Hugging Face's logo

Davlan
/
afro-xlmr-large-75L

Fill-Mask
Transformers
PyTorch
Safetensors
xlm-roberta
Generated from Trainer
Inference Endpoints
Model card Files Community
1
Davlan commited on
Commit
ea35c81
•
1 Parent(s): 6e178e1

Upload 10 files

Browse files
Files changed (11) hide show
  1. .gitattributes +1 -0
  2. all_results.json +15 -0
  3. config.json +28 -0
  4. eval_results.json +10 -0
  5. sentencepiece.bpe.model +3 -0
  6. special_tokens_map.json +15 -0
  7. tokenizer.json +3 -0
  8. tokenizer_config.json +19 -0
  9. train_results.json +8 -0
  10. trainer_state.json +4993 -0
  11. training_args.bin +3 -0
.gitattributes CHANGED
@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
+ tokenizer.json filter=lfs diff=lfs merge=lfs -text
all_results.json ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 3.0,
3
+ "eval_accuracy": 0.7995252983185633,
4
+ "eval_loss": 0.9682623147964478,
5
+ "eval_runtime": 32.6923,
6
+ "eval_samples": 13416,
7
+ "eval_samples_per_second": 410.371,
8
+ "eval_steps_per_second": 8.565,
9
+ "perplexity": 2.6333645226958677,
10
+ "train_loss": 0.06289897245738281,
11
+ "train_runtime": 14736.1316,
12
+ "train_samples": 13116428,
13
+ "train_samples_per_second": 2670.259,
14
+ "train_steps_per_second": 27.815
15
+ }
config.json ADDED
@@ -0,0 +1,28 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "xlm-roberta-large",
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": 1024,
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 4096,
15
+ "layer_norm_eps": 1e-05,
16
+ "max_position_embeddings": 514,
17
+ "model_type": "xlm-roberta",
18
+ "num_attention_heads": 16,
19
+ "num_hidden_layers": 24,
20
+ "output_past": true,
21
+ "pad_token_id": 1,
22
+ "position_embedding_type": "absolute",
23
+ "torch_dtype": "float32",
24
+ "transformers_version": "4.32.0.dev0",
25
+ "type_vocab_size": 1,
26
+ "use_cache": true,
27
+ "vocab_size": 250002
28
+ }
eval_results.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 3.0,
3
+ "eval_accuracy": 0.7995252983185633,
4
+ "eval_loss": 0.9682623147964478,
5
+ "eval_runtime": 32.6923,
6
+ "eval_samples": 13416,
7
+ "eval_samples_per_second": 410.371,
8
+ "eval_steps_per_second": 8.565,
9
+ "perplexity": 2.6333645226958677
10
+ }
sentencepiece.bpe.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cfc8146abe2a0488e9e2a0c56de7952f7c11ab059eca145a0a727afce0db2865
3
+ size 5069051
special_tokens_map.json ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<s>",
3
+ "cls_token": "<s>",
4
+ "eos_token": "</s>",
5
+ "mask_token": {
6
+ "content": "<mask>",
7
+ "lstrip": true,
8
+ "normalized": false,
9
+ "rstrip": false,
10
+ "single_word": false
11
+ },
12
+ "pad_token": "<pad>",
13
+ "sep_token": "</s>",
14
+ "unk_token": "<unk>"
15
+ }
tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:62c24cdc13d4c9952d63718d6c9fa4c287974249e16b7ade6d5a85e7bbb75626
3
+ size 17082660
tokenizer_config.json ADDED
@@ -0,0 +1,19 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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": 512,
15
+ "pad_token": "<pad>",
16
+ "sep_token": "</s>",
17
+ "tokenizer_class": "XLMRobertaTokenizer",
18
+ "unk_token": "<unk>"
19
+ }
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 3.0,
3
+ "train_loss": 0.06289897245738281,
4
+ "train_runtime": 14736.1316,
5
+ "train_samples": 13116428,
6
+ "train_samples_per_second": 2670.259,
7
+ "train_steps_per_second": 27.815
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,4993 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 3.0,
5
+ "global_step": 409890,
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": 5e-05,
13
+ "loss": 2.0235,
14
+ "step": 500
15
+ },
16
+ {
17
+ "epoch": 0.01,
18
+ "learning_rate": 5e-05,
19
+ "loss": 1.871,
20
+ "step": 1000
21
+ },
22
+ {
23
+ "epoch": 0.01,
24
+ "learning_rate": 5e-05,
25
+ "loss": 1.7723,
26
+ "step": 1500
27
+ },
28
+ {
29
+ "epoch": 0.01,
30
+ "learning_rate": 5e-05,
31
+ "loss": 1.7105,
32
+ "step": 2000
33
+ },
34
+ {
35
+ "epoch": 0.02,
36
+ "learning_rate": 5e-05,
37
+ "loss": 1.6615,
38
+ "step": 2500
39
+ },
40
+ {
41
+ "epoch": 0.02,
42
+ "learning_rate": 5e-05,
43
+ "loss": 1.6266,
44
+ "step": 3000
45
+ },
46
+ {
47
+ "epoch": 0.03,
48
+ "learning_rate": 5e-05,
49
+ "loss": 1.5964,
50
+ "step": 3500
51
+ },
52
+ {
53
+ "epoch": 0.03,
54
+ "learning_rate": 5e-05,
55
+ "loss": 1.5695,
56
+ "step": 4000
57
+ },
58
+ {
59
+ "epoch": 0.03,
60
+ "learning_rate": 5e-05,
61
+ "loss": 1.5454,
62
+ "step": 4500
63
+ },
64
+ {
65
+ "epoch": 0.04,
66
+ "learning_rate": 5e-05,
67
+ "loss": 1.529,
68
+ "step": 5000
69
+ },
70
+ {
71
+ "epoch": 0.04,
72
+ "learning_rate": 5e-05,
73
+ "loss": 1.505,
74
+ "step": 5500
75
+ },
76
+ {
77
+ "epoch": 0.04,
78
+ "learning_rate": 5e-05,
79
+ "loss": 1.4823,
80
+ "step": 6000
81
+ },
82
+ {
83
+ "epoch": 0.05,
84
+ "learning_rate": 5e-05,
85
+ "loss": 1.463,
86
+ "step": 6500
87
+ },
88
+ {
89
+ "epoch": 0.05,
90
+ "learning_rate": 5e-05,
91
+ "loss": 1.4563,
92
+ "step": 7000
93
+ },
94
+ {
95
+ "epoch": 0.05,
96
+ "learning_rate": 5e-05,
97
+ "loss": 1.4428,
98
+ "step": 7500
99
+ },
100
+ {
101
+ "epoch": 0.06,
102
+ "learning_rate": 5e-05,
103
+ "loss": 1.4384,
104
+ "step": 8000
105
+ },
106
+ {
107
+ "epoch": 0.06,
108
+ "learning_rate": 5e-05,
109
+ "loss": 1.4163,
110
+ "step": 8500
111
+ },
112
+ {
113
+ "epoch": 0.07,
114
+ "learning_rate": 5e-05,
115
+ "loss": 1.4174,
116
+ "step": 9000
117
+ },
118
+ {
119
+ "epoch": 0.07,
120
+ "learning_rate": 5e-05,
121
+ "loss": 1.407,
122
+ "step": 9500
123
+ },
124
+ {
125
+ "epoch": 0.07,
126
+ "learning_rate": 5e-05,
127
+ "loss": 1.3931,
128
+ "step": 10000
129
+ },
130
+ {
131
+ "epoch": 0.08,
132
+ "learning_rate": 5e-05,
133
+ "loss": 1.3871,
134
+ "step": 10500
135
+ },
136
+ {
137
+ "epoch": 0.08,
138
+ "learning_rate": 5e-05,
139
+ "loss": 1.3877,
140
+ "step": 11000
141
+ },
142
+ {
143
+ "epoch": 0.08,
144
+ "learning_rate": 5e-05,
145
+ "loss": 1.3757,
146
+ "step": 11500
147
+ },
148
+ {
149
+ "epoch": 0.09,
150
+ "learning_rate": 5e-05,
151
+ "loss": 1.3676,
152
+ "step": 12000
153
+ },
154
+ {
155
+ "epoch": 0.09,
156
+ "learning_rate": 5e-05,
157
+ "loss": 1.3567,
158
+ "step": 12500
159
+ },
160
+ {
161
+ "epoch": 0.1,
162
+ "learning_rate": 5e-05,
163
+ "loss": 1.3547,
164
+ "step": 13000
165
+ },
166
+ {
167
+ "epoch": 0.1,
168
+ "learning_rate": 5e-05,
169
+ "loss": 1.3508,
170
+ "step": 13500
171
+ },
172
+ {
173
+ "epoch": 0.1,
174
+ "learning_rate": 5e-05,
175
+ "loss": 1.3469,
176
+ "step": 14000
177
+ },
178
+ {
179
+ "epoch": 0.11,
180
+ "learning_rate": 5e-05,
181
+ "loss": 1.3389,
182
+ "step": 14500
183
+ },
184
+ {
185
+ "epoch": 0.11,
186
+ "learning_rate": 5e-05,
187
+ "loss": 1.3319,
188
+ "step": 15000
189
+ },
190
+ {
191
+ "epoch": 0.11,
192
+ "learning_rate": 5e-05,
193
+ "loss": 1.3231,
194
+ "step": 15500
195
+ },
196
+ {
197
+ "epoch": 0.12,
198
+ "learning_rate": 5e-05,
199
+ "loss": 1.325,
200
+ "step": 16000
201
+ },
202
+ {
203
+ "epoch": 0.12,
204
+ "learning_rate": 5e-05,
205
+ "loss": 1.3141,
206
+ "step": 16500
207
+ },
208
+ {
209
+ "epoch": 0.12,
210
+ "learning_rate": 5e-05,
211
+ "loss": 1.3211,
212
+ "step": 17000
213
+ },
214
+ {
215
+ "epoch": 0.13,
216
+ "learning_rate": 5e-05,
217
+ "loss": 1.3068,
218
+ "step": 17500
219
+ },
220
+ {
221
+ "epoch": 0.13,
222
+ "learning_rate": 5e-05,
223
+ "loss": 1.3025,
224
+ "step": 18000
225
+ },
226
+ {
227
+ "epoch": 0.14,
228
+ "learning_rate": 5e-05,
229
+ "loss": 1.2992,
230
+ "step": 18500
231
+ },
232
+ {
233
+ "epoch": 0.14,
234
+ "learning_rate": 5e-05,
235
+ "loss": 1.3018,
236
+ "step": 19000
237
+ },
238
+ {
239
+ "epoch": 0.14,
240
+ "learning_rate": 5e-05,
241
+ "loss": 1.2995,
242
+ "step": 19500
243
+ },
244
+ {
245
+ "epoch": 0.15,
246
+ "learning_rate": 5e-05,
247
+ "loss": 1.2837,
248
+ "step": 20000
249
+ },
250
+ {
251
+ "epoch": 0.15,
252
+ "learning_rate": 5e-05,
253
+ "loss": 1.282,
254
+ "step": 20500
255
+ },
256
+ {
257
+ "epoch": 0.15,
258
+ "learning_rate": 5e-05,
259
+ "loss": 1.2885,
260
+ "step": 21000
261
+ },
262
+ {
263
+ "epoch": 0.16,
264
+ "learning_rate": 5e-05,
265
+ "loss": 1.2807,
266
+ "step": 21500
267
+ },
268
+ {
269
+ "epoch": 0.16,
270
+ "learning_rate": 5e-05,
271
+ "loss": 1.2813,
272
+ "step": 22000
273
+ },
274
+ {
275
+ "epoch": 0.16,
276
+ "learning_rate": 5e-05,
277
+ "loss": 1.2784,
278
+ "step": 22500
279
+ },
280
+ {
281
+ "epoch": 0.17,
282
+ "learning_rate": 5e-05,
283
+ "loss": 1.2701,
284
+ "step": 23000
285
+ },
286
+ {
287
+ "epoch": 0.17,
288
+ "learning_rate": 5e-05,
289
+ "loss": 1.2683,
290
+ "step": 23500
291
+ },
292
+ {
293
+ "epoch": 0.18,
294
+ "learning_rate": 5e-05,
295
+ "loss": 1.2728,
296
+ "step": 24000
297
+ },
298
+ {
299
+ "epoch": 0.18,
300
+ "learning_rate": 5e-05,
301
+ "loss": 1.2631,
302
+ "step": 24500
303
+ },
304
+ {
305
+ "epoch": 0.18,
306
+ "learning_rate": 5e-05,
307
+ "loss": 1.2591,
308
+ "step": 25000
309
+ },
310
+ {
311
+ "epoch": 0.19,
312
+ "learning_rate": 5e-05,
313
+ "loss": 1.2545,
314
+ "step": 25500
315
+ },
316
+ {
317
+ "epoch": 0.19,
318
+ "learning_rate": 5e-05,
319
+ "loss": 1.2521,
320
+ "step": 26000
321
+ },
322
+ {
323
+ "epoch": 0.19,
324
+ "learning_rate": 5e-05,
325
+ "loss": 1.2515,
326
+ "step": 26500
327
+ },
328
+ {
329
+ "epoch": 0.2,
330
+ "learning_rate": 5e-05,
331
+ "loss": 1.2482,
332
+ "step": 27000
333
+ },
334
+ {
335
+ "epoch": 0.2,
336
+ "learning_rate": 5e-05,
337
+ "loss": 1.2478,
338
+ "step": 27500
339
+ },
340
+ {
341
+ "epoch": 0.2,
342
+ "learning_rate": 5e-05,
343
+ "loss": 1.2418,
344
+ "step": 28000
345
+ },
346
+ {
347
+ "epoch": 0.21,
348
+ "learning_rate": 5e-05,
349
+ "loss": 1.2429,
350
+ "step": 28500
351
+ },
352
+ {
353
+ "epoch": 0.21,
354
+ "learning_rate": 5e-05,
355
+ "loss": 1.2419,
356
+ "step": 29000
357
+ },
358
+ {
359
+ "epoch": 0.22,
360
+ "learning_rate": 5e-05,
361
+ "loss": 1.2349,
362
+ "step": 29500
363
+ },
364
+ {
365
+ "epoch": 0.22,
366
+ "learning_rate": 5e-05,
367
+ "loss": 1.2397,
368
+ "step": 30000
369
+ },
370
+ {
371
+ "epoch": 0.22,
372
+ "learning_rate": 5e-05,
373
+ "loss": 1.2308,
374
+ "step": 30500
375
+ },
376
+ {
377
+ "epoch": 0.23,
378
+ "learning_rate": 5e-05,
379
+ "loss": 1.2292,
380
+ "step": 31000
381
+ },
382
+ {
383
+ "epoch": 0.23,
384
+ "learning_rate": 5e-05,
385
+ "loss": 1.2251,
386
+ "step": 31500
387
+ },
388
+ {
389
+ "epoch": 0.23,
390
+ "learning_rate": 5e-05,
391
+ "loss": 1.2331,
392
+ "step": 32000
393
+ },
394
+ {
395
+ "epoch": 0.24,
396
+ "learning_rate": 5e-05,
397
+ "loss": 1.2259,
398
+ "step": 32500
399
+ },
400
+ {
401
+ "epoch": 0.24,
402
+ "learning_rate": 5e-05,
403
+ "loss": 1.2273,
404
+ "step": 33000
405
+ },
406
+ {
407
+ "epoch": 0.25,
408
+ "learning_rate": 5e-05,
409
+ "loss": 1.2195,
410
+ "step": 33500
411
+ },
412
+ {
413
+ "epoch": 0.25,
414
+ "learning_rate": 5e-05,
415
+ "loss": 1.2194,
416
+ "step": 34000
417
+ },
418
+ {
419
+ "epoch": 0.25,
420
+ "learning_rate": 5e-05,
421
+ "loss": 1.2178,
422
+ "step": 34500
423
+ },
424
+ {
425
+ "epoch": 0.26,
426
+ "learning_rate": 5e-05,
427
+ "loss": 1.2169,
428
+ "step": 35000
429
+ },
430
+ {
431
+ "epoch": 0.26,
432
+ "learning_rate": 5e-05,
433
+ "loss": 1.2171,
434
+ "step": 35500
435
+ },
436
+ {
437
+ "epoch": 0.26,
438
+ "learning_rate": 5e-05,
439
+ "loss": 1.2041,
440
+ "step": 36000
441
+ },
442
+ {
443
+ "epoch": 0.27,
444
+ "learning_rate": 5e-05,
445
+ "loss": 1.2183,
446
+ "step": 36500
447
+ },
448
+ {
449
+ "epoch": 0.27,
450
+ "learning_rate": 5e-05,
451
+ "loss": 1.2112,
452
+ "step": 37000
453
+ },
454
+ {
455
+ "epoch": 0.27,
456
+ "learning_rate": 5e-05,
457
+ "loss": 1.206,
458
+ "step": 37500
459
+ },
460
+ {
461
+ "epoch": 0.28,
462
+ "learning_rate": 5e-05,
463
+ "loss": 1.1992,
464
+ "step": 38000
465
+ },
466
+ {
467
+ "epoch": 0.28,
468
+ "learning_rate": 5e-05,
469
+ "loss": 1.2065,
470
+ "step": 38500
471
+ },
472
+ {
473
+ "epoch": 0.29,
474
+ "learning_rate": 5e-05,
475
+ "loss": 1.2097,
476
+ "step": 39000
477
+ },
478
+ {
479
+ "epoch": 0.29,
480
+ "learning_rate": 5e-05,
481
+ "loss": 1.2004,
482
+ "step": 39500
483
+ },
484
+ {
485
+ "epoch": 0.29,
486
+ "learning_rate": 5e-05,
487
+ "loss": 1.2041,
488
+ "step": 40000
489
+ },
490
+ {
491
+ "epoch": 0.3,
492
+ "learning_rate": 5e-05,
493
+ "loss": 1.2011,
494
+ "step": 40500
495
+ },
496
+ {
497
+ "epoch": 0.3,
498
+ "learning_rate": 5e-05,
499
+ "loss": 1.2008,
500
+ "step": 41000
501
+ },
502
+ {
503
+ "epoch": 0.3,
