Training in progress, step 18000, checkpoint
Browse files- checkpoint-18000/config.json +62 -0
- checkpoint-18000/model.safetensors +3 -0
- checkpoint-18000/optimizer.pt +3 -0
- checkpoint-18000/preprocessor_config.json +26 -0
- checkpoint-18000/rng_state.pth +3 -0
- checkpoint-18000/scheduler.pt +3 -0
- checkpoint-18000/trainer_state.json +1293 -0
- checkpoint-18000/training_args.bin +3 -0
checkpoint-18000/config.json
ADDED
|
@@ -0,0 +1,62 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "hwm21/230824-detr-segment-detect-unmasked",
|
| 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-18000/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9982aaada308d73dd3959211be72a1461fd5a41cd3852683cd7511ebb35cdb3d
|
| 3 |
+
size 166904424
|
checkpoint-18000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a3eefa8ea1761c6e5a4810949139b6864399efa8aae3cd976536918d2536812b
|
| 3 |
+
size 331310970
|
checkpoint-18000/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-18000/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:31a4f5340660dc3dfa10268247f323b44c88d2437c64b03caf593f63094ab657
|
| 3 |
+
size 14244
|
checkpoint-18000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:74f137215ffa9568091192a428944b7d4e25740a18756aa22d97fc285ed65bec
|
| 3 |
+
size 1064
|
checkpoint-18000/trainer_state.json
ADDED
|
@@ -0,0 +1,1293 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 276.9230769230769,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 18000,
|
| 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": 28.85655403137207,
|
| 14 |
+
"learning_rate": 9.9812e-06,
|
| 15 |
+
"loss": 1.0794,
|
| 16 |
+
"step": 100
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 3.076923076923077,
|
| 20 |
+
"grad_norm": 40.480037689208984,
|
| 21 |
+
"learning_rate": 9.961400000000001e-06,
|
| 22 |
+
"loss": 1.1026,
|
| 23 |
+
"step": 200
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 4.615384615384615,
|
| 27 |
+
"grad_norm": 22.348379135131836,
|
| 28 |
+
"learning_rate": 9.941600000000002e-06,
|
| 29 |
+
"loss": 1.1211,
|
| 30 |
+
"step": 300
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 6.153846153846154,
|
| 34 |
+
"grad_norm": 49.462806701660156,
|
| 35 |
+
"learning_rate": 9.921600000000002e-06,
|
| 36 |
+
"loss": 1.1217,
|
| 37 |
+
"step": 400
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 7.6923076923076925,
|
| 41 |
+
"grad_norm": 37.26523208618164,
|
| 42 |
+
"learning_rate": 9.901600000000002e-06,
|
| 43 |
+
"loss": 1.0993,
|
| 44 |
+
"step": 500
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 9.23076923076923,
|
| 48 |
+
"grad_norm": 265.8311767578125,
|
| 49 |
+
"learning_rate": 9.8816e-06,
|
| 50 |
+
"loss": 1.1576,
|
| 51 |
+
"step": 600
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 10.76923076923077,
|
| 55 |
+
"grad_norm": 48.439048767089844,
|
| 56 |
+
"learning_rate": 9.8616e-06,
|
| 57 |
+
"loss": 1.1029,
|
| 58 |
+
"step": 700
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 12.307692307692308,
|
| 62 |
+
"grad_norm": 200.80213928222656,
|
| 63 |
+
"learning_rate": 9.8416e-06,
|
| 64 |
+
"loss": 1.1375,
|
| 65 |
+
"step": 800
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 13.846153846153847,
|
| 69 |
+
"grad_norm": 54.096126556396484,
|
| 70 |
+
"learning_rate": 9.821600000000001e-06,
|
| 71 |
+
"loss": 1.1134,
|
| 72 |
+
"step": 900
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 15.384615384615385,
|
| 76 |
+
"grad_norm": 72.55537414550781,
|
| 77 |
+
"learning_rate": 9.801600000000001e-06,
|
| 78 |
+
"loss": 1.1364,
|
| 79 |
+
"step": 1000
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 16.923076923076923,
|
| 83 |
+
"grad_norm": 45.86960983276367,
|
| 84 |
+
"learning_rate": 9.781600000000001e-06,
|
| 85 |
+
"loss": 1.1259,
|
| 86 |
+
"step": 1100
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 18.46153846153846,
|
| 90 |
+
"grad_norm": 44.21461868286133,
|
| 91 |
+
"learning_rate": 9.761600000000002e-06,
|
| 92 |
+
"loss": 1.1144,
|
| 93 |
+
"step": 1200
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 20.0,
|
| 97 |
+
"grad_norm": 49.014503479003906,
|
| 98 |
+
"learning_rate": 9.741600000000002e-06,
|
| 99 |
+
"loss": 1.1151,
|
| 100 |
+
"step": 1300
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 21.53846153846154,
|
| 104 |
+
"grad_norm": 62.7434196472168,
|
| 105 |
+
"learning_rate": 9.7216e-06,
|
| 106 |
+
"loss": 1.1129,
|
| 107 |
+
"step": 1400
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 23.076923076923077,
|
| 111 |
+
"grad_norm": 48.607032775878906,
|
| 112 |
+
"learning_rate": 9.7016e-06,
|
| 113 |
+
"loss": 1.1189,
|
| 114 |
+
"step": 1500
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 24.615384615384617,
|
| 118 |
+
"grad_norm": 43.288368225097656,
|
| 119 |
+
"learning_rate": 9.6816e-06,
|
| 120 |
+
"loss": 1.102,
|
| 121 |
+
"step": 1600
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 26.153846153846153,
|
| 125 |
+
"grad_norm": 33.83266067504883,
|
| 126 |
+
"learning_rate": 9.6616e-06,
|
| 127 |
+
"loss": 1.1091,
|
| 128 |
+
"step": 1700
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 27.692307692307693,
|
| 132 |
+
"grad_norm": 54.347755432128906,
