Hugging Face's logo

nbroad
/
deberta-v3-base-company-names

Token Classification
Transformers
PyTorch
deberta-v2
Generated from Trainer
Eval Results
Model card Files
xet
Community
4
nbroad commited on
Commit
6fd5484
·
verified ·
1 Parent(s): dda49f8

Model save

Browse files
Files changed (13) hide show
  1. README.md +69 -0
  2. added_tokens.json +3 -0
  3. all_results.json +17 -0
  4. config.json +46 -0
  5. eval_results.json +12 -0
  6. pytorch_model.bin +3 -0
  7. special_tokens_map.json +15 -0
  8. spm.model +3 -0
  9. tokenizer.json +0 -0
  10. tokenizer_config.json +58 -0
  11. train_results.json +8 -0
  12. trainer_state.json +3886 -0
  13. training_args.bin +3 -0
README.md ADDED
@@ -0,0 +1,69 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: mit
3
+ base_model: microsoft/deberta-v3-base
4
+ tags:
5
+ - generated_from_trainer
6
+ metrics:
7
+ - precision
8
+ - recall
9
+ - f1
10
+ - accuracy
11
+ model-index:
12
+ - name: deberta-v3-base-company-names
13
+ results: []
14
+ ---
15
+
16
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
17
+ should probably proofread and complete it, then remove this comment. -->
18
+
19
+ # deberta-v3-base-company-names
20
+
21
+ This model is a fine-tuned version of [microsoft/deberta-v3-base](https://huggingface.co/microsoft/deberta-v3-base) on an unknown dataset.
22
+ It achieves the following results on the evaluation set:
23
+ - Loss: 0.0693
24
+ - Precision: 0.7740
25
+ - Recall: 0.7963
26
+ - F1: 0.7850
27
+ - Accuracy: 0.9769
28
+
29
+ ## Model description
30
+
31
+ More information needed
32
+
33
+ ## Intended uses & limitations
34
+
35
+ More information needed
36
+
37
+ ## Training and evaluation data
38
+
39
+ More information needed
40
+
41
+ ## Training procedure
42
+
43
+ ### Training hyperparameters
44
+
45
+ The following hyperparameters were used during training:
46
+ - learning_rate: 8e-05
47
+ - train_batch_size: 48
48
+ - eval_batch_size: 8
49
+ - seed: 42
50
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
51
+ - lr_scheduler_type: linear
52
+ - lr_scheduler_warmup_ratio: 0.1
53
+ - num_epochs: 3.0
54
+
55
+ ### Training results
56
+
57
+ | Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
58
+ |:-------------:|:-----:|:----:|:---------------:|:---------:|:------:|:------:|:--------:|
59
+ | 0.0752 | 1.0 | 2126 | 0.0664 | 0.7416 | 0.7979 | 0.7687 | 0.9757 |
60
+ | 0.0484 | 2.0 | 4252 | 0.0652 | 0.7725 | 0.7903 | 0.7813 | 0.9768 |
61
+ | 0.0415 | 3.0 | 6378 | 0.0693 | 0.7740 | 0.7963 | 0.7850 | 0.9769 |
62
+
63
+
64
+ ### Framework versions
65
+
66
+ - Transformers 4.34.1
67
+ - Pytorch 2.0.1+cu117
68
+ - Datasets 2.16.1
69
+ - Tokenizers 0.14.1
added_tokens.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ {
2
+ "[MASK]": 128000
3
+ }
all_results.json ADDED
@@ -0,0 +1,17 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 3.0,
3
+ "eval_accuracy": 0.9769126125154315,
4
+ "eval_f1": 0.7849694196330357,
5
+ "eval_loss": 0.06933891773223877,
6
+ "eval_precision": 0.7739696312364425,
7
+ "eval_recall": 0.7962863774326013,
8
+ "eval_runtime": 13.9655,
9
+ "eval_samples": 14160,
10
+ "eval_samples_per_second": 1013.925,
11
+ "eval_steps_per_second": 126.741,
12
+ "train_loss": 0.06623676680927928,
13
+ "train_runtime": 486.7718,
14
+ "train_samples": 102018,
15
+ "train_samples_per_second": 628.742,
16
+ "train_steps_per_second": 13.103
17
+ }
config.json ADDED
@@ -0,0 +1,46 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "microsoft/deberta-v3-base",
3
+ "architectures": [
4
+ "DebertaV2ForTokenClassification"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.1,
7
+ "finetuning_task": "ner",
8
+ "hidden_act": "gelu",
9
+ "hidden_dropout_prob": 0.1,
10
+ "hidden_size": 768,
11
+ "id2label": {
12
+ "0": "O",
13
+ "1": "B-ORG",
14
+ "2": "I-ORG"
15
+ },
16
+ "initializer_range": 0.02,
17
+ "intermediate_size": 3072,
18
+ "label2id": {
19
+ "B-ORG": 1,
20
+ "I-ORG": 2,
21
+ "O": 0
22
+ },
23
+ "layer_norm_eps": 1e-07,
24
+ "max_position_embeddings": 512,
25
+ "max_relative_positions": -1,
26
+ "model_type": "deberta-v2",
27
+ "norm_rel_ebd": "layer_norm",
28
+ "num_attention_heads": 12,
29
+ "num_hidden_layers": 12,
30
+ "pad_token_id": 0,
31
+ "pooler_dropout": 0,
32
+ "pooler_hidden_act": "gelu",
33
+ "pooler_hidden_size": 768,
34
+ "pos_att_type": [
35
+ "p2c",
36
+ "c2p"
37
+ ],
38
+ "position_biased_input": false,
39
+ "position_buckets": 256,
40
+ "relative_attention": true,
41
+ "share_att_key": true,
42
+ "torch_dtype": "float32",
43
+ "transformers_version": "4.34.1",
44
+ "type_vocab_size": 0,
45
+ "vocab_size": 128100
46
+ }
eval_results.json ADDED
@@ -0,0 +1,12 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 3.0,
3
+ "eval_accuracy": 0.9769126125154315,
4
+ "eval_f1": 0.7849694196330357,
5
+ "eval_loss": 0.06933891773223877,
6
+ "eval_precision": 0.7739696312364425,
7
+ "eval_recall": 0.7962863774326013,
8
+ "eval_runtime": 13.9655,
9
+ "eval_samples": 14160,
10
+ "eval_samples_per_second": 1013.925,
11
+ "eval_steps_per_second": 126.741
12
+ }
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e02340e7d410ab34158a51b7c343233adc46cde5dbad9b0afd48987f063b4077
3
+ size 735404397
special_tokens_map.json ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "[CLS]",
3
+ "cls_token": "[CLS]",
4
+ "eos_token": "[SEP]",
5
+ "mask_token": "[MASK]",
6
+ "pad_token": "[PAD]",
7
+ "sep_token": "[SEP]",
8
+ "unk_token": {
9
+ "content": "[UNK]",
10
+ "lstrip": false,
11
+ "normalized": true,
12
+ "rstrip": false,
13
+ "single_word": false
14
+ }
15
+ }
spm.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c679fbf93643d19aab7ee10c0b99e460bdbc02fedf34b92b05af343b4af586fd
3
+ size 2464616
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer_config.json ADDED
@@ -0,0 +1,58 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "[PAD]",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "1": {
12
+ "content": "[CLS]",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "2": {
20
+ "content": "[SEP]",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "3": {
28
+ "content": "[UNK]",
29
+ "lstrip": false,
30
+ "normalized": true,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "128000": {
36
+ "content": "[MASK]",
37
+ "lstrip": false,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ }
43
+ },
44
+ "bos_token": "[CLS]",
45
+ "clean_up_tokenization_spaces": true,
46
+ "cls_token": "[CLS]",
47
+ "do_lower_case": false,
48
+ "eos_token": "[SEP]",
49
+ "mask_token": "[MASK]",
50
+ "model_max_length": 1000000000000000019884624838656,
51
+ "pad_token": "[PAD]",
52
+ "sep_token": "[SEP]",
53
+ "sp_model_kwargs": {},
54
+ "split_by_punct": false,
55
+ "tokenizer_class": "DebertaV2Tokenizer",
56
+ "unk_token": "[UNK]",
57
+ "vocab_type": "spm"
58
+ }
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 3.0,
3
+ "train_loss": 0.06623676680927928,
4
+ "train_runtime": 486.7718,
5
+ "train_samples": 102018,
6
+ "train_samples_per_second": 628.742,
7
+ "train_steps_per_second": 13.103
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,3886 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 3.0,
5
+ "eval_steps": 500,
6
+ "global_step": 6378,
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": 8.77742946708464e-07,
14
+ "loss": 1.1515,
15
+ "step": 10
16
+ },
17
+ {
18
+ "epoch": 0.01,
19
+ "learning_rate": 2.1316614420062697e-06,
20
+ "loss": 1.0942,
21
+ "step": 20
22
+ },
23
+ {
24
+ "epoch": 0.01,
25
+ "learning_rate": 3.3855799373040753e-06,
26
+ "loss": 0.825,
27
+ "step": 30
28
+ },
29
+ {
30
+ "epoch": 0.02,
31
+ "learning_rate": 4.639498432601881e-06,
32
+ "loss": 0.5304,
33
+ "step": 40
34
+ },
35
+ {
36
+ "epoch": 0.02,
37
+ "learning_rate": 5.8934169278996865e-06,
38
+ "loss": 0.3473,
39
+ "step": 50
40
+ },
41
+ {
42
+ "epoch": 0.03,
43
+ "learning_rate": 7.147335423197492e-06,
44
+ "loss": 0.2823,
45
+ "step": 60
46
+ },
47
+ {
48
+ "epoch": 0.03,
49
+ "learning_rate": 8.4012539184953e-06,
50
+ "loss": 0.2224,
51
+ "step": 70
52
+ },
53
+ {
54
+ "epoch": 0.04,
55
+ "learning_rate": 9.655172413793105e-06,
56
+ "loss": 0.1987,
57
+ "step": 80
58
+ },
59
+ {
60
+ "epoch": 0.04,
61
+ "learning_rate": 1.0909090909090909e-05,
62
+ "loss": 0.1516,
63
+ "step": 90
64
+ },
65
+ {
66
+ "epoch": 0.05,
67
+ "learning_rate": 1.2163009404388715e-05,
68
+ "loss": 0.1241,
69
+ "step": 100
70
+ },
71
+ {
72
+ "epoch": 0.05,
73
+ "learning_rate": 1.341692789968652e-05,
74
+ "loss": 0.1199,
75
+ "step": 110
76
+ },
77
+ {
78
+ "epoch": 0.06,
79
+ "learning_rate": 1.4670846394984329e-05,
80
+ "loss": 0.1113,
81
+ "step": 120
82
+ },
83
+ {
84
+ "epoch": 0.06,
85
+ "learning_rate": 1.5924764890282133e-05,
86
+ "loss": 0.105,
87
+ "step": 130
88
+ },
89
+ {
90
+ "epoch": 0.07,
91
+ "learning_rate": 1.717868338557994e-05,
92
+ "loss": 0.0879,
93
+ "step": 140
94
+ },
95
+ {
96
+ "epoch": 0.07,
97
+ "learning_rate": 1.8432601880877744e-05,
98
+ "loss": 0.105,
99
+ "step": 150
100
+ },
101
+ {
102
+ "epoch": 0.08,
103
+ "learning_rate": 1.968652037617555e-05,
104
+ "loss": 0.1043,
105
+ "step": 160
106
+ },
107
+ {
108
+ "epoch": 0.08,
109
+ "learning_rate": 2.0940438871473355e-05,
110
+ "loss": 0.1107,
111
+ "step": 170
112
+ },
113
+ {
114
+ "epoch": 0.08,
115
+ "learning_rate": 2.2194357366771163e-05,
116
+ "loss": 0.1098,
117
+ "step": 180
118
+ },
119
+ {
120
+ "epoch": 0.09,
121
+ "learning_rate": 2.3448275862068967e-05,
122
+ "loss": 0.0981,
123
+ "step": 190
124
+ },
125
+ {
126
+ "epoch": 0.09,
127
+ "learning_rate": 2.4702194357366774e-05,
128
+ "loss": 0.0919,
129
+ "step": 200
130
+ },
131
+ {
132
+ "epoch": 0.1,
133
+ "learning_rate": 2.595611285266458e-05,
134
+ "loss": 0.0782,
135
+ "step": 210
136
+ },
137
+ {
138
+ "epoch": 0.1,
139
+ "learning_rate": 2.7210031347962385e-05,
140
+ "loss": 0.0843,
141
+ "step": 220
142
+ },
143
+ {
144
+ "epoch": 0.11,
145
+ "learning_rate": 2.8463949843260192e-05,
146
+ "loss": 0.1103,
147
+ "step": 230
148
+ },
149
+ {
150
+ "epoch": 0.11,
151
+ "learning_rate": 2.9717868338557996e-05,
152
+ "loss": 0.0937,
153
+ "step": 240
154
+ },
155
+ {
156
+ "epoch": 0.12,
157
+ "learning_rate": 3.097178683385581e-05,
158
+ "loss": 0.092,
159
+ "step": 250
160
+ },
161
+ {
162
+ "epoch": 0.12,
163
+ "learning_rate": 3.222570532915361e-05,
164
+ "loss": 0.0871,
165
+ "step": 260
166
+ },
167
+ {
168
+ "epoch": 0.13,
169
+ "learning_rate": 3.3479623824451415e-05,
170
+ "loss": 0.0908,
171
+ "step": 270
172
+ },
173
+ {
174
+ "epoch": 0.13,
175
+ "learning_rate": 3.4733542319749215e-05,
176
+ "loss": 0.0849,
177
+ "step": 280
178
+ },
179
+ {
180
+ "epoch": 0.14,
181
+ "learning_rate": 3.598746081504703e-05,
182
+ "loss": 0.0972,
183
+ "step": 290
184
+ },
185
+ {
186
+ "epoch": 0.14,
187
+ "learning_rate": 3.724137931034483e-05,
188
+ "loss": 0.0819,
189
+ "step": 300
190
+ },
191
+ {
192
+ "epoch": 0.15,
193
+ "learning_rate": 3.849529780564264e-05,
194
+ "loss": 0.085,
195
+ "step": 310
196
+ },
197
+ {
198
+ "epoch": 0.15,
199
+ "learning_rate": 3.974921630094044e-05,
200
+ "loss": 0.0736,
201
+ "step": 320
202
+ },
203
+ {
204
+ "epoch": 0.16,
205
+ "learning_rate": 4.1003134796238245e-05,
206
+ "loss": 0.0918,
207
+ "step": 330
208
+ },
209
+ {
210
+ "epoch": 0.16,
211
+ "learning_rate": 4.225705329153606e-05,
212
+ "loss": 0.087,
213
+ "step": 340
214
+ },
215
+ {
216
+ "epoch": 0.16,
217
+ "learning_rate": 4.351097178683386e-05,
218
+ "loss": 0.0958,
219
