Hugging Face's logo

Language-Media-Lab
/
mt5-small-ain-jpn-mt

Translation
Transformers
PyTorch
Japanese
Ainu (Japan)
mt5
text2text-generation
Inference Endpoints
Model card Files Community
1
koki commited on
Commit
4ff5334
•
1 Parent(s): 6ab278e

First model version

Browse files
Files changed (10) hide show
  1. config.json +28 -0
  2. optimizer.pt +3 -0
  3. pytorch_model.bin +3 -0
  4. rng_state.pth +3 -0
  5. scheduler.pt +3 -0
  6. special_tokens_map.json +1 -0
  7. spiece.model +3 -0
  8. tokenizer_config.json +1 -0
  9. trainer_state.json +4562 -0
  10. training_args.bin +3 -0
config.json ADDED
@@ -0,0 +1,28 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "google/mt5-small",
3
+ "architectures": [
4
+ "MT5ForConditionalGeneration"
5
+ ],
6
+ "d_ff": 1024,
7
+ "d_kv": 64,
8
+ "d_model": 512,
9
+ "decoder_start_token_id": 0,
10
+ "dropout_rate": 0.1,
11
+ "eos_token_id": 1,
12
+ "feed_forward_proj": "gated-gelu",
13
+ "initializer_factor": 1.0,
14
+ "is_encoder_decoder": true,
15
+ "layer_norm_epsilon": 1e-06,
16
+ "model_type": "mt5",
17
+ "num_decoder_layers": 8,
18
+ "num_heads": 6,
19
+ "num_layers": 8,
20
+ "pad_token_id": 0,
21
+ "relative_attention_num_buckets": 32,
22
+ "tie_word_embeddings": false,
23
+ "tokenizer_class": "T5Tokenizer",
24
+ "torch_dtype": "float32",
25
+ "transformers_version": "4.17.0.dev0",
26
+ "use_cache": true,
27
+ "vocab_size": 250100
28
+ }
optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a19295eedeb9af10063ba04c81e5d886eac558c78fcbf1c333c0fcdac9185f91
3
+ size 2401427529
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ff45c05255ec44a25ffcfa01245751e87892ee03f8d7af90a2f693a1a5c117c9
3
+ size 1200743045
rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2fa6d6acba3a387a74e2512985fef7c05af9eabbf101c09cc8394722cbd9e922
3
+ size 14503
scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a63682dbf4106718ce5aacc63e4a8efa7ba2241e4b09ebf11627b9e0191c290e
3
+ size 623
special_tokens_map.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"eos_token": "</s>", "unk_token": "<unk>", "pad_token": "<pad>"}
spiece.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ef78f86560d809067d12bac6c09f19a462cb3af3f54d2b8acbba26e1433125d6
3
+ size 4309802
tokenizer_config.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"eos_token": "</s>", "unk_token": "<unk>", "pad_token": "<pad>", "extra_ids": 0, "additional_special_tokens": null, "sp_model_kwargs": {}, "special_tokens_map_file": "/root/.cache/huggingface/transformers/685ac0ca8568ec593a48b61b0a3c272beee9bc194a3c7241d15dcadb5f875e53.f76030f3ec1b96a8199b2593390c610e76ca8028ef3d24680000619ffb646276", "name_or_path": "google/mt5-small", "tokenizer_class": "T5Tokenizer"}
trainer_state.json ADDED
@@ -0,0 +1,4562 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 5.0,
5
+ "global_step": 375680,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.01,
12
+ "learning_rate": 0.00019998669080068145,
13
+ "loss": 10.2481,
14
+ "step": 500
15
+ },
16
+ {
17
+ "epoch": 0.01,
18
+ "learning_rate": 0.00019997338160136289,
19
+ "loss": 5.6824,
20
+ "step": 1000
21
+ },
22
+ {
23
+ "epoch": 0.02,
24
+ "learning_rate": 0.0001999600724020443,
25
+ "loss": 5.172,
26
+ "step": 1500
27
+ },
28
+ {
29
+ "epoch": 0.03,
30
+ "learning_rate": 0.00019994676320272574,
31
+ "loss": 5.0986,
32
+ "step": 2000
33
+ },
34
+ {
35
+ "epoch": 0.03,
36
+ "learning_rate": 0.00019993345400340717,
37
+ "loss": 4.9605,
38
+ "step": 2500
39
+ },
40
+ {
41
+ "epoch": 0.04,
42
+ "learning_rate": 0.0001999201448040886,
43
+ "loss": 4.7834,
44
+ "step": 3000
45
+ },
46
+ {
47
+ "epoch": 0.05,
48
+ "learning_rate": 0.00019990683560477005,
49
+ "loss": 4.7696,
50
+ "step": 3500
51
+ },
52
+ {
53
+ "epoch": 0.05,
54
+ "learning_rate": 0.00019989352640545146,
55
+ "loss": 4.6645,
56
+ "step": 4000
57
+ },
58
+ {
59
+ "epoch": 0.06,
60
+ "learning_rate": 0.00019988021720613287,
61
+ "loss": 4.5375,
62
+ "step": 4500
63
+ },
64
+ {
65
+ "epoch": 0.07,
66
+ "learning_rate": 0.0001998669080068143,
67
+ "loss": 4.4608,
68
+ "step": 5000
69
+ },
70
+ {
71
+ "epoch": 0.07,
72
+ "learning_rate": 0.00019985359880749575,
73
+ "loss": 4.3304,
74
+ "step": 5500
75
+ },
76
+ {
77
+ "epoch": 0.08,
78
+ "learning_rate": 0.00019984028960817719,
79
+ "loss": 4.2935,
80
+ "step": 6000
81
+ },
82
+ {
83
+ "epoch": 0.09,
84
+ "learning_rate": 0.0001998269804088586,
85
+ "loss": 4.0207,
86
+ "step": 6500
87
+ },
88
+ {
89
+ "epoch": 0.09,
90
+ "learning_rate": 0.00019981367120954004,
91
+ "loss": 4.1862,
92
+ "step": 7000
93
+ },
94
+ {
95
+ "epoch": 0.1,
96
+ "learning_rate": 0.00019980036201022147,
97
+ "loss": 4.1547,
98
+ "step": 7500
99
+ },
100
+ {
101
+ "epoch": 0.11,
102
+ "learning_rate": 0.0001997870528109029,
103
+ "loss": 4.0028,
104
+ "step": 8000
105
+ },
106
+ {
107
+ "epoch": 0.11,
108
+ "learning_rate": 0.00019977374361158435,
109
+ "loss": 3.9894,
110
+ "step": 8500
111
+ },
112
+ {
113
+ "epoch": 0.12,
114
+ "learning_rate": 0.00019976043441226576,
115
+ "loss": 4.0722,
116
+ "step": 9000
117
+ },
118
+ {
119
+ "epoch": 0.13,
120
+ "learning_rate": 0.0001997471252129472,
121
+ "loss": 3.8993,
122
+ "step": 9500
123
+ },
124
+ {
125
+ "epoch": 0.13,
126
+ "learning_rate": 0.00019973381601362864,
127
+ "loss": 3.8414,
128
+ "step": 10000
129
+ },
130
+ {
131
+ "epoch": 0.14,
132
+ "learning_rate": 0.00019972050681431008,
133
+ "loss": 3.8328,
134
+ "step": 10500
135
+ },
136
+ {
137
+ "epoch": 0.15,
138
+ "learning_rate": 0.00019970719761499149,
139
+ "loss": 3.731,
140
+ "step": 11000
141
+ },
142
+ {
143
+ "epoch": 0.15,
144
+ "learning_rate": 0.00019969388841567292,
145
+ "loss": 3.8229,
146
+ "step": 11500
147
+ },
148
+ {
149
+ "epoch": 0.16,
150
+ "learning_rate": 0.00019968057921635434,
151
+ "loss": 3.6892,
152
+ "step": 12000
153
+ },
154
+ {
155
+ "epoch": 0.17,
156
+ "learning_rate": 0.00019966727001703577,
157
+ "loss": 3.5272,
158
+ "step": 12500
159
+ },
160
+ {
161
+ "epoch": 0.17,
162
+ "learning_rate": 0.0001996539608177172,
163
+ "loss": 3.7993,
164
+ "step": 13000
165
+ },
166
+ {
167
+ "epoch": 0.18,
168
+ "learning_rate": 0.00019964065161839865,
169
+ "loss": 3.5962,
170
+ "step": 13500
171
+ },
172
+ {
173
+ "epoch": 0.19,
174
+ "learning_rate": 0.0001996273424190801,
175
+ "loss": 3.6104,
176
+ "step": 14000
177
+ },
178
+ {
179
+ "epoch": 0.19,
180
+ "learning_rate": 0.0001996140332197615,
181
+ "loss": 3.5948,
182
+ "step": 14500
183
+ },
184
+ {
185
+ "epoch": 0.2,
186
+ "learning_rate": 0.00019960072402044294,
187
+ "loss": 3.5836,
188
+ "step": 15000
189
+ },
190
+ {
191
+ "epoch": 0.21,
192
+ "learning_rate": 0.00019958741482112438,
193
+ "loss": 3.5037,
194
+ "step": 15500
195
+ },
196
+ {
197
+ "epoch": 0.21,
198
+ "learning_rate": 0.00019957410562180581,
199
+ "loss": 3.5228,
200
+ "step": 16000
201
+ },
202
+ {
203
+ "epoch": 0.22,
204
+ "learning_rate": 0.00019956079642248725,
205
+ "loss": 3.4802,
206
+ "step": 16500
207
+ },
208
+ {
209
+ "epoch": 0.23,
210
+ "learning_rate": 0.00019954748722316866,
211
+ "loss": 3.3982,
212
+ "step": 17000
213
+ },
214
+ {
215
+ "epoch": 0.23,
216
+ "learning_rate": 0.0001995341780238501,
217
+ "loss": 3.4461,
218
+ "step": 17500
219
+ },
220
+ {
221
+ "epoch": 0.24,
222
+ "learning_rate": 0.0001995208688245315,
223
+ "loss": 3.426,
224
+ "step": 18000
225
+ },
226
+ {
227
+ "epoch": 0.25,
228
+ "learning_rate": 0.00019950755962521295,
229
+ "loss": 3.5081,
230
+ "step": 18500
231
+ },
232
+ {
233
+ "epoch": 0.25,
234
+ "learning_rate": 0.0001994942504258944,
235
+ "loss": 3.3901,
236
+ "step": 19000
237
+ },
238
+ {
239
+ "epoch": 0.26,
240
+ "learning_rate": 0.0001994809412265758,
241
+ "loss": 3.394,
242
+ "step": 19500
243
+ },
244
+ {
245
+ "epoch": 0.27,
246
+ "learning_rate": 0.00019946763202725724,
247
+ "loss": 3.2572,
248
+ "step": 20000
249
+ },
250
+ {
251
+ "epoch": 0.27,
252
+ "learning_rate": 0.00019945432282793868,
253
+ "loss": 3.1521,
254
+ "step": 20500
255
+ },
256
+ {
257
+ "epoch": 0.28,
258
+ "learning_rate": 0.00019944101362862011,
259
+ "loss": 3.3596,
260
+ "step": 21000
261
+ },
262
+ {
263
+ "epoch": 0.29,
264
+ "learning_rate": 0.00019942770442930155,
265
+ "loss": 3.2875,
266
+ "step": 21500
267
+ },
268
+ {
269
+ "epoch": 0.29,
270
+ "learning_rate": 0.00019941439522998296,
271
+ "loss": 3.2877,
272
+ "step": 22000
273
+ },
274
+ {
275
+ "epoch": 0.3,
276
+ "learning_rate": 0.0001994010860306644,
277
+ "loss": 3.2511,
278
+ "step": 22500
279
+ },
280
+ {
281
+ "epoch": 0.31,
282
+ "learning_rate": 0.00019938777683134584,
283
+ "loss": 3.2489,
284
+ "step": 23000
285
+ },
286
+ {
287
+ "epoch": 0.31,
288
+ "learning_rate": 0.00019937446763202728,
289
+ "loss": 3.189,
290
+ "step": 23500
291
+ },
292
+ {
293
+ "epoch": 0.32,
294
+ "learning_rate": 0.00019936115843270872,
295
+ "loss": 3.1982,
296
+ "step": 24000
297
+ },
298
+ {
299
+ "epoch": 0.33,
300
+ "learning_rate": 0.00019934784923339013,
301
+ "loss": 3.2822,
302
+ "step": 24500
303
+ },
304
+ {
305
+ "epoch": 0.33,
306
+ "learning_rate": 0.00019933454003407156,
307
+ "loss": 3.0918,
308
+ "step": 25000
309
+ },
310
+ {
311
+ "epoch": 0.34,
312
+ "learning_rate": 0.00019932123083475298,
313
+ "loss": 3.1034,
314
+ "step": 25500
315
+ },
316
+ {
317
+ "epoch": 0.35,
318
+ "learning_rate": 0.00019930792163543441,
319
+ "loss": 3.1271,
320
+ "step": 26000
321
+ },
322
+ {
323
+ "epoch": 0.35,
324
+ "learning_rate": 0.00019929461243611585,
325
+ "loss": 3.0492,
326
+ "step": 26500
327
+ },
328
+ {
329
+ "epoch": 0.36,
330
+ "learning_rate": 0.0001992813032367973,
331
+ "loss": 3.1039,
332
+ "step": 27000
333
+ },
334
+ {
335
+ "epoch": 0.37,
336
+ "learning_rate": 0.0001992679940374787,
337
+ "loss": 3.1402,
338
+ "step": 27500
339
+ },
340
+ {
341
+ "epoch": 0.37,
342
+ "learning_rate": 0.00019925468483816014,
343
+ "loss": 3.1043,
344
+ "step": 28000
345
+ },
346
+ {
347
+ "epoch": 0.38,
348
+ "learning_rate": 0.00019924137563884158,
349
+ "loss": 3.1407,
350
+ "step": 28500
351
+ },
352
+ {
353
+ "epoch": 0.39,
354
+ "learning_rate": 0.00019922806643952302,
355
+ "loss": 3.0666,
356
+ "step": 29000
357
+ },
358
+ {
359
+ "epoch": 0.39,
360
+ "learning_rate": 0.00019921475724020445,
361
+ "loss": 3.0374,
362
+ "step": 29500
363
+ },
364
+ {
365
+ "epoch": 0.4,
366
+ "learning_rate": 0.00019920144804088586,
367
+ "loss": 3.0348,
368
+ "step": 30000
369
+ },
370
+ {
371
+ "epoch": 0.41,
372
+ "learning_rate": 0.0001991881388415673,
373
+ "loss": 3.1047,
374
+ "step": 30500
375
+ },
376
+ {
377
+ "epoch": 0.41,
378
+ "learning_rate": 0.00019917482964224874,
379
+ "loss": 3.0446,
380
+ "step": 31000
381
+ },
382
+ {
383
+ "epoch": 0.42,
384
+ "learning_rate": 0.00019916152044293018,
385
+ "loss": 3.0877,
386
+ "step": 31500
387
+ },
388
+ {
389
+ "epoch": 0.43,
390
+ "learning_rate": 0.00019914821124361162,
391
+ "loss": 2.926,
392
+ "step": 32000
393
+ },
394
+ {
395
+ "epoch": 0.43,
396
+ "learning_rate": 0.000199134902044293,
397
+ "loss": 3.0038,
398
+ "step": 32500
399
+ },
400
+ {
401
+ "epoch": 0.44,
402
+ "learning_rate": 0.00019912159284497444,
403
+ "loss": 3.0531,
404
+ "step": 33000
405
+ },
406
+ {
407
+ "epoch": 0.45,
408
+ "learning_rate": 0.00019910828364565588,
409
+ "loss": 3.0204,
410
+ "step": 33500
411
+ },
412
+ {
413
+ "epoch": 0.45,
414
+ "learning_rate": 0.00019909497444633732,
415
+ "loss": 3.0724,
416
+ "step": 34000
417
+ },
418
+ {
419
+ "epoch": 0.46,
420
+ "learning_rate": 0.00019908166524701875,
421
+ "loss": 3.1021,
422
+ "step": 34500
423
+ },
424
+ {
425
+ "epoch": 0.47,
426
+ "learning_rate": 0.00019906835604770017,
427
+ "loss": 3.0148,
428
+ "step": 35000
429
+ },
430
+ {
431
+ "epoch": 0.47,
432
+ "learning_rate": 0.0001990550468483816,
433
+ "loss": 2.9517,
434
+ "step": 35500
435
+ },
436
+ {
437
+ "epoch": 0.48,
438
+ "learning_rate": 0.00019904173764906304,
439
+ "loss": 2.9411,
440
+ "step": 36000
441
+ },
442
+ {
443
+ "epoch": 0.49,
444
+ "learning_rate": 0.00019902842844974448,
445
+ "loss": 2.9533,
446
+ "step": 36500
447
+ },
448
+ {
449
+ "epoch": 0.49,
450
+ "learning_rate": 0.00019901511925042592,
451
+ "loss": 2.9113,
452
+ "step": 37000
453
+ },
454
+ {
455
+ "epoch": 0.5,
456
+ "learning_rate": 0.00019900181005110733,
457
+ "loss": 2.9085,
458
+ "step": 37500
459
+ },
460
+ {
461
+ "epoch": 0.51,
462
+ "learning_rate": 0.00019898850085178877,
463
+ "loss": 2.9903,
464
+ "step": 38000
465
+ },
466
+ {
467
+ "epoch": 0.51,
468
+ "learning_rate": 0.0001989751916524702,
469
+ "loss": 2.8475,
470
+ "step": 38500
471
+ },
472
+ {
473
+ "epoch": 0.52,
474
+ "learning_rate": 0.00019896188245315164,
475
+ "loss": 2.9614,
476
+ "step": 39000
477
+ },
478
+ {
479
+ "epoch": 0.53,
480
+ "learning_rate": 0.00019894857325383305,
481
+ "loss": 2.9126,
482
+ "step": 39500
483
+ },
484
+ {
485
+ "epoch": 0.53,
486
+ "learning_rate": 0.0001989352640545145,
487
+ "loss": 2.919,
488
+ "step": 40000
489
+ },
490
+ {
491
