Hugging Face's logo

CLMBR
/
full-transformer-0

Text Generation
Transformers
PyTorch
opt
Generated from Trainer
Inference Endpoints
text-generation-inference
Model card Files Community
abhinavp commited on
Commit
670ccf0
•
1 Parent(s): f0e993c

Training in progress, epoch 0, checkpoint

Browse files
Files changed (11) hide show
  1. checkpoint-381600/config.json +28 -0
  2. checkpoint-381600/generation_config.json +7 -0
  3. checkpoint-381600/optimizer.pt +3 -0
  4. checkpoint-381600/pytorch_model.bin +3 -0
  5. checkpoint-381600/rng_state.pth +3 -0
  6. checkpoint-381600/scheduler.pt +3 -0
  7. checkpoint-381600/special_tokens_map.json +5 -0
  8. checkpoint-381600/tokenizer.json +0 -0
  9. checkpoint-381600/tokenizer_config.json +12 -0
  10. checkpoint-381600/trainer_state.json +4535 -0
  11. checkpoint-381600/training_args.bin +3 -0
checkpoint-381600/config.json ADDED
@@ -0,0 +1,28 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_remove_final_layer_norm": false,
3
+ "activation_function": "relu",
4
+ "architectures": [
5
+ "OPTForCausalLM"
6
+ ],
7
+ "attention_dropout": 0.0,
8
+ "bos_token_id": 2,
9
+ "do_layer_norm_before": true,
10
+ "dropout": 0.1,
11
+ "enable_bias": true,
12
+ "eos_token_id": 2,
13
+ "ffn_dim": 768,
14
+ "hidden_size": 768,
15
+ "init_std": 0.02,
16
+ "layer_norm_elementwise_affine": true,
17
+ "layerdrop": 0.0,
18
+ "max_position_embeddings": 512,
19
+ "model_type": "opt",
20
+ "num_attention_heads": 8,
21
+ "num_hidden_layers": 8,
22
+ "pad_token_id": 1,
23
+ "torch_dtype": "float32",
24
+ "transformers_version": "4.33.3",
25
+ "use_cache": true,
26
+ "vocab_size": 50002,
27
+ "word_embed_proj_dim": 768
28
+ }
checkpoint-381600/generation_config.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 2,
4
+ "eos_token_id": 2,
5
+ "pad_token_id": 1,
6
+ "transformers_version": "4.33.3"
7
+ }
checkpoint-381600/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2fa02715cae5674a0dfef04ea9ce5c7754f57e8bfeda02dcfd6bd88a4ad94a43
3
+ size 537476677
checkpoint-381600/pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3ab271cd1a3a99949eaa9a096374a297887ac594bd5388c9eb4c986d59cded28
3
+ size 268727709
checkpoint-381600/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:981d6101f1d1ee82531ae9473bf41bfe7b08c7366faa5e7a9d34f9f9c5594c71
3
+ size 14511
checkpoint-381600/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:50b84567574eb74fb86fa84690ccca5e4d75e9475549b777b4cc362253a152fd
3
+ size 627
checkpoint-381600/special_tokens_map.json ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ {
2
+ "eos_token": "<eos>",
3
+ "pad_token": "<pad>",
4
+ "unk_token": "<unk>"
5
+ }
checkpoint-381600/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-381600/tokenizer_config.json ADDED
@@ -0,0 +1,12 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "clean_up_tokenization_spaces": true,
3
+ "eos_token": "<eos>",
4
+ "max_length": null,
5
+ "model_max_length": 512,
6
+ "pad_to_multiple_of": null,
7
+ "pad_token": "<pad>",
8
+ "pad_token_type_id": 0,
9
+ "padding_side": "right",
10
+ "tokenizer_class": "PreTrainedTokenizerFast",
11
+ "unk_token": "<unk>"
12
+ }
checkpoint-381600/trainer_state.json ADDED
@@ -0,0 +1,4535 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 3.8858225345611572,
3
+ "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/full2/transformer/0/checkpoints/checkpoint-381600",
4
+ "epoch": 0.025000606015738065,
5
+ "eval_steps": 10,
6
+ "global_step": 381600,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.0,
13
+ "learning_rate": 4.999998362119627e-05,
14
+ "loss": 11.0453,
15
+ "step": 1
16
+ },
17
+ {
18
+ "epoch": 0.0,
19
+ "learning_rate": 4.999161405248948e-05,
20
+ "loss": 6.8203,
21
+ "step": 512
22
+ },
23
+ {
24
+ "epoch": 0.0,
25
+ "learning_rate": 4.998322810497896e-05,
26
+ "loss": 6.1803,
27
+ "step": 1024
28
+ },
29
+ {
30
+ "epoch": 0.0,
31
+ "learning_rate": 4.997484215746844e-05,
32
+ "loss": 5.976,
33
+ "step": 1536
34
+ },
35
+ {
36
+ "epoch": 0.0,
37
+ "learning_rate": 4.996645620995792e-05,
38
+ "loss": 5.8136,
39
+ "step": 2048
40
+ },
41
+ {
42
+ "epoch": 0.0,
43
+ "learning_rate": 4.99580702624474e-05,
44
+ "loss": 5.7055,
45
+ "step": 2560
46
+ },
47
+ {
48
+ "epoch": 0.0,
49
+ "learning_rate": 4.994968431493688e-05,
50
+ "loss": 5.6058,
51
+ "step": 3072
52
+ },
53
+ {
54
+ "epoch": 0.0,
55
+ "learning_rate": 4.994129836742636e-05,
56
+ "loss": 5.5376,
57
+ "step": 3584
58
+ },
59
+ {
60
+ "epoch": 0.0,
61
+ "learning_rate": 4.993291241991584e-05,
62
+ "loss": 5.4512,
63
+ "step": 4096
64
+ },
65
+ {
66
+ "epoch": 0.0,
67
+ "learning_rate": 4.992452647240532e-05,
68
+ "loss": 5.403,
69
+ "step": 4608
70
+ },
71
+ {
72
+ "epoch": 0.0,
73
+ "learning_rate": 4.99161405248948e-05,
74
+ "loss": 5.3536,
75
+ "step": 5120
76
+ },
77
+ {
78
+ "epoch": 0.0,
79
+ "learning_rate": 4.990775457738428e-05,
80
+ "loss": 5.33,
81
+ "step": 5632
82
+ },
83
+ {
84
+ "epoch": 0.0,
85
+ "learning_rate": 4.989938500867749e-05,
86
+ "loss": 5.2746,
87
+ "step": 6144
88
+ },
89
+ {
90
+ "epoch": 0.0,
91
+ "learning_rate": 4.989099906116697e-05,
92
+ "loss": 5.211,
93
+ "step": 6656
94
+ },
95
+ {
96
+ "epoch": 0.0,
97
+ "learning_rate": 4.988261311365645e-05,
98
+ "loss": 5.1872,
99
+ "step": 7168
100
+ },
101
+ {
102
+ "epoch": 0.0,
103
+ "learning_rate": 4.987422716614593e-05,
104
+ "loss": 5.1491,
105
+ "step": 7680
106
+ },
107
+ {
108
+ "epoch": 0.0,
109
+ "learning_rate": 4.986584121863541e-05,
110
+ "loss": 5.1193,
111
+ "step": 8192
112
+ },
113
+ {
114
+ "epoch": 0.0,
115
+ "learning_rate": 4.985745527112489e-05,
116
+ "loss": 5.0928,
117
+ "step": 8704
118
+ },
119
+ {
120
+ "epoch": 0.0,
121
+ "learning_rate": 4.984906932361437e-05,
122
+ "loss": 5.0671,
123
+ "step": 9216
124
+ },
125
+ {
126
+ "epoch": 0.0,
127
+ "learning_rate": 4.984068337610385e-05,
128
+ "loss": 5.0268,
129
+ "step": 9728
130
+ },
131
+ {
132
+ "epoch": 0.0,
133
+ "learning_rate": 4.983229742859333e-05,
134
+ "loss": 5.0218,
135
+ "step": 10240
136
+ },
137
+ {
138
+ "epoch": 0.0,
139
+ "learning_rate": 4.9823927859886547e-05,
140
+ "loss": 4.986,
141
+ "step": 10752
142
+ },
143
+ {
144
+ "epoch": 0.0,
145
+ "learning_rate": 4.9815541912376026e-05,
146
+ "loss": 4.9659,
147
+ "step": 11264
148
+ },
149
+ {
150
+ "epoch": 0.0,
151
+ "learning_rate": 4.9807155964865506e-05,
152
+ "loss": 4.9377,
153
+ "step": 11776
154
+ },
155
+ {
156
+ "epoch": 0.0,
157
+ "learning_rate": 4.9798770017354986e-05,
158
+ "loss": 4.9353,
159
+ "step": 12288
160
+ },
161
+ {
162
+ "epoch": 0.0,
163
+ "learning_rate": 4.9790400448648195e-05,
164
+ "loss": 4.9084,
165
+ "step": 12800
166
+ },
167
+ {
168
+ "epoch": 0.0,
169
+ "learning_rate": 4.9782014501137675e-05,
170
+ "loss": 4.8814,
171
+ "step": 13312
172
+ },
173
+ {
174
+ "epoch": 0.0,
175
+ "learning_rate": 4.9773628553627155e-05,
176
+ "loss": 4.8641,
177
+ "step": 13824
178
+ },
179
+ {
180
+ "epoch": 0.0,
181
+ "learning_rate": 4.9765242606116635e-05,
182
+ "loss": 4.8465,
183
+ "step": 14336
184
+ },
185
+ {
186
+ "epoch": 0.0,
187
+ "learning_rate": 4.9756873037409844e-05,
188
+ "loss": 4.8346,
189
+ "step": 14848
190
+ },
191
+ {
192
+ "epoch": 0.01,
193
+ "learning_rate": 4.9748487089899324e-05,
194
+ "loss": 4.8112,
195
+ "step": 15360
196
+ },
197
+ {
198
+ "epoch": 0.01,
199
+ "learning_rate": 4.9740117521192533e-05,
200
+ "loss": 4.8046,
201
+ "step": 15872
202
+ },
203
+ {
204
+ "epoch": 0.01,
205
+ "learning_rate": 4.9731731573682013e-05,
206
+ "loss": 4.7911,
207
+ "step": 16384
208
+ },
209
+ {
210
+ "epoch": 0.01,
211
+ "learning_rate": 4.97233456261715e-05,
212
+ "loss": 4.7771,
213
+ "step": 16896
214
+ },
215
+ {
216
+ "epoch": 0.01,
217
+ "learning_rate": 4.971495967866098e-05,
218
+ "loss": 4.7744,
219
+ "step": 17408
220
+ },
221
+ {
222
+ "epoch": 0.01,
223
+ "learning_rate": 4.970657373115046e-05,
224
+ "loss": 4.7607,
225
+ "step": 17920
226
+ },
227
+ {
228
+ "epoch": 0.01,
229
+ "learning_rate": 4.969818778363994e-05,
230
+ "loss": 4.7448,
231
+ "step": 18432
232
+ },
233
+ {
234
+ "epoch": 0.01,
235
+ "learning_rate": 4.968980183612942e-05,
236
+ "loss": 4.7186,
237
+ "step": 18944
238
+ },
239
+ {
240
+ "epoch": 0.01,
241
+ "learning_rate": 4.96814158886189e-05,
242
+ "loss": 4.714,
243
+ "step": 19456
244
+ },
245
+ {
246
+ "epoch": 0.01,
247
+ "learning_rate": 4.967304631991211e-05,
248
+ "loss": 4.6948,
249
+ "step": 19968
250
+ },
251
+ {
252
+ "epoch": 0.01,
253
+ "learning_rate": 4.966466037240159e-05,
254
+ "loss": 4.6941,
255
+ "step": 20480
256
+ },
257
+ {
258
+ "epoch": 0.01,
259
+ "learning_rate": 4.965627442489107e-05,
260
+ "loss": 4.6788,
261
+ "step": 20992
262
+ },
263
+ {
264
+ "epoch": 0.01,
265
+ "learning_rate": 4.964788847738054e-05,
266
+ "loss": 4.6731,
267
+ "step": 21504
268
+ },
269
+ {
270
+ "epoch": 0.01,
271
+ "learning_rate": 4.963950252987002e-05,
272
+ "loss": 4.656,
273
+ "step": 22016
274
+ },
275
+ {
276
+ "epoch": 0.01,
277
+ "learning_rate": 4.963113296116324e-05,
278
+ "loss": 4.6628,
279
+ "step": 22528
280
+ },
281
+ {
282
+ "epoch": 0.01,
283
+ "learning_rate": 4.962274701365272e-05,
284
+ "loss": 4.6451,
285
+ "step": 23040
286
+ },
287
+ {
288
+ "epoch": 0.01,
289
+ "learning_rate": 4.96143610661422e-05,
290
+ "loss": 4.6473,
291
+ "step": 23552
292
+ },
293
+ {
294
+ "epoch": 0.01,
295
+ "learning_rate": 4.960597511863168e-05,
296
+ "loss": 4.6369,
297
+ "step": 24064
298
+ },
299
+ {
300
+ "epoch": 0.01,
301
+ "learning_rate": 4.959758917112116e-05,
302
+ "loss": 4.609,
303
+ "step": 24576
304
+ },
305
+ {
306
+ "epoch": 0.01,
307
+ "learning_rate": 4.95892359812181e-05,
308
+ "loss": 4.602,
309
+ "step": 25088
310
+ },
311
+ {
312
+ "epoch": 0.01,
313
+ "learning_rate": 4.958085003370758e-05,
314
+ "loss": 4.616,
315
+ "step": 25600
316
+ },
317
+ {
318
+ "epoch": 0.01,
319
+ "learning_rate": 4.957246408619706e-05,
320
+ "loss": 4.6008,
321
+ "step": 26112
322
+ },
323
+ {
324
+ "epoch": 0.01,
325
+ "learning_rate": 4.956407813868654e-05,
326
+ "loss": 4.5844,
327
+ "step": 26624
328
+ },
329
+ {
330
+ "epoch": 0.01,
331
+ "learning_rate": 4.9555692191176016e-05,
332
+ "loss": 4.5682,
333
+ "step": 27136
334
+ },
335
+ {
336
+ "epoch": 0.01,
337
+ "learning_rate": 4.9547306243665496e-05,
338
+ "loss": 4.5748,
339
+ "step": 27648
340
+ },
341
+ {
342
+ "epoch": 0.01,
343
+ "learning_rate": 4.9538920296154976e-05,
344
+ "loss": 4.5575,
345
+ "step": 28160
346
+ },
347
+ {
348
+ "epoch": 0.01,
349
+ "learning_rate": 4.9530534348644456e-05,
350
+ "loss": 4.5697,
351
+ "step": 28672
352
+ },
353
+ {
354
+ "epoch": 0.01,
355
+ "learning_rate": 4.952216477993767e-05,
356
+ "loss": 4.5357,
357
+ "step": 29184
358
+ },
359
+ {
360
+ "epoch": 0.01,
361
+ "learning_rate": 4.951377883242715e-05,
362
+ "loss": 4.5472,
363
+ "step": 29696
364
+ },
365
+ {
366
+ "epoch": 0.01,
367
+ "learning_rate": 4.950539288491663e-05,
368
+ "loss": 4.5447,
369
+ "step": 30208
370
+ },
371
+ {
372
+ "epoch": 0.01,
373
+ "learning_rate": 4.949700693740611e-05,
374
+ "loss": 4.5208,
375
+ "step": 30720
376
+ },
377
+ {
378
+ "epoch": 0.01,
379
+ "learning_rate": 4.948863736869932e-05,
380
+ "loss": 4.5379,
381
+ "step": 31232
382
+ },
383
+ {
384
+ "epoch": 0.01,
385
+ "learning_rate": 4.94802514211888e-05,
386
+ "loss": 4.5087,
387
+ "step": 31744
388
+ },
389
+ {
390
+ "epoch": 0.01,
391
+ "learning_rate": 4.947186547367828e-05,
392
+ "loss": 4.5105,
393
+ "step": 32256
394
+ },
395
+ {
396
+ "epoch": 0.01,
397
+ "learning_rate": 4.946347952616776e-05,
398
+ "loss": 4.4967,
399
+ "step": 32768
400
+ },
401
+ {
402
+ "epoch": 0.01,
403
+ "learning_rate": 4.945510995746097e-05,
404
+ "loss": 4.513,
405
+ "step": 33280
406
+ },
407
+ {
408
+ "epoch": 0.01,
409
+ "learning_rate": 4.944672400995045e-05,
410
+ "loss": 4.486,
411
+ "step": 33792
412
+ },
413
+ {
414
+ "epoch": 0.01,
415
+ "learning_rate": 4.943833806243993e-05,
416
+ "loss": 4.4883,
417
+ "step": 34304
418
+ },
419
+ {
420
+ "epoch": 0.01,
421
+ "learning_rate": 4.942995211492941e-05,
422
+ "loss": 4.4747,
423
+ "step": 34816
424
+ },
425
+ {
426
+ "epoch": 0.01,
427
+ "learning_rate": 4.942156616741889e-05,
428
+ "loss": 4.4701,
429
+ "step": 35328
430
+ },
431
+ {
432
+ "epoch": 0.01,
433
+ "learning_rate": 4.9413196598712105e-05,
434
+ "loss": 4.4835,
435
+ "step": 35840
436
+ },
437
+ {
438
+ "epoch": 0.01,
439
+ "learning_rate": 4.940482703000532e-05,
440
+ "loss": 4.4759,
441
+ "step": 36352
442
+ },
443
+ {
444
+ "epoch": 0.01,
445
+ "learning_rate": 4.9396441082494794e-05,
446
+ "loss": 4.4659,
447
+ "step": 36864
448
+ },
449
+ {
450
+ "epoch": 0.01,
451
+ "learning_rate": 4.9388055134984274e-05,
452
+ "loss": 4.4653,
453
+ "step": 37376
454
+ },
455
+ {
456
+ "epoch": 0.01,
457
+ "learning_rate": 4.9379669187473754e-05,
458
+ "loss": 4.4719,
459
+ "step": 37888
460
+ },
461
+ {
462
+ "epoch": 0.01,
463