+ "step": 350
220
+ },
221
+ {
222
+ "epoch": 0.17,
223
+ "learning_rate": 4.476489028213166e-05,
224
+ "loss": 0.1006,
225
+ "step": 360
226
+ },
227
+ {
228
+ "epoch": 0.17,
229
+ "learning_rate": 4.601880877742947e-05,
230
+ "loss": 0.0849,
231
+ "step": 370
232
+ },
233
+ {
234
+ "epoch": 0.18,
235
+ "learning_rate": 4.727272727272728e-05,
236
+ "loss": 0.0772,
237
+ "step": 380
238
+ },
239
+ {
240
+ "epoch": 0.18,
241
+ "learning_rate": 4.852664576802508e-05,
242
+ "loss": 0.0852,
243
+ "step": 390
244
+ },
245
+ {
246
+ "epoch": 0.19,
247
+ "learning_rate": 4.978056426332288e-05,
248
+ "loss": 0.0865,
249
+ "step": 400
250
+ },
251
+ {
252
+ "epoch": 0.19,
253
+ "learning_rate": 5.10344827586207e-05,
254
+ "loss": 0.0936,
255
+ "step": 410
256
+ },
257
+ {
258
+ "epoch": 0.2,
259
+ "learning_rate": 5.2288401253918504e-05,
260
+ "loss": 0.1028,
261
+ "step": 420
262
+ },
263
+ {
264
+ "epoch": 0.2,
265
+ "learning_rate": 5.3542319749216304e-05,
266
+ "loss": 0.0849,
267
+ "step": 430
268
+ },
269
+ {
270
+ "epoch": 0.21,
271
+ "learning_rate": 5.4796238244514105e-05,
272
+ "loss": 0.0935,
273
+ "step": 440
274
+ },
275
+ {
276
+ "epoch": 0.21,
277
+ "learning_rate": 5.605015673981192e-05,
278
+ "loss": 0.0838,
279
+ "step": 450
280
+ },
281
+ {
282
+ "epoch": 0.22,
283
+ "learning_rate": 5.7304075235109726e-05,
284
+ "loss": 0.0851,
285
+ "step": 460
286
+ },
287
+ {
288
+ "epoch": 0.22,
289
+ "learning_rate": 5.855799373040753e-05,
290
+ "loss": 0.0916,
291
+ "step": 470
292
+ },
293
+ {
294
+ "epoch": 0.23,
295
+ "learning_rate": 5.9811912225705334e-05,
296
+ "loss": 0.0926,
297
+ "step": 480
298
+ },
299
+ {
300
+ "epoch": 0.23,
301
+ "learning_rate": 6.106583072100315e-05,
302
+ "loss": 0.0742,
303
+ "step": 490
304
+ },
305
+ {
306
+ "epoch": 0.24,
307
+ "learning_rate": 6.231974921630095e-05,
308
+ "loss": 0.0882,
309
+ "step": 500
310
+ },
311
+ {
312
+ "epoch": 0.24,
313
+ "learning_rate": 6.357366771159875e-05,
314
+ "loss": 0.0789,
315
+ "step": 510
316
+ },
317
+ {
318
+ "epoch": 0.24,
319
+ "learning_rate": 6.482758620689655e-05,
320
+ "loss": 0.0847,
321
+ "step": 520
322
+ },
323
+ {
324
+ "epoch": 0.25,
325
+ "learning_rate": 6.608150470219436e-05,
326
+ "loss": 0.0764,
327
+ "step": 530
328
+ },
329
+ {
330
+ "epoch": 0.25,
331
+ "learning_rate": 6.733542319749216e-05,
332
+ "loss": 0.0963,
333
+ "step": 540
334
+ },
335
+ {
336
+ "epoch": 0.26,
337
+ "learning_rate": 6.858934169278998e-05,
338
+ "loss": 0.0695,
339
+ "step": 550
340
+ },
341
+ {
342
+ "epoch": 0.26,
343
+ "learning_rate": 6.984326018808778e-05,
344
+ "loss": 0.0693,
345
+ "step": 560
346
+ },
347
+ {
348
+ "epoch": 0.27,
349
+ "learning_rate": 7.109717868338559e-05,
350
+ "loss": 0.0908,
351
+ "step": 570
352
+ },
353
+ {
354
+ "epoch": 0.27,
355
+ "learning_rate": 7.23510971786834e-05,
356
+ "loss": 0.0779,
357
+ "step": 580
358
+ },
359
+ {
360
+ "epoch": 0.28,
361
+ "learning_rate": 7.36050156739812e-05,
362
+ "loss": 0.0825,
363
+ "step": 590
364
+ },
365
+ {
366
+ "epoch": 0.28,
367
+ "learning_rate": 7.485893416927901e-05,
368
+ "loss": 0.0813,
369
+ "step": 600
370
+ },
371
+ {
372
+ "epoch": 0.29,
373
+ "learning_rate": 7.611285266457681e-05,
374
+ "loss": 0.072,
375
+ "step": 610
376
+ },
377
+ {
378
+ "epoch": 0.29,
379
+ "learning_rate": 7.736677115987461e-05,
380
+ "loss": 0.0813,
381
+ "step": 620
382
+ },
383
+ {
384
+ "epoch": 0.3,
385
+ "learning_rate": 7.862068965517242e-05,
386
+ "loss": 0.0906,
387
+ "step": 630
388
+ },
389
+ {
390
+ "epoch": 0.3,
391
+ "learning_rate": 7.987460815047022e-05,
392
+ "loss": 0.0868,
393
+ "step": 640
394
+ },
395
+ {
396
+ "epoch": 0.31,
397
+ "learning_rate": 7.987456445993032e-05,
398
+ "loss": 0.0675,
399
+ "step": 650
400
+ },
401
+ {
402
+ "epoch": 0.31,
403
+ "learning_rate": 7.973519163763066e-05,
404
+ "loss": 0.0942,
405
+ "step": 660
406
+ },
407
+ {
408
+ "epoch": 0.32,
409
+ "learning_rate": 7.959581881533101e-05,
410
+ "loss": 0.0671,
411
+ "step": 670
412
+ },
413
+ {
414
+ "epoch": 0.32,
415
+ "learning_rate": 7.945644599303136e-05,
416
+ "loss": 0.0841,
417
+ "step": 680
418
+ },
419
+ {
420
+ "epoch": 0.32,
421
+ "learning_rate": 7.931707317073171e-05,
422
+ "loss": 0.0778,
423
+ "step": 690
424
+ },
425
+ {
426
+ "epoch": 0.33,
427
+ "learning_rate": 7.917770034843206e-05,
428
+ "loss": 0.0899,
429
+ "step": 700
430
+ },
431
+ {
432
+ "epoch": 0.33,
433
+ "learning_rate": 7.90383275261324e-05,
434
+ "loss": 0.0925,
435
+ "step": 710
436
+ },
437
+ {
438
+ "epoch": 0.34,
439
+ "learning_rate": 7.889895470383276e-05,
440
+ "loss": 0.0822,
441
+ "step": 720
442
+ },
443
+ {
444
+ "epoch": 0.34,
445
+ "learning_rate": 7.87595818815331e-05,
446
+ "loss": 0.0837,
447
+ "step": 730
448
+ },
449
+ {
450
+ "epoch": 0.35,
451
+ "learning_rate": 7.862020905923346e-05,
452
+ "loss": 0.077,
453
+ "step": 740
454
+ },
455
+ {
456
+ "epoch": 0.35,
457
+ "learning_rate": 7.84808362369338e-05,
458
+ "loss": 0.0705,
459
+ "step": 750
460
+ },
461
+ {
462
+ "epoch": 0.36,
463
+ "learning_rate": 7.834146341463415e-05,
464
+ "loss": 0.0849,
465
+ "step": 760
466
+ },
467
+ {
468
+ "epoch": 0.36,
469
+ "learning_rate": 7.82020905923345e-05,
470
+ "loss": 0.0792,
471
+ "step": 770
472
+ },
473
+ {
474
+ "epoch": 0.37,
475
+ "learning_rate": 7.806271777003485e-05,
476
+ "loss": 0.0786,
477
+ "step": 780
478
+ },
479
+ {
480
+ "epoch": 0.37,
481
+ "learning_rate": 7.79233449477352e-05,
482
+ "loss": 0.0862,
483
+ "step": 790
484
+ },
485
+ {
486
+ "epoch": 0.38,
487
+ "learning_rate": 7.778397212543555e-05,
488
+ "loss": 0.0828,
489
+ "step": 800
490
+ },
491
+ {
492
+ "epoch": 0.38,
493
+ "learning_rate": 7.76445993031359e-05,
494
+ "loss": 0.0918,
495
+ "step": 810
496
+ },
497
+ {
498
+ "epoch": 0.39,
499
+ "learning_rate": 7.750522648083624e-05,
500
+ "loss": 0.0749,
501
+ "step": 820
502
+ },
503
+ {
504
+ "epoch": 0.39,
505
+ "learning_rate": 7.736585365853659e-05,
506
+ "loss": 0.083,
507
+ "step": 830
508
+ },
509
+ {
510
+ "epoch": 0.4,
511
+ "learning_rate": 7.722648083623694e-05,
512
+ "loss": 0.0666,
513
+ "step": 840
514
+ },
515
+ {
516
+ "epoch": 0.4,
517
+ "learning_rate": 7.708710801393729e-05,
518
+ "loss": 0.0811,
519
+ "step": 850
520
+ },
521
+ {
522
+ "epoch": 0.4,
523
+ "learning_rate": 7.694773519163764e-05,
524
+ "loss": 0.0877,
525
+ "step": 860
526
+ },
527
+ {
528
+ "epoch": 0.41,
529
+ "learning_rate": 7.680836236933799e-05,
530
+ "loss": 0.0825,
531
+ "step": 870
532
+ },
533
+ {
534
+ "epoch": 0.41,
535
+ "learning_rate": 7.666898954703834e-05,
536
+ "loss": 0.0759,
537
+ "step": 880
538
+ },
539
+ {
540
+ "epoch": 0.42,
541
+ "learning_rate": 7.652961672473867e-05,
542
+ "loss": 0.0723,
543
+ "step": 890
544
+ },
545
+ {
546
+ "epoch": 0.42,
547
+ "learning_rate": 7.639024390243902e-05,
548
+ "loss": 0.0763,
549
+ "step": 900
550
+ },
551
+ {
552
+ "epoch": 0.43,
553
+ "learning_rate": 7.625087108013937e-05,
554
+ "loss": 0.0716,
555
+ "step": 910
556
+ },
557
+ {
558
+ "epoch": 0.43,
559
+ "learning_rate": 7.611149825783972e-05,
560
+ "loss": 0.0698,
561
+ "step": 920
562
+ },
563
+ {
564
+ "epoch": 0.44,
565
+ "learning_rate": 7.597212543554007e-05,
566
+ "loss": 0.0873,
567
+ "step": 930
568
+ },
569
+ {
570
+ "epoch": 0.44,
571
+ "learning_rate": 7.583275261324042e-05,
572
+ "loss": 0.0701,
573
+ "step": 940
574
+ },
575
+ {
576
+ "epoch": 0.45,
577
+ "learning_rate": 7.569337979094077e-05,
578
+ "loss": 0.0775,
579
+ "step": 950
580
+ },
581
+ {
582
+ "epoch": 0.45,
583
+ "learning_rate": 7.555400696864112e-05,
584
+ "loss": 0.0705,
585
+ "step": 960
586
+ },
587
+ {
588
+ "epoch": 0.46,
589
+ "learning_rate": 7.541463414634147e-05,
590
+ "loss": 0.0792,
591
+ "step": 970
592
+ },
593
+ {
594
+ "epoch": 0.46,
595
+ "learning_rate": 7.527526132404182e-05,
596
+ "loss": 0.0695,
597
+ "step": 980
598
+ },
599
+ {
600
+ "epoch": 0.47,
601
+ "learning_rate": 7.513588850174217e-05,
602
+ "loss": 0.0934,
603
+ "step": 990
604
+ },
605
+ {
606
+ "epoch": 0.47,
607
+ "learning_rate": 7.499651567944252e-05,
608
+ "loss": 0.0903,
609
+ "step": 1000
610
+ },
611
+ {
612
+ "epoch": 0.48,
613
+ "learning_rate": 7.485714285714287e-05,
614
+ "loss": 0.0863,
615
+ "step": 1010
616
+ },
617
+ {
618
+ "epoch": 0.48,
619
+ "learning_rate": 7.471777003484322e-05,
620
+ "loss": 0.0801,
621
+ "step": 1020
622
+ },
623
+ {
624
+ "epoch": 0.48,
625
+ "learning_rate": 7.457839721254357e-05,
626
+ "loss": 0.0744,
627
+ "step": 1030
628
+ },
629
+ {
630
+ "epoch": 0.49,
631
+ "learning_rate": 7.443902439024392e-05,
632
+ "loss": 0.075,
633
+ "step": 1040
634
+ },
635
+ {
636
+ "epoch": 0.49,
637
+ "learning_rate": 7.429965156794425e-05,
638
+ "loss": 0.0711,
639
+ "step": 1050
640
+ },
641
+ {
642
+ "epoch": 0.5,
643
+ "learning_rate": 7.41602787456446e-05,
644
+ "loss": 0.0904,
645
+ "step": 1060
646
+ },
647
+ {
648
+ "epoch": 0.5,
649
+ "learning_rate": 7.402090592334495e-05,
650
+ "loss": 0.0771,
651
+ "step": 1070
652
+ },
653
+ {
654
+ "epoch": 0.51,
655
+ "learning_rate": 7.38815331010453e-05,
656
+ "loss": 0.0707,
657
+ "step": 1080
658
+ },
659
+ {
660
+ "epoch": 0.51,
661
+ "learning_rate": 7.374216027874565e-05,
662
+ "loss": 0.0655,
663
+ "step": 1090
664
+ },
665
+ {
666
+ "epoch": 0.52,
667
+ "learning_rate": 7.3602787456446e-05,
668
+ "loss": 0.077,
669
+ "step": 1100
670
+ },
671
+ {
672
+ "epoch": 0.52,
673
+ "learning_rate": 7.346341463414635e-05,
674
+ "loss": 0.0775,
675
+ "step": 1110
676
+ },
677
+ {
678
+ "epoch": 0.53,
679
+ "learning_rate": 7.33240418118467e-05,
680
+ "loss": 0.0825,
681
+ "step": 1120
682
+ },
683
+ {
684
+ "epoch": 0.53,
685
+ "learning_rate": 7.318466898954704e-05,
686
+ "loss": 0.0704,
687
+ "step": 1130
688
+ },
689
+ {
690
+ "epoch": 0.54,
691
+ "learning_rate": 7.304529616724739e-05,
692
+ "loss": 0.0614,
693
+ "step": 1140
694
+ },
695
+ {
696
+ "epoch": 0.54,
697
+ "learning_rate": 7.290592334494774e-05,
698
+ "loss": 0.0685,
699
+ "step": 1150
700
+ },
701
+ {
702
+ "epoch": 0.55,
703
+ "learning_rate": 7.276655052264809e-05,
704
+ "loss": 0.0766,
705
+ "step": 1160
706
+ },
707
+ {
708
+ "epoch": 0.55,
709
+ "learning_rate": 7.262717770034844e-05,
710
+ "loss": 0.0746,
711
+ "step": 1170
712
+ },
713
+ {
714
+ "epoch": 0.56,
715
+ "learning_rate": 7.248780487804878e-05,
716
+ "loss": 0.0723,
717
+ "step": 1180
718
+ },
719
+ {
720
+ "epoch": 0.56,
721
+ "learning_rate": 7.234843205574913e-05,
722
+ "loss": 0.0824,
723
+ "step": 1190
724
+ },
725
+ {
726
+ "epoch": 0.56,
727
+ "learning_rate": 7.220905923344948e-05,
728
+ "loss": 0.0708,
729
+ "step": 1200
730
+ },
731
+ {
732
+ "epoch": 0.57,
733
+ "learning_rate": 7.206968641114983e-05,
734
+ "loss": 0.0745,
735
+ "step": 1210
736
+ },
737
+ {
738
+ "epoch": 0.57,
739
+ "learning_rate": 7.193031358885018e-05,
740
+ "loss": 0.0743,
741
+ "step": 1220
742
+ },
743
+ {
744
+ "epoch": 0.58,
745
+ "learning_rate": 7.179094076655053e-05,
746
+ "loss": 0.0792,
747
+ "step": 1230
748
+ },
749
+ {
750
+ "epoch": 0.58,
751
+ "learning_rate": 7.165156794425088e-05,
752
+ "loss": 0.0729,
753
+ "step": 1240
754
+ },
755
+ {
756
+ "epoch": 0.59,
757
+ "learning_rate": 7.151219512195123e-05,
758
+ "loss": 0.0822,
759
+ "step": 1250
760
+ },
761
+ {
762
+ "epoch": 0.59,
763
+ "learning_rate": 7.137282229965158e-05,
764
+ "loss": 0.0766,
765
+ "step": 1260
766
+ },
767
+ {
768
+ "epoch": 0.6,
769
+ "learning_rate": 7.123344947735193e-05,
770
+ "loss": 0.07,