|
| 133 |
+
"learning_rate": 9.641600000000001e-06,
|
| 134 |
+
"loss": 1.1478,
|
| 135 |
+
"step": 1800
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 29.23076923076923,
|
| 139 |
+
"grad_norm": 68.06307220458984,
|
| 140 |
+
"learning_rate": 9.621600000000001e-06,
|
| 141 |
+
"loss": 1.1384,
|
| 142 |
+
"step": 1900
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 30.76923076923077,
|
| 146 |
+
"grad_norm": 24.473976135253906,
|
| 147 |
+
"learning_rate": 9.601600000000001e-06,
|
| 148 |
+
"loss": 1.1218,
|
| 149 |
+
"step": 2000
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 32.30769230769231,
|
| 153 |
+
"grad_norm": 33.08134460449219,
|
| 154 |
+
"learning_rate": 9.5818e-06,
|
| 155 |
+
"loss": 1.0807,
|
| 156 |
+
"step": 2100
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 33.84615384615385,
|
| 160 |
+
"grad_norm": 32.798397064208984,
|
| 161 |
+
"learning_rate": 9.5618e-06,
|
| 162 |
+
"loss": 1.1548,
|
| 163 |
+
"step": 2200
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 35.38461538461539,
|
| 167 |
+
"grad_norm": 49.251651763916016,
|
| 168 |
+
"learning_rate": 9.5418e-06,
|
| 169 |
+
"loss": 1.1056,
|
| 170 |
+
"step": 2300
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 36.92307692307692,
|
| 174 |
+
"grad_norm": 47.84526062011719,
|
| 175 |
+
"learning_rate": 9.521800000000001e-06,
|
| 176 |
+
"loss": 1.169,
|
| 177 |
+
"step": 2400
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 38.46153846153846,
|
| 181 |
+
"grad_norm": 66.61646270751953,
|
| 182 |
+
"learning_rate": 9.501800000000001e-06,
|
| 183 |
+
"loss": 1.1505,
|
| 184 |
+
"step": 2500
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 40.0,
|
| 188 |
+
"grad_norm": 84.58892059326172,
|
| 189 |
+
"learning_rate": 9.481800000000001e-06,
|
| 190 |
+
"loss": 1.1117,
|
| 191 |
+
"step": 2600
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 41.53846153846154,
|
| 195 |
+
"grad_norm": 86.26738739013672,
|
| 196 |
+
"learning_rate": 9.461800000000001e-06,
|
| 197 |
+
"loss": 1.0914,
|
| 198 |
+
"step": 2700
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 43.07692307692308,
|
| 202 |
+
"grad_norm": 106.57470703125,
|
| 203 |
+
"learning_rate": 9.441800000000002e-06,
|
| 204 |
+
"loss": 1.1232,
|
| 205 |
+
"step": 2800
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 44.61538461538461,
|
| 209 |
+
"grad_norm": 133.2903289794922,
|
| 210 |
+
"learning_rate": 9.4218e-06,
|
| 211 |
+
"loss": 1.1544,
|
| 212 |
+
"step": 2900
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 46.15384615384615,
|
| 216 |
+
"grad_norm": 75.39508819580078,
|
| 217 |
+
"learning_rate": 9.4018e-06,
|
| 218 |
+
"loss": 1.1699,
|
| 219 |
+
"step": 3000
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 47.69230769230769,
|
| 223 |
+
"grad_norm": 45.91976547241211,
|
| 224 |
+
"learning_rate": 9.3818e-06,
|
| 225 |
+
"loss": 1.1652,
|
| 226 |
+
"step": 3100
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 49.23076923076923,
|
| 230 |
+
"grad_norm": 37.38508605957031,
|
| 231 |
+
"learning_rate": 9.3618e-06,
|
| 232 |
+
"loss": 1.1386,
|
| 233 |
+
"step": 3200
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 50.76923076923077,
|
| 237 |
+
"grad_norm": 106.42298126220703,
|
| 238 |
+
"learning_rate": 9.341800000000001e-06,
|
| 239 |
+
"loss": 1.0903,
|
| 240 |
+
"step": 3300
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 52.30769230769231,
|
| 244 |
+
"grad_norm": 59.256103515625,
|
| 245 |
+
"learning_rate": 9.321800000000001e-06,
|
| 246 |
+
"loss": 1.1266,
|
| 247 |
+
"step": 3400
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 53.84615384615385,
|
| 251 |
+
"grad_norm": 61.493003845214844,
|
| 252 |
+
"learning_rate": 9.301800000000001e-06,
|
| 253 |
+
"loss": 1.1047,
|
| 254 |
+
"step": 3500
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 55.38461538461539,
|
| 258 |
+
"grad_norm": 149.27308654785156,
|
| 259 |
+
"learning_rate": 9.281800000000001e-06,
|
| 260 |
+
"loss": 1.1132,
|
| 261 |
+
"step": 3600
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 56.92307692307692,
|
| 265 |
+
"grad_norm": 81.58853912353516,
|
| 266 |
+
"learning_rate": 9.2618e-06,
|
| 267 |
+
"loss": 1.083,
|
| 268 |
+
"step": 3700
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 58.46153846153846,
|
| 272 |
+
"grad_norm": 39.950843811035156,
|
| 273 |
+
"learning_rate": 9.2418e-06,
|
| 274 |
+
"loss": 1.09,
|
| 275 |
+
"step": 3800
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 60.0,
|
| 279 |
+
"grad_norm": 66.99285888671875,
|
| 280 |
+
"learning_rate": 9.2218e-06,
|
| 281 |
+
"loss": 1.1161,
|
| 282 |
+
"step": 3900
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 61.53846153846154,
|
| 286 |
+
"grad_norm": 28.769887924194336,
|
| 287 |
+
"learning_rate": 9.2018e-06,
|
| 288 |
+
"loss": 1.1061,
|
| 289 |
+
"step": 4000
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 63.07692307692308,
|
| 293 |
+
"grad_norm": 37.438724517822266,
|
| 294 |
+
"learning_rate": 9.1818e-06,
|
| 295 |
+
"loss": 1.0812,
|
| 296 |
+
"step": 4100
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 64.61538461538461,
|
| 300 |
+
"grad_norm": 44.86701965332031,
