NewVersion
Browse files- ConstructionBertV2/distilBertTrain_ConstructionManagement/checkpoint-90000/config.json +24 -0
- ConstructionBertV2/distilBertTrain_ConstructionManagement/checkpoint-90000/model.safetensors +3 -0
- ConstructionBertV2/distilBertTrain_ConstructionManagement/checkpoint-90000/optimizer.pt +3 -0
- ConstructionBertV2/distilBertTrain_ConstructionManagement/checkpoint-90000/rng_state.pth +3 -0
- ConstructionBertV2/distilBertTrain_ConstructionManagement/checkpoint-90000/scheduler.pt +3 -0
- ConstructionBertV2/distilBertTrain_ConstructionManagement/checkpoint-90000/trainer_state.json +1293 -0
- ConstructionBertV2/distilBertTrain_ConstructionManagement/checkpoint-90000/training_args.bin +3 -0
- ConstructionBertV2/distilBertTrain_ConstructionManagement/checkpoint-95691/config.json +24 -0
- ConstructionBertV2/distilBertTrain_ConstructionManagement/checkpoint-95691/model.safetensors +3 -0
- ConstructionBertV2/distilBertTrain_ConstructionManagement/checkpoint-95691/optimizer.pt +3 -0
- ConstructionBertV2/distilBertTrain_ConstructionManagement/checkpoint-95691/rng_state.pth +3 -0
- ConstructionBertV2/distilBertTrain_ConstructionManagement/checkpoint-95691/scheduler.pt +3 -0
- ConstructionBertV2/distilBertTrain_ConstructionManagement/checkpoint-95691/trainer_state.json +1370 -0
- ConstructionBertV2/distilBertTrain_ConstructionManagement/checkpoint-95691/training_args.bin +3 -0
- ConstructionBertV2/distilBert_ConstructionManagement/config.json +24 -0
- ConstructionBertV2/distilBert_ConstructionManagement/model.safetensors +3 -0
- ConstructionBertV2/distilBert_ConstructionManagement/training_args.bin +3 -0
ConstructionBertV2/distilBertTrain_ConstructionManagement/checkpoint-90000/config.json
ADDED
@@ -0,0 +1,24 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "distilbert-base-uncased",
|
3 |
+
"activation": "gelu",
|
4 |
+
"architectures": [
|
5 |
+
"DistilBertForMaskedLM"
|
6 |
+
],
|
7 |
+
"attention_dropout": 0.1,
|
8 |
+
"dim": 768,
|
9 |
+
"dropout": 0.1,
|
10 |
+
"hidden_dim": 3072,
|
11 |
+
"initializer_range": 0.02,
|
12 |
+
"max_position_embeddings": 512,
|
13 |
+
"model_type": "distilbert",
|
14 |
+
"n_heads": 12,
|
15 |
+
"n_layers": 6,
|
16 |
+
"pad_token_id": 0,
|
17 |
+
"qa_dropout": 0.1,
|
18 |
+
"seq_classif_dropout": 0.2,
|
19 |
+
"sinusoidal_pos_embds": false,
|
20 |
+
"tie_weights_": true,
|
21 |
+
"torch_dtype": "float32",
|
22 |
+
"transformers_version": "4.49.0",
|
23 |
+
"vocab_size": 30522
|
24 |
+
}
|
ConstructionBertV2/distilBertTrain_ConstructionManagement/checkpoint-90000/model.safetensors
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:e362cc080765894fe54394e4b02bb4d0bd31e732071f7815a57d098ed4dfbda6
|
3 |
+
size 267954768
|
ConstructionBertV2/distilBertTrain_ConstructionManagement/checkpoint-90000/optimizer.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:b0384e3f196c79e115973aff8f071df93595d825590ec502cb4ba54970abb825
|
3 |
+
size 535972683
|
ConstructionBertV2/distilBertTrain_ConstructionManagement/checkpoint-90000/rng_state.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:af4e8441ed26b367a486d420b8bc551d8b12466317d910d7c0508695d3833c28
|
3 |
+
size 14645
|
ConstructionBertV2/distilBertTrain_ConstructionManagement/checkpoint-90000/scheduler.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:12fbd96348bdeb3dd2b0bf4f63f44ed154c14a4a6ff40f5a0530cd37334d1aba
|
3 |
+
size 1465
|
ConstructionBertV2/distilBertTrain_ConstructionManagement/checkpoint-90000/trainer_state.json
ADDED
@@ -0,0 +1,1293 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": null,
|
3 |
+
"best_model_checkpoint": null,
|
4 |
+
"epoch": 2.82158196695614,
|
5 |
+
"eval_steps": 500,
|
6 |
+
"global_step": 90000,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 0.015675455371978557,
|
13 |
+
"grad_norm": 22.547378540039062,
|
14 |
+
"learning_rate": 4.973874241046702e-05,
|
15 |
+
"loss": 3.0259,
|
16 |
+
"step": 500
|
17 |
+
},
|
18 |
+
{
|
19 |
+
"epoch": 0.031350910743957114,
|
20 |
+
"grad_norm": 15.296708106994629,
|
21 |
+
"learning_rate": 4.947748482093405e-05,
|
22 |
+
"loss": 2.9667,
|
23 |
+
"step": 1000
|
24 |
+
},
|
25 |
+
{
|
26 |
+
"epoch": 0.04702636611593567,
|
27 |
+
"grad_norm": 22.362838745117188,
|
28 |
+
"learning_rate": 4.921622723140107e-05,
|
29 |
+
"loss": 2.9586,
|
30 |
+
"step": 1500
|
31 |
+
},
|
32 |
+
{
|
33 |
+
"epoch": 0.06270182148791423,
|
34 |
+
"grad_norm": 20.58307647705078,
|
35 |
+
"learning_rate": 4.89549696418681e-05,
|
36 |
+
"loss": 2.9069,
|
37 |
+
"step": 2000
|
38 |
+
},
|
39 |
+
{
|
40 |
+
"epoch": 0.07837727685989278,
|
41 |
+
"grad_norm": 14.068394660949707,
|
42 |
+
"learning_rate": 4.869371205233512e-05,
|
43 |
+
"loss": 2.8975,
|
44 |
+
"step": 2500
|
45 |
+
},
|
46 |
+
{
|
47 |
+
"epoch": 0.09405273223187134,
|
48 |
+
"grad_norm": 14.740192413330078,
|
49 |
+
"learning_rate": 4.843245446280215e-05,
|
50 |
+
"loss": 2.8747,
|
51 |
+
"step": 3000
|
52 |
+
},
|
53 |
+
{
|
54 |
+
"epoch": 0.10972818760384989,
|
55 |
+
"grad_norm": 19.214658737182617,
|
56 |
+
"learning_rate": 4.817119687326917e-05,
|
57 |
+
"loss": 2.9392,
|
58 |
+
"step": 3500
|
59 |
+
},
|
60 |
+
{
|
61 |
+
"epoch": 0.12540364297582846,
|
62 |
+
"grad_norm": 20.592220306396484,
|
63 |
+
"learning_rate": 4.79099392837362e-05,
|
64 |
+
"loss": 2.8518,
|
65 |
+
"step": 4000
|
66 |
+
},
|
67 |
+
{
|
68 |
+
"epoch": 0.141079098347807,
|
69 |
+
"grad_norm": 14.96511459350586,
|
70 |
+
"learning_rate": 4.764868169420322e-05,
|
71 |
+
"loss": 2.8684,
|
72 |
+
"step": 4500
|
73 |
+
},
|
74 |
+
{
|
75 |
+
"epoch": 0.15675455371978556,
|
76 |
+
"grad_norm": 20.952829360961914,
|
77 |
+
"learning_rate": 4.738742410467024e-05,
|
78 |
+
"loss": 2.7943,
|
79 |
+
"step": 5000
|
80 |
+
},
|
81 |
+
{
|
82 |
+
"epoch": 0.17243000909176412,
|
83 |
+
"grad_norm": 23.319133758544922,
|
84 |
+
"learning_rate": 4.7126166515137266e-05,
|
85 |
+
"loss": 2.8203,
|
86 |
+
"step": 5500
|
87 |
+
},
|
88 |
+
{
|
89 |
+
"epoch": 0.18810546446374268,
|
90 |
+
"grad_norm": 12.275296211242676,
|
91 |
+
"learning_rate": 4.686490892560429e-05,
|
92 |
+
"loss": 2.8257,
|
93 |
+
"step": 6000
|
94 |
+
},
|
95 |
+
{
|
96 |
+
"epoch": 0.20378091983572122,
|
97 |
+
"grad_norm": 24.349985122680664,
|
98 |
+
"learning_rate": 4.6603651336071315e-05,
|
99 |
+
"loss": 2.7974,
|
100 |
+
"step": 6500
|
101 |
+
},
|
102 |
+
{
|
103 |
+
"epoch": 0.21945637520769978,
|
104 |
+
"grad_norm": 20.776357650756836,
|
105 |
+
"learning_rate": 4.6342393746538335e-05,
|
106 |
+
"loss": 2.7539,
|
107 |
+
"step": 7000
|
108 |
+
},
|
109 |
+
{
|
110 |
+
"epoch": 0.23513183057967835,
|
111 |
+
"grad_norm": 23.954627990722656,
|
112 |
+
"learning_rate": 4.608113615700536e-05,
|
113 |
+
"loss": 2.7877,
|
114 |
+
"step": 7500
|
115 |
+
},
|
116 |
+
{
|
117 |
+
"epoch": 0.2508072859516569,
|
118 |
+
"grad_norm": 14.333002090454102,
|
119 |
+
"learning_rate": 4.5819878567472384e-05,
|
120 |
+
"loss": 2.7571,
|
121 |
+
"step": 8000
|
122 |
+
},
|
123 |
+
{
|
124 |
+
"epoch": 0.2664827413236355,
|
125 |
+
"grad_norm": 17.226673126220703,
|
126 |
+
"learning_rate": 4.555862097793941e-05,
|
127 |
+
"loss": 2.7535,
|
128 |
+
"step": 8500
|
129 |
+
},
|
130 |
+
{
|
131 |
+
"epoch": 0.282158196695614,
|
132 |
+
"grad_norm": 24.639087677001953,
|
133 |
+
"learning_rate": 4.529736338840643e-05,
|
134 |
+
"loss": 2.7562,
|
135 |
+
"step": 9000
|
136 |
+
},
|
137 |
+
{
|
138 |
+
"epoch": 0.29783365206759255,
|
139 |
+
"grad_norm": 15.692695617675781,
|
140 |
+
"learning_rate": 4.503610579887346e-05,
|
141 |
+
"loss": 2.7613,
|
142 |
+
"step": 9500
|
143 |
+
},
|
144 |
+
{
|
145 |
+
"epoch": 0.3135091074395711,
|
146 |
+
"grad_norm": 22.752601623535156,
|
147 |
+
"learning_rate": 4.477484820934048e-05,
|
148 |
+
"loss": 2.7316,
|
149 |
+
"step": 10000
|
150 |
+
},
|
151 |
+
{
|
152 |
+
"epoch": 0.3291845628115497,
|
153 |
+
"grad_norm": 21.343364715576172,
|
154 |
+
"learning_rate": 4.45135906198075e-05,
|
155 |
+
"loss": 2.6913,
|
156 |
+
"step": 10500
|
157 |
+
},
|
158 |
+
{
|
159 |
+
"epoch": 0.34486001818352824,
|
160 |
+
"grad_norm": 23.11554527282715,
|
161 |
+
"learning_rate": 4.4252333030274536e-05,
|
162 |
+
"loss": 2.7792,
|
163 |
+
"step": 11000
|
164 |
+
},
|
165 |
+
{
|
166 |
+
"epoch": 0.3605354735555068,
|
167 |
+
"grad_norm": 20.79396629333496,
|
168 |
+
"learning_rate": 4.399107544074156e-05,
|
169 |
+
"loss": 2.6721,
|
170 |
+
"step": 11500
|
171 |
+
},
|
172 |
+
{
|
173 |
+
"epoch": 0.37621092892748537,
|
174 |
+
"grad_norm": 18.559513092041016,
|
175 |
+
"learning_rate": 4.3729817851208585e-05,
|
176 |
+
"loss": 2.7126,
|
177 |
+
"step": 12000
|
178 |
+
},
|
179 |
+
{
|
180 |
+
"epoch": 0.3918863842994639,
|
181 |
+
"grad_norm": 16.206348419189453,
|
182 |
+
"learning_rate": 4.3468560261675606e-05,
|
183 |
+
"loss": 2.7313,
|
184 |
+
"step": 12500
|
185 |
+
},
|
186 |
+
{
|
187 |
+
"epoch": 0.40756183967144244,
|
188 |
+
"grad_norm": 16.35007667541504,
|
189 |
+
"learning_rate": 4.320730267214263e-05,
|
190 |
+
"loss": 2.755,
|
191 |
+
"step": 13000
|
192 |
+
},
|
193 |
+
{
|
194 |
+
"epoch": 0.423237295043421,
|
195 |
+
"grad_norm": 28.126436233520508,
|
196 |
+
"learning_rate": 4.2946045082609654e-05,
|
197 |
+
"loss": 2.714,
|
198 |
+
"step": 13500
|
199 |
+
},
|
200 |
+
{
|
201 |
+
"epoch": 0.43891275041539957,
|
202 |
+
"grad_norm": 16.460905075073242,
|
203 |
+
"learning_rate": 4.2684787493076675e-05,
|
204 |
+
"loss": 2.6553,
|
205 |
+
"step": 14000
|
206 |
+
},
|
207 |
+
{
|
208 |
+
"epoch": 0.45458820578737813,
|
209 |
+
"grad_norm": 18.516586303710938,
|
210 |
+
"learning_rate": 4.24235299035437e-05,
|
211 |
+
"loss": 2.6515,
|
212 |
+
"step": 14500
|
213 |
+
},
|
214 |
+
{
|
215 |
+
"epoch": 0.4702636611593567,
|
216 |
+
"grad_norm": 15.802982330322266,
|
217 |
+
"learning_rate": 4.2162272314010724e-05,
|
218 |
+
"loss": 2.7276,
|
219 |
+
"step": 15000
|
220 |
+
},
|
221 |
+
{
|
222 |
+
"epoch": 0.48593911653133526,
|
223 |
+
"grad_norm": 14.24789810180664,
|
224 |
+
"learning_rate": 4.190101472447775e-05,
|
225 |
+
"loss": 2.7027,
|
226 |
+
"step": 15500
|
227 |
+
},
|
228 |
+
{
|
229 |
+
"epoch": 0.5016145719033138,
|
230 |
+
"grad_norm": 15.48758316040039,
|
231 |
+
"learning_rate": 4.163975713494477e-05,
|
232 |
+
"loss": 2.6882,
|
233 |
+
"step": 16000
|
234 |
+
},
|
235 |
+
{
|
236 |
+
"epoch": 0.5172900272752924,
|
237 |
+
"grad_norm": 16.42922592163086,
|
238 |
+
"learning_rate": 4.13784995454118e-05,
|
239 |
+
"loss": 2.6656,
|
240 |
+
"step": 16500
|
241 |
+
},
|
242 |
+
{
|
243 |
+
"epoch": 0.532965482647271,
|
244 |
+
"grad_norm": 14.301593780517578,
|
245 |
+
"learning_rate": 4.111724195587882e-05,
|
246 |
+
"loss": 2.6626,
|
247 |
+
"step": 17000
|
248 |
+
},
|
249 |
+
{
|
250 |
+