771
+ "step": 1270
772
+ },
773
+ {
774
+ "epoch": 0.6,
775
+ "learning_rate": 7.109407665505227e-05,
776
+ "loss": 0.0818,
777
+ "step": 1280
778
+ },
779
+ {
780
+ "epoch": 0.61,
781
+ "learning_rate": 7.095470383275262e-05,
782
+ "loss": 0.0748,
783
+ "step": 1290
784
+ },
785
+ {
786
+ "epoch": 0.61,
787
+ "learning_rate": 7.081533101045297e-05,
788
+ "loss": 0.0727,
789
+ "step": 1300
790
+ },
791
+ {
792
+ "epoch": 0.62,
793
+ "learning_rate": 7.067595818815332e-05,
794
+ "loss": 0.0798,
795
+ "step": 1310
796
+ },
797
+ {
798
+ "epoch": 0.62,
799
+ "learning_rate": 7.053658536585367e-05,
800
+ "loss": 0.0736,
801
+ "step": 1320
802
+ },
803
+ {
804
+ "epoch": 0.63,
805
+ "learning_rate": 7.039721254355402e-05,
806
+ "loss": 0.065,
807
+ "step": 1330
808
+ },
809
+ {
810
+ "epoch": 0.63,
811
+ "learning_rate": 7.025783972125437e-05,
812
+ "loss": 0.0699,
813
+ "step": 1340
814
+ },
815
+ {
816
+ "epoch": 0.63,
817
+ "learning_rate": 7.011846689895471e-05,
818
+ "loss": 0.066,
819
+ "step": 1350
820
+ },
821
+ {
822
+ "epoch": 0.64,
823
+ "learning_rate": 6.997909407665505e-05,
824
+ "loss": 0.0703,
825
+ "step": 1360
826
+ },
827
+ {
828
+ "epoch": 0.64,
829
+ "learning_rate": 6.98397212543554e-05,
830
+ "loss": 0.0801,
831
+ "step": 1370
832
+ },
833
+ {
834
+ "epoch": 0.65,
835
+ "learning_rate": 6.970034843205575e-05,
836
+ "loss": 0.0749,
837
+ "step": 1380
838
+ },
839
+ {
840
+ "epoch": 0.65,
841
+ "learning_rate": 6.95609756097561e-05,
842
+ "loss": 0.0658,
843
+ "step": 1390
844
+ },
845
+ {
846
+ "epoch": 0.66,
847
+ "learning_rate": 6.942160278745645e-05,
848
+ "loss": 0.0607,
849
+ "step": 1400
850
+ },
851
+ {
852
+ "epoch": 0.66,
853
+ "learning_rate": 6.92822299651568e-05,
854
+ "loss": 0.07,
855
+ "step": 1410
856
+ },
857
+ {
858
+ "epoch": 0.67,
859
+ "learning_rate": 6.914285714285715e-05,
860
+ "loss": 0.0747,
861
+ "step": 1420
862
+ },
863
+ {
864
+ "epoch": 0.67,
865
+ "learning_rate": 6.90034843205575e-05,
866
+ "loss": 0.0731,
867
+ "step": 1430
868
+ },
869
+ {
870
+ "epoch": 0.68,
871
+ "learning_rate": 6.886411149825785e-05,
872
+ "loss": 0.0703,
873
+ "step": 1440
874
+ },
875
+ {
876
+ "epoch": 0.68,
877
+ "learning_rate": 6.87247386759582e-05,
878
+ "loss": 0.0634,
879
+ "step": 1450
880
+ },
881
+ {
882
+ "epoch": 0.69,
883
+ "learning_rate": 6.858536585365855e-05,
884
+ "loss": 0.0748,
885
+ "step": 1460
886
+ },
887
+ {
888
+ "epoch": 0.69,
889
+ "learning_rate": 6.84459930313589e-05,
890
+ "loss": 0.0797,
891
+ "step": 1470
892
+ },
893
+ {
894
+ "epoch": 0.7,
895
+ "learning_rate": 6.830662020905925e-05,
896
+ "loss": 0.0899,
897
+ "step": 1480
898
+ },
899
+ {
900
+ "epoch": 0.7,
901
+ "learning_rate": 6.81672473867596e-05,
902
+ "loss": 0.0904,
903
+ "step": 1490
904
+ },
905
+ {
906
+ "epoch": 0.71,
907
+ "learning_rate": 6.802787456445995e-05,
908
+ "loss": 0.0723,
909
+ "step": 1500
910
+ },
911
+ {
912
+ "epoch": 0.71,
913
+ "learning_rate": 6.788850174216028e-05,
914
+ "loss": 0.0623,
915
+ "step": 1510
916
+ },
917
+ {
918
+ "epoch": 0.71,
919
+ "learning_rate": 6.774912891986063e-05,
920
+ "loss": 0.0698,
921
+ "step": 1520
922
+ },
923
+ {
924
+ "epoch": 0.72,
925
+ "learning_rate": 6.760975609756098e-05,
926
+ "loss": 0.0739,
927
+ "step": 1530
928
+ },
929
+ {
930
+ "epoch": 0.72,
931
+ "learning_rate": 6.747038327526133e-05,
932
+ "loss": 0.0655,
933
+ "step": 1540
934
+ },
935
+ {
936
+ "epoch": 0.73,
937
+ "learning_rate": 6.733101045296168e-05,
938
+ "loss": 0.0869,
939
+ "step": 1550
940
+ },
941
+ {
942
+ "epoch": 0.73,
943
+ "learning_rate": 6.719163763066203e-05,
944
+ "loss": 0.0708,
945
+ "step": 1560
946
+ },
947
+ {
948
+ "epoch": 0.74,
949
+ "learning_rate": 6.705226480836238e-05,
950
+ "loss": 0.0661,
951
+ "step": 1570
952
+ },
953
+ {
954
+ "epoch": 0.74,
955
+ "learning_rate": 6.691289198606273e-05,
956
+ "loss": 0.0795,
957
+ "step": 1580
958
+ },
959
+ {
960
+ "epoch": 0.75,
961
+ "learning_rate": 6.677351916376307e-05,
962
+ "loss": 0.0694,
963
+ "step": 1590
964
+ },
965
+ {
966
+ "epoch": 0.75,
967
+ "learning_rate": 6.663414634146341e-05,
968
+ "loss": 0.0923,
969
+ "step": 1600
970
+ },
971
+ {
972
+ "epoch": 0.76,
973
+ "learning_rate": 6.649477351916376e-05,
974
+ "loss": 0.0765,
975
+ "step": 1610
976
+ },
977
+ {
978
+ "epoch": 0.76,
979
+ "learning_rate": 6.635540069686411e-05,
980
+ "loss": 0.0703,
981
+ "step": 1620
982
+ },
983
+ {
984
+ "epoch": 0.77,
985
+ "learning_rate": 6.621602787456446e-05,
986
+ "loss": 0.0757,
987
+ "step": 1630
988
+ },
989
+ {
990
+ "epoch": 0.77,
991
+ "learning_rate": 6.607665505226481e-05,
992
+ "loss": 0.0761,
993
+ "step": 1640
994
+ },
995
+ {
996
+ "epoch": 0.78,
997
+ "learning_rate": 6.593728222996516e-05,
998
+ "loss": 0.0721,
999
+ "step": 1650
1000
+ },
1001
+ {
1002
+ "epoch": 0.78,
1003
+ "learning_rate": 6.579790940766551e-05,
1004
+ "loss": 0.077,
1005
+ "step": 1660
1006
+ },
1007
+ {
1008
+ "epoch": 0.79,
1009
+ "learning_rate": 6.565853658536585e-05,
1010
+ "loss": 0.0598,
1011
+ "step": 1670
1012
+ },
1013
+ {
1014
+ "epoch": 0.79,
1015
+ "learning_rate": 6.55191637630662e-05,
1016
+ "loss": 0.0727,
1017
+ "step": 1680
1018
+ },
1019
+ {
1020
+ "epoch": 0.79,
1021
+ "learning_rate": 6.537979094076655e-05,
1022
+ "loss": 0.0827,
1023
+ "step": 1690
1024
+ },
1025
+ {
1026
+ "epoch": 0.8,
1027
+ "learning_rate": 6.52404181184669e-05,
1028
+ "loss": 0.0725,
1029
+ "step": 1700
1030
+ },
1031
+ {
1032
+ "epoch": 0.8,
1033
+ "learning_rate": 6.510104529616725e-05,
1034
+ "loss": 0.0749,
1035
+ "step": 1710
1036
+ },
1037
+ {
1038
+ "epoch": 0.81,
1039
+ "learning_rate": 6.49616724738676e-05,
1040
+ "loss": 0.0773,
1041
+ "step": 1720
1042
+ },
1043
+ {
1044
+ "epoch": 0.81,
1045
+ "learning_rate": 6.482229965156795e-05,
1046
+ "loss": 0.081,
1047
+ "step": 1730
1048
+ },
1049
+ {
1050
+ "epoch": 0.82,
1051
+ "learning_rate": 6.46829268292683e-05,
1052
+ "loss": 0.0824,
1053
+ "step": 1740
1054
+ },
1055
+ {
1056
+ "epoch": 0.82,
1057
+ "learning_rate": 6.454355400696865e-05,
1058
+ "loss": 0.0662,
1059
+ "step": 1750
1060
+ },
1061
+ {
1062
+ "epoch": 0.83,
1063
+ "learning_rate": 6.4404181184669e-05,
1064
+ "loss": 0.0548,
1065
+ "step": 1760
1066
+ },
1067
+ {
1068
+ "epoch": 0.83,
1069
+ "learning_rate": 6.426480836236935e-05,
1070
+ "loss": 0.0847,
1071
+ "step": 1770
1072
+ },
1073
+ {
1074
+ "epoch": 0.84,
1075
+ "learning_rate": 6.41254355400697e-05,
1076
+ "loss": 0.0758,
1077
+ "step": 1780
1078
+ },
1079
+ {
1080
+ "epoch": 0.84,
1081
+ "learning_rate": 6.398606271777004e-05,
1082
+ "loss": 0.0636,
1083
+ "step": 1790
1084
+ },
1085
+ {
1086
+ "epoch": 0.85,
1087
+ "learning_rate": 6.38466898954704e-05,
1088
+ "loss": 0.0615,
1089
+ "step": 1800
1090
+ },
1091
+ {
1092
+ "epoch": 0.85,
1093
+ "learning_rate": 6.370731707317074e-05,
1094
+ "loss": 0.0672,
1095
+ "step": 1810
1096
+ },
1097
+ {
1098
+ "epoch": 0.86,
1099
+ "learning_rate": 6.356794425087109e-05,
1100
+ "loss": 0.0858,
1101
+ "step": 1820
1102
+ },
1103
+ {
1104
+ "epoch": 0.86,
1105
+ "learning_rate": 6.342857142857143e-05,
1106
+ "loss": 0.0699,
1107
+ "step": 1830
1108
+ },
1109
+ {
1110
+ "epoch": 0.87,
1111
+ "learning_rate": 6.328919860627178e-05,
1112
+ "loss": 0.0608,
1113
+ "step": 1840
1114
+ },
1115
+ {
1116
+ "epoch": 0.87,
1117
+ "learning_rate": 6.314982578397213e-05,
1118
+ "loss": 0.0749,
1119
+ "step": 1850
1120
+ },
1121
+ {
1122
+ "epoch": 0.87,
1123
+ "learning_rate": 6.301045296167248e-05,
1124
+ "loss": 0.083,
1125
+ "step": 1860
1126
+ },
1127
+ {
1128
+ "epoch": 0.88,
1129
+ "learning_rate": 6.287108013937283e-05,
1130
+ "loss": 0.0763,
1131
+ "step": 1870
1132
+ },
1133
+ {
1134
+ "epoch": 0.88,
1135
+ "learning_rate": 6.273170731707318e-05,
1136
+ "loss": 0.0634,
1137
+ "step": 1880
1138
+ },
1139
+ {
1140
+ "epoch": 0.89,
1141
+ "learning_rate": 6.259233449477353e-05,
1142
+ "loss": 0.0595,
1143
+ "step": 1890
1144
+ },
1145
+ {
1146
+ "epoch": 0.89,
1147
+ "learning_rate": 6.245296167247386e-05,
1148
+ "loss": 0.0555,
1149
+ "step": 1900
1150
+ },
1151
+ {
1152
+ "epoch": 0.9,
1153
+ "learning_rate": 6.231358885017421e-05,
1154
+ "loss": 0.0607,
1155
+ "step": 1910
1156
+ },
1157
+ {
1158
+ "epoch": 0.9,
1159
+ "learning_rate": 6.217421602787456e-05,
1160
+ "loss": 0.078,
1161
+ "step": 1920
1162
+ },
1163
+ {
1164
+ "epoch": 0.91,
1165
+ "learning_rate": 6.203484320557491e-05,
1166
+ "loss": 0.0894,
1167
+ "step": 1930
1168
+ },
1169
+ {
1170
+ "epoch": 0.91,
1171
+ "learning_rate": 6.189547038327526e-05,
1172
+ "loss": 0.0672,
1173
+ "step": 1940
1174
+ },
1175
+ {
1176
+ "epoch": 0.92,
1177
+ "learning_rate": 6.175609756097561e-05,
1178
+ "loss": 0.067,
1179
+ "step": 1950
1180
+ },
1181
+ {
1182
+ "epoch": 0.92,
1183
+ "learning_rate": 6.161672473867596e-05,
1184
+ "loss": 0.0733,
1185
+ "step": 1960
1186
+ },
1187
+ {
1188
+ "epoch": 0.93,
1189
+ "learning_rate": 6.147735191637631e-05,
1190
+ "loss": 0.0634,
1191
+ "step": 1970
1192
+ },
1193
+ {
1194
+ "epoch": 0.93,
1195
+ "learning_rate": 6.133797909407666e-05,
1196
+ "loss": 0.0795,
1197
+ "step": 1980
1198
+ },
1199
+ {
1200
+ "epoch": 0.94,
1201
+ "learning_rate": 6.119860627177701e-05,
1202
+ "loss": 0.0757,
1203
+ "step": 1990
1204
+ },
1205
+ {
1206
+ "epoch": 0.94,
1207
+ "learning_rate": 6.105923344947736e-05,
1208
+ "loss": 0.0769,
1209
+ "step": 2000
1210
+ },
1211
+ {
1212
+ "epoch": 0.95,
1213
+ "learning_rate": 6.09198606271777e-05,
1214
+ "loss": 0.0834,
1215
+ "step": 2010
1216
+ },
1217
+ {
1218
+ "epoch": 0.95,
1219
+ "learning_rate": 6.078048780487805e-05,
1220
+ "loss": 0.0655,
1221
+ "step": 2020
1222
+ },
1223
+ {
1224
+ "epoch": 0.95,
1225
+ "learning_rate": 6.06411149825784e-05,
1226
+ "loss": 0.0669,
1227
+ "step": 2030
1228
+ },
1229
+ {
1230
+ "epoch": 0.96,
1231
+ "learning_rate": 6.050174216027875e-05,
1232
+ "loss": 0.0723,
1233
+ "step": 2040
1234
+ },
1235
+ {
1236
+ "epoch": 0.96,
1237
+ "learning_rate": 6.03623693379791e-05,
1238
+ "loss": 0.0748,
1239
+ "step": 2050
1240
+ },
1241
+ {
1242
+ "epoch": 0.97,
1243
+ "learning_rate": 6.0222996515679443e-05,
1244
+ "loss": 0.077,
1245
+ "step": 2060
1246
+ },
1247
+ {
1248
+ "epoch": 0.97,
1249
+ "learning_rate": 6.008362369337979e-05,
1250
+ "loss": 0.0669,
1251
+ "step": 2070
1252
+ },
1253
+ {
1254
+ "epoch": 0.98,
1255
+ "learning_rate": 5.994425087108014e-05,
1256
+ "loss": 0.0818,
1257
+ "step": 2080
1258
+ },
1259
+ {
1260
+ "epoch": 0.98,
1261
+ "learning_rate": 5.980487804878049e-05,
1262
+ "loss": 0.0667,
1263
+ "step": 2090
1264
+ },
1265
+ {
1266
+ "epoch": 0.99,
1267
+ "learning_rate": 5.966550522648084e-05,
1268
+ "loss": 0.0674,
1269
+ "step": 2100
1270
+ },
1271
+ {
1272
+ "epoch": 0.99,
1273
+ "learning_rate": 5.952613240418119e-05,
1274
+ "loss": 0.0902,
1275
+ "step": 2110
1276
+ },
1277
+ {
1278
+ "epoch": 1.0,
1279
+ "learning_rate": 5.938675958188154e-05,
1280
+ "loss": 0.0752,
1281
+ "step": 2120
1282
+ },
1283
+ {
1284
+ "epoch": 1.0,
1285
+ "eval_accuracy": 0.975717042417269,
1286
+ "eval_f1": 0.7687279607809409,
1287
+ "eval_loss": 0.0664275586605072,
1288
+ "eval_precision": 0.7416196481911715,
1289
+ "eval_recall": 0.7978932333511873,
1290
+ "eval_runtime": 14.1977,
1291
+ "eval_samples_per_second": 997.344,
1292
+ "eval_steps_per_second": 124.668,
1293
+ "step": 2126
1294
+ },
1295
+ {
1296
+ "epoch": 1.0,
1297
+ "learning_rate": 5.924738675958189e-05,
1298
+ "loss": 0.0914,
1299
+ "step": 2130