|
| 301 |
+
"learning_rate": 9.161800000000001e-06,
|
| 302 |
+
"loss": 1.0923,
|
| 303 |
+
"step": 4200
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 66.15384615384616,
|
| 307 |
+
"grad_norm": 64.42572021484375,
|
| 308 |
+
"learning_rate": 9.142000000000002e-06,
|
| 309 |
+
"loss": 1.0781,
|
| 310 |
+
"step": 4300
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 67.6923076923077,
|
| 314 |
+
"grad_norm": 51.16594696044922,
|
| 315 |
+
"learning_rate": 9.122e-06,
|
| 316 |
+
"loss": 1.0446,
|
| 317 |
+
"step": 4400
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 69.23076923076923,
|
| 321 |
+
"grad_norm": 39.816341400146484,
|
| 322 |
+
"learning_rate": 9.102e-06,
|
| 323 |
+
"loss": 1.1083,
|
| 324 |
+
"step": 4500
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 70.76923076923077,
|
| 328 |
+
"grad_norm": 57.10240173339844,
|
| 329 |
+
"learning_rate": 9.082e-06,
|
| 330 |
+
"loss": 1.0748,
|
| 331 |
+
"step": 4600
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 72.3076923076923,
|
| 335 |
+
"grad_norm": 98.26517486572266,
|
| 336 |
+
"learning_rate": 9.062e-06,
|
| 337 |
+
"loss": 1.0624,
|
| 338 |
+
"step": 4700
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 73.84615384615384,
|
| 342 |
+
"grad_norm": 23.59007453918457,
|
| 343 |
+
"learning_rate": 9.042e-06,
|
| 344 |
+
"loss": 1.0644,
|
| 345 |
+
"step": 4800
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 75.38461538461539,
|
| 349 |
+
"grad_norm": 34.98348617553711,
|
| 350 |
+
"learning_rate": 9.022000000000001e-06,
|
| 351 |
+
"loss": 1.0859,
|
| 352 |
+
"step": 4900
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 76.92307692307692,
|
| 356 |
+
"grad_norm": 65.94945526123047,
|
| 357 |
+
"learning_rate": 9.002000000000001e-06,
|
| 358 |
+
"loss": 1.0603,
|
| 359 |
+
"step": 5000
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 78.46153846153847,
|
| 363 |
+
"grad_norm": 49.95417404174805,
|
| 364 |
+
"learning_rate": 8.982000000000001e-06,
|
| 365 |
+
"loss": 1.0915,
|
| 366 |
+
"step": 5100
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 80.0,
|
| 370 |
+
"grad_norm": 57.85556411743164,
|
| 371 |
+
"learning_rate": 8.962e-06,
|
| 372 |
+
"loss": 1.0787,
|
| 373 |
+
"step": 5200
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 81.53846153846153,
|
| 377 |
+
"grad_norm": 26.018638610839844,
|
| 378 |
+
"learning_rate": 8.942e-06,
|
| 379 |
+
"loss": 1.0632,
|
| 380 |
+
"step": 5300
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 83.07692307692308,
|
| 384 |
+
"grad_norm": 58.61573028564453,
|
| 385 |
+
"learning_rate": 8.922e-06,
|
| 386 |
+
"loss": 1.075,
|
| 387 |
+
"step": 5400
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 84.61538461538461,
|
| 391 |
+
"grad_norm": 122.04191589355469,
|
| 392 |
+
"learning_rate": 8.902e-06,
|
| 393 |
+
"loss": 1.0724,
|
| 394 |
+
"step": 5500
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 86.15384615384616,
|
| 398 |
+
"grad_norm": 78.37217712402344,
|
| 399 |
+
"learning_rate": 8.882e-06,
|
| 400 |
+
"loss": 1.0925,
|
| 401 |
+
"step": 5600
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 87.6923076923077,
|
| 405 |
+
"grad_norm": 54.941184997558594,
|
| 406 |
+
"learning_rate": 8.862000000000001e-06,
|
| 407 |
+
"loss": 1.1195,
|
| 408 |
+
"step": 5700
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 89.23076923076923,
|
| 412 |
+
"grad_norm": 61.23499298095703,
|
| 413 |
+
"learning_rate": 8.842000000000001e-06,
|
| 414 |
+
"loss": 1.0685,
|
| 415 |
+
"step": 5800
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 90.76923076923077,
|
| 419 |
+
"grad_norm": 89.12802124023438,
|
| 420 |
+
"learning_rate": 8.822000000000001e-06,
|
| 421 |
+
"loss": 1.0899,
|
| 422 |
+
"step": 5900
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 92.3076923076923,
|
| 426 |
+
"grad_norm": 209.87484741210938,
|
| 427 |
+
"learning_rate": 8.802e-06,
|
| 428 |
+
"loss": 1.0771,
|
| 429 |
+
"step": 6000
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 93.84615384615384,
|
| 433 |
+
"grad_norm": 55.978092193603516,
|
| 434 |
+
"learning_rate": 8.782e-06,
|
| 435 |
+
"loss": 1.0657,
|
| 436 |
+
"step": 6100
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 95.38461538461539,
|
| 440 |
+
"grad_norm": 76.39286041259766,
|
| 441 |
+
"learning_rate": 8.762e-06,
|
| 442 |
+
"loss": 1.0653,
|
| 443 |
+
"step": 6200
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 96.92307692307692,
|
| 447 |
+
"grad_norm": 26.763763427734375,
|
| 448 |
+
"learning_rate": 8.742e-06,
|
| 449 |
+
"loss": 1.0844,
|
| 450 |
+
"step": 6300
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 98.46153846153847,
|
| 454 |
+
"grad_norm": 37.72929382324219,
|
| 455 |
+
"learning_rate": 8.722e-06,
|
| 456 |
+
"loss": 1.0753,
|
| 457 |
+
"step": 6400
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 100.0,
|
| 461 |
+
"grad_norm": 33.482521057128906,
|
| 462 |
+
"learning_rate": 8.702e-06,
|
| 463 |
+
"loss": 1.065,
|
| 464 |
+
"step": 6500
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 101.53846153846153,
|
| 468 |
+
"grad_norm": 38.44742965698242,
|
| 469 |
+
"learning_rate": 8.682000000000001e-06,
|
| 470 |
+