"epoch": 0.5486409380192495,
|
251 |
+
"grad_norm": 17.24234390258789,
|
252 |
+
"learning_rate": 4.085598436634584e-05,
|
253 |
+
"loss": 2.6605,
|
254 |
+
"step": 17500
|
255 |
+
},
|
256 |
+
{
|
257 |
+
"epoch": 0.564316393391228,
|
258 |
+
"grad_norm": 19.45075035095215,
|
259 |
+
"learning_rate": 4.059472677681287e-05,
|
260 |
+
"loss": 2.6424,
|
261 |
+
"step": 18000
|
262 |
+
},
|
263 |
+
{
|
264 |
+
"epoch": 0.5799918487632065,
|
265 |
+
"grad_norm": 14.483268737792969,
|
266 |
+
"learning_rate": 4.033346918727989e-05,
|
267 |
+
"loss": 2.7018,
|
268 |
+
"step": 18500
|
269 |
+
},
|
270 |
+
{
|
271 |
+
"epoch": 0.5956673041351851,
|
272 |
+
"grad_norm": 15.697615623474121,
|
273 |
+
"learning_rate": 4.007221159774692e-05,
|
274 |
+
"loss": 2.6693,
|
275 |
+
"step": 19000
|
276 |
+
},
|
277 |
+
{
|
278 |
+
"epoch": 0.6113427595071637,
|
279 |
+
"grad_norm": 23.107515335083008,
|
280 |
+
"learning_rate": 3.981095400821394e-05,
|
281 |
+
"loss": 2.6445,
|
282 |
+
"step": 19500
|
283 |
+
},
|
284 |
+
{
|
285 |
+
"epoch": 0.6270182148791422,
|
286 |
+
"grad_norm": 15.768012046813965,
|
287 |
+
"learning_rate": 3.954969641868097e-05,
|
288 |
+
"loss": 2.589,
|
289 |
+
"step": 20000
|
290 |
+
},
|
291 |
+
{
|
292 |
+
"epoch": 0.6426936702511208,
|
293 |
+
"grad_norm": 24.025400161743164,
|
294 |
+
"learning_rate": 3.928843882914799e-05,
|
295 |
+
"loss": 2.6753,
|
296 |
+
"step": 20500
|
297 |
+
},
|
298 |
+
{
|
299 |
+
"epoch": 0.6583691256230993,
|
300 |
+
"grad_norm": 13.370205879211426,
|
301 |
+
"learning_rate": 3.9027181239615015e-05,
|
302 |
+
"loss": 2.5993,
|
303 |
+
"step": 21000
|
304 |
+
},
|
305 |
+
{
|
306 |
+
"epoch": 0.6740445809950779,
|
307 |
+
"grad_norm": 15.455653190612793,
|
308 |
+
"learning_rate": 3.8765923650082036e-05,
|
309 |
+
"loss": 2.6198,
|
310 |
+
"step": 21500
|
311 |
+
},
|
312 |
+
{
|
313 |
+
"epoch": 0.6897200363670565,
|
314 |
+
"grad_norm": 23.32875633239746,
|
315 |
+
"learning_rate": 3.8504666060549064e-05,
|
316 |
+
"loss": 2.6426,
|
317 |
+
"step": 22000
|
318 |
+
},
|
319 |
+
{
|
320 |
+
"epoch": 0.705395491739035,
|
321 |
+
"grad_norm": 23.468122482299805,
|
322 |
+
"learning_rate": 3.8243408471016085e-05,
|
323 |
+
"loss": 2.6068,
|
324 |
+
"step": 22500
|
325 |
+
},
|
326 |
+
{
|
327 |
+
"epoch": 0.7210709471110136,
|
328 |
+
"grad_norm": 16.411842346191406,
|
329 |
+
"learning_rate": 3.7982150881483106e-05,
|
330 |
+
"loss": 2.5992,
|
331 |
+
"step": 23000
|
332 |
+
},
|
333 |
+
{
|
334 |
+
"epoch": 0.7367464024829922,
|
335 |
+
"grad_norm": 28.372682571411133,
|
336 |
+
"learning_rate": 3.772089329195013e-05,
|
337 |
+
"loss": 2.6194,
|
338 |
+
"step": 23500
|
339 |
+
},
|
340 |
+
{
|
341 |
+
"epoch": 0.7524218578549707,
|
342 |
+
"grad_norm": 16.02229118347168,
|
343 |
+
"learning_rate": 3.7459635702417154e-05,
|
344 |
+
"loss": 2.5892,
|
345 |
+
"step": 24000
|
346 |
+
},
|
347 |
+
{
|
348 |
+
"epoch": 0.7680973132269493,
|
349 |
+
"grad_norm": 19.23042106628418,
|
350 |
+
"learning_rate": 3.719837811288418e-05,
|
351 |
+
"loss": 2.5984,
|
352 |
+
"step": 24500
|
353 |
+
},
|
354 |
+
{
|
355 |
+
"epoch": 0.7837727685989277,
|
356 |
+
"grad_norm": 23.105224609375,
|
357 |
+
"learning_rate": 3.69371205233512e-05,
|
358 |
+
"loss": 2.5461,
|
359 |
+
"step": 25000
|
360 |
+
},
|
361 |
+
{
|
362 |
+
"epoch": 0.7994482239709063,
|
363 |
+
"grad_norm": 12.95092487335205,
|
364 |
+
"learning_rate": 3.667586293381823e-05,
|
365 |
+
"loss": 2.5959,
|
366 |
+
"step": 25500
|
367 |
+
},
|
368 |
+
{
|
369 |
+
"epoch": 0.8151236793428849,
|
370 |
+
"grad_norm": 25.173099517822266,
|
371 |
+
"learning_rate": 3.641460534428525e-05,
|
372 |
+
"loss": 2.6026,
|
373 |
+
"step": 26000
|
374 |
+
},
|
375 |
+
{
|
376 |
+
"epoch": 0.8307991347148634,
|
377 |
+
"grad_norm": 16.152605056762695,
|
378 |
+
"learning_rate": 3.615334775475228e-05,
|
379 |
+
"loss": 2.5875,
|
380 |
+
"step": 26500
|
381 |
+
},
|
382 |
+
{
|
383 |
+
"epoch": 0.846474590086842,
|
384 |
+
"grad_norm": 14.907052993774414,
|
385 |
+
"learning_rate": 3.58920901652193e-05,
|
386 |
+
"loss": 2.6018,
|
387 |
+
"step": 27000
|
388 |
+
},
|
389 |
+
{
|
390 |
+
"epoch": 0.8621500454588206,
|
391 |
+
"grad_norm": 16.335224151611328,
|
392 |
+
"learning_rate": 3.563083257568632e-05,
|
393 |
+
"loss": 2.5612,
|
394 |
+
"step": 27500
|
395 |
+
},
|
396 |
+
{
|
397 |
+
"epoch": 0.8778255008307991,
|
398 |
+
"grad_norm": 15.978832244873047,
|
399 |
+
"learning_rate": 3.536957498615335e-05,
|
400 |
+
"loss": 2.6013,
|
401 |
+
"step": 28000
|
402 |
+
},
|
403 |
+
{
|
404 |
+
"epoch": 0.8935009562027777,
|
405 |
+
"grad_norm": 17.37552261352539,
|
406 |
+
"learning_rate": 3.510831739662037e-05,
|
407 |
+
"loss": 2.5882,
|
408 |
+
"step": 28500
|
409 |
+
},
|
410 |
+
{
|
411 |
+
"epoch": 0.9091764115747563,
|
412 |
+
"grad_norm": 18.65119171142578,
|
413 |
+
"learning_rate": 3.48470598070874e-05,
|
414 |
+
"loss": 2.5373,
|
415 |
+
"step": 29000
|
416 |
+
},
|
417 |
+
{
|
418 |
+
"epoch": 0.9248518669467348,
|
419 |
+
"grad_norm": 14.954788208007812,
|
420 |
+
"learning_rate": 3.458580221755442e-05,
|
421 |
+
"loss": 2.5115,
|
422 |
+
"step": 29500
|
423 |
+
},
|
424 |
+
{
|
425 |
+
"epoch": 0.9405273223187134,
|
426 |
+
"grad_norm": 15.86545467376709,
|
427 |
+
"learning_rate": 3.4324544628021445e-05,
|
428 |
+
"loss": 2.5989,
|
429 |
+
"step": 30000
|
430 |
+
},
|
431 |
+
{
|
432 |
+
"epoch": 0.956202777690692,
|
433 |
+
"grad_norm": 15.822210311889648,
|
434 |
+
"learning_rate": 3.4063287038488466e-05,
|
435 |
+
"loss": 2.5548,
|
436 |
+
"step": 30500
|
437 |
+
},
|
438 |
+
{
|
439 |
+
"epoch": 0.9718782330626705,
|
440 |
+
"grad_norm": 18.378137588500977,
|
441 |
+
"learning_rate": 3.3802029448955494e-05,
|
442 |
+
"loss": 2.5634,
|
443 |
+
"step": 31000
|
444 |
+
},
|
445 |
+
{
|
446 |
+
"epoch": 0.9875536884346491,
|
447 |
+
"grad_norm": 18.04598045349121,
|
448 |
+
"learning_rate": 3.3540771859422515e-05,
|
449 |
+
"loss": 2.5167,
|
450 |
+
"step": 31500
|
451 |
+
},
|
452 |
+
{
|
453 |
+
"epoch": 1.0032291438066276,
|
454 |
+
"grad_norm": 13.649733543395996,
|
455 |
+
"learning_rate": 3.327951426988954e-05,
|
456 |
+
"loss": 2.4968,
|
457 |
+
"step": 32000
|
458 |
+
},
|
459 |
+
{
|
460 |
+
"epoch": 1.018904599178606,
|
461 |
+
"grad_norm": 30.151395797729492,
|
462 |
+
"learning_rate": 3.301825668035657e-05,
|
463 |
+
"loss": 2.4899,
|
464 |
+
"step": 32500
|
465 |
+
},
|
466 |
+
{
|
467 |
+
"epoch": 1.0345800545505848,
|
468 |
+
"grad_norm": 18.006633758544922,
|
469 |
+
"learning_rate": 3.275699909082359e-05,
|
470 |
+
"loss": 2.513,
|
471 |
+
"step": 33000
|
472 |
+
},
|
473 |
+
{
|
474 |
+
"epoch": 1.0502555099225632,
|
475 |
+
"grad_norm": 18.782691955566406,
|
476 |
+
"learning_rate": 3.249574150129062e-05,
|
477 |
+
"loss": 2.4657,
|
478 |
+
"step": 33500
|
479 |
+
},
|
480 |
+
{
|
481 |
+
"epoch": 1.065930965294542,
|
482 |
+
"grad_norm": 14.228589057922363,
|
483 |
+
"learning_rate": 3.223448391175764e-05,
|
484 |
+
"loss": 2.4519,
|
485 |
+
"step": 34000
|
486 |
+
},
|
487 |
+
{
|
488 |
+
"epoch": 1.0816064206665204,
|
489 |
+
"grad_norm": 14.239832878112793,
|
490 |
+
"learning_rate": 3.197322632222467e-05,
|
491 |
+
"loss": 2.3959,
|
492 |
+
"step": 34500
|
493 |
+
},
|
494 |
+
{
|
495 |
+
"epoch": 1.0972818760384988,
|
496 |
+
"grad_norm": 20.227951049804688,
|
497 |
+
"learning_rate": 3.171196873269169e-05,
|
498 |
+
"loss": 2.4191,
|
499 |
+
"step": 35000
|
500 |
+
},
|
501 |
+
{
|
502 |
+
"epoch": 1.1129573314104775,
|
503 |
+
"grad_norm": 16.465883255004883,
|
504 |
+
"learning_rate": 3.145071114315871e-05,
|
505 |
+
"loss": 2.5173,
|
506 |
+
"step": 35500
|
507 |
+
},
|
508 |
+
{
|
509 |
+
"epoch": 1.1286327867824562,
|
510 |
+
"grad_norm": 28.69437599182129,
|
511 |
+
"learning_rate": 3.118945355362574e-05,
|
512 |
+
"loss": 2.5291,
|
513 |
+
"step": 36000
|
514 |
+
},
|
515 |
+
{
|
516 |
+
"epoch": 1.1443082421544346,
|
517 |
+
"grad_norm": 17.486122131347656,
|
518 |
+
"learning_rate": 3.092819596409276e-05,
|
519 |
+
"loss": 2.5009,
|
520 |
+
"step": 36500
|
521 |
+
},
|
522 |
+
{
|
523 |
+
"epoch": 1.159983697526413,
|
524 |
+
"grad_norm": 15.846352577209473,
|
525 |
+
"learning_rate": 3.0666938374559785e-05,
|
526 |
+
"loss": 2.4499,
|
527 |
+
"step": 37000
|
528 |
+
},
|
529 |
+
{
|
530 |
+
"epoch": 1.1756591528983917,
|
531 |
+
"grad_norm": 17.680450439453125,
|
532 |
+
"learning_rate": 3.0405680785026806e-05,
|
533 |
+
"loss": 2.4111,
|
534 |
+
"step": 37500
|
535 |
+
},
|
536 |
+
{
|
537 |
+
"epoch": 1.1913346082703702,
|
538 |
+
"grad_norm": 23.131710052490234,
|
539 |
+
"learning_rate": 3.014442319549383e-05,
|
540 |
+
"loss": 2.4946,
|
541 |
+
"step": 38000
|
542 |
+
},
|
543 |
+
{
|
544 |
+
"epoch": 1.2070100636423489,
|
545 |
+
"grad_norm": 23.71167755126953,
|
546 |
+
"learning_rate": 2.9883165605960855e-05,
|
547 |
+
"loss": 2.458,
|
548 |
+
"step": 38500
|
549 |
+
},
|
550 |
+
{
|
551 |
+
"epoch": 1.2226855190143273,
|
552 |
+
"grad_norm": 17.49135971069336,
|
553 |
+
"learning_rate": 2.962190801642788e-05,
|
554 |
+
"loss": 2.4595,
|
555 |
+
"step": 39000
|
556 |
+
},
|
557 |
+
{
|
558 |
+
"epoch": 1.238360974386306,
|
559 |
+
"grad_norm": 15.8767728805542,
|
560 |
+
"learning_rate": 2.9360650426894903e-05,
|
561 |
+
"loss": 2.4972,
|
562 |
+
"step": 39500
|
563 |
+
},
|
564 |
+
{
|
565 |
+
"epoch": 1.2540364297582844,
|
566 |
+
"grad_norm": 24.60512924194336,
|
567 |
+
"learning_rate": 2.9099392837361928e-05,
|
568 |
+
"loss": 2.4128,
|
569 |
+
"step": 40000
|
570 |
+
},
|
571 |
+
{
|
572 |
+
"epoch": 1.269711885130263,
|
573 |
+
"grad_norm": 21.219585418701172,
|
574 |
+
"learning_rate": 2.8838135247828952e-05,
|
575 |
+
"loss": 2.4715,
|
576 |
+
"step": 40500
|
577 |
+
},
|
578 |
+
{
|
579 |
+
"epoch": 1.2853873405022416,
|
580 |
+
"grad_norm": 15.823412895202637,
|
581 |
+
"learning_rate": 2.8576877658295976e-05,
|
582 |
+
"loss": 2.4254,
|
583 |
+
"step": 41000
|
584 |
+
},
|
585 |
+
{
|
586 |
+
"epoch": 1.3010627958742202,
|
587 |
+
"grad_norm": 15.92619514465332,
|
588 |
+
"learning_rate": 2.8315620068763e-05,
|
589 |
+
"loss": 2.4241,
|
590 |
+
"step": 41500
|
591 |
+
},
|
592 |
+
{
|
593 |
+
"epoch": 1.3167382512461987,
|
594 |
+
"grad_norm": 18.505760192871094,
|
595 |
+
"learning_rate": 2.805436247923002e-05,
|
596 |
+
"loss": 2.4412,
|
597 |
+
"step": 42000
|
598 |
+
},
|
599 |
+
{
|
600 |
+
"epoch": 1.3324137066181772,
|
601 |
+
"grad_norm": 18.413311004638672,
|
602 |
+
"learning_rate": 2.7793104889697046e-05,
|
603 |
+
"loss": 2.4697,
|
604 |
+
"step": 42500
|
605 |
+
},
|
606 |
+
{
|
607 |
+
"epoch": 1.3480891619901558,
|
608 |
+
"grad_norm": 20.015159606933594,
|
609 |
+
"learning_rate": 2.753184730016407e-05,
|
610 |
+
"loss": 2.4047,
|
611 |
+
"step": 43000
|
612 |
