Hugging Face's logo

minseok0809
/
t5-small-scratch-iwslt2017

Text2Text Generation
Transformers
TensorBoard
Safetensors
t5
Generated from Trainer
text-generation-inference
Inference Endpoints
Model card Files Metrics Community
minseok0809 commited on
Commit
02c9e8d
•
1 Parent(s): 3c3c6c8

End of training

Browse files
Files changed (9) hide show
  1. README.md +6 -0
  2. all_results.json +23 -0
  3. eval_results.json +10 -0
  4. generated_predictions.txt +0 -0
  5. generation_config.json +0 -1
  6. predict_results.json +9 -0
  7. runs/May25_02-06-37_0d573eeffc83/events.out.tfevents.1716642100.0d573eeffc83.985437.1 +3 -0
  8. train_results.json +9 -0
  9. trainer_state.json +2296 -0
README.md CHANGED
@@ -3,6 +3,8 @@ license: apache-2.0
3
  base_model: google-t5/t5-small
4
  tags:
5
  - generated_from_trainer
 
 
6
  model-index:
7
  - name: t5-small-scratch-iwslt3007
8
  results: []
@@ -14,6 +16,10 @@ should probably proofread and complete it, then remove this comment. -->
14
  # t5-small-scratch-iwslt3007
15
 
16
  This model is a fine-tuned version of [google-t5/t5-small](https://huggingface.co/google-t5/t5-small) on an unknown dataset.
 
 
 
 
17
 
18
  ## Model description
19
 
 
3
  base_model: google-t5/t5-small
4
  tags:
5
  - generated_from_trainer
6
+ metrics:
7
+ - bleu
8
  model-index:
9
  - name: t5-small-scratch-iwslt3007
10
  results: []
 
16
  # t5-small-scratch-iwslt3007
17
 
18
  This model is a fine-tuned version of [google-t5/t5-small](https://huggingface.co/google-t5/t5-small) on an unknown dataset.
19
+ It achieves the following results on the evaluation set:
20
+ - Loss: 2.0647
21
+ - Bleu: 0.2786
22
+ - Gen Len: 26.3908
23
 
24
  ## Model description
25
 
all_results.json ADDED
@@ -0,0 +1,23 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 50.0,
3
+ "eval_bleu": 0.2786,
4
+ "eval_gen_len": 26.3908,
5
+ "eval_loss": 2.0646986961364746,
6
+ "eval_runtime": 5.0721,
7
+ "eval_samples": 888,
8
+ "eval_samples_per_second": 175.074,
9
+ "eval_steps_per_second": 1.38,
10
+ "predict_bleu": 0.2849,
11
+ "predict_gen_len": 22.3754,
12
+ "predict_loss": 2.105224132537842,
13
+ "predict_runtime": 36.2619,
14
+ "predict_samples": 8079,
15
+ "predict_samples_per_second": 222.796,
16
+ "predict_steps_per_second": 1.765,
17
+ "total_flos": 2.8277448966144e+17,
18
+ "train_loss": 2.6664713644307803,
19
+ "train_runtime": 39272.3209,
20
+ "train_samples": 206112,
21
+ "train_samples_per_second": 262.414,
22
+ "train_steps_per_second": 4.101
23
+ }
eval_results.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 50.0,
3
+ "eval_bleu": 0.2786,
4
+ "eval_gen_len": 26.3908,
5
+ "eval_loss": 2.0646986961364746,
6
+ "eval_runtime": 5.0721,
7
+ "eval_samples": 888,
8
+ "eval_samples_per_second": 175.074,
9
+ "eval_steps_per_second": 1.38
10
+ }
generated_predictions.txt ADDED
The diff for this file is too large to render. See raw diff
 
generation_config.json CHANGED
@@ -1,5 +1,4 @@
1
  {
2
- "_from_model_config": true,
3
  "decoder_start_token_id": 0,
4
  "eos_token_id": 1,
5
  "pad_token_id": 0,
 