+ "learning_rate": 4.9371283239963234e-05,
464
+ "loss": 4.446,
465
+ "step": 38400
466
+ },
467
+ {
468
+ "epoch": 0.01,
469
+ "learning_rate": 4.936291367125644e-05,
470
+ "loss": 4.4601,
471
+ "step": 38912
472
+ },
473
+ {
474
+ "epoch": 0.01,
475
+ "learning_rate": 4.935452772374592e-05,
476
+ "loss": 4.4346,
477
+ "step": 39424
478
+ },
479
+ {
480
+ "epoch": 0.01,
481
+ "learning_rate": 4.93461417762354e-05,
482
+ "loss": 4.4338,
483
+ "step": 39936
484
+ },
485
+ {
486
+ "epoch": 0.01,
487
+ "learning_rate": 4.933775582872488e-05,
488
+ "loss": 4.4317,
489
+ "step": 40448
490
+ },
491
+ {
492
+ "epoch": 0.01,
493
+ "learning_rate": 4.932936988121436e-05,
494
+ "loss": 4.4345,
495
+ "step": 40960
496
+ },
497
+ {
498
+ "epoch": 0.01,
499
+ "learning_rate": 4.932098393370384e-05,
500
+ "loss": 4.4255,
501
+ "step": 41472
502
+ },
503
+ {
504
+ "epoch": 0.01,
505
+ "learning_rate": 4.931259798619332e-05,
506
+ "loss": 4.4334,
507
+ "step": 41984
508
+ },
509
+ {
510
+ "epoch": 0.01,
511
+ "learning_rate": 4.930421203868281e-05,
512
+ "loss": 4.418,
513
+ "step": 42496
514
+ },
515
+ {
516
+ "epoch": 0.01,
517
+ "learning_rate": 4.929584246997602e-05,
518
+ "loss": 4.3969,
519
+ "step": 43008
520
+ },
521
+ {
522
+ "epoch": 0.01,
523
+ "learning_rate": 4.92874565224655e-05,
524
+ "loss": 4.3995,
525
+ "step": 43520
526
+ },
527
+ {
528
+ "epoch": 0.01,
529
+ "learning_rate": 4.927907057495498e-05,
530
+ "loss": 4.411,
531
+ "step": 44032
532
+ },
533
+ {
534
+ "epoch": 0.01,
535
+ "learning_rate": 4.927070100624819e-05,
536
+ "loss": 4.3954,
537
+ "step": 44544
538
+ },
539
+ {
540
+ "epoch": 0.01,
541
+ "learning_rate": 4.92623314375414e-05,
542
+ "loss": 4.3944,
543
+ "step": 45056
544
+ },
545
+ {
546
+ "epoch": 0.01,
547
+ "learning_rate": 4.925394549003088e-05,
548
+ "loss": 4.3981,
549
+ "step": 45568
550
+ },
551
+ {
552
+ "epoch": 0.02,
553
+ "learning_rate": 4.924555954252036e-05,
554
+ "loss": 4.3813,
555
+ "step": 46080
556
+ },
557
+ {
558
+ "epoch": 0.02,
559
+ "learning_rate": 4.923717359500984e-05,
560
+ "loss": 4.3868,
561
+ "step": 46592
562
+ },
563
+ {
564
+ "epoch": 0.02,
565
+ "learning_rate": 4.9228804026303046e-05,
566
+ "loss": 4.3807,
567
+ "step": 47104
568
+ },
569
+ {
570
+ "epoch": 0.02,
571
+ "learning_rate": 4.922041807879253e-05,
572
+ "loss": 4.3858,
573
+ "step": 47616
574
+ },
575
+ {
576
+ "epoch": 0.02,
577
+ "learning_rate": 4.921203213128201e-05,
578
+ "loss": 4.3832,
579
+ "step": 48128
580
+ },
581
+ {
582
+ "epoch": 0.02,
583
+ "learning_rate": 4.920364618377149e-05,
584
+ "loss": 4.3773,
585
+ "step": 48640
586
+ },
587
+ {
588
+ "epoch": 0.02,
589
+ "learning_rate": 4.919526023626097e-05,
590
+ "loss": 4.3577,
591
+ "step": 49152
592
+ },
593
+ {
594
+ "epoch": 0.02,
595
+ "learning_rate": 4.918687428875045e-05,
596
+ "loss": 4.3646,
597
+ "step": 49664
598
+ },
599
+ {
600
+ "epoch": 0.02,
601
+ "learning_rate": 4.917848834123993e-05,
602
+ "loss": 4.3568,
603
+ "step": 50176
604
+ },
605
+ {
606
+ "epoch": 0.02,
607
+ "learning_rate": 4.917010239372941e-05,
608
+ "loss": 4.3604,
609
+ "step": 50688
610
+ },
611
+ {
612
+ "epoch": 0.02,
613
+ "learning_rate": 4.916171644621889e-05,
614
+ "loss": 4.3648,
615
+ "step": 51200
616
+ },
617
+ {
618
+ "epoch": 0.02,
619
+ "learning_rate": 4.915333049870837e-05,
620
+ "loss": 4.3437,
621
+ "step": 51712
622
+ },
623
+ {
624
+ "epoch": 0.02,
625
+ "learning_rate": 4.914494455119785e-05,
626
+ "loss": 4.3483,
627
+ "step": 52224
628
+ },
629
+ {
630
+ "epoch": 0.02,
631
+ "learning_rate": 4.913657498249106e-05,
632
+ "loss": 4.3495,
633
+ "step": 52736
634
+ },
635
+ {
636
+ "epoch": 0.02,
637
+ "learning_rate": 4.912818903498054e-05,
638
+ "loss": 4.3432,
639
+ "step": 53248
640
+ },
641
+ {
642
+ "epoch": 0.02,
643
+ "learning_rate": 4.911980308747002e-05,
644
+ "loss": 4.3316,
645
+ "step": 53760
646
+ },
647
+ {
648
+ "epoch": 0.02,
649
+ "learning_rate": 4.91114171399595e-05,
650
+ "loss": 4.335,
651
+ "step": 54272
652
+ },
653
+ {
654
+ "epoch": 0.02,
655
+ "learning_rate": 4.910303119244898e-05,
656
+ "loss": 4.322,
657
+ "step": 54784
658
+ },
659
+ {
660
+ "epoch": 0.02,
661
+ "learning_rate": 4.909464524493846e-05,
662
+ "loss": 4.3218,
663
+ "step": 55296
664
+ },
665
+ {
666
+ "epoch": 0.02,
667
+ "learning_rate": 4.908627567623168e-05,
668
+ "loss": 4.3195,
669
+ "step": 55808
670
+ },
671
+ {
672
+ "epoch": 0.02,
673
+ "learning_rate": 4.907788972872115e-05,
674
+ "loss": 4.3196,
675
+ "step": 56320
676
+ },
677
+ {
678
+ "epoch": 0.02,
679
+ "learning_rate": 4.906950378121063e-05,
680
+ "loss": 4.314,
681
+ "step": 56832
682
+ },
683
+ {
684
+ "epoch": 0.02,
685
+ "learning_rate": 4.906111783370011e-05,
686
+ "loss": 4.3183,
687
+ "step": 57344
688
+ },
689
+ {
690
+ "epoch": 0.02,
691
+ "learning_rate": 4.905273188618959e-05,
692
+ "loss": 4.3097,
693
+ "step": 57856
694
+ },
695
+ {
696
+ "epoch": 0.02,
697
+ "learning_rate": 4.904434593867907e-05,
698
+ "loss": 4.3127,
699
+ "step": 58368
700
+ },
701
+ {
702
+ "epoch": 0.02,
703
+ "learning_rate": 4.903595999116855e-05,
704
+ "loss": 4.3213,
705
+ "step": 58880
706
+ },
707
+ {
708
+ "epoch": 0.02,
709
+ "learning_rate": 4.902759042246176e-05,
710
+ "loss": 4.3152,
711
+ "step": 59392
712
+ },
713
+ {
714
+ "epoch": 0.02,
715
+ "learning_rate": 4.901920447495124e-05,
716
+ "loss": 4.3094,
717
+ "step": 59904
718
+ },
719
+ {
720
+ "epoch": 0.02,
721
+ "learning_rate": 4.901081852744072e-05,
722
+ "loss": 4.2991,
723
+ "step": 60416
724
+ },
725
+ {
726
+ "epoch": 0.02,
727
+ "learning_rate": 4.90024325799302e-05,
728
+ "loss": 4.2987,
729
+ "step": 60928
730
+ },
731
+ {
732
+ "epoch": 0.02,
733
+ "learning_rate": 4.8994046632419686e-05,
734
+ "loss": 4.2997,
735
+ "step": 61440
736
+ },
737
+ {
738
+ "epoch": 0.02,
739
+ "learning_rate": 4.8985660684909166e-05,
740
+ "loss": 4.2858,
741
+ "step": 61952
742
+ },
743
+ {
744
+ "epoch": 0.02,
745
+ "learning_rate": 4.8977291116202375e-05,
746
+ "loss": 4.2887,
747
+ "step": 62464
748
+ },
749
+ {
750
+ "epoch": 0.02,
751
+ "learning_rate": 4.8968905168691855e-05,
752
+ "loss": 4.2851,
753
+ "step": 62976
754
+ },
755
+ {
756
+ "epoch": 0.02,
757
+ "learning_rate": 4.8960519221181335e-05,
758
+ "loss": 4.2893,
759
+ "step": 63488
760
+ },
761
+ {
762
+ "epoch": 0.02,
763
+ "learning_rate": 4.8952133273670815e-05,
764
+ "loss": 4.2801,
765
+ "step": 64000
766
+ },
767
+ {
768
+ "epoch": 0.02,
769
+ "learning_rate": 4.8943747326160294e-05,
770
+ "loss": 4.2895,
771
+ "step": 64512
772
+ },
773
+ {
774
+ "epoch": 0.02,
775
+ "learning_rate": 4.8935377757453504e-05,
776
+ "loss": 4.2816,
777
+ "step": 65024
778
+ },
779
+ {
780
+ "epoch": 0.02,
781
+ "learning_rate": 4.8926991809942984e-05,
782
+ "loss": 4.2735,
783
+ "step": 65536
784
+ },
785
+ {
786
+ "epoch": 0.02,
787
+ "learning_rate": 4.8918605862432463e-05,
788
+ "loss": 4.2684,
789
+ "step": 66048
790
+ },
791
+ {
792
+ "epoch": 0.02,
793
+ "learning_rate": 4.8910219914921943e-05,
794
+ "loss": 4.2846,
795
+ "step": 66560
796
+ },
797
+ {
798
+ "epoch": 0.02,
799
+ "learning_rate": 4.8901833967411423e-05,
800
+ "loss": 4.2742,
801
+ "step": 67072
802
+ },
803
+ {
804
+ "epoch": 0.02,
805
+ "learning_rate": 4.88934480199009e-05,
806
+ "loss": 4.2679,
807
+ "step": 67584
808
+ },
809
+ {
810
+ "epoch": 0.02,
811
+ "learning_rate": 4.888506207239038e-05,
812
+ "loss": 4.2691,
813
+ "step": 68096
814
+ },
815
+ {
816
+ "epoch": 0.02,
817
+ "learning_rate": 4.88766925036836e-05,
818
+ "loss": 4.2702,
819
+ "step": 68608
820
+ },
821
+ {
822
+ "epoch": 0.02,
823
+ "learning_rate": 4.886830655617308e-05,
824
+ "loss": 4.2672,
825
+ "step": 69120
826
+ },
827
+ {
828
+ "epoch": 0.02,
829
+ "learning_rate": 4.885992060866256e-05,
830
+ "loss": 4.2696,
831
+ "step": 69632
832
+ },
833
+ {
834
+ "epoch": 0.02,
835
+ "learning_rate": 4.885155103995577e-05,
836
+ "loss": 4.2588,
837
+ "step": 70144
838
+ },
839
+ {
840
+ "epoch": 0.02,
841
+ "learning_rate": 4.884316509244525e-05,
842
+ "loss": 4.2571,
843
+ "step": 70656
844
+ },
845
+ {
846
+ "epoch": 0.02,
847
+ "learning_rate": 4.883477914493473e-05,
848
+ "loss": 4.2581,
849
+ "step": 71168
850
+ },
851
+ {
852
+ "epoch": 0.02,
853
+ "learning_rate": 4.882639319742421e-05,
854
+ "loss": 4.25,
855
+ "step": 71680
856
+ },
857
+ {
858
+ "epoch": 0.02,
859
+ "learning_rate": 4.881800724991369e-05,
860
+ "loss": 4.2452,
861
+ "step": 72192
862
+ },
863
+ {
864
+ "epoch": 0.02,
865
+ "learning_rate": 4.880962130240316e-05,
866
+ "loss": 4.245,
867
+ "step": 72704
868
+ },
869
+ {
870
+ "epoch": 0.02,
871
+ "learning_rate": 4.880123535489264e-05,
872
+ "loss": 4.2489,
873
+ "step": 73216
874
+ },
875
+ {
876
+ "epoch": 0.02,
877
+ "learning_rate": 4.879286578618586e-05,
878
+ "loss": 4.2394,
879
+ "step": 73728
880
+ },
881
+ {
882
+ "epoch": 0.02,
883
+ "learning_rate": 4.878447983867534e-05,
884
+ "loss": 4.2489,
885
+ "step": 74240
886
+ },
887
+ {
888
+ "epoch": 0.02,
889
+ "learning_rate": 4.877609389116482e-05,
890
+ "loss": 4.2412,
891
+ "step": 74752
892
+ },
893
+ {
894
+ "epoch": 0.02,
895
+ "learning_rate": 4.87677079436543e-05,
896
+ "loss": 4.2424,
897
+ "step": 75264
898
+ },
899
+ {
900
+ "epoch": 0.02,
901
+ "learning_rate": 4.875933837494751e-05,
902
+ "loss": 4.2462,
903
+ "step": 75776
904
+ },
905
+ {
906
+ "epoch": 0.02,
907
+ "learning_rate": 4.8750952427436986e-05,
908
+ "loss": 4.2278,
909
+ "step": 76288
910
+ },
911
+ {
912
+ "epoch": 0.03,
913
+ "eval_loss": 4.198204517364502,
914
+ "eval_runtime": 302.7571,
915
+ "eval_samples_per_second": 1260.387,
916
+ "eval_steps_per_second": 39.388,
917
+ "step": 76320
918
+ },
919
+ {
920
+ "epoch": 1.0,
921
+ "learning_rate": 4.8742566479926466e-05,
922
+ "loss": 4.2153,
923
+ "step": 76800
924
+ },
925
+ {
926
+ "epoch": 1.0,
927
+ "learning_rate": 4.8734180532415946e-05,
928
+ "loss": 4.2138,
929
+ "step": 77312
930
+ },
931
+ {
932
+ "epoch": 1.0,
933
+ "learning_rate": 4.872581096370916e-05,
934
+ "loss": 4.238,
935
+ "step": 77824
936
+ },
937
+ {
938
+ "epoch": 1.0,
939
+ "learning_rate": 4.8717425016198635e-05,
940
+ "loss": 4.2213,
941
+ "step": 78336
942
+ },
943
+ {
944
+ "epoch": 1.0,
945
+ "learning_rate": 4.8709039068688115e-05,
946
+ "loss": 4.2241,
947
+ "step": 78848
948
+ },
949
+ {
950
+ "epoch": 1.0,
951
+ "learning_rate": 4.8700653121177595e-05,
952
+ "loss": 4.2182,
953
+ "step": 79360
954
+ },
955
+ {
956
+ "epoch": 1.0,
957
+ "learning_rate": 4.8692267173667075e-05,
958
+ "loss": 4.2109,
959
+ "step": 79872
960
+ },
961
+ {
962
+ "epoch": 1.0,
963
+ "learning_rate": 4.8683881226156555e-05,
964
+ "loss": 4.202,
965
+ "step": 80384
966
+ },
967
+ {
968
+ "epoch": 1.0,
969
+ "learning_rate": 4.867551165744977e-05,
970
+ "loss": 4.2111,
971
+ "step": 80896
972
+ },
973
+ {
974
+ "epoch": 1.0,
975
+ "learning_rate": 4.866712570993925e-05,
976
+ "loss": 4.2031,
977
+ "step": 81408
978
+ },
979
+ {
980
+ "epoch": 1.0,
981
+ "learning_rate": 4.865873976242873e-05,
982
+ "loss": 4.2263,
983
+ "step": 81920
984
+ },
985
+ {
986
+ "epoch": 1.0,
987
+ "learning_rate": 4.865035381491821e-05,
988
+ "loss": 4.2114,
989
+ "step": 82432
990
+ },
991
+ {
992
+ "epoch": 1.0,
993
+ "learning_rate": 4.864198424621142e-05,
994
+ "loss": 4.1925,
995
+ "step": 82944
996
+ },
997
+ {
998
+ "epoch": 1.0,
999
+ "learning_rate": 4.86335982987009e-05,
1000
+ "loss": 4.1953,
1001
+ "step": 83456
1002
+ },
1003
+ {
1004
+ "epoch": 1.0,
1005
+ "learning_rate": 4.862521235119038e-05,
1006
+ "loss": 4.1971,
1007
+ "step": 83968
1008
+ },
1009
+ {
1010
+ "epoch": 1.0,
1011
+ "learning_rate": 4.861682640367986e-05,
1012
+ "loss": 4.1843,
1013
+ "step": 84480
1014
+ },
1015
+ {
1016
+ "epoch": 1.0,
1017
+ "learning_rate": 4.860844045616934e-05,
1018
+ "loss": 4.1969,
1019
+ "step": 84992
1020
+ },
1021
+ {
1022
+ "epoch": 1.0,
1023
+ "learning_rate": 4.860005450865882e-05,
1024
+ "loss": 4.1913,
1025
+ "step": 85504
1026
+ },
1027
+ {
1028
+ "epoch": 1.0,
1029
+ "learning_rate": 4.85916685611483e-05,
1030
+ "loss": 4.1805,
1031
+ "step": 86016
1032
+ },
1033
+ {
1034
+ "epoch": 1.0,
1035
+ "learning_rate": 4.858328261363778e-05,
1036
+ "loss": 4.2049,
1037
+ "step": 86528
1038
+ },
1039
+ {
1040
+ "epoch": 1.0,
1041
+ "learning_rate": 4.857489666612726e-05,
1042
+ "loss": 4.1852,
1043
+ "step": 87040
1044
+ },
1045
+ {
1046
+ "epoch": 1.0,
1047
+ "learning_rate": 4.856651071861674e-05,
1048
+ "loss": 4.186,
1049
+ "step": 87552
1050
+ },
1051
+ {
1052
+ "epoch": 1.0,
1053
+ "learning_rate": 4.8558141149909955e-05,
1054
+ "loss": 4.1768,
1055
+ "step": 88064
1056
+ },
1057
+ {
1058
+ "epoch": 1.0,
1059
+ "learning_rate": 4.8549755202399435e-05,
1060
+ "loss": 4.1956,
1061
+ "step": 88576
1062
+ },
1063
+ {
1064
+ "epoch": 1.0,
1065
+ "learning_rate": 4.8541369254888915e-05,
1066
+ "loss": 4.1737,
1067
+ "step": 89088
1068
+ },
1069
+ {
1070
+ "epoch": 1.0,
1071