+
},
|
613 |
+
{
|
614 |
+
"epoch": 1.3637646173621345,
|
615 |
+
"grad_norm": 20.301258087158203,
|
616 |
+
"learning_rate": 2.7270589710631094e-05,
|
617 |
+
"loss": 2.4139,
|
618 |
+
"step": 43500
|
619 |
+
},
|
620 |
+
{
|
621 |
+
"epoch": 1.379440072734113,
|
622 |
+
"grad_norm": 14.145316123962402,
|
623 |
+
"learning_rate": 2.700933212109812e-05,
|
624 |
+
"loss": 2.4226,
|
625 |
+
"step": 44000
|
626 |
+
},
|
627 |
+
{
|
628 |
+
"epoch": 1.3951155281060914,
|
629 |
+
"grad_norm": 12.221750259399414,
|
630 |
+
"learning_rate": 2.6748074531565143e-05,
|
631 |
+
"loss": 2.3895,
|
632 |
+
"step": 44500
|
633 |
+
},
|
634 |
+
{
|
635 |
+
"epoch": 1.41079098347807,
|
636 |
+
"grad_norm": 15.449772834777832,
|
637 |
+
"learning_rate": 2.6486816942032167e-05,
|
638 |
+
"loss": 2.4323,
|
639 |
+
"step": 45000
|
640 |
+
},
|
641 |
+
{
|
642 |
+
"epoch": 1.4264664388500485,
|
643 |
+
"grad_norm": 15.565759658813477,
|
644 |
+
"learning_rate": 2.622555935249919e-05,
|
645 |
+
"loss": 2.4612,
|
646 |
+
"step": 45500
|
647 |
+
},
|
648 |
+
{
|
649 |
+
"epoch": 1.4421418942220272,
|
650 |
+
"grad_norm": 12.723085403442383,
|
651 |
+
"learning_rate": 2.5964301762966216e-05,
|
652 |
+
"loss": 2.404,
|
653 |
+
"step": 46000
|
654 |
+
},
|
655 |
+
{
|
656 |
+
"epoch": 1.4578173495940057,
|
657 |
+
"grad_norm": 27.950130462646484,
|
658 |
+
"learning_rate": 2.570304417343324e-05,
|
659 |
+
"loss": 2.4505,
|
660 |
+
"step": 46500
|
661 |
+
},
|
662 |
+
{
|
663 |
+
"epoch": 1.4734928049659843,
|
664 |
+
"grad_norm": 16.8266544342041,
|
665 |
+
"learning_rate": 2.544178658390026e-05,
|
666 |
+
"loss": 2.4143,
|
667 |
+
"step": 47000
|
668 |
+
},
|
669 |
+
{
|
670 |
+
"epoch": 1.4891682603379628,
|
671 |
+
"grad_norm": 18.774965286254883,
|
672 |
+
"learning_rate": 2.5180528994367285e-05,
|
673 |
+
"loss": 2.4196,
|
674 |
+
"step": 47500
|
675 |
+
},
|
676 |
+
{
|
677 |
+
"epoch": 1.5048437157099412,
|
678 |
+
"grad_norm": 19.893604278564453,
|
679 |
+
"learning_rate": 2.4919271404834313e-05,
|
680 |
+
"loss": 2.458,
|
681 |
+
"step": 48000
|
682 |
+
},
|
683 |
+
{
|
684 |
+
"epoch": 1.52051917108192,
|
685 |
+
"grad_norm": 17.030399322509766,
|
686 |
+
"learning_rate": 2.4658013815301337e-05,
|
687 |
+
"loss": 2.4234,
|
688 |
+
"step": 48500
|
689 |
+
},
|
690 |
+
{
|
691 |
+
"epoch": 1.5361946264538986,
|
692 |
+
"grad_norm": 15.073714256286621,
|
693 |
+
"learning_rate": 2.439675622576836e-05,
|
694 |
+
"loss": 2.358,
|
695 |
+
"step": 49000
|
696 |
+
},
|
697 |
+
{
|
698 |
+
"epoch": 1.551870081825877,
|
699 |
+
"grad_norm": 15.815202713012695,
|
700 |
+
"learning_rate": 2.4135498636235385e-05,
|
701 |
+
"loss": 2.3797,
|
702 |
+
"step": 49500
|
703 |
+
},
|
704 |
+
{
|
705 |
+
"epoch": 1.5675455371978555,
|
706 |
+
"grad_norm": 12.330552101135254,
|
707 |
+
"learning_rate": 2.387424104670241e-05,
|
708 |
+
"loss": 2.3955,
|
709 |
+
"step": 50000
|
710 |
+
},
|
711 |
+
{
|
712 |
+
"epoch": 1.5832209925698342,
|
713 |
+
"grad_norm": 20.456411361694336,
|
714 |
+
"learning_rate": 2.3612983457169434e-05,
|
715 |
+
"loss": 2.3957,
|
716 |
+
"step": 50500
|
717 |
+
},
|
718 |
+
{
|
719 |
+
"epoch": 1.5988964479418128,
|
720 |
+
"grad_norm": 13.23128604888916,
|
721 |
+
"learning_rate": 2.3351725867636455e-05,
|
722 |
+
"loss": 2.4407,
|
723 |
+
"step": 51000
|
724 |
+
},
|
725 |
+
{
|
726 |
+
"epoch": 1.6145719033137913,
|
727 |
+
"grad_norm": 24.75179100036621,
|
728 |
+
"learning_rate": 2.309046827810348e-05,
|
729 |
+
"loss": 2.3367,
|
730 |
+
"step": 51500
|
731 |
+
},
|
732 |
+
{
|
733 |
+
"epoch": 1.6302473586857698,
|
734 |
+
"grad_norm": 14.501784324645996,
|
735 |
+
"learning_rate": 2.2829210688570503e-05,
|
736 |
+
"loss": 2.4223,
|
737 |
+
"step": 52000
|
738 |
+
},
|
739 |
+
{
|
740 |
+
"epoch": 1.6459228140577484,
|
741 |
+
"grad_norm": 18.642641067504883,
|
742 |
+
"learning_rate": 2.2567953099037528e-05,
|
743 |
+
"loss": 2.3998,
|
744 |
+
"step": 52500
|
745 |
+
},
|
746 |
+
{
|
747 |
+
"epoch": 1.6615982694297269,
|
748 |
+
"grad_norm": 19.8629207611084,
|
749 |
+
"learning_rate": 2.2306695509504552e-05,
|
750 |
+
"loss": 2.3973,
|
751 |
+
"step": 53000
|
752 |
+
},
|
753 |
+
{
|
754 |
+
"epoch": 1.6772737248017053,
|
755 |
+
"grad_norm": 14.32746696472168,
|
756 |
+
"learning_rate": 2.2045437919971576e-05,
|
757 |
+
"loss": 2.3911,
|
758 |
+
"step": 53500
|
759 |
+
},
|
760 |
+
{
|
761 |
+
"epoch": 1.692949180173684,
|
762 |
+
"grad_norm": 21.35317039489746,
|
763 |
+
"learning_rate": 2.17841803304386e-05,
|
764 |
+
"loss": 2.3891,
|
765 |
+
"step": 54000
|
766 |
+
},
|
767 |
+
{
|
768 |
+
"epoch": 1.7086246355456627,
|
769 |
+
"grad_norm": 21.998720169067383,
|
770 |
+
"learning_rate": 2.1522922740905625e-05,
|
771 |
+
"loss": 2.3615,
|
772 |
+
"step": 54500
|
773 |
+
},
|
774 |
+
{
|
775 |
+
"epoch": 1.7243000909176411,
|
776 |
+
"grad_norm": 16.640851974487305,
|
777 |
+
"learning_rate": 2.126166515137265e-05,
|
778 |
+
"loss": 2.4245,
|
779 |
+
"step": 55000
|
780 |
+
},
|
781 |
+
{
|
782 |
+
"epoch": 1.7399755462896196,
|
783 |
+
"grad_norm": 15.640762329101562,
|
784 |
+
"learning_rate": 2.100040756183967e-05,
|
785 |
+
"loss": 2.3663,
|
786 |
+
"step": 55500
|
787 |
+
},
|
788 |
+
{
|
789 |
+
"epoch": 1.7556510016615983,
|
790 |
+
"grad_norm": 25.16714859008789,
|
791 |
+
"learning_rate": 2.0739149972306694e-05,
|
792 |
+
"loss": 2.3653,
|
793 |
+
"step": 56000
|
794 |
+
},
|
795 |
+
{
|
796 |
+
"epoch": 1.771326457033577,
|
797 |
+
"grad_norm": 13.939558029174805,
|
798 |
+
"learning_rate": 2.047789238277372e-05,
|
799 |
+
"loss": 2.3162,
|
800 |
+
"step": 56500
|
801 |
+
},
|
802 |
+
{
|
803 |
+
"epoch": 1.7870019124055554,
|
804 |
+
"grad_norm": 17.95723533630371,
|
805 |
+
"learning_rate": 2.0216634793240743e-05,
|
806 |
+
"loss": 2.3965,
|
807 |
+
"step": 57000
|
808 |
+
},
|
809 |
+
{
|
810 |
+
"epoch": 1.8026773677775338,
|
811 |
+
"grad_norm": 16.54608917236328,
|
812 |
+
"learning_rate": 1.9955377203707767e-05,
|
813 |
+
"loss": 2.3554,
|
814 |
+
"step": 57500
|
815 |
+
},
|
816 |
+
{
|
817 |
+
"epoch": 1.8183528231495125,
|
818 |
+
"grad_norm": 15.683633804321289,
|
819 |
+
"learning_rate": 1.969411961417479e-05,
|
820 |
+
"loss": 2.3475,
|
821 |
+
"step": 58000
|
822 |
+
},
|
823 |
+
{
|
824 |
+
"epoch": 1.8340282785214912,
|
825 |
+
"grad_norm": 15.327407836914062,
|
826 |
+
"learning_rate": 1.943286202464182e-05,
|
827 |
+
"loss": 2.3796,
|
828 |
+
"step": 58500
|
829 |
+
},
|
830 |
+
{
|
831 |
+
"epoch": 1.8497037338934696,
|
832 |
+
"grad_norm": 18.815610885620117,
|
833 |
+
"learning_rate": 1.9171604435108843e-05,
|
834 |
+
"loss": 2.3763,
|
835 |
+
"step": 59000
|
836 |
+
},
|
837 |
+
{
|
838 |
+
"epoch": 1.865379189265448,
|
839 |
+
"grad_norm": 18.846534729003906,
|
840 |
+
"learning_rate": 1.8910346845575864e-05,
|
841 |
+
"loss": 2.3869,
|
842 |
+
"step": 59500
|
843 |
+
},
|
844 |
+
{
|
845 |
+
"epoch": 1.8810546446374268,
|
846 |
+
"grad_norm": 24.27644157409668,
|
847 |
+
"learning_rate": 1.864908925604289e-05,
|
848 |
+
"loss": 2.3079,
|
849 |
+
"step": 60000
|
850 |
+
},
|
851 |
+
{
|
852 |
+
"epoch": 1.8967301000094052,
|
853 |
+
"grad_norm": 19.45649528503418,
|
854 |
+
"learning_rate": 1.8387831666509913e-05,
|
855 |
+
"loss": 2.379,
|
856 |
+
"step": 60500
|
857 |
+
},
|
858 |
+
{
|
859 |
+
"epoch": 1.9124055553813837,
|
860 |
+
"grad_norm": 19.247732162475586,
|
861 |
+
"learning_rate": 1.8126574076976937e-05,
|
862 |
+
"loss": 2.3454,
|
863 |
+
"step": 61000
|
864 |
+
},
|
865 |
+
{
|
866 |
+
"epoch": 1.9280810107533624,
|
867 |
+
"grad_norm": 16.541532516479492,
|
868 |
+
"learning_rate": 1.786531648744396e-05,
|
869 |
+
"loss": 2.3611,
|
870 |
+
"step": 61500
|
871 |
+
},
|
872 |
+
{
|
873 |
+
"epoch": 1.943756466125341,
|
874 |
+
"grad_norm": 19.154760360717773,
|
875 |
+
"learning_rate": 1.7604058897910986e-05,
|
876 |
+
"loss": 2.2952,
|
877 |
+
"step": 62000
|
878 |
+
},
|
879 |
+
{
|
880 |
+
"epoch": 1.9594319214973195,
|
881 |
+
"grad_norm": 13.325096130371094,
|
882 |
+
"learning_rate": 1.734280130837801e-05,
|
883 |
+
"loss": 2.3158,
|
884 |
+
"step": 62500
|
885 |
+
},
|
886 |
+
{
|
887 |
+
"epoch": 1.975107376869298,
|
888 |
+
"grad_norm": 12.556410789489746,
|
889 |
+
"learning_rate": 1.7081543718845034e-05,
|
890 |
+
"loss": 2.3199,
|
891 |
+
"step": 63000
|
892 |
+
},
|
893 |
+
{
|
894 |
+
"epoch": 1.9907828322412766,
|
895 |
+
"grad_norm": 17.374223709106445,
|
896 |
+
"learning_rate": 1.682028612931206e-05,
|
897 |
+
"loss": 2.3241,
|
898 |
+
"step": 63500
|
899 |
+
},
|
900 |
+
{
|
901 |
+
"epoch": 2.0064582876132553,
|
902 |
+
"grad_norm": 23.0031795501709,
|
903 |
+
"learning_rate": 1.6559028539779083e-05,
|
904 |
+
"loss": 2.2653,
|
905 |
+
"step": 64000
|
906 |
+
},
|
907 |
+
{
|
908 |
+
"epoch": 2.0221337429852335,
|
909 |
+
"grad_norm": 24.6336612701416,
|
910 |
+
"learning_rate": 1.6297770950246104e-05,
|
911 |
+
"loss": 2.3009,
|
912 |
+
"step": 64500
|
913 |
+
},
|
914 |
+
{
|
915 |
+
"epoch": 2.037809198357212,
|
916 |
+
"grad_norm": 21.2884521484375,
|
917 |
+
"learning_rate": 1.6036513360713128e-05,
|
918 |
+
"loss": 2.3493,
|
919 |
+
"step": 65000
|
920 |
+
},
|
921 |
+
{
|
922 |
+
"epoch": 2.053484653729191,
|
923 |
+
"grad_norm": 13.374622344970703,
|
924 |
+
"learning_rate": 1.5775255771180152e-05,
|
925 |
+
"loss": 2.2815,
|
926 |
+
"step": 65500
|
927 |
+
},
|
928 |
+
{
|
929 |
+
"epoch": 2.0691601091011695,
|
930 |
+
"grad_norm": 20.058076858520508,
|
931 |
+
"learning_rate": 1.5513998181647176e-05,
|
932 |
+
"loss": 2.2794,
|
933 |
+
"step": 66000
|
934 |
+
},
|
935 |
+
{
|
936 |
+
"epoch": 2.0848355644731478,
|
937 |
+
"grad_norm": 22.910066604614258,
|
938 |
+
"learning_rate": 1.52527405921142e-05,
|
939 |
+
"loss": 2.3015,
|
940 |
+
"step": 66500
|
941 |
+
},
|
942 |
+
{
|
943 |
+
"epoch": 2.1005110198451264,
|
944 |
+
"grad_norm": 22.905941009521484,
|
945 |
+
"learning_rate": 1.4991483002581225e-05,
|
946 |
+
"loss": 2.2263,
|
947 |
+
"step": 67000
|
948 |
+
},
|
949 |
+
{
|
950 |
+
"epoch": 2.116186475217105,
|
951 |
+
"grad_norm": 15.105734825134277,
|
952 |
+
"learning_rate": 1.473022541304825e-05,
|
953 |
+
"loss": 2.2947,
|
954 |
+
"step": 67500
|
955 |
+
},
|
956 |
+
{
|
957 |
+
"epoch": 2.131861930589084,
|
958 |
+
"grad_norm": 23.665620803833008,
|
959 |
+
"learning_rate": 1.4468967823515272e-05,
|
960 |
+
"loss": 2.2644,
|
961 |
+
"step": 68000
|
962 |
+
},
|
963 |
+
{
|
964 |
+
"epoch": 2.147537385961062,
|
965 |
+
"grad_norm": 13.619988441467285,
|
966 |
+
"learning_rate": 1.4207710233982296e-05,
|
967 |
+
"loss": 2.3118,
|
968 |
+
"step": 68500
|
969 |
+
},
|
970 |
+
{
|
971 |
+
"epoch": 2.1632128413330407,
|
972 |
+
"grad_norm": 21.304323196411133,
|
973 |
+
"learning_rate": 1.3946452644449324e-05,
|
974 |
