:40,  7.65s/it, lr: 1.0e-04 loss: 5.222e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2082/30000 [4:42:24<59:17:40,  7.65s/it, lr: 1.0e-04 loss: 3.818e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2082/30000 [4:42:24<59:17:40,  7.65s/it, lr: 1.0e-04 loss: 3.818e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2083/30000 [4:42:24<62:46:15,  8.09s/it, lr: 1.0e-04 loss: 3.818e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2083/30000 [4:42:24<62:46:15,  8.09s/it, lr: 1.0e-04 loss: 3.818e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2083/30000 [4:42:34<62:46:15,  8.09s/it, lr: 1.0e-04 loss: 6.762e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2083/30000 [4:42:34<62:46:15,  8.09s/it, lr: 1.0e-04 loss: 6.762e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2084/30000 [4:42:34<65:30:20,  8.45s/it, lr: 1.0e-04 loss: 6.762e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2084/30000 [4:42:34<65:30:20,  8.45s/it, lr: 1.0e-04 loss: 6.762e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2084/30000 [4:42:40<65:30:20,  8.45s/it, lr: 1.0e-04 loss: 2.635e-01]
Xiang_wan_2_2_lora:   7%|##7                                    | 2084/30000 [4:42:40<65:30:20,  8.45s/it, lr: 1.0e-04 loss: 2.635e-01]
Xiang_wan_2_2_lora:   7%|##7                                    | 2085/30000 [4:42:40<61:30:46,  7.93s/it, lr: 1.0e-04 loss: 2.635e-01]
Xiang_wan_2_2_lora:   7%|##7                                    | 2085/30000 [4:42:40<61:30:46,  7.93s/it, lr: 1.0e-04 loss: 2.635e-01]
Xiang_wan_2_2_lora:   7%|##7                                    | 2085/30000 [4:42:47<61:30:46,  7.93s/it, lr: 1.0e-04 loss: 9.018e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2085/30000 [4:42:47<61:30:46,  7.93s/it, lr: 1.0e-04 loss: 9.018e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2086/30000 [4:42:47<58:25:48,  7.54s/it, lr: 1.0e-04 loss: 9.018e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2086/30000 [4:42:47<58:25:48,  7.54s/it, lr: 1.0e-04 loss: 9.018e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2086/30000 [4:42:54<58:25:48,  7.54s/it, lr: 1.0e-04 loss: 7.641e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2086/30000 [4:42:54<58:25:48,  7.54s/it, lr: 1.0e-04 loss: 7.641e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2087/30000 [4:42:54<56:19:04,  7.26s/it, lr: 1.0e-04 loss: 7.641e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2087/30000 [4:42:54<56:19:04,  7.26s/it, lr: 1.0e-04 loss: 7.641e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2087/30000 [4:43:00<56:19:04,  7.26s/it, lr: 1.0e-04 loss: 9.617e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2087/30000 [4:43:00<56:19:04,  7.26s/it, lr: 1.0e-04 loss: 9.617e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2088/30000 [4:43:00<54:50:20,  7.07s/it, lr: 1.0e-04 loss: 9.617e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2088/30000 [4:43:00<54:50:20,  7.07s/it, lr: 1.0e-04 loss: 9.617e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2088/30000 [4:43:10<54:50:20,  7.07s/it, lr: 1.0e-04 loss: 5.610e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2088/30000 [4:43:10<54:50:20,  7.07s/it, lr: 1.0e-04 loss: 5.610e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2089/30000 [4:43:10<60:23:20,  7.79s/it, lr: 1.0e-04 loss: 5.610e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2089/30000 [4:43:10<60:23:20,  7.79s/it, lr: 1.0e-04 loss: 5.610e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2089/30000 [4:43:19<60:23:20,  7.79s/it, lr: 1.0e-04 loss: 4.911e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2089/30000 [4:43:19<60:23:20,  7.79s/it, lr: 1.0e-04 loss: 4.911e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2090/30000 [4:43:26<60:23:12,  7.79s/it, lr: 1.0e-04 loss: 6.624e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2090/30000 [4:43:26<60:23:12,  7.79s/it, lr: 1.0e-04 loss: 6.624e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2091/30000 [4:43:26<44:31:15,  5.74s/it, lr: 1.0e-04 loss: 6.624e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2091/30000 [4:43:26<44:31:15,  5.74s/it, lr: 1.0e-04 loss: 6.624e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2091/30000 [4:43:32<44:31:15,  5.74s/it, lr: 1.0e-04 loss: 1.119e-01]
Xiang_wan_2_2_lora:   7%|##7                                    | 2091/30000 [4:43:32<44:31:15,  5.74s/it, lr: 1.0e-04 loss: 1.119e-01]
Xiang_wan_2_2_lora:   7%|##7                                    | 2092/30000 [4:43:32<46:11:03,  5.96s/it, lr: 1.0e-04 loss: 1.119e-01]
Xiang_wan_2_2_lora:   7%|##7                                    | 2092/30000 [4:43:32<46:11:03,  5.96s/it, lr: 1.0e-04 loss: 1.119e-01]
Xiang_wan_2_2_lora:   7%|##7                                    | 2092/30000 [4:43:39<46:11:03,  5.96s/it, lr: 1.0e-04 loss: 5.658e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2092/30000 [4:43:39<46:11:03,  5.96s/it, lr: 1.0e-04 loss: 5.658e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2093/30000 [4:43:39<47:28:48,  6.12s/it, lr: 1.0e-04 loss: 5.658e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2093/30000 [4:43:39<47:28:48,  6.12s/it, lr: 1.0e-04 loss: 5.658e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2093/30000 [4:43:46<47:28:48,  6.12s/it, lr: 1.0e-04 loss: 5.822e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2093/30000 [4:43:46<47:28:48,  6.12s/it, lr: 1.0e-04 loss: 5.822e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2094/30000 [4:43:46<48:33:40,  6.26s/it, lr: 1.0e-04 loss: 5.822e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2094/30000 [4:43:46<48:33:40,  6.26s/it, lr: 1.0e-04 loss: 5.822e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2094/30000 [4:43:55<48:33:40,  6.26s/it, lr: 1.0e-04 loss: 4.940e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2094/30000 [4:43:55<48:33:40,  6.26s/it, lr: 1.0e-04 loss: 4.940e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2095/30000 [4:43:55<55:08:41,  7.11s/it, lr: 1.0e-04 loss: 4.940e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2095/30000 [4:43:55<55:08:41,  7.11s/it, lr: 1.0e-04 loss: 4.940e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2095/30000 [4:44:01<55:08:41,  7.11s/it, lr: 1.0e-04 loss: 9.047e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2095/30000 [4:44:01<55:08:41,  7.11s/it, lr: 1.0e-04 loss: 9.047e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2096/30000 [4:44:01<54:03:59,  6.98s/it, lr: 1.0e-04 loss: 9.047e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2096/30000 [4:44:01<54:03:59,  6.98s/it, lr: 1.0e-04 loss: 9.047e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2096/30000 [4:44:08<54:03:59,  6.98s/it, lr: 1.0e-04 loss: 6.817e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2096/30000 [4:44:08<54:03:59,  6.98s/it, lr: 1.0e-04 loss: 6.817e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2097/30000 [4:44:08<53:45:27,  6.94s/it, lr: 1.0e-04 loss: 6.817e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2097/30000 [4:44:08<53:45:27,  6.94s/it, lr: 1.0e-04 loss: 6.817e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2097/30000 [4:44:18<53:45:27,  6.94s/it, lr: 1.0e-04 loss: 1.026e-01]
Xiang_wan_2_2_lora:   7%|##7                                    | 2097/30000 [4:44:18<53:45:27,  6.94s/it, lr: 1.0e-04 loss: 1.026e-01]
Xiang_wan_2_2_lora:   7%|##7                                    | 2098/30000 [4:44:18<59:11:05,  7.64s/it, lr: 1.0e-04 loss: 1.026e-01]
Xiang_wan_2_2_lora:   7%|##7                                    | 2098/30000 [4:44:18<59:11:05,  7.64s/it, lr: 1.0e-04 loss: 1.026e-01]
Xiang_wan_2_2_lora:   7%|##7                                    | 2098/30000 [4:44:24<59:11:05,  7.64s/it, lr: 1.0e-04 loss: 9.272e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2098/30000 [4:44:24<59:11:05,  7.64s/it, lr: 1.0e-04 loss: 9.272e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2099/30000 [4:44:24<56:44:11,  7.32s/it, lr: 1.0e-04 loss: 9.272e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2099/30000 [4:44:24<56:44:11,  7.32s/it, lr: 1.0e-04 loss: 9.272e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2099/30000 [4:44:34<56:44:11,  7.32s/it, lr: 1.0e-04 loss: 5.227e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2099/30000 [4:44:34<56:44:11,  7.32s/it, lr: 1.0e-04 loss: 5.227e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2100/30000 [4:44:40<56:44:03,  7.32s/it, lr: 1.0e-04 loss: 6.087e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2100/30000 [4:44:40<56:44:03,  7.32s/it, lr: 1.0e-04 loss: 6.087e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2101/30000 [4:44:40<42:27:25,  5.48s/it, lr: 1.0e-04 loss: 6.087e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2101/30000 [4:44:40<42:27:25,  5.48s/it, lr: 1.0e-04 loss: 6.087e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2101/30000 [4:44:49<42:27:25,  5.48s/it, lr: 1.0e-04 loss: 5.199e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2101/30000 [4:44:49<42:27:25,  5.48s/it, lr: 1.0e-04 loss: 5.199e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2102/30000 [4:44:49<49:49:56,  6.43s/it, lr: 1.0e-04 loss: 5.199e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2102/30000 [4:44:49<49:49:56,  6.43s/it, lr: 1.0e-04 loss: 5.199e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2102/30000 [4:44:59<49:49:56,  6.43s/it, lr: 1.0e-04 loss: 4.141e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2102/30000 [4:44:59<49:49:56,  6.43s/it, lr: 1.0e-04 loss: 4.141e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2103/30000 [4:44:59<55:34:56,  7.17s/it, lr: 1.0e-04 loss: 4.141e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2103/30000 [4:44:59<55:34:56,  7.17s/it, lr: 1.0e-04 loss: 4.141e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2103/30000 [4:45:06<55:34:56,  7.17s/it, lr: 1.0e-04 loss: 7.309e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2103/30000 [4:45:06<55:34:56,  7.17s/it, lr: 1.0e-04 loss: 7.309e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2104/30000 [4:45:06<55:00:33,  7.10s/it, lr: 1.0e-04 loss: 7.309e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2104/30000 [4:45:06<55:00:33,  7.10s/it, lr: 1.0e-04 loss: 7.309e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2104/30000 [4:45:12<55:00:33,  7.10s/it, lr: 1.0e-04 loss: 7.983e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2104/30000 [4:45:12<55:00:33,  7.10s/it, lr: 1.0e-04 loss: 7.983e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2105/30000 [4:45:12<53:48:26,  6.94s/it, lr: 1.0e-04 loss: 7.983e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2105/30000 [4:45:12<53:48:26,  6.94s/it, lr: 1.0e-04 loss: 7.983e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2105/30000 [4:45:22<53:48:26,  6.94s/it, lr: 1.0e-04 loss: 5.567e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2105/30000 [4:45:22<53:48:26,  6.94s/it, lr: 1.0e-04 loss: 5.567e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2106/30000 [4:45:22<59:14:49,  7.65s/it, lr: 1.0e-04 loss: 5.567e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2106/30000 [4:45:22<59:14:49,  7.65s/it, lr: 1.0e-04 loss: 5.567e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2106/30000 [4:45:28<59:14:49,  7.65s/it, lr: 1.0e-04 loss: 5.658e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2106/30000 [4:45:28<59:14:49,  7.65s/it, lr: 1.0e-04 loss: 5.658e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2107/30000 [4:45:28<56:47:03,  7.33s/it, lr: 1.0e-04 loss: 5.658e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2107/30000 [4:45:28<56:47:03,  7.33s/it, lr: 1.0e-04 loss: 5.658e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2107/30000 [4:45:38<56:47:03,  7.33s/it, lr: 1.0e-04 loss: 6.979e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2107/30000 [4:45:38<56:47:03,  7.33s/it, lr: 1.0e-04 loss: 6.979e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2108/30000 [4:45:38<61:22:49,  7.92s/it, lr: 1.0e-04 loss: 6.979e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2108/30000 [4:45:38<61:22:49,  7.92s/it, lr: 1.0e-04 loss: 6.979e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2108/30000 [4:45:47<61:22:49,  7.92s/it, lr: 1.0e-04 loss: 5.709e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2108/30000 [4:45:47<61:22:49,  7.92s/it, lr: 1.0e-04 loss: 5.709e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2109/30000 [4:45:47<64:28:09,  8.32s/it, lr: 1.0e-04 loss: 5.709e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2109/30000 [4:45:47<64:28:09,  8.32s/it, lr: 1.0e-04 loss: 5.709e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2109/30000 [4:45:56<64:28:09,  8.32s/it, lr: 1.0e-04 loss: 6.863e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2109/30000 [4:45:56<64:28:09,  8.32s/it, lr: 1.0e-04 loss: 6.863e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2110/30000 [4:46:06<64:28:01,  8.32s/it, lr: 1.0e-04 loss: 5.298e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2110/30000 [4:46:06<64:28:01,  8.32s/it, lr: 1.0e-04 loss: 5.298e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2111/30000 [4:46:06<51:48:37,  6.69s/it, lr: 1.0e-04 loss: 5.298e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2111/30000 [4:46:06<51:48:37,  6.69s/it, lr: 1.0e-04 loss: 5.298e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2111/30000 [4:46:15<51:48:37,  6.69s/it, lr: 1.0e-04 loss: 6.791e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2111/30000 [4:46:15<51:48:37,  6.69s/it, lr: 1.0e-04 loss: 6.791e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2112/30000 [4:46:15<56:54:29,  7.35s/it, lr: 1.0e-04 loss: 6.791e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2112/30000 [4:46:15<56:54:29,  7.35s/it, lr: 1.0e-04 loss: 6.791e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2112/30000 [4:46:22<56:54:29,  7.35s/it, lr: 1.0e-04 loss: 6.773e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2112/30000 [4:46:22<56:54:29,  7.35s/it, lr: 1.0e-04 loss: 6.773e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2113/30000 [4:46:22<55:18:24,  7.14s/it, lr: 1.0e-04 loss: 6.773e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2113/30000 [4:46:22<55:18:24,  7.14s/it, lr: 1.0e-04 loss: 6.773e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2113/30000 [4:46:28<55:18:24,  7.14s/it, lr: 1.0e-04 loss: 6.676e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2113/30000 [4:46:28<55:18:24,  7.14s/it, lr: 1.0e-04 loss: 6.676e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2114/30000 [4:46:28<54:17:11,  7.01s/it, lr: 1.0e-04 loss: 6.676e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2114/30000 [4:46:28<54:17:11,  7.01s/it, lr: 1.0e-04 loss: 6.676e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2114/30000 [4:46:35<54:17:11,  7.01s/it, lr: 1.0e-04 loss: 5.317e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2114/30000 [4:46:35<54:17:11,  7.01s/it, lr: 1.0e-04 loss: 5.317e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2115/30000 [4:46:35<53:23:39,  6.89s/it, lr: 1.0e-04 loss: 5.317e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2115/30000 [4:46:35<53:23:39,  6.89s/it, lr: 1.0e-04 loss: 5.317e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2115/30000 [4:46:44<53:23:39,  6.89s/it, lr: 1.0e-04 loss: 7.718e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2115/30000 [4:46:44<53:23:39,  6.89s/it, lr: 1.0e-04 loss: 7.718e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2116/30000 [4:46:44<58:53:17,  7.60s/it, lr: 1.0e-04 loss: 7.718e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2116/30000 [4:46:44<58:53:17,  7.60s/it, lr: 1.0e-04 loss: 7.718e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2116/30000 [4:46:51<58:53:17,  7.60s/it, lr: 1.0e-04 loss: 7.499e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2116/30000 [4:46:51<58:53:17,  7.60s/it, lr: 1.0e-04 loss: 7.499e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2117/30000 [4:46:51<56:32:44,  7.30s/it, lr: 1.0e-04 loss: 7.499e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2117/30000 [4:46:51<56:32:44,  7.30s/it, lr: 1.0e-04 loss: 7.499e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2117/30000 [4:47:00<56:32:44,  7.30s/it, lr: 1.0e-04 loss: 8.620e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2117/30000 [4:47:00<56:32:44,  7.30s/it, lr: 1.0e-04 loss: 8.620e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2118/30000 [4:47:00<61:18:40,  7.92s/it, lr: 1.0e-04 loss: 8.620e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2118/30000 [4:47:00<61:18:40,  7.92s/it, lr: 1.0e-04 loss: 8.620e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2118/30000 [4:47:10<61:18:40,  7.92s/it, lr: 1.0e-04 loss: 4.591e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2118/30000 [4:47:10<61:18:40,  7.92s/it, lr: 1.0e-04 loss: 4.591e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2119/30000 [4:47:10<65:07:15,  8.41s/it, lr: 1.0e-04 loss: 4.591e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2119/30000 [4:47:10<65:07:15,  8.41s/it, lr: 1.0e-04 loss: 4.591e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2119/30000 [4:47:19<65:07:15,  8.41s/it, lr: 1.0e-04 loss: 6.687e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2119/30000 [4:47:19<65:07:15,  8.41s/it, lr: 1.0e-04 loss: 6.687e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2120/30000 [4:47:29<65:07:06,  8.41s/it, lr: 1.0e-04 loss: 8.031e-01]
Xiang_wan_2_2_lora:   7%|##7                                    | 2120/30000 [4:47:29<65:07:06,  8.41s/it, lr: 1.0e-04 loss: 8.031e-01]
Xiang_wan_2_2_lora:   7%|##7                                    | 2121/30000 [4:47:29<51:49:00,  6.69s/it, lr: 1.0e-04 loss: 8.031e-01]
Xiang_wan_2_2_lora:   7%|##7                                    | 2121/30000 [4:47:29<51:49:00,  6.69s/it, lr: 1.0e-04 loss: 8.031e-01]
Xiang_wan_2_2_lora:   7%|##7                                    | 2121/30000 [4:47:38<51:49:00,  6.69s/it, lr: 1.0e-04 loss: 7.047e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2121/30000 [4:47:38<51:49:00,  6.69s/it, lr: 1.0e-04 loss: 7.047e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2122/30000 [4:47:38<56:54:34,  7.35s/it, lr: 1.0e-04 loss: 7.047e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2122/30000 [4:47:38<56:54:34,  7.35s/it, lr: 1.0e-04 loss: 7.047e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2122/30000 [4:47:44<56:54:34,  7.35s/it, lr: 1.0e-04 loss: 7.424e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2122/30000 [4:47:44<56:54:34,  7.35s/it, lr: 1.0e-04 loss: 7.424e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2123/30000 [4:47:44<55:20:01,  7.15s/it, lr: 1.0e-04 loss: 7.424e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2123/30000 [4:47:44<55:20:01,  7.15s/it, lr: 1.0e-04 loss: 7.424e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2123/30000 [4:47:51<55:20:01,  7.15s/it, lr: 1.0e-04 loss: 1.071e-01]
Xiang_wan_2_2_lora:   7%|##7                                    | 2123/30000 [4:47:51<55:20:01,  7.15s/it, lr: 1.0e-04 loss: 1.071e-01]
Xiang_wan_2_2_lora:   7%|##7                                    | 2124/30000 [4:47:51<54:12:18,  7.00s/it, lr: 1.0e-04 loss: 1.071e-01]
Xiang_wan_2_2_lora:   7%|##7                                    | 2124/30000 [4:47:51<54:12:18,  7.00s/it, lr: 1.0e-04 loss: 1.071e-01]
Xiang_wan_2_2_lora:   7%|##7                                    | 2124/30000 [4:47:58<54:12:18,  7.00s/it, lr: 1.0e-04 loss: 9.119e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2124/30000 [4:47:58<54:12:18,  7.00s/it, lr: 1.0e-04 loss: 9.119e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2125/30000 [4:47:58<53:15:15,  6.88s/it, lr: 1.0e-04 loss: 9.119e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2125/30000 [4:47:58<53:15:15,  6.88s/it, lr: 1.0e-04 loss: 9.119e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2125/30000 [4:48:05<53:15:15,  6.88s/it, lr: 1.0e-04 loss: 9.432e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2125/30000 [4:48:05<53:15:15,  6.88s/it, lr: 1.0e-04 loss: 9.432e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2126/30000 [4:48:05<53:17:31,  6.88s/it, lr: 1.0e-04 loss: 9.432e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2126/30000 [4:48:05<53:17:31,  6.88s/it, lr: 1.0e-04 loss: 9.432e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2126/30000 [4:48:14<53:17:31,  6.88s/it, lr: 1.0e-04 loss: 3.541e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2126/30000 [4:48:14<53:17:31,  6.88s/it, lr: 1.0e-04 loss: 3.541e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2127/30000 [4:48:14<58:33:20,  7.56s/it, lr: 1.0e-04 loss: 3.541e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2127/30000 [4:48:14<58:33:20,  7.56s/it, lr: 1.0e-04 loss: 3.541e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2127/30000 [4:48:20<58:33:20,  7.56s/it, lr: 1.0e-04 loss: 9.436e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2127/30000 [4:48:20<58:33:20,  7.56s/it, lr: 1.0e-04 loss: 9.436e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2128/30000 [4:48:20<56:24:52,  7.29s/it, lr: 1.0e-04 loss: 9.436e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2128/30000 [4:48:20<56:24:52,  7.29s/it, lr: 1.0e-04 loss: 9.436e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2128/30000 [4:48:30<56:24:52,  7.29s/it, lr: 1.0e-04 loss: 6.765e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2128/30000 [4:48:30<56:24:52,  7.29s/it, lr: 1.0e-04 loss: 6.765e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2129/30000 [4:48:30<60:57:29,  7.87s/it, lr: 1.0e-04 loss: 6.765e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2129/30000 [4:48:30<60:57:29,  7.87s/it, lr: 1.0e-04 loss: 6.765e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2129/30000 [4:48:39<60:57:29,  7.87s/it, lr: 1.0e-04 loss: 6.385e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2129/30000 [4:48:39<60:57:29,  7.87s/it, lr: 1.0e-04 loss: 6.385e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2130/30000 [4:48:48<60:57:22,  7.87s/it, lr: 1.0e-04 loss: 4.868e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2130/30000 [4:48:48<60:57:22,  7.87s/it, lr: 1.0e-04 loss: 4.868e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2131/30000 [4:48:48<49:31:42,  6.40s/it, lr: 1.0e-04 loss: 4.868e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2131/30000 [4:48:48<49:31:42,  6.40s/it, lr: 1.0e-04 loss: 4.868e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2131/30000 [4:48:58<49:31:42,  6.40s/it, lr: 1.0e-04 loss: 6.132e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2131/30000 [4:48:58<49:31:42,  6.40s/it, lr: 1.0e-04 loss: 6.132e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2132/30000 [4:48:58<55:11:18,  7.13s/it, lr: 1.0e-04 loss: 6.132e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2132/30000 [4:48:58<55:11:18,  7.13s/it, lr: 1.0e-04 loss: 6.132e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2132/30000 [4:49:04<55:11:18,  7.13s/it, lr: 1.0e-04 loss: 4.145e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2132/30000 [4:49:04<55:11:18,  7.13s/it, lr: 1.0e-04 loss: 4.145e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2133/30000 [4:49:04<54:27:27,  7.04s/it, lr: 1.0e-04 loss: 4.145e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2133/30000 [4:49:04<54:27:27,  7.04s/it, lr: 1.0e-04 loss: 4.145e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2133/30000 [4:49:11<54:27:27,  7.04s/it, lr: 1.0e-04 loss: 8.498e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2133/30000 [4:49:11<54:27:27,  7.04s/it, lr: 1.0e-04 loss: 8.498e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2134/30000 [4:49:11<53:31:00,  6.91s/it, lr: 1.0e-04 loss: 8.498e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2134/30000 [4:49:11<53:31:00,  6.91s/it, lr: 1.0e-04 loss: 8.498e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2134/30000 [4:49:18<53:31:00,  6.91s/it, lr: 1.0e-04 loss: 4.122e-01]
Xiang_wan_2_2_lora:   7%|##7                                    | 2134/30000 [4:49:18<53:31:00,  6.91s/it, lr: 1.0e-04 loss: 4.122e-01]
Xiang_wan_2_2_lora:   7%|##7                                    | 2135/30000 [4:49:18<52:41:45,  6.81s/it, lr: 1.0e-04 loss: 4.122e-01]
Xiang_wan_2_2_lora:   7%|##7                                    | 2135/30000 [4:49:18<52:41:45,  6.81s/it, lr: 1.0e-04 loss: 4.122e-01]
Xiang_wan_2_2_lora:   7%|##7                                    | 2135/30000 [4:49:24<52:41:45,  6.81s/it, lr: 1.0e-04 loss: 8.838e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2135/30000 [4:49:24<52:41:45,  6.81s/it, lr: 1.0e-04 loss: 8.838e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2136/30000 [4:49:24<52:18:20,  6.76s/it, lr: 1.0e-04 loss: 8.838e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2136/30000 [4:49:24<52:18:20,  6.76s/it, lr: 1.0e-04 loss: 8.838e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2136/30000 [4:49:31<52:18:20,  6.76s/it, lr: 1.0e-04 loss: 1.202e-01]
Xiang_wan_2_2_lora:   7%|##7                                    | 2136/30000 [4:49:31<52:18:20,  6.76s/it, lr: 1.0e-04 loss: 1.202e-01]
Xiang_wan_2_2_lora:   7%|##7                                    | 2137/30000 [4:49:31<51:54:49,  6.71s/it, lr: 1.0e-04 loss: 1.202e-01]
Xiang_wan_2_2_lora:   7%|##7                                    | 2137/30000 [4:49:31<51:54:49,  6.71s/it, lr: 1.0e-04 loss: 1.202e-01]
Xiang_wan_2_2_lora:   7%|##7                                    | 2137/30000 [4:49:37<51:54:49,  6.71s/it, lr: 1.0e-04 loss: 1.064e-01]
Xiang_wan_2_2_lora:   7%|##7                                    | 2137/30000 [4:49:37<51:54:49,  6.71s/it, lr: 1.0e-04 loss: 1.064e-01]
Xiang_wan_2_2_lora:   7%|##7                                    | 2138/30000 [4:49:37<51:45:07,  6.69s/it, lr: 1.0e-04 loss: 1.064e-01]
Xiang_wan_2_2_lora:   7%|##7                                    | 2138/30000 [4:49:37<51:45:07,  6.69s/it, lr: 1.0e-04 loss: 1.064e-01]
Xiang_wan_2_2_lora:   7%|##7                                    | 2138/30000 [4:49:44<51:45:07,  6.69s/it, lr: 1.0e-04 loss: 5.007e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2138/30000 [4:49:44<51:45:07,  6.69s/it, lr: 1.0e-04 loss: 5.007e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2139/30000 [4:49:44<51:22:03,  6.64s/it, lr: 1.0e-04 loss: 5.007e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2139/30000 [4:49:44<51:22:03,  6.64s/it, lr: 1.0e-04 loss: 5.007e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2139/30000 [4:49:53<51:22:03,  6.64s/it, lr: 1.0e-04 loss: 6.154e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2139/30000 [4:49:53<51:22:03,  6.64s/it, lr: 1.0e-04 loss: 6.154e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2140/30000 [4:50:03<51:21:57,  6.64s/it, lr: 1.0e-04 loss: 6.757e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2140/30000 [4:50:03<51:21:57,  6.64s/it, lr: 1.0e-04 loss: 6.757e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2141/30000 [4:50:03<44:39:50,  5.77s/it, lr: 1.0e-04 loss: 6.757e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2141/30000 [4:50:03<44:39:50,  5.77s/it, lr: 1.0e-04 loss: 6.757e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2141/30000 [4:50:12<44:39:50,  5.77s/it, lr: 1.0e-04 loss: 7.933e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2141/30000 [4:50:12<44:39:50,  5.77s/it, lr: 1.0e-04 loss: 7.933e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2142/30000 [4:50:12<51:33:00,  6.66s/it, lr: 1.0e-04 loss: 7.933e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2142/30000 [4:50:12<51:33:00,  6.66s/it, lr: 1.0e-04 loss: 7.933e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2142/30000 [4:50:21<51:33:00,  6.66s/it, lr: 1.0e-04 loss: 1.128e-01]
Xiang_wan_2_2_lora:   7%|##7                                    | 2142/30000 [4:50:21<51:33:00,  6.66s/it, lr: 1.0e-04 loss: 1.128e-01]
Xiang_wan_2_2_lora:   7%|##7                                    | 2143/30000 [4:50:21<56:52:32,  7.35s/it, lr: 1.0e-04 loss: 1.128e-01]
Xiang_wan_2_2_lora:   7%|##7                                    | 2143/30000 [4:50:21<56:52:32,  7.35s/it, lr: 1.0e-04 loss: 1.128e-01]
Xiang_wan_2_2_lora:   7%|##7                                    | 2143/30000 [4:50:31<56:52:32,  7.35s/it, lr: 1.0e-04 loss: 5.475e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2143/30000 [4:50:31<56:52:32,  7.35s/it, lr: 1.0e-04 loss: 5.475e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2144/30000 [4:50:31<61:04:42,  7.89s/it, lr: 1.0e-04 loss: 5.475e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2144/30000 [4:50:31<61:04:42,  7.89s/it, lr: 1.0e-04 loss: 5.475e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2144/30000 [4:50:37<61:04:42,  7.89s/it, lr: 1.0e-04 loss: 1.087e-01]
Xiang_wan_2_2_lora:   7%|##7                                    | 2144/30000 [4:50:37<61:04:42,  7.89s/it, lr: 1.0e-04 loss: 1.087e-01]
Xiang_wan_2_2_lora:   7%|##7                                    | 2145/30000 [4:50:37<58:11:36,  7.52s/it, lr: 1.0e-04 loss: 1.087e-01]
Xiang_wan_2_2_lora:   7%|##7                                    | 2145/30000 [4:50:37<58:11:36,  7.52s/it, lr: 1.0e-04 loss: 1.087e-01]
Xiang_wan_2_2_lora:   7%|##7                                    | 2145/30000 [4:50:47<58:11:36,  7.52s/it, lr: 1.0e-04 loss: 6.701e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2145/30000 [4:50:47<58:11:36,  7.52s/it, lr: 1.0e-04 loss: 6.701e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2146/30000 [4:50:47<62:09:58,  8.03s/it, lr: 1.0e-04 loss: 6.701e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2146/30000 [4:50:47<62:09:58,  8.03s/it, lr: 1.0e-04 loss: 6.701e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2146/30000 [4:50:56<62:09:58,  8.03s/it, lr: 1.0e-04 loss: 8.397e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2146/30000 [4:50:56<62:09:58,  8.03s/it, lr: 1.0e-04 loss: 8.397e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2147/30000 [4:50:56<65:06:13,  8.41s/it, lr: 1.0e-04 loss: 8.397e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2147/30000 [4:50:56<65:06:13,  8.41s/it, lr: 1.0e-04 loss: 8.397e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2147/30000 [4:51:03<65:06:13,  8.41s/it, lr: 1.0e-04 loss: 1.124e-01]
Xiang_wan_2_2_lora:   7%|##7                                    | 2147/30000 [4:51:03<65:06:13,  8.41s/it, lr: 1.0e-04 loss: 1.124e-01]
Xiang_wan_2_2_lora:   7%|##7                                    | 2148/30000 [4:51:03<61:37:09,  7.96s/it, lr: 1.0e-04 loss: 1.124e-01]
Xiang_wan_2_2_lora:   7%|##7                                    | 2148/30000 [4:51:03<61:37:09,  7.96s/it, lr: 1.0e-04 loss: 1.124e-01]
Xiang_wan_2_2_lora:   7%|##7                                    | 2148/30000 [4:51:10<61:37:09,  7.96s/it, lr: 1.0e-04 loss: 1.502e-01]
Xiang_wan_2_2_lora:   7%|##7                                    | 2148/30000 [4:51:10<61:37:09,  7.96s/it, lr: 1.0e-04 loss: 1.502e-01]
Xiang_wan_2_2_lora:   7%|##7                                    | 2149/30000 [4:51:10<58:26:36,  7.55s/it, lr: 1.0e-04 loss: 1.502e-01]
Xiang_wan_2_2_lora:   7%|##7                                    | 2149/30000 [4:51:10<58:26:36,  7.55s/it, lr: 1.0e-04 loss: 1.502e-01]
Xiang_wan_2_2_lora:   7%|##7                                    | 2149/30000 [4:51:19<58:26:36,  7.55s/it, lr: 1.0e-04 loss: 5.550e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2149/30000 [4:51:19<58:26:36,  7.55s/it, lr: 1.0e-04 loss: 5.550e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2150/30000 [4:51:25<58:26:29,  7.55s/it, lr: 1.0e-04 loss: 8.143e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2150/30000 [4:51:25<58:26:29,  7.55s/it, lr: 1.0e-04 loss: 8.143e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2151/30000 [4:51:25<43:19:18,  5.60s/it, lr: 1.0e-04 loss: 8.143e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2151/30000 [4:51:25<43:19:18,  5.60s/it, lr: 1.0e-04 loss: 8.143e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2151/30000 [4:51:35<43:19:18,  5.60s/it, lr: 1.0e-04 loss: 7.567e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2151/30000 [4:51:35<43:19:18,  5.60s/it, lr: 1.0e-04 loss: 7.567e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2152/30000 [4:51:35<50:24:27,  6.52s/it, lr: 1.0e-04 loss: 7.567e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2152/30000 [4:51:35<50:24:27,  6.52s/it, lr: 1.0e-04 loss: 7.567e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2152/30000 [4:51:44<50:24:27,  6.52s/it, lr: 1.0e-04 loss: 5.440e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2152/30000 [4:51:44<50:24:27,  6.52s/it, lr: 1.0e-04 loss: 5.440e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2153/30000 [4:51:44<56:05:22,  7.25s/it, lr: 1.0e-04 loss: 5.440e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2153/30000 [4:51:44<56:05:22,  7.25s/it, lr: 1.0e-04 loss: 5.440e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2153/30000 [4:51:51<56:05:22,  7.25s/it, lr: 1.0e-04 loss: 6.876e-02]
Xiang_wan_2_2_lora:   7%|##7                                    | 2153/30000 [4:51:51<56:05:22,  7.25s/it, lr: 1.0e-04 loss: 6.876e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2154/30000 [4:51:51<54:43:54,  7.08s/it, lr: 1.0e-04 loss: 6.876e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2154/30000 [4:51:51<54:43:54,  7.08s/it, lr: 1.0e-04 loss: 6.876e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2154/30000 [4:51:57<54:43:54,  7.08s/it, lr: 1.0e-04 loss: 4.027e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2154/30000 [4:51:57<54:43:54,  7.08s/it, lr: 1.0e-04 loss: 4.027e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2155/30000 [4:51:57<53:37:40,  6.93s/it, lr: 1.0e-04 loss: 4.027e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2155/30000 [4:51:57<53:37:40,  6.93s/it, lr: 1.0e-04 loss: 4.027e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2155/30000 [4:52:04<53:37:40,  6.93s/it, lr: 1.0e-04 loss: 1.519e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2155/30000 [4:52:04<53:37:40,  6.93s/it, lr: 1.0e-04 loss: 1.519e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2156/30000 [4:52:04<53:16:37,  6.89s/it, lr: 1.0e-04 loss: 1.519e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2156/30000 [4:52:04<53:16:37,  6.89s/it, lr: 1.0e-04 loss: 1.519e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2156/30000 [4:52:11<53:16:37,  6.89s/it, lr: 1.0e-04 loss: 7.748e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2156/30000 [4:52:11<53:16:37,  6.89s/it, lr: 1.0e-04 loss: 7.748e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2157/30000 [4:52:11<52:38:13,  6.81s/it, lr: 1.0e-04 loss: 7.748e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2157/30000 [4:52:11<52:38:13,  6.81s/it, lr: 1.0e-04 loss: 7.748e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2157/30000 [4:52:20<52:38:13,  6.81s/it, lr: 1.0e-04 loss: 8.386e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2157/30000 [4:52:20<52:38:13,  6.81s/it, lr: 1.0e-04 loss: 8.386e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2158/30000 [4:52:20<58:23:59,  7.55s/it, lr: 1.0e-04 loss: 8.386e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2158/30000 [4:52:20<58:23:59,  7.55s/it, lr: 1.0e-04 loss: 8.386e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2158/30000 [4:52:27<58:23:59,  7.55s/it, lr: 1.0e-04 loss: 7.475e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2158/30000 [4:52:27<58:23:59,  7.55s/it, lr: 1.0e-04 loss: 7.475e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2159/30000 [4:52:27<56:08:49,  7.26s/it, lr: 1.0e-04 loss: 7.475e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2159/30000 [4:52:27<56:08:49,  7.26s/it, lr: 1.0e-04 loss: 7.475e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2159/30000 [4:52:36<56:08:49,  7.26s/it, lr: 1.0e-04 loss: 6.611e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2159/30000 [4:52:36<56:08:49,  7.26s/it, lr: 1.0e-04 loss: 6.611e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2160/30000 [4:52:42<56:08:42,  7.26s/it, lr: 1.0e-04 loss: 6.538e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2160/30000 [4:52:42<56:08:42,  7.26s/it, lr: 1.0e-04 loss: 6.538e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2161/30000 [4:52:42<42:05:43,  5.44s/it, lr: 1.0e-04 loss: 6.538e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2161/30000 [4:52:42<42:05:43,  5.44s/it, lr: 1.0e-04 loss: 6.538e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2161/30000 [4:52:52<42:05:43,  5.44s/it, lr: 1.0e-04 loss: 1.006e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2161/30000 [4:52:52<42:05:43,  5.44s/it, lr: 1.0e-04 loss: 1.006e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2162/30000 [4:52:52<49:20:22,  6.38s/it, lr: 1.0e-04 loss: 1.006e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2162/30000 [4:52:52<49:20:22,  6.38s/it, lr: 1.0e-04 loss: 1.006e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2162/30000 [4:53:01<49:20:22,  6.38s/it, lr: 1.0e-04 loss: 4.332e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2162/30000 [4:53:01<49:20:22,  6.38s/it, lr: 1.0e-04 loss: 4.332e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2163/30000 [4:53:01<55:41:05,  7.20s/it, lr: 1.0e-04 loss: 4.332e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2163/30000 [4:53:01<55:41:05,  7.20s/it, lr: 1.0e-04 loss: 4.332e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2163/30000 [4:53:08<55:41:05,  7.20s/it, lr: 1.0e-04 loss: 8.556e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2163/30000 [4:53:08<55:41:05,  7.20s/it, lr: 1.0e-04 loss: 8.556e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2164/30000 [4:53:08<54:24:22,  7.04s/it, lr: 1.0e-04 loss: 8.556e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2164/30000 [4:53:08<54:24:22,  7.04s/it, lr: 1.0e-04 loss: 8.556e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2164/30000 [4:53:17<54:24:22,  7.04s/it, lr: 1.0e-04 loss: 5.108e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2164/30000 [4:53:17<54:24:22,  7.04s/it, lr: 1.0e-04 loss: 5.108e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2165/30000 [4:53:17<59:17:55,  7.67s/it, lr: 1.0e-04 loss: 5.108e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2165/30000 [4:53:17<59:17:55,  7.67s/it, lr: 1.0e-04 loss: 5.108e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2165/30000 [4:53:24<59:17:55,  7.67s/it, lr: 1.0e-04 loss: 7.845e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2165/30000 [4:53:24<59:17:55,  7.67s/it, lr: 1.0e-04 loss: 7.845e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2166/30000 [4:53:24<56:54:04,  7.36s/it, lr: 1.0e-04 loss: 7.845e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2166/30000 [4:53:24<56:54:04,  7.36s/it, lr: 1.0e-04 loss: 7.845e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2166/30000 [4:53:30<56:54:04,  7.36s/it, lr: 1.0e-04 loss: 8.271e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2166/30000 [4:53:30<56:54:04,  7.36s/it, lr: 1.0e-04 loss: 8.271e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2167/30000 [4:53:30<55:01:41,  7.12s/it, lr: 1.0e-04 loss: 8.271e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2167/30000 [4:53:30<55:01:41,  7.12s/it, lr: 1.0e-04 loss: 8.271e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2167/30000 [4:53:37<55:01:41,  7.12s/it, lr: 1.0e-04 loss: 1.076e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2167/30000 [4:53:37<55:01:41,  7.12s/it, lr: 1.0e-04 loss: 1.076e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2168/30000 [4:53:37<53:42:49,  6.95s/it, lr: 1.0e-04 loss: 1.076e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2168/30000 [4:53:37<53:42:49,  6.95s/it, lr: 1.0e-04 loss: 1.076e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2168/30000 [4:53:43<53:42:49,  6.95s/it, lr: 1.0e-04 loss: 9.256e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2168/30000 [4:53:43<53:42:49,  6.95s/it, lr: 1.0e-04 loss: 9.256e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2169/30000 [4:53:43<52:46:52,  6.83s/it, lr: 1.0e-04 loss: 9.256e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2169/30000 [4:53:43<52:46:52,  6.83s/it, lr: 1.0e-04 loss: 9.256e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2169/30000 [4:53:53<52:46:52,  6.83s/it, lr: 1.0e-04 loss: 8.782e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2169/30000 [4:53:53<52:46:52,  6.83s/it, lr: 1.0e-04 loss: 8.782e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2170/30000 [4:53:59<52:46:45,  6.83s/it, lr: 1.0e-04 loss: 6.773e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2170/30000 [4:53:59<52:46:45,  6.83s/it, lr: 1.0e-04 loss: 6.773e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2171/30000 [4:53:59<40:17:14,  5.21s/it, lr: 1.0e-04 loss: 6.773e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2171/30000 [4:53:59<40:17:14,  5.21s/it, lr: 1.0e-04 loss: 6.773e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2171/30000 [4:54:06<40:17:14,  5.21s/it, lr: 1.0e-04 loss: 8.222e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2171/30000 [4:54:06<40:17:14,  5.21s/it, lr: 1.0e-04 loss: 8.222e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2172/30000 [4:54:06<42:50:42,  5.54s/it, lr: 1.0e-04 loss: 8.222e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2172/30000 [4:54:06<42:50:42,  5.54s/it, lr: 1.0e-04 loss: 8.222e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2172/30000 [4:54:15<42:50:42,  5.54s/it, lr: 1.0e-04 loss: 4.641e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2172/30000 [4:54:15<42:50:42,  5.54s/it, lr: 1.0e-04 loss: 4.641e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2173/30000 [4:54:15<50:25:21,  6.52s/it, lr: 1.0e-04 loss: 4.641e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2173/30000 [4:54:15<50:25:21,  6.52s/it, lr: 1.0e-04 loss: 4.641e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2173/30000 [4:54:22<50:25:21,  6.52s/it, lr: 1.0e-04 loss: 9.508e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2173/30000 [4:54:22<50:25:21,  6.52s/it, lr: 1.0e-04 loss: 9.508e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2174/30000 [4:54:22<50:32:04,  6.54s/it, lr: 1.0e-04 loss: 9.508e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2174/30000 [4:54:22<50:32:04,  6.54s/it, lr: 1.0e-04 loss: 9.508e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2174/30000 [4:54:28<50:32:04,  6.54s/it, lr: 1.0e-04 loss: 2.848e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2174/30000 [4:54:28<50:32:04,  6.54s/it, lr: 1.0e-04 loss: 2.848e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2175/30000 [4:54:28<50:25:29,  6.52s/it, lr: 1.0e-04 loss: 2.848e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2175/30000 [4:54:28<50:25:29,  6.52s/it, lr: 1.0e-04 loss: 2.848e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2175/30000 [4:54:35<50:25:29,  6.52s/it, lr: 1.0e-04 loss: 1.060e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2175/30000 [4:54:35<50:25:29,  6.52s/it, lr: 1.0e-04 loss: 1.060e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2176/30000 [4:54:35<50:31:13,  6.54s/it, lr: 1.0e-04 loss: 1.060e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2176/30000 [4:54:35<50:31:13,  6.54s/it, lr: 1.0e-04 loss: 1.060e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2176/30000 [4:54:44<50:31:13,  6.54s/it, lr: 1.0e-04 loss: 5.227e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2176/30000 [4:54:44<50:31:13,  6.54s/it, lr: 1.0e-04 loss: 5.227e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2177/30000 [4:54:44<56:37:14,  7.33s/it, lr: 1.0e-04 loss: 5.227e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2177/30000 [4:54:44<56:37:14,  7.33s/it, lr: 1.0e-04 loss: 5.227e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2177/30000 [4:54:54<56:37:14,  7.33s/it, lr: 1.0e-04 loss: 5.268e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2177/30000 [4:54:54<56:37:14,  7.33s/it, lr: 1.0e-04 loss: 5.268e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2178/30000 [4:54:54<61:38:16,  7.98s/it, lr: 1.0e-04 loss: 5.268e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2178/30000 [4:54:54<61:38:16,  7.98s/it, lr: 1.0e-04 loss: 5.268e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2178/30000 [4:55:00<61:38:16,  7.98s/it, lr: 1.0e-04 loss: 5.902e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2178/30000 [4:55:00<61:38:16,  7.98s/it, lr: 1.0e-04 loss: 5.902e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2179/30000 [4:55:00<58:25:49,  7.56s/it, lr: 1.0e-04 loss: 5.902e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2179/30000 [4:55:00<58:25:49,  7.56s/it, lr: 1.0e-04 loss: 5.902e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2179/30000 [4:55:07<58:25:49,  7.56s/it, lr: 1.0e-04 loss: 9.671e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2179/30000 [4:55:07<58:25:49,  7.56s/it, lr: 1.0e-04 loss: 9.671e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2180/30000 [4:55:13<58:25:41,  7.56s/it, lr: 1.0e-04 loss: 8.008e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2180/30000 [4:55:13<58:25:41,  7.56s/it, lr: 1.0e-04 loss: 8.008e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2181/30000 [4:55:13<43:21:34,  5.61s/it, lr: 1.0e-04 loss: 8.008e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2181/30000 [4:55:13<43:21:34,  5.61s/it, lr: 1.0e-04 loss: 8.008e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2181/30000 [4:55:23<43:21:34,  5.61s/it, lr: 1.0e-04 loss: 6.902e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2181/30000 [4:55:23<43:21:34,  5.61s/it, lr: 1.0e-04 loss: 6.902e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2182/30000 [4:55:23<50:24:25,  6.52s/it, lr: 1.0e-04 loss: 6.902e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2182/30000 [4:55:23<50:24:25,  6.52s/it, lr: 1.0e-04 loss: 6.902e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2182/30000 [4:55:32<50:24:25,  6.52s/it, lr: 1.0e-04 loss: 7.279e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2182/30000 [4:55:32<50:24:25,  6.52s/it, lr: 1.0e-04 loss: 7.279e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2183/30000 [4:55:32<56:02:36,  7.25s/it, lr: 1.0e-04 loss: 7.279e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2183/30000 [4:55:32<56:02:36,  7.25s/it, lr: 1.0e-04 loss: 7.279e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2183/30000 [4:55:39<56:02:36,  7.25s/it, lr: 1.0e-04 loss: 9.545e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2183/30000 [4:55:39<56:02:36,  7.25s/it, lr: 1.0e-04 loss: 9.545e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2184/30000 [4:55:39<54:40:52,  7.08s/it, lr: 1.0e-04 loss: 9.545e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2184/30000 [4:55:39<54:40:52,  7.08s/it, lr: 1.0e-04 loss: 9.545e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2184/30000 [4:55:45<54:40:52,  7.08s/it, lr: 1.0e-04 loss: 4.958e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2184/30000 [4:55:45<54:40:52,  7.08s/it, lr: 1.0e-04 loss: 4.958e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2185/30000 [4:55:45<54:03:02,  7.00s/it, lr: 1.0e-04 loss: 4.958e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2185/30000 [4:55:45<54:03:02,  7.00s/it, lr: 1.0e-04 loss: 4.958e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2185/30000 [4:55:52<54:03:02,  7.00s/it, lr: 1.0e-04 loss: 1.096e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2185/30000 [4:55:52<54:03:02,  7.00s/it, lr: 1.0e-04 loss: 1.096e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2186/30000 [4:55:52<53:03:07,  6.87s/it, lr: 1.0e-04 loss: 1.096e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2186/30000 [4:55:52<53:03:07,  6.87s/it, lr: 1.0e-04 loss: 1.096e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2186/30000 [4:56:01<53:03:07,  6.87s/it, lr: 1.0e-04 loss: 5.162e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2186/30000 [4:56:01<53:03:07,  6.87s/it, lr: 1.0e-04 loss: 5.162e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2187/30000 [4:56:01<58:28:16,  7.57s/it, lr: 1.0e-04 loss: 5.162e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2187/30000 [4:56:01<58:28:16,  7.57s/it, lr: 1.0e-04 loss: 5.162e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2187/30000 [4:56:11<58:28:16,  7.57s/it, lr: 1.0e-04 loss: 8.984e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2187/30000 [4:56:11<58:28:16,  7.57s/it, lr: 1.0e-04 loss: 8.984e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2188/30000 [4:56:11<62:24:47,  8.08s/it, lr: 1.0e-04 loss: 8.984e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2188/30000 [4:56:11<62:24:47,  8.08s/it, lr: 1.0e-04 loss: 8.984e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2188/30000 [4:56:20<62:24:47,  8.08s/it, lr: 1.0e-04 loss: 1.647e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2188/30000 [4:56:20<62:24:47,  8.08s/it, lr: 1.0e-04 loss: 1.647e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2189/30000 [4:56:20<65:14:07,  8.44s/it, lr: 1.0e-04 loss: 1.647e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2189/30000 [4:56:20<65:14:07,  8.44s/it, lr: 1.0e-04 loss: 1.647e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2189/30000 [4:56:26<65:14:07,  8.44s/it, lr: 1.0e-04 loss: 7.504e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2189/30000 [4:56:26<65:14:07,  8.44s/it, lr: 1.0e-04 loss: 7.504e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2190/30000 [4:56:33<65:13:58,  8.44s/it, lr: 1.0e-04 loss: 3.099e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2190/30000 [4:56:33<65:13:58,  8.44s/it, lr: 1.0e-04 loss: 3.099e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2191/30000 [4:56:33<47:04:28,  6.09s/it, lr: 1.0e-04 loss: 3.099e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2191/30000 [4:56:33<47:04:28,  6.09s/it, lr: 1.0e-04 loss: 3.099e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2191/30000 [4:56:40<47:04:28,  6.09s/it, lr: 1.0e-04 loss: 1.053e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2191/30000 [4:56:40<47:04:28,  6.09s/it, lr: 1.0e-04 loss: 1.053e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2192/30000 [4:56:40<48:31:16,  6.28s/it, lr: 1.0e-04 loss: 1.053e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2192/30000 [4:56:40<48:31:16,  6.28s/it, lr: 1.0e-04 loss: 1.053e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2192/30000 [4:56:49<48:31:16,  6.28s/it, lr: 1.0e-04 loss: 6.690e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2192/30000 [4:56:49<48:31:16,  6.28s/it, lr: 1.0e-04 loss: 6.690e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2193/30000 [4:56:49<54:38:10,  7.07s/it, lr: 1.0e-04 loss: 6.690e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2193/30000 [4:56:49<54:38:10,  7.07s/it, lr: 1.0e-04 loss: 6.690e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2193/30000 [4:56:56<54:38:10,  7.07s/it, lr: 1.0e-04 loss: 7.501e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2193/30000 [4:56:56<54:38:10,  7.07s/it, lr: 1.0e-04 loss: 7.501e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2194/30000 [4:56:56<53:28:58,  6.92s/it, lr: 1.0e-04 loss: 7.501e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2194/30000 [4:56:56<53:28:58,  6.92s/it, lr: 1.0e-04 loss: 7.501e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2194/30000 [4:57:02<53:28:58,  6.92s/it, lr: 1.0e-04 loss: 8.527e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2194/30000 [4:57:02<53:28:58,  6.92s/it, lr: 1.0e-04 loss: 8.527e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2195/30000 [4:57:02<52:33:54,  6.81s/it, lr: 1.0e-04 loss: 8.527e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2195/30000 [4:57:02<52:33:54,  6.81s/it, lr: 1.0e-04 loss: 8.527e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2195/30000 [4:57:09<52:33:54,  6.81s/it, lr: 1.0e-04 loss: 9.131e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2195/30000 [4:57:09<52:33:54,  6.81s/it, lr: 1.0e-04 loss: 9.131e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2196/30000 [4:57:09<51:56:20,  6.72s/it, lr: 1.0e-04 loss: 9.131e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2196/30000 [4:57:09<51:56:20,  6.72s/it, lr: 1.0e-04 loss: 9.131e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2196/30000 [4:57:15<51:56:20,  6.72s/it, lr: 1.0e-04 loss: 1.050e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2196/30000 [4:57:15<51:56:20,  6.72s/it, lr: 1.0e-04 loss: 1.050e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2197/30000 [4:57:15<51:36:02,  6.68s/it, lr: 1.0e-04 loss: 1.050e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2197/30000 [4:57:15<51:36:02,  6.68s/it, lr: 1.0e-04 loss: 1.050e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2197/30000 [4:57:22<51:36:02,  6.68s/it, lr: 1.0e-04 loss: 9.221e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2197/30000 [4:57:22<51:36:02,  6.68s/it, lr: 1.0e-04 loss: 9.221e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2198/30000 [4:57:22<51:15:13,  6.64s/it, lr: 1.0e-04 loss: 9.221e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2198/30000 [4:57:22<51:15:13,  6.64s/it, lr: 1.0e-04 loss: 9.221e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2198/30000 [4:57:31<51:15:13,  6.64s/it, lr: 1.0e-04 loss: 4.688e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2198/30000 [4:57:31<51:15:13,  6.64s/it, lr: 1.0e-04 loss: 4.688e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2199/30000 [4:57:31<57:48:07,  7.48s/it, lr: 1.0e-04 loss: 4.688e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2199/30000 [4:57:31<57:48:07,  7.48s/it, lr: 1.0e-04 loss: 4.688e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2199/30000 [4:57:41<57:48:07,  7.48s/it, lr: 1.0e-04 loss: 6.522e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2199/30000 [4:57:41<57:48:07,  7.48s/it, lr: 1.0e-04 loss: 6.522e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2200/30000 [4:57:50<57:47:59,  7.48s/it, lr: 1.0e-04 loss: 6.224e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2200/30000 [4:57:50<57:47:59,  7.48s/it, lr: 1.0e-04 loss: 6.224e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2201/30000 [4:57:50<47:53:54,  6.20s/it, lr: 1.0e-04 loss: 6.224e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2201/30000 [4:57:50<47:53:54,  6.20s/it, lr: 1.0e-04 loss: 6.224e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2201/30000 [4:57:59<47:53:54,  6.20s/it, lr: 1.0e-04 loss: 7.199e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2201/30000 [4:57:59<47:53:54,  6.20s/it, lr: 1.0e-04 loss: 7.199e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2202/30000 [4:57:59<53:54:31,  6.98s/it, lr: 1.0e-04 loss: 7.199e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2202/30000 [4:57:59<53:54:31,  6.98s/it, lr: 1.0e-04 loss: 7.199e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2202/30000 [4:58:09<53:54:31,  6.98s/it, lr: 1.0e-04 loss: 5.377e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2202/30000 [4:58:09<53:54:31,  6.98s/it, lr: 1.0e-04 loss: 5.377e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2203/30000 [4:58:09<58:41:16,  7.60s/it, lr: 1.0e-04 loss: 5.377e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2203/30000 [4:58:09<58:41:16,  7.60s/it, lr: 1.0e-04 loss: 5.377e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2203/30000 [4:58:18<58:41:16,  7.60s/it, lr: 1.0e-04 loss: 5.564e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2203/30000 [4:58:18<58:41:16,  7.60s/it, lr: 1.0e-04 loss: 5.564e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2204/30000 [4:58:18<62:21:46,  8.08s/it, lr: 1.0e-04 loss: 5.564e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2204/30000 [4:58:18<62:21:46,  8.08s/it, lr: 1.0e-04 loss: 5.564e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2204/30000 [4:58:25<62:21:46,  8.08s/it, lr: 1.0e-04 loss: 1.030e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2204/30000 [4:58:25<62:21:46,  8.08s/it, lr: 1.0e-04 loss: 1.030e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2205/30000 [4:58:25<59:11:19,  7.67s/it, lr: 1.0e-04 loss: 1.030e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2205/30000 [4:58:25<59:11:19,  7.67s/it, lr: 1.0e-04 loss: 1.030e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2205/30000 [4:58:31<59:11:19,  7.67s/it, lr: 1.0e-04 loss: 7.443e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2205/30000 [4:58:31<59:11:19,  7.67s/it, lr: 1.0e-04 loss: 7.443e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2206/30000 [4:58:31<56:48:51,  7.36s/it, lr: 1.0e-04 loss: 7.443e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2206/30000 [4:58:31<56:48:51,  7.36s/it, lr: 1.0e-04 loss: 7.443e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2206/30000 [4:58:41<56:48:51,  7.36s/it, lr: 1.0e-04 loss: 5.249e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2206/30000 [4:58:41<56:48:51,  7.36s/it, lr: 1.0e-04 loss: 5.249e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2207/30000 [4:58:41<61:35:48,  7.98s/it, lr: 1.0e-04 loss: 5.249e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2207/30000 [4:58:41<61:35:48,  7.98s/it, lr: 1.0e-04 loss: 5.249e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2207/30000 [4:58:50<61:35:48,  7.98s/it, lr: 1.0e-04 loss: 7.846e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2207/30000 [4:58:50<61:35:48,  7.98s/it, lr: 1.0e-04 loss: 7.846e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2208/30000 [4:58:50<64:40:08,  8.38s/it, lr: 1.0e-04 loss: 7.846e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2208/30000 [4:58:50<64:40:08,  8.38s/it, lr: 1.0e-04 loss: 7.846e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2208/30000 [4:59:00<64:40:08,  8.38s/it, lr: 1.0e-04 loss: 6.902e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2208/30000 [4:59:00<64:40:08,  8.38s/it, lr: 1.0e-04 loss: 6.902e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2209/30000 [4:59:00<66:52:07,  8.66s/it, lr: 1.0e-04 loss: 6.902e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2209/30000 [4:59:00<66:52:07,  8.66s/it, lr: 1.0e-04 loss: 6.902e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2209/30000 [4:59:06<66:52:07,  8.66s/it, lr: 1.0e-04 loss: 6.971e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2209/30000 [4:59:06<66:52:07,  8.66s/it, lr: 1.0e-04 loss: 6.971e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2210/30000 [4:59:13<66:51:58,  8.66s/it, lr: 1.0e-04 loss: 7.345e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2210/30000 [4:59:13<66:51:58,  8.66s/it, lr: 1.0e-04 loss: 7.345e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2211/30000 [4:59:13<47:58:18,  6.21s/it, lr: 1.0e-04 loss: 7.345e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2211/30000 [4:59:13<47:58:18,  6.21s/it, lr: 1.0e-04 loss: 7.345e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2211/30000 [4:59:19<47:58:18,  6.21s/it, lr: 1.0e-04 loss: 8.217e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2211/30000 [4:59:19<47:58:18,  6.21s/it, lr: 1.0e-04 loss: 8.217e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2212/30000 [4:59:19<48:43:13,  6.31s/it, lr: 1.0e-04 loss: 8.217e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2212/30000 [4:59:19<48:43:13,  6.31s/it, lr: 1.0e-04 loss: 8.217e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2212/30000 [4:59:29<48:43:13,  6.31s/it, lr: 1.0e-04 loss: 4.573e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2212/30000 [4:59:29<48:43:13,  6.31s/it, lr: 1.0e-04 loss: 4.573e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2213/30000 [4:59:29<54:49:55,  7.10s/it, lr: 1.0e-04 loss: 4.573e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2213/30000 [4:59:29<54:49:55,  7.10s/it, lr: 1.0e-04 loss: 4.573e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2213/30000 [4:59:36<54:49:55,  7.10s/it, lr: 1.0e-04 loss: 1.067e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2213/30000 [4:59:36<54:49:55,  7.10s/it, lr: 1.0e-04 loss: 1.067e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2214/30000 [4:59:36<54:16:04,  7.03s/it, lr: 1.0e-04 loss: 1.067e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2214/30000 [4:59:36<54:16:04,  7.03s/it, lr: 1.0e-04 loss: 1.067e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2214/30000 [4:59:42<54:16:04,  7.03s/it, lr: 1.0e-04 loss: 8.328e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2214/30000 [4:59:42<54:16:04,  7.03s/it, lr: 1.0e-04 loss: 8.328e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2215/30000 [4:59:42<53:16:41,  6.90s/it, lr: 1.0e-04 loss: 8.328e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2215/30000 [4:59:42<53:16:41,  6.90s/it, lr: 1.0e-04 loss: 8.328e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2215/30000 [4:59:52<53:16:41,  6.90s/it, lr: 1.0e-04 loss: 9.162e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2215/30000 [4:59:52<53:16:41,  6.90s/it, lr: 1.0e-04 loss: 9.162e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2216/30000 [4:59:52<58:40:38,  7.60s/it, lr: 1.0e-04 loss: 9.162e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2216/30000 [4:59:52<58:40:38,  7.60s/it, lr: 1.0e-04 loss: 9.162e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2216/30000 [4:59:58<58:40:38,  7.60s/it, lr: 1.0e-04 loss: 5.040e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2216/30000 [4:59:58<58:40:38,  7.60s/it, lr: 1.0e-04 loss: 5.040e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2217/30000 [4:59:58<56:23:47,  7.31s/it, lr: 1.0e-04 loss: 5.040e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2217/30000 [4:59:58<56:23:47,  7.31s/it, lr: 1.0e-04 loss: 5.040e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2217/30000 [5:00:05<56:23:47,  7.31s/it, lr: 1.0e-04 loss: 1.110e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2217/30000 [5:00:05<56:23:47,  7.31s/it, lr: 1.0e-04 loss: 1.110e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2218/30000 [5:00:05<54:51:46,  7.11s/it, lr: 1.0e-04 loss: 1.110e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2218/30000 [5:00:05<54:51:46,  7.11s/it, lr: 1.0e-04 loss: 1.110e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2218/30000 [5:00:11<54:51:46,  7.11s/it, lr: 1.0e-04 loss: 5.301e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2218/30000 [5:00:11<54:51:46,  7.11s/it, lr: 1.0e-04 loss: 5.301e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2219/30000 [5:00:11<53:32:48,  6.94s/it, lr: 1.0e-04 loss: 5.301e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2219/30000 [5:00:11<53:32:48,  6.94s/it, lr: 1.0e-04 loss: 5.301e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2219/30000 [5:00:21<53:32:48,  6.94s/it, lr: 1.0e-04 loss: 8.420e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2219/30000 [5:00:21<53:32:48,  6.94s/it, lr: 1.0e-04 loss: 8.420e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2220/30000 [5:00:28<53:32:41,  6.94s/it, lr: 1.0e-04 loss: 7.763e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2220/30000 [5:00:28<53:32:41,  6.94s/it, lr: 1.0e-04 loss: 7.763e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2221/30000 [5:00:28<41:21:47,  5.36s/it, lr: 1.0e-04 loss: 7.763e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2221/30000 [5:00:28<41:21:47,  5.36s/it, lr: 1.0e-04 loss: 7.763e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2221/30000 [5:00:37<41:21:47,  5.36s/it, lr: 1.0e-04 loss: 8.414e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2221/30000 [5:00:37<41:21:47,  5.36s/it, lr: 1.0e-04 loss: 8.414e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2222/30000 [5:00:37<48:55:47,  6.34s/it, lr: 1.0e-04 loss: 8.414e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2222/30000 [5:00:37<48:55:47,  6.34s/it, lr: 1.0e-04 loss: 8.414e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2222/30000 [5:00:44<48:55:47,  6.34s/it, lr: 1.0e-04 loss: 5.441e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2222/30000 [5:00:44<48:55:47,  6.34s/it, lr: 1.0e-04 loss: 5.441e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2223/30000 [5:00:44<49:35:09,  6.43s/it, lr: 1.0e-04 loss: 5.441e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2223/30000 [5:00:44<49:35:09,  6.43s/it, lr: 1.0e-04 loss: 5.441e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2223/30000 [5:00:53<49:35:09,  6.43s/it, lr: 1.0e-04 loss: 5.906e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2223/30000 [5:00:53<49:35:09,  6.43s/it, lr: 1.0e-04 loss: 5.906e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2224/30000 [5:00:53<55:43:56,  7.22s/it, lr: 1.0e-04 loss: 5.906e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2224/30000 [5:00:53<55:43:56,  7.22s/it, lr: 1.0e-04 loss: 5.906e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2224/30000 [5:01:02<55:43:56,  7.22s/it, lr: 1.0e-04 loss: 4.158e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2224/30000 [5:01:02<55:43:56,  7.22s/it, lr: 1.0e-04 loss: 4.158e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2225/30000 [5:01:02<60:18:13,  7.82s/it, lr: 1.0e-04 loss: 4.158e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2225/30000 [5:01:02<60:18:13,  7.82s/it, lr: 1.0e-04 loss: 4.158e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2225/30000 [5:01:12<60:18:13,  7.82s/it, lr: 1.0e-04 loss: 1.156e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2225/30000 [5:01:12<60:18:13,  7.82s/it, lr: 1.0e-04 loss: 1.156e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2226/30000 [5:01:12<63:38:40,  8.25s/it, lr: 1.0e-04 loss: 1.156e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2226/30000 [5:01:12<63:38:40,  8.25s/it, lr: 1.0e-04 loss: 1.156e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2226/30000 [5:01:21<63:38:40,  8.25s/it, lr: 1.0e-04 loss: 7.898e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2226/30000 [5:01:21<63:38:40,  8.25s/it, lr: 1.0e-04 loss: 7.898e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2227/30000 [5:01:21<66:15:56,  8.59s/it, lr: 1.0e-04 loss: 7.898e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2227/30000 [5:01:21<66:15:56,  8.59s/it, lr: 1.0e-04 loss: 7.898e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2227/30000 [5:01:28<66:15:56,  8.59s/it, lr: 1.0e-04 loss: 1.144e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2227/30000 [5:01:28<66:15:56,  8.59s/it, lr: 1.0e-04 loss: 1.144e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2228/30000 [5:01:28<61:46:41,  8.01s/it, lr: 1.0e-04 loss: 1.144e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2228/30000 [5:01:28<61:46:41,  8.01s/it, lr: 1.0e-04 loss: 1.144e-01]
Xiang_wan_2_2_lora:   7%|##8                                    | 2228/30000 [5:01:37<61:46:41,  8.01s/it, lr: 1.0e-04 loss: 5.106e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2228/30000 [5:01:37<61:46:41,  8.01s/it, lr: 1.0e-04 loss: 5.106e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2229/30000 [5:01:37<65:10:03,  8.45s/it, lr: 1.0e-04 loss: 5.106e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2229/30000 [5:01:37<65:10:03,  8.45s/it, lr: 1.0e-04 loss: 5.106e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2229/30000 [5:01:44<65:10:03,  8.45s/it, lr: 1.0e-04 loss: 9.038e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2229/30000 [5:01:44<65:10:03,  8.45s/it, lr: 1.0e-04 loss: 9.038e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2230/30000 [5:01:50<65:09:55,  8.45s/it, lr: 1.0e-04 loss: 6.380e-02]
Xiang_wan_2_2_lora:   7%|##8                                    | 2230/30000 [5:01:50<65:09:55,  8.45s/it, lr: 1.0e-04 loss: 6.380e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2231/30000 [5:01:50<46:51:57,  6.08s/it, lr: 1.0e-04 loss: 6.380e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2231/30000 [5:01:50<46:51:57,  6.08s/it, lr: 1.0e-04 loss: 6.380e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2231/30000 [5:02:00<46:51:57,  6.08s/it, lr: 1.0e-04 loss: 7.856e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2231/30000 [5:02:00<46:51:57,  6.08s/it, lr: 1.0e-04 loss: 7.856e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2232/30000 [5:02:00<52:58:36,  6.87s/it, lr: 1.0e-04 loss: 7.856e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2232/30000 [5:02:00<52:58:36,  6.87s/it, lr: 1.0e-04 loss: 7.856e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2232/30000 [5:02:09<52:58:36,  6.87s/it, lr: 1.0e-04 loss: 3.730e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2232/30000 [5:02:09<52:58:36,  6.87s/it, lr: 1.0e-04 loss: 3.730e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2233/30000 [5:02:09<57:54:27,  7.51s/it, lr: 1.0e-04 loss: 3.730e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2233/30000 [5:02:09<57:54:27,  7.51s/it, lr: 1.0e-04 loss: 3.730e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2233/30000 [5:02:16<57:54:27,  7.51s/it, lr: 1.0e-04 loss: 1.043e-01]
Xiang_wan_2_2_lora:   7%|##9                                    | 2233/30000 [5:02:16<57:54:27,  7.51s/it, lr: 1.0e-04 loss: 1.043e-01]
Xiang_wan_2_2_lora:   7%|##9                                    | 2234/30000 [5:02:16<56:00:26,  7.26s/it, lr: 1.0e-04 loss: 1.043e-01]
Xiang_wan_2_2_lora:   7%|##9                                    | 2234/30000 [5:02:16<56:00:26,  7.26s/it, lr: 1.0e-04 loss: 1.043e-01]
Xiang_wan_2_2_lora:   7%|##9                                    | 2234/30000 [5:02:22<56:00:26,  7.26s/it, lr: 1.0e-04 loss: 6.959e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2234/30000 [5:02:22<56:00:26,  7.26s/it, lr: 1.0e-04 loss: 6.959e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2235/30000 [5:02:22<54:30:37,  7.07s/it, lr: 1.0e-04 loss: 6.959e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2235/30000 [5:02:22<54:30:37,  7.07s/it, lr: 1.0e-04 loss: 6.959e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2235/30000 [5:02:32<54:30:37,  7.07s/it, lr: 1.0e-04 loss: 6.189e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2235/30000 [5:02:32<54:30:37,  7.07s/it, lr: 1.0e-04 loss: 6.189e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2236/30000 [5:02:32<59:56:04,  7.77s/it, lr: 1.0e-04 loss: 6.189e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2236/30000 [5:02:32<59:56:04,  7.77s/it, lr: 1.0e-04 loss: 6.189e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2236/30000 [5:02:38<59:56:04,  7.77s/it, lr: 1.0e-04 loss: 6.181e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2236/30000 [5:02:38<59:56:04,  7.77s/it, lr: 1.0e-04 loss: 6.181e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2237/30000 [5:02:38<57:14:55,  7.42s/it, lr: 1.0e-04 loss: 6.181e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2237/30000 [5:02:38<57:14:55,  7.42s/it, lr: 1.0e-04 loss: 6.181e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2237/30000 [5:02:48<57:14:55,  7.42s/it, lr: 1.0e-04 loss: 6.631e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2237/30000 [5:02:48<57:14:55,  7.42s/it, lr: 1.0e-04 loss: 6.631e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2238/30000 [5:02:48<61:29:56,  7.97s/it, lr: 1.0e-04 loss: 6.631e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2238/30000 [5:02:48<61:29:56,  7.97s/it, lr: 1.0e-04 loss: 6.631e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2238/30000 [5:02:54<61:29:56,  7.97s/it, lr: 1.0e-04 loss: 4.010e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2238/30000 [5:02:54<61:29:56,  7.97s/it, lr: 1.0e-04 loss: 4.010e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2239/30000 [5:02:54<58:25:28,  7.58s/it, lr: 1.0e-04 loss: 4.010e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2239/30000 [5:02:54<58:25:28,  7.58s/it, lr: 1.0e-04 loss: 4.010e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2239/30000 [5:03:03<58:25:28,  7.58s/it, lr: 1.0e-04 loss: 7.920e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2239/30000 [5:03:03<58:25:28,  7.58s/it, lr: 1.0e-04 loss: 7.920e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2240/30000 [5:03:13<58:25:21,  7.58s/it, lr: 1.0e-04 loss: 4.245e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2240/30000 [5:03:13<58:25:21,  7.58s/it, lr: 1.0e-04 loss: 4.245e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2241/30000 [5:03:13<48:10:59,  6.25s/it, lr: 1.0e-04 loss: 4.245e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2241/30000 [5:03:13<48:10:59,  6.25s/it, lr: 1.0e-04 loss: 4.245e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2241/30000 [5:03:19<48:10:59,  6.25s/it, lr: 1.0e-04 loss: 7.446e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2241/30000 [5:03:19<48:10:59,  6.25s/it, lr: 1.0e-04 loss: 7.446e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2242/30000 [5:03:19<48:52:24,  6.34s/it, lr: 1.0e-04 loss: 7.446e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2242/30000 [5:03:19<48:52:24,  6.34s/it, lr: 1.0e-04 loss: 7.446e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2242/30000 [5:03:26<48:52:24,  6.34s/it, lr: 1.0e-04 loss: 6.374e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2242/30000 [5:03:26<48:52:24,  6.34s/it, lr: 1.0e-04 loss: 6.374e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2243/30000 [5:03:26<49:21:37,  6.40s/it, lr: 1.0e-04 loss: 6.374e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2243/30000 [5:03:26<49:21:37,  6.40s/it, lr: 1.0e-04 loss: 6.374e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2243/30000 [5:03:33<49:21:37,  6.40s/it, lr: 1.0e-04 loss: 6.518e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2243/30000 [5:03:33<49:21:37,  6.40s/it, lr: 1.0e-04 loss: 6.518e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2244/30000 [5:03:33<50:18:56,  6.53s/it, lr: 1.0e-04 loss: 6.518e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2244/30000 [5:03:33<50:18:56,  6.53s/it, lr: 1.0e-04 loss: 6.518e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2244/30000 [5:03:42<50:18:56,  6.53s/it, lr: 1.0e-04 loss: 2.447e-01]
Xiang_wan_2_2_lora:   7%|##9                                    | 2244/30000 [5:03:42<50:18:56,  6.53s/it, lr: 1.0e-04 loss: 2.447e-01]
Xiang_wan_2_2_lora:   7%|##9                                    | 2245/30000 [5:03:42<56:24:30,  7.32s/it, lr: 1.0e-04 loss: 2.447e-01]
Xiang_wan_2_2_lora:   7%|##9                                    | 2245/30000 [5:03:42<56:24:30,  7.32s/it, lr: 1.0e-04 loss: 2.447e-01]
Xiang_wan_2_2_lora:   7%|##9                                    | 2245/30000 [5:03:49<56:24:30,  7.32s/it, lr: 1.0e-04 loss: 8.972e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2245/30000 [5:03:49<56:24:30,  7.32s/it, lr: 1.0e-04 loss: 8.972e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2246/30000 [5:03:49<54:39:54,  7.09s/it, lr: 1.0e-04 loss: 8.972e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2246/30000 [5:03:49<54:39:54,  7.09s/it, lr: 1.0e-04 loss: 8.972e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2246/30000 [5:03:55<54:39:54,  7.09s/it, lr: 1.0e-04 loss: 6.282e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2246/30000 [5:03:55<54:39:54,  7.09s/it, lr: 1.0e-04 loss: 6.282e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2247/30000 [5:03:55<53:33:24,  6.95s/it, lr: 1.0e-04 loss: 6.282e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2247/30000 [5:03:55<53:33:24,  6.95s/it, lr: 1.0e-04 loss: 6.282e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2247/30000 [5:04:05<53:33:24,  6.95s/it, lr: 1.0e-04 loss: 7.495e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2247/30000 [5:04:05<53:33:24,  6.95s/it, lr: 1.0e-04 loss: 7.495e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2248/30000 [5:04:05<58:48:42,  7.63s/it, lr: 1.0e-04 loss: 7.495e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2248/30000 [5:04:05<58:48:42,  7.63s/it, lr: 1.0e-04 loss: 7.495e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2248/30000 [5:04:14<58:48:42,  7.63s/it, lr: 1.0e-04 loss: 5.062e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2248/30000 [5:04:14<58:48:42,  7.63s/it, lr: 1.0e-04 loss: 5.062e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2249/30000 [5:04:14<62:43:11,  8.14s/it, lr: 1.0e-04 loss: 5.062e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2249/30000 [5:04:14<62:43:11,  8.14s/it, lr: 1.0e-04 loss: 5.062e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2249/30000 [5:04:23<62:43:11,  8.14s/it, lr: 1.0e-04 loss: 6.340e-02]
Xiang_wan_2_2_lora:   7%|##9                                    | 2249/30000 [5:04:23<62:43:11,  8.14s/it, lr: 1.0e-04 loss: 6.340e-02]
+Saving at step 2250
+Saved checkpoint to /home/featurize/ai-toolkit/output/Xiang_wan_2_2_lora/Xiang_wan_2_2_lora_000002250.safetensors
+Saved optimizer to /home/featurize/ai-toolkit/output/Xiang_wan_2_2_lora/optimizer.pt
+
Xiang_wan_2_2_lora:   8%|##9                                    | 2250/30000 [5:04:30<62:43:03,  8.14s/it, lr: 1.0e-04 loss: 3.857e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2250/30000 [5:04:30<62:43:03,  8.14s/it, lr: 1.0e-04 loss: 3.857e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2251/30000 [5:04:30<46:14:22,  6.00s/it, lr: 1.0e-04 loss: 3.857e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2251/30000 [5:04:30<46:14:22,  6.00s/it, lr: 1.0e-04 loss: 3.857e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2251/30000 [5:04:40<46:14:22,  6.00s/it, lr: 1.0e-04 loss: 9.484e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2251/30000 [5:04:40<46:14:22,  6.00s/it, lr: 1.0e-04 loss: 9.484e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2252/30000 [5:04:40<52:32:56,  6.82s/it, lr: 1.0e-04 loss: 9.484e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2252/30000 [5:04:40<52:32:56,  6.82s/it, lr: 1.0e-04 loss: 9.484e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2252/30000 [5:04:46<52:32:56,  6.82s/it, lr: 1.0e-04 loss: 7.122e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2252/30000 [5:04:46<52:32:56,  6.82s/it, lr: 1.0e-04 loss: 7.122e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2253/30000 [5:04:46<52:13:58,  6.78s/it, lr: 1.0e-04 loss: 7.122e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2253/30000 [5:04:46<52:13:58,  6.78s/it, lr: 1.0e-04 loss: 7.122e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2253/30000 [5:04:56<52:13:58,  6.78s/it, lr: 1.0e-04 loss: 7.011e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2253/30000 [5:04:56<52:13:58,  6.78s/it, lr: 1.0e-04 loss: 7.011e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2254/30000 [5:04:56<57:32:18,  7.47s/it, lr: 1.0e-04 loss: 7.011e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2254/30000 [5:04:56<57:32:18,  7.47s/it, lr: 1.0e-04 loss: 7.011e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2254/30000 [5:05:02<57:32:18,  7.47s/it, lr: 1.0e-04 loss: 9.005e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2254/30000 [5:05:02<57:32:18,  7.47s/it, lr: 1.0e-04 loss: 9.005e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2255/30000 [5:05:02<55:35:00,  7.21s/it, lr: 1.0e-04 loss: 9.005e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2255/30000 [5:05:02<55:35:00,  7.21s/it, lr: 1.0e-04 loss: 9.005e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2255/30000 [5:05:09<55:35:00,  7.21s/it, lr: 1.0e-04 loss: 7.600e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2255/30000 [5:05:09<55:35:00,  7.21s/it, lr: 1.0e-04 loss: 7.600e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2256/30000 [5:05:09<54:06:12,  7.02s/it, lr: 1.0e-04 loss: 7.600e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2256/30000 [5:05:09<54:06:12,  7.02s/it, lr: 1.0e-04 loss: 7.600e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2256/30000 [5:05:15<54:06:12,  7.02s/it, lr: 1.0e-04 loss: 6.811e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2256/30000 [5:05:15<54:06:12,  7.02s/it, lr: 1.0e-04 loss: 6.811e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2257/30000 [5:05:15<52:57:09,  6.87s/it, lr: 1.0e-04 loss: 6.811e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2257/30000 [5:05:15<52:57:09,  6.87s/it, lr: 1.0e-04 loss: 6.811e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2257/30000 [5:05:22<52:57:09,  6.87s/it, lr: 1.0e-04 loss: 1.214e-01]
Xiang_wan_2_2_lora:   8%|##9                                    | 2257/30000 [5:05:22<52:57:09,  6.87s/it, lr: 1.0e-04 loss: 1.214e-01]
Xiang_wan_2_2_lora:   8%|##9                                    | 2258/30000 [5:05:22<52:47:42,  6.85s/it, lr: 1.0e-04 loss: 1.214e-01]
Xiang_wan_2_2_lora:   8%|##9                                    | 2258/30000 [5:05:22<52:47:42,  6.85s/it, lr: 1.0e-04 loss: 1.214e-01]
Xiang_wan_2_2_lora:   8%|##9                                    | 2258/30000 [5:05:31<52:47:42,  6.85s/it, lr: 1.0e-04 loss: 6.640e-01]
Xiang_wan_2_2_lora:   8%|##9                                    | 2258/30000 [5:05:31<52:47:42,  6.85s/it, lr: 1.0e-04 loss: 6.640e-01]
Xiang_wan_2_2_lora:   8%|##9                                    | 2259/30000 [5:05:31<58:27:56,  7.59s/it, lr: 1.0e-04 loss: 6.640e-01]
Xiang_wan_2_2_lora:   8%|##9                                    | 2259/30000 [5:05:31<58:27:56,  7.59s/it, lr: 1.0e-04 loss: 6.640e-01]
Xiang_wan_2_2_lora:   8%|##9                                    | 2259/30000 [5:05:38<58:27:56,  7.59s/it, lr: 1.0e-04 loss: 8.257e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2259/30000 [5:05:38<58:27:56,  7.59s/it, lr: 1.0e-04 loss: 8.257e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2260/30000 [5:05:47<58:27:48,  7.59s/it, lr: 1.0e-04 loss: 5.016e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2260/30000 [5:05:47<58:27:48,  7.59s/it, lr: 1.0e-04 loss: 5.016e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2261/30000 [5:05:47<48:06:52,  6.24s/it, lr: 1.0e-04 loss: 5.016e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2261/30000 [5:05:47<48:06:52,  6.24s/it, lr: 1.0e-04 loss: 5.016e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2261/30000 [5:05:54<48:06:52,  6.24s/it, lr: 1.0e-04 loss: 9.368e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2261/30000 [5:05:54<48:06:52,  6.24s/it, lr: 1.0e-04 loss: 9.368e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2262/30000 [5:05:54<48:44:27,  6.33s/it, lr: 1.0e-04 loss: 9.368e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2262/30000 [5:05:54<48:44:27,  6.33s/it, lr: 1.0e-04 loss: 9.368e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2262/30000 [5:06:03<48:44:27,  6.33s/it, lr: 1.0e-04 loss: 5.338e-01]
Xiang_wan_2_2_lora:   8%|##9                                    | 2262/30000 [5:06:03<48:44:27,  6.33s/it, lr: 1.0e-04 loss: 5.338e-01]
Xiang_wan_2_2_lora:   8%|##9                                    | 2263/30000 [5:06:03<54:39:51,  7.09s/it, lr: 1.0e-04 loss: 5.338e-01]
Xiang_wan_2_2_lora:   8%|##9                                    | 2263/30000 [5:06:03<54:39:51,  7.09s/it, lr: 1.0e-04 loss: 5.338e-01]
Xiang_wan_2_2_lora:   8%|##9                                    | 2263/30000 [5:06:12<54:39:51,  7.09s/it, lr: 1.0e-04 loss: 5.124e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2263/30000 [5:06:12<54:39:51,  7.09s/it, lr: 1.0e-04 loss: 5.124e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2264/30000 [5:06:12<59:19:50,  7.70s/it, lr: 1.0e-04 loss: 5.124e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2264/30000 [5:06:12<59:19:50,  7.70s/it, lr: 1.0e-04 loss: 5.124e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2264/30000 [5:06:22<59:19:50,  7.70s/it, lr: 1.0e-04 loss: 2.430e-01]
Xiang_wan_2_2_lora:   8%|##9                                    | 2264/30000 [5:06:22<59:19:50,  7.70s/it, lr: 1.0e-04 loss: 2.430e-01]
Xiang_wan_2_2_lora:   8%|##9                                    | 2265/30000 [5:06:22<63:12:42,  8.20s/it, lr: 1.0e-04 loss: 2.430e-01]
Xiang_wan_2_2_lora:   8%|##9                                    | 2265/30000 [5:06:22<63:12:42,  8.20s/it, lr: 1.0e-04 loss: 2.430e-01]
Xiang_wan_2_2_lora:   8%|##9                                    | 2265/30000 [5:06:31<63:12:42,  8.20s/it, lr: 1.0e-04 loss: 7.313e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2265/30000 [5:06:31<63:12:42,  8.20s/it, lr: 1.0e-04 loss: 7.313e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2266/30000 [5:06:31<65:41:00,  8.53s/it, lr: 1.0e-04 loss: 7.313e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2266/30000 [5:06:31<65:41:00,  8.53s/it, lr: 1.0e-04 loss: 7.313e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2266/30000 [5:06:38<65:41:00,  8.53s/it, lr: 1.0e-04 loss: 5.026e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2266/30000 [5:06:38<65:41:00,  8.53s/it, lr: 1.0e-04 loss: 5.026e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2267/30000 [5:06:38<61:20:08,  7.96s/it, lr: 1.0e-04 loss: 5.026e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2267/30000 [5:06:38<61:20:08,  7.96s/it, lr: 1.0e-04 loss: 5.026e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2267/30000 [5:06:47<61:20:08,  7.96s/it, lr: 1.0e-04 loss: 6.346e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2267/30000 [5:06:47<61:20:08,  7.96s/it, lr: 1.0e-04 loss: 6.346e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2268/30000 [5:06:47<64:24:24,  8.36s/it, lr: 1.0e-04 loss: 6.346e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2268/30000 [5:06:47<64:24:24,  8.36s/it, lr: 1.0e-04 loss: 6.346e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2268/30000 [5:06:57<64:24:24,  8.36s/it, lr: 1.0e-04 loss: 2.950e-01]
Xiang_wan_2_2_lora:   8%|##9                                    | 2268/30000 [5:06:57<64:24:24,  8.36s/it, lr: 1.0e-04 loss: 2.950e-01]
Xiang_wan_2_2_lora:   8%|##9                                    | 2269/30000 [5:06:57<66:37:16,  8.65s/it, lr: 1.0e-04 loss: 2.950e-01]
Xiang_wan_2_2_lora:   8%|##9                                    | 2269/30000 [5:06:57<66:37:16,  8.65s/it, lr: 1.0e-04 loss: 2.950e-01]
Xiang_wan_2_2_lora:   8%|##9                                    | 2269/30000 [5:07:03<66:37:16,  8.65s/it, lr: 1.0e-04 loss: 8.531e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2269/30000 [5:07:03<66:37:16,  8.65s/it, lr: 1.0e-04 loss: 8.531e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2270/30000 [5:07:10<66:37:07,  8.65s/it, lr: 1.0e-04 loss: 6.291e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2270/30000 [5:07:10<66:37:07,  8.65s/it, lr: 1.0e-04 loss: 6.291e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2271/30000 [5:07:10<47:36:44,  6.18s/it, lr: 1.0e-04 loss: 6.291e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2271/30000 [5:07:10<47:36:44,  6.18s/it, lr: 1.0e-04 loss: 6.291e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2271/30000 [5:07:16<47:36:44,  6.18s/it, lr: 1.0e-04 loss: 8.417e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2271/30000 [5:07:16<47:36:44,  6.18s/it, lr: 1.0e-04 loss: 8.417e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2272/30000 [5:07:16<48:11:51,  6.26s/it, lr: 1.0e-04 loss: 8.417e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2272/30000 [5:07:16<48:11:51,  6.26s/it, lr: 1.0e-04 loss: 8.417e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2272/30000 [5:07:23<48:11:51,  6.26s/it, lr: 1.0e-04 loss: 4.690e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2272/30000 [5:07:23<48:11:51,  6.26s/it, lr: 1.0e-04 loss: 4.690e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2273/30000 [5:07:23<49:12:52,  6.39s/it, lr: 1.0e-04 loss: 4.690e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2273/30000 [5:07:23<49:12:52,  6.39s/it, lr: 1.0e-04 loss: 4.690e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2273/30000 [5:07:32<49:12:52,  6.39s/it, lr: 1.0e-04 loss: 6.659e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2273/30000 [5:07:32<49:12:52,  6.39s/it, lr: 1.0e-04 loss: 6.659e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2274/30000 [5:07:32<55:16:01,  7.18s/it, lr: 1.0e-04 loss: 6.659e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2274/30000 [5:07:32<55:16:01,  7.18s/it, lr: 1.0e-04 loss: 6.659e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2274/30000 [5:07:39<55:16:01,  7.18s/it, lr: 1.0e-04 loss: 1.263e-01]
Xiang_wan_2_2_lora:   8%|##9                                    | 2274/30000 [5:07:39<55:16:01,  7.18s/it, lr: 1.0e-04 loss: 1.263e-01]
Xiang_wan_2_2_lora:   8%|##9                                    | 2275/30000 [5:07:39<53:50:03,  6.99s/it, lr: 1.0e-04 loss: 1.263e-01]
Xiang_wan_2_2_lora:   8%|##9                                    | 2275/30000 [5:07:39<53:50:03,  6.99s/it, lr: 1.0e-04 loss: 1.263e-01]
Xiang_wan_2_2_lora:   8%|##9                                    | 2275/30000 [5:07:48<53:50:03,  6.99s/it, lr: 1.0e-04 loss: 6.378e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2275/30000 [5:07:48<53:50:03,  6.99s/it, lr: 1.0e-04 loss: 6.378e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2276/30000 [5:07:48<58:56:39,  7.65s/it, lr: 1.0e-04 loss: 6.378e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2276/30000 [5:07:48<58:56:39,  7.65s/it, lr: 1.0e-04 loss: 6.378e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2276/30000 [5:07:57<58:56:39,  7.65s/it, lr: 1.0e-04 loss: 8.759e-01]
Xiang_wan_2_2_lora:   8%|##9                                    | 2276/30000 [5:07:57<58:56:39,  7.65s/it, lr: 1.0e-04 loss: 8.759e-01]
Xiang_wan_2_2_lora:   8%|##9                                    | 2277/30000 [5:07:57<62:36:16,  8.13s/it, lr: 1.0e-04 loss: 8.759e-01]
Xiang_wan_2_2_lora:   8%|##9                                    | 2277/30000 [5:07:57<62:36:16,  8.13s/it, lr: 1.0e-04 loss: 8.759e-01]
Xiang_wan_2_2_lora:   8%|##9                                    | 2277/30000 [5:08:07<62:36:16,  8.13s/it, lr: 1.0e-04 loss: 8.397e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2277/30000 [5:08:07<62:36:16,  8.13s/it, lr: 1.0e-04 loss: 8.397e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2278/30000 [5:08:07<65:18:05,  8.48s/it, lr: 1.0e-04 loss: 8.397e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2278/30000 [5:08:07<65:18:05,  8.48s/it, lr: 1.0e-04 loss: 8.397e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2278/30000 [5:08:16<65:18:05,  8.48s/it, lr: 1.0e-04 loss: 3.319e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2278/30000 [5:08:16<65:18:05,  8.48s/it, lr: 1.0e-04 loss: 3.319e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2279/30000 [5:08:16<67:12:44,  8.73s/it, lr: 1.0e-04 loss: 3.319e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2279/30000 [5:08:16<67:12:44,  8.73s/it, lr: 1.0e-04 loss: 3.319e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2279/30000 [5:08:25<67:12:44,  8.73s/it, lr: 1.0e-04 loss: 6.425e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2279/30000 [5:08:25<67:12:44,  8.73s/it, lr: 1.0e-04 loss: 6.425e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2280/30000 [5:08:35<67:12:35,  8.73s/it, lr: 1.0e-04 loss: 4.408e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2280/30000 [5:08:35<67:12:35,  8.73s/it, lr: 1.0e-04 loss: 4.408e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2281/30000 [5:08:35<52:52:12,  6.87s/it, lr: 1.0e-04 loss: 4.408e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2281/30000 [5:08:35<52:52:12,  6.87s/it, lr: 1.0e-04 loss: 4.408e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2281/30000 [5:08:44<52:52:12,  6.87s/it, lr: 1.0e-04 loss: 4.880e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2281/30000 [5:08:44<52:52:12,  6.87s/it, lr: 1.0e-04 loss: 4.880e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2282/30000 [5:08:44<57:36:29,  7.48s/it, lr: 1.0e-04 loss: 4.880e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2282/30000 [5:08:44<57:36:29,  7.48s/it, lr: 1.0e-04 loss: 4.880e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2282/30000 [5:08:51<57:36:29,  7.48s/it, lr: 1.0e-04 loss: 7.924e-01]
Xiang_wan_2_2_lora:   8%|##9                                    | 2282/30000 [5:08:51<57:36:29,  7.48s/it, lr: 1.0e-04 loss: 7.924e-01]
Xiang_wan_2_2_lora:   8%|##9                                    | 2283/30000 [5:08:51<55:45:49,  7.24s/it, lr: 1.0e-04 loss: 7.924e-01]
Xiang_wan_2_2_lora:   8%|##9                                    | 2283/30000 [5:08:51<55:45:49,  7.24s/it, lr: 1.0e-04 loss: 7.924e-01]
Xiang_wan_2_2_lora:   8%|##9                                    | 2283/30000 [5:08:57<55:45:49,  7.24s/it, lr: 1.0e-04 loss: 7.197e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2283/30000 [5:08:57<55:45:49,  7.24s/it, lr: 1.0e-04 loss: 7.197e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2284/30000 [5:08:57<54:17:54,  7.05s/it, lr: 1.0e-04 loss: 7.197e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2284/30000 [5:08:57<54:17:54,  7.05s/it, lr: 1.0e-04 loss: 7.197e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2284/30000 [5:09:04<54:17:54,  7.05s/it, lr: 1.0e-04 loss: 8.627e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2284/30000 [5:09:04<54:17:54,  7.05s/it, lr: 1.0e-04 loss: 8.627e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2285/30000 [5:09:04<53:07:51,  6.90s/it, lr: 1.0e-04 loss: 8.627e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2285/30000 [5:09:04<53:07:51,  6.90s/it, lr: 1.0e-04 loss: 8.627e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2285/30000 [5:09:10<53:07:51,  6.90s/it, lr: 1.0e-04 loss: 7.669e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2285/30000 [5:09:10<53:07:51,  6.90s/it, lr: 1.0e-04 loss: 7.669e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2286/30000 [5:09:10<52:19:43,  6.80s/it, lr: 1.0e-04 loss: 7.669e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2286/30000 [5:09:10<52:19:43,  6.80s/it, lr: 1.0e-04 loss: 7.669e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2286/30000 [5:09:17<52:19:43,  6.80s/it, lr: 1.0e-04 loss: 6.567e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2286/30000 [5:09:17<52:19:43,  6.80s/it, lr: 1.0e-04 loss: 6.567e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2287/30000 [5:09:17<51:41:15,  6.71s/it, lr: 1.0e-04 loss: 6.567e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2287/30000 [5:09:17<51:41:15,  6.71s/it, lr: 1.0e-04 loss: 6.567e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2287/30000 [5:09:27<51:41:15,  6.71s/it, lr: 1.0e-04 loss: 5.822e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2287/30000 [5:09:27<51:41:15,  6.71s/it, lr: 1.0e-04 loss: 5.822e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2288/30000 [5:09:27<58:37:33,  7.62s/it, lr: 1.0e-04 loss: 5.822e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2288/30000 [5:09:27<58:37:33,  7.62s/it, lr: 1.0e-04 loss: 5.822e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2288/30000 [5:09:33<58:37:33,  7.62s/it, lr: 1.0e-04 loss: 8.534e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2288/30000 [5:09:33<58:37:33,  7.62s/it, lr: 1.0e-04 loss: 8.534e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2289/30000 [5:09:33<56:23:10,  7.33s/it, lr: 1.0e-04 loss: 8.534e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2289/30000 [5:09:33<56:23:10,  7.33s/it, lr: 1.0e-04 loss: 8.534e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2289/30000 [5:09:42<56:23:10,  7.33s/it, lr: 1.0e-04 loss: 5.755e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2289/30000 [5:09:42<56:23:10,  7.33s/it, lr: 1.0e-04 loss: 5.755e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2290/30000 [5:09:49<56:23:02,  7.33s/it, lr: 1.0e-04 loss: 6.304e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2290/30000 [5:09:49<56:23:02,  7.33s/it, lr: 1.0e-04 loss: 6.304e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2291/30000 [5:09:49<42:11:23,  5.48s/it, lr: 1.0e-04 loss: 6.304e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2291/30000 [5:09:49<42:11:23,  5.48s/it, lr: 1.0e-04 loss: 6.304e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2291/30000 [5:09:58<42:11:23,  5.48s/it, lr: 1.0e-04 loss: 5.962e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2291/30000 [5:09:58<42:11:23,  5.48s/it, lr: 1.0e-04 loss: 5.962e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2292/30000 [5:09:58<49:22:58,  6.42s/it, lr: 1.0e-04 loss: 5.962e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2292/30000 [5:09:58<49:22:58,  6.42s/it, lr: 1.0e-04 loss: 5.962e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2292/30000 [5:10:08<49:22:58,  6.42s/it, lr: 1.0e-04 loss: 2.867e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2292/30000 [5:10:08<49:22:58,  6.42s/it, lr: 1.0e-04 loss: 2.867e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2293/30000 [5:10:08<55:16:15,  7.18s/it, lr: 1.0e-04 loss: 2.867e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2293/30000 [5:10:08<55:16:15,  7.18s/it, lr: 1.0e-04 loss: 2.867e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2293/30000 [5:10:14<55:16:15,  7.18s/it, lr: 1.0e-04 loss: 9.666e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2293/30000 [5:10:14<55:16:15,  7.18s/it, lr: 1.0e-04 loss: 9.666e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2294/30000 [5:10:14<53:53:37,  7.00s/it, lr: 1.0e-04 loss: 9.666e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2294/30000 [5:10:14<53:53:37,  7.00s/it, lr: 1.0e-04 loss: 9.666e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2294/30000 [5:10:21<53:53:37,  7.00s/it, lr: 1.0e-04 loss: 7.088e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2294/30000 [5:10:21<53:53:37,  7.00s/it, lr: 1.0e-04 loss: 7.088e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2295/30000 [5:10:21<53:22:54,  6.94s/it, lr: 1.0e-04 loss: 7.088e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2295/30000 [5:10:21<53:22:54,  6.94s/it, lr: 1.0e-04 loss: 7.088e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2295/30000 [5:10:30<53:22:54,  6.94s/it, lr: 1.0e-04 loss: 8.830e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2295/30000 [5:10:30<53:22:54,  6.94s/it, lr: 1.0e-04 loss: 8.830e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2296/30000 [5:10:30<58:39:27,  7.62s/it, lr: 1.0e-04 loss: 8.830e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2296/30000 [5:10:30<58:39:27,  7.62s/it, lr: 1.0e-04 loss: 8.830e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2296/30000 [5:10:37<58:39:27,  7.62s/it, lr: 1.0e-04 loss: 1.298e-01]
Xiang_wan_2_2_lora:   8%|##9                                    | 2296/30000 [5:10:37<58:39:27,  7.62s/it, lr: 1.0e-04 loss: 1.298e-01]
Xiang_wan_2_2_lora:   8%|##9                                    | 2297/30000 [5:10:37<56:18:34,  7.32s/it, lr: 1.0e-04 loss: 1.298e-01]
Xiang_wan_2_2_lora:   8%|##9                                    | 2297/30000 [5:10:37<56:18:34,  7.32s/it, lr: 1.0e-04 loss: 1.298e-01]
Xiang_wan_2_2_lora:   8%|##9                                    | 2297/30000 [5:10:46<56:18:34,  7.32s/it, lr: 1.0e-04 loss: 5.130e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2297/30000 [5:10:46<56:18:34,  7.32s/it, lr: 1.0e-04 loss: 5.130e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2298/30000 [5:10:46<60:47:07,  7.90s/it, lr: 1.0e-04 loss: 5.130e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2298/30000 [5:10:46<60:47:07,  7.90s/it, lr: 1.0e-04 loss: 5.130e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2298/30000 [5:10:56<60:47:07,  7.90s/it, lr: 1.0e-04 loss: 2.577e-01]
Xiang_wan_2_2_lora:   8%|##9                                    | 2298/30000 [5:10:56<60:47:07,  7.90s/it, lr: 1.0e-04 loss: 2.577e-01]
Xiang_wan_2_2_lora:   8%|##9                                    | 2299/30000 [5:10:56<64:04:10,  8.33s/it, lr: 1.0e-04 loss: 2.577e-01]
Xiang_wan_2_2_lora:   8%|##9                                    | 2299/30000 [5:10:56<64:04:10,  8.33s/it, lr: 1.0e-04 loss: 2.577e-01]
Xiang_wan_2_2_lora:   8%|##9                                    | 2299/30000 [5:11:05<64:04:10,  8.33s/it, lr: 1.0e-04 loss: 6.873e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2299/30000 [5:11:05<64:04:10,  8.33s/it, lr: 1.0e-04 loss: 6.873e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2300/30000 [5:11:14<64:04:02,  8.33s/it, lr: 1.0e-04 loss: 6.720e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2300/30000 [5:11:14<64:04:02,  8.33s/it, lr: 1.0e-04 loss: 6.720e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2301/30000 [5:11:14<51:06:46,  6.64s/it, lr: 1.0e-04 loss: 6.720e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2301/30000 [5:11:14<51:06:46,  6.64s/it, lr: 1.0e-04 loss: 6.720e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2301/30000 [5:11:21<51:06:46,  6.64s/it, lr: 1.0e-04 loss: 1.195e-01]
Xiang_wan_2_2_lora:   8%|##9                                    | 2301/30000 [5:11:21<51:06:46,  6.64s/it, lr: 1.0e-04 loss: 1.195e-01]
Xiang_wan_2_2_lora:   8%|##9                                    | 2302/30000 [5:11:21<50:53:51,  6.62s/it, lr: 1.0e-04 loss: 1.195e-01]
Xiang_wan_2_2_lora:   8%|##9                                    | 2302/30000 [5:11:21<50:53:51,  6.62s/it, lr: 1.0e-04 loss: 1.195e-01]
Xiang_wan_2_2_lora:   8%|##9                                    | 2302/30000 [5:11:30<50:53:51,  6.62s/it, lr: 1.0e-04 loss: 5.018e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2302/30000 [5:11:30<50:53:51,  6.62s/it, lr: 1.0e-04 loss: 5.018e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2303/30000 [5:11:30<56:47:37,  7.38s/it, lr: 1.0e-04 loss: 5.018e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2303/30000 [5:11:30<56:47:37,  7.38s/it, lr: 1.0e-04 loss: 5.018e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2303/30000 [5:11:40<56:47:37,  7.38s/it, lr: 1.0e-04 loss: 1.132e-01]
Xiang_wan_2_2_lora:   8%|##9                                    | 2303/30000 [5:11:40<56:47:37,  7.38s/it, lr: 1.0e-04 loss: 1.132e-01]
Xiang_wan_2_2_lora:   8%|##9                                    | 2304/30000 [5:11:40<60:52:39,  7.91s/it, lr: 1.0e-04 loss: 1.132e-01]
Xiang_wan_2_2_lora:   8%|##9                                    | 2304/30000 [5:11:40<60:52:39,  7.91s/it, lr: 1.0e-04 loss: 1.132e-01]
Xiang_wan_2_2_lora:   8%|##9                                    | 2304/30000 [5:11:49<60:52:39,  7.91s/it, lr: 1.0e-04 loss: 1.096e-01]
Xiang_wan_2_2_lora:   8%|##9                                    | 2304/30000 [5:11:49<60:52:39,  7.91s/it, lr: 1.0e-04 loss: 1.096e-01]
Xiang_wan_2_2_lora:   8%|##9                                    | 2305/30000 [5:11:49<63:56:52,  8.31s/it, lr: 1.0e-04 loss: 1.096e-01]
Xiang_wan_2_2_lora:   8%|##9                                    | 2305/30000 [5:11:49<63:56:52,  8.31s/it, lr: 1.0e-04 loss: 1.096e-01]
Xiang_wan_2_2_lora:   8%|##9                                    | 2305/30000 [5:11:58<63:56:52,  8.31s/it, lr: 1.0e-04 loss: 6.533e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2305/30000 [5:11:58<63:56:52,  8.31s/it, lr: 1.0e-04 loss: 6.533e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2306/30000 [5:11:58<66:18:39,  8.62s/it, lr: 1.0e-04 loss: 6.533e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2306/30000 [5:11:58<66:18:39,  8.62s/it, lr: 1.0e-04 loss: 6.533e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2306/30000 [5:12:05<66:18:39,  8.62s/it, lr: 1.0e-04 loss: 1.162e-01]
Xiang_wan_2_2_lora:   8%|##9                                    | 2306/30000 [5:12:05<66:18:39,  8.62s/it, lr: 1.0e-04 loss: 1.162e-01]
Xiang_wan_2_2_lora:   8%|##9                                    | 2307/30000 [5:12:05<61:44:13,  8.03s/it, lr: 1.0e-04 loss: 1.162e-01]
Xiang_wan_2_2_lora:   8%|##9                                    | 2307/30000 [5:12:05<61:44:13,  8.03s/it, lr: 1.0e-04 loss: 1.162e-01]
Xiang_wan_2_2_lora:   8%|##9                                    | 2307/30000 [5:12:11<61:44:13,  8.03s/it, lr: 1.0e-04 loss: 8.508e-02]
Xiang_wan_2_2_lora:   8%|##9                                    | 2307/30000 [5:12:11<61:44:13,  8.03s/it, lr: 1.0e-04 loss: 8.508e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2308/30000 [5:12:11<58:18:48,  7.58s/it, lr: 1.0e-04 loss: 8.508e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2308/30000 [5:12:11<58:18:48,  7.58s/it, lr: 1.0e-04 loss: 8.508e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2308/30000 [5:12:18<58:18:48,  7.58s/it, lr: 1.0e-04 loss: 1.213e-01]
Xiang_wan_2_2_lora:   8%|###                                    | 2308/30000 [5:12:18<58:18:48,  7.58s/it, lr: 1.0e-04 loss: 1.213e-01]
Xiang_wan_2_2_lora:   8%|###                                    | 2309/30000 [5:12:18<55:56:10,  7.27s/it, lr: 1.0e-04 loss: 1.213e-01]
Xiang_wan_2_2_lora:   8%|###                                    | 2309/30000 [5:12:18<55:56:10,  7.27s/it, lr: 1.0e-04 loss: 1.213e-01]
Xiang_wan_2_2_lora:   8%|###                                    | 2309/30000 [5:12:25<55:56:10,  7.27s/it, lr: 1.0e-04 loss: 1.026e-01]
Xiang_wan_2_2_lora:   8%|###                                    | 2309/30000 [5:12:25<55:56:10,  7.27s/it, lr: 1.0e-04 loss: 1.026e-01]
Xiang_wan_2_2_lora:   8%|###                                    | 2310/30000 [5:12:31<55:56:02,  7.27s/it, lr: 1.0e-04 loss: 1.428e-01]
Xiang_wan_2_2_lora:   8%|###                                    | 2310/30000 [5:12:31<55:56:02,  7.27s/it, lr: 1.0e-04 loss: 1.428e-01]
Xiang_wan_2_2_lora:   8%|###                                    | 2311/30000 [5:12:31<41:45:56,  5.43s/it, lr: 1.0e-04 loss: 1.428e-01]
Xiang_wan_2_2_lora:   8%|###                                    | 2311/30000 [5:12:31<41:45:56,  5.43s/it, lr: 1.0e-04 loss: 1.428e-01]