1300
+ },
1301
+ {
1302
+ "epoch": 1.01,
1303
+ "learning_rate": 5.9108013937282234e-05,
1304
+ "loss": 0.0682,
1305
+ "step": 2140
1306
+ },
1307
+ {
1308
+ "epoch": 1.01,
1309
+ "learning_rate": 5.896864111498258e-05,
1310
+ "loss": 0.0591,
1311
+ "step": 2150
1312
+ },
1313
+ {
1314
+ "epoch": 1.02,
1315
+ "learning_rate": 5.882926829268293e-05,
1316
+ "loss": 0.0622,
1317
+ "step": 2160
1318
+ },
1319
+ {
1320
+ "epoch": 1.02,
1321
+ "learning_rate": 5.868989547038328e-05,
1322
+ "loss": 0.0692,
1323
+ "step": 2170
1324
+ },
1325
+ {
1326
+ "epoch": 1.03,
1327
+ "learning_rate": 5.855052264808363e-05,
1328
+ "loss": 0.054,
1329
+ "step": 2180
1330
+ },
1331
+ {
1332
+ "epoch": 1.03,
1333
+ "learning_rate": 5.841114982578398e-05,
1334
+ "loss": 0.0494,
1335
+ "step": 2190
1336
+ },
1337
+ {
1338
+ "epoch": 1.03,
1339
+ "learning_rate": 5.827177700348433e-05,
1340
+ "loss": 0.0648,
1341
+ "step": 2200
1342
+ },
1343
+ {
1344
+ "epoch": 1.04,
1345
+ "learning_rate": 5.813240418118467e-05,
1346
+ "loss": 0.0639,
1347
+ "step": 2210
1348
+ },
1349
+ {
1350
+ "epoch": 1.04,
1351
+ "learning_rate": 5.799303135888502e-05,
1352
+ "loss": 0.0595,
1353
+ "step": 2220
1354
+ },
1355
+ {
1356
+ "epoch": 1.05,
1357
+ "learning_rate": 5.785365853658537e-05,
1358
+ "loss": 0.0548,
1359
+ "step": 2230
1360
+ },
1361
+ {
1362
+ "epoch": 1.05,
1363
+ "learning_rate": 5.7714285714285716e-05,
1364
+ "loss": 0.0542,
1365
+ "step": 2240
1366
+ },
1367
+ {
1368
+ "epoch": 1.06,
1369
+ "learning_rate": 5.7574912891986066e-05,
1370
+ "loss": 0.0527,
1371
+ "step": 2250
1372
+ },
1373
+ {
1374
+ "epoch": 1.06,
1375
+ "learning_rate": 5.7435540069686416e-05,
1376
+ "loss": 0.0477,
1377
+ "step": 2260
1378
+ },
1379
+ {
1380
+ "epoch": 1.07,
1381
+ "learning_rate": 5.7296167247386765e-05,
1382
+ "loss": 0.0699,
1383
+ "step": 2270
1384
+ },
1385
+ {
1386
+ "epoch": 1.07,
1387
+ "learning_rate": 5.7156794425087115e-05,
1388
+ "loss": 0.0546,
1389
+ "step": 2280
1390
+ },
1391
+ {
1392
+ "epoch": 1.08,
1393
+ "learning_rate": 5.701742160278746e-05,
1394
+ "loss": 0.0543,
1395
+ "step": 2290
1396
+ },
1397
+ {
1398
+ "epoch": 1.08,
1399
+ "learning_rate": 5.687804878048781e-05,
1400
+ "loss": 0.052,
1401
+ "step": 2300
1402
+ },
1403
+ {
1404
+ "epoch": 1.09,
1405
+ "learning_rate": 5.673867595818816e-05,
1406
+ "loss": 0.0614,
1407
+ "step": 2310
1408
+ },
1409
+ {
1410
+ "epoch": 1.09,
1411
+ "learning_rate": 5.659930313588851e-05,
1412
+ "loss": 0.0692,
1413
+ "step": 2320
1414
+ },
1415
+ {
1416
+ "epoch": 1.1,
1417
+ "learning_rate": 5.6459930313588856e-05,
1418
+ "loss": 0.058,
1419
+ "step": 2330
1420
+ },
1421
+ {
1422
+ "epoch": 1.1,
1423
+ "learning_rate": 5.6320557491289206e-05,
1424
+ "loss": 0.064,
1425
+ "step": 2340
1426
+ },
1427
+ {
1428
+ "epoch": 1.11,
1429
+ "learning_rate": 5.6181184668989555e-05,
1430
+ "loss": 0.062,
1431
+ "step": 2350
1432
+ },
1433
+ {
1434
+ "epoch": 1.11,
1435
+ "learning_rate": 5.6041811846689905e-05,
1436
+ "loss": 0.0627,
1437
+ "step": 2360
1438
+ },
1439
+ {
1440
+ "epoch": 1.11,
1441
+ "learning_rate": 5.590243902439025e-05,
1442
+ "loss": 0.0613,
1443
+ "step": 2370
1444
+ },
1445
+ {
1446
+ "epoch": 1.12,
1447
+ "learning_rate": 5.57630662020906e-05,
1448
+ "loss": 0.0653,
1449
+ "step": 2380
1450
+ },
1451
+ {
1452
+ "epoch": 1.12,
1453
+ "learning_rate": 5.562369337979095e-05,
1454
+ "loss": 0.0592,
1455
+ "step": 2390
1456
+ },
1457
+ {
1458
+ "epoch": 1.13,
1459
+ "learning_rate": 5.54843205574913e-05,
1460
+ "loss": 0.0563,
1461
+ "step": 2400
1462
+ },
1463
+ {
1464
+ "epoch": 1.13,
1465
+ "learning_rate": 5.5344947735191646e-05,
1466
+ "loss": 0.0516,
1467
+ "step": 2410
1468
+ },
1469
+ {
1470
+ "epoch": 1.14,
1471
+ "learning_rate": 5.520557491289199e-05,
1472
+ "loss": 0.0619,
1473
+ "step": 2420
1474
+ },
1475
+ {
1476
+ "epoch": 1.14,
1477
+ "learning_rate": 5.506620209059234e-05,
1478
+ "loss": 0.0698,
1479
+ "step": 2430
1480
+ },
1481
+ {
1482
+ "epoch": 1.15,
1483
+ "learning_rate": 5.492682926829269e-05,
1484
+ "loss": 0.0541,
1485
+ "step": 2440
1486
+ },
1487
+ {
1488
+ "epoch": 1.15,
1489
+ "learning_rate": 5.478745644599303e-05,
1490
+ "loss": 0.0459,
1491
+ "step": 2450
1492
+ },
1493
+ {
1494
+ "epoch": 1.16,
1495
+ "learning_rate": 5.464808362369338e-05,
1496
+ "loss": 0.0544,
1497
+ "step": 2460
1498
+ },
1499
+ {
1500
+ "epoch": 1.16,
1501
+ "learning_rate": 5.450871080139373e-05,
1502
+ "loss": 0.0533,
1503
+ "step": 2470
1504
+ },
1505
+ {
1506
+ "epoch": 1.17,
1507
+ "learning_rate": 5.436933797909408e-05,
1508
+ "loss": 0.0597,
1509
+ "step": 2480
1510
+ },
1511
+ {
1512
+ "epoch": 1.17,
1513
+ "learning_rate": 5.422996515679443e-05,
1514
+ "loss": 0.0553,
1515
+ "step": 2490
1516
+ },
1517
+ {
1518
+ "epoch": 1.18,
1519
+ "learning_rate": 5.409059233449478e-05,
1520
+ "loss": 0.0583,
1521
+ "step": 2500
1522
+ },
1523
+ {
1524
+ "epoch": 1.18,
1525
+ "learning_rate": 5.395121951219513e-05,
1526
+ "loss": 0.0552,
1527
+ "step": 2510
1528
+ },
1529
+ {
1530
+ "epoch": 1.19,
1531
+ "learning_rate": 5.381184668989547e-05,
1532
+ "loss": 0.0582,
1533
+ "step": 2520
1534
+ },
1535
+ {
1536
+ "epoch": 1.19,
1537
+ "learning_rate": 5.367247386759582e-05,
1538
+ "loss": 0.059,
1539
+ "step": 2530
1540
+ },
1541
+ {
1542
+ "epoch": 1.19,
1543
+ "learning_rate": 5.353310104529617e-05,
1544
+ "loss": 0.0657,
1545
+ "step": 2540
1546
+ },
1547
+ {
1548
+ "epoch": 1.2,
1549
+ "learning_rate": 5.339372822299652e-05,
1550
+ "loss": 0.0563,
1551
+ "step": 2550
1552
+ },
1553
+ {
1554
+ "epoch": 1.2,
1555
+ "learning_rate": 5.325435540069687e-05,
1556
+ "loss": 0.0568,
1557
+ "step": 2560
1558
+ },
1559
+ {
1560
+ "epoch": 1.21,
1561
+ "learning_rate": 5.311498257839722e-05,
1562
+ "loss": 0.067,
1563
+ "step": 2570
1564
+ },
1565
+ {
1566
+ "epoch": 1.21,
1567
+ "learning_rate": 5.297560975609757e-05,
1568
+ "loss": 0.064,
1569
+ "step": 2580
1570
+ },
1571
+ {
1572
+ "epoch": 1.22,
1573
+ "learning_rate": 5.283623693379792e-05,
1574
+ "loss": 0.0546,
1575
+ "step": 2590
1576
+ },
1577
+ {
1578
+ "epoch": 1.22,
1579
+ "learning_rate": 5.2696864111498255e-05,
1580
+ "loss": 0.0575,
1581
+ "step": 2600
1582
+ },
1583
+ {
1584
+ "epoch": 1.23,
1585
+ "learning_rate": 5.2557491289198605e-05,
1586
+ "loss": 0.0497,
1587
+ "step": 2610
1588
+ },
1589
+ {
1590
+ "epoch": 1.23,
1591
+ "learning_rate": 5.2418118466898955e-05,
1592
+ "loss": 0.0616,
1593
+ "step": 2620
1594
+ },
1595
+ {
1596
+ "epoch": 1.24,
1597
+ "learning_rate": 5.2278745644599304e-05,
1598
+ "loss": 0.0498,
1599
+ "step": 2630
1600
+ },
1601
+ {
1602
+ "epoch": 1.24,
1603
+ "learning_rate": 5.2139372822299654e-05,
1604
+ "loss": 0.0622,
1605
+ "step": 2640
1606
+ },
1607
+ {
1608
+ "epoch": 1.25,
1609
+ "learning_rate": 5.2000000000000004e-05,
1610
+ "loss": 0.0542,
1611
+ "step": 2650
1612
+ },
1613
+ {
1614
+ "epoch": 1.25,
1615
+ "learning_rate": 5.186062717770035e-05,
1616
+ "loss": 0.0625,
1617
+ "step": 2660
1618
+ },
1619
+ {
1620
+ "epoch": 1.26,
1621
+ "learning_rate": 5.17212543554007e-05,
1622
+ "loss": 0.0494,
1623
+ "step": 2670
1624
+ },
1625
+ {
1626
+ "epoch": 1.26,
1627
+ "learning_rate": 5.1581881533101046e-05,
1628
+ "loss": 0.0511,
1629
+ "step": 2680
1630
+ },
1631
+ {
1632
+ "epoch": 1.27,
1633
+ "learning_rate": 5.1442508710801395e-05,
1634
+ "loss": 0.0562,
1635
+ "step": 2690
1636
+ },
1637
+ {
1638
+ "epoch": 1.27,
1639
+ "learning_rate": 5.1303135888501745e-05,
1640
+ "loss": 0.067,
1641
+ "step": 2700
1642
+ },
1643
+ {
1644
+ "epoch": 1.27,
1645
+ "learning_rate": 5.1163763066202095e-05,
1646
+ "loss": 0.0528,
1647
+ "step": 2710
1648
+ },
1649
+ {
1650
+ "epoch": 1.28,
1651
+ "learning_rate": 5.1024390243902444e-05,
1652
+ "loss": 0.0677,
1653
+ "step": 2720
1654
+ },
1655
+ {
1656
+ "epoch": 1.28,
1657
+ "learning_rate": 5.0885017421602794e-05,
1658
+ "loss": 0.0668,
1659
+ "step": 2730
1660
+ },
1661
+ {
1662
+ "epoch": 1.29,
1663
+ "learning_rate": 5.0745644599303143e-05,
1664
+ "loss": 0.0677,
1665
+ "step": 2740
1666
+ },
1667
+ {
1668
+ "epoch": 1.29,
1669
+ "learning_rate": 5.060627177700349e-05,
1670
+ "loss": 0.0446,
1671
+ "step": 2750
1672
+ },
1673
+ {
1674
+ "epoch": 1.3,
1675
+ "learning_rate": 5.0466898954703836e-05,
1676
+ "loss": 0.0474,
1677
+ "step": 2760
1678
+ },
1679
+ {
1680
+ "epoch": 1.3,
1681
+ "learning_rate": 5.0327526132404186e-05,
1682
+ "loss": 0.0634,
1683
+ "step": 2770
1684
+ },
1685
+ {
1686
+ "epoch": 1.31,
1687
+ "learning_rate": 5.0188153310104535e-05,
1688
+ "loss": 0.0588,
1689
+ "step": 2780
1690
+ },
1691
+ {
1692
+ "epoch": 1.31,
1693
+ "learning_rate": 5.0048780487804885e-05,
1694
+ "loss": 0.0587,
1695
+ "step": 2790
1696
+ },
1697
+ {
1698
+ "epoch": 1.32,
1699
+ "learning_rate": 4.9909407665505234e-05,
1700
+ "loss": 0.0653,
1701
+ "step": 2800
1702
+ },
1703
+ {
1704
+ "epoch": 1.32,
1705
+ "learning_rate": 4.9770034843205584e-05,
1706
+ "loss": 0.057,
1707
+ "step": 2810
1708
+ },
1709
+ {
1710
+ "epoch": 1.33,
1711
+ "learning_rate": 4.9630662020905934e-05,
1712
+ "loss": 0.0584,
1713
+ "step": 2820
1714
+ },
1715
+ {
1716
+ "epoch": 1.33,
1717
+ "learning_rate": 4.949128919860627e-05,
1718
+ "loss": 0.0485,
1719
+ "step": 2830
1720
+ },
1721
+ {
1722
+ "epoch": 1.34,
1723
+ "learning_rate": 4.935191637630662e-05,
1724
+ "loss": 0.0595,
1725
+ "step": 2840
1726
+ },
1727
+ {
1728
+ "epoch": 1.34,
1729
+ "learning_rate": 4.921254355400697e-05,
1730
+ "loss": 0.0595,
1731
+ "step": 2850
1732
+ },
1733
+ {
1734
+ "epoch": 1.35,
1735
+ "learning_rate": 4.907317073170732e-05,
1736
+ "loss": 0.0536,
1737
+ "step": 2860
1738
+ },
1739
+ {
1740
+ "epoch": 1.35,
1741
+ "learning_rate": 4.893379790940767e-05,
1742
+ "loss": 0.0674,
1743
+ "step": 2870
1744
+ },
1745
+ {
1746
+ "epoch": 1.35,
1747
+ "learning_rate": 4.879442508710802e-05,
1748
+ "loss": 0.061,
1749
+ "step": 2880
1750
+ },
1751
+ {
1752
+ "epoch": 1.36,
1753
+ "learning_rate": 4.865505226480837e-05,
1754
+ "loss": 0.0528,
1755
+ "step": 2890
1756
+ },
1757
+ {
1758
+ "epoch": 1.36,
1759
+ "learning_rate": 4.851567944250872e-05,
1760
+ "loss": 0.0531,
1761
+ "step": 2900
1762
+ },
1763
+ {
1764
+ "epoch": 1.37,
1765
+ "learning_rate": 4.837630662020906e-05,
1766
+ "loss": 0.0558,
1767
+ "step": 2910
1768
+ },
1769
+ {
1770
+ "epoch": 1.37,
1771
+ "learning_rate": 4.823693379790941e-05,
1772
+ "loss": 0.0708,
1773
+ "step": 2920
1774
+ },
1775
+ {
1776
+ "epoch": 1.38,
1777
+ "learning_rate": 4.809756097560976e-05,
1778
+ "loss": 0.0595,
1779
+ "step": 2930
1780
+ },
1781
+ {
1782
+ "epoch": 1.38,
1783
+ "learning_rate": 4.795818815331011e-05,
1784
+ "loss": 0.0486,
1785
+ "step": 2940
1786
+ },
1787
+ {
1788
+ "epoch": 1.39,
1789
+ "learning_rate": 4.781881533101046e-05,
1790
+ "loss": 0.0644,
1791
+ "step": 2950
1792
+ },
1793
+ {
1794
+ "epoch": 1.39,
1795
+ "learning_rate": 4.767944250871081e-05,
1796
+ "loss": 0.0525,
1797
+ "step": 2960
1798
+ },
1799
+ {
1800
+ "epoch": 1.4,
1801
+ "learning_rate": 4.754006968641116e-05,
1802
+ "loss": 0.0704,
1803
+ "step": 2970
1804
+ },
1805
+ {
1806
+ "epoch": 1.4,
1807
+ "learning_rate": 4.740069686411151e-05,
1808
+ "loss": 0.0612,
1809
+ "step": 2980
1810
+ },
1811
+ {
1812
+ "epoch": 1.41,
1813
+ "learning_rate": 4.726132404181185e-05,
1814
+ "loss": 0.061,
1815
+ "step": 2990
1816
+ },
1817
+ {
1818
+ "epoch": 1.41,
1819
+ "learning_rate": 4.71219512195122e-05,
1820
+ "loss": 0.0616,
1821
+ "step": 3000