+ "learning_rate": 4.8532983307378395e-05,
1072
+ "loss": 4.1744,
1073
+ "step": 89600
1074
+ },
1075
+ {
1076
+ "epoch": 1.0,
1077
+ "learning_rate": 4.8524597359867875e-05,
1078
+ "loss": 4.1703,
1079
+ "step": 90112
1080
+ },
1081
+ {
1082
+ "epoch": 1.0,
1083
+ "learning_rate": 4.8516227791161084e-05,
1084
+ "loss": 4.1688,
1085
+ "step": 90624
1086
+ },
1087
+ {
1088
+ "epoch": 1.0,
1089
+ "learning_rate": 4.8507841843650564e-05,
1090
+ "loss": 4.1673,
1091
+ "step": 91136
1092
+ },
1093
+ {
1094
+ "epoch": 1.01,
1095
+ "learning_rate": 4.8499455896140044e-05,
1096
+ "loss": 4.166,
1097
+ "step": 91648
1098
+ },
1099
+ {
1100
+ "epoch": 1.01,
1101
+ "learning_rate": 4.8491069948629524e-05,
1102
+ "loss": 4.1693,
1103
+ "step": 92160
1104
+ },
1105
+ {
1106
+ "epoch": 1.01,
1107
+ "learning_rate": 4.8482684001119e-05,
1108
+ "loss": 4.1671,
1109
+ "step": 92672
1110
+ },
1111
+ {
1112
+ "epoch": 1.01,
1113
+ "learning_rate": 4.847429805360848e-05,
1114
+ "loss": 4.1652,
1115
+ "step": 93184
1116
+ },
1117
+ {
1118
+ "epoch": 1.01,
1119
+ "learning_rate": 4.8465912106097964e-05,
1120
+ "loss": 4.1701,
1121
+ "step": 93696
1122
+ },
1123
+ {
1124
+ "epoch": 1.01,
1125
+ "learning_rate": 4.845754253739117e-05,
1126
+ "loss": 4.1644,
1127
+ "step": 94208
1128
+ },
1129
+ {
1130
+ "epoch": 1.01,
1131
+ "learning_rate": 4.844915658988065e-05,
1132
+ "loss": 4.1653,
1133
+ "step": 94720
1134
+ },
1135
+ {
1136
+ "epoch": 1.01,
1137
+ "learning_rate": 4.844077064237013e-05,
1138
+ "loss": 4.1492,
1139
+ "step": 95232
1140
+ },
1141
+ {
1142
+ "epoch": 1.01,
1143
+ "learning_rate": 4.843238469485961e-05,
1144
+ "loss": 4.1523,
1145
+ "step": 95744
1146
+ },
1147
+ {
1148
+ "epoch": 1.01,
1149
+ "learning_rate": 4.842399874734909e-05,
1150
+ "loss": 4.1433,
1151
+ "step": 96256
1152
+ },
1153
+ {
1154
+ "epoch": 1.01,
1155
+ "learning_rate": 4.841561279983857e-05,
1156
+ "loss": 4.1509,
1157
+ "step": 96768
1158
+ },
1159
+ {
1160
+ "epoch": 1.01,
1161
+ "learning_rate": 4.840724323113178e-05,
1162
+ "loss": 4.1423,
1163
+ "step": 97280
1164
+ },
1165
+ {
1166
+ "epoch": 1.01,
1167
+ "learning_rate": 4.839885728362126e-05,
1168
+ "loss": 4.1517,
1169
+ "step": 97792
1170
+ },
1171
+ {
1172
+ "epoch": 1.01,
1173
+ "learning_rate": 4.839047133611074e-05,
1174
+ "loss": 4.1339,
1175
+ "step": 98304
1176
+ },
1177
+ {
1178
+ "epoch": 1.01,
1179
+ "learning_rate": 4.838208538860022e-05,
1180
+ "loss": 4.1537,
1181
+ "step": 98816
1182
+ },
1183
+ {
1184
+ "epoch": 1.01,
1185
+ "learning_rate": 4.83736994410897e-05,
1186
+ "loss": 4.1457,
1187
+ "step": 99328
1188
+ },
1189
+ {
1190
+ "epoch": 1.01,
1191
+ "learning_rate": 4.836531349357918e-05,
1192
+ "loss": 4.1453,
1193
+ "step": 99840
1194
+ },
1195
+ {
1196
+ "epoch": 1.01,
1197
+ "learning_rate": 4.83569439248724e-05,
1198
+ "loss": 4.1428,
1199
+ "step": 100352
1200
+ },
1201
+ {
1202
+ "epoch": 1.01,
1203
+ "learning_rate": 4.834855797736188e-05,
1204
+ "loss": 4.1279,
1205
+ "step": 100864
1206
+ },
1207
+ {
1208
+ "epoch": 1.01,
1209
+ "learning_rate": 4.834017202985136e-05,
1210
+ "loss": 4.13,
1211
+ "step": 101376
1212
+ },
1213
+ {
1214
+ "epoch": 1.01,
1215
+ "learning_rate": 4.833178608234084e-05,
1216
+ "loss": 4.1439,
1217
+ "step": 101888
1218
+ },
1219
+ {
1220
+ "epoch": 1.01,
1221
+ "learning_rate": 4.832340013483032e-05,
1222
+ "loss": 4.1345,
1223
+ "step": 102400
1224
+ },
1225
+ {
1226
+ "epoch": 1.01,
1227
+ "learning_rate": 4.83150141873198e-05,
1228
+ "loss": 4.1282,
1229
+ "step": 102912
1230
+ },
1231
+ {
1232
+ "epoch": 1.01,
1233
+ "learning_rate": 4.830662823980928e-05,
1234
+ "loss": 4.1195,
1235
+ "step": 103424
1236
+ },
1237
+ {
1238
+ "epoch": 1.01,
1239
+ "learning_rate": 4.8298258671102486e-05,
1240
+ "loss": 4.1293,
1241
+ "step": 103936
1242
+ },
1243
+ {
1244
+ "epoch": 1.01,
1245
+ "learning_rate": 4.8289872723591966e-05,
1246
+ "loss": 4.1114,
1247
+ "step": 104448
1248
+ },
1249
+ {
1250
+ "epoch": 1.01,
1251
+ "learning_rate": 4.8281486776081446e-05,
1252
+ "loss": 4.137,
1253
+ "step": 104960
1254
+ },
1255
+ {
1256
+ "epoch": 1.01,
1257
+ "learning_rate": 4.8273100828570926e-05,
1258
+ "loss": 4.1076,
1259
+ "step": 105472
1260
+ },
1261
+ {
1262
+ "epoch": 1.01,
1263
+ "learning_rate": 4.8264714881060406e-05,
1264
+ "loss": 4.127,
1265
+ "step": 105984
1266
+ },
1267
+ {
1268
+ "epoch": 1.01,
1269
+ "learning_rate": 4.8256328933549886e-05,
1270
+ "loss": 4.1235,
1271
+ "step": 106496
1272
+ },
1273
+ {
1274
+ "epoch": 1.01,
1275
+ "learning_rate": 4.8247942986039366e-05,
1276
+ "loss": 4.1103,
1277
+ "step": 107008
1278
+ },
1279
+ {
1280
+ "epoch": 1.01,
1281
+ "learning_rate": 4.823957341733258e-05,
1282
+ "loss": 4.1243,
1283
+ "step": 107520
1284
+ },
1285
+ {
1286
+ "epoch": 1.01,
1287
+ "learning_rate": 4.823118746982206e-05,
1288
+ "loss": 4.1121,
1289
+ "step": 108032
1290
+ },
1291
+ {
1292
+ "epoch": 1.01,
1293
+ "learning_rate": 4.8222801522311535e-05,
1294
+ "loss": 4.1111,
1295
+ "step": 108544
1296
+ },
1297
+ {
1298
+ "epoch": 1.01,
1299
+ "learning_rate": 4.8214415574801015e-05,
1300
+ "loss": 4.0977,
1301
+ "step": 109056
1302
+ },
1303
+ {
1304
+ "epoch": 1.01,
1305
+ "learning_rate": 4.8206029627290495e-05,
1306
+ "loss": 4.1234,
1307
+ "step": 109568
1308
+ },
1309
+ {
1310
+ "epoch": 1.01,
1311
+ "learning_rate": 4.8197643679779975e-05,
1312
+ "loss": 4.0986,
1313
+ "step": 110080
1314
+ },
1315
+ {
1316
+ "epoch": 1.01,
1317
+ "learning_rate": 4.8189257732269455e-05,
1318
+ "loss": 4.107,
1319
+ "step": 110592
1320
+ },
1321
+ {
1322
+ "epoch": 1.01,
1323
+ "learning_rate": 4.8180888163562664e-05,
1324
+ "loss": 4.0888,
1325
+ "step": 111104
1326
+ },
1327
+ {
1328
+ "epoch": 1.01,
1329
+ "learning_rate": 4.8172502216052144e-05,
1330
+ "loss": 4.1007,
1331
+ "step": 111616
1332
+ },
1333
+ {
1334
+ "epoch": 1.01,
1335
+ "learning_rate": 4.8164116268541624e-05,
1336
+ "loss": 4.1112,
1337
+ "step": 112128
1338
+ },
1339
+ {
1340
+ "epoch": 1.01,
1341
+ "learning_rate": 4.8155730321031104e-05,
1342
+ "loss": 4.1088,
1343
+ "step": 112640
1344
+ },
1345
+ {
1346
+ "epoch": 1.01,
1347
+ "learning_rate": 4.814736075232432e-05,
1348
+ "loss": 4.1031,
1349
+ "step": 113152
1350
+ },
1351
+ {
1352
+ "epoch": 1.01,
1353
+ "learning_rate": 4.81389748048138e-05,
1354
+ "loss": 4.1086,
1355
+ "step": 113664
1356
+ },
1357
+ {
1358
+ "epoch": 1.01,
1359
+ "learning_rate": 4.813058885730328e-05,
1360
+ "loss": 4.1114,
1361
+ "step": 114176
1362
+ },
1363
+ {
1364
+ "epoch": 1.01,
1365
+ "learning_rate": 4.812220290979276e-05,
1366
+ "loss": 4.0973,
1367
+ "step": 114688
1368
+ },
1369
+ {
1370
+ "epoch": 1.01,
1371
+ "learning_rate": 4.811381696228224e-05,
1372
+ "loss": 4.1069,
1373
+ "step": 115200
1374
+ },
1375
+ {
1376
+ "epoch": 1.01,
1377
+ "learning_rate": 4.810544739357545e-05,
1378
+ "loss": 4.0937,
1379
+ "step": 115712
1380
+ },
1381
+ {
1382
+ "epoch": 1.01,
1383
+ "learning_rate": 4.809706144606493e-05,
1384
+ "loss": 4.0938,
1385
+ "step": 116224
1386
+ },
1387
+ {
1388
+ "epoch": 1.01,
1389
+ "learning_rate": 4.808867549855441e-05,
1390
+ "loss": 4.0892,
1391
+ "step": 116736
1392
+ },
1393
+ {
1394
+ "epoch": 1.01,
1395
+ "learning_rate": 4.808028955104389e-05,
1396
+ "loss": 4.0989,
1397
+ "step": 117248
1398
+ },
1399
+ {
1400
+ "epoch": 1.01,
1401
+ "learning_rate": 4.80719199823371e-05,
1402
+ "loss": 4.0985,
1403
+ "step": 117760
1404
+ },
1405
+ {
1406
+ "epoch": 1.01,
1407
+ "learning_rate": 4.806353403482658e-05,
1408
+ "loss": 4.1027,
1409
+ "step": 118272
1410
+ },
1411
+ {
1412
+ "epoch": 1.01,
1413
+ "learning_rate": 4.805514808731606e-05,
1414
+ "loss": 4.0917,
1415
+ "step": 118784
1416
+ },
1417
+ {
1418
+ "epoch": 1.01,
1419
+ "learning_rate": 4.804676213980554e-05,
1420
+ "loss": 4.072,
1421
+ "step": 119296
1422
+ },
1423
+ {
1424
+ "epoch": 1.01,
1425
+ "learning_rate": 4.803839257109875e-05,
1426
+ "loss": 4.0766,
1427
+ "step": 119808
1428
+ },
1429
+ {
1430
+ "epoch": 1.01,
1431
+ "learning_rate": 4.803000662358823e-05,
1432
+ "loss": 4.0888,
1433
+ "step": 120320
1434
+ },
1435
+ {
1436
+ "epoch": 1.01,
1437
+ "learning_rate": 4.802162067607771e-05,
1438
+ "loss": 4.0822,
1439
+ "step": 120832
1440
+ },
1441
+ {
1442
+ "epoch": 1.01,
1443
+ "learning_rate": 4.801323472856719e-05,
1444
+ "loss": 4.0807,
1445
+ "step": 121344
1446
+ },
1447
+ {
1448
+ "epoch": 1.01,
1449
+ "learning_rate": 4.80048651598604e-05,
1450
+ "loss": 4.0884,
1451
+ "step": 121856
1452
+ },
1453
+ {
1454
+ "epoch": 1.02,
1455
+ "learning_rate": 4.799647921234988e-05,
1456
+ "loss": 4.0742,
1457
+ "step": 122368
1458
+ },
1459
+ {
1460
+ "epoch": 1.02,
1461
+ "learning_rate": 4.798809326483936e-05,
1462
+ "loss": 4.0768,
1463
+ "step": 122880
1464
+ },
1465
+ {
1466
+ "epoch": 1.02,
1467
+ "learning_rate": 4.797970731732884e-05,
1468
+ "loss": 4.0781,
1469
+ "step": 123392
1470
+ },
1471
+ {
1472
+ "epoch": 1.02,
1473
+ "learning_rate": 4.797133774862205e-05,
1474
+ "loss": 4.0878,
1475
+ "step": 123904
1476
+ },
1477
+ {
1478
+ "epoch": 1.02,
1479
+ "learning_rate": 4.796295180111153e-05,
1480
+ "loss": 4.0841,
1481
+ "step": 124416
1482
+ },
1483
+ {
1484
+ "epoch": 1.02,
1485
+ "learning_rate": 4.795456585360101e-05,
1486
+ "loss": 4.0889,
1487
+ "step": 124928
1488
+ },
1489
+ {
1490
+ "epoch": 1.02,
1491
+ "learning_rate": 4.794617990609049e-05,
1492
+ "loss": 4.0642,
1493
+ "step": 125440
1494
+ },
1495
+ {
1496
+ "epoch": 1.02,
1497
+ "learning_rate": 4.793781033738371e-05,
1498
+ "loss": 4.073,
1499
+ "step": 125952
1500
+ },
1501
+ {
1502
+ "epoch": 1.02,
1503
+ "learning_rate": 4.792942438987319e-05,
1504
+ "loss": 4.0714,
1505
+ "step": 126464
1506
+ },
1507
+ {
1508
+ "epoch": 1.02,
1509
+ "learning_rate": 4.792103844236267e-05,
1510
+ "loss": 4.0711,
1511
+ "step": 126976
1512
+ },
1513
+ {
1514
+ "epoch": 1.02,
1515
+ "learning_rate": 4.791265249485215e-05,
1516
+ "loss": 4.0835,
1517
+ "step": 127488
1518
+ },
1519
+ {
1520
+ "epoch": 1.02,
1521
+ "learning_rate": 4.790426654734163e-05,
1522
+ "loss": 4.0634,
1523
+ "step": 128000
1524
+ },
1525
+ {
1526
+ "epoch": 1.02,
1527
+ "learning_rate": 4.7895896978634836e-05,
1528
+ "loss": 4.0669,
1529
+ "step": 128512
1530
+ },
1531
+ {
1532
+ "epoch": 1.02,
1533
+ "learning_rate": 4.7887511031124316e-05,
1534
+ "loss": 4.074,
1535
+ "step": 129024
1536
+ },
1537
+ {
1538
+ "epoch": 1.02,
1539
+ "learning_rate": 4.7879125083613796e-05,
1540
+ "loss": 4.0708,
1541
+ "step": 129536
1542
+ },
1543
+ {
1544
+ "epoch": 1.02,
1545
+ "learning_rate": 4.7870739136103276e-05,
1546
+ "loss": 4.0543,
1547
+ "step": 130048
1548
+ },
1549
+ {
1550
+ "epoch": 1.02,
1551
+ "learning_rate": 4.7862369567396485e-05,
1552
+ "loss": 4.0656,
1553
+ "step": 130560
1554
+ },
1555
+ {
1556
+ "epoch": 1.02,
1557
+ "learning_rate": 4.7853983619885965e-05,
1558
+ "loss": 4.0573,
1559
+ "step": 131072
1560
+ },
1561
+ {
1562
+ "epoch": 1.02,
1563
+ "learning_rate": 4.7845597672375445e-05,
1564
+ "loss": 4.0569,
1565
+ "step": 131584
1566
+ },
1567
+ {
1568
+ "epoch": 1.02,
1569
+ "learning_rate": 4.7837211724864925e-05,
1570
+ "loss": 4.0536,
1571
+ "step": 132096
1572
+ },
1573
+ {
1574
+ "epoch": 1.02,
1575
+ "learning_rate": 4.782882577735441e-05,
1576
+ "loss": 4.0531,
1577
+ "step": 132608
1578
+ },
1579
+ {
1580
+ "epoch": 1.02,
1581
+ "learning_rate": 4.782045620864762e-05,
1582
+ "loss": 4.0534,
1583
+ "step": 133120
1584
+ },
1585
+ {
1586
+ "epoch": 1.02,
1587
+ "learning_rate": 4.78120702611371e-05,
1588
+ "loss": 4.0642,
1589
+ "step": 133632
1590
+ },
1591
+ {
1592
+ "epoch": 1.02,
1593
+ "learning_rate": 4.780368431362658e-05,
1594
+ "loss": 4.0494,
1595
+ "step": 134144
1596
+ },
1597
+ {
1598
+ "epoch": 1.02,
1599
+ "learning_rate": 4.779529836611606e-05,
1600
+ "loss": 4.0583,
1601
+ "step": 134656
1602
+ },
1603
+ {
1604
+ "epoch": 1.02,
1605
+ "learning_rate": 4.778692879740927e-05,
1606
+ "loss": 4.0687,
1607
+ "step": 135168
1608
+ },
1609
+ {
1610
+ "epoch": 1.02,
1611
+ "learning_rate": 4.777854284989875e-05,
1612
+ "loss": 4.0629,
1613
+ "step": 135680
1614
+ },
1615
+ {
1616
+ "epoch": 1.02,
1617
+ "learning_rate": 4.777017328119196e-05,
1618
+ "loss": 4.0586,
1619
+ "step": 136192
1620
+ },
1621
+ {
1622
+ "epoch": 1.02,
1623
+ "learning_rate": 4.776178733368144e-05,
1624
+ "loss": 4.0525,
1625
+ "step": 136704
1626
+ },
1627
+ {
1628
+ "epoch": 1.02,
1629
+ "learning_rate": 4.775340138617092e-05,
1630
+ "loss": 4.0492,
1631
+ "step": 137216
1632
+ },
1633
+ {
1634
+ "epoch": 1.02,
1635
+ "learning_rate": 4.77450154386604e-05,
1636
+ "loss": 4.0526,
1637
+ "step": 137728
1638
+ },
1639
+ {
1640
+ "epoch": 1.02,
1641
+ "learning_rate": 4.773662949114988e-05,
1642
+ "loss": 4.0443,
1643
+ "step": 138240
1644
+ },
1645
+ {
1646
+ "epoch": 1.02,
1647
+ "learning_rate": 4.7728259922443094e-05,
1648
+ "loss": 4.0453,
1649
+ "step": 138752
1650
+ },
1651
+ {
1652
