Hugging Face's logo

cjvt
/
muril-en-hi-codemixed

Fill-Mask
Transformers
PyTorch
English
Hindi
multilingual
bert
generated_from_trainer
Inference Endpoints
Model card Files Community
1
Matej U commited on
Commit
3183e00
•
1 Parent(s): 5748b6d

first version

Browse files
Files changed (11) hide show
  1. README.md +21 -1
  2. all_results.json +15 -0
  3. config.json +27 -0
  4. eval_results.json +10 -0
  5. pytorch_model.bin +3 -0
  6. special_tokens_map.json +19 -0
  7. tokenizer.json +0 -0
  8. tokenizer_config.json +24 -0
  9. train_results.json +8 -0
  10. trainer_state.json +5242 -0
  11. training_args.bin +3 -0
README.md CHANGED
@@ -1,3 +1,23 @@
1
  ---
2
- license: cc-by-sa-4.0
 
 
 
 
 
3
  ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  ---
2
+ tags:
3
+ - generated_from_trainer
4
+ language:
5
+ - en
6
+ - hi
7
+ licence: cc-by-sa-4.0
8
  ---
9
+
10
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
11
+ should probably proofread and complete it, then remove this comment. -->
12
+
13
+
14
+ # muril-en-hi-codemixed
15
+
16
+ muril-en-hi-codemixed is a masked language model, based on the [MuRIL](https://huggingface.co/google/muril-base-cased) multilingual model.
17
+
18
+ muril-en-hi-codemixed replaces the tokenizer, vocabulary and the embeddings layer of the MuRIL model.
19
+ The tokenizer and vocabulary used are the same as in the [roberta-en-hi-codemixed](https://huggingface.co/cjvt/roberta-en-hi-codemixed) model.
20
+ The new embedding weights were initialized from the MuRIL embeddings.
21
+
22
+ The new muril-en-hi-codemixed model was further pre-trained for two epochs on the same codemixed English and Hindi corpora
23
+ as the [roberta-en-hi-codemixed](https://huggingface.co/cjvt/roberta-en-hi-codemixed) model.
all_results.json ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 2.0,
3
+ "eval_accuracy": 0.6136725432898202,
4
+ "eval_loss": 1.900390625,
5
+ "eval_runtime": 4886.2339,
6
+ "eval_samples": 558499,
7
+ "eval_samples_per_second": 114.301,
8
+ "eval_steps_per_second": 0.893,
9
+ "perplexity": 6.688506629955475,
10
+ "train_loss": 0.02994945877374942,
11
+ "train_runtime": 24111.6962,
12
+ "train_samples": 110444789,
13
+ "train_samples_per_second": 9161.097,
14
+ "train_steps_per_second": 35.786
15
+ }
config.json ADDED
@@ -0,0 +1,27 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "./enhi-from-muril",
3
+ "architectures": [
4
+ "BertForMaskedLM"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.1,
7
+ "classifier_dropout": null,
8
+ "embedding_size": 768,
9
+ "gradient_checkpointing": false,
10
+ "hidden_act": "gelu",
11
+ "hidden_dropout_prob": 0.1,
12
+ "hidden_size": 768,
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 3072,
15
+ "layer_norm_eps": 1e-12,
16
+ "max_position_embeddings": 512,
17
+ "model_type": "bert",
18
+ "num_attention_heads": 12,
19
+ "num_hidden_layers": 12,
20
+ "pad_token_id": 0,
21
+ "position_embedding_type": "absolute",
22
+ "torch_dtype": "float16",
23
+ "transformers_version": "4.22.0.dev0",
24
+ "type_vocab_size": 2,
25
+ "use_cache": true,
26
+ "vocab_size": 40005
27
+ }
eval_results.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 2.0,
3
+ "eval_accuracy": 0.6136725432898202,
4
+ "eval_loss": 1.900390625,
5
+ "eval_runtime": 4886.2339,
6
+ "eval_samples": 558499,
7
+ "eval_samples_per_second": 114.301,
8
+ "eval_steps_per_second": 0.893,
9
+ "perplexity": 6.688506629955475
10
+ }
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:264d3ef4865c70c2e1ebdc80a4a7fbc03d1c85038c9172f1dca553227f72bc3c
3
+ size 233646335
special_tokens_map.json ADDED
@@ -0,0 +1,19 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "additional_special_tokens": [
3
+ "<s>NOTUSED",
4
+ "</s>NOTUSED"
5
+ ],
6
+ "bos_token": "<s>",
7
+ "cls_token": "<s>",
8
+ "eos_token": "</s>",
9
+ "mask_token": {
10
+ "content": "<mask>",
11
+ "lstrip": true,
12
+ "normalized": true,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "pad_token": "<pad>",
17
+ "sep_token": "</s>",
18
+ "unk_token": "<unk>"
19
+ }
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer_config.json ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "additional_special_tokens": [
3
+ "<s>NOTUSED",
4
+ "</s>NOTUSED"
5
+ ],
6
+ "bos_token": "<s>",
7
+ "cls_token": "<s>",
8
+ "eos_token": "</s>",
9
+ "mask_token": {
10
+ "__type": "AddedToken",
11
+ "content": "<mask>",
12
+ "lstrip": true,
13
+ "normalized": true,
14
+ "rstrip": false,
15
+ "single_word": false
16
+ },
17
+ "name_or_path": "./enhi-from-muril",
18
+ "pad_token": "<pad>",
19
+ "sep_token": "</s>",
20
+ "sp_model_kwargs": {},
21
+ "special_tokens_map_file": null,
22
+ "tokenizer_class": "CamembertTokenizer",
23
+ "unk_token": "<unk>"
24
+ }
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 2.0,
3
+ "train_loss": 0.02994945877374942,
4
+ "train_runtime": 24111.6962,
5
+ "train_samples": 110444789,
6
+ "train_samples_per_second": 9161.097,
7
+ "train_steps_per_second": 35.786
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,5242 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 2.0,
5
+ "global_step": 862850,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.0,
12
+ "learning_rate": 4.112668990819203e-05,
13
+ "loss": 8.0034,
14
+ "step": 1000
15
+ },
16
+ {
17
+ "epoch": 0.0,
18
+ "learning_rate": 4.533733258325905e-05,
19
+ "loss": 7.5343,
20
+ "step": 2000
21
+ },
22
+ {
23
+ "epoch": 0.01,
24
+ "learning_rate": 4.7778615089460934e-05,
25
+ "loss": 7.2281,
26
+ "step": 3000
27
+ },
28
+ {
29
+ "epoch": 0.01,
30
+ "learning_rate": 4.9509713131658946e-05,
31
+ "loss": 6.5225,
32
+ "step": 4000
33
+ },
34
+ {
35
+ "epoch": 0.01,
36
+ "learning_rate": 5e-05,
37
+ "loss": 6.0666,
38
+ "step": 5000
39
+ },
40
+ {
41
+ "epoch": 0.01,
42
+ "learning_rate": 5e-05,
43
+ "loss": 5.7596,
44
+ "step": 6000
45
+ },
46
+ {
47
+ "epoch": 0.02,
48
+ "learning_rate": 5e-05,
49
+ "loss": 5.5413,
50
+ "step": 7000
51
+ },
52
+ {
53
+ "epoch": 0.02,
54
+ "learning_rate": 5e-05,
55
+ "loss": 5.3661,
56
+ "step": 8000
57
+ },
58
+ {
59
+ "epoch": 0.02,
60
+ "learning_rate": 5e-05,
61
+ "loss": 5.2017,
62
+ "step": 9000
63
+ },
64
+ {
65
+ "epoch": 0.02,
66
+ "learning_rate": 5e-05,
67
+ "loss": 5.052,
68
+ "step": 10000
69
+ },
70
+ {
71
+ "epoch": 0.03,
72
+ "learning_rate": 5e-05,
73
+ "loss": 4.9005,
74
+ "step": 11000
75
+ },
76
+ {
77
+ "epoch": 0.03,
78
+ "learning_rate": 5e-05,
79
+ "loss": 4.7597,
80
+ "step": 12000
81
+ },
82
+ {
83
+ "epoch": 0.03,
84
+ "learning_rate": 5e-05,
85
+ "loss": 4.6439,
86
+ "step": 13000
87
+ },
88
+ {
89
+ "epoch": 0.03,
90
+ "learning_rate": 5e-05,
91
+ "loss": 4.5544,
92
+ "step": 14000
93
+ },
94
+ {
95
+ "epoch": 0.03,
96
+ "learning_rate": 5e-05,
97
+ "loss": 4.4742,
98
+ "step": 15000
99
+ },
100
+ {
101
+ "epoch": 0.04,
102
+ "learning_rate": 5e-05,
103
+ "loss": 4.4085,
104
+ "step": 16000
105
+ },
106
+ {
107
+ "epoch": 0.04,
108
+ "learning_rate": 5e-05,
109
+ "loss": 4.3437,
110
+ "step": 17000
111
+ },
112
+ {
113
+ "epoch": 0.04,
114
+ "learning_rate": 5e-05,
115
+ "loss": 4.2957,
116
+ "step": 18000
117
+ },
118
+ {
119
+ "epoch": 0.04,
120
+ "learning_rate": 5e-05,
121
+ "loss": 4.2381,
122
+ "step": 19000
123
+ },
124
+ {
125
+ "epoch": 0.05,
126
+ "learning_rate": 5e-05,
127
+ "loss": 4.1846,
128
+ "step": 20000
129
+ },
130
+ {
131
+ "epoch": 0.05,
132
+ "learning_rate": 5e-05,
133
+ "loss": 4.1399,
134
+ "step": 21000
135
+ },
136
+ {
137
+ "epoch": 0.05,
138
+ "learning_rate": 5e-05,
139
+ "loss": 4.1097,
140
+ "step": 22000
141
+ },
142
+ {
143
+ "epoch": 0.05,
144
+ "learning_rate": 5e-05,
145
+ "loss": 4.0513,
146
+ "step": 23000
147
+ },
148
+ {
149
+ "epoch": 0.06,
150
+ "learning_rate": 5e-05,
151
+ "loss": 4.0177,
152
+ "step": 24000
153
+ },
154
+ {
155
+ "epoch": 0.06,
156
+ "learning_rate": 5e-05,
157
+ "loss": 3.9807,
158
+ "step": 25000
159
+ },
160
+ {
161
+ "epoch": 0.06,
162
+ "learning_rate": 5e-05,
163
+ "loss": 3.9417,
164
+ "step": 26000
165
+ },
166
+ {
167
+ "epoch": 0.06,
168
+ "learning_rate": 5e-05,
169
+ "loss": 3.9119,
170
+ "step": 27000
171
+ },
172
+ {
173
+ "epoch": 0.06,
174
+ "learning_rate": 5e-05,
175
+ "loss": 3.8789,
176
+ "step": 28000
177
+ },
178
+ {
179
+ "epoch": 0.07,
180
+ "learning_rate": 5e-05,
181
+ "loss": 3.8467,
182
+ "step": 29000
183
+ },
184
+ {
185
+ "epoch": 0.07,
186
+ "learning_rate": 5e-05,
187
+ "loss": 3.8162,
188
+ "step": 30000
189
+ },
190
+ {
191
+ "epoch": 0.07,
192
+ "learning_rate": 5e-05,
193
+ "loss": 3.7815,
194
+ "step": 31000
195
+ },
196
+ {
197
+ "epoch": 0.07,
198
+ "learning_rate": 5e-05,
199
+ "loss": 3.755,
200
+ "step": 32000
201
+ },
202
+ {
203
+ "epoch": 0.08,
204
+ "learning_rate": 5e-05,
205
+ "loss": 3.7321,
206
+ "step": 33000
207
+ },
208
+ {
209
+ "epoch": 0.08,
210
+ "learning_rate": 5e-05,
211
+ "loss": 3.7082,
212
+ "step": 34000
213
+ },
214
+ {
215
+ "epoch": 0.08,
216
+ "learning_rate": 5e-05,
217
+ "loss": 3.6796,
218
+ "step": 35000
219
+ },
220
+ {
221
+ "epoch": 0.08,
222
+ "learning_rate": 5e-05,
223
+ "loss": 3.6481,
224
+ "step": 36000
225
+ },
226
+ {
227
+ "epoch": 0.09,
228
+ "learning_rate": 5e-05,
229
+ "loss": 3.6342,
230
+ "step": 37000
231
+ },
232
+ {
233
+ "epoch": 0.09,
234
+ "learning_rate": 5e-05,
235
+ "loss": 3.5971,
236
+ "step": 38000
237
+ },
238
+ {
239
+ "epoch": 0.09,
240
+ "learning_rate": 5e-05,
241
+ "loss": 3.5824,
242
+ "step": 39000
243
+ },
244
+ {
245
+ "epoch": 0.09,
246
+ "learning_rate": 5e-05,
247
+ "loss": 3.5602,
248
+ "step": 40000
249
+ },
250
+ {
251
+ "epoch": 0.1,
252
+ "learning_rate": 5e-05,
253
+ "loss": 3.5299,
254
+ "step": 41000
255
+ },
256
+ {
257
+ "epoch": 0.1,
258
+ "learning_rate": 5e-05,
259
+ "loss": 3.5166,
260
+ "step": 42000
261
+ },
262
+ {
263
+ "epoch": 0.1,
264
+ "learning_rate": 5e-05,
265
+ "loss": 3.494,
266
+ "step": 43000
267
+ },
268
+ {
269
+ "epoch": 0.1,
270
+ "learning_rate": 5e-05,
271
+ "loss": 3.4759,
272
+ "step": 44000
273
+ },
274
+ {
275
+ "epoch": 0.1,
276
+ "learning_rate": 5e-05,
277
+ "loss": 3.4558,
278
+ "step": 45000
279
+ },
280
+ {
281
+ "epoch": 0.11,
282
+ "learning_rate": 5e-05,
283
+ "loss": 3.4323,
284
+ "step": 46000
285
+ },
286
+ {
287
+ "epoch": 0.11,
288
+ "learning_rate": 5e-05,
289
+ "loss": 3.415,
290
+ "step": 47000
291
+ },
292
+ {
293
+ "epoch": 0.11,
294
+ "learning_rate": 5e-05,
295
+ "loss": 3.396,
296
+ "step": 48000
297
+ },
298
+ {
299
+ "epoch": 0.11,
300
+ "learning_rate": 5e-05,
301
+ "loss": 3.3864,
302
+ "step": 49000
303
+ },
304
+ {
305
+ "epoch": 0.12,
306
+ "learning_rate": 5e-05,
307
+ "loss": 3.3646,
308
+ "step": 50000
309
+ },
310
+ {
311
+ "epoch": 0.12,
312
+ "learning_rate": 5e-05,
313
+ "loss": 3.3441,
314
+ "step": 51000
315
+ },
316
+ {
317
+ "epoch": 0.12,
318
+ "learning_rate": 5e-05,
319
+ "loss": 3.3341,
320
+ "step": 52000
321
+ },
322
+ {
323
+ "epoch": 0.12,
324
+ "learning_rate": 5e-05,
325
+ "loss": 3.3227,
326
+ "step": 53000
327
+ },
328
+ {
329
+ "epoch": 0.13,
330
+ "learning_rate": 5e-05,
331
+ "loss": 3.3056,
332
+ "step": 54000
333
+ },
334
+ {
335
+ "epoch": 0.13,
336
+ "learning_rate": 5e-05,
337
+ "loss": 3.289,
338
+ "step": 55000
339
+ },
340
+ {
341
+ "epoch": 0.13,
342
+ "learning_rate": 5e-05,
343
+ "loss": 3.2696,
344
+ "step": 56000
345
+ },
346
+ {
347
+ "epoch": 0.13,
348
+ "learning_rate": 5e-05,
349
+ "loss": 3.258,
350
+ "step": 57000
351
+ },
352
+ {
353
+ "epoch": 0.13,
354
+ "learning_rate": 5e-05,
355
+ "loss": 3.2438,
356
+ "step": 58000
357
+ },
358
+ {
359
+ "epoch": 0.14,
360
+ "learning_rate": 5e-05,
361
+ "loss": 3.223,
362
+ "step": 59000
363
+ },
364
+ {
365
+ "epoch": 0.14,
366
+ "learning_rate": 5e-05,
367
+ "loss": 3.2181,
368
+ "step": 60000
369
+ },
370
+ {
371
+ "epoch": 0.14,
372
+ "learning_rate": 5e-05,
373
+ "loss": 3.2049,
374
+ "step": 61000
375
+ },
376
+ {
377
+ "epoch": 0.14,
378
+ "learning_rate": 5e-05,
379
+ "loss": 3.1886,
380
+ "step": 62000
381
+ },
382
+ {
383
+ "epoch": 0.15,
384
+ "learning_rate": 5e-05,
385
+ "loss": 3.183,
386
+ "step": 63000
387
+ },
388
+ {
389
+ "epoch": 0.15,
390
+ "learning_rate": 5e-05,
391
+ "loss": 3.1623,
392
+ "step": 64000
393
+ },
394
+ {
395
+ "epoch": 0.15,
396
+ "learning_rate": 5e-05,
397
+ "loss": 3.1553,
398
+ "step": 65000
399
+ },
400
+ {
401
+ "epoch": 0.15,
402
+ "learning_rate": 5e-05,
403
+ "loss": 3.1407,
404
+ "step": 66000
405
+ },
406
+ {
407
+ "epoch": 0.16,
408
+ "learning_rate": 5e-05,
409
+ "loss": 3.1316,
410
+ "step": 67000
411
+ },
412
+ {
413
+ "epoch": 0.16,
414
+ "learning_rate": 5e-05,
415
+ "loss": 3.1209,
416
+ "step": 68000
417
+ },
418
+ {
419
+ "epoch": 0.16,
420
+ "learning_rate": 5e-05,
421
+ "loss": 3.1047,
422
+ "step": 69000
423
+ },
424
+ {
425
+ "epoch": 0.16,
426
+ "learning_rate": 5e-05,
427
+ "loss": 3.0984,
428
+ "step": 70000
429
+ },
430
+ {
431
+ "epoch": 0.16,
432
+ "learning_rate": 5e-05,
433
+ "loss": 3.0901,
434
+ "step": 71000
435
+ },
436
+ {
437
+ "epoch": 0.17,
438
+ "learning_rate": 5e-05,
439
+ "loss": 3.0713,
440
+ "step": 72000
441
+ },
442
+ {
443
+ "epoch": 0.17,
444
+ "learning_rate": 5e-05,
445
+ "loss": 3.0674,
446
+ "step": 73000
447
+ },
448
+ {
449
+ "epoch": 0.17,
450
+ "learning_rate": 5e-05,
451
+ "loss": 3.0568,
452
+ "step": 74000
453
+ },
454
+ {
455
+ "epoch": 0.17,
456
+ "learning_rate": 5e-05,
457
+ "loss": 3.0448,
458