Xiang_wan_2_2_lora:   8%|###                                    | 2311/30000 [5:12:41<41:45:56,  5.43s/it, lr: 1.0e-04 loss: 4.555e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2311/30000 [5:12:41<41:45:56,  5.43s/it, lr: 1.0e-04 loss: 4.555e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2312/30000 [5:12:41<49:03:44,  6.38s/it, lr: 1.0e-04 loss: 4.555e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2312/30000 [5:12:41<49:03:44,  6.38s/it, lr: 1.0e-04 loss: 4.555e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2312/30000 [5:12:47<49:03:44,  6.38s/it, lr: 1.0e-04 loss: 5.898e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2312/30000 [5:12:47<49:03:44,  6.38s/it, lr: 1.0e-04 loss: 5.898e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2313/30000 [5:12:47<49:19:38,  6.41s/it, lr: 1.0e-04 loss: 5.898e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2313/30000 [5:12:47<49:19:38,  6.41s/it, lr: 1.0e-04 loss: 5.898e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2313/30000 [5:12:56<49:19:38,  6.41s/it, lr: 1.0e-04 loss: 8.764e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2313/30000 [5:12:56<49:19:38,  6.41s/it, lr: 1.0e-04 loss: 8.764e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2314/30000 [5:12:56<55:15:30,  7.19s/it, lr: 1.0e-04 loss: 8.764e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2314/30000 [5:12:56<55:15:30,  7.19s/it, lr: 1.0e-04 loss: 8.764e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2314/30000 [5:13:03<55:15:30,  7.19s/it, lr: 1.0e-04 loss: 5.844e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2314/30000 [5:13:03<55:15:30,  7.19s/it, lr: 1.0e-04 loss: 5.844e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2315/30000 [5:13:03<53:55:49,  7.01s/it, lr: 1.0e-04 loss: 5.844e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2315/30000 [5:13:03<53:55:49,  7.01s/it, lr: 1.0e-04 loss: 5.844e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2315/30000 [5:13:15<53:55:49,  7.01s/it, lr: 1.0e-04 loss: 8.581e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2315/30000 [5:13:15<53:55:49,  7.01s/it, lr: 1.0e-04 loss: 8.581e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2316/30000 [5:13:15<62:35:15,  8.14s/it, lr: 1.0e-04 loss: 8.581e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2316/30000 [5:13:15<62:35:15,  8.14s/it, lr: 1.0e-04 loss: 8.581e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2316/30000 [5:13:25<62:35:15,  8.14s/it, lr: 1.0e-04 loss: 3.412e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2316/30000 [5:13:25<62:35:15,  8.14s/it, lr: 1.0e-04 loss: 3.412e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2317/30000 [5:13:25<65:49:06,  8.56s/it, lr: 1.0e-04 loss: 3.412e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2317/30000 [5:13:25<65:49:06,  8.56s/it, lr: 1.0e-04 loss: 3.412e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2317/30000 [5:13:32<65:49:06,  8.56s/it, lr: 1.0e-04 loss: 8.950e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2317/30000 [5:13:32<65:49:06,  8.56s/it, lr: 1.0e-04 loss: 8.950e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2318/30000 [5:13:32<63:13:07,  8.22s/it, lr: 1.0e-04 loss: 8.950e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2318/30000 [5:13:32<63:13:07,  8.22s/it, lr: 1.0e-04 loss: 8.950e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2318/30000 [5:13:39<63:13:07,  8.22s/it, lr: 1.0e-04 loss: 6.998e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2318/30000 [5:13:39<63:13:07,  8.22s/it, lr: 1.0e-04 loss: 6.998e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2319/30000 [5:13:39<60:57:47,  7.93s/it, lr: 1.0e-04 loss: 6.998e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2319/30000 [5:13:39<60:57:47,  7.93s/it, lr: 1.0e-04 loss: 6.998e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2319/30000 [5:13:46<60:57:47,  7.93s/it, lr: 1.0e-04 loss: 7.740e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2319/30000 [5:13:46<60:57:47,  7.93s/it, lr: 1.0e-04 loss: 7.740e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2320/30000 [5:13:53<60:57:39,  7.93s/it, lr: 1.0e-04 loss: 7.404e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2320/30000 [5:13:53<60:57:39,  7.93s/it, lr: 1.0e-04 loss: 7.404e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2321/30000 [5:13:53<44:55:40,  5.84s/it, lr: 1.0e-04 loss: 7.404e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2321/30000 [5:13:53<44:55:40,  5.84s/it, lr: 1.0e-04 loss: 7.404e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2321/30000 [5:14:02<44:55:40,  5.84s/it, lr: 1.0e-04 loss: 6.468e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2321/30000 [5:14:02<44:55:40,  5.84s/it, lr: 1.0e-04 loss: 6.468e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2322/30000 [5:14:02<51:40:12,  6.72s/it, lr: 1.0e-04 loss: 6.468e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2322/30000 [5:14:02<51:40:12,  6.72s/it, lr: 1.0e-04 loss: 6.468e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2322/30000 [5:14:12<51:40:12,  6.72s/it, lr: 1.0e-04 loss: 5.908e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2322/30000 [5:14:12<51:40:12,  6.72s/it, lr: 1.0e-04 loss: 5.908e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2323/30000 [5:14:12<56:53:40,  7.40s/it, lr: 1.0e-04 loss: 5.908e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2323/30000 [5:14:12<56:53:40,  7.40s/it, lr: 1.0e-04 loss: 5.908e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2323/30000 [5:14:21<56:53:40,  7.40s/it, lr: 1.0e-04 loss: 6.084e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2323/30000 [5:14:21<56:53:40,  7.40s/it, lr: 1.0e-04 loss: 6.084e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2324/30000 [5:14:21<61:02:35,  7.94s/it, lr: 1.0e-04 loss: 6.084e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2324/30000 [5:14:21<61:02:35,  7.94s/it, lr: 1.0e-04 loss: 6.084e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2324/30000 [5:14:31<61:02:35,  7.94s/it, lr: 1.0e-04 loss: 3.455e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2324/30000 [5:14:31<61:02:35,  7.94s/it, lr: 1.0e-04 loss: 3.455e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2325/30000 [5:14:31<64:25:16,  8.38s/it, lr: 1.0e-04 loss: 3.455e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2325/30000 [5:14:31<64:25:16,  8.38s/it, lr: 1.0e-04 loss: 3.455e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2325/30000 [5:14:37<64:25:16,  8.38s/it, lr: 1.0e-04 loss: 1.119e-01]
Xiang_wan_2_2_lora:   8%|###                                    | 2325/30000 [5:14:37<64:25:16,  8.38s/it, lr: 1.0e-04 loss: 1.119e-01]
Xiang_wan_2_2_lora:   8%|###                                    | 2326/30000 [5:14:37<60:27:54,  7.87s/it, lr: 1.0e-04 loss: 1.119e-01]
Xiang_wan_2_2_lora:   8%|###                                    | 2326/30000 [5:14:37<60:27:54,  7.87s/it, lr: 1.0e-04 loss: 1.119e-01]
Xiang_wan_2_2_lora:   8%|###                                    | 2326/30000 [5:14:46<60:27:54,  7.87s/it, lr: 1.0e-04 loss: 3.702e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2326/30000 [5:14:46<60:27:54,  7.87s/it, lr: 1.0e-04 loss: 3.702e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2327/30000 [5:14:46<63:40:19,  8.28s/it, lr: 1.0e-04 loss: 3.702e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2327/30000 [5:14:46<63:40:19,  8.28s/it, lr: 1.0e-04 loss: 3.702e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2327/30000 [5:14:53<63:40:19,  8.28s/it, lr: 1.0e-04 loss: 9.213e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2327/30000 [5:14:53<63:40:19,  8.28s/it, lr: 1.0e-04 loss: 9.213e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2328/30000 [5:14:53<59:43:48,  7.77s/it, lr: 1.0e-04 loss: 9.213e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2328/30000 [5:14:53<59:43:48,  7.77s/it, lr: 1.0e-04 loss: 9.213e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2328/30000 [5:15:02<59:43:48,  7.77s/it, lr: 1.0e-04 loss: 5.293e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2328/30000 [5:15:02<59:43:48,  7.77s/it, lr: 1.0e-04 loss: 5.293e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2329/30000 [5:15:02<63:28:43,  8.26s/it, lr: 1.0e-04 loss: 5.293e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2329/30000 [5:15:02<63:28:43,  8.26s/it, lr: 1.0e-04 loss: 5.293e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2329/30000 [5:15:12<63:28:43,  8.26s/it, lr: 1.0e-04 loss: 6.479e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2329/30000 [5:15:12<63:28:43,  8.26s/it, lr: 1.0e-04 loss: 6.479e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2330/30000 [5:15:18<63:28:35,  8.26s/it, lr: 1.0e-04 loss: 3.559e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2330/30000 [5:15:18<63:28:35,  8.26s/it, lr: 1.0e-04 loss: 3.559e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2331/30000 [5:15:18<46:03:27,  5.99s/it, lr: 1.0e-04 loss: 3.559e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2331/30000 [5:15:18<46:03:27,  5.99s/it, lr: 1.0e-04 loss: 3.559e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2331/30000 [5:15:28<46:03:27,  5.99s/it, lr: 1.0e-04 loss: 7.473e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2331/30000 [5:15:28<46:03:27,  5.99s/it, lr: 1.0e-04 loss: 7.473e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2332/30000 [5:15:28<52:37:31,  6.85s/it, lr: 1.0e-04 loss: 7.473e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2332/30000 [5:15:28<52:37:31,  6.85s/it, lr: 1.0e-04 loss: 7.473e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2332/30000 [5:15:37<52:37:31,  6.85s/it, lr: 1.0e-04 loss: 2.596e-01]
Xiang_wan_2_2_lora:   8%|###                                    | 2332/30000 [5:15:37<52:37:31,  6.85s/it, lr: 1.0e-04 loss: 2.596e-01]
Xiang_wan_2_2_lora:   8%|###                                    | 2333/30000 [5:15:37<57:40:12,  7.50s/it, lr: 1.0e-04 loss: 2.596e-01]
Xiang_wan_2_2_lora:   8%|###                                    | 2333/30000 [5:15:37<57:40:12,  7.50s/it, lr: 1.0e-04 loss: 2.596e-01]
Xiang_wan_2_2_lora:   8%|###                                    | 2333/30000 [5:15:46<57:40:12,  7.50s/it, lr: 1.0e-04 loss: 6.746e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2333/30000 [5:15:46<57:40:12,  7.50s/it, lr: 1.0e-04 loss: 6.746e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2334/30000 [5:15:46<61:19:57,  7.98s/it, lr: 1.0e-04 loss: 6.746e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2334/30000 [5:15:46<61:19:57,  7.98s/it, lr: 1.0e-04 loss: 6.746e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2334/30000 [5:15:53<61:19:57,  7.98s/it, lr: 1.0e-04 loss: 5.536e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2334/30000 [5:15:53<61:19:57,  7.98s/it, lr: 1.0e-04 loss: 5.536e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2335/30000 [5:15:53<58:22:38,  7.60s/it, lr: 1.0e-04 loss: 5.536e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2335/30000 [5:15:53<58:22:38,  7.60s/it, lr: 1.0e-04 loss: 5.536e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2335/30000 [5:16:02<58:22:38,  7.60s/it, lr: 1.0e-04 loss: 1.011e-01]
Xiang_wan_2_2_lora:   8%|###                                    | 2335/30000 [5:16:02<58:22:38,  7.60s/it, lr: 1.0e-04 loss: 1.011e-01]
Xiang_wan_2_2_lora:   8%|###                                    | 2336/30000 [5:16:02<62:15:31,  8.10s/it, lr: 1.0e-04 loss: 1.011e-01]
Xiang_wan_2_2_lora:   8%|###                                    | 2336/30000 [5:16:02<62:15:31,  8.10s/it, lr: 1.0e-04 loss: 1.011e-01]
Xiang_wan_2_2_lora:   8%|###                                    | 2336/30000 [5:16:09<62:15:31,  8.10s/it, lr: 1.0e-04 loss: 7.546e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2336/30000 [5:16:09<62:15:31,  8.10s/it, lr: 1.0e-04 loss: 7.546e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2337/30000 [5:16:09<58:59:57,  7.68s/it, lr: 1.0e-04 loss: 7.546e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2337/30000 [5:16:09<58:59:57,  7.68s/it, lr: 1.0e-04 loss: 7.546e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2337/30000 [5:16:18<58:59:57,  7.68s/it, lr: 1.0e-04 loss: 7.782e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2337/30000 [5:16:18<58:59:57,  7.68s/it, lr: 1.0e-04 loss: 7.782e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2338/30000 [5:16:18<62:34:09,  8.14s/it, lr: 1.0e-04 loss: 7.782e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2338/30000 [5:16:18<62:34:09,  8.14s/it, lr: 1.0e-04 loss: 7.782e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2338/30000 [5:16:28<62:34:09,  8.14s/it, lr: 1.0e-04 loss: 1.593e-01]
Xiang_wan_2_2_lora:   8%|###                                    | 2338/30000 [5:16:28<62:34:09,  8.14s/it, lr: 1.0e-04 loss: 1.593e-01]
Xiang_wan_2_2_lora:   8%|###                                    | 2339/30000 [5:16:28<65:21:22,  8.51s/it, lr: 1.0e-04 loss: 1.593e-01]
Xiang_wan_2_2_lora:   8%|###                                    | 2339/30000 [5:16:28<65:21:22,  8.51s/it, lr: 1.0e-04 loss: 1.593e-01]
Xiang_wan_2_2_lora:   8%|###                                    | 2339/30000 [5:16:35<65:21:22,  8.51s/it, lr: 1.0e-04 loss: 1.138e-01]
Xiang_wan_2_2_lora:   8%|###                                    | 2339/30000 [5:16:35<65:21:22,  8.51s/it, lr: 1.0e-04 loss: 1.138e-01]
Xiang_wan_2_2_lora:   8%|###                                    | 2340/30000 [5:16:41<65:21:13,  8.51s/it, lr: 1.0e-04 loss: 9.668e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2340/30000 [5:16:41<65:21:13,  8.51s/it, lr: 1.0e-04 loss: 9.668e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2341/30000 [5:16:41<46:54:05,  6.10s/it, lr: 1.0e-04 loss: 9.668e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2341/30000 [5:16:41<46:54:05,  6.10s/it, lr: 1.0e-04 loss: 9.668e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2341/30000 [5:16:50<46:54:05,  6.10s/it, lr: 1.0e-04 loss: 8.497e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2341/30000 [5:16:50<46:54:05,  6.10s/it, lr: 1.0e-04 loss: 8.497e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2342/30000 [5:16:50<52:53:55,  6.89s/it, lr: 1.0e-04 loss: 8.497e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2342/30000 [5:16:50<52:53:55,  6.89s/it, lr: 1.0e-04 loss: 8.497e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2342/30000 [5:16:57<52:53:55,  6.89s/it, lr: 1.0e-04 loss: 5.364e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2342/30000 [5:16:57<52:53:55,  6.89s/it, lr: 1.0e-04 loss: 5.364e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2343/30000 [5:16:57<52:13:10,  6.80s/it, lr: 1.0e-04 loss: 5.364e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2343/30000 [5:16:57<52:13:10,  6.80s/it, lr: 1.0e-04 loss: 5.364e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2343/30000 [5:17:06<52:13:10,  6.80s/it, lr: 1.0e-04 loss: 5.573e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2343/30000 [5:17:06<52:13:10,  6.80s/it, lr: 1.0e-04 loss: 5.573e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2344/30000 [5:17:06<57:23:05,  7.47s/it, lr: 1.0e-04 loss: 5.573e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2344/30000 [5:17:06<57:23:05,  7.47s/it, lr: 1.0e-04 loss: 5.573e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2344/30000 [5:17:13<57:23:05,  7.47s/it, lr: 1.0e-04 loss: 1.809e-01]
Xiang_wan_2_2_lora:   8%|###                                    | 2344/30000 [5:17:13<57:23:05,  7.47s/it, lr: 1.0e-04 loss: 1.809e-01]
Xiang_wan_2_2_lora:   8%|###                                    | 2345/30000 [5:17:13<55:30:45,  7.23s/it, lr: 1.0e-04 loss: 1.809e-01]
Xiang_wan_2_2_lora:   8%|###                                    | 2345/30000 [5:17:13<55:30:45,  7.23s/it, lr: 1.0e-04 loss: 1.809e-01]
Xiang_wan_2_2_lora:   8%|###                                    | 2345/30000 [5:17:19<55:30:45,  7.23s/it, lr: 1.0e-04 loss: 1.365e-01]
Xiang_wan_2_2_lora:   8%|###                                    | 2345/30000 [5:17:19<55:30:45,  7.23s/it, lr: 1.0e-04 loss: 1.365e-01]
Xiang_wan_2_2_lora:   8%|###                                    | 2346/30000 [5:17:19<53:56:56,  7.02s/it, lr: 1.0e-04 loss: 1.365e-01]
Xiang_wan_2_2_lora:   8%|###                                    | 2346/30000 [5:17:19<53:56:56,  7.02s/it, lr: 1.0e-04 loss: 1.365e-01]
Xiang_wan_2_2_lora:   8%|###                                    | 2346/30000 [5:17:26<53:56:56,  7.02s/it, lr: 1.0e-04 loss: 5.342e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2346/30000 [5:17:26<53:56:56,  7.02s/it, lr: 1.0e-04 loss: 5.342e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2347/30000 [5:17:26<53:31:17,  6.97s/it, lr: 1.0e-04 loss: 5.342e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2347/30000 [5:17:26<53:31:17,  6.97s/it, lr: 1.0e-04 loss: 5.342e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2347/30000 [5:17:35<53:31:17,  6.97s/it, lr: 1.0e-04 loss: 7.102e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2347/30000 [5:17:35<53:31:17,  6.97s/it, lr: 1.0e-04 loss: 7.102e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2348/30000 [5:17:35<58:42:23,  7.64s/it, lr: 1.0e-04 loss: 7.102e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2348/30000 [5:17:35<58:42:23,  7.64s/it, lr: 1.0e-04 loss: 7.102e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2348/30000 [5:17:45<58:42:23,  7.64s/it, lr: 1.0e-04 loss: 5.032e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2348/30000 [5:17:45<58:42:23,  7.64s/it, lr: 1.0e-04 loss: 5.032e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2349/30000 [5:17:45<62:27:00,  8.13s/it, lr: 1.0e-04 loss: 5.032e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2349/30000 [5:17:45<62:27:00,  8.13s/it, lr: 1.0e-04 loss: 5.032e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2349/30000 [5:17:51<62:27:00,  8.13s/it, lr: 1.0e-04 loss: 8.792e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2349/30000 [5:17:51<62:27:00,  8.13s/it, lr: 1.0e-04 loss: 8.792e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2350/30000 [5:18:00<62:26:52,  8.13s/it, lr: 1.0e-04 loss: 3.389e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2350/30000 [5:18:00<62:26:52,  8.13s/it, lr: 1.0e-04 loss: 3.389e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2351/30000 [5:18:00<50:12:22,  6.54s/it, lr: 1.0e-04 loss: 3.389e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2351/30000 [5:18:00<50:12:22,  6.54s/it, lr: 1.0e-04 loss: 3.389e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2351/30000 [5:18:10<50:12:22,  6.54s/it, lr: 1.0e-04 loss: 6.618e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2351/30000 [5:18:10<50:12:22,  6.54s/it, lr: 1.0e-04 loss: 6.618e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2352/30000 [5:18:10<55:31:39,  7.23s/it, lr: 1.0e-04 loss: 6.618e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2352/30000 [5:18:10<55:31:39,  7.23s/it, lr: 1.0e-04 loss: 6.618e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2352/30000 [5:18:16<55:31:39,  7.23s/it, lr: 1.0e-04 loss: 7.188e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2352/30000 [5:18:16<55:31:39,  7.23s/it, lr: 1.0e-04 loss: 7.188e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2353/30000 [5:18:16<54:11:07,  7.06s/it, lr: 1.0e-04 loss: 7.188e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2353/30000 [5:18:16<54:11:07,  7.06s/it, lr: 1.0e-04 loss: 7.188e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2353/30000 [5:18:26<54:11:07,  7.06s/it, lr: 1.0e-04 loss: 7.869e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2353/30000 [5:18:26<54:11:07,  7.06s/it, lr: 1.0e-04 loss: 7.869e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2354/30000 [5:18:26<58:59:22,  7.68s/it, lr: 1.0e-04 loss: 7.869e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2354/30000 [5:18:26<58:59:22,  7.68s/it, lr: 1.0e-04 loss: 7.869e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2354/30000 [5:18:35<58:59:22,  7.68s/it, lr: 1.0e-04 loss: 1.548e-01]
Xiang_wan_2_2_lora:   8%|###                                    | 2354/30000 [5:18:35<58:59:22,  7.68s/it, lr: 1.0e-04 loss: 1.548e-01]
Xiang_wan_2_2_lora:   8%|###                                    | 2355/30000 [5:18:35<62:51:50,  8.19s/it, lr: 1.0e-04 loss: 1.548e-01]
Xiang_wan_2_2_lora:   8%|###                                    | 2355/30000 [5:18:35<62:51:50,  8.19s/it, lr: 1.0e-04 loss: 1.548e-01]
Xiang_wan_2_2_lora:   8%|###                                    | 2355/30000 [5:18:45<62:51:50,  8.19s/it, lr: 1.0e-04 loss: 7.341e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2355/30000 [5:18:45<62:51:50,  8.19s/it, lr: 1.0e-04 loss: 7.341e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2356/30000 [5:18:45<65:27:14,  8.52s/it, lr: 1.0e-04 loss: 7.341e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2356/30000 [5:18:45<65:27:14,  8.52s/it, lr: 1.0e-04 loss: 7.341e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2356/30000 [5:18:51<65:27:14,  8.52s/it, lr: 1.0e-04 loss: 1.907e-01]
Xiang_wan_2_2_lora:   8%|###                                    | 2356/30000 [5:18:51<65:27:14,  8.52s/it, lr: 1.0e-04 loss: 1.907e-01]
Xiang_wan_2_2_lora:   8%|###                                    | 2357/30000 [5:18:51<60:58:32,  7.94s/it, lr: 1.0e-04 loss: 1.907e-01]
Xiang_wan_2_2_lora:   8%|###                                    | 2357/30000 [5:18:51<60:58:32,  7.94s/it, lr: 1.0e-04 loss: 1.907e-01]
Xiang_wan_2_2_lora:   8%|###                                    | 2357/30000 [5:18:58<60:58:32,  7.94s/it, lr: 1.0e-04 loss: 8.868e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2357/30000 [5:18:58<60:58:32,  7.94s/it, lr: 1.0e-04 loss: 8.868e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2358/30000 [5:18:58<58:01:43,  7.56s/it, lr: 1.0e-04 loss: 8.868e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2358/30000 [5:18:58<58:01:43,  7.56s/it, lr: 1.0e-04 loss: 8.868e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2358/30000 [5:19:04<58:01:43,  7.56s/it, lr: 1.0e-04 loss: 6.552e-01]
Xiang_wan_2_2_lora:   8%|###                                    | 2358/30000 [5:19:04<58:01:43,  7.56s/it, lr: 1.0e-04 loss: 6.552e-01]
Xiang_wan_2_2_lora:   8%|###                                    | 2359/30000 [5:19:04<55:36:45,  7.24s/it, lr: 1.0e-04 loss: 6.552e-01]
Xiang_wan_2_2_lora:   8%|###                                    | 2359/30000 [5:19:04<55:36:45,  7.24s/it, lr: 1.0e-04 loss: 6.552e-01]
Xiang_wan_2_2_lora:   8%|###                                    | 2359/30000 [5:19:14<55:36:45,  7.24s/it, lr: 1.0e-04 loss: 8.895e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2359/30000 [5:19:14<55:36:45,  7.24s/it, lr: 1.0e-04 loss: 8.895e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2360/30000 [5:19:20<55:36:37,  7.24s/it, lr: 1.0e-04 loss: 6.212e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2360/30000 [5:19:20<55:36:37,  7.24s/it, lr: 1.0e-04 loss: 6.212e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2361/30000 [5:19:20<41:42:00,  5.43s/it, lr: 1.0e-04 loss: 6.212e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2361/30000 [5:19:20<41:42:00,  5.43s/it, lr: 1.0e-04 loss: 6.212e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2361/30000 [5:19:30<41:42:00,  5.43s/it, lr: 1.0e-04 loss: 8.888e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2361/30000 [5:19:30<41:42:00,  5.43s/it, lr: 1.0e-04 loss: 8.888e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2362/30000 [5:19:30<49:29:47,  6.45s/it, lr: 1.0e-04 loss: 8.888e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2362/30000 [5:19:30<49:29:47,  6.45s/it, lr: 1.0e-04 loss: 8.888e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2362/30000 [5:19:39<49:29:47,  6.45s/it, lr: 1.0e-04 loss: 8.992e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2362/30000 [5:19:39<49:29:47,  6.45s/it, lr: 1.0e-04 loss: 8.992e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2363/30000 [5:19:39<55:14:22,  7.20s/it, lr: 1.0e-04 loss: 8.992e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2363/30000 [5:19:39<55:14:22,  7.20s/it, lr: 1.0e-04 loss: 8.992e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2363/30000 [5:19:48<55:14:22,  7.20s/it, lr: 1.0e-04 loss: 7.023e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2363/30000 [5:19:48<55:14:22,  7.20s/it, lr: 1.0e-04 loss: 7.023e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2364/30000 [5:19:48<59:44:36,  7.78s/it, lr: 1.0e-04 loss: 7.023e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2364/30000 [5:19:48<59:44:36,  7.78s/it, lr: 1.0e-04 loss: 7.023e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2364/30000 [5:19:58<59:44:36,  7.78s/it, lr: 1.0e-04 loss: 3.908e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2364/30000 [5:19:58<59:44:36,  7.78s/it, lr: 1.0e-04 loss: 3.908e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2365/30000 [5:19:58<63:04:24,  8.22s/it, lr: 1.0e-04 loss: 3.908e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2365/30000 [5:19:58<63:04:24,  8.22s/it, lr: 1.0e-04 loss: 3.908e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2365/30000 [5:20:04<63:04:24,  8.22s/it, lr: 1.0e-04 loss: 1.051e-01]
Xiang_wan_2_2_lora:   8%|###                                    | 2365/30000 [5:20:04<63:04:24,  8.22s/it, lr: 1.0e-04 loss: 1.051e-01]
Xiang_wan_2_2_lora:   8%|###                                    | 2366/30000 [5:20:04<59:37:28,  7.77s/it, lr: 1.0e-04 loss: 1.051e-01]
Xiang_wan_2_2_lora:   8%|###                                    | 2366/30000 [5:20:04<59:37:28,  7.77s/it, lr: 1.0e-04 loss: 1.051e-01]
Xiang_wan_2_2_lora:   8%|###                                    | 2366/30000 [5:20:11<59:37:28,  7.77s/it, lr: 1.0e-04 loss: 4.728e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2366/30000 [5:20:11<59:37:28,  7.77s/it, lr: 1.0e-04 loss: 4.728e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2367/30000 [5:20:11<56:48:07,  7.40s/it, lr: 1.0e-04 loss: 4.728e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2367/30000 [5:20:11<56:48:07,  7.40s/it, lr: 1.0e-04 loss: 4.728e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2367/30000 [5:20:17<56:48:07,  7.40s/it, lr: 1.0e-04 loss: 7.548e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2367/30000 [5:20:17<56:48:07,  7.40s/it, lr: 1.0e-04 loss: 7.548e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2368/30000 [5:20:17<55:00:55,  7.17s/it, lr: 1.0e-04 loss: 7.548e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2368/30000 [5:20:17<55:00:55,  7.17s/it, lr: 1.0e-04 loss: 7.548e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2368/30000 [5:20:24<55:00:55,  7.17s/it, lr: 1.0e-04 loss: 6.564e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2368/30000 [5:20:24<55:00:55,  7.17s/it, lr: 1.0e-04 loss: 6.564e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2369/30000 [5:20:24<53:39:36,  6.99s/it, lr: 1.0e-04 loss: 6.564e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2369/30000 [5:20:24<53:39:36,  6.99s/it, lr: 1.0e-04 loss: 6.564e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2369/30000 [5:20:31<53:39:36,  6.99s/it, lr: 1.0e-04 loss: 7.486e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2369/30000 [5:20:31<53:39:36,  6.99s/it, lr: 1.0e-04 loss: 7.486e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2370/30000 [5:20:40<53:39:29,  6.99s/it, lr: 1.0e-04 loss: 6.551e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2370/30000 [5:20:40<53:39:29,  6.99s/it, lr: 1.0e-04 loss: 6.551e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2371/30000 [5:20:40<45:28:12,  5.92s/it, lr: 1.0e-04 loss: 6.551e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2371/30000 [5:20:40<45:28:12,  5.92s/it, lr: 1.0e-04 loss: 6.551e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2371/30000 [5:20:47<45:28:12,  5.92s/it, lr: 1.0e-04 loss: 8.481e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2371/30000 [5:20:47<45:28:12,  5.92s/it, lr: 1.0e-04 loss: 8.481e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2372/30000 [5:20:47<46:46:41,  6.10s/it, lr: 1.0e-04 loss: 8.481e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2372/30000 [5:20:47<46:46:41,  6.10s/it, lr: 1.0e-04 loss: 8.481e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2372/30000 [5:20:56<46:46:41,  6.10s/it, lr: 1.0e-04 loss: 5.344e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2372/30000 [5:20:56<46:46:41,  6.10s/it, lr: 1.0e-04 loss: 5.344e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2373/30000 [5:20:56<53:02:27,  6.91s/it, lr: 1.0e-04 loss: 5.344e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2373/30000 [5:20:56<53:02:27,  6.91s/it, lr: 1.0e-04 loss: 5.344e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2373/30000 [5:21:05<53:02:27,  6.91s/it, lr: 1.0e-04 loss: 4.950e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2373/30000 [5:21:05<53:02:27,  6.91s/it, lr: 1.0e-04 loss: 4.950e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2374/30000 [5:21:05<58:08:53,  7.58s/it, lr: 1.0e-04 loss: 4.950e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2374/30000 [5:21:05<58:08:53,  7.58s/it, lr: 1.0e-04 loss: 4.950e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2374/30000 [5:21:15<58:08:53,  7.58s/it, lr: 1.0e-04 loss: 8.039e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2374/30000 [5:21:15<58:08:53,  7.58s/it, lr: 1.0e-04 loss: 8.039e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2375/30000 [5:21:15<61:47:49,  8.05s/it, lr: 1.0e-04 loss: 8.039e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2375/30000 [5:21:15<61:47:49,  8.05s/it, lr: 1.0e-04 loss: 8.039e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2375/30000 [5:21:21<61:47:49,  8.05s/it, lr: 1.0e-04 loss: 8.060e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2375/30000 [5:21:21<61:47:49,  8.05s/it, lr: 1.0e-04 loss: 8.060e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2376/30000 [5:21:21<58:28:57,  7.62s/it, lr: 1.0e-04 loss: 8.060e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2376/30000 [5:21:21<58:28:57,  7.62s/it, lr: 1.0e-04 loss: 8.060e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2376/30000 [5:21:28<58:28:57,  7.62s/it, lr: 1.0e-04 loss: 7.896e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2376/30000 [5:21:28<58:28:57,  7.62s/it, lr: 1.0e-04 loss: 7.896e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2377/30000 [5:21:28<56:35:23,  7.38s/it, lr: 1.0e-04 loss: 7.896e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2377/30000 [5:21:28<56:35:23,  7.38s/it, lr: 1.0e-04 loss: 7.896e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2377/30000 [5:21:37<56:35:23,  7.38s/it, lr: 1.0e-04 loss: 5.841e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2377/30000 [5:21:37<56:35:23,  7.38s/it, lr: 1.0e-04 loss: 5.841e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2378/30000 [5:21:37<61:01:52,  7.95s/it, lr: 1.0e-04 loss: 5.841e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2378/30000 [5:21:37<61:01:52,  7.95s/it, lr: 1.0e-04 loss: 5.841e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2378/30000 [5:21:44<61:01:52,  7.95s/it, lr: 1.0e-04 loss: 4.418e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2378/30000 [5:21:44<61:01:52,  7.95s/it, lr: 1.0e-04 loss: 4.418e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2379/30000 [5:21:44<58:00:19,  7.56s/it, lr: 1.0e-04 loss: 4.418e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2379/30000 [5:21:44<58:00:19,  7.56s/it, lr: 1.0e-04 loss: 4.418e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2379/30000 [5:21:51<58:00:19,  7.56s/it, lr: 1.0e-04 loss: 6.813e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2379/30000 [5:21:51<58:00:19,  7.56s/it, lr: 1.0e-04 loss: 6.813e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2380/30000 [5:22:00<58:00:11,  7.56s/it, lr: 1.0e-04 loss: 5.024e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2380/30000 [5:22:00<58:00:11,  7.56s/it, lr: 1.0e-04 loss: 5.024e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2381/30000 [5:22:00<47:45:40,  6.23s/it, lr: 1.0e-04 loss: 5.024e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2381/30000 [5:22:00<47:45:40,  6.23s/it, lr: 1.0e-04 loss: 5.024e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2381/30000 [5:22:09<47:45:40,  6.23s/it, lr: 1.0e-04 loss: 8.476e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2381/30000 [5:22:09<47:45:40,  6.23s/it, lr: 1.0e-04 loss: 8.476e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2382/30000 [5:22:09<53:42:38,  7.00s/it, lr: 1.0e-04 loss: 8.476e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2382/30000 [5:22:09<53:42:38,  7.00s/it, lr: 1.0e-04 loss: 8.476e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2382/30000 [5:22:18<53:42:38,  7.00s/it, lr: 1.0e-04 loss: 3.924e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2382/30000 [5:22:18<53:42:38,  7.00s/it, lr: 1.0e-04 loss: 3.924e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2383/30000 [5:22:18<58:14:55,  7.59s/it, lr: 1.0e-04 loss: 3.924e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2383/30000 [5:22:18<58:14:55,  7.59s/it, lr: 1.0e-04 loss: 3.924e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2383/30000 [5:22:28<58:14:55,  7.59s/it, lr: 1.0e-04 loss: 5.055e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2383/30000 [5:22:28<58:14:55,  7.59s/it, lr: 1.0e-04 loss: 5.055e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2384/30000 [5:22:28<61:58:16,  8.08s/it, lr: 1.0e-04 loss: 5.055e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2384/30000 [5:22:28<61:58:16,  8.08s/it, lr: 1.0e-04 loss: 5.055e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2384/30000 [5:22:37<61:58:16,  8.08s/it, lr: 1.0e-04 loss: 6.201e-02]
Xiang_wan_2_2_lora:   8%|###                                    | 2384/30000 [5:22:37<61:58:16,  8.08s/it, lr: 1.0e-04 loss: 6.201e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2385/30000 [5:22:37<65:07:44,  8.49s/it, lr: 1.0e-04 loss: 6.201e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2385/30000 [5:22:37<65:07:44,  8.49s/it, lr: 1.0e-04 loss: 6.201e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2385/30000 [5:22:47<65:07:44,  8.49s/it, lr: 1.0e-04 loss: 5.204e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2385/30000 [5:22:47<65:07:44,  8.49s/it, lr: 1.0e-04 loss: 5.204e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2386/30000 [5:22:47<67:00:05,  8.73s/it, lr: 1.0e-04 loss: 5.204e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2386/30000 [5:22:47<67:00:05,  8.73s/it, lr: 1.0e-04 loss: 5.204e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2386/30000 [5:22:56<67:00:05,  8.73s/it, lr: 1.0e-04 loss: 5.444e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2386/30000 [5:22:56<67:00:05,  8.73s/it, lr: 1.0e-04 loss: 5.444e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2387/30000 [5:22:56<68:22:11,  8.91s/it, lr: 1.0e-04 loss: 5.444e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2387/30000 [5:22:56<68:22:11,  8.91s/it, lr: 1.0e-04 loss: 5.444e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2387/30000 [5:23:05<68:22:11,  8.91s/it, lr: 1.0e-04 loss: 7.780e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2387/30000 [5:23:05<68:22:11,  8.91s/it, lr: 1.0e-04 loss: 7.780e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2388/30000 [5:23:05<69:22:01,  9.04s/it, lr: 1.0e-04 loss: 7.780e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2388/30000 [5:23:05<69:22:01,  9.04s/it, lr: 1.0e-04 loss: 7.780e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2388/30000 [5:23:15<69:22:01,  9.04s/it, lr: 1.0e-04 loss: 4.765e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2388/30000 [5:23:15<69:22:01,  9.04s/it, lr: 1.0e-04 loss: 4.765e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2389/30000 [5:23:15<69:57:01,  9.12s/it, lr: 1.0e-04 loss: 4.765e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2389/30000 [5:23:15<69:57:01,  9.12s/it, lr: 1.0e-04 loss: 4.765e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2389/30000 [5:23:21<69:57:01,  9.12s/it, lr: 1.0e-04 loss: 7.335e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2389/30000 [5:23:21<69:57:01,  9.12s/it, lr: 1.0e-04 loss: 7.335e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2390/30000 [5:23:28<69:56:52,  9.12s/it, lr: 1.0e-04 loss: 1.278e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2390/30000 [5:23:28<69:56:52,  9.12s/it, lr: 1.0e-04 loss: 1.278e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2391/30000 [5:23:28<49:22:43,  6.44s/it, lr: 1.0e-04 loss: 1.278e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2391/30000 [5:23:28<49:22:43,  6.44s/it, lr: 1.0e-04 loss: 1.278e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2391/30000 [5:23:37<49:22:43,  6.44s/it, lr: 1.0e-04 loss: 6.668e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2391/30000 [5:23:37<49:22:43,  6.44s/it, lr: 1.0e-04 loss: 6.668e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2392/30000 [5:23:37<55:09:49,  7.19s/it, lr: 1.0e-04 loss: 6.668e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2392/30000 [5:23:37<55:09:49,  7.19s/it, lr: 1.0e-04 loss: 6.668e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2392/30000 [5:23:44<55:09:49,  7.19s/it, lr: 1.0e-04 loss: 1.562e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2392/30000 [5:23:44<55:09:49,  7.19s/it, lr: 1.0e-04 loss: 1.562e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2393/30000 [5:23:44<53:48:21,  7.02s/it, lr: 1.0e-04 loss: 1.562e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2393/30000 [5:23:44<53:48:21,  7.02s/it, lr: 1.0e-04 loss: 1.562e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2393/30000 [5:23:53<53:48:21,  7.02s/it, lr: 1.0e-04 loss: 6.447e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2393/30000 [5:23:53<53:48:21,  7.02s/it, lr: 1.0e-04 loss: 6.447e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2394/30000 [5:23:53<58:39:10,  7.65s/it, lr: 1.0e-04 loss: 6.447e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2394/30000 [5:23:53<58:39:10,  7.65s/it, lr: 1.0e-04 loss: 6.447e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2394/30000 [5:24:03<58:39:10,  7.65s/it, lr: 1.0e-04 loss: 6.852e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2394/30000 [5:24:03<58:39:10,  7.65s/it, lr: 1.0e-04 loss: 6.852e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2395/30000 [5:24:03<62:13:06,  8.11s/it, lr: 1.0e-04 loss: 6.852e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2395/30000 [5:24:03<62:13:06,  8.11s/it, lr: 1.0e-04 loss: 6.852e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2395/30000 [5:24:09<62:13:06,  8.11s/it, lr: 1.0e-04 loss: 9.518e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2395/30000 [5:24:09<62:13:06,  8.11s/it, lr: 1.0e-04 loss: 9.518e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2396/30000 [5:24:09<58:53:26,  7.68s/it, lr: 1.0e-04 loss: 9.518e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2396/30000 [5:24:09<58:53:26,  7.68s/it, lr: 1.0e-04 loss: 9.518e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2396/30000 [5:24:18<58:53:26,  7.68s/it, lr: 1.0e-04 loss: 7.577e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2396/30000 [5:24:18<58:53:26,  7.68s/it, lr: 1.0e-04 loss: 7.577e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2397/30000 [5:24:18<62:18:01,  8.13s/it, lr: 1.0e-04 loss: 7.577e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2397/30000 [5:24:18<62:18:01,  8.13s/it, lr: 1.0e-04 loss: 7.577e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2397/30000 [5:24:28<62:18:01,  8.13s/it, lr: 1.0e-04 loss: 5.515e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2397/30000 [5:24:28<62:18:01,  8.13s/it, lr: 1.0e-04 loss: 5.515e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2398/30000 [5:24:28<65:01:34,  8.48s/it, lr: 1.0e-04 loss: 5.515e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2398/30000 [5:24:28<65:01:34,  8.48s/it, lr: 1.0e-04 loss: 5.515e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2398/30000 [5:24:34<65:01:34,  8.48s/it, lr: 1.0e-04 loss: 6.202e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2398/30000 [5:24:34<65:01:34,  8.48s/it, lr: 1.0e-04 loss: 6.202e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2399/30000 [5:24:34<60:33:22,  7.90s/it, lr: 1.0e-04 loss: 6.202e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2399/30000 [5:24:34<60:33:22,  7.90s/it, lr: 1.0e-04 loss: 6.202e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2399/30000 [5:24:44<60:33:22,  7.90s/it, lr: 1.0e-04 loss: 7.689e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2399/30000 [5:24:44<60:33:22,  7.90s/it, lr: 1.0e-04 loss: 7.689e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2400/30000 [5:24:50<60:33:14,  7.90s/it, lr: 1.0e-04 loss: 7.840e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2400/30000 [5:24:50<60:33:14,  7.90s/it, lr: 1.0e-04 loss: 7.840e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2401/30000 [5:24:50<44:19:39,  5.78s/it, lr: 1.0e-04 loss: 7.840e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2401/30000 [5:24:50<44:19:39,  5.78s/it, lr: 1.0e-04 loss: 7.840e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2401/30000 [5:25:00<44:19:39,  5.78s/it, lr: 1.0e-04 loss: 6.872e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2401/30000 [5:25:00<44:19:39,  5.78s/it, lr: 1.0e-04 loss: 6.872e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2402/30000 [5:25:00<51:04:09,  6.66s/it, lr: 1.0e-04 loss: 6.872e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2402/30000 [5:25:00<51:04:09,  6.66s/it, lr: 1.0e-04 loss: 6.872e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2402/30000 [5:25:09<51:04:09,  6.66s/it, lr: 1.0e-04 loss: 5.260e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2402/30000 [5:25:09<51:04:09,  6.66s/it, lr: 1.0e-04 loss: 5.260e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2403/30000 [5:25:09<56:14:43,  7.34s/it, lr: 1.0e-04 loss: 5.260e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2403/30000 [5:25:09<56:14:43,  7.34s/it, lr: 1.0e-04 loss: 5.260e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2403/30000 [5:25:15<56:14:43,  7.34s/it, lr: 1.0e-04 loss: 7.831e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2403/30000 [5:25:15<56:14:43,  7.34s/it, lr: 1.0e-04 loss: 7.831e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2404/30000 [5:25:15<54:37:55,  7.13s/it, lr: 1.0e-04 loss: 7.831e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2404/30000 [5:25:15<54:37:55,  7.13s/it, lr: 1.0e-04 loss: 7.831e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2404/30000 [5:25:25<54:37:55,  7.13s/it, lr: 1.0e-04 loss: 6.891e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2404/30000 [5:25:25<54:37:55,  7.13s/it, lr: 1.0e-04 loss: 6.891e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2405/30000 [5:25:25<59:07:32,  7.71s/it, lr: 1.0e-04 loss: 6.891e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2405/30000 [5:25:25<59:07:32,  7.71s/it, lr: 1.0e-04 loss: 6.891e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2405/30000 [5:25:34<59:07:32,  7.71s/it, lr: 1.0e-04 loss: 9.543e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2405/30000 [5:25:34<59:07:32,  7.71s/it, lr: 1.0e-04 loss: 9.543e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2406/30000 [5:25:34<62:46:06,  8.19s/it, lr: 1.0e-04 loss: 9.543e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2406/30000 [5:25:34<62:46:06,  8.19s/it, lr: 1.0e-04 loss: 9.543e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2406/30000 [5:25:44<62:46:06,  8.19s/it, lr: 1.0e-04 loss: 7.059e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2406/30000 [5:25:44<62:46:06,  8.19s/it, lr: 1.0e-04 loss: 7.059e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2407/30000 [5:25:44<65:42:13,  8.57s/it, lr: 1.0e-04 loss: 7.059e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2407/30000 [5:25:44<65:42:13,  8.57s/it, lr: 1.0e-04 loss: 7.059e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2407/30000 [5:25:50<65:42:13,  8.57s/it, lr: 1.0e-04 loss: 8.673e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2407/30000 [5:25:50<65:42:13,  8.57s/it, lr: 1.0e-04 loss: 8.673e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2408/30000 [5:25:50<61:18:06,  8.00s/it, lr: 1.0e-04 loss: 8.673e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2408/30000 [5:25:50<61:18:06,  8.00s/it, lr: 1.0e-04 loss: 8.673e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2408/30000 [5:25:59<61:18:06,  8.00s/it, lr: 1.0e-04 loss: 6.518e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2408/30000 [5:25:59<61:18:06,  8.00s/it, lr: 1.0e-04 loss: 6.518e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2409/30000 [5:25:59<64:08:34,  8.37s/it, lr: 1.0e-04 loss: 6.518e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2409/30000 [5:25:59<64:08:34,  8.37s/it, lr: 1.0e-04 loss: 6.518e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2409/30000 [5:26:09<64:08:34,  8.37s/it, lr: 1.0e-04 loss: 7.671e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2409/30000 [5:26:09<64:08:34,  8.37s/it, lr: 1.0e-04 loss: 7.671e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2410/30000 [5:26:18<64:08:26,  8.37s/it, lr: 1.0e-04 loss: 3.805e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2410/30000 [5:26:18<64:08:26,  8.37s/it, lr: 1.0e-04 loss: 3.805e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2411/30000 [5:26:18<51:07:43,  6.67s/it, lr: 1.0e-04 loss: 3.805e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2411/30000 [5:26:18<51:07:43,  6.67s/it, lr: 1.0e-04 loss: 3.805e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2411/30000 [5:26:25<51:07:43,  6.67s/it, lr: 1.0e-04 loss: 6.468e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2411/30000 [5:26:25<51:07:43,  6.67s/it, lr: 1.0e-04 loss: 6.468e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2412/30000 [5:26:25<50:56:59,  6.65s/it, lr: 1.0e-04 loss: 6.468e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2412/30000 [5:26:25<50:56:59,  6.65s/it, lr: 1.0e-04 loss: 6.468e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2412/30000 [5:26:34<50:56:59,  6.65s/it, lr: 1.0e-04 loss: 7.518e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2412/30000 [5:26:34<50:56:59,  6.65s/it, lr: 1.0e-04 loss: 7.518e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2413/30000 [5:26:34<56:09:24,  7.33s/it, lr: 1.0e-04 loss: 7.518e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2413/30000 [5:26:34<56:09:24,  7.33s/it, lr: 1.0e-04 loss: 7.518e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2413/30000 [5:26:41<56:09:24,  7.33s/it, lr: 1.0e-04 loss: 7.563e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2413/30000 [5:26:41<56:09:24,  7.33s/it, lr: 1.0e-04 loss: 7.563e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2414/30000 [5:26:41<54:39:19,  7.13s/it, lr: 1.0e-04 loss: 7.563e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2414/30000 [5:26:41<54:39:19,  7.13s/it, lr: 1.0e-04 loss: 7.563e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2414/30000 [5:26:50<54:39:19,  7.13s/it, lr: 1.0e-04 loss: 4.000e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2414/30000 [5:26:50<54:39:19,  7.13s/it, lr: 1.0e-04 loss: 4.000e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2415/30000 [5:26:50<59:34:20,  7.77s/it, lr: 1.0e-04 loss: 4.000e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2415/30000 [5:26:50<59:34:20,  7.77s/it, lr: 1.0e-04 loss: 4.000e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2415/30000 [5:26:57<59:34:20,  7.77s/it, lr: 1.0e-04 loss: 1.040e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2415/30000 [5:26:57<59:34:20,  7.77s/it, lr: 1.0e-04 loss: 1.040e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2416/30000 [5:26:57<57:01:05,  7.44s/it, lr: 1.0e-04 loss: 1.040e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2416/30000 [5:26:57<57:01:05,  7.44s/it, lr: 1.0e-04 loss: 1.040e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2416/30000 [5:27:03<57:01:05,  7.44s/it, lr: 1.0e-04 loss: 1.750e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2416/30000 [5:27:03<57:01:05,  7.44s/it, lr: 1.0e-04 loss: 1.750e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2417/30000 [5:27:03<54:56:35,  7.17s/it, lr: 1.0e-04 loss: 1.750e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2417/30000 [5:27:03<54:56:35,  7.17s/it, lr: 1.0e-04 loss: 1.750e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2417/30000 [5:27:12<54:56:35,  7.17s/it, lr: 1.0e-04 loss: 9.716e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2417/30000 [5:27:12<54:56:35,  7.17s/it, lr: 1.0e-04 loss: 9.716e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2418/30000 [5:27:12<59:43:13,  7.79s/it, lr: 1.0e-04 loss: 9.716e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2418/30000 [5:27:12<59:43:13,  7.79s/it, lr: 1.0e-04 loss: 9.716e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2418/30000 [5:27:22<59:43:13,  7.79s/it, lr: 1.0e-04 loss: 9.349e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2418/30000 [5:27:22<59:43:13,  7.79s/it, lr: 1.0e-04 loss: 9.349e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2419/30000 [5:27:22<63:25:02,  8.28s/it, lr: 1.0e-04 loss: 9.349e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2419/30000 [5:27:22<63:25:02,  8.28s/it, lr: 1.0e-04 loss: 9.349e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2419/30000 [5:27:31<63:25:02,  8.28s/it, lr: 1.0e-04 loss: 6.311e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2419/30000 [5:27:31<63:25:02,  8.28s/it, lr: 1.0e-04 loss: 6.311e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2420/30000 [5:27:38<63:24:54,  8.28s/it, lr: 1.0e-04 loss: 5.809e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2420/30000 [5:27:38<63:24:54,  8.28s/it, lr: 1.0e-04 loss: 5.809e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2421/30000 [5:27:38<45:58:10,  6.00s/it, lr: 1.0e-04 loss: 5.809e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2421/30000 [5:27:38<45:58:10,  6.00s/it, lr: 1.0e-04 loss: 5.809e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2421/30000 [5:27:47<45:58:10,  6.00s/it, lr: 1.0e-04 loss: 7.355e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2421/30000 [5:27:47<45:58:10,  6.00s/it, lr: 1.0e-04 loss: 7.355e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2422/30000 [5:27:47<52:36:41,  6.87s/it, lr: 1.0e-04 loss: 7.355e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2422/30000 [5:27:47<52:36:41,  6.87s/it, lr: 1.0e-04 loss: 7.355e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2422/30000 [5:27:57<52:36:41,  6.87s/it, lr: 1.0e-04 loss: 5.483e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2422/30000 [5:27:57<52:36:41,  6.87s/it, lr: 1.0e-04 loss: 5.483e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2423/30000 [5:27:57<57:31:25,  7.51s/it, lr: 1.0e-04 loss: 5.483e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2423/30000 [5:27:57<57:31:25,  7.51s/it, lr: 1.0e-04 loss: 5.483e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2423/30000 [5:28:03<57:31:25,  7.51s/it, lr: 1.0e-04 loss: 9.662e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2423/30000 [5:28:03<57:31:25,  7.51s/it, lr: 1.0e-04 loss: 9.662e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2424/30000 [5:28:03<55:39:04,  7.27s/it, lr: 1.0e-04 loss: 9.662e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2424/30000 [5:28:03<55:39:04,  7.27s/it, lr: 1.0e-04 loss: 9.662e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2424/30000 [5:28:13<55:39:04,  7.27s/it, lr: 1.0e-04 loss: 4.064e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2424/30000 [5:28:13<55:39:04,  7.27s/it, lr: 1.0e-04 loss: 4.064e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2425/30000 [5:28:13<60:01:13,  7.84s/it, lr: 1.0e-04 loss: 4.064e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2425/30000 [5:28:13<60:01:13,  7.84s/it, lr: 1.0e-04 loss: 4.064e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2425/30000 [5:28:22<60:01:13,  7.84s/it, lr: 1.0e-04 loss: 5.688e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2425/30000 [5:28:22<60:01:13,  7.84s/it, lr: 1.0e-04 loss: 5.688e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2426/30000 [5:28:22<63:19:24,  8.27s/it, lr: 1.0e-04 loss: 5.688e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2426/30000 [5:28:22<63:19:24,  8.27s/it, lr: 1.0e-04 loss: 5.688e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2426/30000 [5:28:31<63:19:24,  8.27s/it, lr: 1.0e-04 loss: 4.157e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2426/30000 [5:28:31<63:19:24,  8.27s/it, lr: 1.0e-04 loss: 4.157e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2427/30000 [5:28:31<65:41:25,  8.58s/it, lr: 1.0e-04 loss: 4.157e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2427/30000 [5:28:31<65:41:25,  8.58s/it, lr: 1.0e-04 loss: 4.157e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2427/30000 [5:28:41<65:41:25,  8.58s/it, lr: 1.0e-04 loss: 9.265e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2427/30000 [5:28:41<65:41:25,  8.58s/it, lr: 1.0e-04 loss: 9.265e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2428/30000 [5:28:41<67:14:43,  8.78s/it, lr: 1.0e-04 loss: 9.265e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2428/30000 [5:28:41<67:14:43,  8.78s/it, lr: 1.0e-04 loss: 9.265e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2428/30000 [5:28:50<67:14:43,  8.78s/it, lr: 1.0e-04 loss: 8.253e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2428/30000 [5:28:50<67:14:43,  8.78s/it, lr: 1.0e-04 loss: 8.253e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2429/30000 [5:28:50<68:36:49,  8.96s/it, lr: 1.0e-04 loss: 8.253e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2429/30000 [5:28:50<68:36:49,  8.96s/it, lr: 1.0e-04 loss: 8.253e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2429/30000 [5:28:57<68:36:49,  8.96s/it, lr: 1.0e-04 loss: 9.157e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2429/30000 [5:28:57<68:36:49,  8.96s/it, lr: 1.0e-04 loss: 9.157e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2430/30000 [5:29:06<68:36:40,  8.96s/it, lr: 1.0e-04 loss: 7.352e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2430/30000 [5:29:06<68:36:40,  8.96s/it, lr: 1.0e-04 loss: 7.352e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2431/30000 [5:29:06<53:34:38,  7.00s/it, lr: 1.0e-04 loss: 7.352e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2431/30000 [5:29:06<53:34:38,  7.00s/it, lr: 1.0e-04 loss: 7.352e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2431/30000 [5:29:13<53:34:38,  7.00s/it, lr: 1.0e-04 loss: 9.538e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2431/30000 [5:29:13<53:34:38,  7.00s/it, lr: 1.0e-04 loss: 9.538e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2432/30000 [5:29:13<52:46:20,  6.89s/it, lr: 1.0e-04 loss: 9.538e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2432/30000 [5:29:13<52:46:20,  6.89s/it, lr: 1.0e-04 loss: 9.538e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2432/30000 [5:29:19<52:46:20,  6.89s/it, lr: 1.0e-04 loss: 5.437e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2432/30000 [5:29:19<52:46:20,  6.89s/it, lr: 1.0e-04 loss: 5.437e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2433/30000 [5:29:19<52:10:39,  6.81s/it, lr: 1.0e-04 loss: 5.437e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2433/30000 [5:29:19<52:10:39,  6.81s/it, lr: 1.0e-04 loss: 5.437e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2433/30000 [5:29:29<52:10:39,  6.81s/it, lr: 1.0e-04 loss: 5.085e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2433/30000 [5:29:29<52:10:39,  6.81s/it, lr: 1.0e-04 loss: 5.085e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2434/30000 [5:29:29<57:28:47,  7.51s/it, lr: 1.0e-04 loss: 5.085e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2434/30000 [5:29:29<57:28:47,  7.51s/it, lr: 1.0e-04 loss: 5.085e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2434/30000 [5:29:35<57:28:47,  7.51s/it, lr: 1.0e-04 loss: 8.311e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2434/30000 [5:29:35<57:28:47,  7.51s/it, lr: 1.0e-04 loss: 8.311e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2435/30000 [5:29:35<55:30:32,  7.25s/it, lr: 1.0e-04 loss: 8.311e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2435/30000 [5:29:35<55:30:32,  7.25s/it, lr: 1.0e-04 loss: 8.311e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2435/30000 [5:29:45<55:30:32,  7.25s/it, lr: 1.0e-04 loss: 6.555e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2435/30000 [5:29:45<55:30:32,  7.25s/it, lr: 1.0e-04 loss: 6.555e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2436/30000 [5:29:45<59:56:25,  7.83s/it, lr: 1.0e-04 loss: 6.555e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2436/30000 [5:29:45<59:56:25,  7.83s/it, lr: 1.0e-04 loss: 6.555e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2436/30000 [5:29:51<59:56:25,  7.83s/it, lr: 1.0e-04 loss: 1.330e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2436/30000 [5:29:51<59:56:25,  7.83s/it, lr: 1.0e-04 loss: 1.330e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2437/30000 [5:29:51<57:54:53,  7.56s/it, lr: 1.0e-04 loss: 1.330e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2437/30000 [5:29:51<57:54:53,  7.56s/it, lr: 1.0e-04 loss: 1.330e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2437/30000 [5:30:01<57:54:53,  7.56s/it, lr: 1.0e-04 loss: 9.511e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2437/30000 [5:30:01<57:54:53,  7.56s/it, lr: 1.0e-04 loss: 9.511e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2438/30000 [5:30:01<61:46:41,  8.07s/it, lr: 1.0e-04 loss: 9.511e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2438/30000 [5:30:01<61:46:41,  8.07s/it, lr: 1.0e-04 loss: 9.511e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2438/30000 [5:30:07<61:46:41,  8.07s/it, lr: 1.0e-04 loss: 5.867e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2438/30000 [5:30:07<61:46:41,  8.07s/it, lr: 1.0e-04 loss: 5.867e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2439/30000 [5:30:07<58:23:57,  7.63s/it, lr: 1.0e-04 loss: 5.867e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2439/30000 [5:30:07<58:23:57,  7.63s/it, lr: 1.0e-04 loss: 5.867e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2439/30000 [5:30:17<58:23:57,  7.63s/it, lr: 1.0e-04 loss: 1.058e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2439/30000 [5:30:17<58:23:57,  7.63s/it, lr: 1.0e-04 loss: 1.058e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2440/30000 [5:30:26<58:23:49,  7.63s/it, lr: 1.0e-04 loss: 4.443e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2440/30000 [5:30:26<58:23:49,  7.63s/it, lr: 1.0e-04 loss: 4.443e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2441/30000 [5:30:26<48:02:30,  6.28s/it, lr: 1.0e-04 loss: 4.443e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2441/30000 [5:30:26<48:02:30,  6.28s/it, lr: 1.0e-04 loss: 4.443e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2441/30000 [5:30:35<48:02:30,  6.28s/it, lr: 1.0e-04 loss: 6.326e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2441/30000 [5:30:35<48:02:30,  6.28s/it, lr: 1.0e-04 loss: 6.326e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2442/30000 [5:30:35<53:49:00,  7.03s/it, lr: 1.0e-04 loss: 6.326e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2442/30000 [5:30:35<53:49:00,  7.03s/it, lr: 1.0e-04 loss: 6.326e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2442/30000 [5:30:45<53:49:00,  7.03s/it, lr: 1.0e-04 loss: 5.866e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2442/30000 [5:30:45<53:49:00,  7.03s/it, lr: 1.0e-04 loss: 5.866e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2443/30000 [5:30:45<58:32:02,  7.65s/it, lr: 1.0e-04 loss: 5.866e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2443/30000 [5:30:45<58:32:02,  7.65s/it, lr: 1.0e-04 loss: 5.866e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2443/30000 [5:30:51<58:32:02,  7.65s/it, lr: 1.0e-04 loss: 1.201e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2443/30000 [5:30:51<58:32:02,  7.65s/it, lr: 1.0e-04 loss: 1.201e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2444/30000 [5:30:51<56:22:06,  7.36s/it, lr: 1.0e-04 loss: 1.201e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2444/30000 [5:30:51<56:22:06,  7.36s/it, lr: 1.0e-04 loss: 1.201e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2444/30000 [5:30:58<56:22:06,  7.36s/it, lr: 1.0e-04 loss: 5.447e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2444/30000 [5:30:58<56:22:06,  7.36s/it, lr: 1.0e-04 loss: 5.447e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2445/30000 [5:30:58<55:09:14,  7.21s/it, lr: 1.0e-04 loss: 5.447e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2445/30000 [5:30:58<55:09:14,  7.21s/it, lr: 1.0e-04 loss: 5.447e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2445/30000 [5:31:05<55:09:14,  7.21s/it, lr: 1.0e-04 loss: 9.072e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2445/30000 [5:31:05<55:09:14,  7.21s/it, lr: 1.0e-04 loss: 9.072e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2446/30000 [5:31:05<53:38:11,  7.01s/it, lr: 1.0e-04 loss: 9.072e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2446/30000 [5:31:05<53:38:11,  7.01s/it, lr: 1.0e-04 loss: 9.072e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2446/30000 [5:31:14<53:38:11,  7.01s/it, lr: 1.0e-04 loss: 8.990e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2446/30000 [5:31:14<53:38:11,  7.01s/it, lr: 1.0e-04 loss: 8.990e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2447/30000 [5:31:14<58:42:47,  7.67s/it, lr: 1.0e-04 loss: 8.990e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2447/30000 [5:31:14<58:42:47,  7.67s/it, lr: 1.0e-04 loss: 8.990e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2447/30000 [5:31:20<58:42:47,  7.67s/it, lr: 1.0e-04 loss: 8.578e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2447/30000 [5:31:20<58:42:47,  7.67s/it, lr: 1.0e-04 loss: 8.578e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2448/30000 [5:31:20<56:11:19,  7.34s/it, lr: 1.0e-04 loss: 8.578e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2448/30000 [5:31:20<56:11:19,  7.34s/it, lr: 1.0e-04 loss: 8.578e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2448/30000 [5:31:27<56:11:19,  7.34s/it, lr: 1.0e-04 loss: 1.463e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2448/30000 [5:31:27<56:11:19,  7.34s/it, lr: 1.0e-04 loss: 1.463e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2449/30000 [5:31:27<54:14:54,  7.09s/it, lr: 1.0e-04 loss: 1.463e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2449/30000 [5:31:27<54:14:54,  7.09s/it, lr: 1.0e-04 loss: 1.463e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2449/30000 [5:31:36<54:14:54,  7.09s/it, lr: 1.0e-04 loss: 7.852e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2449/30000 [5:31:36<54:14:54,  7.09s/it, lr: 1.0e-04 loss: 7.852e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2450/30000 [5:31:43<54:14:47,  7.09s/it, lr: 1.0e-04 loss: 1.089e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2450/30000 [5:31:43<54:14:47,  7.09s/it, lr: 1.0e-04 loss: 1.089e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2451/30000 [5:31:43<40:50:56,  5.34s/it, lr: 1.0e-04 loss: 1.089e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2451/30000 [5:31:43<40:50:56,  5.34s/it, lr: 1.0e-04 loss: 1.089e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2451/30000 [5:31:52<40:50:56,  5.34s/it, lr: 1.0e-04 loss: 8.515e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2451/30000 [5:31:52<40:50:56,  5.34s/it, lr: 1.0e-04 loss: 8.515e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2452/30000 [5:31:52<48:39:23,  6.36s/it, lr: 1.0e-04 loss: 8.515e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2452/30000 [5:31:52<48:39:23,  6.36s/it, lr: 1.0e-04 loss: 8.515e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2452/30000 [5:31:59<48:39:23,  6.36s/it, lr: 1.0e-04 loss: 1.082e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2452/30000 [5:31:59<48:39:23,  6.36s/it, lr: 1.0e-04 loss: 1.082e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2453/30000 [5:31:59<49:03:37,  6.41s/it, lr: 1.0e-04 loss: 1.082e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2453/30000 [5:31:59<49:03:37,  6.41s/it, lr: 1.0e-04 loss: 1.082e-01]
Xiang_wan_2_2_lora:   8%|###1                                   | 2453/30000 [5:32:05<49:03:37,  6.41s/it, lr: 1.0e-04 loss: 9.704e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2453/30000 [5:32:05<49:03:37,  6.41s/it, lr: 1.0e-04 loss: 9.704e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2454/30000 [5:32:05<49:23:06,  6.45s/it, lr: 1.0e-04 loss: 9.704e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2454/30000 [5:32:05<49:23:06,  6.45s/it, lr: 1.0e-04 loss: 9.704e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2454/30000 [5:32:12<49:23:06,  6.45s/it, lr: 1.0e-04 loss: 2.947e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2454/30000 [5:32:12<49:23:06,  6.45s/it, lr: 1.0e-04 loss: 2.947e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2455/30000 [5:32:12<49:36:57,  6.48s/it, lr: 1.0e-04 loss: 2.947e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2455/30000 [5:32:12<49:36:57,  6.48s/it, lr: 1.0e-04 loss: 2.947e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2455/30000 [5:32:21<49:36:57,  6.48s/it, lr: 1.0e-04 loss: 7.272e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2455/30000 [5:32:21<49:36:57,  6.48s/it, lr: 1.0e-04 loss: 7.272e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2456/30000 [5:32:21<55:45:03,  7.29s/it, lr: 1.0e-04 loss: 7.272e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2456/30000 [5:32:21<55:45:03,  7.29s/it, lr: 1.0e-04 loss: 7.272e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2456/30000 [5:32:31<55:45:03,  7.29s/it, lr: 1.0e-04 loss: 5.506e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2456/30000 [5:32:31<55:45:03,  7.29s/it, lr: 1.0e-04 loss: 5.506e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2457/30000 [5:32:31<60:12:36,  7.87s/it, lr: 1.0e-04 loss: 5.506e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2457/30000 [5:32:31<60:12:36,  7.87s/it, lr: 1.0e-04 loss: 5.506e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2457/30000 [5:32:40<60:12:36,  7.87s/it, lr: 1.0e-04 loss: 6.094e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2457/30000 [5:32:40<60:12:36,  7.87s/it, lr: 1.0e-04 loss: 6.094e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2458/30000 [5:32:40<63:35:11,  8.31s/it, lr: 1.0e-04 loss: 6.094e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2458/30000 [5:32:40<63:35:11,  8.31s/it, lr: 1.0e-04 loss: 6.094e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2458/30000 [5:32:49<63:35:11,  8.31s/it, lr: 1.0e-04 loss: 9.924e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2458/30000 [5:32:49<63:35:11,  8.31s/it, lr: 1.0e-04 loss: 9.924e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2459/30000 [5:32:49<65:53:35,  8.61s/it, lr: 1.0e-04 loss: 9.924e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2459/30000 [5:32:49<65:53:35,  8.61s/it, lr: 1.0e-04 loss: 9.924e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2459/30000 [5:32:59<65:53:35,  8.61s/it, lr: 1.0e-04 loss: 8.328e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2459/30000 [5:32:59<65:53:35,  8.61s/it, lr: 1.0e-04 loss: 8.328e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2460/30000 [5:33:08<65:53:27,  8.61s/it, lr: 1.0e-04 loss: 4.241e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2460/30000 [5:33:08<65:53:27,  8.61s/it, lr: 1.0e-04 loss: 4.241e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2461/30000 [5:33:08<52:01:51,  6.80s/it, lr: 1.0e-04 loss: 4.241e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2461/30000 [5:33:08<52:01:51,  6.80s/it, lr: 1.0e-04 loss: 4.241e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2461/30000 [5:33:15<52:01:51,  6.80s/it, lr: 1.0e-04 loss: 8.460e-02]
Xiang_wan_2_2_lora:   8%|###1                                   | 2461/30000 [5:33:15<52:01:51,  6.80s/it, lr: 1.0e-04 loss: 8.460e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2462/30000 [5:33:15<51:33:12,  6.74s/it, lr: 1.0e-04 loss: 8.460e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2462/30000 [5:33:15<51:33:12,  6.74s/it, lr: 1.0e-04 loss: 8.460e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2462/30000 [5:33:21<51:33:12,  6.74s/it, lr: 1.0e-04 loss: 4.835e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2462/30000 [5:33:21<51:33:12,  6.74s/it, lr: 1.0e-04 loss: 4.835e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2463/30000 [5:33:21<51:14:06,  6.70s/it, lr: 1.0e-04 loss: 4.835e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2463/30000 [5:33:21<51:14:06,  6.70s/it, lr: 1.0e-04 loss: 4.835e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2463/30000 [5:33:31<51:14:06,  6.70s/it, lr: 1.0e-04 loss: 4.602e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2463/30000 [5:33:31<51:14:06,  6.70s/it, lr: 1.0e-04 loss: 4.602e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2464/30000 [5:33:31<56:38:09,  7.40s/it, lr: 1.0e-04 loss: 4.602e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2464/30000 [5:33:31<56:38:09,  7.40s/it, lr: 1.0e-04 loss: 4.602e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2464/30000 [5:33:40<56:38:09,  7.40s/it, lr: 1.0e-04 loss: 7.118e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2464/30000 [5:33:40<56:38:09,  7.40s/it, lr: 1.0e-04 loss: 7.118e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2465/30000 [5:33:40<60:45:26,  7.94s/it, lr: 1.0e-04 loss: 7.118e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2465/30000 [5:33:40<60:45:26,  7.94s/it, lr: 1.0e-04 loss: 7.118e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2465/30000 [5:33:46<60:45:26,  7.94s/it, lr: 1.0e-04 loss: 8.973e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2465/30000 [5:33:46<60:45:26,  7.94s/it, lr: 1.0e-04 loss: 8.973e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2466/30000 [5:33:46<57:48:55,  7.56s/it, lr: 1.0e-04 loss: 8.973e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2466/30000 [5:33:46<57:48:55,  7.56s/it, lr: 1.0e-04 loss: 8.973e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2466/30000 [5:33:53<57:48:55,  7.56s/it, lr: 1.0e-04 loss: 8.074e-01]
Xiang_wan_2_2_lora:   8%|###2                                   | 2466/30000 [5:33:53<57:48:55,  7.56s/it, lr: 1.0e-04 loss: 8.074e-01]
Xiang_wan_2_2_lora:   8%|###2                                   | 2467/30000 [5:33:53<56:08:05,  7.34s/it, lr: 1.0e-04 loss: 8.074e-01]
Xiang_wan_2_2_lora:   8%|###2                                   | 2467/30000 [5:33:53<56:08:05,  7.34s/it, lr: 1.0e-04 loss: 8.074e-01]
Xiang_wan_2_2_lora:   8%|###2                                   | 2467/30000 [5:34:00<56:08:05,  7.34s/it, lr: 1.0e-04 loss: 7.954e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2467/30000 [5:34:00<56:08:05,  7.34s/it, lr: 1.0e-04 loss: 7.954e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2468/30000 [5:34:00<54:27:07,  7.12s/it, lr: 1.0e-04 loss: 7.954e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2468/30000 [5:34:00<54:27:07,  7.12s/it, lr: 1.0e-04 loss: 7.954e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2468/30000 [5:34:06<54:27:07,  7.12s/it, lr: 1.0e-04 loss: 7.142e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2468/30000 [5:34:06<54:27:07,  7.12s/it, lr: 1.0e-04 loss: 7.142e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2469/30000 [5:34:06<53:01:03,  6.93s/it, lr: 1.0e-04 loss: 7.142e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2469/30000 [5:34:06<53:01:03,  6.93s/it, lr: 1.0e-04 loss: 7.142e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2469/30000 [5:34:16<53:01:03,  6.93s/it, lr: 1.0e-04 loss: 6.592e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2469/30000 [5:34:16<53:01:03,  6.93s/it, lr: 1.0e-04 loss: 6.592e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2470/30000 [5:34:22<53:00:57,  6.93s/it, lr: 1.0e-04 loss: 8.030e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2470/30000 [5:34:22<53:00:57,  6.93s/it, lr: 1.0e-04 loss: 8.030e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2471/30000 [5:34:22<40:10:58,  5.25s/it, lr: 1.0e-04 loss: 8.030e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2471/30000 [5:34:22<40:10:58,  5.25s/it, lr: 1.0e-04 loss: 8.030e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2471/30000 [5:34:32<40:10:58,  5.25s/it, lr: 1.0e-04 loss: 6.375e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2471/30000 [5:34:32<40:10:58,  5.25s/it, lr: 1.0e-04 loss: 6.375e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2472/30000 [5:34:32<47:49:47,  6.25s/it, lr: 1.0e-04 loss: 6.375e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2472/30000 [5:34:32<47:49:47,  6.25s/it, lr: 1.0e-04 loss: 6.375e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2472/30000 [5:34:41<47:49:47,  6.25s/it, lr: 1.0e-04 loss: 4.027e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2472/30000 [5:34:41<47:49:47,  6.25s/it, lr: 1.0e-04 loss: 4.027e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2473/30000 [5:34:41<53:49:17,  7.04s/it, lr: 1.0e-04 loss: 4.027e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2473/30000 [5:34:41<53:49:17,  7.04s/it, lr: 1.0e-04 loss: 4.027e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2473/30000 [5:34:47<53:49:17,  7.04s/it, lr: 1.0e-04 loss: 1.083e-01]
Xiang_wan_2_2_lora:   8%|###2                                   | 2473/30000 [5:34:47<53:49:17,  7.04s/it, lr: 1.0e-04 loss: 1.083e-01]
Xiang_wan_2_2_lora:   8%|###2                                   | 2474/30000 [5:34:47<52:49:02,  6.91s/it, lr: 1.0e-04 loss: 1.083e-01]
Xiang_wan_2_2_lora:   8%|###2                                   | 2474/30000 [5:34:47<52:49:02,  6.91s/it, lr: 1.0e-04 loss: 1.083e-01]
Xiang_wan_2_2_lora:   8%|###2                                   | 2474/30000 [5:34:54<52:49:02,  6.91s/it, lr: 1.0e-04 loss: 6.372e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2474/30000 [5:34:54<52:49:02,  6.91s/it, lr: 1.0e-04 loss: 6.372e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2475/30000 [5:34:54<52:42:38,  6.89s/it, lr: 1.0e-04 loss: 6.372e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2475/30000 [5:34:54<52:42:38,  6.89s/it, lr: 1.0e-04 loss: 6.372e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2475/30000 [5:35:04<52:42:38,  6.89s/it, lr: 1.0e-04 loss: 7.272e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2475/30000 [5:35:04<52:42:38,  6.89s/it, lr: 1.0e-04 loss: 7.272e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2476/30000 [5:35:04<57:57:12,  7.58s/it, lr: 1.0e-04 loss: 7.272e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2476/30000 [5:35:04<57:57:12,  7.58s/it, lr: 1.0e-04 loss: 7.272e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2476/30000 [5:35:10<57:57:12,  7.58s/it, lr: 1.0e-04 loss: 5.947e-01]
Xiang_wan_2_2_lora:   8%|###2                                   | 2476/30000 [5:35:10<57:57:12,  7.58s/it, lr: 1.0e-04 loss: 5.947e-01]
Xiang_wan_2_2_lora:   8%|###2                                   | 2477/30000 [5:35:10<55:48:09,  7.30s/it, lr: 1.0e-04 loss: 5.947e-01]
Xiang_wan_2_2_lora:   8%|###2                                   | 2477/30000 [5:35:10<55:48:09,  7.30s/it, lr: 1.0e-04 loss: 5.947e-01]
Xiang_wan_2_2_lora:   8%|###2                                   | 2477/30000 [5:35:20<55:48:09,  7.30s/it, lr: 1.0e-04 loss: 7.467e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2477/30000 [5:35:20<55:48:09,  7.30s/it, lr: 1.0e-04 loss: 7.467e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2478/30000 [5:35:20<60:22:44,  7.90s/it, lr: 1.0e-04 loss: 7.467e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2478/30000 [5:35:20<60:22:44,  7.90s/it, lr: 1.0e-04 loss: 7.467e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2478/30000 [5:35:29<60:22:44,  7.90s/it, lr: 1.0e-04 loss: 4.109e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2478/30000 [5:35:29<60:22:44,  7.90s/it, lr: 1.0e-04 loss: 4.109e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2479/30000 [5:35:29<63:42:15,  8.33s/it, lr: 1.0e-04 loss: 4.109e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2479/30000 [5:35:29<63:42:15,  8.33s/it, lr: 1.0e-04 loss: 4.109e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2479/30000 [5:35:38<63:42:15,  8.33s/it, lr: 1.0e-04 loss: 1.236e-01]
Xiang_wan_2_2_lora:   8%|###2                                   | 2479/30000 [5:35:38<63:42:15,  8.33s/it, lr: 1.0e-04 loss: 1.236e-01]
Xiang_wan_2_2_lora:   8%|###2                                   | 2480/30000 [5:35:45<63:42:07,  8.33s/it, lr: 1.0e-04 loss: 9.133e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2480/30000 [5:35:45<63:42:07,  8.33s/it, lr: 1.0e-04 loss: 9.133e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2481/30000 [5:35:45<46:10:34,  6.04s/it, lr: 1.0e-04 loss: 9.133e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2481/30000 [5:35:45<46:10:34,  6.04s/it, lr: 1.0e-04 loss: 9.133e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2481/30000 [5:35:52<46:10:34,  6.04s/it, lr: 1.0e-04 loss: 7.784e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2481/30000 [5:35:52<46:10:34,  6.04s/it, lr: 1.0e-04 loss: 7.784e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2482/30000 [5:35:52<47:39:42,  6.24s/it, lr: 1.0e-04 loss: 7.784e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2482/30000 [5:35:52<47:39:42,  6.24s/it, lr: 1.0e-04 loss: 7.784e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2482/30000 [5:35:58<47:39:42,  6.24s/it, lr: 1.0e-04 loss: 6.232e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2482/30000 [5:35:58<47:39:42,  6.24s/it, lr: 1.0e-04 loss: 6.232e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2483/30000 [5:35:58<48:11:30,  6.30s/it, lr: 1.0e-04 loss: 6.232e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2483/30000 [5:35:58<48:11:30,  6.30s/it, lr: 1.0e-04 loss: 6.232e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2483/30000 [5:36:08<48:11:30,  6.30s/it, lr: 1.0e-04 loss: 7.169e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2483/30000 [5:36:08<48:11:30,  6.30s/it, lr: 1.0e-04 loss: 7.169e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2484/30000 [5:36:08<54:26:24,  7.12s/it, lr: 1.0e-04 loss: 7.169e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2484/30000 [5:36:08<54:26:24,  7.12s/it, lr: 1.0e-04 loss: 7.169e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2484/30000 [5:36:17<54:26:24,  7.12s/it, lr: 1.0e-04 loss: 3.407e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2484/30000 [5:36:17<54:26:24,  7.12s/it, lr: 1.0e-04 loss: 3.407e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2485/30000 [5:36:17<59:10:09,  7.74s/it, lr: 1.0e-04 loss: 3.407e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2485/30000 [5:36:17<59:10:09,  7.74s/it, lr: 1.0e-04 loss: 3.407e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2485/30000 [5:36:26<59:10:09,  7.74s/it, lr: 1.0e-04 loss: 7.469e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2485/30000 [5:36:26<59:10:09,  7.74s/it, lr: 1.0e-04 loss: 7.469e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2486/30000 [5:36:26<62:36:40,  8.19s/it, lr: 1.0e-04 loss: 7.469e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2486/30000 [5:36:26<62:36:40,  8.19s/it, lr: 1.0e-04 loss: 7.469e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2486/30000 [5:36:36<62:36:40,  8.19s/it, lr: 1.0e-04 loss: 1.073e-01]
Xiang_wan_2_2_lora:   8%|###2                                   | 2486/30000 [5:36:36<62:36:40,  8.19s/it, lr: 1.0e-04 loss: 1.073e-01]
Xiang_wan_2_2_lora:   8%|###2                                   | 2487/30000 [5:36:36<65:06:12,  8.52s/it, lr: 1.0e-04 loss: 1.073e-01]
Xiang_wan_2_2_lora:   8%|###2                                   | 2487/30000 [5:36:36<65:06:12,  8.52s/it, lr: 1.0e-04 loss: 1.073e-01]
Xiang_wan_2_2_lora:   8%|###2                                   | 2487/30000 [5:36:45<65:06:12,  8.52s/it, lr: 1.0e-04 loss: 6.248e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2487/30000 [5:36:45<65:06:12,  8.52s/it, lr: 1.0e-04 loss: 6.248e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2488/30000 [5:36:45<66:56:18,  8.76s/it, lr: 1.0e-04 loss: 6.248e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2488/30000 [5:36:45<66:56:18,  8.76s/it, lr: 1.0e-04 loss: 6.248e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2488/30000 [5:36:52<66:56:18,  8.76s/it, lr: 1.0e-04 loss: 1.539e-01]
Xiang_wan_2_2_lora:   8%|###2                                   | 2488/30000 [5:36:52<66:56:18,  8.76s/it, lr: 1.0e-04 loss: 1.539e-01]
Xiang_wan_2_2_lora:   8%|###2                                   | 2489/30000 [5:36:52<62:12:02,  8.14s/it, lr: 1.0e-04 loss: 1.539e-01]
Xiang_wan_2_2_lora:   8%|###2                                   | 2489/30000 [5:36:52<62:12:02,  8.14s/it, lr: 1.0e-04 loss: 1.539e-01]
Xiang_wan_2_2_lora:   8%|###2                                   | 2489/30000 [5:36:58<62:12:02,  8.14s/it, lr: 1.0e-04 loss: 5.916e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2489/30000 [5:36:58<62:12:02,  8.14s/it, lr: 1.0e-04 loss: 5.916e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2490/30000 [5:37:08<62:11:53,  8.14s/it, lr: 1.0e-04 loss: 8.669e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2490/30000 [5:37:08<62:11:53,  8.14s/it, lr: 1.0e-04 loss: 8.669e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2491/30000 [5:37:08<50:01:59,  6.55s/it, lr: 1.0e-04 loss: 8.669e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2491/30000 [5:37:08<50:01:59,  6.55s/it, lr: 1.0e-04 loss: 8.669e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2491/30000 [5:37:17<50:01:59,  6.55s/it, lr: 1.0e-04 loss: 6.866e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2491/30000 [5:37:17<50:01:59,  6.55s/it, lr: 1.0e-04 loss: 6.866e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2492/30000 [5:37:17<55:15:21,  7.23s/it, lr: 1.0e-04 loss: 6.866e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2492/30000 [5:37:17<55:15:21,  7.23s/it, lr: 1.0e-04 loss: 6.866e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2492/30000 [5:37:26<55:15:21,  7.23s/it, lr: 1.0e-04 loss: 3.968e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2492/30000 [5:37:26<55:15:21,  7.23s/it, lr: 1.0e-04 loss: 3.968e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2493/30000 [5:37:26<59:33:26,  7.79s/it, lr: 1.0e-04 loss: 3.968e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2493/30000 [5:37:26<59:33:26,  7.79s/it, lr: 1.0e-04 loss: 3.968e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2493/30000 [5:37:36<59:33:26,  7.79s/it, lr: 1.0e-04 loss: 7.974e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2493/30000 [5:37:36<59:33:26,  7.79s/it, lr: 1.0e-04 loss: 7.974e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2494/30000 [5:37:36<62:41:36,  8.21s/it, lr: 1.0e-04 loss: 7.974e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2494/30000 [5:37:36<62:41:36,  8.21s/it, lr: 1.0e-04 loss: 7.974e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2494/30000 [5:37:42<62:41:36,  8.21s/it, lr: 1.0e-04 loss: 3.003e-01]
Xiang_wan_2_2_lora:   8%|###2                                   | 2494/30000 [5:37:42<62:41:36,  8.21s/it, lr: 1.0e-04 loss: 3.003e-01]
Xiang_wan_2_2_lora:   8%|###2                                   | 2495/30000 [5:37:42<59:14:23,  7.75s/it, lr: 1.0e-04 loss: 3.003e-01]
Xiang_wan_2_2_lora:   8%|###2                                   | 2495/30000 [5:37:42<59:14:23,  7.75s/it, lr: 1.0e-04 loss: 3.003e-01]
Xiang_wan_2_2_lora:   8%|###2                                   | 2495/30000 [5:37:49<59:14:23,  7.75s/it, lr: 1.0e-04 loss: 8.896e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2495/30000 [5:37:49<59:14:23,  7.75s/it, lr: 1.0e-04 loss: 8.896e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2496/30000 [5:37:49<56:36:34,  7.41s/it, lr: 1.0e-04 loss: 8.896e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2496/30000 [5:37:49<56:36:34,  7.41s/it, lr: 1.0e-04 loss: 8.896e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2496/30000 [5:37:56<56:36:34,  7.41s/it, lr: 1.0e-04 loss: 1.021e-01]
Xiang_wan_2_2_lora:   8%|###2                                   | 2496/30000 [5:37:56<56:36:34,  7.41s/it, lr: 1.0e-04 loss: 1.021e-01]
Xiang_wan_2_2_lora:   8%|###2                                   | 2497/30000 [5:37:56<55:20:42,  7.24s/it, lr: 1.0e-04 loss: 1.021e-01]
Xiang_wan_2_2_lora:   8%|###2                                   | 2497/30000 [5:37:56<55:20:42,  7.24s/it, lr: 1.0e-04 loss: 1.021e-01]
Xiang_wan_2_2_lora:   8%|###2                                   | 2497/30000 [5:38:02<55:20:42,  7.24s/it, lr: 1.0e-04 loss: 8.720e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2497/30000 [5:38:02<55:20:42,  7.24s/it, lr: 1.0e-04 loss: 8.720e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2498/30000 [5:38:02<53:46:36,  7.04s/it, lr: 1.0e-04 loss: 8.720e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2498/30000 [5:38:02<53:46:36,  7.04s/it, lr: 1.0e-04 loss: 8.720e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2498/30000 [5:38:09<53:46:36,  7.04s/it, lr: 1.0e-04 loss: 7.704e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2498/30000 [5:38:09<53:46:36,  7.04s/it, lr: 1.0e-04 loss: 7.704e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2499/30000 [5:38:09<52:36:16,  6.89s/it, lr: 1.0e-04 loss: 7.704e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2499/30000 [5:38:09<52:36:16,  6.89s/it, lr: 1.0e-04 loss: 7.704e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2499/30000 [5:38:18<52:36:16,  6.89s/it, lr: 1.0e-04 loss: 5.174e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2499/30000 [5:38:18<52:36:16,  6.89s/it, lr: 1.0e-04 loss: 5.174e-02]
+Saving at step 2500
+Saved checkpoint to /home/featurize/ai-toolkit/output/Xiang_wan_2_2_lora/Xiang_wan_2_2_lora_000002500.safetensors
+Saved optimizer to /home/featurize/ai-toolkit/output/Xiang_wan_2_2_lora/optimizer.pt
+
Xiang_wan_2_2_lora:   8%|###2                                   | 2500/30000 [5:38:25<52:36:09,  6.89s/it, lr: 1.0e-04 loss: 6.874e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2500/30000 [5:38:25<52:36:09,  6.89s/it, lr: 1.0e-04 loss: 6.874e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2501/30000 [5:38:25<40:32:31,  5.31s/it, lr: 1.0e-04 loss: 6.874e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2501/30000 [5:38:25<40:32:31,  5.31s/it, lr: 1.0e-04 loss: 6.874e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2501/30000 [5:38:31<40:32:31,  5.31s/it, lr: 1.0e-04 loss: 8.758e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2501/30000 [5:38:31<40:32:31,  5.31s/it, lr: 1.0e-04 loss: 8.758e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2502/30000 [5:38:31<42:50:40,  5.61s/it, lr: 1.0e-04 loss: 8.758e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2502/30000 [5:38:31<42:50:40,  5.61s/it, lr: 1.0e-04 loss: 8.758e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2502/30000 [5:38:39<42:50:40,  5.61s/it, lr: 1.0e-04 loss: 7.007e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2502/30000 [5:38:39<42:50:40,  5.61s/it, lr: 1.0e-04 loss: 7.007e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2503/30000 [5:38:39<46:42:55,  6.12s/it, lr: 1.0e-04 loss: 7.007e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2503/30000 [5:38:39<46:42:55,  6.12s/it, lr: 1.0e-04 loss: 7.007e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2503/30000 [5:38:49<46:42:55,  6.12s/it, lr: 1.0e-04 loss: 9.396e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2503/30000 [5:38:49<46:42:55,  6.12s/it, lr: 1.0e-04 loss: 9.396e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2504/30000 [5:38:49<55:15:37,  7.24s/it, lr: 1.0e-04 loss: 9.396e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2504/30000 [5:38:49<55:15:37,  7.24s/it, lr: 1.0e-04 loss: 9.396e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2504/30000 [5:38:59<55:15:37,  7.24s/it, lr: 1.0e-04 loss: 5.368e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2504/30000 [5:38:59<55:15:37,  7.24s/it, lr: 1.0e-04 loss: 5.368e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2505/30000 [5:38:59<59:57:39,  7.85s/it, lr: 1.0e-04 loss: 5.368e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2505/30000 [5:38:59<59:57:39,  7.85s/it, lr: 1.0e-04 loss: 5.368e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2505/30000 [5:39:08<59:57:39,  7.85s/it, lr: 1.0e-04 loss: 7.041e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2505/30000 [5:39:08<59:57:39,  7.85s/it, lr: 1.0e-04 loss: 7.041e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2506/30000 [5:39:08<63:24:14,  8.30s/it, lr: 1.0e-04 loss: 7.041e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2506/30000 [5:39:08<63:24:14,  8.30s/it, lr: 1.0e-04 loss: 7.041e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2506/30000 [5:39:15<63:24:14,  8.30s/it, lr: 1.0e-04 loss: 7.855e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2506/30000 [5:39:15<63:24:14,  8.30s/it, lr: 1.0e-04 loss: 7.855e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2507/30000 [5:39:15<59:41:17,  7.82s/it, lr: 1.0e-04 loss: 7.855e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2507/30000 [5:39:15<59:41:17,  7.82s/it, lr: 1.0e-04 loss: 7.855e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2507/30000 [5:39:22<59:41:17,  7.82s/it, lr: 1.0e-04 loss: 8.412e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2507/30000 [5:39:22<59:41:17,  7.82s/it, lr: 1.0e-04 loss: 8.412e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2508/30000 [5:39:22<57:41:05,  7.55s/it, lr: 1.0e-04 loss: 8.412e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2508/30000 [5:39:22<57:41:05,  7.55s/it, lr: 1.0e-04 loss: 8.412e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2508/30000 [5:39:28<57:41:05,  7.55s/it, lr: 1.0e-04 loss: 5.883e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2508/30000 [5:39:28<57:41:05,  7.55s/it, lr: 1.0e-04 loss: 5.883e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2509/30000 [5:39:28<55:43:52,  7.30s/it, lr: 1.0e-04 loss: 5.883e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2509/30000 [5:39:28<55:43:52,  7.30s/it, lr: 1.0e-04 loss: 5.883e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2509/30000 [5:39:38<55:43:52,  7.30s/it, lr: 1.0e-04 loss: 1.084e-01]
Xiang_wan_2_2_lora:   8%|###2                                   | 2509/30000 [5:39:38<55:43:52,  7.30s/it, lr: 1.0e-04 loss: 1.084e-01]
Xiang_wan_2_2_lora:   8%|###2                                   | 2510/30000 [5:39:44<55:43:45,  7.30s/it, lr: 1.0e-04 loss: 8.835e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2510/30000 [5:39:44<55:43:45,  7.30s/it, lr: 1.0e-04 loss: 8.835e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2511/30000 [5:39:44<41:46:46,  5.47s/it, lr: 1.0e-04 loss: 8.835e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2511/30000 [5:39:44<41:46:46,  5.47s/it, lr: 1.0e-04 loss: 8.835e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2511/30000 [5:39:51<41:46:46,  5.47s/it, lr: 1.0e-04 loss: 5.713e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2511/30000 [5:39:51<41:46:46,  5.47s/it, lr: 1.0e-04 loss: 5.713e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2512/30000 [5:39:51<43:49:11,  5.74s/it, lr: 1.0e-04 loss: 5.713e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2512/30000 [5:39:51<43:49:11,  5.74s/it, lr: 1.0e-04 loss: 5.713e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2512/30000 [5:40:00<43:49:11,  5.74s/it, lr: 1.0e-04 loss: 5.245e-01]
Xiang_wan_2_2_lora:   8%|###2                                   | 2512/30000 [5:40:00<43:49:11,  5.74s/it, lr: 1.0e-04 loss: 5.245e-01]
Xiang_wan_2_2_lora:   8%|###2                                   | 2513/30000 [5:40:00<50:57:41,  6.67s/it, lr: 1.0e-04 loss: 5.245e-01]
Xiang_wan_2_2_lora:   8%|###2                                   | 2513/30000 [5:40:00<50:57:41,  6.67s/it, lr: 1.0e-04 loss: 5.245e-01]
Xiang_wan_2_2_lora:   8%|###2                                   | 2513/30000 [5:40:10<50:57:41,  6.67s/it, lr: 1.0e-04 loss: 7.388e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2513/30000 [5:40:10<50:57:41,  6.67s/it, lr: 1.0e-04 loss: 7.388e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2514/30000 [5:40:10<56:35:48,  7.41s/it, lr: 1.0e-04 loss: 7.388e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2514/30000 [5:40:10<56:35:48,  7.41s/it, lr: 1.0e-04 loss: 7.388e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2514/30000 [5:40:16<56:35:48,  7.41s/it, lr: 1.0e-04 loss: 9.887e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2514/30000 [5:40:16<56:35:48,  7.41s/it, lr: 1.0e-04 loss: 9.887e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2515/30000 [5:40:16<55:24:56,  7.26s/it, lr: 1.0e-04 loss: 9.887e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2515/30000 [5:40:16<55:24:56,  7.26s/it, lr: 1.0e-04 loss: 9.887e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2515/30000 [5:40:26<55:24:56,  7.26s/it, lr: 1.0e-04 loss: 7.749e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2515/30000 [5:40:26<55:24:56,  7.26s/it, lr: 1.0e-04 loss: 7.749e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2516/30000 [5:40:26<59:53:50,  7.85s/it, lr: 1.0e-04 loss: 7.749e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2516/30000 [5:40:26<59:53:50,  7.85s/it, lr: 1.0e-04 loss: 7.749e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2516/30000 [5:40:32<59:53:50,  7.85s/it, lr: 1.0e-04 loss: 7.403e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2516/30000 [5:40:32<59:53:50,  7.85s/it, lr: 1.0e-04 loss: 7.403e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2517/30000 [5:40:32<57:12:53,  7.49s/it, lr: 1.0e-04 loss: 7.403e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2517/30000 [5:40:32<57:12:53,  7.49s/it, lr: 1.0e-04 loss: 7.403e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2517/30000 [5:40:42<57:12:53,  7.49s/it, lr: 1.0e-04 loss: 6.250e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2517/30000 [5:40:42<57:12:53,  7.49s/it, lr: 1.0e-04 loss: 6.250e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2518/30000 [5:40:42<61:15:06,  8.02s/it, lr: 1.0e-04 loss: 6.250e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2518/30000 [5:40:42<61:15:06,  8.02s/it, lr: 1.0e-04 loss: 6.250e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2518/30000 [5:40:51<61:15:06,  8.02s/it, lr: 1.0e-04 loss: 4.211e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2518/30000 [5:40:51<61:15:06,  8.02s/it, lr: 1.0e-04 loss: 4.211e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2519/30000 [5:40:51<64:22:23,  8.43s/it, lr: 1.0e-04 loss: 4.211e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2519/30000 [5:40:51<64:22:23,  8.43s/it, lr: 1.0e-04 loss: 4.211e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2519/30000 [5:40:58<64:22:23,  8.43s/it, lr: 1.0e-04 loss: 7.679e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2519/30000 [5:40:58<64:22:23,  8.43s/it, lr: 1.0e-04 loss: 7.679e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2520/30000 [5:41:07<64:22:14,  8.43s/it, lr: 1.0e-04 loss: 6.343e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2520/30000 [5:41:07<64:22:14,  8.43s/it, lr: 1.0e-04 loss: 6.343e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2521/30000 [5:41:07<51:24:26,  6.73s/it, lr: 1.0e-04 loss: 6.343e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2521/30000 [5:41:07<51:24:26,  6.73s/it, lr: 1.0e-04 loss: 6.343e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2521/30000 [5:41:14<51:24:26,  6.73s/it, lr: 1.0e-04 loss: 8.572e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2521/30000 [5:41:14<51:24:26,  6.73s/it, lr: 1.0e-04 loss: 8.572e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2522/30000 [5:41:14<51:12:28,  6.71s/it, lr: 1.0e-04 loss: 8.572e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2522/30000 [5:41:14<51:12:28,  6.71s/it, lr: 1.0e-04 loss: 8.572e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2522/30000 [5:41:21<51:12:28,  6.71s/it, lr: 1.0e-04 loss: 1.163e-01]
Xiang_wan_2_2_lora:   8%|###2                                   | 2522/30000 [5:41:21<51:12:28,  6.71s/it, lr: 1.0e-04 loss: 1.163e-01]
Xiang_wan_2_2_lora:   8%|###2                                   | 2523/30000 [5:41:21<51:29:07,  6.75s/it, lr: 1.0e-04 loss: 1.163e-01]
Xiang_wan_2_2_lora:   8%|###2                                   | 2523/30000 [5:41:21<51:29:07,  6.75s/it, lr: 1.0e-04 loss: 1.163e-01]
Xiang_wan_2_2_lora:   8%|###2                                   | 2523/30000 [5:41:30<51:29:07,  6.75s/it, lr: 1.0e-04 loss: 7.400e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2523/30000 [5:41:30<51:29:07,  6.75s/it, lr: 1.0e-04 loss: 7.400e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2524/30000 [5:41:30<56:52:47,  7.45s/it, lr: 1.0e-04 loss: 7.400e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2524/30000 [5:41:30<56:52:47,  7.45s/it, lr: 1.0e-04 loss: 7.400e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2524/30000 [5:41:37<56:52:47,  7.45s/it, lr: 1.0e-04 loss: 6.716e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2524/30000 [5:41:37<56:52:47,  7.45s/it, lr: 1.0e-04 loss: 6.716e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2525/30000 [5:41:37<55:11:44,  7.23s/it, lr: 1.0e-04 loss: 6.716e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2525/30000 [5:41:37<55:11:44,  7.23s/it, lr: 1.0e-04 loss: 6.716e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2525/30000 [5:41:43<55:11:44,  7.23s/it, lr: 1.0e-04 loss: 8.883e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2525/30000 [5:41:43<55:11:44,  7.23s/it, lr: 1.0e-04 loss: 8.883e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2526/30000 [5:41:43<53:42:21,  7.04s/it, lr: 1.0e-04 loss: 8.883e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2526/30000 [5:41:43<53:42:21,  7.04s/it, lr: 1.0e-04 loss: 8.883e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2526/30000 [5:41:50<53:42:21,  7.04s/it, lr: 1.0e-04 loss: 2.966e-01]
Xiang_wan_2_2_lora:   8%|###2                                   | 2526/30000 [5:41:50<53:42:21,  7.04s/it, lr: 1.0e-04 loss: 2.966e-01]
Xiang_wan_2_2_lora:   8%|###2                                   | 2527/30000 [5:41:50<52:42:52,  6.91s/it, lr: 1.0e-04 loss: 2.966e-01]
Xiang_wan_2_2_lora:   8%|###2                                   | 2527/30000 [5:41:50<52:42:52,  6.91s/it, lr: 1.0e-04 loss: 2.966e-01]
Xiang_wan_2_2_lora:   8%|###2                                   | 2527/30000 [5:41:56<52:42:52,  6.91s/it, lr: 1.0e-04 loss: 7.847e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2527/30000 [5:41:56<52:42:52,  6.91s/it, lr: 1.0e-04 loss: 7.847e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2528/30000 [5:41:56<51:57:08,  6.81s/it, lr: 1.0e-04 loss: 7.847e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2528/30000 [5:41:56<51:57:08,  6.81s/it, lr: 1.0e-04 loss: 7.847e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2528/30000 [5:42:03<51:57:08,  6.81s/it, lr: 1.0e-04 loss: 8.323e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2528/30000 [5:42:03<51:57:08,  6.81s/it, lr: 1.0e-04 loss: 8.323e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2529/30000 [5:42:03<51:25:16,  6.74s/it, lr: 1.0e-04 loss: 8.323e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2529/30000 [5:42:03<51:25:16,  6.74s/it, lr: 1.0e-04 loss: 8.323e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2529/30000 [5:42:10<51:25:16,  6.74s/it, lr: 1.0e-04 loss: 1.024e-01]
Xiang_wan_2_2_lora:   8%|###2                                   | 2529/30000 [5:42:10<51:25:16,  6.74s/it, lr: 1.0e-04 loss: 1.024e-01]
Xiang_wan_2_2_lora:   8%|###2                                   | 2530/30000 [5:42:17<51:25:09,  6.74s/it, lr: 1.0e-04 loss: 8.745e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2530/30000 [5:42:17<51:25:09,  6.74s/it, lr: 1.0e-04 loss: 8.745e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2531/30000 [5:42:17<39:32:22,  5.18s/it, lr: 1.0e-04 loss: 8.745e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2531/30000 [5:42:17<39:32:22,  5.18s/it, lr: 1.0e-04 loss: 8.745e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2531/30000 [5:42:23<39:32:22,  5.18s/it, lr: 1.0e-04 loss: 6.525e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2531/30000 [5:42:23<39:32:22,  5.18s/it, lr: 1.0e-04 loss: 6.525e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2532/30000 [5:42:23<42:21:33,  5.55s/it, lr: 1.0e-04 loss: 6.525e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2532/30000 [5:42:23<42:21:33,  5.55s/it, lr: 1.0e-04 loss: 6.525e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2532/30000 [5:42:33<42:21:33,  5.55s/it, lr: 1.0e-04 loss: 3.145e-01]
Xiang_wan_2_2_lora:   8%|###2                                   | 2532/30000 [5:42:33<42:21:33,  5.55s/it, lr: 1.0e-04 loss: 3.145e-01]
Xiang_wan_2_2_lora:   8%|###2                                   | 2533/30000 [5:42:33<49:55:24,  6.54s/it, lr: 1.0e-04 loss: 3.145e-01]
Xiang_wan_2_2_lora:   8%|###2                                   | 2533/30000 [5:42:33<49:55:24,  6.54s/it, lr: 1.0e-04 loss: 3.145e-01]
Xiang_wan_2_2_lora:   8%|###2                                   | 2533/30000 [5:42:39<49:55:24,  6.54s/it, lr: 1.0e-04 loss: 6.560e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2533/30000 [5:42:39<49:55:24,  6.54s/it, lr: 1.0e-04 loss: 6.560e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2534/30000 [5:42:39<50:18:25,  6.59s/it, lr: 1.0e-04 loss: 6.560e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2534/30000 [5:42:39<50:18:25,  6.59s/it, lr: 1.0e-04 loss: 6.560e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2534/30000 [5:42:46<50:18:25,  6.59s/it, lr: 1.0e-04 loss: 3.506e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2534/30000 [5:42:46<50:18:25,  6.59s/it, lr: 1.0e-04 loss: 3.506e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2535/30000 [5:42:46<50:27:54,  6.61s/it, lr: 1.0e-04 loss: 3.506e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2535/30000 [5:42:46<50:27:54,  6.61s/it, lr: 1.0e-04 loss: 3.506e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2535/30000 [5:42:55<50:27:54,  6.61s/it, lr: 1.0e-04 loss: 6.478e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2535/30000 [5:42:55<50:27:54,  6.61s/it, lr: 1.0e-04 loss: 6.478e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2536/30000 [5:42:55<56:31:45,  7.41s/it, lr: 1.0e-04 loss: 6.478e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2536/30000 [5:42:55<56:31:45,  7.41s/it, lr: 1.0e-04 loss: 6.478e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2536/30000 [5:43:02<56:31:45,  7.41s/it, lr: 1.0e-04 loss: 9.884e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2536/30000 [5:43:02<56:31:45,  7.41s/it, lr: 1.0e-04 loss: 9.884e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2537/30000 [5:43:02<54:39:02,  7.16s/it, lr: 1.0e-04 loss: 9.884e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2537/30000 [5:43:02<54:39:02,  7.16s/it, lr: 1.0e-04 loss: 9.884e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2537/30000 [5:43:09<54:39:02,  7.16s/it, lr: 1.0e-04 loss: 8.974e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2537/30000 [5:43:09<54:39:02,  7.16s/it, lr: 1.0e-04 loss: 8.974e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2538/30000 [5:43:09<54:04:30,  7.09s/it, lr: 1.0e-04 loss: 8.974e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2538/30000 [5:43:09<54:04:30,  7.09s/it, lr: 1.0e-04 loss: 8.974e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2538/30000 [5:43:16<54:04:30,  7.09s/it, lr: 1.0e-04 loss: 5.378e-02]
Xiang_wan_2_2_lora:   8%|###2                                   | 2538/30000 [5:43:16<54:04:30,  7.09s/it, lr: 1.0e-04 loss: 5.378e-02]
Xiang_wan_2_2_lora:   8%|###3                                   | 2539/30000 [5:43:16<53:03:07,  6.95s/it, lr: 1.0e-04 loss: 5.378e-02]
Xiang_wan_2_2_lora:   8%|###3                                   | 2539/30000 [5:43:16<53:03:07,  6.95s/it, lr: 1.0e-04 loss: 5.378e-02]
Xiang_wan_2_2_lora:   8%|###3                                   | 2539/30000 [5:43:25<53:03:07,  6.95s/it, lr: 1.0e-04 loss: 5.793e-02]
Xiang_wan_2_2_lora:   8%|###3                                   | 2539/30000 [5:43:25<53:03:07,  6.95s/it, lr: 1.0e-04 loss: 5.793e-02]
Xiang_wan_2_2_lora:   8%|###3                                   | 2540/30000 [5:43:34<53:03:00,  6.95s/it, lr: 1.0e-04 loss: 1.135e-01]
Xiang_wan_2_2_lora:   8%|###3                                   | 2540/30000 [5:43:34<53:03:00,  6.95s/it, lr: 1.0e-04 loss: 1.135e-01]
Xiang_wan_2_2_lora:   8%|###3                                   | 2541/30000 [5:43:34<45:06:23,  5.91s/it, lr: 1.0e-04 loss: 1.135e-01]
Xiang_wan_2_2_lora:   8%|###3                                   | 2541/30000 [5:43:34<45:06:23,  5.91s/it, lr: 1.0e-04 loss: 1.135e-01]
Xiang_wan_2_2_lora:   8%|###3                                   | 2541/30000 [5:43:41<45:06:23,  5.91s/it, lr: 1.0e-04 loss: 8.668e-02]
Xiang_wan_2_2_lora:   8%|###3                                   | 2541/30000 [5:43:41<45:06:23,  5.91s/it, lr: 1.0e-04 loss: 8.668e-02]
Xiang_wan_2_2_lora:   8%|###3                                   | 2542/30000 [5:43:41<46:30:55,  6.10s/it, lr: 1.0e-04 loss: 8.668e-02]
Xiang_wan_2_2_lora:   8%|###3                                   | 2542/30000 [5:43:41<46:30:55,  6.10s/it, lr: 1.0e-04 loss: 8.668e-02]
Xiang_wan_2_2_lora:   8%|###3                                   | 2542/30000 [5:43:47<46:30:55,  6.10s/it, lr: 1.0e-04 loss: 7.074e-02]
Xiang_wan_2_2_lora:   8%|###3                                   | 2542/30000 [5:43:47<46:30:55,  6.10s/it, lr: 1.0e-04 loss: 7.074e-02]
Xiang_wan_2_2_lora:   8%|###3                                   | 2543/30000 [5:43:48<47:31:52,  6.23s/it, lr: 1.0e-04 loss: 7.074e-02]
Xiang_wan_2_2_lora:   8%|###3                                   | 2543/30000 [5:43:48<47:31:52,  6.23s/it, lr: 1.0e-04 loss: 7.074e-02]
Xiang_wan_2_2_lora:   8%|###3                                   | 2543/30000 [5:43:54<47:31:52,  6.23s/it, lr: 1.0e-04 loss: 9.345e-02]
Xiang_wan_2_2_lora:   8%|###3                                   | 2543/30000 [5:43:54<47:31:52,  6.23s/it, lr: 1.0e-04 loss: 9.345e-02]
Xiang_wan_2_2_lora:   8%|###3                                   | 2544/30000 [5:43:54<48:21:49,  6.34s/it, lr: 1.0e-04 loss: 9.345e-02]
Xiang_wan_2_2_lora:   8%|###3                                   | 2544/30000 [5:43:54<48:21:49,  6.34s/it, lr: 1.0e-04 loss: 9.345e-02]
Xiang_wan_2_2_lora:   8%|###3                                   | 2544/30000 [5:44:04<48:21:49,  6.34s/it, lr: 1.0e-04 loss: 7.688e-01]
Xiang_wan_2_2_lora:   8%|###3                                   | 2544/30000 [5:44:04<48:21:49,  6.34s/it, lr: 1.0e-04 loss: 7.688e-01]
Xiang_wan_2_2_lora:   8%|###3                                   | 2545/30000 [5:44:04<55:00:48,  7.21s/it, lr: 1.0e-04 loss: 7.688e-01]
Xiang_wan_2_2_lora:   8%|###3                                   | 2545/30000 [5:44:04<55:00:48,  7.21s/it, lr: 1.0e-04 loss: 7.688e-01]
Xiang_wan_2_2_lora:   8%|###3                                   | 2545/30000 [5:44:10<55:00:48,  7.21s/it, lr: 1.0e-04 loss: 8.534e-02]
Xiang_wan_2_2_lora:   8%|###3                                   | 2545/30000 [5:44:10<55:00:48,  7.21s/it, lr: 1.0e-04 loss: 8.534e-02]
Xiang_wan_2_2_lora:   8%|###3                                   | 2546/30000 [5:44:10<53:39:48,  7.04s/it, lr: 1.0e-04 loss: 8.534e-02]
Xiang_wan_2_2_lora:   8%|###3                                   | 2546/30000 [5:44:10<53:39:48,  7.04s/it, lr: 1.0e-04 loss: 8.534e-02]
Xiang_wan_2_2_lora:   8%|###3                                   | 2546/30000 [5:44:17<53:39:48,  7.04s/it, lr: 1.0e-04 loss: 3.754e-01]
Xiang_wan_2_2_lora:   8%|###3                                   | 2546/30000 [5:44:17<53:39:48,  7.04s/it, lr: 1.0e-04 loss: 3.754e-01]
Xiang_wan_2_2_lora:   8%|###3                                   | 2547/30000 [5:44:17<52:32:05,  6.89s/it, lr: 1.0e-04 loss: 3.754e-01]
Xiang_wan_2_2_lora:   8%|###3                                   | 2547/30000 [5:44:17<52:32:05,  6.89s/it, lr: 1.0e-04 loss: 3.754e-01]
Xiang_wan_2_2_lora:   8%|###3                                   | 2547/30000 [5:44:26<52:32:05,  6.89s/it, lr: 1.0e-04 loss: 9.817e-02]
Xiang_wan_2_2_lora:   8%|###3                                   | 2547/30000 [5:44:26<52:32:05,  6.89s/it, lr: 1.0e-04 loss: 9.817e-02]
Xiang_wan_2_2_lora:   8%|###3                                   | 2548/30000 [5:44:26<57:53:49,  7.59s/it, lr: 1.0e-04 loss: 9.817e-02]
Xiang_wan_2_2_lora:   8%|###3                                   | 2548/30000 [5:44:26<57:53:49,  7.59s/it, lr: 1.0e-04 loss: 9.817e-02]
Xiang_wan_2_2_lora:   8%|###3                                   | 2548/30000 [5:44:33<57:53:49,  7.59s/it, lr: 1.0e-04 loss: 5.839e-02]
Xiang_wan_2_2_lora:   8%|###3                                   | 2548/30000 [5:44:33<57:53:49,  7.59s/it, lr: 1.0e-04 loss: 5.839e-02]
Xiang_wan_2_2_lora:   8%|###3                                   | 2549/30000 [5:44:33<55:40:54,  7.30s/it, lr: 1.0e-04 loss: 5.839e-02]
Xiang_wan_2_2_lora:   8%|###3                                   | 2549/30000 [5:44:33<55:40:54,  7.30s/it, lr: 1.0e-04 loss: 5.839e-02]
Xiang_wan_2_2_lora:   8%|###3                                   | 2549/30000 [5:44:42<55:40:54,  7.30s/it, lr: 1.0e-04 loss: 5.178e-02]
Xiang_wan_2_2_lora:   8%|###3                                   | 2549/30000 [5:44:42<55:40:54,  7.30s/it, lr: 1.0e-04 loss: 5.178e-02]
Xiang_wan_2_2_lora:   8%|###3                                   | 2550/30000 [5:44:49<55:40:46,  7.30s/it, lr: 1.0e-04 loss: 5.872e-02]
Xiang_wan_2_2_lora:   8%|###3                                   | 2550/30000 [5:44:49<55:40:46,  7.30s/it, lr: 1.0e-04 loss: 5.872e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2551/30000 [5:44:49<41:48:01,  5.48s/it, lr: 1.0e-04 loss: 5.872e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2551/30000 [5:44:49<41:48:01,  5.48s/it, lr: 1.0e-04 loss: 5.872e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2551/30000 [5:44:55<41:48:01,  5.48s/it, lr: 1.0e-04 loss: 1.027e-01]
Xiang_wan_2_2_lora:   9%|###3                                   | 2551/30000 [5:44:55<41:48:01,  5.48s/it, lr: 1.0e-04 loss: 1.027e-01]
Xiang_wan_2_2_lora:   9%|###3                                   | 2552/30000 [5:44:55<43:48:11,  5.75s/it, lr: 1.0e-04 loss: 1.027e-01]
Xiang_wan_2_2_lora:   9%|###3                                   | 2552/30000 [5:44:55<43:48:11,  5.75s/it, lr: 1.0e-04 loss: 1.027e-01]
