Training in progress, step 19500, checkpoint
Browse files- checkpoint-19500/config.json +62 -0
- checkpoint-19500/model.safetensors +3 -0
- checkpoint-19500/optimizer.pt +3 -0
- checkpoint-19500/preprocessor_config.json +26 -0
- checkpoint-19500/rng_state.pth +3 -0
- checkpoint-19500/scheduler.pt +3 -0
- checkpoint-19500/trainer_state.json +1398 -0
- checkpoint-19500/training_args.bin +3 -0
checkpoint-19500/config.json
ADDED
|
@@ -0,0 +1,62 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "facebook/detr-resnet-50-dc5",
|
| 3 |
+
"activation_dropout": 0.0,
|
| 4 |
+
"activation_function": "relu",
|
| 5 |
+
"architectures": [
|
| 6 |
+
"DetrForObjectDetection"
|
| 7 |
+
],
|
| 8 |
+
"attention_dropout": 0.0,
|
| 9 |
+
"auxiliary_loss": false,
|
| 10 |
+
"backbone": "resnet50",
|
| 11 |
+
"backbone_config": null,
|
| 12 |
+
"backbone_kwargs": {
|
| 13 |
+
"in_chans": 3,
|
| 14 |
+
"out_indices": [
|
| 15 |
+
1,
|
| 16 |
+
2,
|
| 17 |
+
3,
|
| 18 |
+
4
|
| 19 |
+
],
|
| 20 |
+
"output_stride": 16
|
| 21 |
+
},
|
| 22 |
+
"bbox_cost": 5,
|
| 23 |
+
"bbox_loss_coefficient": 5,
|
| 24 |
+
"class_cost": 1,
|
| 25 |
+
"classifier_dropout": 0.0,
|
| 26 |
+
"d_model": 256,
|
| 27 |
+
"decoder_attention_heads": 8,
|
| 28 |
+
"decoder_ffn_dim": 2048,
|
| 29 |
+
"decoder_layerdrop": 0.0,
|
| 30 |
+
"decoder_layers": 6,
|
| 31 |
+
"dice_loss_coefficient": 1,
|
| 32 |
+
"dilation": true,
|
| 33 |
+
"dropout": 0.1,
|
| 34 |
+
"encoder_attention_heads": 8,
|
| 35 |
+
"encoder_ffn_dim": 2048,
|
| 36 |
+
"encoder_layerdrop": 0.0,
|
| 37 |
+
"encoder_layers": 6,
|
| 38 |
+
"eos_coefficient": 0.1,
|
| 39 |
+
"giou_cost": 2,
|
| 40 |
+
"giou_loss_coefficient": 2,
|
| 41 |
+
"id2label": {
|
| 42 |
+
"0": "segment"
|
| 43 |
+
},
|
| 44 |
+
"init_std": 0.02,
|
| 45 |
+
"init_xavier_std": 1.0,
|
| 46 |
+
"is_encoder_decoder": true,
|
| 47 |
+
"label2id": {
|
| 48 |
+
"segment": 0
|
| 49 |
+
},
|
| 50 |
+
"mask_loss_coefficient": 1,
|
| 51 |
+
"max_position_embeddings": 1024,
|
| 52 |
+
"model_type": "detr",
|
| 53 |
+
"num_channels": 3,
|
| 54 |
+
"num_hidden_layers": 6,
|
| 55 |
+
"num_queries": 500,
|
| 56 |
+
"position_embedding_type": "sine",
|
| 57 |
+
"scale_embedding": false,
|
| 58 |
+
"torch_dtype": "float32",
|
| 59 |
+
"transformers_version": "4.44.2",
|
| 60 |
+
"use_pretrained_backbone": true,
|
| 61 |
+
"use_timm_backbone": true
|
| 62 |
+
}
|
checkpoint-19500/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:60cbd0d7fdaf966273c2e59cb420c8d192922405f48902894abd8c7bbb97fa08
|
| 3 |
+
size 166904424
|
checkpoint-19500/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:323cdbd8d9b8b34c1f91040df016debeb6745a74d53f15bb9d813c470cc07527
|
| 3 |
+
size 331310970
|
checkpoint-19500/preprocessor_config.json
ADDED
|
@@ -0,0 +1,26 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"do_convert_annotations": true,
|
| 3 |
+
"do_normalize": true,
|
| 4 |
+
"do_pad": true,
|
| 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": "DetrImageProcessor",
|
| 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 |
+
"longest_edge": 1333,
|
| 24 |
+
"shortest_edge": 800
|
| 25 |
+
}
|
| 26 |
+
}
|
checkpoint-19500/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ad6a6a3550f03a80c6b9d507fe486d2264a622a574ba285f8e1d5281eeb4b284
|
| 3 |
+
size 14244
|
checkpoint-19500/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c55effe3c0e4f5100a42262d9b2a1ff91095aef99c0edc20263856afcd1fe6b3
|
| 3 |
+
size 1064
|
checkpoint-19500/trainer_state.json
ADDED
|
@@ -0,0 +1,1398 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 300.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 19500,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 1.5384615384615383,
|
| 13 |
+
"grad_norm": 134.7136688232422,
|
| 14 |
+
"learning_rate": 9.981600000000001e-06,
|
| 15 |
+
"loss": 5.6198,
|
| 16 |
+
"step": 100
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 3.076923076923077,
|
| 20 |
+
"grad_norm": 115.98611450195312,
|
| 21 |
+
"learning_rate": 9.961600000000001e-06,
|
| 22 |
+
"loss": 3.5959,
|
| 23 |
+
"step": 200
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 4.615384615384615,
|
| 27 |
+
"grad_norm": 114.23514556884766,
|
| 28 |
+
"learning_rate": 9.941600000000002e-06,
|
| 29 |
+
"loss": 3.3845,
|
| 30 |
+
"step": 300
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 6.153846153846154,
|
| 34 |
+
"grad_norm": 140.5648193359375,
|
| 35 |
+
"learning_rate": 9.921600000000002e-06,
|
| 36 |
+
"loss": 3.1494,
|
| 37 |
+
"step": 400
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 7.6923076923076925,
|
| 41 |
+
"grad_norm": 117.4485855102539,
|
| 42 |
+
"learning_rate": 9.901600000000002e-06,
|
| 43 |
+
"loss": 3.0773,
|
| 44 |
+
"step": 500
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 9.23076923076923,
|
| 48 |
+
"grad_norm": 389.71490478515625,
|
| 49 |
+
"learning_rate": 9.8816e-06,
|
| 50 |
+
"loss": 3.2116,
|
| 51 |
+
"step": 600
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 10.76923076923077,
|
| 55 |
+
"grad_norm": 107.11251831054688,
|
| 56 |
+
"learning_rate": 9.8616e-06,
|
| 57 |
+
"loss": 3.0471,
|
| 58 |
+
"step": 700
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 12.307692307692308,
|
| 62 |
+
"grad_norm": 85.49571228027344,
|
| 63 |
+
"learning_rate": 9.8416e-06,
|
| 64 |
+
"loss": 3.0201,
|
| 65 |
+
"step": 800
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 13.846153846153847,
|
| 69 |
+
"grad_norm": 121.62274932861328,
|
| 70 |
+
"learning_rate": 9.821600000000001e-06,
|
| 71 |
+
"loss": 2.9355,
|
| 72 |
+
"step": 900
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 15.384615384615385,
|
| 76 |
+
"grad_norm": 64.66451263427734,
|
| 77 |
+
"learning_rate": 9.801600000000001e-06,
|
| 78 |
+
"loss": 2.9637,
|
| 79 |
+
"step": 1000
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 16.923076923076923,
