Hugging Face's logo

usmiva
/
bert-web-bg

Fill-Mask
Transformers
PyTorch
Safetensors
Bulgarian
bert
feature-extraction
Generated from Trainer
Model card Files
xet
Community
1
usmiva commited on
Commit
76c04f1
·
1 Parent(s): c9b213f

Upload 13 files

Browse files
Files changed (13) hide show
  1. README.md +53 -1
  2. added_tokens.json +3 -0
  3. all_results.json +15 -0
  4. config.json +25 -0
  5. eval_results.json +10 -0
  6. pytorch_model.bin +3 -0
  7. special_tokens_map.json +7 -0
  8. tokenizer.json +0 -0
  9. tokenizer_config.json +15 -0
  10. train_results.json +8 -0
  11. trainer_state.json +3907 -0
  12. training_args.bin +3 -0
  13. vocab.txt +0 -0
README.md CHANGED
@@ -1,3 +1,55 @@
1
  ---
2
- license: apache-2.0
 
 
 
 
 
 
3
  ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  ---
2
+ tags:
3
+ - generated_from_trainer
4
+ metrics:
5
+ - accuracy
6
+ model-index:
7
+ - name: bg-news-bert
8
+ results: []
9
  ---
10
+
11
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
12
+ should probably proofread and complete it, then remove this comment. -->
13
+
14
+ # bg-news-bert
15
+
16
+ This model is a fine-tuned version of [](https://huggingface.co/) on an unknown dataset.
17
+ It achieves the following results on the evaluation set:
18
+ - Loss: 1.4510
19
+ - Accuracy: 0.6906
20
+
21
+ ## Model description
22
+
23
+ More information needed
24
+
25
+ ## Intended uses & limitations
26
+
27
+ More information needed
28
+
29
+ ## Training and evaluation data
30
+
31
+ More information needed
32
+
33
+ ## Training procedure
34
+
35
+ ### Training hyperparameters
36
+
37
+ The following hyperparameters were used during training:
38
+ - learning_rate: 5e-05
39
+ - train_batch_size: 32
40
+ - eval_batch_size: 32
41
+ - seed: 42
42
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
43
+ - lr_scheduler_type: linear
44
+ - num_epochs: 3.0
45
+
46
+ ### Training results
47
+
48
+
49
+
50
+ ### Framework versions
51
+
52
+ - Transformers 4.22.0.dev0
53
+ - Pytorch 1.11.0
54
+ - Datasets 2.2.1
55
+ - Tokenizers 0.12.1
added_tokens.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ {
2
+ "[PAD]": 30000
3
+ }
all_results.json ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 3.0,
3
+ "eval_accuracy": 0.6906063124235521,
4
+ "eval_loss": 1.4509799480438232,
5
+ "eval_runtime": 5230.4957,
6
+ "eval_samples": 432388,
7
+ "eval_samples_per_second": 82.667,
8
+ "eval_steps_per_second": 2.584,
9
+ "perplexity": 4.267294193497874,
10
+ "train_loss": 3.0939811468297327,
11
+ "train_runtime": 276726.3152,
12
+ "train_samples": 3455772,
13
+ "train_samples_per_second": 37.464,
14
+ "train_steps_per_second": 1.171
15
+ }
config.json ADDED
@@ -0,0 +1,25 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "/home/ivamarinova/bg-news-bert/checkpoint-323500/",
3
+ "architectures": [
4
+ "BertModel"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.1,
7
+ "classifier_dropout": null,
8
+ "hidden_act": "gelu",
9
+ "hidden_dropout_prob": 0.1,
10
+ "hidden_size": 768,
11
+ "initializer_range": 0.02,
12
+ "intermediate_size": 3072,
13
+ "layer_norm_eps": 1e-12,
14
+ "max_position_embeddings": 512,
15
+ "model_type": "bert",
16
+ "num_attention_heads": 12,
17
+ "num_hidden_layers": 12,
18
+ "pad_token_id": 0,
19
+ "position_embedding_type": "absolute",
20
+ "torch_dtype": "float32",
21
+ "transformers_version": "4.22.0.dev0",
22
+ "type_vocab_size": 2,
23
+ "use_cache": true,
24
+ "vocab_size": 30001
25
+ }
eval_results.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 3.0,
3
+ "eval_accuracy": 0.6906063124235521,
4
+ "eval_loss": 1.4509799480438232,
5
+ "eval_runtime": 5230.4957,
6
+ "eval_samples": 432388,
7
+ "eval_samples_per_second": 82.667,
8
+ "eval_steps_per_second": 2.584,
9
+ "perplexity": 4.267294193497874
10
+ }
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7956df50261fe0553affc7702b7c007f983307f4011928756f3e1a9c78cda855
3
+ size 436395057
special_tokens_map.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "cls_token": "[CLS]",
3
+ "mask_token": "[MASK]",
4
+ "pad_token": "[PAD]",
5
+ "sep_token": "[SEP]",
6
+ "unk_token": "[UNK]"
7
+ }
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer_config.json ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "cls_token": "[CLS]",
3
+ "do_basic_tokenize": true,
4
+ "do_lower_case": true,
5
+ "mask_token": "[MASK]",
6
+ "name_or_path": "/home/ivamarinova/bert-bg",
7
+ "never_split": null,
8
+ "pad_token": "[PAD]",
9
+ "sep_token": "[SEP]",
10
+ "special_tokens_map_file": null,
11
+ "strip_accents": null,
12
+ "tokenize_chinese_chars": true,
13
+ "tokenizer_class": "BertTokenizer",
14
+ "unk_token": "[UNK]"
15
+ }
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 3.0,
3
+ "train_loss": 3.0939811468297327,
4
+ "train_runtime": 276726.3152,
5
+ "train_samples": 3455772,
6
+ "train_samples_per_second": 37.464,
7
+ "train_steps_per_second": 1.171
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,3907 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 3.0,
5
+ "global_step": 323979,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.0,
12
+ "learning_rate": 4.992283450470556e-05,
13
+ "loss": 7.2271,
14
+ "step": 500
15
+ },
16
+ {
17
+ "epoch": 0.01,
18
+ "learning_rate": 4.98456690094111e-05,
19
+ "loss": 6.7719,
20
+ "step": 1000
21
+ },
22
+ {
23
+ "epoch": 0.01,
24
+ "learning_rate": 4.976850351411666e-05,
25
+ "loss": 6.6629,
26
+ "step": 1500
27
+ },
28
+ {
29
+ "epoch": 0.02,
30
+ "learning_rate": 4.969133801882221e-05,
31
+ "loss": 6.6057,
32
+ "step": 2000
33
+ },
34
+ {
35
+ "epoch": 0.02,
36
+ "learning_rate": 4.961417252352776e-05,
37
+ "loss": 6.5557,
38
+ "step": 2500
39
+ },
40
+ {
41
+ "epoch": 0.03,
42
+ "learning_rate": 4.953700702823331e-05,
43
+ "loss": 6.5339,
44
+ "step": 3000
45
+ },
46
+ {
47
+ "epoch": 0.03,
48
+ "learning_rate": 4.945984153293887e-05,
49
+ "loss": 6.4981,
50
+ "step": 3500
51
+ },
52
+ {
53
+ "epoch": 0.04,
54
+ "learning_rate": 4.938267603764441e-05,
55
+ "loss": 6.4864,
56
+ "step": 4000
57
+ },
58
+ {
59
+ "epoch": 0.04,
60
+ "learning_rate": 4.930551054234997e-05,
61
+ "loss": 6.4665,
62
+ "step": 4500
63
+ },
64
+ {
65
+ "epoch": 0.05,
66
+ "learning_rate": 4.922834504705553e-05,
67
+ "loss": 6.4375,
68
+ "step": 5000
69
+ },
70
+ {
71
+ "epoch": 0.05,
72
+ "learning_rate": 4.915117955176107e-05,
73
+ "loss": 6.4367,
74
+ "step": 5500
75
+ },
76
+ {
77
+ "epoch": 0.06,
78
+ "learning_rate": 4.907401405646663e-05,
79
+ "loss": 6.4276,
80
+ "step": 6000
81
+ },
82
+ {
83
+ "epoch": 0.06,
84
+ "learning_rate": 4.899684856117218e-05,
85
+ "loss": 6.4075,
86
+ "step": 6500
87
+ },
88
+ {
89
+ "epoch": 0.06,
90
+ "learning_rate": 4.891968306587773e-05,
91
+ "loss": 6.3899,
92
+ "step": 7000
93
+ },
94
+ {
95
+ "epoch": 0.07,
96
+ "learning_rate": 4.884251757058328e-05,
97
+ "loss": 6.3858,
98
+ "step": 7500
99
+ },
100
+ {
101
+ "epoch": 0.07,
102
+ "learning_rate": 4.8765352075288836e-05,
103
+ "loss": 6.3791,
104
+ "step": 8000
105
+ },
106
+ {
107
+ "epoch": 0.08,
108
+ "learning_rate": 4.868818657999438e-05,
109
+ "loss": 6.3676,
110
+ "step": 8500
111
+ },
112
+ {
113
+ "epoch": 0.08,
114
+ "learning_rate": 4.861102108469994e-05,
115
+ "loss": 6.3539,
116
+ "step": 9000
117
+ },
118
+ {
119
+ "epoch": 0.09,
120
+ "learning_rate": 4.853385558940549e-05,
121
+ "loss": 6.3511,
122
+ "step": 9500
123
+ },
124
+ {
125
+ "epoch": 0.09,
126
+ "learning_rate": 4.845669009411104e-05,
127
+ "loss": 6.3434,
128
+ "step": 10000
129
+ },
130
+ {
131
+ "epoch": 0.1,
132
+ "learning_rate": 4.8379524598816595e-05,
133
+ "loss": 6.3393,
134
+ "step": 10500
135
+ },
136
+ {
137
+ "epoch": 0.1,
138
+ "learning_rate": 4.8302359103522146e-05,
139
+ "loss": 6.3319,
140
+ "step": 11000
141
+ },
142
+ {
143
+ "epoch": 0.11,
144
+ "learning_rate": 4.8225193608227696e-05,
145
+ "loss": 6.3237,
146
+ "step": 11500
147
+ },
148
+ {
149
+ "epoch": 0.11,
150
+ "learning_rate": 4.8148028112933246e-05,
151
+ "loss": 6.3252,
152
+ "step": 12000
153
+ },
154
+ {
155
+ "epoch": 0.12,
156
+ "learning_rate": 4.8070862617638804e-05,
157
+ "loss": 6.31,
158
+ "step": 12500
159
+ },
160
+ {
161
+ "epoch": 0.12,
162
+ "learning_rate": 4.799369712234435e-05,
163
+ "loss": 6.3092,
164
+ "step": 13000
165
+ },
166
+ {
167
+ "epoch": 0.13,
168
+ "learning_rate": 4.7916531627049905e-05,
169
+ "loss": 6.3098,
170
+ "step": 13500
171
+ },
172
+ {
173
+ "epoch": 0.13,
174
+ "learning_rate": 4.7839366131755455e-05,
175
+ "loss": 6.3044,
176
+ "step": 14000
177
+ },
178
+ {
179
+ "epoch": 0.13,
180
+ "learning_rate": 4.7762200636461005e-05,
181
+ "loss": 6.297,
182
+ "step": 14500
183
+ },
184
+ {
185
+ "epoch": 0.14,
186
+ "learning_rate": 4.768503514116656e-05,
187
+ "loss": 6.2934,
188
+ "step": 15000
189
+ },
190
+ {
191
+ "epoch": 0.14,
192
+ "learning_rate": 4.760786964587211e-05,
193
+ "loss": 6.2945,
194
+ "step": 15500
195
+ },
196
+ {
197
+ "epoch": 0.15,
198
+ "learning_rate": 4.7530704150577663e-05,
199
+ "loss": 6.2808,
200
+ "step": 16000
201
+ },
202
+ {
203
+ "epoch": 0.15,
204
+ "learning_rate": 4.7453538655283214e-05,
205
+ "loss": 6.2806,
206
+ "step": 16500
207
+ },
208
+ {
209
+ "epoch": 0.16,
210
+ "learning_rate": 4.737637315998877e-05,
211
+ "loss": 6.2805,
212
+ "step": 17000
213
+ },
214
+ {
215
+ "epoch": 0.16,
216
+ "learning_rate": 4.7299207664694315e-05,
217
+ "loss": 6.2715,
218
+ "step": 17500
219
+ },
220
+ {
221
+ "epoch": 0.17,
222
+ "learning_rate": 4.722204216939987e-05,
223
+ "loss": 6.264,
224
+ "step": 18000
225
+ },
226
+ {
227
+ "epoch": 0.17,
228
+ "learning_rate": 4.714487667410542e-05,
229
+ "loss": 6.2593,
230
+ "step": 18500
231
+ },
232
+ {
233
+ "epoch": 0.18,
234
+ "learning_rate": 4.706771117881097e-05,
235
+ "loss": 6.2522,
236
+ "step": 19000
237
+ },
238
+ {
239
+ "epoch": 0.18,
240
+ "learning_rate": 4.699054568351653e-05,
241
+ "loss": 6.2625,
242
+ "step": 19500
243
+ },
244
+ {
245
+ "epoch": 0.19,
246
+ "learning_rate": 4.691338018822208e-05,
247
+ "loss": 6.2572,
248
+ "step": 20000
249
+ },
250
+ {
251
+ "epoch": 0.19,
252
+ "learning_rate": 4.683621469292763e-05,
253
+ "loss": 6.2562,
254
+ "step": 20500
255
+ },
256
+ {
257
+ "epoch": 0.19,
258
+ "learning_rate": 4.675904919763318e-05,
259
+ "loss": 6.2434,
260
+ "step": 21000
261
+ },
262
+ {
263
+ "epoch": 0.2,
264
+ "learning_rate": 4.668188370233874e-05,
265
+ "loss": 6.2465,
266
+ "step": 21500
267
+ },
268
+ {
269
+ "epoch": 0.2,
270
+ "learning_rate": 4.660471820704428e-05,
271
+ "loss": 6.2381,
272
+ "step": 22000
273
+ },
274
+ {
275
+ "epoch": 0.21,
276
+ "learning_rate": 4.652755271174984e-05,
277
+ "loss": 6.2361,
278
+ "step": 22500
279
+ },
280
+ {
281
+ "epoch": 0.21,
282
+ "learning_rate": 4.645038721645539e-05,
283
+ "loss": 6.2318,
284
+ "step": 23000
285
+ },
286
+ {
287
+ "epoch": 0.22,
288
+ "learning_rate": 4.637322172116094e-05,
289
+ "loss": 6.2407,
290
+ "step": 23500
291
+ },
292
+ {
293
+ "epoch": 0.22,
294
+ "learning_rate": 4.629605622586649e-05,
295
+ "loss": 6.2403,
296
+ "step": 24000
297
+ },
298
+ {
299
+ "epoch": 0.23,
300
+ "learning_rate": 4.621889073057205e-05,
301
+ "loss": 6.2229,
302
+ "step": 24500
303
+ },
304
+ {
305
+ "epoch": 0.23,
306
+ "learning_rate": 4.61417252352776e-05,
307
+ "loss": 6.231,
308
+ "step": 25000
309
+ },
310
+ {
311
+ "epoch": 0.24,
312
+ "learning_rate": 4.606455973998315e-05,