1
  {
 
2
  "decoder_start_token_id": 0,
3
  "eos_token_id": 1,
4
  "pad_token_id": 0,
predict_results.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "predict_bleu": 0.2849,
3
+ "predict_gen_len": 22.3754,
4
+ "predict_loss": 2.105224132537842,
5
+ "predict_runtime": 36.2619,
6
+ "predict_samples": 8079,
7
+ "predict_samples_per_second": 222.796,
8
+ "predict_steps_per_second": 1.765
9
+ }
runs/May25_02-06-37_0d573eeffc83/events.out.tfevents.1716642100.0d573eeffc83.985437.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e376ad0687395de443ce4ae9c9ff080e1e3476b166342489128a2bff1a6c730b
3
+ size 465
train_results.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 50.0,
3
+ "total_flos": 2.8277448966144e+17,
4
+ "train_loss": 2.6664713644307803,
5
+ "train_runtime": 39272.3209,
6
+ "train_samples": 206112,
7
+ "train_samples_per_second": 262.414,
8
+ "train_steps_per_second": 4.101
9
+ }
trainer_state.json ADDED
@@ -0,0 +1,2296 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 50.0,
5
+ "eval_steps": 500,
6
+ "global_step": 161050,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.1552312946289972,
13
+ "grad_norm": 2.5311179161071777,
14
+ "learning_rate": 4.9844768705371e-05,
15
+ "loss": 7.3069,
16
+ "step": 500
17
+ },
18
+ {
19
+ "epoch": 0.3104625892579944,
20
+ "grad_norm": 1.969574213027954,
21
+ "learning_rate": 4.968953741074201e-05,
22
+ "loss": 6.5393,
23
+ "step": 1000
24
+ },
25
+ {
26
+ "epoch": 0.4656938838869916,
27
+ "grad_norm": 2.073662757873535,
28
+ "learning_rate": 4.953430611611301e-05,
29
+ "loss": 6.1794,
30
+ "step": 1500
31
+ },
32
+ {
33
+ "epoch": 0.6209251785159888,
34
+ "grad_norm": 1.9679251909255981,
35
+ "learning_rate": 4.937907482148401e-05,
36
+ "loss": 5.8754,
37
+ "step": 2000
38
+ },
39
+ {
40
+ "epoch": 0.776156473144986,
41
+ "grad_norm": 2.1763973236083984,
42
+ "learning_rate": 4.922384352685502e-05,
43
+ "loss": 5.6265,
44
+ "step": 2500
45
+ },
46
+ {
47
+ "epoch": 0.9313877677739832,
48
+ "grad_norm": 2.12394642829895,
49
+ "learning_rate": 4.906861223222602e-05,
50
+ "loss": 5.4302,
51
+ "step": 3000
52
+ },
53
+ {
54
+ "epoch": 1.0866190624029803,
55
+ "grad_norm": 2.004790782928467,
56
+ "learning_rate": 4.891338093759702e-05,
57
+ "loss": 5.2491,
58
+ "step": 3500
59
+ },
60
+ {
61
+ "epoch": 1.2418503570319777,
62
+ "grad_norm": 1.9632400274276733,
63
+ "learning_rate": 4.875814964296802e-05,
64
+ "loss": 5.1163,
65
+ "step": 4000
66
+ },
67
+ {
68
+ "epoch": 1.397081651660975,
69
+ "grad_norm": 2.08528733253479,
70
+ "learning_rate": 4.8602918348339024e-05,
71
+ "loss": 5.0118,
72
+ "step": 4500
73
+ },
74
+ {
75
+ "epoch": 1.552312946289972,
76
+ "grad_norm": 2.016941547393799,
77
+ "learning_rate": 4.8447687053710025e-05,
78
+ "loss": 4.9387,
79
+ "step": 5000
80
+ },
81
+ {
82
+ "epoch": 1.7075442409189692,
83
+ "grad_norm": 2.050320863723755,
84
+ "learning_rate": 4.829245575908103e-05,
85
+ "loss": 4.8555,
86
+ "step": 5500
87
+ },
88
+ {
89
+ "epoch": 1.8627755355479665,
90
+ "grad_norm": 2.028923273086548,
91
+ "learning_rate": 4.8137224464452034e-05,
92
+ "loss": 4.801,
93
+ "step": 6000
94
+ },
95
+ {
96
+ "epoch": 2.018006830176964,
97
+ "grad_norm": 2.3016574382781982,
98
+ "learning_rate": 4.7981993169823035e-05,
99
+ "loss": 4.7415,
100
+ "step": 6500
101
+ },
102
+ {
103
+ "epoch": 2.1732381248059607,
104
+ "grad_norm": 2.362550735473633,
105
+ "learning_rate": 4.782676187519404e-05,
106
+ "loss": 4.6673,
107
+ "step": 7000
108
+ },
109
+ {
110
+ "epoch": 2.328469419434958,
111
+ "grad_norm": 2.2543234825134277,
112
+ "learning_rate": 4.7671530580565043e-05,
113
+ "loss": 4.6305,
114
+ "step": 7500
115
+ },
116
+ {
117
+ "epoch": 2.4837007140639553,
118
+ "grad_norm": 2.4154460430145264,
119
+ "learning_rate": 4.7516299285936044e-05,
120
+ "loss": 4.5969,
121
+ "step": 8000
122
+ },
123
+ {
124
+ "epoch": 2.6389320086929526,
125
+ "grad_norm": 2.489636182785034,
126
+ "learning_rate": 4.736106799130705e-05,
127
+ "loss": 4.5431,
128
+ "step": 8500
129
+ },
130
+ {
131
+ "epoch": 2.79416330332195,
132
+ "grad_norm": 2.3382210731506348,
133
+ "learning_rate": 4.720583669667805e-05,
134
+ "loss": 4.5128,
135
+ "step": 9000
136
+ },
137
+ {
138
+ "epoch": 2.949394597950947,
139
+ "grad_norm": 2.371870756149292,
140
+ "learning_rate": 4.7050605402049054e-05,
141
+ "loss": 4.481,
142
+ "step": 9500
143
+ },
144
+ {
145
+ "epoch": 3.104625892579944,
146
+ "grad_norm": 2.4025254249572754,
147
+ "learning_rate": 4.689537410742006e-05,
148
+ "loss": 4.4316,
149
+ "step": 10000
150
+ },
151
+ {
152
+ "epoch": 3.2598571872089415,
153
+ "grad_norm": 2.3801567554473877,
154
+ "learning_rate": 4.674014281279106e-05,
155
+ "loss": 4.392,
156
+ "step": 10500
157
+ },
158
+ {
159
+ "epoch": 3.4150884818379383,
160
+ "grad_norm": 2.7669832706451416,
161
+ "learning_rate": 4.6584911518162063e-05,
162
+ "loss": 4.3718,
163
+ "step": 11000
164
+ },
165
+ {
166
+ "epoch": 3.5703197764669357,
167
+ "grad_norm": 2.3554067611694336,
168
+ "learning_rate": 4.6429680223533064e-05,
169
+ "loss": 4.3374,
170
+ "step": 11500
171
+ },
172
+ {
173
+ "epoch": 3.725551071095933,
174
+ "grad_norm": 2.672555923461914,
175
+ "learning_rate": 4.6274448928904065e-05,
176
+ "loss": 4.2903,
177
+ "step": 12000
178
+ },
179
+ {
180
+ "epoch": 3.8807823657249303,
181
+ "grad_norm": 3.017108678817749,
182
+ "learning_rate": 4.611921763427507e-05,
183
+ "loss": 4.2753,
184
+ "step": 12500
185
+ },
186
+ {
187
+ "epoch": 4.036013660353928,
188
+ "grad_norm": 2.9149186611175537,
189
+ "learning_rate": 4.5963986339646074e-05,
190
+ "loss": 4.2174,
191
+ "step": 13000
192
+ },
193
+ {
194
+ "epoch": 4.191244954982925,
195
+ "grad_norm": 2.5483713150024414,
196
+ "learning_rate": 4.5808755045017075e-05,
197
+ "loss": 4.1771,
198
+ "step": 13500
199
+ },
200
+ {
201
+ "epoch": 4.346476249611921,
202
+ "grad_norm": 2.8469910621643066,
203
+ "learning_rate": 4.565352375038808e-05,
204
+ "loss": 4.1374,
205
+ "step": 14000
206
+ },
207
+ {
208
+ "epoch": 4.501707544240919,
209
+ "grad_norm": 2.5643463134765625,
210
+ "learning_rate": 4.5498292455759083e-05,
211
+ "loss": 4.1179,
212
+ "step": 14500
213
+ },
214
+ {
215
+ "epoch": 4.656938838869916,
216
+ "grad_norm": 2.949036121368408,
217
+ "learning_rate": 4.5343061161130084e-05,
218
+ "loss": 4.0832,
219
+ "step": 15000
220
+ },
221
+ {
222
+ "epoch": 4.812170133498913,
223
+ "grad_norm": 2.8270998001098633,
224
+ "learning_rate": 4.518782986650109e-05,
225
+ "loss": 4.066,
226
+ "step": 15500
227
+ },
228
+ {
229
+ "epoch": 4.967401428127911,
230
+ "grad_norm": 2.839745283126831,
231
+ "learning_rate": 4.503259857187209e-05,
232
+ "loss": 4.0297,
233
+ "step": 16000
234
+ },
235
+ {
236
+ "epoch": 5.122632722756908,
237
+ "grad_norm": 2.8617799282073975,
238
+ "learning_rate": 4.4877367277243094e-05,
239
+ "loss": 3.9769,
240
+ "step": 16500
241
+ },
242
+ {
243
+ "epoch": 5.277864017385905,
244
+ "grad_norm": 2.988461494445801,
245
+ "learning_rate": 4.47221359826141e-05,
246
+ "loss": 3.9605,
247
+ "step": 17000
248
+ },
249
+ {
250
+ "epoch": 5.433095312014903,
251
+ "grad_norm": 2.9263391494750977,
252
+ "learning_rate": 4.45669046879851e-05,
253
+ "loss": 3.9326,
254
+ "step": 17500
255
+ },
256
+ {
257
+ "epoch": 5.5883266066439,
258
+ "grad_norm": 3.013662099838257,
259
+ "learning_rate": 4.4411673393356104e-05,
260
+ "loss": 3.9117,
261
+ "step": 18000
262
+ },
263
+ {
264
+ "epoch": 5.743557901272896,
265
+ "grad_norm": 3.1931121349334717,
266
+ "learning_rate": 4.4256442098727104e-05,
267
+ "loss": 3.8977,
268
+ "step": 18500
269
+ },
270
+ {
271
+ "epoch": 5.898789195901894,
272
+ "grad_norm": 3.0486295223236084,
273
+ "learning_rate": 4.4101210804098105e-05,
274
+ "loss": 3.8594,
275
+ "step": 19000
276
+ },
277
+ {
278
+ "epoch": 6.054020490530891,
279
+ "grad_norm": 3.090029001235962,
280
+ "learning_rate": 4.3945979509469106e-05,
281
+ "loss": 3.8299,
282
+ "step": 19500
283
+ },
284
+ {
285
+ "epoch": 6.209251785159888,
286
+ "grad_norm": 2.7931933403015137,
287
+ "learning_rate": 4.3790748214840114e-05,
288
+ "loss": 3.7919,
289
+ "step": 20000
290
+ },
291
+ {
292
+ "epoch": 6.364483079788886,
293
+ "grad_norm": 3.0710182189941406,
294
+ "learning_rate": 4.3635516920211115e-05,
295
+ "loss": 3.7688,
296
+ "step": 20500
297
+ },
298
+ {
299
+ "epoch": 6.519714374417883,
300
+ "grad_norm": 3.5505247116088867,
301
+ "learning_rate": 4.3480285625582116e-05,
302
+ "loss": 3.7545,
303
+ "step": 21000
304
+ },
305
+ {
306
+ "epoch": 6.674945669046879,
307
+ "grad_norm": 3.376587390899658,
308
+ "learning_rate": 4.3325054330953124e-05,
309
+ "loss": 3.7285,
310
+ "step": 21500
311
+ },
312
+ {
313