|
| 83 |
+
"grad_norm": 182.4657440185547,
|
| 84 |
+
"learning_rate": 9.781600000000001e-06,
|
| 85 |
+
"loss": 2.9819,
|
| 86 |
+
"step": 1100
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 18.46153846153846,
|
| 90 |
+
"grad_norm": 97.80529022216797,
|
| 91 |
+
"learning_rate": 9.761600000000002e-06,
|
| 92 |
+
"loss": 2.9486,
|
| 93 |
+
"step": 1200
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 20.0,
|
| 97 |
+
"grad_norm": 220.0562744140625,
|
| 98 |
+
"learning_rate": 9.741600000000002e-06,
|
| 99 |
+
"loss": 2.8608,
|
| 100 |
+
"step": 1300
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 21.53846153846154,
|
| 104 |
+
"grad_norm": 95.53397369384766,
|
| 105 |
+
"learning_rate": 9.7216e-06,
|
| 106 |
+
"loss": 2.8322,
|
| 107 |
+
"step": 1400
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 23.076923076923077,
|
| 111 |
+
"grad_norm": 67.54853057861328,
|
| 112 |
+
"learning_rate": 9.7016e-06,
|
| 113 |
+
"loss": 2.9429,
|
| 114 |
+
"step": 1500
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 24.615384615384617,
|
| 118 |
+
"grad_norm": 214.53131103515625,
|
| 119 |
+
"learning_rate": 9.6816e-06,
|
| 120 |
+
"loss": 2.7927,
|
| 121 |
+
"step": 1600
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 26.153846153846153,
|
| 125 |
+
"grad_norm": 293.3318786621094,
|
| 126 |
+
"learning_rate": 9.6616e-06,
|
| 127 |
+
"loss": 2.7665,
|
| 128 |
+
"step": 1700
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 27.692307692307693,
|
| 132 |
+
"grad_norm": 216.3682861328125,
|
| 133 |
+
"learning_rate": 9.641600000000001e-06,
|
| 134 |
+
"loss": 2.8309,
|
| 135 |
+
"step": 1800
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 29.23076923076923,
|
| 139 |
+
"grad_norm": 168.0605010986328,
|
| 140 |
+
"learning_rate": 9.621600000000001e-06,
|
| 141 |
+
"loss": 2.8433,
|
| 142 |
+
"step": 1900
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 30.76923076923077,
|
| 146 |
+
"grad_norm": 103.49143981933594,
|
| 147 |
+
"learning_rate": 9.601600000000001e-06,
|
| 148 |
+
"loss": 2.7606,
|
| 149 |
+
"step": 2000
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 32.30769230769231,
|
| 153 |
+
"grad_norm": 116.6761474609375,
|
| 154 |
+
"learning_rate": 9.581600000000002e-06,
|
| 155 |
+
"loss": 2.6595,
|
| 156 |
+
"step": 2100
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 33.84615384615385,
|
| 160 |
+
"grad_norm": 176.0087432861328,
|
| 161 |
+
"learning_rate": 9.5616e-06,
|
| 162 |
+
"loss": 2.632,
|
| 163 |
+
"step": 2200
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 35.38461538461539,
|
| 167 |
+
"grad_norm": 120.39679718017578,
|
| 168 |
+
"learning_rate": 9.5416e-06,
|
| 169 |
+
"loss": 2.5941,
|
| 170 |
+
"step": 2300
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 36.92307692307692,
|
| 174 |
+
"grad_norm": 100.69256591796875,
|
| 175 |
+
"learning_rate": 9.5216e-06,
|
| 176 |
+
"loss": 2.6007,
|
| 177 |
+
"step": 2400
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 38.46153846153846,
|
| 181 |
+
"grad_norm": 254.54440307617188,
|
| 182 |
+
"learning_rate": 9.5016e-06,
|
| 183 |
+
"loss": 2.5548,
|
| 184 |
+
"step": 2500
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 40.0,
|
| 188 |
+
"grad_norm": 113.618896484375,
|
| 189 |
+
"learning_rate": 9.4816e-06,
|
| 190 |
+
"loss": 2.5157,
|
| 191 |
+
"step": 2600
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 41.53846153846154,
|
| 195 |
+
"grad_norm": 825.5616455078125,
|
| 196 |
+
"learning_rate": 9.461600000000001e-06,
|
| 197 |
+
"loss": 2.4832,
|
| 198 |
+
"step": 2700
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 43.07692307692308,
|
| 202 |
+
"grad_norm": 58.83940505981445,
|
| 203 |
+
"learning_rate": 9.441600000000001e-06,
|
| 204 |
+
"loss": 2.4393,
|
| 205 |
+
"step": 2800
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 44.61538461538461,
|
| 209 |
+
"grad_norm": 89.66182708740234,
|
| 210 |
+
"learning_rate": 9.421600000000001e-06,
|
| 211 |
+
"loss": 2.4381,
|
| 212 |
+
"step": 2900
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 46.15384615384615,
|
| 216 |
+
"grad_norm": 242.81410217285156,
|
| 217 |
+
"learning_rate": 9.4016e-06,
|
| 218 |
+
"loss": 2.3978,
|
| 219 |
+
"step": 3000
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 47.69230769230769,
|
| 223 |
+
"grad_norm": 70.43364715576172,
|
| 224 |
+
"learning_rate": 9.3816e-06,
|
| 225 |
+
"loss": 2.3782,
|
| 226 |
+
"step": 3100
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 49.23076923076923,
|
| 230 |
+
"grad_norm": 244.31356811523438,
|
| 231 |
+
"learning_rate": 9.3616e-06,
|
| 232 |
+
"loss": 2.3319,
|
| 233 |
+
"step": 3200
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 50.76923076923077,
|
| 237 |
+
"grad_norm": 59.848697662353516,
|
| 238 |
+
"learning_rate": 9.3416e-06,
|
| 239 |
+
"loss": 2.3205,
|
| 240 |
+
"step": 3300
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 52.30769230769231,
|
| 244 |
+
"grad_norm": 109.26374053955078,
|
| 245 |
+
"learning_rate": 9.3216e-06,
|
| 246 |
+
"loss": 2.2943,
|
| 247 |
+
"step": 3400
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 53.84615384615385,
|
| 251 |
+
"grad_norm": 307.2849426269531,
|
| 252 |
+
"learning_rate": 9.301600000000001e-06,
|
| 253 |
+
"loss": 2.2913,
|
| 254 |
+
"step": 3500
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 55.38461538461539,
|
| 258 |
+
"grad_norm": 69.93729400634766,
|
| 259 |
+
"learning_rate": 9.281600000000001e-06,
|
| 260 |
+
"loss": 2.2166,
|
| 261 |
+
"step": 3600
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 56.92307692307692,
|
| 265 |
+
"grad_norm": 83.3631591796875,
|
| 266 |
+
"learning_rate": 9.261600000000001e-06,
|
| 267 |
+
"loss": 2.2718,
|
| 268 |
+
"step": 3700
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 58.46153846153846,
|
| 272 |
+