"loss": 1.0583,
|
| 471 |
+
"step": 6600
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 103.07692307692308,
|
| 475 |
+
"grad_norm": 44.009483337402344,
|
| 476 |
+
"learning_rate": 8.662000000000001e-06,
|
| 477 |
+
"loss": 1.0613,
|
| 478 |
+
"step": 6700
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 104.61538461538461,
|
| 482 |
+
"grad_norm": 53.22964096069336,
|
| 483 |
+
"learning_rate": 8.642e-06,
|
| 484 |
+
"loss": 1.079,
|
| 485 |
+
"step": 6800
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 106.15384615384616,
|
| 489 |
+
"grad_norm": 32.453739166259766,
|
| 490 |
+
"learning_rate": 8.622e-06,
|
| 491 |
+
"loss": 1.0564,
|
| 492 |
+
"step": 6900
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 107.6923076923077,
|
| 496 |
+
"grad_norm": 56.385257720947266,
|
| 497 |
+
"learning_rate": 8.602e-06,
|
| 498 |
+
"loss": 1.0638,
|
| 499 |
+
"step": 7000
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 109.23076923076923,
|
| 503 |
+
"grad_norm": 75.76576232910156,
|
| 504 |
+
"learning_rate": 8.582e-06,
|
| 505 |
+
"loss": 1.0451,
|
| 506 |
+
"step": 7100
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 110.76923076923077,
|
| 510 |
+
"grad_norm": 43.19460678100586,
|
| 511 |
+
"learning_rate": 8.562e-06,
|
| 512 |
+
"loss": 1.0651,
|
| 513 |
+
"step": 7200
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 112.3076923076923,
|
| 517 |
+
"grad_norm": 130.46304321289062,
|
| 518 |
+
"learning_rate": 8.542e-06,
|
| 519 |
+
"loss": 1.0489,
|
| 520 |
+
"step": 7300
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 113.84615384615384,
|
| 524 |
+
"grad_norm": 81.6989974975586,
|
| 525 |
+
"learning_rate": 8.522e-06,
|
| 526 |
+
"loss": 1.0275,
|
| 527 |
+
"step": 7400
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 115.38461538461539,
|
| 531 |
+
"grad_norm": 125.24699401855469,
|
| 532 |
+
"learning_rate": 8.502000000000001e-06,
|
| 533 |
+
"loss": 1.0498,
|
| 534 |
+
"step": 7500
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 116.92307692307692,
|
| 538 |
+
"grad_norm": 54.86334228515625,
|
| 539 |
+
"learning_rate": 8.482e-06,
|
| 540 |
+
"loss": 1.0327,
|
| 541 |
+
"step": 7600
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 118.46153846153847,
|
| 545 |
+
"grad_norm": 42.080753326416016,
|
| 546 |
+
"learning_rate": 8.462e-06,
|
| 547 |
+
"loss": 1.0425,
|
| 548 |
+
"step": 7700
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 120.0,
|
| 552 |
+
"grad_norm": 73.9937973022461,
|
| 553 |
+
"learning_rate": 8.442e-06,
|
| 554 |
+
"loss": 1.0433,
|
| 555 |
+
"step": 7800
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 121.53846153846153,
|
| 559 |
+
"grad_norm": 107.1471939086914,
|
| 560 |
+
"learning_rate": 8.422e-06,
|
| 561 |
+
"loss": 1.0573,
|
| 562 |
+
"step": 7900
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 123.07692307692308,
|
| 566 |
+
"grad_norm": 239.5729217529297,
|
| 567 |
+
"learning_rate": 8.402e-06,
|
| 568 |
+
"loss": 1.0388,
|
| 569 |
+
"step": 8000
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 124.61538461538461,
|
| 573 |
+
"grad_norm": 39.29356384277344,
|
| 574 |
+
"learning_rate": 8.382e-06,
|
| 575 |
+
"loss": 1.0647,
|
| 576 |
+
"step": 8100
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 126.15384615384616,
|
| 580 |
+
"grad_norm": 146.09300231933594,
|
| 581 |
+
"learning_rate": 8.362e-06,
|
| 582 |
+
"loss": 1.0166,
|
| 583 |
+
"step": 8200
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 127.6923076923077,
|
| 587 |
+
"grad_norm": 90.30326080322266,
|
| 588 |
+
"learning_rate": 8.342e-06,
|
| 589 |
+
"loss": 1.0677,
|
| 590 |
+
"step": 8300
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 129.23076923076923,
|
| 594 |
+
"grad_norm": 20.28913116455078,
|
| 595 |
+
"learning_rate": 8.322000000000001e-06,
|
| 596 |
+
"loss": 1.0504,
|
| 597 |
+
"step": 8400
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 130.76923076923077,
|
| 601 |
+
"grad_norm": 44.32182312011719,
|
| 602 |
+
"learning_rate": 8.302000000000001e-06,
|
| 603 |
+
"loss": 1.0628,
|
| 604 |
+
"step": 8500
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 132.30769230769232,
|
| 608 |
+
"grad_norm": 54.270172119140625,
|
| 609 |
+
"learning_rate": 8.282000000000001e-06,
|
| 610 |
+
"loss": 1.0361,
|
| 611 |
+
"step": 8600
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 133.84615384615384,
|
| 615 |
+
"grad_norm": 66.50403594970703,
|
| 616 |
+
"learning_rate": 8.262000000000002e-06,
|
| 617 |
+
"loss": 1.029,
|
| 618 |
+
"step": 8700
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 135.3846153846154,
|
| 622 |
+
"grad_norm": 36.06448745727539,
|
| 623 |
+
"learning_rate": 8.242000000000002e-06,
|
| 624 |
+
"loss": 1.0412,
|
| 625 |
+
"step": 8800
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 136.92307692307693,
|
| 629 |
+
"grad_norm": 61.25200653076172,
|
| 630 |
+
"learning_rate": 8.222000000000002e-06,
|
| 631 |
+
"loss": 1.0436,
|
| 632 |
+
"step": 8900
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 138.46153846153845,
|
| 636 |
+
"grad_norm": 75.82115936279297,
|
| 637 |
+
"learning_rate": 8.202e-06,
|
| 638 |
+
"loss": 1.0203,
|
| 639 |
+
"step": 9000