+ "epoch": 6.830176963675877,
314
+ "grad_norm": 4.988298416137695,
315
+ "learning_rate": 4.3169823036324124e-05,
316
+ "loss": 3.7192,
317
+ "step": 22000
318
+ },
319
+ {
320
+ "epoch": 6.985408258304874,
321
+ "grad_norm": 3.1776342391967773,
322
+ "learning_rate": 4.3014591741695125e-05,
323
+ "loss": 3.6964,
324
+ "step": 22500
325
+ },
326
+ {
327
+ "epoch": 7.140639552933871,
328
+ "grad_norm": 3.3717660903930664,
329
+ "learning_rate": 4.285936044706613e-05,
330
+ "loss": 3.6509,
331
+ "step": 23000
332
+ },
333
+ {
334
+ "epoch": 7.295870847562869,
335
+ "grad_norm": 3.224043607711792,
336
+ "learning_rate": 4.2704129152437134e-05,
337
+ "loss": 3.6125,
338
+ "step": 23500
339
+ },
340
+ {
341
+ "epoch": 7.451102142191866,
342
+ "grad_norm": 3.1459500789642334,
343
+ "learning_rate": 4.2548897857808135e-05,
344
+ "loss": 3.6209,
345
+ "step": 24000
346
+ },
347
+ {
348
+ "epoch": 7.606333436820863,
349
+ "grad_norm": 3.296095132827759,
350
+ "learning_rate": 4.239366656317914e-05,
351
+ "loss": 3.5956,
352
+ "step": 24500
353
+ },
354
+ {
355
+ "epoch": 7.761564731449861,
356
+ "grad_norm": 3.7389063835144043,
357
+ "learning_rate": 4.2238435268550144e-05,
358
+ "loss": 3.5743,
359
+ "step": 25000
360
+ },
361
+ {
362
+ "epoch": 7.916796026078858,
363
+ "grad_norm": 3.22495174407959,
364
+ "learning_rate": 4.2083203973921145e-05,
365
+ "loss": 3.5633,
366
+ "step": 25500
367
+ },
368
+ {
369
+ "epoch": 8.072027320707855,
370
+ "grad_norm": 3.2648046016693115,
371
+ "learning_rate": 4.192797267929215e-05,
372
+ "loss": 3.5174,
373
+ "step": 26000
374
+ },
375
+ {
376
+ "epoch": 8.227258615336853,
377
+ "grad_norm": 3.516303539276123,
378
+ "learning_rate": 4.177274138466315e-05,
379
+ "loss": 3.4799,
380
+ "step": 26500
381
+ },
382
+ {
383
+ "epoch": 8.38248990996585,
384
+ "grad_norm": 3.50242280960083,
385
+ "learning_rate": 4.1617510090034154e-05,
386
+ "loss": 3.4678,
387
+ "step": 27000
388
+ },
389
+ {
390
+ "epoch": 8.537721204594845,
391
+ "grad_norm": 3.312835454940796,
392
+ "learning_rate": 4.1462278795405155e-05,
393
+ "loss": 3.4551,
394
+ "step": 27500
395
+ },
396
+ {
397
+ "epoch": 8.692952499223843,
398
+ "grad_norm": 3.0370001792907715,
399
+ "learning_rate": 4.1307047500776156e-05,
400
+ "loss": 3.4393,
401
+ "step": 28000
402
+ },
403
+ {
404
+ "epoch": 8.84818379385284,
405
+ "grad_norm": 3.387784242630005,
406
+ "learning_rate": 4.115181620614716e-05,
407
+ "loss": 3.415,
408
+ "step": 28500
409
+ },
410
+ {
411
+ "epoch": 9.003415088481837,
412
+ "grad_norm": 3.6201112270355225,
413
+ "learning_rate": 4.0996584911518165e-05,
414
+ "loss": 3.418,
415
+ "step": 29000
416
+ },
417
+ {
418
+ "epoch": 9.158646383110835,
419
+ "grad_norm": 3.2895004749298096,
420
+ "learning_rate": 4.0841353616889166e-05,
421
+ "loss": 3.3479,
422
+ "step": 29500
423
+ },
424
+ {
425
+ "epoch": 9.313877677739832,
426
+ "grad_norm": 4.205784320831299,
427
+ "learning_rate": 4.0686122322260166e-05,
428
+ "loss": 3.3433,
429
+ "step": 30000
430
+ },
431
+ {
432
+ "epoch": 9.46910897236883,
433
+ "grad_norm": 4.602181434631348,
434
+ "learning_rate": 4.0530891027631174e-05,
435
+ "loss": 3.3223,
436
+ "step": 30500
437
+ },
438
+ {
439
+ "epoch": 9.624340266997827,
440
+ "grad_norm": 3.7277743816375732,
441
+ "learning_rate": 4.0375659733002175e-05,
442
+ "loss": 3.3039,
443
+ "step": 31000
444
+ },
445
+ {
446
+ "epoch": 9.779571561626824,
447
+ "grad_norm": 3.491147994995117,
448
+ "learning_rate": 4.0220428438373176e-05,
449
+ "loss": 3.2895,
450
+ "step": 31500
451
+ },
452
+ {
453
+ "epoch": 9.934802856255821,
454
+ "grad_norm": 3.2562263011932373,
455
+ "learning_rate": 4.0065197143744184e-05,
456
+ "loss": 3.2754,
457
+ "step": 32000
458
+ },
459
+ {
460
+ "epoch": 10.090034150884819,
461
+ "grad_norm": 3.8227834701538086,
462
+ "learning_rate": 3.9909965849115185e-05,
463
+ "loss": 3.239,
464
+ "step": 32500
465
+ },
466
+ {
467
+ "epoch": 10.245265445513816,
468
+ "grad_norm": 3.5297625064849854,
469
+ "learning_rate": 3.9754734554486186e-05,
470
+ "loss": 3.2036,
471
+ "step": 33000
472
+ },
473
+ {
474
+ "epoch": 10.400496740142813,
475
+ "grad_norm": 3.621588945388794,
476
+ "learning_rate": 3.959950325985719e-05,
477
+ "loss": 3.1944,
478
+ "step": 33500
479
+ },
480
+ {
481
+ "epoch": 10.55572803477181,
482
+ "grad_norm": 3.2394556999206543,
483
+ "learning_rate": 3.9444271965228194e-05,
484
+ "loss": 3.1961,
485
+ "step": 34000
486
+ },
487
+ {
488
+ "epoch": 10.710959329400808,
489
+ "grad_norm": 3.6555140018463135,
490
+ "learning_rate": 3.9289040670599195e-05,
491
+ "loss": 3.1687,
492
+ "step": 34500
493
+ },
494
+ {
495
+ "epoch": 10.866190624029805,
496
+ "grad_norm": 3.675229787826538,
497
+ "learning_rate": 3.9133809375970196e-05,
498
+ "loss": 3.1664,
499
+ "step": 35000
500
+ },
501
+ {
502
+ "epoch": 11.021421918658802,
503
+ "grad_norm": 3.4863462448120117,
504
+ "learning_rate": 3.89785780813412e-05,
505
+ "loss": 3.1411,
506
+ "step": 35500
507
+ },
508
+ {
509
+ "epoch": 11.176653213287798,
510
+ "grad_norm": 3.564634084701538,
511
+ "learning_rate": 3.88233467867122e-05,
512
+ "loss": 3.0942,
513
+ "step": 36000
514
+ },
515
+ {
516
+ "epoch": 11.331884507916795,
517
+ "grad_norm": 3.6480257511138916,
518
+ "learning_rate": 3.8668115492083206e-05,
519
+ "loss": 3.0866,
520
+ "step": 36500
521
+ },
522
+ {
523
+ "epoch": 11.487115802545793,
524
+ "grad_norm": 3.6294379234313965,
525
+ "learning_rate": 3.8512884197454207e-05,
526
+ "loss": 3.075,
527
+ "step": 37000
528
+ },
529
+ {
530
+ "epoch": 11.64234709717479,
531
+ "grad_norm": 3.706338882446289,
532
+ "learning_rate": 3.8357652902825214e-05,
533
+ "loss": 3.0711,
534
+ "step": 37500
535
+ },
536
+ {
537
+ "epoch": 11.797578391803787,
538
+ "grad_norm": 3.5022192001342773,
539
+ "learning_rate": 3.8202421608196215e-05,
540
+ "loss": 3.0543,
541
+ "step": 38000
542
+ },
543
+ {
544
+ "epoch": 11.952809686432785,
545
+ "grad_norm": 3.718973159790039,
546
+ "learning_rate": 3.8047190313567216e-05,
547
+ "loss": 3.0499,
548
+ "step": 38500
549
+ },
550
+ {
551
+ "epoch": 12.108040981061782,
552
+ "grad_norm": 3.5916123390197754,
553
+ "learning_rate": 3.7891959018938224e-05,
554
+ "loss": 3.0056,
555
+ "step": 39000
556
+ },
557
+ {
558
+ "epoch": 12.26327227569078,
559
+ "grad_norm": 3.694257974624634,
560
+ "learning_rate": 3.7736727724309225e-05,
561
+ "loss": 2.9916,
562
+ "step": 39500
563
+ },
564
+ {
565
+ "epoch": 12.418503570319777,
566
+ "grad_norm": 3.5255913734436035,
567
+ "learning_rate": 3.7581496429680226e-05,
568
+ "loss": 2.9823,
569
+ "step": 40000
570
+ },
571
+ {
572
+ "epoch": 12.573734864948774,
573
+ "grad_norm": 3.4169323444366455,
574
+ "learning_rate": 3.742626513505123e-05,
575
+ "loss": 2.9726,
576
+ "step": 40500
577
+ },
578
+ {
579
+ "epoch": 12.728966159577771,
580
+ "grad_norm": 3.2919702529907227,
581
+ "learning_rate": 3.7271033840422234e-05,
582
+ "loss": 2.9586,
583
+ "step": 41000
584
+ },
585
+ {
586
+ "epoch": 12.884197454206769,
587
+ "grad_norm": 3.52095890045166,
588
+ "learning_rate": 3.7115802545793235e-05,
589
+ "loss": 2.9576,
590
+ "step": 41500
591
+ },
592
+ {
593
+ "epoch": 13.039428748835766,
594
+ "grad_norm": 4.317255020141602,
595
+ "learning_rate": 3.6960571251164236e-05,
596
+ "loss": 2.936,
597
+ "step": 42000
598
+ },
599
+ {
600
+ "epoch": 13.194660043464763,
601
+ "grad_norm": 3.2749974727630615,
602
+ "learning_rate": 3.680533995653524e-05,
603
+ "loss": 2.894,
604
+ "step": 42500
605
+ },
606
+ {
607
+ "epoch": 13.34989133809376,
608
+ "grad_norm": 4.149952411651611,
609
+ "learning_rate": 3.665010866190624e-05,
610
+ "loss": 2.8892,
611
+ "step": 43000
612
+ },
613
+ {
614
+ "epoch": 13.505122632722756,
615
+ "grad_norm": 3.8253960609436035,
616
+ "learning_rate": 3.6494877367277246e-05,
617
+ "loss": 2.8957,
618
+ "step": 43500
619
+ },
620
+ {
621
+ "epoch": 13.660353927351753,
622
+ "grad_norm": 3.3494088649749756,
623
+ "learning_rate": 3.6339646072648247e-05,
624
+ "loss": 2.8821,
625
+ "step": 44000
626
+ },
627
+ {
628
+ "epoch": 13.81558522198075,
629
+ "grad_norm": 3.426492691040039,
630
+ "learning_rate": 3.618441477801925e-05,
631
+ "loss": 2.8655,
632
+ "step": 44500
633
+ },
634
+ {
635
+ "epoch": 13.970816516609748,
636
+ "grad_norm": 3.3416435718536377,
637
+ "learning_rate": 3.6029183483390255e-05,
638
+ "loss": 2.8685,
639
+ "step": 45000
640
+ },
641
+ {
642
+ "epoch": 14.126047811238745,
643
+ "grad_norm": 3.6365139484405518,
644
+ "learning_rate": 3.5873952188761256e-05,
645
+ "loss": 2.8184,
646
+ "step": 45500
647
+ },
648
+ {
649
+ "epoch": 14.281279105867743,
650
+ "grad_norm": 3.4732120037078857,
651
+ "learning_rate": 3.571872089413226e-05,
