Upload RT-DETRv2 voucher classifier
Browse files- README.md +14 -14
- checkpoint-1600/config.json +129 -0
- checkpoint-1600/model.safetensors +3 -0
- checkpoint-1600/optimizer.pt +3 -0
- checkpoint-1600/preprocessor_config.json +26 -0
- checkpoint-1600/rng_state.pth +3 -0
- checkpoint-1600/scheduler.pt +3 -0
- checkpoint-1600/trainer_state.json +258 -0
- checkpoint-1600/training_args.bin +3 -0
- checkpoint-1800/config.json +129 -0
- checkpoint-1800/model.safetensors +3 -0
- checkpoint-1800/optimizer.pt +3 -0
- checkpoint-1800/preprocessor_config.json +26 -0
- checkpoint-1800/rng_state.pth +3 -0
- checkpoint-1800/scheduler.pt +3 -0
- checkpoint-1800/trainer_state.json +286 -0
- checkpoint-1800/training_args.bin +3 -0
- checkpoint-2000/model.safetensors +1 -1
- checkpoint-2000/optimizer.pt +1 -1
- checkpoint-2000/rng_state.pth +1 -1
- checkpoint-2000/trainer_state.json +179 -2699
- checkpoint-2000/training_args.bin +1 -1
- checkpoint-2200/config.json +129 -0
- checkpoint-2200/model.safetensors +3 -0
- checkpoint-2200/optimizer.pt +3 -0
- checkpoint-2200/preprocessor_config.json +26 -0
- checkpoint-2200/rng_state.pth +3 -0
- checkpoint-2200/scheduler.pt +3 -0
- checkpoint-2200/trainer_state.json +342 -0
- checkpoint-2200/training_args.bin +3 -0
- checkpoint-2250/config.json +129 -0
- checkpoint-2250/model.safetensors +3 -0
- checkpoint-2250/optimizer.pt +3 -0
- checkpoint-2250/preprocessor_config.json +26 -0
- checkpoint-2250/rng_state.pth +3 -0
- checkpoint-2250/scheduler.pt +3 -0
- checkpoint-2250/trainer_state.json +349 -0
- checkpoint-2250/training_args.bin +3 -0
- model.safetensors +1 -1
- runs/Aug14_17-42-57_2676026c4495/events.out.tfevents.1755193378.2676026c4495.6591.0 +3 -0
- training_args.bin +1 -1
README.md
CHANGED
|
@@ -36,11 +36,11 @@ This model is a fine-tuned version of [PekingU/rtdetr_v2_r101vd](https://hugging
|
|
| 36 |
### Training Details
|
| 37 |
|
| 38 |
**Training Dataset:**
|
| 39 |
-
- **Total Samples**:
|
| 40 |
- **Class Distribution**:
|
| 41 |
-
- **
|
| 42 |
-
- **
|
| 43 |
-
- **
|
| 44 |
|
| 45 |
|
| 46 |
**Training Configuration:**
|
|
@@ -91,15 +91,15 @@ This model is a fine-tuned version of [PekingU/rtdetr_v2_r101vd](https://hugging
|
|
| 91 |
- **Tesoreria receipts mean confidence**: 0.0000 (low)
|
| 92 |
|
| 93 |
**Performance by Object Size:**
|
| 94 |
-
- **Small objects**:
|
| 95 |
-
- **Medium objects**:
|
| 96 |
- **Large objects**: 0.0000
|
| 97 |
|
| 98 |
**Evaluation Dataset:**
|
| 99 |
-
- **Digital invoices**: 157 samples (28.
|
| 100 |
-
- **Fisico receipts**:
|
| 101 |
-
- **Tesoreria receipts**: 133 samples (24.
|
| 102 |
-
- **Total evaluation samples**:
|
| 103 |
|
| 104 |
**Model Configuration:**
|
| 105 |
- **Base model**: PekingU/rtdetr_v2_r101vd
|
|
@@ -114,16 +114,16 @@ This model is a fine-tuned version of [PekingU/rtdetr_v2_r101vd](https://hugging
|
|
| 114 |
- **RAM**: 83.5 GB
|
| 115 |
- **GPU configuration**: A100 optimized
|
| 116 |
|
| 117 |
-
**Training Time**: 1.
|
| 118 |
|
| 119 |
**Training Summary:**
|
| 120 |
-
- **Final training loss**:
|
| 121 |
-
- **Final learning rate**:
|
| 122 |
|
| 123 |
|
| 124 |
### MLflow Tracking
|
| 125 |
|
| 126 |
-
- **MLflow Run ID**:
|
| 127 |
- **MLflow Experiment**: RT-DETRv2_Voucher_Classification
|
| 128 |
|
| 129 |
|
|
|
|
| 36 |
### Training Details
|
| 37 |
|
| 38 |
**Training Dataset:**
|
| 39 |
+
- **Total Samples**: 2178
|
| 40 |
- **Class Distribution**:
|
| 41 |
+
- **digital** (id: 0): 626 samples (28.7%)
|
| 42 |
+
- **tesoreria** (id: 2): 537 samples (24.7%)
|
| 43 |
+
- **fisico** (id: 1): 1015 samples (46.6%)
|
| 44 |
|
| 45 |
|
| 46 |
**Training Configuration:**
|
|
|
|
| 91 |
- **Tesoreria receipts mean confidence**: 0.0000 (low)
|
| 92 |
|
| 93 |
**Performance by Object Size:**
|
| 94 |
+
- **Small objects**: -1.0000
|
| 95 |
+
- **Medium objects**: 0.0000
|
| 96 |
- **Large objects**: 0.0000
|
| 97 |
|
| 98 |
**Evaluation Dataset:**
|
| 99 |
+
- **Digital invoices**: 157 samples (28.8%)
|
| 100 |
+
- **Fisico receipts**: 255 samples (46.8%)
|
| 101 |
+
- **Tesoreria receipts**: 133 samples (24.4%)
|
| 102 |
+
- **Total evaluation samples**: 545
|
| 103 |
|
| 104 |
**Model Configuration:**
|
| 105 |
- **Base model**: PekingU/rtdetr_v2_r101vd
|
|
|
|
| 114 |
- **RAM**: 83.5 GB
|
| 115 |
- **GPU configuration**: A100 optimized
|
| 116 |
|
| 117 |
+
**Training Time**: 1.18 hours
|
| 118 |
|
| 119 |
**Training Summary:**
|
| 120 |
+
- **Final training loss**: 196.0164
|
| 121 |
+
- **Final learning rate**: 6.00e-08
|
| 122 |
|
| 123 |
|
| 124 |
### MLflow Tracking
|
| 125 |
|
| 126 |
+
- **MLflow Run ID**: fe7bd26bd1b344c292d9b485139be46c
|
| 127 |
- **MLflow Experiment**: RT-DETRv2_Voucher_Classification
|
| 128 |
|
| 129 |
|
checkpoint-1600/config.json
ADDED
|
@@ -0,0 +1,129 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"activation_dropout": 0.0,
|
| 3 |
+
"activation_function": "silu",
|
| 4 |
+
"anchor_image_size": null,
|
| 5 |
+
"architectures": [
|
| 6 |
+
"RTDetrV2ForObjectDetection"
|
| 7 |
+
],
|
| 8 |
+
"attention_dropout": 0.0,
|
| 9 |
+
"auxiliary_loss": true,
|
| 10 |
+
"backbone": null,
|
| 11 |
+
"backbone_config": {
|
| 12 |
+
"depths": [
|
| 13 |
+
3,
|
| 14 |
+
4,
|
| 15 |
+
23,
|
| 16 |
+
3
|
| 17 |
+
],
|
| 18 |
+
"downsample_in_bottleneck": false,
|
| 19 |
+
"downsample_in_first_stage": false,
|
| 20 |
+
"embedding_size": 64,
|
| 21 |
+
"hidden_act": "relu",
|
| 22 |
+
"hidden_sizes": [
|
| 23 |
+
256,
|
| 24 |
+
512,
|
| 25 |
+
1024,
|
| 26 |
+
2048
|
| 27 |
+
],
|
| 28 |
+
"layer_type": "bottleneck",
|
| 29 |
+
"model_type": "rt_detr_resnet",
|
| 30 |
+
"num_channels": 3,
|
| 31 |
+
"out_features": [
|
| 32 |
+
"stage2",
|
| 33 |
+
"stage3",
|
| 34 |
+
"stage4"
|
| 35 |
+
],
|
| 36 |
+
"out_indices": [
|
| 37 |
+
2,
|
| 38 |
+
3,
|
| 39 |
+
4
|
| 40 |
+
],
|
| 41 |
+
"stage_names": [
|
| 42 |
+
"stem",
|
| 43 |
+
"stage1",
|
| 44 |
+
"stage2",
|
| 45 |
+
"stage3",
|
| 46 |
+
"stage4"
|
| 47 |
+
],
|
| 48 |
+
"torch_dtype": "float32"
|
| 49 |
+
},
|
| 50 |
+
"backbone_kwargs": null,
|
| 51 |
+
"batch_norm_eps": 1e-05,
|
| 52 |
+
"box_noise_scale": 1.0,
|
| 53 |
+
"d_model": 256,
|
| 54 |
+
"decoder_activation_function": "relu",
|
| 55 |
+
"decoder_attention_heads": 8,
|
| 56 |
+
"decoder_ffn_dim": 1024,
|
| 57 |
+
"decoder_in_channels": [
|
| 58 |
+
384,
|
| 59 |
+
384,
|
| 60 |
+
384
|
| 61 |
+
],
|
| 62 |
+
"decoder_layers": 6,
|
| 63 |
+
"decoder_method": "default",
|
| 64 |
+
"decoder_n_levels": 3,
|
| 65 |
+
"decoder_n_points": 4,
|
| 66 |
+
"decoder_offset_scale": 0.5,
|
| 67 |
+
"disable_custom_kernels": true,
|
| 68 |
+
"dropout": 0.0,
|
| 69 |
+
"encode_proj_layers": [
|
| 70 |
+
2
|
| 71 |
+
],
|
| 72 |
+
"encoder_activation_function": "gelu",
|
| 73 |
+
"encoder_attention_heads": 8,
|
| 74 |
+
"encoder_ffn_dim": 2048,
|
| 75 |
+
"encoder_hidden_dim": 384,
|
| 76 |
+
"encoder_in_channels": [
|
| 77 |
+
512,
|
| 78 |
+
1024,
|
| 79 |
+
2048
|
| 80 |
+
],
|
| 81 |
+
"encoder_layers": 1,
|
| 82 |
+
"eos_coefficient": 0.0001,
|
| 83 |
+
"eval_size": null,
|
| 84 |
+
"feat_strides": [
|
| 85 |
+
8,
|
| 86 |
+
16,
|
| 87 |
+
32
|
| 88 |
+
],
|
| 89 |
+
"focal_loss_alpha": 0.75,
|
| 90 |
+
"focal_loss_gamma": 2.0,
|
| 91 |
+
"freeze_backbone_batch_norms": true,
|
| 92 |
+
"hidden_expansion": 1.0,
|
| 93 |
+
"id2label": {
|
| 94 |
+
"0": "LABEL_0",
|
| 95 |
+
"1": "LABEL_1",
|
| 96 |
+
"2": "LABEL_2"
|
| 97 |
+
},
|
| 98 |
+
"initializer_bias_prior_prob": null,
|
| 99 |
+
"initializer_range": 0.01,
|
| 100 |
+
"is_encoder_decoder": true,
|
| 101 |
+
"label2id": {
|
| 102 |
+
"LABEL_0": 0,
|
| 103 |
+
"LABEL_1": 1,
|
| 104 |
+
"LABEL_2": 2
|
| 105 |
+
},
|
| 106 |
+
"label_noise_ratio": 0.5,
|
| 107 |
+
"layer_norm_eps": 1e-05,
|
| 108 |
+
"learn_initial_query": false,
|
| 109 |
+
"matcher_alpha": 0.25,
|
| 110 |
+
"matcher_bbox_cost": 5.0,
|
| 111 |
+
"matcher_class_cost": 2.0,
|
| 112 |
+
"matcher_gamma": 2.0,
|
| 113 |
+
"matcher_giou_cost": 2.0,
|
| 114 |
+
"model_type": "rt_detr_v2",
|
| 115 |
+
"normalize_before": false,
|
| 116 |
+
"num_denoising": 100,
|
| 117 |
+
"num_feature_levels": 3,
|
| 118 |
+
"num_queries": 300,
|
| 119 |
+
"positional_encoding_temperature": 10000,
|
| 120 |
+
"torch_dtype": "float32",
|
| 121 |
+
"transformers_version": "4.55.0",
|
| 122 |
+
"use_focal_loss": true,
|
| 123 |
+
"use_pretrained_backbone": false,
|
| 124 |
+
"use_timm_backbone": false,
|
| 125 |
+
"weight_loss_bbox": 5.0,
|
| 126 |
+
"weight_loss_giou": 2.0,
|
| 127 |
+
"weight_loss_vfl": 1.0,
|
| 128 |
+
"with_box_refine": true
|
| 129 |
+
}
|
checkpoint-1600/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:226b3771abee1cf8215eee8ff3a6c866930ba69be4144e554d6ad8d45317d5a4
|
| 3 |
+
size 306699044
|
checkpoint-1600/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e429f0d5be726eb26d197c034148dc43ca3de0a5498a36ae73ffb251656d55e0
|
| 3 |
+
size 611580433
|
checkpoint-1600/preprocessor_config.json
ADDED
|
@@ -0,0 +1,26 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"do_convert_annotations": true,
|
| 3 |
+
"do_normalize": false,
|
| 4 |
+
"do_pad": false,
|
| 5 |
+
"do_rescale": true,
|
| 6 |
+
"do_resize": true,
|
| 7 |
+
"format": "coco_detection",
|
| 8 |
+
"image_mean": [
|
| 9 |
+
0.485,
|
| 10 |
+
0.456,
|
| 11 |
+
0.406
|
| 12 |
+
],
|
| 13 |
+
"image_processor_type": "RTDetrImageProcessor",
|
| 14 |
+
"image_std": [
|
| 15 |
+
0.229,
|
| 16 |
+
0.224,
|
| 17 |
+
0.225
|
| 18 |
+
],
|
| 19 |
+
"pad_size": null,
|
| 20 |
+
"resample": 2,
|
| 21 |
+
"rescale_factor": 0.00392156862745098,
|
| 22 |
+
"size": {
|
| 23 |
+
"height": 640,
|
| 24 |
+
"width": 640
|
| 25 |
+
}
|
| 26 |
+
}
|
checkpoint-1600/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0ec603acfcf0594fe5e0f4a5b622df7eaf620ca0be75f37087ee07cf6a8bc746
|
| 3 |
+
size 14244
|
checkpoint-1600/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:294a2ee5d41cd38781c829986780523b3ca2c9ddcff37ad7f87c59ebe379a29a
|
| 3 |
+
size 1064
|
checkpoint-1600/trainer_state.json
ADDED
|
@@ -0,0 +1,258 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 35.55555555555556,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 1600,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 1.1111111111111112,
|
| 14 |
+
"grad_norm": 3663.33642578125,
|
| 15 |
+
"learning_rate": 3.6750000000000003e-07,
|
| 16 |
+
"loss": 1356.1239,
|
| 17 |
+
"step": 50
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 2.2222222222222223,
|
| 21 |
+
"grad_norm": 3973.032470703125,
|
| 22 |
+
"learning_rate": 7.425000000000001e-07,
|
| 23 |
+
"loss": 1275.5178,
|
| 24 |
+
"step": 100
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 3.3333333333333335,
|
| 28 |
+
"grad_norm": 2667.0068359375,
|
| 29 |
+
"learning_rate": 1.1174999999999999e-06,
|
| 30 |
+
"loss": 1123.6059,
|
| 31 |
+
"step": 150
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 4.444444444444445,
|
| 35 |
+
"grad_norm": 2320.48486328125,
|
| 36 |
+
"learning_rate": 1.4925000000000001e-06,
|
| 37 |
+
"loss": 922.8917,
|
| 38 |
+
"step": 200
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 5.555555555555555,
|
| 42 |
+
"grad_norm": 1611.3345947265625,
|
| 43 |
+
"learning_rate": 1.8675000000000001e-06,
|
| 44 |
+
"loss": 714.5638,
|
| 45 |
+
"step": 250
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 6.666666666666667,
|
| 49 |
+
"grad_norm": 1395.7064208984375,
|
| 50 |
+
"learning_rate": 2.2425e-06,
|
| 51 |
+
"loss": 542.1251,
|
| 52 |
+
"step": 300
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 7.777777777777778,
|
| 56 |
+
"grad_norm": 1019.4860229492188,
|
| 57 |
+
"learning_rate": 2.6175e-06,
|
| 58 |
+
"loss": 411.1387,
|
| 59 |
+
"step": 350
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 8.88888888888889,
|
| 63 |
+
"grad_norm": 888.315185546875,
|
| 64 |
+
"learning_rate": 2.9925e-06,
|
| 65 |
+
"loss": 318.4567,
|
| 66 |
+
"step": 400
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 10.0,
|
| 70 |
+
"grad_norm": 2590.113525390625,
|
| 71 |
+
"learning_rate": 3.3675000000000004e-06,
|
| 72 |
+
"loss": 261.6995,
|
| 73 |
+
"step": 450
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 11.11111111111111,
|
| 77 |
+
"grad_norm": 711.677734375,
|
| 78 |
+
"learning_rate": 3.7425e-06,
|
| 79 |
+
"loss": 220.2936,
|
| 80 |
+
"step": 500
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 12.222222222222221,
|
| 84 |
+
"grad_norm": 548.9371948242188,
|
| 85 |
+
"learning_rate": 4.117500000000001e-06,
|
| 86 |
+
"loss": 187.9833,
|
| 87 |
+
"step": 550
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 13.333333333333334,
|
| 91 |
+
"grad_norm": 1127.1611328125,
|
| 92 |
+
"learning_rate": 4.4925e-06,
|
| 93 |
+
"loss": 159.1351,
|
| 94 |
+
"step": 600
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 14.444444444444445,
|
| 98 |
+
"grad_norm": 426.074951171875,
|
| 99 |
+
"learning_rate": 4.8675e-06,
|
| 100 |
+
"loss": 137.1092,
|
| 101 |
+
"step": 650
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 15.555555555555555,
|
| 105 |
+
"grad_norm": 348.842529296875,
|
| 106 |
+
"learning_rate": 5.2425e-06,
|
| 107 |
+
"loss": 119.822,
|
| 108 |
+
"step": 700
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 16.666666666666668,
|
| 112 |
+
"grad_norm": 373.2169189453125,
|
| 113 |
+
"learning_rate": 5.6175e-06,
|
| 114 |
+
"loss": 104.3366,
|
| 115 |
+
"step": 750
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 17.77777777777778,
|
| 119 |
+
"grad_norm": 324.51702880859375,
|
| 120 |
+
"learning_rate": 5.992500000000001e-06,
|
| 121 |
+
"loss": 90.8788,
|
| 122 |
+
"step": 800
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 18.88888888888889,
|
| 126 |
+
"grad_norm": 269.91827392578125,
|
| 127 |
+
"learning_rate": 6.3675e-06,
|
| 128 |
+
"loss": 78.4644,
|
| 129 |
+
"step": 850
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 20.0,
|
| 133 |
+
"grad_norm": 1744.54052734375,
|
| 134 |
+
"learning_rate": 6.7425e-06,
|
| 135 |
+
"loss": 70.3526,
|
| 136 |
+
"step": 900
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 21.11111111111111,
|
| 140 |
+
"grad_norm": 369.39837646484375,
|
| 141 |
+
"learning_rate": 7.1175e-06,
|
| 142 |
+
"loss": 63.9417,
|
| 143 |
+
"step": 950
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 22.22222222222222,
|
| 147 |
+
"grad_norm": 303.95977783203125,
|
| 148 |
+
"learning_rate": 7.4925e-06,
|
| 149 |
+
"loss": 63.4575,
|
| 150 |
+
"step": 1000
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 23.333333333333332,
|
| 154 |
+
"grad_norm": 187.462890625,
|
| 155 |
+
"learning_rate": 7.8675e-06,
|
| 156 |
+
"loss": 54.7417,
|
| 157 |
+
"step": 1050
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 24.444444444444443,
|
| 161 |
+
"grad_norm": 165.56666564941406,
|
| 162 |
+
"learning_rate": 8.2425e-06,
|
| 163 |
+
"loss": 49.6842,
|
| 164 |
+
"step": 1100
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 25.555555555555557,
|
| 168 |
+
"grad_norm": 147.3148193359375,
|
| 169 |
+
"learning_rate": 8.6175e-06,
|
| 170 |
+
"loss": 43.027,
|
| 171 |
+
"step": 1150
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 26.666666666666668,
|
| 175 |
+
"grad_norm": 125.53775024414062,
|
| 176 |
+
"learning_rate": 8.9925e-06,
|
| 177 |
+
"loss": 38.2579,
|
| 178 |
+
"step": 1200
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 27.77777777777778,
|
| 182 |
+
"grad_norm": 109.85810089111328,
|
| 183 |
+
"learning_rate": 9.367500000000001e-06,
|
| 184 |
+
"loss": 34.3957,
|
| 185 |
+
"step": 1250
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 28.88888888888889,
|
| 189 |
+
"grad_norm": 98.328369140625,
|
| 190 |
+
"learning_rate": 9.7425e-06,
|
| 191 |
+
"loss": 31.4378,
|
| 192 |
+
"step": 1300
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 30.0,
|
| 196 |
+
"grad_norm": 109.77750396728516,
|
| 197 |
+
"learning_rate": 1.01175e-05,
|
| 198 |
+
"loss": 28.5084,
|
| 199 |
+
"step": 1350
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 31.11111111111111,
|
| 203 |
+
"grad_norm": 112.47483825683594,
|
| 204 |
+
"learning_rate": 1.04925e-05,
|
| 205 |
+
"loss": 26.1671,
|
| 206 |
+
"step": 1400
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 32.22222222222222,
|
| 210 |
+
"grad_norm": 85.60242462158203,
|
| 211 |
+
"learning_rate": 1.08675e-05,
|
| 212 |
+
"loss": 24.2309,
|
| 213 |
+
"step": 1450
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 33.333333333333336,
|
| 217 |
+
"grad_norm": 73.19799041748047,
|
| 218 |
+
"learning_rate": 1.1242500000000001e-05,
|
| 219 |
+
"loss": 22.6248,
|
| 220 |
+
"step": 1500
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 34.44444444444444,
|
| 224 |
+
"grad_norm": 80.70884704589844,
|
| 225 |
+
"learning_rate": 1.16175e-05,
|
| 226 |
+
"loss": 21.1187,
|
| 227 |
+
"step": 1550
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 35.55555555555556,
|
| 231 |
+
"grad_norm": 110.98326110839844,
|
| 232 |
+
"learning_rate": 1.19925e-05,
|
| 233 |
+
"loss": 20.4828,
|
| 234 |
+
"step": 1600
|
| 235 |
+
}
|
| 236 |
+
],
|
| 237 |
+
"logging_steps": 50,
|
| 238 |
+
"max_steps": 2250,
|
| 239 |
+
"num_input_tokens_seen": 0,
|
| 240 |
+
"num_train_epochs": 50,
|
| 241 |
+
"save_steps": 200,
|
| 242 |
+
"stateful_callbacks": {
|
| 243 |
+
"TrainerControl": {
|
| 244 |
+
"args": {
|
| 245 |
+
"should_epoch_stop": false,
|
| 246 |
+
"should_evaluate": false,
|
| 247 |
+
"should_log": false,
|
| 248 |
+
"should_save": true,
|
| 249 |
+
"should_training_stop": false
|
| 250 |
+
},
|
| 251 |
+
"attributes": {}
|
| 252 |
+
}
|
| 253 |
+
},
|
| 254 |
+
"total_flos": 4.31006450959319e+19,
|
| 255 |
+
"train_batch_size": 24,
|
| 256 |
+
"trial_name": null,
|
| 257 |
+
"trial_params": null
|
| 258 |
+
}
|
checkpoint-1600/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a7b84367094b7487f77de50fba614a6c6667e9cf018b77ee5bfc158268fc5eaf
|
| 3 |
+
size 5368
|
checkpoint-1800/config.json
ADDED
|
@@ -0,0 +1,129 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"activation_dropout": 0.0,
|
| 3 |
+
"activation_function": "silu",
|
| 4 |
+
"anchor_image_size": null,
|
| 5 |
+
"architectures": [
|
| 6 |
+
"RTDetrV2ForObjectDetection"
|
| 7 |
+
],
|
| 8 |
+
"attention_dropout": 0.0,
|
| 9 |
+
"auxiliary_loss": true,
|
| 10 |
+
"backbone": null,
|
| 11 |
+
"backbone_config": {
|
| 12 |
+
"depths": [
|
| 13 |
+
3,
|
| 14 |
+
4,
|
| 15 |
+
23,
|
| 16 |
+
3
|
| 17 |
+
],
|
| 18 |
+
"downsample_in_bottleneck": false,
|
| 19 |
+
"downsample_in_first_stage": false,
|
| 20 |
+
"embedding_size": 64,
|
| 21 |
+
"hidden_act": "relu",
|
| 22 |
+
"hidden_sizes": [
|
| 23 |
+
256,
|
| 24 |
+
512,
|
| 25 |
+
1024,
|
| 26 |
+
2048
|
| 27 |
+
],
|
| 28 |
+
"layer_type": "bottleneck",
|
| 29 |
+
"model_type": "rt_detr_resnet",
|
| 30 |
+
"num_channels": 3,
|
| 31 |
+
"out_features": [
|
| 32 |
+
"stage2",
|
| 33 |
+
"stage3",
|
| 34 |
+
"stage4"
|
| 35 |
+
],
|
| 36 |
+
"out_indices": [
|
| 37 |
+
2,
|
| 38 |
+
3,
|
| 39 |
+
4
|
| 40 |
+
],
|
| 41 |
+
"stage_names": [
|
| 42 |
+
"stem",
|
| 43 |
+
"stage1",
|
| 44 |
+
"stage2",
|
| 45 |
+
"stage3",
|
| 46 |
+
"stage4"
|
| 47 |
+
],
|
| 48 |
+
"torch_dtype": "float32"
|
| 49 |
+
},
|
| 50 |
+
"backbone_kwargs": null,
|
| 51 |
+
"batch_norm_eps": 1e-05,
|
| 52 |
+
"box_noise_scale": 1.0,
|
| 53 |
+
"d_model": 256,
|
| 54 |
+
"decoder_activation_function": "relu",
|
| 55 |
+
"decoder_attention_heads": 8,
|
| 56 |
+
"decoder_ffn_dim": 1024,
|
| 57 |
+
"decoder_in_channels": [
|
| 58 |
+
384,
|
| 59 |
+
384,
|
| 60 |
+
384
|
| 61 |
+
],
|
| 62 |
+
"decoder_layers": 6,
|
| 63 |
+
"decoder_method": "default",
|
| 64 |
+
"decoder_n_levels": 3,
|
| 65 |
+
"decoder_n_points": 4,
|
| 66 |
+
"decoder_offset_scale": 0.5,
|
| 67 |
+
"disable_custom_kernels": true,
|
| 68 |
+
"dropout": 0.0,
|
| 69 |
+
"encode_proj_layers": [
|
| 70 |
+
2
|
| 71 |
+
],
|
| 72 |
+
"encoder_activation_function": "gelu",
|
| 73 |
+