+
"loss": 2.2925,
|
975 |
+
"step": 69000
|
976 |
+
},
|
977 |
+
{
|
978 |
+
"epoch": 2.1788882967050194,
|
979 |
+
"grad_norm": 11.025732040405273,
|
980 |
+
"learning_rate": 1.3685195054916346e-05,
|
981 |
+
"loss": 2.2859,
|
982 |
+
"step": 69500
|
983 |
+
},
|
984 |
+
{
|
985 |
+
"epoch": 2.1945637520769976,
|
986 |
+
"grad_norm": 20.443811416625977,
|
987 |
+
"learning_rate": 1.342393746538337e-05,
|
988 |
+
"loss": 2.3155,
|
989 |
+
"step": 70000
|
990 |
+
},
|
991 |
+
{
|
992 |
+
"epoch": 2.2102392074489763,
|
993 |
+
"grad_norm": 19.622135162353516,
|
994 |
+
"learning_rate": 1.3162679875850395e-05,
|
995 |
+
"loss": 2.3166,
|
996 |
+
"step": 70500
|
997 |
+
},
|
998 |
+
{
|
999 |
+
"epoch": 2.225914662820955,
|
1000 |
+
"grad_norm": 17.77559471130371,
|
1001 |
+
"learning_rate": 1.290142228631742e-05,
|
1002 |
+
"loss": 2.2971,
|
1003 |
+
"step": 71000
|
1004 |
+
},
|
1005 |
+
{
|
1006 |
+
"epoch": 2.2415901181929336,
|
1007 |
+
"grad_norm": 20.948503494262695,
|
1008 |
+
"learning_rate": 1.2640164696784444e-05,
|
1009 |
+
"loss": 2.2901,
|
1010 |
+
"step": 71500
|
1011 |
+
},
|
1012 |
+
{
|
1013 |
+
"epoch": 2.2572655735649123,
|
1014 |
+
"grad_norm": 19.734085083007812,
|
1015 |
+
"learning_rate": 1.2378907107251466e-05,
|
1016 |
+
"loss": 2.2866,
|
1017 |
+
"step": 72000
|
1018 |
+
},
|
1019 |
+
{
|
1020 |
+
"epoch": 2.2729410289368905,
|
1021 |
+
"grad_norm": 14.341938018798828,
|
1022 |
+
"learning_rate": 1.211764951771849e-05,
|
1023 |
+
"loss": 2.2806,
|
1024 |
+
"step": 72500
|
1025 |
+
},
|
1026 |
+
{
|
1027 |
+
"epoch": 2.288616484308869,
|
1028 |
+
"grad_norm": 16.937484741210938,
|
1029 |
+
"learning_rate": 1.1856391928185515e-05,
|
1030 |
+
"loss": 2.2824,
|
1031 |
+
"step": 73000
|
1032 |
+
},
|
1033 |
+
{
|
1034 |
+
"epoch": 2.304291939680848,
|
1035 |
+
"grad_norm": 12.707877159118652,
|
1036 |
+
"learning_rate": 1.1595134338652539e-05,
|
1037 |
+
"loss": 2.2768,
|
1038 |
+
"step": 73500
|
1039 |
+
},
|
1040 |
+
{
|
1041 |
+
"epoch": 2.319967395052826,
|
1042 |
+
"grad_norm": 12.61535358428955,
|
1043 |
+
"learning_rate": 1.1333876749119563e-05,
|
1044 |
+
"loss": 2.2395,
|
1045 |
+
"step": 74000
|
1046 |
+
},
|
1047 |
+
{
|
1048 |
+
"epoch": 2.335642850424805,
|
1049 |
+
"grad_norm": 29.96654510498047,
|
1050 |
+
"learning_rate": 1.1072619159586586e-05,
|
1051 |
+
"loss": 2.2204,
|
1052 |
+
"step": 74500
|
1053 |
+
},
|
1054 |
+
{
|
1055 |
+
"epoch": 2.3513183057967835,
|
1056 |
+
"grad_norm": 14.348071098327637,
|
1057 |
+
"learning_rate": 1.081136157005361e-05,
|
1058 |
+
"loss": 2.2712,
|
1059 |
+
"step": 75000
|
1060 |
+
},
|
1061 |
+
{
|
1062 |
+
"epoch": 2.366993761168762,
|
1063 |
+
"grad_norm": 16.751550674438477,
|
1064 |
+
"learning_rate": 1.0550103980520634e-05,
|
1065 |
+
"loss": 2.1824,
|
1066 |
+
"step": 75500
|
1067 |
+
},
|
1068 |
+
{
|
1069 |
+
"epoch": 2.3826692165407404,
|
1070 |
+
"grad_norm": 13.812084197998047,
|
1071 |
+
"learning_rate": 1.0288846390987659e-05,
|
1072 |
+
"loss": 2.1823,
|
1073 |
+
"step": 76000
|
1074 |
+
},
|
1075 |
+
{
|
1076 |
+
"epoch": 2.398344671912719,
|
1077 |
+
"grad_norm": 19.040403366088867,
|
1078 |
+
"learning_rate": 1.0027588801454681e-05,
|
1079 |
+
"loss": 2.2536,
|
1080 |
+
"step": 76500
|
1081 |
+
},
|
1082 |
+
{
|
1083 |
+
"epoch": 2.4140201272846977,
|
1084 |
+
"grad_norm": 21.43050765991211,
|
1085 |
+
"learning_rate": 9.766331211921707e-06,
|
1086 |
+
"loss": 2.247,
|
1087 |
+
"step": 77000
|
1088 |
+
},
|
1089 |
+
{
|
1090 |
+
"epoch": 2.4296955826566764,
|
1091 |
+
"grad_norm": 17.847736358642578,
|
1092 |
+
"learning_rate": 9.505073622388731e-06,
|
1093 |
+
"loss": 2.2038,
|
1094 |
+
"step": 77500
|
1095 |
+
},
|
1096 |
+
{
|
1097 |
+
"epoch": 2.4453710380286546,
|
1098 |
+
"grad_norm": 16.89531135559082,
|
1099 |
+
"learning_rate": 9.243816032855756e-06,
|
1100 |
+
"loss": 2.2927,
|
1101 |
+
"step": 78000
|
1102 |
+
},
|
1103 |
+
{
|
1104 |
+
"epoch": 2.4610464934006333,
|
1105 |
+
"grad_norm": 21.707191467285156,
|
1106 |
+
"learning_rate": 8.982558443322778e-06,
|
1107 |
+
"loss": 2.2851,
|
1108 |
+
"step": 78500
|
1109 |
+
},
|
1110 |
+
{
|
1111 |
+
"epoch": 2.476721948772612,
|
1112 |
+
"grad_norm": 20.218366622924805,
|
1113 |
+
"learning_rate": 8.721300853789803e-06,
|
1114 |
+
"loss": 2.2716,
|
1115 |
+
"step": 79000
|
1116 |
+
},
|
1117 |
+
{
|
1118 |
+
"epoch": 2.49239740414459,
|
1119 |
+
"grad_norm": 25.17759895324707,
|
1120 |
+
"learning_rate": 8.460043264256827e-06,
|
1121 |
+
"loss": 2.2357,
|
1122 |
+
"step": 79500
|
1123 |
+
},
|
1124 |
+
{
|
1125 |
+
"epoch": 2.508072859516569,
|
1126 |
+
"grad_norm": 17.571487426757812,
|
1127 |
+
"learning_rate": 8.198785674723851e-06,
|
1128 |
+
"loss": 2.1629,
|
1129 |
+
"step": 80000
|
1130 |
+
},
|
1131 |
+
{
|
1132 |
+
"epoch": 2.5237483148885476,
|
1133 |
+
"grad_norm": 19.578426361083984,
|
1134 |
+
"learning_rate": 7.937528085190875e-06,
|
1135 |
+
"loss": 2.2371,
|
1136 |
+
"step": 80500
|
1137 |
+
},
|
1138 |
+
{
|
1139 |
+
"epoch": 2.539423770260526,
|
1140 |
+
"grad_norm": 16.45771026611328,
|
1141 |
+
"learning_rate": 7.676270495657898e-06,
|
1142 |
+
"loss": 2.228,
|
1143 |
+
"step": 81000
|
1144 |
+
},
|
1145 |
+
{
|
1146 |
+
"epoch": 2.5550992256325045,
|
1147 |
+
"grad_norm": 19.219467163085938,
|
1148 |
+
"learning_rate": 7.415012906124922e-06,
|
1149 |
+
"loss": 2.2392,
|
1150 |
+
"step": 81500
|
1151 |
+
},
|
1152 |
+
{
|
1153 |
+
"epoch": 2.570774681004483,
|
1154 |
+
"grad_norm": 21.938257217407227,
|
1155 |
+
"learning_rate": 7.153755316591948e-06,
|
1156 |
+
"loss": 2.2117,
|
1157 |
+
"step": 82000
|
1158 |
+
},
|
1159 |
+
{
|
1160 |
+
"epoch": 2.586450136376462,
|
1161 |
+
"grad_norm": 18.676382064819336,
|
1162 |
+
"learning_rate": 6.892497727058972e-06,
|
1163 |
+
"loss": 2.2385,
|
1164 |
+
"step": 82500
|
1165 |
+
},
|
1166 |
+
{
|
1167 |
+
"epoch": 2.6021255917484405,
|
1168 |
+
"grad_norm": 15.486830711364746,
|
1169 |
+
"learning_rate": 6.631240137525996e-06,
|
1170 |
+
"loss": 2.1963,
|
1171 |
+
"step": 83000
|
1172 |
+
},
|
1173 |
+
{
|
1174 |
+
"epoch": 2.6178010471204187,
|
1175 |
+
"grad_norm": 17.053260803222656,
|
1176 |
+
"learning_rate": 6.369982547993019e-06,
|
1177 |
+
"loss": 2.2028,
|
1178 |
+
"step": 83500
|
1179 |
+
},
|
1180 |
+
{
|
1181 |
+
"epoch": 2.6334765024923974,
|
1182 |
+
"grad_norm": 13.611345291137695,
|
1183 |
+
"learning_rate": 6.108724958460044e-06,
|
1184 |
+
"loss": 2.1783,
|
1185 |
+
"step": 84000
|
1186 |
+
},
|
1187 |
+
{
|
1188 |
+
"epoch": 2.649151957864376,
|
1189 |
+
"grad_norm": 18.15059471130371,
|
1190 |
+
"learning_rate": 5.847467368927068e-06,
|
1191 |
+
"loss": 2.216,
|
1192 |
+
"step": 84500
|
1193 |
+
},
|
1194 |
+
{
|
1195 |
+
"epoch": 2.6648274132363543,
|
1196 |
+
"grad_norm": 14.26778793334961,
|
1197 |
+
"learning_rate": 5.586209779394091e-06,
|
1198 |
+
"loss": 2.2291,
|
1199 |
+
"step": 85000
|
1200 |
+
},
|
1201 |
+
{
|
1202 |
+
"epoch": 2.680502868608333,
|
1203 |
+
"grad_norm": 14.207221031188965,
|
1204 |
+
"learning_rate": 5.324952189861116e-06,
|
1205 |
+
"loss": 2.2301,
|
1206 |
+
"step": 85500
|
1207 |
+
},
|
1208 |
+
{
|
1209 |
+
"epoch": 2.6961783239803117,
|
1210 |
+
"grad_norm": 18.4339542388916,
|
1211 |
+
"learning_rate": 5.06369460032814e-06,
|
1212 |
+
"loss": 2.2454,
|
1213 |
+
"step": 86000
|
1214 |
+
},
|
1215 |
+
{
|
1216 |
+
"epoch": 2.7118537793522903,
|
1217 |
+
"grad_norm": 14.96839427947998,
|
1218 |
+
"learning_rate": 4.802437010795164e-06,
|
1219 |
+
"loss": 2.1817,
|
1220 |
+
"step": 86500
|
1221 |
+
},
|
1222 |
+
{
|
1223 |
+
"epoch": 2.727529234724269,
|
1224 |
+
"grad_norm": 15.77700424194336,
|
1225 |
+
"learning_rate": 4.541179421262188e-06,
|
1226 |
+
"loss": 2.1856,
|
1227 |
+
"step": 87000
|
1228 |
+
},
|
1229 |
+
{
|
1230 |
+
"epoch": 2.7432046900962472,
|
1231 |
+
"grad_norm": 18.85235023498535,
|
1232 |
+
"learning_rate": 4.279921831729212e-06,
|
1233 |
+
"loss": 2.168,
|
1234 |
+
"step": 87500
|
1235 |
+
},
|
1236 |
+
{
|
1237 |
+
"epoch": 2.758880145468226,
|
1238 |
+
"grad_norm": 26.718852996826172,
|
1239 |
+
"learning_rate": 4.018664242196235e-06,
|
1240 |
+
"loss": 2.2026,
|
1241 |
+
"step": 88000
|
1242 |
+
},
|
1243 |
+
{
|
1244 |
+
"epoch": 2.7745556008402046,
|
1245 |
+
"grad_norm": 14.873674392700195,
|
1246 |
+
"learning_rate": 3.7574066526632605e-06,
|
1247 |
+
"loss": 2.1595,
|
1248 |
+
"step": 88500
|
1249 |
+
},
|
1250 |
+
{
|
1251 |
+
"epoch": 2.790231056212183,
|
1252 |
+
"grad_norm": 21.835710525512695,
|
1253 |
+
"learning_rate": 3.4961490631302843e-06,
|
1254 |
+
"loss": 2.1986,
|
1255 |
+
"step": 89000
|
1256 |
+
},
|
1257 |
+
{
|
1258 |
+
"epoch": 2.8059065115841615,
|
1259 |
+
"grad_norm": 17.2098331451416,
|
1260 |
+
"learning_rate": 3.234891473597308e-06,
|
1261 |
+
"loss": 2.2289,
|
1262 |
+
"step": 89500
|
1263 |
+
},
|
1264 |
+
{
|
1265 |
+
"epoch": 2.82158196695614,
|
1266 |
+
"grad_norm": 15.5960111618042,
|
1267 |
+
"learning_rate": 2.973633884064332e-06,
|
1268 |
+
"loss": 2.2339,
|
1269 |
+
"step": 90000
|
1270 |
+
}
|
1271 |
+
],
|
1272 |
+
"logging_steps": 500,
|
1273 |
+
"max_steps": 95691,
|
1274 |
+
"num_input_tokens_seen": 0,
|
1275 |
+
"num_train_epochs": 3,
|
1276 |
+
"save_steps": 10000,
|
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": 2.3860809255545856e+16,
|
1290 |
+
"train_batch_size": 8,
|
1291 |
+
"trial_name": null,
|
1292 |
+
"trial_params": null
|
1293 |
+
}
|
ConstructionBertV2/distilBertTrain_ConstructionManagement/checkpoint-90000/training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:b300cbf6a42b906b10e3e12d38967e1e12f1a1c7aa6f157987dbce7eb0215fff
|
3 |
+
size 5713
|
ConstructionBertV2/distilBertTrain_ConstructionManagement/checkpoint-95691/config.json
ADDED
@@ -0,0 +1,24 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "distilbert-base-uncased",
|
3 |
+
"activation": "gelu",
|
4 |
+
"architectures": [
|
5 |
+
"DistilBertForMaskedLM"
|
6 |
+
],
|
7 |
+
"attention_dropout": 0.1,
|
8 |
+
"dim": 768,
|
9 |
+
"dropout": 0.1,
|
10 |
+
"hidden_dim": 3072,
|
11 |
+
"initializer_range": 0.02,
|
12 |
+
"max_position_embeddings": 512,
|
13 |
+
"model_type": "distilbert",
|
14 |
+
"n_heads": 12,
|
15 |
+
"n_layers": 6,
|
16 |
+
"pad_token_id": 0,
|
17 |
+
"qa_dropout": 0.1,
|
18 |
+
"seq_classif_dropout": 0.2,
|
19 |
+
"sinusoidal_pos_embds": false,
|
20 |
+
"tie_weights_": true,
|
21 |
+
"torch_dtype": "float32",
|
22 |
+
"transformers_version": "4.49.0",
|
23 |
+
"vocab_size": 30522
|
24 |
+
}
|
ConstructionBertV2/distilBertTrain_ConstructionManagement/checkpoint-95691/model.safetensors
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:a3daa13429a8ee4beca3fa677cfd43454133c3399d2c1e70b7da0b098b2efabc
|
3 |
+
size 267954768
|
