Hugging Face's logo

Atnafu
/
amhric_xlmr-small

Fill-Mask
Transformers
PyTorch
xlm-roberta
Generated from Trainer
Inference Endpoints
Model card Files Community
Atnafu commited on
Commit
4dd4dad
•
1 Parent(s): 9080158

initial

Browse files
Files changed (11) hide show
  1. README.md +53 -0
  2. all_results.json +14 -0
  3. config.json +28 -0
  4. eval_results.json +9 -0
  5. pytorch_model.bin +3 -0
  6. sentencepiece.bpe.model +3 -0
  7. special_tokens_map.json +9 -0
  8. tokenizer_config.json +22 -0
  9. train_results.json +8 -0
  10. trainer_state.json +2791 -0
  11. training_args.bin +3 -0
README.md ADDED
@@ -0,0 +1,53 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: afl-3.0
3
+ tags:
4
+ - generated_from_trainer
5
+ model-index:
6
+ - name: amh_small
7
+ results: []
8
+ ---
9
+
10
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
11
+ should probably proofread and complete it, then remove this comment. -->
12
+
13
+ # amh_small
14
+
15
+ This model is a fine-tuned version of [Davlan/afro-xlmr-small](https://huggingface.co/Davlan/afro-xlmr-small) on an unknown dataset.
16
+ It achieves the following results on the evaluation set:
17
+ - Loss: 4.2386
18
+
19
+ ## Model description
20
+
21
+ More information needed
22
+
23
+ ## Intended uses & limitations
24
+
25
+ More information needed
26
+
27
+ ## Training and evaluation data
28
+
29
+ More information needed
30
+
31
+ ## Training procedure
32
+
33
+ ### Training hyperparameters
34
+
35
+ The following hyperparameters were used during training:
36
+ - learning_rate: 5e-05
37
+ - train_batch_size: 8
38
+ - eval_batch_size: 8
39
+ - seed: 42
40
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
41
+ - lr_scheduler_type: linear
42
+ - num_epochs: 5.0
43
+
44
+ ### Training results
45
+
46
+
47
+
48
+ ### Framework versions
49
+
50
+ - Transformers 4.30.2
51
+ - Pytorch 2.0.1+cu118
52
+ - Datasets 2.13.0
53
+ - Tokenizers 0.13.3
all_results.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 5.0,
3
+ "eval_loss": 4.238580703735352,
4
+ "eval_runtime": 104.2188,
5
+ "eval_samples": 18156,
6
+ "eval_samples_per_second": 174.21,
7
+ "eval_steps_per_second": 21.781,
8
+ "perplexity": 69.30941140849554,
9
+ "train_loss": 5.372212270862655,
10
+ "train_runtime": 34970.0124,
11
+ "train_samples": 369106,
12
+ "train_samples_per_second": 52.775,
13
+ "train_steps_per_second": 6.597
14
+ }
config.json ADDED
@@ -0,0 +1,28 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "Davlan/afro-xlmr-small",
3
+ "architectures": [
4
+ "XLMRobertaForMaskedLM"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.1,
7
+ "bos_token_id": 0,
8
+ "classifier_dropout": null,
9
+ "eos_token_id": 2,
10
+ "hidden_act": "gelu",
11
+ "hidden_dropout_prob": 0.1,
12
+ "hidden_size": 768,
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 3072,
15
+ "layer_norm_eps": 1e-05,
16
+ "max_position_embeddings": 514,
17
+ "model_type": "xlm-roberta",
18
+ "num_attention_heads": 12,
19
+ "num_hidden_layers": 12,
20
+ "output_past": true,
21
+ "pad_token_id": 1,
22
+ "position_embedding_type": "absolute",
23
+ "torch_dtype": "float32",
24
+ "transformers_version": "4.30.2",
25
+ "type_vocab_size": 1,
26
+ "use_cache": true,
27
+ "vocab_size": 250002
28
+ }
eval_results.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 5.0,
3
+ "eval_loss": 4.238580703735352,
4
+ "eval_runtime": 104.2188,
5
+ "eval_samples": 18156,
6
+ "eval_samples_per_second": 174.21,
7
+ "eval_steps_per_second": 21.781,
8
+ "perplexity": 69.30941140849554
9
+ }
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a9faf3455b22a3ba6f77b969b263833854fefb9e467b14eb5ecf4de83e7aa3e7
3
+ size 1113254457
sentencepiece.bpe.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:962d9ea4dc5fa781966758e4ff09a2ce618864c4a9e19c5bcb3887457e022d51
3
+ size 5479707
special_tokens_map.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<s>",
3
+ "cls_token": "<s>",
4
+ "eos_token": "</s>",
5
+ "mask_token": "<mask>",
6
+ "pad_token": "<pad>",
7
+ "sep_token": "</s>",
8
+ "unk_token": "<unk>"
9
+ }
tokenizer_config.json ADDED
@@ -0,0 +1,22 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<s>",
3
+ "clean_up_tokenization_spaces": true,
4
+ "cls_token": "<s>",
5
+ "eos_token": "</s>",
6
+ "mask_token": {
7
+ "__type": "AddedToken",
8
+ "content": "<mask>",
9
+ "lstrip": true,
10
+ "normalized": true,
11
+ "rstrip": false,
12
+ "single_word": false
13
+ },
14
+ "model_max_length": 1000000000000000019884624838656,
15
+ "pad_token": "<pad>",
16
+ "sep_token": "</s>",
17
+ "sp_model_kwargs": {},
18
+ "tokenizer_class": "XLMRobertaTokenizer",
19
+ "tokenizer_file": null,
20
+ "unk_token": "<unk>",
21
+ "use_fast": true
22
+ }
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 5.0,
3
+ "train_loss": 5.372212270862655,
4
+ "train_runtime": 34970.0124,
5
+ "train_samples": 369106,
6
+ "train_samples_per_second": 52.775,
7
+ "train_steps_per_second": 6.597
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,2791 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 5.0,
5
+ "global_step": 230695,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.01,
12
+ "learning_rate": 4.989293222653287e-05,
13
+ "loss": 9.2291,
14
+ "step": 500
15
+ },
16
+ {
17
+ "epoch": 0.02,
18
+ "learning_rate": 4.978456403476452e-05,
19
+ "loss": 8.5435,
20
+ "step": 1000
21
+ },
22
+ {
23
+ "epoch": 0.03,
24
+ "learning_rate": 4.967619584299617e-05,
25
+ "loss": 8.4546,
26
+ "step": 1500
27
+ },
28
+ {
29
+ "epoch": 0.04,
30
+ "learning_rate": 4.9567827651227814e-05,
31
+ "loss": 8.3564,
32
+ "step": 2000
33
+ },
34
+ {
35
+ "epoch": 0.05,
36
+ "learning_rate": 4.945945945945946e-05,
37
+ "loss": 8.3088,
38
+ "step": 2500
39
+ },
40
+ {
41
+ "epoch": 0.07,
42
+ "learning_rate": 4.935109126769111e-05,
43
+ "loss": 8.2156,
44
+ "step": 3000
45
+ },
46
+ {
47
+ "epoch": 0.08,
48
+ "learning_rate": 4.924272307592276e-05,
49
+ "loss": 8.1808,
50
+ "step": 3500
51
+ },
52
+ {
53
+ "epoch": 0.09,
54
+ "learning_rate": 4.9134354884154406e-05,
55
+ "loss": 8.1035,
56
+ "step": 4000
57
+ },
58
+ {
59
+ "epoch": 0.1,
60
+ "learning_rate": 4.902598669238606e-05,
61
+ "loss": 8.032,
62
+ "step": 4500
63
+ },
64
+ {
65
+ "epoch": 0.11,
66
+ "learning_rate": 4.89176185006177e-05,
67
+ "loss": 7.9508,
68
+ "step": 5000
69
+ },
70
+ {
71
+ "epoch": 0.12,
72
+ "learning_rate": 4.8809250308849345e-05,
73
+ "loss": 7.8717,
74
+ "step": 5500
75
+ },
76
+ {
77
+ "epoch": 0.13,
78
+ "learning_rate": 4.8700882117081e-05,
79
+ "loss": 7.8179,
80
+ "step": 6000
81
+ },
82
+ {
83
+ "epoch": 0.14,
84
+ "learning_rate": 4.8592513925312644e-05,
85
+ "loss": 7.729,
86
+ "step": 6500
87
+ },
88
+ {
89
+ "epoch": 0.15,
90
+ "learning_rate": 4.84841457335443e-05,
91
+ "loss": 7.7062,
92
+ "step": 7000
93
+ },
94
+ {
95
+ "epoch": 0.16,
96
+ "learning_rate": 4.837577754177594e-05,
97
+ "loss": 7.623,
98
+ "step": 7500
99
+ },
100
+ {
101
+ "epoch": 0.17,
102
+ "learning_rate": 4.826740935000759e-05,
103
+ "loss": 7.5969,
104
+ "step": 8000
105
+ },
106
+ {
107
+ "epoch": 0.18,
108
+ "learning_rate": 4.8159041158239236e-05,
109
+ "loss": 7.5306,
110
+ "step": 8500
111
+ },
112
+ {
113
+ "epoch": 0.2,
114
+ "learning_rate": 4.805067296647088e-05,
115
+ "loss": 7.4785,
116
+ "step": 9000
117
+ },
118
+ {
119
+ "epoch": 0.21,
120
+ "learning_rate": 4.794252151108607e-05,
121
+ "loss": 7.4348,
122
+ "step": 9500
123
+ },
124
+ {
125
+ "epoch": 0.22,
126
+ "learning_rate": 4.7834153319317714e-05,
127
+ "loss": 7.399,
128
+ "step": 10000
129
+ },
130
+ {
131
+ "epoch": 0.23,
132
+ "learning_rate": 4.772578512754937e-05,
133
+ "loss": 7.3493,
134
+ "step": 10500
135
+ },
136
+ {
137
+ "epoch": 0.24,
138
+ "learning_rate": 4.7617416935781014e-05,
139
+ "loss": 7.287,
140
+ "step": 11000
141
+ },
142
+ {
143
+ "epoch": 0.25,
144
+ "learning_rate": 4.750904874401266e-05,
145
+ "loss": 7.2254,
146
+ "step": 11500
147
+ },
148
+ {
149
+ "epoch": 0.26,
150
+ "learning_rate": 4.7400897288627846e-05,
151
+ "loss": 7.1961,
152
+ "step": 12000
153
+ },
154
+ {
155
+ "epoch": 0.27,
156
+ "learning_rate": 4.729252909685949e-05,
157
+ "loss": 7.1299,
158
+ "step": 12500
159
+ },
160
+ {
161
+ "epoch": 0.28,
162
+ "learning_rate": 4.718416090509114e-05,
163