+ "step": 75000
459
+ },
460
+ {
461
+ "epoch": 0.18,
462
+ "learning_rate": 5e-05,
463
+ "loss": 3.0282,
464
+ "step": 76000
465
+ },
466
+ {
467
+ "epoch": 0.18,
468
+ "learning_rate": 5e-05,
469
+ "loss": 3.0217,
470
+ "step": 77000
471
+ },
472
+ {
473
+ "epoch": 0.18,
474
+ "learning_rate": 5e-05,
475
+ "loss": 3.0215,
476
+ "step": 78000
477
+ },
478
+ {
479
+ "epoch": 0.18,
480
+ "learning_rate": 5e-05,
481
+ "loss": 3.0051,
482
+ "step": 79000
483
+ },
484
+ {
485
+ "epoch": 0.19,
486
+ "learning_rate": 5e-05,
487
+ "loss": 2.9998,
488
+ "step": 80000
489
+ },
490
+ {
491
+ "epoch": 0.19,
492
+ "learning_rate": 5e-05,
493
+ "loss": 2.9937,
494
+ "step": 81000
495
+ },
496
+ {
497
+ "epoch": 0.19,
498
+ "learning_rate": 5e-05,
499
+ "loss": 2.9874,
500
+ "step": 82000
501
+ },
502
+ {
503
+ "epoch": 0.19,
504
+ "learning_rate": 5e-05,
505
+ "loss": 2.9712,
506
+ "step": 83000
507
+ },
508
+ {
509
+ "epoch": 0.19,
510
+ "learning_rate": 5e-05,
511
+ "loss": 2.9656,
512
+ "step": 84000
513
+ },
514
+ {
515
+ "epoch": 0.2,
516
+ "learning_rate": 5e-05,
517
+ "loss": 2.9621,
518
+ "step": 85000
519
+ },
520
+ {
521
+ "epoch": 0.2,
522
+ "learning_rate": 5e-05,
523
+ "loss": 2.9509,
524
+ "step": 86000
525
+ },
526
+ {
527
+ "epoch": 0.2,
528
+ "learning_rate": 5e-05,
529
+ "loss": 2.9423,
530
+ "step": 87000
531
+ },
532
+ {
533
+ "epoch": 0.2,
534
+ "learning_rate": 5e-05,
535
+ "loss": 2.9436,
536
+ "step": 88000
537
+ },
538
+ {
539
+ "epoch": 0.21,
540
+ "learning_rate": 5e-05,
541
+ "loss": 2.9297,
542
+ "step": 89000
543
+ },
544
+ {
545
+ "epoch": 0.21,
546
+ "learning_rate": 5e-05,
547
+ "loss": 2.9205,
548
+ "step": 90000
549
+ },
550
+ {
551
+ "epoch": 0.21,
552
+ "learning_rate": 5e-05,
553
+ "loss": 2.9103,
554
+ "step": 91000
555
+ },
556
+ {
557
+ "epoch": 0.21,
558
+ "learning_rate": 5e-05,
559
+ "loss": 2.9019,
560
+ "step": 92000
561
+ },
562
+ {
563
+ "epoch": 0.22,
564
+ "learning_rate": 5e-05,
565
+ "loss": 2.8987,
566
+ "step": 93000
567
+ },
568
+ {
569
+ "epoch": 0.22,
570
+ "learning_rate": 5e-05,
571
+ "loss": 2.8913,
572
+ "step": 94000
573
+ },
574
+ {
575
+ "epoch": 0.22,
576
+ "learning_rate": 5e-05,
577
+ "loss": 2.8863,
578
+ "step": 95000
579
+ },
580
+ {
581
+ "epoch": 0.22,
582
+ "learning_rate": 5e-05,
583
+ "loss": 2.8781,
584
+ "step": 96000
585
+ },
586
+ {
587
+ "epoch": 0.22,
588
+ "learning_rate": 5e-05,
589
+ "loss": 2.8695,
590
+ "step": 97000
591
+ },
592
+ {
593
+ "epoch": 0.23,
594
+ "learning_rate": 5e-05,
595
+ "loss": 2.8731,
596
+ "step": 98000
597
+ },
598
+ {
599
+ "epoch": 0.23,
600
+ "learning_rate": 5e-05,
601
+ "loss": 2.8595,
602
+ "step": 99000
603
+ },
604
+ {
605
+ "epoch": 0.23,
606
+ "learning_rate": 5e-05,
607
+ "loss": 2.8548,
608
+ "step": 100000
609
+ },
610
+ {
611
+ "epoch": 0.23,
612
+ "learning_rate": 5e-05,
613
+ "loss": 2.8471,
614
+ "step": 101000
615
+ },
616
+ {
617
+ "epoch": 0.24,
618
+ "learning_rate": 5e-05,
619
+ "loss": 2.84,
620
+ "step": 102000
621
+ },
622
+ {
623
+ "epoch": 0.24,
624
+ "learning_rate": 5e-05,
625
+ "loss": 2.8328,
626
+ "step": 103000
627
+ },
628
+ {
629
+ "epoch": 0.24,
630
+ "learning_rate": 5e-05,
631
+ "loss": 2.8317,
632
+ "step": 104000
633
+ },
634
+ {
635
+ "epoch": 0.24,
636
+ "learning_rate": 5e-05,
637
+ "loss": 2.8278,
638
+ "step": 105000
639
+ },
640
+ {
641
+ "epoch": 0.25,
642
+ "learning_rate": 5e-05,
643
+ "loss": 2.821,
644
+ "step": 106000
645
+ },
646
+ {
647
+ "epoch": 0.25,
648
+ "learning_rate": 5e-05,
649
+ "loss": 2.8145,
650
+ "step": 107000
651
+ },
652
+ {
653
+ "epoch": 0.25,
654
+ "learning_rate": 5e-05,
655
+ "loss": 2.8083,
656
+ "step": 108000
657
+ },
658
+ {
659
+ "epoch": 0.25,
660
+ "learning_rate": 5e-05,
661
+ "loss": 2.7986,
662
+ "step": 109000
663
+ },
664
+ {
665
+ "epoch": 0.25,
666
+ "learning_rate": 5e-05,
667
+ "loss": 2.7942,
668
+ "step": 110000
669
+ },
670
+ {
671
+ "epoch": 0.26,
672
+ "learning_rate": 5e-05,
673
+ "loss": 2.7862,
674
+ "step": 111000
675
+ },
676
+ {
677
+ "epoch": 0.26,
678
+ "learning_rate": 5e-05,
679
+ "loss": 2.782,
680
+ "step": 112000
681
+ },
682
+ {
683
+ "epoch": 0.26,
684
+ "learning_rate": 5e-05,
685
+ "loss": 2.7773,
686
+ "step": 113000
687
+ },
688
+ {
689
+ "epoch": 0.26,
690
+ "learning_rate": 5e-05,
691
+ "loss": 2.7767,
692
+ "step": 114000
693
+ },
694
+ {
695
+ "epoch": 0.27,
696
+ "learning_rate": 5e-05,
697
+ "loss": 2.771,
698
+ "step": 115000
699
+ },
700
+ {
701
+ "epoch": 0.27,
702
+ "learning_rate": 5e-05,
703
+ "loss": 2.7649,
704
+ "step": 116000
705
+ },
706
+ {
707
+ "epoch": 0.27,
708
+ "learning_rate": 5e-05,
709
+ "loss": 2.7632,
710
+ "step": 117000
711
+ },
712
+ {
713
+ "epoch": 0.27,
714
+ "learning_rate": 5e-05,
715
+ "loss": 2.755,
716
+ "step": 118000
717
+ },
718
+ {
719
+ "epoch": 0.28,
720
+ "learning_rate": 5e-05,
721
+ "loss": 2.7415,
722
+ "step": 119000
723
+ },
724
+ {
725
+ "epoch": 0.28,
726
+ "learning_rate": 5e-05,
727
+ "loss": 2.7435,
728
+ "step": 120000
729
+ },
730
+ {
731
+ "epoch": 0.28,
732
+ "learning_rate": 5e-05,
733
+ "loss": 2.7405,
734
+ "step": 121000
735
+ },
736
+ {
737
+ "epoch": 0.28,
738
+ "learning_rate": 5e-05,
739
+ "loss": 2.7348,
740
+ "step": 122000
741
+ },
742
+ {
743
+ "epoch": 0.29,
744
+ "learning_rate": 5e-05,
745
+ "loss": 2.7296,
746
+ "step": 123000
747
+ },
748
+ {
749
+ "epoch": 0.29,
750
+ "learning_rate": 5e-05,
751
+ "loss": 2.7244,
752
+ "step": 124000
753
+ },
754
+ {
755
+ "epoch": 0.29,
756
+ "learning_rate": 5e-05,
757
+ "loss": 2.7208,
758
+ "step": 125000
759
+ },
760
+ {
761
+ "epoch": 0.29,
762
+ "learning_rate": 5e-05,
763
+ "loss": 2.7157,
764
+ "step": 126000
765
+ },
766
+ {
767
+ "epoch": 0.29,
768
+ "learning_rate": 5e-05,
769
+ "loss": 2.7143,
770
+ "step": 127000
771
+ },
772
+ {
773
+ "epoch": 0.3,
774
+ "learning_rate": 5e-05,
775
+ "loss": 2.7108,
776
+ "step": 128000
777
+ },
778
+ {
779
+ "epoch": 0.3,
780
+ "learning_rate": 5e-05,
781
+ "loss": 2.7009,
782
+ "step": 129000
783
+ },
784
+ {
785
+ "epoch": 0.3,
786
+ "learning_rate": 5e-05,
787
+ "loss": 2.7023,
788
+ "step": 130000
789
+ },
790
+ {
791
+ "epoch": 0.3,
792
+ "learning_rate": 5e-05,
793
+ "loss": 2.6931,
794
+ "step": 131000
795
+ },
796
+ {
797
+ "epoch": 0.31,
798
+ "learning_rate": 5e-05,
799
+ "loss": 2.6901,
800
+ "step": 132000
801
+ },
802
+ {
803
+ "epoch": 0.31,
804
+ "learning_rate": 5e-05,
805
+ "loss": 2.6871,
806
+ "step": 133000
807
+ },
808
+ {
809
+ "epoch": 0.31,
810
+ "learning_rate": 5e-05,
811
+ "loss": 2.6843,
812
+ "step": 134000
813
+ },
814
+ {
815
+ "epoch": 0.31,
816
+ "learning_rate": 5e-05,
817
+ "loss": 2.6842,
818
+ "step": 135000
819
+ },
820
+ {
821
+ "epoch": 0.32,
822
+ "learning_rate": 5e-05,
823
+ "loss": 2.6771,
824
+ "step": 136000
825
+ },
826
+ {
827
+ "epoch": 0.32,
828
+ "learning_rate": 5e-05,
829
+ "loss": 2.6708,
830
+ "step": 137000
831
+ },
832
+ {
833
+ "epoch": 0.32,
834
+ "learning_rate": 5e-05,
835
+ "loss": 2.6638,
836
+ "step": 138000
837
+ },
838
+ {
839
+ "epoch": 0.32,
840
+ "learning_rate": 5e-05,
841
+ "loss": 2.6617,
842
+ "step": 139000
843
+ },
844
+ {
845
+ "epoch": 0.32,
846
+ "learning_rate": 5e-05,
847
+ "loss": 2.6626,
848
+ "step": 140000
849
+ },
850
+ {
851
+ "epoch": 0.33,
852
+ "learning_rate": 5e-05,
853
+ "loss": 2.655,
854
+ "step": 141000
855
+ },
856
+ {
857
+ "epoch": 0.33,
858
+ "learning_rate": 5e-05,
859
+ "loss": 2.6576,
860
+ "step": 142000
861
+ },
862
+ {
863
+ "epoch": 0.33,
864
+ "learning_rate": 5e-05,
865
+ "loss": 2.6408,
866
+ "step": 143000
867
+ },
868
+ {
869
+ "epoch": 0.33,
870
+ "learning_rate": 5e-05,
871
+ "loss": 2.6422,
872
+ "step": 144000
873
+ },
874
+ {
875
+ "epoch": 0.34,
876
+ "learning_rate": 5e-05,
877
+ "loss": 2.6416,
878
+ "step": 145000
879
+ },
880
+ {
881
+ "epoch": 0.34,
882
+ "learning_rate": 5e-05,
883
+ "loss": 2.6411,
884
+ "step": 146000
885
+ },
886
+ {
887
+ "epoch": 0.34,
888
+ "learning_rate": 5e-05,
889
+ "loss": 2.6327,
890
+ "step": 147000
891
+ },
892
+ {
893
+ "epoch": 0.34,
894
+ "learning_rate": 5e-05,
895
+ "loss": 2.6334,
896
+ "step": 148000
897
+ },
898
+ {
899
+ "epoch": 0.35,
900
+ "learning_rate": 5e-05,
901
+ "loss": 2.6208,
902
+ "step": 149000
903
+ },
904
+ {
905
+ "epoch": 0.35,
906
+ "learning_rate": 5e-05,
907
+ "loss": 2.6232,
908
+ "step": 150000
909
+ },
910
+ {
911
+ "epoch": 0.35,
912
+ "eval_accuracy": 0.5274870522655163,
913
+ "eval_loss": 2.509765625,
914
+ "eval_runtime": 4883.386,
915
+ "eval_samples_per_second": 114.367,
916
+ "eval_steps_per_second": 0.894,
917
+ "step": 150000
918
+ },
919
+ {
920
+ "epoch": 0.35,
921
+ "learning_rate": 5e-05,
922
+ "loss": 2.6229,
923
+ "step": 151000
924
+ },
925
+ {
926
+ "epoch": 0.35,
927
+ "learning_rate": 5e-05,
928
+ "loss": 2.6185,
929
+ "step": 152000
930
+ },
931
+ {
932
+ "epoch": 0.35,
933
+ "learning_rate": 5e-05,
934
+ "loss": 2.6103,
935
+ "step": 153000
936
+ },
937
+ {
938
+ "epoch": 0.36,
939
+ "learning_rate": 5e-05,
940
+ "loss": 2.6089,
941
+ "step": 154000
942
+ },
943
+ {
944
+ "epoch": 0.36,
945
+ "learning_rate": 5e-05,
946
+ "loss": 2.6157,
947
+ "step": 155000
948
+ },
949
+ {
950
+ "epoch": 0.36,
951
+ "learning_rate": 5e-05,
952
+ "loss": 2.6045,
953
+ "step": 156000
954
+ },
955
+ {
956
+ "epoch": 0.36,
957
+ "learning_rate": 5e-05,
958
+ "loss": 2.6045,
959
+ "step": 157000
960
+ },
961
+ {
962
+ "epoch": 0.37,
963
+ "learning_rate": 5e-05,
964
+ "loss": 2.5938,
965
+ "step": 158000
966
+ },
967
+ {
968
+ "epoch": 0.37,
969
+ "learning_rate": 5e-05,
970
+ "loss": 2.5924,
971
+ "step": 159000
972
+ },
973
+ {
974
+ "epoch": 0.37,
975
+ "learning_rate": 5e-05,
976
+ "loss": 2.5943,
977
+ "step": 160000
978
+ },
979
+ {
980
+ "epoch": 0.37,
981
+ "learning_rate": 5e-05,
982
+ "loss": 2.589,
983
+ "step": 161000
984
+ },
985
+ {
986
+ "epoch": 0.38,
987
+ "learning_rate": 5e-05,
988
+ "loss": 2.5906,
989
+ "step": 162000
990
+ },
991
+ {
992
+ "epoch": 0.38,
993
+ "learning_rate": 5e-05,
994
+ "loss": 2.575,
995
+ "step": 163000
996
+ },
997
+ {
998
+ "epoch": 0.38,
999
+ "learning_rate": 5e-05,
1000
+ "loss": 2.5799,
1001
+ "step": 164000
1002
+ },
1003
+ {
1004
+ "epoch": 0.38,
1005
+ "learning_rate": 5e-05,
1006
+ "loss": 2.576,
1007
+ "step": 165000
1008
+ },
1009
+ {
1010
+ "epoch": 0.38,
1011
+ "learning_rate": 5e-05,
1012
+ "loss": 2.574,
1013
+ "step": 166000
1014
+ },
1015
+ {
1016
+ "epoch": 0.39,
1017
+ "learning_rate": 5e-05,
1018
+ "loss": 2.5697,
1019
+ "step": 167000
1020
+ },
1021
+ {
1022
+ "epoch": 0.39,
1023
+ "learning_rate": 5e-05,
1024
+ "loss": 2.5699,
1025
+ "step": 168000
1026
+ },
1027
+ {
1028
+ "epoch": 0.39,
1029
+ "learning_rate": 5e-05,
1030
+ "loss": 2.5694,
1031
+ "step": 169000
1032
+ },
1033
+ {
1034
+ "epoch": 0.39,
1035
+ "learning_rate": 5e-05,
1036
+ "loss": 2.5629,
1037
+ "step": 170000
1038
+ },
1039
+ {
1040
+ "epoch": 0.4,
1041
+ "learning_rate": 5e-05,
1042
+ "loss": 2.563,
1043
+ "step": 171000
1044
+ },
1045
+ {
1046
+ "epoch": 0.4,
1047
+ "learning_rate": 5e-05,
1048
+ "loss": 2.5618,
1049
+ "step": 172000
1050
+ },
1051
+ {
1052
+ "epoch": 0.4,
1053
+ "learning_rate": 5e-05,
1054
+ "loss": 2.551,
1055
+ "step": 173000
1056
+ },
1057
+ {
1058
+ "epoch": 0.4,
1059
+ "learning_rate": 5e-05,
1060
+ "loss": 2.5494,
1061
+ "step": 174000
1062
+ },
1063
+ {
1064
+ "epoch": 0.41,
1065
+ "learning_rate": 5e-05,
1066
+ "loss": 2.544,
1067
+ "step": 175000
1068
+ },
1069
+ {
1070
+ "epoch": 0.41,
1071
+ "learning_rate": 5e-05,
1072
+ "loss": 2.5483,
1073
+ "step": 176000
1074
+ },
1075
+ {
1076
+ "epoch": 0.41,
1077
+ "learning_rate": 5e-05,
1078
+ "loss": 2.5458,
1079
+ "step": 177000
1080
+ },
1081
+ {
1082
+ "epoch": 0.41,
1083
+ "learning_rate": 5e-05,
1084
+ "loss": 2.5463,
1085
+ "step": 178000
1086
+ },
1087
+ {
1088
+ "epoch": 0.41,
1089
+ "learning_rate": 5e-05,
1090
+ "loss": 2.5353,
1091
+ "step": 179000
1092
+ },
1093
+ {
1094
+ "epoch": 0.42,
1095
+ "learning_rate": 5e-05,
1096
+ "loss": 2.5317,
1097
+ "step": 180000
1098
+ },
1099
+ {
1100
+ "epoch": 0.42,
1101
+ "learning_rate": 5e-05,
1102
+ "loss": 2.5352,
1103
+ "step": 181000
1104
+ },
1105
+ {
1106
+ "epoch": 0.42,
1107
+ "learning_rate": 5e-05,
1108
+ "loss": 2.5295,
1109
+ "step": 182000
1110
+ },
1111
+ {
1112
+ "epoch": 0.42,
1113
+ "learning_rate": 5e-05,
1114
+ "loss": 2.5286,
1115
+ "step": 183000
1116
+ },
1117
+ {
1118
+ "epoch": 0.43,
1119
+ "learning_rate": 5e-05,
1120
+ "loss": 2.5196,
1121
+ "step": 184000
1122
+ },
1123
+ {
1124
+ "epoch": 0.43,
1125
+ "learning_rate": 5e-05,
1126
+ "loss": 2.5222,
1127
+ "step": 185000
1128
+ },
1129
+ {
1130
+ "epoch": 0.43,
1131
+ "learning_rate": 5e-05,
1132
+ "loss": 2.5245,
1133
+ "step": 186000
1134
+ },
1135
+ {
1136
+ "epoch": 0.43,
1137
+ "learning_rate": 5e-05,
1138
+ "loss": 2.5154,
1139
+ "step": 187000
1140
+ },
1141
+ {
1142
+ "epoch": 0.44,
1143
+ "learning_rate": 5e-05,
1144
+ "loss": 2.509,
1145
+ "step": 188000
1146
+ },
1147
+ {
1148
+ "epoch": 0.44,
1149
+ "learning_rate": 5e-05,
1150
+ "loss": 2.5131,
1151
+ "step": 189000
1152
+ },
1153
+ {
1154
+ "epoch": 0.44,
1155
+ "learning_rate": 5e-05,
1156
+ "loss": 2.5072,
1157
+ "step": 190000
1158
+ },
1159
+ {
1160
+ "epoch": 0.44,
1161
+ "learning_rate": 5e-05,
1162
+ "loss": 2.5079,
1163
+ "step": 191000
1164
+ },
1165
+ {
1166
+ "epoch": 0.45,
1167
+ "learning_rate": 5e-05,
