Hugging Face's logo

Graphcore
/
roberta-base-squad

Question Answering
Transformers
PyTorch
Graphcore
roberta
Generated from Trainer
Inference Endpoints
Model card Files Community
1
jimypbr commited on
Commit
f8d558b
•
1 Parent(s): 0e5caf5

End of training

Browse files
Files changed (17) hide show
  1. .gitattributes +1 -0
  2. .gitignore +1 -0
  3. all_results.json +11 -0
  4. config.json +27 -0
  5. eval_nbest_predictions.json +3 -0
  6. eval_predictions.json +0 -0
  7. eval_results.json +6 -0
  8. ipu_config.json +29 -0
  9. merges.txt +0 -0
  10. pytorch_model.bin +3 -0
  11. special_tokens_map.json +1 -0
  12. tokenizer.json +0 -0
  13. tokenizer_config.json +1 -0
  14. train_results.json +8 -0
  15. trainer_state.json +4165 -0
  16. training_args.bin +3 -0
  17. vocab.json +0 -0
.gitattributes CHANGED
@@ -25,3 +25,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
25
  *.zip filter=lfs diff=lfs merge=lfs -text
26
  *.zstandard filter=lfs diff=lfs merge=lfs -text
27
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
25
  *.zip filter=lfs diff=lfs merge=lfs -text
26
  *.zstandard filter=lfs diff=lfs merge=lfs -text
27
  *tfevents* filter=lfs diff=lfs merge=lfs -text
28
+ eval_nbest_predictions.json filter=lfs diff=lfs merge=lfs -text
.gitignore ADDED
@@ -0,0 +1 @@
 
 
1
+ checkpoint-*/
all_results.json ADDED
@@ -0,0 +1,11 @@
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 2.0,
3
+ "eval_exact_match": 85.26963103122044,
4
+ "eval_f1": 91.74547836701485,
5
+ "eval_samples": 10790,
6
+ "train_loss": 1.2527947025022645,
7
+ "train_runtime": 134.5045,
8
+ "train_samples": 88568,
9
+ "train_samples_per_second": 1316.952,
10
+ "train_steps_per_second": 5.13
11
+ }
config.json ADDED
@@ -0,0 +1,27 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "roberta-base",
3
+ "architectures": [
4
+ "PoptorchPipelinedRobertaForQuestionAnswering"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.1,
7
+ "bos_token_id": 0,
8
+ "classifier_dropout": null,
9
+ "eos_token_id": 2,
10
+ "hidden_act": "gelu",
11
+ "hidden_dropout_prob": 0.1,
12
+ "hidden_size": 768,
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 3072,
15
+ "layer_norm_eps": 1e-05,
16
+ "max_position_embeddings": 514,
17
+ "model_type": "roberta",
18
+ "num_attention_heads": 12,
19
+ "num_hidden_layers": 12,
20
+ "pad_token_id": 1,
21
+ "position_embedding_type": "absolute",
22
+ "torch_dtype": "float16",
23
+ "transformers_version": "4.18.0.dev0",
24
+ "type_vocab_size": 1,
25
+ "use_cache": true,
26
+ "vocab_size": 50265
27
+ }
eval_nbest_predictions.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:15b906a072d75ed2f41e09a5bac4004d5f448e1e2bfc65b3055522073e6079e5
3
+ size 45648758
eval_predictions.json ADDED
The diff for this file is too large to render. See raw diff
 
eval_results.json ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 2.0,
3
+ "eval_exact_match": 85.26963103122044,
4
+ "eval_f1": 91.74547836701485,
5
+ "eval_samples": 10790
6
+ }
ipu_config.json ADDED
@@ -0,0 +1,29 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "device_iterations": 1,
3
+ "embedding_serialization_factor": 0,
4
+ "enable_half_first_order_momentum": true,
5
+ "enable_half_partials": true,
6
+ "executable_cache_dir": "./exe_cache",
7
+ "gradient_accumulation_steps": 16,
8
+ "inference_device_iterations": 5,
9
+ "inference_replication_factor": 4,
10
+ "ipus_per_replica": 4,
11
+ "layers_per_ipu": [
12
+ 0,
13
+ 4,
14
+ 4,
15
+ 4
16
+ ],
17
+ "matmul_proportion": [
18
+ 0.25
19
+ ],
20
+ "optimizer_state_offchip": false,
21
+ "optimum_version": "0.1.2",
22
+ "output_mode": "final",
23
+ "profile_dir": "",
24
+ "recompute_checkpoint_every_layer": true,
25
+ "replicated_tensor_sharding": true,
26
+ "replication_factor": 4,
27
+ "seed": 1984,
28
+ "use_popdist": false
29
+ }
merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a725b2e8e7bcb6925174539798407f55f220d82b095dabfb3635a9538fa7c533
3
+ size 248181489
special_tokens_map.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"bos_token": "<s>", "eos_token": "</s>", "unk_token": "<unk>", "sep_token": "</s>", "pad_token": "<pad>", "cls_token": "<s>", "mask_token": {"content": "<mask>", "single_word": false, "lstrip": true, "rstrip": false, "normalized": false}}
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer_config.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"errors": "replace", "bos_token": "<s>", "eos_token": "</s>", "sep_token": "</s>", "cls_token": "<s>", "unk_token": "<unk>", "pad_token": "<pad>", "mask_token": "<mask>", "add_prefix_space": false, "trim_offsets": true, "model_max_length": 512, "special_tokens_map_file": null, "name_or_path": "roberta-base", "tokenizer_class": "RobertaTokenizer"}
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 2.0,
3
+ "train_loss": 1.2527947025022645,
4
+ "train_runtime": 134.5045,
5
+ "train_samples": 88568,
6
+ "train_samples_per_second": 1316.952,
7
+ "train_steps_per_second": 5.13
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,4165 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 2.0,
5
+ "global_step": 690,
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": 3.468208092485549e-07,
13
+ "loss": 6.0,
14
+ "step": 1
15
+ },
16
+ {
17
+ "epoch": 0.01,
18
+ "learning_rate": 6.936416184971098e-07,
19
+ "loss": 6.0195,
20
+ "step": 2
21
+ },
22
+ {
23
+ "epoch": 0.01,
24
+ "learning_rate": 1.0404624277456647e-06,
25
+ "loss": 6.0156,
26
+ "step": 3
27
+ },
28
+ {
29
+ "epoch": 0.01,
30
+ "learning_rate": 1.3872832369942195e-06,
31
+ "loss": 6.0352,
32
+ "step": 4
33
+ },
34
+ {
35
+ "epoch": 0.01,
36
+ "learning_rate": 1.7341040462427746e-06,
37
+ "loss": 6.0273,
38
+ "step": 5
39
+ },
40
+ {
41
+ "epoch": 0.02,
42
+ "learning_rate": 2.0809248554913294e-06,
43
+ "loss": 5.9805,
44
+ "step": 6
45
+ },
46
+ {
47
+ "epoch": 0.02,
48
+ "learning_rate": 2.4277456647398847e-06,
49
+ "loss": 6.0,
50
+ "step": 7
51
+ },
52
+ {
53
+ "epoch": 0.02,
54
+ "learning_rate": 2.774566473988439e-06,
55
+ "loss": 5.9375,
56
+ "step": 8
57
+ },
58
+ {
59
+ "epoch": 0.03,
60
+ "learning_rate": 3.1213872832369943e-06,
61
+ "loss": 5.9023,
62
+ "step": 9
63
+ },
64
+ {
65
+ "epoch": 0.03,
66
+ "learning_rate": 3.468208092485549e-06,
67
+ "loss": 5.9219,
68
+ "step": 10
69
+ },
70
+ {
71
+ "epoch": 0.03,
72
+ "learning_rate": 3.8150289017341036e-06,
73
+ "loss": 5.9531,
74
+ "step": 11
75
+ },
76
+ {
77
+ "epoch": 0.03,
78
+ "learning_rate": 4.161849710982659e-06,
79
+ "loss": 5.9766,
80
+ "step": 12
81
+ },
82
+ {
83
+ "epoch": 0.04,
84
+ "learning_rate": 4.508670520231214e-06,
85
+ "loss": 5.9141,
86
+ "step": 13
87
+ },
88
+ {
89
+ "epoch": 0.04,
90
+ "learning_rate": 4.855491329479769e-06,
91
+ "loss": 5.8828,
92
+ "step": 14
93
+ },
94
+ {
95
+ "epoch": 0.04,
96
+ "learning_rate": 5.202312138728324e-06,
97
+ "loss": 5.9219,
98
+ "step": 15
99
+ },
100
+ {
101
+ "epoch": 0.05,
102
+ "learning_rate": 5.549132947976878e-06,
103
+ "loss": 5.8633,
104
+ "step": 16
105
+ },
106
+ {
107
+ "epoch": 0.05,
108
+ "learning_rate": 5.895953757225434e-06,
109
+ "loss": 5.9062,
110
+ "step": 17
111
+ },
112
+ {
113
+ "epoch": 0.05,
114
+ "learning_rate": 6.242774566473989e-06,
115
+ "loss": 5.8672,
116
+ "step": 18
117
+ },
118
+ {
119
+ "epoch": 0.06,
120
+ "learning_rate": 6.589595375722544e-06,
121
+ "loss": 5.7695,
122
+ "step": 19
123
+ },
124
+ {
125
+ "epoch": 0.06,
126
+ "learning_rate": 6.936416184971098e-06,
127
+ "loss": 5.8008,
128
+ "step": 20
129
+ },
130
+ {
131
+ "epoch": 0.06,
132
+ "learning_rate": 7.283236994219653e-06,
133
+ "loss": 5.75,
134
+ "step": 21
135
+ },
136
+ {
137
+ "epoch": 0.06,
138
+ "learning_rate": 7.630057803468207e-06,
139
+ "loss": 5.6953,
140
+ "step": 22
141
+ },
142
+ {
143
+ "epoch": 0.07,
144
+ "learning_rate": 7.976878612716762e-06,
145
+ "loss": 5.6758,
146
+ "step": 23
147
+ },
148
+ {
149
+ "epoch": 0.07,
150
+ "learning_rate": 8.323699421965318e-06,
151
+ "loss": 5.6055,
152
+ "step": 24
153
+ },
154
+ {
155
+ "epoch": 0.07,
156
+ "learning_rate": 8.670520231213873e-06,
157
+ "loss": 5.5859,
158
+ "step": 25
159
+ },
160
+ {
161
+ "epoch": 0.08,
162
+ "learning_rate": 9.017341040462428e-06,
163
+ "loss": 5.4531,
164
+ "step": 26
165
+ },
166
+ {
167
+ "epoch": 0.08,
168
+ "learning_rate": 9.364161849710983e-06,
169
+ "loss": 5.5078,
170
+ "step": 27
171
+ },
172
+ {
173
+ "epoch": 0.08,
174
+ "learning_rate": 9.710982658959539e-06,
175
+ "loss": 5.3711,
176
+ "step": 28
177
+ },
178
+ {
179
+ "epoch": 0.08,
180
+ "learning_rate": 1.0057803468208092e-05,
181
+ "loss": 5.1953,
182
+ "step": 29
183
+ },
184
+ {
185
+ "epoch": 0.09,
186
+ "learning_rate": 1.0404624277456647e-05,
187
+ "loss": 5.2578,
188
+ "step": 30
189
+ },
190
+ {
191
+ "epoch": 0.09,
192
+ "learning_rate": 1.0751445086705203e-05,
193
+ "loss": 5.0273,
194
+ "step": 31
195
+ },
196
+ {
197
+ "epoch": 0.09,
198
+ "learning_rate": 1.1098265895953756e-05,
199
+ "loss": 4.9766,
200
+ "step": 32
201
+ },
202
+ {
203
+ "epoch": 0.1,
204
+ "learning_rate": 1.1445086705202312e-05,
205
+ "loss": 4.7852,
206
+ "step": 33
207
+ },
208
+ {
209
+ "epoch": 0.1,
210
+ "learning_rate": 1.1791907514450869e-05,
211
+ "loss": 4.8984,
212
+ "step": 34
213
+ },
214
+ {
215
+ "epoch": 0.1,
216
+ "learning_rate": 1.2138728323699422e-05,
217
+ "loss": 4.625,
218
+ "step": 35
219
+ },
220
+ {
221
+ "epoch": 0.1,
222
+ "learning_rate": 1.2485549132947977e-05,
223
+ "loss": 4.5156,
224
+ "step": 36
225
+ },
226
+ {
227
+ "epoch": 0.11,
228
+ "learning_rate": 1.2832369942196533e-05,
229
+ "loss": 4.5508,
230
+ "step": 37
231
+ },
232
+ {
233
+ "epoch": 0.11,
234
+ "learning_rate": 1.3179190751445088e-05,
235
+ "loss": 4.3281,
236
+ "step": 38
237
+ },
238
+ {
239
+ "epoch": 0.11,
240
+ "learning_rate": 1.3526011560693641e-05,
241
+ "loss": 4.5938,
242
+ "step": 39
243
+ },
244
+ {
245
+ "epoch": 0.12,
246
+ "learning_rate": 1.3872832369942197e-05,
247
+ "loss": 4.1172,
248
+ "step": 40
249
+ },
250
+ {
251
+ "epoch": 0.12,
252
+ "learning_rate": 1.4219653179190752e-05,
253
+ "loss": 4.2188,
254
+ "step": 41
255
+ },
256
+ {
257
+ "epoch": 0.12,
258
+ "learning_rate": 1.4566473988439305e-05,
259
+ "loss": 4.0156,
260
+ "step": 42
261
+ },
262
+ {
263
+ "epoch": 0.12,
264
+ "learning_rate": 1.491329479768786e-05,
265
+ "loss": 4.0234,
266
+ "step": 43
267
+ },
268
+ {
269
+ "epoch": 0.13,
270
+ "learning_rate": 1.5260115606936414e-05,
271
+ "loss": 3.9141,
272
+ "step": 44
273
+ },
274
+ {
275
+ "epoch": 0.13,
276
+ "learning_rate": 1.560693641618497e-05,
277
+ "loss": 3.6562,
278
+ "step": 45
279
+ },
280
+ {
281
+ "epoch": 0.13,
282
+ "learning_rate": 1.5953757225433525e-05,
283
+ "loss": 3.6133,
284
+ "step": 46
285
+ },
286
+ {
287
+ "epoch": 0.14,
288
+ "learning_rate": 1.630057803468208e-05,
289
+ "loss": 3.1582,
290
+ "step": 47
291
+ },
292
+ {
293
+ "epoch": 0.14,
294
+ "learning_rate": 1.6647398843930635e-05,
295
+ "loss": 3.248,
296
+ "step": 48
297
+ },
298
+ {
299
+ "epoch": 0.14,
300
+ "learning_rate": 1.6994219653179194e-05,
301
+ "loss": 3.0996,
302
+ "step": 49
303
+ },
304
+ {
305
+ "epoch": 0.14,
306
+ "learning_rate": 1.7341040462427746e-05,
307
+ "loss": 3.1875,
308
+ "step": 50
309
+ },
310
+ {
311
+ "epoch": 0.15,
312
+ "learning_rate": 1.76878612716763e-05,
313
+ "loss": 2.9941,
314
+ "step": 51
315
+ },
316
+ {
317
+ "epoch": 0.15,
318
+ "learning_rate": 1.8034682080924856e-05,
319
+ "loss": 2.9297,
320
+ "step": 52
321
+ },
322
+ {
323
+ "epoch": 0.15,
324
+ "learning_rate": 1.838150289017341e-05,
325
+ "loss": 3.1094,
326
+ "step": 53
327
+ },
328
+ {
329
+ "epoch": 0.16,
330
+ "learning_rate": 1.8728323699421967e-05,
331
+ "loss": 2.7812,
332
+ "step": 54
333
+ },
334
+ {
335
+ "epoch": 0.16,
336
+ "learning_rate": 1.9075144508670522e-05,
337
+ "loss": 1.9785,
338
+ "step": 55
339
+ },
340
+ {