"grad_norm": 48.52922821044922,
|
| 273 |
+
"learning_rate": 9.2416e-06,
|
| 274 |
+
"loss": 2.2645,
|
| 275 |
+
"step": 3800
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 60.0,
|
| 279 |
+
"grad_norm": 51.951324462890625,
|
| 280 |
+
"learning_rate": 9.2218e-06,
|
| 281 |
+
"loss": 2.2416,
|
| 282 |
+
"step": 3900
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 61.53846153846154,
|
| 286 |
+
"grad_norm": 254.99429321289062,
|
| 287 |
+
"learning_rate": 9.2018e-06,
|
| 288 |
+
"loss": 2.2104,
|
| 289 |
+
"step": 4000
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 63.07692307692308,
|
| 293 |
+
"grad_norm": 109.68157958984375,
|
| 294 |
+
"learning_rate": 9.1818e-06,
|
| 295 |
+
"loss": 2.1574,
|
| 296 |
+
"step": 4100
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 64.61538461538461,
|
| 300 |
+
"grad_norm": 85.5855941772461,
|
| 301 |
+
"learning_rate": 9.161800000000001e-06,
|
| 302 |
+
"loss": 2.1513,
|
| 303 |
+
"step": 4200
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 66.15384615384616,
|
| 307 |
+
"grad_norm": 63.89781951904297,
|
| 308 |
+
"learning_rate": 9.141800000000001e-06,
|
| 309 |
+
"loss": 2.1331,
|
| 310 |
+
"step": 4300
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 67.6923076923077,
|
| 314 |
+
"grad_norm": 184.29612731933594,
|
| 315 |
+
"learning_rate": 9.121800000000001e-06,
|
| 316 |
+
"loss": 2.1866,
|
| 317 |
+
"step": 4400
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 69.23076923076923,
|
| 321 |
+
"grad_norm": 108.59600830078125,
|
| 322 |
+
"learning_rate": 9.1018e-06,
|
| 323 |
+
"loss": 2.1453,
|
| 324 |
+
"step": 4500
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 70.76923076923077,
|
| 328 |
+
"grad_norm": 80.51258087158203,
|
| 329 |
+
"learning_rate": 9.0818e-06,
|
| 330 |
+
"loss": 2.1125,
|
| 331 |
+
"step": 4600
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 72.3076923076923,
|
| 335 |
+
"grad_norm": 172.7328643798828,
|
| 336 |
+
"learning_rate": 9.0618e-06,
|
| 337 |
+
"loss": 2.141,
|
| 338 |
+
"step": 4700
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 73.84615384615384,
|
| 342 |
+
"grad_norm": 58.20785903930664,
|
| 343 |
+
"learning_rate": 9.0418e-06,
|
| 344 |
+
"loss": 2.0786,
|
| 345 |
+
"step": 4800
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 75.38461538461539,
|
| 349 |
+
"grad_norm": 72.46143341064453,
|
| 350 |
+
"learning_rate": 9.0218e-06,
|
| 351 |
+
"loss": 2.0914,
|
| 352 |
+
"step": 4900
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 76.92307692307692,
|
| 356 |
+
"grad_norm": 48.18838119506836,
|
| 357 |
+
"learning_rate": 9.0018e-06,
|
| 358 |
+
"loss": 2.0439,
|
| 359 |
+
"step": 5000
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 78.46153846153847,
|
| 363 |
+
"grad_norm": 87.95365905761719,
|
| 364 |
+
"learning_rate": 8.981800000000001e-06,
|
| 365 |
+
"loss": 2.0331,
|
| 366 |
+
"step": 5100
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 80.0,
|
| 370 |
+
"grad_norm": 99.82858276367188,
|
| 371 |
+
"learning_rate": 8.961800000000001e-06,
|
| 372 |
+
"loss": 2.1268,
|
| 373 |
+
"step": 5200
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 81.53846153846153,
|
| 377 |
+
"grad_norm": 85.33229064941406,
|
| 378 |
+
"learning_rate": 8.9418e-06,
|
| 379 |
+
"loss": 2.028,
|
| 380 |
+
"step": 5300
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 83.07692307692308,
|
| 384 |
+
"grad_norm": 83.01776885986328,
|
| 385 |
+
"learning_rate": 8.9218e-06,
|
| 386 |
+
"loss": 2.0179,
|
| 387 |
+
"step": 5400
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 84.61538461538461,
|
| 391 |
+
"grad_norm": 95.0901107788086,
|
| 392 |
+
"learning_rate": 8.9018e-06,
|
| 393 |
+
"loss": 2.0825,
|
| 394 |
+
"step": 5500
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 86.15384615384616,
|
| 398 |
+
"grad_norm": 85.66802978515625,
|
| 399 |
+
"learning_rate": 8.8818e-06,
|
| 400 |
+
"loss": 2.0588,
|
| 401 |
+
"step": 5600
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 87.6923076923077,
|
| 405 |
+
"grad_norm": 69.63407897949219,
|
| 406 |
+
"learning_rate": 8.8618e-06,
|
| 407 |
+
"loss": 2.0298,
|
| 408 |
+
"step": 5700
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 89.23076923076923,
|
| 412 |
+
"grad_norm": 174.88063049316406,
|
| 413 |
+
"learning_rate": 8.8418e-06,
|
| 414 |
+
"loss": 1.9867,
|
| 415 |
+
"step": 5800
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 90.76923076923077,
|
| 419 |
+
"grad_norm": 65.64617156982422,
|
| 420 |
+
"learning_rate": 8.8218e-06,
|
| 421 |
+
"loss": 2.0674,
|
| 422 |
+
"step": 5900
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 92.3076923076923,
|
| 426 |
+
"grad_norm": 152.2218780517578,
|
| 427 |
+
"learning_rate": 8.802e-06,
|
| 428 |
+
"loss": 1.992,
|
| 429 |
+
"step": 6000
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 93.84615384615384,
|
| 433 |
+
"grad_norm": 74.4027328491211,
|
| 434 |
+
"learning_rate": 8.782e-06,
|
| 435 |
+
"loss": 2.0016,
|
| 436 |
+
"step": 6100
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 95.38461538461539,
|
| 440 |
+
"grad_norm": 67.4207992553711,
|
| 441 |
+
"learning_rate": 8.762e-06,
|
| 442 |
+
"loss": 2.0001,
|
| 443 |
+
"step": 6200
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 96.92307692307692,
|
| 447 |
+
"grad_norm": 182.99061584472656,
|
| 448 |
+
"learning_rate": 8.742e-06,
|
| 449 |
+
"loss": 1.9897,
|
| 450 |
+
"step": 6300
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 98.46153846153847,
|
| 454 |
+
"grad_norm": 70.14286804199219,
|
| 455 |
+
"learning_rate": 8.722e-06,
|
| 456 |
+
"loss": 2.0146,
|
| 457 |
+
"step": 6400
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 100.0,
|
| 461 |
+
"grad_norm": 117.57647705078125,
|
| 462 |
+
"learning_rate": 8.702e-06,
|
| 463 |
+
"loss": 1.9298,
|
| 464 |
+
"step": 6500