+ "epoch": 0.54,
492
+ "learning_rate": 0.0001989219548551959,
493
+ "loss": 2.911,
494
+ "step": 40500
495
+ },
496
+ {
497
+ "epoch": 0.55,
498
+ "learning_rate": 0.00019890864565587734,
499
+ "loss": 2.8617,
500
+ "step": 41000
501
+ },
502
+ {
503
+ "epoch": 0.55,
504
+ "learning_rate": 0.00019889533645655878,
505
+ "loss": 2.8817,
506
+ "step": 41500
507
+ },
508
+ {
509
+ "epoch": 0.56,
510
+ "learning_rate": 0.00019888202725724022,
511
+ "loss": 2.9325,
512
+ "step": 42000
513
+ },
514
+ {
515
+ "epoch": 0.57,
516
+ "learning_rate": 0.00019886871805792166,
517
+ "loss": 2.8276,
518
+ "step": 42500
519
+ },
520
+ {
521
+ "epoch": 0.57,
522
+ "learning_rate": 0.00019885540885860307,
523
+ "loss": 2.9285,
524
+ "step": 43000
525
+ },
526
+ {
527
+ "epoch": 0.58,
528
+ "learning_rate": 0.0001988420996592845,
529
+ "loss": 2.7656,
530
+ "step": 43500
531
+ },
532
+ {
533
+ "epoch": 0.59,
534
+ "learning_rate": 0.00019882879045996594,
535
+ "loss": 2.9298,
536
+ "step": 44000
537
+ },
538
+ {
539
+ "epoch": 0.59,
540
+ "learning_rate": 0.00019881548126064738,
541
+ "loss": 2.8953,
542
+ "step": 44500
543
+ },
544
+ {
545
+ "epoch": 0.6,
546
+ "learning_rate": 0.00019880217206132882,
547
+ "loss": 2.8127,
548
+ "step": 45000
549
+ },
550
+ {
551
+ "epoch": 0.61,
552
+ "learning_rate": 0.00019878886286201023,
553
+ "loss": 2.8404,
554
+ "step": 45500
555
+ },
556
+ {
557
+ "epoch": 0.61,
558
+ "learning_rate": 0.00019877555366269167,
559
+ "loss": 2.8233,
560
+ "step": 46000
561
+ },
562
+ {
563
+ "epoch": 0.62,
564
+ "learning_rate": 0.0001987622444633731,
565
+ "loss": 2.9052,
566
+ "step": 46500
567
+ },
568
+ {
569
+ "epoch": 0.63,
570
+ "learning_rate": 0.00019874893526405452,
571
+ "loss": 2.8838,
572
+ "step": 47000
573
+ },
574
+ {
575
+ "epoch": 0.63,
576
+ "learning_rate": 0.00019873562606473596,
577
+ "loss": 2.8243,
578
+ "step": 47500
579
+ },
580
+ {
581
+ "epoch": 0.64,
582
+ "learning_rate": 0.00019872231686541737,
583
+ "loss": 2.7138,
584
+ "step": 48000
585
+ },
586
+ {
587
+ "epoch": 0.65,
588
+ "learning_rate": 0.0001987090076660988,
589
+ "loss": 2.7165,
590
+ "step": 48500
591
+ },
592
+ {
593
+ "epoch": 0.65,
594
+ "learning_rate": 0.00019869569846678024,
595
+ "loss": 2.8223,
596
+ "step": 49000
597
+ },
598
+ {
599
+ "epoch": 0.66,
600
+ "learning_rate": 0.00019868238926746168,
601
+ "loss": 2.8202,
602
+ "step": 49500
603
+ },
604
+ {
605
+ "epoch": 0.67,
606
+ "learning_rate": 0.00019866908006814312,
607
+ "loss": 2.8057,
608
+ "step": 50000
609
+ },
610
+ {
611
+ "epoch": 0.67,
612
+ "learning_rate": 0.00019865577086882453,
613
+ "loss": 2.774,
614
+ "step": 50500
615
+ },
616
+ {
617
+ "epoch": 0.68,
618
+ "learning_rate": 0.00019864246166950597,
619
+ "loss": 2.8091,
620
+ "step": 51000
621
+ },
622
+ {
623
+ "epoch": 0.69,
624
+ "learning_rate": 0.0001986291524701874,
625
+ "loss": 2.785,
626
+ "step": 51500
627
+ },
628
+ {
629
+ "epoch": 0.69,
630
+ "learning_rate": 0.00019861584327086885,
631
+ "loss": 2.6157,
632
+ "step": 52000
633
+ },
634
+ {
635
+ "epoch": 0.7,
636
+ "learning_rate": 0.00019860253407155028,
637
+ "loss": 2.8632,
638
+ "step": 52500
639
+ },
640
+ {
641
+ "epoch": 0.71,
642
+ "learning_rate": 0.0001985892248722317,
643
+ "loss": 2.7628,
644
+ "step": 53000
645
+ },
646
+ {
647
+ "epoch": 0.71,
648
+ "learning_rate": 0.00019857591567291313,
649
+ "loss": 2.8141,
650
+ "step": 53500
651
+ },
652
+ {
653
+ "epoch": 0.72,
654
+ "learning_rate": 0.00019856260647359454,
655
+ "loss": 2.7369,
656
+ "step": 54000
657
+ },
658
+ {
659
+ "epoch": 0.73,
660
+ "learning_rate": 0.00019854929727427598,
661
+ "loss": 2.721,
662
+ "step": 54500
663
+ },
664
+ {
665
+ "epoch": 0.73,
666
+ "learning_rate": 0.00019853598807495742,
667
+ "loss": 2.7448,
668
+ "step": 55000
669
+ },
670
+ {
671
+ "epoch": 0.74,
672
+ "learning_rate": 0.00019852267887563886,
673
+ "loss": 2.6667,
674
+ "step": 55500
675
+ },
676
+ {
677
+ "epoch": 0.75,
678
+ "learning_rate": 0.00019850936967632027,
679
+ "loss": 2.7889,
680
+ "step": 56000
681
+ },
682
+ {
683
+ "epoch": 0.75,
684
+ "learning_rate": 0.0001984960604770017,
685
+ "loss": 2.823,
686
+ "step": 56500
687
+ },
688
+ {
689
+ "epoch": 0.76,
690
+ "learning_rate": 0.00019848275127768315,
691
+ "loss": 2.7421,
692
+ "step": 57000
693
+ },
694
+ {
695
+ "epoch": 0.77,
696
+ "learning_rate": 0.00019846944207836458,
697
+ "loss": 2.8042,
698
+ "step": 57500
699
+ },
700
+ {
701
+ "epoch": 0.77,
702
+ "learning_rate": 0.00019845613287904602,
703
+ "loss": 2.69,
704
+ "step": 58000
705
+ },
706
+ {
707
+ "epoch": 0.78,
708
+ "learning_rate": 0.00019844282367972743,
709
+ "loss": 2.7895,
710
+ "step": 58500
711
+ },
712
+ {
713
+ "epoch": 0.79,
714
+ "learning_rate": 0.00019842951448040887,
715
+ "loss": 2.7649,
716
+ "step": 59000
717
+ },
718
+ {
719
+ "epoch": 0.79,
720
+ "learning_rate": 0.0001984162052810903,
721
+ "loss": 2.7463,
722
+ "step": 59500
723
+ },
724
+ {
725
+ "epoch": 0.8,
726
+ "learning_rate": 0.00019840289608177175,
727
+ "loss": 2.7047,
728
+ "step": 60000
729
+ },
730
+ {
731
+ "epoch": 0.81,
732
+ "learning_rate": 0.00019838958688245316,
733
+ "loss": 2.73,
734
+ "step": 60500
735
+ },
736
+ {
737
+ "epoch": 0.81,
738
+ "learning_rate": 0.00019837627768313457,
739
+ "loss": 2.7422,
740
+ "step": 61000
741
+ },
742
+ {
743
+ "epoch": 0.82,
744
+ "learning_rate": 0.000198362968483816,
745
+ "loss": 2.7778,
746
+ "step": 61500
747
+ },
748
+ {
749
+ "epoch": 0.83,
750
+ "learning_rate": 0.00019834965928449745,
751
+ "loss": 2.5151,
752
+ "step": 62000
753
+ },
754
+ {
755
+ "epoch": 0.83,
756
+ "learning_rate": 0.00019833635008517888,
757
+ "loss": 2.7631,
758
+ "step": 62500
759
+ },
760
+ {
761
+ "epoch": 0.84,
762
+ "learning_rate": 0.00019832304088586032,
763
+ "loss": 2.7046,
764
+ "step": 63000
765
+ },
766
+ {
767
+ "epoch": 0.85,
768
+ "learning_rate": 0.00019830973168654173,
769
+ "loss": 2.6186,
770
+ "step": 63500
771
+ },
772
+ {
773
+ "epoch": 0.85,
774
+ "learning_rate": 0.00019829642248722317,
775
+ "loss": 2.6684,
776
+ "step": 64000
777
+ },
778
+ {
779
+ "epoch": 0.86,
780
+ "learning_rate": 0.0001982831132879046,
781
+ "loss": 2.687,
782
+ "step": 64500
783
+ },
784
+ {
785
+ "epoch": 0.87,
786
+ "learning_rate": 0.00019826980408858605,
787
+ "loss": 2.6079,
788
+ "step": 65000
789
+ },
790
+ {
791
+ "epoch": 0.87,
792
+ "learning_rate": 0.00019825649488926749,
793
+ "loss": 2.6386,
794
+ "step": 65500
795
+ },
796
+ {
797
+ "epoch": 0.88,
798
+ "learning_rate": 0.0001982431856899489,
799
+ "loss": 2.5986,
800
+ "step": 66000
801
+ },
802
+ {
803
+ "epoch": 0.89,
804
+ "learning_rate": 0.00019822987649063033,
805
+ "loss": 2.639,
806
+ "step": 66500
807
+ },
808
+ {
809
+ "epoch": 0.89,
810
+ "learning_rate": 0.00019821656729131177,
811
+ "loss": 2.6357,
812
+ "step": 67000
813
+ },
814
+ {
815
+ "epoch": 0.9,
816
+ "learning_rate": 0.0001982032580919932,
817
+ "loss": 2.6278,
818
+ "step": 67500
819
+ },
820
+ {
821
+ "epoch": 0.91,
822
+ "learning_rate": 0.00019818994889267465,
823
+ "loss": 2.633,
824
+ "step": 68000
825
+ },
826
+ {
827
+ "epoch": 0.91,
828
+ "learning_rate": 0.00019817663969335606,
829
+ "loss": 2.7147,
830
+ "step": 68500
831
+ },
832
+ {
833
+ "epoch": 0.92,
834
+ "learning_rate": 0.00019816333049403747,
835
+ "loss": 2.6298,
836
+ "step": 69000
837
+ },
838
+ {
839
+ "epoch": 0.92,
840
+ "learning_rate": 0.0001981500212947189,
841
+ "loss": 2.6853,
842
+ "step": 69500
843
+ },
844
+ {
845
+ "epoch": 0.93,
846
+ "learning_rate": 0.00019813671209540035,
847
+ "loss": 2.5161,
848
+ "step": 70000
849
+ },
850
+ {
851
+ "epoch": 0.94,
852
+ "learning_rate": 0.00019812340289608179,
853
+ "loss": 2.6841,
854
+ "step": 70500
855
+ },
856
+ {
857
+ "epoch": 0.94,
858
+ "learning_rate": 0.0001981100936967632,
859
+ "loss": 2.7514,
860
+ "step": 71000
861
+ },
862
+ {
863
+ "epoch": 0.95,
864
+ "learning_rate": 0.00019809678449744464,
865
+ "loss": 2.6619,
866
+ "step": 71500
867
+ },
868
+ {
869
+ "epoch": 0.96,
870
+ "learning_rate": 0.00019808347529812607,
871
+ "loss": 2.6089,
872
+ "step": 72000
873
+ },
874
+ {
875
+ "epoch": 0.96,
876
+ "learning_rate": 0.0001980701660988075,
877
+ "loss": 2.5672,
878
+ "step": 72500
879
+ },
880
+ {
881
+ "epoch": 0.97,
882
+ "learning_rate": 0.00019805685689948895,
883
+ "loss": 2.6084,
884
+ "step": 73000
885
+ },
886
+ {
887
+ "epoch": 0.98,
888
+ "learning_rate": 0.00019804354770017036,
889
+ "loss": 2.6425,
890
+ "step": 73500
891
+ },
892
+ {
893
+ "epoch": 0.98,
894
+ "learning_rate": 0.0001980302385008518,
895
+ "loss": 2.5539,
896
+ "step": 74000
897
+ },
898
+ {
899
+ "epoch": 0.99,
900
+ "learning_rate": 0.00019801692930153324,
901
+ "loss": 2.588,
902
+ "step": 74500
903
+ },
904
+ {
905
+ "epoch": 1.0,
906
+ "learning_rate": 0.00019800362010221468,
907
+ "loss": 2.6845,
908
+ "step": 75000
909
+ },
910
+ {
911
+ "epoch": 1.0,
912
+ "eval_loss": 2.289619207382202,
913
+ "eval_runtime": 153.5049,
914
+ "eval_samples_per_second": 61.19,
915
+ "eval_steps_per_second": 61.19,
916
+ "step": 75136
917
+ },
918
+ {
919
+ "epoch": 1.0,
920
+ "learning_rate": 0.00019799031090289609,
921
+ "loss": 2.4516,
922
+ "step": 75500
923
+ },
924
+ {
925
+ "epoch": 1.01,
926
+ "learning_rate": 0.00019797700170357752,
927
+ "loss": 2.3205,
928
+ "step": 76000
929
+ },
930
+ {
931
+ "epoch": 1.02,
932
+ "learning_rate": 0.00019796369250425894,
933
+ "loss": 2.3807,
934
+ "step": 76500
935
+ },
936
+ {
937
+ "epoch": 1.02,
938
+ "learning_rate": 0.00019795038330494037,
939
+ "loss": 2.4826,
940
+ "step": 77000
941
+ },
942
+ {
943
+ "epoch": 1.03,
944
+ "learning_rate": 0.0001979370741056218,
945
+ "loss": 2.4244,
946
+ "step": 77500
947
+ },
948
+ {
949
+ "epoch": 1.04,
950
+ "learning_rate": 0.00019792376490630325,
951
+ "loss": 2.4297,
952
+ "step": 78000
953
+ },
954
+ {
955
+ "epoch": 1.04,
956
+ "learning_rate": 0.0001979104557069847,
957
+ "loss": 2.5217,
958
+ "step": 78500
959
+ },
960
+ {
961
+ "epoch": 1.05,
962
+ "learning_rate": 0.0001978971465076661,
963
+ "loss": 2.4128,
964
+ "step": 79000
965
+ },
966
+ {
967
+ "epoch": 1.06,
968
+ "learning_rate": 0.00019788383730834754,
969
+ "loss": 2.4708,
970
+ "step": 79500
971
+ },
972
+ {
973
+ "epoch": 1.06,
974
+ "learning_rate": 0.00019787052810902898,
975
+ "loss": 2.3967,
976
+ "step": 80000
977
+ },
978
+ {
979
+ "epoch": 1.07,
980
+ "learning_rate": 0.0001978572189097104,
981
+ "loss": 2.5101,
982
+ "step": 80500
983
+ },
984
+ {
985
+ "epoch": 1.08,
986
+ "learning_rate": 0.00019784390971039185,
987
+ "loss": 2.4615,
988
+ "step": 81000
989
+ },
990
+ {
991
+ "epoch": 1.08,
992
+ "learning_rate": 0.00019783060051107326,
993
+ "loss": 2.4171,
994
+ "step": 81500
995
+ },
996
+ {
997
+ "epoch": 1.09,
998
+ "learning_rate": 0.0001978172913117547,
999
+ "loss": 2.4693,
1000
+ "step": 82000
1001
+ },
1002
+ {
1003
+ "epoch": 1.1,
1004
+ "learning_rate": 0.0001978039821124361,
1005
+ "loss": 2.4741,
1006
+ "step": 82500
1007
+ },
1008
+ {
1009
+ "epoch": 1.1,
1010
+ "learning_rate": 0.00019779067291311755,
1011
+ "loss": 2.3287,
1012
+ "step": 83000
1013
+ },
1014
+ {
1015
+ "epoch": 1.11,
1016
+ "learning_rate": 0.000197777363713799,
1017
+ "loss": 2.4375,
1018
+ "step": 83500
1019
+ },
1020
+ {
1021
+ "epoch": 1.12,
1022
+ "learning_rate": 0.0001977640545144804,
1023
+ "loss": 2.4086,
1024
+ "step": 84000
1025
+ },
1026
+ {
1027
+ "epoch": 1.12,
1028
+ "learning_rate": 0.00019775074531516184,
1029
+ "loss": 2.499,
1030
+ "step": 84500
1031
+ },
1032
+ {
1033
+ "epoch": 1.13,
1034
+ "learning_rate": 0.00019773743611584328,
1035
+ "loss": 2.5159,
1036
+ "step": 85000
1037
+ },
1038
+ {
1039
+ "epoch": 1.14,
1040
+ "learning_rate": 0.00019772412691652471,
1041
+ "loss": 2.4019,
1042
+ "step": 85500
1043
+ },
1044
+ {
1045
+ "epoch": 1.14,
1046
+ "learning_rate": 0.00019771081771720615,
1047
+ "loss": 2.399,
1048
+ "step": 86000
1049
+ },
1050
+ {
1051
+ "epoch": 1.15,
1052
+ "learning_rate": 0.00019769750851788756,
1053
+ "loss": 2.5017,
1054
+ "step": 86500
1055
+ },
1056
+ {
1057
+ "epoch": 1.16,
1058
+ "learning_rate": 0.000197684199318569,
1059
+ "loss": 2.4847,
1060
+ "step": 87000
1061
+ },
1062
+ {
1063
+ "epoch": 1.16,
1064
+ "learning_rate": 0.00019767089011925044,
1065
+ "loss": 2.4126,
1066
+ "step": 87500
1067
+ },
1068
+ {
1069
+ "epoch": 1.17,
1070
+ "learning_rate": 0.00019765758091993188,
1071
+ "loss": 2.4905,
1072
+ "step": 88000
1073
+ },
1074
+ {
1075
+ "epoch": 1.18,
1076
+ "learning_rate": 0.00019764427172061332,
1077
+ "loss": 2.3939,
1078
+ "step": 88500
1079
+ },
1080
+ {
1081
+ "epoch": 1.18,
1082
+ "learning_rate": 0.00019763096252129473,
1083
+ "loss": 2.483,
1084
+ "step": 89000
1085
+ },
1086
+ {
1087
+ "epoch": 1.19,
1088
+ "learning_rate": 0.00019761765332197614,
1089
+ "loss": 2.3784,
1090
+ "step": 89500
1091
+ },
1092
+ {
1093
+ "epoch": 1.2,
1094
+ "learning_rate": 0.00019760434412265758,
1095
+ "loss": 2.3567,