313
+ "loss": 6.2295,
314
+ "step": 25500
315
+ },
316
+ {
317
+ "epoch": 0.24,
318
+ "learning_rate": 4.5987394244688706e-05,
319
+ "loss": 6.2117,
320
+ "step": 26000
321
+ },
322
+ {
323
+ "epoch": 0.25,
324
+ "learning_rate": 4.591022874939425e-05,
325
+ "loss": 6.2233,
326
+ "step": 26500
327
+ },
328
+ {
329
+ "epoch": 0.25,
330
+ "learning_rate": 4.583306325409981e-05,
331
+ "loss": 6.2193,
332
+ "step": 27000
333
+ },
334
+ {
335
+ "epoch": 0.25,
336
+ "learning_rate": 4.575589775880536e-05,
337
+ "loss": 6.2166,
338
+ "step": 27500
339
+ },
340
+ {
341
+ "epoch": 0.26,
342
+ "learning_rate": 4.567873226351091e-05,
343
+ "loss": 6.2093,
344
+ "step": 28000
345
+ },
346
+ {
347
+ "epoch": 0.26,
348
+ "learning_rate": 4.560156676821646e-05,
349
+ "loss": 6.2025,
350
+ "step": 28500
351
+ },
352
+ {
353
+ "epoch": 0.27,
354
+ "learning_rate": 4.5524401272922015e-05,
355
+ "loss": 6.2056,
356
+ "step": 29000
357
+ },
358
+ {
359
+ "epoch": 0.27,
360
+ "learning_rate": 4.5447235777627566e-05,
361
+ "loss": 6.2018,
362
+ "step": 29500
363
+ },
364
+ {
365
+ "epoch": 0.28,
366
+ "learning_rate": 4.5370070282333116e-05,
367
+ "loss": 6.2111,
368
+ "step": 30000
369
+ },
370
+ {
371
+ "epoch": 0.28,
372
+ "learning_rate": 4.529290478703867e-05,
373
+ "loss": 6.2045,
374
+ "step": 30500
375
+ },
376
+ {
377
+ "epoch": 0.29,
378
+ "learning_rate": 4.521573929174422e-05,
379
+ "loss": 6.1887,
380
+ "step": 31000
381
+ },
382
+ {
383
+ "epoch": 0.29,
384
+ "learning_rate": 4.5138573796449774e-05,
385
+ "loss": 6.1956,
386
+ "step": 31500
387
+ },
388
+ {
389
+ "epoch": 0.3,
390
+ "learning_rate": 4.5061408301155324e-05,
391
+ "loss": 6.2002,
392
+ "step": 32000
393
+ },
394
+ {
395
+ "epoch": 0.3,
396
+ "learning_rate": 4.4984242805860875e-05,
397
+ "loss": 6.2027,
398
+ "step": 32500
399
+ },
400
+ {
401
+ "epoch": 0.31,
402
+ "learning_rate": 4.4907077310566425e-05,
403
+ "loss": 6.1877,
404
+ "step": 33000
405
+ },
406
+ {
407
+ "epoch": 0.31,
408
+ "learning_rate": 4.482991181527198e-05,
409
+ "loss": 6.1894,
410
+ "step": 33500
411
+ },
412
+ {
413
+ "epoch": 0.31,
414
+ "learning_rate": 4.475274631997753e-05,
415
+ "loss": 6.1861,
416
+ "step": 34000
417
+ },
418
+ {
419
+ "epoch": 0.32,
420
+ "learning_rate": 4.467558082468308e-05,
421
+ "loss": 6.1839,
422
+ "step": 34500
423
+ },
424
+ {
425
+ "epoch": 0.32,
426
+ "learning_rate": 4.459841532938864e-05,
427
+ "loss": 6.1783,
428
+ "step": 35000
429
+ },
430
+ {
431
+ "epoch": 0.33,
432
+ "learning_rate": 4.4521249834094184e-05,
433
+ "loss": 6.1894,
434
+ "step": 35500
435
+ },
436
+ {
437
+ "epoch": 0.33,
438
+ "learning_rate": 4.444408433879974e-05,
439
+ "loss": 6.1811,
440
+ "step": 36000
441
+ },
442
+ {
443
+ "epoch": 0.34,
444
+ "learning_rate": 4.436691884350529e-05,
445
+ "loss": 6.1822,
446
+ "step": 36500
447
+ },
448
+ {
449
+ "epoch": 0.34,
450
+ "learning_rate": 4.428975334821084e-05,
451
+ "loss": 6.1739,
452
+ "step": 37000
453
+ },
454
+ {
455
+ "epoch": 0.35,
456
+ "learning_rate": 4.421258785291639e-05,
457
+ "loss": 6.1715,
458
+ "step": 37500
459
+ },
460
+ {
461
+ "epoch": 0.35,
462
+ "learning_rate": 4.413542235762195e-05,
463
+ "loss": 6.1779,
464
+ "step": 38000
465
+ },
466
+ {
467
+ "epoch": 0.36,
468
+ "learning_rate": 4.4058256862327493e-05,
469
+ "loss": 6.1735,
470
+ "step": 38500
471
+ },
472
+ {
473
+ "epoch": 0.36,
474
+ "learning_rate": 4.398109136703305e-05,
475
+ "loss": 6.1723,
476
+ "step": 39000
477
+ },
478
+ {
479
+ "epoch": 0.37,
480
+ "learning_rate": 4.390392587173861e-05,
481
+ "loss": 6.1703,
482
+ "step": 39500
483
+ },
484
+ {
485
+ "epoch": 0.37,
486
+ "learning_rate": 4.382676037644415e-05,
487
+ "loss": 6.1738,
488
+ "step": 40000
489
+ },
490
+ {
491
+ "epoch": 0.38,
492
+ "learning_rate": 4.374959488114971e-05,
493
+ "loss": 6.1656,
494
+ "step": 40500
495
+ },
496
+ {
497
+ "epoch": 0.38,
498
+ "learning_rate": 4.367242938585526e-05,
499
+ "loss": 6.1722,
500
+ "step": 41000
501
+ },
502
+ {
503
+ "epoch": 0.38,
504
+ "learning_rate": 4.359526389056081e-05,
505
+ "loss": 6.1643,
506
+ "step": 41500
507
+ },
508
+ {
509
+ "epoch": 0.39,
510
+ "learning_rate": 4.351809839526636e-05,
511
+ "loss": 6.1708,
512
+ "step": 42000
513
+ },
514
+ {
515
+ "epoch": 0.39,
516
+ "learning_rate": 4.344093289997192e-05,
517
+ "loss": 6.1624,
518
+ "step": 42500
519
+ },
520
+ {
521
+ "epoch": 0.4,
522
+ "learning_rate": 4.336376740467746e-05,
523
+ "loss": 6.1644,
524
+ "step": 43000
525
+ },
526
+ {
527
+ "epoch": 0.4,
528
+ "learning_rate": 4.328660190938302e-05,
529
+ "loss": 6.1659,
530
+ "step": 43500
531
+ },
532
+ {
533
+ "epoch": 0.41,
534
+ "learning_rate": 4.3209436414088575e-05,
535
+ "loss": 6.1557,
536
+ "step": 44000
537
+ },
538
+ {
539
+ "epoch": 0.41,
540
+ "learning_rate": 4.313227091879412e-05,
541
+ "loss": 6.1448,
542
+ "step": 44500
543
+ },
544
+ {
545
+ "epoch": 0.42,
546
+ "learning_rate": 4.3055105423499676e-05,
547
+ "loss": 6.1532,
548
+ "step": 45000
549
+ },
550
+ {
551
+ "epoch": 0.42,
552
+ "learning_rate": 4.2977939928205227e-05,
553
+ "loss": 6.1445,
554
+ "step": 45500
555
+ },
556
+ {
557
+ "epoch": 0.43,
558
+ "learning_rate": 4.290077443291078e-05,
559
+ "loss": 6.1419,
560
+ "step": 46000
561
+ },
562
+ {
563
+ "epoch": 0.43,
564
+ "learning_rate": 4.282360893761633e-05,
565
+ "loss": 6.1518,
566
+ "step": 46500
567
+ },
568
+ {
569
+ "epoch": 0.44,
570
+ "learning_rate": 4.2746443442321885e-05,
571
+ "loss": 6.1501,
572
+ "step": 47000
573
+ },
574
+ {
575
+ "epoch": 0.44,
576
+ "learning_rate": 4.266927794702743e-05,
577
+ "loss": 6.142,
578
+ "step": 47500
579
+ },
580
+ {
581
+ "epoch": 0.44,
582
+ "learning_rate": 4.2592112451732985e-05,
583
+ "loss": 6.1523,
584
+ "step": 48000
585
+ },
586
+ {
587
+ "epoch": 0.45,
588
+ "learning_rate": 4.2514946956438536e-05,
589
+ "loss": 6.1383,
590
+ "step": 48500
591
+ },
592
+ {
593
+ "epoch": 0.45,
594
+ "learning_rate": 4.2437781461144086e-05,
595
+ "loss": 6.1358,
596
+ "step": 49000
597
+ },
598
+ {
599
+ "epoch": 0.46,
600
+ "learning_rate": 4.2360615965849643e-05,
601
+ "loss": 6.1378,
602
+ "step": 49500
603
+ },
604
+ {
605
+ "epoch": 0.46,
606
+ "learning_rate": 4.2283450470555194e-05,
607
+ "loss": 6.1415,
608
+ "step": 50000
609
+ },
610
+ {
611
+ "epoch": 0.47,
612
+ "learning_rate": 4.2206284975260744e-05,
613
+ "loss": 6.1407,
614
+ "step": 50500
615
+ },
616
+ {
617
+ "epoch": 0.47,
618
+ "learning_rate": 4.2129119479966295e-05,
619
+ "loss": 6.1405,
620
+ "step": 51000
621
+ },
622
+ {
623
+ "epoch": 0.48,
624
+ "learning_rate": 4.205195398467185e-05,
625
+ "loss": 6.1407,
626
+ "step": 51500
627
+ },
628
+ {
629
+ "epoch": 0.48,
630
+ "learning_rate": 4.1974788489377396e-05,
631
+ "loss": 6.1412,
632
+ "step": 52000
633
+ },
634
+ {
635
+ "epoch": 0.49,
636
+ "learning_rate": 4.189762299408295e-05,
637
+ "loss": 6.1413,
638
+ "step": 52500
639
+ },
640
+ {
641
+ "epoch": 0.49,
642
+ "learning_rate": 4.18204574987885e-05,
643
+ "loss": 6.1347,
644
+ "step": 53000
645
+ },
646
+ {
647
+ "epoch": 0.5,
648
+ "learning_rate": 4.1743292003494054e-05,
649
+ "loss": 6.1364,
650
+ "step": 53500
651
+ },
652
+ {
653
+ "epoch": 0.5,
654
+ "learning_rate": 4.166612650819961e-05,
655
+ "loss": 6.1401,
656
+ "step": 54000
657
+ },
658
+ {
659
+ "epoch": 0.5,
660
+ "learning_rate": 4.158896101290516e-05,
661
+ "loss": 6.1357,
662
+ "step": 54500
663
+ },
664
+ {
665
+ "epoch": 0.51,
666
+ "learning_rate": 4.151179551761071e-05,
667
+ "loss": 6.1317,
668
+ "step": 55000
669
+ },
670
+ {
671
+ "epoch": 0.51,
672
+ "learning_rate": 4.143463002231626e-05,
673
+ "loss": 6.1287,
674
+ "step": 55500
675
+ },
676
+ {
677
+ "epoch": 0.52,
678
+ "learning_rate": 4.135746452702182e-05,
679
+ "loss": 6.1378,
680
+ "step": 56000
681
+ },
682
+ {
683
+ "epoch": 0.52,
684
+ "learning_rate": 4.128029903172736e-05,
685
+ "loss": 6.1365,
686
+ "step": 56500
687
+ },
688
+ {
689
+ "epoch": 0.53,
690
+ "learning_rate": 4.120313353643292e-05,
691
+ "loss": 6.1393,
692
+ "step": 57000
693
+ },
694
+ {
695
+ "epoch": 0.53,
696
+ "learning_rate": 4.112596804113847e-05,
697
+ "loss": 6.1246,
698
+ "step": 57500
699
+ },
700
+ {
701
+ "epoch": 0.54,
702
+ "learning_rate": 4.104880254584402e-05,
703
+ "loss": 6.1184,
704
+ "step": 58000
705
+ },
706
+ {
707
+ "epoch": 0.54,
708
+ "learning_rate": 4.097163705054958e-05,
709
+ "loss": 6.1164,
710
+ "step": 58500
711
+ },
712
+ {
713
+ "epoch": 0.55,
714
+ "learning_rate": 4.089447155525513e-05,
715
+ "loss": 6.1308,
716
+ "step": 59000
717
+ },
718
+ {
719
+ "epoch": 0.55,
720
+ "learning_rate": 4.081730605996068e-05,
721
+ "loss": 6.1179,
722
+ "step": 59500
723
+ },
724
+ {
725
+ "epoch": 0.56,
726
+ "learning_rate": 4.074014056466623e-05,
727
+ "loss": 6.125,
728
+ "step": 60000
729
+ },
730
+ {
731
+ "epoch": 0.56,
732
+ "learning_rate": 4.066297506937179e-05,
733
+ "loss": 6.123,
734
+ "step": 60500
735
+ },
736
+ {
737
+ "epoch": 0.56,
738
+ "learning_rate": 4.058580957407733e-05,
739
+ "loss": 6.1191,
740
+ "step": 61000
741
+ },
742
+ {
743
+ "epoch": 0.57,
744
+ "learning_rate": 4.050864407878289e-05,
745
+ "loss": 6.1176,
746
+ "step": 61500
747
+ },
748
+ {
749
+ "epoch": 0.57,
750
+ "learning_rate": 4.043147858348844e-05,
751
+ "loss": 6.1186,
752
+ "step": 62000
753
+ },
754
+ {
755
+ "epoch": 0.58,
756
+ "learning_rate": 4.035431308819399e-05,
757
+ "loss": 6.1197,
758
+ "step": 62500
759
+ },
760
+ {
761
+ "epoch": 0.58,
762
+ "learning_rate": 4.027714759289954e-05,
763
+ "loss": 6.1212,
764
+ "step": 63000
765
+ },
766
+ {
767
+ "epoch": 0.59,
768
+ "learning_rate": 4.0199982097605096e-05,
769
+ "loss": 6.1219,
770
+ "step": 63500
771
+ },
772
+ {
773
+ "epoch": 0.59,
774
+ "learning_rate": 4.0122816602310646e-05,
775
+ "loss": 6.1169,
776
+ "step": 64000
777
+ },
778
+ {
779
+ "epoch": 0.6,
780
+ "learning_rate": 4.00456511070162e-05,
781
+ "loss": 6.1149,
782
+ "step": 64500
783
+ },
784
+ {
785
+ "epoch": 0.6,
786
+ "learning_rate": 3.9968485611721754e-05,
787
+ "loss": 6.118,
788
+ "step": 65000
789
+ },
790
+ {
791
+ "epoch": 0.61,
792
+ "learning_rate": 3.98913201164273e-05,
793
+ "loss": 6.1218,
794
+ "step": 65500
795
+ },
796
+ {
797
+ "epoch": 0.61,
798
+ "learning_rate": 3.9814154621132855e-05,
799
+ "loss": 6.11,
800
+ "step": 66000
801
+ },
802
+ {
803
+ "epoch": 0.62,
804
+ "learning_rate": 3.9736989125838405e-05,
805
+ "loss": 6.1164,
806
+ "step": 66500
807
+ },
808
+ {
809
+ "epoch": 0.62,
810
+ "learning_rate": 3.9659823630543956e-05,
811
+ "loss": 6.1142,
812
+ "step": 67000
813
+ },
814
+ {
815
+ "epoch": 0.63,
816
+ "learning_rate": 3.9582658135249506e-05,
817
+ "loss": 6.1067,
818
+ "step": 67500
819
+ },
820
+ {
821
+ "epoch": 0.63,
822
+ "learning_rate": 3.9505492639955063e-05,
823
+ "loss": 6.1179,
824
+ "step": 68000
825
+ },
826
+ {
827
+ "epoch": 0.63,
828
+ "learning_rate": 3.9428327144660614e-05,
829
+ "loss": 6.1031,
830
+ "step": 68500
831
+ },
832
+ {
833
+ "epoch": 0.64,
834
+ "learning_rate": 3.9351161649366164e-05,
835
+ "loss": 6.1052,
836
+ "step": 69000
837
+ },
838
+ {
839
+ "epoch": 0.64,
840
+ "learning_rate": 3.927399615407172e-05,
841
+ "loss": 6.1057,
842
+ "step": 69500
843
+ },
844
+ {
845
+ "epoch": 0.65,
846
+ "learning_rate": 3.9196830658777265e-05,
847
+ "loss": 6.1035,
848
+ "step": 70000
849
+ },
850
+ {
851
+ "epoch": 0.65,
852