+ "loss": 7.1257,
164
+ "step": 13000
165
+ },
166
+ {
167
+ "epoch": 0.29,
168
+ "learning_rate": 4.7075792713322785e-05,
169
+ "loss": 7.074,
170
+ "step": 13500
171
+ },
172
+ {
173
+ "epoch": 0.3,
174
+ "learning_rate": 4.696742452155444e-05,
175
+ "loss": 7.052,
176
+ "step": 14000
177
+ },
178
+ {
179
+ "epoch": 0.31,
180
+ "learning_rate": 4.685927306616962e-05,
181
+ "loss": 6.9863,
182
+ "step": 14500
183
+ },
184
+ {
185
+ "epoch": 0.33,
186
+ "learning_rate": 4.675090487440127e-05,
187
+ "loss": 6.9617,
188
+ "step": 15000
189
+ },
190
+ {
191
+ "epoch": 0.34,
192
+ "learning_rate": 4.664253668263292e-05,
193
+ "loss": 6.9034,
194
+ "step": 15500
195
+ },
196
+ {
197
+ "epoch": 0.35,
198
+ "learning_rate": 4.653416849086456e-05,
199
+ "loss": 6.8683,
200
+ "step": 16000
201
+ },
202
+ {
203
+ "epoch": 0.36,
204
+ "learning_rate": 4.642601703547975e-05,
205
+ "loss": 6.8567,
206
+ "step": 16500
207
+ },
208
+ {
209
+ "epoch": 0.37,
210
+ "learning_rate": 4.6317865580094936e-05,
211
+ "loss": 6.8268,
212
+ "step": 17000
213
+ },
214
+ {
215
+ "epoch": 0.38,
216
+ "learning_rate": 4.6209497388326575e-05,
217
+ "loss": 6.7758,
218
+ "step": 17500
219
+ },
220
+ {
221
+ "epoch": 0.39,
222
+ "learning_rate": 4.610112919655823e-05,
223
+ "loss": 6.7915,
224
+ "step": 18000
225
+ },
226
+ {
227
+ "epoch": 0.4,
228
+ "learning_rate": 4.5992761004789874e-05,
229
+ "loss": 6.7391,
230
+ "step": 18500
231
+ },
232
+ {
233
+ "epoch": 0.41,
234
+ "learning_rate": 4.588439281302153e-05,
235
+ "loss": 6.7204,
236
+ "step": 19000
237
+ },
238
+ {
239
+ "epoch": 0.42,
240
+ "learning_rate": 4.5776024621253173e-05,
241
+ "loss": 6.6562,
242
+ "step": 19500
243
+ },
244
+ {
245
+ "epoch": 0.43,
246
+ "learning_rate": 4.566765642948482e-05,
247
+ "loss": 6.6638,
248
+ "step": 20000
249
+ },
250
+ {
251
+ "epoch": 0.44,
252
+ "learning_rate": 4.5559288237716466e-05,
253
+ "loss": 6.6292,
254
+ "step": 20500
255
+ },
256
+ {
257
+ "epoch": 0.46,
258
+ "learning_rate": 4.545092004594811e-05,
259
+ "loss": 6.6086,
260
+ "step": 21000
261
+ },
262
+ {
263
+ "epoch": 0.47,
264
+ "learning_rate": 4.5342551854179765e-05,
265
+ "loss": 6.5568,
266
+ "step": 21500
267
+ },
268
+ {
269
+ "epoch": 0.48,
270
+ "learning_rate": 4.523418366241141e-05,
271
+ "loss": 6.5345,
272
+ "step": 22000
273
+ },
274
+ {
275
+ "epoch": 0.49,
276
+ "learning_rate": 4.512581547064306e-05,
277
+ "loss": 6.556,
278
+ "step": 22500
279
+ },
280
+ {
281
+ "epoch": 0.5,
282
+ "learning_rate": 4.501744727887471e-05,
283
+ "loss": 6.5297,
284
+ "step": 23000
285
+ },
286
+ {
287
+ "epoch": 0.51,
288
+ "learning_rate": 4.490907908710636e-05,
289
+ "loss": 6.4959,
290
+ "step": 23500
291
+ },
292
+ {
293
+ "epoch": 0.52,
294
+ "learning_rate": 4.4800710895338e-05,
295
+ "loss": 6.4772,
296
+ "step": 24000
297
+ },
298
+ {
299
+ "epoch": 0.53,
300
+ "learning_rate": 4.469255943995318e-05,
301
+ "loss": 6.4633,
302
+ "step": 24500
303
+ },
304
+ {
305
+ "epoch": 0.54,
306
+ "learning_rate": 4.4584191248184836e-05,
307
+ "loss": 6.4274,
308
+ "step": 25000
309
+ },
310
+ {
311
+ "epoch": 0.55,
312
+ "learning_rate": 4.447582305641648e-05,
313
+ "loss": 6.4137,
314
+ "step": 25500
315
+ },
316
+ {
317
+ "epoch": 0.56,
318
+ "learning_rate": 4.436745486464813e-05,
319
+ "loss": 6.3865,
320
+ "step": 26000
321
+ },
322
+ {
323
+ "epoch": 0.57,
324
+ "learning_rate": 4.4259303409263314e-05,
325
+ "loss": 6.4233,
326
+ "step": 26500
327
+ },
328
+ {
329
+ "epoch": 0.59,
330
+ "learning_rate": 4.415093521749497e-05,
331
+ "loss": 6.3557,
332
+ "step": 27000
333
+ },
334
+ {
335
+ "epoch": 0.6,
336
+ "learning_rate": 4.404278376211015e-05,
337
+ "loss": 6.3371,
338
+ "step": 27500
339
+ },
340
+ {
341
+ "epoch": 0.61,
342
+ "learning_rate": 4.39344155703418e-05,
343
+ "loss": 6.3477,
344
+ "step": 28000
345
+ },
346
+ {
347
+ "epoch": 0.62,
348
+ "learning_rate": 4.382604737857344e-05,
349
+ "loss": 6.3256,
350
+ "step": 28500
351
+ },
352
+ {
353
+ "epoch": 0.63,
354
+ "learning_rate": 4.3717679186805086e-05,
355
+ "loss": 6.3146,
356
+ "step": 29000
357
+ },
358
+ {
359
+ "epoch": 0.64,
360
+ "learning_rate": 4.360931099503674e-05,
361
+ "loss": 6.2812,
362
+ "step": 29500
363
+ },
364
+ {
365
+ "epoch": 0.65,
366
+ "learning_rate": 4.3500942803268385e-05,
367
+ "loss": 6.2653,
368
+ "step": 30000
369
+ },
370
+ {
371
+ "epoch": 0.66,
372
+ "learning_rate": 4.339257461150004e-05,
373
+ "loss": 6.2625,
374
+ "step": 30500
375
+ },
376
+ {
377
+ "epoch": 0.67,
378
+ "learning_rate": 4.3284206419731684e-05,
379
+ "loss": 6.2311,
380
+ "step": 31000
381
+ },
382
+ {
383
+ "epoch": 0.68,
384
+ "learning_rate": 4.317583822796333e-05,
385
+ "loss": 6.2153,
386
+ "step": 31500
387
+ },
388
+ {
389
+ "epoch": 0.69,
390
+ "learning_rate": 4.306747003619498e-05,
391
+ "loss": 6.2252,
392
+ "step": 32000
393
+ },
394
+ {
395
+ "epoch": 0.7,
396
+ "learning_rate": 4.295910184442662e-05,
397
+ "loss": 6.2012,
398
+ "step": 32500
399
+ },
400
+ {
401
+ "epoch": 0.72,
402
+ "learning_rate": 4.2850733652658276e-05,
403
+ "loss": 6.1814,
404
+ "step": 33000
405
+ },
406
+ {
407
+ "epoch": 0.73,
408
+ "learning_rate": 4.274236546088992e-05,
409
+ "loss": 6.1341,
410
+ "step": 33500
411
+ },
412
+ {
413
+ "epoch": 0.74,
414
+ "learning_rate": 4.263399726912157e-05,
415
+ "loss": 6.1333,
416
+ "step": 34000
417
+ },
418
+ {
419
+ "epoch": 0.75,
420
+ "learning_rate": 4.252562907735322e-05,
421
+ "loss": 6.137,
422
+ "step": 34500
423
+ },
424
+ {
425
+ "epoch": 0.76,
426
+ "learning_rate": 4.241726088558486e-05,
427
+ "loss": 6.1115,
428
+ "step": 35000
429
+ },
430
+ {
431
+ "epoch": 0.77,
432
+ "learning_rate": 4.230932616658359e-05,
433
+ "loss": 6.1138,
434
+ "step": 35500
435
+ },
436
+ {
437
+ "epoch": 0.78,
438
+ "learning_rate": 4.2200957974815234e-05,
439
+ "loss": 6.1054,
440
+ "step": 36000
441
+ },
442
+ {
443
+ "epoch": 0.79,
444
+ "learning_rate": 4.209258978304688e-05,
445
+ "loss": 6.0865,
446
+ "step": 36500
447
+ },
448
+ {
449
+ "epoch": 0.8,
450
+ "learning_rate": 4.1984438327662066e-05,
451
+ "loss": 6.0722,
452
+ "step": 37000
453
+ },
454
+ {
455
+ "epoch": 0.81,
456
+ "learning_rate": 4.187607013589371e-05,
457
+ "loss": 6.0801,
458
+ "step": 37500
459
+ },
460
+ {
461
+ "epoch": 0.82,
462
+ "learning_rate": 4.1767701944125365e-05,
463
+ "loss": 6.0378,
464
+ "step": 38000
465
+ },
466
+ {
467
+ "epoch": 0.83,
468
+ "learning_rate": 4.165933375235701e-05,
469
+ "loss": 6.0162,
470
+ "step": 38500
471
+ },
472
+ {
473
+ "epoch": 0.85,
474
+ "learning_rate": 4.155096556058866e-05,
475
+ "loss": 6.0366,
476
+ "step": 39000
477
+ },
478
+ {
479
+ "epoch": 0.86,
480
+ "learning_rate": 4.1442597368820304e-05,
481
+ "loss": 6.0294,
482
+ "step": 39500
483
+ },
484
+ {
485
+ "epoch": 0.87,
486
+ "learning_rate": 4.133422917705195e-05,
487
+ "loss": 6.0124,
488
+ "step": 40000
489
+ },
490
+ {
491
+ "epoch": 0.88,
492
+ "learning_rate": 4.12258609852836e-05,
493
+ "loss": 5.9881,
494
+ "step": 40500
495
+ },
496
+ {
497
+ "epoch": 0.89,
498
+ "learning_rate": 4.111749279351525e-05,
499
+ "loss": 5.9678,
500
+ "step": 41000
501
+ },
502
+ {
503
+ "epoch": 0.9,
504
+ "learning_rate": 4.1009124601746896e-05,
505
+ "loss": 5.9707,
506
+ "step": 41500
507
+ },
508
+ {
509
+ "epoch": 0.91,
510
+ "learning_rate": 4.090097314636208e-05,
511
+ "loss": 5.976,
512
+ "step": 42000
513
+ },
514
+ {
515
+ "epoch": 0.92,
516
+ "learning_rate": 4.079260495459373e-05,
517
+ "loss": 5.9419,
518
+ "step": 42500
519
+ },
520
+ {
521
+ "epoch": 0.93,
522
+ "learning_rate": 4.0684453499208915e-05,
523
+ "loss": 5.9158,
524
+ "step": 43000
525
+ },
526
+ {
527
+ "epoch": 0.94,
528
+ "learning_rate": 4.057608530744057e-05,
529
+ "loss": 5.9611,
530
+ "step": 43500
531
+ },
532
+ {
533
+ "epoch": 0.95,
534
+ "learning_rate": 4.046771711567221e-05,
535
+ "loss": 5.9104,
536
+ "step": 44000
537
+ },
538
+ {
539
+ "epoch": 0.96,
540
+ "learning_rate": 4.035934892390385e-05,
541
+ "loss": 5.914,
542
+ "step": 44500
543
+ },
544
+ {
545
+ "epoch": 0.98,
546
+ "learning_rate": 4.0250980732135506e-05,
547
+ "loss": 5.8819,
548
+ "step": 45000
549
+ },
550
+ {
551
+ "epoch": 0.99,
552
+ "learning_rate": 4.014261254036715e-05,
553
+ "loss": 5.891,
554
+ "step": 45500