341
+ "epoch": 0.16,
342
+ "learning_rate": 1.9421965317919077e-05,
343
+ "loss": 3.4531,
344
+ "step": 56
345
+ },
346
+ {
347
+ "epoch": 0.17,
348
+ "learning_rate": 1.976878612716763e-05,
349
+ "loss": 2.1875,
350
+ "step": 57
351
+ },
352
+ {
353
+ "epoch": 0.17,
354
+ "learning_rate": 2.0115606936416184e-05,
355
+ "loss": 1.9697,
356
+ "step": 58
357
+ },
358
+ {
359
+ "epoch": 0.17,
360
+ "learning_rate": 2.046242774566474e-05,
361
+ "loss": 2.1777,
362
+ "step": 59
363
+ },
364
+ {
365
+ "epoch": 0.17,
366
+ "learning_rate": 2.0809248554913295e-05,
367
+ "loss": 1.8594,
368
+ "step": 60
369
+ },
370
+ {
371
+ "epoch": 0.18,
372
+ "learning_rate": 2.115606936416185e-05,
373
+ "loss": 2.2441,
374
+ "step": 61
375
+ },
376
+ {
377
+ "epoch": 0.18,
378
+ "learning_rate": 2.1502890173410405e-05,
379
+ "loss": 2.3242,
380
+ "step": 62
381
+ },
382
+ {
383
+ "epoch": 0.18,
384
+ "learning_rate": 2.184971098265896e-05,
385
+ "loss": 1.9541,
386
+ "step": 63
387
+ },
388
+ {
389
+ "epoch": 0.19,
390
+ "learning_rate": 2.2196531791907513e-05,
391
+ "loss": 1.7412,
392
+ "step": 64
393
+ },
394
+ {
395
+ "epoch": 0.19,
396
+ "learning_rate": 2.2543352601156068e-05,
397
+ "loss": 2.2969,
398
+ "step": 65
399
+ },
400
+ {
401
+ "epoch": 0.19,
402
+ "learning_rate": 2.2890173410404623e-05,
403
+ "loss": 1.3818,
404
+ "step": 66
405
+ },
406
+ {
407
+ "epoch": 0.19,
408
+ "learning_rate": 2.323699421965318e-05,
409
+ "loss": 1.9668,
410
+ "step": 67
411
+ },
412
+ {
413
+ "epoch": 0.2,
414
+ "learning_rate": 2.3583815028901737e-05,
415
+ "loss": 2.4727,
416
+ "step": 68
417
+ },
418
+ {
419
+ "epoch": 0.2,
420
+ "learning_rate": 2.3930635838150292e-05,
421
+ "loss": 2.0645,
422
+ "step": 69
423
+ },
424
+ {
425
+ "epoch": 0.2,
426
+ "learning_rate": 2.4277456647398844e-05,
427
+ "loss": 2.0312,
428
+ "step": 70
429
+ },
430
+ {
431
+ "epoch": 0.21,
432
+ "learning_rate": 2.46242774566474e-05,
433
+ "loss": 2.0547,
434
+ "step": 71
435
+ },
436
+ {
437
+ "epoch": 0.21,
438
+ "learning_rate": 2.4971098265895955e-05,
439
+ "loss": 1.7764,
440
+ "step": 72
441
+ },
442
+ {
443
+ "epoch": 0.21,
444
+ "learning_rate": 2.531791907514451e-05,
445
+ "loss": 1.6074,
446
+ "step": 73
447
+ },
448
+ {
449
+ "epoch": 0.21,
450
+ "learning_rate": 2.5664739884393065e-05,
451
+ "loss": 1.2148,
452
+ "step": 74
453
+ },
454
+ {
455
+ "epoch": 0.22,
456
+ "learning_rate": 2.601156069364162e-05,
457
+ "loss": 1.3525,
458
+ "step": 75
459
+ },
460
+ {
461
+ "epoch": 0.22,
462
+ "learning_rate": 2.6358381502890176e-05,
463
+ "loss": 1.6963,
464
+ "step": 76
465
+ },
466
+ {
467
+ "epoch": 0.22,
468
+ "learning_rate": 2.6705202312138728e-05,
469
+ "loss": 1.4746,
470
+ "step": 77
471
+ },
472
+ {
473
+ "epoch": 0.23,
474
+ "learning_rate": 2.7052023121387283e-05,
475
+ "loss": 1.8105,
476
+ "step": 78
477
+ },
478
+ {
479
+ "epoch": 0.23,
480
+ "learning_rate": 2.7398843930635838e-05,
481
+ "loss": 1.5195,
482
+ "step": 79
483
+ },
484
+ {
485
+ "epoch": 0.23,
486
+ "learning_rate": 2.7745664739884393e-05,
487
+ "loss": 1.1855,
488
+ "step": 80
489
+ },
490
+ {
491
+ "epoch": 0.23,
492
+ "learning_rate": 2.809248554913295e-05,
493
+ "loss": 1.3115,
494
+ "step": 81
495
+ },
496
+ {
497
+ "epoch": 0.24,
498
+ "learning_rate": 2.8439306358381504e-05,
499
+ "loss": 1.6318,
500
+ "step": 82
501
+ },
502
+ {
503
+ "epoch": 0.24,
504
+ "learning_rate": 2.878612716763006e-05,
505
+ "loss": 2.2656,
506
+ "step": 83
507
+ },
508
+ {
509
+ "epoch": 0.24,
510
+ "learning_rate": 2.913294797687861e-05,
511
+ "loss": 1.8789,
512
+ "step": 84
513
+ },
514
+ {
515
+ "epoch": 0.25,
516
+ "learning_rate": 2.9479768786127166e-05,
517
+ "loss": 1.3555,
518
+ "step": 85
519
+ },
520
+ {
521
+ "epoch": 0.25,
522
+ "learning_rate": 2.982658959537572e-05,
523
+ "loss": 1.1211,
524
+ "step": 86
525
+ },
526
+ {
527
+ "epoch": 0.25,
528
+ "learning_rate": 3.017341040462428e-05,
529
+ "loss": 1.3418,
530
+ "step": 87
531
+ },
532
+ {
533
+ "epoch": 0.26,
534
+ "learning_rate": 3.052023121387283e-05,
535
+ "loss": 1.5234,
536
+ "step": 88
537
+ },
538
+ {
539
+ "epoch": 0.26,
540
+ "learning_rate": 3.086705202312139e-05,
541
+ "loss": 0.8867,
542
+ "step": 89
543
+ },
544
+ {
545
+ "epoch": 0.26,
546
+ "learning_rate": 3.121387283236994e-05,
547
+ "loss": 1.2432,
548
+ "step": 90
549
+ },
550
+ {
551
+ "epoch": 0.26,
552
+ "learning_rate": 3.15606936416185e-05,
553
+ "loss": 1.5752,
554
+ "step": 91
555
+ },
556
+ {
557
+ "epoch": 0.27,
558
+ "learning_rate": 3.190751445086705e-05,
559
+ "loss": 1.541,
560
+ "step": 92
561
+ },
562
+ {
563
+ "epoch": 0.27,
564
+ "learning_rate": 3.225433526011561e-05,
565
+ "loss": 1.0,
566
+ "step": 93
567
+ },
568
+ {
569
+ "epoch": 0.27,
570
+ "learning_rate": 3.260115606936416e-05,
571
+ "loss": 1.4531,
572
+ "step": 94
573
+ },
574
+ {
575
+ "epoch": 0.28,
576
+ "learning_rate": 3.294797687861272e-05,
577
+ "loss": 1.6367,
578
+ "step": 95
579
+ },
580
+ {
581
+ "epoch": 0.28,
582
+ "learning_rate": 3.329479768786127e-05,
583
+ "loss": 1.0898,
584
+ "step": 96
585
+ },
586
+ {
587
+ "epoch": 0.28,
588
+ "learning_rate": 3.364161849710983e-05,
589
+ "loss": 1.8867,
590
+ "step": 97
591
+ },
592
+ {
593
+ "epoch": 0.28,
594
+ "learning_rate": 3.398843930635839e-05,
595
+ "loss": 1.4658,
596
+ "step": 98
597
+ },
598
+ {
599
+ "epoch": 0.29,
600
+ "learning_rate": 3.433526011560694e-05,
601
+ "loss": 1.2217,
602
+ "step": 99
603
+ },
604
+ {
605
+ "epoch": 0.29,
606
+ "learning_rate": 3.468208092485549e-05,
607
+ "loss": 1.3906,
608
+ "step": 100
609
+ },
610
+ {
611
+ "epoch": 0.29,
612
+ "learning_rate": 3.5028901734104043e-05,
613
+ "loss": 1.4932,
614
+ "step": 101
615
+ },
616
+ {
617
+ "epoch": 0.3,
618
+ "learning_rate": 3.53757225433526e-05,
619
+ "loss": 1.0254,
620
+ "step": 102
621
+ },
622
+ {
623
+ "epoch": 0.3,
624
+ "learning_rate": 3.5722543352601154e-05,
625
+ "loss": 1.1143,
626
+ "step": 103
627
+ },
628
+ {
629
+ "epoch": 0.3,
630
+ "learning_rate": 3.606936416184971e-05,
631
+ "loss": 1.6777,
632
+ "step": 104
633
+ },
634
+ {
635
+ "epoch": 0.3,
636
+ "learning_rate": 3.6416184971098265e-05,
637
+ "loss": 0.6465,
638
+ "step": 105
639
+ },
640
+ {
641
+ "epoch": 0.31,
642
+ "learning_rate": 3.676300578034682e-05,
643
+ "loss": 1.375,
644
+ "step": 106
645
+ },
646
+ {
647
+ "epoch": 0.31,
648
+ "learning_rate": 3.7109826589595375e-05,
649
+ "loss": 1.1191,
650
+ "step": 107
651
+ },
652
+ {
653
+ "epoch": 0.31,
654
+ "learning_rate": 3.7456647398843934e-05,
655
+ "loss": 1.8301,
656
+ "step": 108
657
+ },
658
+ {
659
+ "epoch": 0.32,
660
+ "learning_rate": 3.7803468208092486e-05,
661
+ "loss": 1.6553,
662
+ "step": 109
663
+ },
664
+ {
665
+ "epoch": 0.32,
666
+ "learning_rate": 3.8150289017341044e-05,
667
+ "loss": 1.5566,
668
+ "step": 110
669
+ },
670
+ {
671
+ "epoch": 0.32,
672
+ "learning_rate": 3.8497109826589596e-05,
673
+ "loss": 1.6084,
674
+ "step": 111
675
+ },
676
+ {
677
+ "epoch": 0.32,
678
+ "learning_rate": 3.8843930635838155e-05,
679
+ "loss": 1.3359,
680
+ "step": 112
681
+ },
682
+ {
683
+ "epoch": 0.33,
684
+ "learning_rate": 3.9190751445086707e-05,
685
+ "loss": 1.1104,
686
+ "step": 113
687
+ },
688
+ {
689
+ "epoch": 0.33,
690
+ "learning_rate": 3.953757225433526e-05,
691
+ "loss": 0.9771,
692
+ "step": 114
693
+ },
694
+ {
695
+ "epoch": 0.33,
696
+ "learning_rate": 3.988439306358381e-05,
697
+ "loss": 1.4971,
698
+ "step": 115
699
+ },
700
+ {
701
+ "epoch": 0.34,
702
+ "learning_rate": 4.023121387283237e-05,
703
+ "loss": 1.627,
704
+ "step": 116
705
+ },
706
+ {
707
+ "epoch": 0.34,
708
+ "learning_rate": 4.057803468208093e-05,
709
+ "loss": 0.8262,
710
+ "step": 117
711
+ },
712
+ {
713
+ "epoch": 0.34,
714
+ "learning_rate": 4.092485549132948e-05,
715
+ "loss": 1.5273,
716
+ "step": 118
717
+ },
718
+ {
719
+ "epoch": 0.34,
720
+ "learning_rate": 4.127167630057804e-05,
721
+ "loss": 0.6133,
722
+ "step": 119
723
+ },
724
+ {
725
+ "epoch": 0.35,
726
+ "learning_rate": 4.161849710982659e-05,
727
+ "loss": 1.1729,
728
+ "step": 120
729
+ },
730
+ {
731
+ "epoch": 0.35,
732
+ "learning_rate": 4.196531791907515e-05,
733
+ "loss": 0.5942,
734
+ "step": 121
735
+ },
736
+ {
737
+ "epoch": 0.35,
738
+ "learning_rate": 4.23121387283237e-05,
739
+ "loss": 1.0908,
740
+ "step": 122
741
+ },
742
+ {
743
+ "epoch": 0.36,
744
+ "learning_rate": 4.265895953757226e-05,
745
+ "loss": 1.5547,
746
+ "step": 123
747
+ },
748
+ {
749
+ "epoch": 0.36,
750
+ "learning_rate": 4.300578034682081e-05,
751
+ "loss": 0.8452,
752
+ "step": 124
753
+ },
754
+ {
755
+ "epoch": 0.36,
756
+ "learning_rate": 4.335260115606937e-05,
757
+ "loss": 0.7603,
758
+ "step": 125
759
+ },
760
+ {
761
+ "epoch": 0.37,
762
+ "learning_rate": 4.369942196531792e-05,
763
+ "loss": 0.7627,
764
+ "step": 126
765
+ },
766
+ {
767
+ "epoch": 0.37,
768
+ "learning_rate": 4.404624277456647e-05,
769
+ "loss": 0.6387,
770
+ "step": 127
771
+ },
772
+ {
773
+ "epoch": 0.37,
774
+ "learning_rate": 4.4393063583815025e-05,
775
+ "loss": 1.0439,
776
+ "step": 128
777
+ },
778
+ {
779
+ "epoch": 0.37,
780
+ "learning_rate": 4.4739884393063584e-05,
781
+ "loss": 0.8623,
782
+ "step": 129
783
+ },
784
+ {
785
+ "epoch": 0.38,
786
+ "learning_rate": 4.5086705202312136e-05,
787
+ "loss": 1.9482,
788
+ "step": 130
789
+ },
790
+ {
791
+ "epoch": 0.38,
792
+ "learning_rate": 4.5433526011560694e-05,
793
+ "loss": 1.2148,
794
+ "step": 131
795
+ },
796
+ {
797
+ "epoch": 0.38,
798
+ "learning_rate": 4.5780346820809246e-05,
799
+ "loss": 1.249,
800
+ "step": 132
801
+ },
802
+ {
803
+ "epoch": 0.39,
804
+ "learning_rate": 4.6127167630057805e-05,
805
+ "loss": 1.2393,
806
+ "step": 133
807
+ },
808
+ {
809
+ "epoch": 0.39,
810
+ "learning_rate": 4.647398843930636e-05,
811
+ "loss": 1.6406,
812
+ "step": 134
813
+ },
814
+ {
815
+ "epoch": 0.39,
816
+ "learning_rate": 4.6820809248554915e-05,
817
+ "loss": 1.1143,
818
+ "step": 135
819
+ },
820
+ {
821
+ "epoch": 0.39,
822
+ "learning_rate": 4.7167630057803474e-05,
823
+ "loss": 1.543,
824
+ "step": 136
825
+ },
826
+ {
827
+ "epoch": 0.4,
828
+ "learning_rate": 4.7514450867052026e-05,
829
+ "loss": 1.9912,
830
+ "step": 137
831
+ },
832
+ {
833
+ "epoch": 0.4,
834
+ "learning_rate": 4.7861271676300585e-05,
835
+ "loss": 0.9761,
836
+ "step": 138
837
+ },
838
+ {
839
+ "epoch": 0.4,
840
+ "learning_rate": 4.8208092485549136e-05,
841
+ "loss": 1.4238,
842
+ "step": 139
843
+ },
844
+ {
845
+ "epoch": 0.41,
846
+ "learning_rate": 4.855491329479769e-05,
847
+ "loss": 1.04,
848
+ "step": 140
849
+ },
850
+ {
851
+ "epoch": 0.41,
852
+ "learning_rate": 4.890173410404624e-05,
853
+ "loss": 1.2793,
854
+ "step": 141
855
+ },
856
+ {
857
+ "epoch": 0.41,
858
+ "learning_rate": 4.92485549132948e-05,
859
+ "loss": 1.8594,
860
+ "step": 142
861
+ },
862
+ {
863
+ "epoch": 0.41,
864
+ "learning_rate": 4.959537572254335e-05,
865
+ "loss": 0.8799,
866
+ "step": 143
867
+ },
868
+ {
869
+ "epoch": 0.42,
870
+ "learning_rate": 4.994219653179191e-05,
871
+ "loss": 0.2998,
872
+ "step": 144
873
+ },
874
+ {
875
+ "epoch": 0.42,
876
+ "learning_rate": 5.028901734104046e-05,
877
+ "loss": 0.9536,
878
+ "step": 145
879
+ },
880
+ {
881
+ "epoch": 0.42,
882
+ "learning_rate": 5.063583815028902e-05,
883
+ "loss": 1.4346,
884
+ "step": 146
885
+ },
886
+ {
887
+ "epoch": 0.43,
888
+ "learning_rate": 5.098265895953757e-05,
889
+ "loss": 0.7554,
890
+ "step": 147
891
+ },
892
+ {
893
+ "epoch": 0.43,
894
+ "learning_rate": 5.132947976878613e-05,
895
+ "loss": 0.7715,
896
+ "step": 148
897
+ },
898
+ {
899
+ "epoch": 0.43,
900
+ "learning_rate": 5.167630057803468e-05,
901
+ "loss": 1.6406,
902
+ "step": 149
903
+ },
904
+ {
905
+ "epoch": 0.43,
906
+ "learning_rate": 5.202312138728324e-05,
907
+ "loss": 0.6548,
908
+ "step": 150
909
+ },
910
+ {
911
+ "epoch": 0.44,
912