"encoder_attention_heads": 8,
|
| 74 |
+
"encoder_ffn_dim": 2048,
|
| 75 |
+
"encoder_hidden_dim": 384,
|
| 76 |
+
"encoder_in_channels": [
|
| 77 |
+
512,
|
| 78 |
+
1024,
|
| 79 |
+
2048
|
| 80 |
+
],
|
| 81 |
+
"encoder_layers": 1,
|
| 82 |
+
"eos_coefficient": 0.0001,
|
| 83 |
+
"eval_size": null,
|
| 84 |
+
"feat_strides": [
|
| 85 |
+
8,
|
| 86 |
+
16,
|
| 87 |
+
32
|
| 88 |
+
],
|
| 89 |
+
"focal_loss_alpha": 0.75,
|
| 90 |
+
"focal_loss_gamma": 2.0,
|
| 91 |
+
"freeze_backbone_batch_norms": true,
|
| 92 |
+
"hidden_expansion": 1.0,
|
| 93 |
+
"id2label": {
|
| 94 |
+
"0": "LABEL_0",
|
| 95 |
+
"1": "LABEL_1",
|
| 96 |
+
"2": "LABEL_2"
|
| 97 |
+
},
|
| 98 |
+
"initializer_bias_prior_prob": null,
|
| 99 |
+
"initializer_range": 0.01,
|
| 100 |
+
"is_encoder_decoder": true,
|
| 101 |
+
"label2id": {
|
| 102 |
+
"LABEL_0": 0,
|
| 103 |
+
"LABEL_1": 1,
|
| 104 |
+
"LABEL_2": 2
|
| 105 |
+
},
|
| 106 |
+
"label_noise_ratio": 0.5,
|
| 107 |
+
"layer_norm_eps": 1e-05,
|
| 108 |
+
"learn_initial_query": false,
|
| 109 |
+
"matcher_alpha": 0.25,
|
| 110 |
+
"matcher_bbox_cost": 5.0,
|
| 111 |
+
"matcher_class_cost": 2.0,
|
| 112 |
+
"matcher_gamma": 2.0,
|
| 113 |
+
"matcher_giou_cost": 2.0,
|
| 114 |
+
"model_type": "rt_detr_v2",
|
| 115 |
+
"normalize_before": false,
|
| 116 |
+
"num_denoising": 100,
|
| 117 |
+
"num_feature_levels": 3,
|
| 118 |
+
"num_queries": 300,
|
| 119 |
+
"positional_encoding_temperature": 10000,
|
| 120 |
+
"torch_dtype": "float32",
|
| 121 |
+
"transformers_version": "4.55.0",
|
| 122 |
+
"use_focal_loss": true,
|
| 123 |
+
"use_pretrained_backbone": false,
|
| 124 |
+
"use_timm_backbone": false,
|
| 125 |
+
"weight_loss_bbox": 5.0,
|
| 126 |
+
"weight_loss_giou": 2.0,
|
| 127 |
+
"weight_loss_vfl": 1.0,
|
| 128 |
+
"with_box_refine": true
|
| 129 |
+
}
|
checkpoint-1800/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:07ca2b0b335e81ed89a2f7f930b576eee925506e057c93002392f1e2376180f0
|
| 3 |
+
size 306699044
|
checkpoint-1800/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9bd5f81a1ac758d99c2a50f00a75bf78c82cda77f464f12cdda567e7a65306d7
|
| 3 |
+
size 611580433
|
checkpoint-1800/preprocessor_config.json
ADDED
|
@@ -0,0 +1,26 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"do_convert_annotations": true,
|
| 3 |
+
"do_normalize": false,
|
| 4 |
+
"do_pad": false,
|
| 5 |
+
"do_rescale": true,
|
| 6 |
+
"do_resize": true,
|
| 7 |
+
"format": "coco_detection",
|
| 8 |
+
"image_mean": [
|
| 9 |
+
0.485,
|
| 10 |
+
0.456,
|
| 11 |
+
0.406
|
| 12 |
+
],
|
| 13 |
+
"image_processor_type": "RTDetrImageProcessor",
|
| 14 |
+
"image_std": [
|
| 15 |
+
0.229,
|
| 16 |
+
0.224,
|
| 17 |
+
0.225
|
| 18 |
+
],
|
| 19 |
+
"pad_size": null,
|
| 20 |
+
"resample": 2,
|
| 21 |
+
"rescale_factor": 0.00392156862745098,
|
| 22 |
+
"size": {
|
| 23 |
+
"height": 640,
|
| 24 |
+
"width": 640
|
| 25 |
+
}
|
| 26 |
+
}
|
checkpoint-1800/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a0f22c99e3d52b982805eed64c46ed79d1c93982c6433c732f7f4d4a2ab58f93
|
| 3 |
+
size 14244
|
checkpoint-1800/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c7ef919f119174a0bac8b992747548bc0ab5a6be59a5cfc5a6c8633fade536fa
|
| 3 |
+
size 1064
|
checkpoint-1800/trainer_state.json
ADDED
|
@@ -0,0 +1,286 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 40.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 1800,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 1.1111111111111112,
|
| 14 |
+
"grad_norm": 3663.33642578125,
|
| 15 |
+
"learning_rate": 3.6750000000000003e-07,
|
| 16 |
+
"loss": 1356.1239,
|
| 17 |
+
"step": 50
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 2.2222222222222223,
|
| 21 |
+
"grad_norm": 3973.032470703125,
|
| 22 |
+
"learning_rate": 7.425000000000001e-07,
|
| 23 |
+
"loss": 1275.5178,
|
| 24 |
+
"step": 100
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 3.3333333333333335,
|
| 28 |
+
"grad_norm": 2667.0068359375,
|
| 29 |
+
"learning_rate": 1.1174999999999999e-06,
|
| 30 |
+
"loss": 1123.6059,
|
| 31 |
+
"step": 150
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 4.444444444444445,
|
| 35 |
+
"grad_norm": 2320.48486328125,
|
| 36 |
+
"learning_rate": 1.4925000000000001e-06,
|
| 37 |
+
"loss": 922.8917,
|
| 38 |
+
"step": 200
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 5.555555555555555,
|
| 42 |
+
"grad_norm": 1611.3345947265625,
|
| 43 |
+
"learning_rate": 1.8675000000000001e-06,
|
| 44 |
+
"loss": 714.5638,
|
| 45 |
+
"step": 250
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 6.666666666666667,
|
| 49 |
+
"grad_norm": 1395.7064208984375,
|
| 50 |
+
"learning_rate": 2.2425e-06,
|
| 51 |
+
"loss": 542.1251,
|
| 52 |
+
"step": 300
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 7.777777777777778,
|
| 56 |
+
"grad_norm": 1019.4860229492188,
|
| 57 |
+
"learning_rate": 2.6175e-06,
|
| 58 |
+
"loss": 411.1387,
|
| 59 |
+
"step": 350
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 8.88888888888889,
|
| 63 |
+
"grad_norm": 888.315185546875,
|
| 64 |
+
"learning_rate": 2.9925e-06,
|
| 65 |
+
"loss": 318.4567,
|
| 66 |
+
"step": 400
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 10.0,
|
| 70 |
+
"grad_norm": 2590.113525390625,
|
| 71 |
+
"learning_rate": 3.3675000000000004e-06,
|
| 72 |
+
"loss": 261.6995,
|
| 73 |
+
"step": 450
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 11.11111111111111,
|
| 77 |
+
"grad_norm": 711.677734375,
|
| 78 |
+
"learning_rate": 3.7425e-06,
|
| 79 |
+
"loss": 220.2936,
|
| 80 |
+
"step": 500
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 12.222222222222221,
|
| 84 |
+
"grad_norm": 548.9371948242188,
|
| 85 |
+
"learning_rate": 4.117500000000001e-06,
|
| 86 |
+
"loss": 187.9833,
|
| 87 |
+
"step": 550
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 13.333333333333334,
|
| 91 |
+
"grad_norm": 1127.1611328125,
|
| 92 |
+
"learning_rate": 4.4925e-06,
|
| 93 |
+
"loss": 159.1351,
|
| 94 |
+
"step": 600
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 14.444444444444445,
|
| 98 |
+
"grad_norm": 426.074951171875,
|
| 99 |
+
"learning_rate": 4.8675e-06,
|
| 100 |
+
"loss": 137.1092,
|
| 101 |
+
"step": 650
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 15.555555555555555,
|
| 105 |
+
"grad_norm": 348.842529296875,
|
| 106 |
+
"learning_rate": 5.2425e-06,
|
| 107 |
+
"loss": 119.822,
|
| 108 |
+
"step": 700
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 16.666666666666668,
|
| 112 |
+
"grad_norm": 373.2169189453125,
|
| 113 |
+
"learning_rate": 5.6175e-06,
|
| 114 |
+
"loss": 104.3366,
|
| 115 |
+
"step": 750
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 17.77777777777778,
|
| 119 |
+
"grad_norm": 324.51702880859375,
|
| 120 |
+
"learning_rate": 5.992500000000001e-06,
|
| 121 |
+
"loss": 90.8788,
|
| 122 |
+
"step": 800
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 18.88888888888889,
|
| 126 |
+
"grad_norm": 269.91827392578125,
|
| 127 |
+
"learning_rate": 6.3675e-06,
|
| 128 |
+
"loss": 78.4644,
|
| 129 |
+
"step": 850
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 20.0,
|
| 133 |
+
"grad_norm": 1744.54052734375,
|
| 134 |
+
"learning_rate": 6.7425e-06,
|
| 135 |
+
"loss": 70.3526,
|
| 136 |
+
"step": 900
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 21.11111111111111,
|
| 140 |
+
"grad_norm": 369.39837646484375,
|
| 141 |
+
"learning_rate": 7.1175e-06,
|
| 142 |
+
"loss": 63.9417,
|
| 143 |
+
"step": 950
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 22.22222222222222,
|
| 147 |
+
"grad_norm": 303.95977783203125,
|
| 148 |
+
"learning_rate": 7.4925e-06,
|
| 149 |
+
"loss": 63.4575,
|
| 150 |
+
"step": 1000
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 23.333333333333332,
|
| 154 |
+
"grad_norm": 187.462890625,
|
| 155 |
+
"learning_rate": 7.8675e-06,
|
| 156 |
+
"loss": 54.7417,
|
| 157 |
+
"step": 1050
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 24.444444444444443,
|
| 161 |
+
"grad_norm": 165.56666564941406,
|
| 162 |
+
"learning_rate": 8.2425e-06,
|
| 163 |
+
"loss": 49.6842,
|
| 164 |
+
"step": 1100
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 25.555555555555557,
|
| 168 |
+
"grad_norm": 147.3148193359375,
|
| 169 |
+
"learning_rate": 8.6175e-06,
|
| 170 |
+
"loss": 43.027,
|
| 171 |
+
"step": 1150
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 26.666666666666668,
|
| 175 |
+
"grad_norm": 125.53775024414062,
|
| 176 |
+
"learning_rate": 8.9925e-06,
|
| 177 |
+
"loss": 38.2579,
|
| 178 |
+
"step": 1200
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 27.77777777777778,
|
| 182 |
+
"grad_norm": 109.85810089111328,
|
| 183 |
+
"learning_rate": 9.367500000000001e-06,
|
| 184 |
+
"loss": 34.3957,
|
| 185 |
+
"step": 1250
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 28.88888888888889,
|
| 189 |
+
"grad_norm": 98.328369140625,
|
| 190 |
+
"learning_rate": 9.7425e-06,
|
| 191 |
+
"loss": 31.4378,
|
| 192 |
+
"step": 1300
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 30.0,
|
| 196 |
+
"grad_norm": 109.77750396728516,
|
| 197 |
+
"learning_rate": 1.01175e-05,
|
| 198 |
+
"loss": 28.5084,
|
| 199 |
+
"step": 1350
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 31.11111111111111,
|
| 203 |
+
"grad_norm": 112.47483825683594,
|
| 204 |
+
"learning_rate": 1.04925e-05,
|
| 205 |
+
"loss": 26.1671,
|
| 206 |
+
"step": 1400
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 32.22222222222222,
|
| 210 |
+
"grad_norm": 85.60242462158203,
|
| 211 |
+
"learning_rate": 1.08675e-05,
|
| 212 |
+
"loss": 24.2309,
|
| 213 |
+
"step": 1450
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 33.333333333333336,
|
| 217 |
+
"grad_norm": 73.19799041748047,
|
| 218 |
+
"learning_rate": 1.1242500000000001e-05,
|
| 219 |
+
"loss": 22.6248,
|
| 220 |
+
"step": 1500
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 34.44444444444444,
|
| 224 |
+
"grad_norm": 80.70884704589844,
|
| 225 |
+
"learning_rate": 1.16175e-05,
|
| 226 |
+
"loss": 21.1187,
|
| 227 |
+
"step": 1550
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 35.55555555555556,
|
| 231 |
+
"grad_norm": 110.98326110839844,
|
| 232 |
+
"learning_rate": 1.19925e-05,
|
| 233 |
+
"loss": 20.4828,
|
| 234 |
+
"step": 1600
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 36.666666666666664,
|
| 238 |
+
"grad_norm": 113.65286254882812,
|
| 239 |
+
"learning_rate": 1.23675e-05,
|
| 240 |
+
"loss": 19.7366,
|
| 241 |
+
"step": 1650
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 37.77777777777778,
|
| 245 |
+
"grad_norm": 77.65855407714844,
|
| 246 |
+
"learning_rate": 1.27425e-05,
|
| 247 |
+
"loss": 18.6632,
|
| 248 |
+
"step": 1700
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 38.888888888888886,
|
| 252 |
+
"grad_norm": 88.96723175048828,
|
| 253 |
+
"learning_rate": 1.3117500000000001e-05,
|
| 254 |
+
"loss": 18.0793,
|
| 255 |
+
"step": 1750
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 40.0,
|
| 259 |
+
"grad_norm": 79.1690902709961,
|
| 260 |
+
"learning_rate": 1.34925e-05,
|
| 261 |
+
"loss": 17.0667,
|
| 262 |
+
"step": 1800
|
| 263 |
+
}
|
| 264 |
+
],
|
| 265 |
+
"logging_steps": 50,
|
| 266 |
+
"max_steps": 2250,
|
| 267 |
+
"num_input_tokens_seen": 0,
|
| 268 |
+
"num_train_epochs": 50,
|
| 269 |
+
"save_steps": 200,
|
| 270 |
+
"stateful_callbacks": {
|
| 271 |
+
"TrainerControl": {
|
| 272 |
+
"args": {
|
| 273 |
+
"should_epoch_stop": false,
|
| 274 |
+
"should_evaluate": false,
|
| 275 |
+
"should_log": false,
|
| 276 |
+
"should_save": true,
|
| 277 |
+
"should_training_stop": false
|
| 278 |
+
},
|
| 279 |
+
"attributes": {}
|
| 280 |
+
}
|
| 281 |
+
},
|
| 282 |
+
"total_flos": 4.848540943358362e+19,
|
| 283 |
+
"train_batch_size": 24,
|
| 284 |
+
"trial_name": null,
|
| 285 |
+
"trial_params": null
|
| 286 |
+
}
|
checkpoint-1800/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a7b84367094b7487f77de50fba614a6c6667e9cf018b77ee5bfc158268fc5eaf
|
| 3 |
+
size 5368
|
checkpoint-2000/model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 306699044
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:74e332bd7772e2e7ca2e34234d7da49a711abdbfb273e02189615d3a85b39c3b
|
| 3 |
size 306699044
|
checkpoint-2000/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 611580433
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:04408b4cb5317b335a1dfa8034af865e68e9535b6db6f797f7454b76a0dbb0fd
|
| 3 |
size 611580433
|
checkpoint-2000/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14244
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bf38b27cb6d82f34ae1038db7a7ddef1794dabfa258a278a90cdfdf859a01777
|
| 3 |
size 14244
|
checkpoint-2000/trainer_state.json
CHANGED
|
@@ -2,7 +2,7 @@
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
-
"epoch":
|
| 6 |
"eval_steps": 500,
|
| 7 |
"global_step": 2000,
|
| 8 |
"is_hyper_param_search": false,
|
|
@@ -10,2811 +10,291 @@
|
|
| 10 |
"is_world_process_zero": true,
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
-
"epoch":
|
| 14 |
-
"grad_norm":
|
| 15 |
-
"learning_rate": 3.0000000000000004e-08,
|
| 16 |
-
"loss": 1383.1007,
|
| 17 |
-
"step": 5
|
| 18 |
-
},
|
| 19 |
-
{
|
| 20 |
-
"epoch": 0.9523809523809523,
|
| 21 |
-
"grad_norm": 12343.9111328125,
|
| 22 |
-
"learning_rate": 6.75e-08,
|
| 23 |
-
"loss": 1349.0135,
|
| 24 |
-
"step": 10
|
| 25 |
-
},
|
| 26 |
-
{
|
| 27 |
-
"epoch": 1.380952380952381,
|
| 28 |
-
"grad_norm": 10210.48046875,
|
| 29 |
-
"learning_rate": 1.05e-07,
|
| 30 |
-
"loss": 1340.3351,
|
| 31 |
-
"step": 15
|
| 32 |
-
},
|
| 33 |
-
{
|
| 34 |
-
"epoch": 1.8571428571428572,
|
| 35 |
-
"grad_norm": 7633.39599609375,
|
| 36 |
-
"learning_rate": 1.425e-07,
|
| 37 |
-
"loss": 1369.3844,
|
| 38 |
-
"step": 20
|
| 39 |
-
},
|
| 40 |
-
{
|
| 41 |
-
"epoch": 0.5494505494505495,
|
| 42 |
-
"grad_norm": 12217.7294921875,
|
| 43 |
-
"learning_rate": 1.8e-07,
|
| 44 |
-
"loss": 1362.663,
|
| 45 |
-
"step": 25
|
| 46 |
-
},
|
| 47 |
-
{
|
| 48 |
-
"epoch": 0.6593406593406593,
|
| 49 |
-
"grad_norm": 31348.806640625,
|
| 50 |
-
"learning_rate": 2.175e-07,
|
| 51 |
-
"loss": 1361.9012,
|
| 52 |
-
"step": 30
|
| 53 |
-
},
|
| 54 |
-
{
|
| 55 |
-
"epoch": 0.7692307692307693,
|
| 56 |
-
"grad_norm": 7385.05322265625,
|
| 57 |
-
"learning_rate": 2.5500000000000005e-07,
|
| 58 |
-
"loss": 1365.5517,
|
| 59 |
-
"step": 35
|
| 60 |
-
},
|
| 61 |
-
{
|
| 62 |
-
"epoch": 0.8791208791208791,
|
| 63 |
-
"grad_norm": 15594.4677734375,
|
| 64 |
-
"learning_rate": 2.925e-07,
|
| 65 |
-
"loss": 1341.6402,
|
| 66 |
-
"step": 40
|
| 67 |
-
},
|
| 68 |
-
{
|
| 69 |
-
"epoch": 0.989010989010989,
|
| 70 |
-
"grad_norm": 6404.94775390625,
|
| 71 |
-
"learning_rate": 3.2999999999999996e-07,
|
| 72 |
-
"loss": 1343.9832,
|
| 73 |
-
"step": 45
|
| 74 |
-
},
|
| 75 |
-
{
|
| 76 |
-
"epoch": 1.10989010989011,
|
| 77 |
-
"grad_norm": 6927.478515625,
|
| 78 |
"learning_rate": 3.6750000000000003e-07,
|
| 79 |
-
"loss":
|
| 80 |
"step": 50
|
| 81 |
},
|
| 82 |
{
|
| 83 |
-
"epoch":
|
| 84 |
-
"grad_norm":
|
| 85 |
-
"learning_rate":
|
| 86 |
-
"loss":
|
| 87 |
-
"step":
|
| 88 |
-
},
|
| 89 |
-
{
|
| 90 |
-
"epoch": 1.3296703296703296,
|
| 91 |
-
"grad_norm": 5830.10693359375,
|
| 92 |
-
"learning_rate": 4.425e-07,
|
| 93 |
-
"loss": 1333.5215,
|
| 94 |
-
"step": 60
|
| 95 |
-
},
|
| 96 |
-
{
|
| 97 |
-
"epoch": 1.4395604395604396,
|
| 98 |
-
"grad_norm": 5059.5302734375,
|
| 99 |
-
"learning_rate": 4.800000000000001e-07,
|
| 100 |
-
"loss": 1305.5257,
|
| 101 |
-
"step": 65
|
| 102 |
-
},
|
| 103 |
-
{
|
| 104 |
-
"epoch": 1.5494505494505495,
|
| 105 |
-
"grad_norm": 14086.837890625,
|
| 106 |
-
"learning_rate": 5.175e-07,
|
| 107 |
-
"loss": 1293.7212,
|
| 108 |
-
"step": 70
|
| 109 |
-
},
|
| 110 |
-
{
|
| 111 |
-
"epoch": 1.6593406593406592,
|
| 112 |
-
"grad_norm": 4606.49462890625,
|
| 113 |
-
"learning_rate": 5.55e-07,
|
| 114 |
-
"loss": 1293.7292,
|
| 115 |
-
"step": 75
|
| 116 |
-
},
|
| 117 |
-
{
|
| 118 |
-
"epoch": 1.7692307692307692,
|
| 119 |
-
"grad_norm": 7120.25244140625,
|
| 120 |
-
"learning_rate": 5.925e-07,
|
| 121 |
-
"loss": 1283.8888,
|
| 122 |
-
"step": 80
|
| 123 |
-
},
|
| 124 |
-
{
|
| 125 |
-
"epoch": 1.879120879120879,
|
| 126 |
-
"grad_norm": 4402.51513671875,
|
| 127 |
-
"learning_rate": 6.3e-07,
|
| 128 |
-
"loss": 1270.6494,
|
| 129 |
-
"step": 85
|
| 130 |
-
},
|
| 131 |
-
{
|
| 132 |
-
"epoch": 1.989010989010989,
|
| 133 |
-
"grad_norm": 4826.724609375,
|
| 134 |
-
"learning_rate": 6.675e-07,
|
| 135 |
-
"loss": 1263.9209,
|
| 136 |
-
"step": 90
|
| 137 |
-
},
|
| 138 |
-
{
|
| 139 |
-
"epoch": 2.087912087912088,
|
| 140 |
-
"grad_norm": 4356.83056640625,
|
| 141 |
-
"learning_rate": 7.05e-07,
|
| 142 |
-
"loss": 1255.111,
|
| 143 |
-
"step": 95
|
| 144 |
-
},
|
| 145 |
-
{
|
| 146 |
-
"epoch": 2.197802197802198,
|
| 147 |
-
"grad_norm": 6895.61962890625,
|
| 148 |
-
"learning_rate": 7.425000000000001e-07,
|
| 149 |
-
"loss": 1224.4364,
|
| 150 |
-
"step": 100
|
| 151 |
-
},
|
| 152 |
-
{
|
| 153 |
-
"epoch": 2.3076923076923075,
|
| 154 |
-
"grad_norm": 4793.59375,
|
| 155 |
-
"learning_rate": 7.799999999999999e-07,
|
| 156 |
-
"loss": 1211.3252,
|
| 157 |
-
"step": 105
|
| 158 |
-
},
|
| 159 |
-
{
|
| 160 |
-
"epoch": 2.4175824175824174,
|
| 161 |
-
"grad_norm": 6701.6357421875,
|
| 162 |
-
"learning_rate": 8.175e-07,
|
| 163 |
-
"loss": 1183.1627,
|
| 164 |
-
"step": 110
|
| 165 |
-
},
|
| 166 |
-
{
|
| 167 |
-
"epoch": 2.5274725274725274,
|
| 168 |
-
"grad_norm": 4253.81005859375,
|
| 169 |
-
"learning_rate": 8.550000000000001e-07,
|
| 170 |
-
"loss": 1179.29,
|
| 171 |
-
"step": 115
|
| 172 |
-
},
|
| 173 |
-
{
|
| 174 |
-
"epoch": 2.6373626373626373,
|
| 175 |
-
"grad_norm": 4965.86181640625,
|
| 176 |
-
"learning_rate": 8.925e-07,
|
| 177 |
-
"loss": 1168.3469,
|
| 178 |
-
"step": 120
|
| 179 |
-
},
|
| 180 |
-
{
|
| 181 |
-
"epoch": 2.7472527472527473,
|
| 182 |
-
"grad_norm": 3376.816650390625,
|
| 183 |
-
"learning_rate": 9.3e-07,
|
| 184 |
-
"loss": 1139.6433,
|
| 185 |
-
"step": 125
|
| 186 |
-
},
|
| 187 |
-
{
|
| 188 |
-
"epoch": 2.857142857142857,
|
| 189 |
-
"grad_norm": 3226.073486328125,
|
| 190 |
-
"learning_rate": 9.675e-07,
|
| 191 |
-
"loss": 1133.0777,
|
| 192 |
-
"step": 130
|
| 193 |
-
},
|
| 194 |
-
{
|
| 195 |
-
"epoch": 2.967032967032967,
|
| 196 |
-
"grad_norm": 3560.005126953125,
|
| 197 |
-
"learning_rate": 1.0050000000000001e-06,
|
| 198 |
-
"loss": 1113.3471,
|
| 199 |
-
"step": 135
|
| 200 |
-
},
|
| 201 |
-
{
|
| 202 |
-
"epoch": 3.065934065934066,
|
| 203 |
-
"grad_norm": 2917.787109375,
|
| 204 |
-
"learning_rate": 1.0425000000000002e-06,
|
| 205 |
-
"loss": 1094.2038,
|
| 206 |
-
"step": 140
|
| 207 |
-
},
|
| 208 |
-
{
|
| 209 |
-
"epoch": 3.1758241758241756,
|
| 210 |
-
"grad_norm": 3291.990478515625,
|
| 211 |
-
"learning_rate": 1.08e-06,
|
| 212 |
-
"loss": 1069.3087,
|
| 213 |
-
"step": 145
|
| 214 |
-
},
|
| 215 |
-
{
|
| 216 |
-
"epoch": 3.2857142857142856,
|
| 217 |
-
"grad_norm": 3082.956298828125,
|
| 218 |
-
"learning_rate": 1.1174999999999999e-06,
|
| 219 |
-
"loss": 1043.3319,
|
| 220 |
-
"step": 150
|
| 221 |
-
},
|
| 222 |
-
{
|
| 223 |
-
"epoch": 3.3956043956043955,
|
| 224 |
-
"grad_norm": 2947.577392578125,
|
| 225 |
-
"learning_rate": 1.155e-06,
|
| 226 |
-
"loss": 1040.2479,
|
| 227 |
-
"step": 155
|
| 228 |
-
},
|
| 229 |
-
{
|
| 230 |
-
"epoch": 3.5054945054945055,
|
| 231 |
-
"grad_norm": 2917.072021484375,
|
| 232 |
-
"learning_rate": 1.1925e-06,
|
| 233 |
-
"loss": 1014.5039,
|
| 234 |
-
"step": 160
|
| 235 |
-
},
|
| 236 |
-
{
|
| 237 |
-
"epoch": 3.6153846153846154,
|
| 238 |
-
"grad_norm": 2572.1064453125,
|
| 239 |
-
"learning_rate": 1.23e-06,
|
| 240 |
-
"loss": 980.6475,
|
| 241 |
-
"step": 165
|
| 242 |
-
},
|
| 243 |
-
{
|
| 244 |
-
"epoch": 3.7252747252747254,
|
| 245 |
-
"grad_norm": 3038.7041015625,
|
| 246 |
-
"learning_rate": 1.2675000000000001e-06,
|
| 247 |
-
"loss": 969.8687,
|
| 248 |
-
"step": 170