Xiang_wan_2_2_lora:   9%|###3                                   | 2552/30000 [5:45:05<43:48:11,  5.75s/it, lr: 1.0e-04 loss: 5.137e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2552/30000 [5:45:05<43:48:11,  5.75s/it, lr: 1.0e-04 loss: 5.137e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2553/30000 [5:45:05<51:24:34,  6.74s/it, lr: 1.0e-04 loss: 5.137e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2553/30000 [5:45:05<51:24:34,  6.74s/it, lr: 1.0e-04 loss: 5.137e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2553/30000 [5:45:14<51:24:34,  6.74s/it, lr: 1.0e-04 loss: 5.989e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2553/30000 [5:45:14<51:24:34,  6.74s/it, lr: 1.0e-04 loss: 5.989e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2554/30000 [5:45:14<56:46:04,  7.45s/it, lr: 1.0e-04 loss: 5.989e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2554/30000 [5:45:14<56:46:04,  7.45s/it, lr: 1.0e-04 loss: 5.989e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2554/30000 [5:45:21<56:46:04,  7.45s/it, lr: 1.0e-04 loss: 6.708e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2554/30000 [5:45:21<56:46:04,  7.45s/it, lr: 1.0e-04 loss: 6.708e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2555/30000 [5:45:21<54:45:47,  7.18s/it, lr: 1.0e-04 loss: 6.708e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2555/30000 [5:45:21<54:45:47,  7.18s/it, lr: 1.0e-04 loss: 6.708e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2555/30000 [5:45:27<54:45:47,  7.18s/it, lr: 1.0e-04 loss: 8.368e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2555/30000 [5:45:27<54:45:47,  7.18s/it, lr: 1.0e-04 loss: 8.368e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2556/30000 [5:45:27<53:23:45,  7.00s/it, lr: 1.0e-04 loss: 8.368e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2556/30000 [5:45:27<53:23:45,  7.00s/it, lr: 1.0e-04 loss: 8.368e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2556/30000 [5:45:36<53:23:45,  7.00s/it, lr: 1.0e-04 loss: 4.354e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2556/30000 [5:45:36<53:23:45,  7.00s/it, lr: 1.0e-04 loss: 4.354e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2557/30000 [5:45:36<58:14:15,  7.64s/it, lr: 1.0e-04 loss: 4.354e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2557/30000 [5:45:36<58:14:15,  7.64s/it, lr: 1.0e-04 loss: 4.354e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2557/30000 [5:45:43<58:14:15,  7.64s/it, lr: 1.0e-04 loss: 8.623e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2557/30000 [5:45:43<58:14:15,  7.64s/it, lr: 1.0e-04 loss: 8.623e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2558/30000 [5:45:43<55:50:15,  7.33s/it, lr: 1.0e-04 loss: 8.623e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2558/30000 [5:45:43<55:50:15,  7.33s/it, lr: 1.0e-04 loss: 8.623e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2558/30000 [5:45:49<55:50:15,  7.33s/it, lr: 1.0e-04 loss: 5.602e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2558/30000 [5:45:49<55:50:15,  7.33s/it, lr: 1.0e-04 loss: 5.602e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2559/30000 [5:45:49<53:57:44,  7.08s/it, lr: 1.0e-04 loss: 5.602e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2559/30000 [5:45:49<53:57:44,  7.08s/it, lr: 1.0e-04 loss: 5.602e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2559/30000 [5:45:59<53:57:44,  7.08s/it, lr: 1.0e-04 loss: 6.189e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2559/30000 [5:45:59<53:57:44,  7.08s/it, lr: 1.0e-04 loss: 6.189e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2560/30000 [5:46:08<53:57:37,  7.08s/it, lr: 1.0e-04 loss: 4.320e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2560/30000 [5:46:08<53:57:37,  7.08s/it, lr: 1.0e-04 loss: 4.320e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2561/30000 [5:46:08<45:33:14,  5.98s/it, lr: 1.0e-04 loss: 4.320e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2561/30000 [5:46:08<45:33:14,  5.98s/it, lr: 1.0e-04 loss: 4.320e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2561/30000 [5:46:15<45:33:14,  5.98s/it, lr: 1.0e-04 loss: 6.543e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2561/30000 [5:46:15<45:33:14,  5.98s/it, lr: 1.0e-04 loss: 6.543e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2562/30000 [5:46:15<46:44:10,  6.13s/it, lr: 1.0e-04 loss: 6.543e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2562/30000 [5:46:15<46:44:10,  6.13s/it, lr: 1.0e-04 loss: 6.543e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2562/30000 [5:46:21<46:44:10,  6.13s/it, lr: 1.0e-04 loss: 1.337e-01]
Xiang_wan_2_2_lora:   9%|###3                                   | 2562/30000 [5:46:21<46:44:10,  6.13s/it, lr: 1.0e-04 loss: 1.337e-01]
Xiang_wan_2_2_lora:   9%|###3                                   | 2563/30000 [5:46:21<47:26:36,  6.23s/it, lr: 1.0e-04 loss: 1.337e-01]
Xiang_wan_2_2_lora:   9%|###3                                   | 2563/30000 [5:46:21<47:26:36,  6.23s/it, lr: 1.0e-04 loss: 1.337e-01]
Xiang_wan_2_2_lora:   9%|###3                                   | 2563/30000 [5:46:31<47:26:36,  6.23s/it, lr: 1.0e-04 loss: 5.950e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2563/30000 [5:46:31<47:26:36,  6.23s/it, lr: 1.0e-04 loss: 5.950e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2564/30000 [5:46:31<53:45:40,  7.05s/it, lr: 1.0e-04 loss: 5.950e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2564/30000 [5:46:31<53:45:40,  7.05s/it, lr: 1.0e-04 loss: 5.950e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2564/30000 [5:46:37<53:45:40,  7.05s/it, lr: 1.0e-04 loss: 6.153e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2564/30000 [5:46:37<53:45:40,  7.05s/it, lr: 1.0e-04 loss: 6.153e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2565/30000 [5:46:37<52:43:39,  6.92s/it, lr: 1.0e-04 loss: 6.153e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2565/30000 [5:46:37<52:43:39,  6.92s/it, lr: 1.0e-04 loss: 6.153e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2565/30000 [5:46:44<52:43:39,  6.92s/it, lr: 1.0e-04 loss: 7.807e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2565/30000 [5:46:44<52:43:39,  6.92s/it, lr: 1.0e-04 loss: 7.807e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2566/30000 [5:46:44<51:53:36,  6.81s/it, lr: 1.0e-04 loss: 7.807e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2566/30000 [5:46:44<51:53:36,  6.81s/it, lr: 1.0e-04 loss: 7.807e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2566/30000 [5:46:51<51:53:36,  6.81s/it, lr: 1.0e-04 loss: 5.495e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2566/30000 [5:46:51<51:53:36,  6.81s/it, lr: 1.0e-04 loss: 5.495e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2567/30000 [5:46:51<51:55:00,  6.81s/it, lr: 1.0e-04 loss: 5.495e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2567/30000 [5:46:51<51:55:00,  6.81s/it, lr: 1.0e-04 loss: 5.495e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2567/30000 [5:46:57<51:55:00,  6.81s/it, lr: 1.0e-04 loss: 8.852e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2567/30000 [5:46:57<51:55:00,  6.81s/it, lr: 1.0e-04 loss: 8.852e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2568/30000 [5:46:57<51:13:10,  6.72s/it, lr: 1.0e-04 loss: 8.852e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2568/30000 [5:46:57<51:13:10,  6.72s/it, lr: 1.0e-04 loss: 8.852e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2568/30000 [5:47:06<51:13:10,  6.72s/it, lr: 1.0e-04 loss: 6.035e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2568/30000 [5:47:06<51:13:10,  6.72s/it, lr: 1.0e-04 loss: 6.035e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2569/30000 [5:47:06<56:57:51,  7.48s/it, lr: 1.0e-04 loss: 6.035e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2569/30000 [5:47:06<56:57:51,  7.48s/it, lr: 1.0e-04 loss: 6.035e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2569/30000 [5:47:16<56:57:51,  7.48s/it, lr: 1.0e-04 loss: 6.802e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2569/30000 [5:47:16<56:57:51,  7.48s/it, lr: 1.0e-04 loss: 6.802e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2570/30000 [5:47:25<56:57:43,  7.48s/it, lr: 1.0e-04 loss: 2.331e-01]
Xiang_wan_2_2_lora:   9%|###3                                   | 2570/30000 [5:47:25<56:57:43,  7.48s/it, lr: 1.0e-04 loss: 2.331e-01]
Xiang_wan_2_2_lora:   9%|###3                                   | 2571/30000 [5:47:25<47:14:29,  6.20s/it, lr: 1.0e-04 loss: 2.331e-01]
Xiang_wan_2_2_lora:   9%|###3                                   | 2571/30000 [5:47:25<47:14:29,  6.20s/it, lr: 1.0e-04 loss: 2.331e-01]
Xiang_wan_2_2_lora:   9%|###3                                   | 2571/30000 [5:47:34<47:14:29,  6.20s/it, lr: 1.0e-04 loss: 6.104e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2571/30000 [5:47:34<47:14:29,  6.20s/it, lr: 1.0e-04 loss: 6.104e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2572/30000 [5:47:34<53:02:51,  6.96s/it, lr: 1.0e-04 loss: 6.104e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2572/30000 [5:47:34<53:02:51,  6.96s/it, lr: 1.0e-04 loss: 6.104e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2572/30000 [5:47:44<53:02:51,  6.96s/it, lr: 1.0e-04 loss: 8.776e-01]
Xiang_wan_2_2_lora:   9%|###3                                   | 2572/30000 [5:47:44<53:02:51,  6.96s/it, lr: 1.0e-04 loss: 8.776e-01]
Xiang_wan_2_2_lora:   9%|###3                                   | 2573/30000 [5:47:44<57:44:06,  7.58s/it, lr: 1.0e-04 loss: 8.776e-01]
Xiang_wan_2_2_lora:   9%|###3                                   | 2573/30000 [5:47:44<57:44:06,  7.58s/it, lr: 1.0e-04 loss: 8.776e-01]
Xiang_wan_2_2_lora:   9%|###3                                   | 2573/30000 [5:47:53<57:44:06,  7.58s/it, lr: 1.0e-04 loss: 7.653e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2573/30000 [5:47:53<57:44:06,  7.58s/it, lr: 1.0e-04 loss: 7.653e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2574/30000 [5:47:53<61:20:44,  8.05s/it, lr: 1.0e-04 loss: 7.653e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2574/30000 [5:47:53<61:20:44,  8.05s/it, lr: 1.0e-04 loss: 7.653e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2574/30000 [5:48:00<61:20:44,  8.05s/it, lr: 1.0e-04 loss: 7.174e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2574/30000 [5:48:00<61:20:44,  8.05s/it, lr: 1.0e-04 loss: 7.174e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2575/30000 [5:48:00<58:49:42,  7.72s/it, lr: 1.0e-04 loss: 7.174e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2575/30000 [5:48:00<58:49:42,  7.72s/it, lr: 1.0e-04 loss: 7.174e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2575/30000 [5:48:09<58:49:42,  7.72s/it, lr: 1.0e-04 loss: 4.985e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2575/30000 [5:48:09<58:49:42,  7.72s/it, lr: 1.0e-04 loss: 4.985e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2576/30000 [5:48:09<62:25:42,  8.20s/it, lr: 1.0e-04 loss: 4.985e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2576/30000 [5:48:09<62:25:42,  8.20s/it, lr: 1.0e-04 loss: 4.985e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2576/30000 [5:48:16<62:25:42,  8.20s/it, lr: 1.0e-04 loss: 4.117e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2576/30000 [5:48:16<62:25:42,  8.20s/it, lr: 1.0e-04 loss: 4.117e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2577/30000 [5:48:16<58:56:06,  7.74s/it, lr: 1.0e-04 loss: 4.117e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2577/30000 [5:48:16<58:56:06,  7.74s/it, lr: 1.0e-04 loss: 4.117e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2577/30000 [5:48:22<58:56:06,  7.74s/it, lr: 1.0e-04 loss: 8.255e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2577/30000 [5:48:22<58:56:06,  7.74s/it, lr: 1.0e-04 loss: 8.255e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2578/30000 [5:48:22<56:18:11,  7.39s/it, lr: 1.0e-04 loss: 8.255e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2578/30000 [5:48:22<56:18:11,  7.39s/it, lr: 1.0e-04 loss: 8.255e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2578/30000 [5:48:32<56:18:11,  7.39s/it, lr: 1.0e-04 loss: 3.561e-01]
Xiang_wan_2_2_lora:   9%|###3                                   | 2578/30000 [5:48:32<56:18:11,  7.39s/it, lr: 1.0e-04 loss: 3.561e-01]
Xiang_wan_2_2_lora:   9%|###3                                   | 2579/30000 [5:48:32<60:47:40,  7.98s/it, lr: 1.0e-04 loss: 3.561e-01]
Xiang_wan_2_2_lora:   9%|###3                                   | 2579/30000 [5:48:32<60:47:40,  7.98s/it, lr: 1.0e-04 loss: 3.561e-01]
Xiang_wan_2_2_lora:   9%|###3                                   | 2579/30000 [5:48:38<60:47:40,  7.98s/it, lr: 1.0e-04 loss: 8.677e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2579/30000 [5:48:38<60:47:40,  7.98s/it, lr: 1.0e-04 loss: 8.677e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2580/30000 [5:48:48<60:47:32,  7.98s/it, lr: 1.0e-04 loss: 5.367e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2580/30000 [5:48:48<60:47:32,  7.98s/it, lr: 1.0e-04 loss: 5.367e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2581/30000 [5:48:48<49:17:37,  6.47s/it, lr: 1.0e-04 loss: 5.367e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2581/30000 [5:48:48<49:17:37,  6.47s/it, lr: 1.0e-04 loss: 5.367e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2581/30000 [5:48:57<49:17:37,  6.47s/it, lr: 1.0e-04 loss: 6.843e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2581/30000 [5:48:57<49:17:37,  6.47s/it, lr: 1.0e-04 loss: 6.843e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2582/30000 [5:48:57<55:04:54,  7.23s/it, lr: 1.0e-04 loss: 6.843e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2582/30000 [5:48:57<55:04:54,  7.23s/it, lr: 1.0e-04 loss: 6.843e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2582/30000 [5:49:04<55:04:54,  7.23s/it, lr: 1.0e-04 loss: 6.336e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2582/30000 [5:49:04<55:04:54,  7.23s/it, lr: 1.0e-04 loss: 6.336e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2583/30000 [5:49:04<53:51:55,  7.07s/it, lr: 1.0e-04 loss: 6.336e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2583/30000 [5:49:04<53:51:55,  7.07s/it, lr: 1.0e-04 loss: 6.336e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2583/30000 [5:49:13<53:51:55,  7.07s/it, lr: 1.0e-04 loss: 6.741e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2583/30000 [5:49:13<53:51:55,  7.07s/it, lr: 1.0e-04 loss: 6.741e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2584/30000 [5:49:13<58:34:48,  7.69s/it, lr: 1.0e-04 loss: 6.741e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2584/30000 [5:49:13<58:34:48,  7.69s/it, lr: 1.0e-04 loss: 6.741e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2584/30000 [5:49:23<58:34:48,  7.69s/it, lr: 1.0e-04 loss: 4.912e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2584/30000 [5:49:23<58:34:48,  7.69s/it, lr: 1.0e-04 loss: 4.912e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2585/30000 [5:49:23<62:07:40,  8.16s/it, lr: 1.0e-04 loss: 4.912e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2585/30000 [5:49:23<62:07:40,  8.16s/it, lr: 1.0e-04 loss: 4.912e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2585/30000 [5:49:32<62:07:40,  8.16s/it, lr: 1.0e-04 loss: 5.462e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2585/30000 [5:49:32<62:07:40,  8.16s/it, lr: 1.0e-04 loss: 5.462e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2586/30000 [5:49:32<64:38:48,  8.49s/it, lr: 1.0e-04 loss: 5.462e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2586/30000 [5:49:32<64:38:48,  8.49s/it, lr: 1.0e-04 loss: 5.462e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2586/30000 [5:49:41<64:38:48,  8.49s/it, lr: 1.0e-04 loss: 4.730e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2586/30000 [5:49:41<64:38:48,  8.49s/it, lr: 1.0e-04 loss: 4.730e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2587/30000 [5:49:41<66:33:22,  8.74s/it, lr: 1.0e-04 loss: 4.730e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2587/30000 [5:49:41<66:33:22,  8.74s/it, lr: 1.0e-04 loss: 4.730e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2587/30000 [5:49:51<66:33:22,  8.74s/it, lr: 1.0e-04 loss: 5.949e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2587/30000 [5:49:51<66:33:22,  8.74s/it, lr: 1.0e-04 loss: 5.949e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2588/30000 [5:49:51<67:49:16,  8.91s/it, lr: 1.0e-04 loss: 5.949e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2588/30000 [5:49:51<67:49:16,  8.91s/it, lr: 1.0e-04 loss: 5.949e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2588/30000 [5:50:00<67:49:16,  8.91s/it, lr: 1.0e-04 loss: 6.767e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2588/30000 [5:50:00<67:49:16,  8.91s/it, lr: 1.0e-04 loss: 6.767e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2589/30000 [5:50:00<68:42:17,  9.02s/it, lr: 1.0e-04 loss: 6.767e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2589/30000 [5:50:00<68:42:17,  9.02s/it, lr: 1.0e-04 loss: 6.767e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2589/30000 [5:50:09<68:42:17,  9.02s/it, lr: 1.0e-04 loss: 6.964e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2589/30000 [5:50:09<68:42:17,  9.02s/it, lr: 1.0e-04 loss: 6.964e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2590/30000 [5:50:19<68:42:07,  9.02s/it, lr: 1.0e-04 loss: 4.736e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2590/30000 [5:50:19<68:42:07,  9.02s/it, lr: 1.0e-04 loss: 4.736e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2591/30000 [5:50:19<53:30:43,  7.03s/it, lr: 1.0e-04 loss: 4.736e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2591/30000 [5:50:19<53:30:43,  7.03s/it, lr: 1.0e-04 loss: 4.736e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2591/30000 [5:50:25<53:30:43,  7.03s/it, lr: 1.0e-04 loss: 9.352e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2591/30000 [5:50:25<53:30:43,  7.03s/it, lr: 1.0e-04 loss: 9.352e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2592/30000 [5:50:25<52:37:01,  6.91s/it, lr: 1.0e-04 loss: 9.352e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2592/30000 [5:50:25<52:37:01,  6.91s/it, lr: 1.0e-04 loss: 9.352e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2592/30000 [5:50:35<52:37:01,  6.91s/it, lr: 1.0e-04 loss: 1.148e-01]
Xiang_wan_2_2_lora:   9%|###3                                   | 2592/30000 [5:50:35<52:37:01,  6.91s/it, lr: 1.0e-04 loss: 1.148e-01]
Xiang_wan_2_2_lora:   9%|###3                                   | 2593/30000 [5:50:35<57:18:25,  7.53s/it, lr: 1.0e-04 loss: 1.148e-01]
Xiang_wan_2_2_lora:   9%|###3                                   | 2593/30000 [5:50:35<57:18:25,  7.53s/it, lr: 1.0e-04 loss: 1.148e-01]
Xiang_wan_2_2_lora:   9%|###3                                   | 2593/30000 [5:50:41<57:18:25,  7.53s/it, lr: 1.0e-04 loss: 9.777e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2593/30000 [5:50:41<57:18:25,  7.53s/it, lr: 1.0e-04 loss: 9.777e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2594/30000 [5:50:41<55:14:51,  7.26s/it, lr: 1.0e-04 loss: 9.777e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2594/30000 [5:50:41<55:14:51,  7.26s/it, lr: 1.0e-04 loss: 9.777e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2594/30000 [5:50:50<55:14:51,  7.26s/it, lr: 1.0e-04 loss: 4.424e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2594/30000 [5:50:50<55:14:51,  7.26s/it, lr: 1.0e-04 loss: 4.424e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2595/30000 [5:50:50<59:38:04,  7.83s/it, lr: 1.0e-04 loss: 4.424e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2595/30000 [5:50:50<59:38:04,  7.83s/it, lr: 1.0e-04 loss: 4.424e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2595/30000 [5:51:00<59:38:04,  7.83s/it, lr: 1.0e-04 loss: 6.608e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2595/30000 [5:51:00<59:38:04,  7.83s/it, lr: 1.0e-04 loss: 6.608e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2596/30000 [5:51:00<63:01:32,  8.28s/it, lr: 1.0e-04 loss: 6.608e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2596/30000 [5:51:00<63:01:32,  8.28s/it, lr: 1.0e-04 loss: 6.608e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2596/30000 [5:51:07<63:01:32,  8.28s/it, lr: 1.0e-04 loss: 8.184e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2596/30000 [5:51:07<63:01:32,  8.28s/it, lr: 1.0e-04 loss: 8.184e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2597/30000 [5:51:07<60:02:44,  7.89s/it, lr: 1.0e-04 loss: 8.184e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2597/30000 [5:51:07<60:02:44,  7.89s/it, lr: 1.0e-04 loss: 8.184e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2597/30000 [5:51:16<60:02:44,  7.89s/it, lr: 1.0e-04 loss: 4.727e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2597/30000 [5:51:16<60:02:44,  7.89s/it, lr: 1.0e-04 loss: 4.727e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2598/30000 [5:51:16<63:08:46,  8.30s/it, lr: 1.0e-04 loss: 4.727e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2598/30000 [5:51:16<63:08:46,  8.30s/it, lr: 1.0e-04 loss: 4.727e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2598/30000 [5:51:23<63:08:46,  8.30s/it, lr: 1.0e-04 loss: 3.649e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2598/30000 [5:51:23<63:08:46,  8.30s/it, lr: 1.0e-04 loss: 3.649e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2599/30000 [5:51:23<59:29:48,  7.82s/it, lr: 1.0e-04 loss: 3.649e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2599/30000 [5:51:23<59:29:48,  7.82s/it, lr: 1.0e-04 loss: 3.649e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2599/30000 [5:51:29<59:29:48,  7.82s/it, lr: 1.0e-04 loss: 7.665e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2599/30000 [5:51:29<59:29:48,  7.82s/it, lr: 1.0e-04 loss: 7.665e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2600/30000 [5:51:36<59:29:40,  7.82s/it, lr: 1.0e-04 loss: 7.030e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2600/30000 [5:51:36<59:29:40,  7.82s/it, lr: 1.0e-04 loss: 7.030e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2601/30000 [5:51:36<43:46:42,  5.75s/it, lr: 1.0e-04 loss: 7.030e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2601/30000 [5:51:36<43:46:42,  5.75s/it, lr: 1.0e-04 loss: 7.030e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2601/30000 [5:51:45<43:46:42,  5.75s/it, lr: 1.0e-04 loss: 6.829e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2601/30000 [5:51:45<43:46:42,  5.75s/it, lr: 1.0e-04 loss: 6.829e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2602/30000 [5:51:45<50:31:15,  6.64s/it, lr: 1.0e-04 loss: 6.829e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2602/30000 [5:51:45<50:31:15,  6.64s/it, lr: 1.0e-04 loss: 6.829e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2602/30000 [5:51:55<50:31:15,  6.64s/it, lr: 1.0e-04 loss: 1.034e-01]
Xiang_wan_2_2_lora:   9%|###3                                   | 2602/30000 [5:51:55<50:31:15,  6.64s/it, lr: 1.0e-04 loss: 1.034e-01]
Xiang_wan_2_2_lora:   9%|###3                                   | 2603/30000 [5:51:55<55:55:43,  7.35s/it, lr: 1.0e-04 loss: 1.034e-01]
Xiang_wan_2_2_lora:   9%|###3                                   | 2603/30000 [5:51:55<55:55:43,  7.35s/it, lr: 1.0e-04 loss: 1.034e-01]
Xiang_wan_2_2_lora:   9%|###3                                   | 2603/30000 [5:52:01<55:55:43,  7.35s/it, lr: 1.0e-04 loss: 6.759e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2603/30000 [5:52:01<55:55:43,  7.35s/it, lr: 1.0e-04 loss: 6.759e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2604/30000 [5:52:01<54:23:27,  7.15s/it, lr: 1.0e-04 loss: 6.759e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2604/30000 [5:52:01<54:23:27,  7.15s/it, lr: 1.0e-04 loss: 6.759e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2604/30000 [5:52:08<54:23:27,  7.15s/it, lr: 1.0e-04 loss: 4.727e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2604/30000 [5:52:08<54:23:27,  7.15s/it, lr: 1.0e-04 loss: 4.727e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2605/30000 [5:52:08<53:40:19,  7.05s/it, lr: 1.0e-04 loss: 4.727e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2605/30000 [5:52:08<53:40:19,  7.05s/it, lr: 1.0e-04 loss: 4.727e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2605/30000 [5:52:15<53:40:19,  7.05s/it, lr: 1.0e-04 loss: 9.901e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2605/30000 [5:52:15<53:40:19,  7.05s/it, lr: 1.0e-04 loss: 9.901e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2606/30000 [5:52:15<52:31:30,  6.90s/it, lr: 1.0e-04 loss: 9.901e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2606/30000 [5:52:15<52:31:30,  6.90s/it, lr: 1.0e-04 loss: 9.901e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2606/30000 [5:52:24<52:31:30,  6.90s/it, lr: 1.0e-04 loss: 5.354e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2606/30000 [5:52:24<52:31:30,  6.90s/it, lr: 1.0e-04 loss: 5.354e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2607/30000 [5:52:24<57:51:47,  7.60s/it, lr: 1.0e-04 loss: 5.354e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2607/30000 [5:52:24<57:51:47,  7.60s/it, lr: 1.0e-04 loss: 5.354e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2607/30000 [5:52:33<57:51:47,  7.60s/it, lr: 1.0e-04 loss: 5.580e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2607/30000 [5:52:33<57:51:47,  7.60s/it, lr: 1.0e-04 loss: 5.580e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2608/30000 [5:52:33<61:47:37,  8.12s/it, lr: 1.0e-04 loss: 5.580e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2608/30000 [5:52:33<61:47:37,  8.12s/it, lr: 1.0e-04 loss: 5.580e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2608/30000 [5:52:40<61:47:37,  8.12s/it, lr: 1.0e-04 loss: 4.003e-01]
Xiang_wan_2_2_lora:   9%|###3                                   | 2608/30000 [5:52:40<61:47:37,  8.12s/it, lr: 1.0e-04 loss: 4.003e-01]
Xiang_wan_2_2_lora:   9%|###3                                   | 2609/30000 [5:52:40<58:21:14,  7.67s/it, lr: 1.0e-04 loss: 4.003e-01]
Xiang_wan_2_2_lora:   9%|###3                                   | 2609/30000 [5:52:40<58:21:14,  7.67s/it, lr: 1.0e-04 loss: 4.003e-01]
Xiang_wan_2_2_lora:   9%|###3                                   | 2609/30000 [5:52:46<58:21:14,  7.67s/it, lr: 1.0e-04 loss: 1.135e-01]
Xiang_wan_2_2_lora:   9%|###3                                   | 2609/30000 [5:52:46<58:21:14,  7.67s/it, lr: 1.0e-04 loss: 1.135e-01]
Xiang_wan_2_2_lora:   9%|###3                                   | 2610/30000 [5:52:56<58:21:06,  7.67s/it, lr: 1.0e-04 loss: 3.992e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2610/30000 [5:52:56<58:21:06,  7.67s/it, lr: 1.0e-04 loss: 3.992e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2611/30000 [5:52:56<47:57:25,  6.30s/it, lr: 1.0e-04 loss: 3.992e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2611/30000 [5:52:56<47:57:25,  6.30s/it, lr: 1.0e-04 loss: 3.992e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2611/30000 [5:53:03<47:57:25,  6.30s/it, lr: 1.0e-04 loss: 7.763e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2611/30000 [5:53:03<47:57:25,  6.30s/it, lr: 1.0e-04 loss: 7.763e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2612/30000 [5:53:03<48:55:06,  6.43s/it, lr: 1.0e-04 loss: 7.763e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2612/30000 [5:53:03<48:55:06,  6.43s/it, lr: 1.0e-04 loss: 7.763e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2612/30000 [5:53:12<48:55:06,  6.43s/it, lr: 1.0e-04 loss: 3.683e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2612/30000 [5:53:12<48:55:06,  6.43s/it, lr: 1.0e-04 loss: 3.683e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2613/30000 [5:53:12<54:36:43,  7.18s/it, lr: 1.0e-04 loss: 3.683e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2613/30000 [5:53:12<54:36:43,  7.18s/it, lr: 1.0e-04 loss: 3.683e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2613/30000 [5:53:19<54:36:43,  7.18s/it, lr: 1.0e-04 loss: 9.689e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2613/30000 [5:53:19<54:36:43,  7.18s/it, lr: 1.0e-04 loss: 9.689e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2614/30000 [5:53:19<53:28:41,  7.03s/it, lr: 1.0e-04 loss: 9.689e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2614/30000 [5:53:19<53:28:41,  7.03s/it, lr: 1.0e-04 loss: 9.689e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2614/30000 [5:53:25<53:28:41,  7.03s/it, lr: 1.0e-04 loss: 5.245e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2614/30000 [5:53:25<53:28:41,  7.03s/it, lr: 1.0e-04 loss: 5.245e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2615/30000 [5:53:25<52:26:13,  6.89s/it, lr: 1.0e-04 loss: 5.245e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2615/30000 [5:53:25<52:26:13,  6.89s/it, lr: 1.0e-04 loss: 5.245e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2615/30000 [5:53:34<52:26:13,  6.89s/it, lr: 1.0e-04 loss: 7.904e-02]
Xiang_wan_2_2_lora:   9%|###3                                   | 2615/30000 [5:53:34<52:26:13,  6.89s/it, lr: 1.0e-04 loss: 7.904e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2616/30000 [5:53:34<57:36:01,  7.57s/it, lr: 1.0e-04 loss: 7.904e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2616/30000 [5:53:34<57:36:01,  7.57s/it, lr: 1.0e-04 loss: 7.904e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2616/30000 [5:53:44<57:36:01,  7.57s/it, lr: 1.0e-04 loss: 1.792e-01]
Xiang_wan_2_2_lora:   9%|###4                                   | 2616/30000 [5:53:44<57:36:01,  7.57s/it, lr: 1.0e-04 loss: 1.792e-01]
Xiang_wan_2_2_lora:   9%|###4                                   | 2617/30000 [5:53:44<61:25:20,  8.08s/it, lr: 1.0e-04 loss: 1.792e-01]
Xiang_wan_2_2_lora:   9%|###4                                   | 2617/30000 [5:53:44<61:25:20,  8.08s/it, lr: 1.0e-04 loss: 1.792e-01]
Xiang_wan_2_2_lora:   9%|###4                                   | 2617/30000 [5:53:50<61:25:20,  8.08s/it, lr: 1.0e-04 loss: 5.888e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2617/30000 [5:53:50<61:25:20,  8.08s/it, lr: 1.0e-04 loss: 5.888e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2618/30000 [5:53:50<57:57:58,  7.62s/it, lr: 1.0e-04 loss: 5.888e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2618/30000 [5:53:50<57:57:58,  7.62s/it, lr: 1.0e-04 loss: 5.888e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2618/30000 [5:53:57<57:57:58,  7.62s/it, lr: 1.0e-04 loss: 2.075e-01]
Xiang_wan_2_2_lora:   9%|###4                                   | 2618/30000 [5:53:57<57:57:58,  7.62s/it, lr: 1.0e-04 loss: 2.075e-01]
Xiang_wan_2_2_lora:   9%|###4                                   | 2619/30000 [5:53:57<55:34:04,  7.31s/it, lr: 1.0e-04 loss: 2.075e-01]
Xiang_wan_2_2_lora:   9%|###4                                   | 2619/30000 [5:53:57<55:34:04,  7.31s/it, lr: 1.0e-04 loss: 2.075e-01]
Xiang_wan_2_2_lora:   9%|###4                                   | 2619/30000 [5:54:04<55:34:04,  7.31s/it, lr: 1.0e-04 loss: 6.260e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2619/30000 [5:54:04<55:34:04,  7.31s/it, lr: 1.0e-04 loss: 6.260e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2620/30000 [5:54:10<55:33:57,  7.31s/it, lr: 1.0e-04 loss: 4.737e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2620/30000 [5:54:10<55:33:57,  7.31s/it, lr: 1.0e-04 loss: 4.737e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2621/30000 [5:54:10<41:32:44,  5.46s/it, lr: 1.0e-04 loss: 4.737e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2621/30000 [5:54:10<41:32:44,  5.46s/it, lr: 1.0e-04 loss: 4.737e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2621/30000 [5:54:19<41:32:44,  5.46s/it, lr: 1.0e-04 loss: 6.194e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2621/30000 [5:54:19<41:32:44,  5.46s/it, lr: 1.0e-04 loss: 6.194e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2622/30000 [5:54:19<48:41:50,  6.40s/it, lr: 1.0e-04 loss: 6.194e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2622/30000 [5:54:19<48:41:50,  6.40s/it, lr: 1.0e-04 loss: 6.194e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2622/30000 [5:54:29<48:41:50,  6.40s/it, lr: 1.0e-04 loss: 6.428e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2622/30000 [5:54:29<48:41:50,  6.40s/it, lr: 1.0e-04 loss: 6.428e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2623/30000 [5:54:29<54:27:31,  7.16s/it, lr: 1.0e-04 loss: 6.428e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2623/30000 [5:54:29<54:27:31,  7.16s/it, lr: 1.0e-04 loss: 6.428e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2623/30000 [5:54:38<54:27:31,  7.16s/it, lr: 1.0e-04 loss: 4.922e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2623/30000 [5:54:38<54:27:31,  7.16s/it, lr: 1.0e-04 loss: 4.922e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2624/30000 [5:54:38<58:54:46,  7.75s/it, lr: 1.0e-04 loss: 4.922e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2624/30000 [5:54:38<58:54:46,  7.75s/it, lr: 1.0e-04 loss: 4.922e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2624/30000 [5:54:45<58:54:46,  7.75s/it, lr: 1.0e-04 loss: 5.266e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2624/30000 [5:54:45<58:54:46,  7.75s/it, lr: 1.0e-04 loss: 5.266e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2625/30000 [5:54:45<56:24:14,  7.42s/it, lr: 1.0e-04 loss: 5.266e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2625/30000 [5:54:45<56:24:14,  7.42s/it, lr: 1.0e-04 loss: 5.266e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2625/30000 [5:54:54<56:24:14,  7.42s/it, lr: 1.0e-04 loss: 6.350e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2625/30000 [5:54:54<56:24:14,  7.42s/it, lr: 1.0e-04 loss: 6.350e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2626/30000 [5:54:54<60:32:36,  7.96s/it, lr: 1.0e-04 loss: 6.350e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2626/30000 [5:54:54<60:32:36,  7.96s/it, lr: 1.0e-04 loss: 6.350e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2626/30000 [5:55:03<60:32:36,  7.96s/it, lr: 1.0e-04 loss: 3.154e-01]
Xiang_wan_2_2_lora:   9%|###4                                   | 2626/30000 [5:55:03<60:32:36,  7.96s/it, lr: 1.0e-04 loss: 3.154e-01]
Xiang_wan_2_2_lora:   9%|###4                                   | 2627/30000 [5:55:03<63:54:46,  8.41s/it, lr: 1.0e-04 loss: 3.154e-01]
Xiang_wan_2_2_lora:   9%|###4                                   | 2627/30000 [5:55:03<63:54:46,  8.41s/it, lr: 1.0e-04 loss: 3.154e-01]
Xiang_wan_2_2_lora:   9%|###4                                   | 2627/30000 [5:55:13<63:54:46,  8.41s/it, lr: 1.0e-04 loss: 6.655e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2627/30000 [5:55:13<63:54:46,  8.41s/it, lr: 1.0e-04 loss: 6.655e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2628/30000 [5:55:13<66:01:08,  8.68s/it, lr: 1.0e-04 loss: 6.655e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2628/30000 [5:55:13<66:01:08,  8.68s/it, lr: 1.0e-04 loss: 6.655e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2628/30000 [5:55:19<66:01:08,  8.68s/it, lr: 1.0e-04 loss: 6.520e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2628/30000 [5:55:19<66:01:08,  8.68s/it, lr: 1.0e-04 loss: 6.520e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2629/30000 [5:55:19<61:23:51,  8.08s/it, lr: 1.0e-04 loss: 6.520e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2629/30000 [5:55:19<61:23:51,  8.08s/it, lr: 1.0e-04 loss: 6.520e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2629/30000 [5:55:26<61:23:51,  8.08s/it, lr: 1.0e-04 loss: 9.571e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2629/30000 [5:55:26<61:23:51,  8.08s/it, lr: 1.0e-04 loss: 9.571e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2630/30000 [5:55:33<61:23:43,  8.08s/it, lr: 1.0e-04 loss: 4.904e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2630/30000 [5:55:33<61:23:43,  8.08s/it, lr: 1.0e-04 loss: 4.904e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2631/30000 [5:55:33<44:43:38,  5.88s/it, lr: 1.0e-04 loss: 4.904e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2631/30000 [5:55:33<44:43:38,  5.88s/it, lr: 1.0e-04 loss: 4.904e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2631/30000 [5:55:39<44:43:38,  5.88s/it, lr: 1.0e-04 loss: 8.796e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2631/30000 [5:55:39<44:43:38,  5.88s/it, lr: 1.0e-04 loss: 8.796e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2632/30000 [5:55:39<45:53:30,  6.04s/it, lr: 1.0e-04 loss: 8.796e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2632/30000 [5:55:39<45:53:30,  6.04s/it, lr: 1.0e-04 loss: 8.796e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2632/30000 [5:55:46<45:53:30,  6.04s/it, lr: 1.0e-04 loss: 7.635e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2632/30000 [5:55:46<45:53:30,  6.04s/it, lr: 1.0e-04 loss: 7.635e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2633/30000 [5:55:46<46:47:32,  6.16s/it, lr: 1.0e-04 loss: 7.635e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2633/30000 [5:55:46<46:47:32,  6.16s/it, lr: 1.0e-04 loss: 7.635e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2633/30000 [5:55:55<46:47:32,  6.16s/it, lr: 1.0e-04 loss: 5.762e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2633/30000 [5:55:55<46:47:32,  6.16s/it, lr: 1.0e-04 loss: 5.762e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2634/30000 [5:55:55<53:12:32,  7.00s/it, lr: 1.0e-04 loss: 5.762e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2634/30000 [5:55:55<53:12:32,  7.00s/it, lr: 1.0e-04 loss: 5.762e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2634/30000 [5:56:02<53:12:32,  7.00s/it, lr: 1.0e-04 loss: 7.682e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2634/30000 [5:56:02<53:12:32,  7.00s/it, lr: 1.0e-04 loss: 7.682e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2635/30000 [5:56:02<52:52:27,  6.96s/it, lr: 1.0e-04 loss: 7.682e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2635/30000 [5:56:02<52:52:27,  6.96s/it, lr: 1.0e-04 loss: 7.682e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2635/30000 [5:56:11<52:52:27,  6.96s/it, lr: 1.0e-04 loss: 6.743e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2635/30000 [5:56:11<52:52:27,  6.96s/it, lr: 1.0e-04 loss: 6.743e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2636/30000 [5:56:11<58:03:03,  7.64s/it, lr: 1.0e-04 loss: 6.743e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2636/30000 [5:56:11<58:03:03,  7.64s/it, lr: 1.0e-04 loss: 6.743e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2636/30000 [5:56:20<58:03:03,  7.64s/it, lr: 1.0e-04 loss: 1.706e-01]
Xiang_wan_2_2_lora:   9%|###4                                   | 2636/30000 [5:56:20<58:03:03,  7.64s/it, lr: 1.0e-04 loss: 1.706e-01]
Xiang_wan_2_2_lora:   9%|###4                                   | 2637/30000 [5:56:20<61:52:33,  8.14s/it, lr: 1.0e-04 loss: 1.706e-01]
Xiang_wan_2_2_lora:   9%|###4                                   | 2637/30000 [5:56:20<61:52:33,  8.14s/it, lr: 1.0e-04 loss: 1.706e-01]
Xiang_wan_2_2_lora:   9%|###4                                   | 2637/30000 [5:56:30<61:52:33,  8.14s/it, lr: 1.0e-04 loss: 6.726e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2637/30000 [5:56:30<61:52:33,  8.14s/it, lr: 1.0e-04 loss: 6.726e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2638/30000 [5:56:30<64:25:01,  8.48s/it, lr: 1.0e-04 loss: 6.726e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2638/30000 [5:56:30<64:25:01,  8.48s/it, lr: 1.0e-04 loss: 6.726e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2638/30000 [5:56:36<64:25:01,  8.48s/it, lr: 1.0e-04 loss: 6.766e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2638/30000 [5:56:36<64:25:01,  8.48s/it, lr: 1.0e-04 loss: 6.766e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2639/30000 [5:56:36<60:00:19,  7.90s/it, lr: 1.0e-04 loss: 6.766e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2639/30000 [5:56:36<60:00:19,  7.90s/it, lr: 1.0e-04 loss: 6.766e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2639/30000 [5:56:43<60:00:19,  7.90s/it, lr: 1.0e-04 loss: 8.539e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2639/30000 [5:56:43<60:00:19,  7.90s/it, lr: 1.0e-04 loss: 8.539e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2640/30000 [5:56:49<60:00:12,  7.90s/it, lr: 1.0e-04 loss: 3.865e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2640/30000 [5:56:49<60:00:12,  7.90s/it, lr: 1.0e-04 loss: 3.865e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2641/30000 [5:56:49<43:53:19,  5.78s/it, lr: 1.0e-04 loss: 3.865e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2641/30000 [5:56:49<43:53:19,  5.78s/it, lr: 1.0e-04 loss: 3.865e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2641/30000 [5:56:59<43:53:19,  5.78s/it, lr: 1.0e-04 loss: 9.357e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2641/30000 [5:56:59<43:53:19,  5.78s/it, lr: 1.0e-04 loss: 9.357e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2642/30000 [5:56:59<50:47:28,  6.68s/it, lr: 1.0e-04 loss: 9.357e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2642/30000 [5:56:59<50:47:28,  6.68s/it, lr: 1.0e-04 loss: 9.357e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2642/30000 [5:57:05<50:47:28,  6.68s/it, lr: 1.0e-04 loss: 4.684e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2642/30000 [5:57:05<50:47:28,  6.68s/it, lr: 1.0e-04 loss: 4.684e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2643/30000 [5:57:05<50:33:03,  6.65s/it, lr: 1.0e-04 loss: 4.684e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2643/30000 [5:57:05<50:33:03,  6.65s/it, lr: 1.0e-04 loss: 4.684e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2643/30000 [5:57:12<50:33:03,  6.65s/it, lr: 1.0e-04 loss: 7.482e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2643/30000 [5:57:12<50:33:03,  6.65s/it, lr: 1.0e-04 loss: 7.482e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2644/30000 [5:57:12<50:16:29,  6.62s/it, lr: 1.0e-04 loss: 7.482e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2644/30000 [5:57:12<50:16:29,  6.62s/it, lr: 1.0e-04 loss: 7.482e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2644/30000 [5:57:19<50:16:29,  6.62s/it, lr: 1.0e-04 loss: 6.695e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2644/30000 [5:57:19<50:16:29,  6.62s/it, lr: 1.0e-04 loss: 6.695e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2645/30000 [5:57:19<50:21:04,  6.63s/it, lr: 1.0e-04 loss: 6.695e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2645/30000 [5:57:19<50:21:04,  6.63s/it, lr: 1.0e-04 loss: 6.695e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2645/30000 [5:57:25<50:21:04,  6.63s/it, lr: 1.0e-04 loss: 8.713e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2645/30000 [5:57:25<50:21:04,  6.63s/it, lr: 1.0e-04 loss: 8.713e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2646/30000 [5:57:25<50:08:59,  6.60s/it, lr: 1.0e-04 loss: 8.713e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2646/30000 [5:57:25<50:08:59,  6.60s/it, lr: 1.0e-04 loss: 8.713e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2646/30000 [5:57:34<50:08:59,  6.60s/it, lr: 1.0e-04 loss: 6.562e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2646/30000 [5:57:34<50:08:59,  6.60s/it, lr: 1.0e-04 loss: 6.562e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2647/30000 [5:57:34<56:06:25,  7.38s/it, lr: 1.0e-04 loss: 6.562e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2647/30000 [5:57:34<56:06:25,  7.38s/it, lr: 1.0e-04 loss: 6.562e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2647/30000 [5:57:41<56:06:25,  7.38s/it, lr: 1.0e-04 loss: 7.104e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2647/30000 [5:57:41<56:06:25,  7.38s/it, lr: 1.0e-04 loss: 7.104e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2648/30000 [5:57:41<54:20:28,  7.15s/it, lr: 1.0e-04 loss: 7.104e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2648/30000 [5:57:41<54:20:28,  7.15s/it, lr: 1.0e-04 loss: 7.104e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2648/30000 [5:57:48<54:20:28,  7.15s/it, lr: 1.0e-04 loss: 5.703e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2648/30000 [5:57:48<54:20:28,  7.15s/it, lr: 1.0e-04 loss: 5.703e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2649/30000 [5:57:48<53:05:44,  6.99s/it, lr: 1.0e-04 loss: 5.703e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2649/30000 [5:57:48<53:05:44,  6.99s/it, lr: 1.0e-04 loss: 5.703e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2649/30000 [5:57:57<53:05:44,  6.99s/it, lr: 1.0e-04 loss: 9.181e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2649/30000 [5:57:57<53:05:44,  6.99s/it, lr: 1.0e-04 loss: 9.181e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2650/30000 [5:58:06<53:05:37,  6.99s/it, lr: 1.0e-04 loss: 7.416e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2650/30000 [5:58:06<53:05:37,  6.99s/it, lr: 1.0e-04 loss: 7.416e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2651/30000 [5:58:06<45:01:27,  5.93s/it, lr: 1.0e-04 loss: 7.416e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2651/30000 [5:58:06<45:01:27,  5.93s/it, lr: 1.0e-04 loss: 7.416e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2651/30000 [5:58:13<45:01:27,  5.93s/it, lr: 1.0e-04 loss: 1.147e-01]
Xiang_wan_2_2_lora:   9%|###4                                   | 2651/30000 [5:58:13<45:01:27,  5.93s/it, lr: 1.0e-04 loss: 1.147e-01]
Xiang_wan_2_2_lora:   9%|###4                                   | 2652/30000 [5:58:13<46:19:46,  6.10s/it, lr: 1.0e-04 loss: 1.147e-01]
Xiang_wan_2_2_lora:   9%|###4                                   | 2652/30000 [5:58:13<46:19:46,  6.10s/it, lr: 1.0e-04 loss: 1.147e-01]
Xiang_wan_2_2_lora:   9%|###4                                   | 2652/30000 [5:58:22<46:19:46,  6.10s/it, lr: 1.0e-04 loss: 1.798e-01]
Xiang_wan_2_2_lora:   9%|###4                                   | 2652/30000 [5:58:22<46:19:46,  6.10s/it, lr: 1.0e-04 loss: 1.798e-01]
Xiang_wan_2_2_lora:   9%|###4                                   | 2653/30000 [5:58:22<52:37:33,  6.93s/it, lr: 1.0e-04 loss: 1.798e-01]
Xiang_wan_2_2_lora:   9%|###4                                   | 2653/30000 [5:58:22<52:37:33,  6.93s/it, lr: 1.0e-04 loss: 1.798e-01]
Xiang_wan_2_2_lora:   9%|###4                                   | 2653/30000 [5:58:32<52:37:33,  6.93s/it, lr: 1.0e-04 loss: 5.242e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2653/30000 [5:58:32<52:37:33,  6.93s/it, lr: 1.0e-04 loss: 5.242e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2654/30000 [5:58:32<57:35:22,  7.58s/it, lr: 1.0e-04 loss: 5.242e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2654/30000 [5:58:32<57:35:22,  7.58s/it, lr: 1.0e-04 loss: 5.242e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2654/30000 [5:58:41<57:35:22,  7.58s/it, lr: 1.0e-04 loss: 8.401e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2654/30000 [5:58:41<57:35:22,  7.58s/it, lr: 1.0e-04 loss: 8.401e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2655/30000 [5:58:41<61:17:39,  8.07s/it, lr: 1.0e-04 loss: 8.401e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2655/30000 [5:58:41<61:17:39,  8.07s/it, lr: 1.0e-04 loss: 8.401e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2655/30000 [5:58:50<61:17:39,  8.07s/it, lr: 1.0e-04 loss: 8.419e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2655/30000 [5:58:50<61:17:39,  8.07s/it, lr: 1.0e-04 loss: 8.419e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2656/30000 [5:58:50<64:06:32,  8.44s/it, lr: 1.0e-04 loss: 8.419e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2656/30000 [5:58:50<64:06:32,  8.44s/it, lr: 1.0e-04 loss: 8.419e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2656/30000 [5:58:57<64:06:32,  8.44s/it, lr: 1.0e-04 loss: 7.190e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2656/30000 [5:58:57<64:06:32,  8.44s/it, lr: 1.0e-04 loss: 7.190e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2657/30000 [5:58:57<60:32:17,  7.97s/it, lr: 1.0e-04 loss: 7.190e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2657/30000 [5:58:57<60:32:17,  7.97s/it, lr: 1.0e-04 loss: 7.190e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2657/30000 [5:59:04<60:32:17,  7.97s/it, lr: 1.0e-04 loss: 9.019e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2657/30000 [5:59:04<60:32:17,  7.97s/it, lr: 1.0e-04 loss: 9.019e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2658/30000 [5:59:04<57:30:54,  7.57s/it, lr: 1.0e-04 loss: 9.019e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2658/30000 [5:59:04<57:30:54,  7.57s/it, lr: 1.0e-04 loss: 9.019e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2658/30000 [5:59:13<57:30:54,  7.57s/it, lr: 1.0e-04 loss: 3.513e-01]
Xiang_wan_2_2_lora:   9%|###4                                   | 2658/30000 [5:59:13<57:30:54,  7.57s/it, lr: 1.0e-04 loss: 3.513e-01]
Xiang_wan_2_2_lora:   9%|###4                                   | 2659/30000 [5:59:13<61:24:09,  8.08s/it, lr: 1.0e-04 loss: 3.513e-01]
Xiang_wan_2_2_lora:   9%|###4                                   | 2659/30000 [5:59:13<61:24:09,  8.08s/it, lr: 1.0e-04 loss: 3.513e-01]
Xiang_wan_2_2_lora:   9%|###4                                   | 2659/30000 [5:59:20<61:24:09,  8.08s/it, lr: 1.0e-04 loss: 6.951e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2659/30000 [5:59:20<61:24:09,  8.08s/it, lr: 1.0e-04 loss: 6.951e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2660/30000 [5:59:29<61:24:01,  8.08s/it, lr: 1.0e-04 loss: 8.792e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2660/30000 [5:59:29<61:24:01,  8.08s/it, lr: 1.0e-04 loss: 8.792e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2661/30000 [5:59:29<49:25:43,  6.51s/it, lr: 1.0e-04 loss: 8.792e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2661/30000 [5:59:29<49:25:43,  6.51s/it, lr: 1.0e-04 loss: 8.792e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2661/30000 [5:59:38<49:25:43,  6.51s/it, lr: 1.0e-04 loss: 5.743e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2661/30000 [5:59:38<49:25:43,  6.51s/it, lr: 1.0e-04 loss: 5.743e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2662/30000 [5:59:38<54:51:10,  7.22s/it, lr: 1.0e-04 loss: 5.743e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2662/30000 [5:59:38<54:51:10,  7.22s/it, lr: 1.0e-04 loss: 5.743e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2662/30000 [5:59:45<54:51:10,  7.22s/it, lr: 1.0e-04 loss: 1.534e-01]
Xiang_wan_2_2_lora:   9%|###4                                   | 2662/30000 [5:59:45<54:51:10,  7.22s/it, lr: 1.0e-04 loss: 1.534e-01]
Xiang_wan_2_2_lora:   9%|###4                                   | 2663/30000 [5:59:45<53:34:40,  7.06s/it, lr: 1.0e-04 loss: 1.534e-01]
Xiang_wan_2_2_lora:   9%|###4                                   | 2663/30000 [5:59:45<53:34:40,  7.06s/it, lr: 1.0e-04 loss: 1.534e-01]
Xiang_wan_2_2_lora:   9%|###4                                   | 2663/30000 [5:59:54<53:34:40,  7.06s/it, lr: 1.0e-04 loss: 7.140e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2663/30000 [5:59:54<53:34:40,  7.06s/it, lr: 1.0e-04 loss: 7.140e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2664/30000 [5:59:54<58:20:56,  7.68s/it, lr: 1.0e-04 loss: 7.140e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2664/30000 [5:59:54<58:20:56,  7.68s/it, lr: 1.0e-04 loss: 7.140e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2664/30000 [6:00:04<58:20:56,  7.68s/it, lr: 1.0e-04 loss: 6.598e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2664/30000 [6:00:04<58:20:56,  7.68s/it, lr: 1.0e-04 loss: 6.598e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2665/30000 [6:00:04<62:16:53,  8.20s/it, lr: 1.0e-04 loss: 6.598e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2665/30000 [6:00:04<62:16:53,  8.20s/it, lr: 1.0e-04 loss: 6.598e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2665/30000 [6:00:13<62:16:53,  8.20s/it, lr: 1.0e-04 loss: 6.949e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2665/30000 [6:00:13<62:16:53,  8.20s/it, lr: 1.0e-04 loss: 6.949e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2666/30000 [6:00:13<64:48:40,  8.54s/it, lr: 1.0e-04 loss: 6.949e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2666/30000 [6:00:13<64:48:40,  8.54s/it, lr: 1.0e-04 loss: 6.949e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2666/30000 [6:00:20<64:48:40,  8.54s/it, lr: 1.0e-04 loss: 6.615e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2666/30000 [6:00:20<64:48:40,  8.54s/it, lr: 1.0e-04 loss: 6.615e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2667/30000 [6:00:20<60:28:40,  7.97s/it, lr: 1.0e-04 loss: 6.615e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2667/30000 [6:00:20<60:28:40,  7.97s/it, lr: 1.0e-04 loss: 6.615e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2667/30000 [6:00:29<60:28:40,  7.97s/it, lr: 1.0e-04 loss: 5.255e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2667/30000 [6:00:29<60:28:40,  7.97s/it, lr: 1.0e-04 loss: 5.255e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2668/30000 [6:00:29<63:34:21,  8.37s/it, lr: 1.0e-04 loss: 5.255e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2668/30000 [6:00:29<63:34:21,  8.37s/it, lr: 1.0e-04 loss: 5.255e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2668/30000 [6:00:36<63:34:21,  8.37s/it, lr: 1.0e-04 loss: 5.689e-01]
Xiang_wan_2_2_lora:   9%|###4                                   | 2668/30000 [6:00:36<63:34:21,  8.37s/it, lr: 1.0e-04 loss: 5.689e-01]
Xiang_wan_2_2_lora:   9%|###4                                   | 2669/30000 [6:00:36<59:32:57,  7.84s/it, lr: 1.0e-04 loss: 5.689e-01]
Xiang_wan_2_2_lora:   9%|###4                                   | 2669/30000 [6:00:36<59:32:57,  7.84s/it, lr: 1.0e-04 loss: 5.689e-01]
Xiang_wan_2_2_lora:   9%|###4                                   | 2669/30000 [6:00:42<59:32:57,  7.84s/it, lr: 1.0e-04 loss: 7.892e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2669/30000 [6:00:42<59:32:57,  7.84s/it, lr: 1.0e-04 loss: 7.892e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2670/30000 [6:00:49<59:32:50,  7.84s/it, lr: 1.0e-04 loss: 7.576e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2670/30000 [6:00:49<59:32:50,  7.84s/it, lr: 1.0e-04 loss: 7.576e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2671/30000 [6:00:49<43:39:31,  5.75s/it, lr: 1.0e-04 loss: 7.576e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2671/30000 [6:00:49<43:39:31,  5.75s/it, lr: 1.0e-04 loss: 7.576e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2671/30000 [6:00:56<43:39:31,  5.75s/it, lr: 1.0e-04 loss: 9.563e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2671/30000 [6:00:56<43:39:31,  5.75s/it, lr: 1.0e-04 loss: 9.563e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2672/30000 [6:00:56<45:40:13,  6.02s/it, lr: 1.0e-04 loss: 9.563e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2672/30000 [6:00:56<45:40:13,  6.02s/it, lr: 1.0e-04 loss: 9.563e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2672/30000 [6:01:02<45:40:13,  6.02s/it, lr: 1.0e-04 loss: 7.052e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2672/30000 [6:01:02<45:40:13,  6.02s/it, lr: 1.0e-04 loss: 7.052e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2673/30000 [6:01:02<46:35:34,  6.14s/it, lr: 1.0e-04 loss: 7.052e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2673/30000 [6:01:02<46:35:34,  6.14s/it, lr: 1.0e-04 loss: 7.052e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2673/30000 [6:01:09<46:35:34,  6.14s/it, lr: 1.0e-04 loss: 1.048e-01]
Xiang_wan_2_2_lora:   9%|###4                                   | 2673/30000 [6:01:09<46:35:34,  6.14s/it, lr: 1.0e-04 loss: 1.048e-01]
Xiang_wan_2_2_lora:   9%|###4                                   | 2674/30000 [6:01:09<47:23:39,  6.24s/it, lr: 1.0e-04 loss: 1.048e-01]
Xiang_wan_2_2_lora:   9%|###4                                   | 2674/30000 [6:01:09<47:23:39,  6.24s/it, lr: 1.0e-04 loss: 1.048e-01]
Xiang_wan_2_2_lora:   9%|###4                                   | 2674/30000 [6:01:15<47:23:39,  6.24s/it, lr: 1.0e-04 loss: 9.718e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2674/30000 [6:01:15<47:23:39,  6.24s/it, lr: 1.0e-04 loss: 9.718e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2675/30000 [6:01:15<48:09:48,  6.35s/it, lr: 1.0e-04 loss: 9.718e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2675/30000 [6:01:15<48:09:48,  6.35s/it, lr: 1.0e-04 loss: 9.718e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2675/30000 [6:01:22<48:09:48,  6.35s/it, lr: 1.0e-04 loss: 7.842e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2675/30000 [6:01:22<48:09:48,  6.35s/it, lr: 1.0e-04 loss: 7.842e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2676/30000 [6:01:22<48:33:39,  6.40s/it, lr: 1.0e-04 loss: 7.842e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2676/30000 [6:01:22<48:33:39,  6.40s/it, lr: 1.0e-04 loss: 7.842e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2676/30000 [6:01:28<48:33:39,  6.40s/it, lr: 1.0e-04 loss: 7.088e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2676/30000 [6:01:28<48:33:39,  6.40s/it, lr: 1.0e-04 loss: 7.088e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2677/30000 [6:01:28<48:45:19,  6.42s/it, lr: 1.0e-04 loss: 7.088e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2677/30000 [6:01:28<48:45:19,  6.42s/it, lr: 1.0e-04 loss: 7.088e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2677/30000 [6:01:38<48:45:19,  6.42s/it, lr: 1.0e-04 loss: 6.373e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2677/30000 [6:01:38<48:45:19,  6.42s/it, lr: 1.0e-04 loss: 6.373e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2678/30000 [6:01:38<55:14:46,  7.28s/it, lr: 1.0e-04 loss: 6.373e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2678/30000 [6:01:38<55:14:46,  7.28s/it, lr: 1.0e-04 loss: 6.373e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2678/30000 [6:01:44<55:14:46,  7.28s/it, lr: 1.0e-04 loss: 6.454e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2678/30000 [6:01:44<55:14:46,  7.28s/it, lr: 1.0e-04 loss: 6.454e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2679/30000 [6:01:44<53:30:01,  7.05s/it, lr: 1.0e-04 loss: 6.454e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2679/30000 [6:01:44<53:30:01,  7.05s/it, lr: 1.0e-04 loss: 6.454e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2679/30000 [6:01:54<53:30:01,  7.05s/it, lr: 1.0e-04 loss: 5.998e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2679/30000 [6:01:54<53:30:01,  7.05s/it, lr: 1.0e-04 loss: 5.998e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2680/30000 [6:02:00<53:29:54,  7.05s/it, lr: 1.0e-04 loss: 6.797e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2680/30000 [6:02:00<53:29:54,  7.05s/it, lr: 1.0e-04 loss: 6.797e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2681/30000 [6:02:00<40:17:50,  5.31s/it, lr: 1.0e-04 loss: 6.797e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2681/30000 [6:02:00<40:17:50,  5.31s/it, lr: 1.0e-04 loss: 6.797e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2681/30000 [6:02:07<40:17:50,  5.31s/it, lr: 1.0e-04 loss: 8.800e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2681/30000 [6:02:07<40:17:50,  5.31s/it, lr: 1.0e-04 loss: 8.800e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2682/30000 [6:02:07<42:39:08,  5.62s/it, lr: 1.0e-04 loss: 8.800e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2682/30000 [6:02:07<42:39:08,  5.62s/it, lr: 1.0e-04 loss: 8.800e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2682/30000 [6:02:13<42:39:08,  5.62s/it, lr: 1.0e-04 loss: 1.906e-01]
Xiang_wan_2_2_lora:   9%|###4                                   | 2682/30000 [6:02:13<42:39:08,  5.62s/it, lr: 1.0e-04 loss: 1.906e-01]
Xiang_wan_2_2_lora:   9%|###4                                   | 2683/30000 [6:02:13<44:29:42,  5.86s/it, lr: 1.0e-04 loss: 1.906e-01]
Xiang_wan_2_2_lora:   9%|###4                                   | 2683/30000 [6:02:13<44:29:42,  5.86s/it, lr: 1.0e-04 loss: 1.906e-01]
Xiang_wan_2_2_lora:   9%|###4                                   | 2683/30000 [6:02:23<44:29:42,  5.86s/it, lr: 1.0e-04 loss: 8.573e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2683/30000 [6:02:23<44:29:42,  5.86s/it, lr: 1.0e-04 loss: 8.573e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2684/30000 [6:02:23<51:37:07,  6.80s/it, lr: 1.0e-04 loss: 8.573e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2684/30000 [6:02:23<51:37:07,  6.80s/it, lr: 1.0e-04 loss: 8.573e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2684/30000 [6:02:32<51:37:07,  6.80s/it, lr: 1.0e-04 loss: 1.153e-01]
Xiang_wan_2_2_lora:   9%|###4                                   | 2684/30000 [6:02:32<51:37:07,  6.80s/it, lr: 1.0e-04 loss: 1.153e-01]
Xiang_wan_2_2_lora:   9%|###4                                   | 2685/30000 [6:02:32<57:04:48,  7.52s/it, lr: 1.0e-04 loss: 1.153e-01]
Xiang_wan_2_2_lora:   9%|###4                                   | 2685/30000 [6:02:32<57:04:48,  7.52s/it, lr: 1.0e-04 loss: 1.153e-01]
Xiang_wan_2_2_lora:   9%|###4                                   | 2685/30000 [6:02:41<57:04:48,  7.52s/it, lr: 1.0e-04 loss: 6.440e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2685/30000 [6:02:41<57:04:48,  7.52s/it, lr: 1.0e-04 loss: 6.440e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2686/30000 [6:02:41<61:04:50,  8.05s/it, lr: 1.0e-04 loss: 6.440e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2686/30000 [6:02:41<61:04:50,  8.05s/it, lr: 1.0e-04 loss: 6.440e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2686/30000 [6:02:48<61:04:50,  8.05s/it, lr: 1.0e-04 loss: 7.354e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2686/30000 [6:02:48<61:04:50,  8.05s/it, lr: 1.0e-04 loss: 7.354e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2687/30000 [6:02:48<58:22:28,  7.69s/it, lr: 1.0e-04 loss: 7.354e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2687/30000 [6:02:48<58:22:28,  7.69s/it, lr: 1.0e-04 loss: 7.354e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2687/30000 [6:02:55<58:22:28,  7.69s/it, lr: 1.0e-04 loss: 6.432e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2687/30000 [6:02:55<58:22:28,  7.69s/it, lr: 1.0e-04 loss: 6.432e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2688/30000 [6:02:55<55:50:48,  7.36s/it, lr: 1.0e-04 loss: 6.432e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2688/30000 [6:02:55<55:50:48,  7.36s/it, lr: 1.0e-04 loss: 6.432e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2688/30000 [6:03:01<55:50:48,  7.36s/it, lr: 1.0e-04 loss: 6.682e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2688/30000 [6:03:01<55:50:48,  7.36s/it, lr: 1.0e-04 loss: 6.682e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2689/30000 [6:03:01<54:01:24,  7.12s/it, lr: 1.0e-04 loss: 6.682e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2689/30000 [6:03:01<54:01:24,  7.12s/it, lr: 1.0e-04 loss: 6.682e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2689/30000 [6:03:11<54:01:24,  7.12s/it, lr: 1.0e-04 loss: 8.092e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2689/30000 [6:03:11<54:01:24,  7.12s/it, lr: 1.0e-04 loss: 8.092e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2690/30000 [6:03:20<54:01:16,  7.12s/it, lr: 1.0e-04 loss: 4.795e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2690/30000 [6:03:20<54:01:16,  7.12s/it, lr: 1.0e-04 loss: 4.795e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2691/30000 [6:03:20<45:32:22,  6.00s/it, lr: 1.0e-04 loss: 4.795e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2691/30000 [6:03:20<45:32:22,  6.00s/it, lr: 1.0e-04 loss: 4.795e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2691/30000 [6:03:27<45:32:22,  6.00s/it, lr: 1.0e-04 loss: 8.675e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2691/30000 [6:03:27<45:32:22,  6.00s/it, lr: 1.0e-04 loss: 8.675e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2692/30000 [6:03:27<46:39:30,  6.15s/it, lr: 1.0e-04 loss: 8.675e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2692/30000 [6:03:27<46:39:30,  6.15s/it, lr: 1.0e-04 loss: 8.675e-02]
Xiang_wan_2_2_lora:   9%|###4                                   | 2692/30000 [6:03:33<46:39:30,  6.15s/it, lr: 1.0e-04 loss: 1.115e-01]
Xiang_wan_2_2_lora:   9%|###4                                   | 2692/30000 [6:03:33<46:39:30,  6.15s/it, lr: 1.0e-04 loss: 1.115e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2693/30000 [6:03:33<47:18:28,  6.24s/it, lr: 1.0e-04 loss: 1.115e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2693/30000 [6:03:33<47:18:28,  6.24s/it, lr: 1.0e-04 loss: 1.115e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2693/30000 [6:03:42<47:18:28,  6.24s/it, lr: 1.0e-04 loss: 8.771e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2693/30000 [6:03:42<47:18:28,  6.24s/it, lr: 1.0e-04 loss: 8.771e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2694/30000 [6:03:42<53:40:14,  7.08s/it, lr: 1.0e-04 loss: 8.771e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2694/30000 [6:03:42<53:40:14,  7.08s/it, lr: 1.0e-04 loss: 8.771e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2694/30000 [6:03:52<53:40:14,  7.08s/it, lr: 1.0e-04 loss: 4.819e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2694/30000 [6:03:52<53:40:14,  7.08s/it, lr: 1.0e-04 loss: 4.819e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2695/30000 [6:03:52<59:04:26,  7.79s/it, lr: 1.0e-04 loss: 4.819e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2695/30000 [6:03:52<59:04:26,  7.79s/it, lr: 1.0e-04 loss: 4.819e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2695/30000 [6:03:59<59:04:26,  7.79s/it, lr: 1.0e-04 loss: 7.359e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2695/30000 [6:03:59<59:04:26,  7.79s/it, lr: 1.0e-04 loss: 7.359e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2696/30000 [6:03:59<56:32:56,  7.46s/it, lr: 1.0e-04 loss: 7.359e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2696/30000 [6:03:59<56:32:56,  7.46s/it, lr: 1.0e-04 loss: 7.359e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2696/30000 [6:04:08<56:32:56,  7.46s/it, lr: 1.0e-04 loss: 1.767e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2696/30000 [6:04:08<56:32:56,  7.46s/it, lr: 1.0e-04 loss: 1.767e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2697/30000 [6:04:08<60:39:14,  8.00s/it, lr: 1.0e-04 loss: 1.767e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2697/30000 [6:04:08<60:39:14,  8.00s/it, lr: 1.0e-04 loss: 1.767e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2697/30000 [6:04:15<60:39:14,  8.00s/it, lr: 1.0e-04 loss: 8.649e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2697/30000 [6:04:15<60:39:14,  8.00s/it, lr: 1.0e-04 loss: 8.649e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2698/30000 [6:04:15<57:37:10,  7.60s/it, lr: 1.0e-04 loss: 8.649e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2698/30000 [6:04:15<57:37:10,  7.60s/it, lr: 1.0e-04 loss: 8.649e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2698/30000 [6:04:21<57:37:10,  7.60s/it, lr: 1.0e-04 loss: 1.170e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2698/30000 [6:04:21<57:37:10,  7.60s/it, lr: 1.0e-04 loss: 1.170e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2699/30000 [6:04:21<55:15:18,  7.29s/it, lr: 1.0e-04 loss: 1.170e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2699/30000 [6:04:21<55:15:18,  7.29s/it, lr: 1.0e-04 loss: 1.170e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2699/30000 [6:04:30<55:15:18,  7.29s/it, lr: 1.0e-04 loss: 6.005e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2699/30000 [6:04:30<55:15:18,  7.29s/it, lr: 1.0e-04 loss: 6.005e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2700/30000 [6:04:37<55:15:11,  7.29s/it, lr: 1.0e-04 loss: 9.107e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2700/30000 [6:04:37<55:15:11,  7.29s/it, lr: 1.0e-04 loss: 9.107e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2701/30000 [6:04:37<41:21:50,  5.45s/it, lr: 1.0e-04 loss: 9.107e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2701/30000 [6:04:37<41:21:50,  5.45s/it, lr: 1.0e-04 loss: 9.107e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2701/30000 [6:04:49<41:21:50,  5.45s/it, lr: 1.0e-04 loss: 8.729e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2701/30000 [6:04:49<41:21:50,  5.45s/it, lr: 1.0e-04 loss: 8.729e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2702/30000 [6:04:49<51:58:32,  6.85s/it, lr: 1.0e-04 loss: 8.729e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2702/30000 [6:04:49<51:58:32,  6.85s/it, lr: 1.0e-04 loss: 8.729e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2702/30000 [6:05:02<51:58:32,  6.85s/it, lr: 1.0e-04 loss: 3.264e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2702/30000 [6:05:02<51:58:32,  6.85s/it, lr: 1.0e-04 loss: 3.264e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2703/30000 [6:05:02<62:21:15,  8.22s/it, lr: 1.0e-04 loss: 3.264e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2703/30000 [6:05:02<62:21:15,  8.22s/it, lr: 1.0e-04 loss: 3.264e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2703/30000 [6:05:09<62:21:15,  8.22s/it, lr: 1.0e-04 loss: 6.704e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2703/30000 [6:05:09<62:21:15,  8.22s/it, lr: 1.0e-04 loss: 6.704e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2704/30000 [6:05:09<60:46:58,  8.02s/it, lr: 1.0e-04 loss: 6.704e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2704/30000 [6:05:09<60:46:58,  8.02s/it, lr: 1.0e-04 loss: 6.704e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2704/30000 [6:05:19<60:46:58,  8.02s/it, lr: 1.0e-04 loss: 5.665e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2704/30000 [6:05:19<60:46:58,  8.02s/it, lr: 1.0e-04 loss: 5.665e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2705/30000 [6:05:19<64:49:40,  8.55s/it, lr: 1.0e-04 loss: 5.665e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2705/30000 [6:05:19<64:49:40,  8.55s/it, lr: 1.0e-04 loss: 5.665e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2705/30000 [6:05:26<64:49:40,  8.55s/it, lr: 1.0e-04 loss: 8.257e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2705/30000 [6:05:26<64:49:40,  8.55s/it, lr: 1.0e-04 loss: 8.257e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2706/30000 [6:05:26<61:19:09,  8.09s/it, lr: 1.0e-04 loss: 8.257e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2706/30000 [6:05:26<61:19:09,  8.09s/it, lr: 1.0e-04 loss: 8.257e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2706/30000 [6:05:33<61:19:09,  8.09s/it, lr: 1.0e-04 loss: 6.634e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2706/30000 [6:05:33<61:19:09,  8.09s/it, lr: 1.0e-04 loss: 6.634e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2707/30000 [6:05:33<58:21:59,  7.70s/it, lr: 1.0e-04 loss: 6.634e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2707/30000 [6:05:33<58:21:59,  7.70s/it, lr: 1.0e-04 loss: 6.634e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2707/30000 [6:05:39<58:21:59,  7.70s/it, lr: 1.0e-04 loss: 7.906e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2707/30000 [6:05:39<58:21:59,  7.70s/it, lr: 1.0e-04 loss: 7.906e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2708/30000 [6:05:39<56:22:02,  7.44s/it, lr: 1.0e-04 loss: 7.906e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2708/30000 [6:05:39<56:22:02,  7.44s/it, lr: 1.0e-04 loss: 7.906e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2708/30000 [6:05:49<56:22:02,  7.44s/it, lr: 1.0e-04 loss: 8.394e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2708/30000 [6:05:49<56:22:02,  7.44s/it, lr: 1.0e-04 loss: 8.394e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2709/30000 [6:05:49<60:52:06,  8.03s/it, lr: 1.0e-04 loss: 8.394e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2709/30000 [6:05:49<60:52:06,  8.03s/it, lr: 1.0e-04 loss: 8.394e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2709/30000 [6:05:58<60:52:06,  8.03s/it, lr: 1.0e-04 loss: 6.771e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2709/30000 [6:05:58<60:52:06,  8.03s/it, lr: 1.0e-04 loss: 6.771e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2710/30000 [6:06:05<60:51:58,  8.03s/it, lr: 1.0e-04 loss: 4.726e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2710/30000 [6:06:05<60:51:58,  8.03s/it, lr: 1.0e-04 loss: 4.726e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2711/30000 [6:06:05<44:37:07,  5.89s/it, lr: 1.0e-04 loss: 4.726e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2711/30000 [6:06:05<44:37:07,  5.89s/it, lr: 1.0e-04 loss: 4.726e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2711/30000 [6:06:13<44:37:07,  5.89s/it, lr: 1.0e-04 loss: 7.982e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2711/30000 [6:06:13<44:37:07,  5.89s/it, lr: 1.0e-04 loss: 7.982e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2712/30000 [6:06:13<48:12:36,  6.36s/it, lr: 1.0e-04 loss: 7.982e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2712/30000 [6:06:13<48:12:36,  6.36s/it, lr: 1.0e-04 loss: 7.982e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2712/30000 [6:06:23<48:12:36,  6.36s/it, lr: 1.0e-04 loss: 2.837e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2712/30000 [6:06:23<48:12:36,  6.36s/it, lr: 1.0e-04 loss: 2.837e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2713/30000 [6:06:23<56:26:25,  7.45s/it, lr: 1.0e-04 loss: 2.837e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2713/30000 [6:06:23<56:26:25,  7.45s/it, lr: 1.0e-04 loss: 2.837e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2713/30000 [6:06:34<56:26:25,  7.45s/it, lr: 1.0e-04 loss: 6.628e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2713/30000 [6:06:34<56:26:25,  7.45s/it, lr: 1.0e-04 loss: 6.628e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2714/30000 [6:06:34<62:33:26,  8.25s/it, lr: 1.0e-04 loss: 6.628e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2714/30000 [6:06:34<62:33:26,  8.25s/it, lr: 1.0e-04 loss: 6.628e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2714/30000 [6:06:41<62:33:26,  8.25s/it, lr: 1.0e-04 loss: 6.559e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2714/30000 [6:06:41<62:33:26,  8.25s/it, lr: 1.0e-04 loss: 6.559e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2715/30000 [6:06:41<61:04:50,  8.06s/it, lr: 1.0e-04 loss: 6.559e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2715/30000 [6:06:41<61:04:50,  8.06s/it, lr: 1.0e-04 loss: 6.559e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2715/30000 [6:06:52<61:04:50,  8.06s/it, lr: 1.0e-04 loss: 5.849e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2715/30000 [6:06:52<61:04:50,  8.06s/it, lr: 1.0e-04 loss: 5.849e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2716/30000 [6:06:52<65:49:29,  8.69s/it, lr: 1.0e-04 loss: 5.849e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2716/30000 [6:06:52<65:49:29,  8.69s/it, lr: 1.0e-04 loss: 5.849e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2716/30000 [6:07:00<65:49:29,  8.69s/it, lr: 1.0e-04 loss: 5.194e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2716/30000 [6:07:00<65:49:29,  8.69s/it, lr: 1.0e-04 loss: 5.194e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2717/30000 [6:07:00<64:51:22,  8.56s/it, lr: 1.0e-04 loss: 5.194e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2717/30000 [6:07:00<64:51:22,  8.56s/it, lr: 1.0e-04 loss: 5.194e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2717/30000 [6:07:10<64:51:22,  8.56s/it, lr: 1.0e-04 loss: 6.654e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2717/30000 [6:07:10<64:51:22,  8.56s/it, lr: 1.0e-04 loss: 6.654e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2718/30000 [6:07:10<68:35:02,  9.05s/it, lr: 1.0e-04 loss: 6.654e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2718/30000 [6:07:10<68:35:02,  9.05s/it, lr: 1.0e-04 loss: 6.654e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2718/30000 [6:07:21<68:35:02,  9.05s/it, lr: 1.0e-04 loss: 3.411e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2718/30000 [6:07:21<68:35:02,  9.05s/it, lr: 1.0e-04 loss: 3.411e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2719/30000 [6:07:21<71:31:19,  9.44s/it, lr: 1.0e-04 loss: 3.411e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2719/30000 [6:07:21<71:31:19,  9.44s/it, lr: 1.0e-04 loss: 3.411e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2719/30000 [6:07:31<71:31:19,  9.44s/it, lr: 1.0e-04 loss: 4.948e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2719/30000 [6:07:31<71:31:19,  9.44s/it, lr: 1.0e-04 loss: 4.948e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2720/30000 [6:07:38<71:31:09,  9.44s/it, lr: 1.0e-04 loss: 8.224e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2720/30000 [6:07:38<71:31:09,  9.44s/it, lr: 1.0e-04 loss: 8.224e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2721/30000 [6:07:38<52:01:01,  6.86s/it, lr: 1.0e-04 loss: 8.224e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2721/30000 [6:07:38<52:01:01,  6.86s/it, lr: 1.0e-04 loss: 8.224e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2721/30000 [6:07:49<52:01:01,  6.86s/it, lr: 1.0e-04 loss: 5.762e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2721/30000 [6:07:49<52:01:01,  6.86s/it, lr: 1.0e-04 loss: 5.762e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2722/30000 [6:07:49<58:17:48,  7.69s/it, lr: 1.0e-04 loss: 5.762e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2722/30000 [6:07:49<58:17:48,  7.69s/it, lr: 1.0e-04 loss: 5.762e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2722/30000 [6:07:59<58:17:48,  7.69s/it, lr: 1.0e-04 loss: 4.591e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2722/30000 [6:07:59<58:17:48,  7.69s/it, lr: 1.0e-04 loss: 4.591e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2723/30000 [6:07:59<63:31:24,  8.38s/it, lr: 1.0e-04 loss: 4.591e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2723/30000 [6:07:59<63:31:24,  8.38s/it, lr: 1.0e-04 loss: 4.591e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2723/30000 [6:08:09<63:31:24,  8.38s/it, lr: 1.0e-04 loss: 7.761e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2723/30000 [6:08:09<63:31:24,  8.38s/it, lr: 1.0e-04 loss: 7.761e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2724/30000 [6:08:09<67:24:50,  8.90s/it, lr: 1.0e-04 loss: 7.761e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2724/30000 [6:08:09<67:24:50,  8.90s/it, lr: 1.0e-04 loss: 7.761e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2724/30000 [6:08:16<67:24:50,  8.90s/it, lr: 1.0e-04 loss: 6.555e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2724/30000 [6:08:16<67:24:50,  8.90s/it, lr: 1.0e-04 loss: 6.555e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2725/30000 [6:08:16<63:15:46,  8.35s/it, lr: 1.0e-04 loss: 6.555e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2725/30000 [6:08:16<63:15:46,  8.35s/it, lr: 1.0e-04 loss: 6.555e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2725/30000 [6:08:23<63:15:46,  8.35s/it, lr: 1.0e-04 loss: 1.028e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2725/30000 [6:08:23<63:15:46,  8.35s/it, lr: 1.0e-04 loss: 1.028e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2726/30000 [6:08:23<59:30:38,  7.86s/it, lr: 1.0e-04 loss: 1.028e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2726/30000 [6:08:23<59:30:38,  7.86s/it, lr: 1.0e-04 loss: 1.028e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2726/30000 [6:08:32<59:30:38,  7.86s/it, lr: 1.0e-04 loss: 3.344e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2726/30000 [6:08:32<59:30:38,  7.86s/it, lr: 1.0e-04 loss: 3.344e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2727/30000 [6:08:32<62:46:51,  8.29s/it, lr: 1.0e-04 loss: 3.344e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2727/30000 [6:08:32<62:46:51,  8.29s/it, lr: 1.0e-04 loss: 3.344e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2727/30000 [6:08:42<62:46:51,  8.29s/it, lr: 1.0e-04 loss: 1.136e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2727/30000 [6:08:42<62:46:51,  8.29s/it, lr: 1.0e-04 loss: 1.136e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2728/30000 [6:08:42<65:10:23,  8.60s/it, lr: 1.0e-04 loss: 1.136e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2728/30000 [6:08:42<65:10:23,  8.60s/it, lr: 1.0e-04 loss: 1.136e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2728/30000 [6:08:51<65:10:23,  8.60s/it, lr: 1.0e-04 loss: 8.068e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2728/30000 [6:08:51<65:10:23,  8.60s/it, lr: 1.0e-04 loss: 8.068e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2729/30000 [6:08:51<66:54:15,  8.83s/it, lr: 1.0e-04 loss: 8.068e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2729/30000 [6:08:51<66:54:15,  8.83s/it, lr: 1.0e-04 loss: 8.068e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2729/30000 [6:09:00<66:54:15,  8.83s/it, lr: 1.0e-04 loss: 5.111e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2729/30000 [6:09:00<66:54:15,  8.83s/it, lr: 1.0e-04 loss: 5.111e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2730/30000 [6:09:07<66:54:06,  8.83s/it, lr: 1.0e-04 loss: 1.002e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2730/30000 [6:09:07<66:54:06,  8.83s/it, lr: 1.0e-04 loss: 1.002e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2731/30000 [6:09:07<47:48:01,  6.31s/it, lr: 1.0e-04 loss: 1.002e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2731/30000 [6:09:07<47:48:01,  6.31s/it, lr: 1.0e-04 loss: 1.002e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2731/30000 [6:09:16<47:48:01,  6.31s/it, lr: 1.0e-04 loss: 8.801e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2731/30000 [6:09:16<47:48:01,  6.31s/it, lr: 1.0e-04 loss: 8.801e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2732/30000 [6:09:16<53:46:14,  7.10s/it, lr: 1.0e-04 loss: 8.801e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2732/30000 [6:09:16<53:46:14,  7.10s/it, lr: 1.0e-04 loss: 8.801e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2732/30000 [6:09:26<53:46:14,  7.10s/it, lr: 1.0e-04 loss: 4.507e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2732/30000 [6:09:26<53:46:14,  7.10s/it, lr: 1.0e-04 loss: 4.507e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2733/30000 [6:09:26<58:16:04,  7.69s/it, lr: 1.0e-04 loss: 4.507e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2733/30000 [6:09:26<58:16:04,  7.69s/it, lr: 1.0e-04 loss: 4.507e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2733/30000 [6:09:32<58:16:04,  7.69s/it, lr: 1.0e-04 loss: 9.958e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2733/30000 [6:09:32<58:16:04,  7.69s/it, lr: 1.0e-04 loss: 9.958e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2734/30000 [6:09:32<55:53:28,  7.38s/it, lr: 1.0e-04 loss: 9.958e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2734/30000 [6:09:32<55:53:28,  7.38s/it, lr: 1.0e-04 loss: 9.958e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2734/30000 [6:09:39<55:53:28,  7.38s/it, lr: 1.0e-04 loss: 6.498e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2734/30000 [6:09:39<55:53:28,  7.38s/it, lr: 1.0e-04 loss: 6.498e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2735/30000 [6:09:39<54:05:04,  7.14s/it, lr: 1.0e-04 loss: 6.498e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2735/30000 [6:09:39<54:05:04,  7.14s/it, lr: 1.0e-04 loss: 6.498e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2735/30000 [6:09:48<54:05:04,  7.14s/it, lr: 1.0e-04 loss: 6.976e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2735/30000 [6:09:48<54:05:04,  7.14s/it, lr: 1.0e-04 loss: 6.976e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2736/30000 [6:09:48<58:41:33,  7.75s/it, lr: 1.0e-04 loss: 6.976e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2736/30000 [6:09:48<58:41:33,  7.75s/it, lr: 1.0e-04 loss: 6.976e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2736/30000 [6:09:58<58:41:33,  7.75s/it, lr: 1.0e-04 loss: 5.743e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2736/30000 [6:09:58<58:41:33,  7.75s/it, lr: 1.0e-04 loss: 5.743e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2737/30000 [6:09:58<62:10:03,  8.21s/it, lr: 1.0e-04 loss: 5.743e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2737/30000 [6:09:58<62:10:03,  8.21s/it, lr: 1.0e-04 loss: 5.743e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2737/30000 [6:10:04<62:10:03,  8.21s/it, lr: 1.0e-04 loss: 6.205e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2737/30000 [6:10:04<62:10:03,  8.21s/it, lr: 1.0e-04 loss: 6.205e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2738/30000 [6:10:04<58:29:40,  7.72s/it, lr: 1.0e-04 loss: 6.205e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2738/30000 [6:10:04<58:29:40,  7.72s/it, lr: 1.0e-04 loss: 6.205e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2738/30000 [6:10:11<58:29:40,  7.72s/it, lr: 1.0e-04 loss: 7.548e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2738/30000 [6:10:11<58:29:40,  7.72s/it, lr: 1.0e-04 loss: 7.548e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2739/30000 [6:10:11<55:47:14,  7.37s/it, lr: 1.0e-04 loss: 7.548e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2739/30000 [6:10:11<55:47:14,  7.37s/it, lr: 1.0e-04 loss: 7.548e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2739/30000 [6:10:17<55:47:14,  7.37s/it, lr: 1.0e-04 loss: 7.154e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2739/30000 [6:10:17<55:47:14,  7.37s/it, lr: 1.0e-04 loss: 7.154e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2740/30000 [6:10:24<55:47:06,  7.37s/it, lr: 1.0e-04 loss: 5.709e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2740/30000 [6:10:24<55:47:06,  7.37s/it, lr: 1.0e-04 loss: 5.709e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2741/30000 [6:10:24<41:39:44,  5.50s/it, lr: 1.0e-04 loss: 5.709e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2741/30000 [6:10:24<41:39:44,  5.50s/it, lr: 1.0e-04 loss: 5.709e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2741/30000 [6:10:31<41:39:44,  5.50s/it, lr: 1.0e-04 loss: 1.481e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2741/30000 [6:10:31<41:39:44,  5.50s/it, lr: 1.0e-04 loss: 1.481e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2742/30000 [6:10:31<43:42:48,  5.77s/it, lr: 1.0e-04 loss: 1.481e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2742/30000 [6:10:31<43:42:48,  5.77s/it, lr: 1.0e-04 loss: 1.481e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2742/30000 [6:10:37<43:42:48,  5.77s/it, lr: 1.0e-04 loss: 3.475e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2742/30000 [6:10:37<43:42:48,  5.77s/it, lr: 1.0e-04 loss: 3.475e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2743/30000 [6:10:37<45:16:16,  5.98s/it, lr: 1.0e-04 loss: 3.475e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2743/30000 [6:10:37<45:16:16,  5.98s/it, lr: 1.0e-04 loss: 3.475e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2743/30000 [6:10:46<45:16:16,  5.98s/it, lr: 1.0e-04 loss: 8.552e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2743/30000 [6:10:46<45:16:16,  5.98s/it, lr: 1.0e-04 loss: 8.552e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2744/30000 [6:10:46<52:03:58,  6.88s/it, lr: 1.0e-04 loss: 8.552e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2744/30000 [6:10:46<52:03:58,  6.88s/it, lr: 1.0e-04 loss: 8.552e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2744/30000 [6:10:56<52:03:58,  6.88s/it, lr: 1.0e-04 loss: 6.416e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2744/30000 [6:10:56<52:03:58,  6.88s/it, lr: 1.0e-04 loss: 6.416e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2745/30000 [6:10:56<57:13:20,  7.56s/it, lr: 1.0e-04 loss: 6.416e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2745/30000 [6:10:56<57:13:20,  7.56s/it, lr: 1.0e-04 loss: 6.416e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2745/30000 [6:11:02<57:13:20,  7.56s/it, lr: 1.0e-04 loss: 7.587e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2745/30000 [6:11:02<57:13:20,  7.56s/it, lr: 1.0e-04 loss: 7.587e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2746/30000 [6:11:02<55:12:51,  7.29s/it, lr: 1.0e-04 loss: 7.587e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2746/30000 [6:11:02<55:12:51,  7.29s/it, lr: 1.0e-04 loss: 7.587e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2746/30000 [6:11:12<55:12:51,  7.29s/it, lr: 1.0e-04 loss: 3.690e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2746/30000 [6:11:12<55:12:51,  7.29s/it, lr: 1.0e-04 loss: 3.690e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2747/30000 [6:11:12<59:55:47,  7.92s/it, lr: 1.0e-04 loss: 3.690e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2747/30000 [6:11:12<59:55:47,  7.92s/it, lr: 1.0e-04 loss: 3.690e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2747/30000 [6:11:18<59:55:47,  7.92s/it, lr: 1.0e-04 loss: 7.464e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2747/30000 [6:11:18<59:55:47,  7.92s/it, lr: 1.0e-04 loss: 7.464e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2748/30000 [6:11:18<57:03:23,  7.54s/it, lr: 1.0e-04 loss: 7.464e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2748/30000 [6:11:18<57:03:23,  7.54s/it, lr: 1.0e-04 loss: 7.464e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2748/30000 [6:11:28<57:03:23,  7.54s/it, lr: 1.0e-04 loss: 4.104e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2748/30000 [6:11:28<57:03:23,  7.54s/it, lr: 1.0e-04 loss: 4.104e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2749/30000 [6:11:28<60:59:46,  8.06s/it, lr: 1.0e-04 loss: 4.104e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2749/30000 [6:11:28<60:59:46,  8.06s/it, lr: 1.0e-04 loss: 4.104e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2749/30000 [6:11:37<60:59:46,  8.06s/it, lr: 1.0e-04 loss: 5.362e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2749/30000 [6:11:37<60:59:46,  8.06s/it, lr: 1.0e-04 loss: 5.362e-02]
+Saving at step 2750
+Saved checkpoint to /home/featurize/ai-toolkit/output/Xiang_wan_2_2_lora/Xiang_wan_2_2_lora_000002750.safetensors
+Saved optimizer to /home/featurize/ai-toolkit/output/Xiang_wan_2_2_lora/optimizer.pt
+
Xiang_wan_2_2_lora:   9%|###5                                   | 2750/30000 [6:11:47<60:59:38,  8.06s/it, lr: 1.0e-04 loss: 5.846e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2750/30000 [6:11:47<60:59:38,  8.06s/it, lr: 1.0e-04 loss: 5.846e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2751/30000 [6:11:47<49:43:05,  6.57s/it, lr: 1.0e-04 loss: 5.846e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2751/30000 [6:11:47<49:43:05,  6.57s/it, lr: 1.0e-04 loss: 5.846e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2751/30000 [6:11:56<49:43:05,  6.57s/it, lr: 1.0e-04 loss: 1.007e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2751/30000 [6:11:56<49:43:05,  6.57s/it, lr: 1.0e-04 loss: 1.007e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2752/30000 [6:11:56<54:52:41,  7.25s/it, lr: 1.0e-04 loss: 1.007e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2752/30000 [6:11:56<54:52:41,  7.25s/it, lr: 1.0e-04 loss: 1.007e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2752/30000 [6:12:05<54:52:41,  7.25s/it, lr: 1.0e-04 loss: 1.028e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2752/30000 [6:12:05<54:52:41,  7.25s/it, lr: 1.0e-04 loss: 1.028e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2753/30000 [6:12:05<59:03:19,  7.80s/it, lr: 1.0e-04 loss: 1.028e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2753/30000 [6:12:05<59:03:19,  7.80s/it, lr: 1.0e-04 loss: 1.028e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2753/30000 [6:12:15<59:03:19,  7.80s/it, lr: 1.0e-04 loss: 6.732e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2753/30000 [6:12:15<59:03:19,  7.80s/it, lr: 1.0e-04 loss: 6.732e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2754/30000 [6:12:15<62:15:29,  8.23s/it, lr: 1.0e-04 loss: 6.732e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2754/30000 [6:12:15<62:15:29,  8.23s/it, lr: 1.0e-04 loss: 6.732e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2754/30000 [6:12:21<62:15:29,  8.23s/it, lr: 1.0e-04 loss: 8.444e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2754/30000 [6:12:21<62:15:29,  8.23s/it, lr: 1.0e-04 loss: 8.444e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2755/30000 [6:12:21<59:01:53,  7.80s/it, lr: 1.0e-04 loss: 8.444e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2755/30000 [6:12:21<59:01:53,  7.80s/it, lr: 1.0e-04 loss: 8.444e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2755/30000 [6:12:28<59:01:53,  7.80s/it, lr: 1.0e-04 loss: 6.945e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2755/30000 [6:12:28<59:01:53,  7.80s/it, lr: 1.0e-04 loss: 6.945e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2756/30000 [6:12:28<56:28:15,  7.46s/it, lr: 1.0e-04 loss: 6.945e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2756/30000 [6:12:28<56:28:15,  7.46s/it, lr: 1.0e-04 loss: 6.945e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2756/30000 [6:12:35<56:28:15,  7.46s/it, lr: 1.0e-04 loss: 7.338e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2756/30000 [6:12:35<56:28:15,  7.46s/it, lr: 1.0e-04 loss: 7.338e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2757/30000 [6:12:35<55:45:59,  7.37s/it, lr: 1.0e-04 loss: 7.338e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2757/30000 [6:12:35<55:45:59,  7.37s/it, lr: 1.0e-04 loss: 7.338e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2757/30000 [6:12:43<55:45:59,  7.37s/it, lr: 1.0e-04 loss: 8.886e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2757/30000 [6:12:43<55:45:59,  7.37s/it, lr: 1.0e-04 loss: 8.886e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2758/30000 [6:12:43<56:19:28,  7.44s/it, lr: 1.0e-04 loss: 8.886e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2758/30000 [6:12:43<56:19:28,  7.44s/it, lr: 1.0e-04 loss: 8.886e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2758/30000 [6:12:53<56:19:28,  7.44s/it, lr: 1.0e-04 loss: 6.070e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2758/30000 [6:12:53<56:19:28,  7.44s/it, lr: 1.0e-04 loss: 6.070e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2759/30000 [6:12:53<63:09:51,  8.35s/it, lr: 1.0e-04 loss: 6.070e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2759/30000 [6:12:53<63:09:51,  8.35s/it, lr: 1.0e-04 loss: 6.070e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2759/30000 [6:13:04<63:09:51,  8.35s/it, lr: 1.0e-04 loss: 8.478e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2759/30000 [6:13:04<63:09:51,  8.35s/it, lr: 1.0e-04 loss: 8.478e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2760/30000 [6:13:14<63:09:42,  8.35s/it, lr: 1.0e-04 loss: 3.778e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2760/30000 [6:13:14<63:09:42,  8.35s/it, lr: 1.0e-04 loss: 3.778e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2761/30000 [6:13:14<52:33:22,  6.95s/it, lr: 1.0e-04 loss: 3.778e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2761/30000 [6:13:14<52:33:22,  6.95s/it, lr: 1.0e-04 loss: 3.778e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2761/30000 [6:13:22<52:33:22,  6.95s/it, lr: 1.0e-04 loss: 1.136e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2761/30000 [6:13:22<52:33:22,  6.95s/it, lr: 1.0e-04 loss: 1.136e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2762/30000 [6:13:22<54:00:57,  7.14s/it, lr: 1.0e-04 loss: 1.136e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2762/30000 [6:13:22<54:00:57,  7.14s/it, lr: 1.0e-04 loss: 1.136e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2762/30000 [6:13:30<54:00:57,  7.14s/it, lr: 1.0e-04 loss: 6.218e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2762/30000 [6:13:30<54:00:57,  7.14s/it, lr: 1.0e-04 loss: 6.218e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2763/30000 [6:13:30<55:05:33,  7.28s/it, lr: 1.0e-04 loss: 6.218e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2763/30000 [6:13:30<55:05:33,  7.28s/it, lr: 1.0e-04 loss: 6.218e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2763/30000 [6:13:40<55:05:33,  7.28s/it, lr: 1.0e-04 loss: 6.764e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2763/30000 [6:13:40<55:05:33,  7.28s/it, lr: 1.0e-04 loss: 6.764e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2764/30000 [6:13:40<61:40:38,  8.15s/it, lr: 1.0e-04 loss: 6.764e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2764/30000 [6:13:40<61:40:38,  8.15s/it, lr: 1.0e-04 loss: 6.764e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2764/30000 [6:13:49<61:40:38,  8.15s/it, lr: 1.0e-04 loss: 2.040e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2764/30000 [6:13:49<61:40:38,  8.15s/it, lr: 1.0e-04 loss: 2.040e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2765/30000 [6:13:49<61:57:20,  8.19s/it, lr: 1.0e-04 loss: 2.040e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2765/30000 [6:13:49<61:57:20,  8.19s/it, lr: 1.0e-04 loss: 2.040e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2765/30000 [6:13:59<61:57:20,  8.19s/it, lr: 1.0e-04 loss: 5.677e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2765/30000 [6:13:59<61:57:20,  8.19s/it, lr: 1.0e-04 loss: 5.677e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2766/30000 [6:13:59<66:49:38,  8.83s/it, lr: 1.0e-04 loss: 5.677e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2766/30000 [6:13:59<66:49:38,  8.83s/it, lr: 1.0e-04 loss: 5.677e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2766/30000 [6:14:07<66:49:38,  8.83s/it, lr: 1.0e-04 loss: 5.145e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2766/30000 [6:14:07<66:49:38,  8.83s/it, lr: 1.0e-04 loss: 5.145e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2767/30000 [6:14:07<64:26:18,  8.52s/it, lr: 1.0e-04 loss: 5.145e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2767/30000 [6:14:07<64:26:18,  8.52s/it, lr: 1.0e-04 loss: 5.145e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2767/30000 [6:14:16<64:26:18,  8.52s/it, lr: 1.0e-04 loss: 5.681e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2767/30000 [6:14:16<64:26:18,  8.52s/it, lr: 1.0e-04 loss: 5.681e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2768/30000 [6:14:16<67:03:08,  8.86s/it, lr: 1.0e-04 loss: 5.681e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2768/30000 [6:14:16<67:03:08,  8.86s/it, lr: 1.0e-04 loss: 5.681e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2768/30000 [6:14:26<67:03:08,  8.86s/it, lr: 1.0e-04 loss: 5.150e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2768/30000 [6:14:26<67:03:08,  8.86s/it, lr: 1.0e-04 loss: 5.150e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2769/30000 [6:14:26<68:09:04,  9.01s/it, lr: 1.0e-04 loss: 5.150e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2769/30000 [6:14:26<68:09:04,  9.01s/it, lr: 1.0e-04 loss: 5.150e-02]
Xiang_wan_2_2_lora:   9%|###5                                   | 2769/30000 [6:14:33<68:09:04,  9.01s/it, lr: 1.0e-04 loss: 1.115e-01]