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 140.0,
|
| 643 |
+
"grad_norm": 82.14342498779297,
|
| 644 |
+
"learning_rate": 8.182e-06,
|
| 645 |
+
"loss": 1.0414,
|
| 646 |
+
"step": 9100
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 141.53846153846155,
|
| 650 |
+
"grad_norm": 40.55479431152344,
|
| 651 |
+
"learning_rate": 8.162e-06,
|
| 652 |
+
"loss": 1.0394,
|
| 653 |
+
"step": 9200
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 143.07692307692307,
|
| 657 |
+
"grad_norm": 62.4966926574707,
|
| 658 |
+
"learning_rate": 8.142000000000001e-06,
|
| 659 |
+
"loss": 1.0621,
|
| 660 |
+
"step": 9300
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 144.6153846153846,
|
| 664 |
+
"grad_norm": 40.539161682128906,
|
| 665 |
+
"learning_rate": 8.122000000000001e-06,
|
| 666 |
+
"loss": 1.0346,
|
| 667 |
+
"step": 9400
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 146.15384615384616,
|
| 671 |
+
"grad_norm": 65.81285858154297,
|
| 672 |
+
"learning_rate": 8.102000000000001e-06,
|
| 673 |
+
"loss": 1.0561,
|
| 674 |
+
"step": 9500
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 147.69230769230768,
|
| 678 |
+
"grad_norm": 64.11259460449219,
|
| 679 |
+
"learning_rate": 8.082000000000002e-06,
|
| 680 |
+
"loss": 1.0203,
|
| 681 |
+
"step": 9600
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 149.23076923076923,
|
| 685 |
+
"grad_norm": 68.70179748535156,
|
| 686 |
+
"learning_rate": 8.062000000000002e-06,
|
| 687 |
+
"loss": 1.0413,
|
| 688 |
+
"step": 9700
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 150.76923076923077,
|
| 692 |
+
"grad_norm": 49.53428268432617,
|
| 693 |
+
"learning_rate": 8.0422e-06,
|
| 694 |
+
"loss": 1.0199,
|
| 695 |
+
"step": 9800
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 152.30769230769232,
|
| 699 |
+
"grad_norm": 33.95280075073242,
|
| 700 |
+
"learning_rate": 8.022200000000001e-06,
|
| 701 |
+
"loss": 1.0278,
|
| 702 |
+
"step": 9900
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 153.84615384615384,
|
| 706 |
+
"grad_norm": 75.1406021118164,
|
| 707 |
+
"learning_rate": 8.002200000000001e-06,
|
| 708 |
+
"loss": 1.0093,
|
| 709 |
+
"step": 10000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 155.3846153846154,
|
| 713 |
+
"grad_norm": 33.94615173339844,
|
| 714 |
+
"learning_rate": 7.982200000000001e-06,
|
| 715 |
+
"loss": 1.0431,
|
| 716 |
+
"step": 10100
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 156.92307692307693,
|
| 720 |
+
"grad_norm": 40.253597259521484,
|
| 721 |
+
"learning_rate": 7.962200000000001e-06,
|
| 722 |
+
"loss": 0.9965,
|
| 723 |
+
"step": 10200
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 158.46153846153845,
|
| 727 |
+
"grad_norm": 45.14350128173828,
|
| 728 |
+
"learning_rate": 7.942200000000002e-06,
|
| 729 |
+
"loss": 1.0309,
|
| 730 |
+
"step": 10300
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 160.0,
|
| 734 |
+
"grad_norm": 21.002464294433594,
|
| 735 |
+
"learning_rate": 7.922200000000002e-06,
|
| 736 |
+
"loss": 1.0016,
|
| 737 |
+
"step": 10400
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 161.53846153846155,
|
| 741 |
+
"grad_norm": 35.12867736816406,
|
| 742 |
+
"learning_rate": 7.9022e-06,
|
| 743 |
+
"loss": 1.0121,
|
| 744 |
+
"step": 10500
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 163.07692307692307,
|
| 748 |
+
"grad_norm": 79.60009002685547,
|
| 749 |
+
"learning_rate": 7.8822e-06,
|
| 750 |
+
"loss": 1.0231,
|
| 751 |
+
"step": 10600
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 164.6153846153846,
|
| 755 |
+
"grad_norm": 226.0706787109375,
|
| 756 |
+
"learning_rate": 7.8622e-06,
|
| 757 |
+
"loss": 1.0305,
|
| 758 |
+
"step": 10700
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 166.15384615384616,
|
| 762 |
+
"grad_norm": 23.051137924194336,
|
| 763 |
+
"learning_rate": 7.842200000000001e-06,
|
| 764 |
+
"loss": 0.9855,
|
| 765 |
+
"step": 10800
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 167.69230769230768,
|
| 769 |
+
"grad_norm": 45.090057373046875,
|
| 770 |
+
"learning_rate": 7.822200000000001e-06,
|
| 771 |
+
"loss": 1.0162,
|
| 772 |
+
"step": 10900
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 169.23076923076923,
|
| 776 |
+
"grad_norm": 30.125905990600586,
|
| 777 |
+
"learning_rate": 7.802200000000001e-06,
|
| 778 |
+
"loss": 1.0,
|
| 779 |
+
"step": 11000
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 170.76923076923077,
|
| 783 |
+
"grad_norm": 69.0078125,
|
| 784 |
+
"learning_rate": 7.782200000000001e-06,
|
| 785 |
+
"loss": 1.0363,
|
| 786 |
+
"step": 11100
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 172.30769230769232,
|
| 790 |
+
"grad_norm": 31.446842193603516,
|
| 791 |
+
"learning_rate": 7.762200000000002e-06,
|
| 792 |
+
"loss": 1.0092,
|
| 793 |
+
"step": 11200
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 173.84615384615384,
|
| 797 |
+
"grad_norm": 72.3807144165039,
|
| 798 |
+
"learning_rate": 7.7422e-06,
|
| 799 |
+
"loss": 1.0052,
|
| 800 |
+
"step": 11300
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 175.3846153846154,
|
| 804 |
+
"grad_norm": 42.25762176513672,
|
| 805 |
+
"learning_rate": 7.7222e-06,