652
+ "loss": 2.8102,
653
+ "step": 46000
654
+ },
655
+ {
656
+ "epoch": 14.43651040049674,
657
+ "grad_norm": 3.4674530029296875,
658
+ "learning_rate": 3.5563489599503265e-05,
659
+ "loss": 2.8077,
660
+ "step": 46500
661
+ },
662
+ {
663
+ "epoch": 14.591741695125737,
664
+ "grad_norm": 3.3378286361694336,
665
+ "learning_rate": 3.5408258304874266e-05,
666
+ "loss": 2.808,
667
+ "step": 47000
668
+ },
669
+ {
670
+ "epoch": 14.746972989754735,
671
+ "grad_norm": 3.5950191020965576,
672
+ "learning_rate": 3.525302701024527e-05,
673
+ "loss": 2.7938,
674
+ "step": 47500
675
+ },
676
+ {
677
+ "epoch": 14.902204284383732,
678
+ "grad_norm": 3.4225220680236816,
679
+ "learning_rate": 3.5097795715616274e-05,
680
+ "loss": 2.7817,
681
+ "step": 48000
682
+ },
683
+ {
684
+ "epoch": 15.05743557901273,
685
+ "grad_norm": 3.353314161300659,
686
+ "learning_rate": 3.4942564420987275e-05,
687
+ "loss": 2.7783,
688
+ "step": 48500
689
+ },
690
+ {
691
+ "epoch": 15.212666873641727,
692
+ "grad_norm": 3.483828067779541,
693
+ "learning_rate": 3.4787333126358276e-05,
694
+ "loss": 2.7388,
695
+ "step": 49000
696
+ },
697
+ {
698
+ "epoch": 15.367898168270724,
699
+ "grad_norm": 3.295490026473999,
700
+ "learning_rate": 3.463210183172928e-05,
701
+ "loss": 2.7404,
702
+ "step": 49500
703
+ },
704
+ {
705
+ "epoch": 15.523129462899721,
706
+ "grad_norm": 3.7673990726470947,
707
+ "learning_rate": 3.447687053710028e-05,
708
+ "loss": 2.7192,
709
+ "step": 50000
710
+ },
711
+ {
712
+ "epoch": 15.678360757528718,
713
+ "grad_norm": 3.3685121536254883,
714
+ "learning_rate": 3.432163924247128e-05,
715
+ "loss": 2.7341,
716
+ "step": 50500
717
+ },
718
+ {
719
+ "epoch": 15.833592052157716,
720
+ "grad_norm": 3.8389620780944824,
721
+ "learning_rate": 3.416640794784229e-05,
722
+ "loss": 2.7239,
723
+ "step": 51000
724
+ },
725
+ {
726
+ "epoch": 15.988823346786713,
727
+ "grad_norm": 3.8301634788513184,
728
+ "learning_rate": 3.401117665321329e-05,
729
+ "loss": 2.716,
730
+ "step": 51500
731
+ },
732
+ {
733
+ "epoch": 16.14405464141571,
734
+ "grad_norm": 3.7196969985961914,
735
+ "learning_rate": 3.385594535858429e-05,
736
+ "loss": 2.6732,
737
+ "step": 52000
738
+ },
739
+ {
740
+ "epoch": 16.299285936044708,
741
+ "grad_norm": 3.847682476043701,
742
+ "learning_rate": 3.3700714063955296e-05,
743
+ "loss": 2.6561,
744
+ "step": 52500
745
+ },
746
+ {
747
+ "epoch": 16.454517230673705,
748
+ "grad_norm": 3.079871416091919,
749
+ "learning_rate": 3.35454827693263e-05,
750
+ "loss": 2.6765,
751
+ "step": 53000
752
+ },
753
+ {
754
+ "epoch": 16.609748525302702,
755
+ "grad_norm": 3.7256991863250732,
756
+ "learning_rate": 3.33902514746973e-05,
757
+ "loss": 2.6754,
758
+ "step": 53500
759
+ },
760
+ {
761
+ "epoch": 16.7649798199317,
762
+ "grad_norm": 3.556029796600342,
763
+ "learning_rate": 3.3235020180068306e-05,
764
+ "loss": 2.6675,
765
+ "step": 54000
766
+ },
767
+ {
768
+ "epoch": 16.920211114560697,
769
+ "grad_norm": 3.9556143283843994,
770
+ "learning_rate": 3.307978888543931e-05,
771
+ "loss": 2.6507,
772
+ "step": 54500
773
+ },
774
+ {
775
+ "epoch": 17.075442409189694,
776
+ "grad_norm": 3.482832431793213,
777
+ "learning_rate": 3.292455759081031e-05,
778
+ "loss": 2.6343,
779
+ "step": 55000
780
+ },
781
+ {
782
+ "epoch": 17.230673703818688,
783
+ "grad_norm": 4.063329696655273,
784
+ "learning_rate": 3.2769326296181315e-05,
785
+ "loss": 2.6071,
786
+ "step": 55500
787
+ },
788
+ {
789
+ "epoch": 17.385904998447685,
790
+ "grad_norm": 3.9433131217956543,
791
+ "learning_rate": 3.2614095001552316e-05,
792
+ "loss": 2.6064,
793
+ "step": 56000
794
+ },
795
+ {
796
+ "epoch": 17.541136293076683,
797
+ "grad_norm": 3.5680246353149414,
798
+ "learning_rate": 3.245886370692332e-05,
799
+ "loss": 2.6091,
800
+ "step": 56500
801
+ },
802
+ {
803
+ "epoch": 17.69636758770568,
804
+ "grad_norm": 3.5213463306427,
805
+ "learning_rate": 3.230363241229432e-05,
806
+ "loss": 2.6058,
807
+ "step": 57000
808
+ },
809
+ {
810
+ "epoch": 17.851598882334677,
811
+ "grad_norm": 3.5061028003692627,
812
+ "learning_rate": 3.214840111766532e-05,
813
+ "loss": 2.5981,
814
+ "step": 57500
815
+ },
816
+ {
817
+ "epoch": 18.006830176963675,
818
+ "grad_norm": 3.4912548065185547,
819
+ "learning_rate": 3.199316982303632e-05,
820
+ "loss": 2.5998,
821
+ "step": 58000
822
+ },
823
+ {
824
+ "epoch": 18.162061471592672,
825
+ "grad_norm": 3.2917845249176025,
826
+ "learning_rate": 3.183793852840733e-05,
827
+ "loss": 2.5569,
828
+ "step": 58500
829
+ },
830
+ {
831
+ "epoch": 18.31729276622167,
832
+ "grad_norm": 3.707597017288208,
833
+ "learning_rate": 3.168270723377833e-05,
834
+ "loss": 2.555,
835
+ "step": 59000
836
+ },
837
+ {
838
+ "epoch": 18.472524060850667,
839
+ "grad_norm": 3.4082422256469727,
840
+ "learning_rate": 3.152747593914933e-05,
841
+ "loss": 2.5539,
842
+ "step": 59500
843
+ },
844
+ {
845
+ "epoch": 18.627755355479664,
846
+ "grad_norm": 3.6042168140411377,
847
+ "learning_rate": 3.137224464452034e-05,
848
+ "loss": 2.5485,
849
+ "step": 60000
850
+ },
851
+ {
852
+ "epoch": 18.78298665010866,
853
+ "grad_norm": 3.665288209915161,
854
+ "learning_rate": 3.121701334989134e-05,
855
+ "loss": 2.5499,
856
+ "step": 60500
857
+ },
858
+ {
859
+ "epoch": 18.93821794473766,
860
+ "grad_norm": 3.590956211090088,
861
+ "learning_rate": 3.1061782055262346e-05,
862
+ "loss": 2.5479,
863
+ "step": 61000
864
+ },
865
+ {
866
+ "epoch": 19.093449239366656,
867
+ "grad_norm": 3.468674898147583,
868
+ "learning_rate": 3.090655076063335e-05,
869
+ "loss": 2.5156,
870
+ "step": 61500
871
+ },
872
+ {
873
+ "epoch": 19.248680533995653,
874
+ "grad_norm": 3.4389026165008545,
875
+ "learning_rate": 3.075131946600435e-05,
876
+ "loss": 2.5037,
877
+ "step": 62000
878
+ },
879
+ {
880
+ "epoch": 19.40391182862465,
881
+ "grad_norm": 3.568042039871216,
882
+ "learning_rate": 3.0596088171375355e-05,
883
+ "loss": 2.514,
884
+ "step": 62500
885
+ },
886
+ {
887
+ "epoch": 19.559143123253648,
888
+ "grad_norm": 3.448721408843994,
889
+ "learning_rate": 3.0440856876746353e-05,
890
+ "loss": 2.5004,
891
+ "step": 63000
892
+ },
893
+ {
894
+ "epoch": 19.714374417882645,
895
+ "grad_norm": 3.3219761848449707,
896
+ "learning_rate": 3.0285625582117354e-05,
897
+ "loss": 2.5072,
898
+ "step": 63500
899
+ },
900
+ {
901
+ "epoch": 19.869605712511643,
902
+ "grad_norm": 3.5406880378723145,
903
+ "learning_rate": 3.013039428748836e-05,
904
+ "loss": 2.5006,
905
+ "step": 64000
906
+ },
907
+ {
908
+ "epoch": 20.02483700714064,
909
+ "grad_norm": 3.606174945831299,
910
+ "learning_rate": 2.9975162992859363e-05,
911
+ "loss": 2.5015,
912
+ "step": 64500
913
+ },
914
+ {
915
+ "epoch": 20.180068301769637,
916
+ "grad_norm": 3.5096230506896973,
917
+ "learning_rate": 2.9819931698230363e-05,
918
+ "loss": 2.456,
919
+ "step": 65000
920
+ },
921
+ {
922
+ "epoch": 20.335299596398634,
923
+ "grad_norm": 3.478529691696167,
924
+ "learning_rate": 2.966470040360137e-05,
925
+ "loss": 2.4515,
926
+ "step": 65500
927
+ },
928
+ {
929
+ "epoch": 20.490530891027632,
930
+ "grad_norm": 3.3920540809631348,
931
+ "learning_rate": 2.9509469108972372e-05,
932
+ "loss": 2.4625,
933
+ "step": 66000
934
+ },
935
+ {
936
+ "epoch": 20.64576218565663,
937
+ "grad_norm": 3.6235618591308594,
938
+ "learning_rate": 2.9354237814343373e-05,
939
+ "loss": 2.4665,
940
+ "step": 66500
941
+ },
942
+ {
943
+ "epoch": 20.800993480285626,
944
+ "grad_norm": 3.580286979675293,
945
+ "learning_rate": 2.9199006519714377e-05,
946
+ "loss": 2.4577,
947
+ "step": 67000
948
+ },
949
+ {
950
+ "epoch": 20.956224774914624,
951
+ "grad_norm": 3.5689404010772705,
952
+ "learning_rate": 2.9043775225085378e-05,
953
+ "loss": 2.4619,
954
+ "step": 67500
955
+ },
956
+ {
957
+ "epoch": 21.11145606954362,
958
+ "grad_norm": 3.5909037590026855,
959
+ "learning_rate": 2.888854393045638e-05,
960
+ "loss": 2.4148,
961
+ "step": 68000
962
+ },
963
+ {
964
+ "epoch": 21.26668736417262,
965
+ "grad_norm": 3.4171316623687744,
966
+ "learning_rate": 2.8733312635827387e-05,
967
+ "loss": 2.4071,
968
+ "step": 68500
969
+ },
970
+ {
971
+ "epoch": 21.421918658801616,
972
+ "grad_norm": 3.457381248474121,
973
+ "learning_rate": 2.8578081341198388e-05,
974
+ "loss": 2.4169,
975
+ "step": 69000
976
+ },
977
+ {
978
+ "epoch": 21.577149953430613,
979
+ "grad_norm": 3.420381784439087,
980
+ "learning_rate": 2.842285004656939e-05,
981
+ "loss": 2.42,
982
+ "step": 69500
983
+ },
984
+ {
985
+ "epoch": 21.73238124805961,
986
+ "grad_norm": 3.2285988330841064,
987
+ "learning_rate": 2.8267618751940393e-05,
988
+ "loss": 2.4262,
989
+ "step": 70000
990
+ },
991
+ {
992
+ "epoch": 21.887612542688608,