+ "learning_rate": 5.236994219653179e-05,
913
+ "loss": 1.9697,
914
+ "step": 151
915
+ },
916
+ {
917
+ "epoch": 0.44,
918
+ "learning_rate": 5.271676300578035e-05,
919
+ "loss": 1.2061,
920
+ "step": 152
921
+ },
922
+ {
923
+ "epoch": 0.44,
924
+ "learning_rate": 5.30635838150289e-05,
925
+ "loss": 0.8584,
926
+ "step": 153
927
+ },
928
+ {
929
+ "epoch": 0.45,
930
+ "learning_rate": 5.3410404624277455e-05,
931
+ "loss": 0.8721,
932
+ "step": 154
933
+ },
934
+ {
935
+ "epoch": 0.45,
936
+ "learning_rate": 5.3757225433526014e-05,
937
+ "loss": 0.8486,
938
+ "step": 155
939
+ },
940
+ {
941
+ "epoch": 0.45,
942
+ "learning_rate": 5.4104046242774566e-05,
943
+ "loss": 1.0762,
944
+ "step": 156
945
+ },
946
+ {
947
+ "epoch": 0.46,
948
+ "learning_rate": 5.4450867052023124e-05,
949
+ "loss": 0.5454,
950
+ "step": 157
951
+ },
952
+ {
953
+ "epoch": 0.46,
954
+ "learning_rate": 5.4797687861271676e-05,
955
+ "loss": 0.8379,
956
+ "step": 158
957
+ },
958
+ {
959
+ "epoch": 0.46,
960
+ "learning_rate": 5.5144508670520235e-05,
961
+ "loss": 0.5874,
962
+ "step": 159
963
+ },
964
+ {
965
+ "epoch": 0.46,
966
+ "learning_rate": 5.5491329479768787e-05,
967
+ "loss": 0.8701,
968
+ "step": 160
969
+ },
970
+ {
971
+ "epoch": 0.47,
972
+ "learning_rate": 5.5838150289017345e-05,
973
+ "loss": 0.9492,
974
+ "step": 161
975
+ },
976
+ {
977
+ "epoch": 0.47,
978
+ "learning_rate": 5.61849710982659e-05,
979
+ "loss": 1.2998,
980
+ "step": 162
981
+ },
982
+ {
983
+ "epoch": 0.47,
984
+ "learning_rate": 5.6531791907514456e-05,
985
+ "loss": 1.7168,
986
+ "step": 163
987
+ },
988
+ {
989
+ "epoch": 0.48,
990
+ "learning_rate": 5.687861271676301e-05,
991
+ "loss": 0.6938,
992
+ "step": 164
993
+ },
994
+ {
995
+ "epoch": 0.48,
996
+ "learning_rate": 5.7225433526011566e-05,
997
+ "loss": 1.0918,
998
+ "step": 165
999
+ },
1000
+ {
1001
+ "epoch": 0.48,
1002
+ "learning_rate": 5.757225433526012e-05,
1003
+ "loss": 1.1143,
1004
+ "step": 166
1005
+ },
1006
+ {
1007
+ "epoch": 0.48,
1008
+ "learning_rate": 5.791907514450867e-05,
1009
+ "loss": 1.3672,
1010
+ "step": 167
1011
+ },
1012
+ {
1013
+ "epoch": 0.49,
1014
+ "learning_rate": 5.826589595375722e-05,
1015
+ "loss": 1.0215,
1016
+ "step": 168
1017
+ },
1018
+ {
1019
+ "epoch": 0.49,
1020
+ "learning_rate": 5.861271676300578e-05,
1021
+ "loss": 1.4531,
1022
+ "step": 169
1023
+ },
1024
+ {
1025
+ "epoch": 0.49,
1026
+ "learning_rate": 5.895953757225433e-05,
1027
+ "loss": 1.3926,
1028
+ "step": 170
1029
+ },
1030
+ {
1031
+ "epoch": 0.5,
1032
+ "learning_rate": 5.930635838150289e-05,
1033
+ "loss": 1.624,
1034
+ "step": 171
1035
+ },
1036
+ {
1037
+ "epoch": 0.5,
1038
+ "learning_rate": 5.965317919075144e-05,
1039
+ "loss": 1.2578,
1040
+ "step": 172
1041
+ },
1042
+ {
1043
+ "epoch": 0.5,
1044
+ "learning_rate": 6e-05,
1045
+ "loss": 1.1318,
1046
+ "step": 173
1047
+ },
1048
+ {
1049
+ "epoch": 0.5,
1050
+ "learning_rate": 5.988394584139265e-05,
1051
+ "loss": 0.7456,
1052
+ "step": 174
1053
+ },
1054
+ {
1055
+ "epoch": 0.51,
1056
+ "learning_rate": 5.97678916827853e-05,
1057
+ "loss": 1.1855,
1058
+ "step": 175
1059
+ },
1060
+ {
1061
+ "epoch": 0.51,
1062
+ "learning_rate": 5.965183752417795e-05,
1063
+ "loss": 0.6221,
1064
+ "step": 176
1065
+ },
1066
+ {
1067
+ "epoch": 0.51,
1068
+ "learning_rate": 5.9535783365570604e-05,
1069
+ "loss": 1.5713,
1070
+ "step": 177
1071
+ },
1072
+ {
1073
+ "epoch": 0.52,
1074
+ "learning_rate": 5.9419729206963253e-05,
1075
+ "loss": 1.3613,
1076
+ "step": 178
1077
+ },
1078
+ {
1079
+ "epoch": 0.52,
1080
+ "learning_rate": 5.9303675048355896e-05,
1081
+ "loss": 0.7422,
1082
+ "step": 179
1083
+ },
1084
+ {
1085
+ "epoch": 0.52,
1086
+ "learning_rate": 5.918762088974855e-05,
1087
+ "loss": 0.6445,
1088
+ "step": 180
1089
+ },
1090
+ {
1091
+ "epoch": 0.52,
1092
+ "learning_rate": 5.90715667311412e-05,
1093
+ "loss": 1.0869,
1094
+ "step": 181
1095
+ },
1096
+ {
1097
+ "epoch": 0.53,
1098
+ "learning_rate": 5.895551257253385e-05,
1099
+ "loss": 2.543,
1100
+ "step": 182
1101
+ },
1102
+ {
1103
+ "epoch": 0.53,
1104
+ "learning_rate": 5.88394584139265e-05,
1105
+ "loss": 1.0225,
1106
+ "step": 183
1107
+ },
1108
+ {
1109
+ "epoch": 0.53,
1110
+ "learning_rate": 5.872340425531915e-05,
1111
+ "loss": 1.7207,
1112
+ "step": 184
1113
+ },
1114
+ {
1115
+ "epoch": 0.54,
1116
+ "learning_rate": 5.86073500967118e-05,
1117
+ "loss": 1.0156,
1118
+ "step": 185
1119
+ },
1120
+ {
1121
+ "epoch": 0.54,
1122
+ "learning_rate": 5.849129593810445e-05,
1123
+ "loss": 0.917,
1124
+ "step": 186
1125
+ },
1126
+ {
1127
+ "epoch": 0.54,
1128
+ "learning_rate": 5.8375241779497095e-05,
1129
+ "loss": 1.4346,
1130
+ "step": 187
1131
+ },
1132
+ {
1133
+ "epoch": 0.54,
1134
+ "learning_rate": 5.825918762088975e-05,
1135
+ "loss": 1.4922,
1136
+ "step": 188
1137
+ },
1138
+ {
1139
+ "epoch": 0.55,
1140
+ "learning_rate": 5.81431334622824e-05,
1141
+ "loss": 1.3213,
1142
+ "step": 189
1143
+ },
1144
+ {
1145
+ "epoch": 0.55,
1146
+ "learning_rate": 5.8027079303675055e-05,
1147
+ "loss": 1.0684,
1148
+ "step": 190
1149
+ },
1150
+ {
1151
+ "epoch": 0.55,
1152
+ "learning_rate": 5.79110251450677e-05,
1153
+ "loss": 0.8037,
1154
+ "step": 191
1155
+ },
1156
+ {
1157
+ "epoch": 0.56,
1158
+ "learning_rate": 5.7794970986460346e-05,
1159
+ "loss": 1.9531,
1160
+ "step": 192
1161
+ },
1162
+ {
1163
+ "epoch": 0.56,
1164
+ "learning_rate": 5.7678916827853e-05,
1165
+ "loss": 1.0098,
1166
+ "step": 193
1167
+ },
1168
+ {
1169
+ "epoch": 0.56,
1170
+ "learning_rate": 5.756286266924565e-05,
1171
+ "loss": 1.0127,
1172
+ "step": 194
1173
+ },
1174
+ {
1175
+ "epoch": 0.57,
1176
+ "learning_rate": 5.74468085106383e-05,
1177
+ "loss": 0.4705,
1178
+ "step": 195
1179
+ },
1180
+ {
1181
+ "epoch": 0.57,
1182
+ "learning_rate": 5.733075435203095e-05,
1183
+ "loss": 1.1465,
1184
+ "step": 196
1185
+ },
1186
+ {
1187
+ "epoch": 0.57,
1188
+ "learning_rate": 5.72147001934236e-05,
1189
+ "loss": 1.5605,
1190
+ "step": 197
1191
+ },
1192
+ {
1193
+ "epoch": 0.57,
1194
+ "learning_rate": 5.7098646034816254e-05,
1195
+ "loss": 1.6436,
1196
+ "step": 198
1197
+ },
1198
+ {
1199
+ "epoch": 0.58,
1200
+ "learning_rate": 5.6982591876208896e-05,
1201
+ "loss": 0.7319,
1202
+ "step": 199
1203
+ },
1204
+ {
1205
+ "epoch": 0.58,
1206
+ "learning_rate": 5.6866537717601545e-05,
1207
+ "loss": 1.4277,
1208
+ "step": 200
1209
+ },
1210
+ {
1211
+ "epoch": 0.58,
1212
+ "learning_rate": 5.67504835589942e-05,
1213
+ "loss": 1.0391,
1214
+ "step": 201
1215
+ },
1216
+ {
1217
+ "epoch": 0.59,
1218
+ "learning_rate": 5.6634429400386843e-05,
1219
+ "loss": 0.9307,
1220
+ "step": 202
1221
+ },
1222
+ {
1223
+ "epoch": 0.59,
1224
+ "learning_rate": 5.65183752417795e-05,
1225
+ "loss": 1.6367,
1226
+ "step": 203
1227
+ },
1228
+ {
1229
+ "epoch": 0.59,
1230
+ "learning_rate": 5.640232108317215e-05,
1231
+ "loss": 1.2861,
1232
+ "step": 204
1233
+ },
1234
+ {
1235
+ "epoch": 0.59,
1236
+ "learning_rate": 5.62862669245648e-05,
1237
+ "loss": 1.4727,
1238
+ "step": 205
1239
+ },
1240
+ {
1241
+ "epoch": 0.6,
1242
+ "learning_rate": 5.6170212765957446e-05,
1243
+ "loss": 0.9707,
1244
+ "step": 206
1245
+ },
1246
+ {
1247
+ "epoch": 0.6,
1248
+ "learning_rate": 5.6054158607350095e-05,
1249
+ "loss": 0.937,
1250
+ "step": 207
1251
+ },
1252
+ {
1253
+ "epoch": 0.6,
1254
+ "learning_rate": 5.593810444874275e-05,
1255
+ "loss": 1.1699,
1256
+ "step": 208
1257
+ },
1258
+ {
1259
+ "epoch": 0.61,
1260
+ "learning_rate": 5.58220502901354e-05,
1261
+ "loss": 0.8452,
1262
+ "step": 209
1263
+ },
1264
+ {
1265
+ "epoch": 0.61,
1266
+ "learning_rate": 5.570599613152804e-05,
1267
+ "loss": 0.999,
1268
+ "step": 210
1269
+ },
1270
+ {
1271
+ "epoch": 0.61,
1272
+ "learning_rate": 5.55899419729207e-05,
1273
+ "loss": 1.3018,
1274
+ "step": 211
1275
+ },
1276
+ {
1277
+ "epoch": 0.61,
1278
+ "learning_rate": 5.547388781431335e-05,
1279
+ "loss": 0.9292,
1280
+ "step": 212
1281
+ },
1282
+ {
1283
+ "epoch": 0.62,
1284
+ "learning_rate": 5.5357833655706e-05,
1285
+ "loss": 0.998,
1286
+ "step": 213
1287
+ },
1288
+ {
1289
+ "epoch": 0.62,
1290
+ "learning_rate": 5.5241779497098645e-05,
1291
+ "loss": 2.3984,
1292
+ "step": 214
1293
+ },
1294
+ {
1295
+ "epoch": 0.62,
1296
+ "learning_rate": 5.5125725338491294e-05,
1297
+ "loss": 0.9868,
1298
+ "step": 215
1299
+ },
1300
+ {
1301
+ "epoch": 0.63,
1302
+ "learning_rate": 5.500967117988395e-05,
1303
+ "loss": 0.8711,
1304
+ "step": 216
1305
+ },
1306
+ {
1307
+ "epoch": 0.63,
1308
+ "learning_rate": 5.48936170212766e-05,
1309
+ "loss": 1.2734,
1310
+ "step": 217
1311
+ },
1312
+ {
1313
+ "epoch": 0.63,
1314
+ "learning_rate": 5.477756286266924e-05,
1315
+ "loss": 1.6826,
1316
+ "step": 218
1317
+ },
1318
+ {
1319
+ "epoch": 0.63,
1320
+ "learning_rate": 5.46615087040619e-05,
1321
+ "loss": 1.2676,
1322
+ "step": 219
1323
+ },
1324
+ {
1325
+ "epoch": 0.64,
1326
+ "learning_rate": 5.4545454545454546e-05,
1327
+ "loss": 1.0039,
1328
+ "step": 220
1329
+ },
1330
+ {
1331
+ "epoch": 0.64,
1332
+ "learning_rate": 5.44294003868472e-05,
1333
+ "loss": 1.1943,
1334
+ "step": 221
1335
+ },
1336
+ {
1337
+ "epoch": 0.64,
1338
+ "learning_rate": 5.4313346228239844e-05,
1339
+ "loss": 0.9438,
1340
+ "step": 222
1341
+ },
1342
+ {
1343
+ "epoch": 0.65,
1344
+ "learning_rate": 5.419729206963249e-05,
1345
+ "loss": 0.915,
1346
+ "step": 223
1347
+ },
1348
+ {
1349
+ "epoch": 0.65,
1350
+ "learning_rate": 5.408123791102515e-05,
1351
+ "loss": 0.9854,
1352
+ "step": 224
1353
+ },
1354
+ {
1355
+ "epoch": 0.65,
1356
+ "learning_rate": 5.39651837524178e-05,
1357
+ "loss": 0.7554,
1358
+ "step": 225
1359
+ },
1360
+ {
1361
+ "epoch": 0.66,
1362
+ "learning_rate": 5.384912959381045e-05,
1363
+ "loss": 1.1641,
1364
+ "step": 226
1365
+ },
1366
+ {
1367
+ "epoch": 0.66,
1368
+ "learning_rate": 5.3733075435203096e-05,
1369
+ "loss": 1.1104,
1370
+ "step": 227
1371
+ },
1372
+ {
1373
+ "epoch": 0.66,
1374
+ "learning_rate": 5.3617021276595745e-05,
1375
+ "loss": 1.0586,
1376
+ "step": 228
1377
+ },
1378
+ {
1379
+ "epoch": 0.66,
1380
+ "learning_rate": 5.3500967117988394e-05,
1381
+ "loss": 1.2461,
1382
+ "step": 229
1383
+ },
1384
+ {
1385
+ "epoch": 0.67,
1386
+ "learning_rate": 5.338491295938104e-05,
1387
+ "loss": 0.5435,
1388
+ "step": 230
1389
+ },
1390
+ {
1391
+ "epoch": 0.67,
1392
+ "learning_rate": 5.32688588007737e-05,
1393
+ "loss": 0.9458,
1394
+ "step": 231
1395
+ },
1396
+ {
1397
+ "epoch": 0.67,
1398
+ "learning_rate": 5.315280464216635e-05,
1399
+ "loss": 1.1934,
1400
+ "step": 232
1401
+ },
1402
+ {
1403
+ "epoch": 0.68,
1404
+ "learning_rate": 5.303675048355899e-05,
1405
+ "loss": 0.3909,
1406
+ "step": 233
1407
+ },
1408
+ {
1409
+ "epoch": 0.68,
1410
+ "learning_rate": 5.2920696324951646e-05,
1411
+ "loss": 1.0605,
1412
+ "step": 234
1413
+ },
1414
+ {
1415
+ "epoch": 0.68,
1416
+ "learning_rate": 5.2804642166344295e-05,
1417
+ "loss": 1.084,
1418
+ "step": 235
1419
+ },
1420
+ {
1421
+ "epoch": 0.68,
1422
+ "learning_rate": 5.2688588007736944e-05,
1423
+ "loss": 0.4778,
1424
+ "step": 236
1425
+ },
1426
+ {
1427
+ "epoch": 0.69,
1428
+ "learning_rate": 5.257253384912959e-05,
1429
+ "loss": 0.9976,
1430
+ "step": 237
1431
+ },
1432
+ {
1433
+ "epoch": 0.69,
1434
+ "learning_rate": 5.245647969052224e-05,
1435
+ "loss": 0.4556,
1436
+ "step": 238
1437
+ },
1438
+ {
1439
+ "epoch": 0.69,
1440
+ "learning_rate": 5.23404255319149e-05,
1441
+ "loss": 0.8711,
1442
+ "step": 239
1443
+ },
1444
+ {
1445
+ "epoch": 0.7,
1446
+ "learning_rate": 5.222437137330755e-05,
1447
+ "loss": 0.7954,
1448
+ "step": 240
1449
+ },
1450
+ {
1451
+ "epoch": 0.7,
1452
+ "learning_rate": 5.210831721470019e-05,
1453
+ "loss": 0.2117,
1454
+ "step": 241
1455
+ },
1456
+ {
1457
+ "epoch": 0.7,
1458
+ "learning_rate": 5.1992263056092845e-05,
1459
+ "loss": 1.2148,
1460
+ "step": 242
1461
