Hugging Face's logo

bradmin
/
reward-gpt-b6

Text Classification
Transformers
PyTorch
gpt_neox
Generated from Trainer
text-generation-inference
Inference Endpoints
Model card Files Community
bradmin commited on
Commit
58fdab3
•
1 Parent(s): 79f6e66

Training in progress, step 6000, checkpoint

Browse files
Files changed (10) hide show
  1. checkpoint-6000/config.json +36 -0
  2. checkpoint-6000/optimizer.pt +3 -0
  3. checkpoint-6000/pytorch_model.bin +3 -0
  4. checkpoint-6000/rng_state.pth +3 -0
  5. checkpoint-6000/scheduler.pt +3 -0
  6. checkpoint-6000/special_tokens_map.json +11 -0
  7. checkpoint-6000/tokenizer.json +0 -0
  8. checkpoint-6000/tokenizer_config.json +72 -0
  9. checkpoint-6000/trainer_state.json +3727 -0
  10. checkpoint-6000/training_args.bin +3 -0
checkpoint-6000/config.json ADDED
@@ -0,0 +1,36 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "EleutherAI/polyglot-ko-1.3b",
3
+ "architectures": [
4
+ "GPTNeoXForSequenceClassification"
5
+ ],
6
+ "attention_dropout": 0.0,
7
+ "bos_token_id": 0,
8
+ "classifier_dropout": 0.1,
9
+ "eos_token_id": 2,
10
+ "hidden_act": "gelu",
11
+ "hidden_dropout": 0.0,
12
+ "hidden_size": 2048,
13
+ "id2label": {
14
+ "0": "LABEL_0"
15
+ },
16
+ "initializer_range": 0.02,
17
+ "intermediate_size": 8192,
18
+ "label2id": {
19
+ "LABEL_0": 0
20
+ },
21
+ "layer_norm_eps": 1e-05,
22
+ "max_position_embeddings": 2048,
23
+ "model_type": "gpt_neox",
24
+ "num_attention_heads": 16,
25
+ "num_hidden_layers": 24,
26
+ "pad_token_id": 2,
27
+ "rope_scaling": null,
28
+ "rotary_emb_base": 10000,
29
+ "rotary_pct": 0.5,
30
+ "tie_word_embeddings": false,
31
+ "torch_dtype": "float32",
32
+ "transformers_version": "4.34.1",
33
+ "use_cache": true,
34
+ "use_parallel_residual": true,
35
+ "vocab_size": 30080
36
+ }
checkpoint-6000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b41575baf30a2efaad07ac3d04527f630927a8871a31b6cf44774ff01f3b96d3
3
+ size 10161845830
checkpoint-6000/pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3c772cd4beedbc52e4449635f6a5ef3013799421b9ce2dfcd44951909984d358
3
+ size 5080945162
checkpoint-6000/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3fe69834ffc4b48a54d2b6d6b534eab2cee3412219e35fd5cad617b8f23beaf6
3
+ size 14308
checkpoint-6000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5e0c58164c1ee0d607a7a629ed0a056d63b61f2d0d49a4f0d7a436cfb34e8789
3
+ size 1064
checkpoint-6000/special_tokens_map.json ADDED
@@ -0,0 +1,11 @@
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "additional_special_tokens": [
3
+ "<|endoftext|>",
4
+ "<|sep|>",
5
+ "<|acc|>",
6
+ "<|tel|>",
7
+ "<|rrn|>"
8
+ ],
9
+ "eos_token": "<|endoftext|>",
10
+ "pad_token": "<|endoftext|>"
11
+ }
checkpoint-6000/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-6000/tokenizer_config.json ADDED
@@ -0,0 +1,72 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "<|unused0|>",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "1": {
12
+ "content": "<|unused1|>",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "2": {
20
+ "content": "<|endoftext|>",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "3": {
28
+ "content": "<|sep|>",
29
+ "lstrip": false,
30
+ "normalized": false,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "30000": {
36
+ "content": "<|acc|>",
37
+ "lstrip": false,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ },
43
+ "30001": {
44
+ "content": "<|tel|>",
45
+ "lstrip": false,
46
+ "normalized": false,
47
+ "rstrip": false,
48
+ "single_word": false,
49
+ "special": true
50
+ },
51
+ "30002": {
52
+ "content": "<|rrn|>",
53
+ "lstrip": false,
54
+ "normalized": false,
55
+ "rstrip": false,
56
+ "single_word": false,
57
+ "special": true
58
+ }
59
+ },
60
+ "additional_special_tokens": [
61
+ "<|endoftext|>",
62
+ "<|sep|>",
63
+ "<|acc|>",
64
+ "<|tel|>",
65
+ "<|rrn|>"
66
+ ],
67
+ "clean_up_tokenization_spaces": true,
68
+ "eos_token": "<|endoftext|>",
69
+ "model_max_length": 1000000000000000019884624838656,
70
+ "pad_token": "<|endoftext|>",
71
+ "tokenizer_class": "PreTrainedTokenizerFast"
72
+ }
checkpoint-6000/trainer_state.json ADDED
@@ -0,0 +1,3727 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 0.8554807801984715,
5
+ "eval_steps": 500,
6
+ "global_step": 6000,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.0,
13
+ "learning_rate": 8.999954848325688e-06,
14
+ "loss": 0.7277,
15
+ "step": 10
16
+ },
17
+ {
18
+ "epoch": 0.0,
19
+ "learning_rate": 8.999819394208827e-06,
20
+ "loss": 0.6455,
21
+ "step": 20
22
+ },
23
+ {
24
+ "epoch": 0.0,
25
+ "learning_rate": 8.999593640367633e-06,
26
+ "loss": 0.5428,
27
+ "step": 30
28
+ },
29
+ {
30
+ "epoch": 0.01,
31
+ "learning_rate": 8.9992775913324e-06,
32
+ "loss": 0.3486,
33
+ "step": 40
34
+ },
35
+ {
36
+ "epoch": 0.01,
37
+ "learning_rate": 8.998871253445414e-06,
38
+ "loss": 0.3524,
39
+ "step": 50
40
+ },
41
+ {
42
+ "epoch": 0.01,
43
+ "learning_rate": 8.998374634860822e-06,
44
+ "loss": 0.2701,
45
+ "step": 60
46
+ },
47
+ {
48
+ "epoch": 0.01,
49
+ "learning_rate": 8.997787745544478e-06,
50
+ "loss": 0.2106,
51
+ "step": 70
52
+ },
53
+ {
54
+ "epoch": 0.01,
55
+ "learning_rate": 8.997110597273727e-06,
56
+ "loss": 0.2596,
57
+ "step": 80
58
+ },
59
+ {
60
+ "epoch": 0.01,
61
+ "learning_rate": 8.996343203637181e-06,
62
+ "loss": 0.2015,
63
+ "step": 90
64
+ },
65
+ {
66
+ "epoch": 0.01,
67
+ "learning_rate": 8.995485580034448e-06,
68
+ "loss": 0.2462,
69
+ "step": 100
70
+ },
71
+ {
72
+ "epoch": 0.02,
73
+ "learning_rate": 8.99453774367581e-06,
74
+ "loss": 0.2172,
75
+ "step": 110
76
+ },
77
+ {
78
+ "epoch": 0.02,
79
+ "learning_rate": 8.993499713581887e-06,
80
+ "loss": 0.2655,
81
+ "step": 120
82
+ },
83
+ {
84
+ "epoch": 0.02,
85
+ "learning_rate": 8.992371510583257e-06,
86
+ "loss": 0.1823,
87
+ "step": 130
88
+ },
89
+ {
90
+ "epoch": 0.02,
91
+ "learning_rate": 8.991153157320033e-06,
92
+ "loss": 0.227,
93
+ "step": 140
94
+ },
95
+ {
96
+ "epoch": 0.02,
97
+ "learning_rate": 8.989844678241412e-06,
98
+ "loss": 0.2243,
99
+ "step": 150
100
+ },
101
+ {
102
+ "epoch": 0.02,
103
+ "learning_rate": 8.988446099605179e-06,
104
+ "loss": 0.1787,
105
+ "step": 160
106
+ },
107
+ {
108
+ "epoch": 0.02,
109
+ "learning_rate": 8.986957449477188e-06,
110
+ "loss": 0.2211,
111
+ "step": 170
112
+ },
113
+ {
114
+ "epoch": 0.03,
115
+ "learning_rate": 8.98537875773079e-06,
116
+ "loss": 0.1683,
117
+ "step": 180
118
+ },
119
+ {
120
+ "epoch": 0.03,
121
+ "learning_rate": 8.983710056046243e-06,
122
+ "loss": 0.1879,
123
+ "step": 190
124
+ },
125
+ {
126
+ "epoch": 0.03,
127
+ "learning_rate": 8.98195137791007e-06,
128
+ "loss": 0.1809,
129
+ "step": 200
130
+ },
131
+ {
132
+ "epoch": 0.03,
133
+ "learning_rate": 8.980102758614384e-06,
134
+ "loss": 0.2204,
135
+ "step": 210
136
+ },
137
+ {
138
+ "epoch": 0.03,
139
+ "learning_rate": 8.978164235256191e-06,
140
+ "loss": 0.189,
141
+ "step": 220
142
+ },
143
+ {
144
+ "epoch": 0.03,
145
+ "learning_rate": 8.976135846736634e-06,
146
+ "loss": 0.1389,
147
+ "step": 230
148
+ },
149
+ {
150
+ "epoch": 0.03,
151
+ "learning_rate": 8.97401763376022e-06,
152
+ "loss": 0.1231,
153
+ "step": 240
154
+ },
155
+ {
156
+ "epoch": 0.04,
157
+ "learning_rate": 8.971809638833998e-06,
158
+ "loss": 0.2018,
159
+ "step": 250
160
+ },
161
+ {
162
+ "epoch": 0.04,
163
+ "learning_rate": 8.96951190626671e-06,
164
+ "loss": 0.1271,
165
+ "step": 260
166
+ },
167
+ {
168
+ "epoch": 0.04,
169
+ "learning_rate": 8.967124482167896e-06,
170
+ "loss": 0.1364,
171
+ "step": 270
172
+ },
173
+ {
174
+ "epoch": 0.04,
175
+ "learning_rate": 8.964647414446981e-06,
176
+ "loss": 0.1658,
177
+ "step": 280
178
+ },
179
+ {
180
+ "epoch": 0.04,
181
+ "learning_rate": 8.9620807528123e-06,
182
+ "loss": 0.1479,
183
+ "step": 290
184
+ },
185
+ {
186
+ "epoch": 0.04,
187
+ "learning_rate": 8.959424548770104e-06,
188
+ "loss": 0.1353,
189
+ "step": 300
190
+ },
191
+ {
192
+ "epoch": 0.04,
193
+ "learning_rate": 8.95667885562353e-06,
194
+ "loss": 0.1761,
195
+ "step": 310
196
+ },
197
+ {
198
+ "epoch": 0.05,
199
+ "learning_rate": 8.953843728471533e-06,
200
+ "loss": 0.1391,
201
+ "step": 320
202
+ },
203
+ {
204
+ "epoch": 0.05,
205
+ "learning_rate": 8.950919224207774e-06,
206
+ "loss": 0.1776,
207
+ "step": 330
208
+ },
209
+ {
210
+ "epoch": 0.05,
211
+ "learning_rate": 8.94790540151948e-06,
212
+ "loss": 0.2089,
213
+ "step": 340
214
+ },
215
+ {
216
+ "epoch": 0.05,
217
+ "learning_rate": 8.94480232088627e-06,
218
+ "loss": 0.1368,
219
+ "step": 350
220
+ },
221
+ {
222
+ "epoch": 0.05,
223
+ "learning_rate": 8.941610044578937e-06,
224
+ "loss": 0.1501,
225
+ "step": 360
226
+ },
227
+ {
228
+ "epoch": 0.05,
229
+ "learning_rate": 8.938328636658202e-06,
230
+ "loss": 0.1807,
231
+ "step": 370
232
+ },
233
+ {
234
+ "epoch": 0.05,
235
+ "learning_rate": 8.934958162973425e-06,
236
+ "loss": 0.13,
237
+ "step": 380
238
+ },
239
+ {
240
+ "epoch": 0.06,
241
+ "learning_rate": 8.931498691161287e-06,
242
+ "loss": 0.1437,
243
+ "step": 390
244
+ },
245
+ {
246
+ "epoch": 0.06,
247
+ "learning_rate": 8.92795029064443e-06,
248
+ "loss": 0.0828,
249
+ "step": 400
250
+ },
251
+ {
252
+ "epoch": 0.06,
253
+ "learning_rate": 8.924313032630063e-06,
254
+ "loss": 0.1041,
255
+ "step": 410
256
+ },
257
+ {
258
+ "epoch": 0.06,
259
+ "learning_rate": 8.920586990108539e-06,
260
+ "loss": 0.2663,
261
+ "step": 420
262
+ },
263
+ {
264
+ "epoch": 0.06,
265
+ "learning_rate": 8.916772237851882e-06,
266
+ "loss": 0.2055,
267
+ "step": 430
268
+ },
269
+ {
270
+ "epoch": 0.06,
271
+ "learning_rate": 8.912868852412294e-06,
272
+ "loss": 0.151,
273
+ "step": 440
274
+ },
275
+ {
276
+ "epoch": 0.06,
277
+ "learning_rate": 8.908876912120614e-06,
278
+ "loss": 0.1638,
279
+ "step": 450
280
+ },
281
+ {
282
+ "epoch": 0.07,
283
+ "learning_rate": 8.904796497084747e-06,
284
+ "loss": 0.1399,
285
+ "step": 460
286
+ },
287
+ {
288
+ "epoch": 0.07,
289
+ "learning_rate": 8.900627689188059e-06,
290
+ "loss": 0.1615,
291
+ "step": 470
292
+ },
293
+ {
294
+ "epoch": 0.07,
295
+ "learning_rate": 8.89637057208773e-06,
296
+ "loss": 0.1209,
297
+ "step": 480
298
+ },
299
+ {
300
+ "epoch": 0.07,
301
+ "learning_rate": 8.892025231213077e-06,
302
+ "loss": 0.1605,
303
+ "step": 490
304
+ },
305
+ {
306
+ "epoch": 0.07,
307
+ "learning_rate": 8.887591753763842e-06,
308
+ "loss": 0.124,
309
+ "step": 500
310
+ },
311
+ {
312
+ "epoch": 0.07,
313
+ "eval_accuracy": 0.0,
314
+ "eval_loss": 0.13451002538204193,
315
+ "eval_runtime": 446.6893,
316
+ "eval_samples_per_second": 52.338,
317
+ "eval_steps_per_second": 8.724,
318
+ "step": 500
319
+ },
320
+ {
321