1096
+ "step": 90000
1097
+ },
1098
+ {
1099
+ "epoch": 1.2,
1100
+ "learning_rate": 0.00019759103492333901,
1101
+ "loss": 2.4193,
1102
+ "step": 90500
1103
+ },
1104
+ {
1105
+ "epoch": 1.21,
1106
+ "learning_rate": 0.00019757772572402045,
1107
+ "loss": 2.4165,
1108
+ "step": 91000
1109
+ },
1110
+ {
1111
+ "epoch": 1.22,
1112
+ "learning_rate": 0.0001975644165247019,
1113
+ "loss": 2.4415,
1114
+ "step": 91500
1115
+ },
1116
+ {
1117
+ "epoch": 1.22,
1118
+ "learning_rate": 0.0001975511073253833,
1119
+ "loss": 2.417,
1120
+ "step": 92000
1121
+ },
1122
+ {
1123
+ "epoch": 1.23,
1124
+ "learning_rate": 0.00019753779812606474,
1125
+ "loss": 2.3009,
1126
+ "step": 92500
1127
+ },
1128
+ {
1129
+ "epoch": 1.24,
1130
+ "learning_rate": 0.00019752448892674618,
1131
+ "loss": 2.4853,
1132
+ "step": 93000
1133
+ },
1134
+ {
1135
+ "epoch": 1.24,
1136
+ "learning_rate": 0.00019751117972742762,
1137
+ "loss": 2.3344,
1138
+ "step": 93500
1139
+ },
1140
+ {
1141
+ "epoch": 1.25,
1142
+ "learning_rate": 0.00019749787052810905,
1143
+ "loss": 2.4312,
1144
+ "step": 94000
1145
+ },
1146
+ {
1147
+ "epoch": 1.26,
1148
+ "learning_rate": 0.00019748456132879046,
1149
+ "loss": 2.4414,
1150
+ "step": 94500
1151
+ },
1152
+ {
1153
+ "epoch": 1.26,
1154
+ "learning_rate": 0.0001974712521294719,
1155
+ "loss": 2.4249,
1156
+ "step": 95000
1157
+ },
1158
+ {
1159
+ "epoch": 1.27,
1160
+ "learning_rate": 0.00019745794293015334,
1161
+ "loss": 2.356,
1162
+ "step": 95500
1163
+ },
1164
+ {
1165
+ "epoch": 1.28,
1166
+ "learning_rate": 0.00019744463373083478,
1167
+ "loss": 2.3665,
1168
+ "step": 96000
1169
+ },
1170
+ {
1171
+ "epoch": 1.28,
1172
+ "learning_rate": 0.00019743132453151622,
1173
+ "loss": 2.4506,
1174
+ "step": 96500
1175
+ },
1176
+ {
1177
+ "epoch": 1.29,
1178
+ "learning_rate": 0.0001974180153321976,
1179
+ "loss": 2.413,
1180
+ "step": 97000
1181
+ },
1182
+ {
1183
+ "epoch": 1.3,
1184
+ "learning_rate": 0.00019740470613287904,
1185
+ "loss": 2.4709,
1186
+ "step": 97500
1187
+ },
1188
+ {
1189
+ "epoch": 1.3,
1190
+ "learning_rate": 0.00019739139693356048,
1191
+ "loss": 2.3756,
1192
+ "step": 98000
1193
+ },
1194
+ {
1195
+ "epoch": 1.31,
1196
+ "learning_rate": 0.00019737808773424192,
1197
+ "loss": 2.3728,
1198
+ "step": 98500
1199
+ },
1200
+ {
1201
+ "epoch": 1.32,
1202
+ "learning_rate": 0.00019736477853492335,
1203
+ "loss": 2.4195,
1204
+ "step": 99000
1205
+ },
1206
+ {
1207
+ "epoch": 1.32,
1208
+ "learning_rate": 0.00019735146933560476,
1209
+ "loss": 2.454,
1210
+ "step": 99500
1211
+ },
1212
+ {
1213
+ "epoch": 1.33,
1214
+ "learning_rate": 0.0001973381601362862,
1215
+ "loss": 2.3611,
1216
+ "step": 100000
1217
+ },
1218
+ {
1219
+ "epoch": 1.34,
1220
+ "learning_rate": 0.00019732485093696764,
1221
+ "loss": 2.4303,
1222
+ "step": 100500
1223
+ },
1224
+ {
1225
+ "epoch": 1.34,
1226
+ "learning_rate": 0.00019731154173764908,
1227
+ "loss": 2.4555,
1228
+ "step": 101000
1229
+ },
1230
+ {
1231
+ "epoch": 1.35,
1232
+ "learning_rate": 0.00019729823253833052,
1233
+ "loss": 2.3825,
1234
+ "step": 101500
1235
+ },
1236
+ {
1237
+ "epoch": 1.36,
1238
+ "learning_rate": 0.00019728492333901193,
1239
+ "loss": 2.3164,
1240
+ "step": 102000
1241
+ },
1242
+ {
1243
+ "epoch": 1.36,
1244
+ "learning_rate": 0.00019727161413969337,
1245
+ "loss": 2.4844,
1246
+ "step": 102500
1247
+ },
1248
+ {
1249
+ "epoch": 1.37,
1250
+ "learning_rate": 0.0001972583049403748,
1251
+ "loss": 2.4749,
1252
+ "step": 103000
1253
+ },
1254
+ {
1255
+ "epoch": 1.38,
1256
+ "learning_rate": 0.00019724499574105624,
1257
+ "loss": 2.4141,
1258
+ "step": 103500
1259
+ },
1260
+ {
1261
+ "epoch": 1.38,
1262
+ "learning_rate": 0.00019723168654173765,
1263
+ "loss": 2.4732,
1264
+ "step": 104000
1265
+ },
1266
+ {
1267
+ "epoch": 1.39,
1268
+ "learning_rate": 0.0001972183773424191,
1269
+ "loss": 2.4103,
1270
+ "step": 104500
1271
+ },
1272
+ {
1273
+ "epoch": 1.4,
1274
+ "learning_rate": 0.0001972050681431005,
1275
+ "loss": 2.244,
1276
+ "step": 105000
1277
+ },
1278
+ {
1279
+ "epoch": 1.4,
1280
+ "learning_rate": 0.00019719175894378194,
1281
+ "loss": 2.4251,
1282
+ "step": 105500
1283
+ },
1284
+ {
1285
+ "epoch": 1.41,
1286
+ "learning_rate": 0.00019717844974446338,
1287
+ "loss": 2.4125,
1288
+ "step": 106000
1289
+ },
1290
+ {
1291
+ "epoch": 1.42,
1292
+ "learning_rate": 0.00019716514054514482,
1293
+ "loss": 2.4054,
1294
+ "step": 106500
1295
+ },
1296
+ {
1297
+ "epoch": 1.42,
1298
+ "learning_rate": 0.00019715183134582626,
1299
+ "loss": 2.3574,
1300
+ "step": 107000
1301
+ },
1302
+ {
1303
+ "epoch": 1.43,
1304
+ "learning_rate": 0.00019713852214650767,
1305
+ "loss": 2.4949,
1306
+ "step": 107500
1307
+ },
1308
+ {
1309
+ "epoch": 1.44,
1310
+ "learning_rate": 0.0001971252129471891,
1311
+ "loss": 2.3071,
1312
+ "step": 108000
1313
+ },
1314
+ {
1315
+ "epoch": 1.44,
1316
+ "learning_rate": 0.00019711190374787054,
1317
+ "loss": 2.3755,
1318
+ "step": 108500
1319
+ },
1320
+ {
1321
+ "epoch": 1.45,
1322
+ "learning_rate": 0.00019709859454855198,
1323
+ "loss": 2.3994,
1324
+ "step": 109000
1325
+ },
1326
+ {
1327
+ "epoch": 1.46,
1328
+ "learning_rate": 0.00019708528534923342,
1329
+ "loss": 2.425,
1330
+ "step": 109500
1331
+ },
1332
+ {
1333
+ "epoch": 1.46,
1334
+ "learning_rate": 0.00019707197614991483,
1335
+ "loss": 2.3043,
1336
+ "step": 110000
1337
+ },
1338
+ {
1339
+ "epoch": 1.47,
1340
+ "learning_rate": 0.00019705866695059627,
1341
+ "loss": 2.4273,
1342
+ "step": 110500
1343
+ },
1344
+ {
1345
+ "epoch": 1.48,
1346
+ "learning_rate": 0.00019704535775127768,
1347
+ "loss": 2.392,
1348
+ "step": 111000
1349
+ },
1350
+ {
1351
+ "epoch": 1.48,
1352
+ "learning_rate": 0.00019703204855195912,
1353
+ "loss": 2.2706,
1354
+ "step": 111500
1355
+ },
1356
+ {
1357
+ "epoch": 1.49,
1358
+ "learning_rate": 0.00019701873935264056,
1359
+ "loss": 2.3492,
1360
+ "step": 112000
1361
+ },
1362
+ {
1363
+ "epoch": 1.5,
1364
+ "learning_rate": 0.00019700543015332197,
1365
+ "loss": 2.4536,
1366
+ "step": 112500
1367
+ },
1368
+ {
1369
+ "epoch": 1.5,
1370
+ "learning_rate": 0.0001969921209540034,
1371
+ "loss": 2.4908,
1372
+ "step": 113000
1373
+ },
1374
+ {
1375
+ "epoch": 1.51,
1376
+ "learning_rate": 0.00019697881175468484,
1377
+ "loss": 2.2912,
1378
+ "step": 113500
1379
+ },
1380
+ {
1381
+ "epoch": 1.52,
1382
+ "learning_rate": 0.00019696550255536628,
1383
+ "loss": 2.346,
1384
+ "step": 114000
1385
+ },
1386
+ {
1387
+ "epoch": 1.52,
1388
+ "learning_rate": 0.00019695219335604772,
1389
+ "loss": 2.3519,
1390
+ "step": 114500
1391
+ },
1392
+ {
1393
+ "epoch": 1.53,
1394
+ "learning_rate": 0.00019693888415672913,
1395
+ "loss": 2.3962,
1396
+ "step": 115000
1397
+ },
1398
+ {
1399
+ "epoch": 1.54,
1400
+ "learning_rate": 0.00019692557495741057,
1401
+ "loss": 2.2665,
1402
+ "step": 115500
1403
+ },
1404
+ {
1405
+ "epoch": 1.54,
1406
+ "learning_rate": 0.000196912265758092,
1407
+ "loss": 2.3963,
1408
+ "step": 116000
1409
+ },
1410
+ {
1411
+ "epoch": 1.55,
1412
+ "learning_rate": 0.00019689895655877345,
1413
+ "loss": 2.3922,
1414
+ "step": 116500
1415
+ },
1416
+ {
1417
+ "epoch": 1.56,
1418
+ "learning_rate": 0.00019688564735945488,
1419
+ "loss": 2.3997,
1420
+ "step": 117000
1421
+ },
1422
+ {
1423
+ "epoch": 1.56,
1424
+ "learning_rate": 0.0001968723381601363,
1425
+ "loss": 2.391,
1426
+ "step": 117500
1427
+ },
1428
+ {
1429
+ "epoch": 1.57,
1430
+ "learning_rate": 0.00019685902896081773,
1431
+ "loss": 2.4071,
1432
+ "step": 118000
1433
+ },
1434
+ {
1435
+ "epoch": 1.58,
1436
+ "learning_rate": 0.00019684571976149914,
1437
+ "loss": 2.3302,
1438
+ "step": 118500
1439
+ },
1440
+ {
1441
+ "epoch": 1.58,
1442
+ "learning_rate": 0.00019683241056218058,
1443
+ "loss": 2.2663,
1444
+ "step": 119000
1445
+ },
1446
+ {
1447
+ "epoch": 1.59,
1448
+ "learning_rate": 0.00019681910136286202,
1449
+ "loss": 2.4085,
1450
+ "step": 119500
1451
+ },
1452
+ {
1453
+ "epoch": 1.6,
1454
+ "learning_rate": 0.00019680579216354346,
1455
+ "loss": 2.3689,
1456
+ "step": 120000
1457
+ },
1458
+ {
1459
+ "epoch": 1.6,
1460
+ "learning_rate": 0.00019679248296422487,
1461
+ "loss": 2.466,
1462
+ "step": 120500
1463
+ },
1464
+ {
1465
+ "epoch": 1.61,
1466
+ "learning_rate": 0.0001967791737649063,
1467
+ "loss": 2.3651,
1468
+ "step": 121000
1469
+ },
1470
+ {
1471
+ "epoch": 1.62,
1472
+ "learning_rate": 0.00019676586456558775,
1473
+ "loss": 2.3936,
1474
+ "step": 121500
1475
+ },
1476
+ {
1477
+ "epoch": 1.62,
1478
+ "learning_rate": 0.00019675255536626918,
1479
+ "loss": 2.3022,
1480
+ "step": 122000
1481
+ },
1482
+ {
1483
+ "epoch": 1.63,
1484
+ "learning_rate": 0.0001967392461669506,
1485
+ "loss": 2.3889,
1486
+ "step": 122500
1487
+ },
1488
+ {
1489
+ "epoch": 1.64,
1490
+ "learning_rate": 0.00019672593696763203,
1491
+ "loss": 2.2829,
1492
+ "step": 123000
1493
+ },
1494
+ {
1495
+ "epoch": 1.64,
1496
+ "learning_rate": 0.00019671262776831347,
1497
+ "loss": 2.331,
1498
+ "step": 123500
1499
+ },
1500
+ {
1501
+ "epoch": 1.65,
1502
+ "learning_rate": 0.0001966993185689949,
1503
+ "loss": 2.3535,
1504
+ "step": 124000
1505
+ },
1506
+ {
1507
+ "epoch": 1.66,
1508
+ "learning_rate": 0.00019668600936967635,
1509
+ "loss": 2.3339,
1510
+ "step": 124500
1511
+ },
1512
+ {
1513
+ "epoch": 1.66,
1514
+ "learning_rate": 0.00019667270017035776,
1515
+ "loss": 2.3372,
1516
+ "step": 125000
1517
+ },
1518
+ {
1519
+ "epoch": 1.67,
1520
+ "learning_rate": 0.00019665939097103917,
1521
+ "loss": 2.4082,
1522
+ "step": 125500
1523
+ },
1524
+ {
1525
+ "epoch": 1.68,
1526
+ "learning_rate": 0.0001966460817717206,
1527
+ "loss": 2.4428,
1528
+ "step": 126000
1529
+ },
1530
+ {
1531
+ "epoch": 1.68,
1532
+ "learning_rate": 0.00019663277257240205,
1533
+ "loss": 2.3537,
1534
+ "step": 126500
1535
+ },
1536
+ {
1537
+ "epoch": 1.69,
1538
+ "learning_rate": 0.00019661946337308348,
1539
+ "loss": 2.423,
1540
+ "step": 127000
1541
+ },
1542
+ {
1543
+ "epoch": 1.7,
1544
+ "learning_rate": 0.00019660615417376492,
1545
+ "loss": 2.2162,
1546
+ "step": 127500
1547
+ },
1548
+ {
1549
+ "epoch": 1.7,
1550
+ "learning_rate": 0.00019659284497444633,
1551
+ "loss": 2.3798,
1552
+ "step": 128000
1553
+ },
1554
+ {
1555
+ "epoch": 1.71,
1556
+ "learning_rate": 0.00019657953577512777,
1557
+ "loss": 2.3549,
1558
+ "step": 128500
1559
+ },
1560
+ {
1561
+ "epoch": 1.72,
1562
+ "learning_rate": 0.0001965662265758092,
1563
+ "loss": 2.3361,
1564
+ "step": 129000
1565
+ },
1566
+ {
1567
+ "epoch": 1.72,
1568
+ "learning_rate": 0.00019655291737649065,
1569
+ "loss": 2.3528,
1570
+ "step": 129500
1571
+ },
1572
+ {
1573
+ "epoch": 1.73,
1574
+ "learning_rate": 0.00019653960817717209,
1575
+ "loss": 2.3553,
1576
+ "step": 130000
1577
+ },
1578
+ {
1579
+ "epoch": 1.74,
1580
+ "learning_rate": 0.0001965262989778535,
1581
+ "loss": 2.1969,
1582
+ "step": 130500
1583
+ },
1584
+ {
1585
+ "epoch": 1.74,
1586
+ "learning_rate": 0.00019651298977853493,
1587
+ "loss": 2.3564,
1588
+ "step": 131000
1589
+ },
1590
+ {
1591
+ "epoch": 1.75,
1592
+ "learning_rate": 0.00019649968057921637,
1593
+ "loss": 2.3577,
1594
+ "step": 131500
1595
+ },
1596
+ {
1597
+ "epoch": 1.76,
1598
+ "learning_rate": 0.0001964863713798978,
1599
+ "loss": 2.398,
1600
+ "step": 132000
1601
+ },
1602
+ {
1603
+ "epoch": 1.76,
1604
+ "learning_rate": 0.00019647306218057922,
1605
+ "loss": 2.4231,
1606
+ "step": 132500
1607
+ },
1608
+ {
1609
+ "epoch": 1.77,
1610
+ "learning_rate": 0.00019645975298126066,
1611
+ "loss": 2.3658,
1612
+ "step": 133000
1613
+ },
1614
+ {
1615
+ "epoch": 1.78,
1616
+ "learning_rate": 0.00019644644378194207,
1617
+ "loss": 2.3445,
1618
+ "step": 133500
1619
+ },
1620
+ {
1621
+ "epoch": 1.78,
1622
+ "learning_rate": 0.0001964331345826235,
1623
+ "loss": 2.2921,
1624
+ "step": 134000
1625
+ },
1626
+ {
1627
+ "epoch": 1.79,
1628
+ "learning_rate": 0.00019641982538330495,
1629
+ "loss": 2.3487,
1630
+ "step": 134500
1631
+ },
1632
+ {
1633
+ "epoch": 1.8,
1634
+ "learning_rate": 0.00019640651618398639,
1635
+ "loss": 2.2385,
1636
+ "step": 135000
1637
+ },
1638
+ {
1639
+ "epoch": 1.8,
1640
+ "learning_rate": 0.0001963932069846678,
1641
+ "loss": 2.4189,
1642
+ "step": 135500
1643
+ },
1644
+ {
1645
+ "epoch": 1.81,
1646
+ "learning_rate": 0.00019637989778534923,
1647
+ "loss": 2.3459,
1648
+ "step": 136000
1649
+ },
1650
+ {
1651
+ "epoch": 1.82,
1652
+ "learning_rate": 0.00019636658858603067,
1653
+ "loss": 2.3295,
1654
+ "step": 136500
1655
+ },
1656
+ {
1657
+ "epoch": 1.82,
1658
+ "learning_rate": 0.0001963532793867121,
1659
+ "loss": 2.2921,
1660
+ "step": 137000
1661
+ },
1662
+ {
1663
+ "epoch": 1.83,
1664
+ "learning_rate": 0.00019633997018739355,
1665
+ "loss": 2.3225,
1666
+ "step": 137500
1667
+ },
1668
+ {
1669
+ "epoch": 1.84,
1670
+ "learning_rate": 0.00019632666098807496,
1671
+ "loss": 2.4121,
1672
+ "step": 138000
1673
+ },
1674
+ {
1675
+ "epoch": 1.84,
1676
+ "learning_rate": 0.0001963133517887564,