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 101.53846153846153,
|
| 468 |
+
"grad_norm": 80.27694702148438,
|
| 469 |
+
"learning_rate": 8.682000000000001e-06,
|
| 470 |
+
"loss": 1.9364,
|
| 471 |
+
"step": 6600
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 103.07692307692308,
|
| 475 |
+
"grad_norm": 219.95655822753906,
|
| 476 |
+
"learning_rate": 8.662000000000001e-06,
|
| 477 |
+
"loss": 1.9289,
|
| 478 |
+
"step": 6700
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 104.61538461538461,
|
| 482 |
+
"grad_norm": 176.79177856445312,
|
| 483 |
+
"learning_rate": 8.642e-06,
|
| 484 |
+
"loss": 1.9142,
|
| 485 |
+
"step": 6800
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 106.15384615384616,
|
| 489 |
+
"grad_norm": 60.430049896240234,
|
| 490 |
+
"learning_rate": 8.622e-06,
|
| 491 |
+
"loss": 1.8986,
|
| 492 |
+
"step": 6900
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 107.6923076923077,
|
| 496 |
+
"grad_norm": 53.74516296386719,
|
| 497 |
+
"learning_rate": 8.602e-06,
|
| 498 |
+
"loss": 1.9029,
|
| 499 |
+
"step": 7000
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 109.23076923076923,
|
| 503 |
+
"grad_norm": 70.89128112792969,
|
| 504 |
+
"learning_rate": 8.582e-06,
|
| 505 |
+
"loss": 1.9313,
|
| 506 |
+
"step": 7100
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 110.76923076923077,
|
| 510 |
+
"grad_norm": 137.73828125,
|
| 511 |
+
"learning_rate": 8.562e-06,
|
| 512 |
+
"loss": 1.9328,
|
| 513 |
+
"step": 7200
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 112.3076923076923,
|
| 517 |
+
"grad_norm": 86.80708312988281,
|
| 518 |
+
"learning_rate": 8.542e-06,
|
| 519 |
+
"loss": 1.8831,
|
| 520 |
+
"step": 7300
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 113.84615384615384,
|
| 524 |
+
"grad_norm": 52.322608947753906,
|
| 525 |
+
"learning_rate": 8.522200000000001e-06,
|
| 526 |
+
"loss": 1.8679,
|
| 527 |
+
"step": 7400
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 115.38461538461539,
|
| 531 |
+
"grad_norm": 215.9288787841797,
|
| 532 |
+
"learning_rate": 8.5022e-06,
|
| 533 |
+
"loss": 1.8743,
|
| 534 |
+
"step": 7500
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 116.92307692307692,
|
| 538 |
+
"grad_norm": 74.13555908203125,
|
| 539 |
+
"learning_rate": 8.4822e-06,
|
| 540 |
+
"loss": 1.8575,
|
| 541 |
+
"step": 7600
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 118.46153846153847,
|
| 545 |
+
"grad_norm": 138.6452178955078,
|
| 546 |
+
"learning_rate": 8.4622e-06,
|
| 547 |
+
"loss": 1.8942,
|
| 548 |
+
"step": 7700
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 120.0,
|
| 552 |
+
"grad_norm": 57.42049789428711,
|
| 553 |
+
"learning_rate": 8.4422e-06,
|
| 554 |
+
"loss": 1.8777,
|
| 555 |
+
"step": 7800
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 121.53846153846153,
|
| 559 |
+
"grad_norm": 41.08803939819336,
|
| 560 |
+
"learning_rate": 8.4222e-06,
|
| 561 |
+
"loss": 1.8409,
|
| 562 |
+
"step": 7900
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 123.07692307692308,
|
| 566 |
+
"grad_norm": 66.80331420898438,
|
| 567 |
+
"learning_rate": 8.4022e-06,
|
| 568 |
+
"loss": 1.8325,
|
| 569 |
+
"step": 8000
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 124.61538461538461,
|
| 573 |
+
"grad_norm": 69.8762435913086,
|
| 574 |
+
"learning_rate": 8.3822e-06,
|
| 575 |
+
"loss": 1.8685,
|
| 576 |
+
"step": 8100
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 126.15384615384616,
|
| 580 |
+
"grad_norm": 109.0595703125,
|
| 581 |
+
"learning_rate": 8.362200000000001e-06,
|
| 582 |
+
"loss": 1.8423,
|
| 583 |
+
"step": 8200
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 127.6923076923077,
|
| 587 |
+
"grad_norm": 75.782470703125,
|
| 588 |
+
"learning_rate": 8.3422e-06,
|
| 589 |
+
"loss": 1.8278,
|
| 590 |
+
"step": 8300
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 129.23076923076923,
|
| 594 |
+
"grad_norm": 495.26025390625,
|
| 595 |
+
"learning_rate": 8.3222e-06,
|
| 596 |
+
"loss": 1.823,
|
| 597 |
+
"step": 8400
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 130.76923076923077,
|
| 601 |
+
"grad_norm": 132.37982177734375,
|
| 602 |
+
"learning_rate": 8.3022e-06,
|
| 603 |
+
"loss": 1.8122,
|
| 604 |
+
"step": 8500
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 132.30769230769232,
|
| 608 |
+
"grad_norm": 110.41214752197266,
|
| 609 |
+
"learning_rate": 8.2822e-06,
|
| 610 |
+
"loss": 1.798,
|
| 611 |
+
"step": 8600
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 133.84615384615384,
|
| 615 |
+
"grad_norm": 155.9139404296875,
|
| 616 |
+
"learning_rate": 8.2622e-06,
|
| 617 |
+
"loss": 1.8158,
|
| 618 |
+
"step": 8700
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 135.3846153846154,
|
| 622 |
+
"grad_norm": 61.789390563964844,
|
| 623 |
+
"learning_rate": 8.2422e-06,
|
| 624 |
+
"loss": 1.7863,
|
| 625 |
+
"step": 8800
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 136.92307692307693,
|
| 629 |
+
"grad_norm": 86.55162048339844,
|
| 630 |
+
"learning_rate": 8.2222e-06,
|
| 631 |
+
"loss": 1.7971,
|
| 632 |
+
"step": 8900
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 138.46153846153845,
|
| 636 |
+
"grad_norm": 52.087581634521484,
|
| 637 |
+
"learning_rate": 8.202200000000001e-06,
|
| 638 |
+
"loss": 1.7676,
|
| 639 |
+
"step": 9000
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 140.0,
|
| 643 |
+
"grad_norm": 78.67654418945312,
|
| 644 |
+
"learning_rate": 8.1822e-06,
|
| 645 |
+
"loss": 1.8188,
|
| 646 |
+
"step": 9100
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 141.53846153846155,
|
| 650 |
+
"grad_norm": 102.5003890991211,
|
| 651 |
+
"learning_rate": 8.1622e-06,
|
| 652 |
+
"loss": 1.8154,
|
| 653 |
+
"step": 9200
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 143.07692307692307,
|
| 657 |
+
"grad_norm": 112.49629974365234,
|
| 658 |
+