|
| 249 |
-
},
|
| 250 |
-
{
|
| 251 |
-
"epoch": 3.8351648351648353,
|
| 252 |
-
"grad_norm": 2459.19482421875,
|
| 253 |
-
"learning_rate": 1.305e-06,
|
| 254 |
-
"loss": 946.8183,
|
| 255 |
-
"step": 175
|
| 256 |
-
},
|
| 257 |
-
{
|
| 258 |
-
"epoch": 3.9450549450549453,
|
| 259 |
-
"grad_norm": 2485.954345703125,
|
| 260 |
-
"learning_rate": 1.3425e-06,
|
| 261 |
-
"loss": 923.6708,
|
| 262 |
-
"step": 180
|
| 263 |
-
},
|
| 264 |
-
{
|
| 265 |
-
"epoch": 4.043956043956044,
|
| 266 |
-
"grad_norm": 2574.14599609375,
|
| 267 |
-
"learning_rate": 1.38e-06,
|
| 268 |
-
"loss": 906.8975,
|
| 269 |
-
"step": 185
|
| 270 |
-
},
|
| 271 |
-
{
|
| 272 |
-
"epoch": 4.153846153846154,
|
| 273 |
-
"grad_norm": 2498.670166015625,
|
| 274 |
-
"learning_rate": 1.4175e-06,
|
| 275 |
-
"loss": 879.6545,
|
| 276 |
-
"step": 190
|
| 277 |
-
},
|
| 278 |
-
{
|
| 279 |
-
"epoch": 4.263736263736264,
|
| 280 |
-
"grad_norm": 2621.924072265625,
|
| 281 |
-
"learning_rate": 1.455e-06,
|
| 282 |
-
"loss": 859.6516,
|
| 283 |
-
"step": 195
|
| 284 |
-
},
|
| 285 |
-
{
|
| 286 |
-
"epoch": 4.373626373626374,
|
| 287 |
-
"grad_norm": 2298.13671875,
|
| 288 |
-
"learning_rate": 1.4925000000000001e-06,
|
| 289 |
-
"loss": 835.168,
|
| 290 |
-
"step": 200
|
| 291 |
-
},
|
| 292 |
-
{
|
| 293 |
-
"epoch": 4.483516483516484,
|
| 294 |
-
"grad_norm": 2228.95458984375,
|
| 295 |
-
"learning_rate": 1.53e-06,
|
| 296 |
-
"loss": 817.8176,
|
| 297 |
-
"step": 205
|
| 298 |
-
},
|
| 299 |
-
{
|
| 300 |
-
"epoch": 4.593406593406593,
|
| 301 |
-
"grad_norm": 2075.57568359375,
|
| 302 |
-
"learning_rate": 1.5675e-06,
|
| 303 |
-
"loss": 795.5578,
|
| 304 |
-
"step": 210
|
| 305 |
-
},
|
| 306 |
-
{
|
| 307 |
-
"epoch": 4.7032967032967035,
|
| 308 |
-
"grad_norm": 2308.193603515625,
|
| 309 |
-
"learning_rate": 1.605e-06,
|
| 310 |
-
"loss": 769.8123,
|
| 311 |
-
"step": 215
|
| 312 |
-
},
|
| 313 |
-
{
|
| 314 |
-
"epoch": 4.813186813186813,
|
| 315 |
-
"grad_norm": 2495.91259765625,
|
| 316 |
-
"learning_rate": 1.6425e-06,
|
| 317 |
-
"loss": 745.4971,
|
| 318 |
-
"step": 220
|
| 319 |
-
},
|
| 320 |
-
{
|
| 321 |
-
"epoch": 4.923076923076923,
|
| 322 |
-
"grad_norm": 2490.6796875,
|
| 323 |
-
"learning_rate": 1.68e-06,
|
| 324 |
-
"loss": 724.5927,
|
| 325 |
-
"step": 225
|
| 326 |
-
},
|
| 327 |
-
{
|
| 328 |
-
"epoch": 5.021978021978022,
|
| 329 |
-
"grad_norm": 1986.7320556640625,
|
| 330 |
-
"learning_rate": 1.7175e-06,
|
| 331 |
-
"loss": 711.5865,
|
| 332 |
-
"step": 230
|
| 333 |
-
},
|
| 334 |
-
{
|
| 335 |
-
"epoch": 5.131868131868132,
|
| 336 |
-
"grad_norm": 1901.4664306640625,
|
| 337 |
-
"learning_rate": 1.7550000000000001e-06,
|
| 338 |
-
"loss": 692.5993,
|
| 339 |
-
"step": 235
|
| 340 |
-
},
|
| 341 |
-
{
|
| 342 |
-
"epoch": 5.241758241758242,
|
| 343 |
-
"grad_norm": 2567.854248046875,
|
| 344 |
-
"learning_rate": 1.7925e-06,
|
| 345 |
-
"loss": 665.2945,
|
| 346 |
-
"step": 240
|
| 347 |
-
},
|
| 348 |
-
{
|
| 349 |
-
"epoch": 5.351648351648351,
|
| 350 |
-
"grad_norm": 1668.3482666015625,
|
| 351 |
-
"learning_rate": 1.83e-06,
|
| 352 |
-
"loss": 648.977,
|
| 353 |
-
"step": 245
|
| 354 |
-
},
|
| 355 |
-
{
|
| 356 |
-
"epoch": 5.461538461538462,
|
| 357 |
-
"grad_norm": 1845.254150390625,
|
| 358 |
-
"learning_rate": 1.8675000000000001e-06,
|
| 359 |
-
"loss": 627.8005,
|
| 360 |
-
"step": 250
|
| 361 |
-
},
|
| 362 |
-
{
|
| 363 |
-
"epoch": 5.571428571428571,
|
| 364 |
-
"grad_norm": 1811.66845703125,
|
| 365 |
-
"learning_rate": 1.905e-06,
|
| 366 |
-
"loss": 606.1213,
|
| 367 |
-
"step": 255
|
| 368 |
-
},
|
| 369 |
-
{
|
| 370 |
-
"epoch": 5.681318681318682,
|
| 371 |
-
"grad_norm": 1720.802734375,
|
| 372 |
-
"learning_rate": 1.9425e-06,
|
| 373 |
-
"loss": 587.5198,
|
| 374 |
-
"step": 260
|
| 375 |
-
},
|
| 376 |
-
{
|
| 377 |
-
"epoch": 5.791208791208791,
|
| 378 |
-
"grad_norm": 1535.228759765625,
|
| 379 |
-
"learning_rate": 1.98e-06,
|
| 380 |
-
"loss": 570.1765,
|
| 381 |
-
"step": 265
|
| 382 |
-
},
|
| 383 |
-
{
|
| 384 |
-
"epoch": 5.9010989010989015,
|
| 385 |
-
"grad_norm": 1655.658447265625,
|
| 386 |
-
"learning_rate": 2.0175e-06,
|
| 387 |
-
"loss": 551.0042,
|
| 388 |
-
"step": 270
|
| 389 |
-
},
|
| 390 |
-
{
|
| 391 |
-
"epoch": 6.0,
|
| 392 |
-
"grad_norm": 3979.0732421875,
|
| 393 |
-
"learning_rate": 2.0550000000000002e-06,
|
| 394 |
-
"loss": 535.4148,
|
| 395 |
-
"step": 275
|
| 396 |
-
},
|
| 397 |
-
{
|
| 398 |
-
"epoch": 6.1098901098901095,
|
| 399 |
-
"grad_norm": 1727.4771728515625,
|
| 400 |
-
"learning_rate": 2.0925000000000003e-06,
|
| 401 |
-
"loss": 519.8219,
|
| 402 |
-
"step": 280
|
| 403 |
-
},
|
| 404 |
-
{
|
| 405 |
-
"epoch": 6.21978021978022,
|
| 406 |
-
"grad_norm": 1350.4666748046875,
|
| 407 |
-
"learning_rate": 2.13e-06,
|
| 408 |
-
"loss": 510.5004,
|
| 409 |
-
"step": 285
|
| 410 |
-
},
|
| 411 |
-
{
|
| 412 |
-
"epoch": 6.329670329670329,
|
| 413 |
-
"grad_norm": 2242.578857421875,
|
| 414 |
-
"learning_rate": 2.1675e-06,
|
| 415 |
-
"loss": 497.497,
|
| 416 |
-
"step": 290
|
| 417 |
-
},
|
| 418 |
-
{
|
| 419 |
-
"epoch": 6.43956043956044,
|
| 420 |
-
"grad_norm": 1353.7908935546875,
|
| 421 |
-
"learning_rate": 2.205e-06,
|
| 422 |
-
"loss": 476.5748,
|
| 423 |
-
"step": 295
|
| 424 |
-
},
|
| 425 |
-
{
|
| 426 |
-
"epoch": 6.549450549450549,
|
| 427 |
-
"grad_norm": 1527.2796630859375,
|
| 428 |
-
"learning_rate": 2.2425e-06,
|
| 429 |
-
"loss": 468.7679,
|
| 430 |
-
"step": 300
|
| 431 |
-
},
|
| 432 |
-
{
|
| 433 |
-
"epoch": 6.65934065934066,
|
| 434 |
-
"grad_norm": 1145.3853759765625,
|
| 435 |
-
"learning_rate": 2.28e-06,
|
| 436 |
-
"loss": 454.7695,
|
| 437 |
-
"step": 305
|
| 438 |
-
},
|
| 439 |
-
{
|
| 440 |
-
"epoch": 6.769230769230769,
|
| 441 |
-
"grad_norm": 1233.435302734375,
|
| 442 |
-
"learning_rate": 2.3175e-06,
|
| 443 |
-
"loss": 438.3474,
|
| 444 |
-
"step": 310
|
| 445 |
-
},
|
| 446 |
-
{
|
| 447 |
-
"epoch": 6.8791208791208796,
|
| 448 |
-
"grad_norm": 1669.4859619140625,
|
| 449 |
-
"learning_rate": 2.355e-06,
|
| 450 |
-
"loss": 424.2438,
|
| 451 |
-
"step": 315
|
| 452 |
-
},
|
| 453 |
-
{
|
| 454 |
-
"epoch": 6.989010989010989,
|
| 455 |
-
"grad_norm": 1667.7239990234375,
|
| 456 |
-
"learning_rate": 2.3925e-06,
|
| 457 |
-
"loss": 412.884,
|
| 458 |
-
"step": 320
|
| 459 |
-
},
|
| 460 |
-
{
|
| 461 |
-
"epoch": 7.087912087912088,
|
| 462 |
-
"grad_norm": 1087.6927490234375,
|
| 463 |
-
"learning_rate": 2.43e-06,
|
| 464 |
-
"loss": 399.3542,
|
| 465 |
-
"step": 325
|
| 466 |
-
},
|
| 467 |
-
{
|
| 468 |
-
"epoch": 7.197802197802198,
|
| 469 |
-
"grad_norm": 1415.59765625,
|
| 470 |
-
"learning_rate": 2.4675e-06,
|
| 471 |
-
"loss": 387.3436,
|
| 472 |
-
"step": 330
|
| 473 |
-
},
|
| 474 |
-
{
|
| 475 |
-
"epoch": 7.3076923076923075,
|
| 476 |
-
"grad_norm": 1952.5645751953125,
|
| 477 |
-
"learning_rate": 2.505e-06,
|
| 478 |
-
"loss": 377.6118,
|
| 479 |
-
"step": 335
|
| 480 |
-
},
|
| 481 |
-
{
|
| 482 |
-
"epoch": 7.417582417582418,
|
| 483 |
-
"grad_norm": 1404.5712890625,
|
| 484 |
-
"learning_rate": 2.5425000000000002e-06,
|
| 485 |
-
"loss": 371.8316,
|
| 486 |
-
"step": 340
|
| 487 |
-
},
|
| 488 |
-
{
|
| 489 |
-
"epoch": 7.527472527472527,
|
| 490 |
-
"grad_norm": 1185.5135498046875,
|
| 491 |
-
"learning_rate": 2.58e-06,
|
| 492 |
-
"loss": 363.4571,
|
| 493 |
-
"step": 345
|
| 494 |
-
},
|
| 495 |
-
{
|
| 496 |
-
"epoch": 7.637362637362637,
|
| 497 |
-
"grad_norm": 4727.9638671875,
|
| 498 |
-
"learning_rate": 2.6175e-06,
|
| 499 |
-
"loss": 354.3981,
|
| 500 |
-
"step": 350
|
| 501 |
-
},
|
| 502 |
-
{
|
| 503 |
-
"epoch": 7.747252747252747,
|
| 504 |
-
"grad_norm": 937.0252075195312,
|
| 505 |
-
"learning_rate": 2.655e-06,
|
| 506 |
-
"loss": 344.7615,
|
| 507 |
-
"step": 355
|
| 508 |
-
},
|
| 509 |
-
{
|
| 510 |
-
"epoch": 7.857142857142857,
|
| 511 |
-
"grad_norm": 1180.9298095703125,
|
| 512 |
-
"learning_rate": 2.6925e-06,
|
| 513 |
-
"loss": 340.5888,
|
| 514 |
-
"step": 360
|
| 515 |
-
},
|
| 516 |
-
{
|
| 517 |
-
"epoch": 7.967032967032967,
|
| 518 |
-
"grad_norm": 1405.458984375,
|
| 519 |
-
"learning_rate": 2.73e-06,
|
| 520 |
-
"loss": 326.6432,
|
| 521 |
-
"step": 365
|
| 522 |
-
},
|
| 523 |
-
{
|
| 524 |
-
"epoch": 8.065934065934066,
|
| 525 |
-
"grad_norm": 936.8318481445312,
|
| 526 |
-
"learning_rate": 2.7675e-06,
|
| 527 |
-
"loss": 317.4954,
|
| 528 |
-
"step": 370
|
| 529 |
-
},
|
| 530 |
-
{
|
| 531 |
-
"epoch": 8.175824175824175,
|
| 532 |
-
"grad_norm": 888.236328125,
|
| 533 |
-
"learning_rate": 2.8050000000000002e-06,
|
| 534 |
-
"loss": 309.4422,
|
| 535 |
-
"step": 375
|
| 536 |
-
},
|
| 537 |
-
{
|
| 538 |
-
"epoch": 8.285714285714286,
|
| 539 |
-
"grad_norm": 970.7135620117188,
|
| 540 |
-
"learning_rate": 2.8425e-06,
|
| 541 |
-
"loss": 301.2726,
|
| 542 |
-
"step": 380
|
| 543 |
-
},
|
| 544 |
-
{
|
| 545 |
-
"epoch": 8.395604395604396,
|
| 546 |
-
"grad_norm": 1607.8035888671875,
|
| 547 |
-
"learning_rate": 2.88e-06,
|
| 548 |
-
"loss": 295.7361,
|
| 549 |
-
"step": 385
|
| 550 |
-
},
|
| 551 |
-
{
|
| 552 |
-
"epoch": 8.505494505494505,
|
| 553 |
-
"grad_norm": 814.48486328125,
|
| 554 |
-
"learning_rate": 2.9175e-06,
|
| 555 |
-
"loss": 290.0745,
|
| 556 |
-
"step": 390
|
| 557 |
-
},
|
| 558 |
-
{
|
| 559 |
-
"epoch": 8.615384615384615,
|
| 560 |
-
"grad_norm": 803.1909790039062,
|
| 561 |
-
"learning_rate": 2.955e-06,
|
| 562 |
-
"loss": 279.9774,
|
| 563 |
-
"step": 395
|
| 564 |
-
},
|
| 565 |
-
{
|
| 566 |
-
"epoch": 8.725274725274724,
|
| 567 |
-
"grad_norm": 1034.652099609375,
|
| 568 |
-
"learning_rate": 2.9925e-06,
|
| 569 |
-
"loss": 277.7123,
|
| 570 |
-
"step": 400
|
| 571 |
-
},
|
| 572 |
-
{
|
| 573 |
-
"epoch": 8.835164835164836,
|
| 574 |
-
"grad_norm": 993.9649658203125,
|
| 575 |
-
"learning_rate": 3.0300000000000002e-06,
|
| 576 |
-
"loss": 267.6938,
|
| 577 |
-
"step": 405
|
| 578 |
-
},
|
| 579 |
-
{
|
| 580 |
-
"epoch": 8.945054945054945,
|
| 581 |
-
"grad_norm": 1010.8787841796875,
|
| 582 |
-
"learning_rate": 3.0675e-06,
|
| 583 |
-
"loss": 266.7134,
|
| 584 |
-
"step": 410
|
| 585 |
-
},
|
| 586 |
-
{
|
| 587 |
-
"epoch": 9.043956043956044,
|
| 588 |
-
"grad_norm": 762.2285766601562,
|
| 589 |
-
"learning_rate": 3.105e-06,
|
| 590 |
-
"loss": 264.4567,
|
| 591 |
-
"step": 415
|
| 592 |
-
},
|
| 593 |
-
{
|
| 594 |
-
"epoch": 9.153846153846153,
|
| 595 |
-
"grad_norm": 5807.85791015625,
|
| 596 |
-
"learning_rate": 3.1425e-06,
|
| 597 |
-
"loss": 254.0208,
|
| 598 |
-
"step": 420
|
| 599 |
-
},
|
| 600 |
-
{
|
| 601 |
-
"epoch": 9.263736263736265,
|
| 602 |
-
"grad_norm": 717.2619018554688,
|
| 603 |
-
"learning_rate": 3.18e-06,
|
| 604 |
-
"loss": 249.0201,
|
| 605 |
-
"step": 425
|
| 606 |
-
},
|
| 607 |
-
{
|
| 608 |
-
"epoch": 9.373626373626374,
|
| 609 |
-
"grad_norm": 725.6996459960938,
|
| 610 |
-
"learning_rate": 3.2175e-06,
|
| 611 |
-
"loss": 243.3289,
|
| 612 |
-
"step": 430
|
| 613 |
-
},
|
| 614 |
-
{
|
| 615 |
-
"epoch": 9.483516483516484,
|
| 616 |
-
"grad_norm": 761.1790161132812,
|
| 617 |
-
"learning_rate": 3.255e-06,
|
| 618 |
-
"loss": 238.8994,
|
| 619 |
-
"step": 435
|
| 620 |
-
},
|
| 621 |
-
{
|
| 622 |
-
"epoch": 9.593406593406593,
|
| 623 |
-
"grad_norm": 792.3602905273438,
|
| 624 |
-
"learning_rate": 3.2925000000000002e-06,
|
| 625 |
-
"loss": 231.8852,
|
| 626 |
-
"step": 440
|
| 627 |
-
},
|
| 628 |
-
{
|
| 629 |
-
"epoch": 9.703296703296703,
|
| 630 |
-
"grad_norm": 744.7413330078125,
|
| 631 |
-
"learning_rate": 3.3300000000000003e-06,
|
| 632 |
-
"loss": 229.136,
|
| 633 |
-
"step": 445
|
| 634 |
-
},
|
| 635 |
-
{
|
| 636 |
-
"epoch": 9.813186813186814,
|
| 637 |
-
"grad_norm": 673.6207885742188,
|
| 638 |
-
"learning_rate": 3.3675000000000004e-06,
|
| 639 |
-
"loss": 221.0368,
|
| 640 |
-
"step": 450
|
| 641 |
-
},
|
| 642 |
-
{
|
| 643 |
-
"epoch": 9.923076923076923,
|
| 644 |
-
"grad_norm": 966.46630859375,
|
| 645 |
-
"learning_rate": 3.405e-06,
|
| 646 |
-
"loss": 217.8422,
|
| 647 |
-
"step": 455
|
| 648 |
-
},
|
| 649 |
-
{
|
| 650 |
-
"epoch": 10.021978021978022,
|
| 651 |
-
"grad_norm": 740.294921875,
|
| 652 |
-
"learning_rate": 3.4425e-06,
|
| 653 |
-
"loss": 228.7085,
|
| 654 |
-
"step": 460
|
| 655 |
-
},
|
| 656 |
-
{
|
| 657 |
-
"epoch": 10.131868131868131,
|
| 658 |
-
"grad_norm": 629.9981689453125,
|
| 659 |
-
"learning_rate": 3.48e-06,
|
| 660 |
-
"loss": 211.0439,
|
| 661 |
-
"step": 465
|
| 662 |
-
},
|
| 663 |
-
{
|
| 664 |
-
"epoch": 10.241758241758241,
|
| 665 |
-
"grad_norm": 809.6885375976562,
|
| 666 |
-
"learning_rate": 3.5174999999999998e-06,
|
| 667 |
-
"loss": 204.4649,
|
| 668 |
-
"step": 470
|
| 669 |
-
},
|
| 670 |
-
{
|
| 671 |
-
"epoch": 10.351648351648352,
|
| 672 |
-
"grad_norm": 1631.996337890625,
|
| 673 |
-
"learning_rate": 3.555e-06,
|
| 674 |
-
"loss": 204.0642,
|
| 675 |
-
"step": 475
|
| 676 |
-
},
|
| 677 |
-
{
|
| 678 |
-
"epoch": 10.461538461538462,
|
| 679 |
-
"grad_norm": 958.5594482421875,
|
| 680 |
-
"learning_rate": 3.5925e-06,
|
| 681 |
-
"loss": 199.6854,
|
| 682 |
-
"step": 480
|
| 683 |
-
},
|
| 684 |
-
{
|
| 685 |
-
"epoch": 10.571428571428571,
|
| 686 |
-
"grad_norm": 588.5241088867188,
|
| 687 |
-
"learning_rate": 3.63e-06,
|
| 688 |
-
"loss": 196.1547,
|
| 689 |
-
"step": 485
|
| 690 |
-
},
|
| 691 |
-
{
|
| 692 |
-
"epoch": 10.68131868131868,
|
| 693 |
-
"grad_norm": 1439.742919921875,
|
| 694 |
-
"learning_rate": 3.6675e-06,
|
| 695 |
-
"loss": 194.8075,
|
| 696 |
-
"step": 490
|
| 697 |
-
},
|
| 698 |
-
{
|
| 699 |
-
"epoch": 10.791208791208792,
|
| 700 |
-
"grad_norm": 547.2682495117188,
|
| 701 |
-
"learning_rate": 3.705e-06,
|
| 702 |
-
"loss": 190.5422,
|
| 703 |
-
"step": 495
|
| 704 |
-
},
|
| 705 |
-
{
|
| 706 |
-
"epoch": 10.901098901098901,
|
| 707 |
-
"grad_norm": 550.932373046875,
|
| 708 |
-
"learning_rate": 3.7425e-06,
|
| 709 |
-
"loss": 188.0797,
|
| 710 |
-
"step": 500
|
| 711 |
-
},
|
| 712 |
-
{
|
| 713 |
-
"epoch": 11.0,
|
| 714 |
-
"grad_norm": 1168.6212158203125,
|
| 715 |
-
"learning_rate": 3.7800000000000002e-06,
|
| 716 |
-
"loss": 180.8073,
|
| 717 |
-
"step": 505
|
| 718 |
-
},
|
| 719 |
-
{
|
| 720 |
-
"epoch": 11.10989010989011,
|
| 721 |
-
"grad_norm": 613.2390747070312,
|
| 722 |
-
"learning_rate": 3.8175e-06,
|
| 723 |
-
"loss": 181.0391,
|
| 724 |
-
"step": 510
|
| 725 |
-
},
|
| 726 |
-
{
|
| 727 |
-
"epoch": 11.219780219780219,
|
| 728 |
-
"grad_norm": 589.2745361328125,
|
| 729 |
-
"learning_rate": 3.855e-06,
|
| 730 |
-
"loss": 175.6959,
|
| 731 |
-
"step": 515
|
| 732 |
-
},
|
| 733 |
-
{
|
| 734 |
-
"epoch": 11.32967032967033,
|
| 735 |
-
"grad_norm": 531.6744995117188,
|
| 736 |
-
"learning_rate": 3.8925000000000004e-06,
|
| 737 |
-
"loss": 174.1553,
|
| 738 |
-
"step": 520
|
| 739 |
-
},
|
| 740 |
-
{
|
| 741 |
-
"epoch": 11.43956043956044,
|
| 742 |
-
"grad_norm": 514.2449340820312,
|
| 743 |
-
"learning_rate": 3.9300000000000005e-06,
|
| 744 |
-
"loss": 172.9302,
|
| 745 |
-
"step": 525
|
| 746 |
-
},
|
| 747 |
-
{
|
| 748 |
-
"epoch": 11.54945054945055,
|
| 749 |
-
"grad_norm": 532.4719848632812,
|
| 750 |
-
"learning_rate": 3.9675000000000006e-06,
|
| 751 |
-
"loss": 168.4628,
|
| 752 |
-
"step": 530
|
| 753 |
-
},
|
| 754 |
-
{
|
| 755 |
-
"epoch": 11.659340659340659,
|
| 756 |
-
"grad_norm": 475.0870361328125,
|
| 757 |
-
"learning_rate": 4.005000000000001e-06,
|
| 758 |
-
"loss": 164.1821,
|
| 759 |
-
"step": 535
|
| 760 |
-
},
|
| 761 |
-
{
|
| 762 |
-
"epoch": 11.76923076923077,
|
| 763 |
-
"grad_norm": 531.968994140625,
|
| 764 |
-
"learning_rate": 4.042500000000001e-06,
|
| 765 |
-
"loss": 160.9798,
|
| 766 |
-
"step": 540
|
| 767 |
-
},
|
| 768 |
-
{
|
| 769 |
-
"epoch": 11.87912087912088,
|
| 770 |
-
"grad_norm": 1276.7720947265625,
|
| 771 |
-
"learning_rate": 4.080000000000001e-06,
|
| 772 |
-
"loss": 161.732,
|
| 773 |
-
"step": 545
|
| 774 |
-
},
|
| 775 |
-
{
|
| 776 |
-
"epoch": 11.989010989010989,
|
| 777 |
-
"grad_norm": 566.782958984375,
|
| 778 |
-
"learning_rate": 4.117500000000001e-06,
|
| 779 |
-
"loss": 157.6948,
|
| 780 |
-
"step": 550
|
| 781 |
-
},
|
| 782 |
-
{
|
| 783 |
-
"epoch": 12.087912087912088,
|
| 784 |
-
"grad_norm": 749.6525268554688,
|
| 785 |
-
"learning_rate": 4.155000000000001e-06,
|
| 786 |
-
"loss": 157.8984,
|
| 787 |
-
"step": 555
|
| 788 |
-
},
|
| 789 |
-
{
|
| 790 |
-
"epoch": 12.197802197802197,
|
| 791 |
-
"grad_norm": 465.9924011230469,
|
| 792 |
-
"learning_rate": 4.1925e-06,
|
| 793 |
-
"loss": 151.8522,
|
| 794 |
-
"step": 560
|
| 795 |
-
},
|
| 796 |
-
{
|
| 797 |
-
"epoch": 12.307692307692308,
|
| 798 |
-
"grad_norm": 508.3736572265625,
|
| 799 |
-
"learning_rate": 4.229999999999999e-06,
|
| 800 |
-
"loss": 148.1901,
|
| 801 |
-
"step": 565
|
| 802 |
-
},
|
| 803 |
-
{
|
| 804 |
-
"epoch": 12.417582417582418,
|
| 805 |
-
"grad_norm": 1925.7998046875,
|
| 806 |
-
"learning_rate": 4.267499999999999e-06,
|
| 807 |
-
"loss": 148.9536,
|
| 808 |
-
"step": 570
|
| 809 |
-
},
|
| 810 |
-
{
|
| 811 |
-
"epoch": 12.527472527472527,
|
| 812 |
-
"grad_norm": 528.435546875,
|
| 813 |
-
"learning_rate": 4.3049999999999994e-06,
|
| 814 |
-
"loss": 146.6833,
|
| 815 |
-
"step": 575
|
| 816 |
-
},
|
| 817 |
-
{
|
| 818 |
-
"epoch": 12.637362637362637,
|
| 819 |
-
"grad_norm": 629.6632690429688,
|
| 820 |
-
"learning_rate": 4.3424999999999995e-06,
|
| 821 |
-
"loss": 141.5915,
|
| 822 |
-
"step": 580
|
| 823 |
-
},
|
| 824 |
-
{
|
| 825 |
-
"epoch": 12.747252747252748,
|
| 826 |
-
"grad_norm": 475.8123474121094,
|
| 827 |
-
"learning_rate": 4.3799999999999996e-06,
|
| 828 |
-
"loss": 139.1265,
|
| 829 |
-
"step": 585
|
| 830 |
-
},
|
| 831 |
-
{
|
| 832 |
-
"epoch": 12.857142857142858,
|
| 833 |
-
"grad_norm": 545.7217407226562,
|
| 834 |
-
"learning_rate": 4.4175e-06,
|
| 835 |
-
"loss": 135.9861,
|
| 836 |
-
"step": 590
|
| 837 |
-
},
|
| 838 |
-
{
|
| 839 |
-
"epoch": 12.967032967032967,
|
| 840 |
-
"grad_norm": 2502.422607421875,
|
| 841 |
-
"learning_rate": 4.455e-06,
|
| 842 |
-
"loss": 138.432,
|
| 843 |
-
"step": 595
|
| 844 |
-
},
|
| 845 |
-
{
|
| 846 |
-
"epoch": 13.065934065934066,
|
| 847 |
-
"grad_norm": 432.36279296875,
|
| 848 |
-
"learning_rate": 4.4925e-06,
|
| 849 |
-
"loss": 144.0422,
|
| 850 |
-
"step": 600
|
| 851 |
-
},
|
| 852 |
-
{
|
| 853 |
-
"epoch": 13.175824175824175,
|
| 854 |
-
"grad_norm": 429.08685302734375,
|
| 855 |
-
"learning_rate": 4.53e-06,
|
| 856 |
-
"loss": 132.7908,
|
| 857 |
-
"step": 605
|
| 858 |
-
},
|
| 859 |
-
{
|
| 860 |
-
"epoch": 13.285714285714286,
|
| 861 |
-
"grad_norm": 467.1208801269531,
|
| 862 |
-
"learning_rate": 4.5675e-06,
|
| 863 |
-
"loss": 130.9718,
|
| 864 |
-
"step": 610
|
| 865 |
-
},
|
| 866 |
-
{
|
| 867 |
-
"epoch": 13.395604395604396,
|
| 868 |
-
"grad_norm": 1329.150634765625,
|
| 869 |
-
"learning_rate": 4.605e-06,
|
| 870 |
-
"loss": 134.7514,
|
| 871 |
-
"step": 615
|
| 872 |
-
},
|
| 873 |
-
{
|
| 874 |
-
"epoch": 13.505494505494505,
|
| 875 |
-
"grad_norm": 404.3941650390625,
|
| 876 |
-
"learning_rate": 4.6425e-06,
|
| 877 |
-
"loss": 127.4091,
|
| 878 |
-
"step": 620
|
| 879 |
-
},
|
| 880 |
-
{
|
| 881 |
-
"epoch": 13.615384615384615,
|
| 882 |
-
"grad_norm": 439.34234619140625,
|
| 883 |
-
"learning_rate": 4.68e-06,
|
| 884 |
-
"loss": 123.8929,
|
| 885 |
-
"step": 625
|
| 886 |
-
},
|
| 887 |
-
{
|
| 888 |
-
"epoch": 13.725274725274724,
|
| 889 |
-
"grad_norm": 418.9570007324219,
|
| 890 |
-
"learning_rate": 4.7175e-06,
|
| 891 |
-
"loss": 122.8744,
|
| 892 |
-
"step": 630
|
| 893 |
-
},
|
| 894 |
-
{
|
| 895 |
-
"epoch": 13.835164835164836,
|
| 896 |
-
"grad_norm": 396.3748474121094,
|
| 897 |
-
"learning_rate": 4.755e-06,
|
| 898 |
-
"loss": 121.2023,
|
| 899 |
-
"step": 635
|
| 900 |
-
},
|
| 901 |
-
{
|
| 902 |
-
"epoch": 13.945054945054945,
|
| 903 |
-
"grad_norm": 617.2871704101562,
|
| 904 |
-
"learning_rate": 4.7925e-06,
|
| 905 |
-
"loss": 121.1829,
|
| 906 |
-
"step": 640
|
| 907 |
-
},
|
| 908 |
-
{
|
| 909 |
-
"epoch": 14.043956043956044,