+ "learning_rate": 3.911966516348282e-05,
853
+ "loss": 6.1028,
854
+ "step": 70500
855
+ },
856
+ {
857
+ "epoch": 0.66,
858
+ "learning_rate": 3.904249966818837e-05,
859
+ "loss": 6.1133,
860
+ "step": 71000
861
+ },
862
+ {
863
+ "epoch": 0.66,
864
+ "learning_rate": 3.896533417289392e-05,
865
+ "loss": 6.1124,
866
+ "step": 71500
867
+ },
868
+ {
869
+ "epoch": 0.67,
870
+ "learning_rate": 3.8888168677599474e-05,
871
+ "loss": 6.1048,
872
+ "step": 72000
873
+ },
874
+ {
875
+ "epoch": 0.67,
876
+ "learning_rate": 3.881100318230503e-05,
877
+ "loss": 6.0969,
878
+ "step": 72500
879
+ },
880
+ {
881
+ "epoch": 0.68,
882
+ "learning_rate": 3.873383768701058e-05,
883
+ "loss": 6.1106,
884
+ "step": 73000
885
+ },
886
+ {
887
+ "epoch": 0.68,
888
+ "learning_rate": 3.865667219171613e-05,
889
+ "loss": 6.095,
890
+ "step": 73500
891
+ },
892
+ {
893
+ "epoch": 0.69,
894
+ "learning_rate": 3.857950669642169e-05,
895
+ "loss": 6.1035,
896
+ "step": 74000
897
+ },
898
+ {
899
+ "epoch": 0.69,
900
+ "learning_rate": 3.850234120112723e-05,
901
+ "loss": 6.1016,
902
+ "step": 74500
903
+ },
904
+ {
905
+ "epoch": 0.69,
906
+ "learning_rate": 3.842517570583279e-05,
907
+ "loss": 6.1119,
908
+ "step": 75000
909
+ },
910
+ {
911
+ "epoch": 0.7,
912
+ "learning_rate": 3.834801021053834e-05,
913
+ "loss": 6.1011,
914
+ "step": 75500
915
+ },
916
+ {
917
+ "epoch": 0.7,
918
+ "learning_rate": 3.827084471524389e-05,
919
+ "loss": 6.097,
920
+ "step": 76000
921
+ },
922
+ {
923
+ "epoch": 0.71,
924
+ "learning_rate": 3.819367921994944e-05,
925
+ "loss": 6.1007,
926
+ "step": 76500
927
+ },
928
+ {
929
+ "epoch": 0.71,
930
+ "learning_rate": 3.8116513724655e-05,
931
+ "loss": 6.0952,
932
+ "step": 77000
933
+ },
934
+ {
935
+ "epoch": 0.72,
936
+ "learning_rate": 3.803934822936054e-05,
937
+ "loss": 6.0097,
938
+ "step": 77500
939
+ },
940
+ {
941
+ "epoch": 0.72,
942
+ "learning_rate": 3.79621827340661e-05,
943
+ "loss": 5.9232,
944
+ "step": 78000
945
+ },
946
+ {
947
+ "epoch": 0.73,
948
+ "learning_rate": 3.7885017238771656e-05,
949
+ "loss": 5.8762,
950
+ "step": 78500
951
+ },
952
+ {
953
+ "epoch": 0.73,
954
+ "learning_rate": 3.78078517434772e-05,
955
+ "loss": 5.8308,
956
+ "step": 79000
957
+ },
958
+ {
959
+ "epoch": 0.74,
960
+ "learning_rate": 3.773068624818276e-05,
961
+ "loss": 5.7992,
962
+ "step": 79500
963
+ },
964
+ {
965
+ "epoch": 0.74,
966
+ "learning_rate": 3.765352075288831e-05,
967
+ "loss": 5.761,
968
+ "step": 80000
969
+ },
970
+ {
971
+ "epoch": 0.75,
972
+ "learning_rate": 3.757635525759386e-05,
973
+ "loss": 5.7269,
974
+ "step": 80500
975
+ },
976
+ {
977
+ "epoch": 0.75,
978
+ "learning_rate": 3.749918976229941e-05,
979
+ "loss": 5.7042,
980
+ "step": 81000
981
+ },
982
+ {
983
+ "epoch": 0.75,
984
+ "learning_rate": 3.7422024267004966e-05,
985
+ "loss": 5.6826,
986
+ "step": 81500
987
+ },
988
+ {
989
+ "epoch": 0.76,
990
+ "learning_rate": 3.734485877171051e-05,
991
+ "loss": 5.6541,
992
+ "step": 82000
993
+ },
994
+ {
995
+ "epoch": 0.76,
996
+ "learning_rate": 3.7267693276416066e-05,
997
+ "loss": 5.6237,
998
+ "step": 82500
999
+ },
1000
+ {
1001
+ "epoch": 0.77,
1002
+ "learning_rate": 3.719052778112162e-05,
1003
+ "loss": 5.6091,
1004
+ "step": 83000
1005
+ },
1006
+ {
1007
+ "epoch": 0.77,
1008
+ "learning_rate": 3.711336228582717e-05,
1009
+ "loss": 5.5772,
1010
+ "step": 83500
1011
+ },
1012
+ {
1013
+ "epoch": 0.78,
1014
+ "learning_rate": 3.7036196790532724e-05,
1015
+ "loss": 5.5444,
1016
+ "step": 84000
1017
+ },
1018
+ {
1019
+ "epoch": 0.78,
1020
+ "learning_rate": 3.6959031295238275e-05,
1021
+ "loss": 5.5268,
1022
+ "step": 84500
1023
+ },
1024
+ {
1025
+ "epoch": 0.79,
1026
+ "learning_rate": 3.6881865799943825e-05,
1027
+ "loss": 5.4856,
1028
+ "step": 85000
1029
+ },
1030
+ {
1031
+ "epoch": 0.79,
1032
+ "learning_rate": 3.6804700304649376e-05,
1033
+ "loss": 5.4607,
1034
+ "step": 85500
1035
+ },
1036
+ {
1037
+ "epoch": 0.8,
1038
+ "learning_rate": 3.672753480935493e-05,
1039
+ "loss": 5.4276,
1040
+ "step": 86000
1041
+ },
1042
+ {
1043
+ "epoch": 0.8,
1044
+ "learning_rate": 3.6650369314060477e-05,
1045
+ "loss": 5.4144,
1046
+ "step": 86500
1047
+ },
1048
+ {
1049
+ "epoch": 0.81,
1050
+ "learning_rate": 3.6573203818766034e-05,
1051
+ "loss": 5.3782,
1052
+ "step": 87000
1053
+ },
1054
+ {
1055
+ "epoch": 0.81,
1056
+ "learning_rate": 3.6496038323471584e-05,
1057
+ "loss": 5.3337,
1058
+ "step": 87500
1059
+ },
1060
+ {
1061
+ "epoch": 0.81,
1062
+ "learning_rate": 3.6418872828177135e-05,
1063
+ "loss": 5.2709,
1064
+ "step": 88000
1065
+ },
1066
+ {
1067
+ "epoch": 0.82,
1068
+ "learning_rate": 3.634170733288269e-05,
1069
+ "loss": 5.2095,
1070
+ "step": 88500
1071
+ },
1072
+ {
1073
+ "epoch": 0.82,
1074
+ "learning_rate": 3.626454183758824e-05,
1075
+ "loss": 5.138,
1076
+ "step": 89000
1077
+ },
1078
+ {
1079
+ "epoch": 0.83,
1080
+ "learning_rate": 3.618737634229379e-05,
1081
+ "loss": 4.9959,
1082
+ "step": 89500
1083
+ },
1084
+ {
1085
+ "epoch": 0.83,
1086
+ "learning_rate": 3.611021084699934e-05,
1087
+ "loss": 4.8709,
1088
+ "step": 90000
1089
+ },
1090
+ {
1091
+ "epoch": 0.84,
1092
+ "learning_rate": 3.60330453517049e-05,
1093
+ "loss": 4.7619,
1094
+ "step": 90500
1095
+ },
1096
+ {
1097
+ "epoch": 0.84,
1098
+ "learning_rate": 3.5955879856410444e-05,
1099
+ "loss": 4.6625,
1100
+ "step": 91000
1101
+ },
1102
+ {
1103
+ "epoch": 0.85,
1104
+ "learning_rate": 3.5878714361116e-05,
1105
+ "loss": 4.5585,
1106
+ "step": 91500
1107
+ },
1108
+ {
1109
+ "epoch": 0.85,
1110
+ "learning_rate": 3.580154886582155e-05,
1111
+ "loss": 4.4651,
1112
+ "step": 92000
1113
+ },
1114
+ {
1115
+ "epoch": 0.86,
1116
+ "learning_rate": 3.57243833705271e-05,
1117
+ "loss": 4.3757,
1118
+ "step": 92500
1119
+ },
1120
+ {
1121
+ "epoch": 0.86,
1122
+ "learning_rate": 3.564721787523266e-05,
1123
+ "loss": 4.298,
1124
+ "step": 93000
1125
+ },
1126
+ {
1127
+ "epoch": 0.87,
1128
+ "learning_rate": 3.557005237993821e-05,
1129
+ "loss": 4.2163,
1130
+ "step": 93500
1131
+ },
1132
+ {
1133
+ "epoch": 0.87,
1134
+ "learning_rate": 3.549288688464376e-05,
1135
+ "loss": 4.1503,
1136
+ "step": 94000
1137
+ },
1138
+ {
1139
+ "epoch": 0.88,
1140
+ "learning_rate": 3.541572138934931e-05,
1141
+ "loss": 4.0701,
1142
+ "step": 94500
1143
+ },
1144
+ {
1145
+ "epoch": 0.88,
1146
+ "learning_rate": 3.533855589405487e-05,
1147
+ "loss": 4.0051,
1148
+ "step": 95000
1149
+ },
1150
+ {
1151
+ "epoch": 0.88,
1152
+ "learning_rate": 3.526139039876041e-05,
1153
+ "loss": 3.9148,
1154
+ "step": 95500
1155
+ },
1156
+ {
1157
+ "epoch": 0.89,
1158
+ "learning_rate": 3.518422490346597e-05,
1159
+ "loss": 3.7963,
1160
+ "step": 96000
1161
+ },
1162
+ {
1163
+ "epoch": 0.89,
1164
+ "learning_rate": 3.510705940817152e-05,
1165
+ "loss": 3.6717,
1166
+ "step": 96500
1167
+ },
1168
+ {
1169
+ "epoch": 0.9,
1170
+ "learning_rate": 3.502989391287707e-05,
1171
+ "loss": 3.5825,
1172
+ "step": 97000
1173
+ },
1174
+ {
1175
+ "epoch": 0.9,
1176
+ "learning_rate": 3.4952728417582627e-05,
1177
+ "loss": 3.5186,
1178
+ "step": 97500
1179
+ },
1180
+ {
1181
+ "epoch": 0.91,
1182
+ "learning_rate": 3.487556292228818e-05,
1183
+ "loss": 3.4746,
1184
+ "step": 98000
1185
+ },
1186
+ {
1187
+ "epoch": 0.91,
1188
+ "learning_rate": 3.479839742699373e-05,
1189
+ "loss": 3.3886,
1190
+ "step": 98500
1191
+ },
1192
+ {
1193
+ "epoch": 0.92,
1194
+ "learning_rate": 3.472123193169928e-05,
1195
+ "loss": 3.3322,
1196
+ "step": 99000
1197
+ },
1198
+ {
1199
+ "epoch": 0.92,
1200
+ "learning_rate": 3.4644066436404835e-05,
1201
+ "loss": 3.2667,
1202
+ "step": 99500
1203
+ },
1204
+ {
1205
+ "epoch": 0.93,
1206
+ "learning_rate": 3.456690094111038e-05,
1207
+ "loss": 3.2286,
1208
+ "step": 100000
1209
+ },
1210
+ {
1211
+ "epoch": 0.93,
1212
+ "learning_rate": 3.4489735445815936e-05,
1213
+ "loss": 3.1693,
1214
+ "step": 100500
1215
+ },
1216
+ {
1217
+ "epoch": 0.94,
1218
+ "learning_rate": 3.4412569950521486e-05,
1219
+ "loss": 3.1297,
1220
+ "step": 101000
1221
+ },
1222
+ {
1223
+ "epoch": 0.94,
1224
+ "learning_rate": 3.433540445522704e-05,
1225
+ "loss": 3.095,
1226
+ "step": 101500
1227
+ },
1228
+ {
1229
+ "epoch": 0.94,
1230
+ "learning_rate": 3.4258238959932594e-05,
1231
+ "loss": 3.0555,
1232
+ "step": 102000
1233
+ },
1234
+ {
1235
+ "epoch": 0.95,
1236
+ "learning_rate": 3.4181073464638144e-05,
1237
+ "loss": 3.0215,
1238
+ "step": 102500
1239
+ },
1240
+ {
1241
+ "epoch": 0.95,
1242
+ "learning_rate": 3.4103907969343695e-05,
1243
+ "loss": 2.9936,
1244
+ "step": 103000
1245
+ },
1246
+ {
1247
+ "epoch": 0.96,
1248
+ "learning_rate": 3.4026742474049245e-05,
1249
+ "loss": 2.9553,
1250
+ "step": 103500
1251
+ },
1252
+ {
1253
+ "epoch": 0.96,
1254
+ "learning_rate": 3.39495769787548e-05,
1255
+ "loss": 2.9263,
1256
+ "step": 104000
1257
+ },
1258
+ {
1259
+ "epoch": 0.97,
1260
+ "learning_rate": 3.3872411483460346e-05,
1261
+ "loss": 2.9042,
1262
+ "step": 104500
1263
+ },
1264
+ {
1265
+ "epoch": 0.97,
1266
+ "learning_rate": 3.37952459881659e-05,
1267
+ "loss": 2.8779,
1268
+ "step": 105000
1269
+ },
1270
+ {
1271
+ "epoch": 0.98,
1272
+ "learning_rate": 3.3718080492871454e-05,
1273
+ "loss": 2.8483,
1274
+ "step": 105500
1275
+ },
1276
+ {
1277
+ "epoch": 0.98,
1278
+ "learning_rate": 3.3640914997577004e-05,
1279
+ "loss": 2.8281,
1280
+ "step": 106000
1281
+ },
1282
+ {
1283
+ "epoch": 0.99,
1284
+ "learning_rate": 3.3563749502282555e-05,
1285
+ "loss": 2.8081,
1286
+ "step": 106500
1287
+ },
1288
+ {
1289
+ "epoch": 0.99,
1290
+ "learning_rate": 3.348658400698811e-05,
1291
+ "loss": 2.7802,
1292
+ "step": 107000
1293
+ },
1294
+ {
1295
+ "epoch": 1.0,
1296
+ "learning_rate": 3.340941851169366e-05,
1297
+ "loss": 2.7555,
1298
+ "step": 107500
1299
+ },
1300
+ {
1301
+ "epoch": 1.0,
1302
+ "learning_rate": 3.333225301639921e-05,
1303
+ "loss": 2.7429,
1304
+ "step": 108000
1305
+ },
1306
+ {
1307
+ "epoch": 1.0,
1308
+ "learning_rate": 3.325508752110476e-05,
1309
+ "loss": 2.723,
1310
+ "step": 108500
1311
+ },
1312
+ {
1313
+ "epoch": 1.01,
1314
+ "learning_rate": 3.3177922025810313e-05,
1315
+ "loss": 2.7106,
1316
+ "step": 109000
1317
+ },
1318
+ {
1319
+ "epoch": 1.01,
1320
+ "learning_rate": 3.310075653051587e-05,
1321
+ "loss": 2.6834,
1322
+ "step": 109500
1323
+ },
1324
+ {
1325
+ "epoch": 1.02,
1326
+ "learning_rate": 3.302359103522142e-05,
1327
+ "loss": 2.6625,
1328
+ "step": 110000
1329
+ },
1330
+ {
1331
+ "epoch": 1.02,
1332
+ "learning_rate": 3.294642553992697e-05,
1333
+ "loss": 2.6493,
1334
+ "step": 110500
1335
+ },
1336
+ {
1337
+ "epoch": 1.03,
1338
+ "learning_rate": 3.286926004463252e-05,
1339
+ "loss": 2.6323,
1340
+ "step": 111000
1341
+ },
1342
+ {
1343
+ "epoch": 1.03,
1344
+ "learning_rate": 3.279209454933808e-05,
1345
+ "loss": 2.6196,
1346
+ "step": 111500
1347
+ },
1348
+ {
1349
+ "epoch": 1.04,
1350
+ "learning_rate": 3.271492905404363e-05,
1351
+ "loss": 2.6093,
1352
+ "step": 112000
1353
+ },
1354
+ {
1355
+ "epoch": 1.04,
1356
+ "learning_rate": 3.263776355874918e-05,
1357
+ "loss": 2.5883,
1358
+ "step": 112500
1359
+ },
1360
+ {
1361
+ "epoch": 1.05,
1362
+ "learning_rate": 3.256059806345473e-05,
1363
+ "loss": 2.5722,
1364
+ "step": 113000
1365
+ },
1366
+ {
1367
+ "epoch": 1.05,
1368
+ "learning_rate": 3.248343256816028e-05,
1369
+ "loss": 2.5621,
1370
+ "step": 113500
1371