"learning_rate": 8.142200000000001e-06,
|
| 659 |
+
"loss": 1.7687,
|
| 660 |
+
"step": 9300
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 144.6153846153846,
|
| 664 |
+
"grad_norm": 64.63585662841797,
|
| 665 |
+
"learning_rate": 8.122200000000002e-06,
|
| 666 |
+
"loss": 1.7824,
|
| 667 |
+
"step": 9400
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 146.15384615384616,
|
| 671 |
+
"grad_norm": 74.50164794921875,
|
| 672 |
+
"learning_rate": 8.102200000000002e-06,
|
| 673 |
+
"loss": 1.745,
|
| 674 |
+
"step": 9500
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 147.69230769230768,
|
| 678 |
+
"grad_norm": 70.59980010986328,
|
| 679 |
+
"learning_rate": 8.082200000000002e-06,
|
| 680 |
+
"loss": 1.8179,
|
| 681 |
+
"step": 9600
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 149.23076923076923,
|
| 685 |
+
"grad_norm": 275.9650573730469,
|
| 686 |
+
"learning_rate": 8.0622e-06,
|
| 687 |
+
"loss": 1.7744,
|
| 688 |
+
"step": 9700
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 150.76923076923077,
|
| 692 |
+
"grad_norm": 54.21783447265625,
|
| 693 |
+
"learning_rate": 8.0422e-06,
|
| 694 |
+
"loss": 1.7608,
|
| 695 |
+
"step": 9800
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 152.30769230769232,
|
| 699 |
+
"grad_norm": 58.89350128173828,
|
| 700 |
+
"learning_rate": 8.0224e-06,
|
| 701 |
+
"loss": 1.7835,
|
| 702 |
+
"step": 9900
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 153.84615384615384,
|
| 706 |
+
"grad_norm": 90.88015747070312,
|
| 707 |
+
"learning_rate": 8.0024e-06,
|
| 708 |
+
"loss": 1.7567,
|
| 709 |
+
"step": 10000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 155.3846153846154,
|
| 713 |
+
"grad_norm": 45.44640350341797,
|
| 714 |
+
"learning_rate": 7.9824e-06,
|
| 715 |
+
"loss": 1.7399,
|
| 716 |
+
"step": 10100
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 156.92307692307693,
|
| 720 |
+
"grad_norm": 58.44306945800781,
|
| 721 |
+
"learning_rate": 7.9624e-06,
|
| 722 |
+
"loss": 1.7803,
|
| 723 |
+
"step": 10200
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 158.46153846153845,
|
| 727 |
+
"grad_norm": 80.27843475341797,
|
| 728 |
+
"learning_rate": 7.9424e-06,
|
| 729 |
+
"loss": 1.7106,
|
| 730 |
+
"step": 10300
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 160.0,
|
| 734 |
+
"grad_norm": 111.12615203857422,
|
| 735 |
+
"learning_rate": 7.9224e-06,
|
| 736 |
+
"loss": 1.7134,
|
| 737 |
+
"step": 10400
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 161.53846153846155,
|
| 741 |
+
"grad_norm": 63.302520751953125,
|
| 742 |
+
"learning_rate": 7.9024e-06,
|
| 743 |
+
"loss": 1.7031,
|
| 744 |
+
"step": 10500
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 163.07692307692307,
|
| 748 |
+
"grad_norm": 41.779808044433594,
|
| 749 |
+
"learning_rate": 7.882400000000001e-06,
|
| 750 |
+
"loss": 1.7071,
|
| 751 |
+
"step": 10600
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 164.6153846153846,
|
| 755 |
+
"grad_norm": 58.44318389892578,
|
| 756 |
+
"learning_rate": 7.862400000000001e-06,
|
| 757 |
+
"loss": 1.6759,
|
| 758 |
+
"step": 10700
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 166.15384615384616,
|
| 762 |
+
"grad_norm": 36.30861282348633,
|
| 763 |
+
"learning_rate": 7.842400000000001e-06,
|
| 764 |
+
"loss": 1.7012,
|
| 765 |
+
"step": 10800
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 167.69230769230768,
|
| 769 |
+
"grad_norm": 68.7526626586914,
|
| 770 |
+
"learning_rate": 7.822400000000002e-06,
|
| 771 |
+
"loss": 1.6963,
|
| 772 |
+
"step": 10900
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 169.23076923076923,
|
| 776 |
+
"grad_norm": 82.03331756591797,
|
| 777 |
+
"learning_rate": 7.802400000000002e-06,
|
| 778 |
+
"loss": 1.7456,
|
| 779 |
+
"step": 11000
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 170.76923076923077,
|
| 783 |
+
"grad_norm": 53.351959228515625,
|
| 784 |
+
"learning_rate": 7.7824e-06,
|
| 785 |
+
"loss": 1.7284,
|
| 786 |
+
"step": 11100
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 172.30769230769232,
|
| 790 |
+
"grad_norm": 58.400909423828125,
|
| 791 |
+
"learning_rate": 7.7624e-06,
|
| 792 |
+
"loss": 1.7266,
|
| 793 |
+
"step": 11200
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 173.84615384615384,
|
| 797 |
+
"grad_norm": 47.95368576049805,
|
| 798 |
+
"learning_rate": 7.7424e-06,
|
| 799 |
+
"loss": 1.6671,
|
| 800 |
+
"step": 11300
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 175.3846153846154,
|
| 804 |
+
"grad_norm": 54.79304885864258,
|
| 805 |
+
"learning_rate": 7.7224e-06,
|
| 806 |
+
"loss": 1.7092,
|
| 807 |
+
"step": 11400
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 176.92307692307693,
|
| 811 |
+
"grad_norm": 34.120758056640625,
|
| 812 |
+
"learning_rate": 7.702400000000001e-06,
|
| 813 |
+
"loss": 1.724,
|
| 814 |
+
"step": 11500
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 178.46153846153845,
|
| 818 |
+
"grad_norm": 80.98869323730469,
|
| 819 |
+
"learning_rate": 7.682400000000001e-06,
|
| 820 |
+
"loss": 1.6693,
|
| 821 |
+
"step": 11600
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 180.0,
|
| 825 |
+
"grad_norm": 65.80884552001953,
|
| 826 |
+
"learning_rate": 7.662400000000001e-06,
|
| 827 |
+
"loss": 1.6864,
|
| 828 |
+
"step": 11700
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 181.53846153846155,
|
| 832 |
+
"grad_norm": 86.81282806396484,
|
| 833 |
+
"learning_rate": 7.642400000000002e-06,
|
| 834 |
+
"loss": 1.6715,
|
| 835 |
+
"step": 11800
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 183.07692307692307,
|
| 839 |
+
"grad_norm": 63.2209358215332,
|
| 840 |
+
"learning_rate": 7.622400000000001e-06,
|
| 841 |
+
"loss": 1.6884,
|
| 842 |
+
"step": 11900
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 184.6153846153846,
|
| 846 |
+
"grad_norm": 80.72114562988281,