504
+ "learning_rate": 5e-05,
505
+ "loss": 1.2017,
506
+ "step": 41500
507
+ },
508
+ {
509
+ "epoch": 0.31,
510
+ "learning_rate": 5e-05,
511
+ "loss": 1.2022,
512
+ "step": 42000
513
+ },
514
+ {
515
+ "epoch": 0.31,
516
+ "learning_rate": 5e-05,
517
+ "loss": 1.1948,
518
+ "step": 42500
519
+ },
520
+ {
521
+ "epoch": 0.31,
522
+ "learning_rate": 5e-05,
523
+ "loss": 1.197,
524
+ "step": 43000
525
+ },
526
+ {
527
+ "epoch": 0.32,
528
+ "learning_rate": 5e-05,
529
+ "loss": 1.1886,
530
+ "step": 43500
531
+ },
532
+ {
533
+ "epoch": 0.32,
534
+ "learning_rate": 5e-05,
535
+ "loss": 1.1906,
536
+ "step": 44000
537
+ },
538
+ {
539
+ "epoch": 0.33,
540
+ "learning_rate": 5e-05,
541
+ "loss": 1.1881,
542
+ "step": 44500
543
+ },
544
+ {
545
+ "epoch": 0.33,
546
+ "learning_rate": 5e-05,
547
+ "loss": 1.1908,
548
+ "step": 45000
549
+ },
550
+ {
551
+ "epoch": 0.33,
552
+ "learning_rate": 5e-05,
553
+ "loss": 1.1871,
554
+ "step": 45500
555
+ },
556
+ {
557
+ "epoch": 0.34,
558
+ "learning_rate": 5e-05,
559
+ "loss": 1.1857,
560
+ "step": 46000
561
+ },
562
+ {
563
+ "epoch": 0.34,
564
+ "learning_rate": 5e-05,
565
+ "loss": 1.1821,
566
+ "step": 46500
567
+ },
568
+ {
569
+ "epoch": 0.34,
570
+ "learning_rate": 5e-05,
571
+ "loss": 1.1818,
572
+ "step": 47000
573
+ },
574
+ {
575
+ "epoch": 0.35,
576
+ "learning_rate": 5e-05,
577
+ "loss": 1.1781,
578
+ "step": 47500
579
+ },
580
+ {
581
+ "epoch": 0.35,
582
+ "learning_rate": 5e-05,
583
+ "loss": 1.1797,
584
+ "step": 48000
585
+ },
586
+ {
587
+ "epoch": 0.35,
588
+ "learning_rate": 5e-05,
589
+ "loss": 1.1739,
590
+ "step": 48500
591
+ },
592
+ {
593
+ "epoch": 0.36,
594
+ "learning_rate": 5e-05,
595
+ "loss": 1.1817,
596
+ "step": 49000
597
+ },
598
+ {
599
+ "epoch": 0.36,
600
+ "learning_rate": 5e-05,
601
+ "loss": 1.1794,
602
+ "step": 49500
603
+ },
604
+ {
605
+ "epoch": 0.37,
606
+ "learning_rate": 5e-05,
607
+ "loss": 1.1757,
608
+ "step": 50000
609
+ },
610
+ {
611
+ "epoch": 0.37,
612
+ "learning_rate": 5e-05,
613
+ "loss": 1.1701,
614
+ "step": 50500
615
+ },
616
+ {
617
+ "epoch": 0.37,
618
+ "learning_rate": 5e-05,
619
+ "loss": 1.1687,
620
+ "step": 51000
621
+ },
622
+ {
623
+ "epoch": 0.38,
624
+ "learning_rate": 5e-05,
625
+ "loss": 1.1744,
626
+ "step": 51500
627
+ },
628
+ {
629
+ "epoch": 0.38,
630
+ "learning_rate": 5e-05,
631
+ "loss": 1.1692,
632
+ "step": 52000
633
+ },
634
+ {
635
+ "epoch": 0.38,
636
+ "learning_rate": 5e-05,
637
+ "loss": 1.1636,
638
+ "step": 52500
639
+ },
640
+ {
641
+ "epoch": 0.39,
642
+ "learning_rate": 5e-05,
643
+ "loss": 1.1683,
644
+ "step": 53000
645
+ },
646
+ {
647
+ "epoch": 0.39,
648
+ "learning_rate": 5e-05,
649
+ "loss": 1.1601,
650
+ "step": 53500
651
+ },
652
+ {
653
+ "epoch": 0.4,
654
+ "learning_rate": 5e-05,
655
+ "loss": 1.1664,
656
+ "step": 54000
657
+ },
658
+ {
659
+ "epoch": 0.4,
660
+ "learning_rate": 5e-05,
661
+ "loss": 1.1636,
662
+ "step": 54500
663
+ },
664
+ {
665
+ "epoch": 0.4,
666
+ "learning_rate": 5e-05,
667
+ "loss": 1.1635,
668
+ "step": 55000
669
+ },
670
+ {
671
+ "epoch": 0.41,
672
+ "learning_rate": 5e-05,
673
+ "loss": 1.1616,
674
+ "step": 55500
675
+ },
676
+ {
677
+ "epoch": 0.41,
678
+ "learning_rate": 5e-05,
679
+ "loss": 1.1593,
680
+ "step": 56000
681
+ },
682
+ {
683
+ "epoch": 0.41,
684
+ "learning_rate": 5e-05,
685
+ "loss": 1.1598,
686
+ "step": 56500
687
+ },
688
+ {
689
+ "epoch": 0.42,
690
+ "learning_rate": 5e-05,
691
+ "loss": 1.1582,
692
+ "step": 57000
693
+ },
694
+ {
695
+ "epoch": 0.42,
696
+ "learning_rate": 5e-05,
697
+ "loss": 1.1611,
698
+ "step": 57500
699
+ },
700
+ {
701
+ "epoch": 0.42,
702
+ "learning_rate": 5e-05,
703
+ "loss": 1.1571,
704
+ "step": 58000
705
+ },
706
+ {
707
+ "epoch": 0.43,
708
+ "learning_rate": 5e-05,
709
+ "loss": 1.1539,
710
+ "step": 58500
711
+ },
712
+ {
713
+ "epoch": 0.43,
714
+ "learning_rate": 5e-05,
715
+ "loss": 1.1506,
716
+ "step": 59000
717
+ },
718
+ {
719
+ "epoch": 0.44,
720
+ "learning_rate": 5e-05,
721
+ "loss": 1.154,
722
+ "step": 59500
723
+ },
724
+ {
725
+ "epoch": 0.44,
726
+ "learning_rate": 5e-05,
727
+ "loss": 1.1522,
728
+ "step": 60000
729
+ },
730
+ {
731
+ "epoch": 0.44,
732
+ "learning_rate": 5e-05,
733
+ "loss": 1.1549,
734
+ "step": 60500
735
+ },
736
+ {
737
+ "epoch": 0.45,
738
+ "learning_rate": 5e-05,
739
+ "loss": 1.1501,
740
+ "step": 61000
741
+ },
742
+ {
743
+ "epoch": 0.45,
744
+ "learning_rate": 5e-05,
745
+ "loss": 1.1459,
746
+ "step": 61500
747
+ },
748
+ {
749
+ "epoch": 0.45,
750
+ "learning_rate": 5e-05,
751
+ "loss": 1.1544,
752
+ "step": 62000
753
+ },
754
+ {
755
+ "epoch": 0.46,
756
+ "learning_rate": 5e-05,
757
+ "loss": 1.1529,
758
+ "step": 62500
759
+ },
760
+ {
761
+ "epoch": 0.46,
762
+ "learning_rate": 5e-05,
763
+ "loss": 1.1539,
764
+ "step": 63000
765
+ },
766
+ {
767
+ "epoch": 0.46,
768
+ "learning_rate": 5e-05,
769
+ "loss": 1.151,
770
+ "step": 63500
771
+ },
772
+ {
773
+ "epoch": 0.47,
774
+ "learning_rate": 5e-05,
775
+ "loss": 1.1491,
776
+ "step": 64000
777
+ },
778
+ {
779
+ "epoch": 0.47,
780
+ "eval_accuracy": 0.7652064452221591,
781
+ "eval_loss": 1.1646534204483032,
782
+ "eval_runtime": 32.5535,
783
+ "eval_samples_per_second": 412.121,
784
+ "eval_steps_per_second": 8.601,
785
+ "step": 64000
786
+ },
787
+ {
788
+ "epoch": 0.47,
789
+ "learning_rate": 5e-05,
790
+ "loss": 1.1535,
791
+ "step": 64500
792
+ },
793
+ {
794
+ "epoch": 0.48,
795
+ "learning_rate": 5e-05,
796
+ "loss": 1.1521,
797
+ "step": 65000
798
+ },
799
+ {
800
+ "epoch": 0.48,
801
+ "learning_rate": 5e-05,
802
+ "loss": 1.1486,
803
+ "step": 65500
804
+ },
805
+ {
806
+ "epoch": 0.48,
807
+ "learning_rate": 5e-05,
808
+ "loss": 1.1498,
809
+ "step": 66000
810
+ },
811
+ {
812
+ "epoch": 0.49,
813
+ "learning_rate": 5e-05,
814
+ "loss": 1.1513,
815
+ "step": 66500
816
+ },
817
+ {
818
+ "epoch": 0.49,
819
+ "learning_rate": 5e-05,
820
+ "loss": 1.144,
821
+ "step": 67000
822
+ },
823
+ {
824
+ "epoch": 0.49,
825
+ "learning_rate": 5e-05,
826
+ "loss": 1.1417,
827
+ "step": 67500
828
+ },
829
+ {
830
+ "epoch": 0.5,
831
+ "learning_rate": 5e-05,
832
+ "loss": 1.1414,
833
+ "step": 68000
834
+ },
835
+ {
836
+ "epoch": 0.5,
837
+ "learning_rate": 5e-05,
838
+ "loss": 1.1416,
839
+ "step": 68500
840
+ },
841
+ {
842
+ "epoch": 0.51,
843
+ "learning_rate": 5e-05,
844
+ "loss": 1.1459,
845
+ "step": 69000
846
+ },
847
+ {
848
+ "epoch": 0.51,
849
+ "learning_rate": 5e-05,
850
+ "loss": 1.1476,
851
+ "step": 69500
852
+ },
853
+ {
854
+ "epoch": 0.51,
855
+ "learning_rate": 5e-05,
856
+ "loss": 1.1444,
857
+ "step": 70000
858
+ },
859
+ {
860
+ "epoch": 0.52,
861
+ "learning_rate": 5e-05,
862
+ "loss": 1.1402,
863
+ "step": 70500
864
+ },
865
+ {
866
+ "epoch": 0.52,
867
+ "learning_rate": 5e-05,
868
+ "loss": 1.1439,
869
+ "step": 71000
870
+ },
871
+ {
872
+ "epoch": 0.52,
873
+ "learning_rate": 5e-05,
874
+ "loss": 1.1459,
875
+ "step": 71500
876
+ },
877
+ {
878
+ "epoch": 0.53,
879
+ "learning_rate": 5e-05,
880
+ "loss": 1.1338,
881
+ "step": 72000
882
+ },
883
+ {
884
+ "epoch": 0.53,
885
+ "learning_rate": 5e-05,
886
+ "loss": 1.137,
887
+ "step": 72500
888
+ },
889
+ {
890
+ "epoch": 0.53,
891
+ "learning_rate": 5e-05,
892
+ "loss": 1.1375,
893
+ "step": 73000
894
+ },
895
+ {
896
+ "epoch": 0.54,
897
+ "learning_rate": 5e-05,
898
+ "loss": 1.1322,
899
+ "step": 73500
900
+ },
901
+ {
902
+ "epoch": 0.54,
903
+ "learning_rate": 5e-05,
904
+ "loss": 1.1387,
905
+ "step": 74000
906
+ },
907
+ {
908
+ "epoch": 0.55,
909
+ "learning_rate": 5e-05,
910
+ "loss": 1.1421,
911
+ "step": 74500
912
+ },
913
+ {
914
+ "epoch": 0.55,
915
+ "learning_rate": 5e-05,
916
+ "loss": 1.1318,
917
+ "step": 75000
918
+ },
919
+ {
920
+ "epoch": 0.55,
921
+ "learning_rate": 5e-05,
922
+ "loss": 1.1247,
923
+ "step": 75500
924
+ },
925
+ {
926
+ "epoch": 0.56,
927
+ "learning_rate": 5e-05,
928
+ "loss": 1.1313,
929
+ "step": 76000
930
+ },
931
+ {
932
+ "epoch": 0.56,
933
+ "learning_rate": 5e-05,
934
+ "loss": 1.1323,
935
+ "step": 76500
936
+ },
937
+ {
938
+ "epoch": 0.56,
939
+ "learning_rate": 5e-05,
940
+ "loss": 1.1308,
941
+ "step": 77000
942
+ },
943
+ {
944
+ "epoch": 0.57,
945
+ "learning_rate": 5e-05,
946
+ "loss": 1.1294,
947
+ "step": 77500
948
+ },
949
+ {
950
+ "epoch": 0.57,
951
+ "learning_rate": 5e-05,
952
+ "loss": 1.1265,
953
+ "step": 78000
954
+ },
955
+ {
956
+ "epoch": 0.57,
957
+ "learning_rate": 5e-05,
958
+ "loss": 1.1334,
959
+ "step": 78500
960
+ },
961
+ {
962
+ "epoch": 0.58,
963
+ "learning_rate": 5e-05,
964
+ "loss": 1.1273,
965
+ "step": 79000
966
+ },
967
+ {
968
+ "epoch": 0.58,
969
+ "learning_rate": 5e-05,
970
+ "loss": 1.133,
971
+ "step": 79500
972
+ },
973
+ {
974
+ "epoch": 0.59,
975
+ "learning_rate": 5e-05,
976
+ "loss": 1.1316,
977
+ "step": 80000
978
+ },
979
+ {
980
+ "epoch": 0.59,
981
+ "learning_rate": 5e-05,
982
+ "loss": 1.1279,
983
+ "step": 80500
984
+ },
985
+ {
986
+ "epoch": 0.59,
987
+ "learning_rate": 5e-05,
988
+ "loss": 1.1339,
989
+ "step": 81000
990
+ },
991
+ {
992
+ "epoch": 0.6,
993
+ "learning_rate": 5e-05,
994
+ "loss": 1.1336,
995
+ "step": 81500
996
+ },
997
+ {
998
+ "epoch": 0.6,
999
+ "learning_rate": 5e-05,
1000
+ "loss": 1.135,
1001
+ "step": 82000
1002
+ },
1003
+ {
1004
+ "epoch": 0.6,
1005
+ "learning_rate": 5e-05,
1006
+ "loss": 1.1345,
1007
+ "step": 82500
1008
+ },
1009
+ {
1010
+ "epoch": 0.61,
1011
+ "learning_rate": 5e-05,
1012
+ "loss": 1.1287,
1013
+ "step": 83000
1014
+ },
1015
+ {
1016
+ "epoch": 0.61,
1017
+ "learning_rate": 5e-05,
1018
+ "loss": 1.1237,
1019
+ "step": 83500
1020
+ },
1021
+ {
1022
+ "epoch": 0.61,
1023
+ "learning_rate": 5e-05,
1024
+ "loss": 1.1314,
1025
+ "step": 84000
1026
+ },
1027
+ {
1028
+ "epoch": 0.62,
1029
+ "learning_rate": 5e-05,
1030
+ "loss": 1.1262,
1031
+ "step": 84500
1032
+ },
1033
+ {
1034
+ "epoch": 0.62,
1035
+ "learning_rate": 5e-05,
1036
+ "loss": 1.1251,
1037
+ "step": 85000
1038
+ },
1039
+ {
1040
+ "epoch": 0.63,
1041
+ "learning_rate": 5e-05,
1042
+ "loss": 1.1226,
1043
+ "step": 85500
1044
+ },
1045
+ {
1046
+ "epoch": 0.63,
1047
+ "learning_rate": 5e-05,
1048
+ "loss": 1.1208,
1049
+ "step": 86000
1050
+ },
1051
+ {
1052
+ "epoch": 0.63,
1053
+ "learning_rate": 5e-05,
1054
+ "loss": 1.1155,
1055
+ "step": 86500
1056
+ },
1057
+ {
1058
+ "epoch": 0.64,
1059
+ "learning_rate": 5e-05,
1060
+ "loss": 1.1196,
1061
+ "step": 87000
1062
+ },
1063
+ {
1064
+ "epoch": 0.64,
1065
+ "learning_rate": 5e-05,
1066
+ "loss": 1.1153,
1067
+ "step": 87500
1068
+ },
1069
+ {
1070
+ "epoch": 0.64,
1071
+ "learning_rate": 5e-05,
1072
+ "loss": 1.1168,
1073
+ "step": 88000
1074
+ },
1075
+ {
1076
+ "epoch": 0.65,
1077
+ "learning_rate": 5e-05,
1078
+ "loss": 1.1131,
1079
+ "step": 88500
1080
+ },
1081
+ {
1082
+ "epoch": 0.65,
1083
+ "learning_rate": 5e-05,
1084
+ "loss": 1.1106,
1085
+ "step": 89000
1086
+ },
1087
+ {
1088
+ "epoch": 0.66,
1089
+ "learning_rate": 5e-05,
1090
+ "loss": 1.1238,
1091
+ "step": 89500
1092
+ },
1093
+ {
1094
+ "epoch": 0.66,
1095
+ "learning_rate": 5e-05,
1096
+ "loss": 1.1121,
1097
+ "step": 90000
1098
+ },
1099
+ {
1100
+ "epoch": 0.66,
1101
+ "learning_rate": 5e-05,
1102
+ "loss": 1.108,
1103
+ "step": 90500
1104
+ },
1105
+ {
1106
+ "epoch": 0.67,
1107
+ "learning_rate": 5e-05,
1108
+ "loss": 1.1046,
1109
+ "step": 91000
1110
+ },
1111
+ {
1112
+ "epoch": 0.67,
1113
+ "learning_rate": 5e-05,
1114
+ "loss": 1.115,
1115
+ "step": 91500
1116
+ },
1117
+ {
1118
+ "epoch": 0.67,
1119
+ "learning_rate": 5e-05,
1120
+ "loss": 1.1154,
1121
+ "step": 92000
1122
+ },
1123
+ {
1124
+ "epoch": 0.68,
1125
+ "learning_rate": 5e-05,
1126
+ "loss": 1.1146,
1127
+ "step": 92500
1128
+ },
1129
+ {
1130
+ "epoch": 0.68,
1131
+ "learning_rate": 5e-05,
1132
+ "loss": 1.117,
1133
+ "step": 93000
1134
+ },
1135
+ {
1136
+ "epoch": 0.68,
1137
+ "learning_rate": 5e-05,
1138
+ "loss": 1.1117,
1139
+ "step": 93500
1140
+ },
1141
+ {
1142
+ "epoch": 0.69,
1143
+ "learning_rate": 5e-05,
1144
+ "loss": 1.1052,
1145
+ "step": 94000
1146
+ },
1147
+ {
1148
+ "epoch": 0.69,
1149
+ "learning_rate": 5e-05,
1150
+ "loss": 1.1126,
1151
+ "step": 94500
1152
+ },
1153
+ {
1154
+ "epoch": 0.7,
1155
+ "learning_rate": 5e-05,
1156
+ "loss": 1.1102,
1157
+ "step": 95000
1158
+ },
1159
+ {
1160
+ "epoch": 0.7,
1161
+ "learning_rate": 5e-05,
1162
+ "loss": 1.1076,
1163
+ "step": 95500
1164
+ },
1165
+ {
1166
+ "epoch": 0.7,
1167
+ "learning_rate": 5e-05,
1168
+ "loss": 1.1124,
1169
+ "step": 96000
1170
+ },
1171
+ {
1172
+ "epoch": 0.71,