+ "epoch": 1.02,
1653
+ "learning_rate": 4.7719873974932574e-05,
1654
+ "loss": 4.046,
1655
+ "step": 139264
1656
+ },
1657
+ {
1658
+ "epoch": 1.02,
1659
+ "learning_rate": 4.7711488027422054e-05,
1660
+ "loss": 4.0504,
1661
+ "step": 139776
1662
+ },
1663
+ {
1664
+ "epoch": 1.02,
1665
+ "learning_rate": 4.7703102079911534e-05,
1666
+ "loss": 4.047,
1667
+ "step": 140288
1668
+ },
1669
+ {
1670
+ "epoch": 1.02,
1671
+ "learning_rate": 4.7694716132401014e-05,
1672
+ "loss": 4.0485,
1673
+ "step": 140800
1674
+ },
1675
+ {
1676
+ "epoch": 1.02,
1677
+ "learning_rate": 4.7686330184890494e-05,
1678
+ "loss": 4.0481,
1679
+ "step": 141312
1680
+ },
1681
+ {
1682
+ "epoch": 1.02,
1683
+ "learning_rate": 4.767794423737997e-05,
1684
+ "loss": 4.0367,
1685
+ "step": 141824
1686
+ },
1687
+ {
1688
+ "epoch": 1.02,
1689
+ "learning_rate": 4.766957466867318e-05,
1690
+ "loss": 4.0365,
1691
+ "step": 142336
1692
+ },
1693
+ {
1694
+ "epoch": 1.02,
1695
+ "learning_rate": 4.766118872116266e-05,
1696
+ "loss": 4.0551,
1697
+ "step": 142848
1698
+ },
1699
+ {
1700
+ "epoch": 1.02,
1701
+ "learning_rate": 4.7652802773652136e-05,
1702
+ "loss": 4.0489,
1703
+ "step": 143360
1704
+ },
1705
+ {
1706
+ "epoch": 1.02,
1707
+ "learning_rate": 4.7644416826141616e-05,
1708
+ "loss": 4.044,
1709
+ "step": 143872
1710
+ },
1711
+ {
1712
+ "epoch": 1.02,
1713
+ "learning_rate": 4.7636030878631096e-05,
1714
+ "loss": 4.0391,
1715
+ "step": 144384
1716
+ },
1717
+ {
1718
+ "epoch": 1.02,
1719
+ "learning_rate": 4.762766130992431e-05,
1720
+ "loss": 4.0439,
1721
+ "step": 144896
1722
+ },
1723
+ {
1724
+ "epoch": 1.02,
1725
+ "learning_rate": 4.761927536241379e-05,
1726
+ "loss": 4.0417,
1727
+ "step": 145408
1728
+ },
1729
+ {
1730
+ "epoch": 1.02,
1731
+ "learning_rate": 4.761088941490327e-05,
1732
+ "loss": 4.0507,
1733
+ "step": 145920
1734
+ },
1735
+ {
1736
+ "epoch": 1.02,
1737
+ "learning_rate": 4.760250346739275e-05,
1738
+ "loss": 4.0414,
1739
+ "step": 146432
1740
+ },
1741
+ {
1742
+ "epoch": 1.02,
1743
+ "learning_rate": 4.759413389868596e-05,
1744
+ "loss": 4.0368,
1745
+ "step": 146944
1746
+ },
1747
+ {
1748
+ "epoch": 1.02,
1749
+ "learning_rate": 4.758574795117544e-05,
1750
+ "loss": 4.0389,
1751
+ "step": 147456
1752
+ },
1753
+ {
1754
+ "epoch": 1.02,
1755
+ "learning_rate": 4.757736200366492e-05,
1756
+ "loss": 4.0347,
1757
+ "step": 147968
1758
+ },
1759
+ {
1760
+ "epoch": 1.02,
1761
+ "learning_rate": 4.75689760561544e-05,
1762
+ "loss": 4.0279,
1763
+ "step": 148480
1764
+ },
1765
+ {
1766
+ "epoch": 1.02,
1767
+ "learning_rate": 4.756060648744761e-05,
1768
+ "loss": 4.0322,
1769
+ "step": 148992
1770
+ },
1771
+ {
1772
+ "epoch": 1.02,
1773
+ "learning_rate": 4.755222053993709e-05,
1774
+ "loss": 4.0382,
1775
+ "step": 149504
1776
+ },
1777
+ {
1778
+ "epoch": 1.02,
1779
+ "learning_rate": 4.754383459242657e-05,
1780
+ "loss": 4.0225,
1781
+ "step": 150016
1782
+ },
1783
+ {
1784
+ "epoch": 1.02,
1785
+ "learning_rate": 4.753544864491605e-05,
1786
+ "loss": 4.0408,
1787
+ "step": 150528
1788
+ },
1789
+ {
1790
+ "epoch": 1.02,
1791
+ "learning_rate": 4.7527079076209266e-05,
1792
+ "loss": 4.0322,
1793
+ "step": 151040
1794
+ },
1795
+ {
1796
+ "epoch": 1.02,
1797
+ "learning_rate": 4.7518693128698745e-05,
1798
+ "loss": 4.0297,
1799
+ "step": 151552
1800
+ },
1801
+ {
1802
+ "epoch": 1.02,
1803
+ "learning_rate": 4.7510307181188225e-05,
1804
+ "loss": 4.0355,
1805
+ "step": 152064
1806
+ },
1807
+ {
1808
+ "epoch": 1.02,
1809
+ "learning_rate": 4.7501921233677705e-05,
1810
+ "loss": 4.022,
1811
+ "step": 152576
1812
+ },
1813
+ {
1814
+ "epoch": 1.03,
1815
+ "eval_loss": 4.027729034423828,
1816
+ "eval_runtime": 302.5198,
1817
+ "eval_samples_per_second": 1261.375,
1818
+ "eval_steps_per_second": 39.419,
1819
+ "step": 152640
1820
+ },
1821
+ {
1822
+ "epoch": 0.0,
1823
+ "learning_rate": 4.7493535286167185e-05,
1824
+ "loss": 4.0091,
1825
+ "step": 153088
1826
+ },
1827
+ {
1828
+ "epoch": 0.0,
1829
+ "learning_rate": 4.7485149338656665e-05,
1830
+ "loss": 4.0107,
1831
+ "step": 153600
1832
+ },
1833
+ {
1834
+ "epoch": 0.0,
1835
+ "learning_rate": 4.7476763391146145e-05,
1836
+ "loss": 4.0363,
1837
+ "step": 154112
1838
+ },
1839
+ {
1840
+ "epoch": 0.0,
1841
+ "learning_rate": 4.7468377443635625e-05,
1842
+ "loss": 4.0207,
1843
+ "step": 154624
1844
+ },
1845
+ {
1846
+ "epoch": 0.0,
1847
+ "learning_rate": 4.7459991496125105e-05,
1848
+ "loss": 4.0299,
1849
+ "step": 155136
1850
+ },
1851
+ {
1852
+ "epoch": 0.0,
1853
+ "learning_rate": 4.7451605548614585e-05,
1854
+ "loss": 4.0155,
1855
+ "step": 155648
1856
+ },
1857
+ {
1858
+ "epoch": 0.0,
1859
+ "learning_rate": 4.7443219601104065e-05,
1860
+ "loss": 4.0168,
1861
+ "step": 156160
1862
+ },
1863
+ {
1864
+ "epoch": 0.0,
1865
+ "learning_rate": 4.7434833653593545e-05,
1866
+ "loss": 4.0049,
1867
+ "step": 156672
1868
+ },
1869
+ {
1870
+ "epoch": 0.0,
1871
+ "learning_rate": 4.7426447706083025e-05,
1872
+ "loss": 4.0167,
1873
+ "step": 157184
1874
+ },
1875
+ {
1876
+ "epoch": 0.0,
1877
+ "learning_rate": 4.7418061758572505e-05,
1878
+ "loss": 4.0108,
1879
+ "step": 157696
1880
+ },
1881
+ {
1882
+ "epoch": 0.0,
1883
+ "learning_rate": 4.7409675811061985e-05,
1884
+ "loss": 4.0314,
1885
+ "step": 158208
1886
+ },
1887
+ {
1888
+ "epoch": 0.0,
1889
+ "learning_rate": 4.7401289863551465e-05,
1890
+ "loss": 4.0211,
1891
+ "step": 158720
1892
+ },
1893
+ {
1894
+ "epoch": 0.0,
1895
+ "learning_rate": 4.739293667364841e-05,
1896
+ "loss": 3.9997,
1897
+ "step": 159232
1898
+ },
1899
+ {
1900
+ "epoch": 0.0,
1901
+ "learning_rate": 4.738455072613789e-05,
1902
+ "loss": 4.0054,
1903
+ "step": 159744
1904
+ },
1905
+ {
1906
+ "epoch": 0.0,
1907
+ "learning_rate": 4.737616477862737e-05,
1908
+ "loss": 4.0103,
1909
+ "step": 160256
1910
+ },
1911
+ {
1912
+ "epoch": 0.0,
1913
+ "learning_rate": 4.736777883111685e-05,
1914
+ "loss": 3.9944,
1915
+ "step": 160768
1916
+ },
1917
+ {
1918
+ "epoch": 0.0,
1919
+ "learning_rate": 4.735939288360632e-05,
1920
+ "loss": 4.0083,
1921
+ "step": 161280
1922
+ },
1923
+ {
1924
+ "epoch": 0.0,
1925
+ "learning_rate": 4.73510069360958e-05,
1926
+ "loss": 4.0086,
1927
+ "step": 161792
1928
+ },
1929
+ {
1930
+ "epoch": 0.0,
1931
+ "learning_rate": 4.734262098858528e-05,
1932
+ "loss": 3.9988,
1933
+ "step": 162304
1934
+ },
1935
+ {
1936
+ "epoch": 0.0,
1937
+ "learning_rate": 4.733423504107476e-05,
1938
+ "loss": 4.0262,
1939
+ "step": 162816
1940
+ },
1941
+ {
1942
+ "epoch": 0.0,
1943
+ "learning_rate": 4.732586547236797e-05,
1944
+ "loss": 4.0019,
1945
+ "step": 163328
1946
+ },
1947
+ {
1948
+ "epoch": 0.0,
1949
+ "learning_rate": 4.731747952485746e-05,
1950
+ "loss": 4.0083,
1951
+ "step": 163840
1952
+ },
1953
+ {
1954
+ "epoch": 0.0,
1955
+ "learning_rate": 4.730909357734694e-05,
1956
+ "loss": 3.999,
1957
+ "step": 164352
1958
+ },
1959
+ {
1960
+ "epoch": 0.0,
1961
+ "learning_rate": 4.730070762983642e-05,
1962
+ "loss": 4.0161,
1963
+ "step": 164864
1964
+ },
1965
+ {
1966
+ "epoch": 0.0,
1967
+ "learning_rate": 4.729233806112963e-05,
1968
+ "loss": 3.9915,
1969
+ "step": 165376
1970
+ },
1971
+ {
1972
+ "epoch": 0.0,
1973
+ "learning_rate": 4.728395211361911e-05,
1974
+ "loss": 4.0005,
1975
+ "step": 165888
1976
+ },
1977
+ {
1978
+ "epoch": 0.0,
1979
+ "learning_rate": 4.727556616610859e-05,
1980
+ "loss": 3.9929,
1981
+ "step": 166400
1982
+ },
1983
+ {
1984
+ "epoch": 0.0,
1985
+ "learning_rate": 4.726718021859807e-05,
1986
+ "loss": 3.998,
1987
+ "step": 166912
1988
+ },
1989
+ {
1990
+ "epoch": 0.0,
1991
+ "learning_rate": 4.7258810649891277e-05,
1992
+ "loss": 3.9925,
1993
+ "step": 167424
1994
+ },
1995
+ {
1996
+ "epoch": 0.01,
1997
+ "learning_rate": 4.7250424702380757e-05,
1998
+ "loss": 3.9929,
1999
+ "step": 167936
2000
+ },
2001
+ {
2002
+ "epoch": 0.01,
2003
+ "learning_rate": 4.7242038754870237e-05,
2004
+ "loss": 3.9991,
2005
+ "step": 168448
2006
+ },
2007
+ {
2008
+ "epoch": 0.01,
2009
+ "learning_rate": 4.7233652807359716e-05,
2010
+ "loss": 3.997,
2011
+ "step": 168960
2012
+ },
2013
+ {
2014
+ "epoch": 0.01,
2015
+ "learning_rate": 4.7225283238652926e-05,
2016
+ "loss": 3.9921,
2017
+ "step": 169472
2018
+ },
2019
+ {
2020
+ "epoch": 0.01,
2021
+ "learning_rate": 4.721689729114241e-05,
2022
+ "loss": 4.0027,
2023
+ "step": 169984
2024
+ },
2025
+ {
2026
+ "epoch": 0.01,
2027
+ "learning_rate": 4.720851134363189e-05,
2028
+ "loss": 3.9928,
2029
+ "step": 170496
2030
+ },
2031
+ {
2032
+ "epoch": 0.01,
2033
+ "learning_rate": 4.720012539612137e-05,
2034
+ "loss": 3.9981,
2035
+ "step": 171008
2036
+ },
2037
+ {
2038
+ "epoch": 0.01,
2039
+ "learning_rate": 4.719175582741458e-05,
2040
+ "loss": 3.9835,
2041
+ "step": 171520
2042
+ },
2043
+ {
2044
+ "epoch": 0.01,
2045
+ "learning_rate": 4.718336987990406e-05,
2046
+ "loss": 3.9878,
2047
+ "step": 172032
2048
+ },
2049
+ {
2050
+ "epoch": 0.01,
2051
+ "learning_rate": 4.717498393239354e-05,
2052
+ "loss": 3.9749,
2053
+ "step": 172544
2054
+ },
2055
+ {
2056
+ "epoch": 0.01,
2057
+ "learning_rate": 4.716659798488302e-05,
2058
+ "loss": 3.9898,
2059
+ "step": 173056
2060
+ },
2061
+ {
2062
+ "epoch": 0.01,
2063
+ "learning_rate": 4.715822841617623e-05,
2064
+ "loss": 3.9765,
2065
+ "step": 173568
2066
+ },
2067
+ {
2068
+ "epoch": 0.01,
2069
+ "learning_rate": 4.714984246866571e-05,
2070
+ "loss": 3.9905,
2071
+ "step": 174080
2072
+ },
2073
+ {
2074
+ "epoch": 0.01,
2075
+ "learning_rate": 4.714145652115519e-05,
2076
+ "loss": 3.9759,
2077
+ "step": 174592
2078
+ },
2079
+ {
2080
+ "epoch": 0.01,
2081
+ "learning_rate": 4.713307057364467e-05,
2082
+ "loss": 3.9871,
2083
+ "step": 175104
2084
+ },
2085
+ {
2086
+ "epoch": 0.01,
2087
+ "learning_rate": 4.712470100493788e-05,
2088
+ "loss": 3.9899,
2089
+ "step": 175616
2090
+ },
2091
+ {
2092
+ "epoch": 0.01,
2093
+ "learning_rate": 4.7116315057427366e-05,
2094
+ "loss": 3.9842,
2095
+ "step": 176128
2096
+ },
2097
+ {
2098
+ "epoch": 0.01,
2099
+ "learning_rate": 4.7107929109916846e-05,
2100
+ "loss": 3.9856,
2101
+ "step": 176640
2102
+ },
2103
+ {
2104
+ "epoch": 0.01,
2105
+ "learning_rate": 4.7099543162406326e-05,
2106
+ "loss": 3.9708,
2107
+ "step": 177152
2108
+ },
2109
+ {
2110
+ "epoch": 0.01,
2111
+ "learning_rate": 4.7091173593699535e-05,
2112
+ "loss": 3.9721,
2113
+ "step": 177664
2114
+ },
2115
+ {
2116
+ "epoch": 0.01,
2117
+ "learning_rate": 4.7082787646189015e-05,
2118
+ "loss": 3.9872,
2119
+ "step": 178176
2120
+ },
2121
+ {
2122
+ "epoch": 0.01,
2123
+ "learning_rate": 4.7074401698678495e-05,
2124
+ "loss": 3.9778,
2125
+ "step": 178688
2126
+ },
2127
+ {
2128
+ "epoch": 0.01,
2129
+ "learning_rate": 4.7066015751167975e-05,
2130
+ "loss": 3.9775,
2131
+ "step": 179200
2132
+ },
2133
+ {
2134
+ "epoch": 0.01,
2135
+ "learning_rate": 4.7057646182461184e-05,
2136
+ "loss": 3.9644,
2137
+ "step": 179712
2138
+ },
2139
+ {
2140
+ "epoch": 0.01,
2141
+ "learning_rate": 4.704927661375439e-05,
2142
+ "loss": 3.9781,
2143
+ "step": 180224
2144
+ },
2145
+ {
2146
+ "epoch": 0.01,
2147
+ "learning_rate": 4.704089066624387e-05,
2148
+ "loss": 3.9544,
2149
+ "step": 180736
2150
+ },
2151
+ {
2152
+ "epoch": 0.01,
2153
+ "learning_rate": 4.703250471873335e-05,
2154
+ "loss": 3.9878,
2155
+ "step": 181248
2156
+ },
2157
+ {
2158
+ "epoch": 0.01,
2159
+ "learning_rate": 4.702411877122283e-05,
2160
+ "loss": 3.9558,
2161
+ "step": 181760
2162
+ },
2163
+ {
2164
+ "epoch": 0.01,
2165
+ "learning_rate": 4.701573282371232e-05,
2166
+ "loss": 3.9763,
2167
+ "step": 182272
2168
+ },
2169
+ {
2170
+ "epoch": 0.01,
2171
+ "learning_rate": 4.70073468762018e-05,
2172
+ "loss": 3.9773,
2173
+ "step": 182784
2174
+ },
2175
+ {
2176
+ "epoch": 0.01,
2177
+ "learning_rate": 4.699896092869128e-05,
2178
+ "loss": 3.9599,
2179
+ "step": 183296
2180
+ },
2181
+ {
2182
+ "epoch": 0.01,
2183
+ "learning_rate": 4.699057498118076e-05,
2184
+ "loss": 3.9784,
2185
+ "step": 183808
2186
+ },
2187
+ {
2188
+ "epoch": 0.01,
2189
+ "learning_rate": 4.698220541247397e-05,
2190
+ "loss": 3.9628,
2191
+ "step": 184320
2192
+ },
2193
+ {
2194
+ "epoch": 0.01,
2195
+ "learning_rate": 4.697381946496345e-05,
2196
+ "loss": 3.964,
2197
+ "step": 184832
2198
+ },
2199
+ {
2200
+ "epoch": 0.01,
2201
+ "learning_rate": 4.696543351745293e-05,
2202
+ "loss": 3.952,
2203
+ "step": 185344
2204
+ },
2205
+ {
2206
+ "epoch": 0.01,
2207
+ "learning_rate": 4.695704756994241e-05,
2208
+ "loss": 3.9766,
2209
+ "step": 185856
2210
+ },
2211
+ {
2212
+ "epoch": 0.01,
2213
+ "learning_rate": 4.694867800123562e-05,
2214
+ "loss": 3.9557,
2215
+ "step": 186368
2216
+ },
2217
+ {
2218
+ "epoch": 0.01,
2219
+ "learning_rate": 4.69402920537251e-05,
2220
+ "loss": 3.9609,
2221
+ "step": 186880
2222
+ },
2223
+ {
2224
+ "epoch": 0.01,
2225
+ "learning_rate": 4.693190610621458e-05,
2226
+ "loss": 3.9473,
2227
+ "step": 187392
2228
+ },
2229
+ {
2230
+ "epoch": 0.01,