+ },
1462
+ {
1463
+ "epoch": 0.7,
1464
+ "learning_rate": 5.1876208897485494e-05,
1465
+ "loss": 1.1367,
1466
+ "step": 243
1467
+ },
1468
+ {
1469
+ "epoch": 0.71,
1470
+ "learning_rate": 5.176015473887815e-05,
1471
+ "loss": 1.1689,
1472
+ "step": 244
1473
+ },
1474
+ {
1475
+ "epoch": 0.71,
1476
+ "learning_rate": 5.164410058027079e-05,
1477
+ "loss": 0.7588,
1478
+ "step": 245
1479
+ },
1480
+ {
1481
+ "epoch": 0.71,
1482
+ "learning_rate": 5.152804642166344e-05,
1483
+ "loss": 1.0527,
1484
+ "step": 246
1485
+ },
1486
+ {
1487
+ "epoch": 0.72,
1488
+ "learning_rate": 5.14119922630561e-05,
1489
+ "loss": 1.3223,
1490
+ "step": 247
1491
+ },
1492
+ {
1493
+ "epoch": 0.72,
1494
+ "learning_rate": 5.1295938104448746e-05,
1495
+ "loss": 1.0264,
1496
+ "step": 248
1497
+ },
1498
+ {
1499
+ "epoch": 0.72,
1500
+ "learning_rate": 5.1179883945841395e-05,
1501
+ "loss": 0.9502,
1502
+ "step": 249
1503
+ },
1504
+ {
1505
+ "epoch": 0.72,
1506
+ "learning_rate": 5.1063829787234044e-05,
1507
+ "loss": 0.6611,
1508
+ "step": 250
1509
+ },
1510
+ {
1511
+ "epoch": 0.73,
1512
+ "learning_rate": 5.094777562862669e-05,
1513
+ "loss": 0.8711,
1514
+ "step": 251
1515
+ },
1516
+ {
1517
+ "epoch": 0.73,
1518
+ "learning_rate": 5.083172147001935e-05,
1519
+ "loss": 0.7632,
1520
+ "step": 252
1521
+ },
1522
+ {
1523
+ "epoch": 0.73,
1524
+ "learning_rate": 5.071566731141199e-05,
1525
+ "loss": 0.6128,
1526
+ "step": 253
1527
+ },
1528
+ {
1529
+ "epoch": 0.74,
1530
+ "learning_rate": 5.059961315280464e-05,
1531
+ "loss": 1.7568,
1532
+ "step": 254
1533
+ },
1534
+ {
1535
+ "epoch": 0.74,
1536
+ "learning_rate": 5.0483558994197296e-05,
1537
+ "loss": 0.6343,
1538
+ "step": 255
1539
+ },
1540
+ {
1541
+ "epoch": 0.74,
1542
+ "learning_rate": 5.036750483558994e-05,
1543
+ "loss": 1.0986,
1544
+ "step": 256
1545
+ },
1546
+ {
1547
+ "epoch": 0.74,
1548
+ "learning_rate": 5.0251450676982594e-05,
1549
+ "loss": 1.0752,
1550
+ "step": 257
1551
+ },
1552
+ {
1553
+ "epoch": 0.75,
1554
+ "learning_rate": 5.013539651837524e-05,
1555
+ "loss": 0.9414,
1556
+ "step": 258
1557
+ },
1558
+ {
1559
+ "epoch": 0.75,
1560
+ "learning_rate": 5.001934235976789e-05,
1561
+ "loss": 1.0146,
1562
+ "step": 259
1563
+ },
1564
+ {
1565
+ "epoch": 0.75,
1566
+ "learning_rate": 4.990328820116054e-05,
1567
+ "loss": 0.5898,
1568
+ "step": 260
1569
+ },
1570
+ {
1571
+ "epoch": 0.76,
1572
+ "learning_rate": 4.978723404255319e-05,
1573
+ "loss": 0.8062,
1574
+ "step": 261
1575
+ },
1576
+ {
1577
+ "epoch": 0.76,
1578
+ "learning_rate": 4.9671179883945846e-05,
1579
+ "loss": 1.1055,
1580
+ "step": 262
1581
+ },
1582
+ {
1583
+ "epoch": 0.76,
1584
+ "learning_rate": 4.9555125725338495e-05,
1585
+ "loss": 0.8584,
1586
+ "step": 263
1587
+ },
1588
+ {
1589
+ "epoch": 0.77,
1590
+ "learning_rate": 4.943907156673114e-05,
1591
+ "loss": 0.8101,
1592
+ "step": 264
1593
+ },
1594
+ {
1595
+ "epoch": 0.77,
1596
+ "learning_rate": 4.932301740812379e-05,
1597
+ "loss": 0.8618,
1598
+ "step": 265
1599
+ },
1600
+ {
1601
+ "epoch": 0.77,
1602
+ "learning_rate": 4.920696324951644e-05,
1603
+ "loss": 0.5723,
1604
+ "step": 266
1605
+ },
1606
+ {
1607
+ "epoch": 0.77,
1608
+ "learning_rate": 4.90909090909091e-05,
1609
+ "loss": 1.5088,
1610
+ "step": 267
1611
+ },
1612
+ {
1613
+ "epoch": 0.78,
1614
+ "learning_rate": 4.897485493230174e-05,
1615
+ "loss": 0.407,
1616
+ "step": 268
1617
+ },
1618
+ {
1619
+ "epoch": 0.78,
1620
+ "learning_rate": 4.885880077369439e-05,
1621
+ "loss": 0.8809,
1622
+ "step": 269
1623
+ },
1624
+ {
1625
+ "epoch": 0.78,
1626
+ "learning_rate": 4.8742746615087045e-05,
1627
+ "loss": 1.0938,
1628
+ "step": 270
1629
+ },
1630
+ {
1631
+ "epoch": 0.79,
1632
+ "learning_rate": 4.8626692456479694e-05,
1633
+ "loss": 0.7402,
1634
+ "step": 271
1635
+ },
1636
+ {
1637
+ "epoch": 0.79,
1638
+ "learning_rate": 4.8510638297872336e-05,
1639
+ "loss": 1.0674,
1640
+ "step": 272
1641
+ },
1642
+ {
1643
+ "epoch": 0.79,
1644
+ "learning_rate": 4.839458413926499e-05,
1645
+ "loss": 0.4993,
1646
+ "step": 273
1647
+ },
1648
+ {
1649
+ "epoch": 0.79,
1650
+ "learning_rate": 4.827852998065764e-05,
1651
+ "loss": 0.998,
1652
+ "step": 274
1653
+ },
1654
+ {
1655
+ "epoch": 0.8,
1656
+ "learning_rate": 4.81624758220503e-05,
1657
+ "loss": 1.2129,
1658
+ "step": 275
1659
+ },
1660
+ {
1661
+ "epoch": 0.8,
1662
+ "learning_rate": 4.804642166344294e-05,
1663
+ "loss": 1.208,
1664
+ "step": 276
1665
+ },
1666
+ {
1667
+ "epoch": 0.8,
1668
+ "learning_rate": 4.793036750483559e-05,
1669
+ "loss": 1.3545,
1670
+ "step": 277
1671
+ },
1672
+ {
1673
+ "epoch": 0.81,
1674
+ "learning_rate": 4.7814313346228244e-05,
1675
+ "loss": 0.8594,
1676
+ "step": 278
1677
+ },
1678
+ {
1679
+ "epoch": 0.81,
1680
+ "learning_rate": 4.769825918762089e-05,
1681
+ "loss": 0.7935,
1682
+ "step": 279
1683
+ },
1684
+ {
1685
+ "epoch": 0.81,
1686
+ "learning_rate": 4.758220502901354e-05,
1687
+ "loss": 0.8599,
1688
+ "step": 280
1689
+ },
1690
+ {
1691
+ "epoch": 0.81,
1692
+ "learning_rate": 4.746615087040619e-05,
1693
+ "loss": 0.9775,
1694
+ "step": 281
1695
+ },
1696
+ {
1697
+ "epoch": 0.82,
1698
+ "learning_rate": 4.735009671179884e-05,
1699
+ "loss": 0.7212,
1700
+ "step": 282
1701
+ },
1702
+ {
1703
+ "epoch": 0.82,
1704
+ "learning_rate": 4.723404255319149e-05,
1705
+ "loss": 0.9355,
1706
+ "step": 283
1707
+ },
1708
+ {
1709
+ "epoch": 0.82,
1710
+ "learning_rate": 4.711798839458414e-05,
1711
+ "loss": 0.8828,
1712
+ "step": 284
1713
+ },
1714
+ {
1715
+ "epoch": 0.83,
1716
+ "learning_rate": 4.7001934235976794e-05,
1717
+ "loss": 0.7603,
1718
+ "step": 285
1719
+ },
1720
+ {
1721
+ "epoch": 0.83,
1722
+ "learning_rate": 4.688588007736944e-05,
1723
+ "loss": 0.417,
1724
+ "step": 286
1725
+ },
1726
+ {
1727
+ "epoch": 0.83,
1728
+ "learning_rate": 4.6769825918762085e-05,
1729
+ "loss": 0.7583,
1730
+ "step": 287
1731
+ },
1732
+ {
1733
+ "epoch": 0.83,
1734
+ "learning_rate": 4.665377176015474e-05,
1735
+ "loss": 0.5708,
1736
+ "step": 288
1737
+ },
1738
+ {
1739
+ "epoch": 0.84,
1740
+ "learning_rate": 4.653771760154739e-05,
1741
+ "loss": 1.377,
1742
+ "step": 289
1743
+ },
1744
+ {
1745
+ "epoch": 0.84,
1746
+ "learning_rate": 4.642166344294004e-05,
1747
+ "loss": 1.5684,
1748
+ "step": 290
1749
+ },
1750
+ {
1751
+ "epoch": 0.84,
1752
+ "learning_rate": 4.630560928433269e-05,
1753
+ "loss": 0.9072,
1754
+ "step": 291
1755
+ },
1756
+ {
1757
+ "epoch": 0.85,
1758
+ "learning_rate": 4.618955512572534e-05,
1759
+ "loss": 0.8252,
1760
+ "step": 292
1761
+ },
1762
+ {
1763
+ "epoch": 0.85,
1764
+ "learning_rate": 4.607350096711799e-05,
1765
+ "loss": 1.085,
1766
+ "step": 293
1767
+ },
1768
+ {
1769
+ "epoch": 0.85,
1770
+ "learning_rate": 4.595744680851064e-05,
1771
+ "loss": 0.8286,
1772
+ "step": 294
1773
+ },
1774
+ {
1775
+ "epoch": 0.86,
1776
+ "learning_rate": 4.5841392649903284e-05,
1777
+ "loss": 1.1875,
1778
+ "step": 295
1779
+ },
1780
+ {
1781
+ "epoch": 0.86,
1782
+ "learning_rate": 4.572533849129594e-05,
1783
+ "loss": 0.7661,
1784
+ "step": 296
1785
+ },
1786
+ {
1787
+ "epoch": 0.86,
1788
+ "learning_rate": 4.560928433268859e-05,
1789
+ "loss": 0.9819,
1790
+ "step": 297
1791
+ },
1792
+ {
1793
+ "epoch": 0.86,
1794
+ "learning_rate": 4.5493230174081244e-05,
1795
+ "loss": 1.1055,
1796
+ "step": 298
1797
+ },
1798
+ {
1799
+ "epoch": 0.87,
1800
+ "learning_rate": 4.537717601547389e-05,
1801
+ "loss": 0.7915,
1802
+ "step": 299
1803
+ },
1804
+ {
1805
+ "epoch": 0.87,
1806
+ "learning_rate": 4.5261121856866536e-05,
1807
+ "loss": 0.6772,
1808
+ "step": 300
1809
+ },
1810
+ {
1811
+ "epoch": 0.87,
1812
+ "learning_rate": 4.514506769825919e-05,
1813
+ "loss": 0.9434,
1814
+ "step": 301
1815
+ },
1816
+ {
1817
+ "epoch": 0.88,
1818
+ "learning_rate": 4.502901353965184e-05,
1819
+ "loss": 0.6226,
1820
+ "step": 302
1821
+ },
1822
+ {
1823
+ "epoch": 0.88,
1824
+ "learning_rate": 4.491295938104449e-05,
1825
+ "loss": 0.5186,
1826
+ "step": 303
1827
+ },
1828
+ {
1829
+ "epoch": 0.88,
1830
+ "learning_rate": 4.479690522243714e-05,
1831
+ "loss": 0.5127,
1832
+ "step": 304
1833
+ },
1834
+ {
1835
+ "epoch": 0.88,
1836
+ "learning_rate": 4.468085106382979e-05,
1837
+ "loss": 0.998,
1838
+ "step": 305
1839
+ },
1840
+ {
1841
+ "epoch": 0.89,
1842
+ "learning_rate": 4.4564796905222443e-05,
1843
+ "loss": 1.292,
1844
+ "step": 306
1845
+ },
1846
+ {
1847
+ "epoch": 0.89,
1848
+ "learning_rate": 4.4448742746615086e-05,
1849
+ "loss": 1.3428,
1850
+ "step": 307
1851
+ },
1852
+ {
1853
+ "epoch": 0.89,
1854
+ "learning_rate": 4.4332688588007735e-05,
1855
+ "loss": 1.5078,
1856
+ "step": 308
1857
+ },
1858
+ {
1859
+ "epoch": 0.9,
1860
+ "learning_rate": 4.421663442940039e-05,
1861
+ "loss": 0.5425,
1862
+ "step": 309
1863
+ },
1864
+ {
1865
+ "epoch": 0.9,
1866
+ "learning_rate": 4.410058027079303e-05,
1867
+ "loss": 1.1709,
1868
+ "step": 310
1869
+ },
1870
+ {
1871
+ "epoch": 0.9,
1872
+ "learning_rate": 4.398452611218569e-05,
1873
+ "loss": 0.5474,
1874
+ "step": 311
1875
+ },
1876
+ {
1877
+ "epoch": 0.9,
1878
+ "learning_rate": 4.386847195357834e-05,
1879
+ "loss": 0.7163,
1880
+ "step": 312
1881
+ },
1882
+ {
1883
+ "epoch": 0.91,
1884
+ "learning_rate": 4.3752417794970987e-05,
1885
+ "loss": 0.9209,
1886
+ "step": 313
1887
+ },
1888
+ {
1889
+ "epoch": 0.91,
1890
+ "learning_rate": 4.3636363636363636e-05,
1891
+ "loss": 0.4958,
1892
+ "step": 314
1893
+ },
1894
+ {
1895
+ "epoch": 0.91,
1896
+ "learning_rate": 4.3520309477756285e-05,
1897
+ "loss": 0.5449,
1898
+ "step": 315
1899
+ },
1900
+ {
1901
+ "epoch": 0.92,
1902
+ "learning_rate": 4.340425531914894e-05,
1903
+ "loss": 0.416,
1904
+ "step": 316
1905
+ },
1906
+ {
1907
+ "epoch": 0.92,
1908
+ "learning_rate": 4.328820116054159e-05,
1909
+ "loss": 0.7725,
1910
+ "step": 317
1911
+ },
1912
+ {
1913
+ "epoch": 0.92,
1914
+ "learning_rate": 4.317214700193423e-05,
1915
+ "loss": 0.9961,
1916
+ "step": 318
1917
+ },
1918
+ {
1919
+ "epoch": 0.92,
1920
+ "learning_rate": 4.305609284332689e-05,
1921
+ "loss": 1.0986,
1922
+ "step": 319
1923
+ },
1924
+ {
1925
+ "epoch": 0.93,
1926
+ "learning_rate": 4.2940038684719537e-05,
1927
+ "loss": 0.7969,
1928
+ "step": 320
1929
+ },
1930
+ {
1931
+ "epoch": 0.93,
1932
+ "learning_rate": 4.282398452611219e-05,
1933
+ "loss": 0.6665,
1934
+ "step": 321
1935
+ },
1936
+ {
1937
+ "epoch": 0.93,
1938
+ "learning_rate": 4.2707930367504835e-05,
1939
+ "loss": 1.0449,
1940
+ "step": 322
1941
+ },
1942
+ {
1943
+ "epoch": 0.94,
1944
+ "learning_rate": 4.2591876208897484e-05,
1945
+ "loss": 0.5278,
1946
+ "step": 323
1947
+ },
1948
+ {
1949
+ "epoch": 0.94,
1950
+ "learning_rate": 4.247582205029014e-05,
1951
+ "loss": 1.2695,
1952
+ "step": 324
1953
+ },
1954
+ {
1955
+ "epoch": 0.94,
1956
+ "learning_rate": 4.235976789168279e-05,
1957
+ "loss": 1.2998,
1958
+ "step": 325
1959
+ },
1960
+ {
1961
+ "epoch": 0.94,
1962
+ "learning_rate": 4.224371373307543e-05,
1963
+ "loss": 0.5835,
1964
+ "step": 326
1965
+ },
1966
+ {
1967
+ "epoch": 0.95,
1968
+ "learning_rate": 4.2127659574468086e-05,
1969
+ "loss": 0.8564,
1970
+ "step": 327
1971
+ },
1972
+ {
1973
+ "epoch": 0.95,
1974
+ "learning_rate": 4.2011605415860735e-05,
1975
+ "loss": 0.8145,
1976
+ "step": 328
1977
+ },
1978
+ {
1979
+ "epoch": 0.95,
1980
+ "learning_rate": 4.189555125725339e-05,
1981
+ "loss": 0.8462,
1982
+ "step": 329
1983
+ },
1984
+ {
1985
+ "epoch": 0.96,
1986
+ "learning_rate": 4.1779497098646034e-05,
1987
+ "loss": 0.5107,
1988
+ "step": 330
1989
+ },
1990
+ {
1991
+ "epoch": 0.96,
1992
+ "learning_rate": 4.166344294003868e-05,
1993
+ "loss": 0.7842,
1994
+ "step": 331
1995
+ },
1996
+ {
1997
+ "epoch": 0.96,
1998
+ "learning_rate": 4.154738878143134e-05,
1999
+ "loss": 1.0312,
2000
+ "step": 332
2001
+ },
2002
+ {
2003
+ "epoch": 0.97,
2004
+ "learning_rate": 4.143133462282399e-05,