1173
+ "learning_rate": 5e-05,
1174
+ "loss": 1.1104,
1175
+ "step": 96500
1176
+ },
1177
+ {
1178
+ "epoch": 0.71,
1179
+ "learning_rate": 5e-05,
1180
+ "loss": 1.115,
1181
+ "step": 97000
1182
+ },
1183
+ {
1184
+ "epoch": 0.71,
1185
+ "learning_rate": 5e-05,
1186
+ "loss": 1.1071,
1187
+ "step": 97500
1188
+ },
1189
+ {
1190
+ "epoch": 0.72,
1191
+ "learning_rate": 5e-05,
1192
+ "loss": 1.1108,
1193
+ "step": 98000
1194
+ },
1195
+ {
1196
+ "epoch": 0.72,
1197
+ "learning_rate": 5e-05,
1198
+ "loss": 1.1079,
1199
+ "step": 98500
1200
+ },
1201
+ {
1202
+ "epoch": 0.72,
1203
+ "learning_rate": 5e-05,
1204
+ "loss": 1.1019,
1205
+ "step": 99000
1206
+ },
1207
+ {
1208
+ "epoch": 0.73,
1209
+ "learning_rate": 5e-05,
1210
+ "loss": 1.103,
1211
+ "step": 99500
1212
+ },
1213
+ {
1214
+ "epoch": 0.73,
1215
+ "learning_rate": 5e-05,
1216
+ "loss": 1.1,
1217
+ "step": 100000
1218
+ },
1219
+ {
1220
+ "epoch": 0.74,
1221
+ "learning_rate": 5e-05,
1222
+ "loss": 1.1006,
1223
+ "step": 100500
1224
+ },
1225
+ {
1226
+ "epoch": 0.74,
1227
+ "learning_rate": 5e-05,
1228
+ "loss": 1.103,
1229
+ "step": 101000
1230
+ },
1231
+ {
1232
+ "epoch": 0.74,
1233
+ "learning_rate": 5e-05,
1234
+ "loss": 1.1005,
1235
+ "step": 101500
1236
+ },
1237
+ {
1238
+ "epoch": 0.75,
1239
+ "learning_rate": 5e-05,
1240
+ "loss": 1.1016,
1241
+ "step": 102000
1242
+ },
1243
+ {
1244
+ "epoch": 0.75,
1245
+ "learning_rate": 5e-05,
1246
+ "loss": 1.0999,
1247
+ "step": 102500
1248
+ },
1249
+ {
1250
+ "epoch": 0.75,
1251
+ "learning_rate": 5e-05,
1252
+ "loss": 1.101,
1253
+ "step": 103000
1254
+ },
1255
+ {
1256
+ "epoch": 0.76,
1257
+ "learning_rate": 5e-05,
1258
+ "loss": 1.1041,
1259
+ "step": 103500
1260
+ },
1261
+ {
1262
+ "epoch": 0.76,
1263
+ "learning_rate": 5e-05,
1264
+ "loss": 1.1,
1265
+ "step": 104000
1266
+ },
1267
+ {
1268
+ "epoch": 0.76,
1269
+ "learning_rate": 5e-05,
1270
+ "loss": 1.109,
1271
+ "step": 104500
1272
+ },
1273
+ {
1274
+ "epoch": 0.77,
1275
+ "learning_rate": 5e-05,
1276
+ "loss": 1.105,
1277
+ "step": 105000
1278
+ },
1279
+ {
1280
+ "epoch": 0.77,
1281
+ "learning_rate": 5e-05,
1282
+ "loss": 1.1031,
1283
+ "step": 105500
1284
+ },
1285
+ {
1286
+ "epoch": 0.78,
1287
+ "learning_rate": 5e-05,
1288
+ "loss": 1.0965,
1289
+ "step": 106000
1290
+ },
1291
+ {
1292
+ "epoch": 0.78,
1293
+ "learning_rate": 5e-05,
1294
+ "loss": 1.0958,
1295
+ "step": 106500
1296
+ },
1297
+ {
1298
+ "epoch": 0.78,
1299
+ "learning_rate": 5e-05,
1300
+ "loss": 1.0959,
1301
+ "step": 107000
1302
+ },
1303
+ {
1304
+ "epoch": 0.79,
1305
+ "learning_rate": 5e-05,
1306
+ "loss": 1.1022,
1307
+ "step": 107500
1308
+ },
1309
+ {
1310
+ "epoch": 0.79,
1311
+ "learning_rate": 5e-05,
1312
+ "loss": 1.0999,
1313
+ "step": 108000
1314
+ },
1315
+ {
1316
+ "epoch": 0.79,
1317
+ "learning_rate": 5e-05,
1318
+ "loss": 1.1004,
1319
+ "step": 108500
1320
+ },
1321
+ {
1322
+ "epoch": 0.8,
1323
+ "learning_rate": 5e-05,
1324
+ "loss": 1.103,
1325
+ "step": 109000
1326
+ },
1327
+ {
1328
+ "epoch": 0.8,
1329
+ "learning_rate": 5e-05,
1330
+ "loss": 1.0974,
1331
+ "step": 109500
1332
+ },
1333
+ {
1334
+ "epoch": 0.81,
1335
+ "learning_rate": 5e-05,
1336
+ "loss": 1.0986,
1337
+ "step": 110000
1338
+ },
1339
+ {
1340
+ "epoch": 0.81,
1341
+ "learning_rate": 5e-05,
1342
+ "loss": 1.098,
1343
+ "step": 110500
1344
+ },
1345
+ {
1346
+ "epoch": 0.81,
1347
+ "learning_rate": 5e-05,
1348
+ "loss": 1.0951,
1349
+ "step": 111000
1350
+ },
1351
+ {
1352
+ "epoch": 0.82,
1353
+ "learning_rate": 5e-05,
1354
+ "loss": 1.0915,
1355
+ "step": 111500
1356
+ },
1357
+ {
1358
+ "epoch": 0.82,
1359
+ "learning_rate": 5e-05,
1360
+ "loss": 1.0952,
1361
+ "step": 112000
1362
+ },
1363
+ {
1364
+ "epoch": 0.82,
1365
+ "learning_rate": 5e-05,
1366
+ "loss": 1.1,
1367
+ "step": 112500
1368
+ },
1369
+ {
1370
+ "epoch": 0.83,
1371
+ "learning_rate": 5e-05,
1372
+ "loss": 1.099,
1373
+ "step": 113000
1374
+ },
1375
+ {
1376
+ "epoch": 0.83,
1377
+ "learning_rate": 5e-05,
1378
+ "loss": 1.0868,
1379
+ "step": 113500
1380
+ },
1381
+ {
1382
+ "epoch": 0.83,
1383
+ "learning_rate": 5e-05,
1384
+ "loss": 1.0906,
1385
+ "step": 114000
1386
+ },
1387
+ {
1388
+ "epoch": 0.84,
1389
+ "learning_rate": 5e-05,
1390
+ "loss": 1.0844,
1391
+ "step": 114500
1392
+ },
1393
+ {
1394
+ "epoch": 0.84,
1395
+ "learning_rate": 5e-05,
1396
+ "loss": 1.0868,
1397
+ "step": 115000
1398
+ },
1399
+ {
1400
+ "epoch": 0.85,
1401
+ "learning_rate": 5e-05,
1402
+ "loss": 1.0802,
1403
+ "step": 115500
1404
+ },
1405
+ {
1406
+ "epoch": 0.85,
1407
+ "learning_rate": 5e-05,
1408
+ "loss": 1.0863,
1409
+ "step": 116000
1410
+ },
1411
+ {
1412
+ "epoch": 0.85,
1413
+ "learning_rate": 5e-05,
1414
+ "loss": 1.0922,
1415
+ "step": 116500
1416
+ },
1417
+ {
1418
+ "epoch": 0.86,
1419
+ "learning_rate": 5e-05,
1420
+ "loss": 1.0927,
1421
+ "step": 117000
1422
+ },
1423
+ {
1424
+ "epoch": 0.86,
1425
+ "learning_rate": 5e-05,
1426
+ "loss": 1.0822,
1427
+ "step": 117500
1428
+ },
1429
+ {
1430
+ "epoch": 0.86,
1431
+ "learning_rate": 5e-05,
1432
+ "loss": 1.0885,
1433
+ "step": 118000
1434
+ },
1435
+ {
1436
+ "epoch": 0.87,
1437
+ "learning_rate": 5e-05,
1438
+ "loss": 1.0944,
1439
+ "step": 118500
1440
+ },
1441
+ {
1442
+ "epoch": 0.87,
1443
+ "learning_rate": 5e-05,
1444
+ "loss": 1.0775,
1445
+ "step": 119000
1446
+ },
1447
+ {
1448
+ "epoch": 0.87,
1449
+ "learning_rate": 5e-05,
1450
+ "loss": 1.088,
1451
+ "step": 119500
1452
+ },
1453
+ {
1454
+ "epoch": 0.88,
1455
+ "learning_rate": 5e-05,
1456
+ "loss": 1.0893,
1457
+ "step": 120000
1458
+ },
1459
+ {
1460
+ "epoch": 0.88,
1461
+ "learning_rate": 5e-05,
1462
+ "loss": 1.084,
1463
+ "step": 120500
1464
+ },
1465
+ {
1466
+ "epoch": 0.89,
1467
+ "learning_rate": 5e-05,
1468
+ "loss": 1.0937,
1469
+ "step": 121000
1470
+ },
1471
+ {
1472
+ "epoch": 0.89,
1473
+ "learning_rate": 5e-05,
1474
+ "loss": 1.0857,
1475
+ "step": 121500
1476
+ },
1477
+ {
1478
+ "epoch": 0.89,
1479
+ "learning_rate": 5e-05,
1480
+ "loss": 1.089,
1481
+ "step": 122000
1482
+ },
1483
+ {
1484
+ "epoch": 0.9,
1485
+ "learning_rate": 5e-05,
1486
+ "loss": 1.0913,
1487
+ "step": 122500
1488
+ },
1489
+ {
1490
+ "epoch": 0.9,
1491
+ "learning_rate": 5e-05,
1492
+ "loss": 1.0835,
1493
+ "step": 123000
1494
+ },
1495
+ {
1496
+ "epoch": 0.9,
1497
+ "learning_rate": 5e-05,
1498
+ "loss": 1.0889,
1499
+ "step": 123500
1500
+ },
1501
+ {
1502
+ "epoch": 0.91,
1503
+ "learning_rate": 5e-05,
1504
+ "loss": 1.0848,
1505
+ "step": 124000
1506
+ },
1507
+ {
1508
+ "epoch": 0.91,
1509
+ "learning_rate": 5e-05,
1510
+ "loss": 1.0824,
1511
+ "step": 124500
1512
+ },
1513
+ {
1514
+ "epoch": 0.91,
1515
+ "learning_rate": 5e-05,
1516
+ "loss": 1.0849,
1517
+ "step": 125000
1518
+ },
1519
+ {
1520
+ "epoch": 0.92,
1521
+ "learning_rate": 5e-05,
1522
+ "loss": 1.0788,
1523
+ "step": 125500
1524
+ },
1525
+ {
1526
+ "epoch": 0.92,
1527
+ "learning_rate": 5e-05,
1528
+ "loss": 1.0886,
1529
+ "step": 126000
1530
+ },
1531
+ {
1532
+ "epoch": 0.93,
1533
+ "learning_rate": 5e-05,
1534
+ "loss": 1.0891,
1535
+ "step": 126500
1536
+ },
1537
+ {
1538
+ "epoch": 0.93,
1539
+ "learning_rate": 5e-05,
1540
+ "loss": 1.0835,
1541
+ "step": 127000
1542
+ },
1543
+ {
1544
+ "epoch": 0.93,
1545
+ "learning_rate": 5e-05,
1546
+ "loss": 1.0814,
1547
+ "step": 127500
1548
+ },
1549
+ {
1550
+ "epoch": 0.94,
1551
+ "learning_rate": 5e-05,
1552
+ "loss": 1.0833,
1553
+ "step": 128000
1554
+ },
1555
+ {
1556
+ "epoch": 0.94,
1557
+ "eval_accuracy": 0.7811326318893699,
1558
+ "eval_loss": 1.0721635818481445,
1559
+ "eval_runtime": 32.739,
1560
+ "eval_samples_per_second": 409.787,
1561
+ "eval_steps_per_second": 8.553,
1562
+ "step": 128000
1563
+ },
1564
+ {
1565
+ "epoch": 0.94,
1566
+ "learning_rate": 5e-05,
1567
+ "loss": 1.0762,
1568
+ "step": 128500
1569
+ },
1570
+ {
1571
+ "epoch": 0.94,
1572
+ "learning_rate": 5e-05,
1573
+ "loss": 1.0737,
1574
+ "step": 129000
1575
+ },
1576
+ {
1577
+ "epoch": 0.95,
1578
+ "learning_rate": 5e-05,
1579
+ "loss": 1.077,
1580
+ "step": 129500
1581
+ },
1582
+ {
1583
+ "epoch": 0.95,
1584
+ "learning_rate": 5e-05,
1585
+ "loss": 1.0795,
1586
+ "step": 130000
1587
+ },
1588
+ {
1589
+ "epoch": 0.96,
1590
+ "learning_rate": 5e-05,
1591
+ "loss": 1.0785,
1592
+ "step": 130500
1593
+ },
1594
+ {
1595
+ "epoch": 0.96,
1596
+ "learning_rate": 5e-05,
1597
+ "loss": 1.0784,
1598
+ "step": 131000
1599
+ },
1600
+ {
1601
+ "epoch": 0.96,
1602
+ "learning_rate": 5e-05,
1603
+ "loss": 1.0768,
1604
+ "step": 131500
1605
+ },
1606
+ {
1607
+ "epoch": 0.97,
1608
+ "learning_rate": 5e-05,
1609
+ "loss": 1.0787,
1610
+ "step": 132000
1611
+ },
1612
+ {
1613
+ "epoch": 0.97,
1614
+ "learning_rate": 5e-05,
1615
+ "loss": 1.0761,
1616
+ "step": 132500
1617
+ },
1618
+ {
1619
+ "epoch": 0.97,
1620
+ "learning_rate": 5e-05,
1621
+ "loss": 1.0734,
1622
+ "step": 133000
1623
+ },
1624
+ {
1625
+ "epoch": 0.98,
1626
+ "learning_rate": 5e-05,
1627
+ "loss": 1.0781,
1628
+ "step": 133500
1629
+ },
1630
+ {
1631
+ "epoch": 0.98,
1632
+ "learning_rate": 5e-05,
1633
+ "loss": 1.0689,
1634
+ "step": 134000
1635
+ },
1636
+ {
1637
+ "epoch": 0.98,
1638
+ "learning_rate": 5e-05,
1639
+ "loss": 1.0794,
1640
+ "step": 134500
1641
+ },
1642
+ {
1643
+ "epoch": 0.99,
1644
+ "learning_rate": 5e-05,
1645
+ "loss": 1.0747,
1646
+ "step": 135000
1647
+ },
1648
+ {
1649
+ "epoch": 0.99,
1650
+ "learning_rate": 5e-05,
1651
+ "loss": 1.0769,
1652
+ "step": 135500
1653
+ },
1654
+ {
1655
+ "epoch": 1.0,
1656
+ "learning_rate": 5e-05,
1657
+ "loss": 1.0722,
1658
+ "step": 136000
1659
+ },
1660
+ {
1661
+ "epoch": 1.0,
1662
+ "learning_rate": 5e-05,
1663
+ "loss": 1.068,
1664
+ "step": 136500
1665
+ },
1666
+ {
1667
+ "epoch": 1.0,
1668
+ "learning_rate": 5e-05,
1669
+ "loss": 1.0729,
1670
+ "step": 137000
1671
+ },
1672
+ {
1673
+ "epoch": 1.01,
1674
+ "learning_rate": 5e-05,
1675
+ "loss": 1.0686,
1676
+ "step": 137500
1677
+ },
1678
+ {
1679
+ "epoch": 1.01,
1680
+ "learning_rate": 5e-05,
1681
+ "loss": 1.0706,
1682
+ "step": 138000
1683
+ },
1684
+ {
1685
+ "epoch": 1.01,
1686
+ "learning_rate": 5e-05,
1687
+ "loss": 1.0744,
1688
+ "step": 138500
1689
+ },
1690
+ {
1691
+ "epoch": 1.02,
1692
+ "learning_rate": 5e-05,
1693
+ "loss": 1.0701,
1694
+ "step": 139000
1695
+ },
1696
+ {
1697
+ "epoch": 1.02,
1698
+ "learning_rate": 5e-05,
1699
+ "loss": 1.07,
1700
+ "step": 139500
1701
+ },
1702
+ {
1703
+ "epoch": 1.02,
1704
+ "learning_rate": 5e-05,
1705
+ "loss": 1.0699,
1706
+ "step": 140000
1707
+ },
1708
+ {
1709
+ "epoch": 1.03,
1710
+ "learning_rate": 5e-05,
1711
+ "loss": 1.069,
1712
+ "step": 140500
1713
+ },
1714
+ {
1715
+ "epoch": 1.03,
1716
+ "learning_rate": 5e-05,
1717
+ "loss": 1.0741,
1718
+ "step": 141000
1719
+ },
1720
+ {
1721
+ "epoch": 1.04,
1722
+ "learning_rate": 5e-05,
1723
+ "loss": 1.0693,
1724
+ "step": 141500
1725
+ },
1726
+ {
1727
+ "epoch": 1.04,
1728
+ "learning_rate": 5e-05,
1729
+ "loss": 1.0725,
1730
+ "step": 142000
1731
+ },
1732
+ {
1733
+ "epoch": 1.04,
1734
+ "learning_rate": 5e-05,
1735
+ "loss": 1.073,
1736
+ "step": 142500
1737
+ },
1738
+ {
1739
+ "epoch": 1.05,
1740
+ "learning_rate": 5e-05,
1741
+ "loss": 1.0739,
1742
+ "step": 143000
1743
+ },
1744
+ {
1745
+ "epoch": 1.05,
1746
+ "learning_rate": 5e-05,
1747
+ "loss": 1.068,
1748
+ "step": 143500
1749
+ },
1750
+ {
1751
+ "epoch": 1.05,
1752
+ "learning_rate": 5e-05,
1753
+ "loss": 1.0678,
1754
+ "step": 144000
1755
+ },
1756
+ {
1757
+ "epoch": 1.06,
1758
+ "learning_rate": 5e-05,
1759
+ "loss": 1.0679,
1760
+ "step": 144500
1761
+ },
1762
+ {
1763
+ "epoch": 1.06,
1764
+ "learning_rate": 5e-05,
1765
+ "loss": 1.069,
1766
+ "step": 145000
1767
+ },
1768
+ {
1769
+ "epoch": 1.06,
1770
+ "learning_rate": 5e-05,
1771
+ "loss": 1.0707,
1772
+ "step": 145500
1773
+ },
1774
+ {
1775
+ "epoch": 1.07,
1776
+ "learning_rate": 5e-05,
1777
+ "loss": 1.0687,
1778
+ "step": 146000
1779
+ },
1780
+ {
1781
+ "epoch": 1.07,
1782
+ "learning_rate": 5e-05,
1783
+ "loss": 1.073,
1784
+ "step": 146500
1785
+ },
1786
+ {
1787
+ "epoch": 1.08,
1788
+ "learning_rate": 5e-05,
1789
+ "loss": 1.0643,
1790
+ "step": 147000
1791
+ },
1792
+ {
1793
+ "epoch": 1.08,
1794
+ "learning_rate": 5e-05,
1795
+ "loss": 1.0647,
1796
+ "step": 147500
1797
+ },
1798
+ {
1799
+ "epoch": 1.08,
1800
+ "learning_rate": 5e-05,
1801
+ "loss": 1.0671,
1802
+ "step": 148000
1803
+ },
1804
+ {
1805
+ "epoch": 1.09,
1806
+ "learning_rate": 5e-05,
1807
+ "loss": 1.0722,
1808
+ "step": 148500
1809
+ },
1810
+ {
1811
+ "epoch": 1.09,
1812