1677
+ "loss": 2.4108,
1678
+ "step": 138500
1679
+ },
1680
+ {
1681
+ "epoch": 1.85,
1682
+ "learning_rate": 0.00019630004258943784,
1683
+ "loss": 2.4693,
1684
+ "step": 139000
1685
+ },
1686
+ {
1687
+ "epoch": 1.86,
1688
+ "learning_rate": 0.00019628673339011927,
1689
+ "loss": 2.3493,
1690
+ "step": 139500
1691
+ },
1692
+ {
1693
+ "epoch": 1.86,
1694
+ "learning_rate": 0.00019627342419080069,
1695
+ "loss": 2.3214,
1696
+ "step": 140000
1697
+ },
1698
+ {
1699
+ "epoch": 1.87,
1700
+ "learning_rate": 0.00019626011499148212,
1701
+ "loss": 2.3452,
1702
+ "step": 140500
1703
+ },
1704
+ {
1705
+ "epoch": 1.88,
1706
+ "learning_rate": 0.00019624680579216354,
1707
+ "loss": 2.2276,
1708
+ "step": 141000
1709
+ },
1710
+ {
1711
+ "epoch": 1.88,
1712
+ "learning_rate": 0.00019623349659284497,
1713
+ "loss": 2.2652,
1714
+ "step": 141500
1715
+ },
1716
+ {
1717
+ "epoch": 1.89,
1718
+ "learning_rate": 0.0001962201873935264,
1719
+ "loss": 2.406,
1720
+ "step": 142000
1721
+ },
1722
+ {
1723
+ "epoch": 1.9,
1724
+ "learning_rate": 0.00019620687819420785,
1725
+ "loss": 2.2756,
1726
+ "step": 142500
1727
+ },
1728
+ {
1729
+ "epoch": 1.9,
1730
+ "learning_rate": 0.0001961935689948893,
1731
+ "loss": 2.2884,
1732
+ "step": 143000
1733
+ },
1734
+ {
1735
+ "epoch": 1.91,
1736
+ "learning_rate": 0.0001961802597955707,
1737
+ "loss": 2.27,
1738
+ "step": 143500
1739
+ },
1740
+ {
1741
+ "epoch": 1.92,
1742
+ "learning_rate": 0.00019616695059625214,
1743
+ "loss": 2.3682,
1744
+ "step": 144000
1745
+ },
1746
+ {
1747
+ "epoch": 1.92,
1748
+ "learning_rate": 0.00019615364139693358,
1749
+ "loss": 2.2721,
1750
+ "step": 144500
1751
+ },
1752
+ {
1753
+ "epoch": 1.93,
1754
+ "learning_rate": 0.000196140332197615,
1755
+ "loss": 2.3077,
1756
+ "step": 145000
1757
+ },
1758
+ {
1759
+ "epoch": 1.94,
1760
+ "learning_rate": 0.00019612702299829645,
1761
+ "loss": 2.2378,
1762
+ "step": 145500
1763
+ },
1764
+ {
1765
+ "epoch": 1.94,
1766
+ "learning_rate": 0.00019611371379897786,
1767
+ "loss": 2.4656,
1768
+ "step": 146000
1769
+ },
1770
+ {
1771
+ "epoch": 1.95,
1772
+ "learning_rate": 0.0001961004045996593,
1773
+ "loss": 2.2858,
1774
+ "step": 146500
1775
+ },
1776
+ {
1777
+ "epoch": 1.96,
1778
+ "learning_rate": 0.0001960870954003407,
1779
+ "loss": 2.2983,
1780
+ "step": 147000
1781
+ },
1782
+ {
1783
+ "epoch": 1.96,
1784
+ "learning_rate": 0.00019607378620102215,
1785
+ "loss": 2.3561,
1786
+ "step": 147500
1787
+ },
1788
+ {
1789
+ "epoch": 1.97,
1790
+ "learning_rate": 0.0001960604770017036,
1791
+ "loss": 2.3414,
1792
+ "step": 148000
1793
+ },
1794
+ {
1795
+ "epoch": 1.98,
1796
+ "learning_rate": 0.000196047167802385,
1797
+ "loss": 2.279,
1798
+ "step": 148500
1799
+ },
1800
+ {
1801
+ "epoch": 1.98,
1802
+ "learning_rate": 0.00019603385860306644,
1803
+ "loss": 2.3095,
1804
+ "step": 149000
1805
+ },
1806
+ {
1807
+ "epoch": 1.99,
1808
+ "learning_rate": 0.00019602054940374788,
1809
+ "loss": 2.3459,
1810
+ "step": 149500
1811
+ },
1812
+ {
1813
+ "epoch": 2.0,
1814
+ "learning_rate": 0.0001960072402044293,
1815
+ "loss": 2.3968,
1816
+ "step": 150000
1817
+ },
1818
+ {
1819
+ "epoch": 2.0,
1820
+ "eval_loss": 2.083725929260254,
1821
+ "eval_runtime": 150.1089,
1822
+ "eval_samples_per_second": 62.575,
1823
+ "eval_steps_per_second": 62.575,
1824
+ "step": 150272
1825
+ },
1826
+ {
1827
+ "epoch": 2.0,
1828
+ "learning_rate": 0.00019599393100511075,
1829
+ "loss": 2.245,
1830
+ "step": 150500
1831
+ },
1832
+ {
1833
+ "epoch": 2.01,
1834
+ "learning_rate": 0.00019598062180579216,
1835
+ "loss": 2.1479,
1836
+ "step": 151000
1837
+ },
1838
+ {
1839
+ "epoch": 2.02,
1840
+ "learning_rate": 0.0001959673126064736,
1841
+ "loss": 2.1048,
1842
+ "step": 151500
1843
+ },
1844
+ {
1845
+ "epoch": 2.02,
1846
+ "learning_rate": 0.00019595400340715504,
1847
+ "loss": 2.028,
1848
+ "step": 152000
1849
+ },
1850
+ {
1851
+ "epoch": 2.03,
1852
+ "learning_rate": 0.00019594069420783648,
1853
+ "loss": 2.0738,
1854
+ "step": 152500
1855
+ },
1856
+ {
1857
+ "epoch": 2.04,
1858
+ "learning_rate": 0.00019592738500851792,
1859
+ "loss": 2.1684,
1860
+ "step": 153000
1861
+ },
1862
+ {
1863
+ "epoch": 2.04,
1864
+ "learning_rate": 0.00019591407580919933,
1865
+ "loss": 2.0715,
1866
+ "step": 153500
1867
+ },
1868
+ {
1869
+ "epoch": 2.05,
1870
+ "learning_rate": 0.00019590076660988074,
1871
+ "loss": 2.1869,
1872
+ "step": 154000
1873
+ },
1874
+ {
1875
+ "epoch": 2.06,
1876
+ "learning_rate": 0.00019588745741056218,
1877
+ "loss": 2.0887,
1878
+ "step": 154500
1879
+ },
1880
+ {
1881
+ "epoch": 2.06,
1882
+ "learning_rate": 0.00019587414821124361,
1883
+ "loss": 2.0761,
1884
+ "step": 155000
1885
+ },
1886
+ {
1887
+ "epoch": 2.07,
1888
+ "learning_rate": 0.00019586083901192505,
1889
+ "loss": 2.0249,
1890
+ "step": 155500
1891
+ },
1892
+ {
1893
+ "epoch": 2.08,
1894
+ "learning_rate": 0.0001958475298126065,
1895
+ "loss": 1.998,
1896
+ "step": 156000
1897
+ },
1898
+ {
1899
+ "epoch": 2.08,
1900
+ "learning_rate": 0.0001958342206132879,
1901
+ "loss": 2.186,
1902
+ "step": 156500
1903
+ },
1904
+ {
1905
+ "epoch": 2.09,
1906
+ "learning_rate": 0.00019582091141396934,
1907
+ "loss": 2.2012,
1908
+ "step": 157000
1909
+ },
1910
+ {
1911
+ "epoch": 2.1,
1912
+ "learning_rate": 0.00019580760221465078,
1913
+ "loss": 2.1327,
1914
+ "step": 157500
1915
+ },
1916
+ {
1917
+ "epoch": 2.1,
1918
+ "learning_rate": 0.00019579429301533222,
1919
+ "loss": 2.0996,
1920
+ "step": 158000
1921
+ },
1922
+ {
1923
+ "epoch": 2.11,
1924
+ "learning_rate": 0.00019578098381601365,
1925
+ "loss": 2.0994,
1926
+ "step": 158500
1927
+ },
1928
+ {
1929
+ "epoch": 2.12,
1930
+ "learning_rate": 0.00019576767461669506,
1931
+ "loss": 2.057,
1932
+ "step": 159000
1933
+ },
1934
+ {
1935
+ "epoch": 2.12,
1936
+ "learning_rate": 0.0001957543654173765,
1937
+ "loss": 2.0814,
1938
+ "step": 159500
1939
+ },
1940
+ {
1941
+ "epoch": 2.13,
1942
+ "learning_rate": 0.00019574105621805794,
1943
+ "loss": 2.2237,
1944
+ "step": 160000
1945
+ },
1946
+ {
1947
+ "epoch": 2.14,
1948
+ "learning_rate": 0.00019572774701873938,
1949
+ "loss": 2.113,
1950
+ "step": 160500
1951
+ },
1952
+ {
1953
+ "epoch": 2.14,
1954
+ "learning_rate": 0.00019571443781942082,
1955
+ "loss": 2.2014,
1956
+ "step": 161000
1957
+ },
1958
+ {
1959
+ "epoch": 2.15,
1960
+ "learning_rate": 0.0001957011286201022,
1961
+ "loss": 2.1535,
1962
+ "step": 161500
1963
+ },
1964
+ {
1965
+ "epoch": 2.16,
1966
+ "learning_rate": 0.00019568781942078364,
1967
+ "loss": 2.1548,
1968
+ "step": 162000
1969
+ },
1970
+ {
1971
+ "epoch": 2.16,
1972
+ "learning_rate": 0.00019567451022146508,
1973
+ "loss": 2.2726,
1974
+ "step": 162500
1975
+ },
1976
+ {
1977
+ "epoch": 2.17,
1978
+ "learning_rate": 0.00019566120102214652,
1979
+ "loss": 2.0613,
1980
+ "step": 163000
1981
+ },
1982
+ {
1983
+ "epoch": 2.18,
1984
+ "learning_rate": 0.00019564789182282795,
1985
+ "loss": 2.1457,
1986
+ "step": 163500
1987
+ },
1988
+ {
1989
+ "epoch": 2.18,
1990
+ "learning_rate": 0.00019563458262350936,
1991
+ "loss": 2.0963,
1992
+ "step": 164000
1993
+ },
1994
+ {
1995
+ "epoch": 2.19,
1996
+ "learning_rate": 0.0001956212734241908,
1997
+ "loss": 2.1481,
1998
+ "step": 164500
1999
+ },
2000
+ {
2001
+ "epoch": 2.2,
2002
+ "learning_rate": 0.00019560796422487224,
2003
+ "loss": 2.104,
2004
+ "step": 165000
2005
+ },
2006
+ {
2007
+ "epoch": 2.2,
2008
+ "learning_rate": 0.00019559465502555368,
2009
+ "loss": 2.0308,
2010
+ "step": 165500
2011
+ },
2012
+ {
2013
+ "epoch": 2.21,
2014
+ "learning_rate": 0.00019558134582623512,
2015
+ "loss": 2.0471,
2016
+ "step": 166000
2017
+ },
2018
+ {
2019
+ "epoch": 2.22,
2020
+ "learning_rate": 0.00019556803662691653,
2021
+ "loss": 2.1354,
2022
+ "step": 166500
2023
+ },
2024
+ {
2025
+ "epoch": 2.22,
2026
+ "learning_rate": 0.00019555472742759797,
2027
+ "loss": 2.1223,
2028
+ "step": 167000
2029
+ },
2030
+ {
2031
+ "epoch": 2.23,
2032
+ "learning_rate": 0.0001955414182282794,
2033
+ "loss": 2.1158,
2034
+ "step": 167500
2035
+ },
2036
+ {
2037
+ "epoch": 2.24,
2038
+ "learning_rate": 0.00019552810902896084,
2039
+ "loss": 2.2403,
2040
+ "step": 168000
2041
+ },
2042
+ {
2043
+ "epoch": 2.24,
2044
+ "learning_rate": 0.00019551479982964225,
2045
+ "loss": 2.2066,
2046
+ "step": 168500
2047
+ },
2048
+ {
2049
+ "epoch": 2.25,
2050
+ "learning_rate": 0.0001955014906303237,
2051
+ "loss": 2.0638,
2052
+ "step": 169000
2053
+ },
2054
+ {
2055
+ "epoch": 2.26,
2056
+ "learning_rate": 0.0001954881814310051,
2057
+ "loss": 2.25,
2058
+ "step": 169500
2059
+ },
2060
+ {
2061
+ "epoch": 2.26,
2062
+ "learning_rate": 0.00019547487223168654,
2063
+ "loss": 2.1617,
2064
+ "step": 170000
2065
+ },
2066
+ {
2067
+ "epoch": 2.27,
2068
+ "learning_rate": 0.00019546156303236798,
2069
+ "loss": 2.2301,
2070
+ "step": 170500
2071
+ },
2072
+ {
2073
+ "epoch": 2.28,
2074
+ "learning_rate": 0.00019544825383304942,
2075
+ "loss": 2.165,
2076
+ "step": 171000
2077
+ },
2078
+ {
2079
+ "epoch": 2.28,
2080
+ "learning_rate": 0.00019543494463373086,
2081
+ "loss": 2.1395,
2082
+ "step": 171500
2083
+ },
2084
+ {
2085
+ "epoch": 2.29,
2086
+ "learning_rate": 0.00019542163543441227,
2087
+ "loss": 2.1827,
2088
+ "step": 172000
2089
+ },
2090
+ {
2091
+ "epoch": 2.3,
2092
+ "learning_rate": 0.0001954083262350937,
2093
+ "loss": 2.1452,
2094
+ "step": 172500
2095
+ },
2096
+ {
2097
+ "epoch": 2.3,
2098
+ "learning_rate": 0.00019539501703577514,
2099
+ "loss": 2.1515,
2100
+ "step": 173000
2101
+ },
2102
+ {
2103
+ "epoch": 2.31,
2104
+ "learning_rate": 0.00019538170783645658,
2105
+ "loss": 2.1997,
2106
+ "step": 173500
2107
+ },
2108
+ {
2109
+ "epoch": 2.32,
2110
+ "learning_rate": 0.00019536839863713802,
2111
+ "loss": 2.1392,
2112
+ "step": 174000
2113
+ },
2114
+ {
2115
+ "epoch": 2.32,
2116
+ "learning_rate": 0.00019535508943781943,
2117
+ "loss": 2.0866,
2118
+ "step": 174500
2119
+ },
2120
+ {
2121
+ "epoch": 2.33,
2122
+ "learning_rate": 0.00019534178023850087,
2123
+ "loss": 2.0945,
2124
+ "step": 175000
2125
+ },
2126
+ {
2127
+ "epoch": 2.34,
2128
+ "learning_rate": 0.00019532847103918228,
2129
+ "loss": 2.2065,
2130
+ "step": 175500
2131
+ },
2132
+ {
2133
+ "epoch": 2.34,
2134
+ "learning_rate": 0.00019531516183986372,
2135
+ "loss": 2.1329,
2136
+ "step": 176000
2137
+ },
2138
+ {
2139
+ "epoch": 2.35,
2140
+ "learning_rate": 0.00019530185264054516,
2141
+ "loss": 2.165,
2142
+ "step": 176500
2143
+ },
2144
+ {
2145
+ "epoch": 2.36,
2146
+ "learning_rate": 0.00019528854344122657,
2147
+ "loss": 2.1542,
2148
+ "step": 177000
2149
+ },
2150
+ {
2151
+ "epoch": 2.36,
2152
+ "learning_rate": 0.000195275234241908,
2153
+ "loss": 2.1892,
2154
+ "step": 177500
2155
+ },
2156
+ {
2157
+ "epoch": 2.37,
2158
+ "learning_rate": 0.00019526192504258944,
2159
+ "loss": 2.1984,
2160
+ "step": 178000
2161
+ },
2162
+ {
2163
+ "epoch": 2.38,
2164
+ "learning_rate": 0.00019524861584327088,
2165
+ "loss": 2.2268,
2166
+ "step": 178500
2167
+ },
2168
+ {
2169
+ "epoch": 2.38,
2170
+ "learning_rate": 0.00019523530664395232,
2171
+ "loss": 2.0896,
2172
+ "step": 179000
2173
+ },
2174
+ {
2175
+ "epoch": 2.39,
2176
+ "learning_rate": 0.00019522199744463373,
2177
+ "loss": 2.146,
2178
+ "step": 179500
2179
+ },
2180
+ {
2181
+ "epoch": 2.4,
2182
+ "learning_rate": 0.00019520868824531517,
2183
+ "loss": 2.2207,
2184
+ "step": 180000
2185
+ },
2186
+ {
2187
+ "epoch": 2.4,
2188
+ "learning_rate": 0.0001951953790459966,
2189
+ "loss": 2.2925,
2190
+ "step": 180500
2191
+ },
2192
+ {
2193
+ "epoch": 2.41,
2194
+ "learning_rate": 0.00019518206984667805,
2195
+ "loss": 2.1177,
2196
+ "step": 181000
2197
+ },
2198
+ {
2199
+ "epoch": 2.42,
2200
+ "learning_rate": 0.00019516876064735948,
2201
+ "loss": 2.1932,
2202
+ "step": 181500
2203
+ },
2204
+ {
2205
+ "epoch": 2.42,
2206
+ "learning_rate": 0.0001951554514480409,
2207
+ "loss": 2.1481,
2208
+ "step": 182000
2209
+ },
2210
+ {
2211
+ "epoch": 2.43,
2212
+ "learning_rate": 0.00019514214224872233,
2213
+ "loss": 2.0231,
2214
+ "step": 182500
2215
+ },
2216
+ {
2217
+ "epoch": 2.44,
2218
+ "learning_rate": 0.00019512883304940374,
2219
+ "loss": 2.2687,
2220
+ "step": 183000
2221
+ },
2222
+ {
2223
+ "epoch": 2.44,
2224
+ "learning_rate": 0.00019511552385008518,
2225
+ "loss": 2.0866,
2226
+ "step": 183500
2227
+ },
2228
+ {
2229
+ "epoch": 2.45,
2230
+ "learning_rate": 0.00019510221465076662,
2231
+ "loss": 2.1907,
2232
+ "step": 184000
2233
+ },
2234
+ {
2235
+ "epoch": 2.46,
2236
+ "learning_rate": 0.00019508890545144806,
2237
+ "loss": 2.1635,
2238
+ "step": 184500
2239
+ },
2240
+ {
2241
+ "epoch": 2.46,
2242
+ "learning_rate": 0.00019507559625212947,
2243
+ "loss": 2.0831,
2244
+ "step": 185000
2245
+ },
2246
+ {
2247
+ "epoch": 2.47,
2248
+ "learning_rate": 0.0001950622870528109,
2249
+ "loss": 2.1906,
2250
+ "step": 185500
2251
+ },
2252
+ {
2253
+ "epoch": 2.48,
2254
+ "learning_rate": 0.00019504897785349235,