Xiang_wan_2_2_lora:   9%|###5                                   | 2769/30000 [6:14:33<68:09:04,  9.01s/it, lr: 1.0e-04 loss: 1.115e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2770/30000 [6:14:42<68:08:55,  9.01s/it, lr: 1.0e-04 loss: 2.695e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2770/30000 [6:14:42<68:08:55,  9.01s/it, lr: 1.0e-04 loss: 2.695e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2771/30000 [6:14:42<53:05:17,  7.02s/it, lr: 1.0e-04 loss: 2.695e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2771/30000 [6:14:42<53:05:17,  7.02s/it, lr: 1.0e-04 loss: 2.695e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2771/30000 [6:14:51<53:05:17,  7.02s/it, lr: 1.0e-04 loss: 7.041e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2771/30000 [6:14:51<53:05:17,  7.02s/it, lr: 1.0e-04 loss: 7.041e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2772/30000 [6:14:51<57:32:44,  7.61s/it, lr: 1.0e-04 loss: 7.041e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2772/30000 [6:14:51<57:32:44,  7.61s/it, lr: 1.0e-04 loss: 7.041e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2772/30000 [6:14:58<57:32:44,  7.61s/it, lr: 1.0e-04 loss: 8.149e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2772/30000 [6:14:58<57:32:44,  7.61s/it, lr: 1.0e-04 loss: 8.149e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2773/30000 [6:14:58<56:10:16,  7.43s/it, lr: 1.0e-04 loss: 8.149e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2773/30000 [6:14:58<56:10:16,  7.43s/it, lr: 1.0e-04 loss: 8.149e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2773/30000 [6:15:08<56:10:16,  7.43s/it, lr: 1.0e-04 loss: 6.703e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2773/30000 [6:15:08<56:10:16,  7.43s/it, lr: 1.0e-04 loss: 6.703e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2774/30000 [6:15:08<60:08:05,  7.95s/it, lr: 1.0e-04 loss: 6.703e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2774/30000 [6:15:08<60:08:05,  7.95s/it, lr: 1.0e-04 loss: 6.703e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2774/30000 [6:15:17<60:08:05,  7.95s/it, lr: 1.0e-04 loss: 2.821e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2774/30000 [6:15:17<60:08:05,  7.95s/it, lr: 1.0e-04 loss: 2.821e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2775/30000 [6:15:17<62:59:20,  8.33s/it, lr: 1.0e-04 loss: 2.821e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2775/30000 [6:15:17<62:59:20,  8.33s/it, lr: 1.0e-04 loss: 2.821e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2775/30000 [6:15:26<62:59:20,  8.33s/it, lr: 1.0e-04 loss: 5.999e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2775/30000 [6:15:26<62:59:20,  8.33s/it, lr: 1.0e-04 loss: 5.999e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2776/30000 [6:15:26<65:11:28,  8.62s/it, lr: 1.0e-04 loss: 5.999e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2776/30000 [6:15:26<65:11:28,  8.62s/it, lr: 1.0e-04 loss: 5.999e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2776/30000 [6:15:33<65:11:28,  8.62s/it, lr: 1.0e-04 loss: 4.485e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2776/30000 [6:15:33<65:11:28,  8.62s/it, lr: 1.0e-04 loss: 4.485e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2777/30000 [6:15:33<60:42:01,  8.03s/it, lr: 1.0e-04 loss: 4.485e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2777/30000 [6:15:33<60:42:01,  8.03s/it, lr: 1.0e-04 loss: 4.485e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2777/30000 [6:15:42<60:42:01,  8.03s/it, lr: 1.0e-04 loss: 7.042e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2777/30000 [6:15:42<60:42:01,  8.03s/it, lr: 1.0e-04 loss: 7.042e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2778/30000 [6:15:42<63:27:09,  8.39s/it, lr: 1.0e-04 loss: 7.042e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2778/30000 [6:15:42<63:27:09,  8.39s/it, lr: 1.0e-04 loss: 7.042e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2778/30000 [6:15:49<63:27:09,  8.39s/it, lr: 1.0e-04 loss: 2.571e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2778/30000 [6:15:49<63:27:09,  8.39s/it, lr: 1.0e-04 loss: 2.571e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2779/30000 [6:15:49<59:31:35,  7.87s/it, lr: 1.0e-04 loss: 2.571e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2779/30000 [6:15:49<59:31:35,  7.87s/it, lr: 1.0e-04 loss: 2.571e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2779/30000 [6:15:58<59:31:35,  7.87s/it, lr: 1.0e-04 loss: 8.669e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2779/30000 [6:15:58<59:31:35,  7.87s/it, lr: 1.0e-04 loss: 8.669e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2780/30000 [6:16:05<59:31:28,  7.87s/it, lr: 1.0e-04 loss: 1.866e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2780/30000 [6:16:05<59:31:28,  7.87s/it, lr: 1.0e-04 loss: 1.866e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2781/30000 [6:16:05<43:42:42,  5.78s/it, lr: 1.0e-04 loss: 1.866e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2781/30000 [6:16:05<43:42:42,  5.78s/it, lr: 1.0e-04 loss: 1.866e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2781/30000 [6:16:11<43:42:42,  5.78s/it, lr: 1.0e-04 loss: 7.533e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2781/30000 [6:16:11<43:42:42,  5.78s/it, lr: 1.0e-04 loss: 7.533e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2782/30000 [6:16:11<45:08:54,  5.97s/it, lr: 1.0e-04 loss: 7.533e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2782/30000 [6:16:11<45:08:54,  5.97s/it, lr: 1.0e-04 loss: 7.533e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2782/30000 [6:16:18<45:08:54,  5.97s/it, lr: 1.0e-04 loss: 8.920e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2782/30000 [6:16:18<45:08:54,  5.97s/it, lr: 1.0e-04 loss: 8.920e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2783/30000 [6:16:18<46:18:49,  6.13s/it, lr: 1.0e-04 loss: 8.920e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2783/30000 [6:16:18<46:18:49,  6.13s/it, lr: 1.0e-04 loss: 8.920e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2783/30000 [6:16:27<46:18:49,  6.13s/it, lr: 1.0e-04 loss: 5.309e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2783/30000 [6:16:27<46:18:49,  6.13s/it, lr: 1.0e-04 loss: 5.309e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2784/30000 [6:16:27<52:39:13,  6.96s/it, lr: 1.0e-04 loss: 5.309e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2784/30000 [6:16:27<52:39:13,  6.96s/it, lr: 1.0e-04 loss: 5.309e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2784/30000 [6:16:36<52:39:13,  6.96s/it, lr: 1.0e-04 loss: 3.212e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2784/30000 [6:16:36<52:39:13,  6.96s/it, lr: 1.0e-04 loss: 3.212e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2785/30000 [6:16:36<57:44:04,  7.64s/it, lr: 1.0e-04 loss: 3.212e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2785/30000 [6:16:36<57:44:04,  7.64s/it, lr: 1.0e-04 loss: 3.212e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2785/30000 [6:16:43<57:44:04,  7.64s/it, lr: 1.0e-04 loss: 9.006e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2785/30000 [6:16:43<57:44:04,  7.64s/it, lr: 1.0e-04 loss: 9.006e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2786/30000 [6:16:43<55:22:34,  7.33s/it, lr: 1.0e-04 loss: 9.006e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2786/30000 [6:16:43<55:22:34,  7.33s/it, lr: 1.0e-04 loss: 9.006e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2786/30000 [6:16:50<55:22:34,  7.33s/it, lr: 1.0e-04 loss: 5.727e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2786/30000 [6:16:50<55:22:34,  7.33s/it, lr: 1.0e-04 loss: 5.727e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2787/30000 [6:16:50<53:41:57,  7.10s/it, lr: 1.0e-04 loss: 5.727e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2787/30000 [6:16:50<53:41:57,  7.10s/it, lr: 1.0e-04 loss: 5.727e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2787/30000 [6:16:59<53:41:57,  7.10s/it, lr: 1.0e-04 loss: 5.875e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2787/30000 [6:16:59<53:41:57,  7.10s/it, lr: 1.0e-04 loss: 5.875e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2788/30000 [6:16:59<58:50:48,  7.79s/it, lr: 1.0e-04 loss: 5.875e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2788/30000 [6:16:59<58:50:48,  7.79s/it, lr: 1.0e-04 loss: 5.875e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2788/30000 [6:17:08<58:50:48,  7.79s/it, lr: 1.0e-04 loss: 4.025e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2788/30000 [6:17:08<58:50:48,  7.79s/it, lr: 1.0e-04 loss: 4.025e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2789/30000 [6:17:08<62:19:08,  8.24s/it, lr: 1.0e-04 loss: 4.025e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2789/30000 [6:17:08<62:19:08,  8.24s/it, lr: 1.0e-04 loss: 4.025e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2789/30000 [6:17:15<62:19:08,  8.24s/it, lr: 1.0e-04 loss: 5.590e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2789/30000 [6:17:15<62:19:08,  8.24s/it, lr: 1.0e-04 loss: 5.590e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2790/30000 [6:17:21<62:19:00,  8.24s/it, lr: 1.0e-04 loss: 1.024e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2790/30000 [6:17:21<62:19:00,  8.24s/it, lr: 1.0e-04 loss: 1.024e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2791/30000 [6:17:21<45:03:00,  5.96s/it, lr: 1.0e-04 loss: 1.024e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2791/30000 [6:17:21<45:03:00,  5.96s/it, lr: 1.0e-04 loss: 1.024e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2791/30000 [6:17:28<45:03:00,  5.96s/it, lr: 1.0e-04 loss: 8.392e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2791/30000 [6:17:28<45:03:00,  5.96s/it, lr: 1.0e-04 loss: 8.392e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2792/30000 [6:17:28<46:09:36,  6.11s/it, lr: 1.0e-04 loss: 8.392e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2792/30000 [6:17:28<46:09:36,  6.11s/it, lr: 1.0e-04 loss: 8.392e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2792/30000 [6:17:37<46:09:36,  6.11s/it, lr: 1.0e-04 loss: 4.783e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2792/30000 [6:17:37<46:09:36,  6.11s/it, lr: 1.0e-04 loss: 4.783e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2793/30000 [6:17:37<52:28:25,  6.94s/it, lr: 1.0e-04 loss: 4.783e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2793/30000 [6:17:37<52:28:25,  6.94s/it, lr: 1.0e-04 loss: 4.783e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2793/30000 [6:17:47<52:28:25,  6.94s/it, lr: 1.0e-04 loss: 6.119e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2793/30000 [6:17:47<52:28:25,  6.94s/it, lr: 1.0e-04 loss: 6.119e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2794/30000 [6:17:47<57:25:50,  7.60s/it, lr: 1.0e-04 loss: 6.119e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2794/30000 [6:17:47<57:25:50,  7.60s/it, lr: 1.0e-04 loss: 6.119e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2794/30000 [6:17:56<57:25:50,  7.60s/it, lr: 1.0e-04 loss: 9.837e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2794/30000 [6:17:56<57:25:50,  7.60s/it, lr: 1.0e-04 loss: 9.837e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2795/30000 [6:17:56<61:43:25,  8.17s/it, lr: 1.0e-04 loss: 9.837e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2795/30000 [6:17:56<61:43:25,  8.17s/it, lr: 1.0e-04 loss: 9.837e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2795/30000 [6:18:06<61:43:25,  8.17s/it, lr: 1.0e-04 loss: 6.321e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2795/30000 [6:18:06<61:43:25,  8.17s/it, lr: 1.0e-04 loss: 6.321e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2796/30000 [6:18:06<64:16:10,  8.51s/it, lr: 1.0e-04 loss: 6.321e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2796/30000 [6:18:06<64:16:10,  8.51s/it, lr: 1.0e-04 loss: 6.321e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2796/30000 [6:18:15<64:16:10,  8.51s/it, lr: 1.0e-04 loss: 4.714e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2796/30000 [6:18:15<64:16:10,  8.51s/it, lr: 1.0e-04 loss: 4.714e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2797/30000 [6:18:15<66:05:41,  8.75s/it, lr: 1.0e-04 loss: 4.714e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2797/30000 [6:18:15<66:05:41,  8.75s/it, lr: 1.0e-04 loss: 4.714e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2797/30000 [6:18:24<66:05:41,  8.75s/it, lr: 1.0e-04 loss: 8.463e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2797/30000 [6:18:24<66:05:41,  8.75s/it, lr: 1.0e-04 loss: 8.463e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2798/30000 [6:18:24<67:21:21,  8.91s/it, lr: 1.0e-04 loss: 8.463e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2798/30000 [6:18:24<67:21:21,  8.91s/it, lr: 1.0e-04 loss: 8.463e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2798/30000 [6:18:31<67:21:21,  8.91s/it, lr: 1.0e-04 loss: 1.027e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2798/30000 [6:18:31<67:21:21,  8.91s/it, lr: 1.0e-04 loss: 1.027e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2799/30000 [6:18:31<62:05:48,  8.22s/it, lr: 1.0e-04 loss: 1.027e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2799/30000 [6:18:31<62:05:48,  8.22s/it, lr: 1.0e-04 loss: 1.027e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2799/30000 [6:18:40<62:05:48,  8.22s/it, lr: 1.0e-04 loss: 8.172e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2799/30000 [6:18:40<62:05:48,  8.22s/it, lr: 1.0e-04 loss: 8.172e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2800/30000 [6:18:47<62:05:40,  8.22s/it, lr: 1.0e-04 loss: 3.403e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2800/30000 [6:18:47<62:05:40,  8.22s/it, lr: 1.0e-04 loss: 3.403e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2801/30000 [6:18:47<45:18:41,  6.00s/it, lr: 1.0e-04 loss: 3.403e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2801/30000 [6:18:47<45:18:41,  6.00s/it, lr: 1.0e-04 loss: 3.403e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2801/30000 [6:18:53<45:18:41,  6.00s/it, lr: 1.0e-04 loss: 8.080e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2801/30000 [6:18:53<45:18:41,  6.00s/it, lr: 1.0e-04 loss: 8.080e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2802/30000 [6:18:53<46:28:58,  6.15s/it, lr: 1.0e-04 loss: 8.080e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2802/30000 [6:18:53<46:28:58,  6.15s/it, lr: 1.0e-04 loss: 8.080e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2802/30000 [6:19:03<46:28:58,  6.15s/it, lr: 1.0e-04 loss: 4.933e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2802/30000 [6:19:03<46:28:58,  6.15s/it, lr: 1.0e-04 loss: 4.933e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2803/30000 [6:19:03<53:23:51,  7.07s/it, lr: 1.0e-04 loss: 4.933e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2803/30000 [6:19:03<53:23:51,  7.07s/it, lr: 1.0e-04 loss: 4.933e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2803/30000 [6:19:10<53:23:51,  7.07s/it, lr: 1.0e-04 loss: 4.763e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2803/30000 [6:19:10<53:23:51,  7.07s/it, lr: 1.0e-04 loss: 4.763e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2804/30000 [6:19:10<52:42:06,  6.98s/it, lr: 1.0e-04 loss: 4.763e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2804/30000 [6:19:10<52:42:06,  6.98s/it, lr: 1.0e-04 loss: 4.763e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2804/30000 [6:19:16<52:42:06,  6.98s/it, lr: 1.0e-04 loss: 1.229e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2804/30000 [6:19:16<52:42:06,  6.98s/it, lr: 1.0e-04 loss: 1.229e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2805/30000 [6:19:16<51:58:21,  6.88s/it, lr: 1.0e-04 loss: 1.229e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2805/30000 [6:19:16<51:58:21,  6.88s/it, lr: 1.0e-04 loss: 1.229e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2805/30000 [6:19:26<51:58:21,  6.88s/it, lr: 1.0e-04 loss: 8.077e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2805/30000 [6:19:26<51:58:21,  6.88s/it, lr: 1.0e-04 loss: 8.077e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2806/30000 [6:19:26<57:15:51,  7.58s/it, lr: 1.0e-04 loss: 8.077e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2806/30000 [6:19:26<57:15:51,  7.58s/it, lr: 1.0e-04 loss: 8.077e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2806/30000 [6:19:33<57:15:51,  7.58s/it, lr: 1.0e-04 loss: 1.164e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2806/30000 [6:19:33<57:15:51,  7.58s/it, lr: 1.0e-04 loss: 1.164e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2807/30000 [6:19:33<55:24:33,  7.34s/it, lr: 1.0e-04 loss: 1.164e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2807/30000 [6:19:33<55:24:33,  7.34s/it, lr: 1.0e-04 loss: 1.164e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2807/30000 [6:19:42<55:24:33,  7.34s/it, lr: 1.0e-04 loss: 6.021e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2807/30000 [6:19:42<55:24:33,  7.34s/it, lr: 1.0e-04 loss: 6.021e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2808/30000 [6:19:42<59:55:27,  7.93s/it, lr: 1.0e-04 loss: 6.021e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2808/30000 [6:19:42<59:55:27,  7.93s/it, lr: 1.0e-04 loss: 6.021e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2808/30000 [6:19:49<59:55:27,  7.93s/it, lr: 1.0e-04 loss: 1.319e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2808/30000 [6:19:49<59:55:27,  7.93s/it, lr: 1.0e-04 loss: 1.319e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2809/30000 [6:19:49<57:05:54,  7.56s/it, lr: 1.0e-04 loss: 1.319e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2809/30000 [6:19:49<57:05:54,  7.56s/it, lr: 1.0e-04 loss: 1.319e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2809/30000 [6:19:56<57:05:54,  7.56s/it, lr: 1.0e-04 loss: 8.541e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2809/30000 [6:19:56<57:05:54,  7.56s/it, lr: 1.0e-04 loss: 8.541e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2810/30000 [6:20:02<57:05:47,  7.56s/it, lr: 1.0e-04 loss: 8.774e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2810/30000 [6:20:02<57:05:47,  7.56s/it, lr: 1.0e-04 loss: 8.774e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2811/30000 [6:20:02<42:25:59,  5.62s/it, lr: 1.0e-04 loss: 8.774e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2811/30000 [6:20:02<42:25:59,  5.62s/it, lr: 1.0e-04 loss: 8.774e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2811/30000 [6:20:09<42:25:59,  5.62s/it, lr: 1.0e-04 loss: 1.132e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2811/30000 [6:20:09<42:25:59,  5.62s/it, lr: 1.0e-04 loss: 1.132e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2812/30000 [6:20:09<44:20:26,  5.87s/it, lr: 1.0e-04 loss: 1.132e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2812/30000 [6:20:09<44:20:26,  5.87s/it, lr: 1.0e-04 loss: 1.132e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2812/30000 [6:20:15<44:20:26,  5.87s/it, lr: 1.0e-04 loss: 9.976e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2812/30000 [6:20:15<44:20:26,  5.87s/it, lr: 1.0e-04 loss: 9.976e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2813/30000 [6:20:15<45:40:15,  6.05s/it, lr: 1.0e-04 loss: 9.976e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2813/30000 [6:20:15<45:40:15,  6.05s/it, lr: 1.0e-04 loss: 9.976e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2813/30000 [6:20:25<45:40:15,  6.05s/it, lr: 1.0e-04 loss: 7.579e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2813/30000 [6:20:25<45:40:15,  6.05s/it, lr: 1.0e-04 loss: 7.579e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2814/30000 [6:20:25<52:22:40,  6.94s/it, lr: 1.0e-04 loss: 7.579e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2814/30000 [6:20:25<52:22:40,  6.94s/it, lr: 1.0e-04 loss: 7.579e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2814/30000 [6:20:34<52:22:40,  6.94s/it, lr: 1.0e-04 loss: 1.403e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2814/30000 [6:20:34<52:22:40,  6.94s/it, lr: 1.0e-04 loss: 1.403e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2815/30000 [6:20:34<57:29:06,  7.61s/it, lr: 1.0e-04 loss: 1.403e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2815/30000 [6:20:34<57:29:06,  7.61s/it, lr: 1.0e-04 loss: 1.403e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2815/30000 [6:20:43<57:29:06,  7.61s/it, lr: 1.0e-04 loss: 8.313e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2815/30000 [6:20:43<57:29:06,  7.61s/it, lr: 1.0e-04 loss: 8.313e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2816/30000 [6:20:43<61:13:32,  8.11s/it, lr: 1.0e-04 loss: 8.313e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2816/30000 [6:20:43<61:13:32,  8.11s/it, lr: 1.0e-04 loss: 8.313e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2816/30000 [6:20:53<61:13:32,  8.11s/it, lr: 1.0e-04 loss: 5.098e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2816/30000 [6:20:53<61:13:32,  8.11s/it, lr: 1.0e-04 loss: 5.098e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2817/30000 [6:20:53<64:29:14,  8.54s/it, lr: 1.0e-04 loss: 5.098e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2817/30000 [6:20:53<64:29:14,  8.54s/it, lr: 1.0e-04 loss: 5.098e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2817/30000 [6:21:00<64:29:14,  8.54s/it, lr: 1.0e-04 loss: 6.589e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2817/30000 [6:21:00<64:29:14,  8.54s/it, lr: 1.0e-04 loss: 6.589e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2818/30000 [6:21:00<60:09:19,  7.97s/it, lr: 1.0e-04 loss: 6.589e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2818/30000 [6:21:00<60:09:19,  7.97s/it, lr: 1.0e-04 loss: 6.589e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2818/30000 [6:21:09<60:09:19,  7.97s/it, lr: 1.0e-04 loss: 5.263e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2818/30000 [6:21:09<60:09:19,  7.97s/it, lr: 1.0e-04 loss: 5.263e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2819/30000 [6:21:09<63:18:23,  8.38s/it, lr: 1.0e-04 loss: 5.263e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2819/30000 [6:21:09<63:18:23,  8.38s/it, lr: 1.0e-04 loss: 5.263e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2819/30000 [6:21:16<63:18:23,  8.38s/it, lr: 1.0e-04 loss: 8.669e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2819/30000 [6:21:16<63:18:23,  8.38s/it, lr: 1.0e-04 loss: 8.669e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2820/30000 [6:21:25<63:18:15,  8.38s/it, lr: 1.0e-04 loss: 6.205e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2820/30000 [6:21:25<63:18:15,  8.38s/it, lr: 1.0e-04 loss: 6.205e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2821/30000 [6:21:25<50:29:45,  6.69s/it, lr: 1.0e-04 loss: 6.205e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2821/30000 [6:21:25<50:29:45,  6.69s/it, lr: 1.0e-04 loss: 6.205e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2821/30000 [6:21:34<50:29:45,  6.69s/it, lr: 1.0e-04 loss: 5.974e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2821/30000 [6:21:34<50:29:45,  6.69s/it, lr: 1.0e-04 loss: 5.974e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2822/30000 [6:21:34<55:32:55,  7.36s/it, lr: 1.0e-04 loss: 5.974e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2822/30000 [6:21:34<55:32:55,  7.36s/it, lr: 1.0e-04 loss: 5.974e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2822/30000 [6:21:44<55:32:55,  7.36s/it, lr: 1.0e-04 loss: 1.415e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2822/30000 [6:21:44<55:32:55,  7.36s/it, lr: 1.0e-04 loss: 1.415e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2823/30000 [6:21:44<59:32:46,  7.89s/it, lr: 1.0e-04 loss: 1.415e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2823/30000 [6:21:44<59:32:46,  7.89s/it, lr: 1.0e-04 loss: 1.415e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2823/30000 [6:21:50<59:32:46,  7.89s/it, lr: 1.0e-04 loss: 6.570e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2823/30000 [6:21:50<59:32:46,  7.89s/it, lr: 1.0e-04 loss: 6.570e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2824/30000 [6:21:50<57:10:23,  7.57s/it, lr: 1.0e-04 loss: 6.570e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2824/30000 [6:21:50<57:10:23,  7.57s/it, lr: 1.0e-04 loss: 6.570e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2824/30000 [6:21:57<57:10:23,  7.57s/it, lr: 1.0e-04 loss: 6.061e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2824/30000 [6:21:57<57:10:23,  7.57s/it, lr: 1.0e-04 loss: 6.061e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2825/30000 [6:21:57<55:42:02,  7.38s/it, lr: 1.0e-04 loss: 6.061e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2825/30000 [6:21:57<55:42:02,  7.38s/it, lr: 1.0e-04 loss: 6.061e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2825/30000 [6:22:07<55:42:02,  7.38s/it, lr: 1.0e-04 loss: 8.604e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2825/30000 [6:22:07<55:42:02,  7.38s/it, lr: 1.0e-04 loss: 8.604e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2826/30000 [6:22:07<60:07:52,  7.97s/it, lr: 1.0e-04 loss: 8.604e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2826/30000 [6:22:07<60:07:52,  7.97s/it, lr: 1.0e-04 loss: 8.604e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2826/30000 [6:22:13<60:07:52,  7.97s/it, lr: 1.0e-04 loss: 1.712e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2826/30000 [6:22:13<60:07:52,  7.97s/it, lr: 1.0e-04 loss: 1.712e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2827/30000 [6:22:13<57:17:55,  7.59s/it, lr: 1.0e-04 loss: 1.712e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2827/30000 [6:22:13<57:17:55,  7.59s/it, lr: 1.0e-04 loss: 1.712e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2827/30000 [6:22:23<57:17:55,  7.59s/it, lr: 1.0e-04 loss: 9.394e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2827/30000 [6:22:23<57:17:55,  7.59s/it, lr: 1.0e-04 loss: 9.394e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2828/30000 [6:22:23<61:11:10,  8.11s/it, lr: 1.0e-04 loss: 9.394e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2828/30000 [6:22:23<61:11:10,  8.11s/it, lr: 1.0e-04 loss: 9.394e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2828/30000 [6:22:31<61:11:10,  8.11s/it, lr: 1.0e-04 loss: 7.768e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2828/30000 [6:22:31<61:11:10,  8.11s/it, lr: 1.0e-04 loss: 7.768e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2829/30000 [6:22:31<60:29:41,  8.02s/it, lr: 1.0e-04 loss: 7.768e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2829/30000 [6:22:31<60:29:41,  8.02s/it, lr: 1.0e-04 loss: 7.768e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2829/30000 [6:22:41<60:29:41,  8.02s/it, lr: 1.0e-04 loss: 7.974e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2829/30000 [6:22:41<60:29:41,  8.02s/it, lr: 1.0e-04 loss: 7.974e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2830/30000 [6:22:49<60:29:33,  8.02s/it, lr: 1.0e-04 loss: 7.462e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2830/30000 [6:22:49<60:29:33,  8.02s/it, lr: 1.0e-04 loss: 7.462e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2831/30000 [6:22:49<46:20:20,  6.14s/it, lr: 1.0e-04 loss: 7.462e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2831/30000 [6:22:49<46:20:20,  6.14s/it, lr: 1.0e-04 loss: 7.462e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2831/30000 [6:22:57<46:20:20,  6.14s/it, lr: 1.0e-04 loss: 9.142e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2831/30000 [6:22:57<46:20:20,  6.14s/it, lr: 1.0e-04 loss: 9.142e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2832/30000 [6:22:57<50:40:50,  6.72s/it, lr: 1.0e-04 loss: 9.142e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2832/30000 [6:22:57<50:40:50,  6.72s/it, lr: 1.0e-04 loss: 9.142e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2832/30000 [6:23:05<50:40:50,  6.72s/it, lr: 1.0e-04 loss: 7.741e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2832/30000 [6:23:05<50:40:50,  6.72s/it, lr: 1.0e-04 loss: 7.741e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2833/30000 [6:23:05<52:48:33,  7.00s/it, lr: 1.0e-04 loss: 7.741e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2833/30000 [6:23:05<52:48:33,  7.00s/it, lr: 1.0e-04 loss: 7.741e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2833/30000 [6:23:16<52:48:33,  7.00s/it, lr: 1.0e-04 loss: 8.681e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2833/30000 [6:23:16<52:48:33,  7.00s/it, lr: 1.0e-04 loss: 8.681e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2834/30000 [6:23:16<59:57:24,  7.95s/it, lr: 1.0e-04 loss: 8.681e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2834/30000 [6:23:16<59:57:24,  7.95s/it, lr: 1.0e-04 loss: 8.681e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2834/30000 [6:23:25<59:57:24,  7.95s/it, lr: 1.0e-04 loss: 5.977e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2834/30000 [6:23:25<59:57:24,  7.95s/it, lr: 1.0e-04 loss: 5.977e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2835/30000 [6:23:25<62:59:44,  8.35s/it, lr: 1.0e-04 loss: 5.977e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2835/30000 [6:23:25<62:59:44,  8.35s/it, lr: 1.0e-04 loss: 5.977e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2835/30000 [6:23:33<62:59:44,  8.35s/it, lr: 1.0e-04 loss: 9.389e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2835/30000 [6:23:33<62:59:44,  8.35s/it, lr: 1.0e-04 loss: 9.389e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2836/30000 [6:23:33<61:51:12,  8.20s/it, lr: 1.0e-04 loss: 9.389e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2836/30000 [6:23:33<61:51:12,  8.20s/it, lr: 1.0e-04 loss: 9.389e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2836/30000 [6:23:43<61:51:12,  8.20s/it, lr: 1.0e-04 loss: 2.424e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2836/30000 [6:23:43<61:51:12,  8.20s/it, lr: 1.0e-04 loss: 2.424e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2837/30000 [6:23:43<66:57:10,  8.87s/it, lr: 1.0e-04 loss: 2.424e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2837/30000 [6:23:43<66:57:10,  8.87s/it, lr: 1.0e-04 loss: 2.424e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2837/30000 [6:23:53<66:57:10,  8.87s/it, lr: 1.0e-04 loss: 5.074e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2837/30000 [6:23:53<66:57:10,  8.87s/it, lr: 1.0e-04 loss: 5.074e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2838/30000 [6:23:53<68:17:59,  9.05s/it, lr: 1.0e-04 loss: 5.074e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2838/30000 [6:23:53<68:17:59,  9.05s/it, lr: 1.0e-04 loss: 5.074e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2838/30000 [6:24:02<68:17:59,  9.05s/it, lr: 1.0e-04 loss: 5.867e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2838/30000 [6:24:02<68:17:59,  9.05s/it, lr: 1.0e-04 loss: 5.867e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2839/30000 [6:24:02<69:00:18,  9.15s/it, lr: 1.0e-04 loss: 5.867e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2839/30000 [6:24:02<69:00:18,  9.15s/it, lr: 1.0e-04 loss: 5.867e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2839/30000 [6:24:09<69:00:18,  9.15s/it, lr: 1.0e-04 loss: 8.756e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2839/30000 [6:24:09<69:00:18,  9.15s/it, lr: 1.0e-04 loss: 8.756e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2840/30000 [6:24:16<69:00:09,  9.15s/it, lr: 1.0e-04 loss: 9.202e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2840/30000 [6:24:16<69:00:09,  9.15s/it, lr: 1.0e-04 loss: 9.202e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2841/30000 [6:24:16<48:53:10,  6.48s/it, lr: 1.0e-04 loss: 9.202e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2841/30000 [6:24:16<48:53:10,  6.48s/it, lr: 1.0e-04 loss: 9.202e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2841/30000 [6:24:25<48:53:10,  6.48s/it, lr: 1.0e-04 loss: 7.000e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2841/30000 [6:24:25<48:53:10,  6.48s/it, lr: 1.0e-04 loss: 7.000e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2842/30000 [6:24:25<54:19:00,  7.20s/it, lr: 1.0e-04 loss: 7.000e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2842/30000 [6:24:25<54:19:00,  7.20s/it, lr: 1.0e-04 loss: 7.000e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2842/30000 [6:24:35<54:19:00,  7.20s/it, lr: 1.0e-04 loss: 4.557e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2842/30000 [6:24:35<54:19:00,  7.20s/it, lr: 1.0e-04 loss: 4.557e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2843/30000 [6:24:35<58:39:14,  7.78s/it, lr: 1.0e-04 loss: 4.557e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2843/30000 [6:24:35<58:39:14,  7.78s/it, lr: 1.0e-04 loss: 4.557e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2843/30000 [6:24:41<58:39:14,  7.78s/it, lr: 1.0e-04 loss: 1.033e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2843/30000 [6:24:41<58:39:14,  7.78s/it, lr: 1.0e-04 loss: 1.033e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2844/30000 [6:24:41<56:22:33,  7.47s/it, lr: 1.0e-04 loss: 1.033e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2844/30000 [6:24:41<56:22:33,  7.47s/it, lr: 1.0e-04 loss: 1.033e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2844/30000 [6:24:51<56:22:33,  7.47s/it, lr: 1.0e-04 loss: 9.781e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2844/30000 [6:24:51<56:22:33,  7.47s/it, lr: 1.0e-04 loss: 9.781e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2845/30000 [6:24:51<60:14:53,  7.99s/it, lr: 1.0e-04 loss: 9.781e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2845/30000 [6:24:51<60:14:53,  7.99s/it, lr: 1.0e-04 loss: 9.781e-02]
Xiang_wan_2_2_lora:   9%|###6                                   | 2845/30000 [6:25:00<60:14:53,  7.99s/it, lr: 1.0e-04 loss: 1.064e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2845/30000 [6:25:00<60:14:53,  7.99s/it, lr: 1.0e-04 loss: 1.064e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2846/30000 [6:25:00<63:11:39,  8.38s/it, lr: 1.0e-04 loss: 1.064e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2846/30000 [6:25:00<63:11:39,  8.38s/it, lr: 1.0e-04 loss: 1.064e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2846/30000 [6:25:07<63:11:39,  8.38s/it, lr: 1.0e-04 loss: 2.635e-01]
Xiang_wan_2_2_lora:   9%|###6                                   | 2846/30000 [6:25:07<63:11:39,  8.38s/it, lr: 1.0e-04 loss: 2.635e-01]
Xiang_wan_2_2_lora:   9%|###7                                   | 2847/30000 [6:25:07<59:56:00,  7.95s/it, lr: 1.0e-04 loss: 2.635e-01]
Xiang_wan_2_2_lora:   9%|###7                                   | 2847/30000 [6:25:07<59:56:00,  7.95s/it, lr: 1.0e-04 loss: 2.635e-01]
Xiang_wan_2_2_lora:   9%|###7                                   | 2847/30000 [6:25:14<59:56:00,  7.95s/it, lr: 1.0e-04 loss: 6.934e-02]
Xiang_wan_2_2_lora:   9%|###7                                   | 2847/30000 [6:25:14<59:56:00,  7.95s/it, lr: 1.0e-04 loss: 6.934e-02]
Xiang_wan_2_2_lora:   9%|###7                                   | 2848/30000 [6:25:14<57:07:44,  7.57s/it, lr: 1.0e-04 loss: 6.934e-02]
Xiang_wan_2_2_lora:   9%|###7                                   | 2848/30000 [6:25:14<57:07:44,  7.57s/it, lr: 1.0e-04 loss: 6.934e-02]
Xiang_wan_2_2_lora:   9%|###7                                   | 2848/30000 [6:25:23<57:07:44,  7.57s/it, lr: 1.0e-04 loss: 3.178e-02]
Xiang_wan_2_2_lora:   9%|###7                                   | 2848/30000 [6:25:23<57:07:44,  7.57s/it, lr: 1.0e-04 loss: 3.178e-02]
Xiang_wan_2_2_lora:   9%|###7                                   | 2849/30000 [6:25:23<61:05:01,  8.10s/it, lr: 1.0e-04 loss: 3.178e-02]
Xiang_wan_2_2_lora:   9%|###7                                   | 2849/30000 [6:25:23<61:05:01,  8.10s/it, lr: 1.0e-04 loss: 3.178e-02]
Xiang_wan_2_2_lora:   9%|###7                                   | 2849/30000 [6:25:32<61:05:01,  8.10s/it, lr: 1.0e-04 loss: 6.640e-02]
Xiang_wan_2_2_lora:   9%|###7                                   | 2849/30000 [6:25:32<61:05:01,  8.10s/it, lr: 1.0e-04 loss: 6.640e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2850/30000 [6:25:39<61:04:53,  8.10s/it, lr: 1.0e-04 loss: 3.366e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2850/30000 [6:25:39<61:04:53,  8.10s/it, lr: 1.0e-04 loss: 3.366e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2851/30000 [6:25:39<44:32:25,  5.91s/it, lr: 1.0e-04 loss: 3.366e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2851/30000 [6:25:39<44:32:25,  5.91s/it, lr: 1.0e-04 loss: 3.366e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2851/30000 [6:25:45<44:32:25,  5.91s/it, lr: 1.0e-04 loss: 1.002e-01]
Xiang_wan_2_2_lora:  10%|###7                                   | 2851/30000 [6:25:45<44:32:25,  5.91s/it, lr: 1.0e-04 loss: 1.002e-01]
Xiang_wan_2_2_lora:  10%|###7                                   | 2852/30000 [6:25:45<45:45:38,  6.07s/it, lr: 1.0e-04 loss: 1.002e-01]
Xiang_wan_2_2_lora:  10%|###7                                   | 2852/30000 [6:25:45<45:45:38,  6.07s/it, lr: 1.0e-04 loss: 1.002e-01]
Xiang_wan_2_2_lora:  10%|###7                                   | 2852/30000 [6:25:52<45:45:38,  6.07s/it, lr: 1.0e-04 loss: 8.311e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2852/30000 [6:25:52<45:45:38,  6.07s/it, lr: 1.0e-04 loss: 8.311e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2853/30000 [6:25:52<46:41:43,  6.19s/it, lr: 1.0e-04 loss: 8.311e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2853/30000 [6:25:52<46:41:43,  6.19s/it, lr: 1.0e-04 loss: 8.311e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2853/30000 [6:25:59<46:41:43,  6.19s/it, lr: 1.0e-04 loss: 9.301e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2853/30000 [6:25:59<46:41:43,  6.19s/it, lr: 1.0e-04 loss: 9.301e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2854/30000 [6:25:59<47:21:20,  6.28s/it, lr: 1.0e-04 loss: 9.301e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2854/30000 [6:25:59<47:21:20,  6.28s/it, lr: 1.0e-04 loss: 9.301e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2854/30000 [6:26:05<47:21:20,  6.28s/it, lr: 1.0e-04 loss: 7.072e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2854/30000 [6:26:05<47:21:20,  6.28s/it, lr: 1.0e-04 loss: 7.072e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2855/30000 [6:26:05<48:28:40,  6.43s/it, lr: 1.0e-04 loss: 7.072e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2855/30000 [6:26:05<48:28:40,  6.43s/it, lr: 1.0e-04 loss: 7.072e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2855/30000 [6:26:12<48:28:40,  6.43s/it, lr: 1.0e-04 loss: 8.755e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2855/30000 [6:26:12<48:28:40,  6.43s/it, lr: 1.0e-04 loss: 8.755e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2856/30000 [6:26:12<48:39:56,  6.45s/it, lr: 1.0e-04 loss: 8.755e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2856/30000 [6:26:12<48:39:56,  6.45s/it, lr: 1.0e-04 loss: 8.755e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2856/30000 [6:26:18<48:39:56,  6.45s/it, lr: 1.0e-04 loss: 6.177e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2856/30000 [6:26:18<48:39:56,  6.45s/it, lr: 1.0e-04 loss: 6.177e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2857/30000 [6:26:18<48:46:32,  6.47s/it, lr: 1.0e-04 loss: 6.177e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2857/30000 [6:26:18<48:46:32,  6.47s/it, lr: 1.0e-04 loss: 6.177e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2857/30000 [6:26:28<48:46:32,  6.47s/it, lr: 1.0e-04 loss: 9.927e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2857/30000 [6:26:28<48:46:32,  6.47s/it, lr: 1.0e-04 loss: 9.927e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2858/30000 [6:26:28<55:01:45,  7.30s/it, lr: 1.0e-04 loss: 9.927e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2858/30000 [6:26:28<55:01:45,  7.30s/it, lr: 1.0e-04 loss: 9.927e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2858/30000 [6:26:37<55:01:45,  7.30s/it, lr: 1.0e-04 loss: 1.292e-01]
Xiang_wan_2_2_lora:  10%|###7                                   | 2858/30000 [6:26:37<55:01:45,  7.30s/it, lr: 1.0e-04 loss: 1.292e-01]
Xiang_wan_2_2_lora:  10%|###7                                   | 2859/30000 [6:26:37<59:27:19,  7.89s/it, lr: 1.0e-04 loss: 1.292e-01]
Xiang_wan_2_2_lora:  10%|###7                                   | 2859/30000 [6:26:37<59:27:19,  7.89s/it, lr: 1.0e-04 loss: 1.292e-01]
Xiang_wan_2_2_lora:  10%|###7                                   | 2859/30000 [6:26:46<59:27:19,  7.89s/it, lr: 1.0e-04 loss: 6.864e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2859/30000 [6:26:46<59:27:19,  7.89s/it, lr: 1.0e-04 loss: 6.864e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2860/30000 [6:26:56<59:27:11,  7.89s/it, lr: 1.0e-04 loss: 6.559e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2860/30000 [6:26:56<59:27:11,  7.89s/it, lr: 1.0e-04 loss: 6.559e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2861/30000 [6:26:56<48:21:05,  6.41s/it, lr: 1.0e-04 loss: 6.559e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2861/30000 [6:26:56<48:21:05,  6.41s/it, lr: 1.0e-04 loss: 6.559e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2861/30000 [6:27:03<48:21:05,  6.41s/it, lr: 1.0e-04 loss: 8.383e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2861/30000 [6:27:03<48:21:05,  6.41s/it, lr: 1.0e-04 loss: 8.383e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2862/30000 [6:27:03<49:11:06,  6.52s/it, lr: 1.0e-04 loss: 8.383e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2862/30000 [6:27:03<49:11:06,  6.52s/it, lr: 1.0e-04 loss: 8.383e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2862/30000 [6:27:09<49:11:06,  6.52s/it, lr: 1.0e-04 loss: 4.442e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2862/30000 [6:27:09<49:11:06,  6.52s/it, lr: 1.0e-04 loss: 4.442e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2863/30000 [6:27:09<49:19:47,  6.54s/it, lr: 1.0e-04 loss: 4.442e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2863/30000 [6:27:09<49:19:47,  6.54s/it, lr: 1.0e-04 loss: 4.442e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2863/30000 [6:27:16<49:19:47,  6.54s/it, lr: 1.0e-04 loss: 9.184e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2863/30000 [6:27:16<49:19:47,  6.54s/it, lr: 1.0e-04 loss: 9.184e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2864/30000 [6:27:16<49:19:13,  6.54s/it, lr: 1.0e-04 loss: 9.184e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2864/30000 [6:27:16<49:19:13,  6.54s/it, lr: 1.0e-04 loss: 9.184e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2864/30000 [6:27:22<49:19:13,  6.54s/it, lr: 1.0e-04 loss: 6.102e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2864/30000 [6:27:22<49:19:13,  6.54s/it, lr: 1.0e-04 loss: 6.102e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2865/30000 [6:27:22<49:16:03,  6.54s/it, lr: 1.0e-04 loss: 6.102e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2865/30000 [6:27:22<49:16:03,  6.54s/it, lr: 1.0e-04 loss: 6.102e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2865/30000 [6:27:32<49:16:03,  6.54s/it, lr: 1.0e-04 loss: 5.603e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2865/30000 [6:27:32<49:16:03,  6.54s/it, lr: 1.0e-04 loss: 5.603e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2866/30000 [6:27:32<55:16:16,  7.33s/it, lr: 1.0e-04 loss: 5.603e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2866/30000 [6:27:32<55:16:16,  7.33s/it, lr: 1.0e-04 loss: 5.603e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2866/30000 [6:27:41<55:16:16,  7.33s/it, lr: 1.0e-04 loss: 4.893e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2866/30000 [6:27:41<55:16:16,  7.33s/it, lr: 1.0e-04 loss: 4.893e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2867/30000 [6:27:41<59:38:52,  7.91s/it, lr: 1.0e-04 loss: 4.893e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2867/30000 [6:27:41<59:38:52,  7.91s/it, lr: 1.0e-04 loss: 4.893e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2867/30000 [6:27:47<59:38:52,  7.91s/it, lr: 1.0e-04 loss: 7.911e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2867/30000 [6:27:47<59:38:52,  7.91s/it, lr: 1.0e-04 loss: 7.911e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2868/30000 [6:27:47<56:37:47,  7.51s/it, lr: 1.0e-04 loss: 7.911e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2868/30000 [6:27:47<56:37:47,  7.51s/it, lr: 1.0e-04 loss: 7.911e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2868/30000 [6:27:54<56:37:47,  7.51s/it, lr: 1.0e-04 loss: 1.101e-01]
Xiang_wan_2_2_lora:  10%|###7                                   | 2868/30000 [6:27:54<56:37:47,  7.51s/it, lr: 1.0e-04 loss: 1.101e-01]
Xiang_wan_2_2_lora:  10%|###7                                   | 2869/30000 [6:27:54<54:31:12,  7.23s/it, lr: 1.0e-04 loss: 1.101e-01]
Xiang_wan_2_2_lora:  10%|###7                                   | 2869/30000 [6:27:54<54:31:12,  7.23s/it, lr: 1.0e-04 loss: 1.101e-01]
Xiang_wan_2_2_lora:  10%|###7                                   | 2869/30000 [6:28:03<54:31:12,  7.23s/it, lr: 1.0e-04 loss: 7.022e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2869/30000 [6:28:03<54:31:12,  7.23s/it, lr: 1.0e-04 loss: 7.022e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2870/30000 [6:28:13<54:31:05,  7.23s/it, lr: 1.0e-04 loss: 8.043e-01]
Xiang_wan_2_2_lora:  10%|###7                                   | 2870/30000 [6:28:13<54:31:05,  7.23s/it, lr: 1.0e-04 loss: 8.043e-01]
Xiang_wan_2_2_lora:  10%|###7                                   | 2871/30000 [6:28:13<45:40:59,  6.06s/it, lr: 1.0e-04 loss: 8.043e-01]
Xiang_wan_2_2_lora:  10%|###7                                   | 2871/30000 [6:28:13<45:40:59,  6.06s/it, lr: 1.0e-04 loss: 8.043e-01]
Xiang_wan_2_2_lora:  10%|###7                                   | 2871/30000 [6:28:22<45:40:59,  6.06s/it, lr: 1.0e-04 loss: 8.882e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2871/30000 [6:28:22<45:40:59,  6.06s/it, lr: 1.0e-04 loss: 8.882e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2872/30000 [6:28:22<51:51:47,  6.88s/it, lr: 1.0e-04 loss: 8.882e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2872/30000 [6:28:22<51:51:47,  6.88s/it, lr: 1.0e-04 loss: 8.882e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2872/30000 [6:28:32<51:51:47,  6.88s/it, lr: 1.0e-04 loss: 8.423e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2872/30000 [6:28:32<51:51:47,  6.88s/it, lr: 1.0e-04 loss: 8.423e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2873/30000 [6:28:32<56:46:05,  7.53s/it, lr: 1.0e-04 loss: 8.423e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2873/30000 [6:28:32<56:46:05,  7.53s/it, lr: 1.0e-04 loss: 8.423e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2873/30000 [6:28:38<56:46:05,  7.53s/it, lr: 1.0e-04 loss: 7.917e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2873/30000 [6:28:38<56:46:05,  7.53s/it, lr: 1.0e-04 loss: 7.917e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2874/30000 [6:28:38<54:47:39,  7.27s/it, lr: 1.0e-04 loss: 7.917e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2874/30000 [6:28:38<54:47:39,  7.27s/it, lr: 1.0e-04 loss: 7.917e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2874/30000 [6:28:47<54:47:39,  7.27s/it, lr: 1.0e-04 loss: 8.241e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2874/30000 [6:28:47<54:47:39,  7.27s/it, lr: 1.0e-04 loss: 8.241e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2875/30000 [6:28:47<59:01:44,  7.83s/it, lr: 1.0e-04 loss: 8.241e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2875/30000 [6:28:47<59:01:44,  7.83s/it, lr: 1.0e-04 loss: 8.241e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2875/30000 [6:28:57<59:01:44,  7.83s/it, lr: 1.0e-04 loss: 6.629e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2875/30000 [6:28:57<59:01:44,  7.83s/it, lr: 1.0e-04 loss: 6.629e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2876/30000 [6:28:57<62:15:14,  8.26s/it, lr: 1.0e-04 loss: 6.629e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2876/30000 [6:28:57<62:15:14,  8.26s/it, lr: 1.0e-04 loss: 6.629e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2876/30000 [6:29:06<62:15:14,  8.26s/it, lr: 1.0e-04 loss: 2.309e-01]
Xiang_wan_2_2_lora:  10%|###7                                   | 2876/30000 [6:29:06<62:15:14,  8.26s/it, lr: 1.0e-04 loss: 2.309e-01]
Xiang_wan_2_2_lora:  10%|###7                                   | 2877/30000 [6:29:06<65:01:22,  8.63s/it, lr: 1.0e-04 loss: 2.309e-01]
Xiang_wan_2_2_lora:  10%|###7                                   | 2877/30000 [6:29:06<65:01:22,  8.63s/it, lr: 1.0e-04 loss: 2.309e-01]
Xiang_wan_2_2_lora:  10%|###7                                   | 2877/30000 [6:29:13<65:01:22,  8.63s/it, lr: 1.0e-04 loss: 7.962e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2877/30000 [6:29:13<65:01:22,  8.63s/it, lr: 1.0e-04 loss: 7.962e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2878/30000 [6:29:13<60:33:24,  8.04s/it, lr: 1.0e-04 loss: 7.962e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2878/30000 [6:29:13<60:33:24,  8.04s/it, lr: 1.0e-04 loss: 7.962e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2878/30000 [6:29:19<60:33:24,  8.04s/it, lr: 1.0e-04 loss: 8.802e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2878/30000 [6:29:19<60:33:24,  8.04s/it, lr: 1.0e-04 loss: 8.802e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2879/30000 [6:29:19<57:20:01,  7.61s/it, lr: 1.0e-04 loss: 8.802e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2879/30000 [6:29:19<57:20:01,  7.61s/it, lr: 1.0e-04 loss: 8.802e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2879/30000 [6:29:26<57:20:01,  7.61s/it, lr: 1.0e-04 loss: 7.016e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2879/30000 [6:29:26<57:20:01,  7.61s/it, lr: 1.0e-04 loss: 7.016e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2880/30000 [6:29:33<57:19:53,  7.61s/it, lr: 1.0e-04 loss: 7.449e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2880/30000 [6:29:33<57:19:53,  7.61s/it, lr: 1.0e-04 loss: 7.449e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2881/30000 [6:29:33<42:29:23,  5.64s/it, lr: 1.0e-04 loss: 7.449e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2881/30000 [6:29:33<42:29:23,  5.64s/it, lr: 1.0e-04 loss: 7.449e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2881/30000 [6:29:42<42:29:23,  5.64s/it, lr: 1.0e-04 loss: 6.842e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2881/30000 [6:29:42<42:29:23,  5.64s/it, lr: 1.0e-04 loss: 6.842e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2882/30000 [6:29:42<49:16:22,  6.54s/it, lr: 1.0e-04 loss: 6.842e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2882/30000 [6:29:42<49:16:22,  6.54s/it, lr: 1.0e-04 loss: 6.842e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2882/30000 [6:29:51<49:16:22,  6.54s/it, lr: 1.0e-04 loss: 4.943e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2882/30000 [6:29:51<49:16:22,  6.54s/it, lr: 1.0e-04 loss: 4.943e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2883/30000 [6:29:51<54:49:22,  7.28s/it, lr: 1.0e-04 loss: 4.943e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2883/30000 [6:29:51<54:49:22,  7.28s/it, lr: 1.0e-04 loss: 4.943e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2883/30000 [6:29:58<54:49:22,  7.28s/it, lr: 1.0e-04 loss: 1.139e-01]
Xiang_wan_2_2_lora:  10%|###7                                   | 2883/30000 [6:29:58<54:49:22,  7.28s/it, lr: 1.0e-04 loss: 1.139e-01]
Xiang_wan_2_2_lora:  10%|###7                                   | 2884/30000 [6:29:58<53:13:11,  7.07s/it, lr: 1.0e-04 loss: 1.139e-01]
Xiang_wan_2_2_lora:  10%|###7                                   | 2884/30000 [6:29:58<53:13:11,  7.07s/it, lr: 1.0e-04 loss: 1.139e-01]
Xiang_wan_2_2_lora:  10%|###7                                   | 2884/30000 [6:30:07<53:13:11,  7.07s/it, lr: 1.0e-04 loss: 7.294e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2884/30000 [6:30:07<53:13:11,  7.07s/it, lr: 1.0e-04 loss: 7.294e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2885/30000 [6:30:07<58:21:42,  7.75s/it, lr: 1.0e-04 loss: 7.294e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2885/30000 [6:30:07<58:21:42,  7.75s/it, lr: 1.0e-04 loss: 7.294e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2885/30000 [6:30:17<58:21:42,  7.75s/it, lr: 1.0e-04 loss: 8.545e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2885/30000 [6:30:17<58:21:42,  7.75s/it, lr: 1.0e-04 loss: 8.545e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2886/30000 [6:30:17<61:44:49,  8.20s/it, lr: 1.0e-04 loss: 8.545e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2886/30000 [6:30:17<61:44:49,  8.20s/it, lr: 1.0e-04 loss: 8.545e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2886/30000 [6:30:26<61:44:49,  8.20s/it, lr: 1.0e-04 loss: 5.872e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2886/30000 [6:30:26<61:44:49,  8.20s/it, lr: 1.0e-04 loss: 5.872e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2887/30000 [6:30:26<64:17:40,  8.54s/it, lr: 1.0e-04 loss: 5.872e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2887/30000 [6:30:26<64:17:40,  8.54s/it, lr: 1.0e-04 loss: 5.872e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2887/30000 [6:30:35<64:17:40,  8.54s/it, lr: 1.0e-04 loss: 5.765e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2887/30000 [6:30:35<64:17:40,  8.54s/it, lr: 1.0e-04 loss: 5.765e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2888/30000 [6:30:35<66:10:12,  8.79s/it, lr: 1.0e-04 loss: 5.765e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2888/30000 [6:30:35<66:10:12,  8.79s/it, lr: 1.0e-04 loss: 5.765e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2888/30000 [6:30:45<66:10:12,  8.79s/it, lr: 1.0e-04 loss: 1.008e+00]
Xiang_wan_2_2_lora:  10%|###7                                   | 2888/30000 [6:30:45<66:10:12,  8.79s/it, lr: 1.0e-04 loss: 1.008e+00]
Xiang_wan_2_2_lora:  10%|###7                                   | 2889/30000 [6:30:45<67:28:39,  8.96s/it, lr: 1.0e-04 loss: 1.008e+00]
Xiang_wan_2_2_lora:  10%|###7                                   | 2889/30000 [6:30:45<67:28:39,  8.96s/it, lr: 1.0e-04 loss: 1.008e+00]
Xiang_wan_2_2_lora:  10%|###7                                   | 2889/30000 [6:30:54<67:28:39,  8.96s/it, lr: 1.0e-04 loss: 8.718e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2889/30000 [6:30:54<67:28:39,  8.96s/it, lr: 1.0e-04 loss: 8.718e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2890/30000 [6:31:04<67:28:30,  8.96s/it, lr: 1.0e-04 loss: 5.385e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2890/30000 [6:31:04<67:28:30,  8.96s/it, lr: 1.0e-04 loss: 5.385e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2891/30000 [6:31:04<52:46:03,  7.01s/it, lr: 1.0e-04 loss: 5.385e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2891/30000 [6:31:04<52:46:03,  7.01s/it, lr: 1.0e-04 loss: 5.385e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2891/30000 [6:31:10<52:46:03,  7.01s/it, lr: 1.0e-04 loss: 7.557e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2891/30000 [6:31:10<52:46:03,  7.01s/it, lr: 1.0e-04 loss: 7.557e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2892/30000 [6:31:10<52:25:35,  6.96s/it, lr: 1.0e-04 loss: 7.557e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2892/30000 [6:31:10<52:25:35,  6.96s/it, lr: 1.0e-04 loss: 7.557e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2892/30000 [6:31:17<52:25:35,  6.96s/it, lr: 1.0e-04 loss: 1.829e-01]
Xiang_wan_2_2_lora:  10%|###7                                   | 2892/30000 [6:31:17<52:25:35,  6.96s/it, lr: 1.0e-04 loss: 1.829e-01]
Xiang_wan_2_2_lora:  10%|###7                                   | 2893/30000 [6:31:17<51:42:34,  6.87s/it, lr: 1.0e-04 loss: 1.829e-01]
Xiang_wan_2_2_lora:  10%|###7                                   | 2893/30000 [6:31:17<51:42:34,  6.87s/it, lr: 1.0e-04 loss: 1.829e-01]
Xiang_wan_2_2_lora:  10%|###7                                   | 2893/30000 [6:31:24<51:42:34,  6.87s/it, lr: 1.0e-04 loss: 7.012e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2893/30000 [6:31:24<51:42:34,  6.87s/it, lr: 1.0e-04 loss: 7.012e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2894/30000 [6:31:24<51:01:59,  6.78s/it, lr: 1.0e-04 loss: 7.012e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2894/30000 [6:31:24<51:01:59,  6.78s/it, lr: 1.0e-04 loss: 7.012e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2894/30000 [6:31:33<51:01:59,  6.78s/it, lr: 1.0e-04 loss: 4.246e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2894/30000 [6:31:33<51:01:59,  6.78s/it, lr: 1.0e-04 loss: 4.246e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2895/30000 [6:31:33<56:28:30,  7.50s/it, lr: 1.0e-04 loss: 4.246e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2895/30000 [6:31:33<56:28:30,  7.50s/it, lr: 1.0e-04 loss: 4.246e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2895/30000 [6:31:40<56:28:30,  7.50s/it, lr: 1.0e-04 loss: 9.760e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2895/30000 [6:31:40<56:28:30,  7.50s/it, lr: 1.0e-04 loss: 9.760e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2896/30000 [6:31:40<54:37:18,  7.25s/it, lr: 1.0e-04 loss: 9.760e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2896/30000 [6:31:40<54:37:18,  7.25s/it, lr: 1.0e-04 loss: 9.760e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2896/30000 [6:31:49<54:37:18,  7.25s/it, lr: 1.0e-04 loss: 5.778e-01]
Xiang_wan_2_2_lora:  10%|###7                                   | 2896/30000 [6:31:49<54:37:18,  7.25s/it, lr: 1.0e-04 loss: 5.778e-01]
Xiang_wan_2_2_lora:  10%|###7                                   | 2897/30000 [6:31:49<59:04:05,  7.85s/it, lr: 1.0e-04 loss: 5.778e-01]
Xiang_wan_2_2_lora:  10%|###7                                   | 2897/30000 [6:31:49<59:04:05,  7.85s/it, lr: 1.0e-04 loss: 5.778e-01]
Xiang_wan_2_2_lora:  10%|###7                                   | 2897/30000 [6:31:58<59:04:05,  7.85s/it, lr: 1.0e-04 loss: 4.653e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2897/30000 [6:31:58<59:04:05,  7.85s/it, lr: 1.0e-04 loss: 4.653e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2898/30000 [6:31:58<62:27:36,  8.30s/it, lr: 1.0e-04 loss: 4.653e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2898/30000 [6:31:58<62:27:36,  8.30s/it, lr: 1.0e-04 loss: 4.653e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2898/30000 [6:32:05<62:27:36,  8.30s/it, lr: 1.0e-04 loss: 7.726e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2898/30000 [6:32:05<62:27:36,  8.30s/it, lr: 1.0e-04 loss: 7.726e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2899/30000 [6:32:05<58:40:19,  7.79s/it, lr: 1.0e-04 loss: 7.726e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2899/30000 [6:32:05<58:40:19,  7.79s/it, lr: 1.0e-04 loss: 7.726e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2899/30000 [6:32:12<58:40:19,  7.79s/it, lr: 1.0e-04 loss: 7.492e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2899/30000 [6:32:12<58:40:19,  7.79s/it, lr: 1.0e-04 loss: 7.492e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2900/30000 [6:32:21<58:40:11,  7.79s/it, lr: 1.0e-04 loss: 8.003e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2900/30000 [6:32:21<58:40:11,  7.79s/it, lr: 1.0e-04 loss: 8.003e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2901/30000 [6:32:21<47:51:11,  6.36s/it, lr: 1.0e-04 loss: 8.003e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2901/30000 [6:32:21<47:51:11,  6.36s/it, lr: 1.0e-04 loss: 8.003e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2901/30000 [6:32:27<47:51:11,  6.36s/it, lr: 1.0e-04 loss: 8.473e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2901/30000 [6:32:27<47:51:11,  6.36s/it, lr: 1.0e-04 loss: 8.473e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2902/30000 [6:32:27<48:14:05,  6.41s/it, lr: 1.0e-04 loss: 8.473e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2902/30000 [6:32:27<48:14:05,  6.41s/it, lr: 1.0e-04 loss: 8.473e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2902/30000 [6:32:34<48:14:05,  6.41s/it, lr: 1.0e-04 loss: 7.878e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2902/30000 [6:32:34<48:14:05,  6.41s/it, lr: 1.0e-04 loss: 7.878e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2903/30000 [6:32:34<48:36:53,  6.46s/it, lr: 1.0e-04 loss: 7.878e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2903/30000 [6:32:34<48:36:53,  6.46s/it, lr: 1.0e-04 loss: 7.878e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2903/30000 [6:32:41<48:36:53,  6.46s/it, lr: 1.0e-04 loss: 1.156e-01]
Xiang_wan_2_2_lora:  10%|###7                                   | 2903/30000 [6:32:41<48:36:53,  6.46s/it, lr: 1.0e-04 loss: 1.156e-01]
Xiang_wan_2_2_lora:  10%|###7                                   | 2904/30000 [6:32:41<48:52:51,  6.49s/it, lr: 1.0e-04 loss: 1.156e-01]
Xiang_wan_2_2_lora:  10%|###7                                   | 2904/30000 [6:32:41<48:52:51,  6.49s/it, lr: 1.0e-04 loss: 1.156e-01]
Xiang_wan_2_2_lora:  10%|###7                                   | 2904/30000 [6:32:47<48:52:51,  6.49s/it, lr: 1.0e-04 loss: 5.937e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2904/30000 [6:32:47<48:52:51,  6.49s/it, lr: 1.0e-04 loss: 5.937e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2905/30000 [6:32:47<49:04:41,  6.52s/it, lr: 1.0e-04 loss: 5.937e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2905/30000 [6:32:47<49:04:41,  6.52s/it, lr: 1.0e-04 loss: 5.937e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2905/30000 [6:32:54<49:04:41,  6.52s/it, lr: 1.0e-04 loss: 7.795e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2905/30000 [6:32:54<49:04:41,  6.52s/it, lr: 1.0e-04 loss: 7.795e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2906/30000 [6:32:54<49:03:17,  6.52s/it, lr: 1.0e-04 loss: 7.795e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2906/30000 [6:32:54<49:03:17,  6.52s/it, lr: 1.0e-04 loss: 7.795e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2906/30000 [6:33:01<49:03:17,  6.52s/it, lr: 1.0e-04 loss: 7.062e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2906/30000 [6:33:01<49:03:17,  6.52s/it, lr: 1.0e-04 loss: 7.062e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2907/30000 [6:33:01<49:46:45,  6.61s/it, lr: 1.0e-04 loss: 7.062e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2907/30000 [6:33:01<49:46:45,  6.61s/it, lr: 1.0e-04 loss: 7.062e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2907/30000 [6:33:07<49:46:45,  6.61s/it, lr: 1.0e-04 loss: 8.506e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2907/30000 [6:33:07<49:46:45,  6.61s/it, lr: 1.0e-04 loss: 8.506e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2908/30000 [6:33:07<49:34:55,  6.59s/it, lr: 1.0e-04 loss: 8.506e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2908/30000 [6:33:07<49:34:55,  6.59s/it, lr: 1.0e-04 loss: 8.506e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2908/30000 [6:33:17<49:34:55,  6.59s/it, lr: 1.0e-04 loss: 8.589e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2908/30000 [6:33:17<49:34:55,  6.59s/it, lr: 1.0e-04 loss: 8.589e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2909/30000 [6:33:17<55:44:29,  7.41s/it, lr: 1.0e-04 loss: 8.589e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2909/30000 [6:33:17<55:44:29,  7.41s/it, lr: 1.0e-04 loss: 8.589e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2909/30000 [6:33:26<55:44:29,  7.41s/it, lr: 1.0e-04 loss: 7.829e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2909/30000 [6:33:26<55:44:29,  7.41s/it, lr: 1.0e-04 loss: 7.829e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2910/30000 [6:33:35<55:44:21,  7.41s/it, lr: 1.0e-04 loss: 6.120e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2910/30000 [6:33:35<55:44:21,  7.41s/it, lr: 1.0e-04 loss: 6.120e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2911/30000 [6:33:35<46:18:30,  6.15s/it, lr: 1.0e-04 loss: 6.120e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2911/30000 [6:33:35<46:18:30,  6.15s/it, lr: 1.0e-04 loss: 6.120e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2911/30000 [6:33:45<46:18:30,  6.15s/it, lr: 1.0e-04 loss: 6.051e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2911/30000 [6:33:45<46:18:30,  6.15s/it, lr: 1.0e-04 loss: 6.051e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2912/30000 [6:33:45<52:11:33,  6.94s/it, lr: 1.0e-04 loss: 6.051e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2912/30000 [6:33:45<52:11:33,  6.94s/it, lr: 1.0e-04 loss: 6.051e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2912/30000 [6:33:51<52:11:33,  6.94s/it, lr: 1.0e-04 loss: 1.202e-01]
Xiang_wan_2_2_lora:  10%|###7                                   | 2912/30000 [6:33:51<52:11:33,  6.94s/it, lr: 1.0e-04 loss: 1.202e-01]
Xiang_wan_2_2_lora:  10%|###7                                   | 2913/30000 [6:33:51<51:33:27,  6.85s/it, lr: 1.0e-04 loss: 1.202e-01]
Xiang_wan_2_2_lora:  10%|###7                                   | 2913/30000 [6:33:51<51:33:27,  6.85s/it, lr: 1.0e-04 loss: 1.202e-01]
Xiang_wan_2_2_lora:  10%|###7                                   | 2913/30000 [6:33:58<51:33:27,  6.85s/it, lr: 1.0e-04 loss: 9.007e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2913/30000 [6:33:58<51:33:27,  6.85s/it, lr: 1.0e-04 loss: 9.007e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2914/30000 [6:33:58<50:53:23,  6.76s/it, lr: 1.0e-04 loss: 9.007e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2914/30000 [6:33:58<50:53:23,  6.76s/it, lr: 1.0e-04 loss: 9.007e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2914/30000 [6:34:07<50:53:23,  6.76s/it, lr: 1.0e-04 loss: 4.560e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2914/30000 [6:34:07<50:53:23,  6.76s/it, lr: 1.0e-04 loss: 4.560e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2915/30000 [6:34:07<56:41:55,  7.54s/it, lr: 1.0e-04 loss: 4.560e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2915/30000 [6:34:07<56:41:55,  7.54s/it, lr: 1.0e-04 loss: 4.560e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2915/30000 [6:34:17<56:41:55,  7.54s/it, lr: 1.0e-04 loss: 7.277e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2915/30000 [6:34:17<56:41:55,  7.54s/it, lr: 1.0e-04 loss: 7.277e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2916/30000 [6:34:17<60:34:17,  8.05s/it, lr: 1.0e-04 loss: 7.277e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2916/30000 [6:34:17<60:34:17,  8.05s/it, lr: 1.0e-04 loss: 7.277e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2916/30000 [6:34:26<60:34:17,  8.05s/it, lr: 1.0e-04 loss: 6.865e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2916/30000 [6:34:26<60:34:17,  8.05s/it, lr: 1.0e-04 loss: 6.865e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2917/30000 [6:34:26<63:23:45,  8.43s/it, lr: 1.0e-04 loss: 6.865e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2917/30000 [6:34:26<63:23:45,  8.43s/it, lr: 1.0e-04 loss: 6.865e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2917/30000 [6:34:35<63:23:45,  8.43s/it, lr: 1.0e-04 loss: 8.717e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2917/30000 [6:34:35<63:23:45,  8.43s/it, lr: 1.0e-04 loss: 8.717e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2918/30000 [6:34:35<65:23:54,  8.69s/it, lr: 1.0e-04 loss: 8.717e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2918/30000 [6:34:35<65:23:54,  8.69s/it, lr: 1.0e-04 loss: 8.717e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2918/30000 [6:34:45<65:23:54,  8.69s/it, lr: 1.0e-04 loss: 1.092e-01]
Xiang_wan_2_2_lora:  10%|###7                                   | 2918/30000 [6:34:45<65:23:54,  8.69s/it, lr: 1.0e-04 loss: 1.092e-01]
Xiang_wan_2_2_lora:  10%|###7                                   | 2919/30000 [6:34:45<66:50:27,  8.89s/it, lr: 1.0e-04 loss: 1.092e-01]
Xiang_wan_2_2_lora:  10%|###7                                   | 2919/30000 [6:34:45<66:50:27,  8.89s/it, lr: 1.0e-04 loss: 1.092e-01]
Xiang_wan_2_2_lora:  10%|###7                                   | 2919/30000 [6:34:51<66:50:27,  8.89s/it, lr: 1.0e-04 loss: 1.063e-01]
Xiang_wan_2_2_lora:  10%|###7                                   | 2919/30000 [6:34:51<66:50:27,  8.89s/it, lr: 1.0e-04 loss: 1.063e-01]
Xiang_wan_2_2_lora:  10%|###7                                   | 2920/30000 [6:34:58<66:50:18,  8.89s/it, lr: 1.0e-04 loss: 5.257e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2920/30000 [6:34:58<66:50:18,  8.89s/it, lr: 1.0e-04 loss: 5.257e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2921/30000 [6:34:58<47:38:16,  6.33s/it, lr: 1.0e-04 loss: 5.257e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2921/30000 [6:34:58<47:38:16,  6.33s/it, lr: 1.0e-04 loss: 5.257e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2921/30000 [6:35:05<47:38:16,  6.33s/it, lr: 1.0e-04 loss: 6.499e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2921/30000 [6:35:05<47:38:16,  6.33s/it, lr: 1.0e-04 loss: 6.499e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2922/30000 [6:35:05<48:29:36,  6.45s/it, lr: 1.0e-04 loss: 6.499e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2922/30000 [6:35:05<48:29:36,  6.45s/it, lr: 1.0e-04 loss: 6.499e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2922/30000 [6:35:11<48:29:36,  6.45s/it, lr: 1.0e-04 loss: 6.268e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2922/30000 [6:35:11<48:29:36,  6.45s/it, lr: 1.0e-04 loss: 6.268e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2923/30000 [6:35:11<48:39:52,  6.47s/it, lr: 1.0e-04 loss: 6.268e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2923/30000 [6:35:11<48:39:52,  6.47s/it, lr: 1.0e-04 loss: 6.268e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2923/30000 [6:35:18<48:39:52,  6.47s/it, lr: 1.0e-04 loss: 8.995e-02]
Xiang_wan_2_2_lora:  10%|###7                                   | 2923/30000 [6:35:18<48:39:52,  6.47s/it, lr: 1.0e-04 loss: 8.995e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2924/30000 [6:35:18<48:46:39,  6.49s/it, lr: 1.0e-04 loss: 8.995e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2924/30000 [6:35:18<48:46:39,  6.49s/it, lr: 1.0e-04 loss: 8.995e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2924/30000 [6:35:27<48:46:39,  6.49s/it, lr: 1.0e-04 loss: 5.295e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2924/30000 [6:35:27<48:46:39,  6.49s/it, lr: 1.0e-04 loss: 5.295e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2925/30000 [6:35:27<54:51:06,  7.29s/it, lr: 1.0e-04 loss: 5.295e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2925/30000 [6:35:27<54:51:06,  7.29s/it, lr: 1.0e-04 loss: 5.295e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2925/30000 [6:35:36<54:51:06,  7.29s/it, lr: 1.0e-04 loss: 7.843e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2925/30000 [6:35:36<54:51:06,  7.29s/it, lr: 1.0e-04 loss: 7.843e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2926/30000 [6:35:36<59:10:12,  7.87s/it, lr: 1.0e-04 loss: 7.843e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2926/30000 [6:35:36<59:10:12,  7.87s/it, lr: 1.0e-04 loss: 7.843e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2926/30000 [6:35:43<59:10:12,  7.87s/it, lr: 1.0e-04 loss: 9.355e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2926/30000 [6:35:43<59:10:12,  7.87s/it, lr: 1.0e-04 loss: 9.355e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2927/30000 [6:35:43<56:24:17,  7.50s/it, lr: 1.0e-04 loss: 9.355e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2927/30000 [6:35:43<56:24:17,  7.50s/it, lr: 1.0e-04 loss: 9.355e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2927/30000 [6:35:52<56:24:17,  7.50s/it, lr: 1.0e-04 loss: 6.425e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2927/30000 [6:35:52<56:24:17,  7.50s/it, lr: 1.0e-04 loss: 6.425e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2928/30000 [6:35:52<60:26:40,  8.04s/it, lr: 1.0e-04 loss: 6.425e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2928/30000 [6:35:52<60:26:40,  8.04s/it, lr: 1.0e-04 loss: 6.425e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2928/30000 [6:36:02<60:26:40,  8.04s/it, lr: 1.0e-04 loss: 2.656e-01]
Xiang_wan_2_2_lora:  10%|###8                                   | 2928/30000 [6:36:02<60:26:40,  8.04s/it, lr: 1.0e-04 loss: 2.656e-01]
Xiang_wan_2_2_lora:  10%|###8                                   | 2929/30000 [6:36:02<63:17:24,  8.42s/it, lr: 1.0e-04 loss: 2.656e-01]
Xiang_wan_2_2_lora:  10%|###8                                   | 2929/30000 [6:36:02<63:17:24,  8.42s/it, lr: 1.0e-04 loss: 2.656e-01]
Xiang_wan_2_2_lora:  10%|###8                                   | 2929/30000 [6:36:11<63:17:24,  8.42s/it, lr: 1.0e-04 loss: 5.892e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2929/30000 [6:36:11<63:17:24,  8.42s/it, lr: 1.0e-04 loss: 5.892e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2930/30000 [6:36:18<63:17:15,  8.42s/it, lr: 1.0e-04 loss: 5.222e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2930/30000 [6:36:18<63:17:15,  8.42s/it, lr: 1.0e-04 loss: 5.222e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2931/30000 [6:36:18<45:39:18,  6.07s/it, lr: 1.0e-04 loss: 5.222e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2931/30000 [6:36:18<45:39:18,  6.07s/it, lr: 1.0e-04 loss: 5.222e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2931/30000 [6:36:27<45:39:18,  6.07s/it, lr: 1.0e-04 loss: 5.812e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2931/30000 [6:36:27<45:39:18,  6.07s/it, lr: 1.0e-04 loss: 5.812e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2932/30000 [6:36:27<51:35:21,  6.86s/it, lr: 1.0e-04 loss: 5.812e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2932/30000 [6:36:27<51:35:21,  6.86s/it, lr: 1.0e-04 loss: 5.812e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2932/30000 [6:36:33<51:35:21,  6.86s/it, lr: 1.0e-04 loss: 5.197e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2932/30000 [6:36:33<51:35:21,  6.86s/it, lr: 1.0e-04 loss: 5.197e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2933/30000 [6:36:33<50:57:57,  6.78s/it, lr: 1.0e-04 loss: 5.197e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2933/30000 [6:36:33<50:57:57,  6.78s/it, lr: 1.0e-04 loss: 5.197e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2933/30000 [6:36:40<50:57:57,  6.78s/it, lr: 1.0e-04 loss: 8.825e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2933/30000 [6:36:40<50:57:57,  6.78s/it, lr: 1.0e-04 loss: 8.825e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2934/30000 [6:36:40<50:34:55,  6.73s/it, lr: 1.0e-04 loss: 8.825e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2934/30000 [6:36:40<50:34:55,  6.73s/it, lr: 1.0e-04 loss: 8.825e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2934/30000 [6:36:47<50:34:55,  6.73s/it, lr: 1.0e-04 loss: 6.047e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2934/30000 [6:36:47<50:34:55,  6.73s/it, lr: 1.0e-04 loss: 6.047e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2935/30000 [6:36:47<50:17:55,  6.69s/it, lr: 1.0e-04 loss: 6.047e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2935/30000 [6:36:47<50:17:55,  6.69s/it, lr: 1.0e-04 loss: 6.047e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2935/30000 [6:36:53<50:17:55,  6.69s/it, lr: 1.0e-04 loss: 8.738e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2935/30000 [6:36:53<50:17:55,  6.69s/it, lr: 1.0e-04 loss: 8.738e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2936/30000 [6:36:53<50:05:53,  6.66s/it, lr: 1.0e-04 loss: 8.738e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2936/30000 [6:36:53<50:05:53,  6.66s/it, lr: 1.0e-04 loss: 8.738e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2936/30000 [6:37:00<50:05:53,  6.66s/it, lr: 1.0e-04 loss: 5.286e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2936/30000 [6:37:00<50:05:53,  6.66s/it, lr: 1.0e-04 loss: 5.286e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2937/30000 [6:37:00<50:27:17,  6.71s/it, lr: 1.0e-04 loss: 5.286e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2937/30000 [6:37:00<50:27:17,  6.71s/it, lr: 1.0e-04 loss: 5.286e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2937/30000 [6:37:09<50:27:17,  6.71s/it, lr: 1.0e-04 loss: 7.065e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2937/30000 [6:37:09<50:27:17,  6.71s/it, lr: 1.0e-04 loss: 7.065e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2938/30000 [6:37:09<56:07:48,  7.47s/it, lr: 1.0e-04 loss: 7.065e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2938/30000 [6:37:09<56:07:48,  7.47s/it, lr: 1.0e-04 loss: 7.065e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2938/30000 [6:37:19<56:07:48,  7.47s/it, lr: 1.0e-04 loss: 8.216e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2938/30000 [6:37:19<56:07:48,  7.47s/it, lr: 1.0e-04 loss: 8.216e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2939/30000 [6:37:19<60:17:56,  8.02s/it, lr: 1.0e-04 loss: 8.216e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2939/30000 [6:37:19<60:17:56,  8.02s/it, lr: 1.0e-04 loss: 8.216e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2939/30000 [6:37:25<60:17:56,  8.02s/it, lr: 1.0e-04 loss: 8.096e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2939/30000 [6:37:25<60:17:56,  8.02s/it, lr: 1.0e-04 loss: 8.096e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2940/30000 [6:37:32<60:17:48,  8.02s/it, lr: 1.0e-04 loss: 7.201e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2940/30000 [6:37:32<60:17:48,  8.02s/it, lr: 1.0e-04 loss: 7.201e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2941/30000 [6:37:32<44:00:50,  5.86s/it, lr: 1.0e-04 loss: 7.201e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2941/30000 [6:37:32<44:00:50,  5.86s/it, lr: 1.0e-04 loss: 7.201e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2941/30000 [6:37:41<44:00:50,  5.86s/it, lr: 1.0e-04 loss: 5.248e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2941/30000 [6:37:41<44:00:50,  5.86s/it, lr: 1.0e-04 loss: 5.248e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2942/30000 [6:37:41<50:21:03,  6.70s/it, lr: 1.0e-04 loss: 5.248e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2942/30000 [6:37:41<50:21:03,  6.70s/it, lr: 1.0e-04 loss: 5.248e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2942/30000 [6:37:48<50:21:03,  6.70s/it, lr: 1.0e-04 loss: 5.740e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2942/30000 [6:37:48<50:21:03,  6.70s/it, lr: 1.0e-04 loss: 5.740e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2943/30000 [6:37:48<50:10:36,  6.68s/it, lr: 1.0e-04 loss: 5.740e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2943/30000 [6:37:48<50:10:36,  6.68s/it, lr: 1.0e-04 loss: 5.740e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2943/30000 [6:37:57<50:10:36,  6.68s/it, lr: 1.0e-04 loss: 6.563e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2943/30000 [6:37:57<50:10:36,  6.68s/it, lr: 1.0e-04 loss: 6.563e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2944/30000 [6:37:57<55:29:18,  7.38s/it, lr: 1.0e-04 loss: 6.563e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2944/30000 [6:37:57<55:29:18,  7.38s/it, lr: 1.0e-04 loss: 6.563e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2944/30000 [6:38:07<55:29:18,  7.38s/it, lr: 1.0e-04 loss: 5.487e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2944/30000 [6:38:07<55:29:18,  7.38s/it, lr: 1.0e-04 loss: 5.487e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2945/30000 [6:38:07<60:04:35,  7.99s/it, lr: 1.0e-04 loss: 5.487e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2945/30000 [6:38:07<60:04:35,  7.99s/it, lr: 1.0e-04 loss: 5.487e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2945/30000 [6:38:16<60:04:35,  7.99s/it, lr: 1.0e-04 loss: 6.112e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2945/30000 [6:38:16<60:04:35,  7.99s/it, lr: 1.0e-04 loss: 6.112e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2946/30000 [6:38:16<63:00:17,  8.38s/it, lr: 1.0e-04 loss: 6.112e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2946/30000 [6:38:16<63:00:17,  8.38s/it, lr: 1.0e-04 loss: 6.112e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2946/30000 [6:38:23<63:00:17,  8.38s/it, lr: 1.0e-04 loss: 5.295e-01]
Xiang_wan_2_2_lora:  10%|###8                                   | 2946/30000 [6:38:23<63:00:17,  8.38s/it, lr: 1.0e-04 loss: 5.295e-01]
Xiang_wan_2_2_lora:  10%|###8                                   | 2947/30000 [6:38:23<59:07:31,  7.87s/it, lr: 1.0e-04 loss: 5.295e-01]
Xiang_wan_2_2_lora:  10%|###8                                   | 2947/30000 [6:38:23<59:07:31,  7.87s/it, lr: 1.0e-04 loss: 5.295e-01]
Xiang_wan_2_2_lora:  10%|###8                                   | 2947/30000 [6:38:29<59:07:31,  7.87s/it, lr: 1.0e-04 loss: 7.620e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2947/30000 [6:38:29<59:07:31,  7.87s/it, lr: 1.0e-04 loss: 7.620e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2948/30000 [6:38:29<56:16:31,  7.49s/it, lr: 1.0e-04 loss: 7.620e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2948/30000 [6:38:29<56:16:31,  7.49s/it, lr: 1.0e-04 loss: 7.620e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2948/30000 [6:38:39<56:16:31,  7.49s/it, lr: 1.0e-04 loss: 5.161e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2948/30000 [6:38:39<56:16:31,  7.49s/it, lr: 1.0e-04 loss: 5.161e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2949/30000 [6:38:39<60:22:06,  8.03s/it, lr: 1.0e-04 loss: 5.161e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2949/30000 [6:38:39<60:22:06,  8.03s/it, lr: 1.0e-04 loss: 5.161e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2949/30000 [6:38:45<60:22:06,  8.03s/it, lr: 1.0e-04 loss: 9.237e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2949/30000 [6:38:45<60:22:06,  8.03s/it, lr: 1.0e-04 loss: 9.237e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2950/30000 [6:38:52<60:21:58,  8.03s/it, lr: 1.0e-04 loss: 6.180e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2950/30000 [6:38:52<60:21:58,  8.03s/it, lr: 1.0e-04 loss: 6.180e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2951/30000 [6:38:52<43:56:44,  5.85s/it, lr: 1.0e-04 loss: 6.180e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2951/30000 [6:38:52<43:56:44,  5.85s/it, lr: 1.0e-04 loss: 6.180e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2951/30000 [6:38:59<43:56:44,  5.85s/it, lr: 1.0e-04 loss: 8.296e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2951/30000 [6:38:59<43:56:44,  5.85s/it, lr: 1.0e-04 loss: 8.296e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2952/30000 [6:38:59<45:48:33,  6.10s/it, lr: 1.0e-04 loss: 8.296e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2952/30000 [6:38:59<45:48:33,  6.10s/it, lr: 1.0e-04 loss: 8.296e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2952/30000 [6:39:05<45:48:33,  6.10s/it, lr: 1.0e-04 loss: 1.485e-01]
Xiang_wan_2_2_lora:  10%|###8                                   | 2952/30000 [6:39:05<45:48:33,  6.10s/it, lr: 1.0e-04 loss: 1.485e-01]
Xiang_wan_2_2_lora:  10%|###8                                   | 2953/30000 [6:39:05<46:38:31,  6.21s/it, lr: 1.0e-04 loss: 1.485e-01]
Xiang_wan_2_2_lora:  10%|###8                                   | 2953/30000 [6:39:05<46:38:31,  6.21s/it, lr: 1.0e-04 loss: 1.485e-01]
Xiang_wan_2_2_lora:  10%|###8                                   | 2953/30000 [6:39:12<46:38:31,  6.21s/it, lr: 1.0e-04 loss: 9.764e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2953/30000 [6:39:12<46:38:31,  6.21s/it, lr: 1.0e-04 loss: 9.764e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2954/30000 [6:39:12<47:25:44,  6.31s/it, lr: 1.0e-04 loss: 9.764e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2954/30000 [6:39:12<47:25:44,  6.31s/it, lr: 1.0e-04 loss: 9.764e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2954/30000 [6:39:21<47:25:44,  6.31s/it, lr: 1.0e-04 loss: 1.942e-01]
Xiang_wan_2_2_lora:  10%|###8                                   | 2954/30000 [6:39:21<47:25:44,  6.31s/it, lr: 1.0e-04 loss: 1.942e-01]
Xiang_wan_2_2_lora:  10%|###8                                   | 2955/30000 [6:39:21<53:34:16,  7.13s/it, lr: 1.0e-04 loss: 1.942e-01]
Xiang_wan_2_2_lora:  10%|###8                                   | 2955/30000 [6:39:21<53:34:16,  7.13s/it, lr: 1.0e-04 loss: 1.942e-01]
Xiang_wan_2_2_lora:  10%|###8                                   | 2955/30000 [6:39:30<53:34:16,  7.13s/it, lr: 1.0e-04 loss: 7.267e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2955/30000 [6:39:30<53:34:16,  7.13s/it, lr: 1.0e-04 loss: 7.267e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2956/30000 [6:39:30<58:28:39,  7.78s/it, lr: 1.0e-04 loss: 7.267e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2956/30000 [6:39:30<58:28:39,  7.78s/it, lr: 1.0e-04 loss: 7.267e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2956/30000 [6:39:40<58:28:39,  7.78s/it, lr: 1.0e-04 loss: 1.814e-01]
Xiang_wan_2_2_lora:  10%|###8                                   | 2956/30000 [6:39:40<58:28:39,  7.78s/it, lr: 1.0e-04 loss: 1.814e-01]
Xiang_wan_2_2_lora:  10%|###8                                   | 2957/30000 [6:39:40<61:44:51,  8.22s/it, lr: 1.0e-04 loss: 1.814e-01]
Xiang_wan_2_2_lora:  10%|###8                                   | 2957/30000 [6:39:40<61:44:51,  8.22s/it, lr: 1.0e-04 loss: 1.814e-01]
Xiang_wan_2_2_lora:  10%|###8                                   | 2957/30000 [6:39:46<61:44:51,  8.22s/it, lr: 1.0e-04 loss: 8.798e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2957/30000 [6:39:46<61:44:51,  8.22s/it, lr: 1.0e-04 loss: 8.798e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2958/30000 [6:39:46<58:07:45,  7.74s/it, lr: 1.0e-04 loss: 8.798e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2958/30000 [6:39:46<58:07:45,  7.74s/it, lr: 1.0e-04 loss: 8.798e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2958/30000 [6:39:53<58:07:45,  7.74s/it, lr: 1.0e-04 loss: 8.960e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2958/30000 [6:39:53<58:07:45,  7.74s/it, lr: 1.0e-04 loss: 8.960e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2959/30000 [6:39:53<55:28:24,  7.39s/it, lr: 1.0e-04 loss: 8.960e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2959/30000 [6:39:53<55:28:24,  7.39s/it, lr: 1.0e-04 loss: 8.960e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2959/30000 [6:40:02<55:28:24,  7.39s/it, lr: 1.0e-04 loss: 6.964e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2959/30000 [6:40:02<55:28:24,  7.39s/it, lr: 1.0e-04 loss: 6.964e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2960/30000 [6:40:12<55:28:17,  7.39s/it, lr: 1.0e-04 loss: 2.776e-01]
Xiang_wan_2_2_lora:  10%|###8                                   | 2960/30000 [6:40:12<55:28:17,  7.39s/it, lr: 1.0e-04 loss: 2.776e-01]
Xiang_wan_2_2_lora:  10%|###8                                   | 2961/30000 [6:40:12<46:07:35,  6.14s/it, lr: 1.0e-04 loss: 2.776e-01]
Xiang_wan_2_2_lora:  10%|###8                                   | 2961/30000 [6:40:12<46:07:35,  6.14s/it, lr: 1.0e-04 loss: 2.776e-01]
Xiang_wan_2_2_lora:  10%|###8                                   | 2961/30000 [6:40:21<46:07:35,  6.14s/it, lr: 1.0e-04 loss: 5.379e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2961/30000 [6:40:21<46:07:35,  6.14s/it, lr: 1.0e-04 loss: 5.379e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2962/30000 [6:40:21<52:05:38,  6.94s/it, lr: 1.0e-04 loss: 5.379e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2962/30000 [6:40:21<52:05:38,  6.94s/it, lr: 1.0e-04 loss: 5.379e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2962/30000 [6:40:28<52:05:38,  6.94s/it, lr: 1.0e-04 loss: 5.923e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2962/30000 [6:40:28<52:05:38,  6.94s/it, lr: 1.0e-04 loss: 5.923e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2963/30000 [6:40:28<51:21:05,  6.84s/it, lr: 1.0e-04 loss: 5.923e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2963/30000 [6:40:28<51:21:05,  6.84s/it, lr: 1.0e-04 loss: 5.923e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2963/30000 [6:40:34<51:21:05,  6.84s/it, lr: 1.0e-04 loss: 8.219e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2963/30000 [6:40:34<51:21:05,  6.84s/it, lr: 1.0e-04 loss: 8.219e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2964/30000 [6:40:34<50:46:01,  6.76s/it, lr: 1.0e-04 loss: 8.219e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2964/30000 [6:40:34<50:46:01,  6.76s/it, lr: 1.0e-04 loss: 8.219e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2964/30000 [6:40:43<50:46:01,  6.76s/it, lr: 1.0e-04 loss: 3.921e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2964/30000 [6:40:43<50:46:01,  6.76s/it, lr: 1.0e-04 loss: 3.921e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2965/30000 [6:40:43<56:03:05,  7.46s/it, lr: 1.0e-04 loss: 3.921e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2965/30000 [6:40:43<56:03:05,  7.46s/it, lr: 1.0e-04 loss: 3.921e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2965/30000 [6:40:50<56:03:05,  7.46s/it, lr: 1.0e-04 loss: 7.996e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2965/30000 [6:40:50<56:03:05,  7.46s/it, lr: 1.0e-04 loss: 7.996e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2966/30000 [6:40:50<54:16:53,  7.23s/it, lr: 1.0e-04 loss: 7.996e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2966/30000 [6:40:50<54:16:53,  7.23s/it, lr: 1.0e-04 loss: 7.996e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2966/30000 [6:40:57<54:16:53,  7.23s/it, lr: 1.0e-04 loss: 7.801e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2966/30000 [6:40:57<54:16:53,  7.23s/it, lr: 1.0e-04 loss: 7.801e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2967/30000 [6:40:57<53:08:25,  7.08s/it, lr: 1.0e-04 loss: 7.801e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2967/30000 [6:40:57<53:08:25,  7.08s/it, lr: 1.0e-04 loss: 7.801e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2967/30000 [6:41:06<53:08:25,  7.08s/it, lr: 1.0e-04 loss: 7.916e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2967/30000 [6:41:06<53:08:25,  7.08s/it, lr: 1.0e-04 loss: 7.916e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2968/30000 [6:41:06<58:10:28,  7.75s/it, lr: 1.0e-04 loss: 7.916e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2968/30000 [6:41:06<58:10:28,  7.75s/it, lr: 1.0e-04 loss: 7.916e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2968/30000 [6:41:13<58:10:28,  7.75s/it, lr: 1.0e-04 loss: 8.143e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2968/30000 [6:41:13<58:10:28,  7.75s/it, lr: 1.0e-04 loss: 8.143e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2969/30000 [6:41:13<55:23:35,  7.38s/it, lr: 1.0e-04 loss: 8.143e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2969/30000 [6:41:13<55:23:35,  7.38s/it, lr: 1.0e-04 loss: 8.143e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2969/30000 [6:41:19<55:23:35,  7.38s/it, lr: 1.0e-04 loss: 8.808e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2969/30000 [6:41:19<55:23:35,  7.38s/it, lr: 1.0e-04 loss: 8.808e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2970/30000 [6:41:28<55:23:27,  7.38s/it, lr: 1.0e-04 loss: 4.095e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2970/30000 [6:41:28<55:23:27,  7.38s/it, lr: 1.0e-04 loss: 4.095e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2971/30000 [6:41:28<45:54:11,  6.11s/it, lr: 1.0e-04 loss: 4.095e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2971/30000 [6:41:28<45:54:11,  6.11s/it, lr: 1.0e-04 loss: 4.095e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2971/30000 [6:41:38<45:54:11,  6.11s/it, lr: 1.0e-04 loss: 7.037e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2971/30000 [6:41:38<45:54:11,  6.11s/it, lr: 1.0e-04 loss: 7.037e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2972/30000 [6:41:38<51:58:24,  6.92s/it, lr: 1.0e-04 loss: 7.037e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2972/30000 [6:41:38<51:58:24,  6.92s/it, lr: 1.0e-04 loss: 7.037e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2972/30000 [6:41:47<51:58:24,  6.92s/it, lr: 1.0e-04 loss: 8.010e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2972/30000 [6:41:47<51:58:24,  6.92s/it, lr: 1.0e-04 loss: 8.010e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2973/30000 [6:41:47<56:37:50,  7.54s/it, lr: 1.0e-04 loss: 8.010e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2973/30000 [6:41:47<56:37:50,  7.54s/it, lr: 1.0e-04 loss: 8.010e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2973/30000 [6:41:54<56:37:50,  7.54s/it, lr: 1.0e-04 loss: 1.009e-01]
Xiang_wan_2_2_lora:  10%|###8                                   | 2973/30000 [6:41:54<56:37:50,  7.54s/it, lr: 1.0e-04 loss: 1.009e-01]
Xiang_wan_2_2_lora:  10%|###8                                   | 2974/30000 [6:41:54<54:39:23,  7.28s/it, lr: 1.0e-04 loss: 1.009e-01]
Xiang_wan_2_2_lora:  10%|###8                                   | 2974/30000 [6:41:54<54:39:23,  7.28s/it, lr: 1.0e-04 loss: 1.009e-01]
Xiang_wan_2_2_lora:  10%|###8                                   | 2974/30000 [6:42:01<54:39:23,  7.28s/it, lr: 1.0e-04 loss: 8.349e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2974/30000 [6:42:01<54:39:23,  7.28s/it, lr: 1.0e-04 loss: 8.349e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2975/30000 [6:42:01<53:36:23,  7.14s/it, lr: 1.0e-04 loss: 8.349e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2975/30000 [6:42:01<53:36:23,  7.14s/it, lr: 1.0e-04 loss: 8.349e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2975/30000 [6:42:10<53:36:23,  7.14s/it, lr: 1.0e-04 loss: 4.129e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2975/30000 [6:42:10<53:36:23,  7.14s/it, lr: 1.0e-04 loss: 4.129e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2976/30000 [6:42:10<58:24:32,  7.78s/it, lr: 1.0e-04 loss: 4.129e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2976/30000 [6:42:10<58:24:32,  7.78s/it, lr: 1.0e-04 loss: 4.129e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2976/30000 [6:42:16<58:24:32,  7.78s/it, lr: 1.0e-04 loss: 5.208e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2976/30000 [6:42:16<58:24:32,  7.78s/it, lr: 1.0e-04 loss: 5.208e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2977/30000 [6:42:16<55:49:10,  7.44s/it, lr: 1.0e-04 loss: 5.208e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2977/30000 [6:42:16<55:49:10,  7.44s/it, lr: 1.0e-04 loss: 5.208e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2977/30000 [6:42:26<55:49:10,  7.44s/it, lr: 1.0e-04 loss: 6.621e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2977/30000 [6:42:26<55:49:10,  7.44s/it, lr: 1.0e-04 loss: 6.621e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2978/30000 [6:42:26<60:08:21,  8.01s/it, lr: 1.0e-04 loss: 6.621e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2978/30000 [6:42:26<60:08:21,  8.01s/it, lr: 1.0e-04 loss: 6.621e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2978/30000 [6:42:33<60:08:21,  8.01s/it, lr: 1.0e-04 loss: 7.266e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2978/30000 [6:42:33<60:08:21,  8.01s/it, lr: 1.0e-04 loss: 7.266e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2979/30000 [6:42:33<57:08:52,  7.61s/it, lr: 1.0e-04 loss: 7.266e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2979/30000 [6:42:33<57:08:52,  7.61s/it, lr: 1.0e-04 loss: 7.266e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2979/30000 [6:42:42<57:08:52,  7.61s/it, lr: 1.0e-04 loss: 5.846e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2979/30000 [6:42:42<57:08:52,  7.61s/it, lr: 1.0e-04 loss: 5.846e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2980/30000 [6:42:49<57:08:45,  7.61s/it, lr: 1.0e-04 loss: 2.450e-01]
Xiang_wan_2_2_lora:  10%|###8                                   | 2980/30000 [6:42:49<57:08:45,  7.61s/it, lr: 1.0e-04 loss: 2.450e-01]
Xiang_wan_2_2_lora:  10%|###8                                   | 2981/30000 [6:42:49<42:26:17,  5.65s/it, lr: 1.0e-04 loss: 2.450e-01]
Xiang_wan_2_2_lora:  10%|###8                                   | 2981/30000 [6:42:49<42:26:17,  5.65s/it, lr: 1.0e-04 loss: 2.450e-01]
Xiang_wan_2_2_lora:  10%|###8                                   | 2981/30000 [6:42:58<42:26:17,  5.65s/it, lr: 1.0e-04 loss: 5.900e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2981/30000 [6:42:58<42:26:17,  5.65s/it, lr: 1.0e-04 loss: 5.900e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2982/30000 [6:42:58<49:46:27,  6.63s/it, lr: 1.0e-04 loss: 5.900e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2982/30000 [6:42:58<49:46:27,  6.63s/it, lr: 1.0e-04 loss: 5.900e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2982/30000 [6:43:08<49:46:27,  6.63s/it, lr: 1.0e-04 loss: 5.192e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2982/30000 [6:43:08<49:46:27,  6.63s/it, lr: 1.0e-04 loss: 5.192e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2983/30000 [6:43:08<55:15:01,  7.36s/it, lr: 1.0e-04 loss: 5.192e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2983/30000 [6:43:08<55:15:01,  7.36s/it, lr: 1.0e-04 loss: 5.192e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2983/30000 [6:43:14<55:15:01,  7.36s/it, lr: 1.0e-04 loss: 1.077e-01]
Xiang_wan_2_2_lora:  10%|###8                                   | 2983/30000 [6:43:14<55:15:01,  7.36s/it, lr: 1.0e-04 loss: 1.077e-01]
Xiang_wan_2_2_lora:  10%|###8                                   | 2984/30000 [6:43:14<53:47:24,  7.17s/it, lr: 1.0e-04 loss: 1.077e-01]
Xiang_wan_2_2_lora:  10%|###8                                   | 2984/30000 [6:43:14<53:47:24,  7.17s/it, lr: 1.0e-04 loss: 1.077e-01]
Xiang_wan_2_2_lora:  10%|###8                                   | 2984/30000 [6:43:21<53:47:24,  7.17s/it, lr: 1.0e-04 loss: 7.624e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2984/30000 [6:43:21<53:47:24,  7.17s/it, lr: 1.0e-04 loss: 7.624e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2985/30000 [6:43:21<52:29:48,  7.00s/it, lr: 1.0e-04 loss: 7.624e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2985/30000 [6:43:21<52:29:48,  7.00s/it, lr: 1.0e-04 loss: 7.624e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2985/30000 [6:43:30<52:29:48,  7.00s/it, lr: 1.0e-04 loss: 5.850e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2985/30000 [6:43:30<52:29:48,  7.00s/it, lr: 1.0e-04 loss: 5.850e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2986/30000 [6:43:30<57:30:00,  7.66s/it, lr: 1.0e-04 loss: 5.850e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2986/30000 [6:43:30<57:30:00,  7.66s/it, lr: 1.0e-04 loss: 5.850e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2986/30000 [6:43:37<57:30:00,  7.66s/it, lr: 1.0e-04 loss: 5.881e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2986/30000 [6:43:37<57:30:00,  7.66s/it, lr: 1.0e-04 loss: 5.881e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2987/30000 [6:43:37<55:03:10,  7.34s/it, lr: 1.0e-04 loss: 5.881e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2987/30000 [6:43:37<55:03:10,  7.34s/it, lr: 1.0e-04 loss: 5.881e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2987/30000 [6:43:43<55:03:10,  7.34s/it, lr: 1.0e-04 loss: 1.014e-01]
Xiang_wan_2_2_lora:  10%|###8                                   | 2987/30000 [6:43:43<55:03:10,  7.34s/it, lr: 1.0e-04 loss: 1.014e-01]
Xiang_wan_2_2_lora:  10%|###8                                   | 2988/30000 [6:43:43<53:22:16,  7.11s/it, lr: 1.0e-04 loss: 1.014e-01]
Xiang_wan_2_2_lora:  10%|###8                                   | 2988/30000 [6:43:43<53:22:16,  7.11s/it, lr: 1.0e-04 loss: 1.014e-01]
Xiang_wan_2_2_lora:  10%|###8                                   | 2988/30000 [6:43:50<53:22:16,  7.11s/it, lr: 1.0e-04 loss: 8.985e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2988/30000 [6:43:50<53:22:16,  7.11s/it, lr: 1.0e-04 loss: 8.985e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2989/30000 [6:43:50<52:18:30,  6.97s/it, lr: 1.0e-04 loss: 8.985e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2989/30000 [6:43:50<52:18:30,  6.97s/it, lr: 1.0e-04 loss: 8.985e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2989/30000 [6:43:57<52:18:30,  6.97s/it, lr: 1.0e-04 loss: 6.731e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2989/30000 [6:43:57<52:18:30,  6.97s/it, lr: 1.0e-04 loss: 6.731e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2990/30000 [6:44:06<52:18:23,  6.97s/it, lr: 1.0e-04 loss: 7.852e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2990/30000 [6:44:06<52:18:23,  6.97s/it, lr: 1.0e-04 loss: 7.852e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2991/30000 [6:44:06<44:20:33,  5.91s/it, lr: 1.0e-04 loss: 7.852e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2991/30000 [6:44:06<44:20:33,  5.91s/it, lr: 1.0e-04 loss: 7.852e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2991/30000 [6:44:15<44:20:33,  5.91s/it, lr: 1.0e-04 loss: 1.053e-01]
Xiang_wan_2_2_lora:  10%|###8                                   | 2991/30000 [6:44:15<44:20:33,  5.91s/it, lr: 1.0e-04 loss: 1.053e-01]
Xiang_wan_2_2_lora:  10%|###8                                   | 2992/30000 [6:44:15<50:36:47,  6.75s/it, lr: 1.0e-04 loss: 1.053e-01]
Xiang_wan_2_2_lora:  10%|###8                                   | 2992/30000 [6:44:15<50:36:47,  6.75s/it, lr: 1.0e-04 loss: 1.053e-01]
Xiang_wan_2_2_lora:  10%|###8                                   | 2992/30000 [6:44:25<50:36:47,  6.75s/it, lr: 1.0e-04 loss: 4.317e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2992/30000 [6:44:25<50:36:47,  6.75s/it, lr: 1.0e-04 loss: 4.317e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2993/30000 [6:44:25<55:50:16,  7.44s/it, lr: 1.0e-04 loss: 4.317e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2993/30000 [6:44:25<55:50:16,  7.44s/it, lr: 1.0e-04 loss: 4.317e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2993/30000 [6:44:31<55:50:16,  7.44s/it, lr: 1.0e-04 loss: 7.396e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2993/30000 [6:44:31<55:50:16,  7.44s/it, lr: 1.0e-04 loss: 7.396e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2994/30000 [6:44:31<54:10:40,  7.22s/it, lr: 1.0e-04 loss: 7.396e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2994/30000 [6:44:31<54:10:40,  7.22s/it, lr: 1.0e-04 loss: 7.396e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2994/30000 [6:44:41<54:10:40,  7.22s/it, lr: 1.0e-04 loss: 4.441e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2994/30000 [6:44:41<54:10:40,  7.22s/it, lr: 1.0e-04 loss: 4.441e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2995/30000 [6:44:41<58:32:32,  7.80s/it, lr: 1.0e-04 loss: 4.441e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2995/30000 [6:44:41<58:32:32,  7.80s/it, lr: 1.0e-04 loss: 4.441e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2995/30000 [6:44:50<58:32:32,  7.80s/it, lr: 1.0e-04 loss: 7.817e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2995/30000 [6:44:50<58:32:32,  7.80s/it, lr: 1.0e-04 loss: 7.817e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2996/30000 [6:44:50<61:50:36,  8.24s/it, lr: 1.0e-04 loss: 7.817e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2996/30000 [6:44:50<61:50:36,  8.24s/it, lr: 1.0e-04 loss: 7.817e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2996/30000 [6:45:00<61:50:36,  8.24s/it, lr: 1.0e-04 loss: 3.557e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2996/30000 [6:45:00<61:50:36,  8.24s/it, lr: 1.0e-04 loss: 3.557e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2997/30000 [6:45:00<64:38:43,  8.62s/it, lr: 1.0e-04 loss: 3.557e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2997/30000 [6:45:00<64:38:43,  8.62s/it, lr: 1.0e-04 loss: 3.557e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2997/30000 [6:45:06<64:38:43,  8.62s/it, lr: 1.0e-04 loss: 8.004e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2997/30000 [6:45:06<64:38:43,  8.62s/it, lr: 1.0e-04 loss: 8.004e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2998/30000 [6:45:06<60:06:39,  8.01s/it, lr: 1.0e-04 loss: 8.004e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2998/30000 [6:45:06<60:06:39,  8.01s/it, lr: 1.0e-04 loss: 8.004e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2998/30000 [6:45:13<60:06:39,  8.01s/it, lr: 1.0e-04 loss: 2.644e-01]
Xiang_wan_2_2_lora:  10%|###8                                   | 2998/30000 [6:45:13<60:06:39,  8.01s/it, lr: 1.0e-04 loss: 2.644e-01]
Xiang_wan_2_2_lora:  10%|###8                                   | 2999/30000 [6:45:13<57:00:45,  7.60s/it, lr: 1.0e-04 loss: 2.644e-01]
Xiang_wan_2_2_lora:  10%|###8                                   | 2999/30000 [6:45:13<57:00:45,  7.60s/it, lr: 1.0e-04 loss: 2.644e-01]
Xiang_wan_2_2_lora:  10%|###8                                   | 2999/30000 [6:45:19<57:00:45,  7.60s/it, lr: 1.0e-04 loss: 9.843e-02]
Xiang_wan_2_2_lora:  10%|###8                                   | 2999/30000 [6:45:19<57:00:45,  7.60s/it, lr: 1.0e-04 loss: 9.843e-02]
+Saving at step 3000
+Saved checkpoint to /home/featurize/ai-toolkit/output/Xiang_wan_2_2_lora/Xiang_wan_2_2_lora_000003000.safetensors
+Saved optimizer to /home/featurize/ai-toolkit/output/Xiang_wan_2_2_lora/optimizer.pt
+
Xiang_wan_2_2_lora:  10%|###9                                   | 3000/30000 [6:45:29<57:00:38,  7.60s/it, lr: 1.0e-04 loss: 3.844e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3000/30000 [6:45:29<57:00:38,  7.60s/it, lr: 1.0e-04 loss: 3.844e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3001/30000 [6:45:29<47:34:33,  6.34s/it, lr: 1.0e-04 loss: 3.844e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3001/30000 [6:45:29<47:34:33,  6.34s/it, lr: 1.0e-04 loss: 3.844e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3001/30000 [6:45:36<47:34:33,  6.34s/it, lr: 1.0e-04 loss: 8.033e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3001/30000 [6:45:36<47:34:33,  6.34s/it, lr: 1.0e-04 loss: 8.033e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3002/30000 [6:45:36<48:01:20,  6.40s/it, lr: 1.0e-04 loss: 8.033e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3002/30000 [6:45:36<48:01:20,  6.40s/it, lr: 1.0e-04 loss: 8.033e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3002/30000 [6:45:42<48:01:20,  6.40s/it, lr: 1.0e-04 loss: 7.236e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3002/30000 [6:45:42<48:01:20,  6.40s/it, lr: 1.0e-04 loss: 7.236e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3003/30000 [6:45:42<48:20:04,  6.45s/it, lr: 1.0e-04 loss: 7.236e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3003/30000 [6:45:42<48:20:04,  6.45s/it, lr: 1.0e-04 loss: 7.236e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3003/30000 [6:45:51<48:20:04,  6.45s/it, lr: 1.0e-04 loss: 6.540e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3003/30000 [6:45:51<48:20:04,  6.45s/it, lr: 1.0e-04 loss: 6.540e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3004/30000 [6:45:51<54:05:27,  7.21s/it, lr: 1.0e-04 loss: 6.540e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3004/30000 [6:45:51<54:05:27,  7.21s/it, lr: 1.0e-04 loss: 6.540e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3004/30000 [6:46:01<54:05:27,  7.21s/it, lr: 1.0e-04 loss: 4.240e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3004/30000 [6:46:01<54:05:27,  7.21s/it, lr: 1.0e-04 loss: 4.240e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3005/30000 [6:46:01<58:29:35,  7.80s/it, lr: 1.0e-04 loss: 4.240e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3005/30000 [6:46:01<58:29:35,  7.80s/it, lr: 1.0e-04 loss: 4.240e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3005/30000 [6:46:10<58:29:35,  7.80s/it, lr: 1.0e-04 loss: 7.049e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3005/30000 [6:46:10<58:29:35,  7.80s/it, lr: 1.0e-04 loss: 7.049e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3006/30000 [6:46:10<61:43:26,  8.23s/it, lr: 1.0e-04 loss: 7.049e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3006/30000 [6:46:10<61:43:26,  8.23s/it, lr: 1.0e-04 loss: 7.049e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3006/30000 [6:46:19<61:43:26,  8.23s/it, lr: 1.0e-04 loss: 9.697e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3006/30000 [6:46:19<61:43:26,  8.23s/it, lr: 1.0e-04 loss: 9.697e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3007/30000 [6:46:19<64:07:43,  8.55s/it, lr: 1.0e-04 loss: 9.697e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3007/30000 [6:46:19<64:07:43,  8.55s/it, lr: 1.0e-04 loss: 9.697e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3007/30000 [6:46:29<64:07:43,  8.55s/it, lr: 1.0e-04 loss: 5.980e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3007/30000 [6:46:29<64:07:43,  8.55s/it, lr: 1.0e-04 loss: 5.980e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3008/30000 [6:46:29<66:14:08,  8.83s/it, lr: 1.0e-04 loss: 5.980e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3008/30000 [6:46:29<66:14:08,  8.83s/it, lr: 1.0e-04 loss: 5.980e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3008/30000 [6:46:38<66:14:08,  8.83s/it, lr: 1.0e-04 loss: 6.015e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3008/30000 [6:46:38<66:14:08,  8.83s/it, lr: 1.0e-04 loss: 6.015e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3009/30000 [6:46:38<67:24:36,  8.99s/it, lr: 1.0e-04 loss: 6.015e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3009/30000 [6:46:38<67:24:36,  8.99s/it, lr: 1.0e-04 loss: 6.015e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3009/30000 [6:46:45<67:24:36,  8.99s/it, lr: 1.0e-04 loss: 8.272e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3009/30000 [6:46:45<67:24:36,  8.99s/it, lr: 1.0e-04 loss: 8.272e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3010/30000 [6:46:51<67:24:27,  8.99s/it, lr: 1.0e-04 loss: 7.194e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3010/30000 [6:46:51<67:24:27,  8.99s/it, lr: 1.0e-04 loss: 7.194e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3011/30000 [6:46:51<47:51:13,  6.38s/it, lr: 1.0e-04 loss: 7.194e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3011/30000 [6:46:51<47:51:13,  6.38s/it, lr: 1.0e-04 loss: 7.194e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3011/30000 [6:47:01<47:51:13,  6.38s/it, lr: 1.0e-04 loss: 7.198e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3011/30000 [6:47:01<47:51:13,  6.38s/it, lr: 1.0e-04 loss: 7.198e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3012/30000 [6:47:01<53:21:24,  7.12s/it, lr: 1.0e-04 loss: 7.198e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3012/30000 [6:47:01<53:21:24,  7.12s/it, lr: 1.0e-04 loss: 7.198e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3012/30000 [6:47:07<53:21:24,  7.12s/it, lr: 1.0e-04 loss: 6.843e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3012/30000 [6:47:07<53:21:24,  7.12s/it, lr: 1.0e-04 loss: 6.843e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3013/30000 [6:47:07<52:17:56,  6.98s/it, lr: 1.0e-04 loss: 6.843e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3013/30000 [6:47:07<52:17:56,  6.98s/it, lr: 1.0e-04 loss: 6.843e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3013/30000 [6:47:14<52:17:56,  6.98s/it, lr: 1.0e-04 loss: 7.468e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3013/30000 [6:47:14<52:17:56,  6.98s/it, lr: 1.0e-04 loss: 7.468e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3014/30000 [6:47:14<51:21:47,  6.85s/it, lr: 1.0e-04 loss: 7.468e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3014/30000 [6:47:14<51:21:47,  6.85s/it, lr: 1.0e-04 loss: 7.468e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3014/30000 [6:47:21<51:21:47,  6.85s/it, lr: 1.0e-04 loss: 6.638e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3014/30000 [6:47:21<51:21:47,  6.85s/it, lr: 1.0e-04 loss: 6.638e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3015/30000 [6:47:21<51:19:11,  6.85s/it, lr: 1.0e-04 loss: 6.638e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3015/30000 [6:47:21<51:19:11,  6.85s/it, lr: 1.0e-04 loss: 6.638e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3015/30000 [6:47:30<51:19:11,  6.85s/it, lr: 1.0e-04 loss: 6.296e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3015/30000 [6:47:30<51:19:11,  6.85s/it, lr: 1.0e-04 loss: 6.296e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3016/30000 [6:47:30<56:31:49,  7.54s/it, lr: 1.0e-04 loss: 6.296e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3016/30000 [6:47:30<56:31:49,  7.54s/it, lr: 1.0e-04 loss: 6.296e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3016/30000 [6:47:37<56:31:49,  7.54s/it, lr: 1.0e-04 loss: 4.730e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3016/30000 [6:47:37<56:31:49,  7.54s/it, lr: 1.0e-04 loss: 4.730e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3017/30000 [6:47:37<54:27:05,  7.26s/it, lr: 1.0e-04 loss: 4.730e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3017/30000 [6:47:37<54:27:05,  7.26s/it, lr: 1.0e-04 loss: 4.730e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3017/30000 [6:47:43<54:27:05,  7.26s/it, lr: 1.0e-04 loss: 6.358e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3017/30000 [6:47:43<54:27:05,  7.26s/it, lr: 1.0e-04 loss: 6.358e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3018/30000 [6:47:43<52:53:53,  7.06s/it, lr: 1.0e-04 loss: 6.358e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3018/30000 [6:47:43<52:53:53,  7.06s/it, lr: 1.0e-04 loss: 6.358e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3018/30000 [6:47:50<52:53:53,  7.06s/it, lr: 1.0e-04 loss: 1.750e-01]
Xiang_wan_2_2_lora:  10%|###9                                   | 3018/30000 [6:47:50<52:53:53,  7.06s/it, lr: 1.0e-04 loss: 1.750e-01]
Xiang_wan_2_2_lora:  10%|###9                                   | 3019/30000 [6:47:50<51:47:20,  6.91s/it, lr: 1.0e-04 loss: 1.750e-01]
Xiang_wan_2_2_lora:  10%|###9                                   | 3019/30000 [6:47:50<51:47:20,  6.91s/it, lr: 1.0e-04 loss: 1.750e-01]
Xiang_wan_2_2_lora:  10%|###9                                   | 3019/30000 [6:47:56<51:47:20,  6.91s/it, lr: 1.0e-04 loss: 8.298e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3019/30000 [6:47:56<51:47:20,  6.91s/it, lr: 1.0e-04 loss: 8.298e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3020/30000 [6:48:06<51:47:13,  6.91s/it, lr: 1.0e-04 loss: 2.216e-01]
Xiang_wan_2_2_lora:  10%|###9                                   | 3020/30000 [6:48:06<51:47:13,  6.91s/it, lr: 1.0e-04 loss: 2.216e-01]
Xiang_wan_2_2_lora:  10%|###9                                   | 3021/30000 [6:48:06<44:05:04,  5.88s/it, lr: 1.0e-04 loss: 2.216e-01]
Xiang_wan_2_2_lora:  10%|###9                                   | 3021/30000 [6:48:06<44:05:04,  5.88s/it, lr: 1.0e-04 loss: 2.216e-01]
Xiang_wan_2_2_lora:  10%|###9                                   | 3021/30000 [6:48:15<44:05:04,  5.88s/it, lr: 1.0e-04 loss: 6.771e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3021/30000 [6:48:15<44:05:04,  5.88s/it, lr: 1.0e-04 loss: 6.771e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3022/30000 [6:48:15<50:27:14,  6.73s/it, lr: 1.0e-04 loss: 6.771e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3022/30000 [6:48:15<50:27:14,  6.73s/it, lr: 1.0e-04 loss: 6.771e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3022/30000 [6:48:22<50:27:14,  6.73s/it, lr: 1.0e-04 loss: 7.654e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3022/30000 [6:48:22<50:27:14,  6.73s/it, lr: 1.0e-04 loss: 7.654e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3023/30000 [6:48:22<50:48:13,  6.78s/it, lr: 1.0e-04 loss: 7.654e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3023/30000 [6:48:22<50:48:13,  6.78s/it, lr: 1.0e-04 loss: 7.654e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3023/30000 [6:48:28<50:48:13,  6.78s/it, lr: 1.0e-04 loss: 8.262e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3023/30000 [6:48:28<50:48:13,  6.78s/it, lr: 1.0e-04 loss: 8.262e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3024/30000 [6:48:28<50:15:35,  6.71s/it, lr: 1.0e-04 loss: 8.262e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3024/30000 [6:48:28<50:15:35,  6.71s/it, lr: 1.0e-04 loss: 8.262e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3024/30000 [6:48:35<50:15:35,  6.71s/it, lr: 1.0e-04 loss: 3.825e-01]
Xiang_wan_2_2_lora:  10%|###9                                   | 3024/30000 [6:48:35<50:15:35,  6.71s/it, lr: 1.0e-04 loss: 3.825e-01]
Xiang_wan_2_2_lora:  10%|###9                                   | 3025/30000 [6:48:35<50:01:36,  6.68s/it, lr: 1.0e-04 loss: 3.825e-01]
Xiang_wan_2_2_lora:  10%|###9                                   | 3025/30000 [6:48:35<50:01:36,  6.68s/it, lr: 1.0e-04 loss: 3.825e-01]