+ "learning_rate": 5e-05,
1813
+ "loss": 1.0692,
1814
+ "step": 149000
1815
+ },
1816
+ {
1817
+ "epoch": 1.09,
1818
+ "learning_rate": 5e-05,
1819
+ "loss": 1.071,
1820
+ "step": 149500
1821
+ },
1822
+ {
1823
+ "epoch": 1.1,
1824
+ "learning_rate": 5e-05,
1825
+ "loss": 1.071,
1826
+ "step": 150000
1827
+ },
1828
+ {
1829
+ "epoch": 1.1,
1830
+ "learning_rate": 5e-05,
1831
+ "loss": 1.0705,
1832
+ "step": 150500
1833
+ },
1834
+ {
1835
+ "epoch": 1.11,
1836
+ "learning_rate": 5e-05,
1837
+ "loss": 1.0657,
1838
+ "step": 151000
1839
+ },
1840
+ {
1841
+ "epoch": 1.11,
1842
+ "learning_rate": 5e-05,
1843
+ "loss": 1.0744,
1844
+ "step": 151500
1845
+ },
1846
+ {
1847
+ "epoch": 1.11,
1848
+ "learning_rate": 5e-05,
1849
+ "loss": 1.0655,
1850
+ "step": 152000
1851
+ },
1852
+ {
1853
+ "epoch": 1.12,
1854
+ "learning_rate": 5e-05,
1855
+ "loss": 1.0688,
1856
+ "step": 152500
1857
+ },
1858
+ {
1859
+ "epoch": 1.12,
1860
+ "learning_rate": 5e-05,
1861
+ "loss": 1.0594,
1862
+ "step": 153000
1863
+ },
1864
+ {
1865
+ "epoch": 1.12,
1866
+ "learning_rate": 5e-05,
1867
+ "loss": 1.0704,
1868
+ "step": 153500
1869
+ },
1870
+ {
1871
+ "epoch": 1.13,
1872
+ "learning_rate": 5e-05,
1873
+ "loss": 1.0613,
1874
+ "step": 154000
1875
+ },
1876
+ {
1877
+ "epoch": 1.13,
1878
+ "learning_rate": 5e-05,
1879
+ "loss": 1.0671,
1880
+ "step": 154500
1881
+ },
1882
+ {
1883
+ "epoch": 1.13,
1884
+ "learning_rate": 5e-05,
1885
+ "loss": 1.0646,
1886
+ "step": 155000
1887
+ },
1888
+ {
1889
+ "epoch": 1.14,
1890
+ "learning_rate": 5e-05,
1891
+ "loss": 1.0616,
1892
+ "step": 155500
1893
+ },
1894
+ {
1895
+ "epoch": 1.14,
1896
+ "learning_rate": 5e-05,
1897
+ "loss": 1.0631,
1898
+ "step": 156000
1899
+ },
1900
+ {
1901
+ "epoch": 1.15,
1902
+ "learning_rate": 5e-05,
1903
+ "loss": 1.0581,
1904
+ "step": 156500
1905
+ },
1906
+ {
1907
+ "epoch": 1.15,
1908
+ "learning_rate": 5e-05,
1909
+ "loss": 1.0626,
1910
+ "step": 157000
1911
+ },
1912
+ {
1913
+ "epoch": 1.15,
1914
+ "learning_rate": 5e-05,
1915
+ "loss": 1.054,
1916
+ "step": 157500
1917
+ },
1918
+ {
1919
+ "epoch": 1.16,
1920
+ "learning_rate": 5e-05,
1921
+ "loss": 1.0635,
1922
+ "step": 158000
1923
+ },
1924
+ {
1925
+ "epoch": 1.16,
1926
+ "learning_rate": 5e-05,
1927
+ "loss": 1.0572,
1928
+ "step": 158500
1929
+ },
1930
+ {
1931
+ "epoch": 1.16,
1932
+ "learning_rate": 5e-05,
1933
+ "loss": 1.0596,
1934
+ "step": 159000
1935
+ },
1936
+ {
1937
+ "epoch": 1.17,
1938
+ "learning_rate": 5e-05,
1939
+ "loss": 1.0627,
1940
+ "step": 159500
1941
+ },
1942
+ {
1943
+ "epoch": 1.17,
1944
+ "learning_rate": 5e-05,
1945
+ "loss": 1.0647,
1946
+ "step": 160000
1947
+ },
1948
+ {
1949
+ "epoch": 1.17,
1950
+ "learning_rate": 5e-05,
1951
+ "loss": 1.0556,
1952
+ "step": 160500
1953
+ },
1954
+ {
1955
+ "epoch": 1.18,
1956
+ "learning_rate": 5e-05,
1957
+ "loss": 1.0541,
1958
+ "step": 161000
1959
+ },
1960
+ {
1961
+ "epoch": 1.18,
1962
+ "learning_rate": 5e-05,
1963
+ "loss": 1.0534,
1964
+ "step": 161500
1965
+ },
1966
+ {
1967
+ "epoch": 1.19,
1968
+ "learning_rate": 5e-05,
1969
+ "loss": 1.0516,
1970
+ "step": 162000
1971
+ },
1972
+ {
1973
+ "epoch": 1.19,
1974
+ "learning_rate": 5e-05,
1975
+ "loss": 1.0564,
1976
+ "step": 162500
1977
+ },
1978
+ {
1979
+ "epoch": 1.19,
1980
+ "learning_rate": 5e-05,
1981
+ "loss": 1.0497,
1982
+ "step": 163000
1983
+ },
1984
+ {
1985
+ "epoch": 1.2,
1986
+ "learning_rate": 5e-05,
1987
+ "loss": 1.0643,
1988
+ "step": 163500
1989
+ },
1990
+ {
1991
+ "epoch": 1.2,
1992
+ "learning_rate": 5e-05,
1993
+ "loss": 1.0552,
1994
+ "step": 164000
1995
+ },
1996
+ {
1997
+ "epoch": 1.2,
1998
+ "learning_rate": 5e-05,
1999
+ "loss": 1.058,
2000
+ "step": 164500
2001
+ },
2002
+ {
2003
+ "epoch": 1.21,
2004
+ "learning_rate": 5e-05,
2005
+ "loss": 1.0515,
2006
+ "step": 165000
2007
+ },
2008
+ {
2009
+ "epoch": 1.21,
2010
+ "learning_rate": 5e-05,
2011
+ "loss": 1.0529,
2012
+ "step": 165500
2013
+ },
2014
+ {
2015
+ "epoch": 1.21,
2016
+ "learning_rate": 5e-05,
2017
+ "loss": 1.0567,
2018
+ "step": 166000
2019
+ },
2020
+ {
2021
+ "epoch": 1.22,
2022
+ "learning_rate": 5e-05,
2023
+ "loss": 1.0586,
2024
+ "step": 166500
2025
+ },
2026
+ {
2027
+ "epoch": 1.22,
2028
+ "learning_rate": 5e-05,
2029
+ "loss": 1.0576,
2030
+ "step": 167000
2031
+ },
2032
+ {
2033
+ "epoch": 1.23,
2034
+ "learning_rate": 5e-05,
2035
+ "loss": 1.0514,
2036
+ "step": 167500
2037
+ },
2038
+ {
2039
+ "epoch": 1.23,
2040
+ "learning_rate": 5e-05,
2041
+ "loss": 1.054,
2042
+ "step": 168000
2043
+ },
2044
+ {
2045
+ "epoch": 1.23,
2046
+ "learning_rate": 5e-05,
2047
+ "loss": 1.0559,
2048
+ "step": 168500
2049
+ },
2050
+ {
2051
+ "epoch": 1.24,
2052
+ "learning_rate": 5e-05,
2053
+ "loss": 1.0512,
2054
+ "step": 169000
2055
+ },
2056
+ {
2057
+ "epoch": 1.24,
2058
+ "learning_rate": 5e-05,
2059
+ "loss": 1.0609,
2060
+ "step": 169500
2061
+ },
2062
+ {
2063
+ "epoch": 1.24,
2064
+ "learning_rate": 5e-05,
2065
+ "loss": 1.0571,
2066
+ "step": 170000
2067
+ },
2068
+ {
2069
+ "epoch": 1.25,
2070
+ "learning_rate": 5e-05,
2071
+ "loss": 1.0535,
2072
+ "step": 170500
2073
+ },
2074
+ {
2075
+ "epoch": 1.25,
2076
+ "learning_rate": 5e-05,
2077
+ "loss": 1.0533,
2078
+ "step": 171000
2079
+ },
2080
+ {
2081
+ "epoch": 1.26,
2082
+ "learning_rate": 5e-05,
2083
+ "loss": 1.0436,
2084
+ "step": 171500
2085
+ },
2086
+ {
2087
+ "epoch": 1.26,
2088
+ "learning_rate": 5e-05,
2089
+ "loss": 1.0507,
2090
+ "step": 172000
2091
+ },
2092
+ {
2093
+ "epoch": 1.26,
2094
+ "learning_rate": 5e-05,
2095
+ "loss": 1.0522,
2096
+ "step": 172500
2097
+ },
2098
+ {
2099
+ "epoch": 1.27,
2100
+ "learning_rate": 5e-05,
2101
+ "loss": 1.0623,
2102
+ "step": 173000
2103
+ },
2104
+ {
2105
+ "epoch": 1.27,
2106
+ "learning_rate": 5e-05,
2107
+ "loss": 1.0552,
2108
+ "step": 173500
2109
+ },
2110
+ {
2111
+ "epoch": 1.27,
2112
+ "learning_rate": 5e-05,
2113
+ "loss": 1.0512,
2114
+ "step": 174000
2115
+ },
2116
+ {
2117
+ "epoch": 1.28,
2118
+ "learning_rate": 5e-05,
2119
+ "loss": 1.0561,
2120
+ "step": 174500
2121
+ },
2122
+ {
2123
+ "epoch": 1.28,
2124
+ "learning_rate": 5e-05,
2125
+ "loss": 1.0534,
2126
+ "step": 175000
2127
+ },
2128
+ {
2129
+ "epoch": 1.28,
2130
+ "learning_rate": 5e-05,
2131
+ "loss": 1.0453,
2132
+ "step": 175500
2133
+ },
2134
+ {
2135
+ "epoch": 1.29,
2136
+ "learning_rate": 5e-05,
2137
+ "loss": 1.0523,
2138
+ "step": 176000
2139
+ },
2140
+ {
2141
+ "epoch": 1.29,
2142
+ "learning_rate": 5e-05,
2143
+ "loss": 1.0526,
2144
+ "step": 176500
2145
+ },
2146
+ {
2147
+ "epoch": 1.3,
2148
+ "learning_rate": 5e-05,
2149
+ "loss": 1.0498,
2150
+ "step": 177000
2151
+ },
2152
+ {
2153
+ "epoch": 1.3,
2154
+ "learning_rate": 5e-05,
2155
+ "loss": 1.0523,
2156
+ "step": 177500
2157
+ },
2158
+ {
2159
+ "epoch": 1.3,
2160
+ "learning_rate": 5e-05,
2161
+ "loss": 1.0508,
2162
+ "step": 178000
2163
+ },
2164
+ {
2165
+ "epoch": 1.31,
2166
+ "learning_rate": 5e-05,
2167
+ "loss": 1.0492,
2168
+ "step": 178500
2169
+ },
2170
+ {
2171
+ "epoch": 1.31,
2172
+ "learning_rate": 5e-05,
2173
+ "loss": 1.0492,
2174
+ "step": 179000
2175
+ },
2176
+ {
2177
+ "epoch": 1.31,
2178
+ "learning_rate": 5e-05,
2179
+ "loss": 1.0475,
2180
+ "step": 179500
2181
+ },
2182
+ {
2183
+ "epoch": 1.32,
2184
+ "learning_rate": 5e-05,
2185
+ "loss": 1.0491,
2186
+ "step": 180000
2187
+ },
2188
+ {
2189
+ "epoch": 1.32,
2190
+ "learning_rate": 5e-05,
2191
+ "loss": 1.047,
2192
+ "step": 180500
2193
+ },
2194
+ {
2195
+ "epoch": 1.32,
2196
+ "learning_rate": 5e-05,
2197
+ "loss": 1.0509,
2198
+ "step": 181000
2199
+ },
2200
+ {
2201
+ "epoch": 1.33,
2202
+ "learning_rate": 5e-05,
2203
+ "loss": 1.044,
2204
+ "step": 181500
2205
+ },
2206
+ {
2207
+ "epoch": 1.33,
2208
+ "learning_rate": 5e-05,
2209
+ "loss": 1.0461,
2210
+ "step": 182000
2211
+ },
2212
+ {
2213
+ "epoch": 1.34,
2214
+ "learning_rate": 5e-05,
2215
+ "loss": 1.0459,
2216
+ "step": 182500
2217
+ },
2218
+ {
2219
+ "epoch": 1.34,
2220
+ "learning_rate": 5e-05,
2221
+ "loss": 1.0479,
2222
+ "step": 183000
2223
+ },
2224
+ {
2225
+ "epoch": 1.34,
2226
+ "learning_rate": 5e-05,
2227
+ "loss": 1.0478,
2228
+ "step": 183500
2229
+ },
2230
+ {
2231
+ "epoch": 1.35,
2232
+ "learning_rate": 5e-05,
2233
+ "loss": 1.0499,
2234
+ "step": 184000
2235
+ },
2236
+ {
2237
+ "epoch": 1.35,
2238
+ "learning_rate": 5e-05,
2239
+ "loss": 1.0499,
2240
+ "step": 184500
2241
+ },
2242
+ {
2243
+ "epoch": 1.35,
2244
+ "learning_rate": 5e-05,
2245
+ "loss": 1.0542,
2246
+ "step": 185000
2247
+ },
2248
+ {
2249
+ "epoch": 1.36,
2250
+ "learning_rate": 5e-05,
2251
+ "loss": 1.0539,
2252
+ "step": 185500
2253
+ },
2254
+ {
2255
+ "epoch": 1.36,
2256
+ "learning_rate": 5e-05,
2257
+ "loss": 1.0562,
2258
+ "step": 186000
2259
+ },
2260
+ {
2261
+ "epoch": 1.37,
2262
+ "learning_rate": 5e-05,
2263
+ "loss": 1.0478,
2264
+ "step": 186500
2265
+ },
2266
+ {
2267
+ "epoch": 1.37,
2268
+ "learning_rate": 5e-05,
2269
+ "loss": 1.0491,
2270
+ "step": 187000
2271
+ },
2272
+ {
2273
+ "epoch": 1.37,
2274
+ "learning_rate": 5e-05,
2275
+ "loss": 1.0481,
2276
+ "step": 187500
2277
+ },
2278
+ {
2279
+ "epoch": 1.38,
2280
+ "learning_rate": 5e-05,
2281
+ "loss": 1.0525,
2282
+ "step": 188000
2283
+ },
2284
+ {
2285
+ "epoch": 1.38,
2286
+ "learning_rate": 5e-05,
2287
+ "loss": 1.0487,
2288
+ "step": 188500
2289
+ },
2290
+ {
2291
+ "epoch": 1.38,
2292
+ "learning_rate": 5e-05,
2293
+ "loss": 1.0478,
2294
+ "step": 189000
2295
+ },
2296
+ {
2297
+ "epoch": 1.39,
2298
+ "learning_rate": 5e-05,
2299
+ "loss": 1.0454,
2300
+ "step": 189500
2301
+ },
2302
+ {
2303
+ "epoch": 1.39,
2304
+ "learning_rate": 5e-05,
2305
+ "loss": 1.0514,
2306
+ "step": 190000
2307
+ },
2308
+ {
2309
+ "epoch": 1.39,
2310
+ "learning_rate": 5e-05,
2311
+ "loss": 1.0464,
2312
+ "step": 190500
2313
+ },
2314
+ {
2315
+ "epoch": 1.4,
2316
+ "learning_rate": 5e-05,
2317
+ "loss": 1.0503,
2318
+ "step": 191000
2319
+ },
2320
+ {
2321
+ "epoch": 1.4,
2322
+ "learning_rate": 5e-05,
2323
+ "loss": 1.0444,
2324
+ "step": 191500
2325
+ },
2326
+ {
2327
+ "epoch": 1.41,
2328
+ "learning_rate": 5e-05,
2329
+ "loss": 1.0459,
2330
+ "step": 192000
2331
+ },
2332
+ {
2333
+ "epoch": 1.41,
2334
+ "eval_accuracy": 0.7880748504192896,
2335
+ "eval_loss": 1.0307050943374634,
2336
+ "eval_runtime": 32.7044,
2337
+ "eval_samples_per_second": 410.22,
2338
+ "eval_steps_per_second": 8.562,
2339
+ "step": 192000
2340
+ },
2341
+ {
2342
+ "epoch": 1.41,
2343
+ "learning_rate": 5e-05,
2344
+ "loss": 1.0412,
2345
+ "step": 192500
2346
+ },
2347
+ {
2348
+ "epoch": 1.41,
2349
+ "learning_rate": 5e-05,
2350
+ "loss": 1.0555,
2351
+ "step": 193000
2352
+ },
2353
+ {
2354
+ "epoch": 1.42,
2355
+ "learning_rate": 5e-05,
2356
+ "loss": 1.0413,
2357
+ "step": 193500
2358
+ },
2359
+ {
2360
+ "epoch": 1.42,
2361
+ "learning_rate": 5e-05,
2362
+ "loss": 1.0481,
2363
+ "step": 194000
2364
+ },
2365
+ {
2366
+ "epoch": 1.42,
2367
+ "learning_rate": 5e-05,
2368
+ "loss": 1.0471,
2369
+ "step": 194500
2370
+ },
2371
+ {
2372
+ "epoch": 1.43,
2373
+ "learning_rate": 5e-05,
2374
+ "loss": 1.046,
2375
+ "step": 195000
2376
+ },
2377
+ {
2378
+ "epoch": 1.43,
2379
+ "learning_rate": 5e-05,
2380
+ "loss": 1.0478,
2381
+ "step": 195500
2382
+ },
2383
+ {
2384
+ "epoch": 1.43,
2385
+ "learning_rate": 5e-05,
2386
+ "loss": 1.0436,
2387
+ "step": 196000
2388
+ },
2389
+ {
2390
+ "epoch": 1.44,
2391
+ "learning_rate": 5e-05,
2392
+ "loss": 1.0442,
2393
+ "step": 196500
2394
+ },
2395
+ {
2396
+ "epoch": 1.44,
2397
+ "learning_rate": 5e-05,
2398
+ "loss": 1.0473,
2399
+ "step": 197000
2400
+ },
2401
+ {
2402
+ "epoch": 1.45,
2403
+ "learning_rate": 5e-05,
2404
+ "loss": 1.0386,
2405
+ "step": 197500
2406
+ },
2407
+ {
2408
+ "epoch": 1.45,
2409
+ "learning_rate": 5e-05,
2410
+ "loss": 1.0419,
2411
+ "step": 198000
2412
+ },
2413
+ {
2414
+ "epoch": 1.45,
2415
+ "learning_rate": 5e-05,
2416
+ "loss": 1.0435,
2417
+ "step": 198500
2418
+ },
2419
+ {
2420
+ "epoch": 1.46,
2421
+ "learning_rate": 5e-05,
2422
+ "loss": 1.0436,
2423
+ "step": 199000
2424
+ },
2425
+ {
2426
+ "epoch": 1.46,
2427
+ "learning_rate": 5e-05,
2428
+ "loss": 1.0451,
2429
+ "step": 199500
2430
+ },
2431
+ {
2432
+ "epoch": 1.46,
2433
+ "learning_rate": 5e-05,
2434
+ "loss": 1.048,
2435
+ "step": 200000
2436
+ },
2437
+ {
2438
+ "epoch": 1.47,
2439
+ "learning_rate": 5e-05,
2440
+ "loss": 1.0477,
2441
+ "step": 200500
2442
+ },
2443
+ {
2444
+ "epoch": 1.47,
2445
+ "learning_rate": 5e-05,
2446
+ "loss": 1.0453,
2447
+ "step": 201000
2448
+ },
2449
+ {
2450