+ },
1372
+ {
1373
+ "epoch": 1.06,
1374
+ "learning_rate": 3.240626707286584e-05,
1375
+ "loss": 2.5431,
1376
+ "step": 114000
1377
+ },
1378
+ {
1379
+ "epoch": 1.06,
1380
+ "learning_rate": 3.232910157757139e-05,
1381
+ "loss": 2.5357,
1382
+ "step": 114500
1383
+ },
1384
+ {
1385
+ "epoch": 1.06,
1386
+ "learning_rate": 3.225193608227694e-05,
1387
+ "loss": 2.5253,
1388
+ "step": 115000
1389
+ },
1390
+ {
1391
+ "epoch": 1.07,
1392
+ "learning_rate": 3.217477058698249e-05,
1393
+ "loss": 2.5091,
1394
+ "step": 115500
1395
+ },
1396
+ {
1397
+ "epoch": 1.07,
1398
+ "learning_rate": 3.2097605091688046e-05,
1399
+ "loss": 2.5008,
1400
+ "step": 116000
1401
+ },
1402
+ {
1403
+ "epoch": 1.08,
1404
+ "learning_rate": 3.20204395963936e-05,
1405
+ "loss": 2.4897,
1406
+ "step": 116500
1407
+ },
1408
+ {
1409
+ "epoch": 1.08,
1410
+ "learning_rate": 3.194327410109915e-05,
1411
+ "loss": 2.4621,
1412
+ "step": 117000
1413
+ },
1414
+ {
1415
+ "epoch": 1.09,
1416
+ "learning_rate": 3.18661086058047e-05,
1417
+ "loss": 2.4584,
1418
+ "step": 117500
1419
+ },
1420
+ {
1421
+ "epoch": 1.09,
1422
+ "learning_rate": 3.178894311051025e-05,
1423
+ "loss": 2.4491,
1424
+ "step": 118000
1425
+ },
1426
+ {
1427
+ "epoch": 1.1,
1428
+ "learning_rate": 3.1711777615215805e-05,
1429
+ "loss": 2.4336,
1430
+ "step": 118500
1431
+ },
1432
+ {
1433
+ "epoch": 1.1,
1434
+ "learning_rate": 3.1634612119921356e-05,
1435
+ "loss": 2.4147,
1436
+ "step": 119000
1437
+ },
1438
+ {
1439
+ "epoch": 1.11,
1440
+ "learning_rate": 3.1557446624626906e-05,
1441
+ "loss": 2.4093,
1442
+ "step": 119500
1443
+ },
1444
+ {
1445
+ "epoch": 1.11,
1446
+ "learning_rate": 3.148028112933246e-05,
1447
+ "loss": 2.4001,
1448
+ "step": 120000
1449
+ },
1450
+ {
1451
+ "epoch": 1.12,
1452
+ "learning_rate": 3.1403115634038014e-05,
1453
+ "loss": 2.3902,
1454
+ "step": 120500
1455
+ },
1456
+ {
1457
+ "epoch": 1.12,
1458
+ "learning_rate": 3.132595013874356e-05,
1459
+ "loss": 2.3675,
1460
+ "step": 121000
1461
+ },
1462
+ {
1463
+ "epoch": 1.13,
1464
+ "learning_rate": 3.1248784643449115e-05,
1465
+ "loss": 2.3749,
1466
+ "step": 121500
1467
+ },
1468
+ {
1469
+ "epoch": 1.13,
1470
+ "learning_rate": 3.1171619148154665e-05,
1471
+ "loss": 2.3599,
1472
+ "step": 122000
1473
+ },
1474
+ {
1475
+ "epoch": 1.13,
1476
+ "learning_rate": 3.1094453652860216e-05,
1477
+ "loss": 2.3526,
1478
+ "step": 122500
1479
+ },
1480
+ {
1481
+ "epoch": 1.14,
1482
+ "learning_rate": 3.101728815756577e-05,
1483
+ "loss": 2.3341,
1484
+ "step": 123000
1485
+ },
1486
+ {
1487
+ "epoch": 1.14,
1488
+ "learning_rate": 3.094012266227132e-05,
1489
+ "loss": 2.3356,
1490
+ "step": 123500
1491
+ },
1492
+ {
1493
+ "epoch": 1.15,
1494
+ "learning_rate": 3.0862957166976874e-05,
1495
+ "loss": 2.3146,
1496
+ "step": 124000
1497
+ },
1498
+ {
1499
+ "epoch": 1.15,
1500
+ "learning_rate": 3.0785791671682424e-05,
1501
+ "loss": 2.2939,
1502
+ "step": 124500
1503
+ },
1504
+ {
1505
+ "epoch": 1.16,
1506
+ "learning_rate": 3.070862617638798e-05,
1507
+ "loss": 2.2914,
1508
+ "step": 125000
1509
+ },
1510
+ {
1511
+ "epoch": 1.16,
1512
+ "learning_rate": 3.0631460681093525e-05,
1513
+ "loss": 2.2835,
1514
+ "step": 125500
1515
+ },
1516
+ {
1517
+ "epoch": 1.17,
1518
+ "learning_rate": 3.055429518579908e-05,
1519
+ "loss": 2.2782,
1520
+ "step": 126000
1521
+ },
1522
+ {
1523
+ "epoch": 1.17,
1524
+ "learning_rate": 3.0477129690504636e-05,
1525
+ "loss": 2.272,
1526
+ "step": 126500
1527
+ },
1528
+ {
1529
+ "epoch": 1.18,
1530
+ "learning_rate": 3.0399964195210183e-05,
1531
+ "loss": 2.2664,
1532
+ "step": 127000
1533
+ },
1534
+ {
1535
+ "epoch": 1.18,
1536
+ "learning_rate": 3.032279869991574e-05,
1537
+ "loss": 2.2529,
1538
+ "step": 127500
1539
+ },
1540
+ {
1541
+ "epoch": 1.19,
1542
+ "learning_rate": 3.0245633204621287e-05,
1543
+ "loss": 2.2433,
1544
+ "step": 128000
1545
+ },
1546
+ {
1547
+ "epoch": 1.19,
1548
+ "learning_rate": 3.016846770932684e-05,
1549
+ "loss": 2.2377,
1550
+ "step": 128500
1551
+ },
1552
+ {
1553
+ "epoch": 1.19,
1554
+ "learning_rate": 3.009130221403239e-05,
1555
+ "loss": 2.226,
1556
+ "step": 129000
1557
+ },
1558
+ {
1559
+ "epoch": 1.2,
1560
+ "learning_rate": 3.0014136718737945e-05,
1561
+ "loss": 2.2204,
1562
+ "step": 129500
1563
+ },
1564
+ {
1565
+ "epoch": 1.2,
1566
+ "learning_rate": 2.9936971223443496e-05,
1567
+ "loss": 2.2053,
1568
+ "step": 130000
1569
+ },
1570
+ {
1571
+ "epoch": 1.21,
1572
+ "learning_rate": 2.985980572814905e-05,
1573
+ "loss": 2.2028,
1574
+ "step": 130500
1575
+ },
1576
+ {
1577
+ "epoch": 1.21,
1578
+ "learning_rate": 2.9782640232854596e-05,
1579
+ "loss": 2.1984,
1580
+ "step": 131000
1581
+ },
1582
+ {
1583
+ "epoch": 1.22,
1584
+ "learning_rate": 2.970547473756015e-05,
1585
+ "loss": 2.1928,
1586
+ "step": 131500
1587
+ },
1588
+ {
1589
+ "epoch": 1.22,
1590
+ "learning_rate": 2.9628309242265707e-05,
1591
+ "loss": 2.1778,
1592
+ "step": 132000
1593
+ },
1594
+ {
1595
+ "epoch": 1.23,
1596
+ "learning_rate": 2.9551143746971255e-05,
1597
+ "loss": 2.1823,
1598
+ "step": 132500
1599
+ },
1600
+ {
1601
+ "epoch": 1.23,
1602
+ "learning_rate": 2.947397825167681e-05,
1603
+ "loss": 2.1643,
1604
+ "step": 133000
1605
+ },
1606
+ {
1607
+ "epoch": 1.24,
1608
+ "learning_rate": 2.939681275638236e-05,
1609
+ "loss": 2.1642,
1610
+ "step": 133500
1611
+ },
1612
+ {
1613
+ "epoch": 1.24,
1614
+ "learning_rate": 2.9319647261087913e-05,
1615
+ "loss": 2.1537,
1616
+ "step": 134000
1617
+ },
1618
+ {
1619
+ "epoch": 1.25,
1620
+ "learning_rate": 2.9242481765793463e-05,
1621
+ "loss": 2.1488,
1622
+ "step": 134500
1623
+ },
1624
+ {
1625
+ "epoch": 1.25,
1626
+ "learning_rate": 2.9165316270499017e-05,
1627
+ "loss": 2.1471,
1628
+ "step": 135000
1629
+ },
1630
+ {
1631
+ "epoch": 1.25,
1632
+ "learning_rate": 2.9088150775204564e-05,
1633
+ "loss": 2.1342,
1634
+ "step": 135500
1635
+ },
1636
+ {
1637
+ "epoch": 1.26,
1638
+ "learning_rate": 2.9010985279910118e-05,
1639
+ "loss": 2.1308,
1640
+ "step": 136000
1641
+ },
1642
+ {
1643
+ "epoch": 1.26,
1644
+ "learning_rate": 2.8933819784615675e-05,
1645
+ "loss": 2.126,
1646
+ "step": 136500
1647
+ },
1648
+ {
1649
+ "epoch": 1.27,
1650
+ "learning_rate": 2.8856654289321222e-05,
1651
+ "loss": 2.1168,
1652
+ "step": 137000
1653
+ },
1654
+ {
1655
+ "epoch": 1.27,
1656
+ "learning_rate": 2.8779488794026776e-05,
1657
+ "loss": 2.1099,
1658
+ "step": 137500
1659
+ },
1660
+ {
1661
+ "epoch": 1.28,
1662
+ "learning_rate": 2.8702323298732326e-05,
1663
+ "loss": 2.1055,
1664
+ "step": 138000
1665
+ },
1666
+ {
1667
+ "epoch": 1.28,
1668
+ "learning_rate": 2.862515780343788e-05,
1669
+ "loss": 2.1085,
1670
+ "step": 138500
1671
+ },
1672
+ {
1673
+ "epoch": 1.29,
1674
+ "learning_rate": 2.8547992308143427e-05,
1675
+ "loss": 2.0995,
1676
+ "step": 139000
1677
+ },
1678
+ {
1679
+ "epoch": 1.29,
1680
+ "learning_rate": 2.8470826812848984e-05,
1681
+ "loss": 2.0969,
1682
+ "step": 139500
1683
+ },
1684
+ {
1685
+ "epoch": 1.3,
1686
+ "learning_rate": 2.839366131755453e-05,
1687
+ "loss": 2.0776,
1688
+ "step": 140000
1689
+ },
1690
+ {
1691
+ "epoch": 1.3,
1692
+ "learning_rate": 2.8316495822260085e-05,
1693
+ "loss": 2.0799,
1694
+ "step": 140500
1695
+ },
1696
+ {
1697
+ "epoch": 1.31,
1698
+ "learning_rate": 2.8239330326965642e-05,
1699
+ "loss": 2.0776,
1700
+ "step": 141000
1701
+ },
1702
+ {
1703
+ "epoch": 1.31,
1704
+ "learning_rate": 2.816216483167119e-05,
1705
+ "loss": 2.0776,
1706
+ "step": 141500
1707
+ },
1708
+ {
1709
+ "epoch": 1.31,
1710
+ "learning_rate": 2.8084999336376743e-05,
1711
+ "loss": 2.0655,
1712
+ "step": 142000
1713
+ },
1714
+ {
1715
+ "epoch": 1.32,
1716
+ "learning_rate": 2.8007833841082293e-05,
1717
+ "loss": 2.0537,
1718
+ "step": 142500
1719
+ },
1720
+ {
1721
+ "epoch": 1.32,
1722
+ "learning_rate": 2.7930668345787847e-05,
1723
+ "loss": 2.0508,
1724
+ "step": 143000
1725
+ },
1726
+ {
1727
+ "epoch": 1.33,
1728
+ "learning_rate": 2.7853502850493394e-05,
1729
+ "loss": 2.0511,
1730
+ "step": 143500
1731
+ },
1732
+ {
1733
+ "epoch": 1.33,
1734
+ "learning_rate": 2.777633735519895e-05,
1735
+ "loss": 2.04,
1736
+ "step": 144000
1737
+ },
1738
+ {
1739
+ "epoch": 1.34,
1740
+ "learning_rate": 2.76991718599045e-05,
1741
+ "loss": 2.0409,
1742
+ "step": 144500
1743
+ },
1744
+ {
1745
+ "epoch": 1.34,
1746
+ "learning_rate": 2.7622006364610052e-05,
1747
+ "loss": 2.037,
1748
+ "step": 145000
1749
+ },
1750
+ {
1751
+ "epoch": 1.35,
1752
+ "learning_rate": 2.7544840869315603e-05,
1753
+ "loss": 2.0314,
1754
+ "step": 145500
1755
+ },
1756
+ {
1757
+ "epoch": 1.35,
1758
+ "learning_rate": 2.7467675374021157e-05,
1759
+ "loss": 2.0304,
1760
+ "step": 146000
1761
+ },
1762
+ {
1763
+ "epoch": 1.36,
1764
+ "learning_rate": 2.739050987872671e-05,
1765
+ "loss": 2.0209,
1766
+ "step": 146500
1767
+ },
1768
+ {
1769
+ "epoch": 1.36,
1770
+ "learning_rate": 2.731334438343226e-05,
1771
+ "loss": 2.0227,
1772
+ "step": 147000
1773
+ },
1774
+ {
1775
+ "epoch": 1.37,
1776
+ "learning_rate": 2.7236178888137815e-05,
1777
+ "loss": 2.0162,
1778
+ "step": 147500
1779
+ },
1780
+ {
1781
+ "epoch": 1.37,
1782
+ "learning_rate": 2.7159013392843362e-05,
1783
+ "loss": 2.013,
1784
+ "step": 148000
1785
+ },
1786
+ {
1787
+ "epoch": 1.38,
1788
+ "learning_rate": 2.708184789754892e-05,
1789
+ "loss": 2.0051,
1790
+ "step": 148500
1791
+ },
1792
+ {
1793
+ "epoch": 1.38,
1794
+ "learning_rate": 2.7004682402254466e-05,
1795
+ "loss": 1.9949,
1796
+ "step": 149000
1797
+ },
1798
+ {
1799
+ "epoch": 1.38,
1800
+ "learning_rate": 2.692751690696002e-05,
1801
+ "loss": 2.0048,
1802
+ "step": 149500
1803
+ },
1804
+ {
1805
+ "epoch": 1.39,
1806
+ "learning_rate": 2.685035141166557e-05,
1807
+ "loss": 1.9992,
1808
+ "step": 150000
1809
+ },
1810
+ {
1811
+ "epoch": 1.39,
1812
+ "learning_rate": 2.6773185916371124e-05,
1813
+ "loss": 1.9944,
1814
+ "step": 150500
1815
+ },
1816
+ {
1817
+ "epoch": 1.4,
1818
+ "learning_rate": 2.6696020421076678e-05,
1819
+ "loss": 1.9888,
1820
+ "step": 151000
1821
+ },
1822
+ {
1823
+ "epoch": 1.4,
1824
+ "learning_rate": 2.6618854925782228e-05,
1825
+ "loss": 1.9888,
1826
+ "step": 151500
1827
+ },
1828
+ {
1829
+ "epoch": 1.41,
1830
+ "learning_rate": 2.6541689430487782e-05,
1831
+ "loss": 1.9869,
1832
+ "step": 152000
1833
+ },
1834
+ {
1835
+ "epoch": 1.41,
1836
+ "learning_rate": 2.646452393519333e-05,
1837
+ "loss": 1.988,
1838
+ "step": 152500
1839
+ },
1840
+ {
1841
+ "epoch": 1.42,
1842
+ "learning_rate": 2.6387358439898886e-05,
1843
+ "loss": 1.9722,
1844
+ "step": 153000
1845
+ },
1846
+ {
1847
+ "epoch": 1.42,
1848
+ "learning_rate": 2.6310192944604433e-05,
1849
+ "loss": 1.9684,
1850
+ "step": 153500
1851
+ },
1852
+ {
1853
+ "epoch": 1.43,
1854
+ "learning_rate": 2.6233027449309987e-05,
1855
+ "loss": 1.9693,
1856
+ "step": 154000
1857
+ },
1858
+ {
1859
+ "epoch": 1.43,
1860
+ "learning_rate": 2.6155861954015538e-05,
1861
+ "loss": 1.9706,
1862
+ "step": 154500
1863
+ },
1864
+ {
1865
+ "epoch": 1.44,
1866
+ "learning_rate": 2.607869645872109e-05,
1867
+ "loss": 1.9631,
1868
+ "step": 155000
1869
+ },
1870
+ {
1871
+ "epoch": 1.44,
1872
+ "learning_rate": 2.6001530963426645e-05,
1873
+ "loss": 1.9571,
1874
+ "step": 155500
1875
+ },
1876
+ {
1877
+ "epoch": 1.44,
1878
+ "learning_rate": 2.5924365468132196e-05,
1879
+ "loss": 1.9525,
1880
+ "step": 156000