2005
+ "loss": 0.7061,
2006
+ "step": 333
2007
+ },
2008
+ {
2009
+ "epoch": 0.97,
2010
+ "learning_rate": 4.1315280464216636e-05,
2011
+ "loss": 1.1055,
2012
+ "step": 334
2013
+ },
2014
+ {
2015
+ "epoch": 0.97,
2016
+ "learning_rate": 4.1199226305609285e-05,
2017
+ "loss": 0.6479,
2018
+ "step": 335
2019
+ },
2020
+ {
2021
+ "epoch": 0.97,
2022
+ "learning_rate": 4.1083172147001934e-05,
2023
+ "loss": 0.6284,
2024
+ "step": 336
2025
+ },
2026
+ {
2027
+ "epoch": 0.98,
2028
+ "learning_rate": 4.0967117988394583e-05,
2029
+ "loss": 0.7573,
2030
+ "step": 337
2031
+ },
2032
+ {
2033
+ "epoch": 0.98,
2034
+ "learning_rate": 4.085106382978723e-05,
2035
+ "loss": 1.6924,
2036
+ "step": 338
2037
+ },
2038
+ {
2039
+ "epoch": 0.98,
2040
+ "learning_rate": 4.073500967117989e-05,
2041
+ "loss": 1.2656,
2042
+ "step": 339
2043
+ },
2044
+ {
2045
+ "epoch": 0.99,
2046
+ "learning_rate": 4.061895551257254e-05,
2047
+ "loss": 1.209,
2048
+ "step": 340
2049
+ },
2050
+ {
2051
+ "epoch": 0.99,
2052
+ "learning_rate": 4.050290135396518e-05,
2053
+ "loss": 0.6792,
2054
+ "step": 341
2055
+ },
2056
+ {
2057
+ "epoch": 0.99,
2058
+ "learning_rate": 4.0386847195357835e-05,
2059
+ "loss": 0.5415,
2060
+ "step": 342
2061
+ },
2062
+ {
2063
+ "epoch": 0.99,
2064
+ "learning_rate": 4.0270793036750484e-05,
2065
+ "loss": 0.6519,
2066
+ "step": 343
2067
+ },
2068
+ {
2069
+ "epoch": 1.0,
2070
+ "learning_rate": 4.0154738878143133e-05,
2071
+ "loss": 0.5938,
2072
+ "step": 344
2073
+ },
2074
+ {
2075
+ "epoch": 1.0,
2076
+ "learning_rate": 4.003868471953578e-05,
2077
+ "loss": 1.6006,
2078
+ "step": 345
2079
+ },
2080
+ {
2081
+ "epoch": 1.0,
2082
+ "learning_rate": 3.992263056092843e-05,
2083
+ "loss": 1.0391,
2084
+ "step": 346
2085
+ },
2086
+ {
2087
+ "epoch": 1.01,
2088
+ "learning_rate": 3.980657640232109e-05,
2089
+ "loss": 0.5776,
2090
+ "step": 347
2091
+ },
2092
+ {
2093
+ "epoch": 1.01,
2094
+ "learning_rate": 3.9690522243713736e-05,
2095
+ "loss": 0.7881,
2096
+ "step": 348
2097
+ },
2098
+ {
2099
+ "epoch": 1.01,
2100
+ "learning_rate": 3.957446808510638e-05,
2101
+ "loss": 1.1357,
2102
+ "step": 349
2103
+ },
2104
+ {
2105
+ "epoch": 1.01,
2106
+ "learning_rate": 3.9458413926499034e-05,
2107
+ "loss": 0.6836,
2108
+ "step": 350
2109
+ },
2110
+ {
2111
+ "epoch": 1.02,
2112
+ "learning_rate": 3.934235976789168e-05,
2113
+ "loss": 0.5996,
2114
+ "step": 351
2115
+ },
2116
+ {
2117
+ "epoch": 1.02,
2118
+ "learning_rate": 3.922630560928434e-05,
2119
+ "loss": 0.25,
2120
+ "step": 352
2121
+ },
2122
+ {
2123
+ "epoch": 1.02,
2124
+ "learning_rate": 3.911025145067698e-05,
2125
+ "loss": 1.0361,
2126
+ "step": 353
2127
+ },
2128
+ {
2129
+ "epoch": 1.03,
2130
+ "learning_rate": 3.899419729206963e-05,
2131
+ "loss": 0.2573,
2132
+ "step": 354
2133
+ },
2134
+ {
2135
+ "epoch": 1.03,
2136
+ "learning_rate": 3.8878143133462286e-05,
2137
+ "loss": 0.9463,
2138
+ "step": 355
2139
+ },
2140
+ {
2141
+ "epoch": 1.03,
2142
+ "learning_rate": 3.8762088974854935e-05,
2143
+ "loss": 1.1104,
2144
+ "step": 356
2145
+ },
2146
+ {
2147
+ "epoch": 1.03,
2148
+ "learning_rate": 3.8646034816247584e-05,
2149
+ "loss": 0.7646,
2150
+ "step": 357
2151
+ },
2152
+ {
2153
+ "epoch": 1.04,
2154
+ "learning_rate": 3.852998065764023e-05,
2155
+ "loss": 0.5918,
2156
+ "step": 358
2157
+ },
2158
+ {
2159
+ "epoch": 1.04,
2160
+ "learning_rate": 3.841392649903288e-05,
2161
+ "loss": 1.4531,
2162
+ "step": 359
2163
+ },
2164
+ {
2165
+ "epoch": 1.04,
2166
+ "learning_rate": 3.829787234042554e-05,
2167
+ "loss": 0.9487,
2168
+ "step": 360
2169
+ },
2170
+ {
2171
+ "epoch": 1.05,
2172
+ "learning_rate": 3.818181818181818e-05,
2173
+ "loss": 0.9604,
2174
+ "step": 361
2175
+ },
2176
+ {
2177
+ "epoch": 1.05,
2178
+ "learning_rate": 3.806576402321083e-05,
2179
+ "loss": 0.6528,
2180
+ "step": 362
2181
+ },
2182
+ {
2183
+ "epoch": 1.05,
2184
+ "learning_rate": 3.7949709864603485e-05,
2185
+ "loss": 1.0029,
2186
+ "step": 363
2187
+ },
2188
+ {
2189
+ "epoch": 1.06,
2190
+ "learning_rate": 3.783365570599613e-05,
2191
+ "loss": 0.7422,
2192
+ "step": 364
2193
+ },
2194
+ {
2195
+ "epoch": 1.06,
2196
+ "learning_rate": 3.771760154738878e-05,
2197
+ "loss": 0.7471,
2198
+ "step": 365
2199
+ },
2200
+ {
2201
+ "epoch": 1.06,
2202
+ "learning_rate": 3.760154738878143e-05,
2203
+ "loss": 0.938,
2204
+ "step": 366
2205
+ },
2206
+ {
2207
+ "epoch": 1.06,
2208
+ "learning_rate": 3.748549323017408e-05,
2209
+ "loss": 0.6191,
2210
+ "step": 367
2211
+ },
2212
+ {
2213
+ "epoch": 1.07,
2214
+ "learning_rate": 3.736943907156673e-05,
2215
+ "loss": 0.46,
2216
+ "step": 368
2217
+ },
2218
+ {
2219
+ "epoch": 1.07,
2220
+ "learning_rate": 3.725338491295938e-05,
2221
+ "loss": 1.1885,
2222
+ "step": 369
2223
+ },
2224
+ {
2225
+ "epoch": 1.07,
2226
+ "learning_rate": 3.7137330754352035e-05,
2227
+ "loss": 1.3447,
2228
+ "step": 370
2229
+ },
2230
+ {
2231
+ "epoch": 1.08,
2232
+ "learning_rate": 3.7021276595744684e-05,
2233
+ "loss": 0.749,
2234
+ "step": 371
2235
+ },
2236
+ {
2237
+ "epoch": 1.08,
2238
+ "learning_rate": 3.6905222437137326e-05,
2239
+ "loss": 0.4883,
2240
+ "step": 372
2241
+ },
2242
+ {
2243
+ "epoch": 1.08,
2244
+ "learning_rate": 3.678916827852998e-05,
2245
+ "loss": 0.5981,
2246
+ "step": 373
2247
+ },
2248
+ {
2249
+ "epoch": 1.08,
2250
+ "learning_rate": 3.667311411992263e-05,
2251
+ "loss": 0.9893,
2252
+ "step": 374
2253
+ },
2254
+ {
2255
+ "epoch": 1.09,
2256
+ "learning_rate": 3.655705996131529e-05,
2257
+ "loss": 0.7104,
2258
+ "step": 375
2259
+ },
2260
+ {
2261
+ "epoch": 1.09,
2262
+ "learning_rate": 3.644100580270793e-05,
2263
+ "loss": 0.5493,
2264
+ "step": 376
2265
+ },
2266
+ {
2267
+ "epoch": 1.09,
2268
+ "learning_rate": 3.632495164410058e-05,
2269
+ "loss": 0.5688,
2270
+ "step": 377
2271
+ },
2272
+ {
2273
+ "epoch": 1.1,
2274
+ "learning_rate": 3.6208897485493234e-05,
2275
+ "loss": 0.4919,
2276
+ "step": 378
2277
+ },
2278
+ {
2279
+ "epoch": 1.1,
2280
+ "learning_rate": 3.609284332688588e-05,
2281
+ "loss": 0.8325,
2282
+ "step": 379
2283
+ },
2284
+ {
2285
+ "epoch": 1.1,
2286
+ "learning_rate": 3.5976789168278525e-05,
2287
+ "loss": 0.4885,
2288
+ "step": 380
2289
+ },
2290
+ {
2291
+ "epoch": 1.1,
2292
+ "learning_rate": 3.586073500967118e-05,
2293
+ "loss": 1.084,
2294
+ "step": 381
2295
+ },
2296
+ {
2297
+ "epoch": 1.11,
2298
+ "learning_rate": 3.574468085106383e-05,
2299
+ "loss": 0.6152,
2300
+ "step": 382
2301
+ },
2302
+ {
2303
+ "epoch": 1.11,
2304
+ "learning_rate": 3.5628626692456486e-05,
2305
+ "loss": 0.489,
2306
+ "step": 383
2307
+ },
2308
+ {
2309
+ "epoch": 1.11,
2310
+ "learning_rate": 3.551257253384913e-05,
2311
+ "loss": 0.667,
2312
+ "step": 384
2313
+ },
2314
+ {
2315
+ "epoch": 1.12,
2316
+ "learning_rate": 3.539651837524178e-05,
2317
+ "loss": 0.606,
2318
+ "step": 385
2319
+ },
2320
+ {
2321
+ "epoch": 1.12,
2322
+ "learning_rate": 3.528046421663443e-05,
2323
+ "loss": 0.4199,
2324
+ "step": 386
2325
+ },
2326
+ {
2327
+ "epoch": 1.12,
2328
+ "learning_rate": 3.516441005802708e-05,
2329
+ "loss": 0.9307,
2330
+ "step": 387
2331
+ },
2332
+ {
2333
+ "epoch": 1.12,
2334
+ "learning_rate": 3.504835589941973e-05,
2335
+ "loss": 0.8564,
2336
+ "step": 388
2337
+ },
2338
+ {
2339
+ "epoch": 1.13,
2340
+ "learning_rate": 3.493230174081238e-05,
2341
+ "loss": 0.4531,
2342
+ "step": 389
2343
+ },
2344
+ {
2345
+ "epoch": 1.13,
2346
+ "learning_rate": 3.481624758220503e-05,
2347
+ "loss": 0.875,
2348
+ "step": 390
2349
+ },
2350
+ {
2351
+ "epoch": 1.13,
2352
+ "learning_rate": 3.470019342359768e-05,
2353
+ "loss": 1.0645,
2354
+ "step": 391
2355
+ },
2356
+ {
2357
+ "epoch": 1.14,
2358
+ "learning_rate": 3.458413926499033e-05,
2359
+ "loss": 0.7993,
2360
+ "step": 392
2361
+ },
2362
+ {
2363
+ "epoch": 1.14,
2364
+ "learning_rate": 3.446808510638298e-05,
2365
+ "loss": 0.8906,
2366
+ "step": 393
2367
+ },
2368
+ {
2369
+ "epoch": 1.14,
2370
+ "learning_rate": 3.435203094777563e-05,
2371
+ "loss": 1.0947,
2372
+ "step": 394
2373
+ },
2374
+ {
2375
+ "epoch": 1.14,
2376
+ "learning_rate": 3.4235976789168274e-05,
2377
+ "loss": 1.2812,
2378
+ "step": 395
2379
+ },
2380
+ {
2381
+ "epoch": 1.15,
2382
+ "learning_rate": 3.411992263056093e-05,
2383
+ "loss": 0.4214,
2384
+ "step": 396
2385
+ },
2386
+ {
2387
+ "epoch": 1.15,
2388
+ "learning_rate": 3.400386847195358e-05,
2389
+ "loss": 1.1426,
2390
+ "step": 397
2391
+ },
2392
+ {
2393
+ "epoch": 1.15,
2394
+ "learning_rate": 3.388781431334623e-05,
2395
+ "loss": 0.5415,
2396
+ "step": 398
2397
+ },
2398
+ {
2399
+ "epoch": 1.16,
2400
+ "learning_rate": 3.377176015473888e-05,
2401
+ "loss": 1.0723,
2402
+ "step": 399
2403
+ },
2404
+ {
2405
+ "epoch": 1.16,
2406
+ "learning_rate": 3.3655705996131526e-05,
2407
+ "loss": 0.7412,
2408
+ "step": 400
2409
+ },
2410
+ {
2411
+ "epoch": 1.16,
2412
+ "learning_rate": 3.353965183752418e-05,
2413
+ "loss": 0.9688,
2414
+ "step": 401
2415
+ },
2416
+ {
2417
+ "epoch": 1.17,
2418
+ "learning_rate": 3.342359767891683e-05,
2419
+ "loss": 0.5547,
2420
+ "step": 402
2421
+ },
2422
+ {
2423
+ "epoch": 1.17,
2424
+ "learning_rate": 3.330754352030947e-05,
2425
+ "loss": 0.5977,
2426
+ "step": 403
2427
+ },
2428
+ {
2429
+ "epoch": 1.17,
2430
+ "learning_rate": 3.319148936170213e-05,
2431
+ "loss": 0.9434,
2432
+ "step": 404
2433
+ },
2434
+ {
2435
+ "epoch": 1.17,
2436
+ "learning_rate": 3.307543520309478e-05,
2437
+ "loss": 0.9668,
2438
+ "step": 405
2439
+ },
2440
+ {
2441
+ "epoch": 1.18,
2442
+ "learning_rate": 3.2959381044487434e-05,
2443
+ "loss": 0.8994,
2444
+ "step": 406
2445
+ },
2446
+ {
2447
+ "epoch": 1.18,
2448
+ "learning_rate": 3.2843326885880076e-05,
2449
+ "loss": 0.5439,
2450
+ "step": 407
2451
+ },
2452
+ {
2453
+ "epoch": 1.18,
2454
+ "learning_rate": 3.2727272727272725e-05,
2455
+ "loss": 1.168,
2456
+ "step": 408
2457
+ },
2458
+ {
2459
+ "epoch": 1.19,
2460
+ "learning_rate": 3.261121856866538e-05,
2461
+ "loss": 1.292,
2462
+ "step": 409
2463
+ },
2464
+ {
2465
+ "epoch": 1.19,
2466
+ "learning_rate": 3.249516441005803e-05,
2467
+ "loss": 0.6426,
2468
+ "step": 410
2469
+ },
2470
+ {
2471
+ "epoch": 1.19,
2472
+ "learning_rate": 3.237911025145068e-05,
2473
+ "loss": 0.3635,
2474
+ "step": 411
2475
+ },
2476
+ {
2477
+ "epoch": 1.19,
2478
+ "learning_rate": 3.226305609284333e-05,
2479
+ "loss": 0.9268,
2480
+ "step": 412
2481
+ },
2482
+ {
2483
+ "epoch": 1.2,
2484
+ "learning_rate": 3.214700193423598e-05,
2485
+ "loss": 0.4749,
2486
+ "step": 413
2487
+ },
2488
+ {
2489
+ "epoch": 1.2,
2490
+ "learning_rate": 3.203094777562863e-05,
2491
+ "loss": 0.8774,
2492
+ "step": 414
2493
+ },
2494
+ {
2495
+ "epoch": 1.2,
2496
+ "learning_rate": 3.1914893617021275e-05,
2497
+ "loss": 1.0264,
2498
+ "step": 415
2499
+ },
2500
+ {
2501
+ "epoch": 1.21,
2502
+ "learning_rate": 3.1798839458413924e-05,
2503
+ "loss": 0.8276,
2504
+ "step": 416
2505
+ },
2506
+ {
2507
+ "epoch": 1.21,
2508
+ "learning_rate": 3.168278529980658e-05,
2509
+ "loss": 1.0762,
2510
+ "step": 417
2511
+ },
2512
+ {
2513
+ "epoch": 1.21,
2514
+ "learning_rate": 3.156673114119923e-05,
2515
+ "loss": 1.0459,
2516
+ "step": 418
2517
+ },
2518
+ {
2519
+ "epoch": 1.21,
2520
+ "learning_rate": 3.145067698259188e-05,
2521
+ "loss": 0.6641,
2522
+ "step": 419
2523
+ },
2524
+ {
2525
+ "epoch": 1.22,
2526
+ "learning_rate": 3.133462282398453e-05,
2527
+ "loss": 1.9824,
2528
+ "step": 420
2529
+ },
2530
+ {
2531
+ "epoch": 1.22,
2532
+ "learning_rate": 3.1218568665377176e-05,
2533
+ "loss": 1.2617,
2534
+ "step": 421
2535
+ },
2536
+ {
2537
+ "epoch": 1.22,
2538
+ "learning_rate": 3.1102514506769825e-05,
2539
+ "loss": 0.5913,
2540
+ "step": 422
2541
+ },
2542
+ {
2543
+ "epoch": 1.23,
2544
+ "learning_rate": 3.0986460348162474e-05,
2545
+ "loss": 0.6343,
2546
+ "step": 423
2547
+ },
2548
+ {
2549
+ "epoch": 1.23,
2550