+ "epoch": 1.47,
2451
+ "learning_rate": 5e-05,
2452
+ "loss": 1.0435,
2453
+ "step": 201500
2454
+ },
2455
+ {
2456
+ "epoch": 1.48,
2457
+ "learning_rate": 5e-05,
2458
+ "loss": 1.0467,
2459
+ "step": 202000
2460
+ },
2461
+ {
2462
+ "epoch": 1.48,
2463
+ "learning_rate": 5e-05,
2464
+ "loss": 1.044,
2465
+ "step": 202500
2466
+ },
2467
+ {
2468
+ "epoch": 1.49,
2469
+ "learning_rate": 5e-05,
2470
+ "loss": 1.0495,
2471
+ "step": 203000
2472
+ },
2473
+ {
2474
+ "epoch": 1.49,
2475
+ "learning_rate": 5e-05,
2476
+ "loss": 1.0374,
2477
+ "step": 203500
2478
+ },
2479
+ {
2480
+ "epoch": 1.49,
2481
+ "learning_rate": 5e-05,
2482
+ "loss": 1.0421,
2483
+ "step": 204000
2484
+ },
2485
+ {
2486
+ "epoch": 1.5,
2487
+ "learning_rate": 5e-05,
2488
+ "loss": 1.037,
2489
+ "step": 204500
2490
+ },
2491
+ {
2492
+ "epoch": 1.5,
2493
+ "learning_rate": 5e-05,
2494
+ "loss": 1.0383,
2495
+ "step": 205000
2496
+ },
2497
+ {
2498
+ "epoch": 1.5,
2499
+ "learning_rate": 5e-05,
2500
+ "loss": 1.038,
2501
+ "step": 205500
2502
+ },
2503
+ {
2504
+ "epoch": 1.51,
2505
+ "learning_rate": 5e-05,
2506
+ "loss": 1.038,
2507
+ "step": 206000
2508
+ },
2509
+ {
2510
+ "epoch": 1.51,
2511
+ "learning_rate": 5e-05,
2512
+ "loss": 1.046,
2513
+ "step": 206500
2514
+ },
2515
+ {
2516
+ "epoch": 1.52,
2517
+ "learning_rate": 5e-05,
2518
+ "loss": 1.0428,
2519
+ "step": 207000
2520
+ },
2521
+ {
2522
+ "epoch": 1.52,
2523
+ "learning_rate": 5e-05,
2524
+ "loss": 1.0392,
2525
+ "step": 207500
2526
+ },
2527
+ {
2528
+ "epoch": 1.52,
2529
+ "learning_rate": 5e-05,
2530
+ "loss": 1.0415,
2531
+ "step": 208000
2532
+ },
2533
+ {
2534
+ "epoch": 1.53,
2535
+ "learning_rate": 5e-05,
2536
+ "loss": 1.0426,
2537
+ "step": 208500
2538
+ },
2539
+ {
2540
+ "epoch": 1.53,
2541
+ "learning_rate": 5e-05,
2542
+ "loss": 1.0419,
2543
+ "step": 209000
2544
+ },
2545
+ {
2546
+ "epoch": 1.53,
2547
+ "learning_rate": 5e-05,
2548
+ "loss": 1.0495,
2549
+ "step": 209500
2550
+ },
2551
+ {
2552
+ "epoch": 1.54,
2553
+ "learning_rate": 5e-05,
2554
+ "loss": 1.0492,
2555
+ "step": 210000
2556
+ },
2557
+ {
2558
+ "epoch": 1.54,
2559
+ "learning_rate": 5e-05,
2560
+ "loss": 1.0462,
2561
+ "step": 210500
2562
+ },
2563
+ {
2564
+ "epoch": 1.54,
2565
+ "learning_rate": 5e-05,
2566
+ "loss": 1.0383,
2567
+ "step": 211000
2568
+ },
2569
+ {
2570
+ "epoch": 1.55,
2571
+ "learning_rate": 5e-05,
2572
+ "loss": 1.0402,
2573
+ "step": 211500
2574
+ },
2575
+ {
2576
+ "epoch": 1.55,
2577
+ "learning_rate": 5e-05,
2578
+ "loss": 1.0465,
2579
+ "step": 212000
2580
+ },
2581
+ {
2582
+ "epoch": 1.56,
2583
+ "learning_rate": 5e-05,
2584
+ "loss": 1.044,
2585
+ "step": 212500
2586
+ },
2587
+ {
2588
+ "epoch": 1.56,
2589
+ "learning_rate": 5e-05,
2590
+ "loss": 1.0362,
2591
+ "step": 213000
2592
+ },
2593
+ {
2594
+ "epoch": 1.56,
2595
+ "learning_rate": 5e-05,
2596
+ "loss": 1.0354,
2597
+ "step": 213500
2598
+ },
2599
+ {
2600
+ "epoch": 1.57,
2601
+ "learning_rate": 5e-05,
2602
+ "loss": 1.0389,
2603
+ "step": 214000
2604
+ },
2605
+ {
2606
+ "epoch": 1.57,
2607
+ "learning_rate": 5e-05,
2608
+ "loss": 1.0459,
2609
+ "step": 214500
2610
+ },
2611
+ {
2612
+ "epoch": 1.57,
2613
+ "learning_rate": 5e-05,
2614
+ "loss": 1.0503,
2615
+ "step": 215000
2616
+ },
2617
+ {
2618
+ "epoch": 1.58,
2619
+ "learning_rate": 5e-05,
2620
+ "loss": 1.0489,
2621
+ "step": 215500
2622
+ },
2623
+ {
2624
+ "epoch": 1.58,
2625
+ "learning_rate": 5e-05,
2626
+ "loss": 1.0399,
2627
+ "step": 216000
2628
+ },
2629
+ {
2630
+ "epoch": 1.58,
2631
+ "learning_rate": 5e-05,
2632
+ "loss": 1.0444,
2633
+ "step": 216500
2634
+ },
2635
+ {
2636
+ "epoch": 1.59,
2637
+ "learning_rate": 5e-05,
2638
+ "loss": 1.0413,
2639
+ "step": 217000
2640
+ },
2641
+ {
2642
+ "epoch": 1.59,
2643
+ "learning_rate": 5e-05,
2644
+ "loss": 1.0431,
2645
+ "step": 217500
2646
+ },
2647
+ {
2648
+ "epoch": 1.6,
2649
+ "learning_rate": 5e-05,
2650
+ "loss": 1.0456,
2651
+ "step": 218000
2652
+ },
2653
+ {
2654
+ "epoch": 1.6,
2655
+ "learning_rate": 5e-05,
2656
+ "loss": 1.0424,
2657
+ "step": 218500
2658
+ },
2659
+ {
2660
+ "epoch": 1.6,
2661
+ "learning_rate": 5e-05,
2662
+ "loss": 1.0382,
2663
+ "step": 219000
2664
+ },
2665
+ {
2666
+ "epoch": 1.61,
2667
+ "learning_rate": 5e-05,
2668
+ "loss": 1.0418,
2669
+ "step": 219500
2670
+ },
2671
+ {
2672
+ "epoch": 1.61,
2673
+ "learning_rate": 5e-05,
2674
+ "loss": 1.0369,
2675
+ "step": 220000
2676
+ },
2677
+ {
2678
+ "epoch": 1.61,
2679
+ "learning_rate": 5e-05,
2680
+ "loss": 1.0398,
2681
+ "step": 220500
2682
+ },
2683
+ {
2684
+ "epoch": 1.62,
2685
+ "learning_rate": 5e-05,
2686
+ "loss": 1.032,
2687
+ "step": 221000
2688
+ },
2689
+ {
2690
+ "epoch": 1.62,
2691
+ "learning_rate": 5e-05,
2692
+ "loss": 1.0388,
2693
+ "step": 221500
2694
+ },
2695
+ {
2696
+ "epoch": 1.62,
2697
+ "learning_rate": 5e-05,
2698
+ "loss": 1.042,
2699
+ "step": 222000
2700
+ },
2701
+ {
2702
+ "epoch": 1.63,
2703
+ "learning_rate": 5e-05,
2704
+ "loss": 1.0375,
2705
+ "step": 222500
2706
+ },
2707
+ {
2708
+ "epoch": 1.63,
2709
+ "learning_rate": 5e-05,
2710
+ "loss": 1.0454,
2711
+ "step": 223000
2712
+ },
2713
+ {
2714
+ "epoch": 1.64,
2715
+ "learning_rate": 5e-05,
2716
+ "loss": 1.0387,
2717
+ "step": 223500
2718
+ },
2719
+ {
2720
+ "epoch": 1.64,
2721
+ "learning_rate": 5e-05,
2722
+ "loss": 1.0312,
2723
+ "step": 224000
2724
+ },
2725
+ {
2726
+ "epoch": 1.64,
2727
+ "learning_rate": 5e-05,
2728
+ "loss": 1.0339,
2729
+ "step": 224500
2730
+ },
2731
+ {
2732
+ "epoch": 1.65,
2733
+ "learning_rate": 5e-05,
2734
+ "loss": 1.041,
2735
+ "step": 225000
2736
+ },
2737
+ {
2738
+ "epoch": 1.65,
2739
+ "learning_rate": 5e-05,
2740
+ "loss": 1.035,
2741
+ "step": 225500
2742
+ },
2743
+ {
2744
+ "epoch": 1.65,
2745
+ "learning_rate": 5e-05,
2746
+ "loss": 1.0393,
2747
+ "step": 226000
2748
+ },
2749
+ {
2750
+ "epoch": 1.66,
2751
+ "learning_rate": 5e-05,
2752
+ "loss": 1.0321,
2753
+ "step": 226500
2754
+ },
2755
+ {
2756
+ "epoch": 1.66,
2757
+ "learning_rate": 5e-05,
2758
+ "loss": 1.0369,
2759
+ "step": 227000
2760
+ },
2761
+ {
2762
+ "epoch": 1.67,
2763
+ "learning_rate": 5e-05,
2764
+ "loss": 1.036,
2765
+ "step": 227500
2766
+ },
2767
+ {
2768
+ "epoch": 1.67,
2769
+ "learning_rate": 5e-05,
2770
+ "loss": 1.0383,
2771
+ "step": 228000
2772
+ },
2773
+ {
2774
+ "epoch": 1.67,
2775
+ "learning_rate": 5e-05,
2776
+ "loss": 1.0354,
2777
+ "step": 228500
2778
+ },
2779
+ {
2780
+ "epoch": 1.68,
2781
+ "learning_rate": 5e-05,
2782
+ "loss": 1.03,
2783
+ "step": 229000
2784
+ },
2785
+ {
2786
+ "epoch": 1.68,
2787
+ "learning_rate": 5e-05,
2788
+ "loss": 1.0383,
2789
+ "step": 229500
2790
+ },
2791
+ {
2792
+ "epoch": 1.68,
2793
+ "learning_rate": 5e-05,
2794
+ "loss": 1.0335,
2795
+ "step": 230000
2796
+ },
2797
+ {
2798
+ "epoch": 1.69,
2799
+ "learning_rate": 5e-05,
2800
+ "loss": 1.0317,
2801
+ "step": 230500
2802
+ },
2803
+ {
2804
+ "epoch": 1.69,
2805
+ "learning_rate": 5e-05,
2806
+ "loss": 1.0361,
2807
+ "step": 231000
2808
+ },
2809
+ {
2810
+ "epoch": 1.69,
2811
+ "learning_rate": 5e-05,
2812
+ "loss": 1.0406,
2813
+ "step": 231500
2814
+ },
2815
+ {
2816
+ "epoch": 1.7,
2817
+ "learning_rate": 5e-05,
2818
+ "loss": 1.0328,
2819
+ "step": 232000
2820
+ },
2821
+ {
2822
+ "epoch": 1.7,
2823
+ "learning_rate": 5e-05,
2824
+ "loss": 1.034,
2825
+ "step": 232500
2826
+ },
2827
+ {
2828
+ "epoch": 1.71,
2829
+ "learning_rate": 5e-05,
2830
+ "loss": 1.035,
2831
+ "step": 233000
2832
+ },
2833
+ {
2834
+ "epoch": 1.71,
2835
+ "learning_rate": 5e-05,
2836
+ "loss": 1.0353,
2837
+ "step": 233500
2838
+ },
2839
+ {
2840
+ "epoch": 1.71,
2841
+ "learning_rate": 5e-05,
2842
+ "loss": 1.0288,
2843
+ "step": 234000
2844
+ },
2845
+ {
2846
+ "epoch": 1.72,
2847
+ "learning_rate": 5e-05,
2848
+ "loss": 1.031,
2849
+ "step": 234500
2850
+ },
2851
+ {
2852
+ "epoch": 1.72,
2853
+ "learning_rate": 5e-05,
2854
+ "loss": 1.0332,
2855
+ "step": 235000
2856
+ },
2857
+ {
2858
+ "epoch": 1.72,
2859
+ "learning_rate": 5e-05,
2860
+ "loss": 1.0359,
2861
+ "step": 235500
2862
+ },
2863
+ {
2864
+ "epoch": 1.73,
2865
+ "learning_rate": 5e-05,
2866
+ "loss": 1.033,
2867
+ "step": 236000
2868
+ },
2869
+ {
2870
+ "epoch": 1.73,
2871
+ "learning_rate": 5e-05,
2872
+ "loss": 1.0382,
2873
+ "step": 236500
2874
+ },
2875
+ {
2876
+ "epoch": 1.73,
2877
+ "learning_rate": 5e-05,
2878
+ "loss": 1.038,
2879
+ "step": 237000
2880
+ },
2881
+ {
2882
+ "epoch": 1.74,
2883
+ "learning_rate": 5e-05,
2884
+ "loss": 1.0391,
2885
+ "step": 237500
2886
+ },
2887
+ {
2888
+ "epoch": 1.74,
2889
+ "learning_rate": 5e-05,
2890
+ "loss": 1.0317,
2891
+ "step": 238000
2892
+ },
2893
+ {
2894
+ "epoch": 1.75,
2895
+ "learning_rate": 5e-05,
2896
+ "loss": 1.0423,
2897
+ "step": 238500
2898
+ },
2899
+ {
2900
+ "epoch": 1.75,
2901
+ "learning_rate": 5e-05,
2902
+ "loss": 1.0369,
2903
+ "step": 239000
2904
+ },
2905
+ {
2906
+ "epoch": 1.75,
2907
+ "learning_rate": 5e-05,
2908
+ "loss": 1.0419,
2909
+ "step": 239500
2910
+ },
2911
+ {
2912
+ "epoch": 1.76,
2913
+ "learning_rate": 5e-05,
2914
+ "loss": 1.0388,
2915
+ "step": 240000
2916
+ },
2917
+ {
2918
+ "epoch": 1.76,
2919
+ "learning_rate": 5e-05,
2920
+ "loss": 1.0349,
2921
+ "step": 240500
2922
+ },
2923
+ {
2924
+ "epoch": 1.76,
2925
+ "learning_rate": 5e-05,
2926
+ "loss": 1.034,
2927
+ "step": 241000
2928
+ },
2929
+ {
2930
+ "epoch": 1.77,
2931
+ "learning_rate": 5e-05,
2932
+ "loss": 1.0359,
2933
+ "step": 241500
2934
+ },
2935
+ {
2936
+ "epoch": 1.77,
2937
+ "learning_rate": 5e-05,
2938
+ "loss": 1.0423,
2939
+ "step": 242000
2940
+ },
2941
+ {
2942
+ "epoch": 1.77,
2943
+ "learning_rate": 5e-05,
2944
+ "loss": 1.0341,
2945
+ "step": 242500
2946
+ },
2947
+ {
2948
+ "epoch": 1.78,
2949
+ "learning_rate": 5e-05,
2950
+ "loss": 1.0287,
2951
+ "step": 243000
2952
+ },
2953
+ {
2954
+ "epoch": 1.78,
2955
+ "learning_rate": 5e-05,
2956
+ "loss": 1.0347,
2957
+ "step": 243500
2958
+ },
2959
+ {
2960
+ "epoch": 1.79,
2961
+ "learning_rate": 5e-05,
2962
+ "loss": 1.0304,
2963
+ "step": 244000
2964
+ },
2965
+ {
2966
+ "epoch": 1.79,
2967
+ "learning_rate": 5e-05,
2968
+ "loss": 1.0311,
2969
+ "step": 244500
2970
+ },
2971
+ {
2972
+ "epoch": 1.79,
2973
+ "learning_rate": 5e-05,
2974
+ "loss": 1.0364,
2975
+ "step": 245000
2976
+ },
2977
+ {
2978
+ "epoch": 1.8,
2979
+ "learning_rate": 5e-05,
2980
+ "loss": 1.0271,
2981
+ "step": 245500
2982
+ },
2983
+ {
2984
+ "epoch": 1.8,
2985
+ "learning_rate": 5e-05,
2986
+ "loss": 1.0308,
2987
+ "step": 246000
2988
+ },
2989
+ {
2990
+ "epoch": 1.8,
2991
+ "learning_rate": 5e-05,
2992
+ "loss": 1.0319,
2993
+ "step": 246500
2994
+ },
2995
+ {
2996
+ "epoch": 1.81,
2997
+ "learning_rate": 5e-05,
2998
+ "loss": 1.0345,
2999
+ "step": 247000
3000
+ },
3001
+ {
3002
+ "epoch": 1.81,
3003
+ "learning_rate": 5e-05,
3004
+ "loss": 1.0334,
3005
+ "step": 247500
3006
+ },
3007
+ {
3008
+ "epoch": 1.82,
3009
+ "learning_rate": 5e-05,
3010
+ "loss": 1.0353,
3011
+ "step": 248000
3012
+ },
3013
+ {
3014
+ "epoch": 1.82,
3015
+ "learning_rate": 5e-05,
3016
+ "loss": 1.0274,
3017
+ "step": 248500
3018
+ },
3019
+ {
3020
+ "epoch": 1.82,
3021
+ "learning_rate": 5e-05,
3022
+ "loss": 1.032,
3023
+ "step": 249000
3024
+ },
3025
+ {
3026
+ "epoch": 1.83,
3027
+ "learning_rate": 5e-05,
3028
+ "loss": 1.0336,
3029
+ "step": 249500
3030
+ },
3031
+ {
3032
+ "epoch": 1.83,
3033
+ "learning_rate": 5e-05,
3034
+ "loss": 1.029,
3035
+ "step": 250000
3036
+ },
3037
+ {
3038
+ "epoch": 1.83,
3039
+ "learning_rate": 5e-05,
3040
+ "loss": 1.0278,
3041
+ "step": 250500
3042
+ },
3043
+ {
3044
+ "epoch": 1.84,
3045
+ "learning_rate": 5e-05,
3046
+ "loss": 1.0325,
3047
+ "step": 251000
3048
+ },
3049
+ {
3050
+ "epoch": 1.84,
3051
+ "learning_rate": 5e-05,
3052
+ "loss": 1.0329,
3053
+ "step": 251500
3054
+ },
3055
+ {
3056
+ "epoch": 1.84,
3057
+ "learning_rate": 5e-05,
3058
+ "loss": 1.0312,
3059
+ "step": 252000
3060
+ },
3061
+ {
3062
+ "epoch": 1.85,
3063
+ "learning_rate": 5e-05,
3064
+ "loss": 1.0384,
3065
+ "step": 252500
3066
+ },
3067
+ {
3068
+ "epoch": 1.85,
3069
+ "learning_rate": 5e-05,
3070
+ "loss": 1.0322,
3071
+ "step": 253000
3072
+ },
3073
+ {
3074
+ "epoch": 1.86,
3075
+ "learning_rate": 5e-05,
3076
+ "loss": 1.0296,
3077
+ "step": 253500
3078
+ },
3079
+ {
3080
+ "epoch": 1.86,
3081
+ "learning_rate": 5e-05,
3082
+ "loss": 1.0309,
3083
+ "step": 254000
3084
+ },
3085
+ {
3086
+ "epoch": 1.86,
3087
+ "learning_rate": 5e-05,
3088
+ "loss": 1.0259,
3089
+ "step": 254500
3090
+ },
3091
+ {
3092
+ "epoch": 1.87,
3093