|
| 806 |
+
"loss": 1.0183,
|
| 807 |
+
"step": 11400
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 176.92307692307693,
|
| 811 |
+
"grad_norm": 37.23869323730469,
|
| 812 |
+
"learning_rate": 7.7022e-06,
|
| 813 |
+
"loss": 1.0029,
|
| 814 |
+
"step": 11500
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 178.46153846153845,
|
| 818 |
+
"grad_norm": 123.58518981933594,
|
| 819 |
+
"learning_rate": 7.6822e-06,
|
| 820 |
+
"loss": 1.0043,
|
| 821 |
+
"step": 11600
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 180.0,
|
| 825 |
+
"grad_norm": 36.04173278808594,
|
| 826 |
+
"learning_rate": 7.662200000000001e-06,
|
| 827 |
+
"loss": 1.0249,
|
| 828 |
+
"step": 11700
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 181.53846153846155,
|
| 832 |
+
"grad_norm": 54.06993103027344,
|
| 833 |
+
"learning_rate": 7.642200000000001e-06,
|
| 834 |
+
"loss": 1.0203,
|
| 835 |
+
"step": 11800
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 183.07692307692307,
|
| 839 |
+
"grad_norm": 23.067537307739258,
|
| 840 |
+
"learning_rate": 7.6222000000000005e-06,
|
| 841 |
+
"loss": 1.011,
|
| 842 |
+
"step": 11900
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 184.6153846153846,
|
| 846 |
+
"grad_norm": 38.63265609741211,
|
| 847 |
+
"learning_rate": 7.602400000000001e-06,
|
| 848 |
+
"loss": 1.0129,
|
| 849 |
+
"step": 12000
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 186.15384615384616,
|
| 853 |
+
"grad_norm": 67.62531280517578,
|
| 854 |
+
"learning_rate": 7.5824000000000005e-06,
|
| 855 |
+
"loss": 0.9953,
|
| 856 |
+
"step": 12100
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 187.69230769230768,
|
| 860 |
+
"grad_norm": 60.271663665771484,
|
| 861 |
+
"learning_rate": 7.562400000000001e-06,
|
| 862 |
+
"loss": 0.9949,
|
| 863 |
+
"step": 12200
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 189.23076923076923,
|
| 867 |
+
"grad_norm": 85.98783111572266,
|
| 868 |
+
"learning_rate": 7.542400000000001e-06,
|
| 869 |
+
"loss": 1.008,
|
| 870 |
+
"step": 12300
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 190.76923076923077,
|
| 874 |
+
"grad_norm": 60.040924072265625,
|
| 875 |
+
"learning_rate": 7.522400000000001e-06,
|
| 876 |
+
"loss": 0.9893,
|
| 877 |
+
"step": 12400
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 192.30769230769232,
|
| 881 |
+
"grad_norm": 34.971290588378906,
|
| 882 |
+
"learning_rate": 7.5024e-06,
|
| 883 |
+
"loss": 0.999,
|
| 884 |
+
"step": 12500
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 193.84615384615384,
|
| 888 |
+
"grad_norm": 47.212989807128906,
|
| 889 |
+
"learning_rate": 7.4824000000000006e-06,
|
| 890 |
+
"loss": 0.993,
|
| 891 |
+
"step": 12600
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 195.3846153846154,
|
| 895 |
+
"grad_norm": 38.17967987060547,
|
| 896 |
+
"learning_rate": 7.462400000000001e-06,
|
| 897 |
+
"loss": 1.0284,
|
| 898 |
+
"step": 12700
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 196.92307692307693,
|
| 902 |
+
"grad_norm": 47.6786003112793,
|
| 903 |
+
"learning_rate": 7.442400000000001e-06,
|
| 904 |
+
"loss": 0.9636,
|
| 905 |
+
"step": 12800
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 198.46153846153845,
|
| 909 |
+
"grad_norm": 31.550769805908203,
|
| 910 |
+
"learning_rate": 7.422400000000001e-06,
|
| 911 |
+
"loss": 1.0082,
|
| 912 |
+
"step": 12900
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 200.0,
|
| 916 |
+
"grad_norm": 176.42813110351562,
|
| 917 |
+
"learning_rate": 7.4024000000000005e-06,
|
| 918 |
+
"loss": 0.9848,
|
| 919 |
+
"step": 13000
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 201.53846153846155,
|
| 923 |
+
"grad_norm": 26.10734748840332,
|
| 924 |
+
"learning_rate": 7.382600000000001e-06,
|
| 925 |
+
"loss": 1.0191,
|
| 926 |
+
"step": 13100
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 203.07692307692307,
|
| 930 |
+
"grad_norm": 28.24509048461914,
|
| 931 |
+
"learning_rate": 7.3626000000000005e-06,
|
| 932 |
+
"loss": 0.9799,
|
| 933 |
+
"step": 13200
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 204.6153846153846,
|
| 937 |
+
"grad_norm": 27.511566162109375,
|
| 938 |
+
"learning_rate": 7.342600000000001e-06,
|
| 939 |
+
"loss": 0.9935,
|
| 940 |
+
"step": 13300
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 206.15384615384616,
|
| 944 |
+
"grad_norm": 43.895633697509766,
|
| 945 |
+
"learning_rate": 7.322600000000001e-06,
|
| 946 |
+
"loss": 1.0101,
|
| 947 |
+
"step": 13400
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 207.69230769230768,
|
| 951 |
+
"grad_norm": 73.81085968017578,
|
| 952 |
+
"learning_rate": 7.302600000000001e-06,
|
| 953 |
+
"loss": 0.9559,
|
| 954 |
+
"step": 13500
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 209.23076923076923,
|
| 958 |
+
"grad_norm": 76.736083984375,
|
| 959 |
+
"learning_rate": 7.2826e-06,
|
| 960 |
+
"loss": 0.9831,
|
| 961 |
+
"step": 13600
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 210.76923076923077,
|
| 965 |
+
"grad_norm": 29.066478729248047,
|
| 966 |
+
"learning_rate": 7.2626000000000006e-06,
|
| 967 |
+
"loss": 0.9981,
|
| 968 |
+
"step": 13700