993
+ "grad_norm": 3.473289728164673,
994
+ "learning_rate": 2.8112387457311394e-05,
995
+ "loss": 2.413,
996
+ "step": 70500
997
+ },
998
+ {
999
+ "epoch": 22.042843837317605,
1000
+ "grad_norm": 3.272094488143921,
1001
+ "learning_rate": 2.7957156162682395e-05,
1002
+ "loss": 2.4166,
1003
+ "step": 71000
1004
+ },
1005
+ {
1006
+ "epoch": 22.1980751319466,
1007
+ "grad_norm": 3.903505563735962,
1008
+ "learning_rate": 2.7801924868053403e-05,
1009
+ "loss": 2.3766,
1010
+ "step": 71500
1011
+ },
1012
+ {
1013
+ "epoch": 22.353306426575596,
1014
+ "grad_norm": 3.58394193649292,
1015
+ "learning_rate": 2.7646693573424404e-05,
1016
+ "loss": 2.3844,
1017
+ "step": 72000
1018
+ },
1019
+ {
1020
+ "epoch": 22.508537721204593,
1021
+ "grad_norm": 3.6447036266326904,
1022
+ "learning_rate": 2.7491462278795404e-05,
1023
+ "loss": 2.3683,
1024
+ "step": 72500
1025
+ },
1026
+ {
1027
+ "epoch": 22.66376901583359,
1028
+ "grad_norm": 3.740295648574829,
1029
+ "learning_rate": 2.7336230984166412e-05,
1030
+ "loss": 2.3822,
1031
+ "step": 73000
1032
+ },
1033
+ {
1034
+ "epoch": 22.819000310462588,
1035
+ "grad_norm": 3.439695358276367,
1036
+ "learning_rate": 2.7180999689537413e-05,
1037
+ "loss": 2.3893,
1038
+ "step": 73500
1039
+ },
1040
+ {
1041
+ "epoch": 22.974231605091585,
1042
+ "grad_norm": 3.6966145038604736,
1043
+ "learning_rate": 2.7025768394908414e-05,
1044
+ "loss": 2.3803,
1045
+ "step": 74000
1046
+ },
1047
+ {
1048
+ "epoch": 23.129462899720583,
1049
+ "grad_norm": 3.1649584770202637,
1050
+ "learning_rate": 2.687053710027942e-05,
1051
+ "loss": 2.343,
1052
+ "step": 74500
1053
+ },
1054
+ {
1055
+ "epoch": 23.28469419434958,
1056
+ "grad_norm": 3.4965484142303467,
1057
+ "learning_rate": 2.671530580565042e-05,
1058
+ "loss": 2.3351,
1059
+ "step": 75000
1060
+ },
1061
+ {
1062
+ "epoch": 23.439925488978577,
1063
+ "grad_norm": 3.3004109859466553,
1064
+ "learning_rate": 2.656007451102142e-05,
1065
+ "loss": 2.335,
1066
+ "step": 75500
1067
+ },
1068
+ {
1069
+ "epoch": 23.595156783607575,
1070
+ "grad_norm": 3.5990171432495117,
1071
+ "learning_rate": 2.6404843216392428e-05,
1072
+ "loss": 2.3474,
1073
+ "step": 76000
1074
+ },
1075
+ {
1076
+ "epoch": 23.750388078236572,
1077
+ "grad_norm": 3.2564170360565186,
1078
+ "learning_rate": 2.624961192176343e-05,
1079
+ "loss": 2.3447,
1080
+ "step": 76500
1081
+ },
1082
+ {
1083
+ "epoch": 23.90561937286557,
1084
+ "grad_norm": 3.103309392929077,
1085
+ "learning_rate": 2.609438062713443e-05,
1086
+ "loss": 2.3536,
1087
+ "step": 77000
1088
+ },
1089
+ {
1090
+ "epoch": 24.060850667494567,
1091
+ "grad_norm": 3.434351921081543,
1092
+ "learning_rate": 2.5939149332505437e-05,
1093
+ "loss": 2.3314,
1094
+ "step": 77500
1095
+ },
1096
+ {
1097
+ "epoch": 24.216081962123564,
1098
+ "grad_norm": 3.0751280784606934,
1099
+ "learning_rate": 2.5783918037876435e-05,
1100
+ "loss": 2.3113,
1101
+ "step": 78000
1102
+ },
1103
+ {
1104
+ "epoch": 24.37131325675256,
1105
+ "grad_norm": 3.3030903339385986,
1106
+ "learning_rate": 2.5628686743247436e-05,
1107
+ "loss": 2.3074,
1108
+ "step": 78500
1109
+ },
1110
+ {
1111
+ "epoch": 24.52654455138156,
1112
+ "grad_norm": 3.4424233436584473,
1113
+ "learning_rate": 2.5473455448618444e-05,
1114
+ "loss": 2.3074,
1115
+ "step": 79000
1116
+ },
1117
+ {
1118
+ "epoch": 24.681775846010556,
1119
+ "grad_norm": 3.4242780208587646,
1120
+ "learning_rate": 2.5318224153989445e-05,
1121
+ "loss": 2.3127,
1122
+ "step": 79500
1123
+ },
1124
+ {
1125
+ "epoch": 24.837007140639553,
1126
+ "grad_norm": 3.478363275527954,
1127
+ "learning_rate": 2.5162992859360445e-05,
1128
+ "loss": 2.3068,
1129
+ "step": 80000
1130
+ },
1131
+ {
1132
+ "epoch": 24.99223843526855,
1133
+ "grad_norm": 3.381448745727539,
1134
+ "learning_rate": 2.5007761564731453e-05,
1135
+ "loss": 2.3128,
1136
+ "step": 80500
1137
+ },
1138
+ {
1139
+ "epoch": 25.147469729897548,
1140
+ "grad_norm": 3.3895316123962402,
1141
+ "learning_rate": 2.4852530270102454e-05,
1142
+ "loss": 2.2783,
1143
+ "step": 81000
1144
+ },
1145
+ {
1146
+ "epoch": 25.302701024526545,
1147
+ "grad_norm": 3.522529363632202,
1148
+ "learning_rate": 2.469729897547346e-05,
1149
+ "loss": 2.2744,
1150
+ "step": 81500
1151
+ },
1152
+ {
1153
+ "epoch": 25.457932319155542,
1154
+ "grad_norm": 3.2338192462921143,
1155
+ "learning_rate": 2.454206768084446e-05,
1156
+ "loss": 2.2724,
1157
+ "step": 82000
1158
+ },
1159
+ {
1160
+ "epoch": 25.61316361378454,
1161
+ "grad_norm": 3.4074482917785645,
1162
+ "learning_rate": 2.438683638621546e-05,
1163
+ "loss": 2.2792,
1164
+ "step": 82500
1165
+ },
1166
+ {
1167
+ "epoch": 25.768394908413537,
1168
+ "grad_norm": 4.039905548095703,
1169
+ "learning_rate": 2.4231605091586465e-05,
1170
+ "loss": 2.2841,
1171
+ "step": 83000
1172
+ },
1173
+ {
1174
+ "epoch": 25.923626203042534,
1175
+ "grad_norm": 3.547403335571289,
1176
+ "learning_rate": 2.407637379695747e-05,
1177
+ "loss": 2.2819,
1178
+ "step": 83500
1179
+ },
1180
+ {
1181
+ "epoch": 26.07885749767153,
1182
+ "grad_norm": 3.3868236541748047,
1183
+ "learning_rate": 2.392114250232847e-05,
1184
+ "loss": 2.2798,
1185
+ "step": 84000
1186
+ },
1187
+ {
1188
+ "epoch": 26.23408879230053,
1189
+ "grad_norm": 3.4890501499176025,
1190
+ "learning_rate": 2.3765911207699474e-05,
1191
+ "loss": 2.2533,
1192
+ "step": 84500
1193
+ },
1194
+ {
1195
+ "epoch": 26.389320086929526,
1196
+ "grad_norm": 3.4324049949645996,
1197
+ "learning_rate": 2.361067991307048e-05,
1198
+ "loss": 2.2503,
1199
+ "step": 85000
1200
+ },
1201
+ {
1202
+ "epoch": 26.544551381558524,
1203
+ "grad_norm": 3.444091558456421,
1204
+ "learning_rate": 2.345544861844148e-05,
1205
+ "loss": 2.2475,
1206
+ "step": 85500
1207
+ },
1208
+ {
1209
+ "epoch": 26.69978267618752,
1210
+ "grad_norm": 3.4090046882629395,
1211
+ "learning_rate": 2.330021732381248e-05,
1212
+ "loss": 2.2484,
1213
+ "step": 86000
1214
+ },
1215
+ {
1216
+ "epoch": 26.85501397081652,
1217
+ "grad_norm": 3.447425365447998,
1218
+ "learning_rate": 2.3144986029183485e-05,
1219
+ "loss": 2.2488,
1220
+ "step": 86500
1221
+ },
1222
+ {
1223
+ "epoch": 27.010245265445512,
1224
+ "grad_norm": 3.323572874069214,
1225
+ "learning_rate": 2.2989754734554486e-05,
1226
+ "loss": 2.2459,
1227
+ "step": 87000
1228
+ },
1229
+ {
1230
+ "epoch": 27.16547656007451,
1231
+ "grad_norm": 3.2703440189361572,
1232
+ "learning_rate": 2.283452343992549e-05,
1233
+ "loss": 2.2174,
1234
+ "step": 87500
1235
+ },
1236
+ {
1237
+ "epoch": 27.320707854703507,
1238
+ "grad_norm": 3.5882294178009033,
1239
+ "learning_rate": 2.2679292145296494e-05,
1240
+ "loss": 2.2148,
1241
+ "step": 88000
1242
+ },
1243
+ {
1244
+ "epoch": 27.475939149332504,
1245
+ "grad_norm": 3.4588534832000732,
1246
+ "learning_rate": 2.2524060850667495e-05,
1247
+ "loss": 2.2225,
1248
+ "step": 88500
1249
+ },
1250
+ {
1251
+ "epoch": 27.6311704439615,
1252
+ "grad_norm": 3.370229482650757,
1253
+ "learning_rate": 2.23688295560385e-05,
1254
+ "loss": 2.2274,
1255
+ "step": 89000
1256
+ },
1257
+ {
1258
+ "epoch": 27.7864017385905,
1259
+ "grad_norm": 3.271768093109131,
1260
+ "learning_rate": 2.22135982614095e-05,
1261
+ "loss": 2.2272,
1262
+ "step": 89500
1263
+ },
1264
+ {
1265
+ "epoch": 27.941633033219496,
1266
+ "grad_norm": 3.4835493564605713,
1267
+ "learning_rate": 2.20583669667805e-05,
1268
+ "loss": 2.2324,
1269
+ "step": 90000
1270
+ },
1271
+ {
1272
+ "epoch": 28.096864327848493,
1273
+ "grad_norm": 3.6498279571533203,
1274
+ "learning_rate": 2.1903135672151506e-05,
1275
+ "loss": 2.1981,
1276
+ "step": 90500
1277
+ },
1278
+ {
1279
+ "epoch": 28.25209562247749,
1280
+ "grad_norm": 3.356560468673706,
1281
+ "learning_rate": 2.174790437752251e-05,
1282
+ "loss": 2.2064,
1283
+ "step": 91000
1284
+ },
1285
+ {
1286
+ "epoch": 28.407326917106488,
1287
+ "grad_norm": 3.2081451416015625,
1288
+ "learning_rate": 2.1592673082893514e-05,
1289
+ "loss": 2.1893,
1290
+ "step": 91500
1291
+ },
1292
+ {
1293
+ "epoch": 28.562558211735485,
1294
+ "grad_norm": 3.4001569747924805,
1295
+ "learning_rate": 2.1437441788264515e-05,
1296
+ "loss": 2.2014,
1297
+ "step": 92000
1298
+ },
1299
+ {
1300
+ "epoch": 28.717789506364483,
1301
+ "grad_norm": 3.529869794845581,
1302
+ "learning_rate": 2.128221049363552e-05,
1303
+ "loss": 2.2016,
1304
+ "step": 92500
1305
+ },
1306
+ {
1307
+ "epoch": 28.87302080099348,
1308
+ "grad_norm": 3.313579797744751,
1309
+ "learning_rate": 2.112697919900652e-05,
1310
+ "loss": 2.1937,
1311
+ "step": 93000
1312
+ },
1313
+ {
1314
+ "epoch": 29.028252095622477,
1315
+ "grad_norm": 3.336879253387451,
1316
+ "learning_rate": 2.097174790437752e-05,
1317
+ "loss": 2.1945,
1318