+ "epoch": 0.07,
322
+ "learning_rate": 8.883070228708436e-06,
323
+ "loss": 0.1045,
324
+ "step": 510
325
+ },
326
+ {
327
+ "epoch": 0.07,
328
+ "learning_rate": 8.878460746782163e-06,
329
+ "loss": 0.2208,
330
+ "step": 520
331
+ },
332
+ {
333
+ "epoch": 0.08,
334
+ "learning_rate": 8.873763400485387e-06,
335
+ "loss": 0.1483,
336
+ "step": 530
337
+ },
338
+ {
339
+ "epoch": 0.08,
340
+ "learning_rate": 8.868978284081688e-06,
341
+ "loss": 0.1322,
342
+ "step": 540
343
+ },
344
+ {
345
+ "epoch": 0.08,
346
+ "learning_rate": 8.864105493595962e-06,
347
+ "loss": 0.1227,
348
+ "step": 550
349
+ },
350
+ {
351
+ "epoch": 0.08,
352
+ "learning_rate": 8.859145126812497e-06,
353
+ "loss": 0.1371,
354
+ "step": 560
355
+ },
356
+ {
357
+ "epoch": 0.08,
358
+ "learning_rate": 8.854097283273012e-06,
359
+ "loss": 0.1909,
360
+ "step": 570
361
+ },
362
+ {
363
+ "epoch": 0.08,
364
+ "learning_rate": 8.848962064274655e-06,
365
+ "loss": 0.0993,
366
+ "step": 580
367
+ },
368
+ {
369
+ "epoch": 0.08,
370
+ "learning_rate": 8.843739572867977e-06,
371
+ "loss": 0.1053,
372
+ "step": 590
373
+ },
374
+ {
375
+ "epoch": 0.09,
376
+ "learning_rate": 8.838429913854859e-06,
377
+ "loss": 0.1401,
378
+ "step": 600
379
+ },
380
+ {
381
+ "epoch": 0.09,
382
+ "learning_rate": 8.83303319378641e-06,
383
+ "loss": 0.09,
384
+ "step": 610
385
+ },
386
+ {
387
+ "epoch": 0.09,
388
+ "learning_rate": 8.827549520960827e-06,
389
+ "loss": 0.1667,
390
+ "step": 620
391
+ },
392
+ {
393
+ "epoch": 0.09,
394
+ "learning_rate": 8.821979005421223e-06,
395
+ "loss": 0.1069,
396
+ "step": 630
397
+ },
398
+ {
399
+ "epoch": 0.09,
400
+ "learning_rate": 8.816321758953426e-06,
401
+ "loss": 0.1659,
402
+ "step": 640
403
+ },
404
+ {
405
+ "epoch": 0.09,
406
+ "learning_rate": 8.810577895083726e-06,
407
+ "loss": 0.1309,
408
+ "step": 650
409
+ },
410
+ {
411
+ "epoch": 0.09,
412
+ "learning_rate": 8.804747529076592e-06,
413
+ "loss": 0.16,
414
+ "step": 660
415
+ },
416
+ {
417
+ "epoch": 0.1,
418
+ "learning_rate": 8.79883077793238e-06,
419
+ "loss": 0.1041,
420
+ "step": 670
421
+ },
422
+ {
423
+ "epoch": 0.1,
424
+ "learning_rate": 8.792827760384965e-06,
425
+ "loss": 0.1344,
426
+ "step": 680
427
+ },
428
+ {
429
+ "epoch": 0.1,
430
+ "learning_rate": 8.786738596899362e-06,
431
+ "loss": 0.0922,
432
+ "step": 690
433
+ },
434
+ {
435
+ "epoch": 0.1,
436
+ "learning_rate": 8.780563409669322e-06,
437
+ "loss": 0.0813,
438
+ "step": 700
439
+ },
440
+ {
441
+ "epoch": 0.1,
442
+ "learning_rate": 8.774302322614858e-06,
443
+ "loss": 0.1795,
444
+ "step": 710
445
+ },
446
+ {
447
+ "epoch": 0.1,
448
+ "learning_rate": 8.767955461379783e-06,
449
+ "loss": 0.1409,
450
+ "step": 720
451
+ },
452
+ {
453
+ "epoch": 0.1,
454
+ "learning_rate": 8.761522953329163e-06,
455
+ "loss": 0.1362,
456
+ "step": 730
457
+ },
458
+ {
459
+ "epoch": 0.11,
460
+ "learning_rate": 8.75500492754678e-06,
461
+ "loss": 0.1227,
462
+ "step": 740
463
+ },
464
+ {
465
+ "epoch": 0.11,
466
+ "learning_rate": 8.74840151483254e-06,
467
+ "loss": 0.0643,
468
+ "step": 750
469
+ },
470
+ {
471
+ "epoch": 0.11,
472
+ "learning_rate": 8.741712847699832e-06,
473
+ "loss": 0.1895,
474
+ "step": 760
475
+ },
476
+ {
477
+ "epoch": 0.11,
478
+ "learning_rate": 8.734939060372892e-06,
479
+ "loss": 0.0959,
480
+ "step": 770
481
+ },
482
+ {
483
+ "epoch": 0.11,
484
+ "learning_rate": 8.72808028878409e-06,
485
+ "loss": 0.1089,
486
+ "step": 780
487
+ },
488
+ {
489
+ "epoch": 0.11,
490
+ "learning_rate": 8.721136670571216e-06,
491
+ "loss": 0.0959,
492
+ "step": 790
493
+ },
494
+ {
495
+ "epoch": 0.11,
496
+ "learning_rate": 8.714108345074704e-06,
497
+ "loss": 0.1264,
498
+ "step": 800
499
+ },
500
+ {
501
+ "epoch": 0.12,
502
+ "learning_rate": 8.706995453334855e-06,
503
+ "loss": 0.1078,
504
+ "step": 810
505
+ },
506
+ {
507
+ "epoch": 0.12,
508
+ "learning_rate": 8.699798138088988e-06,
509
+ "loss": 0.1278,
510
+ "step": 820
511
+ },
512
+ {
513
+ "epoch": 0.12,
514
+ "learning_rate": 8.692516543768581e-06,
515
+ "loss": 0.1337,
516
+ "step": 830
517
+ },
518
+ {
519
+ "epoch": 0.12,
520
+ "learning_rate": 8.685150816496384e-06,
521
+ "loss": 0.0815,
522
+ "step": 840
523
+ },
524
+ {
525
+ "epoch": 0.12,
526
+ "learning_rate": 8.677701104083467e-06,
527
+ "loss": 0.1033,
528
+ "step": 850
529
+ },
530
+ {
531
+ "epoch": 0.12,
532
+ "learning_rate": 8.670167556026274e-06,
533
+ "loss": 0.1412,
534
+ "step": 860
535
+ },
536
+ {
537
+ "epoch": 0.12,
538
+ "learning_rate": 8.662550323503606e-06,
539
+ "loss": 0.197,
540
+ "step": 870
541
+ },
542
+ {
543
+ "epoch": 0.13,
544
+ "learning_rate": 8.654849559373595e-06,
545
+ "loss": 0.1118,
546
+ "step": 880
547
+ },
548
+ {
549
+ "epoch": 0.13,
550
+ "learning_rate": 8.647065418170644e-06,
551
+ "loss": 0.1132,
552
+ "step": 890
553
+ },
554
+ {
555
+ "epoch": 0.13,
556
+ "learning_rate": 8.639198056102309e-06,
557
+ "loss": 0.0928,
558
+ "step": 900
559
+ },
560
+ {
561
+ "epoch": 0.13,
562
+ "learning_rate": 8.631247631046176e-06,
563
+ "loss": 0.0723,
564
+ "step": 910
565
+ },
566
+ {
567
+ "epoch": 0.13,
568
+ "learning_rate": 8.623214302546691e-06,
569
+ "loss": 0.126,
570
+ "step": 920
571
+ },
572
+ {
573
+ "epoch": 0.13,
574
+ "learning_rate": 8.615098231811958e-06,
575
+ "loss": 0.1363,
576
+ "step": 930
577
+ },
578
+ {
579
+ "epoch": 0.13,
580
+ "learning_rate": 8.606899581710505e-06,
581
+ "loss": 0.1243,
582
+ "step": 940
583
+ },
584
+ {
585
+ "epoch": 0.14,
586
+ "learning_rate": 8.598618516768005e-06,
587
+ "loss": 0.0382,
588
+ "step": 950
589
+ },
590
+ {
591
+ "epoch": 0.14,
592
+ "learning_rate": 8.590255203163995e-06,
593
+ "loss": 0.1295,
594
+ "step": 960
595
+ },
596
+ {
597
+ "epoch": 0.14,
598
+ "learning_rate": 8.581809808728524e-06,
599
+ "loss": 0.1145,
600
+ "step": 970
601
+ },
602
+ {
603
+ "epoch": 0.14,
604
+ "learning_rate": 8.573282502938791e-06,
605
+ "loss": 0.0915,
606
+ "step": 980
607
+ },
608
+ {
609
+ "epoch": 0.14,
610
+ "learning_rate": 8.564673456915746e-06,
611
+ "loss": 0.1918,
612
+ "step": 990
613
+ },
614
+ {
615
+ "epoch": 0.14,
616
+ "learning_rate": 8.55598284342065e-06,
617
+ "loss": 0.1732,
618
+ "step": 1000
619
+ },
620
+ {
621
+ "epoch": 0.14,
622
+ "eval_accuracy": 0.0,
623
+ "eval_loss": 0.12965914607048035,
624
+ "eval_runtime": 448.5256,
625
+ "eval_samples_per_second": 52.124,
626
+ "eval_steps_per_second": 8.688,
627
+ "step": 1000
628
+ },
629
+ {
630
+ "epoch": 0.14,
631
+ "learning_rate": 8.547210836851614e-06,
632
+ "loss": 0.0833,
633
+ "step": 1010
634
+ },
635
+ {
636
+ "epoch": 0.15,
637
+ "learning_rate": 8.538357613240099e-06,
638
+ "loss": 0.1597,
639
+ "step": 1020
640
+ },
641
+ {
642
+ "epoch": 0.15,
643
+ "learning_rate": 8.52942335024738e-06,
644
+ "loss": 0.1041,
645
+ "step": 1030
646
+ },
647
+ {
648
+ "epoch": 0.15,
649
+ "learning_rate": 8.520408227160983e-06,
650
+ "loss": 0.1819,
651
+ "step": 1040
652
+ },
653
+ {
654
+ "epoch": 0.15,
655
+ "learning_rate": 8.511312424891083e-06,
656
+ "loss": 0.1112,
657
+ "step": 1050
658
+ },
659
+ {
660
+ "epoch": 0.15,
661
+ "learning_rate": 8.502136125966886e-06,
662
+ "loss": 0.1372,
663
+ "step": 1060
664
+ },
665
+ {
666
+ "epoch": 0.15,
667
+ "learning_rate": 8.49287951453295e-06,
668
+ "loss": 0.1596,
669
+ "step": 1070
670
+ },
671
+ {
672
+ "epoch": 0.15,
673
+ "learning_rate": 8.483542776345496e-06,
674
+ "loss": 0.1558,
675
+ "step": 1080
676
+ },
677
+ {
678
+ "epoch": 0.16,
679
+ "learning_rate": 8.47412609876869e-06,
680
+ "loss": 0.069,
681
+ "step": 1090
682
+ },
683
+ {
684
+ "epoch": 0.16,
685
+ "learning_rate": 8.464629670770866e-06,
686
+ "loss": 0.1468,
687
+ "step": 1100
688
+ },
689
+ {
690
+ "epoch": 0.16,
691
+ "learning_rate": 8.455053682920747e-06,
692
+ "loss": 0.106,
693
+ "step": 1110
694
+ },
695
+ {
696
+ "epoch": 0.16,
697
+ "learning_rate": 8.445398327383616e-06,
698
+ "loss": 0.1562,
699
+ "step": 1120
700
+ },
701
+ {
702
+ "epoch": 0.16,
703
+ "learning_rate": 8.435663797917459e-06,
704
+ "loss": 0.1404,
705
+ "step": 1130
706
+ },
707
+ {
708
+ "epoch": 0.16,
709
+ "learning_rate": 8.425850289869074e-06,
710
+ "loss": 0.1198,
711
+ "step": 1140
712
+ },
713
+ {
714
+ "epoch": 0.16,
715
+ "learning_rate": 8.415958000170163e-06,
716
+ "loss": 0.107,
717
+ "step": 1150
718
+ },
719
+ {
720
+ "epoch": 0.17,
721
+ "learning_rate": 8.405987127333367e-06,
722
+ "loss": 0.0911,
723
+ "step": 1160
724
+ },
725
+ {
726
+ "epoch": 0.17,
727
+ "learning_rate": 8.395937871448282e-06,
728
+ "loss": 0.066,
729
+ "step": 1170
730
+ },
731
+ {
732
+ "epoch": 0.17,
733
+ "learning_rate": 8.385810434177462e-06,
734
+ "loss": 0.0988,
735
+ "step": 1180
736
+ },
737
+ {
738
+ "epoch": 0.17,
739
+ "learning_rate": 8.375605018752344e-06,
740
+ "loss": 0.1648,
741
+ "step": 1190
742
+ },
743
+ {
744
+ "epoch": 0.17,
745
+ "learning_rate": 8.365321829969196e-06,
746
+ "loss": 0.1124,
747
+ "step": 1200
748
+ },
749
+ {
750
+ "epoch": 0.17,
751
+ "learning_rate": 8.35496107418499e-06,
752
+ "loss": 0.1251,
753
+ "step": 1210
754
+ },
755
+ {
756
+ "epoch": 0.17,
757
+ "learning_rate": 8.34452295931327e-06,
758
+ "loss": 0.1167,
759
+ "step": 1220
760
+ },
761
+ {
762
+ "epoch": 0.18,
763
+ "learning_rate": 8.334007694819972e-06,
764
+ "loss": 0.1238,
765
+ "step": 1230
766
+ },
767
+ {
768
+ "epoch": 0.18,
769
+ "learning_rate": 8.323415491719233e-06,
770
+ "loss": 0.1127,
771
+ "step": 1240
772
+ },
773
+ {
774
+ "epoch": 0.18,
775
+ "learning_rate": 8.312746562569141e-06,
776
+ "loss": 0.0823,
777
+ "step": 1250
778
+ },
779
+ {
780
+ "epoch": 0.18,
781
+ "learning_rate": 8.302001121467481e-06,
782
+ "loss": 0.145,
783
+ "step": 1260
784
+ },
785
+ {
786
+ "epoch": 0.18,
787
+ "learning_rate": 8.291179384047432e-06,
788
+ "loss": 0.1257,
789
+ "step": 1270
790
+ },
791
+ {
792
+ "epoch": 0.18,
793
+ "learning_rate": 8.280281567473247e-06,
794
+ "loss": 0.1014,
795
+ "step": 1280
796
+ },
797
+ {
798
+ "epoch": 0.18,
799
+ "learning_rate": 8.269307890435886e-06,
800
+ "loss": 0.1289,
801
+ "step": 1290
802
+ },
803
+ {
804
+ "epoch": 0.19,
805
+ "learning_rate": 8.258258573148637e-06,
806
+ "loss": 0.104,
807
+ "step": 1300
808
+ },
809
+ {
810
+ "epoch": 0.19,
811
+ "learning_rate": 8.247133837342685e-06,
812
+ "loss": 0.1231,
813
+ "step": 1310
814
+ },
815
+ {
816
+ "epoch": 0.19,
817
+ "learning_rate": 8.235933906262679e-06,
818
+ "loss": 0.1594,
819
+ "step": 1320
820
+ },
821
+ {
822
+ "epoch": 0.19,
823
+ "learning_rate": 8.22465900466223e-06,
824
+ "loss": 0.0999,
825
+ "step": 1330
826
+ },
827
+ {
828
+ "epoch": 0.19,
829
+ "learning_rate": 8.213309358799427e-06,
830
+ "loss": 0.061,