1168
+ "loss": 2.5062,
1169
+ "step": 192000
1170
+ },
1171
+ {
1172
+ "epoch": 0.45,
1173
+ "learning_rate": 5e-05,
1174
+ "loss": 2.5008,
1175
+ "step": 193000
1176
+ },
1177
+ {
1178
+ "epoch": 0.45,
1179
+ "learning_rate": 5e-05,
1180
+ "loss": 2.504,
1181
+ "step": 194000
1182
+ },
1183
+ {
1184
+ "epoch": 0.45,
1185
+ "learning_rate": 5e-05,
1186
+ "loss": 2.5013,
1187
+ "step": 195000
1188
+ },
1189
+ {
1190
+ "epoch": 0.45,
1191
+ "learning_rate": 5e-05,
1192
+ "loss": 2.4979,
1193
+ "step": 196000
1194
+ },
1195
+ {
1196
+ "epoch": 0.46,
1197
+ "learning_rate": 5e-05,
1198
+ "loss": 2.4919,
1199
+ "step": 197000
1200
+ },
1201
+ {
1202
+ "epoch": 0.46,
1203
+ "learning_rate": 5e-05,
1204
+ "loss": 2.4892,
1205
+ "step": 198000
1206
+ },
1207
+ {
1208
+ "epoch": 0.46,
1209
+ "learning_rate": 5e-05,
1210
+ "loss": 2.4897,
1211
+ "step": 199000
1212
+ },
1213
+ {
1214
+ "epoch": 0.46,
1215
+ "learning_rate": 5e-05,
1216
+ "loss": 2.4897,
1217
+ "step": 200000
1218
+ },
1219
+ {
1220
+ "epoch": 0.47,
1221
+ "learning_rate": 5e-05,
1222
+ "loss": 2.4859,
1223
+ "step": 201000
1224
+ },
1225
+ {
1226
+ "epoch": 0.47,
1227
+ "learning_rate": 5e-05,
1228
+ "loss": 2.4766,
1229
+ "step": 202000
1230
+ },
1231
+ {
1232
+ "epoch": 0.47,
1233
+ "learning_rate": 5e-05,
1234
+ "loss": 2.4815,
1235
+ "step": 203000
1236
+ },
1237
+ {
1238
+ "epoch": 0.47,
1239
+ "learning_rate": 5e-05,
1240
+ "loss": 2.4797,
1241
+ "step": 204000
1242
+ },
1243
+ {
1244
+ "epoch": 0.48,
1245
+ "learning_rate": 5e-05,
1246
+ "loss": 2.4782,
1247
+ "step": 205000
1248
+ },
1249
+ {
1250
+ "epoch": 0.48,
1251
+ "learning_rate": 5e-05,
1252
+ "loss": 2.477,
1253
+ "step": 206000
1254
+ },
1255
+ {
1256
+ "epoch": 0.48,
1257
+ "learning_rate": 5e-05,
1258
+ "loss": 2.4714,
1259
+ "step": 207000
1260
+ },
1261
+ {
1262
+ "epoch": 0.48,
1263
+ "learning_rate": 5e-05,
1264
+ "loss": 2.4693,
1265
+ "step": 208000
1266
+ },
1267
+ {
1268
+ "epoch": 0.48,
1269
+ "learning_rate": 5e-05,
1270
+ "loss": 2.4647,
1271
+ "step": 209000
1272
+ },
1273
+ {
1274
+ "epoch": 0.49,
1275
+ "learning_rate": 5e-05,
1276
+ "loss": 2.4675,
1277
+ "step": 210000
1278
+ },
1279
+ {
1280
+ "epoch": 0.49,
1281
+ "learning_rate": 5e-05,
1282
+ "loss": 2.4628,
1283
+ "step": 211000
1284
+ },
1285
+ {
1286
+ "epoch": 0.49,
1287
+ "learning_rate": 5e-05,
1288
+ "loss": 2.458,
1289
+ "step": 212000
1290
+ },
1291
+ {
1292
+ "epoch": 0.49,
1293
+ "learning_rate": 5e-05,
1294
+ "loss": 2.4575,
1295
+ "step": 213000
1296
+ },
1297
+ {
1298
+ "epoch": 0.5,
1299
+ "learning_rate": 5e-05,
1300
+ "loss": 2.4588,
1301
+ "step": 214000
1302
+ },
1303
+ {
1304
+ "epoch": 0.5,
1305
+ "learning_rate": 5e-05,
1306
+ "loss": 2.4541,
1307
+ "step": 215000
1308
+ },
1309
+ {
1310
+ "epoch": 0.5,
1311
+ "learning_rate": 5e-05,
1312
+ "loss": 2.4556,
1313
+ "step": 216000
1314
+ },
1315
+ {
1316
+ "epoch": 0.5,
1317
+ "learning_rate": 5e-05,
1318
+ "loss": 2.4419,
1319
+ "step": 217000
1320
+ },
1321
+ {
1322
+ "epoch": 0.51,
1323
+ "learning_rate": 5e-05,
1324
+ "loss": 2.4505,
1325
+ "step": 218000
1326
+ },
1327
+ {
1328
+ "epoch": 0.51,
1329
+ "learning_rate": 5e-05,
1330
+ "loss": 2.4469,
1331
+ "step": 219000
1332
+ },
1333
+ {
1334
+ "epoch": 0.51,
1335
+ "learning_rate": 5e-05,
1336
+ "loss": 2.4488,
1337
+ "step": 220000
1338
+ },
1339
+ {
1340
+ "epoch": 0.51,
1341
+ "learning_rate": 5e-05,
1342
+ "loss": 2.4493,
1343
+ "step": 221000
1344
+ },
1345
+ {
1346
+ "epoch": 0.51,
1347
+ "learning_rate": 5e-05,
1348
+ "loss": 2.4399,
1349
+ "step": 222000
1350
+ },
1351
+ {
1352
+ "epoch": 0.52,
1353
+ "learning_rate": 5e-05,
1354
+ "loss": 2.4408,
1355
+ "step": 223000
1356
+ },
1357
+ {
1358
+ "epoch": 0.52,
1359
+ "learning_rate": 5e-05,
1360
+ "loss": 2.4413,
1361
+ "step": 224000
1362
+ },
1363
+ {
1364
+ "epoch": 0.52,
1365
+ "learning_rate": 5e-05,
1366
+ "loss": 2.4375,
1367
+ "step": 225000
1368
+ },
1369
+ {
1370
+ "epoch": 0.52,
1371
+ "learning_rate": 5e-05,
1372
+ "loss": 2.4306,
1373
+ "step": 226000
1374
+ },
1375
+ {
1376
+ "epoch": 0.53,
1377
+ "learning_rate": 5e-05,
1378
+ "loss": 2.4354,
1379
+ "step": 227000
1380
+ },
1381
+ {
1382
+ "epoch": 0.53,
1383
+ "learning_rate": 5e-05,
1384
+ "loss": 2.426,
1385
+ "step": 228000
1386
+ },
1387
+ {
1388
+ "epoch": 0.53,
1389
+ "learning_rate": 5e-05,
1390
+ "loss": 2.4307,
1391
+ "step": 229000
1392
+ },
1393
+ {
1394
+ "epoch": 0.53,
1395
+ "learning_rate": 5e-05,
1396
+ "loss": 2.4297,
1397
+ "step": 230000
1398
+ },
1399
+ {
1400
+ "epoch": 0.54,
1401
+ "learning_rate": 5e-05,
1402
+ "loss": 2.4268,
1403
+ "step": 231000
1404
+ },
1405
+ {
1406
+ "epoch": 0.54,
1407
+ "learning_rate": 5e-05,
1408
+ "loss": 2.4249,
1409
+ "step": 232000
1410
+ },
1411
+ {
1412
+ "epoch": 0.54,
1413
+ "learning_rate": 5e-05,
1414
+ "loss": 2.4211,
1415
+ "step": 233000
1416
+ },
1417
+ {
1418
+ "epoch": 0.54,
1419
+ "learning_rate": 5e-05,
1420
+ "loss": 2.4233,
1421
+ "step": 234000
1422
+ },
1423
+ {
1424
+ "epoch": 0.54,
1425
+ "learning_rate": 5e-05,
1426
+ "loss": 2.415,
1427
+ "step": 235000
1428
+ },
1429
+ {
1430
+ "epoch": 0.55,
1431
+ "learning_rate": 5e-05,
1432
+ "loss": 2.4163,
1433
+ "step": 236000
1434
+ },
1435
+ {
1436
+ "epoch": 0.55,
1437
+ "learning_rate": 5e-05,
1438
+ "loss": 2.4185,
1439
+ "step": 237000
1440
+ },
1441
+ {
1442
+ "epoch": 0.55,
1443
+ "learning_rate": 5e-05,
1444
+ "loss": 2.4148,
1445
+ "step": 238000
1446
+ },
1447
+ {
1448
+ "epoch": 0.55,
1449
+ "learning_rate": 5e-05,
1450
+ "loss": 2.4089,
1451
+ "step": 239000
1452
+ },
1453
+ {
1454
+ "epoch": 0.56,
1455
+ "learning_rate": 5e-05,
1456
+ "loss": 2.411,
1457
+ "step": 240000
1458
+ },
1459
+ {
1460
+ "epoch": 0.56,
1461
+ "learning_rate": 5e-05,
1462
+ "loss": 2.4061,
1463
+ "step": 241000
1464
+ },
1465
+ {
1466
+ "epoch": 0.56,
1467
+ "learning_rate": 5e-05,
1468
+ "loss": 2.4056,
1469
+ "step": 242000
1470
+ },
1471
+ {
1472
+ "epoch": 0.56,
1473
+ "learning_rate": 5e-05,
1474
+ "loss": 2.402,
1475
+ "step": 243000
1476
+ },
1477
+ {
1478
+ "epoch": 0.57,
1479
+ "learning_rate": 5e-05,
1480
+ "loss": 2.4042,
1481
+ "step": 244000
1482
+ },
1483
+ {
1484
+ "epoch": 0.57,
1485
+ "learning_rate": 5e-05,
1486
+ "loss": 2.4025,
1487
+ "step": 245000
1488
+ },
1489
+ {
1490
+ "epoch": 0.57,
1491
+ "learning_rate": 5e-05,
1492
+ "loss": 2.3981,
1493
+ "step": 246000
1494
+ },
1495
+ {
1496
+ "epoch": 0.57,
1497
+ "learning_rate": 5e-05,
1498
+ "loss": 2.4005,
1499
+ "step": 247000
1500
+ },
1501
+ {
1502
+ "epoch": 0.57,
1503
+ "learning_rate": 5e-05,
1504
+ "loss": 2.3938,
1505
+ "step": 248000
1506
+ },
1507
+ {
1508
+ "epoch": 0.58,
1509
+ "learning_rate": 5e-05,
1510
+ "loss": 2.3914,
1511
+ "step": 249000
1512
+ },
1513
+ {
1514
+ "epoch": 0.58,
1515
+ "learning_rate": 5e-05,
1516
+ "loss": 2.3901,
1517
+ "step": 250000
1518
+ },
1519
+ {
1520
+ "epoch": 0.58,
1521
+ "learning_rate": 5e-05,
1522
+ "loss": 2.3912,
1523
+ "step": 251000
1524
+ },
1525
+ {
1526
+ "epoch": 0.58,
1527
+ "learning_rate": 5e-05,
1528
+ "loss": 2.3881,
1529
+ "step": 252000
1530
+ },
1531
+ {
1532
+ "epoch": 0.59,
1533
+ "learning_rate": 5e-05,
1534
+ "loss": 2.3904,
1535
+ "step": 253000
1536
+ },
1537
+ {
1538
+ "epoch": 0.59,
1539
+ "learning_rate": 5e-05,
1540
+ "loss": 2.3907,
1541
+ "step": 254000
1542
+ },
1543
+ {
1544
+ "epoch": 0.59,
1545
+ "learning_rate": 5e-05,
1546
+ "loss": 2.3827,
1547
+ "step": 255000
1548
+ },
1549
+ {
1550
+ "epoch": 0.59,
1551
+ "learning_rate": 5e-05,
1552
+ "loss": 2.3832,
1553
+ "step": 256000
1554
+ },
1555
+ {
1556
+ "epoch": 0.6,
1557
+ "learning_rate": 5e-05,
1558
+ "loss": 2.3824,
1559
+ "step": 257000
1560
+ },
1561
+ {
1562
+ "epoch": 0.6,
1563
+ "learning_rate": 5e-05,
1564
+ "loss": 2.3831,
1565
+ "step": 258000
1566
+ },
1567
+ {
1568
+ "epoch": 0.6,
1569
+ "learning_rate": 5e-05,
1570
+ "loss": 2.377,
1571
+ "step": 259000
1572
+ },
1573
+ {
1574
+ "epoch": 0.6,
1575
+ "learning_rate": 5e-05,
1576
+ "loss": 2.3732,
1577
+ "step": 260000
1578
+ },
1579
+ {
1580
+ "epoch": 0.6,
1581
+ "learning_rate": 5e-05,
1582
+ "loss": 2.3753,
1583
+ "step": 261000
1584
+ },
1585
+ {
1586
+ "epoch": 0.61,
1587
+ "learning_rate": 5e-05,
1588
+ "loss": 2.3747,
1589
+ "step": 262000
1590
+ },
1591
+ {
1592
+ "epoch": 0.61,
1593
+ "learning_rate": 5e-05,
1594
+ "loss": 2.3684,
1595
+ "step": 263000
1596
+ },
1597
+ {
1598
+ "epoch": 0.61,
1599
+ "learning_rate": 5e-05,
1600
+ "loss": 2.3716,
1601
+ "step": 264000
1602
+ },
1603
+ {
1604
+ "epoch": 0.61,
1605
+ "learning_rate": 5e-05,
1606
+ "loss": 2.3681,
1607
+ "step": 265000
1608
+ },
1609
+ {
1610
+ "epoch": 0.62,
1611
+ "learning_rate": 5e-05,
1612
+ "loss": 2.3694,
1613
+ "step": 266000
1614
+ },
1615
+ {
1616
+ "epoch": 0.62,
1617
+ "learning_rate": 5e-05,
1618
+ "loss": 2.3622,
1619
+ "step": 267000
1620
+ },
1621
+ {
1622
+ "epoch": 0.62,
1623
+ "learning_rate": 5e-05,
1624
+ "loss": 2.3646,
1625
+ "step": 268000
1626
+ },
1627
+ {
1628
+ "epoch": 0.62,
1629
+ "learning_rate": 5e-05,
1630
+ "loss": 2.367,
1631
+ "step": 269000
1632
+ },
1633
+ {
1634
+ "epoch": 0.63,
1635
+ "learning_rate": 5e-05,
1636
+ "loss": 2.365,
1637
+ "step": 270000
1638
+ },
1639
+ {
1640
+ "epoch": 0.63,
1641
+ "learning_rate": 5e-05,
1642
+ "loss": 2.3608,
1643
+ "step": 271000
1644
+ },
1645
+ {
1646
+ "epoch": 0.63,
1647
+ "learning_rate": 5e-05,
1648
+ "loss": 2.3558,
1649
+ "step": 272000
1650
+ },
1651
+ {
1652
+ "epoch": 0.63,
1653
+ "learning_rate": 5e-05,
1654
+ "loss": 2.3627,
1655
+ "step": 273000
1656
+ },
1657
+ {
1658
+ "epoch": 0.64,
1659
+ "learning_rate": 5e-05,
1660
+ "loss": 2.3541,
1661
+ "step": 274000
1662
+ },
1663
+ {
1664
+ "epoch": 0.64,
1665
+ "learning_rate": 5e-05,
1666
+ "loss": 2.3541,
1667
+ "step": 275000
1668
+ },
1669
+ {
1670
+ "epoch": 0.64,
1671
+ "learning_rate": 5e-05,
1672
+ "loss": 2.3516,
1673
+ "step": 276000
1674
+ },
1675
+ {
1676
+ "epoch": 0.64,
1677
+ "learning_rate": 5e-05,
1678
+ "loss": 2.354,
1679
+ "step": 277000
1680
+ },
1681
+ {
1682
+ "epoch": 0.64,
1683
+ "learning_rate": 5e-05,
1684
+ "loss": 2.3494,
1685
+ "step": 278000
1686
+ },
1687
+ {
1688
+ "epoch": 0.65,
1689
+ "learning_rate": 5e-05,
1690
+ "loss": 2.3529,
1691
+ "step": 279000
1692
+ },
1693
+ {
1694
+ "epoch": 0.65,
1695
+ "learning_rate": 5e-05,
1696
+ "loss": 2.3493,
1697
+ "step": 280000
1698
+ },
1699
+ {
1700
+ "epoch": 0.65,
1701
+ "learning_rate": 5e-05,
1702
+ "loss": 2.3519,
1703
+ "step": 281000
1704
+ },
1705
+ {
1706
+ "epoch": 0.65,
1707
+ "learning_rate": 5e-05,
1708
+ "loss": 2.3442,
1709
+ "step": 282000
1710
+ },
1711
+ {
1712
+ "epoch": 0.66,
1713
+ "learning_rate": 5e-05,
1714
+ "loss": 2.3431,
1715
+ "step": 283000
1716
+ },
1717
+ {
1718
+ "epoch": 0.66,
1719
+ "learning_rate": 5e-05,
1720
+ "loss": 2.3489,
1721
+ "step": 284000
1722
+ },
1723
+ {
1724
+ "epoch": 0.66,
1725
+ "learning_rate": 5e-05,
1726
+ "loss": 2.3446,
1727
+ "step": 285000
1728
+ },
1729
+ {
1730
+ "epoch": 0.66,
1731
+ "learning_rate": 5e-05,
1732
+ "loss": 2.3451,
1733
+ "step": 286000
1734
+ },
1735
+ {
1736
+ "epoch": 0.67,
1737
+ "learning_rate": 5e-05,
1738
+ "loss": 2.3307,
1739
+ "step": 287000
1740
+ },
1741
+ {
1742
+ "epoch": 0.67,
1743
+ "learning_rate": 5e-05,
1744
+ "loss": 2.3336,
1745
+ "step": 288000
1746
+ },
1747
+ {
1748
+ "epoch": 0.67,
1749
+ "learning_rate": 5e-05,
1750
+ "loss": 2.3383,
1751
+ "step": 289000
1752
+ },
1753
+ {
1754
+ "epoch": 0.67,
1755
+ "learning_rate": 5e-05,
1756
+ "loss": 2.3376,
1757
+ "step": 290000
1758
+ },
1759
+ {
1760
+ "epoch": 0.67,
1761
+ "learning_rate": 5e-05,
1762
+ "loss": 2.3336,
1763
+ "step": 291000
1764
+ },
1765
+ {
1766
+ "epoch": 0.68,
1767
+ "learning_rate": 5e-05,
1768
+ "loss": 2.3351,
1769
+ "step": 292000
1770
+ },
1771
+ {
1772
+ "epoch": 0.68,
1773
+ "learning_rate": 5e-05,
1774
+ "loss": 2.3296,
1775
+ "step": 293000
1776
+ },
1777
+ {
1778
+ "epoch": 0.68,
1779
+ "learning_rate": 5e-05,
1780
+ "loss": 2.3295,
1781
+ "step": 294000
1782
+ },
1783
+ {
1784
+ "epoch": 0.68,
1785
+ "learning_rate": 5e-05,
1786
+ "loss": 2.3274,
1787
+ "step": 295000
1788
+ },
1789
+ {
1790
+ "epoch": 0.69,
1791
+ "learning_rate": 5e-05,
1792
+ "loss": 2.3268,
1793
+ "step": 296000
1794
+ },
1795
+ {
1796
+ "epoch": 0.69,
1797
+ "learning_rate": 5e-05,
1798
+ "loss": 2.3289,
1799
+ "step": 297000
1800
+ },
1801
+ {
1802
+ "epoch": 0.69,
1803
+ "learning_rate": 5e-05,
1804
+ "loss": 2.3293,
1805
+ "step": 298000
1806
+ },
1807
+ {
1808
+ "epoch": 0.69,
1809
+ "learning_rate": 5e-05,
1810
+ "loss": 2.3212,
1811
+ "step": 299000
1812
+ },
1813
+ {
1814
+ "epoch": 0.7,
1815
+ "learning_rate": 5e-05,
1816
+ "loss": 2.3189,
1817
+ "step": 300000
1818
+ },
1819
+ {
1820
+ "epoch": 0.7,
1821
+ "eval_accuracy": 0.5684444981680046,
1822
+ "eval_loss": 2.2109375,
1823
+ "eval_runtime": 4880.4554,
1824
+ "eval_samples_per_second": 114.436,
1825
+ "eval_steps_per_second": 0.894,
1826
+ "step": 300000
1827
+ },
1828
+ {
1829
+ "epoch": 0.7,
1830
+ "learning_rate": 5e-05,
1831
+ "loss": 2.3193,
1832
+ "step": 301000
1833
+ },
1834
+ {
1835
+ "epoch": 0.7,
1836
+ "learning_rate": 5e-05,
1837
+ "loss": 2.317,
1838
+ "step": 302000
1839
+ },
1840
+ {
1841
+ "epoch": 0.7,
1842
+ "learning_rate": 5e-05,
1843
+ "loss": 2.3212,
1844
+ "step": 303000
1845
+ },
1846
+ {
1847
+ "epoch": 0.7,
1848