+ "step": 93500
1319
+ },
1320
+ {
1321
+ "epoch": 29.183483390251475,
1322
+ "grad_norm": 3.567126512527466,
1323
+ "learning_rate": 2.0816516609748526e-05,
1324
+ "loss": 2.1729,
1325
+ "step": 94000
1326
+ },
1327
+ {
1328
+ "epoch": 29.338714684880472,
1329
+ "grad_norm": 3.5103540420532227,
1330
+ "learning_rate": 2.066128531511953e-05,
1331
+ "loss": 2.1772,
1332
+ "step": 94500
1333
+ },
1334
+ {
1335
+ "epoch": 29.49394597950947,
1336
+ "grad_norm": 3.479119300842285,
1337
+ "learning_rate": 2.050605402049053e-05,
1338
+ "loss": 2.1786,
1339
+ "step": 95000
1340
+ },
1341
+ {
1342
+ "epoch": 29.649177274138466,
1343
+ "grad_norm": 3.516080856323242,
1344
+ "learning_rate": 2.0350822725861535e-05,
1345
+ "loss": 2.1714,
1346
+ "step": 95500
1347
+ },
1348
+ {
1349
+ "epoch": 29.804408568767464,
1350
+ "grad_norm": 3.3175048828125,
1351
+ "learning_rate": 2.019559143123254e-05,
1352
+ "loss": 2.176,
1353
+ "step": 96000
1354
+ },
1355
+ {
1356
+ "epoch": 29.95963986339646,
1357
+ "grad_norm": 3.5741403102874756,
1358
+ "learning_rate": 2.004036013660354e-05,
1359
+ "loss": 2.1747,
1360
+ "step": 96500
1361
+ },
1362
+ {
1363
+ "epoch": 30.11487115802546,
1364
+ "grad_norm": 3.3326079845428467,
1365
+ "learning_rate": 1.988512884197454e-05,
1366
+ "loss": 2.1465,
1367
+ "step": 97000
1368
+ },
1369
+ {
1370
+ "epoch": 30.270102452654456,
1371
+ "grad_norm": 3.591451406478882,
1372
+ "learning_rate": 1.9729897547345546e-05,
1373
+ "loss": 2.1452,
1374
+ "step": 97500
1375
+ },
1376
+ {
1377
+ "epoch": 30.425333747283453,
1378
+ "grad_norm": 3.386875629425049,
1379
+ "learning_rate": 1.9574666252716547e-05,
1380
+ "loss": 2.1527,
1381
+ "step": 98000
1382
+ },
1383
+ {
1384
+ "epoch": 30.58056504191245,
1385
+ "grad_norm": 3.5038528442382812,
1386
+ "learning_rate": 1.941943495808755e-05,
1387
+ "loss": 2.1562,
1388
+ "step": 98500
1389
+ },
1390
+ {
1391
+ "epoch": 30.735796336541448,
1392
+ "grad_norm": 3.5655734539031982,
1393
+ "learning_rate": 1.9264203663458555e-05,
1394
+ "loss": 2.1579,
1395
+ "step": 99000
1396
+ },
1397
+ {
1398
+ "epoch": 30.891027631170445,
1399
+ "grad_norm": 3.556865692138672,
1400
+ "learning_rate": 1.9108972368829556e-05,
1401
+ "loss": 2.1654,
1402
+ "step": 99500
1403
+ },
1404
+ {
1405
+ "epoch": 31.046258925799442,
1406
+ "grad_norm": 3.171966791152954,
1407
+ "learning_rate": 1.895374107420056e-05,
1408
+ "loss": 2.1375,
1409
+ "step": 100000
1410
+ },
1411
+ {
1412
+ "epoch": 31.20149022042844,
1413
+ "grad_norm": 3.2655224800109863,
1414
+ "learning_rate": 1.8798509779571565e-05,
1415
+ "loss": 2.1242,
1416
+ "step": 100500
1417
+ },
1418
+ {
1419
+ "epoch": 31.356721515057437,
1420
+ "grad_norm": 3.5983667373657227,
1421
+ "learning_rate": 1.8643278484942562e-05,
1422
+ "loss": 2.1241,
1423
+ "step": 101000
1424
+ },
1425
+ {
1426
+ "epoch": 31.511952809686434,
1427
+ "grad_norm": 3.4082624912261963,
1428
+ "learning_rate": 1.8488047190313567e-05,
1429
+ "loss": 2.1286,
1430
+ "step": 101500
1431
+ },
1432
+ {
1433
+ "epoch": 31.66718410431543,
1434
+ "grad_norm": 3.6842904090881348,
1435
+ "learning_rate": 1.833281589568457e-05,
1436
+ "loss": 2.1307,
1437
+ "step": 102000
1438
+ },
1439
+ {
1440
+ "epoch": 31.82241539894443,
1441
+ "grad_norm": 3.506561040878296,
1442
+ "learning_rate": 1.8177584601055575e-05,
1443
+ "loss": 2.128,
1444
+ "step": 102500
1445
+ },
1446
+ {
1447
+ "epoch": 31.977646693573423,
1448
+ "grad_norm": 3.1754329204559326,
1449
+ "learning_rate": 1.8022353306426576e-05,
1450
+ "loss": 2.1546,
1451
+ "step": 103000
1452
+ },
1453
+ {
1454
+ "epoch": 32.132877988202424,
1455
+ "grad_norm": 3.4814705848693848,
1456
+ "learning_rate": 1.786712201179758e-05,
1457
+ "loss": 2.1155,
1458
+ "step": 103500
1459
+ },
1460
+ {
1461
+ "epoch": 32.28810928283142,
1462
+ "grad_norm": 3.7139673233032227,
1463
+ "learning_rate": 1.7711890717168585e-05,
1464
+ "loss": 2.1056,
1465
+ "step": 104000
1466
+ },
1467
+ {
1468
+ "epoch": 32.44334057746042,
1469
+ "grad_norm": 3.498924732208252,
1470
+ "learning_rate": 1.7556659422539586e-05,
1471
+ "loss": 2.1142,
1472
+ "step": 104500
1473
+ },
1474
+ {
1475
+ "epoch": 32.598571872089416,
1476
+ "grad_norm": 3.50618052482605,
1477
+ "learning_rate": 1.7401428127910587e-05,
1478
+ "loss": 2.1133,
1479
+ "step": 105000
1480
+ },
1481
+ {
1482
+ "epoch": 32.75380316671841,
1483
+ "grad_norm": 3.3597817420959473,
1484
+ "learning_rate": 1.724619683328159e-05,
1485
+ "loss": 2.1023,
1486
+ "step": 105500
1487
+ },
1488
+ {
1489
+ "epoch": 32.90903446134741,
1490
+ "grad_norm": 3.1257002353668213,
1491
+ "learning_rate": 1.7090965538652592e-05,
1492
+ "loss": 2.1164,
1493
+ "step": 106000
1494
+ },
1495
+ {
1496
+ "epoch": 33.06426575597641,
1497
+ "grad_norm": 3.2539827823638916,
1498
+ "learning_rate": 1.6935734244023596e-05,
1499
+ "loss": 2.094,
1500
+ "step": 106500
1501
+ },
1502
+ {
1503
+ "epoch": 33.219497050605405,
1504
+ "grad_norm": 3.5922114849090576,
1505
+ "learning_rate": 1.67805029493946e-05,
1506
+ "loss": 2.0797,
1507
+ "step": 107000
1508
+ },
1509
+ {
1510
+ "epoch": 33.3747283452344,
1511
+ "grad_norm": 3.4829044342041016,
1512
+ "learning_rate": 1.66252716547656e-05,
1513
+ "loss": 2.0991,
1514
+ "step": 107500
1515
+ },
1516
+ {
1517
+ "epoch": 33.5299596398634,
1518
+ "grad_norm": 3.402447462081909,
1519
+ "learning_rate": 1.6470040360136606e-05,
1520
+ "loss": 2.095,
1521
+ "step": 108000
1522
+ },
1523
+ {
1524
+ "epoch": 33.6851909344924,
1525
+ "grad_norm": 3.4598045349121094,
1526
+ "learning_rate": 1.6314809065507607e-05,
1527
+ "loss": 2.0866,
1528
+ "step": 108500
1529
+ },
1530
+ {
1531
+ "epoch": 33.840422229121394,
1532
+ "grad_norm": 3.806218147277832,
1533
+ "learning_rate": 1.6159577770878608e-05,
1534
+ "loss": 2.1014,
1535
+ "step": 109000
1536
+ },
1537
+ {
1538
+ "epoch": 33.99565352375039,
1539
+ "grad_norm": 3.7168009281158447,
1540
+ "learning_rate": 1.6004346476249612e-05,
1541
+ "loss": 2.103,
1542
+ "step": 109500
1543
+ },
1544
+ {
1545
+ "epoch": 34.15088481837939,
1546
+ "grad_norm": 3.703636884689331,
1547
+ "learning_rate": 1.5849115181620616e-05,
1548
+ "loss": 2.0691,
1549
+ "step": 110000
1550
+ },
1551
+ {
1552
+ "epoch": 34.306116113008386,
1553
+ "grad_norm": 3.4325764179229736,
1554
+ "learning_rate": 1.5693883886991617e-05,
1555
+ "loss": 2.077,
1556
+ "step": 110500
1557
+ },
1558
+ {
1559
+ "epoch": 34.461347407637376,
1560
+ "grad_norm": 3.5435197353363037,
1561
+ "learning_rate": 1.553865259236262e-05,
1562
+ "loss": 2.0839,
1563
+ "step": 111000
1564
+ },
1565
+ {
1566
+ "epoch": 34.616578702266374,
1567
+ "grad_norm": 3.507366418838501,
1568
+ "learning_rate": 1.5383421297733626e-05,
1569
+ "loss": 2.0709,
1570
+ "step": 111500
1571
+ },
1572
+ {
1573
+ "epoch": 34.77180999689537,
1574
+ "grad_norm": 3.401642084121704,
1575
+ "learning_rate": 1.5228190003104625e-05,
1576
+ "loss": 2.0656,
1577
+ "step": 112000
1578
+ },
1579
+ {
1580
+ "epoch": 34.92704129152437,
1581
+ "grad_norm": 3.4845223426818848,
1582
+ "learning_rate": 1.507295870847563e-05,
1583
+ "loss": 2.0848,
1584
+ "step": 112500
1585
+ },
1586
+ {
1587
+ "epoch": 35.082272586153366,
1588
+ "grad_norm": 3.477248191833496,
1589
+ "learning_rate": 1.4917727413846632e-05,
1590
+ "loss": 2.0595,
1591
+ "step": 113000
1592
+ },
1593
+ {
1594
+ "epoch": 35.23750388078236,
1595
+ "grad_norm": 3.3301658630371094,
1596
+ "learning_rate": 1.4762496119217633e-05,
1597
+ "loss": 2.056,
1598
+ "step": 113500
1599
+ },
1600
+ {
1601
+ "epoch": 35.39273517541136,
1602
+ "grad_norm": 3.5794308185577393,
1603
+ "learning_rate": 1.4607264824588637e-05,
1604
+ "loss": 2.0579,
1605
+ "step": 114000
1606
+ },
1607
+ {
1608
+ "epoch": 35.54796647004036,
1609
+ "grad_norm": 3.680616617202759,
1610
+ "learning_rate": 1.4452033529959642e-05,
1611
+ "loss": 2.0607,
1612
+ "step": 114500
1613
+ },
1614
+ {
1615
+ "epoch": 35.703197764669355,
1616
+ "grad_norm": 3.581709384918213,
1617
+ "learning_rate": 1.4296802235330644e-05,
1618
+ "loss": 2.0583,
1619
+ "step": 115000
1620
+ },
1621
+ {
1622
+ "epoch": 35.85842905929835,
1623
+ "grad_norm": 3.3622961044311523,
1624
+ "learning_rate": 1.4141570940701645e-05,
1625
+ "loss": 2.0646,
1626
+ "step": 115500
1627
+ },
1628
+ {
1629
+ "epoch": 36.01366035392735,
1630
+ "grad_norm": 3.7327592372894287,
1631
+ "learning_rate": 1.398633964607265e-05,
1632
+ "loss": 2.0665,
1633
+ "step": 116000
1634
+ },
1635
+ {
1636
+ "epoch": 36.16889164855635,
1637
+ "grad_norm": 3.1711039543151855,
1638
+ "learning_rate": 1.3831108351443654e-05,
1639
+ "loss": 2.0485,
1640
+ "step": 116500
1641
+ },
1642
+ {
1643
+ "epoch": 36.324122943185344,
1644
+ "grad_norm": 3.367382764816284,