ConstructionBertV2/distilBertTrain_ConstructionManagement/checkpoint-95691/optimizer.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:874743a99e8b265123a917404cf70031df65bb0925b39f20af74ccf1d9aaf3e1
|
3 |
+
size 535972683
|
ConstructionBertV2/distilBertTrain_ConstructionManagement/checkpoint-95691/rng_state.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:7055213a79ced83b484d638967d5a2932f9437f7417dd6c87dbe0a0be488099b
|
3 |
+
size 14645
|
ConstructionBertV2/distilBertTrain_ConstructionManagement/checkpoint-95691/scheduler.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:39de7b6a6ac61ce63fff2f503d6142614f7fda604dd14d36fbc690b131827597
|
3 |
+
size 1465
|
ConstructionBertV2/distilBertTrain_ConstructionManagement/checkpoint-95691/trainer_state.json
ADDED
@@ -0,0 +1,1370 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": null,
|
3 |
+
"best_model_checkpoint": null,
|
4 |
+
"epoch": 3.0,
|
5 |
+
"eval_steps": 500,
|
6 |
+
"global_step": 95691,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 0.015675455371978557,
|
13 |
+
"grad_norm": 22.547378540039062,
|
14 |
+
"learning_rate": 4.973874241046702e-05,
|
15 |
+
"loss": 3.0259,
|
16 |
+
"step": 500
|
17 |
+
},
|
18 |
+
{
|
19 |
+
"epoch": 0.031350910743957114,
|
20 |
+
"grad_norm": 15.296708106994629,
|
21 |
+
"learning_rate": 4.947748482093405e-05,
|
22 |
+
"loss": 2.9667,
|
23 |
+
"step": 1000
|
24 |
+
},
|
25 |
+
{
|
26 |
+
"epoch": 0.04702636611593567,
|
27 |
+
"grad_norm": 22.362838745117188,
|
28 |
+
"learning_rate": 4.921622723140107e-05,
|
29 |
+
"loss": 2.9586,
|
30 |
+
"step": 1500
|
31 |
+
},
|
32 |
+
{
|
33 |
+
"epoch": 0.06270182148791423,
|
34 |
+
"grad_norm": 20.58307647705078,
|
35 |
+
"learning_rate": 4.89549696418681e-05,
|
36 |
+
"loss": 2.9069,
|
37 |
+
"step": 2000
|
38 |
+
},
|
39 |
+
{
|
40 |
+
"epoch": 0.07837727685989278,
|
41 |
+
"grad_norm": 14.068394660949707,
|
42 |
+
"learning_rate": 4.869371205233512e-05,
|
43 |
+
"loss": 2.8975,
|
44 |
+
"step": 2500
|
45 |
+
},
|
46 |
+
{
|
47 |
+
"epoch": 0.09405273223187134,
|
48 |
+
"grad_norm": 14.740192413330078,
|
49 |
+
"learning_rate": 4.843245446280215e-05,
|
50 |
+
"loss": 2.8747,
|
51 |
+
"step": 3000
|
52 |
+
},
|
53 |
+
{
|
54 |
+
"epoch": 0.10972818760384989,
|
55 |
+
"grad_norm": 19.214658737182617,
|
56 |
+
"learning_rate": 4.817119687326917e-05,
|
57 |
+
"loss": 2.9392,
|
58 |
+
"step": 3500
|
59 |
+
},
|
60 |
+
{
|
61 |
+
"epoch": 0.12540364297582846,
|
62 |
+
"grad_norm": 20.592220306396484,
|
63 |
+
"learning_rate": 4.79099392837362e-05,
|
64 |
+
"loss": 2.8518,
|
65 |
+
"step": 4000
|
66 |
+
},
|
67 |
+
{
|
68 |
+
"epoch": 0.141079098347807,
|
69 |
+
"grad_norm": 14.96511459350586,
|
70 |
+
"learning_rate": 4.764868169420322e-05,
|
71 |
+
"loss": 2.8684,
|
72 |
+
"step": 4500
|
73 |
+
},
|
74 |
+
{
|
75 |
+
"epoch": 0.15675455371978556,
|
76 |
+
"grad_norm": 20.952829360961914,
|
77 |
+
"learning_rate": 4.738742410467024e-05,
|
78 |
+
"loss": 2.7943,
|
79 |
+
"step": 5000
|
80 |
+
},
|
81 |
+
{
|
82 |
+
"epoch": 0.17243000909176412,
|
83 |
+
"grad_norm": 23.319133758544922,
|
84 |
+
"learning_rate": 4.7126166515137266e-05,
|
85 |
+
"loss": 2.8203,
|
86 |
+
"step": 5500
|
87 |
+
},
|
88 |
+
{
|
89 |
+
"epoch": 0.18810546446374268,
|
90 |
+
"grad_norm": 12.275296211242676,
|
91 |
+
"learning_rate": 4.686490892560429e-05,
|
92 |
+
"loss": 2.8257,
|
93 |
+
"step": 6000
|
94 |
+
},
|
95 |
+
{
|
96 |
+
"epoch": 0.20378091983572122,
|
97 |
+
"grad_norm": 24.349985122680664,
|
98 |
+
"learning_rate": 4.6603651336071315e-05,
|
99 |
+
"loss": 2.7974,
|
100 |
+
"step": 6500
|
101 |
+
},
|
102 |
+
{
|
103 |
+
"epoch": 0.21945637520769978,
|
104 |
+
"grad_norm": 20.776357650756836,
|
105 |
+
"learning_rate": 4.6342393746538335e-05,
|
106 |
+
"loss": 2.7539,
|
107 |
+
"step": 7000
|
108 |
+
},
|
109 |
+
{
|
110 |
+
"epoch": 0.23513183057967835,
|
111 |
+
"grad_norm": 23.954627990722656,
|
112 |
+
"learning_rate": 4.608113615700536e-05,
|
113 |
+
"loss": 2.7877,
|
114 |
+
"step": 7500
|
115 |
+
},
|
116 |
+
{
|
117 |
+
"epoch": 0.2508072859516569,
|
118 |
+
"grad_norm": 14.333002090454102,
|
119 |
+
"learning_rate": 4.5819878567472384e-05,
|
120 |
+
"loss": 2.7571,
|
121 |
+
"step": 8000
|
122 |
+
},
|
123 |
+
{
|
124 |
+
"epoch": 0.2664827413236355,
|
125 |
+
"grad_norm": 17.226673126220703,
|
126 |
+
"learning_rate": 4.555862097793941e-05,
|
127 |
+
"loss": 2.7535,
|
128 |
+
"step": 8500
|
129 |
+
},
|
130 |
+
{
|
131 |
+
"epoch": 0.282158196695614,
|
132 |
+
"grad_norm": 24.639087677001953,
|
133 |
+
"learning_rate": 4.529736338840643e-05,
|
134 |
+
"loss": 2.7562,
|
135 |
+
"step": 9000
|
136 |
+
},
|
137 |
+
{
|
138 |
+
"epoch": 0.29783365206759255,
|
139 |
+
"grad_norm": 15.692695617675781,
|
140 |
+
"learning_rate": 4.503610579887346e-05,
|
141 |
+
"loss": 2.7613,
|
142 |
+
"step": 9500
|
143 |
+
},
|
144 |
+
{
|
145 |
+
"epoch": 0.3135091074395711,
|
146 |
+
"grad_norm": 22.752601623535156,
|
147 |
+
"learning_rate": 4.477484820934048e-05,
|
148 |
+
"loss": 2.7316,
|
149 |
+
"step": 10000
|
150 |
+
},
|
151 |
+
{
|
152 |
+
"epoch": 0.3291845628115497,
|
153 |
+
"grad_norm": 21.343364715576172,
|
154 |
+
"learning_rate": 4.45135906198075e-05,
|
155 |
+
"loss": 2.6913,
|
156 |
+
"step": 10500
|
157 |
+
},
|
158 |
+
{
|
159 |
+
"epoch": 0.34486001818352824,
|
160 |
+
"grad_norm": 23.11554527282715,
|
161 |
+
"learning_rate": 4.4252333030274536e-05,
|
162 |
+
"loss": 2.7792,
|
163 |
+
"step": 11000
|
164 |
+
},
|
165 |
+
{
|
166 |
+
"epoch": 0.3605354735555068,
|
167 |
+
"grad_norm": 20.79396629333496,
|
168 |
+
"learning_rate": 4.399107544074156e-05,
|
169 |
+
"loss": 2.6721,
|
170 |
+
"step": 11500
|
171 |
+
},
|
172 |
+
{
|
173 |
+
"epoch": 0.37621092892748537,
|
174 |
+
"grad_norm": 18.559513092041016,
|
175 |
+
"learning_rate": 4.3729817851208585e-05,
|
176 |
+
"loss": 2.7126,
|
177 |
+
"step": 12000
|
178 |
+
},
|
179 |
+
{
|
180 |
+
"epoch": 0.3918863842994639,
|
181 |
+
"grad_norm": 16.206348419189453,
|
182 |
+
"learning_rate": 4.3468560261675606e-05,
|
183 |
+
"loss": 2.7313,
|
184 |
+
"step": 12500
|
185 |
+
},
|
186 |
+
{
|
187 |
+
"epoch": 0.40756183967144244,
|
188 |
+
"grad_norm": 16.35007667541504,
|
189 |
+
"learning_rate": 4.320730267214263e-05,
|
190 |
+
"loss": 2.755,
|
191 |
+
"step": 13000
|
192 |
+
},
|
193 |
+
{
|
194 |
+
"epoch": 0.423237295043421,
|
195 |
+
"grad_norm": 28.126436233520508,
|
196 |
+
"learning_rate": 4.2946045082609654e-05,
|
197 |
+
"loss": 2.714,
|
198 |
+
"step": 13500
|
199 |
+
},
|
200 |
+
{
|
201 |
+
"epoch": 0.43891275041539957,
|
202 |
+
"grad_norm": 16.460905075073242,
|
203 |
+
"learning_rate": 4.2684787493076675e-05,
|
204 |
+
"loss": 2.6553,
|
205 |
+
"step": 14000
|
206 |
+
},
|
207 |
+
{
|
208 |
+
"epoch": 0.45458820578737813,
|
209 |
+
"grad_norm": 18.516586303710938,
|
210 |
+
"learning_rate": 4.24235299035437e-05,
|
211 |
+
"loss": 2.6515,
|
212 |
+
"step": 14500
|
213 |
+
},
|
214 |
+
{
|
215 |
+
"epoch": 0.4702636611593567,
|
216 |
+
"grad_norm": 15.802982330322266,
|
217 |
+
"learning_rate": 4.2162272314010724e-05,
|
218 |
+
"loss": 2.7276,
|
219 |
+
"step": 15000
|
220 |
+
},
|
221 |
+
{
|
222 |
+
"epoch": 0.48593911653133526,
|
223 |
+
"grad_norm": 14.24789810180664,
|
224 |
+
"learning_rate": 4.190101472447775e-05,
|
225 |
+
"loss": 2.7027,
|
226 |
+
"step": 15500
|
227 |
+
},
|
228 |
+
{
|
229 |
+
"epoch": 0.5016145719033138,
|
230 |
+
"grad_norm": 15.48758316040039,
|
231 |
+
"learning_rate": 4.163975713494477e-05,
|
232 |
+
"loss": 2.6882,
|
233 |
+
"step": 16000
|
234 |
+
},
|
235 |
+
{
|
236 |
+
"epoch": 0.5172900272752924,
|
237 |
+
"grad_norm": 16.42922592163086,
|
238 |
+
"learning_rate": 4.13784995454118e-05,
|
239 |
+
"loss": 2.6656,
|
240 |
+
"step": 16500
|
241 |
+
},
|
242 |
+
{
|
243 |
+
"epoch": 0.532965482647271,
|
244 |
+
"grad_norm": 14.301593780517578,
|
245 |
+
"learning_rate": 4.111724195587882e-05,
|
246 |
+
"loss": 2.6626,
|
247 |
+
"step": 17000
|
248 |
+
},
|
249 |
+
{
|
250 |
+
"epoch": 0.5486409380192495,
|
251 |
+
"grad_norm": 17.24234390258789,
|
252 |
+
"learning_rate": 4.085598436634584e-05,
|
253 |
+
"loss": 2.6605,
|
254 |
+
"step": 17500
|
255 |
+
},
|
256 |
+
{
|
257 |
+
"epoch": 0.564316393391228,
|
258 |
+
"grad_norm": 19.45075035095215,
|
259 |
+
"learning_rate": 4.059472677681287e-05,
|
260 |
+
"loss": 2.6424,
|
261 |
+
"step": 18000
|
262 |
+
},
|
263 |
+
{
|
264 |
+
"epoch": 0.5799918487632065,
|
265 |
+
"grad_norm": 14.483268737792969,
|
266 |
+
"learning_rate": 4.033346918727989e-05,
|
267 |
+
"loss": 2.7018,
|
268 |
+
"step": 18500
|
269 |
+
},
|
270 |
+
{
|
271 |
+
"epoch": 0.5956673041351851,
|
272 |
+
"grad_norm": 15.697615623474121,
|
273 |
+
"learning_rate": 4.007221159774692e-05,
|
274 |
+
"loss": 2.6693,
|
275 |
+
"step": 19000
|
276 |
+
},
|
277 |
+
{
|
278 |
+
"epoch": 0.6113427595071637,
|
279 |
+
"grad_norm": 23.107515335083008,
|
280 |
+
"learning_rate": 3.981095400821394e-05,
|
281 |
+
"loss": 2.6445,
|
282 |
+
"step": 19500
|
283 |
+
},
|
284 |
+
{
|
285 |
+
"epoch": 0.6270182148791422,
|
286 |
+
"grad_norm": 15.768012046813965,
|
287 |
+
"learning_rate": 3.954969641868097e-05,
|
288 |
+
"loss": 2.589,
|
289 |
+
"step": 20000
|
290 |
+
},
|
291 |
+
{
|
292 |
+
"epoch": 0.6426936702511208,
|
293 |
+
"grad_norm": 24.025400161743164,
|
294 |
+
"learning_rate": 3.928843882914799e-05,
|
295 |
+
"loss": 2.6753,
|
296 |
+
"step": 20500
|
297 |
+
},
|
298 |
+
{
|
299 |
+
"epoch": 0.6583691256230993,
|
300 |
+
"grad_norm": 13.370205879211426,
|
301 |
+
"learning_rate": 3.9027181239615015e-05,
|
302 |
+
"loss": 2.5993,
|
303 |
+
"step": 21000
|
304 |
+
},
|
305 |
+
{
|
306 |
+
"epoch": 0.6740445809950779,
|
307 |
+
"grad_norm": 15.455653190612793,
|
308 |
+
"learning_rate": 3.8765923650082036e-05,
|
309 |
+
"loss": 2.6198,
|
310 |
+
"step": 21500
|
311 |
+
},
|
312 |
+
{
|
313 |
+
"epoch": 0.6897200363670565,
|
314 |
+
"grad_norm": 23.32875633239746,
|
315 |
+
"learning_rate": 3.8504666060549064e-05,
|
316 |
+
"loss": 2.6426,
|
317 |
+
"step": 22000
|
318 |
+
},
|
319 |
+
{
|
320 |
+
"epoch": 0.705395491739035,
|
321 |
+
"grad_norm": 23.468122482299805,
|
322 |
+
"learning_rate": 3.8243408471016085e-05,
|
323 |
+
"loss": 2.6068,
|
324 |
+
"step": 22500
|
325 |
+
},
|
326 |
+
{
|
327 |
+
"epoch": 0.7210709471110136,
|
328 |
+
"grad_norm": 16.411842346191406,
|
329 |
+
"learning_rate": 3.7982150881483106e-05,
|
330 |
+
"loss": 2.5992,
|
331 |
+
"step": 23000
|
332 |
+
},
|
333 |
+
{
|
334 |
+
"epoch": 0.7367464024829922,
|
335 |
+
"grad_norm": 28.372682571411133,
|
336 |
+
"learning_rate": 3.772089329195013e-05,
|
337 |
+
"loss": 2.6194,
|
338 |
+
"step": 23500
|
339 |
+
},
|
340 |
+
{
|
341 |
+
"epoch": 0.7524218578549707,
|
342 |
+
"grad_norm": 16.02229118347168,
|
343 |
+