+ "learning_rate": 5e-05,
1849
+ "loss": 2.317,
1850
+ "step": 304000
1851
+ },
1852
+ {
1853
+ "epoch": 0.71,
1854
+ "learning_rate": 5e-05,
1855
+ "loss": 2.3183,
1856
+ "step": 305000
1857
+ },
1858
+ {
1859
+ "epoch": 0.71,
1860
+ "learning_rate": 5e-05,
1861
+ "loss": 2.3178,
1862
+ "step": 306000
1863
+ },
1864
+ {
1865
+ "epoch": 0.71,
1866
+ "learning_rate": 5e-05,
1867
+ "loss": 2.3085,
1868
+ "step": 307000
1869
+ },
1870
+ {
1871
+ "epoch": 0.71,
1872
+ "learning_rate": 5e-05,
1873
+ "loss": 2.3127,
1874
+ "step": 308000
1875
+ },
1876
+ {
1877
+ "epoch": 0.72,
1878
+ "learning_rate": 5e-05,
1879
+ "loss": 2.3149,
1880
+ "step": 309000
1881
+ },
1882
+ {
1883
+ "epoch": 0.72,
1884
+ "learning_rate": 5e-05,
1885
+ "loss": 2.3091,
1886
+ "step": 310000
1887
+ },
1888
+ {
1889
+ "epoch": 0.72,
1890
+ "learning_rate": 5e-05,
1891
+ "loss": 2.3088,
1892
+ "step": 311000
1893
+ },
1894
+ {
1895
+ "epoch": 0.72,
1896
+ "learning_rate": 5e-05,
1897
+ "loss": 2.3097,
1898
+ "step": 312000
1899
+ },
1900
+ {
1901
+ "epoch": 0.73,
1902
+ "learning_rate": 5e-05,
1903
+ "loss": 2.3131,
1904
+ "step": 313000
1905
+ },
1906
+ {
1907
+ "epoch": 0.73,
1908
+ "learning_rate": 5e-05,
1909
+ "loss": 2.3031,
1910
+ "step": 314000
1911
+ },
1912
+ {
1913
+ "epoch": 0.73,
1914
+ "learning_rate": 5e-05,
1915
+ "loss": 2.302,
1916
+ "step": 315000
1917
+ },
1918
+ {
1919
+ "epoch": 0.73,
1920
+ "learning_rate": 5e-05,
1921
+ "loss": 2.3056,
1922
+ "step": 316000
1923
+ },
1924
+ {
1925
+ "epoch": 0.73,
1926
+ "learning_rate": 5e-05,
1927
+ "loss": 2.3046,
1928
+ "step": 317000
1929
+ },
1930
+ {
1931
+ "epoch": 0.74,
1932
+ "learning_rate": 5e-05,
1933
+ "loss": 2.2997,
1934
+ "step": 318000
1935
+ },
1936
+ {
1937
+ "epoch": 0.74,
1938
+ "learning_rate": 5e-05,
1939
+ "loss": 2.2933,
1940
+ "step": 319000
1941
+ },
1942
+ {
1943
+ "epoch": 0.74,
1944
+ "learning_rate": 5e-05,
1945
+ "loss": 2.2992,
1946
+ "step": 320000
1947
+ },
1948
+ {
1949
+ "epoch": 0.74,
1950
+ "learning_rate": 5e-05,
1951
+ "loss": 2.2947,
1952
+ "step": 321000
1953
+ },
1954
+ {
1955
+ "epoch": 0.75,
1956
+ "learning_rate": 5e-05,
1957
+ "loss": 2.2933,
1958
+ "step": 322000
1959
+ },
1960
+ {
1961
+ "epoch": 0.75,
1962
+ "learning_rate": 5e-05,
1963
+ "loss": 2.295,
1964
+ "step": 323000
1965
+ },
1966
+ {
1967
+ "epoch": 0.75,
1968
+ "learning_rate": 5e-05,
1969
+ "loss": 2.2935,
1970
+ "step": 324000
1971
+ },
1972
+ {
1973
+ "epoch": 0.75,
1974
+ "learning_rate": 5e-05,
1975
+ "loss": 2.2926,
1976
+ "step": 325000
1977
+ },
1978
+ {
1979
+ "epoch": 0.76,
1980
+ "learning_rate": 5e-05,
1981
+ "loss": 2.2887,
1982
+ "step": 326000
1983
+ },
1984
+ {
1985
+ "epoch": 0.76,
1986
+ "learning_rate": 5e-05,
1987
+ "loss": 2.2966,
1988
+ "step": 327000
1989
+ },
1990
+ {
1991
+ "epoch": 0.76,
1992
+ "learning_rate": 5e-05,
1993
+ "loss": 2.2924,
1994
+ "step": 328000
1995
+ },
1996
+ {
1997
+ "epoch": 0.76,
1998
+ "learning_rate": 5e-05,
1999
+ "loss": 2.2897,
2000
+ "step": 329000
2001
+ },
2002
+ {
2003
+ "epoch": 0.76,
2004
+ "learning_rate": 5e-05,
2005
+ "loss": 2.2903,
2006
+ "step": 330000
2007
+ },
2008
+ {
2009
+ "epoch": 0.77,
2010
+ "learning_rate": 5e-05,
2011
+ "loss": 2.282,
2012
+ "step": 331000
2013
+ },
2014
+ {
2015
+ "epoch": 0.77,
2016
+ "learning_rate": 5e-05,
2017
+ "loss": 2.281,
2018
+ "step": 332000
2019
+ },
2020
+ {
2021
+ "epoch": 0.77,
2022
+ "learning_rate": 5e-05,
2023
+ "loss": 2.2903,
2024
+ "step": 333000
2025
+ },
2026
+ {
2027
+ "epoch": 0.77,
2028
+ "learning_rate": 5e-05,
2029
+ "loss": 2.2831,
2030
+ "step": 334000
2031
+ },
2032
+ {
2033
+ "epoch": 0.78,
2034
+ "learning_rate": 5e-05,
2035
+ "loss": 2.2874,
2036
+ "step": 335000
2037
+ },
2038
+ {
2039
+ "epoch": 0.78,
2040
+ "learning_rate": 5e-05,
2041
+ "loss": 2.2722,
2042
+ "step": 336000
2043
+ },
2044
+ {
2045
+ "epoch": 0.78,
2046
+ "learning_rate": 5e-05,
2047
+ "loss": 2.2823,
2048
+ "step": 337000
2049
+ },
2050
+ {
2051
+ "epoch": 0.78,
2052
+ "learning_rate": 5e-05,
2053
+ "loss": 2.2793,
2054
+ "step": 338000
2055
+ },
2056
+ {
2057
+ "epoch": 0.79,
2058
+ "learning_rate": 5e-05,
2059
+ "loss": 2.2839,
2060
+ "step": 339000
2061
+ },
2062
+ {
2063
+ "epoch": 0.79,
2064
+ "learning_rate": 5e-05,
2065
+ "loss": 2.2808,
2066
+ "step": 340000
2067
+ },
2068
+ {
2069
+ "epoch": 0.79,
2070
+ "learning_rate": 5e-05,
2071
+ "loss": 2.2782,
2072
+ "step": 341000
2073
+ },
2074
+ {
2075
+ "epoch": 0.79,
2076
+ "learning_rate": 5e-05,
2077
+ "loss": 2.2751,
2078
+ "step": 342000
2079
+ },
2080
+ {
2081
+ "epoch": 0.8,
2082
+ "learning_rate": 5e-05,
2083
+ "loss": 2.2728,
2084
+ "step": 343000
2085
+ },
2086
+ {
2087
+ "epoch": 0.8,
2088
+ "learning_rate": 5e-05,
2089
+ "loss": 2.2736,
2090
+ "step": 344000
2091
+ },
2092
+ {
2093
+ "epoch": 0.8,
2094
+ "learning_rate": 5e-05,
2095
+ "loss": 2.2761,
2096
+ "step": 345000
2097
+ },
2098
+ {
2099
+ "epoch": 0.8,
2100
+ "learning_rate": 5e-05,
2101
+ "loss": 2.2692,
2102
+ "step": 346000
2103
+ },
2104
+ {
2105
+ "epoch": 0.8,
2106
+ "learning_rate": 5e-05,
2107
+ "loss": 2.269,
2108
+ "step": 347000
2109
+ },
2110
+ {
2111
+ "epoch": 0.81,
2112
+ "learning_rate": 5e-05,
2113
+ "loss": 2.2704,
2114
+ "step": 348000
2115
+ },
2116
+ {
2117
+ "epoch": 0.81,
2118
+ "learning_rate": 5e-05,
2119
+ "loss": 2.2673,
2120
+ "step": 349000
2121
+ },
2122
+ {
2123
+ "epoch": 0.81,
2124
+ "learning_rate": 5e-05,
2125
+ "loss": 2.2657,
2126
+ "step": 350000
2127
+ },
2128
+ {
2129
+ "epoch": 0.81,
2130
+ "learning_rate": 5e-05,
2131
+ "loss": 2.2731,
2132
+ "step": 351000
2133
+ },
2134
+ {
2135
+ "epoch": 0.82,
2136
+ "learning_rate": 5e-05,
2137
+ "loss": 2.2677,
2138
+ "step": 352000
2139
+ },
2140
+ {
2141
+ "epoch": 0.82,
2142
+ "learning_rate": 5e-05,
2143
+ "loss": 2.2628,
2144
+ "step": 353000
2145
+ },
2146
+ {
2147
+ "epoch": 0.82,
2148
+ "learning_rate": 5e-05,
2149
+ "loss": 2.2665,
2150
+ "step": 354000
2151
+ },
2152
+ {
2153
+ "epoch": 0.82,
2154
+ "learning_rate": 5e-05,
2155
+ "loss": 2.2649,
2156
+ "step": 355000
2157
+ },
2158
+ {
2159
+ "epoch": 0.83,
2160
+ "learning_rate": 5e-05,
2161
+ "loss": 2.266,
2162
+ "step": 356000
2163
+ },
2164
+ {
2165
+ "epoch": 0.83,
2166
+ "learning_rate": 5e-05,
2167
+ "loss": 2.2626,
2168
+ "step": 357000
2169
+ },
2170
+ {
2171
+ "epoch": 0.83,
2172
+ "learning_rate": 5e-05,
2173
+ "loss": 2.2605,
2174
+ "step": 358000
2175
+ },
2176
+ {
2177
+ "epoch": 0.83,
2178
+ "learning_rate": 5e-05,
2179
+ "loss": 2.2598,
2180
+ "step": 359000
2181
+ },
2182
+ {
2183
+ "epoch": 0.83,
2184
+ "learning_rate": 5e-05,
2185
+ "loss": 2.2541,
2186
+ "step": 360000
2187
+ },
2188
+ {
2189
+ "epoch": 0.84,
2190
+ "learning_rate": 5e-05,
2191
+ "loss": 2.2527,
2192
+ "step": 361000
2193
+ },
2194
+ {
2195
+ "epoch": 0.84,
2196
+ "learning_rate": 5e-05,
2197
+ "loss": 2.2495,
2198
+ "step": 362000
2199
+ },
2200
+ {
2201
+ "epoch": 0.84,
2202
+ "learning_rate": 5e-05,
2203
+ "loss": 2.2552,
2204
+ "step": 363000
2205
+ },
2206
+ {
2207
+ "epoch": 0.84,
2208
+ "learning_rate": 5e-05,
2209
+ "loss": 2.2575,
2210
+ "step": 364000
2211
+ },
2212
+ {
2213
+ "epoch": 0.85,
2214
+ "learning_rate": 5e-05,
2215
+ "loss": 2.2532,
2216
+ "step": 365000
2217
+ },
2218
+ {
2219
+ "epoch": 0.85,
2220
+ "learning_rate": 5e-05,
2221
+ "loss": 2.2531,
2222
+ "step": 366000
2223
+ },
2224
+ {
2225
+ "epoch": 0.85,
2226
+ "learning_rate": 5e-05,
2227
+ "loss": 2.2512,
2228
+ "step": 367000
2229
+ },
2230
+ {
2231
+ "epoch": 0.85,
2232
+ "learning_rate": 5e-05,
2233
+ "loss": 2.2536,
2234
+ "step": 368000
2235
+ },
2236
+ {
2237
+ "epoch": 0.86,
2238
+ "learning_rate": 5e-05,
2239
+ "loss": 2.25,
2240
+ "step": 369000
2241
+ },
2242
+ {
2243
+ "epoch": 0.86,
2244
+ "learning_rate": 5e-05,
2245
+ "loss": 2.2494,
2246
+ "step": 370000
2247
+ },
2248
+ {
2249
+ "epoch": 0.86,
2250
+ "learning_rate": 5e-05,
2251
+ "loss": 2.2454,
2252
+ "step": 371000
2253
+ },
2254
+ {
2255
+ "epoch": 0.86,
2256
+ "learning_rate": 5e-05,
2257
+ "loss": 2.2416,
2258
+ "step": 372000
2259
+ },
2260
+ {
2261
+ "epoch": 0.86,
2262
+ "learning_rate": 5e-05,
2263
+ "loss": 2.2417,
2264
+ "step": 373000
2265
+ },
2266
+ {
2267
+ "epoch": 0.87,
2268
+ "learning_rate": 5e-05,
2269
+ "loss": 2.2418,
2270
+ "step": 374000
2271
+ },
2272
+ {
2273
+ "epoch": 0.87,
2274
+ "learning_rate": 5e-05,
2275
+ "loss": 2.2467,
2276
+ "step": 375000
2277
+ },
2278
+ {
2279
+ "epoch": 0.87,
2280
+ "learning_rate": 5e-05,
2281
+ "loss": 2.2409,
2282
+ "step": 376000
2283
+ },
2284
+ {
2285
+ "epoch": 0.87,
2286
+ "learning_rate": 5e-05,
2287
+ "loss": 2.2451,
2288
+ "step": 377000
2289
+ },
2290
+ {
2291
+ "epoch": 0.88,
2292
+ "learning_rate": 5e-05,
2293
+ "loss": 2.2397,
2294
+ "step": 378000
2295
+ },
2296
+ {
2297
+ "epoch": 0.88,
2298
+ "learning_rate": 5e-05,
2299
+ "loss": 2.2392,
2300
+ "step": 379000
2301
+ },
2302
+ {
2303
+ "epoch": 0.88,
2304
+ "learning_rate": 5e-05,
2305
+ "loss": 2.2388,
2306
+ "step": 380000
2307
+ },
2308
+ {
2309
+ "epoch": 0.88,
2310
+ "learning_rate": 5e-05,
2311
+ "loss": 2.2369,
2312
+ "step": 381000
2313
+ },
2314
+ {
2315
+ "epoch": 0.89,
2316
+ "learning_rate": 5e-05,
2317
+ "loss": 2.2354,
2318
+ "step": 382000
2319
+ },
2320
+ {
2321
+ "epoch": 0.89,
2322
+ "learning_rate": 5e-05,
2323
+ "loss": 2.2367,
2324
+ "step": 383000
2325
+ },
2326
+ {
2327
+ "epoch": 0.89,
2328
+ "learning_rate": 5e-05,
2329
+ "loss": 2.2383,
2330
+ "step": 384000
2331
+ },
2332
+ {
2333
+ "epoch": 0.89,
2334
+ "learning_rate": 5e-05,
2335
+ "loss": 2.2352,
2336
+ "step": 385000
2337
+ },
2338
+ {
2339
+ "epoch": 0.89,
2340
+ "learning_rate": 5e-05,
2341
+ "loss": 2.236,
2342
+ "step": 386000
2343
+ },
2344
+ {
2345
+ "epoch": 0.9,
2346
+ "learning_rate": 5e-05,
2347
+ "loss": 2.2327,
2348
+ "step": 387000
2349
+ },
2350
+ {
2351
+ "epoch": 0.9,
2352
+ "learning_rate": 5e-05,
2353
+ "loss": 2.2345,
2354
+ "step": 388000
2355
+ },
2356
+ {
2357
+ "epoch": 0.9,
2358
+ "learning_rate": 5e-05,
2359
+ "loss": 2.2361,
2360
+ "step": 389000
2361
+ },
2362
+ {
2363
+ "epoch": 0.9,
2364
+ "learning_rate": 5e-05,
2365
+ "loss": 2.2316,
2366
+ "step": 390000
2367
+ },
2368
+ {
2369
+ "epoch": 0.91,
2370
+ "learning_rate": 5e-05,
2371
+ "loss": 2.2315,
2372
+ "step": 391000
2373
+ },
2374
+ {
2375
+ "epoch": 0.91,
2376
+ "learning_rate": 5e-05,
2377
+ "loss": 2.2315,
2378
+ "step": 392000
2379
+ },
2380
+ {
2381
+ "epoch": 0.91,
2382
+ "learning_rate": 5e-05,
2383
+ "loss": 2.227,
2384
+ "step": 393000
2385
+ },
2386
+ {
2387
+ "epoch": 0.91,
2388
+ "learning_rate": 5e-05,
2389
+ "loss": 2.2307,
2390
+ "step": 394000
2391
+ },
2392
+ {
2393
+ "epoch": 0.92,
2394
+ "learning_rate": 5e-05,
2395
+ "loss": 2.2276,
2396
+ "step": 395000
2397
+ },
2398
+ {
2399
+ "epoch": 0.92,
2400
+ "learning_rate": 5e-05,
2401
+ "loss": 2.219,
2402
+ "step": 396000
2403
+ },
2404
+ {
2405
+ "epoch": 0.92,
2406
+ "learning_rate": 5e-05,
2407
+ "loss": 2.222,
2408
+ "step": 397000
2409
+ },
2410
+ {
2411
+ "epoch": 0.92,
2412
+ "learning_rate": 5e-05,
2413
+ "loss": 2.2262,
2414
+ "step": 398000
2415
+ },
2416
+ {
2417
+ "epoch": 0.92,
2418
+ "learning_rate": 5e-05,
2419
+ "loss": 2.2224,
2420
+ "step": 399000
2421
+ },
2422
+ {
2423
+ "epoch": 0.93,
2424
+ "learning_rate": 5e-05,
2425
+ "loss": 2.217,
2426
+ "step": 400000
2427
+ },
2428
+ {
2429
+ "epoch": 0.93,
2430
+ "learning_rate": 5e-05,
2431
+ "loss": 2.2236,
2432
+ "step": 401000
2433
+ },
2434
+ {
2435
+ "epoch": 0.93,
2436
+ "learning_rate": 5e-05,
2437
+ "loss": 2.2256,
2438
+ "step": 402000
2439
+ },
2440
+ {
2441
+ "epoch": 0.93,
2442
+ "learning_rate": 5e-05,
2443
+ "loss": 2.2181,
2444
+ "step": 403000
2445
+ },
2446
+ {
2447
+ "epoch": 0.94,
2448
+ "learning_rate": 5e-05,
2449
+ "loss": 2.2209,
2450
+ "step": 404000
2451
+ },
2452
+ {
2453
+ "epoch": 0.94,
2454
+ "learning_rate": 5e-05,
2455
+ "loss": 2.2246,
2456
+ "step": 405000
2457
+ },
2458
+ {
2459
+ "epoch": 0.94,
2460
+ "learning_rate": 5e-05,
2461
+ "loss": 2.223,
2462
+ "step": 406000
2463
+ },
2464
+ {
2465
+ "epoch": 0.94,
2466
+ "learning_rate": 5e-05,
2467
+ "loss": 2.2164,
2468
+ "step": 407000
2469
+ },
2470
+ {
2471
+ "epoch": 0.95,
2472
+ "learning_rate": 5e-05,
2473
+ "loss": 2.2105,
2474
+ "step": 408000
2475
+ },
2476
+ {
2477
+ "epoch": 0.95,
2478
+ "learning_rate": 5e-05,
2479
+ "loss": 2.2181,
2480
+ "step": 409000
2481
+ },
2482
+ {
2483
+ "epoch": 0.95,
2484
+ "learning_rate": 5e-05,
2485
+ "loss": 2.2212,
2486
+ "step": 410000
2487
+ },
2488
+ {
2489
+ "epoch": 0.95,
2490
+ "learning_rate": 5e-05,
2491
+ "loss": 2.2138,
2492
+ "step": 411000
2493
+ },
2494
+ {
2495
+ "epoch": 0.95,
2496
+ "learning_rate": 5e-05,
2497
+ "loss": 2.208,
2498
+ "step": 412000
2499
+ },
2500
+ {
2501
+ "epoch": 0.96,
2502
+ "learning_rate": 5e-05,
2503
+ "loss": 2.217,
2504
+ "step": 413000
2505
+ },
2506
+ {
2507
+ "epoch": 0.96,
2508
+ "learning_rate": 5e-05,
2509
+ "loss": 2.2128,
2510
+ "step": 414000
2511
+ },
2512
+ {
2513
+ "epoch": 0.96,
2514
+ "learning_rate": 5e-05,
2515
+ "loss": 2.2139,
2516
+ "step": 415000
2517
+ },
2518
+ {
2519
+ "epoch": 0.96,
2520
+ "learning_rate": 5e-05,
2521
+ "loss": 2.206,
2522
+ "step": 416000
2523
+ },
2524
+ {
2525
+ "epoch": 0.97,
2526
+ "learning_rate": 5e-05,
2527
+ "loss": 2.213,
2528
+ "step": 417000
2529
+ },
2530
+ {
2531
+ "epoch": 0.97,
2532