+ "learning_rate": 3.087040618955513e-05,
2551
+ "loss": 0.5215,
2552
+ "step": 424
2553
+ },
2554
+ {
2555
+ "epoch": 1.23,
2556
+ "learning_rate": 3.075435203094778e-05,
2557
+ "loss": 0.7549,
2558
+ "step": 425
2559
+ },
2560
+ {
2561
+ "epoch": 1.23,
2562
+ "learning_rate": 3.063829787234042e-05,
2563
+ "loss": 0.3704,
2564
+ "step": 426
2565
+ },
2566
+ {
2567
+ "epoch": 1.24,
2568
+ "learning_rate": 3.052224371373308e-05,
2569
+ "loss": 0.6704,
2570
+ "step": 427
2571
+ },
2572
+ {
2573
+ "epoch": 1.24,
2574
+ "learning_rate": 3.0406189555125726e-05,
2575
+ "loss": 0.9639,
2576
+ "step": 428
2577
+ },
2578
+ {
2579
+ "epoch": 1.24,
2580
+ "learning_rate": 3.0290135396518378e-05,
2581
+ "loss": 0.8818,
2582
+ "step": 429
2583
+ },
2584
+ {
2585
+ "epoch": 1.25,
2586
+ "learning_rate": 3.0174081237911027e-05,
2587
+ "loss": 0.6758,
2588
+ "step": 430
2589
+ },
2590
+ {
2591
+ "epoch": 1.25,
2592
+ "learning_rate": 3.0058027079303673e-05,
2593
+ "loss": 0.728,
2594
+ "step": 431
2595
+ },
2596
+ {
2597
+ "epoch": 1.25,
2598
+ "learning_rate": 2.9941972920696325e-05,
2599
+ "loss": 0.6206,
2600
+ "step": 432
2601
+ },
2602
+ {
2603
+ "epoch": 1.26,
2604
+ "learning_rate": 2.9825918762088974e-05,
2605
+ "loss": 0.7578,
2606
+ "step": 433
2607
+ },
2608
+ {
2609
+ "epoch": 1.26,
2610
+ "learning_rate": 2.9709864603481627e-05,
2611
+ "loss": 0.6523,
2612
+ "step": 434
2613
+ },
2614
+ {
2615
+ "epoch": 1.26,
2616
+ "learning_rate": 2.9593810444874276e-05,
2617
+ "loss": 0.4155,
2618
+ "step": 435
2619
+ },
2620
+ {
2621
+ "epoch": 1.26,
2622
+ "learning_rate": 2.9477756286266925e-05,
2623
+ "loss": 0.5464,
2624
+ "step": 436
2625
+ },
2626
+ {
2627
+ "epoch": 1.27,
2628
+ "learning_rate": 2.9361702127659574e-05,
2629
+ "loss": 0.4067,
2630
+ "step": 437
2631
+ },
2632
+ {
2633
+ "epoch": 1.27,
2634
+ "learning_rate": 2.9245647969052226e-05,
2635
+ "loss": 1.3379,
2636
+ "step": 438
2637
+ },
2638
+ {
2639
+ "epoch": 1.27,
2640
+ "learning_rate": 2.9129593810444875e-05,
2641
+ "loss": 1.0723,
2642
+ "step": 439
2643
+ },
2644
+ {
2645
+ "epoch": 1.28,
2646
+ "learning_rate": 2.9013539651837528e-05,
2647
+ "loss": 0.6567,
2648
+ "step": 440
2649
+ },
2650
+ {
2651
+ "epoch": 1.28,
2652
+ "learning_rate": 2.8897485493230173e-05,
2653
+ "loss": 0.4177,
2654
+ "step": 441
2655
+ },
2656
+ {
2657
+ "epoch": 1.28,
2658
+ "learning_rate": 2.8781431334622826e-05,
2659
+ "loss": 0.4412,
2660
+ "step": 442
2661
+ },
2662
+ {
2663
+ "epoch": 1.28,
2664
+ "learning_rate": 2.8665377176015475e-05,
2665
+ "loss": 0.3015,
2666
+ "step": 443
2667
+ },
2668
+ {
2669
+ "epoch": 1.29,
2670
+ "learning_rate": 2.8549323017408127e-05,
2671
+ "loss": 0.7876,
2672
+ "step": 444
2673
+ },
2674
+ {
2675
+ "epoch": 1.29,
2676
+ "learning_rate": 2.8433268858800773e-05,
2677
+ "loss": 0.7197,
2678
+ "step": 445
2679
+ },
2680
+ {
2681
+ "epoch": 1.29,
2682
+ "learning_rate": 2.8317214700193422e-05,
2683
+ "loss": 1.1719,
2684
+ "step": 446
2685
+ },
2686
+ {
2687
+ "epoch": 1.3,
2688
+ "learning_rate": 2.8201160541586074e-05,
2689
+ "loss": 1.2109,
2690
+ "step": 447
2691
+ },
2692
+ {
2693
+ "epoch": 1.3,
2694
+ "learning_rate": 2.8085106382978723e-05,
2695
+ "loss": 0.301,
2696
+ "step": 448
2697
+ },
2698
+ {
2699
+ "epoch": 1.3,
2700
+ "learning_rate": 2.7969052224371376e-05,
2701
+ "loss": 0.7065,
2702
+ "step": 449
2703
+ },
2704
+ {
2705
+ "epoch": 1.3,
2706
+ "learning_rate": 2.785299806576402e-05,
2707
+ "loss": 1.5684,
2708
+ "step": 450
2709
+ },
2710
+ {
2711
+ "epoch": 1.31,
2712
+ "learning_rate": 2.7736943907156674e-05,
2713
+ "loss": 0.6211,
2714
+ "step": 451
2715
+ },
2716
+ {
2717
+ "epoch": 1.31,
2718
+ "learning_rate": 2.7620889748549323e-05,
2719
+ "loss": 0.7876,
2720
+ "step": 452
2721
+ },
2722
+ {
2723
+ "epoch": 1.31,
2724
+ "learning_rate": 2.7504835589941975e-05,
2725
+ "loss": 0.7261,
2726
+ "step": 453
2727
+ },
2728
+ {
2729
+ "epoch": 1.32,
2730
+ "learning_rate": 2.738878143133462e-05,
2731
+ "loss": 0.6401,
2732
+ "step": 454
2733
+ },
2734
+ {
2735
+ "epoch": 1.32,
2736
+ "learning_rate": 2.7272727272727273e-05,
2737
+ "loss": 0.4714,
2738
+ "step": 455
2739
+ },
2740
+ {
2741
+ "epoch": 1.32,
2742
+ "learning_rate": 2.7156673114119922e-05,
2743
+ "loss": 0.6025,
2744
+ "step": 456
2745
+ },
2746
+ {
2747
+ "epoch": 1.32,
2748
+ "learning_rate": 2.7040618955512575e-05,
2749
+ "loss": 0.708,
2750
+ "step": 457
2751
+ },
2752
+ {
2753
+ "epoch": 1.33,
2754
+ "learning_rate": 2.6924564796905224e-05,
2755
+ "loss": 0.7397,
2756
+ "step": 458
2757
+ },
2758
+ {
2759
+ "epoch": 1.33,
2760
+ "learning_rate": 2.6808510638297873e-05,
2761
+ "loss": 0.8315,
2762
+ "step": 459
2763
+ },
2764
+ {
2765
+ "epoch": 1.33,
2766
+ "learning_rate": 2.669245647969052e-05,
2767
+ "loss": 0.7134,
2768
+ "step": 460
2769
+ },
2770
+ {
2771
+ "epoch": 1.34,
2772
+ "learning_rate": 2.6576402321083174e-05,
2773
+ "loss": 0.4951,
2774
+ "step": 461
2775
+ },
2776
+ {
2777
+ "epoch": 1.34,
2778
+ "learning_rate": 2.6460348162475823e-05,
2779
+ "loss": 0.8564,
2780
+ "step": 462
2781
+ },
2782
+ {
2783
+ "epoch": 1.34,
2784
+ "learning_rate": 2.6344294003868472e-05,
2785
+ "loss": 0.6763,
2786
+ "step": 463
2787
+ },
2788
+ {
2789
+ "epoch": 1.34,
2790
+ "learning_rate": 2.622823984526112e-05,
2791
+ "loss": 0.5425,
2792
+ "step": 464
2793
+ },
2794
+ {
2795
+ "epoch": 1.35,
2796
+ "learning_rate": 2.6112185686653773e-05,
2797
+ "loss": 1.7285,
2798
+ "step": 465
2799
+ },
2800
+ {
2801
+ "epoch": 1.35,
2802
+ "learning_rate": 2.5996131528046423e-05,
2803
+ "loss": 1.1133,
2804
+ "step": 466
2805
+ },
2806
+ {
2807
+ "epoch": 1.35,
2808
+ "learning_rate": 2.5880077369439075e-05,
2809
+ "loss": 0.8857,
2810
+ "step": 467
2811
+ },
2812
+ {
2813
+ "epoch": 1.36,
2814
+ "learning_rate": 2.576402321083172e-05,
2815
+ "loss": 0.4885,
2816
+ "step": 468
2817
+ },
2818
+ {
2819
+ "epoch": 1.36,
2820
+ "learning_rate": 2.5647969052224373e-05,
2821
+ "loss": 0.5234,
2822
+ "step": 469
2823
+ },
2824
+ {
2825
+ "epoch": 1.36,
2826
+ "learning_rate": 2.5531914893617022e-05,
2827
+ "loss": 1.1143,
2828
+ "step": 470
2829
+ },
2830
+ {
2831
+ "epoch": 1.37,
2832
+ "learning_rate": 2.5415860735009674e-05,
2833
+ "loss": 1.0,
2834
+ "step": 471
2835
+ },
2836
+ {
2837
+ "epoch": 1.37,
2838
+ "learning_rate": 2.529980657640232e-05,
2839
+ "loss": 0.4348,
2840
+ "step": 472
2841
+ },
2842
+ {
2843
+ "epoch": 1.37,
2844
+ "learning_rate": 2.518375241779497e-05,
2845
+ "loss": 1.0801,
2846
+ "step": 473
2847
+ },
2848
+ {
2849
+ "epoch": 1.37,
2850
+ "learning_rate": 2.506769825918762e-05,
2851
+ "loss": 0.8135,
2852
+ "step": 474
2853
+ },
2854
+ {
2855
+ "epoch": 1.38,
2856
+ "learning_rate": 2.495164410058027e-05,
2857
+ "loss": 0.4106,
2858
+ "step": 475
2859
+ },
2860
+ {
2861
+ "epoch": 1.38,
2862
+ "learning_rate": 2.4835589941972923e-05,
2863
+ "loss": 0.7329,
2864
+ "step": 476
2865
+ },
2866
+ {
2867
+ "epoch": 1.38,
2868
+ "learning_rate": 2.471953578336557e-05,
2869
+ "loss": 1.46,
2870
+ "step": 477
2871
+ },
2872
+ {
2873
+ "epoch": 1.39,
2874
+ "learning_rate": 2.460348162475822e-05,
2875
+ "loss": 0.2554,
2876
+ "step": 478
2877
+ },
2878
+ {
2879
+ "epoch": 1.39,
2880
+ "learning_rate": 2.448742746615087e-05,
2881
+ "loss": 0.6641,
2882
+ "step": 479
2883
+ },
2884
+ {
2885
+ "epoch": 1.39,
2886
+ "learning_rate": 2.4371373307543522e-05,
2887
+ "loss": 0.3413,
2888
+ "step": 480
2889
+ },
2890
+ {
2891
+ "epoch": 1.39,
2892
+ "learning_rate": 2.4255319148936168e-05,
2893
+ "loss": 0.5776,
2894
+ "step": 481
2895
+ },
2896
+ {
2897
+ "epoch": 1.4,
2898
+ "learning_rate": 2.413926499032882e-05,
2899
+ "loss": 0.5396,
2900
+ "step": 482
2901
+ },
2902
+ {
2903
+ "epoch": 1.4,
2904
+ "learning_rate": 2.402321083172147e-05,
2905
+ "loss": 0.6875,
2906
+ "step": 483
2907
+ },
2908
+ {
2909
+ "epoch": 1.4,
2910
+ "learning_rate": 2.3907156673114122e-05,
2911
+ "loss": 0.7529,
2912
+ "step": 484
2913
+ },
2914
+ {
2915
+ "epoch": 1.41,
2916
+ "learning_rate": 2.379110251450677e-05,
2917
+ "loss": 0.5298,
2918
+ "step": 485
2919
+ },
2920
+ {
2921
+ "epoch": 1.41,
2922
+ "learning_rate": 2.367504835589942e-05,
2923
+ "loss": 1.1924,
2924
+ "step": 486
2925
+ },
2926
+ {
2927
+ "epoch": 1.41,
2928
+ "learning_rate": 2.355899419729207e-05,
2929
+ "loss": 0.479,
2930
+ "step": 487
2931
+ },
2932
+ {
2933
+ "epoch": 1.41,
2934
+ "learning_rate": 2.344294003868472e-05,
2935
+ "loss": 0.73,
2936
+ "step": 488
2937
+ },
2938
+ {
2939
+ "epoch": 1.42,
2940
+ "learning_rate": 2.332688588007737e-05,
2941
+ "loss": 0.7144,
2942
+ "step": 489
2943
+ },
2944
+ {
2945
+ "epoch": 1.42,
2946
+ "learning_rate": 2.321083172147002e-05,
2947
+ "loss": 0.205,
2948
+ "step": 490
2949
+ },
2950
+ {
2951
+ "epoch": 1.42,
2952
+ "learning_rate": 2.309477756286267e-05,
2953
+ "loss": 1.4053,
2954
+ "step": 491
2955
+ },
2956
+ {
2957
+ "epoch": 1.43,
2958
+ "learning_rate": 2.297872340425532e-05,
2959
+ "loss": 0.5771,
2960
+ "step": 492
2961
+ },
2962
+ {
2963
+ "epoch": 1.43,
2964
+ "learning_rate": 2.286266924564797e-05,
2965
+ "loss": 1.2969,
2966
+ "step": 493
2967
+ },
2968
+ {
2969
+ "epoch": 1.43,
2970
+ "learning_rate": 2.2746615087040622e-05,
2971
+ "loss": 0.7012,
2972
+ "step": 494
2973
+ },
2974
+ {
2975
+ "epoch": 1.43,
2976
+ "learning_rate": 2.2630560928433268e-05,
2977
+ "loss": 0.4915,
2978
+ "step": 495
2979
+ },
2980
+ {
2981
+ "epoch": 1.44,
2982
+ "learning_rate": 2.251450676982592e-05,
2983
+ "loss": 0.7959,
2984
+ "step": 496
2985
+ },
2986
+ {
2987
+ "epoch": 1.44,
2988
+ "learning_rate": 2.239845261121857e-05,
2989
+ "loss": 1.3574,
2990
+ "step": 497
2991
+ },
2992
+ {
2993
+ "epoch": 1.44,
2994
+ "learning_rate": 2.2282398452611222e-05,
2995
+ "loss": 0.9609,
2996
+ "step": 498
2997
+ },
2998
+ {
2999
+ "epoch": 1.45,
3000
+ "learning_rate": 2.2166344294003867e-05,
3001
+ "loss": 1.0781,
3002
+ "step": 499
3003
+ },
3004
+ {
3005
+ "epoch": 1.45,
3006
+ "learning_rate": 2.2050290135396516e-05,
3007
+ "loss": 0.8276,
3008
+ "step": 500
3009
+ },
3010
+ {
3011
+ "epoch": 1.45,
3012
+ "learning_rate": 2.193423597678917e-05,
3013
+ "loss": 0.6729,
3014
+ "step": 501
3015
+ },
3016
+ {
3017
+ "epoch": 1.46,
3018
+ "learning_rate": 2.1818181818181818e-05,
3019
+ "loss": 0.6982,
3020
+ "step": 502
3021
+ },
3022
+ {
3023
+ "epoch": 1.46,
3024
+ "learning_rate": 2.170212765957447e-05,
3025
+ "loss": 0.2954,
3026
+ "step": 503
3027
+ },
3028
+ {
3029
+ "epoch": 1.46,
3030
+ "learning_rate": 2.1586073500967116e-05,
3031
+ "loss": 0.9126,
3032
+ "step": 504
3033
+ },
3034
+ {
3035
+ "epoch": 1.46,
3036
+ "learning_rate": 2.1470019342359768e-05,
3037
+ "loss": 0.7505,
3038
+ "step": 505
3039
+ },
3040
+ {
3041
+ "epoch": 1.47,
3042
+ "learning_rate": 2.1353965183752417e-05,
3043
+ "loss": 0.792,
3044
+ "step": 506
3045
+ },
3046
+ {
3047
+ "epoch": 1.47,
3048
+ "learning_rate": 2.123791102514507e-05,
3049
+ "loss": 0.406,
3050
+ "step": 507
3051
+ },
3052
+ {
3053
+ "epoch": 1.47,
3054
+ "learning_rate": 2.1121856866537715e-05,
3055
+ "loss": 1.3525,
3056
+ "step": 508
3057
+ },
3058
+ {
3059
+ "epoch": 1.48,
3060
+ "learning_rate": 2.1005802707930368e-05,
3061
+ "loss": 0.4583,
3062
+ "step": 509
3063
+ },
3064
+ {
3065
+ "epoch": 1.48,
3066
+ "learning_rate": 2.0889748549323017e-05,
3067
+ "loss": 0.5293,
3068
+ "step": 510
3069
+ },
3070
+ {
3071
+ "epoch": 1.48,
3072
+ "learning_rate": 2.077369439071567e-05,
3073
+ "loss": 0.7393,
3074
+ "step": 511
3075
+ },
3076
+ {
3077
+ "epoch": 1.48,
3078
+ "learning_rate": 2.0657640232108318e-05,
3079
+ "loss": 0.877,
3080
+ "step": 512
3081
+ },
3082
+ {
3083
+ "epoch": 1.49,
3084
+ "learning_rate": 2.0541586073500967e-05,
3085
+ "loss": 0.4531,
3086
+ "step": 513
3087
+ },
3088
+ {
3089
+ "epoch": 1.49,
3090
+ "learning_rate": 2.0425531914893616e-05,
3091