1822
+ },
1823
+ {
1824
+ "epoch": 1.42,
1825
+ "learning_rate": 4.698257839721254e-05,
1826
+ "loss": 0.0682,
1827
+ "step": 3010
1828
+ },
1829
+ {
1830
+ "epoch": 1.42,
1831
+ "learning_rate": 4.684320557491289e-05,
1832
+ "loss": 0.0509,
1833
+ "step": 3020
1834
+ },
1835
+ {
1836
+ "epoch": 1.43,
1837
+ "learning_rate": 4.670383275261324e-05,
1838
+ "loss": 0.0587,
1839
+ "step": 3030
1840
+ },
1841
+ {
1842
+ "epoch": 1.43,
1843
+ "learning_rate": 4.656445993031359e-05,
1844
+ "loss": 0.0562,
1845
+ "step": 3040
1846
+ },
1847
+ {
1848
+ "epoch": 1.43,
1849
+ "learning_rate": 4.642508710801394e-05,
1850
+ "loss": 0.0606,
1851
+ "step": 3050
1852
+ },
1853
+ {
1854
+ "epoch": 1.44,
1855
+ "learning_rate": 4.628571428571429e-05,
1856
+ "loss": 0.0576,
1857
+ "step": 3060
1858
+ },
1859
+ {
1860
+ "epoch": 1.44,
1861
+ "learning_rate": 4.6146341463414634e-05,
1862
+ "loss": 0.0461,
1863
+ "step": 3070
1864
+ },
1865
+ {
1866
+ "epoch": 1.45,
1867
+ "learning_rate": 4.600696864111498e-05,
1868
+ "loss": 0.0582,
1869
+ "step": 3080
1870
+ },
1871
+ {
1872
+ "epoch": 1.45,
1873
+ "learning_rate": 4.586759581881533e-05,
1874
+ "loss": 0.0583,
1875
+ "step": 3090
1876
+ },
1877
+ {
1878
+ "epoch": 1.46,
1879
+ "learning_rate": 4.572822299651568e-05,
1880
+ "loss": 0.054,
1881
+ "step": 3100
1882
+ },
1883
+ {
1884
+ "epoch": 1.46,
1885
+ "learning_rate": 4.558885017421603e-05,
1886
+ "loss": 0.0495,
1887
+ "step": 3110
1888
+ },
1889
+ {
1890
+ "epoch": 1.47,
1891
+ "learning_rate": 4.544947735191638e-05,
1892
+ "loss": 0.0565,
1893
+ "step": 3120
1894
+ },
1895
+ {
1896
+ "epoch": 1.47,
1897
+ "learning_rate": 4.531010452961673e-05,
1898
+ "loss": 0.0439,
1899
+ "step": 3130
1900
+ },
1901
+ {
1902
+ "epoch": 1.48,
1903
+ "learning_rate": 4.5170731707317074e-05,
1904
+ "loss": 0.055,
1905
+ "step": 3140
1906
+ },
1907
+ {
1908
+ "epoch": 1.48,
1909
+ "learning_rate": 4.5031358885017424e-05,
1910
+ "loss": 0.0626,
1911
+ "step": 3150
1912
+ },
1913
+ {
1914
+ "epoch": 1.49,
1915
+ "learning_rate": 4.4891986062717773e-05,
1916
+ "loss": 0.0628,
1917
+ "step": 3160
1918
+ },
1919
+ {
1920
+ "epoch": 1.49,
1921
+ "learning_rate": 4.475261324041812e-05,
1922
+ "loss": 0.0627,
1923
+ "step": 3170
1924
+ },
1925
+ {
1926
+ "epoch": 1.5,
1927
+ "learning_rate": 4.461324041811847e-05,
1928
+ "loss": 0.0544,
1929
+ "step": 3180
1930
+ },
1931
+ {
1932
+ "epoch": 1.5,
1933
+ "learning_rate": 4.447386759581882e-05,
1934
+ "loss": 0.0464,
1935
+ "step": 3190
1936
+ },
1937
+ {
1938
+ "epoch": 1.51,
1939
+ "learning_rate": 4.433449477351917e-05,
1940
+ "loss": 0.0494,
1941
+ "step": 3200
1942
+ },
1943
+ {
1944
+ "epoch": 1.51,
1945
+ "learning_rate": 4.419512195121952e-05,
1946
+ "loss": 0.052,
1947
+ "step": 3210
1948
+ },
1949
+ {
1950
+ "epoch": 1.51,
1951
+ "learning_rate": 4.405574912891986e-05,
1952
+ "loss": 0.0553,
1953
+ "step": 3220
1954
+ },
1955
+ {
1956
+ "epoch": 1.52,
1957
+ "learning_rate": 4.391637630662021e-05,
1958
+ "loss": 0.0565,
1959
+ "step": 3230
1960
+ },
1961
+ {
1962
+ "epoch": 1.52,
1963
+ "learning_rate": 4.377700348432056e-05,
1964
+ "loss": 0.0576,
1965
+ "step": 3240
1966
+ },
1967
+ {
1968
+ "epoch": 1.53,
1969
+ "learning_rate": 4.3637630662020907e-05,
1970
+ "loss": 0.0613,
1971
+ "step": 3250
1972
+ },
1973
+ {
1974
+ "epoch": 1.53,
1975
+ "learning_rate": 4.3498257839721256e-05,
1976
+ "loss": 0.0601,
1977
+ "step": 3260
1978
+ },
1979
+ {
1980
+ "epoch": 1.54,
1981
+ "learning_rate": 4.3358885017421606e-05,
1982
+ "loss": 0.0566,
1983
+ "step": 3270
1984
+ },
1985
+ {
1986
+ "epoch": 1.54,
1987
+ "learning_rate": 4.3219512195121955e-05,
1988
+ "loss": 0.0551,
1989
+ "step": 3280
1990
+ },
1991
+ {
1992
+ "epoch": 1.55,
1993
+ "learning_rate": 4.3080139372822305e-05,
1994
+ "loss": 0.0478,
1995
+ "step": 3290
1996
+ },
1997
+ {
1998
+ "epoch": 1.55,
1999
+ "learning_rate": 4.294076655052265e-05,
2000
+ "loss": 0.0561,
2001
+ "step": 3300
2002
+ },
2003
+ {
2004
+ "epoch": 1.56,
2005
+ "learning_rate": 4.2801393728223e-05,
2006
+ "loss": 0.0577,
2007
+ "step": 3310
2008
+ },
2009
+ {
2010
+ "epoch": 1.56,
2011
+ "learning_rate": 4.266202090592335e-05,
2012
+ "loss": 0.0533,
2013
+ "step": 3320
2014
+ },
2015
+ {
2016
+ "epoch": 1.57,
2017
+ "learning_rate": 4.25226480836237e-05,
2018
+ "loss": 0.0603,
2019
+ "step": 3330
2020
+ },
2021
+ {
2022
+ "epoch": 1.57,
2023
+ "learning_rate": 4.2383275261324046e-05,
2024
+ "loss": 0.0533,
2025
+ "step": 3340
2026
+ },
2027
+ {
2028
+ "epoch": 1.58,
2029
+ "learning_rate": 4.2243902439024396e-05,
2030
+ "loss": 0.0718,
2031
+ "step": 3350
2032
+ },
2033
+ {
2034
+ "epoch": 1.58,
2035
+ "learning_rate": 4.2104529616724746e-05,
2036
+ "loss": 0.0557,
2037
+ "step": 3360
2038
+ },
2039
+ {
2040
+ "epoch": 1.59,
2041
+ "learning_rate": 4.1965156794425095e-05,
2042
+ "loss": 0.0656,
2043
+ "step": 3370
2044
+ },
2045
+ {
2046
+ "epoch": 1.59,
2047
+ "learning_rate": 4.182578397212544e-05,
2048
+ "loss": 0.0608,
2049
+ "step": 3380
2050
+ },
2051
+ {
2052
+ "epoch": 1.59,
2053
+ "learning_rate": 4.168641114982579e-05,
2054
+ "loss": 0.0588,
2055
+ "step": 3390
2056
+ },
2057
+ {
2058
+ "epoch": 1.6,
2059
+ "learning_rate": 4.154703832752614e-05,
2060
+ "loss": 0.0556,
2061
+ "step": 3400
2062
+ },
2063
+ {
2064
+ "epoch": 1.6,
2065
+ "learning_rate": 4.140766550522649e-05,
2066
+ "loss": 0.0615,
2067
+ "step": 3410
2068
+ },
2069
+ {
2070
+ "epoch": 1.61,
2071
+ "learning_rate": 4.1268292682926837e-05,
2072
+ "loss": 0.0524,
2073
+ "step": 3420
2074
+ },
2075
+ {
2076
+ "epoch": 1.61,
2077
+ "learning_rate": 4.1128919860627186e-05,
2078
+ "loss": 0.0688,
2079
+ "step": 3430
2080
+ },
2081
+ {
2082
+ "epoch": 1.62,
2083
+ "learning_rate": 4.0989547038327536e-05,
2084
+ "loss": 0.053,
2085
+ "step": 3440
2086
+ },
2087
+ {
2088
+ "epoch": 1.62,
2089
+ "learning_rate": 4.085017421602787e-05,
2090
+ "loss": 0.0657,
2091
+ "step": 3450
2092
+ },
2093
+ {
2094
+ "epoch": 1.63,
2095
+ "learning_rate": 4.071080139372822e-05,
2096
+ "loss": 0.0641,
2097
+ "step": 3460
2098
+ },
2099
+ {
2100
+ "epoch": 1.63,
2101
+ "learning_rate": 4.057142857142857e-05,
2102
+ "loss": 0.0628,
2103
+ "step": 3470
2104
+ },
2105
+ {
2106
+ "epoch": 1.64,
2107
+ "learning_rate": 4.043205574912892e-05,
2108
+ "loss": 0.0599,
2109
+ "step": 3480
2110
+ },
2111
+ {
2112
+ "epoch": 1.64,
2113
+ "learning_rate": 4.029268292682927e-05,
2114
+ "loss": 0.0512,
2115
+ "step": 3490
2116
+ },
2117
+ {
2118
+ "epoch": 1.65,
2119
+ "learning_rate": 4.015331010452962e-05,
2120
+ "loss": 0.0552,
2121
+ "step": 3500
2122
+ },
2123
+ {
2124
+ "epoch": 1.65,
2125
+ "learning_rate": 4.001393728222997e-05,
2126
+ "loss": 0.0572,
2127
+ "step": 3510
2128
+ },
2129
+ {
2130
+ "epoch": 1.66,
2131
+ "learning_rate": 3.987456445993032e-05,
2132
+ "loss": 0.063,
2133
+ "step": 3520
2134
+ },
2135
+ {
2136
+ "epoch": 1.66,
2137
+ "learning_rate": 3.973519163763067e-05,
2138
+ "loss": 0.0634,
2139
+ "step": 3530
2140
+ },
2141
+ {
2142
+ "epoch": 1.67,
2143
+ "learning_rate": 3.959581881533102e-05,
2144
+ "loss": 0.061,
2145
+ "step": 3540
2146
+ },
2147
+ {
2148
+ "epoch": 1.67,
2149
+ "learning_rate": 3.945644599303136e-05,
2150
+ "loss": 0.0543,
2151
+ "step": 3550
2152
+ },
2153
+ {
2154
+ "epoch": 1.67,
2155
+ "learning_rate": 3.931707317073171e-05,
2156
+ "loss": 0.0698,
2157
+ "step": 3560
2158
+ },
2159
+ {
2160
+ "epoch": 1.68,
2161
+ "learning_rate": 3.917770034843206e-05,
2162
+ "loss": 0.0638,
2163
+ "step": 3570
2164
+ },
2165
+ {
2166
+ "epoch": 1.68,
2167
+ "learning_rate": 3.903832752613241e-05,
2168
+ "loss": 0.0584,
2169
+ "step": 3580
2170
+ },
2171
+ {
2172
+ "epoch": 1.69,
2173
+ "learning_rate": 3.889895470383275e-05,
2174
+ "loss": 0.0532,
2175
+ "step": 3590
2176
+ },
2177
+ {
2178
+ "epoch": 1.69,
2179
+ "learning_rate": 3.87595818815331e-05,
2180
+ "loss": 0.0547,
2181
+ "step": 3600
2182
+ },
2183
+ {
2184
+ "epoch": 1.7,
2185
+ "learning_rate": 3.862020905923345e-05,
2186
+ "loss": 0.0624,
2187
+ "step": 3610
2188
+ },
2189
+ {
2190
+ "epoch": 1.7,
2191
+ "learning_rate": 3.84808362369338e-05,
2192
+ "loss": 0.0597,
2193
+ "step": 3620
2194
+ },
2195
+ {
2196
+ "epoch": 1.71,
2197
+ "learning_rate": 3.8341463414634145e-05,
2198
+ "loss": 0.0636,
2199
+ "step": 3630
2200
+ },
2201
+ {
2202
+ "epoch": 1.71,
2203
+ "learning_rate": 3.8202090592334494e-05,
2204
+ "loss": 0.0603,
2205
+ "step": 3640
2206
+ },
2207
+ {
2208
+ "epoch": 1.72,
2209
+ "learning_rate": 3.8062717770034844e-05,
2210
+ "loss": 0.0618,
2211
+ "step": 3650
2212
+ },
2213
+ {
2214
+ "epoch": 1.72,
2215
+ "learning_rate": 3.7923344947735194e-05,
2216
+ "loss": 0.059,
2217
+ "step": 3660
2218
+ },
2219
+ {
2220
+ "epoch": 1.73,
2221
+ "learning_rate": 3.778397212543554e-05,
2222
+ "loss": 0.0674,
2223
+ "step": 3670
2224
+ },
2225
+ {
2226
+ "epoch": 1.73,
2227
+ "learning_rate": 3.764459930313589e-05,
2228
+ "loss": 0.0628,
2229
+ "step": 3680
2230
+ },
2231
+ {
2232
+ "epoch": 1.74,
2233
+ "learning_rate": 3.750522648083624e-05,
2234
+ "loss": 0.0623,
2235
+ "step": 3690
2236
+ },
2237
+ {
2238
+ "epoch": 1.74,
2239
+ "learning_rate": 3.736585365853659e-05,
2240
+ "loss": 0.0545,
2241
+ "step": 3700
2242
+ },
2243
+ {
2244
+ "epoch": 1.75,
2245
+ "learning_rate": 3.7226480836236935e-05,
2246
+ "loss": 0.0559,
2247
+ "step": 3710
2248
+ },
2249
+ {
2250
+ "epoch": 1.75,
2251
+ "learning_rate": 3.7087108013937285e-05,
2252
+ "loss": 0.0563,
2253
+ "step": 3720
2254
+ },
2255
+ {
2256
+ "epoch": 1.75,
2257
+ "learning_rate": 3.6947735191637634e-05,
2258
+ "loss": 0.0634,
2259
+ "step": 3730
2260
+ },
2261
+ {
2262
+ "epoch": 1.76,
2263
+ "learning_rate": 3.680836236933798e-05,
2264
+ "loss": 0.0624,
2265
+ "step": 3740
2266
+ },
2267
+ {
2268
+ "epoch": 1.76,
2269
+ "learning_rate": 3.666898954703833e-05,
2270
+ "loss": 0.0485,
2271
+ "step": 3750
2272
+ },
2273
+ {
2274
+ "epoch": 1.77,
2275
+ "learning_rate": 3.6529616724738676e-05,
2276
+ "loss": 0.0564,
2277
+ "step": 3760
2278
+ },
2279
+ {
2280
+ "epoch": 1.77,
2281
+ "learning_rate": 3.6390243902439026e-05,
2282
+ "loss": 0.0605,
2283
+ "step": 3770
2284
+ },
2285
+ {
2286
+ "epoch": 1.78,
2287
+ "learning_rate": 3.6250871080139376e-05,
2288
+ "loss": 0.0603,
2289
+ "step": 3780
2290
+ },
2291
+ {
2292
+ "epoch": 1.78,
2293
+ "learning_rate": 3.6111498257839725e-05,
2294
+ "loss": 0.0635,
2295
+ "step": 3790
2296
+ },
2297
+ {
2298
+ "epoch": 1.79,
2299
+ "learning_rate": 3.5972125435540075e-05,
2300
+ "loss": 0.056,
2301
+ "step": 3800
2302
+ },
2303
+ {
2304
+ "epoch": 1.79,
2305
+ "learning_rate": 3.5832752613240425e-05,
2306
+ "loss": 0.0578,
2307
+ "step": 3810
2308
+ },
2309
+ {
2310
+ "epoch": 1.8,
2311
+ "learning_rate": 3.569337979094077e-05,
2312
+ "loss": 0.0574,
2313
+ "step": 3820
2314
+ },
2315
+ {
2316
+ "epoch": 1.8,
2317
+ "learning_rate": 3.555400696864112e-05,
2318
+ "loss": 0.0649,
2319
+ "step": 3830
2320
+ },
2321
+ {
2322
+ "epoch": 1.81,
2323
+ "learning_rate": 3.541463414634147e-05,
2324
+ "loss": 0.0566,
2325
+ "step": 3840
2326
+ },
2327
+ {
2328
+ "epoch": 1.81,
2329
+ "learning_rate": 3.5275261324041816e-05,
2330
+ "loss": 0.0562,
2331
+ "step": 3850
2332
+ },
2333
+ {
2334
+ "epoch": 1.82,
2335
+ "learning_rate": 3.513588850174216e-05,
2336
+ "loss": 0.0475,
2337
+ "step": 3860
2338
+ },
2339
+ {
2340
+ "epoch": 1.82,
2341
+ "learning_rate": 3.499651567944251e-05,
2342
+ "loss": 0.0559,
2343
+ "step": 3870