2231
+ "learning_rate": 4.692352015870406e-05,
2232
+ "loss": 3.9563,
2233
+ "step": 187904
2234
+ },
2235
+ {
2236
+ "epoch": 0.01,
2237
+ "learning_rate": 4.6915150589997267e-05,
2238
+ "loss": 3.9664,
2239
+ "step": 188416
2240
+ },
2241
+ {
2242
+ "epoch": 0.01,
2243
+ "learning_rate": 4.690676464248675e-05,
2244
+ "loss": 3.9695,
2245
+ "step": 188928
2246
+ },
2247
+ {
2248
+ "epoch": 0.01,
2249
+ "learning_rate": 4.689837869497623e-05,
2250
+ "loss": 3.9581,
2251
+ "step": 189440
2252
+ },
2253
+ {
2254
+ "epoch": 0.01,
2255
+ "learning_rate": 4.688999274746571e-05,
2256
+ "loss": 3.9721,
2257
+ "step": 189952
2258
+ },
2259
+ {
2260
+ "epoch": 0.01,
2261
+ "learning_rate": 4.688162317875892e-05,
2262
+ "loss": 3.9701,
2263
+ "step": 190464
2264
+ },
2265
+ {
2266
+ "epoch": 0.01,
2267
+ "learning_rate": 4.68732372312484e-05,
2268
+ "loss": 3.9617,
2269
+ "step": 190976
2270
+ },
2271
+ {
2272
+ "epoch": 0.01,
2273
+ "learning_rate": 4.686485128373788e-05,
2274
+ "loss": 3.9624,
2275
+ "step": 191488
2276
+ },
2277
+ {
2278
+ "epoch": 0.01,
2279
+ "learning_rate": 4.685646533622736e-05,
2280
+ "loss": 3.9597,
2281
+ "step": 192000
2282
+ },
2283
+ {
2284
+ "epoch": 0.01,
2285
+ "learning_rate": 4.684809576752057e-05,
2286
+ "loss": 3.9542,
2287
+ "step": 192512
2288
+ },
2289
+ {
2290
+ "epoch": 0.01,
2291
+ "learning_rate": 4.683970982001005e-05,
2292
+ "loss": 3.954,
2293
+ "step": 193024
2294
+ },
2295
+ {
2296
+ "epoch": 0.01,
2297
+ "learning_rate": 4.683132387249953e-05,
2298
+ "loss": 3.96,
2299
+ "step": 193536
2300
+ },
2301
+ {
2302
+ "epoch": 0.01,
2303
+ "learning_rate": 4.682293792498901e-05,
2304
+ "loss": 3.9627,
2305
+ "step": 194048
2306
+ },
2307
+ {
2308
+ "epoch": 0.01,
2309
+ "learning_rate": 4.681456835628222e-05,
2310
+ "loss": 3.9699,
2311
+ "step": 194560
2312
+ },
2313
+ {
2314
+ "epoch": 0.01,
2315
+ "learning_rate": 4.680618240877171e-05,
2316
+ "loss": 3.9544,
2317
+ "step": 195072
2318
+ },
2319
+ {
2320
+ "epoch": 0.01,
2321
+ "learning_rate": 4.679779646126119e-05,
2322
+ "loss": 3.9353,
2323
+ "step": 195584
2324
+ },
2325
+ {
2326
+ "epoch": 0.01,
2327
+ "learning_rate": 4.678941051375067e-05,
2328
+ "loss": 3.9439,
2329
+ "step": 196096
2330
+ },
2331
+ {
2332
+ "epoch": 0.01,
2333
+ "learning_rate": 4.6781040945043876e-05,
2334
+ "loss": 3.9539,
2335
+ "step": 196608
2336
+ },
2337
+ {
2338
+ "epoch": 0.01,
2339
+ "learning_rate": 4.6772654997533356e-05,
2340
+ "loss": 3.9505,
2341
+ "step": 197120
2342
+ },
2343
+ {
2344
+ "epoch": 0.01,
2345
+ "learning_rate": 4.6764269050022836e-05,
2346
+ "loss": 3.9479,
2347
+ "step": 197632
2348
+ },
2349
+ {
2350
+ "epoch": 0.01,
2351
+ "learning_rate": 4.6755883102512316e-05,
2352
+ "loss": 3.9543,
2353
+ "step": 198144
2354
+ },
2355
+ {
2356
+ "epoch": 0.02,
2357
+ "learning_rate": 4.6747513533805525e-05,
2358
+ "loss": 3.9489,
2359
+ "step": 198656
2360
+ },
2361
+ {
2362
+ "epoch": 0.02,
2363
+ "learning_rate": 4.6739127586295005e-05,
2364
+ "loss": 3.9393,
2365
+ "step": 199168
2366
+ },
2367
+ {
2368
+ "epoch": 0.02,
2369
+ "learning_rate": 4.6730741638784485e-05,
2370
+ "loss": 3.9497,
2371
+ "step": 199680
2372
+ },
2373
+ {
2374
+ "epoch": 0.02,
2375
+ "learning_rate": 4.6722355691273965e-05,
2376
+ "loss": 3.9589,
2377
+ "step": 200192
2378
+ },
2379
+ {
2380
+ "epoch": 0.02,
2381
+ "learning_rate": 4.6713986122567174e-05,
2382
+ "loss": 3.9531,
2383
+ "step": 200704
2384
+ },
2385
+ {
2386
+ "epoch": 0.02,
2387
+ "learning_rate": 4.670560017505666e-05,
2388
+ "loss": 3.9583,
2389
+ "step": 201216
2390
+ },
2391
+ {
2392
+ "epoch": 0.02,
2393
+ "learning_rate": 4.669721422754614e-05,
2394
+ "loss": 3.9415,
2395
+ "step": 201728
2396
+ },
2397
+ {
2398
+ "epoch": 0.02,
2399
+ "learning_rate": 4.668882828003562e-05,
2400
+ "loss": 3.9446,
2401
+ "step": 202240
2402
+ },
2403
+ {
2404
+ "epoch": 0.02,
2405
+ "learning_rate": 4.668045871132883e-05,
2406
+ "loss": 3.9413,
2407
+ "step": 202752
2408
+ },
2409
+ {
2410
+ "epoch": 0.02,
2411
+ "learning_rate": 4.667207276381831e-05,
2412
+ "loss": 3.9457,
2413
+ "step": 203264
2414
+ },
2415
+ {
2416
+ "epoch": 0.02,
2417
+ "learning_rate": 4.666368681630779e-05,
2418
+ "loss": 3.9545,
2419
+ "step": 203776
2420
+ },
2421
+ {
2422
+ "epoch": 0.02,
2423
+ "learning_rate": 4.665530086879727e-05,
2424
+ "loss": 3.9385,
2425
+ "step": 204288
2426
+ },
2427
+ {
2428
+ "epoch": 0.02,
2429
+ "learning_rate": 4.664693130009048e-05,
2430
+ "loss": 3.9397,
2431
+ "step": 204800
2432
+ },
2433
+ {
2434
+ "epoch": 0.02,
2435
+ "learning_rate": 4.663854535257996e-05,
2436
+ "loss": 3.9476,
2437
+ "step": 205312
2438
+ },
2439
+ {
2440
+ "epoch": 0.02,
2441
+ "learning_rate": 4.663015940506944e-05,
2442
+ "loss": 3.9488,
2443
+ "step": 205824
2444
+ },
2445
+ {
2446
+ "epoch": 0.02,
2447
+ "learning_rate": 4.662177345755892e-05,
2448
+ "loss": 3.9306,
2449
+ "step": 206336
2450
+ },
2451
+ {
2452
+ "epoch": 0.02,
2453
+ "learning_rate": 4.6613420267655863e-05,
2454
+ "loss": 3.9381,
2455
+ "step": 206848
2456
+ },
2457
+ {
2458
+ "epoch": 0.02,
2459
+ "learning_rate": 4.660503432014534e-05,
2460
+ "loss": 3.9392,
2461
+ "step": 207360
2462
+ },
2463
+ {
2464
+ "epoch": 0.02,
2465
+ "learning_rate": 4.659664837263482e-05,
2466
+ "loss": 3.9339,
2467
+ "step": 207872
2468
+ },
2469
+ {
2470
+ "epoch": 0.02,
2471
+ "learning_rate": 4.65882624251243e-05,
2472
+ "loss": 3.9298,
2473
+ "step": 208384
2474
+ },
2475
+ {
2476
+ "epoch": 0.02,
2477
+ "learning_rate": 4.657987647761378e-05,
2478
+ "loss": 3.9321,
2479
+ "step": 208896
2480
+ },
2481
+ {
2482
+ "epoch": 0.02,
2483
+ "learning_rate": 4.657149053010326e-05,
2484
+ "loss": 3.9358,
2485
+ "step": 209408
2486
+ },
2487
+ {
2488
+ "epoch": 0.02,
2489
+ "learning_rate": 4.656310458259274e-05,
2490
+ "loss": 3.938,
2491
+ "step": 209920
2492
+ },
2493
+ {
2494
+ "epoch": 0.02,
2495
+ "learning_rate": 4.655471863508222e-05,
2496
+ "loss": 3.929,
2497
+ "step": 210432
2498
+ },
2499
+ {
2500
+ "epoch": 0.02,
2501
+ "learning_rate": 4.654634906637543e-05,
2502
+ "loss": 3.9371,
2503
+ "step": 210944
2504
+ },
2505
+ {
2506
+ "epoch": 0.02,
2507
+ "learning_rate": 4.653796311886491e-05,
2508
+ "loss": 3.9474,
2509
+ "step": 211456
2510
+ },
2511
+ {
2512
+ "epoch": 0.02,
2513
+ "learning_rate": 4.652957717135439e-05,
2514
+ "loss": 3.949,
2515
+ "step": 211968
2516
+ },
2517
+ {
2518
+ "epoch": 0.02,
2519
+ "learning_rate": 4.65212076026476e-05,
2520
+ "loss": 3.9373,
2521
+ "step": 212480
2522
+ },
2523
+ {
2524
+ "epoch": 0.02,
2525
+ "learning_rate": 4.651282165513708e-05,
2526
+ "loss": 3.9367,
2527
+ "step": 212992
2528
+ },
2529
+ {
2530
+ "epoch": 0.02,
2531
+ "learning_rate": 4.650443570762657e-05,
2532
+ "loss": 3.9314,
2533
+ "step": 213504
2534
+ },
2535
+ {
2536
+ "epoch": 0.02,
2537
+ "learning_rate": 4.649604976011605e-05,
2538
+ "loss": 3.9355,
2539
+ "step": 214016
2540
+ },
2541
+ {
2542
+ "epoch": 0.02,
2543
+ "learning_rate": 4.648766381260553e-05,
2544
+ "loss": 3.9252,
2545
+ "step": 214528
2546
+ },
2547
+ {
2548
+ "epoch": 0.02,
2549
+ "learning_rate": 4.647929424389874e-05,
2550
+ "loss": 3.9278,
2551
+ "step": 215040
2552
+ },
2553
+ {
2554
+ "epoch": 0.02,
2555
+ "learning_rate": 4.647090829638822e-05,
2556
+ "loss": 3.9285,
2557
+ "step": 215552
2558
+ },
2559
+ {
2560
+ "epoch": 0.02,
2561
+ "learning_rate": 4.64625223488777e-05,
2562
+ "loss": 3.938,
2563
+ "step": 216064
2564
+ },
2565
+ {
2566
+ "epoch": 0.02,
2567
+ "learning_rate": 4.645413640136718e-05,
2568
+ "loss": 3.9302,
2569
+ "step": 216576
2570
+ },
2571
+ {
2572
+ "epoch": 0.02,
2573
+ "learning_rate": 4.644575045385665e-05,
2574
+ "loss": 3.9327,
2575
+ "step": 217088
2576
+ },
2577
+ {
2578
+ "epoch": 0.02,
2579
+ "learning_rate": 4.643736450634613e-05,
2580
+ "loss": 3.9322,
2581
+ "step": 217600
2582
+ },
2583
+ {
2584
+ "epoch": 0.02,
2585
+ "learning_rate": 4.642897855883561e-05,
2586
+ "loss": 3.9229,
2587
+ "step": 218112
2588
+ },
2589
+ {
2590
+ "epoch": 0.02,
2591
+ "learning_rate": 4.642059261132509e-05,
2592
+ "loss": 3.9205,
2593
+ "step": 218624
2594
+ },
2595
+ {
2596
+ "epoch": 0.02,
2597
+ "learning_rate": 4.6412223042618306e-05,
2598
+ "loss": 3.9408,
2599
+ "step": 219136
2600
+ },
2601
+ {
2602
+ "epoch": 0.02,
2603
+ "learning_rate": 4.6403837095107786e-05,
2604
+ "loss": 3.9383,
2605
+ "step": 219648
2606
+ },
2607
+ {
2608
+ "epoch": 0.02,
2609
+ "learning_rate": 4.6395451147597266e-05,
2610
+ "loss": 3.9303,
2611
+ "step": 220160
2612
+ },
2613
+ {
2614
+ "epoch": 0.02,
2615
+ "learning_rate": 4.6387081578890475e-05,
2616
+ "loss": 3.9268,
2617
+ "step": 220672
2618
+ },
2619
+ {
2620
+ "epoch": 0.02,
2621
+ "learning_rate": 4.637871201018369e-05,
2622
+ "loss": 3.9271,
2623
+ "step": 221184
2624
+ },
2625
+ {
2626
+ "epoch": 0.02,
2627
+ "learning_rate": 4.637032606267317e-05,
2628
+ "loss": 3.9306,
2629
+ "step": 221696
2630
+ },
2631
+ {
2632
+ "epoch": 0.02,
2633
+ "learning_rate": 4.636194011516265e-05,
2634
+ "loss": 3.9418,
2635
+ "step": 222208
2636
+ },
2637
+ {
2638
+ "epoch": 0.02,
2639
+ "learning_rate": 4.6353554167652124e-05,
2640
+ "loss": 3.929,
2641
+ "step": 222720
2642
+ },
2643
+ {
2644
+ "epoch": 0.02,
2645
+ "learning_rate": 4.6345168220141604e-05,
2646
+ "loss": 3.9228,
2647
+ "step": 223232
2648
+ },
2649
+ {
2650
+ "epoch": 0.02,
2651
+ "learning_rate": 4.6336782272631084e-05,
2652
+ "loss": 3.9264,
2653
+ "step": 223744
2654
+ },
2655
+ {
2656
+ "epoch": 0.02,
2657
+ "learning_rate": 4.6328396325120564e-05,
2658
+ "loss": 3.9232,
2659
+ "step": 224256
2660
+ },
2661
+ {
2662
+ "epoch": 0.02,
2663
+ "learning_rate": 4.6320010377610043e-05,
2664
+ "loss": 3.9204,
2665
+ "step": 224768
2666
+ },
2667
+ {
2668
+ "epoch": 0.02,
2669
+ "learning_rate": 4.631164080890326e-05,
2670
+ "loss": 3.9203,
2671
+ "step": 225280
2672
+ },
2673
+ {
2674
+ "epoch": 0.02,
2675
+ "learning_rate": 4.630325486139274e-05,
2676
+ "loss": 3.9275,
2677
+ "step": 225792
2678
+ },
2679
+ {
2680
+ "epoch": 0.02,
2681
+ "learning_rate": 4.629486891388222e-05,
2682
+ "loss": 3.9163,
2683
+ "step": 226304
2684
+ },
2685
+ {
2686
+ "epoch": 0.02,
2687
+ "learning_rate": 4.62864829663717e-05,
2688
+ "loss": 3.9289,
2689
+ "step": 226816
2690
+ },
2691
+ {
2692
+ "epoch": 0.02,
2693
+ "learning_rate": 4.627811339766491e-05,
2694
+ "loss": 3.9262,
2695
+ "step": 227328
2696
+ },
2697
+ {
2698
+ "epoch": 0.02,
2699
+ "learning_rate": 4.626972745015439e-05,
2700
+ "loss": 3.9209,
2701
+ "step": 227840
2702
+ },
2703
+ {
2704
+ "epoch": 0.02,
2705
+ "learning_rate": 4.626134150264387e-05,
2706
+ "loss": 3.9269,
2707
+ "step": 228352
2708
+ },
2709
+ {
2710
+ "epoch": 0.02,
2711
+ "learning_rate": 4.625295555513335e-05,
2712
+ "loss": 3.9114,
2713
+ "step": 228864
2714
+ },
2715
+ {
2716
+ "epoch": 0.03,
2717
+ "eval_loss": 3.95308780670166,
2718
+ "eval_runtime": 304.6541,
2719
+ "eval_samples_per_second": 1252.539,
2720
+ "eval_steps_per_second": 39.143,
2721
+ "step": 228960
2722
+ },
2723
+ {
2724
+ "epoch": 1.0,
2725
+ "learning_rate": 4.624458598642656e-05,
2726
+ "loss": 3.915,
2727
+ "step": 229376
2728
+ },
2729
+ {
2730
+ "epoch": 1.0,
2731
+ "learning_rate": 4.623620003891604e-05,
2732
+ "loss": 3.9042,
2733
+ "step": 229888
2734
+ },
2735
+ {
2736
+ "epoch": 1.0,
2737
+ "learning_rate": 4.622781409140552e-05,
2738
+ "loss": 3.9287,
2739
+ "step": 230400
2740
+ },
2741
+ {
2742
+ "epoch": 1.0,
2743
+ "learning_rate": 4.6219428143895e-05,
2744
+ "loss": 3.9135,
2745
+ "step": 230912
2746
+ },
2747
+ {
2748
+ "epoch": 1.0,
2749
+ "learning_rate": 4.621105857518821e-05,
2750
+ "loss": 3.9283,
2751
+ "step": 231424
2752
+ },
2753
+ {
2754
+ "epoch": 1.0,
2755
+ "learning_rate": 4.620267262767769e-05,
2756
+ "loss": 3.9081,
2757
+ "step": 231936
2758
+ },
2759
+ {
2760
+ "epoch": 1.0,
2761
+ "learning_rate": 4.619428668016717e-05,
2762
+ "loss": 3.9171,
2763
+ "step": 232448
2764
+ },
2765
+ {
2766
+ "epoch": 1.0,
2767
+ "learning_rate": 4.618591711146038e-05,
2768
+ "loss": 3.8989,
2769
+ "step": 232960
2770
+ },
2771
+ {
2772
+ "epoch": 1.0,
2773
+ "learning_rate": 4.617753116394986e-05,
2774
+ "loss": 3.9165,
2775
+ "step": 233472
2776
+ },
2777
+ {
2778
+ "epoch": 1.0,
2779
+ "learning_rate": 4.616914521643934e-05,
2780
+ "loss": 3.9041,
2781
+ "step": 233984
2782
+ },
2783
+ {
2784
+ "epoch": 1.0,
2785
+ "learning_rate": 4.616077564773255e-05,
2786
+ "loss": 3.9276,
2787
+ "step": 234496
2788
+ },
2789
+ {
2790
+ "epoch": 1.0,
2791
+ "learning_rate": 4.615238970022203e-05,
2792
+ "loss": 3.9171,
2793
+ "step": 235008
2794
+ },
2795
+ {
2796
+ "epoch": 1.0,
2797
+ "learning_rate": 4.614402013151524e-05,
2798
+ "loss": 3.8957,
2799
+ "step": 235520
2800
+ },
2801
+ {
2802
+ "epoch": 1.0,
2803
+ "learning_rate": 4.613563418400472e-05,
2804
+ "loss": 3.9043,
2805
+ "step": 236032
2806
+ },
2807
+ {
2808
+ "epoch": 1.0,
2809