831
+ "step": 1340
832
+ },
833
+ {
834
+ "epoch": 0.19,
835
+ "learning_rate": 8.201885196432273e-06,
836
+ "loss": 0.0618,
837
+ "step": 1350
838
+ },
839
+ {
840
+ "epoch": 0.19,
841
+ "learning_rate": 8.190386746814125e-06,
842
+ "loss": 0.1607,
843
+ "step": 1360
844
+ },
845
+ {
846
+ "epoch": 0.2,
847
+ "learning_rate": 8.1788142406891e-06,
848
+ "loss": 0.0726,
849
+ "step": 1370
850
+ },
851
+ {
852
+ "epoch": 0.2,
853
+ "learning_rate": 8.167167910287428e-06,
854
+ "loss": 0.0594,
855
+ "step": 1380
856
+ },
857
+ {
858
+ "epoch": 0.2,
859
+ "learning_rate": 8.155447989320808e-06,
860
+ "loss": 0.1341,
861
+ "step": 1390
862
+ },
863
+ {
864
+ "epoch": 0.2,
865
+ "learning_rate": 8.14365471297771e-06,
866
+ "loss": 0.1253,
867
+ "step": 1400
868
+ },
869
+ {
870
+ "epoch": 0.2,
871
+ "learning_rate": 8.131788317918651e-06,
872
+ "loss": 0.0865,
873
+ "step": 1410
874
+ },
875
+ {
876
+ "epoch": 0.2,
877
+ "learning_rate": 8.119849042271458e-06,
878
+ "loss": 0.1227,
879
+ "step": 1420
880
+ },
881
+ {
882
+ "epoch": 0.2,
883
+ "learning_rate": 8.10783712562648e-06,
884
+ "loss": 0.1024,
885
+ "step": 1430
886
+ },
887
+ {
888
+ "epoch": 0.21,
889
+ "learning_rate": 8.095752809031783e-06,
890
+ "loss": 0.1495,
891
+ "step": 1440
892
+ },
893
+ {
894
+ "epoch": 0.21,
895
+ "learning_rate": 8.083596334988313e-06,
896
+ "loss": 0.1098,
897
+ "step": 1450
898
+ },
899
+ {
900
+ "epoch": 0.21,
901
+ "learning_rate": 8.071367947445025e-06,
902
+ "loss": 0.1044,
903
+ "step": 1460
904
+ },
905
+ {
906
+ "epoch": 0.21,
907
+ "learning_rate": 8.059067891794002e-06,
908
+ "loss": 0.0757,
909
+ "step": 1470
910
+ },
911
+ {
912
+ "epoch": 0.21,
913
+ "learning_rate": 8.046696414865507e-06,
914
+ "loss": 0.0991,
915
+ "step": 1480
916
+ },
917
+ {
918
+ "epoch": 0.21,
919
+ "learning_rate": 8.034253764923052e-06,
920
+ "loss": 0.141,
921
+ "step": 1490
922
+ },
923
+ {
924
+ "epoch": 0.21,
925
+ "learning_rate": 8.021740191658405e-06,
926
+ "loss": 0.1023,
927
+ "step": 1500
928
+ },
929
+ {
930
+ "epoch": 0.21,
931
+ "eval_accuracy": 0.0,
932
+ "eval_loss": 0.11160164326429367,
933
+ "eval_runtime": 448.3846,
934
+ "eval_samples_per_second": 52.141,
935
+ "eval_steps_per_second": 8.691,
936
+ "step": 1500
937
+ },
938
+ {
939
+ "epoch": 0.22,
940
+ "learning_rate": 8.009155946186583e-06,
941
+ "loss": 0.0782,
942
+ "step": 1510
943
+ },
944
+ {
945
+ "epoch": 0.22,
946
+ "learning_rate": 7.996501281040809e-06,
947
+ "loss": 0.1132,
948
+ "step": 1520
949
+ },
950
+ {
951
+ "epoch": 0.22,
952
+ "learning_rate": 7.983776450167446e-06,
953
+ "loss": 0.042,
954
+ "step": 1530
955
+ },
956
+ {
957
+ "epoch": 0.22,
958
+ "learning_rate": 7.970981708920902e-06,
959
+ "loss": 0.0896,
960
+ "step": 1540
961
+ },
962
+ {
963
+ "epoch": 0.22,
964
+ "learning_rate": 7.958117314058508e-06,
965
+ "loss": 0.1038,
966
+ "step": 1550
967
+ },
968
+ {
969
+ "epoch": 0.22,
970
+ "learning_rate": 7.945183523735359e-06,
971
+ "loss": 0.052,
972
+ "step": 1560
973
+ },
974
+ {
975
+ "epoch": 0.22,
976
+ "learning_rate": 7.932180597499137e-06,
977
+ "loss": 0.1224,
978
+ "step": 1570
979
+ },
980
+ {
981
+ "epoch": 0.23,
982
+ "learning_rate": 7.919108796284908e-06,
983
+ "loss": 0.1473,
984
+ "step": 1580
985
+ },
986
+ {
987
+ "epoch": 0.23,
988
+ "learning_rate": 7.905968382409874e-06,
989
+ "loss": 0.1499,
990
+ "step": 1590
991
+ },
992
+ {
993
+ "epoch": 0.23,
994
+ "learning_rate": 7.89275961956812e-06,
995
+ "loss": 0.0728,
996
+ "step": 1600
997
+ },
998
+ {
999
+ "epoch": 0.23,
1000
+ "learning_rate": 7.879482772825316e-06,
1001
+ "loss": 0.1159,
1002
+ "step": 1610
1003
+ },
1004
+ {
1005
+ "epoch": 0.23,
1006
+ "learning_rate": 7.8661381086134e-06,
1007
+ "loss": 0.1409,
1008
+ "step": 1620
1009
+ },
1010
+ {
1011
+ "epoch": 0.23,
1012
+ "learning_rate": 7.85272589472523e-06,
1013
+ "loss": 0.1065,
1014
+ "step": 1630
1015
+ },
1016
+ {
1017
+ "epoch": 0.23,
1018
+ "learning_rate": 7.839246400309213e-06,
1019
+ "loss": 0.1386,
1020
+ "step": 1640
1021
+ },
1022
+ {
1023
+ "epoch": 0.24,
1024
+ "learning_rate": 7.8256998958639e-06,
1025
+ "loss": 0.0995,
1026
+ "step": 1650
1027
+ },
1028
+ {
1029
+ "epoch": 0.24,
1030
+ "learning_rate": 7.812086653232559e-06,
1031
+ "loss": 0.0947,
1032
+ "step": 1660
1033
+ },
1034
+ {
1035
+ "epoch": 0.24,
1036
+ "learning_rate": 7.798406945597726e-06,
1037
+ "loss": 0.1507,
1038
+ "step": 1670
1039
+ },
1040
+ {
1041
+ "epoch": 0.24,
1042
+ "learning_rate": 7.784661047475712e-06,
1043
+ "loss": 0.0693,
1044
+ "step": 1680
1045
+ },
1046
+ {
1047
+ "epoch": 0.24,
1048
+ "learning_rate": 7.7708492347111e-06,
1049
+ "loss": 0.1194,
1050
+ "step": 1690
1051
+ },
1052
+ {
1053
+ "epoch": 0.24,
1054
+ "learning_rate": 7.756971784471213e-06,
1055
+ "loss": 0.1297,
1056
+ "step": 1700
1057
+ },
1058
+ {
1059
+ "epoch": 0.24,
1060
+ "learning_rate": 7.743028975240545e-06,
1061
+ "loss": 0.0933,
1062
+ "step": 1710
1063
+ },
1064
+ {
1065
+ "epoch": 0.25,
1066
+ "learning_rate": 7.729021086815177e-06,
1067
+ "loss": 0.1029,
1068
+ "step": 1720
1069
+ },
1070
+ {
1071
+ "epoch": 0.25,
1072
+ "learning_rate": 7.71494840029716e-06,
1073
+ "loss": 0.0572,
1074
+ "step": 1730
1075
+ },
1076
+ {
1077
+ "epoch": 0.25,
1078
+ "learning_rate": 7.700811198088876e-06,
1079
+ "loss": 0.0754,
1080
+ "step": 1740
1081
+ },
1082
+ {
1083
+ "epoch": 0.25,
1084
+ "learning_rate": 7.68660976388737e-06,
1085
+ "loss": 0.0957,
1086
+ "step": 1750
1087
+ },
1088
+ {
1089
+ "epoch": 0.25,
1090
+ "learning_rate": 7.672344382678653e-06,
1091
+ "loss": 0.1321,
1092
+ "step": 1760
1093
+ },
1094
+ {
1095
+ "epoch": 0.25,
1096
+ "learning_rate": 7.658015340731994e-06,
1097
+ "loss": 0.0711,
1098
+ "step": 1770
1099
+ },
1100
+ {
1101
+ "epoch": 0.25,
1102
+ "learning_rate": 7.643622925594161e-06,
1103
+ "loss": 0.1211,
1104
+ "step": 1780
1105
+ },
1106
+ {
1107
+ "epoch": 0.26,
1108
+ "learning_rate": 7.629167426083665e-06,
1109
+ "loss": 0.0726,
1110
+ "step": 1790
1111
+ },
1112
+ {
1113
+ "epoch": 0.26,
1114
+ "learning_rate": 7.614649132284948e-06,
1115
+ "loss": 0.0481,
1116
+ "step": 1800
1117
+ },
1118
+ {
1119
+ "epoch": 0.26,
1120
+ "learning_rate": 7.600068335542579e-06,
1121
+ "loss": 0.1569,
1122
+ "step": 1810
1123
+ },
1124
+ {
1125
+ "epoch": 0.26,
1126
+ "learning_rate": 7.585425328455395e-06,
1127
+ "loss": 0.046,
1128
+ "step": 1820
1129
+ },
1130
+ {
1131
+ "epoch": 0.26,
1132
+ "learning_rate": 7.5707204048706365e-06,
1133
+ "loss": 0.0909,
1134
+ "step": 1830
1135
+ },
1136
+ {
1137
+ "epoch": 0.26,
1138
+ "learning_rate": 7.5559538598780434e-06,
1139
+ "loss": 0.1273,
1140
+ "step": 1840
1141
+ },
1142
+ {
1143
+ "epoch": 0.26,
1144
+ "learning_rate": 7.541125989803943e-06,
1145
+ "loss": 0.0636,
1146
+ "step": 1850
1147
+ },
1148
+ {
1149
+ "epoch": 0.27,
1150
+ "learning_rate": 7.526237092205292e-06,
1151
+ "loss": 0.1124,
1152
+ "step": 1860
1153
+ },
1154
+ {
1155
+ "epoch": 0.27,
1156
+ "learning_rate": 7.511287465863719e-06,
1157
+ "loss": 0.1033,
1158
+ "step": 1870
1159
+ },
1160
+ {
1161
+ "epoch": 0.27,
1162
+ "learning_rate": 7.496277410779513e-06,
1163
+ "loss": 0.1055,
1164
+ "step": 1880
1165
+ },
1166
+ {
1167
+ "epoch": 0.27,
1168
+ "learning_rate": 7.481207228165619e-06,
1169
+ "loss": 0.0778,
1170
+ "step": 1890
1171
+ },
1172
+ {
1173
+ "epoch": 0.27,
1174
+ "learning_rate": 7.466077220441581e-06,
1175
+ "loss": 0.0965,
1176
+ "step": 1900
1177
+ },
1178
+ {
1179
+ "epoch": 0.27,
1180
+ "learning_rate": 7.45088769122748e-06,
1181
+ "loss": 0.0561,
1182
+ "step": 1910
1183
+ },
1184
+ {
1185
+ "epoch": 0.27,
1186
+ "learning_rate": 7.435638945337836e-06,
1187
+ "loss": 0.1318,
1188
+ "step": 1920
1189
+ },
1190
+ {
1191
+ "epoch": 0.28,
1192
+ "learning_rate": 7.420331288775502e-06,
1193
+ "loss": 0.0561,
1194
+ "step": 1930
1195
+ },
1196
+ {
1197
+ "epoch": 0.28,
1198
+ "learning_rate": 7.404965028725507e-06,
1199
+ "loss": 0.0997,
1200
+ "step": 1940
1201
+ },
1202
+ {
1203
+ "epoch": 0.28,
1204
+ "learning_rate": 7.3895404735489035e-06,
1205
+ "loss": 0.135,
1206
+ "step": 1950
1207
+ },
1208
+ {
1209
+ "epoch": 0.28,
1210
+ "learning_rate": 7.374057932776579e-06,
1211
+ "loss": 0.0591,
1212
+ "step": 1960
1213
+ },
1214
+ {
1215
+ "epoch": 0.28,
1216
+ "learning_rate": 7.358517717103039e-06,
1217
+ "loss": 0.0574,
1218
+ "step": 1970
1219
+ },
1220
+ {
1221
+ "epoch": 0.28,
1222
+ "learning_rate": 7.342920138380175e-06,
1223
+ "loss": 0.0589,
1224
+ "step": 1980
1225
+ },
1226
+ {
1227
+ "epoch": 0.28,
1228
+ "learning_rate": 7.327265509611007e-06,
1229
+ "loss": 0.0685,
1230
+ "step": 1990
1231
+ },
1232
+ {
1233
+ "epoch": 0.29,
1234
+ "learning_rate": 7.311554144943401e-06,
1235
+ "loss": 0.0386,
1236
+ "step": 2000
1237
+ },
1238
+ {
1239
+ "epoch": 0.29,
1240
+ "eval_accuracy": 0.0,
1241
+ "eval_loss": 0.14258363842964172,
1242
+ "eval_runtime": 447.0432,
1243
+ "eval_samples_per_second": 52.297,
1244
+ "eval_steps_per_second": 8.717,
1245
+ "step": 2000
1246
+ },
1247
+ {
1248
+ "epoch": 0.29,
1249
+ "learning_rate": 7.295786359663766e-06,
1250
+ "loss": 0.0679,
1251
+ "step": 2010
1252
+ },
1253
+ {
1254
+ "epoch": 0.29,
1255
+ "learning_rate": 7.2799624701907275e-06,
1256
+ "loss": 0.1962,
1257
+ "step": 2020
1258
+ },
1259
+ {
1260
+ "epoch": 0.29,
1261
+ "learning_rate": 7.264082794068776e-06,
1262
+ "loss": 0.0871,
1263
+ "step": 2030
1264
+ },
1265
+ {
1266
+ "epoch": 0.29,
1267
+ "learning_rate": 7.248147649961896e-06,
1268
+ "loss": 0.1062,
1269
+ "step": 2040
1270
+ },
1271
+ {
1272
+ "epoch": 0.29,
1273
+ "learning_rate": 7.232157357647171e-06,
1274
+ "loss": 0.0415,
1275
+ "step": 2050
1276
+ },
1277
+ {
1278
+ "epoch": 0.29,
1279
+ "learning_rate": 7.216112238008363e-06,
1280
+ "loss": 0.1604,
1281
+ "step": 2060
1282
+ },
1283
+ {
1284
+ "epoch": 0.3,
1285
+ "learning_rate": 7.200012613029481e-06,
1286
+ "loss": 0.0401,
1287
+ "step": 2070
1288
+ },
1289
+ {
1290
+ "epoch": 0.3,
1291
+ "learning_rate": 7.1838588057883136e-06,
1292
+ "loss": 0.0683,
1293
+ "step": 2080
1294
+ },
1295
+ {
1296
+ "epoch": 0.3,
1297
+ "learning_rate": 7.167651140449947e-06,
1298
+ "loss": 0.0891,
1299
+ "step": 2090
1300
+ },
1301
+ {
1302
+ "epoch": 0.3,
1303
+ "learning_rate": 7.1513899422602585e-06,
1304
+ "loss": 0.1354,
1305
+ "step": 2100
1306
+ },
1307
+ {
1308
+ "epoch": 0.3,
1309
+ "learning_rate": 7.135075537539394e-06,
1310
+ "loss": 0.0904,
1311
+ "step": 2110
1312
+ },
1313
+ {
1314
+ "epoch": 0.3,
1315
+ "learning_rate": 7.118708253675213e-06,
1316
+ "loss": 0.1138,
1317
+ "step": 2120
1318
+ },
1319
+ {
1320
+ "epoch": 0.3,
1321
+ "learning_rate": 7.1022884191167256e-06,