2344
+ },
2345
+ {
2346
+ "epoch": 1.83,
2347
+ "learning_rate": 3.485714285714286e-05,
2348
+ "loss": 0.0487,
2349
+ "step": 3880
2350
+ },
2351
+ {
2352
+ "epoch": 1.83,
2353
+ "learning_rate": 3.471777003484321e-05,
2354
+ "loss": 0.065,
2355
+ "step": 3890
2356
+ },
2357
+ {
2358
+ "epoch": 1.83,
2359
+ "learning_rate": 3.457839721254356e-05,
2360
+ "loss": 0.0655,
2361
+ "step": 3900
2362
+ },
2363
+ {
2364
+ "epoch": 1.84,
2365
+ "learning_rate": 3.443902439024391e-05,
2366
+ "loss": 0.056,
2367
+ "step": 3910
2368
+ },
2369
+ {
2370
+ "epoch": 1.84,
2371
+ "learning_rate": 3.429965156794426e-05,
2372
+ "loss": 0.0547,
2373
+ "step": 3920
2374
+ },
2375
+ {
2376
+ "epoch": 1.85,
2377
+ "learning_rate": 3.4160278745644606e-05,
2378
+ "loss": 0.0573,
2379
+ "step": 3930
2380
+ },
2381
+ {
2382
+ "epoch": 1.85,
2383
+ "learning_rate": 3.402090592334495e-05,
2384
+ "loss": 0.0618,
2385
+ "step": 3940
2386
+ },
2387
+ {
2388
+ "epoch": 1.86,
2389
+ "learning_rate": 3.38815331010453e-05,
2390
+ "loss": 0.0564,
2391
+ "step": 3950
2392
+ },
2393
+ {
2394
+ "epoch": 1.86,
2395
+ "learning_rate": 3.374216027874565e-05,
2396
+ "loss": 0.0547,
2397
+ "step": 3960
2398
+ },
2399
+ {
2400
+ "epoch": 1.87,
2401
+ "learning_rate": 3.3602787456446e-05,
2402
+ "loss": 0.0644,
2403
+ "step": 3970
2404
+ },
2405
+ {
2406
+ "epoch": 1.87,
2407
+ "learning_rate": 3.346341463414634e-05,
2408
+ "loss": 0.0564,
2409
+ "step": 3980
2410
+ },
2411
+ {
2412
+ "epoch": 1.88,
2413
+ "learning_rate": 3.332404181184669e-05,
2414
+ "loss": 0.0633,
2415
+ "step": 3990
2416
+ },
2417
+ {
2418
+ "epoch": 1.88,
2419
+ "learning_rate": 3.318466898954704e-05,
2420
+ "loss": 0.0574,
2421
+ "step": 4000
2422
+ },
2423
+ {
2424
+ "epoch": 1.89,
2425
+ "learning_rate": 3.304529616724739e-05,
2426
+ "loss": 0.0547,
2427
+ "step": 4010
2428
+ },
2429
+ {
2430
+ "epoch": 1.89,
2431
+ "learning_rate": 3.290592334494774e-05,
2432
+ "loss": 0.0665,
2433
+ "step": 4020
2434
+ },
2435
+ {
2436
+ "epoch": 1.9,
2437
+ "learning_rate": 3.276655052264809e-05,
2438
+ "loss": 0.0545,
2439
+ "step": 4030
2440
+ },
2441
+ {
2442
+ "epoch": 1.9,
2443
+ "learning_rate": 3.262717770034844e-05,
2444
+ "loss": 0.0692,
2445
+ "step": 4040
2446
+ },
2447
+ {
2448
+ "epoch": 1.9,
2449
+ "learning_rate": 3.248780487804879e-05,
2450
+ "loss": 0.0648,
2451
+ "step": 4050
2452
+ },
2453
+ {
2454
+ "epoch": 1.91,
2455
+ "learning_rate": 3.234843205574913e-05,
2456
+ "loss": 0.0542,
2457
+ "step": 4060
2458
+ },
2459
+ {
2460
+ "epoch": 1.91,
2461
+ "learning_rate": 3.220905923344948e-05,
2462
+ "loss": 0.0562,
2463
+ "step": 4070
2464
+ },
2465
+ {
2466
+ "epoch": 1.92,
2467
+ "learning_rate": 3.206968641114983e-05,
2468
+ "loss": 0.0531,
2469
+ "step": 4080
2470
+ },
2471
+ {
2472
+ "epoch": 1.92,
2473
+ "learning_rate": 3.193031358885017e-05,
2474
+ "loss": 0.0615,
2475
+ "step": 4090
2476
+ },
2477
+ {
2478
+ "epoch": 1.93,
2479
+ "learning_rate": 3.179094076655052e-05,
2480
+ "loss": 0.0484,
2481
+ "step": 4100
2482
+ },
2483
+ {
2484
+ "epoch": 1.93,
2485
+ "learning_rate": 3.165156794425087e-05,
2486
+ "loss": 0.0492,
2487
+ "step": 4110
2488
+ },
2489
+ {
2490
+ "epoch": 1.94,
2491
+ "learning_rate": 3.151219512195122e-05,
2492
+ "loss": 0.0594,
2493
+ "step": 4120
2494
+ },
2495
+ {
2496
+ "epoch": 1.94,
2497
+ "learning_rate": 3.137282229965157e-05,
2498
+ "loss": 0.0489,
2499
+ "step": 4130
2500
+ },
2501
+ {
2502
+ "epoch": 1.95,
2503
+ "learning_rate": 3.123344947735192e-05,
2504
+ "loss": 0.0619,
2505
+ "step": 4140
2506
+ },
2507
+ {
2508
+ "epoch": 1.95,
2509
+ "learning_rate": 3.1094076655052264e-05,
2510
+ "loss": 0.0491,
2511
+ "step": 4150
2512
+ },
2513
+ {
2514
+ "epoch": 1.96,
2515
+ "learning_rate": 3.0954703832752614e-05,
2516
+ "loss": 0.0627,
2517
+ "step": 4160
2518
+ },
2519
+ {
2520
+ "epoch": 1.96,
2521
+ "learning_rate": 3.0815331010452964e-05,
2522
+ "loss": 0.0503,
2523
+ "step": 4170
2524
+ },
2525
+ {
2526
+ "epoch": 1.97,
2527
+ "learning_rate": 3.067595818815331e-05,
2528
+ "loss": 0.0495,
2529
+ "step": 4180
2530
+ },
2531
+ {
2532
+ "epoch": 1.97,
2533
+ "learning_rate": 3.053658536585366e-05,
2534
+ "loss": 0.064,
2535
+ "step": 4190
2536
+ },
2537
+ {
2538
+ "epoch": 1.98,
2539
+ "learning_rate": 3.0397212543554012e-05,
2540
+ "loss": 0.0634,
2541
+ "step": 4200
2542
+ },
2543
+ {
2544
+ "epoch": 1.98,
2545
+ "learning_rate": 3.0257839721254355e-05,
2546
+ "loss": 0.0582,
2547
+ "step": 4210
2548
+ },
2549
+ {
2550
+ "epoch": 1.98,
2551
+ "learning_rate": 3.0118466898954705e-05,
2552
+ "loss": 0.0654,
2553
+ "step": 4220
2554
+ },
2555
+ {
2556
+ "epoch": 1.99,
2557
+ "learning_rate": 2.9979094076655055e-05,
2558
+ "loss": 0.0524,
2559
+ "step": 4230
2560
+ },
2561
+ {
2562
+ "epoch": 1.99,
2563
+ "learning_rate": 2.9839721254355404e-05,
2564
+ "loss": 0.0613,
2565
+ "step": 4240
2566
+ },
2567
+ {
2568
+ "epoch": 2.0,
2569
+ "learning_rate": 2.970034843205575e-05,
2570
+ "loss": 0.0484,
2571
+ "step": 4250
2572
+ },
2573
+ {
2574
+ "epoch": 2.0,
2575
+ "eval_accuracy": 0.9767567988685532,
2576
+ "eval_f1": 0.7813079163357163,
2577
+ "eval_loss": 0.06523581594228745,
2578
+ "eval_precision": 0.7725130890052356,
2579
+ "eval_recall": 0.7903053026245314,
2580
+ "eval_runtime": 14.0928,
2581
+ "eval_samples_per_second": 1004.765,
2582
+ "eval_steps_per_second": 125.596,
2583
+ "step": 4252
2584
+ },
2585
+ {
2586
+ "epoch": 2.0,
2587
+ "learning_rate": 2.95609756097561e-05,
2588
+ "loss": 0.0381,
2589
+ "step": 4260
2590
+ },
2591
+ {
2592
+ "epoch": 2.01,
2593
+ "learning_rate": 2.942160278745645e-05,
2594
+ "loss": 0.0487,
2595
+ "step": 4270
2596
+ },
2597
+ {
2598
+ "epoch": 2.01,
2599
+ "learning_rate": 2.92822299651568e-05,
2600
+ "loss": 0.046,
2601
+ "step": 4280
2602
+ },
2603
+ {
2604
+ "epoch": 2.02,
2605
+ "learning_rate": 2.9142857142857146e-05,
2606
+ "loss": 0.0403,
2607
+ "step": 4290
2608
+ },
2609
+ {
2610
+ "epoch": 2.02,
2611
+ "learning_rate": 2.9003484320557492e-05,
2612
+ "loss": 0.0403,
2613
+ "step": 4300
2614
+ },
2615
+ {
2616
+ "epoch": 2.03,
2617
+ "learning_rate": 2.886411149825784e-05,
2618
+ "loss": 0.04,
2619
+ "step": 4310
2620
+ },
2621
+ {
2622
+ "epoch": 2.03,
2623
+ "learning_rate": 2.872473867595819e-05,
2624
+ "loss": 0.0481,
2625
+ "step": 4320
2626
+ },
2627
+ {
2628
+ "epoch": 2.04,
2629
+ "learning_rate": 2.8585365853658537e-05,
2630
+ "loss": 0.0436,
2631
+ "step": 4330
2632
+ },
2633
+ {
2634
+ "epoch": 2.04,
2635
+ "learning_rate": 2.8445993031358887e-05,
2636
+ "loss": 0.0363,
2637
+ "step": 4340
2638
+ },
2639
+ {
2640
+ "epoch": 2.05,
2641
+ "learning_rate": 2.8306620209059237e-05,
2642
+ "loss": 0.0476,
2643
+ "step": 4350
2644
+ },
2645
+ {
2646
+ "epoch": 2.05,
2647
+ "learning_rate": 2.8167247386759586e-05,
2648
+ "loss": 0.0398,
2649
+ "step": 4360
2650
+ },
2651
+ {
2652
+ "epoch": 2.06,
2653
+ "learning_rate": 2.8027874564459932e-05,
2654
+ "loss": 0.0443,
2655
+ "step": 4370
2656
+ },
2657
+ {
2658
+ "epoch": 2.06,
2659
+ "learning_rate": 2.7888501742160282e-05,
2660
+ "loss": 0.0451,
2661
+ "step": 4380
2662
+ },
2663
+ {
2664
+ "epoch": 2.06,
2665
+ "learning_rate": 2.774912891986063e-05,
2666
+ "loss": 0.0411,
2667
+ "step": 4390
2668
+ },
2669
+ {
2670
+ "epoch": 2.07,
2671
+ "learning_rate": 2.7609756097560974e-05,
2672
+ "loss": 0.0524,
2673
+ "step": 4400
2674
+ },
2675
+ {
2676
+ "epoch": 2.07,
2677
+ "learning_rate": 2.7470383275261324e-05,
2678
+ "loss": 0.0449,
2679
+ "step": 4410
2680
+ },
2681
+ {
2682
+ "epoch": 2.08,
2683
+ "learning_rate": 2.7331010452961674e-05,
2684
+ "loss": 0.0405,
2685
+ "step": 4420
2686
+ },
2687
+ {
2688
+ "epoch": 2.08,
2689
+ "learning_rate": 2.7191637630662023e-05,
2690
+ "loss": 0.0502,
2691
+ "step": 4430
2692
+ },
2693
+ {
2694
+ "epoch": 2.09,
2695
+ "learning_rate": 2.705226480836237e-05,
2696
+ "loss": 0.0444,
2697
+ "step": 4440
2698
+ },
2699
+ {
2700
+ "epoch": 2.09,
2701
+ "learning_rate": 2.691289198606272e-05,
2702
+ "loss": 0.0371,
2703
+ "step": 4450
2704
+ },
2705
+ {
2706
+ "epoch": 2.1,
2707
+ "learning_rate": 2.677351916376307e-05,
2708
+ "loss": 0.0485,
2709
+ "step": 4460
2710
+ },
2711
+ {
2712
+ "epoch": 2.1,
2713
+ "learning_rate": 2.663414634146342e-05,
2714
+ "loss": 0.0444,
2715
+ "step": 4470
2716
+ },
2717
+ {
2718
+ "epoch": 2.11,
2719
+ "learning_rate": 2.6494773519163765e-05,
2720
+ "loss": 0.0392,
2721
+ "step": 4480
2722
+ },
2723
+ {
2724
+ "epoch": 2.11,
2725
+ "learning_rate": 2.6355400696864114e-05,
2726
+ "loss": 0.0461,
2727
+ "step": 4490
2728
+ },
2729
+ {
2730
+ "epoch": 2.12,
2731
+ "learning_rate": 2.6216027874564464e-05,
2732
+ "loss": 0.0459,
2733
+ "step": 4500
2734
+ },
2735
+ {
2736
+ "epoch": 2.12,
2737
+ "learning_rate": 2.6076655052264814e-05,
2738
+ "loss": 0.0537,
2739
+ "step": 4510
2740
+ },
2741
+ {
2742
+ "epoch": 2.13,
2743
+ "learning_rate": 2.5937282229965156e-05,
2744
+ "loss": 0.0409,
2745
+ "step": 4520
2746
+ },
2747
+ {
2748
+ "epoch": 2.13,
2749
+ "learning_rate": 2.5797909407665506e-05,
2750
+ "loss": 0.0446,
2751
+ "step": 4530
2752
+ },
2753
+ {
2754
+ "epoch": 2.14,
2755
+ "learning_rate": 2.5658536585365856e-05,
2756
+ "loss": 0.0512,
2757
+ "step": 4540
2758
+ },
2759
+ {
2760
+ "epoch": 2.14,
2761
+ "learning_rate": 2.5519163763066205e-05,
2762
+ "loss": 0.0475,
2763
+ "step": 4550
2764
+ },
2765
+ {
2766
+ "epoch": 2.14,
2767
+ "learning_rate": 2.537979094076655e-05,
2768
+ "loss": 0.045,
2769
+ "step": 4560
2770
+ },
2771
+ {
2772
+ "epoch": 2.15,
2773
+ "learning_rate": 2.52404181184669e-05,
2774
+ "loss": 0.0443,
2775
+ "step": 4570
2776
+ },
2777
+ {
2778
+ "epoch": 2.15,
2779
+ "learning_rate": 2.510104529616725e-05,
2780
+ "loss": 0.0409,
2781
+ "step": 4580
2782
+ },
2783
+ {
2784
+ "epoch": 2.16,
2785
+ "learning_rate": 2.49616724738676e-05,
2786
+ "loss": 0.0511,
2787
+ "step": 4590
2788
+ },
2789
+ {
2790
+ "epoch": 2.16,
2791
+ "learning_rate": 2.4822299651567943e-05,
2792
+ "loss": 0.0429,
2793
+ "step": 4600
2794
+ },
2795
+ {
2796
+ "epoch": 2.17,
2797
+ "learning_rate": 2.4682926829268293e-05,
2798
+ "loss": 0.0407,
2799
+ "step": 4610
2800
+ },
2801
+ {
2802
+ "epoch": 2.17,
2803
+ "learning_rate": 2.4543554006968642e-05,
2804
+ "loss": 0.0422,
2805
+ "step": 4620
2806
+ },
2807
+ {
2808
+ "epoch": 2.18,
2809
+ "learning_rate": 2.4404181184668992e-05,
2810
+ "loss": 0.0381,
2811
+ "step": 4630
2812
+ },
2813
+ {
2814
+ "epoch": 2.18,
2815
+ "learning_rate": 2.426480836236934e-05,
2816
+ "loss": 0.0382,
2817
+ "step": 4640
2818
+ },
2819
+ {
2820
+ "epoch": 2.19,
2821
+ "learning_rate": 2.4125435540069688e-05,
2822
+ "loss": 0.044,
2823
+ "step": 4650
2824
+ },
2825
+ {
2826
+ "epoch": 2.19,
2827
+ "learning_rate": 2.3986062717770038e-05,
2828
+ "loss": 0.0462,
2829
+ "step": 4660
2830
+ },
2831
+ {
2832
+ "epoch": 2.2,
2833
+ "learning_rate": 2.3846689895470387e-05,
2834
+ "loss": 0.0453,
2835
+ "step": 4670
2836
+ },
2837
+ {
2838
+ "epoch": 2.2,
2839
+ "learning_rate": 2.3707317073170733e-05,
2840
+ "loss": 0.0449,
2841
+ "step": 4680
2842
+ },
2843
+ {
2844
+ "epoch": 2.21,
2845
+ "learning_rate": 2.3567944250871083e-05,
2846
+ "loss": 0.0404,
2847
+ "step": 4690
2848
+ },
2849
+ {
2850
+ "epoch": 2.21,
2851
+ "learning_rate": 2.3428571428571433e-05,
2852
+ "loss": 0.0375,
2853
+ "step": 4700
2854
+ },
2855
+ {
2856
+ "epoch": 2.22,
2857
+ "learning_rate": 2.3289198606271776e-05,
2858
+ "loss": 0.0422,
2859