+ "learning_rate": 5e-05,
2533
+ "loss": 2.2072,
2534
+ "step": 418000
2535
+ },
2536
+ {
2537
+ "epoch": 0.97,
2538
+ "learning_rate": 5e-05,
2539
+ "loss": 2.2115,
2540
+ "step": 419000
2541
+ },
2542
+ {
2543
+ "epoch": 0.97,
2544
+ "learning_rate": 5e-05,
2545
+ "loss": 2.2064,
2546
+ "step": 420000
2547
+ },
2548
+ {
2549
+ "epoch": 0.98,
2550
+ "learning_rate": 5e-05,
2551
+ "loss": 2.2061,
2552
+ "step": 421000
2553
+ },
2554
+ {
2555
+ "epoch": 0.98,
2556
+ "learning_rate": 5e-05,
2557
+ "loss": 2.2079,
2558
+ "step": 422000
2559
+ },
2560
+ {
2561
+ "epoch": 0.98,
2562
+ "learning_rate": 5e-05,
2563
+ "loss": 2.2088,
2564
+ "step": 423000
2565
+ },
2566
+ {
2567
+ "epoch": 0.98,
2568
+ "learning_rate": 5e-05,
2569
+ "loss": 2.2069,
2570
+ "step": 424000
2571
+ },
2572
+ {
2573
+ "epoch": 0.99,
2574
+ "learning_rate": 5e-05,
2575
+ "loss": 2.2037,
2576
+ "step": 425000
2577
+ },
2578
+ {
2579
+ "epoch": 0.99,
2580
+ "learning_rate": 5e-05,
2581
+ "loss": 2.2074,
2582
+ "step": 426000
2583
+ },
2584
+ {
2585
+ "epoch": 0.99,
2586
+ "learning_rate": 5e-05,
2587
+ "loss": 2.2036,
2588
+ "step": 427000
2589
+ },
2590
+ {
2591
+ "epoch": 0.99,
2592
+ "learning_rate": 5e-05,
2593
+ "loss": 2.202,
2594
+ "step": 428000
2595
+ },
2596
+ {
2597
+ "epoch": 0.99,
2598
+ "learning_rate": 5e-05,
2599
+ "loss": 2.2022,
2600
+ "step": 429000
2601
+ },
2602
+ {
2603
+ "epoch": 1.0,
2604
+ "learning_rate": 5e-05,
2605
+ "loss": 2.2031,
2606
+ "step": 430000
2607
+ },
2608
+ {
2609
+ "epoch": 1.0,
2610
+ "learning_rate": 5e-05,
2611
+ "loss": 2.1998,
2612
+ "step": 431000
2613
+ },
2614
+ {
2615
+ "epoch": 1.0,
2616
+ "learning_rate": 5e-05,
2617
+ "loss": 2.202,
2618
+ "step": 432000
2619
+ },
2620
+ {
2621
+ "epoch": 1.0,
2622
+ "learning_rate": 5e-05,
2623
+ "loss": 2.1916,
2624
+ "step": 433000
2625
+ },
2626
+ {
2627
+ "epoch": 1.01,
2628
+ "learning_rate": 5e-05,
2629
+ "loss": 2.1934,
2630
+ "step": 434000
2631
+ },
2632
+ {
2633
+ "epoch": 1.01,
2634
+ "learning_rate": 5e-05,
2635
+ "loss": 2.1958,
2636
+ "step": 435000
2637
+ },
2638
+ {
2639
+ "epoch": 1.01,
2640
+ "learning_rate": 5e-05,
2641
+ "loss": 2.1933,
2642
+ "step": 436000
2643
+ },
2644
+ {
2645
+ "epoch": 1.01,
2646
+ "learning_rate": 5e-05,
2647
+ "loss": 2.1945,
2648
+ "step": 437000
2649
+ },
2650
+ {
2651
+ "epoch": 1.02,
2652
+ "learning_rate": 5e-05,
2653
+ "loss": 2.1889,
2654
+ "step": 438000
2655
+ },
2656
+ {
2657
+ "epoch": 1.02,
2658
+ "learning_rate": 5e-05,
2659
+ "loss": 2.1929,
2660
+ "step": 439000
2661
+ },
2662
+ {
2663
+ "epoch": 1.02,
2664
+ "learning_rate": 5e-05,
2665
+ "loss": 2.1935,
2666
+ "step": 440000
2667
+ },
2668
+ {
2669
+ "epoch": 1.02,
2670
+ "learning_rate": 5e-05,
2671
+ "loss": 2.1883,
2672
+ "step": 441000
2673
+ },
2674
+ {
2675
+ "epoch": 1.02,
2676
+ "learning_rate": 5e-05,
2677
+ "loss": 2.1882,
2678
+ "step": 442000
2679
+ },
2680
+ {
2681
+ "epoch": 1.03,
2682
+ "learning_rate": 5e-05,
2683
+ "loss": 2.1835,
2684
+ "step": 443000
2685
+ },
2686
+ {
2687
+ "epoch": 1.03,
2688
+ "learning_rate": 5e-05,
2689
+ "loss": 2.1862,
2690
+ "step": 444000
2691
+ },
2692
+ {
2693
+ "epoch": 1.03,
2694
+ "learning_rate": 5e-05,
2695
+ "loss": 2.1869,
2696
+ "step": 445000
2697
+ },
2698
+ {
2699
+ "epoch": 1.03,
2700
+ "learning_rate": 5e-05,
2701
+ "loss": 2.1906,
2702
+ "step": 446000
2703
+ },
2704
+ {
2705
+ "epoch": 1.04,
2706
+ "learning_rate": 5e-05,
2707
+ "loss": 2.1825,
2708
+ "step": 447000
2709
+ },
2710
+ {
2711
+ "epoch": 1.04,
2712
+ "learning_rate": 5e-05,
2713
+ "loss": 2.1902,
2714
+ "step": 448000
2715
+ },
2716
+ {
2717
+ "epoch": 1.04,
2718
+ "learning_rate": 5e-05,
2719
+ "loss": 2.1843,
2720
+ "step": 449000
2721
+ },
2722
+ {
2723
+ "epoch": 1.04,
2724
+ "learning_rate": 5e-05,
2725
+ "loss": 2.1811,
2726
+ "step": 450000
2727
+ },
2728
+ {
2729
+ "epoch": 1.04,
2730
+ "eval_accuracy": 0.5877310925035356,
2731
+ "eval_loss": 2.078125,
2732
+ "eval_runtime": 4896.1088,
2733
+ "eval_samples_per_second": 114.07,
2734
+ "eval_steps_per_second": 0.891,
2735
+ "step": 450000
2736
+ },
2737
+ {
2738
+ "epoch": 1.05,
2739
+ "learning_rate": 5e-05,
2740
+ "loss": 2.1848,
2741
+ "step": 451000
2742
+ },
2743
+ {
2744
+ "epoch": 1.05,
2745
+ "learning_rate": 5e-05,
2746
+ "loss": 2.1788,
2747
+ "step": 452000
2748
+ },
2749
+ {
2750
+ "epoch": 1.05,
2751
+ "learning_rate": 5e-05,
2752
+ "loss": 2.185,
2753
+ "step": 453000
2754
+ },
2755
+ {
2756
+ "epoch": 1.05,
2757
+ "learning_rate": 5e-05,
2758
+ "loss": 2.1827,
2759
+ "step": 454000
2760
+ },
2761
+ {
2762
+ "epoch": 1.05,
2763
+ "learning_rate": 5e-05,
2764
+ "loss": 2.1772,
2765
+ "step": 455000
2766
+ },
2767
+ {
2768
+ "epoch": 1.06,
2769
+ "learning_rate": 5e-05,
2770
+ "loss": 2.1792,
2771
+ "step": 456000
2772
+ },
2773
+ {
2774
+ "epoch": 1.06,
2775
+ "learning_rate": 5e-05,
2776
+ "loss": 2.1799,
2777
+ "step": 457000
2778
+ },
2779
+ {
2780
+ "epoch": 1.06,
2781
+ "learning_rate": 5e-05,
2782
+ "loss": 2.1831,
2783
+ "step": 458000
2784
+ },
2785
+ {
2786
+ "epoch": 1.06,
2787
+ "learning_rate": 5e-05,
2788
+ "loss": 2.1763,
2789
+ "step": 459000
2790
+ },
2791
+ {
2792
+ "epoch": 1.07,
2793
+ "learning_rate": 5e-05,
2794
+ "loss": 2.1773,
2795
+ "step": 460000
2796
+ },
2797
+ {
2798
+ "epoch": 1.07,
2799
+ "learning_rate": 5e-05,
2800
+ "loss": 2.1804,
2801
+ "step": 461000
2802
+ },
2803
+ {
2804
+ "epoch": 1.07,
2805
+ "learning_rate": 5e-05,
2806
+ "loss": 2.1711,
2807
+ "step": 462000
2808
+ },
2809
+ {
2810
+ "epoch": 1.07,
2811
+ "learning_rate": 5e-05,
2812
+ "loss": 2.1788,
2813
+ "step": 463000
2814
+ },
2815
+ {
2816
+ "epoch": 1.08,
2817
+ "learning_rate": 5e-05,
2818
+ "loss": 2.174,
2819
+ "step": 464000
2820
+ },
2821
+ {
2822
+ "epoch": 1.08,
2823
+ "learning_rate": 5e-05,
2824
+ "loss": 2.1708,
2825
+ "step": 465000
2826
+ },
2827
+ {
2828
+ "epoch": 1.08,
2829
+ "learning_rate": 5e-05,
2830
+ "loss": 2.1772,
2831
+ "step": 466000
2832
+ },
2833
+ {
2834
+ "epoch": 1.08,
2835
+ "learning_rate": 5e-05,
2836
+ "loss": 2.174,
2837
+ "step": 467000
2838
+ },
2839
+ {
2840
+ "epoch": 1.08,
2841
+ "learning_rate": 5e-05,
2842
+ "loss": 2.1663,
2843
+ "step": 468000
2844
+ },
2845
+ {
2846
+ "epoch": 1.09,
2847
+ "learning_rate": 5e-05,
2848
+ "loss": 2.1737,
2849
+ "step": 469000
2850
+ },
2851
+ {
2852
+ "epoch": 1.09,
2853
+ "learning_rate": 5e-05,
2854
+ "loss": 2.1696,
2855
+ "step": 470000
2856
+ },
2857
+ {
2858
+ "epoch": 1.09,
2859
+ "learning_rate": 5e-05,
2860
+ "loss": 2.1726,
2861
+ "step": 471000
2862
+ },
2863
+ {
2864
+ "epoch": 1.09,
2865
+ "learning_rate": 5e-05,
2866
+ "loss": 2.1681,
2867
+ "step": 472000
2868
+ },
2869
+ {
2870
+ "epoch": 1.1,
2871
+ "learning_rate": 5e-05,
2872
+ "loss": 2.174,
2873
+ "step": 473000
2874
+ },
2875
+ {
2876
+ "epoch": 1.1,
2877
+ "learning_rate": 5e-05,
2878
+ "loss": 2.1679,
2879
+ "step": 474000
2880
+ },
2881
+ {
2882
+ "epoch": 1.1,
2883
+ "learning_rate": 5e-05,
2884
+ "loss": 2.17,
2885
+ "step": 475000
2886
+ },
2887
+ {
2888
+ "epoch": 1.1,
2889
+ "learning_rate": 5e-05,
2890
+ "loss": 2.1664,
2891
+ "step": 476000
2892
+ },
2893
+ {
2894
+ "epoch": 1.11,
2895
+ "learning_rate": 5e-05,
2896
+ "loss": 2.1657,
2897
+ "step": 477000
2898
+ },
2899
+ {
2900
+ "epoch": 1.11,
2901
+ "learning_rate": 5e-05,
2902
+ "loss": 2.1664,
2903
+ "step": 478000
2904
+ },
2905
+ {
2906
+ "epoch": 1.11,
2907
+ "learning_rate": 5e-05,
2908
+ "loss": 2.1674,
2909
+ "step": 479000
2910
+ },
2911
+ {
2912
+ "epoch": 1.11,
2913
+ "learning_rate": 5e-05,
2914
+ "loss": 2.1616,
2915
+ "step": 480000
2916
+ },
2917
+ {
2918
+ "epoch": 1.11,
2919
+ "learning_rate": 5e-05,
2920
+ "loss": 2.1597,
2921
+ "step": 481000
2922
+ },
2923
+ {
2924
+ "epoch": 1.12,
2925
+ "learning_rate": 5e-05,
2926
+ "loss": 2.161,
2927
+ "step": 482000
2928
+ },
2929
+ {
2930
+ "epoch": 1.12,
2931
+ "learning_rate": 5e-05,
2932
+ "loss": 2.1626,
2933
+ "step": 483000
2934
+ },
2935
+ {
2936
+ "epoch": 1.12,
2937
+ "learning_rate": 5e-05,
2938
+ "loss": 2.1643,
2939
+ "step": 484000
2940
+ },
2941
+ {
2942
+ "epoch": 1.12,
2943
+ "learning_rate": 5e-05,
2944
+ "loss": 2.1597,
2945
+ "step": 485000
2946
+ },
2947
+ {
2948
+ "epoch": 1.13,
2949
+ "learning_rate": 5e-05,
2950
+ "loss": 2.1575,
2951
+ "step": 486000
2952
+ },
2953
+ {
2954
+ "epoch": 1.13,
2955
+ "learning_rate": 5e-05,
2956
+ "loss": 2.1551,
2957
+ "step": 487000
2958
+ },
2959
+ {
2960
+ "epoch": 1.13,
2961
+ "learning_rate": 5e-05,
2962
+ "loss": 2.1561,
2963
+ "step": 488000
2964
+ },
2965
+ {
2966
+ "epoch": 1.13,
2967
+ "learning_rate": 5e-05,
2968
+ "loss": 2.1524,
2969
+ "step": 489000
2970
+ },
2971
+ {
2972
+ "epoch": 1.14,
2973
+ "learning_rate": 5e-05,
2974
+ "loss": 2.1631,
2975
+ "step": 490000
2976
+ },
2977
+ {
2978
+ "epoch": 1.14,
2979
+ "learning_rate": 5e-05,
2980
+ "loss": 2.1553,
2981
+ "step": 491000
2982
+ },
2983
+ {
2984
+ "epoch": 1.14,
2985
+ "learning_rate": 5e-05,
2986
+ "loss": 2.1595,
2987
+ "step": 492000
2988
+ },
2989
+ {
2990
+ "epoch": 1.14,
2991
+ "learning_rate": 5e-05,
2992
+ "loss": 2.1556,
2993
+ "step": 493000
2994
+ },
2995
+ {
2996
+ "epoch": 1.15,
2997
+ "learning_rate": 5e-05,
2998
+ "loss": 2.1572,
2999
+ "step": 494000
3000
+ },
3001
+ {
3002
+ "epoch": 1.15,
3003
+ "learning_rate": 5e-05,
3004
+ "loss": 2.1586,
3005
+ "step": 495000
3006
+ },
3007
+ {
3008
+ "epoch": 1.15,
3009
+ "learning_rate": 5e-05,
3010
+ "loss": 2.161,
3011
+ "step": 496000
3012
+ },
3013
+ {
3014
+ "epoch": 1.15,
3015
+ "learning_rate": 5e-05,
3016
+ "loss": 2.152,
3017
+ "step": 497000
3018
+ },
3019
+ {
3020
+ "epoch": 1.15,
3021
+ "learning_rate": 5e-05,
3022
+ "loss": 2.1611,
3023
+ "step": 498000
3024
+ },
3025
+ {
3026
+ "epoch": 1.16,
3027
+ "learning_rate": 5e-05,
3028
+ "loss": 2.1484,
3029
+ "step": 499000
3030
+ },
3031
+ {
3032
+ "epoch": 1.16,
3033
+ "learning_rate": 5e-05,
3034
+ "loss": 2.1516,
3035
+ "step": 500000
3036
+ },
3037
+ {
3038
+ "epoch": 1.16,
3039
+ "learning_rate": 5e-05,
3040
+ "loss": 2.1504,
3041
+ "step": 501000
3042
+ },
3043
+ {
3044
+ "epoch": 1.16,
3045
+ "learning_rate": 5e-05,
3046
+ "loss": 2.1493,
3047
+ "step": 502000
3048
+ },
3049
+ {
3050
+ "epoch": 1.17,
3051
+ "learning_rate": 5e-05,
3052
+ "loss": 2.1494,
3053
+ "step": 503000
3054
+ },
3055
+ {
3056
+ "epoch": 1.17,
3057
+ "learning_rate": 5e-05,
3058
+ "loss": 2.1507,
3059
+ "step": 504000
3060
+ },
3061
+ {
3062
+ "epoch": 1.17,
3063
+ "learning_rate": 5e-05,
3064
+ "loss": 2.1555,
3065
+ "step": 505000
3066
+ },
3067
+ {
3068
+ "epoch": 1.17,
3069
+ "learning_rate": 5e-05,
3070
+ "loss": 2.1522,
3071
+ "step": 506000
3072
+ },
3073
+ {
3074
+ "epoch": 1.18,
3075
+ "learning_rate": 5e-05,
3076
+ "loss": 2.1477,
3077
+ "step": 507000
3078
+ },
3079
+ {
3080
+ "epoch": 1.18,
3081
+ "learning_rate": 5e-05,
3082
+ "loss": 2.1452,
3083
+ "step": 508000
3084
+ },
3085
+ {
3086
+ "epoch": 1.18,
3087
+ "learning_rate": 5e-05,
3088
+ "loss": 2.1516,
3089
+ "step": 509000
3090
+ },
3091
+ {
3092
+ "epoch": 1.18,
3093
+ "learning_rate": 5e-05,
3094
+ "loss": 2.1486,
3095
+ "step": 510000
3096
+ },
3097
+ {
3098
+ "epoch": 1.18,
3099
+ "learning_rate": 5e-05,
3100
+ "loss": 2.1481,
3101
+ "step": 511000
3102
+ },
3103
+ {
3104
+ "epoch": 1.19,
3105
+ "learning_rate": 5e-05,
3106
+ "loss": 2.1461,
3107
+ "step": 512000
3108
+ },
3109
+ {
3110
+ "epoch": 1.19,
3111
+ "learning_rate": 5e-05,
3112
+ "loss": 2.1393,
3113
+ "step": 513000
3114
+ },
3115
+ {
3116
+ "epoch": 1.19,
3117
+ "learning_rate": 5e-05,
3118
+ "loss": 2.1419,
3119
+ "step": 514000
3120
+ },
3121
+ {
3122
+ "epoch": 1.19,
3123
+ "learning_rate": 5e-05,
3124
+ "loss": 2.1376,
3125
+ "step": 515000
3126
+ },
3127
+ {
3128
+ "epoch": 1.2,
3129
+ "learning_rate": 5e-05,
3130
+ "loss": 2.1486,
3131
+ "step": 516000
3132
+ },
3133
+ {
3134
+ "epoch": 1.2,
3135
+ "learning_rate": 5e-05,
3136
+ "loss": 2.1425,
3137
+ "step": 517000
3138
+ },
3139
+ {
3140
+ "epoch": 1.2,
3141
+ "learning_rate": 5e-05,
3142
+ "loss": 2.147,
3143
+ "step": 518000
3144
+ },
3145
+ {
3146
+ "epoch": 1.2,
3147
+ "learning_rate": 5e-05,
3148
+ "loss": 2.1458,
3149
+ "step": 519000
3150
+ },
3151
+ {
3152
+ "epoch": 1.21,
3153
+ "learning_rate": 5e-05,
3154
+ "loss": 2.1393,
3155
+ "step": 520000
3156
+ },
3157
+ {
3158
+ "epoch": 1.21,
3159
+ "learning_rate": 5e-05,
3160
+ "loss": 2.141,
3161
+ "step": 521000
3162
+ },
3163
+ {
3164
+ "epoch": 1.21,
3165
+ "learning_rate": 5e-05,
3166
+ "loss": 2.1375,
3167
+ "step": 522000
3168
+ },
3169
+ {
3170
+ "epoch": 1.21,
3171
+ "learning_rate": 5e-05,
3172
+ "loss": 2.143,
3173
+ "step": 523000
3174
+ },
3175
+ {
3176
+ "epoch": 1.21,
3177
+ "learning_rate": 5e-05,
3178
+ "loss": 2.1395,
3179
+ "step": 524000
3180
+ },
3181
+ {
3182
+ "epoch": 1.22,
3183
+ "learning_rate": 5e-05,
3184
+ "loss": 2.1376,
3185
+ "step": 525000
3186
+ },
3187
+ {
3188
+ "epoch": 1.22,
3189
+ "learning_rate": 5e-05,
3190
+ "loss": 2.1365,
3191
+ "step": 526000
3192
+ },
3193
+ {
3194
+ "epoch": 1.22,
3195
+ "learning_rate": 5e-05,
3196
+ "loss": 2.1378,
3197
+ "step": 527000
3198
+ },
3199
+ {
3200
+ "epoch": 1.22,
3201
+ "learning_rate": 5e-05,
3202
+ "loss": 2.1361,
3203
+ "step": 528000
3204
+ },
3205
+ {
3206
+ "epoch": 1.23,
3207
+ "learning_rate": 5e-05,
3208
+ "loss": 2.1397,
3209
+ "step": 529000
3210
+ },
3211
+ {
3212
+ "epoch": 1.23,