1881
+ },
1882
+ {
1883
+ "epoch": 1.45,
1884
+ "learning_rate": 2.584719997283775e-05,
1885
+ "loss": 2.0859,
1886
+ "step": 156500
1887
+ },
1888
+ {
1889
+ "epoch": 1.45,
1890
+ "learning_rate": 2.5770034477543296e-05,
1891
+ "loss": 1.986,
1892
+ "step": 157000
1893
+ },
1894
+ {
1895
+ "epoch": 1.46,
1896
+ "learning_rate": 2.5692868982248854e-05,
1897
+ "loss": 1.9676,
1898
+ "step": 157500
1899
+ },
1900
+ {
1901
+ "epoch": 1.46,
1902
+ "learning_rate": 2.56157034869544e-05,
1903
+ "loss": 1.9541,
1904
+ "step": 158000
1905
+ },
1906
+ {
1907
+ "epoch": 1.47,
1908
+ "learning_rate": 2.5538537991659954e-05,
1909
+ "loss": 1.9392,
1910
+ "step": 158500
1911
+ },
1912
+ {
1913
+ "epoch": 1.47,
1914
+ "learning_rate": 2.5461372496365505e-05,
1915
+ "loss": 1.9335,
1916
+ "step": 159000
1917
+ },
1918
+ {
1919
+ "epoch": 1.48,
1920
+ "learning_rate": 2.538420700107106e-05,
1921
+ "loss": 1.9308,
1922
+ "step": 159500
1923
+ },
1924
+ {
1925
+ "epoch": 1.48,
1926
+ "learning_rate": 2.530704150577661e-05,
1927
+ "loss": 1.9314,
1928
+ "step": 160000
1929
+ },
1930
+ {
1931
+ "epoch": 1.49,
1932
+ "learning_rate": 2.5229876010482163e-05,
1933
+ "loss": 1.9259,
1934
+ "step": 160500
1935
+ },
1936
+ {
1937
+ "epoch": 1.49,
1938
+ "learning_rate": 2.5152710515187717e-05,
1939
+ "loss": 1.9202,
1940
+ "step": 161000
1941
+ },
1942
+ {
1943
+ "epoch": 1.5,
1944
+ "learning_rate": 2.5075545019893264e-05,
1945
+ "loss": 1.9131,
1946
+ "step": 161500
1947
+ },
1948
+ {
1949
+ "epoch": 1.5,
1950
+ "learning_rate": 2.4998379524598818e-05,
1951
+ "loss": 1.9097,
1952
+ "step": 162000
1953
+ },
1954
+ {
1955
+ "epoch": 1.5,
1956
+ "learning_rate": 2.4921214029304368e-05,
1957
+ "loss": 1.9126,
1958
+ "step": 162500
1959
+ },
1960
+ {
1961
+ "epoch": 1.51,
1962
+ "learning_rate": 2.4844048534009922e-05,
1963
+ "loss": 1.9031,
1964
+ "step": 163000
1965
+ },
1966
+ {
1967
+ "epoch": 1.51,
1968
+ "learning_rate": 2.4766883038715476e-05,
1969
+ "loss": 1.9042,
1970
+ "step": 163500
1971
+ },
1972
+ {
1973
+ "epoch": 1.52,
1974
+ "learning_rate": 2.4689717543421026e-05,
1975
+ "loss": 1.899,
1976
+ "step": 164000
1977
+ },
1978
+ {
1979
+ "epoch": 1.52,
1980
+ "learning_rate": 2.4612552048126577e-05,
1981
+ "loss": 1.8939,
1982
+ "step": 164500
1983
+ },
1984
+ {
1985
+ "epoch": 1.53,
1986
+ "learning_rate": 2.453538655283213e-05,
1987
+ "loss": 1.8871,
1988
+ "step": 165000
1989
+ },
1990
+ {
1991
+ "epoch": 1.53,
1992
+ "learning_rate": 2.445822105753768e-05,
1993
+ "loss": 1.8973,
1994
+ "step": 165500
1995
+ },
1996
+ {
1997
+ "epoch": 1.54,
1998
+ "learning_rate": 2.438105556224323e-05,
1999
+ "loss": 1.8879,
2000
+ "step": 166000
2001
+ },
2002
+ {
2003
+ "epoch": 1.54,
2004
+ "learning_rate": 2.4303890066948785e-05,
2005
+ "loss": 1.8853,
2006
+ "step": 166500
2007
+ },
2008
+ {
2009
+ "epoch": 1.55,
2010
+ "learning_rate": 2.4226724571654335e-05,
2011
+ "loss": 1.8851,
2012
+ "step": 167000
2013
+ },
2014
+ {
2015
+ "epoch": 1.55,
2016
+ "learning_rate": 2.414955907635989e-05,
2017
+ "loss": 1.8871,
2018
+ "step": 167500
2019
+ },
2020
+ {
2021
+ "epoch": 1.56,
2022
+ "learning_rate": 2.4072393581065443e-05,
2023
+ "loss": 1.8749,
2024
+ "step": 168000
2025
+ },
2026
+ {
2027
+ "epoch": 1.56,
2028
+ "learning_rate": 2.3995228085770993e-05,
2029
+ "loss": 1.8756,
2030
+ "step": 168500
2031
+ },
2032
+ {
2033
+ "epoch": 1.56,
2034
+ "learning_rate": 2.3918062590476544e-05,
2035
+ "loss": 1.8784,
2036
+ "step": 169000
2037
+ },
2038
+ {
2039
+ "epoch": 1.57,
2040
+ "learning_rate": 2.3840897095182098e-05,
2041
+ "loss": 1.8693,
2042
+ "step": 169500
2043
+ },
2044
+ {
2045
+ "epoch": 1.57,
2046
+ "learning_rate": 2.3763731599887648e-05,
2047
+ "loss": 1.8639,
2048
+ "step": 170000
2049
+ },
2050
+ {
2051
+ "epoch": 1.58,
2052
+ "learning_rate": 2.36865661045932e-05,
2053
+ "loss": 1.8625,
2054
+ "step": 170500
2055
+ },
2056
+ {
2057
+ "epoch": 1.58,
2058
+ "learning_rate": 2.3609400609298752e-05,
2059
+ "loss": 1.8659,
2060
+ "step": 171000
2061
+ },
2062
+ {
2063
+ "epoch": 1.59,
2064
+ "learning_rate": 2.3532235114004303e-05,
2065
+ "loss": 1.8591,
2066
+ "step": 171500
2067
+ },
2068
+ {
2069
+ "epoch": 1.59,
2070
+ "learning_rate": 2.3455069618709853e-05,
2071
+ "loss": 1.8524,
2072
+ "step": 172000
2073
+ },
2074
+ {
2075
+ "epoch": 1.6,
2076
+ "learning_rate": 2.337790412341541e-05,
2077
+ "loss": 1.8551,
2078
+ "step": 172500
2079
+ },
2080
+ {
2081
+ "epoch": 1.6,
2082
+ "learning_rate": 2.330073862812096e-05,
2083
+ "loss": 1.8547,
2084
+ "step": 173000
2085
+ },
2086
+ {
2087
+ "epoch": 1.61,
2088
+ "learning_rate": 2.322357313282651e-05,
2089
+ "loss": 1.8481,
2090
+ "step": 173500
2091
+ },
2092
+ {
2093
+ "epoch": 1.61,
2094
+ "learning_rate": 2.3146407637532065e-05,
2095
+ "loss": 1.8484,
2096
+ "step": 174000
2097
+ },
2098
+ {
2099
+ "epoch": 1.62,
2100
+ "learning_rate": 2.3069242142237616e-05,
2101
+ "loss": 1.8412,
2102
+ "step": 174500
2103
+ },
2104
+ {
2105
+ "epoch": 1.62,
2106
+ "learning_rate": 2.2992076646943166e-05,
2107
+ "loss": 1.8404,
2108
+ "step": 175000
2109
+ },
2110
+ {
2111
+ "epoch": 1.63,
2112
+ "learning_rate": 2.291491115164872e-05,
2113
+ "loss": 1.8344,
2114
+ "step": 175500
2115
+ },
2116
+ {
2117
+ "epoch": 1.63,
2118
+ "learning_rate": 2.283774565635427e-05,
2119
+ "loss": 1.8417,
2120
+ "step": 176000
2121
+ },
2122
+ {
2123
+ "epoch": 1.63,
2124
+ "learning_rate": 2.276058016105982e-05,
2125
+ "loss": 1.837,
2126
+ "step": 176500
2127
+ },
2128
+ {
2129
+ "epoch": 1.64,
2130
+ "learning_rate": 2.2683414665765374e-05,
2131
+ "loss": 1.8251,
2132
+ "step": 177000
2133
+ },
2134
+ {
2135
+ "epoch": 1.64,
2136
+ "learning_rate": 2.2606249170470928e-05,
2137
+ "loss": 1.825,
2138
+ "step": 177500
2139
+ },
2140
+ {
2141
+ "epoch": 1.65,
2142
+ "learning_rate": 2.252908367517648e-05,
2143
+ "loss": 1.8313,
2144
+ "step": 178000
2145
+ },
2146
+ {
2147
+ "epoch": 1.65,
2148
+ "learning_rate": 2.2451918179882032e-05,
2149
+ "loss": 1.8324,
2150
+ "step": 178500
2151
+ },
2152
+ {
2153
+ "epoch": 1.66,
2154
+ "learning_rate": 2.2374752684587583e-05,
2155
+ "loss": 1.8244,
2156
+ "step": 179000
2157
+ },
2158
+ {
2159
+ "epoch": 1.66,
2160
+ "learning_rate": 2.2297587189293133e-05,
2161
+ "loss": 1.8218,
2162
+ "step": 179500
2163
+ },
2164
+ {
2165
+ "epoch": 1.67,
2166
+ "learning_rate": 2.2220421693998687e-05,
2167
+ "loss": 1.825,
2168
+ "step": 180000
2169
+ },
2170
+ {
2171
+ "epoch": 1.67,
2172
+ "learning_rate": 2.2143256198704238e-05,
2173
+ "loss": 1.8204,
2174
+ "step": 180500
2175
+ },
2176
+ {
2177
+ "epoch": 1.68,
2178
+ "learning_rate": 2.2066090703409788e-05,
2179
+ "loss": 1.8138,
2180
+ "step": 181000
2181
+ },
2182
+ {
2183
+ "epoch": 1.68,
2184
+ "learning_rate": 2.1988925208115342e-05,
2185
+ "loss": 1.8059,
2186
+ "step": 181500
2187
+ },
2188
+ {
2189
+ "epoch": 1.69,
2190
+ "learning_rate": 2.1911759712820896e-05,
2191
+ "loss": 1.8126,
2192
+ "step": 182000
2193
+ },
2194
+ {
2195
+ "epoch": 1.69,
2196
+ "learning_rate": 2.1834594217526446e-05,
2197
+ "loss": 1.8102,
2198
+ "step": 182500
2199
+ },
2200
+ {
2201
+ "epoch": 1.69,
2202
+ "learning_rate": 2.1757428722232e-05,
2203
+ "loss": 1.8102,
2204
+ "step": 183000
2205
+ },
2206
+ {
2207
+ "epoch": 1.7,
2208
+ "learning_rate": 2.168026322693755e-05,
2209
+ "loss": 1.8105,
2210
+ "step": 183500
2211
+ },
2212
+ {
2213
+ "epoch": 1.7,
2214
+ "learning_rate": 2.16030977316431e-05,
2215
+ "loss": 1.8027,
2216
+ "step": 184000
2217
+ },
2218
+ {
2219
+ "epoch": 1.71,
2220
+ "learning_rate": 2.1525932236348654e-05,
2221
+ "loss": 1.8073,
2222
+ "step": 184500
2223
+ },
2224
+ {
2225
+ "epoch": 1.71,
2226
+ "learning_rate": 2.1448766741054205e-05,
2227
+ "loss": 1.8022,
2228
+ "step": 185000
2229
+ },
2230
+ {
2231
+ "epoch": 1.72,
2232
+ "learning_rate": 2.1371601245759755e-05,
2233
+ "loss": 1.7926,
2234
+ "step": 185500
2235
+ },
2236
+ {
2237
+ "epoch": 1.72,
2238
+ "learning_rate": 2.129443575046531e-05,
2239
+ "loss": 1.7932,
2240
+ "step": 186000
2241
+ },
2242
+ {
2243
+ "epoch": 1.73,
2244
+ "learning_rate": 2.121727025517086e-05,
2245
+ "loss": 1.7964,
2246
+ "step": 186500
2247
+ },
2248
+ {
2249
+ "epoch": 1.73,
2250
+ "learning_rate": 2.1140104759876413e-05,
2251
+ "loss": 1.7932,
2252
+ "step": 187000
2253
+ },
2254
+ {
2255
+ "epoch": 1.74,
2256
+ "learning_rate": 2.1062939264581967e-05,
2257
+ "loss": 1.7926,
2258
+ "step": 187500
2259
+ },
2260
+ {
2261
+ "epoch": 1.74,
2262
+ "learning_rate": 2.0985773769287518e-05,
2263
+ "loss": 1.7899,
2264
+ "step": 188000
2265
+ },
2266
+ {
2267
+ "epoch": 1.75,
2268
+ "learning_rate": 2.0908608273993068e-05,
2269
+ "loss": 1.7837,
2270
+ "step": 188500
2271
+ },
2272
+ {
2273
+ "epoch": 1.75,
2274
+ "learning_rate": 2.0831442778698622e-05,
2275
+ "loss": 1.7836,
2276
+ "step": 189000
2277
+ },
2278
+ {
2279
+ "epoch": 1.75,
2280
+ "learning_rate": 2.0754277283404172e-05,
2281
+ "loss": 1.7863,
2282
+ "step": 189500
2283
+ },
2284
+ {
2285
+ "epoch": 1.76,
2286
+ "learning_rate": 2.0677111788109723e-05,
2287
+ "loss": 1.7759,
2288
+ "step": 190000
2289
+ },
2290
+ {
2291
+ "epoch": 1.76,
2292
+ "learning_rate": 2.0599946292815277e-05,
2293
+ "loss": 1.7749,
2294
+ "step": 190500
2295
+ },
2296
+ {
2297
+ "epoch": 1.77,
2298
+ "learning_rate": 2.0522780797520827e-05,
2299
+ "loss": 1.7745,
2300
+ "step": 191000
2301
+ },
2302
+ {
2303
+ "epoch": 1.77,
2304
+ "learning_rate": 2.0445615302226377e-05,
2305
+ "loss": 1.7678,
2306
+ "step": 191500
2307
+ },
2308
+ {
2309
+ "epoch": 1.78,
2310
+ "learning_rate": 2.0368449806931935e-05,
2311
+ "loss": 1.778,
2312
+ "step": 192000
2313
+ },
2314
+ {
2315
+ "epoch": 1.78,
2316
+ "learning_rate": 2.0291284311637485e-05,
2317
+ "loss": 1.7771,
2318
+ "step": 192500
2319
+ },
2320
+ {
2321
+ "epoch": 1.79,
2322
+ "learning_rate": 2.0214118816343035e-05,
2323
+ "loss": 1.7682,
2324
+ "step": 193000
2325
+ },
2326
+ {
2327
+ "epoch": 1.79,
2328
+ "learning_rate": 2.013695332104859e-05,
2329
+ "loss": 1.7679,
2330
+ "step": 193500
2331
+ },
2332
+ {
2333
+ "epoch": 1.8,
2334
+ "learning_rate": 2.005978782575414e-05,
2335
+ "loss": 1.7712,
2336
+ "step": 194000
2337
+ },
2338
+ {
2339
+ "epoch": 1.8,
2340
+ "learning_rate": 1.998262233045969e-05,
2341
+ "loss": 1.7615,
2342
+ "step": 194500
2343
+ },
2344
+ {
2345
+ "epoch": 1.81,
2346
+ "learning_rate": 1.9905456835165244e-05,
2347
+ "loss": 1.7664,
2348
+ "step": 195000
2349
+ },
2350
+ {
2351
+ "epoch": 1.81,
2352
+ "learning_rate": 1.9828291339870794e-05,
2353
+ "loss": 1.7618,
2354
+ "step": 195500
2355
+ },
2356
+ {
2357
+ "epoch": 1.81,
2358
+ "learning_rate": 1.9751125844576345e-05,
2359
+ "loss": 1.7604,
2360
+ "step": 196000
2361
+ },
2362
+ {
2363
+ "epoch": 1.82,
2364
+ "learning_rate": 1.96739603492819e-05,
2365
+ "loss": 1.754,
2366
+ "step": 196500
2367
+ },
2368
+ {
2369
+ "epoch": 1.82,
2370
+ "learning_rate": 1.9596794853987452e-05,
2371
+ "loss": 1.7563,
2372
+ "step": 197000
2373
+ },
2374
+ {
2375
+ "epoch": 1.83,
2376
+ "learning_rate": 1.9519629358693003e-05,
2377
+ "loss": 1.7573,
2378
+ "step": 197500
2379
+ },
2380
+ {
2381
+ "epoch": 1.83,
2382
+ "learning_rate": 1.9442463863398557e-05,
2383
+ "loss": 1.7473,
2384
+ "step": 198000
2385
+ },
2386
+ {
2387
+ "epoch": 1.84,
2388
+ "learning_rate": 1.9365298368104107e-05,
2389
+ "loss": 1.743,