+ "step": 4710
2860
+ },
2861
+ {
2862
+ "epoch": 2.22,
2863
+ "learning_rate": 2.3149825783972125e-05,
2864
+ "loss": 0.0448,
2865
+ "step": 4720
2866
+ },
2867
+ {
2868
+ "epoch": 2.22,
2869
+ "learning_rate": 2.3010452961672475e-05,
2870
+ "loss": 0.0368,
2871
+ "step": 4730
2872
+ },
2873
+ {
2874
+ "epoch": 2.23,
2875
+ "learning_rate": 2.2871080139372824e-05,
2876
+ "loss": 0.0477,
2877
+ "step": 4740
2878
+ },
2879
+ {
2880
+ "epoch": 2.23,
2881
+ "learning_rate": 2.273170731707317e-05,
2882
+ "loss": 0.0418,
2883
+ "step": 4750
2884
+ },
2885
+ {
2886
+ "epoch": 2.24,
2887
+ "learning_rate": 2.259233449477352e-05,
2888
+ "loss": 0.0372,
2889
+ "step": 4760
2890
+ },
2891
+ {
2892
+ "epoch": 2.24,
2893
+ "learning_rate": 2.245296167247387e-05,
2894
+ "loss": 0.0428,
2895
+ "step": 4770
2896
+ },
2897
+ {
2898
+ "epoch": 2.25,
2899
+ "learning_rate": 2.231358885017422e-05,
2900
+ "loss": 0.0353,
2901
+ "step": 4780
2902
+ },
2903
+ {
2904
+ "epoch": 2.25,
2905
+ "learning_rate": 2.2174216027874566e-05,
2906
+ "loss": 0.046,
2907
+ "step": 4790
2908
+ },
2909
+ {
2910
+ "epoch": 2.26,
2911
+ "learning_rate": 2.2034843205574915e-05,
2912
+ "loss": 0.042,
2913
+ "step": 4800
2914
+ },
2915
+ {
2916
+ "epoch": 2.26,
2917
+ "learning_rate": 2.1895470383275265e-05,
2918
+ "loss": 0.0436,
2919
+ "step": 4810
2920
+ },
2921
+ {
2922
+ "epoch": 2.27,
2923
+ "learning_rate": 2.1756097560975615e-05,
2924
+ "loss": 0.0398,
2925
+ "step": 4820
2926
+ },
2927
+ {
2928
+ "epoch": 2.27,
2929
+ "learning_rate": 2.1616724738675958e-05,
2930
+ "loss": 0.0396,
2931
+ "step": 4830
2932
+ },
2933
+ {
2934
+ "epoch": 2.28,
2935
+ "learning_rate": 2.1477351916376307e-05,
2936
+ "loss": 0.0474,
2937
+ "step": 4840
2938
+ },
2939
+ {
2940
+ "epoch": 2.28,
2941
+ "learning_rate": 2.1337979094076657e-05,
2942
+ "loss": 0.0405,
2943
+ "step": 4850
2944
+ },
2945
+ {
2946
+ "epoch": 2.29,
2947
+ "learning_rate": 2.1198606271777006e-05,
2948
+ "loss": 0.0473,
2949
+ "step": 4860
2950
+ },
2951
+ {
2952
+ "epoch": 2.29,
2953
+ "learning_rate": 2.1059233449477353e-05,
2954
+ "loss": 0.044,
2955
+ "step": 4870
2956
+ },
2957
+ {
2958
+ "epoch": 2.3,
2959
+ "learning_rate": 2.0919860627177702e-05,
2960
+ "loss": 0.0388,
2961
+ "step": 4880
2962
+ },
2963
+ {
2964
+ "epoch": 2.3,
2965
+ "learning_rate": 2.0780487804878052e-05,
2966
+ "loss": 0.0403,
2967
+ "step": 4890
2968
+ },
2969
+ {
2970
+ "epoch": 2.3,
2971
+ "learning_rate": 2.06411149825784e-05,
2972
+ "loss": 0.041,
2973
+ "step": 4900
2974
+ },
2975
+ {
2976
+ "epoch": 2.31,
2977
+ "learning_rate": 2.0501742160278744e-05,
2978
+ "loss": 0.0429,
2979
+ "step": 4910
2980
+ },
2981
+ {
2982
+ "epoch": 2.31,
2983
+ "learning_rate": 2.0362369337979094e-05,
2984
+ "loss": 0.0363,
2985
+ "step": 4920
2986
+ },
2987
+ {
2988
+ "epoch": 2.32,
2989
+ "learning_rate": 2.0222996515679444e-05,
2990
+ "loss": 0.0414,
2991
+ "step": 4930
2992
+ },
2993
+ {
2994
+ "epoch": 2.32,
2995
+ "learning_rate": 2.0083623693379793e-05,
2996
+ "loss": 0.0515,
2997
+ "step": 4940
2998
+ },
2999
+ {
3000
+ "epoch": 2.33,
3001
+ "learning_rate": 1.9944250871080143e-05,
3002
+ "loss": 0.0447,
3003
+ "step": 4950
3004
+ },
3005
+ {
3006
+ "epoch": 2.33,
3007
+ "learning_rate": 1.980487804878049e-05,
3008
+ "loss": 0.0545,
3009
+ "step": 4960
3010
+ },
3011
+ {
3012
+ "epoch": 2.34,
3013
+ "learning_rate": 1.966550522648084e-05,
3014
+ "loss": 0.0416,
3015
+ "step": 4970
3016
+ },
3017
+ {
3018
+ "epoch": 2.34,
3019
+ "learning_rate": 1.9526132404181185e-05,
3020
+ "loss": 0.0387,
3021
+ "step": 4980
3022
+ },
3023
+ {
3024
+ "epoch": 2.35,
3025
+ "learning_rate": 1.9386759581881535e-05,
3026
+ "loss": 0.0332,
3027
+ "step": 4990
3028
+ },
3029
+ {
3030
+ "epoch": 2.35,
3031
+ "learning_rate": 1.9247386759581884e-05,
3032
+ "loss": 0.0409,
3033
+ "step": 5000
3034
+ },
3035
+ {
3036
+ "epoch": 2.36,
3037
+ "learning_rate": 1.910801393728223e-05,
3038
+ "loss": 0.0371,
3039
+ "step": 5010
3040
+ },
3041
+ {
3042
+ "epoch": 2.36,
3043
+ "learning_rate": 1.896864111498258e-05,
3044
+ "loss": 0.0448,
3045
+ "step": 5020
3046
+ },
3047
+ {
3048
+ "epoch": 2.37,
3049
+ "learning_rate": 1.8829268292682926e-05,
3050
+ "loss": 0.0517,
3051
+ "step": 5030
3052
+ },
3053
+ {
3054
+ "epoch": 2.37,
3055
+ "learning_rate": 1.8689895470383276e-05,
3056
+ "loss": 0.0436,
3057
+ "step": 5040
3058
+ },
3059
+ {
3060
+ "epoch": 2.38,
3061
+ "learning_rate": 1.8550522648083626e-05,
3062
+ "loss": 0.0394,
3063
+ "step": 5050
3064
+ },
3065
+ {
3066
+ "epoch": 2.38,
3067
+ "learning_rate": 1.8411149825783975e-05,
3068
+ "loss": 0.046,
3069
+ "step": 5060
3070
+ },
3071
+ {
3072
+ "epoch": 2.38,
3073
+ "learning_rate": 1.827177700348432e-05,
3074
+ "loss": 0.0414,
3075
+ "step": 5070
3076
+ },
3077
+ {
3078
+ "epoch": 2.39,
3079
+ "learning_rate": 1.813240418118467e-05,
3080
+ "loss": 0.0338,
3081
+ "step": 5080
3082
+ },
3083
+ {
3084
+ "epoch": 2.39,
3085
+ "learning_rate": 1.7993031358885017e-05,
3086
+ "loss": 0.046,
3087
+ "step": 5090
3088
+ },
3089
+ {
3090
+ "epoch": 2.4,
3091
+ "learning_rate": 1.7853658536585367e-05,
3092
+ "loss": 0.0481,
3093
+ "step": 5100
3094
+ },
3095
+ {
3096
+ "epoch": 2.4,
3097
+ "learning_rate": 1.7714285714285717e-05,
3098
+ "loss": 0.0428,
3099
+ "step": 5110
3100
+ },
3101
+ {
3102
+ "epoch": 2.41,
3103
+ "learning_rate": 1.7574912891986066e-05,
3104
+ "loss": 0.0365,
3105
+ "step": 5120
3106
+ },
3107
+ {
3108
+ "epoch": 2.41,
3109
+ "learning_rate": 1.7435540069686412e-05,
3110
+ "loss": 0.046,
3111
+ "step": 5130
3112
+ },
3113
+ {
3114
+ "epoch": 2.42,
3115
+ "learning_rate": 1.7296167247386762e-05,
3116
+ "loss": 0.0357,
3117
+ "step": 5140
3118
+ },
3119
+ {
3120
+ "epoch": 2.42,
3121
+ "learning_rate": 1.7156794425087108e-05,
3122
+ "loss": 0.0387,
3123
+ "step": 5150
3124
+ },
3125
+ {
3126
+ "epoch": 2.43,
3127
+ "learning_rate": 1.7017421602787458e-05,
3128
+ "loss": 0.0364,
3129
+ "step": 5160
3130
+ },
3131
+ {
3132
+ "epoch": 2.43,
3133
+ "learning_rate": 1.6878048780487804e-05,
3134
+ "loss": 0.041,
3135
+ "step": 5170
3136
+ },
3137
+ {
3138
+ "epoch": 2.44,
3139
+ "learning_rate": 1.6738675958188154e-05,
3140
+ "loss": 0.0431,
3141
+ "step": 5180
3142
+ },
3143
+ {
3144
+ "epoch": 2.44,
3145
+ "learning_rate": 1.6599303135888503e-05,
3146
+ "loss": 0.0456,
3147
+ "step": 5190
3148
+ },
3149
+ {
3150
+ "epoch": 2.45,
3151
+ "learning_rate": 1.6459930313588853e-05,
3152
+ "loss": 0.0446,
3153
+ "step": 5200
3154
+ },
3155
+ {
3156
+ "epoch": 2.45,
3157
+ "learning_rate": 1.63205574912892e-05,
3158
+ "loss": 0.0349,
3159
+ "step": 5210
3160
+ },
3161
+ {
3162
+ "epoch": 2.46,
3163
+ "learning_rate": 1.618118466898955e-05,
3164
+ "loss": 0.0405,
3165
+ "step": 5220
3166
+ },
3167
+ {
3168
+ "epoch": 2.46,
3169
+ "learning_rate": 1.6041811846689895e-05,
3170
+ "loss": 0.0447,
3171
+ "step": 5230
3172
+ },
3173
+ {
3174
+ "epoch": 2.46,
3175
+ "learning_rate": 1.5902439024390245e-05,
3176
+ "loss": 0.0359,
3177
+ "step": 5240
3178
+ },
3179
+ {
3180
+ "epoch": 2.47,
3181
+ "learning_rate": 1.5763066202090594e-05,
3182
+ "loss": 0.0391,
3183
+ "step": 5250
3184
+ },
3185
+ {
3186
+ "epoch": 2.47,
3187
+ "learning_rate": 1.5623693379790944e-05,
3188
+ "loss": 0.0412,
3189
+ "step": 5260
3190
+ },
3191
+ {
3192
+ "epoch": 2.48,
3193
+ "learning_rate": 1.548432055749129e-05,
3194
+ "loss": 0.0384,
3195
+ "step": 5270
3196
+ },
3197
+ {
3198
+ "epoch": 2.48,
3199
+ "learning_rate": 1.534494773519164e-05,
3200
+ "loss": 0.0417,
3201
+ "step": 5280
3202
+ },
3203
+ {
3204
+ "epoch": 2.49,
3205
+ "learning_rate": 1.5205574912891988e-05,
3206
+ "loss": 0.0442,
3207
+ "step": 5290
3208
+ },
3209
+ {
3210
+ "epoch": 2.49,
3211
+ "learning_rate": 1.5066202090592337e-05,
3212
+ "loss": 0.0462,
3213
+ "step": 5300
3214
+ },
3215
+ {
3216
+ "epoch": 2.5,
3217
+ "learning_rate": 1.4926829268292684e-05,
3218
+ "loss": 0.0357,
3219
+ "step": 5310
3220
+ },
3221
+ {
3222
+ "epoch": 2.5,
3223
+ "learning_rate": 1.4787456445993033e-05,
3224
+ "loss": 0.0431,
3225
+ "step": 5320
3226
+ },
3227
+ {
3228
+ "epoch": 2.51,
3229
+ "learning_rate": 1.4648083623693381e-05,
3230
+ "loss": 0.0464,
3231
+ "step": 5330
3232
+ },
3233
+ {
3234
+ "epoch": 2.51,
3235
+ "learning_rate": 1.4508710801393729e-05,
3236
+ "loss": 0.0324,
3237
+ "step": 5340
3238
+ },
3239
+ {
3240
+ "epoch": 2.52,
3241
+ "learning_rate": 1.4369337979094079e-05,
3242
+ "loss": 0.046,
3243
+ "step": 5350
3244
+ },
3245
+ {
3246
+ "epoch": 2.52,
3247
+ "learning_rate": 1.4229965156794425e-05,
3248
+ "loss": 0.0494,
3249
+ "step": 5360
3250
+ },
3251
+ {
3252
+ "epoch": 2.53,
3253
+ "learning_rate": 1.4090592334494775e-05,
3254
+ "loss": 0.0406,
3255
+ "step": 5370
3256
+ },
3257
+ {
3258
+ "epoch": 2.53,
3259
+ "learning_rate": 1.3951219512195122e-05,
3260
+ "loss": 0.0474,
3261
+ "step": 5380
3262
+ },
3263
+ {
3264
+ "epoch": 2.54,
3265
+ "learning_rate": 1.3811846689895472e-05,
3266
+ "loss": 0.0434,
3267
+ "step": 5390
3268
+ },
3269
+ {
3270
+ "epoch": 2.54,
3271
+ "learning_rate": 1.3672473867595818e-05,
3272
+ "loss": 0.0502,
3273
+ "step": 5400
3274
+ },
3275
+ {
3276
+ "epoch": 2.54,
3277
+ "learning_rate": 1.3533101045296168e-05,
3278
+ "loss": 0.0461,
3279
+ "step": 5410
3280
+ },
3281
+ {
3282
+ "epoch": 2.55,
3283
+ "learning_rate": 1.3393728222996516e-05,
3284
+ "loss": 0.0414,
3285
+ "step": 5420
3286
+ },
3287
+ {
3288
+ "epoch": 2.55,
3289
+ "learning_rate": 1.3254355400696866e-05,
3290
+ "loss": 0.0386,
3291
+ "step": 5430
3292
+ },
3293
+ {
3294
+ "epoch": 2.56,
3295
+ "learning_rate": 1.3114982578397213e-05,
3296
+ "loss": 0.0377,
3297
+ "step": 5440
3298
+ },
3299
+ {
3300
+ "epoch": 2.56,
3301
+ "learning_rate": 1.2975609756097563e-05,
3302
+ "loss": 0.0501,
3303
+ "step": 5450
3304
+ },
3305
+ {
3306
+ "epoch": 2.57,
3307
+ "learning_rate": 1.283623693379791e-05,
3308
+ "loss": 0.0433,
3309
+ "step": 5460
3310
+ },
3311
+ {
3312
+ "epoch": 2.57,
3313
+ "learning_rate": 1.2696864111498259e-05,
3314
+ "loss": 0.0405,
3315
+ "step": 5470
3316
+ },
3317
+ {
3318
+ "epoch": 2.58,
3319
+ "learning_rate": 1.2557491289198607e-05,
3320
+ "loss": 0.044,
3321
+ "step": 5480
3322
+ },
3323
+ {
3324
+ "epoch": 2.58,
3325
+ "learning_rate": 1.2418118466898957e-05,
3326
+ "loss": 0.0463,
3327
+ "step": 5490
3328
+ },
3329
+ {
3330
+ "epoch": 2.59,
3331
+ "learning_rate": 1.2278745644599304e-05,
3332
+ "loss": 0.0442,
3333
+ "step": 5500
3334
+ },
3335
+ {
3336
+ "epoch": 2.59,
3337
+ "learning_rate": 1.2139372822299652e-05,
3338
+ "loss": 0.0422,
3339
+ "step": 5510
3340
+ },
3341
+ {
3342
+ "epoch": 2.6,
3343
+ "learning_rate": 1.2e-05,
3344
+ "loss": 0.0366,
3345
+ "step": 5520
3346
+ },
3347
+ {
3348
+ "epoch": 2.6,
3349
+ "learning_rate": 1.186062717770035e-05,
3350
+ "loss": 0.0283,
3351
+ "step": 5530
3352
+ },
3353
+ {
3354
+ "epoch": 2.61,
3355
+ "learning_rate": 1.1721254355400698e-05,
3356
+ "loss": 0.034,
3357
+ "step": 5540
3358
+ },
3359
+ {
3360
+ "epoch": 2.61,
3361
+ "learning_rate": 1.1581881533101047e-05,
3362
+ "loss": 0.0441,
3363
+ "step": 5550
3364
+ },
3365
+ {
3366
+ "epoch": 2.62,
3367
+ "learning_rate": 1.1442508710801394e-05,
3368
+ "loss": 0.0364,
3369
+ "step": 5560
3370
+ },
3371
+ {
3372
+ "epoch": 2.62,
3373
+ "learning_rate": 1.1303135888501743e-05,
3374
+ "loss": 0.0535,
3375
+ "step": 5570
3376
+ },
3377
+ {
3378
+ "epoch": 2.62,
3379
+ "learning_rate": 1.1163763066202091e-05,
3380