|
| 910 |
-
"grad_norm": 409.8064270019531,
|
| 911 |
-
"learning_rate": 4.83e-06,
|
| 912 |
-
"loss": 117.2707,
|
| 913 |
-
"step": 645
|
| 914 |
-
},
|
| 915 |
-
{
|
| 916 |
-
"epoch": 14.153846153846153,
|
| 917 |
-
"grad_norm": 516.0888671875,
|
| 918 |
-
"learning_rate": 4.8675e-06,
|
| 919 |
-
"loss": 118.2377,
|
| 920 |
-
"step": 650
|
| 921 |
-
},
|
| 922 |
-
{
|
| 923 |
-
"epoch": 14.263736263736265,
|
| 924 |
-
"grad_norm": 505.36480712890625,
|
| 925 |
-
"learning_rate": 4.9050000000000005e-06,
|
| 926 |
-
"loss": 116.6757,
|
| 927 |
-
"step": 655
|
| 928 |
-
},
|
| 929 |
-
{
|
| 930 |
-
"epoch": 14.373626373626374,
|
| 931 |
-
"grad_norm": 449.5859680175781,
|
| 932 |
-
"learning_rate": 4.9425000000000005e-06,
|
| 933 |
-
"loss": 114.7447,
|
| 934 |
-
"step": 660
|
| 935 |
-
},
|
| 936 |
-
{
|
| 937 |
-
"epoch": 14.483516483516484,
|
| 938 |
-
"grad_norm": 420.5121765136719,
|
| 939 |
-
"learning_rate": 4.980000000000001e-06,
|
| 940 |
-
"loss": 113.2027,
|
| 941 |
-
"step": 665
|
| 942 |
-
},
|
| 943 |
-
{
|
| 944 |
-
"epoch": 14.593406593406593,
|
| 945 |
-
"grad_norm": 345.0196838378906,
|
| 946 |
-
"learning_rate": 5.017500000000001e-06,
|
| 947 |
-
"loss": 111.8875,
|
| 948 |
-
"step": 670
|
| 949 |
-
},
|
| 950 |
-
{
|
| 951 |
-
"epoch": 14.703296703296703,
|
| 952 |
-
"grad_norm": 377.2486877441406,
|
| 953 |
-
"learning_rate": 5.055000000000001e-06,
|
| 954 |
-
"loss": 107.6133,
|
| 955 |
-
"step": 675
|
| 956 |
-
},
|
| 957 |
-
{
|
| 958 |
-
"epoch": 14.813186813186814,
|
| 959 |
-
"grad_norm": 384.3365783691406,
|
| 960 |
-
"learning_rate": 5.092500000000001e-06,
|
| 961 |
-
"loss": 109.3438,
|
| 962 |
-
"step": 680
|
| 963 |
-
},
|
| 964 |
-
{
|
| 965 |
-
"epoch": 14.923076923076923,
|
| 966 |
-
"grad_norm": 385.41998291015625,
|
| 967 |
-
"learning_rate": 5.130000000000001e-06,
|
| 968 |
-
"loss": 103.7929,
|
| 969 |
-
"step": 685
|
| 970 |
-
},
|
| 971 |
-
{
|
| 972 |
-
"epoch": 15.021978021978022,
|
| 973 |
-
"grad_norm": 362.2362976074219,
|
| 974 |
-
"learning_rate": 5.1675e-06,
|
| 975 |
-
"loss": 110.7998,
|
| 976 |
-
"step": 690
|
| 977 |
-
},
|
| 978 |
-
{
|
| 979 |
-
"epoch": 15.131868131868131,
|
| 980 |
-
"grad_norm": 351.775146484375,
|
| 981 |
-
"learning_rate": 5.205e-06,
|
| 982 |
-
"loss": 100.9369,
|
| 983 |
-
"step": 695
|
| 984 |
-
},
|
| 985 |
-
{
|
| 986 |
-
"epoch": 15.241758241758241,
|
| 987 |
-
"grad_norm": 390.7868347167969,
|
| 988 |
-
"learning_rate": 5.2425e-06,
|
| 989 |
-
"loss": 103.8922,
|
| 990 |
-
"step": 700
|
| 991 |
-
},
|
| 992 |
-
{
|
| 993 |
-
"epoch": 15.351648351648352,
|
| 994 |
-
"grad_norm": 344.78564453125,
|
| 995 |
-
"learning_rate": 5.279999999999999e-06,
|
| 996 |
-
"loss": 99.1799,
|
| 997 |
-
"step": 705
|
| 998 |
-
},
|
| 999 |
-
{
|
| 1000 |
-
"epoch": 15.461538461538462,
|
| 1001 |
-
"grad_norm": 378.85675048828125,
|
| 1002 |
-
"learning_rate": 5.3174999999999995e-06,
|
| 1003 |
-
"loss": 98.0914,
|
| 1004 |
-
"step": 710
|
| 1005 |
-
},
|
| 1006 |
-
{
|
| 1007 |
-
"epoch": 15.571428571428571,
|
| 1008 |
-
"grad_norm": 342.2828063964844,
|
| 1009 |
-
"learning_rate": 5.3549999999999996e-06,
|
| 1010 |
-
"loss": 97.6728,
|
| 1011 |
-
"step": 715
|
| 1012 |
-
},
|
| 1013 |
-
{
|
| 1014 |
-
"epoch": 15.68131868131868,
|
| 1015 |
-
"grad_norm": 341.1894226074219,
|
| 1016 |
-
"learning_rate": 5.3925e-06,
|
| 1017 |
-
"loss": 98.8686,
|
| 1018 |
-
"step": 720
|
| 1019 |
-
},
|
| 1020 |
-
{
|
| 1021 |
-
"epoch": 15.791208791208792,
|
| 1022 |
-
"grad_norm": 336.378662109375,
|
| 1023 |
-
"learning_rate": 5.43e-06,
|
| 1024 |
-
"loss": 93.9906,
|
| 1025 |
-
"step": 725
|
| 1026 |
-
},
|
| 1027 |
-
{
|
| 1028 |
-
"epoch": 15.901098901098901,
|
| 1029 |
-
"grad_norm": 334.4036560058594,
|
| 1030 |
-
"learning_rate": 5.4675e-06,
|
| 1031 |
-
"loss": 93.7508,
|
| 1032 |
-
"step": 730
|
| 1033 |
-
},
|
| 1034 |
-
{
|
| 1035 |
-
"epoch": 16.0,
|
| 1036 |
-
"grad_norm": 1449.2354736328125,
|
| 1037 |
-
"learning_rate": 5.505e-06,
|
| 1038 |
-
"loss": 96.8374,
|
| 1039 |
-
"step": 735
|
| 1040 |
-
},
|
| 1041 |
-
{
|
| 1042 |
-
"epoch": 16.10989010989011,
|
| 1043 |
-
"grad_norm": 341.5977478027344,
|
| 1044 |
-
"learning_rate": 5.5425e-06,
|
| 1045 |
-
"loss": 90.1757,
|
| 1046 |
-
"step": 740
|
| 1047 |
-
},
|
| 1048 |
-
{
|
| 1049 |
-
"epoch": 16.21978021978022,
|
| 1050 |
-
"grad_norm": 562.0274658203125,
|
| 1051 |
-
"learning_rate": 5.58e-06,
|
| 1052 |
-
"loss": 89.7033,
|
| 1053 |
-
"step": 745
|
| 1054 |
-
},
|
| 1055 |
-
{
|
| 1056 |
-
"epoch": 16.32967032967033,
|
| 1057 |
-
"grad_norm": 341.64959716796875,
|
| 1058 |
-
"learning_rate": 5.6175e-06,
|
| 1059 |
-
"loss": 90.8916,
|
| 1060 |
-
"step": 750
|
| 1061 |
-
},
|
| 1062 |
-
{
|
| 1063 |
-
"epoch": 16.439560439560438,
|
| 1064 |
-
"grad_norm": 305.3785400390625,
|
| 1065 |
-
"learning_rate": 5.655e-06,
|
| 1066 |
-
"loss": 87.8087,
|
| 1067 |
-
"step": 755
|
| 1068 |
-
},
|
| 1069 |
-
{
|
| 1070 |
-
"epoch": 16.54945054945055,
|
| 1071 |
-
"grad_norm": 317.1134338378906,
|
| 1072 |
-
"learning_rate": 5.6925e-06,
|
| 1073 |
-
"loss": 87.2254,
|
| 1074 |
-
"step": 760
|
| 1075 |
-
},
|
| 1076 |
-
{
|
| 1077 |
-
"epoch": 16.65934065934066,
|
| 1078 |
-
"grad_norm": 305.0978088378906,
|
| 1079 |
-
"learning_rate": 5.73e-06,
|
| 1080 |
-
"loss": 86.9592,
|
| 1081 |
-
"step": 765
|
| 1082 |
-
},
|
| 1083 |
-
{
|
| 1084 |
-
"epoch": 16.76923076923077,
|
| 1085 |
-
"grad_norm": 280.41778564453125,
|
| 1086 |
-
"learning_rate": 5.7675e-06,
|
| 1087 |
-
"loss": 83.7875,
|
| 1088 |
-
"step": 770
|
| 1089 |
-
},
|
| 1090 |
-
{
|
| 1091 |
-
"epoch": 16.87912087912088,
|
| 1092 |
-
"grad_norm": 318.2386169433594,
|
| 1093 |
-
"learning_rate": 5.805e-06,
|
| 1094 |
-
"loss": 87.919,
|
| 1095 |
-
"step": 775
|
| 1096 |
-
},
|
| 1097 |
-
{
|
| 1098 |
-
"epoch": 16.98901098901099,
|
| 1099 |
-
"grad_norm": 278.9162902832031,
|
| 1100 |
-
"learning_rate": 5.8425e-06,
|
| 1101 |
-
"loss": 82.1755,
|
| 1102 |
-
"step": 780
|
| 1103 |
-
},
|
| 1104 |
-
{
|
| 1105 |
-
"epoch": 17.087912087912088,
|
| 1106 |
-
"grad_norm": 267.6657409667969,
|
| 1107 |
-
"learning_rate": 5.8800000000000005e-06,
|
| 1108 |
-
"loss": 84.2234,
|
| 1109 |
-
"step": 785
|
| 1110 |
-
},
|
| 1111 |
-
{
|
| 1112 |
-
"epoch": 17.197802197802197,
|
| 1113 |
-
"grad_norm": 286.16162109375,
|
| 1114 |
-
"learning_rate": 5.9175000000000005e-06,
|
| 1115 |
-
"loss": 81.0474,
|
| 1116 |
-
"step": 790
|
| 1117 |
-
},
|
| 1118 |
-
{
|
| 1119 |
-
"epoch": 17.307692307692307,
|
| 1120 |
-
"grad_norm": 288.74151611328125,
|
| 1121 |
-
"learning_rate": 5.955000000000001e-06,
|
| 1122 |
-
"loss": 80.2032,
|
| 1123 |
-
"step": 795
|
| 1124 |
-
},
|
| 1125 |
-
{
|
| 1126 |
-
"epoch": 17.417582417582416,
|
| 1127 |
-
"grad_norm": 293.4186096191406,
|
| 1128 |
-
"learning_rate": 5.992500000000001e-06,
|
| 1129 |
-
"loss": 79.8146,
|
| 1130 |
-
"step": 800
|
| 1131 |
-
},
|
| 1132 |
-
{
|
| 1133 |
-
"epoch": 17.52747252747253,
|
| 1134 |
-
"grad_norm": 294.4632873535156,
|
| 1135 |
-
"learning_rate": 6.030000000000001e-06,
|
| 1136 |
-
"loss": 78.9017,
|
| 1137 |
-
"step": 805
|
| 1138 |
-
},
|
| 1139 |
-
{
|
| 1140 |
-
"epoch": 17.63736263736264,
|
| 1141 |
-
"grad_norm": 278.7751159667969,
|
| 1142 |
-
"learning_rate": 6.067500000000001e-06,
|
| 1143 |
-
"loss": 74.4998,
|
| 1144 |
-
"step": 810
|
| 1145 |
-
},
|
| 1146 |
-
{
|
| 1147 |
-
"epoch": 17.747252747252748,
|
| 1148 |
-
"grad_norm": 281.1651306152344,
|
| 1149 |
-
"learning_rate": 6.105e-06,
|
| 1150 |
-
"loss": 73.4448,
|
| 1151 |
-
"step": 815
|
| 1152 |
-
},
|
| 1153 |
-
{
|
| 1154 |
-
"epoch": 17.857142857142858,
|
| 1155 |
-
"grad_norm": 253.44654846191406,
|
| 1156 |
-
"learning_rate": 6.1425e-06,
|
| 1157 |
-
"loss": 72.5747,
|
| 1158 |
-
"step": 820
|
| 1159 |
-
},
|
| 1160 |
-
{
|
| 1161 |
-
"epoch": 17.967032967032967,
|
| 1162 |
-
"grad_norm": 270.03265380859375,
|
| 1163 |
-
"learning_rate": 6.18e-06,
|
| 1164 |
-
"loss": 72.7589,
|
| 1165 |
-
"step": 825
|
| 1166 |
-
},
|
| 1167 |
-
{
|
| 1168 |
-
"epoch": 18.065934065934066,
|
| 1169 |
-
"grad_norm": 275.12652587890625,
|
| 1170 |
-
"learning_rate": 6.2175e-06,
|
| 1171 |
-
"loss": 72.1071,
|
| 1172 |
-
"step": 830
|
| 1173 |
-
},
|
| 1174 |
-
{
|
| 1175 |
-
"epoch": 18.175824175824175,
|
| 1176 |
-
"grad_norm": 267.95965576171875,
|
| 1177 |
-
"learning_rate": 6.255e-06,
|
| 1178 |
-
"loss": 71.183,
|
| 1179 |
-
"step": 835
|
| 1180 |
-
},
|
| 1181 |
-
{
|
| 1182 |
-
"epoch": 18.285714285714285,
|
| 1183 |
-
"grad_norm": 250.47987365722656,
|
| 1184 |
-
"learning_rate": 6.2925e-06,
|
| 1185 |
-
"loss": 72.3531,
|
| 1186 |
-
"step": 840
|
| 1187 |
-
},
|
| 1188 |
-
{
|
| 1189 |
-
"epoch": 18.395604395604394,
|
| 1190 |
-
"grad_norm": 284.7598876953125,
|
| 1191 |
-
"learning_rate": 6.3299999999999995e-06,
|
| 1192 |
-
"loss": 75.0338,
|
| 1193 |
-
"step": 845
|
| 1194 |
-
},
|
| 1195 |
-
{
|
| 1196 |
-
"epoch": 18.505494505494504,
|
| 1197 |
-
"grad_norm": 303.2804260253906,
|
| 1198 |
-
"learning_rate": 6.3675e-06,
|
| 1199 |
-
"loss": 73.4008,
|
| 1200 |
-
"step": 850
|
| 1201 |
-
},
|
| 1202 |
-
{
|
| 1203 |
-
"epoch": 18.615384615384617,
|
| 1204 |
-
"grad_norm": 245.9633026123047,
|
| 1205 |
-
"learning_rate": 6.405e-06,
|
| 1206 |
-
"loss": 69.7974,
|
| 1207 |
-
"step": 855
|
| 1208 |
-
},
|
| 1209 |
-
{
|
| 1210 |
-
"epoch": 18.725274725274726,
|
| 1211 |
-
"grad_norm": 243.55807495117188,
|
| 1212 |
-
"learning_rate": 6.4425e-06,
|
| 1213 |
-
"loss": 67.0503,
|
| 1214 |
-
"step": 860
|
| 1215 |
-
},
|
| 1216 |
-
{
|
| 1217 |
-
"epoch": 18.835164835164836,
|
| 1218 |
-
"grad_norm": 274.70330810546875,
|
| 1219 |
-
"learning_rate": 6.48e-06,
|
| 1220 |
-
"loss": 67.1543,
|
| 1221 |
-
"step": 865
|
| 1222 |
-
},
|
| 1223 |
-
{
|
| 1224 |
-
"epoch": 18.945054945054945,
|
| 1225 |
-
"grad_norm": 233.40606689453125,
|
| 1226 |
-
"learning_rate": 6.5175e-06,
|
| 1227 |
-
"loss": 64.5814,
|
| 1228 |
-
"step": 870
|
| 1229 |
-
},
|
| 1230 |
-
{
|
| 1231 |
-
"epoch": 19.043956043956044,
|
| 1232 |
-
"grad_norm": 240.22552490234375,
|
| 1233 |
-
"learning_rate": 6.555e-06,
|
| 1234 |
-
"loss": 79.7664,
|
| 1235 |
-
"step": 875
|
| 1236 |
-
},
|
| 1237 |
-
{
|
| 1238 |
-
"epoch": 19.153846153846153,
|
| 1239 |
-
"grad_norm": 242.1979522705078,
|
| 1240 |
-
"learning_rate": 6.5925e-06,
|
| 1241 |
-
"loss": 64.3813,
|
| 1242 |
-
"step": 880
|
| 1243 |
-
},
|
| 1244 |
-
{
|
| 1245 |
-
"epoch": 19.263736263736263,
|
| 1246 |
-
"grad_norm": 231.37562561035156,
|
| 1247 |
-
"learning_rate": 6.63e-06,
|
| 1248 |
-
"loss": 65.6568,
|
| 1249 |
-
"step": 885
|
| 1250 |
-
},
|
| 1251 |
-
{
|
| 1252 |
-
"epoch": 19.373626373626372,
|
| 1253 |
-
"grad_norm": 235.3353271484375,
|
| 1254 |
-
"learning_rate": 6.6675e-06,
|
| 1255 |
-
"loss": 63.7893,
|
| 1256 |
-
"step": 890
|
| 1257 |
-
},
|
| 1258 |
-
{
|
| 1259 |
-
"epoch": 19.483516483516482,
|
| 1260 |
-
"grad_norm": 322.8166198730469,
|
| 1261 |
-
"learning_rate": 6.705e-06,
|
| 1262 |
-
"loss": 64.4515,
|
| 1263 |
-
"step": 895
|
| 1264 |
-
},
|
| 1265 |
-
{
|
| 1266 |
-
"epoch": 19.593406593406595,
|
| 1267 |
-
"grad_norm": 226.3159637451172,
|
| 1268 |
-
"learning_rate": 6.7425e-06,
|
| 1269 |
-
"loss": 63.5289,
|
| 1270 |
-
"step": 900
|
| 1271 |
-
},
|
| 1272 |
-
{
|
| 1273 |
-
"epoch": 19.703296703296704,
|
| 1274 |
-
"grad_norm": 218.580322265625,
|
| 1275 |
-
"learning_rate": 6.78e-06,
|
| 1276 |
-
"loss": 61.9305,
|
| 1277 |
-
"step": 905
|
| 1278 |
-
},
|
| 1279 |
-
{
|
| 1280 |
-
"epoch": 19.813186813186814,
|
| 1281 |
-
"grad_norm": 263.286376953125,
|
| 1282 |
-
"learning_rate": 6.8175e-06,
|
| 1283 |
-
"loss": 58.6838,
|
| 1284 |
-
"step": 910
|
| 1285 |
-
},
|
| 1286 |
-
{
|
| 1287 |
-
"epoch": 19.923076923076923,
|
| 1288 |
-
"grad_norm": 208.91209411621094,
|
| 1289 |
-
"learning_rate": 6.8550000000000004e-06,
|
| 1290 |
-
"loss": 57.9055,
|
| 1291 |
-
"step": 915
|
| 1292 |
-
},
|
| 1293 |
-
{
|
| 1294 |
-
"epoch": 20.021978021978022,
|
| 1295 |
-
"grad_norm": 211.89749145507812,
|
| 1296 |
-
"learning_rate": 6.8925000000000005e-06,
|
| 1297 |
-
"loss": 59.9611,
|
| 1298 |
-
"step": 920
|
| 1299 |
-
},
|
| 1300 |
-
{
|
| 1301 |
-
"epoch": 20.13186813186813,
|
| 1302 |
-
"grad_norm": 203.60804748535156,
|
| 1303 |
-
"learning_rate": 6.9300000000000006e-06,
|
| 1304 |
-
"loss": 57.4417,
|
| 1305 |
-
"step": 925
|
| 1306 |
-
},
|
| 1307 |
-
{
|
| 1308 |
-
"epoch": 20.24175824175824,
|
| 1309 |
-
"grad_norm": 217.0232696533203,
|
| 1310 |
-
"learning_rate": 6.967500000000001e-06,
|
| 1311 |
-
"loss": 57.6954,
|
| 1312 |
-
"step": 930
|
| 1313 |
-
},
|
| 1314 |
-
{
|
| 1315 |
-
"epoch": 20.35164835164835,
|
| 1316 |
-
"grad_norm": 215.05755615234375,
|
| 1317 |
-
"learning_rate": 7.005000000000001e-06,
|
| 1318 |
-
"loss": 58.4159,
|
| 1319 |
-
"step": 935
|
| 1320 |
-
},
|
| 1321 |
-
{
|
| 1322 |
-
"epoch": 20.46153846153846,
|
| 1323 |
-
"grad_norm": 208.3353729248047,
|
| 1324 |
-
"learning_rate": 7.0425e-06,
|
| 1325 |
-
"loss": 54.717,
|
| 1326 |
-
"step": 940
|
| 1327 |
-
},
|
| 1328 |
-
{
|
| 1329 |
-
"epoch": 20.571428571428573,
|
| 1330 |
-
"grad_norm": 205.00282287597656,
|
| 1331 |
-
"learning_rate": 7.08e-06,
|
| 1332 |
-
"loss": 53.7791,
|
| 1333 |
-
"step": 945
|
| 1334 |
-
},
|
| 1335 |
-
{
|
| 1336 |
-
"epoch": 20.681318681318682,
|
| 1337 |
-
"grad_norm": 197.68243408203125,
|
| 1338 |
-
"learning_rate": 7.1175e-06,
|
| 1339 |
-
"loss": 53.9224,
|
| 1340 |
-
"step": 950
|
| 1341 |
-
},
|
| 1342 |
-
{
|
| 1343 |
-
"epoch": 20.791208791208792,
|
| 1344 |
-
"grad_norm": 183.80923461914062,
|
| 1345 |
-
"learning_rate": 7.155e-06,
|
| 1346 |
-
"loss": 51.6971,
|
| 1347 |
-
"step": 955
|
| 1348 |
-
},
|
| 1349 |
-
{
|
| 1350 |
-
"epoch": 20.9010989010989,
|
| 1351 |
-
"grad_norm": 201.22625732421875,
|
| 1352 |
-
"learning_rate": 7.1925e-06,
|
| 1353 |
-
"loss": 50.1511,
|
| 1354 |
-
"step": 960
|
| 1355 |
-
},
|
| 1356 |
-
{
|
| 1357 |
-
"epoch": 21.0,
|
| 1358 |
-
"grad_norm": 433.5864562988281,
|
| 1359 |
-
"learning_rate": 7.23e-06,
|
| 1360 |
-
"loss": 53.6671,
|
| 1361 |
-
"step": 965
|
| 1362 |
-
},
|
| 1363 |
-
{
|
| 1364 |
-
"epoch": 21.10989010989011,
|
| 1365 |
-
"grad_norm": 180.5747833251953,
|
| 1366 |
-
"learning_rate": 7.2675e-06,
|
| 1367 |
-
"loss": 51.5476,
|
| 1368 |
-
"step": 970
|
| 1369 |
-
},
|
| 1370 |
-
{
|
| 1371 |
-
"epoch": 21.21978021978022,
|
| 1372 |
-
"grad_norm": 187.3274383544922,
|
| 1373 |
-
"learning_rate": 7.305e-06,
|
| 1374 |
-
"loss": 49.8905,
|
| 1375 |
-
"step": 975
|
| 1376 |
-
},
|
| 1377 |
-
{
|
| 1378 |
-
"epoch": 21.32967032967033,
|
| 1379 |
-
"grad_norm": 175.39138793945312,
|
| 1380 |
-
"learning_rate": 7.3425000000000004e-06,
|
| 1381 |
-
"loss": 50.4588,
|
| 1382 |
-
"step": 980
|
| 1383 |
-
},
|
| 1384 |
-
{
|
| 1385 |
-
"epoch": 21.439560439560438,
|
| 1386 |
-
"grad_norm": 173.7222900390625,
|
| 1387 |
-
"learning_rate": 7.3800000000000005e-06,
|
| 1388 |
-
"loss": 48.5218,
|
| 1389 |
-
"step": 985
|
| 1390 |
-
},
|
| 1391 |
-
{
|
| 1392 |
-
"epoch": 21.54945054945055,
|
| 1393 |
-
"grad_norm": 429.6211242675781,
|
| 1394 |
-
"learning_rate": 7.4175e-06,
|
| 1395 |
-
"loss": 48.636,
|
| 1396 |
-
"step": 990
|
| 1397 |
-
},
|
| 1398 |
-
{
|
| 1399 |
-
"epoch": 21.65934065934066,
|
| 1400 |
-
"grad_norm": 178.93069458007812,
|
| 1401 |
-
"learning_rate": 7.455e-06,
|
| 1402 |
-
"loss": 49.8087,
|
| 1403 |
-
"step": 995
|
| 1404 |
-
},
|
| 1405 |
-
{
|
| 1406 |
-
"epoch": 21.76923076923077,
|
| 1407 |
-
"grad_norm": 232.20291137695312,
|
| 1408 |
-
"learning_rate": 7.4925e-06,
|
| 1409 |
-
"loss": 47.1187,
|
| 1410 |
-
"step": 1000
|
| 1411 |
-
},
|
| 1412 |
-
{
|
| 1413 |
-
"epoch": 21.87912087912088,
|
| 1414 |
-
"grad_norm": 181.4752655029297,
|
| 1415 |
-
"learning_rate": 7.53e-06,
|
| 1416 |
-
"loss": 51.2776,
|
| 1417 |
-
"step": 1005
|
| 1418 |
-
},
|
| 1419 |
-
{
|
| 1420 |
-
"epoch": 21.98901098901099,
|
| 1421 |
-
"grad_norm": 167.86483764648438,
|
| 1422 |
-
"learning_rate": 7.567499999999999e-06,
|
| 1423 |
-
"loss": 46.6794,
|
| 1424 |
-
"step": 1010
|
| 1425 |
-
},
|
| 1426 |
-
{
|
| 1427 |
-
"epoch": 22.087912087912088,
|
| 1428 |
-
"grad_norm": 172.2127685546875,
|
| 1429 |
-
"learning_rate": 7.605e-06,
|
| 1430 |
-
"loss": 44.9643,
|
| 1431 |
-
"step": 1015
|
| 1432 |
-
},
|
| 1433 |
-
{
|
| 1434 |
-
"epoch": 22.197802197802197,
|
| 1435 |
-
"grad_norm": 184.90206909179688,
|
| 1436 |
-
"learning_rate": 7.6425e-06,
|
| 1437 |
-
"loss": 45.4152,
|
| 1438 |
-
"step": 1020
|
| 1439 |
-
},
|
| 1440 |
-
{
|
| 1441 |
-
"epoch": 22.307692307692307,
|
| 1442 |
-
"grad_norm": 155.5305938720703,
|
| 1443 |
-
"learning_rate": 7.680000000000001e-06,
|
| 1444 |
-
"loss": 45.4304,
|
| 1445 |
-
"step": 1025
|
| 1446 |
-
},
|
| 1447 |
-
{
|
| 1448 |
-
"epoch": 22.417582417582416,
|
| 1449 |
-
"grad_norm": 670.8698120117188,
|
| 1450 |
-
"learning_rate": 7.7175e-06,
|
| 1451 |
-
"loss": 44.4255,
|
| 1452 |
-
"step": 1030
|
| 1453 |
-
},
|
| 1454 |
-
{
|
| 1455 |
-
"epoch": 22.52747252747253,
|
| 1456 |
-
"grad_norm": 162.875244140625,
|
| 1457 |
-
"learning_rate": 7.755000000000001e-06,
|
| 1458 |
-
"loss": 43.0364,
|
| 1459 |
-
"step": 1035
|
| 1460 |
-
},
|
| 1461 |
-
{
|
| 1462 |
-
"epoch": 22.63736263736264,
|
| 1463 |
-
"grad_norm": 159.903076171875,
|
| 1464 |
-
"learning_rate": 7.7925e-06,
|
| 1465 |
-
"loss": 44.4935,
|
| 1466 |
-
"step": 1040
|
| 1467 |
-
},
|
| 1468 |
-
{
|
| 1469 |
-
"epoch": 22.747252747252748,
|
| 1470 |
-
"grad_norm": 161.0757598876953,
|
| 1471 |
-
"learning_rate": 7.830000000000001e-06,
|
| 1472 |
-
"loss": 43.3238,
|
| 1473 |
-
"step": 1045
|
| 1474 |
-
},
|
| 1475 |
-
{
|
| 1476 |
-
"epoch": 22.857142857142858,
|
| 1477 |
-
"grad_norm": 140.3941650390625,
|
| 1478 |
-
"learning_rate": 7.8675e-06,
|
| 1479 |
-
"loss": 42.2547,
|
| 1480 |
-
"step": 1050
|
| 1481 |
-
},
|
| 1482 |
-
{
|
| 1483 |
-
"epoch": 22.967032967032967,
|
| 1484 |
-
"grad_norm": 164.62034606933594,
|
| 1485 |
-
"learning_rate": 7.905000000000001e-06,
|
| 1486 |
-
"loss": 42.7775,
|
| 1487 |
-
"step": 1055
|
| 1488 |
-
},
|
| 1489 |
-
{
|
| 1490 |
-
"epoch": 23.065934065934066,
|
| 1491 |
-
"grad_norm": 148.73211669921875,
|
| 1492 |
-
"learning_rate": 7.942499999999999e-06,
|
| 1493 |
-
"loss": 42.7699,
|
| 1494 |
-
"step": 1060
|
| 1495 |
-
},
|
| 1496 |
-
{
|
| 1497 |
-
"epoch": 23.175824175824175,
|
| 1498 |
-
"grad_norm": 137.26893615722656,
|
| 1499 |
-
"learning_rate": 7.98e-06,
|
| 1500 |
-
"loss": 40.5411,
|
| 1501 |
-
"step": 1065
|
| 1502 |
-
},
|
| 1503 |
-
{
|
| 1504 |
-
"epoch": 23.285714285714285,
|
| 1505 |
-
"grad_norm": 143.8221893310547,
|
| 1506 |
-
"learning_rate": 8.017499999999999e-06,
|
| 1507 |
-
"loss": 39.9001,
|
| 1508 |
-
"step": 1070
|
| 1509 |
-
},
|
| 1510 |
-
{
|
| 1511 |
-
"epoch": 23.395604395604394,
|
| 1512 |
-
"grad_norm": 160.19586181640625,
|
| 1513 |
-
"learning_rate": 8.055e-06,
|
| 1514 |
-
"loss": 39.8523,
|
| 1515 |
-
"step": 1075
|
| 1516 |
-
},
|
| 1517 |
-
{
|
| 1518 |
-
"epoch": 23.505494505494504,
|
| 1519 |
-
"grad_norm": 146.14981079101562,
|
| 1520 |
-
"learning_rate": 8.0925e-06,
|
| 1521 |
-
"loss": 39.7639,
|
| 1522 |
-
"step": 1080
|
| 1523 |
-
},
|
| 1524 |
-
{
|
| 1525 |
-
"epoch": 23.615384615384617,
|
| 1526 |
-
"grad_norm": 132.55712890625,
|
| 1527 |
-
"learning_rate": 8.13e-06,
|
| 1528 |
-
"loss": 39.7051,
|
| 1529 |
-
"step": 1085
|
| 1530 |
-
},
|
| 1531 |
-
{
|
| 1532 |
-
"epoch": 23.725274725274726,
|
| 1533 |
-
"grad_norm": 137.0689239501953,
|
| 1534 |
-
"learning_rate": 8.1675e-06,
|
| 1535 |
-
"loss": 39.1014,
|
| 1536 |
-
"step": 1090
|
| 1537 |
-
},
|
| 1538 |
-
{
|
| 1539 |
-
"epoch": 23.835164835164836,
|
| 1540 |
-
"grad_norm": 125.66040802001953,
|
| 1541 |
-
"learning_rate": 8.205e-06,
|
| 1542 |
-
"loss": 38.4976,
|
| 1543 |
-
"step": 1095
|
| 1544 |
-
},
|
| 1545 |
-
{
|
| 1546 |
-
"epoch": 23.945054945054945,
|
| 1547 |
-
"grad_norm": 118.39325714111328,