"learning_rate": 3.7459635702417154e-05,
|
344 |
+
"loss": 2.5892,
|
345 |
+
"step": 24000
|
346 |
+
},
|
347 |
+
{
|
348 |
+
"epoch": 0.7680973132269493,
|
349 |
+
"grad_norm": 19.23042106628418,
|
350 |
+
"learning_rate": 3.719837811288418e-05,
|
351 |
+
"loss": 2.5984,
|
352 |
+
"step": 24500
|
353 |
+
},
|
354 |
+
{
|
355 |
+
"epoch": 0.7837727685989277,
|
356 |
+
"grad_norm": 23.105224609375,
|
357 |
+
"learning_rate": 3.69371205233512e-05,
|
358 |
+
"loss": 2.5461,
|
359 |
+
"step": 25000
|
360 |
+
},
|
361 |
+
{
|
362 |
+
"epoch": 0.7994482239709063,
|
363 |
+
"grad_norm": 12.95092487335205,
|
364 |
+
"learning_rate": 3.667586293381823e-05,
|
365 |
+
"loss": 2.5959,
|
366 |
+
"step": 25500
|
367 |
+
},
|
368 |
+
{
|
369 |
+
"epoch": 0.8151236793428849,
|
370 |
+
"grad_norm": 25.173099517822266,
|
371 |
+
"learning_rate": 3.641460534428525e-05,
|
372 |
+
"loss": 2.6026,
|
373 |
+
"step": 26000
|
374 |
+
},
|
375 |
+
{
|
376 |
+
"epoch": 0.8307991347148634,
|
377 |
+
"grad_norm": 16.152605056762695,
|
378 |
+
"learning_rate": 3.615334775475228e-05,
|
379 |
+
"loss": 2.5875,
|
380 |
+
"step": 26500
|
381 |
+
},
|
382 |
+
{
|
383 |
+
"epoch": 0.846474590086842,
|
384 |
+
"grad_norm": 14.907052993774414,
|
385 |
+
"learning_rate": 3.58920901652193e-05,
|
386 |
+
"loss": 2.6018,
|
387 |
+
"step": 27000
|
388 |
+
},
|
389 |
+
{
|
390 |
+
"epoch": 0.8621500454588206,
|
391 |
+
"grad_norm": 16.335224151611328,
|
392 |
+
"learning_rate": 3.563083257568632e-05,
|
393 |
+
"loss": 2.5612,
|
394 |
+
"step": 27500
|
395 |
+
},
|
396 |
+
{
|
397 |
+
"epoch": 0.8778255008307991,
|
398 |
+
"grad_norm": 15.978832244873047,
|
399 |
+
"learning_rate": 3.536957498615335e-05,
|
400 |
+
"loss": 2.6013,
|
401 |
+
"step": 28000
|
402 |
+
},
|
403 |
+
{
|
404 |
+
"epoch": 0.8935009562027777,
|
405 |
+
"grad_norm": 17.37552261352539,
|
406 |
+
"learning_rate": 3.510831739662037e-05,
|
407 |
+
"loss": 2.5882,
|
408 |
+
"step": 28500
|
409 |
+
},
|
410 |
+
{
|
411 |
+
"epoch": 0.9091764115747563,
|
412 |
+
"grad_norm": 18.65119171142578,
|
413 |
+
"learning_rate": 3.48470598070874e-05,
|
414 |
+
"loss": 2.5373,
|
415 |
+
"step": 29000
|
416 |
+
},
|
417 |
+
{
|
418 |
+
"epoch": 0.9248518669467348,
|
419 |
+
"grad_norm": 14.954788208007812,
|
420 |
+
"learning_rate": 3.458580221755442e-05,
|
421 |
+
"loss": 2.5115,
|
422 |
+
"step": 29500
|
423 |
+
},
|
424 |
+
{
|
425 |
+
"epoch": 0.9405273223187134,
|
426 |
+
"grad_norm": 15.86545467376709,
|
427 |
+
"learning_rate": 3.4324544628021445e-05,
|
428 |
+
"loss": 2.5989,
|
429 |
+
"step": 30000
|
430 |
+
},
|
431 |
+
{
|
432 |
+
"epoch": 0.956202777690692,
|
433 |
+
"grad_norm": 15.822210311889648,
|
434 |
+
"learning_rate": 3.4063287038488466e-05,
|
435 |
+
"loss": 2.5548,
|
436 |
+
"step": 30500
|
437 |
+
},
|
438 |
+
{
|
439 |
+
"epoch": 0.9718782330626705,
|
440 |
+
"grad_norm": 18.378137588500977,
|
441 |
+
"learning_rate": 3.3802029448955494e-05,
|
442 |
+
"loss": 2.5634,
|
443 |
+
"step": 31000
|
444 |
+
},
|
445 |
+
{
|
446 |
+
"epoch": 0.9875536884346491,
|
447 |
+
"grad_norm": 18.04598045349121,
|
448 |
+
"learning_rate": 3.3540771859422515e-05,
|
449 |
+
"loss": 2.5167,
|
450 |
+
"step": 31500
|
451 |
+
},
|
452 |
+
{
|
453 |
+
"epoch": 1.0032291438066276,
|
454 |
+
"grad_norm": 13.649733543395996,
|
455 |
+
"learning_rate": 3.327951426988954e-05,
|
456 |
+
"loss": 2.4968,
|
457 |
+
"step": 32000
|
458 |
+
},
|
459 |
+
{
|
460 |
+
"epoch": 1.018904599178606,
|
461 |
+
"grad_norm": 30.151395797729492,
|
462 |
+
"learning_rate": 3.301825668035657e-05,
|
463 |
+
"loss": 2.4899,
|
464 |
+
"step": 32500
|
465 |
+
},
|
466 |
+
{
|
467 |
+
"epoch": 1.0345800545505848,
|
468 |
+
"grad_norm": 18.006633758544922,
|
469 |
+
"learning_rate": 3.275699909082359e-05,
|
470 |
+
"loss": 2.513,
|
471 |
+
"step": 33000
|
472 |
+
},
|
473 |
+
{
|
474 |
+
"epoch": 1.0502555099225632,
|
475 |
+
"grad_norm": 18.782691955566406,
|
476 |
+
"learning_rate": 3.249574150129062e-05,
|
477 |
+
"loss": 2.4657,
|
478 |
+
"step": 33500
|
479 |
+
},
|
480 |
+
{
|
481 |
+
"epoch": 1.065930965294542,
|
482 |
+
"grad_norm": 14.228589057922363,
|
483 |
+
"learning_rate": 3.223448391175764e-05,
|
484 |
+
"loss": 2.4519,
|
485 |
+
"step": 34000
|
486 |
+
},
|
487 |
+
{
|
488 |
+
"epoch": 1.0816064206665204,
|
489 |
+
"grad_norm": 14.239832878112793,
|
490 |
+
"learning_rate": 3.197322632222467e-05,
|
491 |
+
"loss": 2.3959,
|
492 |
+
"step": 34500
|
493 |
+
},
|
494 |
+
{
|
495 |
+
"epoch": 1.0972818760384988,
|
496 |
+
"grad_norm": 20.227951049804688,
|
497 |
+
"learning_rate": 3.171196873269169e-05,
|
498 |
+
"loss": 2.4191,
|
499 |
+
"step": 35000
|
500 |
+
},
|
501 |
+
{
|
502 |
+
"epoch": 1.1129573314104775,
|
503 |
+
"grad_norm": 16.465883255004883,
|
504 |
+
"learning_rate": 3.145071114315871e-05,
|
505 |
+
"loss": 2.5173,
|
506 |
+
"step": 35500
|
507 |
+
},
|
508 |
+
{
|
509 |
+
"epoch": 1.1286327867824562,
|
510 |
+
"grad_norm": 28.69437599182129,
|
511 |
+
"learning_rate": 3.118945355362574e-05,
|
512 |
+
"loss": 2.5291,
|
513 |
+
"step": 36000
|
514 |
+
},
|
515 |
+
{
|
516 |
+
"epoch": 1.1443082421544346,
|
517 |
+
"grad_norm": 17.486122131347656,
|
518 |
+
"learning_rate": 3.092819596409276e-05,
|
519 |
+
"loss": 2.5009,
|
520 |
+
"step": 36500
|
521 |
+
},
|
522 |
+
{
|
523 |
+
"epoch": 1.159983697526413,
|
524 |
+
"grad_norm": 15.846352577209473,
|
525 |
+
"learning_rate": 3.0666938374559785e-05,
|
526 |
+
"loss": 2.4499,
|
527 |
+
"step": 37000
|
528 |
+
},
|
529 |
+
{
|
530 |
+
"epoch": 1.1756591528983917,
|
531 |
+
"grad_norm": 17.680450439453125,
|
532 |
+
"learning_rate": 3.0405680785026806e-05,
|
533 |
+
"loss": 2.4111,
|
534 |
+
"step": 37500
|
535 |
+
},
|
536 |
+
{
|
537 |
+
"epoch": 1.1913346082703702,
|
538 |
+
"grad_norm": 23.131710052490234,
|
539 |
+
"learning_rate": 3.014442319549383e-05,
|
540 |
+
"loss": 2.4946,
|
541 |
+
"step": 38000
|
542 |
+
},
|
543 |
+
{
|
544 |
+
"epoch": 1.2070100636423489,
|
545 |
+
"grad_norm": 23.71167755126953,
|
546 |
+
"learning_rate": 2.9883165605960855e-05,
|
547 |
+
"loss": 2.458,
|
548 |
+
"step": 38500
|
549 |
+
},
|
550 |
+
{
|
551 |
+
"epoch": 1.2226855190143273,
|
552 |
+
"grad_norm": 17.49135971069336,
|
553 |
+
"learning_rate": 2.962190801642788e-05,
|
554 |
+
"loss": 2.4595,
|
555 |
+
"step": 39000
|
556 |
+
},
|
557 |
+
{
|
558 |
+
"epoch": 1.238360974386306,
|
559 |
+
"grad_norm": 15.8767728805542,
|
560 |
+
"learning_rate": 2.9360650426894903e-05,
|
561 |
+
"loss": 2.4972,
|
562 |
+
"step": 39500
|
563 |
+
},
|
564 |
+
{
|
565 |
+
"epoch": 1.2540364297582844,
|
566 |
+
"grad_norm": 24.60512924194336,
|
567 |
+
"learning_rate": 2.9099392837361928e-05,
|
568 |
+
"loss": 2.4128,
|
569 |
+
"step": 40000
|
570 |
+
},
|
571 |
+
{
|
572 |
+
"epoch": 1.269711885130263,
|
573 |
+
"grad_norm": 21.219585418701172,
|
574 |
+
"learning_rate": 2.8838135247828952e-05,
|
575 |
+
"loss": 2.4715,
|
576 |
+
"step": 40500
|
577 |
+
},
|
578 |
+
{
|
579 |
+
"epoch": 1.2853873405022416,
|
580 |
+
"grad_norm": 15.823412895202637,
|
581 |
+
"learning_rate": 2.8576877658295976e-05,
|
582 |
+
"loss": 2.4254,
|
583 |
+
"step": 41000
|
584 |
+
},
|
585 |
+
{
|
586 |
+
"epoch": 1.3010627958742202,
|
587 |
+
"grad_norm": 15.92619514465332,
|
588 |
+
"learning_rate": 2.8315620068763e-05,
|
589 |
+
"loss": 2.4241,
|
590 |
+
"step": 41500
|
591 |
+
},
|
592 |
+
{
|
593 |
+
"epoch": 1.3167382512461987,
|
594 |
+
"grad_norm": 18.505760192871094,
|
595 |
+
"learning_rate": 2.805436247923002e-05,
|
596 |
+
"loss": 2.4412,
|
597 |
+
"step": 42000
|
598 |
+
},
|
599 |
+
{
|
600 |
+
"epoch": 1.3324137066181772,
|
601 |
+
"grad_norm": 18.413311004638672,
|
602 |
+
"learning_rate": 2.7793104889697046e-05,
|
603 |
+
"loss": 2.4697,
|
604 |
+
"step": 42500
|
605 |
+
},
|
606 |
+
{
|
607 |
+
"epoch": 1.3480891619901558,
|
608 |
+
"grad_norm": 20.015159606933594,
|
609 |
+
"learning_rate": 2.753184730016407e-05,
|
610 |
+
"loss": 2.4047,
|
611 |
+
"step": 43000
|
612 |
+
},
|
613 |
+
{
|
614 |
+
"epoch": 1.3637646173621345,
|
615 |
+
"grad_norm": 20.301258087158203,
|
616 |
+
"learning_rate": 2.7270589710631094e-05,
|
617 |
+
"loss": 2.4139,
|
618 |
+
"step": 43500
|
619 |
+
},
|
620 |
+
{
|
621 |
+
"epoch": 1.379440072734113,
|
622 |
+
"grad_norm": 14.145316123962402,
|
623 |
+
"learning_rate": 2.700933212109812e-05,
|
624 |
+
"loss": 2.4226,
|
625 |
+
"step": 44000
|
626 |
+
},
|
627 |
+
{
|
628 |
+
"epoch": 1.3951155281060914,
|
629 |
+
"grad_norm": 12.221750259399414,
|
630 |
+
"learning_rate": 2.6748074531565143e-05,
|
631 |
+
"loss": 2.3895,
|
632 |
+
"step": 44500
|
633 |
+
},
|
634 |
+
{
|
635 |
+
"epoch": 1.41079098347807,
|
636 |
+
"grad_norm": 15.449772834777832,
|
637 |
+
"learning_rate": 2.6486816942032167e-05,
|
638 |
+
"loss": 2.4323,
|
639 |
+
"step": 45000
|
640 |
+
},
|
641 |
+
{
|
642 |
+
"epoch": 1.4264664388500485,
|
643 |
+
"grad_norm": 15.565759658813477,
|
644 |
+
"learning_rate": 2.622555935249919e-05,
|
645 |
+
"loss": 2.4612,
|
646 |
+
"step": 45500
|
647 |
+
},
|
648 |
+
{
|
649 |
+
"epoch": 1.4421418942220272,
|
650 |
+
"grad_norm": 12.723085403442383,
|
651 |
+
"learning_rate": 2.5964301762966216e-05,
|
652 |
+
"loss": 2.404,
|
653 |
+
"step": 46000
|
654 |
+
},
|
655 |
+
{
|
656 |
+
"epoch": 1.4578173495940057,
|
657 |
+
"grad_norm": 27.950130462646484,
|
658 |
+
"learning_rate": 2.570304417343324e-05,
|
659 |
+
"loss": 2.4505,
|
660 |
+
"step": 46500
|
661 |
+
},
|
662 |
+
{
|
663 |
+
"epoch": 1.4734928049659843,
|
664 |
+
"grad_norm": 16.8266544342041,
|
665 |
+
"learning_rate": 2.544178658390026e-05,
|
666 |
+
"loss": 2.4143,
|
667 |
+
"step": 47000
|
668 |
+
},
|
669 |
+
{
|
670 |
+
"epoch": 1.4891682603379628,
|
671 |
+
"grad_norm": 18.774965286254883,
|
672 |
+
"learning_rate": 2.5180528994367285e-05,
|
673 |
+
"loss": 2.4196,
|
674 |
+
"step": 47500
|
675 |
+
},
|
676 |
+
{
|
677 |
+
"epoch": 1.5048437157099412,
|
678 |
+
"grad_norm": 19.893604278564453,
|
679 |
+
"learning_rate": 2.4919271404834313e-05,
|
680 |
+
"loss": 2.458,
|
681 |
+
"step": 48000
|
682 |
+
},
|
683 |
+
{
|
684 |
+
"epoch": 1.52051917108192,
|
685 |
+
"grad_norm": 17.030399322509766,
|
686 |
+
"learning_rate": 2.4658013815301337e-05,
|
687 |
+
"loss": 2.4234,
|
688 |
+
"step": 48500
|
689 |
+
},
|
690 |
+
{
|
691 |
+
"epoch": 1.5361946264538986,
|
692 |
+
"grad_norm": 15.073714256286621,
|
693 |
+
"learning_rate": 2.439675622576836e-05,
|
694 |
+
"loss": 2.358,
|
695 |
+
"step": 49000
|
696 |
+
},
|
697 |
+
{
|
698 |
+
"epoch": 1.551870081825877,
|
699 |
+
"grad_norm": 15.815202713012695,
|
700 |
+
"learning_rate": 2.4135498636235385e-05,
|
701 |
+
"loss": 2.3797,
|
702 |
+
"step": 49500
|
703 |
+
},
|
704 |
+
{
|
705 |
+
"epoch": 1.5675455371978555,
|
706 |
+