2255
+ "loss": 2.2034,
2256
+ "step": 186000
2257
+ },
2258
+ {
2259
+ "epoch": 2.48,
2260
+ "learning_rate": 0.00019503566865417378,
2261
+ "loss": 2.2103,
2262
+ "step": 186500
2263
+ },
2264
+ {
2265
+ "epoch": 2.49,
2266
+ "learning_rate": 0.0001950223594548552,
2267
+ "loss": 2.1678,
2268
+ "step": 187000
2269
+ },
2270
+ {
2271
+ "epoch": 2.5,
2272
+ "learning_rate": 0.00019500905025553663,
2273
+ "loss": 2.0417,
2274
+ "step": 187500
2275
+ },
2276
+ {
2277
+ "epoch": 2.5,
2278
+ "learning_rate": 0.00019499574105621807,
2279
+ "loss": 2.1806,
2280
+ "step": 188000
2281
+ },
2282
+ {
2283
+ "epoch": 2.51,
2284
+ "learning_rate": 0.0001949824318568995,
2285
+ "loss": 2.21,
2286
+ "step": 188500
2287
+ },
2288
+ {
2289
+ "epoch": 2.52,
2290
+ "learning_rate": 0.00019496912265758095,
2291
+ "loss": 2.1416,
2292
+ "step": 189000
2293
+ },
2294
+ {
2295
+ "epoch": 2.52,
2296
+ "learning_rate": 0.00019495581345826236,
2297
+ "loss": 2.1792,
2298
+ "step": 189500
2299
+ },
2300
+ {
2301
+ "epoch": 2.53,
2302
+ "learning_rate": 0.00019494250425894377,
2303
+ "loss": 2.2475,
2304
+ "step": 190000
2305
+ },
2306
+ {
2307
+ "epoch": 2.54,
2308
+ "learning_rate": 0.0001949291950596252,
2309
+ "loss": 2.1561,
2310
+ "step": 190500
2311
+ },
2312
+ {
2313
+ "epoch": 2.54,
2314
+ "learning_rate": 0.00019491588586030665,
2315
+ "loss": 2.1854,
2316
+ "step": 191000
2317
+ },
2318
+ {
2319
+ "epoch": 2.55,
2320
+ "learning_rate": 0.00019490257666098808,
2321
+ "loss": 2.2039,
2322
+ "step": 191500
2323
+ },
2324
+ {
2325
+ "epoch": 2.56,
2326
+ "learning_rate": 0.00019488926746166952,
2327
+ "loss": 2.2249,
2328
+ "step": 192000
2329
+ },
2330
+ {
2331
+ "epoch": 2.56,
2332
+ "learning_rate": 0.00019487595826235093,
2333
+ "loss": 2.0889,
2334
+ "step": 192500
2335
+ },
2336
+ {
2337
+ "epoch": 2.57,
2338
+ "learning_rate": 0.00019486264906303237,
2339
+ "loss": 2.2046,
2340
+ "step": 193000
2341
+ },
2342
+ {
2343
+ "epoch": 2.58,
2344
+ "learning_rate": 0.0001948493398637138,
2345
+ "loss": 2.0463,
2346
+ "step": 193500
2347
+ },
2348
+ {
2349
+ "epoch": 2.58,
2350
+ "learning_rate": 0.00019483603066439525,
2351
+ "loss": 2.0919,
2352
+ "step": 194000
2353
+ },
2354
+ {
2355
+ "epoch": 2.59,
2356
+ "learning_rate": 0.00019482272146507669,
2357
+ "loss": 2.1278,
2358
+ "step": 194500
2359
+ },
2360
+ {
2361
+ "epoch": 2.6,
2362
+ "learning_rate": 0.0001948094122657581,
2363
+ "loss": 2.1866,
2364
+ "step": 195000
2365
+ },
2366
+ {
2367
+ "epoch": 2.6,
2368
+ "learning_rate": 0.00019479610306643953,
2369
+ "loss": 2.25,
2370
+ "step": 195500
2371
+ },
2372
+ {
2373
+ "epoch": 2.61,
2374
+ "learning_rate": 0.00019478279386712097,
2375
+ "loss": 2.038,
2376
+ "step": 196000
2377
+ },
2378
+ {
2379
+ "epoch": 2.62,
2380
+ "learning_rate": 0.0001947694846678024,
2381
+ "loss": 2.0665,
2382
+ "step": 196500
2383
+ },
2384
+ {
2385
+ "epoch": 2.62,
2386
+ "learning_rate": 0.00019475617546848382,
2387
+ "loss": 2.0884,
2388
+ "step": 197000
2389
+ },
2390
+ {
2391
+ "epoch": 2.63,
2392
+ "learning_rate": 0.00019474286626916526,
2393
+ "loss": 2.1792,
2394
+ "step": 197500
2395
+ },
2396
+ {
2397
+ "epoch": 2.64,
2398
+ "learning_rate": 0.00019472955706984667,
2399
+ "loss": 2.1597,
2400
+ "step": 198000
2401
+ },
2402
+ {
2403
+ "epoch": 2.64,
2404
+ "learning_rate": 0.0001947162478705281,
2405
+ "loss": 2.1973,
2406
+ "step": 198500
2407
+ },
2408
+ {
2409
+ "epoch": 2.65,
2410
+ "learning_rate": 0.00019470293867120955,
2411
+ "loss": 2.1074,
2412
+ "step": 199000
2413
+ },
2414
+ {
2415
+ "epoch": 2.66,
2416
+ "learning_rate": 0.00019468962947189099,
2417
+ "loss": 2.1212,
2418
+ "step": 199500
2419
+ },
2420
+ {
2421
+ "epoch": 2.66,
2422
+ "learning_rate": 0.0001946763202725724,
2423
+ "loss": 2.1406,
2424
+ "step": 200000
2425
+ },
2426
+ {
2427
+ "epoch": 2.67,
2428
+ "learning_rate": 0.00019466301107325383,
2429
+ "loss": 2.1905,
2430
+ "step": 200500
2431
+ },
2432
+ {
2433
+ "epoch": 2.68,
2434
+ "learning_rate": 0.00019464970187393527,
2435
+ "loss": 2.0908,
2436
+ "step": 201000
2437
+ },
2438
+ {
2439
+ "epoch": 2.68,
2440
+ "learning_rate": 0.0001946363926746167,
2441
+ "loss": 2.1969,
2442
+ "step": 201500
2443
+ },
2444
+ {
2445
+ "epoch": 2.69,
2446
+ "learning_rate": 0.00019462308347529815,
2447
+ "loss": 2.246,
2448
+ "step": 202000
2449
+ },
2450
+ {
2451
+ "epoch": 2.7,
2452
+ "learning_rate": 0.00019460977427597956,
2453
+ "loss": 2.0437,
2454
+ "step": 202500
2455
+ },
2456
+ {
2457
+ "epoch": 2.7,
2458
+ "learning_rate": 0.000194596465076661,
2459
+ "loss": 2.2312,
2460
+ "step": 203000
2461
+ },
2462
+ {
2463
+ "epoch": 2.71,
2464
+ "learning_rate": 0.00019458315587734244,
2465
+ "loss": 2.182,
2466
+ "step": 203500
2467
+ },
2468
+ {
2469
+ "epoch": 2.72,
2470
+ "learning_rate": 0.00019456984667802387,
2471
+ "loss": 2.2265,
2472
+ "step": 204000
2473
+ },
2474
+ {
2475
+ "epoch": 2.72,
2476
+ "learning_rate": 0.00019455653747870529,
2477
+ "loss": 2.1802,
2478
+ "step": 204500
2479
+ },
2480
+ {
2481
+ "epoch": 2.73,
2482
+ "learning_rate": 0.00019454322827938672,
2483
+ "loss": 2.1688,
2484
+ "step": 205000
2485
+ },
2486
+ {
2487
+ "epoch": 2.74,
2488
+ "learning_rate": 0.00019452991908006813,
2489
+ "loss": 2.1946,
2490
+ "step": 205500
2491
+ },
2492
+ {
2493
+ "epoch": 2.74,
2494
+ "learning_rate": 0.00019451660988074957,
2495
+ "loss": 2.2045,
2496
+ "step": 206000
2497
+ },
2498
+ {
2499
+ "epoch": 2.75,
2500
+ "learning_rate": 0.000194503300681431,
2501
+ "loss": 2.116,
2502
+ "step": 206500
2503
+ },
2504
+ {
2505
+ "epoch": 2.76,
2506
+ "learning_rate": 0.00019448999148211245,
2507
+ "loss": 2.046,
2508
+ "step": 207000
2509
+ },
2510
+ {
2511
+ "epoch": 2.76,
2512
+ "learning_rate": 0.0001944766822827939,
2513
+ "loss": 2.1339,
2514
+ "step": 207500
2515
+ },
2516
+ {
2517
+ "epoch": 2.77,
2518
+ "learning_rate": 0.0001944633730834753,
2519
+ "loss": 2.2015,
2520
+ "step": 208000
2521
+ },
2522
+ {
2523
+ "epoch": 2.77,
2524
+ "learning_rate": 0.00019445006388415674,
2525
+ "loss": 2.1834,
2526
+ "step": 208500
2527
+ },
2528
+ {
2529
+ "epoch": 2.78,
2530
+ "learning_rate": 0.00019443675468483817,
2531
+ "loss": 2.147,
2532
+ "step": 209000
2533
+ },
2534
+ {
2535
+ "epoch": 2.79,
2536
+ "learning_rate": 0.0001944234454855196,
2537
+ "loss": 2.1619,
2538
+ "step": 209500
2539
+ },
2540
+ {
2541
+ "epoch": 2.79,
2542
+ "learning_rate": 0.00019441013628620105,
2543
+ "loss": 2.184,
2544
+ "step": 210000
2545
+ },
2546
+ {
2547
+ "epoch": 2.8,
2548
+ "learning_rate": 0.00019439682708688246,
2549
+ "loss": 2.1712,
2550
+ "step": 210500
2551
+ },
2552
+ {
2553
+ "epoch": 2.81,
2554
+ "learning_rate": 0.0001943835178875639,
2555
+ "loss": 2.2269,
2556
+ "step": 211000
2557
+ },
2558
+ {
2559
+ "epoch": 2.81,
2560
+ "learning_rate": 0.0001943702086882453,
2561
+ "loss": 2.1537,
2562
+ "step": 211500
2563
+ },
2564
+ {
2565
+ "epoch": 2.82,
2566
+ "learning_rate": 0.00019435689948892675,
2567
+ "loss": 2.1268,
2568
+ "step": 212000
2569
+ },
2570
+ {
2571
+ "epoch": 2.83,
2572
+ "learning_rate": 0.0001943435902896082,
2573
+ "loss": 2.1349,
2574
+ "step": 212500
2575
+ },
2576
+ {
2577
+ "epoch": 2.83,
2578
+ "learning_rate": 0.0001943302810902896,
2579
+ "loss": 2.0699,
2580
+ "step": 213000
2581
+ },
2582
+ {
2583
+ "epoch": 2.84,
2584
+ "learning_rate": 0.00019431697189097104,
2585
+ "loss": 2.1889,
2586
+ "step": 213500
2587
+ },
2588
+ {
2589
+ "epoch": 2.85,
2590
+ "learning_rate": 0.00019430366269165248,
2591
+ "loss": 2.1516,
2592
+ "step": 214000
2593
+ },
2594
+ {
2595
+ "epoch": 2.85,
2596
+ "learning_rate": 0.0001942903534923339,
2597
+ "loss": 2.1827,
2598
+ "step": 214500
2599
+ },
2600
+ {
2601
+ "epoch": 2.86,
2602
+ "learning_rate": 0.00019427704429301535,
2603
+ "loss": 2.2482,
2604
+ "step": 215000
2605
+ },
2606
+ {
2607
+ "epoch": 2.87,
2608
+ "learning_rate": 0.00019426373509369676,
2609
+ "loss": 2.2037,
2610
+ "step": 215500
2611
+ },
2612
+ {
2613
+ "epoch": 2.87,
2614
+ "learning_rate": 0.0001942504258943782,
2615
+ "loss": 2.1865,
2616
+ "step": 216000
2617
+ },
2618
+ {
2619
+ "epoch": 2.88,
2620
+ "learning_rate": 0.00019423711669505964,
2621
+ "loss": 2.1926,
2622
+ "step": 216500
2623
+ },
2624
+ {
2625
+ "epoch": 2.89,
2626
+ "learning_rate": 0.00019422380749574108,
2627
+ "loss": 2.2028,
2628
+ "step": 217000
2629
+ },
2630
+ {
2631
+ "epoch": 2.89,
2632
+ "learning_rate": 0.00019421049829642252,
2633
+ "loss": 2.2157,
2634
+ "step": 217500
2635
+ },
2636
+ {
2637
+ "epoch": 2.9,
2638
+ "learning_rate": 0.00019419718909710393,
2639
+ "loss": 2.1386,
2640
+ "step": 218000
2641
+ },
2642
+ {
2643
+ "epoch": 2.91,
2644
+ "learning_rate": 0.00019418387989778534,
2645
+ "loss": 2.1053,
2646
+ "step": 218500
2647
+ },
2648
+ {
2649
+ "epoch": 2.91,
2650
+ "learning_rate": 0.00019417057069846678,
2651
+ "loss": 2.1686,
2652
+ "step": 219000
2653
+ },
2654
+ {
2655
+ "epoch": 2.92,
2656
+ "learning_rate": 0.0001941572614991482,
2657
+ "loss": 2.1914,
2658
+ "step": 219500
2659
+ },
2660
+ {
2661
+ "epoch": 2.93,
2662
+ "learning_rate": 0.00019414395229982965,
2663
+ "loss": 2.2137,
2664
+ "step": 220000
2665
+ },
2666
+ {
2667
+ "epoch": 2.93,
2668
+ "learning_rate": 0.0001941306431005111,
2669
+ "loss": 2.0708,
2670
+ "step": 220500
2671
+ },
2672
+ {
2673
+ "epoch": 2.94,
2674
+ "learning_rate": 0.0001941173339011925,
2675
+ "loss": 2.1136,
2676
+ "step": 221000
2677
+ },
2678
+ {
2679
+ "epoch": 2.95,
2680
+ "learning_rate": 0.00019410402470187394,
2681
+ "loss": 2.071,
2682
+ "step": 221500
2683
+ },
2684
+ {
2685
+ "epoch": 2.95,
2686
+ "learning_rate": 0.00019409071550255538,
2687
+ "loss": 2.1486,
2688
+ "step": 222000
2689
+ },
2690
+ {
2691
+ "epoch": 2.96,
2692
+ "learning_rate": 0.00019407740630323682,
2693
+ "loss": 2.1378,
2694
+ "step": 222500
2695
+ },
2696
+ {
2697
+ "epoch": 2.97,
2698
+ "learning_rate": 0.00019406409710391825,
2699
+ "loss": 2.0198,
2700
+ "step": 223000
2701
+ },
2702
+ {
2703
+ "epoch": 2.97,
2704
+ "learning_rate": 0.00019405078790459966,
2705
+ "loss": 2.2202,
2706
+ "step": 223500
2707
+ },
2708
+ {
2709
+ "epoch": 2.98,
2710
+ "learning_rate": 0.0001940374787052811,
2711
+ "loss": 2.1,
2712
+ "step": 224000
2713
+ },
2714
+ {
2715
+ "epoch": 2.99,
2716
+ "learning_rate": 0.00019402416950596254,
2717
+ "loss": 2.1922,
2718
+ "step": 224500
2719
+ },
2720
+ {
2721
+ "epoch": 2.99,
2722
+ "learning_rate": 0.00019401086030664398,
2723
+ "loss": 2.1636,
2724
+ "step": 225000
2725
+ },
2726
+ {
2727
+ "epoch": 3.0,
2728
+ "eval_loss": 2.014349937438965,
2729
+ "eval_runtime": 150.1185,
2730
+ "eval_samples_per_second": 62.571,
2731
+ "eval_steps_per_second": 62.571,
2732
+ "step": 225408
2733
+ },
2734
+ {
2735
+ "epoch": 3.0,
2736
+ "learning_rate": 0.0001939975511073254,
2737
+ "loss": 2.0504,
2738
+ "step": 225500
2739
+ },
2740
+ {
2741
+ "epoch": 3.01,
2742
+ "learning_rate": 0.0001939842419080068,
2743
+ "loss": 1.995,
2744
+ "step": 226000
2745
+ },
2746
+ {
2747
+ "epoch": 3.01,
2748
+ "learning_rate": 0.00019397093270868824,
2749
+ "loss": 1.9558,
2750
+ "step": 226500
2751
+ },
2752
+ {
2753
+ "epoch": 3.02,
2754
+ "learning_rate": 0.00019395762350936968,
2755
+ "loss": 1.9273,
2756
+ "step": 227000
2757
+ },
2758
+ {
2759
+ "epoch": 3.03,
2760
+ "learning_rate": 0.00019394431431005112,
2761
+ "loss": 1.9675,
2762
+ "step": 227500
2763
+ },
2764
+ {
2765
+ "epoch": 3.03,
2766
+ "learning_rate": 0.00019393100511073255,
2767
+ "loss": 1.9863,
2768
+ "step": 228000
2769
+ },
2770
+ {
2771
+ "epoch": 3.04,
2772
+ "learning_rate": 0.00019391769591141396,
2773
+ "loss": 1.9841,
2774
+ "step": 228500
2775
+ },
2776
+ {
2777
+ "epoch": 3.05,
2778
+ "learning_rate": 0.0001939043867120954,
2779
+ "loss": 1.9306,
2780
+ "step": 229000
2781
+ },
2782
+ {
2783
+ "epoch": 3.05,
2784
+ "learning_rate": 0.00019389107751277684,
2785
+ "loss": 1.9797,
2786
+ "step": 229500
2787
+ },
2788
+ {
2789
+ "epoch": 3.06,
2790
+ "learning_rate": 0.00019387776831345828,
2791
+ "loss": 1.8735,
2792
+ "step": 230000
2793
+ },
2794
+ {
2795
+ "epoch": 3.07,
2796
+ "learning_rate": 0.00019386445911413972,
2797
+ "loss": 1.9623,
2798
+ "step": 230500
2799
+ },
2800
+ {
2801
+ "epoch": 3.07,
2802
+ "learning_rate": 0.00019385114991482113,
2803
+ "loss": 1.9795,
2804
+ "step": 231000
2805
+ },
2806
+ {
2807
+ "epoch": 3.08,
2808
+ "learning_rate": 0.00019383784071550257,
2809
+ "loss": 1.862,
2810
+ "step": 231500
2811
+ },
2812
+ {
2813
+ "epoch": 3.09,
2814
+ "learning_rate": 0.000193824531516184,
2815
+ "loss": 2.0137,
2816
+ "step": 232000
2817
+ },
2818
+ {
2819
+ "epoch": 3.09,
2820
+ "learning_rate": 0.00019381122231686544,
2821
+ "loss": 1.9126,
2822
+ "step": 232500
2823
+ },
2824
+ {
2825
+ "epoch": 3.1,
2826
+ "learning_rate": 0.00019379791311754685,
2827
+ "loss": 1.9946,
2828
+ "step": 233000
2829
+ },
2830
+ {
2831
+ "epoch": 3.11,
2832
+ "learning_rate": 0.0001937846039182283,
2833