|
| 1548 |
-
"learning_rate": 8.2425e-06,
|
| 1549 |
-
"loss": 39.2203,
|
| 1550 |
-
"step": 1100
|
| 1551 |
-
},
|
| 1552 |
-
{
|
| 1553 |
-
"epoch": 24.043956043956044,
|
| 1554 |
-
"grad_norm": 131.4208221435547,
|
| 1555 |
-
"learning_rate": 8.28e-06,
|
| 1556 |
-
"loss": 37.6028,
|
| 1557 |
-
"step": 1105
|
| 1558 |
-
},
|
| 1559 |
-
{
|
| 1560 |
-
"epoch": 24.153846153846153,
|
| 1561 |
-
"grad_norm": 131.80392456054688,
|
| 1562 |
-
"learning_rate": 8.3175e-06,
|
| 1563 |
-
"loss": 38.8382,
|
| 1564 |
-
"step": 1110
|
| 1565 |
-
},
|
| 1566 |
-
{
|
| 1567 |
-
"epoch": 24.263736263736263,
|
| 1568 |
-
"grad_norm": 120.43126678466797,
|
| 1569 |
-
"learning_rate": 8.355e-06,
|
| 1570 |
-
"loss": 37.7083,
|
| 1571 |
-
"step": 1115
|
| 1572 |
-
},
|
| 1573 |
-
{
|
| 1574 |
-
"epoch": 24.373626373626372,
|
| 1575 |
-
"grad_norm": 147.37368774414062,
|
| 1576 |
-
"learning_rate": 8.3925e-06,
|
| 1577 |
-
"loss": 36.4178,
|
| 1578 |
-
"step": 1120
|
| 1579 |
-
},
|
| 1580 |
-
{
|
| 1581 |
-
"epoch": 24.483516483516482,
|
| 1582 |
-
"grad_norm": 162.19711303710938,
|
| 1583 |
-
"learning_rate": 8.43e-06,
|
| 1584 |
-
"loss": 36.3008,
|
| 1585 |
-
"step": 1125
|
| 1586 |
-
},
|
| 1587 |
-
{
|
| 1588 |
-
"epoch": 24.593406593406595,
|
| 1589 |
-
"grad_norm": 128.9326171875,
|
| 1590 |
-
"learning_rate": 8.4675e-06,
|
| 1591 |
-
"loss": 35.7747,
|
| 1592 |
-
"step": 1130
|
| 1593 |
-
},
|
| 1594 |
-
{
|
| 1595 |
-
"epoch": 24.703296703296704,
|
| 1596 |
-
"grad_norm": 123.9283218383789,
|
| 1597 |
-
"learning_rate": 8.504999999999999e-06,
|
| 1598 |
-
"loss": 33.9809,
|
| 1599 |
-
"step": 1135
|
| 1600 |
-
},
|
| 1601 |
-
{
|
| 1602 |
-
"epoch": 24.813186813186814,
|
| 1603 |
-
"grad_norm": 125.92423248291016,
|
| 1604 |
-
"learning_rate": 8.5425e-06,
|
| 1605 |
-
"loss": 35.4138,
|
| 1606 |
-
"step": 1140
|
| 1607 |
-
},
|
| 1608 |
-
{
|
| 1609 |
-
"epoch": 24.923076923076923,
|
| 1610 |
-
"grad_norm": 129.04722595214844,
|
| 1611 |
-
"learning_rate": 8.58e-06,
|
| 1612 |
-
"loss": 35.177,
|
| 1613 |
-
"step": 1145
|
| 1614 |
-
},
|
| 1615 |
-
{
|
| 1616 |
-
"epoch": 25.021978021978022,
|
| 1617 |
-
"grad_norm": 126.06185913085938,
|
| 1618 |
-
"learning_rate": 8.6175e-06,
|
| 1619 |
-
"loss": 34.5535,
|
| 1620 |
-
"step": 1150
|
| 1621 |
-
},
|
| 1622 |
-
{
|
| 1623 |
-
"epoch": 25.13186813186813,
|
| 1624 |
-
"grad_norm": 122.82085418701172,
|
| 1625 |
-
"learning_rate": 8.655e-06,
|
| 1626 |
-
"loss": 34.716,
|
| 1627 |
-
"step": 1155
|
| 1628 |
-
},
|
| 1629 |
-
{
|
| 1630 |
-
"epoch": 25.24175824175824,
|
| 1631 |
-
"grad_norm": 110.43183898925781,
|
| 1632 |
-
"learning_rate": 8.6925e-06,
|
| 1633 |
-
"loss": 33.8336,
|
| 1634 |
-
"step": 1160
|
| 1635 |
-
},
|
| 1636 |
-
{
|
| 1637 |
-
"epoch": 25.35164835164835,
|
| 1638 |
-
"grad_norm": 111.39581298828125,
|
| 1639 |
-
"learning_rate": 8.73e-06,
|
| 1640 |
-
"loss": 33.2828,
|
| 1641 |
-
"step": 1165
|
| 1642 |
-
},
|
| 1643 |
-
{
|
| 1644 |
-
"epoch": 25.46153846153846,
|
| 1645 |
-
"grad_norm": 133.24420166015625,
|
| 1646 |
-
"learning_rate": 8.7675e-06,
|
| 1647 |
-
"loss": 33.4885,
|
| 1648 |
-
"step": 1170
|
| 1649 |
-
},
|
| 1650 |
-
{
|
| 1651 |
-
"epoch": 25.571428571428573,
|
| 1652 |
-
"grad_norm": 117.64434051513672,
|
| 1653 |
-
"learning_rate": 8.805e-06,
|
| 1654 |
-
"loss": 33.8398,
|
| 1655 |
-
"step": 1175
|
| 1656 |
-
},
|
| 1657 |
-
{
|
| 1658 |
-
"epoch": 25.681318681318682,
|
| 1659 |
-
"grad_norm": 117.87113952636719,
|
| 1660 |
-
"learning_rate": 8.8425e-06,
|
| 1661 |
-
"loss": 31.6979,
|
| 1662 |
-
"step": 1180
|
| 1663 |
-
},
|
| 1664 |
-
{
|
| 1665 |
-
"epoch": 25.791208791208792,
|
| 1666 |
-
"grad_norm": 115.08634948730469,
|
| 1667 |
-
"learning_rate": 8.88e-06,
|
| 1668 |
-
"loss": 32.0385,
|
| 1669 |
-
"step": 1185
|
| 1670 |
-
},
|
| 1671 |
-
{
|
| 1672 |
-
"epoch": 25.9010989010989,
|
| 1673 |
-
"grad_norm": 127.63653564453125,
|
| 1674 |
-
"learning_rate": 8.9175e-06,
|
| 1675 |
-
"loss": 31.6513,
|
| 1676 |
-
"step": 1190
|
| 1677 |
-
},
|
| 1678 |
-
{
|
| 1679 |
-
"epoch": 26.0,
|
| 1680 |
-
"grad_norm": 283.0741271972656,
|
| 1681 |
-
"learning_rate": 8.955e-06,
|
| 1682 |
-
"loss": 33.2661,
|
| 1683 |
-
"step": 1195
|
| 1684 |
-
},
|
| 1685 |
-
{
|
| 1686 |
-
"epoch": 26.10989010989011,
|
| 1687 |
-
"grad_norm": 110.21710968017578,
|
| 1688 |
-
"learning_rate": 8.9925e-06,
|
| 1689 |
-
"loss": 30.7823,
|
| 1690 |
-
"step": 1200
|
| 1691 |
-
},
|
| 1692 |
-
{
|
| 1693 |
-
"epoch": 26.21978021978022,
|
| 1694 |
-
"grad_norm": 100.66688537597656,
|
| 1695 |
-
"learning_rate": 9.03e-06,
|
| 1696 |
-
"loss": 31.5021,
|
| 1697 |
-
"step": 1205
|
| 1698 |
-
},
|
| 1699 |
-
{
|
| 1700 |
-
"epoch": 26.32967032967033,
|
| 1701 |
-
"grad_norm": 105.822021484375,
|
| 1702 |
-
"learning_rate": 9.067500000000001e-06,
|
| 1703 |
-
"loss": 30.0609,
|
| 1704 |
-
"step": 1210
|
| 1705 |
-
},
|
| 1706 |
-
{
|
| 1707 |
-
"epoch": 26.439560439560438,
|
| 1708 |
-
"grad_norm": 99.89389038085938,
|
| 1709 |
-
"learning_rate": 9.105e-06,
|
| 1710 |
-
"loss": 30.5733,
|
| 1711 |
-
"step": 1215
|
| 1712 |
-
},
|
| 1713 |
-
{
|
| 1714 |
-
"epoch": 26.54945054945055,
|
| 1715 |
-
"grad_norm": 96.10974884033203,
|
| 1716 |
-
"learning_rate": 9.142500000000001e-06,
|
| 1717 |
-
"loss": 29.9534,
|
| 1718 |
-
"step": 1220
|
| 1719 |
-
},
|
| 1720 |
-
{
|
| 1721 |
-
"epoch": 26.65934065934066,
|
| 1722 |
-
"grad_norm": 112.5010986328125,
|
| 1723 |
-
"learning_rate": 9.18e-06,
|
| 1724 |
-
"loss": 30.4645,
|
| 1725 |
-
"step": 1225
|
| 1726 |
-
},
|
| 1727 |
-
{
|
| 1728 |
-
"epoch": 26.76923076923077,
|
| 1729 |
-
"grad_norm": 113.7469253540039,
|
| 1730 |
-
"learning_rate": 9.217500000000001e-06,
|
| 1731 |
-
"loss": 29.8336,
|
| 1732 |
-
"step": 1230
|
| 1733 |
-
},
|
| 1734 |
-
{
|
| 1735 |
-
"epoch": 26.87912087912088,
|
| 1736 |
-
"grad_norm": 99.2851791381836,
|
| 1737 |
-
"learning_rate": 9.255e-06,
|
| 1738 |
-
"loss": 28.8012,
|
| 1739 |
-
"step": 1235
|
| 1740 |
-
},
|
| 1741 |
-
{
|
| 1742 |
-
"epoch": 26.98901098901099,
|
| 1743 |
-
"grad_norm": 92.29454040527344,
|
| 1744 |
-
"learning_rate": 9.292500000000001e-06,
|
| 1745 |
-
"loss": 29.5607,
|
| 1746 |
-
"step": 1240
|
| 1747 |
-
},
|
| 1748 |
-
{
|
| 1749 |
-
"epoch": 27.087912087912088,
|
| 1750 |
-
"grad_norm": 368.2291564941406,
|
| 1751 |
-
"learning_rate": 9.33e-06,
|
| 1752 |
-
"loss": 57.1964,
|
| 1753 |
-
"step": 1245
|
| 1754 |
-
},
|
| 1755 |
-
{
|
| 1756 |
-
"epoch": 27.197802197802197,
|
| 1757 |
-
"grad_norm": 87.99185943603516,
|
| 1758 |
-
"learning_rate": 9.367500000000001e-06,
|
| 1759 |
-
"loss": 29.8244,
|
| 1760 |
-
"step": 1250
|
| 1761 |
-
},
|
| 1762 |
-
{
|
| 1763 |
-
"epoch": 27.307692307692307,
|
| 1764 |
-
"grad_norm": 87.86891174316406,
|
| 1765 |
-
"learning_rate": 9.405e-06,
|
| 1766 |
-
"loss": 29.159,
|
| 1767 |
-
"step": 1255
|
| 1768 |
-
},
|
| 1769 |
-
{
|
| 1770 |
-
"epoch": 27.417582417582416,
|
| 1771 |
-
"grad_norm": 106.52743530273438,
|
| 1772 |
-
"learning_rate": 9.4425e-06,
|
| 1773 |
-
"loss": 28.5417,
|
| 1774 |
-
"step": 1260
|
| 1775 |
-
},
|
| 1776 |
-
{
|
| 1777 |
-
"epoch": 27.52747252747253,
|
| 1778 |
-
"grad_norm": 105.37548828125,
|
| 1779 |
-
"learning_rate": 9.48e-06,
|
| 1780 |
-
"loss": 29.6107,
|
| 1781 |
-
"step": 1265
|
| 1782 |
-
},
|
| 1783 |
-
{
|
| 1784 |
-
"epoch": 27.63736263736264,
|
| 1785 |
-
"grad_norm": 126.07006072998047,
|
| 1786 |
-
"learning_rate": 9.5175e-06,
|
| 1787 |
-
"loss": 28.5252,
|
| 1788 |
-
"step": 1270
|
| 1789 |
-
},
|
| 1790 |
-
{
|
| 1791 |
-
"epoch": 27.747252747252748,
|
| 1792 |
-
"grad_norm": 102.43346405029297,
|
| 1793 |
-
"learning_rate": 9.555e-06,
|
| 1794 |
-
"loss": 26.8626,
|
| 1795 |
-
"step": 1275
|
| 1796 |
-
},
|
| 1797 |
-
{
|
| 1798 |
-
"epoch": 27.857142857142858,
|
| 1799 |
-
"grad_norm": 95.87886047363281,
|
| 1800 |
-
"learning_rate": 9.5925e-06,
|
| 1801 |
-
"loss": 27.4449,
|
| 1802 |
-
"step": 1280
|
| 1803 |
-
},
|
| 1804 |
-
{
|
| 1805 |
-
"epoch": 27.967032967032967,
|
| 1806 |
-
"grad_norm": 183.47979736328125,
|
| 1807 |
-
"learning_rate": 9.630000000000001e-06,
|
| 1808 |
-
"loss": 26.4121,
|
| 1809 |
-
"step": 1285
|
| 1810 |
-
},
|
| 1811 |
-
{
|
| 1812 |
-
"epoch": 28.065934065934066,
|
| 1813 |
-
"grad_norm": 95.40010070800781,
|
| 1814 |
-
"learning_rate": 9.6675e-06,
|
| 1815 |
-
"loss": 35.9095,
|
| 1816 |
-
"step": 1290
|
| 1817 |
-
},
|
| 1818 |
-
{
|
| 1819 |
-
"epoch": 28.175824175824175,
|
| 1820 |
-
"grad_norm": 90.59423828125,
|
| 1821 |
-
"learning_rate": 9.705000000000001e-06,
|
| 1822 |
-
"loss": 26.6954,
|
| 1823 |
-
"step": 1295
|
| 1824 |
-
},
|
| 1825 |
-
{
|
| 1826 |
-
"epoch": 28.285714285714285,
|
| 1827 |
-
"grad_norm": 91.06011199951172,
|
| 1828 |
-
"learning_rate": 9.7425e-06,
|
| 1829 |
-
"loss": 26.3503,
|
| 1830 |
-
"step": 1300
|
| 1831 |
-
},
|
| 1832 |
-
{
|
| 1833 |
-
"epoch": 28.395604395604394,
|
| 1834 |
-
"grad_norm": 88.42830657958984,
|
| 1835 |
-
"learning_rate": 9.780000000000001e-06,
|
| 1836 |
-
"loss": 26.0797,
|
| 1837 |
-
"step": 1305
|
| 1838 |
-
},
|
| 1839 |
-
{
|
| 1840 |
-
"epoch": 28.505494505494504,
|
| 1841 |
-
"grad_norm": 80.00355529785156,
|
| 1842 |
-
"learning_rate": 9.8175e-06,
|
| 1843 |
-
"loss": 25.8798,
|
| 1844 |
-
"step": 1310
|
| 1845 |
-
},
|
| 1846 |
-
{
|
| 1847 |
-
"epoch": 28.615384615384617,
|
| 1848 |
-
"grad_norm": 81.79782104492188,
|
| 1849 |
-
"learning_rate": 9.855000000000001e-06,
|
| 1850 |
-
"loss": 26.4297,
|
| 1851 |
-
"step": 1315
|
| 1852 |
},
|
| 1853 |
{
|
| 1854 |
-
"epoch":
|
| 1855 |
-
"grad_norm":
|
| 1856 |
-
"learning_rate":
|
| 1857 |
-
"loss":
|
| 1858 |
-
"step":
|
| 1859 |
},
|
| 1860 |
{
|
| 1861 |
-
"epoch":
|
| 1862 |
-
"grad_norm":
|
| 1863 |
-
"learning_rate":
|
| 1864 |
-
"loss":
|
| 1865 |
-
"step":
|
| 1866 |
},
|
| 1867 |
{
|
| 1868 |
-
"epoch":
|
| 1869 |
-
"grad_norm":
|
| 1870 |
-
"learning_rate":
|
| 1871 |
-
"loss":
|
| 1872 |
-
"step":
|
| 1873 |
},
|
| 1874 |
{
|
| 1875 |
-
"epoch":
|
| 1876 |
-
"grad_norm":
|
| 1877 |
-
"learning_rate":
|
| 1878 |
-
"loss":
|
| 1879 |
-
"step":
|
| 1880 |
},
|
| 1881 |
{
|
| 1882 |
-
"epoch":
|
| 1883 |
-
"grad_norm":
|
| 1884 |
-
"learning_rate":
|
| 1885 |
-
"loss":
|
| 1886 |
-
"step":
|
| 1887 |
},
|
| 1888 |
{
|
| 1889 |
-
"epoch":
|
| 1890 |
-
"grad_norm":
|
| 1891 |
-
"learning_rate":
|
| 1892 |
-
"loss":
|
| 1893 |
-
"step":
|
| 1894 |
},
|
| 1895 |
{
|
| 1896 |
-
"epoch":
|
| 1897 |
-
"grad_norm":
|
| 1898 |
-
"learning_rate":
|
| 1899 |
-
"loss":
|
| 1900 |
-
"step":
|
| 1901 |
},
|
| 1902 |
{
|
| 1903 |
-
"epoch":
|
| 1904 |
-
"grad_norm":
|
| 1905 |
-
"learning_rate":
|
| 1906 |
-
"loss":
|
| 1907 |
-
"step":
|
| 1908 |
},
|
| 1909 |
{
|
| 1910 |
-
"epoch":
|
| 1911 |
-
"grad_norm":
|
| 1912 |
-
"learning_rate":
|
| 1913 |
-
"loss":
|
| 1914 |
-
"step":
|
| 1915 |
},
|
| 1916 |
{
|
| 1917 |
-
"epoch":
|
| 1918 |
-
"grad_norm":
|
| 1919 |
-
"learning_rate":
|
| 1920 |
-
"loss":
|
| 1921 |
-
"step":
|
| 1922 |
},
|
| 1923 |
{
|
| 1924 |
-
"epoch":
|
| 1925 |
-
"grad_norm":
|
| 1926 |
-
"learning_rate":
|
| 1927 |
-
"loss":
|
| 1928 |
-
"step":
|
| 1929 |
},
|
| 1930 |
{
|
| 1931 |
-
"epoch":
|
| 1932 |
-
"grad_norm":
|
| 1933 |
-
"learning_rate":
|
| 1934 |
-
"loss":
|
| 1935 |
-
"step":
|
| 1936 |
},
|
| 1937 |
{
|
| 1938 |
-
"epoch":
|
| 1939 |
-
"grad_norm":
|
| 1940 |
-
"learning_rate":
|
| 1941 |
-
"loss":
|
| 1942 |
-
"step":
|
| 1943 |
},
|
| 1944 |
{
|
| 1945 |
-
"epoch":
|
| 1946 |
-
"grad_norm":
|
| 1947 |
-
"learning_rate":
|
| 1948 |
-
"loss":
|
| 1949 |
-
"step":
|
| 1950 |
},
|
| 1951 |
{
|
| 1952 |
-
"epoch":
|
| 1953 |
-
"grad_norm":
|
| 1954 |
-
"learning_rate":
|
| 1955 |
-
"loss":
|
| 1956 |
-
"step":
|
| 1957 |
},
|
| 1958 |
{
|
| 1959 |
-
"epoch":
|
| 1960 |
-
"grad_norm":
|
| 1961 |
-
"learning_rate":
|
| 1962 |
-
"loss":
|
| 1963 |
-
"step":
|
| 1964 |
},
|
| 1965 |
{
|
| 1966 |
-
"epoch":
|
| 1967 |
-
"grad_norm":
|
| 1968 |
-
"learning_rate":
|
| 1969 |
-
"loss":
|
| 1970 |
-
"step":
|
| 1971 |
},
|
| 1972 |
{
|
| 1973 |
-
"epoch":
|
| 1974 |
-
"grad_norm":
|
| 1975 |
-
"learning_rate":
|
| 1976 |
-
"loss":
|
| 1977 |
-
"step":
|
| 1978 |
},
|
| 1979 |
{
|
| 1980 |
-
"epoch":
|
| 1981 |
-
"grad_norm":
|
| 1982 |
-
"learning_rate":
|
| 1983 |
-
"loss":
|
| 1984 |
-
"step":
|
| 1985 |
},
|
| 1986 |
{
|
| 1987 |
-
"epoch":
|
| 1988 |
-
"grad_norm":
|
| 1989 |
-
"learning_rate":
|
| 1990 |
-
"loss":
|
| 1991 |
-
"step":
|
| 1992 |
},
|
| 1993 |
{
|
| 1994 |
-
"epoch":
|
| 1995 |
-
"grad_norm":
|
| 1996 |
-
"learning_rate":
|
| 1997 |
-
"loss":
|
| 1998 |
-
"step":
|
| 1999 |
},
|
| 2000 |
{
|
| 2001 |
-
"epoch":
|
| 2002 |
-
"grad_norm":
|
| 2003 |
-
"learning_rate":
|
| 2004 |
-
"loss":
|
| 2005 |
-
"step":
|
| 2006 |
},
|
| 2007 |
{
|
| 2008 |
-
"epoch":
|
| 2009 |
-
"grad_norm":
|
| 2010 |
-
"learning_rate":
|
| 2011 |
-
"loss":
|
| 2012 |
-
"step":
|
| 2013 |
},
|
| 2014 |
{
|
| 2015 |
-
"epoch":
|
| 2016 |
-
"grad_norm":
|
| 2017 |
-
"learning_rate":
|
| 2018 |
-
"loss":
|
| 2019 |
-
"step":
|
| 2020 |
},
|
| 2021 |
{
|
| 2022 |
-
"epoch":
|
| 2023 |
-
"grad_norm":
|
| 2024 |
-
"learning_rate": 1.
|
| 2025 |
-
"loss":
|
| 2026 |
-
"step":
|
| 2027 |
},
|
| 2028 |
{
|
| 2029 |
-
"epoch": 31.
|
| 2030 |
-
"grad_norm":
|
| 2031 |
-
"learning_rate": 1.
|
| 2032 |
-
"loss":
|
| 2033 |
-
"step":
|
| 2034 |
},
|
| 2035 |
{
|
| 2036 |
-
"epoch":
|
| 2037 |
-
"grad_norm":
|
| 2038 |
"learning_rate": 1.08675e-05,
|
| 2039 |
-
"loss":
|
| 2040 |
"step": 1450
|
| 2041 |
},
|
| 2042 |
{
|
| 2043 |
-
"epoch":
|
| 2044 |
-
"grad_norm":
|
| 2045 |
-
"learning_rate": 1.0905e-05,
|
| 2046 |
-
"loss": 21.4437,
|
| 2047 |
-
"step": 1455
|
| 2048 |
-
},
|
| 2049 |
-
{
|
| 2050 |
-
"epoch": 31.76923076923077,
|
| 2051 |
-
"grad_norm": 79.01976013183594,
|
| 2052 |
-
"learning_rate": 1.09425e-05,
|
| 2053 |
-
"loss": 22.4324,
|
| 2054 |
-
"step": 1460
|
| 2055 |
-
},
|
| 2056 |
-
{
|
| 2057 |
-
"epoch": 31.87912087912088,
|
| 2058 |
-
"grad_norm": 94.18357849121094,
|
| 2059 |
-
"learning_rate": 1.098e-05,
|
| 2060 |
-
"loss": 23.0076,
|
| 2061 |
-
"step": 1465
|
| 2062 |
-
},
|
| 2063 |
-
{
|
| 2064 |
-
"epoch": 31.98901098901099,
|
| 2065 |
-
"grad_norm": 76.7142333984375,
|
| 2066 |
-
"learning_rate": 1.10175e-05,
|
| 2067 |
-
"loss": 21.8061,
|
| 2068 |
-
"step": 1470
|
| 2069 |
-
},
|
| 2070 |
-
{
|
| 2071 |
-
"epoch": 32.08791208791209,
|
| 2072 |
-
"grad_norm": 65.66316986083984,
|
| 2073 |
-
"learning_rate": 1.1055e-05,
|
| 2074 |
-
"loss": 22.7557,
|
| 2075 |
-
"step": 1475
|
| 2076 |
-
},
|
| 2077 |
-
{
|
| 2078 |
-
"epoch": 32.1978021978022,
|
| 2079 |
-
"grad_norm": 85.76339721679688,
|
| 2080 |
-
"learning_rate": 1.1092500000000001e-05,
|
| 2081 |
-
"loss": 20.9964,
|
| 2082 |
-
"step": 1480
|
| 2083 |
-
},
|
| 2084 |
-
{
|
| 2085 |
-
"epoch": 32.30769230769231,
|
| 2086 |
-
"grad_norm": 84.71622467041016,
|
| 2087 |
-
"learning_rate": 1.113e-05,
|
| 2088 |
-
"loss": 21.0421,
|
| 2089 |
-
"step": 1485
|
| 2090 |
-
},
|
| 2091 |
-
{
|
| 2092 |
-
"epoch": 32.417582417582416,
|
| 2093 |
-
"grad_norm": 94.56781768798828,
|
| 2094 |
-
"learning_rate": 1.1167500000000001e-05,
|
| 2095 |
-
"loss": 21.6047,
|
| 2096 |
-
"step": 1490
|
| 2097 |
-
},
|
| 2098 |
-
{
|
| 2099 |
-
"epoch": 32.527472527472526,
|
| 2100 |
-
"grad_norm": 71.27996063232422,
|
| 2101 |
-
"learning_rate": 1.1205e-05,
|
| 2102 |
-
"loss": 21.281,
|
| 2103 |
-
"step": 1495
|
| 2104 |
-
},
|
| 2105 |
-
{
|
| 2106 |
-
"epoch": 32.637362637362635,
|
| 2107 |
-
"grad_norm": 58.75380325317383,
|
| 2108 |
"learning_rate": 1.1242500000000001e-05,
|
| 2109 |
-
"loss":
|
| 2110 |
"step": 1500
|
| 2111 |
},
|
| 2112 |
{
|
| 2113 |
-
"epoch":
|
| 2114 |
-
"grad_norm":
|
| 2115 |
-
"learning_rate": 1.128e-05,
|
| 2116 |
-
"loss": 20.9145,
|
| 2117 |
-
"step": 1505
|
| 2118 |
-
},
|
| 2119 |
-
{
|
| 2120 |
-
"epoch": 32.857142857142854,
|
| 2121 |
-
"grad_norm": 109.56021881103516,
|
| 2122 |
-
"learning_rate": 1.13175e-05,
|
| 2123 |
-
"loss": 21.6375,
|
| 2124 |
-
"step": 1510
|
| 2125 |
-
},
|
| 2126 |
-
{
|
| 2127 |
-
"epoch": 32.967032967032964,
|
| 2128 |
-
"grad_norm": 63.0958251953125,
|
| 2129 |
-
"learning_rate": 1.1355e-05,
|
| 2130 |
-
"loss": 21.3934,
|
| 2131 |
-
"step": 1515
|
| 2132 |
-
},
|
| 2133 |
-
{
|
| 2134 |
-
"epoch": 33.065934065934066,
|
| 2135 |
-
"grad_norm": 72.25486755371094,
|
| 2136 |
-
"learning_rate": 1.13925e-05,
|
| 2137 |
-
"loss": 20.5177,
|
| 2138 |
-
"step": 1520
|
| 2139 |
-
},
|
| 2140 |
-
{
|
| 2141 |
-
"epoch": 33.175824175824175,
|
| 2142 |
-
"grad_norm": 79.46709442138672,
|
| 2143 |
-
"learning_rate": 1.143e-05,
|
| 2144 |
-
"loss": 20.1796,
|
| 2145 |
-
"step": 1525
|
| 2146 |
-
},
|
| 2147 |
-
{
|
| 2148 |
-
"epoch": 33.285714285714285,
|
| 2149 |
-
"grad_norm": 80.37206268310547,
|
| 2150 |
-
"learning_rate": 1.14675e-05,
|
| 2151 |
-
"loss": 20.519,
|
| 2152 |
-
"step": 1530
|
| 2153 |
-
},
|
| 2154 |
-
{
|
| 2155 |
-
"epoch": 33.395604395604394,
|
| 2156 |
-
"grad_norm": 88.30254364013672,
|
| 2157 |
-
"learning_rate": 1.1505e-05,
|
| 2158 |
-
"loss": 20.5289,
|
| 2159 |
-
"step": 1535
|
| 2160 |
-
},
|
| 2161 |
-
{
|
| 2162 |
-
"epoch": 33.505494505494504,
|
| 2163 |
-
"grad_norm": 59.99192428588867,
|
| 2164 |
-
"learning_rate": 1.15425e-05,
|
| 2165 |
-
"loss": 20.5064,
|
| 2166 |
-
"step": 1540
|
| 2167 |
-
},
|
| 2168 |
-
{
|
| 2169 |
-
"epoch": 33.61538461538461,
|
| 2170 |
-
"grad_norm": 71.29468536376953,
|
| 2171 |
-
"learning_rate": 1.1580000000000001e-05,
|
| 2172 |
-
"loss": 19.2822,
|
| 2173 |
-
"step": 1545
|
| 2174 |
-
},
|
| 2175 |
-
{
|
| 2176 |
-
"epoch": 33.72527472527472,
|
| 2177 |
-
"grad_norm": 67.42713928222656,
|
| 2178 |
"learning_rate": 1.16175e-05,
|
| 2179 |
-
"loss":
|
| 2180 |
"step": 1550
|
| 2181 |
},
|
| 2182 |
{
|
| 2183 |
-
"epoch":
|
| 2184 |
-
"grad_norm":
|
| 2185 |
-
"learning_rate": 1.1655000000000001e-05,
|
| 2186 |
-
"loss": 19.9386,
|
| 2187 |
-
"step": 1555
|
| 2188 |
-
},
|
| 2189 |
-
{
|
| 2190 |
-
"epoch": 33.94505494505494,
|
| 2191 |
-
"grad_norm": 133.632080078125,
|
| 2192 |
-
"learning_rate": 1.16925e-05,
|
| 2193 |
-
"loss": 20.3614,
|
| 2194 |
-
"step": 1560
|
| 2195 |
-
},
|
| 2196 |
-
{
|
| 2197 |
-
"epoch": 34.043956043956044,
|
| 2198 |
-
"grad_norm": 56.151954650878906,
|
| 2199 |
-
"learning_rate": 1.1730000000000001e-05,
|
| 2200 |
-
"loss": 18.905,
|
| 2201 |
-
"step": 1565
|
| 2202 |
-
},
|
| 2203 |
-
{
|
| 2204 |
-
"epoch": 34.15384615384615,
|
| 2205 |
-
"grad_norm": 72.71345520019531,
|
| 2206 |
-
"learning_rate": 1.17675e-05,
|
| 2207 |
-
"loss": 20.4241,
|
| 2208 |
-
"step": 1570
|
| 2209 |
-
},
|
| 2210 |
-
{
|
| 2211 |
-
"epoch": 34.26373626373626,
|
| 2212 |
-
"grad_norm": 74.98194122314453,
|
| 2213 |
-
"learning_rate": 1.1805000000000001e-05,
|
| 2214 |
-
"loss": 19.3519,
|
| 2215 |
-
"step": 1575
|
| 2216 |
-
},
|
| 2217 |
-
{
|
| 2218 |
-
"epoch": 34.37362637362637,
|
| 2219 |
-
"grad_norm": 67.34236145019531,
|
| 2220 |
-
"learning_rate": 1.18425e-05,
|
| 2221 |
-
"loss": 18.7936,
|
| 2222 |
-
"step": 1580
|
| 2223 |
-
},
|
| 2224 |
-
{
|
| 2225 |
-
"epoch": 34.48351648351648,
|
| 2226 |
-
"grad_norm": 72.32279205322266,
|
| 2227 |
-
"learning_rate": 1.1880000000000001e-05,
|
| 2228 |
-
"loss": 20.0671,
|
| 2229 |
-
"step": 1585
|
| 2230 |
-
},
|
| 2231 |
-
{
|
| 2232 |
-
"epoch": 34.59340659340659,
|
| 2233 |
-
"grad_norm": 54.22994613647461,
|
| 2234 |
-
"learning_rate": 1.19175e-05,
|
| 2235 |
-
"loss": 19.3878,
|
| 2236 |
-
"step": 1590
|
| 2237 |
-
},
|
| 2238 |
-
{
|
| 2239 |
-
"epoch": 34.7032967032967,
|
| 2240 |
-
"grad_norm": 72.89130401611328,
|
| 2241 |
-
"learning_rate": 1.1955000000000002e-05,
|
| 2242 |
-
"loss": 19.4714,
|
| 2243 |
-
"step": 1595
|
| 2244 |
-
},
|
| 2245 |
-
{
|
| 2246 |
-
"epoch": 34.81318681318681,
|
| 2247 |
-
"grad_norm": 65.4543685913086,
|
| 2248 |
"learning_rate": 1.19925e-05,
|
| 2249 |
-
"loss": 20.