+ "loss": 0.0502,
3381
+ "step": 5580
3382
+ },
3383
+ {
3384
+ "epoch": 2.63,
3385
+ "learning_rate": 1.1024390243902441e-05,
3386
+ "loss": 0.0441,
3387
+ "step": 5590
3388
+ },
3389
+ {
3390
+ "epoch": 2.63,
3391
+ "learning_rate": 1.0885017421602789e-05,
3392
+ "loss": 0.0359,
3393
+ "step": 5600
3394
+ },
3395
+ {
3396
+ "epoch": 2.64,
3397
+ "learning_rate": 1.0745644599303138e-05,
3398
+ "loss": 0.0343,
3399
+ "step": 5610
3400
+ },
3401
+ {
3402
+ "epoch": 2.64,
3403
+ "learning_rate": 1.0606271777003485e-05,
3404
+ "loss": 0.0461,
3405
+ "step": 5620
3406
+ },
3407
+ {
3408
+ "epoch": 2.65,
3409
+ "learning_rate": 1.0466898954703834e-05,
3410
+ "loss": 0.0396,
3411
+ "step": 5630
3412
+ },
3413
+ {
3414
+ "epoch": 2.65,
3415
+ "learning_rate": 1.0327526132404182e-05,
3416
+ "loss": 0.0348,
3417
+ "step": 5640
3418
+ },
3419
+ {
3420
+ "epoch": 2.66,
3421
+ "learning_rate": 1.018815331010453e-05,
3422
+ "loss": 0.0428,
3423
+ "step": 5650
3424
+ },
3425
+ {
3426
+ "epoch": 2.66,
3427
+ "learning_rate": 1.0048780487804878e-05,
3428
+ "loss": 0.0394,
3429
+ "step": 5660
3430
+ },
3431
+ {
3432
+ "epoch": 2.67,
3433
+ "learning_rate": 9.909407665505228e-06,
3434
+ "loss": 0.0348,
3435
+ "step": 5670
3436
+ },
3437
+ {
3438
+ "epoch": 2.67,
3439
+ "learning_rate": 9.770034843205576e-06,
3440
+ "loss": 0.0484,
3441
+ "step": 5680
3442
+ },
3443
+ {
3444
+ "epoch": 2.68,
3445
+ "learning_rate": 9.630662020905924e-06,
3446
+ "loss": 0.0452,
3447
+ "step": 5690
3448
+ },
3449
+ {
3450
+ "epoch": 2.68,
3451
+ "learning_rate": 9.491289198606273e-06,
3452
+ "loss": 0.0365,
3453
+ "step": 5700
3454
+ },
3455
+ {
3456
+ "epoch": 2.69,
3457
+ "learning_rate": 9.351916376306621e-06,
3458
+ "loss": 0.0365,
3459
+ "step": 5710
3460
+ },
3461
+ {
3462
+ "epoch": 2.69,
3463
+ "learning_rate": 9.212543554006969e-06,
3464
+ "loss": 0.0368,
3465
+ "step": 5720
3466
+ },
3467
+ {
3468
+ "epoch": 2.7,
3469
+ "learning_rate": 9.073170731707319e-06,
3470
+ "loss": 0.051,
3471
+ "step": 5730
3472
+ },
3473
+ {
3474
+ "epoch": 2.7,
3475
+ "learning_rate": 8.933797909407667e-06,
3476
+ "loss": 0.0454,
3477
+ "step": 5740
3478
+ },
3479
+ {
3480
+ "epoch": 2.7,
3481
+ "learning_rate": 8.794425087108015e-06,
3482
+ "loss": 0.0429,
3483
+ "step": 5750
3484
+ },
3485
+ {
3486
+ "epoch": 2.71,
3487
+ "learning_rate": 8.655052264808364e-06,
3488
+ "loss": 0.0384,
3489
+ "step": 5760
3490
+ },
3491
+ {
3492
+ "epoch": 2.71,
3493
+ "learning_rate": 8.515679442508712e-06,
3494
+ "loss": 0.0391,
3495
+ "step": 5770
3496
+ },
3497
+ {
3498
+ "epoch": 2.72,
3499
+ "learning_rate": 8.37630662020906e-06,
3500
+ "loss": 0.0351,
3501
+ "step": 5780
3502
+ },
3503
+ {
3504
+ "epoch": 2.72,
3505
+ "learning_rate": 8.23693379790941e-06,
3506
+ "loss": 0.05,
3507
+ "step": 5790
3508
+ },
3509
+ {
3510
+ "epoch": 2.73,
3511
+ "learning_rate": 8.097560975609758e-06,
3512
+ "loss": 0.0361,
3513
+ "step": 5800
3514
+ },
3515
+ {
3516
+ "epoch": 2.73,
3517
+ "learning_rate": 7.958188153310104e-06,
3518
+ "loss": 0.0365,
3519
+ "step": 5810
3520
+ },
3521
+ {
3522
+ "epoch": 2.74,
3523
+ "learning_rate": 7.818815331010453e-06,
3524
+ "loss": 0.0414,
3525
+ "step": 5820
3526
+ },
3527
+ {
3528
+ "epoch": 2.74,
3529
+ "learning_rate": 7.679442508710801e-06,
3530
+ "loss": 0.0513,
3531
+ "step": 5830
3532
+ },
3533
+ {
3534
+ "epoch": 2.75,
3535
+ "learning_rate": 7.54006968641115e-06,
3536
+ "loss": 0.0406,
3537
+ "step": 5840
3538
+ },
3539
+ {
3540
+ "epoch": 2.75,
3541
+ "learning_rate": 7.400696864111498e-06,
3542
+ "loss": 0.0449,
3543
+ "step": 5850
3544
+ },
3545
+ {
3546
+ "epoch": 2.76,
3547
+ "learning_rate": 7.261324041811847e-06,
3548
+ "loss": 0.0413,
3549
+ "step": 5860
3550
+ },
3551
+ {
3552
+ "epoch": 2.76,
3553
+ "learning_rate": 7.121951219512196e-06,
3554
+ "loss": 0.0389,
3555
+ "step": 5870
3556
+ },
3557
+ {
3558
+ "epoch": 2.77,
3559
+ "learning_rate": 6.982578397212544e-06,
3560
+ "loss": 0.0362,
3561
+ "step": 5880
3562
+ },
3563
+ {
3564
+ "epoch": 2.77,
3565
+ "learning_rate": 6.843205574912892e-06,
3566
+ "loss": 0.0418,
3567
+ "step": 5890
3568
+ },
3569
+ {
3570
+ "epoch": 2.78,
3571
+ "learning_rate": 6.703832752613241e-06,
3572
+ "loss": 0.0365,
3573
+ "step": 5900
3574
+ },
3575
+ {
3576
+ "epoch": 2.78,
3577
+ "learning_rate": 6.564459930313589e-06,
3578
+ "loss": 0.0377,
3579
+ "step": 5910
3580
+ },
3581
+ {
3582
+ "epoch": 2.78,
3583
+ "learning_rate": 6.425087108013938e-06,
3584
+ "loss": 0.0476,
3585
+ "step": 5920
3586
+ },
3587
+ {
3588
+ "epoch": 2.79,
3589
+ "learning_rate": 6.285714285714286e-06,
3590
+ "loss": 0.0453,
3591
+ "step": 5930
3592
+ },
3593
+ {
3594
+ "epoch": 2.79,
3595
+ "learning_rate": 6.1463414634146346e-06,
3596
+ "loss": 0.0412,
3597
+ "step": 5940
3598
+ },
3599
+ {
3600
+ "epoch": 2.8,
3601
+ "learning_rate": 6.006968641114983e-06,
3602
+ "loss": 0.0493,
3603
+ "step": 5950
3604
+ },
3605
+ {
3606
+ "epoch": 2.8,
3607
+ "learning_rate": 5.867595818815331e-06,
3608
+ "loss": 0.0429,
3609
+ "step": 5960
3610
+ },
3611
+ {
3612
+ "epoch": 2.81,
3613
+ "learning_rate": 5.72822299651568e-06,
3614
+ "loss": 0.0406,
3615
+ "step": 5970
3616
+ },
3617
+ {
3618
+ "epoch": 2.81,
3619
+ "learning_rate": 5.588850174216028e-06,
3620
+ "loss": 0.0423,
3621
+ "step": 5980
3622
+ },
3623
+ {
3624
+ "epoch": 2.82,
3625
+ "learning_rate": 5.449477351916377e-06,
3626
+ "loss": 0.0353,
3627
+ "step": 5990
3628
+ },
3629
+ {
3630
+ "epoch": 2.82,
3631
+ "learning_rate": 5.3101045296167255e-06,
3632
+ "loss": 0.0362,
3633
+ "step": 6000
3634
+ },
3635
+ {
3636
+ "epoch": 2.83,
3637
+ "learning_rate": 5.1707317073170735e-06,
3638
+ "loss": 0.0478,
3639
+ "step": 6010
3640
+ },
3641
+ {
3642
+ "epoch": 2.83,
3643
+ "learning_rate": 5.031358885017422e-06,
3644
+ "loss": 0.0431,
3645
+ "step": 6020
3646
+ },
3647
+ {
3648
+ "epoch": 2.84,
3649
+ "learning_rate": 4.891986062717771e-06,
3650
+ "loss": 0.0354,
3651
+ "step": 6030
3652
+ },
3653
+ {
3654
+ "epoch": 2.84,
3655
+ "learning_rate": 4.752613240418119e-06,
3656
+ "loss": 0.0414,
3657
+ "step": 6040
3658
+ },
3659
+ {
3660
+ "epoch": 2.85,
3661
+ "learning_rate": 4.613240418118467e-06,
3662
+ "loss": 0.0397,
3663
+ "step": 6050
3664
+ },
3665
+ {
3666
+ "epoch": 2.85,
3667
+ "learning_rate": 4.473867595818816e-06,
3668
+ "loss": 0.0351,
3669
+ "step": 6060
3670
+ },
3671
+ {
3672
+ "epoch": 2.86,
3673
+ "learning_rate": 4.334494773519164e-06,
3674
+ "loss": 0.0331,
3675
+ "step": 6070
3676
+ },
3677
+ {
3678
+ "epoch": 2.86,
3679
+ "learning_rate": 4.195121951219512e-06,
3680
+ "loss": 0.0423,
3681
+ "step": 6080
3682
+ },
3683
+ {
3684
+ "epoch": 2.86,
3685
+ "learning_rate": 4.055749128919861e-06,
3686
+ "loss": 0.042,
3687
+ "step": 6090
3688
+ },
3689
+ {
3690
+ "epoch": 2.87,
3691
+ "learning_rate": 3.916376306620209e-06,
3692
+ "loss": 0.0285,
3693
+ "step": 6100
3694
+ },
3695
+ {
3696
+ "epoch": 2.87,
3697
+ "learning_rate": 3.777003484320558e-06,
3698
+ "loss": 0.0447,
3699
+ "step": 6110
3700
+ },
3701
+ {
3702
+ "epoch": 2.88,
3703
+ "learning_rate": 3.6376306620209062e-06,
3704
+ "loss": 0.0389,
3705
+ "step": 6120
3706
+ },
3707
+ {
3708
+ "epoch": 2.88,
3709
+ "learning_rate": 3.4982578397212546e-06,
3710
+ "loss": 0.0395,
3711
+ "step": 6130
3712
+ },
3713
+ {
3714
+ "epoch": 2.89,
3715
+ "learning_rate": 3.358885017421603e-06,
3716
+ "loss": 0.0364,
3717
+ "step": 6140
3718
+ },
3719
+ {
3720
+ "epoch": 2.89,
3721
+ "learning_rate": 3.2195121951219517e-06,
3722
+ "loss": 0.054,
3723
+ "step": 6150
3724
+ },
3725
+ {
3726
+ "epoch": 2.9,
3727
+ "learning_rate": 3.0801393728223e-06,
3728
+ "loss": 0.0442,
3729
+ "step": 6160
3730
+ },
3731
+ {
3732
+ "epoch": 2.9,
3733
+ "learning_rate": 2.9407665505226484e-06,
3734
+ "loss": 0.0413,
3735
+ "step": 6170
3736
+ },
3737
+ {
3738
+ "epoch": 2.91,
3739
+ "learning_rate": 2.8013937282229968e-06,
3740
+ "loss": 0.0406,
3741
+ "step": 6180
3742
+ },
3743
+ {
3744
+ "epoch": 2.91,
3745
+ "learning_rate": 2.6620209059233455e-06,
3746
+ "loss": 0.047,
3747
+ "step": 6190
3748
+ },
3749
+ {
3750
+ "epoch": 2.92,
3751
+ "learning_rate": 2.522648083623694e-06,
3752
+ "loss": 0.0382,
3753
+ "step": 6200
3754
+ },
3755
+ {
3756
+ "epoch": 2.92,
3757
+ "learning_rate": 2.383275261324042e-06,
3758
+ "loss": 0.043,
3759
+ "step": 6210
3760
+ },
3761
+ {
3762
+ "epoch": 2.93,
3763
+ "learning_rate": 2.2439024390243906e-06,
3764
+ "loss": 0.0435,
3765
+ "step": 6220
3766
+ },
3767
+ {
3768
+ "epoch": 2.93,
3769
+ "learning_rate": 2.104529616724739e-06,
3770
+ "loss": 0.0436,
3771
+ "step": 6230
3772
+ },
3773
+ {
3774
+ "epoch": 2.94,
3775
+ "learning_rate": 1.9651567944250873e-06,
3776
+ "loss": 0.0472,
3777
+ "step": 6240
3778
+ },
3779
+ {
3780
+ "epoch": 2.94,
3781
+ "learning_rate": 1.8257839721254357e-06,
3782
+ "loss": 0.0368,
3783
+ "step": 6250
3784
+ },
3785
+ {
3786
+ "epoch": 2.94,
3787
+ "learning_rate": 1.686411149825784e-06,
3788
+ "loss": 0.038,
3789
+ "step": 6260
3790
+ },
3791
+ {
3792
+ "epoch": 2.95,
3793
+ "learning_rate": 1.5470383275261324e-06,
3794
+ "loss": 0.0437,
3795
+ "step": 6270
3796
+ },
3797
+ {
3798
+ "epoch": 2.95,
3799
+ "learning_rate": 1.407665505226481e-06,
3800
+ "loss": 0.0353,
3801
+ "step": 6280
3802
+ },
3803
+ {
3804
+ "epoch": 2.96,
3805
+ "learning_rate": 1.2682926829268293e-06,
3806
+ "loss": 0.044,
3807
+ "step": 6290
3808
+ },
3809
+ {
3810
+ "epoch": 2.96,
3811
+ "learning_rate": 1.1289198606271779e-06,
3812
+ "loss": 0.0358,
3813
+ "step": 6300
3814
+ },
3815
+ {
3816
+ "epoch": 2.97,
3817
+ "learning_rate": 9.895470383275262e-07,
3818
+ "loss": 0.0378,
3819
+ "step": 6310
3820
+ },
3821
+ {
3822
+ "epoch": 2.97,
3823
+ "learning_rate": 8.501742160278746e-07,
3824
+ "loss": 0.0346,
3825
+ "step": 6320
3826
+ },
3827
+ {
3828
+ "epoch": 2.98,
3829
+ "learning_rate": 7.10801393728223e-07,
3830
+ "loss": 0.0456,
3831
+ "step": 6330
3832
+ },
3833
+ {
3834
+ "epoch": 2.98,
3835
+ "learning_rate": 5.714285714285715e-07,
3836
+ "loss": 0.0494,
3837
+ "step": 6340
3838
+ },
3839
+ {
3840
+ "epoch": 2.99,
3841
+ "learning_rate": 4.320557491289199e-07,
3842
+ "loss": 0.0406,
3843
+ "step": 6350
3844
+ },
3845
+ {
3846
+ "epoch": 2.99,
3847
+ "learning_rate": 2.926829268292683e-07,
3848
+ "loss": 0.0386,
3849
+ "step": 6360
3850
+ },
3851
+ {
3852
+ "epoch": 3.0,
3853
+ "learning_rate": 1.5331010452961674e-07,
3854
+ "loss": 0.0415,
3855
+ "step": 6370
3856
+ },
3857
+ {
3858
+ "epoch": 3.0,
3859
+ "eval_accuracy": 0.9769126125154315,
3860
+ "eval_f1": 0.7849694196330357,
3861
+ "eval_loss": 0.06933891773223877,
3862
+ "eval_precision": 0.7739696312364425,
3863
+ "eval_recall": 0.7962863774326013,
3864
+ "eval_runtime": 14.2581,
3865
+ "eval_samples_per_second": 993.122,
3866
+ "eval_steps_per_second": 124.14,
3867
+ "step": 6378
3868
+ },
3869
+ {
3870
+ "epoch": 3.0,
3871
+ "step": 6378,
3872
+ "total_flos": 1.2518895383371872e+16,
3873
+ "train_loss": 0.06623676680927928,
3874
+ "train_runtime": 486.7718,
3875
+ "train_samples_per_second": 628.742,
3876
+ "train_steps_per_second": 13.103
3877
+ }
3878
+ ],
3879
+ "logging_steps": 10,
3880
+ "max_steps": 6378,
3881
+ "num_train_epochs": 3,
3882
+ "save_steps": 500,
3883
+ "total_flos": 1.2518895383371872e+16,
3884
+ "trial_name": null,
3885
+ "trial_params": null
3886
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ba1317f20a1857e524cc648dba69abb42081abb0655cd8877745eb7546d10e47
3
+ size 4091