Xiang_wan_2_2_lora:  10%|###9                                   | 3025/30000 [6:48:44<50:01:36,  6.68s/it, lr: 1.0e-04 loss: 5.670e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3025/30000 [6:48:44<50:01:36,  6.68s/it, lr: 1.0e-04 loss: 5.670e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3026/30000 [6:48:44<55:41:21,  7.43s/it, lr: 1.0e-04 loss: 5.670e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3026/30000 [6:48:44<55:41:21,  7.43s/it, lr: 1.0e-04 loss: 5.670e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3026/30000 [6:48:54<55:41:21,  7.43s/it, lr: 1.0e-04 loss: 3.402e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3026/30000 [6:48:54<55:41:21,  7.43s/it, lr: 1.0e-04 loss: 3.402e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3027/30000 [6:48:54<59:53:27,  7.99s/it, lr: 1.0e-04 loss: 3.402e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3027/30000 [6:48:54<59:53:27,  7.99s/it, lr: 1.0e-04 loss: 3.402e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3027/30000 [6:49:03<59:53:27,  7.99s/it, lr: 1.0e-04 loss: 5.336e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3027/30000 [6:49:03<59:53:27,  7.99s/it, lr: 1.0e-04 loss: 5.336e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3028/30000 [6:49:03<62:47:43,  8.38s/it, lr: 1.0e-04 loss: 5.336e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3028/30000 [6:49:03<62:47:43,  8.38s/it, lr: 1.0e-04 loss: 5.336e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3028/30000 [6:49:09<62:47:43,  8.38s/it, lr: 1.0e-04 loss: 6.426e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3028/30000 [6:49:09<62:47:43,  8.38s/it, lr: 1.0e-04 loss: 6.426e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3029/30000 [6:49:09<58:47:45,  7.85s/it, lr: 1.0e-04 loss: 6.426e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3029/30000 [6:49:09<58:47:45,  7.85s/it, lr: 1.0e-04 loss: 6.426e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3029/30000 [6:49:19<58:47:45,  7.85s/it, lr: 1.0e-04 loss: 7.019e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3029/30000 [6:49:19<58:47:45,  7.85s/it, lr: 1.0e-04 loss: 7.019e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3030/30000 [6:49:28<58:47:37,  7.85s/it, lr: 1.0e-04 loss: 3.460e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3030/30000 [6:49:28<58:47:37,  7.85s/it, lr: 1.0e-04 loss: 3.460e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3031/30000 [6:49:28<47:51:40,  6.39s/it, lr: 1.0e-04 loss: 3.460e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3031/30000 [6:49:28<47:51:40,  6.39s/it, lr: 1.0e-04 loss: 3.460e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3031/30000 [6:49:38<47:51:40,  6.39s/it, lr: 1.0e-04 loss: 7.653e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3031/30000 [6:49:38<47:51:40,  6.39s/it, lr: 1.0e-04 loss: 7.653e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3032/30000 [6:49:38<53:22:19,  7.12s/it, lr: 1.0e-04 loss: 7.653e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3032/30000 [6:49:38<53:22:19,  7.12s/it, lr: 1.0e-04 loss: 7.653e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3032/30000 [6:49:47<53:22:19,  7.12s/it, lr: 1.0e-04 loss: 5.246e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3032/30000 [6:49:47<53:22:19,  7.12s/it, lr: 1.0e-04 loss: 5.246e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3033/30000 [6:49:47<57:42:50,  7.70s/it, lr: 1.0e-04 loss: 5.246e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3033/30000 [6:49:47<57:42:50,  7.70s/it, lr: 1.0e-04 loss: 5.246e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3033/30000 [6:49:54<57:42:50,  7.70s/it, lr: 1.0e-04 loss: 7.818e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3033/30000 [6:49:54<57:42:50,  7.70s/it, lr: 1.0e-04 loss: 7.818e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3034/30000 [6:49:54<55:22:45,  7.39s/it, lr: 1.0e-04 loss: 7.818e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3034/30000 [6:49:54<55:22:45,  7.39s/it, lr: 1.0e-04 loss: 7.818e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3034/30000 [6:50:00<55:22:45,  7.39s/it, lr: 1.0e-04 loss: 6.020e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3034/30000 [6:50:00<55:22:45,  7.39s/it, lr: 1.0e-04 loss: 6.020e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3035/30000 [6:50:00<53:45:06,  7.18s/it, lr: 1.0e-04 loss: 6.020e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3035/30000 [6:50:00<53:45:06,  7.18s/it, lr: 1.0e-04 loss: 6.020e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3035/30000 [6:50:07<53:45:06,  7.18s/it, lr: 1.0e-04 loss: 9.587e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3035/30000 [6:50:07<53:45:06,  7.18s/it, lr: 1.0e-04 loss: 9.587e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3036/30000 [6:50:07<52:28:18,  7.01s/it, lr: 1.0e-04 loss: 9.587e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3036/30000 [6:50:07<52:28:18,  7.01s/it, lr: 1.0e-04 loss: 9.587e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3036/30000 [6:50:13<52:28:18,  7.01s/it, lr: 1.0e-04 loss: 7.517e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3036/30000 [6:50:13<52:28:18,  7.01s/it, lr: 1.0e-04 loss: 7.517e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3037/30000 [6:50:13<51:33:44,  6.88s/it, lr: 1.0e-04 loss: 7.517e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3037/30000 [6:50:13<51:33:44,  6.88s/it, lr: 1.0e-04 loss: 7.517e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3037/30000 [6:50:20<51:33:44,  6.88s/it, lr: 1.0e-04 loss: 9.146e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3037/30000 [6:50:20<51:33:44,  6.88s/it, lr: 1.0e-04 loss: 9.146e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3038/30000 [6:50:20<51:26:28,  6.87s/it, lr: 1.0e-04 loss: 9.146e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3038/30000 [6:50:20<51:26:28,  6.87s/it, lr: 1.0e-04 loss: 9.146e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3038/30000 [6:50:29<51:26:28,  6.87s/it, lr: 1.0e-04 loss: 4.338e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3038/30000 [6:50:29<51:26:28,  6.87s/it, lr: 1.0e-04 loss: 4.338e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3039/30000 [6:50:29<56:45:44,  7.58s/it, lr: 1.0e-04 loss: 4.338e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3039/30000 [6:50:29<56:45:44,  7.58s/it, lr: 1.0e-04 loss: 4.338e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3039/30000 [6:50:39<56:45:44,  7.58s/it, lr: 1.0e-04 loss: 7.452e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3039/30000 [6:50:39<56:45:44,  7.58s/it, lr: 1.0e-04 loss: 7.452e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3040/30000 [6:50:48<56:45:36,  7.58s/it, lr: 1.0e-04 loss: 4.422e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3040/30000 [6:50:48<56:45:36,  7.58s/it, lr: 1.0e-04 loss: 4.422e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3041/30000 [6:50:48<46:45:08,  6.24s/it, lr: 1.0e-04 loss: 4.422e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3041/30000 [6:50:48<46:45:08,  6.24s/it, lr: 1.0e-04 loss: 4.422e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3041/30000 [6:50:55<46:45:08,  6.24s/it, lr: 1.0e-04 loss: 1.000e-01]
Xiang_wan_2_2_lora:  10%|###9                                   | 3041/30000 [6:50:55<46:45:08,  6.24s/it, lr: 1.0e-04 loss: 1.000e-01]
Xiang_wan_2_2_lora:  10%|###9                                   | 3042/30000 [6:50:55<47:28:50,  6.34s/it, lr: 1.0e-04 loss: 1.000e-01]
Xiang_wan_2_2_lora:  10%|###9                                   | 3042/30000 [6:50:55<47:28:50,  6.34s/it, lr: 1.0e-04 loss: 1.000e-01]
Xiang_wan_2_2_lora:  10%|###9                                   | 3042/30000 [6:51:01<47:28:50,  6.34s/it, lr: 1.0e-04 loss: 7.599e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3042/30000 [6:51:01<47:28:50,  6.34s/it, lr: 1.0e-04 loss: 7.599e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3043/30000 [6:51:01<47:54:52,  6.40s/it, lr: 1.0e-04 loss: 7.599e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3043/30000 [6:51:01<47:54:52,  6.40s/it, lr: 1.0e-04 loss: 7.599e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3043/30000 [6:51:08<47:54:52,  6.40s/it, lr: 1.0e-04 loss: 7.796e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3043/30000 [6:51:08<47:54:52,  6.40s/it, lr: 1.0e-04 loss: 7.796e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3044/30000 [6:51:08<48:18:09,  6.45s/it, lr: 1.0e-04 loss: 7.796e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3044/30000 [6:51:08<48:18:09,  6.45s/it, lr: 1.0e-04 loss: 7.796e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3044/30000 [6:51:17<48:18:09,  6.45s/it, lr: 1.0e-04 loss: 6.371e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3044/30000 [6:51:17<48:18:09,  6.45s/it, lr: 1.0e-04 loss: 6.371e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3045/30000 [6:51:17<54:43:51,  7.31s/it, lr: 1.0e-04 loss: 6.371e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3045/30000 [6:51:17<54:43:51,  7.31s/it, lr: 1.0e-04 loss: 6.371e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3045/30000 [6:51:24<54:43:51,  7.31s/it, lr: 1.0e-04 loss: 8.493e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3045/30000 [6:51:24<54:43:51,  7.31s/it, lr: 1.0e-04 loss: 8.493e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3046/30000 [6:51:24<53:19:25,  7.12s/it, lr: 1.0e-04 loss: 8.493e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3046/30000 [6:51:24<53:19:25,  7.12s/it, lr: 1.0e-04 loss: 8.493e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3046/30000 [6:51:33<53:19:25,  7.12s/it, lr: 1.0e-04 loss: 4.765e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3046/30000 [6:51:33<53:19:25,  7.12s/it, lr: 1.0e-04 loss: 4.765e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3047/30000 [6:51:33<58:10:38,  7.77s/it, lr: 1.0e-04 loss: 4.765e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3047/30000 [6:51:33<58:10:38,  7.77s/it, lr: 1.0e-04 loss: 4.765e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3047/30000 [6:51:40<58:10:38,  7.77s/it, lr: 1.0e-04 loss: 8.593e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3047/30000 [6:51:40<58:10:38,  7.77s/it, lr: 1.0e-04 loss: 8.593e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3048/30000 [6:51:40<55:31:29,  7.42s/it, lr: 1.0e-04 loss: 8.593e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3048/30000 [6:51:40<55:31:29,  7.42s/it, lr: 1.0e-04 loss: 8.593e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3048/30000 [6:51:49<55:31:29,  7.42s/it, lr: 1.0e-04 loss: 1.069e-01]
Xiang_wan_2_2_lora:  10%|###9                                   | 3048/30000 [6:51:49<55:31:29,  7.42s/it, lr: 1.0e-04 loss: 1.069e-01]
Xiang_wan_2_2_lora:  10%|###9                                   | 3049/30000 [6:51:49<59:36:29,  7.96s/it, lr: 1.0e-04 loss: 1.069e-01]
Xiang_wan_2_2_lora:  10%|###9                                   | 3049/30000 [6:51:49<59:36:29,  7.96s/it, lr: 1.0e-04 loss: 1.069e-01]
Xiang_wan_2_2_lora:  10%|###9                                   | 3049/30000 [6:51:59<59:36:29,  7.96s/it, lr: 1.0e-04 loss: 5.655e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3049/30000 [6:51:59<59:36:29,  7.96s/it, lr: 1.0e-04 loss: 5.655e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3050/30000 [6:52:05<59:36:21,  7.96s/it, lr: 1.0e-04 loss: 7.125e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3050/30000 [6:52:05<59:36:21,  7.96s/it, lr: 1.0e-04 loss: 7.125e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3051/30000 [6:52:05<43:33:03,  5.82s/it, lr: 1.0e-04 loss: 7.125e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3051/30000 [6:52:05<43:33:03,  5.82s/it, lr: 1.0e-04 loss: 7.125e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3051/30000 [6:52:15<43:33:03,  5.82s/it, lr: 1.0e-04 loss: 6.847e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3051/30000 [6:52:15<43:33:03,  5.82s/it, lr: 1.0e-04 loss: 6.847e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3052/30000 [6:52:15<50:05:16,  6.69s/it, lr: 1.0e-04 loss: 6.847e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3052/30000 [6:52:15<50:05:16,  6.69s/it, lr: 1.0e-04 loss: 6.847e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3052/30000 [6:52:24<50:05:16,  6.69s/it, lr: 1.0e-04 loss: 2.612e-01]
Xiang_wan_2_2_lora:  10%|###9                                   | 3052/30000 [6:52:24<50:05:16,  6.69s/it, lr: 1.0e-04 loss: 2.612e-01]
Xiang_wan_2_2_lora:  10%|###9                                   | 3053/30000 [6:52:24<55:43:18,  7.44s/it, lr: 1.0e-04 loss: 2.612e-01]
Xiang_wan_2_2_lora:  10%|###9                                   | 3053/30000 [6:52:24<55:43:18,  7.44s/it, lr: 1.0e-04 loss: 2.612e-01]
Xiang_wan_2_2_lora:  10%|###9                                   | 3053/30000 [6:52:33<55:43:18,  7.44s/it, lr: 1.0e-04 loss: 9.212e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3053/30000 [6:52:33<55:43:18,  7.44s/it, lr: 1.0e-04 loss: 9.212e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3054/30000 [6:52:33<59:32:50,  7.96s/it, lr: 1.0e-04 loss: 9.212e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3054/30000 [6:52:33<59:32:50,  7.96s/it, lr: 1.0e-04 loss: 9.212e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3054/30000 [6:52:43<59:32:50,  7.96s/it, lr: 1.0e-04 loss: 4.489e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3054/30000 [6:52:43<59:32:50,  7.96s/it, lr: 1.0e-04 loss: 4.489e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3055/30000 [6:52:43<62:22:21,  8.33s/it, lr: 1.0e-04 loss: 4.489e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3055/30000 [6:52:43<62:22:21,  8.33s/it, lr: 1.0e-04 loss: 4.489e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3055/30000 [6:52:52<62:22:21,  8.33s/it, lr: 1.0e-04 loss: 7.679e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3055/30000 [6:52:52<62:22:21,  8.33s/it, lr: 1.0e-04 loss: 7.679e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3056/30000 [6:52:52<64:34:00,  8.63s/it, lr: 1.0e-04 loss: 7.679e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3056/30000 [6:52:52<64:34:00,  8.63s/it, lr: 1.0e-04 loss: 7.679e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3056/30000 [6:52:59<64:34:00,  8.63s/it, lr: 1.0e-04 loss: 6.065e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3056/30000 [6:52:59<64:34:00,  8.63s/it, lr: 1.0e-04 loss: 6.065e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3057/30000 [6:52:59<60:13:41,  8.05s/it, lr: 1.0e-04 loss: 6.065e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3057/30000 [6:52:59<60:13:41,  8.05s/it, lr: 1.0e-04 loss: 6.065e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3057/30000 [6:53:05<60:13:41,  8.05s/it, lr: 1.0e-04 loss: 8.771e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3057/30000 [6:53:05<60:13:41,  8.05s/it, lr: 1.0e-04 loss: 8.771e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3058/30000 [6:53:05<56:59:58,  7.62s/it, lr: 1.0e-04 loss: 8.771e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3058/30000 [6:53:05<56:59:58,  7.62s/it, lr: 1.0e-04 loss: 8.771e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3058/30000 [6:53:15<56:59:58,  7.62s/it, lr: 1.0e-04 loss: 4.893e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3058/30000 [6:53:15<56:59:58,  7.62s/it, lr: 1.0e-04 loss: 4.893e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3059/30000 [6:53:15<60:39:13,  8.10s/it, lr: 1.0e-04 loss: 4.893e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3059/30000 [6:53:15<60:39:13,  8.10s/it, lr: 1.0e-04 loss: 4.893e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3059/30000 [6:53:21<60:39:13,  8.10s/it, lr: 1.0e-04 loss: 8.016e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3059/30000 [6:53:21<60:39:13,  8.10s/it, lr: 1.0e-04 loss: 8.016e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3060/30000 [6:53:28<60:39:05,  8.10s/it, lr: 1.0e-04 loss: 6.195e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3060/30000 [6:53:28<60:39:05,  8.10s/it, lr: 1.0e-04 loss: 6.195e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3061/30000 [6:53:28<44:04:56,  5.89s/it, lr: 1.0e-04 loss: 6.195e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3061/30000 [6:53:28<44:04:56,  5.89s/it, lr: 1.0e-04 loss: 6.195e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3061/30000 [6:53:37<44:04:56,  5.89s/it, lr: 1.0e-04 loss: 7.575e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3061/30000 [6:53:37<44:04:56,  5.89s/it, lr: 1.0e-04 loss: 7.575e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3062/30000 [6:53:37<50:30:59,  6.75s/it, lr: 1.0e-04 loss: 7.575e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3062/30000 [6:53:37<50:30:59,  6.75s/it, lr: 1.0e-04 loss: 7.575e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3062/30000 [6:53:47<50:30:59,  6.75s/it, lr: 1.0e-04 loss: 5.268e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3062/30000 [6:53:47<50:30:59,  6.75s/it, lr: 1.0e-04 loss: 5.268e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3063/30000 [6:53:47<55:30:33,  7.42s/it, lr: 1.0e-04 loss: 5.268e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3063/30000 [6:53:47<55:30:33,  7.42s/it, lr: 1.0e-04 loss: 5.268e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3063/30000 [6:53:56<55:30:33,  7.42s/it, lr: 1.0e-04 loss: 9.085e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3063/30000 [6:53:56<55:30:33,  7.42s/it, lr: 1.0e-04 loss: 9.085e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3064/30000 [6:53:56<59:28:30,  7.95s/it, lr: 1.0e-04 loss: 9.085e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3064/30000 [6:53:56<59:28:30,  7.95s/it, lr: 1.0e-04 loss: 9.085e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3064/30000 [6:54:03<59:28:30,  7.95s/it, lr: 1.0e-04 loss: 6.766e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3064/30000 [6:54:03<59:28:30,  7.95s/it, lr: 1.0e-04 loss: 6.766e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3065/30000 [6:54:03<56:40:15,  7.57s/it, lr: 1.0e-04 loss: 6.766e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3065/30000 [6:54:03<56:40:15,  7.57s/it, lr: 1.0e-04 loss: 6.766e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3065/30000 [6:54:09<56:40:15,  7.57s/it, lr: 1.0e-04 loss: 7.736e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3065/30000 [6:54:09<56:40:15,  7.57s/it, lr: 1.0e-04 loss: 7.736e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3066/30000 [6:54:09<54:25:40,  7.27s/it, lr: 1.0e-04 loss: 7.736e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3066/30000 [6:54:09<54:25:40,  7.27s/it, lr: 1.0e-04 loss: 7.736e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3066/30000 [6:54:16<54:25:40,  7.27s/it, lr: 1.0e-04 loss: 2.904e-01]
Xiang_wan_2_2_lora:  10%|###9                                   | 3066/30000 [6:54:16<54:25:40,  7.27s/it, lr: 1.0e-04 loss: 2.904e-01]
Xiang_wan_2_2_lora:  10%|###9                                   | 3067/30000 [6:54:16<52:52:46,  7.07s/it, lr: 1.0e-04 loss: 2.904e-01]
Xiang_wan_2_2_lora:  10%|###9                                   | 3067/30000 [6:54:16<52:52:46,  7.07s/it, lr: 1.0e-04 loss: 2.904e-01]
Xiang_wan_2_2_lora:  10%|###9                                   | 3067/30000 [6:54:23<52:52:46,  7.07s/it, lr: 1.0e-04 loss: 8.170e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3067/30000 [6:54:23<52:52:46,  7.07s/it, lr: 1.0e-04 loss: 8.170e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3068/30000 [6:54:23<52:31:00,  7.02s/it, lr: 1.0e-04 loss: 8.170e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3068/30000 [6:54:23<52:31:00,  7.02s/it, lr: 1.0e-04 loss: 8.170e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3068/30000 [6:54:32<52:31:00,  7.02s/it, lr: 1.0e-04 loss: 5.406e-01]
Xiang_wan_2_2_lora:  10%|###9                                   | 3068/30000 [6:54:32<52:31:00,  7.02s/it, lr: 1.0e-04 loss: 5.406e-01]
Xiang_wan_2_2_lora:  10%|###9                                   | 3069/30000 [6:54:32<57:36:12,  7.70s/it, lr: 1.0e-04 loss: 5.406e-01]
Xiang_wan_2_2_lora:  10%|###9                                   | 3069/30000 [6:54:32<57:36:12,  7.70s/it, lr: 1.0e-04 loss: 5.406e-01]
Xiang_wan_2_2_lora:  10%|###9                                   | 3069/30000 [6:54:39<57:36:12,  7.70s/it, lr: 1.0e-04 loss: 9.270e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3069/30000 [6:54:39<57:36:12,  7.70s/it, lr: 1.0e-04 loss: 9.270e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3070/30000 [6:54:48<57:36:04,  7.70s/it, lr: 1.0e-04 loss: 2.478e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3070/30000 [6:54:48<57:36:04,  7.70s/it, lr: 1.0e-04 loss: 2.478e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3071/30000 [6:54:48<47:14:36,  6.32s/it, lr: 1.0e-04 loss: 2.478e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3071/30000 [6:54:48<47:14:36,  6.32s/it, lr: 1.0e-04 loss: 2.478e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3071/30000 [6:54:55<47:14:36,  6.32s/it, lr: 1.0e-04 loss: 9.766e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3071/30000 [6:54:55<47:14:36,  6.32s/it, lr: 1.0e-04 loss: 9.766e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3072/30000 [6:54:55<47:46:38,  6.39s/it, lr: 1.0e-04 loss: 9.766e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3072/30000 [6:54:55<47:46:38,  6.39s/it, lr: 1.0e-04 loss: 9.766e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3072/30000 [6:55:04<47:46:38,  6.39s/it, lr: 1.0e-04 loss: 1.485e-01]
Xiang_wan_2_2_lora:  10%|###9                                   | 3072/30000 [6:55:04<47:46:38,  6.39s/it, lr: 1.0e-04 loss: 1.485e-01]
Xiang_wan_2_2_lora:  10%|###9                                   | 3073/30000 [6:55:04<53:38:56,  7.17s/it, lr: 1.0e-04 loss: 1.485e-01]
Xiang_wan_2_2_lora:  10%|###9                                   | 3073/30000 [6:55:04<53:38:56,  7.17s/it, lr: 1.0e-04 loss: 1.485e-01]
Xiang_wan_2_2_lora:  10%|###9                                   | 3073/30000 [6:55:11<53:38:56,  7.17s/it, lr: 1.0e-04 loss: 6.969e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3073/30000 [6:55:11<53:38:56,  7.17s/it, lr: 1.0e-04 loss: 6.969e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3074/30000 [6:55:11<52:26:02,  7.01s/it, lr: 1.0e-04 loss: 6.969e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3074/30000 [6:55:11<52:26:02,  7.01s/it, lr: 1.0e-04 loss: 6.969e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3074/30000 [6:55:17<52:26:02,  7.01s/it, lr: 1.0e-04 loss: 1.689e-01]
Xiang_wan_2_2_lora:  10%|###9                                   | 3074/30000 [6:55:17<52:26:02,  7.01s/it, lr: 1.0e-04 loss: 1.689e-01]
Xiang_wan_2_2_lora:  10%|###9                                   | 3075/30000 [6:55:17<52:11:46,  6.98s/it, lr: 1.0e-04 loss: 1.689e-01]
Xiang_wan_2_2_lora:  10%|###9                                   | 3075/30000 [6:55:17<52:11:46,  6.98s/it, lr: 1.0e-04 loss: 1.689e-01]
Xiang_wan_2_2_lora:  10%|###9                                   | 3075/30000 [6:55:27<52:11:46,  6.98s/it, lr: 1.0e-04 loss: 7.006e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3075/30000 [6:55:27<52:11:46,  6.98s/it, lr: 1.0e-04 loss: 7.006e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3076/30000 [6:55:27<57:19:56,  7.67s/it, lr: 1.0e-04 loss: 7.006e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3076/30000 [6:55:27<57:19:56,  7.67s/it, lr: 1.0e-04 loss: 7.006e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3076/30000 [6:55:36<57:19:56,  7.67s/it, lr: 1.0e-04 loss: 6.548e-02]
Xiang_wan_2_2_lora:  10%|###9                                   | 3076/30000 [6:55:36<57:19:56,  7.67s/it, lr: 1.0e-04 loss: 6.548e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3077/30000 [6:55:36<60:59:22,  8.16s/it, lr: 1.0e-04 loss: 6.548e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3077/30000 [6:55:36<60:59:22,  8.16s/it, lr: 1.0e-04 loss: 6.548e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3077/30000 [6:55:43<60:59:22,  8.16s/it, lr: 1.0e-04 loss: 7.079e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3077/30000 [6:55:43<60:59:22,  8.16s/it, lr: 1.0e-04 loss: 7.079e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3078/30000 [6:55:43<57:46:39,  7.73s/it, lr: 1.0e-04 loss: 7.079e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3078/30000 [6:55:43<57:46:39,  7.73s/it, lr: 1.0e-04 loss: 7.079e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3078/30000 [6:55:52<57:46:39,  7.73s/it, lr: 1.0e-04 loss: 3.602e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3078/30000 [6:55:52<57:46:39,  7.73s/it, lr: 1.0e-04 loss: 3.602e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3079/30000 [6:55:52<61:14:56,  8.19s/it, lr: 1.0e-04 loss: 3.602e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3079/30000 [6:55:52<61:14:56,  8.19s/it, lr: 1.0e-04 loss: 3.602e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3079/30000 [6:56:02<61:14:56,  8.19s/it, lr: 1.0e-04 loss: 5.458e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3079/30000 [6:56:02<61:14:56,  8.19s/it, lr: 1.0e-04 loss: 5.458e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3080/30000 [6:56:08<61:14:47,  8.19s/it, lr: 1.0e-04 loss: 7.115e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3080/30000 [6:56:08<61:14:47,  8.19s/it, lr: 1.0e-04 loss: 7.115e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3081/30000 [6:56:08<44:31:51,  5.96s/it, lr: 1.0e-04 loss: 7.115e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3081/30000 [6:56:08<44:31:51,  5.96s/it, lr: 1.0e-04 loss: 7.115e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3081/30000 [6:56:18<44:31:51,  5.96s/it, lr: 1.0e-04 loss: 5.626e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3081/30000 [6:56:18<44:31:51,  5.96s/it, lr: 1.0e-04 loss: 5.626e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3082/30000 [6:56:18<51:01:47,  6.82s/it, lr: 1.0e-04 loss: 5.626e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3082/30000 [6:56:18<51:01:47,  6.82s/it, lr: 1.0e-04 loss: 5.626e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3082/30000 [6:56:24<51:01:47,  6.82s/it, lr: 1.0e-04 loss: 6.928e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3082/30000 [6:56:24<51:01:47,  6.82s/it, lr: 1.0e-04 loss: 6.928e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3083/30000 [6:56:24<50:35:05,  6.77s/it, lr: 1.0e-04 loss: 6.928e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3083/30000 [6:56:24<50:35:05,  6.77s/it, lr: 1.0e-04 loss: 6.928e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3083/30000 [6:56:31<50:35:05,  6.77s/it, lr: 1.0e-04 loss: 8.118e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3083/30000 [6:56:31<50:35:05,  6.77s/it, lr: 1.0e-04 loss: 8.118e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3084/30000 [6:56:31<50:10:10,  6.71s/it, lr: 1.0e-04 loss: 8.118e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3084/30000 [6:56:31<50:10:10,  6.71s/it, lr: 1.0e-04 loss: 8.118e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3084/30000 [6:56:37<50:10:10,  6.71s/it, lr: 1.0e-04 loss: 2.774e-01]
Xiang_wan_2_2_lora:  10%|####                                   | 3084/30000 [6:56:37<50:10:10,  6.71s/it, lr: 1.0e-04 loss: 2.774e-01]
Xiang_wan_2_2_lora:  10%|####                                   | 3085/30000 [6:56:37<49:51:42,  6.67s/it, lr: 1.0e-04 loss: 2.774e-01]
Xiang_wan_2_2_lora:  10%|####                                   | 3085/30000 [6:56:37<49:51:42,  6.67s/it, lr: 1.0e-04 loss: 2.774e-01]
Xiang_wan_2_2_lora:  10%|####                                   | 3085/30000 [6:56:44<49:51:42,  6.67s/it, lr: 1.0e-04 loss: 6.545e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3085/30000 [6:56:44<49:51:42,  6.67s/it, lr: 1.0e-04 loss: 6.545e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3086/30000 [6:56:44<49:39:46,  6.64s/it, lr: 1.0e-04 loss: 6.545e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3086/30000 [6:56:44<49:39:46,  6.64s/it, lr: 1.0e-04 loss: 6.545e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3086/30000 [6:56:50<49:39:46,  6.64s/it, lr: 1.0e-04 loss: 8.218e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3086/30000 [6:56:50<49:39:46,  6.64s/it, lr: 1.0e-04 loss: 8.218e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3087/30000 [6:56:50<49:30:19,  6.62s/it, lr: 1.0e-04 loss: 8.218e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3087/30000 [6:56:50<49:30:19,  6.62s/it, lr: 1.0e-04 loss: 8.218e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3087/30000 [6:57:02<49:30:19,  6.62s/it, lr: 1.0e-04 loss: 7.926e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3087/30000 [6:57:02<49:30:19,  6.62s/it, lr: 1.0e-04 loss: 7.926e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3088/30000 [6:57:02<57:00:22,  7.63s/it, lr: 1.0e-04 loss: 7.926e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3088/30000 [6:57:02<57:00:22,  7.63s/it, lr: 1.0e-04 loss: 7.926e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3088/30000 [6:57:11<57:00:22,  7.63s/it, lr: 1.0e-04 loss: 6.559e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3088/30000 [6:57:11<57:00:22,  7.63s/it, lr: 1.0e-04 loss: 6.559e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3089/30000 [6:57:11<61:05:51,  8.17s/it, lr: 1.0e-04 loss: 6.559e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3089/30000 [6:57:11<61:05:51,  8.17s/it, lr: 1.0e-04 loss: 6.559e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3089/30000 [6:57:22<61:05:51,  8.17s/it, lr: 1.0e-04 loss: 6.421e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3089/30000 [6:57:22<61:05:51,  8.17s/it, lr: 1.0e-04 loss: 6.421e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3090/30000 [6:57:32<61:05:42,  8.17s/it, lr: 1.0e-04 loss: 4.773e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3090/30000 [6:57:32<61:05:42,  8.17s/it, lr: 1.0e-04 loss: 4.773e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3091/30000 [6:57:32<50:18:45,  6.73s/it, lr: 1.0e-04 loss: 4.773e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3091/30000 [6:57:32<50:18:45,  6.73s/it, lr: 1.0e-04 loss: 4.773e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3091/30000 [6:57:42<50:18:45,  6.73s/it, lr: 1.0e-04 loss: 7.568e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3091/30000 [6:57:42<50:18:45,  6.73s/it, lr: 1.0e-04 loss: 7.568e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3092/30000 [6:57:42<55:39:07,  7.45s/it, lr: 1.0e-04 loss: 7.568e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3092/30000 [6:57:42<55:39:07,  7.45s/it, lr: 1.0e-04 loss: 7.568e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3092/30000 [6:57:48<55:39:07,  7.45s/it, lr: 1.0e-04 loss: 9.108e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3092/30000 [6:57:48<55:39:07,  7.45s/it, lr: 1.0e-04 loss: 9.108e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3093/30000 [6:57:48<54:23:24,  7.28s/it, lr: 1.0e-04 loss: 9.108e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3093/30000 [6:57:48<54:23:24,  7.28s/it, lr: 1.0e-04 loss: 9.108e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3093/30000 [6:57:55<54:23:24,  7.28s/it, lr: 1.0e-04 loss: 9.859e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3093/30000 [6:57:55<54:23:24,  7.28s/it, lr: 1.0e-04 loss: 9.859e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3094/30000 [6:57:55<53:05:28,  7.10s/it, lr: 1.0e-04 loss: 9.859e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3094/30000 [6:57:55<53:05:28,  7.10s/it, lr: 1.0e-04 loss: 9.859e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3094/30000 [6:58:02<53:05:28,  7.10s/it, lr: 1.0e-04 loss: 3.212e-01]
Xiang_wan_2_2_lora:  10%|####                                   | 3094/30000 [6:58:02<53:05:28,  7.10s/it, lr: 1.0e-04 loss: 3.212e-01]
Xiang_wan_2_2_lora:  10%|####                                   | 3095/30000 [6:58:02<52:06:32,  6.97s/it, lr: 1.0e-04 loss: 3.212e-01]
Xiang_wan_2_2_lora:  10%|####                                   | 3095/30000 [6:58:02<52:06:32,  6.97s/it, lr: 1.0e-04 loss: 3.212e-01]
Xiang_wan_2_2_lora:  10%|####                                   | 3095/30000 [6:58:08<52:06:32,  6.97s/it, lr: 1.0e-04 loss: 8.876e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3095/30000 [6:58:08<52:06:32,  6.97s/it, lr: 1.0e-04 loss: 8.876e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3096/30000 [6:58:08<51:22:19,  6.87s/it, lr: 1.0e-04 loss: 8.876e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3096/30000 [6:58:08<51:22:19,  6.87s/it, lr: 1.0e-04 loss: 8.876e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3096/30000 [6:58:18<51:22:19,  6.87s/it, lr: 1.0e-04 loss: 8.379e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3096/30000 [6:58:18<51:22:19,  6.87s/it, lr: 1.0e-04 loss: 8.379e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3097/30000 [6:58:18<57:16:28,  7.66s/it, lr: 1.0e-04 loss: 8.379e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3097/30000 [6:58:18<57:16:28,  7.66s/it, lr: 1.0e-04 loss: 8.379e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3097/30000 [6:58:24<57:16:28,  7.66s/it, lr: 1.0e-04 loss: 6.600e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3097/30000 [6:58:24<57:16:28,  7.66s/it, lr: 1.0e-04 loss: 6.600e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3098/30000 [6:58:24<54:53:29,  7.35s/it, lr: 1.0e-04 loss: 6.600e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3098/30000 [6:58:24<54:53:29,  7.35s/it, lr: 1.0e-04 loss: 6.600e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3098/30000 [6:58:34<54:53:29,  7.35s/it, lr: 1.0e-04 loss: 6.781e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3098/30000 [6:58:34<54:53:29,  7.35s/it, lr: 1.0e-04 loss: 6.781e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3099/30000 [6:58:34<59:23:30,  7.95s/it, lr: 1.0e-04 loss: 6.781e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3099/30000 [6:58:34<59:23:30,  7.95s/it, lr: 1.0e-04 loss: 6.781e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3099/30000 [6:58:40<59:23:30,  7.95s/it, lr: 1.0e-04 loss: 9.565e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3099/30000 [6:58:40<59:23:30,  7.95s/it, lr: 1.0e-04 loss: 9.565e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3100/30000 [6:58:50<59:23:22,  7.95s/it, lr: 1.0e-04 loss: 4.161e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3100/30000 [6:58:50<59:23:22,  7.95s/it, lr: 1.0e-04 loss: 4.161e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3101/30000 [6:58:50<48:12:12,  6.45s/it, lr: 1.0e-04 loss: 4.161e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3101/30000 [6:58:50<48:12:12,  6.45s/it, lr: 1.0e-04 loss: 4.161e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3101/30000 [6:58:56<48:12:12,  6.45s/it, lr: 1.0e-04 loss: 9.515e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3101/30000 [6:58:56<48:12:12,  6.45s/it, lr: 1.0e-04 loss: 9.515e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3102/30000 [6:58:56<48:31:55,  6.50s/it, lr: 1.0e-04 loss: 9.515e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3102/30000 [6:58:56<48:31:55,  6.50s/it, lr: 1.0e-04 loss: 9.515e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3102/30000 [6:59:03<48:31:55,  6.50s/it, lr: 1.0e-04 loss: 6.405e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3102/30000 [6:59:03<48:31:55,  6.50s/it, lr: 1.0e-04 loss: 6.405e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3103/30000 [6:59:03<48:38:00,  6.51s/it, lr: 1.0e-04 loss: 6.405e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3103/30000 [6:59:03<48:38:00,  6.51s/it, lr: 1.0e-04 loss: 6.405e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3103/30000 [6:59:12<48:38:00,  6.51s/it, lr: 1.0e-04 loss: 7.963e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3103/30000 [6:59:12<48:38:00,  6.51s/it, lr: 1.0e-04 loss: 7.963e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3104/30000 [6:59:12<54:18:38,  7.27s/it, lr: 1.0e-04 loss: 7.963e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3104/30000 [6:59:12<54:18:38,  7.27s/it, lr: 1.0e-04 loss: 7.963e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3104/30000 [6:59:19<54:18:38,  7.27s/it, lr: 1.0e-04 loss: 9.094e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3104/30000 [6:59:19<54:18:38,  7.27s/it, lr: 1.0e-04 loss: 9.094e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3105/30000 [6:59:19<53:23:53,  7.15s/it, lr: 1.0e-04 loss: 9.094e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3105/30000 [6:59:19<53:23:53,  7.15s/it, lr: 1.0e-04 loss: 9.094e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3105/30000 [6:59:26<53:23:53,  7.15s/it, lr: 1.0e-04 loss: 9.615e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3105/30000 [6:59:26<53:23:53,  7.15s/it, lr: 1.0e-04 loss: 9.615e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3106/30000 [6:59:26<52:09:44,  6.98s/it, lr: 1.0e-04 loss: 9.615e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3106/30000 [6:59:26<52:09:44,  6.98s/it, lr: 1.0e-04 loss: 9.615e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3106/30000 [6:59:32<52:09:44,  6.98s/it, lr: 1.0e-04 loss: 3.712e-01]
Xiang_wan_2_2_lora:  10%|####                                   | 3106/30000 [6:59:32<52:09:44,  6.98s/it, lr: 1.0e-04 loss: 3.712e-01]
Xiang_wan_2_2_lora:  10%|####                                   | 3107/30000 [6:59:32<51:17:12,  6.87s/it, lr: 1.0e-04 loss: 3.712e-01]
Xiang_wan_2_2_lora:  10%|####                                   | 3107/30000 [6:59:32<51:17:12,  6.87s/it, lr: 1.0e-04 loss: 3.712e-01]
Xiang_wan_2_2_lora:  10%|####                                   | 3107/30000 [6:59:39<51:17:12,  6.87s/it, lr: 1.0e-04 loss: 5.984e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3107/30000 [6:59:39<51:17:12,  6.87s/it, lr: 1.0e-04 loss: 5.984e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3108/30000 [6:59:39<50:41:12,  6.79s/it, lr: 1.0e-04 loss: 5.984e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3108/30000 [6:59:39<50:41:12,  6.79s/it, lr: 1.0e-04 loss: 5.984e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3108/30000 [6:59:48<50:41:12,  6.79s/it, lr: 1.0e-04 loss: 3.644e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3108/30000 [6:59:48<50:41:12,  6.79s/it, lr: 1.0e-04 loss: 3.644e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3109/30000 [6:59:48<56:14:50,  7.53s/it, lr: 1.0e-04 loss: 3.644e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3109/30000 [6:59:48<56:14:50,  7.53s/it, lr: 1.0e-04 loss: 3.644e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3109/30000 [6:59:55<56:14:50,  7.53s/it, lr: 1.0e-04 loss: 7.945e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3109/30000 [6:59:55<56:14:50,  7.53s/it, lr: 1.0e-04 loss: 7.945e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3110/30000 [7:00:04<56:14:42,  7.53s/it, lr: 1.0e-04 loss: 7.147e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3110/30000 [7:00:04<56:14:42,  7.53s/it, lr: 1.0e-04 loss: 7.147e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3111/30000 [7:00:04<46:20:58,  6.21s/it, lr: 1.0e-04 loss: 7.147e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3111/30000 [7:00:04<46:20:58,  6.21s/it, lr: 1.0e-04 loss: 7.147e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3111/30000 [7:00:11<46:20:58,  6.21s/it, lr: 1.0e-04 loss: 7.509e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3111/30000 [7:00:11<46:20:58,  6.21s/it, lr: 1.0e-04 loss: 7.509e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3112/30000 [7:00:11<47:36:24,  6.37s/it, lr: 1.0e-04 loss: 7.509e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3112/30000 [7:00:11<47:36:24,  6.37s/it, lr: 1.0e-04 loss: 7.509e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3112/30000 [7:00:20<47:36:24,  6.37s/it, lr: 1.0e-04 loss: 4.805e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3112/30000 [7:00:20<47:36:24,  6.37s/it, lr: 1.0e-04 loss: 4.805e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3113/30000 [7:00:20<53:24:48,  7.15s/it, lr: 1.0e-04 loss: 4.805e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3113/30000 [7:00:20<53:24:48,  7.15s/it, lr: 1.0e-04 loss: 4.805e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3113/30000 [7:00:27<53:24:48,  7.15s/it, lr: 1.0e-04 loss: 8.881e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3113/30000 [7:00:27<53:24:48,  7.15s/it, lr: 1.0e-04 loss: 8.881e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3114/30000 [7:00:27<52:22:15,  7.01s/it, lr: 1.0e-04 loss: 8.881e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3114/30000 [7:00:27<52:22:15,  7.01s/it, lr: 1.0e-04 loss: 8.881e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3114/30000 [7:00:36<52:22:15,  7.01s/it, lr: 1.0e-04 loss: 3.467e-01]
Xiang_wan_2_2_lora:  10%|####                                   | 3114/30000 [7:00:36<52:22:15,  7.01s/it, lr: 1.0e-04 loss: 3.467e-01]
Xiang_wan_2_2_lora:  10%|####                                   | 3115/30000 [7:00:36<57:10:43,  7.66s/it, lr: 1.0e-04 loss: 3.467e-01]
Xiang_wan_2_2_lora:  10%|####                                   | 3115/30000 [7:00:36<57:10:43,  7.66s/it, lr: 1.0e-04 loss: 3.467e-01]
Xiang_wan_2_2_lora:  10%|####                                   | 3115/30000 [7:00:43<57:10:43,  7.66s/it, lr: 1.0e-04 loss: 7.158e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3115/30000 [7:00:43<57:10:43,  7.66s/it, lr: 1.0e-04 loss: 7.158e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3116/30000 [7:00:43<54:53:55,  7.35s/it, lr: 1.0e-04 loss: 7.158e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3116/30000 [7:00:43<54:53:55,  7.35s/it, lr: 1.0e-04 loss: 7.158e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3116/30000 [7:00:49<54:53:55,  7.35s/it, lr: 1.0e-04 loss: 7.118e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3116/30000 [7:00:49<54:53:55,  7.35s/it, lr: 1.0e-04 loss: 7.118e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3117/30000 [7:00:49<53:19:27,  7.14s/it, lr: 1.0e-04 loss: 7.118e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3117/30000 [7:00:49<53:19:27,  7.14s/it, lr: 1.0e-04 loss: 7.118e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3117/30000 [7:00:56<53:19:27,  7.14s/it, lr: 1.0e-04 loss: 7.528e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3117/30000 [7:00:56<53:19:27,  7.14s/it, lr: 1.0e-04 loss: 7.528e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3118/30000 [7:00:56<52:13:16,  6.99s/it, lr: 1.0e-04 loss: 7.528e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3118/30000 [7:00:56<52:13:16,  6.99s/it, lr: 1.0e-04 loss: 7.528e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3118/30000 [7:01:05<52:13:16,  6.99s/it, lr: 1.0e-04 loss: 4.817e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3118/30000 [7:01:05<52:13:16,  6.99s/it, lr: 1.0e-04 loss: 4.817e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3119/30000 [7:01:05<57:14:16,  7.67s/it, lr: 1.0e-04 loss: 4.817e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3119/30000 [7:01:05<57:14:16,  7.67s/it, lr: 1.0e-04 loss: 4.817e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3119/30000 [7:01:15<57:14:16,  7.67s/it, lr: 1.0e-04 loss: 6.648e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3119/30000 [7:01:15<57:14:16,  7.67s/it, lr: 1.0e-04 loss: 6.648e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3120/30000 [7:01:22<57:14:08,  7.67s/it, lr: 1.0e-04 loss: 1.411e-01]
Xiang_wan_2_2_lora:  10%|####                                   | 3120/30000 [7:01:22<57:14:08,  7.67s/it, lr: 1.0e-04 loss: 1.411e-01]
Xiang_wan_2_2_lora:  10%|####                                   | 3121/30000 [7:01:22<42:23:32,  5.68s/it, lr: 1.0e-04 loss: 1.411e-01]
Xiang_wan_2_2_lora:  10%|####                                   | 3121/30000 [7:01:22<42:23:32,  5.68s/it, lr: 1.0e-04 loss: 1.411e-01]
Xiang_wan_2_2_lora:  10%|####                                   | 3121/30000 [7:01:31<42:23:32,  5.68s/it, lr: 1.0e-04 loss: 4.321e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3121/30000 [7:01:31<42:23:32,  5.68s/it, lr: 1.0e-04 loss: 4.321e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3122/30000 [7:01:31<49:08:00,  6.58s/it, lr: 1.0e-04 loss: 4.321e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3122/30000 [7:01:31<49:08:00,  6.58s/it, lr: 1.0e-04 loss: 4.321e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3122/30000 [7:01:37<49:08:00,  6.58s/it, lr: 1.0e-04 loss: 5.298e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3122/30000 [7:01:37<49:08:00,  6.58s/it, lr: 1.0e-04 loss: 5.298e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3123/30000 [7:01:37<49:08:04,  6.58s/it, lr: 1.0e-04 loss: 5.298e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3123/30000 [7:01:37<49:08:04,  6.58s/it, lr: 1.0e-04 loss: 5.298e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3123/30000 [7:01:44<49:08:04,  6.58s/it, lr: 1.0e-04 loss: 9.807e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3123/30000 [7:01:44<49:08:04,  6.58s/it, lr: 1.0e-04 loss: 9.807e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3124/30000 [7:01:44<49:12:46,  6.59s/it, lr: 1.0e-04 loss: 9.807e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3124/30000 [7:01:44<49:12:46,  6.59s/it, lr: 1.0e-04 loss: 9.807e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3124/30000 [7:01:53<49:12:46,  6.59s/it, lr: 1.0e-04 loss: 1.272e-01]
Xiang_wan_2_2_lora:  10%|####                                   | 3124/30000 [7:01:53<49:12:46,  6.59s/it, lr: 1.0e-04 loss: 1.272e-01]
Xiang_wan_2_2_lora:  10%|####                                   | 3125/30000 [7:01:53<54:45:18,  7.33s/it, lr: 1.0e-04 loss: 1.272e-01]
Xiang_wan_2_2_lora:  10%|####                                   | 3125/30000 [7:01:53<54:45:18,  7.33s/it, lr: 1.0e-04 loss: 1.272e-01]
Xiang_wan_2_2_lora:  10%|####                                   | 3125/30000 [7:02:03<54:45:18,  7.33s/it, lr: 1.0e-04 loss: 5.468e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3125/30000 [7:02:03<54:45:18,  7.33s/it, lr: 1.0e-04 loss: 5.468e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3126/30000 [7:02:03<59:08:13,  7.92s/it, lr: 1.0e-04 loss: 5.468e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3126/30000 [7:02:03<59:08:13,  7.92s/it, lr: 1.0e-04 loss: 5.468e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3126/30000 [7:02:12<59:08:13,  7.92s/it, lr: 1.0e-04 loss: 6.029e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3126/30000 [7:02:12<59:08:13,  7.92s/it, lr: 1.0e-04 loss: 6.029e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3127/30000 [7:02:12<62:41:14,  8.40s/it, lr: 1.0e-04 loss: 6.029e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3127/30000 [7:02:12<62:41:14,  8.40s/it, lr: 1.0e-04 loss: 6.029e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3127/30000 [7:02:22<62:41:14,  8.40s/it, lr: 1.0e-04 loss: 5.966e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3127/30000 [7:02:22<62:41:14,  8.40s/it, lr: 1.0e-04 loss: 5.966e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3128/30000 [7:02:22<64:49:11,  8.68s/it, lr: 1.0e-04 loss: 5.966e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3128/30000 [7:02:22<64:49:11,  8.68s/it, lr: 1.0e-04 loss: 5.966e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3128/30000 [7:02:31<64:49:11,  8.68s/it, lr: 1.0e-04 loss: 4.155e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3128/30000 [7:02:31<64:49:11,  8.68s/it, lr: 1.0e-04 loss: 4.155e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3129/30000 [7:02:31<66:27:37,  8.90s/it, lr: 1.0e-04 loss: 4.155e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3129/30000 [7:02:31<66:27:37,  8.90s/it, lr: 1.0e-04 loss: 4.155e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3129/30000 [7:02:38<66:27:37,  8.90s/it, lr: 1.0e-04 loss: 8.366e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3129/30000 [7:02:38<66:27:37,  8.90s/it, lr: 1.0e-04 loss: 8.366e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3130/30000 [7:02:47<66:27:28,  8.90s/it, lr: 1.0e-04 loss: 5.317e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3130/30000 [7:02:47<66:27:28,  8.90s/it, lr: 1.0e-04 loss: 5.317e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3131/30000 [7:02:47<52:07:09,  6.98s/it, lr: 1.0e-04 loss: 5.317e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3131/30000 [7:02:47<52:07:09,  6.98s/it, lr: 1.0e-04 loss: 5.317e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3131/30000 [7:02:56<52:07:09,  6.98s/it, lr: 1.0e-04 loss: 4.889e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3131/30000 [7:02:56<52:07:09,  6.98s/it, lr: 1.0e-04 loss: 4.889e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3132/30000 [7:02:56<56:30:57,  7.57s/it, lr: 1.0e-04 loss: 4.889e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3132/30000 [7:02:56<56:30:57,  7.57s/it, lr: 1.0e-04 loss: 4.889e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3132/30000 [7:03:06<56:30:57,  7.57s/it, lr: 1.0e-04 loss: 5.771e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3132/30000 [7:03:06<56:30:57,  7.57s/it, lr: 1.0e-04 loss: 5.771e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3133/30000 [7:03:06<59:59:19,  8.04s/it, lr: 1.0e-04 loss: 5.771e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3133/30000 [7:03:06<59:59:19,  8.04s/it, lr: 1.0e-04 loss: 5.771e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3133/30000 [7:03:15<59:59:19,  8.04s/it, lr: 1.0e-04 loss: 5.824e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3133/30000 [7:03:15<59:59:19,  8.04s/it, lr: 1.0e-04 loss: 5.824e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3134/30000 [7:03:15<62:38:28,  8.39s/it, lr: 1.0e-04 loss: 5.824e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3134/30000 [7:03:15<62:38:28,  8.39s/it, lr: 1.0e-04 loss: 5.824e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3134/30000 [7:03:25<62:38:28,  8.39s/it, lr: 1.0e-04 loss: 2.556e-01]
Xiang_wan_2_2_lora:  10%|####                                   | 3134/30000 [7:03:25<62:38:28,  8.39s/it, lr: 1.0e-04 loss: 2.556e-01]
Xiang_wan_2_2_lora:  10%|####                                   | 3135/30000 [7:03:25<65:04:02,  8.72s/it, lr: 1.0e-04 loss: 2.556e-01]
Xiang_wan_2_2_lora:  10%|####                                   | 3135/30000 [7:03:25<65:04:02,  8.72s/it, lr: 1.0e-04 loss: 2.556e-01]
Xiang_wan_2_2_lora:  10%|####                                   | 3135/30000 [7:03:31<65:04:02,  8.72s/it, lr: 1.0e-04 loss: 1.123e-01]
Xiang_wan_2_2_lora:  10%|####                                   | 3135/30000 [7:03:31<65:04:02,  8.72s/it, lr: 1.0e-04 loss: 1.123e-01]
Xiang_wan_2_2_lora:  10%|####                                   | 3136/30000 [7:03:31<60:31:49,  8.11s/it, lr: 1.0e-04 loss: 1.123e-01]
Xiang_wan_2_2_lora:  10%|####                                   | 3136/30000 [7:03:31<60:31:49,  8.11s/it, lr: 1.0e-04 loss: 1.123e-01]
Xiang_wan_2_2_lora:  10%|####                                   | 3136/30000 [7:03:38<60:31:49,  8.11s/it, lr: 1.0e-04 loss: 5.804e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3136/30000 [7:03:38<60:31:49,  8.11s/it, lr: 1.0e-04 loss: 5.804e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3137/30000 [7:03:38<57:09:11,  7.66s/it, lr: 1.0e-04 loss: 5.804e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3137/30000 [7:03:38<57:09:11,  7.66s/it, lr: 1.0e-04 loss: 5.804e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3137/30000 [7:03:47<57:09:11,  7.66s/it, lr: 1.0e-04 loss: 4.960e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3137/30000 [7:03:47<57:09:11,  7.66s/it, lr: 1.0e-04 loss: 4.960e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3138/30000 [7:03:47<60:46:43,  8.15s/it, lr: 1.0e-04 loss: 4.960e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3138/30000 [7:03:47<60:46:43,  8.15s/it, lr: 1.0e-04 loss: 4.960e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3138/30000 [7:03:54<60:46:43,  8.15s/it, lr: 1.0e-04 loss: 5.720e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3138/30000 [7:03:54<60:46:43,  8.15s/it, lr: 1.0e-04 loss: 5.720e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3139/30000 [7:03:54<57:15:19,  7.67s/it, lr: 1.0e-04 loss: 5.720e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3139/30000 [7:03:54<57:15:19,  7.67s/it, lr: 1.0e-04 loss: 5.720e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3139/30000 [7:04:00<57:15:19,  7.67s/it, lr: 1.0e-04 loss: 7.165e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3139/30000 [7:04:00<57:15:19,  7.67s/it, lr: 1.0e-04 loss: 7.165e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3140/30000 [7:04:07<57:15:11,  7.67s/it, lr: 1.0e-04 loss: 1.293e-01]
Xiang_wan_2_2_lora:  10%|####                                   | 3140/30000 [7:04:07<57:15:11,  7.67s/it, lr: 1.0e-04 loss: 1.293e-01]
Xiang_wan_2_2_lora:  10%|####                                   | 3141/30000 [7:04:07<42:15:14,  5.66s/it, lr: 1.0e-04 loss: 1.293e-01]
Xiang_wan_2_2_lora:  10%|####                                   | 3141/30000 [7:04:07<42:15:14,  5.66s/it, lr: 1.0e-04 loss: 1.293e-01]
Xiang_wan_2_2_lora:  10%|####                                   | 3141/30000 [7:04:14<42:15:14,  5.66s/it, lr: 1.0e-04 loss: 9.106e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3141/30000 [7:04:14<42:15:14,  5.66s/it, lr: 1.0e-04 loss: 9.106e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3142/30000 [7:04:14<44:21:10,  5.94s/it, lr: 1.0e-04 loss: 9.106e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3142/30000 [7:04:14<44:21:10,  5.94s/it, lr: 1.0e-04 loss: 9.106e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3142/30000 [7:04:23<44:21:10,  5.94s/it, lr: 1.0e-04 loss: 3.214e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3142/30000 [7:04:23<44:21:10,  5.94s/it, lr: 1.0e-04 loss: 3.214e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3143/30000 [7:04:23<50:52:05,  6.82s/it, lr: 1.0e-04 loss: 3.214e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3143/30000 [7:04:23<50:52:05,  6.82s/it, lr: 1.0e-04 loss: 3.214e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3143/30000 [7:04:30<50:52:05,  6.82s/it, lr: 1.0e-04 loss: 8.562e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3143/30000 [7:04:30<50:52:05,  6.82s/it, lr: 1.0e-04 loss: 8.562e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3144/30000 [7:04:30<50:27:10,  6.76s/it, lr: 1.0e-04 loss: 8.562e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3144/30000 [7:04:30<50:27:10,  6.76s/it, lr: 1.0e-04 loss: 8.562e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3144/30000 [7:04:36<50:27:10,  6.76s/it, lr: 1.0e-04 loss: 6.445e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3144/30000 [7:04:36<50:27:10,  6.76s/it, lr: 1.0e-04 loss: 6.445e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3145/30000 [7:04:36<49:59:31,  6.70s/it, lr: 1.0e-04 loss: 6.445e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3145/30000 [7:04:36<49:59:31,  6.70s/it, lr: 1.0e-04 loss: 6.445e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3145/30000 [7:04:43<49:59:31,  6.70s/it, lr: 1.0e-04 loss: 9.306e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3145/30000 [7:04:43<49:59:31,  6.70s/it, lr: 1.0e-04 loss: 9.306e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3146/30000 [7:04:43<49:36:52,  6.65s/it, lr: 1.0e-04 loss: 9.306e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3146/30000 [7:04:43<49:36:52,  6.65s/it, lr: 1.0e-04 loss: 9.306e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3146/30000 [7:04:49<49:36:52,  6.65s/it, lr: 1.0e-04 loss: 2.634e-01]
Xiang_wan_2_2_lora:  10%|####                                   | 3146/30000 [7:04:49<49:36:52,  6.65s/it, lr: 1.0e-04 loss: 2.634e-01]
Xiang_wan_2_2_lora:  10%|####                                   | 3147/30000 [7:04:49<49:20:08,  6.61s/it, lr: 1.0e-04 loss: 2.634e-01]
Xiang_wan_2_2_lora:  10%|####                                   | 3147/30000 [7:04:49<49:20:08,  6.61s/it, lr: 1.0e-04 loss: 2.634e-01]
Xiang_wan_2_2_lora:  10%|####                                   | 3147/30000 [7:04:58<49:20:08,  6.61s/it, lr: 1.0e-04 loss: 7.010e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3147/30000 [7:04:58<49:20:08,  6.61s/it, lr: 1.0e-04 loss: 7.010e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3148/30000 [7:04:58<55:11:02,  7.40s/it, lr: 1.0e-04 loss: 7.010e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3148/30000 [7:04:58<55:11:02,  7.40s/it, lr: 1.0e-04 loss: 7.010e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3148/30000 [7:05:05<55:11:02,  7.40s/it, lr: 1.0e-04 loss: 1.483e-01]
Xiang_wan_2_2_lora:  10%|####                                   | 3148/30000 [7:05:05<55:11:02,  7.40s/it, lr: 1.0e-04 loss: 1.483e-01]
Xiang_wan_2_2_lora:  10%|####                                   | 3149/30000 [7:05:05<53:21:50,  7.15s/it, lr: 1.0e-04 loss: 1.483e-01]
Xiang_wan_2_2_lora:  10%|####                                   | 3149/30000 [7:05:05<53:21:50,  7.15s/it, lr: 1.0e-04 loss: 1.483e-01]
Xiang_wan_2_2_lora:  10%|####                                   | 3149/30000 [7:05:15<53:21:50,  7.15s/it, lr: 1.0e-04 loss: 5.095e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3149/30000 [7:05:15<53:21:50,  7.15s/it, lr: 1.0e-04 loss: 5.095e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3150/30000 [7:05:21<53:21:42,  7.15s/it, lr: 1.0e-04 loss: 4.261e-02]
Xiang_wan_2_2_lora:  10%|####                                   | 3150/30000 [7:05:21<53:21:42,  7.15s/it, lr: 1.0e-04 loss: 4.261e-02]
Xiang_wan_2_2_lora:  11%|####                                   | 3151/30000 [7:05:21<40:10:54,  5.39s/it, lr: 1.0e-04 loss: 4.261e-02]
Xiang_wan_2_2_lora:  11%|####                                   | 3151/30000 [7:05:21<40:10:54,  5.39s/it, lr: 1.0e-04 loss: 4.261e-02]
Xiang_wan_2_2_lora:  11%|####                                   | 3151/30000 [7:05:30<40:10:54,  5.39s/it, lr: 1.0e-04 loss: 4.517e-02]
Xiang_wan_2_2_lora:  11%|####                                   | 3151/30000 [7:05:30<40:10:54,  5.39s/it, lr: 1.0e-04 loss: 4.517e-02]
Xiang_wan_2_2_lora:  11%|####                                   | 3152/30000 [7:05:30<47:19:47,  6.35s/it, lr: 1.0e-04 loss: 4.517e-02]
Xiang_wan_2_2_lora:  11%|####                                   | 3152/30000 [7:05:30<47:19:47,  6.35s/it, lr: 1.0e-04 loss: 4.517e-02]
Xiang_wan_2_2_lora:  11%|####                                   | 3152/30000 [7:05:40<47:19:47,  6.35s/it, lr: 1.0e-04 loss: 7.040e-02]
Xiang_wan_2_2_lora:  11%|####                                   | 3152/30000 [7:05:40<47:19:47,  6.35s/it, lr: 1.0e-04 loss: 7.040e-02]
Xiang_wan_2_2_lora:  11%|####                                   | 3153/30000 [7:05:40<53:10:14,  7.13s/it, lr: 1.0e-04 loss: 7.040e-02]
Xiang_wan_2_2_lora:  11%|####                                   | 3153/30000 [7:05:40<53:10:14,  7.13s/it, lr: 1.0e-04 loss: 7.040e-02]
Xiang_wan_2_2_lora:  11%|####                                   | 3153/30000 [7:05:46<53:10:14,  7.13s/it, lr: 1.0e-04 loss: 8.962e-02]
Xiang_wan_2_2_lora:  11%|####                                   | 3153/30000 [7:05:46<53:10:14,  7.13s/it, lr: 1.0e-04 loss: 8.962e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3154/30000 [7:05:46<52:04:21,  6.98s/it, lr: 1.0e-04 loss: 8.962e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3154/30000 [7:05:46<52:04:21,  6.98s/it, lr: 1.0e-04 loss: 8.962e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3154/30000 [7:05:56<52:04:21,  6.98s/it, lr: 1.0e-04 loss: 4.148e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3154/30000 [7:05:56<52:04:21,  6.98s/it, lr: 1.0e-04 loss: 4.148e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3155/30000 [7:05:56<56:52:58,  7.63s/it, lr: 1.0e-04 loss: 4.148e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3155/30000 [7:05:56<56:52:58,  7.63s/it, lr: 1.0e-04 loss: 4.148e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3155/30000 [7:06:02<56:52:58,  7.63s/it, lr: 1.0e-04 loss: 9.395e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3155/30000 [7:06:02<56:52:58,  7.63s/it, lr: 1.0e-04 loss: 9.395e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3156/30000 [7:06:02<54:41:25,  7.33s/it, lr: 1.0e-04 loss: 9.395e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3156/30000 [7:06:02<54:41:25,  7.33s/it, lr: 1.0e-04 loss: 9.395e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3156/30000 [7:06:12<54:41:25,  7.33s/it, lr: 1.0e-04 loss: 5.732e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3156/30000 [7:06:12<54:41:25,  7.33s/it, lr: 1.0e-04 loss: 5.732e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3157/30000 [7:06:12<59:26:53,  7.97s/it, lr: 1.0e-04 loss: 5.732e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3157/30000 [7:06:12<59:26:53,  7.97s/it, lr: 1.0e-04 loss: 5.732e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3157/30000 [7:06:21<59:26:53,  7.97s/it, lr: 1.0e-04 loss: 6.352e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3157/30000 [7:06:21<59:26:53,  7.97s/it, lr: 1.0e-04 loss: 6.352e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3158/30000 [7:06:21<62:19:08,  8.36s/it, lr: 1.0e-04 loss: 6.352e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3158/30000 [7:06:21<62:19:08,  8.36s/it, lr: 1.0e-04 loss: 6.352e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3158/30000 [7:06:30<62:19:08,  8.36s/it, lr: 1.0e-04 loss: 4.520e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3158/30000 [7:06:30<62:19:08,  8.36s/it, lr: 1.0e-04 loss: 4.520e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3159/30000 [7:06:30<64:26:48,  8.64s/it, lr: 1.0e-04 loss: 4.520e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3159/30000 [7:06:30<64:26:48,  8.64s/it, lr: 1.0e-04 loss: 4.520e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3159/30000 [7:06:40<64:26:48,  8.64s/it, lr: 1.0e-04 loss: 6.089e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3159/30000 [7:06:40<64:26:48,  8.64s/it, lr: 1.0e-04 loss: 6.089e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3160/30000 [7:06:46<64:26:39,  8.64s/it, lr: 1.0e-04 loss: 8.562e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3160/30000 [7:06:46<64:26:39,  8.64s/it, lr: 1.0e-04 loss: 8.562e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3161/30000 [7:06:46<46:09:29,  6.19s/it, lr: 1.0e-04 loss: 8.562e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3161/30000 [7:06:46<46:09:29,  6.19s/it, lr: 1.0e-04 loss: 8.562e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3161/30000 [7:06:56<46:09:29,  6.19s/it, lr: 1.0e-04 loss: 7.836e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3161/30000 [7:06:56<46:09:29,  6.19s/it, lr: 1.0e-04 loss: 7.836e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3162/30000 [7:06:56<51:59:26,  6.97s/it, lr: 1.0e-04 loss: 7.836e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3162/30000 [7:06:56<51:59:26,  6.97s/it, lr: 1.0e-04 loss: 7.836e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3162/30000 [7:07:02<51:59:26,  6.97s/it, lr: 1.0e-04 loss: 7.360e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3162/30000 [7:07:02<51:59:26,  6.97s/it, lr: 1.0e-04 loss: 7.360e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3163/30000 [7:07:02<51:22:00,  6.89s/it, lr: 1.0e-04 loss: 7.360e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3163/30000 [7:07:02<51:22:00,  6.89s/it, lr: 1.0e-04 loss: 7.360e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3163/30000 [7:07:09<51:22:00,  6.89s/it, lr: 1.0e-04 loss: 6.524e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3163/30000 [7:07:09<51:22:00,  6.89s/it, lr: 1.0e-04 loss: 6.524e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3164/30000 [7:07:09<50:44:20,  6.81s/it, lr: 1.0e-04 loss: 6.524e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3164/30000 [7:07:09<50:44:20,  6.81s/it, lr: 1.0e-04 loss: 6.524e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3164/30000 [7:07:16<50:44:20,  6.81s/it, lr: 1.0e-04 loss: 4.417e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3164/30000 [7:07:16<50:44:20,  6.81s/it, lr: 1.0e-04 loss: 4.417e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3165/30000 [7:07:16<50:48:57,  6.82s/it, lr: 1.0e-04 loss: 4.417e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3165/30000 [7:07:16<50:48:57,  6.82s/it, lr: 1.0e-04 loss: 4.417e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3165/30000 [7:07:22<50:48:57,  6.82s/it, lr: 1.0e-04 loss: 7.242e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3165/30000 [7:07:22<50:48:57,  6.82s/it, lr: 1.0e-04 loss: 7.242e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3166/30000 [7:07:22<50:07:53,  6.73s/it, lr: 1.0e-04 loss: 7.242e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3166/30000 [7:07:22<50:07:53,  6.73s/it, lr: 1.0e-04 loss: 7.242e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3166/30000 [7:07:31<50:07:53,  6.73s/it, lr: 1.0e-04 loss: 3.408e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3166/30000 [7:07:31<50:07:53,  6.73s/it, lr: 1.0e-04 loss: 3.408e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3167/30000 [7:07:31<55:36:35,  7.46s/it, lr: 1.0e-04 loss: 3.408e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3167/30000 [7:07:31<55:36:35,  7.46s/it, lr: 1.0e-04 loss: 3.408e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3167/30000 [7:07:41<55:36:35,  7.46s/it, lr: 1.0e-04 loss: 8.001e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3167/30000 [7:07:41<55:36:35,  7.46s/it, lr: 1.0e-04 loss: 8.001e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3168/30000 [7:07:41<59:45:26,  8.02s/it, lr: 1.0e-04 loss: 8.001e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3168/30000 [7:07:41<59:45:26,  8.02s/it, lr: 1.0e-04 loss: 8.001e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3168/30000 [7:07:50<59:45:26,  8.02s/it, lr: 1.0e-04 loss: 6.199e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3168/30000 [7:07:50<59:45:26,  8.02s/it, lr: 1.0e-04 loss: 6.199e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3169/30000 [7:07:50<62:35:47,  8.40s/it, lr: 1.0e-04 loss: 6.199e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3169/30000 [7:07:50<62:35:47,  8.40s/it, lr: 1.0e-04 loss: 6.199e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3169/30000 [7:07:57<62:35:47,  8.40s/it, lr: 1.0e-04 loss: 7.342e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3169/30000 [7:07:57<62:35:47,  8.40s/it, lr: 1.0e-04 loss: 7.342e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3170/30000 [7:08:06<62:35:38,  8.40s/it, lr: 1.0e-04 loss: 1.137e-01]
Xiang_wan_2_2_lora:  11%|####1                                  | 3170/30000 [7:08:06<62:35:38,  8.40s/it, lr: 1.0e-04 loss: 1.137e-01]
Xiang_wan_2_2_lora:  11%|####1                                  | 3171/30000 [7:08:06<49:49:54,  6.69s/it, lr: 1.0e-04 loss: 1.137e-01]
Xiang_wan_2_2_lora:  11%|####1                                  | 3171/30000 [7:08:06<49:49:54,  6.69s/it, lr: 1.0e-04 loss: 1.137e-01]
Xiang_wan_2_2_lora:  11%|####1                                  | 3171/30000 [7:08:13<49:49:54,  6.69s/it, lr: 1.0e-04 loss: 8.124e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3171/30000 [7:08:13<49:49:54,  6.69s/it, lr: 1.0e-04 loss: 8.124e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3172/30000 [7:08:13<50:07:19,  6.73s/it, lr: 1.0e-04 loss: 8.124e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3172/30000 [7:08:13<50:07:19,  6.73s/it, lr: 1.0e-04 loss: 8.124e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3172/30000 [7:08:19<50:07:19,  6.73s/it, lr: 1.0e-04 loss: 6.858e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3172/30000 [7:08:19<50:07:19,  6.73s/it, lr: 1.0e-04 loss: 6.858e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3173/30000 [7:08:19<49:50:25,  6.69s/it, lr: 1.0e-04 loss: 6.858e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3173/30000 [7:08:19<49:50:25,  6.69s/it, lr: 1.0e-04 loss: 6.858e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3173/30000 [7:08:26<49:50:25,  6.69s/it, lr: 1.0e-04 loss: 9.005e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3173/30000 [7:08:26<49:50:25,  6.69s/it, lr: 1.0e-04 loss: 9.005e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3174/30000 [7:08:26<49:35:15,  6.65s/it, lr: 1.0e-04 loss: 9.005e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3174/30000 [7:08:26<49:35:15,  6.65s/it, lr: 1.0e-04 loss: 9.005e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3174/30000 [7:08:33<49:35:15,  6.65s/it, lr: 1.0e-04 loss: 5.012e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3174/30000 [7:08:33<49:35:15,  6.65s/it, lr: 1.0e-04 loss: 5.012e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3175/30000 [7:08:33<49:19:54,  6.62s/it, lr: 1.0e-04 loss: 5.012e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3175/30000 [7:08:33<49:19:54,  6.62s/it, lr: 1.0e-04 loss: 5.012e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3175/30000 [7:08:42<49:19:54,  6.62s/it, lr: 1.0e-04 loss: 5.680e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3175/30000 [7:08:42<49:19:54,  6.62s/it, lr: 1.0e-04 loss: 5.680e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3176/30000 [7:08:42<55:07:57,  7.40s/it, lr: 1.0e-04 loss: 5.680e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3176/30000 [7:08:42<55:07:57,  7.40s/it, lr: 1.0e-04 loss: 5.680e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3176/30000 [7:08:51<55:07:57,  7.40s/it, lr: 1.0e-04 loss: 4.318e-01]
Xiang_wan_2_2_lora:  11%|####1                                  | 3176/30000 [7:08:51<55:07:57,  7.40s/it, lr: 1.0e-04 loss: 4.318e-01]
Xiang_wan_2_2_lora:  11%|####1                                  | 3177/30000 [7:08:51<59:17:44,  7.96s/it, lr: 1.0e-04 loss: 4.318e-01]
Xiang_wan_2_2_lora:  11%|####1                                  | 3177/30000 [7:08:51<59:17:44,  7.96s/it, lr: 1.0e-04 loss: 4.318e-01]
Xiang_wan_2_2_lora:  11%|####1                                  | 3177/30000 [7:09:01<59:17:44,  7.96s/it, lr: 1.0e-04 loss: 8.224e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3177/30000 [7:09:01<59:17:44,  7.96s/it, lr: 1.0e-04 loss: 8.224e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3178/30000 [7:09:01<62:14:26,  8.35s/it, lr: 1.0e-04 loss: 8.224e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3178/30000 [7:09:01<62:14:26,  8.35s/it, lr: 1.0e-04 loss: 8.224e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3178/30000 [7:09:07<62:14:26,  8.35s/it, lr: 1.0e-04 loss: 7.589e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3178/30000 [7:09:07<62:14:26,  8.35s/it, lr: 1.0e-04 loss: 7.589e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3179/30000 [7:09:07<58:25:54,  7.84s/it, lr: 1.0e-04 loss: 7.589e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3179/30000 [7:09:07<58:25:54,  7.84s/it, lr: 1.0e-04 loss: 7.589e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3179/30000 [7:09:14<58:25:54,  7.84s/it, lr: 1.0e-04 loss: 7.025e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3179/30000 [7:09:14<58:25:54,  7.84s/it, lr: 1.0e-04 loss: 7.025e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3180/30000 [7:09:21<58:25:46,  7.84s/it, lr: 1.0e-04 loss: 8.839e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3180/30000 [7:09:21<58:25:46,  7.84s/it, lr: 1.0e-04 loss: 8.839e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3181/30000 [7:09:21<42:53:11,  5.76s/it, lr: 1.0e-04 loss: 8.839e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3181/30000 [7:09:21<42:53:11,  5.76s/it, lr: 1.0e-04 loss: 8.839e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3181/30000 [7:09:30<42:53:11,  5.76s/it, lr: 1.0e-04 loss: 5.960e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3181/30000 [7:09:30<42:53:11,  5.76s/it, lr: 1.0e-04 loss: 5.960e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3182/30000 [7:09:30<49:28:08,  6.64s/it, lr: 1.0e-04 loss: 5.960e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3182/30000 [7:09:30<49:28:08,  6.64s/it, lr: 1.0e-04 loss: 5.960e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3182/30000 [7:09:37<49:28:08,  6.64s/it, lr: 1.0e-04 loss: 1.016e-01]
Xiang_wan_2_2_lora:  11%|####1                                  | 3182/30000 [7:09:37<49:28:08,  6.64s/it, lr: 1.0e-04 loss: 1.016e-01]
Xiang_wan_2_2_lora:  11%|####1                                  | 3183/30000 [7:09:37<49:26:13,  6.64s/it, lr: 1.0e-04 loss: 1.016e-01]
Xiang_wan_2_2_lora:  11%|####1                                  | 3183/30000 [7:09:37<49:26:13,  6.64s/it, lr: 1.0e-04 loss: 1.016e-01]
Xiang_wan_2_2_lora:  11%|####1                                  | 3183/30000 [7:09:46<49:26:13,  6.64s/it, lr: 1.0e-04 loss: 5.755e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3183/30000 [7:09:46<49:26:13,  6.64s/it, lr: 1.0e-04 loss: 5.755e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3184/30000 [7:09:46<54:48:18,  7.36s/it, lr: 1.0e-04 loss: 5.755e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3184/30000 [7:09:46<54:48:18,  7.36s/it, lr: 1.0e-04 loss: 5.755e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3184/30000 [7:09:55<54:48:18,  7.36s/it, lr: 1.0e-04 loss: 3.337e-01]
Xiang_wan_2_2_lora:  11%|####1                                  | 3184/30000 [7:09:55<54:48:18,  7.36s/it, lr: 1.0e-04 loss: 3.337e-01]
Xiang_wan_2_2_lora:  11%|####1                                  | 3185/30000 [7:09:55<58:57:07,  7.91s/it, lr: 1.0e-04 loss: 3.337e-01]
Xiang_wan_2_2_lora:  11%|####1                                  | 3185/30000 [7:09:55<58:57:07,  7.91s/it, lr: 1.0e-04 loss: 3.337e-01]
Xiang_wan_2_2_lora:  11%|####1                                  | 3185/30000 [7:10:02<58:57:07,  7.91s/it, lr: 1.0e-04 loss: 9.763e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3185/30000 [7:10:02<58:57:07,  7.91s/it, lr: 1.0e-04 loss: 9.763e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3186/30000 [7:10:02<55:58:28,  7.52s/it, lr: 1.0e-04 loss: 9.763e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3186/30000 [7:10:02<55:58:28,  7.52s/it, lr: 1.0e-04 loss: 9.763e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3186/30000 [7:10:09<55:58:28,  7.52s/it, lr: 1.0e-04 loss: 8.486e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3186/30000 [7:10:09<55:58:28,  7.52s/it, lr: 1.0e-04 loss: 8.486e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3187/30000 [7:10:09<54:34:39,  7.33s/it, lr: 1.0e-04 loss: 8.486e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3187/30000 [7:10:09<54:34:39,  7.33s/it, lr: 1.0e-04 loss: 8.486e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3187/30000 [7:10:15<54:34:39,  7.33s/it, lr: 1.0e-04 loss: 1.063e-01]
Xiang_wan_2_2_lora:  11%|####1                                  | 3187/30000 [7:10:15<54:34:39,  7.33s/it, lr: 1.0e-04 loss: 1.063e-01]
Xiang_wan_2_2_lora:  11%|####1                                  | 3188/30000 [7:10:15<52:44:44,  7.08s/it, lr: 1.0e-04 loss: 1.063e-01]
Xiang_wan_2_2_lora:  11%|####1                                  | 3188/30000 [7:10:15<52:44:44,  7.08s/it, lr: 1.0e-04 loss: 1.063e-01]
Xiang_wan_2_2_lora:  11%|####1                                  | 3188/30000 [7:10:22<52:44:44,  7.08s/it, lr: 1.0e-04 loss: 4.546e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3188/30000 [7:10:22<52:44:44,  7.08s/it, lr: 1.0e-04 loss: 4.546e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3189/30000 [7:10:22<51:43:54,  6.95s/it, lr: 1.0e-04 loss: 4.546e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3189/30000 [7:10:22<51:43:54,  6.95s/it, lr: 1.0e-04 loss: 4.546e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3189/30000 [7:10:28<51:43:54,  6.95s/it, lr: 1.0e-04 loss: 7.422e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3189/30000 [7:10:28<51:43:54,  6.95s/it, lr: 1.0e-04 loss: 7.422e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3190/30000 [7:10:38<51:43:47,  6.95s/it, lr: 1.0e-04 loss: 1.425e-01]
Xiang_wan_2_2_lora:  11%|####1                                  | 3190/30000 [7:10:38<51:43:47,  6.95s/it, lr: 1.0e-04 loss: 1.425e-01]
Xiang_wan_2_2_lora:  11%|####1                                  | 3191/30000 [7:10:38<43:58:24,  5.90s/it, lr: 1.0e-04 loss: 1.425e-01]
Xiang_wan_2_2_lora:  11%|####1                                  | 3191/30000 [7:10:38<43:58:24,  5.90s/it, lr: 1.0e-04 loss: 1.425e-01]
Xiang_wan_2_2_lora:  11%|####1                                  | 3191/30000 [7:10:47<43:58:24,  5.90s/it, lr: 1.0e-04 loss: 5.910e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3191/30000 [7:10:47<43:58:24,  5.90s/it, lr: 1.0e-04 loss: 5.910e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3192/30000 [7:10:47<50:11:25,  6.74s/it, lr: 1.0e-04 loss: 5.910e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3192/30000 [7:10:47<50:11:25,  6.74s/it, lr: 1.0e-04 loss: 5.910e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3192/30000 [7:10:56<50:11:25,  6.74s/it, lr: 1.0e-04 loss: 4.386e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3192/30000 [7:10:56<50:11:25,  6.74s/it, lr: 1.0e-04 loss: 4.386e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3193/30000 [7:10:56<55:12:27,  7.41s/it, lr: 1.0e-04 loss: 4.386e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3193/30000 [7:10:56<55:12:27,  7.41s/it, lr: 1.0e-04 loss: 4.386e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3193/30000 [7:11:06<55:12:27,  7.41s/it, lr: 1.0e-04 loss: 6.760e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3193/30000 [7:11:06<55:12:27,  7.41s/it, lr: 1.0e-04 loss: 6.760e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3194/30000 [7:11:06<59:05:29,  7.94s/it, lr: 1.0e-04 loss: 6.760e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3194/30000 [7:11:06<59:05:29,  7.94s/it, lr: 1.0e-04 loss: 6.760e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3194/30000 [7:11:12<59:05:29,  7.94s/it, lr: 1.0e-04 loss: 4.143e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3194/30000 [7:11:12<59:05:29,  7.94s/it, lr: 1.0e-04 loss: 4.143e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3195/30000 [7:11:12<56:46:41,  7.63s/it, lr: 1.0e-04 loss: 4.143e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3195/30000 [7:11:12<56:46:41,  7.63s/it, lr: 1.0e-04 loss: 4.143e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3195/30000 [7:11:19<56:46:41,  7.63s/it, lr: 1.0e-04 loss: 6.474e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3195/30000 [7:11:19<56:46:41,  7.63s/it, lr: 1.0e-04 loss: 6.474e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3196/30000 [7:11:19<54:29:40,  7.32s/it, lr: 1.0e-04 loss: 6.474e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3196/30000 [7:11:19<54:29:40,  7.32s/it, lr: 1.0e-04 loss: 6.474e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3196/30000 [7:11:28<54:29:40,  7.32s/it, lr: 1.0e-04 loss: 5.489e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3196/30000 [7:11:28<54:29:40,  7.32s/it, lr: 1.0e-04 loss: 5.489e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3197/30000 [7:11:28<58:51:34,  7.91s/it, lr: 1.0e-04 loss: 5.489e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3197/30000 [7:11:28<58:51:34,  7.91s/it, lr: 1.0e-04 loss: 5.489e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3197/30000 [7:11:35<58:51:34,  7.91s/it, lr: 1.0e-04 loss: 9.417e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3197/30000 [7:11:35<58:51:34,  7.91s/it, lr: 1.0e-04 loss: 9.417e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3198/30000 [7:11:35<56:05:53,  7.54s/it, lr: 1.0e-04 loss: 9.417e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3198/30000 [7:11:35<56:05:53,  7.54s/it, lr: 1.0e-04 loss: 9.417e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3198/30000 [7:11:41<56:05:53,  7.54s/it, lr: 1.0e-04 loss: 8.695e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3198/30000 [7:11:41<56:05:53,  7.54s/it, lr: 1.0e-04 loss: 8.695e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3199/30000 [7:11:41<53:53:41,  7.24s/it, lr: 1.0e-04 loss: 8.695e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3199/30000 [7:11:41<53:53:41,  7.24s/it, lr: 1.0e-04 loss: 8.695e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3199/30000 [7:11:48<53:53:41,  7.24s/it, lr: 1.0e-04 loss: 5.551e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3199/30000 [7:11:48<53:53:41,  7.24s/it, lr: 1.0e-04 loss: 5.551e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3200/30000 [7:11:55<53:53:34,  7.24s/it, lr: 1.0e-04 loss: 8.109e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3200/30000 [7:11:55<53:53:34,  7.24s/it, lr: 1.0e-04 loss: 8.109e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3201/30000 [7:11:55<40:28:43,  5.44s/it, lr: 1.0e-04 loss: 8.109e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3201/30000 [7:11:55<40:28:43,  5.44s/it, lr: 1.0e-04 loss: 8.109e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3201/30000 [7:12:01<40:28:43,  5.44s/it, lr: 1.0e-04 loss: 8.614e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3201/30000 [7:12:01<40:28:43,  5.44s/it, lr: 1.0e-04 loss: 8.614e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3202/30000 [7:12:01<43:02:23,  5.78s/it, lr: 1.0e-04 loss: 8.614e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3202/30000 [7:12:01<43:02:23,  5.78s/it, lr: 1.0e-04 loss: 8.614e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3202/30000 [7:12:11<43:02:23,  5.78s/it, lr: 1.0e-04 loss: 4.660e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3202/30000 [7:12:11<43:02:23,  5.78s/it, lr: 1.0e-04 loss: 4.660e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3203/30000 [7:12:11<49:45:30,  6.68s/it, lr: 1.0e-04 loss: 4.660e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3203/30000 [7:12:11<49:45:30,  6.68s/it, lr: 1.0e-04 loss: 4.660e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3203/30000 [7:12:17<49:45:30,  6.68s/it, lr: 1.0e-04 loss: 8.178e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3203/30000 [7:12:17<49:45:30,  6.68s/it, lr: 1.0e-04 loss: 8.178e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3204/30000 [7:12:17<49:31:06,  6.65s/it, lr: 1.0e-04 loss: 8.178e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3204/30000 [7:12:17<49:31:06,  6.65s/it, lr: 1.0e-04 loss: 8.178e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3204/30000 [7:12:27<49:31:06,  6.65s/it, lr: 1.0e-04 loss: 5.624e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3204/30000 [7:12:27<49:31:06,  6.65s/it, lr: 1.0e-04 loss: 5.624e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3205/30000 [7:12:27<54:56:53,  7.38s/it, lr: 1.0e-04 loss: 5.624e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3205/30000 [7:12:27<54:56:53,  7.38s/it, lr: 1.0e-04 loss: 5.624e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3205/30000 [7:12:33<54:56:53,  7.38s/it, lr: 1.0e-04 loss: 7.490e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3205/30000 [7:12:33<54:56:53,  7.38s/it, lr: 1.0e-04 loss: 7.490e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3206/30000 [7:12:33<53:18:20,  7.16s/it, lr: 1.0e-04 loss: 7.490e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3206/30000 [7:12:33<53:18:20,  7.16s/it, lr: 1.0e-04 loss: 7.490e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3206/30000 [7:12:40<53:18:20,  7.16s/it, lr: 1.0e-04 loss: 4.453e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3206/30000 [7:12:40<53:18:20,  7.16s/it, lr: 1.0e-04 loss: 4.453e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3207/30000 [7:12:40<51:58:42,  6.98s/it, lr: 1.0e-04 loss: 4.453e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3207/30000 [7:12:40<51:58:42,  6.98s/it, lr: 1.0e-04 loss: 4.453e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3207/30000 [7:12:49<51:58:42,  6.98s/it, lr: 1.0e-04 loss: 6.772e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3207/30000 [7:12:49<51:58:42,  6.98s/it, lr: 1.0e-04 loss: 6.772e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3208/30000 [7:12:49<56:58:44,  7.66s/it, lr: 1.0e-04 loss: 6.772e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3208/30000 [7:12:49<56:58:44,  7.66s/it, lr: 1.0e-04 loss: 6.772e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3208/30000 [7:12:58<56:58:44,  7.66s/it, lr: 1.0e-04 loss: 5.072e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3208/30000 [7:12:58<56:58:44,  7.66s/it, lr: 1.0e-04 loss: 5.072e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3209/30000 [7:12:58<60:41:39,  8.16s/it, lr: 1.0e-04 loss: 5.072e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3209/30000 [7:12:58<60:41:39,  8.16s/it, lr: 1.0e-04 loss: 5.072e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3209/30000 [7:13:05<60:41:39,  8.16s/it, lr: 1.0e-04 loss: 7.468e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3209/30000 [7:13:05<60:41:39,  8.16s/it, lr: 1.0e-04 loss: 7.468e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3210/30000 [7:13:14<60:41:31,  8.16s/it, lr: 1.0e-04 loss: 1.499e-01]
Xiang_wan_2_2_lora:  11%|####1                                  | 3210/30000 [7:13:14<60:41:31,  8.16s/it, lr: 1.0e-04 loss: 1.499e-01]
Xiang_wan_2_2_lora:  11%|####1                                  | 3211/30000 [7:13:14<48:43:50,  6.55s/it, lr: 1.0e-04 loss: 1.499e-01]
Xiang_wan_2_2_lora:  11%|####1                                  | 3211/30000 [7:13:14<48:43:50,  6.55s/it, lr: 1.0e-04 loss: 1.499e-01]
Xiang_wan_2_2_lora:  11%|####1                                  | 3211/30000 [7:13:21<48:43:50,  6.55s/it, lr: 1.0e-04 loss: 9.921e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3211/30000 [7:13:21<48:43:50,  6.55s/it, lr: 1.0e-04 loss: 9.921e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3212/30000 [7:13:21<48:51:01,  6.56s/it, lr: 1.0e-04 loss: 9.921e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3212/30000 [7:13:21<48:51:01,  6.56s/it, lr: 1.0e-04 loss: 9.921e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3212/30000 [7:13:30<48:51:01,  6.56s/it, lr: 1.0e-04 loss: 5.345e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3212/30000 [7:13:30<48:51:01,  6.56s/it, lr: 1.0e-04 loss: 5.345e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3213/30000 [7:13:30<54:15:04,  7.29s/it, lr: 1.0e-04 loss: 5.345e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3213/30000 [7:13:30<54:15:04,  7.29s/it, lr: 1.0e-04 loss: 5.345e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3213/30000 [7:13:40<54:15:04,  7.29s/it, lr: 1.0e-04 loss: 5.168e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3213/30000 [7:13:40<54:15:04,  7.29s/it, lr: 1.0e-04 loss: 5.168e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3214/30000 [7:13:40<58:23:20,  7.85s/it, lr: 1.0e-04 loss: 5.168e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3214/30000 [7:13:40<58:23:20,  7.85s/it, lr: 1.0e-04 loss: 5.168e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3214/30000 [7:13:49<58:23:20,  7.85s/it, lr: 1.0e-04 loss: 1.266e-01]
Xiang_wan_2_2_lora:  11%|####1                                  | 3214/30000 [7:13:49<58:23:20,  7.85s/it, lr: 1.0e-04 loss: 1.266e-01]
Xiang_wan_2_2_lora:  11%|####1                                  | 3215/30000 [7:13:49<61:31:57,  8.27s/it, lr: 1.0e-04 loss: 1.266e-01]
Xiang_wan_2_2_lora:  11%|####1                                  | 3215/30000 [7:13:49<61:31:57,  8.27s/it, lr: 1.0e-04 loss: 1.266e-01]
Xiang_wan_2_2_lora:  11%|####1                                  | 3215/30000 [7:13:59<61:31:57,  8.27s/it, lr: 1.0e-04 loss: 8.233e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3215/30000 [7:13:59<61:31:57,  8.27s/it, lr: 1.0e-04 loss: 8.233e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3216/30000 [7:13:59<64:02:43,  8.61s/it, lr: 1.0e-04 loss: 8.233e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3216/30000 [7:13:59<64:02:43,  8.61s/it, lr: 1.0e-04 loss: 8.233e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3216/30000 [7:14:05<64:02:43,  8.61s/it, lr: 1.0e-04 loss: 4.996e-01]
Xiang_wan_2_2_lora:  11%|####1                                  | 3216/30000 [7:14:05<64:02:43,  8.61s/it, lr: 1.0e-04 loss: 4.996e-01]
Xiang_wan_2_2_lora:  11%|####1                                  | 3217/30000 [7:14:05<60:14:56,  8.10s/it, lr: 1.0e-04 loss: 4.996e-01]
Xiang_wan_2_2_lora:  11%|####1                                  | 3217/30000 [7:14:05<60:14:56,  8.10s/it, lr: 1.0e-04 loss: 4.996e-01]
Xiang_wan_2_2_lora:  11%|####1                                  | 3217/30000 [7:14:12<60:14:56,  8.10s/it, lr: 1.0e-04 loss: 7.349e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3217/30000 [7:14:12<60:14:56,  8.10s/it, lr: 1.0e-04 loss: 7.349e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3218/30000 [7:14:12<56:56:11,  7.65s/it, lr: 1.0e-04 loss: 7.349e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3218/30000 [7:14:12<56:56:11,  7.65s/it, lr: 1.0e-04 loss: 7.349e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3218/30000 [7:14:21<56:56:11,  7.65s/it, lr: 1.0e-04 loss: 5.289e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3218/30000 [7:14:21<56:56:11,  7.65s/it, lr: 1.0e-04 loss: 5.289e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3219/30000 [7:14:21<60:29:25,  8.13s/it, lr: 1.0e-04 loss: 5.289e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3219/30000 [7:14:21<60:29:25,  8.13s/it, lr: 1.0e-04 loss: 5.289e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3219/30000 [7:14:31<60:29:25,  8.13s/it, lr: 1.0e-04 loss: 6.070e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3219/30000 [7:14:31<60:29:25,  8.13s/it, lr: 1.0e-04 loss: 6.070e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3220/30000 [7:14:37<60:29:17,  8.13s/it, lr: 1.0e-04 loss: 6.312e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3220/30000 [7:14:37<60:29:17,  8.13s/it, lr: 1.0e-04 loss: 6.312e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3221/30000 [7:14:37<43:58:51,  5.91s/it, lr: 1.0e-04 loss: 6.312e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3221/30000 [7:14:37<43:58:51,  5.91s/it, lr: 1.0e-04 loss: 6.312e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3221/30000 [7:14:44<43:58:51,  5.91s/it, lr: 1.0e-04 loss: 6.454e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3221/30000 [7:14:44<43:58:51,  5.91s/it, lr: 1.0e-04 loss: 6.454e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3222/30000 [7:14:44<45:08:04,  6.07s/it, lr: 1.0e-04 loss: 6.454e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3222/30000 [7:14:44<45:08:04,  6.07s/it, lr: 1.0e-04 loss: 6.454e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3222/30000 [7:14:53<45:08:04,  6.07s/it, lr: 1.0e-04 loss: 6.562e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3222/30000 [7:14:53<45:08:04,  6.07s/it, lr: 1.0e-04 loss: 6.562e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3223/30000 [7:14:53<51:19:59,  6.90s/it, lr: 1.0e-04 loss: 6.562e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3223/30000 [7:14:53<51:19:59,  6.90s/it, lr: 1.0e-04 loss: 6.562e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3223/30000 [7:15:00<51:19:59,  6.90s/it, lr: 1.0e-04 loss: 1.255e-01]
Xiang_wan_2_2_lora:  11%|####1                                  | 3223/30000 [7:15:00<51:19:59,  6.90s/it, lr: 1.0e-04 loss: 1.255e-01]
Xiang_wan_2_2_lora:  11%|####1                                  | 3224/30000 [7:15:00<50:46:33,  6.83s/it, lr: 1.0e-04 loss: 1.255e-01]
Xiang_wan_2_2_lora:  11%|####1                                  | 3224/30000 [7:15:00<50:46:33,  6.83s/it, lr: 1.0e-04 loss: 1.255e-01]
Xiang_wan_2_2_lora:  11%|####1                                  | 3224/30000 [7:15:09<50:46:33,  6.83s/it, lr: 1.0e-04 loss: 6.065e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3224/30000 [7:15:09<50:46:33,  6.83s/it, lr: 1.0e-04 loss: 6.065e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3225/30000 [7:15:09<56:18:16,  7.57s/it, lr: 1.0e-04 loss: 6.065e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3225/30000 [7:15:09<56:18:16,  7.57s/it, lr: 1.0e-04 loss: 6.065e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3225/30000 [7:15:19<56:18:16,  7.57s/it, lr: 1.0e-04 loss: 1.053e-01]
Xiang_wan_2_2_lora:  11%|####1                                  | 3225/30000 [7:15:19<56:18:16,  7.57s/it, lr: 1.0e-04 loss: 1.053e-01]
Xiang_wan_2_2_lora:  11%|####1                                  | 3226/30000 [7:15:19<60:05:21,  8.08s/it, lr: 1.0e-04 loss: 1.053e-01]
Xiang_wan_2_2_lora:  11%|####1                                  | 3226/30000 [7:15:19<60:05:21,  8.08s/it, lr: 1.0e-04 loss: 1.053e-01]
Xiang_wan_2_2_lora:  11%|####1                                  | 3226/30000 [7:15:25<60:05:21,  8.08s/it, lr: 1.0e-04 loss: 5.040e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3226/30000 [7:15:25<60:05:21,  8.08s/it, lr: 1.0e-04 loss: 5.040e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3227/30000 [7:15:25<56:54:04,  7.65s/it, lr: 1.0e-04 loss: 5.040e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3227/30000 [7:15:25<56:54:04,  7.65s/it, lr: 1.0e-04 loss: 5.040e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3227/30000 [7:15:32<56:54:04,  7.65s/it, lr: 1.0e-04 loss: 8.598e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3227/30000 [7:15:32<56:54:04,  7.65s/it, lr: 1.0e-04 loss: 8.598e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3228/30000 [7:15:32<54:30:33,  7.33s/it, lr: 1.0e-04 loss: 8.598e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3228/30000 [7:15:32<54:30:33,  7.33s/it, lr: 1.0e-04 loss: 8.598e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3228/30000 [7:15:41<54:30:33,  7.33s/it, lr: 1.0e-04 loss: 1.997e-01]
Xiang_wan_2_2_lora:  11%|####1                                  | 3228/30000 [7:15:41<54:30:33,  7.33s/it, lr: 1.0e-04 loss: 1.997e-01]
Xiang_wan_2_2_lora:  11%|####1                                  | 3229/30000 [7:15:41<58:53:26,  7.92s/it, lr: 1.0e-04 loss: 1.997e-01]
Xiang_wan_2_2_lora:  11%|####1                                  | 3229/30000 [7:15:41<58:53:26,  7.92s/it, lr: 1.0e-04 loss: 1.997e-01]
Xiang_wan_2_2_lora:  11%|####1                                  | 3229/30000 [7:15:50<58:53:26,  7.92s/it, lr: 1.0e-04 loss: 6.265e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3229/30000 [7:15:50<58:53:26,  7.92s/it, lr: 1.0e-04 loss: 6.265e-02]
Xiang_wan_2_2_lora:  11%|####1                                  | 3230/30000 [7:16:00<58:53:18,  7.92s/it, lr: 1.0e-04 loss: 7.407e-01]
Xiang_wan_2_2_lora:  11%|####1                                  | 3230/30000 [7:16:00<58:53:18,  7.92s/it, lr: 1.0e-04 loss: 7.407e-01]
Xiang_wan_2_2_lora:  11%|####2                                  | 3231/30000 [7:16:00<47:49:16,  6.43s/it, lr: 1.0e-04 loss: 7.407e-01]
Xiang_wan_2_2_lora:  11%|####2                                  | 3231/30000 [7:16:00<47:49:16,  6.43s/it, lr: 1.0e-04 loss: 7.407e-01]
Xiang_wan_2_2_lora:  11%|####2                                  | 3231/30000 [7:16:07<47:49:16,  6.43s/it, lr: 1.0e-04 loss: 9.677e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3231/30000 [7:16:07<47:49:16,  6.43s/it, lr: 1.0e-04 loss: 9.677e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3232/30000 [7:16:07<48:34:27,  6.53s/it, lr: 1.0e-04 loss: 9.677e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3232/30000 [7:16:07<48:34:27,  6.53s/it, lr: 1.0e-04 loss: 9.677e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3232/30000 [7:16:13<48:34:27,  6.53s/it, lr: 1.0e-04 loss: 7.229e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3232/30000 [7:16:13<48:34:27,  6.53s/it, lr: 1.0e-04 loss: 7.229e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3233/30000 [7:16:13<48:38:38,  6.54s/it, lr: 1.0e-04 loss: 7.229e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3233/30000 [7:16:13<48:38:38,  6.54s/it, lr: 1.0e-04 loss: 7.229e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3233/30000 [7:16:20<48:38:38,  6.54s/it, lr: 1.0e-04 loss: 9.097e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3233/30000 [7:16:20<48:38:38,  6.54s/it, lr: 1.0e-04 loss: 9.097e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3234/30000 [7:16:20<48:36:44,  6.54s/it, lr: 1.0e-04 loss: 9.097e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3234/30000 [7:16:20<48:36:44,  6.54s/it, lr: 1.0e-04 loss: 9.097e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3234/30000 [7:16:26<48:36:44,  6.54s/it, lr: 1.0e-04 loss: 7.096e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3234/30000 [7:16:26<48:36:44,  6.54s/it, lr: 1.0e-04 loss: 7.096e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3235/30000 [7:16:26<48:36:24,  6.54s/it, lr: 1.0e-04 loss: 7.096e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3235/30000 [7:16:26<48:36:24,  6.54s/it, lr: 1.0e-04 loss: 7.096e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3235/30000 [7:16:35<48:36:24,  6.54s/it, lr: 1.0e-04 loss: 7.259e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3235/30000 [7:16:35<48:36:24,  6.54s/it, lr: 1.0e-04 loss: 7.259e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3236/30000 [7:16:35<54:25:00,  7.32s/it, lr: 1.0e-04 loss: 7.259e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3236/30000 [7:16:35<54:25:00,  7.32s/it, lr: 1.0e-04 loss: 7.259e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3236/30000 [7:16:42<54:25:00,  7.32s/it, lr: 1.0e-04 loss: 9.839e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3236/30000 [7:16:42<54:25:00,  7.32s/it, lr: 1.0e-04 loss: 9.839e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3237/30000 [7:16:42<52:54:06,  7.12s/it, lr: 1.0e-04 loss: 9.839e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3237/30000 [7:16:42<52:54:06,  7.12s/it, lr: 1.0e-04 loss: 9.839e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3237/30000 [7:16:51<52:54:06,  7.12s/it, lr: 1.0e-04 loss: 7.400e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3237/30000 [7:16:51<52:54:06,  7.12s/it, lr: 1.0e-04 loss: 7.400e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3238/30000 [7:16:51<57:46:17,  7.77s/it, lr: 1.0e-04 loss: 7.400e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3238/30000 [7:16:51<57:46:17,  7.77s/it, lr: 1.0e-04 loss: 7.400e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3238/30000 [7:16:58<57:46:17,  7.77s/it, lr: 1.0e-04 loss: 5.468e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3238/30000 [7:16:58<57:46:17,  7.77s/it, lr: 1.0e-04 loss: 5.468e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3239/30000 [7:16:58<55:09:15,  7.42s/it, lr: 1.0e-04 loss: 5.468e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3239/30000 [7:16:58<55:09:15,  7.42s/it, lr: 1.0e-04 loss: 5.468e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3239/30000 [7:17:08<55:09:15,  7.42s/it, lr: 1.0e-04 loss: 6.523e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3239/30000 [7:17:08<55:09:15,  7.42s/it, lr: 1.0e-04 loss: 6.523e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3240/30000 [7:17:14<55:09:08,  7.42s/it, lr: 1.0e-04 loss: 1.579e-01]
Xiang_wan_2_2_lora:  11%|####2                                  | 3240/30000 [7:17:14<55:09:08,  7.42s/it, lr: 1.0e-04 loss: 1.579e-01]
Xiang_wan_2_2_lora:  11%|####2                                  | 3241/30000 [7:17:14<41:09:50,  5.54s/it, lr: 1.0e-04 loss: 1.579e-01]
Xiang_wan_2_2_lora:  11%|####2                                  | 3241/30000 [7:17:14<41:09:50,  5.54s/it, lr: 1.0e-04 loss: 1.579e-01]
Xiang_wan_2_2_lora:  11%|####2                                  | 3241/30000 [7:17:23<41:09:50,  5.54s/it, lr: 1.0e-04 loss: 8.124e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3241/30000 [7:17:23<41:09:50,  5.54s/it, lr: 1.0e-04 loss: 8.124e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3242/30000 [7:17:23<48:03:43,  6.47s/it, lr: 1.0e-04 loss: 8.124e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3242/30000 [7:17:23<48:03:43,  6.47s/it, lr: 1.0e-04 loss: 8.124e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3242/30000 [7:17:30<48:03:43,  6.47s/it, lr: 1.0e-04 loss: 6.755e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3242/30000 [7:17:30<48:03:43,  6.47s/it, lr: 1.0e-04 loss: 6.755e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3243/30000 [7:17:30<48:23:05,  6.51s/it, lr: 1.0e-04 loss: 6.755e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3243/30000 [7:17:30<48:23:05,  6.51s/it, lr: 1.0e-04 loss: 6.755e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3243/30000 [7:17:39<48:23:05,  6.51s/it, lr: 1.0e-04 loss: 7.642e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3243/30000 [7:17:39<48:23:05,  6.51s/it, lr: 1.0e-04 loss: 7.642e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3244/30000 [7:17:39<54:04:48,  7.28s/it, lr: 1.0e-04 loss: 7.642e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3244/30000 [7:17:39<54:04:48,  7.28s/it, lr: 1.0e-04 loss: 7.642e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3244/30000 [7:17:49<54:04:48,  7.28s/it, lr: 1.0e-04 loss: 3.110e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3244/30000 [7:17:49<54:04:48,  7.28s/it, lr: 1.0e-04 loss: 3.110e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3245/30000 [7:17:49<58:20:07,  7.85s/it, lr: 1.0e-04 loss: 3.110e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3245/30000 [7:17:49<58:20:07,  7.85s/it, lr: 1.0e-04 loss: 3.110e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3245/30000 [7:17:58<58:20:07,  7.85s/it, lr: 1.0e-04 loss: 4.964e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3245/30000 [7:17:58<58:20:07,  7.85s/it, lr: 1.0e-04 loss: 4.964e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3246/30000 [7:17:58<61:31:24,  8.28s/it, lr: 1.0e-04 loss: 4.964e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3246/30000 [7:17:58<61:31:24,  8.28s/it, lr: 1.0e-04 loss: 4.964e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3246/30000 [7:18:05<61:31:24,  8.28s/it, lr: 1.0e-04 loss: 9.952e-01]
Xiang_wan_2_2_lora:  11%|####2                                  | 3246/30000 [7:18:05<61:31:24,  8.28s/it, lr: 1.0e-04 loss: 9.952e-01]
Xiang_wan_2_2_lora:  11%|####2                                  | 3247/30000 [7:18:05<58:30:11,  7.87s/it, lr: 1.0e-04 loss: 9.952e-01]
Xiang_wan_2_2_lora:  11%|####2                                  | 3247/30000 [7:18:05<58:30:11,  7.87s/it, lr: 1.0e-04 loss: 9.952e-01]
Xiang_wan_2_2_lora:  11%|####2                                  | 3247/30000 [7:18:14<58:30:11,  7.87s/it, lr: 1.0e-04 loss: 8.301e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3247/30000 [7:18:14<58:30:11,  7.87s/it, lr: 1.0e-04 loss: 8.301e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3248/30000 [7:18:14<61:37:14,  8.29s/it, lr: 1.0e-04 loss: 8.301e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3248/30000 [7:18:14<61:37:14,  8.29s/it, lr: 1.0e-04 loss: 8.301e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3248/30000 [7:18:24<61:37:14,  8.29s/it, lr: 1.0e-04 loss: 4.627e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3248/30000 [7:18:24<61:37:14,  8.29s/it, lr: 1.0e-04 loss: 4.627e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3249/30000 [7:18:24<63:56:20,  8.60s/it, lr: 1.0e-04 loss: 4.627e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3249/30000 [7:18:24<63:56:20,  8.60s/it, lr: 1.0e-04 loss: 4.627e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3249/30000 [7:18:30<63:56:20,  8.60s/it, lr: 1.0e-04 loss: 7.773e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3249/30000 [7:18:30<63:56:20,  8.60s/it, lr: 1.0e-04 loss: 7.773e-02]
+Saving at step 3250
+Saved checkpoint to /home/featurize/ai-toolkit/output/Xiang_wan_2_2_lora/Xiang_wan_2_2_lora_000003250.safetensors
+Saved optimizer to /home/featurize/ai-toolkit/output/Xiang_wan_2_2_lora/optimizer.pt
+
Xiang_wan_2_2_lora:  11%|####2                                  | 3250/30000 [7:18:40<63:56:11,  8.60s/it, lr: 1.0e-04 loss: 5.650e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3250/30000 [7:18:40<63:56:11,  8.60s/it, lr: 1.0e-04 loss: 5.650e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3251/30000 [7:18:40<51:11:18,  6.89s/it, lr: 1.0e-04 loss: 5.650e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3251/30000 [7:18:40<51:11:18,  6.89s/it, lr: 1.0e-04 loss: 5.650e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3251/30000 [7:18:48<51:11:18,  6.89s/it, lr: 1.0e-04 loss: 6.303e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3251/30000 [7:18:48<51:11:18,  6.89s/it, lr: 1.0e-04 loss: 6.303e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3252/30000 [7:18:48<52:29:46,  7.07s/it, lr: 1.0e-04 loss: 6.303e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3252/30000 [7:18:48<52:29:46,  7.07s/it, lr: 1.0e-04 loss: 6.303e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3252/30000 [7:18:55<52:29:46,  7.07s/it, lr: 1.0e-04 loss: 1.734e-01]
Xiang_wan_2_2_lora:  11%|####2                                  | 3252/30000 [7:18:55<52:29:46,  7.07s/it, lr: 1.0e-04 loss: 1.734e-01]
Xiang_wan_2_2_lora:  11%|####2                                  | 3253/30000 [7:18:55<53:26:49,  7.19s/it, lr: 1.0e-04 loss: 1.734e-01]
Xiang_wan_2_2_lora:  11%|####2                                  | 3253/30000 [7:18:55<53:26:49,  7.19s/it, lr: 1.0e-04 loss: 1.734e-01]
Xiang_wan_2_2_lora:  11%|####2                                  | 3253/30000 [7:19:05<53:26:49,  7.19s/it, lr: 1.0e-04 loss: 6.855e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3253/30000 [7:19:05<53:26:49,  7.19s/it, lr: 1.0e-04 loss: 6.855e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3254/30000 [7:19:05<59:36:32,  8.02s/it, lr: 1.0e-04 loss: 6.855e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3254/30000 [7:19:05<59:36:32,  8.02s/it, lr: 1.0e-04 loss: 6.855e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3254/30000 [7:19:16<59:36:32,  8.02s/it, lr: 1.0e-04 loss: 5.505e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3254/30000 [7:19:16<59:36:32,  8.02s/it, lr: 1.0e-04 loss: 5.505e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3255/30000 [7:19:16<64:24:08,  8.67s/it, lr: 1.0e-04 loss: 5.505e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3255/30000 [7:19:16<64:24:08,  8.67s/it, lr: 1.0e-04 loss: 5.505e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3255/30000 [7:19:26<64:24:08,  8.67s/it, lr: 1.0e-04 loss: 5.560e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3255/30000 [7:19:26<64:24:08,  8.67s/it, lr: 1.0e-04 loss: 5.560e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3256/30000 [7:19:26<67:57:15,  9.15s/it, lr: 1.0e-04 loss: 5.560e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3256/30000 [7:19:26<67:57:15,  9.15s/it, lr: 1.0e-04 loss: 5.560e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3256/30000 [7:19:34<67:57:15,  9.15s/it, lr: 1.0e-04 loss: 1.729e-01]
Xiang_wan_2_2_lora:  11%|####2                                  | 3256/30000 [7:19:34<67:57:15,  9.15s/it, lr: 1.0e-04 loss: 1.729e-01]
Xiang_wan_2_2_lora:  11%|####2                                  | 3257/30000 [7:19:34<64:36:52,  8.70s/it, lr: 1.0e-04 loss: 1.729e-01]
Xiang_wan_2_2_lora:  11%|####2                                  | 3257/30000 [7:19:34<64:36:52,  8.70s/it, lr: 1.0e-04 loss: 1.729e-01]
Xiang_wan_2_2_lora:  11%|####2                                  | 3257/30000 [7:19:44<64:36:52,  8.70s/it, lr: 1.0e-04 loss: 7.990e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3257/30000 [7:19:44<64:36:52,  8.70s/it, lr: 1.0e-04 loss: 7.990e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3258/30000 [7:19:44<69:13:16,  9.32s/it, lr: 1.0e-04 loss: 7.990e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3258/30000 [7:19:44<69:13:16,  9.32s/it, lr: 1.0e-04 loss: 7.990e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3258/30000 [7:19:51<69:13:16,  9.32s/it, lr: 1.0e-04 loss: 4.664e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3258/30000 [7:19:51<69:13:16,  9.32s/it, lr: 1.0e-04 loss: 4.664e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3259/30000 [7:19:51<64:09:06,  8.64s/it, lr: 1.0e-04 loss: 4.664e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3259/30000 [7:19:51<64:09:06,  8.64s/it, lr: 1.0e-04 loss: 4.664e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3259/30000 [7:20:01<64:09:06,  8.64s/it, lr: 1.0e-04 loss: 5.983e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3259/30000 [7:20:01<64:09:06,  8.64s/it, lr: 1.0e-04 loss: 5.983e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3260/30000 [7:20:07<64:08:58,  8.64s/it, lr: 1.0e-04 loss: 1.204e-01]
Xiang_wan_2_2_lora:  11%|####2                                  | 3260/30000 [7:20:07<64:08:58,  8.64s/it, lr: 1.0e-04 loss: 1.204e-01]
Xiang_wan_2_2_lora:  11%|####2                                  | 3261/30000 [7:20:07<46:08:49,  6.21s/it, lr: 1.0e-04 loss: 1.204e-01]
Xiang_wan_2_2_lora:  11%|####2                                  | 3261/30000 [7:20:07<46:08:49,  6.21s/it, lr: 1.0e-04 loss: 1.204e-01]