|
| 847 |
+
"learning_rate": 7.602400000000001e-06,
|
| 848 |
+
"loss": 1.6781,
|
| 849 |
+
"step": 12000
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 186.15384615384616,
|
| 853 |
+
"grad_norm": 65.09835815429688,
|
| 854 |
+
"learning_rate": 7.5824000000000005e-06,
|
| 855 |
+
"loss": 1.6451,
|
| 856 |
+
"step": 12100
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 187.69230769230768,
|
| 860 |
+
"grad_norm": 51.8597297668457,
|
| 861 |
+
"learning_rate": 7.562400000000001e-06,
|
| 862 |
+
"loss": 1.6425,
|
| 863 |
+
"step": 12200
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 189.23076923076923,
|
| 867 |
+
"grad_norm": 55.06999969482422,
|
| 868 |
+
"learning_rate": 7.542400000000001e-06,
|
| 869 |
+
"loss": 1.6475,
|
| 870 |
+
"step": 12300
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 190.76923076923077,
|
| 874 |
+
"grad_norm": 221.47671508789062,
|
| 875 |
+
"learning_rate": 7.522400000000001e-06,
|
| 876 |
+
"loss": 1.6459,
|
| 877 |
+
"step": 12400
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 192.30769230769232,
|
| 881 |
+
"grad_norm": 64.53801727294922,
|
| 882 |
+
"learning_rate": 7.5024e-06,
|
| 883 |
+
"loss": 1.6149,
|
| 884 |
+
"step": 12500
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 193.84615384615384,
|
| 888 |
+
"grad_norm": 96.24422454833984,
|
| 889 |
+
"learning_rate": 7.4824000000000006e-06,
|
| 890 |
+
"loss": 1.6204,
|
| 891 |
+
"step": 12600
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 195.3846153846154,
|
| 895 |
+
"grad_norm": 44.547306060791016,
|
| 896 |
+
"learning_rate": 7.462400000000001e-06,
|
| 897 |
+
"loss": 1.6151,
|
| 898 |
+
"step": 12700
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 196.92307692307693,
|
| 902 |
+
"grad_norm": 125.61988830566406,
|
| 903 |
+
"learning_rate": 7.442400000000001e-06,
|
| 904 |
+
"loss": 1.6244,
|
| 905 |
+
"step": 12800
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 198.46153846153845,
|
| 909 |
+
"grad_norm": 98.02783966064453,
|
| 910 |
+
"learning_rate": 7.422400000000001e-06,
|
| 911 |
+
"loss": 1.625,
|
| 912 |
+
"step": 12900
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 200.0,
|
| 916 |
+
"grad_norm": 72.0836181640625,
|
| 917 |
+
"learning_rate": 7.4024000000000005e-06,
|
| 918 |
+
"loss": 1.6201,
|
| 919 |
+
"step": 13000
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 201.53846153846155,
|
| 923 |
+
"grad_norm": 157.2918701171875,
|
| 924 |
+
"learning_rate": 7.382400000000001e-06,
|
| 925 |
+
"loss": 1.6065,
|
| 926 |
+
"step": 13100
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 203.07692307692307,
|
| 930 |
+
"grad_norm": 51.04494094848633,
|
| 931 |
+
"learning_rate": 7.362400000000001e-06,
|
| 932 |
+
"loss": 1.6227,
|
| 933 |
+
"step": 13200
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 204.6153846153846,
|
| 937 |
+
"grad_norm": 53.116188049316406,
|
| 938 |
+
"learning_rate": 7.342400000000001e-06,
|
| 939 |
+
"loss": 1.6284,
|
| 940 |
+
"step": 13300
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 206.15384615384616,
|
| 944 |
+
"grad_norm": 80.10623168945312,
|
| 945 |
+
"learning_rate": 7.3224e-06,
|
| 946 |
+
"loss": 1.619,
|
| 947 |
+
"step": 13400
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 207.69230769230768,
|
| 951 |
+
"grad_norm": 249.86141967773438,
|
| 952 |
+
"learning_rate": 7.302400000000001e-06,
|
| 953 |
+
"loss": 1.6018,
|
| 954 |
+
"step": 13500
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 209.23076923076923,
|
| 958 |
+
"grad_norm": 77.1427001953125,
|
| 959 |
+
"learning_rate": 7.282400000000001e-06,
|
| 960 |
+
"loss": 1.6329,
|
| 961 |
+
"step": 13600
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 210.76923076923077,
|
| 965 |
+
"grad_norm": 86.5726318359375,
|
| 966 |
+
"learning_rate": 7.2626000000000006e-06,
|
| 967 |
+
"loss": 1.6112,
|
| 968 |
+
"step": 13700
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 212.30769230769232,
|
| 972 |
+
"grad_norm": 66.48815155029297,
|
| 973 |
+
"learning_rate": 7.242600000000001e-06,
|
| 974 |
+
"loss": 1.6128,
|
| 975 |
+
"step": 13800
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 213.84615384615384,
|
| 979 |
+
"grad_norm": 99.87622833251953,
|
| 980 |
+
"learning_rate": 7.222600000000001e-06,
|
| 981 |
+
"loss": 1.5902,
|
| 982 |
+
"step": 13900
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 215.3846153846154,
|
| 986 |
+
"grad_norm": 78.00489807128906,
|
| 987 |
+
"learning_rate": 7.2026e-06,
|
| 988 |
+
"loss": 1.5964,
|
| 989 |
+
"step": 14000
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 216.92307692307693,
|
| 993 |
+
"grad_norm": 726.4310913085938,
|
| 994 |
+
"learning_rate": 7.1826000000000005e-06,
|
| 995 |
+
"loss": 1.6301,
|
| 996 |
+
"step": 14100
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 218.46153846153845,
|
| 1000 |
+
"grad_norm": 60.30306625366211,
|
| 1001 |
+
"learning_rate": 7.162600000000001e-06,
|
| 1002 |
+
"loss": 1.5653,
|
| 1003 |
+
"step": 14200
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 220.0,
|
| 1007 |
+
"grad_norm": 88.68881225585938,
|
| 1008 |
+
"learning_rate": 7.142600000000001e-06,
|
| 1009 |
+
"loss": 1.5743,
|
| 1010 |
+
"step": 14300
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 221.53846153846155,
|
| 1014 |
+
"grad_norm": 101.28644561767578,
|
| 1015 |
+
"learning_rate": 7.1226e-06,
|
| 1016 |
+
"loss": 1.5589,
|
| 1017 |
+
"step": 14400
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 223.07692307692307,
|
| 1021 |
+
"grad_norm": 248.6482696533203,
|
| 1022 |
+
"learning_rate": 7.1026e-06,
|
| 1023 |
+
"loss": 1.5706,
|
| 1024 |
+
"step": 14500
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 224.6153846153846,
|
| 1028 |
+
"grad_norm": 110.11705780029297,
|
| 1029 |
+
"learning_rate": 7.082600000000001e-06,
|
| 1030 |
+
"loss": 1.5843,