1322
+ "loss": 0.0917,
1323
+ "step": 2130
1324
+ },
1325
+ {
1326
+ "epoch": 0.31,
1327
+ "learning_rate": 7.085816363367499e-06,
1328
+ "loss": 0.0465,
1329
+ "step": 2140
1330
+ },
1331
+ {
1332
+ "epoch": 0.31,
1333
+ "learning_rate": 7.0692924169790395e-06,
1334
+ "loss": 0.0826,
1335
+ "step": 2150
1336
+ },
1337
+ {
1338
+ "epoch": 0.31,
1339
+ "learning_rate": 7.05271691154417e-06,
1340
+ "loss": 0.0623,
1341
+ "step": 2160
1342
+ },
1343
+ {
1344
+ "epoch": 0.31,
1345
+ "learning_rate": 7.036090179690365e-06,
1346
+ "loss": 0.1098,
1347
+ "step": 2170
1348
+ },
1349
+ {
1350
+ "epoch": 0.31,
1351
+ "learning_rate": 7.019412555073087e-06,
1352
+ "loss": 0.0722,
1353
+ "step": 2180
1354
+ },
1355
+ {
1356
+ "epoch": 0.31,
1357
+ "learning_rate": 7.0026843723690755e-06,
1358
+ "loss": 0.1178,
1359
+ "step": 2190
1360
+ },
1361
+ {
1362
+ "epoch": 0.31,
1363
+ "learning_rate": 6.9859059672696475e-06,
1364
+ "loss": 0.0345,
1365
+ "step": 2200
1366
+ },
1367
+ {
1368
+ "epoch": 0.32,
1369
+ "learning_rate": 6.969077676473951e-06,
1370
+ "loss": 0.0445,
1371
+ "step": 2210
1372
+ },
1373
+ {
1374
+ "epoch": 0.32,
1375
+ "learning_rate": 6.95219983768221e-06,
1376
+ "loss": 0.1169,
1377
+ "step": 2220
1378
+ },
1379
+ {
1380
+ "epoch": 0.32,
1381
+ "learning_rate": 6.9352727895889486e-06,
1382
+ "loss": 0.0563,
1383
+ "step": 2230
1384
+ },
1385
+ {
1386
+ "epoch": 0.32,
1387
+ "learning_rate": 6.918296871876196e-06,
1388
+ "loss": 0.0492,
1389
+ "step": 2240
1390
+ },
1391
+ {
1392
+ "epoch": 0.32,
1393
+ "learning_rate": 6.901272425206663e-06,
1394
+ "loss": 0.1094,
1395
+ "step": 2250
1396
+ },
1397
+ {
1398
+ "epoch": 0.32,
1399
+ "learning_rate": 6.884199791216922e-06,
1400
+ "loss": 0.0515,
1401
+ "step": 2260
1402
+ },
1403
+ {
1404
+ "epoch": 0.32,
1405
+ "learning_rate": 6.867079312510525e-06,
1406
+ "loss": 0.1382,
1407
+ "step": 2270
1408
+ },
1409
+ {
1410
+ "epoch": 0.33,
1411
+ "learning_rate": 6.849911332651154e-06,
1412
+ "loss": 0.1041,
1413
+ "step": 2280
1414
+ },
1415
+ {
1416
+ "epoch": 0.33,
1417
+ "learning_rate": 6.832696196155716e-06,
1418
+ "loss": 0.0913,
1419
+ "step": 2290
1420
+ },
1421
+ {
1422
+ "epoch": 0.33,
1423
+ "learning_rate": 6.815434248487425e-06,
1424
+ "loss": 0.0794,
1425
+ "step": 2300
1426
+ },
1427
+ {
1428
+ "epoch": 0.33,
1429
+ "learning_rate": 6.798125836048876e-06,
1430
+ "loss": 0.0864,
1431
+ "step": 2310
1432
+ },
1433
+ {
1434
+ "epoch": 0.33,
1435
+ "learning_rate": 6.780771306175093e-06,
1436
+ "loss": 0.1276,
1437
+ "step": 2320
1438
+ },
1439
+ {
1440
+ "epoch": 0.33,
1441
+ "learning_rate": 6.763371007126555e-06,
1442
+ "loss": 0.1068,
1443
+ "step": 2330
1444
+ },
1445
+ {
1446
+ "epoch": 0.33,
1447
+ "learning_rate": 6.745925288082212e-06,
1448
+ "loss": 0.1093,
1449
+ "step": 2340
1450
+ },
1451
+ {
1452
+ "epoch": 0.34,
1453
+ "learning_rate": 6.728434499132473e-06,
1454
+ "loss": 0.099,
1455
+ "step": 2350
1456
+ },
1457
+ {
1458
+ "epoch": 0.34,
1459
+ "learning_rate": 6.710898991272189e-06,
1460
+ "loss": 0.0841,
1461
+ "step": 2360
1462
+ },
1463
+ {
1464
+ "epoch": 0.34,
1465
+ "learning_rate": 6.693319116393596e-06,
1466
+ "loss": 0.0863,
1467
+ "step": 2370
1468
+ },
1469
+ {
1470
+ "epoch": 0.34,
1471
+ "learning_rate": 6.6756952272792685e-06,
1472
+ "loss": 0.0726,
1473
+ "step": 2380
1474
+ },
1475
+ {
1476
+ "epoch": 0.34,
1477
+ "learning_rate": 6.658027677595027e-06,
1478
+ "loss": 0.1138,
1479
+ "step": 2390
1480
+ },
1481
+ {
1482
+ "epoch": 0.34,
1483
+ "learning_rate": 6.6403168218828494e-06,
1484
+ "loss": 0.102,
1485
+ "step": 2400
1486
+ },
1487
+ {
1488
+ "epoch": 0.34,
1489
+ "learning_rate": 6.622563015553752e-06,
1490
+ "loss": 0.085,
1491
+ "step": 2410
1492
+ },
1493
+ {
1494
+ "epoch": 0.35,
1495
+ "learning_rate": 6.604766614880662e-06,
1496
+ "loss": 0.0646,
1497
+ "step": 2420
1498
+ },
1499
+ {
1500
+ "epoch": 0.35,
1501
+ "learning_rate": 6.586927976991262e-06,
1502
+ "loss": 0.0487,
1503
+ "step": 2430
1504
+ },
1505
+ {
1506
+ "epoch": 0.35,
1507
+ "learning_rate": 6.569047459860825e-06,
1508
+ "loss": 0.0929,
1509
+ "step": 2440
1510
+ },
1511
+ {
1512
+ "epoch": 0.35,
1513
+ "learning_rate": 6.551125422305035e-06,
1514
+ "loss": 0.097,
1515
+ "step": 2450
1516
+ },
1517
+ {
1518
+ "epoch": 0.35,
1519
+ "learning_rate": 6.533162223972784e-06,
1520
+ "loss": 0.0976,
1521
+ "step": 2460
1522
+ },
1523
+ {
1524
+ "epoch": 0.35,
1525
+ "learning_rate": 6.515158225338948e-06,
1526
+ "loss": 0.1232,
1527
+ "step": 2470
1528
+ },
1529
+ {
1530
+ "epoch": 0.35,
1531
+ "learning_rate": 6.497113787697166e-06,
1532
+ "loss": 0.0646,
1533
+ "step": 2480
1534
+ },
1535
+ {
1536
+ "epoch": 0.36,
1537
+ "learning_rate": 6.479029273152582e-06,
1538
+ "loss": 0.0285,
1539
+ "step": 2490
1540
+ },
1541
+ {
1542
+ "epoch": 0.36,
1543
+ "learning_rate": 6.460905044614577e-06,
1544
+ "loss": 0.0716,
1545
+ "step": 2500
1546
+ },
1547
+ {
1548
+ "epoch": 0.36,
1549
+ "eval_accuracy": 0.0,
1550
+ "eval_loss": 0.11192344129085541,
1551
+ "eval_runtime": 452.1105,
1552
+ "eval_samples_per_second": 51.711,
1553
+ "eval_steps_per_second": 8.62,
1554
+ "step": 2500
1555
+ },
1556
+ {
1557
+ "epoch": 0.36,
1558
+ "learning_rate": 6.442741465789492e-06,
1559
+ "loss": 0.0429,
1560
+ "step": 2510
1561
+ },
1562
+ {
1563
+ "epoch": 0.36,
1564
+ "learning_rate": 6.424538901173323e-06,
1565
+ "loss": 0.0803,
1566
+ "step": 2520
1567
+ },
1568
+ {
1569
+ "epoch": 0.36,
1570
+ "learning_rate": 6.406297716044413e-06,
1571
+ "loss": 0.0858,
1572
+ "step": 2530
1573
+ },
1574
+ {
1575
+ "epoch": 0.36,
1576
+ "learning_rate": 6.388018276456118e-06,
1577
+ "loss": 0.0605,
1578
+ "step": 2540
1579
+ },
1580
+ {
1581
+ "epoch": 0.36,
1582
+ "learning_rate": 6.3697009492294605e-06,
1583
+ "loss": 0.0681,
1584
+ "step": 2550
1585
+ },
1586
+ {
1587
+ "epoch": 0.37,
1588
+ "learning_rate": 6.351346101945772e-06,
1589
+ "loss": 0.0607,
1590
+ "step": 2560
1591
+ },
1592
+ {
1593
+ "epoch": 0.37,
1594
+ "learning_rate": 6.332954102939312e-06,
1595
+ "loss": 0.1236,
1596
+ "step": 2570
1597
+ },
1598
+ {
1599
+ "epoch": 0.37,
1600
+ "learning_rate": 6.314525321289882e-06,
1601
+ "loss": 0.0361,
1602
+ "step": 2580
1603
+ },
1604
+ {
1605
+ "epoch": 0.37,
1606
+ "learning_rate": 6.296060126815411e-06,
1607
+ "loss": 0.0547,
1608
+ "step": 2590
1609
+ },
1610
+ {
1611
+ "epoch": 0.37,
1612
+ "learning_rate": 6.277558890064545e-06,
1613
+ "loss": 0.1043,
1614
+ "step": 2600
1615
+ },
1616
+ {
1617
+ "epoch": 0.37,
1618
+ "learning_rate": 6.259021982309201e-06,
1619
+ "loss": 0.084,
1620
+ "step": 2610
1621
+ },
1622
+ {
1623
+ "epoch": 0.37,
1624
+ "learning_rate": 6.240449775537121e-06,
1625
+ "loss": 0.058,
1626
+ "step": 2620
1627
+ },
1628
+ {
1629
+ "epoch": 0.37,
1630
+ "learning_rate": 6.221842642444411e-06,
1631
+ "loss": 0.0862,
1632
+ "step": 2630
1633
+ },
1634
+ {
1635
+ "epoch": 0.38,
1636
+ "learning_rate": 6.20320095642805e-06,
1637
+ "loss": 0.1117,
1638
+ "step": 2640
1639
+ },
1640
+ {
1641
+ "epoch": 0.38,
1642
+ "learning_rate": 6.184525091578414e-06,
1643
+ "loss": 0.1552,
1644
+ "step": 2650
1645
+ },
1646
+ {
1647
+ "epoch": 0.38,
1648
+ "learning_rate": 6.165815422671752e-06,
1649
+ "loss": 0.0642,
1650
+ "step": 2660
1651
+ },
1652
+ {
1653
+ "epoch": 0.38,
1654
+ "learning_rate": 6.147072325162679e-06,
1655
+ "loss": 0.048,
1656
+ "step": 2670
1657
+ },
1658
+ {
1659
+ "epoch": 0.38,
1660
+ "learning_rate": 6.128296175176628e-06,
1661
+ "loss": 0.0966,
1662
+ "step": 2680
1663
+ },
1664
+ {
1665
+ "epoch": 0.38,
1666
+ "learning_rate": 6.109487349502319e-06,
1667
+ "loss": 0.0853,
1668
+ "step": 2690
1669
+ },
1670
+ {
1671
+ "epoch": 0.38,
1672
+ "learning_rate": 6.090646225584179e-06,
1673
+ "loss": 0.0884,
1674
+ "step": 2700
1675
+ },
1676
+ {
1677
+ "epoch": 0.39,
1678
+ "learning_rate": 6.071773181514785e-06,
1679
+ "loss": 0.0887,
1680
+ "step": 2710
1681
+ },
1682
+ {
1683
+ "epoch": 0.39,
1684
+ "learning_rate": 6.052868596027264e-06,
1685
+ "loss": 0.0736,
1686
+ "step": 2720
1687
+ },
1688
+ {
1689
+ "epoch": 0.39,
1690
+ "learning_rate": 6.033932848487701e-06,
1691
+ "loss": 0.1141,
1692
+ "step": 2730
1693
+ },
1694
+ {
1695
+ "epoch": 0.39,
1696
+ "learning_rate": 6.014966318887518e-06,
1697
+ "loss": 0.0481,
1698
+ "step": 2740
1699
+ },
1700
+ {
1701
+ "epoch": 0.39,
1702
+ "learning_rate": 5.995969387835857e-06,
1703
+ "loss": 0.0891,
1704
+ "step": 2750
1705
+ },
1706
+ {
1707
+ "epoch": 0.39,
1708
+ "learning_rate": 5.976942436551939e-06,
1709
+ "loss": 0.0881,
1710
+ "step": 2760
1711
+ },
1712
+ {
1713
+ "epoch": 0.39,
1714
+ "learning_rate": 5.95788584685741e-06,
1715
+ "loss": 0.0967,
1716
+ "step": 2770
1717
+ },
1718
+ {
1719
+ "epoch": 0.4,
1720
+ "learning_rate": 5.938800001168685e-06,
1721
+ "loss": 0.0753,
1722
+ "step": 2780
1723
+ },
1724
+ {
1725
+ "epoch": 0.4,
1726
+ "learning_rate": 5.919685282489269e-06,
1727
+ "loss": 0.0671,
1728
+ "step": 2790
1729
+ },
1730
+ {
1731
+ "epoch": 0.4,
1732
+ "learning_rate": 5.900542074402076e-06,
1733
+ "loss": 0.1066,
1734
+ "step": 2800
1735
+ },
1736
+ {
1737
+ "epoch": 0.4,
1738
+ "learning_rate": 5.881370761061724e-06,
1739
+ "loss": 0.0888,
1740
+ "step": 2810
1741
+ },
1742
+ {
1743
+ "epoch": 0.4,
1744
+ "learning_rate": 5.862171727186837e-06,
1745
+ "loss": 0.0783,
1746
+ "step": 2820
1747
+ },
1748
+ {
1749
+ "epoch": 0.4,
1750
+ "learning_rate": 5.842945358052311e-06,
1751
+ "loss": 0.0945,
1752
+ "step": 2830
1753
+ },
1754
+ {
1755
+ "epoch": 0.4,
1756
+ "learning_rate": 5.823692039481596e-06,
1757
+ "loss": 0.103,
1758
+ "step": 2840
1759
+ },
1760
+ {
1761
+ "epoch": 0.41,
1762
+ "learning_rate": 5.804412157838945e-06,
1763
+ "loss": 0.0861,
1764
+ "step": 2850
1765
+ },
1766
+ {
1767
+ "epoch": 0.41,
1768
+ "learning_rate": 5.785106100021663e-06,
1769
+ "loss": 0.0627,
1770
+ "step": 2860
1771
+ },
1772
+ {
1773
+ "epoch": 0.41,
1774
+ "learning_rate": 5.765774253452345e-06,
1775
+ "loss": 0.0828,
1776
+ "step": 2870
1777
+ },
1778
+ {
1779
+ "epoch": 0.41,
1780
+ "learning_rate": 5.746417006071094e-06,
1781
+ "loss": 0.0629,
1782
+ "step": 2880
1783
+ },
1784
+ {
1785
+ "epoch": 0.41,
1786
+ "learning_rate": 5.727034746327746e-06,
1787
+ "loss": 0.1149,
1788
+ "step": 2890
1789
+ },
1790
+ {
1791
+ "epoch": 0.41,
1792
+ "learning_rate": 5.707627863174073e-06,
1793
+ "loss": 0.0865,
1794
+ "step": 2900
1795
+ },
1796
+ {
1797
+ "epoch": 0.41,
1798
+ "learning_rate": 5.688196746055968e-06,
1799
+ "loss": 0.0598,
1800
+ "step": 2910
1801
+ },
1802
+ {
1803
+ "epoch": 0.42,
1804
+ "learning_rate": 5.668741784905642e-06,
1805
+ "loss": 0.0564,
1806