3213
+ "learning_rate": 5e-05,
3214
+ "loss": 2.1319,
3215
+ "step": 530000
3216
+ },
3217
+ {
3218
+ "epoch": 1.23,
3219
+ "learning_rate": 5e-05,
3220
+ "loss": 2.1341,
3221
+ "step": 531000
3222
+ },
3223
+ {
3224
+ "epoch": 1.23,
3225
+ "learning_rate": 5e-05,
3226
+ "loss": 2.1307,
3227
+ "step": 532000
3228
+ },
3229
+ {
3230
+ "epoch": 1.24,
3231
+ "learning_rate": 5e-05,
3232
+ "loss": 2.1273,
3233
+ "step": 533000
3234
+ },
3235
+ {
3236
+ "epoch": 1.24,
3237
+ "learning_rate": 5e-05,
3238
+ "loss": 2.1317,
3239
+ "step": 534000
3240
+ },
3241
+ {
3242
+ "epoch": 1.24,
3243
+ "learning_rate": 5e-05,
3244
+ "loss": 2.1268,
3245
+ "step": 535000
3246
+ },
3247
+ {
3248
+ "epoch": 1.24,
3249
+ "learning_rate": 5e-05,
3250
+ "loss": 2.1342,
3251
+ "step": 536000
3252
+ },
3253
+ {
3254
+ "epoch": 1.24,
3255
+ "learning_rate": 5e-05,
3256
+ "loss": 2.1372,
3257
+ "step": 537000
3258
+ },
3259
+ {
3260
+ "epoch": 1.25,
3261
+ "learning_rate": 5e-05,
3262
+ "loss": 2.1311,
3263
+ "step": 538000
3264
+ },
3265
+ {
3266
+ "epoch": 1.25,
3267
+ "learning_rate": 5e-05,
3268
+ "loss": 2.1301,
3269
+ "step": 539000
3270
+ },
3271
+ {
3272
+ "epoch": 1.25,
3273
+ "learning_rate": 5e-05,
3274
+ "loss": 2.1283,
3275
+ "step": 540000
3276
+ },
3277
+ {
3278
+ "epoch": 1.25,
3279
+ "learning_rate": 5e-05,
3280
+ "loss": 2.1264,
3281
+ "step": 541000
3282
+ },
3283
+ {
3284
+ "epoch": 1.26,
3285
+ "learning_rate": 5e-05,
3286
+ "loss": 2.1296,
3287
+ "step": 542000
3288
+ },
3289
+ {
3290
+ "epoch": 1.26,
3291
+ "learning_rate": 5e-05,
3292
+ "loss": 2.1277,
3293
+ "step": 543000
3294
+ },
3295
+ {
3296
+ "epoch": 1.26,
3297
+ "learning_rate": 5e-05,
3298
+ "loss": 2.1296,
3299
+ "step": 544000
3300
+ },
3301
+ {
3302
+ "epoch": 1.26,
3303
+ "learning_rate": 5e-05,
3304
+ "loss": 2.1264,
3305
+ "step": 545000
3306
+ },
3307
+ {
3308
+ "epoch": 1.27,
3309
+ "learning_rate": 5e-05,
3310
+ "loss": 2.1244,
3311
+ "step": 546000
3312
+ },
3313
+ {
3314
+ "epoch": 1.27,
3315
+ "learning_rate": 5e-05,
3316
+ "loss": 2.1227,
3317
+ "step": 547000
3318
+ },
3319
+ {
3320
+ "epoch": 1.27,
3321
+ "learning_rate": 5e-05,
3322
+ "loss": 2.1258,
3323
+ "step": 548000
3324
+ },
3325
+ {
3326
+ "epoch": 1.27,
3327
+ "learning_rate": 5e-05,
3328
+ "loss": 2.1238,
3329
+ "step": 549000
3330
+ },
3331
+ {
3332
+ "epoch": 1.27,
3333
+ "learning_rate": 5e-05,
3334
+ "loss": 2.1293,
3335
+ "step": 550000
3336
+ },
3337
+ {
3338
+ "epoch": 1.28,
3339
+ "learning_rate": 5e-05,
3340
+ "loss": 2.1263,
3341
+ "step": 551000
3342
+ },
3343
+ {
3344
+ "epoch": 1.28,
3345
+ "learning_rate": 5e-05,
3346
+ "loss": 2.117,
3347
+ "step": 552000
3348
+ },
3349
+ {
3350
+ "epoch": 1.28,
3351
+ "learning_rate": 5e-05,
3352
+ "loss": 2.1237,
3353
+ "step": 553000
3354
+ },
3355
+ {
3356
+ "epoch": 1.28,
3357
+ "learning_rate": 5e-05,
3358
+ "loss": 2.1251,
3359
+ "step": 554000
3360
+ },
3361
+ {
3362
+ "epoch": 1.29,
3363
+ "learning_rate": 5e-05,
3364
+ "loss": 2.1243,
3365
+ "step": 555000
3366
+ },
3367
+ {
3368
+ "epoch": 1.29,
3369
+ "learning_rate": 5e-05,
3370
+ "loss": 2.1187,
3371
+ "step": 556000
3372
+ },
3373
+ {
3374
+ "epoch": 1.29,
3375
+ "learning_rate": 5e-05,
3376
+ "loss": 2.1179,
3377
+ "step": 557000
3378
+ },
3379
+ {
3380
+ "epoch": 1.29,
3381
+ "learning_rate": 5e-05,
3382
+ "loss": 2.1226,
3383
+ "step": 558000
3384
+ },
3385
+ {
3386
+ "epoch": 1.3,
3387
+ "learning_rate": 5e-05,
3388
+ "loss": 2.1182,
3389
+ "step": 559000
3390
+ },
3391
+ {
3392
+ "epoch": 1.3,
3393
+ "learning_rate": 5e-05,
3394
+ "loss": 2.1159,
3395
+ "step": 560000
3396
+ },
3397
+ {
3398
+ "epoch": 1.3,
3399
+ "learning_rate": 5e-05,
3400
+ "loss": 2.1186,
3401
+ "step": 561000
3402
+ },
3403
+ {
3404
+ "epoch": 1.3,
3405
+ "learning_rate": 5e-05,
3406
+ "loss": 2.116,
3407
+ "step": 562000
3408
+ },
3409
+ {
3410
+ "epoch": 1.3,
3411
+ "learning_rate": 5e-05,
3412
+ "loss": 2.1152,
3413
+ "step": 563000
3414
+ },
3415
+ {
3416
+ "epoch": 1.31,
3417
+ "learning_rate": 5e-05,
3418
+ "loss": 2.1195,
3419
+ "step": 564000
3420
+ },
3421
+ {
3422
+ "epoch": 1.31,
3423
+ "learning_rate": 5e-05,
3424
+ "loss": 2.1144,
3425
+ "step": 565000
3426
+ },
3427
+ {
3428
+ "epoch": 1.31,
3429
+ "learning_rate": 5e-05,
3430
+ "loss": 2.125,
3431
+ "step": 566000
3432
+ },
3433
+ {
3434
+ "epoch": 1.31,
3435
+ "learning_rate": 5e-05,
3436
+ "loss": 2.1127,
3437
+ "step": 567000
3438
+ },
3439
+ {
3440
+ "epoch": 1.32,
3441
+ "learning_rate": 5e-05,
3442
+ "loss": 2.1177,
3443
+ "step": 568000
3444
+ },
3445
+ {
3446
+ "epoch": 1.32,
3447
+ "learning_rate": 5e-05,
3448
+ "loss": 2.1148,
3449
+ "step": 569000
3450
+ },
3451
+ {
3452
+ "epoch": 1.32,
3453
+ "learning_rate": 5e-05,
3454
+ "loss": 2.1185,
3455
+ "step": 570000
3456
+ },
3457
+ {
3458
+ "epoch": 1.32,
3459
+ "learning_rate": 5e-05,
3460
+ "loss": 2.115,
3461
+ "step": 571000
3462
+ },
3463
+ {
3464
+ "epoch": 1.33,
3465
+ "learning_rate": 5e-05,
3466
+ "loss": 2.1127,
3467
+ "step": 572000
3468
+ },
3469
+ {
3470
+ "epoch": 1.33,
3471
+ "learning_rate": 5e-05,
3472
+ "loss": 2.1075,
3473
+ "step": 573000
3474
+ },
3475
+ {
3476
+ "epoch": 1.33,
3477
+ "learning_rate": 5e-05,
3478
+ "loss": 2.111,
3479
+ "step": 574000
3480
+ },
3481
+ {
3482
+ "epoch": 1.33,
3483
+ "learning_rate": 5e-05,
3484
+ "loss": 2.1139,
3485
+ "step": 575000
3486
+ },
3487
+ {
3488
+ "epoch": 1.34,
3489
+ "learning_rate": 5e-05,
3490
+ "loss": 2.1092,
3491
+ "step": 576000
3492
+ },
3493
+ {
3494
+ "epoch": 1.34,
3495
+ "learning_rate": 5e-05,
3496
+ "loss": 2.1126,
3497
+ "step": 577000
3498
+ },
3499
+ {
3500
+ "epoch": 1.34,
3501
+ "learning_rate": 5e-05,
3502
+ "loss": 2.112,
3503
+ "step": 578000
3504
+ },
3505
+ {
3506
+ "epoch": 1.34,
3507
+ "learning_rate": 5e-05,
3508
+ "loss": 2.1093,
3509
+ "step": 579000
3510
+ },
3511
+ {
3512
+ "epoch": 1.34,
3513
+ "learning_rate": 5e-05,
3514
+ "loss": 2.1008,
3515
+ "step": 580000
3516
+ },
3517
+ {
3518
+ "epoch": 1.35,
3519
+ "learning_rate": 5e-05,
3520
+ "loss": 2.1119,
3521
+ "step": 581000
3522
+ },
3523
+ {
3524
+ "epoch": 1.35,
3525
+ "learning_rate": 5e-05,
3526
+ "loss": 2.1105,
3527
+ "step": 582000
3528
+ },
3529
+ {
3530
+ "epoch": 1.35,
3531
+ "learning_rate": 5e-05,
3532
+ "loss": 2.1078,
3533
+ "step": 583000
3534
+ },
3535
+ {
3536
+ "epoch": 1.35,
3537
+ "learning_rate": 5e-05,
3538
+ "loss": 2.1109,
3539
+ "step": 584000
3540
+ },
3541
+ {
3542
+ "epoch": 1.36,
3543
+ "learning_rate": 5e-05,
3544
+ "loss": 2.1093,
3545
+ "step": 585000
3546
+ },
3547
+ {
3548
+ "epoch": 1.36,
3549
+ "learning_rate": 5e-05,
3550
+ "loss": 2.1047,
3551
+ "step": 586000
3552
+ },
3553
+ {
3554
+ "epoch": 1.36,
3555
+ "learning_rate": 5e-05,
3556
+ "loss": 2.1038,
3557
+ "step": 587000
3558
+ },
3559
+ {
3560
+ "epoch": 1.36,
3561
+ "learning_rate": 5e-05,
3562
+ "loss": 2.1074,
3563
+ "step": 588000
3564
+ },
3565
+ {
3566
+ "epoch": 1.37,
3567
+ "learning_rate": 5e-05,
3568
+ "loss": 2.105,
3569
+ "step": 589000
3570
+ },
3571
+ {
3572
+ "epoch": 1.37,
3573
+ "learning_rate": 5e-05,
3574
+ "loss": 2.1114,
3575
+ "step": 590000
3576
+ },
3577
+ {
3578
+ "epoch": 1.37,
3579
+ "learning_rate": 5e-05,
3580
+ "loss": 2.1002,
3581
+ "step": 591000
3582
+ },
3583
+ {
3584
+ "epoch": 1.37,
3585
+ "learning_rate": 5e-05,
3586
+ "loss": 2.0985,
3587
+ "step": 592000
3588
+ },
3589
+ {
3590
+ "epoch": 1.37,
3591
+ "learning_rate": 5e-05,
3592
+ "loss": 2.1004,
3593
+ "step": 593000
3594
+ },
3595
+ {
3596
+ "epoch": 1.38,
3597
+ "learning_rate": 5e-05,
3598
+ "loss": 2.1008,
3599
+ "step": 594000
3600
+ },
3601
+ {
3602
+ "epoch": 1.38,
3603
+ "learning_rate": 5e-05,
3604
+ "loss": 2.0989,
3605
+ "step": 595000
3606
+ },
3607
+ {
3608
+ "epoch": 1.38,
3609
+ "learning_rate": 5e-05,
3610
+ "loss": 2.102,
3611
+ "step": 596000
3612
+ },
3613
+ {
3614
+ "epoch": 1.38,
3615
+ "learning_rate": 5e-05,
3616
+ "loss": 2.1039,
3617
+ "step": 597000
3618
+ },
3619
+ {
3620
+ "epoch": 1.39,
3621
+ "learning_rate": 5e-05,
3622
+ "loss": 2.0943,
3623
+ "step": 598000
3624
+ },
3625
+ {
3626
+ "epoch": 1.39,
3627
+ "learning_rate": 5e-05,
3628
+ "loss": 2.1032,
3629
+ "step": 599000
3630
+ },
3631
+ {
3632
+ "epoch": 1.39,
3633
+ "learning_rate": 5e-05,
3634
+ "loss": 2.1048,
3635
+ "step": 600000
3636
+ },
3637
+ {
3638
+ "epoch": 1.39,
3639
+ "eval_accuracy": 0.6001050830453463,
3640
+ "eval_loss": 1.994140625,
3641
+ "eval_runtime": 4891.4227,
3642
+ "eval_samples_per_second": 114.179,
3643
+ "eval_steps_per_second": 0.892,
3644
+ "step": 600000
3645
+ },
3646
+ {
3647
+ "epoch": 1.39,
3648
+ "learning_rate": 5e-05,
3649
+ "loss": 2.0963,
3650
+ "step": 601000
3651
+ },
3652
+ {
3653
+ "epoch": 1.4,
3654
+ "learning_rate": 5e-05,
3655
+ "loss": 2.1025,
3656
+ "step": 602000
3657
+ },
3658
+ {
3659
+ "epoch": 1.4,
3660
+ "learning_rate": 5e-05,
3661
+ "loss": 2.0995,
3662
+ "step": 603000
3663
+ },
3664
+ {
3665
+ "epoch": 1.4,
3666
+ "learning_rate": 5e-05,
3667
+ "loss": 2.1012,
3668
+ "step": 604000
3669
+ },
3670
+ {
3671
+ "epoch": 1.4,
3672
+ "learning_rate": 5e-05,
3673
+ "loss": 2.0948,
3674
+ "step": 605000
3675
+ },
3676
+ {
3677
+ "epoch": 1.4,
3678
+ "learning_rate": 5e-05,
3679
+ "loss": 2.0975,
3680
+ "step": 606000
3681
+ },
3682
+ {
3683
+ "epoch": 1.41,
3684
+ "learning_rate": 5e-05,
3685
+ "loss": 2.0955,
3686
+ "step": 607000
3687
+ },
3688
+ {
3689
+ "epoch": 1.41,
3690
+ "learning_rate": 5e-05,
3691
+ "loss": 2.0978,
3692
+ "step": 608000
3693
+ },
3694
+ {
3695
+ "epoch": 1.41,
3696
+ "learning_rate": 5e-05,
3697
+ "loss": 2.0966,
3698
+ "step": 609000
3699
+ },
3700
+ {
3701
+ "epoch": 1.41,
3702
+ "learning_rate": 5e-05,
3703
+ "loss": 2.0921,
3704
+ "step": 610000
3705
+ },
3706
+ {
3707
+ "epoch": 1.42,
3708
+ "learning_rate": 5e-05,
3709
+ "loss": 2.092,
3710
+ "step": 611000
3711
+ },
3712
+ {
3713
+ "epoch": 1.42,
3714
+ "learning_rate": 5e-05,
3715
+ "loss": 2.0944,
3716
+ "step": 612000
3717
+ },
3718
+ {
3719
+ "epoch": 1.42,
3720
+ "learning_rate": 5e-05,
3721
+ "loss": 2.095,
3722
+ "step": 613000
3723
+ },
3724
+ {
3725
+ "epoch": 1.42,
3726
+ "learning_rate": 5e-05,
3727
+ "loss": 2.0939,
3728
+ "step": 614000
3729
+ },
3730
+ {
3731
+ "epoch": 1.43,
3732
+ "learning_rate": 5e-05,
3733
+ "loss": 2.0952,
3734
+ "step": 615000
3735
+ },
3736
+ {
3737
+ "epoch": 1.43,
3738
+ "learning_rate": 5e-05,
3739
+ "loss": 2.0928,
3740
+ "step": 616000
3741
+ },
3742
+ {
3743
+ "epoch": 1.43,
3744
+ "learning_rate": 5e-05,
3745
+ "loss": 2.0932,
3746
+ "step": 617000
3747
+ },
3748
+ {
3749
+ "epoch": 1.43,
3750
+ "learning_rate": 5e-05,
3751
+ "loss": 2.0881,
3752
+ "step": 618000
3753
+ },
3754
+ {
3755
+ "epoch": 1.43,
3756
+ "learning_rate": 5e-05,
3757
+ "loss": 2.0912,
3758
+ "step": 619000
3759
+ },
3760
+ {
3761
+ "epoch": 1.44,
3762
+ "learning_rate": 5e-05,
3763
+ "loss": 2.0909,
3764
+ "step": 620000
3765
+ },
3766
+ {
3767
+ "epoch": 1.44,
3768
+ "learning_rate": 5e-05,
3769
+ "loss": 2.0856,
3770
+ "step": 621000
3771
+ },
3772
+ {
3773
+ "epoch": 1.44,
3774
+ "learning_rate": 5e-05,
3775
+ "loss": 2.0906,
3776
+ "step": 622000
3777
+ },
3778
+ {
3779
+ "epoch": 1.44,
3780
+ "learning_rate": 5e-05,
3781
+ "loss": 2.0905,
3782
+ "step": 623000
3783
+ },
3784
+ {
3785
+ "epoch": 1.45,
3786
+ "learning_rate": 5e-05,
3787
+ "loss": 2.0824,
3788
+ "step": 624000
3789
+ },
3790
+ {
3791
+ "epoch": 1.45,
3792
+ "learning_rate": 5e-05,
3793
+ "loss": 2.0874,
3794
+ "step": 625000
3795
+ },
3796
+ {
3797
+ "epoch": 1.45,
3798
+ "learning_rate": 5e-05,
3799
+ "loss": 2.0894,
3800
+ "step": 626000
3801
+ },
3802
+ {
3803
+ "epoch": 1.45,
3804
+ "learning_rate": 5e-05,
3805
+ "loss": 2.0824,
3806
+ "step": 627000
3807
+ },
3808
+ {
3809
+ "epoch": 1.46,
3810
+ "learning_rate": 5e-05,
3811
+ "loss": 2.0874,
3812
+ "step": 628000
3813
+ },
3814
+ {
3815
+ "epoch": 1.46,
3816
+ "learning_rate": 5e-05,
3817
+ "loss": 2.088,
3818
+ "step": 629000
3819
+ },
3820
+ {
3821
+ "epoch": 1.46,
3822
+ "learning_rate": 5e-05,
3823
+ "loss": 2.088,
3824
+ "step": 630000
3825
+ },
3826
+ {
3827
+ "epoch": 1.46,
3828
+ "learning_rate": 5e-05,
3829
+ "loss": 2.0815,
3830
+ "step": 631000
3831
+ },
3832
+ {
3833
+ "epoch": 1.46,
3834
+ "learning_rate": 5e-05,
3835
+ "loss": 2.0835,
3836
+ "step": 632000
3837
+ },
3838
+ {
3839
+ "epoch": 1.47,
3840
+ "learning_rate": 5e-05,
3841
+ "loss": 2.0886,
3842
+ "step": 633000
3843
+ },
3844
+ {
3845
+ "epoch": 1.47,
3846
+ "learning_rate": 5e-05,
3847
+ "loss": 2.0842,
3848
+ "step": 634000
3849
+ },
3850
+ {
3851
+ "epoch": 1.47,
3852
+ "learning_rate": 5e-05,
3853
+ "loss": 2.0814,
3854
+ "step": 635000
3855
+ },
3856
+ {
3857
+ "epoch": 1.47,
3858
+ "learning_rate": 5e-05,
3859
+ "loss": 2.0822,
3860
+ "step": 636000
3861
+ },
3862
+ {
3863
+ "epoch": 1.48,
3864
+ "learning_rate": 5e-05,
3865
+ "loss": 2.0811,
3866
+ "step": 637000
3867
+ },
3868
+ {
3869
+ "epoch": 1.48,
3870
+ "learning_rate": 5e-05,
3871
+ "loss": 2.0883,
3872
+ "step": 638000
3873
+ },
3874
+ {
3875
+ "epoch": 1.48,
3876
+ "learning_rate": 5e-05,
3877
+ "loss": 2.0858,
3878
+ "step": 639000
3879
+ },
3880
+ {
3881
+ "epoch": 1.48,
3882
+ "learning_rate": 5e-05,
3883
+ "loss": 2.0841,
3884
+ "step": 640000
3885
+ },
3886
+ {
3887
+ "epoch": 1.49,
3888
+ "learning_rate": 5e-05,
3889
+ "loss": 2.0893,
3890
+ "step": 641000
3891
+ },