+ "loss": 1.9526,
2834
+ "step": 233500
2835
+ },
2836
+ {
2837
+ "epoch": 3.11,
2838
+ "learning_rate": 0.0001937712947189097,
2839
+ "loss": 1.954,
2840
+ "step": 234000
2841
+ },
2842
+ {
2843
+ "epoch": 3.12,
2844
+ "learning_rate": 0.00019375798551959114,
2845
+ "loss": 1.9714,
2846
+ "step": 234500
2847
+ },
2848
+ {
2849
+ "epoch": 3.13,
2850
+ "learning_rate": 0.00019374467632027258,
2851
+ "loss": 1.917,
2852
+ "step": 235000
2853
+ },
2854
+ {
2855
+ "epoch": 3.13,
2856
+ "learning_rate": 0.00019373136712095402,
2857
+ "loss": 2.0031,
2858
+ "step": 235500
2859
+ },
2860
+ {
2861
+ "epoch": 3.14,
2862
+ "learning_rate": 0.00019371805792163546,
2863
+ "loss": 1.9253,
2864
+ "step": 236000
2865
+ },
2866
+ {
2867
+ "epoch": 3.15,
2868
+ "learning_rate": 0.00019370474872231687,
2869
+ "loss": 1.8825,
2870
+ "step": 236500
2871
+ },
2872
+ {
2873
+ "epoch": 3.15,
2874
+ "learning_rate": 0.0001936914395229983,
2875
+ "loss": 1.9689,
2876
+ "step": 237000
2877
+ },
2878
+ {
2879
+ "epoch": 3.16,
2880
+ "learning_rate": 0.00019367813032367974,
2881
+ "loss": 2.0113,
2882
+ "step": 237500
2883
+ },
2884
+ {
2885
+ "epoch": 3.17,
2886
+ "learning_rate": 0.00019366482112436118,
2887
+ "loss": 2.0767,
2888
+ "step": 238000
2889
+ },
2890
+ {
2891
+ "epoch": 3.17,
2892
+ "learning_rate": 0.00019365151192504262,
2893
+ "loss": 2.0196,
2894
+ "step": 238500
2895
+ },
2896
+ {
2897
+ "epoch": 3.18,
2898
+ "learning_rate": 0.00019363820272572403,
2899
+ "loss": 1.9681,
2900
+ "step": 239000
2901
+ },
2902
+ {
2903
+ "epoch": 3.19,
2904
+ "learning_rate": 0.00019362489352640547,
2905
+ "loss": 2.0794,
2906
+ "step": 239500
2907
+ },
2908
+ {
2909
+ "epoch": 3.19,
2910
+ "learning_rate": 0.00019361158432708688,
2911
+ "loss": 1.9347,
2912
+ "step": 240000
2913
+ },
2914
+ {
2915
+ "epoch": 3.2,
2916
+ "learning_rate": 0.00019359827512776832,
2917
+ "loss": 1.8842,
2918
+ "step": 240500
2919
+ },
2920
+ {
2921
+ "epoch": 3.21,
2922
+ "learning_rate": 0.00019358496592844976,
2923
+ "loss": 2.1145,
2924
+ "step": 241000
2925
+ },
2926
+ {
2927
+ "epoch": 3.21,
2928
+ "learning_rate": 0.00019357165672913117,
2929
+ "loss": 2.0187,
2930
+ "step": 241500
2931
+ },
2932
+ {
2933
+ "epoch": 3.22,
2934
+ "learning_rate": 0.0001935583475298126,
2935
+ "loss": 1.9615,
2936
+ "step": 242000
2937
+ },
2938
+ {
2939
+ "epoch": 3.23,
2940
+ "learning_rate": 0.00019354503833049404,
2941
+ "loss": 2.0395,
2942
+ "step": 242500
2943
+ },
2944
+ {
2945
+ "epoch": 3.23,
2946
+ "learning_rate": 0.00019353172913117548,
2947
+ "loss": 1.9914,
2948
+ "step": 243000
2949
+ },
2950
+ {
2951
+ "epoch": 3.24,
2952
+ "learning_rate": 0.00019351841993185692,
2953
+ "loss": 1.9202,
2954
+ "step": 243500
2955
+ },
2956
+ {
2957
+ "epoch": 3.25,
2958
+ "learning_rate": 0.00019350511073253833,
2959
+ "loss": 1.9856,
2960
+ "step": 244000
2961
+ },
2962
+ {
2963
+ "epoch": 3.25,
2964
+ "learning_rate": 0.00019349180153321977,
2965
+ "loss": 1.9678,
2966
+ "step": 244500
2967
+ },
2968
+ {
2969
+ "epoch": 3.26,
2970
+ "learning_rate": 0.0001934784923339012,
2971
+ "loss": 1.9491,
2972
+ "step": 245000
2973
+ },
2974
+ {
2975
+ "epoch": 3.27,
2976
+ "learning_rate": 0.00019346518313458264,
2977
+ "loss": 2.048,
2978
+ "step": 245500
2979
+ },
2980
+ {
2981
+ "epoch": 3.27,
2982
+ "learning_rate": 0.00019345187393526408,
2983
+ "loss": 2.0294,
2984
+ "step": 246000
2985
+ },
2986
+ {
2987
+ "epoch": 3.28,
2988
+ "learning_rate": 0.0001934385647359455,
2989
+ "loss": 1.9502,
2990
+ "step": 246500
2991
+ },
2992
+ {
2993
+ "epoch": 3.29,
2994
+ "learning_rate": 0.0001934252555366269,
2995
+ "loss": 1.9172,
2996
+ "step": 247000
2997
+ },
2998
+ {
2999
+ "epoch": 3.29,
3000
+ "learning_rate": 0.00019341194633730834,
3001
+ "loss": 2.0995,
3002
+ "step": 247500
3003
+ },
3004
+ {
3005
+ "epoch": 3.3,
3006
+ "learning_rate": 0.00019339863713798978,
3007
+ "loss": 1.8991,
3008
+ "step": 248000
3009
+ },
3010
+ {
3011
+ "epoch": 3.31,
3012
+ "learning_rate": 0.00019338532793867122,
3013
+ "loss": 1.9673,
3014
+ "step": 248500
3015
+ },
3016
+ {
3017
+ "epoch": 3.31,
3018
+ "learning_rate": 0.00019337201873935266,
3019
+ "loss": 1.8967,
3020
+ "step": 249000
3021
+ },
3022
+ {
3023
+ "epoch": 3.32,
3024
+ "learning_rate": 0.00019335870954003407,
3025
+ "loss": 1.9235,
3026
+ "step": 249500
3027
+ },
3028
+ {
3029
+ "epoch": 3.33,
3030
+ "learning_rate": 0.0001933454003407155,
3031
+ "loss": 2.0222,
3032
+ "step": 250000
3033
+ },
3034
+ {
3035
+ "epoch": 3.33,
3036
+ "learning_rate": 0.00019333209114139695,
3037
+ "loss": 1.9349,
3038
+ "step": 250500
3039
+ },
3040
+ {
3041
+ "epoch": 3.34,
3042
+ "learning_rate": 0.00019331878194207838,
3043
+ "loss": 1.9891,
3044
+ "step": 251000
3045
+ },
3046
+ {
3047
+ "epoch": 3.35,
3048
+ "learning_rate": 0.0001933054727427598,
3049
+ "loss": 2.0221,
3050
+ "step": 251500
3051
+ },
3052
+ {
3053
+ "epoch": 3.35,
3054
+ "learning_rate": 0.00019329216354344123,
3055
+ "loss": 1.92,
3056
+ "step": 252000
3057
+ },
3058
+ {
3059
+ "epoch": 3.36,
3060
+ "learning_rate": 0.00019327885434412267,
3061
+ "loss": 1.9894,
3062
+ "step": 252500
3063
+ },
3064
+ {
3065
+ "epoch": 3.37,
3066
+ "learning_rate": 0.0001932655451448041,
3067
+ "loss": 2.1004,
3068
+ "step": 253000
3069
+ },
3070
+ {
3071
+ "epoch": 3.37,
3072
+ "learning_rate": 0.00019325223594548555,
3073
+ "loss": 2.0359,
3074
+ "step": 253500
3075
+ },
3076
+ {
3077
+ "epoch": 3.38,
3078
+ "learning_rate": 0.00019323892674616696,
3079
+ "loss": 1.9138,
3080
+ "step": 254000
3081
+ },
3082
+ {
3083
+ "epoch": 3.39,
3084
+ "learning_rate": 0.00019322561754684837,
3085
+ "loss": 2.0319,
3086
+ "step": 254500
3087
+ },
3088
+ {
3089
+ "epoch": 3.39,
3090
+ "learning_rate": 0.0001932123083475298,
3091
+ "loss": 1.9884,
3092
+ "step": 255000
3093
+ },
3094
+ {
3095
+ "epoch": 3.4,
3096
+ "learning_rate": 0.00019319899914821125,
3097
+ "loss": 1.9501,
3098
+ "step": 255500
3099
+ },
3100
+ {
3101
+ "epoch": 3.41,
3102
+ "learning_rate": 0.00019318568994889268,
3103
+ "loss": 1.9661,
3104
+ "step": 256000
3105
+ },
3106
+ {
3107
+ "epoch": 3.41,
3108
+ "learning_rate": 0.00019317238074957412,
3109
+ "loss": 2.1242,
3110
+ "step": 256500
3111
+ },
3112
+ {
3113
+ "epoch": 3.42,
3114
+ "learning_rate": 0.00019315907155025553,
3115
+ "loss": 1.9356,
3116
+ "step": 257000
3117
+ },
3118
+ {
3119
+ "epoch": 3.43,
3120
+ "learning_rate": 0.00019314576235093697,
3121
+ "loss": 2.0207,
3122
+ "step": 257500
3123
+ },
3124
+ {
3125
+ "epoch": 3.43,
3126
+ "learning_rate": 0.0001931324531516184,
3127
+ "loss": 2.0019,
3128
+ "step": 258000
3129
+ },
3130
+ {
3131
+ "epoch": 3.44,
3132
+ "learning_rate": 0.00019311914395229985,
3133
+ "loss": 1.9281,
3134
+ "step": 258500
3135
+ },
3136
+ {
3137
+ "epoch": 3.45,
3138
+ "learning_rate": 0.00019310583475298129,
3139
+ "loss": 2.0126,
3140
+ "step": 259000
3141
+ },
3142
+ {
3143
+ "epoch": 3.45,
3144
+ "learning_rate": 0.0001930925255536627,
3145
+ "loss": 1.9973,
3146
+ "step": 259500
3147
+ },
3148
+ {
3149
+ "epoch": 3.46,
3150
+ "learning_rate": 0.00019307921635434413,
3151
+ "loss": 2.0036,
3152
+ "step": 260000
3153
+ },
3154
+ {
3155
+ "epoch": 3.47,
3156
+ "learning_rate": 0.00019306590715502557,
3157
+ "loss": 1.9307,
3158
+ "step": 260500
3159
+ },
3160
+ {
3161
+ "epoch": 3.47,
3162
+ "learning_rate": 0.000193052597955707,
3163
+ "loss": 2.0548,
3164
+ "step": 261000
3165
+ },
3166
+ {
3167
+ "epoch": 3.48,
3168
+ "learning_rate": 0.00019303928875638842,
3169
+ "loss": 1.9384,
3170
+ "step": 261500
3171
+ },
3172
+ {
3173
+ "epoch": 3.49,
3174
+ "learning_rate": 0.00019302597955706986,
3175
+ "loss": 2.0036,
3176
+ "step": 262000
3177
+ },
3178
+ {
3179
+ "epoch": 3.49,
3180
+ "learning_rate": 0.00019301267035775127,
3181
+ "loss": 1.9239,
3182
+ "step": 262500
3183
+ },
3184
+ {
3185
+ "epoch": 3.5,
3186
+ "learning_rate": 0.0001929993611584327,
3187
+ "loss": 1.9539,
3188
+ "step": 263000
3189
+ },
3190
+ {
3191
+ "epoch": 3.51,
3192
+ "learning_rate": 0.00019298605195911415,
3193
+ "loss": 2.0679,
3194
+ "step": 263500
3195
+ },
3196
+ {
3197
+ "epoch": 3.51,
3198
+ "learning_rate": 0.00019297274275979559,
3199
+ "loss": 1.9274,
3200
+ "step": 264000
3201
+ },
3202
+ {
3203
+ "epoch": 3.52,
3204
+ "learning_rate": 0.000192959433560477,
3205
+ "loss": 1.935,
3206
+ "step": 264500
3207
+ },
3208
+ {
3209
+ "epoch": 3.53,
3210
+ "learning_rate": 0.00019294612436115843,
3211
+ "loss": 2.1159,
3212
+ "step": 265000
3213
+ },
3214
+ {
3215
+ "epoch": 3.53,
3216
+ "learning_rate": 0.00019293281516183987,
3217
+ "loss": 2.0521,
3218
+ "step": 265500
3219
+ },
3220
+ {
3221
+ "epoch": 3.54,
3222
+ "learning_rate": 0.0001929195059625213,
3223
+ "loss": 2.1259,
3224
+ "step": 266000
3225
+ },
3226
+ {
3227
+ "epoch": 3.55,
3228
+ "learning_rate": 0.00019290619676320275,
3229
+ "loss": 2.0283,
3230
+ "step": 266500
3231
+ },
3232
+ {
3233
+ "epoch": 3.55,
3234
+ "learning_rate": 0.00019289288756388416,
3235
+ "loss": 2.0267,
3236
+ "step": 267000
3237
+ },
3238
+ {
3239
+ "epoch": 3.56,
3240
+ "learning_rate": 0.0001928795783645656,
3241
+ "loss": 1.9899,
3242
+ "step": 267500
3243
+ },
3244
+ {
3245
+ "epoch": 3.57,
3246
+ "learning_rate": 0.00019286626916524704,
3247
+ "loss": 1.9344,
3248
+ "step": 268000
3249
+ },
3250
+ {
3251
+ "epoch": 3.57,
3252
+ "learning_rate": 0.00019285295996592845,
3253
+ "loss": 2.1067,
3254
+ "step": 268500
3255
+ },
3256
+ {
3257
+ "epoch": 3.58,
3258
+ "learning_rate": 0.00019283965076660989,
3259
+ "loss": 2.0983,
3260
+ "step": 269000
3261
+ },
3262
+ {
3263
+ "epoch": 3.59,
3264
+ "learning_rate": 0.00019282634156729132,
3265
+ "loss": 2.0504,
3266
+ "step": 269500
3267
+ },
3268
+ {
3269
+ "epoch": 3.59,
3270
+ "learning_rate": 0.00019281303236797273,
3271
+ "loss": 1.9778,
3272
+ "step": 270000
3273
+ },
3274
+ {
3275
+ "epoch": 3.6,
3276
+ "learning_rate": 0.00019279972316865417,
3277
+ "loss": 2.0056,
3278
+ "step": 270500
3279
+ },
3280
+ {
3281
+ "epoch": 3.61,
3282
+ "learning_rate": 0.0001927864139693356,
3283
+ "loss": 1.9893,
3284
+ "step": 271000
3285
+ },
3286
+ {
3287
+ "epoch": 3.61,
3288
+ "learning_rate": 0.00019277310477001705,
3289
+ "loss": 1.9795,
3290
+ "step": 271500
3291
+ },
3292
+ {
3293
+ "epoch": 3.62,
3294
+ "learning_rate": 0.0001927597955706985,
3295
+ "loss": 1.9398,
3296
+ "step": 272000
3297
+ },
3298
+ {
3299
+ "epoch": 3.63,
3300
+ "learning_rate": 0.0001927464863713799,
3301
+ "loss": 2.0212,
3302
+ "step": 272500
3303
+ },
3304
+ {
3305
+ "epoch": 3.63,
3306
+ "learning_rate": 0.00019273317717206134,
3307
+ "loss": 1.9342,
3308
+ "step": 273000
3309
+ },
3310
+ {
3311
+ "epoch": 3.64,
3312
+ "learning_rate": 0.00019271986797274277,
3313
+ "loss": 1.9867,
3314
+ "step": 273500
3315
+ },
3316
+ {
3317
+ "epoch": 3.65,
3318
+ "learning_rate": 0.0001927065587734242,
3319
+ "loss": 2.0927,
3320
+ "step": 274000
3321
+ },
3322
+ {
3323
+ "epoch": 3.65,
3324
+ "learning_rate": 0.00019269324957410565,
3325
+ "loss": 2.0769,
3326
+ "step": 274500
3327
+ },
3328
+ {
3329
+ "epoch": 3.66,
3330
+ "learning_rate": 0.00019267994037478706,
3331
+ "loss": 1.9088,
3332
+ "step": 275000
3333
+ },
3334
+ {
3335
+ "epoch": 3.67,
3336
+ "learning_rate": 0.0001926666311754685,
3337
+ "loss": 2.068,
3338
+ "step": 275500
3339
+ },
3340
+ {
3341
+ "epoch": 3.67,
3342
+ "learning_rate": 0.0001926533219761499,
3343
+ "loss": 1.9976,
3344
+ "step": 276000
3345
+ },
3346
+ {
3347
+ "epoch": 3.68,
3348
+ "learning_rate": 0.00019264001277683135,
3349
+ "loss": 1.9657,
3350
+ "step": 276500
3351
+ },
3352
+ {
3353
+ "epoch": 3.69,
3354
+ "learning_rate": 0.0001926267035775128,
3355
+ "loss": 2.0356,
3356
+ "step": 277000
3357
+ },
3358
+ {
3359
+ "epoch": 3.69,
3360
+ "learning_rate": 0.0001926133943781942,
3361
+ "loss": 1.9525,
3362
+ "step": 277500
3363
+ },
3364
+ {
3365
+ "epoch": 3.7,
3366
+ "learning_rate": 0.00019260008517887564,
3367
+ "loss": 1.9918,
3368
+ "step": 278000
3369
+ },
3370
+ {
3371
+ "epoch": 3.71,
3372
+ "learning_rate": 0.00019258677597955707,
3373
+ "loss": 1.9586,
3374
+ "step": 278500
3375
+ },
3376
+ {
3377
+ "epoch": 3.71,
3378
+ "learning_rate": 0.0001925734667802385,
3379
+ "loss": 1.9425,
3380
+ "step": 279000
3381
+ },
3382
+ {
3383
+ "epoch": 3.72,
3384
+ "learning_rate": 0.00019256015758091995,
3385
+ "loss": 1.9502,
3386
+ "step": 279500
3387
+ },
3388
+ {
3389
+ "epoch": 3.73,
3390
+ "learning_rate": 0.00019254684838160136,
3391
+ "loss": 1.8562,
3392
+ "step": 280000
3393
+ },
3394
+ {
3395
+ "epoch": 3.73,
3396
+ "learning_rate": 0.0001925335391822828,
3397
+ "loss": 2.0119,
3398
+ "step": 280500
3399
+ },
3400
+ {
3401
+ "epoch": 3.74,
3402
+ "learning_rate": 0.00019252022998296424,
3403
+ "loss": 2.0455,
3404
+ "step": 281000
3405
+ },
3406
+ {
3407
+ "epoch": 3.75,
3408
+ "learning_rate": 0.00019250692078364568,
3409
+ "loss": 2.039,
3410
+ "step": 281500
3411
+ },
3412
+ {
3413
+ "epoch": 3.75,