+ "step": 2920
1807
+ },
1808
+ {
1809
+ "epoch": 0.42,
1810
+ "learning_rate": 5.649263370133794e-06,
1811
+ "loss": 0.0993,
1812
+ "step": 2930
1813
+ },
1814
+ {
1815
+ "epoch": 0.42,
1816
+ "learning_rate": 5.629761892621775e-06,
1817
+ "loss": 0.0692,
1818
+ "step": 2940
1819
+ },
1820
+ {
1821
+ "epoch": 0.42,
1822
+ "learning_rate": 5.610237743713742e-06,
1823
+ "loss": 0.0787,
1824
+ "step": 2950
1825
+ },
1826
+ {
1827
+ "epoch": 0.42,
1828
+ "learning_rate": 5.590691315208815e-06,
1829
+ "loss": 0.045,
1830
+ "step": 2960
1831
+ },
1832
+ {
1833
+ "epoch": 0.42,
1834
+ "learning_rate": 5.571122999353203e-06,
1835
+ "loss": 0.0221,
1836
+ "step": 2970
1837
+ },
1838
+ {
1839
+ "epoch": 0.42,
1840
+ "learning_rate": 5.551533188832339e-06,
1841
+ "loss": 0.065,
1842
+ "step": 2980
1843
+ },
1844
+ {
1845
+ "epoch": 0.43,
1846
+ "learning_rate": 5.531922276763001e-06,
1847
+ "loss": 0.0351,
1848
+ "step": 2990
1849
+ },
1850
+ {
1851
+ "epoch": 0.43,
1852
+ "learning_rate": 5.512290656685415e-06,
1853
+ "loss": 0.0655,
1854
+ "step": 3000
1855
+ },
1856
+ {
1857
+ "epoch": 0.43,
1858
+ "eval_accuracy": 0.0,
1859
+ "eval_loss": 0.12717388570308685,
1860
+ "eval_runtime": 447.1731,
1861
+ "eval_samples_per_second": 52.282,
1862
+ "eval_steps_per_second": 8.715,
1863
+ "step": 3000
1864
+ },
1865
+ {
1866
+ "epoch": 0.43,
1867
+ "learning_rate": 5.4926387225553675e-06,
1868
+ "loss": 0.0666,
1869
+ "step": 3010
1870
+ },
1871
+ {
1872
+ "epoch": 0.43,
1873
+ "learning_rate": 5.472966868736292e-06,
1874
+ "loss": 0.0775,
1875
+ "step": 3020
1876
+ },
1877
+ {
1878
+ "epoch": 0.43,
1879
+ "learning_rate": 5.453275489991362e-06,
1880
+ "loss": 0.0906,
1881
+ "step": 3030
1882
+ },
1883
+ {
1884
+ "epoch": 0.43,
1885
+ "learning_rate": 5.433564981475563e-06,
1886
+ "loss": 0.0459,
1887
+ "step": 3040
1888
+ },
1889
+ {
1890
+ "epoch": 0.43,
1891
+ "learning_rate": 5.413835738727768e-06,
1892
+ "loss": 0.1028,
1893
+ "step": 3050
1894
+ },
1895
+ {
1896
+ "epoch": 0.44,
1897
+ "learning_rate": 5.394088157662793e-06,
1898
+ "loss": 0.1357,
1899
+ "step": 3060
1900
+ },
1901
+ {
1902
+ "epoch": 0.44,
1903
+ "learning_rate": 5.374322634563464e-06,
1904
+ "loss": 0.0751,
1905
+ "step": 3070
1906
+ },
1907
+ {
1908
+ "epoch": 0.44,
1909
+ "learning_rate": 5.354539566072651e-06,
1910
+ "loss": 0.1197,
1911
+ "step": 3080
1912
+ },
1913
+ {
1914
+ "epoch": 0.44,
1915
+ "learning_rate": 5.334739349185315e-06,
1916
+ "loss": 0.1061,
1917
+ "step": 3090
1918
+ },
1919
+ {
1920
+ "epoch": 0.44,
1921
+ "learning_rate": 5.314922381240544e-06,
1922
+ "loss": 0.0739,
1923
+ "step": 3100
1924
+ },
1925
+ {
1926
+ "epoch": 0.44,
1927
+ "learning_rate": 5.295089059913573e-06,
1928
+ "loss": 0.0794,
1929
+ "step": 3110
1930
+ },
1931
+ {
1932
+ "epoch": 0.44,
1933
+ "learning_rate": 5.275239783207811e-06,
1934
+ "loss": 0.0515,
1935
+ "step": 3120
1936
+ },
1937
+ {
1938
+ "epoch": 0.45,
1939
+ "learning_rate": 5.255374949446845e-06,
1940
+ "loss": 0.0413,
1941
+ "step": 3130
1942
+ },
1943
+ {
1944
+ "epoch": 0.45,
1945
+ "learning_rate": 5.235494957266457e-06,
1946
+ "loss": 0.0763,
1947
+ "step": 3140
1948
+ },
1949
+ {
1950
+ "epoch": 0.45,
1951
+ "learning_rate": 5.215600205606617e-06,
1952
+ "loss": 0.0887,
1953
+ "step": 3150
1954
+ },
1955
+ {
1956
+ "epoch": 0.45,
1957
+ "learning_rate": 5.1956910937034785e-06,
1958
+ "loss": 0.0472,
1959
+ "step": 3160
1960
+ },
1961
+ {
1962
+ "epoch": 0.45,
1963
+ "learning_rate": 5.175768021081368e-06,
1964
+ "loss": 0.0852,
1965
+ "step": 3170
1966
+ },
1967
+ {
1968
+ "epoch": 0.45,
1969
+ "learning_rate": 5.15583138754477e-06,
1970
+ "loss": 0.0981,
1971
+ "step": 3180
1972
+ },
1973
+ {
1974
+ "epoch": 0.45,
1975
+ "learning_rate": 5.135881593170299e-06,
1976
+ "loss": 0.0239,
1977
+ "step": 3190
1978
+ },
1979
+ {
1980
+ "epoch": 0.46,
1981
+ "learning_rate": 5.1159190382986725e-06,
1982
+ "loss": 0.0749,
1983
+ "step": 3200
1984
+ },
1985
+ {
1986
+ "epoch": 0.46,
1987
+ "learning_rate": 5.095944123526683e-06,
1988
+ "loss": 0.0929,
1989
+ "step": 3210
1990
+ },
1991
+ {
1992
+ "epoch": 0.46,
1993
+ "learning_rate": 5.075957249699149e-06,
1994
+ "loss": 0.0315,
1995
+ "step": 3220
1996
+ },
1997
+ {
1998
+ "epoch": 0.46,
1999
+ "learning_rate": 5.05595881790088e-06,
2000
+ "loss": 0.0884,
2001
+ "step": 3230
2002
+ },
2003
+ {
2004
+ "epoch": 0.46,
2005
+ "learning_rate": 5.03594922944862e-06,
2006
+ "loss": 0.0381,
2007
+ "step": 3240
2008
+ },
2009
+ {
2010
+ "epoch": 0.46,
2011
+ "learning_rate": 5.015928885883002e-06,
2012
+ "loss": 0.0473,
2013
+ "step": 3250
2014
+ },
2015
+ {
2016
+ "epoch": 0.46,
2017
+ "learning_rate": 4.995898188960485e-06,
2018
+ "loss": 0.0673,
2019
+ "step": 3260
2020
+ },
2021
+ {
2022
+ "epoch": 0.47,
2023
+ "learning_rate": 4.975857540645292e-06,
2024
+ "loss": 0.061,
2025
+ "step": 3270
2026
+ },
2027
+ {
2028
+ "epoch": 0.47,
2029
+ "learning_rate": 4.955807343101347e-06,
2030
+ "loss": 0.1168,
2031
+ "step": 3280
2032
+ },
2033
+ {
2034
+ "epoch": 0.47,
2035
+ "learning_rate": 4.935747998684199e-06,
2036
+ "loss": 0.0772,
2037
+ "step": 3290
2038
+ },
2039
+ {
2040
+ "epoch": 0.47,
2041
+ "learning_rate": 4.9156799099329535e-06,
2042
+ "loss": 0.0439,
2043
+ "step": 3300
2044
+ },
2045
+ {
2046
+ "epoch": 0.47,
2047
+ "learning_rate": 4.8956034795621925e-06,
2048
+ "loss": 0.0715,
2049
+ "step": 3310
2050
+ },
2051
+ {
2052
+ "epoch": 0.47,
2053
+ "learning_rate": 4.8755191104538905e-06,
2054
+ "loss": 0.0766,
2055
+ "step": 3320
2056
+ },
2057
+ {
2058
+ "epoch": 0.47,
2059
+ "learning_rate": 4.855427205649333e-06,
2060
+ "loss": 0.0304,
2061
+ "step": 3330
2062
+ },
2063
+ {
2064
+ "epoch": 0.48,
2065
+ "learning_rate": 4.835328168341029e-06,
2066
+ "loss": 0.0465,
2067
+ "step": 3340
2068
+ },
2069
+ {
2070
+ "epoch": 0.48,
2071
+ "learning_rate": 4.815222401864617e-06,
2072
+ "loss": 0.1278,
2073
+ "step": 3350
2074
+ },
2075
+ {
2076
+ "epoch": 0.48,
2077
+ "learning_rate": 4.795110309690772e-06,
2078
+ "loss": 0.0688,
2079
+ "step": 3360
2080
+ },
2081
+ {
2082
+ "epoch": 0.48,
2083
+ "learning_rate": 4.7749922954171075e-06,
2084
+ "loss": 0.0719,
2085
+ "step": 3370
2086
+ },
2087
+ {
2088
+ "epoch": 0.48,
2089
+ "learning_rate": 4.754868762760083e-06,
2090
+ "loss": 0.0588,
2091
+ "step": 3380
2092
+ },
2093
+ {
2094
+ "epoch": 0.48,
2095
+ "learning_rate": 4.734740115546896e-06,
2096
+ "loss": 0.0889,
2097
+ "step": 3390
2098
+ },
2099
+ {
2100
+ "epoch": 0.48,
2101
+ "learning_rate": 4.714606757707377e-06,
2102
+ "loss": 0.0488,
2103
+ "step": 3400
2104
+ },
2105
+ {
2106
+ "epoch": 0.49,
2107
+ "learning_rate": 4.69446909326589e-06,
2108
+ "loss": 0.0752,
2109
+ "step": 3410
2110
+ },
2111
+ {
2112
+ "epoch": 0.49,
2113
+ "learning_rate": 4.674327526333218e-06,
2114
+ "loss": 0.0838,
2115
+ "step": 3420
2116
+ },
2117
+ {
2118
+ "epoch": 0.49,
2119
+ "learning_rate": 4.654182461098462e-06,
2120
+ "loss": 0.0794,
2121
+ "step": 3430
2122
+ },
2123
+ {
2124
+ "epoch": 0.49,
2125
+ "learning_rate": 4.63403430182092e-06,
2126
+ "loss": 0.0691,
2127
+ "step": 3440
2128
+ },
2129
+ {
2130
+ "epoch": 0.49,
2131
+ "learning_rate": 4.61388345282198e-06,
2132
+ "loss": 0.0814,
2133
+ "step": 3450
2134
+ },
2135
+ {
2136
+ "epoch": 0.49,
2137
+ "learning_rate": 4.59373031847701e-06,
2138
+ "loss": 0.04,
2139
+ "step": 3460
2140
+ },
2141
+ {
2142
+ "epoch": 0.49,
2143
+ "learning_rate": 4.573575303207232e-06,
2144
+ "loss": 0.1552,
2145
+ "step": 3470
2146
+ },
2147
+ {
2148
+ "epoch": 0.5,
2149
+ "learning_rate": 4.5534188114716195e-06,
2150
+ "loss": 0.0742,
2151
+ "step": 3480
2152
+ },
2153
+ {
2154
+ "epoch": 0.5,
2155
+ "learning_rate": 4.533261247758775e-06,
2156
+ "loss": 0.0525,
2157
+ "step": 3490
2158
+ },
2159
+ {
2160
+ "epoch": 0.5,
2161
+ "learning_rate": 4.5131030165788055e-06,
2162
+ "loss": 0.0725,
2163
+ "step": 3500
2164
+ },
2165
+ {
2166
+ "epoch": 0.5,
2167
+ "eval_accuracy": 0.0,
2168
+ "eval_loss": 0.09580323845148087,
2169
+ "eval_runtime": 449.6777,
2170
+ "eval_samples_per_second": 51.991,
2171
+ "eval_steps_per_second": 8.666,
2172
+ "step": 3500
2173
+ },
2174
+ {
2175
+ "epoch": 0.5,
2176
+ "learning_rate": 4.49294452245522e-06,
2177
+ "loss": 0.0529,
2178
+ "step": 3510
2179
+ },
2180
+ {
2181
+ "epoch": 0.5,
2182
+ "learning_rate": 4.4727861699168e-06,
2183
+ "loss": 0.053,
2184
+ "step": 3520
2185
+ },
2186
+ {
2187
+ "epoch": 0.5,
2188
+ "learning_rate": 4.45262836348949e-06,
2189
+ "loss": 0.0941,
2190
+ "step": 3530
2191
+ },
2192
+ {
2193
+ "epoch": 0.5,
2194
+ "learning_rate": 4.4324715076882714e-06,
2195
+ "loss": 0.0586,
2196
+ "step": 3540
2197
+ },
2198
+ {
2199
+ "epoch": 0.51,
2200
+ "learning_rate": 4.412316007009046e-06,
2201
+ "loss": 0.0536,
2202
+ "step": 3550
2203
+ },
2204
+ {
2205
+ "epoch": 0.51,
2206
+ "learning_rate": 4.392162265920531e-06,
2207
+ "loss": 0.0935,
2208
+ "step": 3560
2209
+ },
2210
+ {
2211
+ "epoch": 0.51,
2212
+ "learning_rate": 4.372010688856124e-06,
2213
+ "loss": 0.0655,
2214
+ "step": 3570
2215
+ },
2216
+ {
2217
+ "epoch": 0.51,
2218
+ "learning_rate": 4.351861680205804e-06,
2219
+ "loss": 0.085,
2220
+ "step": 3580
2221
+ },
2222
+ {
2223
+ "epoch": 0.51,
2224
+ "learning_rate": 4.331715644308002e-06,
2225
+ "loss": 0.0822,
2226
+ "step": 3590
2227
+ },
2228
+ {
2229
+ "epoch": 0.51,
2230
+ "learning_rate": 4.3115729854415e-06,
2231
+ "loss": 0.0196,
2232
+ "step": 3600
2233
+ },
2234
+ {
2235
+ "epoch": 0.51,
2236
+ "learning_rate": 4.291434107817306e-06,
2237
+ "loss": 0.0476,
2238
+ "step": 3610
2239
+ },
2240
+ {
2241
+ "epoch": 0.52,
2242
+ "learning_rate": 4.271299415570548e-06,
2243
+ "loss": 0.0322,
2244
+ "step": 3620
2245
+ },
2246
+ {
2247
+ "epoch": 0.52,
2248
+ "learning_rate": 4.251169312752372e-06,
2249
+ "loss": 0.1195,
2250
+ "step": 3630
2251
+ },
2252
+ {
2253
+ "epoch": 0.52,
2254
+ "learning_rate": 4.231044203321815e-06,
2255
+ "loss": 0.0776,
2256
+ "step": 3640
2257
+ },
2258
+ {
2259
+ "epoch": 0.52,
2260
+ "learning_rate": 4.2109244911377206e-06,
2261
+ "loss": 0.0563,
2262
+ "step": 3650
2263
+ },
2264
+ {
2265
+ "epoch": 0.52,
2266
+ "learning_rate": 4.190810579950615e-06,
2267
+ "loss": 0.1301,
2268
+ "step": 3660
2269
+ },
2270
+ {
2271
+ "epoch": 0.52,
2272
+ "learning_rate": 4.17070287339462e-06,
2273
+ "loss": 0.0751,
2274
+ "step": 3670
2275
+ },
2276
+ {
2277
+ "epoch": 0.52,
2278
+ "learning_rate": 4.150601774979341e-06,
2279
+ "loss": 0.0594,
2280
+ "step": 3680
2281
+ },
2282
+ {
2283
+ "epoch": 0.53,
2284
+ "learning_rate": 4.130507688081777e-06,
2285
+ "loss": 0.0534,
2286
+ "step": 3690
2287
+ },
2288
+ {
2289
+ "epoch": 0.53,