+ "learning_rate": 4.61272482364942e-05,
2810
+ "loss": 3.9104,
2811
+ "step": 236544
2812
+ },
2813
+ {
2814
+ "epoch": 1.0,
2815
+ "learning_rate": 4.611886228898368e-05,
2816
+ "loss": 3.8943,
2817
+ "step": 237056
2818
+ },
2819
+ {
2820
+ "epoch": 1.0,
2821
+ "learning_rate": 4.611047634147316e-05,
2822
+ "loss": 3.9046,
2823
+ "step": 237568
2824
+ },
2825
+ {
2826
+ "epoch": 1.0,
2827
+ "learning_rate": 4.610209039396265e-05,
2828
+ "loss": 3.9067,
2829
+ "step": 238080
2830
+ },
2831
+ {
2832
+ "epoch": 1.0,
2833
+ "learning_rate": 4.609370444645213e-05,
2834
+ "loss": 3.8982,
2835
+ "step": 238592
2836
+ },
2837
+ {
2838
+ "epoch": 1.0,
2839
+ "learning_rate": 4.6085318498941607e-05,
2840
+ "loss": 3.9294,
2841
+ "step": 239104
2842
+ },
2843
+ {
2844
+ "epoch": 1.0,
2845
+ "learning_rate": 4.6076932551431087e-05,
2846
+ "loss": 3.902,
2847
+ "step": 239616
2848
+ },
2849
+ {
2850
+ "epoch": 1.0,
2851
+ "learning_rate": 4.6068546603920566e-05,
2852
+ "loss": 3.9048,
2853
+ "step": 240128
2854
+ },
2855
+ {
2856
+ "epoch": 1.0,
2857
+ "learning_rate": 4.6060177035213776e-05,
2858
+ "loss": 3.9026,
2859
+ "step": 240640
2860
+ },
2861
+ {
2862
+ "epoch": 1.0,
2863
+ "learning_rate": 4.6051791087703256e-05,
2864
+ "loss": 3.9106,
2865
+ "step": 241152
2866
+ },
2867
+ {
2868
+ "epoch": 1.0,
2869
+ "learning_rate": 4.6043405140192735e-05,
2870
+ "loss": 3.8958,
2871
+ "step": 241664
2872
+ },
2873
+ {
2874
+ "epoch": 1.0,
2875
+ "learning_rate": 4.6035019192682215e-05,
2876
+ "loss": 3.9016,
2877
+ "step": 242176
2878
+ },
2879
+ {
2880
+ "epoch": 1.0,
2881
+ "learning_rate": 4.6026633245171695e-05,
2882
+ "loss": 3.8971,
2883
+ "step": 242688
2884
+ },
2885
+ {
2886
+ "epoch": 1.0,
2887
+ "learning_rate": 4.6018263676464904e-05,
2888
+ "loss": 3.9002,
2889
+ "step": 243200
2890
+ },
2891
+ {
2892
+ "epoch": 1.0,
2893
+ "learning_rate": 4.6009877728954384e-05,
2894
+ "loss": 3.8941,
2895
+ "step": 243712
2896
+ },
2897
+ {
2898
+ "epoch": 1.01,
2899
+ "learning_rate": 4.6001491781443864e-05,
2900
+ "loss": 3.8956,
2901
+ "step": 244224
2902
+ },
2903
+ {
2904
+ "epoch": 1.01,
2905
+ "learning_rate": 4.5993105833933344e-05,
2906
+ "loss": 3.8977,
2907
+ "step": 244736
2908
+ },
2909
+ {
2910
+ "epoch": 1.01,
2911
+ "learning_rate": 4.598471988642283e-05,
2912
+ "loss": 3.9027,
2913
+ "step": 245248
2914
+ },
2915
+ {
2916
+ "epoch": 1.01,
2917
+ "learning_rate": 4.597633393891231e-05,
2918
+ "loss": 3.9027,
2919
+ "step": 245760
2920
+ },
2921
+ {
2922
+ "epoch": 1.01,
2923
+ "learning_rate": 4.5967947991401784e-05,
2924
+ "loss": 3.9028,
2925
+ "step": 246272
2926
+ },
2927
+ {
2928
+ "epoch": 1.01,
2929
+ "learning_rate": 4.5959578422695e-05,
2930
+ "loss": 3.8961,
2931
+ "step": 246784
2932
+ },
2933
+ {
2934
+ "epoch": 1.01,
2935
+ "learning_rate": 4.595119247518448e-05,
2936
+ "loss": 3.9025,
2937
+ "step": 247296
2938
+ },
2939
+ {
2940
+ "epoch": 1.01,
2941
+ "learning_rate": 4.594280652767395e-05,
2942
+ "loss": 3.8846,
2943
+ "step": 247808
2944
+ },
2945
+ {
2946
+ "epoch": 1.01,
2947
+ "learning_rate": 4.593442058016343e-05,
2948
+ "loss": 3.899,
2949
+ "step": 248320
2950
+ },
2951
+ {
2952
+ "epoch": 1.01,
2953
+ "learning_rate": 4.592603463265291e-05,
2954
+ "loss": 3.8816,
2955
+ "step": 248832
2956
+ },
2957
+ {
2958
+ "epoch": 1.01,
2959
+ "learning_rate": 4.591764868514239e-05,
2960
+ "loss": 3.8918,
2961
+ "step": 249344
2962
+ },
2963
+ {
2964
+ "epoch": 1.01,
2965
+ "learning_rate": 4.590926273763187e-05,
2966
+ "loss": 3.8831,
2967
+ "step": 249856
2968
+ },
2969
+ {
2970
+ "epoch": 1.01,
2971
+ "learning_rate": 4.590087679012135e-05,
2972
+ "loss": 3.9,
2973
+ "step": 250368
2974
+ },
2975
+ {
2976
+ "epoch": 1.01,
2977
+ "learning_rate": 4.58925236002183e-05,
2978
+ "loss": 3.8821,
2979
+ "step": 250880
2980
+ },
2981
+ {
2982
+ "epoch": 1.01,
2983
+ "learning_rate": 4.5884137652707785e-05,
2984
+ "loss": 3.8901,
2985
+ "step": 251392
2986
+ },
2987
+ {
2988
+ "epoch": 1.01,
2989
+ "learning_rate": 4.587575170519726e-05,
2990
+ "loss": 3.9007,
2991
+ "step": 251904
2992
+ },
2993
+ {
2994
+ "epoch": 1.01,
2995
+ "learning_rate": 4.586736575768674e-05,
2996
+ "loss": 3.893,
2997
+ "step": 252416
2998
+ },
2999
+ {
3000
+ "epoch": 1.01,
3001
+ "learning_rate": 4.585897981017622e-05,
3002
+ "loss": 3.8952,
3003
+ "step": 252928
3004
+ },
3005
+ {
3006
+ "epoch": 1.01,
3007
+ "learning_rate": 4.585061024146943e-05,
3008
+ "loss": 3.8799,
3009
+ "step": 253440
3010
+ },
3011
+ {
3012
+ "epoch": 1.01,
3013
+ "learning_rate": 4.584222429395891e-05,
3014
+ "loss": 3.8795,
3015
+ "step": 253952
3016
+ },
3017
+ {
3018
+ "epoch": 1.01,
3019
+ "learning_rate": 4.583383834644839e-05,
3020
+ "loss": 3.8934,
3021
+ "step": 254464
3022
+ },
3023
+ {
3024
+ "epoch": 1.01,
3025
+ "learning_rate": 4.582545239893787e-05,
3026
+ "loss": 3.8872,
3027
+ "step": 254976
3028
+ },
3029
+ {
3030
+ "epoch": 1.01,
3031
+ "learning_rate": 4.581706645142735e-05,
3032
+ "loss": 3.8842,
3033
+ "step": 255488
3034
+ },
3035
+ {
3036
+ "epoch": 1.01,
3037
+ "learning_rate": 4.580868050391683e-05,
3038
+ "loss": 3.8778,
3039
+ "step": 256000
3040
+ },
3041
+ {
3042
+ "epoch": 1.01,
3043
+ "learning_rate": 4.5800310935210036e-05,
3044
+ "loss": 3.8886,
3045
+ "step": 256512
3046
+ },
3047
+ {
3048
+ "epoch": 1.01,
3049
+ "learning_rate": 4.579192498769952e-05,
3050
+ "loss": 3.8676,
3051
+ "step": 257024
3052
+ },
3053
+ {
3054
+ "epoch": 1.01,
3055
+ "learning_rate": 4.5783539040189e-05,
3056
+ "loss": 3.8898,
3057
+ "step": 257536
3058
+ },
3059
+ {
3060
+ "epoch": 1.01,
3061
+ "learning_rate": 4.577515309267848e-05,
3062
+ "loss": 3.8708,
3063
+ "step": 258048
3064
+ },
3065
+ {
3066
+ "epoch": 1.01,
3067
+ "learning_rate": 4.576676714516796e-05,
3068
+ "loss": 3.882,
3069
+ "step": 258560
3070
+ },
3071
+ {
3072
+ "epoch": 1.01,
3073
+ "learning_rate": 4.575838119765744e-05,
3074
+ "loss": 3.892,
3075
+ "step": 259072
3076
+ },
3077
+ {
3078
+ "epoch": 1.01,
3079
+ "learning_rate": 4.574999525014692e-05,
3080
+ "loss": 3.8722,
3081
+ "step": 259584
3082
+ },
3083
+ {
3084
+ "epoch": 1.01,
3085
+ "learning_rate": 4.57416093026364e-05,
3086
+ "loss": 3.8855,
3087
+ "step": 260096
3088
+ },
3089
+ {
3090
+ "epoch": 1.01,
3091
+ "learning_rate": 4.573323973392961e-05,
3092
+ "loss": 3.8773,
3093
+ "step": 260608
3094
+ },
3095
+ {
3096
+ "epoch": 1.01,
3097
+ "learning_rate": 4.572485378641909e-05,
3098
+ "loss": 3.8731,
3099
+ "step": 261120
3100
+ },
3101
+ {
3102
+ "epoch": 1.01,
3103
+ "learning_rate": 4.57164842177123e-05,
3104
+ "loss": 3.8693,
3105
+ "step": 261632
3106
+ },
3107
+ {
3108
+ "epoch": 1.01,
3109
+ "learning_rate": 4.570809827020178e-05,
3110
+ "loss": 3.8849,
3111
+ "step": 262144
3112
+ },
3113
+ {
3114
+ "epoch": 1.01,
3115
+ "learning_rate": 4.569971232269126e-05,
3116
+ "loss": 3.8712,
3117
+ "step": 262656
3118
+ },
3119
+ {
3120
+ "epoch": 1.01,
3121
+ "learning_rate": 4.569132637518074e-05,
3122
+ "loss": 3.8743,
3123
+ "step": 263168
3124
+ },
3125
+ {
3126
+ "epoch": 1.01,
3127
+ "learning_rate": 4.568294042767022e-05,
3128
+ "loss": 3.8621,
3129
+ "step": 263680
3130
+ },
3131
+ {
3132
+ "epoch": 1.01,
3133
+ "learning_rate": 4.567455448015971e-05,
3134
+ "loss": 3.867,
3135
+ "step": 264192
3136
+ },
3137
+ {
3138
+ "epoch": 1.01,
3139
+ "learning_rate": 4.566616853264919e-05,
3140
+ "loss": 3.8787,
3141
+ "step": 264704
3142
+ },
3143
+ {
3144
+ "epoch": 1.01,
3145
+ "learning_rate": 4.5657798963942396e-05,
3146
+ "loss": 3.8837,
3147
+ "step": 265216
3148
+ },
3149
+ {
3150
+ "epoch": 1.01,
3151
+ "learning_rate": 4.5649413016431876e-05,
3152
+ "loss": 3.8725,
3153
+ "step": 265728
3154
+ },
3155
+ {
3156
+ "epoch": 1.01,
3157
+ "learning_rate": 4.5641027068921356e-05,
3158
+ "loss": 3.8872,
3159
+ "step": 266240
3160
+ },
3161
+ {
3162
+ "epoch": 1.01,
3163
+ "learning_rate": 4.5632641121410836e-05,
3164
+ "loss": 3.8814,
3165
+ "step": 266752
3166
+ },
3167
+ {
3168
+ "epoch": 1.01,
3169
+ "learning_rate": 4.5624255173900316e-05,
3170
+ "loss": 3.8741,
3171
+ "step": 267264
3172
+ },
3173
+ {
3174
+ "epoch": 1.01,
3175
+ "learning_rate": 4.561586922638979e-05,
3176
+ "loss": 3.8793,
3177
+ "step": 267776
3178
+ },
3179
+ {
3180
+ "epoch": 1.01,
3181
+ "learning_rate": 4.560748327887927e-05,
3182
+ "loss": 3.8739,
3183
+ "step": 268288
3184
+ },
3185
+ {
3186
+ "epoch": 1.01,
3187
+ "learning_rate": 4.5599113710172485e-05,
3188
+ "loss": 3.8727,
3189
+ "step": 268800
3190
+ },
3191
+ {
3192
+ "epoch": 1.01,
3193
+ "learning_rate": 4.5590727762661965e-05,
3194
+ "loss": 3.8701,
3195
+ "step": 269312
3196
+ },
3197
+ {
3198
+ "epoch": 1.01,
3199
+ "learning_rate": 4.5582341815151445e-05,
3200
+ "loss": 3.8744,
3201
+ "step": 269824
3202
+ },
3203
+ {
3204
+ "epoch": 1.01,
3205
+ "learning_rate": 4.5573955867640925e-05,
3206
+ "loss": 3.8806,
3207
+ "step": 270336
3208
+ },
3209
+ {
3210
+ "epoch": 1.01,
3211
+ "learning_rate": 4.5565569920130405e-05,
3212
+ "loss": 3.8784,
3213
+ "step": 270848
3214
+ },
3215
+ {
3216
+ "epoch": 1.01,
3217
+ "learning_rate": 4.5557200351423614e-05,
3218
+ "loss": 3.873,
3219
+ "step": 271360
3220
+ },
3221
+ {
3222
+ "epoch": 1.01,
3223
+ "learning_rate": 4.5548814403913094e-05,
3224
+ "loss": 3.8545,
3225
+ "step": 271872
3226
+ },
3227
+ {
3228
+ "epoch": 1.01,
3229
+ "learning_rate": 4.5540428456402574e-05,
3230
+ "loss": 3.8604,
3231
+ "step": 272384
3232
+ },
3233
+ {
3234
+ "epoch": 1.01,
3235
+ "learning_rate": 4.5532042508892054e-05,
3236
+ "loss": 3.8699,
3237
+ "step": 272896
3238
+ },
3239
+ {
3240
+ "epoch": 1.01,
3241
+ "learning_rate": 4.5523656561381534e-05,
3242
+ "loss": 3.8693,
3243
+ "step": 273408
3244
+ },
3245
+ {
3246
+ "epoch": 1.01,
3247
+ "learning_rate": 4.5515270613871014e-05,
3248
+ "loss": 3.8658,
3249
+ "step": 273920
3250
+ },
3251
+ {
3252
+ "epoch": 1.01,
3253
+ "learning_rate": 4.5506884666360494e-05,
3254
+ "loss": 3.8705,
3255
+ "step": 274432
3256
+ },
3257
+ {
3258
+ "epoch": 1.02,
3259
+ "learning_rate": 4.54985150976537e-05,
3260
+ "loss": 3.867,
3261
+ "step": 274944
3262
+ },
3263
+ {
3264
+ "epoch": 1.02,
3265
+ "learning_rate": 4.549012915014318e-05,
3266
+ "loss": 3.8598,
3267
+ "step": 275456
3268
+ },
3269
+ {
3270
+ "epoch": 1.02,
3271
+ "learning_rate": 4.548174320263266e-05,
3272
+ "loss": 3.8654,
3273
+ "step": 275968
3274
+ },
3275
+ {
3276
+ "epoch": 1.02,
3277
+ "learning_rate": 4.547335725512214e-05,
3278
+ "loss": 3.8724,
3279
+ "step": 276480
3280
+ },
3281
+ {
3282
+ "epoch": 1.02,
3283
+ "learning_rate": 4.546497130761162e-05,
3284
+ "loss": 3.8707,
3285
+ "step": 276992
3286
+ },
3287
+ {
3288
+ "epoch": 1.02,
3289
+ "learning_rate": 4.545660173890484e-05,
3290
+ "loss": 3.8805,
3291
+ "step": 277504
3292
+ },
3293
+ {
3294
+ "epoch": 1.02,
3295
+ "learning_rate": 4.544821579139432e-05,
3296
+ "loss": 3.8625,
3297
+ "step": 278016
3298
+ },
3299
+ {
3300
+ "epoch": 1.02,
3301
+ "learning_rate": 4.54398298438838e-05,
3302
+ "loss": 3.8603,
3303
+ "step": 278528
3304
+ },
3305
+ {
3306
+ "epoch": 1.02,
3307
+ "learning_rate": 4.543144389637328e-05,
3308
+ "loss": 3.8639,
3309
+ "step": 279040
3310
+ },
3311
+ {
3312
+ "epoch": 1.02,
3313
+ "learning_rate": 4.542305794886276e-05,
3314
+ "loss": 3.8637,
3315
+ "step": 279552
3316
+ },
3317
+ {
3318
+ "epoch": 1.02,
3319
+ "learning_rate": 4.541467200135224e-05,
3320
+ "loss": 3.8751,
3321
+ "step": 280064
3322
+ },
3323
+ {
3324
+ "epoch": 1.02,
3325
+ "learning_rate": 4.540630243264545e-05,
3326
+ "loss": 3.858,
3327
+ "step": 280576
3328
+ },
3329
+ {
3330
+ "epoch": 1.02,
3331
+ "learning_rate": 4.539791648513493e-05,
3332
+ "loss": 3.8618,
3333
+ "step": 281088
3334
+ },
3335
+ {
3336
+ "epoch": 1.02,
3337
+ "learning_rate": 4.538953053762441e-05,
3338
+ "loss": 3.8658,
3339
+ "step": 281600
3340
+ },
3341
+ {
3342
+ "epoch": 1.02,
3343
+ "learning_rate": 4.538114459011389e-05,
3344
+ "loss": 3.8699,
3345
+ "step": 282112
3346
+ },
3347
+ {
3348
+ "epoch": 1.02,
3349
+ "learning_rate": 4.537275864260337e-05,
3350
+ "loss": 3.8519,
3351
+ "step": 282624
3352
+ },
3353
+ {
3354
+ "epoch": 1.02,
3355
+ "learning_rate": 4.5364389073896576e-05,
3356
+ "loss": 3.8596,
3357
+ "step": 283136
3358
+ },
3359
+ {
3360
+ "epoch": 1.02,
3361
+ "learning_rate": 4.535600312638606e-05,
3362
+ "loss": 3.861,
3363
+ "step": 283648
3364
+ },
3365
+ {
3366
+ "epoch": 1.02,
3367
+ "learning_rate": 4.534761717887554e-05,
3368
+ "loss": 3.8543,
3369
+ "step": 284160
3370
+ },
3371
+ {
3372
+ "epoch": 1.02,
3373
+ "learning_rate": 4.533923123136502e-05,
3374
+ "loss": 3.8479,
3375
+ "step": 284672
3376
+ },
3377
+ {
3378
+ "epoch": 1.02,
3379
+ "learning_rate": 4.53308452838545e-05,
3380
+ "loss": 3.8571,
3381
+ "step": 285184
3382
+ },
3383
+ {
3384
+ "epoch": 1.02,
3385
+ "learning_rate": 4.5322459336343976e-05,
3386
+ "loss": 3.8591,
3387
+ "step": 285696
3388
+ },