+ "learning_rate": 5e-05,
3094
+ "loss": 1.0361,
3095
+ "step": 255000
3096
+ },
3097
+ {
3098
+ "epoch": 1.87,
3099
+ "learning_rate": 5e-05,
3100
+ "loss": 1.0302,
3101
+ "step": 255500
3102
+ },
3103
+ {
3104
+ "epoch": 1.87,
3105
+ "learning_rate": 5e-05,
3106
+ "loss": 1.0314,
3107
+ "step": 256000
3108
+ },
3109
+ {
3110
+ "epoch": 1.87,
3111
+ "eval_accuracy": 0.7907370033516936,
3112
+ "eval_loss": 1.0169389247894287,
3113
+ "eval_runtime": 32.7988,
3114
+ "eval_samples_per_second": 409.04,
3115
+ "eval_steps_per_second": 8.537,
3116
+ "step": 256000
3117
+ },
3118
+ {
3119
+ "epoch": 1.88,
3120
+ "learning_rate": 5e-05,
3121
+ "loss": 1.0349,
3122
+ "step": 256500
3123
+ },
3124
+ {
3125
+ "epoch": 1.88,
3126
+ "learning_rate": 5e-05,
3127
+ "loss": 1.0369,
3128
+ "step": 257000
3129
+ },
3130
+ {
3131
+ "epoch": 1.88,
3132
+ "learning_rate": 5e-05,
3133
+ "loss": 1.0345,
3134
+ "step": 257500
3135
+ },
3136
+ {
3137
+ "epoch": 1.89,
3138
+ "learning_rate": 5e-05,
3139
+ "loss": 1.0275,
3140
+ "step": 258000
3141
+ },
3142
+ {
3143
+ "epoch": 1.89,
3144
+ "learning_rate": 5e-05,
3145
+ "loss": 1.0342,
3146
+ "step": 258500
3147
+ },
3148
+ {
3149
+ "epoch": 1.9,
3150
+ "learning_rate": 5e-05,
3151
+ "loss": 1.0287,
3152
+ "step": 259000
3153
+ },
3154
+ {
3155
+ "epoch": 1.9,
3156
+ "learning_rate": 5e-05,
3157
+ "loss": 1.0238,
3158
+ "step": 259500
3159
+ },
3160
+ {
3161
+ "epoch": 1.9,
3162
+ "learning_rate": 5e-05,
3163
+ "loss": 1.0258,
3164
+ "step": 260000
3165
+ },
3166
+ {
3167
+ "epoch": 1.91,
3168
+ "learning_rate": 5e-05,
3169
+ "loss": 1.029,
3170
+ "step": 260500
3171
+ },
3172
+ {
3173
+ "epoch": 1.91,
3174
+ "learning_rate": 5e-05,
3175
+ "loss": 1.0277,
3176
+ "step": 261000
3177
+ },
3178
+ {
3179
+ "epoch": 1.91,
3180
+ "learning_rate": 5e-05,
3181
+ "loss": 1.026,
3182
+ "step": 261500
3183
+ },
3184
+ {
3185
+ "epoch": 1.92,
3186
+ "learning_rate": 5e-05,
3187
+ "loss": 1.0265,
3188
+ "step": 262000
3189
+ },
3190
+ {
3191
+ "epoch": 1.92,
3192
+ "learning_rate": 5e-05,
3193
+ "loss": 1.0283,
3194
+ "step": 262500
3195
+ },
3196
+ {
3197
+ "epoch": 1.92,
3198
+ "learning_rate": 5e-05,
3199
+ "loss": 1.0283,
3200
+ "step": 263000
3201
+ },
3202
+ {
3203
+ "epoch": 1.93,
3204
+ "learning_rate": 5e-05,
3205
+ "loss": 1.0268,
3206
+ "step": 263500
3207
+ },
3208
+ {
3209
+ "epoch": 1.93,
3210
+ "learning_rate": 5e-05,
3211
+ "loss": 1.032,
3212
+ "step": 264000
3213
+ },
3214
+ {
3215
+ "epoch": 1.94,
3216
+ "learning_rate": 5e-05,
3217
+ "loss": 1.0233,
3218
+ "step": 264500
3219
+ },
3220
+ {
3221
+ "epoch": 1.94,
3222
+ "learning_rate": 5e-05,
3223
+ "loss": 1.0257,
3224
+ "step": 265000
3225
+ },
3226
+ {
3227
+ "epoch": 1.94,
3228
+ "learning_rate": 5e-05,
3229
+ "loss": 1.0286,
3230
+ "step": 265500
3231
+ },
3232
+ {
3233
+ "epoch": 1.95,
3234
+ "learning_rate": 5e-05,
3235
+ "loss": 1.0226,
3236
+ "step": 266000
3237
+ },
3238
+ {
3239
+ "epoch": 1.95,
3240
+ "learning_rate": 5e-05,
3241
+ "loss": 1.0265,
3242
+ "step": 266500
3243
+ },
3244
+ {
3245
+ "epoch": 1.95,
3246
+ "learning_rate": 5e-05,
3247
+ "loss": 1.0245,
3248
+ "step": 267000
3249
+ },
3250
+ {
3251
+ "epoch": 1.96,
3252
+ "learning_rate": 5e-05,
3253
+ "loss": 1.0231,
3254
+ "step": 267500
3255
+ },
3256
+ {
3257
+ "epoch": 1.96,
3258
+ "learning_rate": 5e-05,
3259
+ "loss": 1.0259,
3260
+ "step": 268000
3261
+ },
3262
+ {
3263
+ "epoch": 1.97,
3264
+ "learning_rate": 5e-05,
3265
+ "loss": 1.0302,
3266
+ "step": 268500
3267
+ },
3268
+ {
3269
+ "epoch": 1.97,
3270
+ "learning_rate": 5e-05,
3271
+ "loss": 1.0316,
3272
+ "step": 269000
3273
+ },
3274
+ {
3275
+ "epoch": 1.97,
3276
+ "learning_rate": 5e-05,
3277
+ "loss": 1.0314,
3278
+ "step": 269500
3279
+ },
3280
+ {
3281
+ "epoch": 1.98,
3282
+ "learning_rate": 5e-05,
3283
+ "loss": 1.0329,
3284
+ "step": 270000
3285
+ },
3286
+ {
3287
+ "epoch": 1.98,
3288
+ "learning_rate": 5e-05,
3289
+ "loss": 1.0275,
3290
+ "step": 270500
3291
+ },
3292
+ {
3293
+ "epoch": 1.98,
3294
+ "learning_rate": 5e-05,
3295
+ "loss": 1.0308,
3296
+ "step": 271000
3297
+ },
3298
+ {
3299
+ "epoch": 1.99,
3300
+ "learning_rate": 5e-05,
3301
+ "loss": 1.022,
3302
+ "step": 271500
3303
+ },
3304
+ {
3305
+ "epoch": 1.99,
3306
+ "learning_rate": 5e-05,
3307
+ "loss": 1.0346,
3308
+ "step": 272000
3309
+ },
3310
+ {
3311
+ "epoch": 1.99,
3312
+ "learning_rate": 5e-05,
3313
+ "loss": 1.0275,
3314
+ "step": 272500
3315
+ },
3316
+ {
3317
+ "epoch": 2.0,
3318
+ "learning_rate": 5e-05,
3319
+ "loss": 1.0283,
3320
+ "step": 273000
3321
+ },
3322
+ {
3323
+ "epoch": 2.0,
3324
+ "learning_rate": 5e-05,
3325
+ "loss": 1.0255,
3326
+ "step": 273500
3327
+ },
3328
+ {
3329
+ "epoch": 2.01,
3330
+ "learning_rate": 5e-05,
3331
+ "loss": 1.0261,
3332
+ "step": 274000
3333
+ },
3334
+ {
3335
+ "epoch": 2.01,
3336
+ "learning_rate": 5e-05,
3337
+ "loss": 1.0219,
3338
+ "step": 274500
3339
+ },
3340
+ {
3341
+ "epoch": 2.01,
3342
+ "learning_rate": 5e-05,
3343
+ "loss": 1.0295,
3344
+ "step": 275000
3345
+ },
3346
+ {
3347
+ "epoch": 2.02,
3348
+ "learning_rate": 5e-05,
3349
+ "loss": 1.0239,
3350
+ "step": 275500
3351
+ },
3352
+ {
3353
+ "epoch": 2.02,
3354
+ "learning_rate": 5e-05,
3355
+ "loss": 1.0221,
3356
+ "step": 276000
3357
+ },
3358
+ {
3359
+ "epoch": 2.02,
3360
+ "learning_rate": 5e-05,
3361
+ "loss": 1.0216,
3362
+ "step": 276500
3363
+ },
3364
+ {
3365
+ "epoch": 2.03,
3366
+ "learning_rate": 5e-05,
3367
+ "loss": 1.0266,
3368
+ "step": 277000
3369
+ },
3370
+ {
3371
+ "epoch": 2.03,
3372
+ "learning_rate": 5e-05,
3373
+ "loss": 1.027,
3374
+ "step": 277500
3375
+ },
3376
+ {
3377
+ "epoch": 2.03,
3378
+ "learning_rate": 5e-05,
3379
+ "loss": 1.0235,
3380
+ "step": 278000
3381
+ },
3382
+ {
3383
+ "epoch": 2.04,
3384
+ "learning_rate": 5e-05,
3385
+ "loss": 1.0259,
3386
+ "step": 278500
3387
+ },
3388
+ {
3389
+ "epoch": 2.04,
3390
+ "learning_rate": 5e-05,
3391
+ "loss": 1.0209,
3392
+ "step": 279000
3393
+ },
3394
+ {
3395
+ "epoch": 2.05,
3396
+ "learning_rate": 5e-05,
3397
+ "loss": 1.0202,
3398
+ "step": 279500
3399
+ },
3400
+ {
3401
+ "epoch": 2.05,
3402
+ "learning_rate": 5e-05,
3403
+ "loss": 1.0295,
3404
+ "step": 280000
3405
+ },
3406
+ {
3407
+ "epoch": 2.05,
3408
+ "learning_rate": 5e-05,
3409
+ "loss": 1.0236,
3410
+ "step": 280500
3411
+ },
3412
+ {
3413
+ "epoch": 2.06,
3414
+ "learning_rate": 5e-05,
3415
+ "loss": 1.0292,
3416
+ "step": 281000
3417
+ },
3418
+ {
3419
+ "epoch": 2.06,
3420
+ "learning_rate": 5e-05,
3421
+ "loss": 1.0226,
3422
+ "step": 281500
3423
+ },
3424
+ {
3425
+ "epoch": 2.06,
3426
+ "learning_rate": 5e-05,
3427
+ "loss": 1.0308,
3428
+ "step": 282000
3429
+ },
3430
+ {
3431
+ "epoch": 2.07,
3432
+ "learning_rate": 5e-05,
3433
+ "loss": 1.0219,
3434
+ "step": 282500
3435
+ },
3436
+ {
3437
+ "epoch": 2.07,
3438
+ "learning_rate": 5e-05,
3439
+ "loss": 1.023,
3440
+ "step": 283000
3441
+ },
3442
+ {
3443
+ "epoch": 2.07,
3444
+ "learning_rate": 5e-05,
3445
+ "loss": 1.0251,
3446
+ "step": 283500
3447
+ },
3448
+ {
3449
+ "epoch": 2.08,
3450
+ "learning_rate": 5e-05,
3451
+ "loss": 1.0301,
3452
+ "step": 284000
3453
+ },
3454
+ {
3455
+ "epoch": 2.08,
3456
+ "learning_rate": 5e-05,
3457
+ "loss": 1.0293,
3458
+ "step": 284500
3459
+ },
3460
+ {
3461
+ "epoch": 2.09,
3462
+ "learning_rate": 5e-05,
3463
+ "loss": 1.0169,
3464
+ "step": 285000
3465
+ },
3466
+ {
3467
+ "epoch": 2.09,
3468
+ "learning_rate": 5e-05,
3469
+ "loss": 1.0204,
3470
+ "step": 285500
3471
+ },
3472
+ {
3473
+ "epoch": 2.09,
3474
+ "learning_rate": 5e-05,
3475
+ "loss": 1.0182,
3476
+ "step": 286000
3477
+ },
3478
+ {
3479
+ "epoch": 2.1,
3480
+ "learning_rate": 5e-05,
3481
+ "loss": 1.016,
3482
+ "step": 286500
3483
+ },
3484
+ {
3485
+ "epoch": 2.1,
3486
+ "learning_rate": 5e-05,
3487
+ "loss": 1.0174,
3488
+ "step": 287000
3489
+ },
3490
+ {
3491
+ "epoch": 2.1,
3492
+ "learning_rate": 5e-05,
3493
+ "loss": 1.0148,
3494
+ "step": 287500
3495
+ },
3496
+ {
3497
+ "epoch": 2.11,
3498
+ "learning_rate": 5e-05,
3499
+ "loss": 1.0224,
3500
+ "step": 288000
3501
+ },
3502
+ {
3503
+ "epoch": 2.11,
3504
+ "learning_rate": 5e-05,
3505
+ "loss": 1.0201,
3506
+ "step": 288500
3507
+ },
3508
+ {
3509
+ "epoch": 2.12,
3510
+ "learning_rate": 5e-05,
3511
+ "loss": 1.0219,
3512
+ "step": 289000
3513
+ },
3514
+ {
3515
+ "epoch": 2.12,
3516
+ "learning_rate": 5e-05,
3517
+ "loss": 1.0172,
3518
+ "step": 289500
3519
+ },
3520
+ {
3521
+ "epoch": 2.12,
3522
+ "learning_rate": 5e-05,
3523
+ "loss": 1.0165,
3524
+ "step": 290000
3525
+ },
3526
+ {
3527
+ "epoch": 2.13,
3528
+ "learning_rate": 5e-05,
3529
+ "loss": 1.0227,
3530
+ "step": 290500
3531
+ },
3532
+ {
3533
+ "epoch": 2.13,
3534
+ "learning_rate": 5e-05,
3535
+ "loss": 1.0217,
3536
+ "step": 291000
3537
+ },
3538
+ {
3539
+ "epoch": 2.13,
3540
+ "learning_rate": 5e-05,
3541
+ "loss": 1.0239,
3542
+ "step": 291500
3543
+ },
3544
+ {
3545
+ "epoch": 2.14,
3546
+ "learning_rate": 5e-05,
3547
+ "loss": 1.0192,
3548
+ "step": 292000
3549
+ },
3550
+ {
3551
+ "epoch": 2.14,
3552
+ "learning_rate": 5e-05,
3553
+ "loss": 1.0196,
3554
+ "step": 292500
3555
+ },
3556
+ {
3557
+ "epoch": 2.14,
3558
+ "learning_rate": 5e-05,
3559
+ "loss": 1.0214,
3560
+ "step": 293000
3561
+ },
3562
+ {
3563
+ "epoch": 2.15,
3564
+ "learning_rate": 5e-05,
3565
+ "loss": 1.0215,
3566
+ "step": 293500
3567
+ },
3568
+ {
3569
+ "epoch": 2.15,
3570
+ "learning_rate": 5e-05,
3571
+ "loss": 1.0198,
3572
+ "step": 294000
3573
+ },
3574
+ {
3575
+ "epoch": 2.16,
3576
+ "learning_rate": 5e-05,
3577
+ "loss": 1.0244,
3578
+ "step": 294500
3579
+ },
3580
+ {
3581
+ "epoch": 2.16,
3582
+ "learning_rate": 5e-05,
3583
+ "loss": 1.0201,
3584
+ "step": 295000
3585
+ },
3586
+ {
3587
+ "epoch": 2.16,
3588
+ "learning_rate": 5e-05,
3589
+ "loss": 1.0123,
3590
+ "step": 295500
3591
+ },
3592
+ {
3593
+ "epoch": 2.17,
3594
+ "learning_rate": 5e-05,
3595
+ "loss": 1.0124,
3596
+ "step": 296000
3597
+ },
3598
+ {
3599
+ "epoch": 2.17,
3600
+ "learning_rate": 5e-05,
3601
+ "loss": 1.0179,
3602
+ "step": 296500
3603
+ },
3604
+ {
3605
+ "epoch": 2.17,
3606
+ "learning_rate": 5e-05,
3607
+ "loss": 1.0177,
3608
+ "step": 297000
3609
+ },
3610
+ {
3611
+ "epoch": 2.18,
3612
+ "learning_rate": 5e-05,
3613
+ "loss": 1.0145,
3614
+ "step": 297500
3615
+ },
3616
+ {
3617
+ "epoch": 2.18,
3618
+ "learning_rate": 5e-05,
3619
+ "loss": 1.0174,
3620
+ "step": 298000
3621
+ },
3622
+ {
3623
+ "epoch": 2.18,
3624
+ "learning_rate": 5e-05,
3625
+ "loss": 1.0144,
3626
+ "step": 298500
3627
+ },
3628
+ {
3629
+ "epoch": 2.19,
3630
+ "learning_rate": 5e-05,
3631
+ "loss": 1.0209,
3632
+ "step": 299000
3633
+ },
3634
+ {
3635
+ "epoch": 2.19,
3636
+ "learning_rate": 5e-05,
3637
+ "loss": 1.0219,
3638
+ "step": 299500
3639
+ },
3640
+ {
3641
+ "epoch": 2.2,
3642
+ "learning_rate": 5e-05,
3643
+ "loss": 1.0247,
3644
+ "step": 300000
3645
+ },
3646
+ {
3647
+ "epoch": 2.2,
3648
+ "learning_rate": 5e-05,
3649
+ "loss": 1.018,
3650
+ "step": 300500
3651
+ },
3652
+ {
3653
+ "epoch": 2.2,
3654
+ "learning_rate": 5e-05,
3655
+ "loss": 1.0213,
3656
+ "step": 301000
3657
+ },
3658
+ {
3659
+ "epoch": 2.21,
3660
+ "learning_rate": 5e-05,
3661
+ "loss": 1.026,
3662
+ "step": 301500
3663
+ },
3664
+ {
3665
+ "epoch": 2.21,
3666
+ "learning_rate": 5e-05,
3667
+ "loss": 1.0211,
3668
+ "step": 302000
3669
+ },
3670
+ {
3671
+ "epoch": 2.21,
3672
+ "learning_rate": 5e-05,
3673
+ "loss": 1.0174,
3674
+ "step": 302500
3675
+ },
3676
+ {
3677
+ "epoch": 2.22,
3678
+ "learning_rate": 5e-05,
3679
+ "loss": 1.0153,
3680
+ "step": 303000
3681
+ },
3682
+ {
3683
+ "epoch": 2.22,
3684
+ "learning_rate": 5e-05,
3685
+ "loss": 1.0168,
3686
+ "step": 303500
3687
+ },
3688
+ {
3689
+ "epoch": 2.22,
3690
+ "learning_rate": 5e-05,
3691
+ "loss": 1.0128,
3692
+ "step": 304000
3693
+ },
3694
+ {
3695
+ "epoch": 2.23,
3696
+ "learning_rate": 5e-05,
3697
+ "loss": 1.0127,
3698
+ "step": 304500
3699
+ },
3700
+ {
3701
+ "epoch": 2.23,
3702
+ "learning_rate": 5e-05,
3703
+ "loss": 1.0144,
3704
+ "step": 305000
3705
+ },
3706
+ {
3707
+ "epoch": 2.24,
3708
+ "learning_rate": 5e-05,
3709
+ "loss": 1.0126,
3710
+ "step": 305500
3711
+ },
3712
+ {
3713
+ "epoch": 2.24,
3714
+ "learning_rate": 5e-05,
3715
+ "loss": 1.0119,
3716
+ "step": 306000
3717
+ },
3718
+ {
3719
+ "epoch": 2.24,
3720
+ "learning_rate": 5e-05,
3721
+ "loss": 1.007,
3722
+ "step": 306500
3723
+ },
3724
+ {
3725
+ "epoch": 2.25,
3726
+ "learning_rate": 5e-05,
3727
+ "loss": 1.0151,
3728
+ "step": 307000
3729
+ },
3730
+ {
3731