2390
+ "step": 198500
2391
+ },
2392
+ {
2393
+ "epoch": 1.84,
2394
+ "learning_rate": 1.9288132872809657e-05,
2395
+ "loss": 1.7482,
2396
+ "step": 199000
2397
+ },
2398
+ {
2399
+ "epoch": 1.85,
2400
+ "learning_rate": 1.921096737751521e-05,
2401
+ "loss": 1.7395,
2402
+ "step": 199500
2403
+ },
2404
+ {
2405
+ "epoch": 1.85,
2406
+ "learning_rate": 1.9133801882220762e-05,
2407
+ "loss": 1.7428,
2408
+ "step": 200000
2409
+ },
2410
+ {
2411
+ "epoch": 1.86,
2412
+ "learning_rate": 1.9056636386926312e-05,
2413
+ "loss": 1.7448,
2414
+ "step": 200500
2415
+ },
2416
+ {
2417
+ "epoch": 1.86,
2418
+ "learning_rate": 1.8979470891631866e-05,
2419
+ "loss": 1.7461,
2420
+ "step": 201000
2421
+ },
2422
+ {
2423
+ "epoch": 1.87,
2424
+ "learning_rate": 1.890230539633742e-05,
2425
+ "loss": 1.7332,
2426
+ "step": 201500
2427
+ },
2428
+ {
2429
+ "epoch": 1.87,
2430
+ "learning_rate": 1.882513990104297e-05,
2431
+ "loss": 1.7409,
2432
+ "step": 202000
2433
+ },
2434
+ {
2435
+ "epoch": 1.88,
2436
+ "learning_rate": 1.8747974405748524e-05,
2437
+ "loss": 1.7361,
2438
+ "step": 202500
2439
+ },
2440
+ {
2441
+ "epoch": 1.88,
2442
+ "learning_rate": 1.8670808910454074e-05,
2443
+ "loss": 1.7425,
2444
+ "step": 203000
2445
+ },
2446
+ {
2447
+ "epoch": 1.88,
2448
+ "learning_rate": 1.8593643415159625e-05,
2449
+ "loss": 1.7305,
2450
+ "step": 203500
2451
+ },
2452
+ {
2453
+ "epoch": 1.89,
2454
+ "learning_rate": 1.851647791986518e-05,
2455
+ "loss": 1.7352,
2456
+ "step": 204000
2457
+ },
2458
+ {
2459
+ "epoch": 1.89,
2460
+ "learning_rate": 1.843931242457073e-05,
2461
+ "loss": 1.7351,
2462
+ "step": 204500
2463
+ },
2464
+ {
2465
+ "epoch": 1.9,
2466
+ "learning_rate": 1.836214692927628e-05,
2467
+ "loss": 1.7237,
2468
+ "step": 205000
2469
+ },
2470
+ {
2471
+ "epoch": 1.9,
2472
+ "learning_rate": 1.8284981433981833e-05,
2473
+ "loss": 1.729,
2474
+ "step": 205500
2475
+ },
2476
+ {
2477
+ "epoch": 1.91,
2478
+ "learning_rate": 1.8207815938687384e-05,
2479
+ "loss": 1.7275,
2480
+ "step": 206000
2481
+ },
2482
+ {
2483
+ "epoch": 1.91,
2484
+ "learning_rate": 1.8130650443392938e-05,
2485
+ "loss": 1.725,
2486
+ "step": 206500
2487
+ },
2488
+ {
2489
+ "epoch": 1.92,
2490
+ "learning_rate": 1.805348494809849e-05,
2491
+ "loss": 1.7216,
2492
+ "step": 207000
2493
+ },
2494
+ {
2495
+ "epoch": 1.92,
2496
+ "learning_rate": 1.7976319452804042e-05,
2497
+ "loss": 1.7218,
2498
+ "step": 207500
2499
+ },
2500
+ {
2501
+ "epoch": 1.93,
2502
+ "learning_rate": 1.7899153957509592e-05,
2503
+ "loss": 1.7187,
2504
+ "step": 208000
2505
+ },
2506
+ {
2507
+ "epoch": 1.93,
2508
+ "learning_rate": 1.7821988462215146e-05,
2509
+ "loss": 1.7173,
2510
+ "step": 208500
2511
+ },
2512
+ {
2513
+ "epoch": 1.94,
2514
+ "learning_rate": 1.7744822966920696e-05,
2515
+ "loss": 1.7228,
2516
+ "step": 209000
2517
+ },
2518
+ {
2519
+ "epoch": 1.94,
2520
+ "learning_rate": 1.7667657471626247e-05,
2521
+ "loss": 1.7113,
2522
+ "step": 209500
2523
+ },
2524
+ {
2525
+ "epoch": 1.94,
2526
+ "learning_rate": 1.75904919763318e-05,
2527
+ "loss": 1.7072,
2528
+ "step": 210000
2529
+ },
2530
+ {
2531
+ "epoch": 1.95,
2532
+ "learning_rate": 1.751332648103735e-05,
2533
+ "loss": 1.7174,
2534
+ "step": 210500
2535
+ },
2536
+ {
2537
+ "epoch": 1.95,
2538
+ "learning_rate": 1.74361609857429e-05,
2539
+ "loss": 1.707,
2540
+ "step": 211000
2541
+ },
2542
+ {
2543
+ "epoch": 1.96,
2544
+ "learning_rate": 1.7358995490448455e-05,
2545
+ "loss": 1.7147,
2546
+ "step": 211500
2547
+ },
2548
+ {
2549
+ "epoch": 1.96,
2550
+ "learning_rate": 1.728182999515401e-05,
2551
+ "loss": 1.7156,
2552
+ "step": 212000
2553
+ },
2554
+ {
2555
+ "epoch": 1.97,
2556
+ "learning_rate": 1.720466449985956e-05,
2557
+ "loss": 1.7024,
2558
+ "step": 212500
2559
+ },
2560
+ {
2561
+ "epoch": 1.97,
2562
+ "learning_rate": 1.7127499004565113e-05,
2563
+ "loss": 1.7038,
2564
+ "step": 213000
2565
+ },
2566
+ {
2567
+ "epoch": 1.98,
2568
+ "learning_rate": 1.7050333509270664e-05,
2569
+ "loss": 1.7006,
2570
+ "step": 213500
2571
+ },
2572
+ {
2573
+ "epoch": 1.98,
2574
+ "learning_rate": 1.6973168013976214e-05,
2575
+ "loss": 1.6952,
2576
+ "step": 214000
2577
+ },
2578
+ {
2579
+ "epoch": 1.99,
2580
+ "learning_rate": 1.6896002518681768e-05,
2581
+ "loss": 1.7124,
2582
+ "step": 214500
2583
+ },
2584
+ {
2585
+ "epoch": 1.99,
2586
+ "learning_rate": 1.681883702338732e-05,
2587
+ "loss": 1.7015,
2588
+ "step": 215000
2589
+ },
2590
+ {
2591
+ "epoch": 2.0,
2592
+ "learning_rate": 1.674167152809287e-05,
2593
+ "loss": 1.6964,
2594
+ "step": 215500
2595
+ },
2596
+ {
2597
+ "epoch": 2.0,
2598
+ "learning_rate": 1.6664506032798423e-05,
2599
+ "loss": 1.6972,
2600
+ "step": 216000
2601
+ },
2602
+ {
2603
+ "epoch": 2.0,
2604
+ "learning_rate": 1.6587340537503977e-05,
2605
+ "loss": 1.6966,
2606
+ "step": 216500
2607
+ },
2608
+ {
2609
+ "epoch": 2.01,
2610
+ "learning_rate": 1.6510175042209527e-05,
2611
+ "loss": 1.6976,
2612
+ "step": 217000
2613
+ },
2614
+ {
2615
+ "epoch": 2.01,
2616
+ "learning_rate": 1.643300954691508e-05,
2617
+ "loss": 1.6848,
2618
+ "step": 217500
2619
+ },
2620
+ {
2621
+ "epoch": 2.02,
2622
+ "learning_rate": 1.635584405162063e-05,
2623
+ "loss": 1.6875,
2624
+ "step": 218000
2625
+ },
2626
+ {
2627
+ "epoch": 2.02,
2628
+ "learning_rate": 1.627867855632618e-05,
2629
+ "loss": 1.6929,
2630
+ "step": 218500
2631
+ },
2632
+ {
2633
+ "epoch": 2.03,
2634
+ "learning_rate": 1.6201513061031735e-05,
2635
+ "loss": 1.6858,
2636
+ "step": 219000
2637
+ },
2638
+ {
2639
+ "epoch": 2.03,
2640
+ "learning_rate": 1.6124347565737286e-05,
2641
+ "loss": 1.6863,
2642
+ "step": 219500
2643
+ },
2644
+ {
2645
+ "epoch": 2.04,
2646
+ "learning_rate": 1.6047182070442836e-05,
2647
+ "loss": 1.6858,
2648
+ "step": 220000
2649
+ },
2650
+ {
2651
+ "epoch": 2.04,
2652
+ "learning_rate": 1.597001657514839e-05,
2653
+ "loss": 1.6841,
2654
+ "step": 220500
2655
+ },
2656
+ {
2657
+ "epoch": 2.05,
2658
+ "learning_rate": 1.5892851079853944e-05,
2659
+ "loss": 1.6815,
2660
+ "step": 221000
2661
+ },
2662
+ {
2663
+ "epoch": 2.05,
2664
+ "learning_rate": 1.5815685584559494e-05,
2665
+ "loss": 1.6829,
2666
+ "step": 221500
2667
+ },
2668
+ {
2669
+ "epoch": 2.06,
2670
+ "learning_rate": 1.5738520089265045e-05,
2671
+ "loss": 1.6897,
2672
+ "step": 222000
2673
+ },
2674
+ {
2675
+ "epoch": 2.06,
2676
+ "learning_rate": 1.56613545939706e-05,
2677
+ "loss": 1.6817,
2678
+ "step": 222500
2679
+ },
2680
+ {
2681
+ "epoch": 2.06,
2682
+ "learning_rate": 1.558418909867615e-05,
2683
+ "loss": 1.6822,
2684
+ "step": 223000
2685
+ },
2686
+ {
2687
+ "epoch": 2.07,
2688
+ "learning_rate": 1.5507023603381703e-05,
2689
+ "loss": 1.6784,
2690
+ "step": 223500
2691
+ },
2692
+ {
2693
+ "epoch": 2.07,
2694
+ "learning_rate": 1.5429858108087253e-05,
2695
+ "loss": 1.6743,
2696
+ "step": 224000
2697
+ },
2698
+ {
2699
+ "epoch": 2.08,
2700
+ "learning_rate": 1.5352692612792804e-05,
2701
+ "loss": 1.6865,
2702
+ "step": 224500
2703
+ },
2704
+ {
2705
+ "epoch": 2.08,
2706
+ "learning_rate": 1.5275527117498357e-05,
2707
+ "loss": 1.671,
2708
+ "step": 225000
2709
+ },
2710
+ {
2711
+ "epoch": 2.09,
2712
+ "learning_rate": 1.5198361622203908e-05,
2713
+ "loss": 1.671,
2714
+ "step": 225500
2715
+ },
2716
+ {
2717
+ "epoch": 2.09,
2718
+ "learning_rate": 1.5121196126909462e-05,
2719
+ "loss": 1.6757,
2720
+ "step": 226000
2721
+ },
2722
+ {
2723
+ "epoch": 2.1,
2724
+ "learning_rate": 1.5044030631615014e-05,
2725
+ "loss": 1.6741,
2726
+ "step": 226500
2727
+ },
2728
+ {
2729
+ "epoch": 2.1,
2730
+ "learning_rate": 1.4966865136320566e-05,
2731
+ "loss": 1.6719,
2732
+ "step": 227000
2733
+ },
2734
+ {
2735
+ "epoch": 2.11,
2736
+ "learning_rate": 1.4889699641026116e-05,
2737
+ "loss": 1.671,
2738
+ "step": 227500
2739
+ },
2740
+ {
2741
+ "epoch": 2.11,
2742
+ "learning_rate": 1.4812534145731668e-05,
2743
+ "loss": 1.6658,
2744
+ "step": 228000
2745
+ },
2746
+ {
2747
+ "epoch": 2.12,
2748
+ "learning_rate": 1.473536865043722e-05,
2749
+ "loss": 1.6649,
2750
+ "step": 228500
2751
+ },
2752
+ {
2753
+ "epoch": 2.12,
2754
+ "learning_rate": 1.4658203155142771e-05,
2755
+ "loss": 1.6693,
2756
+ "step": 229000
2757
+ },
2758
+ {
2759
+ "epoch": 2.13,
2760
+ "learning_rate": 1.4581037659848323e-05,
2761
+ "loss": 1.6604,
2762
+ "step": 229500
2763
+ },
2764
+ {
2765
+ "epoch": 2.13,
2766
+ "learning_rate": 1.4503872164553875e-05,
2767
+ "loss": 1.6655,
2768
+ "step": 230000
2769
+ },
2770
+ {
2771
+ "epoch": 2.13,
2772
+ "learning_rate": 1.4426706669259429e-05,
2773
+ "loss": 1.6564,
2774
+ "step": 230500
2775
+ },
2776
+ {
2777
+ "epoch": 2.14,
2778
+ "learning_rate": 1.4349541173964981e-05,
2779
+ "loss": 1.6664,
2780
+ "step": 231000
2781
+ },
2782
+ {
2783
+ "epoch": 2.14,
2784
+ "learning_rate": 1.4272375678670533e-05,
2785
+ "loss": 1.6537,
2786
+ "step": 231500
2787
+ },
2788
+ {
2789
+ "epoch": 2.15,
2790
+ "learning_rate": 1.4195210183376084e-05,
2791
+ "loss": 1.6616,
2792
+ "step": 232000
2793
+ },
2794
+ {
2795
+ "epoch": 2.15,
2796
+ "learning_rate": 1.4118044688081636e-05,
2797
+ "loss": 1.6579,
2798
+ "step": 232500
2799
+ },
2800
+ {
2801
+ "epoch": 2.16,
2802
+ "learning_rate": 1.4040879192787188e-05,
2803
+ "loss": 1.6617,
2804
+ "step": 233000
2805
+ },
2806
+ {
2807
+ "epoch": 2.16,
2808
+ "learning_rate": 1.3963713697492738e-05,
2809
+ "loss": 1.652,
2810
+ "step": 233500
2811
+ },
2812
+ {
2813
+ "epoch": 2.17,
2814
+ "learning_rate": 1.388654820219829e-05,
2815
+ "loss": 1.6544,
2816
+ "step": 234000
2817
+ },
2818
+ {
2819
+ "epoch": 2.17,
2820
+ "learning_rate": 1.3809382706903843e-05,
2821
+ "loss": 1.6504,
2822
+ "step": 234500
2823
+ },
2824
+ {
2825
+ "epoch": 2.18,
2826
+ "learning_rate": 1.3732217211609393e-05,
2827
+ "loss": 1.6574,
2828
+ "step": 235000
2829
+ },
2830
+ {
2831
+ "epoch": 2.18,
2832
+ "learning_rate": 1.3655051716314949e-05,
2833
+ "loss": 1.6495,
2834
+ "step": 235500
2835
+ },
2836
+ {
2837
+ "epoch": 2.19,
2838
+ "learning_rate": 1.35778862210205e-05,
2839
+ "loss": 1.6519,
2840
+ "step": 236000
2841
+ },
2842
+ {
2843
+ "epoch": 2.19,
2844
+ "learning_rate": 1.3500720725726051e-05,
2845
+ "loss": 1.6521,
2846
+ "step": 236500
2847
+ },
2848
+ {
2849
+ "epoch": 2.19,
2850
+ "learning_rate": 1.3423555230431603e-05,
2851
+ "loss": 1.6535,
2852
+ "step": 237000
2853
+ },
2854
+ {
2855
+ "epoch": 2.2,
2856
+ "learning_rate": 1.3346389735137155e-05,
2857
+ "loss": 1.6527,
2858
+ "step": 237500
2859
+ },
2860
+ {
2861
+ "epoch": 2.2,
2862
+ "learning_rate": 1.3269224239842706e-05,
2863
+ "loss": 1.6459,
2864
+ "step": 238000
2865
+ },
2866
+ {
2867
+ "epoch": 2.21,
2868
+ "learning_rate": 1.3192058744548258e-05,
2869
+ "loss": 1.6444,
2870
+ "step": 238500
2871
+ },
2872
+ {
2873
+ "epoch": 2.21,
2874
+ "learning_rate": 1.311489324925381e-05,
2875
+ "loss": 1.6443,
2876
+ "step": 239000
2877
+ },
2878
+ {
2879
+ "epoch": 2.22,
2880
+ "learning_rate": 1.303772775395936e-05,
2881
+ "loss": 1.6448,
2882
+ "step": 239500
2883
+ },
2884
+ {
2885
+ "epoch": 2.22,
2886
+ "learning_rate": 1.2960562258664913e-05,
2887
+ "loss": 1.6469,
2888
+ "step": 240000
2889
+ },
2890
+ {
2891
+ "epoch": 2.23,
2892
+ "learning_rate": 1.2883396763370466e-05,
2893
+ "loss": 1.6368,
2894
+ "step": 240500
2895
+ },
2896
+ {
2897
+ "epoch": 2.23,
2898
+ "learning_rate": 1.2806231268076018e-05,
2899