|
| 2250 |
"step": 1600
|
| 2251 |
},
|
| 2252 |
{
|
| 2253 |
-
"epoch":
|
| 2254 |
-
"grad_norm":
|
| 2255 |
-
"learning_rate": 1.2030000000000002e-05,
|
| 2256 |
-
"loss": 18.6467,
|
| 2257 |
-
"step": 1605
|
| 2258 |
-
},
|
| 2259 |
-
{
|
| 2260 |
-
"epoch": 35.02197802197802,
|
| 2261 |
-
"grad_norm": 82.81990051269531,
|
| 2262 |
-
"learning_rate": 1.2067500000000001e-05,
|
| 2263 |
-
"loss": 26.0395,
|
| 2264 |
-
"step": 1610
|
| 2265 |
-
},
|
| 2266 |
-
{
|
| 2267 |
-
"epoch": 35.13186813186813,
|
| 2268 |
-
"grad_norm": 62.68242645263672,
|
| 2269 |
-
"learning_rate": 1.2105000000000002e-05,
|
| 2270 |
-
"loss": 19.2074,
|
| 2271 |
-
"step": 1615
|
| 2272 |
-
},
|
| 2273 |
-
{
|
| 2274 |
-
"epoch": 35.24175824175824,
|
| 2275 |
-
"grad_norm": 79.71977996826172,
|
| 2276 |
-
"learning_rate": 1.2142500000000001e-05,
|
| 2277 |
-
"loss": 20.164,
|
| 2278 |
-
"step": 1620
|
| 2279 |
-
},
|
| 2280 |
-
{
|
| 2281 |
-
"epoch": 35.35164835164835,
|
| 2282 |
-
"grad_norm": 82.37974548339844,
|
| 2283 |
-
"learning_rate": 1.2180000000000002e-05,
|
| 2284 |
-
"loss": 18.7425,
|
| 2285 |
-
"step": 1625
|
| 2286 |
-
},
|
| 2287 |
-
{
|
| 2288 |
-
"epoch": 35.46153846153846,
|
| 2289 |
-
"grad_norm": 68.9244155883789,
|
| 2290 |
-
"learning_rate": 1.22175e-05,
|
| 2291 |
-
"loss": 18.8626,
|
| 2292 |
-
"step": 1630
|
| 2293 |
-
},
|
| 2294 |
-
{
|
| 2295 |
-
"epoch": 35.57142857142857,
|
| 2296 |
-
"grad_norm": 72.45255279541016,
|
| 2297 |
-
"learning_rate": 1.2254999999999999e-05,
|
| 2298 |
-
"loss": 18.4994,
|
| 2299 |
-
"step": 1635
|
| 2300 |
-
},
|
| 2301 |
-
{
|
| 2302 |
-
"epoch": 35.68131868131868,
|
| 2303 |
-
"grad_norm": 76.19093322753906,
|
| 2304 |
-
"learning_rate": 1.22925e-05,
|
| 2305 |
-
"loss": 18.911,
|
| 2306 |
-
"step": 1640
|
| 2307 |
-
},
|
| 2308 |
-
{
|
| 2309 |
-
"epoch": 35.79120879120879,
|
| 2310 |
-
"grad_norm": 50.99027633666992,
|
| 2311 |
-
"learning_rate": 1.2329999999999999e-05,
|
| 2312 |
-
"loss": 18.2682,
|
| 2313 |
-
"step": 1645
|
| 2314 |
-
},
|
| 2315 |
-
{
|
| 2316 |
-
"epoch": 35.9010989010989,
|
| 2317 |
-
"grad_norm": 79.6258316040039,
|
| 2318 |
"learning_rate": 1.23675e-05,
|
| 2319 |
-
"loss": 19.
|
| 2320 |
"step": 1650
|
| 2321 |
},
|
| 2322 |
{
|
| 2323 |
-
"epoch":
|
| 2324 |
-
"grad_norm":
|
| 2325 |
-
"learning_rate": 1.2404999999999999e-05,
|
| 2326 |
-
"loss": 18.7113,
|
| 2327 |
-
"step": 1655
|
| 2328 |
-
},
|
| 2329 |
-
{
|
| 2330 |
-
"epoch": 36.10989010989011,
|
| 2331 |
-
"grad_norm": 52.13261032104492,
|
| 2332 |
-
"learning_rate": 1.24425e-05,
|
| 2333 |
-
"loss": 17.3056,
|
| 2334 |
-
"step": 1660
|
| 2335 |
-
},
|
| 2336 |
-
{
|
| 2337 |
-
"epoch": 36.21978021978022,
|
| 2338 |
-
"grad_norm": 64.57988739013672,
|
| 2339 |
-
"learning_rate": 1.2479999999999999e-05,
|
| 2340 |
-
"loss": 17.8365,
|
| 2341 |
-
"step": 1665
|
| 2342 |
-
},
|
| 2343 |
-
{
|
| 2344 |
-
"epoch": 36.32967032967033,
|
| 2345 |
-
"grad_norm": 111.00614929199219,
|
| 2346 |
-
"learning_rate": 1.25175e-05,
|
| 2347 |
-
"loss": 18.5584,
|
| 2348 |
-
"step": 1670
|
| 2349 |
-
},
|
| 2350 |
-
{
|
| 2351 |
-
"epoch": 36.43956043956044,
|
| 2352 |
-
"grad_norm": 77.30622863769531,
|
| 2353 |
-
"learning_rate": 1.2555e-05,
|
| 2354 |
-
"loss": 18.0333,
|
| 2355 |
-
"step": 1675
|
| 2356 |
-
},
|
| 2357 |
-
{
|
| 2358 |
-
"epoch": 36.54945054945055,
|
| 2359 |
-
"grad_norm": 60.33699417114258,
|
| 2360 |
-
"learning_rate": 1.25925e-05,
|
| 2361 |
-
"loss": 18.5811,
|
| 2362 |
-
"step": 1680
|
| 2363 |
-
},
|
| 2364 |
-
{
|
| 2365 |
-
"epoch": 36.65934065934066,
|
| 2366 |
-
"grad_norm": 84.54650115966797,
|
| 2367 |
-
"learning_rate": 1.263e-05,
|
| 2368 |
-
"loss": 18.7701,
|
| 2369 |
-
"step": 1685
|
| 2370 |
-
},
|
| 2371 |
-
{
|
| 2372 |
-
"epoch": 36.76923076923077,
|
| 2373 |
-
"grad_norm": 73.11846923828125,
|
| 2374 |
-
"learning_rate": 1.26675e-05,
|
| 2375 |
-
"loss": 18.816,
|
| 2376 |
-
"step": 1690
|
| 2377 |
-
},
|
| 2378 |
-
{
|
| 2379 |
-
"epoch": 36.879120879120876,
|
| 2380 |
-
"grad_norm": 61.516761779785156,
|
| 2381 |
-
"learning_rate": 1.2705e-05,
|
| 2382 |
-
"loss": 17.5013,
|
| 2383 |
-
"step": 1695
|
| 2384 |
-
},
|
| 2385 |
-
{
|
| 2386 |
-
"epoch": 36.98901098901099,
|
| 2387 |
-
"grad_norm": 74.83867645263672,
|
| 2388 |
"learning_rate": 1.27425e-05,
|
| 2389 |
-
"loss":
|
| 2390 |
"step": 1700
|
| 2391 |
},
|
| 2392 |
{
|
| 2393 |
-
"epoch":
|
| 2394 |
-
"grad_norm":
|
| 2395 |
-
"learning_rate": 1.278e-05,
|
| 2396 |
-
"loss": 20.9143,
|
| 2397 |
-
"step": 1705
|
| 2398 |
-
},
|
| 2399 |
-
{
|
| 2400 |
-
"epoch": 37.1978021978022,
|
| 2401 |
-
"grad_norm": 59.48684310913086,
|
| 2402 |
-
"learning_rate": 1.28175e-05,
|
| 2403 |
-
"loss": 17.725,
|
| 2404 |
-
"step": 1710
|
| 2405 |
-
},
|
| 2406 |
-
{
|
| 2407 |
-
"epoch": 37.30769230769231,
|
| 2408 |
-
"grad_norm": 71.78469848632812,
|
| 2409 |
-
"learning_rate": 1.2855e-05,
|
| 2410 |
-
"loss": 17.7316,
|
| 2411 |
-
"step": 1715
|
| 2412 |
-
},
|
| 2413 |
-
{
|
| 2414 |
-
"epoch": 37.417582417582416,
|
| 2415 |
-
"grad_norm": 82.73149871826172,
|
| 2416 |
-
"learning_rate": 1.28925e-05,
|
| 2417 |
-
"loss": 17.8607,
|
| 2418 |
-
"step": 1720
|
| 2419 |
-
},
|
| 2420 |
-
{
|
| 2421 |
-
"epoch": 37.527472527472526,
|
| 2422 |
-
"grad_norm": 104.99292755126953,
|
| 2423 |
-
"learning_rate": 1.293e-05,
|
| 2424 |
-
"loss": 17.7329,
|
| 2425 |
-
"step": 1725
|
| 2426 |
-
},
|
| 2427 |
-
{
|
| 2428 |
-
"epoch": 37.637362637362635,
|
| 2429 |
-
"grad_norm": 56.97990417480469,
|
| 2430 |
-
"learning_rate": 1.29675e-05,
|
| 2431 |
-
"loss": 17.7514,
|
| 2432 |
-
"step": 1730
|
| 2433 |
-
},
|
| 2434 |
-
{
|
| 2435 |
-
"epoch": 37.747252747252745,
|
| 2436 |
-
"grad_norm": 76.46744537353516,
|
| 2437 |
-
"learning_rate": 1.3005e-05,
|
| 2438 |
-
"loss": 17.467,
|
| 2439 |
-
"step": 1735
|
| 2440 |
-
},
|
| 2441 |
-
{
|
| 2442 |
-
"epoch": 37.857142857142854,
|
| 2443 |
-
"grad_norm": 67.00051879882812,
|
| 2444 |
-
"learning_rate": 1.3042500000000001e-05,
|
| 2445 |
-
"loss": 17.9124,
|
| 2446 |
-
"step": 1740
|
| 2447 |
-
},
|
| 2448 |
-
{
|
| 2449 |
-
"epoch": 37.967032967032964,
|
| 2450 |
-
"grad_norm": 73.53226470947266,
|
| 2451 |
-
"learning_rate": 1.308e-05,
|
| 2452 |
-
"loss": 17.2669,
|
| 2453 |
-
"step": 1745
|
| 2454 |
-
},
|
| 2455 |
-
{
|
| 2456 |
-
"epoch": 38.065934065934066,
|
| 2457 |
-
"grad_norm": 58.128150939941406,
|
| 2458 |
"learning_rate": 1.3117500000000001e-05,
|
| 2459 |
-
"loss": 18.
|
| 2460 |
"step": 1750
|
| 2461 |
},
|
| 2462 |
{
|
| 2463 |
-
"epoch":
|
| 2464 |
-
"grad_norm":
|
| 2465 |
-
"learning_rate": 1.3155e-05,
|
| 2466 |
-
"loss": 17.0461,
|
| 2467 |
-
"step": 1755
|
| 2468 |
-
},
|
| 2469 |
-
{
|
| 2470 |
-
"epoch": 38.285714285714285,
|
| 2471 |
-
"grad_norm": 64.85994720458984,
|
| 2472 |
-
"learning_rate": 1.31925e-05,
|
| 2473 |
-
"loss": 16.7421,
|
| 2474 |
-
"step": 1760
|
| 2475 |
-
},
|
| 2476 |
-
{
|
| 2477 |
-
"epoch": 38.395604395604394,
|
| 2478 |
-
"grad_norm": 65.9932861328125,
|
| 2479 |
-
"learning_rate": 1.323e-05,
|
| 2480 |
-
"loss": 16.7541,
|
| 2481 |
-
"step": 1765
|
| 2482 |
-
},
|
| 2483 |
-
{
|
| 2484 |
-
"epoch": 38.505494505494504,
|
| 2485 |
-
"grad_norm": 60.61685562133789,
|
| 2486 |
-
"learning_rate": 1.32675e-05,
|
| 2487 |
-
"loss": 17.6106,
|
| 2488 |
-
"step": 1770
|
| 2489 |
-
},
|
| 2490 |
-
{
|
| 2491 |
-
"epoch": 38.61538461538461,
|
| 2492 |
-
"grad_norm": 67.54483032226562,
|
| 2493 |
-
"learning_rate": 1.3305e-05,
|
| 2494 |
-
"loss": 16.9768,
|
| 2495 |
-
"step": 1775
|
| 2496 |
-
},
|
| 2497 |
-
{
|
| 2498 |
-
"epoch": 38.72527472527472,
|
| 2499 |
-
"grad_norm": 74.80374908447266,
|
| 2500 |
-
"learning_rate": 1.33425e-05,
|
| 2501 |
-
"loss": 16.6762,
|
| 2502 |
-
"step": 1780
|
| 2503 |
-
},
|
| 2504 |
-
{
|
| 2505 |
-
"epoch": 38.83516483516483,
|
| 2506 |
-
"grad_norm": 50.045692443847656,
|
| 2507 |
-
"learning_rate": 1.338e-05,
|
| 2508 |
-
"loss": 16.3018,
|
| 2509 |
-
"step": 1785
|
| 2510 |
-
},
|
| 2511 |
-
{
|
| 2512 |
-
"epoch": 38.94505494505494,
|
| 2513 |
-
"grad_norm": 75.77816009521484,
|
| 2514 |
-
"learning_rate": 1.34175e-05,
|
| 2515 |
-
"loss": 17.1515,
|
| 2516 |
-
"step": 1790
|
| 2517 |
-
},
|
| 2518 |
-
{
|
| 2519 |
-
"epoch": 39.043956043956044,
|
| 2520 |
-
"grad_norm": 48.32164764404297,
|
| 2521 |
-
"learning_rate": 1.3455e-05,
|
| 2522 |
-
"loss": 15.5191,
|
| 2523 |
-
"step": 1795
|
| 2524 |
-
},
|
| 2525 |
-
{
|
| 2526 |
-
"epoch": 39.15384615384615,
|
| 2527 |
-
"grad_norm": 69.80590057373047,
|
| 2528 |
"learning_rate": 1.34925e-05,
|
| 2529 |
-
"loss":
|
| 2530 |
"step": 1800
|
| 2531 |
},
|
| 2532 |
{
|
| 2533 |
-
"epoch":
|
| 2534 |
-
"grad_norm":
|
| 2535 |
-
"learning_rate": 1.3530000000000001e-05,
|
| 2536 |
-
"loss": 16.3254,
|
| 2537 |
-
"step": 1805
|
| 2538 |
-
},
|
| 2539 |
-
{
|
| 2540 |
-
"epoch": 39.37362637362637,
|
| 2541 |
-
"grad_norm": 52.71421432495117,
|
| 2542 |
-
"learning_rate": 1.35675e-05,
|
| 2543 |
-
"loss": 16.1323,
|
| 2544 |
-
"step": 1810
|
| 2545 |
-
},
|
| 2546 |
-
{
|
| 2547 |
-
"epoch": 39.48351648351648,
|
| 2548 |
-
"grad_norm": 68.22982025146484,
|
| 2549 |
-
"learning_rate": 1.3605000000000001e-05,
|
| 2550 |
-
"loss": 16.3548,
|
| 2551 |
-
"step": 1815
|
| 2552 |
-
},
|
| 2553 |
-
{
|
| 2554 |
-
"epoch": 39.59340659340659,
|
| 2555 |
-
"grad_norm": 82.97122192382812,
|
| 2556 |
-
"learning_rate": 1.36425e-05,
|
| 2557 |
-
"loss": 16.154,
|
| 2558 |
-
"step": 1820
|
| 2559 |
-
},
|
| 2560 |
-
{
|
| 2561 |
-
"epoch": 39.7032967032967,
|
| 2562 |
-
"grad_norm": 78.51126861572266,
|
| 2563 |
-
"learning_rate": 1.3680000000000001e-05,
|
| 2564 |
-
"loss": 16.1856,
|
| 2565 |
-
"step": 1825
|
| 2566 |
-
},
|
| 2567 |
-
{
|
| 2568 |
-
"epoch": 39.81318681318681,
|
| 2569 |
-
"grad_norm": 84.22386932373047,
|
| 2570 |
-
"learning_rate": 1.37175e-05,
|
| 2571 |
-
"loss": 16.1977,
|
| 2572 |
-
"step": 1830
|
| 2573 |
-
},
|
| 2574 |
-
{
|
| 2575 |
-
"epoch": 39.92307692307692,
|
| 2576 |
-
"grad_norm": 72.61677551269531,
|
| 2577 |
-
"learning_rate": 1.3755000000000001e-05,
|
| 2578 |
-
"loss": 15.5414,
|
| 2579 |
-
"step": 1835
|
| 2580 |
-
},
|
| 2581 |
-
{
|
| 2582 |
-
"epoch": 40.02197802197802,
|
| 2583 |
-
"grad_norm": 60.4188117980957,
|
| 2584 |
-
"learning_rate": 1.37925e-05,
|
| 2585 |
-
"loss": 17.9055,
|
| 2586 |
-
"step": 1840
|
| 2587 |
-
},
|
| 2588 |
-
{
|
| 2589 |
-
"epoch": 40.13186813186813,
|
| 2590 |
-
"grad_norm": 55.08159637451172,
|
| 2591 |
-
"learning_rate": 1.3830000000000001e-05,
|
| 2592 |
-
"loss": 15.7148,
|
| 2593 |
-
"step": 1845
|
| 2594 |
-
},
|
| 2595 |
-
{
|
| 2596 |
-
"epoch": 40.24175824175824,
|
| 2597 |
-
"grad_norm": 58.98347091674805,
|
| 2598 |
"learning_rate": 1.38675e-05,
|
| 2599 |
-
"loss": 16.
|
| 2600 |
"step": 1850
|
| 2601 |
},
|
| 2602 |
{
|
| 2603 |
-
"epoch":
|
| 2604 |
-
"grad_norm":
|
| 2605 |
-
"learning_rate": 1.3905000000000002e-05,
|
| 2606 |
-
"loss": 15.9047,
|
| 2607 |
-
"step": 1855
|
| 2608 |
-
},
|
| 2609 |
-
{
|
| 2610 |
-
"epoch": 40.46153846153846,
|
| 2611 |
-
"grad_norm": 64.99933624267578,
|
| 2612 |
-
"learning_rate": 1.39425e-05,
|
| 2613 |
-
"loss": 15.9342,
|
| 2614 |
-
"step": 1860
|
| 2615 |
-
},
|
| 2616 |
-
{
|
| 2617 |
-
"epoch": 40.57142857142857,
|
| 2618 |
-
"grad_norm": 63.66476058959961,
|
| 2619 |
-
"learning_rate": 1.3980000000000002e-05,
|
| 2620 |
-
"loss": 15.6676,
|
| 2621 |
-
"step": 1865
|
| 2622 |
-
},
|
| 2623 |
-
{
|
| 2624 |
-
"epoch": 40.68131868131868,
|
| 2625 |
-
"grad_norm": 62.42832565307617,
|
| 2626 |
-
"learning_rate": 1.4017500000000001e-05,
|
| 2627 |
-
"loss": 16.0192,
|
| 2628 |
-
"step": 1870
|
| 2629 |
-
},
|
| 2630 |
-
{
|
| 2631 |
-
"epoch": 40.79120879120879,
|
| 2632 |
-
"grad_norm": 64.26334381103516,
|
| 2633 |
-
"learning_rate": 1.4055000000000002e-05,
|
| 2634 |
-
"loss": 15.4637,
|
| 2635 |
-
"step": 1875
|
| 2636 |
-
},
|
| 2637 |
-
{
|
| 2638 |
-
"epoch": 40.9010989010989,
|
| 2639 |
-
"grad_norm": 53.68131637573242,
|
| 2640 |
-
"learning_rate": 1.4092500000000001e-05,
|
| 2641 |
-
"loss": 15.4847,
|
| 2642 |
-
"step": 1880
|
| 2643 |
-
},
|
| 2644 |
-
{
|
| 2645 |
-
"epoch": 41.0,
|
| 2646 |
-
"grad_norm": 133.40896606445312,
|
| 2647 |
-
"learning_rate": 1.413e-05,
|
| 2648 |
-
"loss": 15.1,
|
| 2649 |
-
"step": 1885
|
| 2650 |
-
},
|
| 2651 |
-
{
|
| 2652 |
-
"epoch": 41.10989010989011,
|
| 2653 |
-
"grad_norm": 64.47783660888672,
|
| 2654 |
-
"learning_rate": 1.4167500000000001e-05,
|
| 2655 |
-
"loss": 15.8261,
|
| 2656 |
-
"step": 1890
|
| 2657 |
-
},
|
| 2658 |
-
{
|
| 2659 |
-
"epoch": 41.21978021978022,
|
| 2660 |
-
"grad_norm": 53.95432662963867,
|
| 2661 |
-
"learning_rate": 1.4205e-05,
|
| 2662 |
-
"loss": 15.3971,
|
| 2663 |
-
"step": 1895
|
| 2664 |
-
},
|
| 2665 |
-
{
|
| 2666 |
-
"epoch": 41.32967032967033,
|
| 2667 |
-
"grad_norm": 60.42184829711914,
|
| 2668 |
"learning_rate": 1.4242500000000001e-05,
|
| 2669 |
-
"loss":
|
| 2670 |
"step": 1900
|
| 2671 |
},
|
| 2672 |
{
|
| 2673 |
-
"epoch":
|
| 2674 |
-
"grad_norm":
|
| 2675 |
-
"learning_rate": 1.428e-05,
|
| 2676 |
-
"loss": 16.5622,
|
| 2677 |
-
"step": 1905
|
| 2678 |
-
},
|
| 2679 |
-
{
|
| 2680 |
-
"epoch": 41.54945054945055,
|
| 2681 |
-
"grad_norm": 71.76943969726562,
|
| 2682 |
-
"learning_rate": 1.4317500000000001e-05,
|
| 2683 |
-
"loss": 15.9214,
|
| 2684 |
-
"step": 1910
|
| 2685 |
-
},
|
| 2686 |
-
{
|
| 2687 |
-
"epoch": 41.65934065934066,
|
| 2688 |
-
"grad_norm": 56.51178741455078,
|
| 2689 |
-
"learning_rate": 1.4355e-05,
|
| 2690 |
-
"loss": 15.283,
|
| 2691 |
-
"step": 1915
|
| 2692 |
-
},
|
| 2693 |
-
{
|
| 2694 |
-
"epoch": 41.76923076923077,
|
| 2695 |
-
"grad_norm": 77.75347900390625,
|
| 2696 |
-
"learning_rate": 1.43925e-05,
|
| 2697 |
-
"loss": 15.6863,
|
| 2698 |
-
"step": 1920
|
| 2699 |
-
},
|
| 2700 |
-
{
|
| 2701 |
-
"epoch": 41.879120879120876,
|
| 2702 |
-
"grad_norm": 69.0566635131836,
|
| 2703 |
-
"learning_rate": 1.4429999999999999e-05,
|
| 2704 |
-
"loss": 14.7517,
|
| 2705 |
-
"step": 1925
|
| 2706 |
-
},
|
| 2707 |
-
{
|
| 2708 |
-
"epoch": 41.98901098901099,
|
| 2709 |
-
"grad_norm": 52.313724517822266,
|
| 2710 |
-
"learning_rate": 1.44675e-05,
|
| 2711 |
-
"loss": 14.5892,
|
| 2712 |
-
"step": 1930
|
| 2713 |
-
},
|
| 2714 |
-
{
|
| 2715 |
-
"epoch": 42.08791208791209,
|
| 2716 |
-
"grad_norm": 46.05543518066406,
|
| 2717 |
-
"learning_rate": 1.4505e-05,
|
| 2718 |
-
"loss": 17.2186,
|
| 2719 |
-
"step": 1935
|
| 2720 |
-
},
|
| 2721 |
-
{
|
| 2722 |
-
"epoch": 42.1978021978022,
|
| 2723 |
-
"grad_norm": 60.843135833740234,
|
| 2724 |
-
"learning_rate": 1.45425e-05,
|
| 2725 |
-
"loss": 15.052,
|
| 2726 |
-
"step": 1940
|
| 2727 |
-
},
|
| 2728 |
-
{
|
| 2729 |
-
"epoch": 42.30769230769231,
|
| 2730 |
-
"grad_norm": 69.2349853515625,
|
| 2731 |
-
"learning_rate": 1.458e-05,
|
| 2732 |
-
"loss": 14.8454,
|
| 2733 |
-
"step": 1945
|
| 2734 |
-
},
|
| 2735 |
-
{
|
| 2736 |
-
"epoch": 42.417582417582416,
|
| 2737 |
-
"grad_norm": 63.5760383605957,
|
| 2738 |
"learning_rate": 1.46175e-05,
|
| 2739 |
-
"loss": 15.