+ "loss": 1.2695,
3092
+ "step": 514
3093
+ },
3094
+ {
3095
+ "epoch": 1.49,
3096
+ "learning_rate": 2.030947775628627e-05,
3097
+ "loss": 0.8027,
3098
+ "step": 515
3099
+ },
3100
+ {
3101
+ "epoch": 1.5,
3102
+ "learning_rate": 2.0193423597678918e-05,
3103
+ "loss": 0.9111,
3104
+ "step": 516
3105
+ },
3106
+ {
3107
+ "epoch": 1.5,
3108
+ "learning_rate": 2.0077369439071567e-05,
3109
+ "loss": 0.6533,
3110
+ "step": 517
3111
+ },
3112
+ {
3113
+ "epoch": 1.5,
3114
+ "learning_rate": 1.9961315280464216e-05,
3115
+ "loss": 0.4072,
3116
+ "step": 518
3117
+ },
3118
+ {
3119
+ "epoch": 1.5,
3120
+ "learning_rate": 1.9845261121856868e-05,
3121
+ "loss": 0.9634,
3122
+ "step": 519
3123
+ },
3124
+ {
3125
+ "epoch": 1.51,
3126
+ "learning_rate": 1.9729206963249517e-05,
3127
+ "loss": 0.8545,
3128
+ "step": 520
3129
+ },
3130
+ {
3131
+ "epoch": 1.51,
3132
+ "learning_rate": 1.961315280464217e-05,
3133
+ "loss": 0.7363,
3134
+ "step": 521
3135
+ },
3136
+ {
3137
+ "epoch": 1.51,
3138
+ "learning_rate": 1.9497098646034815e-05,
3139
+ "loss": 0.6431,
3140
+ "step": 522
3141
+ },
3142
+ {
3143
+ "epoch": 1.52,
3144
+ "learning_rate": 1.9381044487427468e-05,
3145
+ "loss": 0.5728,
3146
+ "step": 523
3147
+ },
3148
+ {
3149
+ "epoch": 1.52,
3150
+ "learning_rate": 1.9264990328820117e-05,
3151
+ "loss": 0.3672,
3152
+ "step": 524
3153
+ },
3154
+ {
3155
+ "epoch": 1.52,
3156
+ "learning_rate": 1.914893617021277e-05,
3157
+ "loss": 0.9966,
3158
+ "step": 525
3159
+ },
3160
+ {
3161
+ "epoch": 1.52,
3162
+ "learning_rate": 1.9032882011605415e-05,
3163
+ "loss": 0.6738,
3164
+ "step": 526
3165
+ },
3166
+ {
3167
+ "epoch": 1.53,
3168
+ "learning_rate": 1.8916827852998064e-05,
3169
+ "loss": 0.5571,
3170
+ "step": 527
3171
+ },
3172
+ {
3173
+ "epoch": 1.53,
3174
+ "learning_rate": 1.8800773694390716e-05,
3175
+ "loss": 0.5581,
3176
+ "step": 528
3177
+ },
3178
+ {
3179
+ "epoch": 1.53,
3180
+ "learning_rate": 1.8684719535783365e-05,
3181
+ "loss": 0.8774,
3182
+ "step": 529
3183
+ },
3184
+ {
3185
+ "epoch": 1.54,
3186
+ "learning_rate": 1.8568665377176018e-05,
3187
+ "loss": 0.8042,
3188
+ "step": 530
3189
+ },
3190
+ {
3191
+ "epoch": 1.54,
3192
+ "learning_rate": 1.8452611218568663e-05,
3193
+ "loss": 1.0947,
3194
+ "step": 531
3195
+ },
3196
+ {
3197
+ "epoch": 1.54,
3198
+ "learning_rate": 1.8336557059961316e-05,
3199
+ "loss": 0.7212,
3200
+ "step": 532
3201
+ },
3202
+ {
3203
+ "epoch": 1.54,
3204
+ "learning_rate": 1.8220502901353965e-05,
3205
+ "loss": 0.8325,
3206
+ "step": 533
3207
+ },
3208
+ {
3209
+ "epoch": 1.55,
3210
+ "learning_rate": 1.8104448742746617e-05,
3211
+ "loss": 0.4131,
3212
+ "step": 534
3213
+ },
3214
+ {
3215
+ "epoch": 1.55,
3216
+ "learning_rate": 1.7988394584139263e-05,
3217
+ "loss": 0.5898,
3218
+ "step": 535
3219
+ },
3220
+ {
3221
+ "epoch": 1.55,
3222
+ "learning_rate": 1.7872340425531915e-05,
3223
+ "loss": 0.9541,
3224
+ "step": 536
3225
+ },
3226
+ {
3227
+ "epoch": 1.56,
3228
+ "learning_rate": 1.7756286266924564e-05,
3229
+ "loss": 0.3879,
3230
+ "step": 537
3231
+ },
3232
+ {
3233
+ "epoch": 1.56,
3234
+ "learning_rate": 1.7640232108317216e-05,
3235
+ "loss": 0.6372,
3236
+ "step": 538
3237
+ },
3238
+ {
3239
+ "epoch": 1.56,
3240
+ "learning_rate": 1.7524177949709866e-05,
3241
+ "loss": 0.4324,
3242
+ "step": 539
3243
+ },
3244
+ {
3245
+ "epoch": 1.57,
3246
+ "learning_rate": 1.7408123791102515e-05,
3247
+ "loss": 0.6602,
3248
+ "step": 540
3249
+ },
3250
+ {
3251
+ "epoch": 1.57,
3252
+ "learning_rate": 1.7292069632495164e-05,
3253
+ "loss": 0.3093,
3254
+ "step": 541
3255
+ },
3256
+ {
3257
+ "epoch": 1.57,
3258
+ "learning_rate": 1.7176015473887816e-05,
3259
+ "loss": 0.9644,
3260
+ "step": 542
3261
+ },
3262
+ {
3263
+ "epoch": 1.57,
3264
+ "learning_rate": 1.7059961315280465e-05,
3265
+ "loss": 0.5732,
3266
+ "step": 543
3267
+ },
3268
+ {
3269
+ "epoch": 1.58,
3270
+ "learning_rate": 1.6943907156673114e-05,
3271
+ "loss": 0.7998,
3272
+ "step": 544
3273
+ },
3274
+ {
3275
+ "epoch": 1.58,
3276
+ "learning_rate": 1.6827852998065763e-05,
3277
+ "loss": 1.0518,
3278
+ "step": 545
3279
+ },
3280
+ {
3281
+ "epoch": 1.58,
3282
+ "learning_rate": 1.6711798839458415e-05,
3283
+ "loss": 0.8706,
3284
+ "step": 546
3285
+ },
3286
+ {
3287
+ "epoch": 1.59,
3288
+ "learning_rate": 1.6595744680851064e-05,
3289
+ "loss": 0.6118,
3290
+ "step": 547
3291
+ },
3292
+ {
3293
+ "epoch": 1.59,
3294
+ "learning_rate": 1.6479690522243717e-05,
3295
+ "loss": 1.1592,
3296
+ "step": 548
3297
+ },
3298
+ {
3299
+ "epoch": 1.59,
3300
+ "learning_rate": 1.6363636363636363e-05,
3301
+ "loss": 0.8555,
3302
+ "step": 549
3303
+ },
3304
+ {
3305
+ "epoch": 1.59,
3306
+ "learning_rate": 1.6247582205029015e-05,
3307
+ "loss": 0.6455,
3308
+ "step": 550
3309
+ },
3310
+ {
3311
+ "epoch": 1.6,
3312
+ "learning_rate": 1.6131528046421664e-05,
3313
+ "loss": 0.272,
3314
+ "step": 551
3315
+ },
3316
+ {
3317
+ "epoch": 1.6,
3318
+ "learning_rate": 1.6015473887814316e-05,
3319
+ "loss": 0.9355,
3320
+ "step": 552
3321
+ },
3322
+ {
3323
+ "epoch": 1.6,
3324
+ "learning_rate": 1.5899419729206962e-05,
3325
+ "loss": 0.6943,
3326
+ "step": 553
3327
+ },
3328
+ {
3329
+ "epoch": 1.61,
3330
+ "learning_rate": 1.5783365570599614e-05,
3331
+ "loss": 1.0205,
3332
+ "step": 554
3333
+ },
3334
+ {
3335
+ "epoch": 1.61,
3336
+ "learning_rate": 1.5667311411992263e-05,
3337
+ "loss": 0.8364,
3338
+ "step": 555
3339
+ },
3340
+ {
3341
+ "epoch": 1.61,
3342
+ "learning_rate": 1.5551257253384912e-05,
3343
+ "loss": 1.0449,
3344
+ "step": 556
3345
+ },
3346
+ {
3347
+ "epoch": 1.61,
3348
+ "learning_rate": 1.5435203094777565e-05,
3349
+ "loss": 0.4841,
3350
+ "step": 557
3351
+ },
3352
+ {
3353
+ "epoch": 1.62,
3354
+ "learning_rate": 1.531914893617021e-05,
3355
+ "loss": 0.9956,
3356
+ "step": 558
3357
+ },
3358
+ {
3359
+ "epoch": 1.62,
3360
+ "learning_rate": 1.5203094777562863e-05,
3361
+ "loss": 0.6279,
3362
+ "step": 559
3363
+ },
3364
+ {
3365
+ "epoch": 1.62,
3366
+ "learning_rate": 1.5087040618955514e-05,
3367
+ "loss": 0.3101,
3368
+ "step": 560
3369
+ },
3370
+ {
3371
+ "epoch": 1.63,
3372
+ "learning_rate": 1.4970986460348163e-05,
3373
+ "loss": 0.6255,
3374
+ "step": 561
3375
+ },
3376
+ {
3377
+ "epoch": 1.63,
3378
+ "learning_rate": 1.4854932301740813e-05,
3379
+ "loss": 1.3223,
3380
+ "step": 562
3381
+ },
3382
+ {
3383
+ "epoch": 1.63,
3384
+ "learning_rate": 1.4738878143133462e-05,
3385
+ "loss": 0.22,
3386
+ "step": 563
3387
+ },
3388
+ {
3389
+ "epoch": 1.63,
3390
+ "learning_rate": 1.4622823984526113e-05,
3391
+ "loss": 0.4963,
3392
+ "step": 564
3393
+ },
3394
+ {
3395
+ "epoch": 1.64,
3396
+ "learning_rate": 1.4506769825918764e-05,
3397
+ "loss": 1.002,
3398
+ "step": 565
3399
+ },
3400
+ {
3401
+ "epoch": 1.64,
3402
+ "learning_rate": 1.4390715667311413e-05,
3403
+ "loss": 0.9609,
3404
+ "step": 566
3405
+ },
3406
+ {
3407
+ "epoch": 1.64,
3408
+ "learning_rate": 1.4274661508704064e-05,
3409
+ "loss": 0.7744,
3410
+ "step": 567
3411
+ },
3412
+ {
3413
+ "epoch": 1.65,
3414
+ "learning_rate": 1.4158607350096711e-05,
3415
+ "loss": 0.542,
3416
+ "step": 568
3417
+ },
3418
+ {
3419
+ "epoch": 1.65,
3420
+ "learning_rate": 1.4042553191489362e-05,
3421
+ "loss": 0.668,
3422
+ "step": 569
3423
+ },
3424
+ {
3425
+ "epoch": 1.65,
3426
+ "learning_rate": 1.392649903288201e-05,
3427
+ "loss": 0.8008,
3428
+ "step": 570
3429
+ },
3430
+ {
3431
+ "epoch": 1.66,
3432
+ "learning_rate": 1.3810444874274661e-05,
3433
+ "loss": 0.5571,
3434
+ "step": 571
3435
+ },
3436
+ {
3437
+ "epoch": 1.66,
3438
+ "learning_rate": 1.369439071566731e-05,
3439
+ "loss": 0.8682,
3440
+ "step": 572
3441
+ },
3442
+ {
3443
+ "epoch": 1.66,
3444
+ "learning_rate": 1.3578336557059961e-05,
3445
+ "loss": 0.5957,
3446
+ "step": 573
3447
+ },
3448
+ {
3449
+ "epoch": 1.66,
3450
+ "learning_rate": 1.3462282398452612e-05,
3451
+ "loss": 0.5708,
3452
+ "step": 574
3453
+ },
3454
+ {
3455
+ "epoch": 1.67,
3456
+ "learning_rate": 1.334622823984526e-05,
3457
+ "loss": 0.7559,
3458
+ "step": 575
3459
+ },
3460
+ {
3461
+ "epoch": 1.67,
3462
+ "learning_rate": 1.3230174081237912e-05,
3463
+ "loss": 0.4141,
3464
+ "step": 576
3465
+ },
3466
+ {
3467
+ "epoch": 1.67,
3468
+ "learning_rate": 1.311411992263056e-05,
3469
+ "loss": 1.0518,
3470
+ "step": 577
3471
+ },
3472
+ {
3473
+ "epoch": 1.68,
3474
+ "learning_rate": 1.2998065764023211e-05,
3475
+ "loss": 1.3086,
3476
+ "step": 578
3477
+ },
3478
+ {
3479
+ "epoch": 1.68,
3480
+ "learning_rate": 1.288201160541586e-05,
3481
+ "loss": 0.8892,
3482
+ "step": 579
3483
+ },
3484
+ {
3485
+ "epoch": 1.68,
3486
+ "learning_rate": 1.2765957446808511e-05,
3487
+ "loss": 0.447,
3488
+ "step": 580
3489
+ },
3490
+ {
3491
+ "epoch": 1.68,
3492
+ "learning_rate": 1.264990328820116e-05,
3493
+ "loss": 0.6396,
3494
+ "step": 581
3495
+ },
3496
+ {
3497
+ "epoch": 1.69,
3498
+ "learning_rate": 1.253384912959381e-05,
3499
+ "loss": 1.0215,
3500
+ "step": 582
3501
+ },
3502
+ {
3503
+ "epoch": 1.69,
3504
+ "learning_rate": 1.2417794970986461e-05,
3505
+ "loss": 0.5806,
3506
+ "step": 583
3507
+ },
3508
+ {
3509
+ "epoch": 1.69,
3510
+ "learning_rate": 1.230174081237911e-05,
3511
+ "loss": 1.3672,
3512
+ "step": 584
3513
+ },
3514
+ {
3515
+ "epoch": 1.7,
3516
+ "learning_rate": 1.2185686653771761e-05,
3517
+ "loss": 0.5278,
3518
+ "step": 585
3519
+ },
3520
+ {
3521
+ "epoch": 1.7,
3522
+ "learning_rate": 1.206963249516441e-05,
3523
+ "loss": 0.8687,
3524
+ "step": 586
3525
+ },
3526
+ {
3527
+ "epoch": 1.7,
3528
+ "learning_rate": 1.1953578336557061e-05,
3529
+ "loss": 0.8115,
3530
+ "step": 587
3531
+ },
3532
+ {
3533
+ "epoch": 1.7,
3534
+ "learning_rate": 1.183752417794971e-05,
3535
+ "loss": 0.292,
3536
+ "step": 588
3537
+ },
3538
+ {
3539
+ "epoch": 1.71,
3540
+ "learning_rate": 1.172147001934236e-05,
3541
+ "loss": 1.0811,
3542
+ "step": 589
3543
+ },
3544
+ {
3545
+ "epoch": 1.71,
3546
+ "learning_rate": 1.160541586073501e-05,
3547
+ "loss": 0.6865,
3548
+ "step": 590
3549
+ },
3550
+ {
3551
+ "epoch": 1.71,
3552
+ "learning_rate": 1.148936170212766e-05,
3553
+ "loss": 1.0137,
3554
+ "step": 591
3555
+ },
3556
+ {
3557
+ "epoch": 1.72,
3558
+ "learning_rate": 1.1373307543520311e-05,
3559
+ "loss": 1.6211,
3560
+ "step": 592
3561
+ },
3562
+ {
3563
+ "epoch": 1.72,
3564
+ "learning_rate": 1.125725338491296e-05,
3565
+ "loss": 0.8965,
3566
+ "step": 593
3567
+ },
3568
+ {
3569
+ "epoch": 1.72,
3570
+ "learning_rate": 1.1141199226305611e-05,
3571
+ "loss": 0.2993,
3572
+ "step": 594
3573
+ },
3574
+ {
3575
+ "epoch": 1.72,
3576
+ "learning_rate": 1.1025145067698258e-05,
3577
+ "loss": 0.5991,
3578
+ "step": 595
3579
+ },
3580
+ {
3581
+ "epoch": 1.73,
3582
+ "learning_rate": 1.0909090909090909e-05,
3583
+ "loss": 0.6768,
3584
+ "step": 596
3585
+ },
3586
+ {
3587
+ "epoch": 1.73,
3588
+ "learning_rate": 1.0793036750483558e-05,
3589
+ "loss": 0.6045,
3590
+ "step": 597
3591
+ },
3592
+ {
3593
+ "epoch": 1.73,
3594
+ "learning_rate": 1.0676982591876209e-05,
3595
+ "loss": 0.3665,
3596
+ "step": 598
3597
+ },
3598
+ {
3599
+ "epoch": 1.74,
3600
+ "learning_rate": 1.0560928433268858e-05,
3601
+ "loss": 1.4131,
3602
+ "step": 599
3603
+ },
3604
+ {
3605
+ "epoch": 1.74,
3606
+ "learning_rate": 1.0444874274661508e-05,
3607
+ "loss": 1.0,
3608
+ "step": 600
3609
+ },
3610
+ {
3611
+ "epoch": 1.74,
3612
+ "learning_rate": 1.0328820116054159e-05,
3613
+ "loss": 0.6123,
3614
+ "step": 601
3615
+ },
3616
+ {
3617
+ "epoch": 1.74,
3618
+ "learning_rate": 1.0212765957446808e-05,
3619
+ "loss": 0.541,
3620
+ "step": 602
3621
+ },
3622
+ {
3623
+ "epoch": 1.75,
3624
+ "learning_rate": 1.0096711798839459e-05,
3625
+ "loss": 0.7446,
3626
+ "step": 603
3627
+ },
3628
+ {
3629
+ "epoch": 1.75,
3630
+ "learning_rate": 9.980657640232108e-06,
3631
+ "loss": 0.4622,
3632
+ "step": 604
3633
+ },
3634