+ "epoch": 2.25,
3732
+ "learning_rate": 5e-05,
3733
+ "loss": 1.0186,
3734
+ "step": 307500
3735
+ },
3736
+ {
3737
+ "epoch": 2.25,
3738
+ "learning_rate": 5e-05,
3739
+ "loss": 1.0147,
3740
+ "step": 308000
3741
+ },
3742
+ {
3743
+ "epoch": 2.26,
3744
+ "learning_rate": 5e-05,
3745
+ "loss": 1.0116,
3746
+ "step": 308500
3747
+ },
3748
+ {
3749
+ "epoch": 2.26,
3750
+ "learning_rate": 5e-05,
3751
+ "loss": 1.0095,
3752
+ "step": 309000
3753
+ },
3754
+ {
3755
+ "epoch": 2.27,
3756
+ "learning_rate": 5e-05,
3757
+ "loss": 1.009,
3758
+ "step": 309500
3759
+ },
3760
+ {
3761
+ "epoch": 2.27,
3762
+ "learning_rate": 5e-05,
3763
+ "loss": 1.0073,
3764
+ "step": 310000
3765
+ },
3766
+ {
3767
+ "epoch": 2.27,
3768
+ "learning_rate": 5e-05,
3769
+ "loss": 1.0147,
3770
+ "step": 310500
3771
+ },
3772
+ {
3773
+ "epoch": 2.28,
3774
+ "learning_rate": 5e-05,
3775
+ "loss": 1.0102,
3776
+ "step": 311000
3777
+ },
3778
+ {
3779
+ "epoch": 2.28,
3780
+ "learning_rate": 5e-05,
3781
+ "loss": 1.0063,
3782
+ "step": 311500
3783
+ },
3784
+ {
3785
+ "epoch": 2.28,
3786
+ "learning_rate": 5e-05,
3787
+ "loss": 1.0043,
3788
+ "step": 312000
3789
+ },
3790
+ {
3791
+ "epoch": 2.29,
3792
+ "learning_rate": 5e-05,
3793
+ "loss": 1.0066,
3794
+ "step": 312500
3795
+ },
3796
+ {
3797
+ "epoch": 2.29,
3798
+ "learning_rate": 5e-05,
3799
+ "loss": 1.0122,
3800
+ "step": 313000
3801
+ },
3802
+ {
3803
+ "epoch": 2.29,
3804
+ "learning_rate": 5e-05,
3805
+ "loss": 1.0172,
3806
+ "step": 313500
3807
+ },
3808
+ {
3809
+ "epoch": 2.3,
3810
+ "learning_rate": 5e-05,
3811
+ "loss": 1.0115,
3812
+ "step": 314000
3813
+ },
3814
+ {
3815
+ "epoch": 2.3,
3816
+ "learning_rate": 5e-05,
3817
+ "loss": 1.0113,
3818
+ "step": 314500
3819
+ },
3820
+ {
3821
+ "epoch": 2.31,
3822
+ "learning_rate": 5e-05,
3823
+ "loss": 1.008,
3824
+ "step": 315000
3825
+ },
3826
+ {
3827
+ "epoch": 2.31,
3828
+ "learning_rate": 5e-05,
3829
+ "loss": 1.0094,
3830
+ "step": 315500
3831
+ },
3832
+ {
3833
+ "epoch": 2.31,
3834
+ "learning_rate": 5e-05,
3835
+ "loss": 1.0179,
3836
+ "step": 316000
3837
+ },
3838
+ {
3839
+ "epoch": 2.32,
3840
+ "learning_rate": 5e-05,
3841
+ "loss": 1.0109,
3842
+ "step": 316500
3843
+ },
3844
+ {
3845
+ "epoch": 2.32,
3846
+ "learning_rate": 5e-05,
3847
+ "loss": 1.0127,
3848
+ "step": 317000
3849
+ },
3850
+ {
3851
+ "epoch": 2.32,
3852
+ "learning_rate": 5e-05,
3853
+ "loss": 1.0131,
3854
+ "step": 317500
3855
+ },
3856
+ {
3857
+ "epoch": 2.33,
3858
+ "learning_rate": 5e-05,
3859
+ "loss": 1.0109,
3860
+ "step": 318000
3861
+ },
3862
+ {
3863
+ "epoch": 2.33,
3864
+ "learning_rate": 5e-05,
3865
+ "loss": 1.0117,
3866
+ "step": 318500
3867
+ },
3868
+ {
3869
+ "epoch": 2.33,
3870
+ "learning_rate": 5e-05,
3871
+ "loss": 1.0176,
3872
+ "step": 319000
3873
+ },
3874
+ {
3875
+ "epoch": 2.34,
3876
+ "learning_rate": 5e-05,
3877
+ "loss": 1.009,
3878
+ "step": 319500
3879
+ },
3880
+ {
3881
+ "epoch": 2.34,
3882
+ "learning_rate": 5e-05,
3883
+ "loss": 1.013,
3884
+ "step": 320000
3885
+ },
3886
+ {
3887
+ "epoch": 2.34,
3888
+ "eval_accuracy": 0.794592400932352,
3889
+ "eval_loss": 0.9925464987754822,
3890
+ "eval_runtime": 32.6868,
3891
+ "eval_samples_per_second": 410.441,
3892
+ "eval_steps_per_second": 8.566,
3893
+ "step": 320000
3894
+ },
3895
+ {
3896
+ "epoch": 2.35,
3897
+ "learning_rate": 5e-05,
3898
+ "loss": 1.0104,
3899
+ "step": 320500
3900
+ },
3901
+ {
3902
+ "epoch": 2.35,
3903
+ "learning_rate": 5e-05,
3904
+ "loss": 1.014,
3905
+ "step": 321000
3906
+ },
3907
+ {
3908
+ "epoch": 2.35,
3909
+ "learning_rate": 5e-05,
3910
+ "loss": 1.0136,
3911
+ "step": 321500
3912
+ },
3913
+ {
3914
+ "epoch": 2.36,
3915
+ "learning_rate": 5e-05,
3916
+ "loss": 1.0077,
3917
+ "step": 322000
3918
+ },
3919
+ {
3920
+ "epoch": 2.36,
3921
+ "learning_rate": 5e-05,
3922
+ "loss": 1.0157,
3923
+ "step": 322500
3924
+ },
3925
+ {
3926
+ "epoch": 2.36,
3927
+ "learning_rate": 5e-05,
3928
+ "loss": 1.0133,
3929
+ "step": 323000
3930
+ },
3931
+ {
3932
+ "epoch": 2.37,
3933
+ "learning_rate": 5e-05,
3934
+ "loss": 1.0109,
3935
+ "step": 323500
3936
+ },
3937
+ {
3938
+ "epoch": 2.37,
3939
+ "learning_rate": 5e-05,
3940
+ "loss": 1.0082,
3941
+ "step": 324000
3942
+ },
3943
+ {
3944
+ "epoch": 2.38,
3945
+ "learning_rate": 5e-05,
3946
+ "loss": 1.0085,
3947
+ "step": 324500
3948
+ },
3949
+ {
3950
+ "epoch": 2.38,
3951
+ "learning_rate": 5e-05,
3952
+ "loss": 1.0052,
3953
+ "step": 325000
3954
+ },
3955
+ {
3956
+ "epoch": 2.38,
3957
+ "learning_rate": 5e-05,
3958
+ "loss": 1.0053,
3959
+ "step": 325500
3960
+ },
3961
+ {
3962
+ "epoch": 2.39,
3963
+ "learning_rate": 5e-05,
3964
+ "loss": 1.0065,
3965
+ "step": 326000
3966
+ },
3967
+ {
3968
+ "epoch": 2.39,
3969
+ "learning_rate": 5e-05,
3970
+ "loss": 1.0095,
3971
+ "step": 326500
3972
+ },
3973
+ {
3974
+ "epoch": 2.39,
3975
+ "learning_rate": 5e-05,
3976
+ "loss": 1.0104,
3977
+ "step": 327000
3978
+ },
3979
+ {
3980
+ "epoch": 2.4,
3981
+ "learning_rate": 5e-05,
3982
+ "loss": 1.0096,
3983
+ "step": 327500
3984
+ },
3985
+ {
3986
+ "epoch": 2.4,
3987
+ "learning_rate": 5e-05,
3988
+ "loss": 1.0092,
3989
+ "step": 328000
3990
+ },
3991
+ {
3992
+ "epoch": 2.4,
3993
+ "learning_rate": 5e-05,
3994
+ "loss": 1.0091,
3995
+ "step": 328500
3996
+ },
3997
+ {
3998
+ "epoch": 2.41,
3999
+ "learning_rate": 5e-05,
4000
+ "loss": 1.0063,
4001
+ "step": 329000
4002
+ },
4003
+ {
4004
+ "epoch": 2.41,
4005
+ "learning_rate": 5e-05,
4006
+ "loss": 1.0108,
4007
+ "step": 329500
4008
+ },
4009
+ {
4010
+ "epoch": 2.42,
4011
+ "learning_rate": 5e-05,
4012
+ "loss": 1.0093,
4013
+ "step": 330000
4014
+ },
4015
+ {
4016
+ "epoch": 2.42,
4017
+ "learning_rate": 5e-05,
4018
+ "loss": 1.0087,
4019
+ "step": 330500
4020
+ },
4021
+ {
4022
+ "epoch": 2.42,
4023
+ "learning_rate": 5e-05,
4024
+ "loss": 1.0118,
4025
+ "step": 331000
4026
+ },
4027
+ {
4028
+ "epoch": 2.43,
4029
+ "learning_rate": 5e-05,
4030
+ "loss": 1.011,
4031
+ "step": 331500
4032
+ },
4033
+ {
4034
+ "epoch": 2.43,
4035
+ "learning_rate": 5e-05,
4036
+ "loss": 1.009,
4037
+ "step": 332000
4038
+ },
4039
+ {
4040
+ "epoch": 2.43,
4041
+ "learning_rate": 5e-05,
4042
+ "loss": 1.0128,
4043
+ "step": 332500
4044
+ },
4045
+ {
4046
+ "epoch": 2.44,
4047
+ "learning_rate": 5e-05,
4048
+ "loss": 1.0027,
4049
+ "step": 333000
4050
+ },
4051
+ {
4052
+ "epoch": 2.44,
4053
+ "learning_rate": 5e-05,
4054
+ "loss": 1.008,
4055
+ "step": 333500
4056
+ },
4057
+ {
4058
+ "epoch": 2.44,
4059
+ "learning_rate": 5e-05,
4060
+ "loss": 1.0017,
4061
+ "step": 334000
4062
+ },
4063
+ {
4064
+ "epoch": 2.45,
4065
+ "learning_rate": 5e-05,
4066
+ "loss": 1.0039,
4067
+ "step": 334500
4068
+ },
4069
+ {
4070
+ "epoch": 2.45,
4071
+ "learning_rate": 5e-05,
4072
+ "loss": 1.0058,
4073
+ "step": 335000
4074
+ },
4075
+ {
4076
+ "epoch": 2.46,
4077
+ "learning_rate": 5e-05,
4078
+ "loss": 1.0124,
4079
+ "step": 335500
4080
+ },
4081
+ {
4082
+ "epoch": 2.46,
4083
+ "learning_rate": 5e-05,
4084
+ "loss": 1.0042,
4085
+ "step": 336000
4086
+ },
4087
+ {
4088
+ "epoch": 2.46,
4089
+ "learning_rate": 5e-05,
4090
+ "loss": 1.0025,
4091
+ "step": 336500
4092
+ },
4093
+ {
4094
+ "epoch": 2.47,
4095
+ "learning_rate": 5e-05,
4096
+ "loss": 1.003,
4097
+ "step": 337000
4098
+ },
4099
+ {
4100
+ "epoch": 2.47,
4101
+ "learning_rate": 5e-05,
4102
+ "loss": 1.0034,
4103
+ "step": 337500
4104
+ },
4105
+ {
4106
+ "epoch": 2.47,
4107
+ "learning_rate": 5e-05,
4108
+ "loss": 1.008,
4109
+ "step": 338000
4110
+ },
4111
+ {
4112
+ "epoch": 2.48,
4113
+ "learning_rate": 5e-05,
4114
+ "loss": 1.0097,
4115
+ "step": 338500
4116
+ },
4117
+ {
4118
+ "epoch": 2.48,
4119
+ "learning_rate": 5e-05,
4120
+ "loss": 1.0007,
4121
+ "step": 339000
4122
+ },
4123
+ {
4124
+ "epoch": 2.48,
4125
+ "learning_rate": 5e-05,
4126
+ "loss": 0.9979,
4127
+ "step": 339500
4128
+ },
4129
+ {
4130
+ "epoch": 2.49,
4131
+ "learning_rate": 5e-05,
4132
+ "loss": 1.0044,
4133
+ "step": 340000
4134
+ },
4135
+ {
4136
+ "epoch": 2.49,
4137
+ "learning_rate": 5e-05,
4138
+ "loss": 1.001,
4139
+ "step": 340500
4140
+ },
4141
+ {
4142
+ "epoch": 2.5,
4143
+ "learning_rate": 5e-05,
4144
+ "loss": 1.0032,
4145
+ "step": 341000
4146
+ },
4147
+ {
4148
+ "epoch": 2.5,
4149
+ "learning_rate": 5e-05,
4150
+ "loss": 1.0039,
4151
+ "step": 341500
4152
+ },
4153
+ {
4154
+ "epoch": 2.5,
4155
+ "learning_rate": 5e-05,
4156
+ "loss": 1.0087,
4157
+ "step": 342000
4158
+ },
4159
+ {
4160
+ "epoch": 2.51,
4161
+ "learning_rate": 5e-05,
4162
+ "loss": 1.0036,
4163
+ "step": 342500
4164
+ },
4165
+ {
4166
+ "epoch": 2.51,
4167
+ "learning_rate": 5e-05,
4168
+ "loss": 1.0086,
4169
+ "step": 343000
4170
+ },
4171
+ {
4172
+ "epoch": 2.51,
4173
+ "learning_rate": 5e-05,
4174
+ "loss": 1.0025,
4175
+ "step": 343500
4176
+ },
4177
+ {
4178
+ "epoch": 2.52,
4179
+ "learning_rate": 5e-05,
4180
+ "loss": 1.0081,
4181
+ "step": 344000
4182
+ },
4183
+ {
4184
+ "epoch": 2.52,
4185
+ "learning_rate": 5e-05,
4186
+ "loss": 1.0109,
4187
+ "step": 344500
4188
+ },
4189
+ {
4190
+ "epoch": 2.53,
4191
+ "learning_rate": 5e-05,
4192
+ "loss": 1.0124,
4193
+ "step": 345000
4194
+ },
4195
+ {
4196
+ "epoch": 2.53,
4197
+ "learning_rate": 5e-05,
4198
+ "loss": 1.0137,
4199
+ "step": 345500
4200
+ },
4201
+ {
4202
+ "epoch": 2.53,
4203
+ "learning_rate": 5e-05,
4204
+ "loss": 1.0027,
4205
+ "step": 346000
4206
+ },
4207
+ {
4208
+ "epoch": 2.54,
4209
+ "learning_rate": 5e-05,
4210
+ "loss": 1.0042,
4211
+ "step": 346500
4212
+ },
4213
+ {
4214
+ "epoch": 2.54,
4215
+ "learning_rate": 5e-05,
4216
+ "loss": 1.0065,
4217
+ "step": 347000
4218
+ },
4219
+ {
4220
+ "epoch": 2.54,
4221
+ "learning_rate": 5e-05,
4222
+ "loss": 1.0166,
4223
+ "step": 347500
4224
+ },
4225
+ {
4226
+ "epoch": 2.55,
4227
+ "learning_rate": 5e-05,
4228
+ "loss": 1.0162,
4229
+ "step": 348000
4230
+ },
4231
+ {
4232
+ "epoch": 2.55,
4233
+ "learning_rate": 5e-05,
4234
+ "loss": 1.0116,
4235
+ "step": 348500
4236
+ },
4237
+ {
4238
+ "epoch": 2.55,
4239
+ "learning_rate": 5e-05,
4240
+ "loss": 1.009,
4241
+ "step": 349000
4242
+ },
4243
+ {
4244
+ "epoch": 2.56,
4245
+ "learning_rate": 5e-05,
4246
+ "loss": 1.0118,
4247
+ "step": 349500
4248
+ },
4249
+ {
4250
+ "epoch": 2.56,
4251
+ "learning_rate": 5e-05,
4252
+ "loss": 1.0088,
4253
+ "step": 350000
4254
+ },
4255
+ {
4256
+ "epoch": 2.57,
4257
+ "learning_rate": 5e-05,
4258
+ "loss": 1.002,
4259
+ "step": 350500
4260
+ },
4261
+ {
4262
+ "epoch": 2.57,
4263
+ "learning_rate": 5e-05,
4264
+ "loss": 1.009,
4265
+ "step": 351000
4266
+ },
4267
+ {
4268
+ "epoch": 2.57,
4269
+ "learning_rate": 5e-05,
4270
+ "loss": 1.0109,
4271
+ "step": 351500
4272
+ },
4273
+ {
4274
+ "epoch": 2.58,
4275
+ "learning_rate": 5e-05,
4276
+ "loss": 1.0129,
4277
+ "step": 352000
4278
+ },
4279
+ {
4280
+ "epoch": 2.58,
4281
+ "learning_rate": 5e-05,
4282
+ "loss": 1.0085,
4283
+ "step": 352500
4284
+ },
4285
+ {
4286
+ "epoch": 2.58,
4287
+ "learning_rate": 5e-05,
4288
+ "loss": 1.0026,
4289
+ "step": 353000
4290
+ },
4291
+ {
4292
+ "epoch": 2.59,
4293
+ "learning_rate": 5e-05,
4294
+ "loss": 1.0009,
4295
+ "step": 353500
4296
+ },
4297
+ {
4298
+ "epoch": 2.59,
4299
+ "learning_rate": 5e-05,
4300
+ "loss": 1.0081,
4301
+ "step": 354000
4302
+ },
4303
+ {
4304
+ "epoch": 2.59,
4305
+ "learning_rate": 5e-05,
4306
+ "loss": 1.0086,
4307
+ "step": 354500
4308
+ },
4309
+ {
4310
+ "epoch": 2.6,
4311
+ "learning_rate": 5e-05,
4312
+ "loss": 1.0068,
4313
+ "step": 355000
4314
+ },
4315
+ {
4316
+ "epoch": 2.6,
4317
+ "learning_rate": 5e-05,
4318
+ "loss": 1.0118,
4319
+ "step": 355500
4320
+ },
4321
+ {
4322
+ "epoch": 2.61,
4323
+ "learning_rate": 5e-05,
4324
+ "loss": 1.0039,
4325
+ "step": 356000
4326
+ },
4327
+ {
4328
+ "epoch": 2.61,
4329
+ "learning_rate": 5e-05,
4330
+ "loss": 1.0015,
4331
+ "step": 356500
4332
+ },
4333
+ {
4334
+ "epoch": 2.61,
4335
+ "learning_rate": 5e-05,
4336
+ "loss": 1.0046,
4337
+ "step": 357000
4338
+ },
4339
+ {
4340
+ "epoch": 2.62,
4341
+ "learning_rate": 5e-05,
4342
+ "loss": 1.0015,
4343
+ "step": 357500
4344
+ },
4345
+ {
4346
+ "epoch": 2.62,
4347
+ "learning_rate": 5e-05,
4348
+ "loss": 1.0091,
4349
+ "step": 358000
4350
+ },
4351
+ {
4352
+ "epoch": 2.62,
4353
+ "learning_rate": 5e-05,
4354
+ "loss": 1.0128,
4355
+ "step": 358500
4356
+ },
4357
+ {
4358
+ "epoch": 2.63,
4359
+ "learning_rate": 5e-05,
4360
+ "loss": 1.005,
4361
+ "step": 359000
4362
+ },
4363
+ {
4364
+ "epoch": 2.63,
4365
+ "learning_rate": 5e-05,
4366
+ "loss": 1.0034,
4367
+ "step": 359500
4368
+ },
4369