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 212.30769230769232,
|
| 972 |
+
"grad_norm": 34.84987258911133,
|
| 973 |
+
"learning_rate": 7.242600000000001e-06,
|
| 974 |
+
"loss": 1.0009,
|
| 975 |
+
"step": 13800
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 213.84615384615384,
|
| 979 |
+
"grad_norm": 145.1222686767578,
|
| 980 |
+
"learning_rate": 7.222600000000001e-06,
|
| 981 |
+
"loss": 1.0315,
|
| 982 |
+
"step": 13900
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 215.3846153846154,
|
| 986 |
+
"grad_norm": 96.69234466552734,
|
| 987 |
+
"learning_rate": 7.2026e-06,
|
| 988 |
+
"loss": 1.0134,
|
| 989 |
+
"step": 14000
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 216.92307692307693,
|
| 993 |
+
"grad_norm": 43.14187240600586,
|
| 994 |
+
"learning_rate": 7.1826000000000005e-06,
|
| 995 |
+
"loss": 0.9885,
|
| 996 |
+
"step": 14100
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 218.46153846153845,
|
| 1000 |
+
"grad_norm": 28.49744415283203,
|
| 1001 |
+
"learning_rate": 7.162600000000001e-06,
|
| 1002 |
+
"loss": 0.9996,
|
| 1003 |
+
"step": 14200
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 220.0,
|
| 1007 |
+
"grad_norm": 72.76812744140625,
|
| 1008 |
+
"learning_rate": 7.142600000000001e-06,
|
| 1009 |
+
"loss": 0.998,
|
| 1010 |
+
"step": 14300
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 221.53846153846155,
|
| 1014 |
+
"grad_norm": 71.88528442382812,
|
| 1015 |
+
"learning_rate": 7.1226e-06,
|
| 1016 |
+
"loss": 0.9691,
|
| 1017 |
+
"step": 14400
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 223.07692307692307,
|
| 1021 |
+
"grad_norm": 43.25485610961914,
|
| 1022 |
+
"learning_rate": 7.1026e-06,
|
| 1023 |
+
"loss": 0.9797,
|
| 1024 |
+
"step": 14500
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 224.6153846153846,
|
| 1028 |
+
"grad_norm": 41.34707260131836,
|
| 1029 |
+
"learning_rate": 7.082600000000001e-06,
|
| 1030 |
+
"loss": 0.9957,
|
| 1031 |
+
"step": 14600
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 226.15384615384616,
|
| 1035 |
+
"grad_norm": 44.02409744262695,
|
| 1036 |
+
"learning_rate": 7.062600000000001e-06,
|
| 1037 |
+
"loss": 0.9775,
|
| 1038 |
+
"step": 14700
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 227.69230769230768,
|
| 1042 |
+
"grad_norm": 41.718284606933594,
|
| 1043 |
+
"learning_rate": 7.0426e-06,
|
| 1044 |
+
"loss": 0.9835,
|
| 1045 |
+
"step": 14800
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 229.23076923076923,
|
| 1049 |
+
"grad_norm": 56.27934646606445,
|
| 1050 |
+
"learning_rate": 7.0226e-06,
|
| 1051 |
+
"loss": 1.0033,
|
| 1052 |
+
"step": 14900
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 230.76923076923077,
|
| 1056 |
+
"grad_norm": 45.04922866821289,
|
| 1057 |
+
"learning_rate": 7.0026000000000005e-06,
|
| 1058 |
+
"loss": 0.9841,
|
| 1059 |
+
"step": 15000
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 232.30769230769232,
|
| 1063 |
+
"grad_norm": 31.639402389526367,
|
| 1064 |
+
"learning_rate": 6.982600000000001e-06,
|
| 1065 |
+
"loss": 0.9965,
|
| 1066 |
+
"step": 15100
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 233.84615384615384,
|
| 1070 |
+
"grad_norm": 55.72714614868164,
|
| 1071 |
+
"learning_rate": 6.9626e-06,
|
| 1072 |
+
"loss": 0.985,
|
| 1073 |
+
"step": 15200
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 235.3846153846154,
|
| 1077 |
+
"grad_norm": 27.006019592285156,
|
| 1078 |
+
"learning_rate": 6.9426e-06,
|
| 1079 |
+
"loss": 0.9853,
|
| 1080 |
+
"step": 15300
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 236.92307692307693,
|
| 1084 |
+
"grad_norm": 44.53865432739258,
|
| 1085 |
+
"learning_rate": 6.9226000000000004e-06,
|
| 1086 |
+
"loss": 0.9502,
|
| 1087 |
+
"step": 15400
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 238.46153846153845,
|
| 1091 |
+
"grad_norm": 60.436180114746094,
|
| 1092 |
+
"learning_rate": 6.902600000000001e-06,
|
| 1093 |
+
"loss": 0.9849,
|
| 1094 |
+
"step": 15500
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 240.0,
|
| 1098 |
+
"grad_norm": 29.967527389526367,
|
| 1099 |
+
"learning_rate": 6.8826e-06,
|
| 1100 |
+
"loss": 0.9524,
|
| 1101 |
+
"step": 15600
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 241.53846153846155,
|
| 1105 |
+
"grad_norm": 60.00186538696289,
|
| 1106 |
+
"learning_rate": 6.8626e-06,
|
| 1107 |
+
"loss": 0.9521,
|
| 1108 |
+
"step": 15700
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 243.07692307692307,
|
| 1112 |
+
"grad_norm": 46.52066421508789,
|
| 1113 |
+
"learning_rate": 6.8426e-06,
|
| 1114 |
+
"loss": 0.9909,
|
| 1115 |
+
"step": 15800
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 244.6153846153846,
|
| 1119 |
+
"grad_norm": 27.765243530273438,
|
| 1120 |
+
"learning_rate": 6.8226000000000005e-06,
|
| 1121 |
+
"loss": 0.9546,
|
| 1122 |
+
"step": 15900
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 246.15384615384616,
|
| 1126 |
+
"grad_norm": 36.32876968383789,
|
| 1127 |
+
"learning_rate": 6.802600000000001e-06,
|
| 1128 |
+
"loss": 0.9812,
|
| 1129 |
+
"step": 16000
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 247.69230769230768,
|
| 1133 |
+
"grad_norm": 43.65727615356445,