3892
+ {
3893
+ "epoch": 1.49,
3894
+ "learning_rate": 5e-05,
3895
+ "loss": 2.0807,
3896
+ "step": 642000
3897
+ },
3898
+ {
3899
+ "epoch": 1.49,
3900
+ "learning_rate": 5e-05,
3901
+ "loss": 2.0841,
3902
+ "step": 643000
3903
+ },
3904
+ {
3905
+ "epoch": 1.49,
3906
+ "learning_rate": 5e-05,
3907
+ "loss": 2.0803,
3908
+ "step": 644000
3909
+ },
3910
+ {
3911
+ "epoch": 1.5,
3912
+ "learning_rate": 5e-05,
3913
+ "loss": 2.086,
3914
+ "step": 645000
3915
+ },
3916
+ {
3917
+ "epoch": 1.5,
3918
+ "learning_rate": 5e-05,
3919
+ "loss": 2.0816,
3920
+ "step": 646000
3921
+ },
3922
+ {
3923
+ "epoch": 1.5,
3924
+ "learning_rate": 5e-05,
3925
+ "loss": 2.083,
3926
+ "step": 647000
3927
+ },
3928
+ {
3929
+ "epoch": 1.5,
3930
+ "learning_rate": 5e-05,
3931
+ "loss": 2.0751,
3932
+ "step": 648000
3933
+ },
3934
+ {
3935
+ "epoch": 1.5,
3936
+ "learning_rate": 5e-05,
3937
+ "loss": 2.0836,
3938
+ "step": 649000
3939
+ },
3940
+ {
3941
+ "epoch": 1.51,
3942
+ "learning_rate": 5e-05,
3943
+ "loss": 2.0757,
3944
+ "step": 650000
3945
+ },
3946
+ {
3947
+ "epoch": 1.51,
3948
+ "learning_rate": 5e-05,
3949
+ "loss": 2.0795,
3950
+ "step": 651000
3951
+ },
3952
+ {
3953
+ "epoch": 1.51,
3954
+ "learning_rate": 5e-05,
3955
+ "loss": 2.0755,
3956
+ "step": 652000
3957
+ },
3958
+ {
3959
+ "epoch": 1.51,
3960
+ "learning_rate": 5e-05,
3961
+ "loss": 2.0811,
3962
+ "step": 653000
3963
+ },
3964
+ {
3965
+ "epoch": 1.52,
3966
+ "learning_rate": 5e-05,
3967
+ "loss": 2.0769,
3968
+ "step": 654000
3969
+ },
3970
+ {
3971
+ "epoch": 1.52,
3972
+ "learning_rate": 5e-05,
3973
+ "loss": 2.0737,
3974
+ "step": 655000
3975
+ },
3976
+ {
3977
+ "epoch": 1.52,
3978
+ "learning_rate": 5e-05,
3979
+ "loss": 2.0803,
3980
+ "step": 656000
3981
+ },
3982
+ {
3983
+ "epoch": 1.52,
3984
+ "learning_rate": 5e-05,
3985
+ "loss": 2.0739,
3986
+ "step": 657000
3987
+ },
3988
+ {
3989
+ "epoch": 1.53,
3990
+ "learning_rate": 5e-05,
3991
+ "loss": 2.0783,
3992
+ "step": 658000
3993
+ },
3994
+ {
3995
+ "epoch": 1.53,
3996
+ "learning_rate": 5e-05,
3997
+ "loss": 2.0744,
3998
+ "step": 659000
3999
+ },
4000
+ {
4001
+ "epoch": 1.53,
4002
+ "learning_rate": 5e-05,
4003
+ "loss": 2.0696,
4004
+ "step": 660000
4005
+ },
4006
+ {
4007
+ "epoch": 1.53,
4008
+ "learning_rate": 5e-05,
4009
+ "loss": 2.0785,
4010
+ "step": 661000
4011
+ },
4012
+ {
4013
+ "epoch": 1.53,
4014
+ "learning_rate": 5e-05,
4015
+ "loss": 2.0742,
4016
+ "step": 662000
4017
+ },
4018
+ {
4019
+ "epoch": 1.54,
4020
+ "learning_rate": 5e-05,
4021
+ "loss": 2.0715,
4022
+ "step": 663000
4023
+ },
4024
+ {
4025
+ "epoch": 1.54,
4026
+ "learning_rate": 5e-05,
4027
+ "loss": 2.0705,
4028
+ "step": 664000
4029
+ },
4030
+ {
4031
+ "epoch": 1.54,
4032
+ "learning_rate": 5e-05,
4033
+ "loss": 2.0698,
4034
+ "step": 665000
4035
+ },
4036
+ {
4037
+ "epoch": 1.54,
4038
+ "learning_rate": 5e-05,
4039
+ "loss": 2.0703,
4040
+ "step": 666000
4041
+ },
4042
+ {
4043
+ "epoch": 1.55,
4044
+ "learning_rate": 5e-05,
4045
+ "loss": 2.075,
4046
+ "step": 667000
4047
+ },
4048
+ {
4049
+ "epoch": 1.55,
4050
+ "learning_rate": 5e-05,
4051
+ "loss": 2.0694,
4052
+ "step": 668000
4053
+ },
4054
+ {
4055
+ "epoch": 1.55,
4056
+ "learning_rate": 5e-05,
4057
+ "loss": 2.0708,
4058
+ "step": 669000
4059
+ },
4060
+ {
4061
+ "epoch": 1.55,
4062
+ "learning_rate": 5e-05,
4063
+ "loss": 2.0656,
4064
+ "step": 670000
4065
+ },
4066
+ {
4067
+ "epoch": 1.56,
4068
+ "learning_rate": 5e-05,
4069
+ "loss": 2.069,
4070
+ "step": 671000
4071
+ },
4072
+ {
4073
+ "epoch": 1.56,
4074
+ "learning_rate": 5e-05,
4075
+ "loss": 2.0724,
4076
+ "step": 672000
4077
+ },
4078
+ {
4079
+ "epoch": 1.56,
4080
+ "learning_rate": 5e-05,
4081
+ "loss": 2.0699,
4082
+ "step": 673000
4083
+ },
4084
+ {
4085
+ "epoch": 1.56,
4086
+ "learning_rate": 5e-05,
4087
+ "loss": 2.0695,
4088
+ "step": 674000
4089
+ },
4090
+ {
4091
+ "epoch": 1.56,
4092
+ "learning_rate": 5e-05,
4093
+ "loss": 2.068,
4094
+ "step": 675000
4095
+ },
4096
+ {
4097
+ "epoch": 1.57,
4098
+ "learning_rate": 5e-05,
4099
+ "loss": 2.0671,
4100
+ "step": 676000
4101
+ },
4102
+ {
4103
+ "epoch": 1.57,
4104
+ "learning_rate": 5e-05,
4105
+ "loss": 2.0716,
4106
+ "step": 677000
4107
+ },
4108
+ {
4109
+ "epoch": 1.57,
4110
+ "learning_rate": 5e-05,
4111
+ "loss": 2.0689,
4112
+ "step": 678000
4113
+ },
4114
+ {
4115
+ "epoch": 1.57,
4116
+ "learning_rate": 5e-05,
4117
+ "loss": 2.062,
4118
+ "step": 679000
4119
+ },
4120
+ {
4121
+ "epoch": 1.58,
4122
+ "learning_rate": 5e-05,
4123
+ "loss": 2.0733,
4124
+ "step": 680000
4125
+ },
4126
+ {
4127
+ "epoch": 1.58,
4128
+ "learning_rate": 5e-05,
4129
+ "loss": 2.0671,
4130
+ "step": 681000
4131
+ },
4132
+ {
4133
+ "epoch": 1.58,
4134
+ "learning_rate": 5e-05,
4135
+ "loss": 2.0676,
4136
+ "step": 682000
4137
+ },
4138
+ {
4139
+ "epoch": 1.58,
4140
+ "learning_rate": 5e-05,
4141
+ "loss": 2.0651,
4142
+ "step": 683000
4143
+ },
4144
+ {
4145
+ "epoch": 1.59,
4146
+ "learning_rate": 5e-05,
4147
+ "loss": 2.0594,
4148
+ "step": 684000
4149
+ },
4150
+ {
4151
+ "epoch": 1.59,
4152
+ "learning_rate": 5e-05,
4153
+ "loss": 2.0658,
4154
+ "step": 685000
4155
+ },
4156
+ {
4157
+ "epoch": 1.59,
4158
+ "learning_rate": 5e-05,
4159
+ "loss": 2.0686,
4160
+ "step": 686000
4161
+ },
4162
+ {
4163
+ "epoch": 1.59,
4164
+ "learning_rate": 5e-05,
4165
+ "loss": 2.0685,
4166
+ "step": 687000
4167
+ },
4168
+ {
4169
+ "epoch": 1.59,
4170
+ "learning_rate": 5e-05,
4171
+ "loss": 2.0632,
4172
+ "step": 688000
4173
+ },
4174
+ {
4175
+ "epoch": 1.6,
4176
+ "learning_rate": 5e-05,
4177
+ "loss": 2.0662,
4178
+ "step": 689000
4179
+ },
4180
+ {
4181
+ "epoch": 1.6,
4182
+ "learning_rate": 5e-05,
4183
+ "loss": 2.0615,
4184
+ "step": 690000
4185
+ },
4186
+ {
4187
+ "epoch": 1.6,
4188
+ "learning_rate": 5e-05,
4189
+ "loss": 2.0546,
4190
+ "step": 691000
4191
+ },
4192
+ {
4193
+ "epoch": 1.6,
4194
+ "learning_rate": 5e-05,
4195
+ "loss": 2.0593,
4196
+ "step": 692000
4197
+ },
4198
+ {
4199
+ "epoch": 1.61,
4200
+ "learning_rate": 5e-05,
4201
+ "loss": 2.0599,
4202
+ "step": 693000
4203
+ },
4204
+ {
4205
+ "epoch": 1.61,
4206
+ "learning_rate": 5e-05,
4207
+ "loss": 2.0635,
4208
+ "step": 694000
4209
+ },
4210
+ {
4211
+ "epoch": 1.61,
4212
+ "learning_rate": 5e-05,
4213
+ "loss": 2.0621,
4214
+ "step": 695000
4215
+ },
4216
+ {
4217
+ "epoch": 1.61,
4218
+ "learning_rate": 5e-05,
4219
+ "loss": 2.0654,
4220
+ "step": 696000
4221
+ },
4222
+ {
4223
+ "epoch": 1.62,
4224
+ "learning_rate": 5e-05,
4225
+ "loss": 2.0576,
4226
+ "step": 697000
4227
+ },
4228
+ {
4229
+ "epoch": 1.62,
4230
+ "learning_rate": 5e-05,
4231
+ "loss": 2.0563,
4232
+ "step": 698000
4233
+ },
4234
+ {
4235
+ "epoch": 1.62,
4236
+ "learning_rate": 5e-05,
4237
+ "loss": 2.059,
4238
+ "step": 699000
4239
+ },
4240
+ {
4241
+ "epoch": 1.62,
4242
+ "learning_rate": 5e-05,
4243
+ "loss": 2.0616,
4244
+ "step": 700000
4245
+ },
4246
+ {
4247
+ "epoch": 1.62,
4248
+ "learning_rate": 5e-05,
4249
+ "loss": 2.0642,
4250
+ "step": 701000
4251
+ },
4252
+ {
4253
+ "epoch": 1.63,
4254
+ "learning_rate": 5e-05,
4255
+ "loss": 2.052,
4256
+ "step": 702000
4257
+ },
4258
+ {
4259
+ "epoch": 1.63,
4260
+ "learning_rate": 5e-05,
4261
+ "loss": 2.0569,
4262
+ "step": 703000
4263
+ },
4264
+ {
4265
+ "epoch": 1.63,
4266
+ "learning_rate": 5e-05,
4267
+ "loss": 2.0617,
4268
+ "step": 704000
4269
+ },
4270
+ {
4271
+ "epoch": 1.63,
4272
+ "learning_rate": 5e-05,
4273
+ "loss": 2.0613,
4274
+ "step": 705000
4275
+ },
4276
+ {
4277
+ "epoch": 1.64,
4278
+ "learning_rate": 5e-05,
4279
+ "loss": 2.0593,
4280
+ "step": 706000
4281
+ },
4282
+ {
4283
+ "epoch": 1.64,
4284
+ "learning_rate": 5e-05,
4285
+ "loss": 2.0582,
4286
+ "step": 707000
4287
+ },
4288
+ {
4289
+ "epoch": 1.64,
4290
+ "learning_rate": 5e-05,
4291
+ "loss": 2.0587,
4292
+ "step": 708000
4293
+ },
4294
+ {
4295
+ "epoch": 1.64,
4296
+ "learning_rate": 5e-05,
4297
+ "loss": 2.0518,
4298
+ "step": 709000
4299
+ },
4300
+ {
4301
+ "epoch": 1.65,
4302
+ "learning_rate": 5e-05,
4303
+ "loss": 2.0585,
4304
+ "step": 710000
4305
+ },
4306
+ {
4307
+ "epoch": 1.65,
4308
+ "learning_rate": 5e-05,
4309
+ "loss": 2.052,
4310
+ "step": 711000
4311
+ },
4312
+ {
4313
+ "epoch": 1.65,
4314
+ "learning_rate": 5e-05,
4315
+ "loss": 2.056,
4316
+ "step": 712000
4317
+ },
4318
+ {
4319
+ "epoch": 1.65,
4320
+ "learning_rate": 5e-05,
4321
+ "loss": 2.0559,
4322
+ "step": 713000
4323
+ },
4324
+ {
4325
+ "epoch": 1.65,
4326
+ "learning_rate": 5e-05,
4327
+ "loss": 2.0538,
4328
+ "step": 714000
4329
+ },
4330
+ {
4331
+ "epoch": 1.66,
4332
+ "learning_rate": 5e-05,
4333
+ "loss": 2.0566,
4334
+ "step": 715000
4335
+ },
4336
+ {
4337
+ "epoch": 1.66,
4338
+ "learning_rate": 5e-05,
4339
+ "loss": 2.0519,
4340
+ "step": 716000
4341
+ },
4342
+ {
4343
+ "epoch": 1.66,
4344
+ "learning_rate": 5e-05,
4345
+ "loss": 2.058,
4346
+ "step": 717000
4347
+ },
4348
+ {
4349
+ "epoch": 1.66,
4350
+ "learning_rate": 5e-05,
4351
+ "loss": 2.0627,
4352
+ "step": 718000
4353
+ },
4354
+ {
4355
+ "epoch": 1.67,
4356
+ "learning_rate": 5e-05,
4357
+ "loss": 2.0476,
4358
+ "step": 719000
4359
+ },
4360
+ {
4361
+ "epoch": 1.67,
4362
+ "learning_rate": 5e-05,
4363
+ "loss": 2.0484,
4364
+ "step": 720000
4365
+ },
4366
+ {
4367
+ "epoch": 1.67,
4368
+ "learning_rate": 5e-05,
4369
+ "loss": 2.0545,
4370
+ "step": 721000
4371
+ },
4372
+ {
4373
+ "epoch": 1.67,
4374
+ "learning_rate": 5e-05,
4375
+ "loss": 2.0558,
4376
+ "step": 722000
4377
+ },
4378
+ {
4379
+ "epoch": 1.68,
4380
+ "learning_rate": 5e-05,
4381
+ "loss": 2.0547,
4382
+ "step": 723000
4383
+ },
4384
+ {
4385
+ "epoch": 1.68,
4386
+ "learning_rate": 5e-05,
4387
+ "loss": 2.0501,
4388
+ "step": 724000
4389
+ },
4390
+ {
4391
+ "epoch": 1.68,
4392
+ "learning_rate": 5e-05,
4393
+ "loss": 2.0562,
4394
+ "step": 725000
4395
+ },
4396
+ {
4397
+ "epoch": 1.68,
4398
+ "learning_rate": 5e-05,
4399
+ "loss": 2.0467,
4400
+ "step": 726000
4401
+ },
4402
+ {
4403
+ "epoch": 1.69,
4404
+ "learning_rate": 5e-05,
4405
+ "loss": 2.0438,
4406
+ "step": 727000
4407
+ },
4408
+ {
4409
+ "epoch": 1.69,
4410
+ "learning_rate": 5e-05,
4411
+ "loss": 2.0458,
4412
+ "step": 728000
4413
+ },
4414
+ {
4415
+ "epoch": 1.69,
4416
+ "learning_rate": 5e-05,
4417
+ "loss": 2.0504,
4418
+ "step": 729000
4419
+ },
4420
+ {
4421
+ "epoch": 1.69,
4422
+ "learning_rate": 5e-05,
4423
+ "loss": 2.0501,
4424
+ "step": 730000
4425
+ },
4426
+ {
4427
+ "epoch": 1.69,
4428
+ "learning_rate": 5e-05,
4429
+ "loss": 2.0498,
4430
+ "step": 731000
4431
+ },
4432
+ {
4433
+ "epoch": 1.7,
4434
+ "learning_rate": 5e-05,
4435
+ "loss": 2.0495,
4436
+ "step": 732000
4437
+ },
4438
+ {
4439
+ "epoch": 1.7,
4440
+ "learning_rate": 5e-05,
4441
+ "loss": 2.0494,
4442
+ "step": 733000
4443
+ },
4444
+ {
4445
+ "epoch": 1.7,
4446
+ "learning_rate": 5e-05,
4447
+ "loss": 2.0469,
4448
+ "step": 734000
4449
+ },
4450
+ {
4451
+ "epoch": 1.7,
4452
+ "learning_rate": 5e-05,
4453
+ "loss": 2.0503,
4454
+ "step": 735000
4455
+ },
4456
+ {
4457
+ "epoch": 1.71,
4458
+ "learning_rate": 5e-05,
4459
+ "loss": 2.0492,
4460
+ "step": 736000
4461
+ },
4462
+ {
4463
+ "epoch": 1.71,
4464
+ "learning_rate": 5e-05,
4465
+ "loss": 2.0445,
4466
+ "step": 737000
4467
+ },
4468
+ {
4469
+ "epoch": 1.71,
4470
+ "learning_rate": 5e-05,
4471
+ "loss": 2.044,
4472
+ "step": 738000
4473
+ },
4474
+ {
4475
+ "epoch": 1.71,
4476
+ "learning_rate": 5e-05,
4477
+ "loss": 2.0418,
4478
+ "step": 739000
4479
+ },
4480
+ {
4481
+ "epoch": 1.72,
4482
+ "learning_rate": 5e-05,
4483
+ "loss": 2.0473,
4484
+ "step": 740000
4485
+ },
4486
+ {
4487
+ "epoch": 1.72,
4488
+ "learning_rate": 5e-05,
4489
+ "loss": 2.0503,
4490
+ "step": 741000
4491
+ },
4492
+ {
4493
+ "epoch": 1.72,
4494
+ "learning_rate": 5e-05,
4495
+ "loss": 2.0513,
4496
+ "step": 742000
4497
+ },
4498
+ {
4499
+ "epoch": 1.72,
4500
+ "learning_rate": 5e-05,
4501
+ "loss": 2.0492,
4502
+ "step": 743000
4503
+ },
4504
+ {
4505
+ "epoch": 1.72,
4506
+ "learning_rate": 5e-05,
4507
+ "loss": 2.0471,
4508
+ "step": 744000
4509
+ },
4510
+ {
4511
+ "epoch": 1.73,
4512
+ "learning_rate": 5e-05,
4513
+ "loss": 2.0357,
4514
+ "step": 745000
4515
+ },
4516
+ {
4517
+ "epoch": 1.73,
4518
+ "learning_rate": 5e-05,
4519
+ "loss": 2.045,
4520
+ "step": 746000
4521
+ },
4522
+ {
4523
+ "epoch": 1.73,
4524
+ "learning_rate": 5e-05,
4525
+ "loss": 2.042,
4526
+ "step": 747000
4527
+ },
4528
+ {
4529
+ "epoch": 1.73,
4530
+ "learning_rate": 5e-05,
4531
+ "loss": 2.0485,
4532
+ "step": 748000
4533
+ },
4534
+ {
4535
+ "epoch": 1.74,
4536
+ "learning_rate": 5e-05,
4537
+ "loss": 2.043,
4538
+ "step": 749000
4539
+ },
4540
+ {
4541
+ "epoch": 1.74,
4542
+ "learning_rate": 5e-05,
4543
+ "loss": 2.0345,
4544
+ "step": 750000
4545
+ },
4546
+ {
4547
+ "epoch": 1.74,
4548
+ "eval_accuracy": 0.6086964451170365,
4549
+ "eval_loss": 1.935546875,
4550
+ "eval_runtime": 4905.4193,
4551
+ "eval_samples_per_second": 113.853,
4552
+ "eval_steps_per_second": 0.89,
4553
+ "step": 750000
4554
+ },
4555
+ {
4556
+ "epoch": 1.74,
4557
+ "learning_rate": 5e-05,
4558
+ "loss": 2.0407,
4559
+ "step": 751000
4560
+ },
4561
+ {
4562
+ "epoch": 1.74,
4563
+ "learning_rate": 5e-05,
4564
+ "loss": 2.042,
4565
+ "step": 752000
4566
+ },
4567
+ {
4568
+ "epoch": 1.75,
4569
+ "learning_rate": 5e-05,
4570
+ "loss": 2.0446,
4571
+ "step": 753000