555
+ },
556
+ {
557
+ "epoch": 1.0,
558
+ "learning_rate": 4.0034244348598806e-05,
559
+ "loss": 5.8627,
560
+ "step": 46000
561
+ },
562
+ {
563
+ "epoch": 1.01,
564
+ "learning_rate": 3.992587615683045e-05,
565
+ "loss": 5.8258,
566
+ "step": 46500
567
+ },
568
+ {
569
+ "epoch": 1.02,
570
+ "learning_rate": 3.98175079650621e-05,
571
+ "loss": 5.8346,
572
+ "step": 47000
573
+ },
574
+ {
575
+ "epoch": 1.03,
576
+ "learning_rate": 3.970957324606082e-05,
577
+ "loss": 5.8269,
578
+ "step": 47500
579
+ },
580
+ {
581
+ "epoch": 1.04,
582
+ "learning_rate": 3.960120505429247e-05,
583
+ "loss": 5.8253,
584
+ "step": 48000
585
+ },
586
+ {
587
+ "epoch": 1.05,
588
+ "learning_rate": 3.949283686252411e-05,
589
+ "loss": 5.8178,
590
+ "step": 48500
591
+ },
592
+ {
593
+ "epoch": 1.06,
594
+ "learning_rate": 3.938446867075576e-05,
595
+ "loss": 5.7758,
596
+ "step": 49000
597
+ },
598
+ {
599
+ "epoch": 1.07,
600
+ "learning_rate": 3.927610047898741e-05,
601
+ "loss": 5.7911,
602
+ "step": 49500
603
+ },
604
+ {
605
+ "epoch": 1.08,
606
+ "learning_rate": 3.9167732287219056e-05,
607
+ "loss": 5.7876,
608
+ "step": 50000
609
+ },
610
+ {
611
+ "epoch": 1.09,
612
+ "learning_rate": 3.905936409545071e-05,
613
+ "loss": 5.7695,
614
+ "step": 50500
615
+ },
616
+ {
617
+ "epoch": 1.11,
618
+ "learning_rate": 3.895121264006589e-05,
619
+ "loss": 5.7523,
620
+ "step": 51000
621
+ },
622
+ {
623
+ "epoch": 1.12,
624
+ "learning_rate": 3.884284444829754e-05,
625
+ "loss": 5.7571,
626
+ "step": 51500
627
+ },
628
+ {
629
+ "epoch": 1.13,
630
+ "learning_rate": 3.873447625652919e-05,
631
+ "loss": 5.7947,
632
+ "step": 52000
633
+ },
634
+ {
635
+ "epoch": 1.14,
636
+ "learning_rate": 3.8626108064760834e-05,
637
+ "loss": 5.7286,
638
+ "step": 52500
639
+ },
640
+ {
641
+ "epoch": 1.15,
642
+ "learning_rate": 3.851773987299248e-05,
643
+ "loss": 5.7481,
644
+ "step": 53000
645
+ },
646
+ {
647
+ "epoch": 1.16,
648
+ "learning_rate": 3.8409371681224126e-05,
649
+ "loss": 5.7334,
650
+ "step": 53500
651
+ },
652
+ {
653
+ "epoch": 1.17,
654
+ "learning_rate": 3.830100348945578e-05,
655
+ "loss": 5.7235,
656
+ "step": 54000
657
+ },
658
+ {
659
+ "epoch": 1.18,
660
+ "learning_rate": 3.8192852034070965e-05,
661
+ "loss": 5.6954,
662
+ "step": 54500
663
+ },
664
+ {
665
+ "epoch": 1.19,
666
+ "learning_rate": 3.808448384230261e-05,
667
+ "loss": 5.7148,
668
+ "step": 55000
669
+ },
670
+ {
671
+ "epoch": 1.2,
672
+ "learning_rate": 3.797611565053426e-05,
673
+ "loss": 5.6986,
674
+ "step": 55500
675
+ },
676
+ {
677
+ "epoch": 1.21,
678
+ "learning_rate": 3.7867747458765904e-05,
679
+ "loss": 5.693,
680
+ "step": 56000
681
+ },
682
+ {
683
+ "epoch": 1.22,
684
+ "learning_rate": 3.775937926699755e-05,
685
+ "loss": 5.6841,
686
+ "step": 56500
687
+ },
688
+ {
689
+ "epoch": 1.24,
690
+ "learning_rate": 3.7651011075229197e-05,
691
+ "loss": 5.6561,
692
+ "step": 57000
693
+ },
694
+ {
695
+ "epoch": 1.25,
696
+ "learning_rate": 3.754264288346085e-05,
697
+ "loss": 5.634,
698
+ "step": 57500
699
+ },
700
+ {
701
+ "epoch": 1.26,
702
+ "learning_rate": 3.7434274691692496e-05,
703
+ "loss": 5.641,
704
+ "step": 58000
705
+ },
706
+ {
707
+ "epoch": 1.27,
708
+ "learning_rate": 3.732612323630768e-05,
709
+ "loss": 5.6495,
710
+ "step": 58500
711
+ },
712
+ {
713
+ "epoch": 1.28,
714
+ "learning_rate": 3.721775504453933e-05,
715
+ "loss": 5.6402,
716
+ "step": 59000
717
+ },
718
+ {
719
+ "epoch": 1.29,
720
+ "learning_rate": 3.7109386852770975e-05,
721
+ "loss": 5.5963,
722
+ "step": 59500
723
+ },
724
+ {
725
+ "epoch": 1.3,
726
+ "learning_rate": 3.700101866100262e-05,
727
+ "loss": 5.6276,
728
+ "step": 60000
729
+ },
730
+ {
731
+ "epoch": 1.31,
732
+ "learning_rate": 3.6892650469234274e-05,
733
+ "loss": 5.6346,
734
+ "step": 60500
735
+ },
736
+ {
737
+ "epoch": 1.32,
738
+ "learning_rate": 3.678428227746592e-05,
739
+ "loss": 5.6149,
740
+ "step": 61000
741
+ },
742
+ {
743
+ "epoch": 1.33,
744
+ "learning_rate": 3.6676130822081106e-05,
745
+ "loss": 5.5947,
746
+ "step": 61500
747
+ },
748
+ {
749
+ "epoch": 1.34,
750
+ "learning_rate": 3.656776263031275e-05,
751
+ "loss": 5.5881,
752
+ "step": 62000
753
+ },
754
+ {
755
+ "epoch": 1.35,
756
+ "learning_rate": 3.64593944385444e-05,
757
+ "loss": 5.5827,
758
+ "step": 62500
759
+ },
760
+ {
761
+ "epoch": 1.37,
762
+ "learning_rate": 3.635102624677605e-05,
763
+ "loss": 5.5667,
764
+ "step": 63000
765
+ },
766
+ {
767
+ "epoch": 1.38,
768
+ "learning_rate": 3.624265805500769e-05,
769
+ "loss": 5.5804,
770
+ "step": 63500
771
+ },
772
+ {
773
+ "epoch": 1.39,
774
+ "learning_rate": 3.6134289863239344e-05,
775
+ "loss": 5.5845,
776
+ "step": 64000
777
+ },
778
+ {
779
+ "epoch": 1.4,
780
+ "learning_rate": 3.602592167147099e-05,
781
+ "loss": 5.5827,
782
+ "step": 64500
783
+ },
784
+ {
785
+ "epoch": 1.41,
786
+ "learning_rate": 3.591755347970264e-05,
787
+ "loss": 5.5525,
788
+ "step": 65000
789
+ },
790
+ {
791
+ "epoch": 1.42,
792
+ "learning_rate": 3.580918528793429e-05,
793
+ "loss": 5.5575,
794
+ "step": 65500
795
+ },
796
+ {
797
+ "epoch": 1.43,
798
+ "learning_rate": 3.5700817096165936e-05,
799
+ "loss": 5.5406,
800
+ "step": 66000
801
+ },
802
+ {
803
+ "epoch": 1.44,
804
+ "learning_rate": 3.559266564078112e-05,
805
+ "loss": 5.5457,
806
+ "step": 66500
807
+ },
808
+ {
809
+ "epoch": 1.45,
810
+ "learning_rate": 3.548429744901277e-05,
811
+ "loss": 5.5298,
812
+ "step": 67000
813
+ },
814
+ {
815
+ "epoch": 1.46,
816
+ "learning_rate": 3.5375929257244415e-05,
817
+ "loss": 5.5226,
818
+ "step": 67500
819
+ },
820
+ {
821
+ "epoch": 1.47,
822
+ "learning_rate": 3.526756106547606e-05,
823
+ "loss": 5.4872,
824
+ "step": 68000
825
+ },
826
+ {
827
+ "epoch": 1.48,
828
+ "learning_rate": 3.5159192873707714e-05,
829
+ "loss": 5.496,
830
+ "step": 68500
831
+ },
832
+ {
833
+ "epoch": 1.5,
834
+ "learning_rate": 3.5051041418322894e-05,
835
+ "loss": 5.4846,
836
+ "step": 69000
837
+ },
838
+ {
839
+ "epoch": 1.51,
840
+ "learning_rate": 3.494267322655455e-05,
841
+ "loss": 5.494,
842
+ "step": 69500
843
+ },
844
+ {
845
+ "epoch": 1.52,
846
+ "learning_rate": 3.483430503478619e-05,
847
+ "loss": 5.4753,
848
+ "step": 70000
849
+ },
850
+ {
851
+ "epoch": 1.53,
852
+ "learning_rate": 3.472593684301784e-05,
853
+ "loss": 5.5017,
854
+ "step": 70500
855
+ },
856
+ {
857
+ "epoch": 1.54,
858
+ "learning_rate": 3.4618002124016566e-05,
859
+ "loss": 5.489,
860
+ "step": 71000
861
+ },
862
+ {
863
+ "epoch": 1.55,
864
+ "learning_rate": 3.450963393224821e-05,
865
+ "loss": 5.4613,
866
+ "step": 71500
867
+ },
868
+ {
869
+ "epoch": 1.56,
870
+ "learning_rate": 3.440126574047985e-05,
871
+ "loss": 5.4371,
872
+ "step": 72000
873
+ },
874
+ {
875
+ "epoch": 1.57,
876
+ "learning_rate": 3.4292897548711504e-05,
877
+ "loss": 5.4453,
878
+ "step": 72500
879
+ },
880
+ {
881
+ "epoch": 1.58,
882
+ "learning_rate": 3.418452935694315e-05,
883
+ "loss": 5.4485,
884
+ "step": 73000
885
+ },
886
+ {
887
+ "epoch": 1.59,
888
+ "learning_rate": 3.40761611651748e-05,
889
+ "loss": 5.4734,
890
+ "step": 73500
891
+ },
892
+ {
893
+ "epoch": 1.6,
894
+ "learning_rate": 3.396779297340645e-05,
895
+ "loss": 5.4371,
896
+ "step": 74000
897
+ },
898
+ {
899
+ "epoch": 1.61,
900
+ "learning_rate": 3.3859641518021636e-05,
901
+ "loss": 5.4367,
902
+ "step": 74500
903
+ },
904
+ {
905
+ "epoch": 1.63,
906
+ "learning_rate": 3.375127332625328e-05,
907
+ "loss": 5.4137,
908
+ "step": 75000
909
+ },
910
+ {
911
+ "epoch": 1.64,
912
+ "learning_rate": 3.364290513448493e-05,
913
+ "loss": 5.4397,
914
+ "step": 75500
915
+ },
916
+ {
917
+ "epoch": 1.65,
918
+ "learning_rate": 3.3534536942716575e-05,
919
+ "loss": 5.4075,
920
+ "step": 76000
921
+ },
922
+ {
923
+ "epoch": 1.66,
924
+ "learning_rate": 3.342616875094822e-05,
925
+ "loss": 5.426,
926
+ "step": 76500
927
+ },
928
+ {
929
+ "epoch": 1.67,
930
+ "learning_rate": 3.3317800559179874e-05,
931
+ "loss": 5.4072,
932
+ "step": 77000
933
+ },
934
+ {
935
+ "epoch": 1.68,
936
+ "learning_rate": 3.320943236741152e-05,
937
+ "loss": 5.3822,
938
+ "step": 77500
939
+ },
940
+ {
941
+ "epoch": 1.69,
942
+ "learning_rate": 3.3101064175643166e-05,
943
+ "loss": 5.3867,
944
+ "step": 78000
945
+ },
946
+ {
947