Xiang_wan_2_2_lora:  11%|####2                                  | 3261/30000 [7:20:14<46:08:49,  6.21s/it, lr: 1.0e-04 loss: 7.525e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3261/30000 [7:20:14<46:08:49,  6.21s/it, lr: 1.0e-04 loss: 7.525e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3262/30000 [7:20:14<46:43:56,  6.29s/it, lr: 1.0e-04 loss: 7.525e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3262/30000 [7:20:14<46:43:56,  6.29s/it, lr: 1.0e-04 loss: 7.525e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3262/30000 [7:20:21<46:43:56,  6.29s/it, lr: 1.0e-04 loss: 6.058e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3262/30000 [7:20:21<46:43:56,  6.29s/it, lr: 1.0e-04 loss: 6.058e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3263/30000 [7:20:21<47:19:36,  6.37s/it, lr: 1.0e-04 loss: 6.058e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3263/30000 [7:20:21<47:19:36,  6.37s/it, lr: 1.0e-04 loss: 6.058e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3263/30000 [7:20:27<47:19:36,  6.37s/it, lr: 1.0e-04 loss: 6.782e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3263/30000 [7:20:27<47:19:36,  6.37s/it, lr: 1.0e-04 loss: 6.782e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3264/30000 [7:20:27<47:40:53,  6.42s/it, lr: 1.0e-04 loss: 6.782e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3264/30000 [7:20:27<47:40:53,  6.42s/it, lr: 1.0e-04 loss: 6.782e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3264/30000 [7:20:37<47:40:53,  6.42s/it, lr: 1.0e-04 loss: 4.132e-01]
Xiang_wan_2_2_lora:  11%|####2                                  | 3264/30000 [7:20:37<47:40:53,  6.42s/it, lr: 1.0e-04 loss: 4.132e-01]
Xiang_wan_2_2_lora:  11%|####2                                  | 3265/30000 [7:20:37<54:10:39,  7.30s/it, lr: 1.0e-04 loss: 4.132e-01]
Xiang_wan_2_2_lora:  11%|####2                                  | 3265/30000 [7:20:37<54:10:39,  7.30s/it, lr: 1.0e-04 loss: 4.132e-01]
Xiang_wan_2_2_lora:  11%|####2                                  | 3265/30000 [7:20:43<54:10:39,  7.30s/it, lr: 1.0e-04 loss: 6.915e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3265/30000 [7:20:43<54:10:39,  7.30s/it, lr: 1.0e-04 loss: 6.915e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3266/30000 [7:20:43<52:39:54,  7.09s/it, lr: 1.0e-04 loss: 6.915e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3266/30000 [7:20:43<52:39:54,  7.09s/it, lr: 1.0e-04 loss: 6.915e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3266/30000 [7:20:53<52:39:54,  7.09s/it, lr: 1.0e-04 loss: 3.377e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3266/30000 [7:20:53<52:39:54,  7.09s/it, lr: 1.0e-04 loss: 3.377e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3267/30000 [7:20:53<57:27:27,  7.74s/it, lr: 1.0e-04 loss: 3.377e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3267/30000 [7:20:53<57:27:27,  7.74s/it, lr: 1.0e-04 loss: 3.377e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3267/30000 [7:21:02<57:27:27,  7.74s/it, lr: 1.0e-04 loss: 6.853e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3267/30000 [7:21:02<57:27:27,  7.74s/it, lr: 1.0e-04 loss: 6.853e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3268/30000 [7:21:02<60:58:33,  8.21s/it, lr: 1.0e-04 loss: 6.853e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3268/30000 [7:21:02<60:58:33,  8.21s/it, lr: 1.0e-04 loss: 6.853e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3268/30000 [7:21:11<60:58:33,  8.21s/it, lr: 1.0e-04 loss: 5.195e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3268/30000 [7:21:11<60:58:33,  8.21s/it, lr: 1.0e-04 loss: 5.195e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3269/30000 [7:21:11<63:28:52,  8.55s/it, lr: 1.0e-04 loss: 5.195e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3269/30000 [7:21:11<63:28:52,  8.55s/it, lr: 1.0e-04 loss: 5.195e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3269/30000 [7:21:18<63:28:52,  8.55s/it, lr: 1.0e-04 loss: 7.422e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3269/30000 [7:21:18<63:28:52,  8.55s/it, lr: 1.0e-04 loss: 7.422e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3270/30000 [7:21:24<63:28:44,  8.55s/it, lr: 1.0e-04 loss: 1.785e-01]
Xiang_wan_2_2_lora:  11%|####2                                  | 3270/30000 [7:21:24<63:28:44,  8.55s/it, lr: 1.0e-04 loss: 1.785e-01]
Xiang_wan_2_2_lora:  11%|####2                                  | 3271/30000 [7:21:24<45:34:40,  6.14s/it, lr: 1.0e-04 loss: 1.785e-01]
Xiang_wan_2_2_lora:  11%|####2                                  | 3271/30000 [7:21:24<45:34:40,  6.14s/it, lr: 1.0e-04 loss: 1.785e-01]
Xiang_wan_2_2_lora:  11%|####2                                  | 3271/30000 [7:21:34<45:34:40,  6.14s/it, lr: 1.0e-04 loss: 6.259e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3271/30000 [7:21:34<45:34:40,  6.14s/it, lr: 1.0e-04 loss: 6.259e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3272/30000 [7:21:34<51:18:21,  6.91s/it, lr: 1.0e-04 loss: 6.259e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3272/30000 [7:21:34<51:18:21,  6.91s/it, lr: 1.0e-04 loss: 6.259e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3272/30000 [7:21:43<51:18:21,  6.91s/it, lr: 1.0e-04 loss: 5.263e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3272/30000 [7:21:43<51:18:21,  6.91s/it, lr: 1.0e-04 loss: 5.263e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3273/30000 [7:21:43<56:31:04,  7.61s/it, lr: 1.0e-04 loss: 5.263e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3273/30000 [7:21:43<56:31:04,  7.61s/it, lr: 1.0e-04 loss: 5.263e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3273/30000 [7:21:50<56:31:04,  7.61s/it, lr: 1.0e-04 loss: 8.785e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3273/30000 [7:21:50<56:31:04,  7.61s/it, lr: 1.0e-04 loss: 8.785e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3274/30000 [7:21:50<54:22:14,  7.32s/it, lr: 1.0e-04 loss: 8.785e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3274/30000 [7:21:50<54:22:14,  7.32s/it, lr: 1.0e-04 loss: 8.785e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3274/30000 [7:21:56<54:22:14,  7.32s/it, lr: 1.0e-04 loss: 4.837e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3274/30000 [7:21:56<54:22:14,  7.32s/it, lr: 1.0e-04 loss: 4.837e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3275/30000 [7:21:56<52:53:55,  7.13s/it, lr: 1.0e-04 loss: 4.837e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3275/30000 [7:21:56<52:53:55,  7.13s/it, lr: 1.0e-04 loss: 4.837e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3275/30000 [7:22:06<52:53:55,  7.13s/it, lr: 1.0e-04 loss: 5.179e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3275/30000 [7:22:06<52:53:55,  7.13s/it, lr: 1.0e-04 loss: 5.179e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3276/30000 [7:22:06<57:28:03,  7.74s/it, lr: 1.0e-04 loss: 5.179e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3276/30000 [7:22:06<57:28:03,  7.74s/it, lr: 1.0e-04 loss: 5.179e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3276/30000 [7:22:12<57:28:03,  7.74s/it, lr: 1.0e-04 loss: 7.149e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3276/30000 [7:22:12<57:28:03,  7.74s/it, lr: 1.0e-04 loss: 7.149e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3277/30000 [7:22:12<55:03:21,  7.42s/it, lr: 1.0e-04 loss: 7.149e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3277/30000 [7:22:12<55:03:21,  7.42s/it, lr: 1.0e-04 loss: 7.149e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3277/30000 [7:22:19<55:03:21,  7.42s/it, lr: 1.0e-04 loss: 9.147e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3277/30000 [7:22:19<55:03:21,  7.42s/it, lr: 1.0e-04 loss: 9.147e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3278/30000 [7:22:19<53:05:10,  7.15s/it, lr: 1.0e-04 loss: 9.147e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3278/30000 [7:22:19<53:05:10,  7.15s/it, lr: 1.0e-04 loss: 9.147e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3278/30000 [7:22:28<53:05:10,  7.15s/it, lr: 1.0e-04 loss: 5.330e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3278/30000 [7:22:28<53:05:10,  7.15s/it, lr: 1.0e-04 loss: 5.330e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3279/30000 [7:22:28<57:49:34,  7.79s/it, lr: 1.0e-04 loss: 5.330e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3279/30000 [7:22:28<57:49:34,  7.79s/it, lr: 1.0e-04 loss: 5.330e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3279/30000 [7:22:35<57:49:34,  7.79s/it, lr: 1.0e-04 loss: 9.289e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3279/30000 [7:22:35<57:49:34,  7.79s/it, lr: 1.0e-04 loss: 9.289e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3280/30000 [7:22:42<57:49:26,  7.79s/it, lr: 1.0e-04 loss: 1.286e-01]
Xiang_wan_2_2_lora:  11%|####2                                  | 3280/30000 [7:22:42<57:49:26,  7.79s/it, lr: 1.0e-04 loss: 1.286e-01]
Xiang_wan_2_2_lora:  11%|####2                                  | 3281/30000 [7:22:42<42:31:52,  5.73s/it, lr: 1.0e-04 loss: 1.286e-01]
Xiang_wan_2_2_lora:  11%|####2                                  | 3281/30000 [7:22:42<42:31:52,  5.73s/it, lr: 1.0e-04 loss: 1.286e-01]
Xiang_wan_2_2_lora:  11%|####2                                  | 3281/30000 [7:22:51<42:31:52,  5.73s/it, lr: 1.0e-04 loss: 5.629e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3281/30000 [7:22:51<42:31:52,  5.73s/it, lr: 1.0e-04 loss: 5.629e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3282/30000 [7:22:51<49:00:56,  6.60s/it, lr: 1.0e-04 loss: 5.629e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3282/30000 [7:22:51<49:00:56,  6.60s/it, lr: 1.0e-04 loss: 5.629e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3282/30000 [7:22:58<49:00:56,  6.60s/it, lr: 1.0e-04 loss: 6.790e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3282/30000 [7:22:58<49:00:56,  6.60s/it, lr: 1.0e-04 loss: 6.790e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3283/30000 [7:22:58<49:09:58,  6.62s/it, lr: 1.0e-04 loss: 6.790e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3283/30000 [7:22:58<49:09:58,  6.62s/it, lr: 1.0e-04 loss: 6.790e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3283/30000 [7:23:07<49:09:58,  6.62s/it, lr: 1.0e-04 loss: 6.595e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3283/30000 [7:23:07<49:09:58,  6.62s/it, lr: 1.0e-04 loss: 6.595e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3284/30000 [7:23:07<54:22:45,  7.33s/it, lr: 1.0e-04 loss: 6.595e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3284/30000 [7:23:07<54:22:45,  7.33s/it, lr: 1.0e-04 loss: 6.595e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3284/30000 [7:23:13<54:22:45,  7.33s/it, lr: 1.0e-04 loss: 5.374e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3284/30000 [7:23:13<54:22:45,  7.33s/it, lr: 1.0e-04 loss: 5.374e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3285/30000 [7:23:13<52:57:51,  7.14s/it, lr: 1.0e-04 loss: 5.374e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3285/30000 [7:23:13<52:57:51,  7.14s/it, lr: 1.0e-04 loss: 5.374e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3285/30000 [7:23:20<52:57:51,  7.14s/it, lr: 1.0e-04 loss: 8.220e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3285/30000 [7:23:20<52:57:51,  7.14s/it, lr: 1.0e-04 loss: 8.220e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3286/30000 [7:23:20<51:39:15,  6.96s/it, lr: 1.0e-04 loss: 8.220e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3286/30000 [7:23:20<51:39:15,  6.96s/it, lr: 1.0e-04 loss: 8.220e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3286/30000 [7:23:27<51:39:15,  6.96s/it, lr: 1.0e-04 loss: 6.341e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3286/30000 [7:23:27<51:39:15,  6.96s/it, lr: 1.0e-04 loss: 6.341e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3287/30000 [7:23:27<50:47:25,  6.84s/it, lr: 1.0e-04 loss: 6.341e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3287/30000 [7:23:27<50:47:25,  6.84s/it, lr: 1.0e-04 loss: 6.341e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3287/30000 [7:23:33<50:47:25,  6.84s/it, lr: 1.0e-04 loss: 7.688e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3287/30000 [7:23:33<50:47:25,  6.84s/it, lr: 1.0e-04 loss: 7.688e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3288/30000 [7:23:33<50:42:10,  6.83s/it, lr: 1.0e-04 loss: 7.688e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3288/30000 [7:23:33<50:42:10,  6.83s/it, lr: 1.0e-04 loss: 7.688e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3288/30000 [7:23:43<50:42:10,  6.83s/it, lr: 1.0e-04 loss: 6.735e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3288/30000 [7:23:43<50:42:10,  6.83s/it, lr: 1.0e-04 loss: 6.735e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3289/30000 [7:23:43<56:10:33,  7.57s/it, lr: 1.0e-04 loss: 6.735e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3289/30000 [7:23:43<56:10:33,  7.57s/it, lr: 1.0e-04 loss: 6.735e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3289/30000 [7:23:52<56:10:33,  7.57s/it, lr: 1.0e-04 loss: 8.699e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3289/30000 [7:23:52<56:10:33,  7.57s/it, lr: 1.0e-04 loss: 8.699e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3290/30000 [7:24:01<56:10:25,  7.57s/it, lr: 1.0e-04 loss: 6.496e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3290/30000 [7:24:01<56:10:25,  7.57s/it, lr: 1.0e-04 loss: 6.496e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3291/30000 [7:24:01<46:17:20,  6.24s/it, lr: 1.0e-04 loss: 6.496e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3291/30000 [7:24:01<46:17:20,  6.24s/it, lr: 1.0e-04 loss: 6.496e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3291/30000 [7:24:11<46:17:20,  6.24s/it, lr: 1.0e-04 loss: 6.535e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3291/30000 [7:24:11<46:17:20,  6.24s/it, lr: 1.0e-04 loss: 6.535e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3292/30000 [7:24:11<51:58:13,  7.01s/it, lr: 1.0e-04 loss: 6.535e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3292/30000 [7:24:11<51:58:13,  7.01s/it, lr: 1.0e-04 loss: 6.535e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3292/30000 [7:24:17<51:58:13,  7.01s/it, lr: 1.0e-04 loss: 1.130e-01]
Xiang_wan_2_2_lora:  11%|####2                                  | 3292/30000 [7:24:17<51:58:13,  7.01s/it, lr: 1.0e-04 loss: 1.130e-01]
Xiang_wan_2_2_lora:  11%|####2                                  | 3293/30000 [7:24:17<51:06:09,  6.89s/it, lr: 1.0e-04 loss: 1.130e-01]
Xiang_wan_2_2_lora:  11%|####2                                  | 3293/30000 [7:24:17<51:06:09,  6.89s/it, lr: 1.0e-04 loss: 1.130e-01]
Xiang_wan_2_2_lora:  11%|####2                                  | 3293/30000 [7:24:26<51:06:09,  6.89s/it, lr: 1.0e-04 loss: 6.817e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3293/30000 [7:24:26<51:06:09,  6.89s/it, lr: 1.0e-04 loss: 6.817e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3294/30000 [7:24:26<55:51:19,  7.53s/it, lr: 1.0e-04 loss: 6.817e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3294/30000 [7:24:26<55:51:19,  7.53s/it, lr: 1.0e-04 loss: 6.817e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3294/30000 [7:24:33<55:51:19,  7.53s/it, lr: 1.0e-04 loss: 1.092e-01]
Xiang_wan_2_2_lora:  11%|####2                                  | 3294/30000 [7:24:33<55:51:19,  7.53s/it, lr: 1.0e-04 loss: 1.092e-01]
Xiang_wan_2_2_lora:  11%|####2                                  | 3295/30000 [7:24:33<54:20:03,  7.32s/it, lr: 1.0e-04 loss: 1.092e-01]
Xiang_wan_2_2_lora:  11%|####2                                  | 3295/30000 [7:24:33<54:20:03,  7.32s/it, lr: 1.0e-04 loss: 1.092e-01]
Xiang_wan_2_2_lora:  11%|####2                                  | 3295/30000 [7:24:40<54:20:03,  7.32s/it, lr: 1.0e-04 loss: 1.016e-01]
Xiang_wan_2_2_lora:  11%|####2                                  | 3295/30000 [7:24:40<54:20:03,  7.32s/it, lr: 1.0e-04 loss: 1.016e-01]
Xiang_wan_2_2_lora:  11%|####2                                  | 3296/30000 [7:24:40<52:40:57,  7.10s/it, lr: 1.0e-04 loss: 1.016e-01]
Xiang_wan_2_2_lora:  11%|####2                                  | 3296/30000 [7:24:40<52:40:57,  7.10s/it, lr: 1.0e-04 loss: 1.016e-01]
Xiang_wan_2_2_lora:  11%|####2                                  | 3296/30000 [7:24:49<52:40:57,  7.10s/it, lr: 1.0e-04 loss: 9.072e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3296/30000 [7:24:49<52:40:57,  7.10s/it, lr: 1.0e-04 loss: 9.072e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3297/30000 [7:24:49<57:31:59,  7.76s/it, lr: 1.0e-04 loss: 9.072e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3297/30000 [7:24:49<57:31:59,  7.76s/it, lr: 1.0e-04 loss: 9.072e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3297/30000 [7:24:56<57:31:59,  7.76s/it, lr: 1.0e-04 loss: 9.381e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3297/30000 [7:24:56<57:31:59,  7.76s/it, lr: 1.0e-04 loss: 9.381e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3298/30000 [7:24:56<55:14:50,  7.45s/it, lr: 1.0e-04 loss: 9.381e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3298/30000 [7:24:56<55:14:50,  7.45s/it, lr: 1.0e-04 loss: 9.381e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3298/30000 [7:25:05<55:14:50,  7.45s/it, lr: 1.0e-04 loss: 7.124e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3298/30000 [7:25:05<55:14:50,  7.45s/it, lr: 1.0e-04 loss: 7.124e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3299/30000 [7:25:05<59:26:35,  8.01s/it, lr: 1.0e-04 loss: 7.124e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3299/30000 [7:25:05<59:26:35,  8.01s/it, lr: 1.0e-04 loss: 7.124e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3299/30000 [7:25:12<59:26:35,  8.01s/it, lr: 1.0e-04 loss: 1.069e-01]
Xiang_wan_2_2_lora:  11%|####2                                  | 3299/30000 [7:25:12<59:26:35,  8.01s/it, lr: 1.0e-04 loss: 1.069e-01]
Xiang_wan_2_2_lora:  11%|####2                                  | 3300/30000 [7:25:21<59:26:27,  8.01s/it, lr: 1.0e-04 loss: 8.534e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3300/30000 [7:25:21<59:26:27,  8.01s/it, lr: 1.0e-04 loss: 8.534e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3301/30000 [7:25:21<48:10:34,  6.50s/it, lr: 1.0e-04 loss: 8.534e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3301/30000 [7:25:21<48:10:34,  6.50s/it, lr: 1.0e-04 loss: 8.534e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3301/30000 [7:25:31<48:10:34,  6.50s/it, lr: 1.0e-04 loss: 6.572e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3301/30000 [7:25:31<48:10:34,  6.50s/it, lr: 1.0e-04 loss: 6.572e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3302/30000 [7:25:31<53:28:20,  7.21s/it, lr: 1.0e-04 loss: 6.572e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3302/30000 [7:25:31<53:28:20,  7.21s/it, lr: 1.0e-04 loss: 6.572e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3302/30000 [7:25:40<53:28:20,  7.21s/it, lr: 1.0e-04 loss: 5.772e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3302/30000 [7:25:40<53:28:20,  7.21s/it, lr: 1.0e-04 loss: 5.772e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3303/30000 [7:25:40<58:10:45,  7.85s/it, lr: 1.0e-04 loss: 5.772e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3303/30000 [7:25:40<58:10:45,  7.85s/it, lr: 1.0e-04 loss: 5.772e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3303/30000 [7:25:50<58:10:45,  7.85s/it, lr: 1.0e-04 loss: 7.373e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3303/30000 [7:25:50<58:10:45,  7.85s/it, lr: 1.0e-04 loss: 7.373e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3304/30000 [7:25:50<61:10:46,  8.25s/it, lr: 1.0e-04 loss: 7.373e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3304/30000 [7:25:50<61:10:46,  8.25s/it, lr: 1.0e-04 loss: 7.373e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3304/30000 [7:25:59<61:10:46,  8.25s/it, lr: 1.0e-04 loss: 4.664e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3304/30000 [7:25:59<61:10:46,  8.25s/it, lr: 1.0e-04 loss: 4.664e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3305/30000 [7:25:59<63:29:16,  8.56s/it, lr: 1.0e-04 loss: 4.664e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3305/30000 [7:25:59<63:29:16,  8.56s/it, lr: 1.0e-04 loss: 4.664e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3305/30000 [7:26:08<63:29:16,  8.56s/it, lr: 1.0e-04 loss: 4.592e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3305/30000 [7:26:08<63:29:16,  8.56s/it, lr: 1.0e-04 loss: 4.592e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3306/30000 [7:26:08<65:11:54,  8.79s/it, lr: 1.0e-04 loss: 4.592e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3306/30000 [7:26:08<65:11:54,  8.79s/it, lr: 1.0e-04 loss: 4.592e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3306/30000 [7:26:15<65:11:54,  8.79s/it, lr: 1.0e-04 loss: 7.056e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3306/30000 [7:26:15<65:11:54,  8.79s/it, lr: 1.0e-04 loss: 7.056e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3307/30000 [7:26:15<60:31:41,  8.16s/it, lr: 1.0e-04 loss: 7.056e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3307/30000 [7:26:15<60:31:41,  8.16s/it, lr: 1.0e-04 loss: 7.056e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3307/30000 [7:26:21<60:31:41,  8.16s/it, lr: 1.0e-04 loss: 7.291e-02]
Xiang_wan_2_2_lora:  11%|####2                                  | 3307/30000 [7:26:21<60:31:41,  8.16s/it, lr: 1.0e-04 loss: 7.291e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3308/30000 [7:26:21<57:00:28,  7.69s/it, lr: 1.0e-04 loss: 7.291e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3308/30000 [7:26:21<57:00:28,  7.69s/it, lr: 1.0e-04 loss: 7.291e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3308/30000 [7:26:31<57:00:28,  7.69s/it, lr: 1.0e-04 loss: 8.623e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3308/30000 [7:26:31<57:00:28,  7.69s/it, lr: 1.0e-04 loss: 8.623e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3309/30000 [7:26:31<60:34:40,  8.17s/it, lr: 1.0e-04 loss: 8.623e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3309/30000 [7:26:31<60:34:40,  8.17s/it, lr: 1.0e-04 loss: 8.623e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3309/30000 [7:26:40<60:34:40,  8.17s/it, lr: 1.0e-04 loss: 8.435e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3309/30000 [7:26:40<60:34:40,  8.17s/it, lr: 1.0e-04 loss: 8.435e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3310/30000 [7:26:50<60:34:32,  8.17s/it, lr: 1.0e-04 loss: 1.378e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3310/30000 [7:26:50<60:34:32,  8.17s/it, lr: 1.0e-04 loss: 1.378e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3311/30000 [7:26:50<48:39:15,  6.56s/it, lr: 1.0e-04 loss: 1.378e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3311/30000 [7:26:50<48:39:15,  6.56s/it, lr: 1.0e-04 loss: 1.378e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3311/30000 [7:26:59<48:39:15,  6.56s/it, lr: 1.0e-04 loss: 6.090e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3311/30000 [7:26:59<48:39:15,  6.56s/it, lr: 1.0e-04 loss: 6.090e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3312/30000 [7:26:59<53:50:27,  7.26s/it, lr: 1.0e-04 loss: 6.090e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3312/30000 [7:26:59<53:50:27,  7.26s/it, lr: 1.0e-04 loss: 6.090e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3312/30000 [7:27:08<53:50:27,  7.26s/it, lr: 1.0e-04 loss: 8.603e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3312/30000 [7:27:08<53:50:27,  7.26s/it, lr: 1.0e-04 loss: 8.603e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3313/30000 [7:27:09<57:54:35,  7.81s/it, lr: 1.0e-04 loss: 8.603e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3313/30000 [7:27:09<57:54:35,  7.81s/it, lr: 1.0e-04 loss: 8.603e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3313/30000 [7:27:18<57:54:35,  7.81s/it, lr: 1.0e-04 loss: 6.458e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3313/30000 [7:27:18<57:54:35,  7.81s/it, lr: 1.0e-04 loss: 6.458e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3314/30000 [7:27:18<61:01:42,  8.23s/it, lr: 1.0e-04 loss: 6.458e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3314/30000 [7:27:18<61:01:42,  8.23s/it, lr: 1.0e-04 loss: 6.458e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3314/30000 [7:27:24<61:01:42,  8.23s/it, lr: 1.0e-04 loss: 1.329e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3314/30000 [7:27:24<61:01:42,  8.23s/it, lr: 1.0e-04 loss: 1.329e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3315/30000 [7:27:24<57:37:13,  7.77s/it, lr: 1.0e-04 loss: 1.329e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3315/30000 [7:27:24<57:37:13,  7.77s/it, lr: 1.0e-04 loss: 1.329e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3315/30000 [7:27:31<57:37:13,  7.77s/it, lr: 1.0e-04 loss: 1.112e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3315/30000 [7:27:31<57:37:13,  7.77s/it, lr: 1.0e-04 loss: 1.112e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3316/30000 [7:27:31<54:59:14,  7.42s/it, lr: 1.0e-04 loss: 1.112e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3316/30000 [7:27:31<54:59:14,  7.42s/it, lr: 1.0e-04 loss: 1.112e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3316/30000 [7:27:40<54:59:14,  7.42s/it, lr: 1.0e-04 loss: 4.692e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3316/30000 [7:27:40<54:59:14,  7.42s/it, lr: 1.0e-04 loss: 4.692e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3317/30000 [7:27:40<59:23:54,  8.01s/it, lr: 1.0e-04 loss: 4.692e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3317/30000 [7:27:40<59:23:54,  8.01s/it, lr: 1.0e-04 loss: 4.692e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3317/30000 [7:27:50<59:23:54,  8.01s/it, lr: 1.0e-04 loss: 4.806e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3317/30000 [7:27:50<59:23:54,  8.01s/it, lr: 1.0e-04 loss: 4.806e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3318/30000 [7:27:50<62:26:11,  8.42s/it, lr: 1.0e-04 loss: 4.806e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3318/30000 [7:27:50<62:26:11,  8.42s/it, lr: 1.0e-04 loss: 4.806e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3318/30000 [7:27:59<62:26:11,  8.42s/it, lr: 1.0e-04 loss: 5.683e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3318/30000 [7:27:59<62:26:11,  8.42s/it, lr: 1.0e-04 loss: 5.683e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3319/30000 [7:27:59<64:32:44,  8.71s/it, lr: 1.0e-04 loss: 5.683e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3319/30000 [7:27:59<64:32:44,  8.71s/it, lr: 1.0e-04 loss: 5.683e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3319/30000 [7:28:09<64:32:44,  8.71s/it, lr: 1.0e-04 loss: 7.897e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3319/30000 [7:28:09<64:32:44,  8.71s/it, lr: 1.0e-04 loss: 7.897e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3320/30000 [7:28:18<64:32:35,  8.71s/it, lr: 1.0e-04 loss: 4.078e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3320/30000 [7:28:18<64:32:35,  8.71s/it, lr: 1.0e-04 loss: 4.078e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3321/30000 [7:28:18<50:56:09,  6.87s/it, lr: 1.0e-04 loss: 4.078e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3321/30000 [7:28:18<50:56:09,  6.87s/it, lr: 1.0e-04 loss: 4.078e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3321/30000 [7:28:25<50:56:09,  6.87s/it, lr: 1.0e-04 loss: 5.200e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3321/30000 [7:28:25<50:56:09,  6.87s/it, lr: 1.0e-04 loss: 5.200e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3322/30000 [7:28:25<50:25:44,  6.81s/it, lr: 1.0e-04 loss: 5.200e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3322/30000 [7:28:25<50:25:44,  6.81s/it, lr: 1.0e-04 loss: 5.200e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3322/30000 [7:28:31<50:25:44,  6.81s/it, lr: 1.0e-04 loss: 6.031e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3322/30000 [7:28:31<50:25:44,  6.81s/it, lr: 1.0e-04 loss: 6.031e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3323/30000 [7:28:31<50:08:21,  6.77s/it, lr: 1.0e-04 loss: 6.031e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3323/30000 [7:28:31<50:08:21,  6.77s/it, lr: 1.0e-04 loss: 6.031e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3323/30000 [7:28:38<50:08:21,  6.77s/it, lr: 1.0e-04 loss: 1.063e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3323/30000 [7:28:38<50:08:21,  6.77s/it, lr: 1.0e-04 loss: 1.063e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3324/30000 [7:28:38<49:58:40,  6.74s/it, lr: 1.0e-04 loss: 1.063e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3324/30000 [7:28:38<49:58:40,  6.74s/it, lr: 1.0e-04 loss: 1.063e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3324/30000 [7:28:48<49:58:40,  6.74s/it, lr: 1.0e-04 loss: 5.398e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3324/30000 [7:28:48<49:58:40,  6.74s/it, lr: 1.0e-04 loss: 5.398e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3325/30000 [7:28:48<55:55:22,  7.55s/it, lr: 1.0e-04 loss: 5.398e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3325/30000 [7:28:48<55:55:22,  7.55s/it, lr: 1.0e-04 loss: 5.398e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3325/30000 [7:28:57<55:55:22,  7.55s/it, lr: 1.0e-04 loss: 5.583e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3325/30000 [7:28:57<55:55:22,  7.55s/it, lr: 1.0e-04 loss: 5.583e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3326/30000 [7:28:57<59:41:34,  8.06s/it, lr: 1.0e-04 loss: 5.583e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3326/30000 [7:28:57<59:41:34,  8.06s/it, lr: 1.0e-04 loss: 5.583e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3326/30000 [7:29:06<59:41:34,  8.06s/it, lr: 1.0e-04 loss: 5.029e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3326/30000 [7:29:06<59:41:34,  8.06s/it, lr: 1.0e-04 loss: 5.029e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3327/30000 [7:29:06<62:26:44,  8.43s/it, lr: 1.0e-04 loss: 5.029e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3327/30000 [7:29:06<62:26:44,  8.43s/it, lr: 1.0e-04 loss: 5.029e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3327/30000 [7:29:13<62:26:44,  8.43s/it, lr: 1.0e-04 loss: 8.744e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3327/30000 [7:29:13<62:26:44,  8.43s/it, lr: 1.0e-04 loss: 8.744e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3328/30000 [7:29:13<58:23:52,  7.88s/it, lr: 1.0e-04 loss: 8.744e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3328/30000 [7:29:13<58:23:52,  7.88s/it, lr: 1.0e-04 loss: 8.744e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3328/30000 [7:29:20<58:23:52,  7.88s/it, lr: 1.0e-04 loss: 6.901e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3328/30000 [7:29:20<58:23:52,  7.88s/it, lr: 1.0e-04 loss: 6.901e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3329/30000 [7:29:20<55:38:56,  7.51s/it, lr: 1.0e-04 loss: 6.901e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3329/30000 [7:29:20<55:38:56,  7.51s/it, lr: 1.0e-04 loss: 6.901e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3329/30000 [7:29:26<55:38:56,  7.51s/it, lr: 1.0e-04 loss: 8.630e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3329/30000 [7:29:26<55:38:56,  7.51s/it, lr: 1.0e-04 loss: 8.630e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3330/30000 [7:29:33<55:38:48,  7.51s/it, lr: 1.0e-04 loss: 4.232e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3330/30000 [7:29:33<55:38:48,  7.51s/it, lr: 1.0e-04 loss: 4.232e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3331/30000 [7:29:33<41:19:40,  5.58s/it, lr: 1.0e-04 loss: 4.232e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3331/30000 [7:29:33<41:19:40,  5.58s/it, lr: 1.0e-04 loss: 4.232e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3331/30000 [7:29:40<41:19:40,  5.58s/it, lr: 1.0e-04 loss: 1.024e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3331/30000 [7:29:40<41:19:40,  5.58s/it, lr: 1.0e-04 loss: 1.024e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3332/30000 [7:29:40<43:35:35,  5.88s/it, lr: 1.0e-04 loss: 1.024e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3332/30000 [7:29:40<43:35:35,  5.88s/it, lr: 1.0e-04 loss: 1.024e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3332/30000 [7:29:49<43:35:35,  5.88s/it, lr: 1.0e-04 loss: 3.761e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3332/30000 [7:29:49<43:35:35,  5.88s/it, lr: 1.0e-04 loss: 3.761e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3333/30000 [7:29:49<50:11:19,  6.78s/it, lr: 1.0e-04 loss: 3.761e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3333/30000 [7:29:49<50:11:19,  6.78s/it, lr: 1.0e-04 loss: 3.761e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3333/30000 [7:29:55<50:11:19,  6.78s/it, lr: 1.0e-04 loss: 7.310e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3333/30000 [7:29:55<50:11:19,  6.78s/it, lr: 1.0e-04 loss: 7.310e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3334/30000 [7:29:55<49:46:29,  6.72s/it, lr: 1.0e-04 loss: 7.310e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3334/30000 [7:29:55<49:46:29,  6.72s/it, lr: 1.0e-04 loss: 7.310e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3334/30000 [7:30:05<49:46:29,  6.72s/it, lr: 1.0e-04 loss: 2.384e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3334/30000 [7:30:05<49:46:29,  6.72s/it, lr: 1.0e-04 loss: 2.384e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3335/30000 [7:30:05<55:08:00,  7.44s/it, lr: 1.0e-04 loss: 2.384e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3335/30000 [7:30:05<55:08:00,  7.44s/it, lr: 1.0e-04 loss: 2.384e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3335/30000 [7:30:11<55:08:00,  7.44s/it, lr: 1.0e-04 loss: 7.751e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3335/30000 [7:30:11<55:08:00,  7.44s/it, lr: 1.0e-04 loss: 7.751e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3336/30000 [7:30:11<53:15:21,  7.19s/it, lr: 1.0e-04 loss: 7.751e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3336/30000 [7:30:11<53:15:21,  7.19s/it, lr: 1.0e-04 loss: 7.751e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3336/30000 [7:30:21<53:15:21,  7.19s/it, lr: 1.0e-04 loss: 8.455e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3336/30000 [7:30:21<53:15:21,  7.19s/it, lr: 1.0e-04 loss: 8.455e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3337/30000 [7:30:21<57:43:37,  7.79s/it, lr: 1.0e-04 loss: 8.455e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3337/30000 [7:30:21<57:43:37,  7.79s/it, lr: 1.0e-04 loss: 8.455e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3337/30000 [7:30:30<57:43:37,  7.79s/it, lr: 1.0e-04 loss: 9.245e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3337/30000 [7:30:30<57:43:37,  7.79s/it, lr: 1.0e-04 loss: 9.245e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3338/30000 [7:30:30<61:03:25,  8.24s/it, lr: 1.0e-04 loss: 9.245e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3338/30000 [7:30:30<61:03:25,  8.24s/it, lr: 1.0e-04 loss: 9.245e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3338/30000 [7:30:36<61:03:25,  8.24s/it, lr: 1.0e-04 loss: 6.726e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3338/30000 [7:30:36<61:03:25,  8.24s/it, lr: 1.0e-04 loss: 6.726e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3339/30000 [7:30:36<57:26:38,  7.76s/it, lr: 1.0e-04 loss: 6.726e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3339/30000 [7:30:36<57:26:38,  7.76s/it, lr: 1.0e-04 loss: 6.726e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3339/30000 [7:30:43<57:26:38,  7.76s/it, lr: 1.0e-04 loss: 8.943e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3339/30000 [7:30:43<57:26:38,  7.76s/it, lr: 1.0e-04 loss: 8.943e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3340/30000 [7:30:53<57:26:31,  7.76s/it, lr: 1.0e-04 loss: 4.925e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3340/30000 [7:30:53<57:26:31,  7.76s/it, lr: 1.0e-04 loss: 4.925e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3341/30000 [7:30:53<47:00:34,  6.35s/it, lr: 1.0e-04 loss: 4.925e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3341/30000 [7:30:53<47:00:34,  6.35s/it, lr: 1.0e-04 loss: 4.925e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3341/30000 [7:30:59<47:00:34,  6.35s/it, lr: 1.0e-04 loss: 7.062e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3341/30000 [7:30:59<47:00:34,  6.35s/it, lr: 1.0e-04 loss: 7.062e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3342/30000 [7:30:59<47:20:18,  6.39s/it, lr: 1.0e-04 loss: 7.062e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3342/30000 [7:30:59<47:20:18,  6.39s/it, lr: 1.0e-04 loss: 7.062e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3342/30000 [7:31:09<47:20:18,  6.39s/it, lr: 1.0e-04 loss: 5.435e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3342/30000 [7:31:09<47:20:18,  6.39s/it, lr: 1.0e-04 loss: 5.435e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3343/30000 [7:31:09<52:56:29,  7.15s/it, lr: 1.0e-04 loss: 5.435e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3343/30000 [7:31:09<52:56:29,  7.15s/it, lr: 1.0e-04 loss: 5.435e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3343/30000 [7:31:15<52:56:29,  7.15s/it, lr: 1.0e-04 loss: 8.076e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3343/30000 [7:31:15<52:56:29,  7.15s/it, lr: 1.0e-04 loss: 8.076e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3344/30000 [7:31:15<51:50:22,  7.00s/it, lr: 1.0e-04 loss: 8.076e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3344/30000 [7:31:15<51:50:22,  7.00s/it, lr: 1.0e-04 loss: 8.076e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3344/30000 [7:31:24<51:50:22,  7.00s/it, lr: 1.0e-04 loss: 4.219e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3344/30000 [7:31:24<51:50:22,  7.00s/it, lr: 1.0e-04 loss: 4.219e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3345/30000 [7:31:24<56:41:23,  7.66s/it, lr: 1.0e-04 loss: 4.219e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3345/30000 [7:31:24<56:41:23,  7.66s/it, lr: 1.0e-04 loss: 4.219e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3345/30000 [7:31:31<56:41:23,  7.66s/it, lr: 1.0e-04 loss: 1.023e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3345/30000 [7:31:31<56:41:23,  7.66s/it, lr: 1.0e-04 loss: 1.023e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3346/30000 [7:31:31<54:26:44,  7.35s/it, lr: 1.0e-04 loss: 1.023e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3346/30000 [7:31:31<54:26:44,  7.35s/it, lr: 1.0e-04 loss: 1.023e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3346/30000 [7:31:38<54:26:44,  7.35s/it, lr: 1.0e-04 loss: 6.245e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3346/30000 [7:31:38<54:26:44,  7.35s/it, lr: 1.0e-04 loss: 6.245e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3347/30000 [7:31:38<53:19:26,  7.20s/it, lr: 1.0e-04 loss: 6.245e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3347/30000 [7:31:38<53:19:26,  7.20s/it, lr: 1.0e-04 loss: 6.245e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3347/30000 [7:31:47<53:19:26,  7.20s/it, lr: 1.0e-04 loss: 7.734e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3347/30000 [7:31:47<53:19:26,  7.20s/it, lr: 1.0e-04 loss: 7.734e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3348/30000 [7:31:47<57:49:43,  7.81s/it, lr: 1.0e-04 loss: 7.734e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3348/30000 [7:31:47<57:49:43,  7.81s/it, lr: 1.0e-04 loss: 7.734e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3348/30000 [7:31:57<57:49:43,  7.81s/it, lr: 1.0e-04 loss: 2.657e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3348/30000 [7:31:57<57:49:43,  7.81s/it, lr: 1.0e-04 loss: 2.657e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3349/30000 [7:31:57<61:13:00,  8.27s/it, lr: 1.0e-04 loss: 2.657e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3349/30000 [7:31:57<61:13:00,  8.27s/it, lr: 1.0e-04 loss: 2.657e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3349/30000 [7:32:06<61:13:00,  8.27s/it, lr: 1.0e-04 loss: 8.638e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3349/30000 [7:32:06<61:13:00,  8.27s/it, lr: 1.0e-04 loss: 8.638e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3350/30000 [7:32:12<61:12:52,  8.27s/it, lr: 1.0e-04 loss: 1.051e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3350/30000 [7:32:12<61:12:52,  8.27s/it, lr: 1.0e-04 loss: 1.051e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3351/30000 [7:32:12<44:28:58,  6.01s/it, lr: 1.0e-04 loss: 1.051e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3351/30000 [7:32:12<44:28:58,  6.01s/it, lr: 1.0e-04 loss: 1.051e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3351/30000 [7:32:19<44:28:58,  6.01s/it, lr: 1.0e-04 loss: 7.579e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3351/30000 [7:32:19<44:28:58,  6.01s/it, lr: 1.0e-04 loss: 7.579e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3352/30000 [7:32:19<45:28:38,  6.14s/it, lr: 1.0e-04 loss: 7.579e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3352/30000 [7:32:19<45:28:38,  6.14s/it, lr: 1.0e-04 loss: 7.579e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3352/30000 [7:32:28<45:28:38,  6.14s/it, lr: 1.0e-04 loss: 6.359e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3352/30000 [7:32:28<45:28:38,  6.14s/it, lr: 1.0e-04 loss: 6.359e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3353/30000 [7:32:28<51:36:21,  6.97s/it, lr: 1.0e-04 loss: 6.359e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3353/30000 [7:32:28<51:36:21,  6.97s/it, lr: 1.0e-04 loss: 6.359e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3353/30000 [7:32:35<51:36:21,  6.97s/it, lr: 1.0e-04 loss: 6.344e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3353/30000 [7:32:35<51:36:21,  6.97s/it, lr: 1.0e-04 loss: 6.344e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3354/30000 [7:32:35<50:54:49,  6.88s/it, lr: 1.0e-04 loss: 6.344e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3354/30000 [7:32:35<50:54:49,  6.88s/it, lr: 1.0e-04 loss: 6.344e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3354/30000 [7:32:44<50:54:49,  6.88s/it, lr: 1.0e-04 loss: 5.001e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3354/30000 [7:32:44<50:54:49,  6.88s/it, lr: 1.0e-04 loss: 5.001e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3355/30000 [7:32:44<56:20:48,  7.61s/it, lr: 1.0e-04 loss: 5.001e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3355/30000 [7:32:44<56:20:48,  7.61s/it, lr: 1.0e-04 loss: 5.001e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3355/30000 [7:32:54<56:20:48,  7.61s/it, lr: 1.0e-04 loss: 7.250e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3355/30000 [7:32:54<56:20:48,  7.61s/it, lr: 1.0e-04 loss: 7.250e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3356/30000 [7:32:54<60:04:39,  8.12s/it, lr: 1.0e-04 loss: 7.250e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3356/30000 [7:32:54<60:04:39,  8.12s/it, lr: 1.0e-04 loss: 7.250e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3356/30000 [7:33:03<60:04:39,  8.12s/it, lr: 1.0e-04 loss: 3.714e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3356/30000 [7:33:03<60:04:39,  8.12s/it, lr: 1.0e-04 loss: 3.714e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3357/30000 [7:33:03<62:45:20,  8.48s/it, lr: 1.0e-04 loss: 3.714e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3357/30000 [7:33:03<62:45:20,  8.48s/it, lr: 1.0e-04 loss: 3.714e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3357/30000 [7:33:13<62:45:20,  8.48s/it, lr: 1.0e-04 loss: 8.672e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3357/30000 [7:33:13<62:45:20,  8.48s/it, lr: 1.0e-04 loss: 8.672e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3358/30000 [7:33:13<64:39:09,  8.74s/it, lr: 1.0e-04 loss: 8.672e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3358/30000 [7:33:13<64:39:09,  8.74s/it, lr: 1.0e-04 loss: 8.672e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3358/30000 [7:33:22<64:39:09,  8.74s/it, lr: 1.0e-04 loss: 4.067e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3358/30000 [7:33:22<64:39:09,  8.74s/it, lr: 1.0e-04 loss: 4.067e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3359/30000 [7:33:22<65:59:51,  8.92s/it, lr: 1.0e-04 loss: 4.067e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3359/30000 [7:33:22<65:59:51,  8.92s/it, lr: 1.0e-04 loss: 4.067e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3359/30000 [7:33:31<65:59:51,  8.92s/it, lr: 1.0e-04 loss: 1.419e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3359/30000 [7:33:31<65:59:51,  8.92s/it, lr: 1.0e-04 loss: 1.419e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3360/30000 [7:33:41<65:59:42,  8.92s/it, lr: 1.0e-04 loss: 4.939e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3360/30000 [7:33:41<65:59:42,  8.92s/it, lr: 1.0e-04 loss: 4.939e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3361/30000 [7:33:41<51:33:54,  6.97s/it, lr: 1.0e-04 loss: 4.939e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3361/30000 [7:33:41<51:33:54,  6.97s/it, lr: 1.0e-04 loss: 4.939e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3361/30000 [7:33:50<51:33:54,  6.97s/it, lr: 1.0e-04 loss: 6.457e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3361/30000 [7:33:50<51:33:54,  6.97s/it, lr: 1.0e-04 loss: 6.457e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3362/30000 [7:33:50<56:19:50,  7.61s/it, lr: 1.0e-04 loss: 6.457e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3362/30000 [7:33:50<56:19:50,  7.61s/it, lr: 1.0e-04 loss: 6.457e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3362/30000 [7:33:57<56:19:50,  7.61s/it, lr: 1.0e-04 loss: 5.568e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3362/30000 [7:33:57<56:19:50,  7.61s/it, lr: 1.0e-04 loss: 5.568e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3363/30000 [7:33:57<54:24:11,  7.35s/it, lr: 1.0e-04 loss: 5.568e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3363/30000 [7:33:57<54:24:11,  7.35s/it, lr: 1.0e-04 loss: 5.568e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3363/30000 [7:34:06<54:24:11,  7.35s/it, lr: 1.0e-04 loss: 6.276e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3363/30000 [7:34:06<54:24:11,  7.35s/it, lr: 1.0e-04 loss: 6.276e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3364/30000 [7:34:06<58:18:06,  7.88s/it, lr: 1.0e-04 loss: 6.276e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3364/30000 [7:34:06<58:18:06,  7.88s/it, lr: 1.0e-04 loss: 6.276e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3364/30000 [7:34:13<58:18:06,  7.88s/it, lr: 1.0e-04 loss: 6.003e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3364/30000 [7:34:13<58:18:06,  7.88s/it, lr: 1.0e-04 loss: 6.003e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3365/30000 [7:34:13<55:37:33,  7.52s/it, lr: 1.0e-04 loss: 6.003e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3365/30000 [7:34:13<55:37:33,  7.52s/it, lr: 1.0e-04 loss: 6.003e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3365/30000 [7:34:22<55:37:33,  7.52s/it, lr: 1.0e-04 loss: 1.318e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3365/30000 [7:34:22<55:37:33,  7.52s/it, lr: 1.0e-04 loss: 1.318e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3366/30000 [7:34:22<59:23:15,  8.03s/it, lr: 1.0e-04 loss: 1.318e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3366/30000 [7:34:22<59:23:15,  8.03s/it, lr: 1.0e-04 loss: 1.318e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3366/30000 [7:34:31<59:23:15,  8.03s/it, lr: 1.0e-04 loss: 7.016e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3366/30000 [7:34:31<59:23:15,  8.03s/it, lr: 1.0e-04 loss: 7.016e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3367/30000 [7:34:31<62:16:39,  8.42s/it, lr: 1.0e-04 loss: 7.016e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3367/30000 [7:34:31<62:16:39,  8.42s/it, lr: 1.0e-04 loss: 7.016e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3367/30000 [7:34:41<62:16:39,  8.42s/it, lr: 1.0e-04 loss: 4.411e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3367/30000 [7:34:41<62:16:39,  8.42s/it, lr: 1.0e-04 loss: 4.411e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3368/30000 [7:34:41<64:13:14,  8.68s/it, lr: 1.0e-04 loss: 4.411e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3368/30000 [7:34:41<64:13:14,  8.68s/it, lr: 1.0e-04 loss: 4.411e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3368/30000 [7:34:47<64:13:14,  8.68s/it, lr: 1.0e-04 loss: 5.694e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3368/30000 [7:34:47<64:13:14,  8.68s/it, lr: 1.0e-04 loss: 5.694e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3369/30000 [7:34:47<59:38:07,  8.06s/it, lr: 1.0e-04 loss: 5.694e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3369/30000 [7:34:47<59:38:07,  8.06s/it, lr: 1.0e-04 loss: 5.694e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3369/30000 [7:34:57<59:38:07,  8.06s/it, lr: 1.0e-04 loss: 7.348e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3369/30000 [7:34:57<59:38:07,  8.06s/it, lr: 1.0e-04 loss: 7.348e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3370/30000 [7:35:03<59:37:59,  8.06s/it, lr: 1.0e-04 loss: 6.750e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3370/30000 [7:35:03<59:37:59,  8.06s/it, lr: 1.0e-04 loss: 6.750e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3371/30000 [7:35:03<43:29:26,  5.88s/it, lr: 1.0e-04 loss: 6.750e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3371/30000 [7:35:03<43:29:26,  5.88s/it, lr: 1.0e-04 loss: 6.750e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3371/30000 [7:35:10<43:29:26,  5.88s/it, lr: 1.0e-04 loss: 8.413e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3371/30000 [7:35:10<43:29:26,  5.88s/it, lr: 1.0e-04 loss: 8.413e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3372/30000 [7:35:10<44:40:42,  6.04s/it, lr: 1.0e-04 loss: 8.413e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3372/30000 [7:35:10<44:40:42,  6.04s/it, lr: 1.0e-04 loss: 8.413e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3372/30000 [7:35:19<44:40:42,  6.04s/it, lr: 1.0e-04 loss: 1.390e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3372/30000 [7:35:19<44:40:42,  6.04s/it, lr: 1.0e-04 loss: 1.390e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3373/30000 [7:35:19<51:02:09,  6.90s/it, lr: 1.0e-04 loss: 1.390e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3373/30000 [7:35:19<51:02:09,  6.90s/it, lr: 1.0e-04 loss: 1.390e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3373/30000 [7:35:29<51:02:09,  6.90s/it, lr: 1.0e-04 loss: 6.175e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3373/30000 [7:35:29<51:02:09,  6.90s/it, lr: 1.0e-04 loss: 6.175e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3374/30000 [7:35:29<55:57:27,  7.57s/it, lr: 1.0e-04 loss: 6.175e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3374/30000 [7:35:29<55:57:27,  7.57s/it, lr: 1.0e-04 loss: 6.175e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3374/30000 [7:35:35<55:57:27,  7.57s/it, lr: 1.0e-04 loss: 5.773e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3374/30000 [7:35:35<55:57:27,  7.57s/it, lr: 1.0e-04 loss: 5.773e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3375/30000 [7:35:35<53:49:18,  7.28s/it, lr: 1.0e-04 loss: 5.773e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3375/30000 [7:35:35<53:49:18,  7.28s/it, lr: 1.0e-04 loss: 5.773e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3375/30000 [7:35:44<53:49:18,  7.28s/it, lr: 1.0e-04 loss: 8.230e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3375/30000 [7:35:44<53:49:18,  7.28s/it, lr: 1.0e-04 loss: 8.230e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3376/30000 [7:35:44<58:09:58,  7.87s/it, lr: 1.0e-04 loss: 8.230e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3376/30000 [7:35:44<58:09:58,  7.87s/it, lr: 1.0e-04 loss: 8.230e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3376/30000 [7:35:51<58:09:58,  7.87s/it, lr: 1.0e-04 loss: 8.858e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3376/30000 [7:35:51<58:09:58,  7.87s/it, lr: 1.0e-04 loss: 8.858e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3377/30000 [7:35:51<56:05:43,  7.59s/it, lr: 1.0e-04 loss: 8.858e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3377/30000 [7:35:51<56:05:43,  7.59s/it, lr: 1.0e-04 loss: 8.858e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3377/30000 [7:36:01<56:05:43,  7.59s/it, lr: 1.0e-04 loss: 5.873e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3377/30000 [7:36:01<56:05:43,  7.59s/it, lr: 1.0e-04 loss: 5.873e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3378/30000 [7:36:01<59:48:57,  8.09s/it, lr: 1.0e-04 loss: 5.873e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3378/30000 [7:36:01<59:48:57,  8.09s/it, lr: 1.0e-04 loss: 5.873e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3378/30000 [7:36:10<59:48:57,  8.09s/it, lr: 1.0e-04 loss: 3.397e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3378/30000 [7:36:10<59:48:57,  8.09s/it, lr: 1.0e-04 loss: 3.397e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3379/30000 [7:36:10<62:32:28,  8.46s/it, lr: 1.0e-04 loss: 3.397e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3379/30000 [7:36:10<62:32:28,  8.46s/it, lr: 1.0e-04 loss: 3.397e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3379/30000 [7:36:17<62:32:28,  8.46s/it, lr: 1.0e-04 loss: 6.791e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3379/30000 [7:36:17<62:32:28,  8.46s/it, lr: 1.0e-04 loss: 6.791e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3380/30000 [7:36:26<62:32:19,  8.46s/it, lr: 1.0e-04 loss: 4.576e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3380/30000 [7:36:26<62:32:19,  8.46s/it, lr: 1.0e-04 loss: 4.576e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3381/30000 [7:36:26<49:37:18,  6.71s/it, lr: 1.0e-04 loss: 4.576e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3381/30000 [7:36:26<49:37:18,  6.71s/it, lr: 1.0e-04 loss: 4.576e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3381/30000 [7:36:32<49:37:18,  6.71s/it, lr: 1.0e-04 loss: 7.721e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3381/30000 [7:36:32<49:37:18,  6.71s/it, lr: 1.0e-04 loss: 7.721e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3382/30000 [7:36:32<49:24:58,  6.68s/it, lr: 1.0e-04 loss: 7.721e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3382/30000 [7:36:32<49:24:58,  6.68s/it, lr: 1.0e-04 loss: 7.721e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3382/30000 [7:36:39<49:24:58,  6.68s/it, lr: 1.0e-04 loss: 6.323e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3382/30000 [7:36:39<49:24:58,  6.68s/it, lr: 1.0e-04 loss: 6.323e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3383/30000 [7:36:39<49:08:44,  6.65s/it, lr: 1.0e-04 loss: 6.323e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3383/30000 [7:36:39<49:08:44,  6.65s/it, lr: 1.0e-04 loss: 6.323e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3383/30000 [7:36:48<49:08:44,  6.65s/it, lr: 1.0e-04 loss: 6.240e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3383/30000 [7:36:48<49:08:44,  6.65s/it, lr: 1.0e-04 loss: 6.240e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3384/30000 [7:36:48<54:32:30,  7.38s/it, lr: 1.0e-04 loss: 6.240e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3384/30000 [7:36:48<54:32:30,  7.38s/it, lr: 1.0e-04 loss: 6.240e-02]
Xiang_wan_2_2_lora:  11%|####3                                  | 3384/30000 [7:36:58<54:32:30,  7.38s/it, lr: 1.0e-04 loss: 1.792e-01]
Xiang_wan_2_2_lora:  11%|####3                                  | 3384/30000 [7:36:58<54:32:30,  7.38s/it, lr: 1.0e-04 loss: 1.792e-01]
Xiang_wan_2_2_lora:  11%|####4                                  | 3385/30000 [7:36:58<58:58:04,  7.98s/it, lr: 1.0e-04 loss: 1.792e-01]
Xiang_wan_2_2_lora:  11%|####4                                  | 3385/30000 [7:36:58<58:58:04,  7.98s/it, lr: 1.0e-04 loss: 1.792e-01]
Xiang_wan_2_2_lora:  11%|####4                                  | 3385/30000 [7:37:07<58:58:04,  7.98s/it, lr: 1.0e-04 loss: 7.863e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3385/30000 [7:37:07<58:58:04,  7.98s/it, lr: 1.0e-04 loss: 7.863e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3386/30000 [7:37:07<61:56:14,  8.38s/it, lr: 1.0e-04 loss: 7.863e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3386/30000 [7:37:07<61:56:14,  8.38s/it, lr: 1.0e-04 loss: 7.863e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3386/30000 [7:37:17<61:56:14,  8.38s/it, lr: 1.0e-04 loss: 2.302e-01]
Xiang_wan_2_2_lora:  11%|####4                                  | 3386/30000 [7:37:17<61:56:14,  8.38s/it, lr: 1.0e-04 loss: 2.302e-01]
Xiang_wan_2_2_lora:  11%|####4                                  | 3387/30000 [7:37:17<63:58:15,  8.65s/it, lr: 1.0e-04 loss: 2.302e-01]
Xiang_wan_2_2_lora:  11%|####4                                  | 3387/30000 [7:37:17<63:58:15,  8.65s/it, lr: 1.0e-04 loss: 2.302e-01]
Xiang_wan_2_2_lora:  11%|####4                                  | 3387/30000 [7:37:23<63:58:15,  8.65s/it, lr: 1.0e-04 loss: 8.956e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3387/30000 [7:37:23<63:58:15,  8.65s/it, lr: 1.0e-04 loss: 8.956e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3388/30000 [7:37:23<59:22:53,  8.03s/it, lr: 1.0e-04 loss: 8.956e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3388/30000 [7:37:23<59:22:53,  8.03s/it, lr: 1.0e-04 loss: 8.956e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3388/30000 [7:37:30<59:22:53,  8.03s/it, lr: 1.0e-04 loss: 6.314e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3388/30000 [7:37:30<59:22:53,  8.03s/it, lr: 1.0e-04 loss: 6.314e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3389/30000 [7:37:30<56:12:48,  7.60s/it, lr: 1.0e-04 loss: 6.314e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3389/30000 [7:37:30<56:12:48,  7.60s/it, lr: 1.0e-04 loss: 6.314e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3389/30000 [7:37:36<56:12:48,  7.60s/it, lr: 1.0e-04 loss: 1.012e-01]
Xiang_wan_2_2_lora:  11%|####4                                  | 3389/30000 [7:37:36<56:12:48,  7.60s/it, lr: 1.0e-04 loss: 1.012e-01]
Xiang_wan_2_2_lora:  11%|####4                                  | 3390/30000 [7:37:43<56:12:40,  7.60s/it, lr: 1.0e-04 loss: 8.074e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3390/30000 [7:37:43<56:12:40,  7.60s/it, lr: 1.0e-04 loss: 8.074e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3391/30000 [7:37:43<41:44:45,  5.65s/it, lr: 1.0e-04 loss: 8.074e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3391/30000 [7:37:43<41:44:45,  5.65s/it, lr: 1.0e-04 loss: 8.074e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3391/30000 [7:37:50<41:44:45,  5.65s/it, lr: 1.0e-04 loss: 9.859e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3391/30000 [7:37:50<41:44:45,  5.65s/it, lr: 1.0e-04 loss: 9.859e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3392/30000 [7:37:50<43:54:36,  5.94s/it, lr: 1.0e-04 loss: 9.859e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3392/30000 [7:37:50<43:54:36,  5.94s/it, lr: 1.0e-04 loss: 9.859e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3392/30000 [7:37:56<43:54:36,  5.94s/it, lr: 1.0e-04 loss: 5.438e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3392/30000 [7:37:56<43:54:36,  5.94s/it, lr: 1.0e-04 loss: 5.438e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3393/30000 [7:37:56<45:06:56,  6.10s/it, lr: 1.0e-04 loss: 5.438e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3393/30000 [7:37:56<45:06:56,  6.10s/it, lr: 1.0e-04 loss: 5.438e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3393/30000 [7:38:03<45:06:56,  6.10s/it, lr: 1.0e-04 loss: 8.361e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3393/30000 [7:38:03<45:06:56,  6.10s/it, lr: 1.0e-04 loss: 8.361e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3394/30000 [7:38:03<45:59:55,  6.22s/it, lr: 1.0e-04 loss: 8.361e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3394/30000 [7:38:03<45:59:55,  6.22s/it, lr: 1.0e-04 loss: 8.361e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3394/30000 [7:38:12<45:59:55,  6.22s/it, lr: 1.0e-04 loss: 7.742e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3394/30000 [7:38:12<45:59:55,  6.22s/it, lr: 1.0e-04 loss: 7.742e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3395/30000 [7:38:12<52:23:02,  7.09s/it, lr: 1.0e-04 loss: 7.742e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3395/30000 [7:38:12<52:23:02,  7.09s/it, lr: 1.0e-04 loss: 7.742e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3395/30000 [7:38:19<52:23:02,  7.09s/it, lr: 1.0e-04 loss: 8.716e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3395/30000 [7:38:19<52:23:02,  7.09s/it, lr: 1.0e-04 loss: 8.716e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3396/30000 [7:38:19<51:16:17,  6.94s/it, lr: 1.0e-04 loss: 8.716e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3396/30000 [7:38:19<51:16:17,  6.94s/it, lr: 1.0e-04 loss: 8.716e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3396/30000 [7:38:25<51:16:17,  6.94s/it, lr: 1.0e-04 loss: 6.735e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3396/30000 [7:38:25<51:16:17,  6.94s/it, lr: 1.0e-04 loss: 6.735e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3397/30000 [7:38:25<50:32:44,  6.84s/it, lr: 1.0e-04 loss: 6.735e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3397/30000 [7:38:25<50:32:44,  6.84s/it, lr: 1.0e-04 loss: 6.735e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3397/30000 [7:38:35<50:32:44,  6.84s/it, lr: 1.0e-04 loss: 8.825e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3397/30000 [7:38:35<50:32:44,  6.84s/it, lr: 1.0e-04 loss: 8.825e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3398/30000 [7:38:35<55:49:10,  7.55s/it, lr: 1.0e-04 loss: 8.825e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3398/30000 [7:38:35<55:49:10,  7.55s/it, lr: 1.0e-04 loss: 8.825e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3398/30000 [7:38:41<55:49:10,  7.55s/it, lr: 1.0e-04 loss: 8.770e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3398/30000 [7:38:41<55:49:10,  7.55s/it, lr: 1.0e-04 loss: 8.770e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3399/30000 [7:38:41<53:45:13,  7.27s/it, lr: 1.0e-04 loss: 8.770e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3399/30000 [7:38:41<53:45:13,  7.27s/it, lr: 1.0e-04 loss: 8.770e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3399/30000 [7:38:51<53:45:13,  7.27s/it, lr: 1.0e-04 loss: 4.774e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3399/30000 [7:38:51<53:45:13,  7.27s/it, lr: 1.0e-04 loss: 4.774e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3400/30000 [7:39:00<53:45:05,  7.27s/it, lr: 1.0e-04 loss: 4.153e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3400/30000 [7:39:00<53:45:05,  7.27s/it, lr: 1.0e-04 loss: 4.153e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3401/30000 [7:39:00<44:56:04,  6.08s/it, lr: 1.0e-04 loss: 4.153e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3401/30000 [7:39:00<44:56:04,  6.08s/it, lr: 1.0e-04 loss: 4.153e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3401/30000 [7:39:07<44:56:04,  6.08s/it, lr: 1.0e-04 loss: 7.276e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3401/30000 [7:39:07<44:56:04,  6.08s/it, lr: 1.0e-04 loss: 7.276e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3402/30000 [7:39:07<45:51:44,  6.21s/it, lr: 1.0e-04 loss: 7.276e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3402/30000 [7:39:07<45:51:44,  6.21s/it, lr: 1.0e-04 loss: 7.276e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3402/30000 [7:39:16<45:51:44,  6.21s/it, lr: 1.0e-04 loss: 1.121e-01]
Xiang_wan_2_2_lora:  11%|####4                                  | 3402/30000 [7:39:16<45:51:44,  6.21s/it, lr: 1.0e-04 loss: 1.121e-01]
Xiang_wan_2_2_lora:  11%|####4                                  | 3403/30000 [7:39:16<51:48:13,  7.01s/it, lr: 1.0e-04 loss: 1.121e-01]
Xiang_wan_2_2_lora:  11%|####4                                  | 3403/30000 [7:39:16<51:48:13,  7.01s/it, lr: 1.0e-04 loss: 1.121e-01]
Xiang_wan_2_2_lora:  11%|####4                                  | 3403/30000 [7:39:25<51:48:13,  7.01s/it, lr: 1.0e-04 loss: 4.536e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3403/30000 [7:39:25<51:48:13,  7.01s/it, lr: 1.0e-04 loss: 4.536e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3404/30000 [7:39:25<56:26:48,  7.64s/it, lr: 1.0e-04 loss: 4.536e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3404/30000 [7:39:25<56:26:48,  7.64s/it, lr: 1.0e-04 loss: 4.536e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3404/30000 [7:39:32<56:26:48,  7.64s/it, lr: 1.0e-04 loss: 3.198e-01]
Xiang_wan_2_2_lora:  11%|####4                                  | 3404/30000 [7:39:32<56:26:48,  7.64s/it, lr: 1.0e-04 loss: 3.198e-01]
Xiang_wan_2_2_lora:  11%|####4                                  | 3405/30000 [7:39:32<54:17:08,  7.35s/it, lr: 1.0e-04 loss: 3.198e-01]
Xiang_wan_2_2_lora:  11%|####4                                  | 3405/30000 [7:39:32<54:17:08,  7.35s/it, lr: 1.0e-04 loss: 3.198e-01]
Xiang_wan_2_2_lora:  11%|####4                                  | 3405/30000 [7:39:41<54:17:08,  7.35s/it, lr: 1.0e-04 loss: 6.356e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3405/30000 [7:39:41<54:17:08,  7.35s/it, lr: 1.0e-04 loss: 6.356e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3406/30000 [7:39:41<58:26:19,  7.91s/it, lr: 1.0e-04 loss: 6.356e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3406/30000 [7:39:41<58:26:19,  7.91s/it, lr: 1.0e-04 loss: 6.356e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3406/30000 [7:39:48<58:26:19,  7.91s/it, lr: 1.0e-04 loss: 1.583e-01]
Xiang_wan_2_2_lora:  11%|####4                                  | 3406/30000 [7:39:48<58:26:19,  7.91s/it, lr: 1.0e-04 loss: 1.583e-01]
Xiang_wan_2_2_lora:  11%|####4                                  | 3407/30000 [7:39:48<56:07:01,  7.60s/it, lr: 1.0e-04 loss: 1.583e-01]
Xiang_wan_2_2_lora:  11%|####4                                  | 3407/30000 [7:39:48<56:07:01,  7.60s/it, lr: 1.0e-04 loss: 1.583e-01]
Xiang_wan_2_2_lora:  11%|####4                                  | 3407/30000 [7:39:55<56:07:01,  7.60s/it, lr: 1.0e-04 loss: 7.893e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3407/30000 [7:39:55<56:07:01,  7.60s/it, lr: 1.0e-04 loss: 7.893e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3408/30000 [7:39:55<53:46:25,  7.28s/it, lr: 1.0e-04 loss: 7.893e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3408/30000 [7:39:55<53:46:25,  7.28s/it, lr: 1.0e-04 loss: 7.893e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3408/30000 [7:40:01<53:46:25,  7.28s/it, lr: 1.0e-04 loss: 5.719e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3408/30000 [7:40:01<53:46:25,  7.28s/it, lr: 1.0e-04 loss: 5.719e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3409/30000 [7:40:01<52:16:51,  7.08s/it, lr: 1.0e-04 loss: 5.719e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3409/30000 [7:40:01<52:16:51,  7.08s/it, lr: 1.0e-04 loss: 5.719e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3409/30000 [7:40:08<52:16:51,  7.08s/it, lr: 1.0e-04 loss: 7.856e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3409/30000 [7:40:08<52:16:51,  7.08s/it, lr: 1.0e-04 loss: 7.856e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3410/30000 [7:40:17<52:16:44,  7.08s/it, lr: 1.0e-04 loss: 6.117e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3410/30000 [7:40:17<52:16:44,  7.08s/it, lr: 1.0e-04 loss: 6.117e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3411/30000 [7:40:17<44:06:15,  5.97s/it, lr: 1.0e-04 loss: 6.117e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3411/30000 [7:40:17<44:06:15,  5.97s/it, lr: 1.0e-04 loss: 6.117e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3411/30000 [7:40:24<44:06:15,  5.97s/it, lr: 1.0e-04 loss: 8.107e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3411/30000 [7:40:24<44:06:15,  5.97s/it, lr: 1.0e-04 loss: 8.107e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3412/30000 [7:40:24<45:12:09,  6.12s/it, lr: 1.0e-04 loss: 8.107e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3412/30000 [7:40:24<45:12:09,  6.12s/it, lr: 1.0e-04 loss: 8.107e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3412/30000 [7:40:33<45:12:09,  6.12s/it, lr: 1.0e-04 loss: 3.396e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3412/30000 [7:40:33<45:12:09,  6.12s/it, lr: 1.0e-04 loss: 3.396e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3413/30000 [7:40:33<51:24:23,  6.96s/it, lr: 1.0e-04 loss: 3.396e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3413/30000 [7:40:33<51:24:23,  6.96s/it, lr: 1.0e-04 loss: 3.396e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3413/30000 [7:40:39<51:24:23,  6.96s/it, lr: 1.0e-04 loss: 1.144e-01]
Xiang_wan_2_2_lora:  11%|####4                                  | 3413/30000 [7:40:39<51:24:23,  6.96s/it, lr: 1.0e-04 loss: 1.144e-01]
Xiang_wan_2_2_lora:  11%|####4                                  | 3414/30000 [7:40:39<50:32:21,  6.84s/it, lr: 1.0e-04 loss: 1.144e-01]
Xiang_wan_2_2_lora:  11%|####4                                  | 3414/30000 [7:40:39<50:32:21,  6.84s/it, lr: 1.0e-04 loss: 1.144e-01]
Xiang_wan_2_2_lora:  11%|####4                                  | 3414/30000 [7:40:46<50:32:21,  6.84s/it, lr: 1.0e-04 loss: 6.418e-01]
Xiang_wan_2_2_lora:  11%|####4                                  | 3414/30000 [7:40:46<50:32:21,  6.84s/it, lr: 1.0e-04 loss: 6.418e-01]
Xiang_wan_2_2_lora:  11%|####4                                  | 3415/30000 [7:40:46<50:27:51,  6.83s/it, lr: 1.0e-04 loss: 6.418e-01]
Xiang_wan_2_2_lora:  11%|####4                                  | 3415/30000 [7:40:46<50:27:51,  6.83s/it, lr: 1.0e-04 loss: 6.418e-01]
Xiang_wan_2_2_lora:  11%|####4                                  | 3415/30000 [7:40:55<50:27:51,  6.83s/it, lr: 1.0e-04 loss: 5.131e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3415/30000 [7:40:55<50:27:51,  6.83s/it, lr: 1.0e-04 loss: 5.131e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3416/30000 [7:40:55<55:36:18,  7.53s/it, lr: 1.0e-04 loss: 5.131e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3416/30000 [7:40:55<55:36:18,  7.53s/it, lr: 1.0e-04 loss: 5.131e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3416/30000 [7:41:05<55:36:18,  7.53s/it, lr: 1.0e-04 loss: 3.767e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3416/30000 [7:41:05<55:36:18,  7.53s/it, lr: 1.0e-04 loss: 3.767e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3417/30000 [7:41:05<59:25:32,  8.05s/it, lr: 1.0e-04 loss: 3.767e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3417/30000 [7:41:05<59:25:32,  8.05s/it, lr: 1.0e-04 loss: 3.767e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3417/30000 [7:41:14<59:25:32,  8.05s/it, lr: 1.0e-04 loss: 5.936e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3417/30000 [7:41:14<59:25:32,  8.05s/it, lr: 1.0e-04 loss: 5.936e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3418/30000 [7:41:14<62:17:55,  8.44s/it, lr: 1.0e-04 loss: 5.936e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3418/30000 [7:41:14<62:17:55,  8.44s/it, lr: 1.0e-04 loss: 5.936e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3418/30000 [7:41:21<62:17:55,  8.44s/it, lr: 1.0e-04 loss: 1.109e-01]
Xiang_wan_2_2_lora:  11%|####4                                  | 3418/30000 [7:41:21<62:17:55,  8.44s/it, lr: 1.0e-04 loss: 1.109e-01]
Xiang_wan_2_2_lora:  11%|####4                                  | 3419/30000 [7:41:21<58:13:47,  7.89s/it, lr: 1.0e-04 loss: 1.109e-01]
Xiang_wan_2_2_lora:  11%|####4                                  | 3419/30000 [7:41:21<58:13:47,  7.89s/it, lr: 1.0e-04 loss: 1.109e-01]
Xiang_wan_2_2_lora:  11%|####4                                  | 3419/30000 [7:41:30<58:13:47,  7.89s/it, lr: 1.0e-04 loss: 4.651e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3419/30000 [7:41:30<58:13:47,  7.89s/it, lr: 1.0e-04 loss: 4.651e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3420/30000 [7:41:39<58:13:39,  7.89s/it, lr: 1.0e-04 loss: 4.828e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3420/30000 [7:41:39<58:13:39,  7.89s/it, lr: 1.0e-04 loss: 4.828e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3421/30000 [7:41:39<47:20:21,  6.41s/it, lr: 1.0e-04 loss: 4.828e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3421/30000 [7:41:39<47:20:21,  6.41s/it, lr: 1.0e-04 loss: 4.828e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3421/30000 [7:41:46<47:20:21,  6.41s/it, lr: 1.0e-04 loss: 7.814e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3421/30000 [7:41:46<47:20:21,  6.41s/it, lr: 1.0e-04 loss: 7.814e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3422/30000 [7:41:46<48:05:12,  6.51s/it, lr: 1.0e-04 loss: 7.814e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3422/30000 [7:41:46<48:05:12,  6.51s/it, lr: 1.0e-04 loss: 7.814e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3422/30000 [7:41:53<48:05:12,  6.51s/it, lr: 1.0e-04 loss: 6.771e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3422/30000 [7:41:53<48:05:12,  6.51s/it, lr: 1.0e-04 loss: 6.771e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3423/30000 [7:41:53<48:16:26,  6.54s/it, lr: 1.0e-04 loss: 6.771e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3423/30000 [7:41:53<48:16:26,  6.54s/it, lr: 1.0e-04 loss: 6.771e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3423/30000 [7:41:59<48:16:26,  6.54s/it, lr: 1.0e-04 loss: 8.869e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3423/30000 [7:41:59<48:16:26,  6.54s/it, lr: 1.0e-04 loss: 8.869e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3424/30000 [7:41:59<48:19:23,  6.55s/it, lr: 1.0e-04 loss: 8.869e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3424/30000 [7:41:59<48:19:23,  6.55s/it, lr: 1.0e-04 loss: 8.869e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3424/30000 [7:42:06<48:19:23,  6.55s/it, lr: 1.0e-04 loss: 9.832e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3424/30000 [7:42:06<48:19:23,  6.55s/it, lr: 1.0e-04 loss: 9.832e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3425/30000 [7:42:06<48:21:55,  6.55s/it, lr: 1.0e-04 loss: 9.832e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3425/30000 [7:42:06<48:21:55,  6.55s/it, lr: 1.0e-04 loss: 9.832e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3425/30000 [7:42:15<48:21:55,  6.55s/it, lr: 1.0e-04 loss: 7.096e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3425/30000 [7:42:15<48:21:55,  6.55s/it, lr: 1.0e-04 loss: 7.096e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3426/30000 [7:42:15<54:03:32,  7.32s/it, lr: 1.0e-04 loss: 7.096e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3426/30000 [7:42:15<54:03:32,  7.32s/it, lr: 1.0e-04 loss: 7.096e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3426/30000 [7:42:25<54:03:32,  7.32s/it, lr: 1.0e-04 loss: 4.275e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3426/30000 [7:42:25<54:03:32,  7.32s/it, lr: 1.0e-04 loss: 4.275e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3427/30000 [7:42:25<58:22:28,  7.91s/it, lr: 1.0e-04 loss: 4.275e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3427/30000 [7:42:25<58:22:28,  7.91s/it, lr: 1.0e-04 loss: 4.275e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3427/30000 [7:42:34<58:22:28,  7.91s/it, lr: 1.0e-04 loss: 3.910e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3427/30000 [7:42:34<58:22:28,  7.91s/it, lr: 1.0e-04 loss: 3.910e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3428/30000 [7:42:34<61:36:26,  8.35s/it, lr: 1.0e-04 loss: 3.910e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3428/30000 [7:42:34<61:36:26,  8.35s/it, lr: 1.0e-04 loss: 3.910e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3428/30000 [7:42:41<61:36:26,  8.35s/it, lr: 1.0e-04 loss: 5.437e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3428/30000 [7:42:41<61:36:26,  8.35s/it, lr: 1.0e-04 loss: 5.437e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3429/30000 [7:42:41<57:45:56,  7.83s/it, lr: 1.0e-04 loss: 5.437e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3429/30000 [7:42:41<57:45:56,  7.83s/it, lr: 1.0e-04 loss: 5.437e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3429/30000 [7:42:47<57:45:56,  7.83s/it, lr: 1.0e-04 loss: 8.245e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3429/30000 [7:42:47<57:45:56,  7.83s/it, lr: 1.0e-04 loss: 8.245e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3430/30000 [7:42:54<57:45:48,  7.83s/it, lr: 1.0e-04 loss: 1.197e-01]
Xiang_wan_2_2_lora:  11%|####4                                  | 3430/30000 [7:42:54<57:45:48,  7.83s/it, lr: 1.0e-04 loss: 1.197e-01]
Xiang_wan_2_2_lora:  11%|####4                                  | 3431/30000 [7:42:54<42:26:41,  5.75s/it, lr: 1.0e-04 loss: 1.197e-01]
Xiang_wan_2_2_lora:  11%|####4                                  | 3431/30000 [7:42:54<42:26:41,  5.75s/it, lr: 1.0e-04 loss: 1.197e-01]
Xiang_wan_2_2_lora:  11%|####4                                  | 3431/30000 [7:43:00<42:26:41,  5.75s/it, lr: 1.0e-04 loss: 1.027e-01]
Xiang_wan_2_2_lora:  11%|####4                                  | 3431/30000 [7:43:00<42:26:41,  5.75s/it, lr: 1.0e-04 loss: 1.027e-01]
Xiang_wan_2_2_lora:  11%|####4                                  | 3432/30000 [7:43:00<43:56:00,  5.95s/it, lr: 1.0e-04 loss: 1.027e-01]
Xiang_wan_2_2_lora:  11%|####4                                  | 3432/30000 [7:43:00<43:56:00,  5.95s/it, lr: 1.0e-04 loss: 1.027e-01]
Xiang_wan_2_2_lora:  11%|####4                                  | 3432/30000 [7:43:10<43:56:00,  5.95s/it, lr: 1.0e-04 loss: 4.569e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3432/30000 [7:43:10<43:56:00,  5.95s/it, lr: 1.0e-04 loss: 4.569e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3433/30000 [7:43:10<50:20:39,  6.82s/it, lr: 1.0e-04 loss: 4.569e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3433/30000 [7:43:10<50:20:39,  6.82s/it, lr: 1.0e-04 loss: 4.569e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3433/30000 [7:43:16<50:20:39,  6.82s/it, lr: 1.0e-04 loss: 7.821e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3433/30000 [7:43:16<50:20:39,  6.82s/it, lr: 1.0e-04 loss: 7.821e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3434/30000 [7:43:16<49:56:20,  6.77s/it, lr: 1.0e-04 loss: 7.821e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3434/30000 [7:43:16<49:56:20,  6.77s/it, lr: 1.0e-04 loss: 7.821e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3434/30000 [7:43:26<49:56:20,  6.77s/it, lr: 1.0e-04 loss: 3.488e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3434/30000 [7:43:26<49:56:20,  6.77s/it, lr: 1.0e-04 loss: 3.488e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3435/30000 [7:43:26<55:13:57,  7.48s/it, lr: 1.0e-04 loss: 3.488e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3435/30000 [7:43:26<55:13:57,  7.48s/it, lr: 1.0e-04 loss: 3.488e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3435/30000 [7:43:32<55:13:57,  7.48s/it, lr: 1.0e-04 loss: 8.329e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3435/30000 [7:43:32<55:13:57,  7.48s/it, lr: 1.0e-04 loss: 8.329e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3436/30000 [7:43:32<53:27:30,  7.24s/it, lr: 1.0e-04 loss: 8.329e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3436/30000 [7:43:32<53:27:30,  7.24s/it, lr: 1.0e-04 loss: 8.329e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3436/30000 [7:43:42<53:27:30,  7.24s/it, lr: 1.0e-04 loss: 5.294e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3436/30000 [7:43:42<53:27:30,  7.24s/it, lr: 1.0e-04 loss: 5.294e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3437/30000 [7:43:42<58:11:05,  7.89s/it, lr: 1.0e-04 loss: 5.294e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3437/30000 [7:43:42<58:11:05,  7.89s/it, lr: 1.0e-04 loss: 5.294e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3437/30000 [7:43:48<58:11:05,  7.89s/it, lr: 1.0e-04 loss: 8.859e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3437/30000 [7:43:48<58:11:05,  7.89s/it, lr: 1.0e-04 loss: 8.859e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3438/30000 [7:43:48<55:21:06,  7.50s/it, lr: 1.0e-04 loss: 8.859e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3438/30000 [7:43:48<55:21:06,  7.50s/it, lr: 1.0e-04 loss: 8.859e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3438/30000 [7:43:58<55:21:06,  7.50s/it, lr: 1.0e-04 loss: 7.525e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3438/30000 [7:43:58<55:21:06,  7.50s/it, lr: 1.0e-04 loss: 7.525e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3439/30000 [7:43:58<59:11:47,  8.02s/it, lr: 1.0e-04 loss: 7.525e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3439/30000 [7:43:58<59:11:47,  8.02s/it, lr: 1.0e-04 loss: 7.525e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3439/30000 [7:44:07<59:11:47,  8.02s/it, lr: 1.0e-04 loss: 5.586e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3439/30000 [7:44:07<59:11:47,  8.02s/it, lr: 1.0e-04 loss: 5.586e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3440/30000 [7:44:16<59:11:39,  8.02s/it, lr: 1.0e-04 loss: 2.366e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3440/30000 [7:44:16<59:11:39,  8.02s/it, lr: 1.0e-04 loss: 2.366e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3441/30000 [7:44:16<47:56:52,  6.50s/it, lr: 1.0e-04 loss: 2.366e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3441/30000 [7:44:16<47:56:52,  6.50s/it, lr: 1.0e-04 loss: 2.366e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3441/30000 [7:44:23<47:56:52,  6.50s/it, lr: 1.0e-04 loss: 8.235e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3441/30000 [7:44:23<47:56:52,  6.50s/it, lr: 1.0e-04 loss: 8.235e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3442/30000 [7:44:23<48:04:41,  6.52s/it, lr: 1.0e-04 loss: 8.235e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3442/30000 [7:44:23<48:04:41,  6.52s/it, lr: 1.0e-04 loss: 8.235e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3442/30000 [7:44:29<48:04:41,  6.52s/it, lr: 1.0e-04 loss: 6.698e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3442/30000 [7:44:29<48:04:41,  6.52s/it, lr: 1.0e-04 loss: 6.698e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3443/30000 [7:44:29<48:03:56,  6.52s/it, lr: 1.0e-04 loss: 6.698e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3443/30000 [7:44:29<48:03:56,  6.52s/it, lr: 1.0e-04 loss: 6.698e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3443/30000 [7:44:39<48:03:56,  6.52s/it, lr: 1.0e-04 loss: 6.169e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3443/30000 [7:44:39<48:03:56,  6.52s/it, lr: 1.0e-04 loss: 6.169e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3444/30000 [7:44:39<53:39:12,  7.27s/it, lr: 1.0e-04 loss: 6.169e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3444/30000 [7:44:39<53:39:12,  7.27s/it, lr: 1.0e-04 loss: 6.169e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3444/30000 [7:44:46<53:39:12,  7.27s/it, lr: 1.0e-04 loss: 6.895e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3444/30000 [7:44:46<53:39:12,  7.27s/it, lr: 1.0e-04 loss: 6.895e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3445/30000 [7:44:46<52:39:39,  7.14s/it, lr: 1.0e-04 loss: 6.895e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3445/30000 [7:44:46<52:39:39,  7.14s/it, lr: 1.0e-04 loss: 6.895e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3445/30000 [7:44:55<52:39:39,  7.14s/it, lr: 1.0e-04 loss: 4.082e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3445/30000 [7:44:55<52:39:39,  7.14s/it, lr: 1.0e-04 loss: 4.082e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3446/30000 [7:44:55<57:08:56,  7.75s/it, lr: 1.0e-04 loss: 4.082e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3446/30000 [7:44:55<57:08:56,  7.75s/it, lr: 1.0e-04 loss: 4.082e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3446/30000 [7:45:04<57:08:56,  7.75s/it, lr: 1.0e-04 loss: 4.252e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3446/30000 [7:45:04<57:08:56,  7.75s/it, lr: 1.0e-04 loss: 4.252e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3447/30000 [7:45:04<60:37:49,  8.22s/it, lr: 1.0e-04 loss: 4.252e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3447/30000 [7:45:04<60:37:49,  8.22s/it, lr: 1.0e-04 loss: 4.252e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3447/30000 [7:45:11<60:37:49,  8.22s/it, lr: 1.0e-04 loss: 8.456e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3447/30000 [7:45:11<60:37:49,  8.22s/it, lr: 1.0e-04 loss: 8.456e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3448/30000 [7:45:11<57:03:34,  7.74s/it, lr: 1.0e-04 loss: 8.456e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3448/30000 [7:45:11<57:03:34,  7.74s/it, lr: 1.0e-04 loss: 8.456e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3448/30000 [7:45:17<57:03:34,  7.74s/it, lr: 1.0e-04 loss: 5.916e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3448/30000 [7:45:17<57:03:34,  7.74s/it, lr: 1.0e-04 loss: 5.916e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3449/30000 [7:45:17<54:32:19,  7.39s/it, lr: 1.0e-04 loss: 5.916e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3449/30000 [7:45:17<54:32:19,  7.39s/it, lr: 1.0e-04 loss: 5.916e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3449/30000 [7:45:27<54:32:19,  7.39s/it, lr: 1.0e-04 loss: 6.568e-02]
Xiang_wan_2_2_lora:  11%|####4                                  | 3449/30000 [7:45:27<54:32:19,  7.39s/it, lr: 1.0e-04 loss: 6.568e-02]
Xiang_wan_2_2_lora:  12%|####4                                  | 3450/30000 [7:45:33<54:32:11,  7.39s/it, lr: 1.0e-04 loss: 7.302e-02]
Xiang_wan_2_2_lora:  12%|####4                                  | 3450/30000 [7:45:33<54:32:11,  7.39s/it, lr: 1.0e-04 loss: 7.302e-02]
Xiang_wan_2_2_lora:  12%|####4                                  | 3451/30000 [7:45:33<40:44:57,  5.53s/it, lr: 1.0e-04 loss: 7.302e-02]
Xiang_wan_2_2_lora:  12%|####4                                  | 3451/30000 [7:45:33<40:44:57,  5.53s/it, lr: 1.0e-04 loss: 7.302e-02]
Xiang_wan_2_2_lora:  12%|####4                                  | 3451/30000 [7:45:43<40:44:57,  5.53s/it, lr: 1.0e-04 loss: 5.961e-02]
Xiang_wan_2_2_lora:  12%|####4                                  | 3451/30000 [7:45:43<40:44:57,  5.53s/it, lr: 1.0e-04 loss: 5.961e-02]
Xiang_wan_2_2_lora:  12%|####4                                  | 3452/30000 [7:45:43<47:57:35,  6.50s/it, lr: 1.0e-04 loss: 5.961e-02]
Xiang_wan_2_2_lora:  12%|####4                                  | 3452/30000 [7:45:43<47:57:35,  6.50s/it, lr: 1.0e-04 loss: 5.961e-02]
Xiang_wan_2_2_lora:  12%|####4                                  | 3452/30000 [7:45:52<47:57:35,  6.50s/it, lr: 1.0e-04 loss: 6.871e-02]
Xiang_wan_2_2_lora:  12%|####4                                  | 3452/30000 [7:45:52<47:57:35,  6.50s/it, lr: 1.0e-04 loss: 6.871e-02]
Xiang_wan_2_2_lora:  12%|####4                                  | 3453/30000 [7:45:52<53:18:53,  7.23s/it, lr: 1.0e-04 loss: 6.871e-02]
Xiang_wan_2_2_lora:  12%|####4                                  | 3453/30000 [7:45:52<53:18:53,  7.23s/it, lr: 1.0e-04 loss: 6.871e-02]
Xiang_wan_2_2_lora:  12%|####4                                  | 3453/30000 [7:46:01<53:18:53,  7.23s/it, lr: 1.0e-04 loss: 5.381e-02]
Xiang_wan_2_2_lora:  12%|####4                                  | 3453/30000 [7:46:01<53:18:53,  7.23s/it, lr: 1.0e-04 loss: 5.381e-02]
Xiang_wan_2_2_lora:  12%|####4                                  | 3454/30000 [7:46:01<57:30:23,  7.80s/it, lr: 1.0e-04 loss: 5.381e-02]
Xiang_wan_2_2_lora:  12%|####4                                  | 3454/30000 [7:46:01<57:30:23,  7.80s/it, lr: 1.0e-04 loss: 5.381e-02]
Xiang_wan_2_2_lora:  12%|####4                                  | 3454/30000 [7:46:11<57:30:23,  7.80s/it, lr: 1.0e-04 loss: 8.205e-02]
Xiang_wan_2_2_lora:  12%|####4                                  | 3454/30000 [7:46:11<57:30:23,  7.80s/it, lr: 1.0e-04 loss: 8.205e-02]
Xiang_wan_2_2_lora:  12%|####4                                  | 3455/30000 [7:46:11<60:44:43,  8.24s/it, lr: 1.0e-04 loss: 8.205e-02]
Xiang_wan_2_2_lora:  12%|####4                                  | 3455/30000 [7:46:11<60:44:43,  8.24s/it, lr: 1.0e-04 loss: 8.205e-02]
Xiang_wan_2_2_lora:  12%|####4                                  | 3455/30000 [7:46:20<60:44:43,  8.24s/it, lr: 1.0e-04 loss: 5.246e-02]
Xiang_wan_2_2_lora:  12%|####4                                  | 3455/30000 [7:46:20<60:44:43,  8.24s/it, lr: 1.0e-04 loss: 5.246e-02]
Xiang_wan_2_2_lora:  12%|####4                                  | 3456/30000 [7:46:20<63:05:53,  8.56s/it, lr: 1.0e-04 loss: 5.246e-02]
Xiang_wan_2_2_lora:  12%|####4                                  | 3456/30000 [7:46:20<63:05:53,  8.56s/it, lr: 1.0e-04 loss: 5.246e-02]
Xiang_wan_2_2_lora:  12%|####4                                  | 3456/30000 [7:46:27<63:05:53,  8.56s/it, lr: 1.0e-04 loss: 6.850e-02]
Xiang_wan_2_2_lora:  12%|####4                                  | 3456/30000 [7:46:27<63:05:53,  8.56s/it, lr: 1.0e-04 loss: 6.850e-02]
Xiang_wan_2_2_lora:  12%|####4                                  | 3457/30000 [7:46:27<59:01:14,  8.00s/it, lr: 1.0e-04 loss: 6.850e-02]
Xiang_wan_2_2_lora:  12%|####4                                  | 3457/30000 [7:46:27<59:01:14,  8.00s/it, lr: 1.0e-04 loss: 6.850e-02]
Xiang_wan_2_2_lora:  12%|####4                                  | 3457/30000 [7:46:33<59:01:14,  8.00s/it, lr: 1.0e-04 loss: 9.609e-02]
Xiang_wan_2_2_lora:  12%|####4                                  | 3457/30000 [7:46:33<59:01:14,  8.00s/it, lr: 1.0e-04 loss: 9.609e-02]
Xiang_wan_2_2_lora:  12%|####4                                  | 3458/30000 [7:46:33<55:53:05,  7.58s/it, lr: 1.0e-04 loss: 9.609e-02]
Xiang_wan_2_2_lora:  12%|####4                                  | 3458/30000 [7:46:33<55:53:05,  7.58s/it, lr: 1.0e-04 loss: 9.609e-02]
Xiang_wan_2_2_lora:  12%|####4                                  | 3458/30000 [7:46:43<55:53:05,  7.58s/it, lr: 1.0e-04 loss: 9.670e-02]
Xiang_wan_2_2_lora:  12%|####4                                  | 3458/30000 [7:46:43<55:53:05,  7.58s/it, lr: 1.0e-04 loss: 9.670e-02]
Xiang_wan_2_2_lora:  12%|####4                                  | 3459/30000 [7:46:43<59:41:40,  8.10s/it, lr: 1.0e-04 loss: 9.670e-02]
Xiang_wan_2_2_lora:  12%|####4                                  | 3459/30000 [7:46:43<59:41:40,  8.10s/it, lr: 1.0e-04 loss: 9.670e-02]
Xiang_wan_2_2_lora:  12%|####4                                  | 3459/30000 [7:46:52<59:41:40,  8.10s/it, lr: 1.0e-04 loss: 6.744e-02]
Xiang_wan_2_2_lora:  12%|####4                                  | 3459/30000 [7:46:52<59:41:40,  8.10s/it, lr: 1.0e-04 loss: 6.744e-02]
Xiang_wan_2_2_lora:  12%|####4                                  | 3460/30000 [7:46:59<59:41:32,  8.10s/it, lr: 1.0e-04 loss: 7.523e-02]
Xiang_wan_2_2_lora:  12%|####4                                  | 3460/30000 [7:46:59<59:41:32,  8.10s/it, lr: 1.0e-04 loss: 7.523e-02]
Xiang_wan_2_2_lora:  12%|####4                                  | 3461/30000 [7:46:59<43:28:56,  5.90s/it, lr: 1.0e-04 loss: 7.523e-02]
Xiang_wan_2_2_lora:  12%|####4                                  | 3461/30000 [7:46:59<43:28:56,  5.90s/it, lr: 1.0e-04 loss: 7.523e-02]
Xiang_wan_2_2_lora:  12%|####4                                  | 3461/30000 [7:47:05<43:28:56,  5.90s/it, lr: 1.0e-04 loss: 7.326e-02]
Xiang_wan_2_2_lora:  12%|####4                                  | 3461/30000 [7:47:05<43:28:56,  5.90s/it, lr: 1.0e-04 loss: 7.326e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3462/30000 [7:47:05<44:35:25,  6.05s/it, lr: 1.0e-04 loss: 7.326e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3462/30000 [7:47:05<44:35:25,  6.05s/it, lr: 1.0e-04 loss: 7.326e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3462/30000 [7:47:12<44:35:25,  6.05s/it, lr: 1.0e-04 loss: 1.012e+00]
Xiang_wan_2_2_lora:  12%|####5                                  | 3462/30000 [7:47:12<44:35:25,  6.05s/it, lr: 1.0e-04 loss: 1.012e+00]
Xiang_wan_2_2_lora:  12%|####5                                  | 3463/30000 [7:47:12<45:39:06,  6.19s/it, lr: 1.0e-04 loss: 1.012e+00]
Xiang_wan_2_2_lora:  12%|####5                                  | 3463/30000 [7:47:12<45:39:06,  6.19s/it, lr: 1.0e-04 loss: 1.012e+00]
Xiang_wan_2_2_lora:  12%|####5                                  | 3463/30000 [7:47:18<45:39:06,  6.19s/it, lr: 1.0e-04 loss: 8.477e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3463/30000 [7:47:18<45:39:06,  6.19s/it, lr: 1.0e-04 loss: 8.477e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3464/30000 [7:47:19<46:21:58,  6.29s/it, lr: 1.0e-04 loss: 8.477e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3464/30000 [7:47:19<46:21:58,  6.29s/it, lr: 1.0e-04 loss: 8.477e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3464/30000 [7:47:28<46:21:58,  6.29s/it, lr: 1.0e-04 loss: 5.221e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3464/30000 [7:47:28<46:21:58,  6.29s/it, lr: 1.0e-04 loss: 5.221e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3465/30000 [7:47:28<52:39:57,  7.15s/it, lr: 1.0e-04 loss: 5.221e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3465/30000 [7:47:28<52:39:57,  7.15s/it, lr: 1.0e-04 loss: 5.221e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3465/30000 [7:47:34<52:39:57,  7.15s/it, lr: 1.0e-04 loss: 1.141e-01]
Xiang_wan_2_2_lora:  12%|####5                                  | 3465/30000 [7:47:34<52:39:57,  7.15s/it, lr: 1.0e-04 loss: 1.141e-01]
Xiang_wan_2_2_lora:  12%|####5                                  | 3466/30000 [7:47:34<51:32:23,  6.99s/it, lr: 1.0e-04 loss: 1.141e-01]
Xiang_wan_2_2_lora:  12%|####5                                  | 3466/30000 [7:47:34<51:32:23,  6.99s/it, lr: 1.0e-04 loss: 1.141e-01]
Xiang_wan_2_2_lora:  12%|####5                                  | 3466/30000 [7:47:44<51:32:23,  6.99s/it, lr: 1.0e-04 loss: 3.830e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3466/30000 [7:47:44<51:32:23,  6.99s/it, lr: 1.0e-04 loss: 3.830e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3467/30000 [7:47:44<56:52:30,  7.72s/it, lr: 1.0e-04 loss: 3.830e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3467/30000 [7:47:44<56:52:30,  7.72s/it, lr: 1.0e-04 loss: 3.830e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3467/30000 [7:47:50<56:52:30,  7.72s/it, lr: 1.0e-04 loss: 7.117e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3467/30000 [7:47:50<56:52:30,  7.72s/it, lr: 1.0e-04 loss: 7.117e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3468/30000 [7:47:50<54:19:23,  7.37s/it, lr: 1.0e-04 loss: 7.117e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3468/30000 [7:47:50<54:19:23,  7.37s/it, lr: 1.0e-04 loss: 7.117e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3468/30000 [7:48:00<54:19:23,  7.37s/it, lr: 1.0e-04 loss: 1.845e-01]
Xiang_wan_2_2_lora:  12%|####5                                  | 3468/30000 [7:48:00<54:19:23,  7.37s/it, lr: 1.0e-04 loss: 1.845e-01]
Xiang_wan_2_2_lora:  12%|####5                                  | 3469/30000 [7:48:00<58:35:33,  7.95s/it, lr: 1.0e-04 loss: 1.845e-01]
Xiang_wan_2_2_lora:  12%|####5                                  | 3469/30000 [7:48:00<58:35:33,  7.95s/it, lr: 1.0e-04 loss: 1.845e-01]
Xiang_wan_2_2_lora:  12%|####5                                  | 3469/30000 [7:48:06<58:35:33,  7.95s/it, lr: 1.0e-04 loss: 6.948e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3469/30000 [7:48:06<58:35:33,  7.95s/it, lr: 1.0e-04 loss: 6.948e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3470/30000 [7:48:16<58:35:25,  7.95s/it, lr: 1.0e-04 loss: 6.360e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3470/30000 [7:48:16<58:35:25,  7.95s/it, lr: 1.0e-04 loss: 6.360e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3471/30000 [7:48:16<47:29:46,  6.45s/it, lr: 1.0e-04 loss: 6.360e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3471/30000 [7:48:16<47:29:46,  6.45s/it, lr: 1.0e-04 loss: 6.360e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3471/30000 [7:48:22<47:29:46,  6.45s/it, lr: 1.0e-04 loss: 8.789e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3471/30000 [7:48:22<47:29:46,  6.45s/it, lr: 1.0e-04 loss: 8.789e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3472/30000 [7:48:22<47:39:29,  6.47s/it, lr: 1.0e-04 loss: 8.789e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3472/30000 [7:48:22<47:39:29,  6.47s/it, lr: 1.0e-04 loss: 8.789e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3472/30000 [7:48:32<47:39:29,  6.47s/it, lr: 1.0e-04 loss: 5.256e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3472/30000 [7:48:32<47:39:29,  6.47s/it, lr: 1.0e-04 loss: 5.256e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3473/30000 [7:48:32<53:05:50,  7.21s/it, lr: 1.0e-04 loss: 5.256e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3473/30000 [7:48:32<53:05:50,  7.21s/it, lr: 1.0e-04 loss: 5.256e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3473/30000 [7:48:44<53:05:50,  7.21s/it, lr: 1.0e-04 loss: 7.641e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3473/30000 [7:48:44<53:05:50,  7.21s/it, lr: 1.0e-04 loss: 7.641e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3474/30000 [7:48:44<60:00:32,  8.14s/it, lr: 1.0e-04 loss: 7.641e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3474/30000 [7:48:44<60:00:32,  8.14s/it, lr: 1.0e-04 loss: 7.641e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3474/30000 [7:48:53<60:00:32,  8.14s/it, lr: 1.0e-04 loss: 6.059e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3474/30000 [7:48:53<60:00:32,  8.14s/it, lr: 1.0e-04 loss: 6.059e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3475/30000 [7:48:53<63:10:21,  8.57s/it, lr: 1.0e-04 loss: 6.059e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3475/30000 [7:48:53<63:10:21,  8.57s/it, lr: 1.0e-04 loss: 6.059e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3475/30000 [7:49:01<63:10:21,  8.57s/it, lr: 1.0e-04 loss: 9.320e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3475/30000 [7:49:01<63:10:21,  8.57s/it, lr: 1.0e-04 loss: 9.320e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3476/30000 [7:49:01<60:45:55,  8.25s/it, lr: 1.0e-04 loss: 9.320e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3476/30000 [7:49:01<60:45:55,  8.25s/it, lr: 1.0e-04 loss: 9.320e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3476/30000 [7:49:08<60:45:55,  8.25s/it, lr: 1.0e-04 loss: 6.821e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3476/30000 [7:49:08<60:45:55,  8.25s/it, lr: 1.0e-04 loss: 6.821e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3477/30000 [7:49:08<58:28:56,  7.94s/it, lr: 1.0e-04 loss: 6.821e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3477/30000 [7:49:08<58:28:56,  7.94s/it, lr: 1.0e-04 loss: 6.821e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3477/30000 [7:49:18<58:28:56,  7.94s/it, lr: 1.0e-04 loss: 4.541e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3477/30000 [7:49:18<58:28:56,  7.94s/it, lr: 1.0e-04 loss: 4.541e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3478/30000 [7:49:18<62:12:21,  8.44s/it, lr: 1.0e-04 loss: 4.541e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3478/30000 [7:49:18<62:12:21,  8.44s/it, lr: 1.0e-04 loss: 4.541e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3478/30000 [7:49:24<62:12:21,  8.44s/it, lr: 1.0e-04 loss: 4.936e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3478/30000 [7:49:24<62:12:21,  8.44s/it, lr: 1.0e-04 loss: 4.936e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3479/30000 [7:49:24<58:33:53,  7.95s/it, lr: 1.0e-04 loss: 4.936e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3479/30000 [7:49:24<58:33:53,  7.95s/it, lr: 1.0e-04 loss: 4.936e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3479/30000 [7:49:34<58:33:53,  7.95s/it, lr: 1.0e-04 loss: 5.658e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3479/30000 [7:49:34<58:33:53,  7.95s/it, lr: 1.0e-04 loss: 5.658e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3480/30000 [7:49:43<58:33:45,  7.95s/it, lr: 1.0e-04 loss: 4.076e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3480/30000 [7:49:43<58:33:45,  7.95s/it, lr: 1.0e-04 loss: 4.076e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3481/30000 [7:49:43<47:30:42,  6.45s/it, lr: 1.0e-04 loss: 4.076e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3481/30000 [7:49:43<47:30:42,  6.45s/it, lr: 1.0e-04 loss: 4.076e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3481/30000 [7:49:50<47:30:42,  6.45s/it, lr: 1.0e-04 loss: 6.260e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3481/30000 [7:49:50<47:30:42,  6.45s/it, lr: 1.0e-04 loss: 6.260e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3482/30000 [7:49:50<48:13:14,  6.55s/it, lr: 1.0e-04 loss: 6.260e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3482/30000 [7:49:50<48:13:14,  6.55s/it, lr: 1.0e-04 loss: 6.260e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3482/30000 [7:49:56<48:13:14,  6.55s/it, lr: 1.0e-04 loss: 8.528e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3482/30000 [7:49:56<48:13:14,  6.55s/it, lr: 1.0e-04 loss: 8.528e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3483/30000 [7:49:56<48:19:58,  6.56s/it, lr: 1.0e-04 loss: 8.528e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3483/30000 [7:49:56<48:19:58,  6.56s/it, lr: 1.0e-04 loss: 8.528e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3483/30000 [7:50:03<48:19:58,  6.56s/it, lr: 1.0e-04 loss: 8.305e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3483/30000 [7:50:03<48:19:58,  6.56s/it, lr: 1.0e-04 loss: 8.305e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3484/30000 [7:50:03<48:14:46,  6.55s/it, lr: 1.0e-04 loss: 8.305e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3484/30000 [7:50:03<48:14:46,  6.55s/it, lr: 1.0e-04 loss: 8.305e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3484/30000 [7:50:10<48:14:46,  6.55s/it, lr: 1.0e-04 loss: 2.861e-01]
Xiang_wan_2_2_lora:  12%|####5                                  | 3484/30000 [7:50:10<48:14:46,  6.55s/it, lr: 1.0e-04 loss: 2.861e-01]
Xiang_wan_2_2_lora:  12%|####5                                  | 3485/30000 [7:50:10<48:18:35,  6.56s/it, lr: 1.0e-04 loss: 2.861e-01]
Xiang_wan_2_2_lora:  12%|####5                                  | 3485/30000 [7:50:10<48:18:35,  6.56s/it, lr: 1.0e-04 loss: 2.861e-01]
Xiang_wan_2_2_lora:  12%|####5                                  | 3485/30000 [7:50:16<48:18:35,  6.56s/it, lr: 1.0e-04 loss: 8.170e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3485/30000 [7:50:16<48:18:35,  6.56s/it, lr: 1.0e-04 loss: 8.170e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3486/30000 [7:50:16<48:25:49,  6.58s/it, lr: 1.0e-04 loss: 8.170e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3486/30000 [7:50:16<48:25:49,  6.58s/it, lr: 1.0e-04 loss: 8.170e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3486/30000 [7:50:25<48:25:49,  6.58s/it, lr: 1.0e-04 loss: 5.478e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3486/30000 [7:50:25<48:25:49,  6.58s/it, lr: 1.0e-04 loss: 5.478e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3487/30000 [7:50:25<54:07:35,  7.35s/it, lr: 1.0e-04 loss: 5.478e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3487/30000 [7:50:25<54:07:35,  7.35s/it, lr: 1.0e-04 loss: 5.478e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3487/30000 [7:50:32<54:07:35,  7.35s/it, lr: 1.0e-04 loss: 8.424e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3487/30000 [7:50:32<54:07:35,  7.35s/it, lr: 1.0e-04 loss: 8.424e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3488/30000 [7:50:32<52:37:18,  7.15s/it, lr: 1.0e-04 loss: 8.424e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3488/30000 [7:50:32<52:37:18,  7.15s/it, lr: 1.0e-04 loss: 8.424e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3488/30000 [7:50:41<52:37:18,  7.15s/it, lr: 1.0e-04 loss: 2.473e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3488/30000 [7:50:41<52:37:18,  7.15s/it, lr: 1.0e-04 loss: 2.473e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3489/30000 [7:50:41<57:05:42,  7.75s/it, lr: 1.0e-04 loss: 2.473e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3489/30000 [7:50:41<57:05:42,  7.75s/it, lr: 1.0e-04 loss: 2.473e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3489/30000 [7:50:51<57:05:42,  7.75s/it, lr: 1.0e-04 loss: 5.292e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3489/30000 [7:50:51<57:05:42,  7.75s/it, lr: 1.0e-04 loss: 5.292e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3490/30000 [7:50:57<57:05:35,  7.75s/it, lr: 1.0e-04 loss: 6.240e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3490/30000 [7:50:57<57:05:35,  7.75s/it, lr: 1.0e-04 loss: 6.240e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3491/30000 [7:50:57<42:00:18,  5.70s/it, lr: 1.0e-04 loss: 6.240e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3491/30000 [7:50:57<42:00:18,  5.70s/it, lr: 1.0e-04 loss: 6.240e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3491/30000 [7:51:07<42:00:18,  5.70s/it, lr: 1.0e-04 loss: 5.369e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3491/30000 [7:51:07<42:00:18,  5.70s/it, lr: 1.0e-04 loss: 5.369e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3492/30000 [7:51:07<48:36:35,  6.60s/it, lr: 1.0e-04 loss: 5.369e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3492/30000 [7:51:07<48:36:35,  6.60s/it, lr: 1.0e-04 loss: 5.369e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3492/30000 [7:51:13<48:36:35,  6.60s/it, lr: 1.0e-04 loss: 4.594e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3492/30000 [7:51:13<48:36:35,  6.60s/it, lr: 1.0e-04 loss: 4.594e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3493/30000 [7:51:13<48:33:25,  6.59s/it, lr: 1.0e-04 loss: 4.594e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3493/30000 [7:51:13<48:33:25,  6.59s/it, lr: 1.0e-04 loss: 4.594e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3493/30000 [7:51:20<48:33:25,  6.59s/it, lr: 1.0e-04 loss: 8.240e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3493/30000 [7:51:20<48:33:25,  6.59s/it, lr: 1.0e-04 loss: 8.240e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3494/30000 [7:51:20<48:27:15,  6.58s/it, lr: 1.0e-04 loss: 8.240e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3494/30000 [7:51:20<48:27:15,  6.58s/it, lr: 1.0e-04 loss: 8.240e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3494/30000 [7:51:26<48:27:15,  6.58s/it, lr: 1.0e-04 loss: 1.022e-01]
Xiang_wan_2_2_lora:  12%|####5                                  | 3494/30000 [7:51:26<48:27:15,  6.58s/it, lr: 1.0e-04 loss: 1.022e-01]
Xiang_wan_2_2_lora:  12%|####5                                  | 3495/30000 [7:51:26<48:28:45,  6.58s/it, lr: 1.0e-04 loss: 1.022e-01]
Xiang_wan_2_2_lora:  12%|####5                                  | 3495/30000 [7:51:26<48:28:45,  6.58s/it, lr: 1.0e-04 loss: 1.022e-01]
Xiang_wan_2_2_lora:  12%|####5                                  | 3495/30000 [7:51:36<48:28:45,  6.58s/it, lr: 1.0e-04 loss: 6.983e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3495/30000 [7:51:36<48:28:45,  6.58s/it, lr: 1.0e-04 loss: 6.983e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3496/30000 [7:51:36<54:23:55,  7.39s/it, lr: 1.0e-04 loss: 6.983e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3496/30000 [7:51:36<54:23:55,  7.39s/it, lr: 1.0e-04 loss: 6.983e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3496/30000 [7:51:45<54:23:55,  7.39s/it, lr: 1.0e-04 loss: 4.006e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3496/30000 [7:51:45<54:23:55,  7.39s/it, lr: 1.0e-04 loss: 4.006e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3497/30000 [7:51:45<58:56:40,  8.01s/it, lr: 1.0e-04 loss: 4.006e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3497/30000 [7:51:45<58:56:40,  8.01s/it, lr: 1.0e-04 loss: 4.006e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3497/30000 [7:51:52<58:56:40,  8.01s/it, lr: 1.0e-04 loss: 9.191e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3497/30000 [7:51:52<58:56:40,  8.01s/it, lr: 1.0e-04 loss: 9.191e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3498/30000 [7:51:52<55:53:11,  7.59s/it, lr: 1.0e-04 loss: 9.191e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3498/30000 [7:51:52<55:53:11,  7.59s/it, lr: 1.0e-04 loss: 9.191e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3498/30000 [7:52:01<55:53:11,  7.59s/it, lr: 1.0e-04 loss: 7.478e-01]
Xiang_wan_2_2_lora:  12%|####5                                  | 3498/30000 [7:52:01<55:53:11,  7.59s/it, lr: 1.0e-04 loss: 7.478e-01]