+ "loss": 1.6462,
2900
+ "step": 241000
2901
+ },
2902
+ {
2903
+ "epoch": 2.24,
2904
+ "learning_rate": 1.272906577278157e-05,
2905
+ "loss": 1.6395,
2906
+ "step": 241500
2907
+ },
2908
+ {
2909
+ "epoch": 2.24,
2910
+ "learning_rate": 1.2651900277487123e-05,
2911
+ "loss": 1.6396,
2912
+ "step": 242000
2913
+ },
2914
+ {
2915
+ "epoch": 2.25,
2916
+ "learning_rate": 1.2574734782192673e-05,
2917
+ "loss": 1.6442,
2918
+ "step": 242500
2919
+ },
2920
+ {
2921
+ "epoch": 2.25,
2922
+ "learning_rate": 1.2497569286898225e-05,
2923
+ "loss": 1.6417,
2924
+ "step": 243000
2925
+ },
2926
+ {
2927
+ "epoch": 2.25,
2928
+ "learning_rate": 1.2420403791603777e-05,
2929
+ "loss": 1.6377,
2930
+ "step": 243500
2931
+ },
2932
+ {
2933
+ "epoch": 2.26,
2934
+ "learning_rate": 1.234323829630933e-05,
2935
+ "loss": 1.6415,
2936
+ "step": 244000
2937
+ },
2938
+ {
2939
+ "epoch": 2.26,
2940
+ "learning_rate": 1.2266072801014882e-05,
2941
+ "loss": 1.6313,
2942
+ "step": 244500
2943
+ },
2944
+ {
2945
+ "epoch": 2.27,
2946
+ "learning_rate": 1.2188907305720434e-05,
2947
+ "loss": 1.6325,
2948
+ "step": 245000
2949
+ },
2950
+ {
2951
+ "epoch": 2.27,
2952
+ "learning_rate": 1.2111741810425984e-05,
2953
+ "loss": 1.6376,
2954
+ "step": 245500
2955
+ },
2956
+ {
2957
+ "epoch": 2.28,
2958
+ "learning_rate": 1.2034576315131536e-05,
2959
+ "loss": 1.6412,
2960
+ "step": 246000
2961
+ },
2962
+ {
2963
+ "epoch": 2.28,
2964
+ "learning_rate": 1.195741081983709e-05,
2965
+ "loss": 1.6296,
2966
+ "step": 246500
2967
+ },
2968
+ {
2969
+ "epoch": 2.29,
2970
+ "learning_rate": 1.188024532454264e-05,
2971
+ "loss": 1.6288,
2972
+ "step": 247000
2973
+ },
2974
+ {
2975
+ "epoch": 2.29,
2976
+ "learning_rate": 1.1803079829248193e-05,
2977
+ "loss": 1.6349,
2978
+ "step": 247500
2979
+ },
2980
+ {
2981
+ "epoch": 2.3,
2982
+ "learning_rate": 1.1725914333953745e-05,
2983
+ "loss": 1.6317,
2984
+ "step": 248000
2985
+ },
2986
+ {
2987
+ "epoch": 2.3,
2988
+ "learning_rate": 1.1648748838659295e-05,
2989
+ "loss": 1.6306,
2990
+ "step": 248500
2991
+ },
2992
+ {
2993
+ "epoch": 2.31,
2994
+ "learning_rate": 1.1571583343364849e-05,
2995
+ "loss": 1.6322,
2996
+ "step": 249000
2997
+ },
2998
+ {
2999
+ "epoch": 2.31,
3000
+ "learning_rate": 1.1494417848070401e-05,
3001
+ "loss": 1.6334,
3002
+ "step": 249500
3003
+ },
3004
+ {
3005
+ "epoch": 2.31,
3006
+ "learning_rate": 1.1417252352775952e-05,
3007
+ "loss": 1.6261,
3008
+ "step": 250000
3009
+ },
3010
+ {
3011
+ "epoch": 2.32,
3012
+ "learning_rate": 1.1340086857481504e-05,
3013
+ "loss": 1.6245,
3014
+ "step": 250500
3015
+ },
3016
+ {
3017
+ "epoch": 2.32,
3018
+ "learning_rate": 1.1262921362187056e-05,
3019
+ "loss": 1.6254,
3020
+ "step": 251000
3021
+ },
3022
+ {
3023
+ "epoch": 2.33,
3024
+ "learning_rate": 1.1185755866892608e-05,
3025
+ "loss": 1.6244,
3026
+ "step": 251500
3027
+ },
3028
+ {
3029
+ "epoch": 2.33,
3030
+ "learning_rate": 1.110859037159816e-05,
3031
+ "loss": 1.6203,
3032
+ "step": 252000
3033
+ },
3034
+ {
3035
+ "epoch": 2.34,
3036
+ "learning_rate": 1.1031424876303712e-05,
3037
+ "loss": 1.6255,
3038
+ "step": 252500
3039
+ },
3040
+ {
3041
+ "epoch": 2.34,
3042
+ "learning_rate": 1.0954259381009263e-05,
3043
+ "loss": 1.6187,
3044
+ "step": 253000
3045
+ },
3046
+ {
3047
+ "epoch": 2.35,
3048
+ "learning_rate": 1.0877093885714815e-05,
3049
+ "loss": 1.6242,
3050
+ "step": 253500
3051
+ },
3052
+ {
3053
+ "epoch": 2.35,
3054
+ "learning_rate": 1.0799928390420368e-05,
3055
+ "loss": 1.6218,
3056
+ "step": 254000
3057
+ },
3058
+ {
3059
+ "epoch": 2.36,
3060
+ "learning_rate": 1.0722762895125919e-05,
3061
+ "loss": 1.6202,
3062
+ "step": 254500
3063
+ },
3064
+ {
3065
+ "epoch": 2.36,
3066
+ "learning_rate": 1.0645597399831471e-05,
3067
+ "loss": 1.6138,
3068
+ "step": 255000
3069
+ },
3070
+ {
3071
+ "epoch": 2.37,
3072
+ "learning_rate": 1.0568431904537023e-05,
3073
+ "loss": 1.6146,
3074
+ "step": 255500
3075
+ },
3076
+ {
3077
+ "epoch": 2.37,
3078
+ "learning_rate": 1.0491266409242575e-05,
3079
+ "loss": 1.608,
3080
+ "step": 256000
3081
+ },
3082
+ {
3083
+ "epoch": 2.38,
3084
+ "learning_rate": 1.0414100913948127e-05,
3085
+ "loss": 1.62,
3086
+ "step": 256500
3087
+ },
3088
+ {
3089
+ "epoch": 2.38,
3090
+ "learning_rate": 1.033693541865368e-05,
3091
+ "loss": 1.6238,
3092
+ "step": 257000
3093
+ },
3094
+ {
3095
+ "epoch": 2.38,
3096
+ "learning_rate": 1.025976992335923e-05,
3097
+ "loss": 1.6198,
3098
+ "step": 257500
3099
+ },
3100
+ {
3101
+ "epoch": 2.39,
3102
+ "learning_rate": 1.0182604428064782e-05,
3103
+ "loss": 1.6165,
3104
+ "step": 258000
3105
+ },
3106
+ {
3107
+ "epoch": 2.39,
3108
+ "learning_rate": 1.0105438932770334e-05,
3109
+ "loss": 1.615,
3110
+ "step": 258500
3111
+ },
3112
+ {
3113
+ "epoch": 2.4,
3114
+ "learning_rate": 1.0028273437475886e-05,
3115
+ "loss": 1.6119,
3116
+ "step": 259000
3117
+ },
3118
+ {
3119
+ "epoch": 2.4,
3120
+ "learning_rate": 9.951107942181438e-06,
3121
+ "loss": 1.6029,
3122
+ "step": 259500
3123
+ },
3124
+ {
3125
+ "epoch": 2.41,
3126
+ "learning_rate": 9.87394244688699e-06,
3127
+ "loss": 1.6108,
3128
+ "step": 260000
3129
+ },
3130
+ {
3131
+ "epoch": 2.41,
3132
+ "learning_rate": 9.796776951592541e-06,
3133
+ "loss": 1.6152,
3134
+ "step": 260500
3135
+ },
3136
+ {
3137
+ "epoch": 2.42,
3138
+ "learning_rate": 9.719611456298095e-06,
3139
+ "loss": 1.6118,
3140
+ "step": 261000
3141
+ },
3142
+ {
3143
+ "epoch": 2.42,
3144
+ "learning_rate": 9.642445961003647e-06,
3145
+ "loss": 1.6092,
3146
+ "step": 261500
3147
+ },
3148
+ {
3149
+ "epoch": 2.43,
3150
+ "learning_rate": 9.565280465709197e-06,
3151
+ "loss": 1.6072,
3152
+ "step": 262000
3153
+ },
3154
+ {
3155
+ "epoch": 2.43,
3156
+ "learning_rate": 9.48811497041475e-06,
3157
+ "loss": 1.6117,
3158
+ "step": 262500
3159
+ },
3160
+ {
3161
+ "epoch": 2.44,
3162
+ "learning_rate": 9.410949475120302e-06,
3163
+ "loss": 1.6077,
3164
+ "step": 263000
3165
+ },
3166
+ {
3167
+ "epoch": 2.44,
3168
+ "learning_rate": 9.333783979825854e-06,
3169
+ "loss": 1.6082,
3170
+ "step": 263500
3171
+ },
3172
+ {
3173
+ "epoch": 2.44,
3174
+ "learning_rate": 9.256618484531406e-06,
3175
+ "loss": 1.6054,
3176
+ "step": 264000
3177
+ },
3178
+ {
3179
+ "epoch": 2.45,
3180
+ "learning_rate": 9.179452989236958e-06,
3181
+ "loss": 1.6012,
3182
+ "step": 264500
3183
+ },
3184
+ {
3185
+ "epoch": 2.45,
3186
+ "learning_rate": 9.102287493942508e-06,
3187
+ "loss": 1.6091,
3188
+ "step": 265000
3189
+ },
3190
+ {
3191
+ "epoch": 2.46,
3192
+ "learning_rate": 9.02512199864806e-06,
3193
+ "loss": 1.6026,
3194
+ "step": 265500
3195
+ },
3196
+ {
3197
+ "epoch": 2.46,
3198
+ "learning_rate": 8.947956503353613e-06,
3199
+ "loss": 1.6038,
3200
+ "step": 266000
3201
+ },
3202
+ {
3203
+ "epoch": 2.47,
3204
+ "learning_rate": 8.870791008059165e-06,
3205
+ "loss": 1.5986,
3206
+ "step": 266500
3207
+ },
3208
+ {
3209
+ "epoch": 2.47,
3210
+ "learning_rate": 8.793625512764717e-06,
3211
+ "loss": 1.5989,
3212
+ "step": 267000
3213
+ },
3214
+ {
3215
+ "epoch": 2.48,
3216
+ "learning_rate": 8.716460017470269e-06,
3217
+ "loss": 1.6037,
3218
+ "step": 267500
3219
+ },
3220
+ {
3221
+ "epoch": 2.48,
3222
+ "learning_rate": 8.63929452217582e-06,
3223
+ "loss": 1.5999,
3224
+ "step": 268000
3225
+ },
3226
+ {
3227
+ "epoch": 2.49,
3228
+ "learning_rate": 8.562129026881373e-06,
3229
+ "loss": 1.5942,
3230
+ "step": 268500
3231
+ },
3232
+ {
3233
+ "epoch": 2.49,
3234
+ "learning_rate": 8.484963531586925e-06,
3235
+ "loss": 1.6006,
3236
+ "step": 269000
3237
+ },
3238
+ {
3239
+ "epoch": 2.5,
3240
+ "learning_rate": 8.407798036292476e-06,
3241
+ "loss": 1.5911,
3242
+ "step": 269500
3243
+ },
3244
+ {
3245
+ "epoch": 2.5,
3246
+ "learning_rate": 8.330632540998028e-06,
3247
+ "loss": 1.5997,
3248
+ "step": 270000
3249
+ },
3250
+ {
3251
+ "epoch": 2.5,
3252
+ "learning_rate": 8.25346704570358e-06,
3253
+ "loss": 1.5983,
3254
+ "step": 270500
3255
+ },
3256
+ {
3257
+ "epoch": 2.51,
3258
+ "learning_rate": 8.176301550409132e-06,
3259
+ "loss": 1.5966,
3260
+ "step": 271000
3261
+ },
3262
+ {
3263
+ "epoch": 2.51,
3264
+ "learning_rate": 8.099136055114684e-06,
3265
+ "loss": 1.5899,
3266
+ "step": 271500
3267
+ },
3268
+ {
3269
+ "epoch": 2.52,
3270
+ "learning_rate": 8.021970559820236e-06,
3271
+ "loss": 1.599,
3272
+ "step": 272000
3273
+ },
3274
+ {
3275
+ "epoch": 2.52,
3276
+ "learning_rate": 7.944805064525787e-06,
3277
+ "loss": 1.5894,
3278
+ "step": 272500
3279
+ },
3280
+ {
3281
+ "epoch": 2.53,
3282
+ "learning_rate": 7.86763956923134e-06,
3283
+ "loss": 1.5969,
3284
+ "step": 273000
3285
+ },
3286
+ {
3287
+ "epoch": 2.53,
3288
+ "learning_rate": 7.790474073936891e-06,
3289
+ "loss": 1.5976,
3290
+ "step": 273500
3291
+ },
3292
+ {
3293
+ "epoch": 2.54,
3294
+ "learning_rate": 7.713308578642443e-06,
3295
+ "loss": 1.5907,
3296
+ "step": 274000
3297
+ },
3298
+ {
3299
+ "epoch": 2.54,
3300
+ "learning_rate": 7.636143083347995e-06,
3301
+ "loss": 1.5812,
3302
+ "step": 274500
3303
+ },
3304
+ {
3305
+ "epoch": 2.55,
3306
+ "learning_rate": 7.5589775880535464e-06,
3307
+ "loss": 1.5889,
3308
+ "step": 275000
3309
+ },
3310
+ {
3311
+ "epoch": 2.55,
3312
+ "learning_rate": 7.481812092759099e-06,
3313
+ "loss": 1.5857,
3314
+ "step": 275500
3315
+ },
3316
+ {
3317
+ "epoch": 2.56,
3318
+ "learning_rate": 7.4046465974646515e-06,
3319
+ "loss": 1.5849,
3320
+ "step": 276000
3321
+ },
3322
+ {
3323
+ "epoch": 2.56,
3324
+ "learning_rate": 7.327481102170203e-06,
3325
+ "loss": 1.5893,
3326
+ "step": 276500
3327
+ },
3328
+ {
3329
+ "epoch": 2.56,
3330
+ "learning_rate": 7.250315606875754e-06,
3331
+ "loss": 1.5901,
3332
+ "step": 277000
3333
+ },
3334
+ {
3335
+ "epoch": 2.57,
3336
+ "learning_rate": 7.173150111581306e-06,
3337
+ "loss": 1.5812,
3338
+ "step": 277500
3339
+ },
3340
+ {
3341
+ "epoch": 2.57,
3342
+ "learning_rate": 7.095984616286859e-06,
3343
+ "loss": 1.5969,
3344
+ "step": 278000
3345
+ },
3346
+ {
3347
+ "epoch": 2.58,
3348
+ "learning_rate": 7.01881912099241e-06,
3349
+ "loss": 1.5894,
3350
+ "step": 278500
3351
+ },
3352
+ {
3353
+ "epoch": 2.58,
3354
+ "learning_rate": 6.9416536256979625e-06,
3355
+ "loss": 1.5812,
3356
+ "step": 279000
3357
+ },
3358
+ {
3359
+ "epoch": 2.59,
3360
+ "learning_rate": 6.864488130403514e-06,
3361
+ "loss": 1.5908,
3362
+ "step": 279500
3363
+ },
3364
+ {
3365
+ "epoch": 2.59,
3366
+ "learning_rate": 6.787322635109065e-06,
3367
+ "loss": 1.5869,
3368
+ "step": 280000
3369
+ },
3370
+ {
3371
+ "epoch": 2.6,
3372
+ "learning_rate": 6.710157139814618e-06,
3373
+ "loss": 1.5863,
3374
+ "step": 280500
3375
+ },
3376
+ {
3377
+ "epoch": 2.6,
3378
+ "learning_rate": 6.63299164452017e-06,
3379
+ "loss": 1.5842,
3380
+ "step": 281000
3381
+ },
3382
+ {
3383
+ "epoch": 2.61,
3384
+ "learning_rate": 6.5558261492257214e-06,
3385
+ "loss": 1.5794,
3386
+ "step": 281500
3387
+ },
3388
+ {
3389
+ "epoch": 2.61,
3390
+ "learning_rate": 6.4786606539312736e-06,
3391
+ "loss": 1.5896,
3392
+ "step": 282000
3393
+ },
3394
+ {
3395
+ "epoch": 2.62,
3396
+ "learning_rate": 6.401495158636825e-06,
3397
+ "loss": 1.5837,
3398
+ "step": 282500
3399
+ },
3400
+ {
3401
+ "epoch": 2.62,
3402
+ "learning_rate": 6.324329663342378e-06,
3403
+ "loss": 1.579,
3404
+ "step": 283000
3405
+ },
3406
+ {
3407
+ "epoch": 2.63,
3408
+ "learning_rate": 6.24716416804793e-06,
3409
+ "loss": 1.5811,