+ "epoch": 1.7,
948
+ "learning_rate": 3.299291272025835e-05,
949
+ "loss": 5.4032,
950
+ "step": 78500
951
+ },
952
+ {
953
+ "epoch": 1.71,
954
+ "learning_rate": 3.288454452849e-05,
955
+ "loss": 5.3875,
956
+ "step": 79000
957
+ },
958
+ {
959
+ "epoch": 1.72,
960
+ "learning_rate": 3.2776176336721645e-05,
961
+ "loss": 5.382,
962
+ "step": 79500
963
+ },
964
+ {
965
+ "epoch": 1.73,
966
+ "learning_rate": 3.266802488133683e-05,
967
+ "loss": 5.3784,
968
+ "step": 80000
969
+ },
970
+ {
971
+ "epoch": 1.74,
972
+ "learning_rate": 3.255965668956848e-05,
973
+ "loss": 5.3607,
974
+ "step": 80500
975
+ },
976
+ {
977
+ "epoch": 1.76,
978
+ "learning_rate": 3.2451288497800124e-05,
979
+ "loss": 5.3523,
980
+ "step": 81000
981
+ },
982
+ {
983
+ "epoch": 1.77,
984
+ "learning_rate": 3.234292030603178e-05,
985
+ "loss": 5.3397,
986
+ "step": 81500
987
+ },
988
+ {
989
+ "epoch": 1.78,
990
+ "learning_rate": 3.223455211426342e-05,
991
+ "loss": 5.3615,
992
+ "step": 82000
993
+ },
994
+ {
995
+ "epoch": 1.79,
996
+ "learning_rate": 3.2126183922495076e-05,
997
+ "loss": 5.3337,
998
+ "step": 82500
999
+ },
1000
+ {
1001
+ "epoch": 1.8,
1002
+ "learning_rate": 3.2017815730726716e-05,
1003
+ "loss": 5.3297,
1004
+ "step": 83000
1005
+ },
1006
+ {
1007
+ "epoch": 1.81,
1008
+ "learning_rate": 3.190944753895836e-05,
1009
+ "loss": 5.3278,
1010
+ "step": 83500
1011
+ },
1012
+ {
1013
+ "epoch": 1.82,
1014
+ "learning_rate": 3.1801079347190015e-05,
1015
+ "loss": 5.3225,
1016
+ "step": 84000
1017
+ },
1018
+ {
1019
+ "epoch": 1.83,
1020
+ "learning_rate": 3.1692927891805194e-05,
1021
+ "loss": 5.3084,
1022
+ "step": 84500
1023
+ },
1024
+ {
1025
+ "epoch": 1.84,
1026
+ "learning_rate": 3.158477643642038e-05,
1027
+ "loss": 5.3226,
1028
+ "step": 85000
1029
+ },
1030
+ {
1031
+ "epoch": 1.85,
1032
+ "learning_rate": 3.1476408244652034e-05,
1033
+ "loss": 5.3311,
1034
+ "step": 85500
1035
+ },
1036
+ {
1037
+ "epoch": 1.86,
1038
+ "learning_rate": 3.136804005288368e-05,
1039
+ "loss": 5.3289,
1040
+ "step": 86000
1041
+ },
1042
+ {
1043
+ "epoch": 1.87,
1044
+ "learning_rate": 3.1259671861115326e-05,
1045
+ "loss": 5.3217,
1046
+ "step": 86500
1047
+ },
1048
+ {
1049
+ "epoch": 1.89,
1050
+ "learning_rate": 3.115152040573051e-05,
1051
+ "loss": 5.3092,
1052
+ "step": 87000
1053
+ },
1054
+ {
1055
+ "epoch": 1.9,
1056
+ "learning_rate": 3.104315221396216e-05,
1057
+ "loss": 5.3042,
1058
+ "step": 87500
1059
+ },
1060
+ {
1061
+ "epoch": 1.91,
1062
+ "learning_rate": 3.0934784022193805e-05,
1063
+ "loss": 5.3147,
1064
+ "step": 88000
1065
+ },
1066
+ {
1067
+ "epoch": 1.92,
1068
+ "learning_rate": 3.082641583042545e-05,
1069
+ "loss": 5.2633,
1070
+ "step": 88500
1071
+ },
1072
+ {
1073
+ "epoch": 1.93,
1074
+ "learning_rate": 3.0718047638657104e-05,
1075
+ "loss": 5.2619,
1076
+ "step": 89000
1077
+ },
1078
+ {
1079
+ "epoch": 1.94,
1080
+ "learning_rate": 3.0609896183272284e-05,
1081
+ "loss": 5.3033,
1082
+ "step": 89500
1083
+ },
1084
+ {
1085
+ "epoch": 1.95,
1086
+ "learning_rate": 3.0501527991503937e-05,
1087
+ "loss": 5.2377,
1088
+ "step": 90000
1089
+ },
1090
+ {
1091
+ "epoch": 1.96,
1092
+ "learning_rate": 3.0393159799735583e-05,
1093
+ "loss": 5.266,
1094
+ "step": 90500
1095
+ },
1096
+ {
1097
+ "epoch": 1.97,
1098
+ "learning_rate": 3.0284791607967233e-05,
1099
+ "loss": 5.2726,
1100
+ "step": 91000
1101
+ },
1102
+ {
1103
+ "epoch": 1.98,
1104
+ "learning_rate": 3.017642341619888e-05,
1105
+ "loss": 5.2806,
1106
+ "step": 91500
1107
+ },
1108
+ {
1109
+ "epoch": 1.99,
1110
+ "learning_rate": 3.0068271960814065e-05,
1111
+ "loss": 5.2758,
1112
+ "step": 92000
1113
+ },
1114
+ {
1115
+ "epoch": 2.0,
1116
+ "learning_rate": 2.995990376904571e-05,
1117
+ "loss": 5.2737,
1118
+ "step": 92500
1119
+ },
1120
+ {
1121
+ "epoch": 2.02,
1122
+ "learning_rate": 2.9851535577277358e-05,
1123
+ "loss": 5.2566,
1124
+ "step": 93000
1125
+ },
1126
+ {
1127
+ "epoch": 2.03,
1128
+ "learning_rate": 2.9743384121892544e-05,
1129
+ "loss": 5.2196,
1130
+ "step": 93500
1131
+ },
1132
+ {
1133
+ "epoch": 2.04,
1134
+ "learning_rate": 2.9635015930124194e-05,
1135
+ "loss": 5.2356,
1136
+ "step": 94000
1137
+ },
1138
+ {
1139
+ "epoch": 2.05,
1140
+ "learning_rate": 2.952664773835584e-05,
1141
+ "loss": 5.2116,
1142
+ "step": 94500
1143
+ },
1144
+ {
1145
+ "epoch": 2.06,
1146
+ "learning_rate": 2.9418279546587486e-05,
1147
+ "loss": 5.235,
1148
+ "step": 95000
1149
+ },
1150
+ {
1151
+ "epoch": 2.07,
1152
+ "learning_rate": 2.9309911354819136e-05,
1153
+ "loss": 5.2186,
1154
+ "step": 95500
1155
+ },
1156
+ {
1157
+ "epoch": 2.08,
1158
+ "learning_rate": 2.9201543163050782e-05,
1159
+ "loss": 5.2145,
1160
+ "step": 96000
1161
+ },
1162
+ {
1163
+ "epoch": 2.09,
1164
+ "learning_rate": 2.909317497128243e-05,
1165
+ "loss": 5.1935,
1166
+ "step": 96500
1167
+ },
1168
+ {
1169
+ "epoch": 2.1,
1170
+ "learning_rate": 2.8984806779514078e-05,
1171
+ "loss": 5.2145,
1172
+ "step": 97000
1173
+ },
1174
+ {
1175
+ "epoch": 2.11,
1176
+ "learning_rate": 2.8876438587745724e-05,
1177
+ "loss": 5.2262,
1178
+ "step": 97500
1179
+ },
1180
+ {
1181
+ "epoch": 2.12,
1182
+ "learning_rate": 2.876828713236091e-05,
1183
+ "loss": 5.211,
1184
+ "step": 98000
1185
+ },
1186
+ {
1187
+ "epoch": 2.13,
1188
+ "learning_rate": 2.8659918940592557e-05,
1189
+ "loss": 5.1676,
1190
+ "step": 98500
1191
+ },
1192
+ {
1193
+ "epoch": 2.15,
1194
+ "learning_rate": 2.8551550748824206e-05,
1195
+ "loss": 5.1791,
1196
+ "step": 99000
1197
+ },
1198
+ {
1199
+ "epoch": 2.16,
1200
+ "learning_rate": 2.8443182557055852e-05,
1201
+ "loss": 5.1993,
1202
+ "step": 99500
1203
+ },
1204
+ {
1205
+ "epoch": 2.17,
1206
+ "learning_rate": 2.833503110167104e-05,
1207
+ "loss": 5.1698,
1208
+ "step": 100000
1209
+ },
1210
+ {
1211
+ "epoch": 2.18,
1212
+ "learning_rate": 2.8226662909902685e-05,
1213
+ "loss": 5.1661,
1214
+ "step": 100500
1215
+ },
1216
+ {
1217
+ "epoch": 2.19,
1218
+ "learning_rate": 2.8118294718134335e-05,
1219
+ "loss": 5.1772,
1220
+ "step": 101000
1221
+ },
1222
+ {
1223
+ "epoch": 2.2,
1224
+ "learning_rate": 2.800992652636598e-05,
1225
+ "loss": 5.1629,
1226
+ "step": 101500
1227
+ },
1228
+ {
1229
+ "epoch": 2.21,
1230
+ "learning_rate": 2.7901558334597634e-05,
1231
+ "loss": 5.1636,
1232
+ "step": 102000
1233
+ },
1234
+ {
1235
+ "epoch": 2.22,
1236
+ "learning_rate": 2.7793406879212813e-05,
1237
+ "loss": 5.1611,
1238
+ "step": 102500
1239
+ },
1240
+ {
1241
+ "epoch": 2.23,
1242
+ "learning_rate": 2.7685038687444463e-05,
1243
+ "loss": 5.172,
1244
+ "step": 103000
1245
+ },
1246
+ {
1247
+ "epoch": 2.24,
1248
+ "learning_rate": 2.757667049567611e-05,
1249
+ "loss": 5.1525,
1250
+ "step": 103500
1251
+ },
1252
+ {
1253
+ "epoch": 2.25,
1254
+ "learning_rate": 2.7468302303907756e-05,
1255
+ "loss": 5.1161,
1256
+ "step": 104000
1257
+ },
1258
+ {
1259
+ "epoch": 2.26,
1260
+ "learning_rate": 2.735993411213941e-05,
1261
+ "loss": 5.159,
1262
+ "step": 104500
1263
+ },
1264
+ {
1265
+ "epoch": 2.28,
1266
+ "learning_rate": 2.725178265675459e-05,
1267
+ "loss": 5.1327,
1268
+ "step": 105000
1269
+ },
1270
+ {
1271
+ "epoch": 2.29,
1272
+ "learning_rate": 2.7143414464986238e-05,
1273
+ "loss": 5.1263,
1274
+ "step": 105500
1275
+ },
1276
+ {
1277
+ "epoch": 2.3,
1278
+ "learning_rate": 2.7035046273217884e-05,
1279
+ "loss": 5.1287,
1280
+ "step": 106000
1281
+ },
1282
+ {
1283
+ "epoch": 2.31,
1284
+ "learning_rate": 2.6926678081449537e-05,
1285
+ "loss": 5.1209,
1286
+ "step": 106500
1287
+ },
1288
+ {
1289
+ "epoch": 2.32,
1290
+ "learning_rate": 2.681830988968118e-05,
1291
+ "loss": 5.1138,
1292
+ "step": 107000
1293
+ },
1294
+ {
1295
+ "epoch": 2.33,
1296
+ "learning_rate": 2.6709941697912833e-05,
1297
+ "loss": 5.1045,
1298
+ "step": 107500
1299
+ },
1300
+ {
1301
+ "epoch": 2.34,
1302
+ "learning_rate": 2.660157350614448e-05,
1303
+ "loss": 5.1258,
1304
+ "step": 108000
1305
+ },
1306
+ {
1307
+ "epoch": 2.35,
1308
+ "learning_rate": 2.6493205314376125e-05,
1309
+ "loss": 5.1379,
1310
+ "step": 108500
1311
+ },
1312
+ {
1313
+ "epoch": 2.36,
1314
+ "learning_rate": 2.6384837122607775e-05,
1315
+ "loss": 5.0773,
1316
+ "step": 109000
1317
+ },
1318
+ {
1319
+ "epoch": 2.37,
1320
+ "learning_rate": 2.6276685667222954e-05,