Xiang_wan_2_2_lora:  12%|####5                                  | 3499/30000 [7:52:01<59:29:22,  8.08s/it, lr: 1.0e-04 loss: 7.478e-01]
Xiang_wan_2_2_lora:  12%|####5                                  | 3499/30000 [7:52:01<59:29:22,  8.08s/it, lr: 1.0e-04 loss: 7.478e-01]
Xiang_wan_2_2_lora:  12%|####5                                  | 3499/30000 [7:52:08<59:29:22,  8.08s/it, lr: 1.0e-04 loss: 1.433e-01]
Xiang_wan_2_2_lora:  12%|####5                                  | 3499/30000 [7:52:08<59:29:22,  8.08s/it, lr: 1.0e-04 loss: 1.433e-01]
+Saving at step 3500
+Saved checkpoint to /home/featurize/ai-toolkit/output/Xiang_wan_2_2_lora/Xiang_wan_2_2_lora_000003500.safetensors
+Saved optimizer to /home/featurize/ai-toolkit/output/Xiang_wan_2_2_lora/optimizer.pt
+
Xiang_wan_2_2_lora:  12%|####5                                  | 3500/30000 [7:52:15<59:29:13,  8.08s/it, lr: 1.0e-04 loss: 1.093e-01]
Xiang_wan_2_2_lora:  12%|####5                                  | 3500/30000 [7:52:15<59:29:13,  8.08s/it, lr: 1.0e-04 loss: 1.093e-01]
Xiang_wan_2_2_lora:  12%|####5                                  | 3501/30000 [7:52:15<43:50:40,  5.96s/it, lr: 1.0e-04 loss: 1.093e-01]
Xiang_wan_2_2_lora:  12%|####5                                  | 3501/30000 [7:52:15<43:50:40,  5.96s/it, lr: 1.0e-04 loss: 1.093e-01]
Xiang_wan_2_2_lora:  12%|####5                                  | 3501/30000 [7:52:24<43:50:40,  5.96s/it, lr: 1.0e-04 loss: 6.944e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3501/30000 [7:52:24<43:50:40,  5.96s/it, lr: 1.0e-04 loss: 6.944e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3502/30000 [7:52:24<50:12:23,  6.82s/it, lr: 1.0e-04 loss: 6.944e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3502/30000 [7:52:24<50:12:23,  6.82s/it, lr: 1.0e-04 loss: 6.944e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3502/30000 [7:52:31<50:12:23,  6.82s/it, lr: 1.0e-04 loss: 6.803e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3502/30000 [7:52:31<50:12:23,  6.82s/it, lr: 1.0e-04 loss: 6.803e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3503/30000 [7:52:31<49:42:42,  6.75s/it, lr: 1.0e-04 loss: 6.803e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3503/30000 [7:52:31<49:42:42,  6.75s/it, lr: 1.0e-04 loss: 6.803e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3503/30000 [7:52:40<49:42:42,  6.75s/it, lr: 1.0e-04 loss: 6.793e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3503/30000 [7:52:40<49:42:42,  6.75s/it, lr: 1.0e-04 loss: 6.793e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3504/30000 [7:52:40<54:48:11,  7.45s/it, lr: 1.0e-04 loss: 6.793e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3504/30000 [7:52:40<54:48:11,  7.45s/it, lr: 1.0e-04 loss: 6.793e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3504/30000 [7:52:49<54:48:11,  7.45s/it, lr: 1.0e-04 loss: 5.631e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3504/30000 [7:52:49<54:48:11,  7.45s/it, lr: 1.0e-04 loss: 5.631e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3505/30000 [7:52:49<58:39:01,  7.97s/it, lr: 1.0e-04 loss: 5.631e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3505/30000 [7:52:49<58:39:01,  7.97s/it, lr: 1.0e-04 loss: 5.631e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3505/30000 [7:52:56<58:39:01,  7.97s/it, lr: 1.0e-04 loss: 7.975e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3505/30000 [7:52:56<58:39:01,  7.97s/it, lr: 1.0e-04 loss: 7.975e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3506/30000 [7:52:56<55:44:13,  7.57s/it, lr: 1.0e-04 loss: 7.975e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3506/30000 [7:52:56<55:44:13,  7.57s/it, lr: 1.0e-04 loss: 7.975e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3506/30000 [7:53:03<55:44:13,  7.57s/it, lr: 1.0e-04 loss: 6.681e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3506/30000 [7:53:03<55:44:13,  7.57s/it, lr: 1.0e-04 loss: 6.681e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3507/30000 [7:53:03<53:35:51,  7.28s/it, lr: 1.0e-04 loss: 6.681e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3507/30000 [7:53:03<53:35:51,  7.28s/it, lr: 1.0e-04 loss: 6.681e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3507/30000 [7:53:09<53:35:51,  7.28s/it, lr: 1.0e-04 loss: 8.011e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3507/30000 [7:53:09<53:35:51,  7.28s/it, lr: 1.0e-04 loss: 8.011e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3508/30000 [7:53:09<52:48:25,  7.18s/it, lr: 1.0e-04 loss: 8.011e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3508/30000 [7:53:09<52:48:25,  7.18s/it, lr: 1.0e-04 loss: 8.011e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3508/30000 [7:53:16<52:48:25,  7.18s/it, lr: 1.0e-04 loss: 2.818e-01]
Xiang_wan_2_2_lora:  12%|####5                                  | 3508/30000 [7:53:16<52:48:25,  7.18s/it, lr: 1.0e-04 loss: 2.818e-01]
Xiang_wan_2_2_lora:  12%|####5                                  | 3509/30000 [7:53:16<51:35:08,  7.01s/it, lr: 1.0e-04 loss: 2.818e-01]
Xiang_wan_2_2_lora:  12%|####5                                  | 3509/30000 [7:53:16<51:35:08,  7.01s/it, lr: 1.0e-04 loss: 2.818e-01]
Xiang_wan_2_2_lora:  12%|####5                                  | 3509/30000 [7:53:23<51:35:08,  7.01s/it, lr: 1.0e-04 loss: 1.072e-01]
Xiang_wan_2_2_lora:  12%|####5                                  | 3509/30000 [7:53:23<51:35:08,  7.01s/it, lr: 1.0e-04 loss: 1.072e-01]
Xiang_wan_2_2_lora:  12%|####5                                  | 3510/30000 [7:53:32<51:35:01,  7.01s/it, lr: 1.0e-04 loss: 2.572e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3510/30000 [7:53:32<51:35:01,  7.01s/it, lr: 1.0e-04 loss: 2.572e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3511/30000 [7:53:32<43:36:33,  5.93s/it, lr: 1.0e-04 loss: 2.572e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3511/30000 [7:53:32<43:36:33,  5.93s/it, lr: 1.0e-04 loss: 2.572e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3511/30000 [7:53:39<43:36:33,  5.93s/it, lr: 1.0e-04 loss: 9.637e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3511/30000 [7:53:39<43:36:33,  5.93s/it, lr: 1.0e-04 loss: 9.637e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3512/30000 [7:53:39<45:05:07,  6.13s/it, lr: 1.0e-04 loss: 9.637e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3512/30000 [7:53:39<45:05:07,  6.13s/it, lr: 1.0e-04 loss: 9.637e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3512/30000 [7:53:48<45:05:07,  6.13s/it, lr: 1.0e-04 loss: 5.194e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3512/30000 [7:53:48<45:05:07,  6.13s/it, lr: 1.0e-04 loss: 5.194e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3513/30000 [7:53:48<51:06:43,  6.95s/it, lr: 1.0e-04 loss: 5.194e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3513/30000 [7:53:48<51:06:43,  6.95s/it, lr: 1.0e-04 loss: 5.194e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3513/30000 [7:53:55<51:06:43,  6.95s/it, lr: 1.0e-04 loss: 6.865e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3513/30000 [7:53:55<51:06:43,  6.95s/it, lr: 1.0e-04 loss: 6.865e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3514/30000 [7:53:55<50:26:54,  6.86s/it, lr: 1.0e-04 loss: 6.865e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3514/30000 [7:53:55<50:26:54,  6.86s/it, lr: 1.0e-04 loss: 6.865e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3514/30000 [7:54:04<50:26:54,  6.86s/it, lr: 1.0e-04 loss: 4.391e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3514/30000 [7:54:04<50:26:54,  6.86s/it, lr: 1.0e-04 loss: 4.391e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3515/30000 [7:54:04<55:48:02,  7.58s/it, lr: 1.0e-04 loss: 4.391e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3515/30000 [7:54:04<55:48:02,  7.58s/it, lr: 1.0e-04 loss: 4.391e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3515/30000 [7:54:11<55:48:02,  7.58s/it, lr: 1.0e-04 loss: 9.641e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3515/30000 [7:54:11<55:48:02,  7.58s/it, lr: 1.0e-04 loss: 9.641e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3516/30000 [7:54:11<53:44:17,  7.30s/it, lr: 1.0e-04 loss: 9.641e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3516/30000 [7:54:11<53:44:17,  7.30s/it, lr: 1.0e-04 loss: 9.641e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3516/30000 [7:54:17<53:44:17,  7.30s/it, lr: 1.0e-04 loss: 7.512e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3516/30000 [7:54:17<53:44:17,  7.30s/it, lr: 1.0e-04 loss: 7.512e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3517/30000 [7:54:17<52:02:56,  7.08s/it, lr: 1.0e-04 loss: 7.512e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3517/30000 [7:54:17<52:02:56,  7.08s/it, lr: 1.0e-04 loss: 7.512e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3517/30000 [7:54:26<52:02:56,  7.08s/it, lr: 1.0e-04 loss: 7.735e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3517/30000 [7:54:26<52:02:56,  7.08s/it, lr: 1.0e-04 loss: 7.735e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3518/30000 [7:54:26<56:53:07,  7.73s/it, lr: 1.0e-04 loss: 7.735e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3518/30000 [7:54:26<56:53:07,  7.73s/it, lr: 1.0e-04 loss: 7.735e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3518/30000 [7:54:36<56:53:07,  7.73s/it, lr: 1.0e-04 loss: 3.936e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3518/30000 [7:54:36<56:53:07,  7.73s/it, lr: 1.0e-04 loss: 3.936e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3519/30000 [7:54:36<60:10:21,  8.18s/it, lr: 1.0e-04 loss: 3.936e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3519/30000 [7:54:36<60:10:21,  8.18s/it, lr: 1.0e-04 loss: 3.936e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3519/30000 [7:54:42<60:10:21,  8.18s/it, lr: 1.0e-04 loss: 8.532e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3519/30000 [7:54:42<60:10:21,  8.18s/it, lr: 1.0e-04 loss: 8.532e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3520/30000 [7:54:52<60:10:13,  8.18s/it, lr: 1.0e-04 loss: 4.771e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3520/30000 [7:54:52<60:10:13,  8.18s/it, lr: 1.0e-04 loss: 4.771e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3521/30000 [7:54:52<48:11:31,  6.55s/it, lr: 1.0e-04 loss: 4.771e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3521/30000 [7:54:52<48:11:31,  6.55s/it, lr: 1.0e-04 loss: 4.771e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3521/30000 [7:54:58<48:11:31,  6.55s/it, lr: 1.0e-04 loss: 1.000e-01]
Xiang_wan_2_2_lora:  12%|####5                                  | 3521/30000 [7:54:58<48:11:31,  6.55s/it, lr: 1.0e-04 loss: 1.000e-01]
Xiang_wan_2_2_lora:  12%|####5                                  | 3522/30000 [7:54:58<48:17:20,  6.57s/it, lr: 1.0e-04 loss: 1.000e-01]
Xiang_wan_2_2_lora:  12%|####5                                  | 3522/30000 [7:54:58<48:17:20,  6.57s/it, lr: 1.0e-04 loss: 1.000e-01]
Xiang_wan_2_2_lora:  12%|####5                                  | 3522/30000 [7:55:05<48:17:20,  6.57s/it, lr: 1.0e-04 loss: 5.779e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3522/30000 [7:55:05<48:17:20,  6.57s/it, lr: 1.0e-04 loss: 5.779e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3523/30000 [7:55:05<48:41:51,  6.62s/it, lr: 1.0e-04 loss: 5.779e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3523/30000 [7:55:05<48:41:51,  6.62s/it, lr: 1.0e-04 loss: 5.779e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3523/30000 [7:55:14<48:41:51,  6.62s/it, lr: 1.0e-04 loss: 7.724e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3523/30000 [7:55:14<48:41:51,  6.62s/it, lr: 1.0e-04 loss: 7.724e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3524/30000 [7:55:14<54:02:47,  7.35s/it, lr: 1.0e-04 loss: 7.724e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3524/30000 [7:55:14<54:02:47,  7.35s/it, lr: 1.0e-04 loss: 7.724e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3524/30000 [7:55:24<54:02:47,  7.35s/it, lr: 1.0e-04 loss: 4.044e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3524/30000 [7:55:24<54:02:47,  7.35s/it, lr: 1.0e-04 loss: 4.044e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3525/30000 [7:55:24<58:02:10,  7.89s/it, lr: 1.0e-04 loss: 4.044e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3525/30000 [7:55:24<58:02:10,  7.89s/it, lr: 1.0e-04 loss: 4.044e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3525/30000 [7:55:33<58:02:10,  7.89s/it, lr: 1.0e-04 loss: 5.492e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3525/30000 [7:55:33<58:02:10,  7.89s/it, lr: 1.0e-04 loss: 5.492e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3526/30000 [7:55:33<61:10:30,  8.32s/it, lr: 1.0e-04 loss: 5.492e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3526/30000 [7:55:33<61:10:30,  8.32s/it, lr: 1.0e-04 loss: 5.492e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3526/30000 [7:55:42<61:10:30,  8.32s/it, lr: 1.0e-04 loss: 4.936e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3526/30000 [7:55:42<61:10:30,  8.32s/it, lr: 1.0e-04 loss: 4.936e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3527/30000 [7:55:42<63:15:12,  8.60s/it, lr: 1.0e-04 loss: 4.936e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3527/30000 [7:55:42<63:15:12,  8.60s/it, lr: 1.0e-04 loss: 4.936e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3527/30000 [7:55:52<63:15:12,  8.60s/it, lr: 1.0e-04 loss: 5.663e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3527/30000 [7:55:52<63:15:12,  8.60s/it, lr: 1.0e-04 loss: 5.663e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3528/30000 [7:55:52<64:48:55,  8.81s/it, lr: 1.0e-04 loss: 5.663e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3528/30000 [7:55:52<64:48:55,  8.81s/it, lr: 1.0e-04 loss: 5.663e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3528/30000 [7:55:58<64:48:55,  8.81s/it, lr: 1.0e-04 loss: 2.608e-01]
Xiang_wan_2_2_lora:  12%|####5                                  | 3528/30000 [7:55:58<64:48:55,  8.81s/it, lr: 1.0e-04 loss: 2.608e-01]
Xiang_wan_2_2_lora:  12%|####5                                  | 3529/30000 [7:55:58<59:54:37,  8.15s/it, lr: 1.0e-04 loss: 2.608e-01]
Xiang_wan_2_2_lora:  12%|####5                                  | 3529/30000 [7:55:58<59:54:37,  8.15s/it, lr: 1.0e-04 loss: 2.608e-01]
Xiang_wan_2_2_lora:  12%|####5                                  | 3529/30000 [7:56:05<59:54:37,  8.15s/it, lr: 1.0e-04 loss: 7.204e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3529/30000 [7:56:05<59:54:37,  8.15s/it, lr: 1.0e-04 loss: 7.204e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3530/30000 [7:56:14<59:54:29,  8.15s/it, lr: 1.0e-04 loss: 3.917e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3530/30000 [7:56:14<59:54:29,  8.15s/it, lr: 1.0e-04 loss: 3.917e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3531/30000 [7:56:14<48:14:23,  6.56s/it, lr: 1.0e-04 loss: 3.917e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3531/30000 [7:56:14<48:14:23,  6.56s/it, lr: 1.0e-04 loss: 3.917e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3531/30000 [7:56:24<48:14:23,  6.56s/it, lr: 1.0e-04 loss: 7.545e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3531/30000 [7:56:24<48:14:23,  6.56s/it, lr: 1.0e-04 loss: 7.545e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3532/30000 [7:56:24<53:18:51,  7.25s/it, lr: 1.0e-04 loss: 7.545e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3532/30000 [7:56:24<53:18:51,  7.25s/it, lr: 1.0e-04 loss: 7.545e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3532/30000 [7:56:34<53:18:51,  7.25s/it, lr: 1.0e-04 loss: 3.945e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3532/30000 [7:56:34<53:18:51,  7.25s/it, lr: 1.0e-04 loss: 3.945e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3533/30000 [7:56:34<59:36:01,  8.11s/it, lr: 1.0e-04 loss: 3.945e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3533/30000 [7:56:34<59:36:01,  8.11s/it, lr: 1.0e-04 loss: 3.945e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3533/30000 [7:56:42<59:36:01,  8.11s/it, lr: 1.0e-04 loss: 8.335e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3533/30000 [7:56:42<59:36:01,  8.11s/it, lr: 1.0e-04 loss: 8.335e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3534/30000 [7:56:42<58:30:07,  7.96s/it, lr: 1.0e-04 loss: 8.335e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3534/30000 [7:56:42<58:30:07,  7.96s/it, lr: 1.0e-04 loss: 8.335e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3534/30000 [7:56:52<58:30:07,  7.96s/it, lr: 1.0e-04 loss: 2.212e-01]
Xiang_wan_2_2_lora:  12%|####5                                  | 3534/30000 [7:56:52<58:30:07,  7.96s/it, lr: 1.0e-04 loss: 2.212e-01]
Xiang_wan_2_2_lora:  12%|####5                                  | 3535/30000 [7:56:52<63:13:41,  8.60s/it, lr: 1.0e-04 loss: 2.212e-01]
Xiang_wan_2_2_lora:  12%|####5                                  | 3535/30000 [7:56:52<63:13:41,  8.60s/it, lr: 1.0e-04 loss: 2.212e-01]
Xiang_wan_2_2_lora:  12%|####5                                  | 3535/30000 [7:57:00<63:13:41,  8.60s/it, lr: 1.0e-04 loss: 8.380e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3535/30000 [7:57:00<63:13:41,  8.60s/it, lr: 1.0e-04 loss: 8.380e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3536/30000 [7:57:00<61:03:51,  8.31s/it, lr: 1.0e-04 loss: 8.380e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3536/30000 [7:57:00<61:03:51,  8.31s/it, lr: 1.0e-04 loss: 8.380e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3536/30000 [7:57:10<61:03:51,  8.31s/it, lr: 1.0e-04 loss: 8.777e-01]
Xiang_wan_2_2_lora:  12%|####5                                  | 3536/30000 [7:57:10<61:03:51,  8.31s/it, lr: 1.0e-04 loss: 8.777e-01]
Xiang_wan_2_2_lora:  12%|####5                                  | 3537/30000 [7:57:10<65:08:36,  8.86s/it, lr: 1.0e-04 loss: 8.777e-01]
Xiang_wan_2_2_lora:  12%|####5                                  | 3537/30000 [7:57:10<65:08:36,  8.86s/it, lr: 1.0e-04 loss: 8.777e-01]
Xiang_wan_2_2_lora:  12%|####5                                  | 3537/30000 [7:57:18<65:08:36,  8.86s/it, lr: 1.0e-04 loss: 8.804e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3537/30000 [7:57:18<65:08:36,  8.86s/it, lr: 1.0e-04 loss: 8.804e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3538/30000 [7:57:18<63:24:35,  8.63s/it, lr: 1.0e-04 loss: 8.804e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3538/30000 [7:57:18<63:24:35,  8.63s/it, lr: 1.0e-04 loss: 8.804e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3538/30000 [7:57:28<63:24:35,  8.63s/it, lr: 1.0e-04 loss: 5.277e-02]
Xiang_wan_2_2_lora:  12%|####5                                  | 3538/30000 [7:57:28<63:24:35,  8.63s/it, lr: 1.0e-04 loss: 5.277e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3539/30000 [7:57:28<66:53:40,  9.10s/it, lr: 1.0e-04 loss: 5.277e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3539/30000 [7:57:28<66:53:40,  9.10s/it, lr: 1.0e-04 loss: 5.277e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3539/30000 [7:57:38<66:53:40,  9.10s/it, lr: 1.0e-04 loss: 5.676e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3539/30000 [7:57:38<66:53:40,  9.10s/it, lr: 1.0e-04 loss: 5.676e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3540/30000 [7:57:48<66:53:30,  9.10s/it, lr: 1.0e-04 loss: 2.265e-01]
Xiang_wan_2_2_lora:  12%|####6                                  | 3540/30000 [7:57:48<66:53:30,  9.10s/it, lr: 1.0e-04 loss: 2.265e-01]
Xiang_wan_2_2_lora:  12%|####6                                  | 3541/30000 [7:57:48<53:34:33,  7.29s/it, lr: 1.0e-04 loss: 2.265e-01]
Xiang_wan_2_2_lora:  12%|####6                                  | 3541/30000 [7:57:48<53:34:33,  7.29s/it, lr: 1.0e-04 loss: 2.265e-01]
Xiang_wan_2_2_lora:  12%|####6                                  | 3541/30000 [7:57:59<53:34:33,  7.29s/it, lr: 1.0e-04 loss: 6.646e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3541/30000 [7:57:59<53:34:33,  7.29s/it, lr: 1.0e-04 loss: 6.646e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3542/30000 [7:57:59<59:01:43,  8.03s/it, lr: 1.0e-04 loss: 6.646e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3542/30000 [7:57:59<59:01:43,  8.03s/it, lr: 1.0e-04 loss: 6.646e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3542/30000 [7:58:06<59:01:43,  8.03s/it, lr: 1.0e-04 loss: 2.054e-01]
Xiang_wan_2_2_lora:  12%|####6                                  | 3542/30000 [7:58:06<59:01:43,  8.03s/it, lr: 1.0e-04 loss: 2.054e-01]
Xiang_wan_2_2_lora:  12%|####6                                  | 3543/30000 [7:58:06<58:02:00,  7.90s/it, lr: 1.0e-04 loss: 2.054e-01]
Xiang_wan_2_2_lora:  12%|####6                                  | 3543/30000 [7:58:06<58:02:00,  7.90s/it, lr: 1.0e-04 loss: 2.054e-01]
Xiang_wan_2_2_lora:  12%|####6                                  | 3543/30000 [7:58:14<58:02:00,  7.90s/it, lr: 1.0e-04 loss: 7.166e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3543/30000 [7:58:14<58:02:00,  7.90s/it, lr: 1.0e-04 loss: 7.166e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3544/30000 [7:58:14<57:17:33,  7.80s/it, lr: 1.0e-04 loss: 7.166e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3544/30000 [7:58:14<57:17:33,  7.80s/it, lr: 1.0e-04 loss: 7.166e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3544/30000 [7:58:22<57:17:33,  7.80s/it, lr: 1.0e-04 loss: 1.252e-01]
Xiang_wan_2_2_lora:  12%|####6                                  | 3544/30000 [7:58:22<57:17:33,  7.80s/it, lr: 1.0e-04 loss: 1.252e-01]
Xiang_wan_2_2_lora:  12%|####6                                  | 3545/30000 [7:58:22<57:35:27,  7.84s/it, lr: 1.0e-04 loss: 1.252e-01]
Xiang_wan_2_2_lora:  12%|####6                                  | 3545/30000 [7:58:22<57:35:27,  7.84s/it, lr: 1.0e-04 loss: 1.252e-01]
Xiang_wan_2_2_lora:  12%|####6                                  | 3545/30000 [7:58:29<57:35:27,  7.84s/it, lr: 1.0e-04 loss: 7.774e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3545/30000 [7:58:29<57:35:27,  7.84s/it, lr: 1.0e-04 loss: 7.774e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3546/30000 [7:58:29<56:59:30,  7.76s/it, lr: 1.0e-04 loss: 7.774e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3546/30000 [7:58:29<56:59:30,  7.76s/it, lr: 1.0e-04 loss: 7.774e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3546/30000 [7:58:37<56:59:30,  7.76s/it, lr: 1.0e-04 loss: 5.530e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3546/30000 [7:58:37<56:59:30,  7.76s/it, lr: 1.0e-04 loss: 5.530e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3547/30000 [7:58:37<56:27:46,  7.68s/it, lr: 1.0e-04 loss: 5.530e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3547/30000 [7:58:37<56:27:46,  7.68s/it, lr: 1.0e-04 loss: 5.530e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3547/30000 [7:58:47<56:27:46,  7.68s/it, lr: 1.0e-04 loss: 8.584e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3547/30000 [7:58:47<56:27:46,  7.68s/it, lr: 1.0e-04 loss: 8.584e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3548/30000 [7:58:47<61:56:30,  8.43s/it, lr: 1.0e-04 loss: 8.584e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3548/30000 [7:58:47<61:56:30,  8.43s/it, lr: 1.0e-04 loss: 8.584e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3548/30000 [7:58:54<61:56:30,  8.43s/it, lr: 1.0e-04 loss: 9.303e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3548/30000 [7:58:54<61:56:30,  8.43s/it, lr: 1.0e-04 loss: 9.303e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3549/30000 [7:58:54<60:00:44,  8.17s/it, lr: 1.0e-04 loss: 9.303e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3549/30000 [7:58:54<60:00:44,  8.17s/it, lr: 1.0e-04 loss: 9.303e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3549/30000 [7:59:01<60:00:44,  8.17s/it, lr: 1.0e-04 loss: 1.104e-01]
Xiang_wan_2_2_lora:  12%|####6                                  | 3549/30000 [7:59:01<60:00:44,  8.17s/it, lr: 1.0e-04 loss: 1.104e-01]
Xiang_wan_2_2_lora:  12%|####6                                  | 3550/30000 [7:59:11<60:00:36,  8.17s/it, lr: 1.0e-04 loss: 9.618e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3550/30000 [7:59:11<60:00:36,  8.17s/it, lr: 1.0e-04 loss: 9.618e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3551/30000 [7:59:11<49:09:07,  6.69s/it, lr: 1.0e-04 loss: 9.618e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3551/30000 [7:59:11<49:09:07,  6.69s/it, lr: 1.0e-04 loss: 9.618e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3551/30000 [7:59:20<49:09:07,  6.69s/it, lr: 1.0e-04 loss: 5.556e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3551/30000 [7:59:20<49:09:07,  6.69s/it, lr: 1.0e-04 loss: 5.556e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3552/30000 [7:59:20<54:04:08,  7.36s/it, lr: 1.0e-04 loss: 5.556e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3552/30000 [7:59:20<54:04:08,  7.36s/it, lr: 1.0e-04 loss: 5.556e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3552/30000 [7:59:30<54:04:08,  7.36s/it, lr: 1.0e-04 loss: 5.981e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3552/30000 [7:59:30<54:04:08,  7.36s/it, lr: 1.0e-04 loss: 5.981e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3553/30000 [7:59:30<58:20:40,  7.94s/it, lr: 1.0e-04 loss: 5.981e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3553/30000 [7:59:30<58:20:40,  7.94s/it, lr: 1.0e-04 loss: 5.981e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3553/30000 [7:59:37<58:20:40,  7.94s/it, lr: 1.0e-04 loss: 9.810e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3553/30000 [7:59:37<58:20:40,  7.94s/it, lr: 1.0e-04 loss: 9.810e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3554/30000 [7:59:37<55:45:30,  7.59s/it, lr: 1.0e-04 loss: 9.810e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3554/30000 [7:59:37<55:45:30,  7.59s/it, lr: 1.0e-04 loss: 9.810e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3554/30000 [7:59:43<55:45:30,  7.59s/it, lr: 1.0e-04 loss: 8.770e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3554/30000 [7:59:43<55:45:30,  7.59s/it, lr: 1.0e-04 loss: 8.770e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3555/30000 [7:59:43<53:35:10,  7.29s/it, lr: 1.0e-04 loss: 8.770e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3555/30000 [7:59:43<53:35:10,  7.29s/it, lr: 1.0e-04 loss: 8.770e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3555/30000 [7:59:53<53:35:10,  7.29s/it, lr: 1.0e-04 loss: 7.100e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3555/30000 [7:59:53<53:35:10,  7.29s/it, lr: 1.0e-04 loss: 7.100e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3556/30000 [7:59:53<57:51:33,  7.88s/it, lr: 1.0e-04 loss: 7.100e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3556/30000 [7:59:53<57:51:33,  7.88s/it, lr: 1.0e-04 loss: 7.100e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3556/30000 [8:00:02<57:51:33,  7.88s/it, lr: 1.0e-04 loss: 1.319e-01]
Xiang_wan_2_2_lora:  12%|####6                                  | 3556/30000 [8:00:02<57:51:33,  7.88s/it, lr: 1.0e-04 loss: 1.319e-01]
Xiang_wan_2_2_lora:  12%|####6                                  | 3557/30000 [8:00:02<60:59:19,  8.30s/it, lr: 1.0e-04 loss: 1.319e-01]
Xiang_wan_2_2_lora:  12%|####6                                  | 3557/30000 [8:00:02<60:59:19,  8.30s/it, lr: 1.0e-04 loss: 1.319e-01]
Xiang_wan_2_2_lora:  12%|####6                                  | 3557/30000 [8:00:11<60:59:19,  8.30s/it, lr: 1.0e-04 loss: 4.899e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3557/30000 [8:00:11<60:59:19,  8.30s/it, lr: 1.0e-04 loss: 4.899e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3558/30000 [8:00:11<63:09:18,  8.60s/it, lr: 1.0e-04 loss: 4.899e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3558/30000 [8:00:11<63:09:18,  8.60s/it, lr: 1.0e-04 loss: 4.899e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3558/30000 [8:00:18<63:09:18,  8.60s/it, lr: 1.0e-04 loss: 8.273e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3558/30000 [8:00:18<63:09:18,  8.60s/it, lr: 1.0e-04 loss: 8.273e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3559/30000 [8:00:18<58:39:52,  7.99s/it, lr: 1.0e-04 loss: 8.273e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3559/30000 [8:00:18<58:39:52,  7.99s/it, lr: 1.0e-04 loss: 8.273e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3559/30000 [8:00:27<58:39:52,  7.99s/it, lr: 1.0e-04 loss: 6.446e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3559/30000 [8:00:27<58:39:52,  7.99s/it, lr: 1.0e-04 loss: 6.446e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3560/30000 [8:00:37<58:39:44,  7.99s/it, lr: 1.0e-04 loss: 4.752e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3560/30000 [8:00:37<58:39:44,  7.99s/it, lr: 1.0e-04 loss: 4.752e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3561/30000 [8:00:37<47:36:50,  6.48s/it, lr: 1.0e-04 loss: 4.752e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3561/30000 [8:00:37<47:36:50,  6.48s/it, lr: 1.0e-04 loss: 4.752e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3561/30000 [8:00:46<47:36:50,  6.48s/it, lr: 1.0e-04 loss: 6.227e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3561/30000 [8:00:46<47:36:50,  6.48s/it, lr: 1.0e-04 loss: 6.227e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3562/30000 [8:00:46<52:41:20,  7.17s/it, lr: 1.0e-04 loss: 6.227e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3562/30000 [8:00:46<52:41:20,  7.17s/it, lr: 1.0e-04 loss: 6.227e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3562/30000 [8:00:53<52:41:20,  7.17s/it, lr: 1.0e-04 loss: 7.467e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3562/30000 [8:00:53<52:41:20,  7.17s/it, lr: 1.0e-04 loss: 7.467e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3563/30000 [8:00:53<51:36:05,  7.03s/it, lr: 1.0e-04 loss: 7.467e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3563/30000 [8:00:53<51:36:05,  7.03s/it, lr: 1.0e-04 loss: 7.467e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3563/30000 [8:00:59<51:36:05,  7.03s/it, lr: 1.0e-04 loss: 7.644e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3563/30000 [8:00:59<51:36:05,  7.03s/it, lr: 1.0e-04 loss: 7.644e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3564/30000 [8:00:59<50:33:08,  6.88s/it, lr: 1.0e-04 loss: 7.644e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3564/30000 [8:00:59<50:33:08,  6.88s/it, lr: 1.0e-04 loss: 7.644e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3564/30000 [8:01:08<50:33:08,  6.88s/it, lr: 1.0e-04 loss: 5.393e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3564/30000 [8:01:08<50:33:08,  6.88s/it, lr: 1.0e-04 loss: 5.393e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3565/30000 [8:01:08<55:27:30,  7.55s/it, lr: 1.0e-04 loss: 5.393e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3565/30000 [8:01:08<55:27:30,  7.55s/it, lr: 1.0e-04 loss: 5.393e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3565/30000 [8:01:15<55:27:30,  7.55s/it, lr: 1.0e-04 loss: 7.926e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3565/30000 [8:01:15<55:27:30,  7.55s/it, lr: 1.0e-04 loss: 7.926e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3566/30000 [8:01:15<53:24:02,  7.27s/it, lr: 1.0e-04 loss: 7.926e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3566/30000 [8:01:15<53:24:02,  7.27s/it, lr: 1.0e-04 loss: 7.926e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3566/30000 [8:01:21<53:24:02,  7.27s/it, lr: 1.0e-04 loss: 1.534e-01]
Xiang_wan_2_2_lora:  12%|####6                                  | 3566/30000 [8:01:21<53:24:02,  7.27s/it, lr: 1.0e-04 loss: 1.534e-01]
Xiang_wan_2_2_lora:  12%|####6                                  | 3567/30000 [8:01:21<51:49:23,  7.06s/it, lr: 1.0e-04 loss: 1.534e-01]
Xiang_wan_2_2_lora:  12%|####6                                  | 3567/30000 [8:01:21<51:49:23,  7.06s/it, lr: 1.0e-04 loss: 1.534e-01]
Xiang_wan_2_2_lora:  12%|####6                                  | 3567/30000 [8:01:31<51:49:23,  7.06s/it, lr: 1.0e-04 loss: 1.027e-01]
Xiang_wan_2_2_lora:  12%|####6                                  | 3567/30000 [8:01:31<51:49:23,  7.06s/it, lr: 1.0e-04 loss: 1.027e-01]
Xiang_wan_2_2_lora:  12%|####6                                  | 3568/30000 [8:01:31<56:54:58,  7.75s/it, lr: 1.0e-04 loss: 1.027e-01]
Xiang_wan_2_2_lora:  12%|####6                                  | 3568/30000 [8:01:31<56:54:58,  7.75s/it, lr: 1.0e-04 loss: 1.027e-01]
Xiang_wan_2_2_lora:  12%|####6                                  | 3568/30000 [8:01:40<56:54:58,  7.75s/it, lr: 1.0e-04 loss: 5.944e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3568/30000 [8:01:40<56:54:58,  7.75s/it, lr: 1.0e-04 loss: 5.944e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3569/30000 [8:01:40<60:19:06,  8.22s/it, lr: 1.0e-04 loss: 5.944e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3569/30000 [8:01:40<60:19:06,  8.22s/it, lr: 1.0e-04 loss: 5.944e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3569/30000 [8:01:49<60:19:06,  8.22s/it, lr: 1.0e-04 loss: 8.498e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3569/30000 [8:01:49<60:19:06,  8.22s/it, lr: 1.0e-04 loss: 8.498e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3570/30000 [8:01:56<60:18:58,  8.22s/it, lr: 1.0e-04 loss: 4.722e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3570/30000 [8:01:56<60:18:58,  8.22s/it, lr: 1.0e-04 loss: 4.722e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3571/30000 [8:01:56<43:43:27,  5.96s/it, lr: 1.0e-04 loss: 4.722e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3571/30000 [8:01:56<43:43:27,  5.96s/it, lr: 1.0e-04 loss: 4.722e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3571/30000 [8:02:03<43:43:27,  5.96s/it, lr: 1.0e-04 loss: 8.004e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3571/30000 [8:02:03<43:43:27,  5.96s/it, lr: 1.0e-04 loss: 8.004e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3572/30000 [8:02:03<44:46:58,  6.10s/it, lr: 1.0e-04 loss: 8.004e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3572/30000 [8:02:03<44:46:58,  6.10s/it, lr: 1.0e-04 loss: 8.004e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3572/30000 [8:02:12<44:46:58,  6.10s/it, lr: 1.0e-04 loss: 8.866e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3572/30000 [8:02:12<44:46:58,  6.10s/it, lr: 1.0e-04 loss: 8.866e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3573/30000 [8:02:12<50:50:59,  6.93s/it, lr: 1.0e-04 loss: 8.866e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3573/30000 [8:02:12<50:50:59,  6.93s/it, lr: 1.0e-04 loss: 8.866e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3573/30000 [8:02:21<50:50:59,  6.93s/it, lr: 1.0e-04 loss: 7.576e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3573/30000 [8:02:21<50:50:59,  6.93s/it, lr: 1.0e-04 loss: 7.576e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3574/30000 [8:02:21<55:37:41,  7.58s/it, lr: 1.0e-04 loss: 7.576e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3574/30000 [8:02:21<55:37:41,  7.58s/it, lr: 1.0e-04 loss: 7.576e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3574/30000 [8:02:28<55:37:41,  7.58s/it, lr: 1.0e-04 loss: 9.541e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3574/30000 [8:02:28<55:37:41,  7.58s/it, lr: 1.0e-04 loss: 9.541e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3575/30000 [8:02:28<54:07:21,  7.37s/it, lr: 1.0e-04 loss: 9.541e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3575/30000 [8:02:28<54:07:21,  7.37s/it, lr: 1.0e-04 loss: 9.541e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3575/30000 [8:02:37<54:07:21,  7.37s/it, lr: 1.0e-04 loss: 7.422e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3575/30000 [8:02:37<54:07:21,  7.37s/it, lr: 1.0e-04 loss: 7.422e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3576/30000 [8:02:37<58:11:57,  7.93s/it, lr: 1.0e-04 loss: 7.422e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3576/30000 [8:02:37<58:11:57,  7.93s/it, lr: 1.0e-04 loss: 7.422e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3576/30000 [8:02:44<58:11:57,  7.93s/it, lr: 1.0e-04 loss: 2.622e-01]
Xiang_wan_2_2_lora:  12%|####6                                  | 3576/30000 [8:02:44<58:11:57,  7.93s/it, lr: 1.0e-04 loss: 2.622e-01]
Xiang_wan_2_2_lora:  12%|####6                                  | 3577/30000 [8:02:44<55:19:31,  7.54s/it, lr: 1.0e-04 loss: 2.622e-01]
Xiang_wan_2_2_lora:  12%|####6                                  | 3577/30000 [8:02:44<55:19:31,  7.54s/it, lr: 1.0e-04 loss: 2.622e-01]
Xiang_wan_2_2_lora:  12%|####6                                  | 3577/30000 [8:02:51<55:19:31,  7.54s/it, lr: 1.0e-04 loss: 7.835e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3577/30000 [8:02:51<55:19:31,  7.54s/it, lr: 1.0e-04 loss: 7.835e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3578/30000 [8:02:51<53:18:39,  7.26s/it, lr: 1.0e-04 loss: 7.835e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3578/30000 [8:02:51<53:18:39,  7.26s/it, lr: 1.0e-04 loss: 7.835e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3578/30000 [8:03:00<53:18:39,  7.26s/it, lr: 1.0e-04 loss: 5.517e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3578/30000 [8:03:00<53:18:39,  7.26s/it, lr: 1.0e-04 loss: 5.517e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3579/30000 [8:03:00<57:35:16,  7.85s/it, lr: 1.0e-04 loss: 5.517e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3579/30000 [8:03:00<57:35:16,  7.85s/it, lr: 1.0e-04 loss: 5.517e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3579/30000 [8:03:09<57:35:16,  7.85s/it, lr: 1.0e-04 loss: 6.121e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3579/30000 [8:03:09<57:35:16,  7.85s/it, lr: 1.0e-04 loss: 6.121e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3580/30000 [8:03:18<57:35:09,  7.85s/it, lr: 1.0e-04 loss: 4.336e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3580/30000 [8:03:18<57:35:09,  7.85s/it, lr: 1.0e-04 loss: 4.336e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3581/30000 [8:03:18<46:53:42,  6.39s/it, lr: 1.0e-04 loss: 4.336e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3581/30000 [8:03:18<46:53:42,  6.39s/it, lr: 1.0e-04 loss: 4.336e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3581/30000 [8:03:28<46:53:42,  6.39s/it, lr: 1.0e-04 loss: 6.092e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3581/30000 [8:03:28<46:53:42,  6.39s/it, lr: 1.0e-04 loss: 6.092e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3582/30000 [8:03:28<52:11:52,  7.11s/it, lr: 1.0e-04 loss: 6.092e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3582/30000 [8:03:28<52:11:52,  7.11s/it, lr: 1.0e-04 loss: 6.092e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3582/30000 [8:03:35<52:11:52,  7.11s/it, lr: 1.0e-04 loss: 1.767e-01]
Xiang_wan_2_2_lora:  12%|####6                                  | 3582/30000 [8:03:35<52:11:52,  7.11s/it, lr: 1.0e-04 loss: 1.767e-01]
Xiang_wan_2_2_lora:  12%|####6                                  | 3583/30000 [8:03:35<51:38:54,  7.04s/it, lr: 1.0e-04 loss: 1.767e-01]
Xiang_wan_2_2_lora:  12%|####6                                  | 3583/30000 [8:03:35<51:38:54,  7.04s/it, lr: 1.0e-04 loss: 1.767e-01]
Xiang_wan_2_2_lora:  12%|####6                                  | 3583/30000 [8:03:41<51:38:54,  7.04s/it, lr: 1.0e-04 loss: 8.342e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3583/30000 [8:03:41<51:38:54,  7.04s/it, lr: 1.0e-04 loss: 8.342e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3584/30000 [8:03:41<50:40:38,  6.91s/it, lr: 1.0e-04 loss: 8.342e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3584/30000 [8:03:41<50:40:38,  6.91s/it, lr: 1.0e-04 loss: 8.342e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3584/30000 [8:03:50<50:40:38,  6.91s/it, lr: 1.0e-04 loss: 1.296e-01]
Xiang_wan_2_2_lora:  12%|####6                                  | 3584/30000 [8:03:50<50:40:38,  6.91s/it, lr: 1.0e-04 loss: 1.296e-01]
Xiang_wan_2_2_lora:  12%|####6                                  | 3585/30000 [8:03:50<55:28:59,  7.56s/it, lr: 1.0e-04 loss: 1.296e-01]
Xiang_wan_2_2_lora:  12%|####6                                  | 3585/30000 [8:03:50<55:28:59,  7.56s/it, lr: 1.0e-04 loss: 1.296e-01]
Xiang_wan_2_2_lora:  12%|####6                                  | 3585/30000 [8:03:57<55:28:59,  7.56s/it, lr: 1.0e-04 loss: 8.676e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3585/30000 [8:03:57<55:28:59,  7.56s/it, lr: 1.0e-04 loss: 8.676e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3586/30000 [8:03:57<53:28:27,  7.29s/it, lr: 1.0e-04 loss: 8.676e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3586/30000 [8:03:57<53:28:27,  7.29s/it, lr: 1.0e-04 loss: 8.676e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3586/30000 [8:04:06<53:28:27,  7.29s/it, lr: 1.0e-04 loss: 3.482e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3586/30000 [8:04:06<53:28:27,  7.29s/it, lr: 1.0e-04 loss: 3.482e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3587/30000 [8:04:06<57:39:28,  7.86s/it, lr: 1.0e-04 loss: 3.482e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3587/30000 [8:04:06<57:39:28,  7.86s/it, lr: 1.0e-04 loss: 3.482e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3587/30000 [8:04:13<57:39:28,  7.86s/it, lr: 1.0e-04 loss: 7.880e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3587/30000 [8:04:13<57:39:28,  7.86s/it, lr: 1.0e-04 loss: 7.880e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3588/30000 [8:04:13<54:51:39,  7.48s/it, lr: 1.0e-04 loss: 7.880e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3588/30000 [8:04:13<54:51:39,  7.48s/it, lr: 1.0e-04 loss: 7.880e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3588/30000 [8:04:19<54:51:39,  7.48s/it, lr: 1.0e-04 loss: 4.456e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3588/30000 [8:04:19<54:51:39,  7.48s/it, lr: 1.0e-04 loss: 4.456e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3589/30000 [8:04:19<52:40:00,  7.18s/it, lr: 1.0e-04 loss: 4.456e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3589/30000 [8:04:19<52:40:00,  7.18s/it, lr: 1.0e-04 loss: 4.456e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3589/30000 [8:04:29<52:40:00,  7.18s/it, lr: 1.0e-04 loss: 5.404e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3589/30000 [8:04:29<52:40:00,  7.18s/it, lr: 1.0e-04 loss: 5.404e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3590/30000 [8:04:38<52:39:53,  7.18s/it, lr: 1.0e-04 loss: 4.834e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3590/30000 [8:04:38<52:39:53,  7.18s/it, lr: 1.0e-04 loss: 4.834e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3591/30000 [8:04:38<44:14:24,  6.03s/it, lr: 1.0e-04 loss: 4.834e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3591/30000 [8:04:38<44:14:24,  6.03s/it, lr: 1.0e-04 loss: 4.834e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3591/30000 [8:04:47<44:14:24,  6.03s/it, lr: 1.0e-04 loss: 6.990e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3591/30000 [8:04:47<44:14:24,  6.03s/it, lr: 1.0e-04 loss: 6.990e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3592/30000 [8:04:47<50:09:34,  6.84s/it, lr: 1.0e-04 loss: 6.990e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3592/30000 [8:04:47<50:09:34,  6.84s/it, lr: 1.0e-04 loss: 6.990e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3592/30000 [8:04:57<50:09:34,  6.84s/it, lr: 1.0e-04 loss: 6.902e-01]
Xiang_wan_2_2_lora:  12%|####6                                  | 3592/30000 [8:04:57<50:09:34,  6.84s/it, lr: 1.0e-04 loss: 6.902e-01]
Xiang_wan_2_2_lora:  12%|####6                                  | 3593/30000 [8:04:57<54:48:17,  7.47s/it, lr: 1.0e-04 loss: 6.902e-01]
Xiang_wan_2_2_lora:  12%|####6                                  | 3593/30000 [8:04:57<54:48:17,  7.47s/it, lr: 1.0e-04 loss: 6.902e-01]
Xiang_wan_2_2_lora:  12%|####6                                  | 3593/30000 [8:05:06<54:48:17,  7.47s/it, lr: 1.0e-04 loss: 6.562e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3593/30000 [8:05:06<54:48:17,  7.47s/it, lr: 1.0e-04 loss: 6.562e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3594/30000 [8:05:06<58:35:08,  7.99s/it, lr: 1.0e-04 loss: 6.562e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3594/30000 [8:05:06<58:35:08,  7.99s/it, lr: 1.0e-04 loss: 6.562e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3594/30000 [8:05:15<58:35:08,  7.99s/it, lr: 1.0e-04 loss: 1.032e-01]
Xiang_wan_2_2_lora:  12%|####6                                  | 3594/30000 [8:05:15<58:35:08,  7.99s/it, lr: 1.0e-04 loss: 1.032e-01]
Xiang_wan_2_2_lora:  12%|####6                                  | 3595/30000 [8:05:15<61:14:43,  8.35s/it, lr: 1.0e-04 loss: 1.032e-01]
Xiang_wan_2_2_lora:  12%|####6                                  | 3595/30000 [8:05:15<61:14:43,  8.35s/it, lr: 1.0e-04 loss: 1.032e-01]
Xiang_wan_2_2_lora:  12%|####6                                  | 3595/30000 [8:05:25<61:14:43,  8.35s/it, lr: 1.0e-04 loss: 8.119e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3595/30000 [8:05:25<61:14:43,  8.35s/it, lr: 1.0e-04 loss: 8.119e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3596/30000 [8:05:25<63:16:26,  8.63s/it, lr: 1.0e-04 loss: 8.119e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3596/30000 [8:05:25<63:16:26,  8.63s/it, lr: 1.0e-04 loss: 8.119e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3596/30000 [8:05:31<63:16:26,  8.63s/it, lr: 1.0e-04 loss: 5.431e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3596/30000 [8:05:31<63:16:26,  8.63s/it, lr: 1.0e-04 loss: 5.431e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3597/30000 [8:05:31<58:55:33,  8.03s/it, lr: 1.0e-04 loss: 5.431e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3597/30000 [8:05:31<58:55:33,  8.03s/it, lr: 1.0e-04 loss: 5.431e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3597/30000 [8:05:41<58:55:33,  8.03s/it, lr: 1.0e-04 loss: 5.880e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3597/30000 [8:05:41<58:55:33,  8.03s/it, lr: 1.0e-04 loss: 5.880e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3598/30000 [8:05:41<61:54:29,  8.44s/it, lr: 1.0e-04 loss: 5.880e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3598/30000 [8:05:41<61:54:29,  8.44s/it, lr: 1.0e-04 loss: 5.880e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3598/30000 [8:05:50<61:54:29,  8.44s/it, lr: 1.0e-04 loss: 4.264e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3598/30000 [8:05:50<61:54:29,  8.44s/it, lr: 1.0e-04 loss: 4.264e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3599/30000 [8:05:50<63:54:38,  8.71s/it, lr: 1.0e-04 loss: 4.264e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3599/30000 [8:05:50<63:54:38,  8.71s/it, lr: 1.0e-04 loss: 4.264e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3599/30000 [8:05:57<63:54:38,  8.71s/it, lr: 1.0e-04 loss: 6.999e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3599/30000 [8:05:57<63:54:38,  8.71s/it, lr: 1.0e-04 loss: 6.999e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3600/30000 [8:06:03<63:54:29,  8.71s/it, lr: 1.0e-04 loss: 5.492e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3600/30000 [8:06:03<63:54:29,  8.71s/it, lr: 1.0e-04 loss: 5.492e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3601/30000 [8:06:03<45:36:11,  6.22s/it, lr: 1.0e-04 loss: 5.492e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3601/30000 [8:06:03<45:36:11,  6.22s/it, lr: 1.0e-04 loss: 5.492e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3601/30000 [8:06:10<45:36:11,  6.22s/it, lr: 1.0e-04 loss: 9.295e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3601/30000 [8:06:10<45:36:11,  6.22s/it, lr: 1.0e-04 loss: 9.295e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3602/30000 [8:06:10<46:14:26,  6.31s/it, lr: 1.0e-04 loss: 9.295e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3602/30000 [8:06:10<46:14:26,  6.31s/it, lr: 1.0e-04 loss: 9.295e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3602/30000 [8:06:19<46:14:26,  6.31s/it, lr: 1.0e-04 loss: 4.384e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3602/30000 [8:06:19<46:14:26,  6.31s/it, lr: 1.0e-04 loss: 4.384e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3603/30000 [8:06:19<51:58:10,  7.09s/it, lr: 1.0e-04 loss: 4.384e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3603/30000 [8:06:19<51:58:10,  7.09s/it, lr: 1.0e-04 loss: 4.384e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3603/30000 [8:06:28<51:58:10,  7.09s/it, lr: 1.0e-04 loss: 6.239e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3603/30000 [8:06:28<51:58:10,  7.09s/it, lr: 1.0e-04 loss: 6.239e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3604/30000 [8:06:28<56:29:03,  7.70s/it, lr: 1.0e-04 loss: 6.239e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3604/30000 [8:06:28<56:29:03,  7.70s/it, lr: 1.0e-04 loss: 6.239e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3604/30000 [8:06:35<56:29:03,  7.70s/it, lr: 1.0e-04 loss: 1.921e-01]
Xiang_wan_2_2_lora:  12%|####6                                  | 3604/30000 [8:06:35<56:29:03,  7.70s/it, lr: 1.0e-04 loss: 1.921e-01]
Xiang_wan_2_2_lora:  12%|####6                                  | 3605/30000 [8:06:35<54:43:24,  7.46s/it, lr: 1.0e-04 loss: 1.921e-01]
Xiang_wan_2_2_lora:  12%|####6                                  | 3605/30000 [8:06:35<54:43:24,  7.46s/it, lr: 1.0e-04 loss: 1.921e-01]
Xiang_wan_2_2_lora:  12%|####6                                  | 3605/30000 [8:06:42<54:43:24,  7.46s/it, lr: 1.0e-04 loss: 6.860e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3605/30000 [8:06:42<54:43:24,  7.46s/it, lr: 1.0e-04 loss: 6.860e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3606/30000 [8:06:42<52:48:35,  7.20s/it, lr: 1.0e-04 loss: 6.860e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3606/30000 [8:06:42<52:48:35,  7.20s/it, lr: 1.0e-04 loss: 6.860e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3606/30000 [8:06:48<52:48:35,  7.20s/it, lr: 1.0e-04 loss: 3.162e-01]
Xiang_wan_2_2_lora:  12%|####6                                  | 3606/30000 [8:06:48<52:48:35,  7.20s/it, lr: 1.0e-04 loss: 3.162e-01]
Xiang_wan_2_2_lora:  12%|####6                                  | 3607/30000 [8:06:48<51:24:12,  7.01s/it, lr: 1.0e-04 loss: 3.162e-01]
Xiang_wan_2_2_lora:  12%|####6                                  | 3607/30000 [8:06:48<51:24:12,  7.01s/it, lr: 1.0e-04 loss: 3.162e-01]
Xiang_wan_2_2_lora:  12%|####6                                  | 3607/30000 [8:06:58<51:24:12,  7.01s/it, lr: 1.0e-04 loss: 4.810e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3607/30000 [8:06:58<51:24:12,  7.01s/it, lr: 1.0e-04 loss: 4.810e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3608/30000 [8:06:58<56:19:39,  7.68s/it, lr: 1.0e-04 loss: 4.810e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3608/30000 [8:06:58<56:19:39,  7.68s/it, lr: 1.0e-04 loss: 4.810e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3608/30000 [8:07:07<56:19:39,  7.68s/it, lr: 1.0e-04 loss: 7.253e-01]
Xiang_wan_2_2_lora:  12%|####6                                  | 3608/30000 [8:07:07<56:19:39,  7.68s/it, lr: 1.0e-04 loss: 7.253e-01]
Xiang_wan_2_2_lora:  12%|####6                                  | 3609/30000 [8:07:07<59:59:50,  8.18s/it, lr: 1.0e-04 loss: 7.253e-01]
Xiang_wan_2_2_lora:  12%|####6                                  | 3609/30000 [8:07:07<59:59:50,  8.18s/it, lr: 1.0e-04 loss: 7.253e-01]
Xiang_wan_2_2_lora:  12%|####6                                  | 3609/30000 [8:07:16<59:59:50,  8.18s/it, lr: 1.0e-04 loss: 5.583e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3609/30000 [8:07:16<59:59:50,  8.18s/it, lr: 1.0e-04 loss: 5.583e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3610/30000 [8:07:26<59:59:42,  8.18s/it, lr: 1.0e-04 loss: 5.627e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3610/30000 [8:07:26<59:59:42,  8.18s/it, lr: 1.0e-04 loss: 5.627e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3611/30000 [8:07:26<48:11:06,  6.57s/it, lr: 1.0e-04 loss: 5.627e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3611/30000 [8:07:26<48:11:06,  6.57s/it, lr: 1.0e-04 loss: 5.627e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3611/30000 [8:07:32<48:11:06,  6.57s/it, lr: 1.0e-04 loss: 7.265e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3611/30000 [8:07:32<48:11:06,  6.57s/it, lr: 1.0e-04 loss: 7.265e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3612/30000 [8:07:32<48:17:06,  6.59s/it, lr: 1.0e-04 loss: 7.265e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3612/30000 [8:07:32<48:17:06,  6.59s/it, lr: 1.0e-04 loss: 7.265e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3612/30000 [8:07:42<48:17:06,  6.59s/it, lr: 1.0e-04 loss: 7.621e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3612/30000 [8:07:42<48:17:06,  6.59s/it, lr: 1.0e-04 loss: 7.621e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3613/30000 [8:07:42<53:47:28,  7.34s/it, lr: 1.0e-04 loss: 7.621e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3613/30000 [8:07:42<53:47:28,  7.34s/it, lr: 1.0e-04 loss: 7.621e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3613/30000 [8:07:48<53:47:28,  7.34s/it, lr: 1.0e-04 loss: 9.164e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3613/30000 [8:07:48<53:47:28,  7.34s/it, lr: 1.0e-04 loss: 9.164e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3614/30000 [8:07:48<52:26:23,  7.15s/it, lr: 1.0e-04 loss: 9.164e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3614/30000 [8:07:48<52:26:23,  7.15s/it, lr: 1.0e-04 loss: 9.164e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3614/30000 [8:07:58<52:26:23,  7.15s/it, lr: 1.0e-04 loss: 4.482e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3614/30000 [8:07:58<52:26:23,  7.15s/it, lr: 1.0e-04 loss: 4.482e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3615/30000 [8:07:58<56:49:06,  7.75s/it, lr: 1.0e-04 loss: 4.482e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3615/30000 [8:07:58<56:49:06,  7.75s/it, lr: 1.0e-04 loss: 4.482e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3615/30000 [8:08:07<56:49:06,  7.75s/it, lr: 1.0e-04 loss: 7.870e-02]
Xiang_wan_2_2_lora:  12%|####6                                  | 3615/30000 [8:08:07<56:49:06,  7.75s/it, lr: 1.0e-04 loss: 7.870e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3616/30000 [8:08:07<60:19:40,  8.23s/it, lr: 1.0e-04 loss: 7.870e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3616/30000 [8:08:07<60:19:40,  8.23s/it, lr: 1.0e-04 loss: 7.870e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3616/30000 [8:08:14<60:19:40,  8.23s/it, lr: 1.0e-04 loss: 6.606e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3616/30000 [8:08:14<60:19:40,  8.23s/it, lr: 1.0e-04 loss: 6.606e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3617/30000 [8:08:14<56:50:24,  7.76s/it, lr: 1.0e-04 loss: 6.606e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3617/30000 [8:08:14<56:50:24,  7.76s/it, lr: 1.0e-04 loss: 6.606e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3617/30000 [8:08:23<56:50:24,  7.76s/it, lr: 1.0e-04 loss: 6.418e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3617/30000 [8:08:23<56:50:24,  7.76s/it, lr: 1.0e-04 loss: 6.418e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3618/30000 [8:08:23<60:12:00,  8.21s/it, lr: 1.0e-04 loss: 6.418e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3618/30000 [8:08:23<60:12:00,  8.21s/it, lr: 1.0e-04 loss: 6.418e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3618/30000 [8:08:30<60:12:00,  8.21s/it, lr: 1.0e-04 loss: 6.295e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3618/30000 [8:08:30<60:12:00,  8.21s/it, lr: 1.0e-04 loss: 6.295e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3619/30000 [8:08:30<56:47:50,  7.75s/it, lr: 1.0e-04 loss: 6.295e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3619/30000 [8:08:30<56:47:50,  7.75s/it, lr: 1.0e-04 loss: 6.295e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3619/30000 [8:08:39<56:47:50,  7.75s/it, lr: 1.0e-04 loss: 9.662e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3619/30000 [8:08:39<56:47:50,  7.75s/it, lr: 1.0e-04 loss: 9.662e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3620/30000 [8:08:49<56:47:42,  7.75s/it, lr: 1.0e-04 loss: 4.182e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3620/30000 [8:08:49<56:47:42,  7.75s/it, lr: 1.0e-04 loss: 4.182e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3621/30000 [8:08:49<46:30:50,  6.35s/it, lr: 1.0e-04 loss: 4.182e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3621/30000 [8:08:49<46:30:50,  6.35s/it, lr: 1.0e-04 loss: 4.182e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3621/30000 [8:08:55<46:30:50,  6.35s/it, lr: 1.0e-04 loss: 9.873e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3621/30000 [8:08:55<46:30:50,  6.35s/it, lr: 1.0e-04 loss: 9.873e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3622/30000 [8:08:55<46:58:48,  6.41s/it, lr: 1.0e-04 loss: 9.873e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3622/30000 [8:08:55<46:58:48,  6.41s/it, lr: 1.0e-04 loss: 9.873e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3622/30000 [8:09:02<46:58:48,  6.41s/it, lr: 1.0e-04 loss: 1.707e-01]
Xiang_wan_2_2_lora:  12%|####7                                  | 3622/30000 [8:09:02<46:58:48,  6.41s/it, lr: 1.0e-04 loss: 1.707e-01]
Xiang_wan_2_2_lora:  12%|####7                                  | 3623/30000 [8:09:02<47:11:33,  6.44s/it, lr: 1.0e-04 loss: 1.707e-01]
Xiang_wan_2_2_lora:  12%|####7                                  | 3623/30000 [8:09:02<47:11:33,  6.44s/it, lr: 1.0e-04 loss: 1.707e-01]
Xiang_wan_2_2_lora:  12%|####7                                  | 3623/30000 [8:09:08<47:11:33,  6.44s/it, lr: 1.0e-04 loss: 8.913e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3623/30000 [8:09:08<47:11:33,  6.44s/it, lr: 1.0e-04 loss: 8.913e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3624/30000 [8:09:08<47:26:05,  6.47s/it, lr: 1.0e-04 loss: 8.913e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3624/30000 [8:09:08<47:26:05,  6.47s/it, lr: 1.0e-04 loss: 8.913e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3624/30000 [8:09:15<47:26:05,  6.47s/it, lr: 1.0e-04 loss: 7.752e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3624/30000 [8:09:15<47:26:05,  6.47s/it, lr: 1.0e-04 loss: 7.752e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3625/30000 [8:09:15<47:27:40,  6.48s/it, lr: 1.0e-04 loss: 7.752e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3625/30000 [8:09:15<47:27:40,  6.48s/it, lr: 1.0e-04 loss: 7.752e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3625/30000 [8:09:24<47:27:40,  6.48s/it, lr: 1.0e-04 loss: 7.332e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3625/30000 [8:09:24<47:27:40,  6.48s/it, lr: 1.0e-04 loss: 7.332e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3626/30000 [8:09:24<53:23:25,  7.29s/it, lr: 1.0e-04 loss: 7.332e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3626/30000 [8:09:24<53:23:25,  7.29s/it, lr: 1.0e-04 loss: 7.332e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3626/30000 [8:09:31<53:23:25,  7.29s/it, lr: 1.0e-04 loss: 5.582e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3626/30000 [8:09:31<53:23:25,  7.29s/it, lr: 1.0e-04 loss: 5.582e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3627/30000 [8:09:31<51:50:01,  7.08s/it, lr: 1.0e-04 loss: 5.582e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3627/30000 [8:09:31<51:50:01,  7.08s/it, lr: 1.0e-04 loss: 5.582e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3627/30000 [8:09:38<51:50:01,  7.08s/it, lr: 1.0e-04 loss: 9.570e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3627/30000 [8:09:38<51:50:01,  7.08s/it, lr: 1.0e-04 loss: 9.570e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3628/30000 [8:09:38<51:19:15,  7.01s/it, lr: 1.0e-04 loss: 9.570e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3628/30000 [8:09:38<51:19:15,  7.01s/it, lr: 1.0e-04 loss: 9.570e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3628/30000 [8:09:47<51:19:15,  7.01s/it, lr: 1.0e-04 loss: 5.249e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3628/30000 [8:09:47<51:19:15,  7.01s/it, lr: 1.0e-04 loss: 5.249e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3629/30000 [8:09:47<56:11:08,  7.67s/it, lr: 1.0e-04 loss: 5.249e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3629/30000 [8:09:47<56:11:08,  7.67s/it, lr: 1.0e-04 loss: 5.249e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3629/30000 [8:09:56<56:11:08,  7.67s/it, lr: 1.0e-04 loss: 7.031e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3629/30000 [8:09:56<56:11:08,  7.67s/it, lr: 1.0e-04 loss: 7.031e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3630/30000 [8:10:06<56:11:01,  7.67s/it, lr: 1.0e-04 loss: 4.147e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3630/30000 [8:10:06<56:11:01,  7.67s/it, lr: 1.0e-04 loss: 4.147e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3631/30000 [8:10:06<46:10:43,  6.30s/it, lr: 1.0e-04 loss: 4.147e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3631/30000 [8:10:06<46:10:43,  6.30s/it, lr: 1.0e-04 loss: 4.147e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3631/30000 [8:10:15<46:10:43,  6.30s/it, lr: 1.0e-04 loss: 5.552e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3631/30000 [8:10:15<46:10:43,  6.30s/it, lr: 1.0e-04 loss: 5.552e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3632/30000 [8:10:15<51:40:29,  7.06s/it, lr: 1.0e-04 loss: 5.552e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3632/30000 [8:10:15<51:40:29,  7.06s/it, lr: 1.0e-04 loss: 5.552e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3632/30000 [8:10:24<51:40:29,  7.06s/it, lr: 1.0e-04 loss: 1.358e-01]
Xiang_wan_2_2_lora:  12%|####7                                  | 3632/30000 [8:10:24<51:40:29,  7.06s/it, lr: 1.0e-04 loss: 1.358e-01]
Xiang_wan_2_2_lora:  12%|####7                                  | 3633/30000 [8:10:24<56:04:30,  7.66s/it, lr: 1.0e-04 loss: 1.358e-01]
Xiang_wan_2_2_lora:  12%|####7                                  | 3633/30000 [8:10:24<56:04:30,  7.66s/it, lr: 1.0e-04 loss: 1.358e-01]
Xiang_wan_2_2_lora:  12%|####7                                  | 3633/30000 [8:10:34<56:04:30,  7.66s/it, lr: 1.0e-04 loss: 6.493e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3633/30000 [8:10:34<56:04:30,  7.66s/it, lr: 1.0e-04 loss: 6.493e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3634/30000 [8:10:34<59:26:24,  8.12s/it, lr: 1.0e-04 loss: 6.493e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3634/30000 [8:10:34<59:26:24,  8.12s/it, lr: 1.0e-04 loss: 6.493e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3634/30000 [8:10:40<59:26:24,  8.12s/it, lr: 1.0e-04 loss: 7.050e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3634/30000 [8:10:40<59:26:24,  8.12s/it, lr: 1.0e-04 loss: 7.050e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3635/30000 [8:10:40<56:42:31,  7.74s/it, lr: 1.0e-04 loss: 7.050e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3635/30000 [8:10:40<56:42:31,  7.74s/it, lr: 1.0e-04 loss: 7.050e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3635/30000 [8:10:50<56:42:31,  7.74s/it, lr: 1.0e-04 loss: 8.281e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3635/30000 [8:10:50<56:42:31,  7.74s/it, lr: 1.0e-04 loss: 8.281e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3636/30000 [8:10:50<60:01:26,  8.20s/it, lr: 1.0e-04 loss: 8.281e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3636/30000 [8:10:50<60:01:26,  8.20s/it, lr: 1.0e-04 loss: 8.281e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3636/30000 [8:10:56<60:01:26,  8.20s/it, lr: 1.0e-04 loss: 3.104e-01]
Xiang_wan_2_2_lora:  12%|####7                                  | 3636/30000 [8:10:56<60:01:26,  8.20s/it, lr: 1.0e-04 loss: 3.104e-01]
Xiang_wan_2_2_lora:  12%|####7                                  | 3637/30000 [8:10:56<56:31:14,  7.72s/it, lr: 1.0e-04 loss: 3.104e-01]
Xiang_wan_2_2_lora:  12%|####7                                  | 3637/30000 [8:10:56<56:31:14,  7.72s/it, lr: 1.0e-04 loss: 3.104e-01]
Xiang_wan_2_2_lora:  12%|####7                                  | 3637/30000 [8:11:06<56:31:14,  7.72s/it, lr: 1.0e-04 loss: 8.130e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3637/30000 [8:11:06<56:31:14,  7.72s/it, lr: 1.0e-04 loss: 8.130e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3638/30000 [8:11:06<60:00:45,  8.20s/it, lr: 1.0e-04 loss: 8.130e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3638/30000 [8:11:06<60:00:45,  8.20s/it, lr: 1.0e-04 loss: 8.130e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3638/30000 [8:11:12<60:00:45,  8.20s/it, lr: 1.0e-04 loss: 9.414e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3638/30000 [8:11:12<60:00:45,  8.20s/it, lr: 1.0e-04 loss: 9.414e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3639/30000 [8:11:12<56:37:28,  7.73s/it, lr: 1.0e-04 loss: 9.414e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3639/30000 [8:11:12<56:37:28,  7.73s/it, lr: 1.0e-04 loss: 9.414e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3639/30000 [8:11:22<56:37:28,  7.73s/it, lr: 1.0e-04 loss: 8.162e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3639/30000 [8:11:22<56:37:28,  7.73s/it, lr: 1.0e-04 loss: 8.162e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3640/30000 [8:11:28<56:37:20,  7.73s/it, lr: 1.0e-04 loss: 4.699e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3640/30000 [8:11:28<56:37:20,  7.73s/it, lr: 1.0e-04 loss: 4.699e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3641/30000 [8:11:28<41:50:47,  5.72s/it, lr: 1.0e-04 loss: 4.699e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3641/30000 [8:11:28<41:50:47,  5.72s/it, lr: 1.0e-04 loss: 4.699e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3641/30000 [8:11:37<41:50:47,  5.72s/it, lr: 1.0e-04 loss: 7.602e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3641/30000 [8:11:37<41:50:47,  5.72s/it, lr: 1.0e-04 loss: 7.602e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3642/30000 [8:11:38<48:23:00,  6.61s/it, lr: 1.0e-04 loss: 7.602e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3642/30000 [8:11:38<48:23:00,  6.61s/it, lr: 1.0e-04 loss: 7.602e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3642/30000 [8:11:47<48:23:00,  6.61s/it, lr: 1.0e-04 loss: 2.812e-01]
Xiang_wan_2_2_lora:  12%|####7                                  | 3642/30000 [8:11:47<48:23:00,  6.61s/it, lr: 1.0e-04 loss: 2.812e-01]
Xiang_wan_2_2_lora:  12%|####7                                  | 3643/30000 [8:11:47<53:54:12,  7.36s/it, lr: 1.0e-04 loss: 2.812e-01]
Xiang_wan_2_2_lora:  12%|####7                                  | 3643/30000 [8:11:47<53:54:12,  7.36s/it, lr: 1.0e-04 loss: 2.812e-01]
Xiang_wan_2_2_lora:  12%|####7                                  | 3643/30000 [8:11:54<53:54:12,  7.36s/it, lr: 1.0e-04 loss: 6.968e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3643/30000 [8:11:54<53:54:12,  7.36s/it, lr: 1.0e-04 loss: 6.968e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3644/30000 [8:11:54<52:22:00,  7.15s/it, lr: 1.0e-04 loss: 6.968e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3644/30000 [8:11:54<52:22:00,  7.15s/it, lr: 1.0e-04 loss: 6.968e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3644/30000 [8:12:00<52:22:00,  7.15s/it, lr: 1.0e-04 loss: 6.920e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3644/30000 [8:12:00<52:22:00,  7.15s/it, lr: 1.0e-04 loss: 6.920e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3645/30000 [8:12:00<51:12:14,  6.99s/it, lr: 1.0e-04 loss: 6.920e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3645/30000 [8:12:00<51:12:14,  6.99s/it, lr: 1.0e-04 loss: 6.920e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3645/30000 [8:12:10<51:12:14,  6.99s/it, lr: 1.0e-04 loss: 7.651e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3645/30000 [8:12:10<51:12:14,  6.99s/it, lr: 1.0e-04 loss: 7.651e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3646/30000 [8:12:10<56:15:48,  7.69s/it, lr: 1.0e-04 loss: 7.651e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3646/30000 [8:12:10<56:15:48,  7.69s/it, lr: 1.0e-04 loss: 7.651e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3646/30000 [8:12:19<56:15:48,  7.69s/it, lr: 1.0e-04 loss: 1.027e-01]
Xiang_wan_2_2_lora:  12%|####7                                  | 3646/30000 [8:12:19<56:15:48,  7.69s/it, lr: 1.0e-04 loss: 1.027e-01]
Xiang_wan_2_2_lora:  12%|####7                                  | 3647/30000 [8:12:19<59:47:44,  8.17s/it, lr: 1.0e-04 loss: 1.027e-01]
Xiang_wan_2_2_lora:  12%|####7                                  | 3647/30000 [8:12:19<59:47:44,  8.17s/it, lr: 1.0e-04 loss: 1.027e-01]
Xiang_wan_2_2_lora:  12%|####7                                  | 3647/30000 [8:12:28<59:47:44,  8.17s/it, lr: 1.0e-04 loss: 9.136e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3647/30000 [8:12:28<59:47:44,  8.17s/it, lr: 1.0e-04 loss: 9.136e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3648/30000 [8:12:28<62:18:54,  8.51s/it, lr: 1.0e-04 loss: 9.136e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3648/30000 [8:12:28<62:18:54,  8.51s/it, lr: 1.0e-04 loss: 9.136e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3648/30000 [8:12:35<62:18:54,  8.51s/it, lr: 1.0e-04 loss: 9.894e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3648/30000 [8:12:35<62:18:54,  8.51s/it, lr: 1.0e-04 loss: 9.894e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3649/30000 [8:12:35<58:08:04,  7.94s/it, lr: 1.0e-04 loss: 9.894e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3649/30000 [8:12:35<58:08:04,  7.94s/it, lr: 1.0e-04 loss: 9.894e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3649/30000 [8:12:44<58:08:04,  7.94s/it, lr: 1.0e-04 loss: 7.011e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3649/30000 [8:12:44<58:08:04,  7.94s/it, lr: 1.0e-04 loss: 7.011e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3650/30000 [8:12:54<58:07:56,  7.94s/it, lr: 1.0e-04 loss: 8.578e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3650/30000 [8:12:54<58:07:56,  7.94s/it, lr: 1.0e-04 loss: 8.578e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3651/30000 [8:12:54<47:06:29,  6.44s/it, lr: 1.0e-04 loss: 8.578e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3651/30000 [8:12:54<47:06:29,  6.44s/it, lr: 1.0e-04 loss: 8.578e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3651/30000 [8:13:00<47:06:29,  6.44s/it, lr: 1.0e-04 loss: 9.452e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3651/30000 [8:13:00<47:06:29,  6.44s/it, lr: 1.0e-04 loss: 9.452e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3652/30000 [8:13:00<47:20:16,  6.47s/it, lr: 1.0e-04 loss: 9.452e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3652/30000 [8:13:00<47:20:16,  6.47s/it, lr: 1.0e-04 loss: 9.452e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3652/30000 [8:13:07<47:20:16,  6.47s/it, lr: 1.0e-04 loss: 8.431e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3652/30000 [8:13:07<47:20:16,  6.47s/it, lr: 1.0e-04 loss: 8.431e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3653/30000 [8:13:07<47:27:17,  6.48s/it, lr: 1.0e-04 loss: 8.431e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3653/30000 [8:13:07<47:27:17,  6.48s/it, lr: 1.0e-04 loss: 8.431e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3653/30000 [8:13:13<47:27:17,  6.48s/it, lr: 1.0e-04 loss: 7.529e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3653/30000 [8:13:13<47:27:17,  6.48s/it, lr: 1.0e-04 loss: 7.529e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3654/30000 [8:13:13<47:41:35,  6.52s/it, lr: 1.0e-04 loss: 7.529e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3654/30000 [8:13:13<47:41:35,  6.52s/it, lr: 1.0e-04 loss: 7.529e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3654/30000 [8:13:20<47:41:35,  6.52s/it, lr: 1.0e-04 loss: 8.258e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3654/30000 [8:13:20<47:41:35,  6.52s/it, lr: 1.0e-04 loss: 8.258e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3655/30000 [8:13:20<47:51:27,  6.54s/it, lr: 1.0e-04 loss: 8.258e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3655/30000 [8:13:20<47:51:27,  6.54s/it, lr: 1.0e-04 loss: 8.258e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3655/30000 [8:13:29<47:51:27,  6.54s/it, lr: 1.0e-04 loss: 7.973e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3655/30000 [8:13:29<47:51:27,  6.54s/it, lr: 1.0e-04 loss: 7.973e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3656/30000 [8:13:29<53:33:38,  7.32s/it, lr: 1.0e-04 loss: 7.973e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3656/30000 [8:13:29<53:33:38,  7.32s/it, lr: 1.0e-04 loss: 7.973e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3656/30000 [8:13:36<53:33:38,  7.32s/it, lr: 1.0e-04 loss: 1.657e-01]
Xiang_wan_2_2_lora:  12%|####7                                  | 3656/30000 [8:13:36<53:33:38,  7.32s/it, lr: 1.0e-04 loss: 1.657e-01]
Xiang_wan_2_2_lora:  12%|####7                                  | 3657/30000 [8:13:36<52:03:51,  7.12s/it, lr: 1.0e-04 loss: 1.657e-01]
Xiang_wan_2_2_lora:  12%|####7                                  | 3657/30000 [8:13:36<52:03:51,  7.12s/it, lr: 1.0e-04 loss: 1.657e-01]
Xiang_wan_2_2_lora:  12%|####7                                  | 3657/30000 [8:13:43<52:03:51,  7.12s/it, lr: 1.0e-04 loss: 5.820e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3657/30000 [8:13:43<52:03:51,  7.12s/it, lr: 1.0e-04 loss: 5.820e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3658/30000 [8:13:43<51:21:53,  7.02s/it, lr: 1.0e-04 loss: 5.820e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3658/30000 [8:13:43<51:21:53,  7.02s/it, lr: 1.0e-04 loss: 5.820e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3658/30000 [8:13:52<51:21:53,  7.02s/it, lr: 1.0e-04 loss: 5.608e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3658/30000 [8:13:52<51:21:53,  7.02s/it, lr: 1.0e-04 loss: 5.608e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3659/30000 [8:13:52<56:13:11,  7.68s/it, lr: 1.0e-04 loss: 5.608e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3659/30000 [8:13:52<56:13:11,  7.68s/it, lr: 1.0e-04 loss: 5.608e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3659/30000 [8:13:58<56:13:11,  7.68s/it, lr: 1.0e-04 loss: 7.451e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3659/30000 [8:13:58<56:13:11,  7.68s/it, lr: 1.0e-04 loss: 7.451e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3660/30000 [8:14:08<56:13:03,  7.68s/it, lr: 1.0e-04 loss: 8.125e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3660/30000 [8:14:08<56:13:03,  7.68s/it, lr: 1.0e-04 loss: 8.125e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3661/30000 [8:14:08<45:58:40,  6.28s/it, lr: 1.0e-04 loss: 8.125e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3661/30000 [8:14:08<45:58:40,  6.28s/it, lr: 1.0e-04 loss: 8.125e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3661/30000 [8:14:14<45:58:40,  6.28s/it, lr: 1.0e-04 loss: 8.410e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3661/30000 [8:14:14<45:58:40,  6.28s/it, lr: 1.0e-04 loss: 8.410e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3662/30000 [8:14:14<46:38:23,  6.37s/it, lr: 1.0e-04 loss: 8.410e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3662/30000 [8:14:14<46:38:23,  6.37s/it, lr: 1.0e-04 loss: 8.410e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3662/30000 [8:14:24<46:38:23,  6.37s/it, lr: 1.0e-04 loss: 3.922e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3662/30000 [8:14:24<46:38:23,  6.37s/it, lr: 1.0e-04 loss: 3.922e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3663/30000 [8:14:24<52:07:34,  7.13s/it, lr: 1.0e-04 loss: 3.922e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3663/30000 [8:14:24<52:07:34,  7.13s/it, lr: 1.0e-04 loss: 3.922e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3663/30000 [8:14:30<52:07:34,  7.13s/it, lr: 1.0e-04 loss: 7.070e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3663/30000 [8:14:30<52:07:34,  7.13s/it, lr: 1.0e-04 loss: 7.070e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3664/30000 [8:14:30<51:09:17,  6.99s/it, lr: 1.0e-04 loss: 7.070e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3664/30000 [8:14:30<51:09:17,  6.99s/it, lr: 1.0e-04 loss: 7.070e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3664/30000 [8:14:37<51:09:17,  6.99s/it, lr: 1.0e-04 loss: 1.692e-01]
Xiang_wan_2_2_lora:  12%|####7                                  | 3664/30000 [8:14:37<51:09:17,  6.99s/it, lr: 1.0e-04 loss: 1.692e-01]
Xiang_wan_2_2_lora:  12%|####7                                  | 3665/30000 [8:14:37<50:45:40,  6.94s/it, lr: 1.0e-04 loss: 1.692e-01]
Xiang_wan_2_2_lora:  12%|####7                                  | 3665/30000 [8:14:37<50:45:40,  6.94s/it, lr: 1.0e-04 loss: 1.692e-01]
Xiang_wan_2_2_lora:  12%|####7                                  | 3665/30000 [8:14:46<50:45:40,  6.94s/it, lr: 1.0e-04 loss: 6.319e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3665/30000 [8:14:46<50:45:40,  6.94s/it, lr: 1.0e-04 loss: 6.319e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3666/30000 [8:14:46<55:43:07,  7.62s/it, lr: 1.0e-04 loss: 6.319e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3666/30000 [8:14:46<55:43:07,  7.62s/it, lr: 1.0e-04 loss: 6.319e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3666/30000 [8:14:53<55:43:07,  7.62s/it, lr: 1.0e-04 loss: 1.094e-01]
Xiang_wan_2_2_lora:  12%|####7                                  | 3666/30000 [8:14:53<55:43:07,  7.62s/it, lr: 1.0e-04 loss: 1.094e-01]
Xiang_wan_2_2_lora:  12%|####7                                  | 3667/30000 [8:14:53<53:24:01,  7.30s/it, lr: 1.0e-04 loss: 1.094e-01]
Xiang_wan_2_2_lora:  12%|####7                                  | 3667/30000 [8:14:53<53:24:01,  7.30s/it, lr: 1.0e-04 loss: 1.094e-01]
Xiang_wan_2_2_lora:  12%|####7                                  | 3667/30000 [8:15:02<53:24:01,  7.30s/it, lr: 1.0e-04 loss: 9.011e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3667/30000 [8:15:02<53:24:01,  7.30s/it, lr: 1.0e-04 loss: 9.011e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3668/30000 [8:15:02<57:39:23,  7.88s/it, lr: 1.0e-04 loss: 9.011e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3668/30000 [8:15:02<57:39:23,  7.88s/it, lr: 1.0e-04 loss: 9.011e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3668/30000 [8:15:12<57:39:23,  7.88s/it, lr: 1.0e-04 loss: 2.461e-01]
Xiang_wan_2_2_lora:  12%|####7                                  | 3668/30000 [8:15:12<57:39:23,  7.88s/it, lr: 1.0e-04 loss: 2.461e-01]
Xiang_wan_2_2_lora:  12%|####7                                  | 3669/30000 [8:15:12<60:42:16,  8.30s/it, lr: 1.0e-04 loss: 2.461e-01]
Xiang_wan_2_2_lora:  12%|####7                                  | 3669/30000 [8:15:12<60:42:16,  8.30s/it, lr: 1.0e-04 loss: 2.461e-01]
Xiang_wan_2_2_lora:  12%|####7                                  | 3669/30000 [8:15:21<60:42:16,  8.30s/it, lr: 1.0e-04 loss: 4.748e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3669/30000 [8:15:21<60:42:16,  8.30s/it, lr: 1.0e-04 loss: 4.748e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3670/30000 [8:15:27<60:42:07,  8.30s/it, lr: 1.0e-04 loss: 5.000e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3670/30000 [8:15:27<60:42:07,  8.30s/it, lr: 1.0e-04 loss: 5.000e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3671/30000 [8:15:27<43:58:01,  6.01s/it, lr: 1.0e-04 loss: 5.000e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3671/30000 [8:15:27<43:58:01,  6.01s/it, lr: 1.0e-04 loss: 5.000e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3671/30000 [8:15:37<43:58:01,  6.01s/it, lr: 1.0e-04 loss: 5.653e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3671/30000 [8:15:37<43:58:01,  6.01s/it, lr: 1.0e-04 loss: 5.653e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3672/30000 [8:15:37<49:55:33,  6.83s/it, lr: 1.0e-04 loss: 5.653e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3672/30000 [8:15:37<49:55:33,  6.83s/it, lr: 1.0e-04 loss: 5.653e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3672/30000 [8:15:46<49:55:33,  6.83s/it, lr: 1.0e-04 loss: 5.312e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3672/30000 [8:15:46<49:55:33,  6.83s/it, lr: 1.0e-04 loss: 5.312e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3673/30000 [8:15:46<55:02:03,  7.53s/it, lr: 1.0e-04 loss: 5.312e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3673/30000 [8:15:46<55:02:03,  7.53s/it, lr: 1.0e-04 loss: 5.312e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3673/30000 [8:15:53<55:02:03,  7.53s/it, lr: 1.0e-04 loss: 8.150e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3673/30000 [8:15:53<55:02:03,  7.53s/it, lr: 1.0e-04 loss: 8.150e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3674/30000 [8:15:53<53:07:50,  7.27s/it, lr: 1.0e-04 loss: 8.150e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3674/30000 [8:15:53<53:07:50,  7.27s/it, lr: 1.0e-04 loss: 8.150e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3674/30000 [8:15:59<53:07:50,  7.27s/it, lr: 1.0e-04 loss: 6.824e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3674/30000 [8:15:59<53:07:50,  7.27s/it, lr: 1.0e-04 loss: 6.824e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3675/30000 [8:15:59<51:34:22,  7.05s/it, lr: 1.0e-04 loss: 6.824e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3675/30000 [8:15:59<51:34:22,  7.05s/it, lr: 1.0e-04 loss: 6.824e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3675/30000 [8:16:06<51:34:22,  7.05s/it, lr: 1.0e-04 loss: 9.018e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3675/30000 [8:16:06<51:34:22,  7.05s/it, lr: 1.0e-04 loss: 9.018e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3676/30000 [8:16:06<50:35:05,  6.92s/it, lr: 1.0e-04 loss: 9.018e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3676/30000 [8:16:06<50:35:05,  6.92s/it, lr: 1.0e-04 loss: 9.018e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3676/30000 [8:16:15<50:35:05,  6.92s/it, lr: 1.0e-04 loss: 4.240e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3676/30000 [8:16:15<50:35:05,  6.92s/it, lr: 1.0e-04 loss: 4.240e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3677/30000 [8:16:15<55:32:43,  7.60s/it, lr: 1.0e-04 loss: 4.240e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3677/30000 [8:16:15<55:32:43,  7.60s/it, lr: 1.0e-04 loss: 4.240e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3677/30000 [8:16:25<55:32:43,  7.60s/it, lr: 1.0e-04 loss: 7.212e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3677/30000 [8:16:25<55:32:43,  7.60s/it, lr: 1.0e-04 loss: 7.212e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3678/30000 [8:16:25<59:12:58,  8.10s/it, lr: 1.0e-04 loss: 7.212e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3678/30000 [8:16:25<59:12:58,  8.10s/it, lr: 1.0e-04 loss: 7.212e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3678/30000 [8:16:34<59:12:58,  8.10s/it, lr: 1.0e-04 loss: 5.340e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3678/30000 [8:16:34<59:12:58,  8.10s/it, lr: 1.0e-04 loss: 5.340e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3679/30000 [8:16:34<61:51:57,  8.46s/it, lr: 1.0e-04 loss: 5.340e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3679/30000 [8:16:34<61:51:57,  8.46s/it, lr: 1.0e-04 loss: 5.340e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3679/30000 [8:16:41<61:51:57,  8.46s/it, lr: 1.0e-04 loss: 7.943e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3679/30000 [8:16:41<61:51:57,  8.46s/it, lr: 1.0e-04 loss: 7.943e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3680/30000 [8:16:50<61:51:48,  8.46s/it, lr: 1.0e-04 loss: 6.200e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3680/30000 [8:16:50<61:51:48,  8.46s/it, lr: 1.0e-04 loss: 6.200e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3681/30000 [8:16:50<49:07:50,  6.72s/it, lr: 1.0e-04 loss: 6.200e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3681/30000 [8:16:50<49:07:50,  6.72s/it, lr: 1.0e-04 loss: 6.200e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3681/30000 [8:16:57<49:07:50,  6.72s/it, lr: 1.0e-04 loss: 6.668e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3681/30000 [8:16:57<49:07:50,  6.72s/it, lr: 1.0e-04 loss: 6.668e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3682/30000 [8:16:57<48:51:01,  6.68s/it, lr: 1.0e-04 loss: 6.668e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3682/30000 [8:16:57<48:51:01,  6.68s/it, lr: 1.0e-04 loss: 6.668e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3682/30000 [8:17:06<48:51:01,  6.68s/it, lr: 1.0e-04 loss: 5.309e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3682/30000 [8:17:06<48:51:01,  6.68s/it, lr: 1.0e-04 loss: 5.309e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3683/30000 [8:17:06<53:52:03,  7.37s/it, lr: 1.0e-04 loss: 5.309e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3683/30000 [8:17:06<53:52:03,  7.37s/it, lr: 1.0e-04 loss: 5.309e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3683/30000 [8:17:15<53:52:03,  7.37s/it, lr: 1.0e-04 loss: 6.105e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3683/30000 [8:17:15<53:52:03,  7.37s/it, lr: 1.0e-04 loss: 6.105e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3684/30000 [8:17:15<57:54:38,  7.92s/it, lr: 1.0e-04 loss: 6.105e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3684/30000 [8:17:15<57:54:38,  7.92s/it, lr: 1.0e-04 loss: 6.105e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3684/30000 [8:17:25<57:54:38,  7.92s/it, lr: 1.0e-04 loss: 3.944e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3684/30000 [8:17:25<57:54:38,  7.92s/it, lr: 1.0e-04 loss: 3.944e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3685/30000 [8:17:25<60:45:44,  8.31s/it, lr: 1.0e-04 loss: 3.944e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3685/30000 [8:17:25<60:45:44,  8.31s/it, lr: 1.0e-04 loss: 3.944e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3685/30000 [8:17:31<60:45:44,  8.31s/it, lr: 1.0e-04 loss: 1.037e-01]
Xiang_wan_2_2_lora:  12%|####7                                  | 3685/30000 [8:17:31<60:45:44,  8.31s/it, lr: 1.0e-04 loss: 1.037e-01]
Xiang_wan_2_2_lora:  12%|####7                                  | 3686/30000 [8:17:31<57:14:18,  7.83s/it, lr: 1.0e-04 loss: 1.037e-01]
Xiang_wan_2_2_lora:  12%|####7                                  | 3686/30000 [8:17:31<57:14:18,  7.83s/it, lr: 1.0e-04 loss: 1.037e-01]
Xiang_wan_2_2_lora:  12%|####7                                  | 3686/30000 [8:17:41<57:14:18,  7.83s/it, lr: 1.0e-04 loss: 4.808e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3686/30000 [8:17:41<57:14:18,  7.83s/it, lr: 1.0e-04 loss: 4.808e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3687/30000 [8:17:41<60:26:53,  8.27s/it, lr: 1.0e-04 loss: 4.808e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3687/30000 [8:17:41<60:26:53,  8.27s/it, lr: 1.0e-04 loss: 4.808e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3687/30000 [8:17:50<60:26:53,  8.27s/it, lr: 1.0e-04 loss: 5.010e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3687/30000 [8:17:50<60:26:53,  8.27s/it, lr: 1.0e-04 loss: 5.010e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3688/30000 [8:17:50<63:28:14,  8.68s/it, lr: 1.0e-04 loss: 5.010e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3688/30000 [8:17:50<63:28:14,  8.68s/it, lr: 1.0e-04 loss: 5.010e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3688/30000 [8:18:00<63:28:14,  8.68s/it, lr: 1.0e-04 loss: 3.246e-01]
Xiang_wan_2_2_lora:  12%|####7                                  | 3688/30000 [8:18:00<63:28:14,  8.68s/it, lr: 1.0e-04 loss: 3.246e-01]
Xiang_wan_2_2_lora:  12%|####7                                  | 3689/30000 [8:18:00<64:49:13,  8.87s/it, lr: 1.0e-04 loss: 3.246e-01]
Xiang_wan_2_2_lora:  12%|####7                                  | 3689/30000 [8:18:00<64:49:13,  8.87s/it, lr: 1.0e-04 loss: 3.246e-01]
Xiang_wan_2_2_lora:  12%|####7                                  | 3689/30000 [8:18:09<64:49:13,  8.87s/it, lr: 1.0e-04 loss: 4.217e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3689/30000 [8:18:09<64:49:13,  8.87s/it, lr: 1.0e-04 loss: 4.217e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3690/30000 [8:18:16<64:49:04,  8.87s/it, lr: 1.0e-04 loss: 5.252e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3690/30000 [8:18:16<64:49:04,  8.87s/it, lr: 1.0e-04 loss: 5.252e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3691/30000 [8:18:16<46:08:30,  6.31s/it, lr: 1.0e-04 loss: 5.252e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3691/30000 [8:18:16<46:08:30,  6.31s/it, lr: 1.0e-04 loss: 5.252e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3691/30000 [8:18:25<46:08:30,  6.31s/it, lr: 1.0e-04 loss: 4.761e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3691/30000 [8:18:25<46:08:30,  6.31s/it, lr: 1.0e-04 loss: 4.761e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3692/30000 [8:18:25<51:43:57,  7.08s/it, lr: 1.0e-04 loss: 4.761e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3692/30000 [8:18:25<51:43:57,  7.08s/it, lr: 1.0e-04 loss: 4.761e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3692/30000 [8:18:34<51:43:57,  7.08s/it, lr: 1.0e-04 loss: 4.679e-02]
Xiang_wan_2_2_lora:  12%|####7                                  | 3692/30000 [8:18:34<51:43:57,  7.08s/it, lr: 1.0e-04 loss: 4.679e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3693/30000 [8:18:34<56:07:23,  7.68s/it, lr: 1.0e-04 loss: 4.679e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3693/30000 [8:18:34<56:07:23,  7.68s/it, lr: 1.0e-04 loss: 4.679e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3693/30000 [8:18:41<56:07:23,  7.68s/it, lr: 1.0e-04 loss: 8.281e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3693/30000 [8:18:41<56:07:23,  7.68s/it, lr: 1.0e-04 loss: 8.281e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3694/30000 [8:18:41<54:07:22,  7.41s/it, lr: 1.0e-04 loss: 8.281e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3694/30000 [8:18:41<54:07:22,  7.41s/it, lr: 1.0e-04 loss: 8.281e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3694/30000 [8:18:51<54:07:22,  7.41s/it, lr: 1.0e-04 loss: 4.723e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3694/30000 [8:18:51<54:07:22,  7.41s/it, lr: 1.0e-04 loss: 4.723e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3695/30000 [8:18:51<58:27:59,  8.00s/it, lr: 1.0e-04 loss: 4.723e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3695/30000 [8:18:51<58:27:59,  8.00s/it, lr: 1.0e-04 loss: 4.723e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3695/30000 [8:18:57<58:27:59,  8.00s/it, lr: 1.0e-04 loss: 1.068e-01]
Xiang_wan_2_2_lora:  12%|####8                                  | 3695/30000 [8:18:57<58:27:59,  8.00s/it, lr: 1.0e-04 loss: 1.068e-01]
Xiang_wan_2_2_lora:  12%|####8                                  | 3696/30000 [8:18:57<55:35:02,  7.61s/it, lr: 1.0e-04 loss: 1.068e-01]
Xiang_wan_2_2_lora:  12%|####8                                  | 3696/30000 [8:18:57<55:35:02,  7.61s/it, lr: 1.0e-04 loss: 1.068e-01]
Xiang_wan_2_2_lora:  12%|####8                                  | 3696/30000 [8:19:06<55:35:02,  7.61s/it, lr: 1.0e-04 loss: 1.091e-01]
Xiang_wan_2_2_lora:  12%|####8                                  | 3696/30000 [8:19:06<55:35:02,  7.61s/it, lr: 1.0e-04 loss: 1.091e-01]
Xiang_wan_2_2_lora:  12%|####8                                  | 3697/30000 [8:19:06<59:03:25,  8.08s/it, lr: 1.0e-04 loss: 1.091e-01]
Xiang_wan_2_2_lora:  12%|####8                                  | 3697/30000 [8:19:06<59:03:25,  8.08s/it, lr: 1.0e-04 loss: 1.091e-01]
Xiang_wan_2_2_lora:  12%|####8                                  | 3697/30000 [8:19:13<59:03:25,  8.08s/it, lr: 1.0e-04 loss: 7.326e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3697/30000 [8:19:13<59:03:25,  8.08s/it, lr: 1.0e-04 loss: 7.326e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3698/30000 [8:19:13<55:50:06,  7.64s/it, lr: 1.0e-04 loss: 7.326e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3698/30000 [8:19:13<55:50:06,  7.64s/it, lr: 1.0e-04 loss: 7.326e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3698/30000 [8:19:22<55:50:06,  7.64s/it, lr: 1.0e-04 loss: 4.683e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3698/30000 [8:19:22<55:50:06,  7.64s/it, lr: 1.0e-04 loss: 4.683e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3699/30000 [8:19:22<59:18:30,  8.12s/it, lr: 1.0e-04 loss: 4.683e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3699/30000 [8:19:22<59:18:30,  8.12s/it, lr: 1.0e-04 loss: 4.683e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3699/30000 [8:19:29<59:18:30,  8.12s/it, lr: 1.0e-04 loss: 7.657e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3699/30000 [8:19:29<59:18:30,  8.12s/it, lr: 1.0e-04 loss: 7.657e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3700/30000 [8:19:35<59:18:22,  8.12s/it, lr: 1.0e-04 loss: 4.812e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3700/30000 [8:19:35<59:18:22,  8.12s/it, lr: 1.0e-04 loss: 4.812e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3701/30000 [8:19:35<43:02:13,  5.89s/it, lr: 1.0e-04 loss: 4.812e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3701/30000 [8:19:35<43:02:13,  5.89s/it, lr: 1.0e-04 loss: 4.812e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3701/30000 [8:19:45<43:02:13,  5.89s/it, lr: 1.0e-04 loss: 6.645e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3701/30000 [8:19:45<43:02:13,  5.89s/it, lr: 1.0e-04 loss: 6.645e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3702/30000 [8:19:45<49:15:40,  6.74s/it, lr: 1.0e-04 loss: 6.645e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3702/30000 [8:19:45<49:15:40,  6.74s/it, lr: 1.0e-04 loss: 6.645e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3702/30000 [8:19:54<49:15:40,  6.74s/it, lr: 1.0e-04 loss: 1.515e-01]
Xiang_wan_2_2_lora:  12%|####8                                  | 3702/30000 [8:19:54<49:15:40,  6.74s/it, lr: 1.0e-04 loss: 1.515e-01]
Xiang_wan_2_2_lora:  12%|####8                                  | 3703/30000 [8:19:54<54:42:31,  7.49s/it, lr: 1.0e-04 loss: 1.515e-01]
Xiang_wan_2_2_lora:  12%|####8                                  | 3703/30000 [8:19:54<54:42:31,  7.49s/it, lr: 1.0e-04 loss: 1.515e-01]
Xiang_wan_2_2_lora:  12%|####8                                  | 3703/30000 [8:20:01<54:42:31,  7.49s/it, lr: 1.0e-04 loss: 7.379e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3703/30000 [8:20:01<54:42:31,  7.49s/it, lr: 1.0e-04 loss: 7.379e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3704/30000 [8:20:01<53:01:12,  7.26s/it, lr: 1.0e-04 loss: 7.379e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3704/30000 [8:20:01<53:01:12,  7.26s/it, lr: 1.0e-04 loss: 7.379e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3704/30000 [8:20:08<53:01:12,  7.26s/it, lr: 1.0e-04 loss: 6.101e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3704/30000 [8:20:08<53:01:12,  7.26s/it, lr: 1.0e-04 loss: 6.101e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3705/30000 [8:20:08<51:30:43,  7.05s/it, lr: 1.0e-04 loss: 6.101e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3705/30000 [8:20:08<51:30:43,  7.05s/it, lr: 1.0e-04 loss: 6.101e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3705/30000 [8:20:17<51:30:43,  7.05s/it, lr: 1.0e-04 loss: 9.210e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3705/30000 [8:20:17<51:30:43,  7.05s/it, lr: 1.0e-04 loss: 9.210e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3706/30000 [8:20:17<56:10:55,  7.69s/it, lr: 1.0e-04 loss: 9.210e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3706/30000 [8:20:17<56:10:55,  7.69s/it, lr: 1.0e-04 loss: 9.210e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3706/30000 [8:20:23<56:10:55,  7.69s/it, lr: 1.0e-04 loss: 7.207e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3706/30000 [8:20:23<56:10:55,  7.69s/it, lr: 1.0e-04 loss: 7.207e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3707/30000 [8:20:23<53:42:14,  7.35s/it, lr: 1.0e-04 loss: 7.207e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3707/30000 [8:20:23<53:42:14,  7.35s/it, lr: 1.0e-04 loss: 7.207e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3707/30000 [8:20:30<53:42:14,  7.35s/it, lr: 1.0e-04 loss: 9.203e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3707/30000 [8:20:30<53:42:14,  7.35s/it, lr: 1.0e-04 loss: 9.203e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3708/30000 [8:20:30<52:00:16,  7.12s/it, lr: 1.0e-04 loss: 9.203e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3708/30000 [8:20:30<52:00:16,  7.12s/it, lr: 1.0e-04 loss: 9.203e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3708/30000 [8:20:39<52:00:16,  7.12s/it, lr: 1.0e-04 loss: 4.677e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3708/30000 [8:20:39<52:00:16,  7.12s/it, lr: 1.0e-04 loss: 4.677e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3709/30000 [8:20:39<56:42:43,  7.77s/it, lr: 1.0e-04 loss: 4.677e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3709/30000 [8:20:39<56:42:43,  7.77s/it, lr: 1.0e-04 loss: 4.677e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3709/30000 [8:20:46<56:42:43,  7.77s/it, lr: 1.0e-04 loss: 9.142e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3709/30000 [8:20:46<56:42:43,  7.77s/it, lr: 1.0e-04 loss: 9.142e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3710/30000 [8:20:53<56:42:36,  7.77s/it, lr: 1.0e-04 loss: 6.055e-01]
Xiang_wan_2_2_lora:  12%|####8                                  | 3710/30000 [8:20:53<56:42:36,  7.77s/it, lr: 1.0e-04 loss: 6.055e-01]
Xiang_wan_2_2_lora:  12%|####8                                  | 3711/30000 [8:20:53<41:45:01,  5.72s/it, lr: 1.0e-04 loss: 6.055e-01]
Xiang_wan_2_2_lora:  12%|####8                                  | 3711/30000 [8:20:53<41:45:01,  5.72s/it, lr: 1.0e-04 loss: 6.055e-01]
Xiang_wan_2_2_lora:  12%|####8                                  | 3711/30000 [8:20:59<41:45:01,  5.72s/it, lr: 1.0e-04 loss: 1.139e-01]
Xiang_wan_2_2_lora:  12%|####8                                  | 3711/30000 [8:20:59<41:45:01,  5.72s/it, lr: 1.0e-04 loss: 1.139e-01]
Xiang_wan_2_2_lora:  12%|####8                                  | 3712/30000 [8:20:59<43:18:44,  5.93s/it, lr: 1.0e-04 loss: 1.139e-01]
Xiang_wan_2_2_lora:  12%|####8                                  | 3712/30000 [8:20:59<43:18:44,  5.93s/it, lr: 1.0e-04 loss: 1.139e-01]
Xiang_wan_2_2_lora:  12%|####8                                  | 3712/30000 [8:21:09<43:18:44,  5.93s/it, lr: 1.0e-04 loss: 5.598e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3712/30000 [8:21:09<43:18:44,  5.93s/it, lr: 1.0e-04 loss: 5.598e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3713/30000 [8:21:09<49:41:49,  6.81s/it, lr: 1.0e-04 loss: 5.598e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3713/30000 [8:21:09<49:41:49,  6.81s/it, lr: 1.0e-04 loss: 5.598e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3713/30000 [8:21:15<49:41:49,  6.81s/it, lr: 1.0e-04 loss: 6.065e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3713/30000 [8:21:15<49:41:49,  6.81s/it, lr: 1.0e-04 loss: 6.065e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3714/30000 [8:21:15<49:27:03,  6.77s/it, lr: 1.0e-04 loss: 6.065e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3714/30000 [8:21:15<49:27:03,  6.77s/it, lr: 1.0e-04 loss: 6.065e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3714/30000 [8:21:22<49:27:03,  6.77s/it, lr: 1.0e-04 loss: 1.682e-01]
Xiang_wan_2_2_lora:  12%|####8                                  | 3714/30000 [8:21:22<49:27:03,  6.77s/it, lr: 1.0e-04 loss: 1.682e-01]
Xiang_wan_2_2_lora:  12%|####8                                  | 3715/30000 [8:21:22<49:00:38,  6.71s/it, lr: 1.0e-04 loss: 1.682e-01]
Xiang_wan_2_2_lora:  12%|####8                                  | 3715/30000 [8:21:22<49:00:38,  6.71s/it, lr: 1.0e-04 loss: 1.682e-01]
Xiang_wan_2_2_lora:  12%|####8                                  | 3715/30000 [8:21:28<49:00:38,  6.71s/it, lr: 1.0e-04 loss: 7.387e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3715/30000 [8:21:28<49:00:38,  6.71s/it, lr: 1.0e-04 loss: 7.387e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3716/30000 [8:21:28<48:50:12,  6.69s/it, lr: 1.0e-04 loss: 7.387e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3716/30000 [8:21:28<48:50:12,  6.69s/it, lr: 1.0e-04 loss: 7.387e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3716/30000 [8:21:38<48:50:12,  6.69s/it, lr: 1.0e-04 loss: 3.600e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3716/30000 [8:21:38<48:50:12,  6.69s/it, lr: 1.0e-04 loss: 3.600e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3717/30000 [8:21:38<54:13:39,  7.43s/it, lr: 1.0e-04 loss: 3.600e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3717/30000 [8:21:38<54:13:39,  7.43s/it, lr: 1.0e-04 loss: 3.600e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3717/30000 [8:21:44<54:13:39,  7.43s/it, lr: 1.0e-04 loss: 9.142e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3717/30000 [8:21:44<54:13:39,  7.43s/it, lr: 1.0e-04 loss: 9.142e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3718/30000 [8:21:44<53:03:01,  7.27s/it, lr: 1.0e-04 loss: 9.142e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3718/30000 [8:21:44<53:03:01,  7.27s/it, lr: 1.0e-04 loss: 9.142e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3718/30000 [8:21:51<53:03:01,  7.27s/it, lr: 1.0e-04 loss: 6.025e-01]
Xiang_wan_2_2_lora:  12%|####8                                  | 3718/30000 [8:21:51<53:03:01,  7.27s/it, lr: 1.0e-04 loss: 6.025e-01]
Xiang_wan_2_2_lora:  12%|####8                                  | 3719/30000 [8:21:51<51:21:10,  7.03s/it, lr: 1.0e-04 loss: 6.025e-01]
Xiang_wan_2_2_lora:  12%|####8                                  | 3719/30000 [8:21:51<51:21:10,  7.03s/it, lr: 1.0e-04 loss: 6.025e-01]
Xiang_wan_2_2_lora:  12%|####8                                  | 3719/30000 [8:22:00<51:21:10,  7.03s/it, lr: 1.0e-04 loss: 6.508e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3719/30000 [8:22:00<51:21:10,  7.03s/it, lr: 1.0e-04 loss: 6.508e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3720/30000 [8:22:07<51:21:03,  7.03s/it, lr: 1.0e-04 loss: 5.809e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3720/30000 [8:22:07<51:21:03,  7.03s/it, lr: 1.0e-04 loss: 5.809e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3721/30000 [8:22:07<38:48:37,  5.32s/it, lr: 1.0e-04 loss: 5.809e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3721/30000 [8:22:07<38:48:37,  5.32s/it, lr: 1.0e-04 loss: 5.809e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3721/30000 [8:22:13<38:48:37,  5.32s/it, lr: 1.0e-04 loss: 7.840e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3721/30000 [8:22:13<38:48:37,  5.32s/it, lr: 1.0e-04 loss: 7.840e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3722/30000 [8:22:13<41:05:01,  5.63s/it, lr: 1.0e-04 loss: 7.840e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3722/30000 [8:22:13<41:05:01,  5.63s/it, lr: 1.0e-04 loss: 7.840e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3722/30000 [8:22:20<41:05:01,  5.63s/it, lr: 1.0e-04 loss: 1.145e-01]
Xiang_wan_2_2_lora:  12%|####8                                  | 3722/30000 [8:22:20<41:05:01,  5.63s/it, lr: 1.0e-04 loss: 1.145e-01]
Xiang_wan_2_2_lora:  12%|####8                                  | 3723/30000 [8:22:20<42:45:18,  5.86s/it, lr: 1.0e-04 loss: 1.145e-01]
Xiang_wan_2_2_lora:  12%|####8                                  | 3723/30000 [8:22:20<42:45:18,  5.86s/it, lr: 1.0e-04 loss: 1.145e-01]
Xiang_wan_2_2_lora:  12%|####8                                  | 3723/30000 [8:22:29<42:45:18,  5.86s/it, lr: 1.0e-04 loss: 7.649e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3723/30000 [8:22:29<42:45:18,  5.86s/it, lr: 1.0e-04 loss: 7.649e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3724/30000 [8:22:29<49:40:41,  6.81s/it, lr: 1.0e-04 loss: 7.649e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3724/30000 [8:22:29<49:40:41,  6.81s/it, lr: 1.0e-04 loss: 7.649e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3724/30000 [8:22:36<49:40:41,  6.81s/it, lr: 1.0e-04 loss: 6.785e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3724/30000 [8:22:36<49:40:41,  6.81s/it, lr: 1.0e-04 loss: 6.785e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3725/30000 [8:22:36<49:50:06,  6.83s/it, lr: 1.0e-04 loss: 6.785e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3725/30000 [8:22:36<49:50:06,  6.83s/it, lr: 1.0e-04 loss: 6.785e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3725/30000 [8:22:46<49:50:06,  6.83s/it, lr: 1.0e-04 loss: 7.426e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3725/30000 [8:22:46<49:50:06,  6.83s/it, lr: 1.0e-04 loss: 7.426e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3726/30000 [8:22:46<55:01:24,  7.54s/it, lr: 1.0e-04 loss: 7.426e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3726/30000 [8:22:46<55:01:24,  7.54s/it, lr: 1.0e-04 loss: 7.426e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3726/30000 [8:22:52<55:01:24,  7.54s/it, lr: 1.0e-04 loss: 6.800e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3726/30000 [8:22:52<55:01:24,  7.54s/it, lr: 1.0e-04 loss: 6.800e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3727/30000 [8:22:52<52:53:38,  7.25s/it, lr: 1.0e-04 loss: 6.800e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3727/30000 [8:22:52<52:53:38,  7.25s/it, lr: 1.0e-04 loss: 6.800e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3727/30000 [8:22:59<52:53:38,  7.25s/it, lr: 1.0e-04 loss: 1.194e-01]
Xiang_wan_2_2_lora:  12%|####8                                  | 3727/30000 [8:22:59<52:53:38,  7.25s/it, lr: 1.0e-04 loss: 1.194e-01]
Xiang_wan_2_2_lora:  12%|####8                                  | 3728/30000 [8:22:59<51:28:33,  7.05s/it, lr: 1.0e-04 loss: 1.194e-01]
Xiang_wan_2_2_lora:  12%|####8                                  | 3728/30000 [8:22:59<51:28:33,  7.05s/it, lr: 1.0e-04 loss: 1.194e-01]
Xiang_wan_2_2_lora:  12%|####8                                  | 3728/30000 [8:23:05<51:28:33,  7.05s/it, lr: 1.0e-04 loss: 6.227e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3728/30000 [8:23:05<51:28:33,  7.05s/it, lr: 1.0e-04 loss: 6.227e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3729/30000 [8:23:05<50:21:44,  6.90s/it, lr: 1.0e-04 loss: 6.227e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3729/30000 [8:23:05<50:21:44,  6.90s/it, lr: 1.0e-04 loss: 6.227e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3729/30000 [8:23:12<50:21:44,  6.90s/it, lr: 1.0e-04 loss: 8.386e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3729/30000 [8:23:12<50:21:44,  6.90s/it, lr: 1.0e-04 loss: 8.386e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3730/30000 [8:23:18<50:21:37,  6.90s/it, lr: 1.0e-04 loss: 1.142e-01]
Xiang_wan_2_2_lora:  12%|####8                                  | 3730/30000 [8:23:18<50:21:37,  6.90s/it, lr: 1.0e-04 loss: 1.142e-01]
Xiang_wan_2_2_lora:  12%|####8                                  | 3731/30000 [8:23:18<38:17:09,  5.25s/it, lr: 1.0e-04 loss: 1.142e-01]
Xiang_wan_2_2_lora:  12%|####8                                  | 3731/30000 [8:23:18<38:17:09,  5.25s/it, lr: 1.0e-04 loss: 1.142e-01]