|
| 1031 |
+
"step": 14600
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 226.15384615384616,
|
| 1035 |
+
"grad_norm": 67.07064819335938,
|
| 1036 |
+
"learning_rate": 7.062600000000001e-06,
|
| 1037 |
+
"loss": 1.5643,
|
| 1038 |
+
"step": 14700
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 227.69230769230768,
|
| 1042 |
+
"grad_norm": 97.37787628173828,
|
| 1043 |
+
"learning_rate": 7.0426e-06,
|
| 1044 |
+
"loss": 1.5424,
|
| 1045 |
+
"step": 14800
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 229.23076923076923,
|
| 1049 |
+
"grad_norm": 53.19269561767578,
|
| 1050 |
+
"learning_rate": 7.0226e-06,
|
| 1051 |
+
"loss": 1.5644,
|
| 1052 |
+
"step": 14900
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 230.76923076923077,
|
| 1056 |
+
"grad_norm": 213.32289123535156,
|
| 1057 |
+
"learning_rate": 7.0026000000000005e-06,
|
| 1058 |
+
"loss": 1.5541,
|
| 1059 |
+
"step": 15000
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 232.30769230769232,
|
| 1063 |
+
"grad_norm": 105.63107299804688,
|
| 1064 |
+
"learning_rate": 6.982600000000001e-06,
|
| 1065 |
+
"loss": 1.5518,
|
| 1066 |
+
"step": 15100
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 233.84615384615384,
|
| 1070 |
+
"grad_norm": 43.85015106201172,
|
| 1071 |
+
"learning_rate": 6.9626e-06,
|
| 1072 |
+
"loss": 1.5385,
|
| 1073 |
+
"step": 15200
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 235.3846153846154,
|
| 1077 |
+
"grad_norm": 135.48651123046875,
|
| 1078 |
+
"learning_rate": 6.9426e-06,
|
| 1079 |
+
"loss": 1.5093,
|
| 1080 |
+
"step": 15300
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 236.92307692307693,
|
| 1084 |
+
"grad_norm": 73.94020080566406,
|
| 1085 |
+
"learning_rate": 6.9226000000000004e-06,
|
| 1086 |
+
"loss": 1.537,
|
| 1087 |
+
"step": 15400
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 238.46153846153845,
|
| 1091 |
+
"grad_norm": 157.08612060546875,
|
| 1092 |
+
"learning_rate": 6.902600000000001e-06,
|
| 1093 |
+
"loss": 1.5757,
|
| 1094 |
+
"step": 15500
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 240.0,
|
| 1098 |
+
"grad_norm": 79.48292541503906,
|
| 1099 |
+
"learning_rate": 6.8826e-06,
|
| 1100 |
+
"loss": 1.5214,
|
| 1101 |
+
"step": 15600
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 241.53846153846155,
|
| 1105 |
+
"grad_norm": 59.73866653442383,
|
| 1106 |
+
"learning_rate": 6.8626e-06,
|
| 1107 |
+
"loss": 1.5166,
|
| 1108 |
+
"step": 15700
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 243.07692307692307,
|
| 1112 |
+
"grad_norm": 97.34918212890625,
|
| 1113 |
+
"learning_rate": 6.8426e-06,
|
| 1114 |
+
"loss": 1.5145,
|
| 1115 |
+
"step": 15800
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 244.6153846153846,
|
| 1119 |
+
"grad_norm": 198.27203369140625,
|
| 1120 |
+
"learning_rate": 6.8226000000000005e-06,
|
| 1121 |
+
"loss": 1.5237,
|
| 1122 |
+
"step": 15900
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 246.15384615384616,
|
| 1126 |
+
"grad_norm": 38.80288314819336,
|
| 1127 |
+
"learning_rate": 6.802600000000001e-06,
|
| 1128 |
+
"loss": 1.5118,
|
| 1129 |
+
"step": 16000
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 247.69230769230768,
|
| 1133 |
+
"grad_norm": 89.98989868164062,
|
| 1134 |
+
"learning_rate": 6.7828000000000005e-06,
|
| 1135 |
+
"loss": 1.5171,
|
| 1136 |
+
"step": 16100
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 249.23076923076923,
|
| 1140 |
+
"grad_norm": 55.13770294189453,
|
| 1141 |
+
"learning_rate": 6.762800000000001e-06,
|
| 1142 |
+
"loss": 1.5044,
|
| 1143 |
+
"step": 16200
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 250.76923076923077,
|
| 1147 |
+
"grad_norm": 34.529293060302734,
|
| 1148 |
+
"learning_rate": 6.7428e-06,
|
| 1149 |
+
"loss": 1.4981,
|
| 1150 |
+
"step": 16300
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 252.30769230769232,
|
| 1154 |
+
"grad_norm": 79.73957824707031,
|
| 1155 |
+
"learning_rate": 6.7228e-06,
|
| 1156 |
+
"loss": 1.501,
|
| 1157 |
+
"step": 16400
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 253.84615384615384,
|
| 1161 |
+
"grad_norm": 51.99100875854492,
|
| 1162 |
+
"learning_rate": 6.7028000000000004e-06,
|
| 1163 |
+
"loss": 1.4968,
|
| 1164 |
+
"step": 16500
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 255.3846153846154,
|
| 1168 |
+
"grad_norm": 47.11521530151367,
|
| 1169 |
+
"learning_rate": 6.682800000000001e-06,
|
| 1170 |
+
"loss": 1.4961,
|
| 1171 |
+
"step": 16600
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 256.9230769230769,
|
| 1175 |
+
"grad_norm": 70.4249496459961,
|
| 1176 |
+
"learning_rate": 6.6628e-06,
|
| 1177 |
+
"loss": 1.5162,
|
| 1178 |
+
"step": 16700
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 258.46153846153845,
|
| 1182 |
+
"grad_norm": 56.895957946777344,
|
| 1183 |
+
"learning_rate": 6.6428e-06,
|
| 1184 |
+
"loss": 1.5187,
|
| 1185 |
+
"step": 16800
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 260.0,
|
| 1189 |
+
"grad_norm": 34.502071380615234,
|
| 1190 |
+
"learning_rate": 6.623000000000001e-06,
|
| 1191 |
+
"loss": 1.5325,
|
| 1192 |
+
"step": 16900
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 261.53846153846155,
|
| 1196 |
+
"grad_norm": 66.03074645996094,
|
| 1197 |
+
"learning_rate": 6.603e-06,
|
| 1198 |
+
"loss": 1.5103,
|
| 1199 |
+
"step": 17000
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 263.0769230769231,
|
| 1203 |
+
"grad_norm": 110.48493957519531,
|
| 1204 |
+
"learning_rate": 6.583e-06,
|
| 1205 |
+
"loss": 1.4571,
|
| 1206 |
+
"step": 17100
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 264.61538461538464,
|
| 1210 |
+
"grad_norm": 55.59686279296875,
|
| 1211 |
+
"learning_rate": 6.5630000000000005e-06,
|
| 1212 |
+
"loss": 1.5147,
|
| 1213 |
+
"step": 17200
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 266.15384615384613,