3389
+ {
3390
+ "epoch": 1.02,
3391
+ "learning_rate": 4.531408976763719e-05,
3392
+ "loss": 3.8586,
3393
+ "step": 286208
3394
+ },
3395
+ {
3396
+ "epoch": 1.02,
3397
+ "learning_rate": 4.530570382012667e-05,
3398
+ "loss": 3.8529,
3399
+ "step": 286720
3400
+ },
3401
+ {
3402
+ "epoch": 1.02,
3403
+ "learning_rate": 4.529731787261615e-05,
3404
+ "loss": 3.8563,
3405
+ "step": 287232
3406
+ },
3407
+ {
3408
+ "epoch": 1.02,
3409
+ "learning_rate": 4.5288931925105625e-05,
3410
+ "loss": 3.8716,
3411
+ "step": 287744
3412
+ },
3413
+ {
3414
+ "epoch": 1.02,
3415
+ "learning_rate": 4.5280545977595105e-05,
3416
+ "loss": 3.8697,
3417
+ "step": 288256
3418
+ },
3419
+ {
3420
+ "epoch": 1.02,
3421
+ "learning_rate": 4.527217640888832e-05,
3422
+ "loss": 3.8632,
3423
+ "step": 288768
3424
+ },
3425
+ {
3426
+ "epoch": 1.02,
3427
+ "learning_rate": 4.52637904613778e-05,
3428
+ "loss": 3.8587,
3429
+ "step": 289280
3430
+ },
3431
+ {
3432
+ "epoch": 1.02,
3433
+ "learning_rate": 4.525540451386728e-05,
3434
+ "loss": 3.8559,
3435
+ "step": 289792
3436
+ },
3437
+ {
3438
+ "epoch": 1.02,
3439
+ "learning_rate": 4.524701856635676e-05,
3440
+ "loss": 3.8548,
3441
+ "step": 290304
3442
+ },
3443
+ {
3444
+ "epoch": 1.02,
3445
+ "learning_rate": 4.523863261884624e-05,
3446
+ "loss": 3.8483,
3447
+ "step": 290816
3448
+ },
3449
+ {
3450
+ "epoch": 1.02,
3451
+ "learning_rate": 4.523026305013945e-05,
3452
+ "loss": 3.8519,
3453
+ "step": 291328
3454
+ },
3455
+ {
3456
+ "epoch": 1.02,
3457
+ "learning_rate": 4.522187710262893e-05,
3458
+ "loss": 3.852,
3459
+ "step": 291840
3460
+ },
3461
+ {
3462
+ "epoch": 1.02,
3463
+ "learning_rate": 4.521349115511841e-05,
3464
+ "loss": 3.8612,
3465
+ "step": 292352
3466
+ },
3467
+ {
3468
+ "epoch": 1.02,
3469
+ "learning_rate": 4.520510520760789e-05,
3470
+ "loss": 3.8535,
3471
+ "step": 292864
3472
+ },
3473
+ {
3474
+ "epoch": 1.02,
3475
+ "learning_rate": 4.51967356389011e-05,
3476
+ "loss": 3.8604,
3477
+ "step": 293376
3478
+ },
3479
+ {
3480
+ "epoch": 1.02,
3481
+ "learning_rate": 4.518834969139058e-05,
3482
+ "loss": 3.8522,
3483
+ "step": 293888
3484
+ },
3485
+ {
3486
+ "epoch": 1.02,
3487
+ "learning_rate": 4.517996374388006e-05,
3488
+ "loss": 3.8507,
3489
+ "step": 294400
3490
+ },
3491
+ {
3492
+ "epoch": 1.02,
3493
+ "learning_rate": 4.517157779636954e-05,
3494
+ "loss": 3.8452,
3495
+ "step": 294912
3496
+ },
3497
+ {
3498
+ "epoch": 1.02,
3499
+ "learning_rate": 4.516319184885902e-05,
3500
+ "loss": 3.8666,
3501
+ "step": 295424
3502
+ },
3503
+ {
3504
+ "epoch": 1.02,
3505
+ "learning_rate": 4.5154822280152234e-05,
3506
+ "loss": 3.8648,
3507
+ "step": 295936
3508
+ },
3509
+ {
3510
+ "epoch": 1.02,
3511
+ "learning_rate": 4.5146436332641714e-05,
3512
+ "loss": 3.8551,
3513
+ "step": 296448
3514
+ },
3515
+ {
3516
+ "epoch": 1.02,
3517
+ "learning_rate": 4.5138050385131194e-05,
3518
+ "loss": 3.8512,
3519
+ "step": 296960
3520
+ },
3521
+ {
3522
+ "epoch": 1.02,
3523
+ "learning_rate": 4.5129664437620674e-05,
3524
+ "loss": 3.856,
3525
+ "step": 297472
3526
+ },
3527
+ {
3528
+ "epoch": 1.02,
3529
+ "learning_rate": 4.5121278490110154e-05,
3530
+ "loss": 3.8536,
3531
+ "step": 297984
3532
+ },
3533
+ {
3534
+ "epoch": 1.02,
3535
+ "learning_rate": 4.5112892542599634e-05,
3536
+ "loss": 3.8672,
3537
+ "step": 298496
3538
+ },
3539
+ {
3540
+ "epoch": 1.02,
3541
+ "learning_rate": 4.5104506595089114e-05,
3542
+ "loss": 3.8496,
3543
+ "step": 299008
3544
+ },
3545
+ {
3546
+ "epoch": 1.02,
3547
+ "learning_rate": 4.509613702638232e-05,
3548
+ "loss": 3.8537,
3549
+ "step": 299520
3550
+ },
3551
+ {
3552
+ "epoch": 1.02,
3553
+ "learning_rate": 4.50877510788718e-05,
3554
+ "loss": 3.8513,
3555
+ "step": 300032
3556
+ },
3557
+ {
3558
+ "epoch": 1.02,
3559
+ "learning_rate": 4.507936513136128e-05,
3560
+ "loss": 3.8557,
3561
+ "step": 300544
3562
+ },
3563
+ {
3564
+ "epoch": 1.02,
3565
+ "learning_rate": 4.507097918385076e-05,
3566
+ "loss": 3.8441,
3567
+ "step": 301056
3568
+ },
3569
+ {
3570
+ "epoch": 1.02,
3571
+ "learning_rate": 4.506259323634024e-05,
3572
+ "loss": 3.8465,
3573
+ "step": 301568
3574
+ },
3575
+ {
3576
+ "epoch": 1.02,
3577
+ "learning_rate": 4.505422366763345e-05,
3578
+ "loss": 3.8545,
3579
+ "step": 302080
3580
+ },
3581
+ {
3582
+ "epoch": 1.02,
3583
+ "learning_rate": 4.504583772012293e-05,
3584
+ "loss": 3.8464,
3585
+ "step": 302592
3586
+ },
3587
+ {
3588
+ "epoch": 1.02,
3589
+ "learning_rate": 4.503745177261242e-05,
3590
+ "loss": 3.8544,
3591
+ "step": 303104
3592
+ },
3593
+ {
3594
+ "epoch": 1.02,
3595
+ "learning_rate": 4.50290658251019e-05,
3596
+ "loss": 3.8558,
3597
+ "step": 303616
3598
+ },
3599
+ {
3600
+ "epoch": 1.02,
3601
+ "learning_rate": 4.502067987759138e-05,
3602
+ "loss": 3.8488,
3603
+ "step": 304128
3604
+ },
3605
+ {
3606
+ "epoch": 1.02,
3607
+ "learning_rate": 4.501231030888459e-05,
3608
+ "loss": 3.8515,
3609
+ "step": 304640
3610
+ },
3611
+ {
3612
+ "epoch": 1.02,
3613
+ "learning_rate": 4.500392436137407e-05,
3614
+ "loss": 3.8457,
3615
+ "step": 305152
3616
+ },
3617
+ {
3618
+ "epoch": 1.03,
3619
+ "eval_loss": 3.9113528728485107,
3620
+ "eval_runtime": 304.0304,
3621
+ "eval_samples_per_second": 1255.108,
3622
+ "eval_steps_per_second": 39.223,
3623
+ "step": 305280
3624
+ },
3625
+ {
3626
+ "epoch": 0.0,
3627
+ "learning_rate": 4.499553841386355e-05,
3628
+ "loss": 3.835,
3629
+ "step": 305664
3630
+ },
3631
+ {
3632
+ "epoch": 0.0,
3633
+ "learning_rate": 4.498715246635303e-05,
3634
+ "loss": 3.8334,
3635
+ "step": 306176
3636
+ },
3637
+ {
3638
+ "epoch": 0.0,
3639
+ "learning_rate": 4.497876651884251e-05,
3640
+ "loss": 3.855,
3641
+ "step": 306688
3642
+ },
3643
+ {
3644
+ "epoch": 0.0,
3645
+ "learning_rate": 4.497038057133199e-05,
3646
+ "loss": 3.8469,
3647
+ "step": 307200
3648
+ },
3649
+ {
3650
+ "epoch": 0.0,
3651
+ "learning_rate": 4.496199462382146e-05,
3652
+ "loss": 3.8579,
3653
+ "step": 307712
3654
+ },
3655
+ {
3656
+ "epoch": 0.0,
3657
+ "learning_rate": 4.495360867631094e-05,
3658
+ "loss": 3.8369,
3659
+ "step": 308224
3660
+ },
3661
+ {
3662
+ "epoch": 0.0,
3663
+ "learning_rate": 4.494522272880042e-05,
3664
+ "loss": 3.8458,
3665
+ "step": 308736
3666
+ },
3667
+ {
3668
+ "epoch": 0.0,
3669
+ "learning_rate": 4.49368367812899e-05,
3670
+ "loss": 3.8284,
3671
+ "step": 309248
3672
+ },
3673
+ {
3674
+ "epoch": 0.0,
3675
+ "learning_rate": 4.4928467212583116e-05,
3676
+ "loss": 3.8483,
3677
+ "step": 309760
3678
+ },
3679
+ {
3680
+ "epoch": 0.0,
3681
+ "learning_rate": 4.4920081265072596e-05,
3682
+ "loss": 3.8335,
3683
+ "step": 310272
3684
+ },
3685
+ {
3686
+ "epoch": 0.0,
3687
+ "learning_rate": 4.4911695317562076e-05,
3688
+ "loss": 3.8516,
3689
+ "step": 310784
3690
+ },
3691
+ {
3692
+ "epoch": 0.0,
3693
+ "learning_rate": 4.4903309370051556e-05,
3694
+ "loss": 3.849,
3695
+ "step": 311296
3696
+ },
3697
+ {
3698
+ "epoch": 0.0,
3699
+ "learning_rate": 4.4894939801344765e-05,
3700
+ "loss": 3.8236,
3701
+ "step": 311808
3702
+ },
3703
+ {
3704
+ "epoch": 0.0,
3705
+ "learning_rate": 4.4886553853834245e-05,
3706
+ "loss": 3.8379,
3707
+ "step": 312320
3708
+ },
3709
+ {
3710
+ "epoch": 0.0,
3711
+ "learning_rate": 4.4878167906323725e-05,
3712
+ "loss": 3.8383,
3713
+ "step": 312832
3714
+ },
3715
+ {
3716
+ "epoch": 0.0,
3717
+ "learning_rate": 4.4869781958813205e-05,
3718
+ "loss": 3.8259,
3719
+ "step": 313344
3720
+ },
3721
+ {
3722
+ "epoch": 0.0,
3723
+ "learning_rate": 4.4861412390106414e-05,
3724
+ "loss": 3.8361,
3725
+ "step": 313856
3726
+ },
3727
+ {
3728
+ "epoch": 0.0,
3729
+ "learning_rate": 4.4853026442595894e-05,
3730
+ "loss": 3.8371,
3731
+ "step": 314368
3732
+ },
3733
+ {
3734
+ "epoch": 0.0,
3735
+ "learning_rate": 4.4844640495085374e-05,
3736
+ "loss": 3.8334,
3737
+ "step": 314880
3738
+ },
3739
+ {
3740
+ "epoch": 0.0,
3741
+ "learning_rate": 4.4836254547574854e-05,
3742
+ "loss": 3.8543,
3743
+ "step": 315392
3744
+ },
3745
+ {
3746
+ "epoch": 0.0,
3747
+ "learning_rate": 4.482788497886807e-05,
3748
+ "loss": 3.8338,
3749
+ "step": 315904
3750
+ },
3751
+ {
3752
+ "epoch": 0.0,
3753
+ "learning_rate": 4.481949903135755e-05,
3754
+ "loss": 3.8351,
3755
+ "step": 316416
3756
+ },
3757
+ {
3758
+ "epoch": 0.0,
3759
+ "learning_rate": 4.481111308384703e-05,
3760
+ "loss": 3.8378,
3761
+ "step": 316928
3762
+ },
3763
+ {
3764
+ "epoch": 0.0,
3765
+ "learning_rate": 4.480272713633651e-05,
3766
+ "loss": 3.8412,
3767
+ "step": 317440
3768
+ },
3769
+ {
3770
+ "epoch": 0.0,
3771
+ "learning_rate": 4.479435756762972e-05,
3772
+ "loss": 3.8288,
3773
+ "step": 317952
3774
+ },
3775
+ {
3776
+ "epoch": 0.0,
3777
+ "learning_rate": 4.47859716201192e-05,
3778
+ "loss": 3.8402,
3779
+ "step": 318464
3780
+ },
3781
+ {
3782
+ "epoch": 0.0,
3783
+ "learning_rate": 4.477758567260868e-05,
3784
+ "loss": 3.8259,
3785
+ "step": 318976
3786
+ },
3787
+ {
3788
+ "epoch": 0.0,
3789
+ "learning_rate": 4.476919972509816e-05,
3790
+ "loss": 3.8298,
3791
+ "step": 319488
3792
+ },
3793
+ {
3794
+ "epoch": 0.0,
3795
+ "learning_rate": 4.476083015639137e-05,
3796
+ "loss": 3.8273,
3797
+ "step": 320000
3798
+ },
3799
+ {
3800
+ "epoch": 0.0,
3801
+ "learning_rate": 4.475244420888085e-05,
3802
+ "loss": 3.8292,
3803
+ "step": 320512
3804
+ },
3805
+ {
3806
+ "epoch": 0.01,
3807
+ "learning_rate": 4.474405826137033e-05,
3808
+ "loss": 3.8315,
3809
+ "step": 321024
3810
+ },
3811
+ {
3812
+ "epoch": 0.01,
3813
+ "learning_rate": 4.473567231385981e-05,
3814
+ "loss": 3.8365,
3815
+ "step": 321536
3816
+ },
3817
+ {
3818
+ "epoch": 0.01,
3819
+ "learning_rate": 4.4727302745153024e-05,
3820
+ "loss": 3.8369,
3821
+ "step": 322048
3822
+ },
3823
+ {
3824
+ "epoch": 0.01,
3825
+ "learning_rate": 4.4718916797642504e-05,
3826
+ "loss": 3.8362,
3827
+ "step": 322560
3828
+ },
3829
+ {
3830
+ "epoch": 0.01,
3831
+ "learning_rate": 4.4710530850131984e-05,
3832
+ "loss": 3.8281,
3833
+ "step": 323072
3834
+ },
3835
+ {
3836
+ "epoch": 0.01,
3837
+ "learning_rate": 4.4702144902621464e-05,
3838
+ "loss": 3.8381,
3839
+ "step": 323584
3840
+ },
3841
+ {
3842
+ "epoch": 0.01,
3843
+ "learning_rate": 4.469377533391467e-05,
3844
+ "loss": 3.8242,
3845
+ "step": 324096
3846
+ },
3847
+ {
3848
+ "epoch": 0.01,
3849
+ "learning_rate": 4.468538938640415e-05,
3850
+ "loss": 3.8267,
3851
+ "step": 324608
3852
+ },
3853
+ {
3854
+ "epoch": 0.01,
3855
+ "learning_rate": 4.467700343889363e-05,
3856
+ "loss": 3.8179,
3857
+ "step": 325120
3858
+ },
3859
+ {
3860
+ "epoch": 0.01,
3861
+ "learning_rate": 4.466861749138311e-05,
3862
+ "loss": 3.8251,
3863
+ "step": 325632
3864
+ },
3865
+ {
3866
+ "epoch": 0.01,
3867
+ "learning_rate": 4.466024792267632e-05,
3868
+ "loss": 3.8208,
3869
+ "step": 326144
3870
+ },
3871
+ {
3872
+ "epoch": 0.01,
3873
+ "learning_rate": 4.46518619751658e-05,
3874
+ "loss": 3.8341,
3875
+ "step": 326656
3876
+ },
3877
+ {
3878
+ "epoch": 0.01,
3879
+ "learning_rate": 4.464347602765528e-05,
3880
+ "loss": 3.8167,
3881
+ "step": 327168
3882
+ },
3883
+ {
3884
+ "epoch": 0.01,
3885
+ "learning_rate": 4.463509008014476e-05,
3886
+ "loss": 3.8231,
3887
+ "step": 327680
3888
+ },
3889
+ {
3890
+ "epoch": 0.01,
3891
+ "learning_rate": 4.462670413263425e-05,
3892
+ "loss": 3.8397,
3893
+ "step": 328192
3894
+ },
3895
+ {
3896
+ "epoch": 0.01,
3897
+ "learning_rate": 4.461833456392746e-05,
3898
+ "loss": 3.8293,
3899
+ "step": 328704
3900
+ },
3901
+ {
3902
+ "epoch": 0.01,
3903
+ "learning_rate": 4.460994861641694e-05,
3904
+ "loss": 3.8285,
3905
+ "step": 329216
3906
+ },
3907
+ {
3908
+ "epoch": 0.01,
3909
+ "learning_rate": 4.460156266890642e-05,
3910
+ "loss": 3.8197,
3911
+ "step": 329728
3912
+ },
3913
+ {
3914
+ "epoch": 0.01,
3915
+ "learning_rate": 4.45931767213959e-05,
3916
+ "loss": 3.8095,
3917
+ "step": 330240
3918
+ },
3919
+ {
3920
+ "epoch": 0.01,
3921
+ "learning_rate": 4.4584807152689106e-05,
3922
+ "loss": 3.8317,
3923
+ "step": 330752
3924
+ },
3925
+ {
3926
+ "epoch": 0.01,
3927
+ "learning_rate": 4.4576421205178586e-05,
3928
+ "loss": 3.8243,
3929
+ "step": 331264
3930
+ },
3931
+ {
3932
+ "epoch": 0.01,
3933
+ "learning_rate": 4.4568035257668066e-05,
3934
+ "loss": 3.819,
3935
+ "step": 331776
3936
+ },
3937
+ {
3938
+ "epoch": 0.01,
3939
+ "learning_rate": 4.4559649310157546e-05,
3940
+ "loss": 3.8164,
3941
+ "step": 332288
3942
+ },
3943
+ {
3944
+ "epoch": 0.01,
3945
+ "learning_rate": 4.4551279741450755e-05,
3946
+ "loss": 3.8259,
3947
+ "step": 332800
3948
+ },
3949
+ {
3950
+ "epoch": 0.01,
3951
+ "learning_rate": 4.4542893793940235e-05,
3952
+ "loss": 3.8029,
3953
+ "step": 333312
3954
+ },
3955
+ {
3956
+ "epoch": 0.01,
3957
+ "learning_rate": 4.4534507846429715e-05,
3958
+ "loss": 3.8282,
3959
+ "step": 333824
3960
+ },
3961
+ {
3962
+ "epoch": 0.01,
3963
+ "learning_rate": 4.45261218989192e-05,
3964
+ "loss": 3.8056,
3965
+ "step": 334336
3966
+ },
3967
+ {
3968