2290
+ "learning_rate": 4.110421015938227e-06,
2291
+ "loss": 0.0672,
2292
+ "step": 3700
2293
+ },
2294
+ {
2295
+ "epoch": 0.53,
2296
+ "learning_rate": 4.090342161636191e-06,
2297
+ "loss": 0.0722,
2298
+ "step": 3710
2299
+ },
2300
+ {
2301
+ "epoch": 0.53,
2302
+ "learning_rate": 4.070271528106287e-06,
2303
+ "loss": 0.0633,
2304
+ "step": 3720
2305
+ },
2306
+ {
2307
+ "epoch": 0.53,
2308
+ "learning_rate": 4.050209518114164e-06,
2309
+ "loss": 0.1139,
2310
+ "step": 3730
2311
+ },
2312
+ {
2313
+ "epoch": 0.53,
2314
+ "learning_rate": 4.030156534252416e-06,
2315
+ "loss": 0.0358,
2316
+ "step": 3740
2317
+ },
2318
+ {
2319
+ "epoch": 0.53,
2320
+ "learning_rate": 4.010112978932511e-06,
2321
+ "loss": 0.0972,
2322
+ "step": 3750
2323
+ },
2324
+ {
2325
+ "epoch": 0.54,
2326
+ "learning_rate": 3.990079254376704e-06,
2327
+ "loss": 0.0397,
2328
+ "step": 3760
2329
+ },
2330
+ {
2331
+ "epoch": 0.54,
2332
+ "learning_rate": 3.97005576260998e-06,
2333
+ "loss": 0.0627,
2334
+ "step": 3770
2335
+ },
2336
+ {
2337
+ "epoch": 0.54,
2338
+ "learning_rate": 3.9500429054519696e-06,
2339
+ "loss": 0.0617,
2340
+ "step": 3780
2341
+ },
2342
+ {
2343
+ "epoch": 0.54,
2344
+ "learning_rate": 3.930041084508903e-06,
2345
+ "loss": 0.0681,
2346
+ "step": 3790
2347
+ },
2348
+ {
2349
+ "epoch": 0.54,
2350
+ "learning_rate": 3.910050701165534e-06,
2351
+ "loss": 0.0434,
2352
+ "step": 3800
2353
+ },
2354
+ {
2355
+ "epoch": 0.54,
2356
+ "learning_rate": 3.890072156577101e-06,
2357
+ "loss": 0.0437,
2358
+ "step": 3810
2359
+ },
2360
+ {
2361
+ "epoch": 0.54,
2362
+ "learning_rate": 3.870105851661263e-06,
2363
+ "loss": 0.0673,
2364
+ "step": 3820
2365
+ },
2366
+ {
2367
+ "epoch": 0.55,
2368
+ "learning_rate": 3.850152187090063e-06,
2369
+ "loss": 0.0627,
2370
+ "step": 3830
2371
+ },
2372
+ {
2373
+ "epoch": 0.55,
2374
+ "learning_rate": 3.8302115632818865e-06,
2375
+ "loss": 0.0969,
2376
+ "step": 3840
2377
+ },
2378
+ {
2379
+ "epoch": 0.55,
2380
+ "learning_rate": 3.810284380393421e-06,
2381
+ "loss": 0.0379,
2382
+ "step": 3850
2383
+ },
2384
+ {
2385
+ "epoch": 0.55,
2386
+ "learning_rate": 3.790371038311635e-06,
2387
+ "loss": 0.086,
2388
+ "step": 3860
2389
+ },
2390
+ {
2391
+ "epoch": 0.55,
2392
+ "learning_rate": 3.7704719366457428e-06,
2393
+ "loss": 0.0953,
2394
+ "step": 3870
2395
+ },
2396
+ {
2397
+ "epoch": 0.55,
2398
+ "learning_rate": 3.75058747471919e-06,
2399
+ "loss": 0.027,
2400
+ "step": 3880
2401
+ },
2402
+ {
2403
+ "epoch": 0.55,
2404
+ "learning_rate": 3.730718051561647e-06,
2405
+ "loss": 0.0792,
2406
+ "step": 3890
2407
+ },
2408
+ {
2409
+ "epoch": 0.56,
2410
+ "learning_rate": 3.710864065900986e-06,
2411
+ "loss": 0.0777,
2412
+ "step": 3900
2413
+ },
2414
+ {
2415
+ "epoch": 0.56,
2416
+ "learning_rate": 3.6910259161552983e-06,
2417
+ "loss": 0.1068,
2418
+ "step": 3910
2419
+ },
2420
+ {
2421
+ "epoch": 0.56,
2422
+ "learning_rate": 3.67120400042488e-06,
2423
+ "loss": 0.0759,
2424
+ "step": 3920
2425
+ },
2426
+ {
2427
+ "epoch": 0.56,
2428
+ "learning_rate": 3.6513987164842605e-06,
2429
+ "loss": 0.0391,
2430
+ "step": 3930
2431
+ },
2432
+ {
2433
+ "epoch": 0.56,
2434
+ "learning_rate": 3.631610461774208e-06,
2435
+ "loss": 0.0546,
2436
+ "step": 3940
2437
+ },
2438
+ {
2439
+ "epoch": 0.56,
2440
+ "learning_rate": 3.6118396333937564e-06,
2441
+ "loss": 0.0817,
2442
+ "step": 3950
2443
+ },
2444
+ {
2445
+ "epoch": 0.56,
2446
+ "learning_rate": 3.5920866280922445e-06,
2447
+ "loss": 0.1293,
2448
+ "step": 3960
2449
+ },
2450
+ {
2451
+ "epoch": 0.57,
2452
+ "learning_rate": 3.572351842261341e-06,
2453
+ "loss": 0.0762,
2454
+ "step": 3970
2455
+ },
2456
+ {
2457
+ "epoch": 0.57,
2458
+ "learning_rate": 3.5526356719271034e-06,
2459
+ "loss": 0.0385,
2460
+ "step": 3980
2461
+ },
2462
+ {
2463
+ "epoch": 0.57,
2464
+ "learning_rate": 3.532938512742018e-06,
2465
+ "loss": 0.0444,
2466
+ "step": 3990
2467
+ },
2468
+ {
2469
+ "epoch": 0.57,
2470
+ "learning_rate": 3.5132607599770728e-06,
2471
+ "loss": 0.0759,
2472
+ "step": 4000
2473
+ },
2474
+ {
2475
+ "epoch": 0.57,
2476
+ "eval_accuracy": 0.0,
2477
+ "eval_loss": 0.09747572988271713,
2478
+ "eval_runtime": 449.16,
2479
+ "eval_samples_per_second": 52.05,
2480
+ "eval_steps_per_second": 8.676,
2481
+ "step": 4000
2482
+ },
2483
+ {
2484
+ "epoch": 0.57,
2485
+ "learning_rate": 3.493602808513816e-06,
2486
+ "loss": 0.0295,
2487
+ "step": 4010
2488
+ },
2489
+ {
2490
+ "epoch": 0.57,
2491
+ "learning_rate": 3.4739650528364323e-06,
2492
+ "loss": 0.0367,
2493
+ "step": 4020
2494
+ },
2495
+ {
2496
+ "epoch": 0.57,
2497
+ "learning_rate": 3.4543478870238354e-06,
2498
+ "loss": 0.047,
2499
+ "step": 4030
2500
+ },
2501
+ {
2502
+ "epoch": 0.58,
2503
+ "learning_rate": 3.4347517047417474e-06,
2504
+ "loss": 0.0243,
2505
+ "step": 4040
2506
+ },
2507
+ {
2508
+ "epoch": 0.58,
2509
+ "learning_rate": 3.4151768992348124e-06,
2510
+ "loss": 0.1458,
2511
+ "step": 4050
2512
+ },
2513
+ {
2514
+ "epoch": 0.58,
2515
+ "learning_rate": 3.3956238633186902e-06,
2516
+ "loss": 0.11,
2517
+ "step": 4060
2518
+ },
2519
+ {
2520
+ "epoch": 0.58,
2521
+ "learning_rate": 3.376092989372188e-06,
2522
+ "loss": 0.0348,
2523
+ "step": 4070
2524
+ },
2525
+ {
2526
+ "epoch": 0.58,
2527
+ "learning_rate": 3.3565846693293755e-06,
2528
+ "loss": 0.0407,
2529
+ "step": 4080
2530
+ },
2531
+ {
2532
+ "epoch": 0.58,
2533
+ "learning_rate": 3.337099294671723e-06,
2534
+ "loss": 0.0623,
2535
+ "step": 4090
2536
+ },
2537
+ {
2538
+ "epoch": 0.58,
2539
+ "learning_rate": 3.3176372564202514e-06,
2540
+ "loss": 0.0472,
2541
+ "step": 4100
2542
+ },
2543
+ {
2544
+ "epoch": 0.59,
2545
+ "learning_rate": 3.298198945127675e-06,
2546
+ "loss": 0.0584,
2547
+ "step": 4110
2548
+ },
2549
+ {
2550
+ "epoch": 0.59,
2551
+ "learning_rate": 3.2787847508705756e-06,
2552
+ "loss": 0.0522,
2553
+ "step": 4120
2554
+ },
2555
+ {
2556
+ "epoch": 0.59,
2557
+ "learning_rate": 3.2593950632415603e-06,
2558
+ "loss": 0.0596,
2559
+ "step": 4130
2560
+ },
2561
+ {
2562
+ "epoch": 0.59,
2563
+ "learning_rate": 3.24003027134146e-06,
2564
+ "loss": 0.0293,
2565
+ "step": 4140
2566
+ },
2567
+ {
2568
+ "epoch": 0.59,
2569
+ "learning_rate": 3.2206907637715064e-06,
2570
+ "loss": 0.0649,
2571
+ "step": 4150
2572
+ },
2573
+ {
2574
+ "epoch": 0.59,
2575
+ "learning_rate": 3.201376928625543e-06,
2576
+ "loss": 0.0601,
2577
+ "step": 4160
2578
+ },
2579
+ {
2580
+ "epoch": 0.59,
2581
+ "learning_rate": 3.182089153482235e-06,
2582
+ "loss": 0.0341,
2583
+ "step": 4170
2584
+ },
2585
+ {
2586
+ "epoch": 0.6,
2587
+ "learning_rate": 3.162827825397289e-06,
2588
+ "loss": 0.0459,
2589
+ "step": 4180
2590
+ },
2591
+ {
2592
+ "epoch": 0.6,
2593
+ "learning_rate": 3.1435933308956882e-06,
2594
+ "loss": 0.0224,
2595
+ "step": 4190
2596
+ },
2597
+ {
2598
+ "epoch": 0.6,
2599
+ "learning_rate": 3.124386055963936e-06,
2600
+ "loss": 0.0288,
2601
+ "step": 4200
2602
+ },
2603
+ {
2604
+ "epoch": 0.6,
2605
+ "learning_rate": 3.105206386042308e-06,
2606
+ "loss": 0.0997,
2607
+ "step": 4210
2608
+ },
2609
+ {
2610
+ "epoch": 0.6,
2611
+ "learning_rate": 3.0860547060171217e-06,
2612
+ "loss": 0.0825,
2613
+ "step": 4220
2614
+ },
2615
+ {
2616
+ "epoch": 0.6,
2617
+ "learning_rate": 3.0669314002130036e-06,
2618
+ "loss": 0.0409,
2619
+ "step": 4230
2620
+ },
2621
+ {
2622
+ "epoch": 0.6,
2623
+ "learning_rate": 3.047836852385192e-06,
2624
+ "loss": 0.0429,
2625
+ "step": 4240
2626
+ },
2627
+ {
2628
+ "epoch": 0.61,
2629
+ "learning_rate": 3.028771445711819e-06,
2630
+ "loss": 0.0614,
2631
+ "step": 4250
2632
+ },
2633
+ {
2634
+ "epoch": 0.61,
2635
+ "learning_rate": 3.0097355627862352e-06,
2636
+ "loss": 0.0406,
2637
+ "step": 4260
2638
+ },
2639
+ {
2640
+ "epoch": 0.61,
2641
+ "learning_rate": 2.99072958560932e-06,
2642
+ "loss": 0.0391,
2643
+ "step": 4270
2644
+ },
2645
+ {
2646
+ "epoch": 0.61,
2647
+ "learning_rate": 2.971753895581829e-06,
2648
+ "loss": 0.0983,
2649
+ "step": 4280
2650
+ },
2651
+ {
2652
+ "epoch": 0.61,
2653
+ "learning_rate": 2.952808873496727e-06,
2654
+ "loss": 0.0237,
2655
+ "step": 4290
2656
+ },
2657
+ {
2658
+ "epoch": 0.61,
2659
+ "learning_rate": 2.9338948995315536e-06,
2660
+ "loss": 0.0713,
2661
+ "step": 4300
2662
+ },
2663
+ {
2664
+ "epoch": 0.61,
2665
+ "learning_rate": 2.9150123532407973e-06,
2666
+ "loss": 0.0285,
2667
+ "step": 4310
2668
+ },
2669
+ {
2670
+ "epoch": 0.62,
2671
+ "learning_rate": 2.8961616135482686e-06,
2672
+ "loss": 0.0724,
2673
+ "step": 4320
2674
+ },
2675
+ {
2676
+ "epoch": 0.62,
2677
+ "learning_rate": 2.8773430587395077e-06,
2678
+ "loss": 0.0603,
2679
+ "step": 4330
2680
+ },
2681
+ {
2682
+ "epoch": 0.62,
2683
+ "learning_rate": 2.858557066454183e-06,
2684
+ "loss": 0.09,
2685
+ "step": 4340
2686
+ },
2687
+ {
2688
+ "epoch": 0.62,
2689
+ "learning_rate": 2.83980401367852e-06,
2690
+ "loss": 0.0203,
2691
+ "step": 4350
2692
+ },
2693
+ {
2694
+ "epoch": 0.62,
2695
+ "learning_rate": 2.8210842767377328e-06,
2696
+ "loss": 0.0579,
2697
+ "step": 4360
2698
+ },
2699
+ {
2700
+ "epoch": 0.62,
2701
+ "learning_rate": 2.8023982312884707e-06,
2702
+ "loss": 0.0249,
2703
+ "step": 4370
2704
+ },
2705
+ {
2706
+ "epoch": 0.62,
2707
+ "learning_rate": 2.783746252311286e-06,
2708
+ "loss": 0.0508,
2709
+ "step": 4380
2710
+ },
2711
+ {
2712
+ "epoch": 0.63,
2713
+ "learning_rate": 2.7651287141031027e-06,
2714
+ "loss": 0.0213,
2715
+ "step": 4390
2716
+ },
2717
+ {
2718
+ "epoch": 0.63,
2719
+ "learning_rate": 2.746545990269708e-06,
2720
+ "loss": 0.086,
2721
+ "step": 4400
2722
+ },
2723
+ {
2724
+ "epoch": 0.63,
2725
+ "learning_rate": 2.727998453718255e-06,
2726
+ "loss": 0.0348,
2727
+ "step": 4410
2728
+ },
2729
+ {
2730
+ "epoch": 0.63,
2731
+ "learning_rate": 2.709486476649779e-06,
2732
+ "loss": 0.1003,
2733
+ "step": 4420
2734
+ },
2735
+ {
2736
+ "epoch": 0.63,
2737
+ "learning_rate": 2.6910104305517307e-06,
2738
+ "loss": 0.0035,
2739
+ "step": 4430
2740
+ },
2741
+ {
2742
+ "epoch": 0.63,
2743
+ "learning_rate": 2.672570686190513e-06,
2744
+ "loss": 0.0776,
2745
+ "step": 4440
2746
+ },
2747
+ {
2748
+ "epoch": 0.63,
2749
+ "learning_rate": 2.6541676136040553e-06,
2750
+ "loss": 0.0812,
2751
+ "step": 4450
2752
+ },
2753
+ {
2754
+ "epoch": 0.64,
2755
+ "learning_rate": 2.635801582094372e-06,
2756
+ "loss": 0.0662,
2757
+ "step": 4460
2758
+ },
2759
+ {
2760
+ "epoch": 0.64,
2761
+ "learning_rate": 2.6174729602201664e-06,
2762
+ "loss": 0.0873,
2763
+ "step": 4470
2764
+ },
2765
+ {
2766
+ "epoch": 0.64,
2767
+ "learning_rate": 2.5991821157894182e-06,
2768
+ "loss": 0.0223,
2769
+ "step": 4480
2770
+ },
2771
+ {
2772
+ "epoch": 0.64,
2773