3414
+ "learning_rate": 0.00019249361158432711,
3415
+ "loss": 2.0412,
3416
+ "step": 282000
3417
+ },
3418
+ {
3419
+ "epoch": 3.76,
3420
+ "learning_rate": 0.00019248030238500853,
3421
+ "loss": 2.1478,
3422
+ "step": 282500
3423
+ },
3424
+ {
3425
+ "epoch": 3.77,
3426
+ "learning_rate": 0.00019246699318568994,
3427
+ "loss": 1.941,
3428
+ "step": 283000
3429
+ },
3430
+ {
3431
+ "epoch": 3.77,
3432
+ "learning_rate": 0.00019245368398637138,
3433
+ "loss": 1.9749,
3434
+ "step": 283500
3435
+ },
3436
+ {
3437
+ "epoch": 3.78,
3438
+ "learning_rate": 0.0001924403747870528,
3439
+ "loss": 2.0763,
3440
+ "step": 284000
3441
+ },
3442
+ {
3443
+ "epoch": 3.79,
3444
+ "learning_rate": 0.00019242706558773425,
3445
+ "loss": 2.0318,
3446
+ "step": 284500
3447
+ },
3448
+ {
3449
+ "epoch": 3.79,
3450
+ "learning_rate": 0.0001924137563884157,
3451
+ "loss": 2.0111,
3452
+ "step": 285000
3453
+ },
3454
+ {
3455
+ "epoch": 3.8,
3456
+ "learning_rate": 0.0001924004471890971,
3457
+ "loss": 1.9364,
3458
+ "step": 285500
3459
+ },
3460
+ {
3461
+ "epoch": 3.81,
3462
+ "learning_rate": 0.00019238713798977854,
3463
+ "loss": 2.0628,
3464
+ "step": 286000
3465
+ },
3466
+ {
3467
+ "epoch": 3.81,
3468
+ "learning_rate": 0.00019237382879045998,
3469
+ "loss": 2.1622,
3470
+ "step": 286500
3471
+ },
3472
+ {
3473
+ "epoch": 3.82,
3474
+ "learning_rate": 0.00019236051959114142,
3475
+ "loss": 2.049,
3476
+ "step": 287000
3477
+ },
3478
+ {
3479
+ "epoch": 3.83,
3480
+ "learning_rate": 0.00019234721039182285,
3481
+ "loss": 2.0455,
3482
+ "step": 287500
3483
+ },
3484
+ {
3485
+ "epoch": 3.83,
3486
+ "learning_rate": 0.00019233390119250426,
3487
+ "loss": 1.991,
3488
+ "step": 288000
3489
+ },
3490
+ {
3491
+ "epoch": 3.84,
3492
+ "learning_rate": 0.0001923205919931857,
3493
+ "loss": 1.9896,
3494
+ "step": 288500
3495
+ },
3496
+ {
3497
+ "epoch": 3.85,
3498
+ "learning_rate": 0.00019230728279386714,
3499
+ "loss": 2.1045,
3500
+ "step": 289000
3501
+ },
3502
+ {
3503
+ "epoch": 3.85,
3504
+ "learning_rate": 0.00019229397359454858,
3505
+ "loss": 2.0259,
3506
+ "step": 289500
3507
+ },
3508
+ {
3509
+ "epoch": 3.86,
3510
+ "learning_rate": 0.00019228066439523,
3511
+ "loss": 2.0056,
3512
+ "step": 290000
3513
+ },
3514
+ {
3515
+ "epoch": 3.87,
3516
+ "learning_rate": 0.0001922673551959114,
3517
+ "loss": 1.9663,
3518
+ "step": 290500
3519
+ },
3520
+ {
3521
+ "epoch": 3.87,
3522
+ "learning_rate": 0.00019225404599659284,
3523
+ "loss": 2.0817,
3524
+ "step": 291000
3525
+ },
3526
+ {
3527
+ "epoch": 3.88,
3528
+ "learning_rate": 0.00019224073679727428,
3529
+ "loss": 2.1459,
3530
+ "step": 291500
3531
+ },
3532
+ {
3533
+ "epoch": 3.89,
3534
+ "learning_rate": 0.00019222742759795572,
3535
+ "loss": 2.0219,
3536
+ "step": 292000
3537
+ },
3538
+ {
3539
+ "epoch": 3.89,
3540
+ "learning_rate": 0.00019221411839863715,
3541
+ "loss": 2.0752,
3542
+ "step": 292500
3543
+ },
3544
+ {
3545
+ "epoch": 3.9,
3546
+ "learning_rate": 0.00019220080919931856,
3547
+ "loss": 2.0652,
3548
+ "step": 293000
3549
+ },
3550
+ {
3551
+ "epoch": 3.91,
3552
+ "learning_rate": 0.0001921875,
3553
+ "loss": 2.0201,
3554
+ "step": 293500
3555
+ },
3556
+ {
3557
+ "epoch": 3.91,
3558
+ "learning_rate": 0.00019217419080068144,
3559
+ "loss": 2.018,
3560
+ "step": 294000
3561
+ },
3562
+ {
3563
+ "epoch": 3.92,
3564
+ "learning_rate": 0.00019216088160136288,
3565
+ "loss": 2.0404,
3566
+ "step": 294500
3567
+ },
3568
+ {
3569
+ "epoch": 3.93,
3570
+ "learning_rate": 0.00019214757240204432,
3571
+ "loss": 2.0208,
3572
+ "step": 295000
3573
+ },
3574
+ {
3575
+ "epoch": 3.93,
3576
+ "learning_rate": 0.00019213426320272573,
3577
+ "loss": 1.9643,
3578
+ "step": 295500
3579
+ },
3580
+ {
3581
+ "epoch": 3.94,
3582
+ "learning_rate": 0.00019212095400340717,
3583
+ "loss": 2.0174,
3584
+ "step": 296000
3585
+ },
3586
+ {
3587
+ "epoch": 3.95,
3588
+ "learning_rate": 0.0001921076448040886,
3589
+ "loss": 1.991,
3590
+ "step": 296500
3591
+ },
3592
+ {
3593
+ "epoch": 3.95,
3594
+ "learning_rate": 0.00019209433560477004,
3595
+ "loss": 2.0344,
3596
+ "step": 297000
3597
+ },
3598
+ {
3599
+ "epoch": 3.96,
3600
+ "learning_rate": 0.00019208102640545145,
3601
+ "loss": 2.1951,
3602
+ "step": 297500
3603
+ },
3604
+ {
3605
+ "epoch": 3.97,
3606
+ "learning_rate": 0.0001920677172061329,
3607
+ "loss": 2.0148,
3608
+ "step": 298000
3609
+ },
3610
+ {
3611
+ "epoch": 3.97,
3612
+ "learning_rate": 0.0001920544080068143,
3613
+ "loss": 1.9864,
3614
+ "step": 298500
3615
+ },
3616
+ {
3617
+ "epoch": 3.98,
3618
+ "learning_rate": 0.00019204109880749574,
3619
+ "loss": 2.0757,
3620
+ "step": 299000
3621
+ },
3622
+ {
3623
+ "epoch": 3.99,
3624
+ "learning_rate": 0.00019202778960817718,
3625
+ "loss": 2.0382,
3626
+ "step": 299500
3627
+ },
3628
+ {
3629
+ "epoch": 3.99,
3630
+ "learning_rate": 0.00019201448040885862,
3631
+ "loss": 2.0325,
3632
+ "step": 300000
3633
+ },
3634
+ {
3635
+ "epoch": 4.0,
3636
+ "learning_rate": 0.00019200117120954006,
3637
+ "loss": 2.0484,
3638
+ "step": 300500
3639
+ },
3640
+ {
3641
+ "epoch": 4.0,
3642
+ "eval_loss": 1.9609955549240112,
3643
+ "eval_runtime": 152.8744,
3644
+ "eval_samples_per_second": 61.443,
3645
+ "eval_steps_per_second": 61.443,
3646
+ "step": 300544
3647
+ },
3648
+ {
3649
+ "epoch": 4.01,
3650
+ "learning_rate": 0.00019198786201022147,
3651
+ "loss": 1.835,
3652
+ "step": 301000
3653
+ },
3654
+ {
3655
+ "epoch": 4.01,
3656
+ "learning_rate": 0.0001919745528109029,
3657
+ "loss": 1.8727,
3658
+ "step": 301500
3659
+ },
3660
+ {
3661
+ "epoch": 4.02,
3662
+ "learning_rate": 0.00019196124361158434,
3663
+ "loss": 1.7339,
3664
+ "step": 302000
3665
+ },
3666
+ {
3667
+ "epoch": 4.03,
3668
+ "learning_rate": 0.00019194793441226578,
3669
+ "loss": 1.8193,
3670
+ "step": 302500
3671
+ },
3672
+ {
3673
+ "epoch": 4.03,
3674
+ "learning_rate": 0.00019193462521294722,
3675
+ "loss": 1.8356,
3676
+ "step": 303000
3677
+ },
3678
+ {
3679
+ "epoch": 4.04,
3680
+ "learning_rate": 0.00019192131601362863,
3681
+ "loss": 1.8941,
3682
+ "step": 303500
3683
+ },
3684
+ {
3685
+ "epoch": 4.05,
3686
+ "learning_rate": 0.00019190800681431007,
3687
+ "loss": 1.8211,
3688
+ "step": 304000
3689
+ },
3690
+ {
3691
+ "epoch": 4.05,
3692
+ "learning_rate": 0.00019189469761499148,
3693
+ "loss": 1.9222,
3694
+ "step": 304500
3695
+ },
3696
+ {
3697
+ "epoch": 4.06,
3698
+ "learning_rate": 0.00019188138841567292,
3699
+ "loss": 1.7926,
3700
+ "step": 305000
3701
+ },
3702
+ {
3703
+ "epoch": 4.07,
3704
+ "learning_rate": 0.00019186807921635436,
3705
+ "loss": 1.8561,
3706
+ "step": 305500
3707
+ },
3708
+ {
3709
+ "epoch": 4.07,
3710
+ "learning_rate": 0.00019185477001703577,
3711
+ "loss": 1.7667,
3712
+ "step": 306000
3713
+ },
3714
+ {
3715
+ "epoch": 4.08,
3716
+ "learning_rate": 0.0001918414608177172,
3717
+ "loss": 1.803,
3718
+ "step": 306500
3719
+ },
3720
+ {
3721
+ "epoch": 4.09,
3722
+ "learning_rate": 0.00019182815161839864,
3723
+ "loss": 1.7807,
3724
+ "step": 307000
3725
+ },
3726
+ {
3727
+ "epoch": 4.09,
3728
+ "learning_rate": 0.00019181484241908008,
3729
+ "loss": 1.8113,
3730
+ "step": 307500
3731
+ },
3732
+ {
3733
+ "epoch": 4.1,
3734
+ "learning_rate": 0.00019180153321976152,
3735
+ "loss": 1.8102,
3736
+ "step": 308000
3737
+ },
3738
+ {
3739
+ "epoch": 4.11,
3740
+ "learning_rate": 0.00019178822402044293,
3741
+ "loss": 1.8823,
3742
+ "step": 308500
3743
+ },
3744
+ {
3745
+ "epoch": 4.11,
3746
+ "learning_rate": 0.00019177491482112437,
3747
+ "loss": 1.7312,
3748
+ "step": 309000
3749
+ },
3750
+ {
3751
+ "epoch": 4.12,
3752
+ "learning_rate": 0.0001917616056218058,
3753
+ "loss": 1.8165,
3754
+ "step": 309500
3755
+ },
3756
+ {
3757
+ "epoch": 4.13,
3758
+ "learning_rate": 0.00019174829642248724,
3759
+ "loss": 1.8188,
3760
+ "step": 310000
3761
+ },
3762
+ {
3763
+ "epoch": 4.13,
3764
+ "learning_rate": 0.00019173498722316868,
3765
+ "loss": 1.9123,
3766
+ "step": 310500
3767
+ },
3768
+ {
3769
+ "epoch": 4.14,
3770
+ "learning_rate": 0.0001917216780238501,
3771
+ "loss": 1.8474,
3772
+ "step": 311000
3773
+ },
3774
+ {
3775
+ "epoch": 4.15,
3776
+ "learning_rate": 0.0001917083688245315,
3777
+ "loss": 1.8434,
3778
+ "step": 311500
3779
+ },
3780
+ {
3781
+ "epoch": 4.15,
3782
+ "learning_rate": 0.00019169505962521294,
3783
+ "loss": 1.832,
3784
+ "step": 312000
3785
+ },
3786
+ {
3787
+ "epoch": 4.16,
3788
+ "learning_rate": 0.00019168175042589438,
3789
+ "loss": 1.7631,
3790
+ "step": 312500
3791
+ },
3792
+ {
3793
+ "epoch": 4.17,
3794
+ "learning_rate": 0.00019166844122657582,
3795
+ "loss": 1.8939,
3796
+ "step": 313000
3797
+ },
3798
+ {
3799
+ "epoch": 4.17,
3800
+ "learning_rate": 0.00019165513202725726,
3801
+ "loss": 1.8447,
3802
+ "step": 313500
3803
+ },
3804
+ {
3805
+ "epoch": 4.18,
3806
+ "learning_rate": 0.00019164182282793867,
3807
+ "loss": 1.8582,
3808
+ "step": 314000
3809
+ },
3810
+ {
3811
+ "epoch": 4.19,
3812
+ "learning_rate": 0.0001916285136286201,
3813
+ "loss": 1.8515,
3814
+ "step": 314500
3815
+ },
3816
+ {
3817
+ "epoch": 4.19,
3818
+ "learning_rate": 0.00019161520442930154,
3819
+ "loss": 1.7869,
3820
+ "step": 315000
3821
+ },
3822
+ {
3823
+ "epoch": 4.2,
3824
+ "learning_rate": 0.00019160189522998298,
3825
+ "loss": 1.9908,
3826
+ "step": 315500
3827
+ },
3828
+ {
3829
+ "epoch": 4.21,
3830
+ "learning_rate": 0.0001915885860306644,
3831
+ "loss": 1.8234,
3832
+ "step": 316000
3833
+ },
3834
+ {
3835
+ "epoch": 4.21,
3836
+ "learning_rate": 0.00019157527683134583,
3837
+ "loss": 1.8885,
3838
+ "step": 316500
3839
+ },
3840
+ {
3841
+ "epoch": 4.22,
3842
+ "learning_rate": 0.00019156196763202727,
3843
+ "loss": 1.8792,
3844
+ "step": 317000
3845
+ },
3846
+ {
3847
+ "epoch": 4.23,
3848
+ "learning_rate": 0.0001915486584327087,
3849
+ "loss": 1.9137,
3850
+ "step": 317500
3851
+ },
3852
+ {
3853
+ "epoch": 4.23,
3854
+ "learning_rate": 0.00019153534923339015,
3855
+ "loss": 1.8785,
3856
+ "step": 318000
3857
+ },
3858
+ {
3859
+ "epoch": 4.24,
3860
+ "learning_rate": 0.00019152204003407156,
3861
+ "loss": 1.8575,
3862
+ "step": 318500
3863
+ },
3864
+ {
3865
+ "epoch": 4.25,
3866
+ "learning_rate": 0.00019150873083475297,
3867
+ "loss": 1.8303,
3868
+ "step": 319000
3869
+ },
3870
+ {
3871
+ "epoch": 4.25,
3872
+ "learning_rate": 0.0001914954216354344,
3873
+ "loss": 1.7941,
3874
+ "step": 319500
3875
+ },
3876
+ {
3877
+ "epoch": 4.26,
3878
+ "learning_rate": 0.00019148211243611585,
3879
+ "loss": 1.9304,
3880
+ "step": 320000
3881
+ },
3882
+ {
3883
+ "epoch": 4.27,
3884
+ "learning_rate": 0.00019146880323679728,
3885
+ "loss": 1.8178,
3886
+ "step": 320500
3887
+ },
3888
+ {
3889
+ "epoch": 4.27,
3890
+ "learning_rate": 0.00019145549403747872,
3891
+ "loss": 1.9175,
3892
+ "step": 321000
3893
+ },
3894
+ {
3895
+ "epoch": 4.28,
3896
+ "learning_rate": 0.00019144218483816013,
3897
+ "loss": 1.8657,
3898
+ "step": 321500
3899
+ },
3900
+ {
3901
+ "epoch": 4.29,
3902
+ "learning_rate": 0.00019142887563884157,
3903
+ "loss": 1.821,
3904
+ "step": 322000
3905
+ },
3906
+ {
3907
+ "epoch": 4.29,
3908
+ "learning_rate": 0.000191415566439523,
3909
+ "loss": 1.7716,
3910
+ "step": 322500
3911
+ },
3912
+ {
3913
+ "epoch": 4.3,
3914
+ "learning_rate": 0.00019140225724020445,
3915
+ "loss": 1.8235,
3916
+ "step": 323000
3917
+ },
3918
+ {
3919
+ "epoch": 4.31,
3920
+ "learning_rate": 0.00019138894804088589,
3921
+ "loss": 1.8991,
3922
+ "step": 323500
3923
+ },
3924
+ {
3925
+ "epoch": 4.31,
3926
+ "learning_rate": 0.0001913756388415673,
3927
+ "loss": 1.9076,
3928
+ "step": 324000
3929
+ },
3930
+ {
3931
+ "epoch": 4.32,
3932
+ "learning_rate": 0.00019136232964224873,
3933
+ "loss": 1.9073,
3934
+ "step": 324500
3935
+ },
3936
+ {
3937
+ "epoch": 4.33,
3938
+ "learning_rate": 0.00019134902044293017,
3939
+ "loss": 2.0163,
3940
+ "step": 325000
3941
+ },
3942
+ {
3943
+ "epoch": 4.33,
3944
+ "learning_rate": 0.0001913357112436116,
3945
+ "loss": 1.8352,
3946
+ "step": 325500
3947
+ },
3948
+ {
3949
+ "epoch": 4.34,
3950
+ "learning_rate": 0.00019132240204429302,
3951
+ "loss": 1.8498,
3952
+ "step": 326000
3953
+ },
3954
+ {
3955
+ "epoch": 4.35,
3956
+ "learning_rate": 0.00019130909284497446,
3957
+ "loss": 1.9367,
3958
+ "step": 326500
3959
+ },
3960
+ {
3961
+ "epoch": 4.35,
3962
+ "learning_rate": 0.00019129578364565587,
3963
+ "loss": 1.847,
3964
+ "step": 327000
3965
+ },
3966
+ {
3967
+ "epoch": 4.36,
3968
+ "learning_rate": 0.0001912824744463373,
3969
+ "loss": 1.8725,
3970
+ "step": 327500
3971
+ },
3972
+ {
3973
+ "epoch": 4.37,
3974
+ "learning_rate": 0.00019126916524701875,
3975
+ "loss": 1.9263,
3976
+ "step": 328000
3977
+ },
3978
+ {
3979
+ "epoch": 4.37,
3980
+ "learning_rate": 0.00019125585604770019,
3981
+ "loss": 1.8871,
3982
+ "step": 328500
3983
+ },
3984
+ {
3985
+ "epoch": 4.38,
3986
+ "learning_rate": 0.0001912425468483816,
3987
+ "loss": 1.8387,
3988
+ "step": 329000
3989
+ },
3990
+ {
3991
+ "epoch": 4.39,