+ "epoch": 0.01,
3969
+ "learning_rate": 4.451775233021241e-05,
3970
+ "loss": 3.8206,
3971
+ "step": 334848
3972
+ },
3973
+ {
3974
+ "epoch": 0.01,
3975
+ "learning_rate": 4.450936638270189e-05,
3976
+ "loss": 3.8253,
3977
+ "step": 335360
3978
+ },
3979
+ {
3980
+ "epoch": 0.01,
3981
+ "learning_rate": 4.450098043519137e-05,
3982
+ "loss": 3.8134,
3983
+ "step": 335872
3984
+ },
3985
+ {
3986
+ "epoch": 0.01,
3987
+ "learning_rate": 4.449259448768085e-05,
3988
+ "loss": 3.8176,
3989
+ "step": 336384
3990
+ },
3991
+ {
3992
+ "epoch": 0.01,
3993
+ "learning_rate": 4.448420854017033e-05,
3994
+ "loss": 3.8168,
3995
+ "step": 336896
3996
+ },
3997
+ {
3998
+ "epoch": 0.01,
3999
+ "learning_rate": 4.447583897146354e-05,
4000
+ "loss": 3.8084,
4001
+ "step": 337408
4002
+ },
4003
+ {
4004
+ "epoch": 0.01,
4005
+ "learning_rate": 4.446745302395302e-05,
4006
+ "loss": 3.815,
4007
+ "step": 337920
4008
+ },
4009
+ {
4010
+ "epoch": 0.01,
4011
+ "learning_rate": 4.44590670764425e-05,
4012
+ "loss": 3.8175,
4013
+ "step": 338432
4014
+ },
4015
+ {
4016
+ "epoch": 0.01,
4017
+ "learning_rate": 4.445068112893198e-05,
4018
+ "loss": 3.812,
4019
+ "step": 338944
4020
+ },
4021
+ {
4022
+ "epoch": 0.01,
4023
+ "learning_rate": 4.444231156022519e-05,
4024
+ "loss": 3.8119,
4025
+ "step": 339456
4026
+ },
4027
+ {
4028
+ "epoch": 0.01,
4029
+ "learning_rate": 4.443392561271467e-05,
4030
+ "loss": 3.8028,
4031
+ "step": 339968
4032
+ },
4033
+ {
4034
+ "epoch": 0.01,
4035
+ "learning_rate": 4.4425539665204156e-05,
4036
+ "loss": 3.8024,
4037
+ "step": 340480
4038
+ },
4039
+ {
4040
+ "epoch": 0.01,
4041
+ "learning_rate": 4.4417153717693636e-05,
4042
+ "loss": 3.816,
4043
+ "step": 340992
4044
+ },
4045
+ {
4046
+ "epoch": 0.01,
4047
+ "learning_rate": 4.4408784148986845e-05,
4048
+ "loss": 3.8252,
4049
+ "step": 341504
4050
+ },
4051
+ {
4052
+ "epoch": 0.01,
4053
+ "learning_rate": 4.4400398201476325e-05,
4054
+ "loss": 3.808,
4055
+ "step": 342016
4056
+ },
4057
+ {
4058
+ "epoch": 0.01,
4059
+ "learning_rate": 4.4392012253965805e-05,
4060
+ "loss": 3.8275,
4061
+ "step": 342528
4062
+ },
4063
+ {
4064
+ "epoch": 0.01,
4065
+ "learning_rate": 4.4383626306455285e-05,
4066
+ "loss": 3.8235,
4067
+ "step": 343040
4068
+ },
4069
+ {
4070
+ "epoch": 0.01,
4071
+ "learning_rate": 4.4375240358944764e-05,
4072
+ "loss": 3.8112,
4073
+ "step": 343552
4074
+ },
4075
+ {
4076
+ "epoch": 0.01,
4077
+ "learning_rate": 4.4366870790237974e-05,
4078
+ "loss": 3.8146,
4079
+ "step": 344064
4080
+ },
4081
+ {
4082
+ "epoch": 0.01,
4083
+ "learning_rate": 4.4358484842727454e-05,
4084
+ "loss": 3.8205,
4085
+ "step": 344576
4086
+ },
4087
+ {
4088
+ "epoch": 0.01,
4089
+ "learning_rate": 4.4350098895216933e-05,
4090
+ "loss": 3.8117,
4091
+ "step": 345088
4092
+ },
4093
+ {
4094
+ "epoch": 0.01,
4095
+ "learning_rate": 4.4341712947706413e-05,
4096
+ "loss": 3.8096,
4097
+ "step": 345600
4098
+ },
4099
+ {
4100
+ "epoch": 0.01,
4101
+ "learning_rate": 4.433334337899962e-05,
4102
+ "loss": 3.8166,
4103
+ "step": 346112
4104
+ },
4105
+ {
4106
+ "epoch": 0.01,
4107
+ "learning_rate": 4.43249574314891e-05,
4108
+ "loss": 3.8161,
4109
+ "step": 346624
4110
+ },
4111
+ {
4112
+ "epoch": 0.01,
4113
+ "learning_rate": 4.431657148397859e-05,
4114
+ "loss": 3.8193,
4115
+ "step": 347136
4116
+ },
4117
+ {
4118
+ "epoch": 0.01,
4119
+ "learning_rate": 4.430818553646807e-05,
4120
+ "loss": 3.8107,
4121
+ "step": 347648
4122
+ },
4123
+ {
4124
+ "epoch": 0.01,
4125
+ "learning_rate": 4.429981596776128e-05,
4126
+ "loss": 3.7951,
4127
+ "step": 348160
4128
+ },
4129
+ {
4130
+ "epoch": 0.01,
4131
+ "learning_rate": 4.429143002025076e-05,
4132
+ "loss": 3.7979,
4133
+ "step": 348672
4134
+ },
4135
+ {
4136
+ "epoch": 0.01,
4137
+ "learning_rate": 4.428304407274024e-05,
4138
+ "loss": 3.8096,
4139
+ "step": 349184
4140
+ },
4141
+ {
4142
+ "epoch": 0.01,
4143
+ "learning_rate": 4.427465812522971e-05,
4144
+ "loss": 3.8106,
4145
+ "step": 349696
4146
+ },
4147
+ {
4148
+ "epoch": 0.01,
4149
+ "learning_rate": 4.426628855652293e-05,
4150
+ "loss": 3.8077,
4151
+ "step": 350208
4152
+ },
4153
+ {
4154
+ "epoch": 0.01,
4155
+ "learning_rate": 4.425790260901241e-05,
4156
+ "loss": 3.8102,
4157
+ "step": 350720
4158
+ },
4159
+ {
4160
+ "epoch": 0.02,
4161
+ "learning_rate": 4.424951666150189e-05,
4162
+ "loss": 3.8051,
4163
+ "step": 351232
4164
+ },
4165
+ {
4166
+ "epoch": 0.02,
4167
+ "learning_rate": 4.424113071399136e-05,
4168
+ "loss": 3.802,
4169
+ "step": 351744
4170
+ },
4171
+ {
4172
+ "epoch": 0.02,
4173
+ "learning_rate": 4.4232761145284576e-05,
4174
+ "loss": 3.8071,
4175
+ "step": 352256
4176
+ },
4177
+ {
4178
+ "epoch": 0.02,
4179
+ "learning_rate": 4.4224375197774056e-05,
4180
+ "loss": 3.8079,
4181
+ "step": 352768
4182
+ },
4183
+ {
4184
+ "epoch": 0.02,
4185
+ "learning_rate": 4.421598925026354e-05,
4186
+ "loss": 3.8154,
4187
+ "step": 353280
4188
+ },
4189
+ {
4190
+ "epoch": 0.02,
4191
+ "learning_rate": 4.4207603302753016e-05,
4192
+ "loss": 3.8166,
4193
+ "step": 353792
4194
+ },
4195
+ {
4196
+ "epoch": 0.02,
4197
+ "learning_rate": 4.419923373404623e-05,
4198
+ "loss": 3.8104,
4199
+ "step": 354304
4200
+ },
4201
+ {
4202
+ "epoch": 0.02,
4203
+ "learning_rate": 4.419084778653571e-05,
4204
+ "loss": 3.7991,
4205
+ "step": 354816
4206
+ },
4207
+ {
4208
+ "epoch": 0.02,
4209
+ "learning_rate": 4.4182461839025185e-05,
4210
+ "loss": 3.8107,
4211
+ "step": 355328
4212
+ },
4213
+ {
4214
+ "epoch": 0.02,
4215
+ "learning_rate": 4.4174075891514665e-05,
4216
+ "loss": 3.8024,
4217
+ "step": 355840
4218
+ },
4219
+ {
4220
+ "epoch": 0.02,
4221
+ "learning_rate": 4.416570632280788e-05,
4222
+ "loss": 3.8161,
4223
+ "step": 356352
4224
+ },
4225
+ {
4226
+ "epoch": 0.02,
4227
+ "learning_rate": 4.415732037529736e-05,
4228
+ "loss": 3.803,
4229
+ "step": 356864
4230
+ },
4231
+ {
4232
+ "epoch": 0.02,
4233
+ "learning_rate": 4.4148934427786834e-05,
4234
+ "loss": 3.7983,
4235
+ "step": 357376
4236
+ },
4237
+ {
4238
+ "epoch": 0.02,
4239
+ "learning_rate": 4.4140548480276314e-05,
4240
+ "loss": 3.8141,
4241
+ "step": 357888
4242
+ },
4243
+ {
4244
+ "epoch": 0.02,
4245
+ "learning_rate": 4.413217891156953e-05,
4246
+ "loss": 3.8097,
4247
+ "step": 358400
4248
+ },
4249
+ {
4250
+ "epoch": 0.02,
4251
+ "learning_rate": 4.412379296405901e-05,
4252
+ "loss": 3.8001,
4253
+ "step": 358912
4254
+ },
4255
+ {
4256
+ "epoch": 0.02,
4257
+ "learning_rate": 4.411540701654849e-05,
4258
+ "loss": 3.8043,
4259
+ "step": 359424
4260
+ },
4261
+ {
4262
+ "epoch": 0.02,
4263
+ "learning_rate": 4.410702106903797e-05,
4264
+ "loss": 3.8048,
4265
+ "step": 359936
4266
+ },
4267
+ {
4268
+ "epoch": 0.02,
4269
+ "learning_rate": 4.4098651500331186e-05,
4270
+ "loss": 3.7885,
4271
+ "step": 360448
4272
+ },
4273
+ {
4274
+ "epoch": 0.02,
4275
+ "learning_rate": 4.409026555282066e-05,
4276
+ "loss": 3.7919,
4277
+ "step": 360960
4278
+ },
4279
+ {
4280
+ "epoch": 0.02,
4281
+ "learning_rate": 4.408187960531014e-05,
4282
+ "loss": 3.8042,
4283
+ "step": 361472
4284
+ },
4285
+ {
4286
+ "epoch": 0.02,
4287
+ "learning_rate": 4.407349365779962e-05,
4288
+ "loss": 3.7999,
4289
+ "step": 361984
4290
+ },
4291
+ {
4292
+ "epoch": 0.02,
4293
+ "learning_rate": 4.4065124089092835e-05,
4294
+ "loss": 3.8018,
4295
+ "step": 362496
4296
+ },
4297
+ {
4298
+ "epoch": 0.02,
4299
+ "learning_rate": 4.405673814158231e-05,
4300
+ "loss": 3.7961,
4301
+ "step": 363008
4302
+ },
4303
+ {
4304
+ "epoch": 0.02,
4305
+ "learning_rate": 4.404835219407179e-05,
4306
+ "loss": 3.7967,
4307
+ "step": 363520
4308
+ },
4309
+ {
4310
+ "epoch": 0.02,
4311
+ "learning_rate": 4.403996624656127e-05,
4312
+ "loss": 3.8159,
4313
+ "step": 364032
4314
+ },
4315
+ {
4316
+ "epoch": 0.02,
4317
+ "learning_rate": 4.4031596677854484e-05,
4318
+ "loss": 3.8152,
4319
+ "step": 364544
4320
+ },
4321
+ {
4322
+ "epoch": 0.02,
4323
+ "learning_rate": 4.4023210730343963e-05,
4324
+ "loss": 3.8079,
4325
+ "step": 365056
4326
+ },
4327
+ {
4328
+ "epoch": 0.02,
4329
+ "learning_rate": 4.4014824782833443e-05,
4330
+ "loss": 3.8029,
4331
+ "step": 365568
4332
+ },
4333
+ {
4334
+ "epoch": 0.02,
4335
+ "learning_rate": 4.4006438835322923e-05,
4336
+ "loss": 3.796,
4337
+ "step": 366080
4338
+ },
4339
+ {
4340
+ "epoch": 0.02,
4341
+ "learning_rate": 4.399806926661613e-05,
4342
+ "loss": 3.799,
4343
+ "step": 366592
4344
+ },
4345
+ {
4346
+ "epoch": 0.02,
4347
+ "learning_rate": 4.398968331910561e-05,
4348
+ "loss": 3.7937,
4349
+ "step": 367104
4350
+ },
4351
+ {
4352
+ "epoch": 0.02,
4353
+ "learning_rate": 4.398129737159509e-05,
4354
+ "loss": 3.7995,
4355
+ "step": 367616
4356
+ },
4357
+ {
4358
+ "epoch": 0.02,
4359
+ "learning_rate": 4.397291142408457e-05,
4360
+ "loss": 3.7963,
4361
+ "step": 368128
4362
+ },
4363
+ {
4364
+ "epoch": 0.02,
4365
+ "learning_rate": 4.396454185537778e-05,
4366
+ "loss": 3.8055,
4367
+ "step": 368640
4368
+ },
4369
+ {
4370
+ "epoch": 0.02,
4371
+ "learning_rate": 4.395615590786726e-05,
4372
+ "loss": 3.7993,
4373
+ "step": 369152
4374
+ },
4375
+ {
4376
+ "epoch": 0.02,
4377
+ "learning_rate": 4.394776996035674e-05,
4378
+ "loss": 3.8023,
4379
+ "step": 369664
4380
+ },
4381
+ {
4382
+ "epoch": 0.02,
4383
+ "learning_rate": 4.393938401284622e-05,
4384
+ "loss": 3.7984,
4385
+ "step": 370176
4386
+ },
4387
+ {
4388
+ "epoch": 0.02,
4389
+ "learning_rate": 4.39309980653357e-05,
4390
+ "loss": 3.7931,
4391
+ "step": 370688
4392
+ },
4393
+ {
4394
+ "epoch": 0.02,
4395
+ "learning_rate": 4.392262849662892e-05,
4396
+ "loss": 3.7927,
4397
+ "step": 371200
4398
+ },
4399
+ {
4400
+ "epoch": 0.02,
4401
+ "learning_rate": 4.39142425491184e-05,
4402
+ "loss": 3.8059,
4403
+ "step": 371712
4404
+ },
4405
+ {
4406
+ "epoch": 0.02,
4407
+ "learning_rate": 4.390585660160788e-05,
4408
+ "loss": 3.8094,
4409
+ "step": 372224
4410
+ },
4411
+ {
4412
+ "epoch": 0.02,
4413
+ "learning_rate": 4.389747065409736e-05,
4414
+ "loss": 3.8024,
4415
+ "step": 372736
4416
+ },
4417
+ {
4418
+ "epoch": 0.02,
4419
+ "learning_rate": 4.3889101085390566e-05,
4420
+ "loss": 3.7935,
4421
+ "step": 373248
4422
+ },
4423
+ {
4424
+ "epoch": 0.02,
4425
+ "learning_rate": 4.3880715137880046e-05,
4426
+ "loss": 3.8035,
4427
+ "step": 373760
4428
+ },
4429
+ {
4430
+ "epoch": 0.02,
4431
+ "learning_rate": 4.3872329190369526e-05,
4432
+ "loss": 3.7955,
4433
+ "step": 374272
4434
+ },
4435
+ {
4436
+ "epoch": 0.02,
4437
+ "learning_rate": 4.3863943242859006e-05,
4438
+ "loss": 3.8105,
4439
+ "step": 374784
4440
+ },
4441
+ {
4442
+ "epoch": 0.02,
4443
+ "learning_rate": 4.3855573674152215e-05,
4444
+ "loss": 3.7968,
4445
+ "step": 375296
4446
+ },
4447
+ {
4448
+ "epoch": 0.02,
4449
+ "learning_rate": 4.3847187726641695e-05,
4450
+ "loss": 3.7977,
4451
+ "step": 375808
4452
+ },
4453
+ {
4454
+ "epoch": 0.02,
4455
+ "learning_rate": 4.3838801779131175e-05,
4456
+ "loss": 3.7959,
4457
+ "step": 376320
4458
+ },
4459
+ {
4460
+ "epoch": 0.02,
4461
+ "learning_rate": 4.3830415831620655e-05,
4462
+ "loss": 3.8041,
4463
+ "step": 376832
4464
+ },
4465
+ {
4466
+ "epoch": 0.02,
4467
+ "learning_rate": 4.382204626291387e-05,
4468
+ "loss": 3.7876,
4469
+ "step": 377344
4470
+ },
4471
+ {
4472
+ "epoch": 0.02,
4473
+ "learning_rate": 4.381366031540335e-05,
4474
+ "loss": 3.7929,
4475
+ "step": 377856
4476
+ },
4477
+ {
4478
+ "epoch": 0.02,
4479
+ "learning_rate": 4.380527436789283e-05,
4480
+ "loss": 3.7995,
4481
+ "step": 378368
4482
+ },
4483
+ {
4484
+ "epoch": 0.02,
4485
+ "learning_rate": 4.379688842038231e-05,
4486
+ "loss": 3.7924,
4487
+ "step": 378880
4488
+ },
4489
+ {
4490
+ "epoch": 0.02,
4491
+ "learning_rate": 4.378851885167552e-05,
4492
+ "loss": 3.8002,
4493
+ "step": 379392
4494
+ },
4495
+ {
4496
+ "epoch": 0.02,
4497
+ "learning_rate": 4.3780132904165e-05,
4498
+ "loss": 3.8013,
4499
+ "step": 379904
4500
+ },
4501
+ {
4502
+ "epoch": 0.02,
4503
+ "learning_rate": 4.377174695665448e-05,
4504
+ "loss": 3.7984,
4505
+ "step": 380416
4506
+ },
4507
+ {
4508
+ "epoch": 0.02,
4509
+ "learning_rate": 4.376336100914396e-05,
4510
+ "loss": 3.7933,
4511
+ "step": 380928
4512
+ },
4513
+ {
4514
+ "epoch": 0.02,
4515
+ "learning_rate": 4.375499144043717e-05,
4516
+ "loss": 3.7937,
4517
+ "step": 381440
4518
+ },
4519
+ {
4520
+ "epoch": 0.03,
4521
+ "eval_loss": 3.8858225345611572,
4522
+ "eval_runtime": 335.7172,
4523
+ "eval_samples_per_second": 1136.644,
4524
+ "eval_steps_per_second": 35.521,
4525
+ "step": 381600
4526
+ }
4527
+ ],
4528
+ "logging_steps": 512,
4529
+ "max_steps": 3052726,
4530
+ "num_train_epochs": 9223372036854775807,
4531
+ "save_steps": 10,
4532
+ "total_flos": 2.6558396388066816e+17,
4533
+ "trial_name": null,
4534
+ "trial_params": null
4535
+ }
checkpoint-381600/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c3acce838bf5be98e3878a84eef79455a1368123828b51c81d719ad33b6b820b
3
+ size 4219