+ "learning_rate": 2.5809294158520197e-06,
2774
+ "loss": 0.0421,
2775
+ "step": 4490
2776
+ },
2777
+ {
2778
+ "epoch": 0.64,
2779
+ "learning_rate": 2.5627152266923973e-06,
2780
+ "loss": 0.061,
2781
+ "step": 4500
2782
+ },
2783
+ {
2784
+ "epoch": 0.64,
2785
+ "eval_accuracy": 0.0,
2786
+ "eval_loss": 0.127749502658844,
2787
+ "eval_runtime": 447.4087,
2788
+ "eval_samples_per_second": 52.254,
2789
+ "eval_steps_per_second": 8.71,
2790
+ "step": 4500
2791
+ },
2792
+ {
2793
+ "epoch": 0.64,
2794
+ "learning_rate": 2.5445399138221693e-06,
2795
+ "loss": 0.0794,
2796
+ "step": 4510
2797
+ },
2798
+ {
2799
+ "epoch": 0.64,
2800
+ "learning_rate": 2.5264038419728033e-06,
2801
+ "loss": 0.0787,
2802
+ "step": 4520
2803
+ },
2804
+ {
2805
+ "epoch": 0.65,
2806
+ "learning_rate": 2.5083073750883057e-06,
2807
+ "loss": 0.0568,
2808
+ "step": 4530
2809
+ },
2810
+ {
2811
+ "epoch": 0.65,
2812
+ "learning_rate": 2.4902508763179086e-06,
2813
+ "loss": 0.0291,
2814
+ "step": 4540
2815
+ },
2816
+ {
2817
+ "epoch": 0.65,
2818
+ "learning_rate": 2.472234708008795e-06,
2819
+ "loss": 0.0688,
2820
+ "step": 4550
2821
+ },
2822
+ {
2823
+ "epoch": 0.65,
2824
+ "learning_rate": 2.4542592316988106e-06,
2825
+ "loss": 0.0574,
2826
+ "step": 4560
2827
+ },
2828
+ {
2829
+ "epoch": 0.65,
2830
+ "learning_rate": 2.436324808109227e-06,
2831
+ "loss": 0.0695,
2832
+ "step": 4570
2833
+ },
2834
+ {
2835
+ "epoch": 0.65,
2836
+ "learning_rate": 2.4184317971374866e-06,
2837
+ "loss": 0.1057,
2838
+ "step": 4580
2839
+ },
2840
+ {
2841
+ "epoch": 0.65,
2842
+ "learning_rate": 2.4005805578499946e-06,
2843
+ "loss": 0.0234,
2844
+ "step": 4590
2845
+ },
2846
+ {
2847
+ "epoch": 0.66,
2848
+ "learning_rate": 2.3827714484748997e-06,
2849
+ "loss": 0.0852,
2850
+ "step": 4600
2851
+ },
2852
+ {
2853
+ "epoch": 0.66,
2854
+ "learning_rate": 2.3650048263949187e-06,
2855
+ "loss": 0.0613,
2856
+ "step": 4610
2857
+ },
2858
+ {
2859
+ "epoch": 0.66,
2860
+ "learning_rate": 2.3472810481401524e-06,
2861
+ "loss": 0.0783,
2862
+ "step": 4620
2863
+ },
2864
+ {
2865
+ "epoch": 0.66,
2866
+ "learning_rate": 2.329600469380944e-06,
2867
+ "loss": 0.0732,
2868
+ "step": 4630
2869
+ },
2870
+ {
2871
+ "epoch": 0.66,
2872
+ "learning_rate": 2.3119634449207286e-06,
2873
+ "loss": 0.0728,
2874
+ "step": 4640
2875
+ },
2876
+ {
2877
+ "epoch": 0.66,
2878
+ "learning_rate": 2.294370328688919e-06,
2879
+ "loss": 0.1218,
2880
+ "step": 4650
2881
+ },
2882
+ {
2883
+ "epoch": 0.66,
2884
+ "learning_rate": 2.2768214737338107e-06,
2885
+ "loss": 0.0373,
2886
+ "step": 4660
2887
+ },
2888
+ {
2889
+ "epoch": 0.67,
2890
+ "learning_rate": 2.2593172322154817e-06,
2891
+ "loss": 0.0424,
2892
+ "step": 4670
2893
+ },
2894
+ {
2895
+ "epoch": 0.67,
2896
+ "learning_rate": 2.24185795539874e-06,
2897
+ "loss": 0.0805,
2898
+ "step": 4680
2899
+ },
2900
+ {
2901
+ "epoch": 0.67,
2902
+ "learning_rate": 2.224443993646062e-06,
2903
+ "loss": 0.0652,
2904
+ "step": 4690
2905
+ },
2906
+ {
2907
+ "epoch": 0.67,
2908
+ "learning_rate": 2.207075696410577e-06,
2909
+ "loss": 0.0688,
2910
+ "step": 4700
2911
+ },
2912
+ {
2913
+ "epoch": 0.67,
2914
+ "learning_rate": 2.1897534122290383e-06,
2915
+ "loss": 0.0305,
2916
+ "step": 4710
2917
+ },
2918
+ {
2919
+ "epoch": 0.67,
2920
+ "learning_rate": 2.1724774887148365e-06,
2921
+ "loss": 0.0273,
2922
+ "step": 4720
2923
+ },
2924
+ {
2925
+ "epoch": 0.67,
2926
+ "learning_rate": 2.1552482725510296e-06,
2927
+ "loss": 0.051,
2928
+ "step": 4730
2929
+ },
2930
+ {
2931
+ "epoch": 0.68,
2932
+ "learning_rate": 2.1380661094833717e-06,
2933
+ "loss": 0.0281,
2934
+ "step": 4740
2935
+ },
2936
+ {
2937
+ "epoch": 0.68,
2938
+ "learning_rate": 2.120931344313393e-06,
2939
+ "loss": 0.0893,
2940
+ "step": 4750
2941
+ },
2942
+ {
2943
+ "epoch": 0.68,
2944
+ "learning_rate": 2.1038443208914606e-06,
2945
+ "loss": 0.0333,
2946
+ "step": 4760
2947
+ },
2948
+ {
2949
+ "epoch": 0.68,
2950
+ "learning_rate": 2.0868053821098957e-06,
2951
+ "loss": 0.0247,
2952
+ "step": 4770
2953
+ },
2954
+ {
2955
+ "epoch": 0.68,
2956
+ "learning_rate": 2.0698148698960826e-06,
2957
+ "loss": 0.0781,
2958
+ "step": 4780
2959
+ },
2960
+ {
2961
+ "epoch": 0.68,
2962
+ "learning_rate": 2.0528731252056063e-06,
2963
+ "loss": 0.0861,
2964
+ "step": 4790
2965
+ },
2966
+ {
2967
+ "epoch": 0.68,
2968
+ "learning_rate": 2.035980488015422e-06,
2969
+ "loss": 0.0862,
2970
+ "step": 4800
2971
+ },
2972
+ {
2973
+ "epoch": 0.69,
2974
+ "learning_rate": 2.0191372973170156e-06,
2975
+ "loss": 0.012,
2976
+ "step": 4810
2977
+ },
2978
+ {
2979
+ "epoch": 0.69,
2980
+ "learning_rate": 2.0023438911096174e-06,
2981
+ "loss": 0.0289,
2982
+ "step": 4820
2983
+ },
2984
+ {
2985
+ "epoch": 0.69,
2986
+ "learning_rate": 1.9856006063934057e-06,
2987
+ "loss": 0.0076,
2988
+ "step": 4830
2989
+ },
2990
+ {
2991
+ "epoch": 0.69,
2992
+ "learning_rate": 1.9689077791627564e-06,
2993
+ "loss": 0.0455,
2994
+ "step": 4840
2995
+ },
2996
+ {
2997
+ "epoch": 0.69,
2998
+ "learning_rate": 1.952265744399489e-06,
2999
+ "loss": 0.1003,
3000
+ "step": 4850
3001
+ },
3002
+ {
3003
+ "epoch": 0.69,
3004
+ "learning_rate": 1.935674836066151e-06,
3005
+ "loss": 0.1002,
3006
+ "step": 4860
3007
+ },
3008
+ {
3009
+ "epoch": 0.69,
3010
+ "learning_rate": 1.9191353870993173e-06,
3011
+ "loss": 0.098,
3012
+ "step": 4870
3013
+ },
3014
+ {
3015
+ "epoch": 0.7,
3016
+ "learning_rate": 1.902647729402903e-06,
3017
+ "loss": 0.0318,
3018
+ "step": 4880
3019
+ },
3020
+ {
3021
+ "epoch": 0.7,
3022
+ "learning_rate": 1.8862121938415108e-06,
3023
+ "loss": 0.0779,
3024
+ "step": 4890
3025
+ },
3026
+ {
3027
+ "epoch": 0.7,
3028
+ "learning_rate": 1.8698291102337817e-06,
3029
+ "loss": 0.013,
3030
+ "step": 4900
3031
+ },
3032
+ {
3033
+ "epoch": 0.7,
3034
+ "learning_rate": 1.8534988073457876e-06,
3035
+ "loss": 0.0663,
3036
+ "step": 4910
3037
+ },
3038
+ {
3039
+ "epoch": 0.7,
3040
+ "learning_rate": 1.8372216128844242e-06,
3041
+ "loss": 0.0701,
3042
+ "step": 4920
3043
+ },
3044
+ {
3045
+ "epoch": 0.7,
3046
+ "learning_rate": 1.820997853490837e-06,
3047
+ "loss": 0.027,
3048
+ "step": 4930
3049
+ },
3050
+ {
3051
+ "epoch": 0.7,
3052
+ "learning_rate": 1.8048278547338745e-06,
3053
+ "loss": 0.0587,
3054
+ "step": 4940
3055
+ },
3056
+ {
3057
+ "epoch": 0.71,
3058
+ "learning_rate": 1.788711941103541e-06,
3059
+ "loss": 0.1021,
3060
+ "step": 4950
3061
+ },
3062
+ {
3063
+ "epoch": 0.71,
3064
+ "learning_rate": 1.7726504360044994e-06,
3065
+ "loss": 0.0984,
3066
+ "step": 4960
3067
+ },
3068
+ {
3069
+ "epoch": 0.71,
3070
+ "learning_rate": 1.7566436617495665e-06,
3071
+ "loss": 0.0771,
3072
+ "step": 4970
3073
+ },
3074
+ {
3075
+ "epoch": 0.71,
3076
+ "learning_rate": 1.7406919395532607e-06,
3077
+ "loss": 0.0779,
3078
+ "step": 4980
3079
+ },
3080
+ {
3081
+ "epoch": 0.71,
3082
+ "learning_rate": 1.7247955895253417e-06,
3083
+ "loss": 0.0265,
3084
+ "step": 4990
3085
+ },
3086
+ {
3087
+ "epoch": 0.71,
3088
+ "learning_rate": 1.7089549306643985e-06,
3089
+ "loss": 0.0804,
3090
+ "step": 5000
3091
+ },
3092
+ {
3093
+ "epoch": 0.71,
3094
+ "eval_accuracy": 0.0,
3095
+ "eval_loss": 0.10553926229476929,
3096
+ "eval_runtime": 447.3537,
3097
+ "eval_samples_per_second": 52.261,
3098
+ "eval_steps_per_second": 8.711,
3099
+ "step": 5000
3100
+ },
3101
+ {
3102
+ "epoch": 0.71,
3103
+ "learning_rate": 1.6931702808514368e-06,
3104
+ "loss": 0.1135,
3105
+ "step": 5010
3106
+ },
3107
+ {
3108
+ "epoch": 0.72,
3109
+ "learning_rate": 1.6774419568435118e-06,
3110
+ "loss": 0.0521,
3111
+ "step": 5020
3112
+ },
3113
+ {
3114
+ "epoch": 0.72,
3115
+ "learning_rate": 1.6617702742673597e-06,
3116
+ "loss": 0.048,
3117
+ "step": 5030
3118
+ },
3119
+ {
3120
+ "epoch": 0.72,
3121
+ "learning_rate": 1.6461555476130759e-06,
3122
+ "loss": 0.0534,
3123
+ "step": 5040
3124
+ },
3125
+ {
3126
+ "epoch": 0.72,
3127
+ "learning_rate": 1.630598090227791e-06,
3128
+ "loss": 0.0355,
3129
+ "step": 5050
3130
+ },
3131
+ {
3132
+ "epoch": 0.72,
3133
+ "learning_rate": 1.615098214309398e-06,
3134
+ "loss": 0.0472,
3135
+ "step": 5060
3136
+ },
3137
+ {
3138
+ "epoch": 0.72,
3139
+ "learning_rate": 1.5996562309002698e-06,
3140
+ "loss": 0.0735,
3141
+ "step": 5070
3142
+ },
3143
+ {
3144
+ "epoch": 0.72,
3145
+ "learning_rate": 1.5842724498810353e-06,
3146
+ "loss": 0.0217,
3147
+ "step": 5080
3148
+ },
3149
+ {
3150
+ "epoch": 0.73,
3151
+ "learning_rate": 1.5689471799643443e-06,
3152
+ "loss": 0.0754,
3153
+ "step": 5090
3154
+ },
3155
+ {
3156
+ "epoch": 0.73,
3157
+ "learning_rate": 1.5536807286886871e-06,
3158
+ "loss": 0.0841,
3159
+ "step": 5100
3160
+ },
3161
+ {
3162
+ "epoch": 0.73,
3163
+ "learning_rate": 1.53847340241221e-06,
3164
+ "loss": 0.0436,
3165
+ "step": 5110
3166
+ },
3167
+ {
3168
+ "epoch": 0.73,
3169
+ "learning_rate": 1.5233255063065798e-06,
3170
+ "loss": 0.0282,
3171
+ "step": 5120
3172
+ },
3173
+ {
3174
+ "epoch": 0.73,
3175
+ "learning_rate": 1.5082373443508486e-06,
3176
+ "loss": 0.036,
3177
+ "step": 5130
3178
+ },
3179
+ {
3180
+ "epoch": 0.73,
3181
+ "learning_rate": 1.4932092193253592e-06,
3182
+ "loss": 0.0263,
3183
+ "step": 5140
3184
+ },
3185
+ {
3186
+ "epoch": 0.73,
3187
+ "learning_rate": 1.4782414328056736e-06,
3188
+ "loss": 0.1591,
3189
+ "step": 5150
3190
+ },
3191
+ {
3192
+ "epoch": 0.74,
3193
+ "learning_rate": 1.4633342851565094e-06,
3194
+ "loss": 0.0599,
3195
+ "step": 5160
3196
+ },
3197
+ {
3198
+ "epoch": 0.74,
3199
+ "learning_rate": 1.4484880755257256e-06,
3200
+ "loss": 0.1064,
3201
+ "step": 5170
3202
+ },
3203
+ {
3204
+ "epoch": 0.74,
3205
+ "learning_rate": 1.4337031018383067e-06,
3206
+ "loss": 0.0376,
3207
+ "step": 5180
3208
+ },
3209
+ {
3210
+ "epoch": 0.74,
3211
+ "learning_rate": 1.418979660790396e-06,
3212
+ "loss": 0.0918,
3213
+ "step": 5190
3214
+ },
3215
+ {
3216
+ "epoch": 0.74,
3217
+ "learning_rate": 1.4043180478433327e-06,
3218
+ "loss": 0.057,
3219
+ "step": 5200
3220
+ },
3221
+ {
3222
+ "epoch": 0.74,
3223
+ "learning_rate": 1.3897185572177244e-06,
3224
+ "loss": 0.0387,
3225
+ "step": 5210
3226
+ },
3227
+ {
3228
+ "epoch": 0.74,
3229
+ "learning_rate": 1.375181481887551e-06,
3230
+ "loss": 0.0826,
3231
+ "step": 5220
3232
+ },
3233
+ {
3234
+ "epoch": 0.75,
3235
+ "learning_rate": 1.360707113574272e-06,
3236
+ "loss": 0.043,
3237
+ "step": 5230
3238
+ },
3239
+ {
3240
+ "epoch": 0.75,
3241
+ "learning_rate": 1.3462957427409852e-06,
3242
+ "loss": 0.0494,
3243
+ "step": 5240
3244
+ },
3245
+ {
3246
+ "epoch": 0.75,
3247
+ "learning_rate": 1.3319476585865862e-06,
3248
+ "loss": 0.0351,
3249
+ "step": 5250
3250
+ },
3251
+ {
3252
+ "epoch": 0.75,
3253