3992
+ "learning_rate": 0.00019122923764906303,
3993
+ "loss": 1.8359,
3994
+ "step": 329500
3995
+ },
3996
+ {
3997
+ "epoch": 4.39,
3998
+ "learning_rate": 0.00019121592844974447,
3999
+ "loss": 1.967,
4000
+ "step": 330000
4001
+ },
4002
+ {
4003
+ "epoch": 4.4,
4004
+ "learning_rate": 0.0001912026192504259,
4005
+ "loss": 1.9031,
4006
+ "step": 330500
4007
+ },
4008
+ {
4009
+ "epoch": 4.41,
4010
+ "learning_rate": 0.00019118931005110735,
4011
+ "loss": 1.9016,
4012
+ "step": 331000
4013
+ },
4014
+ {
4015
+ "epoch": 4.41,
4016
+ "learning_rate": 0.00019117600085178876,
4017
+ "loss": 1.8559,
4018
+ "step": 331500
4019
+ },
4020
+ {
4021
+ "epoch": 4.42,
4022
+ "learning_rate": 0.0001911626916524702,
4023
+ "loss": 1.8603,
4024
+ "step": 332000
4025
+ },
4026
+ {
4027
+ "epoch": 4.43,
4028
+ "learning_rate": 0.00019114938245315164,
4029
+ "loss": 1.9245,
4030
+ "step": 332500
4031
+ },
4032
+ {
4033
+ "epoch": 4.43,
4034
+ "learning_rate": 0.00019113607325383305,
4035
+ "loss": 1.8627,
4036
+ "step": 333000
4037
+ },
4038
+ {
4039
+ "epoch": 4.44,
4040
+ "learning_rate": 0.00019112276405451449,
4041
+ "loss": 1.8681,
4042
+ "step": 333500
4043
+ },
4044
+ {
4045
+ "epoch": 4.45,
4046
+ "learning_rate": 0.00019110945485519592,
4047
+ "loss": 1.9043,
4048
+ "step": 334000
4049
+ },
4050
+ {
4051
+ "epoch": 4.45,
4052
+ "learning_rate": 0.00019109614565587733,
4053
+ "loss": 1.8662,
4054
+ "step": 334500
4055
+ },
4056
+ {
4057
+ "epoch": 4.46,
4058
+ "learning_rate": 0.00019108283645655877,
4059
+ "loss": 1.9518,
4060
+ "step": 335000
4061
+ },
4062
+ {
4063
+ "epoch": 4.47,
4064
+ "learning_rate": 0.0001910695272572402,
4065
+ "loss": 1.8168,
4066
+ "step": 335500
4067
+ },
4068
+ {
4069
+ "epoch": 4.47,
4070
+ "learning_rate": 0.00019105621805792165,
4071
+ "loss": 1.9287,
4072
+ "step": 336000
4073
+ },
4074
+ {
4075
+ "epoch": 4.48,
4076
+ "learning_rate": 0.0001910429088586031,
4077
+ "loss": 1.8791,
4078
+ "step": 336500
4079
+ },
4080
+ {
4081
+ "epoch": 4.49,
4082
+ "learning_rate": 0.0001910295996592845,
4083
+ "loss": 1.8938,
4084
+ "step": 337000
4085
+ },
4086
+ {
4087
+ "epoch": 4.49,
4088
+ "learning_rate": 0.00019101629045996594,
4089
+ "loss": 1.9088,
4090
+ "step": 337500
4091
+ },
4092
+ {
4093
+ "epoch": 4.5,
4094
+ "learning_rate": 0.00019100298126064737,
4095
+ "loss": 1.9197,
4096
+ "step": 338000
4097
+ },
4098
+ {
4099
+ "epoch": 4.51,
4100
+ "learning_rate": 0.0001909896720613288,
4101
+ "loss": 2.082,
4102
+ "step": 338500
4103
+ },
4104
+ {
4105
+ "epoch": 4.51,
4106
+ "learning_rate": 0.00019097636286201025,
4107
+ "loss": 1.8672,
4108
+ "step": 339000
4109
+ },
4110
+ {
4111
+ "epoch": 4.52,
4112
+ "learning_rate": 0.00019096305366269166,
4113
+ "loss": 1.9084,
4114
+ "step": 339500
4115
+ },
4116
+ {
4117
+ "epoch": 4.53,
4118
+ "learning_rate": 0.0001909497444633731,
4119
+ "loss": 1.912,
4120
+ "step": 340000
4121
+ },
4122
+ {
4123
+ "epoch": 4.53,
4124
+ "learning_rate": 0.0001909364352640545,
4125
+ "loss": 1.9065,
4126
+ "step": 340500
4127
+ },
4128
+ {
4129
+ "epoch": 4.54,
4130
+ "learning_rate": 0.00019092312606473595,
4131
+ "loss": 1.819,
4132
+ "step": 341000
4133
+ },
4134
+ {
4135
+ "epoch": 4.55,
4136
+ "learning_rate": 0.0001909098168654174,
4137
+ "loss": 1.9697,
4138
+ "step": 341500
4139
+ },
4140
+ {
4141
+ "epoch": 4.55,
4142
+ "learning_rate": 0.0001908965076660988,
4143
+ "loss": 1.9895,
4144
+ "step": 342000
4145
+ },
4146
+ {
4147
+ "epoch": 4.56,
4148
+ "learning_rate": 0.00019088319846678024,
4149
+ "loss": 1.933,
4150
+ "step": 342500
4151
+ },
4152
+ {
4153
+ "epoch": 4.57,
4154
+ "learning_rate": 0.00019086988926746167,
4155
+ "loss": 1.8534,
4156
+ "step": 343000
4157
+ },
4158
+ {
4159
+ "epoch": 4.57,
4160
+ "learning_rate": 0.0001908565800681431,
4161
+ "loss": 1.948,
4162
+ "step": 343500
4163
+ },
4164
+ {
4165
+ "epoch": 4.58,
4166
+ "learning_rate": 0.00019084327086882455,
4167
+ "loss": 1.9291,
4168
+ "step": 344000
4169
+ },
4170
+ {
4171
+ "epoch": 4.59,
4172
+ "learning_rate": 0.00019082996166950596,
4173
+ "loss": 1.813,
4174
+ "step": 344500
4175
+ },
4176
+ {
4177
+ "epoch": 4.59,
4178
+ "learning_rate": 0.0001908166524701874,
4179
+ "loss": 1.9467,
4180
+ "step": 345000
4181
+ },
4182
+ {
4183
+ "epoch": 4.6,
4184
+ "learning_rate": 0.00019080334327086884,
4185
+ "loss": 1.9648,
4186
+ "step": 345500
4187
+ },
4188
+ {
4189
+ "epoch": 4.6,
4190
+ "learning_rate": 0.00019079003407155028,
4191
+ "loss": 1.9408,
4192
+ "step": 346000
4193
+ },
4194
+ {
4195
+ "epoch": 4.61,
4196
+ "learning_rate": 0.00019077672487223171,
4197
+ "loss": 2.0716,
4198
+ "step": 346500
4199
+ },
4200
+ {
4201
+ "epoch": 4.62,
4202
+ "learning_rate": 0.00019076341567291313,
4203
+ "loss": 1.8179,
4204
+ "step": 347000
4205
+ },
4206
+ {
4207
+ "epoch": 4.62,
4208
+ "learning_rate": 0.00019075010647359454,
4209
+ "loss": 1.9267,
4210
+ "step": 347500
4211
+ },
4212
+ {
4213
+ "epoch": 4.63,
4214
+ "learning_rate": 0.00019073679727427598,
4215
+ "loss": 1.8477,
4216
+ "step": 348000
4217
+ },
4218
+ {
4219
+ "epoch": 4.64,
4220
+ "learning_rate": 0.0001907234880749574,
4221
+ "loss": 1.8807,
4222
+ "step": 348500
4223
+ },
4224
+ {
4225
+ "epoch": 4.64,
4226
+ "learning_rate": 0.00019071017887563885,
4227
+ "loss": 1.9378,
4228
+ "step": 349000
4229
+ },
4230
+ {
4231
+ "epoch": 4.65,
4232
+ "learning_rate": 0.0001906968696763203,
4233
+ "loss": 1.9433,
4234
+ "step": 349500
4235
+ },
4236
+ {
4237
+ "epoch": 4.66,
4238
+ "learning_rate": 0.0001906835604770017,
4239
+ "loss": 1.9764,
4240
+ "step": 350000
4241
+ },
4242
+ {
4243
+ "epoch": 4.66,
4244
+ "learning_rate": 0.00019067025127768314,
4245
+ "loss": 1.9097,
4246
+ "step": 350500
4247
+ },
4248
+ {
4249
+ "epoch": 4.67,
4250
+ "learning_rate": 0.00019065694207836458,
4251
+ "loss": 1.8964,
4252
+ "step": 351000
4253
+ },
4254
+ {
4255
+ "epoch": 4.68,
4256
+ "learning_rate": 0.00019064363287904601,
4257
+ "loss": 1.8199,
4258
+ "step": 351500
4259
+ },
4260
+ {
4261
+ "epoch": 4.68,
4262
+ "learning_rate": 0.00019063032367972745,
4263
+ "loss": 1.8813,
4264
+ "step": 352000
4265
+ },
4266
+ {
4267
+ "epoch": 4.69,
4268
+ "learning_rate": 0.00019061701448040886,
4269
+ "loss": 1.9906,
4270
+ "step": 352500
4271
+ },
4272
+ {
4273
+ "epoch": 4.7,
4274
+ "learning_rate": 0.0001906037052810903,
4275
+ "loss": 1.9196,
4276
+ "step": 353000
4277
+ },
4278
+ {
4279
+ "epoch": 4.7,
4280
+ "learning_rate": 0.00019059039608177174,
4281
+ "loss": 2.023,
4282
+ "step": 353500
4283
+ },
4284
+ {
4285
+ "epoch": 4.71,
4286
+ "learning_rate": 0.00019057708688245318,
4287
+ "loss": 2.0203,
4288
+ "step": 354000
4289
+ },
4290
+ {
4291
+ "epoch": 4.72,
4292
+ "learning_rate": 0.0001905637776831346,
4293
+ "loss": 1.9758,
4294
+ "step": 354500
4295
+ },
4296
+ {
4297
+ "epoch": 4.72,
4298
+ "learning_rate": 0.000190550468483816,
4299
+ "loss": 1.8171,
4300
+ "step": 355000
4301
+ },
4302
+ {
4303
+ "epoch": 4.73,
4304
+ "learning_rate": 0.00019053715928449744,
4305
+ "loss": 2.0035,
4306
+ "step": 355500
4307
+ },
4308
+ {
4309
+ "epoch": 4.74,
4310
+ "learning_rate": 0.00019052385008517888,
4311
+ "loss": 1.9566,
4312
+ "step": 356000
4313
+ },
4314
+ {
4315
+ "epoch": 4.74,
4316
+ "learning_rate": 0.00019051054088586032,
4317
+ "loss": 1.8568,
4318
+ "step": 356500
4319
+ },
4320
+ {
4321
+ "epoch": 4.75,
4322
+ "learning_rate": 0.00019049723168654175,
4323
+ "loss": 2.0018,
4324
+ "step": 357000
4325
+ },
4326
+ {
4327
+ "epoch": 4.76,
4328
+ "learning_rate": 0.00019048392248722316,
4329
+ "loss": 1.8047,
4330
+ "step": 357500
4331
+ },
4332
+ {
4333
+ "epoch": 4.76,
4334
+ "learning_rate": 0.0001904706132879046,
4335
+ "loss": 1.8328,
4336
+ "step": 358000
4337
+ },
4338
+ {
4339
+ "epoch": 4.77,
4340
+ "learning_rate": 0.00019045730408858604,
4341
+ "loss": 1.9396,
4342
+ "step": 358500
4343
+ },
4344
+ {
4345
+ "epoch": 4.78,
4346
+ "learning_rate": 0.00019044399488926748,
4347
+ "loss": 1.8235,
4348
+ "step": 359000
4349
+ },
4350
+ {
4351
+ "epoch": 4.78,
4352
+ "learning_rate": 0.00019043068568994892,
4353
+ "loss": 1.8668,
4354
+ "step": 359500
4355
+ },
4356
+ {
4357
+ "epoch": 4.79,
4358
+ "learning_rate": 0.00019041737649063033,
4359
+ "loss": 2.0224,
4360
+ "step": 360000
4361
+ },
4362
+ {
4363
+ "epoch": 4.8,
4364
+ "learning_rate": 0.00019040406729131177,
4365
+ "loss": 1.9938,
4366
+ "step": 360500
4367
+ },
4368
+ {
4369
+ "epoch": 4.8,
4370
+ "learning_rate": 0.0001903907580919932,
4371
+ "loss": 1.9797,
4372
+ "step": 361000
4373
+ },
4374
+ {
4375
+ "epoch": 4.81,
4376
+ "learning_rate": 0.00019037744889267464,
4377
+ "loss": 1.9694,
4378
+ "step": 361500
4379
+ },
4380
+ {
4381
+ "epoch": 4.82,
4382
+ "learning_rate": 0.00019036413969335605,
4383
+ "loss": 2.0293,
4384
+ "step": 362000
4385
+ },
4386
+ {
4387
+ "epoch": 4.82,
4388
+ "learning_rate": 0.0001903508304940375,
4389
+ "loss": 1.9625,
4390
+ "step": 362500
4391
+ },
4392
+ {
4393
+ "epoch": 4.83,
4394
+ "learning_rate": 0.0001903375212947189,
4395
+ "loss": 1.9487,
4396
+ "step": 363000
4397
+ },
4398
+ {
4399
+ "epoch": 4.84,
4400
+ "learning_rate": 0.00019032421209540034,
4401
+ "loss": 1.9184,
4402
+ "step": 363500
4403
+ },
4404
+ {
4405
+ "epoch": 4.84,
4406
+ "learning_rate": 0.00019031090289608178,
4407
+ "loss": 2.0013,
4408
+ "step": 364000
4409
+ },
4410
+ {
4411
+ "epoch": 4.85,
4412
+ "learning_rate": 0.00019029759369676322,
4413
+ "loss": 1.9245,
4414
+ "step": 364500
4415
+ },
4416
+ {
4417
+ "epoch": 4.86,
4418
+ "learning_rate": 0.00019028428449744466,
4419
+ "loss": 1.8413,
4420
+ "step": 365000
4421
+ },
4422
+ {
4423
+ "epoch": 4.86,
4424
+ "learning_rate": 0.00019027097529812607,
4425
+ "loss": 1.8837,
4426
+ "step": 365500
4427
+ },
4428
+ {
4429
+ "epoch": 4.87,
4430
+ "learning_rate": 0.0001902576660988075,
4431
+ "loss": 2.0188,
4432
+ "step": 366000
4433
+ },
4434
+ {
4435
+ "epoch": 4.88,
4436
+ "learning_rate": 0.00019024435689948894,
4437
+ "loss": 1.9003,
4438
+ "step": 366500
4439
+ },
4440
+ {
4441
+ "epoch": 4.88,
4442
+ "learning_rate": 0.00019023104770017038,
4443
+ "loss": 1.9519,
4444
+ "step": 367000
4445
+ },
4446
+ {
4447
+ "epoch": 4.89,
4448
+ "learning_rate": 0.0001902177385008518,
4449
+ "loss": 1.9312,
4450
+ "step": 367500
4451
+ },
4452
+ {
4453
+ "epoch": 4.9,
4454
+ "learning_rate": 0.00019020442930153323,
4455
+ "loss": 1.8173,
4456
+ "step": 368000
4457
+ },
4458
+ {
4459
+ "epoch": 4.9,
4460
+ "learning_rate": 0.00019019112010221467,
4461
+ "loss": 1.9114,
4462
+ "step": 368500
4463
+ },
4464
+ {
4465
+ "epoch": 4.91,
4466
+ "learning_rate": 0.00019017781090289608,
4467
+ "loss": 1.9343,
4468
+ "step": 369000
4469
+ },
4470
+ {
4471
+ "epoch": 4.92,
4472
+ "learning_rate": 0.00019016450170357752,
4473
+ "loss": 1.906,
4474
+ "step": 369500
4475
+ },
4476
+ {
4477
+ "epoch": 4.92,
4478
+ "learning_rate": 0.00019015119250425896,
4479
+ "loss": 1.9389,
4480
+ "step": 370000
4481
+ },
4482
+ {
4483
+ "epoch": 4.93,
4484
+ "learning_rate": 0.00019013788330494037,
4485
+ "loss": 1.917,
4486
+ "step": 370500
4487
+ },
4488
+ {
4489
+ "epoch": 4.94,
4490
+ "learning_rate": 0.0001901245741056218,
4491
+ "loss": 1.9064,
4492
+ "step": 371000
4493
+ },
4494
+ {
4495
+ "epoch": 4.94,
4496
+ "learning_rate": 0.00019011126490630324,
4497
+ "loss": 1.9388,
4498
+ "step": 371500
4499
+ },
4500
+ {
4501
+ "epoch": 4.95,
4502
+ "learning_rate": 0.00019009795570698468,
4503
+ "loss": 1.8737,
4504
+ "step": 372000
4505
+ },
4506
+ {
4507
+ "epoch": 4.96,
4508
+ "learning_rate": 0.00019008464650766612,
4509
+ "loss": 1.9446,
4510
+ "step": 372500
4511
+ },
4512
+ {
4513
+ "epoch": 4.96,
4514
+ "learning_rate": 0.00019007133730834753,
4515
+ "loss": 1.9625,
4516
+ "step": 373000
4517
+ },
4518
+ {
4519
+ "epoch": 4.97,
4520
+ "learning_rate": 0.00019005802810902897,
4521
+ "loss": 1.8929,
4522
+ "step": 373500
4523
+ },
4524
+ {
4525
+ "epoch": 4.98,
4526
+ "learning_rate": 0.0001900447189097104,
4527
+ "loss": 1.9421,
4528
+ "step": 374000
4529
+ },
4530
+ {
4531
+ "epoch": 4.98,
4532
+ "learning_rate": 0.00019003140971039184,
4533
+ "loss": 2.0236,
4534
+ "step": 374500
4535
+ },
4536
+ {
4537
+ "epoch": 4.99,
4538
+ "learning_rate": 0.00019001810051107328,
4539
+ "loss": 2.0012,
4540
+ "step": 375000
4541
+ },
4542
+ {
4543
+ "epoch": 5.0,
4544
+ "learning_rate": 0.0001900047913117547,
4545
+ "loss": 2.0569,
4546
+ "step": 375500
4547
+ },
4548
+ {
4549
+ "epoch": 5.0,
4550
+ "eval_loss": 1.9444704055786133,
4551
+ "eval_runtime": 150.8043,
4552
+ "eval_samples_per_second": 62.286,
4553
+ "eval_steps_per_second": 62.286,
4554
+ "step": 375680
4555
+ }
4556
+ ],
4557
+ "max_steps": 7513600,
4558
+ "num_train_epochs": 100,
4559
+ "total_flos": 6244677543475200.0,
4560
+ "trial_name": null,
4561
+ "trial_params": null
4562
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c73a3a5d6da773158a7357071a288cb77e5c92f1a9a1eaa4711ed357e4b14e02
3
+ size 3183