1321
+ "loss": 5.1031,
1322
+ "step": 109500
1323
+ },
1324
+ {
1325
+ "epoch": 2.38,
1326
+ "learning_rate": 2.6168317475454607e-05,
1327
+ "loss": 5.0969,
1328
+ "step": 110000
1329
+ },
1330
+ {
1331
+ "epoch": 2.39,
1332
+ "learning_rate": 2.6059949283686254e-05,
1333
+ "loss": 5.0868,
1334
+ "step": 110500
1335
+ },
1336
+ {
1337
+ "epoch": 2.41,
1338
+ "learning_rate": 2.5951581091917903e-05,
1339
+ "loss": 5.0737,
1340
+ "step": 111000
1341
+ },
1342
+ {
1343
+ "epoch": 2.42,
1344
+ "learning_rate": 2.5843429636533083e-05,
1345
+ "loss": 5.0793,
1346
+ "step": 111500
1347
+ },
1348
+ {
1349
+ "epoch": 2.43,
1350
+ "learning_rate": 2.5735061444764736e-05,
1351
+ "loss": 5.0712,
1352
+ "step": 112000
1353
+ },
1354
+ {
1355
+ "epoch": 2.44,
1356
+ "learning_rate": 2.5626693252996382e-05,
1357
+ "loss": 5.0871,
1358
+ "step": 112500
1359
+ },
1360
+ {
1361
+ "epoch": 2.45,
1362
+ "learning_rate": 2.5518325061228032e-05,
1363
+ "loss": 5.064,
1364
+ "step": 113000
1365
+ },
1366
+ {
1367
+ "epoch": 2.46,
1368
+ "learning_rate": 2.5409956869459678e-05,
1369
+ "loss": 5.0995,
1370
+ "step": 113500
1371
+ },
1372
+ {
1373
+ "epoch": 2.47,
1374
+ "learning_rate": 2.5301588677691324e-05,
1375
+ "loss": 5.0805,
1376
+ "step": 114000
1377
+ },
1378
+ {
1379
+ "epoch": 2.48,
1380
+ "learning_rate": 2.5193220485922974e-05,
1381
+ "loss": 5.0899,
1382
+ "step": 114500
1383
+ },
1384
+ {
1385
+ "epoch": 2.49,
1386
+ "learning_rate": 2.508485229415462e-05,
1387
+ "loss": 5.0422,
1388
+ "step": 115000
1389
+ },
1390
+ {
1391
+ "epoch": 2.5,
1392
+ "learning_rate": 2.4976700838769806e-05,
1393
+ "loss": 5.0442,
1394
+ "step": 115500
1395
+ },
1396
+ {
1397
+ "epoch": 2.51,
1398
+ "learning_rate": 2.4868332647001453e-05,
1399
+ "loss": 5.0505,
1400
+ "step": 116000
1401
+ },
1402
+ {
1403
+ "epoch": 2.52,
1404
+ "learning_rate": 2.4759964455233102e-05,
1405
+ "loss": 5.0718,
1406
+ "step": 116500
1407
+ },
1408
+ {
1409
+ "epoch": 2.54,
1410
+ "learning_rate": 2.4651812999848285e-05,
1411
+ "loss": 5.0682,
1412
+ "step": 117000
1413
+ },
1414
+ {
1415
+ "epoch": 2.55,
1416
+ "learning_rate": 2.454344480807993e-05,
1417
+ "loss": 5.0645,
1418
+ "step": 117500
1419
+ },
1420
+ {
1421
+ "epoch": 2.56,
1422
+ "learning_rate": 2.443507661631158e-05,
1423
+ "loss": 5.0679,
1424
+ "step": 118000
1425
+ },
1426
+ {
1427
+ "epoch": 2.57,
1428
+ "learning_rate": 2.432670842454323e-05,
1429
+ "loss": 5.0509,
1430
+ "step": 118500
1431
+ },
1432
+ {
1433
+ "epoch": 2.58,
1434
+ "learning_rate": 2.4218340232774877e-05,
1435
+ "loss": 5.0678,
1436
+ "step": 119000
1437
+ },
1438
+ {
1439
+ "epoch": 2.59,
1440
+ "learning_rate": 2.4109972041006526e-05,
1441
+ "loss": 5.0296,
1442
+ "step": 119500
1443
+ },
1444
+ {
1445
+ "epoch": 2.6,
1446
+ "learning_rate": 2.4001603849238173e-05,
1447
+ "loss": 5.0411,
1448
+ "step": 120000
1449
+ },
1450
+ {
1451
+ "epoch": 2.61,
1452
+ "learning_rate": 2.389323565746982e-05,
1453
+ "loss": 5.0054,
1454
+ "step": 120500
1455
+ },
1456
+ {
1457
+ "epoch": 2.62,
1458
+ "learning_rate": 2.378486746570147e-05,
1459
+ "loss": 5.0269,
1460
+ "step": 121000
1461
+ },
1462
+ {
1463
+ "epoch": 2.63,
1464
+ "learning_rate": 2.367671601031665e-05,
1465
+ "loss": 5.0299,
1466
+ "step": 121500
1467
+ },
1468
+ {
1469
+ "epoch": 2.64,
1470
+ "learning_rate": 2.35683478185483e-05,
1471
+ "loss": 5.0179,
1472
+ "step": 122000
1473
+ },
1474
+ {
1475
+ "epoch": 2.66,
1476
+ "learning_rate": 2.3460196363163487e-05,
1477
+ "loss": 5.0287,
1478
+ "step": 122500
1479
+ },
1480
+ {
1481
+ "epoch": 2.67,
1482
+ "learning_rate": 2.3351828171395134e-05,
1483
+ "loss": 5.0147,
1484
+ "step": 123000
1485
+ },
1486
+ {
1487
+ "epoch": 2.68,
1488
+ "learning_rate": 2.324345997962678e-05,
1489
+ "loss": 5.0182,
1490
+ "step": 123500
1491
+ },
1492
+ {
1493
+ "epoch": 2.69,
1494
+ "learning_rate": 2.313509178785843e-05,
1495
+ "loss": 4.9981,
1496
+ "step": 124000
1497
+ },
1498
+ {
1499
+ "epoch": 2.7,
1500
+ "learning_rate": 2.3026723596090076e-05,
1501
+ "loss": 4.9984,
1502
+ "step": 124500
1503
+ },
1504
+ {
1505
+ "epoch": 2.71,
1506
+ "learning_rate": 2.2918355404321725e-05,
1507
+ "loss": 4.9986,
1508
+ "step": 125000
1509
+ },
1510
+ {
1511
+ "epoch": 2.72,
1512
+ "learning_rate": 2.280998721255337e-05,
1513
+ "loss": 4.9822,
1514
+ "step": 125500
1515
+ },
1516
+ {
1517
+ "epoch": 2.73,
1518
+ "learning_rate": 2.270161902078502e-05,
1519
+ "loss": 4.9897,
1520
+ "step": 126000
1521
+ },
1522
+ {
1523
+ "epoch": 2.74,
1524
+ "learning_rate": 2.2593467565400208e-05,
1525
+ "loss": 5.0029,
1526
+ "step": 126500
1527
+ },
1528
+ {
1529
+ "epoch": 2.75,
1530
+ "learning_rate": 2.248531611001539e-05,
1531
+ "loss": 4.9913,
1532
+ "step": 127000
1533
+ },
1534
+ {
1535
+ "epoch": 2.76,
1536
+ "learning_rate": 2.2376947918247037e-05,
1537
+ "loss": 4.9867,
1538
+ "step": 127500
1539
+ },
1540
+ {
1541
+ "epoch": 2.77,
1542
+ "learning_rate": 2.2268579726478686e-05,
1543
+ "loss": 4.9702,
1544
+ "step": 128000
1545
+ },
1546
+ {
1547
+ "epoch": 2.79,
1548
+ "learning_rate": 2.2160211534710333e-05,
1549
+ "loss": 5.0042,
1550
+ "step": 128500
1551
+ },
1552
+ {
1553
+ "epoch": 2.8,
1554
+ "learning_rate": 2.205184334294198e-05,
1555
+ "loss": 4.9958,
1556
+ "step": 129000
1557
+ },
1558
+ {
1559
+ "epoch": 2.81,
1560
+ "learning_rate": 2.1943691887557165e-05,
1561
+ "loss": 4.969,
1562
+ "step": 129500
1563
+ },
1564
+ {
1565
+ "epoch": 2.82,
1566
+ "learning_rate": 2.183532369578881e-05,
1567
+ "loss": 4.9888,
1568
+ "step": 130000
1569
+ },
1570
+ {
1571
+ "epoch": 2.83,
1572
+ "learning_rate": 2.172695550402046e-05,
1573
+ "loss": 4.9643,
1574
+ "step": 130500
1575
+ },
1576
+ {
1577
+ "epoch": 2.84,
1578
+ "learning_rate": 2.1618804048635647e-05,
1579
+ "loss": 4.9661,
1580
+ "step": 131000
1581
+ },
1582
+ {
1583
+ "epoch": 2.85,
1584
+ "learning_rate": 2.1510435856867294e-05,
1585
+ "loss": 4.9511,
1586
+ "step": 131500
1587
+ },
1588
+ {
1589
+ "epoch": 2.86,
1590
+ "learning_rate": 2.140206766509894e-05,
1591
+ "loss": 4.9586,
1592
+ "step": 132000
1593
+ },
1594
+ {
1595
+ "epoch": 2.87,
1596
+ "learning_rate": 2.129369947333059e-05,
1597
+ "loss": 4.9791,
1598
+ "step": 132500
1599
+ },
1600
+ {
1601
+ "epoch": 2.88,
1602
+ "learning_rate": 2.118533128156224e-05,
1603
+ "loss": 4.9457,
1604
+ "step": 133000
1605
+ },
1606
+ {
1607
+ "epoch": 2.89,
1608
+ "learning_rate": 2.1076963089793885e-05,
1609
+ "loss": 4.9577,
1610
+ "step": 133500
1611
+ },
1612
+ {
1613
+ "epoch": 2.9,
1614
+ "learning_rate": 2.096859489802553e-05,
1615
+ "loss": 4.9646,
1616
+ "step": 134000
1617
+ },
1618
+ {
1619
+ "epoch": 2.92,
1620
+ "learning_rate": 2.086022670625718e-05,
1621
+ "loss": 4.9457,
1622
+ "step": 134500
1623
+ },
1624
+ {
1625
+ "epoch": 2.93,
1626
+ "learning_rate": 2.0751858514488827e-05,
1627
+ "loss": 4.9405,
1628
+ "step": 135000
1629
+ },
1630
+ {
1631
+ "epoch": 2.94,
1632
+ "learning_rate": 2.064370705910401e-05,
1633
+ "loss": 4.9576,
1634
+ "step": 135500
1635
+ },
1636
+ {
1637
+ "epoch": 2.95,
1638
+ "learning_rate": 2.053533886733566e-05,
1639
+ "loss": 4.9494,
1640
+ "step": 136000
1641
+ },
1642
+ {
1643
+ "epoch": 2.96,
1644
+ "learning_rate": 2.042697067556731e-05,
1645
+ "loss": 4.9348,
1646
+ "step": 136500
1647
+ },
1648
+ {
1649
+ "epoch": 2.97,
1650
+ "learning_rate": 2.0318819220182492e-05,
1651
+ "loss": 4.938,
1652
+ "step": 137000
1653
+ },
1654
+ {
1655
+ "epoch": 2.98,
1656
+ "learning_rate": 2.021045102841414e-05,
1657
+ "loss": 4.9657,
1658
+ "step": 137500
1659
+ },
1660
+ {
1661
+ "epoch": 2.99,
1662
+ "learning_rate": 2.0102082836645788e-05,
1663
+ "loss": 4.9181,
1664
+ "step": 138000
1665
+ },
1666
+ {
1667
+ "epoch": 3.0,
1668
+ "learning_rate": 1.9993714644877438e-05,
1669
+ "loss": 4.9687,
1670
+ "step": 138500
1671
+ },
1672
+ {
1673
+ "epoch": 3.01,
1674
+ "learning_rate": 1.9885346453109084e-05,
1675
+ "loss": 4.9063,
1676
+ "step": 139000
1677
+ },
1678
+ {
1679
+ "epoch": 3.02,
1680
+ "learning_rate": 1.977697826134073e-05,
1681
+ "loss": 4.9311,
1682
+ "step": 139500
1683
+ },
1684
+ {
1685
+ "epoch": 3.03,
1686
+ "learning_rate": 1.966861006957238e-05,
1687
+ "loss": 4.887,
1688
+ "step": 140000
1689
+ },
1690
+ {
1691