4572
+ },
4573
+ {
4574
+ "epoch": 1.75,
4575
+ "learning_rate": 5e-05,
4576
+ "loss": 2.0447,
4577
+ "step": 754000
4578
+ },
4579
+ {
4580
+ "epoch": 1.75,
4581
+ "learning_rate": 5e-05,
4582
+ "loss": 2.0402,
4583
+ "step": 755000
4584
+ },
4585
+ {
4586
+ "epoch": 1.75,
4587
+ "learning_rate": 5e-05,
4588
+ "loss": 2.0392,
4589
+ "step": 756000
4590
+ },
4591
+ {
4592
+ "epoch": 1.75,
4593
+ "learning_rate": 5e-05,
4594
+ "loss": 2.037,
4595
+ "step": 757000
4596
+ },
4597
+ {
4598
+ "epoch": 1.76,
4599
+ "learning_rate": 5e-05,
4600
+ "loss": 2.0341,
4601
+ "step": 758000
4602
+ },
4603
+ {
4604
+ "epoch": 1.76,
4605
+ "learning_rate": 5e-05,
4606
+ "loss": 2.0381,
4607
+ "step": 759000
4608
+ },
4609
+ {
4610
+ "epoch": 1.76,
4611
+ "learning_rate": 5e-05,
4612
+ "loss": 2.0374,
4613
+ "step": 760000
4614
+ },
4615
+ {
4616
+ "epoch": 1.76,
4617
+ "learning_rate": 5e-05,
4618
+ "loss": 2.0347,
4619
+ "step": 761000
4620
+ },
4621
+ {
4622
+ "epoch": 1.77,
4623
+ "learning_rate": 5e-05,
4624
+ "loss": 2.0345,
4625
+ "step": 762000
4626
+ },
4627
+ {
4628
+ "epoch": 1.77,
4629
+ "learning_rate": 5e-05,
4630
+ "loss": 2.036,
4631
+ "step": 763000
4632
+ },
4633
+ {
4634
+ "epoch": 1.77,
4635
+ "learning_rate": 5e-05,
4636
+ "loss": 2.0357,
4637
+ "step": 764000
4638
+ },
4639
+ {
4640
+ "epoch": 1.77,
4641
+ "learning_rate": 5e-05,
4642
+ "loss": 2.0386,
4643
+ "step": 765000
4644
+ },
4645
+ {
4646
+ "epoch": 1.78,
4647
+ "learning_rate": 5e-05,
4648
+ "loss": 2.0363,
4649
+ "step": 766000
4650
+ },
4651
+ {
4652
+ "epoch": 1.78,
4653
+ "learning_rate": 5e-05,
4654
+ "loss": 2.0406,
4655
+ "step": 767000
4656
+ },
4657
+ {
4658
+ "epoch": 1.78,
4659
+ "learning_rate": 5e-05,
4660
+ "loss": 2.0332,
4661
+ "step": 768000
4662
+ },
4663
+ {
4664
+ "epoch": 1.78,
4665
+ "learning_rate": 5e-05,
4666
+ "loss": 2.0358,
4667
+ "step": 769000
4668
+ },
4669
+ {
4670
+ "epoch": 1.78,
4671
+ "learning_rate": 5e-05,
4672
+ "loss": 2.0359,
4673
+ "step": 770000
4674
+ },
4675
+ {
4676
+ "epoch": 1.79,
4677
+ "learning_rate": 5e-05,
4678
+ "loss": 2.0429,
4679
+ "step": 771000
4680
+ },
4681
+ {
4682
+ "epoch": 1.79,
4683
+ "learning_rate": 5e-05,
4684
+ "loss": 2.0402,
4685
+ "step": 772000
4686
+ },
4687
+ {
4688
+ "epoch": 1.79,
4689
+ "learning_rate": 5e-05,
4690
+ "loss": 2.0372,
4691
+ "step": 773000
4692
+ },
4693
+ {
4694
+ "epoch": 1.79,
4695
+ "learning_rate": 5e-05,
4696
+ "loss": 2.0343,
4697
+ "step": 774000
4698
+ },
4699
+ {
4700
+ "epoch": 1.8,
4701
+ "learning_rate": 5e-05,
4702
+ "loss": 2.0351,
4703
+ "step": 775000
4704
+ },
4705
+ {
4706
+ "epoch": 1.8,
4707
+ "learning_rate": 5e-05,
4708
+ "loss": 2.0366,
4709
+ "step": 776000
4710
+ },
4711
+ {
4712
+ "epoch": 1.8,
4713
+ "learning_rate": 5e-05,
4714
+ "loss": 2.0364,
4715
+ "step": 777000
4716
+ },
4717
+ {
4718
+ "epoch": 1.8,
4719
+ "learning_rate": 5e-05,
4720
+ "loss": 2.034,
4721
+ "step": 778000
4722
+ },
4723
+ {
4724
+ "epoch": 1.81,
4725
+ "learning_rate": 5e-05,
4726
+ "loss": 2.0313,
4727
+ "step": 779000
4728
+ },
4729
+ {
4730
+ "epoch": 1.81,
4731
+ "learning_rate": 5e-05,
4732
+ "loss": 2.0281,
4733
+ "step": 780000
4734
+ },
4735
+ {
4736
+ "epoch": 1.81,
4737
+ "learning_rate": 5e-05,
4738
+ "loss": 2.0316,
4739
+ "step": 781000
4740
+ },
4741
+ {
4742
+ "epoch": 1.81,
4743
+ "learning_rate": 5e-05,
4744
+ "loss": 2.0298,
4745
+ "step": 782000
4746
+ },
4747
+ {
4748
+ "epoch": 1.81,
4749
+ "learning_rate": 5e-05,
4750
+ "loss": 2.0292,
4751
+ "step": 783000
4752
+ },
4753
+ {
4754
+ "epoch": 1.82,
4755
+ "learning_rate": 5e-05,
4756
+ "loss": 2.0269,
4757
+ "step": 784000
4758
+ },
4759
+ {
4760
+ "epoch": 1.82,
4761
+ "learning_rate": 5e-05,
4762
+ "loss": 2.0309,
4763
+ "step": 785000
4764
+ },
4765
+ {
4766
+ "epoch": 1.82,
4767
+ "learning_rate": 5e-05,
4768
+ "loss": 2.0316,
4769
+ "step": 786000
4770
+ },
4771
+ {
4772
+ "epoch": 1.82,
4773
+ "learning_rate": 5e-05,
4774
+ "loss": 2.0379,
4775
+ "step": 787000
4776
+ },
4777
+ {
4778
+ "epoch": 1.83,
4779
+ "learning_rate": 5e-05,
4780
+ "loss": 2.0298,
4781
+ "step": 788000
4782
+ },
4783
+ {
4784
+ "epoch": 1.83,
4785
+ "learning_rate": 5e-05,
4786
+ "loss": 2.028,
4787
+ "step": 789000
4788
+ },
4789
+ {
4790
+ "epoch": 1.83,
4791
+ "learning_rate": 5e-05,
4792
+ "loss": 2.0278,
4793
+ "step": 790000
4794
+ },
4795
+ {
4796
+ "epoch": 1.83,
4797
+ "learning_rate": 5e-05,
4798
+ "loss": 2.0316,
4799
+ "step": 791000
4800
+ },
4801
+ {
4802
+ "epoch": 1.84,
4803
+ "learning_rate": 5e-05,
4804
+ "loss": 2.0269,
4805
+ "step": 792000
4806
+ },
4807
+ {
4808
+ "epoch": 1.84,
4809
+ "learning_rate": 5e-05,
4810
+ "loss": 2.031,
4811
+ "step": 793000
4812
+ },
4813
+ {
4814
+ "epoch": 1.84,
4815
+ "learning_rate": 5e-05,
4816
+ "loss": 2.0281,
4817
+ "step": 794000
4818
+ },
4819
+ {
4820
+ "epoch": 1.84,
4821
+ "learning_rate": 5e-05,
4822
+ "loss": 2.0237,
4823
+ "step": 795000
4824
+ },
4825
+ {
4826
+ "epoch": 1.85,
4827
+ "learning_rate": 5e-05,
4828
+ "loss": 2.0244,
4829
+ "step": 796000
4830
+ },
4831
+ {
4832
+ "epoch": 1.85,
4833
+ "learning_rate": 5e-05,
4834
+ "loss": 2.0262,
4835
+ "step": 797000
4836
+ },
4837
+ {
4838
+ "epoch": 1.85,
4839
+ "learning_rate": 5e-05,
4840
+ "loss": 2.0275,
4841
+ "step": 798000
4842
+ },
4843
+ {
4844
+ "epoch": 1.85,
4845
+ "learning_rate": 5e-05,
4846
+ "loss": 2.0253,
4847
+ "step": 799000
4848
+ },
4849
+ {
4850
+ "epoch": 1.85,
4851
+ "learning_rate": 5e-05,
4852
+ "loss": 2.0336,
4853
+ "step": 800000
4854
+ },
4855
+ {
4856
+ "epoch": 1.86,
4857
+ "learning_rate": 5e-05,
4858
+ "loss": 2.029,
4859
+ "step": 801000
4860
+ },
4861
+ {
4862
+ "epoch": 1.86,
4863
+ "learning_rate": 5e-05,
4864
+ "loss": 2.0271,
4865
+ "step": 802000
4866
+ },
4867
+ {
4868
+ "epoch": 1.86,
4869
+ "learning_rate": 5e-05,
4870
+ "loss": 2.0221,
4871
+ "step": 803000
4872
+ },
4873
+ {
4874
+ "epoch": 1.86,
4875
+ "learning_rate": 5e-05,
4876
+ "loss": 2.0287,
4877
+ "step": 804000
4878
+ },
4879
+ {
4880
+ "epoch": 1.87,
4881
+ "learning_rate": 5e-05,
4882
+ "loss": 2.0264,
4883
+ "step": 805000
4884
+ },
4885
+ {
4886
+ "epoch": 1.87,
4887
+ "learning_rate": 5e-05,
4888
+ "loss": 2.0263,
4889
+ "step": 806000
4890
+ },
4891
+ {
4892
+ "epoch": 1.87,
4893
+ "learning_rate": 5e-05,
4894
+ "loss": 2.0242,
4895
+ "step": 807000
4896
+ },
4897
+ {
4898
+ "epoch": 1.87,
4899
+ "learning_rate": 5e-05,
4900
+ "loss": 2.0221,
4901
+ "step": 808000
4902
+ },
4903
+ {
4904
+ "epoch": 1.88,
4905
+ "learning_rate": 5e-05,
4906
+ "loss": 2.0242,
4907
+ "step": 809000
4908
+ },
4909
+ {
4910
+ "epoch": 1.88,
4911
+ "learning_rate": 5e-05,
4912
+ "loss": 2.0212,
4913
+ "step": 810000
4914
+ },
4915
+ {
4916
+ "epoch": 1.88,
4917
+ "learning_rate": 5e-05,
4918
+ "loss": 2.0213,
4919
+ "step": 811000
4920
+ },
4921
+ {
4922
+ "epoch": 1.88,
4923
+ "learning_rate": 5e-05,
4924
+ "loss": 2.0182,
4925
+ "step": 812000
4926
+ },
4927
+ {
4928
+ "epoch": 1.88,
4929
+ "learning_rate": 5e-05,
4930
+ "loss": 2.0195,
4931
+ "step": 813000
4932
+ },
4933
+ {
4934
+ "epoch": 1.89,
4935
+ "learning_rate": 5e-05,
4936
+ "loss": 2.0245,
4937
+ "step": 814000
4938
+ },
4939
+ {
4940
+ "epoch": 1.89,
4941
+ "learning_rate": 5e-05,
4942
+ "loss": 2.018,
4943
+ "step": 815000
4944
+ },
4945
+ {
4946
+ "epoch": 1.89,
4947
+ "learning_rate": 5e-05,
4948
+ "loss": 2.0219,
4949
+ "step": 816000
4950
+ },
4951
+ {
4952
+ "epoch": 1.89,
4953
+ "learning_rate": 5e-05,
4954
+ "loss": 2.0221,
4955
+ "step": 817000
4956
+ },
4957
+ {
4958
+ "epoch": 1.9,
4959
+ "learning_rate": 5e-05,
4960
+ "loss": 2.019,
4961
+ "step": 818000
4962
+ },
4963
+ {
4964
+ "epoch": 1.9,
4965
+ "learning_rate": 5e-05,
4966
+ "loss": 2.0199,
4967
+ "step": 819000
4968
+ },
4969
+ {
4970
+ "epoch": 1.9,
4971
+ "learning_rate": 5e-05,
4972
+ "loss": 2.0166,
4973
+ "step": 820000
4974
+ },
4975
+ {
4976
+ "epoch": 1.9,
4977
+ "learning_rate": 5e-05,
4978
+ "loss": 2.0192,
4979
+ "step": 821000
4980
+ },
4981
+ {
4982
+ "epoch": 1.91,
4983
+ "learning_rate": 5e-05,
4984
+ "loss": 2.0217,
4985
+ "step": 822000
4986
+ },
4987
+ {
4988
+ "epoch": 1.91,
4989
+ "learning_rate": 5e-05,
4990
+ "loss": 2.0179,
4991
+ "step": 823000
4992
+ },
4993
+ {
4994
+ "epoch": 1.91,
4995
+ "learning_rate": 5e-05,
4996
+ "loss": 2.0242,
4997
+ "step": 824000
4998
+ },
4999
+ {
5000
+ "epoch": 1.91,
5001
+ "learning_rate": 5e-05,
5002
+ "loss": 2.0158,
5003
+ "step": 825000
5004
+ },
5005
+ {
5006
+ "epoch": 1.91,
5007
+ "learning_rate": 5e-05,
5008
+ "loss": 2.013,
5009
+ "step": 826000
5010
+ },
5011
+ {
5012
+ "epoch": 1.92,
5013
+ "learning_rate": 5e-05,
5014
+ "loss": 2.0196,
5015
+ "step": 827000
5016
+ },
5017
+ {
5018
+ "epoch": 1.92,
5019
+ "learning_rate": 5e-05,
5020
+ "loss": 2.02,
5021
+ "step": 828000
5022
+ },
5023
+ {
5024
+ "epoch": 1.92,
5025
+ "learning_rate": 5e-05,
5026
+ "loss": 2.021,
5027
+ "step": 829000
5028
+ },
5029
+ {
5030
+ "epoch": 1.92,
5031
+ "learning_rate": 5e-05,
5032
+ "loss": 2.0192,
5033
+ "step": 830000
5034
+ },
5035
+ {
5036
+ "epoch": 1.93,
5037
+ "learning_rate": 5e-05,
5038
+ "loss": 2.0191,
5039
+ "step": 831000
5040
+ },
5041
+ {
5042
+ "epoch": 1.93,
5043
+ "learning_rate": 5e-05,
5044
+ "loss": 2.0184,
5045
+ "step": 832000
5046
+ },
5047
+ {
5048
+ "epoch": 1.93,
5049
+ "learning_rate": 5e-05,
5050
+ "loss": 2.0196,
5051
+ "step": 833000
5052
+ },
5053
+ {
5054
+ "epoch": 1.93,
5055
+ "learning_rate": 5e-05,
5056
+ "loss": 2.0183,
5057
+ "step": 834000
5058
+ },
5059
+ {
5060
+ "epoch": 1.94,
5061
+ "learning_rate": 5e-05,
5062
+ "loss": 2.0134,
5063
+ "step": 835000
5064
+ },
5065
+ {
5066
+ "epoch": 1.94,
5067
+ "learning_rate": 5e-05,
5068
+ "loss": 2.0163,
5069
+ "step": 836000
5070
+ },
5071
+ {
5072
+ "epoch": 1.94,
5073
+ "learning_rate": 5e-05,
5074
+ "loss": 2.0113,
5075
+ "step": 837000
5076
+ },
5077
+ {
5078
+ "epoch": 1.94,
5079
+ "learning_rate": 5e-05,
5080
+ "loss": 2.0206,
5081
+ "step": 838000
5082
+ },
5083
+ {
5084
+ "epoch": 1.94,
5085
+ "learning_rate": 5e-05,
5086
+ "loss": 2.0166,
5087
+ "step": 839000
5088
+ },
5089
+ {
5090
+ "epoch": 1.95,
5091
+ "learning_rate": 5e-05,
5092
+ "loss": 2.0181,
5093
+ "step": 840000
5094
+ },
5095
+ {
5096
+ "epoch": 1.95,
5097
+ "learning_rate": 5e-05,
5098
+ "loss": 2.0168,
5099
+ "step": 841000
5100
+ },
5101
+ {
5102
+ "epoch": 1.95,
5103
+ "learning_rate": 5e-05,
5104
+ "loss": 2.01,
5105
+ "step": 842000
5106
+ },
5107
+ {
5108
+ "epoch": 1.95,
5109
+ "learning_rate": 5e-05,
5110
+ "loss": 2.0153,
5111
+ "step": 843000
5112
+ },
5113
+ {
5114
+ "epoch": 1.96,
5115
+ "learning_rate": 5e-05,
5116
+ "loss": 2.0084,
5117
+ "step": 844000
5118
+ },
5119
+ {
5120
+ "epoch": 1.96,
5121
+ "learning_rate": 5e-05,
5122
+ "loss": 2.0169,
5123
+ "step": 845000
5124
+ },
5125
+ {
5126
+ "epoch": 1.96,
5127
+ "learning_rate": 5e-05,
5128
+ "loss": 2.0142,
5129
+ "step": 846000
5130
+ },
5131
+ {
5132
+ "epoch": 1.96,
5133
+ "learning_rate": 5e-05,
5134
+ "loss": 2.0132,
5135
+ "step": 847000
5136
+ },
5137
+ {
5138
+ "epoch": 1.97,
5139
+ "learning_rate": 5e-05,
5140
+ "loss": 2.0126,
5141
+ "step": 848000
5142
+ },
5143
+ {
5144
+ "epoch": 1.97,
5145
+ "learning_rate": 5e-05,
5146
+ "loss": 2.0121,
5147
+ "step": 849000
5148
+ },
5149
+ {
5150
+ "epoch": 1.97,
5151
+ "learning_rate": 5e-05,
5152
+ "loss": 2.0122,
5153
+ "step": 850000
5154
+ },
5155
+ {
5156
+ "epoch": 1.97,
5157
+ "learning_rate": 5e-05,
5158
+ "loss": 2.0149,
5159
+ "step": 851000
5160
+ },
5161
+ {
5162
+ "epoch": 1.97,
5163
+ "learning_rate": 5e-05,
5164
+ "loss": 2.0125,
5165
+ "step": 852000
5166
+ },
5167
+ {
5168
+ "epoch": 1.98,
5169
+ "learning_rate": 5e-05,
5170
+ "loss": 2.0074,
5171
+ "step": 853000
5172
+ },
5173
+ {
5174
+ "epoch": 1.98,
5175
+ "learning_rate": 5e-05,
5176
+ "loss": 2.0136,
5177
+ "step": 854000
5178
+ },
5179
+ {
5180
+ "epoch": 1.98,
5181
+ "learning_rate": 5e-05,
5182
+ "loss": 2.0118,
5183
+ "step": 855000
5184
+ },
5185
+ {
5186
+ "epoch": 1.98,
5187
+ "learning_rate": 5e-05,
5188
+ "loss": 2.0139,
5189
+ "step": 856000
5190
+ },
5191
+ {
5192
+ "epoch": 1.99,
5193
+ "learning_rate": 5e-05,
5194
+ "loss": 2.0158,
5195
+ "step": 857000
5196
+ },
5197
+ {
5198
+ "epoch": 1.99,
5199
+ "learning_rate": 5e-05,
5200
+ "loss": 2.0105,
5201
+ "step": 858000
5202
+ },
5203
+ {
5204
+ "epoch": 1.99,
5205
+ "learning_rate": 5e-05,
5206
+ "loss": 2.0075,
5207
+ "step": 859000
5208
+ },
5209
+ {
5210
+ "epoch": 1.99,
5211
+ "learning_rate": 5e-05,
5212
+ "loss": 2.007,
5213
+ "step": 860000
5214
+ },
5215
+ {
5216
+ "epoch": 2.0,
5217
+ "learning_rate": 5e-05,
5218
+ "loss": 2.0143,
5219
+ "step": 861000
5220
+ },
5221
+ {
5222
+ "epoch": 2.0,
5223
+ "learning_rate": 5e-05,
5224
+ "loss": 2.0067,
5225
+ "step": 862000
5226
+ },
5227
+ {
5228
+ "epoch": 2.0,
5229
+ "step": 862850,
5230
+ "total_flos": 1.5713271682523202e+19,
5231
+ "train_loss": 0.02994945877374942,
5232
+ "train_runtime": 24111.6962,
5233
+ "train_samples_per_second": 9161.097,
5234
+ "train_steps_per_second": 35.786
5235
+ }
5236
+ ],
5237
+ "max_steps": 862850,
5238
+ "num_train_epochs": 2,
5239
+ "total_flos": 1.5713271682523202e+19,
5240
+ "trial_name": null,
5241
+ "trial_params": null
5242
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c3cdeeac5c64ef58a3919f4edcd09f7bc8ad3fbe1f2741a9fedbf10c91f96ab5
3
+ size 4411