"grad_norm": 12.330552101135254,
|
707 |
+
"learning_rate": 2.387424104670241e-05,
|
708 |
+
"loss": 2.3955,
|
709 |
+
"step": 50000
|
710 |
+
},
|
711 |
+
{
|
712 |
+
"epoch": 1.5832209925698342,
|
713 |
+
"grad_norm": 20.456411361694336,
|
714 |
+
"learning_rate": 2.3612983457169434e-05,
|
715 |
+
"loss": 2.3957,
|
716 |
+
"step": 50500
|
717 |
+
},
|
718 |
+
{
|
719 |
+
"epoch": 1.5988964479418128,
|
720 |
+
"grad_norm": 13.23128604888916,
|
721 |
+
"learning_rate": 2.3351725867636455e-05,
|
722 |
+
"loss": 2.4407,
|
723 |
+
"step": 51000
|
724 |
+
},
|
725 |
+
{
|
726 |
+
"epoch": 1.6145719033137913,
|
727 |
+
"grad_norm": 24.75179100036621,
|
728 |
+
"learning_rate": 2.309046827810348e-05,
|
729 |
+
"loss": 2.3367,
|
730 |
+
"step": 51500
|
731 |
+
},
|
732 |
+
{
|
733 |
+
"epoch": 1.6302473586857698,
|
734 |
+
"grad_norm": 14.501784324645996,
|
735 |
+
"learning_rate": 2.2829210688570503e-05,
|
736 |
+
"loss": 2.4223,
|
737 |
+
"step": 52000
|
738 |
+
},
|
739 |
+
{
|
740 |
+
"epoch": 1.6459228140577484,
|
741 |
+
"grad_norm": 18.642641067504883,
|
742 |
+
"learning_rate": 2.2567953099037528e-05,
|
743 |
+
"loss": 2.3998,
|
744 |
+
"step": 52500
|
745 |
+
},
|
746 |
+
{
|
747 |
+
"epoch": 1.6615982694297269,
|
748 |
+
"grad_norm": 19.8629207611084,
|
749 |
+
"learning_rate": 2.2306695509504552e-05,
|
750 |
+
"loss": 2.3973,
|
751 |
+
"step": 53000
|
752 |
+
},
|
753 |
+
{
|
754 |
+
"epoch": 1.6772737248017053,
|
755 |
+
"grad_norm": 14.32746696472168,
|
756 |
+
"learning_rate": 2.2045437919971576e-05,
|
757 |
+
"loss": 2.3911,
|
758 |
+
"step": 53500
|
759 |
+
},
|
760 |
+
{
|
761 |
+
"epoch": 1.692949180173684,
|
762 |
+
"grad_norm": 21.35317039489746,
|
763 |
+
"learning_rate": 2.17841803304386e-05,
|
764 |
+
"loss": 2.3891,
|
765 |
+
"step": 54000
|
766 |
+
},
|
767 |
+
{
|
768 |
+
"epoch": 1.7086246355456627,
|
769 |
+
"grad_norm": 21.998720169067383,
|
770 |
+
"learning_rate": 2.1522922740905625e-05,
|
771 |
+
"loss": 2.3615,
|
772 |
+
"step": 54500
|
773 |
+
},
|
774 |
+
{
|
775 |
+
"epoch": 1.7243000909176411,
|
776 |
+
"grad_norm": 16.640851974487305,
|
777 |
+
"learning_rate": 2.126166515137265e-05,
|
778 |
+
"loss": 2.4245,
|
779 |
+
"step": 55000
|
780 |
+
},
|
781 |
+
{
|
782 |
+
"epoch": 1.7399755462896196,
|
783 |
+
"grad_norm": 15.640762329101562,
|
784 |
+
"learning_rate": 2.100040756183967e-05,
|
785 |
+
"loss": 2.3663,
|
786 |
+
"step": 55500
|
787 |
+
},
|
788 |
+
{
|
789 |
+
"epoch": 1.7556510016615983,
|
790 |
+
"grad_norm": 25.16714859008789,
|
791 |
+
"learning_rate": 2.0739149972306694e-05,
|
792 |
+
"loss": 2.3653,
|
793 |
+
"step": 56000
|
794 |
+
},
|
795 |
+
{
|
796 |
+
"epoch": 1.771326457033577,
|
797 |
+
"grad_norm": 13.939558029174805,
|
798 |
+
"learning_rate": 2.047789238277372e-05,
|
799 |
+
"loss": 2.3162,
|
800 |
+
"step": 56500
|
801 |
+
},
|
802 |
+
{
|
803 |
+
"epoch": 1.7870019124055554,
|
804 |
+
"grad_norm": 17.95723533630371,
|
805 |
+
"learning_rate": 2.0216634793240743e-05,
|
806 |
+
"loss": 2.3965,
|
807 |
+
"step": 57000
|
808 |
+
},
|
809 |
+
{
|
810 |
+
"epoch": 1.8026773677775338,
|
811 |
+
"grad_norm": 16.54608917236328,
|
812 |
+
"learning_rate": 1.9955377203707767e-05,
|
813 |
+
"loss": 2.3554,
|
814 |
+
"step": 57500
|
815 |
+
},
|
816 |
+
{
|
817 |
+
"epoch": 1.8183528231495125,
|
818 |
+
"grad_norm": 15.683633804321289,
|
819 |
+
"learning_rate": 1.969411961417479e-05,
|
820 |
+
"loss": 2.3475,
|
821 |
+
"step": 58000
|
822 |
+
},
|
823 |
+
{
|
824 |
+
"epoch": 1.8340282785214912,
|
825 |
+
"grad_norm": 15.327407836914062,
|
826 |
+
"learning_rate": 1.943286202464182e-05,
|
827 |
+
"loss": 2.3796,
|
828 |
+
"step": 58500
|
829 |
+
},
|
830 |
+
{
|
831 |
+
"epoch": 1.8497037338934696,
|
832 |
+
"grad_norm": 18.815610885620117,
|
833 |
+
"learning_rate": 1.9171604435108843e-05,
|
834 |
+
"loss": 2.3763,
|
835 |
+
"step": 59000
|
836 |
+
},
|
837 |
+
{
|
838 |
+
"epoch": 1.865379189265448,
|
839 |
+
"grad_norm": 18.846534729003906,
|
840 |
+
"learning_rate": 1.8910346845575864e-05,
|
841 |
+
"loss": 2.3869,
|
842 |
+
"step": 59500
|
843 |
+
},
|
844 |
+
{
|
845 |
+
"epoch": 1.8810546446374268,
|
846 |
+
"grad_norm": 24.27644157409668,
|
847 |
+
"learning_rate": 1.864908925604289e-05,
|
848 |
+
"loss": 2.3079,
|
849 |
+
"step": 60000
|
850 |
+
},
|
851 |
+
{
|
852 |
+
"epoch": 1.8967301000094052,
|
853 |
+
"grad_norm": 19.45649528503418,
|
854 |
+
"learning_rate": 1.8387831666509913e-05,
|
855 |
+
"loss": 2.379,
|
856 |
+
"step": 60500
|
857 |
+
},
|
858 |
+
{
|
859 |
+
"epoch": 1.9124055553813837,
|
860 |
+
"grad_norm": 19.247732162475586,
|
861 |
+
"learning_rate": 1.8126574076976937e-05,
|
862 |
+
"loss": 2.3454,
|
863 |
+
"step": 61000
|
864 |
+
},
|
865 |
+
{
|
866 |
+
"epoch": 1.9280810107533624,
|
867 |
+
"grad_norm": 16.541532516479492,
|
868 |
+
"learning_rate": 1.786531648744396e-05,
|
869 |
+
"loss": 2.3611,
|
870 |
+
"step": 61500
|
871 |
+
},
|
872 |
+
{
|
873 |
+
"epoch": 1.943756466125341,
|
874 |
+
"grad_norm": 19.154760360717773,
|
875 |
+
"learning_rate": 1.7604058897910986e-05,
|
876 |
+
"loss": 2.2952,
|
877 |
+
"step": 62000
|
878 |
+
},
|
879 |
+
{
|
880 |
+
"epoch": 1.9594319214973195,
|
881 |
+
"grad_norm": 13.325096130371094,
|
882 |
+
"learning_rate": 1.734280130837801e-05,
|
883 |
+
"loss": 2.3158,
|
884 |
+
"step": 62500
|
885 |
+
},
|
886 |
+
{
|
887 |
+
"epoch": 1.975107376869298,
|
888 |
+
"grad_norm": 12.556410789489746,
|
889 |
+
"learning_rate": 1.7081543718845034e-05,
|
890 |
+
"loss": 2.3199,
|
891 |
+
"step": 63000
|
892 |
+
},
|
893 |
+
{
|
894 |
+
"epoch": 1.9907828322412766,
|
895 |
+
"grad_norm": 17.374223709106445,
|
896 |
+
"learning_rate": 1.682028612931206e-05,
|
897 |
+
"loss": 2.3241,
|
898 |
+
"step": 63500
|
899 |
+
},
|
900 |
+
{
|
901 |
+
"epoch": 2.0064582876132553,
|
902 |
+
"grad_norm": 23.0031795501709,
|
903 |
+
"learning_rate": 1.6559028539779083e-05,
|
904 |
+
"loss": 2.2653,
|
905 |
+
"step": 64000
|
906 |
+
},
|
907 |
+
{
|
908 |
+
"epoch": 2.0221337429852335,
|
909 |
+
"grad_norm": 24.6336612701416,
|
910 |
+
"learning_rate": 1.6297770950246104e-05,
|
911 |
+
"loss": 2.3009,
|
912 |
+
"step": 64500
|
913 |
+
},
|
914 |
+
{
|
915 |
+
"epoch": 2.037809198357212,
|
916 |
+
"grad_norm": 21.2884521484375,
|
917 |
+
"learning_rate": 1.6036513360713128e-05,
|
918 |
+
"loss": 2.3493,
|
919 |
+
"step": 65000
|
920 |
+
},
|
921 |
+
{
|
922 |
+
"epoch": 2.053484653729191,
|
923 |
+
"grad_norm": 13.374622344970703,
|
924 |
+
"learning_rate": 1.5775255771180152e-05,
|
925 |
+
"loss": 2.2815,
|
926 |
+
"step": 65500
|
927 |
+
},
|
928 |
+
{
|
929 |
+
"epoch": 2.0691601091011695,
|
930 |
+
"grad_norm": 20.058076858520508,
|
931 |
+
"learning_rate": 1.5513998181647176e-05,
|
932 |
+
"loss": 2.2794,
|
933 |
+
"step": 66000
|
934 |
+
},
|
935 |
+
{
|
936 |
+
"epoch": 2.0848355644731478,
|
937 |
+
"grad_norm": 22.910066604614258,
|
938 |
+
"learning_rate": 1.52527405921142e-05,
|
939 |
+
"loss": 2.3015,
|
940 |
+
"step": 66500
|
941 |
+
},
|
942 |
+
{
|
943 |
+
"epoch": 2.1005110198451264,
|
944 |
+
"grad_norm": 22.905941009521484,
|
945 |
+
"learning_rate": 1.4991483002581225e-05,
|
946 |
+
"loss": 2.2263,
|
947 |
+
"step": 67000
|
948 |
+
},
|
949 |
+
{
|
950 |
+
"epoch": 2.116186475217105,
|
951 |
+
"grad_norm": 15.105734825134277,
|
952 |
+
"learning_rate": 1.473022541304825e-05,
|
953 |
+
"loss": 2.2947,
|
954 |
+
"step": 67500
|
955 |
+
},
|
956 |
+
{
|
957 |
+
"epoch": 2.131861930589084,
|
958 |
+
"grad_norm": 23.665620803833008,
|
959 |
+
"learning_rate": 1.4468967823515272e-05,
|
960 |
+
"loss": 2.2644,
|
961 |
+
"step": 68000
|
962 |
+
},
|
963 |
+
{
|
964 |
+
"epoch": 2.147537385961062,
|
965 |
+
"grad_norm": 13.619988441467285,
|
966 |
+
"learning_rate": 1.4207710233982296e-05,
|
967 |
+
"loss": 2.3118,
|
968 |
+
"step": 68500
|
969 |
+
},
|
970 |
+
{
|
971 |
+
"epoch": 2.1632128413330407,
|
972 |
+
"grad_norm": 21.304323196411133,
|
973 |
+
"learning_rate": 1.3946452644449324e-05,
|
974 |
+
"loss": 2.2925,
|
975 |
+
"step": 69000
|
976 |
+
},
|
977 |
+
{
|
978 |
+
"epoch": 2.1788882967050194,
|
979 |
+
"grad_norm": 11.025732040405273,
|
980 |
+
"learning_rate": 1.3685195054916346e-05,
|
981 |
+
"loss": 2.2859,
|
982 |
+
"step": 69500
|
983 |
+
},
|
984 |
+
{
|
985 |
+
"epoch": 2.1945637520769976,
|
986 |
+
"grad_norm": 20.443811416625977,
|
987 |
+
"learning_rate": 1.342393746538337e-05,
|
988 |
+
"loss": 2.3155,
|
989 |
+
"step": 70000
|
990 |
+
},
|
991 |
+
{
|
992 |
+
"epoch": 2.2102392074489763,
|
993 |
+
"grad_norm": 19.622135162353516,
|
994 |
+
"learning_rate": 1.3162679875850395e-05,
|
995 |
+
"loss": 2.3166,
|
996 |
+
"step": 70500
|
997 |
+
},
|
998 |
+
{
|
999 |
+
"epoch": 2.225914662820955,
|
1000 |
+
"grad_norm": 17.77559471130371,
|
1001 |
+
"learning_rate": 1.290142228631742e-05,
|
1002 |
+
"loss": 2.2971,
|
1003 |
+
"step": 71000
|
1004 |
+
},
|
1005 |
+
{
|
1006 |
+
"epoch": 2.2415901181929336,
|
1007 |
+
"grad_norm": 20.948503494262695,
|
1008 |
+
"learning_rate": 1.2640164696784444e-05,
|
1009 |
+
"loss": 2.2901,
|
1010 |
+
"step": 71500
|
1011 |
+
},
|
1012 |
+
{
|
1013 |
+
"epoch": 2.2572655735649123,
|
1014 |
+
"grad_norm": 19.734085083007812,
|
1015 |
+
"learning_rate": 1.2378907107251466e-05,
|
1016 |
+
"loss": 2.2866,
|
1017 |
+
"step": 72000
|
1018 |
+
},
|
1019 |
+
{
|
1020 |
+
"epoch": 2.2729410289368905,
|
1021 |
+
"grad_norm": 14.341938018798828,
|
1022 |
+
"learning_rate": 1.211764951771849e-05,
|
1023 |
+
"loss": 2.2806,
|
1024 |
+
"step": 72500
|
1025 |
+
},
|
1026 |
+
{
|
1027 |
+
"epoch": 2.288616484308869,
|
1028 |
+
"grad_norm": 16.937484741210938,
|
1029 |
+
"learning_rate": 1.1856391928185515e-05,
|
1030 |
+
"loss": 2.2824,
|
1031 |
+
"step": 73000
|
1032 |
+
},
|
1033 |
+
{
|
1034 |
+
"epoch": 2.304291939680848,
|
1035 |
+
"grad_norm": 12.707877159118652,
|
1036 |
+
"learning_rate": 1.1595134338652539e-05,
|
1037 |
+
"loss": 2.2768,
|
1038 |
+
"step": 73500
|
1039 |
+
},
|
1040 |
+
{
|
1041 |
+
"epoch": 2.319967395052826,
|
1042 |
+
"grad_norm": 12.61535358428955,
|
1043 |
+
"learning_rate": 1.1333876749119563e-05,
|
1044 |
+
"loss": 2.2395,
|
1045 |
+
"step": 74000
|
1046 |
+
},
|
1047 |
+
{
|
1048 |
+
"epoch": 2.335642850424805,
|
1049 |
+
"grad_norm": 29.96654510498047,
|
1050 |
+
"learning_rate": 1.1072619159586586e-05,
|
1051 |
+
"loss": 2.2204,
|
1052 |
+
"step": 74500
|
1053 |
+
},
|
1054 |
+
{
|
1055 |
+
"epoch": 2.3513183057967835,
|
1056 |
+
"grad_norm": 14.348071098327637,
|
1057 |
+
"learning_rate": 1.081136157005361e-05,
|
1058 |
+
"loss": 2.2712,
|
1059 |
+
"step": 75000
|
1060 |
+
},
|
1061 |
+
{
|
1062 |
+
"epoch": 2.366993761168762,
|
1063 |
+
"grad_norm": 16.751550674438477,
|
1064 |
+
"learning_rate": 1.0550103980520634e-05,
|
1065 |
+
"loss": 2.1824,