+ "epoch": 3.05,
1692
+ "learning_rate": 1.956024187780403e-05,
1693
+ "loss": 4.9098,
1694
+ "step": 140500
1695
+ },
1696
+ {
1697
+ "epoch": 3.06,
1698
+ "learning_rate": 1.9451873686035676e-05,
1699
+ "loss": 4.8869,
1700
+ "step": 141000
1701
+ },
1702
+ {
1703
+ "epoch": 3.07,
1704
+ "learning_rate": 1.9343505494267325e-05,
1705
+ "loss": 4.9259,
1706
+ "step": 141500
1707
+ },
1708
+ {
1709
+ "epoch": 3.08,
1710
+ "learning_rate": 1.9235137302498972e-05,
1711
+ "loss": 4.8914,
1712
+ "step": 142000
1713
+ },
1714
+ {
1715
+ "epoch": 3.09,
1716
+ "learning_rate": 1.9126769110730618e-05,
1717
+ "loss": 4.8805,
1718
+ "step": 142500
1719
+ },
1720
+ {
1721
+ "epoch": 3.1,
1722
+ "learning_rate": 1.9018617655345804e-05,
1723
+ "loss": 4.8959,
1724
+ "step": 143000
1725
+ },
1726
+ {
1727
+ "epoch": 3.11,
1728
+ "learning_rate": 1.891024946357745e-05,
1729
+ "loss": 4.9006,
1730
+ "step": 143500
1731
+ },
1732
+ {
1733
+ "epoch": 3.12,
1734
+ "learning_rate": 1.8802098008192637e-05,
1735
+ "loss": 4.8775,
1736
+ "step": 144000
1737
+ },
1738
+ {
1739
+ "epoch": 3.13,
1740
+ "learning_rate": 1.8693729816424286e-05,
1741
+ "loss": 4.8962,
1742
+ "step": 144500
1743
+ },
1744
+ {
1745
+ "epoch": 3.14,
1746
+ "learning_rate": 1.858536162465593e-05,
1747
+ "loss": 4.8837,
1748
+ "step": 145000
1749
+ },
1750
+ {
1751
+ "epoch": 3.15,
1752
+ "learning_rate": 1.847699343288758e-05,
1753
+ "loss": 4.8987,
1754
+ "step": 145500
1755
+ },
1756
+ {
1757
+ "epoch": 3.16,
1758
+ "learning_rate": 1.836862524111923e-05,
1759
+ "loss": 4.8875,
1760
+ "step": 146000
1761
+ },
1762
+ {
1763
+ "epoch": 3.18,
1764
+ "learning_rate": 1.8260257049350875e-05,
1765
+ "loss": 4.8839,
1766
+ "step": 146500
1767
+ },
1768
+ {
1769
+ "epoch": 3.19,
1770
+ "learning_rate": 1.815188885758252e-05,
1771
+ "loss": 4.8876,
1772
+ "step": 147000
1773
+ },
1774
+ {
1775
+ "epoch": 3.2,
1776
+ "learning_rate": 1.804352066581417e-05,
1777
+ "loss": 4.9015,
1778
+ "step": 147500
1779
+ },
1780
+ {
1781
+ "epoch": 3.21,
1782
+ "learning_rate": 1.793515247404582e-05,
1783
+ "loss": 4.8669,
1784
+ "step": 148000
1785
+ },
1786
+ {
1787
+ "epoch": 3.22,
1788
+ "learning_rate": 1.782721775504454e-05,
1789
+ "loss": 4.9037,
1790
+ "step": 148500
1791
+ },
1792
+ {
1793
+ "epoch": 3.23,
1794
+ "learning_rate": 1.771884956327619e-05,
1795
+ "loss": 4.8772,
1796
+ "step": 149000
1797
+ },
1798
+ {
1799
+ "epoch": 3.24,
1800
+ "learning_rate": 1.7610481371507836e-05,
1801
+ "loss": 4.8542,
1802
+ "step": 149500
1803
+ },
1804
+ {
1805
+ "epoch": 3.25,
1806
+ "learning_rate": 1.7502113179739485e-05,
1807
+ "loss": 4.8429,
1808
+ "step": 150000
1809
+ },
1810
+ {
1811
+ "epoch": 3.26,
1812
+ "learning_rate": 1.7393961724354668e-05,
1813
+ "loss": 4.8799,
1814
+ "step": 150500
1815
+ },
1816
+ {
1817
+ "epoch": 3.27,
1818
+ "learning_rate": 1.7285593532586318e-05,
1819
+ "loss": 4.8603,
1820
+ "step": 151000
1821
+ },
1822
+ {
1823
+ "epoch": 3.28,
1824
+ "learning_rate": 1.7177225340817964e-05,
1825
+ "loss": 4.8515,
1826
+ "step": 151500
1827
+ },
1828
+ {
1829
+ "epoch": 3.29,
1830
+ "learning_rate": 1.706885714904961e-05,
1831
+ "loss": 4.8639,
1832
+ "step": 152000
1833
+ },
1834
+ {
1835
+ "epoch": 3.31,
1836
+ "learning_rate": 1.696048895728126e-05,
1837
+ "loss": 4.8507,
1838
+ "step": 152500
1839
+ },
1840
+ {
1841
+ "epoch": 3.32,
1842
+ "learning_rate": 1.6852120765512906e-05,
1843
+ "loss": 4.8426,
1844
+ "step": 153000
1845
+ },
1846
+ {
1847
+ "epoch": 3.33,
1848
+ "learning_rate": 1.6743752573744556e-05,
1849
+ "loss": 4.8277,
1850
+ "step": 153500
1851
+ },
1852
+ {
1853
+ "epoch": 3.34,
1854
+ "learning_rate": 1.6635384381976205e-05,
1855
+ "loss": 4.8831,
1856
+ "step": 154000
1857
+ },
1858
+ {
1859
+ "epoch": 3.35,
1860
+ "learning_rate": 1.6527016190207852e-05,
1861
+ "loss": 4.8377,
1862
+ "step": 154500
1863
+ },
1864
+ {
1865
+ "epoch": 3.36,
1866
+ "learning_rate": 1.6418647998439498e-05,
1867
+ "loss": 4.8549,
1868
+ "step": 155000
1869
+ },
1870
+ {
1871
+ "epoch": 3.37,
1872
+ "learning_rate": 1.6310279806671148e-05,
1873
+ "loss": 4.8403,
1874
+ "step": 155500
1875
+ },
1876
+ {
1877
+ "epoch": 3.38,
1878
+ "learning_rate": 1.6201911614902797e-05,
1879
+ "loss": 4.8526,
1880
+ "step": 156000
1881
+ },
1882
+ {
1883
+ "epoch": 3.39,
1884
+ "learning_rate": 1.609354342313444e-05,
1885
+ "loss": 4.8472,
1886
+ "step": 156500
1887
+ },
1888
+ {
1889
+ "epoch": 3.4,
1890
+ "learning_rate": 1.5985391967749626e-05,
1891
+ "loss": 4.8356,
1892
+ "step": 157000
1893
+ },
1894
+ {
1895
+ "epoch": 3.41,
1896
+ "learning_rate": 1.5877023775981276e-05,
1897
+ "loss": 4.8552,
1898
+ "step": 157500
1899
+ },
1900
+ {
1901
+ "epoch": 3.42,
1902
+ "learning_rate": 1.5768655584212922e-05,
1903
+ "loss": 4.8528,
1904
+ "step": 158000
1905
+ },
1906
+ {
1907
+ "epoch": 3.44,
1908
+ "learning_rate": 1.566028739244457e-05,
1909
+ "loss": 4.8505,
1910
+ "step": 158500
1911
+ },
1912
+ {
1913
+ "epoch": 3.45,
1914
+ "learning_rate": 1.5552135937059755e-05,
1915
+ "loss": 4.8677,
1916
+ "step": 159000
1917
+ },
1918
+ {
1919
+ "epoch": 3.46,
1920
+ "learning_rate": 1.54437677452914e-05,
1921
+ "loss": 4.844,
1922
+ "step": 159500
1923
+ },
1924
+ {
1925
+ "epoch": 3.47,
1926
+ "learning_rate": 1.533539955352305e-05,
1927
+ "loss": 4.8299,
1928
+ "step": 160000
1929
+ },
1930
+ {
1931
+ "epoch": 3.48,
1932
+ "learning_rate": 1.5227248098138235e-05,
1933
+ "loss": 4.8197,
1934
+ "step": 160500
1935
+ },
1936
+ {
1937
+ "epoch": 3.49,
1938
+ "learning_rate": 1.5118879906369885e-05,
1939
+ "loss": 4.8289,
1940
+ "step": 161000
1941
+ },
1942
+ {
1943
+ "epoch": 3.5,
1944
+ "learning_rate": 1.501051171460153e-05,
1945
+ "loss": 4.8317,
1946
+ "step": 161500
1947
+ },
1948
+ {
1949
+ "epoch": 3.51,
1950
+ "learning_rate": 1.4902360259216716e-05,
1951
+ "loss": 4.8262,
1952
+ "step": 162000
1953
+ },
1954
+ {
1955
+ "epoch": 3.52,
1956
+ "learning_rate": 1.4793992067448365e-05,
1957
+ "loss": 4.8332,
1958
+ "step": 162500
1959
+ },
1960
+ {
1961
+ "epoch": 3.53,
1962
+ "learning_rate": 1.468562387568001e-05,
1963
+ "loss": 4.8332,
1964
+ "step": 163000
1965
+ },
1966
+ {
1967
+ "epoch": 3.54,
1968
+ "learning_rate": 1.457725568391166e-05,
1969
+ "loss": 4.839,
1970
+ "step": 163500
1971
+ },
1972
+ {
1973
+ "epoch": 3.55,
1974
+ "learning_rate": 1.4468887492143307e-05,
1975
+ "loss": 4.8062,
1976
+ "step": 164000
1977
+ },
1978
+ {
1979
+ "epoch": 3.57,
1980
+ "learning_rate": 1.4360519300374955e-05,
1981
+ "loss": 4.8388,
1982
+ "step": 164500
1983
+ },
1984
+ {
1985
+ "epoch": 3.58,
1986
+ "learning_rate": 1.4252151108606602e-05,
1987
+ "loss": 4.7874,
1988
+ "step": 165000
1989
+ },
1990
+ {
1991
+ "epoch": 3.59,
1992
+ "learning_rate": 1.414378291683825e-05,
1993
+ "loss": 4.8314,
1994
+ "step": 165500
1995
+ },
1996
+ {
1997
+ "epoch": 3.6,
1998
+ "learning_rate": 1.4035414725069897e-05,
1999
+ "loss": 4.8281,
2000
+ "step": 166000
2001
+ },
2002
+ {
2003
+ "epoch": 3.61,
2004
+ "learning_rate": 1.3927046533301547e-05,
2005
+ "loss": 4.805,
2006
+ "step": 166500
2007
+ },
2008
+ {
2009
+ "epoch": 3.62,
2010
+ "learning_rate": 1.3818678341533195e-05,
2011
+ "loss": 4.8146,
2012
+ "step": 167000
2013
+ },
2014
+ {
2015
+ "epoch": 3.63,
2016
+ "learning_rate": 1.3710310149764841e-05,
2017
+ "loss": 4.8093,
2018
+ "step": 167500
2019
+ },
2020
+ {
2021
+ "epoch": 3.64,
2022
+ "learning_rate": 1.360194195799649e-05,
2023
+ "loss": 4.8045,
2024
+ "step": 168000
2025
+ },
2026
+ {
2027
+ "epoch": 3.65,
2028
+ "learning_rate": 1.3493790502611675e-05,
2029
+ "loss": 4.812,
2030
+ "step": 168500
2031
+ },
2032
+ {
2033
+ "epoch": 3.66,
2034
+ "learning_rate": 1.338542231084332e-05,
2035
+ "loss": 4.8153,
2036
+ "step": 169000
2037
+ },
2038
+ {
2039
+ "epoch": 3.67,
2040
+ "learning_rate": 1.327705411907497e-05,
2041
+ "loss": 4.7973,
2042
+ "step": 169500
2043
+ },
2044
+ {
2045
+ "epoch": 3.68,
2046
+ "learning_rate": 1.3168685927306618e-05,
2047
+ "loss": 4.7848,
2048
+ "step": 170000
2049
+ },
2050
+ {
2051
+ "epoch": 3.7,
2052
+ "learning_rate": 1.3060317735538265e-05,
2053
+ "loss": 4.7788,
2054
+ "step": 170500
2055
+ },
2056
+ {
2057
+ "epoch": 3.71,
2058
+ "learning_rate": 1.2951949543769915e-05,
2059
+ "loss": 4.7904,