|
| 1134 |
+
"learning_rate": 6.7826e-06,
|
| 1135 |
+
"loss": 0.9669,
|
| 1136 |
+
"step": 16100
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 249.23076923076923,
|
| 1140 |
+
"grad_norm": 37.54214096069336,
|
| 1141 |
+
"learning_rate": 6.7626e-06,
|
| 1142 |
+
"loss": 0.9502,
|
| 1143 |
+
"step": 16200
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 250.76923076923077,
|
| 1147 |
+
"grad_norm": 62.36497497558594,
|
| 1148 |
+
"learning_rate": 6.7426000000000005e-06,
|
| 1149 |
+
"loss": 0.9875,
|
| 1150 |
+
"step": 16300
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 252.30769230769232,
|
| 1154 |
+
"grad_norm": 76.69297790527344,
|
| 1155 |
+
"learning_rate": 6.722600000000001e-06,
|
| 1156 |
+
"loss": 0.9471,
|
| 1157 |
+
"step": 16400
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 253.84615384615384,
|
| 1161 |
+
"grad_norm": 48.03883743286133,
|
| 1162 |
+
"learning_rate": 6.7026e-06,
|
| 1163 |
+
"loss": 0.9804,
|
| 1164 |
+
"step": 16500
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 255.3846153846154,
|
| 1168 |
+
"grad_norm": 83.64213562011719,
|
| 1169 |
+
"learning_rate": 6.6826e-06,
|
| 1170 |
+
"loss": 0.9469,
|
| 1171 |
+
"step": 16600
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 256.9230769230769,
|
| 1175 |
+
"grad_norm": 32.0591926574707,
|
| 1176 |
+
"learning_rate": 6.6626e-06,
|
| 1177 |
+
"loss": 0.9718,
|
| 1178 |
+
"step": 16700
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 258.46153846153845,
|
| 1182 |
+
"grad_norm": 317.7740783691406,
|
| 1183 |
+
"learning_rate": 6.6426000000000006e-06,
|
| 1184 |
+
"loss": 0.9825,
|
| 1185 |
+
"step": 16800
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 260.0,
|
| 1189 |
+
"grad_norm": 34.885589599609375,
|
| 1190 |
+
"learning_rate": 6.6226e-06,
|
| 1191 |
+
"loss": 0.9286,
|
| 1192 |
+
"step": 16900
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 261.53846153846155,
|
| 1196 |
+
"grad_norm": 44.93685531616211,
|
| 1197 |
+
"learning_rate": 6.6026e-06,
|
| 1198 |
+
"loss": 0.9758,
|
| 1199 |
+
"step": 17000
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 263.0769230769231,
|
| 1203 |
+
"grad_norm": 62.70928192138672,
|
| 1204 |
+
"learning_rate": 6.5826e-06,
|
| 1205 |
+
"loss": 0.9254,
|
| 1206 |
+
"step": 17100
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 264.61538461538464,
|
| 1210 |
+
"grad_norm": 40.158775329589844,
|
| 1211 |
+
"learning_rate": 6.5626000000000005e-06,
|
| 1212 |
+
"loss": 0.9905,
|
| 1213 |
+
"step": 17200
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 266.15384615384613,
|
| 1217 |
+
"grad_norm": 43.55058670043945,
|
| 1218 |
+
"learning_rate": 6.5426e-06,
|
| 1219 |
+
"loss": 0.9529,
|
| 1220 |
+
"step": 17300
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 267.6923076923077,
|
| 1224 |
+
"grad_norm": 46.61787796020508,
|
| 1225 |
+
"learning_rate": 6.5226e-06,
|
| 1226 |
+
"loss": 0.9784,
|
| 1227 |
+
"step": 17400
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 269.2307692307692,
|
| 1231 |
+
"grad_norm": 55.41309356689453,
|
| 1232 |
+
"learning_rate": 6.5026e-06,
|
| 1233 |
+
"loss": 0.9268,
|
| 1234 |
+
"step": 17500
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 270.7692307692308,
|
| 1238 |
+
"grad_norm": 55.73357391357422,
|
| 1239 |
+
"learning_rate": 6.4826e-06,
|
| 1240 |
+
"loss": 0.9792,
|
| 1241 |
+
"step": 17600
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 272.3076923076923,
|
| 1245 |
+
"grad_norm": 20.340187072753906,
|
| 1246 |
+
"learning_rate": 6.4626e-06,
|
| 1247 |
+
"loss": 0.9295,
|
| 1248 |
+
"step": 17700
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 273.84615384615387,
|
| 1252 |
+
"grad_norm": 38.640380859375,
|
| 1253 |
+
"learning_rate": 6.442600000000001e-06,
|
| 1254 |
+
"loss": 0.9414,
|
| 1255 |
+
"step": 17800
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 275.38461538461536,
|
| 1259 |
+
"grad_norm": 65.87943267822266,
|
| 1260 |
+
"learning_rate": 6.422600000000001e-06,
|
| 1261 |
+
"loss": 0.9438,
|
| 1262 |
+
"step": 17900
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 276.9230769230769,
|
| 1266 |
+
"grad_norm": 24.6571102142334,
|
| 1267 |
+
"learning_rate": 6.402600000000001e-06,
|
| 1268 |
+
"loss": 0.9481,
|
| 1269 |
+
"step": 18000
|
| 1270 |
+
}
|
| 1271 |
+
],
|
| 1272 |
+
"logging_steps": 100,
|
| 1273 |
+
"max_steps": 50000,
|
| 1274 |
+
"num_input_tokens_seen": 0,
|
| 1275 |
+
"num_train_epochs": 770,
|
| 1276 |
+
"save_steps": 500,
|
| 1277 |
+
"stateful_callbacks": {
|
| 1278 |
+
"TrainerControl": {
|
| 1279 |
+
"args": {
|
| 1280 |
+
"should_epoch_stop": false,
|
| 1281 |
+
"should_evaluate": false,
|
| 1282 |
+
"should_log": false,
|
| 1283 |
+
"should_save": true,
|
| 1284 |
+
"should_training_stop": false
|
| 1285 |
+
},
|
| 1286 |
+
"attributes": {}
|
| 1287 |
+
}
|
| 1288 |
+
},
|
| 1289 |
+
"total_flos": 6.880376991744e+19,
|
| 1290 |
+
"train_batch_size": 8,
|
| 1291 |
+
"trial_name": null,
|
| 1292 |
+
"trial_params": null
|
| 1293 |
+
}
|
checkpoint-18000/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:02b350f6f7a17a22819550620918a491d9cf6859434196530c490daf6bb911f9
|
| 3 |
+
size 5240
|