|
1066 |
+
"step": 75500
|
1067 |
+
},
|
1068 |
+
{
|
1069 |
+
"epoch": 2.3826692165407404,
|
1070 |
+
"grad_norm": 13.812084197998047,
|
1071 |
+
"learning_rate": 1.0288846390987659e-05,
|
1072 |
+
"loss": 2.1823,
|
1073 |
+
"step": 76000
|
1074 |
+
},
|
1075 |
+
{
|
1076 |
+
"epoch": 2.398344671912719,
|
1077 |
+
"grad_norm": 19.040403366088867,
|
1078 |
+
"learning_rate": 1.0027588801454681e-05,
|
1079 |
+
"loss": 2.2536,
|
1080 |
+
"step": 76500
|
1081 |
+
},
|
1082 |
+
{
|
1083 |
+
"epoch": 2.4140201272846977,
|
1084 |
+
"grad_norm": 21.43050765991211,
|
1085 |
+
"learning_rate": 9.766331211921707e-06,
|
1086 |
+
"loss": 2.247,
|
1087 |
+
"step": 77000
|
1088 |
+
},
|
1089 |
+
{
|
1090 |
+
"epoch": 2.4296955826566764,
|
1091 |
+
"grad_norm": 17.847736358642578,
|
1092 |
+
"learning_rate": 9.505073622388731e-06,
|
1093 |
+
"loss": 2.2038,
|
1094 |
+
"step": 77500
|
1095 |
+
},
|
1096 |
+
{
|
1097 |
+
"epoch": 2.4453710380286546,
|
1098 |
+
"grad_norm": 16.89531135559082,
|
1099 |
+
"learning_rate": 9.243816032855756e-06,
|
1100 |
+
"loss": 2.2927,
|
1101 |
+
"step": 78000
|
1102 |
+
},
|
1103 |
+
{
|
1104 |
+
"epoch": 2.4610464934006333,
|
1105 |
+
"grad_norm": 21.707191467285156,
|
1106 |
+
"learning_rate": 8.982558443322778e-06,
|
1107 |
+
"loss": 2.2851,
|
1108 |
+
"step": 78500
|
1109 |
+
},
|
1110 |
+
{
|
1111 |
+
"epoch": 2.476721948772612,
|
1112 |
+
"grad_norm": 20.218366622924805,
|
1113 |
+
"learning_rate": 8.721300853789803e-06,
|
1114 |
+
"loss": 2.2716,
|
1115 |
+
"step": 79000
|
1116 |
+
},
|
1117 |
+
{
|
1118 |
+
"epoch": 2.49239740414459,
|
1119 |
+
"grad_norm": 25.17759895324707,
|
1120 |
+
"learning_rate": 8.460043264256827e-06,
|
1121 |
+
"loss": 2.2357,
|
1122 |
+
"step": 79500
|
1123 |
+
},
|
1124 |
+
{
|
1125 |
+
"epoch": 2.508072859516569,
|
1126 |
+
"grad_norm": 17.571487426757812,
|
1127 |
+
"learning_rate": 8.198785674723851e-06,
|
1128 |
+
"loss": 2.1629,
|
1129 |
+
"step": 80000
|
1130 |
+
},
|
1131 |
+
{
|
1132 |
+
"epoch": 2.5237483148885476,
|
1133 |
+
"grad_norm": 19.578426361083984,
|
1134 |
+
"learning_rate": 7.937528085190875e-06,
|
1135 |
+
"loss": 2.2371,
|
1136 |
+
"step": 80500
|
1137 |
+
},
|
1138 |
+
{
|
1139 |
+
"epoch": 2.539423770260526,
|
1140 |
+
"grad_norm": 16.45771026611328,
|
1141 |
+
"learning_rate": 7.676270495657898e-06,
|
1142 |
+
"loss": 2.228,
|
1143 |
+
"step": 81000
|
1144 |
+
},
|
1145 |
+
{
|
1146 |
+
"epoch": 2.5550992256325045,
|
1147 |
+
"grad_norm": 19.219467163085938,
|
1148 |
+
"learning_rate": 7.415012906124922e-06,
|
1149 |
+
"loss": 2.2392,
|
1150 |
+
"step": 81500
|
1151 |
+
},
|
1152 |
+
{
|
1153 |
+
"epoch": 2.570774681004483,
|
1154 |
+
"grad_norm": 21.938257217407227,
|
1155 |
+
"learning_rate": 7.153755316591948e-06,
|
1156 |
+
"loss": 2.2117,
|
1157 |
+
"step": 82000
|
1158 |
+
},
|
1159 |
+
{
|
1160 |
+
"epoch": 2.586450136376462,
|
1161 |
+
"grad_norm": 18.676382064819336,
|
1162 |
+
"learning_rate": 6.892497727058972e-06,
|
1163 |
+
"loss": 2.2385,
|
1164 |
+
"step": 82500
|
1165 |
+
},
|
1166 |
+
{
|
1167 |
+
"epoch": 2.6021255917484405,
|
1168 |
+
"grad_norm": 15.486830711364746,
|
1169 |
+
"learning_rate": 6.631240137525996e-06,
|
1170 |
+
"loss": 2.1963,
|
1171 |
+
"step": 83000
|
1172 |
+
},
|
1173 |
+
{
|
1174 |
+
"epoch": 2.6178010471204187,
|
1175 |
+
"grad_norm": 17.053260803222656,
|
1176 |
+
"learning_rate": 6.369982547993019e-06,
|
1177 |
+
"loss": 2.2028,
|
1178 |
+
"step": 83500
|
1179 |
+
},
|
1180 |
+
{
|
1181 |
+
"epoch": 2.6334765024923974,
|
1182 |
+
"grad_norm": 13.611345291137695,
|
1183 |
+
"learning_rate": 6.108724958460044e-06,
|
1184 |
+
"loss": 2.1783,
|
1185 |
+
"step": 84000
|
1186 |
+
},
|
1187 |
+
{
|
1188 |
+
"epoch": 2.649151957864376,
|
1189 |
+
"grad_norm": 18.15059471130371,
|
1190 |
+
"learning_rate": 5.847467368927068e-06,
|
1191 |
+
"loss": 2.216,
|
1192 |
+
"step": 84500
|
1193 |
+
},
|
1194 |
+
{
|
1195 |
+
"epoch": 2.6648274132363543,
|
1196 |
+
"grad_norm": 14.26778793334961,
|
1197 |
+
"learning_rate": 5.586209779394091e-06,
|
1198 |
+
"loss": 2.2291,
|
1199 |
+
"step": 85000
|
1200 |
+
},
|
1201 |
+
{
|
1202 |
+
"epoch": 2.680502868608333,
|
1203 |
+
"grad_norm": 14.207221031188965,
|
1204 |
+
"learning_rate": 5.324952189861116e-06,
|
1205 |
+
"loss": 2.2301,
|
1206 |
+
"step": 85500
|
1207 |
+
},
|
1208 |
+
{
|
1209 |
+
"epoch": 2.6961783239803117,
|
1210 |
+
"grad_norm": 18.4339542388916,
|
1211 |
+
"learning_rate": 5.06369460032814e-06,
|
1212 |
+
"loss": 2.2454,
|
1213 |
+
"step": 86000
|
1214 |
+
},
|
1215 |
+
{
|
1216 |
+
"epoch": 2.7118537793522903,
|
1217 |
+
"grad_norm": 14.96839427947998,
|
1218 |
+
"learning_rate": 4.802437010795164e-06,
|
1219 |
+
"loss": 2.1817,
|
1220 |
+
"step": 86500
|
1221 |
+
},
|
1222 |
+
{
|
1223 |
+
"epoch": 2.727529234724269,
|
1224 |
+
"grad_norm": 15.77700424194336,
|
1225 |
+
"learning_rate": 4.541179421262188e-06,
|
1226 |
+
"loss": 2.1856,
|
1227 |
+
"step": 87000
|
1228 |
+
},
|
1229 |
+
{
|
1230 |
+
"epoch": 2.7432046900962472,
|
1231 |
+
"grad_norm": 18.85235023498535,
|
1232 |
+
"learning_rate": 4.279921831729212e-06,
|
1233 |
+
"loss": 2.168,
|
1234 |
+
"step": 87500
|
1235 |
+
},
|
1236 |
+
{
|
1237 |
+
"epoch": 2.758880145468226,
|
1238 |
+
"grad_norm": 26.718852996826172,
|
1239 |
+
"learning_rate": 4.018664242196235e-06,
|
1240 |
+
"loss": 2.2026,
|
1241 |
+
"step": 88000
|
1242 |
+
},
|
1243 |
+
{
|
1244 |
+
"epoch": 2.7745556008402046,
|
1245 |
+
"grad_norm": 14.873674392700195,
|
1246 |
+
"learning_rate": 3.7574066526632605e-06,
|
1247 |
+
"loss": 2.1595,
|
1248 |
+
"step": 88500
|
1249 |
+
},
|
1250 |
+
{
|
1251 |
+
"epoch": 2.790231056212183,
|
1252 |
+
"grad_norm": 21.835710525512695,
|
1253 |
+
"learning_rate": 3.4961490631302843e-06,
|
1254 |
+
"loss": 2.1986,
|
1255 |
+
"step": 89000
|
1256 |
+
},
|
1257 |
+
{
|
1258 |
+
"epoch": 2.8059065115841615,
|
1259 |
+
"grad_norm": 17.2098331451416,
|
1260 |
+
"learning_rate": 3.234891473597308e-06,
|
1261 |
+
"loss": 2.2289,
|
1262 |
+
"step": 89500
|
1263 |
+
},
|
1264 |
+
{
|
1265 |
+
"epoch": 2.82158196695614,
|
1266 |
+
"grad_norm": 15.5960111618042,
|
1267 |
+
"learning_rate": 2.973633884064332e-06,
|
1268 |
+
"loss": 2.2339,
|
1269 |
+
"step": 90000
|
1270 |
+
},
|
1271 |
+
{
|
1272 |
+
"epoch": 2.8372574223281184,
|
1273 |
+
"grad_norm": 14.342427253723145,
|
1274 |
+
"learning_rate": 2.7123762945313563e-06,
|
1275 |
+
"loss": 2.2128,
|
1276 |
+
"step": 90500
|
1277 |
+
},
|
1278 |
+
{
|
1279 |
+
"epoch": 2.852932877700097,
|
1280 |
+
"grad_norm": 25.278989791870117,
|
1281 |
+
"learning_rate": 2.45111870499838e-06,
|
1282 |
+
"loss": 2.2061,
|
1283 |
+
"step": 91000
|
1284 |
+
},
|
1285 |
+
{
|
1286 |
+
"epoch": 2.8686083330720757,
|
1287 |
+
"grad_norm": 22.495084762573242,
|
1288 |
+
"learning_rate": 2.1898611154654044e-06,
|
1289 |
+
"loss": 2.2115,
|
1290 |
+
"step": 91500
|
1291 |
+
},
|
1292 |
+
{
|
1293 |
+
"epoch": 2.8842837884440544,
|
1294 |
+
"grad_norm": 13.67044734954834,
|
1295 |
+
"learning_rate": 1.9286035259324287e-06,
|
1296 |
+
"loss": 2.1698,
|
1297 |
+
"step": 92000
|
1298 |
+
},
|
1299 |
+
{
|
1300 |
+
"epoch": 2.899959243816033,
|
1301 |
+
"grad_norm": 15.999963760375977,
|
1302 |
+
"learning_rate": 1.6673459363994524e-06,
|
1303 |
+
"loss": 2.2169,
|
1304 |
+
"step": 92500
|
1305 |
+
},
|
1306 |
+
{
|
1307 |
+
"epoch": 2.9156346991880113,
|
1308 |
+
"grad_norm": 25.277389526367188,
|
1309 |
+
"learning_rate": 1.4060883468664766e-06,
|
1310 |
+
"loss": 2.1946,
|
1311 |
+
"step": 93000
|
1312 |
+
},
|
1313 |
+
{
|
1314 |
+
"epoch": 2.93131015455999,
|
1315 |
+
"grad_norm": 15.683512687683105,
|
1316 |
+
"learning_rate": 1.1448307573335007e-06,
|
1317 |
+
"loss": 2.1701,
|
1318 |
+
"step": 93500
|
1319 |
+
},
|
1320 |
+
{
|
1321 |
+
"epoch": 2.9469856099319687,
|
1322 |
+
"grad_norm": 19.357261657714844,
|
1323 |
+
"learning_rate": 8.835731678005248e-07,
|
1324 |
+
"loss": 2.1759,
|
1325 |
+
"step": 94000
|
1326 |
+
},
|
1327 |
+
{
|
1328 |
+
"epoch": 2.962661065303947,
|
1329 |
+
"grad_norm": 13.178261756896973,
|
1330 |
+
"learning_rate": 6.223155782675487e-07,
|
1331 |
+
"loss": 2.1992,
|
1332 |
+
"step": 94500
|
1333 |
+
},
|
1334 |
+
{
|
1335 |
+
"epoch": 2.9783365206759256,
|
1336 |
+
"grad_norm": 19.615259170532227,
|
1337 |
+
"learning_rate": 3.6105798873457274e-07,
|
1338 |
+
"loss": 2.2599,
|
1339 |
+
"step": 95000
|
1340 |
+
},
|
1341 |
+
{
|
1342 |
+
"epoch": 2.9940119760479043,
|
1343 |
+
"grad_norm": 15.04670524597168,
|
1344 |
+
"learning_rate": 9.98003992015968e-08,
|
1345 |
+
"loss": 2.2026,
|
1346 |
+
"step": 95500
|
1347 |
+
}
|
1348 |
+
],
|
1349 |
+
"logging_steps": 500,
|
1350 |
+
"max_steps": 95691,
|
1351 |
+
"num_input_tokens_seen": 0,
|
1352 |
+
"num_train_epochs": 3,
|
1353 |
+
"save_steps": 10000,
|
1354 |
+
"stateful_callbacks": {
|
1355 |
+
"TrainerControl": {
|
1356 |
+
"args": {
|
1357 |
+
"should_epoch_stop": false,
|
1358 |
+
"should_evaluate": false,
|
1359 |
+
"should_log": false,
|
1360 |
+
"should_save": true,
|
1361 |
+
"should_training_stop": true
|
1362 |
+
},
|
1363 |
+
"attributes": {}
|
1364 |
+
}
|
1365 |
+
},
|
1366 |
+
"total_flos": 2.5369520913363456e+16,
|
1367 |
+
"train_batch_size": 8,
|
1368 |
+
"trial_name": null,
|
1369 |
+
"trial_params": null
|
1370 |
+
}
|
ConstructionBertV2/distilBertTrain_ConstructionManagement/checkpoint-95691/training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:b300cbf6a42b906b10e3e12d38967e1e12f1a1c7aa6f157987dbce7eb0215fff
|
3 |
+
size 5713
|
ConstructionBertV2/distilBert_ConstructionManagement/config.json
ADDED
@@ -0,0 +1,24 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "distilbert-base-uncased",
|
3 |
+
"activation": "gelu",
|
4 |
+
"architectures": [
|
5 |
+
"DistilBertForMaskedLM"
|
6 |
+
],
|
7 |
+
"attention_dropout": 0.1,
|
8 |
+
"dim": 768,
|
9 |
+
"dropout": 0.1,
|
10 |
+
"hidden_dim": 3072,
|
11 |
+
"initializer_range": 0.02,
|
12 |
+
"max_position_embeddings": 512,
|
13 |
+
"model_type": "distilbert",
|
14 |
+
"n_heads": 12,
|
15 |
+
"n_layers": 6,
|
16 |
+
"pad_token_id": 0,
|
17 |
+
"qa_dropout": 0.1,
|
18 |
+
"seq_classif_dropout": 0.2,
|
19 |
+
"sinusoidal_pos_embds": false,
|
20 |
+
"tie_weights_": true,
|
21 |
+
"torch_dtype": "float32",
|
22 |
+
"transformers_version": "4.49.0",
|
23 |
+
"vocab_size": 30522
|
24 |
+
}
|
ConstructionBertV2/distilBert_ConstructionManagement/model.safetensors
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:a3daa13429a8ee4beca3fa677cfd43454133c3399d2c1e70b7da0b098b2efabc
|
3 |
+
size 267954768
|
ConstructionBertV2/distilBert_ConstructionManagement/training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:b300cbf6a42b906b10e3e12d38967e1e12f1a1c7aa6f157987dbce7eb0215fff
|
3 |
+
size 5713
|