3410
+ "step": 283500
3411
+ },
3412
+ {
3413
+ "epoch": 2.63,
3414
+ "learning_rate": 6.169998672753481e-06,
3415
+ "loss": 1.5833,
3416
+ "step": 284000
3417
+ },
3418
+ {
3419
+ "epoch": 2.63,
3420
+ "learning_rate": 6.092833177459033e-06,
3421
+ "loss": 1.5814,
3422
+ "step": 284500
3423
+ },
3424
+ {
3425
+ "epoch": 2.64,
3426
+ "learning_rate": 6.015667682164585e-06,
3427
+ "loss": 1.5826,
3428
+ "step": 285000
3429
+ },
3430
+ {
3431
+ "epoch": 2.64,
3432
+ "learning_rate": 5.938502186870137e-06,
3433
+ "loss": 1.5786,
3434
+ "step": 285500
3435
+ },
3436
+ {
3437
+ "epoch": 2.65,
3438
+ "learning_rate": 5.861336691575689e-06,
3439
+ "loss": 1.575,
3440
+ "step": 286000
3441
+ },
3442
+ {
3443
+ "epoch": 2.65,
3444
+ "learning_rate": 5.784171196281241e-06,
3445
+ "loss": 1.5753,
3446
+ "step": 286500
3447
+ },
3448
+ {
3449
+ "epoch": 2.66,
3450
+ "learning_rate": 5.707005700986793e-06,
3451
+ "loss": 1.5788,
3452
+ "step": 287000
3453
+ },
3454
+ {
3455
+ "epoch": 2.66,
3456
+ "learning_rate": 5.629840205692344e-06,
3457
+ "loss": 1.5777,
3458
+ "step": 287500
3459
+ },
3460
+ {
3461
+ "epoch": 2.67,
3462
+ "learning_rate": 5.5526747103978964e-06,
3463
+ "loss": 1.5719,
3464
+ "step": 288000
3465
+ },
3466
+ {
3467
+ "epoch": 2.67,
3468
+ "learning_rate": 5.4755092151034486e-06,
3469
+ "loss": 1.5772,
3470
+ "step": 288500
3471
+ },
3472
+ {
3473
+ "epoch": 2.68,
3474
+ "learning_rate": 5.398343719809e-06,
3475
+ "loss": 1.5733,
3476
+ "step": 289000
3477
+ },
3478
+ {
3479
+ "epoch": 2.68,
3480
+ "learning_rate": 5.321178224514552e-06,
3481
+ "loss": 1.5791,
3482
+ "step": 289500
3483
+ },
3484
+ {
3485
+ "epoch": 2.69,
3486
+ "learning_rate": 5.244012729220104e-06,
3487
+ "loss": 1.5711,
3488
+ "step": 290000
3489
+ },
3490
+ {
3491
+ "epoch": 2.69,
3492
+ "learning_rate": 5.166847233925655e-06,
3493
+ "loss": 1.5727,
3494
+ "step": 290500
3495
+ },
3496
+ {
3497
+ "epoch": 2.69,
3498
+ "learning_rate": 5.089681738631208e-06,
3499
+ "loss": 1.57,
3500
+ "step": 291000
3501
+ },
3502
+ {
3503
+ "epoch": 2.7,
3504
+ "learning_rate": 5.0125162433367596e-06,
3505
+ "loss": 1.5743,
3506
+ "step": 291500
3507
+ },
3508
+ {
3509
+ "epoch": 2.7,
3510
+ "learning_rate": 4.935350748042312e-06,
3511
+ "loss": 1.5749,
3512
+ "step": 292000
3513
+ },
3514
+ {
3515
+ "epoch": 2.71,
3516
+ "learning_rate": 4.858185252747864e-06,
3517
+ "loss": 1.5697,
3518
+ "step": 292500
3519
+ },
3520
+ {
3521
+ "epoch": 2.71,
3522
+ "learning_rate": 4.781019757453416e-06,
3523
+ "loss": 1.573,
3524
+ "step": 293000
3525
+ },
3526
+ {
3527
+ "epoch": 2.72,
3528
+ "learning_rate": 4.703854262158967e-06,
3529
+ "loss": 1.5754,
3530
+ "step": 293500
3531
+ },
3532
+ {
3533
+ "epoch": 2.72,
3534
+ "learning_rate": 4.626688766864519e-06,
3535
+ "loss": 1.5677,
3536
+ "step": 294000
3537
+ },
3538
+ {
3539
+ "epoch": 2.73,
3540
+ "learning_rate": 4.5495232715700714e-06,
3541
+ "loss": 1.5689,
3542
+ "step": 294500
3543
+ },
3544
+ {
3545
+ "epoch": 2.73,
3546
+ "learning_rate": 4.472357776275623e-06,
3547
+ "loss": 1.5688,
3548
+ "step": 295000
3549
+ },
3550
+ {
3551
+ "epoch": 2.74,
3552
+ "learning_rate": 4.395192280981175e-06,
3553
+ "loss": 1.5759,
3554
+ "step": 295500
3555
+ },
3556
+ {
3557
+ "epoch": 2.74,
3558
+ "learning_rate": 4.318026785686727e-06,
3559
+ "loss": 1.5662,
3560
+ "step": 296000
3561
+ },
3562
+ {
3563
+ "epoch": 2.75,
3564
+ "learning_rate": 4.240861290392278e-06,
3565
+ "loss": 1.5725,
3566
+ "step": 296500
3567
+ },
3568
+ {
3569
+ "epoch": 2.75,
3570
+ "learning_rate": 4.16369579509783e-06,
3571
+ "loss": 1.5712,
3572
+ "step": 297000
3573
+ },
3574
+ {
3575
+ "epoch": 2.75,
3576
+ "learning_rate": 4.0865302998033825e-06,
3577
+ "loss": 1.5632,
3578
+ "step": 297500
3579
+ },
3580
+ {
3581
+ "epoch": 2.76,
3582
+ "learning_rate": 4.0093648045089346e-06,
3583
+ "loss": 1.5631,
3584
+ "step": 298000
3585
+ },
3586
+ {
3587
+ "epoch": 2.76,
3588
+ "learning_rate": 3.932199309214486e-06,
3589
+ "loss": 1.5639,
3590
+ "step": 298500
3591
+ },
3592
+ {
3593
+ "epoch": 2.77,
3594
+ "learning_rate": 3.855033813920038e-06,
3595
+ "loss": 1.5672,
3596
+ "step": 299000
3597
+ },
3598
+ {
3599
+ "epoch": 2.77,
3600
+ "learning_rate": 3.77786831862559e-06,
3601
+ "loss": 1.5691,
3602
+ "step": 299500
3603
+ },
3604
+ {
3605
+ "epoch": 2.78,
3606
+ "learning_rate": 3.7007028233311418e-06,
3607
+ "loss": 1.5654,
3608
+ "step": 300000
3609
+ },
3610
+ {
3611
+ "epoch": 2.78,
3612
+ "learning_rate": 3.623537328036694e-06,
3613
+ "loss": 1.559,
3614
+ "step": 300500
3615
+ },
3616
+ {
3617
+ "epoch": 2.79,
3618
+ "learning_rate": 3.5463718327422456e-06,
3619
+ "loss": 1.5709,
3620
+ "step": 301000
3621
+ },
3622
+ {
3623
+ "epoch": 2.79,
3624
+ "learning_rate": 3.469206337447798e-06,
3625
+ "loss": 1.5669,
3626
+ "step": 301500
3627
+ },
3628
+ {
3629
+ "epoch": 2.8,
3630
+ "learning_rate": 3.39204084215335e-06,
3631
+ "loss": 1.5635,
3632
+ "step": 302000
3633
+ },
3634
+ {
3635
+ "epoch": 2.8,
3636
+ "learning_rate": 3.314875346858901e-06,
3637
+ "loss": 1.5659,
3638
+ "step": 302500
3639
+ },
3640
+ {
3641
+ "epoch": 2.81,
3642
+ "learning_rate": 3.2377098515644536e-06,
3643
+ "loss": 1.5623,
3644
+ "step": 303000
3645
+ },
3646
+ {
3647
+ "epoch": 2.81,
3648
+ "learning_rate": 3.1605443562700053e-06,
3649
+ "loss": 1.5638,
3650
+ "step": 303500
3651
+ },
3652
+ {
3653
+ "epoch": 2.81,
3654
+ "learning_rate": 3.083378860975557e-06,
3655
+ "loss": 1.5644,
3656
+ "step": 304000
3657
+ },
3658
+ {
3659
+ "epoch": 2.82,
3660
+ "learning_rate": 3.006213365681109e-06,
3661
+ "loss": 1.5593,
3662
+ "step": 304500
3663
+ },
3664
+ {
3665
+ "epoch": 2.82,
3666
+ "learning_rate": 2.929047870386661e-06,
3667
+ "loss": 1.5592,
3668
+ "step": 305000
3669
+ },
3670
+ {
3671
+ "epoch": 2.83,
3672
+ "learning_rate": 2.851882375092213e-06,
3673
+ "loss": 1.5571,
3674
+ "step": 305500
3675
+ },
3676
+ {
3677
+ "epoch": 2.83,
3678
+ "learning_rate": 2.774716879797765e-06,
3679
+ "loss": 1.5574,
3680
+ "step": 306000
3681
+ },
3682
+ {
3683
+ "epoch": 2.84,
3684
+ "learning_rate": 2.6975513845033168e-06,
3685
+ "loss": 1.557,
3686
+ "step": 306500
3687
+ },
3688
+ {
3689
+ "epoch": 2.84,
3690
+ "learning_rate": 2.6203858892088685e-06,
3691
+ "loss": 1.5555,
3692
+ "step": 307000
3693
+ },
3694
+ {
3695
+ "epoch": 2.85,
3696
+ "learning_rate": 2.5432203939144206e-06,
3697
+ "loss": 1.5559,
3698
+ "step": 307500
3699
+ },
3700
+ {
3701
+ "epoch": 2.85,
3702
+ "learning_rate": 2.4660548986199723e-06,
3703
+ "loss": 1.562,
3704
+ "step": 308000
3705
+ },
3706
+ {
3707
+ "epoch": 2.86,
3708
+ "learning_rate": 2.3888894033255244e-06,
3709
+ "loss": 1.5609,
3710
+ "step": 308500
3711
+ },
3712
+ {
3713
+ "epoch": 2.86,
3714
+ "learning_rate": 2.311723908031076e-06,
3715
+ "loss": 1.5584,
3716
+ "step": 309000
3717
+ },
3718
+ {
3719
+ "epoch": 2.87,
3720
+ "learning_rate": 2.234558412736628e-06,
3721
+ "loss": 1.5618,
3722
+ "step": 309500
3723
+ },
3724
+ {
3725
+ "epoch": 2.87,
3726
+ "learning_rate": 2.15739291744218e-06,
3727
+ "loss": 1.5539,
3728
+ "step": 310000
3729
+ },
3730
+ {
3731
+ "epoch": 2.88,
3732
+ "learning_rate": 2.080227422147732e-06,
3733
+ "loss": 1.5567,
3734
+ "step": 310500
3735
+ },
3736
+ {
3737
+ "epoch": 2.88,
3738
+ "learning_rate": 2.0030619268532837e-06,
3739
+ "loss": 1.5585,
3740
+ "step": 311000
3741
+ },
3742
+ {
3743
+ "epoch": 2.88,
3744
+ "learning_rate": 1.925896431558836e-06,
3745
+ "loss": 1.5562,
3746
+ "step": 311500
3747
+ },
3748
+ {
3749
+ "epoch": 2.89,
3750
+ "learning_rate": 1.8487309362643875e-06,
3751
+ "loss": 1.5575,
3752
+ "step": 312000
3753
+ },
3754
+ {
3755
+ "epoch": 2.89,
3756
+ "learning_rate": 1.7715654409699394e-06,
3757
+ "loss": 1.5587,
3758
+ "step": 312500
3759
+ },
3760
+ {
3761
+ "epoch": 2.9,
3762
+ "learning_rate": 1.6943999456754913e-06,
3763
+ "loss": 1.5602,
3764
+ "step": 313000
3765
+ },
3766
+ {
3767
+ "epoch": 2.9,
3768
+ "learning_rate": 1.6172344503810433e-06,
3769
+ "loss": 1.562,
3770
+ "step": 313500
3771
+ },
3772
+ {
3773
+ "epoch": 2.91,
3774
+ "learning_rate": 1.5400689550865952e-06,
3775
+ "loss": 1.5581,
3776
+ "step": 314000
3777
+ },
3778
+ {
3779
+ "epoch": 2.91,
3780
+ "learning_rate": 1.462903459792147e-06,
3781
+ "loss": 1.5536,
3782
+ "step": 314500
3783
+ },
3784
+ {
3785
+ "epoch": 2.92,
3786
+ "learning_rate": 1.385737964497699e-06,
3787
+ "loss": 1.5588,
3788
+ "step": 315000
3789
+ },
3790
+ {
3791
+ "epoch": 2.92,
3792
+ "learning_rate": 1.3085724692032509e-06,
3793
+ "loss": 1.5569,
3794
+ "step": 315500
3795
+ },
3796
+ {
3797
+ "epoch": 2.93,
3798
+ "learning_rate": 1.2314069739088028e-06,
3799
+ "loss": 1.5536,
3800
+ "step": 316000
3801
+ },
3802
+ {
3803
+ "epoch": 2.93,
3804
+ "learning_rate": 1.1542414786143547e-06,
3805
+ "loss": 1.5558,
3806
+ "step": 316500
3807
+ },
3808
+ {
3809
+ "epoch": 2.94,
3810
+ "learning_rate": 1.0770759833199066e-06,
3811
+ "loss": 1.5585,
3812
+ "step": 317000
3813
+ },
3814
+ {
3815
+ "epoch": 2.94,
3816
+ "learning_rate": 9.999104880254585e-07,
3817
+ "loss": 1.5542,
3818
+ "step": 317500
3819
+ },
3820
+ {
3821
+ "epoch": 2.94,
3822
+ "learning_rate": 9.227449927310104e-07,
3823
+ "loss": 1.5529,
3824
+ "step": 318000
3825
+ },
3826
+ {
3827
+ "epoch": 2.95,
3828
+ "learning_rate": 8.455794974365622e-07,
3829
+ "loss": 1.5605,
3830
+ "step": 318500
3831
+ },
3832
+ {
3833
+ "epoch": 2.95,
3834
+ "learning_rate": 7.684140021421142e-07,
3835
+ "loss": 1.5496,
3836
+ "step": 319000
3837
+ },
3838
+ {
3839
+ "epoch": 2.96,
3840
+ "learning_rate": 6.91248506847666e-07,
3841
+ "loss": 1.5489,
3842
+ "step": 319500
3843
+ },
3844
+ {
3845
+ "epoch": 2.96,
3846
+ "learning_rate": 6.140830115532179e-07,
3847
+ "loss": 1.5593,
3848
+ "step": 320000
3849
+ },
3850
+ {
3851
+ "epoch": 2.97,
3852
+ "learning_rate": 5.3691751625877e-07,
3853
+ "loss": 1.5583,
3854
+ "step": 320500
3855
+ },
3856
+ {
3857
+ "epoch": 2.97,
3858
+ "learning_rate": 4.597520209643218e-07,
3859
+ "loss": 1.5566,
3860
+ "step": 321000
3861
+ },
3862
+ {
3863
+ "epoch": 2.98,
3864
+ "learning_rate": 3.825865256698737e-07,
3865
+ "loss": 1.557,
3866
+ "step": 321500
3867
+ },
3868
+ {
3869
+ "epoch": 2.98,
3870
+ "learning_rate": 3.0542103037542556e-07,
3871
+ "loss": 1.5617,
3872
+ "step": 322000
3873
+ },
3874
+ {
3875
+ "epoch": 2.99,
3876
+ "learning_rate": 2.282555350809775e-07,
3877
+ "loss": 1.5574,
3878
+ "step": 322500
3879
+ },
3880
+ {
3881
+ "epoch": 2.99,
3882
+ "learning_rate": 1.5109003978652938e-07,
3883
+ "loss": 1.5566,
3884
+ "step": 323000
3885
+ },
3886
+ {
3887
+ "epoch": 3.0,
3888
+ "learning_rate": 7.392454449208128e-08,
3889
+ "loss": 1.551,
3890
+ "step": 323500
3891
+ },
3892
+ {
3893
+ "epoch": 3.0,
3894
+ "step": 323979,
3895
+ "total_flos": 2.728710874004779e+18,
3896
+ "train_loss": 3.0939811468297327,
3897
+ "train_runtime": 276726.3152,
3898
+ "train_samples_per_second": 37.464,
3899
+ "train_steps_per_second": 1.171
3900
+ }
3901
+ ],
3902
+ "max_steps": 323979,
3903
+ "num_train_epochs": 3,
3904
+ "total_flos": 2.728710874004779e+18,
3905
+ "trial_name": null,
3906
+ "trial_params": null
3907
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:71f7d314091ff901c466dc083cc6a47e66a7f0b7a3970ab63e8c2f39cec6d0e9
3
+ size 3375
vocab.txt ADDED
The diff for this file is too large to render. See raw diff