1645
+ "learning_rate": 1.3675877056814653e-05,
1646
+ "loss": 2.0385,
1647
+ "step": 117000
1648
+ },
1649
+ {
1650
+ "epoch": 36.47935423781434,
1651
+ "grad_norm": 3.5484557151794434,
1652
+ "learning_rate": 1.3520645762185657e-05,
1653
+ "loss": 2.0345,
1654
+ "step": 117500
1655
+ },
1656
+ {
1657
+ "epoch": 36.63458553244334,
1658
+ "grad_norm": 3.6032321453094482,
1659
+ "learning_rate": 1.3365414467556662e-05,
1660
+ "loss": 2.0468,
1661
+ "step": 118000
1662
+ },
1663
+ {
1664
+ "epoch": 36.789816827072336,
1665
+ "grad_norm": 4.026392936706543,
1666
+ "learning_rate": 1.3210183172927663e-05,
1667
+ "loss": 2.0481,
1668
+ "step": 118500
1669
+ },
1670
+ {
1671
+ "epoch": 36.94504812170133,
1672
+ "grad_norm": 3.644474983215332,
1673
+ "learning_rate": 1.3054951878298665e-05,
1674
+ "loss": 2.0447,
1675
+ "step": 119000
1676
+ },
1677
+ {
1678
+ "epoch": 37.10027941633033,
1679
+ "grad_norm": 3.28544545173645,
1680
+ "learning_rate": 1.289972058366967e-05,
1681
+ "loss": 2.0358,
1682
+ "step": 119500
1683
+ },
1684
+ {
1685
+ "epoch": 37.25551071095933,
1686
+ "grad_norm": 3.2941246032714844,
1687
+ "learning_rate": 1.274448928904067e-05,
1688
+ "loss": 2.0084,
1689
+ "step": 120000
1690
+ },
1691
+ {
1692
+ "epoch": 37.410742005588325,
1693
+ "grad_norm": 3.7883927822113037,
1694
+ "learning_rate": 1.2589257994411675e-05,
1695
+ "loss": 2.0329,
1696
+ "step": 120500
1697
+ },
1698
+ {
1699
+ "epoch": 37.56597330021732,
1700
+ "grad_norm": 3.23514986038208,
1701
+ "learning_rate": 1.2434026699782676e-05,
1702
+ "loss": 2.0297,
1703
+ "step": 121000
1704
+ },
1705
+ {
1706
+ "epoch": 37.72120459484632,
1707
+ "grad_norm": 4.0385894775390625,
1708
+ "learning_rate": 1.227879540515368e-05,
1709
+ "loss": 2.0347,
1710
+ "step": 121500
1711
+ },
1712
+ {
1713
+ "epoch": 37.87643588947532,
1714
+ "grad_norm": 3.6746459007263184,
1715
+ "learning_rate": 1.2123564110524683e-05,
1716
+ "loss": 2.0386,
1717
+ "step": 122000
1718
+ },
1719
+ {
1720
+ "epoch": 38.031667184104315,
1721
+ "grad_norm": 3.5813486576080322,
1722
+ "learning_rate": 1.1968332815895685e-05,
1723
+ "loss": 2.0275,
1724
+ "step": 122500
1725
+ },
1726
+ {
1727
+ "epoch": 38.18689847873331,
1728
+ "grad_norm": 3.3303112983703613,
1729
+ "learning_rate": 1.1813101521266688e-05,
1730
+ "loss": 2.0086,
1731
+ "step": 123000
1732
+ },
1733
+ {
1734
+ "epoch": 38.34212977336231,
1735
+ "grad_norm": 3.5184788703918457,
1736
+ "learning_rate": 1.165787022663769e-05,
1737
+ "loss": 2.016,
1738
+ "step": 123500
1739
+ },
1740
+ {
1741
+ "epoch": 38.49736106799131,
1742
+ "grad_norm": 3.505143642425537,
1743
+ "learning_rate": 1.1502638932008695e-05,
1744
+ "loss": 2.0179,
1745
+ "step": 124000
1746
+ },
1747
+ {
1748
+ "epoch": 38.652592362620304,
1749
+ "grad_norm": 3.432009696960449,
1750
+ "learning_rate": 1.1347407637379696e-05,
1751
+ "loss": 2.02,
1752
+ "step": 124500
1753
+ },
1754
+ {
1755
+ "epoch": 38.8078236572493,
1756
+ "grad_norm": 3.268419027328491,
1757
+ "learning_rate": 1.1192176342750698e-05,
1758
+ "loss": 2.0185,
1759
+ "step": 125000
1760
+ },
1761
+ {
1762
+ "epoch": 38.9630549518783,
1763
+ "grad_norm": 3.895332098007202,
1764
+ "learning_rate": 1.1036945048121703e-05,
1765
+ "loss": 2.0213,
1766
+ "step": 125500
1767
+ },
1768
+ {
1769
+ "epoch": 39.118286246507296,
1770
+ "grad_norm": 3.481630325317383,
1771
+ "learning_rate": 1.0881713753492705e-05,
1772
+ "loss": 1.9974,
1773
+ "step": 126000
1774
+ },
1775
+ {
1776
+ "epoch": 39.27351754113629,
1777
+ "grad_norm": 3.3944613933563232,
1778
+ "learning_rate": 1.0726482458863706e-05,
1779
+ "loss": 2.0012,
1780
+ "step": 126500
1781
+ },
1782
+ {
1783
+ "epoch": 39.42874883576529,
1784
+ "grad_norm": 4.5043182373046875,
1785
+ "learning_rate": 1.057125116423471e-05,
1786
+ "loss": 2.0044,
1787
+ "step": 127000
1788
+ },
1789
+ {
1790
+ "epoch": 39.58398013039429,
1791
+ "grad_norm": 3.6557388305664062,
1792
+ "learning_rate": 1.0416019869605713e-05,
1793
+ "loss": 2.0059,
1794
+ "step": 127500
1795
+ },
1796
+ {
1797
+ "epoch": 39.739211425023285,
1798
+ "grad_norm": 3.3339288234710693,
1799
+ "learning_rate": 1.0260788574976716e-05,
1800
+ "loss": 2.0018,
1801
+ "step": 128000
1802
+ },
1803
+ {
1804
+ "epoch": 39.89444271965228,
1805
+ "grad_norm": 3.376677989959717,
1806
+ "learning_rate": 1.0105557280347718e-05,
1807
+ "loss": 2.0102,
1808
+ "step": 128500
1809
+ },
1810
+ {
1811
+ "epoch": 40.04967401428128,
1812
+ "grad_norm": 3.601083278656006,
1813
+ "learning_rate": 9.950325985718721e-06,
1814
+ "loss": 2.0019,
1815
+ "step": 129000
1816
+ },
1817
+ {
1818
+ "epoch": 40.20490530891028,
1819
+ "grad_norm": 3.458583116531372,
1820
+ "learning_rate": 9.795094691089725e-06,
1821
+ "loss": 1.9888,
1822
+ "step": 129500
1823
+ },
1824
+ {
1825
+ "epoch": 40.360136603539274,
1826
+ "grad_norm": 3.3313772678375244,
1827
+ "learning_rate": 9.639863396460728e-06,
1828
+ "loss": 1.9947,
1829
+ "step": 130000
1830
+ },
1831
+ {
1832
+ "epoch": 40.51536789816827,
1833
+ "grad_norm": 3.6328794956207275,
1834
+ "learning_rate": 9.484632101831729e-06,
1835
+ "loss": 1.985,
1836
+ "step": 130500
1837
+ },
1838
+ {
1839
+ "epoch": 40.67059919279727,
1840
+ "grad_norm": 3.165858030319214,
1841
+ "learning_rate": 9.329400807202733e-06,
1842
+ "loss": 1.9962,
1843
+ "step": 131000
1844
+ },
1845
+ {
1846
+ "epoch": 40.825830487426266,
1847
+ "grad_norm": 3.330787420272827,
1848
+ "learning_rate": 9.174169512573736e-06,
1849
+ "loss": 1.9982,
1850
+ "step": 131500
1851
+ },
1852
+ {
1853
+ "epoch": 40.981061782055264,
1854
+ "grad_norm": 3.3106303215026855,
1855
+ "learning_rate": 9.018938217944738e-06,
1856
+ "loss": 1.9974,
1857
+ "step": 132000
1858
+ },
1859
+ {
1860
+ "epoch": 41.13629307668426,
1861
+ "grad_norm": 3.49682354927063,
1862
+ "learning_rate": 8.863706923315741e-06,
1863
+ "loss": 1.9706,
1864
+ "step": 132500
1865
+ },
1866
+ {
1867
+ "epoch": 41.29152437131326,
1868
+ "grad_norm": 3.306004047393799,
1869
+ "learning_rate": 8.708475628686744e-06,
1870
+ "loss": 1.9795,
1871
+ "step": 133000
1872
+ },
1873
+ {
1874
+ "epoch": 41.446755665942256,
1875
+ "grad_norm": 3.715866804122925,
1876
+ "learning_rate": 8.553244334057746e-06,
1877
+ "loss": 1.9736,
1878
+ "step": 133500
1879
+ },
1880
+ {
1881
+ "epoch": 41.60198696057125,
1882
+ "grad_norm": 3.6723625659942627,
1883
+ "learning_rate": 8.398013039428749e-06,
1884
+ "loss": 1.9844,
1885
+ "step": 134000
1886
+ },
1887
+ {
1888
+ "epoch": 41.75721825520025,
1889
+ "grad_norm": 3.4786245822906494,
1890
+ "learning_rate": 8.242781744799752e-06,
1891
+ "loss": 1.9902,
1892
+ "step": 134500
1893
+ },
1894
+ {
1895
+ "epoch": 41.91244954982925,
1896
+ "grad_norm": 3.588338851928711,
1897
+ "learning_rate": 8.087550450170754e-06,
1898
+ "loss": 1.9829,
1899
+ "step": 135000
1900
+ },
1901
+ {
1902
+ "epoch": 42.067680844458245,
1903
+ "grad_norm": 3.7299747467041016,
1904
+ "learning_rate": 7.932319155541758e-06,
1905
+ "loss": 1.9803,
1906
+ "step": 135500
1907
+ },
1908
+ {
1909
+ "epoch": 42.22291213908724,
1910
+ "grad_norm": 3.1764190196990967,
1911
+ "learning_rate": 7.77708786091276e-06,
1912
+ "loss": 1.968,
1913
+ "step": 136000
1914
+ },
1915
+ {
1916
+ "epoch": 42.37814343371624,
1917
+ "grad_norm": 3.7980525493621826,
1918
+ "learning_rate": 7.621856566283764e-06,
1919
+ "loss": 1.9643,
1920
+ "step": 136500
1921
+ },
1922
+ {
1923
+ "epoch": 42.53337472834524,
1924
+ "grad_norm": 3.3517909049987793,
1925
+ "learning_rate": 7.466625271654766e-06,
1926
+ "loss": 1.9786,
1927
+ "step": 137000
1928
+ },
1929
+ {
1930
+ "epoch": 42.688606022974234,
1931
+ "grad_norm": 3.4372501373291016,
1932
+ "learning_rate": 7.311393977025768e-06,
1933
+ "loss": 1.9649,
1934
+ "step": 137500
1935
+ },
1936
+ {
1937
+ "epoch": 42.84383731760323,
1938
+ "grad_norm": 3.6716599464416504,
1939
+ "learning_rate": 7.156162682396772e-06,
1940
+ "loss": 1.9805,
1941
+ "step": 138000
1942
+ },
1943
+ {
1944
+ "epoch": 42.99906861223223,
1945
+ "grad_norm": 3.4524483680725098,
1946
+ "learning_rate": 7.000931387767774e-06,
1947
+ "loss": 1.9822,
1948
+ "step": 138500
1949
+ },
1950
+ {
1951
+ "epoch": 43.154299906861226,
1952
+ "grad_norm": 3.7548680305480957,
1953
+ "learning_rate": 6.845700093138777e-06,
1954
+ "loss": 1.9648,
1955
+ "step": 139000
1956
+ },
1957
+ {
1958
+ "epoch": 43.30953120149022,
1959
+ "grad_norm": 3.3365092277526855,
1960
+ "learning_rate": 6.69046879850978e-06,
1961
+ "loss": 1.9602,
1962
+ "step": 139500
1963
+ },
1964
+ {
1965
+ "epoch": 43.46476249611922,
1966
+ "grad_norm": 3.7342677116394043,
1967
+ "learning_rate": 6.535237503880783e-06,
1968
+ "loss": 1.9615,
1969
+ "step": 140000
1970
+ },