2060
+ "step": 171000
2061
+ },
2062
+ {
2063
+ "epoch": 3.72,
2064
+ "learning_rate": 1.284358135200156e-05,
2065
+ "loss": 4.8071,
2066
+ "step": 171500
2067
+ },
2068
+ {
2069
+ "epoch": 3.73,
2070
+ "learning_rate": 1.273521316023321e-05,
2071
+ "loss": 4.7705,
2072
+ "step": 172000
2073
+ },
2074
+ {
2075
+ "epoch": 3.74,
2076
+ "learning_rate": 1.2627061704848394e-05,
2077
+ "loss": 4.7761,
2078
+ "step": 172500
2079
+ },
2080
+ {
2081
+ "epoch": 3.75,
2082
+ "learning_rate": 1.251869351308004e-05,
2083
+ "loss": 4.7745,
2084
+ "step": 173000
2085
+ },
2086
+ {
2087
+ "epoch": 3.76,
2088
+ "learning_rate": 1.2410325321311688e-05,
2089
+ "loss": 4.8037,
2090
+ "step": 173500
2091
+ },
2092
+ {
2093
+ "epoch": 3.77,
2094
+ "learning_rate": 1.2301957129543338e-05,
2095
+ "loss": 4.7899,
2096
+ "step": 174000
2097
+ },
2098
+ {
2099
+ "epoch": 3.78,
2100
+ "learning_rate": 1.2193805674158522e-05,
2101
+ "loss": 4.7707,
2102
+ "step": 174500
2103
+ },
2104
+ {
2105
+ "epoch": 3.79,
2106
+ "learning_rate": 1.2085437482390169e-05,
2107
+ "loss": 4.7628,
2108
+ "step": 175000
2109
+ },
2110
+ {
2111
+ "epoch": 3.8,
2112
+ "learning_rate": 1.1977069290621818e-05,
2113
+ "loss": 4.8077,
2114
+ "step": 175500
2115
+ },
2116
+ {
2117
+ "epoch": 3.81,
2118
+ "learning_rate": 1.1868701098853464e-05,
2119
+ "loss": 4.7716,
2120
+ "step": 176000
2121
+ },
2122
+ {
2123
+ "epoch": 3.83,
2124
+ "learning_rate": 1.1760549643468649e-05,
2125
+ "loss": 4.7809,
2126
+ "step": 176500
2127
+ },
2128
+ {
2129
+ "epoch": 3.84,
2130
+ "learning_rate": 1.1652181451700297e-05,
2131
+ "loss": 4.7824,
2132
+ "step": 177000
2133
+ },
2134
+ {
2135
+ "epoch": 3.85,
2136
+ "learning_rate": 1.1543813259931947e-05,
2137
+ "loss": 4.7673,
2138
+ "step": 177500
2139
+ },
2140
+ {
2141
+ "epoch": 3.86,
2142
+ "learning_rate": 1.1435445068163593e-05,
2143
+ "loss": 4.7763,
2144
+ "step": 178000
2145
+ },
2146
+ {
2147
+ "epoch": 3.87,
2148
+ "learning_rate": 1.132707687639524e-05,
2149
+ "loss": 4.7798,
2150
+ "step": 178500
2151
+ },
2152
+ {
2153
+ "epoch": 3.88,
2154
+ "learning_rate": 1.1218708684626889e-05,
2155
+ "loss": 4.7793,
2156
+ "step": 179000
2157
+ },
2158
+ {
2159
+ "epoch": 3.89,
2160
+ "learning_rate": 1.1110557229242073e-05,
2161
+ "loss": 4.7544,
2162
+ "step": 179500
2163
+ },
2164
+ {
2165
+ "epoch": 3.9,
2166
+ "learning_rate": 1.1002189037473721e-05,
2167
+ "loss": 4.7661,
2168
+ "step": 180000
2169
+ },
2170
+ {
2171
+ "epoch": 3.91,
2172
+ "learning_rate": 1.0893820845705369e-05,
2173
+ "loss": 4.7634,
2174
+ "step": 180500
2175
+ },
2176
+ {
2177
+ "epoch": 3.92,
2178
+ "learning_rate": 1.0785452653937017e-05,
2179
+ "loss": 4.784,
2180
+ "step": 181000
2181
+ },
2182
+ {
2183
+ "epoch": 3.93,
2184
+ "learning_rate": 1.0677084462168665e-05,
2185
+ "loss": 4.7577,
2186
+ "step": 181500
2187
+ },
2188
+ {
2189
+ "epoch": 3.94,
2190
+ "learning_rate": 1.0568716270400313e-05,
2191
+ "loss": 4.7507,
2192
+ "step": 182000
2193
+ },
2194
+ {
2195
+ "epoch": 3.96,
2196
+ "learning_rate": 1.0460348078631961e-05,
2197
+ "loss": 4.7365,
2198
+ "step": 182500
2199
+ },
2200
+ {
2201
+ "epoch": 3.97,
2202
+ "learning_rate": 1.0351979886863609e-05,
2203
+ "loss": 4.7678,
2204
+ "step": 183000
2205
+ },
2206
+ {
2207
+ "epoch": 3.98,
2208
+ "learning_rate": 1.0243828431478793e-05,
2209
+ "loss": 4.7701,
2210
+ "step": 183500
2211
+ },
2212
+ {
2213
+ "epoch": 3.99,
2214
+ "learning_rate": 1.0135460239710441e-05,
2215
+ "loss": 4.7522,
2216
+ "step": 184000
2217
+ },
2218
+ {
2219
+ "epoch": 4.0,
2220
+ "learning_rate": 1.0027092047942088e-05,
2221
+ "loss": 4.7454,
2222
+ "step": 184500
2223
+ },
2224
+ {
2225
+ "epoch": 4.01,
2226
+ "learning_rate": 9.918723856173737e-06,
2227
+ "loss": 4.7236,
2228
+ "step": 185000
2229
+ },
2230
+ {
2231
+ "epoch": 4.02,
2232
+ "learning_rate": 9.810355664405383e-06,
2233
+ "loss": 4.7409,
2234
+ "step": 185500
2235
+ },
2236
+ {
2237
+ "epoch": 4.03,
2238
+ "learning_rate": 9.701987472637031e-06,
2239
+ "loss": 4.74,
2240
+ "step": 186000
2241
+ },
2242
+ {
2243
+ "epoch": 4.04,
2244
+ "learning_rate": 9.593836017252216e-06,
2245
+ "loss": 4.7351,
2246
+ "step": 186500
2247
+ },
2248
+ {
2249
+ "epoch": 4.05,
2250
+ "learning_rate": 9.485467825483864e-06,
2251
+ "loss": 4.7128,
2252
+ "step": 187000
2253
+ },
2254
+ {
2255
+ "epoch": 4.06,
2256
+ "learning_rate": 9.377316370099049e-06,
2257
+ "loss": 4.7234,
2258
+ "step": 187500
2259
+ },
2260
+ {
2261
+ "epoch": 4.07,
2262
+ "learning_rate": 9.268948178330696e-06,
2263
+ "loss": 4.7467,
2264
+ "step": 188000
2265
+ },
2266
+ {
2267
+ "epoch": 4.09,
2268
+ "learning_rate": 9.160579986562344e-06,
2269
+ "loss": 4.7249,
2270
+ "step": 188500
2271
+ },
2272
+ {
2273
+ "epoch": 4.1,
2274
+ "learning_rate": 9.052211794793992e-06,
2275
+ "loss": 4.7212,
2276
+ "step": 189000
2277
+ },
2278
+ {
2279
+ "epoch": 4.11,
2280
+ "learning_rate": 8.94384360302564e-06,
2281
+ "loss": 4.7368,
2282
+ "step": 189500
2283
+ },
2284
+ {
2285
+ "epoch": 4.12,
2286
+ "learning_rate": 8.835475411257288e-06,
2287
+ "loss": 4.7173,
2288
+ "step": 190000
2289
+ },
2290
+ {
2291
+ "epoch": 4.13,
2292
+ "learning_rate": 8.727107219488936e-06,
2293
+ "loss": 4.748,
2294
+ "step": 190500
2295
+ },
2296
+ {
2297
+ "epoch": 4.14,
2298
+ "learning_rate": 8.618739027720584e-06,
2299
+ "loss": 4.7198,
2300
+ "step": 191000
2301
+ },
2302
+ {
2303
+ "epoch": 4.15,
2304
+ "learning_rate": 8.510370835952232e-06,
2305
+ "loss": 4.7166,
2306
+ "step": 191500
2307
+ },
2308
+ {
2309
+ "epoch": 4.16,
2310
+ "learning_rate": 8.402219380567417e-06,
2311
+ "loss": 4.7247,
2312
+ "step": 192000
2313
+ },
2314
+ {
2315
+ "epoch": 4.17,
2316
+ "learning_rate": 8.293851188799063e-06,
2317
+ "loss": 4.7295,
2318
+ "step": 192500
2319
+ },
2320
+ {
2321
+ "epoch": 4.18,
2322
+ "learning_rate": 8.185482997030712e-06,
2323
+ "loss": 4.7058,
2324
+ "step": 193000
2325
+ },
2326
+ {
2327
+ "epoch": 4.19,
2328
+ "learning_rate": 8.077114805262359e-06,
2329
+ "loss": 4.6972,
2330
+ "step": 193500
2331
+ },
2332
+ {
2333
+ "epoch": 4.2,
2334
+ "learning_rate": 7.968963349877543e-06,
2335
+ "loss": 4.7311,
2336
+ "step": 194000
2337
+ },
2338
+ {
2339
+ "epoch": 4.22,
2340
+ "learning_rate": 7.860595158109193e-06,
2341
+ "loss": 4.7356,
2342
+ "step": 194500
2343
+ },
2344
+ {
2345
+ "epoch": 4.23,
2346
+ "learning_rate": 7.752443702724378e-06,
2347
+ "loss": 4.7023,
2348
+ "step": 195000
2349
+ },
2350
+ {
2351
+ "epoch": 4.24,
2352
+ "learning_rate": 7.644292247339562e-06,
2353
+ "loss": 4.7178,
2354
+ "step": 195500
2355
+ },
2356
+ {
2357
+ "epoch": 4.25,
2358
+ "learning_rate": 7.535924055571208e-06,
2359
+ "loss": 4.7288,
2360
+ "step": 196000
2361
+ },
2362
+ {
2363
+ "epoch": 4.26,
2364
+ "learning_rate": 7.427555863802857e-06,
2365
+ "loss": 4.735,
2366
+ "step": 196500
2367
+ },
2368
+ {
2369
+ "epoch": 4.27,
2370
+ "learning_rate": 7.319187672034505e-06,
2371
+ "loss": 4.742,
2372
+ "step": 197000
2373
+ },
2374
+ {
2375
+ "epoch": 4.28,
2376
+ "learning_rate": 7.210819480266152e-06,
2377
+ "loss": 4.7199,
2378
+ "step": 197500
2379
+ },
2380
+ {
2381
+ "epoch": 4.29,
2382
+ "learning_rate": 7.102451288497801e-06,
2383
+ "loss": 4.7203,
2384
+ "step": 198000
2385
+ },
2386
+ {
2387
+ "epoch": 4.3,
2388
+ "learning_rate": 6.994083096729448e-06,
2389
+ "loss": 4.6962,
2390
+ "step": 198500
2391
+ },
2392
+ {
2393
+ "epoch": 4.31,
2394
+ "learning_rate": 6.885714904961097e-06,
2395
+ "loss": 4.7219,
2396
+ "step": 199000
2397
+ },
2398
+ {
2399
+ "epoch": 4.32,
2400
+ "learning_rate": 6.777563449576281e-06,
2401
+ "loss": 4.705,
2402
+ "step": 199500
2403
+ },
2404
+ {
2405
+ "epoch": 4.33,
2406
+ "learning_rate": 6.6691952578079285e-06,
2407
+ "loss": 4.7085,
2408
+ "step": 200000
2409
+ },
2410
+ {
2411
+ "epoch": 4.35,
2412
+ "learning_rate": 6.560827066039576e-06,
2413
+ "loss": 4.7082,
2414
+ "step": 200500
2415
+ },
2416
+ {
2417
+ "epoch": 4.36,
2418
+ "learning_rate": 6.4524588742712235e-06,
2419
+ "loss": 4.705,
2420
+ "step": 201000
2421
+ },
2422
+ {
2423
+ "epoch": 4.37,
2424
+ "learning_rate": 6.344090682502872e-06,
2425
+ "loss": 4.6962,
2426
+ "step": 201500
2427
+ },
2428
+ {
2429
+ "epoch": 4.38,
2430
+ "learning_rate": 6.235722490734519e-06,