|
| 2740 |
"step": 1950
|
| 2741 |
},
|
| 2742 |
{
|
| 2743 |
-
"epoch":
|
| 2744 |
-
"grad_norm":
|
| 2745 |
-
"learning_rate": 1.4655e-05,
|
| 2746 |
-
"loss": 15.7076,
|
| 2747 |
-
"step": 1955
|
| 2748 |
-
},
|
| 2749 |
-
{
|
| 2750 |
-
"epoch": 42.637362637362635,
|
| 2751 |
-
"grad_norm": 67.10089111328125,
|
| 2752 |
-
"learning_rate": 1.46925e-05,
|
| 2753 |
-
"loss": 14.6074,
|
| 2754 |
-
"step": 1960
|
| 2755 |
-
},
|
| 2756 |
-
{
|
| 2757 |
-
"epoch": 42.747252747252745,
|
| 2758 |
-
"grad_norm": 56.15253448486328,
|
| 2759 |
-
"learning_rate": 1.473e-05,
|
| 2760 |
-
"loss": 14.7393,
|
| 2761 |
-
"step": 1965
|
| 2762 |
-
},
|
| 2763 |
-
{
|
| 2764 |
-
"epoch": 42.857142857142854,
|
| 2765 |
-
"grad_norm": 62.48304748535156,
|
| 2766 |
-
"learning_rate": 1.47675e-05,
|
| 2767 |
-
"loss": 15.0895,
|
| 2768 |
-
"step": 1970
|
| 2769 |
-
},
|
| 2770 |
-
{
|
| 2771 |
-
"epoch": 42.967032967032964,
|
| 2772 |
-
"grad_norm": 64.34083557128906,
|
| 2773 |
-
"learning_rate": 1.4805e-05,
|
| 2774 |
-
"loss": 14.4191,
|
| 2775 |
-
"step": 1975
|
| 2776 |
-
},
|
| 2777 |
-
{
|
| 2778 |
-
"epoch": 43.065934065934066,
|
| 2779 |
-
"grad_norm": 72.964111328125,
|
| 2780 |
-
"learning_rate": 1.48425e-05,
|
| 2781 |
-
"loss": 15.1835,
|
| 2782 |
-
"step": 1980
|
| 2783 |
-
},
|
| 2784 |
-
{
|
| 2785 |
-
"epoch": 43.175824175824175,
|
| 2786 |
-
"grad_norm": 64.65727233886719,
|
| 2787 |
-
"learning_rate": 1.488e-05,
|
| 2788 |
-
"loss": 14.8041,
|
| 2789 |
-
"step": 1985
|
| 2790 |
-
},
|
| 2791 |
-
{
|
| 2792 |
-
"epoch": 43.285714285714285,
|
| 2793 |
-
"grad_norm": 58.626461029052734,
|
| 2794 |
-
"learning_rate": 1.49175e-05,
|
| 2795 |
-
"loss": 15.048,
|
| 2796 |
-
"step": 1990
|
| 2797 |
-
},
|
| 2798 |
-
{
|
| 2799 |
-
"epoch": 43.395604395604394,
|
| 2800 |
-
"grad_norm": 66.43981170654297,
|
| 2801 |
-
"learning_rate": 1.4955e-05,
|
| 2802 |
-
"loss": 14.752,
|
| 2803 |
-
"step": 1995
|
| 2804 |
-
},
|
| 2805 |
-
{
|
| 2806 |
-
"epoch": 43.505494505494504,
|
| 2807 |
-
"grad_norm": 62.531463623046875,
|
| 2808 |
"learning_rate": 1.4992500000000001e-05,
|
| 2809 |
-
"loss":
|
| 2810 |
"step": 2000
|
| 2811 |
}
|
| 2812 |
],
|
| 2813 |
-
"logging_steps":
|
| 2814 |
-
"max_steps":
|
| 2815 |
"num_input_tokens_seen": 0,
|
| 2816 |
"num_train_epochs": 50,
|
| 2817 |
-
"save_steps":
|
| 2818 |
"stateful_callbacks": {
|
| 2819 |
"TrainerControl": {
|
| 2820 |
"args": {
|
|
@@ -2827,7 +307,7 @@
|
|
| 2827 |
"attributes": {}
|
| 2828 |
}
|
| 2829 |
},
|
| 2830 |
-
"total_flos": 5.
|
| 2831 |
"train_batch_size": 24,
|
| 2832 |
"trial_name": null,
|
| 2833 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 44.44444444444444,
|
| 6 |
"eval_steps": 500,
|
| 7 |
"global_step": 2000,
|
| 8 |
"is_hyper_param_search": false,
|
|
|
|
| 10 |
"is_world_process_zero": true,
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
+
"epoch": 1.1111111111111112,
|
| 14 |
+
"grad_norm": 3663.33642578125,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 15 |
"learning_rate": 3.6750000000000003e-07,
|
| 16 |
+
"loss": 1356.1239,
|
| 17 |
"step": 50
|
| 18 |
},
|
| 19 |
{
|
| 20 |
+
"epoch": 2.2222222222222223,
|
| 21 |
+
"grad_norm": 3973.032470703125,
|
| 22 |
+
"learning_rate": 7.425000000000001e-07,
|
| 23 |
+
"loss": 1275.5178,
|
| 24 |
+
"step": 100
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 25 |
},
|
| 26 |
{
|
| 27 |
+
"epoch": 3.3333333333333335,
|
| 28 |
+
"grad_norm": 2667.0068359375,
|
| 29 |
+
"learning_rate": 1.1174999999999999e-06,
|
| 30 |
+
"loss": 1123.6059,
|
| 31 |
+
"step": 150
|
| 32 |
},
|
| 33 |
{
|
| 34 |
+
"epoch": 4.444444444444445,
|
| 35 |
+
"grad_norm": 2320.48486328125,
|
| 36 |
+
"learning_rate": 1.4925000000000001e-06,
|
| 37 |
+
"loss": 922.8917,
|
| 38 |
+
"step": 200
|
| 39 |
},
|
| 40 |
{
|
| 41 |
+
"epoch": 5.555555555555555,
|
| 42 |
+
"grad_norm": 1611.3345947265625,
|
| 43 |
+
"learning_rate": 1.8675000000000001e-06,
|
| 44 |
+
"loss": 714.5638,
|
| 45 |
+
"step": 250
|
| 46 |
},
|
| 47 |
{
|
| 48 |
+
"epoch": 6.666666666666667,
|
| 49 |
+
"grad_norm": 1395.7064208984375,
|
| 50 |
+
"learning_rate": 2.2425e-06,
|
| 51 |
+
"loss": 542.1251,
|
| 52 |
+
"step": 300
|
| 53 |
},
|
| 54 |
{
|
| 55 |
+
"epoch": 7.777777777777778,
|
| 56 |
+
"grad_norm": 1019.4860229492188,
|
| 57 |
+
"learning_rate": 2.6175e-06,
|
| 58 |
+
"loss": 411.1387,
|
| 59 |
+
"step": 350
|
| 60 |
},
|
| 61 |
{
|
| 62 |
+
"epoch": 8.88888888888889,
|
| 63 |
+
"grad_norm": 888.315185546875,
|
| 64 |
+
"learning_rate": 2.9925e-06,
|
| 65 |
+
"loss": 318.4567,
|
| 66 |
+
"step": 400
|
| 67 |
},
|
| 68 |
{
|
| 69 |
+
"epoch": 10.0,
|
| 70 |
+
"grad_norm": 2590.113525390625,
|
| 71 |
+
"learning_rate": 3.3675000000000004e-06,
|
| 72 |
+
"loss": 261.6995,
|
| 73 |
+
"step": 450
|
| 74 |
},
|
| 75 |
{
|
| 76 |
+
"epoch": 11.11111111111111,
|
| 77 |
+
"grad_norm": 711.677734375,
|
| 78 |
+
"learning_rate": 3.7425e-06,
|
| 79 |
+
"loss": 220.2936,
|
| 80 |
+
"step": 500
|
| 81 |
},
|
| 82 |
{
|
| 83 |
+
"epoch": 12.222222222222221,
|
| 84 |
+
"grad_norm": 548.9371948242188,
|
| 85 |
+
"learning_rate": 4.117500000000001e-06,
|
| 86 |
+
"loss": 187.9833,
|
| 87 |
+
"step": 550
|
| 88 |
},
|
| 89 |
{
|
| 90 |
+
"epoch": 13.333333333333334,
|
| 91 |
+
"grad_norm": 1127.1611328125,
|
| 92 |
+
"learning_rate": 4.4925e-06,
|
| 93 |
+
"loss": 159.1351,
|
| 94 |
+
"step": 600
|
| 95 |
},
|
| 96 |
{
|
| 97 |
+
"epoch": 14.444444444444445,
|
| 98 |
+
"grad_norm": 426.074951171875,
|
| 99 |
+
"learning_rate": 4.8675e-06,
|
| 100 |
+
"loss": 137.1092,
|
| 101 |
+
"step": 650
|
| 102 |
},
|
| 103 |
{
|
| 104 |
+
"epoch": 15.555555555555555,
|
| 105 |
+
"grad_norm": 348.842529296875,
|
| 106 |
+
"learning_rate": 5.2425e-06,
|
| 107 |
+
"loss": 119.822,
|
| 108 |
+
"step": 700
|
| 109 |
},
|
| 110 |
{
|
| 111 |
+
"epoch": 16.666666666666668,
|
| 112 |
+
"grad_norm": 373.2169189453125,
|
| 113 |
+
"learning_rate": 5.6175e-06,
|
| 114 |
+
"loss": 104.3366,
|
| 115 |
+
"step": 750
|
| 116 |
},
|
| 117 |
{
|
| 118 |
+
"epoch": 17.77777777777778,
|
| 119 |
+
"grad_norm": 324.51702880859375,
|
| 120 |
+
"learning_rate": 5.992500000000001e-06,
|
| 121 |
+
"loss": 90.8788,
|
| 122 |
+
"step": 800
|
| 123 |
},
|
| 124 |
{
|
| 125 |
+
"epoch": 18.88888888888889,
|
| 126 |
+
"grad_norm": 269.91827392578125,
|
| 127 |
+
"learning_rate": 6.3675e-06,
|
| 128 |
+
"loss": 78.4644,
|
| 129 |
+
"step": 850
|
| 130 |
},
|
| 131 |
{
|
| 132 |
+
"epoch": 20.0,
|
| 133 |
+
"grad_norm": 1744.54052734375,
|
| 134 |
+
"learning_rate": 6.7425e-06,
|
| 135 |
+
"loss": 70.3526,
|
| 136 |
+
"step": 900
|
| 137 |
},
|
| 138 |
{
|
| 139 |
+
"epoch": 21.11111111111111,
|
| 140 |
+
"grad_norm": 369.39837646484375,
|
| 141 |
+
"learning_rate": 7.1175e-06,
|
| 142 |
+
"loss": 63.9417,
|
| 143 |
+
"step": 950
|
| 144 |
},
|
| 145 |
{
|
| 146 |
+
"epoch": 22.22222222222222,
|
| 147 |
+
"grad_norm": 303.95977783203125,
|
| 148 |
+
"learning_rate": 7.4925e-06,
|
| 149 |
+
"loss": 63.4575,
|
| 150 |
+
"step": 1000
|
| 151 |
},
|
| 152 |
{
|
| 153 |
+
"epoch": 23.333333333333332,
|
| 154 |
+
"grad_norm": 187.462890625,
|
| 155 |
+
"learning_rate": 7.8675e-06,
|
| 156 |
+
"loss": 54.7417,
|
| 157 |
+
"step": 1050
|
| 158 |
},
|
| 159 |
{
|
| 160 |
+
"epoch": 24.444444444444443,
|
| 161 |
+
"grad_norm": 165.56666564941406,
|
| 162 |
+
"learning_rate": 8.2425e-06,
|
| 163 |
+
"loss": 49.6842,
|
| 164 |
+
"step": 1100
|
| 165 |
},
|
| 166 |
{
|
| 167 |
+
"epoch": 25.555555555555557,
|
| 168 |
+
"grad_norm": 147.3148193359375,
|
| 169 |
+
"learning_rate": 8.6175e-06,
|
| 170 |
+
"loss": 43.027,
|
| 171 |
+
"step": 1150
|
| 172 |
},
|
| 173 |
{
|
| 174 |
+
"epoch": 26.666666666666668,
|
| 175 |
+
"grad_norm": 125.53775024414062,
|
| 176 |
+
"learning_rate": 8.9925e-06,
|
| 177 |
+
"loss": 38.2579,
|
| 178 |
+
"step": 1200
|
| 179 |
},
|
| 180 |
{
|
| 181 |
+
"epoch": 27.77777777777778,
|
| 182 |
+
"grad_norm": 109.85810089111328,
|
| 183 |
+
"learning_rate": 9.367500000000001e-06,
|
| 184 |
+
"loss": 34.3957,
|
| 185 |
+
"step": 1250
|
| 186 |
},
|
| 187 |
{
|
| 188 |
+
"epoch": 28.88888888888889,
|
| 189 |
+
"grad_norm": 98.328369140625,
|
| 190 |
+
"learning_rate": 9.7425e-06,
|
| 191 |
+
"loss": 31.4378,
|
| 192 |
+
"step": 1300
|
| 193 |
},
|
| 194 |
{
|
| 195 |
+
"epoch": 30.0,
|
| 196 |
+
"grad_norm": 109.77750396728516,
|
| 197 |
+
"learning_rate": 1.01175e-05,
|
| 198 |
+
"loss": 28.5084,
|
| 199 |
+
"step": 1350
|
| 200 |
},
|
| 201 |
{
|
| 202 |
+
"epoch": 31.11111111111111,
|
| 203 |
+
"grad_norm": 112.47483825683594,
|
| 204 |
+
"learning_rate": 1.04925e-05,
|
| 205 |
+
"loss": 26.1671,
|
| 206 |
+
"step": 1400
|
| 207 |
},
|
| 208 |
{
|
| 209 |
+
"epoch": 32.22222222222222,
|
| 210 |
+
"grad_norm": 85.60242462158203,
|
| 211 |
"learning_rate": 1.08675e-05,
|
| 212 |
+
"loss": 24.2309,
|
| 213 |
"step": 1450
|
| 214 |
},
|
| 215 |
{
|
| 216 |
+
"epoch": 33.333333333333336,
|
| 217 |
+
"grad_norm": 73.19799041748047,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 218 |
"learning_rate": 1.1242500000000001e-05,
|
| 219 |
+
"loss": 22.6248,
|
| 220 |
"step": 1500
|
| 221 |
},
|
| 222 |
{
|
| 223 |
+
"epoch": 34.44444444444444,
|
| 224 |
+
"grad_norm": 80.70884704589844,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 225 |
"learning_rate": 1.16175e-05,
|
| 226 |
+
"loss": 21.1187,
|
| 227 |
"step": 1550
|
| 228 |
},
|
| 229 |
{
|
| 230 |
+
"epoch": 35.55555555555556,
|
| 231 |
+
"grad_norm": 110.98326110839844,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 232 |
"learning_rate": 1.19925e-05,
|
| 233 |
+
"loss": 20.4828,
|
| 234 |
"step": 1600
|
| 235 |
},
|
| 236 |
{
|
| 237 |
+
"epoch": 36.666666666666664,
|
| 238 |
+
"grad_norm": 113.65286254882812,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 239 |
"learning_rate": 1.23675e-05,
|
| 240 |
+
"loss": 19.7366,
|
| 241 |
"step": 1650
|
| 242 |
},
|
| 243 |
{
|
| 244 |
+
"epoch": 37.77777777777778,
|
| 245 |
+
"grad_norm": 77.65855407714844,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 246 |
"learning_rate": 1.27425e-05,
|
| 247 |
+
"loss": 18.6632,
|
| 248 |
"step": 1700
|
| 249 |
},
|
| 250 |
{
|
| 251 |
+
"epoch": 38.888888888888886,
|
| 252 |
+
"grad_norm": 88.96723175048828,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 253 |
"learning_rate": 1.3117500000000001e-05,
|
| 254 |
+
"loss": 18.0793,
|
| 255 |
"step": 1750
|
| 256 |
},
|
| 257 |
{
|
| 258 |
+
"epoch": 40.0,
|
| 259 |
+
"grad_norm": 79.1690902709961,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 260 |
"learning_rate": 1.34925e-05,
|
| 261 |
+
"loss": 17.0667,
|
| 262 |
"step": 1800
|
| 263 |
},
|
| 264 |
{
|
| 265 |
+
"epoch": 41.111111111111114,
|
| 266 |
+
"grad_norm": 93.60108184814453,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 267 |
"learning_rate": 1.38675e-05,
|
| 268 |
+
"loss": 16.6106,
|
| 269 |
"step": 1850
|
| 270 |
},
|
| 271 |
{
|
| 272 |
+
"epoch": 42.22222222222222,
|
| 273 |
+
"grad_norm": 66.16129302978516,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 274 |
"learning_rate": 1.4242500000000001e-05,
|
| 275 |
+
"loss": 16.4905,
|
| 276 |
"step": 1900
|
| 277 |
},
|
| 278 |
{
|
| 279 |
+
"epoch": 43.333333333333336,
|
| 280 |
+
"grad_norm": 62.41362380981445,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 281 |
"learning_rate": 1.46175e-05,
|
| 282 |
+
"loss": 15.6427,
|
| 283 |
"step": 1950
|
| 284 |
},
|
| 285 |
{
|
| 286 |
+
"epoch": 44.44444444444444,
|
| 287 |
+
"grad_norm": 107.30168151855469,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 288 |
"learning_rate": 1.4992500000000001e-05,
|
| 289 |
+
"loss": 15.0731,
|
| 290 |
"step": 2000
|
| 291 |
}
|
| 292 |
],
|
| 293 |
+
"logging_steps": 50,
|
| 294 |
+
"max_steps": 2250,
|
| 295 |
"num_input_tokens_seen": 0,
|
| 296 |
"num_train_epochs": 50,
|
| 297 |
+
"save_steps": 200,
|
| 298 |
"stateful_callbacks": {
|
| 299 |
"TrainerControl": {
|
| 300 |
"args": {
|
|
|
|
| 307 |
"attributes": {}
|
| 308 |
}
|
| 309 |
},
|
| 310 |
+
"total_flos": 5.387467985017897e+19,
|
| 311 |
"train_batch_size": 24,
|
| 312 |
"trial_name": null,
|
| 313 |
"trial_params": null
|
checkpoint-2000/training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5368
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a7b84367094b7487f77de50fba614a6c6667e9cf018b77ee5bfc158268fc5eaf
|
| 3 |
size 5368
|
checkpoint-2200/config.json
ADDED
|
@@ -0,0 +1,129 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"activation_dropout": 0.0,
|
| 3 |
+
"activation_function": "silu",
|
| 4 |
+
"anchor_image_size": null,
|
| 5 |
+
"architectures": [
|
| 6 |
+
"RTDetrV2ForObjectDetection"
|
| 7 |
+
],
|
| 8 |
+
"attention_dropout": 0.0,
|
| 9 |
+
"auxiliary_loss": true,
|
| 10 |
+
"backbone": null,
|
| 11 |
+
"backbone_config": {
|
| 12 |
+
"depths": [
|
| 13 |
+
3,
|
| 14 |
+
4,
|
| 15 |
+
23,
|
| 16 |
+
3
|
| 17 |
+
],
|
| 18 |
+
"downsample_in_bottleneck": false,
|
| 19 |
+
"downsample_in_first_stage": false,
|
| 20 |
+
"embedding_size": 64,
|
| 21 |
+
"hidden_act": "relu",
|
| 22 |
+
"hidden_sizes": [
|
| 23 |
+
256,
|
| 24 |
+
512,
|
| 25 |
+
1024,
|
| 26 |
+
2048
|
| 27 |
+
],
|
| 28 |
+
"layer_type": "bottleneck",
|
| 29 |
+
"model_type": "rt_detr_resnet",
|
| 30 |
+
"num_channels": 3,
|
| 31 |
+
"out_features": [
|
| 32 |
+
"stage2",
|
| 33 |
+
"stage3",
|
| 34 |
+
"stage4"
|
| 35 |
+
],
|
| 36 |
+
"out_indices": [
|
| 37 |
+
2,
|
| 38 |
+
3,
|
| 39 |
+
4
|
| 40 |
+
],
|
| 41 |
+
"stage_names": [
|
| 42 |
+
"stem",
|
| 43 |
+
"stage1",
|
| 44 |
+
"stage2",
|
| 45 |
+
"stage3",
|
| 46 |
+
"stage4"
|
| 47 |
+
],
|
| 48 |
+
"torch_dtype": "float32"
|
| 49 |
+
},
|
| 50 |
+
"backbone_kwargs": null,
|
| 51 |
+
"batch_norm_eps": 1e-05,
|
| 52 |
+
"box_noise_scale": 1.0,
|
| 53 |
+
"d_model": 256,
|
| 54 |
+
"decoder_activation_function": "relu",
|
| 55 |
+
"decoder_attention_heads": 8,
|
| 56 |
+
"decoder_ffn_dim": 1024,
|
| 57 |
+
"decoder_in_channels": [
|
| 58 |
+
384,
|
| 59 |
+
384,
|
| 60 |
+
384
|
| 61 |
+
],
|
| 62 |
+
"decoder_layers": 6,
|
| 63 |
+
"decoder_method": "default",
|
| 64 |
+
"decoder_n_levels": 3,
|
| 65 |
+
"decoder_n_points": 4,
|
| 66 |
+
"decoder_offset_scale": 0.5,
|
| 67 |
+
"disable_custom_kernels": true,
|
| 68 |
+
"dropout": 0.0,
|
| 69 |
+
"encode_proj_layers": [
|
| 70 |
+
2
|
| 71 |
+
],
|
| 72 |
+
"encoder_activation_function": "gelu",
|
| 73 |
+
"encoder_attention_heads": 8,
|
| 74 |
+
"encoder_ffn_dim": 2048,
|
| 75 |
+
"encoder_hidden_dim": 384,
|
| 76 |
+
"encoder_in_channels": [
|
| 77 |
+
512,
|
| 78 |
+
1024,
|
| 79 |
+
2048
|
| 80 |
+
],
|
| 81 |
+
"encoder_layers": 1,
|
| 82 |
+
"eos_coefficient": 0.0001,
|
| 83 |
+
"eval_size": null,
|
| 84 |
+
"feat_strides": [
|
| 85 |
+
8,
|
| 86 |
+
16,
|
| 87 |
+
32
|
| 88 |
+
],
|
| 89 |
+
"focal_loss_alpha": 0.75,
|
| 90 |
+
"focal_loss_gamma": 2.0,
|
| 91 |
+
"freeze_backbone_batch_norms": true,
|
| 92 |
+
"hidden_expansion": 1.0,
|
| 93 |
+
"id2label": {
|
| 94 |
+
"0": "LABEL_0",
|
| 95 |
+
"1": "LABEL_1",
|
| 96 |
+
"2": "LABEL_2"
|
| 97 |
+
},
|
| 98 |
+
"initializer_bias_prior_prob": null,
|
| 99 |
+
"initializer_range": 0.01,
|
| 100 |
+
"is_encoder_decoder": true,
|
| 101 |
+
"label2id": {
|
| 102 |
+
"LABEL_0": 0,
|
| 103 |
+
"LABEL_1": 1,
|
| 104 |
+
"LABEL_2": 2
|
| 105 |
+
},
|
| 106 |
+
"label_noise_ratio": 0.5,
|
| 107 |
+
"layer_norm_eps": 1e-05,
|
| 108 |
+
"learn_initial_query": false,
|
| 109 |
+
"matcher_alpha": 0.25,
|
| 110 |
+
"matcher_bbox_cost": 5.0,
|
| 111 |
+
"matcher_class_cost": 2.0,
|
| 112 |
+
"matcher_gamma": 2.0,
|
| 113 |
+
"matcher_giou_cost": 2.0,
|
| 114 |
+
"model_type": "rt_detr_v2",
|
| 115 |
+
"normalize_before": false,
|
| 116 |
+
"num_denoising": 100,
|
| 117 |
+
"num_feature_levels": 3,
|
| 118 |
+
"num_queries": 300,
|
| 119 |
+
"positional_encoding_temperature": 10000,
|
| 120 |
+
"torch_dtype": "float32",
|
| 121 |
+
"transformers_version": "4.55.0",
|
| 122 |
+
"use_focal_loss": true,
|
| 123 |
+
"use_pretrained_backbone": false,
|
| 124 |
+
"use_timm_backbone": false,
|
| 125 |
+
"weight_loss_bbox": 5.0,
|
| 126 |
+
"weight_loss_giou": 2.0,
|
| 127 |
+
"weight_loss_vfl": 1.0,
|
| 128 |
+
"with_box_refine": true
|
| 129 |
+
}
|
checkpoint-2200/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2cb217c7b7445eb9473cddef9f4ea779004ca224f4615b6f69aa70c7cc0b781f
|
| 3 |
+
size 306699044
|
checkpoint-2200/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8ca9f9b9c554485c8d65a25449ddeb6582be923a8692ab873d9c1e6a21062298
|
| 3 |
+
size 611580433
|
checkpoint-2200/preprocessor_config.json
ADDED
|
@@ -0,0 +1,26 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"do_convert_annotations": true,
|
| 3 |
+
"do_normalize": false,
|
| 4 |
+
"do_pad": false,
|
| 5 |
+
"do_rescale": true,
|
| 6 |
+
"do_resize": true,
|
| 7 |
+
"format": "coco_detection",
|
| 8 |
+
"image_mean": [
|
| 9 |
+
0.485,
|
| 10 |
+
0.456,
|
| 11 |
+
0.406
|
| 12 |
+
],
|
| 13 |
+
"image_processor_type": "RTDetrImageProcessor",
|
| 14 |
+
"image_std": [
|
| 15 |
+
0.229,
|
| 16 |
+
0.224,
|
| 17 |
+
0.225
|
| 18 |
+
],
|
| 19 |
+
"pad_size": null,
|
| 20 |
+
"resample": 2,
|
| 21 |
+
"rescale_factor": 0.00392156862745098,
|
| 22 |
+
"size": {
|
| 23 |
+
"height": 640,
|
| 24 |
+
"width": 640
|
| 25 |
+
}
|
| 26 |
+
}
|
checkpoint-2200/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:90a75f76146ccf9511e324998dc7ec6df7764e7adf2655174e5f732a90e23392
|
| 3 |
+
size 14244
|
checkpoint-2200/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4be819b970a626d9fdb18122878b575ce46f2fa37fc360325a23e4cfdc87bcd1
|
| 3 |
+
size 1064
|
checkpoint-2200/trainer_state.json
ADDED
|
@@ -0,0 +1,342 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 48.888888888888886,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 2200,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 1.1111111111111112,
|
| 14 |
+
"grad_norm": 3663.33642578125,
|
| 15 |
+
"learning_rate": 3.6750000000000003e-07,
|
| 16 |
+
"loss": 1356.1239,
|
| 17 |
+
"step": 50
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 2.2222222222222223,
|
| 21 |
+
"grad_norm": 3973.032470703125,
|
| 22 |
+
"learning_rate": 7.425000000000001e-07,
|
| 23 |
+
"loss": 1275.5178,
|
| 24 |
+
"step": 100
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 3.3333333333333335,
|
| 28 |
+
"grad_norm": 2667.0068359375,
|
| 29 |
+
"learning_rate": 1.1174999999999999e-06,
|
| 30 |
+
"loss": 1123.6059,
|
| 31 |
+
"step": 150
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 4.444444444444445,
|
| 35 |
+
"grad_norm": 2320.48486328125,
|
| 36 |
+
"learning_rate": 1.4925000000000001e-06,
|
| 37 |
+
"loss": 922.8917,
|
| 38 |
+
"step": 200
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 5.555555555555555,
|
| 42 |
+
"grad_norm": 1611.3345947265625,
|
| 43 |
+
"learning_rate": 1.8675000000000001e-06,
|
| 44 |
+
"loss": 714.5638,
|
| 45 |
+
"step": 250
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 6.666666666666667,
|
| 49 |
+
"grad_norm": 1395.7064208984375,
|
| 50 |
+
"learning_rate": 2.2425e-06,
|
| 51 |
+
"loss": 542.1251,
|
| 52 |
+
"step": 300
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 7.777777777777778,
|
| 56 |
+
"grad_norm": 1019.4860229492188,
|
| 57 |
+
"learning_rate": 2.6175e-06,
|
| 58 |
+
"loss": 411.1387,
|
| 59 |
+
"step": 350
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 8.88888888888889,
|
| 63 |
+
"grad_norm": 888.315185546875,
|
| 64 |
+
"learning_rate": 2.9925e-06,
|
| 65 |
+
"loss": 318.4567,
|
| 66 |
+
"step": 400
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 10.0,
|
| 70 |
+
"grad_norm": 2590.113525390625,
|
| 71 |
+
"learning_rate": 3.3675000000000004e-06,
|
| 72 |
+
"loss": 261.6995,
|
| 73 |
+
"step": 450
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 11.11111111111111,
|
| 77 |
+
"grad_norm": 711.677734375,
|
| 78 |
+
"learning_rate": 3.7425e-06,
|
| 79 |
+
"loss": 220.2936,
|
| 80 |
+
"step": 500
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 12.222222222222221,
|
| 84 |
+
"grad_norm": 548.9371948242188,
|
| 85 |
+
"learning_rate": 4.117500000000001e-06,
|
| 86 |
+
"loss": 187.9833,
|
| 87 |
+
"step": 550
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 13.333333333333334,
|
| 91 |
+
"grad_norm": 1127.1611328125,
|
| 92 |
+
"learning_rate": 4.4925e-06,
|
| 93 |
+
"loss": 159.1351,
|
| 94 |
+
"step": 600
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 14.444444444444445,
|
| 98 |
+
"grad_norm": 426.074951171875,
|
| 99 |
+
"learning_rate": 4.8675e-06,
|
| 100 |
+
"loss": 137.1092,
|
| 101 |
+
"step": 650
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 15.555555555555555,
|
| 105 |
+
"grad_norm": 348.842529296875,
|
| 106 |
+
"learning_rate": 5.2425e-06,
|
| 107 |
+
"loss": 119.822,
|
| 108 |
+
"step": 700
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 16.666666666666668,
|
| 112 |
+
"grad_norm": 373.2169189453125,
|
| 113 |
+
"learning_rate": 5.6175e-06,
|
| 114 |
+
"loss": 104.3366,
|
| 115 |
+
"step": 750
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 17.77777777777778,
|
| 119 |
+
"grad_norm": 324.51702880859375,
|
| 120 |
+
"learning_rate": 5.992500000000001e-06,
|
| 121 |
+
"loss": 90.8788,
|
| 122 |
+
"step": 800
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 18.88888888888889,
|
| 126 |
+
"grad_norm": 269.91827392578125,
|
| 127 |
+
"learning_rate": 6.3675e-06,
|
| 128 |
+
"loss": 78.4644,
|
| 129 |
+
"step": 850
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 20.0,
|
| 133 |
+
"grad_norm": 1744.54052734375,
|
| 134 |
+
"learning_rate": 6.7425e-06,
|
| 135 |
+
"loss": 70.3526,
|
| 136 |
+
"step": 900
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 21.11111111111111,
|
| 140 |
+
"grad_norm": 369.39837646484375,
|
| 141 |
+
"learning_rate": 7.1175e-06,
|
| 142 |
+
"loss": 63.9417,
|
| 143 |
+
"step": 950
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 22.22222222222222,
|
| 147 |
+
"grad_norm": 303.95977783203125,
|
| 148 |
+
"learning_rate": 7.4925e-06,
|
| 149 |
+
"loss": 63.4575,
|
| 150 |
+
"step": 1000
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 23.333333333333332,
|
| 154 |
+
"grad_norm": 187.462890625,
|
| 155 |
+
"learning_rate": 7.8675e-06,
|
| 156 |
+
"loss": 54.7417,
|
| 157 |
+
"step": 1050
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 24.444444444444443,
|
| 161 |