1971
+ {
1972
+ "epoch": 43.61999379074822,
1973
+ "grad_norm": 3.863659381866455,
1974
+ "learning_rate": 6.380006209251785e-06,
1975
+ "loss": 1.9672,
1976
+ "step": 140500
1977
+ },
1978
+ {
1979
+ "epoch": 43.775225085377215,
1980
+ "grad_norm": 3.3871288299560547,
1981
+ "learning_rate": 6.224774914622788e-06,
1982
+ "loss": 1.9663,
1983
+ "step": 141000
1984
+ },
1985
+ {
1986
+ "epoch": 43.93045638000621,
1987
+ "grad_norm": 3.8932695388793945,
1988
+ "learning_rate": 6.069543619993791e-06,
1989
+ "loss": 1.9654,
1990
+ "step": 141500
1991
+ },
1992
+ {
1993
+ "epoch": 44.08568767463521,
1994
+ "grad_norm": 3.8590352535247803,
1995
+ "learning_rate": 5.914312325364793e-06,
1996
+ "loss": 1.9528,
1997
+ "step": 142000
1998
+ },
1999
+ {
2000
+ "epoch": 44.2409189692642,
2001
+ "grad_norm": 3.2460689544677734,
2002
+ "learning_rate": 5.759081030735797e-06,
2003
+ "loss": 1.9557,
2004
+ "step": 142500
2005
+ },
2006
+ {
2007
+ "epoch": 44.3961502638932,
2008
+ "grad_norm": 3.6797304153442383,
2009
+ "learning_rate": 5.6038497361067995e-06,
2010
+ "loss": 1.9552,
2011
+ "step": 143000
2012
+ },
2013
+ {
2014
+ "epoch": 44.551381558522195,
2015
+ "grad_norm": 3.47025203704834,
2016
+ "learning_rate": 5.448618441477802e-06,
2017
+ "loss": 1.9548,
2018
+ "step": 143500
2019
+ },
2020
+ {
2021
+ "epoch": 44.70661285315119,
2022
+ "grad_norm": 3.376739978790283,
2023
+ "learning_rate": 5.293387146848805e-06,
2024
+ "loss": 1.9584,
2025
+ "step": 144000
2026
+ },
2027
+ {
2028
+ "epoch": 44.86184414778019,
2029
+ "grad_norm": 3.447455644607544,
2030
+ "learning_rate": 5.138155852219808e-06,
2031
+ "loss": 1.9524,
2032
+ "step": 144500
2033
+ },
2034
+ {
2035
+ "epoch": 45.01707544240919,
2036
+ "grad_norm": 3.5013909339904785,
2037
+ "learning_rate": 4.98292455759081e-06,
2038
+ "loss": 1.9537,
2039
+ "step": 145000
2040
+ },
2041
+ {
2042
+ "epoch": 45.172306737038184,
2043
+ "grad_norm": 3.857957124710083,
2044
+ "learning_rate": 4.827693262961813e-06,
2045
+ "loss": 1.9446,
2046
+ "step": 145500
2047
+ },
2048
+ {
2049
+ "epoch": 45.32753803166718,
2050
+ "grad_norm": 3.5706942081451416,
2051
+ "learning_rate": 4.672461968332816e-06,
2052
+ "loss": 1.9512,
2053
+ "step": 146000
2054
+ },
2055
+ {
2056
+ "epoch": 45.48276932629618,
2057
+ "grad_norm": 3.5100932121276855,
2058
+ "learning_rate": 4.5172306737038195e-06,
2059
+ "loss": 1.9573,
2060
+ "step": 146500
2061
+ },
2062
+ {
2063
+ "epoch": 45.638000620925176,
2064
+ "grad_norm": 3.7259585857391357,
2065
+ "learning_rate": 4.361999379074821e-06,
2066
+ "loss": 1.9461,
2067
+ "step": 147000
2068
+ },
2069
+ {
2070
+ "epoch": 45.79323191555417,
2071
+ "grad_norm": 3.6146180629730225,
2072
+ "learning_rate": 4.206768084445825e-06,
2073
+ "loss": 1.946,
2074
+ "step": 147500
2075
+ },
2076
+ {
2077
+ "epoch": 45.94846321018317,
2078
+ "grad_norm": 3.5233962535858154,
2079
+ "learning_rate": 4.051536789816827e-06,
2080
+ "loss": 1.9454,
2081
+ "step": 148000
2082
+ },
2083
+ {
2084
+ "epoch": 46.10369450481217,
2085
+ "grad_norm": 3.628469944000244,
2086
+ "learning_rate": 3.89630549518783e-06,
2087
+ "loss": 1.9394,
2088
+ "step": 148500
2089
+ },
2090
+ {
2091
+ "epoch": 46.258925799441165,
2092
+ "grad_norm": 3.5431318283081055,
2093
+ "learning_rate": 3.7410742005588326e-06,
2094
+ "loss": 1.9319,
2095
+ "step": 149000
2096
+ },
2097
+ {
2098
+ "epoch": 46.41415709407016,
2099
+ "grad_norm": 3.5288734436035156,
2100
+ "learning_rate": 3.5858429059298356e-06,
2101
+ "loss": 1.9436,
2102
+ "step": 149500
2103
+ },
2104
+ {
2105
+ "epoch": 46.56938838869916,
2106
+ "grad_norm": 3.5967633724212646,
2107
+ "learning_rate": 3.4306116113008387e-06,
2108
+ "loss": 1.9393,
2109
+ "step": 150000
2110
+ },
2111
+ {
2112
+ "epoch": 46.72461968332816,
2113
+ "grad_norm": 3.546693801879883,
2114
+ "learning_rate": 3.275380316671841e-06,
2115
+ "loss": 1.9502,
2116
+ "step": 150500
2117
+ },
2118
+ {
2119
+ "epoch": 46.879850977957155,
2120
+ "grad_norm": 3.5913896560668945,
2121
+ "learning_rate": 3.120149022042844e-06,
2122
+ "loss": 1.9514,
2123
+ "step": 151000
2124
+ },
2125
+ {
2126
+ "epoch": 47.03508227258615,
2127
+ "grad_norm": 3.8581769466400146,
2128
+ "learning_rate": 2.964917727413847e-06,
2129
+ "loss": 1.9383,
2130
+ "step": 151500
2131
+ },
2132
+ {
2133
+ "epoch": 47.19031356721515,
2134
+ "grad_norm": 3.577950954437256,
2135
+ "learning_rate": 2.8096864327848496e-06,
2136
+ "loss": 1.9267,
2137
+ "step": 152000
2138
+ },
2139
+ {
2140
+ "epoch": 47.34554486184415,
2141
+ "grad_norm": 3.6761834621429443,
2142
+ "learning_rate": 2.6544551381558526e-06,
2143
+ "loss": 1.9418,
2144
+ "step": 152500
2145
+ },
2146
+ {
2147
+ "epoch": 47.500776156473144,
2148
+ "grad_norm": 3.6187174320220947,
2149
+ "learning_rate": 2.4992238435268553e-06,
2150
+ "loss": 1.9362,
2151
+ "step": 153000
2152
+ },
2153
+ {
2154
+ "epoch": 47.65600745110214,
2155
+ "grad_norm": 3.6399784088134766,
2156
+ "learning_rate": 2.343992548897858e-06,
2157
+ "loss": 1.9375,
2158
+ "step": 153500
2159
+ },
2160
+ {
2161
+ "epoch": 47.81123874573114,
2162
+ "grad_norm": 3.451267957687378,
2163
+ "learning_rate": 2.188761254268861e-06,
2164
+ "loss": 1.9349,
2165
+ "step": 154000
2166
+ },
2167
+ {
2168
+ "epoch": 47.966470040360136,
2169
+ "grad_norm": 3.334773540496826,
2170
+ "learning_rate": 2.0335299596398635e-06,
2171
+ "loss": 1.9418,
2172
+ "step": 154500
2173
+ },
2174
+ {
2175
+ "epoch": 48.12170133498913,
2176
+ "grad_norm": 3.2175698280334473,
2177
+ "learning_rate": 1.8782986650108664e-06,
2178
+ "loss": 1.9339,
2179
+ "step": 155000
2180
+ },
2181
+ {
2182
+ "epoch": 48.27693262961813,
2183
+ "grad_norm": 3.638575315475464,
2184
+ "learning_rate": 1.723067370381869e-06,
2185
+ "loss": 1.9272,
2186
+ "step": 155500
2187
+ },
2188
+ {
2189
+ "epoch": 48.43216392424713,
2190
+ "grad_norm": 3.2722606658935547,
2191
+ "learning_rate": 1.567836075752872e-06,
2192
+ "loss": 1.9311,
2193
+ "step": 156000
2194
+ },
2195
+ {
2196
+ "epoch": 48.587395218876125,
2197
+ "grad_norm": 3.3455374240875244,
2198
+ "learning_rate": 1.4126047811238747e-06,
2199
+ "loss": 1.9318,
2200
+ "step": 156500
2201
+ },
2202
+ {
2203
+ "epoch": 48.74262651350512,
2204
+ "grad_norm": 3.634216785430908,
2205
+ "learning_rate": 1.2573734864948775e-06,
2206
+ "loss": 1.9335,
2207
+ "step": 157000
2208
+ },
2209
+ {
2210
+ "epoch": 48.89785780813412,
2211
+ "grad_norm": 3.6753644943237305,
2212
+ "learning_rate": 1.1021421918658803e-06,
2213
+ "loss": 1.9397,
2214
+ "step": 157500
2215
+ },
2216
+ {
2217
+ "epoch": 49.05308910276312,
2218
+ "grad_norm": 3.3982911109924316,
2219
+ "learning_rate": 9.469108972368831e-07,
2220
+ "loss": 1.9312,
2221
+ "step": 158000
2222
+ },
2223
+ {
2224
+ "epoch": 49.208320397392114,
2225
+ "grad_norm": 3.403062343597412,
2226
+ "learning_rate": 7.916796026078859e-07,
2227
+ "loss": 1.9299,
2228
+ "step": 158500
2229
+ },
2230
+ {
2231
+ "epoch": 49.36355169202111,
2232
+ "grad_norm": 3.539356231689453,
2233
+ "learning_rate": 6.364483079788885e-07,
2234
+ "loss": 1.9314,
2235
+ "step": 159000
2236
+ },
2237
+ {
2238
+ "epoch": 49.51878298665011,
2239
+ "grad_norm": 3.549177885055542,
2240
+ "learning_rate": 4.812170133498913e-07,
2241
+ "loss": 1.9192,
2242
+ "step": 159500
2243
+ },
2244
+ {
2245
+ "epoch": 49.674014281279106,
2246
+ "grad_norm": 3.6044740676879883,
2247
+ "learning_rate": 3.259857187208941e-07,
2248
+ "loss": 1.9283,
2249
+ "step": 160000
2250
+ },
2251
+ {
2252
+ "epoch": 49.829245575908104,
2253
+ "grad_norm": 3.24859619140625,
2254
+ "learning_rate": 1.7075442409189693e-07,
2255
+ "loss": 1.9268,
2256
+ "step": 160500
2257
+ },
2258
+ {
2259
+ "epoch": 49.9844768705371,
2260
+ "grad_norm": 3.433534860610962,
2261
+ "learning_rate": 1.552312946289972e-08,
2262
+ "loss": 1.9389,
2263
+ "step": 161000
2264
+ },
2265
+ {
2266
+ "epoch": 50.0,
2267
+ "step": 161050,
2268
+ "total_flos": 2.8277448966144e+17,
2269
+ "train_loss": 2.6664713644307803,
2270
+ "train_runtime": 39272.3209,
2271
+ "train_samples_per_second": 262.414,
2272
+ "train_steps_per_second": 4.101
2273
+ }
2274
+ ],
2275
+ "logging_steps": 500,
2276
+ "max_steps": 161050,
2277
+ "num_input_tokens_seen": 0,
2278
+ "num_train_epochs": 50,
2279
+ "save_steps": 500,
2280
+ "stateful_callbacks": {
2281
+ "TrainerControl": {
2282
+ "args": {
2283
+ "should_epoch_stop": false,
2284
+ "should_evaluate": false,
2285
+ "should_log": false,
2286
+ "should_save": true,
2287
+ "should_training_stop": true
2288
+ },
2289
+ "attributes": {}
2290
+ }
2291
+ },
2292
+ "total_flos": 2.8277448966144e+17,
2293
+ "train_batch_size": 64,
2294
+ "trial_name": null,
2295
+ "trial_params": null
2296
+ }