2431
+ "loss": 4.706,
2432
+ "step": 202000
2433
+ },
2434
+ {
2435
+ "epoch": 4.39,
2436
+ "learning_rate": 6.127354298966167e-06,
2437
+ "loss": 4.7202,
2438
+ "step": 202500
2439
+ },
2440
+ {
2441
+ "epoch": 4.4,
2442
+ "learning_rate": 6.018986107197815e-06,
2443
+ "loss": 4.7005,
2444
+ "step": 203000
2445
+ },
2446
+ {
2447
+ "epoch": 4.41,
2448
+ "learning_rate": 5.910617915429464e-06,
2449
+ "loss": 4.701,
2450
+ "step": 203500
2451
+ },
2452
+ {
2453
+ "epoch": 4.42,
2454
+ "learning_rate": 5.802466460044648e-06,
2455
+ "loss": 4.6878,
2456
+ "step": 204000
2457
+ },
2458
+ {
2459
+ "epoch": 4.43,
2460
+ "learning_rate": 5.694098268276296e-06,
2461
+ "loss": 4.7178,
2462
+ "step": 204500
2463
+ },
2464
+ {
2465
+ "epoch": 4.44,
2466
+ "learning_rate": 5.585730076507944e-06,
2467
+ "loss": 4.6609,
2468
+ "step": 205000
2469
+ },
2470
+ {
2471
+ "epoch": 4.45,
2472
+ "learning_rate": 5.4773618847395916e-06,
2473
+ "loss": 4.7039,
2474
+ "step": 205500
2475
+ },
2476
+ {
2477
+ "epoch": 4.46,
2478
+ "learning_rate": 5.369210429354776e-06,
2479
+ "loss": 4.7098,
2480
+ "step": 206000
2481
+ },
2482
+ {
2483
+ "epoch": 4.48,
2484
+ "learning_rate": 5.260842237586424e-06,
2485
+ "loss": 4.7026,
2486
+ "step": 206500
2487
+ },
2488
+ {
2489
+ "epoch": 4.49,
2490
+ "learning_rate": 5.152474045818072e-06,
2491
+ "loss": 4.6955,
2492
+ "step": 207000
2493
+ },
2494
+ {
2495
+ "epoch": 4.5,
2496
+ "learning_rate": 5.044105854049719e-06,
2497
+ "loss": 4.7102,
2498
+ "step": 207500
2499
+ },
2500
+ {
2501
+ "epoch": 4.51,
2502
+ "learning_rate": 4.935737662281367e-06,
2503
+ "loss": 4.6803,
2504
+ "step": 208000
2505
+ },
2506
+ {
2507
+ "epoch": 4.52,
2508
+ "learning_rate": 4.827586206896552e-06,
2509
+ "loss": 4.7065,
2510
+ "step": 208500
2511
+ },
2512
+ {
2513
+ "epoch": 4.53,
2514
+ "learning_rate": 4.7192180151282e-06,
2515
+ "loss": 4.7135,
2516
+ "step": 209000
2517
+ },
2518
+ {
2519
+ "epoch": 4.54,
2520
+ "learning_rate": 4.6108498233598475e-06,
2521
+ "loss": 4.7195,
2522
+ "step": 209500
2523
+ },
2524
+ {
2525
+ "epoch": 4.55,
2526
+ "learning_rate": 4.5024816315914954e-06,
2527
+ "loss": 4.7074,
2528
+ "step": 210000
2529
+ },
2530
+ {
2531
+ "epoch": 4.56,
2532
+ "learning_rate": 4.39433017620668e-06,
2533
+ "loss": 4.6901,
2534
+ "step": 210500
2535
+ },
2536
+ {
2537
+ "epoch": 4.57,
2538
+ "learning_rate": 4.286178720821865e-06,
2539
+ "loss": 4.7073,
2540
+ "step": 211000
2541
+ },
2542
+ {
2543
+ "epoch": 4.58,
2544
+ "learning_rate": 4.177810529053512e-06,
2545
+ "loss": 4.7132,
2546
+ "step": 211500
2547
+ },
2548
+ {
2549
+ "epoch": 4.59,
2550
+ "learning_rate": 4.06944233728516e-06,
2551
+ "loss": 4.6883,
2552
+ "step": 212000
2553
+ },
2554
+ {
2555
+ "epoch": 4.61,
2556
+ "learning_rate": 3.9610741455168085e-06,
2557
+ "loss": 4.6894,
2558
+ "step": 212500
2559
+ },
2560
+ {
2561
+ "epoch": 4.62,
2562
+ "learning_rate": 3.852705953748456e-06,
2563
+ "loss": 4.6735,
2564
+ "step": 213000
2565
+ },
2566
+ {
2567
+ "epoch": 4.63,
2568
+ "learning_rate": 3.744337761980104e-06,
2569
+ "loss": 4.6991,
2570
+ "step": 213500
2571
+ },
2572
+ {
2573
+ "epoch": 4.64,
2574
+ "learning_rate": 3.6359695702117514e-06,
2575
+ "loss": 4.6894,
2576
+ "step": 214000
2577
+ },
2578
+ {
2579
+ "epoch": 4.65,
2580
+ "learning_rate": 3.5276013784433993e-06,
2581
+ "loss": 4.6913,
2582
+ "step": 214500
2583
+ },
2584
+ {
2585
+ "epoch": 4.66,
2586
+ "learning_rate": 3.4192331866750473e-06,
2587
+ "loss": 4.6822,
2588
+ "step": 215000
2589
+ },
2590
+ {
2591
+ "epoch": 4.67,
2592
+ "learning_rate": 3.3108649949066948e-06,
2593
+ "loss": 4.7032,
2594
+ "step": 215500
2595
+ },
2596
+ {
2597
+ "epoch": 4.68,
2598
+ "learning_rate": 3.2027135395218794e-06,
2599
+ "loss": 4.6878,
2600
+ "step": 216000
2601
+ },
2602
+ {
2603
+ "epoch": 4.69,
2604
+ "learning_rate": 3.0943453477535277e-06,
2605
+ "loss": 4.6907,
2606
+ "step": 216500
2607
+ },
2608
+ {
2609
+ "epoch": 4.7,
2610
+ "learning_rate": 2.9859771559851752e-06,
2611
+ "loss": 4.7031,
2612
+ "step": 217000
2613
+ },
2614
+ {
2615
+ "epoch": 4.71,
2616
+ "learning_rate": 2.877608964216823e-06,
2617
+ "loss": 4.6638,
2618
+ "step": 217500
2619
+ },
2620
+ {
2621
+ "epoch": 4.72,
2622
+ "learning_rate": 2.769240772448471e-06,
2623
+ "loss": 4.6622,
2624
+ "step": 218000
2625
+ },
2626
+ {
2627
+ "epoch": 4.74,
2628
+ "learning_rate": 2.6610893170636553e-06,
2629
+ "loss": 4.6783,
2630
+ "step": 218500
2631
+ },
2632
+ {
2633
+ "epoch": 4.75,
2634
+ "learning_rate": 2.5527211252953036e-06,
2635
+ "loss": 4.7027,
2636
+ "step": 219000
2637
+ },
2638
+ {
2639
+ "epoch": 4.76,
2640
+ "learning_rate": 2.444352933526951e-06,
2641
+ "loss": 4.6681,
2642
+ "step": 219500
2643
+ },
2644
+ {
2645
+ "epoch": 4.77,
2646
+ "learning_rate": 2.335984741758599e-06,
2647
+ "loss": 4.6985,
2648
+ "step": 220000
2649
+ },
2650
+ {
2651
+ "epoch": 4.78,
2652
+ "learning_rate": 2.2278332863737837e-06,
2653
+ "loss": 4.693,
2654
+ "step": 220500
2655
+ },
2656
+ {
2657
+ "epoch": 4.79,
2658
+ "learning_rate": 2.1196818309889683e-06,
2659
+ "loss": 4.6522,
2660
+ "step": 221000
2661
+ },
2662
+ {
2663
+ "epoch": 4.8,
2664
+ "learning_rate": 2.0113136392206162e-06,
2665
+ "loss": 4.6671,
2666
+ "step": 221500
2667
+ },
2668
+ {
2669
+ "epoch": 4.81,
2670
+ "learning_rate": 1.902945447452264e-06,
2671
+ "loss": 4.6676,
2672
+ "step": 222000
2673
+ },
2674
+ {
2675
+ "epoch": 4.82,
2676
+ "learning_rate": 1.7945772556839117e-06,
2677
+ "loss": 4.6786,
2678
+ "step": 222500
2679
+ },
2680
+ {
2681
+ "epoch": 4.83,
2682
+ "learning_rate": 1.6862090639155596e-06,
2683
+ "loss": 4.673,
2684
+ "step": 223000
2685
+ },
2686
+ {
2687
+ "epoch": 4.84,
2688
+ "learning_rate": 1.5778408721472075e-06,
2689
+ "loss": 4.6652,
2690
+ "step": 223500
2691
+ },
2692
+ {
2693
+ "epoch": 4.85,
2694
+ "learning_rate": 1.4694726803788553e-06,
2695
+ "loss": 4.666,
2696
+ "step": 224000
2697
+ },
2698
+ {
2699
+ "epoch": 4.87,
2700
+ "learning_rate": 1.3613212249940399e-06,
2701
+ "loss": 4.7056,
2702
+ "step": 224500
2703
+ },
2704
+ {
2705
+ "epoch": 4.88,
2706
+ "learning_rate": 1.2529530332256876e-06,
2707
+ "loss": 4.6693,
2708
+ "step": 225000
2709
+ },
2710
+ {
2711
+ "epoch": 4.89,
2712
+ "learning_rate": 1.1445848414573355e-06,
2713
+ "loss": 4.6729,
2714
+ "step": 225500
2715
+ },
2716
+ {
2717
+ "epoch": 4.9,
2718
+ "learning_rate": 1.0362166496889832e-06,
2719
+ "loss": 4.6815,
2720
+ "step": 226000
2721
+ },
2722
+ {
2723
+ "epoch": 4.91,
2724
+ "learning_rate": 9.278484579206312e-07,
2725
+ "loss": 4.6881,
2726
+ "step": 226500
2727
+ },
2728
+ {
2729
+ "epoch": 4.92,
2730
+ "learning_rate": 8.194802661522789e-07,
2731
+ "loss": 4.6657,
2732
+ "step": 227000
2733
+ },
2734
+ {
2735
+ "epoch": 4.93,
2736
+ "learning_rate": 7.111120743839268e-07,
2737
+ "loss": 4.6779,
2738
+ "step": 227500
2739
+ },
2740
+ {
2741
+ "epoch": 4.94,
2742
+ "learning_rate": 6.027438826155747e-07,
2743
+ "loss": 4.6812,
2744
+ "step": 228000
2745
+ },
2746
+ {
2747
+ "epoch": 4.95,
2748
+ "learning_rate": 4.943756908472226e-07,
2749
+ "loss": 4.6778,
2750
+ "step": 228500
2751
+ },
2752
+ {
2753
+ "epoch": 4.96,
2754
+ "learning_rate": 3.862242354624071e-07,
2755
+ "loss": 4.6756,
2756
+ "step": 229000
2757
+ },
2758
+ {
2759
+ "epoch": 4.97,
2760
+ "learning_rate": 2.778560436940549e-07,
2761
+ "loss": 4.673,
2762
+ "step": 229500
2763
+ },
2764
+ {
2765
+ "epoch": 4.98,
2766
+ "learning_rate": 1.6948785192570278e-07,
2767
+ "loss": 4.6745,
2768
+ "step": 230000
2769
+ },
2770
+ {
2771
+ "epoch": 5.0,
2772
+ "learning_rate": 6.111966015735062e-08,
2773
+ "loss": 4.6569,
2774
+ "step": 230500
2775
+ },
2776
+ {
2777
+ "epoch": 5.0,
2778
+ "step": 230695,
2779
+ "total_flos": 4.869967130385408e+17,
2780
+ "train_loss": 5.372212270862655,
2781
+ "train_runtime": 34970.0124,
2782
+ "train_samples_per_second": 52.775,
2783
+ "train_steps_per_second": 6.597
2784
+ }
2785
+ ],
2786
+ "max_steps": 230695,
2787
+ "num_train_epochs": 5,
2788
+ "total_flos": 4.869967130385408e+17,
2789
+ "trial_name": null,
2790
+ "trial_params": null
2791
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:69daec732a7e1a919f92fe02edf47c56cb077c892a10f6eea6ae656cd0e15379
3
+ size 3963