+
"grad_norm": 165.56666564941406,
|
| 162 |
+
"learning_rate": 8.2425e-06,
|
| 163 |
+
"loss": 49.6842,
|
| 164 |
+
"step": 1100
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 25.555555555555557,
|
| 168 |
+
"grad_norm": 147.3148193359375,
|
| 169 |
+
"learning_rate": 8.6175e-06,
|
| 170 |
+
"loss": 43.027,
|
| 171 |
+
"step": 1150
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 26.666666666666668,
|
| 175 |
+
"grad_norm": 125.53775024414062,
|
| 176 |
+
"learning_rate": 8.9925e-06,
|
| 177 |
+
"loss": 38.2579,
|
| 178 |
+
"step": 1200
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 27.77777777777778,
|
| 182 |
+
"grad_norm": 109.85810089111328,
|
| 183 |
+
"learning_rate": 9.367500000000001e-06,
|
| 184 |
+
"loss": 34.3957,
|
| 185 |
+
"step": 1250
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 28.88888888888889,
|
| 189 |
+
"grad_norm": 98.328369140625,
|
| 190 |
+
"learning_rate": 9.7425e-06,
|
| 191 |
+
"loss": 31.4378,
|
| 192 |
+
"step": 1300
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 30.0,
|
| 196 |
+
"grad_norm": 109.77750396728516,
|
| 197 |
+
"learning_rate": 1.01175e-05,
|
| 198 |
+
"loss": 28.5084,
|
| 199 |
+
"step": 1350
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 31.11111111111111,
|
| 203 |
+
"grad_norm": 112.47483825683594,
|
| 204 |
+
"learning_rate": 1.04925e-05,
|
| 205 |
+
"loss": 26.1671,
|
| 206 |
+
"step": 1400
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 32.22222222222222,
|
| 210 |
+
"grad_norm": 85.60242462158203,
|
| 211 |
+
"learning_rate": 1.08675e-05,
|
| 212 |
+
"loss": 24.2309,
|
| 213 |
+
"step": 1450
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 33.333333333333336,
|
| 217 |
+
"grad_norm": 73.19799041748047,
|
| 218 |
+
"learning_rate": 1.1242500000000001e-05,
|
| 219 |
+
"loss": 22.6248,
|
| 220 |
+
"step": 1500
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 34.44444444444444,
|
| 224 |
+
"grad_norm": 80.70884704589844,
|
| 225 |
+
"learning_rate": 1.16175e-05,
|
| 226 |
+
"loss": 21.1187,
|
| 227 |
+
"step": 1550
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 35.55555555555556,
|
| 231 |
+
"grad_norm": 110.98326110839844,
|
| 232 |
+
"learning_rate": 1.19925e-05,
|
| 233 |
+
"loss": 20.4828,
|
| 234 |
+
"step": 1600
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 36.666666666666664,
|
| 238 |
+
"grad_norm": 113.65286254882812,
|
| 239 |
+
"learning_rate": 1.23675e-05,
|
| 240 |
+
"loss": 19.7366,
|
| 241 |
+
"step": 1650
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 37.77777777777778,
|
| 245 |
+
"grad_norm": 77.65855407714844,
|
| 246 |
+
"learning_rate": 1.27425e-05,
|
| 247 |
+
"loss": 18.6632,
|
| 248 |
+
"step": 1700
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 38.888888888888886,
|
| 252 |
+
"grad_norm": 88.96723175048828,
|
| 253 |
+
"learning_rate": 1.3117500000000001e-05,
|
| 254 |
+
"loss": 18.0793,
|
| 255 |
+
"step": 1750
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 40.0,
|
| 259 |
+
"grad_norm": 79.1690902709961,
|
| 260 |
+
"learning_rate": 1.34925e-05,
|
| 261 |
+
"loss": 17.0667,
|
| 262 |
+
"step": 1800
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 41.111111111111114,
|
| 266 |
+
"grad_norm": 93.60108184814453,
|
| 267 |
+
"learning_rate": 1.38675e-05,
|
| 268 |
+
"loss": 16.6106,
|
| 269 |
+
"step": 1850
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 42.22222222222222,
|
| 273 |
+
"grad_norm": 66.16129302978516,
|
| 274 |
+
"learning_rate": 1.4242500000000001e-05,
|
| 275 |
+
"loss": 16.4905,
|
| 276 |
+
"step": 1900
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 43.333333333333336,
|
| 280 |
+
"grad_norm": 62.41362380981445,
|
| 281 |
+
"learning_rate": 1.46175e-05,
|
| 282 |
+
"loss": 15.6427,
|
| 283 |
+
"step": 1950
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 44.44444444444444,
|
| 287 |
+
"grad_norm": 107.30168151855469,
|
| 288 |
+
"learning_rate": 1.4992500000000001e-05,
|
| 289 |
+
"loss": 15.0731,
|
| 290 |
+
"step": 2000
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 45.55555555555556,
|
| 294 |
+
"grad_norm": 100.18666076660156,
|
| 295 |
+
"learning_rate": 1.2060000000000001e-05,
|
| 296 |
+
"loss": 14.6973,
|
| 297 |
+
"step": 2050
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 46.666666666666664,
|
| 301 |
+
"grad_norm": 68.78209686279297,
|
| 302 |
+
"learning_rate": 9.06e-06,
|
| 303 |
+
"loss": 13.8928,
|
| 304 |
+
"step": 2100
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 47.77777777777778,
|
| 308 |
+
"grad_norm": 65.85958099365234,
|
| 309 |
+
"learning_rate": 6.0600000000000004e-06,
|
| 310 |
+
"loss": 13.4318,
|
| 311 |
+
"step": 2150
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 48.888888888888886,
|
| 315 |
+
"grad_norm": 72.31432342529297,
|
| 316 |
+
"learning_rate": 3.06e-06,
|
| 317 |
+
"loss": 12.6856,
|
| 318 |
+
"step": 2200
|
| 319 |
+
}
|
| 320 |
+
],
|
| 321 |
+
"logging_steps": 50,
|
| 322 |
+
"max_steps": 2250,
|
| 323 |
+
"num_input_tokens_seen": 0,
|
| 324 |
+
"num_train_epochs": 50,
|
| 325 |
+
"save_steps": 200,
|
| 326 |
+
"stateful_callbacks": {
|
| 327 |
+
"TrainerControl": {
|
| 328 |
+
"args": {
|
| 329 |
+
"should_epoch_stop": false,
|
| 330 |
+
"should_evaluate": false,
|
| 331 |
+
"should_log": false,
|
| 332 |
+
"should_save": true,
|
| 333 |
+
"should_training_stop": false
|
| 334 |
+
},
|
| 335 |
+
"attributes": {}
|
| 336 |
+
}
|
| 337 |
+
},
|
| 338 |
+
"total_flos": 5.926395026677432e+19,
|
| 339 |
+
"train_batch_size": 24,
|
| 340 |
+
"trial_name": null,
|
| 341 |
+
"trial_params": null
|
| 342 |
+
}
|
checkpoint-2200/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a7b84367094b7487f77de50fba614a6c6667e9cf018b77ee5bfc158268fc5eaf
|
| 3 |
+
size 5368
|
checkpoint-2250/config.json
ADDED
|
@@ -0,0 +1,129 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"activation_dropout": 0.0,
|
| 3 |
+
"activation_function": "silu",
|
| 4 |
+
"anchor_image_size": null,
|
| 5 |
+
"architectures": [
|
| 6 |
+
"RTDetrV2ForObjectDetection"
|
| 7 |
+
],
|
| 8 |
+
"attention_dropout": 0.0,
|
| 9 |
+
"auxiliary_loss": true,
|
| 10 |
+
"backbone": null,
|
| 11 |
+
"backbone_config": {
|
| 12 |
+
"depths": [
|
| 13 |
+
3,
|
| 14 |
+
4,
|
| 15 |
+
23,
|
| 16 |
+
3
|
| 17 |
+
],
|
| 18 |
+
"downsample_in_bottleneck": false,
|
| 19 |
+
"downsample_in_first_stage": false,
|
| 20 |
+
"embedding_size": 64,
|
| 21 |
+
"hidden_act": "relu",
|
| 22 |
+
"hidden_sizes": [
|
| 23 |
+
256,
|
| 24 |
+
512,
|
| 25 |
+
1024,
|
| 26 |
+
2048
|
| 27 |
+
],
|
| 28 |
+
"layer_type": "bottleneck",
|
| 29 |
+
"model_type": "rt_detr_resnet",
|
| 30 |
+
"num_channels": 3,
|
| 31 |
+
"out_features": [
|
| 32 |
+
"stage2",
|
| 33 |
+
"stage3",
|
| 34 |
+
"stage4"
|
| 35 |
+
],
|
| 36 |
+
"out_indices": [
|
| 37 |
+
2,
|
| 38 |
+
3,
|
| 39 |
+
4
|
| 40 |
+
],
|
| 41 |
+
"stage_names": [
|
| 42 |
+
"stem",
|
| 43 |
+
"stage1",
|
| 44 |
+
"stage2",
|
| 45 |
+
"stage3",
|
| 46 |
+
"stage4"
|
| 47 |
+
],
|
| 48 |
+
"torch_dtype": "float32"
|
| 49 |
+
},
|
| 50 |
+
"backbone_kwargs": null,
|
| 51 |
+
"batch_norm_eps": 1e-05,
|
| 52 |
+
"box_noise_scale": 1.0,
|
| 53 |
+
"d_model": 256,
|
| 54 |
+
"decoder_activation_function": "relu",
|
| 55 |
+
"decoder_attention_heads": 8,
|
| 56 |
+
"decoder_ffn_dim": 1024,
|
| 57 |
+
"decoder_in_channels": [
|
| 58 |
+
384,
|
| 59 |
+
384,
|
| 60 |
+
384
|
| 61 |
+
],
|
| 62 |
+
"decoder_layers": 6,
|
| 63 |
+
"decoder_method": "default",
|
| 64 |
+
"decoder_n_levels": 3,
|
| 65 |
+
"decoder_n_points": 4,
|
| 66 |
+
"decoder_offset_scale": 0.5,
|
| 67 |
+
"disable_custom_kernels": true,
|
| 68 |
+
"dropout": 0.0,
|
| 69 |
+
"encode_proj_layers": [
|
| 70 |
+
2
|
| 71 |
+
],
|
| 72 |
+
"encoder_activation_function": "gelu",
|
| 73 |
+
"encoder_attention_heads": 8,
|
| 74 |
+
"encoder_ffn_dim": 2048,
|
| 75 |
+
"encoder_hidden_dim": 384,
|
| 76 |
+
"encoder_in_channels": [
|
| 77 |
+
512,
|
| 78 |
+
1024,
|
| 79 |
+
2048
|
| 80 |
+
],
|
| 81 |
+
"encoder_layers": 1,
|
| 82 |
+
"eos_coefficient": 0.0001,
|
| 83 |
+
"eval_size": null,
|
| 84 |
+
"feat_strides": [
|
| 85 |
+
8,
|
| 86 |
+
16,
|
| 87 |
+
32
|
| 88 |
+
],
|
| 89 |
+
"focal_loss_alpha": 0.75,
|
| 90 |
+
"focal_loss_gamma": 2.0,
|
| 91 |
+
"freeze_backbone_batch_norms": true,
|
| 92 |
+
"hidden_expansion": 1.0,
|
| 93 |
+
"id2label": {
|
| 94 |
+
"0": "LABEL_0",
|
| 95 |
+
"1": "LABEL_1",
|
| 96 |
+
"2": "LABEL_2"
|
| 97 |
+
},
|
| 98 |
+
"initializer_bias_prior_prob": null,
|
| 99 |
+
"initializer_range": 0.01,
|
| 100 |
+
"is_encoder_decoder": true,
|
| 101 |
+
"label2id": {
|
| 102 |
+
"LABEL_0": 0,
|
| 103 |
+
"LABEL_1": 1,
|
| 104 |
+
"LABEL_2": 2
|
| 105 |
+
},
|
| 106 |
+
"label_noise_ratio": 0.5,
|
| 107 |
+
"layer_norm_eps": 1e-05,
|
| 108 |
+
"learn_initial_query": false,
|
| 109 |
+
"matcher_alpha": 0.25,
|
| 110 |
+
"matcher_bbox_cost": 5.0,
|
| 111 |
+
"matcher_class_cost": 2.0,
|
| 112 |
+
"matcher_gamma": 2.0,
|
| 113 |
+
"matcher_giou_cost": 2.0,
|
| 114 |
+
"model_type": "rt_detr_v2",
|
| 115 |
+
"normalize_before": false,
|
| 116 |
+
"num_denoising": 100,
|
| 117 |
+
"num_feature_levels": 3,
|
| 118 |
+
"num_queries": 300,
|
| 119 |
+
"positional_encoding_temperature": 10000,
|
| 120 |
+
"torch_dtype": "float32",
|
| 121 |
+
"transformers_version": "4.55.0",
|
| 122 |
+
"use_focal_loss": true,
|
| 123 |
+
"use_pretrained_backbone": false,
|
| 124 |
+
"use_timm_backbone": false,
|
| 125 |
+
"weight_loss_bbox": 5.0,
|
| 126 |
+
"weight_loss_giou": 2.0,
|
| 127 |
+
"weight_loss_vfl": 1.0,
|
| 128 |
+
"with_box_refine": true
|
| 129 |
+
}
|
checkpoint-2250/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d3a52d4cd4386295eedbfb267bc679eca4b27864d745fff06694c0f9dbf823a6
|
| 3 |
+
size 306699044
|
checkpoint-2250/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:05dd6b927d3bf0bbfe4b8e26ceb5b08121784a39348599920d07011f715be702
|
| 3 |
+
size 611580433
|
checkpoint-2250/preprocessor_config.json
ADDED
|
@@ -0,0 +1,26 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"do_convert_annotations": true,
|
| 3 |
+
"do_normalize": false,
|
| 4 |
+
"do_pad": false,
|
| 5 |
+
"do_rescale": true,
|
| 6 |
+
"do_resize": true,
|
| 7 |
+
"format": "coco_detection",
|
| 8 |
+
"image_mean": [
|
| 9 |
+
0.485,
|
| 10 |
+
0.456,
|
| 11 |
+
0.406
|
| 12 |
+
],
|
| 13 |
+
"image_processor_type": "RTDetrImageProcessor",
|
| 14 |
+
"image_std": [
|
| 15 |
+
0.229,
|
| 16 |
+
0.224,
|
| 17 |
+
0.225
|
| 18 |
+
],
|
| 19 |
+
"pad_size": null,
|
| 20 |
+
"resample": 2,
|
| 21 |
+
"rescale_factor": 0.00392156862745098,
|
| 22 |
+
"size": {
|
| 23 |
+
"height": 640,
|
| 24 |
+
"width": 640
|
| 25 |
+
}
|
| 26 |
+
}
|
checkpoint-2250/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:66787fd9597d0c4424fa82a41803d924585f86f960101c754c7f5cfada26d864
|
| 3 |
+
size 14244
|
checkpoint-2250/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e97b09861aa673666d96648b6cbe41d7c63c06fd70e7ca49c43cf61264897a95
|
| 3 |
+
size 1064
|
checkpoint-2250/trainer_state.json
ADDED
|
@@ -0,0 +1,349 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 50.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 2250,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 1.1111111111111112,
|
| 14 |
+
"grad_norm": 3663.33642578125,
|
| 15 |
+
"learning_rate": 3.6750000000000003e-07,
|
| 16 |
+
"loss": 1356.1239,
|
| 17 |
+
"step": 50
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 2.2222222222222223,
|
| 21 |
+
"grad_norm": 3973.032470703125,
|
| 22 |
+
"learning_rate": 7.425000000000001e-07,
|
| 23 |
+
"loss": 1275.5178,
|
| 24 |
+
"step": 100
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 3.3333333333333335,
|
| 28 |
+
"grad_norm": 2667.0068359375,
|
| 29 |
+
"learning_rate": 1.1174999999999999e-06,
|
| 30 |
+
"loss": 1123.6059,
|
| 31 |
+
"step": 150
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 4.444444444444445,
|
| 35 |
+
"grad_norm": 2320.48486328125,
|
| 36 |
+
"learning_rate": 1.4925000000000001e-06,
|
| 37 |
+
"loss": 922.8917,
|
| 38 |
+
"step": 200
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 5.555555555555555,
|
| 42 |
+
"grad_norm": 1611.3345947265625,
|
| 43 |
+
"learning_rate": 1.8675000000000001e-06,
|
| 44 |
+
"loss": 714.5638,
|
| 45 |
+
"step": 250
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 6.666666666666667,
|
| 49 |
+
"grad_norm": 1395.7064208984375,
|
| 50 |
+
"learning_rate": 2.2425e-06,
|
| 51 |
+
"loss": 542.1251,
|
| 52 |
+
"step": 300
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 7.777777777777778,
|
| 56 |
+
"grad_norm": 1019.4860229492188,
|
| 57 |
+
"learning_rate": 2.6175e-06,
|
| 58 |
+
"loss": 411.1387,
|
| 59 |
+
"step": 350
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 8.88888888888889,
|
| 63 |
+
"grad_norm": 888.315185546875,
|
| 64 |
+
"learning_rate": 2.9925e-06,
|
| 65 |
+
"loss": 318.4567,
|
| 66 |
+
"step": 400
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 10.0,
|
| 70 |
+
"grad_norm": 2590.113525390625,
|
| 71 |
+
"learning_rate": 3.3675000000000004e-06,
|
| 72 |
+
"loss": 261.6995,
|
| 73 |
+
"step": 450
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 11.11111111111111,
|
| 77 |
+
"grad_norm": 711.677734375,
|
| 78 |
+
"learning_rate": 3.7425e-06,
|
| 79 |
+
"loss": 220.2936,
|
| 80 |
+
"step": 500
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 12.222222222222221,
|
| 84 |
+
"grad_norm": 548.9371948242188,
|
| 85 |
+
"learning_rate": 4.117500000000001e-06,
|
| 86 |
+
"loss": 187.9833,
|
| 87 |
+
"step": 550
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 13.333333333333334,
|
| 91 |
+
"grad_norm": 1127.1611328125,
|
| 92 |
+
"learning_rate": 4.4925e-06,
|
| 93 |
+
"loss": 159.1351,
|
| 94 |
+
"step": 600
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 14.444444444444445,
|
| 98 |
+
"grad_norm": 426.074951171875,
|
| 99 |
+
"learning_rate": 4.8675e-06,
|
| 100 |
+
"loss": 137.1092,
|
| 101 |
+
"step": 650
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 15.555555555555555,
|
| 105 |
+
"grad_norm": 348.842529296875,
|
| 106 |
+
"learning_rate": 5.2425e-06,
|
| 107 |
+
"loss": 119.822,
|
| 108 |
+
"step": 700
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 16.666666666666668,
|
| 112 |
+
"grad_norm": 373.2169189453125,
|
| 113 |
+
"learning_rate": 5.6175e-06,
|
| 114 |
+
"loss": 104.3366,
|
| 115 |
+
"step": 750
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 17.77777777777778,
|
| 119 |
+
"grad_norm": 324.51702880859375,
|
| 120 |
+
"learning_rate": 5.992500000000001e-06,
|
| 121 |
+
"loss": 90.8788,
|
| 122 |
+
"step": 800
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 18.88888888888889,
|
| 126 |
+
"grad_norm": 269.91827392578125,
|
| 127 |
+
"learning_rate": 6.3675e-06,
|
| 128 |
+
"loss": 78.4644,
|
| 129 |
+
"step": 850
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 20.0,
|
| 133 |
+
"grad_norm": 1744.54052734375,
|
| 134 |
+
"learning_rate": 6.7425e-06,
|
| 135 |
+
"loss": 70.3526,
|
| 136 |
+
"step": 900
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 21.11111111111111,
|
| 140 |
+
"grad_norm": 369.39837646484375,
|
| 141 |
+
"learning_rate": 7.1175e-06,
|
| 142 |
+
"loss": 63.9417,
|
| 143 |
+
"step": 950
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 22.22222222222222,
|
| 147 |
+
"grad_norm": 303.95977783203125,
|
| 148 |
+
"learning_rate": 7.4925e-06,
|
| 149 |
+
"loss": 63.4575,
|
| 150 |
+
"step": 1000
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 23.333333333333332,
|
| 154 |
+
"grad_norm": 187.462890625,
|
| 155 |
+
"learning_rate": 7.8675e-06,
|
| 156 |
+
"loss": 54.7417,
|
| 157 |
+
"step": 1050
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 24.444444444444443,
|
| 161 |
+
"grad_norm": 165.56666564941406,
|
| 162 |
+
"learning_rate": 8.2425e-06,
|
| 163 |
+
"loss": 49.6842,
|
| 164 |
+
"step": 1100
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 25.555555555555557,
|
| 168 |
+
"grad_norm": 147.3148193359375,
|
| 169 |
+
"learning_rate": 8.6175e-06,
|
| 170 |
+
"loss": 43.027,
|
| 171 |
+
"step": 1150
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 26.666666666666668,
|
| 175 |
+
"grad_norm": 125.53775024414062,
|
| 176 |
+
"learning_rate": 8.9925e-06,
|
| 177 |
+
"loss": 38.2579,
|
| 178 |
+
"step": 1200
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 27.77777777777778,
|
| 182 |
+
"grad_norm": 109.85810089111328,
|
| 183 |
+
"learning_rate": 9.367500000000001e-06,
|
| 184 |
+
"loss": 34.3957,
|
| 185 |
+
"step": 1250
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 28.88888888888889,
|
| 189 |
+
"grad_norm": 98.328369140625,
|
| 190 |
+
"learning_rate": 9.7425e-06,
|
| 191 |
+
"loss": 31.4378,
|
| 192 |
+
"step": 1300
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 30.0,
|
| 196 |
+
"grad_norm": 109.77750396728516,
|
| 197 |
+
"learning_rate": 1.01175e-05,
|
| 198 |
+
"loss": 28.5084,
|
| 199 |
+
"step": 1350
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 31.11111111111111,
|
| 203 |
+
"grad_norm": 112.47483825683594,
|
| 204 |
+
"learning_rate": 1.04925e-05,
|
| 205 |
+
"loss": 26.1671,
|
| 206 |
+
"step": 1400
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 32.22222222222222,
|
| 210 |
+
"grad_norm": 85.60242462158203,
|
| 211 |
+
"learning_rate": 1.08675e-05,
|
| 212 |
+
"loss": 24.2309,
|
| 213 |
+
"step": 1450
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 33.333333333333336,
|
| 217 |
+
"grad_norm": 73.19799041748047,
|
| 218 |
+
"learning_rate": 1.1242500000000001e-05,
|
| 219 |
+
"loss": 22.6248,
|
| 220 |
+
"step": 1500
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 34.44444444444444,
|
| 224 |
+
"grad_norm": 80.70884704589844,
|
| 225 |
+
"learning_rate": 1.16175e-05,
|
| 226 |
+
"loss": 21.1187,
|
| 227 |
+
"step": 1550
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 35.55555555555556,
|
| 231 |
+
"grad_norm": 110.98326110839844,
|
| 232 |
+
"learning_rate": 1.19925e-05,
|
| 233 |
+
"loss": 20.4828,
|
| 234 |
+
"step": 1600
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 36.666666666666664,
|
| 238 |
+
"grad_norm": 113.65286254882812,
|
| 239 |
+
"learning_rate": 1.23675e-05,
|
| 240 |
+
"loss": 19.7366,
|
| 241 |
+
"step": 1650
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 37.77777777777778,
|
| 245 |
+
"grad_norm": 77.65855407714844,
|
| 246 |
+
"learning_rate": 1.27425e-05,
|
| 247 |
+
"loss": 18.6632,
|
| 248 |
+
"step": 1700
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 38.888888888888886,
|
| 252 |
+
"grad_norm": 88.96723175048828,
|
| 253 |
+
"learning_rate": 1.3117500000000001e-05,
|
| 254 |
+
"loss": 18.0793,
|
| 255 |
+
"step": 1750
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 40.0,
|
| 259 |
+
"grad_norm": 79.1690902709961,
|
| 260 |
+
"learning_rate": 1.34925e-05,
|
| 261 |
+
"loss": 17.0667,
|
| 262 |
+
"step": 1800
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 41.111111111111114,
|
| 266 |
+
"grad_norm": 93.60108184814453,
|
| 267 |
+
"learning_rate": 1.38675e-05,
|
| 268 |
+
"loss": 16.6106,
|
| 269 |
+
"step": 1850
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 42.22222222222222,
|
| 273 |
+
"grad_norm": 66.16129302978516,
|
| 274 |
+
"learning_rate": 1.4242500000000001e-05,
|
| 275 |
+
"loss": 16.4905,
|
| 276 |
+
"step": 1900
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 43.333333333333336,
|
| 280 |
+
"grad_norm": 62.41362380981445,
|
| 281 |
+
"learning_rate": 1.46175e-05,
|
| 282 |
+
"loss": 15.6427,
|
| 283 |
+
"step": 1950
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 44.44444444444444,
|
| 287 |
+
"grad_norm": 107.30168151855469,
|
| 288 |
+
"learning_rate": 1.4992500000000001e-05,
|
| 289 |
+
"loss": 15.0731,
|
| 290 |
+
"step": 2000
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 45.55555555555556,
|
| 294 |
+
"grad_norm": 100.18666076660156,
|
| 295 |
+
"learning_rate": 1.2060000000000001e-05,
|
| 296 |
+
"loss": 14.6973,
|
| 297 |
+
"step": 2050
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 46.666666666666664,
|
| 301 |
+
"grad_norm": 68.78209686279297,
|
| 302 |
+
"learning_rate": 9.06e-06,
|
| 303 |
+
"loss": 13.8928,
|
| 304 |
+
"step": 2100
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 47.77777777777778,
|
| 308 |
+
"grad_norm": 65.85958099365234,
|
| 309 |
+
"learning_rate": 6.0600000000000004e-06,
|
| 310 |
+
"loss": 13.4318,
|
| 311 |
+
"step": 2150
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 48.888888888888886,
|
| 315 |
+
"grad_norm": 72.31432342529297,
|
| 316 |
+
"learning_rate": 3.06e-06,
|
| 317 |
+
"loss": 12.6856,
|
| 318 |
+
"step": 2200
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 50.0,
|
| 322 |
+
"grad_norm": 50.271549224853516,
|
| 323 |
+
"learning_rate": 6.000000000000001e-08,
|
| 324 |
+
"loss": 12.0923,
|
| 325 |
+
"step": 2250
|
| 326 |
+
}
|
| 327 |
+
],
|
| 328 |
+
"logging_steps": 50,
|
| 329 |
+
"max_steps": 2250,
|
| 330 |
+
"num_input_tokens_seen": 0,
|
| 331 |
+
"num_train_epochs": 50,
|
| 332 |
+
"save_steps": 200,
|
| 333 |
+
"stateful_callbacks": {
|
| 334 |
+
"TrainerControl": {
|
| 335 |
+
"args": {
|
| 336 |
+
"should_epoch_stop": false,
|
| 337 |
+
"should_evaluate": false,
|
| 338 |
+
"should_log": false,
|
| 339 |
+
"should_save": true,
|
| 340 |
+
"should_training_stop": true
|
| 341 |
+
},
|
| 342 |
+
"attributes": {}
|
| 343 |
+
}
|
| 344 |
+
},
|
| 345 |
+
"total_flos": 6.060676179197952e+19,
|
| 346 |
+
"train_batch_size": 24,
|
| 347 |
+
"trial_name": null,
|
| 348 |
+
"trial_params": null
|
| 349 |
+
}
|
checkpoint-2250/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a7b84367094b7487f77de50fba614a6c6667e9cf018b77ee5bfc158268fc5eaf
|
| 3 |
+
size 5368
|
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 306699044
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d3a52d4cd4386295eedbfb267bc679eca4b27864d745fff06694c0f9dbf823a6
|
| 3 |
size 306699044
|
runs/Aug14_17-42-57_2676026c4495/events.out.tfevents.1755193378.2676026c4495.6591.0
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6333d2a8e567a7e0ba6820d48e67bd616d919e804c37bbbe27e4f49f18a3f884
|
| 3 |
+
size 16890
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5368
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a7b84367094b7487f77de50fba614a6c6667e9cf018b77ee5bfc158268fc5eaf
|
| 3 |
size 5368
|