Xiang_wan_2_2_lora:  12%|####8                                  | 3731/30000 [8:23:25<38:17:09,  5.25s/it, lr: 1.0e-04 loss: 6.069e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3731/30000 [8:23:25<38:17:09,  5.25s/it, lr: 1.0e-04 loss: 6.069e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3732/30000 [8:23:25<40:44:21,  5.58s/it, lr: 1.0e-04 loss: 6.069e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3732/30000 [8:23:25<40:44:21,  5.58s/it, lr: 1.0e-04 loss: 6.069e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3732/30000 [8:23:34<40:44:21,  5.58s/it, lr: 1.0e-04 loss: 5.122e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3732/30000 [8:23:34<40:44:21,  5.58s/it, lr: 1.0e-04 loss: 5.122e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3733/30000 [8:23:34<48:15:56,  6.62s/it, lr: 1.0e-04 loss: 5.122e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3733/30000 [8:23:34<48:15:56,  6.62s/it, lr: 1.0e-04 loss: 5.122e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3733/30000 [8:23:41<48:15:56,  6.62s/it, lr: 1.0e-04 loss: 6.332e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3733/30000 [8:23:41<48:15:56,  6.62s/it, lr: 1.0e-04 loss: 6.332e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3734/30000 [8:23:41<48:12:31,  6.61s/it, lr: 1.0e-04 loss: 6.332e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3734/30000 [8:23:41<48:12:31,  6.61s/it, lr: 1.0e-04 loss: 6.332e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3734/30000 [8:23:50<48:12:31,  6.61s/it, lr: 1.0e-04 loss: 4.764e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3734/30000 [8:23:50<48:12:31,  6.61s/it, lr: 1.0e-04 loss: 4.764e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3735/30000 [8:23:50<53:42:02,  7.36s/it, lr: 1.0e-04 loss: 4.764e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3735/30000 [8:23:50<53:42:02,  7.36s/it, lr: 1.0e-04 loss: 4.764e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3735/30000 [8:24:00<53:42:02,  7.36s/it, lr: 1.0e-04 loss: 7.809e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3735/30000 [8:24:00<53:42:02,  7.36s/it, lr: 1.0e-04 loss: 7.809e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3736/30000 [8:24:00<57:52:22,  7.93s/it, lr: 1.0e-04 loss: 7.809e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3736/30000 [8:24:00<57:52:22,  7.93s/it, lr: 1.0e-04 loss: 7.809e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3736/30000 [8:24:06<57:52:22,  7.93s/it, lr: 1.0e-04 loss: 5.825e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3736/30000 [8:24:06<57:52:22,  7.93s/it, lr: 1.0e-04 loss: 5.825e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3737/30000 [8:24:06<54:54:49,  7.53s/it, lr: 1.0e-04 loss: 5.825e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3737/30000 [8:24:06<54:54:49,  7.53s/it, lr: 1.0e-04 loss: 5.825e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3737/30000 [8:24:13<54:54:49,  7.53s/it, lr: 1.0e-04 loss: 5.304e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3737/30000 [8:24:13<54:54:49,  7.53s/it, lr: 1.0e-04 loss: 5.304e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3738/30000 [8:24:13<52:55:45,  7.26s/it, lr: 1.0e-04 loss: 5.304e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3738/30000 [8:24:13<52:55:45,  7.26s/it, lr: 1.0e-04 loss: 5.304e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3738/30000 [8:24:22<52:55:45,  7.26s/it, lr: 1.0e-04 loss: 1.264e-01]
Xiang_wan_2_2_lora:  12%|####8                                  | 3738/30000 [8:24:22<52:55:45,  7.26s/it, lr: 1.0e-04 loss: 1.264e-01]
Xiang_wan_2_2_lora:  12%|####8                                  | 3739/30000 [8:24:22<57:23:11,  7.87s/it, lr: 1.0e-04 loss: 1.264e-01]
Xiang_wan_2_2_lora:  12%|####8                                  | 3739/30000 [8:24:22<57:23:11,  7.87s/it, lr: 1.0e-04 loss: 1.264e-01]
Xiang_wan_2_2_lora:  12%|####8                                  | 3739/30000 [8:24:32<57:23:11,  7.87s/it, lr: 1.0e-04 loss: 6.431e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3739/30000 [8:24:32<57:23:11,  7.87s/it, lr: 1.0e-04 loss: 6.431e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3740/30000 [8:24:41<57:23:03,  7.87s/it, lr: 1.0e-04 loss: 4.496e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3740/30000 [8:24:41<57:23:03,  7.87s/it, lr: 1.0e-04 loss: 4.496e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3741/30000 [8:24:41<46:41:43,  6.40s/it, lr: 1.0e-04 loss: 4.496e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3741/30000 [8:24:41<46:41:43,  6.40s/it, lr: 1.0e-04 loss: 4.496e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3741/30000 [8:24:48<46:41:43,  6.40s/it, lr: 1.0e-04 loss: 1.142e-01]
Xiang_wan_2_2_lora:  12%|####8                                  | 3741/30000 [8:24:48<46:41:43,  6.40s/it, lr: 1.0e-04 loss: 1.142e-01]
Xiang_wan_2_2_lora:  12%|####8                                  | 3742/30000 [8:24:48<47:02:07,  6.45s/it, lr: 1.0e-04 loss: 1.142e-01]
Xiang_wan_2_2_lora:  12%|####8                                  | 3742/30000 [8:24:48<47:02:07,  6.45s/it, lr: 1.0e-04 loss: 1.142e-01]
Xiang_wan_2_2_lora:  12%|####8                                  | 3742/30000 [8:24:54<47:02:07,  6.45s/it, lr: 1.0e-04 loss: 7.557e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3742/30000 [8:24:54<47:02:07,  6.45s/it, lr: 1.0e-04 loss: 7.557e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3743/30000 [8:24:54<47:16:11,  6.48s/it, lr: 1.0e-04 loss: 7.557e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3743/30000 [8:24:54<47:16:11,  6.48s/it, lr: 1.0e-04 loss: 7.557e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3743/30000 [8:25:01<47:16:11,  6.48s/it, lr: 1.0e-04 loss: 7.209e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3743/30000 [8:25:01<47:16:11,  6.48s/it, lr: 1.0e-04 loss: 7.209e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3744/30000 [8:25:01<47:27:59,  6.51s/it, lr: 1.0e-04 loss: 7.209e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3744/30000 [8:25:01<47:27:59,  6.51s/it, lr: 1.0e-04 loss: 7.209e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3744/30000 [8:25:10<47:27:59,  6.51s/it, lr: 1.0e-04 loss: 4.340e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3744/30000 [8:25:10<47:27:59,  6.51s/it, lr: 1.0e-04 loss: 4.340e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3745/30000 [8:25:10<53:09:29,  7.29s/it, lr: 1.0e-04 loss: 4.340e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3745/30000 [8:25:10<53:09:29,  7.29s/it, lr: 1.0e-04 loss: 4.340e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3745/30000 [8:25:19<53:09:29,  7.29s/it, lr: 1.0e-04 loss: 1.052e-01]
Xiang_wan_2_2_lora:  12%|####8                                  | 3745/30000 [8:25:19<53:09:29,  7.29s/it, lr: 1.0e-04 loss: 1.052e-01]
Xiang_wan_2_2_lora:  12%|####8                                  | 3746/30000 [8:25:19<57:24:02,  7.87s/it, lr: 1.0e-04 loss: 1.052e-01]
Xiang_wan_2_2_lora:  12%|####8                                  | 3746/30000 [8:25:19<57:24:02,  7.87s/it, lr: 1.0e-04 loss: 1.052e-01]
Xiang_wan_2_2_lora:  12%|####8                                  | 3746/30000 [8:25:29<57:24:02,  7.87s/it, lr: 1.0e-04 loss: 5.324e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3746/30000 [8:25:29<57:24:02,  7.87s/it, lr: 1.0e-04 loss: 5.324e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3747/30000 [8:25:29<60:31:17,  8.30s/it, lr: 1.0e-04 loss: 5.324e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3747/30000 [8:25:29<60:31:17,  8.30s/it, lr: 1.0e-04 loss: 5.324e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3747/30000 [8:25:38<60:31:17,  8.30s/it, lr: 1.0e-04 loss: 6.152e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3747/30000 [8:25:38<60:31:17,  8.30s/it, lr: 1.0e-04 loss: 6.152e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3748/30000 [8:25:38<63:13:48,  8.67s/it, lr: 1.0e-04 loss: 6.152e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3748/30000 [8:25:38<63:13:48,  8.67s/it, lr: 1.0e-04 loss: 6.152e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3748/30000 [8:25:48<63:13:48,  8.67s/it, lr: 1.0e-04 loss: 4.207e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3748/30000 [8:25:48<63:13:48,  8.67s/it, lr: 1.0e-04 loss: 4.207e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3749/30000 [8:25:48<64:41:22,  8.87s/it, lr: 1.0e-04 loss: 4.207e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3749/30000 [8:25:48<64:41:22,  8.87s/it, lr: 1.0e-04 loss: 4.207e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3749/30000 [8:25:57<64:41:22,  8.87s/it, lr: 1.0e-04 loss: 5.258e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3749/30000 [8:25:57<64:41:22,  8.87s/it, lr: 1.0e-04 loss: 5.258e-02]
+Saving at step 3750
+Saved checkpoint to /home/featurize/ai-toolkit/output/Xiang_wan_2_2_lora/Xiang_wan_2_2_lora_000003750.safetensors
+Saved optimizer to /home/featurize/ai-toolkit/output/Xiang_wan_2_2_lora/optimizer.pt
+
Xiang_wan_2_2_lora:  12%|####8                                  | 3750/30000 [8:26:04<64:41:14,  8.87s/it, lr: 1.0e-04 loss: 4.808e-02]
Xiang_wan_2_2_lora:  12%|####8                                  | 3750/30000 [8:26:04<64:41:14,  8.87s/it, lr: 1.0e-04 loss: 4.808e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3751/30000 [8:26:04<46:30:45,  6.38s/it, lr: 1.0e-04 loss: 4.808e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3751/30000 [8:26:04<46:30:45,  6.38s/it, lr: 1.0e-04 loss: 4.808e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3751/30000 [8:26:10<46:30:45,  6.38s/it, lr: 1.0e-04 loss: 8.759e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3751/30000 [8:26:10<46:30:45,  6.38s/it, lr: 1.0e-04 loss: 8.759e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3752/30000 [8:26:10<46:43:11,  6.41s/it, lr: 1.0e-04 loss: 8.759e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3752/30000 [8:26:10<46:43:11,  6.41s/it, lr: 1.0e-04 loss: 8.759e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3752/30000 [8:26:20<46:43:11,  6.41s/it, lr: 1.0e-04 loss: 5.842e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3752/30000 [8:26:20<46:43:11,  6.41s/it, lr: 1.0e-04 loss: 5.842e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3753/30000 [8:26:20<52:10:37,  7.16s/it, lr: 1.0e-04 loss: 5.842e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3753/30000 [8:26:20<52:10:37,  7.16s/it, lr: 1.0e-04 loss: 5.842e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3753/30000 [8:26:26<52:10:37,  7.16s/it, lr: 1.0e-04 loss: 7.941e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3753/30000 [8:26:26<52:10:37,  7.16s/it, lr: 1.0e-04 loss: 7.941e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3754/30000 [8:26:26<50:56:13,  6.99s/it, lr: 1.0e-04 loss: 7.941e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3754/30000 [8:26:26<50:56:13,  6.99s/it, lr: 1.0e-04 loss: 7.941e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3754/30000 [8:26:35<50:56:13,  6.99s/it, lr: 1.0e-04 loss: 6.323e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3754/30000 [8:26:35<50:56:13,  6.99s/it, lr: 1.0e-04 loss: 6.323e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3755/30000 [8:26:35<55:42:44,  7.64s/it, lr: 1.0e-04 loss: 6.323e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3755/30000 [8:26:35<55:42:44,  7.64s/it, lr: 1.0e-04 loss: 6.323e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3755/30000 [8:26:42<55:42:44,  7.64s/it, lr: 1.0e-04 loss: 7.996e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3755/30000 [8:26:42<55:42:44,  7.64s/it, lr: 1.0e-04 loss: 7.996e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3756/30000 [8:26:42<53:32:58,  7.35s/it, lr: 1.0e-04 loss: 7.996e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3756/30000 [8:26:42<53:32:58,  7.35s/it, lr: 1.0e-04 loss: 7.996e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3756/30000 [8:26:51<53:32:58,  7.35s/it, lr: 1.0e-04 loss: 4.105e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3756/30000 [8:26:51<53:32:58,  7.35s/it, lr: 1.0e-04 loss: 4.105e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3757/30000 [8:26:51<57:38:20,  7.91s/it, lr: 1.0e-04 loss: 4.105e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3757/30000 [8:26:51<57:38:20,  7.91s/it, lr: 1.0e-04 loss: 4.105e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3757/30000 [8:26:58<57:38:20,  7.91s/it, lr: 1.0e-04 loss: 8.601e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3757/30000 [8:26:58<57:38:20,  7.91s/it, lr: 1.0e-04 loss: 8.601e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3758/30000 [8:26:58<55:21:31,  7.59s/it, lr: 1.0e-04 loss: 8.601e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3758/30000 [8:26:58<55:21:31,  7.59s/it, lr: 1.0e-04 loss: 8.601e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3758/30000 [8:27:05<55:21:31,  7.59s/it, lr: 1.0e-04 loss: 6.032e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3758/30000 [8:27:05<55:21:31,  7.59s/it, lr: 1.0e-04 loss: 6.032e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3759/30000 [8:27:05<53:01:55,  7.28s/it, lr: 1.0e-04 loss: 6.032e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3759/30000 [8:27:05<53:01:55,  7.28s/it, lr: 1.0e-04 loss: 6.032e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3759/30000 [8:27:14<53:01:55,  7.28s/it, lr: 1.0e-04 loss: 6.404e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3759/30000 [8:27:14<53:01:55,  7.28s/it, lr: 1.0e-04 loss: 6.404e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3760/30000 [8:27:21<53:01:48,  7.28s/it, lr: 1.0e-04 loss: 8.884e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3760/30000 [8:27:21<53:01:48,  7.28s/it, lr: 1.0e-04 loss: 8.884e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3761/30000 [8:27:21<39:44:32,  5.45s/it, lr: 1.0e-04 loss: 8.884e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3761/30000 [8:27:21<39:44:32,  5.45s/it, lr: 1.0e-04 loss: 8.884e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3761/30000 [8:27:27<39:44:32,  5.45s/it, lr: 1.0e-04 loss: 1.034e-01]
Xiang_wan_2_2_lora:  13%|####8                                  | 3761/30000 [8:27:27<39:44:32,  5.45s/it, lr: 1.0e-04 loss: 1.034e-01]
Xiang_wan_2_2_lora:  13%|####8                                  | 3762/30000 [8:27:27<41:43:09,  5.72s/it, lr: 1.0e-04 loss: 1.034e-01]
Xiang_wan_2_2_lora:  13%|####8                                  | 3762/30000 [8:27:27<41:43:09,  5.72s/it, lr: 1.0e-04 loss: 1.034e-01]
Xiang_wan_2_2_lora:  13%|####8                                  | 3762/30000 [8:27:36<41:43:09,  5.72s/it, lr: 1.0e-04 loss: 9.553e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3762/30000 [8:27:36<41:43:09,  5.72s/it, lr: 1.0e-04 loss: 9.553e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3763/30000 [8:27:36<48:36:04,  6.67s/it, lr: 1.0e-04 loss: 9.553e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3763/30000 [8:27:36<48:36:04,  6.67s/it, lr: 1.0e-04 loss: 9.553e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3763/30000 [8:27:46<48:36:04,  6.67s/it, lr: 1.0e-04 loss: 6.978e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3763/30000 [8:27:46<48:36:04,  6.67s/it, lr: 1.0e-04 loss: 6.978e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3764/30000 [8:27:46<53:50:05,  7.39s/it, lr: 1.0e-04 loss: 6.978e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3764/30000 [8:27:46<53:50:05,  7.39s/it, lr: 1.0e-04 loss: 6.978e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3764/30000 [8:27:53<53:50:05,  7.39s/it, lr: 1.0e-04 loss: 8.529e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3764/30000 [8:27:53<53:50:05,  7.39s/it, lr: 1.0e-04 loss: 8.529e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3765/30000 [8:27:53<52:40:29,  7.23s/it, lr: 1.0e-04 loss: 8.529e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3765/30000 [8:27:53<52:40:29,  7.23s/it, lr: 1.0e-04 loss: 8.529e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3765/30000 [8:27:59<52:40:29,  7.23s/it, lr: 1.0e-04 loss: 9.396e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3765/30000 [8:27:59<52:40:29,  7.23s/it, lr: 1.0e-04 loss: 9.396e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3766/30000 [8:27:59<51:15:04,  7.03s/it, lr: 1.0e-04 loss: 9.396e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3766/30000 [8:27:59<51:15:04,  7.03s/it, lr: 1.0e-04 loss: 9.396e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3766/30000 [8:28:09<51:15:04,  7.03s/it, lr: 1.0e-04 loss: 5.594e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3766/30000 [8:28:09<51:15:04,  7.03s/it, lr: 1.0e-04 loss: 5.594e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3767/30000 [8:28:09<56:14:06,  7.72s/it, lr: 1.0e-04 loss: 5.594e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3767/30000 [8:28:09<56:14:06,  7.72s/it, lr: 1.0e-04 loss: 5.594e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3767/30000 [8:28:18<56:14:06,  7.72s/it, lr: 1.0e-04 loss: 6.137e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3767/30000 [8:28:18<56:14:06,  7.72s/it, lr: 1.0e-04 loss: 6.137e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3768/30000 [8:28:18<59:41:01,  8.19s/it, lr: 1.0e-04 loss: 6.137e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3768/30000 [8:28:18<59:41:01,  8.19s/it, lr: 1.0e-04 loss: 6.137e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3768/30000 [8:28:27<59:41:01,  8.19s/it, lr: 1.0e-04 loss: 8.557e-01]
Xiang_wan_2_2_lora:  13%|####8                                  | 3768/30000 [8:28:27<59:41:01,  8.19s/it, lr: 1.0e-04 loss: 8.557e-01]
Xiang_wan_2_2_lora:  13%|####8                                  | 3769/30000 [8:28:27<62:04:46,  8.52s/it, lr: 1.0e-04 loss: 8.557e-01]
Xiang_wan_2_2_lora:  13%|####8                                  | 3769/30000 [8:28:27<62:04:46,  8.52s/it, lr: 1.0e-04 loss: 8.557e-01]
Xiang_wan_2_2_lora:  13%|####8                                  | 3769/30000 [8:28:37<62:04:46,  8.52s/it, lr: 1.0e-04 loss: 5.669e-02]
Xiang_wan_2_2_lora:  13%|####8                                  | 3769/30000 [8:28:37<62:04:46,  8.52s/it, lr: 1.0e-04 loss: 5.669e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3770/30000 [8:28:43<62:04:38,  8.52s/it, lr: 1.0e-04 loss: 8.373e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3770/30000 [8:28:43<62:04:38,  8.52s/it, lr: 1.0e-04 loss: 8.373e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3771/30000 [8:28:43<44:40:37,  6.13s/it, lr: 1.0e-04 loss: 8.373e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3771/30000 [8:28:43<44:40:37,  6.13s/it, lr: 1.0e-04 loss: 8.373e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3771/30000 [8:28:50<44:40:37,  6.13s/it, lr: 1.0e-04 loss: 6.534e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3771/30000 [8:28:50<44:40:37,  6.13s/it, lr: 1.0e-04 loss: 6.534e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3772/30000 [8:28:50<45:30:15,  6.25s/it, lr: 1.0e-04 loss: 6.534e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3772/30000 [8:28:50<45:30:15,  6.25s/it, lr: 1.0e-04 loss: 6.534e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3772/30000 [8:28:57<45:30:15,  6.25s/it, lr: 1.0e-04 loss: 1.231e-01]
Xiang_wan_2_2_lora:  13%|####9                                  | 3772/30000 [8:28:57<45:30:15,  6.25s/it, lr: 1.0e-04 loss: 1.231e-01]
Xiang_wan_2_2_lora:  13%|####9                                  | 3773/30000 [8:28:57<46:39:35,  6.40s/it, lr: 1.0e-04 loss: 1.231e-01]
Xiang_wan_2_2_lora:  13%|####9                                  | 3773/30000 [8:28:57<46:39:35,  6.40s/it, lr: 1.0e-04 loss: 1.231e-01]
Xiang_wan_2_2_lora:  13%|####9                                  | 3773/30000 [8:29:03<46:39:35,  6.40s/it, lr: 1.0e-04 loss: 6.643e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3773/30000 [8:29:03<46:39:35,  6.40s/it, lr: 1.0e-04 loss: 6.643e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3774/30000 [8:29:03<46:48:34,  6.43s/it, lr: 1.0e-04 loss: 6.643e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3774/30000 [8:29:03<46:48:34,  6.43s/it, lr: 1.0e-04 loss: 6.643e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3774/30000 [8:29:12<46:48:34,  6.43s/it, lr: 1.0e-04 loss: 4.232e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3774/30000 [8:29:12<46:48:34,  6.43s/it, lr: 1.0e-04 loss: 4.232e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3775/30000 [8:29:12<52:45:57,  7.24s/it, lr: 1.0e-04 loss: 4.232e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3775/30000 [8:29:12<52:45:57,  7.24s/it, lr: 1.0e-04 loss: 4.232e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3775/30000 [8:29:19<52:45:57,  7.24s/it, lr: 1.0e-04 loss: 8.775e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3775/30000 [8:29:19<52:45:57,  7.24s/it, lr: 1.0e-04 loss: 8.775e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3776/30000 [8:29:19<51:24:07,  7.06s/it, lr: 1.0e-04 loss: 8.775e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3776/30000 [8:29:19<51:24:07,  7.06s/it, lr: 1.0e-04 loss: 8.775e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3776/30000 [8:29:25<51:24:07,  7.06s/it, lr: 1.0e-04 loss: 2.809e-01]
Xiang_wan_2_2_lora:  13%|####9                                  | 3776/30000 [8:29:25<51:24:07,  7.06s/it, lr: 1.0e-04 loss: 2.809e-01]
Xiang_wan_2_2_lora:  13%|####9                                  | 3777/30000 [8:29:25<50:16:39,  6.90s/it, lr: 1.0e-04 loss: 2.809e-01]
Xiang_wan_2_2_lora:  13%|####9                                  | 3777/30000 [8:29:25<50:16:39,  6.90s/it, lr: 1.0e-04 loss: 2.809e-01]
Xiang_wan_2_2_lora:  13%|####9                                  | 3777/30000 [8:29:32<50:16:39,  6.90s/it, lr: 1.0e-04 loss: 9.697e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3777/30000 [8:29:32<50:16:39,  6.90s/it, lr: 1.0e-04 loss: 9.697e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3778/30000 [8:29:32<49:26:25,  6.79s/it, lr: 1.0e-04 loss: 9.697e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3778/30000 [8:29:32<49:26:25,  6.79s/it, lr: 1.0e-04 loss: 9.697e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3778/30000 [8:29:39<49:26:25,  6.79s/it, lr: 1.0e-04 loss: 9.841e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3778/30000 [8:29:39<49:26:25,  6.79s/it, lr: 1.0e-04 loss: 9.841e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3779/30000 [8:29:39<49:05:48,  6.74s/it, lr: 1.0e-04 loss: 9.841e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3779/30000 [8:29:39<49:05:48,  6.74s/it, lr: 1.0e-04 loss: 9.841e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3779/30000 [8:29:45<49:05:48,  6.74s/it, lr: 1.0e-04 loss: 6.360e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3779/30000 [8:29:45<49:05:48,  6.74s/it, lr: 1.0e-04 loss: 6.360e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3780/30000 [8:29:55<49:05:41,  6.74s/it, lr: 1.0e-04 loss: 2.207e-01]
Xiang_wan_2_2_lora:  13%|####9                                  | 3780/30000 [8:29:55<49:05:41,  6.74s/it, lr: 1.0e-04 loss: 2.207e-01]
Xiang_wan_2_2_lora:  13%|####9                                  | 3781/30000 [8:29:55<42:08:31,  5.79s/it, lr: 1.0e-04 loss: 2.207e-01]
Xiang_wan_2_2_lora:  13%|####9                                  | 3781/30000 [8:29:55<42:08:31,  5.79s/it, lr: 1.0e-04 loss: 2.207e-01]
Xiang_wan_2_2_lora:  13%|####9                                  | 3781/30000 [8:30:04<42:08:31,  5.79s/it, lr: 1.0e-04 loss: 8.110e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3781/30000 [8:30:04<42:08:31,  5.79s/it, lr: 1.0e-04 loss: 8.110e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3782/30000 [8:30:04<48:30:00,  6.66s/it, lr: 1.0e-04 loss: 8.110e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3782/30000 [8:30:04<48:30:00,  6.66s/it, lr: 1.0e-04 loss: 8.110e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3782/30000 [8:30:13<48:30:00,  6.66s/it, lr: 1.0e-04 loss: 5.271e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3782/30000 [8:30:13<48:30:00,  6.66s/it, lr: 1.0e-04 loss: 5.271e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3783/30000 [8:30:13<53:35:58,  7.36s/it, lr: 1.0e-04 loss: 5.271e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3783/30000 [8:30:13<53:35:58,  7.36s/it, lr: 1.0e-04 loss: 5.271e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3783/30000 [8:30:20<53:35:58,  7.36s/it, lr: 1.0e-04 loss: 9.355e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3783/30000 [8:30:20<53:35:58,  7.36s/it, lr: 1.0e-04 loss: 9.355e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3784/30000 [8:30:20<52:02:43,  7.15s/it, lr: 1.0e-04 loss: 9.355e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3784/30000 [8:30:20<52:02:43,  7.15s/it, lr: 1.0e-04 loss: 9.355e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3784/30000 [8:30:27<52:02:43,  7.15s/it, lr: 1.0e-04 loss: 8.342e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3784/30000 [8:30:27<52:02:43,  7.15s/it, lr: 1.0e-04 loss: 8.342e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3785/30000 [8:30:27<50:59:37,  7.00s/it, lr: 1.0e-04 loss: 8.342e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3785/30000 [8:30:27<50:59:37,  7.00s/it, lr: 1.0e-04 loss: 8.342e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3785/30000 [8:30:36<50:59:37,  7.00s/it, lr: 1.0e-04 loss: 5.592e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3785/30000 [8:30:36<50:59:37,  7.00s/it, lr: 1.0e-04 loss: 5.592e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3786/30000 [8:30:36<55:46:11,  7.66s/it, lr: 1.0e-04 loss: 5.592e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3786/30000 [8:30:36<55:46:11,  7.66s/it, lr: 1.0e-04 loss: 5.592e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3786/30000 [8:30:45<55:46:11,  7.66s/it, lr: 1.0e-04 loss: 7.057e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3786/30000 [8:30:45<55:46:11,  7.66s/it, lr: 1.0e-04 loss: 7.057e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3787/30000 [8:30:45<59:10:40,  8.13s/it, lr: 1.0e-04 loss: 7.057e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3787/30000 [8:30:45<59:10:40,  8.13s/it, lr: 1.0e-04 loss: 7.057e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3787/30000 [8:30:55<59:10:40,  8.13s/it, lr: 1.0e-04 loss: 4.741e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3787/30000 [8:30:55<59:10:40,  8.13s/it, lr: 1.0e-04 loss: 4.741e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3788/30000 [8:30:55<62:03:26,  8.52s/it, lr: 1.0e-04 loss: 4.741e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3788/30000 [8:30:55<62:03:26,  8.52s/it, lr: 1.0e-04 loss: 4.741e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3788/30000 [8:31:01<62:03:26,  8.52s/it, lr: 1.0e-04 loss: 5.294e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3788/30000 [8:31:01<62:03:26,  8.52s/it, lr: 1.0e-04 loss: 5.294e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3789/30000 [8:31:01<57:47:34,  7.94s/it, lr: 1.0e-04 loss: 5.294e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3789/30000 [8:31:01<57:47:34,  7.94s/it, lr: 1.0e-04 loss: 5.294e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3789/30000 [8:31:08<57:47:34,  7.94s/it, lr: 1.0e-04 loss: 6.233e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3789/30000 [8:31:08<57:47:34,  7.94s/it, lr: 1.0e-04 loss: 6.233e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3790/30000 [8:31:17<57:47:26,  7.94s/it, lr: 1.0e-04 loss: 5.857e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3790/30000 [8:31:17<57:47:26,  7.94s/it, lr: 1.0e-04 loss: 5.857e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3791/30000 [8:31:17<46:53:34,  6.44s/it, lr: 1.0e-04 loss: 5.857e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3791/30000 [8:31:17<46:53:34,  6.44s/it, lr: 1.0e-04 loss: 5.857e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3791/30000 [8:31:24<46:53:34,  6.44s/it, lr: 1.0e-04 loss: 6.140e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3791/30000 [8:31:24<46:53:34,  6.44s/it, lr: 1.0e-04 loss: 6.140e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3792/30000 [8:31:24<47:09:40,  6.48s/it, lr: 1.0e-04 loss: 6.140e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3792/30000 [8:31:24<47:09:40,  6.48s/it, lr: 1.0e-04 loss: 6.140e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3792/30000 [8:31:30<47:09:40,  6.48s/it, lr: 1.0e-04 loss: 7.590e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3792/30000 [8:31:30<47:09:40,  6.48s/it, lr: 1.0e-04 loss: 7.590e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3793/30000 [8:31:30<47:20:12,  6.50s/it, lr: 1.0e-04 loss: 7.590e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3793/30000 [8:31:30<47:20:12,  6.50s/it, lr: 1.0e-04 loss: 7.590e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3793/30000 [8:31:37<47:20:12,  6.50s/it, lr: 1.0e-04 loss: 6.897e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3793/30000 [8:31:37<47:20:12,  6.50s/it, lr: 1.0e-04 loss: 6.897e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3794/30000 [8:31:37<47:28:55,  6.52s/it, lr: 1.0e-04 loss: 6.897e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3794/30000 [8:31:37<47:28:55,  6.52s/it, lr: 1.0e-04 loss: 6.897e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3794/30000 [8:31:46<47:28:55,  6.52s/it, lr: 1.0e-04 loss: 3.576e-01]
Xiang_wan_2_2_lora:  13%|####9                                  | 3794/30000 [8:31:46<47:28:55,  6.52s/it, lr: 1.0e-04 loss: 3.576e-01]
Xiang_wan_2_2_lora:  13%|####9                                  | 3795/30000 [8:31:46<53:31:16,  7.35s/it, lr: 1.0e-04 loss: 3.576e-01]
Xiang_wan_2_2_lora:  13%|####9                                  | 3795/30000 [8:31:46<53:31:16,  7.35s/it, lr: 1.0e-04 loss: 3.576e-01]
Xiang_wan_2_2_lora:  13%|####9                                  | 3795/30000 [8:31:53<53:31:16,  7.35s/it, lr: 1.0e-04 loss: 1.177e-01]
Xiang_wan_2_2_lora:  13%|####9                                  | 3795/30000 [8:31:53<53:31:16,  7.35s/it, lr: 1.0e-04 loss: 1.177e-01]
Xiang_wan_2_2_lora:  13%|####9                                  | 3796/30000 [8:31:53<51:54:42,  7.13s/it, lr: 1.0e-04 loss: 1.177e-01]
Xiang_wan_2_2_lora:  13%|####9                                  | 3796/30000 [8:31:53<51:54:42,  7.13s/it, lr: 1.0e-04 loss: 1.177e-01]
Xiang_wan_2_2_lora:  13%|####9                                  | 3796/30000 [8:32:00<51:54:42,  7.13s/it, lr: 1.0e-04 loss: 5.643e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3796/30000 [8:32:00<51:54:42,  7.13s/it, lr: 1.0e-04 loss: 5.643e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3797/30000 [8:32:00<50:46:20,  6.98s/it, lr: 1.0e-04 loss: 5.643e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3797/30000 [8:32:00<50:46:20,  6.98s/it, lr: 1.0e-04 loss: 5.643e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3797/30000 [8:32:09<50:46:20,  6.98s/it, lr: 1.0e-04 loss: 7.596e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3797/30000 [8:32:09<50:46:20,  6.98s/it, lr: 1.0e-04 loss: 7.596e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3798/30000 [8:32:09<55:42:03,  7.65s/it, lr: 1.0e-04 loss: 7.596e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3798/30000 [8:32:09<55:42:03,  7.65s/it, lr: 1.0e-04 loss: 7.596e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3798/30000 [8:32:18<55:42:03,  7.65s/it, lr: 1.0e-04 loss: 7.214e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3798/30000 [8:32:18<55:42:03,  7.65s/it, lr: 1.0e-04 loss: 7.214e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3799/30000 [8:32:18<59:13:23,  8.14s/it, lr: 1.0e-04 loss: 7.214e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3799/30000 [8:32:18<59:13:23,  8.14s/it, lr: 1.0e-04 loss: 7.214e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3799/30000 [8:32:27<59:13:23,  8.14s/it, lr: 1.0e-04 loss: 4.010e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3799/30000 [8:32:27<59:13:23,  8.14s/it, lr: 1.0e-04 loss: 4.010e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3800/30000 [8:32:34<59:13:14,  8.14s/it, lr: 1.0e-04 loss: 4.208e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3800/30000 [8:32:34<59:13:14,  8.14s/it, lr: 1.0e-04 loss: 4.208e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3801/30000 [8:32:34<43:04:26,  5.92s/it, lr: 1.0e-04 loss: 4.208e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3801/30000 [8:32:34<43:04:26,  5.92s/it, lr: 1.0e-04 loss: 4.208e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3801/30000 [8:32:41<43:04:26,  5.92s/it, lr: 1.0e-04 loss: 8.875e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3801/30000 [8:32:41<43:04:26,  5.92s/it, lr: 1.0e-04 loss: 8.875e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3802/30000 [8:32:41<44:15:31,  6.08s/it, lr: 1.0e-04 loss: 8.875e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3802/30000 [8:32:41<44:15:31,  6.08s/it, lr: 1.0e-04 loss: 8.875e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3802/30000 [8:32:50<44:15:31,  6.08s/it, lr: 1.0e-04 loss: 1.924e-01]
Xiang_wan_2_2_lora:  13%|####9                                  | 3802/30000 [8:32:50<44:15:31,  6.08s/it, lr: 1.0e-04 loss: 1.924e-01]
Xiang_wan_2_2_lora:  13%|####9                                  | 3803/30000 [8:32:50<50:44:06,  6.97s/it, lr: 1.0e-04 loss: 1.924e-01]
Xiang_wan_2_2_lora:  13%|####9                                  | 3803/30000 [8:32:50<50:44:06,  6.97s/it, lr: 1.0e-04 loss: 1.924e-01]
Xiang_wan_2_2_lora:  13%|####9                                  | 3803/30000 [8:32:59<50:44:06,  6.97s/it, lr: 1.0e-04 loss: 6.766e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3803/30000 [8:32:59<50:44:06,  6.97s/it, lr: 1.0e-04 loss: 6.766e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3804/30000 [8:32:59<55:24:09,  7.61s/it, lr: 1.0e-04 loss: 6.766e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3804/30000 [8:32:59<55:24:09,  7.61s/it, lr: 1.0e-04 loss: 6.766e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3804/30000 [8:33:09<55:24:09,  7.61s/it, lr: 1.0e-04 loss: 8.133e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3804/30000 [8:33:09<55:24:09,  7.61s/it, lr: 1.0e-04 loss: 8.133e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3805/30000 [8:33:09<58:47:19,  8.08s/it, lr: 1.0e-04 loss: 8.133e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3805/30000 [8:33:09<58:47:19,  8.08s/it, lr: 1.0e-04 loss: 8.133e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3805/30000 [8:33:15<58:47:19,  8.08s/it, lr: 1.0e-04 loss: 7.976e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3805/30000 [8:33:15<58:47:19,  8.08s/it, lr: 1.0e-04 loss: 7.976e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3806/30000 [8:33:15<55:47:13,  7.67s/it, lr: 1.0e-04 loss: 7.976e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3806/30000 [8:33:15<55:47:13,  7.67s/it, lr: 1.0e-04 loss: 7.976e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3806/30000 [8:33:25<55:47:13,  7.67s/it, lr: 1.0e-04 loss: 4.101e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3806/30000 [8:33:25<55:47:13,  7.67s/it, lr: 1.0e-04 loss: 4.101e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3807/30000 [8:33:25<59:11:57,  8.14s/it, lr: 1.0e-04 loss: 4.101e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3807/30000 [8:33:25<59:11:57,  8.14s/it, lr: 1.0e-04 loss: 4.101e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3807/30000 [8:33:34<59:11:57,  8.14s/it, lr: 1.0e-04 loss: 8.457e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3807/30000 [8:33:34<59:11:57,  8.14s/it, lr: 1.0e-04 loss: 8.457e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3808/30000 [8:33:34<62:07:58,  8.54s/it, lr: 1.0e-04 loss: 8.457e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3808/30000 [8:33:34<62:07:58,  8.54s/it, lr: 1.0e-04 loss: 8.457e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3808/30000 [8:33:44<62:07:58,  8.54s/it, lr: 1.0e-04 loss: 7.839e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3808/30000 [8:33:44<62:07:58,  8.54s/it, lr: 1.0e-04 loss: 7.839e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3809/30000 [8:33:44<63:46:27,  8.77s/it, lr: 1.0e-04 loss: 7.839e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3809/30000 [8:33:44<63:46:27,  8.77s/it, lr: 1.0e-04 loss: 7.839e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3809/30000 [8:33:50<63:46:27,  8.77s/it, lr: 1.0e-04 loss: 1.137e-01]
Xiang_wan_2_2_lora:  13%|####9                                  | 3809/30000 [8:33:50<63:46:27,  8.77s/it, lr: 1.0e-04 loss: 1.137e-01]
Xiang_wan_2_2_lora:  13%|####9                                  | 3810/30000 [8:33:58<63:46:18,  8.77s/it, lr: 1.0e-04 loss: 4.794e-01]
Xiang_wan_2_2_lora:  13%|####9                                  | 3810/30000 [8:33:58<63:46:18,  8.77s/it, lr: 1.0e-04 loss: 4.794e-01]
Xiang_wan_2_2_lora:  13%|####9                                  | 3811/30000 [8:33:58<47:20:33,  6.51s/it, lr: 1.0e-04 loss: 4.794e-01]
Xiang_wan_2_2_lora:  13%|####9                                  | 3811/30000 [8:33:58<47:20:33,  6.51s/it, lr: 1.0e-04 loss: 4.794e-01]
Xiang_wan_2_2_lora:  13%|####9                                  | 3811/30000 [8:34:06<47:20:33,  6.51s/it, lr: 1.0e-04 loss: 6.776e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3811/30000 [8:34:06<47:20:33,  6.51s/it, lr: 1.0e-04 loss: 6.776e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3812/30000 [8:34:06<49:14:51,  6.77s/it, lr: 1.0e-04 loss: 6.776e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3812/30000 [8:34:06<49:14:51,  6.77s/it, lr: 1.0e-04 loss: 6.776e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3812/30000 [8:34:13<49:14:51,  6.77s/it, lr: 1.0e-04 loss: 8.028e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3812/30000 [8:34:13<49:14:51,  6.77s/it, lr: 1.0e-04 loss: 8.028e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3813/30000 [8:34:13<50:38:29,  6.96s/it, lr: 1.0e-04 loss: 8.028e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3813/30000 [8:34:13<50:38:29,  6.96s/it, lr: 1.0e-04 loss: 8.028e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3813/30000 [8:34:21<50:38:29,  6.96s/it, lr: 1.0e-04 loss: 7.502e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3813/30000 [8:34:21<50:38:29,  6.96s/it, lr: 1.0e-04 loss: 7.502e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3814/30000 [8:34:21<51:47:52,  7.12s/it, lr: 1.0e-04 loss: 7.502e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3814/30000 [8:34:21<51:47:52,  7.12s/it, lr: 1.0e-04 loss: 7.502e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3814/30000 [8:34:28<51:47:52,  7.12s/it, lr: 1.0e-04 loss: 1.446e-01]
Xiang_wan_2_2_lora:  13%|####9                                  | 3814/30000 [8:34:28<51:47:52,  7.12s/it, lr: 1.0e-04 loss: 1.446e-01]
Xiang_wan_2_2_lora:  13%|####9                                  | 3815/30000 [8:34:28<52:41:37,  7.24s/it, lr: 1.0e-04 loss: 1.446e-01]
Xiang_wan_2_2_lora:  13%|####9                                  | 3815/30000 [8:34:28<52:41:37,  7.24s/it, lr: 1.0e-04 loss: 1.446e-01]
Xiang_wan_2_2_lora:  13%|####9                                  | 3815/30000 [8:34:39<52:41:37,  7.24s/it, lr: 1.0e-04 loss: 7.022e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3815/30000 [8:34:39<52:41:37,  7.24s/it, lr: 1.0e-04 loss: 7.022e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3816/30000 [8:34:39<59:02:56,  8.12s/it, lr: 1.0e-04 loss: 7.022e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3816/30000 [8:34:39<59:02:56,  8.12s/it, lr: 1.0e-04 loss: 7.022e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3816/30000 [8:34:49<59:02:56,  8.12s/it, lr: 1.0e-04 loss: 3.377e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3816/30000 [8:34:49<59:02:56,  8.12s/it, lr: 1.0e-04 loss: 3.377e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3817/30000 [8:34:49<63:39:53,  8.75s/it, lr: 1.0e-04 loss: 3.377e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3817/30000 [8:34:49<63:39:53,  8.75s/it, lr: 1.0e-04 loss: 3.377e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3817/30000 [8:35:00<63:39:53,  8.75s/it, lr: 1.0e-04 loss: 6.089e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3817/30000 [8:35:00<63:39:53,  8.75s/it, lr: 1.0e-04 loss: 6.089e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3818/30000 [8:35:00<68:12:56,  9.38s/it, lr: 1.0e-04 loss: 6.089e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3818/30000 [8:35:00<68:12:56,  9.38s/it, lr: 1.0e-04 loss: 6.089e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3818/30000 [8:35:09<68:12:56,  9.38s/it, lr: 1.0e-04 loss: 4.684e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3818/30000 [8:35:09<68:12:56,  9.38s/it, lr: 1.0e-04 loss: 4.684e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3819/30000 [8:35:09<68:29:15,  9.42s/it, lr: 1.0e-04 loss: 4.684e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3819/30000 [8:35:09<68:29:15,  9.42s/it, lr: 1.0e-04 loss: 4.684e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3819/30000 [8:35:19<68:29:15,  9.42s/it, lr: 1.0e-04 loss: 5.964e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3819/30000 [8:35:19<68:29:15,  9.42s/it, lr: 1.0e-04 loss: 5.964e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3820/30000 [8:35:25<68:29:05,  9.42s/it, lr: 1.0e-04 loss: 7.727e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3820/30000 [8:35:25<68:29:05,  9.42s/it, lr: 1.0e-04 loss: 7.727e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3821/30000 [8:35:25<48:16:59,  6.64s/it, lr: 1.0e-04 loss: 7.727e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3821/30000 [8:35:25<48:16:59,  6.64s/it, lr: 1.0e-04 loss: 7.727e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3821/30000 [8:35:35<48:16:59,  6.64s/it, lr: 1.0e-04 loss: 4.335e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3821/30000 [8:35:35<48:16:59,  6.64s/it, lr: 1.0e-04 loss: 4.335e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3822/30000 [8:35:35<53:14:28,  7.32s/it, lr: 1.0e-04 loss: 4.335e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3822/30000 [8:35:35<53:14:28,  7.32s/it, lr: 1.0e-04 loss: 4.335e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3822/30000 [8:35:42<53:14:28,  7.32s/it, lr: 1.0e-04 loss: 6.680e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3822/30000 [8:35:42<53:14:28,  7.32s/it, lr: 1.0e-04 loss: 6.680e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3823/30000 [8:35:42<53:12:10,  7.32s/it, lr: 1.0e-04 loss: 6.680e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3823/30000 [8:35:42<53:12:10,  7.32s/it, lr: 1.0e-04 loss: 6.680e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3823/30000 [8:35:50<53:12:10,  7.32s/it, lr: 1.0e-04 loss: 6.979e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3823/30000 [8:35:50<53:12:10,  7.32s/it, lr: 1.0e-04 loss: 6.979e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3824/30000 [8:35:50<54:03:24,  7.43s/it, lr: 1.0e-04 loss: 6.979e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3824/30000 [8:35:50<54:03:24,  7.43s/it, lr: 1.0e-04 loss: 6.979e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3824/30000 [8:35:58<54:03:24,  7.43s/it, lr: 1.0e-04 loss: 1.654e-01]
Xiang_wan_2_2_lora:  13%|####9                                  | 3824/30000 [8:35:58<54:03:24,  7.43s/it, lr: 1.0e-04 loss: 1.654e-01]
Xiang_wan_2_2_lora:  13%|####9                                  | 3825/30000 [8:35:58<55:51:40,  7.68s/it, lr: 1.0e-04 loss: 1.654e-01]
Xiang_wan_2_2_lora:  13%|####9                                  | 3825/30000 [8:35:58<55:51:40,  7.68s/it, lr: 1.0e-04 loss: 1.654e-01]
Xiang_wan_2_2_lora:  13%|####9                                  | 3825/30000 [8:36:09<55:51:40,  7.68s/it, lr: 1.0e-04 loss: 5.968e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3825/30000 [8:36:09<55:51:40,  7.68s/it, lr: 1.0e-04 loss: 5.968e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3826/30000 [8:36:09<61:39:58,  8.48s/it, lr: 1.0e-04 loss: 5.968e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3826/30000 [8:36:09<61:39:58,  8.48s/it, lr: 1.0e-04 loss: 5.968e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3826/30000 [8:36:19<61:39:58,  8.48s/it, lr: 1.0e-04 loss: 6.861e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3826/30000 [8:36:19<61:39:58,  8.48s/it, lr: 1.0e-04 loss: 6.861e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3827/30000 [8:36:19<65:44:27,  9.04s/it, lr: 1.0e-04 loss: 6.861e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3827/30000 [8:36:19<65:44:27,  9.04s/it, lr: 1.0e-04 loss: 6.861e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3827/30000 [8:36:30<65:44:27,  9.04s/it, lr: 1.0e-04 loss: 6.443e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3827/30000 [8:36:30<65:44:27,  9.04s/it, lr: 1.0e-04 loss: 6.443e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3828/30000 [8:36:30<68:53:20,  9.48s/it, lr: 1.0e-04 loss: 6.443e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3828/30000 [8:36:30<68:53:20,  9.48s/it, lr: 1.0e-04 loss: 6.443e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3828/30000 [8:36:40<68:53:20,  9.48s/it, lr: 1.0e-04 loss: 9.831e-01]
Xiang_wan_2_2_lora:  13%|####9                                  | 3828/30000 [8:36:40<68:53:20,  9.48s/it, lr: 1.0e-04 loss: 9.831e-01]
Xiang_wan_2_2_lora:  13%|####9                                  | 3829/30000 [8:36:40<70:50:52,  9.75s/it, lr: 1.0e-04 loss: 9.831e-01]
Xiang_wan_2_2_lora:  13%|####9                                  | 3829/30000 [8:36:40<70:50:52,  9.75s/it, lr: 1.0e-04 loss: 9.831e-01]
Xiang_wan_2_2_lora:  13%|####9                                  | 3829/30000 [8:36:48<70:50:52,  9.75s/it, lr: 1.0e-04 loss: 7.675e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3829/30000 [8:36:48<70:50:52,  9.75s/it, lr: 1.0e-04 loss: 7.675e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3830/30000 [8:36:57<70:50:42,  9.75s/it, lr: 1.0e-04 loss: 1.848e-01]
Xiang_wan_2_2_lora:  13%|####9                                  | 3830/30000 [8:36:57<70:50:42,  9.75s/it, lr: 1.0e-04 loss: 1.848e-01]
Xiang_wan_2_2_lora:  13%|####9                                  | 3831/30000 [8:36:57<53:57:42,  7.42s/it, lr: 1.0e-04 loss: 1.848e-01]
Xiang_wan_2_2_lora:  13%|####9                                  | 3831/30000 [8:36:57<53:57:42,  7.42s/it, lr: 1.0e-04 loss: 1.848e-01]
Xiang_wan_2_2_lora:  13%|####9                                  | 3831/30000 [8:37:07<53:57:42,  7.42s/it, lr: 1.0e-04 loss: 8.541e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3831/30000 [8:37:07<53:57:42,  7.42s/it, lr: 1.0e-04 loss: 8.541e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3832/30000 [8:37:07<59:11:58,  8.14s/it, lr: 1.0e-04 loss: 8.541e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3832/30000 [8:37:07<59:11:58,  8.14s/it, lr: 1.0e-04 loss: 8.541e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3832/30000 [8:37:16<59:11:58,  8.14s/it, lr: 1.0e-04 loss: 4.564e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3832/30000 [8:37:16<59:11:58,  8.14s/it, lr: 1.0e-04 loss: 4.564e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3833/30000 [8:37:16<59:23:41,  8.17s/it, lr: 1.0e-04 loss: 4.564e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3833/30000 [8:37:16<59:23:41,  8.17s/it, lr: 1.0e-04 loss: 4.564e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3833/30000 [8:37:26<59:23:41,  8.17s/it, lr: 1.0e-04 loss: 5.416e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3833/30000 [8:37:26<59:23:41,  8.17s/it, lr: 1.0e-04 loss: 5.416e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3834/30000 [8:37:26<63:07:52,  8.69s/it, lr: 1.0e-04 loss: 5.416e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3834/30000 [8:37:26<63:07:52,  8.69s/it, lr: 1.0e-04 loss: 5.416e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3834/30000 [8:37:32<63:07:52,  8.69s/it, lr: 1.0e-04 loss: 5.715e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3834/30000 [8:37:32<63:07:52,  8.69s/it, lr: 1.0e-04 loss: 5.715e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3835/30000 [8:37:32<59:02:33,  8.12s/it, lr: 1.0e-04 loss: 5.715e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3835/30000 [8:37:32<59:02:33,  8.12s/it, lr: 1.0e-04 loss: 5.715e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3835/30000 [8:37:39<59:02:33,  8.12s/it, lr: 1.0e-04 loss: 6.761e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3835/30000 [8:37:39<59:02:33,  8.12s/it, lr: 1.0e-04 loss: 6.761e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3836/30000 [8:37:39<56:03:08,  7.71s/it, lr: 1.0e-04 loss: 6.761e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3836/30000 [8:37:39<56:03:08,  7.71s/it, lr: 1.0e-04 loss: 6.761e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3836/30000 [8:37:46<56:03:08,  7.71s/it, lr: 1.0e-04 loss: 5.091e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3836/30000 [8:37:46<56:03:08,  7.71s/it, lr: 1.0e-04 loss: 5.091e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3837/30000 [8:37:46<53:43:49,  7.39s/it, lr: 1.0e-04 loss: 5.091e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3837/30000 [8:37:46<53:43:49,  7.39s/it, lr: 1.0e-04 loss: 5.091e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3837/30000 [8:37:56<53:43:49,  7.39s/it, lr: 1.0e-04 loss: 6.330e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3837/30000 [8:37:56<53:43:49,  7.39s/it, lr: 1.0e-04 loss: 6.330e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3838/30000 [8:37:56<59:25:08,  8.18s/it, lr: 1.0e-04 loss: 6.330e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3838/30000 [8:37:56<59:25:08,  8.18s/it, lr: 1.0e-04 loss: 6.330e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3838/30000 [8:38:03<59:25:08,  8.18s/it, lr: 1.0e-04 loss: 7.575e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3838/30000 [8:38:03<59:25:08,  8.18s/it, lr: 1.0e-04 loss: 7.575e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3839/30000 [8:38:03<58:20:23,  8.03s/it, lr: 1.0e-04 loss: 7.575e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3839/30000 [8:38:03<58:20:23,  8.03s/it, lr: 1.0e-04 loss: 7.575e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3839/30000 [8:38:14<58:20:23,  8.03s/it, lr: 1.0e-04 loss: 8.184e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3839/30000 [8:38:14<58:20:23,  8.03s/it, lr: 1.0e-04 loss: 8.184e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3840/30000 [8:38:25<58:20:15,  8.03s/it, lr: 1.0e-04 loss: 4.945e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3840/30000 [8:38:25<58:20:15,  8.03s/it, lr: 1.0e-04 loss: 4.945e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3841/30000 [8:38:25<48:53:28,  6.73s/it, lr: 1.0e-04 loss: 4.945e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3841/30000 [8:38:25<48:53:28,  6.73s/it, lr: 1.0e-04 loss: 4.945e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3841/30000 [8:38:32<48:53:28,  6.73s/it, lr: 1.0e-04 loss: 8.992e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3841/30000 [8:38:32<48:53:28,  6.73s/it, lr: 1.0e-04 loss: 8.992e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3842/30000 [8:38:32<50:30:12,  6.95s/it, lr: 1.0e-04 loss: 8.992e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3842/30000 [8:38:32<50:30:12,  6.95s/it, lr: 1.0e-04 loss: 8.992e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3842/30000 [8:38:40<50:30:12,  6.95s/it, lr: 1.0e-04 loss: 6.722e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3842/30000 [8:38:40<50:30:12,  6.95s/it, lr: 1.0e-04 loss: 6.722e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3843/30000 [8:38:40<51:42:29,  7.12s/it, lr: 1.0e-04 loss: 6.722e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3843/30000 [8:38:40<51:42:29,  7.12s/it, lr: 1.0e-04 loss: 6.722e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3843/30000 [8:38:48<51:42:29,  7.12s/it, lr: 1.0e-04 loss: 7.054e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3843/30000 [8:38:48<51:42:29,  7.12s/it, lr: 1.0e-04 loss: 7.054e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3844/30000 [8:38:48<52:30:58,  7.23s/it, lr: 1.0e-04 loss: 7.054e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3844/30000 [8:38:48<52:30:58,  7.23s/it, lr: 1.0e-04 loss: 7.054e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3844/30000 [8:38:55<52:30:58,  7.23s/it, lr: 1.0e-04 loss: 8.854e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3844/30000 [8:38:55<52:30:58,  7.23s/it, lr: 1.0e-04 loss: 8.854e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3845/30000 [8:38:55<53:11:26,  7.32s/it, lr: 1.0e-04 loss: 8.854e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3845/30000 [8:38:55<53:11:26,  7.32s/it, lr: 1.0e-04 loss: 8.854e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3845/30000 [8:39:03<53:11:26,  7.32s/it, lr: 1.0e-04 loss: 8.343e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3845/30000 [8:39:03<53:11:26,  7.32s/it, lr: 1.0e-04 loss: 8.343e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3846/30000 [8:39:03<53:44:00,  7.40s/it, lr: 1.0e-04 loss: 8.343e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3846/30000 [8:39:03<53:44:00,  7.40s/it, lr: 1.0e-04 loss: 8.343e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3846/30000 [8:39:13<53:44:00,  7.40s/it, lr: 1.0e-04 loss: 9.435e-02]
Xiang_wan_2_2_lora:  13%|####9                                  | 3846/30000 [8:39:13<53:44:00,  7.40s/it, lr: 1.0e-04 loss: 9.435e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3847/30000 [8:39:13<59:55:30,  8.25s/it, lr: 1.0e-04 loss: 9.435e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3847/30000 [8:39:13<59:55:30,  8.25s/it, lr: 1.0e-04 loss: 9.435e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3847/30000 [8:39:24<59:55:30,  8.25s/it, lr: 1.0e-04 loss: 6.184e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3847/30000 [8:39:24<59:55:30,  8.25s/it, lr: 1.0e-04 loss: 6.184e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3848/30000 [8:39:24<66:00:51,  9.09s/it, lr: 1.0e-04 loss: 6.184e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3848/30000 [8:39:24<66:00:51,  9.09s/it, lr: 1.0e-04 loss: 6.184e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3848/30000 [8:39:35<66:00:51,  9.09s/it, lr: 1.0e-04 loss: 1.126e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3848/30000 [8:39:35<66:00:51,  9.09s/it, lr: 1.0e-04 loss: 1.126e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3849/30000 [8:39:35<68:49:10,  9.47s/it, lr: 1.0e-04 loss: 1.126e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3849/30000 [8:39:35<68:49:10,  9.47s/it, lr: 1.0e-04 loss: 1.126e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3849/30000 [8:39:42<68:49:10,  9.47s/it, lr: 1.0e-04 loss: 9.903e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3849/30000 [8:39:42<68:49:10,  9.47s/it, lr: 1.0e-04 loss: 9.903e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3850/30000 [8:39:50<68:49:01,  9.47s/it, lr: 1.0e-04 loss: 1.004e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3850/30000 [8:39:50<68:49:01,  9.47s/it, lr: 1.0e-04 loss: 1.004e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3851/30000 [8:39:50<50:00:01,  6.88s/it, lr: 1.0e-04 loss: 1.004e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3851/30000 [8:39:50<50:00:01,  6.88s/it, lr: 1.0e-04 loss: 1.004e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3851/30000 [8:40:00<50:00:01,  6.88s/it, lr: 1.0e-04 loss: 9.629e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3851/30000 [8:40:00<50:00:01,  6.88s/it, lr: 1.0e-04 loss: 9.629e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3852/30000 [8:40:00<56:06:24,  7.72s/it, lr: 1.0e-04 loss: 9.629e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3852/30000 [8:40:00<56:06:24,  7.72s/it, lr: 1.0e-04 loss: 9.629e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3852/30000 [8:40:08<56:06:24,  7.72s/it, lr: 1.0e-04 loss: 5.967e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3852/30000 [8:40:08<56:06:24,  7.72s/it, lr: 1.0e-04 loss: 5.967e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3853/30000 [8:40:08<55:55:09,  7.70s/it, lr: 1.0e-04 loss: 5.967e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3853/30000 [8:40:08<55:55:09,  7.70s/it, lr: 1.0e-04 loss: 5.967e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3853/30000 [8:40:18<55:55:09,  7.70s/it, lr: 1.0e-04 loss: 5.799e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3853/30000 [8:40:18<55:55:09,  7.70s/it, lr: 1.0e-04 loss: 5.799e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3854/30000 [8:40:18<60:58:54,  8.40s/it, lr: 1.0e-04 loss: 5.799e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3854/30000 [8:40:18<60:58:54,  8.40s/it, lr: 1.0e-04 loss: 5.799e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3854/30000 [8:40:29<60:58:54,  8.40s/it, lr: 1.0e-04 loss: 5.032e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3854/30000 [8:40:29<60:58:54,  8.40s/it, lr: 1.0e-04 loss: 5.032e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3855/30000 [8:40:29<66:40:01,  9.18s/it, lr: 1.0e-04 loss: 5.032e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3855/30000 [8:40:29<66:40:01,  9.18s/it, lr: 1.0e-04 loss: 5.032e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3855/30000 [8:40:37<66:40:01,  9.18s/it, lr: 1.0e-04 loss: 9.925e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3855/30000 [8:40:37<66:40:01,  9.18s/it, lr: 1.0e-04 loss: 9.925e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3856/30000 [8:40:37<63:19:06,  8.72s/it, lr: 1.0e-04 loss: 9.925e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3856/30000 [8:40:37<63:19:06,  8.72s/it, lr: 1.0e-04 loss: 9.925e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3856/30000 [8:40:44<63:19:06,  8.72s/it, lr: 1.0e-04 loss: 8.041e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3856/30000 [8:40:44<63:19:06,  8.72s/it, lr: 1.0e-04 loss: 8.041e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3857/30000 [8:40:44<60:56:46,  8.39s/it, lr: 1.0e-04 loss: 8.041e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3857/30000 [8:40:44<60:56:46,  8.39s/it, lr: 1.0e-04 loss: 8.041e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3857/30000 [8:40:55<60:56:46,  8.39s/it, lr: 1.0e-04 loss: 5.986e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3857/30000 [8:40:55<60:56:46,  8.39s/it, lr: 1.0e-04 loss: 5.986e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3858/30000 [8:40:55<65:01:54,  8.96s/it, lr: 1.0e-04 loss: 5.986e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3858/30000 [8:40:55<65:01:54,  8.96s/it, lr: 1.0e-04 loss: 5.986e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3858/30000 [8:41:05<65:01:54,  8.96s/it, lr: 1.0e-04 loss: 4.448e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3858/30000 [8:41:05<65:01:54,  8.96s/it, lr: 1.0e-04 loss: 4.448e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3859/30000 [8:41:05<68:01:32,  9.37s/it, lr: 1.0e-04 loss: 4.448e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3859/30000 [8:41:05<68:01:32,  9.37s/it, lr: 1.0e-04 loss: 4.448e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3859/30000 [8:41:21<68:01:32,  9.37s/it, lr: 1.0e-04 loss: 7.862e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3859/30000 [8:41:21<68:01:32,  9.37s/it, lr: 1.0e-04 loss: 7.862e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3860/30000 [8:41:31<68:01:22,  9.37s/it, lr: 1.0e-04 loss: 9.697e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3860/30000 [8:41:31<68:01:22,  9.37s/it, lr: 1.0e-04 loss: 9.697e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3861/30000 [8:41:31<52:47:10,  7.27s/it, lr: 1.0e-04 loss: 9.697e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3861/30000 [8:41:31<52:47:10,  7.27s/it, lr: 1.0e-04 loss: 9.697e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3861/30000 [8:41:38<52:47:10,  7.27s/it, lr: 1.0e-04 loss: 6.809e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3861/30000 [8:41:38<52:47:10,  7.27s/it, lr: 1.0e-04 loss: 6.809e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3862/30000 [8:41:38<53:31:05,  7.37s/it, lr: 1.0e-04 loss: 6.809e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3862/30000 [8:41:38<53:31:05,  7.37s/it, lr: 1.0e-04 loss: 6.809e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3862/30000 [8:41:46<53:31:05,  7.37s/it, lr: 1.0e-04 loss: 8.255e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3862/30000 [8:41:46<53:31:05,  7.37s/it, lr: 1.0e-04 loss: 8.255e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3863/30000 [8:41:46<54:01:30,  7.44s/it, lr: 1.0e-04 loss: 8.255e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3863/30000 [8:41:46<54:01:30,  7.44s/it, lr: 1.0e-04 loss: 8.255e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3863/30000 [8:41:56<54:01:30,  7.44s/it, lr: 1.0e-04 loss: 4.880e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3863/30000 [8:41:56<54:01:30,  7.44s/it, lr: 1.0e-04 loss: 4.880e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3864/30000 [8:41:56<58:19:23,  8.03s/it, lr: 1.0e-04 loss: 4.880e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3864/30000 [8:41:56<58:19:23,  8.03s/it, lr: 1.0e-04 loss: 4.880e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3864/30000 [8:42:02<58:19:23,  8.03s/it, lr: 1.0e-04 loss: 7.248e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3864/30000 [8:42:02<58:19:23,  8.03s/it, lr: 1.0e-04 loss: 7.248e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3865/30000 [8:42:02<55:51:24,  7.69s/it, lr: 1.0e-04 loss: 7.248e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3865/30000 [8:42:02<55:51:24,  7.69s/it, lr: 1.0e-04 loss: 7.248e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3865/30000 [8:42:09<55:51:24,  7.69s/it, lr: 1.0e-04 loss: 1.151e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3865/30000 [8:42:09<55:51:24,  7.69s/it, lr: 1.0e-04 loss: 1.151e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3866/30000 [8:42:09<53:56:16,  7.43s/it, lr: 1.0e-04 loss: 1.151e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3866/30000 [8:42:09<53:56:16,  7.43s/it, lr: 1.0e-04 loss: 1.151e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3866/30000 [8:42:16<53:56:16,  7.43s/it, lr: 1.0e-04 loss: 6.439e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3866/30000 [8:42:16<53:56:16,  7.43s/it, lr: 1.0e-04 loss: 6.439e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3867/30000 [8:42:16<52:09:41,  7.19s/it, lr: 1.0e-04 loss: 6.439e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3867/30000 [8:42:16<52:09:41,  7.19s/it, lr: 1.0e-04 loss: 6.439e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3867/30000 [8:42:25<52:09:41,  7.19s/it, lr: 1.0e-04 loss: 5.542e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3867/30000 [8:42:25<52:09:41,  7.19s/it, lr: 1.0e-04 loss: 5.542e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3868/30000 [8:42:25<57:23:16,  7.91s/it, lr: 1.0e-04 loss: 5.542e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3868/30000 [8:42:25<57:23:16,  7.91s/it, lr: 1.0e-04 loss: 5.542e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3868/30000 [8:42:32<57:23:16,  7.91s/it, lr: 1.0e-04 loss: 7.919e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3868/30000 [8:42:32<57:23:16,  7.91s/it, lr: 1.0e-04 loss: 7.919e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3869/30000 [8:42:32<54:36:57,  7.52s/it, lr: 1.0e-04 loss: 7.919e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3869/30000 [8:42:32<54:36:57,  7.52s/it, lr: 1.0e-04 loss: 7.919e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3869/30000 [8:42:39<54:36:57,  7.52s/it, lr: 1.0e-04 loss: 1.049e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3869/30000 [8:42:39<54:36:57,  7.52s/it, lr: 1.0e-04 loss: 1.049e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3870/30000 [8:42:49<54:36:49,  7.52s/it, lr: 1.0e-04 loss: 1.245e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3870/30000 [8:42:49<54:36:49,  7.52s/it, lr: 1.0e-04 loss: 1.245e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3871/30000 [8:42:49<45:17:35,  6.24s/it, lr: 1.0e-04 loss: 1.245e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3871/30000 [8:42:49<45:17:35,  6.24s/it, lr: 1.0e-04 loss: 1.245e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3871/30000 [8:42:56<45:17:35,  6.24s/it, lr: 1.0e-04 loss: 7.111e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3871/30000 [8:42:56<45:17:35,  6.24s/it, lr: 1.0e-04 loss: 7.111e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3872/30000 [8:42:56<47:37:40,  6.56s/it, lr: 1.0e-04 loss: 7.111e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3872/30000 [8:42:56<47:37:40,  6.56s/it, lr: 1.0e-04 loss: 7.111e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3872/30000 [8:43:04<47:37:40,  6.56s/it, lr: 1.0e-04 loss: 7.538e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3872/30000 [8:43:04<47:37:40,  6.56s/it, lr: 1.0e-04 loss: 7.538e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3873/30000 [8:43:04<49:30:49,  6.82s/it, lr: 1.0e-04 loss: 7.538e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3873/30000 [8:43:04<49:30:49,  6.82s/it, lr: 1.0e-04 loss: 7.538e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3873/30000 [8:43:11<49:30:49,  6.82s/it, lr: 1.0e-04 loss: 1.346e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3873/30000 [8:43:11<49:30:49,  6.82s/it, lr: 1.0e-04 loss: 1.346e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3874/30000 [8:43:11<49:55:52,  6.88s/it, lr: 1.0e-04 loss: 1.346e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3874/30000 [8:43:11<49:55:52,  6.88s/it, lr: 1.0e-04 loss: 1.346e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3874/30000 [8:43:18<49:55:52,  6.88s/it, lr: 1.0e-04 loss: 1.718e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3874/30000 [8:43:18<49:55:52,  6.88s/it, lr: 1.0e-04 loss: 1.718e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3875/30000 [8:43:18<50:57:48,  7.02s/it, lr: 1.0e-04 loss: 1.718e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3875/30000 [8:43:18<50:57:48,  7.02s/it, lr: 1.0e-04 loss: 1.718e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3875/30000 [8:43:29<50:57:48,  7.02s/it, lr: 1.0e-04 loss: 8.102e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3875/30000 [8:43:29<50:57:48,  7.02s/it, lr: 1.0e-04 loss: 8.102e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3876/30000 [8:43:29<58:02:17,  8.00s/it, lr: 1.0e-04 loss: 8.102e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3876/30000 [8:43:29<58:02:17,  8.00s/it, lr: 1.0e-04 loss: 8.102e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3876/30000 [8:43:36<58:02:17,  8.00s/it, lr: 1.0e-04 loss: 1.037e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3876/30000 [8:43:36<58:02:17,  8.00s/it, lr: 1.0e-04 loss: 1.037e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3877/30000 [8:43:36<57:21:30,  7.90s/it, lr: 1.0e-04 loss: 1.037e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3877/30000 [8:43:36<57:21:30,  7.90s/it, lr: 1.0e-04 loss: 1.037e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3877/30000 [8:43:47<57:21:30,  7.90s/it, lr: 1.0e-04 loss: 6.682e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3877/30000 [8:43:47<57:21:30,  7.90s/it, lr: 1.0e-04 loss: 6.682e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3878/30000 [8:43:47<63:48:46,  8.79s/it, lr: 1.0e-04 loss: 6.682e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3878/30000 [8:43:47<63:48:46,  8.79s/it, lr: 1.0e-04 loss: 6.682e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3878/30000 [8:43:58<63:48:46,  8.79s/it, lr: 1.0e-04 loss: 1.656e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3878/30000 [8:43:58<63:48:46,  8.79s/it, lr: 1.0e-04 loss: 1.656e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3879/30000 [8:43:58<67:30:50,  9.30s/it, lr: 1.0e-04 loss: 1.656e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3879/30000 [8:43:58<67:30:50,  9.30s/it, lr: 1.0e-04 loss: 1.656e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3879/30000 [8:44:07<67:30:50,  9.30s/it, lr: 1.0e-04 loss: 7.327e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3879/30000 [8:44:07<67:30:50,  9.30s/it, lr: 1.0e-04 loss: 7.327e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3880/30000 [8:44:15<67:30:41,  9.30s/it, lr: 1.0e-04 loss: 8.288e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3880/30000 [8:44:15<67:30:41,  9.30s/it, lr: 1.0e-04 loss: 8.288e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3881/30000 [8:44:15<49:19:48,  6.80s/it, lr: 1.0e-04 loss: 8.288e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3881/30000 [8:44:15<49:19:48,  6.80s/it, lr: 1.0e-04 loss: 8.288e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3881/30000 [8:44:23<49:19:48,  6.80s/it, lr: 1.0e-04 loss: 6.032e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3881/30000 [8:44:23<49:19:48,  6.80s/it, lr: 1.0e-04 loss: 6.032e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3882/30000 [8:44:23<50:58:40,  7.03s/it, lr: 1.0e-04 loss: 6.032e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3882/30000 [8:44:23<50:58:40,  7.03s/it, lr: 1.0e-04 loss: 6.032e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3882/30000 [8:44:30<50:58:40,  7.03s/it, lr: 1.0e-04 loss: 4.238e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3882/30000 [8:44:30<50:58:40,  7.03s/it, lr: 1.0e-04 loss: 4.238e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3883/30000 [8:44:30<51:58:56,  7.17s/it, lr: 1.0e-04 loss: 4.238e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3883/30000 [8:44:30<51:58:56,  7.17s/it, lr: 1.0e-04 loss: 4.238e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3883/30000 [8:44:41<51:58:56,  7.17s/it, lr: 1.0e-04 loss: 7.458e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3883/30000 [8:44:41<51:58:56,  7.17s/it, lr: 1.0e-04 loss: 7.458e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3884/30000 [8:44:41<58:25:57,  8.05s/it, lr: 1.0e-04 loss: 7.458e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3884/30000 [8:44:41<58:25:57,  8.05s/it, lr: 1.0e-04 loss: 7.458e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3884/30000 [8:44:49<58:25:57,  8.05s/it, lr: 1.0e-04 loss: 1.147e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3884/30000 [8:44:49<58:25:57,  8.05s/it, lr: 1.0e-04 loss: 1.147e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3885/30000 [8:44:49<59:20:57,  8.18s/it, lr: 1.0e-04 loss: 1.147e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3885/30000 [8:44:49<59:20:57,  8.18s/it, lr: 1.0e-04 loss: 1.147e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3885/30000 [8:45:00<59:20:57,  8.18s/it, lr: 1.0e-04 loss: 6.624e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3885/30000 [8:45:00<59:20:57,  8.18s/it, lr: 1.0e-04 loss: 6.624e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3886/30000 [8:45:00<64:07:21,  8.84s/it, lr: 1.0e-04 loss: 6.624e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3886/30000 [8:45:00<64:07:21,  8.84s/it, lr: 1.0e-04 loss: 6.624e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3886/30000 [8:45:10<64:07:21,  8.84s/it, lr: 1.0e-04 loss: 5.776e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3886/30000 [8:45:10<64:07:21,  8.84s/it, lr: 1.0e-04 loss: 5.776e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3887/30000 [8:45:10<67:41:54,  9.33s/it, lr: 1.0e-04 loss: 5.776e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3887/30000 [8:45:10<67:41:54,  9.33s/it, lr: 1.0e-04 loss: 5.776e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3887/30000 [8:45:21<67:41:54,  9.33s/it, lr: 1.0e-04 loss: 6.433e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3887/30000 [8:45:21<67:41:54,  9.33s/it, lr: 1.0e-04 loss: 6.433e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3888/30000 [8:45:21<70:15:36,  9.69s/it, lr: 1.0e-04 loss: 6.433e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3888/30000 [8:45:21<70:15:36,  9.69s/it, lr: 1.0e-04 loss: 6.433e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3888/30000 [8:45:29<70:15:36,  9.69s/it, lr: 1.0e-04 loss: 7.642e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3888/30000 [8:45:29<70:15:36,  9.69s/it, lr: 1.0e-04 loss: 7.642e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3889/30000 [8:45:29<66:13:27,  9.13s/it, lr: 1.0e-04 loss: 7.642e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3889/30000 [8:45:29<66:13:27,  9.13s/it, lr: 1.0e-04 loss: 7.642e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3889/30000 [8:45:35<66:13:27,  9.13s/it, lr: 1.0e-04 loss: 6.868e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3889/30000 [8:45:35<66:13:27,  9.13s/it, lr: 1.0e-04 loss: 6.868e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3890/30000 [8:45:46<66:13:18,  9.13s/it, lr: 1.0e-04 loss: 2.259e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3890/30000 [8:45:46<66:13:18,  9.13s/it, lr: 1.0e-04 loss: 2.259e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3891/30000 [8:45:46<53:09:24,  7.33s/it, lr: 1.0e-04 loss: 2.259e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3891/30000 [8:45:46<53:09:24,  7.33s/it, lr: 1.0e-04 loss: 2.259e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3891/30000 [8:45:52<53:09:24,  7.33s/it, lr: 1.0e-04 loss: 6.780e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3891/30000 [8:45:52<53:09:24,  7.33s/it, lr: 1.0e-04 loss: 6.780e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3892/30000 [8:45:52<51:53:41,  7.16s/it, lr: 1.0e-04 loss: 6.780e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3892/30000 [8:45:52<51:53:41,  7.16s/it, lr: 1.0e-04 loss: 6.780e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3892/30000 [8:46:03<51:53:41,  7.16s/it, lr: 1.0e-04 loss: 9.993e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3892/30000 [8:46:03<51:53:41,  7.16s/it, lr: 1.0e-04 loss: 9.993e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3893/30000 [8:46:03<58:22:05,  8.05s/it, lr: 1.0e-04 loss: 9.993e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3893/30000 [8:46:03<58:22:05,  8.05s/it, lr: 1.0e-04 loss: 9.993e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3893/30000 [8:46:11<58:22:05,  8.05s/it, lr: 1.0e-04 loss: 8.070e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3893/30000 [8:46:11<58:22:05,  8.05s/it, lr: 1.0e-04 loss: 8.070e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3894/30000 [8:46:11<57:35:26,  7.94s/it, lr: 1.0e-04 loss: 8.070e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3894/30000 [8:46:11<57:35:26,  7.94s/it, lr: 1.0e-04 loss: 8.070e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3894/30000 [8:46:21<57:35:26,  7.94s/it, lr: 1.0e-04 loss: 7.985e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3894/30000 [8:46:21<57:35:26,  7.94s/it, lr: 1.0e-04 loss: 7.985e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3895/30000 [8:46:21<62:26:30,  8.61s/it, lr: 1.0e-04 loss: 7.985e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3895/30000 [8:46:21<62:26:30,  8.61s/it, lr: 1.0e-04 loss: 7.985e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3895/30000 [8:46:31<62:26:30,  8.61s/it, lr: 1.0e-04 loss: 5.986e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3895/30000 [8:46:31<62:26:30,  8.61s/it, lr: 1.0e-04 loss: 5.986e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3896/30000 [8:46:31<66:11:48,  9.13s/it, lr: 1.0e-04 loss: 5.986e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3896/30000 [8:46:31<66:11:48,  9.13s/it, lr: 1.0e-04 loss: 5.986e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3896/30000 [8:46:39<66:11:48,  9.13s/it, lr: 1.0e-04 loss: 4.900e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3896/30000 [8:46:39<66:11:48,  9.13s/it, lr: 1.0e-04 loss: 4.900e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3897/30000 [8:46:39<63:14:46,  8.72s/it, lr: 1.0e-04 loss: 4.900e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3897/30000 [8:46:39<63:14:46,  8.72s/it, lr: 1.0e-04 loss: 4.900e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3897/30000 [8:46:49<63:14:46,  8.72s/it, lr: 1.0e-04 loss: 6.772e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3897/30000 [8:46:49<63:14:46,  8.72s/it, lr: 1.0e-04 loss: 6.772e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3898/30000 [8:46:49<66:39:04,  9.19s/it, lr: 1.0e-04 loss: 6.772e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3898/30000 [8:46:49<66:39:04,  9.19s/it, lr: 1.0e-04 loss: 6.772e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3898/30000 [8:46:57<66:39:04,  9.19s/it, lr: 1.0e-04 loss: 9.090e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3898/30000 [8:46:57<66:39:04,  9.19s/it, lr: 1.0e-04 loss: 9.090e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3899/30000 [8:46:57<63:14:52,  8.72s/it, lr: 1.0e-04 loss: 9.090e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3899/30000 [8:46:57<63:14:52,  8.72s/it, lr: 1.0e-04 loss: 9.090e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3899/30000 [8:47:04<63:14:52,  8.72s/it, lr: 1.0e-04 loss: 8.723e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3899/30000 [8:47:04<63:14:52,  8.72s/it, lr: 1.0e-04 loss: 8.723e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3900/30000 [8:47:10<63:14:43,  8.72s/it, lr: 1.0e-04 loss: 5.704e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3900/30000 [8:47:10<63:14:43,  8.72s/it, lr: 1.0e-04 loss: 5.704e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3901/30000 [8:47:10<45:18:15,  6.25s/it, lr: 1.0e-04 loss: 5.704e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3901/30000 [8:47:10<45:18:15,  6.25s/it, lr: 1.0e-04 loss: 5.704e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3901/30000 [8:47:18<45:18:15,  6.25s/it, lr: 1.0e-04 loss: 8.189e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3901/30000 [8:47:18<45:18:15,  6.25s/it, lr: 1.0e-04 loss: 8.189e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3902/30000 [8:47:18<47:52:42,  6.60s/it, lr: 1.0e-04 loss: 8.189e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3902/30000 [8:47:18<47:52:42,  6.60s/it, lr: 1.0e-04 loss: 8.189e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3902/30000 [8:47:25<47:52:42,  6.60s/it, lr: 1.0e-04 loss: 9.620e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3902/30000 [8:47:25<47:52:42,  6.60s/it, lr: 1.0e-04 loss: 9.620e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3903/30000 [8:47:25<47:56:09,  6.61s/it, lr: 1.0e-04 loss: 9.620e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3903/30000 [8:47:25<47:56:09,  6.61s/it, lr: 1.0e-04 loss: 9.620e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3903/30000 [8:47:31<47:56:09,  6.61s/it, lr: 1.0e-04 loss: 7.090e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3903/30000 [8:47:31<47:56:09,  6.61s/it, lr: 1.0e-04 loss: 7.090e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3904/30000 [8:47:31<47:56:17,  6.61s/it, lr: 1.0e-04 loss: 7.090e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3904/30000 [8:47:31<47:56:17,  6.61s/it, lr: 1.0e-04 loss: 7.090e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3904/30000 [8:47:38<47:56:17,  6.61s/it, lr: 1.0e-04 loss: 8.146e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3904/30000 [8:47:38<47:56:17,  6.61s/it, lr: 1.0e-04 loss: 8.146e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3905/30000 [8:47:38<47:57:25,  6.62s/it, lr: 1.0e-04 loss: 8.146e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3905/30000 [8:47:38<47:57:25,  6.62s/it, lr: 1.0e-04 loss: 8.146e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3905/30000 [8:47:45<47:57:25,  6.62s/it, lr: 1.0e-04 loss: 6.510e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3905/30000 [8:47:45<47:57:25,  6.62s/it, lr: 1.0e-04 loss: 6.510e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3906/30000 [8:47:45<47:48:34,  6.60s/it, lr: 1.0e-04 loss: 6.510e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3906/30000 [8:47:45<47:48:34,  6.60s/it, lr: 1.0e-04 loss: 6.510e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3906/30000 [8:47:54<47:48:34,  6.60s/it, lr: 1.0e-04 loss: 4.071e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3906/30000 [8:47:54<47:48:34,  6.60s/it, lr: 1.0e-04 loss: 4.071e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3907/30000 [8:47:54<53:37:04,  7.40s/it, lr: 1.0e-04 loss: 4.071e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3907/30000 [8:47:54<53:37:04,  7.40s/it, lr: 1.0e-04 loss: 4.071e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3907/30000 [8:48:04<53:37:04,  7.40s/it, lr: 1.0e-04 loss: 7.544e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3907/30000 [8:48:04<53:37:04,  7.40s/it, lr: 1.0e-04 loss: 7.544e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3908/30000 [8:48:04<58:14:34,  8.04s/it, lr: 1.0e-04 loss: 7.544e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3908/30000 [8:48:04<58:14:34,  8.04s/it, lr: 1.0e-04 loss: 7.544e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3908/30000 [8:48:10<58:14:34,  8.04s/it, lr: 1.0e-04 loss: 5.993e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3908/30000 [8:48:10<58:14:34,  8.04s/it, lr: 1.0e-04 loss: 5.993e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3909/30000 [8:48:10<55:33:13,  7.67s/it, lr: 1.0e-04 loss: 5.993e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3909/30000 [8:48:10<55:33:13,  7.67s/it, lr: 1.0e-04 loss: 5.993e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3909/30000 [8:48:18<55:33:13,  7.67s/it, lr: 1.0e-04 loss: 9.230e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3909/30000 [8:48:18<55:33:13,  7.67s/it, lr: 1.0e-04 loss: 9.230e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3910/30000 [8:48:26<55:33:06,  7.67s/it, lr: 1.0e-04 loss: 7.162e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3910/30000 [8:48:26<55:33:06,  7.67s/it, lr: 1.0e-04 loss: 7.162e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3911/30000 [8:48:26<43:00:39,  5.94s/it, lr: 1.0e-04 loss: 7.162e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3911/30000 [8:48:26<43:00:39,  5.94s/it, lr: 1.0e-04 loss: 7.162e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3911/30000 [8:48:33<43:00:39,  5.94s/it, lr: 1.0e-04 loss: 7.473e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3911/30000 [8:48:33<43:00:39,  5.94s/it, lr: 1.0e-04 loss: 7.473e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3912/30000 [8:48:33<44:48:36,  6.18s/it, lr: 1.0e-04 loss: 7.473e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3912/30000 [8:48:33<44:48:36,  6.18s/it, lr: 1.0e-04 loss: 7.473e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3912/30000 [8:48:41<44:48:36,  6.18s/it, lr: 1.0e-04 loss: 5.973e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3912/30000 [8:48:41<44:48:36,  6.18s/it, lr: 1.0e-04 loss: 5.973e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3913/30000 [8:48:41<47:45:17,  6.59s/it, lr: 1.0e-04 loss: 5.973e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3913/30000 [8:48:41<47:45:17,  6.59s/it, lr: 1.0e-04 loss: 5.973e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3913/30000 [8:48:48<47:45:17,  6.59s/it, lr: 1.0e-04 loss: 6.319e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3913/30000 [8:48:48<47:45:17,  6.59s/it, lr: 1.0e-04 loss: 6.319e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3914/30000 [8:48:48<49:59:26,  6.90s/it, lr: 1.0e-04 loss: 6.319e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3914/30000 [8:48:48<49:59:26,  6.90s/it, lr: 1.0e-04 loss: 6.319e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3914/30000 [8:48:57<49:59:26,  6.90s/it, lr: 1.0e-04 loss: 8.518e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3914/30000 [8:48:57<49:59:26,  6.90s/it, lr: 1.0e-04 loss: 8.518e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3915/30000 [8:48:57<53:03:34,  7.32s/it, lr: 1.0e-04 loss: 8.518e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3915/30000 [8:48:57<53:03:34,  7.32s/it, lr: 1.0e-04 loss: 8.518e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3915/30000 [8:49:06<53:03:34,  7.32s/it, lr: 1.0e-04 loss: 4.898e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3915/30000 [8:49:06<53:03:34,  7.32s/it, lr: 1.0e-04 loss: 4.898e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3916/30000 [8:49:06<57:11:57,  7.89s/it, lr: 1.0e-04 loss: 4.898e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3916/30000 [8:49:06<57:11:57,  7.89s/it, lr: 1.0e-04 loss: 4.898e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3916/30000 [8:49:13<57:11:57,  7.89s/it, lr: 1.0e-04 loss: 5.395e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3916/30000 [8:49:13<57:11:57,  7.89s/it, lr: 1.0e-04 loss: 5.395e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3917/30000 [8:49:13<54:46:38,  7.56s/it, lr: 1.0e-04 loss: 5.395e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3917/30000 [8:49:13<54:46:38,  7.56s/it, lr: 1.0e-04 loss: 5.395e-01]
Xiang_wan_2_2_lora:  13%|#####                                  | 3917/30000 [8:49:19<54:46:38,  7.56s/it, lr: 1.0e-04 loss: 7.530e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3917/30000 [8:49:19<54:46:38,  7.56s/it, lr: 1.0e-04 loss: 7.530e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3918/30000 [8:49:19<52:48:01,  7.29s/it, lr: 1.0e-04 loss: 7.530e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3918/30000 [8:49:19<52:48:01,  7.29s/it, lr: 1.0e-04 loss: 7.530e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3918/30000 [8:49:26<52:48:01,  7.29s/it, lr: 1.0e-04 loss: 6.087e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3918/30000 [8:49:26<52:48:01,  7.29s/it, lr: 1.0e-04 loss: 6.087e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3919/30000 [8:49:26<51:23:46,  7.09s/it, lr: 1.0e-04 loss: 6.087e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3919/30000 [8:49:26<51:23:46,  7.09s/it, lr: 1.0e-04 loss: 6.087e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3919/30000 [8:49:33<51:23:46,  7.09s/it, lr: 1.0e-04 loss: 6.451e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3919/30000 [8:49:33<51:23:46,  7.09s/it, lr: 1.0e-04 loss: 6.451e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3920/30000 [8:49:42<51:23:39,  7.09s/it, lr: 1.0e-04 loss: 5.767e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3920/30000 [8:49:42<51:23:39,  7.09s/it, lr: 1.0e-04 loss: 5.767e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3921/30000 [8:49:42<43:29:05,  6.00s/it, lr: 1.0e-04 loss: 5.767e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3921/30000 [8:49:42<43:29:05,  6.00s/it, lr: 1.0e-04 loss: 5.767e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3921/30000 [8:49:51<43:29:05,  6.00s/it, lr: 1.0e-04 loss: 5.120e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3921/30000 [8:49:51<43:29:05,  6.00s/it, lr: 1.0e-04 loss: 5.120e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3922/30000 [8:49:51<49:27:58,  6.83s/it, lr: 1.0e-04 loss: 5.120e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3922/30000 [8:49:51<49:27:58,  6.83s/it, lr: 1.0e-04 loss: 5.120e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3922/30000 [8:50:01<49:27:58,  6.83s/it, lr: 1.0e-04 loss: 5.552e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3922/30000 [8:50:01<49:27:58,  6.83s/it, lr: 1.0e-04 loss: 5.552e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3923/30000 [8:50:01<54:33:50,  7.53s/it, lr: 1.0e-04 loss: 5.552e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3923/30000 [8:50:01<54:33:50,  7.53s/it, lr: 1.0e-04 loss: 5.552e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3923/30000 [8:50:08<54:33:50,  7.53s/it, lr: 1.0e-04 loss: 8.523e-02]
Xiang_wan_2_2_lora:  13%|#####                                  | 3923/30000 [8:50:08<54:33:50,  7.53s/it, lr: 1.0e-04 loss: 8.523e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3924/30000 [8:50:08<52:49:43,  7.29s/it, lr: 1.0e-04 loss: 8.523e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3924/30000 [8:50:08<52:49:43,  7.29s/it, lr: 1.0e-04 loss: 8.523e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3924/30000 [8:50:14<52:49:43,  7.29s/it, lr: 1.0e-04 loss: 5.178e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3924/30000 [8:50:14<52:49:43,  7.29s/it, lr: 1.0e-04 loss: 5.178e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3925/30000 [8:50:14<51:33:04,  7.12s/it, lr: 1.0e-04 loss: 5.178e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3925/30000 [8:50:14<51:33:04,  7.12s/it, lr: 1.0e-04 loss: 5.178e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3925/30000 [8:50:21<51:33:04,  7.12s/it, lr: 1.0e-04 loss: 9.179e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3925/30000 [8:50:21<51:33:04,  7.12s/it, lr: 1.0e-04 loss: 9.179e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3926/30000 [8:50:21<50:37:26,  6.99s/it, lr: 1.0e-04 loss: 9.179e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3926/30000 [8:50:21<50:37:26,  6.99s/it, lr: 1.0e-04 loss: 9.179e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3926/30000 [8:50:30<50:37:26,  6.99s/it, lr: 1.0e-04 loss: 1.146e-01]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3926/30000 [8:50:30<50:37:26,  6.99s/it, lr: 1.0e-04 loss: 1.146e-01]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3927/30000 [8:50:30<55:41:58,  7.69s/it, lr: 1.0e-04 loss: 1.146e-01]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3927/30000 [8:50:30<55:41:58,  7.69s/it, lr: 1.0e-04 loss: 1.146e-01]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3927/30000 [8:50:40<55:41:58,  7.69s/it, lr: 1.0e-04 loss: 5.500e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3927/30000 [8:50:40<55:41:58,  7.69s/it, lr: 1.0e-04 loss: 5.500e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3928/30000 [8:50:40<59:20:14,  8.19s/it, lr: 1.0e-04 loss: 5.500e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3928/30000 [8:50:40<59:20:14,  8.19s/it, lr: 1.0e-04 loss: 5.500e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3928/30000 [8:50:49<59:20:14,  8.19s/it, lr: 1.0e-04 loss: 8.143e-01]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3928/30000 [8:50:49<59:20:14,  8.19s/it, lr: 1.0e-04 loss: 8.143e-01]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3929/30000 [8:50:49<62:08:14,  8.58s/it, lr: 1.0e-04 loss: 8.143e-01]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3929/30000 [8:50:49<62:08:14,  8.58s/it, lr: 1.0e-04 loss: 8.143e-01]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3929/30000 [8:50:59<62:08:14,  8.58s/it, lr: 1.0e-04 loss: 8.493e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3929/30000 [8:50:59<62:08:14,  8.58s/it, lr: 1.0e-04 loss: 8.493e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3930/30000 [8:51:06<62:08:06,  8.58s/it, lr: 1.0e-04 loss: 5.680e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3930/30000 [8:51:06<62:08:06,  8.58s/it, lr: 1.0e-04 loss: 5.680e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3931/30000 [8:51:06<44:48:58,  6.19s/it, lr: 1.0e-04 loss: 5.680e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3931/30000 [8:51:06<44:48:58,  6.19s/it, lr: 1.0e-04 loss: 5.680e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3931/30000 [8:51:15<44:48:58,  6.19s/it, lr: 1.0e-04 loss: 5.945e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3931/30000 [8:51:15<44:48:58,  6.19s/it, lr: 1.0e-04 loss: 5.945e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3932/30000 [8:51:15<50:30:08,  6.97s/it, lr: 1.0e-04 loss: 5.945e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3932/30000 [8:51:15<50:30:08,  6.97s/it, lr: 1.0e-04 loss: 5.945e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3932/30000 [8:51:24<50:30:08,  6.97s/it, lr: 1.0e-04 loss: 3.249e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3932/30000 [8:51:24<50:30:08,  6.97s/it, lr: 1.0e-04 loss: 3.249e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3933/30000 [8:51:24<55:01:06,  7.60s/it, lr: 1.0e-04 loss: 3.249e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3933/30000 [8:51:24<55:01:06,  7.60s/it, lr: 1.0e-04 loss: 3.249e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3933/30000 [8:51:31<55:01:06,  7.60s/it, lr: 1.0e-04 loss: 8.207e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3933/30000 [8:51:31<55:01:06,  7.60s/it, lr: 1.0e-04 loss: 8.207e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3934/30000 [8:51:31<53:10:42,  7.34s/it, lr: 1.0e-04 loss: 8.207e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3934/30000 [8:51:31<53:10:42,  7.34s/it, lr: 1.0e-04 loss: 8.207e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3934/30000 [8:51:40<53:10:42,  7.34s/it, lr: 1.0e-04 loss: 9.485e-01]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3934/30000 [8:51:40<53:10:42,  7.34s/it, lr: 1.0e-04 loss: 9.485e-01]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3935/30000 [8:51:40<57:24:09,  7.93s/it, lr: 1.0e-04 loss: 9.485e-01]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3935/30000 [8:51:40<57:24:09,  7.93s/it, lr: 1.0e-04 loss: 9.485e-01]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3935/30000 [8:51:47<57:24:09,  7.93s/it, lr: 1.0e-04 loss: 7.391e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3935/30000 [8:51:47<57:24:09,  7.93s/it, lr: 1.0e-04 loss: 7.391e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3936/30000 [8:51:47<54:47:08,  7.57s/it, lr: 1.0e-04 loss: 7.391e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3936/30000 [8:51:47<54:47:08,  7.57s/it, lr: 1.0e-04 loss: 7.391e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3936/30000 [8:51:56<54:47:08,  7.57s/it, lr: 1.0e-04 loss: 6.121e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3936/30000 [8:51:56<54:47:08,  7.57s/it, lr: 1.0e-04 loss: 6.121e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3937/30000 [8:51:56<58:32:40,  8.09s/it, lr: 1.0e-04 loss: 6.121e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3937/30000 [8:51:56<58:32:40,  8.09s/it, lr: 1.0e-04 loss: 6.121e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3937/30000 [8:52:03<58:32:40,  8.09s/it, lr: 1.0e-04 loss: 7.507e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3937/30000 [8:52:03<58:32:40,  8.09s/it, lr: 1.0e-04 loss: 7.507e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3938/30000 [8:52:03<56:09:29,  7.76s/it, lr: 1.0e-04 loss: 7.507e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3938/30000 [8:52:03<56:09:29,  7.76s/it, lr: 1.0e-04 loss: 7.507e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3938/30000 [8:52:14<56:09:29,  7.76s/it, lr: 1.0e-04 loss: 4.307e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3938/30000 [8:52:14<56:09:29,  7.76s/it, lr: 1.0e-04 loss: 4.307e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3939/30000 [8:52:14<61:28:04,  8.49s/it, lr: 1.0e-04 loss: 4.307e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3939/30000 [8:52:14<61:28:04,  8.49s/it, lr: 1.0e-04 loss: 4.307e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3939/30000 [8:52:24<61:28:04,  8.49s/it, lr: 1.0e-04 loss: 8.772e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3939/30000 [8:52:24<61:28:04,  8.49s/it, lr: 1.0e-04 loss: 8.772e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3940/30000 [8:52:32<61:27:56,  8.49s/it, lr: 1.0e-04 loss: 7.842e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3940/30000 [8:52:32<61:27:56,  8.49s/it, lr: 1.0e-04 loss: 7.842e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3941/30000 [8:52:32<46:07:33,  6.37s/it, lr: 1.0e-04 loss: 7.842e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3941/30000 [8:52:32<46:07:33,  6.37s/it, lr: 1.0e-04 loss: 7.842e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3941/30000 [8:52:40<46:07:33,  6.37s/it, lr: 1.0e-04 loss: 8.626e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3941/30000 [8:52:40<46:07:33,  6.37s/it, lr: 1.0e-04 loss: 8.626e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3942/30000 [8:52:40<48:41:58,  6.73s/it, lr: 1.0e-04 loss: 8.626e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3942/30000 [8:52:40<48:41:58,  6.73s/it, lr: 1.0e-04 loss: 8.626e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3942/30000 [8:52:50<48:41:58,  6.73s/it, lr: 1.0e-04 loss: 4.801e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3942/30000 [8:52:50<48:41:58,  6.73s/it, lr: 1.0e-04 loss: 4.801e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3943/30000 [8:52:50<55:46:53,  7.71s/it, lr: 1.0e-04 loss: 4.801e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3943/30000 [8:52:50<55:46:53,  7.71s/it, lr: 1.0e-04 loss: 4.801e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3943/30000 [8:53:01<55:46:53,  7.71s/it, lr: 1.0e-04 loss: 8.442e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3943/30000 [8:53:01<55:46:53,  7.71s/it, lr: 1.0e-04 loss: 8.442e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3944/30000 [8:53:01<61:21:17,  8.48s/it, lr: 1.0e-04 loss: 8.442e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3944/30000 [8:53:01<61:21:17,  8.48s/it, lr: 1.0e-04 loss: 8.442e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3944/30000 [8:53:12<61:21:17,  8.48s/it, lr: 1.0e-04 loss: 4.049e-01]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3944/30000 [8:53:12<61:21:17,  8.48s/it, lr: 1.0e-04 loss: 4.049e-01]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3945/30000 [8:53:12<66:34:36,  9.20s/it, lr: 1.0e-04 loss: 4.049e-01]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3945/30000 [8:53:12<66:34:36,  9.20s/it, lr: 1.0e-04 loss: 4.049e-01]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3945/30000 [8:53:19<66:34:36,  9.20s/it, lr: 1.0e-04 loss: 1.433e-01]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3945/30000 [8:53:19<66:34:36,  9.20s/it, lr: 1.0e-04 loss: 1.433e-01]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3946/30000 [8:53:19<62:41:48,  8.66s/it, lr: 1.0e-04 loss: 1.433e-01]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3946/30000 [8:53:19<62:41:48,  8.66s/it, lr: 1.0e-04 loss: 1.433e-01]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3946/30000 [8:53:30<62:41:48,  8.66s/it, lr: 1.0e-04 loss: 1.650e-01]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3946/30000 [8:53:30<62:41:48,  8.66s/it, lr: 1.0e-04 loss: 1.650e-01]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3947/30000 [8:53:30<67:28:40,  9.32s/it, lr: 1.0e-04 loss: 1.650e-01]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3947/30000 [8:53:30<67:28:40,  9.32s/it, lr: 1.0e-04 loss: 1.650e-01]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3947/30000 [8:53:41<67:28:40,  9.32s/it, lr: 1.0e-04 loss: 5.508e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3947/30000 [8:53:41<67:28:40,  9.32s/it, lr: 1.0e-04 loss: 5.508e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3948/30000 [8:53:41<70:32:43,  9.75s/it, lr: 1.0e-04 loss: 5.508e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3948/30000 [8:53:41<70:32:43,  9.75s/it, lr: 1.0e-04 loss: 5.508e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3948/30000 [8:53:51<70:32:43,  9.75s/it, lr: 1.0e-04 loss: 4.693e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3948/30000 [8:53:51<70:32:43,  9.75s/it, lr: 1.0e-04 loss: 4.693e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3949/30000 [8:53:51<72:02:36,  9.96s/it, lr: 1.0e-04 loss: 4.693e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3949/30000 [8:53:51<72:02:36,  9.96s/it, lr: 1.0e-04 loss: 4.693e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3949/30000 [8:54:02<72:02:36,  9.96s/it, lr: 1.0e-04 loss: 7.213e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3949/30000 [8:54:02<72:02:36,  9.96s/it, lr: 1.0e-04 loss: 7.213e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3950/30000 [8:54:13<72:02:26,  9.96s/it, lr: 1.0e-04 loss: 6.915e-01]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3950/30000 [8:54:13<72:02:26,  9.96s/it, lr: 1.0e-04 loss: 6.915e-01]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3951/30000 [8:54:13<56:58:59,  7.88s/it, lr: 1.0e-04 loss: 6.915e-01]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3951/30000 [8:54:13<56:58:59,  7.88s/it, lr: 1.0e-04 loss: 6.915e-01]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3951/30000 [8:54:24<56:58:59,  7.88s/it, lr: 1.0e-04 loss: 7.940e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3951/30000 [8:54:24<56:58:59,  7.88s/it, lr: 1.0e-04 loss: 7.940e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3952/30000 [8:54:24<61:54:51,  8.56s/it, lr: 1.0e-04 loss: 7.940e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3952/30000 [8:54:24<61:54:51,  8.56s/it, lr: 1.0e-04 loss: 7.940e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3952/30000 [8:54:35<61:54:51,  8.56s/it, lr: 1.0e-04 loss: 1.669e-01]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3952/30000 [8:54:35<61:54:51,  8.56s/it, lr: 1.0e-04 loss: 1.669e-01]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3953/30000 [8:54:35<66:46:06,  9.23s/it, lr: 1.0e-04 loss: 1.669e-01]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3953/30000 [8:54:35<66:46:06,  9.23s/it, lr: 1.0e-04 loss: 1.669e-01]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3953/30000 [8:54:43<66:46:06,  9.23s/it, lr: 1.0e-04 loss: 7.150e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3953/30000 [8:54:43<66:46:06,  9.23s/it, lr: 1.0e-04 loss: 7.150e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3954/30000 [8:54:43<65:31:10,  9.06s/it, lr: 1.0e-04 loss: 7.150e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3954/30000 [8:54:43<65:31:10,  9.06s/it, lr: 1.0e-04 loss: 7.150e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3954/30000 [8:54:51<65:31:10,  9.06s/it, lr: 1.0e-04 loss: 4.338e-01]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3954/30000 [8:54:51<65:31:10,  9.06s/it, lr: 1.0e-04 loss: 4.338e-01]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3955/30000 [8:54:51<63:34:04,  8.79s/it, lr: 1.0e-04 loss: 4.338e-01]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3955/30000 [8:54:51<63:34:04,  8.79s/it, lr: 1.0e-04 loss: 4.338e-01]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3955/30000 [8:55:02<63:34:04,  8.79s/it, lr: 1.0e-04 loss: 6.184e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3955/30000 [8:55:02<63:34:04,  8.79s/it, lr: 1.0e-04 loss: 6.184e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3956/30000 [8:55:02<67:19:57,  9.31s/it, lr: 1.0e-04 loss: 6.184e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3956/30000 [8:55:02<67:19:57,  9.31s/it, lr: 1.0e-04 loss: 6.184e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3956/30000 [8:55:13<67:19:57,  9.31s/it, lr: 1.0e-04 loss: 4.387e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3956/30000 [8:55:13<67:19:57,  9.31s/it, lr: 1.0e-04 loss: 4.387e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3957/30000 [8:55:13<70:12:06,  9.70s/it, lr: 1.0e-04 loss: 4.387e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3957/30000 [8:55:13<70:12:06,  9.70s/it, lr: 1.0e-04 loss: 4.387e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3957/30000 [8:55:23<70:12:06,  9.70s/it, lr: 1.0e-04 loss: 7.258e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3957/30000 [8:55:23<70:12:06,  9.70s/it, lr: 1.0e-04 loss: 7.258e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3958/30000 [8:55:23<72:19:32, 10.00s/it, lr: 1.0e-04 loss: 7.258e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3958/30000 [8:55:23<72:19:32, 10.00s/it, lr: 1.0e-04 loss: 7.258e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3958/30000 [8:55:35<72:19:32, 10.00s/it, lr: 1.0e-04 loss: 2.987e-01]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3958/30000 [8:55:35<72:19:32, 10.00s/it, lr: 1.0e-04 loss: 2.987e-01]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3959/30000 [8:55:35<75:30:25, 10.44s/it, lr: 1.0e-04 loss: 2.987e-01]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3959/30000 [8:55:35<75:30:25, 10.44s/it, lr: 1.0e-04 loss: 2.987e-01]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3959/30000 [8:55:44<75:30:25, 10.44s/it, lr: 1.0e-04 loss: 6.974e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3959/30000 [8:55:44<75:30:25, 10.44s/it, lr: 1.0e-04 loss: 6.974e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3960/30000 [8:55:54<75:30:15, 10.44s/it, lr: 1.0e-04 loss: 4.949e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3960/30000 [8:55:54<75:30:15, 10.44s/it, lr: 1.0e-04 loss: 4.949e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3961/30000 [8:55:54<58:23:31,  8.07s/it, lr: 1.0e-04 loss: 4.949e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3961/30000 [8:55:54<58:23:31,  8.07s/it, lr: 1.0e-04 loss: 4.949e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3961/30000 [8:56:02<58:23:31,  8.07s/it, lr: 1.0e-04 loss: 6.737e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3961/30000 [8:56:02<58:23:31,  8.07s/it, lr: 1.0e-04 loss: 6.737e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3962/30000 [8:56:02<58:07:26,  8.04s/it, lr: 1.0e-04 loss: 6.737e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3962/30000 [8:56:02<58:07:26,  8.04s/it, lr: 1.0e-04 loss: 6.737e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3962/30000 [8:56:13<58:07:26,  8.04s/it, lr: 1.0e-04 loss: 3.857e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3962/30000 [8:56:13<58:07:26,  8.04s/it, lr: 1.0e-04 loss: 3.857e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3963/30000 [8:56:13<63:35:43,  8.79s/it, lr: 1.0e-04 loss: 3.857e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3963/30000 [8:56:13<63:35:43,  8.79s/it, lr: 1.0e-04 loss: 3.857e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3963/30000 [8:56:25<63:35:43,  8.79s/it, lr: 1.0e-04 loss: 4.823e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3963/30000 [8:56:25<63:35:43,  8.79s/it, lr: 1.0e-04 loss: 4.823e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3964/30000 [8:56:25<69:43:30,  9.64s/it, lr: 1.0e-04 loss: 4.823e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3964/30000 [8:56:25<69:43:30,  9.64s/it, lr: 1.0e-04 loss: 4.823e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3964/30000 [8:56:36<69:43:30,  9.64s/it, lr: 1.0e-04 loss: 5.777e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3964/30000 [8:56:36<69:43:30,  9.64s/it, lr: 1.0e-04 loss: 5.777e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3965/30000 [8:56:36<73:30:41, 10.16s/it, lr: 1.0e-04 loss: 5.777e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3965/30000 [8:56:36<73:30:41, 10.16s/it, lr: 1.0e-04 loss: 5.777e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3965/30000 [8:56:45<73:30:41, 10.16s/it, lr: 1.0e-04 loss: 9.935e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3965/30000 [8:56:45<73:30:41, 10.16s/it, lr: 1.0e-04 loss: 9.935e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3966/30000 [8:56:45<70:04:10,  9.69s/it, lr: 1.0e-04 loss: 9.935e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3966/30000 [8:56:45<70:04:10,  9.69s/it, lr: 1.0e-04 loss: 9.935e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3966/30000 [8:56:56<70:04:10,  9.69s/it, lr: 1.0e-04 loss: 5.120e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3966/30000 [8:56:56<70:04:10,  9.69s/it, lr: 1.0e-04 loss: 5.120e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3967/30000 [8:56:56<72:48:48, 10.07s/it, lr: 1.0e-04 loss: 5.120e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3967/30000 [8:56:56<72:48:48, 10.07s/it, lr: 1.0e-04 loss: 5.120e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3967/30000 [8:57:05<72:48:48, 10.07s/it, lr: 1.0e-04 loss: 8.006e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3967/30000 [8:57:05<72:48:48, 10.07s/it, lr: 1.0e-04 loss: 8.006e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3968/30000 [8:57:05<70:58:28,  9.82s/it, lr: 1.0e-04 loss: 8.006e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3968/30000 [8:57:05<70:58:28,  9.82s/it, lr: 1.0e-04 loss: 8.006e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3968/30000 [8:57:16<70:58:28,  9.82s/it, lr: 1.0e-04 loss: 3.771e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3968/30000 [8:57:16<70:58:28,  9.82s/it, lr: 1.0e-04 loss: 3.771e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3969/30000 [8:57:16<73:30:54, 10.17s/it, lr: 1.0e-04 loss: 3.771e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3969/30000 [8:57:16<73:30:54, 10.17s/it, lr: 1.0e-04 loss: 3.771e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3969/30000 [8:57:24<73:30:54, 10.17s/it, lr: 1.0e-04 loss: 7.908e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3969/30000 [8:57:24<73:30:54, 10.17s/it, lr: 1.0e-04 loss: 7.908e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3970/30000 [8:57:34<73:30:44, 10.17s/it, lr: 1.0e-04 loss: 6.044e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3970/30000 [8:57:34<73:30:44, 10.17s/it, lr: 1.0e-04 loss: 6.044e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3971/30000 [8:57:34<56:05:23,  7.76s/it, lr: 1.0e-04 loss: 6.044e-02]
Xiang_wan_2_2_lora:  13%|#####1                                 | 3971/30000 [8:57:34<56:05:23,  7.76s/it, lr: 1.0e-04 loss: 6.044e-02]
\ No newline at end of file