+ {
4370
+ "epoch": 2.63,
4371
+ "learning_rate": 5e-05,
4372
+ "loss": 1.0061,
4373
+ "step": 360000
4374
+ },
4375
+ {
4376
+ "epoch": 2.64,
4377
+ "learning_rate": 5e-05,
4378
+ "loss": 1.0004,
4379
+ "step": 360500
4380
+ },
4381
+ {
4382
+ "epoch": 2.64,
4383
+ "learning_rate": 5e-05,
4384
+ "loss": 1.009,
4385
+ "step": 361000
4386
+ },
4387
+ {
4388
+ "epoch": 2.65,
4389
+ "learning_rate": 5e-05,
4390
+ "loss": 1.0038,
4391
+ "step": 361500
4392
+ },
4393
+ {
4394
+ "epoch": 2.65,
4395
+ "learning_rate": 5e-05,
4396
+ "loss": 1.0017,
4397
+ "step": 362000
4398
+ },
4399
+ {
4400
+ "epoch": 2.65,
4401
+ "learning_rate": 5e-05,
4402
+ "loss": 1.0012,
4403
+ "step": 362500
4404
+ },
4405
+ {
4406
+ "epoch": 2.66,
4407
+ "learning_rate": 5e-05,
4408
+ "loss": 1.0019,
4409
+ "step": 363000
4410
+ },
4411
+ {
4412
+ "epoch": 2.66,
4413
+ "learning_rate": 5e-05,
4414
+ "loss": 0.9962,
4415
+ "step": 363500
4416
+ },
4417
+ {
4418
+ "epoch": 2.66,
4419
+ "learning_rate": 5e-05,
4420
+ "loss": 1.0034,
4421
+ "step": 364000
4422
+ },
4423
+ {
4424
+ "epoch": 2.67,
4425
+ "learning_rate": 5e-05,
4426
+ "loss": 1.0062,
4427
+ "step": 364500
4428
+ },
4429
+ {
4430
+ "epoch": 2.67,
4431
+ "learning_rate": 5e-05,
4432
+ "loss": 1.0059,
4433
+ "step": 365000
4434
+ },
4435
+ {
4436
+ "epoch": 2.68,
4437
+ "learning_rate": 5e-05,
4438
+ "loss": 1.0079,
4439
+ "step": 365500
4440
+ },
4441
+ {
4442
+ "epoch": 2.68,
4443
+ "learning_rate": 5e-05,
4444
+ "loss": 1.0083,
4445
+ "step": 366000
4446
+ },
4447
+ {
4448
+ "epoch": 2.68,
4449
+ "learning_rate": 5e-05,
4450
+ "loss": 1.0113,
4451
+ "step": 366500
4452
+ },
4453
+ {
4454
+ "epoch": 2.69,
4455
+ "learning_rate": 5e-05,
4456
+ "loss": 1.0042,
4457
+ "step": 367000
4458
+ },
4459
+ {
4460
+ "epoch": 2.69,
4461
+ "learning_rate": 5e-05,
4462
+ "loss": 1.0047,
4463
+ "step": 367500
4464
+ },
4465
+ {
4466
+ "epoch": 2.69,
4467
+ "learning_rate": 5e-05,
4468
+ "loss": 1.0012,
4469
+ "step": 368000
4470
+ },
4471
+ {
4472
+ "epoch": 2.7,
4473
+ "learning_rate": 5e-05,
4474
+ "loss": 1.0021,
4475
+ "step": 368500
4476
+ },
4477
+ {
4478
+ "epoch": 2.7,
4479
+ "learning_rate": 5e-05,
4480
+ "loss": 1.0043,
4481
+ "step": 369000
4482
+ },
4483
+ {
4484
+ "epoch": 2.7,
4485
+ "learning_rate": 5e-05,
4486
+ "loss": 1.0016,
4487
+ "step": 369500
4488
+ },
4489
+ {
4490
+ "epoch": 2.71,
4491
+ "learning_rate": 5e-05,
4492
+ "loss": 1.0022,
4493
+ "step": 370000
4494
+ },
4495
+ {
4496
+ "epoch": 2.71,
4497
+ "learning_rate": 5e-05,
4498
+ "loss": 0.9942,
4499
+ "step": 370500
4500
+ },
4501
+ {
4502
+ "epoch": 2.72,
4503
+ "learning_rate": 5e-05,
4504
+ "loss": 1.0004,
4505
+ "step": 371000
4506
+ },
4507
+ {
4508
+ "epoch": 2.72,
4509
+ "learning_rate": 5e-05,
4510
+ "loss": 0.998,
4511
+ "step": 371500
4512
+ },
4513
+ {
4514
+ "epoch": 2.72,
4515
+ "learning_rate": 5e-05,
4516
+ "loss": 0.9981,
4517
+ "step": 372000
4518
+ },
4519
+ {
4520
+ "epoch": 2.73,
4521
+ "learning_rate": 5e-05,
4522
+ "loss": 0.9983,
4523
+ "step": 372500
4524
+ },
4525
+ {
4526
+ "epoch": 2.73,
4527
+ "learning_rate": 5e-05,
4528
+ "loss": 0.996,
4529
+ "step": 373000
4530
+ },
4531
+ {
4532
+ "epoch": 2.73,
4533
+ "learning_rate": 5e-05,
4534
+ "loss": 0.9905,
4535
+ "step": 373500
4536
+ },
4537
+ {
4538
+ "epoch": 2.74,
4539
+ "learning_rate": 5e-05,
4540
+ "loss": 0.9982,
4541
+ "step": 374000
4542
+ },
4543
+ {
4544
+ "epoch": 2.74,
4545
+ "learning_rate": 5e-05,
4546
+ "loss": 0.992,
4547
+ "step": 374500
4548
+ },
4549
+ {
4550
+ "epoch": 2.74,
4551
+ "learning_rate": 5e-05,
4552
+ "loss": 1.0035,
4553
+ "step": 375000
4554
+ },
4555
+ {
4556
+ "epoch": 2.75,
4557
+ "learning_rate": 5e-05,
4558
+ "loss": 0.9998,
4559
+ "step": 375500
4560
+ },
4561
+ {
4562
+ "epoch": 2.75,
4563
+ "learning_rate": 5e-05,
4564
+ "loss": 0.9997,
4565
+ "step": 376000
4566
+ },
4567
+ {
4568
+ "epoch": 2.76,
4569
+ "learning_rate": 5e-05,
4570
+ "loss": 0.995,
4571
+ "step": 376500
4572
+ },
4573
+ {
4574
+ "epoch": 2.76,
4575
+ "learning_rate": 5e-05,
4576
+ "loss": 0.9972,
4577
+ "step": 377000
4578
+ },
4579
+ {
4580
+ "epoch": 2.76,
4581
+ "learning_rate": 5e-05,
4582
+ "loss": 0.9965,
4583
+ "step": 377500
4584
+ },
4585
+ {
4586
+ "epoch": 2.77,
4587
+ "learning_rate": 5e-05,
4588
+ "loss": 0.9999,
4589
+ "step": 378000
4590
+ },
4591
+ {
4592
+ "epoch": 2.77,
4593
+ "learning_rate": 5e-05,
4594
+ "loss": 0.9981,
4595
+ "step": 378500
4596
+ },
4597
+ {
4598
+ "epoch": 2.77,
4599
+ "learning_rate": 5e-05,
4600
+ "loss": 1.0018,
4601
+ "step": 379000
4602
+ },
4603
+ {
4604
+ "epoch": 2.78,
4605
+ "learning_rate": 5e-05,
4606
+ "loss": 0.9977,
4607
+ "step": 379500
4608
+ },
4609
+ {
4610
+ "epoch": 2.78,
4611
+ "learning_rate": 5e-05,
4612
+ "loss": 0.9947,
4613
+ "step": 380000
4614
+ },
4615
+ {
4616
+ "epoch": 2.78,
4617
+ "learning_rate": 5e-05,
4618
+ "loss": 0.9954,
4619
+ "step": 380500
4620
+ },
4621
+ {
4622
+ "epoch": 2.79,
4623
+ "learning_rate": 5e-05,
4624
+ "loss": 0.9954,
4625
+ "step": 381000
4626
+ },
4627
+ {
4628
+ "epoch": 2.79,
4629
+ "learning_rate": 5e-05,
4630
+ "loss": 1.002,
4631
+ "step": 381500
4632
+ },
4633
+ {
4634
+ "epoch": 2.8,
4635
+ "learning_rate": 5e-05,
4636
+ "loss": 0.9944,
4637
+ "step": 382000
4638
+ },
4639
+ {
4640
+ "epoch": 2.8,
4641
+ "learning_rate": 5e-05,
4642
+ "loss": 0.9964,
4643
+ "step": 382500
4644
+ },
4645
+ {
4646
+ "epoch": 2.8,
4647
+ "learning_rate": 5e-05,
4648
+ "loss": 1.0025,
4649
+ "step": 383000
4650
+ },
4651
+ {
4652
+ "epoch": 2.81,
4653
+ "learning_rate": 5e-05,
4654
+ "loss": 1.0045,
4655
+ "step": 383500
4656
+ },
4657
+ {
4658
+ "epoch": 2.81,
4659
+ "learning_rate": 5e-05,
4660
+ "loss": 0.9984,
4661
+ "step": 384000
4662
+ },
4663
+ {
4664
+ "epoch": 2.81,
4665
+ "eval_accuracy": 0.7979610372060427,
4666
+ "eval_loss": 0.9764020442962646,
4667
+ "eval_runtime": 32.733,
4668
+ "eval_samples_per_second": 409.862,
4669
+ "eval_steps_per_second": 8.554,
4670
+ "step": 384000
4671
+ },
4672
+ {
4673
+ "epoch": 2.81,
4674
+ "learning_rate": 5e-05,
4675
+ "loss": 0.9979,
4676
+ "step": 384500
4677
+ },
4678
+ {
4679
+ "epoch": 2.82,
4680
+ "learning_rate": 5e-05,
4681
+ "loss": 0.991,
4682
+ "step": 385000
4683
+ },
4684
+ {
4685
+ "epoch": 2.82,
4686
+ "learning_rate": 5e-05,
4687
+ "loss": 0.9959,
4688
+ "step": 385500
4689
+ },
4690
+ {
4691
+ "epoch": 2.83,
4692
+ "learning_rate": 5e-05,
4693
+ "loss": 0.999,
4694
+ "step": 386000
4695
+ },
4696
+ {
4697
+ "epoch": 2.83,
4698
+ "learning_rate": 5e-05,
4699
+ "loss": 0.9982,
4700
+ "step": 386500
4701
+ },
4702
+ {
4703
+ "epoch": 2.83,
4704
+ "learning_rate": 5e-05,
4705
+ "loss": 1.0048,
4706
+ "step": 387000
4707
+ },
4708
+ {
4709
+ "epoch": 2.84,
4710
+ "learning_rate": 5e-05,
4711
+ "loss": 0.9964,
4712
+ "step": 387500
4713
+ },
4714
+ {
4715
+ "epoch": 2.84,
4716
+ "learning_rate": 5e-05,
4717
+ "loss": 0.9956,
4718
+ "step": 388000
4719
+ },
4720
+ {
4721
+ "epoch": 2.84,
4722
+ "learning_rate": 5e-05,
4723
+ "loss": 0.997,
4724
+ "step": 388500
4725
+ },
4726
+ {
4727
+ "epoch": 2.85,
4728
+ "learning_rate": 5e-05,
4729
+ "loss": 0.9915,
4730
+ "step": 389000
4731
+ },
4732
+ {
4733
+ "epoch": 2.85,
4734
+ "learning_rate": 5e-05,
4735
+ "loss": 0.9912,
4736
+ "step": 389500
4737
+ },
4738
+ {
4739
+ "epoch": 2.85,
4740
+ "learning_rate": 5e-05,
4741
+ "loss": 0.9944,
4742
+ "step": 390000
4743
+ },
4744
+ {
4745
+ "epoch": 2.86,
4746
+ "learning_rate": 5e-05,
4747
+ "loss": 0.9965,
4748
+ "step": 390500
4749
+ },
4750
+ {
4751
+ "epoch": 2.86,
4752
+ "learning_rate": 5e-05,
4753
+ "loss": 0.9899,
4754
+ "step": 391000
4755
+ },
4756
+ {
4757
+ "epoch": 2.87,
4758
+ "learning_rate": 5e-05,
4759
+ "loss": 1.0008,
4760
+ "step": 391500
4761
+ },
4762
+ {
4763
+ "epoch": 2.87,
4764
+ "learning_rate": 5e-05,
4765
+ "loss": 0.9898,
4766
+ "step": 392000
4767
+ },
4768
+ {
4769
+ "epoch": 2.87,
4770
+ "learning_rate": 5e-05,
4771
+ "loss": 0.9999,
4772
+ "step": 392500
4773
+ },
4774
+ {
4775
+ "epoch": 2.88,
4776
+ "learning_rate": 5e-05,
4777
+ "loss": 0.9974,
4778
+ "step": 393000
4779
+ },
4780
+ {
4781
+ "epoch": 2.88,
4782
+ "learning_rate": 5e-05,
4783
+ "loss": 0.9946,
4784
+ "step": 393500
4785
+ },
4786
+ {
4787
+ "epoch": 2.88,
4788
+ "learning_rate": 5e-05,
4789
+ "loss": 0.996,
4790
+ "step": 394000
4791
+ },
4792
+ {
4793
+ "epoch": 2.89,
4794
+ "learning_rate": 5e-05,
4795
+ "loss": 0.9921,
4796
+ "step": 394500
4797
+ },
4798
+ {
4799
+ "epoch": 2.89,
4800
+ "learning_rate": 5e-05,
4801
+ "loss": 0.9977,
4802
+ "step": 395000
4803
+ },
4804
+ {
4805
+ "epoch": 2.89,
4806
+ "learning_rate": 5e-05,
4807
+ "loss": 1.0009,
4808
+ "step": 395500
4809
+ },
4810
+ {
4811
+ "epoch": 2.9,
4812
+ "learning_rate": 5e-05,
4813
+ "loss": 0.9879,
4814
+ "step": 396000
4815
+ },
4816
+ {
4817
+ "epoch": 2.9,
4818
+ "learning_rate": 5e-05,
4819
+ "loss": 0.999,
4820
+ "step": 396500
4821
+ },
4822
+ {
4823
+ "epoch": 2.91,
4824
+ "learning_rate": 5e-05,
4825
+ "loss": 0.9994,
4826
+ "step": 397000
4827
+ },
4828
+ {
4829
+ "epoch": 2.91,
4830
+ "learning_rate": 5e-05,
4831
+ "loss": 0.9971,
4832
+ "step": 397500
4833
+ },
4834
+ {
4835
+ "epoch": 2.91,
4836
+ "learning_rate": 5e-05,
4837
+ "loss": 0.9913,
4838
+ "step": 398000
4839
+ },
4840
+ {
4841
+ "epoch": 2.92,
4842
+ "learning_rate": 5e-05,
4843
+ "loss": 0.9971,
4844
+ "step": 398500
4845
+ },
4846
+ {
4847
+ "epoch": 2.92,
4848
+ "learning_rate": 5e-05,
4849
+ "loss": 1.003,
4850
+ "step": 399000
4851
+ },
4852
+ {
4853
+ "epoch": 2.92,
4854
+ "learning_rate": 5e-05,
4855
+ "loss": 0.9903,
4856
+ "step": 399500
4857
+ },
4858
+ {
4859
+ "epoch": 2.93,
4860
+ "learning_rate": 5e-05,
4861
+ "loss": 0.9929,
4862
+ "step": 400000
4863
+ },
4864
+ {
4865
+ "epoch": 2.93,
4866
+ "learning_rate": 5e-05,
4867
+ "loss": 0.9959,
4868
+ "step": 400500
4869
+ },
4870
+ {
4871
+ "epoch": 2.93,
4872
+ "learning_rate": 5e-05,
4873
+ "loss": 0.9972,
4874
+ "step": 401000
4875
+ },
4876
+ {
4877
+ "epoch": 2.94,
4878
+ "learning_rate": 5e-05,
4879
+ "loss": 0.9964,
4880
+ "step": 401500
4881
+ },
4882
+ {
4883
+ "epoch": 2.94,
4884
+ "learning_rate": 5e-05,
4885
+ "loss": 0.9963,
4886
+ "step": 402000
4887
+ },
4888
+ {
4889
+ "epoch": 2.95,
4890
+ "learning_rate": 5e-05,
4891
+ "loss": 0.9954,
4892
+ "step": 402500
4893
+ },
4894
+ {
4895
+ "epoch": 2.95,
4896
+ "learning_rate": 5e-05,
4897
+ "loss": 0.993,
4898
+ "step": 403000
4899
+ },
4900
+ {
4901
+ "epoch": 2.95,
4902
+ "learning_rate": 5e-05,
4903
+ "loss": 0.9961,
4904
+ "step": 403500
4905
+ },
4906
+ {
4907
+ "epoch": 2.96,
4908
+ "learning_rate": 5e-05,
4909
+ "loss": 0.9967,
4910
+ "step": 404000
4911
+ },
4912
+ {
4913
+ "epoch": 2.96,
4914
+ "learning_rate": 5e-05,
4915
+ "loss": 1.0006,
4916
+ "step": 404500
4917
+ },
4918
+ {
4919
+ "epoch": 2.96,
4920
+ "learning_rate": 5e-05,
4921
+ "loss": 0.9901,
4922
+ "step": 405000
4923
+ },
4924
+ {
4925
+ "epoch": 2.97,
4926
+ "learning_rate": 5e-05,
4927
+ "loss": 0.9988,
4928
+ "step": 405500
4929
+ },
4930
+ {
4931
+ "epoch": 2.97,
4932
+ "learning_rate": 5e-05,
4933
+ "loss": 1.0004,
4934
+ "step": 406000
4935
+ },
4936
+ {
4937
+ "epoch": 2.98,
4938
+ "learning_rate": 5e-05,
4939
+ "loss": 1.0004,
4940
+ "step": 406500
4941
+ },
4942
+ {
4943
+ "epoch": 2.98,
4944
+ "learning_rate": 5e-05,
4945
+ "loss": 0.991,
4946
+ "step": 407000
4947
+ },
4948
+ {
4949
+ "epoch": 2.98,
4950
+ "learning_rate": 5e-05,
4951
+ "loss": 0.9978,
4952
+ "step": 407500
4953
+ },
4954
+ {
4955
+ "epoch": 2.99,
4956
+ "learning_rate": 5e-05,
4957
+ "loss": 0.9928,
4958
+ "step": 408000
4959
+ },
4960
+ {
4961
+ "epoch": 2.99,
4962
+ "learning_rate": 5e-05,
4963
+ "loss": 1.0039,
4964
+ "step": 408500
4965
+ },
4966
+ {
4967
+ "epoch": 2.99,
4968
+ "learning_rate": 5e-05,
4969
+ "loss": 0.993,
4970
+ "step": 409000
4971
+ },
4972
+ {
4973
+ "epoch": 3.0,
4974
+ "learning_rate": 5e-05,
4975
+ "loss": 0.9899,
4976
+ "step": 409500
4977
+ },
4978
+ {
4979
+ "epoch": 3.0,
4980
+ "step": 409890,
4981
+ "total_flos": 3.6701198802265047e+19,
4982
+ "train_loss": 0.06289897245738281,
4983
+ "train_runtime": 14736.1316,
4984
+ "train_samples_per_second": 2670.259,
4985
+ "train_steps_per_second": 27.815
4986
+ }
4987
+ ],
4988
+ "max_steps": 409890,
4989
+ "num_train_epochs": 3,
4990
+ "total_flos": 3.6701198802265047e+19,
4991
+ "trial_name": null,
4992
+ "trial_params": null
4993
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3f74619408913b68950ccea8d7c1522769d6133dcde2c89ad160c2f74a8be92c
3
+ size 5563