+ "learning_rate": 1.317663149039977e-06,
3254
+ "loss": 0.0078,
3255
+ "step": 5260
3256
+ },
3257
+ {
3258
+ "epoch": 0.75,
3259
+ "learning_rate": 1.3034425007542776e-06,
3260
+ "loss": 0.0252,
3261
+ "step": 5270
3262
+ },
3263
+ {
3264
+ "epoch": 0.75,
3265
+ "learning_rate": 1.289285999101078e-06,
3266
+ "loss": 0.0471,
3267
+ "step": 5280
3268
+ },
3269
+ {
3270
+ "epoch": 0.75,
3271
+ "learning_rate": 1.2751939281647143e-06,
3272
+ "loss": 0.0454,
3273
+ "step": 5290
3274
+ },
3275
+ {
3276
+ "epoch": 0.76,
3277
+ "learning_rate": 1.2611665707365612e-06,
3278
+ "loss": 0.0639,
3279
+ "step": 5300
3280
+ },
3281
+ {
3282
+ "epoch": 0.76,
3283
+ "learning_rate": 1.2472042083093647e-06,
3284
+ "loss": 0.1117,
3285
+ "step": 5310
3286
+ },
3287
+ {
3288
+ "epoch": 0.76,
3289
+ "learning_rate": 1.2333071210715855e-06,
3290
+ "loss": 0.0734,
3291
+ "step": 5320
3292
+ },
3293
+ {
3294
+ "epoch": 0.76,
3295
+ "learning_rate": 1.2194755879017843e-06,
3296
+ "loss": 0.0301,
3297
+ "step": 5330
3298
+ },
3299
+ {
3300
+ "epoch": 0.76,
3301
+ "learning_rate": 1.2057098863630184e-06,
3302
+ "loss": 0.0352,
3303
+ "step": 5340
3304
+ },
3305
+ {
3306
+ "epoch": 0.76,
3307
+ "learning_rate": 1.1920102926972738e-06,
3308
+ "loss": 0.0521,
3309
+ "step": 5350
3310
+ },
3311
+ {
3312
+ "epoch": 0.76,
3313
+ "learning_rate": 1.1783770818199275e-06,
3314
+ "loss": 0.0261,
3315
+ "step": 5360
3316
+ },
3317
+ {
3318
+ "epoch": 0.77,
3319
+ "learning_rate": 1.1648105273142198e-06,
3320
+ "loss": 0.0568,
3321
+ "step": 5370
3322
+ },
3323
+ {
3324
+ "epoch": 0.77,
3325
+ "learning_rate": 1.151310901425776e-06,
3326
+ "loss": 0.0203,
3327
+ "step": 5380
3328
+ },
3329
+ {
3330
+ "epoch": 0.77,
3331
+ "learning_rate": 1.1378784750571335e-06,
3332
+ "loss": 0.083,
3333
+ "step": 5390
3334
+ },
3335
+ {
3336
+ "epoch": 0.77,
3337
+ "learning_rate": 1.1245135177623087e-06,
3338
+ "loss": 0.0551,
3339
+ "step": 5400
3340
+ },
3341
+ {
3342
+ "epoch": 0.77,
3343
+ "learning_rate": 1.1112162977413925e-06,
3344
+ "loss": 0.0085,
3345
+ "step": 5410
3346
+ },
3347
+ {
3348
+ "epoch": 0.77,
3349
+ "learning_rate": 1.0979870818351596e-06,
3350
+ "loss": 0.0411,
3351
+ "step": 5420
3352
+ },
3353
+ {
3354
+ "epoch": 0.77,
3355
+ "learning_rate": 1.0848261355197223e-06,
3356
+ "loss": 0.0857,
3357
+ "step": 5430
3358
+ },
3359
+ {
3360
+ "epoch": 0.78,
3361
+ "learning_rate": 1.0717337229011941e-06,
3362
+ "loss": 0.0282,
3363
+ "step": 5440
3364
+ },
3365
+ {
3366
+ "epoch": 0.78,
3367
+ "learning_rate": 1.058710106710401e-06,
3368
+ "loss": 0.0527,
3369
+ "step": 5450
3370
+ },
3371
+ {
3372
+ "epoch": 0.78,
3373
+ "learning_rate": 1.0457555482975958e-06,
3374
+ "loss": 0.0702,
3375
+ "step": 5460
3376
+ },
3377
+ {
3378
+ "epoch": 0.78,
3379
+ "learning_rate": 1.0328703076272267e-06,
3380
+ "loss": 0.0824,
3381
+ "step": 5470
3382
+ },
3383
+ {
3384
+ "epoch": 0.78,
3385
+ "learning_rate": 1.0200546432727093e-06,
3386
+ "loss": 0.0826,
3387
+ "step": 5480
3388
+ },
3389
+ {
3390
+ "epoch": 0.78,
3391
+ "learning_rate": 1.007308812411247e-06,
3392
+ "loss": 0.0542,
3393
+ "step": 5490
3394
+ },
3395
+ {
3396
+ "epoch": 0.78,
3397
+ "learning_rate": 9.94633070818662e-07,
3398
+ "loss": 0.0886,
3399
+ "step": 5500
3400
+ },
3401
+ {
3402
+ "epoch": 0.78,
3403
+ "eval_accuracy": 0.0,
3404
+ "eval_loss": 0.10989828407764435,
3405
+ "eval_runtime": 447.8187,
3406
+ "eval_samples_per_second": 52.206,
3407
+ "eval_steps_per_second": 8.702,
3408
+ "step": 5500
3409
+ },
3410
+ {
3411
+ "epoch": 0.79,
3412
+ "learning_rate": 9.8202767286427e-07,
3413
+ "loss": 0.0689,
3414
+ "step": 5510
3415
+ },
3416
+ {
3417
+ "epoch": 0.79,
3418
+ "learning_rate": 9.69492871505768e-07,
3419
+ "loss": 0.0212,
3420
+ "step": 5520
3421
+ },
3422
+ {
3423
+ "epoch": 0.79,
3424
+ "learning_rate": 9.570289182841664e-07,
3425
+ "loss": 0.0596,
3426
+ "step": 5530
3427
+ },
3428
+ {
3429
+ "epoch": 0.79,
3430
+ "learning_rate": 9.44636063318732e-07,
3431
+ "loss": 0.0136,
3432
+ "step": 5540
3433
+ },
3434
+ {
3435
+ "epoch": 0.79,
3436
+ "learning_rate": 9.323145553019798e-07,
3437
+ "loss": 0.0335,
3438
+ "step": 5550
3439
+ },
3440
+ {
3441
+ "epoch": 0.79,
3442
+ "learning_rate": 9.200646414946699e-07,
3443
+ "loss": 0.041,
3444
+ "step": 5560
3445
+ },
3446
+ {
3447
+ "epoch": 0.79,
3448
+ "learning_rate": 9.078865677208571e-07,
3449
+ "loss": 0.0312,
3450
+ "step": 5570
3451
+ },
3452
+ {
3453
+ "epoch": 0.8,
3454
+ "learning_rate": 8.957805783629492e-07,
3455
+ "loss": 0.0638,
3456
+ "step": 5580
3457
+ },
3458
+ {
3459
+ "epoch": 0.8,
3460
+ "learning_rate": 8.837469163568095e-07,
3461
+ "loss": 0.0227,
3462
+ "step": 5590
3463
+ },
3464
+ {
3465
+ "epoch": 0.8,
3466
+ "learning_rate": 8.717858231868755e-07,
3467
+ "loss": 0.0649,
3468
+ "step": 5600
3469
+ },
3470
+ {
3471
+ "epoch": 0.8,
3472
+ "learning_rate": 8.59897538881317e-07,
3473
+ "loss": 0.0137,
3474
+ "step": 5610
3475
+ },
3476
+ {
3477
+ "epoch": 0.8,
3478
+ "learning_rate": 8.480823020072212e-07,
3479
+ "loss": 0.0243,
3480
+ "step": 5620
3481
+ },
3482
+ {
3483
+ "epoch": 0.8,
3484
+ "learning_rate": 8.363403496657983e-07,
3485
+ "loss": 0.0273,
3486
+ "step": 5630
3487
+ },
3488
+ {
3489
+ "epoch": 0.8,
3490
+ "learning_rate": 8.246719174876306e-07,
3491
+ "loss": 0.0572,
3492
+ "step": 5640
3493
+ },
3494
+ {
3495
+ "epoch": 0.81,
3496
+ "learning_rate": 8.13077239627939e-07,
3497
+ "loss": 0.0629,
3498
+ "step": 5650
3499
+ },
3500
+ {
3501
+ "epoch": 0.81,
3502
+ "learning_rate": 8.015565487618886e-07,
3503
+ "loss": 0.0406,
3504
+ "step": 5660
3505
+ },
3506
+ {
3507
+ "epoch": 0.81,
3508
+ "learning_rate": 7.901100760799157e-07,
3509
+ "loss": 0.0314,
3510
+ "step": 5670
3511
+ },
3512
+ {
3513
+ "epoch": 0.81,
3514
+ "learning_rate": 7.78738051283088e-07,
3515
+ "loss": 0.0787,
3516
+ "step": 5680
3517
+ },
3518
+ {
3519
+ "epoch": 0.81,
3520
+ "learning_rate": 7.674407025785011e-07,
3521
+ "loss": 0.1002,
3522
+ "step": 5690
3523
+ },
3524
+ {
3525
+ "epoch": 0.81,
3526
+ "learning_rate": 7.5621825667469e-07,
3527
+ "loss": 0.1132,
3528
+ "step": 5700
3529
+ },
3530
+ {
3531
+ "epoch": 0.81,
3532
+ "learning_rate": 7.4507093877709e-07,
3533
+ "loss": 0.0525,
3534
+ "step": 5710
3535
+ },
3536
+ {
3537
+ "epoch": 0.82,
3538
+ "learning_rate": 7.33998972583506e-07,
3539
+ "loss": 0.035,
3540
+ "step": 5720
3541
+ },
3542
+ {
3543
+ "epoch": 0.82,
3544
+ "learning_rate": 7.230025802796343e-07,
3545
+ "loss": 0.0541,
3546
+ "step": 5730
3547
+ },
3548
+ {
3549
+ "epoch": 0.82,
3550
+ "learning_rate": 7.120819825345952e-07,
3551
+ "loss": 0.0394,
3552
+ "step": 5740
3553
+ },
3554
+ {
3555
+ "epoch": 0.82,
3556
+ "learning_rate": 7.01237398496509e-07,
3557
+ "loss": 0.0442,
3558
+ "step": 5750
3559
+ },
3560
+ {
3561
+ "epoch": 0.82,
3562
+ "learning_rate": 6.904690457881011e-07,
3563
+ "loss": 0.066,
3564
+ "step": 5760
3565
+ },
3566
+ {
3567
+ "epoch": 0.82,
3568
+ "learning_rate": 6.797771405023263e-07,
3569
+ "loss": 0.032,
3570
+ "step": 5770
3571
+ },
3572
+ {
3573
+ "epoch": 0.82,
3574
+ "learning_rate": 6.69161897198043e-07,
3575
+ "loss": 0.0657,
3576
+ "step": 5780
3577
+ },
3578
+ {
3579
+ "epoch": 0.83,
3580
+ "learning_rate": 6.586235288956968e-07,
3581
+ "loss": 0.0427,
3582
+ "step": 5790
3583
+ },
3584
+ {
3585
+ "epoch": 0.83,
3586
+ "learning_rate": 6.481622470730558e-07,
3587
+ "loss": 0.0726,
3588
+ "step": 5800
3589
+ },
3590
+ {
3591
+ "epoch": 0.83,
3592
+ "learning_rate": 6.377782616609594e-07,
3593
+ "loss": 0.0229,
3594
+ "step": 5810
3595
+ },
3596
+ {
3597
+ "epoch": 0.83,
3598
+ "learning_rate": 6.274717810391067e-07,
3599
+ "loss": 0.0292,
3600
+ "step": 5820
3601
+ },
3602
+ {
3603
+ "epoch": 0.83,
3604
+ "learning_rate": 6.172430120318808e-07,
3605
+ "loss": 0.0189,
3606
+ "step": 5830
3607
+ },
3608
+ {
3609
+ "epoch": 0.83,
3610
+ "learning_rate": 6.070921599041899e-07,
3611
+ "loss": 0.0462,
3612
+ "step": 5840
3613
+ },
3614
+ {
3615
+ "epoch": 0.83,
3616
+ "learning_rate": 5.970194283573546e-07,
3617
+ "loss": 0.0274,
3618
+ "step": 5850
3619
+ },
3620
+ {
3621
+ "epoch": 0.84,
3622
+ "learning_rate": 5.870250195250159e-07,
3623
+ "loss": 0.0131,
3624
+ "step": 5860
3625
+ },
3626
+ {
3627
+ "epoch": 0.84,
3628
+ "learning_rate": 5.771091339690828e-07,
3629
+ "loss": 0.0304,
3630
+ "step": 5870
3631
+ },
3632
+ {
3633
+ "epoch": 0.84,
3634
+ "learning_rate": 5.672719706757034e-07,
3635
+ "loss": 0.0352,
3636
+ "step": 5880
3637
+ },
3638
+ {
3639
+ "epoch": 0.84,
3640
+ "learning_rate": 5.575137270512739e-07,
3641
+ "loss": 0.0677,
3642
+ "step": 5890
3643
+ },
3644
+ {
3645
+ "epoch": 0.84,
3646
+ "learning_rate": 5.478345989184799e-07,
3647
+ "loss": 0.0441,
3648
+ "step": 5900
3649
+ },
3650
+ {
3651
+ "epoch": 0.84,
3652
+ "learning_rate": 5.382347805123599e-07,
3653
+ "loss": 0.0603,
3654
+ "step": 5910
3655
+ },
3656
+ {
3657
+ "epoch": 0.84,
3658
+ "learning_rate": 5.28714464476415e-07,
3659
+ "loss": 0.0078,
3660
+ "step": 5920
3661
+ },
3662
+ {
3663
+ "epoch": 0.85,
3664
+ "learning_rate": 5.192738418587362e-07,
3665
+ "loss": 0.0277,
3666
+ "step": 5930
3667
+ },
3668
+ {
3669
+ "epoch": 0.85,
3670
+ "learning_rate": 5.099131021081786e-07,
3671
+ "loss": 0.0442,
3672
+ "step": 5940
3673
+ },
3674
+ {
3675
+ "epoch": 0.85,
3676
+ "learning_rate": 5.006324330705489e-07,
3677
+ "loss": 0.0308,
3678
+ "step": 5950
3679
+ },
3680
+ {
3681
+ "epoch": 0.85,
3682
+ "learning_rate": 4.914320209848486e-07,
3683
+ "loss": 0.0695,
3684
+ "step": 5960
3685
+ },
3686
+ {
3687
+ "epoch": 0.85,
3688
+ "learning_rate": 4.823120504795239e-07,
3689
+ "loss": 0.0592,
3690
+ "step": 5970
3691
+ },
3692
+ {
3693
+ "epoch": 0.85,
3694
+ "learning_rate": 4.732727045687714e-07,
3695
+ "loss": 0.0112,
3696
+ "step": 5980
3697
+ },
3698
+ {
3699
+ "epoch": 0.85,
3700
+ "learning_rate": 4.643141646488575e-07,
3701
+ "loss": 0.0376,
3702
+ "step": 5990
3703
+ },
3704
+ {
3705
+ "epoch": 0.86,
3706
+ "learning_rate": 4.554366104944838e-07,
3707
+ "loss": 0.0591,
3708
+ "step": 6000
3709
+ },
3710
+ {
3711
+ "epoch": 0.86,
3712
+ "eval_accuracy": 0.0,
3713
+ "eval_loss": 0.127273827791214,
3714
+ "eval_runtime": 447.0287,
3715
+ "eval_samples_per_second": 52.299,
3716
+ "eval_steps_per_second": 8.718,
3717
+ "step": 6000
3718
+ }
3719
+ ],
3720
+ "logging_steps": 10,
3721
+ "max_steps": 7013,
3722
+ "num_train_epochs": 1,
3723
+ "save_steps": 500,
3724
+ "total_flos": 0.0,
3725
+ "trial_name": null,
3726
+ "trial_params": null
3727
+ }
checkpoint-6000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cbd473640b089575e3b93f40caac6d96b1006088b463e865229fad92539da1e0
3
+ size 4600