|
| 1217 |
+
"grad_norm": 72.83611297607422,
|
| 1218 |
+
"learning_rate": 6.543000000000001e-06,
|
| 1219 |
+
"loss": 1.48,
|
| 1220 |
+
"step": 17300
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 267.6923076923077,
|
| 1224 |
+
"grad_norm": 54.794437408447266,
|
| 1225 |
+
"learning_rate": 6.523e-06,
|
| 1226 |
+
"loss": 1.461,
|
| 1227 |
+
"step": 17400
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 269.2307692307692,
|
| 1231 |
+
"grad_norm": 41.280731201171875,
|
| 1232 |
+
"learning_rate": 6.503e-06,
|
| 1233 |
+
"loss": 1.4188,
|
| 1234 |
+
"step": 17500
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 270.7692307692308,
|
| 1238 |
+
"grad_norm": 81.90428924560547,
|
| 1239 |
+
"learning_rate": 6.4830000000000004e-06,
|
| 1240 |
+
"loss": 1.4256,
|
| 1241 |
+
"step": 17600
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 272.3076923076923,
|
| 1245 |
+
"grad_norm": 67.94799041748047,
|
| 1246 |
+
"learning_rate": 6.463000000000001e-06,
|
| 1247 |
+
"loss": 1.4295,
|
| 1248 |
+
"step": 17700
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 273.84615384615387,
|
| 1252 |
+
"grad_norm": 77.92097473144531,
|
| 1253 |
+
"learning_rate": 6.443e-06,
|
| 1254 |
+
"loss": 1.4575,
|
| 1255 |
+
"step": 17800
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 275.38461538461536,
|
| 1259 |
+
"grad_norm": 83.06784057617188,
|
| 1260 |
+
"learning_rate": 6.423e-06,
|
| 1261 |
+
"loss": 1.434,
|
| 1262 |
+
"step": 17900
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 276.9230769230769,
|
| 1266 |
+
"grad_norm": 60.48369598388672,
|
| 1267 |
+
"learning_rate": 6.403e-06,
|
| 1268 |
+
"loss": 1.4496,
|
| 1269 |
+
"step": 18000
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 278.46153846153845,
|
| 1273 |
+
"grad_norm": 47.32282257080078,
|
| 1274 |
+
"learning_rate": 6.3830000000000006e-06,
|
| 1275 |
+
"loss": 1.4149,
|
| 1276 |
+
"step": 18100
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 280.0,
|
| 1280 |
+
"grad_norm": 134.77822875976562,
|
| 1281 |
+
"learning_rate": 6.363e-06,
|
| 1282 |
+
"loss": 1.3949,
|
| 1283 |
+
"step": 18200
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 281.53846153846155,
|
| 1287 |
+
"grad_norm": 72.09223937988281,
|
| 1288 |
+
"learning_rate": 6.343e-06,
|
| 1289 |
+
"loss": 1.4298,
|
| 1290 |
+
"step": 18300
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 283.0769230769231,
|
| 1294 |
+
"grad_norm": 95.94608306884766,
|
| 1295 |
+
"learning_rate": 6.323e-06,
|
| 1296 |
+
"loss": 1.4189,
|
| 1297 |
+
"step": 18400
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 284.61538461538464,
|
| 1301 |
+
"grad_norm": 79.95153045654297,
|
| 1302 |
+
"learning_rate": 6.3030000000000005e-06,
|
| 1303 |
+
"loss": 1.4066,
|
| 1304 |
+
"step": 18500
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 286.15384615384613,
|
| 1308 |
+
"grad_norm": 165.74986267089844,
|
| 1309 |
+
"learning_rate": 6.283e-06,
|
| 1310 |
+
"loss": 1.4231,
|
| 1311 |
+
"step": 18600
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 287.6923076923077,
|
| 1315 |
+
"grad_norm": 105.84443664550781,
|
| 1316 |
+
"learning_rate": 6.263e-06,
|
| 1317 |
+
"loss": 1.3851,
|
| 1318 |
+
"step": 18700
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 289.2307692307692,
|
| 1322 |
+
"grad_norm": 48.660987854003906,
|
| 1323 |
+
"learning_rate": 6.243e-06,
|
| 1324 |
+
"loss": 1.4234,
|
| 1325 |
+
"step": 18800
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 290.7692307692308,
|
| 1329 |
+
"grad_norm": 53.28395462036133,
|
| 1330 |
+
"learning_rate": 6.223e-06,
|
| 1331 |
+
"loss": 1.3861,
|
| 1332 |
+
"step": 18900
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 292.3076923076923,
|
| 1336 |
+
"grad_norm": 54.91928482055664,
|
| 1337 |
+
"learning_rate": 6.203e-06,
|
| 1338 |
+
"loss": 1.3799,
|
| 1339 |
+
"step": 19000
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 293.84615384615387,
|
| 1343 |
+
"grad_norm": 69.98273468017578,
|
| 1344 |
+
"learning_rate": 6.183e-06,
|
| 1345 |
+
"loss": 1.4223,
|
| 1346 |
+
"step": 19100
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 295.38461538461536,
|
| 1350 |
+
"grad_norm": 55.819786071777344,
|
| 1351 |
+
"learning_rate": 6.163e-06,
|
| 1352 |
+
"loss": 1.3747,
|
| 1353 |
+
"step": 19200
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 296.9230769230769,
|
| 1357 |
+
"grad_norm": 64.86967468261719,
|
| 1358 |
+
"learning_rate": 6.143e-06,
|
| 1359 |
+
"loss": 1.4014,
|
| 1360 |
+
"step": 19300
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 298.46153846153845,
|
| 1364 |
+
"grad_norm": 96.05943298339844,
|
| 1365 |
+
"learning_rate": 6.123e-06,
|
| 1366 |
+
"loss": 1.3517,
|
| 1367 |
+
"step": 19400
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 300.0,
|
| 1371 |
+
"grad_norm": 37.66103744506836,
|
| 1372 |
+
"learning_rate": 6.103e-06,
|
| 1373 |
+
"loss": 1.3625,
|
| 1374 |
+
"step": 19500
|
| 1375 |
+
}
|
| 1376 |
+
],
|
| 1377 |
+
"logging_steps": 100,
|
| 1378 |
+
"max_steps": 50000,
|
| 1379 |
+
"num_input_tokens_seen": 0,
|
| 1380 |
+
"num_train_epochs": 770,
|
| 1381 |
+
"save_steps": 500,
|
| 1382 |
+
"stateful_callbacks": {
|
| 1383 |
+
"TrainerControl": {
|
| 1384 |
+
"args": {
|
| 1385 |
+
"should_epoch_stop": false,
|
| 1386 |
+
"should_evaluate": false,
|
| 1387 |
+
"should_log": false,
|
| 1388 |
+
"should_save": true,
|
| 1389 |
+
"should_training_stop": false
|
| 1390 |
+
},
|
| 1391 |
+
"attributes": {}
|
| 1392 |
+
}
|
| 1393 |
+
},
|
| 1394 |
+
"total_flos": 7.453741741056e+19,
|
| 1395 |
+
"train_batch_size": 8,
|
| 1396 |
+
"trial_name": null,
|
| 1397 |
+
"trial_params": null
|
| 1398 |
+
}
|
checkpoint-19500/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ee77453dae7076acf919eb7bcd8ccb5a082573a80b4f2c85a6a7d7ad1edcbedf
|
| 3 |
+
size 5240
|