+ {
3635
+ "epoch": 1.75,
3636
+ "learning_rate": 9.864603481624759e-06,
3637
+ "loss": 0.4734,
3638
+ "step": 605
3639
+ },
3640
+ {
3641
+ "epoch": 1.76,
3642
+ "learning_rate": 9.748549323017408e-06,
3643
+ "loss": 0.5908,
3644
+ "step": 606
3645
+ },
3646
+ {
3647
+ "epoch": 1.76,
3648
+ "learning_rate": 9.632495164410058e-06,
3649
+ "loss": 0.8159,
3650
+ "step": 607
3651
+ },
3652
+ {
3653
+ "epoch": 1.76,
3654
+ "learning_rate": 9.516441005802707e-06,
3655
+ "loss": 0.2949,
3656
+ "step": 608
3657
+ },
3658
+ {
3659
+ "epoch": 1.77,
3660
+ "learning_rate": 9.400386847195358e-06,
3661
+ "loss": 0.4207,
3662
+ "step": 609
3663
+ },
3664
+ {
3665
+ "epoch": 1.77,
3666
+ "learning_rate": 9.284332688588009e-06,
3667
+ "loss": 0.8066,
3668
+ "step": 610
3669
+ },
3670
+ {
3671
+ "epoch": 1.77,
3672
+ "learning_rate": 9.168278529980658e-06,
3673
+ "loss": 0.5264,
3674
+ "step": 611
3675
+ },
3676
+ {
3677
+ "epoch": 1.77,
3678
+ "learning_rate": 9.052224371373309e-06,
3679
+ "loss": 0.2637,
3680
+ "step": 612
3681
+ },
3682
+ {
3683
+ "epoch": 1.78,
3684
+ "learning_rate": 8.936170212765958e-06,
3685
+ "loss": 0.8501,
3686
+ "step": 613
3687
+ },
3688
+ {
3689
+ "epoch": 1.78,
3690
+ "learning_rate": 8.820116054158608e-06,
3691
+ "loss": 0.7407,
3692
+ "step": 614
3693
+ },
3694
+ {
3695
+ "epoch": 1.78,
3696
+ "learning_rate": 8.704061895551257e-06,
3697
+ "loss": 1.3408,
3698
+ "step": 615
3699
+ },
3700
+ {
3701
+ "epoch": 1.79,
3702
+ "learning_rate": 8.588007736943908e-06,
3703
+ "loss": 1.96,
3704
+ "step": 616
3705
+ },
3706
+ {
3707
+ "epoch": 1.79,
3708
+ "learning_rate": 8.471953578336557e-06,
3709
+ "loss": 0.6353,
3710
+ "step": 617
3711
+ },
3712
+ {
3713
+ "epoch": 1.79,
3714
+ "learning_rate": 8.355899419729208e-06,
3715
+ "loss": 1.1045,
3716
+ "step": 618
3717
+ },
3718
+ {
3719
+ "epoch": 1.79,
3720
+ "learning_rate": 8.239845261121858e-06,
3721
+ "loss": 1.373,
3722
+ "step": 619
3723
+ },
3724
+ {
3725
+ "epoch": 1.8,
3726
+ "learning_rate": 8.123791102514507e-06,
3727
+ "loss": 0.9888,
3728
+ "step": 620
3729
+ },
3730
+ {
3731
+ "epoch": 1.8,
3732
+ "learning_rate": 8.007736943907158e-06,
3733
+ "loss": 0.4907,
3734
+ "step": 621
3735
+ },
3736
+ {
3737
+ "epoch": 1.8,
3738
+ "learning_rate": 7.891682785299807e-06,
3739
+ "loss": 0.4163,
3740
+ "step": 622
3741
+ },
3742
+ {
3743
+ "epoch": 1.81,
3744
+ "learning_rate": 7.775628626692456e-06,
3745
+ "loss": 0.7407,
3746
+ "step": 623
3747
+ },
3748
+ {
3749
+ "epoch": 1.81,
3750
+ "learning_rate": 7.659574468085105e-06,
3751
+ "loss": 0.6738,
3752
+ "step": 624
3753
+ },
3754
+ {
3755
+ "epoch": 1.81,
3756
+ "learning_rate": 7.543520309477757e-06,
3757
+ "loss": 0.7217,
3758
+ "step": 625
3759
+ },
3760
+ {
3761
+ "epoch": 1.81,
3762
+ "learning_rate": 7.427466150870407e-06,
3763
+ "loss": 0.3872,
3764
+ "step": 626
3765
+ },
3766
+ {
3767
+ "epoch": 1.82,
3768
+ "learning_rate": 7.3114119922630565e-06,
3769
+ "loss": 0.7197,
3770
+ "step": 627
3771
+ },
3772
+ {
3773
+ "epoch": 1.82,
3774
+ "learning_rate": 7.195357833655706e-06,
3775
+ "loss": 1.125,
3776
+ "step": 628
3777
+ },
3778
+ {
3779
+ "epoch": 1.82,
3780
+ "learning_rate": 7.0793036750483554e-06,
3781
+ "loss": 1.0352,
3782
+ "step": 629
3783
+ },
3784
+ {
3785
+ "epoch": 1.83,
3786
+ "learning_rate": 6.963249516441005e-06,
3787
+ "loss": 0.6802,
3788
+ "step": 630
3789
+ },
3790
+ {
3791
+ "epoch": 1.83,
3792
+ "learning_rate": 6.847195357833655e-06,
3793
+ "loss": 0.6938,
3794
+ "step": 631
3795
+ },
3796
+ {
3797
+ "epoch": 1.83,
3798
+ "learning_rate": 6.731141199226306e-06,
3799
+ "loss": 1.0049,
3800
+ "step": 632
3801
+ },
3802
+ {
3803
+ "epoch": 1.83,
3804
+ "learning_rate": 6.615087040618956e-06,
3805
+ "loss": 0.5635,
3806
+ "step": 633
3807
+ },
3808
+ {
3809
+ "epoch": 1.84,
3810
+ "learning_rate": 6.499032882011606e-06,
3811
+ "loss": 0.7109,
3812
+ "step": 634
3813
+ },
3814
+ {
3815
+ "epoch": 1.84,
3816
+ "learning_rate": 6.3829787234042555e-06,
3817
+ "loss": 1.1309,
3818
+ "step": 635
3819
+ },
3820
+ {
3821
+ "epoch": 1.84,
3822
+ "learning_rate": 6.266924564796905e-06,
3823
+ "loss": 0.5781,
3824
+ "step": 636
3825
+ },
3826
+ {
3827
+ "epoch": 1.85,
3828
+ "learning_rate": 6.150870406189555e-06,
3829
+ "loss": 0.7534,
3830
+ "step": 637
3831
+ },
3832
+ {
3833
+ "epoch": 1.85,
3834
+ "learning_rate": 6.034816247582205e-06,
3835
+ "loss": 0.6694,
3836
+ "step": 638
3837
+ },
3838
+ {
3839
+ "epoch": 1.85,
3840
+ "learning_rate": 5.918762088974855e-06,
3841
+ "loss": 0.9985,
3842
+ "step": 639
3843
+ },
3844
+ {
3845
+ "epoch": 1.86,
3846
+ "learning_rate": 5.802707930367505e-06,
3847
+ "loss": 0.7793,
3848
+ "step": 640
3849
+ },
3850
+ {
3851
+ "epoch": 1.86,
3852
+ "learning_rate": 5.6866537717601556e-06,
3853
+ "loss": 0.5532,
3854
+ "step": 641
3855
+ },
3856
+ {
3857
+ "epoch": 1.86,
3858
+ "learning_rate": 5.5705996131528054e-06,
3859
+ "loss": 0.4526,
3860
+ "step": 642
3861
+ },
3862
+ {
3863
+ "epoch": 1.86,
3864
+ "learning_rate": 5.4545454545454545e-06,
3865
+ "loss": 1.2031,
3866
+ "step": 643
3867
+ },
3868
+ {
3869
+ "epoch": 1.87,
3870
+ "learning_rate": 5.338491295938104e-06,
3871
+ "loss": 0.9697,
3872
+ "step": 644
3873
+ },
3874
+ {
3875
+ "epoch": 1.87,
3876
+ "learning_rate": 5.222437137330754e-06,
3877
+ "loss": 1.1504,
3878
+ "step": 645
3879
+ },
3880
+ {
3881
+ "epoch": 1.87,
3882
+ "learning_rate": 5.106382978723404e-06,
3883
+ "loss": 0.3545,
3884
+ "step": 646
3885
+ },
3886
+ {
3887
+ "epoch": 1.88,
3888
+ "learning_rate": 4.990328820116054e-06,
3889
+ "loss": 0.7798,
3890
+ "step": 647
3891
+ },
3892
+ {
3893
+ "epoch": 1.88,
3894
+ "learning_rate": 4.874274661508704e-06,
3895
+ "loss": 1.0352,
3896
+ "step": 648
3897
+ },
3898
+ {
3899
+ "epoch": 1.88,
3900
+ "learning_rate": 4.758220502901354e-06,
3901
+ "loss": 0.9839,
3902
+ "step": 649
3903
+ },
3904
+ {
3905
+ "epoch": 1.88,
3906
+ "learning_rate": 4.642166344294004e-06,
3907
+ "loss": 1.3174,
3908
+ "step": 650
3909
+ },
3910
+ {
3911
+ "epoch": 1.89,
3912
+ "learning_rate": 4.526112185686654e-06,
3913
+ "loss": 0.7246,
3914
+ "step": 651
3915
+ },
3916
+ {
3917
+ "epoch": 1.89,
3918
+ "learning_rate": 4.410058027079304e-06,
3919
+ "loss": 1.2266,
3920
+ "step": 652
3921
+ },
3922
+ {
3923
+ "epoch": 1.89,
3924
+ "learning_rate": 4.294003868471954e-06,
3925
+ "loss": 0.5737,
3926
+ "step": 653
3927
+ },
3928
+ {
3929
+ "epoch": 1.9,
3930
+ "learning_rate": 4.177949709864604e-06,
3931
+ "loss": 0.6113,
3932
+ "step": 654
3933
+ },
3934
+ {
3935
+ "epoch": 1.9,
3936
+ "learning_rate": 4.061895551257254e-06,
3937
+ "loss": 0.8057,
3938
+ "step": 655
3939
+ },
3940
+ {
3941
+ "epoch": 1.9,
3942
+ "learning_rate": 3.945841392649904e-06,
3943
+ "loss": 0.3179,
3944
+ "step": 656
3945
+ },
3946
+ {
3947
+ "epoch": 1.9,
3948
+ "learning_rate": 3.829787234042553e-06,
3949
+ "loss": 0.3594,
3950
+ "step": 657
3951
+ },
3952
+ {
3953
+ "epoch": 1.91,
3954
+ "learning_rate": 3.7137330754352033e-06,
3955
+ "loss": 0.3843,
3956
+ "step": 658
3957
+ },
3958
+ {
3959
+ "epoch": 1.91,
3960
+ "learning_rate": 3.597678916827853e-06,
3961
+ "loss": 0.7217,
3962
+ "step": 659
3963
+ },
3964
+ {
3965
+ "epoch": 1.91,
3966
+ "learning_rate": 3.4816247582205027e-06,
3967
+ "loss": 0.772,
3968
+ "step": 660
3969
+ },
3970
+ {
3971
+ "epoch": 1.92,
3972
+ "learning_rate": 3.365570599613153e-06,
3973
+ "loss": 0.7642,
3974
+ "step": 661
3975
+ },
3976
+ {
3977
+ "epoch": 1.92,
3978
+ "learning_rate": 3.249516441005803e-06,
3979
+ "loss": 0.4045,
3980
+ "step": 662
3981
+ },
3982
+ {
3983
+ "epoch": 1.92,
3984
+ "learning_rate": 3.1334622823984527e-06,
3985
+ "loss": 0.8843,
3986
+ "step": 663
3987
+ },
3988
+ {
3989
+ "epoch": 1.92,
3990
+ "learning_rate": 3.0174081237911026e-06,
3991
+ "loss": 0.48,
3992
+ "step": 664
3993
+ },
3994
+ {
3995
+ "epoch": 1.93,
3996
+ "learning_rate": 2.9013539651837524e-06,
3997
+ "loss": 0.5723,
3998
+ "step": 665
3999
+ },
4000
+ {
4001
+ "epoch": 1.93,
4002
+ "learning_rate": 2.7852998065764027e-06,
4003
+ "loss": 0.5728,
4004
+ "step": 666
4005
+ },
4006
+ {
4007
+ "epoch": 1.93,
4008
+ "learning_rate": 2.669245647969052e-06,
4009
+ "loss": 0.8955,
4010
+ "step": 667
4011
+ },
4012
+ {
4013
+ "epoch": 1.94,
4014
+ "learning_rate": 2.553191489361702e-06,
4015
+ "loss": 0.5557,
4016
+ "step": 668
4017
+ },
4018
+ {
4019
+ "epoch": 1.94,
4020
+ "learning_rate": 2.437137330754352e-06,
4021
+ "loss": 0.9785,
4022
+ "step": 669
4023
+ },
4024
+ {
4025
+ "epoch": 1.94,
4026
+ "learning_rate": 2.321083172147002e-06,
4027
+ "loss": 0.397,
4028
+ "step": 670
4029
+ },
4030
+ {
4031
+ "epoch": 1.94,
4032
+ "learning_rate": 2.205029013539652e-06,
4033
+ "loss": 0.5137,
4034
+ "step": 671
4035
+ },
4036
+ {
4037
+ "epoch": 1.95,
4038
+ "learning_rate": 2.088974854932302e-06,
4039
+ "loss": 0.4404,
4040
+ "step": 672
4041
+ },
4042
+ {
4043
+ "epoch": 1.95,
4044
+ "learning_rate": 1.972920696324952e-06,
4045
+ "loss": 0.6499,
4046
+ "step": 673
4047
+ },
4048
+ {
4049
+ "epoch": 1.95,
4050
+ "learning_rate": 1.8568665377176017e-06,
4051
+ "loss": 0.4558,
4052
+ "step": 674
4053
+ },
4054
+ {
4055
+ "epoch": 1.96,
4056
+ "learning_rate": 1.7408123791102513e-06,
4057
+ "loss": 1.3662,
4058
+ "step": 675
4059
+ },
4060
+ {
4061
+ "epoch": 1.96,
4062
+ "learning_rate": 1.6247582205029014e-06,
4063
+ "loss": 1.3398,
4064
+ "step": 676
4065
+ },
4066
+ {
4067
+ "epoch": 1.96,
4068
+ "learning_rate": 1.5087040618955513e-06,
4069
+ "loss": 1.0703,
4070
+ "step": 677
4071
+ },
4072
+ {
4073
+ "epoch": 1.97,
4074
+ "learning_rate": 1.3926499032882014e-06,
4075
+ "loss": 0.208,
4076
+ "step": 678
4077
+ },
4078
+ {
4079
+ "epoch": 1.97,
4080
+ "learning_rate": 1.276595744680851e-06,
4081
+ "loss": 0.5503,
4082
+ "step": 679
4083
+ },
4084
+ {
4085
+ "epoch": 1.97,
4086
+ "learning_rate": 1.160541586073501e-06,
4087
+ "loss": 0.5762,
4088
+ "step": 680
4089
+ },
4090
+ {
4091
+ "epoch": 1.97,
4092
+ "learning_rate": 1.044487427466151e-06,
4093
+ "loss": 0.627,
4094
+ "step": 681
4095
+ },
4096
+ {
4097
+ "epoch": 1.98,
4098
+ "learning_rate": 9.284332688588008e-07,
4099
+ "loss": 0.5713,
4100
+ "step": 682
4101
+ },
4102
+ {
4103
+ "epoch": 1.98,
4104
+ "learning_rate": 8.123791102514507e-07,
4105
+ "loss": 0.6621,
4106
+ "step": 683
4107
+ },
4108
+ {
4109
+ "epoch": 1.98,
4110
+ "learning_rate": 6.963249516441007e-07,
4111
+ "loss": 0.6255,
4112
+ "step": 684
4113
+ },
4114
+ {
4115
+ "epoch": 1.99,
4116
+ "learning_rate": 5.802707930367505e-07,
4117
+ "loss": 1.2549,
4118
+ "step": 685
4119
+ },
4120
+ {
4121
+ "epoch": 1.99,
4122
+ "learning_rate": 4.642166344294004e-07,
4123
+ "loss": 1.0869,
4124
+ "step": 686
4125
+ },
4126
+ {
4127
+ "epoch": 1.99,
4128
+ "learning_rate": 3.4816247582205034e-07,
4129
+ "loss": 0.5698,
4130
+ "step": 687
4131
+ },
4132
+ {
4133
+ "epoch": 1.99,
4134
+ "learning_rate": 2.321083172147002e-07,
4135
+ "loss": 0.52,
4136
+ "step": 688
4137
+ },
4138
+ {
4139
+ "epoch": 2.0,
4140
+ "learning_rate": 1.160541586073501e-07,
4141
+ "loss": 0.4097,
4142
+ "step": 689
4143
+ },
4144
+ {
4145
+ "epoch": 2.0,
4146
+ "learning_rate": 0.0,
4147
+ "loss": 0.9941,
4148
+ "step": 690
4149
+ },
4150
+ {
4151
+ "epoch": 2.0,
4152
+ "step": 690,
4153
+ "total_flos": 2.215462037272658e+18,
4154
+ "train_loss": 1.2527947025022645,
4155
+ "train_runtime": 134.5045,
4156
+ "train_samples_per_second": 1316.952,
4157
+ "train_steps_per_second": 5.13
4158
+ }
4159
+ ],
4160
+ "max_steps": 690,
4161
+ "num_train_epochs": 2,
4162
+ "total_flos": 2.215462037272658e+18,
4163
+ "trial_name": null,
4164
+ "trial_params": null
4165
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4d85f45c5246dc6793ad38b497a1c80251b60b76b381fb289b524e9fec8d0f86
3
+ size 2671
vocab.json ADDED
The diff for this file is too large to render. See raw diff